ABSTRACT: Rationale: Although there is no available theory or effective technology to differentiate, subhealth symptoms caused by prolonged exposure of incorrect diet or humid environment are triggers for chronic disease, which become a serious problem of human health. These symptoms are similar as dampness in traditional Chinese medicine (TCM). They can lead to dampness constitution (DC) and dampness syndrome (DS), which easily resulting in a variety of human immune and metabolic diseases, especially hyperlipidemia in human beings. However, its clinical diagnosis still mainly depends on the doctor’s judgement and constitution scales. It is urgent to develop an objective, accurate and effective clinical diagnosis technology to meet this significant clinical challenge. Methods: Coupling TiO2 enrichment method and exosome proteomics, we obtained high-quality serum exosome proteins and proteome profiling in healthy subjects with balanced constitution (BC) and typical dampness groups through data-independent acquisition mass spectrometry (DIA-MS). Two groups of potential diagnostic markers of dampness were identified by machine learning algorithm. These proteins were verified by the expression changes in BC, hyperlipidemia subjects with balanced constitution (HBC) and with dampness constitution (HDC) groups. Furthermore, these proteins were also verified by the expression changes before and after 16 week’s use of Fuling-zexie decoction, a well-known traditional Chinese medicine prescription in conditioning dampness. Results: Serum exosome proteome can be saved as identity of dampness syndrome. By comparing the exosome proteome of the individuals from balanced constitution and dampness groups, 89 differentially-expressed exosome proteins were identified, which were involved in dampness-relevant biological processes, such as immune activity, inflammation and lipid metabolism. Finally, a panel of fourteen proteins included group A (IGA2, IGHV1-69, IGHV3-38, IGHV4-28, IGKV3-15, IGKV4-1) and group B proteins (CRP, FAN3, F9, F10, GP5, SERPING1, SPP2, PCSK9) screened through random forest approach could efficiently differentiate dampness from healthy subjects with balanced constitution (BC) and typical dampness groups. To further verify the panel composed of these fourteen proteins, we analyzed 41 serum exosome proteomes contained BC, HBC and HDC groups, showing that combination of six proteins in group A and six proteins in group B could accurately distinguish between HDC and BC, and eight proteins in group B could accurately distinguish between HBC and HDC. Moreover, we analyzed 30 serum exosome proteomes from the dampness groups with hyperlipidemia before and after 16 weeks’ treatment with traditional Chinese medicine Fuling-zexie decoction or placebo. The results showed that ten proteins were dampness specifical proteins and changed in the direction of improvement with the removal of dampness. Finally, we identified thirteen proteins as potential diagnostic biomarkers. Conclusion: These results demonstrated the clinical applicability of our serum exosome proteomics in characterizing and differentiating dampness. The differential dampness specific diagnostic markers obtained from the serum exosome proteomics followed by machine learning approach have potential application value in identification of dampness.
Zhimin Yang 
