The blood exposome and its role in discovering causes of disease

Abstract

Background: Since 2001, researchers have examined the human genome (G) mainly to discover causes of disease, despite evidence that G explains relatively little risk. We posit that unexplained disease risks are caused by the exposome (E; representing all exposures) and G × E interactions. Thus, etiologic research has been hampered by scientists' continuing reliance on low-tech methods to characterize E compared with high-tech omics for characterizing G.

Objectives: Because exposures are inherently chemical in nature and arise from both endogenous and exogenous sources, blood specimens can be used to characterize exposomes. To explore the "blood exposome" and its connection to disease, we sought human blood concentrations of many chemicals, along with their sources, evidence of chronic-disease risks, and numbers of metabolic pathways.

Methods: From the literature we obtained human blood concentrations of 1,561 small molecules and metals derived from foods, drugs, pollutants, and endogenous processes. We mapped chemical similarities after weighting by blood concentrations, disease-risk citations, and numbers of human metabolic pathways.

Results: Blood concentrations spanned 11 orders of magnitude and were indistinguishable for endogenous and food chemicals and drugs, whereas those of pollutants were 1,000 times lower. Chemical similarities mapped by disease risks were equally distributed by source categories, but those mapped by metabolic pathways were dominated by endogenous molecules and essential nutrients.

Conclusions: For studies of disease etiology, the complexity of human exposures motivates characterization of the blood exposome, which includes all biologically active chemicals. Because most small molecules in blood are not human metabolites, investigations of causal pathways should expand beyond the endogenous metabolome.

Figure 1

Risk factors for exposures that contribute to chronic-disease mortality. The chart was compiled from World Health Organization estimates of exposures affecting 50 million global deaths in 2010 (Lim et al. 2012). (Because some risk factors may be correlated, the indicated percentages are approximate.)

Figure 2

Small molecules and metals in human blood. Each curve represents the cumulative distribution of chemical concentrations from a particular source category (pollutants, n = 94; drugs, n = 49; food chemicals, n = 195; endogenous chemicals, n = 1,223). Abbreviations: BDE 100, 2,2′,4,4′,6-pentabromo­diphenyl ether; DDE, 1,1-bis-(4-chlorophenyl)-2,2-dichloroethene; OCDD, 1,2,3,4,6,7,8,9-octachloro­oxanthrene; PCB 170, 2,2′,3,3′,4,4′,5-heptachloro-1,1′-biphenyl.

Figure 3

Chemical-similarity maps of small molecules and metals in human blood (Tanimoto coefficient ≥ 0.7; symbol color represents the source category). (A) All chemicals (n = 1,561; symbol size reflects the blood/serum concentration). Map locations: AA, leucotrienes; AB, perfluorinated compounds; AC, alkylamines; AD, pteridines; AE, pyrimidine nucleotides; AF, aliphatic amino acids and derivatives; AG, sphingo­lipids; AH, organochlorine pesticides; AI, prenol lipids; AJ, sulfur compounds; AK, flavonoids; AL, pyrroles and indoles; AM, pyridines; AN, alkaloids; AO, benzoic acids and phenols; AP, eicosanoids; AQ, fatty acids and fatty amines; AR, steroids; AS, organic acids; AT, monosaccharides; AU, phosphates; AV, alcohols; AW, fatty acid esters and conjugates; AX, polychlorinated biphenyls; AY, simple aromatics; AZ, chlorinated dioxins and furans; BA, sulfates and nitrites/nitrates; BB, purine nucleotides; BC, aromatic amino acids and derivatives; BD, benzoic acids and phenols. (B) Matching chemicals from (A) cited in studies of chronic-disease risks (n = 336; symbol size reflects the number of citations). Map locations: 1, selenium; 2, nitric oxide; 3, folic acid; 4, vitamin B12; 5, metformin; 6, cotinine; 7, lead; 8, bilirubin; 9, atorvastatin; 10, ascorbic acid; 11, thyroxine; 12, nor­epinephrine; 13, aspirin; 14, eicosapentaenoic acid; 15, magnesium; 16, calcium; 17, sodium; 18, uric acid; 19, creatinine; 20, l-arginine; 21, homocysteine; 22, l-methionine; 23, l-valine; 24, β-carotene; 25, vitamin A; 26, vitamin D3; 27, cholesterol; 28, simvastatin; 29, aldosterone; 30, cortisol; 31, testosterone; 32, malon­dialdehyde; 33, ᴅ-glucose; 34, estradiol; 35, PCBs; 36, ethanol. (C) Matching chemicals from (A) having human metabolic pathways (n = 658; symbol size reflects the number of pathways). Map locations: 1, adenosine triphosphate; 2, hydrogen peroxide; 3, adeno­sine diphosphate; 4, guano­sine diphosphate; 5, guanosine triphosphate; 6, NADPH; 7, cyclic AMP; 8, adenosine monophosphate; 9, NADH; 10, NAD; 11, FAD; 12, manganese; 13, sodium; 14, calcium; 15, zinc; 16, magnesium; 17, potassium; 18, norepinephrine; 19, epinephrine; 20, ʟ‑phenyl­alanine; 21, ʟ‑tyrosine; 22, dopamine; 23, palmitic acid; 24, cholesterol; 25, ʟ‑glutamic acid; 26, adenine; 27, ʟ‑aspartic acid; 28, oxoglutaric acid; 29, pyruvic acid; 30, phosphate; 31, pyrophosphate; 32, formic acid; 33, uridine 5′-mono­phosphate; 34, uridine 5′-diphosphate; 35, ʟ‑arginine; 36, ʟ‑alanine; 37, ʟ‑cysteine; 38, ʟ‑serine; 39, arachodonic acid; 40, α-linolenic acid.