THESIS

Ml! lfllllluzllfllflillll W" Win21“

EVIDENCE FOR CELL SURFACE ASSOCIATED NUCLEIC ACID:
A STUDY USING WHOLE CELL ELECTROPHORESIS
ON THE 8-180 MOUSE TUMOR

BY

David Anthony Juckett

A DISSERTATION

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

DOCTOR OF PHILOSOPHY

Department of Biophysics

1979

ABSTRACT
EVIDENCE FOR CELL SURFACE ASSOCIATED NUCLEIC ACID:

A STUDY USING WHOLE CELL ELECTROPHORESIS
ON THE 8-180 MOUSE TUMOR

BY

David Anthony Juckett

A whole cell electrophoresis apparatus was built which
was capable of measuring the average electrophoretic mobility
of a suspension of cells in 3-5 minutes. This was used to measure
the mobility of 8-180 mouse ascitic tumor cells under various
conditions. The mobility was determined as a function of the age
of the tumor and treatment with cisvdichlorodiammeplatinum(II)
(cisplatin). The cisplatin treatment of mice bearing the 8-180
caused a reduction in mobility of the tumor cells by 10-20:. The
cells, after removal from the animals, were incubated with several
enzymes to see if any of them would mimic the effect of cisplatin
treatment. In the 5-180 tumor, RNase mimicked the effect of cisplatin
by lowering the mobility of controls to the level of the cisplatin
treated cells. The RNase had no effect on the platinum treated cells.
In'the S-lBO-J tumor, both RNase and DNase mimicked the effect of
cisplatin treatment. Several experiments were then performed to
establish the presence of nucleic acids on the cell surface, which
was implied by the first enzyme studies. pH profiles, ionic strength
profiles, binding of anti-DNA antibody, binding of 'platinum-thymine-

blues', incubation with Sepharose immobilized DNase and Agarose

David Anthony Juckett

immobilized RNase, incubation with restriction endonucleases

Eco-R1 and Ban 1, enzyme kinetics and in vitro incubations with
cisplatin and other antitumor drugs all gave support for cell
surface nucleic acids. These nucleic acids appeared to be in

the process of flux through the membrane with a turnover half-

time of about 2 hours. Inhibition of cell nucleic acid metabolism
resulted in the loss of their expression on the surface of the cell.
The possible roles of these surface nucleic acids are discussed,

particularly with regard to their possible immunological roles.

ACKNOWLEDGMENTS

I would like to thank Dr. Barnett Rosenberg for his support
and guidance throughout my work. I would like to thank Mrs.
Loretta Van Camp for her continual assistance and her good
ideas. I am grateful to Dr. Estelle McGroarty, Dr. Ronald Patterson,
and Dr. Clifford Welsch for serving on my committee. Finally
I would like to thank the many mice that gave their lives for

this work.

ii

TABLE OF CONTENTS

 

 

 

 

 

LIST 0]? TABLES ------------------------------ — - vi
LIST OF FIGURES ------------------- __ Vii
Introduction ------—---—- __ -----...-.. 1
Literature Review ............................................. 6
I. Platinum Research - —___=_ ........... 5

A. Biological Effects
B. Macromolecular Interactions

II. Whole Cell Electrophoresis ......................... 11
A. Theory
B. Effects of pH and Ionic Strength
C. Effects of Enzymes and Other Surface Effectors
D. Changes in Tumor Cells

 

III. Tumor Membranes ................ _ ...... 19
A. Changes in Cell Behavior
B. Membrane Components
l.Glycoproteins
2.Clycolipids
3.?roteins
4.Lectins
5.Antigens

IV. Surface Nucleic Acids ............................... 26
A. Surface RNA
B. Surface DNA
C. Cell Released Nucleic Acids

 

 

 

 

 

 

 

Development of Electrophoresis Apparatus _ _ 35
Materials and Methods ------------ _ --- 43
I. Buffers and Solutions -— _ _ 43
II. Cells - —— — 43
III.anymes-- - 45
IV. Tissue Culture Media for Incubations --------------- 45
V. Enzyme Treatment ----------- 46
VI. Cell Pathology-u- ----------------------- 46
VII. Analysis of Raw Electrophoretic Mobility Data ----- 46
VIII. Computer-- ------------------------ 48

 

iii

9383

 

 

 

 

Outline --- -- -------------- _ __ _ 49
Results ------------------- — _ 52
I. Cell Pathology ----------—----— — —— 53
II. Electrophoresis -------------- _ 57

 

A. Initial Experiments

B. Furthe

C.

D.

E.

l.

S.
l.
2.
3.
4.

5.
6.

7.

Mobility as a function of the tumor
age
The distribution of mobilities
Enzyme incubations
Mobilities as a function of tumor age
and enzyme treatment - S-lBO-J
Enzyme kinetics

Experiments to Test for Nucleic Acids 65
pH profiles of the 8-180 and S-lBO—J
Ionic strength profiles

Incubation with Sepharose bead
immobilized DNase and RNase
Ylatinum-thymine-blue' attachment to
tne cell membrane

Incubation with anti-DNA antisera
Incubation with the restriction
enzymes Eco-R1 and Bam-I

A spectrophotometric study

Binding and Origin 81

1.
2.

3.

Cellular Debris

Lability of nucleic acids with trypsin
neuraminidase, and low ionic strength
In vitro studies

Other Experiments 92

l.
2.

3.

Binding of anti-tumor antisera to
'interim 8-180'

Animals injected with nuclease treated
cells

Tumor supernatant and antibody production

Other Cell Types 99

1.
2.
3.
4.
5.
6.

8-180 and S-lSO-J solid tumors
P-388 and P-815 ascitic tumors
Liver and fetal cells

AKR spontaneous leukemia

Tissue culture cells V79 and LlZlO
Summary of cell types

Summary of Conclusions 104

iv

9‘89

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Discussion ------ 106
1. Proof of Nucleic Acids —— 106
II. Overlapping Effects of RNase and DNase -- _ 107
III. Lability of Surface Nucleic Acids and Their ------- 103
Turnover
IV. Size of the Nucleic Acid Group -- __ 109
V. Solid Tumors ------- — 112
VI. The P—815 Mastocytoma —== 113
VII.Results of Other Cells — _=——— ——— - 113
VIII.Roles of Cell Surface Nucleic Acids —-— —— 115
IX. Future Experiments - 120
X. The Effect of cisplatin —— —— — 122
Appendix I - 123
Appendix II ------------- — —= 128
Appendix III ------------------------- 129
Appendix IV =— —= — 131
Bibliography ----------- —- 133

 

LIST OF TABLES

Table page

I The effects of enzyme incubation on the mobility 61
of the 5-180 tumor.

II Incubation of cells with immobilized RNase and 72
DNase.

III The binding of the 'platinum-thymine-blues' to 74
the S-lBO-J.

IV The incubation of cells with anti-DNA antisera. 74

V The effect of restriction endonuclease Eco-R1 76

and nuclease $1 on cell mobility.

VI Incubation of cells with restriction endonuclease 77
Ram I.

VII The effect of cell debris incubation on the 5-180-J 81
mobility.‘

VIII Multiple enzyme incubations. 82

IX The effect of ionic strength and temperature on 85

the mobility of the S-lBO-J.

X Incubation of the S-lBO-J with sera from tumor 93
bearing animals.

XI Day of death of mice injected with S-lBO-J 95
incubated in RNase and DNAse.

XII The effect of RNAse and DNAse on the mobility of 100
'interim 8-180' and S-lBO-J in the solid form

XIII Nuclease effect on the mobility of various cell 100
types.
XIV Summary of nuclease effects on several cell types. 103

vi

10.

11.

LIST OF FI GURES

Diagram of mirror movement.

Diagram of measuring chamber.

Diagram of chamber and electrode circuit.
Circuit diagram of electrophoresis apparatus.

Percent of major cell types versus age of tumor
in normal tumor, a, and platinum treated tumor,b.

The electrophoretic mobility of 5-180 tumor cells
with and without platinum treatment as a function
of tumor age. Bars represent the standard error.

Distribution of mobility for the S-180-J. a)
cisplatin treated tumor, hatched line is host
cells, solid is tumor. b) untreated tumor, tumor
cells only. The ordinate is number of cells.

Mobility as a function of tumor age for control and

platinum treated cells. Squares represent RNase
incubation, triangles DNase incubation, and
circles are incubation controls.

Kinetics of enzyme action upon mobility of the
S-lBO-J. Top curve in each figure is the
incubation control.

Plot of mobility versus pH of the measuring buffer
for control (C) and platinum treated (Pt) 8-180.

Plot of mobiltiy versus pH for the S-lBO-J.

Solid data points are control cells, hollow points
are Pt treated. Squares are RNase incubated,
triangles are DNase incubated, and circles are
incubation controls.

vii

page

37

38

38

42

55

58

59

64

66

68

Figure

12.

13.

I4.

15.

16.

17.

18.

page

Log of electrophoretic mobility versus the square 70
root of the ionic strength of the measuring media.

a) Control cells, b) platinum treated tumor.

R,P,N, represent RNase incubation, protease

incubation, and neuraminidase incubation respectively

As in Figure 12 except this is S-lBO-J, R,D,N 71
represent incubations with RNase, DNase and
neuraminidase respectively.

UV spectra of supernatant from cells (S-lBO-J) 79
incubated under various conditions. see text
for details.

Plot of mobility versus time in in vitro 88
incubations. Cells were removed from animals,

preincubated with cisplatin (Pt), or with

Actinomycin-D (Act-D), for forty minutes before

time zero. Time shown is time in normal tissue

culture media. R represents RNase incubation of

samples taken during the course of the in vitro
incubations.

In vitro experiments as described in Figure 15. 90
The control curve is repeated in each figure as
reference. a) Normal S-lSO-J were preincubated
with RNase and then allowed to grow in vitro.

b) similar to (a) except cells were preincubated
with DNase. c) Squares; cells were preincubated
with mercaptopurine, 0.5mg/ml. Crosses; cells were
incubated with methotrexate, 0.5mg/ml.

d) the RNase curve of Figure 15 is repeated.
Triangles; cells were preincubated with lmg/ml
puromycin.

Mobility of sheep red blood cells as a function of 97
sera taken from mice injected with (C) sheep

RBC's only, (8) sheep RBC's + S-lBO-J supernatant,

and (R) sheep RBC's + tumor supernatant

incubated with RNase.

Model of the relative orientation of the slip 111
plane to a hypothetical nucleic acid complex.

viii

Figure

19.

20.

21.

Plot of mobility versus the reciprocol of the
square root of the ionic strength of the
measuring buffer. Solid circles are the 5-180
controls, open circles are the 3-180 RNase
incubated, and trianglis are red blood cell data
from Heard and Seaman .

Model of cell surface with thickness and the
potential curves from charges in the membrane
in a) normal ionic strength and b) low ionic

strength.

Plot of mobilities of sheep RBC's as a function
of sera dilutions. Closed circles are normal
rabbit sera and open circles are anti-sheep
RBC antisera.

ix

page

125

126

132

Introduction

In the last ten years the development of the successful
treatment of many cancers with cis-dichlorodiammineplatinum—II
(Cisplatin) has led to many investigations of its mode of action
in tumor regression. Most of this research falls into one of
three catagories; studies of platinum chemistry, results of clinical
trials, and interactions of platinum compounds with DNA.

Examination of this new drug's interaction with cellular biochemistry
is noticably lacking and cisplatin's effect on the cell membrane
is virtually unknown.

The first purpose of this study is to examine a particular
characteristic associated with the cell membrane as a function
of platinum treatment. That characteristic is the average membrane
surface charge which is derived from the total cellular electro-
phoretic mobility. This is measured on viable cells in physiological
buffer solutions.

The average surface charge is the sum of all surface groups
which are charged at physiological pH. This average charge represents
very little by itself, but it can be used to determine the nature
of the individual charged groups which are present on the outer
surface of the cell. These charged groups are mainly negative,
consisting of carboxylic groups and phosphates. Proteins in the

membrane contain negatively charged carboxylic acid groups,

1

non-charged sulfhydral groups, and positively charged amines.3O
Sugar moieties probably contribute the majority of the negative
charged carboxylic acid residues, particularly in the form of sialic

acid.24

Phosphates are present in the lipids of the membrane and also
in some surface macromolecules, and sulfates are also present in
small amounts. When the removal of one or another of these groups can
be identified by changes in cell electrophoresis then a
physiologically significant piece of information about the cell emerges.

To determine the types of groups present, the technique of
whole cell electrophoresis is used in conjunction with controlled
surface effectors. These surface effectors are used to selectively
remove or alter known groups on the surface in order to establish
their presence. The most commonly used effectors are digestive
enzymes specific for saccharide, peptide, or nucleic acid bonds.
Other effectors that can be used include chargximacromolecules
which electrostatically bind to the surface groups, antibodies
specific to surface antigens, and reagents which specifically alter
exposed molecules. Once groups are determined, their relative
concentrations on the surface can be followed as a function of
cellular treatments, such as chemotherapy, immunotherapy, radiation
treatment, etc. This technique is used in this study to examine the
5-180 mouse tumor system as a function of platinum chemotherapy.

When a tumor bearing animal is treated with a therapeutic
dose of cisplatin the electrophoretic mobility of the tumor cells
decrease by approximately 20%. This change occurs over the first

24 hours and remains depressed unless the tumor begins to grow again.

This phenomenon occurs simultaneously with an invasion of the
tumor by neutrophils of the host, followed in later days by large
numbers of macrophages. This host response may be due to
inflammation caused by platinum therapy, by dying cells, or by

an increased immune response to the tumor.

It was this last possibility, perhaps the least likely of the
three, that spurred interest in examining the phenomenon of the
reduced cell surface charge. This reduced charge could lower the
electrostatic repulsion between cells and increase cell-cell
interaction resulting in greater phagocytic activity by host cells.
Since most tumors show increased surface charge37 they could be
protected from the cell-cell interaction necessary for phagocytosis.

Changes in surface charge could also represent a loss or addition
of one or more specific groups, such as receptors, transport
enzymes, catalytic enzymes, glycoprotein or mucopolysaccharide
antigens etc. These types of changes would effect the cell's inter-
action with it’s environment and, therefore, could alter the immune
response.

Due to length of time it takes for the mobility to decrease,
approximately 1 day, the possibility of cisplatin binding to
the membrane and thereby altering the charge was virtually ruled
out. Cisplatin is also a neutral molecule at physiological chloride
ion concentration and its surface binding, therefore, would go
unmeasured. The most probable route for the drug would be to enter
the dell where the changes it induced resulted later in a surface

alteration. The purpose of this thesis was to find what these

alterations of the membrane were and to determine, if possible,
whether these alterations were functional or simply incidental.

In addition to the interest in cisplatin's effect we were
intrigued by another phenomenon which was observed on the surface of
tumor cells that had been stained with the 'platinum-pyrimidine-
blues'88. Electron microscopy studies showed patches on the surfaces
of tumor cells which were attributed to nucleic acids. These patches
were susceptible to nucleases and to neuraminidase, but were not
present at all on non-tumorigenic cells. It seemed likely that we
could detect surface nucleic acids with whole cell electrophoresis
due to their net negative charge at pH 7.4. If they could be detected,
then it would corroborate the staining studies which implicated
nucleic acids on the very periphery of the cell surface - a highly
unlikely and unusual location. Therefore, the second purpose of
this work is to examine cells electrophoretically for cell surface
nucleic acids.

The experimental techniques of this dissertation project are
mostly limited to those of whole cell electrophoresis. It appeared
to be a very effective way to examine very small amounts of a surface
moiety on viable cells especially after cell surface nucleic acids
were implicated. This thesis presents some evidence for a new
phenomenon, speculates on it's role, and suggests some further
experiments. I first present some backround information on whole
cell electrophoresis; the relevant areas of cancer research, particular-
ly tumor cell surfaces and tumor antigens; platinum chemotherapy; and

the work to date on extracellular RNA and DNA. This is followed by

my experimental results and then a discussion section in which I
examine the possible roles of cell surface nucleic acids in light

of some other reports in the literature.

Literature Review

I. Platinum Research

gig:dichlorodiammineplatinum-II, cisplatin, has proven to be
one of a group of very potent platinum base anti-tumor drugs which
is currently being used in human chemotherapy protocols. It's mode
of action is still not known although there is a large amount of
evidence for the primary lesion being an attack on DNA. Initial
studies of it's binding to protein and RNA show little or no effect
on their synthesis, however the number of such studies is small.
The majority of the published work on platinum compounds in cancer
research are concentrated in three areas; platinum chemistry,
platinum interactions with DNA,and clinical results. This thesis
does not address the chemical or clinical aspects but rather,

the biological areas and macromolecular interactions.

A. Biological Effects
The first effect noted on biological systems, was the

filamentation of Escherichia coli which led to the discovery of

1

 

cisplatin's potential use for tumor chemotherapy. The E. coli
continued to metabolize and to grow, but they did not divide. This
immediately led to the investigation of platinum interactions with

DNA, since similar effects can be induced with ultraviolet irradiation,
X-ray irradiation, or alkylating agents - all DNA effectors. Platinum

interactions with ribonucleoproteins in filamentous E. coli did not

6

\

I

seem to interfere with protein synthesis, 5 so, DNA appeared
to be the target molecule.

Lysogenic strains of E.coli could be activated into phage
production by incubation with very low levels of platinum compounds.3
The level of anti-tumor activity correlated well with the ability to
induce phage in the lysogenic E. coli.

Cisplatin is also a mutagen. Beck and Brubaker4 showed that
reversion of mutants back to the prototrophic condition occurred
at a rate 105 times greater than spontaneous reversions. It was later
shown that the induced mutation is a base substitution rather than
a frame shift.5 In general, the platinum anti-tumor drugs exhibit
a correlation between mutagenicity and cytotoxicity.5'6

The carcinogenic potential of platinum drugs has been recently
tested7 and shown to be positive for the production of benign tumors
in mice. Lung adenomas were produced in 100% of animals treated with
cisplatin and, niconjunction with croton oil, skin papillomas were
induced, some of which turned malignant. These results are indicative
of moderate carcinogenic activity.

These biological effects all suggest that DNA is the molecule
attacked by the platinum drugs. Other effects on cells and animals
are not so straightforward but suggest other concommittant roles
for platinum action.

When used in animals as an anti-cancer drug, cisplatin causes
tumor regression which is probably assisted by an enhanced immune

response. Conran8 demonstrated that immunoajmulants can increase the

efficacy of therapy, and that immunodepressants decrease efficacy.

Sodhi9 showed that within a few hours after platinum treatment of

a tumor the host macrophages attach in greater numbers to the tumor

cells. This occurs despite cisplatin's apparent immunosuppressive

properties as shown by its ability to reduce hemolytic plaque

forming spleen cells and lymphocyte blastogensis in mice.10
Many of the cells after treatment turn into giantcells with

several nuclei,ll'12

very reminiscent of the filamentous growth

in bacteria. In Erhlich ascites cells exhibiting this giantcell
growth, the synthesis of DNA is depressed, yet there are several
copies per cell. Both RNA and protein syntheSis is above normal,13
which may be due to these extra copies of DNA. In general, these
cells are blocked in the 62 phase of the cell cycle and are unable
to continue on to mitosis.

Respiratory inhibition of chicken heart mitochondria occurs
with cisplatin but only at doses far in excess of therapeutic levels.
This probably means that respiration is not inhibited in normally
treated tumors. Zakharova,l4 however, showed that adenosine tri-
phosphatase can be inhibited at levels of cisplatin closer to

treatment values used. This could play a role in cellular energy

levels and could enhance tumor regression.

B. Macromolecular Interactions
DNA has been studied the most thoroughly of all macromolecules

2,

in its binding with platinum. Studies indicate that DNA synthesis
is inhibited, but RNA and protein synthesis is not. This was taken

to mean that platinum's main effect was on the DNA and, further, that

this effect was due to platinum binding to DNA.

Pascoe and Roberts 16'17 showed that the amount of cisplatin
bound to DNA, mRNA, rRNA, and tRNA was about equal by weight.
Approximately 1 platinum molecule is bound per 105 bases. In the
case of protein, there was 1 platinum.per 106 amino acids. They make
the point that since DNA is so large it contains many more platinum
residues per molecule than the smaller RNA molecules and therefore
poses a greater threat to the information coded on the DNA. Whereas
RNA can be transcribed from short pieces of DNA, DNA synthesis would
require an intact path for DNA polymerase. If this path was blocked
by platinum, synthesis would be stopped or slowed down while repair
occurred, if repair could occur. This, unfortunately, does not address
the problem of giant, multinucleated cells which appear to be
undergoing DNA synthesis but are unable to divide.

The binding of platinum to DNA is fastest at the guanosine
residue,18 with the N-7 site the most prevalent attachment point.

The N-l, 0-6 of guanine and the 6-NH2 of adenine are some of the
other possible binding sites. The active gi§_configuration platinum
drugs have the ability to crosslink in both an inter and intra strand
fashion between sites on neighboring nucleotides,19 thereby creating
a physical block.20

DNA repair appears to exist for some platinum induced lesions
because introduction of caffeine, an excision repair inhibitor,
reduces survival in chinese hamster cells after cisplatin treatment.21

Although postreplicative repair may be involved in the removal of

lesions, it appears that most repair is the caffeine sensitive

10

excision repair which repairs the DNA before replication

occurs.19 Differences in cells' ability to repair their DNA may
explain why 'normal' tissues are not as effected by platinum as
tumor tissues. The tumor cells may lack a repair process making them
more\nflnerable to the effect of platinum binding.

The interaction of platinum with RNA and protein is poorly
understood. Early work showed that platinum does bind to both, but
that it did not seem to effect the new synthesis of either. werk
exploring the possibility that platinum binding reduces the function—
ality of RNA or protein has not been done. Hermann et a1.22 did
show significant physical aterations in double stranded synthetic
RNA, and Robins et al.23 indicate binding to serum proteins and
some cellular proteins. Borer and Nicolau 102 indicatxla significant
binding of cisplatin to protiens of the red blood cell membrane
using shifts in fluorescence, and they suspect the platinum is bound
to amino acids other than the fluorochromic ones. This suggests that
platinum could have a significant membrane effect before it enters
the cell. Electron Spin Resonance spectroscopy studies of Sinha et

al.103 also showed tight binding of cisplatin to proteins in the

membranes of red blood cells and the P-815 mastocytoma. Wolf et al.104
fimfiher showed cisplatin exists in three fractions in the blood;
bound to cells, serum proteins, and free. They showed that the binding
to the red cell and serum proteins is very strong. The significance of
platinum binding to protein and ribonucleic acid is not known,
however, possible roles include interference' with enzyme activity or

functionality of newly synthesized protein, sequestering platinum in

a protective manner, and specialized transport of platinum.

II. Whole Cell Electrophoresis

A. Theory

When cells are suspended in an isotonic solution in which an
electric field has been established, they migrate toward the positive
source of the field. This was not supprising to early investigators
because it was known that the cell membrane contained phospholipids,
glycolipids, glycoproteins and proteins which can have both
negative and positive charges at physiologic pH.

The surfaces of cells are probably a complicated mixture of
these different groups. These cell surfaces undergo turnover of their
components, expand and contract with movement, growth, and cell-cell
interactions, and are not flat surfaces but have depth probably more
like a molecular sponge.24 At some point at the very periphery of
the cell is a surface contour called the slip plane where a physical
demarcation exists between the charged surface molecules, with their
tightly associated counterions, and thaions in the solution bathing
the cell. The negative charges on the surface are balanced by
positive charges in solution. These counterions of the extracellular
space crowd up near the surface and their concentration falls off
exponentially along radial directions from the cell. Most of the
counterions lie beyond the slip plane (Appendix I, Figure 20) and
therefore there is a potential (zeta potential) at this plane
representing the unbalanced charges at the surface. When the cell is
placed in an electric field the force of the field acts on the charges

at this plane and the cell moves toward the positive electrode.

11

12

Its velocity depends upon the zeta potential at the slip plane and
it is this velocity which is taken as the electrophoretic mobility.

The relationship used to obtain the actual surface potential or
surface charge from the mobility has been discussed by many authors
(see review by Abramsonzs). The Smoluchowski equation28 relates

mobility, u, to the zeta potential, f,

e 8"

MW

where a’is the dielectric constant and“ is the viscosity. The
valuea- of these two constants are taken to be those of the bathing
solution.(This may be a source of error.26) This equation is only
valid for particles or cells or a suffiiknt size,(approximately

10 micrometers or larger) and it assumes that their surfaces are
impenetrable to ions.

The surface charge, 0', is related to the mobility by the equation
ll: '9':

Kc
where l/K is the Debye-Huckel double layer thickness, and represents
the distance over which the potential drops to l/e.

These equations are only approximations and have been modified
by several authors to take into account membrane penetrability,27
microvilli and surface contours,24 pH changes at the surface and
membrane conductivity.28 However, the inaccuracy of the relationship
of electrophoretic mobility to surface potential or charge is not of

major importance in the use of the technique.

Most of the information obtained through whole cell electrophoresis

13

comes from measuring differences in mobility which represent small

differences in the amount of charged moieties on the cell surface.

An average mammalian cell exhibits 108 - 109 net negative charges

on it's surface an measured by electrophoresis. Changes of 10% are

easily measured. This represents the gain or loss of 107 - 108

charges or approximately 10.16 moles. The nature of these changes
are determined by specific treatment of the membrane, usually with
enzymes which will cleave off known groups.

The surface charges of particles, bacteria, viruses, sperm
cells, blood cells, and tissue cells both normal and transformed,
have been examined with whole cell electrophoresis by many

28,29,30

investigators over the past 40-50 years. It wasn't until

. . . . . 31
the discovery of an enzyme, neuraminidase, from Vibrio cholerae ,

 

. . . . 32
which was shown to cleave Sialic aCid groups from the cell surfaces
that interest in the technique rose. Sialic acids were found in

24

great amounts on the surface of some cells, ' representing

as much as 80% of the surface charge on red blood cells. These sialic
acids were found to be bound to glycoproteins 34 and glycolipidsBs'36
through glycosidic bonds. The carboxylic group on the sialic acid
remained free, and at pH 7.0 contributed a negative charge to the
cell surface. Cells of all types were then examined in order to

see if they had specific sialic acid profiles. Of particular

interest was the turnover rate, the topography of residues, the levels
during phases of the cell cycle, the role in antigen-antibody

7
interaction and possible changes in transformed cells.3

14

8. Effects of pH and Ionic Strength

The existence of a group like sialic acid with a pK between 3.0
and 4.0 had been expected by investigators because the isoelectric
point of many cells is in that range. The isoelectric point was
first determined by Kozawa 38 in 1914 by measuring the mobility
of red blood cells in buffers of varying pH. At a particular
pH the cells stopped moving and below that pH they reversed
direction. The point where there was no movement, and therefore,
no net charge on the cell,vus their isoelectric point. On many cells
this occurred between pH 3.0 and 4.0. Heard and Seaman24 examined
the red blood cell in great detail with a variety of buffers and
ionic strengths. They determined ranges of pH over which the red
cell was stable as a function of the ionic strength.

Altering the ionic strength of the measuring solution and
making it isotonic with a nonionic molecule, like sorbitol, is
another way to obtain more information from the mobility measurements.

In normal ionic strength buffers the membrane groups that lie
as little as 10 A under the surface do not contribute to the zeta
potential. This is due to the abundance of counterions which can
diffuse into the membrane and neutralize the charges. By lowering the
ionic strength the electric potential of the membrane charges
extendsfarther away from the cell due to the lack of counterions.
The groups deeper in the membrane then contribute more to the
surface charge and the zeta potential.24 Mobility measurements at
lower ionic strengths therefore have components which represent

charges deeper than 10 3. ( For greater detail, see Appendix I.)

15

The red cell shows a membrane ionic strength profile which suggests
that it has little thickness observable with mobility measurements,
. . . 39,40

whereas the oppOSite is true for other mammalian cells.

(See Appendix I )

C. Effects of Enzymes and Other Surface Effectors

Enzyme studies improved the value of electrophoretic mobility
measurements because specific groups could be removed and the
resulting changes in surface charge noted. Neuraminidase and
proteolytic enzymes showed thatsialic acid and proteins or
glycoproteins accounted for a large amount of the surface negativity
of the red cell.41

The use of ribonuclease by Weiss and Mayhew71 showed that there
is probably RNA on the surfaces of some lymphocytes and tumor cells.
More will be mentioned of this in a later section.

Mehrishi 30 treated cells with 6,6'-dithiodinicotinic acid
which binds specifically to free -SH groups. Upon binding, the
molecule is cleaved. The half remaining bound to the cell has a free
carboxylic acid group which adds a net negative charge to the
cell. The other half is in solution and it has a characteristic
absorption peak at 344 nm. Using this method, Mehrishi42 showed
that no detectable levels of -SH groups existed on red blood cells,
however, it changed the surface charge of lymphocytes by 10%,
Ehrlich ascites by 15%, and platelets by 5%.

Mehrishi also showed that the platelet has a membrane
susceptibility to alkaline phosphatase which removes about 8%

of the cell's negative charge.30

16

In certain cells hyaluronidase also removes surface coat material
and lowers the mobility. Sato et al.40 showed that tumor cells
treated with X-ray irradiation underwent a surface reorientation
that moved the hyaluronidase sensitive sites deeper in the membrane.
They determined this by lowering the ionic strength of the measuring
buffer to probe the charges deeper in the membrane.

Other treatments of cells have been used besides those designed
to discover specific groups on the surface. Mehrishi 42 has incubated
cells with the synthetic polypeptides; polylysine, polyarginine, and
polyornithine and found that cells can be made less negative, zero,
or even positive depending on the concentration of these molecules.
Cells of similar mobilities required differing amounts of polycations
to reduce their charge. This may be due to varying topographical
distributions of charges on the surface. Many of these polycations
also have anti-tumor properties. It was suggested that the reduction
in surface charge could result in increased phagocytosis by host cells.

In similar experiments, Mitchell and Cater43 showed reduction
in mobility by polycations, ptomaines, inflammatory mediators,
antisera, and heparin. Heparin served to reverse the mobility
lowering of polyamines and protamine binding, and to partly reverse
the effect of antisera on BP8 tumor cells. Their most interesting
discovery is that lymph node cells from animals immunized with BP8
tumor cells could be divided into two populations by polyamines,
one negatively charged and one positively charged. In a later paper,44
they showed that only the negatively charged cells could confer
immunity to BP8 tumor cells in C3H mice and protect them from a

subsequent tumor challenge

17

Lymphocytes:higeneral show a heterogeneous distribution of
electrophoretic mobilities. It has been shown that the T and B
cell populations can be separated by preparative electrophoresis.4s
Ware et al.46 have examined these two populations and have shown that
they have different mobilities andxifferent susceptibilities to
neuraminidase. They also showed that the T-cells can have subpopulations
which suggests that there is a correlation between the different
functions of T-lymphocytes and their surface properties.
Conconavalin-A has been used in many membrane studes and in a
few electrophoretic studies. Blume et al.47 showed that the binding
of Conconavalin-A to thymocytes resulted in an increase of mobility
which they suggested was due to cellular response rather than the
binding of Conconavalin-A alone. The effect occurred at levels of
Conconavalin-A of 10-12 ug/ml which amounted to 220 molecules per
106 cells. Therefore Conconavalin-A probably induced a change in the
thymocytes' metabolism which resulted in an alteration of the membrane
organization. In this case mobility measurements serve as a very

sensitive .monitor of binding activity.

D. Changes in Tumor Cells

In 1956 Ambrose48 reported a difference between the mobility of
normal and tumor kidney cells. The tumor cells exhibited a higher
surface charge than the normal cells. Others 37 showed that some
tumor cells had the same mobility as normal cells but were more
susceptible to neuraminidase. Fuhrmann49 showed this for liver cells.
The proliferating normal liver cells and hepatoma cells had higher

mobilities than nonaproliferating normal cells but the tumor cells

18

were more sensitive to neuraminidase. Other reports37 suggested a
similar correlation, but after several other cases were examined
there did not seem to be a distinct correlation between malignancy
and surface charge level. Vassar50 noted differences between
mesenchymal tumor cells and carcinomas in some cases, and there was
a difference in the sensitivities to neuraminidase but there was no
difference between the normal and malignant cells examined. Weiss51
also examined cell surface charge as a function of the degree of
malignancy and found no significant differences. In general, however,
any change in mobility associated with tumorigenesis is usually an

. 52
increase, which is also typical of fast growing cells.

III. Tumor Membranes

It has long been hoped that there would be a simple, measureable
difference between tumor cells and normal cells. The search for this
key difference has been extensive, but no such difference appears
to exist. The membranes of tumor cells have been examined ever
since early studies indicated that contact inhibition was lost in
tumor cell culture553, and that tumor cells were less adhesive.54
This search has led to the examination of surface proteins, glyco-
proteins, glycolipids, glycosaminoglycans, lectin agglutinability,
lectin receptor mobiltiy, membrane bound enzymes, membrane transport,
proteolysis, tumor associated antigens, the H-2 antigen complex,
antigen shedding, and the different cell-cell interactions. Each
of these areas, as well as others, have volumes of reports addressing
the issue of tumor cell membranes and several good reviews exist which

55'56157'58 SO far, however, the

attempt to piece it all together.
only safe conclusions that can be drawn are that there are a number
of changes that sometimes occur in tumorigenesis but they are
not substantially different from the changes normal cells undergo
during rapid growth or in mitosis.

This does not mean that these findings are unimportant because
the membrane is still the mediator between intracellular and extra-
cellular activities. The interaction with the environment that normal
cells undergo during mitosis and rap1d growth is critical to the

understanding of tumor cell interactions with the host even if the

basic cause of the cancerous state lies more deeplywwithin the cell.

19

20

The following is a brief review of the currently known changes
in membranes associated with neoplastic transformation. It is a

synopsis of the material in the major reviews cited earlier.

A. Changes in Cell Behavior

Malignant transformation always shows some departure from the
normal pattern of contact inhibition. Each cell type has its own pattern
of migration and behavior when it makes contact with neighboring cells.
Malignancy often introduces subtle changes in the normal migration
and growth of these cells as viewed in vitro. 53 Epithelial and
glial cells, which usually form strongcell-cell adhesions, stop their
locomotion and contract, show after transformation, a diminished
tendency for-intercellular adhesions. Fibroblasts show variable
patterns after transformation, with increased overlapping, less
adhesions to other cells, and 'continued.membrane ruffling. Malignant
lymphoma cells appear less mobile than normal lymphoblastoid cells.
In conjunction with contact inhibition type behavior, tumor cells often
lose density-dependent growth restrictions and in vitro many are
capable of growing in suspension with no anchorage dependence.

These behavioral changes in vitro have not been thoroughly
reproduced in vivo, so it is not known if changes in migratory patterns
and contact interactions are important in the tumor-host system,

particularly in metastasis.

B. Membrane Components
A variety of membrane alterations occur in the malignant state,

particularly in glycolipids, glycoproteins and proteins.

21

l. Glycoproteins

The carbohydrate portions of glycoproteins undergo changes
in size and in sialic acid content. Sialyl tranferase has been shown
to be less in some cells, resulting in incomplete synthesis of
certain large glycoproteins.35 0n the other hand, small molecular weight
glycoproteins have more sialic acids in transformed cells and in
fast growing normal cells.55 Virus transformed, non-tumorigenic
kidney fibroblasts showed longer and more highly branched
oligosaccharides of the surface glycoproteins. 56 This has also
been found in tumor cells after virus infection. Similar earlier
work seemed to indicate larger glycopeptide fragments in tumor cells
and mitotic cells when the cells were treated with trypsin and pronase.
This work is comparable to that of the virus induced tumors.

Probably the most agreed upon change in glycoproteins is the loss of
”a ZOOK-ZSOK glycoprotein58 which goes under several names;
fibronectin, LETS (large external trypsin sensitive) protein, Z protein,
galactoprotein,and fibroblast surface antigen. This protein exists
in high levels in membranes of normal cells but transiently disappears
during mitosis. It can be removed with trypsin and its disappearance
occurs simultaneously with induction of the cell into the cell cycle.
This had led several investigators to suggest that this protein is
involved with growth control. Tumor cells appear to be lacking in
this membrane protein but they also may expel it into the medium.
Lectin binding sites have been found on this protein, and the protein
appears to interact with the cell's fibrillar structures. It has even

been proposed that the lectin cap formation in tumor cells is due to the

22

lectins binding to less competive, mobile proteins which are
normally not bound when the LETS protein is available. 58 Future work
will probably elucidate the role of this protein.

2. Glycolipids

The glycolipids in the membrane of tumor cells seem to be less
complex than those of normal cells with fewer molecules present and
they are not accessible to antibodies, enzymes and lectins. This
was determined on the NIL hamster cells by Hakonori.59 Most of the
virus transformed cells examined also showed similar reductions in
complex glycolipids. As normal cells grow and contact each other there
is an increased synthesis of these glycolipids, suggesting that they
may be involved in growth control. Spontaneous tumors of the mouse
do not show this reduction unless the cells are cultured for several
generations§5which raises thaquestion of relevance between
in vitro and in vivo findings.

In general, there are measurable changes in both glycoproteins and
glycolipids between normal, transformed, transformed and tumorigenic
cells, and cells during development.60 These changes are not specific to
cancer, but may be more correlated to levels of differentiation and
cell growth rate.

3. Proteins

The proteins that are altered in tumor cell membranes (besides
glycoproteins) are transport enzymes, adenyl cyclase, Na+ - K+ - ATPase,
proteasesand glycosidases.55'56

Sugar entry increases in transformed cells as does entry of certain

. 55 + + . .
amino aCids and phosphates. Na and K movement is increased due to

23

the 4-5 fold increase in Na+ - KI - ATPase activity. It has been
suggested that malignancy is due to the increase in cell transport
which also brings in molecules which stimulate growth.55 It is perhaps
easier to envision the increased transport as a secondary response to
increased energy needs of malignant cells.

Elevated levels of glycosidase activity in human tumors with
respect to similar normal tissues have been reported, however these
measurements are made on cell homogenates. It is postulated that they
can influence both the tumor cell origin and the surrounding cells.
The increased levels are seen in some transformed tissue culture
lines but not all. 55

Protease changes have been reported and raise some interesting
questions. Since protease treatment of normal cells can induce the
mitotic cycle, remove LETS protein, alter other glycoproteins, change
glucose transport, increase agglutinability with lectins, and make
them more reactive to antibodies, a role of proteases in cell cycle
control has been hypothesized.57 A tumor cell could be releasing
proteases at precise moments in the cell cycle which stimulates the
next round of division. Tumor cells also release a plasminogen
activator which normal subconfluent cells release. However, unlike the
normal cells the tumor cells release it continuously. Roblin61 has
shown that tumor growth serum depleted in plasminogen results in major
morphological changes in tumor cells back to their normal configuration.
This indicates that the tumor activated plasmin causes surface
alterations of the tumor cells, which may be important to the trans-

formed state. The importance of protease action on the membrane

24

could be great. Antigen sites could be altered or their steric
orientation changed, membrane permeability, membrane-cytoplasmic
communication, surface receptors, etc., could all be significantly
influenced by protease action. Further work in the area should be
very enlightening to growth and cell-cell interactions.

4. Lectins

The agglutination of tumor cells by lectins has been known
for several years. The capping phenomenonand.the deduced
greater mobility of binding sites has been intriguing, but no real
meaning has been attributed to the phenomenon. Mannino and Burger62
suggest that the surface change detected by agglutination commits the
cell to enter the next round of the cell cycle. Since normal cells in
mitosis exhibit agglutinability there does seem to be a connection.
The question is whether or not it is a causal relationship. They
blocked the Con-A binding sites of 3T3 cells but did not agglutinate
the cells and showed that if the blocking agent was added in early
mitoses the cells finished the cell cycle and entered G6. The removal
of the block allowed the cells to enter the cycle again. Thissuggested
that a stop signal was necessary for cell control and that stop
receptor was associated with lectin binding sites. This raised the
question of what the stop signal molecule might be.

5. Antigens.

The antigen alterations of tumor cells is another large area
of investigation, especially for the possibilities of effective

63,64

anti-tumor immunotherapy. Tumor specific antigens have been

55
demonstrated that are capable of eliciting an immune response.

25

Viral antigens are usually present in virus transformed cells and fetal
type surface antigens are also present on some tumors. 65 The
fetal antigens, carcinoembryonic antigen (CEA) and alpha-fetoprotein
(¢&FP) are released by certain human tumors but their significance is
not known. They are being used as a monitor for recurrent tumors in
some cases. The release of antigens like CEA and ¢xFP has spurred interest
in the idea of antigen shedding by tumors.66 Some results have shown
that antigens released by tumors combine with antibody and form
complexes that act as blocking factors which suppress the response of
sensitized lymphocytes. Shed antigensalone are also claimed to be
inhibitory to sensitized lymphocytes6§ The possiblity of large numbers
of released antigens simply overwhelming the immune system is yet
another theory of tumor protection.66

The field of tumor immunology is immense65 but it seems that the
most interesting work is not on the antigen sites of tumors but in the
host's response system. The possible suppression activities of tumor
cells is one very important aspect which has only recently been
examined extensively.67 Another related area is the importance of the
histocompatibility complex H-2 or HL-A. There is evidence68 that they
are involved in antigen recognition by the cells of the immune system.
This cell-cell interaction at the membrane level, particularly the
necessary conformations of antigen and 8-2 complex could be an area for
new breakthroughs and of importance to tumor immunology.

I have outlined briefly only the major changes in the membranes
of tumor cells. Many internal differences have been cataloged. Many of

them are probably the result of tumorigenesis and not the cause, as are

some of these membrane alterations.

IV. Surface Nucleic Acids

A. Surface RNA
In 1952, Lansing and Rosenthal69 showed evidence for ribonucleic
acid on the surface of Arbacia eggs through light microscopy

staining with toluidin blue. They also used the plant Elodeacnnadensis

 

and measured calcium oxalate formation before and after ribonuclease
treatment. The peripheral cell edges lost staining ability after
RNase treatment in the first case and less calcium oxalate crystals
formed after RNase treatment in the second case. They suggested
that a membrane RNA molecule was important as a calcium repository
and could possible serve as a transmembrane carrier for certain
molecules.

Chaudhari and Lieberman7O found RNA on the surface of liver
nuclei by using particle electrophoresis and treatment with RNase.
The RNA was only on the nuclei of liver cells of partially
hepatectomized rats, where the cells are undergoing much greater DNA
synthesis rates.

The next report of surface nucleic acids was made by Weiss
and Mayhew71. They used whole cell electrophoresis to show that
cultured LlZlO tumors and RPMI no.41 cells exhibited ribonuclease
sensitive sites on their outer surfaces. Mouse, human and chicken
erythrocytes did not change after RNase treatment, whereas all
cells measured were sensitive to neuraminidase, giving a reduction
in mobility.72 Further work of theirs72 indicated that these RNase

senSitive Sites were able to bind Ca ions, and at low concentrations

26

27

of Ca++ ions, these were the only surface binding sites. They suggested
that there was membrane bound RNA and that it was a structural
component with one of it's functions being a reservoir of divalent
cations.

They continued to study this surface RNA39 and showed that it
did not easily wash off cells, nor could it be added to cells by
debris from broken cells. The rate of cell metabolism and culture
division time did, however, determine the level of surface RNA.

Cells deprived of serum in their tissue culture media lost a
significant amount of charge due to RNA, whereas this starVation did
not effect the level of neuraminidase sensitive sites. This led
them to suggest that the surface RNA was a relatively labile group
but it was not an adventitious group.

Mayhew and Weiss71 also showed this RNase sensitivity on mouse
lymphocytes, polymorphonuclear leukocytes, thymus cells, liver cells,
and in $37 ascites cells. The tumor cells and transformed tissue
culture cells exhibited the greatest amount.

Mayhew52 also showed that the surface charge of tissue culture
cells (RPMI no. 41) increased during the M phase of the cell cycle.
An increased level of surface sialic acids seemed to be responsible
for this change. When non-synchronized cells were examined at
different growth rates, however, the sialic acid content was constant
even though the surface charge increased with increasing growth rates.
This, and their previously described wonk,suggested that fast growing
cells have an increased amount of negatively charged groups which

are susceptible to RNase removal but not to neuraminidase removal.

28

They further substantiated that hypothesis by examining normal
and rapidly proliferating lymphoid cells in the mouse.73 The rapidly
growing, moreimmature cells contained RNA on their surface while the
normal controls did not.

Weiss et al. continued this work by exploring the relative density

of these RNase and neuraminidase sensitive sites on the surface

of Ehrlich ascites cells.74'75

The cells were fixed in glutaraldehyde
and treated with positively charged, colloidal iron hydroxide
particles. The cells were sectioned and viewed by electron
microscopy. When cells were treated with RNase - leaving behind the
sialic acid residues - the distribution of the particles was a uniform
Poisson distribution. When the cells were treated with neuraminidase
- leaving the RNA molecules - the particles were found in clusters
and the clusters were distributed in a uniform Poisson distribution.
When both enzymes were used, the remaining 66% of the negativity
of the cell surface was not strong enough to hold any particles
bound. From this they suggested that sialic acid on the cell was
clustered in small patches, each patch with enough negativity,rv30
electrons, to hold one particle bound. The RNA was clustered in
larger patches, each patch capable of binding several particles. The
remainder of the cell's negative charge was distributed evenly
and its density wasn't great enough to keep a colloidal iron particle
bound.

Other electrophoresis work which has demonstrated RNase
susceptibility on cell surfaces was done by Woo and Cater76. They
were examining antigens on adult liver cells, fetal liver cells, and

hepatomas in the mouse. During their studies they showed that one

29

of the similarities between the fetal and hepatoma cells was a
susceptibility to RNase treatment. RNase did not effect normal liver
cells but decreased the mobility of fetal cells by 17% and hepatoma
cells by 29%. They also examined the BP8 ascites tumor and lymph

node cells and showed a 39% reduction in mobility with RNase in the
tumor cells as compared to a 4.8% reduction in normal lymphnode cells
and 13% reduction in lymph node cells from animals immunized with

BP8 cells. They spent considerable effort in the attempt to wash

the RNA off of the cells, with no effect, and to make sure that

the RNase enzyme was not binding and causing reduced mobility. They
suggested the RNA was real and hypothesized that it might simply

be due to mitotically active cells or might serve a role in cell-cell
recognition or information transfer.

In another electrophoresis study, Droege et a1.77 examined
chicken lymphocytesfrom blood, spleen, bursa, and thymus. The blood
lymphocytes showed a 22% reduction in mobility after RNase treatment,
whereas none of the lymphocytes from other organs were sensitive
to RNase. They suggesed that this RNA could serve as an endogenous
adjuvant in the lymphocytes response to antigen.

Besides the electrophoreSis work there have been a few other
studies suggesting surface RNA. During the isolation of liver

membranes investigators have found 1-2%78 and 2-4%79'80

RNA in
their membrane preparations. The problem with this is that this
quantity of RNA could easily be contamination. In fact, most

researchensattributed RNA in their membrane preparations to

. . 8 . .
contamination. DaVidova and Shapot 1 examined more closely this RNA

30

in membrane preparations using rat liver and hepatoma cells. They
found that the plasma membrane-bound RNA contained different size
components than the endoplasmic reticulum RNA. They also added a

known amount of radiolabelled contaminant RNA to their liver homogenate
and determined that only 4-7% of the membrane associated RNA was
contaminated by cytoplasmic RNA. They suggested from the isolation
techniques that the RNA is bound to protein and that this is

complexed with lipid.

Using L cells Glick82 examined the RNA associated with membrane
preparations and compared it to ribosomal RNA. She found that
the base compositions and profiles on sucrose gradients were
similar. However, the synthesis rate of surface RNA was greater
(as measured by uptake of 38 uridine) and the surface RNA was less
sensitive to Actinomycin-D and to RNase digestion. She suggested
that protein producing ribosomes are associated with the cell outer
membrane and that they synthesized specific types of proteins. The
question which was not addressed by this study was whether the
membrane RNA examined originated from the inside or the outside
of the membrane.

The possibility of an RNA-protein complex of some sort located
on the surface was supported by Rieber and Bacalao.83 They labelled
Chinese hamster ovary cells with 32P-phosphate and 3H-leucine
and then treated them with trypsin. The trypsin treatment released
radioactive phosphate which was sensitive to RNase treatment and
also contained protein. With several methods they tried to assure

that the trypsin was not getting into the cell and releasing cellular

31

phosphate containing macromolecules. When cells were incubated

with Btz-cAMP-testosterone they showed a greater amount of a higher
molecular weight phosphate containing-RNase sensitive macromolecule.
They suggested that this membrane ribonucleoprotein may be associated

with cell proliferation and differentiation.

B. Surface DNA

In 1970, Singer and Reid84 examined the mucoid coats of cervical
cells under the electron microscope and found fibrous bundles and
filaments. Using a method to remove mucoid coats from cells they
examined rat thymus cells. 85 After using DNA extraction methods on
the cell coats they found a DNA which had a higher, more abrupt
melting temperature but lower hyperchromicity than intracellular
DNA. This DNA was shorter and aggregated into bundles and rings
and had a greater buoyant density. They suggested that it might be
partially single stranded and may contain repeating sequences. Reid86
presentedan elaborate argument that this surface DNA is involved with
cellular communication in an electronic network fashion.

Lerner et al.87 then reported a plasma membrane associated DNA
in diploid human lymphocytes. The DNA was isolated from.membrane
fractions and shown to be smaller in size than nuclear DNA. It

replicated during G at a rate 16% of the S-phase rate, whereas the

1

nuclear DNA's replication rate in G was only 3% of the s-phase

1
rate. They suggested that this DNA was different from the nuclear
DNA and was not a membrane contaminant. The membrane DNA was susceptible

to DNase while on the membrane, indicating that itvms in an exposed

32

position. Finally, they noted that the number of DNA molecules in the
membrane corresponded about one to one with the number of bound

IgG molecules. They suggested that this membrane DNA could be involved
in immunoglobulin synthesis, gene amplification, or immunologically
associated information transfer between cells.

In 1975, Aggarwal et al.88 suggested the presence of DNA on the
surface of tumor cells. Using platinum-pyrimidine complexes as
electron dense stains for electron microscopy, they noted patches
on the surface of tumor cells which could be removed by neuraminidase
or DNase, but not RNase, trypsin, or hyaluronidase. Eight tumorigenic
cell types showed patches, while 11 non-tumorigenic tested negative.
They suggested that DNA was present on the outer surface of tumor
cells, and was bound in association with sialic acids.‘Rosenberg121
suggested that this surface DNA could have an immunosuppressive role
by acting to mask tumor antigens on the cell surface.

Further work along this line by McAllister et al.89 suggested
that other cells besides tumors also stained on the surface, and that
the technique needed refinement. She also suggested that under
certain conditions other molecules than nucleic acids could be stained.

Aggarwal strengthened his argument by using an autoradiographic
technique in conjunction with electron microscopy.90 He labelled cells
with 3H thymidine and plotted the number of grains versus distance
from the nucleus. He compared this to a similar plot of the densities
of platinum staining versus distance. There were distinct peaks inside
of the nucleus, at the nuclear membrane, and at the plasma membrane.

Russell and Golub91 found a subpopulation of leukemic spleen

33

Cells capable of suppressing anti-sheep erythrocyte antibody response
of normal mouse AKR spleen cells. They showed that this subpopulation
of cells had DNA located on their surface. The suppressor population
could be isolated by passing the cells down a column containing
bound anti-calf thymus DNA antibody. Treatment of these cells with
DNase eliminated the suppressor activity. In the assay for suppressor
activity cell contact was required, which could mean an physical
transfer of DNA or a product was necessary for suppression.
Hollinshead and Stewart92 have also made an observation involving
membrane DNA. They found a DNase sensitive factor on the membranes
of cells from human Oat cell lung cancer. This appears to be involved
as an inhibitor, protecting the tumor cells from an immune response

directed against the cell's tumor associated antigens.

C. Cell Released Nucleic Acid.

There have been many reports of RNA molecules that have an
effect on cells. The term immune-RNA or informational-RNA or
transfer-RNA is used to indicate RNA which is extracted from cells,
usually sensitized cells, and given to other cells or injected into
animals. Transfer of specific immunity has been demonstrated by this
technique as well as induction of differention. Several reviews

93,94,95

exist on this subject but they do not discuss the release of

RNA from cells. A few studies have shown a release of an RNA-DNA

complex which will be discussed below.

The in vitro release of DNA from lymphocytes,97'101'loo'105'99

98
bacteria,96 frog auricles,97 and the water mould Allomyces arbuscula

- .. 98
has been demonstrated in recent years. Khandjian et al. and Stroun

 

34

et al.97 described a complex of RNA-DNA and protein as being released.
Stroun et al.97 have shown that extracellular concentrations released
from lymphocytes attained a steady state level which was recognized
and regulated by the cell. They showed that the RNA-DNA-protein complex
was quite stable and required drastic proceedures for removal of the
RNA. The complex was not susceptible to RNase degredation. Only after
phenol extraction, DNase digestion, and chloroform and sodium
perchlorate extraction was the remaining nucleic acid RNase sensitive.97
They also showed that this complex was capable of extracellular
DNA synthesis.

Rogers100 described a DNA which was excreted by phytohemagglutinin-
stimulated lymphocytes. He showed that the DNA excreted was
homologous with only 10% of the nuclear DNA and was smaller than the
nuclear DNA. He followed the course of labelled DNA from the nucleus
to the supernatant on normal and stimulated lymphocytes.101 The DNA
traveled from the nucleus to the supernatant during the course of 3
days. In the stimulated cells 3-4 times the amount of the DNA
was excreted. He suggested that this excretion of a particular
segment of the cell DNA may be very important in the activation of
lymphocytes.

The DNA released by lymphocytes has also been shown to attach

105

to red blood cells and cause rosette formation. The significance

of this is not known.

Developement of ElectrOphoresis Apparatus

I. Backround and Overall Design

The technique of whole cell electrophoresis consists of placing
a cell in a solution containing an electric field and determining
its induced velocity. The cell migrates toward one of the electrodes
with a velocity directly related to the value of its average
surface charge. Under normal conditions cells are negatively charged
and therefore migrate toward the positive electrode. The electro-
phoretic mobility is measured as velocity per unit electric field,
u/sec/Volt/cm. The standard technique for determining this mobility
was to watch a single cell and determine the time it took to travel
a set distance thereby obtaining its velocity. This is a tedious
proceedure because several cells had to be examined in this way
to obtain a relevent representation of the population. This also
exposed the cells to the electric field for prolonged times. In
order to be able to collect data rapidly on a large number of cells
a new technique was necessary.

A new method of measurement was developed using a technique
first described by Costa and Penniman.106 It consisted of
mechanically moving the image of the cells, which are viewed
through a microsc0pe, in a direction opposite to that induced by the
electric field. When the two are equal the cells appear to stand
still and the mechanical movement just equals the field induced
velocity of the cell. Their apparatus consisted of using a prism
oriented in the optical path of the microscope which shifted the

cell image with delicately controlled minute rotations. These were

35

36

controlled with relatively elaborate electronic servo systems.

This technique was altered for the present apparatus by
incorporating a sliding mirror which, as shown in Figure ‘1 , serves
to shift the image at the reticle located in the eyepiece of the
microscope. The mirror is driven by a motor through a worm gear and
the speed of the motor is accurately controlled and monitored.

When a measurement of cellular mobility is made a 'cloud' of cells

is watched and the velocity of the mirror is adjusted until the
'cloud' appears stationary. The motor speed is then directly
proportional to cell velocity. This method allows for rapid deter-
minations of large numbers of cells thereby obtaining statistically
reliable measurements with a minimum of exposure for the cells.

(A five minute period is usually sufficient to obtain a mobility value
with a coeficient of variation of 1-22)

To further speed the data taking process a television camera was
attached to the eyepiece so that the cells could be viewed on a TV

monitor, reducing the eyestrain from prolonged microscope work.

II. Measuring Chamber and Electrodes.

The measuring chamber was obtained from Thomas Inc. and is the
f1at,parallel plate design as shown in Figure ’2 . The distance
between the inner surfaces of the plates is 400 micrometers. This
distance times the width of the chamber defines the cross sectional
area of the chamber and equals 0.048 cmz. A water jacket was added to
the chamber, Figure 3, and the chamber is connected to the dual
purpose electrode chambers/fill and drain ports. Stopcocks allow the
sample to be added and drained or the electrodes to be added to the

circuit when the electric field is applied.

37

 
 
   

motor
objective

ndrror

:33 grid in eyepiece

light from objective

/

 

 

 

 

1f

Image/2x7 image Shift ’1

plane / 1’
(grid) " )J , J
’ n

I J ‘3
’ ' mirror mount
I I
’ I
4:;

 

 

mirror displacement

Figure 1. Diagram of mirror movement.

width 400 um

I

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

top view (1, 4" I"// 770
x J
side view CL; fi—D
‘ XX f/
4;, 2 thin plates
Figure 2. Diagram of measuring chamber.
fill
waste
N .
water jacket
stopcock/a ’77 ‘I:; r
electode
chamber F’ chamber

 

 

 

 

Amp Power

ML supply

Figure 3. Diagram of chamber and electrode circuit.

 

 

 

 

 

 

 

 

39

The field strength inside the chamber cannot be measured
directly and therefore it is derived by measuring the electric
current flowing in the electrode circuit (the ammeter in Figure 3 ).
The currents flowing in all elements of the circuit are identical
therefore this represents the current flowing in the measuring
chamber.

The electric field is derived from the current by the following
equation:

L

E=KA’

where E is the electric field, I is the current, K is the conductivity
of the buffer, and A is the cross sectional area of the chamber.

K is determined with the following equation:

I
K = Ibuff . K
std

where a 0.1 N KCl of known conductivity of 1.119, at 18°C is used

std ,

to determine K. The standard is placed in the chamber first and the
current is measured at a fixed voltage between the electrodes. The
standard solution is then replaced with the buffer and the current
is again measured.

The electrodes are Agl AgCl reversible electrodes in saturated
KCl solutions. They are made by plating AgCl onto silver wire in a
0.1 N HCl solution with a current of 10-15 mA for 20 minutes.
The saturated KCl solution in the electrode chambers is separated
from the physiological buffer solution in the measuring chamber by
plugs of glass wool. Risks of cell damage due to leaking of the KCl
through the plugs during measurements is reduced by 10-15 cm

connecting tubes from the measuring chamber to the electrode chambers.

40

111. Mirror Movement

The microscope used for this apparatus was the NikmMS inverted
microscope. It had the advantage of a right angle bend design in the
Optical path from objective to eyepiece. This allowed for placement
of the movable 45° mirror as shown in Figure l . The motor attached
through the worm gear to the mirror contains a gear reduction unit
and a built in tachometer to monitor motor speed. The motor input
and tachometer output are attached to a speed control power supply.
The motor and power supply were purchased from Servo-Tek Inc. The
tachometer output is also monitored with a digital volt meter (DVM)
and is used to determine cell velocity.

The mirror velocity is proportional to the velocity of the
cells with a multiplicative constant given by the power of the lens
objective. The mirror velocity, in turn, is proportional to the
r.p.m. of the motor with a proportionality constant given by the
pitch of the worm gear.

The mirror movement was calibrated with a stopwatch. The stopwatch
timing of a stationary particle in the field of view as it was moved
by the mirror gave an actual value for the velocity. The tachometer
output from the mirror is also related to the velocity observed by the
equation,

0.3545

Tach voltage x Y

= velocity

where the factor 0.3545 is a constant obtained from motor speed,
the gear ratio, and conversion factors. After calibration it was
determined that this factor was off by 1.692 and was therefore adjusted

to the value 0.3605. (Y is the power of the objective.)

41
The overall schematic for the switches and interconnections is
shown in figure 4 . Provisions are made to switch polarity of the
electrode voltage and reverse direction of the motor rotation.
A chart recorder is used to monitor electode current while a DVM
Bused fortmmtachometer voltage. Further details and operating

instructions are given in Appendix III.

42

 
 

 

=5.

.maumwmmmu mammnonmouuooam no Eaummfiv uflnuuwo .v wunmfih

3

 

 

 

:7“: _ _.!

J...
1.34... ..

 

 

 

.2239

 

.29:

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

A AAAL l
vvvvvv V

 

 

 

 

 

 

 

 

 

 

 

 

 

Materials and Methods

I. Buffers and Solutions
The buffer used for all electrophoresis measurements was a
citrate-phosphate base buffer with a pH range from 2.2 to 8.0.
(NaCl = 0.137M; KCl = 0.004M; NaZHPO4 = 0.011M; Citric acid =
0.006M for a pH = 7.4) For pH = 6,citric acid = 0.0029M and NaZHP04=
0.01M 3 for pH = 4.6, citric acid = 0.0052M and NaZHPOA = 0.009M;
for pH = 3.0, citric acid = 0.016M, NaZHPO4 = 0.006M.
When the ionic strength of the solutions were changed,.isotonic
(52) mannitol was added tothis buffer.
All incubations with enzymes were performed in a modified
Hanks solution (NaCl = 0.137M; KCl = 0.005M; MgSO

4

CaCl2 = 0.00081M; glucose = 0.011M; NaHZPO4 = 0.001M; and

Nazi-1P04 = 0.0045M). The bicarbonate has been removed to eliminate

= 0.0016M;

the need for a CO2 atmosphere.
When cells are removed from the animals they are diluted
with a 0.9% NaCl solution.
The perfusion buffer used during the isolation of cells from

the liver, fetus, thymus, spleen and solid S-180 consists of NaCl at

0.08 M, EDTA at 0.027M and sodium citrate at 0.027M.

II. Cells

The S-180 mouse tumor cells are grown in ascitic form in the
peritoneal cavity of the ICR mouse. 2, 4, or 8 x 106 cells are
injected into the animal and then at later times the animal was
sacrificed and the cells washed out of the peritoneal cavity with

saline. Twenty four hours after injection of the cells is denoted

43

44

as day l and so on from there.

The animals were injected with cisplatin on day 1 (day 3 in
some experiments) with a dosage of 7 mg/Kg. This cisplatin was
dissolved in saline and injected under sterile conditions with an
approximate injection volume of 0.5 ml.

After the cells were removed from the animals they were washed
by centrifugation to remove blood cells and were stored On ice for
short periods before incubations and befoneelectrophoretic mobility
measurements.

The P388 and P815 cell lines were grown in the DEA/2 mouse strain.
AKR spleen and thymus cells were obtained from.leukemic ( those
showing physical signs of sickness) and preleukemic (showing no
signs of sickness) AKR mice. The LlZlO and V79 cells were obtained
from tissue culture. Blood cells were obtained from mice via intra-
orbital bleeding.

To obtain- liver cells in suspension the animal was sacrificed
via cervical dislocation and the thoracic cavity was opened. The animal
was then perfused through the heart with citrate-EDTA-saline. Livers
of successfully perfused animals were removed and minced finely with
scissors. These pieces were then forced through a 40 mesh stainless
steel screen and the resulting cell and tissue mixture was then
allowed to sediment for approximately one minute at l g and the
supernatant was removed. This supernatant contained the whole cells
and the debris. The cells were cleaned by centrifugation in citrate-
saline.

The same proceedure was followed for spleen, thymus, fetal, and

the solid S-180 cells.

45

III. Enzymes

The enzymes used in the study were obtained from different

sources:
RNase A Miles Laboratories 36-511 5X crystalized -protease free
Sigma Type I " "
Sigma Type XI-A best grade "
Worthington RAF-OFA phosphate free "
DNase I Sigma Type I chromatographically purified
best grade
ICN 100575 2X crystalized

Neuraminidase Sigma Type V

Protease Calbiochemical -B grade
Trypsin Nutritional biochemicals l-300
Hyaluronidase Sigma Type IV

Immobilized DNase Worthington

Immobilized RNase Miles Laboratories

Eco Rl. Miles Laboratories

Enzyme concentrations for RNase and DNase, unless otherwise
noted, are 0.2mg/ml RNase, Sigma Type XI-A, and 0.1mg/m1 DNase, Sigma
Type I. The other types listed above were tried but were not found

to be different in their effects, so one type was used thereafter.

IV. Tissue Culture Media for Incubations

For the long term incubations in vitro the cells are placed
into tissue culture medflnn-Both of the following were used;
Medium 199 with Hanks base and 10% calf serum, and NCTC 135 with

Earles balanced salt + L-glutamine and 10% fetal calf serum.

46

v, Enzyme Treatment

Enzymes were dissolved in modified Hanks buffer and the cells
added. Incubation was at 37°C for various times in closed vials
containing no air with glass beads added or in open tubes. The
former were placed in a 37°C hot air incubator on a rotating mixer,
the latter was incubated at 37°C in a water bath. After incubation
the cells were washed once by centrifugation then placed on ice.
VI. Cell PathoIOgy

The cell population and cell morphology was primarily analyzed
by using smears on glass slidesstained with Wrights Stain. Peroxidase
Stain specific for neutrophils was also used and in a special
examination performed by Dr. Steven Stockham, veterinary pathologist,
methylene blue and PAP smears were also used. Peritoneal washes
were spun down and resuspended in a minimum fluid in order to
Obtain a cell suspension suitable for smearing onto slides. Slides
were examined with light microscopy.

VII. Analysis of Raw Electrophoretic Mobility Data.

Measurements from the RPM apparatus consist of pairs of left
and right readings. The cloud of cells is nulled with electric
field set for migration to the left and then repeated with the
electric field reversed. This proceedure of alternating the field
is continued until a representative sample is taken. There are
two numbers obtained for either a left or a right measurement, the
current reading and the tachometer voltage from the mirror motor.

These numbers are then entered into a computer program which converts

the current into an electric field value and the tachometer value

into cell velocity. Dividing the velocity by the electric field gives

47

the electrophoretic mobility in the|usual form of u/sec/Volt/cm.
The left and right mobility values are averaged to obtain
an overall mobility. A special kind of averaging is performed to
remove a bias in measuring that would occur due to changes in the
drift of cells in the chamber, to currents caused by sedimenting
cells, to the motion of cells due to their own mobility or due to
the electroosmotic effects. The problem arises from the serial
nature in which thaneasurements are recorded,ie. left, right, left,
right, etc. Normally the first left and right pair are measured
and averaged, and the second pair and so on. The final average is
taken from the average of these pair values. If all of the factors
I listed above are constant during the measurement of each pair
then their average is a truerepresentation of the cell mobility.
However, if these factors are changing during the measurement of
any one pair the value of the mobility will be slightly in error.
It is therefore obvious that the arbitrary choice of pairs in the
overall left-right sequence adds an artificial constraint. To
attempt to alleviate this, the 'left' should be partially averaged
with the previous 'right' in the train of measurements and the
'right' with the subsequent 'left' with modifications for the first
and last measurements. Therefore, a measurement is averaged with
the previous and subsequent values in the train giving % weights
to each of its neighbors. This proceedure does not alter the total
average value of the mobility, which is the average of all the
measurements, no matter what the conditions. It only reduces the
variation caused by the extraneous forces acting in time intervals

shorter than the time it takes to measure one pair of values.

48

VIII. Computer

The mobility measurements were computed on the CDC 6500
computer located at the Computer Center of Michigan State University.
The current readings and the tachometer voltages are entered
into a computer program and the mobilites are computed, the
average taken, the standard deviation computed, and the

deviations from the mean plotted.

Outline

In this section I will present a summary outline of the
different experiments that will follow in the Results section.

Some of the early experiments were designed to observe and
quantitate phenomena. The experiments following those were to
elucidate the observed lowered mobility after platinum treatment
and the susceptibility of the cell membranes to RNase and DNase
incubation. In particular, experiments were designed to test an
hypothesis of cell surface nucleic acids in several different ways
using whole cell electrophoresis.

In several of the reports in the literature, which were
mentioned earlier, cell surface RNA and DNA was described. Dr.
Weiss' laboratory performed many experiments with cell surface
RNA. They described its appearance on tumor cells and stimulated
lymphocytes, its correlation to cell growth rate, and it distribution
on the surface. They used electrophoresis to observe this surface
RNA. Other investigators discovered DNA on the surface mostly
using membrane preparations, surface staining, and radioactive
labelling techniques. They made comparisons to nuclear DNA by size,
homology, and melting characteristics.

In this study the first electrophoretic mobility observations
were on tumor cells as a function of tumor age in the animals.
Normal tumors were then compared to platinum treated tumors. There
was a distinct difference in mobility,so further observations were
made in an attempt to elucidate the reasons for the difference.

pH profiles were run to compare their isoelectric points, different

49

50

enzyme incubations were performed to determine if a specific
membrane group was effected by platinum treatment, and ionic strength
profiles were obtained to probe into the membrane.

These observations led to the hypothesis that cell surface
nucleic acids were present on the surfaces of the S-180 tumor cells
and that treatment with cis-Pt(II) caused this group to disappear.

To help substantiate the hypothesis, enzyme kinetics were
determined for RNase and DNase to assure a true enzymatic process
was occuring. Enzymes bound to large Sepharose and Agarose beads
were incubate with cells to insure only surface action. The
restriction endonuclease Eco-R1 was incubated with cells to further
strengthen the surface DNA hypothesis. Cells were exposed to anti-
DNA sera to see if there was any membrane binding. Positively charged
'platinum-thymine-blues' were allowed to bind to cells. Those cells
with surface nucleic acids would bind this molecule most readily
and their mobility would drop.

All of these experiments supported the hypothesis of surface
nucleic acids. Other experiments using neuraminidase and trypsin
and low ionic strength showed that these nucleic acids are only loosely
associated to the membrane. In vitro experiments further showed that
the nucleic acids probably have a turnover rate on the membrane of
2-4 hours. Other in vitro work indicates that cis-Pt(II), methotrexate,
mercaptopurine, and Actinomycin-D act in a similar way to cause
thaloss of this surface nucleic acid, probably by inhibiting its
synthesis within the cell.

Next, some immunological tests were performed to see if an

obvious role for the surface nucleic acids was evident.

51

These experiments lent little evidence one way aranother.

Finally, other cell types were examined. The solid S-l80,
the P388 and P815 ascitic tumors, the AKR spontaneous leukemia,
normal liver, fetal, spleen, and thymus, and the V79 and LlZlO tissue
culture cells. The normal cells showed no surface nucleic acids, the
cells of lymphocyte origin seemed to have only DNA on their surface

while other tumor cells contained RNA or both.

Results

The cell type used for the majority of this study was the
mouse S-180 tumor. This tumor has been serially transplanted in
ICR mice in our lab for several years. This study involves three
different strains of the S-180 tumor. The first, called S-180,
was the tumor used in our lab until the Summer of 1978. At this
point the animals were diagnosed as diseased and new mice and a new
tumor were procurred. This tumor was used until December 1978 and
is refered to as'interim 5-180'. The third tumor was obtained in
January 1979. This tumor has characteristics closer to the original
5-180 strain. It had been obtained from Japan, therfore it is
refered to as the S-l80-J. We sought this tumor because of difficulties
which existed with the 'interim 5-180'. The 'interim S-180' was too
sensitive to cis-Pt(II) and was not a good tumor for drug screening.
It also had a lower electrophoretic mobility and less sensitivity
to RNase, which was the most intersting electrophoretic property
of the original S-180 tumor. The S-lBO-J had an intermediate
sensitivity and allowed me to continue my examination.

Notes on conventions of wording:

- platinum treated - means neoplastic cells removed from animals
which have received a theraputic dose of cis-Pt(II).

- incubated - as in 'platinum incubated' or'RNase incubated'
This means cells that have been incubated in vitro.

- mobility - is used to refer to whole cell electrophoretic
mobility.

52

53
I. Cell Pathology

A. 5-180

Samples of tumor cells removed from animals were placed on
slides and stained with Wrights stain. These were examined at
different stages of tumor growth with and without platinum treatment.

I examined these cells with light microscopytand noted that
the normal tumor cells were all uniform in size (20-40um), contained
large nuclei and little cytoplasm, and consdtuted at least 952 of
the total cell population. The platinum treated cells were '
different. The cells, which I believed to be neoplastic, were often
somewhat smaller or very much larger with many nuclei, they contained
‘mnre vacuoles, and their percentage of the total was considerably
less. A large population of smaller cells were present (lO-lSum).
These cells appeared to be leukocytes, in particular, neutrophils.

A peroxidase stain was used to determine if these cells were
neutrophils. This stain specifically reacts with the peroxidase
enzyme present within the cell and leaves a dark green to black residue
at thelocation of large concentrations of the enzyme. The platinum
treated tumor did contain a very large pOpulation of neutrOphils,
which increased with time after platinum treatment.

This response led me to consider the possibility of a host
interaction after platinum treatment. Neutrophils are involved in
inflammatory reactions but are also capable of tumor cell phagocytosis.
A further examination of the cell types present after platinum
treatment was necessary and I arranged with a pathologist at MSU

to do this for me.

54

His findings, which are presented in the next section, were
on the 'interim 8-180' tumor strain. There were distinct differences
between the S-l80 and the 'interim S-l80'. The S-l80 cells, from the
time of their injection into the animal, contained very few host cells,
were more regular in size and shape, and were less sensitive to
cis-Pt(II) treatment. The 'interim S-l80' seemed more antigenic
because ofthe neutrophils and macrophages that appeared immediately

after injection of the tumor.

B. 'Interim 5-180'

During the course of this tumor growth samples were obtained
and given to Dr. Steve Stockham, a pathologist at MSU. He determined
the cell populations in both the control tumor and platinum treated
tumor as a function of time. Figure 5 is a compilation of the
information he returned.

The control tumor begins with relatively large populations
of macrophagesand neutrophils but their numbers decline with
respect to the tumor cells as time progresses. The platinum treated
tumor, however, shows a large decline in neoplastic cells and an
overwhelming amount of host cells. In the first couple of days
neutrophils predominate but are then followed by macrophages.

Neutrophils have been shown to be cytotoxic to tumor cells,

but only if they outnumber the tumor cells.107’108’109

Here, then
is a possible second line of defense after platinum treatment. The
platinum's first effect is upon the cell itself to slow its growth

and perhaps second to alter its surface making it more susceptible

to host phagocytes.

 

55

 

 

- a.
80 .__./.
/ ..... NEOPLASTICCELLS
d C
5” .... muomcts
x .... NEUTROPIIILS
40- 0
>4
20- In \;'\--I "0.,
.k\ 0......
\‘\ ........................
‘5 —————— ‘ ————— A
1 z i 6 3 lb
30- u.
50‘
x
40‘
20..

 

 

 

Figure 5. Percent of major cell types versus age of tumor in normal

tumor, a, and platinum treated tumor, b.

56
Macrophages are also considered important in the host's
defense against tumors. This large macrophage to tumor cell ratio
in later days after platinum treatment might be the third strike

against the remaining neoplastic cells.

57
II. Electrophoresis

A. Observations with Electrophoresis

l. Mobilities as a function of the tumor age.

One of the first studies was the examination of the electrOphoretic
mobilities of the S-180 tumou as a function of time after injection
of the tumor into the mice. One set of animals received the tumor
on day 0 and the other received the tumor on day 0 and 7mg/kg
cis-Pt(II) on day 1. Figure 6 shows the mobilites up through 9
days of tumor growth. (After the 9th day the control becomes very
large and bloody, while the platinum treated animals contain very
few cells.) There is a striking reduction in mobility in the ureted
cells beginning about 7 hours after treatment. For the first 4 hours
there is no change suggesting that cis-Pt(II) does not lower the
mobility by simply binding to the membrane. As mentioned previously
the treated cell population contains several cell types, where the

neoplastic cells are in the minority. When measuring the mobility,

only the large, neoplastic cells are Visually chosen for measurement.

2. The distribution of mobilities

The apparatus used in this study to measure the mobilities
determines the average for the population in view. In order to determine
if this was a valid representation of their mobilities, individual
cells were timed. Samples of control cells and platinum treated cells
were measured using a stopwatch to measure their mobilities. Figure 7
shows histograms of both samples. Forty cells were measured in each
case and in the platinum treated cells the size of the cell was also
recorded. This resulted in two populations appearing, the small host

cells and the large neoplastic cells. The spread on the neoplastic cells

58

.Howwm owmpgum on» ucmmowmow whom .09... H055 mo coauocom m we
ucofiumom» EscaumHm #593?» one 4.5.3 3de H055 omalm mo ESTES. uauowonmouuuodm can. .0 ouomflm

 

 

 

a a e a m as . a - n N _ a
2.-
.. .. .. 94..
J1 L...
m1 u .. as... a m
m

  

25:8

 

 

59

 

 

 

 

 

 

4.3 4.2 «4.1 -l.0 -0.9
mobility

Figure 7. Distributions of mobility for the S-lBO-J. a) cisplatin
treated tumor, hatched line is host cells, solid is tumor.
b) untreated tumor, tumor cells only. The ordinate is

number of cells.

60
is large but they appear to be unimodal so the averaged measurement

obtained from the electrophoresis apparatus is meaningful.

3. Enzyme incubations.

In Table I the three 5-180 strains, both control and platinum
treated, are listed versus incubation with the enzymes, neuraminidase,
protease, hyaluronidase, DNase, and RNase. In the 5-180 and
'interim 8-180' RNase was the only enzyme incubation which mimicked

the effect of platinum treatment in the control and had no effect

on the platinum treated cell. In the S-l80-J both RNase and DNase
could mimic the effect of platinum treatment on cell mobility.

Both protease and neuraminidase lowered the mobilities of the controls
and treated cells, but each enzyme lowered the mobilities of the

control and treated cells to a common level.

4. Mobilities as a function of tumor age and enzyme treatment - S-lBO-J.
The time course of the mobilities of the S-lBO-J is shown in

Figure 8 . The enzymes RNase and DNase at each day are also shown.

The platinum.treatment occured at day 3 instead of day l which meant

that fewer cures were seen and the tumor resumed normal growth sooner.

As can be seen in the platinum treated animals the RNase and DNase

susceptibility returns with time. This is the first evidence that

the cells are acquiring these groups and that the groups may be

associated with growth rate.

61

.owuoc omwauwnuo mamas: mwowwm pudendum yawn»
new: omumaa on Had: megawawooz .mwowuw pudendum on» who mammcucuwma cm owumfia mooam> one

.AHHvumumwo wawan and: oouemwu coon m>en menu madfiecm scum mwoeou one “am + you ooumua magma one

.uouuoo no Ho. mo Ho>oa woomowmficwwm a be
maouucoo w>wuooamou Hmong scum acouommao haucmuwMacmum one mxuwwmumm no so“: omumaa mozam> one

.cemazwu mm: aromalm cam now one mmmcowa we: omaum mam wow pom: ommououa one .mwnoz

 

 

 

 

 

 

 

 

. Aec.v mo.a Aeo.v mo.a «aszo.v am.o Amoleqm.o Aao.v so.a
Amo.v m~.~ «amo.v ~o.~ «Amc.v mo.~ xaoao.v ~m.o amo.emm.o amao.v ma.~
. : a~mo.v ma.o «aafio.v ea.o . Ammo.v mm.o
. Am~o.v oo.~ «Amc.v om.o .aamo.v on.o u Amao.v mm.o
- Amo.v ec.z Aqno.v mo.H «Awmo.v mm.o «Ammo.v mm.o ammo.v mo.a
Amo.v om.~ Aco.v em.a «Aoao.v eo.~ «Aem.ov em.o xamao.v hm.o Aaao.v ~m.~
.~m>= ommzn ommzm .cMEduamz wmmmuowm Howucoo

 

 

 

um + waowaum
stomaum

ua+.omaum appease.
. OQHIW adhwuflw.

um+ omHIm
omalm

.Hoaou omHIm on» no huwawnos on» :0 mcofluonoocfi ofihwco mo muomwmc one .H wanes

62

 

.mdowucoo cowuonoocw owe
moauuwu one .coHumnoocw mmmzo moamcmfiuu .cowuonooca mmmzm “Coachman mowoovm
.mHHoo ooumowu Eocwueam one dowucou won mom wean» mo cowuucom a mo auflafinoz .m chomflm

N _ a»:

“I'll!“

 

3:32.. 3:23

63

5. Enzyme kinetics

The enzyme RNase is positively charged at physiological pH and
therefore it could bind to the cell membrane and lower the mobility.
The enzyme action was examined in such a way that this possibility
was eliminated. In Figure 9 a,the change in mobility as a function
of time for RNase action on the S-lBO-J is shown. Samples were taken
at increasing times after the cells were exposed to the enzyme.
These samples were then placed on ice until the final sample was
taken, then the enzyme was washed out of all tubes simultaneously.
This proceedure rules out electrostatic binding because this kind
of binding occurs at ice temperatures as well or even better than
at 37° C, whereas ezymatic action occurs only at higher temperatures.
The curve clearly shows that RNase action is stopped by ice
temperatures in the early samples.

Similar kinetics for DNase and RNase+DNase are shown in Figure
9 b,c. The RNase, DNase, and RNase+DNase kinetics are first order
curves with rate constants of 0.065, 0.074, and 0.161 min"1
respectively. The noteworthy element of these rate constants is the
additivity for the two enzymes. The rate constant for RNase+DNase
is almost exactly the sum of each of their separate rate constants.

The level to which the mobility drops is about equal for both
RNase and DNase and RNase+DNase. From this and the additivity of
their rates it is probable that they are both acting on the same

site.

64

3

an

   
  

.Howucou cowuonooca ecu ma
owoofim some ca o>uoo mos .hlomanm can no muedanoa com: deduce mahwcm mo moaumcax .m gunmen

«use:
8 _= e 3 an as S a 3 an em =—

 

p

323 + 3:...

O——.

 

h—C—d
l

 

 

3...

—

O

—
O

MIMI

 

65

B. Further Experiments to Test for Nucleic Acids

1. pH profiles of the S-180 and S-lBO-J.

When the cells are measured in buffers of varying pH a profile
is obtained which usually shows an isoelectric point where the cells
have no net charge. This point is approximated by interpolating
between negative and positive mobilities. In Figure 10.the pH
profiles for control and platinum treated 5-180 cells are shown. The
isoelectric' points are 3.4 and 3.9 respectively. This shift to
higher pH in the platinum treated cells could be due to the loss of
a group with a pK lower than 3.4. A rough calculuion can be done
to show the approximate pK of this group.

At pH 7.0, wheee the profiles are relatively flat, the
platinum treated cells have about 202 lower mobility. Therefore,
roughly 802 of the charge of the control has an average pK of 3.9 .
(that of the treated cells). The remaining 20% of the charge must have
a pK low enough such that when it is averaged with the remaining
802 a pK of 3.4 is obtained. Such a pK of the missing group computes
out to be about 1.4 .

The calculation assumes that there is a group missing from the
platinum treated cell surfaces rather than groups added, such as
positively charge molecules. (This is a reasonable assumption - see
Appendix II) This also assumes that the pK's of different groups
average into‘one cellular pK in a linearly additive fashion.

A pK of 1-2 for a surface moiety does not allow too many
possibilities. The two groups that come to mind are phosphates and

sulfates. Sulfates are not commonly known in biological molecules

66

+0.5
0
0
l 3 5 7 pll
-0.5
0
Pt
*1.“ \o
c
\ .
-l.5

Figure 10. Plot of mobility versus pH of the measuring buffer for
control(C) and platinum treated (Pt) S-180.

67

but phosphates are. Phosphates are located in membrane phospholipids
and are part of the backbone of polynucleic acids. The pK of
ribonucleic acid is around 1.0 and pK of the phospholipid
phosphates is between 1.0 and 2.0.

The major phospholipids of mammalian cell membranes are
phosphotidyl ethanolamine, phosphotidyl choline, and sphingomyelin.
All of these are zwitterions with no net charge and with the
positive group the most exterior group. The lack of an exposed
negative charge would eliminate the possibility that the loss of
one of one of these groups was responsible for the effect seen
in Figure 10.

The possibility of surface ribonucleic acids existing on
certain cell types was shown by other authors( see section
IV of the Literature Review), so an examination of the 5-180
using various enzymes including RNase was undertaken.

In Figure 11, the pH profile for the S-180-J tumor is give.
The curve of solid circles is the untreated control, the open
circles are the platinum treated cells, and the squares and
triangles represent cells incubated with RNase or DNase
respectively. Only the control is significantly different from
the other five curves. This and other evidence which will be
presented indicates a similarity between platinum treatment and

incubation with nucleases.

68

\Wfla

-l.5

Figure 11. Plot of mobility versus pH for the S-180-J. Solid data
points are control cells, hollow points are Pt treated.
Squares are RNase incubated, triangles are DNase incubated
and circles are incubation controls.

69
2. Ionic strength profiles

The Literature Review section on electrophoresis and Appendix I
discuss the rationale for measuring mobilities over varying ionic
strengths. Basically, it allows groups deeper within the membrane's
three dimensional structure to exert their potential upon the
zeta potential of the slip plums In the Appendix, I discuss the
ionic strenth profiles of the S-180 control and RNase treated cells
and I conclude that it is reasonable to suggest that the RNase
susceptible group is localized on the periphery of the cell surface.

In Figures 12 and 13 the ionic strength profiles of 5-180 and
S-lBO-J are given for.control and platinum treatment with RNase,
protease and neuraminidase incubation of S-l8O and RNase, DNase,
and neuraminidase incubation of S-l80-J.

Most notably in both cell strains there is a smaller percent
drop in mobility at lower ionic strengths and concommitantly an
increase in slope of the curves for the nuclease treated cells.
Both of these can be argued to indicate the peripheral nanme of the
nuclease susceptible group or its possible lability.

Again it is seen in this experiment that neuriminidase and
protease lowers the control and the platinum treated cells to a common
level. So these enzymes probably take the proposed nucleic acid off

the control cells during their normal action on the membrane.

7O

.>Hm>..nuoommow cowumnooce ommpficdnmwooc one .coaumnooce unmouowm .noflmnmocfl
ommzm acomowmow .z.m.m .HoEou pounce» Ecwumam 3 £360 Howunou Am 6.2.65 mcewommmfi
on”. «0 numcmuum Dance 0:» mo uoou mumoom o5 momuo> languages oauwuonmouuooam m0 m3 .2. ouzmflm

2.3 .3;

+5”.

 

a. e elm» 1|;

--=I
_.

ice—l

4.4/40"
/
h0fi

' I

9.

’T

L
L

/

a we-

. a m/w -
/w

c " z
/H7I—o—¢ i——o——I
s—ol—od—n l—0-4
i l I I I l l
“2
'T
#04 P—.
///
I I I

 

 

71

 

.>Ho>wuoommou mumoacflamuooc one .ommzn
.ommzm sua3 chHumnoocw ucomowmmw 2.0.x .nlomHIm me was» umooxo NH shaman ca ufl .ma whomfim

.239 5.7

1......» an...»

72

3. Incubation with Sepharose bead immobilized DNase and RNase

One of the arguments against using free DNase and RNase enzymes
is the possibility that they can enter the cell. This could result
in effects other than those directed upon the surface. In Table II
the results of an incubation with Sepharose bound DNase and Agarose
bound RNase are shown. After the incubation of the cells with the
bound enzymes Ficoll is added to a final concentration of 152. When
the mixture of cells and beads are centrifuged in this solution,
only the beads sediment. The cells in the supernatant are then washed
free of the Ficoll. Mobilities are shown for Ficoll wash alone and
plain Sepharose beads with Ficoll wash to show that there is no
major effect due to the incubation proceedure.

The effect of bound enzyme action is evident. Following with

regular unbound enzymes shows no further changes.

Table II., Incubation of cells with immobilized RNase and DNase.

9.93521 .3329. less.
3-180-J -1.143 (.011) -l.048(.028)x -l.046 (.013)“
" +Ficoll wash -l.l4 (.01) ND ND
" +Sepharose beads -l.17 (.03) ND ND
and Ficoll
" +Seph-DNase -1.07 (.02)c ND ‘1-05 ('022)¢
" +Agar-RNase -l.08 (.023)C -1.06 (.032)c ND

Notes: Those values with the c are significantly different from the
S-lBO-J control without any enzyme treatment.

Cells are incubated with the bound enzyme for 45 minutes
with a Seph-DNaee concentration of approx. 400 units/ml
(see Wothington catalog) and a Agar-RNase concentration of
approx. 6 Units/m1 (see Miles Lab catalog)

ND - not determined

* - significance level of .01

73

4. 'Platinum-thymine-blue' attachment to the cell membrane.

'Platinum-pyrimidine-blues' bind to nucleic acids, probably
due to their polycation properties and their pyrimidine base
composition. 'platinum-thymine-blue' was incubated with S-l80-J
cells at a concentration of 10ug/m1 for 30 minutes at 37°C. Table
III’ shows the effect of 'platinum-blues' on the mobilities of
control and platinum treated cells as well as these cells incubated
first with RNase or DNase and then with 'platinum-blues'. Cells
treated with cis-Pt(II) or any cells incubated with nucleases
showed virtually no binding effects of 'platinum-thymine-blue:
This reinforces the evidence that there are surface nucleic acids

whibh are removed by cis-Pt(II) as well as nucleases.

5. Incubation with anti-DNA antisera.

Dr. Edward Golub generously gave me rabbit anti-DNA antisera
which was made by injecting calf thymus DNA plus bovine serum albumin
into rabbits. The titre of this anti-sera is low due to the low
antigenicity of nucleic acids.

When antibodies bind to specific cell surface components the
mobility of the cells is reduced. This is probably due to the covering
of some charges with a relatively non charged molecule. (See Appendix IV)

S-180-J control and platinum treated cells were incubated with
anti-DNA antisera and the results are shown in Table IV . This
basically shows that there is a small reduction in control cells at
the dilutions shown but not in the platinum treated cells. Incubation

with normal rabbit sera at the lowest dilution also had no effect.

74

Table III. The binding of 'platinum-thyminrblues' to the S-180-J.

Enzyme Incubation

 

 

Control RNase DNase
S-180-J Control -l.21 (.02) -l.O8 (.04) -1.08 (.015)
'Pt-thymine-blue'zadded -1.11 (.02) -1.06 (.02) -1.05 (.02)
S-l80-J Cis-Pt(II) treated -l.08 (.03) -l.07 (.02) -1.08 (.02)
'Pt-Thymine-blue added -l.06 (.04) -l.11 (.01) -1.04 (.01)

Notes: Pt-thymine-blue was incubated with the cells for 10 minutes at 37°C
at a concentration of lOug/ml

Parenthetic values are one standard deviation

Table IV. The incubation of cells with anti-DNA antisera.

 

Control Dilutions of Anti-DNA antisera
1:10 1:5 1:27 u
SelSE-J -1.23 (.003) -l.18 (.010)‘ -1.17 (.003)" -l.l6 (.04)
Pt treated -1011 (.03) - -1008 (.02) -

S-180-J

Notes: When treated with a 1:2 dilution of normal rabbit sera the
S-l80-J's mobility was -1.22 (.014)

Anti-sera was incubated at 23°C for 30 minutes, then the cells
were washed once by centrifugation.

* - significance level of .01

75

6. Incubation with the restriction enzyme Eco-R1 and Bam I.

In the S-l80-J the sensitivity to RNase is accompanied by a
sensitivity to DNase. The possibility of surface DNA is further
examined here by use of the specific endonuclease restriction enzyme
Eco-R1. This enzyme recognizes the following double strand sequence:

G/AATT c
c TTAA/G

and cleaves the double stranded DNA at the slash marks. A
demonstrable effect of this enzyme on the mobility of S-lBO-J cells
would stranghten the implication of surface DNA and would suggest
it is double stranded.

In Table ‘I.Eco-R1 is shown to be active in control cells, but
not in platinum treated cells and if EDTA is added to remove the
activating ion, Mg”, the effect disappears.

Also shown in Table V is the effect of Nuclease S-l which is
specific for single stranded RNA or DNA. At very large concentrations
of the enzyme no effect is seen. This supports the results of Eco-R1
in indicating that the surface nucleic acid is double stranded.

In Table VI, the effect of restriction enzyme Bam I is shown.

An effect is noticable at lower enzyme concentrations than seen

in Eco-R1. Both of these enzymes are used at extremely high concen-

trations, however, and probably do damage the cells' nuclei.

76

Table V. The effect of restriction endonuclease Eco-R1 and nuclease
$1 on cell mobility.

Eco-R1 Activityfi

 

 

 

 

Control 20,000units 10,000units 5,000units

S-180-J * *

+10mM Mgso4 -1.20 (.03) -1.03 (.03) -l.08 (.01) -l.16 (.03)

+10mM EDTA -1023 (002) -1021 (003) -1020 (.01) -1021 (004)
S-180-J Pt

treated day l

+lOmM MgSOa -l.15 (.02) -l.lS (.02) -l.18 (.02) -l.l9 (.01)

+10mM EDTA -1017 (001) -1017 (002) -1015 (002) -1014 (.03)
S-l80-J Pt

treated day 2

+10mM MgSO4 -l.05 (.02) -1.03 (.02) -l.05 (.02) -l.09 (.03)

Notes: the platinum treated cells on day 1 show a curious phenomena
which I see occasionally. The overall mobility does not drop
but the sensitivity to enzymes disappears. By day two the
full effect of the platinum seems to have taken effect.

Parenthetic values are one standard deviation.

Nuclease 81 Activity
Control 330,000units/ml 125,000 units/m1

S‘lSO'J ‘1022 (002) -1023 (.02) -1024 (.01)

77

Table VI. Incubation of cells with restriction endonuclease Bam I.

 

units units units
Control 2000 ml 500 T 100 m1
Untreated
+Mg -1.18(.02) -l.O(.Ol) -l.09(.01) -l.18(.02)
+EDTA -1.19(.O3) -l.12(.03) -l.21(.02) -l.23(.04)
Pt treated
+Mg -l.O7(.02) -l.09(.04) -l.05(.02) -1.05(.02)
+EDTA -l.054(.04) -1.09(.02) -1.07(.Ol) -l.06(.03)

7. A spectrophotometric study.
I examined the possibility of detecting nucleic acids
as they were being removed from the cells in the presence of the
nucleases. The procedure was to incubate the cells with enzymes
and remove the supernatant, filter it, and examine the ultraviolet
absorption spectra for the characteristic 260nm absorption peak for
nucleic acids.
In Figure l4,curve (d) is the difference in the absorbence of
the supernatants of cells incubated with RNase and without at 37°C
for 30 minutes. 2.2 x 108 cells were incubated in each sample. The peak
= 1.88, A

absorbence is 0.62 and A = 31.3. Using a

260/280 260/230

standard curve of RNA absorption versus concentration, the absorbance

of 0.62 represents 1.7 mM nucleic acids in solution.

78

This particular incubation proceedure has a major problem:
that RNase is only in one tube. If RNase releases material from
inside the cell, that would cause an inflated reading. Therefore,
the same cells were washed and both, treated with RNgse. In this
case the cells previously treated with RNase should have no surface
nucleic acids and the previous controls should. So the latter
should release more than the former, while the backround should be
about constant.

In FigurelA, curve(a),(b),(c) these results are shown. Curve (a)
represents the previous controls, curve (b) those previously treated
with RNase, and curve (c) the difference. In this case the difference
is very small and one might think not even significant, but a
quick caloflation will show that it could be about one or two orders
of magnitude too big.

From the electrophoretic mobility the number of surface charges
influencing the zeta potential can be calculated.72 For these cells
that is about 108-109 charges. If we:are generous and suggest that
10 times that many are undetected by electrophoresis then that

10 charges per cell surface. RNase removes approximately

implies 10
102 of these or 109 charges. In this study 2 x 108 cells were used
which implies 2 x 1017 nucleic acid 'charges' should be released
specifically from the membrane. That is only 0.001 mM, one hundred
times less than observed spectrophotmetrically.

This experiment points out two things. The first is that in
normal incubation many nucleic acid molecules are released either

during regular metabolism ( the absorbance for cells incubated without

RNase was 0.29 at 260nm.) or from dead cells. The second is that

79

.mawouoo How axon com
.mcowuwpcoo moowum> Hops: venomous“ anomaumv maaoo scum unnumcwomom mo muuummm >3 .vH ouomah

5-. can can :3 K EN ea :2

 

~._

8O

measuring released nucleic acids during membrane enzyme incubation
is not sensitive enough to detect the changes.

If the 0.11 mM nucleic acid levels detected here are real it
would suggest very large complexes of nucleic acids attached to
the membrane and probably sticking out into the extracellular space.
These large complexes would have many interior nucleic acids that

would not contribute to the surface charge and would go undetected.

81

C. Binding and Origin

1. Cellular debris

One of the possible origins for surface nucleic acids is
debris from.broken cells adhering to intact cells. The normal
handling of cells during the experiment argues against this. Cell
breakage most likely occurs during centrifugation and when the cells
are stored on ice. Yet after enzyme treatment and after they have
been centrifuged and vortexed, the mobility remains depressed. There
seems to be no effect due to the ever present debris.

To strengthen this argument, however, cells were exposed to
homogenized S-lBO-J tumor cells. A hand homogenizer was used to break
up about 107 cells. Normal cells and enzyme treated cells were
incubated with this debris for 45 minutes at ice temperatures
(the temperature that the cells were normally stored at in typical
experiments). This temperature would also facilitate electrostatic
binding. They were then washed normally and their mobilites measured.
In Table VII, the results are shown. The debris did. not bind to
enzyme treated cells and in no case did it raise the mobility,

which would result if nucleic acids bound to the surface.

Table VII. The effect of cell debris incubation on S-180-J mobility.

control RNase DNase
S-180-J -l.23 (.01) -l.10 (.02) -l.O7 (.04)
" + incubation with i 4’ '4
cell debris -l.14 (.02) -1.08 (.032 -l.06 (.03)

Note: The arrows indicate that after the cells were treated with
RNase and DNase they were then exposed to the cell debris.
and remeasured.

82

2. Lability of nucleic acids with trypsin, neuraminidase and low
ionic strength.

In Table VIII a,b the 5-180-J cells are incubated with
neuraminidase and trypsin then each are followed with incubations
with RNase or DNase. The action of either trypsin or neuraminidase
lowers the mobility to a specifed level for each enzyme, but
incubation with either nuclease does not change this. This seems
to indicate that nuclease susceptible groups are loosely bound
and depend upon sialic acid residues and/or protein molecules
to remain attached to the cell.

If the ionic strength of the cell bathing solution is lowered
the electrostatic forces between surface molecules change.
Attractive forces increase and repelling forces increase. Of course,
in low ionic strength buffers transmembrane ionic fluxes will be
drastically altered too. However, since the membrane is negatively
charged and DNA and RNA are negatively charged perhaps lowered ionic

strength can cause the nucleic acids to come off.

83

Table VIII. Multiple enzyme incubations.

Table VIII a

 

 

 

 

. . . Hanks Hanks + Trypsin
5 min. incubation -EI§Z'1.02) _1.07 (.03)
t J' ‘N
Followed by Hanks Hanks+RNase Hanks+DNase
20 min. inCUbation ’1024 (001) -1006 (.01) -1008 (.03)

Table VIII b

 

 

 

 

55 min incubation Hanks Hanks + Neuraminidase
-1027 (002) -0092 (002)
l l \.
Foiéozig Egcubation Hanks Hanks+RNase Hanks+DNase
’1027 (.02) -0091 (.02) -0093 (001)

Notes: Trypsin concentration used was 0.005mg/ml
Parenthetic values are standard deviations.

Neuraminidase concentration was 0.5mg/ml

84

In Table Ix a,b two experiments are shown. Cells were incubated
at 3 different temperatures in part (a), and it appears that 23°C is
the most detrimental to the cells' mobility both for the 100% saline
control and the 1s saline-99% isotonic mannitol test. At all
temperatures the lower ionic strengths lowered mobility, but the
least effected were cells at 37°C. This is possibly due to cell
metabolism reducing membrane loses. In part (b), a 23°C experiment
is shown. Here it is shown that the effect of low ionic strength
saline on the cells occurs over time. Following these incubations with
RNase results in bringing them all to about one common level. So
the most labile group to low ionic strength is the nucleic acid
moiety.

The fact that trypsin, neuraminidase, and low ionic strength
can remove this nucleic acid group leads me to suggest that this
group must be weakly bound to the membrane. One explanation for this
could be that it is in a process of flux through the membrane and it

only exists on the surface for a short time.

3. In vitro studies

One of the early questions involving the effect of platinum
treatment upon the cells was whether the platinum acted directly
upon the cells to cause the mobility drop or through some host
response. In order to test this, in vitro experiments were performed
where the cells were removed from an animal and incubated with

platinum or other compounds in Hanks buffer and then placed into

85

Table TX. The effect of ionic strength and temperature on the
mobility of the S-180-J.

30 min. incubation
S-lBO-J saline

" in 12 saline
992 mannitol

 

23°incubation
5 minutes
S-180-J in saline

" in 12 sal
992 man

 

15 minutes
S-180-J saline

" lZsal/992man

30 minutes
S-180-J saline

" IZsal/992man

 

 

 

“1008 (.02)

Tab 1e D! b

 

02212.1.
-1020 (002)

-1013 (002)

'1018 ( 003)
-1.09 (.02)

-1.15 (.01)
*1.04 (.02)

 

Table Ix a
Temperature
2°C 23°C 3_7_°_C_
-1021 (001) '1016 (001) “1.21 (004)

“1003 (.02) -1013 (002)

RNase
-1004 (004)

-1003 (001)

-1.05 (.01)
-1.02 (.03)

-100 (.03)
”1001 (001)

86

tissue culture media for up to 22 hours. The culture medium was
changed at regular intervals of 1 to 2 hours. Samples were removed using
sterile technique and the mobility was measured.

The curves in Figure 15 show the way the mobility changes with
time. The control cells show a continual decline and alchatacteristic
dip between 4 and 8 hours. The effect of cis-Pt(II) is to increase
the rate of decline to an even lower value with only a slight dip.
The RNase treatment of controls (open circles ) is a curve similar
to that of cis-Pt(II) beginning at about 4 hours. Therefore, the
effect of platinum in vitro is to cause the loss of this surface
nucleic acid.

An RNA inhibitor, Actinomycin-D is also shown (dotted line)
in this figure. It too causes the reduction to a level similar to
RNase and platinum. Its effect is slower, but this may be dose
dependent.

Other inhibitors were also tried to further substantiate the
nucleic acid hypothesis and its possible origin inside the cell.

In Figure 16 c the effects of mercaptopurine, methotrexate are
shown. They produce an effect similar to platinum by lowering the
mobility to that produced by RNase. Both of these drugs are DNA
synthesis inhibitors and are used in antitumor therapy. Puromycin, a
protein synthesis inhibitor, is shown in Figure 16 d. It lowers the
mobility much further than the other drugs but in the process also
removes the RNase susceptible group. At the five hour mark the bar
under the triangle indicates the mobility of that five hour sample
after RNase treatment-there is no change. It was shown,before,that the

loss of protein causes the loss of the nuclease susceptible groups.

87

Figure 15. Plot of mobility versus time in in vitro incubations.
Cells were removed from animals pre-incubated with
cisplatin, (Pt), or with Actinomycin-D, (Act-D), for
forty minutes before time zero. Time shown is time in
normal tissue culture media. R represents RNase incubation
of samples taken during the course of the in vitro
incubations.

 

88

.mu unease

 

 

89

Figure 16. In vitro experiments as described in Figure 15.
The control curve is repeated in each figure as
reference. a) Normal S-l80-J were preincubated
with RNase and then allowed to grow in vitro.

b) similar to (a) except cells were preincubated

with DNase. c) Squares: cells were preincubated

with mercaptopurine 0.5mg/m1. Crosses: cells were
incubated with methotrexate 0.5mg/ml.

d) The RNase curve of Figure 15 is repeated.

Triangles: cells were preincubated with lmg/ml
puromycin.

 

 

 

9O

 

 

 

 

 

 

a. b.
4.2- 0 ‘
\.c \J:
\ \
[it 0‘. {A}. f‘o
’1'] ' I.i . .... 'o/ d 0"” \‘ ‘ . /
". " O/‘/ A‘ If. ./
J °d \x
-l.0' “‘
o 5 i 0 5 10 o 4 8
hours
c. d.
-l.2- - o
\o
\.
\
‘10] 'I 'l .\ ./
A 0"./
o, _ _\_ - _ 0‘ ’ ’ fl
° ~‘-°' ’ ’ rum
4.0" = '
Re
a I i I é I \A

Figure 16.

 

 

91

In this case the main point is that puromycin causes more membrane
changes than the loss of nucleic acids.

In general, the action of these drugs supports a nucleic acid
origin for these nuclease susceptible groups and a link to cell
metabolism.

This is further supported by Figures 16 a,b. In these experiments
cells were first treated with RNase or DNase and then placed in
tissue culture media. Within 4hours the mobility increases and
the curves join the curve of the control cells. It appears that
this group is renewed on the surface and the cell may be continually

producing this material and exuding it into the media.

92

D. Uther Experiments

The results shown up to this point allow me to suggest that
there are nucleic acids on the cell surface, that they are probably
loosely bound, and their presence is linked to the metabolism
of the cell. There is no evidence which could be used to indicate
a function or purpose to these nucleic acids. Whereas my goal was
not to determine their function I did attempt a few 'quick and dirty'

experiments which I will include here for completeness.

1. Binding of antitumor sera to 'interim 5-180'.

Mice were given normal injections of tumor and platinum and
7 days later they were bled and their serum expressed and pooled.
Two types of sera were obtained,( T-sera) from. animals with tumor
only and (Pt-sera) from animal with platinum treated tumor. Each
sera was then incubated with control cells, platinum treated cells
and the RNase incubation of each of these.

In Table )( results are shown over varying concentrations of
sera for the control and control + RNase, and at l to 5 dilutions
for the platinum treated cells. The results are shown as differences
from those cells incubated without sera.

Basically, the results show nothing. The only interesting
thing is the difference between the RNase incubated controls
treated with T-sera versus Pt-sera. It is interesting because of the
hypothesis that the nucleic acids may exist as antigen masks. If
this were true there should be antibody in platinum treated animals
to somemdng under the RNA since these cells contain no surface RNA

after platinum therapy. There is the slightest hint of extra binding

93

Table x. Incubation of the S-180-J with sera from tumor bearing

animals.

S-180-J
+ T-sera

+Pt-sera
S-l80-J l RNase incubated
+ T-sera

+ Pt sera

S-l80-J Pt treated
+ T-sera
+ Pt-sera

S-180-J Pt treated
[RNase incubated
+T-sera

+Pt-sera

1:5

0.056 (.01)
0.046 (.008)

0.026 (.014)
0.056 (.01)

0.075 (.01)
0.085 (.02)

0.08 (.02)
0.09 (.03)

Serum Dilutions

1:10 1:20
0.06 (.01) 0.04 (.01)
0.04 (.01) 0.03 (.01)
0.00 (.04) 0.00 (.01)

0.055 (.005) 0.025 (.005)

94

to the RNase treated cells by something in the Pt-sera. However,

why is there a similar amount of binding of T-sera and Pt-sera to
control cells? From a different point of view, in the comparison of
T-sera binding to control cells and RNase + control cells there is
less binding of T-sera to RNase incubated cells. This would imply
that there is antibody in T-sera to this RNase susceptible group.

The data is too weak to dwell on it any longer.

2.Animals injected with nuclease treated cells.

In persuing the hypothesis of antigen masking animals were
injected with S-180-J tumor cells treated with RNase and DNase. The
animals were then followed until day of death and any increased life
span over controls was noted. The results of three experiments are
shown in Table XI a,b,c. In Table XIa,S-l80-J cells were preincubated
with RNase and DNAse at lmg/ml for one hour at 37°C. Their viability,
as measured by trypan blue,was 802 for control, 762 for DNase, and
682 for RN ase. Cell concentrations were then adjusted to 106 viable
cells per sample and unjected into 10 mice each. Average day of death
is shown in the table and suggest that the RNase treated cells were
more susceptible to a host response.

The low viabilities suggested that the incubations were too
harsh on the cells so another test was run. This time the cells were
incubated with the standard enzyme concentrations: 0.2 mg/ml RNase,
0.1 mg/ml DNase, for 20 minutes at 37°C. According to the kinetics of
Figure 13 this is enough time to just remove the surface groups.
Viabilities were all greater than 902 and as is shown in Table XIb,there

is no significant increase in day of death over the control group.

95

Table XI. Day of death of mice injected with s-180-J incubated
in RNase and DNase

S-180-J incubated 1 hr.
Hanks

" +Rnase lmg/ml

"+DNase 1mg/ml

S-180-J incubated 20 min.
Hanks

" +Rnase 0.2mg/ml

" +DNase 0.1mg/ml

S-180-J incubated 35 min.
Hanks

" +Sepharose Dnase

Hanks/Ficoll wash

Table XI a

 

Table .X1 b

Average Day of death Survivors
excluding survivors

19.6 (1.84)

15.3 (3.9)

20.4 (4.43)

13.8 (2.35)

11.25 (3.1)

12.4 (2.9)

Table X} C

14.0 (2.3) 0
16.3 (3.4)(17‘ILS) 0
12'8 (1'92221211s) 0
15.5 (2.8) 0

" + Seph-DNase/ Ficoll wash

Notes: Parenthetic values are one standard deviation
ILS - increased life span

96

A third experiment was run using Sepharose bound DNase. Cells were
incubated for 35 minutes in DNase-Sepharose and injected into the
animals. In Table XIc the results are shown. In case I the Sepharose
beads with the DNase were not washed out but injected with the cells.
In case 11, 152 Ficoll was used to remove the beads as was described
earlier. In both cases a slight increased life span (ILS) is noted
but its significance is questionable.

These results would seem to indicate that if there is an
antigen masking phenomena it can not be determined by this type
of experiment. There is even question if the mice can mount an immune
response in 7-12 days and if they could it does not preclude the

cells from making more nucleic acids to mask their antigens.

3. Tumor supernatant and antibody production.

Another possible function of the surface nucleic acids is
that of immune suppression. As others have suggested nucleic acids
may be exuded by the cell into the extracellular space. This could
transmit a suppressive message to the lymphocyte population .

To test this, sheep red blood cells were injected into mice
intraperitoneally. One set of animals received RBC's alone, another
.RBC's plus the supernatant of S-180-J cells incubated in vitro,
and a third RBC's plus the supernatant treated with RNase. These
animals were older and therfore immunocompetent. After two sets of
injections 7 days apart the animals were bled and the sera expressed.
The mobility of sheep RBC's was then determined after incubation
with different dilutions of these sera.

In Figure 17 the results are shown. After each incubation

97

.ummzm no“: penthouse uncumcwomoo moss» + n.0mm mecca Amy one .ucmum:
summon biomaum + n.0mm moonm Amy .haco m.Umm moonm ADV and: beacons“ cues
sown noxmu show no scauocom m we maamo pecan pow moose no auflawnoz .na cuomam

.38 s\_ ex ex .\_ .\_

 

=4:

3..

98

another incubation occured using anti-mouse IgG antibody. This would
bind to any bound antibody on the sheep RBC's and increase the
reduction in the mobility.

The curve labeled C shows the amount of antibody against the
sheep RBC's. When the blood cells were injected with supernatant
the results of antibody production are on curve S. When the -super-
natant was treated with RNase, curve S-R, there is no change from
curve 8. This does show some reduction in antibody titre, about
1/2, due to the presence of the tumor supernatant, but RNase does
not effect it.

This does not rule out a nucleic acid containing suppressor
substance, however, because a nucleoprotein complex could easily

be insensitive to RNase as shown by Stroun et a1 97.

99

E. Other Cell Types
Up to this point all of the experiments were performed with
one of the 5-180 ascitic tumor strains. In this section I will

describe some work with other tumors and some normal cells.

1. S-180 and S-180-J solid tumors.

One very interesting question is raised by using the ascitic
form of the S-l80: Does the surface nucleic acid require single
cell suspension?

The S-180 tumor also grows very well in the solid form if
it is injected subcutaneously. Tumors were, therfore, grown in the
subaxillary region of the ICR mice. From one to two weeks after
injection they were removed as described in materials and methods
using a sodium citrate - EDTA - saline bathing solution and separated
into cell suspensions and then treated with enzymes.

In Table XI the results are shown. Both the 'interim S-180'
and the S-180-J show the same sensitivities as their respective
ascitic forms, ie. the'interim S-l80' is senélive to RNase while
the S-180-J is sensitive to both RNase and DNase. Also included in
this table is an S-180-J ascitic tumor processed with the sodium
citrate-EDTA-saline buffer in a similar manner as the solid to assure'

the buffer did not effect the cells.

2. P388 and P815 ascitic tumors.

The P388 is a lymphocytic leukemia and the P815 is a mastocytoma
both of which are syngeneic to the DEA/2 mouse strain. They were
grown intraperitoneally in the DBA/Z mice. Both of these were tested

for RNase and DNase sensitivities and the results are shown in Table XIII

100

Table XII. The effect of RNase and DNase on the mobility of
'interim 5-180' and S-lSO-j in the solid form.

Control RNase DNase
Solid 'interim S-180' -1.00 (.02) -o.95 (.01)¢ -0.99 (.005)
'A‘ 1':
$6114 s-1aoe3 -1.16 (.013) -1.03 (.02) -1.02 (.003)

S-180-J ascitic * *
Na. Citrate-mTA-Salo -1020 (.005) ’1008 (.05) -1003 (002)

Note: ¢ denotes significance at the 0.05 leve1.in all tables on this
page 0

Table XIII. Nuclease effect on the mobility of various cell types.

Control RNase DNase
P388 -l.048 (.01) —1.045 (.010) -0.978 (.012)‘
t *

P815 -l.113 (.006) -1.05 (.008) -1.032 (.007)
Fetal cells -l.l41 (.008) -1.116 (.004) -l.124 (.010)
Liver cells -1.09 (.02) -l.09 (.02) -1.06 (.04)
AKR preleukemic

spleen -l.10 (.02) -1.07 (.01) -l.08 (.04)

“MS -1011 (.03) '1011 (.02) ’1008 (.05)
AKR leukemic

Spleen -l.07 (.04) -1.05 (.04) -1.03 (.03)

Thymus -1.05 (.03) -1.05 (.06) -0.98 (.04)°
V-79 tissue culture -l.25 (.05) -1.24 (.05) -1.27 (.03)

C - RNase C - DNase

L1210 -.0036 (.006) 0.037 (.006)“
L1210 virus infected 0.0233 (.005)¢ 0.051 (.010)c

L1210 resistant to Pt -0.021 (.002)‘ -0.024 (.006)¢

101

Both were sensitive to DNase, but only the P815 was sensitive to

RNase.

3. Liver and fetal cells.

These cells, like the solid tumors were removed and broken
into single cell suspensions as described earlier. The fetal cells
were from pregnant mice 10-14 days into gestation. The embryos .were
removed from the placenta andwere dissociated into individual cells.
Some of the livers were from these same pregnant mice and were
noticeably larger than normal livers. The results are in Table XII.

There wasa lack of effect of ' RNase and DNase on these 'normal' cells.

4. AKR spontaneous leukemia
AKR mice were obtained from Dr. Golub, Purdue University. After

a certain age these mice spontaneously form a lymphoid leukemia.
Both preleukemic and leukemic animals were examined. Their spleens
and thymuses were removed, dissociated into individual cells, and
measured to determine their mobility. Again, DNase and RNase were
used to determine if there was any sensitivity. This was particularly
interesting because Dr. Golub had shown that a population of
suppressor cells in the leukemic mice had DNA on their surface.

In Table XIIIthe results support his finding by showing a DNase

sensitivity in the leukemic animals.

5. Tissue culture cells V79 and L1210.
Two tissue culture cells were examined, the non-tumorigenic
V79 and the leukemic L1210. Three different types of L1210 were

examined, two of them are grown in 0.3 ug/ml and 1.0 ug/ml ofcusplatin

102

respectively. The 0.3 ug/ml L1210 is a strain resistant to cisplatin.
The 1.0ug/m1 strain is a resistant strain infected with a virus.

The results, as given in Table XIII, show differences for
these three strains. The normal L1210 has a DNase sensitivity, the
virus infected cells have a slight RNase sensitivity as well as
DNase and the resistant strain shows a significant increase in
mobility when incubated with these enzymes. This increase is a
result which I cannot interpret.

The V79 cells show no sensitivity to either enzyme.(See TableXIII)

6. Summary of cell types

In Table XIV all the cells examined are listed. A plus
sign is placed by a cell type if it responded to the enzyme and
the difference using the t-test was significant at the 0.01 level

unless otherwise noted.

103

Table XIV. Summary of nuclease effects on several cell types.

S-180

'inflérim S-180'
‘interim S-180' solid
S-180-J

S-180-J solid

P815 ascitic
P388 ascitic

AKR leukemic
spleen

thymus

AKR preleukemic
spleen

thymus

Fetal cells

Liver

V-79 tissue culture
L1210 tissue culture
L1210 virus infected
L1210 Pt resistant

Notes: c indicates significant at the .05 level.

RNase

+ + + +

+¢

+

DNase

+ indicates a significant effect of the enzymes at a level

of .01

- indicates no significant effect of the enzyme treatment.

104

F. Summary of Conclusions

1.Treatment of tumor bearing animals with.CiSPlatin results in

a sustained lowering of the tumor cell electrophoretic mobility.

2. The pH profiles of control and cisplatin treated cells suggests
the loss of a charged group with a low pK value from the platinum
treated cells.

3. Incubating the S-180 with RNase or the S-180-J with RNase or DNase
lowers the control mobility to the level of platinum treated cells
but does not effect the nobility of the platinum treated cells.

4. RNase or DNase separately or together lower the mobility of the
S-180-J to the same level. Therefore they probably act upon the same
site. Enzyme kinetics support this by showing the additivity of
reaction rates.

5. The enzyme kinetics also represent a tne enzymatic reaction and
not simple adsorption.

6. Preincubation of cells with either neuraminidase or trypsin or
low ionic strength removes the nuclease susceptible groups.

7. From ionic strength profiles the nuclease susceptible groups
appear to be located at the very periphery of the cells. The pre-
incubation with neuraminidase and trypsin supports this.

8. Incubation with cell debris, however, does not result in
increased mobility after nuclease treatment.

9.The ability of 'platinum-thymine-blue' to bind to control cells
but not to cisplatin or enzyme treated cells further supports the
hypothesis of cell surface nucleic acids.

10. anti-DNA antisera binding to the S-180-J again implies cell surface

nucleic acids.

105

11. The restriction enzyme,Eco-R1,acts in a similar way as DNase,
therfore offering some further' evidence for surface nucleic acids.
12. Sepharose bound DNase and RNase limits their action to the cell
surface thereby precluding secondary membrane charge changes due to the
enzyme acting inside the cells.

13. Measuring the nucleic acids released using spectrophotometric
methods could be 100 times too insensitive.

14. In vitro studies indicate that.cisplatin , actinomycin-D,
methotrexate, and mercaptopurine are effective in mimicking

nuclease incubation. The effect is not immediate but occurs about

6 hours after exposure to the drugs. The protein inhibitor puromycin
removes the nuclease sensitive groups but also has other mobility
lowering effects.

15. After removing the surface nucleic acids with enzymes and
allowing the cells to grow in vitro the surface charge returns to

the control level in about 4 hours.

16. Similarly, in platinum treated tumors in vivo that begin to

grow again the RNase and DNase sensitivity returns. So this sensitivity
is probably linked to cell metabolism.

17. No conclusive evidence is presented regarding the role of cell
surface nucleic acids.

18. Other tumors also show nucleic acids on their surface, with DNA
more commonly found on cells of lymphatic origin. The non tumorigenic

cells tested showed no sensitivity to nucleases.

Discussion
The work presented here adds new evidence to the slowly
growing story of cell surface nucleic acids. An old tumor line,
the S-180, has provided an excellent model for this study because
of its relatively large amounts of apparent RNA-DNA on the surface.
Perhaps this tumor will make a good model cell for examining the

origins,and for physically characterizing these surface groups.

I. Proof of Surface Nucleic acids

There is no absolute proof for cell surface RNA or DNA
presented here. There is a lot of evidence suggesting such molecules,
however. The action of RNase and DNase is the most thoroughly
examined and probably the best evidence and it is strengthened
by the Sepharose-DNase and Agarose-RNase and the positiVe action
of Eco-R1 and Sam I endonuclease restriction enzymes.

The specific binding of the 'platinum-blues' to only the cells
expected to have nucleic acids and the similar binding of anti-DNA
antisera to untreated cells lend further indirect evidence of the
existence of membrane nucleic acids.

In the in vitro studies, the RNA or DNA synthesis inhibitors
specifically cause the loss of surface charge equal to that which could
be released by incubation with RNase. Purcmycin and cycloheximide
(not shown) cause lowered mobility which results in the loss of RNase
susceptible groups but they also cause further reductions probably
due to the loss of membrane proteins. The exact mimicking of RNase
by the nucleic acid synthesis inhibitors is further evidence of

surface nucleic acids.

106

107

II. Overlapping effects of RNase and DNase.

The interlocking effects of RNase and DNase raise interesting
questions. Examination ofthue enzyme has shown no protease contamination
which rules out that form of common denominator. The responses of
different cell types, some DNase sensitive only, some RNase sensitive
only, also tend to contradict a common contaminant as the cause
of the overlapping actions of DNase and RNase.

If they each are acting only on DNA and RNA respectively
then the surface of the cell must harbor a nucleic acid complex.
Furthermore, this complex must be bound to the membrane in such a
way that partial degradation of either RNA or DNA would release it.
An electrostatic binding would fit this requirement. Once part of
the negative charge is removed by RNase or DNAse the remaining

charges could be insufficient to keep the complex bound.

108

III. Lability of Surface Nucleic Acids and Their Turnover

The ease with which this nucleic acid group comes off with
neuraminidase or trypsin or low ionic strength is another interesting
phenomena. I suspect, therefore, that this nucleic acid is on the
very periphery of the cell, weakly bound to the most exterior
membrane components. Its conformation is such, though, that it
allows neuraminidase and protease to act on their binding sites.

The turnover rate demonstrated in the in vitro studies and
this relative weak surface binding supports the possibility of

the exuded nucleoprotein complex described by Stroun et a197,

Rogersloo, and Kandjian et a1.9°. The fact that it can be detected
on the surface may be due to temporary binding and the amount
detected may be due to synthesis rate or to variations in membrane
affinity.

The extracellular nucleoprotein found by Stroun et al.97 was
virtually insensitive to degradation by RNAse or DNase. This
raises two points. The first is that it could explain the negative
results of the experiment in section II-D-3 of the Results.
The antibody titre for sheep RBC's was lowered by tumor supernatant
but did not return to normal after RNase incubation. If the
complex was insensitive to RNase then that is what would be expected.
This leads to the second point. If the complex is not sensitive
to RNase why does the electrophoretic mobility of the cell decrease
upon incubation with RNase? This forces the existance of both a
susceptible and an unsusceptible RNA fraction in the complex to

satisfy these observations. While this is possible there is no

evidence supporting it.

109

The results I have shown do not directly indicate a nucleic acid
molecule excreted by the S-l80. The evidence does show that their
presence on the surface is linked to cell metabolism, particularly
nucleic acid metabolism. This is shown by incubation with
methotrexate, mercaptopurine, and Actinomycin-D. After incubation
with these inhibitors, the nucleic acid disappears from the
surface over a time period of about 4 hours. If the surface nucleic
acids are removed with enzymes, they reappear in a similar period
of 2-4 hours. These times are shorter than one cell cycle which
is about 24 hours for these cells. This nucleic acid flux either means
these nucleic acids are being degraded and need to be replaced or

that they are being released from the cell.

IV. Size of Nucleic Acid Group

In the Results section II-B—7, I calculated the amount of
nucleic acids released into the Isupernatant by RNase treatment
to be 100 times greater than that detected by whole cell electrophoresis
when measured spectrophotometrically. I also said that if it
was real, than the nucleic acid would have to exist in a large
complex which sits on the surface. Using this ratio of 1:100 an
approximation of the size of this proposed complex can be calculated.
Several non trivial assumptions must be made, however.

Let us assume a nucleic acid complex in the shape of a sphere of
radius R. Let us further assume that only the charges facing away
from the membrane and close to the slip plane will contribute to the
zeta potential. Let us assume that only those charges subtended by

a solid angle of 60° contribute significantly. (See Figure 1‘8 ). This

110

.woameoo pace
caoauoc Hobauonuomhc m cu ocean mean on» mo ccfluoucofiwc o>auoaow on» no Homo: .mH gunman

. .1 44mm

. .
.
:wuuowocoaaw |I\\u\\iw.
w. saga—Eco \\ (M...
xPIINI\\\\. cwom because I
u I
.

o
.

   

women mean

111

turns out to be only 0.047 of the total area of the sphere.

Now, let X be the total number (or volume) of nucleotides in
the sphere. The volume of a sphere is g4fR3 where we can let R
be the radius in 'nucleotide units'. The volume in 'nucleotide units'
is equal to X, so X = -l3‘-1‘I'R3 and it follows that R = [%7,- ]y3 .
The surface area is the 411' R2 or 4‘1T[Z—,'T X]? . Of this surface
area only 0.047 contributes to the zeta potential or 0.047 0 4°[%fi:°]°i
This simplifies to 0.227 [X]°°, which represents the number of
nucleotides in the sphere that contribute to the zeta potential.

Now, using this we can compute what size spherical nucleic
acid complex would be necessary to contain an amount of total charges
100 times greater than the charges that contribute to the zeta

potential. 80, to get this we set up the simple proportion

equation:

 

0.227 [xf’ = 1 .
x 100

Simplifying, this becomes X = [22.713 or X = 11,697 . This
implies a molecular weight of 3.5 x 10°. This is smaller than most
viral genomes, larger than most messenger RNA molecules, but not out
of the range of either an RNA or DNA molecule.

The major assumptions :in this calculation are: 1.) The
complex is a close packed sphere: 2.) the complex is only made
up of nucleic acids: 3.) the slip plane is not bent around the complex.
The assumption of a sphere is not that crucial and a relatively
standard one. The assumption of only nucleic acids is not that
bad either. If we assume that 252 of the complex is protein then
the molecular weight of the nucleic acid would only be reduced to

1.48 x 106 daltons and the protein component would be 0.24 x 106

112

daltons (about the size of a LETS protein)? But the last assumption
could make dramatic changes in the calculation if it were altered.
However, the concept of the slip plane is, in itself, an assumption
and several correction factors have been calculated to make
adjustments for membrane topography, such.as large lumps. Basically,
these corrections require recognizing that the mobility is lower

than would be ecpected when the surface has bumps.2° This supports

the model of the slip plane'missing'some of the charges, and therefore
supports the assumption of only part of the surface nucleotides
contributing to the zeta potential.

Calculating a size that is within reason is somewhat satisfying
and even gives some hope for simple spectrophotometric studies. The
main problem is in detecting small differences in large numbers
because the cells normally appear to excrete nucleic acids or

nucleotides.

V. Solid Tumors

The similarity of response of 5-180 solid tumors to their
ascitic counterparts after RNase and DNase treatment is almost
unexpected. The disruptive procedures necessary to take the tumor
mass to a single cell suspension would seem strong enough to leave
a membrane highly scarred. Since these surface nucleic acids are
removable with a variety of enzymes and low ionic strength it is
strange that they survive these isolation proceedures. The role of
calcium ions here is probably very important. Weiss72 has Shown that
calcium ions are closely associated With cell surface ribonucleic

acids. Calcium is also important in cell to cell adhesion 110’111

113

solute permeability through the hyaluronic acid matrixllz, in the
interaction of antibody and compliment113, and in cell proliferation
mechanismslla, to name a few. Calcium chelating agents are routinely
used when isolating single cells from tissue because it reduces
cell-cell adhesion. The chelation of calcium does not seem to
destabilize the membrane nucleic acid because EDTA and sodium citrate
do not effect ascitic S-l80. Perhaps removing the calcium makes the
membrane components less rigid and therefore more plastic and

resiliant. Thus this would assure more cell survival and more intact

membranes during the physical shearing of tissue separations.

VI. The P815 mastocytoma.

The P815 mastocytoma would make another good model for cell
surface RNA and DNA. The effect of both nucleases exists with this
cell although its not as pronounced as in the S-180 tumor. The
major benefit, though,is that this tumor is syngeneic with the
DBA/Z mouse. This allows much greater versatility in designing
immunolOgical experiments and removes doubt associated with non-

syngeneic tumors like the S-180.

VII. Results of Other Cells.

Most of the cells tested exhibit nucleic acid on the cell
surface if they are tumorigeneic and no nucleic acid of they are
normal. Other investigators have shown that a cell does not have

to be transformed to exhibit cell surface nucleic acids77'°3'°4'87

96,97.98 WeiSS73 has shown that rapidly

or to excrete nucleic acids.
87
growing lymphoid cells have RNA on their surface; Lerner showed that

stimulated lymphocytes have DNA associated with their membrane: and

114

Rogersloo found that these stimulated lymphocytes excreted DNA.

Cater and Woo76 indicate, by electrophoresis, that fetal liver cells
have RNA on their surface, about half as much as hepatoma cells.
I did not find nuclease susceptibility on mixed fetal cells or
on thymocytes or splenocytes of the preleukemic AKR mouse. These
lymphocytes, though, were not specifically activated to a proliferative
state which is probably crucial, and the fetal cells were not care-
fully divided intc tissue types which may be important.

More cells will have to be examined, particularly moreLnormal cells.
The problem is obtaining normal tissue that can be dissociated into
individual cells while keeping the cells intact. There are few
normal cells that are in suspension like ascitic tumors. Of course,
there are few spontaneous tumors that are in suspension either. The
proceedure of tissue separation will also have to be examined

and perhaps improved.

115

VIII. Roles of Surface Nucleic Acids

The evidence, both weak and strong, available at this point
is as follows.

RNase or DNase susceptible groups have been shown on tumor

71,72 ,88,91 73,87 73.76

cells, stimulated lymphocytes,

84

fetal liver cells,
and sperm cells.

In each case, as in thisstudy, the investigators have concluded
that these groups were nucleic acids.

In this study, the presence of these surface nucleic acids was
linked to the cell's nucleic acid metabolism, and furthermore, they
exhibited turnover on the membrane.

Studies have indicated a nucleoprotein complex released from
cells.°3'°7'°°

AKR leukemic mice have a suppressor T-cell popluation exhibiting
surface DNA.91

Human lung cancers have an immune inhibitory substance on their
membrane which has been reported to be a nucleoprotein.92

Could this surface nucleic acid have a role in tumor growth or
immune suppression? At this point there are only conjectures, but let
me outline some of the possibilites.

1. No“ role at all - These nucleic acids could be simply debris
exuded by very rapidly growing cells. Most of the work to data could
not argue againstthis because the concentration of effort has been
on proving it's existance not on it's function.

2. Surface functions - Their role could be fulfilled by their

location on or in the membrane.

116

a.) antigen masking - due to their low antigenicity
nucleic acids could cover tumor antigens.

b.) immunosuppressive molecules - They could be transmitted
to lymphocytes along with tumor antigens and when inside the lymphocyte
suppress antibody reproduction or lymphocyte mitogenesis.

c.) The surface nucleic acids could be stimulators of the
very cell they are located on. These surface groups could act as
receptors which,upcn activation,are incorporated into the cell
where their translated products act. This could be a mitogenic
action or a differentiation or dedifferentiation signal, etc.

3. Intercellular release - these nucleic acids could transmit
information to other cells.

a.) immunosuppressive - analogous to 2,b., these groups
could be released as generalized or specific immune suppressors
which act on any of a number of cells: lymphocytes (killer, suppressor,
helper,cytotoxic), macrophages, or neutrophils.

b.) mitogenic - analogous to 2,c., these nucleic
acids could activate nearby tumor cells by releasing a
growth factor stimulating the tumor growth . They could also act as a
transforming agent to nearby normal cells.

c.) promotor of angiogenesis - these released nucleic acids
could activate tumor vascularization by interacting with endothelial
cells of the capillary walls and stimulating their proliferation.

d.) embryonic communication - the interrelationship betweeen
embryo cells could be partly mediated by nuclecprotein complexes. The
appearance of these complexes during tumor growth may be due

to dedifferentiation and there may not be a target for them to act upon.

117

( The converse of this is also interesting. If the tumor cell is
a dedifferentiated cell trying to interact with the other cells
around it of a different place and time developmentally, then it
would not receive the correct feedback signal for normal

differentiation. It would be stuck in its current state forever.)

The problem with the science of tumor immunology is that none
of these possibilities can be definitely ruled out. The factors
that tumors excrete, that lymphocytes and macrophages excrete, the
chemical nature of their antigens, their mode of binding and killing
etc. are not well characterized. The most common technique for tumor
antigen extraction, for example, presumes that it is protein.65
They dissolve the membrane in BM KCl then salt out the soluble fraction
with ammonium sulfate. This only recovers 15-492 of the antigenic
activity and makes one wonder if there is something lost because
of the design of the technique, such as a nucleic acid component.

The idea of antigen masking by surface nucleic acids could be
a possibility because of the low antigenicity of nucleic acids.
(Note: antibodies do form to the phosphate backbone of polynuclectides,
called NG-I antibodies,115 and crossreact with most types of nucleic
acids) This hypod1esis requires that the cell use a potentially valuable
molecule in a capacity which does not use its specialized information
carrying capacity. It also requires that the cell knows what substance
on its surface is antigenic and that it is in an environment such

that it should cover this antigen. These are amazing aSsumptions

118

considering the nature of tumor cells. Tumor cells are aberrant
growths within an organism and are not parasites which have
invaded an organism. Because they do not have a life cycle outside
of the organism, and because they have developed from cells within
the organism, they have not been exposed to evolutionary pressures.
Without this evolutionary development they would not have special
defense mechanisms, nor the ability to select defenses appropriate
to their environment. Therefore, the notion that the tumor cell
is capable of masking antigens seems untenable on teleological
grounds.

A more likely role than passive antigen masking for these
nucleic acids would be active immunosuppression. This idea
uses the information carrying potential of RNA and DNA. This
kind of information transfer need not be limited to immune suppression
but could be involved in many signals transmitted between cells.
A nucleic acid excreted by a tumor could be part of an embryological
communication system. Stimulated lymphocytes could excrete a signal
which recruits other lymphocytes into specific stimulation. Activation
of macrophages could occur through nucleic acids. Many soluble factors
circulate in the blood sera and many 'growth factors', 'inhibitory
factors', 'suppression factors', 'transfer factors', 'activation

55'56'58 and only partially

factors', have been postulated
characterized. The possibility that they contain nucleic acid should
be kept in mind by investigators when examining these factors.

The blood serum is obviously accustomed to the presence of

nucleic acids because RNase circulates there. Some cancer patients

have elevated levels of RNase117 perhaps to remove breakdown products,or

119

perhaps in response to some form of nucleic acid 'factor'. A serum
protein which binds DNA is also elevated during malignancylla’llg
but its role is currently unknown. The role of excretion of cellular

nucleic acids and their interactions with the organism await further

investigation for clarification.

There is some evidence, however, that the nucleic acid has a
function on the cell surface. Work of Russell and Golub91 shows an immune
suppression activity of AKR leukemic cells when these cells are
allowed to make cant-act with sensitized normal spleen cells. These
cells also contain a surface DNA moiety, which when removed stops
suppression. In a communication from Dr. Golub, he states that the
suppressor cells have a specific antigenic marker - a histocompatability
marker - which is also necessary for suppression. This information
along with the H-2 restriction (lymphocytes must recognize an H-Z
antigen site on affector and effected cells) supports the idea
of nucleic acids being localized on the cell surface for their action.
when two compatible cells make intimate contact the informational
nucleic acid could betransferred.from the surface of one of the cells
to the interior of the other.

A similar process could occur during macrophage antigen collection.
The macrophage is believed to transmit antigenic material, probably
with a carrier, to lymphocytes. It is conceivable that it could
transmit a suppressor molecule concurrently,or,.that the suppressor
molecule could directly inhibit the macrophage. Currently, the only

support for a surface nucleic acid having a suppressive role is the

120

work of Russell and Golub?1Whether it is an immunosuppressant;
whether its action depends upon its location on the membrane or in
solution; or whether it is an artifact of fast growing cells will

have to be answered by future experiments.

IX. Future Experiments

It seems that a sufficient amount of data exists to support
the fact that there are cell surface nucleic acids. From this
point I believe the emphasis in reasearch should focus on two areas.
The first is isolating and characterizing this RNA and DNA - some of
this has already been started. The second is to determine it's
role, if any.

Experiments to accomplish the first goal will probably be

forced to use the techniques currently being used by Rogers,100

Lerner,87 Click,82 Reid,85 and Stroun97. These involve membrane
preparations, nucleic acid extraction and isolation, sucrose density
gradient separations and characterizations, melting point determinations,
and homology studies.

The techniques that I would like to see used are to make proteins
from the RNA as well as RNA and proteins from the DNA. These proteins
may have recognizable properties which can give important clues to
the purpose of the nucleic acids. Roger’s101 work indicates that
only a small portion of the total nuclear DNA shows up being excreted
by lymphocytes. Such a specific piece of DNA may be translatable and
transcribable into Rxognizable proteins. These proteins could be

analyzed in a variety of ways. They could be assayed for immuno-

suppressive properties using a technique similar to Golub'sgl; they

121

could be sequenced and compared to known protein sequences, or

they could be added to tissue cultures to observe growth changes that
they might induce. Use of the 5-180 tumor system may be useful for
some of this nucleic acid isolation and characterization because

of the large quantities present on this cell type.

Experiments to ascertain a role for these surface nucleic acids
will probably start with immunology. Since the first indication of
a role is in immunosuppression, this should be used as a starting
point. While Golub's laboratory pursues the AKR leukemia system
other tumor systems should be explored.

The P815 mastocytoma, syngeneic with the DEA/2 mouse offers a
possible system. I have found both DNase.and RNase sensitivity on
this tumor, although small, so it offers a potential yet to be
explored. DNA-RNA could be extracted from the membrane or from the
supernatant and used in immunosuppressant experiments. Extracted
nucleic acid could be used in vitro in hemolytic plaque assays with
sensitized spleen cells, in macrophage or lymphocyte versus tumor
cytotoxicity assays or in vivo to determine it's suppressive effects
on foreign antigens. If extracted nucleic acids do not produce a
response in the in vitro assays, whole tumor cells could be used. This
would be necessary of H-Q restriction was involved or if other
surface components were important.

The use of the S-180 cell for the examination of nucleic acid
role may be limited due to the unknown character of this tumor.
Although, since it is a non-syngeneic tumor to all mice, it may
represent a system.which requires tremendous immunosuppressive

qualities to overcome major histocompatibility-transplantation antigens.

122

X. The Effect of cisplatin.

No new modes of action for cisplatin were derived from this
study. The drug did have a major effect upon the cell surface nucleic
acid but so did other antitumor drugs specific for nucleic acid
synthesis.

The apparent connection between cell surface nucleic acids and
tumor cells, and their removal by the action of antitumor drugs, and
the possible immunosuppressive role of cell surface DNA, is
all very interesting. These findings suggest that the antitumor
drugs, like cisplatin, may have other effects than the direct

interference with nuclear DNA synthesis.

APPENDICES

Appendix I

The Helmholtz-Smoluchowski equation relates mobility, u,
to the potential at the surface of a slip plane, f , with delectric

constant, 5 , and viscosity,‘ .

LL: [6 (1)

41:41

The surface charge that produces the potential is given by

the Debye-Huckel equation for large, smooth particles.

:: C i 5L (2)

‘Hr
where cr'is the surface charge and k is the Debye-Huckel constant

which can be defined by the equation,
-‘<X
Y = (P6. (3)

which gives the potential, Y , at any point X away from the surface.
l/K is in units of distance and is often called the thickness of
the ionic double layer which surrounds the particle. For large

smooth particles equations (1) and (2) can be combined to givelzo,
and l/K can be related to the ionic strength of the bathing solution:

VK : c“/2513- (5)

where z is the ionic vaiency and c is the concentration in
millimolar. 11K is then given in Angstroms.
Substituting (5) into (4) we can say that the mobility is

proportional to the reciprocal of the square root of the ionic

123

124

concentration.

Ll o( <1“. Vic—f (L)

When the mobility is plotted versus l/fa a straight line should
occur, if 0' remains constant.

In Figure 19 the mobilities of red blood cells (from Heard and
Seamangfi) and the 5-180 are plotted. The red blood cells give a
good straight line with an intercept of zero (infinite ionic
concentration - no mobility). The tumor cells, however, do not give
a good straight line but their mobilities are lower than expected,
indicative of increasing 0‘. Therefore they do not qualify as smooth
particles.

Another equation can be derived which can take this into
account. In Figure 20, a hypothetical membrane is diagrammed. This
membrane has thickness with charges inbedded within it and counterions
from solution which can diffuse into it. The charges that lie on
any of a series of imaginary planes within the membrane contribute
to the zeta potential depending on their distance in and the ionic
strength of the bathing solution. At low ionic strength there are
fewer counterions and the inner charges exert more influence
(figure 20,b). Haydan 27 has developed a model to simplify the
contribution of the inner charges. He defines a potential, #3, due to
the charges at the surface and another potential of the same magnitude
located a distance d beneath the surface. This arbitrary internal
potential is an average of the potentials of all the imaginary
planes within the membrane. Its distance from the surface is such

that it equals the surface potential. With these assumptions he

125

.vmcmasom poo swoon Scum cusp Haoo pooHn own one moamcmwuu pom .pouonpocw omozm
omanm on» one moaowwo ammo .mHowucoo omalm on» who moaoufio pwaom .Hommsn mcfluswooe
on» mo camcouum cacow ecu mo noon chosen 0:» mo Hoooumwoou on» msmuo> huwafinos no uon .mH owsmwm

e
Z 2 3 IR 9 o ~

     

 

Sn

 

 

All'llflflfl

126

u“—
m

surface

/ . : 2n, plane

arbitrary plane
within membrane

Figure 20. Model of cell surface with thickness and the potential
curves from charges in the membrane in a) normal ionic
strength and b) low ionic strength.

127

arrives at an equation relating the potential \h,, the distance d,
and the Debye-Huckel constant K to the measured surface charge 0'.
This equation can be simplified to a first order approximation which

gives

-k:d_
(rec “Poe. <7)

If this is substituted into equation (4) and rearranged and the

proportionalbity between K and {E'substituted it gives
he 46?. .4 -ic .L ”Mic. (8)

If lni-E is ignored an error of less than 107: is added but for
out purposes this is acceptable.‘£,is taken as a constant so a relation
appears which is linear between ln u and {E with a slope of d.
In Figures 11 and 12 of Results section, the mobilities versus ionic
strength are plotted as In u versus {32 The slopes of these curves
represent d, the distance between the surface and an interior plane
which represents the interior charges. For Figure ll.a; d,was calculated
for the control and RNase incubated cells. The value for the control
is 4.1K and for the RNase treated cells 5.83.

Even though the value ,d, is an arbitrary construct representing
a vaguely defined depth in the membrane it can be used for comparison
purposes. The RNase incubated d value is 40% greater than the control.
This suggests a greater influence of deeper charges in the RNase
incubated cells, perhaps due to the loss of charges that predominate

near the surface .

Appendix II

The pH profile curves shown in Figures 8 and 9 show a
shift in isoelectric point to higher pH and an overall uplifting
of the curve due to platinum treatment. Besides the loss of a group
of pK 1-2 as argued in the Results section, the shift in isoelectric
point could be attributed to the addition of a positively charged
group with a pK of 6.3 . This, however, does not follow from the
profiles shown. A pK of 6.3 would indicate a group positively charged
below pH6.3 but neutral above pH 6.3 and while this would explain
the differences in the curves at low pH's it does not explain the

much larger differences at pH 7.4 .

128

Appendix III

Operating Proceedure

1. Turn on Dual power supply, chart recorder, DVM(sa:on DC Volts)
TV, TV camera, microscope light, water bath pump( cooling water from
tap should have water bath cold- about 16°C), and turn motor speed
control from Standby to ON. Make sure that the water bath settles to
an operating temperature of 18°C.

2. Turn stopcocks to drain and pass distilled water through system.

3. Fill chamber with 0.1 M KCl and turn stopcocks to connect chamber
with electrodes.

4. Set electrode power supply to about 200V and turn S-3 left or
right. The chart recorder pen should move towards one side or the
other. (S-Z should be in position D). Turn voltage control on
reference power supply until pen of chart recorder zeros. (chart
recorder zero should be set in the middle for convenience). Turn
S-3 to center position, turn S-Z to E, and place DVM on DC amperes
and read milliamps off meter. Turn S-Z back to D.

5. Plush out KCl, add measuring buffer, and turn stopcocks to
electrodes. Repeat proceedure in step 4 to obtain amperage through
buffer.

6. Set reference power supply until DVM reads desired reference
current, usually 6mA, Turn S-2 to D, and DVM to DC voltage setting.
The chart recorder will now read from 5.5 to 6.5 mA with 6mA at the
the zero point in the center.

7. Flush out buffer, add new buffer and cells to be measured, close
stopcocks. Make sure that the microscope is focused at 1/5 of the
distance between the two inner surfaces of the measuring chamber.
the chamber is 400 um thick, 0 is considered at the inner glass
surface closest to the observer.( The inner surface is easy to find
because it is usually slightly dirty.)

8. Turn 5-3 to the left and S-l to the left. Adjust motor speed control
until the cells appear stationary. Put Sn3 switch at the center
point and record tachometer voltage from the DVM and the chamber
current from the chart recorder. Then turn S-l to the center position.
Do the same for the right hand settings of S-3 and S-l. Repeat these
right and left pairs of meaurements until satisfied a representative

129

130

sample is obtained - about 6-10 times.

9. Plush out sample and put in new sample or clean chamber with
distilled water or detergent.

Note: Settings A,B,D on 5-2 are designed to allow monitoring of
electrode voltage, reference voltage, and motor control voltage
respectively.

Appendix IV

In this discussion I will show the mobility lowering ability
of antisera when it binds to antigens on the cell surface. Sheep
red blood cells, in Alsevers solution and Rabbit anti-sheep BBC
antisera were obtained from the Colorado Serum Co. Varying
concentrations of this sera were incubateiwith washed red cells
and the mobilites measured. Incubation was performed in a barbitone
buffered saline for 30 minutes at 23°C. Normal rabbit sera was used
as a control. In Figure 21 the results are shown. There is a large
drop in mobility with increasing binding of immunOglobulin. At
the highest concentrations of sera the cells lysed,(sera was heat
inactivated to remove compliment activity) probably due to residual
compliment.

The onset of a definitive change in mobility occurs at a higher
concentration of antisera than would be necessary in a compliment
fixation test, so it is not as sensitive as compliment fixation.
However, for those cells that are not susceptible to compliment

induced lysis, this is one way to determine antibody binding.

131

132

.muonwucm 0mm moonmifiucm one mououdo ammo use when Hannah Hmauoc
mum moaowwo pomoHU .chaucudp snow «0 noduoccu o no n.0mm moose mo mowuwaflnoa no uon .HN owsmwm

2.3.2:. 23

as as as s.

 

 

 

BIBLIOGRAPHY

Bibliography

1. B. Rosenberg, L. Van.Camp, and T. Krigas, Nature, 205(1965)698.
2. J.A. Howle and G.R. Gale J. Bact. 103(1970)259.

3. S. Reslova, Chem.-Biol. Interact. 4(197l-l972)66.

 

4. D.J. Beck and R.R. Brubaker, Mutat. Res 27(1975)181.

5. D.J. Beck and J.E. Fisch, submitted for publication (1978), referrred
to in Biochimie l9(l978)233.

6. R. Lecointe, J.P.Macquet, J.L. Butour and C.Paoletti, Mutation Res.
48(1977)139.

7. W.R. Leoposd, E.C. Miller, and J.A. Miller, Cancer Res.,39(l979)913.

8. P.B. Conran, and B. Rosenberg, "Advances in Antimicrobial and
Antineoplastic Chemotherapy", Vol II, Univ. Park Press, Balt.,
p.235, 1972.

9. A. Khan, ianlatinum coordination Complexes in Cancer Chemotherapy"
(Ed. by T.A. Connors and J.J. Roberts) Springer-Verlag, Heidelberg,
1) 124919740

10. A. Sodhi Abs. of the Toulouse Mtg on floord. Chem and Cancer Chemo.
JUIY 1978.

11. A. Sodhi, J. of Clin. One. and Hema., 7(1977)569.
12. E. Heinen, see ref #10.

13. M. Gosalvez, M. Blanco, J. Hunter, M. Miko, B. Chance,
Eur. J. of Cancer, l9(l974)567.

14. Kahkarova, Tatianenko, Moshkovsky, Raykhman, Kondratyeva,
published in 1977 in Russian - english abstract.

15. H.C. Harder, B.Rosenberg, Int. J. of Cancer 6(1970)207.

l6. J.M. Pascoe and J.J. Roberts,Biochem. Pharmacol 23(1974)1345.

133

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

134
J.M. Pascoe and J.J. Roberts, Biochem. Pharmocol., 23(1974)1359.

S, Mansey, B. Rosenberg and A.J. Thomson, Journal of the Amer.
Chem. Soc., 95(1973)1633.

J.J. Roberts and A.J. Thomson, In;"Nucleic Acid Research and
Molecular Biology", p. 71, 1979.

H.W.van den Berg and J.J. Roberts, Mutation Res. 33(1978)lll.
J. Froehlich, R. Srivastava, and G. Eickhorn, Abs. see ref 10.
D. Herman, W. Guschlbauer, Abs. see ref 10.

A.B. Robins, Abs. see ref 10.

D.H. Heard and G.V.G. Seaman, J. Gen. Physiol., 43(1960)635.

H.A. Abramson, "Electrokinetic Phenomena and Their Application
to Biology and Medicine". The Chemical Catalog Co. New York, 1934.

J. Henniker, Rev. Mod. Phys. 21(1949)322.
D.A. Haydon, Biochim. Biophys. Acta, 50(1961)450.

C.C. Briton,Jr. and M.A.Lauffer, In: " Electrophoresis, Theory,
Methods, and Applications.", Ed. M.Bier, 1959. p.427.

"Cell Electrophoresis”, Ed., E.J. Ambrose, J.A. Churchill Ltd.,
London, 1965.

J.N. Mehrishi, In: Prog. in Bioph. and Mol. Biol. 25(1972)l.

F.M. Burnet, J.F. McCrea, and J.D. Stone, Brit. J. Exptl.
Pathol., 27(1946)228.

 

G.L. Ada and J.D. Stone, Brit. J. Expt. Pathol., 31(1950)263.

E. Klenk, "Ciba Found. Symp. Cheme. Biol. Mucopolysaccharides."
p 311, 1958.

H.C. Wu, E. Merxan, P.H.Black, Biochem. J., 8(1969)2509.

C.C. Sweeley, G. Dawson, In: "Red Cell Membrane", Eds., G.A.
Jamieson and T.J. Greenwalt, 1969, p172.

S. Hakomori, Proc. Natl. Acad. Sci.,USA, 67(1970)l74l.

L. Weiss, J. Natl. Cancer Inst. 50(1973)3.

135

38. S. Kozawa, Biochem Z. 60(1974)l46.
39. E. Mayhew and L. Weiss, Expt. Cell Res. 50(1968)441.
40. C. Sato, K. Kojima, K. Nishizawa, Biochim. Biophys. Acta,470(l977)446

41. G.V.G. Seaman and G.M.W. Cook, In; Cell elctrophoresis, see ref 29.
p. 480

42. J.N. Mehrishi, D.R. Grassetti, Egtggg, 224(1969)563.

43. D.M. Mitchel and D.B. Cater, Br. J. Exp. Path. 52(1971)152.

44. C.B. Thoman and D.B. Cater, Br. J. Exp Path. 53(1972)111.

45. A.L.Griffich, N.Catsimpoolas, N.H.Wortis, life Sciences 16:1693.
46. S.A. Smith and B.R.Ware Biophysical J. 21(1978)179a.

47. P.Blume A. Mallen, R.Knox, G.V.F. Seaman, Nature 271(1978)378.
48. E.J.Ambeose, A.M. Jones, J.H.B. Lowick, Natugg 177(1956)576.

49. C.F. Fuhrmann, In:'tell Elctrophoresis: see ref 29, p. 92.

50. P.S. Vassar, G.V.F. Seaman, D.B. Brooks, In: Proc. 7th Canadian
Cancer Research Conf. 1967, p265.

51. L.Weiss, T.S. Hauschika, Int. J. Cancer 6(1970)270.
52. EoMaYhew, JoGeno PhLSiO]. 49(1966)7170 I
54. D.R. ComanJCancer Res., 4(1944)625.

53. J. Ponten, Contact Inhibition, in: Cancer: a Comprehensive Treatise
Vol III, Plenum Press, 1975. p55.

55. J.C. Robbins and G.L. Nicolson, Surfaces of Normal and Transformed
Cells, in: Cancer, A Comprehensive Treatise. p. 3.

56. M.C. Click, Cell Surfaces Changes Associated with Malignancy,
In: Fundamental Aspects of Metastasis, Ed. L.Weiss, 1976.

57. L. Weiss, In: "Fundamental Aspects of Metastasis", p. 25.

58. C. G. Gahmberg, In: Cell Surface Reviews, Vol. 3, Eds., G. Post
and G. L. Nicolson, 1977.

59. S. Hakomori, Adv. Cancer Res. 18(1973)265.

60. T.Muramatsu. G. Cachelin, J.F. Nicolas, H.Condamine, H.Jakob,
F.Jacob, Proc. Natl. Acad. Sci. USA, 75(1978)2315.

 

61.

62.

63.

64.

65.

67.
66.
68.
69.
70.
71.
72.
73.
74.
75.
76.

77.

78.
79.
80.
81.

82.

83.

136

R. Roblin, I.-N.Chou, P.H. Black, Adv. Cancer Res. 22(1975)

R.J. Mannino, M.M. Burger,In: Regulation of Growth and
Differentiated Function in Eukaryote Cells, Ed. G.P. Talwar, 1975.

E.M. Hersh, J.U. Gutterman, G.M. Mavligit., Adv. Int. Med.
22(1977)145.

C.W. Parker, Pharm. Rev. 25(1973)325.

 

J.E. Castro, In: Immunological Aspects of Cancer, Ed. J.E. Castro,
1978, p. 10

~D.Noar, Adv. Cancer Res. 29(1979)45.

M.R. Price and R.W. Baldwin, In: Cell Surface Reviews,; 423.

J.F.A.P. Miller, Adv. Cancer Res, 29(1979)45.

 

A.I. Lansing, T.R. Rosenthal J.Cell. Comp. Physiol. 40(1952)337.

S. Chaudhare and I. Lieberman, Expt. Cell Res. 40(1965)527.
L. Weiss and E. Mayhew, J. Cell. Physiol. 68(1966)345.

L. Weiss and E. Mayhew, J. Cell Physiol. 69(1968)281.

M. Bennett, c. Mayhew, L. Weiss, J. Cell Phys. 74(1969)l83
L. Weiss, 0.3. Jung, R. Zeigel, Int. J. Cancer, 9(1972)48.
L. Weiss, and R. Zeigel, J. Theor. Biol., 34(1972)21.

J. Woo and D.B. Cater, Biochem J. 128(1972)1273.

W. Droege, D. Malchow, J.L. Strominger, Eur. J. Immunol.,
2(1972)156.

D.F. Wallach, P.H.Zahler,Proc.Natl.Acad.Sci.USA,56(1967)1552.
T. Takeuchi and H.Terryana, Expt. Cell Res. 40(1965)32.

M.C. Glick, L.WarrenJ Proc.Natl.Acad.Sci.USA,63(l969)563.

 

S.Ya. Davidova, V.S. Shapot, FEBS letters 6(1970)349.

M.C. Click, In: Biogenesis and Turnover of Membrane Macro-

molecules, Ed. J.S.Cook, 1976.

M. Rieber, J. Bacalao, Proc.Natl.Acad.Sci.USA,71(l974)4960.

 

84.
85.
86.
87.

88.

89.

90.
91.

92.

93.
94.
95.

96.

97.

98.
99.
100.

101.

137

A. Singer, B.L. Reid. J. Reprod. Fertil. 23(1970)249.

B.L. Reid, P.M. Blackwell, Aust.J. Med. Technol.,3(l972)121.
B.L. Ried, Biosystems, 5(1974)207.

R.A. Lerner, w. Meinke, D.A. Goldstein, 33153, 68(1971)1212.

S.K. Aggarwal, R.W.Wagner, P.K.McAllister, B. Rosenberg,
Proc.Natl.Acad.Sci.USA,72(l975)928.

P.K. McAllister, B. Rosenberg, S.K. Aggarwal, R.W. Wagner.
J. of Clin Hema. 0ncol., 7(2)(1977)717.

S.K. Aggarwal, J. Histochem. Cytochem., 25(1977)359.
J.L. Russell, E.S. Golub, Proc.Natl.Acad.Sci.USA,75(1978)6211.

A.C.Hollinshead, Proceedings, IX International Symp. Comp.
Leukemia and Related Diseases.0ct. 1979. preprint.

"Immune RNA in Neoplasia"
S. Sell, J. Mendelsohn, Arch. Pathol. Lab. Med. 102(1878)217.
A.A. Gottlieb, R.H.Schwartz, Cell. Immunology, 5(1972)34l.

A. Hollinshead, T.Stewart, In: International Cancer Congress,
Symp.XII,Ed. Chacon, Pergamon Press, Ltd. In Press.l978.

M. Stroun, P. Anker, M.Beljanski, J.Henri, C. Lederrey, M. tha,
P.A. Maurice, Cancer Res. 38(1978)3546.

E. Khandjian, G. Turian, Differentiation 5(1976)l7l.
P. Anker, M. Stroun, P.A. Maurice, Canc. Res. 36(1976)2832.
J.C. Rogers, J. Expt Med., 143(1976)1249.

J.C. Rogers, Proc. Natl.Acad.Sci.USA,73(1976)3211.

102. 0.L. Horer, C.Nicolau, FEBS Letters l4(l97l)262.

103.

104.

B.K. Sinha, C.F. Chingne11,Biochem.Biophys.Res.Com.,86(1979)1051.

W. Wolf, W.Cole, R.C. Manaka, B. Odenheimer, Chem-Biol. Inter.
22(1978)353.

105. G. Politis, M.C.Plassara, H.Thomou-Politi, Nature, 257(1975)485.

106. P.J. Goetz, J.C. Penniman,Jr., Amer. Lab. Oct 1976,p21.

138

107. D. Williams, V.Esquenazi, R.Cirocca, J.Jensen, J. Immunol,
ll6(l976)554.

 

108. R.P. Gale, J. Zighelboim, J. Dmmunol., 114(1975)1047.

109. A.H. Pickaver, N.A.Ratcliffe, A.B. Williams, H. Smith,
Nature(new Biol)235(l972)186.

110. I. Vlodavski, L.Sachs, Exp. Cell Res. 105(1977)l79.
111. N. Zakai, G.Kulka, A.Loyter3 Proc.Natl.Acad.Sci.USA, 74(1977)24l7.
112. M.A. Napier, N.M. Hadlery Proc.Natl.Acad.Sci.USA.,75(1978)2261.

113. W.T. Shearer, J.P. Atkinson, C.W. Parker, J. Immunol. ll7(l976)973

 

114. S.H.H. Swierenga, J.A. Whitefield, S.Narasaki, Proc.Natl.
Acad. Sci.USA, 75(1978)6069.

 

115. j. Panijel, M.C. Delaunez, Biochem.Biophys.Res.Com.
129(1967)158.

 

116. I. Kamo, He. Friedman, Adv, Canc. Res., 25(1977)27l.

117. R. H. Kottel, S.O. Hock, R.G. Parsons, J.A. Hoch, Br. J. Cancer.
38(1978)280.

118. R.G. Parson. P.M. Aldenderfer, R.Kowal, J. Natl. Canc. Insts.
63(1979)43.

119. R.G. Parsons, R.L. Longmire, 5.0. Hoch, J.A. Hoch, Canc. Res.
37(1977)692.

120. J. Th.G. Overbeek, and J. Lijklema, In: "Electrophoresis,
Theory, Method, Applications." l959.p.l.

121. B. Rosenberg, Cancer Chemotherapy Rep.59(l975)589.

 

Figure

12.

13.

14.

15.

16.

17.

18.

page

Log of electrophoretic mobility versus the square 70
root of the ionic strength of the measuring media.

a) Control cells, b) platinum treated tumor.

R,P,N, represent RNase incubation, protease

incubation, and neuraminidase incubation respectively

As in Figure 12 except this is S-180-J, R,D,N 71
represent incubations with RNase, DNase and
neuraminidase respectively.

UV spectra of supernatant from cells (S-180-J) 79
incubated under various conditions. see text
for details.

Plot of mobility versus time in in vitro 88
incubations. Cells were removed from animals,

preincubated with cisplatin (Pt), or with

Actinomycin-D (Act-D), for forty minutes before

time zero. Time shown is time in normal tissue

culture media. R represents RNase incubation of

samples taken during the course of the in vitro
incubations.

In vitro experiments as described in Figure 15. 90
The control curve is repeated in each figure as
reference. a) Normal S-lBO-J were preincubated
with RNase and then allowed to grow in vitro.

b) similar to (a) except cells were preincubated
with DNase. c) Squares; cells were preincubated
with mercaptopurine, 0.5mg/m1. Crosses; cells were
incubated with methotrexate, 0.5mg/m1.

d) the RNase curve of Figure 15 is repeated.
Triangles; cells were preincubated with lmg/ml
puromycin.

Mobility of sheep red blood cells as a function of 97
sera taken from mice injected with (C) sheep

RBC's only: (S) sheep RBC's + S-lBO-J supernatant,

and (R) sheep RBC's + tumor supernatant

incubated with RNase.

Model of the relative orientation of the slip 111
plane to a hypothetical nucleic acid complex.

viii

 

"Ii'iiiiiiiiiiiiiiii