Nischarin expression may have differing roles in male and female melanoma patients

Due to the development of resistance to previously effective therapies, there is a constant need for novel treatment modalities for metastatic melanoma. Nischarin (NISCH) is a druggable scaffolding protein reported as a tumor suppressor and a positive prognostic marker in breast and ovarian cancers through regulation of cancer cell survival, motility and invasion. The aim of this study was to examine the expression and potential role of nischarin in melanoma. We found that nischarin expression was decreased in melanoma tissues compared to the uninvolved skin, and this was attributed to the presence of microdeletions and hyper-methylation of the NISCH promoter in the tumor tissue. In addition to the previously reported cytoplasmic and membranous localization, we observed nischarin in the nuclei in melanoma patients’ tissues. NISCH expression in primary melanoma had favorable prognostic value for female patients, but, unexpectedly, high NISCH expression predicted worse prognosis for males. Gene set enrichment analysis suggested significant sex-related disparities in predicted association of NISCH with several signaling pathways, as well as with different tumor immune infiltrate composition in male and female patients. Taken together, our results imply that nischarin may have a role in melanoma progression, but that fine-tuning of the pathways it regulates is sex-dependent. Nischarin is a tumor suppressor whose role has not been investigated in melanoma. Nischarin expression was downregulated in melanoma tissue compared to the normal skin. Nischarin had the opposite prognostic value in male and female melanoma patients. Nischarin association with signaling pathways differed in females and males. Our findings challenge the current view of nischarin as a universal tumor suppressor. Nischarin is a tumor suppressor whose role has not been investigated in melanoma. Nischarin expression was downregulated in melanoma tissue compared to the normal skin. Nischarin had the opposite prognostic value in male and female melanoma patients. Nischarin association with signaling pathways differed in females and males. Our findings challenge the current view of nischarin as a universal tumor suppressor.


Introduction
Malignant melanoma is the deadliest form of skin cancer, with increasing incidence in both men and women [1]. Annually, about 1.7% of all newly diagnosed primary malignant cancer cases and approximately 0.6% of all cancer deaths worldwide are due to the cutaneous melanoma [2]. If diagnosed and treated early, melanoma is curable in 93% of patients [1], but the patients with distant metastases have a dire prognosis. In the last decade, advancements in targeted and immunotherapies, particularly MAP kinase pathway inhibitors (BRAF, MEK) and immune checkpoint blockers (CTLA-4, PD-1), have dramatically increased median survival for patients with non-resectable metastatic disease from 6-10 months to 24 months [3]. Unfortunately, significant proportion of patients does not respond to treatment, or develop resistance and eventually relapse [4]. In addition, many melanoma patients worldwide do not have access to first-line innovative therapeutics [5]. Therefore, there is an urgent need for both novel therapeutic targets and affordable treatment strategies for melanoma.
Nischarin (NISCH, Imidazoline receptor 1 (IR1), Imidazoline receptor antisera selected (IRAS)) is a scaffolding protein, first cloned from the human hippocampal library twenty years ago [6]. It was first identified as a cytosolic protein anchored to the inner layer of the plasma membrane, bound to alpha 5 beta 1 (α5β1) integrin and involved in cytoskeletal organization [7]. Nischarin is also of great interest as a target in drug discovery as a non-adrenergic imidazoline receptor and has been shown to couple to several transduction pathways including cAMP pathway, G protein coupled receptor signaling, MAPKs ERK and c-Jun and nitric oxide [8]. Data from the Human Protein Atlas (HPA) repository indicate that the NISCH is expressed in almost all human tissues, with the highest expression in cerebellum, adrenal gland, bronchus, rectum, gallbladder, heart muscle, and skin [9]. In cancer, NISCH has been defined as a tumor suppressor, with decreased expression in breast and ovarian cancer compared to the healthy tissue [10,11]. NISCH gene is located at the 3p21.1, in the region described as a putative tumor suppressor cluster [12]. The tumor suppressive role of NISCH stems from its ability to regulate cytoskeletal signaling through interaction with the integrin α5 and formation of focal adhesions and invadopodia [7,[13][14][15]. This ultimately has an impact on cancer cell survival, motility and invasion. Nischarin has also been shown to alter cancer cell energy metabolism through LKB1-AMPK-mTOR axis [16]. As two major obstacles in melanoma treatment -invasive potential and metabolic plasticity -are potentially linked to NISCH, NISCH is a great target for investigation in the context of melanoma progression and treatment. Importantly, several Food and Drug Administration (FDA) approved NISCH agonists are currently in clinical use [8] which might enable quick integration into melanoma patients treatment options.
Nischarin expression and role in melanoma have not been investigated to date. This study aimed to assess NISCH mRNA and protein expression, transcriptional regulation, NISCH prognostic value, as well as its potential role in melanoma progression, by examining publicly available datasets and melanoma patient samples, and taking into account sexrelated differences.

Materials and methods
Detailed methods are available in Supplemental Material.

NISCH expression during melanoma progression
We first examined NISCH mRNA expression in three publicly available datasets containing samples of normal skin, nevi and primary and metastatic melanoma tissue. Primary melanoma tumor tissue expressed significantly less NISCH mRNA than nevi or normal skin; while decrease in expression in nevi compared to the skin was significant in GSE3189 [17] and presented as a trend in GSE46517 [18] (Fig. 1a). There was no difference in NISCH mRNA expression between the primary tumor and the metastatic sites when the metastases samples were grouped into lymph node metastases and other metastases (Fig. 1b). To validate these findings, we examined melanoma tissues from patients operated at our institution (IORS cohort). While there was no difference in expression between primary tumors and other metastatic sites, lymph node metastases had significantly lower NISCH expression compared to primary melanoma samples (Fig. 1c).
Next, we examined the protein expression and localization of NISCH in the samples described in the HPA and in our patient samples. In the HPA that uses the HPA023189 antibody, NISCH was detected in healthy skin (Fig. 2a), with high expression in melanocytes, keratinocytes and 1 3 Langerhans cells and medium expression in fibroblasts. The localization of the protein was described as cytoplasmic and membranous, which corresponds to the so far described functions of the protein. In the melanoma samples in the HPA, NISCH was described as weakly expressed in the 50% of the samples, and intriguingly in 75% of the samples its localization was described as nuclear (Fig. 2b). In IORS primary melanoma patient samples NISCH could be observed both in the dermis and the epidermis, while in the dermal compartment it was present both in the cytoplasm and the nuclei (Fig. 2c). Vimentin and H&E staining were performed in consecutive sections to label the mesenchymal compartment. NISCH protein expression could also be detected in the lymph node and distant metastases, and the amount of protein corresponded to the level of mRNA (Fig. 2d). In the metastatic samples too, the staining could be observed both in the cytoplasm and the nucleus (Fig. 2e). These results imply that in melanoma NISCH could have so far unreported role in the nucleus.

Regulation of NISCH gene expression in melanoma
As we observed a decrease in NISCH mRNA expression in primary tumors compared to nevi in both IORS patient cohort and publicly available datasets, as well as possible decrease in expression in lymph node metastases compared to primary tumors, we next examined the mechanisms of NISCH down-regulation in melanoma. In breast cancer, loss of NISCH expression was reported as a consequence of a loss of heterozygosity (LOH) on the chromosomal locus 3p21 and microdeletions in introns 2, 3 and 6 of the NISCH gene [14]. In ovarian cancer, silencing was reported to be mainly due to the NISCH promoter hyper-methylation [10]. In melanoma, we first analyzed NISCH gene copy-number in TCGA SKCM and DFCI [19] datasets. These datasets contain only metastatic melanoma samples for copy-number analysis. NISCH shallow deletions were detected in around 20% of TCGA SKCM and 15% of DFCI samples. Samples in which shallow deletions were detected had significantly lower NISCH expression (Fig. 3a, p = 0.0001 for TCGA SKCM, p = 0.0183 for DFCI) and samples with the gain of gene copies had higher NISCH expression compared to the diploid samples (Fig. 3a, p = 0.0004 for TCGA SKCM). In our patient cohort, we detected microdeletions in the NISCH gene promoter (ΔKCT below − 0.65) in 30% of lymph node samples and in 40% of other metastases samples (Fig. 3b). Collectively, these results imply that in metastatic melanoma microdeletions may be a common mechanism for decreased NISCH mRNA expression.
Next, we analyzed the methylation status of the NISCH promoter in the TCGA SKCM and GSE86355 [20] datasets. In the TCGA SKCM dataset, the methylation status did not NISCH mRNA expression in (a) benign and primary melanoma tissues included in GSE3189 and GSE46517 datasets, and (b) primary tumors and metastatic sites from TCGA SKCM and GSE46517 datasets. Data were analyzed by one-way ANOVA with Tukey's mul-tiple comparisons test. * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001 (c) Relative NISCH mRNA levels in samples of IORS patients measured by qRT-PCR and analyzed using one-way ANOVA with Dunnett's multiple comparisons test. * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001 significantly differ in melanoma compared to the normal skin samples, but in GSE86355 dataset there was a trend of increased methylation with tumor progression and in metastatic melanoma samples NISCH promoter was significantly more methylated than in nevi (Fig. 3c, ANOVA p = 0.0254, Dunett's multiple comparisons test p = 0.0405). In the IORS patient cohort, we detected hyper-methylation of the NISCH promoter in 3/8 primary tumor samples compared to nevi, 8/10 lymph node metastases, and all 5 of other metastases samples (Fig. 3b, d). In conclusion, both microdeletions in the NISCH gene and hyper-methylation of the NISCH promoter are possible mechanisms of NISCH mRNA decrease in the lymph nodes and the distant metastatic sites in melanoma. In the primary tumor samples only hyper-methylation was present.
In addition, we looked into the NISCH gene mutation status in the TCGA and DFCI melanoma samples and observed that 3.86% of all TCGA SKCM samples (2/76 primary, 12/287 metastatic melanoma) and 5.79% DFCI samples (7/121 metastatic melanoma) had a mutation present. Majority of NISCH mutations in the TCGA SKCM dataset (for which data on the sex was available) were detected in males -two samples in the primary tumor group that were mutated were both males, and 9 out of 12 metastatic samples with the detected mutation (Fig. 3e). These sex-related differences in NISCH mutational status are important to be noted, as they may have functional implications.

Sex differences in the prognostic value of NISCH expression in melanoma
As expression of NISCH was decreased in melanoma samples compared to normal skin and nevi, we sought to determine whether NISCH levels had any association with clinical or pathological patient data. Unlike in breast and ovarian cancers where NISCH mRNA and protein expression decreased with grade [10,11], there was no difference in NISCH mRNA levels in tumors classified by Clark's level, Breslow depth, mitotic rate or cancer stage in the TCGA SKCM and GSE46517 datasets (not shown). It should be noted that these results must be interpreted with caution due to the small number of samples (102 samples in TCGA SKCM and 31 sample in GSE46517 dataset) and their uneven distribution (most samples were stage 2, and only a few were stage 4). Based on the NISCH expression level, patients were classified into two groups and association between prognosis and gene expression was examined. Using the best NISCH expression cut-off, which refers to the NISCH expression value that yields maximal difference with regard to survival between the high and low NISCH expression groups at the lowest log-rank p-value, NISCH was, surprisingly, an unfavorable prognostic marker in melanoma (HR = 2.05, 95% CI of ratio 0.8629-4.871, p = 0.0496) (Fig. 4a). As this was in contrast with so far reported prognostic value for NISCH in breast and ovarian cancers [10,14], we re-analyzed the prognostic value in males and females separately. We found that NISCH was a favorable prognostic marker in female melanoma patients (HR = 0.3116, 95% CI of ratio 0.09046-1.073, p = 0.0446), and unfavorable in male melanoma patients (HR = 3.738, 95% CI of ratio1.112-12.57, p = 0.0019) (Fig. 4c). This difference was not due to the different expression of NISCH in males and females, as the levels of NISCH mRNA were comparable (Fig. 4b). Analysis of the additional dataset GSE46517 confirmed that NISCH prognostic value differed between male and female patients (Supplemental Fig. S1). Taken together, our findings imply that NISCH may have a sex-dependent role in melanoma progression and that its role in cancer progression may be tumor-specific.

Nischarin associated gene networks
To examine the observed sex-related disparity in NISCH prognostic value in melanoma, we performed gene set enrichment analysis (GSEA) using Hallmark, KEGG and Reactome gene sets in males and females separately. The majority of signaling pathways involved in regulation of metabolism (mTORC1 signaling, gluconeogenesis, glycolysis, tricarboxylic acid (TCA) cycle, oxidative phosphorylation), cell cycle checkpoints, unfolded protein response and ABC protein transport were significantly negatively associated with increasing NISCH expression for both sexes (Fig. 5a). On the other hand, there were signatures that were inversely associated with NISCH in males and females. The most prominent differences were in the Hedgehog and WNT signaling pathways and WNT ligand biogenesis and trafficking, which were positively associated with high NISCH expression in females and negatively in males. In males, NISCH had a negative association with distinct metabolic pathways as well as cell cycle checkpoint and nucleotide excision repair (NER) (Fig. 5b, supplemental table S3). The effect of patient sex on WNT gene family in melanoma was especially notable since majority of WNT genes had strong correlation with high NISCH expression in female patients, but not in males (Fig. 5c). Our analysis implies that NISCH has a role in melanoma metabolic plasticity, but that the fine-tuning of the pathways is sex-dependent.

Association of NISCH with immune infiltration
It is well documented that anticancer immune response differs in men and women (previously reviewed [21]), and that sexassociated molecular differences in tumor mutation burden in males have an impact on immune response and survival of melanoma patients [22]. To examine the possible impact of the immune cell content on the sex-dependent NISCH role in melanoma, we downloaded Infiltration Estimation file for all TCGA SKCM primary tumors and formed four groups based on the NISCH expression and sex to match groups used in the survival analysis (Supplemental Fig. S2). Six tools that allow inter-sample comparisons of the same cell type (TIMER, CIBERSORT-ABS, EPIC, quanTIseq, xCell, and MCP-counter) come with a set of cell type signatures ranging from 6 immune cell types to 64 immune and non-immune cell types. Out of all the examined immune infiltrate cell types, males with high NISCH expression -that had the worst prognosis -had significantly higher levels of tumor infiltrating B cells than the males with low NISCH expression according to 4 out of 5 methods with this cell type listed ( Fig. 6a and Supplemental Fig. S2). Memory B cell counts, analyzed only in CIBERSORT-ABS and xCell, were higher in NISCH high males compared to NISCH low males, but also higher than in females independently of NISCH levels according to CIBER-SORT-ABS (Fig. 6b). Almost no differences in immune cell infiltration were noted between high and low NISCH female tumor samples. To examine this phenomenon in our patient cohort we have performed fluorescent immunohistochemistry on the primary tumor samples (female n = 3, male n = 7) for the B cell marker CD20 and nischarin (Fig. 6c). There were very few CD20 positive cells in female tumor samples (Fig. 6c, upper panel). In tumor samples from male patients in regions positive for NISCH expression there was evident CD20 + cell infiltration (Fig. 6c, lower panel). These findings are in line with the in silico analysis.
It has been reported that in melanoma the presence of distinct tumor-infiltrating B cell subsets may be associated with patient outcome [23]. However, functional phenotype and B cell receptor repertoire profile need to be validated by traditional tumor imaging techniques before drawing final conclusions. Taken together, our results imply that there is a sex-dependent difference in active signaling pathways and a subset of the immune infiltrate that associate with NISCH expression in melanoma.

Effects of nischarin agonist on viability of melanoma cell lines
As we found in the GSEA that NISCH was negatively associated with pathways involved in regulation of cell metabolism and cell cycle progression we hypothesized that nischarin agonists may inhibit melanoma growth. To test this hypothesis we first examined the expression of NISCH in three melanoma cell lines: A375, 518a2 and HTB140 and found that all three melanoma cell lines expressed it (Fig. 7a). Next, we treated melanoma cells with increasing concentrations of nischarin agonist rilmenidine for 72 h and measured viability by MTT assay (Fig. 7b). Rilmenidine dose-dependently reduced melanoma cell viability, and cells with BRAF V600E mutation -A375 and 518a2 -had lower IC50 values for rilmenidine compared to BRAF WT cells HTB140 (52 μM and 66 μM, and 148 μM, respectively). This is in line with greater metabolic activity of BRAF-mutated cells.

Discussion
Immune checkpoint inhibitor therapy and targeted BRAF and MEK inhibitors have revolutionized melanoma treatment in the past decade, but only a subset of patients have complete and durable response to therapy. In search for new therapeutic targets, we examined the expression, regulation and potential role of nischarin in progression of melanoma. NISCH was described as a tumor-suppressive protein for its role in regulation of the cell migration and invasiveness [11,15]. NISCH expression has been shown to decrease with progression in breast and ovarian cancer [10,11,24] leading to increased cell migration, tumor growth, and cell survival [11]. Restoring NISCH levels in breast cancer cells reduced tumor growth in vivo and cell migration in vitro by inhibiting cancer cells ability to attach to the extracellular matrix [15], decreased FAK phosphorylation and prevented ERK activation [14,25]. Since ERK activation is one of the main mechanisms of resistance to BRAF inhibitors [26], we hypothesized that NISCH may be a good candidate to investigate in metastatic melanoma, given that there are several approved agonists for this receptor. Fig. 3 Regulation of nischarin gene expression. (a) Somatic copynumber alteration (CNA) analysis of metastatic melanoma samples from TCGA SKCM and DFCI datasets analyzed by one-way ANOVA with Dunnett's multiple comparisons test. * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001 (b) Levels of NISCH expression in IORS patient samples with indicated presence of microdeletions and NISCH promoter hyper-methylation. (c) NISCH promoter methylation status in TCGA SKCM and GSE86355 datasets (d) and in IORS cohort. All data were analyzed using one-way ANOVA with Dunnett's multiple comparisons test. * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001 (e) Overview of primary and metastatic melanoma samples from TCGA SKCM dataset with detected NISCH mutation ◂ In this study, we examined NISCH expression in publicly available datasets and samples collected at our institution -in normal human skin, nevi, primary and metastatic tumor samples -and found that it is expressed at both mRNA and protein level, and that its expression is decreased in tumors compared to skin and nevi. Surprisingly, unlike in breast and ovarian cancers, where NISCH mRNA and protein expression decreased with grade [10,11], we did not find any association of NISCH mRNA levels in melanoma tumors with clinical or pathological patient data. Opposite of the dominantly cytoplasmic and membranous localization that was reported in breast cancer cells, NISCH was also detectable around and in the nucleus in melanoma cells, similarly to its localization in neurons [27], cells that share common embryonic origin with melanocytes [28]. Nuclear localization of NISCH has not been previously described in cancer, and may imply difference in function in epithelial-and neuroectodermal-derived cells, the aspect that needs further study.
Regulation of NISCH expression in melanoma was in line with the previously reported mechanisms, indicating presence of microdeletions and hyper-methylation of the NISCH promoter in the tumor tissue. NISCH promoter hyper-methylation was reported in 30% of breast and 36.7% of ovarian cancer tissues [10,14]. In lung cancer, NISCH gene was also hyper-methylated and inactivated [29]. In our study, NISCH promoter hyper-methylation was detected in several primary melanoma samples, but its impact on NISCH levels was the most pronounced in metastatic melanoma samples. Both IORS experimental cohort and analysis of the GSE86355 data indicated that in majority of metastatic melanoma samples NISCH promoter was hypermethylated compared to nevi (13/15 and 20/28, respectively), indicating that this is a common occurrence in melanoma. This observation was also in concordance with the study that examined aberrant promoter methylation in a panel of neural crest-derived tumors, where NISCH promoter was hyper-methylated in 80% of melanoma samples [30]. Copy number alteration was the other proposed mechanism of NISCH expression regulation in cancer [11]. In our study, microdeletions were present in metastatic samples with evidently lower NISCH levels, implying that in metastatic melanoma this may be an important mechanism of NISCH downregulation. Microdeletions and hypermethylation of NISCH promoter are not mutually exclusive events since both could be seen in the same patient in IORS cohort. Co-occurrence of different mechanisms of NISCH inactivation was also observed in TCGA SKCM dataset (not shown), but also in ovarian cancer where both NISCH promoter hyper-methylation and LOH occurred in the same tumor samples [10]. Knowing that malignant melanoma carries one of the highest mutational burdens [31], it is Fig. 4 The effect of nischarin mRNA expression on melanoma patients' overall survival. (a) TCGA SKCM primary melanoma samples were used for Kaplan-Meier overall survival analysis based on the best NISCH expression cut-off level (11.8 FPKM) according to the Human Protein Atlas. (b) These samples were further divided based on patient sex and difference between NISCH mRNA levels in females and males was analyzed using t-test. * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001 (c) Additional survival analysis was performed for newly formed groups with the best NISCH expression cut-off levels for females (8.23 FPKM) and males (12.7 FPKM) (c) Correlation analysis between NISCH and WNT genes based on GSEA ranking score from previous analysis of the TCGA SKCM dataset (Pearson correlation). Red color refers to stronger positive correlation, and blue color indicates negative correlation ◂ not surprising that NISCH mutations were detected in both primary and metastatic melanoma samples in in silico analyzed datasets. The majority of NISCH mutations in TCGA SKCM dataset were detected in males. This was not unexpected as male melanoma patients have greater mutational burden overall [32], but functional implications of NISCH mutations need further examination.
Surprisingly, NISCH appeared to be an unfavorable prognostic marker in melanoma, unlike in ovarian cancer where patients with higher NISCH expression had better overall survival [10] and in breast cancer where patients with higher NISCH expression had increased recurrence-free survival [14]. As this was an unexpected finding, and bearing in mind that breast and ovarian cancers were examined in females, we analyzed the effects of NISCH mRNA expression level in primary melanoma in females and males separately. Strikingly, in females NISCH was a favorable prognostic marker, while in males it was unfavorable. Sex differences in cancer driver genes and sex-influenced biomarkers of cancer patient outcome are present across a broad range of tumor types [33], and our findings indicate that tumor-suppressive role that has been reported for NISCH may not be universal for female and male patients. In melanoma, women have up to 30% advantage in all aspects of the disease progression, most likely due to underlying biological sex difference in tumor-host interactions [34]. Female melanoma patients also have lower incidence and better survival compared to males (that persists after adjustment for various confounding variables such as stage, body site, age and metastatic status) [35,36]. It is not completely understood why female melanoma patients have better prognosis (reviewed in [37]). Men are more likely to be diagnosed later, as women more often report for medical visits. There are also important differences in pharmacokinetics (drug absorption, distribution, metabolism and elimination), immune response and hormone status between the sexes that influence the progression and response to therapy. Estrogen has a protective role in the females, and the difference between sexes decreases with age when estrogen levels decrease in the postmenopausal period.
We did not observe the difference in the levels of NISCH expression in males and females in publicly available datasets or in our cohort, and we hypothesize that tumor-host interactions influence NISCH prognostic role. Indeed, in NISCH loss of function mutant mice (with leucine-rich repeat (LRR) domain deletion that prevents NISCH interaction with AMPK [16]), male and female mice showed divergent glucose metabolism and sexual dimorphism in insulin sensitivity [38].
Since NISCH has been predominantly investigated in female cancer patients, and its effects in males are unknown, we performed GSEA in male and female melanoma samples separately. Our analysis revealed that regardless of sex NISCH expression had significant correlation with many metabolic pathways. This is not surprising since it has been shown that NISCH plays an important physiological role in metabolic homeostasis [39]. Groups of genes involved in mTORC1 signaling, regulation of glucose metabolism and TCA cycle had a negative correlation with NISCH both in Fig. 6 Association of nischarin expression with tumor immune profiles in melanoma samples. Differences in levels of tumor infiltrating (a) B cells (according to EPIC, quanTIseq, xCell, and MCPcounter) and (b) memory B cell (CIBERSORT-ABS and xCell) in primary melanoma TCGA SKCM samples divided into "high" and "low NISCH" groups for both sexes were examined using one-way ANOVA with Tukey's multiple comparisons test. * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001. (c) Representative images of female (upper panel) and male melanoma primary tumor tissue (lower panel) stained for CD20 B cell marker (red) and NISCH (green). Nuclei stained with DAPI, scale bar 100 μm females and males. Mammalian target of rapamycin complex 1 (mTORC1) serves as a central regulator of cell metabolism, growth, proliferation, invasion, and survival [40], and has been found to be almost universally activated in melanoma [41]. Regulation of mTORC1 at the signaling level includes its suppression by LKB1 through activation of AMPK [42]. Previous study showed that interaction between LKB1 and NISCH inhibits cell migration and breast tumor progression [43], but more recently NISCH connection to LKB1-AMPK pathway was established in mutant mouse model with NISCH deletion. NISCH mutant mice showed significantly decreased activity of AMPK [38], suggesting that NISCH is important regulator of AMPK-mTOR signaling pathway. Many metabolic processes differ in males and females, with great contribution from the X-linked metabolic genes (reviewed in [44]), and NISCH association with regulation of cell metabolism may have different effects in male and female patients. Of importance, there were distinct signaling pathways with opposite association dependent on the patient sex. NER and cell cycle checkpoint regulation were negatively associated with higher NISCH expression in males, but not in females. As NER is responsible for repair of UV-induced DNA induced damage this may contribute to the higher mutation burden frequently observed in male melanoma patients [45]. In females, but not in males NISCH was positively associated with WNT signaling, especially with WNT8B, WNT6, WNT3A, and WNT8D (Fig. 5c). Although it has been previously reported that WNT6 and WNT8 both play a critical role in the neural crest induction during embryonic development [46], their role in melanoma remains unclear. On the other hand, elevated nuclear beta catenin and activation of the WNT/beta catenin signaling by WNT3A in melanoma was associated with reduced proliferation and improved survival [47] and BRAF inhibitor-induced apoptosis was mediated, in part, through WNT3A [48,49]. This could be one possible explanation for positive association of high NISCH expression with better survival in female patients.
Elevated WNT/beta catenin signaling has also been reported to prevent anti-tumor immunity in melanoma [50] and decreased T cell infiltration [51]. The only significant difference between NISCH high and NISCH low tumors in males based on our in silico analysis of the tumor immune infiltrate was in the increased presence of memory B cells in NISCH high tumors. We did not observe any significant differences in immune infiltrate in female patient tumors based on NISCH expression. Although our patient cohort had low sample number (n = 10 primary tumors) we examined the presence of CD20 + B cells in the tumor tissue, and found that indeed in NISCH positive regions in male tumors there was CD20 + infiltrate, while in females we could detect very few CD20 + cells. This result needs further validation on a larger patient cohort by immunohistochemistry and analysis of the functional type of the B cell infiltrate, which was beyond the scope of the present study.
Ultimately, pathways negatively associated with NISCH expression control melanoma growth and progression, and NISCH agonization may be a novel therapeutic approach. We found that NISCH agonist rilmenidine significantly decreases melanoma cell viability in vitro. These results lay a ground for further investigation of the potential of NISCH agonists for repurposing in melanoma treatment.
In conclusion, our study brings three important and novel findings that open new avenues for investigation of the nischarin role in cancer. First, we report a previously unreported localization of nischarin in the cell nucleus in melanoma tissues. Second, we highlight the role of sex in melanoma biology and its impact on nischarin role in melanoma progression, aspect not previously addressed in mechanistic studies of nischarin in cancer. This finding questions the tumor suppressive role of nischarin and warrants examination of its prognostic role in other cancer types since NISCH was previously investigated only in female cancers. Third, we report that NISCH agonist rilmenidine decreases melanoma cell viability. Limitations of this study include a relatively small cohort of patient samples and the fact that it is was performed as a single-center study, thus validation of the obtained results in a multi-center setting is needed.
Consent to publish Not applicable.