Skip to main navigation Skip to search Skip to main content

Multimodal MRI and multiomics reveal high-risk neurophenotype in brain-gut circuits as therapeutic target for Crohn's disease

  • Xuehua Li
  • , Ruonan Zhang
  • , Xiaodi Shen
  • , Shaochun Lin
  • , Chuhuai Wang
  • , Weikai Zheng
  • , Weimiao Kong
  • , Zihao Lin
  • , Rongchang Li
  • , Caiguang Liu
  • , Xin Cao
  • , Li Huang
  • , Weitao He
  • , Jinjiang Lin
  • , Haijie Wang
  • , Zhoulei Li
  • , Marietta Iacucci
  • , Subrata Ghosh
  • , Zhenpeng Peng
  • , Canhui Sun
  • Minhu Chen, Guang Yang, Yejun Wang, Ren Mao, Shi Ting Feng
  • The First Affiliated Hospital of Sun Yat-Sen University
  • Shenzhen University
  • East China Normal University
  • Shenzhen Qianhai Taikang Hospital

Research output: Contribution to journalArticlepeer-review

Abstract

The brain-gut axis shapes Crohn's disease (CD) pathogenesis, yet CD-associated neurophenotypes lack defined clinical and mechanistic significance. This work aimed to define these neurophenotypes, assess their prognostic impact, elucidate neurophenotype-driven progression mechanisms using multi-omics, and validate their therapeutic potential in vivo. 109 CD patients were prospectively recruited from two centers and underwent baseline multimodal neuro-MRI, MR enterography, ileocolonoscopy, and fecal/blood sample collection. The neurophenotypes were characterized using a multimodal neuro-MRI model developed from 13 of 13,870 features. 83 patients were followed up for disease progression, with repeated brain-gut assessments. Multi-omics (fecal microbiome/metabolomics, serum metabolomics/neurotransmitters) were used to decode the mechanisms underlying neurophenotype-driven progression. Dextran sulfate sodium (DSS)-induced colitis mice with different neurophenotypes were treated with repeated transcranial magnetic stimulation (rTMS) to explore its therapeutic potential. Multimodal neuro-MRI accurately mapped CD-specific neural signatures, stratifying patients into high-risk (neurophenotype score ≥ 0.45) and low-risk neurophenotypes with robust cross-center validity (training cohort AUC = 0.842, test cohort AUC = 0.824). High-risk neurophenotype was identified as an independent predictor of accelerated disease progression (Hazard ratio = 15.46, p = 0.030), independent of intestinal inflammation. Integrated multi-omics revealed that the high-risk neurophenotype contributed to CD progression through microbial-metabolic-neurotransmitter networks where tryptophan emerged as the central regulatory hub. Serum tryptophan levels were negatively correlated with neurophenotype severity and intestinal disease progression. rTMS targeting high-risk neurophenotype in DSS-induced colitis mice elevated tryptophan levels and ameliorated intestinal disease activity. This study redefines the high-risk neurophenotype as a sustained pathogenic driver rather than a mere phenomenon, proposing brain-gut axis modulation as a promising therapeutic strategy distinct from conventional anti-inflammatory approaches.

Original languageEnglish
Article numbere70122
Pages (from-to)1-20
Number of pages20
JournalInterdisciplinary Medicine
DOIs
Publication statusAccepted/In press - 3 Apr 2026

Keywords

  • Crohn's disease
  • Intestinal disease progression
  • Multimodal MRI
  • Neurophenotype
  • [APCMicrobiome]
  • [Medicine]

Fingerprint

Dive into the research topics of 'Multimodal MRI and multiomics reveal high-risk neurophenotype in brain-gut circuits as therapeutic target for Crohn's disease'. Together they form a unique fingerprint.

Cite this