F. prausnitzii potentially modulates the association between citrus intake and depression

Abstract

Background: The gut microbiome modulates the effects of diet on host health, but it remains unclear which specific foods and microbial features interact to influence risk of depression. To understand this interplay, we leveraged decades of dietary and depression data from a longitudinal cohort of women (n = 32,427), along with fecal metagenomics and plasma metabolomics from a substudy (n = 207) nested in this cohort, as well as an independent validation cohort of men (n = 307).

Results: We report that citrus intake and its components are prospectively associated with a lower risk of depression and altered abundance of 15 gut microbial species, including enriched Faecalibacterium prausnitzii. In turn, we found a lower abundance of F. prausnitzii and its metabolic pathway, S-adenosyl-L-methionine (SAM) cycle I in participants with depression. To explore causality, we found that lower SAM production by F. prausnitzii may decrease intestinal monoamine oxidase A gene expression implicated in serotonin and dopamine synthesis.

Conclusions: These data underscore the role of diet in the prevention of depression and offer a plausible explanation for how the intestinal microbiome modulates the influence of citrus on mental health. Video Abstract.

Keywords: Citrus fruits; Depression; Gut microbiome; Metabolomics; Metagenomics; Transcriptomics.

Fig. 1

Identification of gut microbial species and metabolic pathways that may modulate relationships between citrus intake and depression. To understand the role of the gut microbiome in modulating the association between citrus intake and depression, we primarily leveraged data from the prospective Nurses’ Health Study II (NHS2), comprised of 32,427 middle-aged women, free of depression at baseline, who provided detailed depression, health, and dietary data every 2–4 years between 2003 and 2017. From 2013 to 2014, a nested sub-study of 207 women from the NHS2, the Mind Body Study (MBS), collected up to four stool samples, a fasting blood sample, dietary information, and detailed depression symptom questionnaires. Stool samples were then profiled through metagenomics (MGX), and blood was analyzed by untargeted metabolomics (MBX). In the analysis phase, we first prospectively assessed associations between citrus consumption with risk of depression and gut microbial species. Then, we related these species to depression status. Finally, we identified microbial metabolic pathways that could explain how microbial modulation of diet could plausibly influence mental health. Our findings were validated in the Men’s Lifestyle Validation Study (MLVS), a parallel cohort of men to the MBS

Fig. 2

Citrus intake is inversely associated with depression, possibly via modulation of F. prausnitzii. A Greater intake of citrus is prospectively associated with a decreased risk of depression (p-trend 0.001) after adjustment for covariates including age, BMI, activity, smoking status, menopausal hormone therapy, total caloric intake, intake of alcohol, comorbidities (history of diabetes mellitus, hypertension, and dyslipidemia), social network level, median family income, overall diet quality, and sleep hours. B Global gut microbial taxonomic profiles vary according to the presence/absence of citrus intake (R.² = 0.005, P = 0.001, adonis2). C Citrus consumption (assessed in 2011) is prospectively associated with 15 gut microbial species (assessed in 2013, linear mixed effects model, FDR q < 0.25). A circular cladogram of all species using GraPhlAn2[38] revealed that certain microbial species, such as Faecalibacterium prausnitzii and Bifidobacterium longum, showed an increase in abundance, while others, such as Acidaminococcus intestini and Parabacteroides merdae, exhibited a decrease. D Participants with depression have a lower relative abundance of F. prausnitzii compared to those without depression (linear mixed effects model, FDR q = 0.05, adjusted for age, BMI, calorie intake, antibiotics, alcohol, Bristol stool score, and diet quality. E In the independent MLVS cohort, greater abundance of F. prausnitzii was associated with a composite biomarker score for depression (β = 5.78, p-value 0.04)

Fig. 3

Metabolic pathways by which F. prausnitzii may influence depression. A Volcano plot demonstrating metagenomic carriage by 4098 pathways from all microbial species in relation to depression. S-Adenosyl-L-methionine cycle I pathway, encoded by F. prausnitzii, was among the strongest hits. B S-Adenosyl-L-methionine cycle I pathway of F. prausnitzii is reduced in depression (generalized linear mixed effects model, FDR q = 0.01). C In the SAM cycle pathway, S-adenosyl-L-homocysteine (SAH) is hydrolyzed to S-ribosyl-L-homocysteine by adenosylhomocysteine nucleosidase (EC 3.2.2.9) and then converted to L-homocysteine by S-ribosylhomocysteine lyase (EC 4.4.1.21). L-homocysteine is methylated to L-methionine using a methyl group from methylated folate, catalyzed by cobalamin-independent methionine synthase (EC 2.1.1.14). The SAM cycle is completed with the regeneration of SAM by SAM synthetase (EC 2.5.1.6). Abundance of the EC 3.2.2.9, EC 4.4.1.21, and EC 2.5.1.6 were significantly associated with depression (FDR q 0.004, 0.006, and 0.004, respectively). D Naringenin, a flavonoid specific to citrus fruit, was associated with an increased abundance of SAM synthetase, the rate-limiting step in SAM production (β = 0.00001, p-value = 0.038). E Greater abundance of S-adenosyl-L-methionine pathway was associated with lower monoamine oxidase A (MAOA) gene expression in the colon, which plays a role degradation of serotonin and dopamine