Identification of homer protein homolog 3 as a prognostic marker of colon adenocarcinoma

Abstract

Background

Homer protein homolog 3 (HOMER3), a factor implicated in both physiological and pathological processes, has been studied extensively to determine the relationship between its expression level and the prognosis of various malignancies. However, the significance and clinicopathological role of HOMER3 in colorectal adenocarcinoma remain unclear.

Methods

In this study, bioinformatics techniques were used to find the correlation between high HOMER3 expression levels and clinicopathological features of colorectal adenocarcinoma (COAD) patients.

Results

Cellular experiments confirmed the differential expression of HOMER3 in tumor cells compared to normal cells. HOMER3 overexpression was significantly associated with COAD staging and carcinoembryonic antigen (CEA) levels. Patients with high HOMER3 expression levels have a poor prognosis. HOMER3 expression levels can be distinguished more accurately between tumor and non-tumor tissues (AUC = 0.634). The HOMER3 gene variation rate in COAD tissue was 0.7 %. Moreover, 16 of the 22 DNA methylation sites in HOMER3 were associated with COAD prognosis. Our findings confirmed that HOMER3 was positively correlated with immune cell infiltration and immune checkpoints (PD-1, CTLA-4, LMTK3, and LAG3) in COAD, Specifically, we will clearly state that while there is statistical significance, the actual strength of the correlations is weak. During KEGG enrichment analysis, HOMER3 was enriched along with DLG4 and SHANK1 in glutamatergic synapses. Additionally, upstream microRNAs that could bind to HOMER3 were predicted. These findings suggest that HOMER3 might be involved in COAD development and immune regulation.

Conclusions

HOMER3 acts as a potential biomarker that can facilitate innovative developments in the diagnosis and prognostic assessment of COAD.

Highlights

• •In the present study, the potential of HOMER3 in the diagnosis and prognosis of COAD has been revealed to us.
• •And, we also correlated HOMER3 gene methylation with the prognosis of COAD. HOMER3 may be involved not only in the development of COAD but also may be involved in immune regulation.

Abbreviations

HOMER3

Homer protein homolog 3

COAD

Colorectal adenocarcinoma

CRCC

Olorectal cancer

CEAC

Arcinoembryonic antigen

AML

Acute myeloid leukemia

ESCC

Esophageal squamous cell carcinoma

ROC

Receiver operating characteristic

1. Introduction

Colorectal cancer (CRC) is the second most common cause of cancer deaths in the United States and has caused widespread devastation globally [1]. Colon adenocarcinoma (COAD), the most commonly diagnosed histologic subtype of CRC, often results in patients having an inferior quality of life [2,3]. Despite advances in treatments, including targeted therapies, COAD patients still receive a poor prognosis [4]. COAD development often involves various molecules and complex signaling pathways both in vivo and in vitro [4]. Recently, a growing number of researchers have focused on expressed genes abnormally in tumor tissues to identify potential genes for the early diagnosis and prognostic assessment of cancer [5,6].

The HOMER protein family, also known as the cytoplasmic scaffold protein family, was first reported as a new class of dendritic proteins that mediates a new cellular mechanism for regulating metabolic glutamate signal transduction [7]. Continued research has led to the systematic classification of the HOMER protein family into three members (homer protein homolog 1–3), each with multiple subtypes resulting from alternative splicing [8]. Homer protein homolog 3 (HOMER3), as a member of the HOMER family, was reported to be overexpressed in acute myeloid leukemia (AML) samples in 2008 [9]. Subsequently, HOMER3 was proven to be involved in the growth and differentiation of the leukemic cell line [10]. Abnormal HOMER3 expression has been reported in esophageal squamous cell carcinoma (ESCC), indicating its tumorigenic role in this cancer type [11]. In addition, HOMER3 occurs in gliomas, where the tumor-promoting effect caused by the up-regulation of WW domain-binding protein 2 (WBP2) might be mediated by the interaction between α-enolase (ENO1), HOMER3, and WBP2 [7]. Recently, HOMER3 has been reported as a biomarker in breast, bladder, and hepatocellular carcinoma [12,13], indicating its potential therapeutic significance across various tumors. However, the mechanism of action of HOMER3 in human CRC, particularly COAD, remains unclear.

In this study, we explored the role of HOMER3 as a biomarker in the diagnosis and prognosis of COAD and examined its correlation with immune infiltration and checkpoints. HOMER3 mutations and DNA methylation levels in COAD were assessed. In addition, the potential pathways and upstream miRNAs of HOMER3 involved in COAD onset and development were predicted.

2. Materials and methods

2.1. Comparison of HOMER3 expression levels

TIMER2.0 (http://timer.cistrome.org/) is a comprehensive resource for the analysis of immune infiltration across different cancer types [14,15]. This article primarily uses it to study the differential expression of HOMER3 between tumor and adjacent normal tissues. Additionally, TIMER was utilized to reveal the correlation between HOMER3 and immune cell infiltration (including CD8⁺ T cells, macrophages, neutrophils, and dendritic cells) in COAD. Gene expression levels were displayed using log2 TPM, and the analysis was based on the TCGA-UCEC dataset (N = 562). Data from 480 COAD tissues, 41 non-cancerous colon tissues, and 41 paired non-cancerous RNA-seq gene expression profiles and clinical data were retrieved and downloaded from the TCGA public database (https://www.cancer.gov/). The data were normalized, and an expression matrix for HOMER3 in COAD tissues was plotted.

2.2. Clinical correlation analysis of HOMER3 and COAD

Extract critical clinical information from COAD patients, excluding samples with missing values and those that are not evaluable, and use the "ggplot2" package in R software to analyze the correlation between HOMER3 and clinicopathological features. The clinical data include age, gender, clinical TNM staging, pathological staging, CEA levels, survival time, and survival status. The TCGA prognosis profiles were obtained from the UCSC database. All malignant tumors with fewer than 10 samples were excluded. The relationship between HOMER3 expression in various cancer species and prognosis was studied. The Survminer R package was used for proportional hazards hypothesis testing (log-rank test) and fitting survival regression. The analysis results of OS (overall survival), DSS (disease-specific survival), DFI (disease-free interval), and PFI (progression-free interval) were generated with Kaplan-Meier plots using the survival package. We used the "survminer" and "ggplot2" packages, along with R (pROC package, timeROC package, and survival package) to generate and analyze diagnostic receiver operating characteristic (ROC) curves, time-related diagnostic curves, and nomogram models. Risk ratios (HR) and 95 % confidence intervals were calculated using univariate survival analysis.

2.3. HOMER3 gene alterations in COAD patients

HOMER3 genome profiles in cBioPortal [16,17] were based on two datasets, i.e., “TCGA (https://xenabrowser.net/datapages/?dataset=TCGA), COAD Firehose Legacy”, and “DFCI, CellReports2016”. Kaplan-Meier plots were generated, differences in survival curves were tested on a logarithmic scale, and differences (P < 0.05) were found to be statistically significant.

2.4. DNA methylation of HOMER3

HOMER3 DNA methylation sites were analyzed using the SurvivalMeth database [18]. Moreover, the prognostic value of HOMER3 methylation levels was assessed by determining overall survival (OS).

2.5. Correlation analysis of HOMER3 levels with immune infiltration and checkpoints

The correlation between HOMER3 and immune infiltration and cell markers in COAD tissues was carried out using the Tumor Immune Estimation Resource (TIMER, https://cistrome.shinyapps.io/timer/) [16,17]. We further explored the relationship between the expression of HEMOR3 in COAD and the level of immune cell infiltration. Recently, a method called ESTIMATE has been used to predict the infiltration of non-tumor cells, which calculates an immunological score based on certain gene expression patterns of immune cells. In this study, the ESTIMATE R package (version 1.0.13) was used to test the stromal score, immune score, and ESTIMATE score derived from gene expression for each patient in each tumor type. The Wilcoxon test was employed for the differential analysis of immune cell content and immune cells with p < 0.05 were screened out. Secondly, using the "limma", "ggplot2", "ggpubr", and "ggExtra" packages, a Spearman test was conducted to analyze the correlation between the content of immune cells and the expression level of HOMER3 mRNA. Immune cells with p < 0.05 were selected to obtain HOMER3-related immune cells.

2.6. Gene ontology and Kyoto encyclopedia of genes and genomes analyses

HOMER3 protein-protein interaction networks (PPIs) were downloaded from the STRING database (STRING: functional protein association networks (string-db.org)). The top 10 functional pairs of genes were selected by assessing protein interaction scores for gene ontology (GO) enrichment and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis to study the function of HOMER3.

2.7. Cell lines and cell culture

Five colorectal cancer cell lines (SW480, HT-29, LOVO, SW620, and HCT116) and healthy colon epithelial cells (NCM460, CCD841, and FHC) were purchased from the cell bank at the Shanghai Institute of Biological Sciences (Shanghai, China). SW480, FHC, CCD-841, and HCT116 cells were grown in DMEM medium (MA0212, Meilunbio, China) with 10 % FBS (PWL001, Meilunbio, China) and 1 % penicillin/streptomycin (MA0110, Meilun, China), whereas HT-29, LOVO, SW620, and NCM460 were cultured in RPMI-1640 medium (MA0215, Meilunbio, China) with 10 % FBS and 1 % penicillin/streptomycin. All cells were grown at 37 °C in a humidified atmosphere with 5 % CO₂.

2.8. Quantitative real-time polymerase chain reactions (qRT -PCR)

Total RNA was extracted from cell lines using TRIzol reagent (Vazyme Biotech Co., Ltd.) in accordance with the manufacturer's instructions. RNA transcription was performed for cDNA synthesis using HiScript RT SuperMix for qPCR (Vazyme Biotech Co., Ltd.) and SYBR Green (Vazyme Biotech Co., Ltd.), and generated products were analyzed via real-time quantitative PCR (qRT-PCR). The qRT-PCR procedure was performed at 95 °C for 10 min, followed by 40 cycles at 95 °C for 10 s and 60 °C for 30 s. HOMER3 expression levels were normalized to those of GAPDH, and relative expression levels were calculated using the 2^−ΔΔCt method. The sequences of primers and siRNAs used for qPCR were as follows: HOMER3 forward primer, 5ʹ-CGCACTCACTGTCTCCTATT-3ʹ; HOMER3 reverse primer, 5ʹ-GGAACTTCTCGGCAAACT-3ʹ; SANIL forward primer, 5ʹ-CCCCAATCGGAACCTAACT-3ʹ; SANIL reverse primer, GACAGAGTCCCAGATGAGCA; VIM forward primer, GAGAACTTTGCCTTGAAGC; VIM reverse primer, TCCAGCAGCTTCCTGTAGGT; N-Cad forward primer, CAGAGAGTCGCCAAATGTCA; N-Cad reverse primer, TTCACAAGTCTCGGCCTCTT; E-Cad forward primer, TTGAGTGTCCTGCACAGAGG; E-Cad reverse primer, GAGGGAGCTGAGTGAACCTG; siHOMER3 forward primer, 5ʹ- GGACAUGACCAUCCAUGAATT-3ʹ; siHOMER3 reverse primers, 5ʹ- UUCAUGGAUGGUCAUGUCCTT-3ʹ; GAPDH forward primer, 5ʹ-CTGACTTCAACAGCGACACC-3ʹ; GAPDH reverse primer, 5ʹ-TGCTGTAGCCAAATTCGTTGT-3ʹ.

2.9. Silencing of HOMER3

Silencing experiments targeting the HOMER3 gene were performed using short RNAs (siRNA), and transfection was performed using Lipofectamine 2000 (Invitrogen, USA) reagents. The medium was replaced with complete medium 6 h after transfection. Cells were either reset or used for RNA extraction in subsequent experiments 48 h after siRNA transfection. After confirming the impact of effective siRNA interference on the HOMER3 gene in three replicated studies, siRNA was employed in additional tests.

2.10. Wound healing test

The transfected cells were resuspended, counted, and re-inoculated into 6-well plates with 1 × 10⁶ cells per well. After 24 h, the cell monolayer was scratched and photographed. Subsequently, cell migration was observed and recorded under the microscope every 24 h.

2.11. Transwell test

Matrix glue (BD Biosciences, MA) was added to the upper chamber containing the invasion group. After 30 min in the incubator, the excess matrix glue solution was removed. Transfected cells were resuspended in serum-free medium, adjusted to a concentration of 1 × 10^{^6} cells/mL, and 200 μL of the cell suspension was added to the upper chamber. Additionally, 600 μL complete medium was added to the lower chamber. After 24 h, the upper chamber was removed, and excess cells and matrix glue in the upper chamber were wiped off. It was washed twice with PBS, fixed with paraformaldehyde for 10 min, stained with 0.1 % crystal violet (C8470, Solarbio, China) for 30 min, and finally observed and photographed under a microscope.

2.12. Cell proliferation assays

Cell proliferation was measured using the CCK-8 assay. Briefly, 2 × 10⁴ cells were inoculated into a 96-well plate and cultured for 24, 48, and 72 h. Next, 10 μL/well of CCK-8 solution (Japanese Tonkindo) was added to each well, and the plate was incubated at 37 °C for 3 h. Absorbance was measured at 450 nm using a Bio-Rad microplate reader (USA).

2.13. Western blot assay

RIPA reagent (MA0151, Meilun, China) was used to extract proteins from cells, while BCA reagent was used to determine protein concentrations. Subsequently, 10 % SDS-PAGE was chosen based on the band size of the intended protein. The protein was subsequently transferred to a PVDF membrane and blocked for 1 h and 30 min at room temperature with 5 % milk. GAPDH (10494-1-AP, PTG, 1:20000), CHDH (17356-1-AP, PTG, 1:5000), E-cad (ab231303, Abcam, 1:10000), N-cad (ab76011, Abcam, 1:1000), VIM (ab92547, Abcam, 1:5000), HOMER3 (ab97438, Abcam, 1:1000), and SNAIL (ab180714, Abcam, 1:1000) primary antibodies were incubated with the protein overnight at 4 °C. The membrane was then treated for 1 h at room temperature with an anti-mouse secondary antibody and anti-rabbit secondary antibody, and results were analyzed using an ECL kit and ImageJ software. All Western Blot results were confirmed by repeating the experiment three times to ensure reproducibility and consistency.

2.14. Statistical analysis

This experiment used R language packages (version 3.6.3) and GraphPad Prism software (version 6.02) for statistical analysis of the data. Statistical analyses were automatically performed by the corresponding packages, with the significance level α for all hypothesis tests set at 0.05. Statistical significance was specifically defined as follows: ① During survival gene screening, P < 0.05 was considered statistically significant; ② For univariate and multivariate Cox regression analyses (including tumor stage, tumor grade, and survival time), P < 0.05 was considered statistically significant, with P < 0.01 as the threshold for selection in this experiment; ③ For the analysis of clinical characteristic correlations, P < 0.05 was considered statistically significant; ④ For differential expression analysis, a fold change (logFC) > 1 and an adjusted P-value (P-adjust) < 0.05 were considered statistically significant; ⑤ For GO enrichment analysis based on R language, P < 0.05 and an adjusted P-value (q-value) < 0.05 were considered statistically significant; ⑥ For pathway enrichment analysis based on GSEA, FDR q < 0.25 and NOMP <0.05 were considered statistically significant; ⑦ For immune cell infiltration analysis, P < 0.05 was considered statistically significant. Differences between groups were compared based on data type using the Wilcoxon rank-sum test or Student's t-test. Correlations were determined using Pearson or Spearman correlation tests. The differences between survival curves were assessed using the log-rank test and were found to be (P < 0.05) statistically significant.

3. Results

3.1. High HOMER3 expression levels in COAD patients

HOMER3 expression was analyzed in 20 human cancers. HOMER3 was significantly upregulated in 17 cancer types, including bladder and urethral carcinoma, breast invasive carcinoma, and COAD, when compared to healthy tissues (Fig. 1A). Elevated HOMER3 expression was observed in both non-matching and pairwise comparisons(Fig. 1B and C).

Fig. 1.

Overexpression of HOMER3 in COAD. Based on the TCGA dataset, HOMER3 expression levels were examined in 20 types of tumor tissues and non-tumor tissues (A), 480 COAD tissues and 41 healthy colon tissues (B), and 41 COAD tissues and their paired adjacent healthy colon tissues (C); HOMER3 expression levels were higher in COAD tissues.***P < 0.001, **P < 0.01, *P < 0.05; ns, no statistical difference.

3.2. Correlation of HOMER3 expression with the clinical staging and prognosis of COAD patients

HOMER3 expression was significantly correlated with the T (Fig. 2A), N (Fig. 2B), M (Fig. 2C), and pathologic (Fig. 2D) stages. Patients with advanced-stage disease had higher HOMER3 levels. HOMER3 expression was lower in older patients (Fig. 2E), and high HOMER3 expression corresponded to higher CEA levels (Fig. 2F). Kaplan-Meier survival analysis showed that higher HOMER3 expression in COAD patients tended to be associated with a shorter OS (P = 0.025), disease-specific survival (DSS) (P = 0.005), and progression-free interval (PFI) (P = 0.001) (Fig. 3). In summary, HOMER3 overexpression was often associated with a poorer prognosis.

Fig. 2.

Clinical correlation between HOMER3 expression levels and COAD. HOMER3 expression levels were. remarkably correlated with the T stage (A), M stage (B), N stage (C), pathologic stage (D), age (E), and CEA level (F).***P < 0.001, **P < 0.01, *P < 0.05; ns, no statistical difference.

Fig. 3.

Assessment of the prognostic value of HOMER3 mRNA levels in COAD patients using Kaplan–Meier plots. Curves for overall survival (A), disease-specific survival (B), and progression-free interval (C) were plotted at a threshold of P < 0.05 to compare COAD patients with high (red) and low (blue) HOMER3 expression levels. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

3.3. HOMER3 overexpression could potentially be used for predicting the diagnosis and prognosis of COAD

A nomogram model based on clinical data and HOMER3 expression levels can be applied to predict the survival probability of a patient in clinical settings (Fig. 4A). ROC curve analysis indicated that HOMER3 expression exhibited partial diagnostic potential (AUC = 0.634) (Fig. 4B). Meanwhile, the time-dependent survival ROC curve showed that all AUC values were above 0.6, except for the 5-year survival rate, which was close to 0.594 (Fig. 4C).

Fig. 4.

ROC analysis and a nomogram model of HOMER3. (A) The nomogram model integrates clinicopathologic factors and HOMER3 levels to predict 1-, 3- and 5-year survival probability. (B) Use of the ROC curve to distinguish tumor tissues from healthy tissues. (C) Time-dependent survival ROC curve analysis for predicting 1-, 3- and 5-year survival rates.

3.4. Genetic alterations in HOMER3 in COAD patients

The analysis included two datasets from TCGA, Firehose Legacy, and DFCI, Cell Reports 2016, totaling 1259 COAD patients. The proportion of HOMER3 genetic alterations in COAD patients was 1.3 % (Fig. 5A), with alteration rates ranging from 0.65 % to 0.81 % (Fig. 5B). Log-rank tests revealed no significant differences in OS (P = 0.658) (Fig. 5C) and disease-free survival (P = 0.206) (Fig. 5D) between patients with and without alterations.

Fig. 5.

Genetic alterations in HOMER3 in COAD patients. (A) An OncoPrint visual summary illustrates the alterations in a HOMER3 query. (B) Summary of alterations in HOMER3 in COAD from DFCI, Cell Report 2016 (a); and TCGA, Firehose Legacy (b). Kaplan–Meier plots comparing overall (C) and disease-free (D) survival in patients with/without HOMER3 gene alterations.

3.5. HOMER3 methylation in COAD patients

Survivvalmeth predictions identified 22 methylated CpG sites in HOMER3 among COAD patients, with cg08894891 and cg08105378 displaying the highest DNA methylation levels (Fig. 6A). Among these 22 sites, 16 sites showed prognostic relevance (cg19530551, cg23310850, cg11921270, cg23264278, cg27586581, cg02774856, cg21725265, cg24836583, cg18397882, cg02592062, cg08886546, cg11601336, cg24733637, cg08894891, cg06872721and cg08105378) (Table 1). Based on median risk, COAD patients were divided into high-risk and low-risk groups. The OS survival curve showed that patients with high-risk scores were more likely to receive a poor prognosis (Fig. 6B).

Fig. 6.

DNA methylation levels in HOMER3 in COAD. (A) Visualization of the correlation between the methylation level and HOMER3 expression level. (B) Kaplan-Meier plots were used to compare overall survival between patients with high and low-risk score.

Table 1.

Effect of hypermethylation level on prognosis in COAD.

Probe ID	Average of tumor samples	Average of normal samples	Fold-change	P-value
cg19530551	0.068725	0.0219	3.138076	1.67E-11
cg23310850	0.059685	0.023008	2.59408	1.82E-10
cg11921270	0.050583	0.023473	2.154977	1.41E-08
cg23264278	0.126726	0.058837	2.153844	4.38E-12
cg27586581	0.034817	0.017319	2.010289	4.29E-05
cg02774856	0.0773	0.052415	1.474766	2.79E-15
cg21725265	0.083787	0.057966	1.445434	1.83E-07
cg24836583	0.032082	0.022839	1.404675	1.69E-03
cg18397882	0.15422	0.112114	1.37556	2.68E-09
cg02592062	0.621277	0.478964	1.297126	1.66E-28
cg08886546	0.133166	0.11452	1.162819	3.26E-04
cg11601336	0.707955	0.642452	1.101958	1.54E-11
cg24733637	0.478366	0.438721	1.090363	5.64E-04
cg16899036	0.400694	0.386004	1.038055	1.92E-01
cg12568707	0.63947	0.633602	1.009261	7.68E-01
cg19788957	0.817658	0.812624	1.006194	4.60E-01
cg25592107	0.605962	0.602712	1.005392	8.56E-01
cg06739873	0.845842	0.856787	0.987225	3.02E-01
cg11300838	0.667153	0.678396	0.983427	4.54E-01
cg08894891	0.926993	0.962515	0.963094	1.02E-06
cg06872721	0.831766	0.866724	0.959666	3.86E-04
cg08105378	0.87626	0.939513	0.932674	2.33E-08

3.6. Correlation between HOMER3 expression and immune cell infiltration

Based on the analysis results from the TIMER database, the expression level of HOMER3 in colorectal cancer (COAD) patients is significantly positively correlated with the number and infiltration level of various immune cells. Specifically, the expression level of HOMER3 is positively correlated with the number of CD4⁺ T cells, macrophages, neutrophils, and dendritic cells. In addition, the expression of HOMER3 is positively correlated with tumor purity, while it is negatively correlated with the number of B cells (Fig. 7). Further analysis has revealed significant correlations between HOMER3 expression and various immune cell biomarkers. These immune cells include B cells, CD8⁺ T cells, Tfh cells (follicular helper T cells), Th1 cells, Th2 cells, Th9 cells, Th17 cells, Th22 cells, Treg cells (regulatory T cells), M1 macrophages, M2 macrophages, tumor-associated macrophages (TAM), natural killer cells, neutrophils, and dendritic cells. These correlations are detailed in (Table 2), providing a more comprehensive perspective on the role of HOMER3 in the tumor immune microenvironment.

Fig. 7.

Relationship between HOMER3 expression levels and immune cell infiltration in COAD.

Table 2.

Correlation analysis between HOMER3 expression and immune cell markers in COAD.

Immune cell	Biomaker	Cor	p value
B cell	CD19	0.156	c
	CD20(KRT20)	−0.131	b
	CD38	0.119	a
CD8+ T cell	CD8A	0.166	c
	CD8B	0.149	b
Tfh	BCL6	0.333	c
	ICOS	0.021	0.855
	CXCR5	0.2	c
Th1	T-bet(TBX21)	0.199	c
	STAT1	0.18	c
	STAT4	0.196	c
	IL12RB2	0.171	c
	WSX1(IL27RA)	0.255	c
	IFN-γ(IFNG)	0.054	0.643
	TNK-α(TNF)	−0.005	0.968
Th2	CCR3	−0.008	0.945
	GATA3	0.396	c
	STAT5A	0.247	c
	STAT6	0.101	a
Th9	IRF4	0.1	a
	PU.1(SPI1)	0.428	c
	TGFBR2	0.159	c
Th17	IL-17A	−0.238	c
	IL-21R	0.336	c
	IL-23R	−0.213	c
	STAT3	0.092	a
Th22	AHR	0.065	0.571
	CCR10	0.208	c
Treg	CCR8	0.241	c
	CD25(IL2RA)	0.179	c
	FOXP3	0.315	c
M1 macrophage	COX2(PTGS2)	0.327	b
	INOS(NOS2)	−0.167	c
	IRF5	0.35	c
M2 macrophage	ARG1	0.17	0.134
	CD206(MRC1)	0.272	c
	CD115(CSF1R)	0.373	c
TAM	PDCD1LG2	0.291	c
	CD80	0.163	c
	CD40	0.321	c
	TLR7	0.241	c
Natural killer cell	CD7	0.239	c
	KIR3DL1	0.104	a
	XCL1	0.177	c
Neutrophil	CD11b(ITGAM)	0.406	c
	CD15(FUT4)	−0.231	c
	CD66b(CEACAM8)	−0.052	0.647
Dendritic cell	CD1C	0.227	c
	CD11c(ITGAX)	0.366	c
	CD141(THBD)	0.406	c

Note: Tfh, follicular helper T cell; Th, T helper cell; Treg, regulatory T cell, TAM, tumor-associated macrophage.

^aP < 0.05.

^bP < 0.01.

^cP < 0.001.

3.7. Correlation between HOMER3 expression and immune checkpoints

In our study, we delved into the mechanisms by which HOMER3 contributes to the immune escape of tumors. To achieve this, we selected several key immune checkpoint molecules as indicators, including PD-1 (also known as PDCD-1), CTLA-4, LMTK3, and LAG3. These molecules are well-known for their critical roles in regulating immune responses and are often implicated in the processes that allow tumors to evade immune detection and destruction.Through meticulous analysis, we found that there is a significant positive correlation between the expression levels of these immune checkpoint molecules and the levels of HOMER3 in the tumor cells. Specifically, the data revealed that as the expression of HOMER3 increases, so does the expression of PD-1, CTLA-4, LMTK3, and LAG3. This correlation suggests that HOMER3 may play a role in upregulating these immune checkpoints, which in turn could facilitate the immune evasion strategies employed by the tumor cells (Fig. 8). The results of our analysis are visually represented in Fig. 8, where the strong positive correlations between HOMER3 expression and each of the immune checkpoint molecules are depicted. These findings indicate that HOMER3 could be a pivotal factor in the complex network of interactions that govern the immune response in the tumor microenvironment, potentially providing a novel target for therapeutic intervention in the treatment of colorectal cancer and other types of cancer.

Fig. 8.

Correlation of HOMER3 expression with PD-1, CTLA-4, LMTK3, and LAG3 expression levels in COAD. Assessing the correlation of HOMER3 expression levels with PD-1 (A), CTLA-4 (B), LMTK3 (C), and LAG3(D) levels in COAD using TCGA databases.

3.8. GO and KEGG pathway analyses

Based on the top ten predicted functional molecules with high scores, we plotted the PPI network patterns of HOMER3 (Fig. 9A and B). Subsequently, KEGG pathway analysis demonstrated high HOMER3, DLG4, and SHANK1 enrichment in glutamatergic synapses (Fig. 9C; Table 3). Furthermore, we established a significant correlation between HOMER3 expression and DLG4 and SHANK and the association between overexpression of DLG4 and SHANK1 and COAD prognosis (Fig. 9D–G). The GO enrichment analysis landscape was illustrated in terms of biological process function, cellular component function, and molecular function (Table 3).

Fig. 9.

Investigating functional annotations and predicted signaling pathways. (A) Examining HOMER3-interacting proteins in COAD represented visually in a bubble chart. (B) Annotation of 10 functional partner genes of HOMER3. (C) GO term and KEGG pathway enrichment analyses. (D) Assessment of the relationship between HOMER3 and DLG4. (E) Survival analysis of DLG4 in COAD. (F) The relevance of the relationship between HOMER3 and SHANK1. (G) Survival analysis of SHANK1 in COAD.

Table 3.

GO and KEGG enrichment analyses of HOMER3 and functional partner genes in COAD.

ONTOLOGY	ID	Description	GeneRatio	BgRatio	pvalue	p.adjust	qvalue
BP	GO:0007215	glutamate receptor signaling pathway	8/11	100/18670	8.29e-17	2.98e-14	1.40e-14
BP	GO:0007216	G protein-coupled glutamate receptor signaling pathway	5/11	15/18670	7.32e-14	1.32e-11	6.21e-12
BP	GO:0051966	regulation of synaptic transmission, glutamatergic	5/11	70/18670	2.91e-10	3.49e-08	1.64e-08
BP	GO:0035249	synaptic transmission, glutamatergic	5/11	93/18670	1.24e-09	1.12e-07	5.26e-08
BP	GO:0031644	regulation of neurological system process	5/11	137/18670	8.82e-09	5.90e-07	2.78e-07
CC	GO:0014069	postsynaptic density	10/11	324/19717	1.36e-17	7.18e-16	3.19e-16
CC	GO:0032279	asymmetric synapse	10/11	328/19717	1.54e-17	7.18e-16	3.19e-16
CC	GO:0099572	postsynaptic specialization	10/11	348/19717	2.79e-17	7.18e-16	3.19e-16
CC	GO:0098984	neuron to neuron synapse	10/11	350/19717	2.96e-17	7.18e-16	3.19e-16
CC	GO:0045211	postsynaptic membrane	9/11	323/19717	4.06e-15	7.89e-14	3.51e-14
MF	GO:0035254	glutamate receptor binding	6/11	47/17697	1.15e-13	6.45e-12	3.52e-12
MF	GO:0030159	receptor signaling complex scaffold activity	3/11	23/17697	3.14e-07	7.63e-06	4.16e-06
MF	GO:0001664	G protein-coupled receptor binding	5/11	280/17697	4.09e-07	7.63e-06	4.16e-06
MF	GO:0032947	protein-containing complex scaffold activity	3/11	29/17697	6.47e-07	9.06e-06	4.94e-06
MF	GO:0035255	ionotropic glutamate receptor binding	3/11	32/17697	8.77e-07	9.83e-06	5.36e-06
KEGG	hsa04724	Glutamatergic synapse	10/11	114/8076	2.29e-18	2.26e-16	1.88e-16
KEGG	hsa04068	FoxO signaling pathway	5/11	131/8076	4.45e-07	2.20e-05	1.82e-05
KEGG	hsa04720	Long-term potentiation	3/11	67/8076	8.59e-05	0.003	0.002
KEGG	hsa04540	Gap junction	3/11	88/8076	1.94e-04	0.005	0.004
KEGG	hsa04723	Retrograde endocannabinoid signaling	3/11	148/8076	8.93e-04	0.018	0.015

3.9. Predicted upstream miRNAs for HOMER3

Prediction results identified 10 upstream miRNAs regulating HOMER3, forming a miRNA-HOMER3 network (Fig. 10A). Further analysis suggested hsa-miR-580-3p, hsa-miR-625-5p, and hsa-miR-3940-3p negatively regulate HOMER3. In COAD patients, hsa-miR-3940-3p is hypothesized to suppress HOMER3 in normal states (Fig. 10B).

Fig. 10.

Identification of miR-3940-3p as a potential upstream miRNA of HOMER3 in COAD. (A) The miRNA-HOMER3 regulatory network was established using Cytoscape software. (B) Analysis of the correlation between expression levels of predicted miRNAs in COAD using the starBase database.

3.10. Using qRT-PCR and Western blot to validate HOMER3 expression in cell lines

We examined HOMER3 expression in COAD cell lines (SW480, HT-29, LOVO, SW620, and HCT116) and healthy colon epithelial cells (NCM460, CCD841, and FHC). The results demonstrated significantly high expression levels of HOMER3 in the COAD cell lines SW480, HT-29, LOVO, and HCT116, which was consistent with previous predictions (Fig. 11).

Fig. 11.

qRT -PCR and Western blot analysis of HOMER3 expression in COAD cell lines (SW480, HT-29, LOVO, SW620, and HCT116) and healthy human colonic epithelial cells (NCM460, CCD841 and FHC). *P < 0.05.

3.11. HOMER3 silencing inhibited cell growth, migration, and invasion in COAD cell lines

The above result aligns with previous research indicating that LOVO cells consistently expressed HOMER3 at high levels, justifying their selection for further experiments. The HOMER3 gene was silenced using short interfering RNA to investigate how gene silencing affects colorectal cancer development. CCK-8 results showed significant inhibition of the proliferation of HOMER3-silenced cells in the LOVO cell line compared to negative controls and blank groups (Fig. 12 A). Wound healing assays conducted at 0, 24, and 48 h showed reduced cell migration ability in LOVO cells under the microscope when the HOMER3 gene was silenced, indicating that high HOMER3 expression promoted cancer cell migration (Fig. 12 B). We then performed transwell experiments using LOVO cells. Transwell experiments further confirmed these findings, as the results showed decreased LOVO cell numbers upon HOMER3 gene silencing in invasion and migration experiments, confirming the role of HOMER3 in promoting cancer cell migration (Fig. 12C).

Fig. 12.

HOMER3 silencing inhibits LOVO cell growth, migration, and invasion in COAD cell lines. (A) CCK-8 assay results in LOVO cells and (B, C) transpore invasion/migration assay results with LOVO cells after HOMER3 silencing. **P < 0.01, ***P < 0.001.

3.12. HOMER3 gene affects tumor progression by influencing the EMT process

We used short interfering RNA to silence HOMER3 to examine if HOMER3 promotes tumor progression by altering the EMT process. RT-qPCR and WB analyses were used to confirm the effects of HOMER3 gene silencing (Fig. 13. A, B). Next, we measured the expression levels of N-cad, SNAIL, VIM, and E-cad. The findings indicated that the silencing of HOMER3 was accompanied by a rise in the RNA and protein levels of E-cad. The RNA and protein expression levels of N-cad, SNAIL, and VIM were lower (Fig. 13. C, D). These results imply that HOMER3 knockdown slows tumor growth. Thus, HOMER3 may facilitate the EMT process in colorectal cancer, thereby facilitating tumor invasion and migration.

Fig. 13.

Effect of HOMER3 on the EMT process occurring in cancerous cells. (A, B) Verification of transfection efficiency of si-HOMER3 in LOVO cells using RT-qPCR and Western Blot, **P ＜ 0.01. (C, D) Changes in RNA and protein expression levels of several classic EMT genes after HOMER3 gene knockdown. ***P ＜ 0.001.

4. Discussion

The correlation between HOMER3 and tumorigenesis has become increasingly evident in recent years. In an initial study, HOMER3 was found to be expressed at increased levels in AML blasts [1]. Subsequent studies revealed that HOMER3 affected As2O3-induced apoptosis by inhibiting Bcl2 expression in leukemia cell lines [10]. In addition, HOMER3 has been identified as a candidate target gene for genomic aberrations in patients with ESCC [11,19]. Recent studies have shown that HOMER3 significantly enhanced the tumorigenicity of glioma cells in vivo by interacting with WBP2. The association between HOMER3 overexpression and patient prognosis in triple-negative breast cancer has also been revealed [7,13]. HOMER3 silencing significantly inhibited the proliferation and cell invasion of bladder cancer cells [12]. We performed a comprehensive pan-cancer analysis of HOMER3 and found that it was upregulated in most tumors; however, no relevant studies in COAD were found, which necessitated a thorough investigation.

This study explored the high expression levels of HOMER3 in COAD tissues using the TCGA database. Our analysis demonstrated that a high expression of HOMER3 in COAD was correlated with poorer prognosis and advanced tumor stages. ROC curve analysis indicates that HOMER3 has tumor diagnostic ability, and we developed a nomogram-enhanced COAD diagnosis. In conclusion, HOMER3 can be used as a COAD diagnosis and prognosis biomarker.

Although tumorigenesis is generally linked with genetic alterations, which are often associated with a poor prognosis, HOMER3 genetic alterations represent only 0.7 % of COAD cases and are not significantly associated with poor OS. Similarly, DNA methylation, a common cancer-related epigenetic mechanism, is strongly associated with cancer development. HOMER3 may play an important role in the link between methylation and prognosis of COAD patients. We identified 16 hypermethylated sites in the HOMER3 sequence that are associated with poor OS.

The occurrence and development of COAD, intricately involves the immune system, as it is an inflammation-related cancer. In order to address the limited treatment options available to patients with advanced disease, several immunological approaches, including dendritic cell-based (DC) therapy and immune checkpoint inhibitors, have emerged as effective treatments [20]. Our study underscores the potential of HOMER3 in immunotherapy treatments for COAD. HOMER3 is positively correlated with various immune cells in COAD patients, including CD4+T cells, macrophages, neutrophils, and dendritic cells.

In our study, HOMER3 showed a significant correlation with immune cell biomarkers. For example, we found that HOMER3 was significantly associated with TAMs, which are commonly thought to promote tumorigenesis and development by modulating the adaptive immune response [20]. In contrast, regulatory T cells (Tregs) play a key role in tumor regulation in COAD [21,22]; we have identified a positive correlation between HOMER3 and Tregs.

Immune checkpoint inhibitors (ICIS) show great potential for improving the survival of cancer patients [20]. In addition to the common immune checkpoints PD-1 and CTLA-4, there is increasing evidence that LMTK3 and LAG3 promote tumorigenesis in various cancers and play an important role in numerous signaling pathways [23,24]. The optimal expression of immune checkpoints is crucial to ensure that immunotherapy is effective [25]. Our findings showed a significant positive correlation between HOMER3 and PD-1, CTLA-4, LMTK3, and LAG3 in COAD patients. This discovery inspires further exploration into the clinical application of HOMER3 in ICIS therapies.

In the study of colorectal adenocarcinoma (COAD), we explored the correlation between Homer3 and immune checkpoints such as PD-1, CTLA-4, LMTK3, and LAG3. The results showed that there is a certain weak immunological correlation between Homer3 and these immune checkpoints, but this correlation has certain limitations. Firstly, we found that the correlation coefficients between Homer3 and CTLA-4, LMTK3, and LAG3 were low (r < 0.3), indicating that the association is weak in intensity. This may suggest that the immunomodulatory role of Homer3 in COAD may not be a dominant factor, but may be influenced by other factors or complex pathways. Secondly, due to the limitation of the current study sample size, we cannot exclude the possibility of incidental associations. A larger sample size study may help further validate these correlations. In addition, our study did not reveal the specific mechanism of action between Homer3 and CTLA-4, LMTK3, and LAG3. Homer3 may be involved in the immune regulation process of COAD, but its specific role may be indirectly related to these immune checkpoints, or interact with other unknown genes and pathways. Finally, although our study found a weak correlation between Homer3 and immune checkpoints, this does not mean that Homer3 can be directly used for immunotherapy of COAD. Further functional and mechanistic studies are still needed to clarify the role of Homer3 in immune regulation of COAD and its potential therapeutic value. In summary, the weak immunological correlation and limitations between Homer3 and CTLA-4, LMTK3, and LAG3 in COAD suggest that when exploring the role of Homer3 in COAD, it is necessary to consider the possible influence of other factors and complex pathways, and to conduct more research to clarify its specific mechanism of action.

In recent years, advancements in immunotherapy have attracted increasing attention towards the role of glutamatergic synapses in cancer therapy [26,27]. KEGG pathway enrichment analysis revealed that HOMER3 and its functional partners, i.e., DLG4 and SHANK1, were enriched in the glutamatergic synaptic pathway. Moreover, we also determined the importance of DLG4 and SHANK1 expression for COAD prognosis. These results suggest that HOMER3, DLG4, and SHANK1 have synergistic effects on immune regulation and tumor development.

Our GO enrichment analysis results have revealed that HOMER3 and its interacting genes are extensively involved in the glutamate receptor signaling pathway, synaptogenesis, and development. Aberrant endosomal to membrane recycling of G protein-coupled receptors (GPCR), a process taking postsynaptic density into account, might play a key role in cancer growth and development [28]. Overexpression of MCM6 and/or Ki-67 has shown great potential for the pathological diagnosis of anaplastic oligodendroglioma, and postsynaptic specialization has been significantly enriched in this process [29]. In addition, neuron-to-neuron synapse involvement has been implicated in the genesis of neuroendocrine tumors [30]. These results provide insight into the potential role of HOMER3 in COAD and serve as a promising avenue for future research.

The continuous progress in experimental techniques [31] has resulted in the regulation of gene expression by ncRNAs receiving increased attention [17,32]. Among the first discovered ncRNAs, miRNAs play an indispensable role in various diseases [33,34]. We reveal that hsa-miR-3940-3p might exhibit the most potential as an upstream regulatory miRNA of HOMER3 in COAD.

Our study found that HOMER3 is highly expressed in colorectal cancer. In order to identify changes in certain EMT-related markers and regulators at the RNA and protein levels, we first knocked down HOMER3 in CRC cells. Our findings showed that HOMER3 could facilitate the growth of tumors by encouraging the development of EMT in colorectal cancer. Further investigations into the specific mechanism by which HOMER3 affects the EMT process and its regulation by upstream miRNA, especially hsa-miR-3940-3p, are crucial. RNA sequencing of HOMER3 silenced cells must be conducted to identify enriched differential genes, related signaling pathways, or transcription factors. The expression of these differential genes was detected by qPCR and WB. Appropriate genes were selected for further knockdown or overexpression experiments to improve our understanding of the specific mechanism by which HOMER3 affected the EMT process of colorectal cancer and its regulation by hsa-miR-3940-3p.

Although the mechanism of action of the HOMER3 gene in tumors and its association with tumor proliferation and prognosis have been reported, its potential as a serum marker for tumor diagnosis remains unexplored. HOMER3 has been implicated in physiological and pathological processes and plays an important role in carcinogenesis. Therefore, future research on HOMER3 holds significant promise. Although there are hardly any reports on the physiological and biochemical properties of HOMER3, previous studies have shown that HOMER3 can potentially be used as a malignancy biomarker. Investigating the structure of HOMER3 and developing specific small molecule inhibitors or HOMER3 activators could pave the way for novel therapeutic strategies against metabolic diseases and cancers.

Finally, this study is the first to explore the diagnostic and prognostic value of HOMER3 expression in COAD. However, further in vitro and in vivo experiments on HOMER3 need to be conducted to understand the specific mechanism of action of HOMER3 thoroughly.

5. Conclusions

Our findings revealed the diagnostic and prognostic potential of HOMER3 in COAD. Moreover, the correlation between HOMER3 gene methylation and COAD prognosis was assessed. HOMER3 plays a role in both COAD development and immune regulation. In addition, our predictions regarding the functions and upstream regulatory mechanism of HOMER3 in COAD provide new insights into COAD pathogenesis. Consequently, HOMER3 can be used as a biomarker to predict the prognosis of patients with COAD and to develop new clinical treatments.

Funding sources

This work was supported by the Integrated Project of Major Research Plan of National Natural Science Foundation of China (No.92249303).

Ethics statement

The cell lines eight colorectal cancer cell lines (SW480, HT-29, LOVO, SW620, and HCT116) and healthy colon epithelial cells (NCM460, CCD841, and FHC) were purchased from the cell bank at the Shanghai Institute of Biological Sciences (Shanghai, China).The authors declare that they have no known competing financial interests or personalrelationships that could have appeared to influence the work reported in this paper.

CRediT authorship contribution statement

Min Luo: Writing – original draft, Software. Cheng Zhao: Formal analysis, Data curation. Yanhua Zhao: Investigation. yin wang: Methodology, Formal analysis. Peifeng Li: Writing – review & editing, Funding acquisition.

Declaration of competing interest

All authors disclosed no relevant relationships.

Appendix A. Supplementary data

The following are the Supplementary data to this article.

figs1.