Serotonin-induced MMP-13 Production Is Mediated via Phospholipase C, Protein Kinase C, and ERK1/2 in Rat Uterine Smooth Muscle Cells*

Serotonin (5-hydroxytryptamine; 5-HT), acting via the 5-HT 2A receptor, up-regulates the transcription and pro- duction of interstitial collagenase (matrix metallopro-teinase-13; MMP-13), a critical enzyme responsible for maintaining the integrity of the uterus, after parturi-tion. Serotonin treatment of rat uterine myometrial smooth muscle cells induced inositol phosphate (IP) turnover, which was abolished by the 5-HT 2A receptor- specific antagonists ketanserin and spiperone. The phospholipase C (PLC) inhibitors U73122 and D609 attenuated serotonin-mediated-IP turnover with a corresponding inhibition of MMP-13 protein production. Subsequent recovery of both MMP-13 protein expression and IP generation was seen following the removal of D609. Protein kinase C (PKC) activators, the diacylglyc-erol analogue 1,2-dioctanoyl- sn -glycerol and phorbol myristate acetate (PMA), mimicked the effect of serotonin on MMP-13 protein expression; prolonged PMA treatment (which down-regulates PKC) lowered MMP-13 protein levels. The PKC-specific inhibitors bisindolylmaleimide I, calphostin C, CGP 41251, and the PKC (cid:1)

The maintenance of the three-dimensional architecture of the mammalian uterus is carefully regulated during pregnancy and after parturition. Production of fibrillar collagens (Types I and III) is up-regulated 10-fold in the uterus during pregnancy (1,2). These collagens possess a rigid superhelical structure, which is responsible for providing the tensile strength necessary to accommodate the force exerted by the growing fetus. After parturition, a precisely programmed non-necrotic and non-inflammatory process of connective tissue remodeling occurs, which results in the rapid and dramatic degradation of the collagen amassed and returns the uterus to reproductive competence (3). This transition is initiated by the enzyme interstitial collagenase (matrix metalloproteinase-13; MMP-13), 1 a member of the matrix metalloproteinase family. It acts as not only the rate-limiting step but also the committed step in the catalysis of the degradation of extracellular collagen (4,5). The expression of MMP-13 occurs during postpartum involution and is not present either during the pregnancy or even within hours prior to parturition. The myometrial smooth muscle cell of the uterus has been identified not only as the source for MMP-13 (6) but also that of its natural substrate, collagen (7). Thus, the myometrial smooth muscle cell provides an interesting paradigm for not only the synthetic needs of the uterus during pregnancy but also the degradative events that occur postpartum.
Our laboratory has focused on MMP-13 and the regulatory mechanisms involved in the restructuring of the postpartum uterus by this critical enzyme. To this end, we have developed a primary cell culture system with rat myometrial smooth muscle cells (SMCs) that produce MMP-13 (8,9). We have found that the indoleamine serotonin (5-hydroxytryptamine, 5-HT) is an obligate inducer of MMP-13 in both rat and human myometrial SMCs (8,10). Nuclear run-on analysis demonstrated that serotonin treatment in rat myometrial SMCs increased MMP-13 gene transcription by 6-fold, which corresponded to maximal steady-state mRNA and protein production following exposure to serotonin (11). Additionally, we have determined that interleukin-1 (IL-1) is induced in a serotonindependent manner and is itself required for the production of MMP-13 (12). The requisite IL-1 presence acts to not only induce autocrine expression of the genes for IL-1␣ and ␤ but also that of transin (rat stromelysin) and 92 kDa gelatinase (type IV collagenase), 2 as well as the gene for IL-6. 3 Interestingly, we have also found that serotonin acts to down-regulate the production of genes responsible for the hypertrophic nature of the uterus during pregnancy. Serotonin exerts negative regulatory effects upon the genes for types I and III collagen, fibronectin, and the anti-protease ␣ 2 -macroglobulin (11,13). Of particular note, a single serotonin receptor * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
ʈ To whom correspondence should be addressed. Tel.: 518-262-8791; Fax: 518-262-6161; E-mail: melenda@mail.amc.edu. is responsible for mediating both the positive-and negativeregulated serotonin-dependent processes in myometrial smooth muscle cells: the 5-HT 2A subtype (14). Although these divergent biochemical processes are mediated through a single receptor subtype, they possess different postreceptor mechanisms. Specifically, while the serotonin-dependent upregulated genes, MMP-13, transin, 92 kDa gelatinase, IL-1, and IL-6, appear to require IL-1, the down-regulated genes, types I and III collagen, fibronectin, and ␣ 2 -macroglobulin, are independent of IL-1 influence. Furthermore, phorbol esters, known activators of protein kinase C (PKC), mimic the effects of serotonin on the serotonin-dependent-induced genes while having no effect on the serotonin-dependent down-regulated genes (13,15). Conversely, 8-bromoadenosine 3Ј:5Ј-cyclic monophosphate (8-bromo-cyclic AMP) mimics the effects of serotonin on the negatively regulated gene products for type I collagen and ␣ 2 -macroglobulin (11,13). Thus, the serotonin-mediated induction of MMP-13 and accompanying down-regulation of collagen appears to be via two distinct signaling pathways that must be stringently controlled to preserve the integrity of the uterus during pregnancy.
The 5-HT 2A receptor subtype possesses the 7 transmembrane domains characteristic of G-protein-coupled receptors. Binding of serotonin has been shown to increase intracellular Ca 2ϩ levels (16,17) and to stimulate phospholipase C (PLC) for the production of inositol phosphates (18 -21) and diacylglycerol (DAG), both activators of PKC. Like other G-protein-coupled receptors, this activation of PKC through the 5-HT 2A receptor stimulates the mitogen-activated protein kinase (MAPK) cascade, leading to the activation of the extracellular signal-regulated kinases 1 and 2 (ERK1/2) (22)(23)(24). Although many postreceptor signaling cascades have been described for the 5-HT 2A receptor in other systems, little is known about the signaling pathway(s) involved in serotonin-mediated induction of MMP-13 in rat myometrial SMCs. This study indicates that serotonin-induced MMP-13 protein production requires the sequential activation of a phospholipase C pathway, protein kinase C, and ERK1/2.
Cell Culture-Primary cell cultures of rat myometrial smooth muscle cells were obtained from 4-day postpartum uteri of Sprague-Dawley rats and used in all experiments. Uteri were minced, washed, and then incubated in 0.05% trypsin in DMEM containing 30 mM HEPES, 200 units/ml penicillin, and 200 g/ml streptomycin for 1 h at 37°C. The uterine tissue was then pipetted for ϳ20 min/uterus with a 10-ml pipette and then centrifuged at 500 rpm for 5 min. The supernatant was retained, and the cells were pelleted following centrifugation at 1200 rpm for 10 min, rinsed, and recentrifuged at 1200 rpm for 10 min. The cells were then resuspended in DMEM supplemented with 10% FBS (v/v), 30 mM HEPES, 200 units/ml penicillin, and 200 g/ml streptomycin. Cells were plated at a density of 100,000 cells/cm 2 and maintained in DMEM supplemented with 10% FBS (v/v), 30 mM HEPES, 200 units/ml penicillin, and 200 g/ml streptomycin. Upon reaching confluence, the media was then changed to DMEM supplemented with 10% serotonin-free FBS (CS-FBS) (v/v) that had been adsorbed with dextran-coated charcoal (10,25), 30 mM HEPES, 200 units/ml penicillin, 200 g/ml streptomycin, and 100 ng/ml LPS and were maintained under these conditions until experimental treatments began.
Inositol Phosphate Assay-Cells were labeled for 18 -24 h with 0.25 Ci/ml myo-[ 3 H]inositol in inositol-free and serum-free DMEM supplemented with 30 mM HEPES, 200 units/ml penicillin, 200 g/ml streptomycin, and 100 ng/ml LPS. Immediately prior to experimental treatments, cells were washed with 1-2 ml of inositol-free DMEM. Cells were pretreated in 0.5 ml of the same media containing 20 mM LiCl and, if necessary, the appropriate inhibitors and vehicles for the inhibitors for 30 -60 min. 5 M serotonin (or appropriate concentration of agonist) was then added to challenge the cells (1 h, 37°C). After the appropriate incubation time had passed, the cells were washed with 2 ml of phosphate-buffered saline (PBS) and aspirated, and 0.75 ml of ice-cold 20 mM formic acid was then added to stop the reaction. The cells were then placed at Ϫ20°C for Ն1 h.  (26,27).
Protein Extraction of MMP-13-Primary cultures of myometrial smooth muscle cells upon reaching confluence were incubated with appropriate compounds, including vehicle controls, in inositol-free and serum-free DMEM for varying times dependent upon treatments, and the media was then removed. Each sample of media was incubated overnight at 4°C with a 50-l slurry of heparin-Sepharose beads in 0.05 M Tris/0.01 M CaCl 2 /0.2 M NaCl per 1.5-2.0 ml of media to preferentially bind the MMP-13 protein. Dextran sulfate (50 g/ml 0.05 M Tris, 0.01 M CaCl 2 ) was then used to elute the collagenase off the heparin-Sepharose beads.
PKC Isoforms-For protein extraction of PKC isoforms, cells were pretreated for 1 h with the inhibitors and the appropriate vehicle controls, followed by serotonin for 7.5 min, rinsed 2ϫ with ice-cold PBS and scraped off 75 cm 2 flasks with 700 l of lysis buffer containing protease inhibitors (5 mM benzamidine, 50 g/ml leupeptin, 50 g/ml aprotinin, 50 g/ml trypsin inhibitor, 5 g/ml pepstatin A, 1 mM phenylmethylsulfonyl fluoride, 20 mM NaF, 1 mM Na 3 VO 4 , 1 mM paranitrophenyl phosphate, and 5 mM imidazole). The cells were then sonicated (for 40 s at 50% duty cycle and with pulsation) and spun at 4°C for 45 min at 55,000 rpm. The cytosolic fraction (supernatant) was retained while the pellet was resuspended in 200 l of lysis buffer and 1% Triton X-100. The tubes were then vortexed, left on ice for 20 min, and spun at 14,000 rpm in a microcentrifuge at 4°C for 5 min. The membrane fraction (supernatant) was retained and the insoluble fraction (pellet) resuspended in 2ϫ sample buffer. Protein determination assays were performed with the Pierce BCA protein assay kit or the Bio-Rad Laboratories protein assay.
ERK1/2-The ERK1/2 proteins were isolated from cells plated in 75-cm 2 flasks. Inhibitors and vehicle controls were preincubated for 1 h, and then serotonin was added for 7.5 min, with the exception of the time course studies, and the cells were rinsed 2ϫ with ice-cold PBS and then scraped with 250 l of PBS. The cells were spun down for 30 min at 14,000 rpm at 4°C and then lysed with hypotonic buffer (20 mM HEPES, 10 mM KCl, 1 mM EDTA, 10% glycerol, 20 mM NaF, and 1 mM dithiothreitol) containing protease inhibitors (3 g/ml aprotinin, 0.1 mM Na 3 VO 4 , and 1 mM phenylmethylsulfonyl fluoride) and 0.2% Nonidet P-40 for 10 min on ice. The samples were spun at 13,000 rpm for 2 min at 4°C and the supernatant retained as the cytosolic fraction. The pellet was resuspended in high salt hypotonic buffer (hypotonic buffer as described above with 20% glycerol and 420 mM NaCl) and protease inhibitors and rocked overnight at 4°C. The next day, the samples were spun at 4°C for 10 min at 13,000 rpm and the supernatant retained as the nuclear fraction. Protein determination assays were performed as noted above.

Serotonin-induced Up-regulation of MMP-13
Is PLC-dependent Effects of Serotonin, DOI, and 5-HT 2A Receptor Antagonists on Inositol Phosphate Production-Serotonin-mediated MMP-13 production occurs through the 5-HT 2A receptor, and this receptor is known to mediate phospholipase C activation (18 -21). We have previously demonstrated that 5-HT 2A receptor-selective agonists (1-(2,5-dimethoxy-4-iodophenyl)-2aminopropane HCL (DOI) and quipazine) and antagonists (ketanserin and spiperone) can mimic and inhibit, respectively, the effects of serotonin on MMP-13 production (14). To determine if binding of serotonin to the 5-HT 2A receptor (which ultimately results in MMP-13 up-regulation) induces PLC activity, rat uterine smooth muscle cells were challenged with serotonin (5 M) or the 5-HT 2A receptor agonist DOI (5 M), and IP generation was quantified. Serotonin increased inositol phosphate production to levels ϳ10-fold over basal (Fig. 1A). This serotonin-mediated response was blocked in the presence of the 5-HT 2A receptor-specific antagonists ketanserin (1 M) or spiperone (1 M) (Fig. 1B). The receptor agonist DOI mimicked the effects of serotonin on IP production (Fig. 1A), an effect that was attenuated by spiperone (Fig. 1C). These results indicate that in rat myometrial smooth muscle cells, serotonin can activate PLC through the 5-HT 2A receptor subtype.
Effects of the PI-PLC Inhibitor U73122 and Its Inactive Analogue U73343 on Inositol Phosphate Turnover and MMP-13 Protein Expression-To further confirm the role of phospholipase C in the production of MMP-13, the PLC inhibitor U73122 and its inactive analogue U73343 were used to determine their effects on serotonin-induced inositol phosphate turnover. U73122 inhibited the release of inositol phosphates ϳ50% while the inactive analogue U73343 had no effect (Fig.  1D). This inhibitory response was dose-dependent with maximal inhibition at 20 M U73122. At this same concentration, U73122 was able to attenuate the serotonin-dependent production of MMP-13 protein expression following serotonin treatment while U73343 did not exert a significant effect on MMP-13 protein levels (Fig. 1, E and F). The two bands seen in the MMP-13 standard are representative of the pro-form of MMP-13 and its active counterpart, which has approximately a 10-kDa loss in mass.
Effects of the PC-PLC Inhibitor D609 on Inositol Phosphate Turnover and MMP-13 Protein Levels-Initially, we utilized the PC-PLC inhibitor tricyclodecan-9-yl-xanthogenate (D609) as a negative control for our PI-PLC activity assays. Interestingly, measurements of IP generation in the presence of serotonin revealed that D609 depressed inositol phosphate production in a dose-dependent manner ( Fig. 2A); the maximal effect was seen at 250 M. This inhibition maximally decreased IP turnover by ϳ50% (Fig. 2, A and B), as seen with U73122. Similarly, MMP-13 protein production was attenuated by D609 and reached levels comparable to those of the negative control by 150 M D609 (Fig. 2C). Although the cells were viable after D609 treatment, we wanted to determine if there had been some detrimental effects to their cellular functions that might account for the inhibitory effects on IP production. To that end, we examined if the effects of D609 on serotonin-stimulated PLC activity and MMP-13 production were reversible. D609 was incubated overnight and then washed out of the cells prior to treatment with serotonin. Recovery of serotonin-mediated IP production was seen 1-day post-D609 treatment and IP turnover was higher than that of the control (Fig. 2D, left). However, in the days following, the levels of IP production equalized between the cells that had been treated with D609 and controls. With respect to D609 effects on MMP-13 levels, protein production recovered within 6 h of D609 removal and serotonin challenge (Fig. 2D, right). The particular control sample (ϩ5HT) at 6 h with nearly undetectable MMP-13 production was likely due to the occasional differences seen in varying primary cell culture preparations. These results suggest that PC-PLC may regulate IP generation and MMP-13 production in rat uterine SMCs by an as yet unknown mechanism.

Role of PKC in MMP-13 Expression
DOG and PMA Regulate MMP-13 Expression-The second messenger DAG, produced by the activation of phospholipase C, is a known activator of protein kinase C. To examine the role of PKC in MMP-13 production, cells were challenged with DOG, a cell-permeable DAG analogue, for 6 h. DOG was able to induce the production of MMP-13 to a level comparable to that of serotonin (Fig. 3A). Wilcox et al. (15) have shown that shortterm exposure to phorbol myristate acetate mimics the effects of serotonin on MMP-13 protein production in myometrial smooth muscle cells. On the other hand, prolonged exposure to PMA is known to deplete PKC (29 -31). Cells were pretreated for 24 h with 25 nM PMA to down-regulate phorbol estersensitive PKC isoforms and then challenged with either 5 M serotonin or 25 nM PMA for 6 h. Control cells were not exposed to PMA and were treated in concert with the pretreated cells, and the media collected after 6 h. Acute exposure to PMA stimulates the production of higher levels of MMP-13 protein than that of the serotonin-stimulated response (Fig. 3A). However, cells that had been pretreated with PMA to deplete PKC protein expressed significantly lower levels of MMP-13 protein as compared with cells exposed to either serotonin or PMA for only 6 h. Following down-regulation with PMA, subsequent agonist challenge with either serotonin or PMA did not elicit a comparable response to that of controls. These results support a role for PKC in the regulation of MMP-13 expression in rat myometrial SMCs.  (Fig. 3B). These isoforms represent all three classes of PKC, classical (␣, ␤; cPKC), novel (␦, ⑀; nPKC) and atypical (/, ; aPKC). The PKC isoforms and ␦ appear to be the most highly expressed within the rat myometrium while the ␥ isoform was undetectable.
Protein Kinase C-␣, Ϫ␤, Ϫ␦, Ϫ⑀, and -Isoform Expression in Cytosol and Membrane Fractions following Serotonin Treatment-The translocation of PKC isoforms from the cytosol to the membrane is considered to be a determinant of PKC activation (32). To identify which PKC isoforms in particular participate in the signaling cascade initiated by serotonin, which results in MMP-13 production, we examined the expression of various PKC isoforms in both the cytosol and membrane fractions of cell lysates following serotonin treatment. Upon the addition of serotonin, there was a marked decrease (Ϫ21%) in PKC␦ expression in the cytosolic fraction of cell lysates and a corresponding increase (ϩ32%) in the membrane fraction (Fig.  3C). PKC⑀ also showed a similar expression pattern following serotonin treatment with increased translocation to the mem-brane (ϩ36%). On the other hand, PKC␣ expression was present in the cytosolic fractions of all the experimental treatments with an increase in expression after exposure to serotonin. However, PKC␣ expression was nearly undetectable in the membrane fractions of the cell lysates of rat myometrial smooth muscle cells. Neither PKC␤ nor PKC exhibited translocation from the cytosol to the membrane following serotonin treatment. These results are consistent with a role for the nPKC isoforms ␦ and/or ⑀ in the signaling pathway from PLC to MMP-13 protein production.

Effects of the Protein Kinase C Inhibitors Bisindolylmaleimide I, CGP 41251, and Calphostin C on MMP-13 Protein Production-
The PKC activator DOG up-regulated the production of MMP-13 in rat myometrial smooth muscle cells. Activation of PKC by acute exposure to PMA, and the depletion of PKC by prolonged PMA exposure resulted in the up-regulation and down-regulation, respectively, of MMP-13 protein levels. These factors then led to further examination of the role of PKC in this system. Bisindolylmaleimide I and CGP 41251, two potent Cells were grown and maintained as described in the legend to Fig. 1. A, DOG, a DAG analogue, was added for 6 h and MMP-13 protein extracted as described under "Experimental Procedures." Cells were pretreated with 25 nM PMA for 24 h prior to serotonin/PMA challenge for an additional 6 h. Control cells were not exposed to a pretreatment of PMA. MMP-13 protein was collected and Western blot analysis performed as previously described (n ϭ 2). B, PKC protein was isolated as described under "Experimental Procedures." 400 g of total protein of cell lysates was run on a 7.5% SDS-PAGE gel for Western blot analysis and antibody incubations (PKC-MW kDa: ␣-77, ␤-77, ␥-78, ␦-78, ⑀-83, -74, -74, -78, and -68) were performed in a Hoeffer PR 150 Deca-Probe Incubation Manifold (n ϭ 2). C, protein fractions of cell lysates were run on a 7.5% SDS-PAGE gel for Western blot analysis (n ϭ 2). Numbers in the table are densitometry readings that correspond to percent change from control Ϯ S.E. N.D., not determined experiment). S.E. for (D609ϩ5HT) samples were Ͻ0.5 and therefore, not visible on the graph. *, p Ͻ 0.05; comparison to control (ϩ5HT) (A) and comparison to control (Ϫ5HT) (B). C, Western blot analysis was performed on cells pretreated with varying concentrations of D609 for 1 h prior to serotonin challenge. After 6 h, the media was removed and collagenase protein extracted with heparin-Sepharose beads, and equal volumes were run on a 12% SDS-PAGE gel. Data represent the mean Ϯ S.E. of triplicate determinations (n ϭ 3) (3 wells/treatment/experiment). ⌬ and *, p Ͻ 0.05 where ⌬ is comparison to the control, Ϫ5HT and * is comparison to the control, ϩ5HT. D, for the recovery treatments, the cells were incubated overnight with D609 (250 M). D609 was removed the next day, the cells washed, and serotonin was added for the times indicated. IP assays were performed on the cells (D, left panel), and MMP-13 protein was extracted as described in the text (D, right panel). staurosporine analogues that act as competitive inhibitors on the ATP-binding site of PKC and are more selective for the cPKCs, were able to inhibit the serotonin-mediated signal transduction pathway and attenuated the MMP-13 protein production at a concentration of 10 nM (Fig. 4, A and B). Calphostin C, a competitive inhibitor of the diacylglycerolbinding site on the regulatory domain of PKC, was also able to significantly inhibit the production of MMP-13 protein at a concentration of 10 nM (Fig. 4C). These results strongly suggest that a DAG-activated PKC is necessary for serotonin-mediated MMP-13 up-regulation in rat uterine SMCs.
Effects of the Protein Kinase C-␦ Isoform-selective Inhibitor Rottlerin on MMP-13 Protein Levels-Due to the fact that the PKC isoform ␦ is one of the highly expressed isoforms in the postpartum rat uterus, rottlerin, a PKC␦-selective inhibitor, was used to examine the role of that isoform in serotoninmediated signaling. Rottlerin was able to attenuate serotonininduced MMP-13 protein production in a dose-dependent manner (Fig. 5). At a concentration of 500 nM, rottlerin was able to inhibit MMP-13 protein accumulation to levels similar to that of the negative control. These data, in concert with those in Fig.  3, provide evidence for PKC, possibly the ␦ isoform, in regulating the signaling cascade initiated by serotonin in the production of MMP-13.
Serotonin-mediated ERK Phosphorylation and Effects of the MEK Inhibitor PD98059 -PKC has been shown to activate, coupled to the 5-HT 2A receptor, the MAP kinase signaling pathway of which ERK1/2 is a downstream effector (22)(23)(24). Dual phosphorylation of ERK1/2 by MEK has been correlated to the activation of ERK1/2 (33)(34)(35). To determine whether ERK1/2 activity is stimulated by serotonin, ERK1/2 phosphorylation over time was measured. ERK1/2 phosphorylation was detected 5 min after exposure to serotonin with maximal phosphorylation achieved within 10 min (Fig. 6A) within both the cytosolic and nuclear fractions. ERK1/2 phosphorylation then diminished at 15 min and returned to basal levels within 30 min following serotonin treatment. This result indicates that serotonin links to the MAP kinase cascade, resulting in ERK1/2 activation.
Having determined that serotonin induced ERK activation, we wanted to determine if ERK phosphorylation was required for MMP-13 protein production. If ERK phosphorylation is necessary for MMP-13 expression, blocking the activation of ERK should in turn attenuate the effects of serotonin in the up-regulation of MMP-13. We found that the MEK inhibitor PD98059, with the inhibition of ERK1/2, was able to suppress MMP-13 protein expression at a concentration of 25 M (Fig.  6B). Furthermore, PD98059 also abrogated the effects of sero- tonin on ERK1/2 activation at that same concentration (Fig.  6C). These data further implicates the MAP kinase pathway and ERK1/2 in regulating serotonin-mediated MMP-13 expression.
Effects of PKC Inhibitors on ERK Phosphorylation by Serotonin-The PKC inhibitors rottlerin and calphostin C were used to further correlate the relationship between PKC, the MAP kinase cascade, and MMP-13 protein expression. If ERK lies downstream of PKC along the signaling pathway initiated by serotonin, PKC inhibitors that attenuate MMP-13 protein production should correspondingly inhibit ERK phosphorylation. Both rottlerin (500 nM) and calphostin C (50 nM) were able to inhibit the activation of ERK following serotonin treatment (Fig. 6D). Taken together, these results with those shown in Figs. 4 and 5, strongly suggest that PKC mediates serotonin induction of MMP-13 protein levels through a downstream effector, ERK1/2.

DISCUSSION
The purpose of these studies was to identify and order the major players in the signal transduction pathway that is initiated by serotonin binding to the 5-HT 2A receptor and results in the production of MMP-13. The 5-HT 2A receptor is a G-proteincoupled receptor (36) that has been linked to PLC in a variety of cell systems, including the rat cerebrum, aorta, and uterus (18,20,21). Previous studies from this laboratory have demonstrated that the 5-HT 2A receptor is the only serotonin receptor subtype mRNA detectable in rat uterine smooth muscle cells (14). Furthermore, this is the receptor that differentially regulates expression of a number of genes. Serotonin up-regulated genes include IL-1 (12), IL-6, 3 transin, and 92 kDa gelatinase, 2 while types I and III collagen, fibronectin, and ␣ 2macroglobulin are down-regulated by serotonin (11,13). While our previous studies have focused on the transcriptional regulation and enzymatic activity of MMP-13 (11,15,37), here we have extended those observations to the level of immunoreactive protein. The results of this study further clarify the signaling mechanisms by which serotonin, acting through the 5-HT 2A receptor, mediate the production of MMP-13 and demonstrate that phospholipase C, its downstream effector PKC and the MAP kinase cascade are integral components in this cell system.
Our present results indicate that PLC activation is required for MMP-13 protein expression. As illustrated in Fig. 1, serotonin was able to markedly increase IP production specifically through the 5-HT 2A receptor subtype. Selective and well-characterized 5-HT 2A receptor agonists (38,39) were able to mimic the effects of serotonin on IP production and conversely, 5-HT 2A receptor selective antagonists (40,41) abrogated the effects of serotonin on IP turnover. Consistent with the hypothesis that PLC is a crucial element of this signaling pathway, the PLC-selective inhibitor U73122 abolished the increased IP turnover and decreased the production of MMP-13 protein. Surprisingly, we found that the putative phosphatidylcholinespecific PLC inhibitor D609 (42,43) significantly and dose-dependently exerted inhibitory effects upon both the phosphatidylinositol-specific PLC pathway and MMP-13 production in a reversible manner, as shown in Fig. 2. There have been recent reports in which D609 was found to not only exert effects upon PC-PLC but also PLD (44,45) and PI-PLC (46,47). These results do not preclude the possible role of PC-PLC or PLD in MMP-13 protein expression. However, our results can serve to further emphasize the role of DAG in regulating MMP-13 expression since DAG is a major product of both PLC, whether it is derived from PI-or PC-PLC, and PLD activity. Furthermore, a DAG analogue alone was able to up-regulate MMP-13 protein expression to levels comparable to that of serotonin.
Following a logical progression from PLC to the activation of PKC by DAG and previous studies where PMA was found to stimulate MMP-13 expression in human (48) and rabbit (49) fibroblasts, as well as rat myometrial smooth muscle cells (15), we examined the possible role of PKC in mediating serotonin signaling. Although expression of PKC isoforms following prolonged PMA exposure were not examined in this study, phorbol ester-induced down-regulation of PKC isoforms has previously been demonstrated in both the human and rat myometrium (50,51). In our rat myometrial smooth muscle cell system, exposure to PMA for 24 h was sufficient to attenuate the production of MMP-13 protein. Furthermore, the PKC-selective inhibitors bisindolylmaleimide I, CGP 41251, and calphostin C, which have different sites of action upon PKC, were all able to abrogate the effects of serotonin on MMP-13 protein level. These inhibitors exert their effects on either the ATPbinding site (bisindolylmaleimide I, CGP 41251) or the DAGbinding site of PKC (calphostin C), respectively, and provided further evidence that serotonin-mediated MMP-13 production follows a traditional signaling pathway from PLC to PKC.
In identifying which PKC isoforms are present in primary cultures of 4-day postpartum rat myometrial SMCs, we found members from all three classes: classical (␣, ␤), novel (␦, ⑀), and atypical (/, ). Of these three PKC subtypes, only the classical FIG. 5. Effects of the PKC␦-selective inhibitor rottlerin on interstitial collagenase protein production. The cells were pretreated with varying concentrations of rottlerin for 1 h prior to serotonin challenge. After 6 h, the media was collected and interstitial collagenase protein was extracted with heparin-Sepharose beads, and equal volumes were run on a 12% SDS-PAGE gel. Data represent the mean Ϯ S.E. of quintuple determinations (n ϭ 5) (3 wells/treatment/ experiment). ⌬ and *, p Ͻ 0.05; ⌬ is comparison to the control (Ϫ5HT) and * is comparison to the control, ϩ5HT. and novel isoforms possess a DAG requirement. Both DAG analogues, the phorbol ester PMA and DOG, were able to mimic the effects of serotonin on MMP-13 production. This indicated a likely role for one or more DAG-regulated PKC isoform(s) and, taking advantage of the PKC␦-selective inhibitor rottlerin, we focused on the highly expressed cPKC isoform ␦. As seen in Fig. 5, rottlerin significantly inhibited MMP-13 protein production. To determine if other PKC isoforms participated in the signaling pathway, we examined PKC isoform translocation, as a measure of activity, from the cytosol to the membrane upon serotonin treatment.
Expression of both the ␦ and ⑀ isoform diminished within the cytosol while showing a corresponding increase within the membrane fraction. These studies are consistent with a PKC requirement for MMP-13 expression and implicate the PKC␦ isoform in this signal transduction pathway. Although neither of the two classical PKC isoforms, ␣ or ␤, showed detectable translocation to the membrane following serotonin treatment, we cannot dismiss the possibility that they may indeed contribute to the overall induction of MMP-13 since bisindolylmaleimide I and CGP 41251 are selective for the classical PKC isoforms. There may also be a role for PKC⑀ in this signaling mechanism as evidenced by its increased expression at the membrane following serotonin treatment. Despite the expression of the atypical PKC isoforms in this cell system, it seems unlikely that they influence serotonin-mediated MMP-13 pro-duction in light of the potent inhibitory effects of calphostin C, a competitive inhibitor for the DAG-binding site.
The activation of the MAP kinase cascade by PKC has been shown to be involved in 5-HT 2A receptor subtype signaling events (22)(23)(24). Furthermore, the PKC␦ isoform has been FIG. 6. Effects of serotonin, PD98059, and PKC inhibitors on ERK phosphorylation. A, serotonin was added as indicated. Cytoplasmic and nuclear fractions were separated as previously described, and the protein fractions were run on a 10% SDS-PAGE gel for Western blot analysis (n ϭ 3). B, PD98059 was added 1 h prior to serotonin treatment and MMP-13 protein extracted after 6 h of serotonin treatment (n ϭ 4). ⌬ and *, p Ͻ 0.05; ⌬ is comparison to the control, Ϫ5HT and * is comparison to the control, ϩ5HT. C, serotonin was added for 7.5 min for optimal detection of ERK phosphorylation (n ϭ 2) Samples were run on 12% (MMP-13) and 10% (phospho-ERK) SDS-PAGE gels, respectively. D, rottlerin (500 nM) and calphostin C (50 nM; light-activated) were preincubated for 1 h prior to the addition of serotonin (7.5 min)(n ϭ 2). shown in a variety of other cell types, mainly epithelial in origin but not in myometrial SMCs, to activate the ERK pathway (52)(53)(54). Ueda et al. (55) were able to demonstrate this same effect in COS1 cells transfected with constitutively active PKC␦. Previous studies have shown that increased MMP-13 mRNA expression in chondrocytes (56,57), osteoblasts (58), and synoviocytes (59) are dependent upon ERK1/2 activation. In contrast, other studies have shown that activated ERK1/2 inhibited MMP-13 expression in human dermal fibroblasts (60), and the inhibition of ERK1/2 activity in transformed keratinocytes had no effect upon MMP-13 expression (61). Our studies found that ERK1/2 activation occurs rapidly and subsequently diminishes by 30 min. Utilizing the MEK inhibitor PD98059, we were able to block both ERK phosphorylation and MMP-13 protein production at the same concentrations. This demonstrates that not only does the MAP kinase cascade occur as a result of serotonin treatment but also that ERK phosphorylation is necessary for MMP-13 protein production. These disparate data between cellular systems may be, in part, due to the distinct functions that MMP-13 has in mammalian physiology. The nature of uterine biology and the drive to return to reproductive competence may necessitate the divergent signaling regulation from that seen in wound healing and other physiological processes. The similarities seen in signaling mechanisms associated with MMP-13 expression in uterine remodeling and the pathology of arthritis may perhaps be correlated with the critical role of cytokines in both these processes (12,62).
To better delineate the relationship between PKC and ERK1/2, the PKC inhibitors rottlerin and calphostin C were used to examine their effects on ERK phosphorylation. Both inhibitors were indeed able to attenuate ERK1/2 activation by serotonin indicating that PKC activation precedes that of ERK1/2 and that the MAP kinase cascade has an integral part in the overall maintenance of uterine integrity. Activated ERK1/2 levels were increased in both the cytosolic and nuclear fractions following serotonin treatment. While phosphorylated nuclear ERK1/2 may modulate gene transcription, there may be other possible functions for cytoplasmic ERK1/2 in MMP-13 expression and secretion. Liu et al. (63) demonstrated that intracellular levels of MMP-9 and MMP-2 did not parallel the suppression of MMP-9/MMP-2 secretion by dominant negative Ras and thus, indicated that there may be a post-transcriptional mechanism for ERK1/2 regulation of MMP secretion. In terms of what may be occurring downstream of ERK1/2 activation, recent studies from our laboratory have identified the AP-1 site on the rat MMP-13 promoter, along with the extended palindrome that surrounds it, as a requirement for MMP-13 transcription (64). Although the AP-1 site is necessary, it is not sufficient for MMP-13 production.
The role that serotonin-dependent IL-1 expression plays in the above signaling pathway has not been addressed in the present study. Liacini et al. (56) have recently shown that the IL-1-induced MMP-13 expression in articular chondrocytes can be reduced by inhibitors to ERK1/2. Interestingly, Mengshol et al. (65) found that IL-1 induction of MMP-13 in chondrocytes was almost completely abolished by the p38 inhibitor SB203580 whereas the MEK inhibitor PD98059 had no effect on IL-1-dependent MMP-13 protein expression. It may be that IL-1 mediates the phosphorylation of AP-1 binding proteins through p38, which then bind to the AP-1 site in response to serotonin, while serotonin mediates its responses via ERK1/2. Indeed, IL-1 gene expression was demonstrated to be more dependent upon the activation of p38 than of ERK1/2 (66). Another possibility is suggested by recent studies from Reunanen et al. (67) where they found that p38 activation not only induces MMP-1 (collagenase-1) expression but also stabilizes the mRNA by ϳ15-fold. Furthermore, an earlier report demonstrated ERK1/2 activation alone potently up-regulates MMP-1 mRNA levels whereas p38 activation has no effect on the MMP-1 promoter (68). IL-1 itself has been shown to stabilize human collagenase mRNA (69). It is likely that IL-1 modulates MMP-13 production in an autocrine manner that augments the effects of serotonin and thus, amplifies the signaling cascade in myometrial SMCs.
Little is known about other possible post-transcriptional mechanisms that may regulate MMP-13 expression and secretion in myometrial smooth muscle cells. With respect to other MMPs, MMP-9 secretion can be inhibited independently of alterations of mRNA levels in cells expressing a membraneanchored glycoprotein that possesses protease inhibitor-like domains (70). Regulation of MMP-9 secretion in carcinoma cells was determined to be dependent upon the efficiency of translation (71). It is evident by the complex nature of the signaling pathways involved that there may be manifold points of regulation by these effector molecules.
In summary, our overall model (Fig. 7) proposes that serotonin initiates a signal transduction cascade that is mediated by the 5-HT 2A receptor subtype in myometrial smooth muscle cells (14) and causes a subsequent activation of PLC through a G-protein to produce DAG. The resultant DAG in turn activates PKC, where the ␦ isoform is one likely participant, and PKC then propagates this cascade through ERK1/2 activation, which results in the production of MMP-13. These results demonstrate the multiple regulatory factors required for the stringent control of MMP-13 production in the context of ensuring a timely initiation of postpartum events resulting in the degradation of collagen, as such, may logically contribute to the prevention of inappropriate collagenolysis in the uterus, and the subsequent return of the uterus to reproductive competence.