Myelin and lymphocyte protein serves as a prognostic biomarker and is closely associated with the tumor microenvironment in the nephroblastoma

Abstract Nephroblastoma, also known as Wilms' tumor (WT), is the most common renal tumor that occurs in children. Although the efficacy of treatment has been significantly improved by a series of comprehensive treatments, some patients still have poor prognosis. Myelin and lymphocyte (MAL) protein, a highly hydrophobic integrated membrane‐bound protein, has been implicated in many tumors and is also closely linked to kidney development. However, the relationship between MAL and WT has not yet been elucidated. Therefore, we attempted to evaluate the feasibility of MAL as a promising prognosis factor for WT. The differential expression of MAL was investigated using TARGET database and was verified using the Gene Expression Omnibus database and real‐time quantitative PCR. The prognostic ability of MAL was determined using Kaplan–Meier and Cox regression analyses. Pearson correlation analysis was applied to explore the relationship between MAL expression and methylation sites. The ESTIMATE and CIBERSORT algorithms showed that MAL expression was associated with the WT tumor microenvironment. Gene Set Enrichment Analysis (GSEA) indicated that multiple signaling pathways closely associated with tumorigenesis were differentially enriched between the high‐ and low‐MAL groups. In conclusion, our study comprehensively explored the potential of MAL as a prognosis factor for WT. Meanwhile, we also demonstrated that MAL, as a prognostic factor for WT, may be closely related to the tumor microenvironment.


| INTRODUCTION
Nephroblastoma, also known as Wilms' tumor (WT), is the renal tumor with the highest incidence in children, accounting for 90% of all childhood renal tumors. 1 It has been shown that about 1 in 10,000 children have had this disease, with the highest incidence at the age of 3. 2 At present, the overall survival (OS) rate of patients with WT has increased to more than 90% through a series of comprehensive standard therapies. 3,4 However, some patients still have poor prognosis because of cancer recurrence and metastasis, 5 and a quarter of survivors experience severe chronic diseases associated with the antitumor therapy they receive. 6 Therefore, new and promising prognosis biomarkers are needed to explore for patients with WT and provide a reference for the implementation of clinical decision-making.
As a primitive, pluripotent embryonic renal precursors malignancy, a large number of studies have shown that the occurrence of WT was closely related to early renal dysplasia. 7 The kidneys began to develop through a series of complex tissue interactions between mesoderm derivatives around the fifth week of pregnancy. In this process, the transformation of mesenchymal progenitor cells into epithelial cells was crucial, which was finely regulated by a gene network composed of a series of genes. Mutations of genes in this process, such as the WT1 gene and related genes in the WNT pathway, often resulted kidney dysplasia thus caused a series of kidney related diseases. [8][9][10][11] Therefore, the normal development of the kidney was of great significance for the exemption of kidney-related diseases. Frank et al. 12 found that during the whole development of kidney, the high expression level of MAL would be maintained until adulthood. And MAL played a role of structural protein in the apical membrane of specialized epithelium such as renal tubules. Its functions of differentiation and transportation may play an important role in the development of kidney. 12,13 MAL had been reported to play an important role in a variety of tumors, [14][15][16][17] but the role of MAL gene in nephroblastoma had not been reported yet. Therefore, we perceived that it was necessary to explore the features of MAL in WT and conduct further studies.
In this study, we tried to clarify the relationship between MAL expression and the prognosis of WT through related bioinformatics methods, and at the same time explored its relationship with tumor microenvironment and related signaling pathways. The results from our research were expected to provide new perspectives for exploring the prognosis factors for WT and guide for the implementation of clinical decision-making.

between WT and normal tissues
First, the fragments per kilobase of exon model per million mapped reads (FPKM) RNA-seq data of 124 primary WT and six adjacent normal tissues were download by using TCGAbiolinks R package. Further, corresponding clinical data were acquired from UCSC Xena (https://xenab rowser.net). Next, the Emsembl database (http://asia.ensem bl.org/index.html) was used to convert Ensembl IDs into gene symbols. The expression file of the MAL gene was extracted for further analysis. Differential expression of MAL between WT and adjacent normal tissues was identified using a t-test. Finally, Gene Expression Omnibus database (GEO. http://www. ncbi.nlm.nih.gov/geo/) was used to verify the accuracy of the results, including GSE2712, GSE1115, and GSE73209. Detailed information of these validation datasets is shown in Table 1.

| Prognostic value analysis
The expression profile was processed in log2 (normalized value + 1) data format. Survival analysis was performed using the survminer R package. The log-rank test was used to confirm statistical significance, and the basis of grouping was dependent on the median value of MAL expression. The comparison between MAL expression and other clinical characteristics on the survival impact of patients with WT was explored by univariate and multivariate Cox analyses (age, gender, and stage).

| Exploration of relationship between MAL expression and tumor microenvironment
We further explored the relationship between MAL expression and tumor microenvironment in order to understand the difference of tumor microenvironment between high and low-MAL groups. The tumor microenvironment was identified as the cellular environment where cancerous cells were present, which was composed of a series of cell types, including mesenchymal cells, immune cells, inflammatory mediators, endothelial cells, and extracellular matrix (ECM) molecules. 18 The stromal score, immune score, ESTIMATE score, and tumor purity of each sample were calculated in the R platform using the ESTIMATE algorithm. The difference between high and low-MAL groups was further compared using the Wilcoxon test. KM analysis was performed to explore the survival influence of tumor microenvironmentrelated scores. The basis of grouping was determined by its median value. The infiltration fraction of 22 immune cell types in each sample was estimated by performing the CIBERSORT algorithm. The Wilcoxon test was conducted to compare the difference between high and low-MAL groups.

| Differential signaling pathways demonstrate through GSEA between the high and low-MAL groups
The differentially activated signaling pathways between the high-and low-MAL groups were explored in the R platform. The ordered gene list was identified using the edgeR R package, and Gene Set Enrichment Analysis (GSEA) was performed using the gseKEGG function of clusterProfiler R package. 19

| Real-time quantitative PCR analysis
We performed real-time quantitative PCR (RT-qPCR) in vitro to show the expression of MAL in WT. The clinical specimens were from three adjacent normal tissues and three WT tissues who had been pathologically identified at the Department of Pathology, First Affiliated Hospital of Guangxi Medical University, from January 2019 to December 2019, and processed suitably. RT-qPCR was performed using PC33-2 in an ABI 7500 cycler (Applied Biosystems, Aidlab Biotechnologies Co., Ltd). The differential expression of MAL in WT tissues relative to the adjacent normal tissues was calculated using the 2 −ΔΔ Ct method, and GAPDH was identified as the internal control. The primers of MAL and GAPDH were synthesized by Aidlab Biotechnologies and the sequences were as follows: MAL: Forward 5′-CGACTTGCTCTTCATCTTTGAG-3′ and Reverse 5′ ATGTACAG GATGATCAAGGTGG-3′; GAPDH: Forward 5′-AGAAGGCTGGGGCTCATTTG-3′ and reverse 5′-AGGGGCCATCCACAGTC TTC-3′. All experiments were repeated at least three times.

| Statistical analysis
All statistical analyses were carried out using the R platform (version: 3.61, http://www.r-proje ct.org/). The relationship between MAL expression and pathological features was analyzed using the Kruskal-Wallis and Wilcoxon rank tests. The minifi R package was used to process the original data and calculate the β value. The IlluminaHumanMethylation450kanno. ilmn12.hg19 R package was used for the conversion of gene symbols. Two-tailed p value < 0.05 was considered statistically significant in all statistical analysis: *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001.

| Differential expression in MAL between WT and normal tissues
The results of RT-qPCR analysis are shown in Figure 1, and the expression of MAL was lower in WT than in adjacent normal tissues (p < 0.05, Figure 1), which showed the differential expression of MAL in adjacent normal tissues and WT in vitro.
The expression of MAL in the TARGET, GSE2712, GSE11151, and GSE73209 datasets also showed significant statistical differences (Figure 2A-D, p < 0.05). MAL was inadequately expressed in WT, and it was not significantly correlated with age, gender, stage, and histologic type ( Figure 3A-D).

| Survival outcomes and univariate COX analysis
KM analysis showed that the low-MAL group had worse prognosis than the high-MAL group in WT (p = 8.81e-03, HR = 0.471, 95% CI: 0.27-0.822, Figure 4). In addition, univariate and multivariate Cox analyses proved that MAL was an independent prognostic factor for overall survival (OS), with an HR of 0.582 (p = 2.35E-02, Table 2).

| Correlation analysis between MAL expression and methylation
We found that there were 17 methylation sites in the MAL gene. The β value of the methylation site in the MAL gene is shown in Figure 5A. The correlation analysis showed that MAL expression was related to cg03566174 (R = 0.21, p = 2.04E-02, Figure 5B, Figure S1) and cg05314420 (R = −0.265, p = 3.17e-03, Figure 5C, Figure S1) methylation sites. Kaplan-Meier analysis revealed that the cg05314420 methylation site of MAL was associated with OS (p = 2.29e-02, HR = 2.188, 95% CI: 1.225-3.907, Figure 5D,E).

| MAL expression and tumor microenvironment
The results of the immune microenvironment score indicated that the immune score, stromal score, and ESTIMATE score in the high-MAL group were statistically higher than those in the low-MAL group, while the tumor purity was opposite (p < 0.05, Figure 6A-D). KM analysis showed that the stromal score and ESTIMATE score presented protective factors, while tumor purity showed a risk factor, which indirectly reflected the low-MAL group with a worse prognosis ( Figure 7A-D).
Additionally, the infiltration analysis of 22 immune cell components of each WT sample revealed that CD8 T cells, resting CD4 memory T cells, monocytes, M2 macrophages, resting dendritic cells were significantly different between the high-and low-MAL groups via Wilcoxon analysis(p < 0.05, Figure 8), which indicated that there were certain differences in immune cell infiltration in tumor microenvironment between high-and low-MAL groups, which may affect the prognosis of WT patients to a certain extent.

| Differential signaling pathways demonstrate through GSEA between the high-and low-MAL groups
We performed GSEA between low-and high-MAL groups and the result showed that some important signaling pathways existed differentially enriched between the two groups, such as Wnt signaling pathways and PPAR signaling pathways (NES > 1, p Adjust <0.05, Figure 9A-C).

| DISCUSSION
As an integrated-membrane protein with multiple biological functions and playing a significant role in cell differentiation and proliferation, MAL has been extensively studied in relation to numerous tumors. 20,21 MAL not only inhibited the progression of cancer, 15,22,23 but in some cases, it also promoted the development of cancer. 17,[24][25][26] Moreover, the role of MAL in the kidney has been mentioned in relevant studies. 21,27,28 However, there have been few studies on the role of MAL in WT. In our study, we found that MAL was under-expressed in WT, and it was not significantly correlated with age, gender, or stage. Meanwhile, RT-qPCR in vitro also confirmed the low expression of MAL in WT, which provided us with a preliminary understanding of the role of MAL in WT.
Considering the difference of MAL expression between WT and adjacent normal tissues, the survival status difference of WT was further analyzed based on the grouping of MAL expression values. Meanwhile, the univariate and multivariate Cox analyses were also used to explore the influence of MAL expression on prognosis of WT and it was shown to be an independent prognostic factor in our analysis. Therefore, in the subsequent analysis, we conducted a series of explorations based on the expression difference of MAL in WT to understand the potential impact of MAL as a prognostic factor.
The level of gene methylation can affect the expression of genes, and methylation of MAL has been extensively demonstrated in a variety of cancers. 15,16,22,23,29 Therefore, in order to explore the methylation level of MAL in WT, we used Pearson correlation analysis to investigate the relationship between MAL expression and methylation sites in WT. The results showed that the expression of MAL was closely related to the methylation sites of cg05314420 and cg03566174. It was worth noting that cg05314420 was positively correlated with the expression of MAL, while cg03566174 was negatively correlated with the expression of MAL. Therefore, the decrease of MAL expression may be the result of the combined effects of multiple methylation sites. Further, our analysis showed that methylation sites of cg05314420 could affect the OS of patients with WT, which indicated that the degree of methylation of MAL may be one of the reasons why MAL is a prognostic factor of WT. Unfortunately, the related methylation sites of MAL in WT had not been reported, but which would provide us with a new idea for the study of WT.
Not only the characteristics of the tumor itself, but also the microenvironment of the tumor affects the progression of the tumor. There is growing evidence showing that the tumor microenvironment is critical to the occurrence and prognosis of tumors. 30

T A B L E 2
The results of univariate and multivariable Cox regression analyses between high-and low-MAL group by the ESTIMATE algorithm and CIBERSORT algorithm. The results of ESTIMATE algorithm showed that patients with low-MAL expression turned out to have lower stromal score, immune, and ESTIMATE scores, while purity of tumor was the opposite. The patients with higher stromal score or lower tumor purity had more favorable prognostic. This is consistent with previous studies. 36,37 Meanwhile, the results of CIBERSORT algorithm also indicated that the expression of MAL was closely correlated with the infiltration of CD8 T cells, resting CD4 memory T cells, monocytes, M2 macrophages, and resting dendritic cells, which suggested to some extent that MAL, as a prognostic factor of WT, may be closely related to tumor microenvironment. Due to the low mutation load and immunogenicity of WT, it was not sensitive to most immunotherapy methods. However, because of the low toxicity of immunotherapy, the exploration of immunotherapy approaches for WT has never stopped. 35,38 Therefore, it was hoped that our study could bring some ideas for the follow-up research of WT. To investigate the differential activated signaling pathways between the high-and low-MAL groups, we performed GSEA. The results showed that the differential activated signaling pathways between the high-and low-MAL groups were important in cancer development, such as the Wnt signaling pathway and PPAP signaling pathway. The Wnt signaling pathway is a highly conserved pathway throughout evolution, and the complexity and function of the pathway is of vital importance in human embryonic development and growth. 39,40 Activation of the Wnt signaling pathway is very common in tumors and is abundant in other diseases. 41,42 In terms of cancer, Fan et al. demonstrated that RTL1 could promote melanoma proliferation by regulating the Wnt/β-catenin signaling pathway 43 and Korbut et al. also revealed the Wnt/β-catenin signaling pathway regulated by Tiam1 could contribute the metastasis of thyroid cancer. 44 In breast carcinoma, ALX4 has been shown to be a downregulated epigenetic tumor suppressor gene that could also inhibit the progression of breast cancer via the Wnt/β-catenin pathway. 45 Moreover, the mutations of related genes in the Wnt signaling pathway play a critical role in WT, [46][47][48] which suggests the importance of the Wnt signaling pathway in WT. Peroxisome proliferator-activated receptors (PPARs) are nuclear transcription factors that are related to cellular growth and differentiation, energy metabolism, insulin sensitization, and tumor regulation. 49,50 PPARs include three major members: PPARɑ, PPAR β/δ, and PPARγ, which have different tissue distribution and functions. 51 In the kidney, PPARɑ regulates the homeostasis of energy metabolism 52 and PPARγ maintains lipid and glucose homeostasis and is important for the control of renal function. 53 Although little research has been done on the effect of the PPAR signaling pathway in the WT, our study would provide a novel view for the study of the WT and PPAP signaling pathways. The results of GSEA showed that multiple signaling pathways closely associated with tumorigenesis were differentially enriched between the high-and low-MAL groups, which greatly demonstrated the different biological status between the high-and low-MAL groups, and further indicated the feasibility of MAL as a potential prognostic factor for WT.
This study comprehensively explored the potential of MAL as a prognosis factor for WT through complex bioinformatics analysis. Meanwhile, we also demonstrated that MAL, as a prognostic factor for WT, may be closely related to the tumor microenvironment. However, some limitations remain to be considered in our research. First, due to a lack of support from more in vitro and in vivo experiments, the credibility of tumor microenvironmentrelated analysis is a challenge. Second, the data in our study comes from public databases, so there are unknown risks in grouping, blinding, and hidden bias. Therefore, although we have demonstrated the potential of MAL as a prognostic factor for WT in various ways, a prospective, well-designed clinical trial is awaiting further validation of our findings. F I G U R E 6 The relationship of between the expression file of MAL with immune score, stromal score, ESTIMATE score, and Tumor purity of tumor microenvironment. The high-MAL group exhibited higher immune score, higher stromal score, higher ESTIMATE score, and lower tumor purity compared to low-MAL group. *means p < 0.05, **means p < 518 0.01, and ***means p < 0.001. An ns means not statistically significant F I G U R E 8 Differential infiltration analysis of 22 immune cell components in patients with WT. The red represents high-MAL group, and blue represents low-MAL group. The X axis represents the type of immune cells, and the Y axis represents the average level of immune cell infiltration in different MAL groups. *means p < 0.05, **means p < 0.01, and ***means p < 0.001. An ns means not statistically significant F I G U R E 7 KM survival analysis of immune score, stromal score, ESTIMATE score, and tumor purity in patients with WT

| CONCLUSION
In this study, we explored the potential of MAL as a prognosis factor for WT through related bioinformatics analysis. Meanwhile, we also demonstrated that MAL, as a prognostic factor for WT, may be closely related to the tumor microenvironment. Our results were expected to provide a new perspective for exploring prognostic factors of WT.