 

a .. ‘
mania
A ~Jt

 

5&4
. EﬁmmtrJ
13%”

“v 5"
140311.:

in.

fl 1.
I452}..-
52.33. .

.\ .

I.
$3.4? :3.

w I ,3" 1}
5755-31 .?:I 61"
....; L 5. 39..

.7
..
.. ..

 

w... .1; , , .1 _
.mhz—i: hr 3 1 .11Mgwwww: i ”hum“.

_ .
. . . y sit

).‘,
ffab7dll

MICHIGANLIBRARIES
STATE UNIVERSITY
EAST LANSING, MICH 48824-1048

This is to certify that the
dissertation entitled

SOY REDUCES COLON CANCER RISK IN HUMANS AND

PhD.

 

RATS

presented by

DEEPA GOWRI THIAGARAJAN

has been accepted towards fulﬁllment
of the requirements for the

degree in Human Nutrition

 

C

Major Professor’s Signature

5 —-/O -- 3.005

'

 

Date

MSU is an Afﬁrmative Action/Equal Opportunity Institution

 

 

c-.---.-A--—l-

PLACE IN RETURN Box to remove this checkout from your record.

TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE

DATE DUE

DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2/05 alClﬁC/DateDueJndd-pjs

 

 

SOY REDUCES COLON CANCER RISK IN HUMANS AND RATS

By

DEEPA GOWRI THIAGARAJAN

A DISSERTATION

Submitted to
Michigan State University
in partial fulﬁllment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Food Science and Human Nutrition

2005

ABSTRACT
SOY REDUCES COLON CANCER RISK IN HUMANS AND RATS
By

DEEPA GOWRI THIAGARAJAN

Three studies were conducted to determine 1) the inﬂuence of age and history of
colonic polyps, cancer or FAP on colonic epithelial cell proliferation indices in hmnans,
2) the efﬁcacy of soy protein isolate (SPI) supplements to reduce intermediate biomarkers
of colon cancer (CC) risk in humans at risk for CC, and 3) the inﬂuence of soy products
on development of aberrant crypt foci (ACF) and on colonic epithelial cell proliferation
in ADM-initiated rats. In the ﬁrst study, subjects (N =46) were grouped based on age and
clinical history of polyps, F AP and CC. Colon biopsies were obtained by colonoscopy
and cell proliferation was measured by proliferating cell nuclear antigen (PCNA) and Ki-
67 immunohistochemistry (IHC) and relative PCNA protein levels by slot-blot
quantiﬁcation. PCNA labeling index (LI), PCNA proliferating zone (PZ), Ki-67 P2 and
PCNA protein levels increased with age and history of polyps, F AP or CC (P < 0.05).
The PCNA LI and P2, Ki-67 P2, and PCNA protein in colonic mucosa were highly
correlated with each other and were highly correlated with risk of CC which indicates
that these biomarkers have utility as intermediate biomarkers in chemoprevention studies.

The second experiment was a double-blind, prospective study designed to
determine if consumption of one of two supplements containing 38 g/d of SPI with 58 mg
of total genistein or 40 g/d of calcium caseinate for one year would inﬂuence colonic
epithelial cell proliferation as measured by PCNA and Ki-67 labeling and PCNA protein

expression in colonic epithelial cells. Colonic biopsies from were obtained from subjects

(N=42; age = 34-70; previous history of adenomatous polyps or CC) before and after
supplementation and used for IHC detection of PCNA and Ki-67 antigens in colon crypts
and for determination of relative PCNA protein expression by slot-blotting. There were
statistically signiﬁcant reductions in both PCNA LI and PCNA P2 in colonic mucosa of
subjects that consumed the soy supplements, but not in subjects who consumed casein at
the end of the study period. The relative expression of PCNA protein in colonic mucosa
was signiﬁcantly reduced by consumption of either of the soy supplements, but was not
inﬂuenced by consumption of the casein supplement.

The third study investigated the potential of different soy products with varying
levels of phytochemicals to reduce aberrant crypt foci (ACF) development and colonic
epithelial cell proliferation in carcinogen-initiated rats. The dietary treatments (n = 15
rats/treatment) were full fat soy ﬂakes (SF K), de-fatted soy ﬂour (SF L), soy concentrate
(SC), SC plus 150 mg/kg genistein (GEN), and SC plus 5 g/kg calcium (CAL). After 12
weeks of dietary treatment, rats consuming SC had the greatest (P < 0.05) numbers of
total ACF, large ACF and total aberrant crypts among all treatment groups. The GEN
and CAL treatments resulted in the lowest (P < 0.05) numbers of ACF. Rats fed SFK had
the lowest (P < 0.05) PCNA LI and P2. Rats fed SFL and SF K had signiﬁcantly lower
levels of PCNA protein when compared to rats consuming SC, GEN, or CAL diets.
Quantiﬁed PCNA protein expression positively correlated with LI and P2 (r = 0.81, r =
0.65, respectively at P<0.0001) analyses of PCNA by IHC. In conclusion, the reductions
we observed in colonic epithelial cell proliferation in humans and animals consuming soy
in these studies are characteristic of alterations that would be associated with an overall

decrease in risk for CC.

Copyright by
DEEPA GOWRI THIAGARAJAN
2005

Dedicated to

Amma and Appa, Dr. Shanthi and Professor Thiagarajan

ACKNOWLEDGMENTS

The following people and organizations are gratefully acknowledged for their
contribution to the successful completion of this dissertation and to an enriched
experience during my research: Dr. James Mayle and Dr. Radhika Srinivasan from the
Department of Medicine at MSU, and Ms. Gail Rouhier from Ingham Regional Medical
Center. Neither the studies nor the dissertation itself would have been possible without
the study subjects, so I express my gratitude to those persons who participated in the
research. The United Soybean Board generously provided ﬁnancial support for this
research for which I am grateful. I am also thankful to MSU for direct ﬁnancial aid
through fellowships, awards, and travel grants to make my graduate studies possible.

My greatest debt is to Prof. Maurice R. Bennink, who has been a dedicated
advisor, a judicious mentor, and the kindest friend. I would like to express my gratitude
to him for his support, patience, and encouragement throughout my graduate studies. It is
not often that one ﬁnds an advisor and colleague that always ﬁnds the time to listen to the
little problems and roadblocks that unavoidably crop up in the course of performing
research. His technical and editorial advice was essential to the completion of this
dissertation and he has taught me innumerable lessons and insights on the workings of
academic research in general. His conﬁdence allowed me the extraordinary privilege of
exploring my mind's potential. Thank you, Dr. Bennink, trusted mentor, for the promise
delivered.

My appreciation also goes to Dr. Dale Romsos for his comments and criticisms
during the completion of this study. I am especially indebted to Dr. Romsos for his

valuable advice, help and suggestions during my research and completion of my

vi

dissertation. It became distinctly easier to ﬁnish with his helpﬁilness and conﬁdence in
my ability to carry this project through. He is most deﬁnitely an outstanding scholar, a
superlative guide, and the consummate critic and collaborator.

My thanks also go to committee members Dr. Lorraine Weatherspoon and Dr.
James Mayle, for reading previous drafts of this dissertation and providing many valuable
comments that improved the presentation and content of this dissertation.

I was fortunate to have the opportunity to work with a group of energetic people
in Dr. Maurice Bennink’s and Dr. Les Bourquin’s labs. I am appreciative of my fellow
graduate students, Elizabeth Rondini, Elizabeth Seymour and Bing Wang, for making my
time at MSU an enjoyable experience. I relish every moment that we have worked
together including all those late night lab activities! The ﬁ'iendships of Dr. Ross Santell
and Mr. Richard Grabiel led to many interesting and good-spirited discussions relating to
this research and are greatly treasured. I am thankful for all their collective
encouragement to ﬁnish this dissertation.

During my graduate program, many faculty and staff in the Department of Food
Science and Human Nutrition assisted and encouraged me in various ways. I am
especially grateful to DrS. Mark Uebersax, Won Song, Joseph Shroeder, Bill Helferich
and Gale Strasburg for all that they have taught me. I also thank the students whom I was
privileged to teach and from whom I also learned much.

I would like to make a special acknowledgement to my colleagues at MBI
International and at the Institute of International Agriculture, several of whom are my
closest friends. They have always been there for me, listening to me rejoice, complain,

and ponder my way through the Ph.D. study.

vii

I owe so much to my family, especially my parents, Shanthi and Thiagarajan. This
dissertation is dedicated to Amma and Appa, who taught me the importance of a solid
educational background as the foundation and means for achieving success in life.
Without their love, support, and understanding, the accomplishment of receiving a
Doctorate of Philosophy degree would not have been possible. Throughout my life they
have provided me with words of encouragement so that I would continue to seek self-
improvement, life-long pursuit of learning, and lead a productive life. They are excellent
role models and my goal is to emulate their behavior and to follow in their footsteps. For
always being there when I needed them most, and never once complaining about how
infrequently I visit, they deserve far more credit than I can ever give them. Thank you for
believing in me.

Finally, I would like to express my deepest gratitude to Les Bourquin for
everything from technical support to emotional support. I could not have done it without

you. Thank you for being there for me from the very beginning.

viii

TABLE OF CONTENTS

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

LIST OF TABLES xi
LIST OF FIGURES xiii
LIST OF ABBREVIATIONS xiv
CHAPTER I. INTRODUCTION 1
CHAPTER 11. REVIEW OF LITERATURE 6
A. Epidemiology: Incidence patterns and distribution of colon cancer ................ 6
B. Genetic basis for colorectal carcinogenesis 7
1. The multistage hypothesis 7
2. Familial adenomatous polyposis (FAP) 9
3. Hereditary non-polyposis colorectal cancer (HNPCC) 10
C. Morphology of colorectal neoplasia precursor and associated lesions of
colorectal cancer 10
1. Hyperproliferation and Aberrant Crypt Foci 11
2. Colo-rectal Micro-Adenomas 12
3. Hyperplastic Polyps 13
4. Adenomas l4
5. Severe Dysplasia, Carcinoma In Situ 15
6. “De Novo” Carcinoma l6
7. Invasive Colorectal Cancer 16
D. Role of environmental factors in colorectal cancer 16
1. Role of diet in etiology and prevention of colon carcinogenesis ................ 18
1.1. Effect of dietary mutagens and carcinogenesis 18
1.2. Reduction of contact between colon mucosa and carcinogens .......... 19
1.3. Role of colonic microﬂora 20
1.4. Prevention of cellular damage and antioxidants 21
1.5. Fat consumption and caloric intake 23
2. Diet and Chemoprevention of colon cancer 24
2.1. Chemoprevention 24
2.2. Soy consumption and colon cancer 25
2.3. Role of calcium in colorectal chemoprevention 50
E. Biomarkers and models for chemopreventive studies 55
1. Surrogate end point biomarkers (SEB) 55
2. Cell proliferation biomarkers 56
3. Differentiation markers 58
4. Apototic markers 58
5. Aberrant crypt foci 59
6. Animal models 60
F. Conclusion 62

 

CHAPTER III. CELL PROLIFERATION INDICES IN COLONIC CRYPTS OF

 

 

 

 

 

HUMANS AT VARYING AGES AND RISKS FOR COLON CANCER ................ 65
A. ABSTRACT 65
B. INTRODUCTION 66
C. MATERIALS AND METHODS 69
D. RESULTS 74
E. DISCUSSION 79
F. CONCLUSION _ 88

 

CHAPTER IV. BIOMARKERS FOR COLON CANCER RISK BEFORE AND
AFTER CONSUMING SOY OR CASEIN SUPPLEMENTS FOR ONE YEAR 94

 

 

 

 

A. ABSTRACT 94
B. INTRODUCTION 95
C. MATERIALS AND METHODS 99
D. RESULTS 107
E. DISCUSSION 110

 

CHAPTER V. EFFECT OF SOY FLAKES, SOY FLOUR, SOY CONCENTRATE,
GENISTEIN AND CALCIUM ON ABERRANT CRYPT FOCI DEVELOPMENT
AND COLONIC EPITHELIAL CELL PROLIFERATION IN RATS INJECT ED

 

 

 

 

 

 

 

WITH AOM 125
A. ABSTRACT 125
B. INTRODUCTION 126
C. MATERIALS AND METHODS 131
D. RESULTS 139
E. DISCUSSION 143
CHAPTER VI. SUMMARY AND CONCLUSIONS 159
CHAPTER VII. RECOMMENDATIONS FOR FUTURE STUDIES ................... 162

BIBLIOGRAPHY 168

 

Table 1.

Table 2.

Table 3.

Table 4.

Table 5.

Table 6.

Table 7.

Table 8.

Table 9.

Table 10.

Table 11.

Table 12.

Table 13.

Table 14.

Table 15.

Table 16.

LIST OF TABLES

Percentage composition of soy protein products (moisture-free basis) ........... 31
Estimated isoﬂavone contents of selected soy protein products ..................... 32

Crypt heights and Proliferation indices — PCNA and Ki-67 LI & PZ in
colon of subjects at varying age and risk for colon cancer ...................... 89

PCNA compartmental labeling indices (for basal, middle and top crypt com—
partments) in colon of subjects at varying age and risk for colon cancer ...... 90

Ki-67 compartmental labeling indices (for basal, middle and top crypt com-
partments) in colon of subjects at varying age and risk for colon cancer ...... 91

Relative PCNA levels in colonic mucosa of subjects at varying age and risk
for colon cancer92

Pearson correlation coefﬁcients for Ki-67 & PCNA LI (total crypt LI) and
PZ, PCNA middle LI and relative PCNA levels in colonic mucosa of

subjects at varying age and risk for colon cancer .................................. 93
Nutrient Composition of Dietary Supplements ................................... 119
Characteristics of Study Participants at Baseline ................................. 120
Daily Intakes of Selected Dietary Nutrients by Study Participants at the
beginning (TB) and ending (TB) of each treatment ............................... 121
Whole Colonic Crypt Proliferation Indices (PCNA and Ki—67) of subjects. 122
PCNA compartmental proliferation indices in colonic crypts of subjects... 123
Effect of Casein, SPIl plus Calcium and SP1 minus Calcium on the mean
beginning and ending relative PCNA (Slot-blot methodology) protein

levels in colonic mucosa ............................................................. 124
Composition of diets ................................................................. 152
Isoﬂavone concentrations in rat diets .............................................. 153

Effect of diet on preneoplastic lesions in the colon of rats treated with
AOM ................................................................................... 154

xi

Table 17. Effect of diet on colonic crypt height and epithelial PCNA proliferation
indices in rats treated with AOM ................................................... 155

Table 18. Effect of diet on colonic crypt height and epithelial Ki-67 proliferation
indices in rats treated with AOM ................................................... 156

Table 19. Effect of diet on quantiﬁed colonic PCNA levels in rats treated with
AOM ................................................................................... 157

Table 20. Correlation coefﬁcients for total ACF, quantiﬁed PCNA levels,

PCNA LI and PZ and Ki-67 LI and P2 in colonic mucosa of rats
injected with AOM .................................................................. 158

xii

LIST OF FIGURES

Figure 1. Stages of soy bean processing .......................................................................... 30

xiii

LIST OF ABBREVIATIONS

ACF .............................................................................. Aberrant Crypt Foci
ADME ..................................... Absorption, Distribution, Metabolism and Excretion
AEC ........................................................................... Aminoethyl Carbazole
AICR/WCRF ....... American Institute for Cancer Research/World Cancer Research Fund
AIN 93G ......................................... American Institute of Nutrition 93 Growth diet
AOM ................................................................................... Azoxymethane
AP .............................................................................. Adenomatous Polyps
APC ................................................................... Adenomatous Polyposis Coli
AUC ............................................................................ Area Under the Curve
BBI ........................................................................... Bowman-Birk Inhibitor
BMI .................................................................................. Body Mass Index
BSA ........................................................................ Bovine Serum Albumim
CAL ............................................................................................ Calcium
CC ....................................................................................... Colon Cancer
CH ........................................................................................ Crypt Height
COX .................................................................................. Cyclooxygenase
DCC ............................................................. i ........... Deleted in Colon Cancer
DM .................................................................................. Diabetes Mellitus
DMH .............................................................................. Dimethylhydrazine
EDTA ......................................................... Ethylene Diamine Tetra Acetic acid
EGF ........................................................................ Epidermal Growth Factor
FAP ............................................................... Familial Adenomatous Polyposis

xiv

GEN ........................................................................................... Genistein

hMLHl ................................................................ Human Mut—L Homologue 1
HNPCC ............................................ Hereditary Non-Polyposis Colorectal Cancer
HRT ................................................................ Hormone Replacement Therapy
IHC ............................................................................ Immunohistochemical
LEF ....................................................................... Lymphoid Enhance Factor
LI ....................................................................................... Labeling Index
MAPK .......................................................... Mitogen Activated Protein Kinase
MOPS ....................................................... 4-Morpholine Propane Sulfonic Acid
NSAID ................................................... Non Steroidal Anti—Inﬂammatory Drugs
PBBI ............................................................... Puriﬁed Bowman-Birk Inhibitor
PBS ...................................................................... Phosphate-Buffered Saline
PCNA .......................................................... Proliferating Cell Nuclear Antigen
PKC ................................................................................. Protein Kinase C
PMSF ............................................................... Phenylmethylsulphonylﬂuoride
PVDF ......................................................................... Polyvinylideneﬂuoride
PZ .................................................................................. Proliferation Zone
RR ....................................................................................... Relative Risks
SAS ...................................................................... Statistical Analysis System
SC .................................................................................... Soy Concentrate
SDS ................................................................... ~ .. . .Sodium Dodecyl Sulphate
SEB ............................................................... Surrogate End- point Biomarkers
SEM ..................................................................... Standard Error of the Mean

XV

SF K ......................................................................................... Soy Flakes

SFL ............................................................................................ Soy Flour
SPI ................................................................................. Soy Protein Isolate
TBS .............................................................................. Tris-Buffered Saline
TCF ....................................................................................... T Cell Factor
TGF .................................................................... Transforming Growth Factor

xvi

CHAPTER I. INTRODUCTION

In the United States, colon cancer (CC) is the third most common form of cancer
in terms of incidence and second most common cause of cancer mortality (American
Cancer Society, 2005). It is estimated that 66-75% of CC incidence could be prevented
by adequate diets and with appropriate changes made to current dietary habits and
lifestyle practices (AICR/WCRF, 1997). Increasing the consumption of foods that contain
anti-carcinogenic phytochemicals is emphasized as a means to reduce incidence of many
diet-related cancers such as CC. Even though there is no strong correlation between any
particular food and CC risk, many cross-cultural epidemiological studies indicate that
consumption of soy foods may contribute to the relatively low rates of colon, breast and
prostate cancers observed in east Asian countries such as China and Japan (Messina et al.,
1994; Barnes, 1998; Messina, 1999; Birt et al., 2001).

The development of CC is a process involving multiple steps, and often is a
consequence of an intimate, but only moderately understood, interplay between genetic
and environmental factors (Fearon and Vogelstein, 1990). As the colorectal epithelium
progresses from a normal phenotype to one that is hyperproliferative, multiple genetic
and epigenetic alterations occur, including activation of proto-oncogenes, inactivation of
tumor suppressor genes, and mutations in mismatch repair genes leading to malignancy
(F earon and Jones, 1992).

Dietary factors can potentially affect each step of tumorigenesis, thereby
modulating the above-mentioned genetic and epigenetic changes. The process of using
natural or synthetic chemical compounds to reverse, suppress, and/or prevent progression

to invasive cancer is referred as “chemo-prevention” (Lippman et al., 1994). Cancer

chemo—prevention studies using animal models provide for strict control of variables, but
inherent Species differences can be a problem, precluding direct extrapolations of
observations from animal studies to humans. Previous studies in humans have tested the
ability of candidate dietary components for their utility as chemo-preventive agents in
colonic carcinogenesis with little success (Einspahr et al., 1997; Barnes et al., 1998;
Krishnan et al., 1998). Cancer intervention trials using human subjects are exceedingly
difﬁcult to conduct because of the slow progressive nature of carcinogenesis and the
large numbers of subjects required to achieve adequate statistical power (Kim and
Mason, 1996; Pfeiffer et al., 2003).

In the last decade, several strategies have been employed to circumvent the
problems associated with conducting cancer intervention trials in humans. One strategy is
to study those individuals who are at high risk (age, history of polyps, FAP) for
developing CC to determine whether a certain chemo-preventive agent can suppress or
prevent cancer development. Another strategy is to use early or intermediate biomarkers
as surrogate end points for cancer risk rather than relying on occurrence or recurrence of
CC (Sharma et al., 2001; Garcea et al., 2003). Alterations in mucosal proliferation
indices, aberrant histopathologic features, abnormal immunologic and biochemical
markers, and more recently, alterations of molecular markers have been extensively
studied for their applicability and utility as intermediate markers in colorectal cancer
chemoprevention trials (Deschner, 1988; Lipkin, 1988; Singh et al., 1993; Bostick et al.,
1995; Greenwald et al., 1995; Toribara and Sleisenger, 1995; Mark, 1996; Thigalingam et

al., 1996; Takayama et al., 1998; Gryfe et al., 1999).

Recognizing carcinogenesis as a multi-step process, one of the ﬁrst indications of
preneoplastic changes in colorectal mucosa is increased epithelial cell proliferation.
Actively dividing cells produce a number of unique proteins. Proliferating cell nuclear
antigen (PCNA) is one protein whose level of synthesis correlates directly with the rate of
cellular proliferation and DNA synthesis. Previous research indicates that PCNA is a
reliable biological marker for proliferative activity in colonic crypts (Malecka—Panas et
al., 1997; Shpitz etal., 1997; Barnes et al., 1998). PCNA expression correlates well with
that of other proliferation markers like bromo-deoxyuridine (BrdU) and Ki-67 antigen
expression in both normal and neoplastic colonic epithelium (Kubota and Kino, 1995;
Yerly-Motta et al., 1999; Wu et al., 2000; Diez et al., 2003). Unfortunately, the
immunohistochemical (IHC) techniques used to determine cell proliferation using PCNA
or other endogenous markers are relatively long and laborious. Hence, there is a need for
a rapid, reliable and valid method of measuring PCNA protein levels in colonic mucosa
that could be utilized in human studies to test for potential chemo-preventive agents. A
simpler and more rapid method for PCNA protein quantiﬁcation may also have utility in
the clinical assessment of colon biopsy specimens.

To date, there has been no published research which compares PCNA and Ki-67
proliferation indices in persons of widely varying ages who have no history of colonic
diseases. A study was designed to assess and directly compare PCNA and Ki-67
proliferation indices (using IHC) in colonic mucosa biopsies obtained from persons who
differed in age (without any history of colonic diseases) and in persons with known risk
factors (history of adenomas, familial adenomatous polyposis and history of CC) for

colon cancer development. The second objective of this study was to use an

immunoblotting method to determine relative levels of PCNA protein in colonic
epithelial cells and determine if PCNA protein expression correlated with proliferation
indices based on IHC techniques.

Epidemiological studies have reported that more frequent consumption of
vegetable protein, including soy protein, is associated with reduced risk for CC. There are
no published studies on the effects of soy protein consumption on colonic epithelial cell
proliferation in humans. A second study was then conducted to determine the potential
chemo-preventive activity of soy protein isolate in humans at risk for CC. This clinical
trial was designed to determine if consuming 38 g of soy protein isolate supplement per
day for one year would reduce CC risk as reﬂected by changes in colonic epithelial cell
proliferation patterns and quantiﬁed PCNA levels (by immunoblotting method) in colonic
mucosa. A secondary objective was to determine if adding or removing calcium from the
soy protein isolate supplement would alter the impact of soy on colonic epithelial cell
proliferation in humans.

Soybeans and soybean-based foods are a good source of several phytochemicals,
including isoﬂavones, phytosterols, saponins, protease inhibitors and phytate that have
been reported to have potential anticancer properties (Messina and Bennink, 1998).
Therefore, the third study was conducted to determine if soy-containing diets having
differing concentrations of phytochemicals (especially isoﬂavones) can reduce
preneoplastic lesions and indices of colonic epithelial cell proliferation (as assessed by
PCNA and Ki-67 IHC in colonic tissue and by PCNA protein quantiﬁcation (by
immunoblotting) in rats initiated with a colon carcinogen (azoxymethane). A further

objective was to determine if changes in colonic epithelial cell proliferation patterns (as

inﬂuenced by dietary treatments) correlated with the numbers of preneoplastic lesions
detected in the colons of rats treated with azoxymethane and subjected to dietary
treatments differing in soy preparations and phytochemical contents. The ﬁnal objective
of this study was to determine if dietary calcium concentration (1 or 5 g/kg of diet) would
inﬂuence preneoplastic lesion development and markers of colonic epithelial cell
proliferation in these rats.

A reduction in colonic epithelial cell proliferation, a reduced Size of the zone of
proliferation in colonic crypts, and a decrease in quantiﬁed PCNA levels in the colon
mucosa (of both animals and humans) after consumption of soy compounds would
suggest that one or more phytochemicals present in soy is responsible for altering the cell
proliferation pattern. Thus, these studies will demonstrate if chemo-preventive agents in
soy inhibit colonic epithelial cell proliferation as reﬂected by changes in PCNA and Ki-
67 expression and total quantity of PCNA protein in colonic mucosa, alterations that

would be associated with an overall decrease in risk for CC.

CHAPTER II. REVIEW OF LITERATURE

A. Epidemiology: Incidence patterns and distribution of colon cancer

Cancer of the colon and rectum is the third most common incident cancer and
cause of death from cancer, throughout the world (AICR/WCRF, 1997). In 2000 alone,
an estimated 945,000 were diagnosed worldwide, accounting for 9.4% of all new cases of
cancer (Parkin, 2004). In the US, colorectal cancer is the third most common incident
cancer and second most common cause of mortality (American Cancer Society, 2005).
The incidence of colorectal cancer in the US. has been declining since 1998 (American
Cancer Society, 2005). However, worldwide incidence of and deaths from colon cancer
(CC) are generally increasing, mostly in the developed countries and in urban areas of the
developing countries (Parkin, 2004). High- risk areas include North America, Europe and
Australia, New Zealand (Parkin, 2004) accounting for over 63% of the total global
incidence. Incidence is also now increasing in what was considered as low risk areas
such as Central and South America, Asia and Africa (AICR/WCRF, 1997).

Worldwide incidence of colon cancer is similar in both men and women (Parkin,
2004). In the United States, colorectal cancer incidence and mortality rates are more than
35% higher in men than in women (American Cancer Society, 2005). In 2002, the age-
adjusted incidence rates for CC were 62.1 and 46.0 cases per 100,000 population in men
and women, respectively. The age-adjusted rates for rectal cancer were 24.8 and 17.4
cases per 100,000 population in men and women, respectively (Ries et al., 2005).
According to migrant and temporal trend studies colorectal cancers are largely
determined by environmental exposures. Colorectal cancer rates have increased in the

migrants that moved from low risk to high- risk countries reﬂecting the host countries

rates within one or two generations (Parkin, 2004). There is at least a 25-fold variation in
colorectal cancer worldwide (Parkin, 2004). Regional difference in incidence rates may
be explained, in large part, by dietary and other environmental differences. Colorectal
cancer has also long been known to occur more frequently in certain families (Macklin,
1960) and there are several rare genetic syndromes that carry a markedly elevated risk
(Gardner, 1951; Veale, 1965; Utsunomiya and Lynch, 1990). Colorectal cancer is thus

casually related to both genes and environment.

B. Genetic basis for colorectal carcinogenesis
1. The multistage hypothesis

Colon cancer development is a multi-stage process in which transformed cells
progress to precancerous lesions (e.g. adenomatous polyps) and, ultimately, carcinomas
during a period spanning several years to decades. The gastrointestinal epithelium is a
complex microenvironment, made up of at least ﬁve interrelated cell types and
characterized by a tightly regulated program of cell proliferation, maturation,
differentiation and apoptosis. Carcinogenesis in colonic mucosa is likely the result of
successive accumulation of multiple genetic mutations, resulting in transformed
phenotype and eventual progression to invasive cancer. Vogelstein et al. (1988) described
speciﬁc mutations in colorectal cancer and deﬁned their relationship to the adenoma-
carcinoma sequence. This led to wide acceptance of the multistep hypothesis as basis for
malignant transformation and gave a genetic perspective to the processes of colon tumor

initiation, promotion and progression.

Fearon and Vogelstein (1990) in their genetic model, indicate a hyperproliferative
colonic epithelium as the ﬁrst step toward CC development. People at high risk for colon
cancer (old age, presence of adenomatous polyps or familial adenomatous polyposi
[FAP]) have increased mucosal cell proliferation throughout their colon. Gene mutations
or deletions accompany the promotion from a hyperproliferative normal crypt to a small
adenoma. In this initiation stage, the adenomatous polyposis coli gene (APC) acts as a
gatekeeper of colonic epithelial proliferation. One key function of the normal APC
protein in colonic epithelial cells is to form a multi-protein complex (with axin/conductin
and glycogen synthase kinase-3 B) that binds to cytosolic B-catenin and targets it for
destruction via the ubiquitin-proteasome pathway (Easwaran et al., 1999). APC gene
mutations typically result in the production of truncated APC protein that is unable to
bind to B-catenin, thereby allowing the cytosolic accumulation of B-catenin. When [3-
catenin levels are elevated, it can translocate to the nucleus, bind to T cell factor
(TCF)/lymphoid enhancer factor (LEF) transcription factors, and stimulate transcription
of a wide variety of gene products via activation of the Wnt signaling pathway (Ilyas and
Tomlinson, 1997). Thus, APC gene mutation is permissive for persistently elevated
concentrations of B-catenin, which ultimately results in abnormal cell proliferation,
differentiation and migration in the colonic mucosa. Mutations of the APC gene are a
frequent and early event of sporadic colon tumorigenesis (Luchtenborg et al., 2004).

Another common genetic event in colon carcinogenesis is mutation of the K-ras
gene (Fearon and Vogelstein, 1990). Transformation to the cancerous state
(adenocarcinoma) is frequently associated with the loss of the DCC (Deleted in Colon

Cancer) gene and/or loss of the p53 gene. The p53 gene encodes for a protein having a

Signiﬁcant homology to cell adhesion family of proteins. This protein can also block
DNA-damaged cells from progressing to the S phase of the cell cycle thereby acting as a
check point to allow for adequate DNA repair before permitting another round of cell
division. Mutated p53 protein binds to and inactivates wild type p53 protein. Mutation
and or deletion of p53 gene plus a K-ras mutation is sufﬁcient to transform a normal cell
to a cancerous one. Approximately 60 - 85% of colon tumors have mutated p53 genes
(Baker et al., 1989; Yamaguchi et al., 1994a; Yamaguchi et al., 1994b).

The DCC gene encodes a plasma membrane protein that is believed to function in
cell-cell communication and/or cell-substrate interactions (Cho et al., 1994). As mucosal
cells travel toward the ﬂat mucosa of the crypt, the DCC encoded protein may pass
signals from the submucosa directing cells to undergo terminal differentiation. There is a
strong relationship between lack of DCC expression, lack of cell differentiation, and
tumorigenesis (Hedrick et al., 1994). Approximately 75% of colorectal tumors and 40%
of adenomas with foci of carcinomatous transformation have a mutated or missing DCC

gene (F earon et al., 1990).

2. Familial adenomatous polyposis (F AP)

Familial adenomatous polyposis is an autosomal dominantly inherited disease,
which was ﬁrst described by Bussey in 1975. This condition is responsible for less than
1% of all colorectal cancers and has a prevalence of about 1 in 10,000 (Bishop and
Thomas, 1990; Bisgaard et al., 1994). The gene responsible for FAP was shown to be
present on chromosome 5 by Leppert et al., in 1987. This gene was later identiﬁed and

cloned and now referred to as the APC or Adenomatous Polyposis Coli gene (Groden et

al., 1991; Nishisho et al., 1991). A germ-line mutation in the APC gene causes FAP,
which is characterized by a permanent imbalance in cell proliferation leading to

hyperproliferative epithelium preceding the formation of adenomas. Individuals with
F AP inherit a single mutated allele of the APC gene that acts in a dominant negative
manner leading to development of dozens to hundreds of colonic adenomas at a very

young age.

3. Hereditary non-polyposis colorectal cancer (HNPCC)

HNPCC is a heritable dominant condition with a high degree of gene penetrance.
A family of four DNA mismatch repair genes have been identiﬁed to be responsible for
the development of HNPCC (Papadopoulos et al., 1994). They are regarded as
oncosuppressor genes, a second copy of one of the four genes being inactivated either by
mutation or loss of the gene. The subsequent breakdown of the DNA repair mechanisms
leads to an accelerated pathway of neoplastic evolution in which gene activation is
preferentially mediated by somatic mutation as opposed to an allelic loss (Leach et al.,

1993).

C. Morphology of colorectal neoplasia precursor and associated lesions of
colorectal cancer
Colon carcinogenesis is a multi-step process in which initiated cells progress
through various stages (6. g. ACF, microadenomas, adenomatous polyps), ultimately

leading to carcinoma in Situ and invasive colon cancer. The various precursor and

associated lesions which differ morphologically from the normal colorectal mucosa are

discussed in the following paragraphs.

1. Hyperproliferation and Aberrant Crypt Foci
Increased colorectal crypt cell proliferation, a well as an expansion of the

proliferative zone toward the luminal surface of the crypts have been regarded as an early
indicators of future risk for the development of colon tumors (Friedman, 1985; Lipkin,
1988; Fearon and Vogelstein, 1990; Srivastava et al., 2001). Hyper proliferation has been
observed in adenomas as compared with benign epithelium, with the highest proliferation
seen in adenomas with high grade dysplasia. In addition, a shift of proliferation toward
the luminal surface of the epithelium occurs in adenomas as well as aberrant crypt foci
(ACF). ACF were ﬁrst described by Bird, in chemically induced rodent models of colon
carcinogenesis when the whole mounts of colon tissue was stained with methylene blue
and examined under low power magniﬁcation (Bird, 1987; McLellan and Bird, 1988).
These lesions have been also identiﬁed in macroscopically normal human colonic
mucosa (Pretlow et al., 1991; Roncucci et al., 1991). ACF are described as usually having
larger than normal crypts, with large cells microscopically elevated above the mucosal
surface, and often Show increased branching and proliferation. They also have been
reported to contain cells displaying features ranging from almost normal to dysplastia,
with a variety of changes in mucin production and goblet cells, suggesting heterogeneity
at an early stage of carcinogenesis (McLellan et al., 1991a; McLellan et al., 1991b;
Cademi et al., 1995; Pretlow, 1995). Hyperproliferative lesions described in the earlier

studies could have been instances of ACF, or at least included ACF. Mutations of K-ras,

APC, and some p53 genes occur in human ACF and in chemically induced ACF in
experimental animals (Smith et al., 1994; Yamashita et al., 1995).

The numbers of ACF can be inﬂuenced by diet and other factors in the colonic
environment. Studies have shown stimulation of ACF by exposure to secondary bile
acids and suppression by primary bile acids. In chemically-induced rodent models of CC
dietary ﬁber, beta-carotene and other retinoids, soy phytochemicals (genistein, inositol
hexaphosphate, protease inhibitors, saponins), calcium, aspirin and other NSAIDS, all
have been demonstrated to inhibit ACF formation (Rao et al., 1992; Thorup et al., 1994;
Pretlow, 1995; Wargovich eta1., 1995a; Wargovich et al., 1995b). Although there is no
direct evidence for ACF transitioning to micro-adenomas, adenomas or carcinomas, they
are being regarded as early morphological changes in colorectal neoplasia. Jen et al.
(1994) reported that only those ACF having morphology similar to micro-adenomas and
APC gene mutations were likely to progress to carcinomas. ACF also appears to develop
in response to certain environmental agents such as dietary factors, alcohol and smoking,
which in some individuals produce non-neoplastic hyperproliferative lesions such as
hyperplastic polyps and may, depending on nature of mutations that occur, result in
development of colorectal tumors. A better understanding of the morphologic and
molecular heterogeneity of ACF is needed to help elucidate the early phases of colorectal

neoplasia.

2. Colo-rectal Micro-Adenomas

These were ﬁrst recognized in 1991 (Roncucci et al.) during a microscopical

examination of ACF. As mentioned earlier, there is no known relationship between ACF

12

and micro-adenomas but it is possible that a minority of ACF are microadenomas at the
time of onset, and that some of these may grow to become endoscopically recognizable
adenomas. The nature of the genetic change in ACF may be an important determinant of
their ultimate progression to other lesion types. Non-neoplastic hyperplastic foci may
develop into hyperplastic polyps, whilst those foci which also have APC mutations are
likely to develop into micro-adenomas, and some of which then progress into macro-
adenomas (Jen et al., 1994). Micro-adenomas have now become an important link in the
morphology of colorectal neoplastic events as well as in the concept of “de-novo”

colorectal cancer.

3. Hyperplastic Polyps

Hyperplastic polyps endoscopically look similar to adenomatous polyps, however
have a distinct histology. Hyperplastic polyps range in size from 1 — 40 mm and are more
frequent in the distal than the proximal colon (Ferrandez et al., 2004). Until recently,
hyperplastic polyps were not considered likely to progress to colorectal cancer. There
was indirect evidence of a relationship of hyperplastic polyps to adenomas and colorectal
cancers due to similarities in their distribution (Isbister, 1993). Furthermore, hyperplastic
polyps in the distal large bowel had been found to be strong indicators for the presence of
colorectal adenomas in the proximal large bowel (N usko et al., 1994; Van Stalk et al.,
1994). Recent studies indicate that hyperplastic polyps are morphologically similar to
serrated adenomas and serrated adenocarcinomas and are likely to be biologically related.
Chan et al. (2003) found that acquisition of a BRAF mutation was associated with the

progression of hyperplastic polyps to serrated adenomas, whereas acquisition of K-ras

l3

mutations was associated with the progression of hyperplastic polyps to an admixed
hyperplastic polyp/adenoma morphology. Hawkins and Ward (2001) demonstrated that
expression of human mut-L homologue l (hMLHl) protein, a DNA mismatch repair
enzyme, was lost in benign hyperplastic polyps and serrated adenomas obtained from 10
of 13 subjects previously diagnosed with microsatellite-unstable CC. This loss of
hMLHl protein expression was associated with methylation of the promoter region of the
hMLHl gene, and concluded that methylation of the hMLHl promoter within these
neoplastic cell populations may be a critical step in the progression to carcinoma
(Hawkins and Ward, 2001). Hyman et al. (2004) conducted a prospective study to assess
the risk of patients having hyperplastic polyps to develop CC, and concluded that these

subjects are, in fact, at high risk for developing CC.

4. Adenomas
These are benign tumors of the intestinal epithelium and in Western populations

are considered as the major precursor of colorectal cancer. The monoclonal origin of
adenomas, that is, their commencement from one stem cell was ﬁrst established by
Fearon et al. (1987). Current evidence suggests that adenomas commence in a subset of
proliferative lesions such as ACF, becoming ﬁrst a micro-adenoma and then a visible
adenoma. Adenomas are classiﬁed according to their appearance as protuberant or
polypoid, or as ﬂat adenomas. Microscopically, adenomas are classiﬁed as tubular,
villous or tubulovillous. Polypoid adenomas are the most common type of adenoma in
Western populations and their evolution appears to be associated with both p53 and K—ras

mutations (Fujimori et al., 1994; Yukawa et al., 1994).

14

Macroscopically ﬂat adenomas were ﬁrst recognized by Muto and co-workers in
Japan (Muto et al., 1985; Minamoto et al., 1994). In Western populations ﬂat adenomas
are much less frequent than polypoid adenomas, although their frequency is uncertain as
they can be difﬁcult to identify during colonoscopy (Matsumoto et al., 1994). They are
probably more common in the distal than in proximal colon, suggesting environmental
exposures to be important in their etiology. F lat adenomas also tend to occur at an earlier
age than polypoid adenomas and are more aggressive, both histologically and clinically,
regarding malignant potential. Furthermore, ras mutations are not expressed in ﬂat
adenomas (Fujimori et al., 1994; Yukawa et al., 1994) but APC gene mutations have been
detected (van Wyk et al., 1999). An epidemiologic overview of these biologic differences
suggest that quantitative rather than qualitative differences of inherited and
environmental exposures are responsible for the different frequency of ﬂat compared to

polypoid adenomas in Japan versus Western populations.

5. Severe Dysplasia, Carcinoma In Situ
Histologic changes intermediate between normal colorectal mucosa and a
neoplastic lesion are termed as “dysplasia”, and these changes have been graded into mild
or low-grade and severe or high-grade dysplasia (O'Brien et al., 1990). The terms “severe
dysplasia” or “high-grade dysplasia” are often preferred to terms “carcinoma in Situ” or
“intramucosal carcinoma” because these lesions probably lack the ability to invade and

metastasize, as it requires further genetic and immunologic changes.

15

6. “De Novo” Carcinoma

These are small non-polypoid cancerous lesions that account for one-third of the
incident cases in Western populations (Kuramoto and Oohara, 1995). There has been
some evidence to believe these cancers originate from normal colorectal mucosal cells
(“de-novo”), and other evidence suggests that these cancers evolve from ﬂat adenomas.
The ﬂat adenoma-adenocarcinoma sequence has also been documented in chemically
induced colon tumors in rats, and about one-third of all neoplasms were of the ﬂat
variety, a proportion similar to “de-novo” cancers in Western populations (Rubio and

Shetye, 1994).

7. Invasive Colorectal Cancer
When cancer cells penetrate the muscle layer, they are regarded as invasive
cancers which are able to Spread, metastasize and cause premature death. The ability to
become invasive and then metastasize appears to involve further somatic mutations and

possibly other host-defense mechanisms which at present are poorly understood.

D. Role of environmental factors in colorectal cancer
Diet has long been regarded as the most important environmental inﬂuence on
colorectal cancer. Multiple confounding interactions among dietary constituents have
made it difﬁcult to determine the relationship between diet and colorectal cancer. Despite
this, there is a growing body of evidence from both human epidemiological studies and
animal studies that have linked consumption of several dietary components to colorectal

cancer risk. Ever since Burkitt (1969) noted an inverse association between ﬁber intake

16

and CC incidence, several prospective and case-control studies have conﬁrmed the
relationship between ﬁber consumption and colorectal neoplasia (Giovannucci et al.,
1992). Increased risk for CC has been associated with high dietary fat intakes,
particularly that of saturated fats present in red meat (Newmark et al., 1991; Potter, 1992;
Giovannucci et al., 1994). Numerous epidemiologic studies have reported that
populations having high vegetable and fruits intakes have a decreased incidence of CC
(Potter, 1993). A number of additional dietary agents, such as vitamins, micronutrients,
and other minor dietary components have been suggested to lower risk for CC. Because
the consumption of ﬁber, fat, red meat, fruits, vegetables, antioxidants, or micronutrients
are all closely linked, it is difﬁcult to point to any one of these as a causative or protective
agent for CC.

Equally difﬁcult is to distinguish between energy intake, and intake of fat (the
most energy dense constituent of the diet) when studying their potential relationships
with CC risk. Cohort studies (Willett et al., 1990; Bostick et al., 1994; Giovannucci et al.,
1994; Goldbohm et al., 1994) have found either no relationship or an inverse relationship
between total energy intake and colon cancer risk. On the other hand, case-control studies
(Potter and McMichael, 1986; Whittemore et al., 1990; Peters et al., 1992) have found
that cases reported greater caloric intake than controls suggesting an association between
high energy intake and increased risk for CC. A case-control study by Slattery et al.
(1997) found an increase in CC risk associated with high energy intake, but only at lower
levels of lifetime vigorous physical activity and higher body mass index (BMI). These
studies suggest it is possible that total energy intake has no simple relationship with colon

cancer risk, but its effect maybe dependent on levels of obesity and physical activity.

17

Type 2 diabetes mellitus (DM) has emerged as another risk factor for CC.
Limburg et al. (2005) examined the relationship between type 2 DM and CC risk in
postmenopausal women participating in the Iowa Women’s Health Study. They found
that women having type 2 DM had a signiﬁcantly increased risk of developing CC when
compared to women with no history of DM (relative risk = 1.4). Cancer subsite-speciﬁc
analysis found that type 2 DM increased the risk for proximal CC (relative risk = 1.9),
but was not associated with increased risk for cancer in the distal colon (relative risk =
1.1) or the rectum (relative risk =' 0.8). Coughlin et al. (2004) examined the relationship
between DM and CC risk in a large cohort of men and women who had no history of
cancer at enrollment in 1982. After 16 years of mortality follow-up, DM was
signiﬁcantly associated with CC in men (relative risk = 1.20) and women (relative risk =
1.24).

Other established non-dietary factors that inﬂuence CC risk include advanced age,
low levels of physical activity, genetic pre-disposition, ulcerative colitis and smoking.
Regular use of aspirin and/or NSAIDS has been associated with a decreased risk for

colon cancer in humans (Waddell et al., 1989; Labayle et al., 1991).

1. Role of diet in etiology and prevention of colon carcinogenesis
1.1. Effect of dietary mutagens and carcinogenesis
Preventing dietary carcinogens and mutagens from interacting with the colon
mucosa and inhibiting initiation of CC has been one of the many approaches employed in
reducing diet induced CC. Carcinogens and mutagens include agents that occur in diet

naturally or agents which are inadvertently added or produced during cooking or food

18

processing. In vitro mutagen assays such as the Salmonella lyphimurium (Ames) test can
identify potential genotoxic agents in the diet (Maron and Ames, 1983). When dose,
threshold effects and in vivo detoxiﬁcation mechanisms were considered, neither the
naturally occurring nor inadvertently added dietary genotoxins accounted for the higher
incidence of CC observed in Western countries compared to Poland, southern Europe,
Japan or South America (Henderson, 1990). Potent mutagens such as polycyclic aromatic
hydrocarbons, can be formed during charcoal broiling of meat (Lijinsky and Ross, 1967;
Barnett, 1976). However, research by Weisburger (1973) and Toth (1980) have Shown
that polycyclic aromatic hydrocarbons do not signiﬁcantly increase the risk for CC.
Subsequent research focused on the formation of heterocyclic aromatic amines during
frying, broiling and toasting of foods. These are considered extremely potent mutagens
(Sugimura, 1992; Weisburger et al., 1994) and their presence in food may be one cause

for the increased CC incidence in Western countries.

1.2. Reduction of contact between colon mucosa and carcinogens
An alternative approach in prevention of mutagenesis is to reduce contact

between a mutagen/carcinogen and the colon mucosa. This can be achieved by diluting
intestinal contents and/or by decreasing residence time of undigested food in the colon.
Some types of dietary ﬁber can fulﬁll both roles. Non-fermented dietary ﬁber
constituents like lignin or poorly-fermented dietary ﬁber constituents like cellulose and
hemicellulose adsorb and absorb water. These ﬁbers plus their associated water dilute
intestinal contents and any carcinogen or mutagen that may be present. Also, dietary

cellulose and hemicellulose decrease intestinal transit time through the colon. Pectins and

19

gurus are often completely fermented by colonic microﬂora and, therefore, contribute
little dilution of carcinogens and mutagens in the colon and do not decrease residence
time of fecal material in the colon. Epidemiological studies suggest that dietary ﬁber is
important in reducing CC incidence (Greenwald et al., 1987; Trock et al., 1990).
However, results from animal studies that tested the efficacy of dietary ﬁber in reducing
chemically-induced CC have been mixed — ﬁnding decreased, no effect, or even an
enhancement of colon carcinogenesis (Glauert et al., 1981). Even though a preponderance
of evidence suggests that dietary ﬁber decreases CC risk (Macquart-Moulin et al., 1986;
Kune et al., 1987; Graham et al., 1988; Zaridze et al., 1993), the extent to which it can
reduce CC incidence is uncertain. In contrast, recent comprehensive prospective studies
examining the role of ﬁber and its components on risk of colorectal neoplasms have
found no protective effect from a high-ﬁber diet against colorectal cancer or adenomas
(Fuchs et al., 1999; Alberts et al., 2000; Bonithon-Kopp et al., 2000; Michaels et al.,

2000; Schatzkin et al., 2000).

1.3. Role of colonic microﬂora
Differences in CC incidence between countries with high and low incidence rates
have been hypothesized to result from differences in colonic bacterial populations or
from differences in bacterial metabolites (Hill et al., 1971). Studies enumerating
facultative and strict anaerobes in colon contents failed to ﬁnd major differences in
numbers or types of bacteria in the colons of persons at high versus low risk for CC
(Moore and Holdeman, 1975; Goldberg et al., 1977). Although types or number of

colonic bacteria were not associated with CC incidence, populations that have a high

20

incidence of CC have greater concentrations of mutagens in their stools (Reddy et al.,
1978; Van Tasslee et al., 1990). Fecapentaenes are bacterial metabolites of unabsorbed
fatty acids and they are mutagenic in bacterial mutation assays. However, fecapentaenes
do not act as initiators of CC (Weisburger et al., 1990) as was originally hypothesized. A
later study showed that fecapentaenes were more likely to function as promoters of CC
(Zarkovic et al., 1993). Colonic concentrations of fecapentaenes are increased with high
fat, low ﬁber diets. Therefore, it is possible that fecapentaenes contribute to the increased

incidence of CC observed in populations consuming high fat, low ﬁber diets.

I. 4. Prevention of cellular damage and antioxidants

Stimulating the activities of hepatic and intestinal “phase II” enzymes that
detoxify xenobiotics can reduce chemically induced CC (Deschner et al., 1991; Yoshimi
et al., 1992). Phase II enzymes conjugate xenobiotics with glutathione, glucuronic acid,
sulfate, or acetate to xenobiotics to reduce their toxicity and to speed their clearance via
bile or urine. Epidemiologic studies often ﬁnd an inverse relationship between CC risk
and consumption of fruits, vegetables and grains (Potter, 1996; Potter, 1999). This
inverse relationship generally is attributed to increased consumption of antioxidants,
ﬁber, and substances that enhance phase II enzymes (Steinmetz and Potter, 1991;
Steinmetz and Potter, 1996). However, the results of the multi-center, prospective Polyp
Prevention Trial found that consumption of a low fat, high ﬁber diet did not reduce polyp
recurrence during a four-year intervention period (Schatzkin et al., 2000). More recent
analysis of carotenoid and vitamin A consumption among a sub-cohort of the Polyp

Prevention Trial indicated that higher serum concentrations and dietary intakes of a-

21

carotene and B-carotene, as well as higher dietary vitamin A consumption, all were
associated with reduced risk for polyp recurrence (Steck-Scott et al., 2004). In another
study comparing plasma antioxidant concentrations and polyp incidence in subjects
undergoing complete colonoscopy, Erhardt et al. (2003) found that plasma lycopene
concentrations were signiﬁcantly lower (35%) in subjects having adenomas as compared
to subjects without adenomas.

Oxidative damage to critical cellular macromolecules can both initiate and
promote CC (Cerutti and Trump, 1981; Wattenberg, 1993; Lippman et al., 1994).
Carotenoids, vitamins C and E, phenolics and selenium, as part of the glutathione
peroxidase enzyme complex, act as general antioxidants at the cellular level. However, it
is difﬁcult to Show that dietary intakes above the minimum requirements offer a
protective effect (Greenberg et al., 1994; MacLennan et al., 1995; Clark et al., 1996).

Inﬂammatory conditions stimulate oxidative damage to cells, production of
growth factors and cell proliferation leading to promotion of CC (Cerutti and Trump,
1981; Baldassarre et al., 2004).The prostanoids - prostaglandin E1, prostaglandin E2 and
leukotriene B4 - are potent proinﬂammatory agents. Inhibition of prostanoid synthesis
can reduce the oxidative damage to cells and tissue and thus inhibit CC. Non steroidal
anti-inﬂammatory drugs (N SAIDS) that inhibit prostanoid synthesis also lead to
regression of colon polyps in humans (Earnest et al., 1992; Giardiello et al., 1993; Reddy
et al., 1993; Pennisi, 1998; Baron et al., 2002) and inhibit chemically induced CC in

rodent models (Labayle et al., 1991; Reddy et al., 1996; Kawamori et al., 1998).

22

I . 5. Fat consumption and caloric intake

Dietary fats, particularly the saturated fat present in red meat, may contribute to
colorectal cancer (Willett et al., 1990). Multiple case-control studies also have
demonstrated an association between high fat consumption and increased CC risk (Potter,
1992; Giovannucci et al., 1994). Animal studies suggest that high fat intake, particularly
saturated fat, induces abnormal proliferation of colonic mucosa and may even lead to
development of aberrant crypt foci, one of the earliest cancer-associated structural
changes in the gut (Newmark et al., 1991). Bile acids whose production is increased by
high fat intake have mitogenic effects upon intestinal epithelial cells (Hill et al., 1971;
Chomchai et al., 1974). Secondary bile acids (deoxycholic and lithocholic acid) are
present in higher concentrations in the colon of populations at increased risk for CC.
Because this population also had an increased fat and decreased ﬁber diet, the exact
relationship of these components to CC risk remains uncertain.

There has been considerable interest in the relationship between calorie intake and
colon carcinogenesis. When investigating their relationship with any chronic disease it is
difﬁcult to distinguish between the energy intake and intake of dietary fat. Thus the
question has been raised whether the effect of dietary fat on colon carcinogenesis is due
to speciﬁc action of fat or to an associated caloric effect (Carrol, 1986; Reddy, 1986).
Several studies have reported that calorie restriction reduces the formation of
spontaneous and chemically induced tumors in mice (Ross and Bras, 1971). Further
studies have Shown that calorie restriction decreased oxidative damage to DNA (Klurfeld
et al., 1987). The effect of a high-fat, semi-puriﬁed diet injested ad libitum and of a 30%

calorie restricted diet on AOM-induced colon carcinogenesis during the post initiation

23

phase was investigated by Reddy et a1. (1987). The animals on the calorie- restricted diet
developed signiﬁcantly fewer colon tumors and had a lower colon tumor incidence than
the rats in the ad libitum-fed group. In another study (Kumar et al., 1990) the effect of
10%, 20%, and 30% calorie restriction was investigated and it was found that the
incidence and multiplicity of colon tumors were signiﬁcantly inhibited in animals fed
20% and 30% calorie-restricted diets. These results indicate calorie restriction elicits a

dose response effect on inhibition of tumor incidence.

2. Diet and Chemoprevention of colon cancer
2.1. Chemoprevention

The epidemiology of colorectal cancer suggests that the most direct method of
preventing colorectal cancers is by identiﬁcation and removal of precursor adenomas.
However, the multi-step nature of colonic neoplasia makes early detection difﬁcult
(Loren et al., 2003). In advanced stages of the cancer, surgical resection of the colon and
adjuvant chemotherapy or radiotherapy is practiced. Because reduced risk of colorectal
neoplasia is associated with increased dietary intakes of various chemicals with cancer-
preventing properties in animal studies, an intense effort is underway to develop effective
and safe chemopreventive agents (Hawk et al., 2004). An ideal chemopreventive agent
would be Simple to administer with extremely low toxicity, as continuous exposure may
be necessary for prevention of colon cancer (Courtney et al., 2004).

A number of promising chemopreventive agents such as dietary calcium, soy
phytochemicals, citrus bioﬂavanoids, cherry anthocyanins, COX 2 inhibitors (N SAIDS),

and others have been identiﬁed which modulate cellular processes such as cellular

24

proliferation, differentiation and apoptosis, and thereby may have efﬁcacy in reducing
CC risk (Hawk et al., 2004). The following sections describe the potential of soy

phytochemicals and calcium to reduce CC risk

2.2. Soy consumption and colon cancer
i. Epidemiological studies
International variations in colon cancer have been attributed, at least in part, to
differences in dietary intake. Asian populations have relatively low rates of breast,
prostate, and colon cancer when compared with Western cultures. Studies of migrant
populations have helped conﬁrm the hypothesis that environmental variables (especially
diet) play a Signiﬁcant role in determining cancer risk. For example, rates of breast cancer
are lower in Japan than in the United States and increase among Japanese migrating to
Hawaii (Parkin, 1989). Asian diets are typically lower in total and saturated fat and
higher in dietary ﬁber than Western diets. Another Signiﬁcant dietary difference between
Asian and Western population is the consumption of soy-containing foods. Soybeans
have long been a major component of the Asian diet, common soy products being
soymilk, tofu, miso, yuba and tempeh. Mean intakes of 20-50 grams of soy foods per day,
has been reported in studies, but the values for soy intake on dry weight basis is much
lower (Wu et al., 1998).
In a review of epidemiologic literature, Messina et al. (1994), examined the
relationship between consumption of soy-containing foods and cancer risk. Among
twenty-ﬁve studies of fermented soy products reviewed, four noted an increased risk,

eighteen no statistically signiﬁcant difference and three found a decreased risk of cancer

25

development of various sites (breast, prostate, colon, lung). Among twenty-six studies of
non-fermented soy products, results were more consistent in showing either a decreased
risk (10 studies) or no statistically signiﬁcant difference (15 studies), with only one study
showing an increased risk of esophageal cancer associated with consumption of fried
bean curd.

In a cross-cultural study of thirty-eight countries, McKeown-Eyseen and Bright-
See (1984) found no association between soybean intake and CC risk. The countries
studied were chosen on the basis of reliable cancer mortality data but were not identiﬁed.
Therefore, it is difﬁcult to evaluate ﬁndings from this study Since only a handful of
countries consume amounts of soy likely to bear any physiological relevance. Haenszel et
al. (1973) reported a Signiﬁcantly increased risk of CC among Japanese born Issai
residing in Hawaii who ingested fermented soybeans (relative risk = 1.6). Tajima and
Tominaga (1985) in a Japanese study, noted an increase in rectal, but not in CC risk, in
persons ingesting miso soup (relative risk = 2.05, P<0.05). These authors found no
relationship between bean curd consumption and either colon or rectal cancer.

Three other case-control studies have been suggestive of a protective effect of soy
consumption although differences were statistically signiﬁcant in only one of these
studies. Poole et al. (1989) in the only case-control study conducted in the United States,
noted a relative risk of 0.53 (conﬁdence interval 0.11-2.66) for CC for persons eating
tofu/soybeans more than once a month, as opposed to less than once a month. Hu et al.
(1991) found a protective effect of soybean products on rectal, but not on colon, cancer
risk in a case-control study conducted in China. Watanabe et al. (1984) found that

consumption of beans and bean curd once or twice a week and three or more times a

26

week, in comparison to not consuming theses foods, was associated with lowered rectal
cancer risk (relative risks = 0.15 and 0.12 respectively). They found a lower, but not
signiﬁcantly decreased risk of CC (relative risk = 0.63).

More detailed discussion of these studies can be found in a (Messina and
Bennink, 1998) review of 8 ecological and case-control studies that examine the
relationship between soy food intake and colorectal cancer. None of these studies
reported a statistically signiﬁcant reduction in CC risk with soy consumption. They
observe that while these studies report a non-statistical reduction in colon cancer risk,
there were a few that associate certain soy product consumption with a higher risk for
CC. However, in most of these studies, soy intake and cancer risk was not the primary
focus of investigation, and there is cause for concern with regards to the completeness
and accuracy of the dietary soy data in these studies. Spector et al. (2003) in her review
of 13 epidemiological studies (3 ecological, 1 cohort and 9 case-control) observed that
these studies were inconclusive with only a suggestion of possible inverse association
between soy products and colorectal cancer. Therefore, it is safe to conclude, while there
is no compelling epidemiologic evidence to support an inverse relationship between soy
intake and risk of colorectal cancer, there exist a need for studies that are designed

speciﬁcally to determine if soy consumption lowers the risk for colon cancer.

ii. Animal studies
Much research has been done regarding the effect of feeding isolated soy protein
and soy isoﬂavones on the development of ACF in rats given a chemical carcinogen. At

least three studies showed that soy diets or isoﬂavones inhibited the formation of ACF

27

(Pereira et al., 1994; Helms and Gallaher, 1995; Masaoka et al., 1998). A dose-dependent
inhibitory effect was also reported in these studies (Pereira et al., 1994; Helms and
Gallaher, 1995; Masaoka et al., 1998). Two other studies failed to show an inhibition of
ACF formation by isoﬂavone supplement. In a study reported by Davies et al. (1999),
isoﬂavones (570 mg/kg diet) did not inhibit formation of ACF when compared to a
casein-based control diet. Gee et al. (2000) performed two experiments with genistein or
soy protein isolate given at different time points relative to carcinogen administration.
Feeding genistein or soy protein isolate prior to the injection of dimethyl hydrazine
(DMH) increased the formation of ACF by 2-3 fold, while there was no effect of
genistein or soy protein isolate when administered after the treatment of DMH.

Most animal experiments have not shown a signiﬁcant protective effect of
soybean products given as a source of protein in chemically-induced colorectal
tumorigenesis (Clinton et al., 1979; McIntosh et al., 1995). The incidence of colorectal
tumors induced by DMH was not lower in rats fed soy protein than those fed meat protein
(McIntosh et al., 1995). An isoﬂavone rich diet (500-600mg /kg diet) did not inhibit
tumorigenesis in either AOM-treated rats or the Min mouse model (Sorensen et al., 1998;
Davies et al., 1999). The Min mouse has Single mutated copy of the APC gene and is a
model for human familial adenomatous polyposis. However, a recent study, found that
soy protein isolate containing 430mg of isoﬂavones per kg diet substantially decreased
AOM-induced colorectal tumors in the F2 generation of rats, with their parents fed the

same diet before mating (Hakkak et al., 2001).

28

iii. Chemical composition of soy and proposed anti-carcinogenic compounds in say

The composition of soybeans may vary somewhat according to variety and
growing conditions. The average proximate composition of soybeans is 40% protein,
20% lipid, 35% carbohydrates and 4.9% ash. Through plant breeding it has been possible
to obtain protein levels ranging between 40% and 45%, and lipid levels between 18% and
20%. Several putative anti-carcinogenic phyto-chemicals have been identiﬁed in
soybeans. These include isoﬂavones, protease inhibitors, phenolic acids, phytosterols,
phytates, and saponins.

The variety of soy products available for human consumption and used for
research differ markedly in their concentrations of macronutrients (protein, fat, and
carbohydrates) and micro-constituents (isoﬂavones, saponins, phytic acid, phytosterols,
vitamins, and minerals; Endres, 2001). Methods of processing can dramatically inﬂuence
the concentrations of isoﬂavones and other phytochemicals present in soy products
(Figure 1). De-fatted soy ﬂour (obtained by hexane extraction of soy ﬂakes) contains
signiﬁcant concentrations of isoﬂavones and other phytochemicals. However, ethanol
extraction of soy ﬂour to produce soy protein concentrate removes the majority of the
isoﬂavones and oligosaccharides that were originally present in soy ﬂour (US.
Department of Agriculture, 1999). Table 1 shows the substantial compositional
differences in macronutrient content among soy ﬂours, concentrates, and isolates. The
chemical composition of soy concentrates or isolates is affected by processing parameters
such as solubilization method and temperature. Either water or alcohol (ethanol) washing
can be used to concentrate the protein fraction by removing sugars and oligosaccharides

from soy. Water and, to some degree, ethanol will also remove some of the low-

29

 

Beans

Cleaning Cracki
Dehulllng Conditioning
Making

m...

011 Extraction

Oil

/ \

   
   

Protein and
Carbohydrate
Reﬁned - -
E t . . . Lee th
xsre'apitirggoznd Grinding Say 0;] l m
50, 1,5,. Soluble Carbohydrate
11;: . r. -. . . . ..:.i Extraction
Extrusion
Granulation
Traditional
{.6 Textured - Soy Concentrate

_ . _ ’1' Flour
5.5.93 15.01%”. f3

  
  

Molecular
“Eight
Extrusion Modiﬁcation
Textured Soy Functional Soy
Concentrate Concentrate

 

 

 

Figure 1. Stages of soy bean processing

30

Table 1. Percentage composition of soy protein products (moisture-free basis)l

 

 

Constituent Defatted flours Concentrates Isolates
and grits
Protein (N x 6.25) 56-59 65-72 90-92
Fat 0.5-1.1 0.5-1.0 0.5-1.0
Crude ﬁber 2.7-3.8 3.5-5.0 0.1-0.2
Soluble ﬁber 2.1-2.2 2.1-5.9 <0.2
Insoluble ﬁber 17-17.6 13.5-20.2 <0.2
Ash 5.4-6.5 4.0-6.5 4.0-5.0
Carbohydrates 32-34 20-22 3-4
(by difference)

 

1 Modiﬁed from Endres, 2001.

molecular-weight peptides (those below 2000 daltons) that contain high amounts of
sulfur amino acids. Alcohol washing also removes most of the isoﬂavones (and other
lipid-soluble phytochemicals such as saponins) from the product (Coward et al., 1993;
Murphy et al., 1999). Table 2 lists the total and individual isoﬂavone contents of selective
soy foods (Jackson and Gilani, 2002).

Researchers often compare effects of ethanol-washed with un-washed soy protein
and assume that any differences observed in health outcomes are due to the isoﬂavone
component. However, saponins, other alcohol soluble constituents, and the protein and
peptide composition could also contribute to disparate health effects of these soy
preparations. The composition of commercialized soy foods varies substantially. Setchell
and Cole (2003) recently evaluated the variations in isoﬂavone levels in soy foods and
soy protein isolates. They found a two- to three-fold variation in total isoﬂavones in soy
protein isolates obtained over three years and ﬁve-fold differences in isoﬂavone levels in
different commercial soy milks. The differences among products are due in part to the

variety of soybeans, differences in growing and storage conditions, and differential food

31

Table 2. Estimated isoﬂavone contents of selected soy protein products“2

 

 

Description Daidzein Genistein Glycitein Total
isoﬂavones

Soy ﬂour, full-fat, roasted 99.27 98.75 16.40 198.95

Soybean ﬂakes defatted 36.97 85.69 14.23 125.82

Soy protein concentrate, aqueous 43.04 55.59 5.16 102.07
washed

Soy protein concentrate, produced 6.83 5.33 1.57 12.47
by alcohol extraction

Soy protein isolate 33.59 59.62 9.47 97.43

 

1 Modiﬁed from Jackson and Gilani, 2002
2 Values based on mg/ 100 g edible portion

processing techniques. Therefore, it is important to accurately assess the chemical
composition and phytochemical contents of soy products when investigating the potential
health effects of these products. The following sections will review evidence on the

potential efﬁcacy of these compounds to inﬂuence CC risk.

a). Isoﬂavones

Isoﬂavones are a group of naturally occurring heterocyclic phenols found in
soybeans and forage plants. Genistein (4’, 5, 7-dihydroxy isoﬂavone), and daidzein (4’,
7-dihydroxy isoﬂavone), occurring as their beta-glucosides genistin and daidzin,
respectively are the principal isoﬂavones found in soybeans consisting as much as 90%
of the total isoﬂavone content (Wang and Murphy, 1994b). Soybeans are a particularly
unique and rich source of isoﬂavones in the human diet, containing genistein and
daidzein at levels up to 3 mg/g of dry matter (Coward et al., 1993). Genistein and
daidzein derivatives usually are present in roughly equal concentrations in soybeans and

soy products. Isoﬂavones are insoluble in solvents typically used for soybean oil

32

extraction. Hence, soy ﬂour (50% protein) is a rich source of isoﬂavones, containing
about 3 mg of total isoﬂavones per gram of dry matter. Other Asian style soy-based
products (soy milk, tofu, soy powder and soy nuts) have total isoﬂavone concentrations
in the range of 1.3 - 3.8 mg/g dry weight. Extraction of soy ﬂour with alcohol to produce
soy concentrate (65% protein) results in low concentrations of isoﬂavones in the
concentrate as isoﬂavones are soluble in alcohol. Soy concentrates prepared by water
extraction contain isoﬂavone concentrations similar to that found in soy ﬂour. A portion
of isoﬂavones are lost during preparation of soy protein isolate (90% protein). Soy
protein isolates prepared by conventional means (alkaline solubilization following acid
precipitation of protein) contain approximately 1 mg or more total isoﬂavones per gram
dry matter. Soy ﬁber contains only small concentrations of isoﬂavones as they are
normally associated with the protein fraction of the soybean.

Soy products that are prepared by processes that do not involve fermentation
normally contain isoﬂavones in their conjugated forms. However, fermentation of soy
products results in the hydrolysis of isoﬁavone glycosides, releasing their aglycone
forms. Tempeh and miso, which are fermented soy products typically consumed in Asian
countries, contain a signiﬁcant quantity of their total isoﬂavones in the aglycone form.
Non-fermented soy-foods typically consumed in Asia, such as soy milk and tofu,
primarily contain isoﬂavone glycosides. Soy sauce contains only a minor concentration
of isoﬂavone glycosides (Coward et al., 1993; Wang and Murphy, 1994a; Franke and

Custer, 1996).

33

b). Nutritional intake

The issue of what constitutes normal dietary levels of isoﬂavone intake,
particularly in Asian populations where soy foods are a staple, is controversial. Based on
urinary isoﬂavone excretion, researchers ﬁrst suggested that typical intakes of
isoﬂavones ranged from 50 to 150 mg/day in Japanese adults (Setchell et al., 1984). This
was later disputed and a more conservative estimate of 50 mg/day was proposed as more
likely (Messina, 1995). There have been no direct attempts to estimate daily intake of
phytoestrogens in Japanese adults, although several have used dietary recall to provide
estimates of soy food intake in persons living in Japan (Wilcox et al., 1995; Nagata et al.,
1998; Wakai et al., 1999), China (Chen et al., 1999), Indonesia (Purba et al., 1999), and
Japanese-Americans in Hawaii and Canada (Wilcox et al., 1995; Maskarinec et al., 1998;
Nagata et al., 1998; Chen et al., 1999; Purba et al., 1999; Wakai et al., 1999). Nagata et
al. (1998) reported that the average daily amount of soy foods consumed by Japanese
adults is 54.4 and 63.6 g for women and men, respectively. However, it should be noted
that there was a huge individual variation. This intake of soy foods corresponds to 8.00 g
and 6.88 g of total soy protein for women and men, respectively. From these ﬁgures, a
reasonable assessment of isoﬂavone intake can be calculated by assuming 2—5 mg
isoﬂavones per gram soy protein (Setchell and Cassidy, 1999). Based on this assumption,
Japanese adults probably consume 15—45 mg of isoﬂavone/day on average. Estimates of
isoﬂavone intakes by Chinese adults are similar (Chen et al., 1999), while it is evident
that Indonesians must ingest much higher levels based on the relatively large quantities of
tofu and tempe consumed (Purba et al., 1999). Recent published ﬁgures Show the median

intakes of these foods to be 125 g/day for elderly people living in Jakarta with the range

34

being 62 and 250 g/day for the 25th and 75th percentiles of intake. This represents a
considerable intake of soy foods and while there is a wide variability in isoﬂavone
content of tempe and tofu, this study would suggest that intakes in excess of 150 mg/day
may be attainable. Interestingly, the incidence of breast and prostate cancers in Indonesia
are considerably lower than observed in Japan and China, and the incidences of these
cancers would appear to be inversely correlated with the isoﬂavone intakes observed in
these three countries. Intakes of isoﬂavones by Asians are much higher than those of
Westerners, which are clearly negligible based upon the extremely low isoﬂavone
concentrations found in urine and plasma. Other studies on isoﬂavone intakes in Asian
countries and in Asian populations in Western countries seem to support these low
estimates of the Westerners (Yuan et al., 1995; Wu et al., 1996; Wu et al., 1998; Chen et

al., 1999).

c). Overview of metabolism

The biologic activity and metabolic fate of dietary soy isoﬂavones differ
depending on their chemical form (Kelly et al., 1993; Cassidy et al., 1994; Xu et al.,
1994; Joannou et al., 1995; Knight and Eden, 1995; Cassidy, 1996). Whereas structure
itself is not a limiting factor for absorption from the gastrointestinal tract, (Hendrich et
al., 1999) the chemical form of the isoﬂavone and its metabolites inﬂuence the extent of
absorption, with aglycones more readily absorbed and more bioavailable than highly
polar conjugated species (Setchell et al., 1984; Izumi et al., 2000; Setchell, 2000).

Following ingestion, the acetyl and malonyl derivatives of genistin and daidzin

are metabolized to genistin and daidzin, which are then hydrolyzed in the large intestine

35

by bacteria, resulting in the removal of the sugar moiety to produce their respective
aglycones, daidzein, and genistein (Setchell et al., 1984; Kelly et al., 1993; Izumi et al.,
2000; Setchell, 2000). Following absorption of the aglycones, these compounds and their
metabolites are readily conjugated in the liver with glucuronic acid and/or sulfate,
circulate enterohepatically with potential metabolism and reabsorption in the intestine
(Barnes et al., 1996). Those isoﬂavones that are not absorbed are excreted in the
unconjugated form in feces (Adlercreutz et al., 1995). The glucuronide fraction, the
predominant conjugate representing up to 90% of circulating isoﬂavones in both rats and
humans, (Holder et al., 1999; Doerge et al., 2000) is considered biologically inactive,
(Cassidy, 1996) whereas the free and sulfated fractions, present at much lower
concentrations, are generally thought to be biologically active. Alternatively, daidzein
may be further metabolized by resident microﬂora in the gastrointestinal tract to equol
and O-desmethylangolensin (via their respective intermediates, dehydroequol and
dihydrodaidzein) (Kelly et al., 1993; Bayer et al., 2001). Similarly, genistein may be
metabolized to 6'-hydroxy-O-desmethylangolensin via the intermediate dihydrogenistein
(Joannou et al., 1995) and further to p-ethyl-phenol (Goldwyn et al., 2000). A recent in
vitro study by Kulling et al. (2000) identiﬁed hydroxylated metabolites of both genistein
and daidzein using liver microsomes from Aroclor-treated male Wistar rats, which the
authors suggested might function as an important metabolic pathway in vivo. Metabolites
of glycitein have recently been tentatively identiﬁed in human urine as 5'-OH-O-
desmethylangolensin and 5'-methoxy-O-desmethylangolensin (Heinonen et al., 2000).
The absorption, distribution, metabolism, and excretion (ADME) of soy

isoﬂavones vary, based on age or gender, and among cultural groups. Interindividual

36

variability has been documented in several studies, (Kelly et al., 1995; Xu et al., 1995;
Gooderham et al., 1996; Morton et al., 1997; Xu et al., 2000) although evidence that
isoﬂavones may induce their own metabolism in some individuals is inconclusive (Lu et
al., 1995; Barnes et al., 1996; Lu et al., 1996; Karr et al., 1997). Investigation of
variability in the ADME of isoﬂavones between males and females has produced
inconsistent results (Lockhart et al., 1978; Kirkman et al., 1995). Absorption and
excretion of isoﬂavones have been reported in infants, with evidence that these may vary
because the underdeveloped gut microﬂora cannot hydrolyze the glucuronide forms
(Cruz et al., 1994; Huggett et al., 1997). Additionally, differences in metabolic pathways
may arise owing to different subpopulations of microﬂora, intestinal transit time, pH, or
redox potential, (Hendrich et al., 1999) factors that are inﬂuenced by diet, drugs
(including antibiotics), bowel disease, surgery, and host immunity (Knight and Eden,
1995). Differences among cultural or geographic populations have not been
systematically investigated, although it has been reported that some individuals are not
able to form equol (Adlercreutz et al., 1991; Adlercreutz et al., 1995; Barnes et al., 1996).
Pharmacokinetic studies demonstrate that after oral ingestion of individual
isoﬂavones, peak plasma concentrations are attained 5—6 hours later for the aglycones
and the clearance from plasma proceeds with a half-life of systemic elimination of 6—8
hours (Setchell et al., 2001 ). Notable differences are seen in the pharmacokinetics of
daidzein and genistein and their corresponding B-glycosides. Plasma concentrations of
genistein are consistently higher than daidzein when equimolar amounts are ingested.
This is attributed to the much greater volume of distribution of daidzein compared with

genistein and its higher clearance rate. The bioavailability of genistein is higher than that

37

of daidzein and the overall bioavailability of isoﬂavones is highest when they are
ingested as aglycones versus 0 ~glycosides, as determined following single-bolus oral
administration (Setchell et al., 2001). The rate of absorption of the aglycones is much
faster than that of the B -glycosides (Setchell et al., 2001), a ﬁnding also conﬁrmed in
another study comparing a fermented and unferrnented soy food product (Izumi et al.,
2000). This would be predicted based on chemical structure, because at normal intestinal
pH the aglycones will be rapidly absorbed by a process of non-ionic passive diffusion.
Setchell et al. (2001) demonstrated that the time required to reach the maximal plasma
concentration of isoﬂavones is signiﬁcantly longer when the B-glycosides are ingested.
However, the aglycones are more vulnerable than the corresponding B -glycosides to
further degradation to an array of other metabolites (Kelly et al., 1993; Joannou et al.,
1995), thus limiting their bioavailability. Setchell also revealed a curvilinear relationship
between the systemic bioavailability (as measured from the AUC of the plasma
concentration curves) and amount of isoﬂavones ingested, at least where food is
concerned (Setchell, 2000). Based on the pharmacokinetics, maximum steady-state
plasma levels are more likely to be attained by repeated ingestion throughout the day of
several servings of soy foods having modest isoﬂavones levels. This contention is
supported by data from studies of infants fed soy formulas where the plasma levels
attained by repeated feeding throughout the day are approximately tenfold higher than
those observed in adults consuming similar quantities of isoﬂavones (Setchell et al.,

1997).

38

d). Biological effects

In vitro studies have indicated that isoﬂavones have a number of interesting
biological effects. Genistein is well known as a weak estrogen receptor agonist with an
activity that is approximately 1/1000th that of estradiol. Genistein has been demonstrated
to inhibit the activities of a number of cellular enzymes involved in cell cycle regulation,
including both receptor and cytosolic protein tyrosine kinases (Akiyama et al., 1987;
Makishima et al., 1991), ribosomal S6 kinase (Linassier et al., 1990), MAP kinases
(Thorbum and Thorburn, 1994), DNA topoisomerases I and II (Constantinou et al.,
1990), epidermal growth factor induced phosphatidyl inositol turn over (Imoto et al.,
1988) and increase transforming growth factor B (TGF [3; Peterson and Barnes, 1996).
Genistein also has been demonstrated to elicit anti-oxidant (Jha et al., 1985; Wei et al.,
1995), anti-proliferative (Yanagihara et al., 1993), and anti-angiogenic (F otsis et al.,
1993; Raines and Ross, 1995) activities. Daidzein is not an effective inhibitor of tyrosine
kinases, but has been found to inhibit the activity of casein kinase II (Higashi and
Ogawara, 1994).

Most of the anti-carcinogenic potential of isoﬂavones has been inferred from
effects of isoﬂavone aglycones on the growth of cancer cells in culture. Several in vitro
studies have shown that genistein inhibits growth of a wide range of both horrnone-
dependent and hormone—independent cancer cell lines with IC50 values ranging from 5—
100 uM (approximately 2—25 ug/ml). In addition, genistein administration has been
shown to inhibit the in vivo metastatic activity of both breast (Scholar and Toewa, 1994)
and prostate cancer cells (Santibanez et al., 1997) independent of effects on cell growth.

Genistein inhibits the growth of colon cancer cell lines HCT-8 (Clark et al., 1989) HCC-

39

48 and HCC-50 (Yanagihara et al., 1993), and caco-2 and HT-29 (Kuo, 1996) in a dose
dependent manner.

The precise mechanisms by which genistein inhibits growth of cancer cells are
still unknown, but several potential mechanisms have been proposed. Inhibition of cell
growth induced by genistein is associated with cell cycle arrest at G2 — M and induction
of differentiation and apoptosis (Matsukawa et al., 1993; Pagliacci et al., 1994; Shao et
al., 1998). The effects of genistein on cell growth are believed to be mediated largely by
the inhibition of key enzymes (tyrosine kinases, MAP kinases, DNA topo-isomerases)
involved in cell cycle regulation. Enzymes in the protein tyrosine kinase family
phosphorylate the tyrosine residues on key proteins involved in signal transduction events
in normal and tumor cells (Akiyama et al., 1987; Makishima et al., 1991). Gensitein
inhibits DNA topoisomerase I and II, which are involved in DNA replication. Peterson
and Barnes (1996) reported that genistein increased the in vitro concentrations of TGF-B
and this may be important given the role of TGF-B in inhibiting the growth of cancer
cells. In Caco-2 cells, genistein induced differentiation as measured by changes in
alkaline phosphates and dipeptidyl peptidase, and this enzyme induction was correlated
with the phosphorylation of three different proteins (Basson et al., 1998). These results
are consistent with earlier studies that showed genistein inhibited gastrin-stimulated
growth in non transformed intestinal crypt cells (Raid et al., 1993) and in colon mucosa
(Majumdar, 1990). The blockade in gastrin stimulation of cell growth by genistein was
mediated by reduced tyrosine phosphorylation of several proteins (Majumdar, 1990; Raid

etaL,l993)

40

In vitro studies indicate that the inhibitory effect of genistein on cancer
development can only be achieved at relatively high concentrations. The ICso required for
growth inhibition is typically greater than 13.2 mol genistein/L (Barnes et al., 1995) and
can vary as high as 90 umol genistein/L (Kuo, 1996). However, plasma levels of
genistein in people consuming soy rich foods have estimated to be in the range of 1—6
pmol genistein/L (Xu et al., 1995; King and Bursill, 1998; Adlercreutz et al., 2000).
Given the short half- life of genistein (4.8—5.7 hours), and the fact that most of plasma
genistein is conjugated with glucuronic acid, a very high intake of and a repeated
exposure to genistein is required to achieve plasma concentrations in the range of in vitro
concentrations required to inhibit cancer cell growth.

Various in vitro and in vivo assays show genistein to be a weak estrogen receptor
agonist, exerting an estrogenic effect ranging from approximately 1 x 10'3 to 1 x 10'5 that
of 17-0 estradiol (Mayr et al., 1992; Markiewicz et al., 1993). Weak estrogens have been
postulated to function as anti-estrogens in vivo (Martin et al., 1978) and thus,
theoretically, may reduce the risk of hormone-dependent cancers such as breast and
prostate cancer (Setchell et al., 1984). However, this effect seems to be tissue speciﬁc and
in tissues such as heart and bones genistein has an estrogenic effect. Traditionally, CC
has not been regarded as being a hormone-dependent cancer, but some evidence suggests
that CC can be hormonally inﬂuenced. Although world wide CC incidence rates among
men and women are similar, there is a gender variation in the sub sites of CC. McMichael
and Potter (1980) have reported a female excess of right-Sided colon cancers and a male
excess of left-sided cancers. This may be due to differences in gut metabolism, colonic

bacterial populations, and fermentation rates between the sexes, which could be mediated

41

ultimately by hormones. Studies have also shown that there was a brief decline in CC
incidence among women compared with men in countries where use of the original high-
dose oral contraceptives were common. McMichael and Potter (1980) also suggested that
oral contraceptive use was associated with low rates of CC in women. Chen et al. (1998)
found that women who used hormone replacement therapy (HRT) in the year prior to
sigmoidoscopy, were only half as likely to have colorectal adenomas compared to the
women who did not use HRT. Similarly, in a case-control study conducted by Newcomb
and Storer (1995) it was found that recent users of HRT had a CC relative risk of 0.54
when compared with post-menopausal women who never used HRT. Grodstein et al.
(1998), in a prospective study involving 59, 000 post-menopausal women, found that
current users of HRT had a relative risk for colorectal cancer of 0.65. In contrast,
Maclennan et al. (1995) had published a meta-analysis on fourteen studies related to HRT
and colon cancer that concluded that HRT was not protective. As soy isoﬂavones have
been shown to exert both estrogenic and anti-estrogenic effects, they may have the
potential to affect CC risk through hormonal mechanisms, if a role for estrogen is
established in the etiology of CC.

As mentioned previously, some research conﬁrms a protective effect of
isoﬂavones against carcinogenesis in animal models. Pereira et al. (1994) tested the
ability of genistein to inhibit the formation of ACF in colons of rats treated with lSmg/kg
body weight AOM and found that the addition of 75 and 150 mg genistein per kg of
semi-puriﬁed diets fed to rats resulted in 29 % and 34 % reductions in ACF respectively.
Although this study Show that genistein inhibits pre cancerous lesions in the colons of

rats, there are none that show if inhibition of ACF development would translate to a

42

reduction in colon tumors. In a related study by Bennink and Om (1998), animals fed
defatted soy ﬂour (total isoﬂavones 873 mg/Kg) resulted in a 47-49 % reduction in
AOM-induced colon tumor incidence and a 60-74 % reduction in tumor burden in
comparison to rats fed soy protein concentrate or casein. Feeding a mixture of genistin,
daidzin and glycitin isoﬂavones (concentrations equivalent to those found in de-fatted
soy ﬂour) added in the soy concentrate diet did not inﬂuence tumor incidence but feeding
genistin (equivalent to 500 mg/kg genistein) added to the soy protein concentrate
increased tumor incidence (P = 0.06) (Bennink and Om, 1998). This result conﬁrms the
study by Rao et al. (1997) who found increased AOM-induced colon carcinogenesis in
male F344 rats when fed diets containing 500mg genistein lkg diet beginning two weeks
prior to, during and after carcinogen administration. However, it should be noted that in
the study by Bennink and Om (1998) adding genistin or an isoﬂavone mixture to soy
concentrate or feeding defatted soy decreased average tumor weight by about 50 %.
These results suggest that genistin alone is sufﬁcient to inhibit tumor growth but not
tumor incidence. From the same study, it can also be concluded that defatted soy ﬂour
(with ethanol soluble phytochemicals) inhibited colon carcinogenesis. Therefore it
appears that other ethanol soluble anti-cancer constituents such as, saponins, phytosterols,
protease inhibitors and phenols that are present in the defatted soy ﬂour may be

responsible for the anti-cancer effect of de-fatted soy ﬂour.

e). Protease inhibitors

Protease inhibitors are proteins found in cereals, nuts, fruits, vegetables, eggs, and

potatoes with typical molecular weights between 8,000 and 10,000 daltons. As much as 6

43

% of all soy protein consists of protease inhibitors, making soy a particularly plentiﬁrl
source of these compounds (Birk, 1990). Troll et al. (1980) initially proposed that
soybean protease inhibitors could suppress cancer. Subsequent research determined that
chymotrypsin-speciﬁc inhibitors, but not trypsin-speciﬁc inhibitors, suppress
transformation in vitro (Kennedy, 1993). Feeding the soybean Bowman-Birk inhibitor
(BBI, chymotrypsin inhibitor) suppressed chemically induced CC in mice (Weed et al.,
1985; St. Clair et al., 1990).

The exact mechanism whereby chymotrypsin inhibitors can prevent cancer is not
yet known. In vitro research demonstrated that BBI acts between the initiation and
transformation stages of carcinogenesis. BBI does not prevent initiation, but it does
irreversibly inhibit critical steps that lead to cell transformation. If a cell is already
transformed, BBI cannot reverse transformation.

BBI has been evaluated as puriﬁed BBI (PBBI) or as an extract of soybeans
enriched in BBI, termed BBI concentrate (BBIC), which is under evaluation in hmnan
trials as an anti-carcinogenic and anti-inﬂammatory agent. PBBI and BBIC have been
shown to inhibit radiation- and chemically- induced malignant transformation in vitro
(Kennedy, 1993) and animal carcinogenesis in DMH induced colon and liver
carcinogenesis in mice (Weed et al., 1985; Billings et al., 1990; St. Clair et al., 1990) and
Spontaneous colon carcinogenesis in mice that are genetically susceptible to the induction
of intestinal carcinogenesis (Kennedy et al., 1996). When Kennedy et al. (2002) fed
soybean extract containing 0.5 % BBIC to Min mice, there was a 40% reduction in the
number of tumors/mouse in the small intestine and 42% reduction in tumorigenesis in the

colon (Lichtenstein et al., 2002). Billings et al. (1990) also reported a 45% reduction in

44

DMH induced colon adenocarcinomas with a diet containing 0.1% BBIC in rats. When
animals are given dietary BBIC, it is known that BBI reaches the colon in an active ﬁrm
(Owen and Jones, 1974) and enters colonic epithelial cells (Locniskar et al., 1985). The
doses of 0.1% and 0.5% BBIC used in the animal studies are comparable to doses that are
being studied in human cancer prevention trials (25-400 C.I. units) (Bland and Britton,
1975).

Additional studies regarding soy as a source of protease inhibitors to help prevent
cancer is warranted. The anti-cancer activity of soy protease inhibitors is ascribed to the
BBI, but the BBI in soy products is reportedly less effective than puriﬁed BBI or BBI
concentrates (Kennedy, 1993). BBI is heat sensitive as protease inhibitor activity is
destroyed by autoclaving and autoclaved BBI concentrate is ineffective in preventing oral
carcinogenesis (Messadi et al., 1986). Therefore, how much BBI activity remains in heat
processed soy foods (i.e., roasted soy nuts, spray dried soy protein isolate, or baked and
cooked products containing soy protein) is not known. The quantity of BBI in soy
concentrates is dependent upon the technique used to produce the concentrate.
Concentrates prepared by aqueous alcohol extraction will have greatly reduced BBI
activity since acidic 60 % ethanol is used to extract BBI in the preparation of BBI
concentrate (Yavelow etal., 1985). Much research will be required to answer questions
regarding the efﬁcacy and safety of protease inhibitors as anti-carcinogens. Identifying

soy products which can deliver biologically active BBI to humans is also needed.

45

f). Phytates

Graf and Eaton (1985) proposed that phytic acid (inositol hexaphosphate, 1P6),
could decrease the incidence of CC. Since then, data from epidemiologic (Graf and
Eaton, 1993), animal and cell culture studies (reviewed by Shamsuddin, 1995;
Shamsuddin et al., 1997) have been published to support this hypothesis.

Several research groups (Pretlow et al., 1992a; Shamsuddin, 1995; Alabaster et
al., 1996) have demonstrated that 1P6 can inhibit chemically-induced CC in animal
models and that 1P6 can inhibit both the initiation and promotion stages of carcinogenesis.
Both in vitro and in vivo studies suggest that 1P6 acts by inducing cell differentiation
(Shamsuddin et al., 1997) or by reducing cell proliferation (Corpet et al., 1997). It is
reported that 1P6 blocks activation of phosphatidyl inositol—3 kinase which in turn inhibits
cell transformation and may account for reduced cell proliferation (Corpet etal., 1997)
and enhanced cell differentiation (Shamsuddin et al., 1997).

There is very little doubt that isolated phytate fed at 1—3% of the diet can inhibit
CC in animal models (Shamsuddin, 1995). However, phytic acid in plant material is
mostly bound and exists as either protein-phytic acid complexes or as protein-phytic acid-
cation complexes (Szwergold et al., 1987). Since complexed 1P6 is less available, it is less
effective as an anti-carcinogen than isolated, neutralized phytate (Shamsuddin et al.,
1997). Thus, it is unlikely that the potential of dietary soy to inhibit colon carcinogenesis

is explained by its phytic acid content.

46

g). Phenolic acids

Some phenolic acids have anti-mutagenic activity in vitro and anti-carcinogenic
activity in vivo. Defatted soy ﬂakes contain about 1.1 mg of total phenolic acids per gram
and soy isolate contains about 0.5 mg phenolic acids per gram. Ferulic, isoferulic, and
caffeic acids are the major phenolic acids present in soybeans (Seo and Morr, 1984) and
they have been shown to inhibit carcinogenesis in animal models (Tanaka et al., 1993)
The anti- cancer activity of phenolic acids is most likely due to their antioxidant
properties and to the ability of phenolic acids to reduce hyperplasia (Stich, 1991; Tanaka
et al., 1993). There is insufﬁcient data to conclude if phenolic acids reduce CC or if
dietary soy can provide sufﬁcient quantities of phenolic acids to inhibit colon

carcinogenesis.

h). Phytosterols

Soybean phytosterols are found primarily in the oil fraction of the bean. Soybean
oil contains about 372 mg of total phytosterols per 100 g of oil (Weihrauch and Gardner,
1978). The main phytosterols in soybeans are B-sitosterol (56%), campesterol (21%), and
stigmasterol (20%). Reﬁning the oil reduces the phytosterol content by about 30%,
whereas reﬁning followed by hydrogenation reduces phytosterol content by about 60%.

There are both epidemiologic and experimental CC studies suggesting that dietary
phytosterols may help prevent CC. Epidemiologic studies indicate that there is an inverse
relationship between dietary phytosterol intake and risk of CC (Nair et al., 1984; Rao and
Janezic, 1992). Raicht et al. (1980) induced CC in rats with N-methyl-N-nitrosourea and

showed that 0.2% dietary B-sitosterol reduced both CC incidence and the number of

47

tumors per rat. Awad et al. (1997) reported that B-sitosterol possibly inhibits the growth
of HT-29 colon cancer cells through altered sphingomyelin metabolism. Growth
inhibition was associated with increased cellular concentrations of ceramide.
Vegetarians of the Seventh-day Adventist religion (a population with low risk for
colon cancer) consume about 350 mg of phytosterols per day. Soybeans contribute no
more than 10% of the total phytosterol intake for Seventh Adventist vegetarians.
Therefore, it is unlikely that soy phytosterols could have a major role in reduction of

colon cancer risk in either vegetarians or non-vegetarians.

i). Saponins

Soybeans are the major dietary source of saponins since saponin (triterpenes)
distribution is limited in commonly consumed foods. The saponin contents of whole
soybean, soybean ﬂour, tofu, soy isolate, and lecithin are 5.6, 2.2, 2.1, 08—2, and 2.9% of
the dry weight, respectively (Oakenfull, 1981). Part of the saponins is removed during oil
extraction and an additional portion is removed from the oil during reﬁning.

Saponins are amphiphillic compounds composed of water soluble sugar residue
attached to a lipid- soluble aglycone (sapogenol). The biological activity of each saponin
depends upon its polarity, acidity and hydrophobicity, characteristics determined by
chemical structure (Milgate and Roberts, 1995). Under invitro conditions, various
saponins have demonstrated antimutagenic, anticarcinogenic and antimetastatic effects
against multiple cell lines (Rao and Gurﬁnkel, 2000). Soyasponins encompasses three
groups of sapogenols, are comprised of a neutral, non polar oleane-lZ-ene triterpene

aglycone with differing sugar moieties linked at one or more glycosylation sites (Yoshiki

48

et al., 1998). Crude soybean saponin mixtures at concentrations ranging from 150 to
2400mg/L inhibit growth of multiple human colon adenocarcinoma cell lines in vitro.
Treatment of human colon cancer cell lines with soyasaponins suppressed growth,
induced morphological alterations, increased multiple markers of differentiation, and
inhibited protein kinase C (PKC) activity (Sung et al., 1995a; Sung et al., 1995b; Oh and
Sung, 2001). Thus, soy saponins exhibit inhibitory effects toward neoplastic cells, actions
that are similar to those of various plant saponins reported to be anti-neoplastic and anti
carcinogenic agents (Rao and Gurﬁnkel, 2000). Following consumption of soy, soy
saponins appear to pass undigested through the small intestine. Saponins may be partially
metabolized to sapogenols by bacterial glycosidases in the colon and the resulting
mixture of saponins and sapogenols can presumably interact with the epithelium
(Gestetner et al., 1968; Gurﬁnkel and Rao, 2003; Hu et al., 2004). Soyasaponins could
therefore potentially modulate colon carcinogenesis.

A study by Koratkar and Rao, (1997) found that a diet containing 3% soy
saponins, when administered 1 week after a series of AOM injections, reduced the
average number of ACF per colon in CFl mice as compared with mice consuming a
control AIN93G diet (2.6 vs 7.67 ACF per colon respectively). As saponins are not
absorbed and remain in the gastrointestinal tract they may facilitate binding of bile acids
to ﬁber (Sidhu and Oakenfull, 1986). There is evidence that bile acids may act as tumor
promoters in the colon (Reddy and Wynder, 1977). Therefore, reducing the concentration
of soluble bile acids in the colon could potentially reduce colon cancer. Better analytical

methods to quantify saponins are needed to aid in anti-cancer research of soy saponins.

49

2.3. Role of calcium in colorectal chemoprevention
i. Epidemiology
In the last decade there has been considerable interest in the study of calcium as a

chemopreventive agent against colorectal cancer. Several epidemiologic studies Show
(Garland et al., 1985; Sorenson et al., 1988) that CC incidence is lower in regions of
western industrialized societies where consumption of calcium-containing dietary
constituents such as dairy products were relatively high. However, epidemiological data
regarding the association between calcium ingestion and colorectal cancer or adenoma
have varied. A number of case-control and cohort studies have suggested that increased
intake of calcium and of vitamin D may be associated with a reduced CC incidence.
However, other studies (Bostick et al., 1993b; Neugut et al., 1996) have not clearly
supported this association. The Iowa Women’s Health Study (Bostick et al., 1993b)
investigated whether a high intake of calcium, Vitamin D or dairy products protected
against colon cancer among women (Lointier et al., 1992; Brenner et al., 1994) over 55
years of age. The relative risks (R) for CC observed for the highest quintile of intake as
compared with the lowest were 0.52 (95% CI 035-084) for calcium and 0.54 (95% CI
0.35-0.84) for vitamin D. However, when incorporated into a multivariate model, the
trends lost statistical signiﬁcance and RR were attenuated (Bostick et al., 1993b). Tseng
et al. (1996) evaluated the association between a number of micronutrients including
calcium and the risk of colorectal neoplasia. In this study, a reduced risk of adenomas
was conﬁned to men with the highest calcium quartile. Another case-control study
evaluated the association between regular usage of supplements containing vitamins A,

C, D and E or calcium with the risk of developing colorectal adenomas (new or

50

recurrent). No protective effect was documented for any of the nutrients examined,
including calcium (Neugut et al., 1996). Overall the epidemiological data on calcium
intake is quite heterogeneous, suggesting a very modest beneﬁcial effect at best.
Newmark et al. (1984) proposed a hypothesis based on the consideration that
consumption of calcium leads to increase in ionic concentration of calcium in the
intestinal lumen. He suggested that luminal calcium reacts with bile acids and dietary
long-chain fatty acids to form insoluble complexes, thereby nullifying the cytotoxic
effects of these lipids. Several experimental animal studies conducted by Wargovich et al.
(1983; 1984) have shown that calcium can ameliorate the damaging effects of these
agents when applied intra-colonically. These ﬁndings were supported in later animal
studies, which found that calcium prevented the cytotoxic effect of intestinal lipids and
prevented increased in intestinal cell proliferation, thereby inﬂuencing the development

of colorectal cancer (Bird, 1986; Rafter et al., 1986).

ii. Mechanism
A number of mechanisms have been postulated to explain the putative

chemopreventive effect of calcium against CC development. Calcium may act through
the binding of bile and fatty acids in the intestinal lumen and decreasing exposure of the
intestinal epithelium to those substances. Bile and fatty acids can produce toxic effects to
the colonic epithelium and cause increased cell proliferation or compensatory hyper-
proliferation secondary to repair responses (Bostick et al., 1993a). In either case,
increased proliferation may ultimately lead to carcinogenesis of the colonic epithelium.

Typically, 40 to 90% of ingested calcium is not absorbed in the small intestine and can

51

 

form calcium salts with bile acids or fatty acids. Calcium salts of bile acids may be less
irritating to the colon mucosa than bile salts formed with sodium or potassium
(Wargovich et al., 1983). Therefore, when high fat diets are consumed and there are more
bile acids passing through the colon, high levels of dietary calcium provide more calcium
potentially available to complex with bile acids. Since calcium salts of bile acids are less
irritating to the colonic mucosa, epithelial cell proliferation is reduced. Thus, calcium
could act in a nonspeciﬁc, non-cellular manner to reduce colon cancer incidence by
reducing cell proliferation.

Calcium may also directly inﬂuence colonic epithelial cell proliferation (Lipkin
and Newmark, 1985) or induce terminal differentiation of colonic epithelial cells
(N ewmark et al., 1984). Increasing the intraluminal concentrations of calcium in the
colon may produce results (decreased cell proliferation and increased cell differentiation)
similar to those observed when colon tissue explants and cells were cultured in medium
containing 2.2 mMol/L calcium (Buset et al., 1986). These data suggest that increased
intraluminal calcium concentrations may decrease cell proliferation and increase cell
differentiation by acting directly on mucosal cells (Buset et al., 1986). Furthermore, it
has been reported that addition of the calcium ionophore A23187 to cell cultures
stimulates proteolysis of the APC protein and induces apoptosis (Browne et al., 1994).
High levels of dietary calcium can attenuate the increased cell proliferation and tyrosine
kinase activity caused by azoxymethane injections in rats (Arlow et al., 1989). These
observations suggest that intraluminal calcium can directly affect mucosal cells in a

manner that is independent of bile salts.

52

Data ﬁ'om the large prospective studies on calcium and colon cancer risk Show
weak, non-signiﬁcant associations, although there is some evidence of a dose-response
relationship (Martinez and Willett, 1998). Some clinical studies have demonstrated an
anti-carcinogenic effect as a result of calcium supplementation. Calcium intervention
trials in humans (Lipkin et al., 1989; Rozen et al., 1989; Steinbach et al., 1994) have
shown that supplementing human diets with calcium can reduce intestinal cell
proliferation (an intermediate biomarker of colorectal cancer risk) in the upper 40% of
colon crypts in people at intermediate or high risk of colon cancer. However, other
studies (Gregoire et al., 1989; Stem et al., 1990) have shown no effect of dietary calcium,
whilst one (Kleibeuker et al., 1993) showed the opposite effect. A review of 17 human
studies of calcium supplementation and colorectal epithelial cell proliferation concluded
that it is unlikely that calcium supplementation can substantially lower total labeling
indices, but it may normalize the distribution of proliferating cells within colon crypts
(Bostick, 1997). A randomized placebo-controlled study reported no effect of calcium
supplementation on rectal mucosal proliferation in patients with previous colon adenomas
(Baron et al., 1995). A double-blind three-year intervention study with calcium (1.6
g/day) and other anti-oxidants in polyp-bearing patients (n = 116) also reported that there
was no overall effect of calcium on polyp growth. However, the study reported that
calcium and the antioxidants prevented formation of new adenomas (Hofstad et al.,
1998).

The effects of calcium supplementation on bile acid content also have been
evaluated clinically. Two different studies documented beneﬁcial calcium-induced

changes in the bile acid composition, which were characterized by a decrease in total as

53

well as secondary bile acid levels in feces (Alberts et al., 1996; Lupton et al., 1996). A
randomized placebo-controlled phase III study (Baron et al., 1999) demonstrated a
moderate (20%) but signiﬁcant reduction in the incidence of new adenomas seen in the
cohort receiving calcium supplementation (1200 mg/d elemental calcium). The adenoma
recurrence rate in the calcium treatment group was 31% and that of the placebo group
was 38% (Baron et al., 1999). Similar results, though not statistically signiﬁcant, were
observed in the smaller European Calcium Fiber Polyp Prevention trial (F aivre et al.,
1999). A recently published analysis of cases from the Nurses’ Health Study and the
Health Professionals Follow-Up Study found that calcium reduced the risk of distal colon
cancer but not of proximal colon cancer (Wu et al., 2002). The study also suggested a
threshold effect whereby most of the beneﬁt was achieved by reaching intakes of 700 to
800 mg/day. Overall, the potential beneﬁcal effects of calcium in reducing colon cancer
risk appear modest.

In rodent models of chemically-induced colon carcinogenesis (promoted by high
fat diets), calcium administration was associated with protection against colorectal
carcinogenesis as evidenced by decreased epithelial cell proliferation, ACF numbers and
tumor formation. High calcium diets (10 g/kg) normalized the distribution of proliferating
colon cells in rats treated with N-methyl-N’-nitro-N-nitrosoguanidine (Reshef et al.,
1990) and inhibited azoxymethane-induced cell proliferation (Li et al., 1998). Other
studies have demonstrated that high-calcium diets reduced aberrant crypt formation
(Pereira et al., 1994; Li et al., 1998; Pierre et al., 2004) and chemically-induced colon
cancer in rats (Appleton et al., 1987; Beatty et al., 1993; Quilliot et al., 1999). Recent

studies in rodent models have further demonstrated that calcium decreases colonic tumors

54

induced by targeted mutations (Yang et al., 1998) or by dietary factors (N ewmark et al.,
2001). Thus, several studies in rodent models have demonstrated that dietary calcium can

reduce development of CC.

E. Biomarkers and models for chemopreventive studies
1. Surrogate end point biomarkers (SEB)

Biomarkers of carcinogenesis are quantiﬁable molecules or changes involved in
physiological or pathological events which occur between exposure to exogenous or
endogenous carcinogens and the subsequent development of cancer (Kelloff et al., 1994).
They can be categorized as molecular, histological, risk and tumor biomarkers. Some of
the biomarkers of early colorectal carcinogenesis can provide a means for diagnosing
very early changes associated with cancer development before actual tumors or even
polyps are present.

These biomarkers are increasingly used in programs of screening,
chemoprevention and chemotherapy for CC. Such biomarkers are termed ‘surrogate
endpoint biomarkers’ (SEBS) and have been identiﬁed for several different stages in the
carcinogenic process. These biomarkers represent a means of monitoring disease
progression without having to wait for true neoplasia and metaplasia to develop. SEBS
are also termed ‘preneoplastic biomarkers’ and Should be distinguished from ‘risk
biomarkers’ and ‘tumor markers’. Of these three types of markers, preneoplastic
biomarkers are the most useful for assessing the efﬁcacy of cancer chemopreventive

agents (Einspahr et al., 1997; Collet et al., 1999; Sharma, 2000). Examples of cellular

55

preneoplastic biomarker measurements include those of proliferation, differentiation and

apoptosis.

2. Cell proliferation biomarkers

Increased epithelial cell proliferation is a common feature of colon
carcinogenesis. Since the discovery of cell—cycle associated proteins such as PCNA and
Ki-67, their expression has been used in several studies (Renehan et al., 2002) for IHC
detection of proliferating cells. Normally, proliferation in colonic crypts is restricted to
the basal 60% of the crypts. As colonocytes migrate up the crypt, these cells undergo
terminal differentiation. Ultimately, after approximately 7 days of transit time to the crypt
surface, colonocytes undergo programmed cell death and are extruded. Under normal
circumstances, cell proliferation does not occur in the upper 40% of the crypts (Lipkin,
1988). An expansion of the proliferative compartment toward the crypt surface
(hyperproliferation in the middle and upper thirds of the crypt) is considered a
preneoplastic biomarker (Lipkin, 1988) and has been described in association with
development of adenomatous polyps and carcinoma (Risio et al., 1991).

Methods used to assess colonic epithelial cell proliferation can be broadly divided
into ‘static techniques’ that reﬂect the proliferation state (percentage of a proliferation-
associated population at any particular time within the colonic crypts) and ‘dynamic
techniques’ which reﬂect cell proliferation rate (including cell cycle phase transit time)
(Renehan et al., 2002). ‘Dynamic’ measurements are more biologically relevant.
However, their assessments require either systemic administration of radioactive agents,

or in vitro incubation of tissue samples with tritiated thymidine or bromodeoxyuridine

56

(BrdU). Therefore, static measurements such as S-phase fraction analysis using ﬂow
cytometry and immunohistochemical (IHC) measurements of cell cycle-associated
intrinsic proteins such as PCNA and Ki-67 antigen are preferred methods because of their
case of use in studies having large numbers of specimens.

PCNA is an evolutionarily-conserved nuclear protein (36 kDa) identiﬁed as an
ancillary factor for DNA delta polymerase. It is necessary for DNA replication and repair
during cell cycle progression. PCNA is detected in cells even at Go and progressively
accumulates during the G to M phases of the cell cycle (Prelich etal., 1987; Toschi and
Bravo, 1988; Zacchetti et al., 2003). Ki-67 is a nuclear antigen that is primarily expressed
during S-phase of the cell cycle (Bruno and Darzynkiewicz, 1992). Therefore, Ki-67
labeling indices reﬂect the number of actively proliferating cells within the crypt,
whereas PCNA indices are indicative of the capacity of cells to continue proliferation.

PCNA and Ki-67 proliferation indices have been used to assess the efﬁcacy of
chemopreventive agents in CC studies and in clinical studies that evaluate the risk for CC
in humans (Baron et al., 1995; Weaver et al., 2000; Grubben et al., 2001; Pfeiffer et al.,
2003; Krishnan et al., 2004). In particular, PCNA expression (as measured by
immunohistochemistry) has been used in human studies to demonstrate a step-wise
increase in proliferation indices during neoplastic progression of colonic lesions (Shpitz
et al., 1997), for predicting the potential for metastases in invasive colorectal carcinoma
(Tanaka et al., 1995; Choi et al., 1997), and to predict the development of metachronous
adenomas (Ottaviani et al., 1999). Studies also have demonstrated increased PCNA
labeling indices by activation of growth factors such as EGF receptor kinase and TGF-a

(Malecka-Panas et al., 1997; Barnes et al., 1998) and by overexpression of cyclins such

57

as Cyclin D (Izawa et al., 2002) in colonic tissue. Furthermore, several studies have
demonstrated that PCNA is an accurate marker for the proliferative activity in colonic
crypts, and also correlates well with other proliferation markers (e. g. BrdU) in both
normal and neoplastic colonic epithelium in humans (Kubota and Kino, 1995; Tanaka et
al., 1995; Yoshikawa and Utsunomiya, 1996; Choi et al., 1997; Shpitz et al., 1997;

McShane et al., 1998; Yerly-Motta et al., 1999; Wu et al., 2000; Diez et al., 2003).

3. Differentiation markers

Cellular differentiation is evaluated through the expression (or lack thereof) of
markers that reﬂect epithelial maturation. Propensity towards an undifferentiated
phenotype is characteristic of neoplastic cells. Molecules that are expressed in a fully-
differentiated epithelial cell may be used as a marker of the pro-differentiating effects of
an intervention. For example, Dilidros biflorus agglutinin (DBA), which labels mucin in
goblet cells, has been used as a marker for colonic epithelial cell differentiation.
Undifferentiated cell phenotypes are associated with failure of oligosaccharide synthesis
and a decrease in DBA staining (Boland et al., 1988). Reduced DBA staining has been
documented in colon adenomas and cancers as well as some familial colon cancer

syndromes (Sam et al., 1990).

4. Apototic markers
Homeostasis of the normal colonic epithelium requires a proper balance between

cell proliferation and cell death (Hall et al., 1994; Ahnen, 1999). When colonocytes

migrate to the luminal surface, they undergo a process of physiologic cell death termed

58

apoptosis and are sloughed into the intestinal lumen. Apoptosis is associated with marked
alterations of cell surface structure (such as reduced cell adhesion) so that cells lose
contact with their neighbors and be sloughed early in the process of apoptosis. Failure of
apoptotic function is a characteristic of neoplastic phenotypes and may play a role in
tumor progression (Bedi et al., 1995).

No markers of cell sloughing exist, but several markers of apoptosis have been
applied to the colon mucosa. However, quantitation of apoptotic end points in colonic
epithelium has proven difﬁcult. This is primarily due to the relatively low rates of
apoptosis observed in colonic epithelial cells and the rapid rate of cell loss after initiation
of apoptosis. Many investigators have used detection of fragmented DNA, a
characteristic of apoptotic cells, and measured by TUNEL immunohistochemistry. Other
methods for detecting apoptosis in human colonic crypts (ﬂow cytometry, caspase
induction, Bcl-2/Bax ratio) have not been sufﬁciently explored to be useful as
reproducible biomarkers of apoptosis in the human colonic crypt. Apoptosis as a
biomarkers for chemopreventive activity in morphologically normal human colonic

epithelium, while of interest and potential importance, remains unvalidated.

5. Aberrant crypt foci
The smallest recognizable histopathologic evidence of colorectal epithelial
carcinogenesis is the ACF or microadenoma. ACF are lesions consisting of large thick
crypts as observed in methelene blue-stained specimens of colon and were originally
found in mice treated with AOM (Bird, 1987). In humans, two types of ACF are

observed: hyperplastic foci (hypercellular but normal appearing cells) and dysplastic foci

59

(contain dysplastic cells Similar to those found in adenomas) (Polyak et al., 1996; Siu et
al., 1997; Takayama et al., 1998). Hyperplastic lesions commonly contain K-ras
mutations but rarely have APC mutations (Pretlow et al., 1993; Jen et al., 1994; Smith et
al., 1994) and are unlikely to progess to neoplasia. Dyplastic lesions, on the other hand,
are thought to be precursors to adenomas. Numerous dysplastic ACFs are found in FAP
patients (Hamilton, 1983). Moreover, the occurrence patterns of ACF and their response
to chemopreventive interventions parallel those of adenomas and carcinomas in rodent
carcinogenesis models (Bird, 1987; Wargovich et al., 1995b; Reddy et al., 1996;
Kawamori et al., 1998). Thus, ACFS represent an excellent early surrogate marker of CC

risk to assess in rodent models and potentially, in humans.

6. Animal models

Animal models represent a pivotal tool in the efﬁcacy evaluation of
chemopreventive agents. Chemical carcinogenesis in rodent models provides
reproducible development of tumors in the intestinal epithelium of rodents that have been
exposed to a chemical compound that can act as a complete carcinogen. Frequently-used
chemical carcinogens include dimethyl hydrazine (DMH), azoxymethane (AOM), and
methoxymethane. DMH undergoes conversion into AOM and azoxymethanol, which are
subsequently conjugated with glucuronic acid and excreted in bile. Bacterial hydrolysis
occurs once the conjugated carcinogens reach the intestinal lumen but is not essential to
their tumorigenic effect (Hartiala, 1973; Fang and Strobel, 1978).

The azoxymethane-injected rat model is considered to be a highly relevant colon

tumorigenesis model system. Although the Speciﬁc etiology of CC and other diet-

60

associated cancers remain unclear, the chemical carcinogen-induced rat colon tumor
system is a valuable model for studying the potential mediators of colonic cell
proliferation, differentiation and apoptosis, which are useful preneoplastic biomarkers of
colon cancer. Evidence to support use of this model includes: i) utilization of AOM
provides a clear distinction between tumor initiation and promotion; ii) carcinogen-
induced K-ras and APC mutations occur as early events in colon carcinogenesis, similar
to humans; iii) the development of tumors is responsive to the amount and type of dietary
fat, and iv) AOM-induced colon tumors closely parallel human colonic neoplasia in
pathologic features (Ahnen, 1985; Reddy et al., 1991; Kakiuchi et al., 1995; Chapkin et
al., 1997; Maltzrnan et al., 1997). In addition, the similarity between the expression of
intracellular modulators of apoptosis (B-catenin, COX-II, bcl-2, bax) in AOM injected
rats and in humans suggest that this experimental system is a good model for studying
mechanisms that regulate apoptosis in humans (Dubois et al., 1996; Hirose et al., 1997;
Takahashi et al., 1998). AOM also induces changes (such as formation of ACF,
adenomas and adenocarcinomas) in the rodent colon that closely replicate, histologically,
the adenoma-carcinoima sequence in humans (Ward et al., 1973). Thus, AOM-induced
rat colonic neoplasms are similar to human colonic tumors in their histologic features and
proliferation characteristics. Given the similarities in gene mutations and gene
expression that exist for rodent and human colon carcinogenesis, this model is a highly

relevant model for preclinical in vivo testing of new chemopreventive agents.

61

F. Conclusion

Colon cancer is a leading contributor to cancer prevalence and mortality in the
Western world (Parkin, 2004), accounting for 15% of all cancers diagnosed annually in
the US. (American Cancer Society, 2005). Identiﬁcation and application of effective
prevention and treatment strategies for this disease represent crucial public health
challenges in industrialized countries. Diet has been recognized as an important
modiﬁable risk factor for CC (AICR/WCRF, 1997). Diet may contain chemopreventive
agents such as soy and calcium. These agents can inhibit the initiation of preneoplastic
lesions by carcinogens or reverse their progression to invasive cancers. Dietary patterns
that emphasize legumes as protein alternatives to meat also may reduce CC risk (Byers et
a1,2002)

Soybeans have been extensively studied and found to be protective against CC in
various model systems. These results are supported by epidemiological studies on Asian
populations and immigrants which demonstrate a protective role of soy consumption
against CC risk. While the protective role of soy has been attributed in part to its high
phytochemical content (e.g. genistein, daidzein, BBI), studies on the efﬁcacy of isolated
soy phytochemicals to protect against CC have yielded mixed results. It is unlikely that
the efﬁcacy of soybeans against CC risk can be attributed to one chemical constituent,
but rather to the synergistic effects of a composite of several chemopreventive
components found in soy.

Despite many epidemiogical studies on the potential of calcium to prevent
colorectal cancer, a protective effect has been difﬁcult to establish. Recent prospective

studies and randomized controlled studies on calcium intake and CC risk support a

62

modest protective effect of about 20-3 0%. Since the inverse relationship between calcium
intake and CC risk depends upon variables such as colon subsite, amount of calcium and
vitamin D exposure, future studies will be needed to examine these variables. There is
signiﬁcant biologic plausibility for a relationship between increased calcium intake and
reduced CC risk and, because calcium is associated with beneﬁts to other organ systems,
calcium supplements should be considered in chemoprevention strategies for CC.

The use of colorectal cancer incidence as an end point in chemoprevention trials
is impractical due to the long time frame of colon carcinogenesis as well as the large
numbers of subjects required to test the efﬁcacy of various interventions. Thus, there is a
need to identify easily-detectable biomarkers of a precancerous state as intermediate
biomarkers that can reliably predict CC risk and possible beneﬁts of intervention in
chemoprevention trials (Einspahr et al., 1997).

Once identiﬁed, potential preneoplastic biomarkers must be painstakingly
validated. This task involves the judicious use of the most suitable animal models which
reﬂect the multifactorial nature of colorectal cancer (Shureiqi et al., 2000). Furthermore,
the relationship between the biomarker and risk of cancer has to be established in
prospective studies in high-risk individuals. Such studies constitute a signiﬁcant
undertaking, but are necessary to bridge the gap between preclinical model systems and
clinical trials.

Among the methods proposed as surrogate end-point biomarkers for the
assessment of CC risk in human populations, the assessment of colon epithelial cell
proliferation and ACF visualization in vivo are very promising. Cell proliferation can be

repeatedly evaluated by obtaining biopsy material routinely taken during sigmoidoscopy.

63

These techniques are both relatively easy and of low cost. ACF visualization in vivo
seems quite promising as an intermediate biomarker of CC risk, but this technique
requires further validation. At the present level of knowledge it appears that these
approaches could be used in conjunction for evaluating CC risk in human populations

and in intervention trials.

64

CHAPTER III. CELL PROLIFERATION INDICES IN COLONIC CRYPTS OF

HUMANS AT VARYING AGES AND RISKS FOR COLON CANCER
A. ABSTRACT

Colon cancer (CC) intervention studies require a sensitive and accurate index to
measure treatment effectiveness. Mucosal cell proliferation before and after treatment is
frequently used to monitor effectiveness. This study was conducted to determine the
correlations between CC risk (subjects varying in age and history of polyps, familial
adenomatous polyps-F AP or CC) and 3 indices of mucosal proliferation: labeling index
(LI), proliferation zone (PZ) and relative proliferating cell nuclear antigen (PCNA)
levels. Subjects (46) were grouped based on age, 20-29 yrs (n=6, mean age of 22.7 i 3.1
years), 30-49 yrs (n=6, mean age of 38.9 i 6.7 years), 50-69 yrs (n=6, mean age of 59.3 :1:
5.9 years), 70-79 yrs (n=7, mean age of 72.1 i 1.9 years) and clinical history, adenomas
(n=8, age range of 44 to 72 years with mean age of 57.9 i 12.4 years), FAP (n=7, age
range of 9 to 46 years with mean age of 23.6 i 13.5 years, 3 males, 4 females) and CC
(n=6, age range of 52 to 74 years with mean age of 65.7 d: 9.0 years). Cell proliferation
in the colonic biopsies was measured by PCNA and Ki-67 immunohistochemistry and
relative PCNA levels by slot-blot quantiﬁcation. The results showed that the PCNA LI,
PCNA PZ, Ki-67 P2 and relative PCNA levels increased with age and history of polyps,
F AP or CC (at P<0.05). There were no signiﬁcant differences in the Ki-67 total LI
among subjects at varying age and history of colonic diseases. This study shows that the
proliferative compartment (as assessed by PCNA and Ki-67) and relative PCNA levels,

but not Ki-67 Ll, were positively correlated with increasing age and CC risk, and

65

therefore useful discriminators to assess CC risk in large scale clinical studies and

chemo-modulation in dietary intervention studies.

B. INTRODUCTION

One of the early events in the multi-stage process of colon carcinogenesis is an
increase in colonic cell proliferation (Vogelstein et al., 1988; Vogelstein and Kinzler,
1993). This is supported by observations that hyper-proliferation occurs in colonic crypts
of subjects at elevated risk for colon cancer (CC) (older age, presence of polyps, family
history of CC, ulcerative colitis, Familial Adenomatous Polyposis) when compared to
young subjects with no history of colon disease (Lipkin, 1984; Lipkin, 1987; Deschner et
al., 1988; Roncucci et al., 1988; Paganelli et al., 1990; Rozen et al., 1990; Paganelli et al.,
1991; Risio et al., 1991; Gerdes et al., 1993; Bostick et al., 1995; Bostick et al., 1997b;
Mathers et al., 1998; Mills etal., 2001; Sinicrope et al., 2004). Indices of cell
proliferation (Labeling Index [LU-fraction of labeled cells in colonic crypt and
Proliferation zone [PZ]-highest labeled cell in colonic crypt) in the colonic mucosa and
the distribution of the proliferating cells in the mucosal crypts are commonly used as
surrogate biomarkers of CC risk in clinical trials. Methods typically used in the past to
assess colonic cellular proliferation have included in viva administration of tritiated
thymidine and BrdU, (to label cells in the S-phase) ex-vivo incubations of colonic
biopsies with BrdU and, more recently, the use of endogenous markers such as
proliferating cell nuclear antigen (PCNA) and Ki-67 antigen. PCNA is a nuclear protein

identiﬁed as an ancillary factor for DNA delta polymerase and detected in all phases of

66

the cell cycle (Bravo and MacDonald-Bravo, 1987). Ki-67 is primarily expressed during
S-phase of the cell cycle and is absent in the G0 phase (Brrmo and Darzynkiewicz, 1992).

In normal colonic mucosa, proliferating cells are restricted to the basal 60% of the
crypts (Lipkin, 1988). Previous research by Lipkin (1971; 1975; 1987) and Deschner
(1975; 1975; 1977; 1985) consistently found increased proliferation indices (both L1 and
expansion in P2) in normal mucosa of subjects with risk factors for CC (family history of
polyposis, previous history of CC and sporadic adenomas) when compared to normal
controls of similar age ranges. Studies using either BrdU or tritiated thymidine to label
proliferating cells have demonstrated hyper-proliferation and expanded proliferative
zones in macroscopically normal rectal mucosa of persons with adenomas (Wilson et al.,
1990; Risio et al., 1991), expanded proliferative zone in persons with a strong family
history of CC (Gerdes et al., 1993), and an increase in L1 (mean mitosis/crypt) in
macroscopically normal colonic mucosa of persons with sporadic adenomas, carcinomas
or FAP when compared to normal subjects (Mills et al., 2001). Deschner et al. (1988)
observed that increased CC risk is more closely related to an expansion in the P2 rather
than to total crypt Ll. Subsequently, several other studies also have indicated that
expansion of the P2, rather than total crypt epithelial cell proliferation, is a better
indicator of increased CC risk (Paganelli et al., 1991; Risio et al., 1991; Gerdes et al.,
1993; Mills et al., 1995; Paspatis et al., 1998; Akedo et al., 2001).

Some researchers have demonstrated inconsistent observations with colonic
epithelial proliferation indices and thus question the relevance of LI and P2 as indicators
of the CC risk (Biasco et al., 1994; McShane et al., 1998). Keku et al. (1998) found

inconsistent association between CC risk, diet and rectal proliferation indices in subjects

67

with adenomas. Sandler et al. (2000) reported that proliferative index did not predict
future colorectal neoplasia and was weakly associated with current history of adenomas
in humans. Differences among the various studies that have examined colonic epithelial
cell proliferation have included variable methods for tissue collection and preparation

(e. g. culture of biopsies for tritiated thymidine and BrdU labeling) and, more importantly,
the use of different endogenous or exogenous proliferation markers (e. g. PCNA, Ki-67,
tritiated thymidine, BrdU, and S-phase fraction by ﬂow cytometry). Each of these
markers detect cells in S-phase and some of these markers (e. g. PCNA, Ki-67) label cells
in other phases of cell cycle. Therefore, the various markers measure different aspects of
cell proliferation cycle and lead to inherently variable observations (Renehan et al.,
2002).

Aging is characterized by an overall increase of epithelial cell proliferation in
colorectal mucosa and by an upward expansion of the proliferative compartment, Similar
to that observed in populations at increased risk for colon cancer. Deschner et al. (1988)
reported that L1 (using tritiated thymidine) in normal mucosa of rectal biopsies of older
persons with no colonic disease (mean age of 55.9 i143, L1 = 7.7%) was signiﬁcantly
greater than that observed in younger persons who had no history of colonic disease
(mean age of 23.3 :1: 1.5, L1 = 5.8%). Roncucci et al. (1988) used tritiated thymidine
labeling to demonstrate an overall increase in rectal epithelial cell proliferation and an
expansion of the proliferative compartment with increasing age of normal mucosa of
subjects who were grouped into categories of 30 to 51, 51 to 65 and 66 to 90 years of age.

deqvist et al. (2002), using DNA-ﬂow cytometry, compared S-phase fraction of colonic

68

epithelial cells of middle-aged persons (mean of 4 H: 8.9 years) and an older age group
(mean of 69 i 8.4 years) and reported that the S-phase fraction increased with age.

To date, there has been no published research which compares PCNA and Ki-67
proliferation indices in persons of widely varying ages who have no history of colonic
diseases. The ﬁrst objective of this study was to use immunohistochemical methods to
assess and directly compare PCNA and Ki-67 proliferation indices in colonic mucosa
biopsies obtained from persons who differed in age and known risk factors for CC
development. Given the laborious nature of IHC techniques used to measure proliferation
markers, there is a need to develop more facile indices of colonic epithelial cell
proliferation. Simpler methods would be highly desirable in dietary intervention and
large scale clinical studies that evaluate the risk for CC. For this reason, the second
objective of this study was to use an immunoblotting method to determine relative levels
of PCNA protein in colonic epithelial cells and determine if these correlate with

proliferation indices based on IHC techniques.

C. MATERIALS AND METHODS
Experimental design
A total of forty-Six subjects (20 males and 26 females) were recruited by the
Division of Gastroenterology at Michigan State University (MSU) clinical center. All
subjects received a remuneration of $ 50 for participation in the study and the study was
approved by the University Committee on Research Involving Human Subjects at MSU.
Subjects were divided into two broad classes a) subjects without a history of

colonic diseases (adenomas, FAP or colon cancer) were considered ‘controls’ and b)

69

subjects with a history of colonic diseases were considered at risk for CC. Control
subjects were referred for endoscopic examination for health maintenance or volunteers
from MSU. They were divided into four age ranges, 20-29 (n=6, mean age of 22.7 d: 3.1
years, 2 males, 4 females), 30-49 (n=6, mean age of 38.9 i 6.7 years, 3 males, 3 females),
50-69 (n=6, mean age of 59.3 i 5.9 years, 4 males, 2 females) and 70-79 (n=7, mean age
of 72.1 i 1.9 years, 3 males, 4 females) with an overall mean age of 49.2 :h 20.0 years.
Subjects (overall mean age of 48.75: 11.6 years) with colonic diseases were divided into
three additional groups on the basis of their history of neoplasia. The groups were,
persons with a history of adenomatous polyps (n=8, age range of 44 to 72 years with
mean age of 57.9 d: 12.4 years, 1 male, 7 females), persons with a history of F AP (n=7,
age range of 9 to 46 years with mean age of 23.6 i 13.5 years, 3 males, 4 females) and
persons with a history of CC (n=6, age range of 52 to 74 years with mean age of 65.7 :t
9.0 years, 4 males, 2 females, three subjects had undergone surgical resections of their
colons in the past). Biopsies were obtained from the same site (30-40 cms from the anal
verge) for all subjects except for two subjects in the FAP group (who had colostomies

and biopsies were obtained from the rectal stump).

Tissue procurement and processing
All subjects received an oral electrolyte bowel preparation, GolytelyTM containing
polyethylene glycol 3350““, NaCl, KCl, NaHCO3, and Na2S04, the evening before
ﬂexible sigmoidoscopy or colonoscopy. Colonic biopsy samples (4) per subject were
Obtained during ﬂexible sigmoidoscopy or colonoscopy. Biopsies approximately 3mm by

8mm were obtained from normal appearing mucosa 30 to 40 cms from the anal verge

70

using jumbo forceps. All biopsies were immediately rinsed in phosphate buffered saline
(pH 7.4). Two samples were placed individually in 0.6 ml microcentriﬁrge tubes and
immediately frozen on dry ice for subsequent slot-blot quantiﬁcation of PCNA. The
remaining samples were pinned ﬂat to cardboard strips and ﬁxed in B5 ﬁxative for 1 hour
and in neutral buffered formalin for 2 — 6 hours. The ﬁxed biopsies were placed between
foam pads in histopath cassettes, dehydrated, inﬁltrated with parafﬁn and embedded in

parafﬁn using standard histologic procedures.

PCNA and Ki-67 IHC analyses

Three micron thick sections of parafﬁn-embedded biopsy samples were cut and
mounted on poly-L-lysine coated glass slides and dried at 58° C for 2 hours. Sections
were cleared of parafﬁn with xylene, hydrated in increasing ratios of waterzethanol, and
treated with iodinezpotassium iodide and 5% sodium thiosulfate to remove mercury.
Samples were then subjected to antigen retrieval (10 mMol/L citrate buffer, pH 6.0, 95
°C) for 20 minutes before immersion in 0.3% hydrogen peroxide for 10 minutes to
quench endogenous peroxidase activity. Following rinses in tris-buffered saline (TBS),
sections were then incubated with primary antibody (1:100 dilution of PC10 [anti-PCNA]
or MMl [anti- Ki—67] respectively in 1% BSA in Tris buffered saline, pH 7.4) overnight
at 4 °C. After subsequent incubations in biotinylated rabbit anti-mouse immunoglobulins
(diluted 1:100, 45 min), followed by peroxidase-conjugated streptavidin (diluted 1:300,
45 min), they were treated with 3-amino-9-ethylcarbazole (ABC) and color development
was monitored for 10—20 minutes until red staining was evident. The sections were

rinsed twice with 10 mMol/L TBS in between incubations with antibodies, streptavidin

71

and ABC. Slides were then counter-stained with hemotoxylin and cover-Slipped using
Faramount mounting medium. Appropriate positive (human tonsil) and negative controls
(omission of primary antibody) were used for evaluation of PCNA and Ki-67 staining.
The primary antibodies were obtained from Novacastra (now Vector Laboratories;
Burlingame, CA), and biotinylated rabbit anti-mouse immunoglobulin, streptavidin,

AEC, and Faramount were obtained from Dako (Carpinteria, CA).

Quantiﬁcation of PCNA and Ki—67 IHC:

PCNA and Ki-67 labeling were quantiﬁedby the percentage and positions of
positively stained (red) cells within full-length crypts in colonic sections. Well-oriented
longitudinally-sectioned full-length crypts (n = 10 per subject) were identiﬁed by light
microscopy. A scorable crypt was deﬁned as one in which the base touched the
muscularis mucosa and had an open lumen at the top. The number and position of PCNA
and Ki-67 labeled cells and crypt heights were recorded for ten full crypts for each
subject to determine cell proliferation parameters. All cells with a red nuclear staining
were counted as positive. The center-most cell at the base of the crypt, and in between the
hemicrypts, was designated as cell number one and the cell at the top of the crypt mouth
was considered the highest cell. A continuous column of cells from cell number one to
the highest cell were counted along each hemicrypt wall of each crypt. Crypt height was
deﬁned as the number of cells per hemicrypt.

Total Crypt Labeling Index (LI) was calculated by dividing the number of labeled
cells in a hemicrypt by the total number of cells in a hemicrypt (crypt height). These

values were averaged for at least 20 hemicrypts (representing 10 full-length crypts) to

72

obtain the L1 for each subject. Compartment Labeling Indexes (basal LI, mid LI and top
LI) were calculated by dividing the number of labeled cells within each third (basal,
middle and top) of each hemicrypt by the number of cells in each third of the respective
hemicrypt. Proliferation zone (PZ) was obtained by dividing the position of the highest
labeled cell from the hemicrypt base by the total number of cells in the hemicrypt. Each
of these PZ values were averaged for at least 20 hemicrypts to obtain the overall PZ for

each subject.

PCNA Measurement by Slot-blot

Colonic biopsy samples were homogenized with 0.5m] of non-denaturing buffer
containing protease inhibitors (20mMol/L MOPS, 150 mMol/L sodium chloride, 1%
vol/vol Triton X-100, 1% wt/vol deoxycholate, 0.1% wt/vol SDS, 1 mMol/L EDTA, 100
uMol/L sodium orthovandate, 250 pMol/L PMSF, 10 ug/ml leupeptin, 1 jig/ml pepstatin)
and total protein concentrations were determined by the biocinchoninic assay using BSA
as the standard. Relative PCNA levels in tissue homogenates were then determined by
immuno-quantiﬁcation using slot-blot methodology. PVDF membranes (Immobilon P
membranes; Millipore, Billerica, MA) were wetted in methanol and then transferred to
PBS. Twenty pg of protein from each sample (dilutions in PBS) was vacuum blotted on
to the PVDF membranes using a slot-blot apparatus (Hoefer Scientiﬁc instruments). The
membranes were held overnight at 4 °C in blocking buffer (3% BSA in TBS with Tween
20). The membranes were then incubated for 2 hours with monoclonal PC-lO antibody
(1 : 1000 dilution of primary antibody in blocking buffer). After rinsing with wash buffer

(TBS with Tween 20) the membranes were incubated in biotinylated secondary antibody

73

(1 : 1000 dilution of goat anti-mouse IgG in wash buffer for one hour). After subsequent
rinses with wash buffer, the membranes were treated with a chemiluminescence reagent
(Dupont NEN research products) and light emission was captured by autoradiography.
Developed autoradiography ﬁlms were quantiﬁed by scanning densitometry. Known
amounts of mouse IgG Fc fragments standard curve ranging from (0.3125ng to 10.00ng)

were included in each blot to allow for normalization of staining intensity across blots.

Statistical Analysis
Crypt height, PCNA and Ki-67 proliferation indices (total crypt LI, PZ, and

compartmental LI for basal, middle, and top compartments of the crypts) and relative
PCNA quantiﬁed by slot-blot method were analde as a completely randomized design
(one-way Analysis of variance) according to the General Linear Models procedure of
SAS (version 6.11). When signiﬁcant group effects were detected (P < 0.05), the least
signiﬁcant difference was used to compare group means. Simple correlation coefﬁcients
between PCNA and Ki-67 labeling indices and the relative PCNA levels and were

calculated and tested for signiﬁcance at P < 0.05 using the correlation procedure of SAS.

D. RESULTS
Proliferation indices with IHC:
Total Crypt Labeling Index (LI)
Total crypt PCNA LI (Table 3) was signiﬁcantly inﬂuenced by age and risk for
CC. Among subjects with no history of colonic disease, total crypt PCNA L1 in the two

oldest age groups (50-69 years and 70-79 years, 29% and 30% of the total cells in the

74

 

crypt had PCNA label, respectively) was signiﬁcantly greater than that observed in
subjects in the younger age groups (20-29 years and 30-49 years, 20% and 23% of the
total cells in the crypt had PCNA label, respectively). Among control subjects with no
history of colonic disease, there was a linear increase in PCNA total LI with increasing
age (individual values, r = 0.81, P<0.0001, data not shown).

The total crypt PCNA LI of subjects in any of the at-risk categories were
signiﬁcantly greater than that observed for all control subjects regardless of age. The total
crypt PCNA L1 in subjects with F AP (40% of the total cells in the crypt had PCNA label)
was signiﬁcantly greater than that observed in any other groups.

Total crypt Ki-67 LI (Table 3) was not inﬂuenced by age or risk category and

averaged 0.080 across all groups.

Proliferation zone

PCNA proliferation zone (Table 3) was signiﬁcantly inﬂuenced by age and risk
for CC. Among control subjects with no history of disease, PCNA proliferation zone was
signiﬁcantly greater in older subjects (50-69 years [PZ=49%] and 70-79 years
[PZ=56%]) when compared to subjects in the younger age groups (20-29 years
[PZ=3 7%] and 30-49 years [PZ=40%]). There was a linear increase in PCNA
proliferation zone with increasing age (r = 0.86, P<0.0001, data not shown).

PCNA proliferation zone was statistically similar for older subjects (50-69 and
70-79 years) with no history of colon disease and subjects having a history of

adenomatous polyps (PZ=55%) or colon cancer (PZ=5 5%). Among all age and risk

75

categories, the largest PCNA proliferation zone was observed in subjects with F AP
(PZ=63%).

Ki-67 proliferation zone (Table 3) was Signiﬁcantly inﬂuenced by age and colonic
disease. Control subjects in the oldest age group (70-79 years) had a greater Ki-67
proliferation zone (PZ=41%) when compared to control subjects aged 20-29 and 30-49
years (P2 = 28% and 31%, respectively). Among control subjects, there was a linear
increase in Ki-67 proliferation zone with increasing age (individual values, r = 0.63,
P<0.0001, data not shown).

Control subjects in the two oldest age groups and subjects in all disease groups
had statistically similar Ki-67 proliferation zones (PZ=35%-41%) and proliferation zones
in each of these groups were signiﬁcantly greater than that observed in the youngest (20-

29) normal subjects (PZ=28%).

Compartmental Labeling Indices (basal LI, mid LI, top L1)

Given the observed changes in total crypt PCNA LI and P2, we were interested in
determining if PCNA labeling patterns in speciﬁc regions of the crypts were inﬂuenced
by age and colonic disease. PCNA labeling was observed primarily in the basal
compartment, with a smaller amount detected in the middle of the crypts and very little in
the top compartment (Table 4). The basal compartment LI was signiﬁcantly inﬂuenced
by age and risk category, but the overall pattern differed somewhat from that observed
for total crypt PCNA LI (Table 3). Among the control subjects, those in the 50-69 year
old age category had the greatest basal L1 (58% of the cells in the basal compartment had

PCNA label) when compared to subjects in the rest of the age groups. We observed a

76

quadratic effect of age on PCNA basal Ll (P < 0.05). This quadratic effect was due to
middle-aged subjects (50—69 years) having greater PCNA basal LI than either the
youngest or oldest subject. Subjects with a history of CC (5 8%) and F AP (57%) had
signiﬁcantly greater basal Ll values compared to subjects with a history of AP (44%).

Middle compartment LI was signiﬁcantly inﬂuenced by age and risk for CC.
Control subjects in the older age groups (50-69 years and 70-79 years) had signiﬁcantly
greater middle compartment L1 (29% and 42% of the cells in the middle compartment
had PCNA label, respectively) when compared to the younger age groups (age 20-29 and
30-49 groups had 13% and 18% of the total cells in the middle compartment with PCNA
label, respectively). Among control subjects who had no history of colonic disease, there
was a linear increase in the PCNA middle LI with increasing age (r = 0.88, P< 0.0001,
data not shown). All subjects with a history of colonic disease had signiﬁcantly greater
middle LI than subjects with no history of colonic diseases.

Top compartment Ll was signiﬁcantly greater in subjects with F AP when
compared to all other groups. Top compartment LI was greater than 6% in subjects with
FAP, but was less than 1% for subjects in all other age and risk categories. Among
controls, there was a linear increase in PCNA top LI with increasing age (P < 0.05).

Although total crypt Ki-67 LI was not inﬂuenced by age or colonic disease, we
were interested in the overall patterns of Ki-67 protein expression among these groups.
Ki-67 labeling was observed primarily in the basal compartment with some labeling in
the middle compartment and very little in the top compartment (Table 5). Ki-67 basal
compartment labeling index was signiﬁcantly inﬂuenced by age and risk for CC. Control

subjects in the younger age groups (20-29 years and 30-49 years) had signiﬁcantly

77

greater basal L1 (21% and 23% of the total cells in the basal compartment had Ki-67
label) when compared to the subjects in the oldest age group (70-79 years [16%] ).
Among control subjects, there was a linear decrease in Ki-67 basal LI with increasing age
(P < 0.05).

When comparing all age and disease groups, there were no overall Signiﬁcant
differences detected in Ki-67 L1 in the middle and top compartments of the colonic
crypts. However, we did observe a linear increase in Ki-67 middle compartment LI with
increasing age (P<0.0001) when only control subjects with no history of colonic disease

were compared.

Crypt height
Crypt height (Table 3) was signiﬁcantly inﬂuenced by age and risk category.
Control subjects in the oldest (70-79 years) and youngest (20-29 years) age categories
had signiﬁcantly taller crypts (average of 60 and 57 cells per hemicrypt, respectively)
than subjects in the intermediate age groups. Subjects with a history of AP had
signiﬁcantly taller (65 cells per hemicrypt) colonic crypts than all subjects with no

history of colonic diseases.

Relative PCNA levels
PCNA protein levels quantiﬁed by the slot-blot method were signiﬁcantly
affected by age and risk for CC (Table 6). Among control subjects, PCNA content
increased with increasing age. Among control subjects having no history of colonic

disease, elderly subjects (70-79 years of age) had approximately 2.5 times as much

78

PCNA protein per 20ug mucosal protein as the youngest subjects (20-29 years of age).
The middle aged groups (30-49 years and 50-69 years) had intermediate PCNA levels.

Subjects with a history of adenomatous polyps, F AP and CC had approximately
twice as much PCNA protein per 20ug mucosal protein as the oldest subjects with no
history of colonic disease (Table 6). Among the risk categories (subjects with AP, FAP,
and CC), subjects with a history of CC had the greatest expression of PCNA in the

colonic mucosa.

Correlation among biomarkers (IHC indices of Ki-6 7 and PCNA and Relative PCNA
levels)

Ki-67 total crypt LI was positively correlated with Ki-67 PZ (r = 0.47, P =
0.0009), but was not Signiﬁcantly correlated to any indices of PCNA expression (Table
7). Ki-67 P2 was signiﬁcantly correlated with PCNA total crypt LI, PCNA middle crypt
LI and PCNA PZ, and tended to be correlated with quantiﬁed PCNA. PCNA total crypt
LI, PCNA PZ, PCNA middle crypt LI and relative PCNA protein quantiﬁed by slot-blot
analysis all were highly correlated with each other (individual values, r = 0.56 — 0.94,

P<0.0001).

E. DISCUSSION

In this study we assessed and compared the proliferative indices of colonic cells
in normal colonic mucosa of subjects at varying age and risk for CC with the use of
PCNA and Ki-67 using traditional immunohistochemistry methods. To the best of our

knowledge, no one has reported PCNA and Ki-67 expression in normal colonic mucosa

79

of subjects at varying ages. An expansion of the proliferative zone as measured by both
PCNA and Ki-67 was observed with increasing age in normal subjects, and also was
observed in subjects, regardless of age, who had a history of colonic diseases
(adenomatous polyps, FAP and CC). We also demonstrated that relative PCNA levels in
colonic epithelium biopsies determined by a simple immunoblot method is a useful
indicator of mucosal proliferative activity in the colon. Quantiﬁed PCNA levels, PCNA
total LI, PCNA P2 and PCNA middle compartment LI of all subjects correlated with
each other and showed a linear increase with increasing age among subjects having no
history of colonic disease.

We observed that both total crypt PCNA LI and P2 increased with age in control
subjects and also increased considerably in subjects who had a history of colonic disease
(Table 3). Since there is a high incidence of CC in aged population, the relationship
between changes in colonic cell proliferation and aging has been the focus of other
studies. Researchers have reported that aging is characterized by an overall increase of
epithelial cell proliferation in colorectal mucosa and by an upward expansion of the
proliferative compartment, similar to that observed in a population at risk for colon
cancer. The contribution of age as a risk factor for CC as measured by proliferation
indices was ﬁrst examined by Deschner et al., in 1988. They report a signiﬁcant increase
(1.3 times) in the L1 (using tritiated thymidine) in normal mucosa of rectal biopsies of
older control groups (age range of 33-78 years, mean age of 55.9 $14.3) when compared
to younger control groups (age range of 22-27, mean age of 23.3 i 1.5). Increases in cell
proliferation in colorectal mucosa of senescent rats have also been shown by other

techniques in animals (Holt and Yeh, 1988; Lee et al., 2000; Xiao et al., 2000; Lee et al.,

80

2001). Roncucci et a1. (1988) used tritiated thymidine autoradiography to demonstrate an
overall increase in rectal epithelial cell proliferation and an expansion of the proliferative
compartment with age in subjects in three age groups of 30 to 51, 51 to 65 and 66 to 90
years. In their study they reported signiﬁcant differences in total labeling index and top
compartment labeling index in normal rectal mucosa of people of 66 years and higher
(12.9% and 0.04%) when compared to the youngest age group of 30 to 50 years of age
(9.5 % and 0.01%). More recently, Sjbqvist et al. (2002) using DNA-ﬂow cytometry
methods demonstrated that the S-phase fraction of colonic epithelial cells (biopsies
obtained from right, mid and left colon) increased linearly with age of normal subjects
(age range of 21-80). They reported that the younger age group (mean of 4lzt 8.9 years)
had a statistically lower S-phase fraction of 2.37 :t: 1.38% when compared to older age
group (mean of 69 :t 8.4 years) with S-phase fraction of 2.91 i 1.66%. In our study, total
crypt PCNA LI was affected by age with the highest indices (29% - 30%) observed in the
older age groups (age range of 50-79 years, mean age of 65.8 :1: 3.9 years). The total crypt
PCNA LI was 1.4 times greater than that observed in subjects in the younger age groups
(age range of 20-49 years, mean age range of 30.8 i 4.9 years). The differences between
these age groups reported in our study with PCNA is comparable to the increases (1.3)
reported by Deschner et al. (1988) and Roncucci et al. (1988) using another proliferation
marker, tritiated thymidine in similar age range of control subjects. Several studies have
also demonstrated that PCNA correlates well with other proliferation markers (e. g. BrdU)
in both normal and neoplastic colonic epithelium in humans (Kubota and Kino, 1995;
Tanaka et al., 1995; Yoshikawa and Utsunomiya, 1996; Shpitz et al., 1997; McShane et

al., 1998; Yerly-Motta et al., 1999; Wu et al., 2000; Diez et al., 2003).

81

Many studies have used IHC methods to demonstrate increased PCNA labeling
indices in neoplastic tissues of adenomatous polyps and colon tumors (Yoshikawa and
Utsunomiya, 1996; Bostick et al., 1997a; Shpitz et al., 1997; Ottaviani et al., 1999; Saito
and Mori, 1999; Wu et al., 2000; Diez et al., 2003). In particular, PCNA expression (as
measured by immunohistochemistry) has been used in human studies to demonstrate a
step-wise increase in proliferation indices during neoplastic progression of colonic
lesions (Shpitz et al., 1997), for predicting the potential for metastases in invasive
colorectal carcinoma (Tanaka et al., 1995; Choi et al., 1997), and to predict the
development of metachronous adenomas (Ottaviani et al., 1999). Studies also have
demonstrated increased PCNA labeling indices along with increases in growth factors
(such as EGF receptor kinase and TGF-a) (Malecka-Panas et al., 1997; Barnes et al.,
1998) and overexpression of cyclins-(such as Cyclin D) (Izawa et al., 2002) in colonic
tissue. In our study the overall total crypt PCNA LI also was signiﬁcantly inﬂuenced by
disease conditions with the greatest indices (39%) observed in normal appearing mucosa
of subjects with FAP when compared to normal mucosa of subjects with a history of CC
(36%) and subjects with a history of adenomatous polyps (34%) (Table 3).The mean total
crypt PCNA LI measured in normal colonic mucosa of our study subjects with colonic
diseases (mean age of 48.73: 11.6 years) were statistically higher than the measurements
observed in subjects in the oldest age group (age range of 70-79 years, mean age of 72.1
:1: 1.9 years). Moreover, in the FAP group, the subjects had an age range of 9 to 46 years
with mean age of 23.6 a: 13.5 years. All subjects with FAP had statistically comparable
total crypt Ll with a mean of 40 % i 1%. These ﬁndings clearly demonstrate the effect of

neoplasia on colonic proliferation in normal mucosa regardless of age in these subjects.

82

Our PCNA IHC results also were consistent with ﬁndings from other studies (Bromely et
al., 1996; Bostick et al., 1997b; Malecka-Panas et al., 1997) on colonic proliferation in
normal-appearing colonic mucosa of subjects with adenomas, FAP and CC when
compared to controls in similar age range. Our observed increases in total crypt PCNA LI
also suggest there is a delay in epithelial cell differentiation with increasing age and risk
for CC

Normally, proliferation in colonic crypts is restricted to the basal 60% of the
crypts. As colonocytes migrate up the crypt, they undergo terminal differentiation.
Ultimately, after approximately 7 days of transit time to the crypt surface, colonocytes
undergo programmed cell death and are extruded. Under normal circumstances, cell
proliferation does not occur in the upper 40% of the crypts (Lipkin, 1971). An expansion
of the proliferative compartment toward the crypt surface (hyperproliferation in the
middle and upper thirds of the crypt) is considered a preneoplastic biomarker (Lipkin,
1988) and has been associated with increased risk for development of adenomatous
polyps and carcinoma (Risio et al., 1991). Expansion of the proliferation zone has been
shown in a variety of neoplastic conditions, most notably in F AP and in colon
tumorigenesis (Lipkin, 1984; Lipkin et al., 1984; Rozen et al., 1991; Rozen, 1992;
Richter et al., 1993; Rooney et al., 1993; Mills et al., 1995; Strater et al., 1995). In
general, expansion of the proliferative compartment indicates that normal mechanisms
controlling colonic epithelial cell proliferation and differentiation are compromised. In
our study an expansion of the proliferative zone as measured by both PCNA and Ki-67
was observed with increasing age in normal subjects with the greatest expansion (56 %

and 41%, respectively) observed in the oldest subjects (mean age 72.1i 1.9 years). This is

83

approximately an increase of 1.5 times in proliferative zone (with both PCNA and Ki-67)
when compared to subjects in the youngest age group (20-29 years with a mean age of
22.7 :1: 3.0 years). The mid and top compartmental indices of PCNA also signiﬁcantly
increased with age (Table 4). The signiﬁcant differences we observed in the PCNA
middle compartment Ll among the groups further conﬁrm that an upward shift in the
proliferative compartment occurred with increasing age and risk for CC. The pattern of
changes we observed in Ki-67 PZ among the groups was generally similar to that
observed for PCNA. However, among older normal subjects or those With AP, CC, or
F AP, PCNA PZ expanded to a greater extent than that of Ki-67. This observation
strongly suggests that PCNA PZ increased both as a consequence of expanded zone of
cell proliferation (indicated by expanded Ki-67 P2) and due to delayed terminal
differentiation of colonic epithelial cells with increasing age and risk for CC. Therefore,
it is important to have age matched subjects in studies of colon cell proliferation. Future
experiments should also more careﬁilly examine mechanisms controlling colonic
epithelial cell differentiation and factors that cause delayed differentiation in older
persons and those who have other risk factors associated with colon cancer development.
The total crypt Ki-67 LI did not Show any age or disease effects, however the mid
compartment L1 as measured by Ki-67 was signiﬁcantly affected by age with the highest
(7%) observed in the oldest age subjects (Table 5). An expansion of the zone of Ki-67
labeling was observed in older normal subjects and those with disease conditions. Our
Ki-67 labeling results are consistent with those observed by other groups. While some
researchers have shown that Ki-67 LI can be used as a valuable prognostic indicator in

cancers including CC (Gerdes et al., 1991; Tungekar et al., 1991; Wintzer et al., 1991;

84

Suzuki et al., 1992; Kang et al., 1997; Akedo et al., 2001; Garrity et al., 2004), others
(Sahin et al., 1994; Green et al., 1998; Paspatis et al., 1998) have demonstrated no
differences in Ki-67 total labeling indices when comparing normal subjects with persons
with CC or high risk for developing CC. However Paspatis et a1. (1998) and Mills et al.
(1995) reported an expansion of proliferative compartment in subjects with neoplasms
and in subjects with F AP respectively despite no differences in the overall total labeling
indices with Ki-67 which is also consistent with ﬁndings reported in our study. Ki-67 is a
cell proliferation marker that is primarily expressed during the S-phase of cell cycle and
absent in Go while PCNA accumulates as the cells progress through G, S, and M phases
of the cell cycle and returns to low levels during Go. Terminally differentiated cells do
not express PCNA. While Ki-67 labeling provides an estimate of total cell proliferation in
tissues (Bruno and Darzynkiewicz, 1992; Zacchetti et al., 2003), unlike PCNA, it does
not necessarily reﬂect a cell’s capacity to undergo another round of cell division (Prelich
et al., 1987; Toschi and Bravo, 1988; Zacchetti et al., 2003). Other studies comparing
expression of both PCNA and Ki-67 have demonstrated consistently higher PCNA
indices, especially in colorectal and breast tumors (McCormick et al., 1993; Gasparini et
al., 1994; Silvestrini, 1994; Yerly-Motta et al., 1999; Le Pessot et al., 2001). The
probable reason for this discordance is variability in expression of both antigens during
the cell cycle, with PCNA expression being more persistent (half-life of 20 hrs, Bravo
and MacDonald-Bravo, 1987; Toschi and Bravo, 1988; Scott et al., 1991) throughout the
cell cycle than that of Ki-67 (half-life 1-1.5 hours, Du Manoir et al., 1991; Gerdes et al.,

1991; Scott et al., 1991; Bruno and Darzynkiewicz, 1992).

85

The cell concentration of PCNA is directly correlated with the proliferative state
of the cell (Prelich et al., 1987). The presence of PCNA in G0 -phase cells would indicate
that these cells maintain the Capacity to go through another round of cell division.
Conversely, the absence of PCNA is indicative that a cell has undergone terminal
differentiation (Bromely et al., 1996). Hence, loss of PCNA expression appears to be
potentially useful as a differentiation marker of colonic epithelial cells. Furthermore,
detection and quantitation of PCNA content can represent proliferative status in the
colonic crypts.

In this study we also demonstrated that quantitation of PCNA content in colonic
epithelium biopsies by a relatively simple immunoblotting method was a useful indicator
of mucosal proliferative activity in the colon. This approach was less laborious and
tedious than the immunohistochemical methods used to measure cell cycle-associated
antigens as indicators of mucosal proliferative activity in the colon. To our knowledge,
no other groups have used quantitative analysis of PCNA protein in normal mucosa of
subjects to measure colonic mucosal proliferation. Izawa et al. (2002) used Western
blotting to measure Cyclin D1 and PCNA expression in colonic polyps of humans.
However, they did not compare expression of these proteins in colonic polyps with
expression of the same proteins in normal-appearing mucosa in their subjects, nor did
they compare their results with subjects who had no history of colonic disease. The slot-
blot method used in our study to assess PCNA has an advantage versus western blotting
methods of being able to screen dozens of samples simultaneously. A disadvantage of the
Slot-blot approach versus immunohistochemical methods is that information on location

of PCNA expression is not measurable with this approach. However, we found that

86

PCNA protein slot-blotting results were highly correlated with PCNA proliferative zone
as well as total crypt and middle compartment labeling indices. We believe that this
approach would be particularly well-suited to the analysis of colonic biopsies from large-
scale clinical trials for studying diet or drug effects on colon cancer risk.

As illustrated in Table 6, relative PCNA levels signiﬁcantly increased with age of
normal subjects and also were dramatically increased in subjects with colonic disease
(history of adenomatous polyps, FAP and CC). Among the control subjects, we observed
that elderly subjects (age range of 70-79 years, mean age of 72.1 i: 1.9 years) had
approximately 2.5 times as much relative PCNA levels in the colonic mucosa as the
youngest subjects (age range of 20-29 years, mean age of 22.7 i 3.0 years). We also
found elevated expression of total PCNA protein in colonic mucosa of subjects with
adenomatous polyps (mean age of 57.8 i 12.4 years), familial adenomatous polyposis
(mean age of 23.6 i 13.4 years) and colon cancer (mean age of 65.7 :1: 9.0 years).
Regardless of age, the greatest relative PCNA protein levels were observed in subjects
with a history of CC, with a 1.4 fold increase when compared to subjects with a history of
F AP and with a 1.6 fold increase when compared to subjects with adenomatous polyps.
These results indicate that the relative PCNA levels in human colonic mucosa accurately
reﬂect proliferative activity in colonic epithelial cells as evidenced by the PCNA
proliferation indices. In addition, relative PCNA levels, PCNA total labeling index,
PCNA proliferation zone and PCNA middle compartment labeling index of all subjects
correlated with each other and showed a linear increase with increasing age among
subjects having no history of colonic disease (Table 7). Furthermore, elevated expression

of total PCNA protein in colonic mucosa of subjects with adenomatous polyps, familial

87

adenomatous polyposis and colon cancer, as well as in older subjects who had no history
of colonic disease, indicates that the relative PCNA protein levels can be a useful
indicator of colon cancer risk. Taken together, the differences in relative PCNA levels as
well as with PCNA proliferation indices reported in our study seems to be consistent with
proliferation patterns observed in studies using other proliferation markers in colonic

mucosa of people at risk for CC.

F. CONCLUSION

There is a need for development of colon cancer intermediate end points or
precursors that can be measured easily, that would require relatively small samples and
which could be used in short intervention trials. This would be an important advance in
understanding colon cancer etiology and assessing interventions to reduce colon cancer
risk. In our study, the relative levels of PCNA in endoscopically normal appearing colon
mucosa were affected by age, history of adenomatous polyps, personal history of CC and
FAP. Although there were no signiﬁcant differences in Ki-67 total LI, PCNA total LI,
P2, and Ki-67 PZ differed signiﬁcantly with age and risk for CC. These correlate with
the differences observed in the relative PCNA levels among the groups. The slot-blot
technique and subsequent use of densitometry to quantitate PCNA levels in colonic
biopsy samples was easy and reproducible. The relative PCNA levels in conjunction with
proliferative compartment analyses could serve to identify groups of high risk population
for CC. In addition this could be used as a sensitive index to study dietary modulations

on colon carcinogenesis.

88

Table 3. Crypt heights1 and Proliferation indices — PCNA and Ki-67 LI2 & P23 in colon
of subjects at varying age and risk for colon cancer

 

Proliferation Indices‘

 

 

 

Categoriess PCNA LI Ki-67 Ll PCNA PZ Ki-67 PZ CH
20-29yrs (6) 0.203“ :0.013 0.081 :0.006 0.375“ :0.024 0.276“ :0.026 57"“ :2
30-49yrs (6) 0.233“ :0.013 0.089 :0.006 0.406“ :0.024 0.309“b :0.026 52“ :2
50-69yrs (6) 0.292b :0.013 0.083 :0.006 0.492b :0.024 0353““ :0.026 53“b :1
70-79yrs (7) 0.301b :0.012 0.079 :0.006 0.562“ :0.022 0.411“ :0.024 60“d :1
AP (8) 0.344“ :0.011 0.075 :0.005 0553““ :0.021 0.379“ :0.022 65“ :1
FAP (7) 0.398d :0.012 0.078 :0.006 0.633d :0.022 0.380“ :0.024 63““ :1
CC (6) 0.360c :0.013 0.081 :0.006 0.548bc :0.024 0357““ :0.026 63““ :2

 

1 CH — Crypt Height, total number of cells in the hemicrypt from the base to the mouth of

the crypt

2 LI - Total Crypt Labeling Index, obtained by dividing the number of labeled cells in a
hemicrypt by crypt height.
3 Pl — Proliferation Zone, obtained by dividing the position of the highest labeled cell
from the hemicrypt base by the crypt height

4 Values are presented as means i SEM. Means in the same column not Sharing a

common superscript are different (P<0.05).
5 Age groups are subjects who have had no history of adenomas or colon cancer or FAP
AP — People with a history of adenomatous polyps

F AP - People with a history of familial adenomatous polyposis

CC - People with a history of colon cancer

89

Table 4. PCNA compartmental labeling indices (for basal, middle and top crypt
compartments) in colon of subjects at varying age and risk for colon cancer

 

3

PCNA Indices”

 

 

 

Categories Basal Mid Top
L1 L1 L1

20-29yrs (6) 0.481“b :0.025 0.129“ :0.026 0.001“ :0.013
30-49yrs (6) 0507““ :0.025 0.182“ :0.026 0.000“ :0.013
50-69yrs (6) 0.578“ :0.025 0.294b :0.026 0.000“ :0.013
70-79yrs (7) 0.432“ :0.023 0.416“ :0.024 0.009“ :0.012
AP (8) 0.442“b :0.021 0.530“ :0.023 0.000“ :0.011
FAP (7) 0574““ :0.023 0.554“ :0.024 0.066b :0.012
CC (6) 0.582“ :0.025 0.492“ :0.026 0.006“ :0.013

 

I Values are presented as means i SEM. Means in the same column not sharing a
common superscript are different (P<0.05).

2 Basal LI— Labeling Index in Basal one-third compartment of the crypt;
Mid LI — Labeling Index in Middle one-third compartment of the crypt;
Top LI — Labeling Index in Top one-third compartment of the crypt;

3 Age groups are subjects who have had no history of adenomas or colon cancer or FAP
AP — People with a history of adenomatous polyps
F AP - People with a history of familial adenomatous polyposis
CC - People with a history of colon cancer

90

Table 5. Ki-67 compartmental labeling indices (for basal, middle and top crypt
compartments) in colon of subjects at varying age and risk for colon cancer.

 

Ki-67 Indices"2

 

 

 

Categories3 Basal Mid Top
L1 L1 L1

20-29yrs (6) 0.214“ :0.012 0.027 i001] 0.000 :t0.000
30-49yrs (6) 0.228c :0.012 0.039 i001 1 0.000 i0.000
50-69yrs (6) 0.188ab “:0.012 0.061 :0.011 0.000 $0.000
70-79yrs (7) 0.1618 i0.011 0.073 :0.011 0.002 i0.000
AP (8) 0.161al i0.010 0.064 :0.010 0.000 :0.000
F AP (7) 0.1663 i001] 0.068 i0.010 0.000 :0.000
CC (6) 0.183“ :0.012 0.059 :0.011 0.000 :0.000

 

' Values are presented as means i- SEM. Means in the same column not sharing a
common superscript are different (P<0.05).

2 Basal LI— Labeling Index in Basal one-third compartment of the crypt;
Mid LI — Labeling Index in Middle one-third compartment of the crypt;

Top LI — Labeling Index in Top one-third compartment of the crypt;

3 Age groups are subjects who have had no history of adenomas or colon cancer or F AP
AP — People with a history of adenomatous polyps

FAP - People with a history of familial adenomatous polyposis

CC - People with a history of colon cancer

91

Table 6. Relative PCNA levels in colonic mucosa of subjects at varying age and risk for
colon cancer

 

 

CategoriesI N Relative PCNA2
l: 20-29yrs 6 3042’ i 59
2: 30-49yrs 6 614““ : 59
3: 50-69yrs 6 550b : 59
4: 70-79yrs 7 763c i 55
5: AP 8 1251d : 51
6: FAP 7 1432“ : 55
7: cc 6 2047f : 59

 

' Age groups are subjects who have had no history of adenomas or colon cancer or FAP
AP — People with a history of adenomatous polyps
FAP - People with a history of familial adenomatous polyposis
CC - People with a history of colon cancer

2 Values (these are relative values based on densitometry per 20 ug mucosal protein) are
presented as means i SEM. Means in the same column not sharing a common
superscript are different (P<0.05).

92

Table 7. Pearson correlation coefﬁcients for Ki-67 & PCNA LI (total crypt LI) and PZ,
PCNA middle L1 and relative PCNA levels in colonic mucosa of subjects at varying age

and risk for colon cancer.

 

 

Indices Ki-67 LI Ki-67 PZ PCNA L1 PCNA PZ Relative
PCNA

Ki-67

LI x

Ki-67 0.47

PZ P = 0.0009 x

PCNA - 0.12 0.42

LI NS' P = 0.0041 x

PCNA - 0.10 0.49 0.90

PZ Ns‘ P = 0.0006 P < 0.0001 x

Relative - 0.09 0.25 0.73 0.56

PCNA Ns' P = 0.0942 P < 0.0001 P < 0.0001 x

PCNA

Middle - 0.16 0.48 0.94 0.90 0.72

LI NS' P = 0.0007 P < 0.0001 P < 0.0001 P < 0.0001

 

1 Not signiﬁcant

93

CHAPTER IV. BIOMARKERS FOR COLON CANCER RISK BEFORE AND
AFTER CONSUMING SOY OR CASEIN SUPPLEMENTS FOR ONE YEAR
A. ABSTRACT
Increased cell proliferation is a common feature of the promotion stage of colon
carcinogenesis. Mucosal cell proliferation before and after dietary treatment or other
interventions is frequently used to monitor treatment effectiveness. A double-blind,
prospective study consisting of forty two subjects (34-70 yrs of age having a history of
adenomatous polyps or colon cancer) was designed to see if consumption of one of two
supplements containing 38 g/d of soy protein isolate (containing 58 mg total genistein) or
supplement containing 40 g/d of calcium caseinate for one year would inﬂuence colonic
epithelial cell proliferation as measured by PCNA and Ki-67 labeling in and PCNA
protein expression in colonic epithelial cells. The two soy supplements differed only in
their calcium contents. Four colonic biopsies from each subject were obtained before and
after supplementation and used for immunohistochemical detection of PCNA and Ki-67
antigens in colon crypts and determination of relative PCNA protein expression by slot-
blotting. There was a statistically signiﬁcant reduction in both PCNA labeling index (Ll)
(beginning versus ending values - soy: 30% and 26% P < 0.05; casein: 28% no change)
and PCNA proliferation zone (PZ) (beginning versus ending values - soy: 53% and 45%
P < 0.05; casein: 49% and 48%) during the course of the experiment by subjects that
consumed soy supplement but not in subjects who consumed casein. There were no
signiﬁcant treatment effects on the total Ki-67 LI (9%) and Ki-67 P2 (46%). The relative
expression of PCNA protein in colonic mucosa was signiﬁcantly (P < 0.05) reduced by

consumption of either of the soy supplements, but was not inﬂuenced by consumption of

94

the casein supplement. Since the presence of PCNA indicates the cells capacity to
undergo cell division, the decrease in PCNA LI and PZ denotes a lowering in the
proliferating capacity of colonic crypts in those subjects that consumed the soy
supplement. This observation indicates that consumption of soy protein isolate
supplements caused changes in colonic epithelial cell proliferation patterns that are

consistent with a reduced risk for developing colon cancer.

B. INTRODUCTION

Colorectal cancer (CC) is the third most common cancer and second most
common cause of mortality in the US. (American Cancer Society, 2005). People with
sporadic adenomatous polyps are known to have signiﬁcantly greater risk for developing
CC compared to the general population (Luebeck and Moolgavkar, 2002). The presence
of proliferating epithelial cells in the upper 40% of colon crypts is considered a good
criteria for assessing high risk for CC (Lipkin, 1987). An upward expansion of the zone
of cell proliferation in colonic epithelium occurs in individuals with sporadic adenomas
of the large intestine (Deschner, 1985; Lipkin, 1987; Wilson et al., 1990; Risio et al.,
1991). Therefore, patients with Sporadic colon adenomas represent a suitable population
for studying if consumption of certain foods will cause a downward shift in the
proliferating compartment of colon crypts and thus, reduce risk of CC.

The role of diet in the etiology of CC has been recognized and extensively studied
in the last few decades (Potter, 1996; Lipkin et al., 1999; Sanderson et al., 2004). A
review by American Institute for Cancer Research (AICR/WCRF, 1997) estimates that

that 67-75% of CC incidence could be prevented by adequate diets. Epidemiological

95

studies have established a positive correlation between colon cancer risk and increased
intake of animal fat (Wynder et al., 1969; Reddy et al., 1978; McKeown-Eyssen and
Bright-See, 1984; Sandler, 1996) and an inverse association between colon cancer risk
and consmnption of diets rich in fruits and vegetables (Steinmetz and Potter, 1991; Thun
et al., 1992; Potter, 1996). Increasing consumption of foods that contain anti-carcinogenic
phytochemicals (e. g. fruits, vegetables, soybeans) is a major public health strategy for
reducing the incidence of many diet-related cancers such as CC. Soybean foods in
particular have been promoted as having the potential to reduce CC risk (Messina and
Bennink, 1998). Soy contains a number of compounds having potential anti-cancer
activity. These compounds include isoﬂavones, protease inhibitors, phytosterols,
saponins and phytate.

Soybeans are a rich and unique source of the isoﬂavones genistein and daidzein
(1-3 mg of total isoﬂavone per gram of dry matter). lsoﬂavones typically occur as
glycosides and are normally associated with the protein fraction of soybeans. The type
and nature of processing can affect the quantity of isoﬂavones in processed soy products.
A portion of the isoﬂavones are lost during preparation of SPI, but certain conventional
means of preparation (with alkaline solubilization and acid precipitation) can produce SPI
which contain 1mg or more total isoﬂavones per gram of dry matter (Lucas and Riaz,
1995)

Numerous rodent studies have examined the potential of dietary soy to reduce
colon cancer development. Hakkak et al. (2001) reported a statistically signiﬁcant
reduction (P<0.05) in AOM-induced CC when male rats were fed soy protein isolate

(SPI) compared to casein. Bennink (2001) demonstrated Signiﬁcant inhibition of colon

96

 

cancer development by feeding full fat soy ﬂour or defatted soy ﬂour to AOM-induced
rats. However, some animal studies reported that soybean protein had no protective
effect against colon carcinogenesis when compared to beef protein (Reddy et al., 1976;
Clinton et al., 1979; McIntosh et al., 1995).

The potential of soy isoﬂavones to reduce cancer risk has been the focus of many
recent studies (Sorensen et al., 1998; Gee et al., 2000; Guo et al., 2004). Genistein has
received particular interest because of its ability to inhibit several enzymes (particularly
tyrosine kinases, (Akiyama et al., 1987) involved in cell cycle regulation. Numerous in
vitro studies have demonstrated that genistein inhibits the growth of a variety of cancer
cell lines at concentrations ranging from 5—50 umol/liter (Yanagihara et al., 1993; Booth
et al., 1999; Wenzel et al., 2000). Several studies have examined the efﬁcacy of
isoﬂavones to reduce colon cancer risk in animal models. Studies conducted by Pereira et
al. (1994) found that addition of 150 mg genistein per kg of diets fed to rats resulted in 34
% reduction in the numbers of aberrant crypt foci (ACF) in rat colons after initiating CC
by azoxymethane (AOM) injection. Helms and Gallaher (1995) found that feeding
genistein (3 70 mg/kg diet) reduced DMH-induced ACF by 35%. Collectively, these
studies suggest that feeding puriﬁed isoﬂavones inhibit preneoplastic biomarkers of colon
carcinogenesis. Increased labeling index (LI) and an upward shiﬁ in the proliferation
zone (PZ) within a colonic crypt are two biomarkers believed to be related to increased
risk for developing CC (Lipkin, 1987; Deschner et al., 1988; Gerdes et al., 1993; Mills et
al., 1995; Akedo et al., 2001). A demonstrated reduction in L1 and P2 by dietary soy

would further substantiate its potential efﬁcacy to reduce CC risk.

97

Another factor that has been extensively studied for its potential to reduce CC risk
is dietary calcium. The association between calcium intake and colorectal remains
uncertain, despite considerable epidemiologic and laboratory research (Baron et al.,
1999). It has been proposed that dietary calcium binds to bile acids in the lumen of the
bowel, inhibiting their potential to promote colon cancer (N ewmark et al., 1984; Van der
Meer and De Vries, 1985; Van der Meer et al., 1990). Rodent studies examining the
inﬂuence of dietary calcium on colon carcinogenesis have reported inhibition of bile-
induced mucosal damage and experimental carcinogenesis (Govers et al., 1994; Piard et
al., 1994). Several randomized human studies have found that supplementary calcium
(range of 1.25 g to 2 g/day) reduces rectal mucosal proliferation (Lipkin and Newmark,
1985; Rozen et al., 1989; Bostick et al., 1995; Wargovich et al., 1995a), although other
human studies have not found this effect (Stern et al., 1990; Armitage et al., 1995; Baron
et al., 1995; Cats et al., 1995; Karagas et al., 1998).

While there have been a number of studies examining the inﬂuence of dietary
calcium on biomarkers of cell proliferation in colonic epithelium in animal models and in
humans, we are aware of no studies examining the effect of soy protein on colonic
epithelial cell proliferation in humans. The objective of this research was to determine if
consuming 38 g/day of soy protein isolate supplement (containing 58 mg total genistein)
would reduce CC risk as reﬂected by changes in colonic epithelial cell proliferation
patterns in humans. A secondary objective was to determine if adding calcirun to the soy

protein isolate supplement would alter its impact on colon epithelial cell proliferation.

98

C. MATERIALS AND METHODS
Experimental Design

This investigation was a 12-month, randomized, double-blind study with parallel
treatment and control groups evaluating the effects of soy protein on colonic epithelial
cell proliferation and other potential parameters of colon cancer risk. Subjects were
randomly assigned either to one of two supplements in the treatment arm (soy protein
isolate with or without added calcium) or the control arm (calcium caseinate). Treatment
assignments were determined by a statistician. Changes in colon epithelial cell
proliferation patterns were measured before and after consuming the supplements for one
year. This research was approved by the Michigan State University Committee for
Research Involving Human Subjects and informed, written consent was obtained from all

study participants.

Subjects
Subjects were eligible for participation in this study if they had a previous or

current history of two or more colonic adenomatous polyps or history of colon cancer and
if they were between the ages of 35-70 years at study entry. Potential subjects were
excluded if they had a history of allergy to soy or soy products, b) milk allergy, c) serious
intestinal disease such as chronic inﬂammatory bowel disease, or d) signiﬁcant diseases
of the cardiovascular, pulmonary, hepatic, gastrointestinal, renal, neurological,
hematological systems. Subjects were primarily recruited from a list of patients from the
gastroenterology unit at the Clinical Center of Michigan State University. Some subjects

were solicited from the greater Lansing area by newspaper advertisement.

99

Advertisements were placed at nearby hospitals and clinics to encourage recruitment.
Potential subjects (n = 1, 200) having a history of colon polyps were screened for
eligibility. Of this group, 400 subjects met the eligibility criteria. After veriﬁcation of
previous history of adenomatous polyps or CC from patient records, seventy subjects
entered the study. Patient histories, physical examinations, ﬂexible sigmoidoscopy or
colonoscopy, colon biopsies and blood samples were obtained for baseline information.
After consuming the supplements for twelve months, blood samples and colon biopsies
were obtained. F orty-two subjects (27 males and 15 females) completed the study.
Subjects were less than 70 years old (age range: 35-70 years), were ambulatory,
did not have any other gastro-intestinal disorders, consumed 2 or fewer aspirins/day, and
were requested to not consume soy products. Subjects provided bimonthly urine samples
to allow monitoring of urinary isoﬂavone excretion to conﬁrm supplement consumption
(by the groups consuming either of the two soy protein isolate-based supplements) and to -
conﬁrm minimal soy consumption by the group consuming the casein-based supplement.
At baseline and at the end of one year, each participant also completed a food frequency

questionnaire.

Dietary Treatments
Age- and gender-matched subjects were randomly assigned to one of the three
treatments: (a) casein beverage [n = 13 subjects completed the study, mean age = 58, 8
males and 5 females] (b) soy protein isolate (SP1) beverage [n = 16, mean age = 59, 13
males and 3 females], and (0) SP1 beverage without calcium [n = 13, mean age = 59, 6

males and 7 females]. The SPI-based supplements were based on SuproTM brand soy

100

protein manufactured by Protein Technologies International (Saint Louis, MO). The SPI
beverage powder was commercially available and contained 0.6 g calcium per packet.
Given the potential effects of calcium on indices of colon epithelial cell proliferation to
be measured in the study, a soy treatment lacking calcium was formulated to allow direct
determination of the effect of soy protein without additional supplemental calcium.
Calcium caseinate was the protein source used in the casein-based supplement. Hence,
the casein-based supplement powder contained 1.01 g calcium per packet. Since the
supplements were formulated to provide nutrients other than protein, they also contained
additional vitamins (A, B1, B2, niacin, B6, C, D and Big) and minerals (magnesium and
zinc). The supplements were formulated to contain equivalent concentrations of all
essential nutrients except calcium. The nutrient composition for each packet (27 g) of
each of the three dietary supplements is presented in Table 8. Subjects were instructed to
consume 2 packets/day of each supplement. In addition to other nutrients, the
supplements each provided approximately 40 g of protein and 200 kcal per day. The two
SPI-based supplements provided approximately 58 mg of genistein (primarily as
glycosides) daily. As a supplement to their normal diets, subjects self-administered these
beverage powder supplements which they consumed twice each day (morning and
evening) for one year. Subjects were provided instructions on how to incorporate the
beverage powder into their typical diet. Typically, the beverage powder supplements
were mixed either in water, juice or carbonated beverages prior to consumption. There

was no other dietary intervention imposed on these free-living subjects.

101

Nutrient intake
Nutrient intake was calculated from food frequency questionnaires (F FQ)
administered at baseline and after consumption of treatment or control beverage for one
year (Health Habits and History Questionnaire developed by Dr. Gladys Block, National

Cancer Institute, DIETSYS-Version 3.0).

Compliance
Subject compliance with the experimental protocol was assessed by: 1) self-
completed consumption log; 2) monitoring unused supplement; 3) phone contacts and 4)

by measuring isoﬂavone excretion by HPLC analysis of bi-monthly urine samples.

Tissue procurement and processing
Colon mucosa biopsies were obtained from each subject at the beginning of the

study prior to starting dietary treatment. Another set of biopsies was obtained from each
subject at the completion of the one-year study period. Subjects received an oral
electrolyte bowel preparation (GolytelyTM containing polyethylene glycol 3350““, NaCl,
KCl, NaHCO3, and N32804) the evening before ﬂexible sigmoidoscopy or colonoscopy.
Colonic biopsy samples (4 per subject) were obtained during ﬂexible sigmoidoscopy or
colonoscopy using “jumbo” forceps. Biopsies (average size 3 X 8 mm) were obtained
from normal appearing mucosa 30-40 cms from the anal verge. All biopsies were
immediately rinsed in phosphate buffered saline (pH 7.4). Two biopsies were placed
individually in 0.6 ml microcentrifuge tubes and immediately frozen on dry ice for

subsequent slot-blot quantiﬁcation of PCNA. The remaining two biopsies were pinned

102

ﬂat to cardboard strips and ﬁxed in B5 ﬁxative for 2 hours, then ﬁxed in neutral buffered
formalin for 2—6 hours. The ﬁxed biopsies were placed between foam pads in histopath
cassettes, dehydrated, inﬁltrated with parafﬁn and embedded with parafﬁn using standard

histologic procedures.

PCNA and Ki—67 IHC analyses

Three micron thick sections of parafﬁn-embedded biopsy samples were cut and
mounted on poly-L-lysine coated glass slides and dried at 58° C for 2 hours. Sections
were cleared of parafﬁn with xylene, hydrated in increasing ratios of waterzethanol, and
treated with iodine2potassium iodide and 5% sodium thiosulfate to remove mercury.
Samples were then subjected to antigen retrieval (10 mMol/L citrate buffer, pH 6.0, 95
°C) for 20 minutes before immersion in 0.3% hydrogen peroxide for 10 minutes to
quench endogenous peroxide activity. Following rinses in tris-buffered saline (TBS),
sections were then incubated with primary antibody (1:100 dilution of PC10 [anti-PCNA]
or MMl [anti- Ki-67] respectively in 1% BSA in Tris buffered saline, pH 7.4) overnight
at 4 °C. Aﬁer subsequent incubations in biotinylated rabbit anti-mouse immunoglobulins
(diluted 1:100, 45 min), followed by peroxidase-conjugated streptavidin (diluted 1:300,
45 min), they were treated with 3-amino-9-ethylcarbazole (ABC) and color development
was monitored for 10—20 minutes until red staining was evident. The sections were
rinsed twice with 10 mMol/L TBS in between incubations with antibodies, streptavidin
and ABC. Slides were then counter-stained with hemotoxylin and cover-slipped using
Faramount mounting medium. Appropriate positive (human tonsil) and negative controls

(omission of primary antibody) were used for evaluation of PCNA and Ki-67 staining.

103

The primary antibodies were obtained from Novacastra (now Vector Laboratories;
Burlingame, CA), and biotinylated rabbit anti-mouse immunoglobulin, streptavidin,

AEC, and Faramount were obtained from Dako (Carpinteria, CA).

Quantiﬁcation of PCNA and Ki—67 Immunohistochemistry

PCNA and Ki-67 labeling were quantiﬁed by the percentage and positions of
positively stained (red) nuclei within full-length crypts in colonic sections. Well-oriented
longitudinally-sectioned full-length crypts (n = 10 per subject) were identiﬁed by light
microscopy. A scorable crypt was deﬁned as one in which the base touched the
muscularis mucosa and had an open lumen at the top. The number and position of PCNA
and Ki-67 labeled cells and crypt heights were recorded for ten full crypts for each
subject to determine cell proliferation parameters. All cells with a red nuclear staining
were counted as positive. The center-most cell at the base of the crypt, and in between the
hemicrypts, was designated as cell number one and the cell at the top of the crypt mouth
was considered the highest cell. A continuous column of cells from cell number one to
the highest cell were counted along each hemicrypt wall of each crypt. Crypt height was
deﬁned as the number of cells per hemicrypt.

Total Crypt Labeling Index (LI) was calculated by dividing the number of labeled
cells in a hemicrypt by the total number of cells in a hemicrypt (crypt height). These
values were averaged for at least 20 hemicrypts (representing 10 full-length crypts) to
obtain the L1 for each subject. Compartment Labeling Indexes (basal LI, mid LI and top
LI) were calculated by dividing the number of labeled cells within each third (basal,

middle and top) of each hemicrypt by the number of cells in each third of the respective

104

hemicrypt. Proliferation zone (PZ) was obtained by dividing the position of the highest
labeled cell from the hemicrypt base by the total number of cells in the hemicrypt. Each
of these PZ values were averaged for at least 20 hemicrypts to obtain the overall PZ for

each subject.

PCNA Measurement by Slot-blot
Slot-blot analysis to determine relative PCNA protein concentration was

conducted on frozen biopsies obtained at the beginning and end of treatment. Biopsies
were homogenized with a glass pestle using 0.5m] of non-denaturing buffer containing
protease inhibitors (20mMol/L MOPS, 150 mMol/L sodium chloride, 1% vol/vol Triton
X-100, 1% wt/vol deoxycholate, 0.1% va01 SDS, 1 mMol/L EDTA, 100 uMol/L
sodium orthovandate, 250 uMol/L PMSF, 10 ug/ml leupeptin, 1 jig/ml pepstatin). Total
protein concentrations in homogenates were determined by the biocinchoninic assay
using BSA as the standard. Relative PCNA levels in tissue homogenates were then
determined by immuno-quantiﬁcation using slot-blot methodology. PVDF membranes
(Immobilon P membranes; Millipore, Billerica, MA) were wetted in methanol and then
transferred to PBS. Twenty pg of protein from each sample (dilutions in PBS) was
vacuum blotted on to the PVDF membranes using a slot-blot apparatus (Hoefer Scientiﬁc
instruments). The membranes were held overnight at 4 °C in blocking buffer (3% BSA in
TBS with Tween 20). The membranes were then incubated overnight with monoclonal
PC-10 antibody (1 : 1000 dilution of primary antibody in blocking buffer). After rinsing
with wash buffer (TBS with Tween 20) the membranes were incubated in biotinylated

secondary antibody (1:1000 dilution of goat anti-mouse IgG in wash buffer for one hour).

105

Blots were rinsed twice with wash buffer and then incubated with streptavidin-peroxidase
for 1 hour. After subsequent rinses with wash buffer, the membranes were treated with a
chemiluminescence reagent (Dupont NEN research products) and light emission was
captured by autoradiography. Developed autoradiography ﬁlms were quantiﬁed by
scanning densitometry. Known amounts of mouse IgG F c fragments (standard curve
ranging from 0.3125ng to 10.00ng) were included in each blot to allow for normalization

of staining intensity across blots.

Statistical Analysis
Crypt height, PCNA and Ki-67 proliferation indices (total crypt LI, P2, and

compartmental LI for basal, middle, and top compartments of the crypts) and relative
PCNA quantiﬁed by the slot-blot method each were statistically analyzed by two-way
Analysis of Variance (treatment and time as the two factors) using the General Linear
Models procedure of SAS (version 6.11, SAS Institute Inc., Cary, NC). When signiﬁcant
treatment or time effects were detected (P < 0.05), the least signiﬁcant difference was
used to compare treatment means. The means and standard errors presented in Tables 4, 5
and 6 were generated from this 2-way analysis of variance. Comparisons of beginning
(TB) and ending (TE) values within each treatment and parameter were conducted using

paired t-tests for each time pair.

106

D. RESULTS
Nutrient Intake from Food Frequency Records

Table 10 provides information on the intakes of selected nutrients, based on the
food frequency questionnaire administered at the beginning and end of the supplement
period, of the 42 participants who completed the study. There were no statistically
signiﬁcant differences among the three groups for any of the dietary variables presented
except for calcium intake in subjects (Table 10). The total calcium intake (both from diet,
multi—vitamin use and beverage supplement) varied across the treatment groups, with the
calcium intake in the casein beverage supplement groups being the highest when

compared to the soy beverage supplement groups.

Whole Crypt Proliferation Indices

Results for whole crypt PCNA and Ki-67 labeling indices are presented in Table
1 1. Crypt height was not inﬂuenced by time (beginning versus end) or by dietary
treatments and, for all subjects, averaged 64 cells per hemicrypt. Prior to initiation of the
dietary treatments (TB values), approximately 30% (PCNA) and 9% (Ki-67) of the
colonic crypt cells stained positively for PCNA and Ki-67, respectively. Total crypt LI
for PCNA and Ki-67 were not signiﬁcantly different among the supplement groups at the
beginning of the study. Subjects who consumed the SP1 with calcium supplement had a
signiﬁcant (P < 0.05) reduction in total crypt PCNA LI from 30.3% initially to 23.6%
after one year of supplementation. There was no statistically signiﬁcant reduction in the
mean total crypt PCNA L1 in subjects that consumed the SP1 without added calcium or

the casein supplements, although subjects consuming the SP1 minus calcium supplement

107

tended (P < 0.10) to have reduced whole crypt PCNA Ll at the end of the study (28.5%)
as compared to the beginning of the treatment period (31.5%). When the two soy-
containing treatments were combined, there was a signiﬁcant (P < 0.05) reduction in total
crypt PCNA LI after one year of supplement consumption. Total crypt labeling with Ki-
67 was not inﬂuenced by supplement or time.

Prior to initiation of the dietary treatments, (T3 values), PCNA and Ki-67 PZ
values did not differ among the treatment groups and averaged (across all treatments)
51% and 45% for PCNA and Ki-67, respectively (Table 11). Subjects who consumed the
SP1 with calcium supplement had a signiﬁcant (P < 0.05) reduction in the mean PCNA
PZ from 55% initially to 44% after one year. There was no statistically signiﬁcant
reduction in the mean total crypt PCNA P2 in subjects that consumed the SP1 without
calcium (beginning value: 49% and ending value: 47%) or the casein (beginning value:
49% and ending value: 47%) supplements. When the two SP1 supplement (+/- calcium)
groups were combined, there was a statistically signiﬁcant reductiOn in the mean PCNA
PZ from 53% to 45% during the course of the study. The mean total crypt Ki-67 P2 was

not inﬂuenced by dietary supplement or by time.

Compartmental Labeling Indices (basal LI, mid LI, top LI)

Results for compartmental LI indices for PCNA are presented in Table 12. Given
the observed changes in total crypt PCNA LI and PZ (Table 12), we were interested in
determining how PCNA labeling patterns in speciﬁc regions of the crypts were
inﬂuenced by consumption of soy protein. In subjects that consumed the SP1 plus

calcium supplement, there were signiﬁcant (P < 0.05) reductions in PCNA L1 values in

108

 

the middle compartment (32% to 19%) and top compartment (4% to 0.5%), and a
tendency (P < 0.10) for reduced PCNA L1 in the basal compartment (55% to 51%),
during the course of this study. Subjects that consumed the SP1 without calcium
supplement also tended (P < 0.10) to have reduced basal compartment PCNA LI at the
end of the study (5 5%) as compared to the beginning value (57%). There was a
statistically signiﬁcant increase in the basal compartment PCNA Ll values (beginning
value: 52% and ending value: 57%) in subjects that consumed casein beverage powder
but no changes were observed in PCNA L1 in the middle or top compartments for
subjects consuming the casein supplement.

When the two SP1 supplement (+/- calcium) groups were combined, there were
signiﬁcant (P < 0.05) reductions of PCNA L1 in the basal compartment (56% to 53%),
middle compartment (34% to 24%) and top compartment 3% to 0.4% during the course

of the study (Table 12).

Relative PCNA Protein Concentration in Colonic Mucosa

Data on PCNA protein expression in the colonic mucosa of subjects are presented
in Table 13. The values for PCNA protein are relative values based on scanning
densitometry and normalized to a known concentration of mouse IgG F c fragments.
PCNA protein expression did not differ among the supplement groups at the beginning of
the study. After one year of supplement consumption, there were signiﬁcant (P < 0.05)
reductions in relative PCNA protein concentrations in subjects who consumed either SPI
plus calcium (10.5% reduction) or SP1 minus calcium (8.3% reduction). Averaged across

both soy supplement groups, there was a 9.5% reduction in relative PCNA protein

109

 

 

expression in colonic mucosa] cells during the course of the study. Relative PCNA
protein concentration in subjects consuming the casein supplement were not Signiﬁcantly
different aﬁer supplementation when compared to the values before initiation of the

study.

E. DISCUSSION

Most human studies on soy consumption and cancer risk have focused on
epidemiological associations, pharmacokinetic properties of soy components, and on the
effects of isoﬂavones on circulating hormone levels. To date, there has been no
intervention trial conducted in htunans to study the potential efficacy of soy or soy-
containing foods to reduce risk of colon cancer. In this study, we evaluated the potential
efﬁcacy of soy protein isolate to reduce intermediate biomarkers (PCNA and Ki-67
expression in colonic crypts) associated with colorectal cancer risk in humans.

PCNA is a cell proliferation marker that accumulates as the cells progress through
G], S, and M phases of the cell cycle and returns to low levels during G0. Terminally
differentiated cells do not express PCNA (Bromely et al., 1996). Studies have shown that
PCNA can be used as a reliable marker for the proliferative activity in colonic crypts
(Malecka-Panas et al., 1997; Barnes et al., 1998; Izawa et al., 2002). Many studies have
used IHC methods to demonstrate increased PCNA labeling indices in neoplastic tissues
of adenomatous polyps and colon tumors (Yoshikawa and Utsunomiya, 1996; Bostick,
1997; Bostick et al., 1997a; Bostick et al., 1997b; Shpitz et al., 1997; Ottaviani et al.,
1999; Saito and Mori, 1999; Wu et al., 2000; Diez et al., 2003). In particular, PCNA

expression (as measured by immunohistochemistry) has been used in human studies to

110

 

demonstrate a step-wise increase in proliferation indices during neoplastic progression of
colonic lesions (Shpitz et al., 1997), for predicting the potential for metastases in invasive
colorectal carcinoma (Tanaka et al., 1995; Choi et al., 1997), and to predict the
development of metachronous adenomas (Ottaviani et al., 1999).

Ki-67 is a cell proliferation marker that is primarily expressed during the S-phase
of cell cycle and is absent in G0. While some researchers have shown that Ki-67 LI can
be used as a valuable prognostic indicator in cancers including CC (Gerdes et al., 1987;
Tungekar et al., 1991; Wintzer et al., 1991; Suzuki et al., 1992; Kang et al., 1997; Akedo
et al., 2001; Garrity et al., 2004), others (Sahin et al., 1994; Green et al., 1998; Paspatis et
al., 1998) have demonstrated no differences in Ki-67 total labeling indices when
comparing normal subjects with persons with CC or high risk for developing CC.
However Paspatis et al. (1998) and Mills et al. (1995) reported an expansion of
proliferative compartment in subjects with neoplasms and in subjects with F AP
respectively despite no differences in the overall total labeling indices with Ki-67.

Normally, proliferation in colonic crypts is restricted to the basal 60% of the
crypts. As colonocytes migrate up the crypt, these cells undergo terminal differentiation.
Ultimately, after approximately 7 days of transit time to the crypt surface, colonocytes
undergo programmed cell death and are extruded. Under normal circumstances, cell
proliferation does not occur in the upper 40% of the crypts (Lipkin, 1971). An expansion
of the proliferative compartment toward the crypt surface (hyperproliferation in the
middle and upper thirds of the crypt) is considered a preneoplastic biomarker (Lipkin,
1988) and has been associated with increased risk for development of adenomatous

polyps and carcinoma (Risio et al., 1991). Expansion of the proliferation zone has been

111

shown in a variety of neoplastic conditions, most notably in FAP and in colon
tumorigenesis (Rozen et al., 1991; Richter et al., 1993; Mills et al., 1995). In general,
expansion of the proliferative compartment indicates that normal mechanisms controlling
colonic epithelial cell proliferation and differentiation are compromised.

In our study we used the immunohistochemical detection of PCNA and Ki-67 to
detect crypt cell proliferation patterns in the biopsies of subjects that consumed soy or
casein beverage for one year. The subjects participating in this study either had a history
of adenomatous polyps or colon cancer and, therefore, represented a group that was at
high risk for colon cancer. We have previously demonstrated (Chapter 3) that humans
who have a history of adenomatous polyps or colon cancer, when compared to age-
matched normal subjects, have increased PCNA LI and increased relative expression of
PCNA protein in colonic crypts. Our overall objective was to determine if consumption
of soy protein supplement with isoﬂavones would reduce these indices of colon cancer
risk.

We observed a signiﬁcant decrease in total crypt PCNA labeling index and
proliferation zone in colonic mucosal biopsies of subjects that consumed the soy protein
supplements for one year. Subjects who consumed the casein supplement did not have
any changes in PCNA or Ki-67 labeling indices in colonic mucosa. It is noteworthy that
no signiﬁcant effects of supplement consumption on Ki-67 L1 or PZ were detected,
whereas soy supplement consumption reduced whole crypt PCNA LI and P2 (Table 11).
Taken together, these observations suggest that consumption of the soy supplement did
not inﬂuence the S-phase fraction of colonic epithelial cells, but did reduce the fraction of

cells capable of undergoing another round of cell division (based on reduced PCNA LI).

112

 

 

Further, reduction of the PCNA PZ by soy supplementation suggests that colonic
epithelial cells underwent terminal cell differentiation relatively earlier than in subjects
consuming the casein-based supplement. Overall, these observations indicate that
consumption of the soy protein supplement is associated with changes in intermediate
biomarkers of colonic epithelial cell proliferation that are consistent with reduced risk for
colon cancer development.

Our ﬁndings are that consumption of SPI consumption reduced colonic epithelial
cell proliferation and, presumably, CC risk. Studies in laboratory animals have shown
that a soy-rich diet inhibits chemically induced carcinogenesis (Pereira et al., 1994;
Koratkar and Rao, 1997; Kennedy et al., 2002). Hakkak et al. (2001) found that feeding
rats a SPI-based diet reduced incidence of AOM-induced colonic tumors when compared
with rats fed a casein-based diet. Rats fed the casein diet had a 50% incidence of colon
tumors compared with 12% on soy protein based diet (Hakkak etal., 2001). Guo et al.
(2004) reported that feeding an isolate of soy protein containing isoﬂavones reduced
colon tumor incidence in ovariectomized estrogen receptor-u knockout female mice
relative to that observed in these mice when fed a casein-based diet.

The precise mechanisms whereby SPI prevents colon cancer in animal models are
unknown. However, factors that are associated with the isolated protein, particularly the
isoﬂavones genistein and daidzein, have been implicated as being important (Messina
and Erdman, 1998). Several potential mechanisms have been proposed for the anti-
carcinogenic activity of isoﬂavones. These proposed mechanisms include a) inhibition of
tyrosine kinases; b) inhibition of proteases; c) anti-oxidant activity, (1) increased synthesis

and decreased degradation of steroid hormone binding globulin; e) weak estrogen

113

agonist/antagonist activity mediated through estrogen receptor a; f) altered hormone
metabolism or action; g) induction of Phase I and II detoxiﬁcation systems; and h)
inhibition of carcinogen activation (Wattenberg, 1985; Barnes et al., 1990; Ronis et al.,
1999). Several studies have indicated that treatment with puriﬁed isoﬂavones or soy-
based diets can induce phase II enzymes such as glutathione S-transferases (GSTS) in the
liver, small intestine and colon (Appelt and Reicks, 1999).

In addition to isoﬂavones, soy contains a number of other biologically active
factors, including saponins, phytosterols, protease inhibitors and inositol hexaphosphate,
that have been postulated to inﬂuence colon carcinogenesis (Messina and Bennink,
1998). Given the composition of SPI used in this study, we are unable to ascertain the
relative contributions of isoﬂavones or the other constituents present in soy to our
observed reductions in colonic epithelial cell proliferation. We have previously reported
that addition of 150 mg genistein/kg of a diet based on soy concentrate led to signiﬁcant
reductions in ACF numbers and PCNA LI and PZ in colonic mucosa of AOM-initiated
rats (Chapter 5). These animal model data suggest that isoﬂavones alone could be
responsible for the effects on intermediate biomarkers observed in this study.

A secondary objective of this study was to assess the impact of supplemental
calcium on intermediate biomarkers for CC risk. There is some evidence that calcium
prevents colorectal carcinogenesis by inﬂuencing a complex series of signaling events
induced at various tiers of colonic cell organization (Lamprecht and Lipkin, 2001).
Several animal studies (Lipkin et al., 1999) and some human studies (Lipkin and
Newmark, 1985; Rozen et al., 1989; Wargovich et al., 1992; Bostick et al., 1995; Holt et

al., 1998) but not all (Gregoire etal., 1989; Stem et al., 1990; Kleibeuker et al., 1993;

114

Baron et al., 1995) show that consumption of calcium and dairy foods reduced colonic
epithelial cell proliferation. Few clinical trials also have suggested that calcium intake
reduced the recurrence of colorectal adenomas (Baron et al., 1999; Bonithon-Kopp etal.,
2000). Overall, the beneﬁcal effects of calcium on CC seems modest and contradictory.
The casein supplement used in this study contained the calcium salt of casein as
the protein source, and therefore, subjects consuming this supplement were receiving an
additional 2 g of calcium per day. When compared with subjects who consumed either
of the soy-based supplements, subjects consuming the casein supplement also reported
greater consumption of calcium from mineral supplements during the end-of—study F F Q.
Despite the greater intake of calcium by subjects consuming the casein treatment, no
signiﬁcant reductions in total crypt PCNA L1 or P2 or relative PCNA protein
concentration were observed in colon biopsies obtained from these persons at the end of
this study. The PCNA proliferation indices of subjects assigned the casein beverage
treatment at the beginning of the study were numerically lower than those of the subjects
assigned to either SPI treatment, but these differences in initial PCNA LI and P2 were
not statistically different among treatments. Thus, we do not believe that the lack of
effect of calcium caseinate on colonic mucosal proliferation indices can be explained by
slight differences in initial PCNA L1 or PZ. Moreover, other studies (Bostick et al.,
1993a; Bostick et al., 1993b; Baron et al., 1995; Weisgerber et al., 1996; Alberts etal.,
1997) have also found no effect of calcium supplementation on colonic epithelial cell
proliferation contrary to what was previously demonstrated (Lipkin and Newmark, 1985;
Wargovich and Lointier, 1987; Rozen et al., 1989; Wargovich et al., 1992). The original

studies of Lipkin and Newmark (1985) and associates all reported Signiﬁcant but modest

115

reductions in colonic proliferation indices in subjects when calcium was provided in
doses of 1.2-1 .5 g/day. However in a recent review article Chia (2004) concluded that
that there is strong evidence of a threshold effect and that calcium intake by humans
greater than 700mg/day does not beneﬁt the colon.

We did observe that subjects consuming SPI plus calcium had greater reductions
in colonic mucosal PCNA LI and P2 when compared to subjects consuming SPI without
calcium. Thus, we can not rule out the possibility that effects of SP1 and calcium on
colonic epithelial cell proliferation may be additive.

In a previous study (Chapter 3) we provided evidence indicating that PCNA
proliferation indices may be a better discriminator than Ki-67 for predicting colon cancer
risk. Colon cancer has long been recognized as an age-related disease (American Cancer
Society, 1998) and research by Roncucci et al. (1988) and Deschner et al. (1988) has
consistently demonstrated an increase in colonic epithelial cell proliferation indices with
increasing age. We previously measured PCNA labeling index and proliferation zone in
human subjects that had known intestinal diseases (CC, FAP, adenomatous polyps) and
in subjects of varying age with no history of colonic disease. We found that both PCNA
LI and PZ increased incrementally with increasing age (from 20 to 70 years of age) in
normal subjects. Persons having a history of adenomatous polyps had PCNA LI and P2
values that were comparable to or greater than that observed in normal subjects ranging
in age from 70-79 years (Chapter 3). The beginning proliferation indices for subjects
consmning the SP1 plus calcium supplement (Table 11) in this study were comparable to
indices observed in subjects in the 70-79 year age group or in subjects with a history of

adenomatous polyps in our previous study (Chapter 3). The end-of-study PCNA LI and

116

 

 

PZ indices of subjects that consumed SP1 in this study compares favorably to that
observed in 50-year-old normal subjects having no history of colonic diseases reported in
our earlier study (Chapter 3). The American Cancer Society (2005) reports that the
percentage of people developing malignant colon cancer at the age range of 55 — 60 years
is much lower than that for seventy-ﬁve years of age. Therefore, we interpret the
downward shift in proliferation zone and decrease in labeling index after consumption of
soy protein isolate as a meaningful decrease in colon cancer risk.

We previously demonstrated that relative PCNA protein concentrations (as
determined by slot-blot analysis) were highly correlated with PCNA L1 and PZ indices in
colonic crypts of humans who varied in age and risk for CC development (Chapter 3). In
this study, we further demonstrated that relative PCNA protein concentrations responded
to dietary supplementation with SP1. This indicates that, like whole crypt PCNA LI and
PZ, relative PCNA protein concentration in the colonic mucosa is responsive to dietary
intervention. Further study is necessary to determine if relative PCNA protein expression
in colonic mucosa is a potentially valid method to predict CC risk of humans that is more
facile than whole-crypt PCNA LI and PZ analyses.

In summary, we demonstrated in this study that supplementing the diets of free-
living adults who were at elevated risk for CC with SP1 caused signiﬁcant changes in
PCNA indices that correlate with reduced risk for CC. Administration of a calcium
caseinate supplement did not inﬂuence whole-crypt PCNA L1 or PZ despite a much
greater intake of calcium. Given that SP1 supplementation reduced PCNA LI and PZ and

total PCNA protein concentrations, but did not inﬂuence Ki-67 L1 or PZ, we conclude

117

 

 

that the major effect of SPI supplementation on colonic epithelial cells was the induction

of more rapid cell differentiation as these cells migrate toward the crypt surface.

118

f

 

Table 8. Nutrient Composition of Dietary Supplements

 

 

 

Intake/ supplement Dietary Beverage Supplement
packetl

Casein SP12 plus SPI minus

Calcium Calcium

Moisture (g) 0.72 0.63 0.80
Protein (g) 20.49 18.76 18.95
Fat (g) 0.25 1.10 1.10
Ash (g) 1.83 2.78 1.10
CHO (g, by difference) 4.21 4.28 4.00
Kcal 101 102 102
Calcium (g) 1.01 0.60 0.05
Magnesium (mg) 53 55 53
Zinc (mg) 3.46 2.41 2.04
Vitamin A (IU) 952 1060 880
Vitamin D (IU) 139 132 130
Vitamin C (mg) 3.16 2.70 3.15
Vitamin B1 (mg) 0.16 0.13 0.22
Vitamin B2 (mg) 0.36 0.45 0.54
Niacin (mg) 0.29 0.40 0.44
Vitamin B6 (mg) 0.21 0.22 0.29
Folacin (pg) 19 74 45
Vitamin B12 (pg) 1.24 2.34 1.29

 

' Two packets were consumed per day.
2 SP1 — Soy Protein Isolate.

119

 

Table 9. Characteristics of Study Participants at Baseline

 

Dietary Beverage Supplement

 

SPII plus SP1 minus

 

 

Characteristics at Baseline Casein Calcium Calcium
(n= 13) (n=16) (n=13)
Male Gender (%) 62 81 46
Mean Age (yrs) 58 59 59
Family/Personal history of colorectal 1/13 0 2/13

carcinoma (number)

 

I SP1 — Soy Protein Isolate.

120

 

Table 10. Daily Intakes of Selected Dietary Nutrients by Study Participants at the
beginning (TB) and ending (T5) of each treatment

 

Dietary Beveragg Supplement

 

 

 

Casein SPIl plus SP1 minus Calcium
Calcium

(n=13) (n=16) (n=13)
.“z::i.‘:.““'°““ T. T. T. T.
Calories (kcal/day) 1705 1641 1934 1654 1651 1747
Protein (g/day) 64 62 75 66 64 76
Total Fat (g/day) 79 68 78 62 71 98
Carbohydrates (g/day) 177 187 219 206 186 212
Calcium from food 752 705 770 882 701 763
(mg/day)
Calcium from mineral 125 409 535 96 50 90
supplement (mg/day)
Calcium from beverage - 2000 - 1200 - 100
treatment (mg/day)
Use of multi-vitamin 17 36 21 15 10 9
mineral pill (%)

 

I SP1 — Soy Protein Isolate

2 TB — Indices prior to consumption of beverage treatment, and TE — Indices after
consumption of beverage treatment for one year

3 Values represent pooled results after consumption of beverage supplement at the end of
the one-year study period.

121

 

Table 11. Whole Colonic Crypt Proliferation Indices“ 7 (PCNA and Ki-67) of subjects

 

Dietagy Beverage Supplement

 

 

SP12 plus SPI minus SP1 +/-
Casein Calcium Calcium Calcium3
(n=13; (n=16; (n=13; (n=29)
Indices mean age=5 8) mean age =59) mean age=59)
Crypt Height
(Pooled value)4 62.0 : 1.4 64.1 : 1.3 65.8 : 1.4
PCNA LIS-TBl 0.286 3: 0.013 0.303 i 0.012 0.315 3: 0.013 0.308 i 0.010
Tel 0.282 i 0.013 0236* i 0.012 0.285i i 0.013 0258* i 0.010
PZ6- TB 0.487 :1: 0.024 0.554 1“ 0.022 0.498 i 0.024 0.529 t 0.017
TE 0.476 at 0.024 0436* i 0.022 0.473 at 0.024 0453* i 0.017
Ki-67 Ll- TB 0.093 i 0.005 0.097 i 0.005 0.095 at 0.005
TE 0.090 i 0.005 0.095 3: 0.005 0.088 1 0.005
PZ- TB 0.460 t 0.019 0.456 3: 0.017 0.460 i 0.018
TE 0.450 t 0.019 0.460 t 0.017 0.465 1- 0.018

 

' Whole colonic crypt proliferation Indices were determined before (TB) and after (TE)
consumption of beverage supplements for one year.

2 SP1 — Soy Protein Isolate.

3 Pooled values after combining soy beverage treatments +/- calcium.

4 Crypt Height — total number of cells in the hemicrypt from the base to the mouth of the
crypt prior to consumption of the dietary treatment. The values presented in this table
are averaged for the TB and TE time points within each treatment because no signiﬁcant
effects of time or treatment were detected.

5 LI — Total Crypt Labeling Index, obtained by dividing the number of labeled cells in a
hemicrypt by crypt height.

6 P2 — Proliferation Zone, obtained by dividing the position of the highest labeled cell
from the hemicrypt base by the crypt height.

7 Values are presented as means i SEM.

Within a treatment and parameter, an asterisk (*) indicates a Signiﬁcant difference (P <
0.05) between TE vs. TB values. A cross (1') indicates a trend (P < 0.10) for a difference
between TE vs. TB values.

122

 

Table 12. PCNA compartmental proliferation indices”7 in colonic crypts of subjects

 

 

Dietary Beverage Supplement

 

Casein SP12 plus SPI minus SP1 +/-
Calcium Calcium Calcium3
(n= 13; (n=16; (n=13; (n=29)

PCNA Indices mean age=58 ) mean age =59) mean age=59)

 

Basal LI4- T13l 0.523 i 0.014 0.549 i 0.013 0.574 :t 0.014 0.560 t 0.010
T121 0574* i 0.014 0.5091 1' 0.013 0.5531’ i 0.014 0529* i 0.010

Mid LIS- TB 0.318 i 0.035 0.318 i 0.031 0.360 2% 0.035 0.337 i 0.024
TE 0.252 :t 0.035 0192* i- 0.031 0.298 1- 0.035 0239* i 0.024

Top LI6- TB 0.015 i 0.009 0.041 i 0.008 0.012 i 0.009 0.028 i 0.006
TE 0.021 1- 0.009 0.005* d: 0.008 0.004 i 0.009 0004* i 0.006

 

' PCNA compartmental proliferation indices in colonic crypts were determined before
(TB) and after (TE) consumption of beverage supplements for one year.

2 SP1 — Soy Protein Isolate.

3 Pooled values aﬁer combining soy beverage treatments i calcium.

4 Basal LI- Labeling Index in Basal '/3'd compartment of the crypt.

5 Mid LI — Labeling Index in Middle 1/3“1 compartment of the crypt.

6 Top LI — Labeling Index in Top “/3” compartment of the crypt.

7 Values are presented as means i SEM.
Within a treatment and parameter, an asterisk (*) indicates a signiﬁcant difference (P <
0.05) between TE vs. TB values. A cross (‘1') indicates a trend (P < 0.10) for a difference
between TE vs. TB values.

123

 

Table 13. Effect of Casein, SPIl plus Calcium and SP1 minus Calcium on the mean
beginning and ending relative PCNA (Slot-blot methodology) protein levels4 in colonic

mucosa

 

 

SPI SPI SP1 +/-
Casein plus Calcium minus Calcium2
Calcium
Indices (n = 13) (n = 16) (n = 13) (n = 29)
TB“ 1215:37 1223:33 1230:37 1226:24
T53 1200:37 1095*:33 1128*:37 1110*:24

 

' SPI — Soy Protein Isolate.

2 Pooled values after combining soy beverage treatments i calcium.

3 PCNA protein expression in colonic mucosa was determined before (TB) and after (TE)
consumption of beverage supplements for one year.

4 Values (these are relative values based on densitometry per 20 pg mucosal protein) are

presented as means i SEM.
An asterisk (*) indicates a Signiﬁcant change (P < 0.05) for TE vs. TB values.

124

 

 

CHAPTER V. EFFECT OF SOY FLAKES, SOY FLOUR, SOY CONCENTRATE,
GENISTEIN AND CALCIUM ON ABERRANT CRYPT FOCI DEVELOPMENT
AND COLONIC EPITHELIAL CELL PROLIFERATION IN RATS INJECTED

WITH AOM
A. ABSTRACT
Increased colonic epithelial cell proliferation and delayed terminal cell
differentiation are commonly associated with increased risk for colon carcinogenesis.

The objective of this study was to investigate the chemopreventive potential of different

soy products against colon carcinogenesis in carcinogen-treated rats as reﬂected by

changes in aberrant crypt foci (ACF) development and colonic epithelial cell
proliferation. Colon cancer was initiated in rats by injecting 15 mg/kg BW AOM
subcutaneously at 21 and 28 days of age. At 35 days of age, rats were placed on the
dietary treatments (n = 15/treatment). The dietary treatments contained soy with differing
degrees of processing: whole soy known as full fat soy ﬂakes (SF K), soy with fat soluble
components removed (de-fatted soy ﬂour, [SFL]) and soy concentrate which has the fat
soluble and aqueous-alcohol soluble components removed (soy concentrate [SC]). Other
dietary treatments included soy concentrate plus 150 mg/kg genistein (GEN), and soy
concentrate plus 5 g/kg calcium (CAL). Rats were fed these diets for twelve weeks, at
which point they were sacriﬁced, the entire colons removed for aberrant crypt foci and
proliferation analyses. Rats consuming SC had the greatest (P < 0.05) numbers of total

ACF, large ACF and total aberrant crypts among all treatment groups. Supplementation

of the soy concentrate diet with either genistein or calcium resulted in the lowest (P <

0.05) numbers of ACF, large ACF and total aberrant crypts. Cell proliferation was

assessed by PCNA and Ki-67 labeling index (LI) and proliferative zone (PZ), and

125

quantiﬁed PCNA levels in colon mucosa. Rats fed SF K had the lowest (P < 0.05) PCNA
LI and P2 (0.180, 0.342) and these indices were signiﬁcantly lower than those observed
in rats consuming SC (0.336, 0.509). The PCNA LI and P2 of rats consuming SF L
(0.203, 0.348), CAL (0.215, 0.388) or GEN (0.222, 0.389) were not statistically different
from each other but were Signiﬁcantly lower than that observed in the SC-fed rats. Rats
consuming SC also had the greatest (P < 0.05) Ki-67 LI and P2 values among all
treatment groups. Quantiﬁed PCNA protein levels (by Slot-blot methodology) in colonic
epithelitun of rats fed SFL or SF K were signiﬁcantly lower than that observed in rats
consuming SC, GEN, or CAL diets. All biomarkers of cell proliferation tested in this
study (PCNA and Ki-67 LI and P2, quantiﬁed PCNA protein) were positively correlated
(P < 0.05) with each other and with numbers of ACF. Overall, these results indicate that
one or more aqueous-alcohol soluble but not fat soluble phytochemicals present in soy is
responsible for the reduction in ACF and reduced colonic epithelial cell proliferation
observed in AOM-treated rats. Furthermore, the protective effects of soy ﬂakes and soy
ﬂour against AOM-induced ACF development were replicated by addition of 150 mg/kg

genistein or 5 g/kg calcium to the SC diet.

B. INTRODUCTION

Colon cancer (CC) is a serious health problem in most developed countries and is
the second most common cause of cancer mortality in the United States (American
Cancer Society, 2005). Asian populations have relatively low rates of breast, prostate,
and CC when compared with Western cultures. Asian diets are typically lower in total

and saturated fat and higher in dietary ﬁber than Western diets. Another Signiﬁcant

126

 

dietary difference between Asian and Western population is the consumption of soy-
containing foods. Published reviews of epidemiologic literature, by McKeown-Eyseen
and Bright-See (1984), Messina et al. (1994), Messina and Bennink (1998), and Spector
et al. (2003) have all examined the relationship between soy food intake and colorectal
cancer. The reviewers observed that very few of the ecological and case-control studies
reported a statistically signiﬁcant reduction in CC risk with soy consumption. Therefore,
it is safe to conclude, that while there is no compelling epidemiologic evidence to support
an inverse relationship between soy intake and risk of colorectal cancer, there exist a
need for studies that are designed speciﬁcally to determine if soy consumption lowers the
risk for CC.

Soybeans and soybean-based foods are a good source of genistein. Genistein and
other soy phytochemicals such as phytosterols, saponins, protease inhibitors and phytate
have been reported to have potential anticancer properties. Several studies (Shamsuddin
et al., 1989; Pretlow et al., 1992b; Pereira et al., 1994; Awad et al., 1997; Koratkar and
Rao, 1997; Kennedy et al., 2002) have demonstrated that feeding isolated soy
phytochemicals in large amounts reduces experimental CC in animals. In a series of
studies conducted by Bennink (2001) it was reported that diets containing either full-fat
soy ﬂour or defatted soy ﬂour reduced colon tumor incidence in male Fisher 344 rats,
whereas soy concentrate did not reduce colon tumor incidence. More recently a study by
Murillo et al. (2004) reported that supplementation with soy bean ﬂour signiﬁcantly
decreased the total number of ACF in the colon of rats injected with AOM, a 53%
inhibition when compared to the control group (rodent chow). However, some animal

studies indicate that soy-containing diets may actually increase CC risk (Rao et al., 1997;

127

Gee et al., 2000). McIntosh et al. (1995) compared the effects of various protein sources
(20g/ 100g diet) on induction of colorectal neoplasia by dimethyl hydrazine (DMH) in
male Sprague Dawley rats and observed numerically higher tumor incidence in rats fed
soy protein than in those fed whey or casein, but these effects were not statistically
different. Davies et a1. (1999) found that diets high in fat, low in calcium, and containing
high-isoﬂavone soy protein did not protect against ACF development in AOM-treated
Fisher 344 rats.

Much of the interest in genistein as a potential anti-carcinogen has been promoted
by its well-established anti-proliferative effects against cancer cells in vitro. Studies have
shown that genistein inhibits the growth of hormone-dependent cells from tumors of the
breast (Peterson and Barnes, 1991) and prostate (Peterson and Barnes, 1993), and also
inhibits proliferation of leukemia (Makishima et al., 1991), melanoma (Rauth et al.,
1997) and gastric cancer cells (Matsukawa et al., 1993; Yanagihara et al., 1993).
Genistein has been shown to block mitosis and stimulate apoptosis in colon cancer cell
lines (Booth et al., 1999; Wenzel etal., 2000). These ﬁndings have led to studies in
animal models that have attempted to demonstrate the anti-carcinogeic effect of puriﬁed
genistein or of soy products containing isoﬂavones. Pereira et al. (1994) observed a
protective effect of genistein (150mg/Kg diet) against induction of ACF (preneoplastic
lesions), which have been shown to correlate with the subsequent development of
adenomas and adenocarcinomas in rat distal colon. However, a study by Rao et a1. (1997)
reported that genistein (250mg/Kg diet) increased colon tumor numbers in AOM-treated

rats.

128

Kennedy et al. (2002) reported that feeding a soybean extract containing the
Bowman-Birk protease inhibitor reduced DMH induced colon carcinogenesis in mice and
rats. Koratkar and Rao (1997) found that a diet containing 3% soy saponins, when
administered after AOM injection, reduced the average number of ACF per colon in CF]
mice as compared to mice consuming the control diet. Investigators have examined the
effect of phytosterols on colonycyte proliferation in rats (Awad et al., 1997) and report a
normalization of cholic acid induced hyperproliferation of colonycytes in rats. Raicht et
al. (1980) examined the growth of methylnitrosurea induced tumors in rats fed 0.2% soy
phytosterols and reported a 60% reduction in tumor incidence with phytosterols. Thus,
several components in soy have been reported to reduce carcinogen-induced
precancerous lesions or tumors in the colon of animal models for human colon cancer.

Dysregulated epithelial cell proliferation within normal-appearing colonic mucosa
is a strong predictor of increased susceptibility to neoplasia (Lipkin, 1988). Measures of
colonic epithelial cell proliferation (Labeling Index and Proliferation Zone) have been
proposed as indices of colorectal cancer risk, or as surrogate endpoint biomarkers for
cancer chemoprevention studies (Scalmati and Lipkin, 1993; Alberts etal., 1994; Syngal
et al., 2000; Kelloff et al., 2004). The transition of colonic mucosal cells from
hyperproliferation to atypia/dysplasia is characterized by development of abnormal
nuclear and/or cellular shapes and development of colon crypts having larger cells, raised
crypt surfaces, and abnormal staining pattern. These foci of abnormal colon crypts have
been termed aberrant crypt foci (ACF), which are the primary pre-cancerous lesion found

in rodents treated with colon carcinogens (Bird et al., 1986).

129

 

Soy foods are one of only a few dietary sources of the isoﬂavone genistein.
Methods of processing can dramatically inﬂuence the concentrations of isoﬂavones and
other phytochemicals present in soy products. De-fatted soy ﬂour (obtained by hexane
solvent extraction of soy ﬂakes) contains signiﬁcant concentrations of isoﬂavones and
other phytochemicals. However, ethanol extraction of soy ﬂour to produce soy protein
concentrate removes the majority of the isoﬂavones and oligosaccharides that were
originally present in soy ﬂour (US. Department of Agriculture, 1999).

We hypothesize that the inhibitory effects of soy-containing diets with
phtochemicals against colon cancer promotion are associated with reduced colonic
epithelial cell proliferation and reduced development of ACF. Therefore, the objective of
this study was to determine if soy-containing diets having differing concentrations of
phytochemicals (especially isoﬂavones) can reduce ACF and indices of colonic epithelial
cell proliferation (as assessed by PCNA and Ki-67 immunohistochemistry in colonic
tissue and by PCNA protein quantiﬁcation by immunoblotting) in AOM-initiated rats. A
further objective was to determine if changes in colon epithelial cell proliferation patterns
(as inﬂuenced by dietary treatments) correlate with ACF formation in rats treated with
AOM. Our ﬁnal objective was to determine if dietary calcium concentration (1 or 5 g/kg
of diet) would inﬂuence ACF development and markers of colonic epithelial cell

proliferation.

130

 

C. MATERIALS AND METHODS '
Materials
Azoxymethane was purchased from Ash-Stevens, Inc (Detroit, MI). Full fat soy
ﬂakes, defatted soy ﬂour and soy concentrate were obtained from Central Soya (Ft
Wayne, IN). Genistein was obtained from Dr. Muralee Nair at Michigan State University
and was synthesized as described previously (Chang et al., 1994). Corn starch and
sucrose were obtained from Michigan State University Food Stores. All other dietary

ingredients were obtained from Dyets Inc. (Bethlehem, PA).

Animals
Male Fischer 344 rats were obtained at 3 weeks of age from Harlan Sprague-
Dawley Inc (Indianapolis). The rats were housed in plastic cages (3 rats/cage) with wood
chip bedding. The animal room was maintained at 21-22 °C and 45% i 10% relative
humidity with a 12-h light-dark cycle. Diets and water were provided ad libitum. This
study was approved by the Michigan State University All-University Committee on

Animal Use and Care.

Diets
All diets used in this study were based on the AIN-93G diet (Reeves et al., 1993)
and the compositions of diets are listed in Table 14. The basal diet was formulated to
maximize the promotion of colon tumorigenesis. The calcium content of AIN-93G diet is
5 g/kg to allow for maximum bone and teeth mineralization in rapidly growing rodents.

This concentration of calcium can have an anti-cancer action in the colon as indicated by

131

a decrease aberrant crypt foci and colonic proliferation indices in rats that were injected
with AOM (Beatty et al., 1993; Li et al., 1998; Pierre et al., 2003). Therefore, we
formulated all diets except the SC+Cal diet to be low (1 g/kg) in dietary calcium so that
the potential anti-cancer action of other dietary ingredients would not be masked by the
dietary calcium levels present in standard AIN-93G diet. An earlier study conducted by
Pereira et al. (1994) reported that 150 mg/kg of genistein added to the AIN-93G diet
signiﬁcantly reduced the number of aberrant crypt foci (by 34% relative to controls) in
F344 rats injected with AOM. Therefore, we added genistein (150 mg/kg) to the soy
concentrate diet to determine if this concentration of genistein would inhibit ACF to the
same extent as the soyﬂour diet.

All diets were formulated to contain approximately 14.8% fat and 20% protein
from the various soy preparations. The concentrations of essential nutrients (except
calcium) were increased by 13% due to the increased energy density of these diets
compared to the standard AIN-93G diet. Diets contained 1 g/kg calcium except for the
SC+Cal diet, which contained 5 g/kg calcium. The ratios of protein, minerals (with the
exception of calcium), vitamins and other essential nutrients to energy were similar to the

standard AlN-93G diet (Reeves et al., 1993).

Experimental design
Each dietary treatment group consisted of 15 rats. All rats, except negative
controls, were injected subcutaneously with azoxymethane (15 mg/kg body wt) at 21 and
28 d of age. Because the time of day when azoxymethane injections are made is

important (Pereira et al., 1994), the injections took place between 0800 and 1000. The

132

 

 

negative control group received saline injections instead of azoxymethane. All rats were
fed the soy concentrate diet from days 21 to 35 days of age and then were switched to
their respective dietary treatments at 35 d of age, and were killed 12 wk later by

overexposure to carbon dioxide followed by exsanguination.

Methods
Isoflavone Analysis
Isoﬂavones were extracted from the soy ﬂakes, soy ﬂour, and soy concentrate by

mixing duplicate 1 g samples of each with 20 mL of 0.1 mol/L hydrochloric acid. Eighty
milliliters of methanol was added and the mixture was sonicated for 20 min, left at room
temperature for 2 h, and gravity ﬁltered with Whatman no. 1 ﬁlter paper (Whatman,
Clifton, NJ). The ﬁltrate was centrifuged at 10,000 X g and the isoﬂavones in the
supernate were separated and quantiﬁed by HPLC. The isoﬂavones were separated with a
reversed-phase column (Microsorb-MV, 25 X 0.46-cm column packed with C-18 Silica
particles of 5-mm diameter and 100-A pore size; Rainin Instrument Co, Woburn, MA) by
using a gradient mobile phase. Solvent A was 10% methanol, 89.9% water, and 0.1%
acetic acid; solvent B was 99.9% methanol and 0.1% acetic acid. The amount of solvent
B was increased linearly from 20% at 0 min to 30% at 2 min and then to 70% at 30 min.
The isoﬂavones were detected at 260nm and quantiﬁed by the external standard

technique.

133

 

Pathology and histology

After rats were sacriﬁced, the colons were removed, cut open longitudinally, and
rinsed in phosphate-buffered saline (PBS, pH 7.4). One cm of the colon (3 cm from the
anal verge) was excised and pinned ﬂat to a cardboard, ﬁxed in B5 ﬁxative for one hour,
placed between foam pads in tissue cassettes, and processed by routine parafﬁn
embedding and sectioning for subsequent IHC. A section of the colon (proximal) was
scraped with a microscope slide to obtain mucosal cells, which were frozen immediately
on dry ice and saved for Slot-blot quantiﬁcation of PCNA. The remaining length of the
colon was ﬁxed in 2% paraformaldehyde-PBS for 1 hour at 4°C and saved for ACF

quantiﬁcation.

Aberrant Crypt Foci analyses
Fixed colons were stained with methylene blue (0.1% in PBS) for 3 min as

previously described (McLellan and Bird, 1988), rinsed in PBS, and stored in 0.4%
formaldehyde-PBS (pH 7.4) at 4°C. AOM-induced lesions appear as darkly stained,
enlarged, and Slightly elevated crypts that can be distinguished from the lightly stained
normal mucosa. The lesions may appear as single aberrant crypts or as foci containing
two or more aberrant crypts. ACF and the number of aberrant crypts per focus were
visualized by stereomicroscopy. The total number of ACF, the numbers of large ACF
(i.e. ACF that contained 4 or more aberrant crypts) and the total number of aberrant
crypts per rat were recorded. Foci with 10 or more aberrant crypts (too many aberrant

crypts per focus to count accurately) were recorded as foci containing more than 10

134

 

 

aberrant crypts. In summing the total numbers of aberrant crypts per rat, we assigned a

value of 10 aberrant crypts to ACF that contained ten or more aberrant crypts.

PCNA and Ki-67 IHC analyses

Immunohistochemistry analyses were conducted on a sub-set of the rats from
each treatment (11 = 10 rats/treatment for No AOM, soy ﬂour, soy concentrate and soy
concentrate plus genistein treatments; n = 9 rats for the soy concentrate plus calcium
treatment; n = 11 rats for the soy ﬂakes treatment). Three-micron sections of paraffin-
embedded biopsy samples were cut and mounted on poly-L-lysine coated glass slides and
dried at 58 °C for 2 hours. Sections were cleared of parafﬁn with xylene, hydrated in
increasing ratios of waterzethanol, and treated with iodinezpotassium iodide and 5%
sodium thiosulfate to remove mercury. Samples were then subjected to antigen retrieval
(10 mMol/L citrate buffer, pH 6.0, 95 °C) for 20 minutes before immersion in 0.3%
hydrogen peroxide for 10 minutes to quench endogenous peroxide activity. Following
rinses in tris-buffered saline (TBS), sections were then incubated with primary antibody
(1:100 dilution of PC10 [anti-PCNA] or MM] [anti- Ki-67] respectively in 1% BSA in
Tris buffered saline, pH 7.4) overnight at 4° C. After subsequent incubations in
biotinylated rabbit anti-mouse immunoglobulins (diluted 1:100, 45 min), followed by
peroxidase-conjugated streptavidin (diluted 1:300, 45 min), sections were treated with 3-
amino-9-ethylcarbazole (ABC) and color development was monitored for 10—20 minutes
until red staining was evident. Slides were then counter-stained with hemotoxylin and
cover-Slipped using Faramount mounting medium. Appropriate positive (human tonsils)

and negative controls (omission of primary antibody) were used for evaluation of PCNA

135

 

and Ki-67 staining. The antibodies were obtained from Novacastra, and biotinylated
rabbit anti-mouse immunoglobulin, streptavidin, AEC, and Faramount were obtained
from Dako (Carpinteria, CA). The sections were rinsed twice with 10 mMol/L TBS in

between incubations with antibodies, streptavidin and ABC.

Quantiﬁcation of PCNA and Ki-67 Immunohistochemistry

PCNA and Ki-67 labeling were quantiﬁed by the percentage and positions of
positively stained (red) cells within full-length crypts in colonic sections. Well-oriented
longitudinally-sectioned full-length crypts (n = 10 per rat) were identiﬁed by light
microscopy. A scorable crypt was deﬁned as one in which the base touched the
muscularis mucosa and had an open lumen at the top. The number and position of PCNA
and Ki-67 labeled cells and crypt heights were recorded for ten full crypts for each rat to
determine cell proliferation parameters. All cells with a red nuclear staining were counted
as positive. The center-most cell at the base of the crypt, and in between the hemicrypts,
was designated as cell number one and the cell at the top of the crypt mouth was
considered the highest cell. A continuous column of cells from cell number one to the
highest cell were counted along each hemicrypt wall of each crypt. Crypt height was
deﬁned as the number of cells per hemicrypt.

Total Crypt Labeling Index (LI) was calculated by dividing the number of labeled
cells in a hemicrypt by the total number of cells in a hemicrypt (crypt height). These
values were averaged for at least 20 hemicrypts (representing 10 full-length crypts) to
obtain the L1 for each rat. Compartment Labeling Indexes (basal LI, mid LI and top LI)

were calculated by dividing the number of labeled cells within each third (basal, middle

 

136

 

and top) of each hemicrypt by the number of cells in each third of the respective

hemicrypt. Proliferation zone (PZ) was obtained by dividing the position of the highest
labeled cell from the hemicrypt base by the total number of cells in the hemicrypt. Each
of these PZ values were averaged for at least 20 hemicrypts to obtain the overall PZ for

each rat.

PCNA Measurement by Slot-blot

Slot-blot analysis was conducted on a sub-set of the rats from each treatment (11 =
10 rats/treatment for No AOM, soy ﬂour, soy concentrate and soy concentrate plus
genistein treatments; n = 9 rats for the soy concentrate plus calcium treatment; n = 11 rats
for the soy ﬂakes treatment). These samples were from the same rats that were used for
IHC determination of PCNA and Ki-67 expression. Mucosal samples were homogenized
with 0.5 ml of non-denaturing buffer containing protease inhibitors (20mMol/L MOPS,
150 mMol/L sodium chloride, 1% vol/vol Triton X-100, 1% wt/vol deoxycholate, 0.1%

va01 SDS, 1 mMol/L EDTA, 100 pMol/L sodium orthovandate, 250 pMol/L PMSF, 10

pg/ml leupeptin, 1 pg/ml pepstatin) and total protein concentrations were determined by
the biocinchoninic assay using BSA as the standard. Relative PCNA levels in mucosal
homogenates were then determined by immuno-quantiﬁcation using slot-blot
methodology. PVDF membranes (Immobilon P membranes; Millipore, Billerica, MA)
were wetted in methanol and then transferred to PBS. Twenty pg of protein from each
sample (dilutions in PBS) was vacuum blotted on to the PVDF membranes using a slot-
blot apparatus (Hoefer Scientiﬁc instruments). The membranes were held overnight at 4

°C in blocking buffer (3% BSA in TBS with Tween 20. The membranes were then

137

incubated for 2 hours with monoclonal PC-10 antibody (1:1000 dilution of primary
antibody in blocking buffer). After rinsing with wash buffer (TBS with Tween 20) the
membranes were incubated in biotinylated secondary antibody (1:1000 dilution of goat
anti-mouse IgG) in wash buffer for one hour. After subsequent rinses with wash buffer,
the membranes were treated with a chemiluminescence reagent (Dupont NEN research
products) and light emission was captured by autoradiography. Developed
autoradiography ﬁlms were quantiﬁed by scanning densitometry. Known amounts of
mouse IgG Fc fragments (standard curve ranging from 0.3125ng to 10.00ng) were

included in each blot to allow for normalization of staining intensity across blots.

Statistical Analysis

Aberrant crypt foci, crypt height, PCNA and Ki-67 proliferation indices (total
crypt LI, PZ, and compartmental LI for basal, middle, and top compartments of the
crypts) and relative PCNA quantiﬁed by the Slot-blot method each were statistically
analyzed as a completely randomized design (one-way Analysis of Variance) using the
General Linear Models procedure of SAS (version 6.11, SAS Institute Inc., Cary, NC).
When signiﬁcant diet effects were detected (P < 0.05), the least Signiﬁcant difference was
used to compare treatment means. Simple correlation coefﬁcients between aberrant crypt
foci, PCNA and Ki-67 labeling indices and the relative PCNA levels and were calculated
and tested for signiﬁcance at P < 0.05 using the correlation procedure of SAS. Rats from
the No AOM treatment were not used in the correlation analysis because they did not

develop ACF.

138

D. RESULTS
Isoﬁavone Composition of Diets

The full fat soy ﬂake-based diet (least processed) and defatted soy ﬂour-based
diet (obtained by hexane extraction of oil from full fat soy ﬂakes) both contained
approximately 490 mg/kg genistein derivatives (Table 15). These genistein derivatives
occurred primarily as glycosides in these diets. Both of these diets also contained
considerable concentrations of daidzein derivatives, with the soy ﬂake-containing diet
having 422 mg/kg daidzein derivatives and the soy ﬂour-containing diet having 334
mg/kg daidzein derivatives. Both of these diets also contained approximately 50 mg/kg
glycitein derivatives. The soy concentrate-based diet, as well as the soy concentrate plus
calcium diet, contained negligible amounts of isoﬂavones (Table 15). The soy
concentrate plus genistein diet contain 151 mg/kg of genistein, most of which was

directly added to the diet during its formulation.

Aberrant Crypt Foci
The effects of dietary treatment on preneoplastic lesions are Shown in Table 16,

which reports the total number of ACF, the number of ACF containing 4 or more aberrant
crypts, and the total number of aberrant crypts per rat. These results Show that rats
consuming diets containing 5 g/kg calcium or 150 mg/kg added genistein had the fewest
colon ACF (70 and 77 ACF, respectively). Rats consuming soy ﬂour and full-fat soy
ﬂakes had Signiﬁcantly fewer ACF than rats consuming soy concentrate (87 and 97
versus 132 ACF), but both groups had signiﬁcantly greater numbers of ACF than rats

consuming soy concentrate plus 5 g/kg calcium. Rats consuming the soy concentrate diet

139

 

 

 

had the greatest numbers of ACF (132 ACF/rat). No ACF were identiﬁed in colons from
the negative control rats that were injected with saline (data not shown).

Rats consuming the soy concentrate-based diet had the greatest (P < 0.05)
numbers of large ACF (those having 4 or more aberrant crypts) and the greatest number
of total aberrant crypts (676 aberrant crypts/rat). Rats consuming soy concentrate plus
either 150 mg/kg genistein or 5 g/kg calcium had the smallest numbers of large ACF (41
and 48 large ACF/rat, respectively). Rats consuming soy ﬂakes, soy concentrate plus
genistein, or soy concentrate plus calcium had the smallest total numbers of aberrant
crypts (397, 336 and 335 aberrant crypts/rat, respectively. Rats consuming soy ﬂakes had

intermediate numbers of large ACF (65/rat) and total aberrant crypts (475/rat).

Proliferation indices with IHC
Crypt Height
Data on crypt heights calculated during PCNA and Ki-67 immunohistochemistry
are presented in Tables 17 and 18, respectively. Among all treatment groups, rats
consuming soy concentrate and soy concentrate plus calcium had the greatest (P < 0.05)
crypt heights, with both groups averaging over 70 cells per hemicrypt. Rats consuming
soy ﬂour and soy concentrate plus genistein had the shortest crypt heights (P < 0.05),

with both groups averaging approximately 61 cells per hemicrypt.

Total Crypt Labeling Index (LI)
Data on PCNA labeling index in colonic epithelium are presented in Table 17.

Animals consuming the soy ﬂakes-based diet had the lowest (P < 0.05) total PCNA

 

140

 

labeling index (18%) among all groups. The greatest PCNA total LI values were
observed in rats consuming soy concentrate (33.6%), followed by rats consuming soy
concentrate without having AOM injections (No AOM group, L1 = 27.5%). Animals
consuming the soy concentrate plus genistein, soy concentrate plus calcium, and soy ﬂour
diets had statistically comparable total labeling indices (22.2, 21.5 and 20.3%,
respectively), and these LI values were Si gniﬁcantly lower than that observed in rats
consuming the soy concentrate diet.

Ki-67 total LI values are presented in Table 18. Among all treatment groups, rats
consuming soy concentrate had the greatest (P < 0.05) Ki-67 labeling index (13.9%).
Rats consuming all other diets as well as rats who were not injected with AOM had
statistically similar Ki-67 total labeling indices, with values ranging from 6.8% for rats

consuming soy ﬂakes to 8.6% for rats consuming soy concentrate plus calcium.

Proliferation Zone

PCNA proliferation zone data are presented in Table 17. The greatest (P < 0.05)
PCNA P2 was observed in rats consuming the soy concentrate-based diet (50.9%) and
the lowest PCNA PZ occurred in rats consuming soy ﬂakes (34.2%) or soy ﬂour (34.8%).
Addition of genistein or calcium to the soy concentrate diet reduced PCNA PZ
signiﬁcantly when compared to rats consuming soy concentrate alone. Rats consuming
soy concentrate without AOM injection (no AOM) had greater (P < 0.05) PCNA PZ
(44.6%) than all treatments except for soy concentrate (with AOM injection).

Ki-67 PZ (Table 18) was greatest (P < 0.05) in rats consmning soy concentrate

(40.6%) and lowest in rats consuming soy ﬂakes (23.3%) or soy ﬂour (24.8%). Rats

141

 

 

consuming soy concentrate plus genistein (29.2%), soy concentrate plus calcium (29.7%)

or soy concentrate without AOM injection (29.8%) all had intermediate Ki-67 PZ values.

C ompartmental Labeling Indices

Data on compartmental labeling indices (basal, middle and top LI values with the
crypts) for PCNA and Ki-67 are presented in Tables 17 and 18, respectively. PCNA
labeling in the basal third of the crypt was greatest (P < 0.05) in rats consuming soy
concentrate either with (60.1%) or without (58.4%) AOM injection. Rats consuming soy
ﬂakes had the lowest (P < 0.05) basal PCNA LI (46.3%). PCNA L1 in the middle third of
the crypt was signiﬁcantly greater in rats consuming soy concentrate (40.7%) than for all
other treatments. Rats consuming soy ﬂakes (7.8%), soy ﬂour (9.4%) or soy concentrate
plus genistein had the lowest PCNA labeling indices in the middle third of the crypts.

Ki-67 L1 in the basal third of the crypt was greatest in rats consuming soy
concentrate (29.0%), lowest for rats consuming soy ﬂakes (19.2%) or soy concentrate
without AOM (19.8%), and was intermediate for the other treatments. Rats consuming
soy concentrate had greater (P < 0.05) Ki-67 L1 in the middle third of the crypts (12.8%)
than all other treatment groups except for rats consuming soy concentrate plus calcium
(7.4%). PCNA and Ki-67 labeling in the top third of colonic crypts was minimal and did

not differ among the treatment groups.

Quantiﬁed PCNA Levels
The results of Slot-blot quantiﬁcation of PCNA protein levels in colonic mucosa

are presented in Table 19. Rats fed soy ﬂakes or defatted soy ﬂour had the lowest (P <

142

 

 

0.05) relative PCNA levels when compared to all other groups. Rats consuming soy
concentrate had signiﬁcantly greater PCNA expression than was observed in all other
treatment groups. Feeding soy concentrate without AOM injection (no AOM group)
resulted in greater (P < 0.05) PCNA expression than was observed in all AOM-injected
groups except for rats consuming soy concentrate. Animals on soy concentrate plus
genistein and on soy concentrate plus calcium had Similar relative PCNA levels (666 and
659), although these values were signiﬁcantly greater than was observed in rats

consuming soy ﬂakes or defatted soy ﬂour.

Correlations among Biomarkers

Table 20 Shows the correlation coefﬁcients for ACF, quantiﬁed PCNA levels,
PCNA total crypt LI and P2, and Ki-67 total crypt LI and P2. These correlations only
i ncl uded rats that were injected with AOM, so the data for the sham-injected control rats
were excluded from this analysis. All correlations among these parameters were
Statistically signiﬁcant. PCNA total crypt LI and PZ were highly correlated (F092).
Quantiﬁed PCNA protein levels also were highly correlated with total crypt PCNA LI
(@082) and PZ (r=0.70). Ki-67 total crypt P2 was also highly correlated with quantiﬁed
I)C1\IA (r=0.74), total crypt PCNA LI (r=0.68), total crypt PCNA PZ (r=0.64) and total

Crypt Ki-67 Pz (r=0.73).
E- DISCUSSION

The major goal of this research was to determine whether diets containing soy

p1”()(iucts (with varying levels of isoﬂavones and other phytochemicals) would inhibit the

143

early stages of AOM-induced colon cancer in F344 rats. Processing methods, such as
alcohol extraction to produce soy protein concentrate, dramatically inﬂuence the
phytochemical content in the resulting soy products. Four diets were designed to study
the effect soy-based diets differing in isoﬂavone and other phytochemical contents on
their potential impact on colonic epithelial cell proliferation and aberrant crypt foci
formation. We also tested the effect of calcium concentration in the diet by including a
treatment that contained 5 g/kg calcium, as compared with the low calcium level (1 g/kg)
in the other treatment diets.
Colon carcinogenesis is characterized by increased colonocyte proliferation and
decreased differentiation and apoptosis. The level of expression of certain proliferation
or differentiation markers (such as PCNA and Ki-67) as well as the location of their
expression within colonic crypts becomes important in assessing the status of
preneoplasia (formation of aberrant crypt foci) and neoplasia of the colon (Yamashita et
a1 - , 1994). Detection of PCNA in the nucleolus of a colonocyte indicates the cell’s
capacity to go through another round of cell division or, in other words, a lack of terminal
di fferentiation. As demonstrated in other studies (Deschner, 1988; Mills et al., 1995;
Akedo et al., 2001), an upward shift in the proliferation zone in conjunction with overall
itler‘eases in total cell proliferation in the colonic crypt is a useful intermediate biomarker
of increased CC risk.
The cell concentration of PCNA is directly correlated with the proliferative state
of the cell, increasing during G1, peaking at the Gl/S-phase interface, decreasing during
G2 and reaching low levels in M- phase and interphase (Prelich et al., 1987). The

pre Sence of PCNA in G0 -phase cells would indicate that these cells maintain the capacity

144

to go through another round of cell division. Terminally differentiated cells do not
express PCNA. Thus, the absence of PCNA indicates that a cell has undergone terminal
differentiation (Bromely et al., 1996). Hence, loss of PCNA expression appears to be
potentially useful as a differentiation marker of colonic epithelial cells. F urtherrnore,
detection and quantitation of PCNA content can represent proliferative status in the
colonic crypts. For example, a dietary treatment that results in an upward Shift in the
PCNA PZ within colonic crypts suggests that terminal cell differentiation is delayed by
this treatment. Ki-67 is a cell proliferation marker that is primarily expressed during the
S -phase of cell cycle and absent in Go while PCNA accumulates as the cells progress
through G], S, and M phases of the cell cycle and returns to low levels during Go. While
Ki-67 labeling provides an estimate of total cell proliferation in tissues (Bruno and
Darzynkiewicz, 1992; Zacchetti et al., 2003) unlike PCNA, it does not necessarily reﬂect
a cell’s capacity to undergo another round of cell division (Prelich et al., 1987; Toschi
and Bravo, 1988; Zacchetti et al., 2003).
In our study the rats that were fed the soy concentrate diet, after initiation of colon
Cal‘scer by AOM, had the largest number of total ACF, large ACF and total aberrant
ct’5Ipts (P<0.05) among all treatment groups. The rats consuming soy concentrate also
had the greatest PCNA and Ki-67 total LI and P2 values among all groups. Conversely,
ITits fed soy ﬂakes, soy ﬂour or soy concentrate plus 150 mg/kg genistein all had
Sigrn'ﬁcantly lower Ki-67 and PCNA LI and PZ values than rats consuming soy
concentrate (Tables 17 and 18). Likewise, rats fed the soy ﬂakes, defatted soy ﬂour and
$03, concentrate plus genistein all had signiﬁcantly lower numbers of ACF, large ACF

and total aberrant crypts when compared to the soy concentrate-fed rats. It is also

145

noteable that addition of 150 mg/kg genistein to the soy concentrate diet was sufﬁcient to
reduce ACF development in rats to an even greater extent than was observed by feeding
soy ﬂakes or soy ﬂour.

Extraction of soy ﬂakes and soy ﬂour with aqueous ethanol to produce soy
concentrate results in loss of phytochemicals such as isoﬂavones (genistein and daidzein),
Bowman-Birk protease inhibitor, and saponins (Wang and Murphy, 1996; US.
Department of Agriculture, 1999). As shown in Table 15, the soy concentrate diet

contained minimal concentrations of isoﬂavones when compared to either the soy ﬂake
or defatted soy ﬂour diet. This may account for the greater promotion of ACF and
greater PCNA and Ki-67 LI and PZ values observed in the soy concentrate-fed group.
The presence of high PCNA mitotic indices and upward shift 0f PZ observed in rats fed
the soy concentrate indicates that rats consuming this diet had a substantial delay in
reminal differentiation of colonic epithelial cells. It also was interesting to observe that
extracting the oil- and fat-soluble components (such as sterols that has been postulated to
pro tect against CC, (Awad et al., 1997) from soy ﬂakes to produce soy ﬂour did not alter
Potential of soy ﬂour to reduce ACF or lower PCNA and Ki-67 LI and P2 indices.
Studies in laboratory animals have Shown that feeding a soy-rich diet inhibits
Che Imically induced carcinogenesis (Pereira et al., 1994; Koratkar and Rao, 1997;
Kenmedy et al., 2002). Hakkak et al. (2001) fed male Sprague Dawley rats AIN-93G
diet 8 formulated with either soy protein isolate or casein as the sole protein source for
their entire life. They observed that rats fed a diet containing SPI had reduced AOM-

lnduced colonic tumor incidence when compared with rats fed a casein-containing diet.

146

Rats fed the casein diet had a 50% incidence of colon tumors compared with 12% on soy
protein based diet.

Genistein, a major isoﬂavone in soy, has been demonstrated to have anti-cancer
effects in many cancer cell culture studies (Yanagihara et al., 1993). Little research has
been reported that conﬁrms a protective effect of isoﬂavones against carcinogenesis in
animal models. Pereira et al. (1994) found that 150 mg/kg of dietary genistein inhibited

the formation of ACF by 34% in the colons of rats treated with azoxymethane when
compared to rats consuming the Standard AIN 93-G diet.

In our study, adding the isoﬂavone genistein to the soy concentrate diet caused
signiﬁcant decreases in the number of ACF, the number of large ACF, and the total
I) timber of aberrant crypts when compared to the soy concentrate-fed rats (Table 16).
B 0th the soy ﬂour-based diet and the full fat soy ﬂake-base diets contained 490 mg/kg
genistein derivatives (primarily as glycosides), but the soy ﬂake diet was less effective,

and the soy ﬂour diet was equally effective, in inhibiting the formation of total ACF
compared to when 150 mg/kg genistein (as the aglycone) was added to the soy

Con centrate diet. The genistein-containing diet was more effective than either soy ﬂakes
01‘ soy ﬂour in inhibiting the formation of large ACF. These ACF results suggest that the
ag l ycone genistein was more effective in preventing early CC promotion than genistein
when present predominantly as glycosides. Feeding pure genistein compared with
geni stin at equimolar doses would unequivocally answer the question of whether

gen i stein glycosides have less efﬁcacy in inhibiting early CC promotion when compared
to the aglycone form. Although genistein has been shown to inhibit the appearance of

prec ancerous lesions in the colon of rats, we do not know if this would translate into a

147

reduction in colon tumors in animals or humans. A study conducted by Rao et al. (1997)
demonstrated that adding genistein to the diet (250 mg/kg) actually increased AOM-
induced colon cancer in rats.

The LI and P2 (both PCNA and Ki-67, Tables 17 and 18) of soy ﬂour, calcium
and genistein fed rats were not statistically different from each other and were lower (P <
0.05) than the L1 and PZ values observed in rats fed soy concentrate. Adding 150 mg/kg

of the isoﬂavone genistein (as the aglycone) to the soy concentrate diets caused a
signiﬁcant decrease in the proliferation indices when compared to feeding soy
concentrate alone. In fact, the PCNA LI and P2 observed in rats fed soy ﬂour, calcium
and genistein (all of which were injected with AOM) all were lower than was observed
even in the sham-injected (no AOM) rats, which were fed the soy concentrate-based diet.
The Ki-67 LI and PZ were similar for all 4 of these treatment groups.

The standard AIN-93G diet has adequate calcium content (5 g/kg) to allow
maximum bone and teeth mineralization in rapidly growing rodents. However, dietary
Cal cium content can have anticancer action in the colon. Others have shown that calcium
Supplementation to the already high-calcium content of AIN-93G diets decreases AOM-
iIlciiaced ACF and tumors in rats (Beatty et al., 1993; Li et al., 1998; Wargovich et al.,
20 O 0; Pierre et al., 2003). Some clinical trials show that supplemental calcium (1 .25 g)

m ay reduce the risk of CC by decreasing cell proliferation in the upper 40% of colon
crB’Iats in people at immediate or high risk for developing CC (Lipkin and Newmark,
198 5; Rozen et al., 1989; Wargovich et al., 1995a). However, 3 other studies examining
calc ium and CC risk did not ﬁnd a reduction in colonic mucosa proliferation rates or a

downward shift in the proliferative compartment with calcium supplementation (1.2 —

148

1.5g, Stern et al., 1990; Baron et al., 1995; Cats et al., 1995). Even though the human data
are less supportive of the hypotheses that high-calcium diets have a protective effect
against CC development, we intentionally kept the dietary calcium content low (all diets
contained 1 g/kg calcium except the ‘calcium’ diet which was the soy concentrate diet
with 5 g/kg calcium) so that the anticancer activity of other dietary ingredients (such as
the soy phytochemicals) would not be masked by dietary calcium amounts used in the
standard AIN-93G diet.
Feeding 5 g/kg calcium to rats lowered (P < 0.05) the proliferative indices (both
PCNA and Ki-67 Ll & PZ) when compared to the soy concentrate-fed rats (whose diet
contained 1 g/kg calcium). The soy concentrate plus calcium diet also had signiﬁcantly
fewer total ACF, large ACF and total aberrant crypts when compared to feeding the soy
c cancentrate diet (which contained 1 g/kg calcium). This effect of adding calcium to the
soy concentrate diet on ACF development and proliferation indices was similar to the
effect of adding genistein to the soy concentrate diet. When Pereira et al. (1994) added
1 5 0 mg/kg genistein to AIN-93G diets containing 5 g/kg calcium, the number of ACF
Was decreased by 34% compared to rats consuming standard AIN-93G diets. In this
Stu dy, the soy concentrate plus genistein diet contained 150 mg/kg genistein and 1 g/kg
cal cium and we observed a 42% reduction in the number of ACF with this diet compared
to rats consuming soy concentrate (with 1 g/kg calcium). Thus, the response of AOM-
ind uced ACF formation to dietary genistein observed in our study was Similar to that
Obs erved by Pereira et al. (1994), although these two studies utilized control diets
diff‘ering in protein source (soy concentrate versus casein) and calcium level (1 g/kg in

our study and 5 g/kg in the study by Pereira et al., 1994). Given our treatment protocol,

149

we were unable to determine if the effects of dietary genistein and calcium might be
additive. Testing the potential additivity of these effects Should be the focus of a future
study. We also are unable to speculate whether diets based on soy ﬂakes or soy ﬂour
would reduce ACF formation when the diet contains a higher calcium level (i.e. 5 g/kg as
in the standard AIN-93G diet). .

The quantity of PCNA protein in colonic mucosa represents a potentially useful
marker for the study of mucosal proliferative activity and therefore, colon cancer risk.
Using slot-blot methodology, we quantiﬁed the PCNA protein levels in the colonic

mucosa of rats in this study. The results obtained from this method indicate that relative
amounts of PCNA quantiﬁed by Slot-blotting reﬂect the dietary induced differences we
0 lrserved by PCNA immunohistochemistry (Tables 17 and 19) results. Rats fed soy ﬂour
and soy ﬂakes had the lowest (P < 0.05) levels of quantiﬁed PCNA (slot-blot method)
among all treatments, whereas rats fed soy concentrate had the greatest (P < 0.05)
quantiﬁed PCNA levels. Furthermore, quantiﬁed PCNA was highly correlated with
PC NA and Ki-67 LI and P2 values (Table 20), demonstrating that Slot-blot quantiﬁcation

Of PCNA represents another potentially useful method to evaluate the efﬁcacy of diet to

lnﬂ Irence colon cancer risk.

In overall conclusion, this study indicates that diets based on soy ﬂour and soy

ﬂak es, as compared with diets based on soy concentrate, are protective against AOM-
ind uced ACF formation. Although the particular phytochemicals present in soy ﬂour and
soy flakes that are responsible for this protective effect were not identiﬁed in this study,

we demonstrated that we could achieve similar levels of protection against ACF

fortl'lation by supplementing a soy concentrate-based diet with 150 mg/kg genistein.

150

Supplementation of the soy concentrate-based diet with calcium to a level of 5 g/kg
calcium in the diet was able to exert a level of protection against ACF formation that was
comparable to that observed when feeding soy ﬂour, soy ﬂakes and soy concentrate plus
genistein. Reductions in ACF formation by diet were associated strongly with reductions
in whole crypt LI and P2 for both PCNA and Ki-67, and also were correlated with

signiﬁcant reductions in PCNA protein expression in colonic mucosa.

151

Table 14. Composition of dietsl

 

 

 

 

Treatment Groups
Ingredients Soy Soy flour Soy Soy concentrate
ﬂakes concentrate + additional
GEN2 CAL3
% by wt
Full Fat Soy Flakes 46.07 - - - -
Soy Flour - 35.79 - - -
Soy Concentrate - - 28.45 28.45 28.45
Genistein - - - 0.015 -
Calcium Carbonate - - - - 1.00
Soy Oil 4.95 14.72 14.78 14.78 14.78
Corn Starch 2.85 3.36 10.64 10.63 9.64

 

I A basal mixture comprised 46.13% of the diet and contained sucrose (40% of diet),

mineral mix (4.2% of diet providing 0.1% Ca by weight in the diet), vitamin mix
(1.13% of diet), methionine (0.23% of diet), L-cystine (0.34% of diet), choline bitartrate
( 0.23% of diet) and butylated hydroxytoluene (0.003% of diet).

GEN — Soy concentrate plus additional genistein (150mg/Kg diet)
CAL — Soy concentrate plus calcium to equal the standard AIN-93G mineral mix (5g
(3ng diet)

152

 

Table 15. Isoﬂavone concentrations in rat diets

 

 

 

 

Diets
Soy concentrate
Soy Soy and SC plus
Isoﬂavone ﬂakes ﬂour calcium Genistein
mg of aglycone/ kg of diet
Malonyl- 212 342 8 8
genistinl
Genistinl 235 142 8 8
Genistein 45 10 1 151
Total geinstein
derivatives 492 494 l 7 167
Malonyl- 126 196 7 7
daidzin+
Daidzinl 182 92 7 7
Daidzein 114 46
Total daidzein
derivative 422 334 19 19
Glycitinl 48 45 3 3
Glycitein 4 ND“ 1 1
Total glycitein
derivatives 52 45 4 4

‘

 

1—
* Expressed as mg of the aglycone, not the glycoside weight
* ND — none detected

153

Table 16. Effect of diet on preneoplastic lesions in the colon of rats treated with AOM

 

 

DietT Preneoplastic lesions“

ACF ACF 2 4 AC Total AC
SFK 97“ : 5 65“ : 3 475b : 24
SFL 87““ : 5 54“ : 3 397“ : 24
SC 132“:5 91“:3 676“:24
SC+GEN 77““ : 5 41“ : 3 336“ : 24
SC+CAL 70d : 5 48““ : 3 335“ : 24

 

I SFK — soy ﬂakes; SFL —- soy ﬂour; SC — soy concentrate; SC+GEN — soy concentrate
plus genistein; SC+CAL — soy concentrate plus calcium.

: ACF - Foci with one or more aberrant crypts/rat; ACF Z 4 AC — ACF with 4 or more
aberrant crypts/rat; total AC — the total number of aberrant crypts/rat.

* Values are presented as means and i SEM (n = 15 rats/treatment). Means in the same
column not Sharing a common superscript are signiﬁcantly different (P < 0.05).

154

 

Table 17. Effect of diet on colonic crypt height and epithelial PCNA proliferation
indices in rats treated with AOM

 

 

 

 

 

 

PCNA Indices,r
Diet“ CH§ Ll PZ
No AOM 66.6“ : 1.2 0.275“ : 0.008 0.446“ : 0.015
SFK 65.6““ : 1.1 0.180“ : 0.008 0.342“ : 0.014
SFL 60.8“ : 1.2 0.203“ : 0.008 0348““ : 0.015
SC 69.7““ : 1.2 0.336“ : 0.008 0.509“ : 0.015
SC+GEN 61.8d : 1.2 0.222“ : 0.008 0.389“ : 0.015
SC+CAL 72.4“ : 1.3 0.215c : 0.008 0.388“ : 0.016
PCNA IndicesT
Diet“ Basal LI Mid LI Top LI
No AOM 0584““ : 0.014 0.233“ : 0.021 0.007 : 0.002
SFK 0.463“ : 0.014 0.078“ : 0.020 0.000 : 0.002
SFL 0.530“ : 0.014 0094““ : 0.021 0.000 : 0.002
SC 0.601“ : 0.014 0.407“ : 0.021 0.000 : 0.002
SC+GEN 0545““ : 0.014 0125““ : 0.021 0.000 : 0.002
SC+CAL 0553““ : 0.015 0.150“ : 0.022 0.000 : 0.002

 

Values are presented as means and i SEM (n = 10 rats/treatment except n = 9 for
SC+CAL and n = 11 for SFK). Means in the same column not sharing the same
superscripts are Signiﬁcantly different (P < 0.05).

No AOM - vehicle treated; SF K — soy ﬂakes; SF L — soy ﬂour; SC — soy concentrate;
SC+GEN — soy concentrate plus genistein; SC+CAL — soy concentrate plus calcium.
CH — Crypt Height, total number of cells in the hemicrypt from the base to the mouth of
the crypt;

LI — Total Crypt Labeling Index, obtained by dividing the number of labeled cells in a
hemicrypt by crypt height.

PZ - Proliferation Zone, obtained by dividing the position of the highest labeled cell
from the hemicrypt base by the crypt height

Basal LI— Labeling Index in Basal one-third compartment of the crypt

Mid Ll — Labeling Index in Middle one-third compartment of the crypt

Top LI — Labeling Index in Top one-third compartment of the crypt

‘

155

Table 18. Effect of diet on colonic crypt height and epithelial Ki-67 proliferation indices
in rats treated with AOM

 

Ki-67 Indices?

 

 

 

 

 

 

 

Diet“ CH§ Ll PZ
No AOM 65.9“ i 1.1 0.076“ a 0.017 0.298“ a 0.020
SFK 66.4“ i 1.1 0.068“ i 0.016 0.233“ a: 0.191
SFL 61.1“ :I: 1.1 0.073“ i 0.017 0248““ i 0.020
SC 70.4“ i 1.1 0.139“ i 0.017 0.406“ at 0.020
SC+GEN 61.4“ i 1.1 0.083“ i 0.017 0.292“ :t 0.020
SC+CAL 72.4“ i 1.2 0.086“ i 0.018 0.297“ a 0.021
Ki-67 Indices“
Diet“ Basal LI Mid Ll Top LI
No AOM 0.198“ at 0.029 0.030“ a 0.023 0.000 a 0.001
SFK 0.192“: 0.028 0.013“ a 0.022 0.001 i 0.001
SFL 0208““ i 0.029 0.021“ a 0.023 0.000 i 0.001
SC 0.290“: 0.029 0.128“ a. 0.023 0.000 a 0.001
SC+GEN 0226““ i- 0.029 0.029“ :t 0.023 0.000 i 0.001
SC+CAL 0228““ a 0.030 0074““ a 0.025 0.000 : 0.001
1.

4.
+

§

Values are presented as means and i SEM (n = 10 rats/treatment except n = 9 for
SC+CAL and n = 11 for SFK). Means in the same column not sharing the same

superscripts are signiﬁcantly different (P < 0.05).
No AOM — vehicle treated; SFK — soy ﬂakes; SFL — soy ﬂour; SC - soy concentrate;
SC+GEN — soy concentrate plus genistein; SC+CAL - soy concentrate plus calcium.
CH - Crypt Height, total number of cells in the hemicrypt from the base to the mouth of
the crypt;
LI — Total Crypt Labeling Index, obtained by dividing the number of labeled cells in a
hemicrypt by crypt height.
PZ — Proliferation Zone, obtained by dividing the position of the highest labeled cell
from the hemicrypt base by the crypt height
Basal LI— Labeling Index in Basal one-third compartment of the crypt
Mid LI — Labeling Index in Middle one-third compartment of the crypt
Top LI — Labeling Index in Top one-third compartment of the crypt

156

Table 19. Effect of diet on quantiﬁed colonic PCNA levels in rats treated with AOM

 

 

Duﬂ PCNA“

No AOM 736“ :18
SFK 535“ :17
SFL 536“ :18
SC 1014“ :18
SC+GEN 666“ :18
SC+CAL 659“ :19

 

I No_AOM — vehicle treated; SF K — soy flakes; SFL — soy ﬂour; SC — soy concentrate;
SC+GEN — soy concentrate plus genistein; SC+CAL — soy concentrate plus calcium.

1 Quantiﬁed PCNA units are relative. Values are presented as means and i SEM (n = 10
rats/treatment except n = 9 for SC+CAL and n = 11 for SFK). Means in the same
column not sharing the same superscripts are signiﬁcantly different (P < 0.05).

157

 

Table 20. Correlation coefﬁcients for total ACF, quantiﬁed PCNA levels, PCNA LI and
P2 and Ki-67 LI and P2 in colonic mucosa of rats injected with AOM

 

 

ParameterT PCNA SB PCNA LI PCNA PZ Ki-67 LI Ki-67 PZ
Total ACF 0.51 0.56 0.52 0.34 0.43
P value 0.0002 <0.0001 0.0001 0.0143 0.0017
PCNA SB - 0.82 0.70 0.53 0.74
P value <0.0001 <0.0001 <0.0001 <0.0001
PCNA LI - - 0.92 0.38 0.68
P value <0.0001 0.0064 <0.0001
PCNA PZ - - - 0.34 0.64
P value 0.0165 <0.0001
Ki-67 LI - - - - 0.73
P value <0.0001

 

T Total ACF — total aberrant crypt foci; PCNA SB — PCNA quantiﬁed by slot blotting;
LI — Labeling Index; PZ — Proliferative Zone. Total N = 50; n = 10 rats/treatment except
n = 9 for SC+CAL and n = 11 for SF K. Rats from the No AOM treatment were not
used in this analysis.

158

 

CHAPTER VI. SUMMARY AND CONCLUSIONS

Colon cancer (CC) incidence as an end point in chemoprevention trials is
impractical due to the length of time required for CC to develop as well as the large
number of subjects required to test the efﬁcacy of various interventions. Thus, there is a
need to identify biomarkers that will reliably predict CC risk and beneﬁts of intervention
trials. The overall goal of this research was to determine the colon cancer chemo-
preventive potential of soy-containing foods in humans at risk for CC and in animals
initiated with a colon carcinogen using intermediate biomarkers rather than appearance of
polyps or cancer as the endpoint. Increased proliferation (as measured by tritiated
thymidine and BrdU labeling) and delayed terminal differentiation in colonic epithelial

cells are commonly found in individuals with increased risk for CC and in rodents treated
with carcinogens that induce CC.

The ﬁrst objective of the ﬁrst study was to determine the correlation between CC
ri 5k and indices of mucosal proliferation (LI and P2 as assessed by IHC detection of
P CNA and Ki-67). Given the laborious nature of IHC techniques used to measure
proliferation markers, there is a need to develop a more facile index of colonic epithelial

Q ell proliferation. A simpler method would be highly desirable in dietary intervention
and large scale clinical studies that evaluate the risk for CC. For this reason, the second
Objective of this research was to use an immunoblotting method to determine relative
levels of PCNA protein in colonic epithelial cells and determine if PCNA protein
Correlates with proliferation indices based on IHC techniques. The ﬁrst study
demonstrated that the proliferative compartment (measured by PCNA and Ki-67 PZ),

PCNA LI, and relative PCNA protein levels were positively correlated with increasing

159

age and CC risk and were therefore considered useful intermediate biomarkers to assess
CC risk. Ki-67 LI was judged to be not useful for assessing CC risk status.
The objective of the second study was to utilize proliferative indices (PCNA and
Ki-67 labeling and quantitative PCNA protein expression) as intermediate biomarkers to
test for reduction of CC risk in humans after soy consumption for one year. A double-
blind, prospective study was conducted to determine if consumption of one of two
supplements containing soy protein isolate with isoﬂavones or a supplement containing
calcium caseinate for one year would inﬂuence colon epithelial cell proliferation. The
two soy supplements differed only in their calcium contents. There were statistically
signiﬁcant reductions in both PCNA LI and PCNA PZ during the course of the
experiment in subjects that consumed soy supplements but not in subjects who consumed
casein. There were no signiﬁcant treatment effects on total Ki-67 L1 or Ki-67 PZ. The
relative expression of PCNA protein in colonic mucosa was signiﬁcantly reduced during
the course of the experiment in subjects who consumed either of the soy supplements, but
Was not inﬂuenced by consumption of the casein supplement. Since the presence of
P CNA indicates the capacity of cells to undergo cell division, the decrease in PCNA LI
and P2 observed in subjects consuming the soy supplements denotes a lowering of the
Q ell proliferation capacity and infers a greater degree of terminal cell differentiation of
Q olonic crypts in those subjects. This study indicated that consumption of supplements
Containing soy protein isolate caused changes in colon epithelial cell proliferation
patterns that are consistent with a reduced risk for developing colon cancer.
The objective of the ﬁnal study was to investigate the CC chemopreventive

potential of different soy products in carcinogen-treated rats as reﬂected by changes in

160

 

ACF and colon epithelial cell proliferation. Rats consuming soy concentrate had the
greatest numbers of total ACF, large ACF and total aberrant crypts among all treatment
groups. Supplementation of the soy concentrate diet with either genistein or calcium
resulted in the lowest numbers of these endpoints. All biomarkers of cell proliferation
tested in this study (PCNA and Ki-67 LI and PZ, quantiﬁed PCNA protein) were
positively correlated with each other and with numbers of ACF. The data from this study
indicates that one or more aqueous—alcohol soluble, but not fat soluble, phytochemicals
present in soy is responsible for the reduction in ACF and reduced colonic epithelial cell
proliferation observed in AOM-treated rats. Furthermore, the protective effects of soy
ﬂakes and soy ﬂour against AOM-induced ACF development were replicated by addition
of 150 mg/kg genistein or 5 g/kg calcium to the soy concentrate diet.

In conclusion, it was demonstrated that PCNA LI and PZ, Ki-67 PZ, and relative
concentrations of PCNA protein in colonic mucosa are highly correlated and are
correlated with risk of CC, which indicates that these biomarkers have utility as
intermediate biomarkers in chemoprevention studies. Consumption of a soy protein
isolate-based supplement for one year reduced risk of CC in humans and feeding soy

ﬂakes and soy ﬂour reduced biomarkers of CC in a rat model of human CC.

161

 

 

CHAPTER VII. RECOMMENDATIONS FOR FUTURE STUDIES

The following studies should be conducted to further assess the potential efﬁcacy

of soy products and speciﬁc soy phytochemicals to reduce colon cancer risk.

A. Examination of the potential efﬁcacy of soy protein isolate, mixed soy
isoﬂavones, and genistin to reduce indices of colon cancer risk (PCNA LI and PZ,
PCNA protein expression) in colonic mucosa of humans at high risk for colon
cancer.

In this dissertation research, I demonstrated that consumption of soy protein
isolate-based supplements for one year by subjects at high risk for colon cancer resulted
in reduced indices of colonic epithelial cell proliferation (Chapter 4). Although the
supplement treatments used in this research were similar in their contents of protein and
most essential nutrients, they did differ signiﬁcantly in calcium concentration. The soy
protein isolate supplement containing calcium was more effective in reducing indices of
cell proliferation than was the soy protein isolate supplement without additional calcium.
Thus, the effects of soy could not be completely separated from potential effects of
calcium in this research. This experiment should be repeated with a similar group of
hi gh-risk individuals who consume supplements designed to directly assess the effects of
soy protein isolate, isoﬂavones, and genistin on indices of colonic epithelial cell
proliferation. The speciﬁc effects of genistin are particularly interesting because this
research also demonstrated that 150 mg of aglycone genistein/kg diet added to a soy
concentrate-based diet signiﬁcantly reduced indices of colonic epithelial cell proliferation

in AOM—treated rats (Chapter 5). The supplement treatments (40 g protein/day) would

I62

 

 

consist of l) soy protein isolate, 2) soy protein concentrate (prepared by alcohol
extraction and thus containing very low concentrations of isoﬂavones), 3) soy protein
concentrate plus mixed isoﬂavones added at concentrations to mimic those in soy protein
isolate, and 4) soy protein concentrate plus genistin at concentrations equal to those in
soy protein isolate. Subjects would consume these supplements in addition to their
regular diets for one year. Colonic mucosal biopsies would be obtained before initiation
of the experiment and after one year of supplementation. Biopsies would be used to

assess colonic epithelial cell proliferation, differentiation and apoptosis.

B. Validation of quantitative PCNA protein expression in human colonic mucosa as
a reliable preneoplastic biomarker for elevated colon cancer risk.

The research reported in Chapter 3 demonstrated that quantiﬁed PCNA protein
concentrations in colonic mucosal biopsies were signiﬁcantly increased in persons of
older age (having no history of colonic disease) and in persons who had a history of
adenomatous polyps, familial adenomatous polyposis, and colon cancer. A particularly
striking observation in this research was the ﬁnding that humans who had a history of
adenomatous polyps, regardless of age, had colonic epithelial cell proliferation patterns
comparable to the oldest subjects having no history of colonic disease. This observation
conﬁrms that colonic epithelial cell proliferation in normal-appearing colonic mucosa is
disregulated in persons who develop adenomatous polyps and more advanced colonic
neoplasia. This research also demonstrated that quantiﬁed PCNA protein concentrations
in colonic mucosa were highly correlated with established IHC markers of colon cancer

risk (PCNA LI and P2). This research should be extended to assess the potential of

163

 

quantitative PCNA protein expression to predict colon cancer risk in a prospective human
clinical study. Although the cost and logistics of conducting such a study solely for the
purpose of validating PCNA protein expression as a biomarker of neoplastic potential, is
unwarranted, the possibility of joining this research objective to an ongoing prospective

study of human colon cancer development should be pursued.

C. Determine if elevated intermediate biomarkers of colon cancer risk translate to
increased tumorigenesis in rodent models of human colon cancer.

The research reported in Chapter 5 demonstrated that intermediate biomarkers of
colon cancer risk (ACF, PCNA LI and PZ, quantiﬁed PCNA protein expression) were
highly correlated in an AOM-induced rat model of human colon cancer. However, the
validity of ACF and other intermediate biomarkers to predict colon cancer development
(adenomcarcinomas) remains in question. In order to determine the potential of
intermediate biomarkers to predict colon cancer risk, a wide range colon cancer incidence
and colon tumor burden must be produced by the dietary treatments. Thus, treatments
used in this experiment should include 1) high-fat (15%) AIN-93G control diet, and
isonitrogenous, isoenergetic diets containing the following protein and phytochemical
sources, 2) full fat soy ﬂakes, 3) defatted soy ﬂour, 4) soy protein isolate, 5) alcohol-
extracted soy concentrate, 6) diet 5 + mixed isoﬂavones to mimic their concentrations
found in diet 2, and 7) diet 5 + genistin to mimic its concentration in diet 2. These diets
would be administered to F 344 rats beginning one week after a second injection of AOM
(15 mg/kg body weight) and fed for a period of 40 weeks. Rats would be sacriﬁced after

10 weeks and 40 weeks of tumor promotion. Intermediate biomarkers of colon cancer

I64

risk would be assessed at the lO-week time point, and colon tumor incidence, burden and
histological grade would be determined at the 40-week time point. Intermediate
biomarkers (PCNA LI and PZ, quantiﬁed PCNA protein expression) also would be
assessed after 40 weeks of tumor promotion to assess any changes that occur in these

indices during the promotion phase.

D. Determine mechanisms of action whereby soy-containing diets and soy
isoﬂavones inﬂuence colonic epithelial cell proliferation, differentiation and
apoptosis.

Although several studies have demonstrated 1) that soy-containing diets can
inﬂuence colon cancer risk in animal models, and 2) that isolated soy phytochemicals
(e. g. isoﬂavones) can inﬂuence signaling pathways and growth of cancer cells in culture,
there remains a dearth of conclusive information on the mechanisms whereby soy-
containing diets and isoﬂavones inﬂuence colon cancer risk. Studies utilizing
carcinogen-induced and knockout rodent models of human colon cancer should be
utilized to systematically evaluate potential mechanisms whereby soy inﬂuences colon
cancer development. The initial focus of these studies should be on the activities of cell
signaling pathways known to be up-regulated in colon carcinogens (map kinase
pathways, Wnt/B-catenin pathway, etc.) in colonic epithelial cells obtained from rodents
exposed to soy ﬂour- or isoﬂavone-containing diets as compared to diets based on soy
concentrate (which is not protective against colon cancer development relative to full fat

soy ﬂakes or de-fatted soy ﬂour). These studies also should utilize gene expression

I65

proﬁling coupled with real-time PCR assays to identify genes that are inﬂuenced by

dietary soy.

I66

 

BIBLIOGRAPHY

167

BIBLIOGRAPHY

Adlercreutz, H., Honjo, A. and Higashi, A. (1991) Urinary excretion of lignans and
isoﬂavonoid phytoestrogens in Japanese men and women consuming a traditional
Japanese diet. Am. J. Clin. Nutr. 54: 1093-1110.

Adlercreutz, H., F otsis, T., Kurzer, M.S., Wahala, K., Makela, T. and Hase, T. (1995)
Isotope dilution gas chromatographic-mass spectrometric method for the
determination of unconjugated lignans and isoﬂavonoids in human feces, with
preliminary results in omnivorous and vegetarian women. Anal. Biochem. 225: 101-
108.

Adlercreutz, H., Mazur, W., Bartels, P., Elomaa, V.V., Watanabe, S., Wahala, K.,
Landstrom, M., Lundin, E., Bergh, A., Damber, J .-E., Aman, P. and Widmark, A.
(2000) Phytoestrogens and prostate disease. J. Nutr. 130: 658S-659S.

Ahnen, DJ. (1985) Are animal models of colon cancer relevant to human diseases? Dig.
Dis. Sci. 30: 1038-1068.

Ahnen, DJ. (1999) Tissue markers of colon cancer risk. Gastrointest. Endosc. 49: 850-
SS9.

AICR/WCRF (1997). Food, nutrition and the prevention of cancer: A global perspective.
American Institute of Cancer Research/World Cancer Research Fund, Washington
DC.

Akedo, 1., Ishikawa, H., Ioka, T., Kaji, 1., Narahara, H., Ishiguro, S., Suzuki, T. and
Otani, T. (2001) Evaluation of epithelial cell proliferation rate in normal-appearing
colonic mucosa as a high-risk marker for colorectal cancer. Cancer Epidemiol.
Biomarkers Prev. 10: 925-930.

Akiyama, T., Ishida, J ., Nakagawa, S., Ogawara, H., Watanabe, S., Itoh, N., Shibuya, M.
and F ukami, Y. (1987) Genistein, a speciﬁc inhibitor of tyrosine-speciﬁc protein
kinases. J. Biol. Chem. 262: 5592—5595.

Alabaster, 0., Tang, Z. and Shivapurkar, N. (1996) Dietary ﬁber and the
chemopreventive modelation of colon carcinogenesis. Mutat. Res. 350: 185-197.

Alberts, D.S., Einspahr, J .G., Aickin, M., Hixson, L., Earnest, D.L., Roe, D. and Powell,
M. (1994) Validation of proliferation indices as surrogate endpoint biomarkers. J.
Cell. Biochem. Suppl. 19: 73-83.

Alberts, D.S., Ritenbaugh, C., Story, J .A., Aickin, M., Rees-McGee, S., Buller, M.K.,
Atwood, J ., Phelps, J ., Ramanujam, P.S., Bellapravalu, S., Patel, J ., Bextinger, L. and

168

Clark, L. (1996) Randomized, double-blinded, placebo-controlled study of effect of
wheat bran ﬁber and calcium on fecal bile acids in patients with resected
adenomatous colon polyps. J. Nat. Cancer Inst. 88: 81-92.

Alberts, D.S., Einspahr, J ., Ritenbaugh, C., Aickin, M., Rees-McGee, S., Atwood, J .,
Emerson, S., Mason-Liddil, N., Bettinger, L., Patel, J ., Bellapravalu, S., Ramanujam,
P.S., Phelps, J. and Clark, L. (1997) The effect of wheat bran ﬁber and calcium
supplementation on rectal mucosal proliferation rates in patients with resected
adenomatous colorectal polyps. Cancer Epidemiol. Biomarkers Prev. 6: 161-169.

Alberts, D.S., Martinez, M.E., Roe, D.J., Guillen-Rodriguez, J .M., Marshall, J .R., van
Leeuwen, J .B., Reid, M.E., Ritenbaugh, C., Vargas, P.A., Bhattacharyya, A.B.,
Earnest, D.L. and Sampliner, RE. (2000) Lack of effect of a high-ﬁber cereal
supplement on the recurrence of colorectal adenomas. Phoenix Colon Cancer
Prevention Physicians' Network. N. Engl. J. Med. 342: 1156-1 162.

American Cancer Society (1998). Cancer Facts and Figures 1998. American Cancer
Society, Atlanta, GA.

American Cancer Society (2005). Colorectal Cancer Facts and Figures Special Edition
2005. American Cancer Society, Atlanta, GA.

Appelt, LC. and Reicks, M.M. (1999) Soy induces phase II enzymes but does not inhibit

dimethylbenz[a] anthracene-induced carcinogenesis in female rats. J. Nutr. 129:
1820-1826.

Appleton, G.V., Davies, P.W., Bristol, J .B. and Williamston, RC. (1987) Inhibition of
intestinal carcinogenesis by dietary supplementation with calcium. Br. J. Surg. 74:
523-525.

Arlow, F.L., Walczak, S.M., Luk, G.D. and Majumdar, AP. (1989) Attenuation of
azoxymethane-induced colonic mucosal omithine decarboxylase and tyrosine kinase
activity by calcium in rats. Cancer Res. 49: 5884-5888.

Armitage, N.C., Rooney, P.S., Gifford, K.A., Clarke, PA. and Hardcastle, J .D. (1995)
The effect of calcium supplements on rectal mucosal proliferation. Br. J. Cancer 71:
186-190.

Awad, A.B., Hernandez, A.Y.T., Fink, CS. and Mendel, S.L. (1997) Effect of dietary
phytosterols on cell proliferation and protein kinase C activity in rat colonic mucosa.
Nutr. Cancer 27: 210-215.

Baker, S.J., Fearon, E.R., Nigro, J .M., Hamilton, S.R., Preisinger, A.C., Jessup, J.M., van
Tuinen, P., Ledbetter, D.H., Barker, D.F., Nakamura, Y., White, R. and Vogelstein,
B. (1989) Chromosome 17 deletions and p53 gene mutations in colorectal
carcinomas. Science 244: 217-221 .

I69

Baldassarre, G., Nicoloso, M.S., Schiappacassi, M., Chimienti, E. and Belletti, B. (2004)
Linking inﬂammation to cell cycle progression. Curr. Pharm. Des. 10: 1653-1666.

Barnes, C.J., Hamby-Mason, R.L., Hemaidan, A. and Lee, M. (1998) Proliferating cell
nuclear antigen and transforming growth factor on expression as biomarkers of colon
cancer chemoprevention in rats. Med. Sci. Res. 26: 659-661.

Barnes, S., Grubbs, C., Setchell, K.D.R. and Carlson, J. (1990). Soybeans inhibit
mammary tumors in nodules of breast cancer. In: Mutagens and carcinogens in the
diet (Pariza, M.), pp. 239-253. Wiley-Liss, New York, NY.

Barnes, 8., Peterson, T.T.G. and Coward, K.D. (1995) Rationale for the use of genistein-
containing soy materials in chemoprevention trials in breast and prostate cancer. J.
Cell. Biochem. Suppl. 22: 181-187.

Barnes, S., Sfakianos, J ., Coward, L. and Kirk, M. (1996) Soy isoﬂavonoids and cancer
prevention. Underlying biochemical and pharmacological issues. Adv. Exp. Med.
Biol. 401: 87-100.

Barnes, S. (1998) Evolution of the health beneﬁts of soy isoﬂavones. Proc. Soc. Exp.
Biol. Med. 217: 386-392.

Barnett, D. (1976) Polycyclic aromatic hydrocarbons and foods. Food Res. Q. 36: 8-12.

Baron, J .A., Tosteson, T.D., Wargovich, M.J., Sandler, R., Mandel, J ., Bond, J ., Haile,
R., Summers, R., van Stolk, R., Rothstein, R. and Weiss, J. (1995) Calcium
supplementation and rectal mucosal proliferation: A randomized controlled trial. J.
Nat. Cancer Inst. 87: 1303-1307.

Baron, J .A., Beach, M., Mandel, J .S., van Stolk, R.U., Haile, R.W., Sandler, R.S.,
Rothstein, R., Summers, R.W., Snover, D.C., Beck, G.J., Bond, J .H. and Greenberg,
ER. (1999) Calcium supplements for the prevention of colorectal adenomas. Calcium
Polyp Prevention Study Group. N. Engl. J. Med. 340: 101-107.

Baron, J .A., Cole, BF and Mott, LA. (2002). Aspirin chemoprevention of colorectal
cancer. AACR 93rd Annual Meeting, San Francisco, CA.

Basson, M.D., Emenaker, NJ. and Rashid, Z. (1998) Effects of modulation of tyrosine
phosphorylation on brush border enzyme activity in human Caco-2 intestinal

epithelial cells. Cell Tissue Res. 292: 553-562.

Bayer, T., Colnot, T. and Dekant, W. (2001) Disposition and biotransformation of the
estrogenic isoﬂavone daidzein in rats. J. Toxicol. Sci. 62: 205-211.

170

Beatty, M.M., Lee, E.Y. and Glauert, HP. (1993) Inﬂuence of dietary calcium and
vitamin D on colon epithelial cell proliferation and I,2-dimethylhydrazine-induced
colon carcinogenesis in rats fed high fat diets. J. Nutr. 123: 144-152.

Bedi, A., Pasricha, P.J. and Akhtar, A.J. (1995) Inhibition of apoptosis during
development of colorectal cancer. Cancer Res. 55: 1811-1816.

Bennink, MR. and Om, AS. (1998) Inhibition of colon cancer (CC) by soy
phytochemicals but not by soy protein. F ASEB J. 12: A655.

Bennink, MR. (2001) Dietary soy reduces colon carcinogenesis in human and rats.
Nutrition and cancer prevention: New insights into the role of phytochemicals. Adv.
Exp. Med. Biol. 492: 11-17.

Biasco, G., Paganelli, G.M., Santucci, R., Brandi, G. and Barbara, L. (1994)
Methodological problems in the use of rectal cell proliferation as a biomarker of
colorectal cancer risk. J. Cell. Biochem. Suppl. 19: 55-60.

Billings, P.C., Newbeme, PM. and Kennedy, AR. (1990) Protease inhibitor suppression
of colon and anal gland carcinogenesis induced by dimethylhydrazine.
Carcinogenesis 11: 10083-11086.

Bird, RP. (1986) Effect of dietary components on the pathobiology of colonic
epithelium: Possible relationship with colon tumorigenesis. Lipids 21: 289-291.

Bird, R.P., Schneider, R., Stamp, D. and Bruce, W.R. (1986) Effect of dietary calcium
and cholic acid on the proliferative indices of murine colonic epithelium.
Carcinogenesis 7: 1657-1661.

Bird, RP. (1987) Observation and quantiﬁcation of aberrant crypts in murine colon
treated with a colon carcinogen: Preliminary ﬁndings. Cancer Lett. 37: 147-151.

Birk, Y. (1990). Structure and function of protease inhibitors from legume seeds and

cereal grains and their interaction with digestive enzymes of stored product insects.
BARD, Bet Dagan, Israel.

Birt, D.F., Hendrich, S. and Wang, W. (2001) Dietary agents in cancer prevention:
Flavonoids and isﬂavonoids. Pharmacol. Ther. 90: 157-177.

Bisgaard, M.L., Fenger, K., Bulow, S., Niebuhr, E. and Mohr, J. (1994) Familial
adenomatuous polyposis (FAP): Frequency, penetrance and mutation rate. Hum.
Mutat. 3: 121-125.

Bishop, D.T. and Thomas, H.J.W. (1990) The genetics of colorectal cancer. Cancer Surv.
4: 585-604.

171

 

Bland, P.W. and Britton, DC. (1975) Colonic lymphoid tissue and its inﬂuence on tumor
induction in DMH-treated rats. Br. J. Cancer 4: 275-280.

Boland, C.R., Montgomery, GK. and Kim, Y.S. (1988) Alterations in human colonic
mucin occurring with cellular differentiation and malignant transformation. Proc.
Natl. Acad. Sci. USA 79: 2051-2055.

Bonithon-Kopp, C., Kronborg, 0., Giacosa, A., Rath, U. and F aivre, J. (2000) Calcium
and ﬁbre supplementation in prevention of colorectal adenoma recurrence: A
randomised intervention trial. Lancet 356: 1300-1306.

Booth, C., Hargreaves, D.F., Hadﬁeld, J.A., McGown, AT and Potten, CS. (1999)
Isoﬂavones inhibit intestinal epithelial cell proliferation and induce apoptosis in vitro.
Br. J. Cancer 80: 1550—1557.

Bostick, R., Potter, J .D., Fosdick, L., Grambsch, P., Lampe, J.W., Wood, J .R., Louis,
T.A., Ganz, R. and Grandits, G. (1993a) Calcium and colorectal epithelial cell
proliferation: A preliminary randomized, double-blinded, placebo-controlled clinical
trial. J. Nat. Cancer Inst. 85: 132-141.

Bostick, R., Potter, J .D., Sellers, T., McKenzie, D., Kushi, L.H. and Folsom, A. (1993b)
Relation of calcium, vitamin D, and dairy food intake to incidence of colon cancer
among older women: The Iowa Women's Health Study. Am. J. Epidemiol. 137: 1302-
1317.

Bostick, R.M., Potter, J .D., Kushi, L.H., Sellers, T.A., Steinmetz, K.A., McKenzie, D.R.,
Gapstur, SM. and Folsom, AR. (1994) Sugar, meat, and fat intake, and non-dietary

risk factors for colon cancer incidence in Iowa women. Cancer Causes Control 5: 38-
52.

Bostick, R.M., Fosdick, L., Wood, J .R., Grambsch, P., Grandits, G.A., Lillemoe, T.J.,
Louis, TA. and Potter, J .D. (1995) Calcium and colorectal epithelial cell proliferation
in sporadic adenoma patients: A randomized, double-blinded, placebo-controlled
clinical trial. J. Nat. Cancer Inst. 87: 1307-1315.

Bostick, RM. (1997) Human studies of calcium supplementation and colorectal epithelial
cell proliferation. Cancer Epidemiol. Biomarkers Prev. 6: 971-980.

Bostick, R.M., F osdick, L., Grandits, G.A., Lillemoe, T.J., Wood, J .R., Grambsch, P.,
Louis, TA. and Potter, J .D. (1997a) Colorectal epithelial cell proliferative kinetics

and risk factors for colon cancer in sporadic adenoma patients. Cancer Epidemiol.
Biomarkers Prev. 6: 1011-1019.

Bostick, R.M., Fosdick, L., Lillemoe, T.J., Ovem, R, Wood, J .R., Grambsch, P., Elmer,
P. and Potter, J .D. (1997b) Methodological ﬁndings and considerations in measuring

l 172

M1
.LL

 

colorectal epithelial cell proliferation in humans. Cancer Epidemiol. Biomarkers Prev.
6: 931-942.

Bravo, R. and MacDonald-Bravo, H. (1987) Existence of two populations of

cyclin/proliferating cell nuclear antigen during the cell cycle: Association with DNA
replication sites. J. Cell. Biol. 105: 1549-1554.

Brenner, C.E., Baker, S.M., Morrison, P.T., Warren, G., Smith, LG. and Lescoe, MK.
(1994) Mutation in the DNA mismatch repair gene homologue hMLHI is associated
with hereditary nonpolyposis colon cancer. Nature 368: 25 8-261.

Bromely, M., Rew, D., Becciolini, A., Balzi, M., Chadwick, C., Hewitt, D., Li, Y.Q. and
Potten, CS. (1996) A comparison of proliferation markers (BrdU, Ki-67, PCNA)

determined at each cell position in the crypts of normal human colonic mucosa. Eur.
J. Histochem. 40: 89-100.

Browne, S.J., Williams, A.C., Hague, A., Butt, A.J. and Paraskeva, C. (1994) Loss of
APC protein expressed by human colonic epithelial cells and the appearance of a

speciﬁc low-molecular-weight form is associated with apoptosis in vitro. Int. J.
Cancer 59: 56-64.

Bruno, S. and Darzynkiewicz, Z. (1992) Cell cycle dependent expression and stability of
the nuclear protein detected by Ki-67 antibody in HL-60 cells. Cell. Prolif. 25: 31-40.

Burkitt, D. (1969) Related disease-related cause? Lancet 2: 1229-1231.

Buset, M., Lipkin, M., Winawer, S., Swaroop, S. and Friedman, E. (1986) Inhibition of
human colonic epithelial cell proliferation in vivo and in vitro by calcium. Cancer
Res. 46: 5426-5430.

Bussey, H.J.R. (1975). Familial polyposis coli: Family studies, histopathology,
differential diagnosis and results of treatment. John Hopkins University Press,
Baltimore, MD.

Byers, T., Nestle, M., McTieman, A., Doyle, C., Currie-Williams, A., Gransler, T. and
Thun, M.J. (2002) American Cancer Society guidelines on nutrition and physical
activity for cancer prevention: Reducing the risk of cancer with healthy food choices
and physical activity. CA Cancer J. Clin. 52: 92-119.

Cademi, G., Giannini, A., Lancioni, L., Luceri, C., Biggeri, A. and Dolara, P. (1995)
Characterisation of aberrant crypt foci in carcinogen-treated rats: Association with

intestinal carcinogenesis. Br. J. Cancer 71: 763-769.

Carrol, K.K. (1986) Dietary fat and cancer: Speciﬁc action or caloric effect? J. Nutr. 116:
1130.

I73

Cassidy, A., Bingham, S. and Setchell, K.D.R. (1994) Biological effects of a diet of soy
protein rich in isoﬂavones on the mentrual cycle of premenopausal women. Am. J.
Clin. Nutr. 60: 333-340.

Cassidy, A. (1996) Physiological effects of phyto-oestrogenes in relation to cancer and
other human health risks. Proc. Nutr. Soc. 55: 399-417.

Cats, A., Kleibeuker, J .H., Van der Meer, R., Kuipers, F ., Sluiter, W.J., Hardonk, M.J.,
Oremus, E.T., Mulder, NH. and De Vries, E.G. (1995) Randomized, double-blinded,
placebo-controlled intervention study with supplemental calcium in families with
hereditary nonpolyposis colorectal cancer. J. Nat. Cancer Inst. 87: 598-603.

Cerutti, RA. and Trump, HF. (1981) Inﬂammation and oxidative stress in
carcinogenesis. Cancer Cells 3: 1-7.

Chan, T.L., Zhao, W., Leung, S.Y. and Yuen, ST (2003) BRAF and KRAS mutations in
colorectal hyperplastic polyps and serrated adenomas. Cancer Res. 63: 4878-4881.

Chang, Y.-C., Nair, M.G., Santell, RC. and Helferich, W.G. (1994) Microwave-mediated
synthesis of anticarcinogenic isoﬂavones from soybeans. J. Agric. Food Chem. 42:
1869-1871.

Chapkin, R.S., Jiang, Y.H., Davidson, LA. and Lupton, JR. (1997) Modulation of
intracellular second messengers by dietary fat during colonic tumor development.
Adv. Exp. Med. Biol. 422: 85-96.

Chen, M.-J., Longnecker, M.P., Morgenstem, H., Lee, E.R., F rankl, H.D. and Haile, R.
(1998) Recent fuse of hormone replacement therapy and the prevalence of colorectal
adenomas. Cancer Epidemiol. Biomarkers Prev. 7: 227-230.

Chen, Z., Zheng, W., Custer, L.J., Dai, Q., Shu, X.O., Jin, F. and Franke, AA. (1999)
Usual dietary consumption of soy foods and its correlation with the excretion rate of

isoﬂavonoids in overnight urine samples among Chinese women in Shanghai. Nutr.
Cancer 33: 82-87.

Chia, V. (2004) Calcium and colorectal cancer: Some questions remain. Nutr. Rev. 62:
115-124.

Cho, K.R., Oliner, J .D., Simons, J .W., Hedrick, L., Fearon, E.R., Preisinger, A.C., Hedge,
P., Silverman, GA. and Vogelstein, B. (1994) The DCC gene: Structural analysis and
mutations in colorectal carcinomas. Genomics 19: 525-531.

Choi, H.J., Jung, I.K., Kim, SS. and Hong, SH. (1997) Proliferating cell nuclear antigen

expression and its relationship to malignancy potential in invasive colorectal
carcinomas. Dis. Colon Rectum 40: 51-59.

174

 

Chomchai, C., Bhadrachari, N. and Nigro, ND. (1974) The effect of bile on the induction
of experimental intestinal tumors in rats. Dis. Colon Rectum 17: 310-312.

Clark, J ., Chu, M. and Calabresi, P. (1989). Growth inhibition of colon cancer cells by
genistein and suramin. Proceedings of the American Association of Cancer Research.

Clark, L.C., Combs, G.F., Jr., Turnbull, B.W., Slate, E.H., Chalker, D.K. and Chow, J.
(1996) Effects of selenium supplementation for cancer prevention in patients with

carcinoma of the skin. A randomized controlled trial. Nutritional Prevention Study
Group. JAMA 276: 1957-1963.

Clinton, S.K., Destree, R.J., Anderson, D.B., Truex, C.R., Imrey, PB. and Visek, W.J.
(1979) 1,2-dimethylhydrazine induced intestinal cancer in rats fed beef or soybean
protein. Nutr. Rep. Int. 20: 335-342.

Collet, J .-P., Sharpe, C., Belzile, E., Boivin, J .-F., Hanley, J. and Abenhaim, L. (1999)
Colorectal cancer prevention by non—steroidal anti-inﬂammatory drugs: Effects of
dosage and timing. Br. J. Cancer 81: 62-68.

Constantinou, A., Kiguchi, K. and Huberrnan, E. (1990) Induction of differentiation and
DNA strand breakage in human HL-60 and K-562 leukemia cells by genistein.
Cancer Res. 50: 2618-2624.

Corpet, D.E., Jacquinet, C., Peiffer, G. and Tache, S. (1997) Insulin injections promote
the growth of aberrant crypt foci in the colon of rats. Nutr. Cancer 27: 316-320.

Coughlin, S.S., Calle, E.E., Teras, L.R., Petrelli, J. and Thun, M.J. (2004) Diabetes
mellitus as a predictor of cancer mortality in a large cohort of US adults. Am. J.
Epidemiol. 159: 1160-1167.

Courtney, E.D.J., Melville, D.M. and Leicester, R.J. (2004) Review article:
Chemoprevention of colorectal cancer. Aliment. Pharmacol. Ther. 19: 1-24.

Coward, L., Barnes, N.C., Setchell, DR. and Barnes, S. (1993) Genistein, daidzein, and
their B-glycoside conjugates: Antitumour isoﬂavones in soybean foods from
American and Asian diets. J. Agric. Food Chem. 41: 1961-1967.

Cruz, M.L.A., Wong, W.W. and Mimouni, F. (1994) Effects of infant nutrition on
cholesterol synthesis rates. Pediatr. Res. 35: 135-140.

Davies, M.J., Bowey, E.A., Adlercreutz, H., Rowland, LR. and Rumsby, PC. (1999)
Effects of soy or rye supplementation of high-fat diets on colon tumor development in

azoxymethane-treated rats. Carcinogenesis 20: 927-931.

Deschner, BE. and Lipkin, M. (1975) Proliferative patterns in colonic mucosa in familial
polyposis. Cancer 35: 413-418.

175

 

 

Deschner, E.E., Long, F .C. and Katz, S. (1975) The detection of aberrant DNA synthesis
in a member of a high-risk cancer family. Am. J. Dig. Dis. 20: 418-424.

Deschner, E.E., Winawer, S.J., Long, F .C. and Boyle, CC. (1977) Early detection of
colonic neoplasia in patients at high risk. Cancer 40: 2625-2631.

Deschner, EB. (1985) Cell kinetic approaches to deﬁning premalignant conditions. Prog.
Clin. Biol. Res. 186: 187-202.

Deschner, BE. (1988) Cell proliferation and colonic neoplasia. Scand. J. Gastroenterol.
Suppl. 151: 94-97.

Deschner, E.E., Godbold, J. and Lynch, HT. (1988) Rectal epithelial cell proliferation in
a group of young adults: Inﬂuence of age and genetic risk for colon cancer. Cancer
61: 2286-2290.

Deschner, E.E., Ruperto, J ., Wong, G. and Mewmark, H.L. (1991) Quercetin and rutin as

inhibitors of axozymethanol-induced colonic neoplasia. Carcinogenesis 12: 1193-
1196.

Diez, M., Ramos, P., Medrano, M.J., Mugiierza, J .M., Villeta, R., Lozano, 0., Escribano,
J ., Noguerales, F., Ruiz, A. and Granell, J. (2003) Preoperatively irradiated rectal
carcinoma: Analysis of the histopathologic response and predictive value of
proliferating cell nuclear antigen immunostaining. Oncology 64: 213-219.

Doerge, D.R., Chang, H.C., Chruchwell, MI. and Holder, CL. (2000) Analysis of soy
isoﬂavone conjugation in vitro and in human blood using liquid chromatography -
mass spectrometry. Drug Metab. Dispos. 28: 298-307.

Du Manoir, S., Guillaud, P., Camus, E., Seigneurin, D. and Brugal, G. (1991) Ki-67
labelling in postrnitotic cells deﬁnes different Ki-67 pathways within the 2c
compartment. Cytometry 12: 455-463.

Dubois, R.N., Radhika, A., Reddy, BS. and Entingh, A.J. (1996) Increased
cyclooxygenase-2 levels in carcinogen-induced rat colonic tumors. Gastroenterol.
110: 1259-1262.

Earnest, D.L., Hixon, L]. and Alberts, BS. (1992) Piroxicam and other cyclooxygenase
inhibitors: Potential for cancer chemoprevention. J. Cell. Biochem. Suppl. 161: 156-
166.

Easwaran, V., Song, V., Polakis, P. and Byers, S. (1999) The ubiquitin-proteasome
pathway and serine kinase activity modulate adenomatous polyposis coli protein-

mediated regulation of B-catenin—lymphocyte enhancer-binding factor signaling. J.
Biol. Chem. 274: 16641-16645.

I76

Einspahr, J .G., Alberts, D.S., Gapstur, S.M., Bostick, R.M., Emerson, SS. and Gemer,
E.W. (1997) Surrogate end-point biomarkers as measures of colon cancer risk and

their use in cancer chemoprevention trials. Cancer Epidemiol. Biomarkers Prev. 6:
37-48.

Endres, J .G. (2001). Soy protein products: Characteristics, nutritional aspects, and
utilization. AOCS Press, Champaign, IL.

Erhardt, J .G., Meisner, C., Bode, J .C. and Bode, C. (2003) Lycopene, B-carotene, and
colorectal adenomas. Am. J. Clin. Nutr. 78: 1219-1224.

Faivre, J ., Bonithon-Kopp, C. and Medicinne, CF. (1999). A randomized trial of calcium
and ﬁber supplementation in the prevention of recurrence of colorectal adenomas.
American Gastroenterological Association Annual Meeting, Orlando, FL.

Fang, W.F. and Strobel, H.W. (1978) Activation of carcinogens and mutagens by rat
colon mucosa. Cancer Res. 38: 2939.

Fearon, E.R., Hamilton, SR. and Vogelstein, B. (1987) Clonal analysis of human
colorectal tumors. Science 238: 193-196.

Fearon, E.R., Cho, K.R., Nigro, J .M., Kern, S.E., Simons, J.W., Ruppert, J.M., Hamilton,
S.R., Preisinger, A.C., Thomas, G. and Kinzler, K.W. (1990) Identiﬁcation of a
chromosome 18q gene that is altered in colorectal cancers. Science 247: 49-56.

Fearon, ER. and Vogelstein, B. (1990) A genetic model for colorectal tumorigenesis.
Cell 61: 759-767.

F earon, ER. and Jones, RA. (1992) Progressing toward a molecular description of
colorectal cancer development. F ASEB J. 6: 2783-2790.

Ferrandez, A., Samowitz, W., DiSario, J .A. and Burt, R.W. (2004) Phenotypic
characteristics and risk of cancer development inhyperplastic polyposis: Case series
and literature review. Am. J. Gastroenterol. 99: 2012-201 8.

Fotsis, T., Pepper, M., Adlercreutz, H., Fleishmann, G., Hase, T., Montesano, R. and
Schweigerer, L. (1993) Genistein, a dietary-derived inhibitor of in vitro angiogenesis.
Proc. Natl. Acad. Sci. USA 90: 2690-2694.

F ranke, AA. and Custer, L]. (1996) Daidzein and genistein concentrations in human
milk after soy consumption. Clin. Chem. 42: 955-964.

Friedman, EA. (1985) A multistage model for human colon carcinoma development
integrating cell culture studies with pathology. Cancer Invest. 5: 453-461.

177

Fuchs, C.S., Colditz, G.A., Stampfer, M.J., Speizer, E.E., Giovannucci, E., Hunter, D.J.,
Rosner, B. and Willet, WC. (1999) Dietary ﬁber and the risk of colorectal cancer and
adenoma in women. N. Engl. J. Med. 340: 169-176.

F ujimori, T., Satonaka, K., Yamamura, I.Y., Nagasako, K. and Maeda, S. (1994) Non-
involvement of ras mutations in ﬂat colorectal adendomas and carcinomas. Int. J.
Cancer 57: 51-55.

Garcea, G., Sharma, R.A., Dennison, A., Steward, W.P., Gescher, A. and Berry, DP.
(2003) Molecular biomarkers of colorectal carcinogenesis and their role in
surveillance and early intervention. Eur. J. Cancer 39: 1041-1052.

Gardner, E]. (1951) A genetic and clinical study of intestinal polyposis, a predisposing
factor for carcinoma of the colon and rectum. Am. J. Hum. Genet. 3: 167-176.

Garland, C., Shekelle, R.B., Barrett-Connor, E., Criqui, M.H., Rossof, AH. and Paul, O.
(1985) Dietary vitamin D and calcium and risk of colorectal cancer: A 19-year
prospective study in men. Lancet 1: 307-309.

Garrity, M.M., Burgart, L.J., Mahoney, M.R., Windschitl, H.E., Salim, M., Wiesenfeld,
M., Krook, J .E., Michalak, J .C., Goldberg, R.M., O’Connell, M.J., Furth, A.F.,
Sargent, D.J., Murphy, L.M., Hill, E., Riehle, D.L., Meyers, CH. and Witzig, TE.
(2004) Prognostic value of proliferation, apoptosis, defective DNA mismatch repair,
and p53 overexpression in patients with resected Dukes’ BZ or C colon cancer: A
North Central Cancer Treatment Group Study. J. Clin. Oncol. 22: 1572-1582.

Gasparini, G., Boracchi, P., Verderio, P. and P., B. (1994) Cell kinetics in human breast
cancer: Comparison between the prognostic value of the cytoﬂuorimetric S-phase
fraction and that of antibodies to Ki-67 and PCNA antigens detected by
immunocytochemistry. Int. J. Cancer 57: 822-829.

Gee, J .M., Notebom, H.P.J.M., Polley, A.C.J. and Johnson, LT. (2000) Increased
induction of aberrant crypt foci by 1,2-dimethylhydrazine in rats fed diets containing
puriﬁed genistein or genistein-rich soya protein. Carcinogenesis 21: 2255-2259.

Gerdes, H., Gillin, J .S., Zimbalist, E., Urmacher, C., Lipkin, M. and Winawer, SJ. (1993)
Expansion of the epithelial proliferative compartment and frequency of adenomatous
polyps in the colon correlate with the strength of family history of colorectal cancer.
Cancer Res. 53: 279-282.

Gerdes, J ., Stein, H., Pileri, S., Rivano, M.T., Gobbi, M., Ralﬂ<iaer, E., Nielsen, K.M.,
Pallesen, G., Bartels, H., Palestro, G. and Delsol, G. (1987) Prognostic relevance of
tumor-cell growth fraction in malignant non-Hodgkin's lymphomas. Lancet 2: 448-
449.

I78

 

Gerdes, J ., Li, L., Schlueter, C., Duchrow, M., Wohlenberg, C., Gerlach, C., Stahmer, 1.,
Kloth, S., Brandt, E. and Flad, H.D. (1991) Immunobiochemical and molecular

biologic characterization of the cell proliferation-associated nuclear antigen that is
deﬁned by monoclonal antibody Ki-67. Am. J. Pathol. 138: 867-873.

Gestetner, B., Birk, Y. and Tencer, Y. (1968) Fate of ingested soybean saponins and the
physiological aspect of their hemolytic activity. J. Agric. Food Chem. 16: 1031-1035.

Giardiello, F.M., Hamilton, S.R., Krush, A.J., Piantadosi, S., Hylind, L.M., Celano, P.,
Booker, C., Robison, R. and Offerhaus, G.J. (1993) Treatment of colonic and rectal
adenomas with sulindac in familial adenomatous. N. Engl. J. Med.

Giovannucci, E., Stamffer, M.J., Colditz, G., Rimm, EB. and Willett, W. (1992)
Relationship of diet to risk of colorectal adenoma in men. J. Nat. Cancer Inst. 84: 91-
98.

Giovannucci, E., Rimm, E.B., Stampfer, M.J., Colditz, G.A., Asherio, A. and Willett,
WC. (1994) Intake of fat, meat and ﬁber in relation to risk of colon cancer in men.
Cancer Res. 54: 2390-2397.

Glauert, H.P., Bennink, MR. and Sander, CH. (1981) Enhancement of 1,2-
dimethylhydrazine-induced colon carcinogenesis in mice by dietary agar. Fd. Cosmet.
Toxicol. 19: 281-286.

Goldberg, M.J., Smith, J .W. and Nichols, R.L. (1977) Comparison of the fecal microﬂora
of Seventh-day Adventists with individuals consuming a general diet. Ann. Surgery
186: 97-100.

Goldbohm, R.A., van den Brandt, P.A., van 't Veer, P., Brants, H.A., Dorant, E.,
Sturmans, F. and Hermus, R.J. (1994) A prospective cohort study on the relation
between meat consumption and the risk of colon cancer. Cancer Res. 54: 718-723.

Goldwyn, S., Lazinsky, A. and Wei, H. (2000) Promotion of health by soy isoﬂavones:
Efﬁcacy, beneﬁt and safety concerns. Drug Metab. Drug Interact. 17: 261-289.

Gooderham, M.J., Adlercreutz, H.H., Ojala, S.T., Wahala, K. and Holub, B]. (1996) A
soy protein isolate rich in genistein and daidzein and its effects on plasma isoﬂavone
concentrations, platelet aggregation, blood lipids and fatty acid composition of
plasma phospholipid in normal men. J. Nutr. 126: 2000-2006.

Govers, M.J.A.P., Terrnont, D.S.M.L., Kleibeuker, J.H., De Vries, E.G.E. and Van der
Meer, R. (1994) Mechanisms of the antiproliferative effect of milk mineral and other
calcium supplements on colonic epithelium. Cancer Res. 54: 95-100.

Gr 3f, E. and Eaton, J .W. (1985) Dietary suppression of colonic cancer. Fiber or phytate?
Cancer 56: 717-718.

179

Graf, E. and Eaton, J .W. (1993) Suppression of colonic cancer by dietary phytic acid.
Nutr. Cancer 19: 11-19.

Graham, S., Marshall, J., Haughey, B., Mittleman, A., Swanson, M., Zielezny, M., Byers,
T., Wilkinson, G. and West, D.W. (1988) Dietary epidemiology of cancer of the
colon in western New York. Am. J. Epidemiol 128: 490-503.

Green, S.E., Chapman, P., Burn, J., Burt, A.D., Bennet, M., Appleton, D.R., Varma, J .S.
and Mathers, J .C. (1998) Colonic epithelial cell proliferation in hereditary non-
polyposis colorectal cancer. Gut 43: 85-92.

Greenberg, E.R., Baron, J.A. and Tosteson, TD. (1994) A clinical trial of antioxidant
vitamins to prevent colorectal adenoma. N. Engl. J. Med. 331: 141-147.

Greenwald, P., Lanza, E. and Eddy, GA. (1987) Dietary ﬁber in the reduction of colon
cancer risk. J. Am. Diet Assoc. 87: 1178-1188.

Greenwald, P., Kelloff, G.J., Boone, CW. and McDonald, SS. (1995) Genetic and
cellular changes in colorectal cancer: Proposed targets of chemopreventive agents.
Cancer Epidemiol. Biomarkers Prev. 7: 691-702.

Gregoire, R.C., Stern, H.S., Yeung, K.S., Stadler, J., Langley, S., F urrer, R. and Bruce,
W.R. (1989) Effect of calcium supplementation on mucosal cell proliferation in high
risk patients for colon cancer. Gut 30: 376-382.

Groden, J ., Thliveris, A., Samowitz, W., Carlson, M., Gelbert, L., Albertsen, H., Joslyn,
G., Stevens, J ., Spirio, L. and Robertson, M. (1991) Identiﬁcation and
characterization of the familial adenomatous polyposis. Cell 66: 589-600.

Grodstein, F., Martinez, M.E., Platz, E.A., Giovannucci, E., Colditz, G.A., Kautzky, M.,
Fuchs, C. and Stampfer, M.J. (1998) Postmenopausal hormone use and risk for
colorectal cancer and adenoma. Ann. Int. Med. 128: 705-712.

Grubben, M.J.A.L., van den Braak, C.C.M., Essenberg, M., Olthof, M., Tangerrnan, A.,
Katan, MB. and Nagengast, FM. (2001) Effect of resistant starch on potential
biomarkers for colonic cancer risk in patients with colonic adenomas: A controlled

trial. Dig. Dis. Sci. 46: 750-756.

Gryfe, R., Di Nicola, N., La], G., Gallinger, S. and Redston, M. (1999) Inherited
colorectal polyposis and cancer risk of the APC 11307K polymorphism. Am. J. Hum.
Genet. 64: 378-384.

Guo, J .Y., Li, X., Browning, J .D.J ., Rottinghaus, G.E., Lubahn, D.B., Constantinou, A.,
Bennink, M. and MacDonald, RS. (2004) Dietary soy isoﬂavones and estrone protect

I80

ovariectomized ERa K0 and wild-type mice from carcinogen-induced colon cancer.
J. Nutr. 1342: 179-182.

Gurﬁnkel, D.M. and Rao, A.V. (2003) Soyasaponins: The relationship between chemical
structure and colon anticarcinogenic activity. Nutr. Cancer 47: 24-33.

Haenszel, W., Berg, J .W., Segi, M., Kurihara, M. and Locke, PB. (1973) Large-bowel
cancer in Hawaiian Japanese. J. Nat. Cancer Inst. 51: 1765-1779.

Hakkak, R., Korourian, S., Ronis, M.J.J., Johnston, J .M. and Badger, TM. (2001) Soy
protein isolate consumption protects against azoxymethane-induced colon tumors in
male rats. Cancer Lett. 166: 27-32.

Hall, P.A., Coates, P.J., Ansari, B. and Hopwood, D. (1994) Regulation of cell number in
the mammalian gastrointestinal tract: The importance of apoptosis. J. Cell Sci. 107:
3569-3577.

Hamilton, S. (1983) Pathogenesis of polyps (adenomas). Dis. Colon Rectum 26: 413-414.

Hartiala, K. (1973) Metabolism of hormones, drugs and other substances by the gut.
Physiol. Rev. 53: 496.

Hawk, E.T., Asad, U. and Viner, J .L. (2004) Colorectal cancer chemoprevention - an
overview of the science. Gastroenterol. 126: 1423-1447.

Hawkins, NJ. and Ward, R.L. (2001) Sporadic colorectal adenomas with microsatellite

instability and their possible origin in hyperplastic polyps and serrated adenomas. J.
Nat. Cancer Inst. 93: 1307-1313.

Hedrick, L., Cho, K.R., Fearon, E.R., Wu, T.C., Kinzler, K.W. and Vogelstein, B. (1994)
The DCC gene product in cellular differentiation and colorectal tumorigenesis. Genes
Dev. 8: 1174-1183.

Heinonen, S., Wahala, K. and Adlercreutz, H.H. (2000) Metabolites of dietary
phytoestrogens daidzein, genistein, and glycitein. J. Nutr. 130: 6838.

Helms, J .R. and Gallaher, DD. (1995) The effect of dietary soy protein isolate and
genistein on the development of preneoplastic lesions (aberrant crypts) in rats. J.
Nutr. 125: 802S.

Henderson, BE. (1990) Summary report of the sixth symposium on cancer registries and
epidemiology in the Paciﬁc basin. J. Nat. Cancer Inst. 82: 1186-1190.

Hendrich, S., Wang, G]. and Lin, H.K. (1999) Isoﬂavone metabolism and
bioavailability. Antioxidant Status, Diet, Nutrition and Health: 211-230.

181

Higashi, K. and Ogawara, H. (1994) Daidzein inhibits insulin- or insulin-like growth
factor-l-mediated signaling in cell cycle progression of Swiss 3T3. Biochemica
Biophysica Acta. 1221: 29-35.

Hill, M.J., Crowther, J .S., Drasar, B.S., Aries, V. and Williams, RE. (1971) Bacteria and
aetiology of cancer of large bowel. Lancet 1: 95-100.

Hirose, Y., Yoshimi, N., Suzui, M., Kawabata, K., Tanaka, T. and Mori, H. (1997)
Expression of bcl-2, bax, and bcl-XL proteins in axoxymethane-induced rat colonic
adenocarcinomas. Molec. Carcinogenesis 19: 25-30.

Hofstad, B., Almendingen, K., Vatn, M., Andersen, S.N., Owen, R.W., Larsen, S. and
Osnes, M. (1998) Growth and recurrence of colorectal polyps: A double-blind 3-year
intervention with calcium and antioxidants. Digestion 59: 148-156.

Holder, C.L., Churchwell, M.I. and Doerge, DR. (1999) Quantiﬁcation of soy
isoﬂavones, genistein, daidzein, and conjugates in rat blood using LC/ES-MS. J.
Agric. Food Chem. 47: 3764-3770.

Holt, RR. and Yeh, KY. (1988) Colonic proliferation is increased in senescent rats.
Gastroenterol. 95: 1 556-1563.

Holt, P.R., Atillasoy, E.G., Gilman, J ., Guss, J ., Moss, S.F., Newmark, H., Fan, K., Yang,
K. and Lipkin, M. (1998) Modulation of abnormal colonic epithelial cell proliferation
and differentiation by low-fat dairy foods: A randomized controlled trial. JAMA 208:
1074-1079.

Hu, J ., Liu, Y., Yu, Y., Zhao, T., Liu, S. and Wang, Q.Q. (1991) Diet and cancer of the
colon and rectum: A case-control study in China. Int. J. Epidemiol. 20: 362-367.

Hu, J ., Reddy, M.B., Hendrich, S. and Murphy, PA. (2004) Soyasaponin I and
sapongenol B have limited absorption by Caco-2 intestinal cells and limited
bioavailability in women. J. Nutr. 134: 1867-1873.

Huggett, A.C., Pridmore, S., Malnoe, A., Haschke, F. and Offord, EA. (1997) Phyto-
oestrogens in soy-based infant formula. Lancet 350: 815-816.

Hyman, N.H., Anderson, P. and Blasyk, H. (2004) Hyperplastic polyps and the risk of
colon cancer. Dis. Colon Rectum 47: 2101-2104.

Ilyas, M. and Tomlinson, I.P.M. (1997) The interactions of APC, E-cadherin and B-
Catenin in tumour development and progression. J. Pathol. 182: 128-137.

Imoto, M., Yamashita, T. and Sawa, T. (1988) Inhibition of cellular phosphatidylinositol
turnover by psitectorigenin. F EBS Lett. 230: 43-46.

182

Isbister, W.H. (1993) Hyperplastic polyps. Aust. NZ J. Surg. 63: 175-180.

Izawa, H., Yamamoto, H., Ikeda, M., Ikeda, K., Fukunaga, H., Yasui, M., Ikenaga, M.,
Sekimoto, M., Monden, T., Matsuura, N. and Monden, M. (2002) Analysis of cyclin
D1 and CDK expression in colonic polyps containing neoplastic foci: A study of
proteins extracted from parafﬁn sections. Oncol. Rep. 9: 1313-1318.

Izumi, T., Piskula, MK. and Osawa, S. (2000) Soy isoﬂavone aglycones are absorbed
faster and in higher amounts than their glucosides in humans. J. Nutr. 130: 1695-
1699.

Jackson, C.-J.C. and Gilani, GS. (2002). Tables of isoﬂavone, coumestan, and lignan

data. In: Phytoestrogens and health (Gilani, GS. and Anderson, J .J .B.), AOCS Press,
Champaign, IL.

Jen, J ., Powell, S.M., Papadopoulos, N., Smith, K.J., Hamilton, S.R., Vogelstein, B. and
Kinzler, K.W. (1994) Molecular determinants of dysplasia in colorectal lesions.
Cancer Res. 54: 5523-5526.

Jha, H.C., von Recklinghausen, G. and Zilliken, F. (1985) Inhibition of in vitro
microsomal lipid peroxidation by isoﬂavonoids. Biochem. Pharmacol. 34: 1367-1369.

Joannou, G.E., Kelly, G.E., Reeder, A.Y., Waring, M.A. and Nelson, C. (1995) A urinary
proﬁle study of dietary phytoestrogens: The identiﬁcation and mode of metabolism of
new isoﬂavonoids. J. Steroid Biochem Molec. Biol. 54: 167-184.

Kakiuchi, H., Watanabe, M., Ushijima, T., Toyota, M., Imai, K., Weisburger, J .H.,
Sugimura, T. and Nagao, M. (1995) Speciﬁc 5'-GGA-3' mutation of the APC gene in

rat colon tumors induced by 2-amino-1methyl-6-phenylimidazo[4,5-b] pyridine. Proc.
Natl. Acad. Sci. USA 92: 910-914.

Kang, M., Mitomi, H., Sada, M., Tokumitsu, Y., Takahashi, Y., Igarashi, M., Katsumata,
T. and Okayasu, I. (1997) Ki-67, p53, and Bel-2 expression of serrated adenomas of
the colon. Am. J. Surg. Pathol. 21: 417-423.

Karagas, M.R., Tosteson, T.D., Greenberg, E.R., Rothstein, R.I., Roebuck, B.D., Herrin,
M. and Ahnen, D. (1998) Effects of milk and milk products on rectal mucosal
proliferation in humans. Cancer Epidemiol. Biomarkers Prev. 7: 757-766.

Karr, S.C., Lampe, J .W., Hutchins, A.M. and Slavin, J.L. (1997) Urinary isoﬂavonoid
excretion in humans is dose dependent at low to moderate levels of soy-protein
consumption. Am. J. Clin. Nutr. 66: 46-51.

Kawamori, T., Rao, C., Seibert, K. and Reddy, B. (1998) Chemopreventive effect of

celecoxib, a speciﬁc cyclooxygenase-2 inhibitor on colon carcinogenesis. Cancer Res.
58: 409-412.

183

Keku, T.O., Galanko, J .A., Murray, S.C., Woosley, J .T. and Sandler, RS. (1998) Rectal
mucosal proliferation, dietary factors, and the risk of colorectal adenomas. Cancer
Epidemiol. Biomarkers Prev. 7: 993-999.

Kelloff, G.J., Boone, C.W., Steele, V., Crowell, J .A., Lubet, R. and Sigman, CC. (1994)
Progress in cancer chemoprevention: Perspectives on agent selection and short-term
clinical trials. Cancer Res. 548: 20158-20248.

Kelloff, G.J., Bast, R.C., Jr., Coffey, D.S., D'Amico, A.V., Kerbel, R.S., Park, J.W.,
Ruddon, R.W., Rustin, G.J.S., Schilsky, R.L., Sigman, CC. and Woude, G.F.V.
(2004) Biomarkers, surrogate end points, and the acceleration of drug development
for cancer prevention and treatment: An update. Clin. Cancer Res. 10: 3881-3884.

Kelly, G.E., Nelson, C., Waring, M.A., Joannou, GE. and Reeder, A.Y. (1993)
Metabolites of dietary (soy) isoﬂavones in human urine. Clin. Chim. Acta. 223: 9-22.

Kelly, G.E., Joannou, G.E., Reeder, A.Y., Nelson, C. and Waring, M.A. (1995) The
variable metabolic response to dietary isoﬂavones in humans. Proc. Soc. Exp. Biol.
Med. 208: 40-43.

Kennedy, AR. (1993). In vitro studies of anticarcinogenic protease inhibitors. In:
Protease inhibitors as cancer chemoprevntive agents (Troll, W. and Kennedy, A.R.),
pp. 65-91. Plenum, New York, NY.

Kennedy, A.R., Beazer-Barclay, Y., Kinzler, K.W. and Newbeme, RM. (1996)
Suppression of carcinogenesis in the intestines of min mice by the soybean-derived
Bowman-Birk inhibitor. Cancer Res. 56: 679-682.

Kennedy, A.R., Billings, P.C., Wan, X.S. and Newbeme, RM. (2002) Effects of
Bowman-Birk inhibitor on rat colon carcinogenesis. Nutr. Cancer 43: 174-186.

Kim, Y.I. and Mason, J .B. (1996) Nutrition chemoprevention and gastrointestinal
cancers: A critical review. Nutr. Rev. 54: 259-279.

King, RA. and Bursill, DB. (1998) Plasma and urinary kinetics of the isoﬂavones
daidzein and genistein after a single soy meal in humans. Am. J. Clin Nutr 67: 867-
872.

Kirkman, L.M., Lampe, J .W., Campbell, D.R., Martinin, MC. and Slavin, J.L. (1995)
Urinary lignan and isoﬂavonoid excretion in men and women consuming vegetable

and soy diets. Nutr. Cancer 24: 1-12.

Kleibeuker, J .H., Welberg, J .W., Mulder, N.H., Van der Meer, R., Cats, A., Limburg,
A.J., Kreumer, W.M., Hardonk, M.J. and De Veries, E.G. (1993) Epithelial cell

184

proliferation in the sigmoid colon of patients with adenomatous polyps increases
during oral calcium supplementation. Br. J. Cancer 67: 500-503.

Klurfeld, D.M., Weber, M.M. and Kritchevsky, D. (1987) Inhibition of chemically
induced mammary and colon tumor production by caloric restriction in rats fed
increased dietary fat. Cancer Res. 47: 2759-2762.

Knight, DC. and Eden, J .A. (1995) Phytoestrogens: A short review. Maturitas 22: 167-
175.

Koratkar, R. and Rao, A.V. (1997) Effects of soya bean saponins on azoxymethane-
induced preneoplastic lesions in the colon of mice. Nutr. Cancer 27: 206-209.

Krishnan, K., Rufﬁn, MT and Brenner, DE. (1998) Clinical models of chemoprevention
for colon cancer. Hematol. Oncol. Clin. North Am. 12: 1079-1113.

Krishnan, K., Aoki, T., Rufﬁn, M.T., Normolle, D.P., Boland, CR. and Brenner, DE.
(2004) Effects of low dose aspirin (81mg) on proliferating cell nuclear antigen and
Amaranthus caudatus labeling in normal-risk and high-risk human subjects for
colorectal cancer. Cancer Detect. Prev. 28: 107-113.

Kubota, O. and Kino, I. (1995) Depressed adenomas of the colon in familial adenomatous
polyposis: Histology, immunohistochemical detection of proliferating cell nuclear
antigen (PCNA), and analysis of the background mucosa. Am. J. Surg. Pathol. 19:
318-327.

Kulling, S.E., Honig, D.M., Simat, T.J. and Metzler, M. (2000) Oxidative in vitro
metabolism of the soy phytoestrogens daidzein and genistein. J. Agric. Food Chem.
48: 4963-4972.

Kumar, S.P., Roy, S.J., Tokumo, K. and Reddy, BS. (1990) Effect of different levels of
calorie restriction on azoxymethane induced colon carcinogenesis in male F344 rats.
Cancer Res. 50: 5761-5766.

Kune, S., Kune, G. and Watson, L. (1987) Case-control study of dietary etiological
factors: The Melbourne Colorectal Cancer Study. Nutr. Cancer 9: 21-42.

Kuo, S.-M. (1996) Antiproliferative potency of structurally distinct dietary ﬂavoniods on
human colon cancer cells. Cancer Lett. 110: 41-48.

Kurarnoto, S. and Oohara, T. (1995) How do colorectal cancers develop? Cancer 75:
1534-1538.

Labayle, D., Fischer, D., Vielh, P., Drouhin, F ., Pariente, A., Bories, C., Duhamel, 0.,

Trousset, M. and Attali, P. (1991) Sulindac causes regression of rectal polyps in
familial adenomatous polyposis. Gastroenterol. 10: 635-639.

185

Lamprecht, SA. and Lipkin, M. (2001) Cellular mechanisms of calcium and vitamin D in
the inhibition of colorectal carcinogenesis. Ann. NY Acad. Sci. 952: 73-87.

Le Pessot, F., Michel, P., Paresy, M., Lemoine, F ., Hellot, M.F., Paillot, B., Scotte, M.,
Peillon, C. and Jacques, H. (2001) Cell proliferation in colorectal adenocarcinomas:
Comparison between Ki-67 immunostaining and bromodeoxyuridine uptake detected
by immunohistochemistry and ﬂow cytometry. Pathol. Res. Pract. 197: 411-418.

Leach, F .S., Nicolaides, N.C., Papadopoulos, N., Lin, B., Jen, J ., Parsons, R., Peltomaki,
P., Sistonen, P., Aaltonen, LA. and Nystrom-Lahti, M. (1993) Mutations of a mutS
homolog in hereditary non polyposis colorectal cancer. Cell 75: 1215-1225.

Lee, H.-M., Greeley, G.H., Jr. and Englander, E.W. (2000) Effects of aging on
expression of genes involved in regulation of proliferation and apoptosis in the
colonic epithelium. Mech. Ageing Dev. 155: 139-155.

Lee, H.-M., Greeley, G.H., Jr. and Englander, E.W. (2001) Age-associated changes in
gene expression patterns in the duodenum and colon of rats. Mech. Ageing Dev. 122:
355-371.

Li, H., Kramer, P.M., Lubet, R.A., Steele, V.E., Kelloff, GI. and Pereira, M.A. (1998)
Effect of calcium on azoxymethane-induced aberrant crypt foci and cell proliferation
in the colon of rats. Cancer Lett. 124: 39-46.

Lichtenstein, G.R., Deren, J ., Katz, K., Kennedy, AR. and Ware, J .H. (2002) The
Bowman-Birk protease inhibitor: A novel therapy for the induction of remission in
patients with active ulcerative colitis. Gastroenterol. 122: A60.

Lijinsky, W. and Ross, A.E. (1967) Production of carcinogenic polynuclear hydrocarbons
in the cooking of food. Food Cosmet. Toxicol. 5: 343-347.

Limburg, P.J., Anderson, K.E., Johnson, T.W., Jacobs, D.R., Jr., Lazovich, D., Hong,
C.P., Nicodemus, K.K. and Folsom, AR. (2005) Diabetes mellitus and subsite-

speciﬁc colorectal cancer risks in the Iowa Women's Health Study. Cancer
Epidemiol. Biomarkers Prev. 14: 133-137.

Linassier, C., Pierre, M., Le Peco, J .-B. and Pierre, J. (1990) Mechanism of action in
NIH-3T3 cells of genistein, an inhibitor of EGF receptor tyrosine kinase activity.
Biochem. Pharmacol. 39: 187-193.

Lipkin, M. (1971) Proliferation and differentiation of normal and neoplastic cells in the
colon of man. Cancer 28: 38-40.

Lipkin, M. (1975) Biology of large bowel cancer. Present status and research frontiers.
Cancer 36: 2319-2324.

186

Lipkin, M. (1984) Method for binary classiﬁcation and risk assessment of individuals
with familial polyposis based on [3H]TdR labelling of epithelial cells in colonic
crypts. Cell Tissue Kinet. 17: 209-222.

Lipkin, M., Blattner, W.A., Gardner, E.J., Burt, R.W., Lynch, H.T., Deschner, E.E.,
Winawer, S. and Fraumeni, J .F ., Jr. (1984) Classiﬁcation and risk assessment of
individuals with familial polyposis, Gardner's syndrome, and familial non-polyposis

colon cancer from [3H]thymidine labeling patterns in colonic epithelial cells. Cancer
Res. 44: 4201-4207.

Lipkin, M. and Newmark, H. (1985) Effect of added dietary calcium on colonic epithelial
cell proliferation in subjects at high risk for familial colonic cancer. N. Engl. J. Med.
313: 1381-1384.

Lipkin, M. (1987) Biomarkers of increased susceptibility to gastrointestinal cancer: Their
development and application to studies of cancer prevention. Gastroenterol. 92: 1083-
1086.

Lipkin, M., Enker, W.E. and Winawer, SJ. (1987) Tritiated-thymidine labeling of rectal
epithelial cells in 'non-prep' biopsies of individuals at increased risk for colonic
neoplasia. Cancer Letters 37: 153-161.

Lipkin, M. (1988) Biomarkers of increased susceptibility to gastrointestinal cancer: New
application to studies of cancer prevention in human subjects. Cancer Res. 48: 23 5-
245.

Lipkin, M., Friedman, E., Winawer, SJ. and Newmark, H. (1989) Colonic epithelial cell
proliferation in responders and nonresponders to supplemental dietary calcium.
Cancer Res. 49: 248-254.

Lipkin, M., Reddy, 8., Newmark, H. and Lamprecht, SA. (1999) Dietary factors in
human colorectal cancer. Annu. Rev. Nutr. 19: 545-586.

Lippman, S.M., Benner, SE. and Hong, W.K. (1994) Cancer chemoprevention. J. Clin.
Oncol. 12: 851-873.

Lockhart, G.L., Jones, BL. and F inney, KR (1978) Determination of coumestrol in

soybeans by high-performance liquid and thin-layer chromatography. Cereal Chem.
55: 967-972.

Locniskar, M., Nauss, K.M., Kaufmann, P. and Newbeme, PM. (1985) Interaction of

dietary fat and route of carcinogen administration on 1,2-dimethyl-hydrazine-induced
colon tumorigenesis in rats. Carcinogenesis 6: 349-354.

187

 

Lointier, P., Wildrick, D.M. and Boman, BM. (1992) The effects of steroid hormones on
a human colon cancer cell line in vitro. Anticancer Res. 12: 1327-1330.

Loren, D.E., Lewis, J. and Kochman, ML. (2003) Colon cancer: Detection and
prevention. Hematol. Oncol. Clin. North Am. 17: 553-574.

Lu, L.J., Grady, J.J., Marshall, M.V., Ramanujam, V.M. and Anderson, KB. (1995)
Altered time course of urinary daidzein and genistein excretion during chronic soya
diet in healthy male subjects. Nutr. Cancer 24: 311-323.

Lu, L.J., Lin, S.N., Grady, J .J ., Nagamani, M. and Anderson, KB. (1996) Altered kinetics
and extent of urinary daidzein and genistein excretion in women during chronic soya
exposure. Nutr. Cancer 26: 289-302.

Lucas, E.W. and Riaz, MN. (1995) Soy protein products: Processing and use. J. Nutr.
125: 573S-580S.

Luchtenborg, M., Weijenberg, M.P., Roemen, G.M., de Bruine, A.P., van den Brandt,
P.A., Lentjes, M.H., Brink, M., van Engeland, M., Goldbohm, RA. and de Goeij,
AF (2004) APC mutations in sporadic colorectal carcinomas from The Netherlands
Cohort Study. Carcinogenesis 25: 1219-1226.

Luebeck, E.G. and Moolgavkar, SH. (2002) Multistage carcinogenesis and the incidence
of colon cancer. Proc. Natl. Acad. Sci. USA 99: 15095-15100.

Lupton, J .R., Steibach, G., Chang, W.C., O’brien, B.C., Wiese, S., Stolzfus, C.L., Glober,
G.A., Wargovich, M.J., McPherson, RS. and Winn, R.J. (1996) Calcium
supplementation modiﬁes the relative amounts of bile acids in bile and affects key
aspects of human colon physiology. J. Nutr. 126: 1421-1428.

Macklin, M. (1960) Inheritance of cancer of the stomach and large intestine. J. Nat.
Cancer Inst. 24: 551-571.

MacLennan, R., Macrae, F., Bain, C., Battistutta, D., Chapuis, P., Gratten, H., Lambert,
J ., Newland, R.C., Ngu, M. and Russell, A. (1995) Randomized trial of intake of fat,
ﬁber, and beta carotene to prevent colorectal adenomas. The Australian Polyp
Prevention Project. J. Nat. Cancer Inst. 87: 1760-1766.

Macquart-Moulin, G., Riboli, E., Comee, J., Charnay, B., Berthezene, P. and Day, N.
(1986) Case-control study on colorectal cancer and diet in Marseilles. Int. J. Cancer
38: 183-191.

Majumdar, A.P.N. (1990) Role of tyrosine kinases in gastrin induction of ornithine
decarboxylase in colonic mucosa. Am. J. Physiol. 259: G626-G630.

188

Makishima, M., Honma, Y., Hozumi, M., Sampi, K., Hattori, M., Umezawa, K. and
Motoyoshi, K. (1991) Effects of inhibitors of protein tyrosine kinase activity and/or
phosphatidylinositol turnover on differentiation of some human myelomonocytic
leukemia cells. Leuk. Res. 15: 701—708.

Malecka-Panas, E., Kordek, R., Biemat, W., Tureaud, J., Liberski, PP. and Majumdar,
AP (1997) Differential activation of total and EGF receptor (EGF-R) tyrosine kinase
(tyr-k) in the rectal mucosa in patients with adenomatous polyps, ulcerative colitis
and colon cancer. Hepato-gastroenterol. 44: 435-440.

Maltzman, T., Whittington, J ., Driggers, L., Stephens, J. and Ahnen, DJ. (1997) AOM-
induced mouse colon tumors do not express full-length APC protein. Carcinogenesis
18: 2435-2439.

Mark, SD. (1996) Deﬁning and analyzing cohorts using molecular markers of cancer
risk. J. Cell. Biochem. Suppl. 25: 69-79.

Markiewicz, L., Garey, J ., Adlercreutz, H. and Gurpide, E. (1993) In vitro bioassays of
non-steriodal phytoestrogens. J. Steroid Biochem Molec. Biol. 45: 399-405.

Maron, D.M. and Ames, B.N. (1983) Revised methods for the Salmonella mutagenicity
test. Mutat. Res. 113: 173-215.

Martin, P.M., Horwitz, K.B., Ryan, D.S. and McGuire, W.L. (1978) Phytoestrogen
interaction with estrogen receptors in human breast cancer cells. Endocrinology 103:
1860-1867.

Martinez, ME. and Willett, WC. (1998) Calcium, vitamin D, and colorectal cancer: A

review of the epidemiologic evidence. Cancer Epidemiol. Biomarkers Prev. 7: 163-
168.

Masaoka, Y., Watanabe, H., Katoh, 0., Ito, A. and Dohi, K. (1998) Effects of miso and
NaCl on the development of colonic aberrant crypt foci induced by azoxymethane in
F344 rats. Nutr. Cancer 32: 25-28.

Maskarinec, G., Singh, S., Meng, L. and Franke, AA. (1998) Dietary soy intake and
urinary isoﬂavone excretion among women from a multiethnic population. Cancer
Epidemiol. Biomarkers Prev. 7: 613-619.

Mathers, J.C., Kooshkghazi, M. and Coaker, J. (1998) Is crypt cell proliferation a reliable
biomarker of intestinal cancer risk? Proc. Nutr. Soc. j57: 43A.

Matsukawa, Y., Marni, N., Sakai, T., Satomi, Y., Yoshida, M., Matsumoto, K., Nishino,

H. and Aoike, A. (1993) Genistein arrests cell cycle progression at Gz—M. Cancer
Res. 53: 1328-1331.

189

Matsumoto, T., Iida, M., Yao, T. and Fujishima, M. (1994) Role of nonpolypoid
neoplastic lesions in the pathogenesis of colorectal cancer. Dis. Colon Rectum 37: 45-
455.

Mayr, U., Butsch, A. and Schneider, S. (1992) Validation of two in vitro test systems for

estrogenic activities with zearalenone, phytoestrogens and cereal extracts. Toxicology
74: 135-149.

McCormick, D., Yu, G, Hobbs, C. and Hall, PA. (1993) The relevance of antibody
concentration to the immunohistological quantiﬁcation of cell proliferation-associated
antigens. Histopathology 22: 543-547.

McIntosh, G.H., Regester, G.O., Le Leu, R.K., Royle, P.J. and Smithers, G.W. (1995)
Dairy proteins protect against dimethylhydrazine-induced intestinal cancers in rats. J.
Nutr. 125: 809-816.

McKeown-Eyssen, GE. and Bright-See, E. (1984) Dietary factors in colon cancer:
International relationships. Nutr. Cancer 6: 160-170.

McLellan, EA. and Bird, RP. (1988) Aberrant crypts: Potential preneoplastic lesions in
the murine colon. Cancer Res. 48: 6187-6192.

McLellan, E.A., Medline, A. and Bird, R.P. (1991a) Dose response and proliferative
characteristics of aberrant crypt foci: Putative preneoplastic lesions in rat colon.
Carcinogenesis 12: 2093-2098.

McLellan, E.A., Medline, A. and Bird, R.P. (1991b) Sequential analyses of the growth
morphological characteristics of aberrant crypt foci: Putative preneoplastic lesions.
Cancer Res. 51: 5270-5274.

McMichael, A.J. and Potter, J .D. (1980) Reproduction, endogenous and exogenous sex
hormones, and colon cancer: A review and hypothesis. J. Natl. Cancer Inst. 65: 1201-
1207.

McShane, L.M., Kulldorff, M., Wargovich, M.J., Woods, C., Purewal, M., Freedman,
L.S., Corle, D.K., Burt, R.W., Mateski, D.J., Lawson, M., Lanza, E., O'Brien, 3.,
Lake, W., Jr., Moler, J. and Schatzkin, A. (1998) An evaluation of rectal mucosal
proliferation measure variability sources in the polyp prevention trial: Can we detect
informative differences among individuals' proliferation measures amid the noise?
Cancer Epidemiol. Biomarkers Prev. 7: 605-612.

Messadi, D.V., Billings, P.C., Shklar, G. and Kennady, AR. (1986) Inhibition of oral
carcinogenesis by a protease inhibitor. J. Nat. Cancer Inst. 76: 447-452.

Messina, M. Persky, V., Setchell, K. D. and Barnes, S. (1994) Soy intake and cancer risk:
A review of the in vitro and in vivo data. Nutr. Cancer 21: 113131.

190

 

Messina, M. (1995) Isoﬂavone intakes by Japanese were overestimated. Am. J. Clin.
Nutr. 62: 645-649.

Messina, M. and Bennink, M. (1998) Soyfoods, isoﬂavones and risk of colonic cancer: A
review of the in vitro and in vivo data. Ballieres Clinical Endocr. Metab. 12: 708-728.

Messina, M. and Erdman, J .W. (1998) The role of soy in preventing and treating chronic
disease. Am. J. Clin. Nutr. 68: 1329S-1344S.

Messina, M.J. (1999) Legumes and soybeans: Overview of their nutritional proﬁles and
health effects. Am. J. Clin. Nutr. 70: 439S-450S.

Michaels, K.B., Giovannucci, E., Joshipura, K.J., Rosner, B.A., Stampfer, M.J., Fuchs,
C.S., Colditz, G.A., Speizer, F .E. and Willet, WC. (2000) Prospective study of fruit
and vegetable consumption and incidence of colon and rectal cancers. J. Nat. Cancer
Inst. 92: 1740-1752.

Milgate, J. and Roberts, D.C.K. (1995) The nutritional and biological signiﬁcance of
saponins. Nutr. Res. 15: 1223-1249.

Mills, S.J., Shepherd, N.A., Hall, P.A., Hastings, A., Mathers, J .C. and Gunn, A. (1995)
Proliferative compartment deregulation in the non-neoplastic colonic epithelium of
familial adenomatous polyposis. Gut 36: 391-394.

Mills, S.J., Mathers, J .C., Chapman, P.D., Burn, J. and Gunn, A. (2001) Colonic crypt
cell proliferation state assessed by whole crypt microdissection in sporadic neoplasia
and familial adenomatous polyposis. Gut 48: 41-46.

Minamoto, T., Sawaguchi, K., Ohta, T., Itoh, T. and Mai, M. (1994) Superﬁcial-type
adenomas and adenocarcinomas of the colon and rectum: A comparative
morphological study. Gastroenterol. 106: 1436-1443.

Moore, W.E.C. and Holdeman, L.V. (1975) Discussion of current bacteriological
investigations of the relationship between intestinal ﬂora, diet, and colon cancer.
Cancer Res. 35: 3418-3420.

Morton, M.S., Matos-Ferreira, A., Abranches-Monteiro, L., Correia, R., Blacklock, N.,
Chan, P.S.F., Cheng, C., Lloyd, S., Chieh-ping, W. and Grifﬁths, K. (1997)
Measurement and metabolism of isoﬂavonids and lignans in the human male. Cancer
Lett.114: 145-151.

Murillo, G., Choi, J .K., Pan, 0., Constantinou, AI. and Mehta, KG. (2004) Efﬁcacy of

garbanzo and soybean ﬂour in suppression of aberrant crypt foci in the colons of CF -
1 mice. Anticancer Res. 24: 3049-3055.

191

Murphy, P.A., Song, T., Buseman, G., Barua, K., Beecher, G.R., Trainer, D. and Holden,
J. (1999) Isoﬂavones in retail and institutional soy foods. J. Agric. Food Chem. 47:
2697-2704. .

Muto, T., Kamiya, J ., Swada, T., Konishi, F., Sugihara, K., Kubota, Y., Adachi, M.,
Agawa, S., Saito, Y. and Morioka, Y. (1985) Small ﬂat adenoma of the large bowel

with special reference to its clinicopathological features. Dis. Colon Rectum 28: 847-
851.

Nagata, C., Takatsuka, N., Kurisu, Y. and Shimizu, H. (1998) Decreased serum total
cholesterol concentration is associated with high intake of soy products in Japanese
men and women. J. Nutr. 128: 209-213.

Nair, P.P., Turjman, N., Kessie, G., Calkins, B., Goodman, G.T., Davidovitz, H. and
Nimmagadda, G. (1984) Diet, nutrition intake, and metabolism in populations at high

and low risk for colon cancer. Dietary cholesterol, B-sitosterol, and stigmasterol. Am.
J. Clin. Nutr. 40: 927-930.

Neugut, A.I., Horvath, K., Whelan, R.L., Terry, M.B., Garbowski, G.C., Bertram, A.,
Forde, K.A., Treat, MR. and Waye, J. (1996) The effect of calcium and vitamin

supplements on the incidence and recurrence of colorectal adenomatous polyps.
Cancer 78: 723-728.

Newcomb, RA. and Storer, BE. (1995) Postmenopausal hormone use and risk of large-
bowel cancer. J. Nat. Cancer Inst. 87: 1067-1071.

Newmark, H., Yang, K., Lipkin, M., Kopelovich, L., Liu, Y., Fan, K. and Shinozaki, H.
(2001) A westem-style diet induces benign and malignant neoplasms in the colon of
normal C57B1/6 mice. Carcinogenesis 22: 1871-1875.

Newmark, H.L., Wargovich, M.J. and Bruce, W.R. (1984) Colon cancer and dietary fat,
phosphate, and calcium: A hypothesis. J. Nat. Cancer Inst. 72: 1323-1325.

Newmark, H.L., Lipkin, M. and Maheshwari, N. (1991) Colonic hyperproliferation
induced in rats by nutritional-stress diets containing four components of a human
westem-style diet. Am. J. Clin. Nutr. 54: 209-214.

Nishisho, 1., Nakamura, Y., Miyoshi, Y., Miki, Y., Ando, H., Horii, A., Koyama, K.,
Utsunomiya, J ., Baba, S. and Hedge, P. (1991) Mutations of chromosome 5q21 genes
in FAP and colorectal cancer patients. Science 253: 665-669.

Nusko, G., Altendorf-Hofmann, A. and Herrnanek, P. (1994). Correlation of polypoid
lesions in the rectosigmoid and proximal colon. Tenth World Congress of

Gastroenterology, Lon Angeles, CA.

Oakenfull, D. (1981) Saponins in food - a review. Food Chem. 7: 19-40.

192

O'Brien, M.J., Winawer, S.J., Zauber, A.G., Gottlieb, L.S., Stemberg, S.S., Diaz, 8.,
Dickersin, G.R., Ewing, S., Geller, S. and Kasimian, D. (1990) The National Polyp
Study. Patient and polyp characteristics associated with high grade dysplasia in
colorectal adenomas. Gastroenterol. 98: 371-379.

Oh, Y. and Sung, MK. (2001) Soybean saponins inhibit cell proliferation by suppressing
PKC activation and induce differentiation of HT-29 human colon adenocarcinoma
cells. Nutr. Cancer 39: 132-138.

Ottaviani, G., Lavezzi, A.M., De Ruberto, F., Giuseppe, F. and Matturri, L. (1999) The
prognostic value of cell proliferation in colorectal adenomas assessed with tritiated
thymidine and anti-proliferating cell nuclear antigen. Cancer Detect. Prev. 23: 57-63.

Owen, R.J. and Jones, AL. (1974) Epithelial cell specialization within human Peyer's
patches: An ultrastructural study of intestinal lymphoid follicles. Gastroenterol. 66:
1 89-203.

Paganelli, G.M., Santucci, R., Biasco, G., Miglioli, M. and Barbara, L. (1990) Effect of
sex and age on rectal cell renewal in humans. Cancer Lett. 53: 117-121.

Paganelli, G.M., Biasco, G., Santucci, R., Brandi, G., Lalli, A.A., Miglioli, M. and
Barbara, L. (1991) Rectal cell proliferation and colorectal cancer risk level in patients
with nonfarnilial adenomatous polyps of the large bowel. Cancer 68: 2451-2454.

Pagliacci, M.C., Smacchia, M., Migliorati, G., Grignani, F ., Riccardi, C. and Nicoletti, I.
(1994) Growth-inhibitory effects of the natural phytoestrogen genistein in MCF-7
human breast cancer cells. Eur. J. Cancer 30A: 1675-1682.

Papadopoulos, N., Nicolaides, N.C., Wei, Y.-F., Ruben, S.M., Carter, K.C., Rosen, C.A.,
Haseltine, W.A., Fleischmann, R.D., Fraser, CM. and Adams, MD. (1994) Mutation
of a mutL homolog in hereditary colon cancer. Science 261: 1625-1629.

Parkin, D.M. (1989) Cancers of the breast, endometrium and ovary: Geographic
correlations. Eur. J. Cancer. Clin. Oncol. 25: 1917-1925.

Parkin, D.M. (2004) International variation. Oncogene 23: 6329-6340.

Paspatis, G.A., Zizi, A., Chlouverakis, G.J., Giannikaki, E.S., Vasilakaki, T.,
Elemenoglou, I. and Karamanolis, DC. (1998) Proliferative patterns of rectal mucosa
as predictors of advanced colonic neoplasms in routinely processed rectal biopsies.

Am. J. Gastroenterol. 93: 1472-1477.

Pennisi, E. (1998) Building a better aspirin. Science 280: 1191-1192.

193

 

Pereira, M.A., Barnes, L.H., Rassman, V.L., Kelloff, G.V. and Steele, V.E. (1994) Use of
azoxymethane-induced foci of aberrant crypts in rat colon to identify potential cancer
chemopreventive agents. Carcinogenesis 15: 1049—1054.

Peters, R.K., Pike, M.C., Garabrant, D. and Mack, TM. (1992) Diet and colon cancer in
Los Angeles County, California. Cancer Causes Control 3: 457-473.

Peterson, G. and Barnes, S. (1991) Genistein inhibition of the growth of human breast

cancer cells: Independence from estrogen receptors and the multi-drug resistance
gene. Biochem. Biophys. Res. Commun. 179: 661—667.

Peterson, G. and Barnes, S. (1993) Genistein and biochanin A inhibit the growth of
human prostate cancer cells but not epidermal growth factor receptor tyrosine
autophosphorylation. Prostate 22: 335—345.

Peterson, G. and Barnes, S. (1996) Genistein inhibits both estrogen and growth factor-

stimulated proliferation of human breast cancer cells. Cell Growth and Differentiation
7: 1345-1351.

Pfeiffer, R., McShane, L.M., Wargovich, M.J., Burt, R.W., Kikendall, W., Lawson, M.,
Lanza, E. and Schatzkin, A. (2003) The effect of a low-fat, high ﬁber, fruit and
vegetable intervention on rectal mucosal proliferation. Cancer 98: 1161 -1 168.

Piard, F., Martin, M., Boutron, M.C., Hillon, P., Hammann, A. and Martin, F. (1994)
Effect of different doses of dietary calcium on murine colonic cell proliferation. Eur.
J. Cancer Prev. 3: 215-221.

Pierre, F., Tache, S., Petit, C.R., Van der Meer, R. and Corpet, DE. (2003) Meat and
cancer: Haemoglobin and haemin in a low-calcium diet promote colorectal
carcinogenesis at the aberrant crypt stage in rats. Carcinogenesis 24: 1683-1690.

Pierre, F., Freeman, A., Tache, S., Van der Meer, R. and Corpet, DE. (2004) Beef meat
and blood sausage promote the formation of azoxymethane-induced mucin-depleted
foci and aberrant crypt foci in rat colons. J. Nutr. 134: 271 1-2716.

Polyak, K., Hamilton, S., Vogelstein, B. and Kinzler, K.W. (1996) Early alteration of
cell-cycle-regulated gene expression in colorectal neoplasis. Am. J. Pathol 149: 381-
387.

Poole, C. (1989) A case-control study of diet and colon cancer. Doctoral Dissertation,
Harvard School of Public Health, Boston, MA.

Potter, J .D. and McMichael, A.J. (1986) Diet and cancer of the colon and rectum: A case-
control study. J. Nat. Cancer Inst. 76: 557-569.

194

Potter, J .D. (1992) Reconciling the epidemiology, physiology, and molecular biology of
colon cancer. JAMA 268: 1573-1577.

Potter, J.D. (1993) Colon cancer: Review of the epidemiology. Epidemiol. Rev. 15: 499-
545.

Potter, J .D. (1996) Nutrition and colorectal cancer. Cancer Causes Control 7: 127-146.

Potter, J .D. (1999) Colorectal cancer: Molecules and populations. J. Nat. Cancer Inst. 91:
916-932.

Prelich, G., Tan, C.K., Kostura, M., Mathews, M.B., So, A.G., Downey, KM. and
Stillman, B. (1987) Functional identity of proliferating cell nuclear antigen and a
DNA polymerase-delta auxillary protein. Nature 326: 517-520.

Pretlow, T., Oriordan, M., Somich, G., Amini, S. and Pretlow, T. (1992a) Aberrant crypts
correlate with tumor incidence in F344 rats treated with azoxymethane and phytate.
Carcinogenesis 13: 1509-1512.

Pretlow, T.P., Barrow, B.J., Ashton, W.S., O'Riordan, M.A., Pretlow, T., Jureisek, J .A.
and Stellato, T. (1991) Aberrant crypts: Putative preneoplastic foci in human colonic
mucosa. Cancer Res. 51: 564-567.

Pretlow, T.P., O'Riordan, M.A., Pretlow, T.G. and Stellato, T.A. (1992b) Aberrant crypts
in human colonic mucosa: Putative preneoplastic lesions. J. Cell. Biochem. Suppl.
16G: 55-62.

Pretlow, T.P., Brasitus, T.A., Fulton, N.C., Cheyer, C. and Kaplan, EL. (1993) K-ras
mutations in putative preneoplastic lesions in human colon. J. Nat. Cancer Inst. 85:
2004-2007.

Pretlow, T.P. (1995) Aberrant crypt foci and K-ras mutations: Earliest recognized players
or innocent bystanders in colon carcinogenesis? Gastroenterol. 108: 600-603.

Purba, M., Lukito, W., Wahlqvist, M.L., Kouris-Blazos, A., Hadisaputro, S., Lestiani, L.,
Wattanapenpaiboon, N. and Kamso, S. (1999) Food intake and eating patterns of

Indonesian elderly before the 1998 economic crisis. Asia Paciﬁc J. Clin. Nutr. 8: 200-
206.

Quilliot, D., Belbraouet, S., Pelletier, X., Gueant, J .L., F loquet, J. and Debry, G. (1999)
Inﬂuence of a high-calcium carbonate diet on the incidence of experimental colon
cancer in rats. Nutr. Cancer 34: 213-219.

Rafter, J.J., Eng, V.W., Furrer, R., Medline, A. and Bruce, W.R. (1986) Effects of
calcium and pH on the mucosal damage produced by deoxycholic acid in the rat
colon. Gut 27: 1320-1329.

195

Raicht, R.F., Cohen, B., Fazzini, E., Sarwal, AN. and Takahashi, M. (1980) Protective
effect of plant sterols against chemically induced colon tumors in rats. Cancer Res.
40: 403-405.

Raid, A., Oliver, B., Abdelrahman, A., Sha'aﬁ, RI. and Hajjar, J.J. (1993) Role of
tyrosine kinase and phosphotyrosine phosphatase in growth of the intestinal crypt cell
(IEC-6) line. Proc. Soc. Exp. Biol. Med. 202: 435-439.

Raines, E.W. and Ross, R. (1995) Biology of atherosclerotic plaque formation: Possible
role of growth factors in lesion development and the potential impact of soy. J. Nutr.
125: 624S-63OS.

Rao, A.V. and Janezic, SA. (1992) The role of dietary phytosterols in colon
carcinogenesis. Nutr. Cancer 18: 43-52.

Rao, A.V., Janezic, S.A., Friday, D. and Kendall, CW. (1992) Dietary cholesterol
enhances the induction and development of colonic preneoplastic lesions in
C57BL/6J and BALB/CJ mice treated with azoxymethane. Caner Lett. 63: 249-257.

Rao, A.V. and Gurﬁnkel, D.M. (2000) The bioactivity of saponins: Triterpenoid and
steroidal glycosides. Drug Metab. Drug Interact 17: 211-235.

Rao, C.V., Wang, C., Simi, B., Lubet, R., Kelloff, G.J., Steele, V. and Reddy, BS. (1997)
Enhancement of experimental colon cancer by genistein. Cancer Res. 57: 3717-3722.

Rauth, S., Kichina, J. and Green, A. (1997) Inhibition of growth and induction of
differentiation of metastatic melanoma cells in vitro by genistein: Chemosensitivity is
regulated by cellular p53. Br. J. Cancer 75: 1559—1566.

Reddy, 3., Rao, C. and Seibert, K. (1996) Evaluation of a cycloxygenase-2 inhibitor for

potential chemopreventive properties in colon carcinogenesis. Cancer Res. 56: 4566-
4569.

Reddy, B.S., Narisawa, T. and Weisburger, J .H. (1976) Effect of a diet with high levels
of protein and fat on colon carcinogenesis in F344 rats treated with 1,2-
dimethylhydrazine. J. Nat. Cancer Inst. 57: 567-569.

Reddy, BS. and Wynder, EL. (1977) Metabolic epidemiology of colon cancer. Fecal

bile acids and neutral sterols in colon cancer patients and patients with adenomatous
polyps. Cancer 39: 2533-2539.

Reddy, B.S., Hedges, A., Laakso, K. and Wynder, EL. (1978) Fecal constituents of a

high-risk North American and a low-risk Finnish population for the development of
large bowel cancer. Cancer Lett. 4: 217-222.

196

Reddy, BS. (1986) Dietary fat and cancer: Speciﬁc action or caloric effect. J. Nutr. 116:
1132-1135. -

Reddy, B.S., Maruyama, H. and Kelloff, G. (1987) Dose-related inhibition of colon
carcinogenesis by dietary piroxicam, a nonsteroidal antiinﬂammatory drug, during
different stages of rat colon tumor development. Cancer Res. 47: 5340-5346.

Reddy, B.S., Burill, C. and Rigotty, J. (1991) Effect of diets high in omega-3 and omega-
6 fatty acids on initiation and post-initiation stages of colon carcinogenesis. Cancer
Res. 51: 1270-1277.

Reddy, B.S., Rao, C.V., Rivenson, A. and Kelloff, G.J. (1993) Inhibitory effect of aspirin
on azoxymethane-induced colon carcinogenesis in F344 rats. Carcinogenesis (Lond.)
14: 1493-1497.

Reeves, P.G., Nielsen, PH. and Fahey, G.C., Jr. (1993) AIN-93 puriﬁed diets for
laboratory rodents: Final report of the American Institute of Nutrition ad hoc writing
committee on the reformulation of the AIN-76A rodent diet. J. Nutr. 123: 1939-1951.

Renehan, A.G., O'Dwyer, S.T., Haboubi, NJ. and Potten, CS. (2002) Early cellular
events in colorectal carcinogenesis. Colorectal Disease 4: 76-89.

Reshef, R., Rozen, P., Fireman, Z., Fine, N., Barzilai, M., Shasha, SM. and Shkolnik, T.
(1990) Effect of a calcium-enriched diet on the colonic epithelial hyperproliferation

induced by N-Methyl-N'-nitro-N-nitrosoguanidine in rats on a low calcium and fat
diet. Cancer Res. 50: 1764-1767.

Richter, A., Yang, K., Richter, F., Lynch, HT. and Lipkin, M. (1993) Morphological and
morphometric measurements in colorectal mucosa of subjects at increased risk for
colonic neoplasia. Cancer Lett. 74: 65-68.

Ries, L.A.G., Eisner, M.P., Kosary, C.L., Hankey, B.F., Miller, B.A., Clegg, L. and
Mariotto, A. (2005). SEER cancer statistics review, 1975-2002. Bethesda, MD,
National Cancer Institute.

Risio, M., Lipkin, M., Candelaresi, G., Bertone, A., Coverlizza, S. and Rossini, F .P.
(1991) Correlations between rectal mucosa cell proliferation and the clinical and

pathological features of nonfamilial neoplasia of the large intestine. Cancer Res. 51 :
1917-1921.

Roncucci, L., Ponz de Leon, M., Scalmati, A., Malagoli, G., Pratissoli, S., Perini, M. and

Chahin, NJ. (1988) The inﬂuence of age on colonic epithelial cell proliferation.
Cancer 62: 2373-2377.

197

Roncucci, L., Stamp, D., Medline, A., Cullen, J.B. and Bruce, W. (1991) Identiﬁcation
and quantiﬁcation of aberrant crypt foci and microadenomas in the human colon.
Hum. Pathol. 22: 287-294.

Ronis, M.J., Rowlands, J .C., Hakkak, R. and Badger, TM. (1999) Altered expression and
glucocorticoid-inducibility of hepatic CYP3A and CYP2B enzymes in male rats fed
diets containing soy protein isolate. J. Nutr. 129: 1958—1965.

Rooney, P.S., Clarke, P.A., Gifford, K.A., Hardcastle, J .D. and Armitage, NC. (1993)
The identiﬁcation of high and low risk group for colorectal cancer using rectal
mucosal crypt cell production rate (CCPR). Br. J. Cancer 68: 172-175.

Ross, M.H. and Bras, G. (1971) Lasting inﬂuence of early caloric restriction on
prevalence of neoplasms in the rat. J. Nutr. 47: 1095-1113.

Rozen, P., F irman, Z., Fine, N., Wax, Y. and Ron, E. (1989) Oral calcium suppresses
increased rectal epithelial proliferation of person at risk for colorectal cancer. Gut 30:
650-655.

Rozen, P., Fireman, Z., Fine, N., Chetrit, A. and Lubin, F. (1990) Rectal epithelial
proliferation characteristics of ﬁrst degree relatives of sporadic colon cancer patients.
Cancer Lett. 51: 127-132.

Rozen, P., Fireman, Z., Fine, N., Chetrit, A. and Zajicek, G. (1991) Rectal epithelial
proliferation in persons post sporadic colorectal neoplasia. Cancer Lett. 59: 1-8.

Rozen, P. (1992) An evaluation of rectal epithelial proliferation measurement as
biomarker of risk for colorectal neoplasia and response in intervention studies. Eur. J.
Cancer Prev. 1: 215-224.

Rubio, C. and Shetye, J. (1994) Flat adendoma-adenocarcinoma sequence in the colon of
rats. Dis. Colon Rectum 37: 1300-1306.

Sahin, A.A., Ro, J .Y., Brown, R.W., Ordonez, N.G., Cleary, K.R., El-Naggar, A.K.,
Wilson, P. and Ayala, A.G. (1994) Assessment of Ki-67-derived tumor proliferative
activity in colorectal adenocarcinomas. Mod. Pathol. 7: 17-22.

Saito, K. and Mori, M. (1999) Morphologic progression of pedunculated submucosal
invasive colon cancer. Dig. Dis. Sci. 44: 1605-1609.

Sam, J .S., Lynch, HT. and Burt, R.W. (1990) Abnormalities of lectin histochemistry in
familial polyposis coli and hereditary nonpolyposis colorectal cancer. Cancer 66:
502-508.

Sanderson, P., Johnson, I.T., Mathers, J.C., Powers, H.J., Downes, C.S., McGlynn, A.P.,
Dare, R., Kampman, E., Pool-Zobel, B.L., Bingham, SA. and Rafter, J .J . (2004)

198

 

Emerging diet-related surrogate end points for colorectal cancer: UK food standards
agency diet and colonic health workshop report. Br. J. Nutr. 91: 315-322.

Sandler, RS. (1996) Epidemiology and risk factors for colorectal cancer. Gastroenterol.
Clin. North Am. 25: 717-735.

Sandler, R.S., Baron, J .A., Tosteson, T.D., Mandel, J .S. and Haile, R.W. (2000) Rectal
mucosal proliferation and risk of colorectal adenomas: Results from a randomized
controlled trial. Cancer Epidemiol. Biomarkers Prev. 9: 653-656.

Santibanez, J .F., Navarro, A. and Martinez, J. (1997) Genistein inhibits proliferation and

in vitro invasive potential of human prostatic cancer cell lines. Anticancer Res. 17:
1199-1204.

Scalmati, A. and Lipkin,»M. (1993) Proliferation and differentiation biomarkers in
colorectal mucosa and their application to chemoprevention studies. Environ. Health
Perspect. 99: 169-173.

Schatzkin, A., Lanza, E., Corle, D., Lance, P., lber, F., Caan, B., Shike, M., Weissfeld, J.,
Burt, R., Cooper, M.R., Kikendall, J .W. and Cahill, J. (2000) Lack of effect of a low-

fat, high-ﬁber diet on the recurrence of colorectal adenomas. Polyp Prevention Trial
Study Group. N. Engl. J. Med. 342: 1149-1155.

Scholar, EM. and Toewa, ML. (1994) Inhibition of invasion murine mammary
carcinoma cells by the tyrosine kinase inhibitor genistein. Cancer Lett. 87: 159-162.

Scott, R.J., Hall, P.A., Haldane, J .S., van Noorden, S., Price, Y., Lane, DP. and Wright,
NA. (1991) A comparison of immunohistochemical markers of cell proliferation with
experimentally determined growth fraction. J. Pathol. 165: 173-178.

Seo, A. and Morr, CV. (1984) Improved high-performance liquid chromatographic
analysis of phenolic acids and isoﬂavonoids from soybean protein products. J. Agric.
Food Chem. 32: 530-533.

Setchell, K.D., Zimmer-Nechemias, L., Cai, J. and Heubi, J .E. (1997) Exposure of infants
to phyto-oestrogens from soy-based infant formula. Lancet 350: 23-27.

Setchell, K.D. and Cassidy, A. (1999) Dietary isoﬂavones: Biological effects and
relevance to human health. J. Nutr. 129: 75 8S—767S.

Setchell, K.D.R., Borriello, S.P., Hulme, P., Kirk, D.N. and Axelson, M. (1984)
Nonsteroidal estrogens of dietary origin: Possible roles in hormone-dependent

disease. Am. J. Clin. Nutr. 40: 569-578.

Setchell, K.D.R. (2000) Absorption and metabolism of soy isoﬂavones - from food to
dietary supplements and adults to infants. J. Nutr. 130: 654S-6558.

199

Setchell, K.D.R., Brown, N.M., Desai, P., Zimmer-Nechemias, L., Wolfe, B.E., Brashear,
W.T., Kirscher, A.S., Cassidy, A. and Heubi, J. (2001) Bioavailability of pure

isoﬂavones in healthy humans and analysis of commercial soy isoﬂavone
supplements. J. Nutr. 131: l362S-137SS.

Setchell, K.D.R. and Cole, SJ. (2003) Variations in isoﬂavone levels in soy foods and

soy protein isolates and issues related to isoﬂavone databases and food labeling. J.
Agric. Food Chem. 51: 4146-4155.

Shamsuddin, A.M., Ullah, A. and Chakravarthy, AK. (1989) Inositol and inositol
hexaphosphate suppress cell proliferation and tumor formation in CD-1 mice.
Carcinogenesis 10: 1461-1463.

Shamsuddin, A.M. (1995) Inositol phosphates have novel anticancer function. J. Nutr.
125: 7258-732S.

Shamsuddin, A.M., Vucenik, I. and Cole, KB. (1997) 1P6: A novel anti-cancer agent.
Life Sci. 61: 343-354.

Shao, Z.-M., Alpaugh, M.L., Fontana, J .A. and Barsky, SH. (1998) Genistein inhibits
proliferation similarly in estrogen receptor-positive and negative human breast
carcinoma cell lines characterized by P21W ”cm induction, Gz/M arrest, and
apoptosis. J. Cell. Biochem. 69: 44-54.

Sharma, RA. (2000) Cancer chemoprevention: A clinical reality. J. Roy. Soc. Med. 93:
518-520.

Sharma, R.A., Manson, M.M., Gescher, A. and Steward, WP. (2001) Colorectal cancer
chemoprevention: Biochemical targets and clinical development of promising agents.
Eur. J. Cancer 37: 12-22.

Shpitz, B., Bomstein, Y., Mekori, Y., Cohen, R., Kaufman, Z., Grankin, M. and
Bemheim, J. (1997) Proliferating cell nuclear antigen as a marker of cell kinetics in

aberrant crypt foci, hyperplastic polyps, adenomas, and adenocarcinomas of the
human colon. Am. J. Surg. 174: 425-430.

Shureiqi, 1., Reddy, P. and Brenner, DE. (2000) Chemoprevention: General perspective.
Crit. Rev. Oncol./Hematol. 33: 157-167.

Sidhu, G.S. andOakenfull, D.G. (1986) A mechanism for the hypocholesterolaemic
activity of saponins. Br. J. Cancer 55: 643-649.

Silvestrini, R. (1994) Cell kinetics: Prognostic and therapeutic implications in human
tumours. Cell. Prolif. 27: 579-596.

200

 

Singh, J ., Kellof, G. and Reddy, BS. (1993) Intermediate biomarkers of colon cancer:
Modulation of expression of ras oncogene by chemopreventive agents during
azoxymethane induced colon carcinogenesis. Carcinogenesis 12: 699-704.

Sinicrope, F .A., Half, E., Morris, J.S., Lynch, P.M., Morrow, J.D., Levin, B., Hawk, E.T.,
Cohen, D.S., Ayers, G.D., Stephens, LC. and Familial Adenomatous Polyposis Study
Group (2004) Cell proliferation and apoptotic indices predict adenoma regression in a

placebo-controlled trial of celecoxib in familial adenomatous polyposis patients.
Cancer Epidemiol. Biomarkers Prev. 13: 920-927.

Siu, I.M., Pretlow, T.G., Amini, SB. and Pretlow, T.P. (1997) Identiﬁcation of dysplasia
in human colonic aberrant crypt foci. Am. J. Pathol. 150: 1805-1813.

Sjoqvist, U., Lofberg, R., Ost, A. and Tribukait, B. (2002) Age-and site-dependent cell
cycle composition of the normal human colonic mucosa. Anticancer Res. 22: 3437-
3442.

Slattery, M.L., Caan, B.J., Potter, J.D., Berry, T.D., Coates, A., Dundan, D. and Edwards,
S.L. (1997) Dietary energy sources and colon cancer risk. Am. J. Epidemiol. 145:
199-210.

Smith, A.J., Stems, H.S., Penner, M., Hay, K., Mitri, A., Bapat, B.V. and Gallinger, S.
(1994) Somatic APC and K-ras codon 12 mutations in aberrant crypt foci from
human colons. Cancer Res. 54: 5527-5530.

Sorensen, I.K., Kristiansen, E., Mortenson, A., Nicolaisen, G.M., Wijnads, J .A., Van
Kranen, H.J. and Van Krijl, CF. (1998) The effects of soy isoﬂavones on the
development of intestinal neoplasia in ApcMin mouse. Cancer Lett. 130: 217-225.

Sorenson, A.W., Slattery, M.L. and Ford, M.H. (1988) Calcium and colon cancer: A
review. Nutr. Cancer 11: 135-145.

Spector, D., Anthony, M., Alexander, D. and Arab, L. (2003) Soy consumption and
colorectal cancer. Nutr. Cancer 47: 1-12.

Srivastava, S., Verma, M. and Henson, DE. (2001) Biomarkers for early detection of
colon cancer. Clin. Cancer Res. 7: 1118-1126.

St. Clair, W.H., Billings, P.C., Care, J ., Keller-McGrandy, C. and Newbeme, P. (1990)
Suppression of dimethylhydrazine-induced carcinogenesis in mice by dietary addition
of the Bowman-Birk protease inhibitor. Cancer Res. 50: 580-586.

Steck-Scott, S., Forman, M.R., Sowell, A., Borkowf, C.B., Albert, P.S., Slattery, M.,

Brewer, B., Caan, B., Paskett, E., Iber, F., Kikendall, W., Marshall, J., Shike, M.,
Weissfeld, J ., Snyder, K., Schnatzkin, A. and Lanza, E. (2004) Carotenoids, vitamin

201

 

A and risk of adenomatous polyp recurrence in the polyp prevention trial. Int. J.
Cancer 112: 295-305.

Steinbach, G., Lupton, J ., Reddy, B.S., Kral, J.G. and Holt, PR. (1994) Effect of calcium
supplementation on rectal epithelial hyperproliferation in intestinal bypass subjects.
Gastroenterol. 106: 1 162-1 167.

Steinmetz, K.A. and Potter, J.D. (1991) Vegetable, fruit, and cancer. I. Epidemiology.
Cancer Causes Control 2: 325-357.

Steinmetz, K.A. and Potter, J .D. (1996) Vegetables, fruit and cancer prevention: A
review. J. Am. Diet. Assoc. 10: 1027-1039.

Stern, H.S., Gregoire, R.C., Kashtan, H., Stadler, J. and Bruce, R.W. (1990) Long-term
effects of dietary calcium on risk markers for colon cancer in patients with familial
polyposis. Surgery 108: 528-533.

Stich, HF. (1991) The beneﬁcial and hazardous effects of simple phenolic compounds.
Mutat. Res. 259: 307-324.

Strater, J ., Koretz, K., Gﬁnthert, AR. and Moller, P. (1995) In situ detection of
enterocytic apoptosis in normal colonic mucosa and in familial adenomatous
polyposis. Gut 37: 819-825.

Sugimura, T. (1992) Multistep carcinogenesis: A 1992 perspective. Science 258: 603-
607.

Sung, M.K., Kendall, C.W., Koo, M.M. and Rao, A.V. (1995a) Effect of soybean
saponins and gypsnophilla saponin on growth and viability of colon carcinoma cells
in culture. Nutr. Cancer 23: 259-270.

Sung, M.K., Kendall, CW. and Rao, A.V. (1995b) Effect of soybean saponins and
gypsophila saponin on morphology of colon carcinoma cells in culture. Nutr. Cancer
23: 259-270.

Suzuki, H., Matsumoto, K. and Terabe, M. (1992) Ki-67 antibody labeling index in
colorectal carcinoma. J. Clin. Gastroenterol. 15: 317-320.

Syngal, S., Clarke, G. and Bandipalliam, P. (2000) Potential roles of genetic biomarkers
in colorectal cancer chemoprevention. J. Cell. Biochem. Suppl. 34: 28-34.

Szwergold, B.S., Graham, RA. and Brown, TR. (1987) Observation of inositol pentakis-

and hexakis-phosphates in mammalian tissues by 31P NMR. Biochem. Biophys. Res.
Commun. 149: 874-881.

202

.7

Tajima, K. and Tominaga, S. (1985) Dietary habits and gastro-intestinal cancers: A

comparative case-control study of stomach and large intestinal cancers in Nagoya,
Japan. Jpn. J. Cancer Res. 76: 705-716.

Takahashi, M., F ukuda, K., Sugimura, T. and Wakabayashi, K. (1998) B-Catenin is
frequently mutated and demonstrates altered cellular location in azoxymethane-
induced rat colon tumors. Cancer Res. 58: 42-46.

Takayama, T., Katsuki, S., Takahashi, Y., Ohi, M., Nojiri, S., Sakamaki, S., Kato, J.,
Kogawa, K., Miyake, H. and Niitsu, Y. (1998) Aberrant crypt foci of the colon as
precursors of adenoma and cancer. N. Engl. J. Med. 339: 1277—1284.

Tanaka, S., Haruma, K., Tatsuta, S., Hiraga, Y., Teixeira, C.R., Shimamoto, F.,
Yoshihara, M., Sumii, K. and Kajiyama, G. (1995) Proliferating cell nuclear antigen
expression correlates with the metastatic potential of submucosal invasive colorectal
carcinoma. Oncology 52: 134-139.

Tanaka, T., Kojima, T., Suzui, M. and Mori, H. (1993) Chemoprevention of colon
carcinogenesis by the natural product of a simple phenolic compound protocatechuic
acid: Suppressing effects on tumor development and biomarkers expression of colon
tumorigenesis. Cancer Res. 53: 3908-3913.

Thigalingam, S., Lengauer, C., Leach, F.S., Schutte, M., Hahn, S.A., Overhauser, J .,
Willson, J.K., Markowiz, S., Hamilton, S.R., Kern, S.E., Kinzler, K.W. and
Vogelstein, B. (1996) Evaluation of candidate tumour suppressor genes on
chromosome 18 in colorectal cancers. Nature Genet. 13.

Thorburn, J. and Thorburn, A. (1994) The tyrosine kinase inhibitor, genistein, prevents
or-adrenergic-induced cardiac muscle cell hypertrophy by inhibiting activation of the
ras-map kinase signalling pathway. Biochem. Biophys. Res. Commun. 202: 1586-

1 591 .

Thorup, 1., Meyer, 0. and Kristansen, E. (1994) Inﬂuence of a dietary ﬁber on
development of dimethylhydrazine-induced aberrant crypt foci and colon tumor
incidence in Wistar rats. Nutr. Cancer 21 : 177-182.

Thun, M.J., Calle, E.E., Namboodiri, M.M., Flanders, W.D., Coates, R.J., Byers, T.,
Boffetta, P., Garﬁnkel, L. and Heath, C.W., Jr. (1992) Risk factors for fatal colon
cancer in a large prospective study. J. Nat. Cancer Inst. 84: 1491-1500.

Toribara, NW. and Sleisenger, M.H. (1995) Screening for colorectal cancer. N. Engl. J.
Med. 332: 861-867.

Toschi, L. and Bravo, R. (1988) Changes in cyclin/proliferating cell nuclear antigen
distribution during DNA repair synthesis. J. Cell Biol. 107: 1623-1638.

203

Toth, B. (1980) Tumorigenesis by benzo(a)pyrene administered intracolonically.
Oncology 37: 77-82.

Trock, B., Lanza, E. and Greenwald, P. (1990) Dietary ﬁber, vegetables, and colon
cancer: Critical review and meta-analyses of the epidemiologic evidence. J. Nat.
Cancer Inst. 82: 650-661.

Troll, W., Wiesner, R., Shellabarger, C.J., Holtzrnan, S. and Stone, J .P. (1980) Soybean
diet lowers breast tumor incidence in irradiated rats. Carcinogenesis 1: 469-472.

Tseng, M., Murray, S.C., Kuper, LL. and Sandler, RS. (1996) Micronutrients and the
risk of colorectal adenomas. Am. J. Epidemiol. 144: 1026-1027.

Tungekar, M.F., Gatter, K.C., Dunnill, MS. and Mason, D.Y. (1991) Ki-67
immunostaining and survival in operable lung cancer. HistOpathology 19: 545-550.

US. Department of Agriculture (1999). USDA-Iowa State University database on the
isoﬂavone content of foods.

Utsunomiya, J. and Lynch, HT. (1990). Hereditary colorectal cancer. Springer-Verlag,
New York, NY.

Van der Meer, R. and De Vries, HT. (1985) Differential binding of glycine- and taurine-
conjugated bile acids to insoluble calcium phosphate. Biochem. J. 229: 265-268.

Van der Meer, R., Welberg, J .W., Kuipers, F ., Kleibeuker, J .H., Mulder, N.H., Termont,
D.S., Vonk, R.J., De Vries, HT. and De Vries, E.G. (1990) Effects of supplemental

dietary calcium on the intestinal association of calcium, phosphate, and bile acids.
Gastroenterol. 99: 1653-1659.

Van Stalk, R.V., Theodors, A. and Beck, G.J. (1994). Distal colonic hyperplastic polyps
are markers for proximal colonic adenomas: A meta-analysis. Tenth World Congress
of Gastroenterology, Los Angeles, CA.

Van Tasslee, R.L., Kingston, D.G.I. and Wilkins, TD. (1990) Metabolism of dietary
genotoxins by the human colonic microﬂora; the fecapentaenes and hetrocyclic
amines. Mutat. Res. 238: 209-221.

van Wyk, R., Slezak, P., Kotze, M.J., Jaramillo, E., Koizumi, K. and Grobbelaar, J.J.
(1999) Multiple APC mutations in sporadic ﬂat colorectal adenomas. Eur. J. Hum.
Genet. 7: 928-932.

Veale, A.M. (1965). Intestinal polyposis. Cambridge University Press, Cambridge, UK.

204

Vogelstein, B., F earon, E.R., Hamilton, S.R., Kern, S.E., Preisinger, A.C., Leppert, M.,
Nakamura, Y., White, R., Smits, A.M. and Bos, J .L. (1988) Genetic alterations during
colorectal-tumor development. N. Engl. J. Med. 319: 525-532.

Vogelstein, B. and Kinzler, K.W. (1993) The multistep nature of cancer. Trends Genet. 9:
138-141.

Waddell, W.R., Ganser, G.F., Cerise, E.J. and Loughry, R.W. (1989) Sulindac for
polyposis of the colon. Am. J. Surg. 157: 175-179.

Wakai, K., Egami, 1., Kato, K., Kawamura, T., Tamakoshi, A., L., Y., Nakayama, T.,
Wada, M. and Ohno, Y. (1999) Dietary intake and sources of isoﬂavones among
Japanese. Nutr. Cancer 33: 139-145.

Wang, H.J. and Murphy, P.A. (1994a) Isoﬂavone content in commercial soybean foods.
J. Agric. Food Chem. 42: 1666-1673.

Wang, H.J. and Murphy, P.A. (1994b) Isoﬂavone composition of American and Japanese
soybeans in Iowa: Effects of variety, crop year and location. J. Agric. Food Chem. 42:
1674-1677.

Wang, H.J. and Murphy, PA. (1996) Mass balance study of isoﬂavones during soybean
processing. J. Agric. Food Chem. 44: 2377-2383.

Ward, J .M., Yamamoto, RS. and Brown, CA. (1973) Pathology of intestinal neoplasm

and other lesions in rats exposed to azoxymethane. J. Nat. Cancer Inst. 51: 1029-
1039.

Wargovich, M., Jimenez, A., McKee, K., Steele, V. and Velasco, M. (2000) Efﬁcacy of
potential chemopreventive agents on rat colon aberrant crypt formation and
progression. Carcinogenesis 21: 1 149-1155.

Wargovich, M.J., Eng, V.W., Newmark, H.L. and Bruce, W.R. (1983) Calcium
ameliorates the toxic effect of deoxycholic acid on colonic epithelium.
Carcinogenesis 4: 1205-1207.

Wargovich, M.J., Eng, V.W. and Newmark, H.L. (1984) Calcium inhibits the damaging
and compensatory proliferative effects of fatty acids on mouse colon epithelium.
Cancer Lett. 23: 253-258.

Wargovich, M.J. and Lointier, PH. (1987) Calcium and vitamin D modulate mouse colon
epithelial proliferation and growth characteristics of a human colon tumor cell line.

Can. J. Physiol. Pharmacol. 65: 472-477.

Wargovich, M.J., Isbell, G., Shabot, M., Winn, R., Lanza, F., Hochman, L., Larson, E.,
Lynch, P., Rooubein, L. and Levin, B. (1992) Calcium supplementation decreases

205

rectal epithelial cell proliferation in subjects with sporadic adenoma. Gastroenterol.
103: 92-97.

Wargovich, M.J., Chen, C.D., Harris, C., Yang, E. and Velasco, M. (1995a) Inhibition of
aberrant crypt growth by non-steroidal anti-inﬂammatory agents and differentiation
agents in the rat colon. Int. J. Cancer 60: 515-519.

Wargovich, M.J., Jimenez, A., McKee, K., Steele, V.E., Velasco, M., Woods, J., Price,
R., Gray, K. and Kelloff, G. (1995b) Efﬁcacy of potential chemopreventive agents on
rat colon aberrant crypt formation and progression. Gastroenterol. 108: A551.

Watanabe, Y., Tada, M., Kawamoto, K., Uozumi, G., Kajiwara, Y., Hayashi, K.,
Yamaguchi, K., Murakami, K., Misaki, F. and Akasaka, Y. (1984) A case-control
study of the rectum and the colon. Nippon Shokakibyo Gakkai Zasshi 81: 185-194.

Wattenberg, L.W. (1985) Chemoprevention of cancer. Cancer Res. 45: 1-8.

Wattenberg, L.W. (1993) Prevention-therapy-basic science and the resolution of the
cancer problem: Presidential address. Cancer Res. 53: 5890-5896.

Weaver, G.A., Tangel, C.T., Krause, J.A., Parﬁtt, M.M., Stragand, J.J., Jenkins, P.L.,
Erb, T.A., Davidson, R.H., Alpem, H.D., Guiney, W.B., Jr. and Higgins, P.J. (2000)
Biomarkers of human colonic cell growth are inﬂuenced differently by a history of
colonic neoplasia and the consumption of acarbose. J. Nutr. 130: 2718-2725.

Weed, H., McGrandy, RB. and Kennedy, AR. (1985) Protection against
dimethylhydrazine-induced adenomatous tumors of the mouse colon by the dietary

addition of an extra of soybeans containing the Bowman-Birk protease inhibitor.
Carcinogenesis 6: 1239-1241.

Wei, H., Bowen, R., Cai, Q., Barnes, S. and Wang, Y. (1995) Antioxidant and anti-
promotional effects of the soybean isoﬂavone genistein. Proc. Soc. Exp. Biol. Med.
208: 124-130.

Weihrauch, J .L. and Gardner, J .M. (1978) Sterol content of foods of plant origin. J. Am.
Diet Assoc. 73: 39-47.

Weisburger, J .H. (1973) Chemical carcinogens and their mode of action in colonic
neoplasia. Dis. Colon Rectum 16: 431-437.

Weisburger, J .H., Jones, R.C., Wang, C.X., Backlund, J .Y., Williams, G.M., Kingston,
D.G., Van Tassell, R.L., Keyes, R.F., Wilkins, TD. and de Wit, RP. (1990)
Carcinogenicity tests of fecapentaene-12 in mice and rats. Cancer Lett. 49: 49-59.

Weisburger, J .H., Rivenson, A. and Reinhardt, J. (1994) Genotoxicity and
carcinogenicity in rats and mice of 2-amino-3,6-dihydro-3-methyl-7h-imidazolo[4,5-

206

 

 

ﬂquinolin-7-one: An intestinal bacterial metabolic of 2-amino-3-methyl-3H-
imidazo[4,5-f]quinoline. J. Nat. Cancer Inst. 86: 25-30.

Weisgerber, U.M., Boeing, H., Owen, R.W., Waldherr, R., Raedsch, R. and Wahrendorf,
J. (1996) Effect of longterm placebo controlled calcium supplementation on

sigmoidal cell proliferation in patients with sporadic adenomatous polyps. Gut 38:
396-402.

Wenzel, U., Kuntz, S., Brendel, MD. and Daniel, H. (2000) Dietary ﬂavone is a potent
apoptosis inducer in human colon carcinoma cells. Cancer Res. 60: 3823-3831.

Whittemore, A.S., Wu-Williams, A.H., Lee, M., Zheng, S., Gallagher, R.P., Jiao, D.A.,
Zhou, L., Wang, X.H., Chen, K. and Jung, D. (1990) Diet, physical activity, and
colorectal cancer among Chinese in North America and China. J. Nat. Cancer Inst.
82: 915-926.

Wilcox, B., Fuchigami, K., Wilcox, D., Kendall, C., Suzuki, M. and Todoriki, H. (1995)
Isoﬂavone intake in Japanese and Japanese-Canadians. Am. J. Clin. Nutr. 61: 901
(Abstract).

Willett, W.C., Stampfer, M.J., Colditz, G., Rosner, BA. and Speizer, RE. (1990)
Relation of meat, fat and ﬁber intake to the risk of colon cancer in a prospective study
among women. N. Engl. J. Med. 323: 1664-1672.

Wilson, R.G., Smith, AN. and Bird, CC. (1990) Immunohistochemical detection of
abnormal cell proliferation in colonic mucosa of subjects with polyps. J. Clin. Pathol.
43: 744-747.

Wintzer, H.O., Zipfel, 1., Schulte-Monting, J ., Hellerich, U. and von Kleist, S. (1991) Ki-
67 immunostaining in human breast tumors and its relationship to prognosis. Cancer
67: 421-428.

Wu, A.H., Ziegler, R.G., Horn-Ross, P.L., Nomura, A.M., West, D.W., Kolonel, L.N.,
Rosenthal, J .F ., Hoover, RN. and Pike, MC. (1996) Tofu and risk of breast cancer in
Asian-Americans. Cancer Epidemiol. Biomarkers Prev. 5: 901-906.

Wu, A.H., Ziegler, R.G., Nomura, A.M., West, D.W., Kolonel, L.N., Hom-Ross, P.L.,
Hoover, RN. and Pike, MC. (1998) Soy intake and risk of breast cancer in Asians
and Asian Americans. Am. J. Clin. Nutr. 68(suppl): 14375-14435.

Wu, J., Mao, Y., He, L., Wang, N., Wu, C., He, Q. and Skog, S. (2000) A new cell
proliferating marker: Cytosolic thymidine kinase as compared to proliferating cell
nuclear antigen in patients with colorectal carcinoma. Anticancer Res. 20: 4815-4820.

Wu, K., Willett, W., Fuchs, C., Colditz, G. and Giovannucci, E. (2002) Calcium intake
and risk of colon cancer in women and men. J. Nat. Cancer Inst. 94: 437-446.

207

Wynder, B.L., Kajitani, T., Ishikawa, S., Dodo, H. and Takano, A. (1969) Environmental
factors of cancer of the colon and rectum. Cancer 23: 1210-1220.

Xiao, Z.-Q., Jaszewski, R. and Majumdar, A.P.N. (2000) Aging enhances G1 phase in the
colonic mucosa of rats. Mech. Ageing Dev. 116: 1-14.

Xu, X., Wang, H.-J., Murphy, P.A., Cook, LR. and Hendrich, S. (1994) Daidzein is a
more bioavailable soymilk isoﬂavone than is genistein in adult women. J. Nutr. 124:
825-832.

Xu, X., Harris, K.S., Wang, H.-j., Murphy, RA. and Hendrich, S. (1995) Bioavailability
of soybean isoﬂavones depends upon gut microﬂora in women. J. Nutr. 125.

Xu, X., Duncan, A.M., Wanagen, KB and Kurzer, MS. (2000) Soy consumption alters
endogenous estrogen metabolism in postmenopausal women. Cancer Epidemiol.
Biomarkers Prev. 9: 781-786.

Yamaguchi, A., Makimoto, K., Goi, T., Takeuchi, K., Maehara, M., Isobe, Y. and
Nakagawara, G. (1994a) Overexpression of p53 protein and proliferative activity in
colorectal adenoma. Oncology 51: 224-227.

Yamaguchi, M., Hayashi, Y., Matsuoka, S., Takahashi, T. and Matsukage, A. (1994b)
Differential effect of p53 on the promoters of mouse DNA polymerase 13 gene and
proliferating-cell-nuclear-antigen gene. Eur. J. Biochem. 221: 227-237.

Yamashita, N., Minamoto, T., Onda, M. and Esumi, H. (1994) Increased cell proliferation

of azoxymethane-induced aberrant crypt foci of rat colon. Jpn. J. Cancer Res. 85:
692-698.

Yamashita, N., Minamoto, T., Ochiai, A., Onda, M. and Esumi, H. (1995) Frequent and
characteristic K-ras activation and absence of p53 protein accumulation in aberrant
crypt foci of the colon. Gastroenterol. 108: 434—440.

Yanagihara, K., Ito, A., Toge, T. and Numoto, M. (1993) Antiproliferative effects of
isoﬂavones on human cancer cell lines established from the gastrointestinal tract.
Cancer Res. 53: 5815—5821.

Yang, K., Edelmann, W., Fan, K.H., Lau, K., Leung, D., Newmark, H., Kucherlapati, R.
and Lipkin, M. (1998) Dietary modulation of carcinoma development in a mouse
model for human familial adenomatous polyposis. Cancer Res. 58: 5713-5717.

Yavelow, J ., Collins, M., Birk, Y., Troll, W. and Kennedy, AR. (1985) Nanomolar

concentrations of Bowman-Birk soybean protease inhibitor suppress x-ray-induced
transformation in vitro. Proc. Natl. Acad. Sci. USA 82: 5395-5399.

208

Yerly-Motta, V., Pavy, J .J ., Bosset, S.F., Arbez-Gindre, F. and Herve, P. (1999) Validity
of PCNA immunostaining in normal and tumor cells: Comparison with Ki-67
labeling. J. Histotechnol. 22: 287-293.

Yoshikawa, R. and Utsunomiya, J. (1996) Cell proliferation kinetics are abnormal in
transitional mucosa adjacent to colorectal carcinoma. Br. J. Surg. 83: 36-39.

Yoshiki, Y., Kudou, Y.S. and Okubo, K. (1998) Relationship between chemical
structures and biological activities of triterpenoid saponins from soybean. Biosci.
Biotechnol. Biochem. 62: 2291-2299.

Yoshimi, N., Wang, A., Morishita, Y., Tanaka, T., Sugie, S., Kawai, K., Yamahara, J.
and Mori, H. (1992) Modifying effects of fungal and herb metabolites on

azoxymethane-induced intestinal carcinogenesis in rats. Jpn. J. Cancer Res. 183:
1273-1278.

Yuan, J .M., Wang, Q.S., Ross, R.K., Henderson, BE. and Yu, MC. (1995) Diet and
breast cancer in Shanghai and Tiajin, China. Br. J. Cancer 71: 1353-1358.

Yukawa, M., Fujimori, T., Maeda, S., Tabuchi, M. and Nagasako, K. (1994) Comparative
clinicopathological and immunohistological study of ras and p53 in ﬂat and polypoid
type colorectal tumours. Gut 35: 1258-1261.

Zacchetti, A., van Garderen, E., Teske, E., Nederbragt, H., Dierendonck, J .H. and
Rutteman, GR. (2003) Validation of the use of proliferation markers in canine
neoplastic and non-neoplastic tissues: Comparison of Ki-67 and proliferating cell

nuclear antigen (PCNA) expression versus in vivo bromodeoxyuridine labelling by
immunohistochemistry. APMIS 111: 430-438.

Zaridze, D., Filipchenko, V., Kustov, V., Serdyuk, V. and Duffy, S. (1993) Diet and
colorectal cancer: Results from two case-control studies in Russia. Eur. J. Cancer
29A: 112-115.

Zarkovic, M., Quin, X., Nakatsuru, Y., Shamsuddin, A.M. and Ishikawa, T. (1993)

Tumor promotion by fecapentaene-12 in a rat colon carcinogenesis model.
Carcinogenesis 14.

209

 

   

1"11111111111111131111111