Sphingosine kinases negatively regulate the expression of matrix metalloproteases (MMP1 and MMP3) and their inhibitor TIMP3 genes via sphingosine 1‐phosphate in extravillous trophoblasts

Abstract Purpose Extracellular matrix remodeling is essential for extravillous trophoblast (EVT) cell migration and invasion during placental development and regulated by matrix metalloproteinases (MMPs) and tissue inhibitors of metalloproteases (TIMPs). Sphingosine kinases (SPHK1 and SPHK2) synthesize sphingosine‐1‐phosphate (S1P), which works either intracellularly or extracellularly via its receptors S1PR1‐5 in an autocrine or paracrine manner. The role of SPHKs/S1P in regulating the expression of MMPs and TIMPs in EVT is mostly unknown and forms the primary objective of the study. Methods HTR‐8/SVneo cells were used as a model of EVT. To inhibit the expression of SPHKs, cells were treated with specific inhibitors, SK1‐I and SKI‐II, or gene‐specific siRNAs. The expressions of MMPs and TIMPs were estimated by qPCR. Results We demonstrated that SPHK1, MMP1‐3, and TIMP1‐3 were highly expressed in HTR‐8/SVneo cells. We found that treatment of cells with SK1‐I, SKI‐II, and knockdown of SPHK1 or SPHK2 increased the expression of MMP1, MMP3, and TIMP3. The addition of extracellular S1P inhibits the upregulation of MMPs and TIMPs in treated cells. Conclusions SPHKs negatively regulate the expression of MMP1, MMP3, and TIMP3. The level of intracellular S1P acts as a negative feedback switch for MMP1, MMP3, and TIMP3 expression in EVT cells.


| INTRODUC TI ON
Extravillous trophoblast (EVT) cell migration is crucial during placental development, spiral artery remodeling, and successful pregnancy outcome. Aberrant cell migration causes various pregnancy-related diseases, such as preeclampsia 1 and placental abnormalities. 2 Extracellular matrix (ECM) remodeling is one of the critical processes for proper cell migration. It is tightly regulated by ECM-degrading enzymes, such as matrix metalloproteases (MMPs) and tissue inhibitors of metalloproteases (TIMPs) family proteins. 3 MMPs are zinc-dependent proteases that are secretory or membrane proteins and get activated during matrix remodeling in various physiological and pathophysiological conditions. 3 There are 24 MMP genes reported in humans that encode for 23 MMP proteins. 3,4 To avoid excessive and deleterious degradation of tissues, TIMPs inhibit activated-MMPs and maintain tissue homeostasis. TIMP family consists of four members, TIMP1, TIMP2, TIMP3, and TIMP4. Except for TIMP3, all TIMPs are secretory proteins. 5 Co-expression of MMPs and TIMPs has been demonstrated in trophoblast cells. 6 MMP2 and MMP7 have a significant role in endometrial tissue remodeling during decidualization. 7,8 Various studies showed that nearly all MMPs were expressed in decidua and cytotrophoblasts. 6,[9][10][11] Sphingosine-1-phosphate (S1P), a sphingolipid, plays a vital role in various cellular processes, including cell proliferation, cell migration, secretion, and apoptosis. [12][13][14] S1P is synthesized intracellularly by the phosphorylation of sphingosine through two conserved Sphingosine kinases (SPHK1 and SPHK2). 15 S1P works either intracellularly as a secondary messenger or through its receptors S1PR1-5 in an autocrine or paracrine manner. 16 Intracellular S1P can regulate the expression of various genes by binding to many intracellular targets, such as histone deacetylase (HDAC), 17 atypical protein kinase C (aPKC), 18 PPARγ, 19 and TNF receptor-associated factor 2 (TRAF2). 20 S1PR1/S1P signaling is required for the expression of MMP2 in bone marrow-derived mesenchymal stromal cells. 21 The expression of MMP2 and MMP9 was upregulated by S1P treatment in pancreatic cancer cells. 22 S1P promotes EVT cell invasion through MMP2 using S1P/S1PR1 signaling. 23  In the present study, we investigated whether SPHK activity and intracellular S1P play any role in the expression of MMPs and TIMPs in EVT cells. Indeed, we found that the expression of MMP1, MMP3, and TIMP3 genes was regulated through the intracellular level of S1P synthesized by SPHK1 and SPHK2.

| Materials
Primers were synthesized from Eurofins, India, and IDT, India. SPHK1 inhibitor SK1-I was purchased from Enzo Life Sciences, USA. S1P and SKI-II were procured from Cayman chemicals. TAMRA-S1P was from Echelon Biosciences. All cell culture reagents were purchased from HyClone, GE Life Sciences. iScript cDNA synthesis kit was from Bio-Rad and DyNAmo ColorFlash SYBR Green qPCR kit was from Thermo Fisher. Control, SPHK1, and SPHK2 specific siRNAs were procured from Sigma-Aldrich.

| Cell culture
HTR-8/SVneo cell line, a first-trimester human EVT cell line, was a kind gift from Dr Charles H Graham. 29 These cells were grown as described previously. 12 Cells were grown in factor-reduced media (basal media + 5% charcoal-stripped FBS [dFBS]) 24 hours prior to the experiment. Cells were kept in serum-starved media (basal media + 0.5% dFBS) and treated with inhibitors for 30 minutes before activating with S1P for the specified time.

| Real-time PCR
Cells were lysed with TRIzol reagent (Invitrogen), and then total RNA was isolated as per the manufacturer's protocol. cDNA was synthesized with an iScript cDNA synthesis kit as per the manufacturer's protocol.
Primers were designed using the NCBI Primer-BLAST tool, and realtime PCR was performed on Roche LightCycler 96 machine, as described previously. 30 The list of primers used in this study are shown in Table S1. The cycling program followed for the reaction included 7 minutes of initial denaturation at 95°C and then 40 cycles with 10 seconds at 95°C and 20 seconds at 60°C. The specificity of the amplicons was analyzed by thermal dissociation curve and agarose gel electrophoresis. Data were normalized against the house-keeping gene, β-actin.

| Gene silencing
Fifty thousand cells/well were grown in 24-well plates. Transfection of SPHK1 and SPHK2 specific siRNAs was performed in duplicate wells using Lipofectamine RNAiMAX transfection reagent as per manufacturer's protocol. siRNA (10 pmol) was added to each well, and the plates were incubated for 48-72 hours with 5% CO 2 in a humidified incubator at 37°C. Scrambled siRNA was used as a negative control.

| Immunofluorescent staining and fluorescence microscopy
HTR-8/SVneo cells were grown in a 24-well plate and incubated with TAMRA-S1P (1 μM) for 1 hour. For nuclear staining, Hoechst 33258 was added to the cell culture media for 20 minutes before capturing the images. After incubation, cells were washed with PBS, and images were taken with Zeiss Axio Observer fluorescence microscope with ZEN software.

| Gelatin zymography
Expression of MMP1 and MMP3 in the HTR-8/SVneo cells was analyzed using gelatin zymography as described previously with modifications. 31 Cells were grown in a 12-well plate and were treated with S1P, SK1-I, and SKI-II as described earlier.

| Statistical analysis
Every experiment was performed three or more times independently with different passages of cells. Statistical analysis was performed using GraphPad Prism 7 software (GraphPad, La Jolla, CA, USA) using one-way ANOVA or two-tailed unpaired t test. The results are calculated as mean ± SD and data are shown as mean + SD.
P-values of <.05 were considered statistically significant.

| Expression profiling of SPHK, MMP, and TIMP genes in HTR-8/SVneo cells
To examine SPHK1 and SPHK2 genes' expression in HTR-8/SVneo cells, we measured the abundance of mRNA by real-time PCR. We found that SPHK1 is highly expressed (~45-fold) as compared to SPHK2 in these cells ( Figure 1A). These data suggest a functional role of SPHK1 in EVT cells.
Matrix remodeling proteins are crucial for cell migration and placental development. We first analyzed the basal expression of these genes in an EVT cell line model, HTR-8/SVneo cells. We could find the expression of MMP1, MMP2, and MMP3 genes in HTR-8/SVneo cells, but MMP7 and MMP9 were not in the detectable range ( Figure 1B). Further, we checked the expression of TIMPs and found that all the TIMP genes (TIMP1-3) except TIMP4 were expressed in EVT. Expression of TIMP3 was ~10-fold lower than that of TIMP1 ( Figure 1C). The differential expression of MMPs and TIMPs suggested a cell-specific role of these genes in EVT.

| Extracellular S1P gets transported in HTR-8/ SVneo cells
Previous studies showed that a high concentration of extracellular S1P significantly increases the intracellular level by 7-fold. 20 To confirm that exogenous S1P can enter the cells, we incubated the cells with fluorescently tagged S1P (TAMRA-S1P, 1 µM). In agreement with previous studies, 20,32 we could observe that TAMRA-S1P was transported into the cells ( Figure 2). The data show that S1P is transported into the cells and it might have a role independent of receptor activation.  Figure 3E). The enhanced activity was reversed upon treatment with S1P ( Figure 3E). These data suggest that SPHK1 or both the isoforms play a crucial role in regulating MMP1 and MMP3 expression.

F I G U R E 1 Expression analysis of
We asked whether the formation of intracellular S1P by SPHKs has a role in regulating MMPs gene expression. To examine the effect of S1P on the expression of MMPs, cells treated with SK1-I and SKI-II were co-incubated with 10 µM S1P, as extracellular S1P can be internalized by the cells. We found that the upregulation of MMP1, MMP3 was significantly reduced in SK1-I or SKI-II-treated cells ( Figure 3A,B). These data suggest that SPHKs-dependent intracellular S1P regulates the gene expression of MMP1 and MMP3.
To identify the specific role of SPHK1 and SPHK2 in regulating the expression of MMPs, we knocked down SPHK1 and SPHK2 genes in HTR-8/SVneo cells using specific SPHK1 and SPHK2 siRNAs. A set of three different siRNAs were used for each SPHK isoform.
Only one siRNA against the SPHK1 gene could inhibit the expression by ~65% ( Figure S1A). All three siRNAs against SPHK2 could knock down the SPHK2 gene by ~92% ( Figure S1B). In contrast to the effect of SKI-1, MMP3 (~7-fold; P < .01) but not MMP1 gene was significantly upregulated after the knockdown of SPHK1 ( Figure 3C). This observation might be due to the partial knockdown of SPHK1 by siRNA. Additionally, MMP1 (~9-fold; P < .01) and MMP3 (~3.6-fold; P < .05) were significantly upregulated after the knockdown of the SPHK2 gene ( Figure 3D). These data indicate that both SPHK1 and SPHK2 regulate the expression of MMP1 and MMP3.

| SPHKs regulate the expression of the TIMP3 in HTR-8/SVneo cells
Co-expression of MMPs and TIMPs has been demonstrated in trophoblast cells. 6 We asked whether SPHK/S1P axis also regulates the expression of TIMPs . When the cells were treated with SK1-I for 24 hours, the expression of the TIMP3 (~6.8-fold; P < .01; Figure 4A right panel) was highly upregulated while TIMP1 and TIMP2 were not affected by SPHK1 inhibition ( Figure 4A  To determine the role of intracellular S1P on the expression of TIMPs, cells treated with SK1-I and SKI-II were incubated with 10 µM S1P. We found that the upregulation of TIMP3 was significantly reduced in SK1-I or SKI-II-treated cells ( Figure 4A,B). These data suggest that SPHKs -dependent intracellular S1P regulates the gene expression of TIMP3.

| S1P alone doesn't regulate the basal expression of MMPs and TIMPs in HTR-8/SVneo cells
To examine whether extracellular S1P regulates the gene expression of MMPs and TIMPs, we treated the cells with S1P (10 µM). We found that S1P did not significantly affect the expression of MMPs and TIMPs after treating the cells with S1P for 24 hours (Figure 5A,B).
These data suggest that extracellular S1P did not control the basal expression of these genes. If the concentration of intracellular S1P is lower than the threshold, the expression of these genes is upregulated.

| D ISCUSS I ON
Trophoblast cells regulate the matrix remodeling via the secretion of various MMPs and TIMPs spatially and temporally. In this study, we evaluated the role of SPHKs and S1P in regulating MMPs/TIMPs expression in EVT cells. We could show that MMP1, MMP2, and MMP3 genes were highly expressed in HTR-8/SVneo cells. Expression of MMP7 and MMP9 was lesser as compared to the other genes in EVT cells. Our data are consistent with a previous report in which MMP2 F I G U R E 5 Extracellular S1P does not control the expression of MMPs and TIMPs expression. HTR-8/SVneo cells were treated with solvent (control), S1P (10 µM), and the expressions of MMPs and TIMPs were measured by real-time PCR. A) Bar diagrams show that extracellular S1P did not regulate the expression of MMP1, MMP2, and MMP3. B) Bar diagrams show that extracellular S1P did not regulate the expression of TIMP1, TIMP2, and TIMP3. The data are means + SD of three independent experiments with different passages of HTR-8/SVneo cells. ns, not significant [Colour figure can be viewed at wileyonlinelibrary.com] expression was observed in EVT cells and MMP9 mainly in villous cytotrophoblast. 37 Another study showed that MMP2 is highly expressed in early trophoblast (6-8 weeks) while MMP9 is highly expressed in late first-trimester trophoblasts (9-12 weeks). 38 MMP1, MMP3, and MMP7 were differentially expressed in trophoblasts throughout the pregnancy. 10 It suggests that the expression of a specific MMP depends on the trophoblasts' type and pregnancy stage.

TIMP1-3 but not TIMP4 was expressed in EVT cells.
In contrast, a study showed that all four TIMPs are expressed in cytotrophoblast cells. 9 It was shown that TIMP2 was highly expressed in trophoblasts. 39 Contrary to this, TIMP1 expression was increased from week 6 to 9 in cytotrophoblast while TIMP2 was undetectable. 11 These studies suggest that differential regulation of Together with our finding, we suggest that SPHKs differentially regulate the expression of MMPs and TIMPs family proteins in a cell type-specific manner. The primary role of SPHKs is to phosphorylate sphingosine and produce intracellular S1P. 15 These data suggest that SPHKs regulate the expression of these genes either independent or dependent on intracellular S1P. A study showed that activation of SPHK1 mediates the upregulation of TIMP1 in human fibroblast cells, and the intracellular level of S1P was increased in SPHK1 overexpressed cells. 28 Previously, it was shown that the addition of exogenous S1P enhanced its intracellular level. 20,32 In agreement with the previous data, we could show that S1P could enter the cells by adding fluorescently labeled S1P to the cell culture medium. Intracellular S1P binds with PPARγ and PG1β complex and regulates the expression of various genes in endothelial cells. 19 Intracellular S1P synthesized by SPHK2 acts as HDAC inhibitor and enhances the expression of c-fos and cyclin-dependent kinase inhibitor p21. 17 Another study showed that SPHK1 and S1P bind with TRAF2 and regulate NF-κB activation. 20 Surprisingly, in this study, we showed that the addition of exogenous S1P inhibits the upregulation of MMP1, MMP3, and TIMP3 , indicating a new role of intracellular S1P. These data suggest that SPHKs-dependent intracellular S1P levels might act as a negative feedback switch and regulate the expression of MMP1, MMP3, and TIMP3, specifically in EVTs. S1P and SPHK1 play an essential role in various cellular processes, including cell migration and invasion in different cell types. 44 S1P inhibits cell migration in C2C12 myoblasts via the S1PR2 receptor. 45 S1P promotes EVT cell invasion through MMP2 using S1P/S1PR1 signaling. 23 In this study, we could not observe the effect of S1P on EVT cell migration ( Figure S2). A study showed that S1P attenuated the migration and thus outgrowth of EVT from the first-trimester placental explant. 46 These contrasting results might be due to the cell type-specific signaling. In preeclampsia, HDAC9 was downregulated, and knockdown of HDAC9 upregulated the expression F I G U R E 6 Proposed model of the role of intracellular S1P. The cartoon shows the role of intracellular S1P in MMP1, MMP3, and TIMP3 expression and regulation [Colour figure can be viewed at wileyonlinelibrary.com] of TIMP3 in HTR-8/SVneo cells. 27 Poor EVT cell migration and invasion were observed in preeclampsia. 47 Overall, we propose that SPHKs-dependent regulation of TIMP3 might play an essential role in preeclampsia.
In conclusion, we found that the level of intracellular S1P acts as a controlling switch for MMP1, MMP3, and TIMP3 expression in EVT cells, suggesting a new role of intracellular S1P in ECM remodeling ( Figure 6). Together with previous studies, we propose that downregulation of SPHKs in pregnancy disorders, such as preeclampsia 40 decreases the intracellular S1P level leading to the activation of TIMP3 27 and MMP1, and MMP3.