| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLES |
Department of Obstetrics and Gynecology, University of Kentucky (K.S.S., T.E.C.), Lexington, Kentucky 40536; and Department of Obstetrics and Gynecology, University of Wisconsin (M.J.B.), Madison, Wisconsin 53715
Address all correspondence and requests for reprints to: Kristen S. Simpson, Ph.D., Department of Obstetrics and Gynecology, MS 331, 800 Rose Street, University of Kentucky, Lexington, Kentucky 40536-0298. E-mail: kss{at}pop.uky.edu
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The TIMPs are a family of proteinase inhibitors that at present includes four proteins (TIMP-1, TIMP-2, TIMP-3, and TIMP-4) (4, 5, 7). TIMP-1 and TIMP-2 are both secreted proteins that have the ability to inhibit the activity of all MMPs (5). The TIMPs have been implicated in a number of functions within the ovary, including MMP inhibition. Both TIMP-1 and TIMP-2 have growth factor-like activity, can cause changes in cell morphology, and have the ability to inhibit angiogenesis, all processes that may contribute to the changes occurring in the cyclic ovary (5). TIMP-1 may have another important function in the ovary, as it has the ability to stimulate steroidogenesis in granulosa cells when complexed with the proenzyme form of cathepsin L (10).
Unique among the TIMPs, TIMP-3 protein is bound to the ECM, giving it the highest potential of altering the ECM immediately surrounding cells. It also has a greater ability to inhibit MMP-9 than either TIMP-1 or TIMP-2. There is evidence suggesting that TIMP-3 is regulated by the cell cycle in some cell types (4), and it may have a role in CL function and/or regression, as its expression in the ovary increased on day 4 of pseudopregnancy in rats and remained fairly high throughout the life span of the CL (11).
Little is known about the function of the most recently identified TIMP, TIMP-4, which is a secreted protein expressed at low levels in the ovary (12, 13). Based upon its amino acid sequence it is more similar to TIMP-2 and TIMP-3 than to TIMP-1 (12, 14) and, like TIMP-2, is capable of binding to MMP-2 (15).
To date, most studies in the rat have examined the expression patterns of the TIMPs in gonadotropin-primed prepubertal animals, which may not reflect the true physiological expression of these proteins in the adult cycling animal. Therefore, the current study was undertaken to examine the spatial and temporal localization of the various TIMP transcripts in rat ovaries collected throughout the 4-day estrous cycle. In addition, we examined TIMP-3 messenger RNA (mRNA) expression in atretic follicles, as overexpression of TIMP-3 has the ability to induce apoptosis in some cells (16, 17).
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
RNA isolation and Northern blot analysis
Total RNA was isolated from individual ovaries using TRIzol
reagent (Life Technologies, Inc., Rockville, MD). Each
ovary was homogenized in 800 µl TRIzol containing 250 µg/ml
glycogen. The manufacturer’s protocol was followed for total RNA
isolation. Concentrations of RNA were determined by
spectrophotometry.
Total RNA (20 µg) from each individual ovary (n = 6/day of
the estrous cycle) was electrophoresed through a 1% agarose gel
containing 2.2 M formaldehyde and then transferred onto a
positively charged nylon membrane (Nytran, Schleicher & Schuell, Inc., Keene, NH). Northern blots were prehybridized in
hybridization solution (Northern Max Prehybridization/Hybridization
Solution, Ambion, Inc., Austin, TX) and subsequently
hybridized overnight with [
-32P]-labeled
complementary DNA (cDNA) probes specific for mouse TIMP-1, -2, or -3 at
a concentration of 2 ng probe/ml hybridization solution (murine TIMP
cDNA plasmids used for probe preparation were supplied by Dr.
Kevin Leco, University of Western Ontario, London, Ontario,
Canada). Probes were synthesized by random priming using the
Rediprime II random primer labeling kit (Amersham Pharmacia Biotech, Piscataway, NJ) and were purified over Mini Quick Spin
DNA columns (Roche Molecular Biochemicals, Indianapolis,
IN). After hybridization, the blots were washed according to standard
protocols at 55 C (18) and exposed to X-OMAT AR x-ray film
(Eastman Kodak Co., Rochester, NY). Blots were
simultaneously probed for TIMP-1 and TIMP-3 mRNA, followed by a second
probing using a TIMP-2-specific probe. Between each experiment, the
probes were stripped from the membranes. All blots were subsequently
hybridized with an 18S ribosomal RNA (rRNA) probe to control for the
amount of RNA loaded per lane. Each RNA sample was analyzed in
duplicate. Relative band densities for the Northern blots were
determined using MetaMorph software (Universal Imaging Corp., West
Chester, PA) and were normalized to the relative expression of the 18S
rRNA.
In situ hybridization
In situ hybridization probes were produced using
plasmids containing murine TIMP-1, TIMP-2, or TIMP-3 cDNA. Plasmids
were linearized using the appropriate restriction enzymes, and the
antisense and sense riboprobes for TIMP-1, TIMP-2, and TIMP-3 were
synthesized from the corresponding linearized plasmid and labeled with
[-35S]UTP using the Maxiscript
in vitro RNA transcription kit from Ambion, Inc. After riboprobe synthesis, the probes were purified over
Sephadex G-50 Quick Spin Columns (Roche Molecular Biochemicals). Approximate lengths of the synthesized probes
were 520 bases (TIMP-1 antisense), 400 bases (TIMP-1 sense), 370 bases
(TIMP-2 antisense and sense probes), and 360 bases (TIMP-3 antisense
and sense probes).
Ovaries were serially sectioned at 12 µm and mounted on Probe-On Plus slides (Fisher Scientific, Pittsburgh, PA). Tissues were fixed in 4% paraformaldehyde in PBS and then sequentially washed in PBS, 0.75% glycine, PBS, and 1.5% triethanolamine with 0.25% acetic anhydride before being dehydrated. Tissues were allowed to hybridize overnight in a humidified chamber (60 C) with a specific TIMP probe in hybridization buffer [50% formamide, 10% dextran sulfate, 20 mM Tris-HCl (pH 7.4), 1 mM EDTA (pH 8.0), 300 mM NaCl, 1 x Denhardt’s solution, 0.1 mg/ml salmon sperm DNA, 0.25 mg/ml yeast total RNA, 0.25 mg/ml yeast transfer RNA, 0.1% SDS, 0.1% sodium thiosulfate, and 100 mM dithiothreitol] containing 3 x 106 cpm probe/slide. Approximately 18–20 h later, slides were washed extensively to remove nonspecifically bound TIMP complementary RNA. Tissues were washed with 2 x SSC (standard saline citrate) buffer (1 x SSC = 0.15 M NaCl and 15 mM sodium citrate; all washes in SSC also contained 10 mM sodium thiosulfate), followed by ribonuclease A treatment (100 µg/ml in Tris-EDTA buffer) for 30 min at 45 C. Slides were then washed in 0.2 x SSC, followed by a wash in 0.1 x SSC for 1 h at 55 C before rinsing in deionized H2O, dehydrating in ethanol, and air-drying. Sections were processed for autoradiography using Kodak NTB2 emulsion (Eastman Kodak Co.) and were stored at 4 C for 1–4 weeks. For visualization of the in situ reaction product, slides were developed in Kodak D19 (1:1) and stained with Gill’s Formulation #2 hematoxylin solution (Fisher Scientific). Tissues were examined with a Nikon Microphot-SA microscope (Nikon, Melville, NY) under bright- and darkfield optics. A sense riboprobe, used as a control for nonspecific binding, was included for each ovary and each time point for the different TIMPs.
One ovary from each of three of the animals was used for in situ hybridization (n = 3). For each TIMP, 16 tissue sections/ovary were analyzed, making a total of 48 tissue sections analyzed for each time point. Four sections per ovary were analyzed using each TIMP sense probe, so that a total of 12 sections/time point were examined.
Laser capture microdissection (LCM)
TIMP-1 mRNA expression was further examined in granulosa cells
using LCM and RT-PCR. One OCT-embedded ovary from each day of the
estrous cycle was sectioned at 10 µm, and the tissue sections were
mounted onto uncharged slides. The tissue sections were stained with
methyl green, and the granulosa cells were visualized and captured
using an Arcturus PixCell II LCM System (Mountain View, CA). Using this
system, cells were captured onto a thermoplastic membrane on optically
transparent caps (CapSure LCM Transfer Film TF-100, Arcturus) using a
laser pulse that melted a small area of the cap membrane onto a
specified area of the tissue (19, 20). The membrane
immediately solidified back onto the cap, taking with it any cells that
were in the specified area of tissue. For laser capture, a 7.5-µm
laser setting was used with pulse intensity and duration settings
(50–80 mwatts and 500–750 µsec) that were the minimum required to
capture one to three cells. After capturing granulosa cells from one
ovary (1000–1500 laser pulses), the cap was placed onto a 0.5-ml
microcentrifuge tube containing 200 µl TRIzol reagent and 250 µg/ml
glycogen. Tubes were inverted multiple times and set on ice until RNA
isolation was performed as described above. After precipitation, the
RNA was air-dried and resuspended in 20 µl diethyl
pyrocarbonate-treated water.
RT-PCR
RNA samples collected by LCM were treated with deoxyribonuclease
I (DNase I; Life Technologies, Inc.) and then reverse
transcribed using the Superscript First Strand Synthesis System
(Life Technologies, Inc.). Five microliters of the RNA
were treated with DNase, and, for the RT reaction, 3.6 µl
DNase-treated RNA were combined with 2.5 µM
oligo(deoxythymidine) primer and 500 µM of each dNTP. The
mixture was then heated at 65 C for 5 min. Subsequently, 1 x RT
buffer [20 mM Tris-HCl (pH 8.4) and 50 mM
KCl], 0.01 M dithiothreitol, 5 mM
MgCl2, and 2 U/µl RNaseOut ribonuclease
inhibitor (Life Technologies, Inc.) were added, and the
mixture was heated at 42 C for 5 min. Fifty units of Superscript
II reverse transcriptase (Life Technologies, Inc.)
were then added, and the reaction was performed at 42 C for 50 min,
after which the reaction was stopped by heating at 70 C for 15 min.
Subsequently, the cDNA for TIMP-1 was amplified by PCR using
TIMP-1-specific primers (upper primer, 5'-TCCCCAGAAATCATCGAGAC-3';
lower primer, 5'-ATGGCTGAACAGGGAAACAC-3') synthesized by Integrated DNA
Technologies (Coralville, IA). For PCR, 5 µl of the 20 µl RT
reaction, 1 x PCR buffer [20 mM Tris-HCl (pH 8.4)
and 50 mM KCl], 1.5 mM
MgCl2, 200 µM of each dNTP, and 1
µM of each gene-specific primer were combined for a final
PCR reaction volume of 50 µl. Hot start PCR was performed using
TaqBead Hot Start Polymerase (1.25 U Taq; Promega Corp.) for 35 cycles according to the following parameters: 94 C
for 30 sec, 50 C for 30 sec, and 72 C for 45 sec in an
Eppendorf Mastercycler Gradient thermal cycler (Westbury,
NY). Negative controls were performed in which either no reverse
transcriptase or no RNA was added to the RT reactions to ensure that
there was no DNA contamination of the RNA samples or reagents. After
PCR amplification, the PCR products were electrophoresed through a 1%
agarose gel and visualized with ethidium bromide staining.
DNA fragmentation assay
Serial sections of the ovaries used for in situ
hybridization were also examined for apoptosis by testing for DNA
fragmentation using the terminal deoxynucleotidyl transferase-mediated
biotin-deoxyuridine triphosphate nick end labeling method. The frozen
tissue sections (12 µM) were prepared by fixing
in 4% paraformaldehyde and rinsing in PBS. DNA fragmentation was
determined using the ApoAlert DNA Fragmentation Assay from
CLONTECH Laboratories, Inc. (Palo Alto, CA). After
performing the assay according to the manufacturer’s instructions, the
slides were mounted with VectaShield Mounting Medium with propidium
iodide (Vector Laboratories, Inc., Burlingame, CA).
Tissues were analyzed on an Eclipse E800 microscope
(Nikon, Melville, NY) for fluorescence.
Data analysis
For the Northern blot analysis, one ovary from each of six
animals was examined on each day of the estrous cycle (n = 6). Two
blots were analyzed for each ovarian RNA sample (i.e. n
= 6 in duplicate), and the levels of transcript were normalized to the
levels of the 18S rRNA and averaged for the two blots before
statistical analysis. Differences in mRNA levels were tested by one-way
ANOVA. Post-hoc group comparisons were performed using
Tukey’s test, with P < 0.05 considered
significant.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
). The days of metestrus, diestrus, and
proestrus did not significantly differ from one another. Two
transcripts were detected for TIMP-2 by Northern blotting, a 3.5-kb and
a 1.0-kb transcript (Fig. 1B). A significant decrease was detected
between metestrus and diestrus for the 3.5-kb transcript, but there was
no change in expression on any other day. Expression levels of the
1.0-kb TIMP-2 transcript did not change throughout the estrous cycle
(P = 0.077 between metestrus and diestrus). TIMP-3
transcript levels were highest on proestrus, but significantly
decreased on the day of estrus (Fig. 1C). Expression of TIMP-3 mRNA on
metestrus and diestrus was intermediate between the high levels at
proestrus and the nadir at estrus.
|
In developing follicles on all days examined, TIMP-1 mRNA was found to
be highly expressed in the theca, but was low to undetectable in the
granulosa layer (Fig. 2). Stromal
labeling was also high in the ovary on all days of the estrous cycle.
Of particular interest was the observation that TIMP-1 exhibited a
high, homogeneous labeling pattern within newly forming CL with a
bright band of hybridization encircling the CL (Fig. 2, A and B).
However, the CL from previous cycles did not have this encircling band
of hybridization and exhibited a more heterogeneous pattern of
hybridization throughout (Fig. 2, A and B), which appeared to decrease
with the apparent age of the CL (i.e. CL from two or more
cycles previous; data not shown). The expression of TIMP-1 mRNA within
the new CL declined by metestrus, although the band of hybridization
encircling the CL remained (Fig. 2, C and D). On metestrus, CL from
previous cycles exhibited a pattern of high, heterogeneous interior
labeling, no surrounding band (Fig. 2D), and a decrease in labeling
with the apparent age of the CL. This pattern of labeling
continued through diestrus (Fig. 2, E and F), but by proestrus, the
bands surrounding the youngest CL were no longer visible, although the
interior of these CL still exhibited a high level of labeling (Fig. 2, G and H).
|
. This technique enabled us to collect a
pure population of granulosa cells from which to isolate RNA. A PCR
product of the expected size was produced from the granulosa cells,
demonstrating the presence of TIMP-1 mRNA in granulosa cells on all
days of the rat estrous cycle (data not shown). Negative controls did
not produce a detectable TIMP-1 PCR product.
|
). The high level of TIMP-2 mRNA present
in the theca of follicles appeared to remain after ovulation, as a
bright ring of hybridization was evident surrounding new CL on the day
of estrus (Fig. 4, A and B). There was also an increase in TIMP-2 mRNA
expression levels within the newly forming CL. CL from previous cycles
expressed TIMP-2 mRNA at higher levels than the newly forming CL, with
localized regions of high labeling within the CL and labeling
surrounding the CL. Levels of TIMP-2 hybridization within the new CL
appeared to decline after estrus to reach low levels by diestrus (Fig. 4, C–F). By proestrus, the mRNA expression pattern within the new CL
was similar to that in CL from previous cycles, with higher overall
expression levels and localized regions of high expression. A unique
finding in this study was that all follicles and CL appeared to have an
encircling ring of expression on all days of the cycle that was of
thecal origin.
|
). In most sections, as with
the other TIMPs, little TIMP-3 mRNA expression was detected in
granulosa cells, but high levels of expression were seen in thecal
tissue. Of note was the observation that some follicles exhibited a low
level of TIMP-3 expression in the granulosa cells (seen in the follicle
F in Fig. 5H). Upon ovulation, a high level of transcript was expressed
throughout the new CL, whereas CL from previous cycles exhibited a high
level of expression in only some cells (Fig. 5, A and B). Older CL from
several cycles ago had only a low level of expression. As opposed to
the other TIMPs, there was no ring of TIMP-3 hybridization surrounding
any of the CL. TIMP-3 expression changed in the new CL by metestrus,
with high levels of mRNA in only some cells, similar to older CL, but
at higher overall levels (Fig. 5, C and D). This pattern of expression
continued through diestrus (Fig. 5, E and F). No differences in TIMP-3
expression could be seen between the most recent CL and older CL on
proestrus, as all showed a high level of expression in certain cells
(data not shown).
|
. TIMP-1 mRNA was highly expressed
throughout the newly formed CL, with a band of expression surrounding
the CL (Fig. 6, A–C). It was also seen at lower levels in the
granulosa cells of follicles (F). In contrast, TIMP-2 mRNA was
expressed at low levels within the newly formed CL, but was very highly
expressed in a band surrounding the CL (Fig. 6, D–F). No hybridization
for TIMP-2 was seen in the granulosa cells of follicles. TIMP-3 mRNA
was expressed in a punctate pattern throughout the new CL (Fig. 6, G–I). Specific hybridization was not detected in tissues hybridized
with the sense riboprobes for TIMP-1, TIMP-2, or TIMP-3 (Fig. 6, C, F,
and I).
|
, G and H, and Fig. 5, G
and H). To investigate these observations further, serial tissue
sections were analyzed for either TIMP-2 or TIMP-3 mRNA by in
situ hybridization or for DNA fragmentation by the terminal
deoxynucleotidyl transferase-mediated biotin-deoxyuridine triphosphate
nick end labeling method. The sections were compared to determine
whether TIMP-3 mRNA was more abundant in the granulosa cells of atretic
vs. healthy follicles and whether the high level of TIMP-2
and TIMP-3 mRNA expression seen in the small follicles was associated
with atresia or luteinization of the follicles. TIMP-3 transcript was
generally observed to be higher in the granulosa cells of healthy
follicles (Fig. 7, A and B). Based upon
their lack of apoptotic cells, their lack of a granulosa cell basement
membrane, and their cellular morphology, the small follicles were
identified as luteinized follicles. Luteinized follicles are a later
stage in the regression of follicles. In luteinized follicles, the
granulosa cells and oocyte have become atretic, regressed, and been
removed, whereas the cells of the thecal layer have hypertrophied and
become more luteal-like (23, 24). Comparisons of adjacent
sections analyzed by in situ hybridization for TIMP-2 and
TIMP-3 showed that the high level of expression of these transcripts
seen on proestrus were in these later stage luteinized follicles, not
in early stage atretic follicles (Fig. 7, C and D; TIMP-2 data not
shown). The difference between a normal follicle and a luteinized
follicle and their expression of TIMP-3 mRNA can be seen in Fig. 8. The normal follicle expressed low
levels of TIMP-3 mRNA in the granulosa cells and high levels in the
theca, whereas the luteinized follicle expressed high levels of TIMP-3
mRNA throughout. Expression of TIMP-3 transcript in the luteinized
follicles declined to lower levels by estrus, but could again be seen
in these follicles by metestrus and diestrus (Fig. 5, D and F). It was
further observed in these studies, that CL from previous cycles
contained a small number of apoptotic cells (data not shown) in
agreement with other studies (25, 26).
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The increase in TIMP-1 mRNA between proestrus and estrus was most likely a result of the LH surge. Previous studies have shown that TIMP-1 mRNA is up-regulated in the ovary by hCG treatment of the gonadotropin-primed immature animal (27, 28, 29). The increase in whole ovarian mRNA expression observed on the morning of estrus in the current study reflects the dramatic increase in TIMP-1 expression in the luteinizing granulosa cells of newly forming CL compared with the low levels of expression in granulosa cells of developing follicles. This observation is in agreement with previous studies performed in the gonadotropin-primed prepubertal rat, where expression of TIMP-1 mRNA in the granulosa layer of preovulatory follicles is at basal or undetectable levels until stimulation by hCG. After an hCG stimulus, there is an induction of TIMP-1 mRNA as the granulosa cells undergo luteinization (28, 30, 31). As controversy exists as to whether the granulosa cells express TIMP-1 mRNA before an LH/hCG stimulus, we used the powerful technique of laser capture microdissection to definitively identify TIMP-1 mRNA in rat granulosa cells from antral follicles on all days of the cycle. The presence of TIMP-1 mRNA in the granulosa cells of developing follicles may be acting to regulate the MMPs and ECM turnover. Alternatively, TIMP-1 may impact steroidogenesis in the granulosa cells, as a TIMP-1/cathepsin L complex increased progesterone in cultured rat granulosa cells (10), and the TIMP-1 knockout mouse has decreased progesterone production on estrus and increased estrogen production on estrus and diestrus (32).
The in situ localization provided several unique observations regarding TIMP-1 expression. First, there was an intense band of TIMP-1 mRNA expression encircling the forming CL. Second, the increase in TIMP-1 mRNA expression in the luteinizing granulosa and thecal cells of the forming CL on estrus was followed by a decrease in mRNA abundance 24 h later on metestrus. These results support the hypothesis that TIMP-1 is important in the newly forming CL, probably for controlling MMP remodeling of the ECM as the postovulatory follicle is transformed into the CL. As noted previously, TIMP-1 may also be crucial for steroidogenesis (10), for controlling angiogenesis as the CL is rapidly vascularized, for regulating growth through its actions as a growth factor, or for altering cellular morphology, all functions that have been attributed to TIMP-1 (5, 6). A role for TIMP-1 in formation of the CL is further supported by data in the pseudopregnant rat, in which TIMP-1 mRNA was highly abundant during luteal formation. In the pseudopregnant model, TIMP-1 mRNA decreased during CL maintenance, then increased again during CL regression, suggesting that TIMP-1 may also be important for luteal regression, yet is only needed at low levels for CL maintenance (11, 33). Although its level and location change, the continued presence of TIMP-1 transcript throughout the life span of CL suggests that TIMP-1 may be needed at a certain basal level in all stages of luteal function.
Of particular interest were the marked species differences in ovarian TIMP-1 mRNA expression between the rat and the mouse. For example, TIMP-1 mRNA in whole ovarian extracts increased between proestrus and estrus in the rat (current study), whereas the levels of TIMP-1 mRNA in the mouse ovary were highest on proestrus and declined by estrus (34). Differences were also noted in the cellular localization of TIMP-1 mRNA. In the rat, TIMP-1 mRNA was highly expressed in the theca and stroma of cycling animals (current study) and gonadotropin-primed immature animals (30, 31). The cycling mouse ovary exhibited a different pattern of expression, with TIMP-1 mRNA detected in the CL and oocyte, whereas it was not observed in the theca, stroma, or granulosa layer (34). This differs somewhat from the PMSG/hCG-treated immature mouse, which expresses TIMP-1 mRNA in the theca, in the granulosa cells of large preovulatory follicles, and in newly forming CL (28). Although some of these differences in cellular localization may reflect the different techniques used for in situ localization of TIMP-1 mRNA (colorimetric vs. radiometric), the disparate results with Northern analysis highlight the species variation and suggest distinct differences in the function of ovarian TIMP-1 between the rat and the mouse.
Unique to the current study was the demonstration that the two TIMP-2 transcripts were differentially expressed in the cycling rat ovary. The 1.0-kb transcript did not change in expression across the cycle, whereas the 3.5-kb transcript decreased between metestrus and diestrus. Although generally considered to be expressed in a low, constitutive manner (11, 28, 34), differences in the expression patterns for the two TIMP-2 transcripts were also observed in the PMSG-treated rat (35). The PMSG-treated rat exhibited a significant decrease in the larger transcript by 24 h after PMSG injection, whereas no change was detected up to 48 h post-PMSG for the smaller transcript (35). Therefore, based upon our data there are slight differences in the expression of ovarian TIMP-2 mRNA over the cycle.
TIMP-2 mRNA was observed in the theca and stroma on all days of the cycle, similar to the localization reported in ovaries from PMSG-primed rats (31). The current study clearly demonstrates that after ovulation, TIMP-2 mRNA was highly up-regulated in the luteinizing thecal layer and became detectable in the luteinizing granulosa layer. The significance of the decrease in expression of the larger TIMP-2 transcript seen on diestrus and the high expression of mRNA in the theca, theca-lutein, and older CL is unknown. Perhaps TIMP-2 is highly abundant in the theca and theca-luteal cells to control basement membrane degradation by the MMPs before and during luteal formation. Alternatively, it is possible that TIMP-2 may be present to activate MMP-2, as a TIMP-2/membrane-type MMP complex is capable of activating MMP-2 in a spatially regulated manner (36). TIMP-2 has also been implicated in a number of functions that are similar to those of TIMP-1, including regulating the MMPs, inhibiting angiogenesis, acting as a growth factor, and altering cellular morphology (5, 6). Of note was the observation that CL from previous cycles had a higher overall level of TIMP-2 expression than the newly formed CL and also had localized regions of high expression. These regions of expression could be areas of more rapid regression within the CL, or they may be associated with remodeling of the vasculature. The high overall level of TIMP-2 mRNA in regressing CL from previous cycles points to a likely role in controlling the MMPs and tissue remodeling during luteal demise.
Expression of TIMP-3 showed an inverse pattern to TIMP-1, decreasing between proestrus and estrus in the whole ovary. These results are similar to those found in cycling mice, where TIMP-3 mRNA levels were highest on early proestrus (34). In other studies, TIMP-3 mRNA decreased in abundance 6 h after PMSG treatment of prepubertal rats and remained low through 48 h post-PMSG (35). Additionally, TIMP-3 mRNA was low during CL formation in early pseudopregnancy, but increased during the period of luteal maintenance and remained high throughout luteal regression (11). It is readily apparent that ovarian TIMP-3 is regulated differently and has different actions than the other TIMPs. The high levels on proestrus suggest that TIMP-3 is either needed for the final stages of follicular maturation before ovulation, or it has a role in regression of CL from the previous cycle. It is more likely that TIMP-3 is important for CL regression, given that TIMP-3 mRNA was not induced, but was decreased by gonadotropin stimulation of follicular development (35) and that TIMP-3 mRNA was high during regression of the pseudopregnant CL (11). In fact, TIMP-3 overexpression in rat aortic smooth muscle cells has been shown to induce cell death through apoptosis (16), a mechanism that may be important for CL regression, as apoptosis has been described in regressing CL (37, 38).
Investigation of TIMP-3 mRNA localization in the present study showed that TIMP-3 mRNA expression in the cycling ovary exhibited a similar pattern of follicular and stromal expression as reported in PMSG-treated and PMSG/hCG-treated prepubertal rats (31). In the cycling mouse, TIMP-3 mRNA was detected in the theca and granulosa of follicles and in CL, but was also present in the oocyte and in an atretic follicle (34). These observations of TIMP-3 mRNA in atretic murine follicles (34), the association between TIMP-3 and apoptosis in smooth muscle cells (16), along with the differential expression of TIMP-3 mRNA in granulosa cells of follicles of approximately the same size led us to investigate the expression of TIMP-3 in relation to the health of the follicle. The current findings that TIMP-3 mRNA was more abundant in granulosa cells of healthy follicles compared with adjacent atretic follicles is in contrast to these other previous reports. It appears that the level of TIMP-3 mRNA expression may reflect whether the follicle remains healthy or becomes apoptotic. As serial sections were not examined in these ovaries, it was impossible to determine the true size of a follicle from a section. Thus, it is possible that when follicles attain a certain size or stage of maturation their granulosa cells obtain the ability to produce TIMP-3. Follicles that are destined to undergo atresia may never reach this advanced stage of maturity, so their granulosa cells may never be able to produce TIMP-3 mRNA.
The differences in the rate of the apoptotic process may explain the lack of TIMP-3 in the granulosa cells of atretic follicles compared with expression in luteal tissue from older CL undergoing regression. Structural regression of the CL is a slow process that takes place over several subsequent cycles (39) as opposed to the comparatively rapid demise of the atretic follicle (40). This difference in the rate of structural change was manifest in a high level of apoptosis in atretic follicles, whereas CL from previous cycles exhibited far fewer apoptotic cells. In addition to its potential role in apoptosis, TIMP-3 has been associated with G1 progression during the cell cycle and with terminal cellular differentiation (41). It is thus feasible that TIMP-3 may play a role in the differentiation of granulosa cells during the final stages of follicular maturation or in transformation of granulosa to luteal cells in the newly forming CL.
One of the intriguing findings in this study was the high level of TIMP-2 and TIMP-3 mRNA expression in small luteinized follicles. Luteinized follicles, which are the thecal remnants of follicles that have undergone atresia, make up a large part of the interstitial tissue (23, 42). The hypertrophied theca of luteinized follicles produce progesterone at levels similar to those in the CL in pseudopregnant rabbits (42). Thus, these structures may influence the functions or characteristics of other ovarian structures. Although neither TIMP-2 nor TIMP-3 mRNA was detected in the granulosa cells of atretic follicles, they were both highly expressed in luteinized follicles and appeared to be significantly up-regulated on proestrus. The high level of TIMP-2 and TIMP-3 mRNA points to a role for these genes in the ECM remodeling and differentiation of luteinized follicles. As these follicles have LH receptors and are steroidogenic (43), TIMP-2 and TIMP-3 may also have important roles in other aspects of these follicles’ physiology, such as steroidogenesis.
In summary, these studies demonstrate for the first time that the TIMPs exhibit dynamic and yet distinct expression patterns within the cycling rat ovary. The high levels of TIMP-1 mRNA in the CL on estrus indicate a likely role for this inhibitor in luteinization. TIMP-2 may act more as a housekeeping gene, maintaining the structural integrity of the ovary during the constant tissue remodeling that takes place throughout the cycle. However, the higher level of one TIMP-2 transcript on metestrus and the changes in tissue distribution of TIMP-2 mRNA over the cycle suggest a more regulated control of this protein’s expression and function. Based upon the expression pattern of TIMP-3, it is reasonable to suggest that it plays a role in CL regression, although during follicular development TIMP-3 may be associated with the differentiation and growth of the follicle. This family of protein inhibitors warrants further study to elucidate the role each of these genes plays in the ovary during the estrous cycle. Most likely, each inhibitor contributes to the regulation of the MMPs, but the findings in other studies of diverse functions for the TIMPs implies that their capabilities within the ovary are complex and multifunctional.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received November 10, 2000.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
A, Chen TT, Wimalasena J, Caudle MR 1993 Cellular localization of luteinizing hormone receptor
immunoreactivity in the ovaries of immature, gonadotropin-primed and
normal cycling rats. Biol Reprod 48:1367–1382[Abstract]
This article has been cited by other articles:
 |
|
 |
|
  N. Tinfo and C. Komar Potential role for peroxisome proliferator-activated receptor {gamma} in regulating luteal lifespan in the rat Reproduction, January 1, 2007; 133(1): 187 - 196. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Jo and T. E. Curry Jr. Luteinizing Hormone-Induced RUNX1 Regulates the Expression of Genes in Granulosa Cells of Rat Periovulatory Follicles Mol. Endocrinol., September 1, 2006; 20(9): 2156 - 2172. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. McDonnel Smedts and T.E. Curry Jr. Expression of Basigin, an Inducer of Matrix Metalloproteinases, in the Rat Ovary Biol Reprod, July 1, 2005; 73(1): 80 - 87. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Q. Li, L. J Bakke, J R. Pursley, and G. W Smith Localization and temporal regulation of tissue inhibitors of metalloproteinases 3 and 4 in bovine preovulatory follicles Reproduction, November 1, 2004; 128(5): 555 - 564. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Okamoto, R. Niu, and S. Yamada Increased expression of tissue inhibitor of metalloproteinase-2 in clear cell carcinoma of the ovary Mol. Hum. Reprod., October 1, 2003; 9(10): 569 - 575. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. S. Simpson, C. M. Komar, and T. E. Curry Jr Localization and Expression of Tissue Inhibitor of Metalloproteinase-4 in the Immature Gonadotropin-Stimulated and Adult Rat Ovary Biol Reprod, January 1, 2003; 68(1): 214 - 221. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. L. Espey and J. S. Richards Temporal and Spatial Patterns of Ovarian Gene Transcription Following an Ovulatory Dose of Gonadotropin in the Rat Biol Reprod, December 1, 2002; 67(6): 1662 - 1670. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. E. Curry Jr and S. E. Wheeler Cellular Localization of Tissue Inhibitors of Metalloproteinases in the Rat Ovary Throughout Pseudopregnancy Biol Reprod, December 1, 2002; 67(6): 1943 - 1951. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. J. Bakke, M. P.D. Dow, C. A. Cassar, M. W. Peters, J. R. Pursley, and G. W. Smith Effect of the Preovulatory Gonadotropin Surge on Matrix Metalloproteinase (MMP)-14, MMP-2, and Tissue Inhibitor of Metalloproteinases-2 Expression Within Bovine Periovulatory Follicular and Luteal Tissue Biol Reprod, June 1, 2002; 66(6): 1627 - 1634. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. M. Komar and T. E. Curry Jr Localization and Expression of Messenger RNAs for the Peroxisome Proliferator-Activated Receptors in Ovarian Tissue from Naturally Cycling and Pseudopregnant Rats Biol Reprod, May 1, 2002; 66(5): 1531 - 1539. [Abstract] [Full Text]
|
|
|
|
|
|
T. E. Curry Jr., L. Song, and S. E. Wheeler Cellular Localization of Gelatinases and Tissue Inhibitors of Metalloproteinases During Follicular Growth, Ovulation, and Early Luteal Formation in the Rat Biol Reprod, September 1, 200 |
