Semaphorin 3C promotes de novo steroidogenesis in prostate cancer cells

Intratumoral androgen biosynthesis contributes to castration-resistant prostate cancer progression in patients treated with androgen deprivation therapy. The molecular mechanisms by which castration-resistant prostate cancer acquires the capacity for androgen biosynthesis to bypass androgen deprivation therapy are not entirely known. Here, we show that semaphorin 3C, a secreted signaling protein that is highly expressed in castration-resistant prostate cancer, can promote steroidogenesis by altering the expression profile of key steroidogenic enzymes. Semaphorin 3C not only upregulates enzymes required for androgen synthesis from dehydroepiandrosterone or de novo from cholesterol but also simultaneously downregulates enzymes involved in the androgen inactivation pathway. These changes in gene expression correlate with increased production of androgens induced by semaphorin 3C in prostate cancer model cells. Moreover, semaphorin 3C upregulates androgen synthesis in LNCaP cell-derived xenograft tumors, likely contributing to the enhanced in vivo tumor growth rate post castration. Furthermore, semaphorin 3C activates sterol regulatory element-binding protein, a transcription factor that upregulates enzymes involved in the synthesis of cholesterol, a sole precursor for de novo steroidogenesis. The ability of semaphorin 3C to promote intratumoral androgen synthesis may be a key mechanism contributing to the reactivation of the androgen receptor pathway in castration-resistant prostate cancer, conferring continued growth under androgen deprivation therapy. These findings identify semaphorin 3C as a potential therapeutic target for suppressing intratumoral steroidogenesis.


Introduction
Despite its initial efficacy, androgen deprivation therapy (ADT) remains a palliative treatment for metastatic prostate cancer, and patients eventually progress to castration-resistant prostate cancer (CRPC) (So et al. 2005, Tilki et al. 2016).CRPC cells are capable of intratumoral androgen synthesis despite the low serum testosterone during ADT (Mostaghel et al. 2007, Dillard et al. 2008, Montgomery et al. 2008), providing a sufficient level of androgens for the reactivation of the androgen receptor (AR) (Goldenberg et al. 1988, Bruchovsky et al. 1989).The molecular pathways that underlie the activation of intratumoral androgen synthesis in PCa cells upon systemic androgen deprivation remain poorly understood.This study is a step toward unraveling the signaling programs that affect steroidogenesis in PCa cells.
Several lines of evidence suggest that activation of various receptor tyrosine kinase (RTK) pathways can stimulate steroidogenesis.For instance, EGF and insulinlike growth factor (IGF) 1 stimulate steroidogenesis in Leydig cells (Sordoillet et al. 1991, Evaul & Hammes 2008).Furthermore, insulin and IGF2 have been shown to increase steroidogenesis in PCa cells (Lubik et al. 2011(Lubik et al. , 2013)).Because of its therapeutic significance, the identification of other drivers of intratumoral steroidogenesis remains an important objective in the study of CRPC.
Semaphorin 3C (SEMA3C) is a signaling protein that acts as an autocrine growth factor that activates various RTKs, such as EGFR, HER2, and MET, in a cognate ligand-independent manner.Also, evidence shows that SEMA3C can promote PCa growth in androgen-deprived conditions (Peacock et al. 2018).In addition, SEMA3C has been shown to play a role in perineural invasion of prostate cancer cells (Maru et al. 2001, DeLancey et al. 2013, Ciftci et al. 2015, Kraus et al. 2019, Yin et al. 2021).However, the functional role of SEMA3C in promoting castration-resistant growth of PCa cells remains unclear.
Herein, we discovered that SEMA3C stimulates androgen biogenesis from dehydroepiandrosterone (DHEA) or de novo from cholesterol in multiple PCa cell lines and our xenograft LNCaP model.This phenomenon is accompanied by the upregulation of multiple steroidogenic enzymes while suppressing the expression of androgen-catabolizing enzymes by SEMA3C.Furthermore, activation of sterol regulatory element-binding protein (SREBP) transcription factors and induction of cholestrogenic enzymes by SEMA3C are consistent with increased intratumoral cholesterol levels by SEMA3C in xenograft LNCaP tumors post castration.Notably, erlotinib, the U.S. Food and Drug Administration-approved epidermal growth factor receptor (EGFR) inhibitor, attenuates the induction of steroidogenic enzymes by SEMA3C, suggesting EGFR mediation in stimulating androgen biosynthesis by SEMA3C.We propose to investigate SEMA3C as a potential therapeutic target that may play a role in promoting CRPC growth through intratumoral androgen biosynthesis, de novo from cholesterol, by triggering RTK signaling pathways.

Cell culture and treatments
LNCaP and 22Rv1 cell lines, purchased from American Type Culture Collection, were authenticated by IDEXX BioAnalytics case #20201/21100-2014 (Columbia, MO, USA).C4-2 cells, kindly provided by Dr Leland W K Chung (MD Anderson Cancer Center, TX, USA), were authenticated by whole-genome and whole-transcriptome sequencing on Illumina Genome Analyzer IIx platform in 2009.LAPC4 cells, a kind gift from Dr Buttyan (Vancouver Prostate Centre, Vancouver, Canada) authenticated by IDEXX BioAnalytics case #27880-2018.All cell lines were maintained in RPMI 1640 media supplemented with 10% (v/v) fetal bovine serum with no antibiotics and were tested for mycoplasma contamination weekly, using a mycoplasma detection kit (InvivoGen, San Diego, CA, USA; Cat.No. rep-pt2).

Endocrine-Related Cancer e230010
Cat. No. R0192).qPCR reactions in three technical replicates were carried out on AB ViiA7 real-time PCR machine using Platinum SYBR Green (Invitrogen, Cat. No. 11744-500), then results were analyzed by ΔΔCt method using GAPDH as an endogenous control.qPCR primer sequences can be found in Supplementary Table 1 (see section on supplementary materials given at the end of this article).

Luciferase assay
Transient transfection of LNCaP cells with 1.2 µg ARR2PB-Luciferase reporter plasmid (a kind gift from Dr Rennie lab at Vancouver Prostate Centre, University of British Columbia) and SREBP-responsive element, SRE synthRE, (Active Motif, Carlsbad, CA, USA; Cat.No. 32192) was conducted using MirusIT-2020 transfection reagent (Mirus, Madison, WI, USA; Cat.No. MIR 5400).Cells were then treated with DHEA 3.4uM, and recombinant SEMA3C 1uM or vehicle for an additional 48 h prior to harvesting for luciferase assay using the Dual-Luciferase Reporter Assay System (Promega, Cat.No. E1960).Luminescence from all samples in technical duplicates was read on TECAN Infinite M200 PRO and was normalized to Renilla luciferase.

Analysis of clinical datasets
UCSC Xena visualization software (https://xena.ucsc.edu/) was used to examine coexpression of SEMA3C and steroidogenic enzymes in the TCGA PRAD prostate cancer dataset.Fold-changes are shown in color, log2 (normalized_count + 1) mean is subtracted per column across 550 samples.cBioPortal plots of the TCGA provisional dataset (https://www.cbioportal.org/)show the degree of correlation in expression of SEMA3C and each of enzymes CYP11A1 and SRD5A2, and AR target FKBP5.Pearson's and Spearman's correlations are indicated.

In vivo study
All animal procedures were conducted according to the standards and regulations of the Canadian Council on Animal Care and approved protocol (A15-0150) by the University of British Columbia Committee on Animal Care.To create LNCaP CRPC xenograft model (Peacock et al. 2018), 1 × 10 6 LNCaP SEMA3C and LNCaP vector cells were subcutaneously inoculated with 0.1 mL Matrigel (Becton Dickinson Labware) in the flank region of 6-to 8-week-old male athymic nude mice (Jackson Laboratory) via a 27-gauge needle under methoxyflurane anesthesia.Mice bearing LNCaP tumors were weekly monitored for body weight, tumor volume, and serum PSA levels.PSA concentration in sera samples obtained from tail vein Endocrine-Related Cancer e230010 incisions was determined by the Roche Diagnostics cobas e411 immunoassay system which is an automated, random access multichannel analyzer for immunological analysis using enhanced chemiluminescent technology.When tumor volumes or serum PSA levels reached over 200 mm 3 or a minimum of 50 ng/mL, castration was performed via the scrotum under isoflurane anesthesia.The experimental endpoint was 4 weeks post castration when subcutaneous tumors were harvested to either snap-freeze in liquid nitrogen to maintain in −80℃ for RNA isolation and steroid analysis (n = 6 per group) or to fix in formalin to prepare tissue microarray.SEMA3C protein expression in LNCaP SEMA3C and LNCaP vector tumors were compared by immunohistochemistry (IHC) staining of formalin-fixed tissues with a discovery DAB Map detection kit.

Lipid and steroid analysis
Steroids and lipid content were isolated from conditioned cell culture media, homogenates of collected cells or xenograft tumor tissues using solid phase extraction (SPE) columns.Isolated steroids or lipids were analyzed with liquid chromatography coupled with mass spectrometry (LC-MS).Details of these procedures are found in the Supplementary Materials.

Statistical analysis
All results were presented as mean ± s.d.from at least three independent experiments.Data from biological replicates were analyzed using the two-tail Student's t-test and differences with P < 0.05 reached statistical significance.

SEMA3C altered the expression of steroidogenic enzymes in PCa cells
We evaluated the effects of SEMA3C to impact the expression profile of the steroidogenic enzymes in castration-sensitive LNCaP and LAPC4 cells and castration-resistant 22Rv1 and C4-2 cells.Comparison of the expression of a battery of steroidogenic genes between LNCaP SEMA3C and LNCaP vector cells showed that a statistically significant induction in mRNA expression of HSD3B1, AKR1C3, and SRD5A2 in the SEMA3C overexpressing cells.Increases of CYP17A1, HSD3B2, HSD17B3, and SRD5A1 did not reach statistical significance (Fig. 1A).SEMA3C overexpression was validated in LNCaP SEMA3C (Appendix Fig. S1a).Also, we treated LNCaP cells with increasing concentrations of recombinant SEMA3C (0.5 and 1 μM) for 24 h.We found that mRNA expression of steroidogenic enzymes HSD3B1, HSD17B3 AKR1C3, and SRD5A2 increased in a dose-dependent manner (Appendix Fig. S1b).Similar to LNCaP cells, LAPC4 cells responded to recombinant SEMA3C treatment (1 μM) by significant induction of HSD3B1, HSD17B3, AKR1C3, and SRD5A2 along with CYP11A1 and HSD3B2 (Fig. 1B).SEMA3C expression is correlated with expression of steroidogenic enzymes CYP11A1 and SRD5A2 as well as the AR target gene, FKBP5, in prostate adenocarcinoma patients according to TCGA PanCancer dataset available on cBioPortal.Also, UCSC Xena confirmed coexpression of SEMA3C with FKBP5, CYP11A1, and SRD5A2 in prostate cancer patients (https:// xena.ucsc.edu)(Appendix Fig. S1c).
Silencing SEMA3C resulted in reduced mRNA expression of HSD3B1, HSD3B2, AKR1C3, SRD5A1, and SRD5A2 along with STAR in C4-2 cells.Conversely, AKR1C1 and AKR1C2 mRNA expression was increased in siSEMA3C-treated C4-2 cells compared to those treated with siScrambled (siScr) (Fig. 1D).However, SEMA3C silencing in 22Rv1 cells only resulted in downregulation of HSD17B3, AKR1C3, and SRD5A1 enzymes and also showed suppression of AKR1C1 and AKR1C2 (Fig. 1E).SEMA3C knockdown was validated in 22Rv1 and C4-2 cells (Appendix Fig. S1e).PLXNB1 is a receptor for SEMA3C, so we investigated whether silencing of this SEMA3C receptor can attenuate the effect of SEMA3C on the expression of steroidogenic enzymes.siRNAmediated silencing of PLXNB1 in 22Rv1 cells resulted in the reduced mRNA expression of both isoforms of HSD3B enzyme (Fig. 1F).To study the potential role of EGFR pathway in the stimulation of steroidogenesis pathway by SEMA3C, we inhibited EGFR activity with erlotinib prior to recombinant SEMA3C treatment of LNCaP cells.EGFR inhibition attenuated the induction https://doi.org/10.1530/ERC-23-0010 Endocrine-Related Cancer e230010 of STAR, HSD3B1, and AKR1C3 by SEMA3C in LNCaP cells (Fig. 1G).EGFR inhibition by erlotinib was confirmed by measuring phosphorylated and total EGFR and MAPK by immunoblotting (Appendix Fig. S1f).

SEMA3C increased androgen synthesis from adrenal precursor, DHEA
Since ADT does not reduce DHEA serum level, it is the most abundant steroid hormone in the serum of PCa patients post ADT.Cancer cells may use this adrenal precursor as a substrate for intratumoral androgen synthesis (Nishiyama et al. 2004, Labrie 2010, Narimoto et al. 2010, Arai et al. 2011).To assess the effect of SEMA3C on the androgen biosynthesis in various PCa cell lines, we incubated cells with DHEA in CSSsupplemented media for 48 h.SEMA3C overexpression resulted in increased testosterone, dihydrotestosterone (DHT), androsterone released by LNCaP cells (Fig. 2A).Also, the intracellular concentration of the same androgens within LNCaP SEMA3C cells was significantly higher compared to LNCaP vector cells (Appendix Fig. S2a).Similarly, SEMA3C overexpression in 22Rv1 cells increased secreted (Fig. 2B) and intracellular (Appendix Fig. S2b) levels of testosterone, DHT and androsterone.Also, LC-MS analysis of steroids isolated from the conditioned media derived from LAPC4 cells treated with recombinant SEMA3C showed comparable results with SEMA3C-overexpressing LNCaP and 22Rv1 cells.However, the increase of androstenedione reached statistical significance, and SEMA3C caused an increase in the 5a-dione level instead of androsterone (Fig. 2C).Silencing SEMA3C suppressed the production of DHT in C4-2 cells (Fig. 2D) and testosterone in 22Rv1 cells (Fig. 2E).In order to rule out a contribution from proliferation, we used PrestoBlue assay to evaluate the proliferative effects of SEMA3C overexpression in LNCaP and 22Rv1 cells and SEMA3C knockdown in C4-2 and 22Rv1 cells, as well as recombinant SEMA3C treatment on LAPC4 cells.We found that at early time points up to 48 h, SEMA3C did not significantly alter the proliferation of PCa cells (Appendix Fig. S2c).Consistent Endocrine-Related Cancer e230010 with these findings, Peacock et al. results showed that the difference between the proliferation rate of SEMA3C-overexpressing LNCaP and empty LNCaP cells did not reach statistical significance sooner than day 5 (Peacock et al. 2018).These data suggest that the impact of SEMA3C on steroid production during the first 48 h is likely not due to SEMA3C's effect on the proliferation of PCa cells.

SEMA3C increased AR-dependent gene expression
We found that treatment with recombinant SEMA3C for 48 h increases mRNA expression of AR target genes, FKBP5, TMPRSS2, and PSA in LNCaP (Fig. 3A) and LAPC4 cells (Fig. 3B).Luciferase assay showed that recombinant SEMA3C treatment of LNCaP cells, transfected with a luciferase reporter plasmid with probasin promoter (ARR2PB), increased AR transcriptional activity (Fig. 3C).RNA-seq analysis in LNCaP SEMA3C vs empty LNCaP found an increase in a number of AR-regulated genes.The differential expression of AR targets is presented as the heatmap (Fig. 3D).

SEMA3C induced de novo steroid synthesis from cholesterol precursor
We assessed the ability of SEMA3C to increase de novo steroidogenesis by treating LNCaP SEMA3C and LNCaP vector Endocrine-Related Cancer e230010 cells with 14 C-acetate in CSS-supplemented media for 48 h.Released steroids in the conditioned media were isolated with SPE columns and radiolabeled steroids were identified and quantified by HPLC coupled with a radiometric Scintillation Analyzer.We used radioisotopelabeled controls including cholesterol, pregnenolone, and DHT to identify the MS peaks of 14 C-labaeled steroids isolated from LNCaP conditioned media.We found that SEMA3C overexpression is correlated with a statistically significant increase in 14 C-pregnenolone, 14 C-testosterone, and 14 C-DHT levels in LNCaP cells (Fig. 4A) (Appendix Fig. S3).Protein expression of a few enzymes in steroidogenesis enzymes was altered by SEMA3C overexpression in LNCaP cells (Fig. 4B).

SEMA3C altered expression of steroidogenic and cholesterogenic genes
Effect of SEMA3C on the expression of enzymes involved in steroidogenesis pathway was confirmed by sequencing whole transcriptome of LNCaP SEMA3C and LNCaP vector cells.DEseq analysis showed that SEMA3C overexpression is correlated with the expression of multiple genes involved in steroidogenesis and cholesterol synthesis pathways.SEMA3C overexpression is associated with increased expression of SREBP targets, FASN, FDPS, and ELOVL7.However, there is a reverse correlation between expression of SEMA3C and androgen-catabolizing enzymes, AKR1C1, AKR1C2, UGT2B10, UGT2B11, UGT2B15, and UGT2B17 (Fig. 6A).Differential expression of these genes is displayed on a volcano plot (Fig. 6B).

SEMA3C enhanced castration-resistant tumor growth through de novo steroidogenesis
SEMA3C overexpressing LNCaP xenograft tumors were developed by inoculating LNCaP SEMA3C and LNCaP vector cells in nude mice, as described in the Materials and methods section.LNCaP SEMA3C xenografts had a higher tumor growth rate 4 weeks post castration than LNCaP vector tumors (Fig. 7A).This result is consistent with Peacock et al. 2018 report, showing that SEMA3C overexpression in LNCaP xenograft tumor increases their growth rate post castration.We validated SEMA3C overexpression by IHC of the LNCaP SEMA3C tumors harvested 4 weeks post castration (Fig. 7B).LC-MS analysis of steroids isolated from tumor homogenates showed that intratumoral concentration of testosterone and DHT significantly increased in LNCaP SEMA3C tumors compared to LNCaP vector tumors (Fig. 7C), consistent with the induction of key steroidogenic enzymes, STAR, HSD3B1, HSD3B2, AKR1C3, HSD17B3, and SRD5A1, along with the downregulation of androgen-inactivating enzymes AKR1C1, AKR1C2, UGT2B10, and UGT2B15 by SEMA3C.Increased PSA mRNA expression by SEMA3C suggests that SEMA3C overexpression induced PSA expression in the xenograft tumors (Fig. 7D).Also, LC-MS analysis of lipid extracts of tumor homogenates showed a statistically significant increase in free cholesterol level in LNCaP SEMA3C tumor samples compared to control tumors (Fig. 7E).Also, SEMA3C overexpression affected intratumoral lipid concentration, whose increase did not reach statistical significance (data not shown).Consistent with in vitro results, SEMA3C induced mRNA expression of enzymes and proteins of the SREBP pathway in LNCaP xenograft tumor cells, HMGCR, HMGCS1, FDPS, LDLr, FASN, and ELOVL6 (Fig. 7F).

Discussion
Intratumoral androgen synthesis is increasingly becoming recognized as a key mechanism underlying the resistance of prostate cancer cells to ADT (Stanbrough et al. 2006, Locke et al. 2009, Mostaghel et al. 2018).However, the cell-extrinsic factors that trigger the molecular mechanism(s) driving steroidogenesis in prostate cancer cells are poorly understood.Here, we demonstrate that SEMA3C, an autocrine prostate cancer growth factor, can induce the expression of key steroidogenic enzymes while simultaneously suppressing the expression of androgen-inactivating enzymes leading to increased synthesis of androgens such as testosterone and DHT from adrenal hormones such as DHEA or de novo from cholesterol.In addition, our findings suggest that Figure 6 (A) Total RNA was isolated from LNCaP SEMA3C and LNCaP vector (n = 3 per group) to be analyzed by RNA-seq.Morpheus software (https://software.broadinstitute.org/morpheus/)was utilized to generate a heatmap for the visualization of differential gene expression.Mean values of respective gene expression from two biological replicates (SEMA3C 1, 2 and Vector 1, 2) were used to provide us with a heat map based on the matrix of values which is mapped to a matrix of colors.(B) A volcano plot was generated to show the statistical significance (P-value) vs the magnitude of change (fold change). https://doi.org/10.1530/ERC-23-0010 Endocrine-Related Cancer e230010 SEMA3C mediates its effects on de novo steroidogenesis, in part, by activating the SREBP transcription factor.Our results suggest that SEMA3C may be acting through its ability to stimulate EGFR signaling since SEMA3C-induced steroidogenesis was suppressed by the presence of erlotinib.
RTK-activating growth factors are a class of cellextrinsic factors that drive steroidogenesis.It has been demonstrated that IGF2 and insulin stimulate steroidogenesis in PCa cells (van der Veeken et al. 2009, Lubik et al. 2011, 2013) while EGF signaling has been shown to stimulate steroidogenesis in Leydig tumor cells (Cameron et al. 1996, Manna et al. 2002, Shiraishi & Ascoli 2006).Our finding that SEMA3C can stimulate steroidogenesis in PCa cells is consistent with these data, since we have previously shown that SEMA3C activates multiple RTKs, such as EGFR, HER2, and MET in a cognate-independent manner (Xu et al. 2017, Peacock et al. 2018) suggesting that SEMA3C may be stimulating steroidogenesis by activating RTKs.Notably, we showed that erlotinib (an inhibitor of EGFR and HER2) attenuates SEMA3C's effects on the induction of cholesterogenic and steroidogenic enzymes in LNCaP cells suggesting that EGFR and HER2 may be the key RTKs mediating e230010

Figure 2
Figure 2 SEMA3C affected steroid synthesis in PCa cells.(A) LNCaP SEMA3C and LNCaP vector cells treated with DHEA 3.4 μM in 5% CSS-supplemented phenol red-free RPMI for 48 h.The concentration of a panel of steroids in the conditioned media derived from treated cells was analyzed by LC-MS.(B) Identical treatment for 22Rv1 SEMA3C vs control cells was followed by an LC-MS analysis to measure the concentration of steroids in the conditioned media.(C) Steroids which were isolated from the conditioned media derived from LAPC4 cells treated with recombinant SEMA3C (1 μM) or vehicle in the presence of DHEA (3.4 μM) for 48 h were analyzed by LC-MS.The concentration of steroids was analyzed in (D) C4-2 and (E) 22Rv1 cells following siRNA-mediated silencing of SEMA3C, in the presence of DHEA (3.4 μM).(n = 3, mean ± s.d.; *P < 0.05, **P < 0.01).

Figure 3
Figure 3SEMA3C induces AR-dependent gene expression.Recombinant SEMA3C 1 μM treatment of (A) LNCaP and (B) LAPC4 cells, supplemented with DHEA 3.4 μM for 24 h upregulated mRNA expression of AR target genes, quantified by RT-qPCR method.(C) LNCaP cells were transiently transfected with ARR2PB reporter plasmid and then were treated with recombinant SEMA3C 1 μM or vehicle in DHEA (3.4 μM)-supplemented media.Luciferase assay showed SEMA3C signaling increased probasin promoter activity compared to vehicle.The whole protein of the treated cells was immunoblotted for AR protein level.(D) Total RNA was isolated from LNCaP SEMA3C and LNCaP vector (n = 3 per group) to be analyzed by RNA-seq.This result shows the differential expression of a set of AR-regulated genes between LNCaP SEMA3C and LNCaP vector cells.(n = 3, mean ± s.d.; *P < 0.05, **P < 0.01).

Figure 4
Figure 4SEMA3C increases de novo steroidogenesis.(A) Steroids in media of LNCaP SEMA3C and LNCaP vector cells following 72 h treatment with 6 μCi/mL 14C-acetate in phenol red-free RPMI with 5% (v/v) CSS were identified and quantified by HPLC and radiometric detection with HPLC-coupled radiometric detection.(B) The whole protein of LNCaP SEMA3C and LNCaP vector was immunoblotted to examine the protein level of enzymes involved in steroidogenesis pathway.We found altered expression of AKR1C3, HSD3B17, HSD3B2, and UGT2B17.(n = 6, mean ± s.d.; *P < 0.05, **P < 0.01).

Figure 7
Figure 7SEMA3C promotes intratumoral androgen synthesis.SEMA3C-overexpressing tumors were developed by inoculation of 1 × 10 6 LNCaP SEMA3C or LNCaP vector in each of two injection sites of male nude mice (n = 6).(A) Tumor volumes (mm 3 ) were measured weekly in both SEMA3C and control groups, before collection 4 weeks post castration (post-Cx).(B) SEMA3C overexpression in LNCaP SEMA3C tumors was validated in RNA and protein levels by RT-qPCR and IHC methods.(C) LC-MS analysis of isolated steroids from both tumor types showed significant induction in testosterone and DHT production by LNCaP SEMA3C tumor cells vs controls.(D) Whole RNA was isolated from tumor lysates, and mRNA expression of PSA and a panel of steroidogenic and androgen-catabolizing enzymes in LNCaP SEMA3C tumors compared to LNCaP vector tumors.(E) SEMA3C gene overexpression in LNCaP tumor cells increased intratumoral concentration of free cholesterol quantified by LC-MS technique (n = 6).(F) mRNA expression of genes coding for enzymes involved in cholesterol synthesis and uptake was quantified in LNCaP SEMA3C compared to LNCaP vector tumor cells (mean ± s.d.; *P < 0.05, **P < 0.01, ***P < 0.001).