Changes in Vasodilator-Stimulated Phosphoprotein Phosphorylation,Profilin-1,and Cofilin-1 in Accreta and Protection by DHA

Introduction

Human placentation involves establishment of an efficient attaching system and blood flow between mother and its fetus. The anchoring villi, which are tunnel-like structures, hold the placenta in close contact to the myometrium. During this tightly regulated process, trophoblasts migrate from the basement membrane, adhere to and penetrate the decidua, and ultimately invade the inner third of the myometrium and the arterioles supplying this region. ^{1 -3} Placental vasodilator-stimulated phosphoprotein (VASP) is expressed throughout pregnancy, with notably high expression in invasive extravillous trophoblasts (EVTs) of anchoring villi and distal columns. ⁴ Aberrant EVT proliferation, migration, and invasion into uterine tissue have been observed in placenta disorders characterized by both insufficient (preeclampsia) and excessive (accreta) invasion. ^{5 -11} In cancer cell proliferation and migration, which closely resemble placenta trophoblast hyperproliferative diseases, differential phosphorylation of VASP at Ser157 (pVASP-S157) and Ser239 (pVASP-S239), profilin-1, and cofilin-1 mediate cellular actin cytoskeletal remodeling of proliferating and migrating cancer cells. ^{12 -19} These proteins and their possible interactions during actin remodeling further regulate the formation of protrusive structures such as lamellipodia and filopodia (horizontally) and invadopodia (vertically down the nuclear axis), ^{16,17,19 -23} which facilitate cell migration and invasion. ^{20,24 -27} To date, these mechanisms have not been explored with respect to EVT hyperproliferation and migration associated with placenta accreta.

Actin remodeling (polymerization, stabilization/depolymerization) is regulated by VASP, profilin-1, and cofilin-1 in a feedback mechanistic manner. Vasodilator-stimulated phosphoprotein is an actin-binding protein (ABP) that contains 2 Ena/VASP homology (EVH) domains separated by a central proline-rich region (PRR). Cyclic adenosine monophosphate–dependent protein kinase preferentially phosphorylates Ser157 within the PRR, increasing actin binding activity and promoting formation of cytoskeleton-shaped membrane protrusions. ²⁸ In contrast, protein kinase-G preferentially phosphorylates Ser239 within the C-terminal EVH2 domain, which promotes F-actin dissociation and inhibits cytoskeleton-shaped membrane protrusions. ^16,19,29,30 Profilin-1 is a ubiquitous ABP that regulates actin polymerization in response to extracellular signals. ^18,24,31 Cofilin-1 (nonmuscle; n-cofilin) is an actin-modulating member of the actin depolymerizing factor/cofilin family widely distributed in intracellular regions that binds to and depolymerizes filamentous F-actin and inhibits polymerization of monomeric G-actin. ^26,32,33 Cofilin-1 has roles in actin dynamics depending upon concentration, phosphorylation status, G-actin, other protein factors, and cellular functional requirements. ^{34 -39} Moreover, cofilin-1 levels vary among cancer cell and tumor types and degree of proliferation and migration. ^{26,40 -42}

Docosahexaenoic acid (DHA), an anti-inflammatory omega-3 fatty acid often taken as a dietary supplement, has been suggested as a pharmacologic agent in placental disorders, which are often associated with inflammation. ^{43 -47} This suggestion has pathophysiologic plausibility because DHA inhibits cancer cell proliferation, migration, invasion, and survival. ^12,48,49 We recently showed that the antiproliferative, antimigratory, and antiviability effects of DHA in cancer cells were mediated through altered profilin-1 and cofilin-1 expression and differential phosphorylation of VASP. ¹² Here, we tested whether DHA modulates trophoblast hyperproliferation and migration through actin homeostasis. Differential VASP phosphorylation and ABP expression in human placenta biopsy tissue were associated with placenta accreta. In vitro experiments explored these relationships with respect to trophoblast proliferation and migration in the context of DHA treatment.

Materials and Methods

Results

Placenta Accreta Tissue Had Increased F-Actin and Altered ABP Expression

Figure 1A shows representative images of normal and pathological placenta biopsy tissues stained for F-actin and ABPs. Average protein expression intensity is depicted adjacent to images of placental sections stained with the relevant antibody (Figure 1B-H). Placenta accreta had significantly increased F-actin compared with normal placentation (Figure 1B, P = .018). Compared with healthy control placenta, accreta had significantly increased levels of VASP (Figure 1C, P = .030), cofilin-1 (Figure 1D, P = .019), and pVASP-S239 (Figure 1E, P = .022). In contrast, focal accreta only had significantly increased levels of cofilin-1 compared to normal placentation (Figure 1D, P < .001). No differences in profilin-1 (Figure 1F) or pVASP-S157 (Figure 1G) levels were observed in accreta or focal accreta compared to healthy placenta.

OPEN IN VIEWER

Figure 1.

Placenta accreta expresses high levels of F-actin, cofilin-1, VASP, and pVASP-S239. F-actin content, profilin-1 and cofilin-1 expression, and VASP phosphorylation were quantified in term human placenta biopsy samples from normal healthy controls, focal accreta, and accreta using florescent microscopy (A, 400× magnification) and NIH ImageJ analysis (B-G). A, F-actin (red) stained with phalloidin; profilin-1, cofilin-1, VASP, pVASP-S239, and pVASP-157 (green) stained with specific antibodies; nuclei (blue) counterstained with DAPI. Data were analyzed by t test followed by 1-way ANOVA and further analyzed by Tukey post hoc. (*) P < .05, and (***) P < .001, in comparison to control; n = 4 tissue sections from different donors. Scale bar is 200 µm. ANOVA indicates analysis of variance; VASP, vasodilator-stimulated phosphoprotein.

Docosahexaenoic Acid Reduced Proliferation and Migration of BeWo Cells

Using BeWo cells as a model trophoblast system, DHA treatment for 24 hours suppressed BeWo cell proliferation to a level that reached significance by 6 days posttreatment (Figure 2A, P < .001). Using a wound assay, wound closure was nearly complete in the control by 48 hours (Figure 2B). However, there was significantly reduced trophoblast migration at 24 hours (P < .05) and 48 hours (P < .001) following treatment with DHA (Figure 2C). There was a significant interaction between treatment and time for both proliferation (P < .001) and migration (P = .038).

OPEN IN VIEWER

Figure 2.

Docosahexaenoic acid supplementation inhibits proliferation and migration of trophoblast-derived choriocarcinoma cells, BeWo. A, BeWo cell proliferation based on cell counts taken at 2, 4, and 6 days following 24 hours treatment with 30 µM DHA or PBS (Control). B, Representative wound assays of BeWo cells treated for 24 hours with 30 µM DHA or PBS prior to wound induction using a 20 μL pipette tip. C, Cell migration (percent wound closure) was quantified from digital images of the same regions taken at 0, 24, and 48 hours posttreatment. Images were captured with an Olympus BX51 compound microscope at 10× magnification and analyzed with NIH ImageJ software. Data were analyzed by 2-way repeated measures ANOVA and by Tukey post hoc. (*) P < .05, and (***) P < .001, n = 3 experiments. Scale bar is 200 µm. ANOVA indicates analysis of variance; DHA, docosahexaenoic acid; PBS, phosphate-buffered saline.

Docosahexaenoic Acid Alters the Expression Profile of ABPs in BeWo Cells

To determine whether the DHA effects on trophoblast proliferation and migration were associated with changes in ABP expression patterns, total protein isolated from BeWo cells treated with DHA or PBS for 24 hours was analyzed by Western blot. At the tested concentration, DHA had no significant effects on VASP (Figure 3A), profilin-1 (Figure 3B), or cofilin-1 (Figure 3C) expression levels. Docosahexaenoic acid differentially affected VASP phosphorylation, with significantly reduced phosphorylation at Ser239 (Figure 3D, P = .022), but no effect on phosphorylation at Ser157 (Figure 3E).

OPEN IN VIEWER

Figure 3.

Docosahexaenoic acid modulates expression of actin-binding proteins in trophoblast-derived choriocarcinoma cells. BeWo cells pretreated for 24 hours with 30 µM DHA or PBS (vehicle control) were harvested and total protein subjected to Western blot analysis of (A) VASP, (B) profilin-1, (C) cofilin-1, (D) pVASP-S239, and (E) pVASP-S157. Data were analyzed by t test followed by 1-way ANOVA and by Tukey post hoc. (*) P < .05, (**) P < .005, and (***) P < .001, n = 3 experiments. ANOVA indicates analysis of variance; DHA, docosahexaenoic acid; PBS, phosphate-buffered saline; VASP, vasodilator-stimulated phosphoprotein.

Docosahexaenoic Acid Induces Apoptosis in BeWo Cells

Docosahexaenoic acid significantly induced expression of the proapoptotic caspase-9 protein in BeWo cells 24 hours after treatment (Figure 4A, P = .005), concurrently suppressed expression of the antiapoptotic BCL-2 protein (Figure 4B, P = .009), and had no effect on proapoptotic Bax protein expression (Figure 4C). Collectively these results indicate that DHA can induce trophoblast apoptosis.

OPEN IN VIEWER

Figure 4.

Docosahexaenoic acid induces apoptosis in trophoblast-derived choriocarcinoma cells. BeWo cells were treated with DHA for 24 hours and harvested for Western blot analysis of apoptosis proteins (A) caspase-9, (B) BCL-2, and (C) Bax. Data were analyzed by t test followed by 1-way ANOVA and further analyzed by Tukey post hoc. (*) P < .05, (**) P < .005, n = 3 experiments. ANOVA indicates analysis of variance; DHA, docosahexaenoic acid.

Discussion

Extravillous trophoblast migration and invasion are highly regulated and critical for successful placentation. Unsurprisingly, there is an incomplete understanding of the numerous factors that contribute to its deregulation, resulting in pathological conditions ^5,10,52 of hyper- (placenta accreta or choriocarcinoma) and hypoinvasion (preeclampsia and intrauterine growth restriction). ^5,6,10,43 Interpretation of the existing data suggests that reestablishing homeostasis earlier in pregnancy during the hyperproliferative trophoblast stage might offer a novel approach to target overinvasive disorders (eg, accreta, percreta, and increta). ¹¹ However, before beginning to test clinically DHA’s potential therapeutic role, we should first identify key proteins and molecules participating in regulation of EVT proliferation, migration, and invasion.

Actin dynamics regulated through ABPs mediate cancer cell proliferation, migration, and invasion during metastasis ^17,20,27 and may be similarly associated with these processes in trophoblasts. Previous data demonstrating temporal changes in placental trophoblast VASP expression and phosphorylation during the course of normal pregnancy suggest an intrinsic role for this phosphoprotein regulating trophoblast function. ⁴ However, phosphorylation specificity was not yet determined. As in cancer cells, dysregulated trophoblast VASP expression or activation (phosphorylation) might be associated with aberrant proliferation and migration, facilitating placental hyperinvasion during accreta. Our assertion is consistent with prior studies demonstrating differential VASP phosphorylation and altered profilin-1- and cofilin-1-mediated actin dynamics regulate proliferation and migration. ^{13,15,39,40,53} Our results demonstrating increased VASP, pVASP-S239, and cofilin-1 in the context of increased F-actin are suggestive of altered actin dynamics as a mechanism regulating trophoblast hyperproliferation and migration through promotion of homeostasis.

Fundamental to understanding accreta pathology is determining whether dysregulated ABPs are causal or a response to trophoblast hyperproliferation and migration. Cofilin-1 participates in actin polymerization by providing adenosine triphosphate-actin monomers to profilin-1 for production of F-actin ^34,37 ; elevated levels suggest potential cofilin-1 involvement in increased F-actin content in accreta and in trophoblast hyperproliferation and migration. Cofilin-1 is associated with certain cancers ^{26,54 -56} and its overexpression and silencing, respectively, exacerbate or abrogate cancer cell hyperproliferation, migration, and invasion, although absence of parallel investigation of actin remodeling and profilin-1 renders comparisons to accreta and trophoblasts incomplete. ^57,58 Previous studies showed increased F-actin, high proliferation, and migration require active participation of profilin-1 and pVASP-S157 in actin dynamics, ^{17,19,24,40,53} although neither of these proteins were changed in our study, despite high levels of F-actin at the late stage of disease. One interpretation of our results is that unchanged levels of profilin-1 and pVASP-S157 in accreta biopsy tissue represent a return of proactin polymerization machinery to basal levels at this stage of disease, with elevated cofilin-1 favoring actin depolymerization mechanics stabilizing actin. ^39,59 However, we cannot exclude the possibility that lack of change in profilin-1 and pVASP-S157 in accreta and unresponsiveness to DHA in vitro might reflect differences in regulation of overinvasiveness between trophoblasts and cancer cells.

Treatment of 3 prostate cancer cell lines with 23,24-dihydrocucurbitacin F suggests antiproliferative effects can occur through actin aggregation and cofilin-actin rod formation, leading to apoptosis. ⁶⁰ Significantly increased pVASP-S239 levels in cancer cells and tissue counter high actin polymerization, hyperproliferation, invasion, and migration. ^13,15,40,61 As shown by our data, increased levels of VASP, pVASP-S239, and cofilin-1 in the context of increased F-actin content suggest a regulatory role of these proteins to counteract trophoblast hypermigration and proliferation.

Docosahexaenoic acid reduces cancer cell aggressiveness, which is associated with changes in actin dynamics and apoptotic profiles. ^40,48 Similar effects on the proliferative and migratory activity of trophoblasts (BeWo) seen through our experiments suggest involvement of common target(s) and machinery. Contrasting with cancer cells, ^40,61 reduced proliferation and migration in BeWo cells treated with DHA were associated with a parallel decrease in VASP S239 phosphorylation and coflin-1 level that trended lower. Although much waits discovery, as in cancer studies, DHA has shown both its potential to temper trophoblast invasive phenotype and its utility as a tool to manipulate ABP expression patterns. Opposite effects on BeWo ABP expression than observed in cancer cells might be related to DHA levels, although we cannot exclude cell-specific differences in responsiveness to DHA.

Evidence from cancer studies associates cofilin-1 with increased proliferation and reduced apoptosis. ^{56 -58} Significantly induced expression of proapoptotic caspase-9 and suppression of antiapoptotic BCL-2, with cofilin-1 trending lower, suggest DHA-modulating effects similar to cofilin-1 knockdown in bladder cancer cells ⁵⁶ and suggests that reduced BeWo cell growth was at least partially due to increased apoptosis. Because accreta tissue are characterized by increased trophoblast proliferation regardless of severity, ¹¹ inducing apoptosis in overproliferative trophoblasts may have therapeutic benefit.

Conclusion

Our study provides evidence that ABPs are active participants in the molecular events regulating the invasive phenotype of the EVT. We propose that increased VASP Ser239 phosphorylation and cofilin-1 might counteract elevated F-actin and high trophoblast proliferation in accreta. In vitro data support proposed roles for DHA-mediated pVASP-S239/cofilin-1 expression and apoptosis, which affected BeWo cell migration and proliferation, respectively; although more in-depth investigation is needed to explore a connection between therapeutic potential of DHA and ABPs in overinvasive disorders.