Plexin-B2 and Semaphorins Do Not Drive Rhabdomyosarcoma Proliferation or Migration

Rhabdomyosarcoma (RMS) is the most common pediatric soft tissue sarcoma for which subsets of patients have longstanding unmet clinical needs. For example, children with alveolar rhabdomyosarcoma and metastases at diagnosis will experience only 8% disease-free 5-year survival for nonlocalized unresectable recurrent disease. Hence, development of novel therapeutic strategies is urgently needed to improve outcomes. The Plexin-Semaphorin pathway is largely unexplored for sarcoma research. However, emerging interest in the Plexin-Semaphorin signaling axis in pediatric sarcomas has led to phase I cooperative group dose-finding clinical trials, now completed (NCT03320330). In this study, we specifically investigated the protein expression of transmembrane receptor Plexin-B2 and its cognate SEMA4C ligands in clinical RMS tumors and cell models. By RNA interferences, we assessed the role of Plexin-B2 in cell growth and cell migration ability in selected alveolar and embryonal RMS cell model systems. Our results affirmed expression of Plexin-B2 across human samples, while also dissecting expression of the different protein subunits of Plexin-B2 along with the assessment of preferred Semaphorin ligands of Plexin-B2. Plexin-B2 knockdown had positive or negative effects on cell growth, which varied by cell model system. Migration assayed after Plexin-B2 knockdown revealed selective cell line specific migration inhibition, which was independent of Plexin-B2 expression level. Overall, these findings are suggestive of context-specific and possibly patient-specific (stochastic) role of Plexin-B2 and SEMA4 ligands in RMS.


Introduction
Rhabdomyosarcoma (RMS) is the most common soft tissue sarcoma in children and accounts for ∼5-10% of all childhood malignancies [1,2]. Histologically, two major subtypes of RMS are described, alveolar (aRMS) and embryonal (eRMS). e aRMS subtype is known to be clinically aggressive and metastatic in nature with unfavorable prognosis [3,4]. Despite technical advancement in the treatment strategies (namely, surgery, radiation therapy, and chemotherapy), the survival rate has stayed unchanged for multiple decades [5,6] with a long-term disease-free survival rate for metastatic aRMS and eRMS of 8% and 43%, respectively [2,5,7]. erefore, development of novel therapeutic strategies is urgently needed to improve cure rates.
Plexins are a family of transmembrane protein receptors that bind to Semaphorin molecules to modulate cellular functions such as cell migration, cell adhesion, and invasion [8,9]. Specifically, Plexin-B2 is involved in cell motility, vessel formation, and central nervous system development [10][11][12] by binding to Semaphorin 4C (SEMA4C). Recent finding reported that Plexin-B2-SEMA4C axis also plays a critical role in recruiting T-cells to germinal centers and regulating the antibody responses [13]. Overexpression of Plexin-B2 and SEMA4C has been implicated in the poor prognosis in the cancers of bone, breast, and brain [8,9,[14][15][16]. Furthermore, in the context of cancer progression, Plexin-B2-SEMA4C signaling cascade can enhance cell invasion and migration by triggering downstream effector pathways such Met, ErbB2, and RhoA-dependent kinase [17]; Le et al., 2015). In order to better understand the disease metastasis, we investigate the previously unexplored Plexin-Semaphorin signaling pathway in RMS and elucidate the role of this membrane protein complex in the context of disease progression.
Here, we investigated roles for Plexin-B2-SEMA4s in RMS (aRMS and eRMS) disease progression; specifically, we examined the effect of silencing Plexin-B2 on cell growth and migration. In our studies, we have affirmed the expression of Plexin-B2 and its cognate ligands SEMA4C, SEMA4D, and SEMA4F across human samples. Furthermore, we worked carefully to understand the isoforms of Plexin-B2 and the functional significance of Semaphorin ligands that bind to Plexin-B2 using small interfering RNAs of Plexin-B2 in rhabdomyosarcoma cell lines as a genetic proof of concept. Our findings show that cell growth was significantly decreased in both aRMS and eRMS cancer cell lines. Interestingly, silencing of Plexin-B2 enhanced cell growth in a primary aRMS cell culture. However, cell migration capacity was reduced in only one aRMS cell line tested, despite all cell lines expressing Plexin-B2. ese results suggest potential patient-to-patient variability for the function of Plexin-Semaphorin signal transduction in RMS that may not be predictable from Plexin-B2 as a biomarker.

Cell Migration Assay after Plexin-B2
Silencing. In order to determine the effect of Plexin-B2 silencing on cell migration, transwell migration assays were performed using Transwell ® permeable supports (Costar, Cat. #3422) with 8 μm polycarbonate membrane. RMS cells, with and without Plexin-B2 knockdown, were serum-starved, resuspended in serum-free media, and seeded in the upper chamber at a density of 5 × 10 4 . e bottom chambers consisted of serumfree media (negative control), with media consisting of 10% FBS (positive control). After 48 hr, cell migration was quantified via colorimetric crystal violet assay. Briefly, after incubation, medium from the inserts was aspirated, which was followed by the removal of non-migrating cells from inner side of the inserts using cotton swab. e migrating cells were fixed in 4% paraformaldehyde and washed in PBS. e cells were then stained with 0.1% crystal violet prepared in 10% methanol. e inserts were washed, air-dried, and subjected to 10% acetic acid to elute the crystal violet dye. e intensity of dye was proportional to the cells migrating to the outer side of inserts. e absorbance of eluted dye was measured at a wavelength of 595 nm using a BioTek plate reader.

Cell Migration Assay Using Real-Time Cell Analysis (RTCA).
e assay was set up on the real-time cell analyzer (xCELLigence RTCA DP, Agilent, Santa Clara, CA) system using CIM-plate 16 (Agilent, Cat. #5665817001). Each test condition was performed in triplicate. RMS cells were serum-starved (16 hr) prior to performing the assay. Wells of the lower chamber of the plate were filled with chemoattractant SEMA4C (R&D Cat. #6125-S4), SEMA4D (R&D 2 Sarcoma Cat. #7470-S4), SEMA4C + SEMA4D (200ng/ml in serumfree media (SFM)), 10% FBS (positive control), and SFM (negative control). e upper chamber was then attached and 50 ul of SFM added to each well. e plate was allowed to equilibrate in the incubator at 37°C and 5% CO2 for 1 hr before taking the baseline cell index (CI) readings for each well. RMS cells (5 × 10 4 cells/100 ul) were plated into each well of the upper chamber of the plate. e cells were allowed to settle at room temperature for 30 minutes before initiating the assay. Readings were taken every 15 minutes for the duration of the assay (48 hr). Numerical data expressed as mean ± standard deviation were calculated automatically by the RTCA software package. All CI values were normalized to that of the SFM (negative control).

Statistical Analyses.
e significance of variation in mean optical density (OD) with treatment was assessed with a linear model of OD in terms of treatment, cell type, and treatment by cell type interaction with a Tukey adjustment to correct for multiple comparisons. All data were analyzed in log units. Statistical testing was two-sided with an experiment-wise significance level of 5%. Results were summarized with the sample size, mean, standard error of the mean, 95% confidence intervals for differences of means, and p value. SAS Version 9.4 for Windows (SAS Institute, Cary, North Carolina) was used throughout.

Plexin-B2 and Its Ligands Are Overexpressed in Majority of RMS Tissues and Correlated with Survival.
To investigate protein expression of Plexin-B2, SEMA4C, SEMA4D, and SEMA4F, we performed IHC on formalin-fixed paraffinembedded mouse and human RMS tissue microarray (TMA). Staining revealed that in both subtypes of RMS (eRMS and aRMS), the protein expressions of Plexin-B2 and its ligands SEMA4D and SEMA4F are elevated (see representative images in Figure 1). It is worthy of note that SEMA4D is the target of pepinemab (VX15/2503), recently investigated in pediatric phase I clinical trials (NCT03320330) ( [18], [19]). IHC scores of Plexin-B2, SEMA4C, SEMA4D, and SEMA4 were calculated based on the percentage of stain present and the stain intensity of tumor cells in the tumor tissues of human patients and Sarcoma mouse models. Quantification of the stains in aRMS, aRMS, and undifferentiated pleiomorphic sarcomas (UPS) is summarized in Table 1. Overall, Plexin-B2 was present in 100% RMS cells in human tumor samples at low intensity in aRMS and at intermediate to high intensity in eRMS. In the mouse models tissues, Plexin-B2 expression was present in 75% of aRMS and 86% of eRMS tissues ranging from weak to strong staining intensity. e ligand SEMA4C was weakly     6 Sarcoma expressed in 67% of human eRMS only. e SEMA4D and SEMA4F expression was seen in 100% of the human tumor cells, albeit at low intensity. We next examined the clinical relevance of Plexin-B2 expression in RMS patients. e Kaplan-Meier curve indicated that high Plexin-B2 correlated with poorer overall survival in RMS patients in comparison to low Plexin-B2, with a statistical significance of 0.033 (Supplementary Figure S1). Overall, the data suggest that high protein and gene expression of Plexin-B2 have unfavorable survival outcomes in RMS patients.

Plexin-B2 Is Highly Expressed in RMS Cell Lines.
Next, to confirm the protein expression of Plexin-B2, we performed immunoblot analyses using available cell culture models of RMS. To fully characterize isoform expression, we probed protein lysates of eRMS (RD and Rh18) and aRMS

Sarcoma
Notably, CF-1, a PDX explant primary tumor cell culture, showed the highest Plexin-B2 expression and RD (eRMS) showed the lowest Plexin-B2 expression compared to all the RMS cell lines. ese results qualified the cell lines tested for further study of Plexin-B2 in rhabdomyosarcoma.

SEMA4C, SEMA4D, and SEMA4F Are Aberrantly
Expressed in RMS Cell Lines. To assess the expression pattern of the Plexin-B2 ligands, we probed the RMS protein lysates for SEMA4C, SEMA4D, and SEMA4F antibodies and detected membrane-bound SEMAs. In contrast to IHC quantification, SEMA4C expression was robust in all the tested RMS cell lines (eRMS and aRMS); SEMA4D expression pattern was aberrant between subtypes (Figures 3(a) and 3(b)). It is worthy of note that the eRMS cell lines, RD and Rh18, showed higher SEMA4D expression compared to aRMS cell lines (Rh3, Rh5, Rh30, and Rh41) and CF-1 showed the lowest SEMA4D expression among the tested cell lines. SEM4F showed a distinct expression pattern with low levels of SEMA4F in eRMS and high levels of SEM4F in aRMS cell lines. SEMA4D and SEMA4F displayed opposite expression trends between eRMS and aRMS cell lines (Figures 3(b) and 3(c)). ese findings indicate that the ligand molecules of Plexin-B2 show distinct autocrine expression pattern in eRMS and aRMS. Specifically, SEMA4D expression is directly proportional to Plexin-B2 expression in the eRMS cell types. In contrast, SEMA4F expression was directly proportional to Plexin-B2 expression in the aRMS.

Efficient Knockdown of Plexin-B2 in RMS Cell Lines.
In order to determine the function of Plexin-B2 in cell growth and migration in RMS, we optimized knockdown of Plexin-B2 using two sets of siRNAs targeting Plexin-B2 and a set of non-targeting (scrambled) siRNA siControls. e siRNA-mediated knockdown of Plexin-B2 was conducted in two eRMS cell lines (Rh18 and RD) and three aRMS cell lines or cell cultures (CF-1, Rh30, and Rh41). Immunoblot analyses confirmed the efficient knockdown of Plexin-B2 in all the tested cell lines with both sets of siRNAs (siplexinB2_01 and siplexinB2_03) as depicted by representative immunoblots in Figure 4, Supplementary Figure S2, and Supplementary Figure S3. Notably, we found that SEMA4C expression increased after Plexin-B2 knockdown. In contrast, both SEMA4D and SEMA4F expressions remained unchanged upon Plexin-B2 knockdown. For the experiments that followed, we selected si_03 siRNAs as an efficient tool for Plexin-B2 knockdown.

Knockdown of Plexin-B2 Reduces Cell Proliferation in Cell
Lines Rh30, Rh41, RD, and Rh18 but Not in CF-1. To evaluate the effect of Plexin-B2 knockdown in cell proliferation, we selected three aRMS cell cultures (CF-1, Rh30, and Rh41) and two eRMS cell lines (Rh18 and RD) and measured cell viability at 24, 48, and 72 hr. Our results showed significant decrease in cell proliferation at 48 hr and 72 hr time points in aRMS cell lines, Rh30 and Rh41 ( Figure 5). Cell growth after Plexin-B2 silencing was significantly decreased in RD and Rh18 cell lines at all time points. Plexin-B2 knockdown significantly (paradoxically) increased cell growth in CF-1. Overall, these results suggest a multifactorial and stochastic effect of Plexin-B2 knockdown on cell growth in RMS. 10 Sarcoma

Neither
To determine whether Plexin-B2 regulates cell migration of RMS cell lines in ligand-independent or serum-responsive manner, we silenced Plexin-B2 using siRNAs in Rh30, Rh41, RD, and CF-1 (as shown in Figure 4 and Supplementary Figure S2) and then evaluated cell migration using transwell assays. e migrating cells were dyed with crystal violet, and the eluted dye intensity was quantified at 595 nm. e knockdown of Plexin-B2 caused a significant reduction in cell migration for only Rh30 (but not Rh41, CF-1, or RD) in the context of 10% fetal bovine serum bait (Figure 6(a)-6(d)) and Supplementary Figure S4). Next, we investigated whether the cognate ligands for Plexin-B2 stimulated migration of the Plexin-B2 expressing cell lines. Using a realtime conduction-based migration monitoring instrument (Figure 7), no change in migration was observed when ectopic SEMA4C or SEMA4D was used as a bait. Overall, these results revealed a multifactorial and stochastic effect of Plexin-B2 knockdown on cell growth in RMS, seemingly independent of SEMA4C or SEMA4D expression.

Discussion
In this study, we explored the role of Plexin-B2-SEMA4 signaling axis in the context of cell growth, cell migration, and invasion in RMS. We show that Plexin-B2 is highly expressed in aRMS and eRMS subtype at protein levels. From retrospective analysis of clinical IRSG-IV samples, high Plexin-B2 expression in new diagnosis samples correlated with poor patient survival in RMS. Furthermore, SEMA4C was robustly expressed in all the tested RMS cell lines, and SEMA4D was prominently expressed in eRMS cell lines. In contrast, SEMA4F was highly expressed in aRMS as compared to eRMS cell lines as revealed by immunoblot analyses. Knockdown of Plexin-B2 resulted in impaired cell growth in all immortalized RMS cell lines. However, the silencing of Plexin-B2 reduced serum-stimulated cell migration capacity in only one aRMS cell line and had no effect on any eRMS cell line; additionally, the Plexin-B2 cognate ligands SEMA4C and SEMA4D had no effect on migration in any of the Plexin-B2 expressing cell lines/cell cultures tested. A potential caveat of our cell viability assays is that Cell-Titer-Glo is a measure of cellular metabolism and therefore an indirect measure of cell viability and that viability and proliferation parameters generally but not always coincide.
Recent findings have reported that Plexin-B-SEMA4 signaling axis is overexpressed and involved in disease progression via cell proliferation and cell migration in bone, breast, and brain cancers [8,9,[15][16][17]20]. Related studies have revealed that SEMA4-Plexin-B signaling axis alters cell morphology and enhances cell migration and cell invasion by activating downstream signaling pathways such as Rho GTPases, RTK-Met, and AKT [8,15,17]. Our results indicate that Plexin-B2 can regulate cell migration in a selected cellular model of alveolar rhabdomyosarcoma. However, future experiments will be required for the assessment of noncanonical ligand dependence and downstream signaling pathways in RMS.
In summary, we have shown that Plexin-B2 can be a prognostic factor for RMS, implying that the Plexin-B-SEMA4 signaling axis might be a potential therapeutic target for rhabdomyosarcoma; however, paradoxical effects of Plexin-B2 knockdown to increase tumor cell growth and only selectively in certain cell lines to decrease tumor cell migration raise significant concerns about therapeutic approaches to Plexin-B-SEMA4 inhibition until the factors underlying this complex biology and clinical biomarkers of response can be developed.

Data Availability
ere are no extended datasets, and all data are presented as figures and tables in the manuscript.

Conflicts of Interest
e authors declare that they have no conflicts of interest.