Cargando…

A precise molecular subtyping of ulcerative colitis reveals the immune heterogeneity and predicts clinical drug responses

BACKGROUND AND AIMS: We sought to identify novel molecular subtypes of ulcerative colitis (UC) based on large-scale cohorts and establish a clinically applicable subtyping system for the precision treatment of the disease. METHODS: Eight microarray profiles containing colon samples from 357 patients...

Descripción completa

Detalles Bibliográficos
Autores principales: Mo, Shaocong, Jin, Bryan, Tseng, Yujen, Lin, Lingxi, Lin, Lishuang, Shen, Xin, Song, Huan, Kong, Mingjia, Luo, Zhongguang, Chu, Yiwei, Jiang, Chen, Cao, Zhiwei, Liu, Jie, Luo, Feifei
Formato: Online Artículo Texto
Lenguaje:English
Publicado: BioMed Central 2023
Materias:
Acceso en línea:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10347743/
https://www.ncbi.nlm.nih.gov/pubmed/37443022
http://dx.doi.org/10.1186/s12967-023-04326-w
Descripción
Sumario:BACKGROUND AND AIMS: We sought to identify novel molecular subtypes of ulcerative colitis (UC) based on large-scale cohorts and establish a clinically applicable subtyping system for the precision treatment of the disease. METHODS: Eight microarray profiles containing colon samples from 357 patients were utilized. Expression heterogeneity was screened out and stable subtypes were identified among UC patients. Immune infiltration pattern and biological agent response were compared among subtypes to assess the value in guiding treatment. The relationship between PRLR and TNFSF13B genes with the highest predictive value was further validated by functional experiments. RESULTS: Three stable molecular subtypes were successfully identified. Immune cell infiltration analysis defined three subtypes as innate immune activated UC (IIA), whole immune activated UC (WIA), and immune homeostasis like UC (IHL). Notably, the response rate towards biological agents (infliximab/vedolizumab) in WIA patients was the lowest (less than 10%), while the response rate in IHL patients was the highest, ranging from 42 to 60%. Among the featured genes of subtypes, the ratio of PRLR to TNFSF13B could effectively screen for IHL UC subtype suitable for biological agent therapies (Area under curve: 0.961–0.986). Furthermore, we demonstrated that PRLR expressed in epithelial cells could inhibit the expression of TNFSF13B in monocyte-derived macrophages through the CXCL1-NF-κB pathway. CONCLUSIONS: We identified three stable UC subtypes with a heterogeneous immune pattern and different response rates towards biological agents for the first time. We also established a precise molecular subtyping system and classifier to predict clinical drug response and provide individualized treatment strategies for UC patients. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1186/s12967-023-04326-w.