Clinical implication and immunological landscape analyses of ANLN in pan‐cancer: A new target for cancer research

Abstract Background Anillin is a F‐actin binding protein (ANLN) mainly involved in the process of cytokinesis and known to be dysregulated in diverse cancers. However, the role of ANLN in pan‐cancer prognosis and tumor immunity remains unclear. Methods Gene expression profiles of 31 solid tumors were downloaded from The Cancer Genome Atlas (TCGA) database. ANLN mRNA and protein expression were quantified using quantitative real‐time PCR (qRT‐PCR) and immunohistochemistry (IHC). Protein expression of ANLN was further confirmed in Human Protein Atlas (HPA) database. Cox regression and Kaplan–Meier analysis were utilized to assess the prognostic value of ANLN in pan‐cancer. The correlation between ANLN and different immune gene markers and infiltration cells was analyzed via ESTIMATE and CIBERSORT. A BLCA immunotherapy cohort: IMvigor (210) was used to confirm the role of ANLN in immune response. Results ANLN upregulation was detected in 21 types of cancers and was associated with poor overall survival (OS), disease‐free interval (DFI), and progression‐free interval (PFI) in most cancers except in THYM (Thymoma). Additionally, correlation analysis revealed a significantly positive association between ANLN expression and tumor mutation burden (TMB), microsatellite instability (MSI), immune cells infiltration. and immune checkpoint genes in various cancers. The BLCA immunotherapy cohort confirmed that patients with higher ANLN level had better immune responses and longer OS. Conclusion ANLN may serve as a prognostic biomarker for pan‐cancer. ANLN upregulation is associated with higher TMB, MSI, and immune cell infiltration in multiple types of tumors, shedding new light for cancer treatment.


| INTRODUCTION
Anillin (ANLN) is a F-actin binding protein originally isolated from Drosophila embryo. 1 It is highly conserved and mainly involved in the cortical cytoskeletal dynamics during cytokinesis and cellularization. 2 Dysregulations of ANLN is observed in a wide range of malignant tumors and its overexpression is closely linked to unfavorable outcomes in colorectal cancer, 3 breast cancer, 4 and liver cancer. 5 However, the role of ANLN in pan-cancer remains still unclear.
The tumor microenvironment (TME) consists of not only cancer cells, but also tumor-associated normal cells, including infiltrating stromal and immune cells. 6 Increasing evidence suggests that TME has multiple effects on tumorigenesis, development, and prognosis. [7][8][9][10][11] However, the possible mechanisms of stromal and immune components remain elusive. Immunotherapy has proved to be a powerful clinical strategy for cancer treatment. 12 Monoclonal antibodies targeting cytotoxic T lymphocytes-associated antigen 4 (CTLA4), programed death-1 (PD-1) receptor, and its ligand (PD-L1) have achieved clinical success in some cancers. [13][14][15] However, only a subset of patients could benefit from current immunotherapies. 16 Thus, the discovery of effective immunotherapy targets is of great clinical significance.
Hence, our study comprehensively explored the prognostic significance of ANLN in 31 tumors based on The Cancer Genome Atlas (TCGA) data. Subsequently, we explored the potential correlations between ANLN and tumor immune characteristics.

| Prognosis analysis
The prognostic values of ANLN, including Overall Survival (OS) and Disease-Free Interval (DFI) and Progression-Free Interval (PFI), was assessed via BLCA immunotherapy cohort confirmed that patients with higher ANLN level had better immune responses and longer OS.
Conclusion: ANLN may serve as a prognostic biomarker for pan-cancer. ANLN upregulation is associated with higher TMB, MSI, and immune cell infiltration in multiple types of tumors, shedding new light for cancer treatment.

K E Y W O R D S
ANLN, immune infiltration, pan-cancer, prognosis, TMB univariate Cox regression model. Samples were clustered into high or low expression cohorts based on the median value of ANLN. Kaplan-Meier (K-M) curves and the forest plots were utilized to compare the survival time differences. All analyses were carried out using the "survival" and "survminer" package. After that, the associations between ANLN and clinical parameters, including age, sex, and the TNM stage, were assessed by the chi-square test or Fisher's exact test. All tests were two-tailed and a p-value of 0.05 or less was considered statistically significant. Microsatellite instability (MSI) is a molecular tumor phenotype of a deficient mismatch repair system. 18 Analysis regarding the correlation between ANLN expression and TMB/MSI was conducted by Pearson correlation analysis and shown in the radar chart using the "fmsb" package.

| Correlation between ANLN expression and tumor immunity
ESTIMATE (Estimation of Stromal and Immune cells in Malignant Tumor tissues using Expression data) is a method for estimating the presence of stromal/immune cells in tumor tissues and tumor purity with high accuracy. 19 This method was applied to explore the association between the immune and stromal cell ratio with ANLN expression. 19 We further explored the correlation between ANLN expression and 22 types of infiltrating immune cells using the "CIBERSORT" algorithm. The association between ANLN expression and common immune checkpoint genes was investigated using the "Reshape2" package and shown in heatmaps by the R "RColorBrewer" package. The Spearman method was utlized to assess the correlation coefficients.

| Collection of an immunotherapybased cohort
After systematically searching for immunotherapy gene expression profiles, an immunotherapeutic cohort: advanced urothelial cancer with intervention of atezolizumab, an anti-PD-L1 antibody (IMvigor210 cohort) with complete clinical information was recruited in our study. 20 The complete information of the cohort is available on http://resea rch-pub.Gene.com/imvig or210 coreb iologies. The raw count data were normalized by the DEseq2 R package and transformed into the TPM value for further study.

| Gene set enrichment analysis
Next, our study explored the potential molecular functions and biological pathways regulated by ANLN. The Gene Set Enrichment Analysis (GSEA) software (http:// www.broad insti tute.org/gsea) was used to investigate the enrichment of GO and KEGG pathways between high ANLN expression and low expression groups according to the median value of ANLN. All analyses were conducted with R package "org.Hs.eg.db," "clusterProfiler," "enrichplot," "DOSE/limma." Gene sets with |NES| > 1, NOM p < 0.05, and FDR q < 0.25 were considered as enrichment significant.

| Experimental validation
Fresh tissue samples of 7 LUAD patients were obtained from the Affiliated Suzhou Hospital of Nanjing Medical University. The mRNA and protein expression of ANLN were quantified using quantitative real-time PCR (qRT-PCR) and immunohistochemistry (IHC) as previously described. 10 Briefly, qRT-PCR was conducted on a LightCycler 480 (Roche 480, Germany) real-time thermal cycler, using SYBR Green dye. Each sample was set with three multiple wells. Assays detected with 5 Ct's less than the negative control (No Template Control, NTC), and with Ct < 35 were included for further data analysis. The melting analysis was added finally to evaluate the specificity of PCR products. Expression levels of ANLN were calculated using 2 −ΔΔCt method and normalized to that of GAPDH mRNA (forward: 5′-GTCTCCTCTGACTTCAACAGCG −3′ and reverse: 5′-ACCACCCTGTTGCTGTAGCCAA −3′). The sequences of the primers used were as follows: forward: 5′-CAGACAGTTCCATCCAAGGGAG-3′ and reverse: 5′-CTTGACAACGCTCTCCAAAGCG-3′. IHC staining for ANLN was accomplished by two professional pathologists. Formalin-fixed human lung tumor tissue sections were sliced into serial sections with a thickness of 5 μm for IHC. An anti-ANLN antibody (anti-ANLN: catalog no., BS-7738R; dilution, 1:200; Bioss Inc.) was used as the primary antibody. PBS was functioned a negative control to keep the specificity of staining. A 3,3′-diaminobenzidine (DAB) and hematoxylin counterstain (ZSGB-Bio, China) were used to visualize antibody staining. Finally, immunostained sections were photographed using a microscope (Olympus Corporation). This study was approved by the Ethics Committee of the hospital. Each participant had signed informed consent in accordance with the Declaration of Helsinki (Ethics Approval No. KL901198).

| Survival analysis of ANLN
Our study investigated the correlation between ANLN expression and patients' OS, DFI, PFI using the TCGA cohort. Kaplan-Meier curves showed that high  Furthermore, we explored the correlation between ANLN expression and different clinical features across pan-cancer. Results revealed that higher expression of ANLN was significantly associated with advanced TNM stage in ACC, BLCA, BRCA, ESCA, KICH, KIRC, KIRP, LUAD, and reversely in LIHC. Interestingly, ANLN expression was higher in stage I compared with stage II (p = 0.008), but lower in stage II compared with stage IV (p = 0.0018) in THCA (Figure 4). ANLN expression also differs greatly in age and gender across different tumors ( Figure S1).

| Tumor mutational burden and microsatellite instability analysis data
TMB and MSI are essential in the process of tumor formation and development. We assessed the association between ANLN expression and TMB/MSI across 31 cancers. Results suggested that elevated ANLN expression was significantly associated with higher TMB in ACC, BLCA, BRCA, COAD, KICH, KIRC, LUAD, LUSC, MESO, PAAD, PRAD, READ, SARC, SKCM, STAD, TGCT, THCA, and UCEC, while increased ANLN level was inversely related to TMB in LGG and

| Correlation between ANLN expression and immune infiltration
Tumor-infiltrating immune cells are important components for the tumor microenvironment, and are closely linked to tumorigenesis and progression. 21 Infiltrating stromal and immune cells in tumor microenvironment play essential roles in disturbing tumor signaling and cancer biology. 6 Hence, we firstly explored the correlation between ANLN expression and tumor-infiltrating immune cells. Our results demonstrated that ANLN expression was significantly associated with the infiltration levels of macrophages M1 in 18 cancer types ( Figure 6, those with p < 0.00001 were shown), CD4 + T cells in 12 cancer types, CD8+ T cells in 9 cancer types, T cells regulatory (Tregs) in 13 cancer types, B cells in 8 cancer types, M2 in 7 cancer types, NK cells in 9 cancer types, and neutrophils in 7 cancer types (those with p < 0.00001 were shown in Figure S2). For instance, in BLCA, ANLN expression was positively correlated with M1 macrophages

F I G U R E 3 Forest plots of hazard ratios of ANLN in pan-cancer. (A) Association between ANLN expression and OS; (B) Association between ANLN expression and DFI; (C) Association between ANLN expression and PFI.
Cox regression analysis. p < 0.05 was considered to be significant. and neutrophil infiltration, but negatively correlated with Tregs. In HNSC, ANLN expression was negatively correlated with CD8+ cells, Tregs, but positively correlated with CD4 + T cells memory resting.
Secondly, our study investigated the association between ANLN expression and the immune and stromal cell ratio. Results showed that ANLN expression is positively correlated with the stromal score and immune score in KIRC, THCA, while negatively correlated with in LUSC, SARC, STAD, and UCEC (Figure 7). In addition, our study observed a statistical positive connection between ANLN expression and stromal score in PRAD but noted a negative connection in BRCA, LIHC, TGCT. Moreover, ANLN expression was negatively linked to immune score in PAAD and SKCM (Figure 7).
Immune checkpoint genes (ICGs) represent a novel target in cancer treatment. Therefore, we further examined the association between ANLN expression and the common immune checkpoint genes. As shown in Figure 8

| Immunotherapeutic response prediction
Considering the strong correlation between ANLN expression and tumor immune landscape, our study next explored the immunotherapeutic predictive ability of ANLN in an anti-PD-L1 cohort (IMvigor210). The Kaplan-Meier curves suggested that increased level of ANLN was significantly connected with prolonged OS ( Figure 9A). In addition, patients with higher ANLN expression level exhibited better immune response ( Figure 9B,C).

| GO and KEGG-based gene set enrichment analysis
Next, GSEA was applied to explore the functional enrichment of ANLN expression between high expression and low expression cohorts. GO enrichment term demonstrated that high expression of ANLN was mainly associated with cell immunity and cell proliferation ( Figure S3). KEGG enrichment term exhibited that high expression of ANLN was positively involved "cell cycle" pathways in ACC, BRCA, KICH, HHSC, KICH, LUAD. MESO, OV, PAAD, PCPG, SARC, and UVM ( Figure S4).

| DISCUSSION
The pan-cancer analysis project launched by TCGA was aimed to illustrate an integrated picture of commonalities and heterogeneities across different tumors. 22 The project interpreted molecular aberrations from genomic, transcriptional, proteomic, and epigenetic levels in large patient cohorts, providing comprehensive information for early diagnosis and treatment strategy. 23 Our study is the first to comprehensively analyze the expression pattern of ANLN and its correlation with patients' prognosis and immune features in 31 different types of cancers.
Our study found that ANLN was upregulated in BLCA, BRCA, CESC, CHOL, COAD, ESCA, HNSC, KICH, KIRC, KIRP, LIHC, LUAD, LUSC, PAAD, PCPG, PRAD, READ, SARC, STAD, THCA, and UCEC. Cox regression and Kaplan-Meier analysis demonstrated that high ANLN expression was associated with shorter OS, DFI, and PFS in KIRP, LIHC, LUAD, and PAAD. This is in line with a recent study that ANLN is a prognostic biomarker for LUAD patients. 24 ANLN is primarily involved in the F I G U R E 7 Association between ANLN expression and the immune and stromal cell ratio in patients' with KIRC, THCA, LUSC, SARC, STAD, UCEC, PRAD, BRCA, LIHC, TGCT, PAAD, and SKCM. p < 0.05 was considered to be significant. process of cytokinesis and plays vital roles in tumorigenesis and progression. 2 ANLN could significantly promote cancer cell growth and migration via targeting miR-217 25 and HMGA2. 26 Upregulation of ANLN could activate RHOA, leading to the activation of phosphatidylinositol 3-kinase-AKT (PI3K/AKT), which promotes cell proliferation and progression. 27 Another study revealed that ANLN might be involved in the metastasis of lung adenocarcinoma by promoting epithelial mesenchymal transformation of tumor cells. 28 In addition, ANLN expression is also positively related to Wnt/β-catenin signaling in gastric cancer. 29 All these biological functions of ANLN may explain its oncogenic role in the above-mentioned cancers. Notably, over-expression of ANLN was correlated with better prognosis in THYM. Future studies are needed to explore its controversial role in human cancer.
TMB is emerging as a promising biomarker for immunotherapy. It refers to the total number of somatic coding mutations, gene insertion, base substitutions, and deletion errors detected per million bases. 30 TMB presents as a useful estimation of tumor neoantigens, which could activate tumor-specific T cells to recognize and kill tumor cells. 30 High TMB are associated with prolonged survival in melanoma, 31 NSCLC, 32 and breast cancer. 33 MSI is a F I G U R E 8 Association between ANLN expression and 47 common immune checkpoints gene levels in pancancer. *p < 0.05, **p < 0.01, ***p < 0.001.

F I G U R E 9 ANLN expression in the role of immunotherapeutic response prediction. (A) Kaplan-Meier curves for patients' OS; (B)
Differences in ANLN expression between non-response and response types; (C) The proportion of patients with response to PD-L1 blockade immunotherapy in low or high ANLN groups. p < 0.05 was considered to be significant. molecular tumor phenotype caused by genomic hypermutability. 34 Recent studies suggested that MSI could serve as an efficient biomarker for immunotherapy. 35 Clinical trials showed that MSI-H patients have better outcomes than microsatellite stable patients, owning to the activation of T lymphocyte recognition of neoantigens. 35,36 Additionally, nearly all patients with MSI-H have high TMB level and could respond more strongly to immunotherapy. 17 Our study found that ANLN expression was positively associated with TMB or MSI in multiple cancers, including BLCA, which was consistent with the results from the BLCA immunotherapy cohort that patients with higher ANLN level had better immune responses and thus demonstrated longer OS. All the above findings suggested that ANLN expression in these tumors may help identify patients who are more likely to benefit from immunotherapy. TME is a complex ecology consisting of tumor cells, tumor vasculature, surrounding immune and inflammatory cells, as well as adipocytes and fibroblasts. 6 Tumor immune microenvironment (iTME), as a core part of the TME, consists of tumor-infiltrating lymphocytes (TILs) and other immune cells, such as macrophages, neutrophils, and dendritic cells. 37 Macrophages is a group of heterogeneous cells in the iTME and is strongly correlated with poor survival in diverse cancers. 38 Tumor-associated macrophages (TAMs) play essential roles in carcinogenesis and progression by providing cytokines and inducing tumor angiogenesis. 39 TAMs could be polarized into two distinct phenotypes: (1) M1 macrophages, which act as an anti-tumor defender; (2) M2 macrophages, which could promote tumor invasion and metastasis. 39 Our study found that ANLN expression was positively associated with M1 macrophage infiltration, but negatively linked to M2 macrophage infiltration. Thus, we assumed that ANLN might participate in the host anti-tumor immune response by inducing M1 macrophages infiltration. However, we also found that ANLN expression was negatively connected to CD8+ T cells in some cancer types. Cytotoxic CD8+ memory T cells in the TME is positively linked to cancer prognosis in diverse cancers. [40][41][42] It could contribute to cancer prognoses by killing tumor cells. 43 This phenomenon could be explained by the existence of other important regulators in the interaction between ANLN and CD8+ T cells. Further studies are needed to investigate the exact mechanism during this anti-tumor progress. Single-cell RNA sequencing (scRNA-seq), which allows for transcriptome-wide analyses of individual cells may help offer new insights in the future. A previous study showed that ANLN along with other regulators are significantly associated with the expression of naïve B cells, regulatory T cells, and neutrophils in head and neck squamous cell carcinoma (HNSCC). 44 However, studies focusing on the functional mechanism of ANLN in immune cell infiltration is rare. Our GO and KEGG pathway analyses revealed that ANLN was mainly enriched in cell cycle, immunity, and metabolic-related pathway, which may explain its oncogenic role in different tumors. Further studies are needed to find the exact functional mechanism between ANLN expression and different immune cells. Moreover, our analyses demonstrated that ANLN expression had a significant correlation with most of the common immune checkpoint markers. We further validated the immunotherapeutic predictive ability of ANLN in an anti-PD-L1 cohort (IMvigor210). Results showed that high ANLN expression was associated with prolonged OS and better immune response, which was consistent with the above definitions. All these above findings confirmed our hypothesis that ANLN could be a robust biomarker for prognosis and immune response prediction.
Several limitations should be considered in our study. First, because of the use of microarray and sequencing data from different databases, systematic bias may occur. Hence, as mentioned above, scRNA-seq or other higher resolution methods could be performed. Second, our study was a retrospective study, so future prospective researches are needed to verify the correlation between ANLN expression and patients' prognosis and tumor immunity. Finally, the precise mechanisms by which ANLN facilitates tumor progression and tumor immunity remain largely obscure. Further mechanistic studies on ANLN at both cellular and molecular levels are needed to fully elucidate its functions.
In conclusion, our study found that ANLN was upregulated in diverse cancers and its aberrant expression was associated with pan-cancer prognosis, MSI, TMB, the tumor immune microenvironment, and immune checkpoint genes. Therefore, ANLN may prove to be a promising biomarker for prognostic judgment and immunotherapy.

AUTHOR CONTRIBUTIONS
JC, QYZ, and XZ conceived and designed this study. LZ, YW, and YH obtained the data. XZ, LZ, YW, and YH analyzed the data. XPW and ZBH helped discuss the results. LZ drafted the manuscript. All authors contributed to the article and approved the submitted version.