Molecular classification of the placebo effect in nausea
Numerous studies have shown that the mere expectation improvement can alleviate symptoms in various conditions. These placebo effects often include reliable changes in central and peripheral organ systems. Here, we tested for the first time whether placebo effects can be monitored and predicted by plasma proteins. In a randomized controlled design, 90 healthy participants were exposed to a 20-min vection stimulus on two separate days and were randomly allocated to placebo treatment or no treatment on the second day. Significant placebo effects on nausea, motion sickness, and gastric activity could be verified. Using state-of-the-art proteomics, 74 differentially regulated proteins were identified in placebo-treated participants as compared to no-treatment controls. Gene ontology (GO) enrichment analyses of these proteins revealed acute-phase proteins as well as microinflammatory proteins to be reliable plasma correlates of the placebo effect. Regression analyses showed that day-adjusted scores of nausea indices in the placebo group were predictable by the identified GO protein signatures. We next identified specific plasma proteins, for which a significant amount of variance could be explained by the experimental factors sex, group, nausea, or their interactions. GO enrichment analyses of these proteins identified grooming behavior as a prominent hit, based on neurexin-1 (NRXN1) and contactin-associated protein-like 4 (CNTNAP4). Finally, Receiver Operator Characteristics (ROC) allowed to identify specific plasma proteins differentiating placebo responders from non-responders. These comprised immunoglobulins (IGHM, IGKV1D-16, IGHV3-23, IGHG1) and MASP2, related to regulation of complement activation, as well as proteins involved in oxidation reduction processes (QSOX1, CP TXN). This proof-of-concept study indicates that plasma proteomics are a promising tool to identify molecular correlates and predictors of the placebo effect in humans.