Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2024 May 22:15:1401733.
doi: 10.3389/fimmu.2024.1401733. eCollection 2024.

Identifying hub genes in response to ustekinumab and the impact of ustekinumab treatment on fibrosis in Crohn's disease

Ying Xu  1 , Shu Wang  1 , Ziping Ye  1 , Hongjie Zhang  1
Affiliations

Identifying hub genes in response to ustekinumab and the impact of ustekinumab treatment on fibrosis in Crohn's disease

Ying Xu et al. Front Immunol. .

Abstract

Introduction: Crohn's disease (CD) is a chronic inflammatory disease. Approximately 50% of patients with CD progressed from inflammation to fibrosis. Currently, there are no effective drugs for treating intestinal fibrosis. Biologic therapies for CD such as ustekinumab have benefited patients; however, up to 30% of patients with CD have no response to initial treatment, and the effect of ustekinumab on intestinal fibrosis is still uncertain. Therefore, it is of great significance to explore the predictive factors of ustekinumab treatment response and the effect of ustekinumab on intestinal fibrosis.

Materials and methods: Public datasets-GSE207465 (blood samples) and GSE112366 and GSE207022 (intestinal samples)-were downloaded and analyzed individually (unmerged) based on the treatment response. Differentially expressed genes (DEGs) were identified by the "limma" R package and changes in immune cell infiltration were determined by the "CIBERSORT" R package in both blood and intestinal samples at week 0 (before treatment). To find predictive factors of ustekinumab treatment response, the weighted gene co-expression network analysis (WGCNA) R package was used to identify hub genes in GSE112366. Hub genes were then verified in GSE207022, and a prediction model was built by random forest algorithm. Furthermore, fibrosis-related gene changes were analyzed in ileal samples before and after treatment with ustekinumab.

Results: (1) Our analysis found that MUC1, DUOX2, LCN2, and PDZK1IP1 were hub genes in GSE112366. GSE207022 revealed that MUC1 (AUC:0.761), LCN2 (AUC:0.79), and PDZK1IP1 (AUC:0.731) were also lower in the response group. Moreover, the random forest model was shown to have strong predictive capabilities in identifying responders (AUC = 0.875). To explore the relationship between intestinal tissue and blood, we found that ITGA4 had lower expression in the intestinal and blood samples of responders. The expression of IL18R1 is also lower in responders' intestines. IL18, the ligand of IL18R1, was also found to have lower expression in the blood samples from responders vs. non-responders. (2) GSE112366 revealed a significant decrease in fibrosis-related module genes (COL4A1, TUBB6, IFITM2, SERPING1, DRAM1, NAMPT, MMP1, ZEB2, ICAM1, PFKFB3, and ACTA2) and fibrosis-related pathways (ECM-receptor interaction and PI3K-AKT pathways) after ustekinumab treatment.

Conclusion: MUC1, LCN2, and PDZK1IP1 were identified as hub genes in intestinal samples, with lower expression indicating a positive prediction of ustekinumab treatment response. Moreover, ITGA4 and IL18/IL18R1 may be involved in the treatment response in blood and intestinal samples. Finally, ustekinumab treatment was shown to significantly alter fibrotic genes and pathways.

Keywords: Crohn’s disease; intestinal fibrosis; transcriptomics; treatment response; ustekinumab.

PubMed Disclaimer

Conflict of interest statement

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Figures

Figure 1
Figure 1
Flow diagram of the data analysis. (A) Analysis of the baseline index. (B) Analysis of the index before and after ustekinumab treatment.
Figure 2
Figure 2
DEGs in the response group compared to the nonresponse group in the ileum at week 0. (A) A volcanic map of the DEGs in GSE112366. (B) A heatmap of the top 10 upregulated and the top 10 downregulated DEGs. (C) The KEGG analysis. (D) The top 10 functional enrichment in BP, CC, and MF analysis. (E) Differences in immune cell infiltration between the response group and the non-response group. Y: response, N: non-response. DEGs, differentially expressed genes; KEGG, Kyoto Encyclopedia of Genes and Genomes. Wilcoxon tests were used for statistics. *p < 0.05.
Figure 3
Figure 3
WGCNA of the GSE112366 dataset. (A) The soft threshold power of WGCNA. The power was 13. (B) The genes with strong correlation were clustered into the same module, and different modules were represented by different colors. (C) The correlation between the modules and the treatment response. (D) The green module was significantly correlated with the treatment response (COR = 0.52, p = 0.0065). WGCNA, weighted gene co-expression network analysis.
Figure 4
Figure 4
Identification of hub genes. (A) The PPI network of genes in green module in the STRING database. (B) The top 10 genes were evaluated by the cytoHubba plugin. (C) The Venn diagram of overlapping between the genes in the green module and DEGs in GSE112366. The overlapping genes are BACE2, PDZK1IP1, KCNE3, LCN2, DUOX2, and MUC1. (D) The correlation between overlapping genes and 22 immune cells. Pearson correlations were used for statistics. DEGs, differentially expressed genes.
Figure 5
Figure 5
Verification of hub genes and establishment of the random forest model. (A) Comparison of the expression of hub genes in the GSE207022 dataset between the response and the non-response group. (B) The area under ROC curve indicates the effectiveness of the hub genes in prediction of treatment response to ustekinumab in the GSE207022 dataset. (C) Relative importance of all features in the current study based on mean decrease in accuracy (left) and mean decrease in Gini index (right) in the GSE207022 dataset. (D) The ROC plot for the random forest model in the GSE207022 dataset, AUROC is 0.875.
Figure 6
Figure 6
The GO analysis and the related regulatory network of hub genes. (A) The top 15 GO terms with the greatest significance of three hub genes. GO: Gene Ontology. (B) The network of transcription factors and three hub genes. (C) The network of miRNAs and three hub genes.
Figure 7
Figure 7
The co-expression genes in ileal and blood samples. (A) The Venn diagram of overlapping DEG genes among the GSE112366, GSE207022, and GSE207465 datasets. The difference of the expression of ITGA4 and IL18R1 between the response group and the nonresponse group in the blood (B), ileum (C), and rectum (D). DEGs, differentially expressed genes; Y: response, N: non-response.
Figure 8
Figure 8
The expression of ITGA4 and IL18R1 in the blood and intestinal samples. The correlation analysis between the expression of ITGA4 and IL18R1 and immune cell abundance in the blood (A), ileum (B), and rectum (C).
Figure 9
Figure 9
DEGs before and after ustekinumab treatment. (A) A volcanic map of the DEGs in the GSE112366 dataset. (B) A heatmap of the top 40 DEGs. (C) The KEGG analysis. (D) The GO enrichment analysis of DEGs in BP, CC, and MF. (E) The estimation of the infiltration of immune cells using the CIBERSORT algorithm. Wilcoxon tests were used for statistics. *p <0.05, **p < 0.01. DEGs, differentially expressed genes; GO, Gene Ontology; KEGG, Kyoto Encyclopedia of Genes and Genomes.
Figure 10
Figure 10
(A) The top 10 pathways from KEGG analysis and the genes enriched in the pathways are visualized. (B) The fibrosis-related pathways with significant changes and the genes enriched in the pathways are visualized. KEGG, Kyoto Encyclopedia of Genes and Genomes.
Figure 11
Figure 11
The fibrosis-related DEGs before (week 0) and after treatment (week 44) with ustekinumab. DEGs, differentially expressed genes. *p < 0.05; **p < 0.01; ***p < 0.001; ns, not significant.
See this image and copyright information in PMC

Similar articles

See all similar articles

Cited by

References

    1. Torres J, Mehandru S, Colombel JF, Peyrin-Biroulet L. Crohn’s disease. Lancet (London England). (2017) 389:1741–55. doi: 10.1016/S0140-6736(16)31711-1 - DOI - PubMed
    1. Feuerstein JD, Cheifetz AS. Crohn disease: epidemiology, diagnosis, and management. Mayo Clin Proc. (2017) 92:1088–103. doi: 10.1016/j.mayocp.2017.04.010 - DOI - PubMed
    1. Baumgart DC, Sandborn WJ. Crohn’s disease. Lancet. (2012) 380:1590–605. doi: 10.1016/S0140-6736(12)60026-9 - DOI - PubMed
    1. D’Alessio S, Ungaro F, Noviello D, Lovisa S, Peyrin-Biroulet L, Danese S. Revisiting fibrosis in inflammatory bowel disease: the gut thickens. Nat Rev Gastroenterol Hepatol. (2022) 19:169–84. doi: 10.1038/s41575-021-00543-0 - DOI - PubMed
    1. Cushing K, Higgins PDR. Management of crohn disease: A review. Jama. (2021) 325:69–80. doi: 10.1001/jama.2020.18936 - DOI - PMC - PubMed