Biomedical Research 27 (5) 233-241, 2006
Effect of heat on synthesis of gelatinases and pro-infl ammatory cytokines in
equine tendinocytes
Yoshinao HOSAKA
1, Sachiko OZOE
1, Rikio KIRISAWA
2, Hiromi UEDA
1, Kazushige TAKEHANA
1 and Mamoru
YAMAGUCHI
1 Laboratory of Veterinary Anatomy, Department of Biosciences and
2 Laboratory of Veterinary Microbiology, Department of Pathobiolo-
gy, Rakuno Gakuen University, Hokkaido 069-6501, Japan and
3 Department of Biosciences, School of Veterinary Medicine, Ohio State
University, Columbus OH43210, USA
(Received 21 July 2006; and 30 August 2006)
ABSTRACT
The aim of this study was to clarify whether matrix metalloproteinases (MMP-2 and -9: gelatinas-
es) and pro-infl ammatory cytokines [tumor necrosis factor (TNF) α and interleukin (IL)-1β] are
induced by heat in tendon tissue in vitro and to test the hypothesis that heat exposure causes ten-
dinocytes to synthesize pro-infl ammatory cytokines and that synthesis of these cytokines, in turn,
leads to up-regulation of synthesis of gelatinases. Isolated tendinocytes from equine superfi cial
digital fl exor tendons were cultured and all experiments were performed on cells passaged 3 or 4
times. In the cells exposed to heat (37 to 45°C, 0 to 60 min), the survival rate decreased sharply
in a temperature- and time-dependent manner, especially at 42 and 45°C. Cells exposed at 40°C,
however, showed little change in survival rate and morphology. Gelatin zymograms revealed that
proMMP-2 and -9 were the only two MMPs remaining in the supernatant of the cultured tendino-
cytes, including that of untreated cells. Addition of TNFα and IL-1β to the culture medium of ten-
dinocytes accelerated proMMP-9 synthesis considerably. Heating the tendinocytes (40°C) led to a
three-fold increase in proMMP-9 synthesis in a short time. Only TNFα was detected in tendino-
cytes after heat exposure for 30 and 60 min. In contrast, IL-1β was under the detectable level in
ELISA. Cooling of heat-exposed cells from 40°C to 37°C considerably down-regulated cellular
proMMP-9 synthesis. Furthermore, proMMP-9 level was greatly reduced in cells treated at lower
temperatures, 20°C and 5°C. These fi ndings support our hypothesis that hyperthermia in the horse
tendon induces tendinocytes to synthesize pro-infl ammatory cytokines and that the synthesis of
these cytokines results in the up-regulation of gelatinases.
Tendon injury, especially in the superfi cial digital
flexor tendon (SDFT), has proved to be a major
problem for racehorses. Recent studies have shown
that about 10–30% of racehorses suffer from ten-
donitis (29). Although the etiology of tendonitis has
been discussed in many reports (2, 19, 25), there is
little information on the mechanism of tendon deg-
radation. Pro-infl ammatory cytokines, such as tumor
necrosis factor (TNF) α and interleukin (IL)-1β, and
gelatinases in matrix metalloproteinases (MMPs) are
involved in degradation of connective tissues (30,
33, 34, 40), and these pro-infl ammatory cytokines
and gelatinases are synthesized in tendinocytes (15,
18, 33). Additionally, TNF α and IL-1β are known
to be initiators of the synthesis of gelatinases and of
the synthesis and release of cytokines (3, 7, 8, 10,
24, 28).
Address correspondence to: Dr. Yoshinao Hosaka
Laboratory of Veterinary Anatomy, Department of Bio-
sciences, School of Veterinary Medicine, Rakuno
Gakuen University, Ebetsu, Hokkaido 069-8501, Japan
Tel: +81-11-388-4866, Fax: +81-11-387-5890
E-mail: hosap@rakuno.ac.jp
Y. Hosaka et al.
234
ed by the outgrowth method. Isolated cells were
cultured at 37°C in 5% CO2 in Dulbecco’s modifi ed
Eagle’s medium (Sigma, St. Louis, MO, USA) with
5% fetal bovine serum (Gibco, Carlsbad, CA, USA)
and antibiotics (Sigma). Isolated cells had the char-
acteristic appearance of fibroblasts; cell processes
protruded in a star-like shape in sparse cell cultures,
and as semi-confl uence was reached cells became
spindle-shaped, producing a parallel array (Fig. 1a).
All experiments were performed on cells from cul-
tures that had undergone 3 or 4 passages (P3 or 4).
Moreover, to test the tendinocytic character, te-
nomodulin, a specifi c marker of tendon tissue (14,
35), was detected by reverse transcriptional-poly-
merase chain reaction (RT-PCR). Sequences for
primers (Sigma Genosys, Hokkaido, Japan) were as
follows: tenomodulin (GenBank ID: AF059407,
product size: 251 bp) forward: 5’-GTC CCT CAA
GTG AAG GTG GA-3’, reverse: 5’-GTT GCA
AGG CAT GAT GAC AC-3’; β-actin (GenBank ID:
AF035774, product size: 488 bp) forward: 5’-TGC
GTG ACA TCA AGG AGA AG-3’, reverse: 5’-
ACA GGT CCT TAC GGA TGT CG-3’. The isolat-
MMPs are a family of zinc-dependent endopepti-
dases that selectively degrade the extracellular ma-
trix (ECM). The MMP family consists of at least 20
enzymes divided into fi ve different groups: collage-
nases, gelatinases, stromelysins, membrane-type
MMPs and other MMPs (36, 37, 43). The gelatinase
group, MMP-2 (gelatinase A, 72 kDa) and MMP-9
(gelatinase B, 92 kDa), appear to play an important
role in tendinopathy, especially in degradation of the
ECM, collagen fi bers and glycosaminoglycans, such
as decorin and biglycan (6, 24). The mechanism un-
derlying the onset of tendonitis has not been fully
clarifi ed; however, exercise-induced heat is a highly
plausible factor.
Tendons provide strong, pliable connections be-
tween muscles and their points of insertion into
bones (19). This tissue exhibits powerful resistance
to tensile force and is also known to be hypovascu-
lar. This strength appears to be closely associated
with the characteristic properties of their ECMs,
which have few capillary networks. During exercise,
the tendon transmits muscular kinetic energy to the
bone, and heat is also generated in the tendon as it
extends and contracts repeatedly at the same time (1,
20, 32). The temperature in the core region of the
tendon reaches 40 to 45°C when a horse is allowed
to gallop (44). The central core of the tendon, which
is the site of most marked temperature increases, is
also the site of degradation and subsequent injury in
both the equine SDFT (42) and human Achilles ten-
don (4). Such a high temperature may not only elic-
it tendon degradation but may also prompt the onset
of tendonitis (5).
The aim of this study was to clarify whether
MMPs (gelatinases) and pro-infl ammatory cytokines
are induced by heat in tendon tissue in vitro and to
test the hypothesis that hyperthermia in the tendon
causes tendinocytes to synthesize pro-infl ammatory
cytokines and that synthesis of these cytokines, in
turn, leads to up-regulation of synthesis of gelatin-
ases.
MATERIALS AND METHODS
This study was performed in accordance with the
Guidelines for Animal Experimentation of Rakuno
Gakuen University, Japan.
Method for isolating tendinocytes and characteristics
of the cells. SDFTs were collected from two healthy
female racehorses (1 and 2 years of age) that had
been euthanized for reasons unrelated to the tendon
or to the musculoskeletal system. Cells were isolat-
Fig. 1 Characteristics of cells used in this study. Isolated
tendinocytes have the characteristic appearance of fi bro-
blasts morphologically (a). Both cells isolated in passage
(P)3 and P4 express tenomodulin (TeM), a specifi c marker
of tendon tissue (b). Bar = 100 µm
Heat effect on gelatinases and cytokines synthesis
235
anti-human proMMP-2 antibody (Santa Cruz Bio-
technology, Santa Cruz, CA, USA) at 1 : 50 dilution
or goat anti-human proMMP-9 antibody (Santa Cruz
Biotechnology) at 1 : 100 dilution for 1 h at room
temperature. After washing again with PBS, they
were incubated with fl uorescein isothiocyanate-con-
jugated rabbit anti-goat IgG (Molecular Probes, Eu-
gene, OR, USA) at 1 : 100 dilution for 30 min at
room temperature in a dark room. Cells on the cov-
erslips were mounted on glass slides using an aque-
ous mounting medium (Permafluor, Immunotech,
Marseille, France) and observed under a laser scan
confocal microscope (Fluoview, Olympus, Tokyo,
Japan).
Exposure to hyperthermia of different temperatures.
Tendinocytes from P3 and P4 passages were seeded
at a cell density of 5 × 10
4 /mL on a 35-mm plastic
plate (TPP) for viability and morphological analy-
ses. For the counting of viable cells, cells were
washed with PBS (-) (Nissui, Tokyo, Japan) and
suspended in trypsin-EDTA (Gibco). Cell suspen-
sions were adjusted to contain 1 × 10
5 cells/mL, and
the suspended cells were warmed in a water bath
for 5, 10, 20, 30 and 60 min at 37, 40, 42 and 45°C.
After heat exposure, the cells were incubated for 20
min at 37°C and transferred to 96-well plates (TPP)
and then incubated for 24 h, during which time via-
ble cells were able to re-adhere to the surface of the
96-well plate.
Quantifi cation of cell survival. Viable cells that had
adhered to the bottoms of 96-well plates were quan-
tifi ed with a cell counting kit (Dojindo, Kumamoto,
Japan) according to the manufacturer’s instructions.
Briefl y, the medium with dead cells was decanted,
and fresh medium containing the reagent equipped
in the commercialized kit was added. The plates
were returned to the incubator at 37°C for 4 h. After
this time, absorbance of the medium was read at
405 and 690 nm with a spectrometer (ImmunoMini
NJ-2300, System Instruments, Tokyo, Japan). Re-
sults were expressed as the percentage cell survival
relative to that of cells kept at 37°C. All data are
given as means ± standard error.
Scanning electron microscopy (SEM). Tendinocytes
were also incubated for 60 min at various tempera-
tures (37, 40, 42 and 45°C) and examined by SEM.
After heat exposure, cells were washed with phos-
phate buffer and post-fi xed with OsO4 at room tem-
perature for 1 h. Conductive staining was carried out
by 1% thiocarboydradize-1% OsO4 treatment (11).
ed cells of both P3 and P4 expressed tenomodulin
mRNA the same as SDFT tissue (Fig. 1b).
Gelatin zymography. To compare the expression
levels of MMPs in vivo and in vitro, SDFT tissue
samples that had been cultured for 48 h and super-
natant were used. SDFT samples and supernatant
were each mixed at ratio of 1 : 4 and 1 : 1 (v/v), re-
spectively, with sample buffer. The sample buffer
consisted of 40 mM Tris, pH 6.8, 5% SDS, 20%
glycerol, and 0.03% bromphenol blue without re-
ducing agent or heating. Five ml of SDFT tissue- or
supernatant-sample buffer mixture was loaded and
electrophoresed through an 8% polyacrylamide gel
containing 0.3% gelatin using an electrophoresis
unit. After electrophoresis of the sample, the gel
was washed and incubated in a reaction buffer
(50 mM Tris-HCl and 10 mM CaCl2, pH 7.4) for
16 h at 37°C. The gels were stained with 0.2% Coo-
massie Brilliant Blue R (Sigma) and destained with
7% acetic acid and 4% methanol to visualize the
unstained proteolytic band. In order to determine the
type of gelatinases observed on the zymograms,
10 mM ethylenediaminetetraacetic acid (EDTA) was
added to the buffer during the incubation period. All
procedures were performed as aforementioned for a
gel without EDTA in the incubation buffer. The mo-
lecular weights of gelatinous bands were estimated
by comparing their electrophoretic migration to that
of protein standards (BioRad, Hercules, CA, USA).
For the cellular MMP assay, cells were incubated
in 35-mm plastic plates (TPP, Trasadingen, Switzer-
land) with pro-inflammatory cytokines, purified
horse TNFα (10 ng/mL) and IL-1β (10 ng/mL), at
37°C for 6 to 72 h, and another batch of cells was
incubated at 40°C for 5 to 60 min. Additionally,
some heat-exposed cells were incubated for 20 min
at various temperatures after 30 min of incubation at
40°C (see below). After pro-infl ammatory cytokine
treatment and heat exposure, supernatants were col-
lected and the gelatin zymogram procedure was car-
ried out as described above. Densities of the bands
were calculated using NIH image.
Immunofluorescent staining of proMMP-2 and -9.
Tendinocytes were grown on coverslips (Fisher
Scientifi c, Pittsburgh, PA, USA) for 48 h for immu-
nostaining. The cells were rinsed with 10 mM phos-
phate-buffered saline (PBS), pH 7.4, and then fi xed
with methanol-acetone (1 : 1) at 4°C for 30 min and
washed with PBS several times. After blocking for
30 min with 3% bovine serum albumin in PBS at
room temperature, cells were incubated with goat
Y. Hosaka et al.
236
The samples were dehydrated and processed for
freeze-drying with t-butyl alcohol substitution (17).
The specimens were examined with a fi eld emission
scanning electron microscope (JSM-6000F, JEOL,
Tokyo, Japan) at an acceleration voltage of 3 kV.
Assay for pro-infl ammatory cytokines after heat ex-
posure. Concentrations of pro-inflammatory cyto-
kines in the supernatant were measured by using a
self-produced sandwich enzyme-linked immunosor-
bent assay (ELISA) for equine TNFα or IL-1β by a
standard method. Mouse monoclonal antibodies
against equine TNFα or IL-1β (21) were used for
capture antibodies. Rabbit polyclonal antibodies
against equine TNFα or IL-1β were used for detect-
ing antibodies. Secondary peroxidase-conjugated
goat anti-rabbit IgG was obtained from Zymed Lab-
oratories (San Francisco, CA, USA). Minimum de-
tectable concentration limit was 1 ng/mL for both
TNFα and IL-1β in the supernatant with the ELISA
kits used in this study.
Effect of heat on synthesis of pro-infl ammatory cyto-
kines by tendinocytes. Due to the very low activity
of proMMP-2, we determined only proMMP-9 (See
Figs. 2 and 4). Heat-exposed cells were incubated
for 20 min at 5, 20 and 37°C after incubation at
40°C for 30 min on ice or in a water bath. Cells that
had not been exposed to heat (maintain at 37°C dur-
ing the experiment) were used as control cells. Then
the supernatant was collected and analyzed by gela-
tin zymography to determine the activity of pro-
MMP-9. Densities of the bands were analyzed using
NIH image software, and the relative ratio of pro-
MMP-9 activities was calculated.
Statistical analysis. A computer program (Stat View
for Windows, version 5.0) was used for the determi-
nation of means and standard errors and for one-
way analysis of variance (ANOVA). Sheffé’s test
was used to compare differences among mean rela-
tive ratios of MMP activities at a signifi cant level of
P < 0.05.
RESULTS
Activity and distribution of gelatinases (MMP-2 and
-9) in vitro
Gelatin zymograms revealed that equine tendon
samples have strong activity of pro- and activated
MMP-2 and -9 (Fig. 2a). In the culture medium
sample, the gelatin zymogram also showed signifi -
cant activities for proMMP-2 and -9, but the activat-
ed forms of both MMPs were hardly detectable,
indicating that auto-activation of these proenzymes
did not occur in vitro (Fig. 2a). EDTA inhibited all
gelatinase activities, confi rming that the bands on
the zymogram represent MMPs (Fig. 2a). Immuno-
histochemical analyses showed both proMMP-2 and
-9 to be distributed in tendinocytes. ProMMP-9
showed positive immunohistochemical reaction, but
proMMP-2 reaction was very weak in the tendino-
cytes (Fig. 2b).
Viability and morphological alteration of tendino-
cytes in hyperthermia model
Percentages of cells surviving after exposure to tem-
peratures of 37 to 45°C for 0 to 60 min are shown
in Fig. 3a. In the cells exposed to heat, the survival
rate decreased sharply in temperature- and time-
dependent manners, especially at 42 and 45°C. The
Fig. 2 Activity and distribution of gelatinases (MMP-2 and -9) in vivo and in vitro. Pro- and activated MMP-2 and -9 show
strong activity in the in vivo sample (SDFT; left lane). In the culture medium sample (middle lane), proMMP-2 and -9 show
enzymic activity; however, enzymic activity for activated MMP-2 and -9 is weak. Zymography gel incubated in the presence
of EDTA (right lane) (a). Both immunopositive reactions for proMMP-2 and -9 are distributed in the tendinocyte, although
proMMP-2 reaction is weak (b). Bar = 200 µm.
Heat effect on gelatinases and cytokines synthesis
237
percentages of tendinocytes that remained viable af-
ter exposure for 10 min to temperatures of 40°C,
42°C and 45°C were 75.3 ± 2.5%, 52.2 ± 4.4 % and
7.7 ± 5.2 %, respectively. At the end of this experi-
ment, after 60 min of heating at 40°C, the survival
rate of tendinocytes was 62.5 ± 2.8%, whereas heat-
ing for 60 min at 42°C and 45°C resulted in a drop
in the cell survival rate to 5.3 ± 3.2% and 1.9 ± 0.6%,
respectively. Electron microscopy revealed that the
cellular structure collapsed in cells exposed to the
highest temperature (45°C) for 60 min, with many
holes in the cell membrane (Fig. 3d). Cells exposed
to 40°C for 60 min, however, showed only a slight
change in survival rate, and multiple cellular projec-
tions were found (Fig. 3c), as if the cells had been
activated by heat.
Effect of pro-infl ammatory cytokines and heat on the
gelatinase activities of tendinocytes
ProMMP-9 synthesis in tendinocytes is strongly in-
duced by pro-infl ammatory cytokines and by heat-
ing. Gelatin zymograms revealed that proMMP-2
and -9 were the only two MMPs remaining in the
supernatant of the cultured tendinocytes, including
that of untreated cells. Neither activated MMP-2 nor
-9 showed change in activity level. Addition of
TNFα and IL-1β to the culture medium of tendino-
cytes accelerated proMMP-9 synthesis considerably
(Fig. 4a, b). On the other hand, the effect of TNFα
or IL-1β on proMMP-2 synthesis was moderate
(Fig. 4a, b). ProMMP-9 synthesis in tendinocytes
was induced by heat in a short time (Fig. 4c, d).
The band density of proMMP-9 at 60 min was
three-fold stronger than the density of the control
cells (0 min).
Effect of heat on pro-infl ammatory cytokine synthe-
sis by tendinocytes
Heated tendinocytes can produce TNFα in a short
time. TNFα was detected in tendinocytes after heat
exposure for 30 and 60 min. In contrast, the concen-
tration of IL-1β was under the detectable level (1 ng/
mL) throughout the experimental period (Fig. 5).
Effect of cooling treatment on proMMP-9 synthesis
by heated tendinocytes
Cooling of heat-exposed tendinocytes reduced the
proMMP-9 level. The proMMP-9 level of heat-ex-
posed cells remaining at 40°C was 2.87-times high-
er than that of the control cells (relative ratio). On
the other hand, cooling of heat-exposed cells from
40°C to 37°C resulted in a considerable decrease in
cellular proMMP-9 synthesis (relative ratio: 2.22
times) compared with that in the cells remaining at
40°C. Furthermore, the proMMP-9 level was re-
duced more in cells cooled to 20°C (1.17 times) and
5°C (1.22 times) than in the control cells (Fig. 6).
Signifi cant differences were found among treatment
temperatures.
DISCUSSION
The results presented in this study suggest that heat
induces tendinocytes to synthesize TNFα and that
synthesis of pro-infl ammatory cytokines (TNFα and
IL-1β) results in up-regulation of proMMP-9.
Severe alterations in the survival rate and mor-
phology of tendinocytes were evident after exposure
to a temperature of 45°C. The cells within tendons
play a leading role in maintaining the ECM through
the synthesis of collagen and other matrix compo-
nents (19). The central core of the tendon, which is
the site of the most marked temperature increases
during exercise, is also the site of degradation and
Fig. 3 Relative ratio of viability and observation of tendi-
nocytes in hyperthermia. In cells exposed to heat, the sur-
vival rate decreases sharply in temperature- and time-
dependent manners, especially at 42 and 45°C (a). Electron
microscopy shows that the cellular structure collapses in
cells exposed to the highest temperature (45°C), with many
holes in the cellular membrane (d). Cells exposed to 40°C,
however, show only a slight change in survival rate, and
multiple cellular projections can be seen (c). b: 37°C
Bar = 10 µm
Y. Hosaka et al.
238
subsequent tendon injury (4, 42, 44). Therefore, a
sharp decrease in cell number caused by heat would
result in a reduced synthetic capability, and altera-
tion of cell metabolism may also endanger the in-
tegrity of the ECM. Heating cells in an in vitro
model as carried out in this study is not completely
analogous to the situation in vivo; however, the re-
Fig. 4 Relative ratio of gelatinase activity after treatment with pro-infl ammatory cytokines or heat exposure. The upper
panels are gelatin zymograms for the supernatant of pro-infl ammatory cytokine-treated cells (a) or heat-exposed (40°C)
cells (c). The gelatin zymogram revealed that proMMP-2 and -9 are the only MMPs remaining in the supernatant of cul-
tured tendinocytes, including that of untreated cells. Treatment of cultured tendinocytes with TNFα or IL-1β accelerated
proMMP-9 (indicated by solid arrowheads) synthesis considerably (a, b). On the other hand, TNFα and IL-1β had little ef-
fect on proMMP-2 synthesis (indicated by empty arrowheads) (a, b). Heat induces proMMP-9 (solid arrowhead) synthesis in
tendinocytes in a short time (c, d).
Fig. 5 Synthesis of pro-infl ammatory cytokines after heat
exposure. TNFα was detected in tendinocytes after heat
exposure (40°C) and increased in the short time. In con-
trast, IL-1β is under a detectable level. Minimum detect-
able limits for ELISA kits used in this study are 1 ng/mL for
both TNFα and IL-1β.
Fig. 6 Activity of proMMP-9 after cooling treatment. Cool-
ing of heat-exposed cells from 40°C to 37°C resulted in
considerable down-regulation of cellular proMMP-9 activity.
Furthermore, proMMP-9 synthesis level was greatly re-
duced in cells treated at lower temperatures, 20°C and 5°C.
Different letters (a, b, and c) indicate signifi cant difference
(P < 0.05).
Heat effect on gelatinases and cytokines synthesis
239
sults of this study suggest that exercise-induced hy-
perthermia might play a role in the pathogenesis of
degenerative core lesions of the tendon.
ProMMP-9 synthesis in tendinocytes was induced
by pro-infl ammatory cytokines and by heating, and
heated cells could produce pro-infl ammatory cyto-
kines. In the present study, the effects of two pro-
infl ammatory cytokines on activities of proMMP-2
and -9 in the culture of equine tendinocytes were
examined at P3 or P4. This is the fi rst report on the
stimulatory effects of TNFα and IL-1β on produc-
tion of MMPs in cultured tendinocytes. The gelatin-
ases MMP-2 and -9 are known to be synthesized
and secreted by several types of cells such as mac-
rophages, mast cells, neutrophils and epidermal ke-
ratinocytes, and fibroblasts (36, 37, 43). Pro-
inflammatory cytokines such as TNFα and IL-1β
play a major role in the process of degradation and
wound healing in connective tissue (13, 41). TNFα
is produced by various cells and enhances many bi-
ological events, including production of MMPs and
stimulation or inhibition of cellular proliferation,
and secreted TNFα stimulates cells to synthesize IL-
1β in a para/autocrine manner (12, 27, 28). Even in
the tendon, these gelatinases and pro-infl ammatory
cytokines are synthesized by tendinocytes and are
involved in the turnover of tendon tissue and in the
maintenance of homeostasis (9, 15, 16, 23). In sev-
eral types of fi broblast-like cells, pro-infl ammatory
cytokines can activate both MMP-2 and -9 (45). In
the present study, both TNFα and IL-1β up-regulat-
ed the production of proMMP-9 but had little effect
on the production of proMMP-2. These observations
suggest that MMP-2 and -9 are differentially regu-
lated in tendinocytes.
This study also showed that heated cells produced
TNFα in a short time. A number of previous studies
have shown that hyperthermia can modulate TNFα
synthesis in many cells (31, 38, 39). Augmentation
of TNFα synthesis has been thought to occur
through several mechanisms, including increased
transcription or translation of TNFα mRNA, altered
secretory events or stability, or heat damage-induced
release of membrane-bound TNFα (22, 38). We
have limited the scope of this study to reveal the
mechanism of induction of TNFα by heat, but clear-
ly there are interactions between hyperthermia and
TNFα secretion by tendinocytes.
Cooling of heat-exposed tendinocytes reduced the
level of proMMP-9. Lowering the temperature from
40°C to 37°C resulted in considerable down-regula-
tion of proMMP-9 activity in heat-exposed cells.
Furthermore, proMMP-9 synthesis level was greatly
reduced in cells treated at lower temperatures, 20°C
and 5°C. In hypovascular tissues such as the tendon,
ligament and epimysium of skeletal muscles, the
rate of movement of materials is thought to be much
slower than in other vascular tissues because materi-
als move by diffusion (19, 46). This means that ef-
fl ux of some biological factors synthesized in these
tissues is diffi cult, and the factors easily accumulate
in the tissue in a short time. Similarly, once heat is
produced in a tendon, a hyperthermic condition is
maintained for a long time because thermo-diffusion
is hardly ever mediated by the bloodstream. The hy-
povascular nature of the tendon might facilitate the
continuance of a hyperthermia condition after exer-
cise and inducement of pro-infl ammatory cytokines
by heat. Therefore, cooling of the tendon after exer-
cise might allow heat-induced synthesis of pro-in-
flammatory cytokines to be controlled and might
inhibit the occurrence or progression of tendonitis.
In conclusion, 1) heat infl uences the survival rate
of cells and cellular morphology, 2) proMMP-9 syn-
thesis in tendinocytes is induced by pro-infl ammato-
ry cytokines and by heating, and heated cells can
produce TNFα in a short time, and 3) cooling of
heat-exposed tendinocytes reduces the proMMP-9
level. Together, these fi ndings support our hypothe-
ses that hyperthermia in the horse tendon induces
synthesis of pro-infl ammatory cytokines by tendino-
cytes and that synthesis of these cytokines results in
up-regulation of synthesis of gelatinases. The results
of this study are of particular clinical importance for
the prevention of tendon degradation possibly by
control of the tendon temperature in the animal. It is
conceivable that cooling the legs of horses after
training is a simple but effective means of prevent-
ing tendinopathy.
Acknowledgements
The authors are grateful to Dr. Nell L. Kennedy,
Rakuno Gakuen University, Japan, and Mrs. Carol
Cochrane-Yamaguchi, Ohio, USA for review of this
article. This study was partly supported by the Sa-
sakawa Scientifi c Research Grant from The Japan
Science Society (F05-403).
REFERENCES
1. Archambault JM, Wiley JP and Bray RC (1995) Exercise
loading of tendons and the development of overuse injuries.
A review of current literature. Sports Med 20, 77–89.
2. Åström M and Westlin N (1994) Blood flow in chronic
Achilles tendinopathy. Clin Orthop Relat Res 308, 166–172.
3. Arechavaleta-Velasco F, Ogando D, Parry S and Vadillo-
Y. Hosaka et al.
240
Ortega F (2002) Production of matrix metalloproteinase-9 in
lipopolysaccharide-stimulated human amnion occurs through
an autocrine and paracrine proinfl ammatory cytokine-depen-
dent system. Biol Report 67, 1952–1958.
4. Arner O, Liindholm A and Orell SR (1959) Histological
changes in subcutaneous rupture of the Achilles tendon. Acta
Chir Scand 116, 484–490.
5. Birch HL, Wilson AM and Goodship AE (1997) The effect
of exercise-induced localised hyperthermia on tendon cell
survival. J Exp Biol 200, 1703–1708.
6. Chakraborti S, Mandal M, Das S, Mandal A and Chakraborti
T (2003) Regulation of matrix metalloproteinases: an over-
view. Mol Cell Biochem 253, 269–285.
7. Chang YC, Yang SF, Lai CC, Liu JY and Hsieh YS (2002)
Regulation of matrix metalloproteinase production by cyto-
kines, pharmacological agents and periodontal pathogens in
human periodontal ligament fi broblast cultures. J Periodontal
Res 37, 196–203.
8. Chu SC, Yang SF, Lue KH, Hsieh YS, Wu CL and Lu KH
(2004) Regulation of gelatinases expression by cytokines, en-
dotoxin, and pharmacological agents in the human osteoar-
thritic knee. Connect Tissue Res 45, 142–150.
9. Dahlgren LA, Mohammed HO and Nixon AJ (2005) Tempo-
ral expression of growth factors and matrix molecules in
healing tendon lesions. J Orthop Res 23, 84–92.
10. Dasu MR, Barrow RE, Spies M and Herndon DN (2003)
Matrix metalloproteinase expression in cytokine stimulated
human dermal fi broblasts. Burns 29, 527–531.
11. Familiari G, Nottola SA, Macchiarelli G, Familiari A and
Motta PM (1992) A technique for exposure of the glycopro-
teic matrix (zona pellucida and mucus) for scanning electron
microscopy. Microsc Res Tech 23, 225–229.
12. Gan X, Wong B, Wright SD and Cai TQ (2001) Production
of matrix metalloproteinase-9 in CaCO-2 cells in response to
infl ammatory stimuli. J Interferon Cytokine Res 21, 93–98.
13. Hatamochi A, Mori K and Ueki H (1994) Role of cytokines
in controlling connective tissue gene expression. Arch Der-
matol Res 287, 115–121.
14. Hiraki Y and Shukunami C (2005) Angiogenesis inhibitors
localized in hypovascular mesenchymal tissues: chondromod-
ulin-I and tenomodulin. Connect Tissue Res 46, 3–11.
15. Hosaka Y, Kirisawa R, Yamamoto E, Ueda H, Iwai H and
Takehana K (2002) Localization of cytokines in tendinocytes
of the superfi cial digital fl exor tendon in the horse. J Vet Med
Sci 64, 945–947.
16. Hosaka Y, Ueda H, Yamasaki T, Suzuki D, Matsuda N and
Takehana K (2005) Structure and component alteration of
rabbit Achilles tendon in tissue culture. Biomed Res 26, 279–
286.
17. Inoue T and Osatake H (1988) A new drying method of
biological specimens for scanning electron microscopy: the
t-butyl alcohol freeze-drying method. Arch Histol Cytol 51,
53–59.
18. Jones GC, Corps AN, Pennington CJ, Clark IM, Edwards
DR, Bradley MM, Hazleman BL and Riley GP (2006) Ex-
pression profi ling of metalloproteinases and tissue inhibitors
of metalloproteinases in normal and degenerate human achil-
les tendon. Arthritis Rheum 54, 832–842.
19. Józsa LG and Kannus P (1997) Structure and Metabolism of
Normal Tendons. In: Human Tendons (Józsa LG, Kannus P,
eds), pp 76–81, Human Kinetics, Champaign, Illinois.
20. Ker RF (1981) Dynamic tensile properties of the plantaris
tendon of sheep (Ovis aries). J Exp Biol 93: 283–302.
21. Kirisawa R, Hashimoto N, Tazaki M, Yamanaka H, Ishii R,
Hagiwara K and Iwai H (2006) Molecular characterization
and functional expression of equine interleukin-1 type I and
type II receptor cDNAs. Vet Immunol Immunopathol 109,
219–231.
22. Klostergaard J, Barta M and Tomasovic SP (1989) Hyper-
thermic modulation of tumor necrosis factor-dependent
monocyte/macrophage tumor cytotoxicity in vitro. J Biol Re-
sponse Mod 8, 262–277.
23. Koskinen SO, Heinemeier KM, Olesen JL, Langberg H and
Kjær M (2004) Physical exercise can infl uence local levels
of matrix metalloproteinases and their inhibitors in tendon-
related connective tissue. J Appl Physiol 96, 861–864.
24. Kuroki K, Stoker AM and Cook JL (2005) Effects of proin-
fl ammatory cytokines on canine articular chondrocytes in a
three-dimensional culture. Am J Vet Res 66, 1187–1196.
25. Leadbetter WB (1992) Cell-matrix response in tendon injury.
Clin Sports Med. 11, 533–578.
26. Malemud CJ (2006) Matrix metalloproteinases (MMPs) in
health and disease: an overview. Front Biosci 11,1696–1701.
27. Mast BA and Schultz GS (1996) Interactions of cytokines,
growth factors, and proteases in acute and chronic wounds.
Wound Repair Regen 4, 411–420.
28. Nee LE, McMorrow T, Campbell E, Slattery C and Ryan MP
(2004) TNFα and IL-1β-mediated regulation of MMP-9 and
TIMP-1 in renal proximal tubular cells. Kidney Int 66, 1376–
1386.
29. Oikawa M and Kasashima Y (2002) The Japanese experience
with tendonitis in racehorses. J Equine Sci 13, 41–56.
30. Oshiro W, Lou J, Xing X, Tu Y and Manske PR (2003) Flex-
or tendon healing in the rat: a histologic and gene expression
study. J Hand Surg (Am) 28, 814–823.
31. Rhind SG, Gannon GA, Shephard RJ, Buguet A, Shek PN
and Radomski MW (2004) Cytokine induction during
exertional hyperthermia is abolished by core temperature
clamping: neuroendocrine regulatory mechanisms. Int J Hy-
perthermia 20, 503–516.
32. Riemersma DJ and Schamhardt HC (1985) In vitro mechani-
cal properties of equine tendons in relation to crosssectional
area and collagen content. Res Vet Sci 39, 263–270.
33. Riley GP, Curry V, DeGroot J, van El B, Verzijl N, Hazle-
man BL and Bank RA (2002) Matrix metalloproteinase activ-
ities and their relationship with collagen remodelling in
tendon pathology. Matrix Biol 21, 185–195.
34. Russell JE and Manske PR (1989) Biochemical evaluation of
organ cultures from primate fl exor tendons. Connect Tissue
Res 23, 51–64.
35. Shukunami C, Oshima Y and Hiraki Y (2005) Chondromod-
ulin-I and tenomodulin: a new class of tissue-specifi c angio-
genesis inhibitors found in hypovascular connective tissues.
Biochem Biophys Res Commun 333, 299–307.
36. Stamenkovic I (2003) Extracellular matrix remodelling: the
role of matrix metalloproteinases. J Pathol 200, 448–464.
37. Sternlicht MD and Werb Z (2001) How matrix metallopro-
teinases regulate cell behavior. Annu Rev Cell Dev Biol 17,
463–516.
38. Tomasovic SP and Klostergaard J (1989) Hyperthermic mod-
ulation of macrophage-tumor cell interactions. Cancer Metas-
tasis Rev 8, 215–229.
39. Tomasovic SP, Lu S and Klostergaard J (1992) Comparative
in vitro studies of the potentiation of tumor necrosis factor
(TNF)-α, TNF-β, and TNF-SAM2 cytotoxicity by hyperther-
mia. J Immunother 11, 85–92.
40. Viitanen M, Bird J, Smith R, Tulamo RM and May SA (2003)
Biochemical characterisation of navicular hyaline cartilage,
Heat effect on gelatinases and cytokines synthesis
241
perthermia as a possible mechanism for tendon degeneration.
J Biomech 27, 899–905.
45. Wong WR, Kossodo S and Kochevar IE (2001) Infl uence of
cytokines on matrix metalloproteinases produced by fi bro-
blasts cultured in monolayer and collagen gels. J Formos
Med Assoc 100, 377–382.
46. Young B and Heath J (2000) Connective tissues. In: Wheat-
ers Functional Histology (Burkitt H et al., eds), pp 97–192,
Churchill Livingstone, Edinburgh, Scotland, UK.
navicular fi brocartilage and the deep digital fl exor tendon in
horses with navicular disease. Res Vet Sci 75, 113–120.
41. Wahl SM, Wong H and McCartney-Francis N (1989) Role of
growth factors in infl ammation and repair. J Cell Biochem
40, 193–199.
42. Webbon PM (1977) A post mortem study of equine digital
fl exor tendons. Equine Vet J 9, 61–67.
43. Werb Z (1997) ECM and cell surface proteolysis: regulating
cellular ecology. Cell 91, 439–442.
44. Wilson AM and Goodship AE (1994) Exercise-induced hy-