Diagnosis, Prognosis, and Drug Target Discovery for Chronic Widespread Pain: A Large Proteogenomic Study, 2025, Chen et al

hotblack

Senior Member (Voting Rights)
Diagnosis, Prognosis, and Drug Target Discovery for Chronic Widespread Pain: A Large Proteogenomic Study

Li Chen, Eoin Kelleher, Ruogu Meng, Duanke Liu, Yuchen Guo, Yunhe Wang, Yaqing Gao, Zhe Huang, Zhu Liang, Shuai Yuan, Chao Zeng, Jun Ma, Yanhui Dong, Anushka Irani, Guanghua Lei, Junqing Xie, Daniel Prieto-Alhambra

Abstract
Chronic widespread pain (CWP) remains challenging due to its heterogeneous causes and complex mechanisms. A total of 2920 plasma proteins are analyzed from 29,254 UK Biobank participants. A total of 256 proteins are identified as cross-sectionally correlated with CWP. A simple (top 10 proteins) and comprehensive (all significant proteins) proteomic-based score (ProtS) is created for CWP diagnosis, both outperforming and improving the existing clinical score (area under the curve, AUC: 0.801, 0.723, and 0.791 alone, and 0.856 and 0.880 in combination).

In addition, the protein score predicted 13-years risk of pain-related traits over the body, including pain onset, progression, and intensity; Moreover, it has stronger associations with nociplastic pain and fibromyalgia compared to nociceptive and neuropathic pain, implying a unique protein signature of different pain mechanisms.

Finally, among 434 candidate proteins prioritized in the observational analysis, 18 are corroborated with causal relevance by Mendelian randomization, and importantly, four (CA14, DPEP1, LGALS3, and TNF) showed potential as novel drug targets repurposed for treating CWP.

Link (Adv. Sci.)
https://doi.org/10.1002/advs.202507691
 
Looked like an interesting paper on data driven drug discovery for chronic pain using a proteome-wide association study (PWAS) and things like gene ontology enrichment analysis on UK Biobank data (protein and genetic data). They used machine learning with a proteomic-based score and clinical score to identify relevant proteins then used open target databases to look for potential therapeutics to modify them.

I haven’t been through it detail but it looks like there’s lots interesting in there and that CA14 popped up when CA10 was identified in DecodeME may be relevant?

A few quotes on proteins identified

In the analysis of widespread chronic pain, the proteins most significantly upregulated compared to controls were CDCP1, IL18R1, GAST, GDF15, ASGR1, CCL7, VSIG2, CHGA, TNFRSF10A, and MRC1. In contrast, the ten most downregulated proteins were CA14, APOF, PON3, SELENOP, EGFR, ITGAV, SCT, CELA2A, IGFBP3, and PPY

Several proteins, including APOF, ASGR1, CA14, CA6, MXRA8, PLTP, PON3, and RARRES2, were shared across multiple pain types

Proteins, including CA14, KIR2DS4, PDCD1, CHI3L1, CNDP1, IL18R1, CRH, GAST, HPGDS, ITGAV, SELENOP, ASAH2, LYPD8, GZMB, and PROCR, were identified as key contributors to widespread chronic pain. Among these, CA14 emerged as the most impactful, reflected by its highest SHAP (SHapley Additive exPlanations) value, suggesting a prominent role in pain prediction. KIR2DS4 and PDCD1 followed as major contributors.

9 proteins (BPIFB2, CA14, COL9A1, CTSO, IFI30, LGALS3, PTN, SFTPD, and TNF) exhibited significant associations with nociplastic pain or fibromyalgia. 2 of the 15 proteins (LEG1 and LRRC37A2) showed significant associations with nociceptive pain or neuropathic pain, indicating that these 14 proteins can effectively distinguish nociplastic pain/fibromyalgia from nociceptive and neuropathic pain up to 9 years in advance (Figure 5C). At the 2021 follow-up, the onset and progression of chronic pain showed significant associations with the 10 proteins (BPIFB2, CA14, COL9A1, CTSO, FAM171B, IFI30, LGALS3, MLN, PTN, and SFTPD).

On drug repurposing

To evaluate the translational potential of our protein prioritization strategy, we systematically interrogated Open target databases for therapeutic agents targeting the 18 proteins identified by MR as causally linked to widespread chronic pain. Of these, four proteins—CA14, DPEP1, LGALS3, and TNF—had existing pharmacological modulators, either approved or in clinical development, encompassing 18 distinct compounds across small molecules, antibodies, and protein-based biologics

CA14, a carbonic anhydrase expressed in neuronal tissues, is targeted by sulthiame, an antiepileptic drug approved for obstructive sleep apnea and seizure disorders. Given the observed non-linear associations between CA14 levels and pain phenotypes, CA14 agonists—rather than inhibitors—may offer greater therapeutic utility in this context

DPEP1, a renal dipeptidase involved in inflammatory responses, is inhibited by cilastatin, a small molecule co-administered with β-lactam antibiotics to reduce renal toxicity in infectious disease settings.

LGALS3 (galectin-3), a regulator of immune signaling and fibrosis, is targeted by several investigational agents, including belapectin, davanat, and olitigaltin, with indications ranging from non-alcoholic steatohepatitis to fibrotic lung disease and cancer.

TNF, a master cytokine in inflammatory signaling, is the target of multiple approved biologics—such as adalimumab, infliximab, and etanercept—used in immune-mediated diseases including rheumatoid arthritis, Crohn's disease, and psoriasis.

Collectively, these findings point to multiple actionable opportunities for drug repurposing in nociplastic pain, supported by proteome-wide causal inference. In particular, TNF antagonists and galectin-3 inhibitors emerge as strong candidates for translational advancement in pain therapeutics

And the conclusion

In conclusion, this study underscores the significant potential of proteomic profiling to advance our understanding and management of chronic pain, particularly nociplastic pain, and fibromyalgia. By coupling machine learning–based prioritization with genetic validation, we identified 18 circulating proteins with putative causal roles in widespread chronic pain, several of which—such as CA14, DDR1, and MLN—also demonstrated colocalized genetic regulation. Importantly, four of these proteins (CA14, DPEP1, LGALS3, and TNF) are targetable by approved or investigational therapeutics, revealing immediate opportunities for drug repurposing. TNF inhibitors and galectin-3 antagonists, in particular, emerge as promising candidates for translation into clinical pain management
 
Last edited:
Back
Top Bottom