$5....
:34. ...,
2. 41.4 . {lusting

- 1.59.,“ .

any .. i
nfifluwfiw.
.1! 2 L: I
, mid.-ps.fimfinhuu.
yarn-rats;

rmw.

.... .
e! I
t. .

4: l'us'. '
‘tt'hbfifi‘
1'31; h- .2.
a. ‘ .

b:

.2... rfiLfl. L .
IVHHwh'VHvaI;

 

212m: .

, MF‘EM'Y
'~r. W?- :

  

A L L ,, 4.9:. c L -L L v «344.5».uu...|h.|1r101
4 .y.. . I 3.0) . L .. . . . . L r all. n 3.1.. .1. inure .; . . .a
r z : ..L. I. . . . .. . Ifloéhi. $3....) Irl.ri.m.v.r}.r

. _ a. ...u,... MfifivvmwmmeRWIfifimmw; : .. .. . LL, .

 

..t.\vf!

 

MW‘IBSS

MCI HIGAN STATE UNIVERSITY LIB

IIIIIII III III IIIII

31293 00570 9088
”BEAR? I
“31390:! State
nivorllty

 

                               

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

This is to certify that the

thesis entitled

BIOMECHANICAL PROPERTIES OF THE CANINE
PATELLAR TENDON AS A FUNCTION OF AGE
AND DIABETES

presented by

Ronald Lynn Lancaster

has been accepted towards fulfillment
of the requirements for

M. S . . Biomechanics
degree in

 

 

/

Major professor

Date April 13, 1990

 

0-7639 MS U is an Affirmative Action/Equal Opportunity Institution

__.___.__,-—--fi———‘ ._H__.

 

PLACE IN RETURN BOX to remove thIc checkout from your record.
TO AVOID FINES Mum on or befor- date due.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

I

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MSU Is An Afflnnafive Action/Equal OpportunIty Institution

 

BIOMECHANICAL PROPEREIES OF THE CANINE
PATELLAR TENDON AS A.FUNCTION OF AGE
AND DIABETES

BY

Ronald Lynn Lancaster

A THESIS

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

'ms'mn or sonnet

Department of Biomechanics

1990

QOSSbQC

ABSTRACT

BIOMECEANICAL PROPERTIES OF THE CANINE
PATELLAR TENDON.AS.A FUNCTION OFLAGE
AND DIABETES

BY

Ronald Lynn Lancaster

Use of the canine patellar tendon in orthopaedic and
pathological research continues to grow. The canine model
is often used to examine human knee ligament reconstruction
techniques. In the present research, age related changes in
the biomechanical, biochemical, biophysical, and
morphological properties of the canine patellar tendon were
investigated. Alterations in these changes caused by
naturally occurring diabetes mellitus were also determined.
While collagen from control tendon had decreased solubility
to pepsin with age, its mechanical properties remained
unchanged. Diabetes caused a decreasing trend in strength
despite a more rapid decrease in solubility compared to age-
matched controls. Thermal stability decreased with age yet
increased with diabetes. Further work is needed to better
understand both the mechanisms of aging in this connective

tissue and alterations caused by diabetes mellitus.

ACKNOWLEDGMENTS

I would like to express my deep appreciation to: Roger
C. Haut, Ph.D, for his guidance and friendship, Linda
Lancaster for her love and support, and my family for their
support throughout my academic career.

I would also like to thank the following people for
their significant contributions to this research: Ms. Aimee
Farquhar and Ms. Valerie Smyth for the biochemical analyses,
and Jane Walsh and Mr. Vance Kincaid, IT, for the
morphological analyses.

This research was partially supported by the Companion
Animal Fund, College of Veterinary Medicine, Michigan State

University, and by the National Institute of Health.

iii

TABLE OE CONTENTS

Page

LIST OF TABLES.... ..... ........ .......... ... ........... vi

LIST OF FIGURES ........................................ vii

LIST OF SYMBOLS ........................................ viii
Chapter

I. INTRODUCTION......... ....... ....... .............. 1

II. SURVEY OF LITERATURE............... .............. 3

CONNECTIVE TISSUE ORIGIN ...... . .............. 3

COLLAGEN SYNTHESIS AND GROWTH...... .......... 3

METABOLIC CHANGES WITH AGE........... ........ 7

BIOMECHANICAL CHANGES WITH AGE ............... 8

BIOCHEMICAL CHANGES WITH AGE.. ............... 10

BIOPHYSICAL CHANGES WITH AGE... ..... . ........ ll

DIABETES MELLITUS ............................ 12

GLYCOSYLATION EFFECTS WITH DIABETES. ......... 15

BIOMECHANICAL ALTERATIONS WITH DIABETES...... l6

BIOCHEMICAL ALTERATIONS WITH DIABETES... ..... 18

BIOPHYSICAL ALTERATIONS WITH DIABETES.. ...... 19

III. METHODS AND MATERIALS.............. .............. 20

SPECIMENS.......................... .......... 20

MECHANICAL TEST PREPARATION......... ......... 23

PRE CONDITIONING...................... ....... 30

TENSILE TESTS.......... ........ .... .......... 32

BIOCHEMISTRY........... ........... . .......... 34

THERMAL STABILITY..... ....................... 37

MORPHOLOGY............ ......... ... ........... 4O

STATISTICAL ANALYSIS ......................... 41

IV. RESULTS ...................... . ................... 42

GEOMETRIC PROPERTIES............ ..... ........ 42

LPHYSIOLOGICAL PROPERTIES.... ....... . ......... 42

STRUCTURAL PROPERTIES ......... . .............. 49

MATERIAL PROPERTIES...... .................... S7

BIOCHEMICAL PROPERTIES ....................... 66

THERMAL SHRINKAGE PROPERTIES.. ............... 72

MORPHOLOGY ................................... 76

iv

V. DISCUSSION ........................ . ..............

VI. REFERENCESOOOOOOO... ..... ......OOOOOOOOO... ......

8O

LIST OF TABLES

Page
Table 1. Animal age, sex, weight, and breed .......... 21
Table 2. Tendon geometric properties...... ........... 43
Table 3. Thermal shrinkage slope data ................ 76

vi

Figure
Figure
Figure

Figure
Figure
Figure
Figure
Figure
Figure
Figure

Figure
Figure

Figure

Figure
Figure
Figure
Figure

Figure
Figure
Figure
Figure

Figure
Figure

Figure
Figure

Figure
Figure

Figure
Figure

..s
010004010142. 0.) N

11.
12.

13.

14.
15.

17.
18.
19.
20.
21.

22.
23.

24.
25.

26.
27.

28.
29.

LIST OF FIGURES

Page

Collagen synthesis (taken from

Bailey 6)........................... . 4
Formation of glucose derived structures

(taken from Cerami et a1. 8) ......... 6
Glucose induced cross-linking (taken

from Cerami et a1. 8)................ 7
Hyperglycemic chain of events ........... l4
Potted tibia in PVC tube...... .......... 25
Potted patella in box grip.. ............ 27
Mounted PPT complex.... ................. 29
Typical strain-time curve...... ..... .... 31
Typical stress- strain curve ............ 33
Typical HIT normalized load- temperature

response of patellar tendon ....... ... 39
Creep as a function of age .............. 44
Physiological stiffness as a function

of age........................ ....... 46
Physiological modulus as a function of

age.................... .............. 48
Failure mechanisms ..................... 50
Failure load as a function of age ....... 52
Stiffness as a function of age... ....... 54
Failure strain energy as a function of

age............................. ..... 56
Failure strain as a function of age..... 58
Tensile strength as a function of age... 60
Tensile modulus as a function of age.... 62
Failure strain energy density as a

function of age ...................... 64
Composite stress-strain curve.. ......... 65
Pepsin soluble fraction of collagen as a

function of age... .................. 67
Insoluble fraction of collagen as a

function of age...... ................ 69
Total collagen content as a function

of age............................... 71
Control tendon Ts as a function of age.. 73
Diabetic tendon Ts as a function of

age.. ....... ......... ................ 74
Fibril diameter histograms .............. 78
Fibril diameter TEM - 2.5 year old ...... 79

vii

ACL
ATP
CHO

GBM
HCl
HC104
HIT

LO

NIH
NaCl

PT
PPT

s.d.
Se
Sed
$11
812
TEM
TS
Ts

PVC
(Y)

LIST 0' SYMBOLS

Cross-sectional area

Anterior cruciate ligament
Adenosine triphosphate
Carbohydrate

Tensile modulus

Glomerular basement membrane (kidneys)
Hydrochloric acid

Perchloric acid

Hydrothermal isometric tension
Stiffness

Failure elongation

Initial length

Molar concentration

National Institute of Health
Sodium chloride

Failure load

Patellar tendon
Patella-patellar tendon-tibia
Strain

Standard deviation

Failure strain energy

Failure strain energy density
Primary slope of HIT curve
Secondary slope of HIT curve
Transmission electron microscopy
Tensile strength

Thermal shrinkage temperature
Acid level pH = -log[H+]
Polyvinyl.chloride

Years

viii

I. INTRODUCTION

Use of the canine model in orthopaedic and pathological
investigations continues to grow. Orthopaedic studies
include evaluation of autogeneic and allogeneic repair
procedures in reconstructive joint surgery (7,11,28). The
canine patellar tendon (PT) is often used in these
investigations because of widespread parallel tissue use in
human knee ligament repair. Jones (23) first introduced the
use of PT as a graft in 1963. Variations on the amount of
graft and location from which it is excised have emerged
since that time (2). The canine model in pathology related
research is less prevalent (9,35) and yet offers a unique
source of genetically transferred disease states common to
man.

Changes in the biomechanical, biochemical, and
biophysical properties of the canine patellar tendon with
both age and diabetes mellitus (often called an age
accelerator) is not well documented. In this study, a
thorough review of previous investigations into the above
changes during aging in connective tissues of several
species has been conducted. Additional review of known
alterations in these properties with naturally occurring and

experimentally induced diabetes mellitus has also been

2
performed. Concurrently biomechanical, biochemical,

hydrothermal isometric tension (HIT), and morphological
examinations were conducted on control (non-diseased) and
diabetic canine patellar tendons. Variations in measured
properties with age and diabetes are documented. This
information may help lead the scientific community to a
better understanding of both the mechanisms of normal aging
and alterations caused by the crippling and sometimes lethal

disease, diabetes mellitus.

II. SURVEY OF LITERATURE

CORRECTIVE TISSUE ORIGIN

As a germ cell develops into an embryo it
differentiates into three layers. The mesenchyme (middle
layer) is the origin of connective tissue. Cells of
mesenchymal origin produce an intercellular matrix which
contains four major classes of macromolecules: collagen,
elastin, proteoglycans, and glycoproteins (lamina and
fibronectin) (39).

Tendon primarily consists of the protein collagen (85%)
with a small amount of proteoglycans and glycoproteins.
These components, in combination, are responsible for the
physical properties of tendon. The effects of proteoglycans
and glycoproteins on the mechanical function of connective
tissues are important but are beyond the scope of this
investigation. Because tendon is largely collagen, a basic
understanding of the development and alterations of collagen

with age is required.

COLLAGEN SYNTHESIS AND GROWTH
Tendon is primarily extracellular tissue with a small
number of cells called fibroblasts. These cells produce

collagen proteins and the other extracellular components

4
mentioned previously; The synthesis of collagen is

schematically illustrated in Figure 1.

     
    
   

W W I...

l ltlllmmfl
0 I

no“.

I
not

   

Immunis-

   
  

' no. a. i'. tn.

mm a "" mmum um

cream:
.mmmn

a

   

 

I Ilflilllll‘. “null!

mum

FIBBOBLAST

g. ..." II,

. cu -
’ Illlltlllltll Ignatwrt
IIICIIIISII

Figure 1. Collagen synthesis (taken from Bailey, 6).

Polypeptide chain synthesis occurs in the fibroblast
cell beginning with the DNA (genes). Three linear peptide
chains [{a1(I))2a2(I)] (type I collagen) combine inside the
cell through intramolecular cross-linking and form a triple
helix. Hydroxylation (adding of an OH molecule) of proline
and lysine residues occurs, along with the enzymatically
controlled (prolylhydroxylase and lysylhydroxylase) addition
of carbohydrates to specific hydroxylysine residues.
Glycosylation of the hydroxylysine molecules then takes
place by the addition of galactosyl and glucosylgalactosyl

residues (carbohydrates). This process occurs through an

.5
oxygen-glycosidic linkage which is enzymatically catalyzed

by collagen galactosyl transferase and collagen glucosyl
transferase (39). The procollagen chain is then extruded
from the cell by reverse pinocytosis. At this point, a
small portion of the chain ends is cleaved by peptidase
through hydrolysis leaving a tropocollagen molecule.

The first stage in the formation of cross-links between
tropocollagen helices is an oxidative deamination (removal
of an amino group - NHZ) of specific lysine or hydroxylysine
residues, located at the nonhelical ends (nitrogen- and
carbon- terminals), by a copper-dependent enyzyme,
lysyloxidase. The resulting aldehydes (monovalent group
with CHO) are believed to condense with e-amino groups of
hydroxylysine which are situated in the helical regions of
neighboring molecules, to form a network of intermolecular
bonds which are assumed to be of the aldimine type. This
results in a highly ordered quarter-staggered arrangement of
.tropocollagen molecules which combine to form fibrils, which
later form collagen fibers (6).

Figure 2 illustrates a second form of cross linking
which is believed to occur in long-lived proteins throughout
the body. These cross-links are formed without the aide of
enzymes and begin by the attachment of a carbohydrate
molecule to an amino group on a protein to form a Schiff
base. The Schiff base formation is unstable and quickly but

reversibly rearranges to a more stable Amadori product. At

6
least twenty such Amadori products have been identified in

human beings (8).

 

 

W\ :‘o
08- '-ON 4- T = —'n—
N-Iv-N fl/ \fl P
N—P—W n-
0.1).." W—I
"—1 ~¢0 Ito-1
II a.”
00—1'
II—I'
I
mm mm comm

 

Figure 2. Formation of glucose derived structures (taken
from Cerami, et al., 8).

After a long duration (months to years) the Amadori
products may slowly dehydrate and rearrange to form a
glucose derived structure of several possible types. These
structures may combine with certain molecules to form
advance glycosylation end products (AGE's). AGE’s may have
the ability to cross-link adjacent proteins. This form of

non-enzymatic cross-linking is called glucosylation (Figure
3).

7

I cannon:
————> -————> —-— -4>
AhflNO AMADORI ace
enoup scuwranse _ Pnooucr
J GLUCOSE
AMHK)GHOUP

W’K
& W6

 

PEIIMNIEIII' /
GLUCOSE DEHIVED AGE HUD! ID
CROSS-LINK IMP PHOIEIN

Figure 3. Glucose induced cross-linking (taken from
Cerami et al. 8).

Collagen is reported to have a half-life of 2.5 years
(41). Degradation of collagen is catalyzed by a collagenase
specific to the type of collagen. Phagocytosis also helps
to remove degrading collagen. Released hydroxyproline (a
major constituent of collagen) is excreted as di- or
tripeptides or as free hydroxyproline which is usually
resorbed (39).

METABOLIC CHANGES WITH AGE

It is well documented that tolerance to carbohydrates
decreases with age (12). Endogenous insulin actually
increases during senescence while its effectiveness
decreases through a reduction in cellular membrane receptor
sites. People over 85 years of age have approximately 3.5

times higher insulin levels two hours after glucose loading

8
than people 30 years of age. (12). This indicates a

decreasing tolerance to insulin by the cells with age. A
negative feedback loop is generated first by an age caused
reduction of insulin receptors on the cell membrane,
followed by a rise in the level of insulin in the blood.
Over time, the hyperinsulism will cause further reduction in
the number of insulin receptors. This illustrates decreased
carbohydrate tolerance with normal aging is probably caused
by decreased effectiveness as opposed to a reduction in the
amount or production of insulin (12).

The high level of extracellular glucose, caused by the
decreased effectiveness of insulin, may cause non-enzymatic
cross-linking of collagen and other proteins in the body
during normal aging. This leads to pathologies commonly
seen in diabetic patients such as stiffened joints,
cataracts, etc. Because of this, some authors believe that
practically all people over 70 years of age develop

diabetic-like features (12).

BIOMECHANICAL CHANGES WITH.AGE

Once collagen is arranged and cross-linked it begins to
support loads or stresses. It appears that the nature of
the loading determines the type of connective tissue formed.
How much stress a tissue can support is documented in two
ways. Structural properties reflect the strength of the
bulk of tissue and include: failure load (P), failure

deformation (L), stiffness (k), and failure strain energy

9
(Se). Material properties are structural properties

normalized for geometry and more directly reflect the
strength of the specific tissue type and include: tensile
strength (TS), failure strain (S), tensile modulus (E), and
failure strain energy density (Sed).

How connective tissue responds to loads and stresses is
the subject of several inVestigations. The tensile
strength, modulus, and strain energy density to tensile
failure of the human anterior cruciate ligament (ACL)
significantly decreases with age (15). Hubbard and Little
(22), on the other hand, mechanically tested palmaris longus
and the extensor hallucis longus tendons from human cadavers
and found no statistically significant changes in the
tensile strength and modulus with age. The former data
compares to that recently reported by Haut, et al., (20),
who saw minimal change in the tensile modulus of a limited
number of split human patellar tendons between the ages of
18 and 63 years.

The degree of age-related change in the mechanical
properties of connective tissues may be tissue specific.
Vasseur, et al., (42) studied the extent of tissue
degeneration and corresponding changes in the mechanical
properties of canine stifle (hind limb knee) joint ligaments
with age. The greatest amount of degeneration occurs in the
cranial cruciate ligament (ACL), while less is found in the
caudel cruciate and collateral ligaments. Only the cranial

cruciate was mechanically tested and has a decreasing

10
modulus, tensile strength, and failure strain energy density

with age. Woo, et al., (47) investigated the effects of
maturation on the rabbit medial collateral ligament and
documents significant increases in the failure load and
energy-absorbing capacity of the femur-ligament-tibia
complex to maturation. The tensile strength and modulus of
the ligament substance increase during maturation, while the
rates of these changes decrease with age.

Many studies of age on the mechanical properties of
connective tissue have been conducted using the rat model.
It is generally agreed that the tensile strength of rat tail
tendon increases to twelve months of age (19,40,46). Vogel
(46) notes a decrease in tensile strength after 12 months of
age, while Haut (19) and Torp (40) find the tensile modulus
of rat tail tendon increases throughout life. Galeski (15)
documents an increase in the modulus of elasticity with age.
Vogel again reports a maximum modulus at twelve months

followed by a gradual decline with age.

BIOCHEMICAL CHANGES WITH AGE

Traditional biochemical investigations of collagen
includes three separate assays. The first involves the
immersion of a specimen in a weak acid solution to determine
the most recently synthesized collagen. This is usually
followed by digesting the specimen with the enzyme pepsin to
quantify the concentration of more mature collagen with

newly formed cross-links. The remaining specimen is termed

11
the amounts of collagen soluble in acid, pepsin, and the

insoluble fraction should equal the total collagen
concentration (for exact biochemical procedures, refer to
section III).

Vogel studied the effects of aging on the biochemical
properties of rat tail tendon and dorsal skin and finds a
strong correlation between tensile strength and the
concentration of collagen insoluble in weak acids (45). Both
insoluble and total collagen in rat skin decrease after
maturity (4 months of age), while these properties increase
in tail tendon until twelve months of age (46). Hamlin, et
al., (16) and Schnider, et al., (33) note an increased
resistance of collagen to collagenase digestion in human
tendon with age. These studies indicate enhancement of
cross-linking with age and suggest a corresponding increase

in tensile strength and modulus of the tissues.

BIOPHYSICAL CHANGES WITH AGE

The structural integrity of collagen with aging has also
been documented from studies of thermal stability (10). It
has been noted that collagen held isometrically will generate
tension when heated as a result of the breakdown in the
ordered crystalline structure to a preferential random coil
state (26). The shape of the hydrothermal isometric tension
(HIT) curve has been correlated with the character of the

cross-linking within collagen.

12
Allain, et al., (1) measured an increased maximum

contraction and a decreased rate of relaxation in rat skin
from maturity to 18 months. A corresponding alteration
occurs in the chemical bonding within collagen, from the
thermally labile dehydro-hydroxylysinonorleucine cross-links
in the 1-5 month, to a more thermally stabile cross-link of
unknown composition (thought to be an aldimine derivative) in
the 5-18 month old rat skin. Despite the increase in cross-
link stability, there is no change in the shrinkage
temperature of rat skin with age.

Mitchell and Rigby (29) find a small increase in
shrinkage temperature of rat tail and human tendon, along
with an increase in thermal stability with aging, as measured
by the rate of tension development during heating. they
infer two possibilities: 1) an increase in cross-link density
with age or, 2) an increase in cross-links that are thermally

stable with age.

DIABETES MELLITUS

Diabetes affects the islands of Langerhans which are
masses of small cells in the pancreas that produce insulin.
Insulin is required by all cells in the body, except the
brain, to mediate the transport of glucose across the cell
membrane. Glucose is used to produce energy (ATP) within the
cell. Without insulin, glucose remains in the extracellular
fluids of the body and causes a chain reaction of events as

illustrated in Figure 4.

13
extracellular fluids of the body and causes a chain reaction

of events as illustrated in Figure 4.

Two forms of insulin dependent diabetes are discussed
in this document. Type I diabetes mellitus is called
naturally occurring, juvenile onset, and/or spontaneous
diabetes. As these names indicate, it is a genetically
transferred disease with symptoms that spontaneously emerge
at different times during the maturation of many species.
The second form of diabetes is experimentally induced in
laboratory animals by chemical poisoning (3) or removal of
the pancreas (in some animals of the present study) to study
the effects of hyperglycemia on several tissues and systems

in the animal.

14

INSULIN fifEICIENCY

ll
INCREASED PLASMA fiEUCOSE AND KETONES
II
INCREASED FILTRATION 8F GLUCOSE AND KETONES
ll
EXCRETION OF GLfifOSE AND KETONES

II

DECREASED PfiASMA VOLUME
11

DECREASED VENOUS RETURN
II

DECREASED CfiFDIAC OUTPUT
ll

DECREASED BfifOD PRESSURE
II

DECREASED BRfiEN BLOOD FLOW

II

BRAIN DAMAGE, COMA, AND DEATH
Figure 4. Hyperglycemic chain of events.

Serious complications and sometimes death may result
from untreated or poorly controlled diabetes. Alterations
in practically all tissues of the body occur as a result of
this disease including: peripheral arterial diseases,
atherosclerosis, osteoarthritis, osteoporosis, and
osteopenia, peridontal diseases, tendonitis, bursitis,
pseudo-scleroderma, loss of sensation or feeling in the
extremities, limited joint mobility, cataracts, kidney
failure, and other pathologies. The degenerative processes

which occur in the diabetic individual may be caused or

15
influenced by genetic abnormalities, metabolic

abnormalities, or both (34).

Human skin samples taken from the hands of juvenile
diabetics are found to have decreased vascular lumen,
accumulation of collagen, and changes in the collagen fibril
arrangement (43). Decreased vascular lumen causes tissue
hypoxia causing further thickening of the skin tissue and
once again illustrates a negative feedback situation in the
diabetic metabolism.

Complications from diabetes often simulate pathologies
found in senility and lead to the expression, age
accelerator. Research was conducted by Hamlin, et al., (17)
in which collagenase digestion of human collagen was used to
predict donor ages. There is remarkable accuracy in the age
prediction of a non-diseased donor while the predicted age
of a diabetic patient is much higher than the true
chronological age. Many authors have remarked that while
the complications of diabetes resemble advancing age, the
disease process is very complicated and is probably not true

accelerated aging (5,17).

GLYCOSYLATION EFFECTS WITH DIABETES

The differences noted between diabetes (lack of
insulin) and normal aging (decreased effectiveness of
insulin) is only part of the story. As mentioned previously
glucose in the extracellular fluid of the body has the

ability to form cross-links between adjacent proteins.

16
Glycosylation of collagen increases with normal aging (33)

and is prematurely present in the connective tissues of
diabetic patients (43).

Non-enzymatic glycosylation of collagen is increased in
experimentally induced diabetic rat aorta (3), skin (24),
and tail tendon (5). It is also increased in human skin and
collagen samples from naturally occurring diabetics (32,33).
Hemoglobin in blood becomes glycosylated by glucose and its
concentration is used to rate the level of metabolic control

of diabetes with insulin (43).

BIOMECHANICAL.ALTERATIONS WITH DIABETES

The rat is by far the most experimentally studied
animal. Experimental diabetes is normally induced by
streptozocin injection which poisons the pancreatic cells.
Subsequent effects of the induced hyperglycemia on many
connective tissues frbm the rat are well documented.

Andreassen, et al., (5) found a 50% increase in tensile
strength of rat tail tendons over controls after 30 days of
experimentally induced diabetes. A second group treated
with insulin therapy had no changes in the mechanical
properties when compared with controls. Galeski, et al.,
(14), using the same model, reports increased stiffness and
modulus after about four months of diabetes. Insulin
therapy prevented all documented changes.

Two additional studies were conducted by Andreassen and

others (3,4) using the diabetic rat model. The first

17
involved mechanically testing aortic tissue in which there

is an increased tensile strength after three months of
hyperglycemia. Insulin therapy again prevented these
changes (3). The same investigators measured collagen
content and strength in granulation tissue of rat skin
formed under hyperglycemic conditions. The results indicate
that untreated diabetic rat skin has a decreased failure
strain energy density after only twenty days. An insulin
treated group showed a decrease in failure strain energy
density for the first seven days, which then returned to the
level of controls by the tenth day. It is concluded that
collagen deposition is decreased during the early phases of
healing in experimental diabetes compared with normoglycemic
animals (4).

Bones from diabetic rats were tested in torsion by
Einhorn, et al., (13). Experimental diabetes decreased the
torsional strength, energy absorption, and stiffness of rat
femurs. Canine femurs from naturally occurring diabetics
were tested in four point bending (9). No statistical
changes were found in the material properties of bone from
naturally occurring diabetics. There is, however, a
decrease in the cross-sectional area caused by osteopenia
which leads to a weakening in the structural integrity of

the femurs.

18
BIOCHEMICAL ALTERATIONS WITH DIABETES

Andreassen, et al., (5) used tritiated sodium
borohydride to reduce pepsin insoluble collagen and measure
the level of glucose attachment to lysine and hydroxylysine
residues. While no increase in reducible cross-links is
noted, glycosylated collagen increased considerably in the
diabetic rat model. There is also reduced collagen content
in experimentally induced diabetic rat aorta after three
months (3). Glomerular basement membranes (GBM) in induced
rats were examined by Romen (30). A decrease in collagen
synthesis with an even larger decrease in catabolism leads
to an increase in GBM thickness.

Smith, et al., (37) looked at skeletal muscle
proteolysis in the diabetic rat model and found a 30%
increase in protein degradation along with a 60% increase in
myofibrillar catabolism after only one day of diabetes. As
with all other experimentally induced diabetic rat tissues,
insulin reversed all changes caused by hyperglycemia.

Human dura mater is a lining around the brain which is
nearly pure collagen. Schnider, et al., (32) examined
human, juvenile diabetic dura mater and found an increased
resistance to cyanogen bromide digestion over non-diseased
tissues. In discussion of their findings they noted an age-
like cross—linking effect caused by the disease.

Human skin from juvenile diabetics has increased cross-
linking determined by decreased solubility to collagenase

and cyanogen bromide (44). Also seen in the naturally

19
occurring diabetic is an increased glycation and collagen

content (44). Kohn, et al., (25) also reports increased
collagen formation in human diabetic skin because
procollagen molecules are found to have abnormal reducible
cross-links while lacking some normal reducible cross-links.
Galeski, et al., (14) measured an increased collagen
fibril diameter of experimentally induced diabetic rat tail

tendon.

BIOPHYSICAL.ALTERATIONS WITH DIABETES

Experimentally induced diabetic rat tail tendon has a
30% increase in maximum contraction force, a 80% increase in
relaxation time after that force is reached, and no changes
in these parameters with insulin therapy as compared with

controls (5).

III . METHODS AND MATERIALS

SPECIMENS

All tissues were obtained from the Michigan State
University Veterinary Clinical Center. Twenty nine canines
were used as control animals. Both sexes were used from
varied breeds and weights. The ages ranged from six months
to fifteen years. Their average weight was 26.318.3 kgs.
and ranged from 10 to 42 kgs.

Seven naturally occurring and two experimentally
induced diabetic canines were also obtained from the MSU
Veterinary Clinical Center. Their ages ranged from 2.5 to 9
years old. Both sexes from varied breeds and weights were
used in this study. 'Their weights ranged from 8.2 to 30.8
kgs. with an average weight of 20,1 7 kgs (Table 1).

Control specimens were acquired at the Veterinary
Clinical Center necropsy laboratory from animals that had
died acutely or following euthanasia for varied reasons.
Examples of acute deaths include: old age, trauma, gastric
dilation, and localized tumors. Medical records and
necropsy results were screened to ensure that the animals
were free from diseases known to cause degradation of

connective tissue.

20

Table

1.

CONTROLS

A

DIABET

*
'k

* Indicates experimentally induced diabetic animals.

0 O O O O O O O O O O O O O O O O O

OOOOOOOOOOU‘IOOOOUIUIOOU'IUICDUT

HLJF‘HFJ
NHI—‘OOkaOQxlmmmmrbowaNNNI-‘I—‘OO

O

13.0
14.0
14.0
15.0

ICS

\DkOKOO‘OSU'IIbNN
OOOOOOOU‘IU'I

Animal age,

YEAR )

U)
[‘1
X

"I3r§fiJZ'UHIKIZ"13IZHIK'USIW'UHJZEZS:ZI§Siz'fifil I

SWZWWWZWW

sex, weight,

and breed.

BREED

 

German Shepard
German Shepard
German Shepard
Schnauzer
Samoyed

Cocker Spaniel
Rottweiler
Chow

Husky

Am. Eskimo

Mix

Collie

Boxer

Black Lab.
Gld. Retriever
Black Lab.

Mix

Irish Setter
German Shepard
Mix

Shelty

Gld. Retriever
Airdale

Irish Setter
Mix

Collie

Black Lab.
Irish Setter

Mix
Beagle

German Shepard
Gld. Retriever
Gld. Retriever
Mix
Gld.
Gld.
Gld.

Retriever
Retriever
Retriever

22
The diabetic animals were donated to the Veterinary

Clinical Center when they became spontaneously diabetic
before twelve weeks of age. From the time of diagnosis
serum glucose was regulated by one to two daily injections
of insulin (NPH or PZI, beef-pork mixture). The dosage was
sufficient to prevent ketonuria, however glucosuria and mild
hyperglycemia were not uncommon. The dogs were exercised
and fed twice daily. Glycosylated hemoglobin values were
determined by the method of Standefer and Eaton (38) and
ranged from 3.6% to 8.4% with a median of 5.1%. Urinary
glucose concentrations determined by the Beckman Dri-STAT
Enzymatic Glucose-NH Endpoint Reagent ranged from 0.3 to 162
mg/mg createnine with a median of 28. The dogs were
maintained at the Veterinary Clinical Center for life (4 to
9 years). Euthansia was performed following completion of
the NIH-supported study (35).

The two experimentally induced diabetics had
pancreatectomies at six months of age and subsequently
required exocrine pancreatic supplements (Viokase, A.H.
Robbins Co., Richmond, VA 23220). They were boarded and
cared for in the same manner as the naturally occurring
diabetic animals until euthanasia was performed (36).

The hind limbs of each canine were removed and frozen
at -200 centigrade within 12 hours of death. The limbs were
kept frozen for a period no greater than nine months. Prior
to mechanical or HIT experiments the limbs were removed and

allowed to thaw overnight at room temperature.

23
Once thawed the patella-patellar tendon-tibia complexes

(PPT) were isolated from a randomly chosen limb of each
animal for mechanical testing and subsequent biochemical
evaluation. Patellar tendons from selected contralateral

sides were also isolated for use in the HIT experiments.

MECHANICAL'TEST PREPARATION

The patellar tendon measurements were made by a single
examiner using vernier calipers. Width and thickness of the
tendon were measured at three points along its length and
averaged. The initial length was taken from the inferior
pole of the patella to the superior-posterior attachment of
the tendon on the tibia. The tendon was wrapped in
physiological saline (0.9% NaCl) soaked gauze and was kept
moistened throughout the testing procedure.

The tibia was cut at the distal end to a length of
approximately six inches. A 1.5 inch diameter, 6 inch long
polyvinyl chloride (PVC) tube was placed vertically on a
paper mat under a ventilation hood (Figure 5). A room
curing fiberglass resin (NAPA brand Fiberglass Reinforced
Plastic Filler 6371) was mixed with compatible hardener and
poured into the PVC tube. The tibia was pushed down into
the resin such that the posterior-inferior condyles rested
on the top edge of the tube. Care was taken to ensure that
the tendon did not come in contact with the resin. Once

properly placed the potted configuration was allowed to

24
harden at room temperature. The tendon remained wrapped in

moistened gauze throughout the potting procedure.

25

 

Figure 5. Potted tibia in PVC tube.

26
Once the tibia was firmly fixed in the PVC tube the

patella was potted. A special stainless steel box-like grip
(1.5x1.5x1.25 inches) was used to hold the patella (Figure
6). A slot cut in the bottom of the box allowed the tendon
to pass through without restricting its motion. The front
lid of the box was removed to pot the patella. The inside
was sprayed with a silicone based lubricant to allow easy
removal of the potted patella once the tensile test was
complete. The same fiberglass resin/hardener mix described
previously was used to secure the patella. The box was
nearly filled with the fiberglass. The patella was pressed
into the resin up to the anterior (tendon-covered) surface
and covered by the front lid of the box. The apex (inferior
pole) of the patella was allowed to rest against the bottom
of the box. The tendon was passed through the slot
described earlier, the front was secured, and the resin was
allowed to harden for'ten to fifteen minutes.

A standard servo-hydraulic tensile testing machine
(Instron model 1331) was used to mechanically test the
complex. A specially designed fixture was built to allow
exact alignment of the tendon along the load axis (Figure

7).

27

 

Figure 6. Potted patella in box grip.

28
The fixture was composed of three horizontal 1/4 in.

stainless steel plates and a vertically mounted rotating
disc of the same material. The bottom plate was secured to
the base of the Instron machine. Screw drives were mounted
between each set of plates which permitted direction changes
toward, away, and to the right and left of the operator.
This x-y table configuration enabled the operator to
precisely align the tendon along the load axis.

The rotating disc welded to the top plate
simultaneously served two important functions. Its primary
function was to provide a secure mounting mechanism for the
potted tibia. Its secondary function was to allow for angle
changes between the tibia and the load axis. An inclusive
angle of 1100 was used to simulate the physiological loading
configuration and allow pure tensile deformation of the PT
without a tearing or cutting action at the tibial
tuberosity. I

The upper part of the loading mechanism included a 2000
lb. load cell (Instron model 1010-AF) mounted in line with
the load axis. Below the load cell was a universal joint
used to correct minor offsets in the PPT alignment. The box
containing the potted patella was connected to the lowest

portion of the universal joint.

29

 

Figure 7. Mounted PPT complex.

30
Physiological saline was heated to 37°.C by a heat

exhanger within a tank placed above the crosshead of the
Instron. Gravity fed saline was passed through a tube to a
drip mechanism which gave the operator control of the saline
as it flowed onto the tendon. The drip tube was positioned
at the top of the tendon where it first exited the box grip
which permitted continuous moistening of the tendon during

the mechanical tests.

PRE CONDITIONING

Each PPT complex was cyclically loaded from 90 (N) to
180 (N) at one hertz for 40 cycles. Actual setting was for a
frequency of 1 hertz, but inaccuracy in a dial potentiometer
indicated the resultant frequency for all tests was slightly
more than 1 hertz. A standard linearly variable displacement
transducer within the Instron actuator measured the grip-to-
grip deformations which occurred during the cyclic tests. A
Nicolet oscilloscope and compatible disk drive collected and
stored the load and deformation voltage data continuously.
Strain versus time curves were generated using the
deformation data. Creep in a tendon was defined as the
change in strain between the first and fortieth cycle of pre—
conditioning. The physiological stiffness was determined by
hand from the slope of the loading segment of the fourtieth
cycle from the load-deformation data as displayed on the
oscilliscopy. Figure 8 illustrates a typical strain time

curve and the points at which the creep values were

31
which the creep values were documented in this study. The

physiological modulus was determined using the relationship:

E a kLo/A.

 

 

 

 

 

 

CHEEP
4- /
STRAIN I/
m.)
2‘Wr
U- L
U ' 20 40

TIME (s)

Figure 8. Typical strain-time curve.

32
TENSILE TESTS

Immediately after pre-conditioning the complex was
loaded to failure at a nominal rate of 100 % of the initial
length of the tendon per second. Continuous load and
deformation data were collected and stored in the same
manner as the cyclic tests. The mechanism of failure in the
PPT complex was visually determined and documented.

Cross-sectional area and initial length of the tendons
were used to normalize the load and deformation data,
respectively. Stress versus strain plots were generated
using the normalized data, and material properties of the
tendon were determined from these plots. The tensile
strength was defined as the peak stress achieved in the PPT
complex during a midsubstance failure and the corresponding
strain value was termed failure strain. The tensile modulus
was determined by hand from the slope in the linear portion
of the stress-strain curve. Figure 9 illustrates the
representation of a typical stress-strain curve and the

location at which the above-parameters were determined.

33

 

 

 

 

 

STRESS 79—“ '
(MPa)
40-
E
s:
\
o- . -.
o 9 18

STRAIN (%)

Figure 9. Typical stress-strain curve.

The load-deformation and stress-strain data were loaded
into a computer program which calculated the area under each
curve using the trapezoidal technique. The failure strain
energy was defined as the area under the load-deformation
curve and the failure strain energy density was the area

under the stress-strain curve to failure.

34
BIOCHEMISTRY

The collagen content in each tendon was determined
through a process of extractions and enzyme digestion.
Three aliquots were used in each of two soluble fractions
and one insoluble fraction collected and analyzed for their
individual collagen content by the method of Schnider and
Kohn (33). A separate assay was conducted to determine the
total collagen content in each tissue and served as a check
for losses during sequential extractions.

For each assay approximately 15-20 mg. of tendon was
frozen in liquid nitrogen, mechanically smashed, lyophilized
overnight, weighed, and placed into small test tubes. The
lipids were removed by agitating the samples overnight in
chloroform-methanol (2:1 V:V) at 4° C.

The following day the samples were centrifuged at 2000
rpm for 1.5 hours. The supernatant was removed and
discarded. The above procedure was repeated two additional
times for a total of 72 hours of extraction.

After removal of the supernatant on the third day, the
sample was dried with nitrogen gas. The total collagen
samples were immediatly hydrolyzed and neutralized by the
procedure to follow.

Once dried the aliquots used for solubility tests were
suspended in two ml. of 0.5 M acetic acid, agitated
overnight at 4° C, and centrifuged for 1.5 hours on the
following day. The supernatants were removed and placed in

a clean, dry test tube. The above procedure was repeated

35
for two additional nights. The supernatants were collected

and pooled. After removal of the supernatant on the third
day the pellet was suspended in 2 m1. of acetic acid and
digested in 2 ml. of pepsin enzyme at a concentration of 1
mg/ml for 18 hours at 4° C. After digestion, a neutral salt
solution of 1.0 M NaCl in 0.05 M Tris-HCl buffer (1:1 V:V)
with a pH of 7.4 was added to the aliquot. Enough salt
solution was added to completely neutralize (pH 7.0) each
sample. The samples were centrifuged for one hour at 4000
rpm. The supernatant was removed and transferred to a
clean, dry test tube. The pellet was rewashed twice with
the salt solution, centrifuged for 20 minutes, and the
supernatants were combined. This was the pepsin soluble
fraction. The remaining sample was the insoluble fraction.
The pepsin soluble fraction was dialyzed in 0.1 M
acetic acid overnight using cellulose membranes (VWR
Scientific, m.w. cutoff 12,000-14,000) to remove excess
salt. All fractions were frozen and lyophilized overnight

and then hydrolyzed.

HYDROLYSIS
2 ml. of 6 M HCl was added to the test tube to dissolve
the precipitate. The sample was incubated overnight at 105°

C in a sealed test tube.

36
NEUTRALIZATION

The following day, the sample was neutralized by the
following procedure: 2-3 drops of methyl-red indicator was
added to the hydrolyzed sample. 2400 pl. of 2.5 M NaOH was
then added and vortexed. Additional NaOH was added until a
pH of 7.0 was reached. The acid- and pepsin-soluble, and
total collagen fractions were diluted to 10 ml. The
insoluble fraction was diluted to 25 ml.

An assay for collagen was conducted on all the samples.
A set of ten standards (1-10 ug/ml) was prepared from a
stock of 1.0 mg hydroxyproline/100 ml. Two water samples
were used for zero standards.

The assay was conducted using 1 ml of sample. Three
solutions were used during the assay preparation including:
3 M HClO4, 0.03 M chloramine-T solution (0.845 g Chloramine-
T in 20 ml H20, 30 ml propanol, 50 ml citrate-acetate
buffer, pH 6.0), and 5 % solution of p-
dimethylaminobenzaldehyde (p-Dab) in propanol (2.0 g/40 ml)
dissolved under heat. 1 ml. of the chloramine-T solution
was added to each test tube, vortexed, and let to settle for
20 minutes. 1 ml. of 3 M HClO4 was then added to each test
tube, vortexed, and let to settle for five minutes. 1 ml.
of p-dab solution was then added, vortexed, and incubated at
60° C. for 18 minutes (according to Stegemann, 38).

Once the test tubes had cooled to room temperature, the
absorbances of the standards and samples were measured on a

Beckman Model DG-spectrophotometer. The water samples were

37
prepared with reagents as above and served as reference.

The absorbances of the samples were compared with the
standard curve for the calculation of hydroxyproline per dry

weight of tissue.

THERMAL STABILITY

The frozen limbs designated for use in the HIT tests
were allowed to thaw at room temperature overnight. The
patellar tendon was excised and immediately placed in room
temperature physiological saline (0.9% NaCl).

Tendons were laid in a pool of saline on a dissecting
table. Through a stereo microscope the strips of tendon
were out along the long axis. The anticipation was for the
strips to approximate individual fascicles, at least having
continuous fibers of collagen. This, however, was not
verified further. Once excised, the tendon strip was.
mounted between flat plate clamps and held isometrically at
12 mm.

The HIT tests were conducted using a specially designed
fixture. The gripped tendon was loaded into the device and
an insulated pyrex cylinder (diameter: 7 cm, length: 20 cm)
filled with 750 ml. of room temperature saline surrounded
the tendon. A thumb screw mechanism was used to apply 0.16
N pre-tension on the tendon. This was measured with a
digital multimeter.

The saline was heated at a rate of 1.5° C/min from 24°

to 95° C by a 209 watt, 1.25 cm wide resistive heating tape.

38
The tape was coated with silicone rubber to increase the

thermal Contact area and to protect from electrical shorts.
A variable a.c. transformer was used to supply power to the
heating element.

The HIT device was calibrated during a previous set of
experiments (11). During the calibration it was discovered
that the heating rate was non-linear due to limitations of
the transformer, use of a temperature sensitive load cell,
and heat loss from the device. For these reasons data
collected past 80° C. are not reported in this study.

The temperature of the saline was measured by a T—type
copper/constantan thermocouple placed near the center of the
bath with the reference electrode in a insulated container
of crushed ice. The voltage signal from the thermocouple
was amplified by a Sensotec (model SA-4) strain gauge
amplifier.

The load cell used to measure tension within the tendon
bundle was a Gould-Statham, model DC 3. As mentioned
previously, this transducer was not temperature compensated,
therefore, a plexiglass shield was placed beneath it to
deflect convected heat rising from the saline bath. In
addition, a carbon fiber rod was used in line between the
top grip and the load cell to minimize heat conduction into
the cell. Voltage readings from the load cell were
amplified by a matching strain gauge amplifier used for

temperature data collection.

39
Once the test was complete, the tendon between the

grips was removed, freeze-dried, and weighed. The dry
weight was used to normalize the load data which was plotted
against temperature data. Three value were determined from
the normalized load-temperature plots as seen in Figure 10.
The thermal shrinkage temperature (Ts) was defined as the
point at which tendon contraction occurred. The primary
(SL1) and secondary (SL2) slopes were also recorded. 'The
primary slope was measured by hand from the initial rise in
the load-temperature curve immediately after tendon
contraction. The secondary slope was measured between

approximately 70° and 80° C.

 

N/
m8

        
  
 

sI2

 

 

 

,4.

Ln'
0

n

Figure 10. Typical HIT normalized load-temperature response
of patellar tendon.

40
MORPHOLOGY

Transmission electron microscopy (TEM) was used to
examine the distribution of collagen of collagen fibril
diameters in 2.5, 4, and 9 year old diabetic and control
tendons. 1.0 cm samples of patellar tendon fixed in buffered
formalin were placed in fresh 3% glutaraldehyde in 0.1 M
phosphate buffer pH 7.4 and postfixed in 1% osmium tetroxide
in 0.1 M phosphate buffer for one hour each at room
temperature. Samples were minced further into 0.5 mm
sections and then sequentially dehydrated in 70%, 95% and
twice in 100% ethyl alcohol. Samples were then embedded in
Epon-Arldite epoxy resin in a cross section orientation and
polymerized at 65° C for 24 hours. 70-90 nm ultra thin
sections of the cross section oriented tendon were cut on a
Dupont diamond knife and supported on 200 mesh copper grids
for TEM. Thin sections were heavy metal stained with
alcoholic uranyl acetate (1% UA in 70% ETOH) and Reynolds
lead citrate. A Phillips 201 TEM at 60 kV was used to
observe and document cross sectioned samples. A
representative random photomicrograph was taken of each
diabetic and age-matched control at 20,000X magnification.

A random representative 5 x 5 inch area was selected for
each 8 x 10-3 enlarged glossy print prepared from TEM
negatives and approximate fibril diameters were digitized in
a point to point mode on a Nikon Joyce Loebl Magiscan with a

1.35 enlargement factor. The final scale factor was 81,000X

41
(20,000-3-1.35X). Statistical analysis of average fibril

diameter was carried out with Magiscan software.
STATISTICAL ANALYSIS

Standard linear regression technique was used to
determine changes in each property measured in this study as
a function of age. In most cases the control data were
sufficient in number to adequately perform the analysis.
The diabetic data were at the most nine in number and
although linear regressions were also performed on this
data, a separate analysis of variance was conducted between
the diabetic data and the age-matched controls. The
diabetic values were plotted onto the linear regression
curves of the control data for visual comparison. Changes
in a property with age were defined when the slope of the
regression line was statistically different from zero at
P<0.05. Correlation of the data to the line was determined
by the regression coefficient r which range from zero (no
correlation), to one (perfect correlation). A Students
paired t-test was conducted between the diabetic and age-
matched controls to determine whether statistically
significant differences (P<0.05) occurred between the two

sample populations.

IV} RESULTS

GEOMETRIC PROPERTIES
There were variations in the geometric features of the
patellar tendons due to the diversity of breeds and activity

levels of the animals used for mechanical testing (Table 2).

PHYSIOLOGICAL PROPERTIES

Three mechanical properties of the tendons were
determined in the estimated physiological range of loading
(90 N-180 N). Creep of the tendon was the amount of stretch
which occurred during forty cycles of pre-conditioning.
Creep in the control tendons had virtually no dependence on
age (P=0.96) (Figure 11). The correlation coefficient of
control, creep data with age was 0.01. Creep of diabetic
tendons was not significantly different than controls at
P=0.11 (Figure 11). Linear regression of this data showed

almost no correlation to the line (r=0.01), and no

significant change with age (P=0.98).

42

43
Table 2. Tendon geometric properties.

 

Controls
Age (Y) Lg (mmlel) CSA (mmleZ) V (mmle3)
0.5 3.39 0.33 1.11
0.8 3.90 0.23 0.89
1.5 4.15 0.34 1.41
1.5 4.20 0.37 1.56
2.0 3.20 0.17 0.53
2.0 1.73 0.15 0.26
2.5 4.01 0.29 1.14
2.5 2.28 0.17 0.38
3.0 3.33 0.23 0.77
4.0 2.16 0.11 0.24
4.0 3.39 0.23 0.79
5.0 3.40 0.27 0.92
5.5 3.92 0.32 1.24
6.0 3.20 0.27 0.85
6.0 3.96 0.24 0.93
7.0 3.75 0.40 1.49
7.0 3.81 0.20 0.78
9.0 4.41 0.22 0.95
9.0 3.64 0.26 0.98
10.0 2.59 0.25 0.66
10.0 2.63 0.13 0.42
11.0 3.42 0.22 0.75
11.0 3.27 0.32 0.82
12.0 3.42 0.29 0.77
.0 3.65' 0.21 0.78
.0 2.70 0.15 0.44
.0 3.98 0.25 0.98
.0 3.34 0.27 0.89
.0 3.51 0.17 0.67
Average 1 s.d. 3.39:0.64 0.24:0.07 0.84:0.33
PM
Age (Y) Lo (mmlel) CSA (mmlez) V (mmxlOé)
2.5 3.21 0.17 0.54
2.5 1.94 0.09 0.26
4.0 3.58 0.32 1.13
5.0 2.76 0.21 0.59
6.0 2.80 0.15 0.41
6.0 3.03 0.19 0.58
9.0 3.09 0.23 0.72
9.0 2.59 0.18 0.45
9.0 3.03 0.24 0.71
Average : s.d. 2.89:0.46 0.20:0.06 0.60:0.25

44

.omm mo cowuocom m on moouo

mUBMQSQ “x
aliuama<

.HH whamwm

“#23200 D

 

 

 

 

.a. v.a+xaoo.ouun I u u
.0. v.H+xuoo.o I»

 

 

 

IXHWVUIS

45
The second parameter determined from pre-conditioning

was the physiological stiffness of the PPT complex. As
Figure 12 illustrates, there was a slight downward trend in
stiffness versus age, but the slope of the regression line
was not statistically different than zero (P=0.65). There
was a weak correlation of the stiffness values to the
regression line as indicated by a low regression coefficient
of r=0.1. While no statistical difference between the
diabetics and age-matched controls was found (P=0.37), all
the diabetic tendon stiffness values fell above the
regression line of the controls, except one experimentally
induced. This indicated a tendency for greater stiffness of
the long term diabetic tendons than controls in the
physiological loading range. There was moderate correlation
of the data to the regression line (r=0.65) and a significant.

increase in the diabetic data with age (P=0.08).

46

.mmm mo :oHquSM m mm mmocmmwum Hmowmoaowmmnm .NH ousmflm

 

 

 

 

 

gl— x gab—.8 D
n§§w=u0<
an t— Nu an o o t N o
_ L . _ _ _ . L _ _ _ . _ . b O
3. coats.» I» I I I
ADV NVN+XM.HII¥
I am
0 I on—
D I an
O D
D
a x \uofi
D h\D\
C IILWII .u n
x 0‘0 \x 0 '0“
X \X D D
\\ x
a \\ x
\ D
\ \u\ D rig
\ X
\
\\
D I am"
act

 

 

 

(“DGDHSSENJflIS

47
The stiffness data were normalized by geometry to

determine age-dependent changes in the physiological modulus
(Figure 13). Like stiffness, the moduli had no significant
changes with age (P=0.65), but a more definite tendency to
increase slightly with age. There was weak correlation of
the data with the regression line (r=0.1). There was no
statistical difference between the diabetic tendon moduli
and controls (P=0.57). Although, once again, the majority
of the diabetic tendons had moduli above the regression line
for controls, suggesting that diabetic tendon material
tended to be stiffer than age-matched controls under
physiological loads. There was poor correlation of the
diabetic data to the regression line (r=0.16) and no

significant changes with age (P=0.7).

48

.omm mo adauocau n no unannoe Hmowmoaowmwnm

.mH museum

 

 

 

 

mucug3a x muznxoo a
nzgucuu<
.— m. .u o o e u a
r _ — — — L — p _ _ n p b h _ o
a I on.
o
u I cos
n. o
u x u
D D D I can
IIhPII
- . o
IIIIII+WII IIIII, x II. II .fiT..II
X III III .III D X
bII .II II .II .II I now
D
n o
x I can
.a. nmn+xu.vnw I I I o
lo. mam+xv.«nu

 

 

PHMBSflIKKNN

49
STRUCTURAL PROPIRIIES

Four types of failure mechanisms occurred in the PPT
complexes: tendon midsubstance rupture (A), bony avulsion of
the patella (B), shear fracture of the patella parallel to
the load axis (C), and tensile fracture of the patella 45°
to the load axis (D) (Figure 14). 41% of the control
tendons failed midsubstance, while 38% failed by bony
avulsion of the patella. While no significant (P=0.3)
changes occurred, there was a tendency for a shift from a
midsubstance in the young to bony avulsions in the old
control PPT complexes. The remaining 21% of the controls
failed in a shear fracture of the patella.

No midsubstance tendon failures occurred in the
diabetic complexes. 78% of the diabetics failed by fracture
of the patella along a plane 45° to the load axis. The

remaining 22% failed by bony avulsion of the patella.

50

 

 

C
U

 

 

 

 

Figure 14. Failure meChanisms.

The structural properties of the canine tendons were
plotted as a function of age. The average loads and standard
deviations of the control and diabetic complexes were
2.5010.9 kN and 2.0310.7 kN respectively. The failure load
had a slightly decreasing trend with age (P=0.36) (Figure
15). The correlation of the data with the regression line
was weak (r=0.18). When analyzed against the age-matched
controls, there was a significant decrease (P=0.03) in the
failure load of the diabetic complexes. Regression of the

diabetic data had a correlation to the line of r=0.32 and

51
P=O.41. The trend was for diabetic tissue to increase in

failure load with age versus a decreasing trend for diabetic

tissue.

52

.omm mo coauocsm m an cmoa musaflmm

.mH magmas

 

 

 

 

 

mung X 33.3.8 D
gg8¢
*— B 9 o o t N o
L L L r L L L L L L L L L L F
X
D
D D
D \I.
D n x on x DLxhuII
\ID\\
3 I\ D
a m II.IIXPIHII mfl. x
D‘ x
\
D
D
D D x D D
D D
D
0

Lo. unna+xvo I» I I I
.0. unpu+xn.nnIlu

 

 

 

00 0'01

ppm-now

53
As with the physiological stiffness of controls, tendon

stiffness prior to failure did not depend on age (P=0.86)
(Figure 16). There was also a weak correlation in this data
to the regression line (r=0.04). The average and standard
deviation values were 308166 kN/m and 288145 kN/m for the
control and diabetic tendons, respectively. The age-matched
controls were stiffer than the diabetic tendons (P=0.05),
which differs from a slight increase noted in the
physiological stiffness of diabetic tendons versus controls.
Correlation of the diabetic data to their regression line

was weak (r=0.03) and no changes occurred with age (P=0.93).

54

.mwm mo cofluocsm m mm mmmcmmwum

.mH magmas

MUENQSO “x “#33260 D

. naameuu<
.— N. L: o o e
F

N

 

“Ow HON+NU.OIIN I I I
«0v NHM+xn.OIIN

 

 

 

 

 

NWGDflSSHNJfllS

55
I Failure strain energy of controls had a non-significant

(P=O.11) decreasing trend with age (Figure 17). There was
better_correlation of this data with the regression line
than parameters discussed previously (r=0.3). The average
and standard deviation values of the control and diabetic
complexes were 16112 (J) and 1217 (J), respectively. When
analyzed with the age-matched controls, no difference was
found (P-O.15) in the energy of the complexes caused by
diabetes. However, most of the diabetic values fell below
the regression line for controls. Correlation of the
diabetic data to their regression line was poor (r=0.2) and

no changes occurred with age (P=0.6).

56

.omm mo aofiuocam m an hmumcm cannum muaaamm .ba munmfim

mug x “.5528 D
2E8¢

 

 

 

x
u n
a I an
a n.
I 9
I am
3. Q38... I» I I I
.2 «93861..» a

 

 

 

VI was":

57
MATERIAL PROPERTIES

The strain at failure was one of four material
properties evaluated as a function of age. Only
midsubstance failures were analyzed to determine material
properties of the tendon. A non-significant (P>0.05)
decrease in failure strain occurred with age (Figure 18).
There was very weak correlation of the data with the
regression line with r=0.04. The average and standard
deviation of the failure strain values was 33.418.5 (%). No
comparison of the control and diabetic strain values was

conducted because there were no midsubstance failures in the

diabetic tendons.

58

.mmm mo cowuoanu m mm cwmuum muaawmm

BHiNch<
Nu .8 o o v
L

.mL mupmfim

 

 

 

a.mm+xmo.OIIM

 

 

IXHMVIMS

59
Tensile strength of a complex was defined as the

maximum stress achieved before rupture. Like the failure
strain, the tensile strength was documented from
midsubstance failures only. As illustrated in Figure 19,
the tensile strength did not change significantly (P>0.05)
with increasing animal age, but a downward trend was noted.
Correlation of the data to the regression line was moderate
(r=0.37). The average and standard deviation of the control
tendons was 111.2136.2 (MPa). Once again, no diabetic
values were documented due to a lack of midsubstance

failures.

60

.omm no cowuucau m on numcmuum «Hamcma .mH muzmam

maiwcmu<

 

 

NMH+Nu.nIIH

 

 

 

88883388§882883889°

PHMISSSUIS

61
The slope taken from the linear portion of the stress-

strain curve was defined previously as the tensile modulus of
the tendon. This was the only material property which had an
increasing but non-significant (P=0.36) trend with age
(Figure 20). Correlation of the moduli data to the
regression line was weak (r=0.18). Diabetic moduli were not
statistically different than the age-matched controls
(P=0.47). Correlation to the diabetic regression line was
moderate (r=0.44) and no statistically significant changes
occurred with age (P=0.24). Yet there was a tendency for the
diabetic tendon modulus to decrease with age, in contrast to
the controls. Average and standard deviation values of the
control and diabetic tendons were 4531126 (MPa) and 454i144

(MPa), respectively.

62

.omm mo cofluocam m mm msasooe oaflmcoa

.om museum

 

 

 

 

 

 

 

«gunman x “wxxnzou D
fizqmcmu< .
e— «L as o u .v N
L L L L L L L L L L L L L L L
a
v / D D
II .II x a
II
In /
le a
II x
D D
D x / Lu
3 IVW/ D D C
n x
, C D / In D .
n 0 Lu
0 D x II
D II
D D j
a a
La. vmm+xvuI|a I I I
8. «.338 I» x

 

 

I'dWI 801000“

63
The last material property determined was the area under

the stress-strain curve called the failure strain energy
density. Only values from midsubstance failures were used,
and therefore comparison with the diabetic complexes was not
conducted. Like the strain energy there was a non-
significant (P>0.05) decrease with age with a moderate
correlation of the data to the regression line (r=0.31)
(Figure 21). This was expected due to the decreasing trend

in both the failure strain and tensile strength with age.

64

.omm mo coauocsu m mm >uwmcmv amumco cwmuum muzaflmm .HN muzmam

38¢

 

 

 

N . @933 . clan»

 

 

Cfl/fl 11188130 19m

65
A slight decrease in modulus prior to failure for the

diabetic complexes versus aged-matched controls was
illustrated in a composite curve of these data (Figure 22).
The curve represents the mean data at every 2% strain for
controls and diabetics. It is shown here without standard
deviations to indicate the qualitative differences between
the stress-strain responses of the diabetic and age-matched

control complexes.

 

 

 

 

no L
“-1
m4 D
D
II- °
00- a +
a '6
”-1 4
I!“
.J
i no n+
TO- DO
80- 0+
Q
so- 9
40- 9
a- I
”d .
I
o-‘U'III'UIUTUIIUV
oaqlluuunlmauaaxa
mm
comma +nlmu

Figure 22. Composite stress-strain curve.

66
BIOCHEMICAL PROPERTIES

Three fractions were assayed from tissue aliquots of
diabetic and control tendon. Total collagen content was also
determined in separate aliquots. Small and sometimes non-
existent traces of acid soluble fractions were found,
indicating that there was very small amounts of newly formed
collagen within these tendons. These data were not reported
further in this document. Collagen concentrations soluble
and insoluble to pepsin, as well as total collagen content,
were determined and evaluated as a function of age and
diabetes.

The pepsin digestable fraction of collagen was an
indicator of the quantity of collagen which was more recently
synthesized. It was obvious (as shown in Figure 23) that the
solubility of collagen to pepsin decreased significantly
(P<0.05) with age, and there was a strong correlation of the
data to the regression line (r=.74). While there was no
statistically significant difference (P=0.5) between the
diabetics and controls, diabetic collagen appeared to become

less soluble at a faster rate than controls.

67

t— H. Li
\L

MUEMMSE x"
fi8<wtmfl=‘

.mmm mo cowuocau m mm cmmmgou no coauomuu 9333 980mm

«#33200 D

 

 

p
0

3L 3.38.7.5 I I I
8. vn+anII~

 

 

 

 

.MN madman

(.III/h) NSDY'I'IOO

68 .
There was a significant increase in fraction'of

collagen insoluble to pepsin with age (P<0.05) along with
good correlation of the data to the regression line (r=0.62)
(Figure 24). There was no significant difference (P=0.2)
between the diabetic and control tendons, however, all but
one experimental diabetic fell above the regression line of
the controls. This suggests that collagen from naturally
occurring diabetic tendons may be more cross-linked than

controls.

69

wow mo coauocnm m mm commaaoo mo cofiuomum mannaomcH .vN ousmwm

 

 

8.593 x 30528 a,
gfieflfi
fl ... _ u 9 o u v u o
L L L L L L L L L L . p L L F L 0
EL 9938.?» I I I I 9
Lo. m.p¢+xm.nn»
I 8

 

 

 

 

{PW/ml NSSTI'IOO

70
The concentration of total collagen measured in the

tendons had no changes with age (P=O.5) or diabetes (P=O.4)
(Figure 25). This indicates that the decreased solubility in
pepsin of the tendons is compensated by an increased
insoluble fraction resulting in little or no change in total

collagen content of patellar tendons with increasing age.

71

.omm no coauocnm m on ucmucoo commaaoo Hmuoa .mN ouomfim

MOBMQSG x ”#23200 D
n§<mcu01

 

8L Sinnéiu

 

 

 

 

D
D
fisaaiassessesaao

 

(WW/QMIKMMOVUJ

72
THERMAL SHRINKAGE PROPERTIES

A limited number of tendon samples were used to obtain
information on the thermal stability of collagen. Linear
regression analyses were performed on the thermal shrinkage
temperatures of control and diabetic tendon (Figures 26 &
27). The analysis of control tendon temperatures showed a
significant (P=0.004) decrease occurred with age. Visual
inspection of Figure 26 illustrates that there may have been
less change in the shrinkage temperature with age than was
statistically determined. There was moderate correlation of
the data to the regression line (r=0.53). Diabetic tendon,
on the other hand, had significantly (P=0.05) increasing
shrinkage temperature with age, and also had moderate
correlation to the regression line (r=0.50). The above
changes with age resulted in a significant difference

(P=0.00) between diabetic and age-matched control tendon.

73

.mmm mo coauocsu

m mm as cocoon Houucoo

WGSECUU(

.mm musmfim

 

 

m.nm+xoo.OInH

 

 

 

KMEMHIVHIQNEL

74

.omm no cowuucsu m on ma cocoon owumnmwo

fil<fl=ufl<

.pm muamfim

 

v.nu+xd.olfl

 

 

 

 

KNEMHIVHIINEL

75
During the HIT preparation it was difficult to separate

the tendon into individual fascicles. As a result a larger
then desired strip of tissue was often tested. While there
was no significant dependence of the primary (Sll, P>O.2) or
secondary ($12, P>O.2) slopes on either diabetes or age
(Table 3), some trends were observed in the data. The
experimental diabetic (2.5 year old) had lower or more
gradual secondary slopes than controls (P<0.05). Four and
five year old naturally occurring diabetics had nearly equal
slopes as controls. Two tailed paired t-tests were used to
determine that no significant differences occurred between

diabetic and control slopes (P>0.05).

76

Table 3. Thermal shrinkage slope data.

*CONTROL *DIABETIC
Mill 5.12. .511 ill;
2.5 9.12 1.40 2.00 0.96
2.5 4.08 1.17 2.11 0.85
2.5 1.59 1.17
2.5 2.08 1.61
2.5 2.44 0.93
2.5 1.92 1.19
4.0 1.73 1.50 1.67 1.68
4.0 1.48 1.18 1.54 1.21
5.0 2.10 1.15 1.61 1.20
5.0 2.08 1.32 2.94 1.29
5.0 1.21 1.13
5.5 2.00 1.52
5.5 1.80 1.46
5.5 1.53 1.00
6.0 2.36 1.14 1.67 1.05
6.0 0.59 0.36 1.85 1.16
6.0 3.23 0.86 2.49 1.49
6.0 1.75 1.21
9.0 3.23 0.96 1.75 1.21
9.0 2.08 0.85 1.97 0.99
9.0 1.43 0.79 1.97 0.91
9.0 3.68 1.14 1.79 1.16
9.0 2.64 0.85

15.0 2.14 0.99
15.0 1.85 1.30

* All slope values are times 10‘2 (N/mg oC)

MORPHOLOGY

Photographs from a transmission electron microscope were
used to examine collagen fibrils within 2.5, 4, and 9 year
old control and diabetic tendons (Figure 28). Computerized
digitization of these photographs was utilized to quantify
fibril diameters. Frequency histograms illustrated the
distribution of fibril diameters within a specimen.

The younger control tendon had a bimodal distribution of

large and small diameter fibers which shifted to more uniform

and smaller average diameter fibrils in the old tendon

77
(Figure 29). The 2.5 and 4 year old control tendons had

larger average fibril diameters than the 9 year old specimen.
While there was no statistically significant (P>0.05)
difference between the average diameter of the 2.5 year old
control and diabetic tendon, there was a clearly defined
grouping of smaller diameter fibrils in the diabetic sample.
The 4 year old diabetic tendon was significantly (P<0.05)
smaller than the control and also had an obvious grouping of
smaller fibril diameters. There was no significant
difference between the average fiber diameter of the 9 year
old control and diabetic tendon. It was interesting that the
trend for a larger group of small diameter fibrils in

diabetics versus controls was not seen at nine years of age.

78

 

 

 

Duvet-Ila . 6333355

 

 

CI
”-1 l
a d

.1 IL!

 

 

 

 

miles-mu. -Imm

 

 

 

:3 M1 Hunts.

a "0m -' 5m ' mum

Figure 28. Fibril diameter histograms.

 

Figure 29.

 

CONTROL

E; 1":zcrow é‘g:c. Ig.
:‘31;%.: 3&2: “‘3‘? .6 :?3& I
I. . '5 6:235. «: 33:7. .
I’8215’o.3::£92‘;r £332 :03; I,

I. .9 3"“ ' g. g;‘ 9;?
”93:36:” ”g '1' :9: :fiifs $333

I.
2.“ ' e I‘ b '

",1. a. «1..., u . 65‘- ‘-
9“ . .1 .3336 u 0....
I53... mm;

D 3964. .B‘ ‘n‘h’! 52933.1

 

DIABETIC (EXPERIMENTAL)

Fibril diameter TEM - 2.5 year olds.

‘V. DISCUSSION

The objective of this study was to document alterations
in the mechanical, biochemical, and thermal properties of the
patellar tendon with age and diabetes using a canine model.
Information of this nature is important because of the
widespread use of patellar tendon in ligament reconstruction
procedures and use of the canine model in diabetes research,
as it relates to alterations in the physical properties of
musculoskeletal connective tissues.

No particular dependence on age was noted in the
geometric properties of the canine PT. This was probably due
to the diversity of breeds used in this study. Since these
were patients from the Veterinary Clinical Center, activity
levels of the dogs were not known and could contribute to
wide variations in the dimensions of the patellar tendons.

The cyclic testing was designed to measure properties of
the tendon under physiological loading. The inclusive angle
in the test fixture of 1100 helped simulate the normal
configuration in the canine stifle joint. The cyclic load
range was estimated to be those in the PT while walking.

This resulted in a slight stretch of the tendon. Hubbard et
al. (21) examined several different canine tendons and found

a rapid change in strain during the first 12 seconds of

80

81
cyclic loading which resulted in a measurable amount of

hysteresis during that time. Little information is currently
available on cyclic or pre-conditioning of tendon as a
function of age. Woo et al. (47) collected data on cyclic
loading of rabbit knee ligament and found an increase in the
area of hysteresis during maturation. Comparable information
was not collected during this study. Diabetes and age caused
little alteration in the behavior of canine tendon during the
initial loading phase as indicated by the constant percentage
of creep in both tissue types. The slight increase in
physiological stiffness and modulus of the diabetic tendon
over controls is consistent with pathologies reported from
human diabetic joints. Limited joint mobility is common in
diabetes (31).

The large percentage of avulsion failures in this study
was different from the numerous midsubstance failures of
bone-ACL-bone preparations reported by Vasseur, et al., (42).
Assuming that the strength of bone degenerates with age, the
slight decrease in failure load may result from a greater
occurrence of avulsion fractures at the patella with specimen
age. This finding agreed with Grood, et al., (15) in which
more avulsion fractures occurred in the human bone-ACL-bone
preparation with advancing age. In an earlier study (20),
tensile failure tests on patella-PT-tibia preparations
exhibited greater occurrence of bone fracture and avulsions-
for more aged specimens, especially after gamma irradiation

sterilization.

82
In cases-of midsubstance failure only, very slight

decreases occurred in tensile strength and failure strain
energy density of the PT with age. The trend with age
compares to Vasseur, et al., (42) for the canine ACL,
although to a much lesser extent. In contrast to Vasseur’s
findings on the ACL, however, the tensile modulus of the PT
remains nearly constant or has a slight rise with age. These
results are better matched to those of Vasseur for the canine
caudate and collateral ligaments. It has been suggested that
these ligaments may exhibit essentially no degenerative
changes because of a superior blood supply. The extra-
articular location of the collateral ligaments may provide
significant blood supply from genicular collateral
circulation and help provide a mechanism of repair from minor
disruptions associated with stresses of normal weight
bearing. Canine collateral ligaments may also be preserved
because of minimal stress during normal ambulation (27).
While actual in vivo loads are not known, physiological
forces in the patellar tendon may in fact exceed those within
the ACL, yet this results in minimal degeneration of the
tendon. The data on ageerelated changes in the canine
patellar tendon follow closely those of Hubbard and Little
(22) for human tendons. Interestingly, the data from
Vasseur, et al., (42) on the canine ACL parallel those of
Grood, et al., (15) for the human ACL.

The tensile fracture of the patella in the majority of

diabetic PPT complexes indicates an inherent weakness in the

83
tensile strength of diabetic bone. This supports earlier

studies by Einhorn (13) and Curcione (9) where both
experimental and naturally occurring diabetes induced
osteopenia and resulted in decreased structural integrity of
rat and canine femurs, respectively. It appears that
degeneration occurs more rapidly in the more metabolically
active bone than tendon. The decreased strength may have
resulted from a greater occurrence of bony avulsion failures
which in turn caused the decreased failure load in the
diabetic PPT complexes.

Alterations different from normal aging have occurred in
connective tissue of juvenile diabetic canines. While the
material properties of control tendons that failed
midsubstance had no changes with age, the tensile modulus of
the diabetic tendon just prior to failure appeared to be
slightly reduced over age-matched controls. This may, in
fact, be a consequence of an incipid failure of underlying
bone and a larger percentage of small diameter collagen
fibrils in the diabetic tendon. This tends to contradict
Andreassen, et al., (5) who reported increased strength in
experimentally induced diabetic rat tail tendon after 30 days
of disease. Yet, like Andreassen, the current study did show
slightly increased physiological stiffness and modulus with
diabetes. The current study, however, is in contradiction
with Galeski, et al., (14) who reported increased collagen
fibril diameters and strength with experimentally induced

diabetes.

84
It is not known exactly what alterations occur in

tissues of animals during the onset of diabetes. Perhaps
chemical changes occur within the fibroblast cell, such as
excessive hydroxylation, that alters the physical make-up of
the collagen chains before expulsion into the extracellular
medium. Differences noted in the mechanical and thermal
stability of experimentally induced diabetic animal models,
and the juvenile diabetics examined in this study, may be due
to several factors. Studies involving experimentally induced
models usually begin with animals which have normal
intracellular and extracellular collagen formation. It is
not known what alterations in the collagen formation may have
already occurred at that point in time, within an age-matched
juvenile diabetic animal. In addition, most experimental
studies are short term, usually less than six months.
Perhaps long term experimental diabetic models would exhibit
alterations in tissue properties that more closely match
those of the juvenile diabetic. As few differences were
noted between the 2.5 year old experimentally induced
diabetics and the older juvenile diabetics in the current
study, perhaps two years in the hyperglycemic state was
enough time to produce comparable degenerative changes.
Biophysical (thermal) and biochemical studies on
collagen, the major strength-bearing component of all
connective tissues, were performed to help understand
mechanisms of age-related change. The documented changes

with age of various physical and chemical properties of

85
collagenous tissues (e.g., decrease in solubility) have been

attributed to alterations in the type of covalent cross-
linking present in collagen. An interesting result from this
study was a moderate decrease in shrinkage temperature of the
canine PT with age. Earlier studies on rat tail and human
wrist tendon (29), as well as rat skin (1), indicate an
increasing thermal stability in the collagen as represented
by an increased maximum contraction force and decreased
relaxation after that force was reached. This resulted in
alterations in the slope of the corresponding isometric
tension-temperature curve with age. The shape of the curve
did not change significantly with age in the current study.
Yet, the slight increase in slope noted in the diabetics with
age parallels the decrease in solubility to pepsin, both of
which indicates increased mechanical stability over age-
matched controls. This is supported by the increased
physiological modulus and stiffness of the diabetic tendons
over age-matched controls. At the same time there was
clearly an increased thermal shrinkage temperature with age
for diabetics, in contrast to controls. Controversy exists
over the relationship of the amount and nature of covalent
cross-linking and the shrinkage temperature of collagen.
While the changes in the biochemical nature of human and rat
tail tendons (29), rat skin (1), and canine patellar tendons
are similar in nature; shrinkage temperatures increased in
human tendons and rat tail tendons, remained constant in rat

skin, and moderately decreased in canine tendon with age.

86
Earlier studies have shown that in animals treated with D-

penicillamine strips of dorsal skin lose contractile ability
(18) due to inhibition in the formation of new intramolecular
cross-links and disruption of pre-existing cross-links. In
vitro exposure of the skin to the cross-linking agent
glutaraldehyde restores shrinkage capacity. Increased
concentrations of the fixative lead to an increased shrinkage
temperature (18). The slight decrease in shrinkage
temperature of the canine patellar tendon with age suggested
a decrease in cross-link integrity or density. This may be
supported by the decreasing trend in the tensile strength
with age. Yet, the amount of collagen insoluble to the
enzyme pepsin was increased with specimen age. The results
suggest that during normal aging of the canine patellar
tendon there was either a decrease in the ratio of thermally
stable to labile cross-links that does not parallel changes
in pepsin solubility, or alterations occurred in the ordered
crystalline structure of the collagen. While increases in
the content of insoluble collagen increase with increased
tensile strength (45) and probably increased modulus, the
modulus of the canine patellar tendons may have remained .
constant with age because of offsetting changes in integrity
of collagen as evidenced by a slight reduction in its
biophysical stability.

Biochemically the diabetic tendon resembled advancing
age with a more rapid increase in collagen insoluble to

pepsin than controls. The small fractions of acid soluble

87
collagen indicates a lack of newly synthesized collagen in

both the diabetic and control tendons. Total collagen
concentration in the the diabetics was the same as controls.
Biophysically the changes caused by diabetes were somewhat
opposite to those of advancing age. A significantly
increasing shrinkage temperature in the juvenile diabetic
tendon indicated an increasing ratio of thermally stable to
labile cross-links, as opposed to the previously mentioned
decreasing ratio in controls. Perhaps hyperglycosylation
induced by hyperglycemia and continued fluctuations in and
out of dehydration during insulin therapy have an as yet
unknown influence on the structural integrity of collagen.
One implication of the data from the present study is:
since the PT did not degenerate significantly with age,
autogeneic and allogeneic grafts from a variety of age groups
may provide a acceptably stiff tissue for reconstruction of a
torn or otherwise damaged ACL. The assumption, of course, is
that age-related changes in the human PT parallel those
documented here in the canine PT. The limited data of Haut,
et al., (20) would tend to support this statement. On the
other hand, these data also suggest relatively less tissue
may be needed for reconstruction during advancing age to
match changes in the aging ACL. These data also provide
information on the mechanical properties of the canine PT to
compare with earlier studies on the ACL by Vasseur (42).
Differences in the aging properties of these two tissues

parallel differences noted by Vasseur et al. for extra versus

88
intra-articular tissues. The data have also provided new

information on correlations of mechanical properties with
biophysical and biochemical alterations in collagen. On this
issue even more research is suggested.

A second implication from this study is that diabetes
mellitus does in fact alter the mechanical, biochemical, and
thermal stability of the canine patellar tendon, but maybe
not to the extent documented for bone tissue (9,13).' This
has been likely supported in the current investigation
because of the increased incidence of bone fracture in
diabetic animals.

Clearly, future research needs to be directed at a better
understanding of the characteristics of cross-linking in
collagen and correlations with mechanical properties of
various connective tissues. Future studies are needed to
more fully understand these biophysical, biochemical, and
biomechanical properties of collagen in the canine patellar

tendon.

 

. REFERENCES

10.

11.

12.

89
REFERENCES

Allain, J.C.; Le Lous, M.: Bazin, 8.; Bailey, A.J.: and

Delaunay, A.: Isometric Tension Developed During Heating
of Collagenous Tissues. Biochimica et Biophysica Acta.

533:147-155, 1978.

Alm, A. and Gillquist, J.: Reconstruction of the
Anterior Cruciate Ligament by Using the Medial Third of
the Patellar Ligament. Acta Chir. Scand. 140:289-296,
1974.

Andreassen, T.T. and Oxland, H.: Changes in collagen and
elastin of the rat aorta induced by experimental
diabetes and food restriction. Acta Endocrinologica
(Copenh). 115:338-344, 1987.

Andreassen, T.T. and Oxland, H.: The Influence of
Experimental Diabetes and Insulin Treatments on the
Biomechanical Properties of Rat Skin Incisional Wounds.
Acta. Chir. Scand. 153:405-409, 1987.

Andreassen, T.T.; Seyer-Hansen, K.; and Bailey, A.J.:
Thermal Stability, Mechanical Properties and Reducible
Cross-Links of Rat Tail Tendon in Experimental Diabetes.
Biochim. et Biophys. 677:313-317, 1981.

Bailey, A.J.: Robins, S.P.; and Balian, G.: Biological
significance of the intermolecular crosslinks of
collagen. Nature. 251:105-108, 1974.

Burks, R.T.; Haut, R.C.; and Lancaster, R.L.:
Biomechanical and Histological Observations on the Dog
Patellar Tendon After Removal of Its Central One-third.
Amer. J. Sports Medicine. 18(2):146-153, 1990.

Cerami, A.: Vlassara, H.; and Brownlee, M.: Glucose and
Aging. Scientific American. 9:90-96, 1987.

Curcione, P.J. and Haut, R.C.: Structural and Material
Properties of the Canine Femur in Juvenile Onset
Diabetes Mellitus. NIH report, Biomedical Student
Research Program, 1989.

Danielson, C.C.: Difference in Thermal Stability of
Type-land Type-III Collagen from Rat Skin. Biochem. J.
203:323-326, 1982.

DeDeyne, P.: The Effects of Gamma Irradiation on
Patellar tendon Allografts. Masters thesis, Michigan
State University, 1989.

Dilman, V.D.: Diseases of Aging. Boston, MA: John Wright
PSG Inc, 1981

l3.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

90

Einhorn, T.A.; Boskey, A.L.; Gundberg, C.M.; Vigorita,
V.J.; Devlin, V.J.; and Beyer, M.M.: The Mineral and
Mechanical Properties of Bone in Chronic Experimental
Diabetes. J. Orth. Res. 6:317-323, 1988.

Galeski, A.: Kastelic, J.: Baer, E.; and Kohn, R.R.:
Mechanical and Structural Changes in Rat Tail Tendon
Induced by Alloxan Diabetes and Aging. J. Biomechanics
10:775-782, 1977.

Grood, E.S.; Noyes, F.R.; and Butler, D.L.: Age Related
Changes in the Mechanical Properties of Knee Ligaments.
Biomechanics Symposium, ASME, AMD-Vol. 23, 1977.

Hamlin, C.R. and Kohn, R.R.: Evidence for Progressive,
Age-related Structural Changes in Post-mature Human
Collagen. Biochim. Biophys. Acta. 236:458-467, 1971.

Hamlin, C.R.; Kohn, R.R.: and Luschin, J.H.: Apparent
Accelerated Aging of Human Collagen in Diabetes
Mellitus. Diabetes, Vol. 24, No. 10:902-904, 1975.

Hardy, P.M.; Nicholls, A.C.; and Rydon, H.N. The Nature
of the Cross—linking of Proteins by Glutaraldehyde. J.
Chem. Soc. Perkin Trans. 1:958, 1976.

Haut, R.C.: Age-dependent Influence of Strain Rate on
the Tensile Failure of Rat-tail Tendon. J. Biomech.
Eng. 105:296-299, 1983.

Haut, R.C.: Powlison; Rutherford, G.N.; and Kateley,
J.R.: Some Effects of Donor Age and Sex on the
Mechanical Properties of Patellar Tendon Graft Tissues.
ASME Adv. in Bioeng, pp. 75-78, 1988.

Hubbard, R.P. and Chun, K.J.: Mechanical Responses of
Tendons to Repeated Extensions and Wait Periods. J.
Biomech. Eng. 110:11-19, 1988.

Hubbard, R.P. and Soutas-Little, R.W.: Mechanical
Properties of Human Tendon and Their Age Dependence.
J. Biomech. Eng. 106:144-150, 1984.

Jones, K.G.: Reconstruction of the Anterior Cruciate
Ligament Using the Central One-third of the Patellar
Ligament. J. Bone Joint Surg. 45Az925-932, 1963.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

91

Joseph, K.T. and Bose, S.M.: Influence of Biological
Ageing on the Stability of Skin Collagen in Albino Rats.
Collagen, John Wiley and Sons, New York, NY, 1962.

Kohn, R.R. and Hensse, S.: Abnormal Collagen in Cultures
of Fibroblasts From Human Beings with Diabetes Mellitus.
Biochemical and Biophysical Res. Communications 76:765-
771, 1977.

Le Lous, M.: Allain, J.C.; Cohen-Solal, L.; and
Maroteaux, P.: Hydrothermal Isometric Tension Curves
from Different Connective Tissues. Role of Collagen
Genetic Types and Noncollagenous Componants. Connective
Tissue Res. 11:199-206, 1983.

Lewis, J.L. and Shybut, G.T. In vivo Forces in the
Collateral Ligaments of Canine Knees. Trans. Orthop.
Res. Soc. 6:4, 1981.

McFarland, E.G.; Morrey, B.F.; An, K.N.; and Wood, M.B.:
The relationship of vascularity and water content to
tensile strength in a patellar tendon replacement of the
anterior cruciate ligament in dogs. The American
Journal of Sports Medicine Vol. 14, No. 6:436-448, 1986.

Mitchell, T.W. and Rigby, B.J.: In Vivo and In Vitro
Aging of Collagen Examined Using an Isometric Melting
Technique. Biochim. Biophys. Acta. 393:531-541, 1975.

Romen, W.; Lange, H.-W.; Hempel, K.; and Heck, T.:
Studies on Collagen Metabolism in Rats. Virchows Arch
[Cell Pathol] 36:313-320, 1981.

Rosenbloom, A.L.; Silverstein, J.H.: Lezotte, D.C.;
Richardson, K.; and McCallum, M.: Joint Mobility in
Childhood Diabetes. The New England Journal of Medicine
Vol. 305, No. 4:191-194, July 23, 1981.

Schnider, S.L. and Kohn, R.R.: Effects of Age and
Diabetes Mellitus on the Solubility and Nonenzymatic
Glucosylation of Human Skin Collagen. J. Clin. Invest.
67:1630-1635, 1981.

Schnider, S.L. and Kohn, R.R.: Glucosylation of Human
Collagen in Aging and Diabetes Mellitus. J. Clin.
Invest. 66:1179-1181, 1980.

Shagan, B.P.: Is Diabetes a Model for Aging? Medical
Clinics of North America Vol. 60, No. 6:1209-1211, 1976.

Shigematsu, Y.; Sweeley, C.C.: Schall, W.D.; and
Gossain, V.: Urinary Organic Acids in Natural Early-
Onset Insulin-Dependent Diabetic Dogs. Acta Paediatr.
Jpn. Vol. 30, No. 3:285-293, 3 June, 1988.

 

36.

37.

38.

39.

40.

41.

42.

43.

44.

45.

46.

92

Smith, 0.; Wong, C.Y.; and Gelfand, R.A.: Skeletal
Muscle Proteolysis in Rats With Acute Streptozocin-
Induced Diabetes. Diabetes 38:1117-1122, Sept. 1989.

Standefer, J.O. and Eaton, R.P.: Evaluation of a Method
for Colorimetric Determination of Glycosylated
Hemoglobin. Clin. Chem. 29:135-140, 1983.

Stegemann, H.: Mikro-vestimmung von Hydroxyproline mit
Chloramin-Tund p-Dimethylamino Benzaldehyd. Physiol.

.Chem. 311(1):41-45, 1958.

Sternberg, M.; Cohen-Forterre, L.; and Peyroux, J.:
Connective Tissue in Diabetes Mellitus: Biochemical
Alterations of the Intercellular Matrix with Special
Reference to Proteoglycans, Collagens and Basement
Membranes. Diabete & Metabolisme 11:27-50, 1985.

Torp, S.; Arridge, R.G.C.; Armeniades, C.D.; and Baer,
E.: Structure Property Relationships in Tendon as a
Function of Age. Colston Papers 26:197-221, 1975.

Uitto, J.; Perejda, A.J.; Grant, G.A.; Rowold, E.A.,
Kilo, C.; and Williamson, J.R.: Glycosylation of Human
Glomerular Basement Membrane Collagen: Increased Content
of Hexose in Ketoamine Linkage and Unaltered
Hydroxylysine-O-Glycosides in Patients with Diabetes.
Connective Tissue Research 10:287-296, 1982.

Vasseur, P.B.; Pool, R.R.; Arnoczky, S.P.; and Lau,
R.E.: Correlative Biomechanical and Histologic Study of
the Cranial Cruciate Ligament in Dogs. Am. J. Vet. Res.
Vol. 46, No. 9:1842-1854, Sept., 1985.

Vera, M.; Shumkov, G.; and Guell, R.: Histological and
Histochemical Skin Changes in Insulin-Dependent Diabetic
Patients with and without Limited Joint Mobility. Acta
diabetol. lat. 24:101-108, 1987.

Vishwanath, V.; Frank, K.E.; Elmets, C.A.; Dauchot,
P.J.; and Monnier, V.M.: Glycation of Skin Collagen in
Type 1 Diabetes Mellitus. Diabetes 35:916-921, Aug.,
1986.

Vogel, H.G.:: Correlation Between Tensile Strength and
Collagen Content in Rat Skin. Effect of Age and
Cortisol Treatment. Connective Tissue Res. 2:177-182,
1974.

Vogel, H.G.: Influence of Maturation and Age on
Mechanical and Biochemical Parameters of Connective
Tissue of Various Organs in the Rat. Connective Tissue
Research 6:161-166, 1978.

 

46.

47.

93

Vogel, H.G.: Influence of Maturation and Age on
Mechanical and Biochemical Parameters of Connective
Tissue of Various Organs in the Rat. Connective Tissue
Research 6:161-166, 1978.

Woo, S.L-Y.; Orlando, C.A.; Gomez, M.A.; Frank, C.B.;
and Akeson, W.H.: Tensile Properties of the Medial
Collateral Ligament as a Function of Age. J. Orth.
Res. 4:133-141, 1986.

IE5

"111111111111