Associations of circulating proteins with lipoprotein profiles: proteomic analyses from the omniheart randomized trial and the atherosclerosis risk in communities (ARIC) study

Authors: KIM, HYUNJU - Johns Hopkins University; LICHTENSTEIN, ALICE - Jean Mayer Human Nutrition Research Center On Aging At Tufts University; GANZ, PETER - University Of California San Francisco (UCSF); MILLER, EDGAR - Johns Hopkins University; CORESH, JOSEF - Johns Hopkins University; APPEL, LAWRENCE - Johns Hopkins University; REBHOLZ, CASEY - Johns Hopkins University

Submitted to: Clinical Proteomics
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 6/20/2023
Publication Date: 7/3/2023
Citation: Kim, H., Lichtenstein, A.H., Ganz, P., Miller, E.R., Coresh, J., Appel, L.J., Rebholz, C.M. 2023. Associations of circulating proteins with lipoprotein profiles: proteomic analyses from the omniheart randomized trial and the atherosclerosis risk in communities (ARIC) study. Clinical Proteomics. https://doi.org/10.1186/s12014-023-09416-x.
DOI: https://doi.org/10.1186/s12014-023-09416-x

Interpretive Summary: There are many dietary patterns that are consistent with supporting heart health. They frequently differ in the relative proportions of fat, carbohydrate and protein. Accurately characterizing the relationship among these different dietary patterns and heart disease risk is challenging. Most of what we know relies on subjective self-reported dietary intake data. An alternate approach is to identify objective biomarkers, in our case groups of blood protein concentrations that reflect diet composition. Using samples from a randomized controlled trial that fed participants diets high in fat, carbohydrate and protein, we identified protein signatures characteristic of each of the three dietary patterns. We then identified healthy protein signatures that were related to serum lipoprotein concentrations associated heart health. The diet-related proteins and lipoprotein associations were then validated in an observational cohort. This work may lead to approaches to modify dietary take to reduce heart disease risk.

Technical Abstract: Background: Within healthy dietary patterns, manipulation of the proportion of macronutrient can reduce CVD risk. However, the biological pathways underlying healthy diet-disease associations are poorly understood. Using an untargeted, large-scale proteomic profiling, we aimed to 1) identify proteins mediating the association between healthy dietary patterns varying in the proportion of macronutrient and lipoproteins, and 2) validate the associations between diet-related proteins and lipoproteins in the Atherosclerosis Risk in Communities (ARIC) Study. Methods: In 140 adults from the OmniHeart trial, a randomized, cross-over, controlled feeding study with 3 intervention periods (carbohydrate-rich; protein-rich; unsaturated fat-rich dietary patterns), 4,958 proteins were quantified at the end of each diet intervention period using an aptamer assay (SomaLogic). We assessed differences in log2-transformed proteins in 3 between 52 diet comparisons using paired t-tests, examined the associations between diet-related proteins and lipoproteins using linear regression, and identified proteins mediating these associations using a causal mediation analysis. Levels of diet-related proteins and lipoprotein associations were validated in the ARIC study (n=11,201) using multivariable linear regression models, adjusting for important confounders. Results: Three between-diet comparisons identified 497 significantly different proteins (protein59 rich vs. carbohydrate-rich=18; unsaturated fat-rich vs. carbohydrate-rich=335; protein-rich vs. unsaturated fat-rich dietary patterns=398). Of these, 9 proteins [apolipoprotein M, afamin, collagen alpha-3(VI) chain, chitinase-3-like protein 1, inhibin beta A chain, palmitoleoyl-protein carboxylesterase NOTUM, cathelicidin antimicrobial peptide, guanylate-binding protein 2, COP9 signalosome complex subunit 7b] were positively associated with lipoproteins [high64 density lipoprotein (HDL)-cholesterol (C)=2; triglyceride=5; non-HDL-C=3; total cholesterol to HDL-C ratio=1]. Another protein, sodium-coupled monocarboxylate transporter 1, was inversely associated with HDL-C and positively associated with total cholesterol to HDL-C ratio. The proportion of the association between diet and lipoproteins mediated by these 10 proteins ranged from 21-98%. All of the associations between diet-related proteins and lipoproteins were significant in the ARIC study, except for afamin. Conclusions: We identified proteins that mediate the association between healthy dietary patterns varying in macronutrients and lipoproteins in a randomized feeding study and an observational study. Trial registration: NCT00051350 at clinicaltrials.gov Keywords: Nutrition; Nutrition/Carbohydrate; Nutrition/Protein; Proteomics; lipoproteins; clinical trials; feeding study; macronutrients; dietary pattern.