Abstract

Background

   Inflammatory bowel diseases, including ulcerative colitis (UC) and
   Crohn’s disease, are some of the most common inflammatory disorders of
   the gastrointestinal tract. The dysfunction of the immune system in the
   intestines is suggested to be the underlying cause of the pathogenesis
   of UC. However, the mechanisms regulating these dysfunctional immune
   cells and inflammatory phenotypes are still unclear.

Methods

   The differential expression analysis on microarray datasets were
   performed including [37]GSE24287, [38]GSE87466, [39]GSE102133, and
   [40]GSE107499, including 376 samples. “Gene Ontology” and “Kyoto
   Encyclopedia of Genes and Genomes” pathway enrichment analyses were
   conducted to identify the common differentially expressed genes (DEGs)
   in these datasets and explore their underlying biological mechanisms.
   Further algorithms like “Cell-type Identification by Estimating
   Relative Subsets of RNA Transcripts” were used to determine the
   infiltration status of immune cells in patients with UC. “Cytoscape”
   and “Gene Set Enrichment Analysis” were used to screen for hub genes
   and to investigate their biological mechanisms. The Tumor Immune
   Estimation Resource database was used to study the correlation between
   hub genes and infiltrating immune cells in patients with UC. A total of
   three hub genes, CCL3, MMP3, and TIMP1, were identified using
   Cytoscape.

Results

   A positive correlation was observed between these hub genes and
   patients with active UC. These genes served as a biomarker for active
   UC. Moreover, a decrease in CCL3, MMP3, and TIMP1 expression was
   observed in the mucosa of the intestine of patients with active UC who
   responded to Golimumab therapy. In addition, results show a significant
   positive correlation between CCL3, MMP3, and TIMP1 expression and
   different immune cell types including dendritic cells, macrophages,
   CD8+ T cells, and neutrophils in patients with colon cancer. Moreover,
   CCL3, MMP3, and TIMP1 expression were strongly correlated with
   different immune cell markers.

Conclusion

   Study results show the involvement of hub genes like CCL3, MMP3, and
   TIMP1 in the pathogenesis of UC. These genes could serve as a novel
   pharmacological regulator of UC. These could be used as a therapeutic
   target for treating patients with UC and may serve as biomarkers for
   immune cell infiltration in colon cancer.

   Keywords: UC, GEO dataset, biomarkers, colon cancer, immune

1. Introduction

   Inflammatory bowel diseases (IBDs) and their two primary subtypes,
   Crohn’s disease (CD) and ulcerative colitis (UC), cause digestive
   disorders and inflammation in the gastrointestinal tract. UC is a
   chronic recurrent IBD characterized by abdominal cramps, pus, or
   mucinous stools. The lesion usually originates in the rectum, continues
   to the proximal colon, and is often confined to the mucosal layer
   ([41]1, [42]2). The course of the disease in patients with UC is long,
   has a higher chance of relapse, and is challenging to treat. Patients
   with UC have a higher risk of developing colorectal cancer ([43]3).
   World Health Organization has classified UC as a refractory disease
   that seriously threatens human health ([44]4, [45]5).

   However, the pathogenesis of UC is complex. Studies have shown the
   involvement of multiple factors such as environmental and psychological
   factors, epithelial barrier defects, and immune disorders, including
   abnormal cytokine secretion and immune response dysregulation in the
   pathogenesis of UC ([46]6, [47]7). Several clinical
   indicators/biomarkers used for diagnosing and monitoring UC lack
   sensitivity and specificity in distinguishing UC from other functional
   intestinal diseases ([48]8). Patients with UC usually undergo the
   active and remission phases. The goals for successful treatment of UC
   include early remission and long-term maintenance to prevent
   recurrences. Currently, the commonly used drugs for treating patients
   with UC include glucocorticoids, immunomodulators, aminosalicylate,
   biological agents, etc. However, few biomarkers can accurately predict
   drug sensitivity/resistance in patients with UC. Failure of the immune
   tolerance mechanism or damaged normal intestinal barrier leads to
   excessive infiltration of immune cells in lamina propria of the
   intestine, which is then exposed to invading luminal antigens ([49]9).
   These activated immune cells secrete chemokines and cytokines including
   interferon-γ, interleukins, and tumor necrosis factor (TNF) to promote
   inflammation ([50]10). Studies have shown that an imbalance between
   inflammation and mucosal immunity is an important feature of colorectal
   cancer, which also increases the risk of colorectal cancer progression
   in patients with UC ([51]11, [52]12). Therefore, it is necessary to
   determine the molecular basis and mechanism of the pathogenesis of UC
   and identify specific biomarkers to diagnose and treat patients with
   UC. This will help address the problem related to the diagnosis and
   treatment of patients with UC.

   Therefore, in this study, we retrieved datasets like [53]GSE87466,
   [54]GSE107499, [55]GSE24287, and [56]GSE102133 related to IBD from
   multiple databases like Gene Expression Omnibus (GEO,
   [57]http://www.ncbi.nih.gov/geo) ([58]13–[59]15). We used algorithms
   and databases like “Gene Set Enrichment Analysis (GSEA),” “Gene
   Ontology (GO),” and “Kyoto Encyclopedia of Genes and Genomes (KEGG)
   pathway” enrichment analyses to determine the potential underlying
   mechanism. We also used the “Cell-type Identification by Estimating
   Relative Subsets of RNA Transcripts (CIBERSORT)” algorithm to determine
   the proportion of 22 immune cell types in tissues of patients with UC.
   We generated Venn diagrams to identify differentially expressed genes
   (DEGs). Finally, we identified three key hub genes using Cytoscape.
   Further, the Tumor Immune Estimation Resource (TIMER) and Gene
   Expression Profiling Interactive Analysis (GEPIA) databases were used
   to determine the potential correlation between the expression of three
   hub genes and the level of immune cell infiltration in colon cancer.

2. Materials and methods

2.1. Patient selection and tissue collection

   Immunohistochemistry (IHC) was performed on 20 UC tissues and paired
   adjacent normal tissues collected between January 2022 and December
   2022 from Xuzhou Central Hospital, Jiangsu, China. All experiments were
   conducted in accordance with the Declaration of Helsinki. Written
   informed consent was obtained from all patients.

2.2. IHC

   IHC was performed as describe previously ([60]16). Briefly, the tissues
   were exposed to primary antibodies like macrophage inflammatory
   protein-1α (CCL3; 1:200, DF8572, Affinity), MMP3 Monoclonal antibody
   (1:200, 66338-1-Ig, Proteintech), and TIMP1 polyclonal antibody (1:200,
   16644-1-AP, Proteintech). The tissues were stained with
   hematoxylin-eosin (HE) using the HE staining Kit (KGA224, keyGEN
   BioTECH, Nanjing, Jiangsu).

2.3. Cell culture

   Normal colon mucosa NCM460 cells were cultured in Dulbecco’s modified
   Eagle’s medium (DMEM; [61]KGM31600; keyGEN BioTECH, Nanjing, Jiangsu.)
   containing 10% fetal bovine serum. 1 X 105 NCM460 cells were cultured
   in 6-well plates and stimulated by 1 ug LPS for 12 h and 24 h at 37°C.
   Proinflammatory cytokines IL-1β and IL-8 in the supernatants of
   different treatment groups were measured by Flow cytometry.

2.4. Western blotting

   Western blotting was performed as described previously ([62]16). The
   membranes/blot were incubated with primary antibodies like CCL3, MMP3,
   TIMP1, and GAPDH (CCL3; 1:1000, DF8572, Affinity; MMP3 Monoclonal
   antibody, 1:5000, 66338-1-Ig, Proteintech; TIMP1 polyclonal antibody,
   1:1000, 16644-1-AP, Proteintech; GAPDH, 1:20000, 60004-1-Ig,
   Proteintech; Rainbow 180 plus Marker, 39.VPM003, VICMED; Universal
   antibody diluent, 39.VP6022-50, VICMED). Next, the membranes/blots were
   incubated with HRP-conjugated secondary antibody for 1 h. We used
   enhanced chemiluminescence (ECL) system (Image Quant LAS 4000 mini,
   Pittsburgh, USA) for detecting the protein bands as per instructions.
   All experiments were performed in triplicates.

2.5. Data acquisition and DEG analysis

   The microarray datasets like [63]GSE87466, [64]GSE107499, [65]GSE24287,
   and [66]GSE102133 were downloaded from GEO as MINiML files. The
   normalized quantiles functions of the “preprocessCore” R package
   (version 3.4.1) were used to normalize the microarray data. A principal
   component analysis plot was constructed to characterize the samples
   before and after removing the batch effect ([67]17). The“linear models
   for microarray data” and “RobustRankAggreg” R package were used to
   identify DEGs and statistically changed genes (adj. p-value < 0.05).
   Finally, we constructed a Venn diagram to identify the common DEGs
   between the datasets.

2.6. Functional enrichment analysis of DEGs and hub genes

   GO enrichment analysis was used to annotate gene sets into the
   biological process (BP), molecular function (MF), and cellular
   component (CC) terms. The KEGG pathway enrichment analysis was
   conducted to identify the signaling pathways related to DEGs. The
   Metascape database
   ([68]https://metascape.org/gp/index.html#/main/step1) was used to
   perform GO and KEGG pathway enrichment analyses ([69]18). p < 0.05 was
   set as the criterion to identify the significantly enriched GO terms
   and KEGG pathways. Additionally, GSEA was performed to study the
   underlying biological mechanisms of UC, and the results were calculated
   with the aid of Sangerbox3.0 ([70]http://sangerbox.com/home.html), a
   web-based portal ([71]19). We divided the samples based on hub gene
   expression into the high (≥ 50%) and low (< 50%) expression groups to
   identify the pathways and molecular mechanisms associated with UC.

2.7. Determining the infiltration of immune cells in UC

   The “CIBERSORT” algorithm is a gene expression-based deconvolution
   method to determine the cellular composition of a tissue using the LM22
   reference set. The “CIBERSORT” algorithm results were calculated using
   Sangerbox3.0 ([72]19).

2.8. Animal model of colitis

   Tissues from the Dextran sulfate sodium (DSS)-induced colitis mouse
   model were donated by Dr. Zhao Mingxia, School of Basic Medicine, Anhui
   Medical University. All procedures involving animals were reviewed and
   approved by the Ethics Committee of Anhui Medical University (No.
   20180353).

2.9. Statistical analysis

   The Wilcoxon signed-rank test and the Kruskal-Wallis test were used for
   performing the expression analysis in patients with UC and control
   individuals. The correlation between the expression of CCL3, MMP3, and
   TIMP1 and immune cells in tumors was determined using Spearman
   correlation. The strength of the correlation was represented as
   R-value: where 0–0.19 denoted “very weak,” 0.20–0.39 denoted “weak,”
   0.40–0.59 denoted “moderate,” 0.60–0.79 denoted “strong,” 0.80–1.00
   denoted “very strong.” P < 0.05 were considered statistically
   significant. * P < 0.05, ** P < 0.01, *** P < 0.001.

3. Results

3.1. Identification of DEGs, functional annotation, and hub gene selection

   Gene expression data obtained from four microarray datasets, GS87466,
   [73]GSE107499, [74]GSE24287, and [75]GSE102133, were used for
   differential expression analysis. We identified 88 DEGs ([76] Figure 1A
   ). These 88 DEGs were subjected to “GO and KEGG enrichment analyses” to
   identify the potential functions and underlying mechanisms associated
   with IBD. Most GO terms of biological process (BP) were focused on the
   inflammatory response, cell migration, as well as chemotaxis of
   neutrophils and granulocytes ([77] Figure 1B ). GO terms of molecular
   function (MF) were focused on the chemokine activity. Moreover, DEGs
   were enriched in pathways like the IL-17 and TNF signaling pathways and
   cytokine-cytokine receptor interaction ([78] Figure 1B ). The STRING
   database ([79]http://string-db.org) was used for predicting the
   protein-protein interaction (PPI) network of target genes. The results
   were visualized using Cytoscape (version 3.7.2). Next, a Cytoscape
   plug-in like “Molecular Complex Detection” was used to identify and
   visualize important modules of the densely connected nodes in the PPI
   network of target genes. Cytoscape plug-in, “cytoHubba”, was used to
   identify the key genes in the PPI network. We identified the top 20
   genes by the intersection of overlapping genes from six algorithms like
   “Edge Percolated Component,” “Maximal Clique Centrality,” “Closeness,”
   “BottleNeck,” “Degree,” and “Maximum Neighborhood Component.” The
   results indicated three genes, excluding CCL3, MMP3, and TIMP1 ([80]
   Figure 1C ).

Figure 1.

   [81]Figure 1
   [82]Open in a new tab

   Identification and functional annotation of DEGs and hub gene
   selection. (A) Venn diagram shows overlapping genes from four gene
   sets. (B) GO and KEGG pathway enrichment analysis of DEGs. (C)
   Identification of three hub genes.

3.2. Immune-infiltrating landscape of UC and biological processes in the hub
genes

   The pathogenesis of UC may be associated with immune dysregulation in
   the mucosa of the intestine in patients with UC. Therefore, to
   understand the pathogenesis of UC, we used CIBERSORT for comparing the
   differences in the level of 22 immune cell types infiltrating in the
   intestinal mucosa of the healthy controls and patients with UC from the
   [83]GSE87466 ([84] Figure 2A ) and [85]GSE107499 ([86] Figure 2B )
   datasets. The results revealed an increase in the level of infiltration
   of immune cells like T follicular helper cells, M1 macrophages,
   neutrophils, and memory-activated CD4^+ T cells, in the intestinal
   mucosa of patients with UC compared to healthy controls. Moreover, high
   infiltration levels of immune cells like plasma cells, active NK cells,
   and M2 macrophages were observed in the intestinal mucosa of healthy
   controls. GSEA was performed to study the potential biological
   processes associated with hub genes like CCL3, MMP3, and TIMP1. The
   results showed that these hub genes were associated with processes like
   innate immune response, positive regulation of the immune system, B
   cell mediated-immune responses, and T cell activation migration in the
   [87]GSE87466 ([88] Figure 2C ) and [89]GSE107499 ([90] Figure 2D )
   datasets. Together, this indicates that these hub genes mediated immune
   response during the disease process in patients with UC.

Figure 2.

   [91]Figure 2
   [92]Open in a new tab

   CIBERSORT was used to determine immune cell infiltration status in
   samples from the [93]GSE87466 and [94]GSE107499 datasets. GSEA was
   performed on CCL3, MMP3, and TIMP1 to identify biological processes.
   (A) Stacked bar plots and box plots show the relative levels of 22
   immune cells in the [95]GSE87466 dataset. (B) Stacked bar plots and
   boxplots show the relative levels of 22 immune cells in the
   [96]GSE107499 dataset. (C, D) GSEA shows the biological processes
   related to CCL3, MMP3, and TIMP1 in [97]GSE87466 and [98]GSE107499. "*"
   corresponds to a significant level of 5%, that is, p<0.05. p<0.05
   indicates that the difference is 95% certain and has statistical
   significance. "**" corresponds to a significant level of 1%, that is,
   p<0.01, which indicates that the difference is 99% certain, and the
   difference is significant, with statistical significance. "***"
   corresponds to a significant level of 0.1%, that is, p<0.001. p<0.001
   indicates that the difference is 99.9%, which is very significant and
   statistically significant.

3.3. Verification of related hub gene expression

   Subsequently, we validated the differential expression of these three
   hub genes in the [99]GSE87466 and [100]GSE107499 datasets. The results
   indicated an increase in CCL3, MMP3, and TIMP1 expression in the colon
   tissue lesions of patients with UC ([101] Figures 3A, B ). The receiver
   operating characteristic (ROC) analysis was performed on these three
   hub genes based on the [102]GSE87466 and [103]GSE107499 datasets. The
   results revealed that the performance of these three genes in
   predicting the diagnosis of patients with UC was good. The area under
   the ROC curve value of these three hub genes was relatively high:
   0.923, 0.973, and 0.992 in [104]GSE87466 ([105] Figure 3C ) and 0.881,
   0.938, and 0.964 in [106]GSE107499 ([107] Figure 3D ), respectively.
   Furthermore, we determine CCL3, MMP3, and TIMP1 expression in patients
   with active UC from the [108]GSE53306 ([109]20) and [110]GSE59071
   ([111]21) datasets. [112]Figures 4A, B show the differences in the
   expression of CCL3, MMP3, and TIMP1 across healthy controls, inactive
   as well as active intestinal mucosal tissues.

Figure 3.

   [113]Figure 3
   [114]Open in a new tab

   Verification of hub genes. (A, B) The expression of CCL3, MMP3, and
   TIMP1 in [115]GSE87466 and [116]GSE107499. (C, D) ROC analysis of CCL3,
   MMP3, and TIMP1 in the [117]GSE87466 and [118]GSE107499 datasets.
   "****" corresponds to 0.01% significant level, that is, p<0.0001,
   p<0.0001, which indicates that the difference is 99.99% certain, and
   the difference is very significant, with statistical significance.

Figure 4.

   [119]Figure 4
   [120]Open in a new tab

   CCL3, MMP3, and TIMP1 expression levels in patients with active UC. (A)
   The expression of CCL3, MMP3, and TIMP1 in the [121]GSE53306 dataset.
   (B) The expression of CCL3, MMP3, and TIMP1 in the [122]GSE59071
   dataset. "*" corresponds to a significant level of 5%, that is, p<0.05.
   p<0.05 indicates that the difference is 95% certain and has statistical
   significance. "**" corresponds to a significant level of 1%, that is,
   p<0.01, which indicates that the difference is 99% certain, and the
   difference is significant, with statistical significance. "***"
   corresponds to a significant level of 0.1%, that is, p<0.001. p<0.001
   indicates that the difference is 99.9%, which is very significant and
   statistically significant. ns means no statistical significance.

3.4. Validation of the expression of CCL3, MMP3, and TIMP1 in UC patients
tissues, enteritis cells and UC mouse model

   We used IHC to study CCL3, MMP3, and the TIMP1 expression in tissues of
   healthy controls and patients with UC for verifying the results
   obtained in patients from the GEO datasets. The results revealed a
   significant increase in CCL3, MMP3, and TIMP1 expression in tissues of
   patients with UC patients compared to healthy controls ([123]
   Figures 5A–C ). Furthermore, we used western blotting to further verify
   the expression of CCL3, MMP3, and TIMP1 in the tissues of patients with
   UC and healthy controls. The results revealed a significant increase in
   CCL3, MMP3, and TIMP1 expression ([124] Figure 5D ). We also induced
   normal colon mucosa NCM460 cells to transform into enteritis cells
   through LPS, and found that after 24 hours of induction, enteritis
   cells could secreted higher inflammatory factor IL-1β and IL-8 ([125]
   Figure 5E ), and the expression levels of CCL3, MMP3 and TIMP1 in
   enteritis cells were significantly up-regulated, which were consistent
   with the expression in the above tissues ([126] Figure 5F ).
   Additionally, we determined CCL3 and MMP3 expression in the UC mouse
   model. [127]Figures 6A–C show a significant increase in CCL3 and MMP3
   expression levels in tissues of the UC mice compared to the control
   group.

Figure 5.

   [128]Figure 5
   [129]Open in a new tab

   Expression of CCL3, MMP3, and TIMP1 in UC patients and induced
   enteritis cells. (A) The expression of CCL3. (B) The expression of
   MMP3. (C) The expression of TIMP1. (D) The protein expression levels of
   CCL3, MMP3 and TIMP1 in UC patients and healthy controls. (E) The
   inflammatory factors IL-8 and IL-1β secretion level after 1 ug LPS
   stimulated NCM460 cells for 0h, 12h and 24h. (F) The protein expression
   levels of CCL3, MMP3 and TIMP1 in NCM460 cells stimulated with LPS for
   0h, 12h and 24h.

Figure 6.

   [130]Figure 6
   [131]Open in a new tab

   CCL3, MMP3, and TIMP1 expression in DSS-induced colitis mouse model.
   (A) HE staining of tissues of DSS-induced colitis mice and the control
   group. (B) The expression of CCL3 in DSS-induced colitis mouse model.
   (C) The expression of MMP3 in DSS-induced colitis mouse model.

3.5. CCL3, MMP3, and TIMP1 can reflect the therapeutic effect of golimumab
(GLM) in UC patients

   GLM is a monoclonal antibody that is widely used for treating patients
   with UC in clinical settings and has been successful in achieving
   mucosal healing, clinical remission, and response in patients
   ([132]22). In the [133]GSE92415 ([134]22) dataset, high CCL3, MMP3, and
   TIMP1 expression levels were observed in the intestinal mucosa of
   patients with active UC in the GLM response and non-response groups
   compared to the healthy controls before GLM treatment ([135] Figure 7A
   ). In patients treated with GLM for 6 weeks, CCL3, MMP3, and TIMP1
   expression levels were increased in the intestinal mucosa of patients
   with active UC in the GLM response and non-response groups compared to
   the healthy controls. Furthermore, a significant decrease in CCL3,
   MMP3, and TIMP1 expression levels was observed in patients in the GLM
   response group compared to the non-response group ([136] Figure 7B ).
   These results suggest that CCL3, MMP3, and TIMP1 expression could be
   used as indicators to determine the effect of GLM on patients with
   active UC.

Figure 7.

   [137]Figure 7
   [138]Open in a new tab

   Expression of CCL3, MMP3, and TIMP1 on GLM in patients with UC. (A)
   Expression of CCL3, MMP3, and TIMP1 in the intestinal mucosa of
   patients with active UC in the non-response and response groups before
   GLM treatment. (B) Expression of CCL3, MMP3, and TIMP1 in the
   intestinal mucosa of patients with active UC after 6 weeks of GLM
   treatment in the response and non-response groups.

3.6. Correlation between CCL3, MMP3, and TIMP1 expression and infiltration of
immune cells in colon cancer

   Studies have demonstrated that chronic intestinal inflammation
   increases the risk of developing colitis-associated cancer in patients
   with UC ([139]23). Hence, we used the TIMER database to analyze the
   correlation between CCL3, MMP3, and TIMP1 expression and immune cell
   infiltration status in patients with colon cancer. A significant
   positive correlation was observed between CCL3 expression and immune
   cells like CD8+ T cells, macrophages neutrophils, and dendritic cells
   (DC) (P = 5.12e^-10, cor = 0.302; P = 6.20e^-11, cor = 0.318; P =
   2.41e^-51, cor = 0.659; P = 3.16e^-29, cor = 0.52, respectively;
   [140]Figure 8A ). MMP3 expression was significantly correlated with
   immune cells like CD8^+ T cells, neutrophils, and DCs (P = 2.29e^-2,
   cor = 0.113; P = 8.34e^-18, cor = 0.411; P = 4.85e^-6, cor =0.226,
   respectively; [141]Figure 8B ). Similarly, we observed a positive
   correlation between TIMP1 expression and the levels of infiltrating
   CD8^+ T cells, macrophages, neutrophils, and DCs (P = 9.75e^-6, cor =
   0.218; P = 1.06e^-14, cor = 0.372; P = 2.85e^-15, cor = 0.381; P =
   1.26e^-13, cor = 0.358, respectively; [142]Figure 8C ). In addition,
   CCL3, MMP3, and TIMP1 expression were negatively correlated with tumor
   purity in colon cancer (P = 2.30e^-8, cor = -0.273; P = 6.68e^-5, cor =
   -0.196 and P = 7.41e^-16, cor = -0.385, respectively). These results
   indicate the significance of CCL3, MMP3, and TIMP1 expression in the
   infiltration of immune cells in colon cancer.

Figure 8.

   [143]Figure 8
   [144]Open in a new tab

   Relationship between CCL3, MMP3, and TIMP1 expression and level of
   immune cell infiltration in colon cancer. (A) CCL3. (B) MMP3. (C)
   TIMP1.

4. Discussion

   UC is an inflammatory disease of the intestine caused by various
   factors and is becoming increasingly common. It is characterized by a
   long clinical course and frequent relapses. Due to the complex nature
   of UC pathogenesis, there is still a lack of accurate molecular
   diagnostic biomarkers for patients with UC. Previous studies have shown
   alterations in the adaptive immune responses as a potential cause of
   UC. Recent studies have focused on the alterations in innate immune
   responses in the mucosa of the intestine as the cause of UC
   pathogenesis ([145]24, [146]25). However, the underlying mechanisms of
   UC pathogenesis remain unclear.

   In this study, we have analyzed [147]GSE87466, [148]GSE107499,
   [149]GSE24287, and [150]GSE102133 datasets downloaded from GEO to
   identify DEGs in patients with UC. GO and KEGG pathway enrichment
   analyses were performed to identify processes and pathways related to
   DEGs. CIBERSORT analysis revealed a significant increase in the levels
   of memory-activated CD4+ T cells, T follicular helper cells, M1
   macrophages, and neutrophils in the colon tissues of patients with UC.
   Finally, we performed comprehensive bioinformatics analysis using
   Cytoscape to identify CCL3, MMP3, and TIMP1 as hub genes and validated
   their expression in the UC mouse model.

   CCL3, a chemokine, belongs to the CC chemokine family and is involved
   in immune surveillance and tolerance. It is an effective chemokine for
   immune cells like lymphocytes and monocytes. Further, CCL3 binds to the
   receptors like CCR4, CCR1, and CCR5 on immune cells like T cells, DC, B
   cells, and eosinophils. Various studies have used in vitro and in vivo
   models and have shown that CCL3 exerts chemotactic effects on
   neutrophils and monocytes ([151]26, [152]27). During
   monocyte-endothelial cell interactions, monocyte binds to intercellular
   adhesion molecule-1 and elevates CCL3 levels in these cells. Thus, the
   levels of CCL3 increase in response to endothelial cell-leukocyte
   interactions, which is an important mechanism for recruiting cells
   during inflammatory immune responses ([153]28–[154]30). Banks et al.
   revealed a correlation between CCL3 expression and severity of
   inflammation, and an increase in CCL3 expression level was correlated
   with increased disease activity ([155]31), consistent with our results.

   MMP3 belongs to the Zinc-dependent endopeptidase family, primarily
   secreted by cancer cells, connective tissues, endothelial cells, immune
   cells like mononuclear macrophages, and neutrophils. Studies have shown
   the involvement of MMP3 in extracellular matrix degradation ([156]32,
   [157]33). MMP3 is a major matrix metalloproteinase detected in the
   mucosa of patients with IBD ([158]34). Anna et al. showed a correlation
   between the levels of MMP3 in the serum of children with UC and disease
   activity. Furthermore, the levels of MMP3 in serum are directly linked
   to disease activity, wherein an increase in MMP3 levels in serum
   increases the disease activity. Thus MMP3 could serve as a new marker
   to determine the activity of UC in children ([159]35). However, TIMP1
   can counteract matrix metalloproteinase (MMP) activity. Mounting
   evidence suggests the involvement of an imbalance in the MMP/TIMP ratio
   in the pathogenesis of diseases like IBD and cancers ([160]36,
   [161]37). Nevertheless, TIMP1 attenuates enhanced MMP activity in IBD
   ([162]38). Multiple studies have shown an increase in TIMP1 expression
   in patients with UC. Further, a significant positive correlation was
   observed between TIMP1 expression and the degree of mucosal injury
   under an endoscope, C-reactive protein levels, disease, and clinical
   activity index values ([163]38–[164]40). These results are consistent
   with our findings.

   Our results showed an increase in CCL3, MMP3, and TIMP1 expression
   levels in the tissues of patients with UC. Further, the CCL3 expression
   level was higher in patients with active UC compared to patients with
   inactive UC using GEO datasets. In addition, high CCL3, MMP3, and TIMP1
   expression levels were observed in patients with UC and DSS-induced UC
   mouse models compared to healthy controls or control mice, thereby
   further validating our results. We performed GSEA, and the results
   revealed that high expression of CCL3, MMP3, and TIMP1 was related to
   innate and B cell-mediated immune response, positive regulation of the
   immune system, T cell activation, and migration. This suggests that
   CCL3, MMP3, and TIMP1 play an immunomodulatory role in UC pathogenesis.
   GLM is a monoclonal antibody against TNFα and is involved in regulating
   the autophagy of cells. It has been widely used for treating adults
   with UC for several years ([165]41). Our results showed that after 6
   weeks of GLM treatment, a significant decrease in CCL3, MMP3, and TIMP1
   expression levels was observed in patients with UC in the GLM response
   group compared to patients with UC in the non-response group. This
   indicates the involvement of GLM in activating autophagy in UC
   ([166]42). Recently, studies have focused on the progression of UC to
   colon cancers. Dysregulation of the immune system and alteration in
   immune homeostasis in the intestine are important features of the
   malignant transformation of UCs ([167]43). Therefore, we next analyzed
   if there was a correlation between the expression of hub genes such as
   CCL3, MMP3, and TIMP1 and infiltrating immune cells in colon cancer.
   The results revealed that CCL3, MMP3, and TIMP1 expression was
   positively correlated with the infiltration status of immune cells like
   DCs, CD8^+ T cells, and neutrophils in colon cancer. Together, these
   results demonstrate the significance of CCL3, MMP3, and TIMP1 in
   recruiting and regulating infiltrating immune cells in colon cancer.

   In summary, we used a combination of bioinformatics tools to identify
   three hub genes in UC, and their expression was validated in patients
   with UC and DSS-induced UC mouse models. These results may aid in
   enhancing our understanding and exploring the mechanisms underlying the
   pathogenesis of UC. This may also help explore potential biomarkers for
   diagnosing and treating patients with UC.

Data availability statement

   The datasets presented in this study can be found in online
   repositories. The names of the repository/repositories and accession
   number(s) can be found below: [168]http://www.ncbi.nih.gov/geo;
   [169]https://metascape.org/gp/index.html#/main/step1;
   [170]http://sangerbox.com/home.html;
   [171]http://sangerbox.com/home.html; [172]http://string-db.org.

Ethics statement

   The studies involving human participants were reviewed and approved by
   the Institutional Ethics Committee of Xuzhou central hospital. The
   patients/participants provided their written informed consent to
   participate in this study. The animal study was reviewed and approved
   by the Institutional Ethics Committee of Xuzhou central hospital.
   Written informed consent was obtained from the owners for the
   participation of their animals in this study.

Author contributions

   ZP and JF: Conceptualization, Data curation, Formal analysis, Funding
   acquisition, Investigation, Methodology, Resources, Supervision,
   Visualization, Writing-original draft, Project administration,
   Writing-review and editing. HL: Formal analysis, Writing-review and
   editing. YF and FZ: Data curation, Investigation, Writing-review and
   editing. FG, GN, and BG: Data curation, Project administration. All
   authors contributed to the article and approved the submitted version.

Acknowledgments