Abstract

Background

   CSMD2 has been reported as a potential prognostic factor in several
   cancers. However, whether CSMD2 affects bladder cancer (BC) remains
   unclear.

Methods

   Public data were obtained from the TCGA
   ([33]https://cancergenome.nih.gov) databases. CSMD2expression and its
   prognostic value were analyzed using bioinformatics methods. CSMD2 mRNA
   level in patients with BC and BC cell lines was evaluated via
   quantitative reverse transcriptase polymerase chain reaction. CSMD2
   protein level in patients with BC was evaluated via
   immunohistochemistry. BC cell lines T24 and UMUC-3 were selected for
   loss-of-function assays targeting CSMD2. Cell viability was determined
   by CCK8 and clone formation experiments. Cell migration and invasion
   were evaluated using Transwell assays. Furthermore, the transcriptome
   of UMUC-3 with CSMD2 knockdown was sequenced to analyze potential
   signaling network pathways. Finally, the TIMER2.0 database was employed
   to identify the correlation between CSMD2 and immune cells in the tumor
   microenvironment.

Results

   CSMD2 expression was up-regulated in BC tissues compared to adjacent
   tissues. High CSMD2 expression was associated with poor survival and
   could serve as an independent predictor for survival in patients with
   BC. Furthermore, down-regulation of CSMD2 notably restrained the
   viability, migration, and invasion abilities of T24 and UMUC-3 cells.
   Moreover, transcriptomic sequencing after CSMD2 knockdown in
   UMUC-3 cells revealed its involvement in the regulation of the
   malignant phenotype in BC. Finally, public databases suggest a
   connection between CSMD2 and immune cell infiltration in BC.

Conclusions

   These findings suggest that CSMD2 may promote proliferation and
   tumorigenicity, and could represent a potential target for improving
   the prognosis of BC.

   Keywords: CSMD2, Prognostic marker, Bladder cancer, Biomarker, Immune
   cells

1. INTRODUCTION

   In the United States, bladder cancer (BC) is the second most prevalent
   cancer of the genitourinary system, with an estimated 82,290 new cases
   and 16,710deaths are estimated in 2023 [[34]1]. Urothelial cell
   carcinoma accounts for ∼90 % of BC cases, while ∼5 % are squamous cell
   carcinoma. Other pathological types, such as adenocarcinoma, are rare
   [[35]2]. Of the new cases, 75 % are classified as non-muscle invasive
   BC (NMIBC), while ∼25 % of the new cases involve BC that has invaded
   the bladder muscle layer or deeper, known as muscle-invasive BC (MIBC)
   [[36]3]. After treatment, 10–15 % of patients with NMIBC progress to
   MIBC [[37]4]. Despite the proven benefits of neoadjuvant chemotherapy
   and adjuvant immunotherapy in improving the survival of patients with
   MIBC, their overall 5-year overall survival (OS) rate following surgery
   is ∼50 % [[38]5,[39]6,[40]7]. Hence, thereit is imperativeto identify
   new biomarkers that can pave the way for innovative approaches to
   diagnose and target BC.

   CUB and sushi multiple domains 2 (CSMD2), located on the short arm of
   human chromosome 1 (1p35.1), encodes a single-pass transmembrane
   protein with a large extracellular domain consisting of CUB and Sushi
   domain repeats. CSMD2 is highly expressed in the developing and mature
   brain and has been linked to schizophrenia [[41]8,[42]9]. Previous
   reports have suggested associations between CSMD2 and the progression
   of various cancers, including small-cell lung cancer [[43]10],
   hepatocellular carcinoma [[44]11], colorectal cancer [[45]12],
   pancreatic cancer [[46]13], and melanoma [[47]14]. In a recent
   pan-cancer analysis study, it was revealed that CSMD2 is upregulated in
   most tumors and exhibits moderate to high diagnostic efficiency, with
   high expression closely associated with poor prognosis in patients with
   tumors [[48]15]. However, these findings have not been experimentally
   validated, and the biological functions of CSMD2 in BC remain largely
   unexplored.

   Immunotherapy is currently considered an effective adjuvant therapy for
   cancer treatment, especially for advanced metastatic cancer [[49]16].
   CSMD2 has the potential to participate in regulating the complement
   cascade reaction of the immune system [[50]15], but its role in BC
   remains unknown. This study aims to provide a preliminary exploration
   of the role of CSMD2 in BC, including its expression in BC, its
   predictive value for clinical prognosis, and the relationship between
   CSMD2 expression and anti-tumor immunity as well as immune evasion in
   the tumor microenvironment.

2. Materials and methods

2.1. Bioinformatics analysis

   First, we downloaded RNA-sequencing data for BC from The Cancer.

   Genome Atlas (TCGA; [51]https://cancergenome.nih.gov) database, which
   included 412 BCE samples and 19 adjacent tissues. We then employed R
   packages [[52]17], such as “ggplot” [[53]18], “rms” v6.3-0, and
   “survival” v3.3.1, to analyze the clinical characteristics and the
   prognostic value of CSMD2 in patients with BC within the TCGA database.
   Additionally, we utilized the R package "DESeq2″ v1.26.0 [[54]19] with
   R v4.2.1 to investigate the correlations between CSMD2 mRNA level and
   all genes in BC. Subsequently, we conducted gene set enrichment
   analysis (GSEA) using the “cluster Profiler” package within the MSigDB
   Collections [[55]20,[56]21]. In this analysis, statistically
   significant results were defined as having a normalized enrichment
   score (|NES|) > 1, a false discovery rate (FDR) < 0.25, and an adjusted
   p-value <0.05.

2.2. Clinical samples

   The 20 pairs of adjacent and cancerous tissue specimens, which were
   employed for quantitative real-time polymerase chain reaction (qRT-PCR)
   and immunohistochemistry (IHC) analysis, were obtained from the First
   Affiliated Hospital of Nanchang University (China). The collection and
   utilization of all biological samples in this study were granted
   approval by the Ethics Committee of the First Affiliated Hospital of
   Nanchang University (#[2022] CDYFYYLK [11–031]5), and were conducted in
   compliance with the informed consent of the participating patients.

2.3. Cell culture

   The human BC cell lines (BIU, EJ, T24, 5637, and UMUC-3) and the
   non-malignanturothelial cell line (SV-HUC-1) were generously provided
   by the Institute of Urology at The First Affiliated Hospital of
   Nanchang University. SV-HUC-1 cells were cultivated in F–12K complete
   medium; T24 cells were grown in medium composed of DMEM (Gibco,
   Billings, MT, USA); whereas UMUC3, BIU, and EJ cells were cultured in
   medium containing MEM (Procell, Wuhan, China). Cell line 5637 was
   maintained in medium containing 1640 (Gibco). All of these culture
   media were supplemented with 10 % fetal calf serum, 0.1 mg/mL
   streptomycin, and 100 U/mL penicillin. The cells were incubated at
   37 °C in an atmosphere with 5 % CO[2].

2.4. Cell transfection

   All of the siRNA sequences in S1 were custom-designed and synthesized
   by RiboBio (Guangzhou, China). A day prior to siRNA transfection,
   2·10^5 cells were seeded in the culture medium without antibiotics in
   six-well plates. Following the manufacturer's protocol, once the cell
   concentration had reached 80 %, the Lipofectamine 2000 kit (Invitrogen,
   Carlsbad, CA, USA) was utilized for transfection, and the efficiency of
   gene knockdown was assessed 24–48h later.

2.5. RNA extraction and qRT-PCR

   Total RNA was extracted using the TRIzol reagent (Invitrogen), and cDNA
   was synthesized using the PrimeScript RT Reagent Kit (Transgen,
   Beijing, China). mRNA levels of CSMD2 were assessed using the SYBR
   Premix Ex Taq system (TaKaRa, Dalian, China), and qRT-PCR analysis was
   conducted using the ABI Prism 7500 Fast RT-PCR System (Applied
   Biosystems, Waltham, MA, USA). Relative expression levels were
   calculated using the comparative (2-^ΔΔCT) method and expressed as
   fold-change. The expression of β-actin was utilized as an internal
   control. The specific primer sequences used are provided in
   [57]Supplementary Table S1. Subsequently, three pairs of BC cell RNAs,
   which were transfected with the CSMD2-knockdown small interfering RNAs
   (siRNAs; si-1, si-2,and the negative control si-nc), were extracted and
   sent to OE Biotech (Shanghai, China) for transcriptome sequencing.

2.6. Western blot assays

   For the extraction of total protein lysates, RIPA lysis buffer
   (Beyotime, Nantong, China) containing protease inhibitor was used.
   Protein concentrations were measured using a BCA Protein Assay Kit
   (Thermo Fisher Scientific, Waltham, MA, USA). Protein samples were
   separated by sodium dodecyl sulfate‐polyacrylamide gel electrophoresis.
   Subsequently, the samples were transferred onto a polyvinylidene
   fluoride membrane, which was then blocked with 5 % BSA for 1.5 h. The
   membrane was subsequently probed with primary antibodies, including
   anti‐CSMD2 (NBP2-92141; Novus, Centennial, CO, USA), β-Tubulin
   (#10094-1-AP; Proteintech, Wuhan, China), anti‐phospho‐AKT
   (phosphorylated at Ser473; Abcam, Cambridge, UK), anti‐AKT (#4685; CST,
   Danvers, MA, USA), anti‐PI3K (#3011S; CST), and anti‐phospho‐PI3K
   (phosphorylated at Y607; Abcam). The samples were incubated overnight
   at 4 °C, followed by washing in Tris‐buffered saline buffer with
   Tween‐20 four times for 8 min. Then, horseradish peroxidase-conjugated
   secondary antibodies were applied for 1.5 h. After washing the
   membrane, protein level bands were visualized using an enhanced
   Electro‐Chemi‐Luminescence kit (Beyotime) on a ChemiDoc MP imaging
   system (Bio‐Rad, Hercules, CA, USA). The relative quantity of proteins
   was determined as the ratio of the protein of interest to an internal
   reference.

2.7. IHC analysis

   The IHC procedure and scoring for CSMD2 level were performed following
   previously established methods [[58]22,[59]23]. Briefly, tissue
   sections affixed to glass slides were first deparaffinized and
   rehydrated using xylene and then submerged in EDTA antigen retrieval
   buffer. They were treated with 3 % hydrogen peroxide and subsequently
   incubated with 1 % bovine serum albumin. Subsequently, they were left
   to incubate overnight at 4 °C with an anti-CSMD2 antibody (1:200
   dilution, BS-8187R; Bioss, Beijing, China). Normal goat serum served as
   a negative control. After washing with phosphor-buffered saline
   (PBS)-Tween (PBS + 1 % Tween), a secondary antibody from Zymed (San
   Francisco, CA, USA) was applied, followed by incubation with the
   avidin-biotin peroxidase complex (Zymed). The tissue sections were then
   immersed in 3,3′-diaminobenzidine, counterstained with 10 % Mayer's
   hematoxylin, dehydrated, and mounted on slides.

   Quantification of CSMD2 for each sample was determined by a pathologist
   who was unaware of the clinical and molecular data, using a modified
   H-score formula:

   H-SCORE = ∑（pi × i） = （percentage of weak intensity × 1) + (percentage
   of moderate intensity × 2) + (percentage of strong intensity × 3)
   [[60]24,[61]25].

2.8. CCK-8 cell proliferation assays

   T24 and UMUC-3 cells were seeded in 96-well cell plates at a density of
   3000 cells per well and cultured under standard conditions. Cell
   viability was assessed every 24 h by adding 10 μL CCK-8reagent to each
   well and incubating them at 37 °C for 2 h. Subsequently, the optical
   density value at a wavelength of 450 nm was measured and recorded.

2.9. Plate clone formation assays

   To prepare a cell suspension, exponential T24 and UMUC-3 cells were
   digested using trypsin. This cell suspension, containing 800 cells per
   dish, was then transferred to 6-well cell plates, each containing 3 mL
   pre-culture medium. The cells were cultured at 37 °C with 5 % CO[2] for
   1–2 weeks, and the culture process was concluded once visible colonies
   had formed. After rinsing with PBS, the cells were fixed with 4 %
   paraformaldehyde for 20 min and stained with 0.1 % crystal violet for
   20 min. Finally, the size and number of colonies were evaluated.

2.10. Transwell chambers assays

   In the cell invasion and migration assays, 8 μm pore-size Transwell
   inserts were employed. For the invasion assay, following a 24-h
   transfection period, 2.10^5 cells were resuspended in 200 μL serum-free
   medium and placed into the upper chamber of Transwell plates, which
   were pre-coated with Matrigel diluted at 1:8. The lower chamber was
   filled with 600 μL medium containing 20 % fetal bovine serum. After
   24 h incubation, the cells in the lower chamber were collected, fixed,
   and stained. The experimental steps for the Transwell migration assay
   are similar to the invasion assays, except that Matrigel is not used in
   the migration assay.

2.11. Statistical analysis

   Differential expression analysis was performed using the Wilcoxon
   signed-rank test in GraphPad Prism v9.2 (GraphPad Software, La Jolla,
   CA, USA) for statistical computations. For sequencing data and mining
   of datasets, such as TCGA, automated statistical analysis was performed
   using statistical packages in the R programming language.

3. RESULTS

3.1. CSMD2 is up-regulated in BC

   First, we analyzed CSMD2 expression using the TCGA database. As shown
   in [62]Fig. 1A and B, CSMD2 exhibits significantly higher expression
   levels in BC tissues in comparison to normal tissues (p < 0.01).
   Furthermore, the data from our 20 pairs of adjacent and cancerous
   tissues revealed that the CSMD2 mRNA level ([63]Fig. 1C) and protein
   expression ([64]Fig. 2A–E) are notably higher in the BC group when
   compared to the normal group. These findings suggest that CSMD2 is
   up-regulated in BC, and may be associated with the progression of this
   condition.

Fig. 1.

   [65]Fig. 1
   [66]Open in a new tab

   CSMD2 mRNA level is upregulated in bladder cancer. (A–B) Analysis of
   CSMD2 expression in unpaired and paired bladder cancer tissues from the
   TCGA database. (C) CSMD2 mRNA level in 20 paired bladder tumor tissues
   and their adjacent normal tissues. (D) Real-time PCR analysis of CSMD2
   expression in SV-HCV-1 and bladder cancer cell lines. *p < 0.05;
   **p < 0.01; ***p < 0.001; ****p < 0.0001.

Fig. 2.

   [67]Fig. 2
   [68]Open in a new tab

   CSMD2 protein level is upregulated in bladder cancer tissues. (A–D)
   Representative images of CSMD2 protein level in normal bladder tissues
   and bladder cancer (upper: magnification × 200; lower:
   magnification × 400). (C) CSMD2 protein level in 20 paired bladder
   tumor tissues and their adjacent normal tissues.

3.2. CSMD2 expression in BC cell lines in vitro

   To assess mRNA and protein levels of CSMD2 in vitro, we selected two BC
   cell lines, i.e., T24 and UMUC-3, with the normal bladder cell line
   SV-HUC-1 serving as the control. Real-time PCR analysis indicated that
   CSMD2 mRNA level was upregulated in BC cells when compared to the
   control cells ([69]Fig. 1D). We selected the T24 and UMUC-3 cells that
   exhibited high CSMD2 expression to conduct the knockdown experiments.

3.3. Clinical value of CSMD2

   We used a chi-squared test to determine the associations between CSMD2
   expression and clinical symptoms in patients with BC. As illustrated in
   [70]Supplementary Table S2, the expression of CSMD2 exhibited positive
   correlations with pathologic T stage, pathologic stage, and histologic
   grade (p < 0.01). Interestingly, univariate analysis revealed that
   CSMD2 expression, TNM stage, T stage, N stage, M stage, and patient age
   had statistically significant prognostic value for overall survival
   (OS; [71]Table 1). Additionally, multivariate analysis demonstrated
   that CSMD2 expression and N stage are independent prognostic factors
   for the OS of patients with BC ([72]Table 1). Moreover, we utilized a
   Kaplan–Meier plot to directly visualize the relationship between CSMD2
   expression and OS. As depicted in [73]Fig. 3A, the results suggested
   that patients with high CSMD2 expression had a poorer OS rate compared
   to patients with low CSMD2 expression (p = 0.024). These findings
   indicate that CSMD2 could serve as an independent predictor for
   patients with BC. Patients with higher CSMD2 expression patients were
   more likely to have higher pathologic stage ([74]Fig. 3B), higher T
   stage ([75]Fig. 3C), higher pathologic N stage ([76]Fig. 3D), and
   higher histologic grade (p < 0.0001; [77]Fig. 3E). Furthermore, we
   utilized the R packages “rms” and “survival” to construct a nomogram
   that incorporated pathologic T, N, and M stages, age, gender, and CSMD2
   expression of patients with BC into a nomogram to predict patient
   survival outcomes at 1, 3, and 5 years ([78]Fig. 3F). The nomogram
   exhibited strong predictive performance for OS rates in BC, and
   calibration plots confirmed its reliability in predicting model
   performance ([79]Fig. 3G).

Table 1.

   Univariate and multivariate analysis of prognostic parameters in
   bladder cancer.
   Characteristics Univariate analysis
     __________________________________________________________________

   Multivariate analysis
     __________________________________________________________________

   p-value HR 95 % CI p-value HR 95 % CI
   CSMD2 (low/high) 0.039 1.37 1.02–1.84 0.003 2.191 1.317–3.642
   TNM stage (I + II/III + IV) <0.001 2.267 1.567–3.281 0.265 0.512
   0.158–1.660
   T stage (T1 + T2/T3 + T4) <0.001 2.157 1.485–3.132 0.081 2.602
   0.889–7.609
   N stage (N0/N1 + N2 + N3) <0.001 2.25 1.649–3.072 0.008 2.065
   1.205–3.538
   M stage (M0/M1) 0.002 3.112 1.491–6.493 0.648 1.25 0.479–3.262
   Age (≤70/>70 years) 0.018 1.424 1.064–1.906 0.45 1.199 0.748–1.922
   Gender (female/male) 0.39 0.868 0.629–1.198
   [80]Open in a new tab

   I, II, III, and IV represent different stages.

   T, size or direct extent of the primary tumor; N, degree of spread to
   regional lymph nodes; M, presence of distant metastasis; CI, confidence
   interval; HR, hazard ratio.

Fig. 3.

   [81]Fig. 3
   [82]Open in a new tab

   The clinical value of CSMD2 in bladder cancer. (A) Association between
   CSMD2 mRNA level and overall survival in patients with bladder cancer
   in the TCGA database (p < 0.05). (B–E) Associations between CSMD2
   expression and pathologic stage, T stage, N stage, and histologic grade
   in the TCGA database. (F–G) Nomogram depicting the impact of CSMD2
   expression on outcomes of patients with bladder cancer and performance
   calibration graph. *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.

3.4. Silencing CSMD2 inhibits BC progression

   We used qRT-PCR to detect the efficiency of CSMD2 knockdown in BC
   cells. Following transfection with the siRNAs si-1 and si-2, the data
   clearly demonstrated a significant downregulation of CSMD2 mRNA levels
   when compared to the si-nc groups in T24 and UMUC-3 cells (p < 0.01;
   [83]Fig. 4A). The results of CCK-8 assays ([84]Fig. 4B and C) indicated
   that the optical density values in the si-1 and si-2 groups were
   significantly lower than those in the si-nc group (p < 0.01). In clone
   formation assays, it was observed that the number of visible colonies
   in the si-1 and si-2 groups was significantly fewer compared to the
   si-nc groups in T24 and UMUC-3 cells (p < 0.01; [85]Fig. 4D and E). The
   findings from invasion and migration assays demonstrated that, in both
   T24 and UMUC-3 cells, the numbers of cells involved in invasion and
   migration were significantly reduced in the si-1 and si-2 groups
   compared to the si-nc group (p < 0.01; [86]Fig. 4F–H).

Fig. 4.

   [87]Fig. 4
   [88]Open in a new tab

   Knockdown of CSMD2 inhibits the tumorigenicity of bladder cancer. (A)
   Knockdown efficiency of siRNA targeting CSMD2 in the bladder cancer
   cell lines UMUC-3 and T24. nc, negative control. (B–C) CCK-8 assays
   reveal significantly reduced cell proliferation in the si-csmd2-1 and
   si-csmd2-2 groups compared to the si-nc group in bladder cancer cells
   (**p < 0.01). (D–E) Clone formation assays demonstrate a remarkable
   decrease in cell clone formation efficiency in the si-csmd2-1 and
   si-csmd2-2 groups compared to the si-nc group in bladder cancer cells
   (**p < 0.01). (F–H) Transwell assays indicate that fewer UMUC-3 and
   T24 cells migrate and invade in the si-csmd2-1 and si-csmd2-2 groups
   compared to the si-nc group (**p < 0.01).

3.5. CSMD2 expression affects the PI3K/AKT signaling pathway

   To determine which pathway might be involved in CSMD2-mediated BC
   progression, we conducted GSEA analysis in the TCGA BC dataset
   (n = 412). As shown in [89]Fig. 5A, our analysis revealed a correlation
   between high CSMD2 expression and the PI3K/AKT pathway. The
   phosphorylation of PI3K and AKT decreased in CSMD2-silenced cells.
   However, it's noteworthy that the down-‐regulation of CSMD2 did not
   significantly impact the protein levels of PI3K and AKT ([90]Fig. 5B).
   These results collectively suggest that the downregulation of CSMD2 is
   associated with decreased BC cell proliferation and invasive capacities
   through the regulation of the PI3K/AKT signaling pathway ([91]Fig. 5C).

Fig. 5.

   [92]Fig. 5
   [93]Open in a new tab

   Down‐regulation of CSMD2 and its association with p‐PI3K and p‐AKT. (A)
   GSEA plot illustrating the positive correlation between CSMD2
   expression and PI3K-AKT gene signatures. (B) Western blotting shows
   that the knockdown of CSMD2 in UMUC-3 and T24 cells significantly
   reduces p‐PI3K and p‐AKT levels compared to the si‐1, si-2, and si‐NC
   groups (Full non-adjusted images of WB shown in S4). (C) The
   downregulation of CSMD2 leads to a decrease in phosphorylation of PI3K
   and AKT, resulting in reduced proliferation, migration, and invasive
   abilities of bladder cancer cells.

3.6. Sequencing analysis after knockdown of CSMD2

   We extracted total RNA from three groups of UMUC-3 cells: the CSMD2
   knockdown group (si-csmd2-1), the negative control group (NC), and a
   control group with no treatment. Transcriptional profiling analysis was
   performed using RNA sequencing to gain comprehensive insights into the
   biological functions of CSMD2 in BC. [94]Supplementary Fig. S3Adisplays
   a clustered heatmap illustrating differentially expressed genes between
   the two groups. Furthermore, a volcano plot depicting the distribution
   of differentially expressed genes between the two groups is shown in
   [95]Supplementary Fig. S3B. Compared to the control group, the CSMD2
   knockdown group exhibited significant upregulation of 615 genes and
   significant downregulation of 455 genes. The CSMD2 gene appears to be
   implicated in immune response through multiple pathways and is
   associated with biological processes such as cell adhesion and growth.
   These insights were derived from gene ontology (GO) and Kyoto
   Encyclopedia of Genes and Genomes (KEGG) analyses performed on the
   differentially expressed genes, as indicated in [96]Fig. 6A and B.
   [97]Fig. 6C shows the protein–protein interaction (PPI) network
   analysis of differential gene expression between the two groups based
   on sequencing data.

Fig. 6.

   [98]Fig. 6
   [99]Open in a new tab

   Analysis of CSMD2 knockdown sequencing results. (A–B) GO and KEGG
   enrichment analysis plots for genes that differ between the two groups.
   BP, biological process; CC, cellular component; MF, molecular function.
   (C) PPI network analysis of the 30 most differentially expressed genes
   from sequencing data. (E) Pathway enrichment analysis comparing CSMD2
   knockdown to control cells. NC, negative control.

   Moreover, according to the GSEA results shown in [100]Fig. 6D, the
   genes in the NC group are primarily enriched in DNA response, DNA
   replication, cell cycle, transcriptional misregulation in cancer, and
   other related signaling pathways. This suggests that CSMD2 may regulate
   tumor development by modulating DNA alterations and the cell cycle.

3.7. Correlation of CSMD2 with immunomarkers in BC

   The positive correlation observed with the infiltration of immune
   evasion-related cells, such as Tregs and cancer-associated fibroblasts
   ([101]Fig. 7A and B) suggests that CSMD2 may be associated with immune
   evasion. We extended our analysis by comparing the infiltration of
   various immune cell subtypes in patients with BC with varying copy
   numbers of the CSMD2 gene, using the “SCNA” module of the TIMER
   database. As shown in [102]Fig. 7C, statistically significant
   differences were noted in the changes in infiltration of CD4^+ T cells,
   macrophages, neutrophils, and dendritic cells across various copy
   number states. This implies that the copy number of CSMD2 might
   influence the infiltration of immune cells. Furthermore, we evaluated
   the impact of immune cell infiltration on the clinical outcomes of
   patients with BC using the TIMER platform
   ([103]http://timer.cistrome.org/). Our findings revealed that lower
   levels of CD8^+ T cells were associated with poorer prognosis in
   patients with BC (p < 0.05; [104]Fig. 7D).

Fig. 7.

   [105]Fig. 7
   [106]Open in a new tab

   CSMD2's association with the immune microenvironment. (A–B)
   Correlations of CSMD2 expression with cancer-associated fibroblasts and
   Tregs based on XCELL and QUANTISEQ. Correlations are depicted with
   partial Spearman's correlation, and statistical significance is based
   on TIMER2.0. (C) Changes in the infiltration of immune subpopulations
   at each copy number status of CSMD2. (D) Clinical survival outcomes of
   patients with bladder cancer in the CD8^+ T cells group.

4. Discussion

   Despite advancements in surgical and anesthesia techniques, as well as
   the widespread use of perioperative chemotherapy, the long-term
   survival rate of patients with BC has remained stagnant for decades
   [[107]6,[108]7]. Simultaneously, through advanced molecular research,
   our understanding of disease biology has significantly expanded. A
   recent study on the prognosis of patients with occult hepatitis B virus
   infection indicated that the development of cirrhosis and subsequent
   hepatocellular carcinoma may be associated with CSMD2 [[109]11].
   Another recent report on a pan-cancer study emphasized that CSMD2 is
   upregulated in most tumors and demonstrates moderate to high diagnostic
   efficiency. High expression of CSMD2 is also closely linked to a poor
   prognosis in cancer patients [[110]15]. However, the additional
   functions of CSMD2 and its mechanisms in cancers have yet to be fully
   elucidated.

   Through our research, we identified abnormal expression of CSMD2 in BC
   using various online resources, datasets, and tissue samples, as well
   as cell lines that we collected. By integrating this data with clinical
   and pathological parameters, we determined that high CSMD2 expression
   is associated with a poorer prognosis in patients with BC, and that the
   expression of CSMD2 holds significant diagnostic value. To further
   validate our bioinformatics analysis results, we used siRNAs to knock
   down CSMD2 and experimentally confirmed its impact on the
   proliferation, migration, and invasion of BC cell lines. Additionally,
   we conducted transcriptomic sequencing to analyze the differentially
   expressed genes and relevant signaling pathway networks associated with
   the immune microenvironment in the CSMD2 knockdown group compared to
   the control group.

   The PI3K/Akt signaling pathway is an intracellular signaling pathway
   that plays a pivotal role in various cellular processes, including cell
   cycle progression, cell proliferation, survival, lifespan, and even
   cancer. Akt, as the central mediator of the PI3K pathway, is a
   serine/threonine kinase that plays a crucial role in multiple cellular
   processes such as cell proliferation, apoptosis, cell migration,
   glucose metabolism, and transcription [[111]26]. The JAK/STAT signaling
   pathway is involved in diverse cellular processes, including immune
   response, cell growth, differentiation, inflammation, and
   hematopoiesis. Dysregulation of this pathway can contribute to the
   development of various diseases, making it an important target for
   therapeutic interventions [[112]27]. In our study, we observed that the
   knockdown of CSMD2 led to significant alterations in the PI3K/AKT and
   JAK/STAT pathways ([113]Fig. 5B), suggesting that CSMD2 may primarily
   regulate tumor functional changes in tumors through the modulation of
   these two signaling pathways. However, further experimental validation
   is required to confirm this hypothesis.

   The tumor microenvironment encompasses the tissues surrounding a tumor,
   including the extracellular matrix, immune cells, and stromal cells
   [[114]28]. In recent years, the tumor microenvironment, particularly
   the immune microenvironment, has received significant attention in
   research. Within the tumor immune microenvironment, various molecules
   and cells can influence tumor development and treatment outcomes
   [[115]29]. CD8^+ T cells play critical roles in immune responses,
   including cytotoxic activity, immune memory, cytokine production,
   regulation of other immune cells, and immune modulation, making them
   vital for anti-tumor immunity. Tregs, on the other hand, exert negative
   regulation on CD8^+ T cells, contributing to immune tolerance, immune
   homeostasis, and the prevention of excessive immune responses
   [[116]30]. In our study, we identified a positive correlation between
   CSMD2 and Tregs infiltration. Additionally, we observed that patients
   with low CD8^+ T cell infiltration had worse prognoses. These findings
   suggest that CSMD2 may regulate Tregs and subsequently impact CD8^+ T
   cells. We also noted a positive correlation between CSMD2 expression
   and immune escape-related tumor-associated fibroblasts. Therefore, we
   hypothesize that CSMD2 may promote tumor progression through immune
   evasion or immune suppression. However, further experimental validation
   is needed to confirm these observations.

   Although we conducted a comprehensive analysis of CSMD2 in BC, there
   are still certain limitations to be addressed. First, we sourced the
   transcriptomic data and clinicopathological information from public
   databases, which introduces data heterogeneity, platform differences,
   and potential gaps in clinical treatment information. Second, the
   immune-related mechanisms of CSMD2 in BC are based on bioinformatics
   predictions and require further experimental validation. In the future,
   we aim to address these limitations and further validate the
   bioinformatics predictions.

   In conclusion, our study demonstrates that CSMD2 is highly expressed in
   BC and is associated with poor prognosis. In addition, we have shown
   for the first time that the downregulation of CSMD2 inhibits
   proliferation, invasion, and migration. These findings suggest that
   CSMD2 may be involved in the development of BC by influencing cell
   proliferation, migration, and the immune microenvironment. Our study
   identifies the pathogenic role of CSMD2 and provides a potential
   therapeutic target for patients with BC.

5. Consent for publication

   All authors consent to the publication of this study.

Data availability

   Publicly available datasets were analyzed in this study. These data can
   be found freely from the TCGA data portal
   ([117]https://portal.gdc.cancer.gov/). The RNA sequencing data
   (BioProject accession number PRJNA 1030890) reported in this paper have
   been deposited into Sequence Read Archive (SRA) database. These data
   will be available in the date of publication.

Funding

   This study was supported by the National Natural Science Foundation of
   P.R. China (Grants #81560419, #81960512, and #81760457), Key Project of
   Natural Science Foundation of Jiangxi Province (#20212ACB206013), and
   Youth Project of Natural Science Foundation of Jiangxi Province
   (#20212BAB216037).

CRediT authorship contribution statement

   Zhijun Yao: Writing – original draft. Hailang Yang: Visualization,
   Software. Xiaoqiang Liu: Investigation, Funding acquisition. Ming
   Jiang: Visualization. Wen Deng: Visualization, Software,
   Conceptualization. Bin Fu: Writing – review & editing, Funding
   acquisition.

Declaration of competing interest

   The authors declare that they have no known competing financial
   interests or personal relationships that could have appeared to
   influence the work reported in this paper.

Acknowledgements