Abstract

   Skin cutaneous melanoma (SKCM) is a highly malignant form of skin
   cancer, known for its unfavorable prognosis and elevated mortality
   rate. RARRES1, a gene responsive to retinoic acid receptors, displays
   varied functions in various cancer types. However, the specific role
   and underlying mechanisms of RARRES1 in SKCM are still unclear.
   [44]GSE15605 was utilized to analyze the expression of RARRES1 in SKCM.
   Subsequently, the TCGA and GEO databases were employed to investigate
   the relationships between RARRES1 and clinicopathological parameters,
   as well as the prognostic implications and diagnostic efficacy of
   RARRES1 in SKCM. GO, KEGG, and GSEA analyses were conducted to explore
   the potential functions of RARRES1. Furthermore, the associations
   between RARRES1 and immune infiltration were examined. Genomic
   alterations and promoter methylation levels of RARRES1 in SKCM were
   assessed using cBioPortal, UALCAN, and the GEO database. Finally,
   RARRES1 expression in SKCM was validated through immunohistochemistry,
   and its functional role in SKCM progression was elucidated via in vivo
   and in vitro experiments. We found that RARRES1 was downregulated in
   SKCM compared with normal tissues, and this low expression was
   associated with worse clinicopathological features and poor prognosis
   of SKCM. The diagnostic efficacy of RARRES1, as determined by ROC
   analysis, was 0.732. Through GO, KEGG, and GSEA enrichment analysis, we
   identified 30 correlated genes and pathways that were mainly enriched
   in the tumor immune microenvironment, proliferation, apoptosis, and
   autophagy. Additionally, RARRES1 expression was found to be positively
   related to the infiltration of various immune cells in SKCM,
   particularly macrophages and T helper cells, among others. Analysis of
   genomic alterations and promoter methylation revealed that shallow
   deletion and hypermethylation of the RARRES1 promoter could lead to
   reduced RARRES1 expression. IHC validation confirmed the downregulation
   of RARRES1 in SKCM. Moreover, overexpression of RARRES1 inhibited the
   proliferation and migration of A375 cells, promoted apoptosis, and
   inhibited autophagic flux. In the mouse xenograft model, RARRES1
   overexpression also suppressed SKCM tumor growth. Collectively, these
   findings suggest that RARRES1 may function as a suppressor and could
   potentially serve as a prognostic biomarker and therapeutic target for
   SKCM.

   Keywords: Skin cutaneous melanoma, RARRES1, Tumor suppressor, Autophagy

   Subject terms: Melanoma, Cell biology, Cancer, Skin cancer, Tumour
   biomarkers

Introduction

   Skin cutaneous melanoma (SKCM) originates from the malignant conversion
   of melanocytes and is considered the most fatal type of skin cancer.
   SKCM accounts for approximately 55,500 deaths each year, making up 0.7%
   of all cancer-related mortalities^[45]1,[46]2. In recent decades, with
   the development of immunotherapy and targeted therapies, the mortality
   rate of SKCM has decreased^[47]3. However, the prognosis for patients,
   especially those with metastases, remains formidable, with an
   approximate 5-year survival rate of only 27%^[48]2. Moreover, the
   incidence of malignant melanoma continues to rise in the world wide,
   with the overall incidence of melanoma in the United States
   experiencing an annual increase of approximately 1.4% over the past
   decade^[49]3,[50]4. Therefore, it is crucial to identify improved
   prognostic biomarkers and therapeutic targets for SKCM.

   Retinoic acid receptor responder 1 (RARRES1), also known as
   tazarotene-induced gene TIG1, is a retinoid responsive gene that was
   initially discovered in psoriatic skin raft cultures following retinoid
   treatment^[51]5. Research has shown that RARRES1 demonstrates
   anti-tumor effects in various cancer types. Additionally,
   hypermethylation of the RARRES1 promoter leads to reduced expression in
   liver cancer, prostate cancer, breast cancer, and colorectal
   cancer^[52]6–[53]9. Notably, RARRES1 plays diverse roles in different
   tissues. Shyu et al. demonstrated that RARRES1 inhibited cell invasion,
   migration, and epithelial-mesenchymal transition in testicular cancer
   cell^[54]10. They also elucidated the role of RARRES1 in inducing
   autophagy activity in cervical cancer cells^[55]7. In addition,
   aberrant RARRES1 expression can induce activation of M1 microphage in
   kidney renal clear cell carcinoma and inhibit cells proliferation in
   colorectal cancer respectively^[56]9,[57]11. Nevertheless, the
   expression and functional role of RARRES1 in SKCM remain unclear and
   need further investigation.

   In this study, we comprehensively evaluated RARRES1 expression and its
   prognostic value in SKCM using data from GEO and TCGA database.
   Besides, mutation, promoter methylation and potential functions of
   RARRES1 in SKCM was investigated through bioinformatic analyses.
   Finally, we confirmed RARRES1 expression in SKCM tissues and
   experimentally explored its biological roles in cell proliferation,
   migration, apoptosis and autophagy.

Material and methods

Expression of RARRES1 in SKCM and correlation analysis of clinical
characteristics

   The [58]GSE15605 dataset was downloaded from the Gene Expression
   Omnibus (GEO, [59]https://www.ncbi.nlm.nih.gov/geo/) database. Using
   the GEO2R tool, we filtered the expression data of RARRES1 in SKCM and
   normal samples. The RNAseq data for SKCM was acquired from the TCGA
   database ([60]https://portal.gdc.cancer.gov) ^[61]12. We extracted data
   in TPM format along with clinical information, excluding samples that
   were normal or lacked clinical information. Subsequently, comprehensive
   data for 470 SKCM patients was obtained. The entire data processing was
   carried out using R (version 4.2.1). For data analysis, we employed the
   stats package and car package, customizing our approach to suit the
   unique characteristics of the data format. Finally, the data was
   visualized using the ggplot2 package.

Prognostic analysis

   Patient information and prognosis data of SKCM was obtained from
   TCGA^[62]12. Kaplan–Meier survival curves were constructed to compare
   the prognosis between the high and low RARRES1 groups. Additionally,
   univariate and multivariate Cox regression analyses were employed to
   assess the risk factors associated with SKCM. The survival package and
   survminer package in R (4.2.1) was used for data processing, the
   ggplot2 package was used for data visualization. The nomogram was
   constructed and visualized using the R package rms. For the ROC
   analysis, the expression data of RARRES1 in the [63]GSE15605 dataset
   was processed by Graphpad prism.

Correlated genes enrichment analysis

   The mRNA expression data of 470 SKCM patients in TCGA database was
   obtained. Pearson correlation was used to identify genes associated
   with RARRES1 expression. The results were filtered, selecting the top
   15 genes with the highest positive correlation and the top 15 genes
   with the highest negative correlation. R package ggplot2 was used to
   create a co-expression heatmap. Then, we selected 900 genes which were
   significantly correlated with RARRES1 (P < 0.05, |R|> 0.5) to perform
   GO and KEGG analysis (Table [64]S1). GO analysis is a widely accepted
   approach for performing comprehensive functional enrichment studies on
   a large scale, encompassing biological processes (BP), molecular
   functions (MF), and cellular components (CC)^[65]13. KEGG serves as a
   highly utilized database for archiving data related to genomes,
   biological pathways, diseases, and pharmaceuticals^[66]14. GO
   annotation analysis and KEGG pathway enrichment analysis of correlated
   genes were performed using R package clusterProfiler. P < 0.05 was
   considered statistically significant. Results of enrichment analysis
   were visualized using ggplot2, igraph, and ggraph.

GSEA analysis

   Gene Set Enrichment Analysis (GSEA) is a powerful analytical technique
   designed to assess whether a predefined group of genes exhibits
   statistically significant and coherent variations between two distinct
   phenotypes^[67]15. The RNAseq data of 470 SKCM patients was analyzed by
   DESeq2 and edgeR. After obtaining gene names and their corresponding
   log fold changes (logFC), we performed GSEA using the clusterProfiler.
   The objective was to unveil substantial disparities in functional and
   pathway profiles between the high- and low-RARRES1 groups.
   h.all.v2022.1.Hs.entrez.gmt [Hallmarks] in MSigDB Collections
   ([68]https://www.gsea-msigdb.org/gsea/msigdb/collections.jsp) was used
   as the reference gene collection. An adjusted P < 0.05 and False
   discovery rate (FDR) < 0.25 were considered as significant enrichment.

Correlation analysis between RARRES1 expression and immune infiltration

   For the information of 470 SKCM patients in TCGA database, immune
   infiltration analysis was performed using the ssGSEA algorithm based on
   that provided in the R package GSVA [1.46.0], using markers of 24
   immune cells to calculate the immune infiltration corresponding to the
   cloud-based data^[69]16,[70]17. The data were processed using R (4.2.1)
   and correlation analysis was performed between the primary variables
   and immune infiltration matrix data, and the results were displayed as
   lollipop plots and analyzed through Spearman correlation analysis. Data
   was visualized using ggplot2.

Analysis of gene mutations and promoter methylation analysis

   cBioPortal database ([71]http://cbioportal.org) was used to analyze the
   profile of genetic alterations in SKCM, amino acid changes in proteins,
   and the effect of different copy number variants (CNV) on RARRES1 mRNA
   expression^[72]18. Furthermore, UALCAN database
   ([73]https://ualcan.path.uab.edu/) was utilized to assess RARRES1
   promoter methylation in SKCM^[74]19. Finally, we downloaded
   [75]GSE120878 dataset from the GEO database. This dataset consists of
   DNA methylation data for 89 SKCM and 73 nevi. Furthermore, EWAS
   database ([76]https://ngdc.cncb.ac.cn/ewas/) was used to retrieve
   methylation sites specifically associated with RARRES1 in SKCM.

Clinical samples

   Paraffin-embedded skin samples were taken from 20 patients (11 male and
   9 female, aged 57–74 years) diagnosed with SKCM and 11 healthy controls
   (6 male and 5 female, aged 60–78 years) diagnosed with intradermal
   nevus. These samples were sourced from the tissue bank of the First
   Affiliated Hospital of Xi’an Jiaotong University. All diagnoses were
   confirmed by two experienced pathologists. Written informed consent was
   diligently obtained from all patients before sample collection and
   analysis. Ethics approval for the study was acquired from the
   Institutional Ethics Committee of Xi’an Jiaotong University.

Immunohistochemistry

   Immunohistochemistry was performed according to a standard procedure.
   The RARRES1 antibody (ab198908, Abcam) and Ki-67 antibody
   ([77]GB111141, Servicebo, China) was used. RARRES1 sections were
   semi-quantitatively categorized into four grades: negative (-), weak
   ( +), moderate (+ +), and strong (+ + +). The results were
   independently evaluated by two pathologists. Ki67 sections were scanned
   by Hamamatsu digital pathology. Image J software was used to quantify
   the Ki67 staining.

Cell culture

   The A375 cells were purchased from ATCC and cultured in Dulbecco’s
   modified Eagle’s medium containing with 10% fetal bovine serum,
   penicillin (100 U/mL), and streptomycin (100 μg/mL) at 37℃ in a
   humidified atmosphere with 5% CO[2].

Plasmid construction and transient transfection

   The RARRES1 expression plasmid was synthesized by MiaoLing Plasmid
   (Wuhan, China). A375 cells were transfected with 2 μg of the RARRES1
   expression plasmid (pCMV-RARRES1(human)-3×FLAG) or the control pCDNA
   vector using PEI for 48 h. The transfection procedures were conducted
   according to the standard protocols.

Lentivirus production and construction of stable RARRES1-overexpressing A375
cell line

   Full length cDNA encoding human RARRES1 was inserted into
   pLV3-CMV-RARRES1(human)-3 × FLAG-CopGFP-Puro to obtain
   RARRES1-overexpressing plasmid pCDH-RARRES1. pCDH-RARRES1 lentivirus
   particles were obtained by co-transfecting pCDH-RARRES1 with packaging
   plasmids (pLP1, pLP2 and pLP-VSVG) into HEK293T cells. A375 cells were
   infected with concentrated pCDH-RARRES1 lentivirus particles and then
   were screened by puromycin to obtain stable RARRES1-overexpressed A375
   cell line.

Cell proliferation assay

   The proliferation rates of A375 cells transfected with pcDNA3.4-Vector
   or pCDNA3.4-RARRES1 were assessed by MTT reagent
   (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide). These
   assays were performed according to the manufacturer’s instructions.

Crystal violet staining

   A total of 500 cells, stably expressing RARRES1 or Vector, were seeded
   into 6 wells plates and cultured for 2 weeks. Colonies were fixed and
   stained by crystal violet according to the standard protocols.

Soft agar clonogenic assay

   The cells were trypsinised, collected after puromycin screening, and
   counted using a haemocytometer. Furthermore, 200 cells/well were seeded
   in 0.7% soft agar in media on a 12-well plate, over a 1.2% soft agar
   underlay. Finally, 1 mL normal medium with was added to cover the top
   agar. They were grown for about 10–14 days and cell clone images were
   captured by a Nexcope NIB620 inverted microscope (Nexcope, Ningbo,
   China).

Cell cycle assay

   After transfected by pcDNA3.4-RARRES1/Vector for 48 h, cells were
   collected and washed with PBS, then cells were fixed with 70% ethanol
   at − 20 °C overnight. 100 μL RNase A solution was added and cells were
   incubated at 37 °C for 30 min. Finally, 400 μL propidium iodide (PI)
   solution was added and incubated at room temperature for 30 min in the
   dark. The DNA content was detected using flow cytometry (NovoCyte flow
   cytometry, ACEA Biosciences).

Transwell assay

   Transwell assays were conducted using Transwell chambers (3422;
   Corning). The lower chamber was filled with 800 μL of complete culture.
   In the upper chamber, 3 × 10^4 cells were seeded using 200 μL of fetal
   bovine serum-free medium. After a 24-h incubation, the migrating cells
   were fixed and then stained with 0.1% aqueous crystal violet.
   Microscopic images were captured, and the cell count in each field was
   documented.

Apoptosis assays

   Cells transfected by pcDNA3.4-RARRES1/Vector, were stained with FITC
   Annexin V Apoptosis Detection Kit according to the manufacturer’s
   instructions. Cells apoptosis was analyzed by flow cytometry (NovoCyte
   flow cytometry, ACEA Biosciences).

Protein extraction and Western blotting

   Western blotting was performed as described previously^[78]20. The
   following antibodies were used in this study: RARRES1 (MA5-26,247,
   Invitrogen), Flag (66,008-4-Ig, proteintech), P53(CST# #9282), BAX
   (50,599-2-Ig, proteintech), BAK (29,552-1-AP, proteintech), BCL-2
   (12,789-1-AP, proteintech), Cyclin D1 (ab16663), Cyclin E1
   (11,554-1-AP, proteintech, China), P62 (CST#23,214), LC3I/II
   (CST#4108), β-actin (CST#3700), GAPDH (CST#5174), β-Tubulin
   ([79]M30109, Abmart, China). The membranes were incubated with HRP
   conjugated secondary antibody at room temperature for 1 h. Finally, the
   chemiluminescent signal was detected using the ultrasensitive ECL kit
   (Pierce Chemical, Rockford, IL, USA). The intensity of the bands was
   quantified by using Image J software (National Institutes of Health,
   Bethesda, MD, USA).

ROS determination

   ROS levels were measured with DHE (Biyotime, #S0063) dyes. A375 cells
   were first seeded into 12-well plates overnight and transfected with
   pcDNA3.4-Vector or pCDNA3.4-RARRES1 for 48 h. Then, cells were
   incubated with DHE solution dissolved in FBS-free medium for 45 min and
   then washed with PBS. Nexcope NIB620 inverted microscope (Nexcope,
   Ningbo, China) was used to take the fluorescent images. Image J
   software was usd to analyze DHE relative fluorescence intensity. (bar
   length = 50 um).

LysoSensor green DND-189 staining

   Lysosomal staining with LysoSensor Green DND-189, was performed
   according to the manufacturer (Yeasen 40767ES50, Shanghai, China).
   After transfection of pCDNA3.4, Lysosensor green was added to the
   medium and incubated for 10 min at 37℃, with 5% CO[2.] Then cells were
   directly transferred to a fluorescent microscope for imaging.

Xenograft tumor model

   A total of 4 × 10^6 cells transfected with pCDH-RARRES1/Vector were
   subcutaneously injected into4-week-old female BALB/c-nude mice on their
   right flanks (6 mice/group). Tumor volume was calculated every 3 days.
   On day 10, the mice were sacrificed and subcutaneous tumors were
   collected. Tumor volumes were calculated using the formula
   V = 1/2 × L × W^2, where "L" represents tumor length and "W" represents
   tumor width. All animal experiment were approved by the Institutional
   Animal Care and Use Committee of Xi’an Jiaotong University.

Statistical analysis

   T-test was used to analyze the differential expression genes between
   SKCM and normal. Kruskal–Wallis Test, Welch t' test and Wilcoxon rank
   sum test were used to assess the differences in RARRES1 expression
   among samples with varying clinical features. Logrank test was used to
   compare the prognosis between the high and low expression groups of
   RARRES1. Sperman’s correlation was carried out to evaluate enrichment
   scores of 20 immune cells and RARRES1 expression separately.
   Experimental data were expressed as the mean ± SEM. Statistical graphs
   and analyses were conducted using GraphPad Prism software. To compare
   two unpaired groups, Student's t-test was employed, and P < 0.05 was
   deemed as statistically significant.

Ethics approval and consent to participate

   Ethics approval for the study was acquired from the Institutional
   Ethics Committee of Xi’an Jiaotong University, and the use of animals
   in our experiments was consistent with ethical requirements.

Results

Expression, clinicopathological correlation, prognostic and diagnostic
potential of RARRES1 in SKCM

   The expression of RARRES1 in SKCM (n = 58) and normal tissues (n = 16)
   was analyzed using [80]GSE15065 dataset. As shown in Fig. [81]1A,
   RARRES1 expression in SKCM was significantly down-regulated compared to
   the normal tissues. Subsequently, TCGA database was used to investigate
   the correlation between RARRES1 expression and clinical characteristics
   of SKCM. We found that RARRES1 was significantly reduced in SKCM
   patients of T4 stage, pathological stage II, age over 60, with melanoma
   ulcer, higher Clark level, higher Breslow thickness and without
   radiotherapy (Fig. [82]1B-H), indicating decreased RARRES1 expression
   was associated with poor clinical features of SKCM. We then assessed
   the correlation between RARRES1 expression and survival of SKCM
   patients. The Kaplan–Meier survival analysis indicated that low
   expression of RARRES1 was associated with unfavorable overall survival
   (OS) or disease specific survival (DSS) (Fig. [83]1I–J). Besides,
   RARRES1 expression significantly affected the OS of SKCM patients
   undergoing immunotherapy, showing an overall correlation with poor OS
   after immunotherapy (HR = 0.6 (0.43–0.83), logrank P = 0. 0021 for all
   anti-PD-1 treatment; HR = 0.049 (0.29–0.83), logrank P = 0.0068 for
   anti-CTLA-4 treatment) (Fig. [84]1K-L). We futher evaluated the factors
   correlated with OS in SKCM, as shown in Table [85]1. Univariate Cox
   regression analysis showed that tumor stage (T, N, M), patient age,
   ulceration, Clark level, Breslow depth, and RARRES1 expression were
   associated with survival. Multivariate Cox regression analysis
   demonstrated that only pathological N stage, Breslow depth and RARRES1
   are independent prognostic factors for survival. For further
   verification, we constructed a nomogram of OS that incorporated
   RARRES1, pathological N stage and Breslow depth, enabling us to predict
   the 1-, 3-, 5-year survival probability in SKCM (Fig. [86]1M). Based on
   the expression data of RARRES1 in [87]GSE15605, the diagnostic value of
   RARRES1 in SKCM was calculated, the area under the curve (AUC) was
   0.732 (P = 0.0048), which illustrated that RARRES1 had a good
   diagnostic performance in SKCM^[88]21 (Fig. [89]1N).

Figure 1.

   [90]Figure 1
   [91]Open in a new tab

   Expression, clinicopathological correlation, prognostic and diagnostic
   potential of RARRES1 in SKCM. (A) RARRES1 expression in 58 SKCM tissues
   and 16 normal tissues in the [92]GSE15605 dataset. (B–H) The
   association of RARRES1 expression and (B) T classification, (C)
   pathological stages, (D) patient’s age, (E) melanoma ulceration, (F)
   melanoma Clark level, (G) Breslow depth, (H) radiation therapy. (I–J)
   Survival curves of OS and DSS. (K–L) The prognostic impact of RARRES1
   on immunotherapy with anti-PD-1 treatment and anti-CLTA-4 treatment.
   (M) Nomogram that integrates RARRES1 expression and other prognostic
   factors in SKCM. (N) ROC curve for evaluating the diagnostic value of
   RARRES1. **P < 0.01 and ***P < 0.001.

Table 1.

   The univariate and multivariate analyses of overall survival.
   Characteristics Number (n) Univariate analysis Multivariate analysis
   HR (95% CI) P-value HR (95% CI) P-value
   T stage (T1-2 vs. T3-4) 362 (119 vs. 243) 2.099 (1.511–2.914)  < 0.001
   0.857 (0.484–1.517) 0.596
   N stage (N0 vs. N1-3) 403 (225 vs. 178) 1.760 (1.310–2.365)  < 0.001
   2.432 (1.623–3.646)  < 0.001
   M stage (M0 vs. M1) 431 (407 vs. 24) 1.902 (1.032–3.506) 0.039 2.059
   (0.822–5.157) 0.123
   Age (< = 60 vs. > 60) 457 (247 vs. 210) 1.663 (1.256–2.201)  < 0.001
   1.197 (0.808–1.772) 0.370
   Melanoma ulceration (No vs. Yes) 314 (147 vs. 167) 2.099 (1.506–2.927)
    < 0.001 1.431 (0.949–2.160) 0.088
   Clark level (I&II&III vs. IV&V) 316 (96 vs. 220) 2.184 (1.520–3.137)
    < 0.001 1.307 (0.801–2.133) 0.284
   Breslow depth (< = 3 vs. > 3) 356 (181 vs. 175) 2.665 (1.948–3.644)
    < 0.001 1.922 (1.142–3.238) 0.014
   RARRES1 (High vs. Low) 457 (229 vs. 228) 1.633 (1.247–2.139)  < 0.001
   1.603 (1.095–2.347) 0.015
   [93]Open in a new tab

Gene correlation and functional enrichment analysis of RARRES1 expression in
SKCM

   We proceeded to employ bioinformatics analysis in order to examine the
   biological function of RARRES1. Initially, we conducted correlation
   analysis between RARRES1 and other genes in the TCGA database.
   Subsequently, we assigned ranks to genes that exhibited a significant
   correlation with RARRES1 expression (P < 0.05). Consequently, we
   identified the top 15 genes that displayed a positive correlation and
   the top 15 genes that displayed a negative correlation with RARRES1, as
   visually depicted in the heatmap (Fig. [94]2A). The GO terms were
   classified into BP, CC and MF. Within the BP category, the enriched
   terms included leukocyte cell–cell adhesion, regulation of T cell
   activation, and positive regulation of leukocyte. In the CC category,
   the enriched terms included external side of the plasma membrane, T
   cell receptor complex, and plasma membrane signaling receptor complex.
   Lastly, in the MF category, the enriched terms included immune receptor
   activity, cytokine receptor activity, and MHC protein complex binding.
   In terms of KEGG pathway enrichment, significant enrichments were
   observed in hematopoietic cell lineage, cell adhesion molecules, and
   cytokines-cytokine receptor interaction (Fig. [95]2B-C). We also
   performed a GSEA analysis to identify the key pathways related to
   RARRES1. GSEA analysis found that 480 data sets satisfied the criteria
   of an FDR < 0.25 and a P < 0.05, as shown in Table [96]S2. The enriched
   pathways encompassed various processes, including mitotic G1 Phase and
   G1/S transition (Fig. [97]2D), caspase cascade (Fig. [98]2E), oxidative
   damage response (Fig. [99]2F), autophagy (Fig. [100]2G), cancer
   immunotherapy by PD1 blockade (Fig. [101]2H), cancer immunotherapy by
   CTLA4 blockade (Fig. [102]2I), modulators of Tcr signaling and T cell
   activation (Fig. [103]2J), natural cell mediated cytotoxicity
   (Fig. [104]2K) and so on.

Figure 2.

   [105]Figure 2
   [106]Open in a new tab

   Gene correlation and functional enrichment analysis of RARRES1
   expression in SKCM. (A) Top 30 genes correlated with RARRES1
   expression. (B) The GO and KEGG enrichment analysis of RARRES1. (C) The
   network results of GO and KEGG pathways. (D-K) GSEA analysis on (D)
   mitotic G1 phase and G1/S transition, (E) caspase cascade, (F)
   oxidative damage response, (G) autophagy, (H) cancer immunotherapy by
   PD1 blockade, (I) cancer immunotherapy by CTLA4 blockade, (J)
   modulators of TCR signaling and T cell activation, (K) natural killer
   cell mediated cytotoxicity.

Correlation between RARRES1 expression and immune cell infiltration in SKCM

   Functional enrichment analysis of RARRES1 revealed that numerous
   pathways were closely associated with immune processes. Furthermore,
   tumor-associated immune cells play important roles in determining tumor
   progression, and the infiltration of immune cells is strongly related
   to the outcome of the tumor^[107]22. Therefore, we assessed the
   correlation between the RARRES1 expression and the extent of immune
   cell infiltration in SKCM. As shown in Fig. [108]3A, RARRES1 expression
   in SKCM was positively correlated with the infiltration of various
   immune cells, indicating the low expression of RARRES1 might be
   unfavorable for immune cell infiltration in SKCM. We further compare
   the difference in immune cell infiltration between RARRES1 high and low
   expression groups. The results demonstrated that SKCM with low
   expression of RARRES1 had less immune cell infiltration
   (Fig. [109]3B-U).

Figure 3.

   [110]Figure 3
   [111]Open in a new tab

   RARRES1-related immune infiltration in SKCM. (A) Lollipop chart of the
   correlation between RARRES1 expression and 24 types of immune cells.
   (B–U) Scatter plot of GPR143 expression and immune cell infiltration:
   (B) activated dendritic cells, (C) B cells, (D) CD8 + T cells, (E)
   cytotoxic cells, (F) dendritic cells, (G) eosinophils, (H) immature
   dendritic cells, (I) Macrophages, (J) neutrophils, (K) NK CD56dim
   cells, (L) NK cells, (M) plasmacytoid dendritic cells, (N) T cells, (O)
   T helper cells, (P) T central memory cells, (Q) T effector memory
   cells, (R) T follicular helper cells, (S) T gamma delta cells, (T) Th1
   cells, (U) Th2 cells. *P < 0.05, ***P < 0.001, ns: no significance.

Mutation and promoter methylation of RARRES1 in SKCM

   Previous studies have demonstrated that the abnormal expression of
   RARRES1 arises due to hypermethylation of the CpG island associated
   with the RARRES1 promoter^[112]23. Thus, we analyzed the genomic
   alteration and promoter methylation of RARRES1 via cBioPortal, UALCAN
   and GEO database. In a total of 443 SKCM samples, the mutation rate of
   RARRES1 was found to be 1.8% (Fig. [113]4A). This implies that there
   were 8 instances of gene alteration, comprising 6 cases of mutation and
   2 cases of deep deletion (Fig. [114]4B). More specifically, the
   missense mutation of RARRES1 in SKCM led to changes in the amino acids
   that constitute the protein (Table [115]2). All 6 mutated nucleotide
   positions were visually represented in Fig. [116]4C. Copy number
   variations (CNV) involve the amplification or deletion of chromosomal
   segments resulting from genomic rearrangements. CNVs play a crucial
   role in influencing gene expression^[117]24. We then investigated the
   association between the RARRES1 expression and CNVs. As shown in
   Fig. [118]4D, shallow deletion maybe responsible for the downregulated
   of RARRES1. In addition, data from UALCAN database showed that the
   promoter methylation of metastatic melanoma was significantly higher
   than that of the normal group, as shown in Fig. [119]4E. However, due
   to the limited sample size of the normal group, we were unable to
   identify any significant differences in the promoter methylation levels
   between normal and SKCM tissues. In order to conduct a more
   comprehensive analysis of the promoter methylation of RARRES1 in SKCM,
   we identified 15 promoter methylation sites of RARRES1 based on the
   [120]GSE120878 dataset. Among these sites, 10 exhibited significantly
   higher levels of methylation in SKCM compared to nevus
   (Fig. [121]5A-J). These findings suggest a strong correlation between
   the downregulation of RARRES1 expression in melanoma and the presence
   of mutations as well as elevated promoter methylation.

Figure 4.

   [122]Figure 4
   [123]Open in a new tab

   Mutation and promoter methylation of RARRES1 in SKCM. (A) Mutation of
   SKCM in 443 SKCM samples. (B) Mutation frequency distribution of
   RARRES1 in SKCM. (C) 6 mutated nucleotide positions in RARRES1 gene.
   (D) RARRES1 expression in different RARRES1 CNV groups. (E) Promoter
   methylation level of RARRES1 in normal tissues, primary SKCM, and
   metastatic SKCM. **** P < 0.0001.

Table 2.

   Mutated nucleotide position of RARRES1 in SKCM.
   Sample ID TCGA-D3-A51T-06 TCGA-GN-A266-06 TCGA-EE-A29M-06
   TCGA-EB-A41A-01 TCGA-EB-A431-01 TCGA-FR-A8YC-06
   Cancer type SKCM SKCM SKCM SKCM SKCM SKCM
   Protein Change G279R W192* G124E T140A W204* R189K
   Mutation Type Missense Nonsense Missense Missense Nonsense Missense
   Variant Type SNP SNP SNP SNP SNP SNP
   Copy # Diploid Diploid Gain Blanks Blanks Shallow Deletion
   Chromosome 3 3 3 3 3 3
   Start position 158,415,517 158,422,676 158,428,691 158,428,644
   158,422,641 158,422,686
   End position 158,415,517 158,422,676 158,428,691 158,428,644
   158,422,641 158,422,686
   Allele Freq (T) 0.3 0.31 0.44 0.47 0.11 0.3
   # Mut in Sample 978 3271 1959 1531 1398 2480
   [124]Open in a new tab

Figure 5.

   [125]Figure 5
   [126]Open in a new tab

   10 high promoter methylation sites in SKCM. (A) cg03269060; (B)
   cg07693931; (C) cg11188318; (D) cg12580509; (E) cg14226182; (F)
   cg18349298; (G) cg08697665; (H) cg09503974; (I) cg09612534; (J)
   cg16708174. *P < 0.05, **P < 0.01, ***P < 0.001, **** P < 0.0001.

Decreased RARRES1 expression was confirmed in SKCM by immunohistochemistry

   To validate the expression of RARRES1, a total of 20 skin samples from
   patients with SKCM and 11 skin samples from individuals with
   intradermal nevus were collected for immunohistochemistry staining. As
   depicted in Fig. [127]6A, RARRES1 was found to be weakly expressed in
   the tumor tissues of SKCM. In contrast, a noticeable increase in the
   expression of RARRES1 was observed in intradermal nevi (Fig. [128]6B).
   Furthermore, a semiquantitative analysis demonstrated a significant
   downregulation of RARRES1 in SKCM (Fig. [129]6C).

Figure 6.

   [130]Figure 6
   [131]Open in a new tab

   Analysis of RARRES1 expression in SKCM and intradermal nevus (IDN)
   tissues. (A) Immunohistochemical staining of RARRES1 in SKCM (bar
   length = 200 μm). (B) Immunohistochemical staining of RARRES1 in
   intradermal nevus (bar length = 200 μm). (C) Semiquantitative analysis
   of RARRES1 staining.

RARRES1 overexpression inhibits A375 cell proliferation and migration

   Based on the above results, we suppose that Rarres1 is a tumor
   suppressor in SKCM. Furthermore, we conducted gain-of-function assays
   in A375 cells by transfected pCDNA-RARRES1/Vector and
   lentiviruses-RARRES1/Vector. The overexpressing efficiency is shown in
   Fig. [132]7A. Results from MTT assay (Fig. [133]7B) showed that RARRES1
   overexpression inhibited A375 cell proliferation. Consistently, crystal
   violet staining (Fig. [134]7C-D) and soft agar assays (Fig. [135]7E-F)
   showed an injured ability of colony formation on RARRES1 upregulated
   A375 cells. Flow cytometry analysis was subsequently conducted to
   assess the impact of RARRES1 on the cell cycle distribution of A375
   cells. As demonstrated in Fig. [136]7G-H, RARRES1 overexpression
   resulted in an increased cell population in the G0/G1 phase and a
   concurrent decrease in the S phase. In addition, we detected the
   expression of cell cycle related protein. Western blotting results
   revealed that RARRES1 overexpression led to a reduction in the
   expression levels of Cyclin D1 and Cyclin E1, as depicted in Fig [137]7
   I and J. As the pathway related to migration and invasion is visible in
   the enrichment analysis, we next analyzed the role of RARES1 on the
   migration or invasion of A375 cells. We directly studied in vitro
   migration in 24-well plates inserted with 8.0-mm-pore transwells.
   Significant decrease in migration was observed in RARRES1
   overexpression cells compared with vector group (Fig. [138]7K-L).
   Consistently, an increased expression of E-cadherin was observed upon
   RARRES1 overexpression (Fig. [139]7M-N). Overall, these findings
   indicate that RARRES1 plays an important role in cell proliferation,
   colony formation, cell cycle and migration, and its overexpression
   significantly suppresses A375 cell proliferation and migration.

Figure 7.

   [140]Figure 7
   [141]Open in a new tab

   Effects of RARRES1 overexpression on A375 cell proliferation and
   migration. (A) Relative protein levels of RARRES1 in A375 cells
   transfected with pCDNA-RARRES1 of Lentivirus-RARRES1. (B) MTT assay at
   6, 24, 48 and 72 h after RARRES1 overexpression. (C–D) Colony forming
   assays after RARRES1 overexpression. (E–F) Soft agar assays after
   RARRES1 overexpression (bar length = 1 mm). (G–H) Cell cycle analysis
   after RARRES1 overexpression. (I-J) Western blotting analysis of cell
   cycle regulators after over expression. (K–L) Transwell migration assay
   after RARRES1 overexpression (bar length = 100 μm). (M–N) Western
   blotting analysis of E-cadherin after RARRES1 overexpression.
   *P < 0.05, **P < 0.01, ***P < 0.001. All assays were performed in
   triplicate.

Overexpression of RARRES1 promotes apoptosis and inhibits autophagic flux in
A375 cell line

   We then investigated the role of RARRES1 in the regulation of cell
   apoptosis. A375 cells transfected with the pCDNA-vector showed
   significantly low baseline apoptosis rates, whereas overexpression of
   RARRES1 noticeably increased the rate of cell apoptosis
   (Fig. [142]8A-B). RARRES1 has been reported to activate p53, thereby
   promoting apoptosis. We subsequently investigated the impact of RARRES1
   on p53 in A375 cells, and the results showed that RARRES1
   overexpression significantly increased the protein levels of p53 and
   its target gene BAX. Additionally, we observed an increase in the
   expression of BAK, another key executor of mitochondrial membrane
   permeabilization in the mitochondrial pathway apoptosis
   (Fig. [143]8C-D). Since ROS plays a crucial role in various signaling
   pathways, such as cell cycle regulation and apoptosis, we further
   explored the effect of RARRES1 on ROS and observed a significant
   increase in ROS levels with RARRES1 overexpression (Fig. [144]8E-F). As
   ROS elevation also promotes autophagy^[145]25, we next accessed whether
   Rarres1 over-expression impacts autophagy. As shown in Fig. [146]8G-H,
   an augmented LC3 conversion from LC3-I to LC3-II in cells
   overexpressing RARRES1 was discovered through immunoblotting assays.
   These findings indicate that the overexpression of RARRES1 led to the
   accumulation of autophagosomes in A375 cells. This could be attributed
   to either an enhancement in autophagy induction or a suppression in
   autophagosome degradation. To elucidate the underlying factor
   contributing to the autophagosome accumulation, we initially assessed
   the expression levels of key autophagy-related proteins. Our results
   demonstrated that RARRES1 overexpression significantly upregulated the
   mRNA expression of ATG5, ATG7, and ATG10 (Fig. [147]8I), demonstrating
   that RARRES1 overexpression induced autophagic activation. Next, we
   investigated whether RARRES1 overexpression hinders downstream
   processes in autophagic flux. We analyzed the expression of the
   SQSTM1/P62 protein, which acts as a selective autophagy receptor and is
   degraded alongside ubiquitinated substrates during autolysosome
   degradation. As demonstrated in Fig. [148]8J-K, we observed an
   accumulation of p62 upon Rarres1 overexpression. This finding aligned
   with the inhibition of autophagic flux during the later stages.
   Subsequently, we hypothesized that the blockage of autophagic flux
   could be attributed to lysosomal dysfunction. Since a low lysosomal pH
   is essential for maintaining proper lysosomal function, we utilized
   Lysosensor DND-189, a pH-dependent probe that exhibits increased
   fluorescence upon acidification, to assess lysosomal pH levels. As
   presented in Fig. [149]8L-M, Rarres1 overexpression led to a reduction
   in fluorescence, indicating a decrease in lysosomal acidification and
   thus contributing to the impairment of autophagic flux. Overall, these
   findings suggest that Rarres1 overexpression stimulated autophagic
   initiation while simultaneously impeding the degradation of autophagic
   cargo during the later stages.

Figure 8.

   [150]Figure 8
   [151]Open in a new tab

   Effects of RARRES1 overexpression on A375 cell apoptosis and autophagy.
   (A–B) Cell apoptosis assay after RARRES1 overexpression. (C–D) Western
   blotting analysis of cell apoptosis biomarkers after RARRES1
   overexpression. (E) Representative images and (F) quantification of ROS
   in A375 cells measured by DHE dye after RARRES1 overexpression (bar
   length = 50 μm). (G–H) Western blotting analysis of LC3I/II after
   RARRES1 overexpression. (I) Relative mRNA levels of autophagy related
   genes after RARRES1 overexpression. (J–K) Western blotting analysis of
   P62 after RARRES1 overexpression. (L) Representative images and (M)
   quantification of lysosomal pH in A375 cells determined by lysosensor
   Green DND-189 after RARRES1 overexpression (bar length = 50 um). (N)
   Schematic diagram illustrating the mechanism by which RARRES1 inhibits
   proliferation and migration of SKCM cells. *P < 0.05, **P < 0.01,
   ***P < 0.001. All assays were performed in triplicate.

   In summary, our findings suggest that RARRES1 overexpression inhibits
   SKCM cell proliferation by inducing ROS-mediated cell cycle arrest,
   apoptosis induction, and autophagy blockade. Furthermore, RARRES1
   overexpression also suppresses the migration of SKCM cells
   (Fig. [152]8N).

Overexpression of RARRES1 suppresses tumor growth in vivo

   Xenograft tumor experiments were conducted to assess the in vivo
   effects of RARRES1. The nude mice were administered a subcutaneous
   injection of stable RARRES1-overexpressing cells. The tumors of
   RARRES1-overexpressing cells exhibited smaller volume and reduced
   weight (Fig. [153]9A-D). The stable overexpression of RARRES1 is
   depicted in Fig. [154]9E. Immunohistochemical staining (IHC)
   demonstrated a significant reduction in the expression of the cell
   proliferation marker Ki67 in the RARRES1 overexpression group
   (Fig. [155]9F-G). These results confirmed that RARRES1 overexpression
   inhibits tumor growth in vivo, consistent with the in vitro
   experimental findings.

Figure 9.

   [156]Figure 9
   [157]Open in a new tab

   Overexpression of RARRES1 suppresses tumor growth in vivo. (A) Pictures
   of nude mice implanted with A375 cells overexpressing RARRES1. (B)
   Pictures of the tumors harvested at the end of the experiment. (C)
   Tumor growth of nude mice (n = 6). (D) The weight of tumors harvested
   (n = 6). (E) Validation of RARRES1 overexpression in tumor tissues by
   Western blotting. (F) Immunohistochemistry staining of Ki-67 in
   xenograft tumor samples (n = 6). Bar length = 50 μm. (G) Quantification
   of Ki-67 staining. *P < 0.05, **P < 0.01, ***P < 0.001.

Discussion

   RARRES1, also referred to as Tazarotene-induced gene 1 (TIG1), plays a
   crucial role in the regulation of various tumorigenic processes.
   Studies have shown a significant downregulation of RARRES1 expression
   in several types of cancers, including prostate, colorectal,
   nasopharyngeal, gastric, endometrial, and testicular cancers.
   Furthermore, it has been observed that the expression level of RARRES1
   is inversely correlated with patient survival in cases of prostate and
   colorectal cancers^[158]9,[159]10,[160]26–[161]29. Nonetheless, RARRES1
   might have a promotive effect on cancer development in specific tumor
   types. For example, Wang X et al. demonstrated that enhanced RARRES1
   expression in inflammatory breast cancer facilitated the growth of
   tumors^[162]30.

   To elucidate the expression and functional roles of RARRES1 in SKCM, we
   obtained data from both the GEO and TCGA databases, followed by
   comprehensive bioinformatic analyses. Our findings revealed a
   significant downregulation of RARRES1 expression in SKCM. Furthermore,
   we observed that low RARRES1 expression was associated with unfavorable
   clinical features and prognostic outcomes, regardless of whether
   patients received immunotherapy. Consequently, RARRES1 emerged as an
   independent prognostic factor for SKCM. Moreover, it held promise as a
   valuable diagnostic marker. These results substantiated the suppressive
   role of RARRES1 in the progression of SKCM, supporting previous studies
   exploring RARRES1 in other tumor types^[163]9,[164]10.

   A recent study has unveiled that the coexistence of tumor-associated
   CD8^+ T cells and CD20^+ B cells is correlated with enhanced survival
   in metastatic melanoma^[165]31. Srour et al. performed
   immunohistochemical staining on samples obtained from melanoma patients
   receiving immune checkpoint inhibitor. The results of their study
   revealed a significant association between immune infiltration and the
   therapeutic efficacy of the treatment^[166]32. In our study, we
   conducted GO, KEGG, and GSEA enrichment analysis, which revealed a
   significant association between RARRES1 expression and the tumor immune
   microenvironment, cell cycle regulation, migration, apoptosis,
   autophagy, and oxidative damage. Furthermore, our findings indicated a
   positive correlation between RARRES1 expression in SKCM and the
   infiltration of various immune cells, including macrophages, CD8^+ T
   cells, B cells, and T helper cells. These results suggest that RARRES1
   may enhance its anti-tumor effect in SKCM by promoting immune cell
   infiltration. Consistent with our findings, a previous study
   demonstrated that RARRES1 induces M1 macrophage activation to exert
   antitumor effects in renal clear cell carcinoma^[167]11.

   We also investigated the underlying mechanism behind the reduced
   expression of RARRES1 in SKCM. Prior research has established a
   correlation between the transcriptional silencing of RARRES1 in
   esophageal, gastric, endometrial, and prostate cancers and epigenetic
   inactivation resulting from promoter
   hypermethylation^[168]23,[169]28,[170]29,[171]33. In accordance with
   the aforementioned studies, our findings demonstrated a notably
   increased level of RARRES1 promoter methylation in metastatic melanoma
   when compared to the control group. Moreover, we identified 10 promoter
   sites that exhibit higher levels of methylation in SKCM as compared to
   nevus. Additionally, we observed that various copy number mutations,
   particularly shallow deletion, have also been associated with decreased
   expression of RARRES1 in SKCM.

   Based on the findings of bioinformatic analysis, we proceeded with
   experimental validation in order to corroborate the expression and
   function of RARRES1 in SKCM. IHC staining showed a significant
   downregulation of RARRES1 in SKCM compared to intradermal nevus. Both
   in vitro and in vivo experiments consistently demonstrated that
   increased expression of RARRES1 suppressed the proliferation of A375
   cells, which aligns with the observations made in a prior study
   conducted on colorectal cancer cells^[172]9. Through flow cytometry
   analysis, we discovered that RARRES1 exerts inhibitory effects on cell
   proliferation by suppressing the G1/S transition, as evidenced by the
   downregulation of Cyclin D1 and Cyclin E1 expression. According to the
   results of the enrichment analysis, it was found that RARRES1 is
   strongly correlated with the migration and invasion of melanoma.
   Therefore, a thorough examination was conducted to investigate the
   influence of RARRES1 on cell migration. Consistent with the initial
   hypothesis, the outcomes of the transwell assay demonstrated that
   overexpression of RARRES1 effectively impeded the migration of A375
   cells, which was accompanied by a notable elevation in the protein
   levels of E-cadherin.

   As previously mentioned, the upregulation of RARRES1 in glomerular
   disease resulted in the upregulation of genes related to the cell death
   pathway. Moreover, the overexpression of RARRES1 in podocytes induced
   the activation of P53, leading to apoptosis and promoting the
   advancement of glomerular disease. These discoveries implyed that
   RARRES1 might have a significant involvement in the process of
   apoptosis^[173]34. As expected, our results revealed that RARRES1
   overexpression induced apoptosis in A375 cells through upregulating the
   expression of pro-apoptotic factors P53, BAK and BAX. In addition,
   studies have demonstrated that ROS plays a key role in
   apoptosis^[174]25,[175]35. Consistently, we also found that RARRES1
   overexpression promoted ROS production, providing another plausible
   explanation for RARRES1's role in facilitating cell apoptosis.

   On the other hand, we found autophagy flux in A375 cell was inhibited
   by RARRES1 overexpression. Autophagy is an extensively conserved
   lysosomal degradation pathway that plays a crucial role in maintaining
   cellular homeostasis by clearing protein aggregates and damaged
   organelles^[176]36. Previous studies have shown that RARRES1 activates
   autophagy to inhibit tumor progression in prostate cancer and cervical
   cell carcinoma^[177]7,[178]37. In our study, we discovered that the
   overexpression of RARRES1 led to an increase in the expression of
   ATG5/7/10 and LCII, suggesting the activation of autophagy during the
   early stages. However, we also observed the accumulation of P62, a
   biomarker associated with the late stages of autophagy. This
   accumulation indicated that the overexpression of RARRES1 inhibited the
   progression of autophagy towards the later stages. Recent research has
   highlighted that the inhibition of autophagy at later stages can lead
   to the excessive buildup of undegraded materials within autophagic
   vacuoles, ultimately resulting in an accelerated process of cell
   death^[179]38. Therefore, it was hypothesized that the inhibition of
   autophagic flux is significantly correlated with the heightened
   prevalence of cellular apoptosis.

Conclusion

   In conclusion, the expression of RARRES1 in SKCM was found to be
   diminished as a result of copy number mutations and high promoter
   methylation. Additionally, it was observed that low RARRES1 expression
   correlated with a poorer overall survival. Enrichment analysis
   suggested that RARRES1 may function as a tumor-suppressive factor by
   regulating various aspects of immune cell infiltration, proliferation,
   migration, apoptosis, and autophagy. Experimental validation further
   confirmed the functional roles of RARRES1 in proliferation, migration,
   apoptosis, and autophagy. This study emphasizes the significance of
   RARRES1 as a valuable prognostic biomarker for SKCM and highlights its
   potential as a promising therapeutic target.

Supplementary Information

   [180]Supplementary Table 1.^ (131.6KB, xlsx)
   [181]Supplementary Table 2.^ (186.3KB, xlsx)

Abbreviations

   SKCM
          Skin cutaneous melanoma

   RARRES1
          Retinoic acid receptor responder 1

   TCGA
          The Cancer Genome Atlas

   GO
          Gene ontology

   KEGG
          Kyoto encyclopedia of genes and genomes

   TIG1
          Tazarotene-induced gene

   IHC
          Immunohistochemistry

   ROC
          Receiver operating curve

   GEO
          Gene expression omnibus

   BP
          Biological processes

   MF
          Molecular functions

   CC
          Cellular components

   FDR
          False discovery rate

   CNV
          Copy number variants

   OS
          Overall survival

   DSS
          Disease specific survival

   HR
          Hazard ratio

   AUC
          Area under the curve

   IDN
          Intradermal nevus

Author contributions

   M.L., G.Z. and R.B. wrote and revised the manuscript. Y.Z. and G.W
   designed the research. M.L., G.Z., R.B. and X.L. performed the
   bioinformatics analysis and experiments. Z.W., K.H., X.Z. and P.Y.
   processed data. All authors reviewed the manuscript. All authors have
   read and consented to the publication of the article.

Funding

   This work was supported by grants from the National Natural Science
   Foundation of China (82273541 and 81972938).

Data availability

   The datasets featured in this study are available in online
   repositories, and the names of these repositories along with their
   respective accession numbers can be located within the article. The
   [182]GSE15605 and [183]GSE120878 datasets were downloaded from the GEO
   ([184]https://www.ncbi.nlm.nih.gov/geo/) database. Patient information
   and prognosis data of SKCM was obtained from TCGA
   ([185]https://www.cancer.gov/ccg/research/genome-sequencing/tcga).

Competing interests

   The authors declare no competing interests.

Footnotes

   Publisher's note

   Springer Nature remains neutral with regard to jurisdictional claims in
   published maps and institutional affiliations.

   These authors contributed equally: Meng Liu, Ruimin Bai and Guanfei
   Zhang.

Contributor Information

   Yan Zheng, Email: zenyan66@126.com.

   Guorong Wang, Email: 13227701303@qq.com.

Supplementary Information

   The online version contains supplementary material available at
   10.1038/s41598-024-65032-1.

References