ANO4 Expression Is a Potential Prognostic Biomarker in Non-Metastasized Clear Cell Renal Cell Carcinoma

Background: Over the past decade, transcriptome profiling has elucidated many pivotal pathways involved in oncogenesis. However, a detailed comprehensive map of tumorigenesis remains an enigma to solve. Propelled research has been devoted to investigating the molecular drivers of clear cell renal cell carcinoma (ccRCC). To add another piece to the puzzle, we evaluated the role of anoctamin 4 (ANO4) expression as a potential prognostic biomarker in non-metastasized ccRCC. Methods: A total of 422 ccRCC patients with the corresponding ANO4 expression and clinicopathological data were obtained from The Cancer Genome Atlas Program (TCGA). Differential expression across several clinicopathological variables was performed. The Kaplan–Meier method was used to assess the impact of ANO4 expression on the overall survival (OS), progression-free interval (PFI), disease-free interval (DFI), and disease-specific survival (DSS). Univariate and multivariate Cox logistic regression analyses were conducted to identify independent factors modulating the aforementioned outcomes. Gene set enrichment analysis (GSEA) was used to discern a set of molecular mechanisms involved in the prognostic signature. Tumor immune microenvironment was estimated using xCell. Results: ANO4 expression was upregulated in tumor samples compared to normal kidney tissue. Albeit the latter finding, low ANO4 expression is associated with advanced clinicopathological variables such as tumor grade, stage, and pT. In addition, low ANO4 expression is linked to shorter OS, PFI, and DSS. Multivariate Cox logistic regression analysis identified ANO4 expression as an independent prognostic variable in OS (HR: 1.686, 95% CI: 1.120–2.540, p = 0.012), PFI (HR: 1.727, 95% CI: 1.103–2.704, p = 0.017), and DSS (HR: 2.688, 95% CI: 1.465–4.934, p = 0.001). GSEA identified the following pathways to be enriched within the low ANO4 expression group: epithelial–mesenchymal transition, G2-M checkpoint, E2F targets, estrogen response, apical junction, glycolysis, hypoxia, coagulation, KRAS, complement, p53, myogenesis, and TNF-α signaling via NF-κB pathways. ANO4 expression correlates significantly with monocyte (ρ = −0.1429, p = 0.0033) and mast cell (ρ = 0.1598, p = 0.001) infiltration. Conclusions: In the presented work, low ANO4 expression is portrayed as a potential poor prognostic factor in non-metastasized ccRCC. Further experimental studies should be directed to shed new light on the exact molecular mechanisms involved.


Introduction
Renal cell carcinoma (RCC) is the most common genitourinary tumor in the United States, with a projected 79,000 new cases and 13,920 deaths in the year 2022 [1]. There is a twofold increase in the likelihood of men being diagnosed with kidney cancer as compared to women [1]. About 60% of cases are detected incidentally and 20% to 30% of patients present with metastatic disease [2]. In most cases, metastatic renal cancer is incurable [2,3].

Clinical and Transcriptomic Data Acquisition and Processing
This work aims to investigate the prognostic utility of ANO4 mRNA expression in non-metastasized ccRCC (KIRC). KIRC clinical and transcriptomic data from The Cancer Genome Atlas Program (TCGA) were accessed using the California Santa Cruz Cancer Genomics Browser (UCSC Xena, http://xena.ucsc.edu, accessed on the 1 October 2022); a web-based platform for visualizing and analyzing public genomic data resources [25]. Experimental genotypic profiling was performed using Illumina HiSeq 2000 RNA sequencing platform to obtain level 3 data. ANO4 expression from tumor samples with normal adjacent tissues was downloaded in an RNA-Seq by expectation maximization (RSEM) normalized count transformed as log 2 (x + 1). Clinicopathological variables incluedg age, gender, American Joint Committee on Cancer (AJCC) stage, International Society of Urologic Pathologists (ISUP) grade alongside the TNM scoring system which comprises tumor size, lymph node involvement, and metastasis status. Primary end points were the overall survival (OS), progression-free interval (PFI), disease-free interval (DFI), and disease-specific survival (DSS). The KIRC cohort was curated by omitting patients with an OS time of 0, metastasis status of M1 and MX, and patients without ANO4 expression data. KIRC cohort was also divided into two subsets based on ANO4 expression (high vs low expression), with a cut-point determined by X-tile software [26].

Statistical Analysis
IBM SPSS statistical package for Windows v.26 (Armonk, NY, USA) and GraphPad prism v.9.3.1 (San Diego, CA, USA) were utilized for statistical analysis and graph generation. Nominal data were presented as frequency (percentage). Mean ± standard deviation of the mean (SD) or standard error of the mean (SEM) were used to present normally distributed continuous variables, while non-normally distributed data were presented as median (interquartile range (IQR)). Normality was evaluated using the Kolmogorov-Smirnov test and the Shapiro-Wilk test aided with quantile-quantile (Q-Q) plots. Comparison between ANO4 expression status against clinicopathological variables was performed as follows: Chi-square test or Fisher's exact test for categorical variables, paired t-test for normally distributed paired samples, unpaired t-test and Welch's corrected unpaired t-test for normally distributed non-paired samples according to variance equality, and finally, Wilcoxon matched pairs test and Mann-Whitney U-test for non-normally distributed data.
Kaplan-Meier survival methods were utilized to evaluate the impact of ANO4 expression status in relation to OS, PFI, DFI, and DSS. Statistical difference across survival curves was detected using a log-rank test reporting the p-value, 95% confidence intervals (95% CI), and hazard ratios (HR  [27]. All statistical tests were two-sided and a p ≤ 0.05 was considered statistically significant.

Gene Set Enrichment Analysis (GSEA)
GSEA was performed to discern a set of molecular mechanisms involved in the prognostic ANO4 signature. GSEA performs genome-wide transcriptional profiling across two expressional groups (high vs low ANO4 expression) against a set of genes representing pivotal processes involved in oncogenesis/tumorigenesis [28]. Three molecular signatures databases were involved in the analysis, namely: hallmark gene sets, the C2 positional gene sets (Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways in cancer), and C5 ontology gene sets (BP (biological process), CC (cellular component), and MF (molecular function)). The number of permutations was set to 1000 with the "gene set" permutation type. The chip platform was Human_UniProt_IDs. Gene sets were considered significantly enriched with an adjusted p value ≤ 0.05 and a false discovery rate (FDR) < 0.25.

Protein-Protein Interaction (PPI) Network Construction
Identification and retrieval of ANO4-related interacting genes were browsed using STRING v.11 (https://string-db.org, accessed on 1 October 2022), an online PPI networking resource based on functional interactions annotated using gene-enrichment analysis, GO/KEGG classification systems, and high-throughput text-mining [29]. Significant interactions were labeled if a combined score ≥ 0.4 was observed. The maximum number of interactions was limited to 50.

Immune Infiltration Analysis
Immune cell infiltration was enumerated from transcriptomes using xCell, an online platform with pre-calculated TCGA immune infiltration estimates that deploys a curve-fitting approach with a novel spillover compensation method [30]. Spearman's correlation test was used to report Spearman's rank correlation coefficient (rho, ρ) and the correlation significance.

Results
The presented work studies the usefulness of ANO4 expression as a potential prognostic biomarker in non-metastasized ccRCC using a modified TCGA-KIRC cohort. It consisted of 422 patients with a median age of 61 (IQR: 51-71) and a male predominance (275, 65.2%). The majority of patients with non-metastasized ccRCC were found to have tumor grade 2 (204, 48.3%), followed by grade 3 (162, 38.4%) and grade 4 (39, 9.2%). In regard to tumor staging, the cohort was mainly distributed across two stages including Stage 1 (243, 57.6%) and stage 3 (121, 28.7%), which was definitely in concordance with the pathological T and N scoring distribution. Table 1 represents the baseline demographical and clinicopathological characteristics of the modified TCGA-KIRC cohort. X-tile software divided the patients based on ANO4 expression harboring two groups (low expression, n = 126 and high expression, n =296). Data are presented as n (%) or median (IQR).

Low ANO4 Expression Is Correlated with Poor Clinicopathological Features in Non-Metastasized ccRCC
In comparison with normal kidney tissue, the expression of ANO4 was significantly upregulated (p < 0.0001) in the non-metastasized KIRC ( Figure 1A). The latter finding was confirmed by analyzing the expression of ANO4 using age and gender-matched participants (p < 0.0001), as illustrated in Figure 1B. Although tumor tissues exhibit a high ANO4 expression compared to normal tissue, intriguingly, low ANO4 expression within tumor samples is associated with advanced demographics and clinicopathological features. ANO4 expression did not exhibit a tractable significant difference across age subgroups (p = 0.1989), as seen in Figure 1C. On the other hand, samples obtained from male participants have a significant low ANO4 expression (p < 0.0001) in contrast to female samples ( Figure 1D). The advanced tumor grade group, including grades 3 and 4, has significant reduced ANO4 expression (p = 0.0497) in comparison to grades 1 and 2 ( Figure 1E). The same pattern tends to be observed also in the tumor stage (p = 0.004, Figure 1F) and pT (p < 0.0048, Figure 1G). Despite the previous associations, a non-significant difference across groups with and without lymph involvement (p = 0.524) was observed ( Figure 1H). Table 2 summarizes the differences between low and high-expression groups based on the dichotomized demographics and clinicopathological features.

Low ANO4 Expression Is Correlated with Poor Clinicopathological Features in non-Metastasized ccRCC
In comparison with normal kidney tissue, the expression of ANO4 was significantly upregulated (p < 0.0001) in the non-metastasized KIRC ( Figure 1A). The latter finding was confirmed by analyzing the expression of ANO4 using age and gender-matched participants (p < 0.0001), as illustrated in Figure 1B. Although tumor tissues exhibit a high ANO4 expression compared to normal tissue, intriguingly, low ANO4 expression within tumor samples is associated with advanced demographics and clinicopathological features. ANO4 expression did not exhibit a tractable significant difference across age subgroups (p = 0.1989), as seen in Figure 1C. On the other hand, samples obtained from male participants have a significant low ANO4 expression (p < 0.0001) in contrast to female samples ( Figure 1D). The advanced tumor grade group, including grades 3 and 4, has significant reduced ANO4 expression (p = 0.0497) in comparison to grades 1 and 2 ( Figure 1E). The same pattern tends to be observed also in the tumor stage (p = 0.004, Figure 1F) and pT (p < 0.0048, Figure 1G). Despite the previous associations, a non-significant difference across groups with and without lymph involvement (p = 0.524) was observed ( Figure 1H). Table  2 summarizes the differences between low and high-expression groups based on the dichotomized demographics and clinicopathological features.  Data are presented as n (%).

ANO4-Related Signaling Pathways Based on GSEA
The first GSEA involved the hallmark gene sets in which a total of 34 out of 50 gene sets were enriched in the low ANO4 expression group, with 21 gene sets having FDR < 25%.     Figure 4B represents the GO results highlighting the top five enriched pathways in each subcategory, including MF, CC, and BP. Figure 5A represents the ANO4-related PPI. The generated PPI contains 11 nodes and 12 edges with an average node degree of 2.18, an average local clustering coefficient of 0.883, and PPI enrichment p-value of 0.329. The nodes include the following PLSCR5, CCDC181, ACOT7, SLC17A8, ASCL1, ASH2L, TSPAN16, DYSF, SPIC, and GAS2L3.

The Correlation between ANO4 Expression and Tumor Immune Infiltrate
Pre-calculated TCGA immune infiltrate estimates were obtained from xCell.

Discussion
Anoctamins (ANO) are a group of anion channels with eight transmembrane domains, and numerous cellular functions [20]. The coding gene family, formerly known as transmembrane proteins with 16 domains (TMEM16), was only just discovered in 2014 thanks to bioinformatic analysis [31]. The new name (anoctamins), proposed by Yang et al., has replaced TMEM16 in GenBank despite objections, and it has been given the HUGO nomenclature seal of approval [31,32]. This family of transmembrane proteins constitutes ten paralogues (ANO1 through ANO10/ TMEM16A to TMEM16K [excluding the J alphabet from the naming]) dispersed across various human tissues [20]. All anoctamins feature eight hydrophobic helices, which have been previously hypothesized to be transmembrane domains according to hydropathy analysis, however, these findings are still debatable [31]. Duran et al. reported that, unlike ANO1 and ANO2 which have a clear Ca +2 activated C -l channel (CaCC) functionality, other members did not pursue such function, since they could not produce a C -l current through Ca +2 activation [33]. Furthermore, ANOs 3 through 7 were determined to be intracellular proteins probably residing in the endoplasmic reticulum without being trafficked to the cell membrane [33]. Despite being scarce, this family has been the subject of an expanding body of literature due to its association with numerous pathologies. To illustrate, ANO1 has been linked to a variety of malignancies. Along with ANO5, which has been connected to specific types of muscular dystrophy, ANO6 and ANO10 have also been linked to Scott syndrome and autosomal recessive spinocerebellar ataxia, respectively [31].
Anoctamins have been involved in a wide variety of cellular functions, including epithelial cell secretion, neuronal activation, smooth muscle contractions, skeletal muscle membrane repair, sensory transduction, and carcinogenesis [20,31,33]. On the molecular level, they behave as CaCCs, along with a scramblase activity, in which they express phospholipids across the membrane from the cytoplasmic side to the extracellular side while requiring Ca +2 [20,31,33]. ANO1 has been the most extensively studied ANO protein and has been shown to mediate Cl −1 secretion in secretory epithelia of the respiratory, gastrointestinal and renal systems, and sweat glands [34]. Under the activation of noxious heat, its role in heat sensation via somatosensory neurons is obvious [35]. Furthermore, ANO1 regulates vascular and bronchial smooth muscle tone. Because of this, ANO1 may play a role in the pathophysiological mechanisms underlying hypertension and asthma, respectively [34,36]. According to Sun et al., the interaction between RANK and ANO1 in osteoclasts causes increased bone resorption and decreased bone mass. In individuals with osteoporosis, this makes ANO1 a viable therapeutic target [37]. ANO1 antagonist can have a synergistic effect with commercially used tocolytics on human uterine smooth muscles [38]. Finally, and perhaps most importantly, it is believed to escalate multiple carcinogenic processes including cellular proliferation, migration, and metastasis [39]. ANO2, a member of the ANO family that is expressed in olfactory sensory cells of the olfactory epithelium may have a function in the olfaction process [40]. It has been found that ANO3 is expressed in the dorsal root ganglia, which controls nociception [34]. ANO5 is most abundantly expressed in the musculoskeletal system including bones, chondrocytes, cardiac, and skeletal muscles [41]. The role of ANO6 as a procoagulant has been confirmed through human and animal studies resulting in Cl −1 influx and swelling of platelets [42]. ANO6 has a well-known scramblase activity, which is also shared with ANO3, ANO4, and ANO7 [43]. Overall, except for ANO1, which has been thoroughly investigated, other members of the ANO family need even more in-depth research, and we are only at the beginning.
ANO1, which is found on locus 11q13, is amplified in multiple tumors, particularly in gastrointestinal stromal tumors (GIST) and head and neck squamous cell carcinoma (HNSCC), in addition to numerous others, such as lung adenocarcinoma, chondroblastoma, esophageal squamous cell carcinoma, salivary gland tumors, oral squamous cell carcinoma, leiomyosarcoma of the uterus, glioma, breast, colorectal and prostate cancer [39,[44][45][46][47][48][49][50][51][52][53][54][55][56][57][58][59][60][61]. ANO1 was a cancer biomarker prior to it is identification as a chloride channel in GIST known as DOG1, and it has received several other names including ORAOV2, and TAOS-2 among oncologists [31,39]. However, its overexpression has, unfortunately, been a poor prognostic factor, leading to greater mortality rates, tumor growth, and invasiveness, as well as its positive correlation with distant metastasis, migration, and tumor grading [45][46][47]50,61]. The carcinogenesis orchestrated by ANO1 has implicated numerous cell signaling pathways. ANO1-mediated proliferation was shown by Duvvuri et al. to be accompanied by extracellular signal-regulated kinase (ERK)1/2 activation, cyclin D1 evocation, and mitogen-activated protein kinase (MAPK) activation [45]. ANO1 upregulates calcium/calmodulin-dependent protein kinase II (CaMKII) and epidermal growth factor receptor (EGFR) expression, the latter of which regulates the MAPK or PI 3 K-AKT pathway. Therefore, ANO1 plays a role in governing cell variability through EGFR-AKT/SRC/MAPK and CaMKII signaling pathways [34]. ezrin-radixin-moesin proteins form a cross-link between the cell membrane and actin filaments of the cytoskeleton, hence its involvement in cell migration. Together with its physical association with ANO1, this might provide a clue to the role of ANO1 in EGF-driven migratory and invasive properties [34]. In addition to its connection to the cytoskeleton, its ability to control cell size aids in the development of shrunken cells that can move via diapedesis through inter-endothelial gaps [39,62]. Another intriguing association lies between the sonic hedgehog signaling pathway and ANO1 which is known to coordinate cellular growth and differentiation [39,63].
Anaplastic thyroid carcinoma appears to acquire its undesirable traits and aggressiveness through ANO1 overexpression; ANO1 knockdown greatly reduces the tumor's aggressive behavior [64]. In gastric cancer, ANO5 has been linked to a negative prognostic role; when it is knocked down, it causes apoptosis, reduces cell proliferation, and arrests the cell cycle at the G1/S transition [65]. According to research by Pan et al., osteosarcomas express ANO5, a pro-tumorigenic factor that increases tumor size, grade, and metastasis. The instability and destruction of nel-like proteins 1 and 2 enable such activity [66]. ANO5 expression is scarce in healthy pancreatic tissue, but it is elevated in pancreatic cancer, where it contributes to the disease's proliferation and migration [67]. ANO7 is only expressed in prostate cells, has an unknown function, and is negatively correlated with the prognosis of prostate cancer. Its low expression has been linked to positive surgical margin, lymph node metastasis, high classical and quantitative Gleason grades, advanced tumor stage, high Ki67 labelling index, and early biochemical recurrence [68].
We focused our bioinformatic research on ANO4 because it is one of the ANO family members that has not received enough attention. It is mostly expressed in the cervix, ovaries, prostate, adrenal glands, and central nervous system [69]. Given this, it has been suggested that it contributes to a variety of neurological diseases, such as schizophrenia, multiple sclerosis, Alzheimer's disease, and anxiety disorders [70][71][72][73][74]. Reichhart et al. reported that ANO4 acts as a Ca 2+ -dependent phospholipid scramblase and monovalent nonselective ion channel [69]. Leitzke et al. reported ANO4 to modulate disintegrinlike metalloproteases ADAM 10 and 17 sheddase activity, evident through the increasing ADAM10 and 17 substrates: transforming growth factor alpha (TGF-α), amphiregulin (AREG), and betacellulin. They also demonstrated that the effects brought upon by the overexpression of ANO4 are due to it is scramblase activity, consequently resulting in diminishing AREG and increased cellular proliferation [75]. Maniero et al. recognized ANO4 as a significant gene expressed in the zona glomerulosa cells despite its contradictory effects on aldosterone secretion [76,77]. To the best of our knowledge, this is the first article to relate the expression of ANO4 with tumor involvement.