Leptin modulated microRNA-628-5p targets Jagged-1 and inhibits prostate cancer hallmarks

MicroRNAs (miRNAs) are single-stranded non-coding RNA molecules that play a regulatory role in gene expression and cancer cell signaling. We previously identified miR-628-5p (miR-628) as a potential biomarker in serum samples from men with prostate cancer (PCa) (Srivastava et al. in Tumour Biol 35:4867–4873, 10.1007/s13277-014-1638-1, 2014). This study examined the detailed cellular phenotypes and pathways regulated by miR-628 in PCa cells. Since obesity is a significant risk factor for PCa, and there is a correlation between levels of the obesity-associated hormone leptin and PCa development, here we investigated the functional relationship between leptin and miR-628 regulation in PCa. We demonstrated that exposure to leptin downregulated the expression of miR-628 and increased cell proliferation/migration in PCa cells. We next studied the effects on cancer-related phenotypes in PCa cells after altering miR-628 expression levels. Enforced expression of miR-628 in PCa cells inhibited cell proliferation, reduced PCa cell survival/migration/invasion/spheroid formation, and decreased markers of cell stemness. Mechanistically, miR-628 binds with the JAG1-3′UTR and inhibits the expression of Jagged-1 (JAG1). JAG1 inhibition by miR-628 downregulated Notch signaling, decreased the expression of Snail/Slug, and modulated epithelial-mesenchymal transition and invasiveness in PC3 cells. Furthermore, expression of miR-628 in PCa cells increased sensitivity towards the drugs enzalutamide and docetaxel by induction of cell apoptosis. Collectively our data suggest that miR-628 is a key regulator of PCa carcinogenesis and is modulated by leptin, offering a novel therapeutic opportunity to inhibit the growth of advanced PCa.

www.nature.com/scientificreports/ literature indicates that interactions with the local adipose milieu are an important driver of PCa malignancy. Adipocytes exhibit both short and long-range interaction with cancer cells and can regulate gene expression in other peripheral tissues through the secretion of adipokines. Adipokines, such as leptin, insulin-like growth factor 1 (IGF1), interleukin-6 (IL-6), and vascular endothelial growth factor (VEGF), are upregulated in obese populations and demonstrated to interfere with several signaling pathways and exert mitogenic effects 9 . Leptin, in particular, is a major adipose cytokine that controls body weight by regulating energy intake and expenditure, and clinical studies have demonstrated a positive correlation between blood leptin levels and PCa development 10 . Evidence suggests that differences in low-grade disease and patient outcomes are also driven by underlying tumor genomic differences, including overexpression or suppression of microRNAs (miRNAs) 11 . MiRNAs bind 3′UTRs (untranslated regions), and it is estimated that 60% of protein-coding genes have predicted miRNAsbinding sites 12 . MiRNAs have critical regulatory roles in gene expression affecting many cellular processes, including growth, proliferation, differentiation, metabolism, and cell death 12 . Aberrant expression and dysregulation of miRNAs have been associated with various human diseases, including cancer, autoimmune diseases, and cardiovascular and neurological disorders 13 . Bioinformatics analysis indicates that a single 3′UTR can be targeted by several miRNAs [14][15][16] . This important characteristic of miRNA biology is called 'genomic redundancy' since a miRNA is predicted to repress the expression of thousands of mRNAs, and each mRNA could be targeted by different miRNAs 14 . Due to their relevance in numerous diseases, including cancers, and their stability in various bio-fluids, they offer a novel opportunity to serve as prognostic and diagnostic biomarkers to identify cancer development and progression 7,15 . Interestingly, the complicated crosstalk between obese-related adipokines, such as leptin, includes a post-transcriptional gene regulation and an altered miRNA transcriptome, which can play a role in tumor proliferation, invasion, apoptosis, and angiogenesis 17 .
MiR-628-5p (miR-628) is located at the 15q21.3 chromosome region and commonly acts as a tumor suppressor 18 . MiR-628 has frequently been implicated as a tumor suppressor in various cancers; however, a few studies have also suggested its involvement in oncogenesis 19 . Our group was the first to report that miR-628 is differentially expressed in serum obtained from PCa patients compared to serum obtained from healthy individuals 20 . Differential expression levels of miR-628 were validated through RT/qPCR, with levels significantly lower in PCa patients when compared to healthy individuals. Decreased expression of miR-628 was observed in both AA and CA PCa populations compared to matched controls, emphasizing the discriminatory power and relevance of this miRNA in PCa 20 . The miR-628 expression has also been reported to reduce PCa cell proliferation by targeting the fibroblast growth factor receptor 2 (FGFR2) signaling pathway 21 .
The present study explored the association of adipokine leptin and miRNA-628 regulation and their contribution to PCa development/suppression. In addition, we investigated the effect of leptin in the modulation of miR-628 expression and PCa cell aggressiveness. We have demonstrated that overexpression of miR-628 in PCa cells inhibited cell proliferation/migration/invasion/spheroid formation by targeting Jagged1, suggesting that miR-628 acts tumor suppressor.
Human leptin (Cat# Z02962-200, GenScript) was reconstituted in water following the manufacturer's instructions. Before leptin treatment, cells were grown in serum-free media for 24 h, and then cells were treated with increasing concentrations of leptin. The effect of leptin on miR-628 expression was analyzed. Jagged-1 peptide (Cat #188-204), Notch ligand peptide (Cat# AS-61298), and a Jagged-1 scrambled control (Cat# AS-64239) were obtained from Anaspec (Ca, USA). Both peptides were reconstituted and utilized according to the manufacturer's instructions. In some experiments, cells were exposed to enzalutamide (Selleckchem; Cat. # S1250) and docetaxel (Selleckchem; Cat. # No. S1148) after transfection with miR-628 or negative control (NC), and the effect of miR-628 on drug sensitivity was determined.

Wound healing assay.
Cells were cultured at a density of 1.5-2.0 × 10 5 per well in 6-well plates in triplicates and transiently transfected with NC mimic (100 nM) or miR-628 mimic (100 nM). After 24 h, a vertical scratch was made using a pipet tip (200 µl). Cells were imaged at 0 and 18 h by using a phase-contrast Zeiss microscope. Wound length was measured utilizing the ZEN software in pixels, and the percentage of wound closure was calculated using the formula [(A 0 -A 24 )/A 0 ] × 100.
Spheroid culture assay. Spheroid culture assay was performed as previously described 24  Migration assay. Permeable cell culture inserts (Thomas Scientific, Swedesboro, NJ, USA) were used to assess cell migration. Cells were transfected with NC mimic (100 nM) or miR-628 mimic (100 nM) and then starved in serum-free media for 24 h. After starvation, cells were trypsinized and suspended in serum-free media at a density of 0.5 × 10 6 cells/ml. Then 300 µl of cell suspension was added to the interior of each insert (triplicates). Complete media with 10% serum was used as an attractant and added to the lower chamber. After 24-48 h of incubation, migrating cells at the bottom of the insert were fixed in 10% methanol and stained with 0.5% crystal violet solution. Inserts were submerged in PBS to wash residual stain and allowed to air dry. Cells were visualized under a light microscope and quantified alternatively by dissolving stained cells in 10% acetic acid. Equal amounts (100 µl/well) of dye/solute mixture were then transferred to a 96 well plate (duplicated per insert), and absorbance was determined at 562 nm using a Fluostar Omega plate reader (BMG Lab Tech, Cary, NC, USA).
Invasion assay. Invasion assay was performed using permeable cell culture inserts (Thomas Scientific, Swedesboro, NJ) coated with 30 µl of Corning ® Matrigel ® Growth Factor Reduced (GFR) Basement Membrane Matrix (Corning, NY). Inserts were incubated for 30 min at 37 °C to allow the matrigel to solidify. We followed a procedure similar to the migration assay described above but utilized matrigel coated inserts.
Statistical analysis. Statistical analysis was performed utilizing the SigmaStat 4.0 software and GraphPad Prism 9. To evaluate differences between experimental variables, we utilized a t-test or one-way or two-way ANOVAs with Tukey's multiple comparisons test. Differences were considered statistically significant at p values equal to or below 0.05 (p < 0.05). Data is presented as the following: ns p > 0.05, *p < 0.05, **p < 0.01, or ***p < 0.001. Data was also plotted using GraphPad Prism 9.

MiR-628 is downregulated in PCa patient tumors and the adipokine leptin regulates its expression in PCa cells.
We first analyzed the basal expression of miR-628 in the human epithelial prostate tissue cell line RWPE-1 (derived from healthy non-cancerous tissue) and in various PCa cell lines of various lineages (LNCaP, PC3, DU145, and MDA-PCa-2b) by RT/qPCR (Fig. 1A). We observed lower expression of miR-628 in the PCa cell lines LNCaP, DU145, and MDA-PCa-2b compared to normal RWPE-1 cells (p < 0.001). However, the expression of miR-628 was not significantly changed between PC3 cells compared with normal prostate RWPE1 cells.
We previously demonstrated that miR-628 was downregulated in a subset of PCa patient sera compared to normal controls and that miR-628 levels could effectively discriminate between disease and non-diseased. We assessed miR-628 expression in a cohort of prostate cancer patient tissues compared to benign adjacent control tissues to further confirm these results. The expression of miR-628 was significantly downregulated (p < 0.0001) in the tumor samples compared to matched benign controls (Fig. 1B). Additionally, we assessed miR-628 levels utilizing the miRNA Target Viewer (miR-TV). This software utilizes the TCGA-PRAD (The Cancer Genome Atlas, Prostate Adenocarcinoma) dataset to show miRNA expression in 52 normal versus 494 tumor tissues 25  www.nature.com/scientificreports/ TCGA-PRAD data revealed that miR-628 expression is significantly downregulated (p < 0.02) in patient tumor samples compared with the control group (Fig. 1S), suggesting that miR-628 expression is downregulated in PCa cells or tumors. Due to the strong correlation between increased levels of adipokine leptin and PCa aggressiveness 10 , we analyzed the effect of leptin on miR-628 regulation. Interestingly, exposure to LNCaP and PC3 cells with 5 ng/ ml, 10 ng/ml, 25 ng/ml, or 50 ng/ml of leptin for 48 h significantly decreased miR-628 expression in both LNCaP and PC3 cells (Fig. 1C, left and right panels). Together, our data indicated that miR-628 is downregulated in PCa tumor samples compared to normal matched controls, and the expression of miR-628 is controlled by adipokine leptin in PCa cells.

Leptin abrogates miR-628 impact on cell proliferation/colony formation in PCa cells .
Since leptin exposure downregulated miR-628 expression in PCa cellular models, we assessed LNCaP and PC3 cell proliferation/survival after transfection with either NC or miR-628 and after treatment with leptin (10 ng/ml) ( Fig. 2A). As expected, miR-628 overexpression decreased cell proliferation in both LNCaP and PC3 cells at all indicated time points (*p < 0.05, **p < 0.01, and ***p < 0.001). Transfection of NC and leptin treatments increased cell survival in LNCaP and PC3 cells when compared to miR-628 transfected cells. Interestingly, miR-628 expression and leptin treatment decreased cell proliferation compared with NC transfected and leptin treated cells, suggesting that leptin downregulates miR-628 expression and hence induced cell proliferation in PCa cells ( Fig. 2A, upper and lower panels). Finally, the impact of leptin on PCa cell colony formation was determined in LNCaP and PC3. Our data further suggest that leptin treatment increased cell colony formation in LNCaP and PC3 cells (Fig. 2B, upper and lower panels). Transfection with miR-628 mimic and leptin treatments decreased cell colony formation, specifically in LNCaP cells indicating that leptin effectively decreases endogenous miR-628 expression and enhances cell colony formation (Fig. 2B upper and lower panels). In conclusion, the adipokine leptin downregulated miR-628 expression and increased cell proliferation/ cell colony formation and miR-628 overexpression partially abrogated the proliferative effects of this hormone.
Overexpression of miR-628 affects PCa cell proliferation. Since the expression of miR-628 is downregulated by leptin in PCa tumors and PCa cells, we asked whether altering miR-628 expression affects PCa cell survival and proliferation. For this purpose, we transfected four PCa cell lines with CY labeled miR-628 mimic or Negative Control (NC) mimics. The resulting expression levels were assessed using a fluorescence microscope to determine relative transfection efficiency (Fig. 3A). We confirmed efficient transfection of both miR-628 mimic and NC in all four cell lines (Fig. 3A). MiR-628 expression was significantly increased after transfection of miR-628 mimic, as shown by RT/qPCR (Fig. 3B).
Further, the effect of miR-628 on cell survival was analyzed by MTT assay on indicated prostate RWPE1 cells and PCa cell lines. We observed a modest reduction in cell survival in RWPE1 cells at 48 h but not significant at the 72 h time point (Fig. 3C, left panel). When we analyzed the impact of miR-628 on cell survival in the PCa cell lines (PC3, LNCaP, and DU145), we found that overexpression of miR-628 significantly reduced cell survival in all indicated PCa cell lines (**p < 0.01, ***p < 0.001) (Fig. 3C). Interestingly, PC3 cells were minimally responsive at earlier time points (Fig. 3C) but were significantly affected at later time points.
We next investigated the effect of miR-628 on PCa cell proliferation using live-cell analysis by Incucyte for 72 or 120 h (Fig. 3D, upper panels), and the results were further confirmed with MTT assay at the end of the www.nature.com/scientificreports/ selected period (Fig. 3D, lower panel). Our data demonstrated that expression of miR-628 mimic (red line) effectively impaired cell proliferation in all the PCa cell lines over time compared to NC control (blue line), and the reduction in cell proliferation was statistically significant when measured by MTT assay (Fig. 3D, lower panel). Collectively our data indicate that miR-628 overexpression reduced cell survival/proliferation in PCa cell lines.

Expression of miR-628 affects cell colony and spheroid formation in PCa cells.
To further characterize the effects of miR-628 overexpression in PCa, we performed a cell colony formation assay after overexpression of miR-628 or NC. Cell colonies were defined with at least 50 cells/colony 10 days after plating (Fig. 4A). We selected PC3, DU145, and LNCaP PCa cells as in previous experiments and added another PCa cell line MDA-PCa-2b. Transfection with miR-628 mimic significantly reduced colony formation in 22Rv1 (50% reduction), PC3 (79% reduction), DU145 (45% reduction), MDA-PCa-2b cells (37% reduction) and LNCaP cells (69% reduction) compared to cells transfected with NC mimic (*p < 0.05, **p < 0.01, ns-not significant) (Fig. 4A, upper and lower panels). It is known that the ability of cells to form spheres is an indicator of the percentage of cells with stem-like characteristics present in a population of cells 26 . To investigate the role of miR-628 in this hallmark of cancer development, we transfected PC3, 22Rv1, MDA-PCa-2b, and DU145 cells with NC control or miR-628 mimic, and cells were grown in non-adherent plates under serum-free conditions, and spheroid development was observed after 10 days (Fig. 4B, upper panel). A significant reduction in both spheroid size and the number was observed after transfection with miR-628 mimic compared to cells transfected with NC in the PCa cell lines (22Rv1, PC3, MDA-PCa-2b, and DU145) (*p<0.05, **p<0.01, ***p<0.001, ns-not significant) (Fig. 4B, upper and lower panels). Furthermore, we studied the expression of CD44 and Nanog, commonly used cancer stem-cell markers 27 after transfection of PC3 cells with NC, miR-628, and miR-628 inhibitor. We observed that PC3 cells transfected with miR-628 mimic had decreased expression of CD44 compared to cells transfected with NC mimic control (Fig. 4C). However, this effect was not observed when the expression of Nanog was evaluated. Interestingly, transfection with miR-628 inhibitor (100 nM) had the opposite effect, increasing the expression of Nanog compared to mimic. The expression of CD44 also increased when PC3 cells were transfected with miR-628 inhibitor www.nature.com/scientificreports/ compared to cells transfected with miR-628 mimic (Fig. 4C). Together, these results indicate that miR-628 affects PCa cell proliferation and reduces cell colony formation and spheroid formation in PCa cells. These data support the potential role of miR-628 in regulating properties related to PCa cell stemness.

MiR-628 targets JAG1-3′UTR and affects JAG1 expression and downstream Notch signaling PCa cells.
To determine the mechanistic role of miR-628 in PCa suppression we used the Target scan software (http:// www. targe tscan. org/ vert_ 72/) to analyze genes possibly targeted by miR-628. Target scan analysis suggests that miR-628 can bind with several targets, including the 3′UTR of JAG1, a ligand of Notch signaling (Fig. 5A, upper panel). JAG1 overexpression has been correlated with PCa disease recurrence, therapy resistance, and metastasis 28,29 . To confirm this, we performed luciferase assays to determine if miR-628 can activate or repress the expression of JAG1. Co-transfection of JAG1-3′UTR-Luc construct with wild-type miR-628 Results were further confirmed with MTT assay (lower panels). At the end of the selected period, absorbance was read at 570 nm (***p < 0.001). Data represent average ± SEM. Experiments were performed in triplicates. www.nature.com/scientificreports/ decreased luciferase activity; however, mutant type miR-628 expression restored luciferase activity significantly in both LNCaP and PC3 cells (Fig. 5A, lower panel). We also assessed JAG1 expression in RWPE-1 and various PCa cell lines. Immunoblotting data shows that the expression of JAG1 protein is variable between RWPE1 and PCa LNCaP, PC3, and DU145 cells (Fig. 5B). Furthermore, to study if JAG1 was differentially expressed between "matched normal" or benign and PCa samples obtained from AA and CA men, we utilized dataset GSE41969 from the GEO and analyzed it using the GEO2R web tool. We had 163 benign and 629 PCa samples (80 benign AA samples, 270 PCa AA samples, 82 benign CA samples, and 369 CA samples). JAG1 was significantly overexpressed in both AA and CA PCa samples compared to their respective normal matched controls (p < 0.0001) (Fig. 5C). JAG1 was also statistically overexpressed in the tumor samples compared to the benign samples, regardless of race (p < 0.0001) (Fig. 2S). We chose PCa cell lines LNCaP and PC3 to continue future studies. We quantified the mRNA levels of JAG1 after transfection of miR-628, and our data indicate that overexpression of miR-628 downregulated mRNA levels of JAG1 and its downstream target ZEB1 significantly in PC3 cells (Fig. 5D, left and right panels). An earlier study suggested that as a Notch receptor ligand, JAG1 promotes epithelial-mesenchymal transition (EMT) 30 . Since miR-628 affects JAG1 expression, we analyzed the effect of miR-628 on the expression of JAG1, and its receptors Notch1, Notch2, Notch3, and Notch downstream target Hes1. Immunoblotting data demonstrated that miR-628 overexpression downregulated the expression of Notch ligand, JAG1, and also downregulated the expression of Notch1, Notch2, Notch3, and Notch downstream marker HES1 in both LNCaP and PC3 cell lines (Fig. 5E) suggesting that miR-628 targets JAG1 and inhibits Notch signaling in PCa cells.

Scientific Reports
Further, we examined the effect of miR-628 overexpression or knockdown on JAG1 cytosolic and nuclear shuttling. We observed a decrease in JAG1 in both nuclear and cytosolic fractions compared to NC controls. JAG1 expression increased when PC3 cells were transfected with miR-628 inhibitor. The expression of the Notch intracellular domain (cleaved notch) revealed that overexpression of miR-628 decreased detection of NICD in the nuclear fraction and inhibitor transfection had opposite effects. Finally, we observed a decrease in downstream target Hes1 in both nuclear and cytosolic fractions of PC3 cells transfected with miR-628. On the other hand, transfection with miR-628 inhibitor increased Hes1 expression compared to miR-628 mimic (Fig. 5F). Together, these results suggest that miR-628 targets JAG1 and its downstream signaling in PCa cells.
To determine if activation of the JAG1-Notch pathway is affected by miR-628 expression, we utilized a JAG1mimicking peptide to initiate activation of this pathway through the measurement of Notch1 cleavage (Fig. 5G). Similar JAG1 peptides have been previously utilized to study the effects of the activation of this pathway under a variety of conditions 31 . We first treated PC3 cells with 100 nM of miR-628 mimic or NC for 48 h and then after initiation of the Jagged-Notch pathway compared to SC. Interestingly, Notch1 cleavage was reduced in PC3 cells transfected with miR-628 compared with PC3 cells treated with NC under the same conditions. We did observe an increase in cleavage 30 min after initiation of the Jagged-Notch pathway using JAG1 peptide compared to SC, with a decrease at 2 h. Next, we assessed PARP cleavage under these same conditions. We observed minimal cleavage in PC3 cells transfected with NC control and treated with either SC or JAG1 peptide. On the other hand, we observed an increase in PARP cleave in PC3 cells transfected with miR 628 and treated with SC. The observed PARP cleavage was reduced after initiation of the JAG1-Notch pathway using the JAG1 peptide, showing a possible override of the effects of miR 628 mimic. Together, these results suggest that miR-628 targets JAG1 and may modulate the expression of its targets.

MiR-628 expression affects PCa cell migration, invasion, and tube formation in HUVEC cells.
Since miR-629 targets the 3′UTR region of JAG-1, we wanted to study the effect of miR-628 overexpression in different downstream targets and PCa cell migration, invasion, and its impact on tube formation in HUVEC cells. Using an in vitro scratch assay, we quantified the effect of miR-628 on PCa cell migration (Fig. 6A). Our data suggest that overexpression of miR-628 significantly reduced cell migration in PC3 (48%) and DU145 (55%) cells compared with NC transfected PC3 (28%) and DU145 (29%) cells (Fig. 6A, left and right panels). The effect of miR-628 on PCa cell migration and invasion was further analyzed by trans-well assay (Fig. 6B,C). MiR-628 overexpression downregulated PC3 cell migration compared to NC mimic control (49% after 24 h and 45% after 48 h) (Fig. 6B) and decreased invasion in PC3 cells transfected with miR-628 (48% after 48 h) compared with NC transfected PC3 cells (Fig. 6C). Mechanistically the expression of Snail/Slug and EMT markers play an important role in cell migration, invasion, and aggressiveness in PCa tumor cells 32,33 . RT/ qPCR data indicate that overexpression of miR-628 inhibits Snail and Slug expression in LNCaP cells and Snail expression in PC3 cells (Fig. 6D). We also assessed Snail and Slug protein expression via immunoblotting in PC3 cells transfected with miR-628 mimic or NC (Fig. 6E), and our data suggest that PC3 cells transfected with miR-628 showed a decreased expression of Slug (0.58) and Snail (0.61) compared to NC (Fig. 6E). PC3 cells transfected with miR-628 had decreased expression of zinc finger E-box-binding homeobox 1 (ZEB1) (0.38) The effect of miR-628 on the expression of E-cadherin (ECAD) and N-cadherin (NCAD) was measured by RT/qPCR in LNCaP and PC3 cells after transfection with 100 nM of miR-628 mimic or NC (left panels). *p < 0.05, ***p < 0.001 (H) E-cadherin and N-cadherin protein expression was analyzed by Western blotting as described in the materials and methods section after transfection with 100 nM of miR-628 mimic or NC. Membranes were cut before hybridization using the molecular weight marker as a reference and imaged using an Azure imaging system. All western blot full images are provided in supplementary information files. (I) Angiogenesis was determined in HUVEC cells transfected with miR-628 or NC (upper panels). Representative images were captured using an inverted light microscope and representative images are shown at 20 × magnification. The tube length and number were determined at 24 h using ImageJ (lower panels; **p < 0.01, ***p < 0.001).  (Fig. 6F). Increased expression of ZEB1 has been shown to enhance motility and invasiveness, thus promoting an aggressive phenotype in PCa cells 34 . We also analyzed the expression of EMT markers E-cadherin and N-cadherin after overexpression of miR-628 by RT/qPCR (Fig. 6G) and immunoblotting (Fig. 6H). Overexpression of miR-628 decreased expression of N-cadherin mRNA levels in PC3 cells, but no significant changes were observed at the protein level. Transfection of miR-628 led to decreased protein expression of N-cadherin in LNCaP cells, but no significant changes were observed when mRNA was measured. Overexpression of miR-628 decreased expression at both the mRNA and protein levels of E-cadherin in LNCaP cells and decreased its protein expression in PC3 cells (Fig. 6G,H). Since effective tumor growth and eventual metastasis are dependent on angiogenesis 35 , we next investigated the role of miR-628 in angiogenesis (Fig. 6I). HUVEC cells were transfected with miR-628 mimic or NC mimic for 24 h, cells were re-plated on matrigel coated plates, and vascular tube formation was monitored after 24 h (Fig. 6I). Our results show that miR-628 significantly reduced tube number (**p < 0.01) and length (pixels) (***p < 0.001) compared to HUVEC cells transfected with NC. These results show that miR-628 expression downregulates PCa cell hallmarks, namely migration, invasion, and angiogenesis.

Expression of miR-628 increases sensitivity towards enzalutamide and docetaxel and induces apoptosis.
Drug resistance is one of the main drivers of adverse outcomes in PCa patients 36 . Drugs such as enzalutamide (ENZ) effectively inhibit androgen receptor (AR) signaling, which plays an important role in the development of PCa disease 37 . Docetaxel (DTX) binds tubulin and stabilizes microtubules, improving overall survival in men suffering from metastatic disease 38 . Since the aberrant expression of miRNAs has been linked to chemoresistance and the development of CRPC 39,40 , we explore the role of miR-628 in drug resistance. Both LNCaP (AR-positive) and PC3 (AR-negative) cells were transfected with NC or miR-628, and cells were then treated with increasing concentrations of Enz (5 µM and 10 µM) or DTX (5 nM and 10 nM), and relative cell survival was measured by MTT assay (Fig. 7A, Fig. 3S). Expression of miR-628 decreased cell survival after 48 h compared with NC transfected cells as expected (Fig. 7A, Fig. 3SA upper and lower panels). LNCaP cells transfected with miR-628 did not show a significant decrease in cell survival when treated with increasing concentrations of DTX as compared to NC transfected and vehicle-treated cells (Fig. 7A, upper panels). In contrast, PC3 cells transfected with miR-628 did show a significant decrease in cell survival after treatment with increasing concentrations of DTX compared with NC transfected and vehicle-treated PC3 cells (Fig. 3SA, lower panels). Moreover, miR-628 overexpression and ENZ treatments significantly decreased LNCaP cell survival compared to miR-628 transfected cells, and no significant ENZ effects were observed in PC3 cells since PC3 cells do not express androgen receptor (AR). In addition, we analyzed the effect of miR-628 on cell apoptosis after treatment www.nature.com/scientificreports/ of DTX and ENZ. Our data suggested that overexpression of miR-628 induced cleaved-PARP expression and decreased pro-Caspase-3 expression, indicating that miR-628 induced cell apoptosis in both LNCaP and PC3 cells (Fig. 7B, Fig. 3SB, upper and lower panels). Interestingly, overexpression of miR-628 combined with DTX or ENZ treatments also induced cleaved-PARP expression and decreased pro-Caspase-3 expression in both LNCaP and PC3 miR-628 cells compared with NC and DTX or ENZ treated cells (Fig. 7B, Fig. 3SB, upper and lower panels). These results suggest that a combination of miR-628 and the clinically approved drugs ENZ and DTX could have synergistic or additive effects in decreasing PCa cell proliferation and possibly enhancing cell death.

Discussion
Obesity is a risk factor for both cancer development and increased mortality from this disease 41 . Due to the ubiquitous nature of adipose tissue, many types of solid tumors grow close to or in direct contact with adipocytes and other adipose-associated cell populations. The specific nature of the reciprocal communication between a developing tumor and adjacent adipose tissue is an area of active study. A growing body of literature indicates that these interactions with the local adipose milieu are important drivers of malignancy 42 . Adipocytes exhibit both short-and long-range interactions with cancer cells and can regulate gene expression in other peripheral tissues through the secretion of adipokines. For example, blood levels of adipokines are associated with PCa development 9 . The complicated crosstalk between obese-related adipokines, such as leptin, includes post-transcriptional gene regulation that may result in an altered miRNA transcriptome that may affect tumor proliferation, invasion, apoptosis, and angiogenesis 17 . Overall, our results show that increased leptin levels can result in a significant decrease in miR-628 expression and that this decrease increases markers of PCa cell aggressiveness. This has serious implications since obesity has already been associated with the risk of developing aggressive disease and recurrence due to decreased biopsy and prostate-specific antigen (PSA) efficiency and overall larger prostate size 43 . Regulation of miRNAs by hormones such as leptin could further explain the complex interaction of adipose and surrounding tissue that contributes to PCa disease, especially in cases where there is a high level of adiposity and adipokine secretion. Our results show that co-treatment with miR-628 and leptin partially abrogated the proliferative effects of this adipokine. A more comprehensive understanding of the role of leptin in regulating miRNAs in PCa is necessary, and the novelty of this discovery could potentially lead to more insights into the development of this disease.
MiRNAs are small non-coding RNA molecules with specific sequences that bind to corresponding target sequences in mRNA or DNA and suppress gene expression 44 . Aberrant expression of miRNAs has been shown to contribute to a wide range of diseases, including cancer 44 . In the current study, we found that miR-628 expression levels were lower in the PCa cell lines compared to non-cancer-derived prostate cells. This is consistent with our previous study 20 showing that patients with PCa have decreased expression of miR-628 compared to controls. In addition, increasing expression or "re-expression" of miR-628 effectively repressed tumorigenic properties in PCa cell lines, namely proliferation, colony formation, migration, and invasion. These results suggest that the decrease or loss in miR-628 expression might play a role in the development of PCa. Interestingly, increased expression of miR-628 in HUVEC cells decreased tube size and number compared to control, effectively abrogating angiogenesis. Increased angiogenesis and lymphangiogenesis are hallmarks of cancer formation, aiding cells in surviving the hypoxic conditions associated with solid tumors 45 . This is a relatively unexplored aspect of miR-628 function and warrants further study.
MiR-628 can target multiple genes, but we chose to explore its interaction with Jagged-1 due to its relevance in PCa and its correlation with aggressive PCa and poor prognosis 28 . MiR-628 directly targets the 3′UTR of JAG1 and decreases the expression of JAG1 at both the mRNA and protein levels. Furthermore, our luciferase gene reporter assay showed that mutations in the interacting region of miR-628 effectively upregulated the JAG1-Luc activity, confirming the binding of miR-628 to the UTR of JAG1. Next, we observed that miR-628 overexpression in PCa cell lines effectively decreased levels of multiple JAG1 downstream targets. JAG1 is one of five canonical ligands of Notch receptors, and the ligand-induced Notch signaling system is a well-characterized signaling system that plays critical roles in development, tissue homeostasis, and disease 29 . Dysregulation of Notch signaling has been found in various diseases, including cancer, and several studies reported the involvement of JAG1 in cancer cell invasiveness, EMT, and metastasis 46 . These results highlight that miR-628 re-expression could potentially abrogate the activation of the JAG1-Notch pathway and decrease EMT, which is essential for tumorigenesis. Notably, the variation in mRNA and protein levels observed in some of the targets could be attributed to various factors, such as post-transcriptional and post-translation regulation and cellular response to highly dynamic scenarios, particularly under stressful conditions, such as a cancer cell microenvironment [46][47][48][49] . We also observed that Nanog levels were not affected by the miR-628 mimic compared to NC control, but a robust overexpression was observed when the miR-628 inhibitor was utilized. It seems that miR-628 effects in Nanog specifically are not as significant when overexpressed but rather when the inhibitor is utilized to decrease miR-628 expression. Although we do not have evidence that miR-628 directly binds to the 3′UTR region of Nanog, we can hypothesize that absence of miR-628 might contribute to changes in the expression of another miR-628 target that regulates Nanog or could be the effects of post-transcriptional modifications.
In our studies, we wanted to see the effects of combining miR-628 mimic and clinically relevant drugs such as ENZ and DTX to study if co-treatment could increase the effectiveness of these drugs. Although PC3 cells are AR-independent cell lines and the effects of ENZ treatment in proliferation were not as significant, we did observe an increase in PARP cleave and a reduction in the pro-caspase 3. The PC3 cells could be utilized to study AR-independent disease and the ENZ resistance model since cancer cells can develop alternative mechanisms, such as GR overexpression, to effectively develop therapeutic strategies to circumvent the effects of AR blockage 50 .
Due to their stability in various body fluids such as blood and serum, miRNAs are used as diagnostic tumorspecific biomarkers in cancer 51 . Our group previously reported a significant decrease in miR-628 in the plasma  20 . This is significant given that the currently utilized biomarker for PCa, PSA, has limited diagnostic usefulness due to the variability in expression in both serum and cancer tissue 52 . This biomarker variability could result in unnecessary biopsy and even removal of the prostate in cases where the disease was not likely to become aggressive 53 . On the other hand, revised recommendations for delayed testing have left the AA males vulnerable to a more aggressive disease with higher mortality in their community and a higher tumor volume despite similar PSA levels to their Caucasian counterparts 54 . This highlights the importance of utilizing additional markers of disease, such as miR-628 to appropriately identify and screen men that could potentially develop more aggressive PCa disease. Overall, our results show that miR-628 is downregulated in PCa patient tissue, and its overexpression reduced key cancer hallmarks, including proliferation, colony formation, invasion, migration, angiogenesis, and decreased drug resistance. We also show that miR 628 directly targets JAG1 transcription and its overexpression downregulated specific downstream targets in the JAG1-NOTCH signaling pathway involved in EMT, as presented in our model (Fig. 7C). Furthermore, leptin downregulated miR-628 and increased cell proliferation, while overexpression of miR-628 could partially abrogate the proliferative effects of this hormone. Overall, miR-628 effects in PCa cells have significant implications for PCa development, and together with our previous work, miR-628 could be further explored as a promising biomarker for the disease.

Data availability
All data generated during the current study are included in this article and will be available from the corresponding author on request.