Secondary Logo

Journal Logo


The microbial metabolite trimethylamine-N-oxide in association with inflammation and microbial dysregulation in three HIV cohorts at various disease stages

Missailidis, Catharinaa; Neogi, Ujjwala,b; Stenvinkel, Peterc; Trøseid, Mariusd,e,f; Nowak, Piotrg,*; Bergman, Petera,*

Author Information
doi: 10.1097/QAD.0000000000001813
  • Free



Despite efficient antiretroviral treatment (ART), HIV-1-infected individuals are at increased risk of cardiovascular disease (CVD) [1], where chronic immune activation and inflammation secondary to HIV-induced translocation of microbial products and metabolites are potential contributory factors. Recent evidence suggests a linkage between compositional changes of intestinal microbiota and dysregulation of the gut immune barrier function in HIV [2–5]. Of note, several studies have observed decreased microbiome diversity in HIV patients that is directly correlated with increased immune activation [3,6–8].

Trimethylamine-N-oxide (TMAO) has recently been identified as a microbial metabolite with pro-atherosclerotic properties [9,10]. Elevated levels of TMAO have been associated with increased cardiovascular risk [9–14]. TMAO is the primary metabolite of trimethylamine (TMA) that is produced by gut bacteria and further converted to TMAO by flavin-containing monooxygenase (FMO3) in the liver [10,13,15,16]. Diets rich in red meat, eggs and full-fat dairy products provide substrate for TMA conversion and may subsequently increase TMAO levels. Although plasma levels of TMAO are governed by diet, FMO3 enzyme activity and renal clearance, growing evidence suggests that the gut microbial composition is directly linked to circulating TMAO levels in a healthy population [10,15,17,18]. Today, there are two main bacterial pathways generating TMA through degradation of dietary substrates in humans: the CutC/CutD pathway [19,20] and the Cnt A/B pathway [21,22]. Notably, genetic components for TMA conversion are detected in Firmicutes, Proteobacteria and Actinobacteria, but appears to be absent in Bacteroidetes [20,23].

Considering the potential role of gut microbiota in HIV pathogenesis and the demonstrated effects of TMAO on the atherosclerotic process, we hypothesized that TMAO production would increase over time as a result of HIV-associated dysbiosis, thus presenting a link between gut microbiota, systemic inflammation and cardiovascular risk in HIV. Exploring this hypothesis, we assessed levels of TMAO in two longitudinal cohorts of primary and chronic HIV-1-infected individuals before and after ART initiation in relation to systemic inflammation and microbial composition of the gut. The role of TMAO in relation to cardiovascular comorbidity was also assessed in a cross-sectional cohort of HIV-1-infected individuals on long-term effective ART.


Study design

We studied HIV1-infected individuals from two longitudinal studies (before and after ART) [8,24] (Table 1) and one cross-sectional study cohort [25] in comparison to healthy controls (Table 2). Circulating TMAO, markers of systemic inflammation, viral load and cellular immunity were assessed in plasma and whole blood. Additionally, in one cohort (chronic HIV-1) [8], the bacterial composition in stool was assessed by high-throughput 16SrRNA sequencing at baseline and after initiation of ART.

Table 1
Table 1:
Patient characteristics in ART-naive HIV cohorts.
Table 2
Table 2:
Patient characteristics in long-term ART treated cohort.

Study cohorts

Primary HIV-infected cohort

Seventeen participants with primary HIV-1 infection (participants of the Quest study) [24,26] were included with sampling within 2 weeks of HIV acquisition and 3–4 months after introduction of ART.

Chronic HIV-1 cohort

Twenty-two participants were included from a study on microbiota and immune status in treatment-naive chronic HIV-1-patients [8] followed for a median of 10 months (IQR 4–15) after ART initiation. Patients were recruited from the HIV outpatient clinic at Karolinska University Hospital, Stockholm, Sweden. Nine sex-matched and age-matched HIV-1-negative individuals were included, consisting of household members and partners of the patients.

Chronic antiretroviral therapy cohort

One hundred and one HIV-1 patients on long-term ART (median 9 years, IQR 5–15) were included from a cross-sectional study of vitamin D levels in HIV-1-infected individuals from the HIV outpatient clinic at South General Hospital, Stockholm, Sweden, 2012 [25]. Six-year follow-up data on time to first cardiovascular events (unstable angina, acute myocardial infarction, transitory ischemic attack and stroke) were registered from medical files. HIV-1-negative healthy controls (n = 30) consisting of staff at the Karolinska University Hospital were included from the same study.

As an external control, 23 sex-matched and age-matched HIV-1-negative individuals were selected from a previously presented population-based cohort in the Stockholm region of Sweden [27].


All participants provided written informed consent and the study was performed in accordance with the Declaration of Helsinki and with regulations from Karolinska Institutet. The Regional Ethical Committee, Stockholm, Sweden reviewed and approved the study protocols and amendments (Karolinska Institutet: Dnr 98-015, 2009-1485-31-3, 2011-1383-31-3 and 244-001),

Plasma HIV-1 RNA quantification and CD4+ cell counts

Analyses of CD4+ cell counts and plasma HIV-1 RNA load were performed at the Karolinska University Laboratory, Stockholm, Sweden with flow cytometry and Cobas Amplicor (Roche Molecular Systems Inc, Branchburg, New Jersey, USA), respectively.


Analysis of plasma levels of TMAO was performed by liquid chromatography/tandem mass spectrometry (LC/MS/MS) at Swedish Metabolomics Centre, Umeå, Sweden utilizing a protocol previously described [27] (PHI, chronic ART with controls and sex-matched and age-matched controls), and in the laboratory of Bevital AS (; Chronic HIV-1 cohort with controls).

Soluble inflammation markers

Limulus Amebocyte Lysate (LAL; Lonza; Basel, Switzerland) and Human sCD14 Quantikine ELISA (R&D, Minneapolis, Minnesota, USA) were used to measure plasma lipopolysaccharide (LPS) and soluble CD14 (sCD14), as described [8,25]

LPS-binding protein (LBP; Hycult Biotech, Uden, The Netherlands), and D-dimer (Technoclone, Vienna, Austria) were determined in plasma samples by ELISA according to the manufacturers’ instructions.

Microbiota analysis

Stool samples were collected and stored at −80°C until further use. Microbiota composition was determined by 16S rRNA sequencing targeting V3–V4 region of 16S gene using Illumina MiSeq and analysed using the established bioinformatics pipeline as described previously [8]. A post hoc analysis was performed with a special focus on gut bacteria with the propensity to produce TMA (defined by in-vitro studies and/or genomic presence of the TMA-producing enzymes CutC/D and Cnt A/B) [19,20,22,23,28,29] (Table 3).

Table 3
Table 3:
Bacteria with the propensity to produce trimethylamine.

Statistical analyses

Data were expressed as median (10th–90th percentile or range) or n (percentage) as appropriate. Univariate comparisons between groups were performed using Mann–Whitney U test for continuous variables. Wilcoxon matched-pairs signed rank test was used to evaluate differences between time-points. Spearman correlation was applied to explore associations between continuous variables.

Determinants of TMAO (baseline, follow-up and delta values) in the PHI and chronic HIV-1 cohorts were assesses by use of repeated measure ANOVA (individually controlling for male sex Caucasian origin, MSM and protease inhibitor use) and simple linear regression (age, CD4+ cell count, CD4+/CD8+ ratio, viral load and inflammatory markers). Multiple linear regression was used to assess determinants of TMAO in the chronic ART cohort [adjusting for age, MSM, estimated glomerular filtration rate (eGFR), years with HIV, ART years, and years with undetectable viral load]. Cox proportional hazard regression was used to assess if baseline TMAO predicted risk of cardiovascular events.

P < 0.05 was considered as statistically significant. Statistical analyses were performed using statistical software SPSS, version 23 (IBM, Armonk, New York, USA).


Patient characteristics and demographics

The median age was similar in the PHI and the chronic HIV-1 cohort with healthy controls (Table 1). The PHI cohort was dominated by MSM of Caucasian origin, whereas the chronic HIV-1 cohort had a more heterogeneous distribution of sex, ethnic origin and sexual orientation (Table 1). All patients in the PHI cohort were started on protease inhibitors. In the chronic HIV-1 cohort, 65% were started on a protease inhibitor and 35% on a nonnucleoside reverse-transcriptase inhibitor (NNRTI)-based regiment (Table 1). Patients in the chronic ART cohort were significantly older than their healthy controls (P = 0.001) and were predominantly MSM of Caucasian origin (Table 2). Twenty-four percent of the HIV-1-infected individuals were on a protease inhibitor and 54% on a NNRTI-containing regiment (Table 2).

Trimethylamine-N-oxide levels were decreased in treatment-naive HIV-1-individuals and increased to normal levels with treatment

Levels of TMAO were lower in untreated HIV individuals and increased significantly after ART, both in primary infected (median 1.28 vs. 2.30 μmol/l, P = 0.040) and chronic HIV-infected individuals (median 2.09 vs 4.47 μmol/l, P < 0.001; Fig. 1a). Despite increased TMAO levels at follow-up in the chronic HIV-1 cohort, levels did not differ significantly from the corresponding healthy controls (P = 0.075; Fig. 1a).

Fig. 1
Fig. 1:
Distribution of plasma trimethylamine-N-oxide levels and correlation to bacterial genera in faecal samples.TMAO levels in primary infected and chronic HIV-1-infected individuals followed from baseline (BL) to follow-up (FU) after 4–10 months ART treatment, compared with HIV-1-infected individuals on long-term ART and HIV-1-negative controls (a). P values are generated by Wilcoxon signed-rank (BL-FU) and Mann–Whitney U test. Spearman correlation heat map of plasma TMAO and relative abundance of genera in faecal sample of HIV-1-patients at baseline (b). Only 15 patients at follow-up because of missing faecal samples. Only bacterial genera with more than 0.001% abundance present in at least 75% of all patient samples were included. ART, antiretroviral therapy; TMAO, trimethylamine-N-oxide.* P value <0.01, ** P value <0.001.

Notably, chronic HIV-1-individuals with longstanding ART had similar levels as their healthy controls (P = 0.89). To account for the difference in age between the chronic ART cohort and their controls, we introduced an age-matched and sex-matched, HIV-1-negative control group (Table 2). Compared with the age-matched and sex-matched controls, chronic ART individuals had significantly lower TMAO levels (P = 0.008) and higher eGFR (P < 0.001; Fig. 1a). However, a one-way ANCOVA conducted to compare TMAO levels found no significant difference in mean TMAO levels (F (1, 106) = 0.563, P = 0.455) between the two cohorts whenever controlling for eGFR.

Plasma trimethylamine-N-oxide and correlations with bacterial composition in stool

Spearman correlation analysis between bacterial taxa and TMAO levels revealed several interesting findings at different taxonomical levels. Thus, at baseline, we found that TMAO levels were inversely correlated with Bacteroidetes (Rho: −0.62, P = 0.002), and positively correlated with Firmicutes (Rho: 0.65, P = 0.001). Additionally, baseline plasma TMAO levels were inversely correlated with the genera Alistipes (Rho: −0.47, P = 0.029), Bacteroides (Rho: −0.58, P = 0.004), Odoribacter (Rho: −0.60, P = 0.003), Parabacteroidetes (Rho: 0.52, P = 0.014) and positively correlated with Prevotella (Rho 0.44, P = 0.039); all members of the Bacteroidetes phyla. Furthermore, the genus Suturella, a member of the Proteobacteria phylum was inversely correlated (Rho: −0.64, P = 0.001; Fig. 1b). However, at follow-up, these associations were no longer present and only a positive association between TMAO and Dialister spp. (Rho: 0.64, P = 0.007), member of the Firmicutes phylum, was observed (Fig. 1b).

Relative abundance of trimethylamine-producing bacteria did not change with antiretroviral therapy initiation and did not correlate with circulating trimethylamine-N-oxide or systemic inflammation

As plasma levels of TMAO are partly regulated by TMA-producing gut bacteria, we further assessed the relative abundance of bacteria with the propensity to produce TMA before and after initiation of ART at genus level (Table 3). The relative abundance of TMA-producing bacteria did not significantly increase after ART, but remained similar to controls (data no shown). Moreover, changes in the relative abundance of TMA-producing bacteria over time (ΔTMA = follow-up − baseline levels) neither correlate with the observed changes in circulating TMAO nor with changes in circulating sCD14, LPS, LBP, and D-dimer (data not shown).

Baseline CD4+/CD8+ ratio, but not viremia, inflammation or use of antiretroviral therapy, predicted increased levels of trimethylamine-N-oxide at follow-up

To assess if immune status, HIV viremia and inflammation predicted TMAO levels at baseline and follow-up, we assessed the individual cohorts by repeated measure ANOVA and linear regression analysis.

In the PHI and chronic HIV-1 cohorts, TMAO levels did not associate with male sex, Caucasian origin or MSM status. Age, viral load, CD4+ cell count or CD4+/CD8+ ratio did not predict TMAO levels in either cohort, nor did they associate with change in TMAO levels (ΔTMAO: follow-up − baseline levels) with one exception: in the chronic HIV-1 cohort, we found that higher baseline CD4+/CD8+ ratio predicted increased ΔTMAO (B = 11.92, P = 0.016). None of the inflammatory markers (sCD14, LPS, LBP, D-dimer) predicted TMAO levels or ΔTMAO in the chronic HIV-1 cohort. Moreover, in the chronic HIV-1 cohort, neither use of protease inhibitor nor NNRTI predicted follow-up TMAO levels, or ΔTMAO.

In the chronic ART cohort, multiple regression analysis demonstrated that neither preART-VL, CD4+ cell count, CD4+/CD8+ ratio, nor baseline CD4+ cell count or CD4+/CD8+ ratio predicted TMAO, whenever controlling for age, MSM, eGFR, years with HIV, years of ART and years with undetectable viral load. Similarly, markers of baseline inflammation (hsCRP, sCD14 and LPS) neither predict TMAO levels nor did use of protease inhibitor or NNRTI have a significant effect on TMAO levels.

Higher trimethylamine-N-oxide levels did not infer a significantly increased risk for cardiovascular events

During 6 years follow-up, there were six registered cardiovascular events in the chronic ART cohort. Although the study was not powered to predict cardiovascular events, the highest tertile of TMAO (>4.93 μmol/l) was nonsignificantly associated with a 2.8-fold risk [hazard ratio 2.76: 95% confidence interval (CI) 0.29–26.70, P = 0.38].

Lack of increase in plasma trimethylamine-N-oxide levels after antiretroviral therapy initiation associated with loss of Bacteroidetes and increased lipopolysaccharide levels

Intrigued by the strong association between baseline CD4+/CD8+ ratio and increased TMAO levels after ART initiation in the chronic HIV-1 cohort we further assessed the cohort with regards to changes in TMAO-levels.

We found that increased TMAO levels after ART initiation (ΔTMAO >1 μmol/l), were present in a subgroup of 10 HIV individuals whereas no increase (Δ TMAO <1 μmol/l) was observed in 11 individuals (Fig. 2a). Individuals with increased TMAO levels had significantly higher CD4+/CD8+ ratio at baseline (Fig. 2b) compared with patients with no increase. The two groups had a similar age and sex distribution but in the group with increased TMAO levels, more individuals were treated with protease inhibitor (80%), compared with the group with no increase (50%; P = 0.210). On the basis of the hypothesis that lower CD4+/CD8+ ratio may reflect a more dysregulated gut microbiota, we compared the microbial composition in individuals with increase versus no increase in ΔTMAO levels in relation to corresponding healthy controls.

Fig. 2
Fig. 2:
Differential change in plasma trimethylamine-N-oxide levels after antiretroviral therapy initiation (ΔTMAO: follow-up – baseline <1) (a) in relation to CD4+/CD8+ ratio at baseline (b), microbial composition (c) and and LPS levels at follow-up (d).Number of participants is based on available samples. P values are generated by Wilcoxon signed-rank test (a) and Mann–Whitney U test (b and d). LPS, lipopolysaccharide.

We found that treatment-naive HIV-individuals were similar to controls in relative abundance of Firmicutes, Bacteroidetes, Proteobacteria and Actinobacteria. After ART initiation, HIV individuals with increased ΔTMAO remained stable in their composition of gut microbiota, whereas individuals with no increase demonstrated a significant reduction in relative abundance of Bacteroidetes (P = 0.023) and a corresponding nonsignificant expansion of Firmicutes and Proteobacteria compared with controls (Fig. 2c).

Moreover, after ART initiation, HIV individuals with no increase had significantly higher levels of LPS compared with individuals with increased ΔTMAO (P = 0.01; Fig. 2d). Of note, LPS levels correlated inversely with relative abundance of Bacteroidetes at follow-up (Rho –0.90, P = 0.037).


Our study explored circulating TMAO levels in primary and chronic HIV-1 before and after ART and in HIV-1-infected individuals on long-term effective ART. In support of our original hypothesis we found that TMAO levels increased over time with ART. However, regardless of treatment length, TMAO levels in HIV-1-infected individuals remained similar to healthy controls and were not significantly affected by treatment regime, immune status or degree of systemic inflammation. Notably gut dysbiosis was more pronounced in patients without increase in TMAO levels after ART initiation.

The observation that HIV-1-infected individuals had similar TMAO levels as healthy controls has been reported by other studies assessing circulating TMAO and CVD in HIV [32–34]. Contrary to population-based studies [9–11,13] these studies provide no conclusive evidence that elevated TMAO has a strong impact on atherosclerotic burden, prevalent CVD or adverse cardiac events in HIV-infected individuals [32–35]. Similarly, although the study was not powered to detect relation to cardiovascular events, we found no significant risk with increasing TMAO levels (P = 0.38) in well controlled HIV-1-infected individuals on long-term ART. Furthermore, our observation suggesting that ART initiation increases TMAO levels was previously reported in a nested case–control study by Haissman et al.[33].

Why does trimethylamine-N-oxide increase after antiretroviral therapy?

It could be argued that the effect might be related to dietary changes and the return of appetite after viral control is established through treatment. However, this argument remains purely hypothetical as we lack data on diet and general well being. Moreover, fatigue and loss of appetite is more pronounced in advanced HIV disease that is not clinically represented in the two study cohorts at baseline.

It could also be argued that ART favours a more TMA-prone microbiome. However, we could not find evidence supporting an expansion of TMA-producing bacteria after ART initiation in chronic HIV-1-infected individuals. Notwithstanding, the data concerning TMA production of gut bacteria are not complete and any such analysis should be interpreted cautiously. Notably, phylogeny appears to be a poor predictor of TMA production [29].

Despite the lack of association between TMA-producers and systemic TMAO-levels after ART we found a positive correlation between plasma TMAO and Firmicutes at baseline and at follow-up and an inverse correlation with members of Bacteroidetes at baseline with the exception of Prevotella spp. Our observations support published data demonstrating a widespread TMA production propensity among members of Proteobacteria and Firmicutes that appears to be missing in Bacteroidetes [19,20,22,23,29]. Further, several studies have observed a positive association between Prevotella spp. and TMAO levels [10,36,37], but the underlying mechanisms remain to be elucidated.

Furthermore, although there is no established link between FMO3 and HIV-1, it could be argued that FMO3 activity might be affected by ART. Of note, a recent publication presented evidence that FMO3 was involved in metabolizing F18, a new experimental NNRTI [38]. To our knowledge, no studies have assessed the role of FMO3 in metabolizing protease inhibitor, or commercially available NNRTI. Whereas Haissman et al.[33] observed that protease inhibitor, but not NNRTI, was associated with increased TMAO levels following ART initiation, we found no significant impact of choice of protease inhibitor or NNRTI on TMAO levels at follow-up (P = 0.17) or changes in TMAO levels (P = 0.21) in the chronic HIV-cohort. Nor did protease inhibitor (P = 0.99) or NNRTI (P = 0.59) predict TMAO levels in the chronic ART cohort. However, although no significant association between protease inhibitor treatment and TMAO levels was established in the chronic HIV-1 cohort, protease inhibitor was the dominating ART in the group with increased TMAO levels at follow-up.

Finally, it is possible that the decreased TMAO levels observed in treatment-naive HIV-1-infected individuals relate to down-regulation of the TMA-converting enzyme FMO3 by gut-related inflammation, as observed in a model of LPS-induced sepsis in mice [39]. Indirect support of this hypothesis is provided by a recent clinical study, which reported reduced TMAO levels in IBD patients with active disease compared with inactive disease and healthy controls [40]. Moreover, we found a more pronounced loss of Bacteroidetes and higher levels of LPS in chronic HIV-1-infected individuals who did not increase their TMAO levels after ART despite viral control. Loss of Bacteroidetes and increase in Proteobacteria has been linked to increased microbial translocation and systemic inflammation in both HIV and inflammatory bowel disease (IBD) [41,42]. Finally, chronic HIV-1-infected individuals with no increase in TMAO had lower CD4+/CD8+ ratios at baseline, suggesting that lack of gain in TMAO might serve as a surrogate-marker for a more dysregulated gut immune barrier function that prevails despite effective ART.

FMO3 downregulation may also be attributed to reduction of the bile acid-activated Farnesoid X receptor (FXR). FXR modulates bile acid homeostasis and inflammation in the gut [43–45], and is involved in inducing FMO3 in mice models [16,36,46]. Importantly, in HIV patients, downregulation of FXR expression has been observed in monocytes and hepatocytes despite functional ART [47,48]. However, the link between FXR-expression and TMAO levels is not yet established in clinical cohorts.

The strength of our study is the consistency in trend observed both in longitudinal and cross-sectional study cohorts representing both primary HIV and individuals with chronic HIV infection at different treatment length and matching healthy controls. Moreover, longitudinal data on microbiome composition provided a unique opportunity to assess its relation to TMAO. The limitations of the study include the relatively low number of patients in each cohort, lack of power to study cardiovascular event rate, lack of information regarding MSM status in the control cohorts and lack of dietary data. However, the healthy controls of the chronic HIV-1 cohort consisted of household members and partners of the patients indicating similar sexual orientation and food intake.

In conclusion, although we cannot exclude that elevated TMAO may contribute to cardiovascular risk in general, our data do not support the hypothesis that TMAO is a significant link between gut dysbiosis, inflammation and cardiovascular risk in HIV-1-infected individuals. TMAO levels appear to be disparately regulated by HIV infection, ART and microbiota, thus limiting its role as a cardiovascular risk marker in HIV-infected individuals.


The study was supported by Stockholm's county council (SLL-KI), Swedish Research Council.

Author contributions: C.M, P.N. and P.B. conceived and designed the experiments; C.M, P.S, P.N. and P.B. coordinated the sample collection. C.M. performed the experiments. C.M, U.N, P.S, M.T, P.N. and P.B. analysed the data. C.M. reviewed and all the authors edited the article.

Conflicts of interest

There are no conflicts of interest.


1. Schouten J, Wit FW, Stolte IG, Kootstra N, van der Valk M, Geerlings SG, et al. AGEhIV Cohort Study GroupCross-sectional comparison of the prevalence of age-associated comorbidities and their risk factors between HIV-infected and uninfected individuals: the AGEhIV Cohort Study. Clin Infect Dis 2014; 59:1787–1797.
2. Vujkovic-Cvijin I, Dunham RM, Iwai S, Maher MC, Albright RG, Broadhurst MJ, et al. Dysbiosis of the gut microbiota is associated with hiv disease progression and tryptophan catabolism. Sci Transl Med 2013; 5: 193ra191.
3. Dillon SM, Lee EJ, Kotter CV, Austin GL, Dong Z, Hecht DK, et al. An altered intestinal mucosal microbiome in HIV-1 infection is associated with mucosal and systemic immune activation and endotoxemia. Mucosal Immunol 2014; 7:983–994.
4. Mutlu EA, Keshavarzian A, Losurdo J, Swanson G, Siewe B, Forsyth C, et al. A compositional look at the human gastrointestinal microbiome and immune activation parameters in HIV infected subjects. PLoS Pathog 2014; 10:e1003829.
5. Perez-Santiago J, Gianella S, Massanella M, Spina CA, Karris MY, Var SR, et al. Gut Lactobacillales are associated with higher CD4 and less microbial translocation during HIV infection. AIDS (London, England) 2013; 27:1921–1931.
6. Dinh DM, Volpe GE, Duffalo C, Bhalchandra S, Tai AK, Kane AV, et al. Intestinal microbiota, microbial translocation, and systemic inflammation in chronic HIV infection. J Infect Dis 2015; 211:19–27.
7. Ellis CL, Ma ZM, Mann SK, Li CS, Wu J, Knight TH, et al. Molecular characterization of stool microbiota in HIV-infected subjects by panbacterial and order-level 16S ribosomal DNA (rDNA) quantification and correlations with immune activation. J Acquir Immune Defic Syndr 2011; 57:363–370.
8. Nowak P, Troseid M, Avershina E, Barqasho B, Neogi U, Holm K, et al. Gut microbiota diversity predicts immune status in HIV-1 infection. AIDS (London, England) 2015; 29:2409–2418.
9. Wang Z, Klipfell E, Bennett BJ, Koeth R, Levison BS, Dugar B, et al. Gut flora metabolism of phosphatidylcholine promotes cardiovascular disease. Nature 2011; 472:57–63.
10. Koeth RA, Wang Z, Levison BS, Buffa JA, Org E, Sheehy BT, et al. Intestinal microbiota metabolism of L-carnitine, a nutrient in red meat, promotes atherosclerosis. Nat Med 2013; 19:576–585.
11. Troseid M, Ueland T, Hov JR, Svardal A, Gregersen I, Dahl CP, et al. Microbiota-dependent metabolite trimethylamine-N-oxide is associated with disease severity and survival of patients with chronic heart failure. J Intern Med 2015; 277:717–726.
12. Tang WH, Wang Z, Shrestha K, Borowski AG, Wu Y, Troughton RW, et al. Intestinal microbiota-dependent phosphatidylcholine metabolites, diastolic dysfunction, and adverse clinical outcomes in chronic systolic heart failure. J Card Fail 2015; 21:91–96.
13. Tang WH, Wang Z, Levison BS, Koeth RA, Britt EB, Fu X, et al. Intestinal microbial metabolism of phosphatidylcholine and cardiovascular risk. N Engl J Med 2013; 368:1575–1584.
14. Stubbs JR, House JA, Ocque AJ, Zhang S, Johnson C, Kimber C, et al. Serum trimethylamine-n-oxide is elevated in CKD and correlates with coronary atherosclerosis burden. J Am Soc Nephrol 2016; 27:305–313.
15. Miller CA, Corbin KD, da Costa KA, Zhang S, Zhao X, Galanko JA, et al. Effect of egg ingestion on trimethylamine-N-oxide production in humans: a randomized, controlled, dose-response study. Am J Clin Nutr 2014; 100:778–786.
16. Bennett BJ, de Aguiar Vallim TQ, Wang Z, Shih DM, Meng Y, Gregory J, et al. Trimethylamine-N-oxide, a metabolite associated with atherosclerosis, exhibits complex genetic and dietary regulation. Cell Metab 2013; 17:49–60.
17. Cho CE, Taesuwan S, Malysheva OV, Bender E, Tulchinsky NF, Yan J, et al. Trimethylamine-N-oxide (TMAO) response to animal source foods varies among healthy young men and is influenced by their gut microbiota composition: a randomized controlled trial. Mol Nutr Food Res 2017; 61:1–10.
18. Gregory JC, Buffa JA, Org E, Wang Z, Levison BS, Zhu W, et al. Transmission of atherosclerosis susceptibility with gut microbial transplantation. J Biol Chem 2015; 290:5647–5660.
19. Martinez-del Campo A, Bodea S, Hamer HA, Marks JA, Haiser HJ, Turnbaugh PJ, et al. Characterization and detection of a widely distributed gene cluster that predicts anaerobic choline utilization by human gut bacteria. mBio 2015; 6: pii: e00042-15.
20. Craciun S, Balskus EP. Microbial conversion of choline to trimethylamine requires a glycyl radical enzyme. Proc Natl Acad Sci U S A 2012; 109:21307–21312.
21. Unemoto T, Hayashi M, Miyaki K, Hayashi M. Formation of trimethylamine from DL-carnitine by Serratia marcescens. Biochim Biophys Acta 1966; 121:220–222.
22. Zhu Y, Jameson E, Crosatti M, Schafer H, Rajakumar K, Bugg TD, et al. Carnitine metabolism to trimethylamine by an unusual Rieske-type oxygenase from human microbiota. Proc Natl Acad Sci U S A 2014; 111:4268–4273.
23. Falony G, Vieira-Silva S, Raes J. Microbiology meets big data: the case of gut microbiota-derived trimethylamine. Annu Rev Microbiol 2015; 69:305–321.
24. Barqasho B, Nowak P, Tjernlund A, Kinloch S, Goh LE, Lampe F, et al. Kinetics of plasma cytokines and chemokines during primary HIV-1 infection and after analytical treatment interruption. HIV Med 2009; 10:94–102.
25. Missailidis C, Hoijer J, Johansson M, Ekstrom L, Bratt G, Hejdeman B, et al. Vitamin D status in well controlled Caucasian HIV patients in relation to inflammatory and metabolic markers–a cross-sectional cohort study in Sweden. Scand J Immunol 2015; 82:55–62.
26. Goh LE, Perrin L, Hoen B, Cooper D, Phillips A, Janossy G, et al. Study protocol for the evaluation of the potential for durable viral suppression after quadruple HAART with or without HIV vaccination: the QUEST study. HIV Clin Trials 2001; 2:438–444.
27. Missailidis C, Hallqvist J, Qureshi AR, Barany P, Heimburger O, Lindholm B, et al. Serum trimethylamine-N-oxide is strongly related to renal function and predicts outcome in chronic kidney disease. PloS One 2016; 11:e0141738.
28. Rath S, Heidrich B, Pieper DH, Vital M. Uncovering the trimethylamine-producing bacteria of the human gut microbiota. Microbiome 2017; 5:54.
29. Romano KA, Vivas EI, Amador-Noguez D, Rey FE. Intestinal microbiota composition modulates choline bioavailability from diet and accumulation of the proatherogenic metabolite trimethylamine-N-oxide. mBio 2015; 6:e02481.
30. Cruden DL, Galask RP. Reduction of trimethylamine oxide to trimethylamine by mobiluncus strains isolated from patients with bacterial vaginosis. Microbial Ecol Health Dis 1988; 1:95–100.
    31. Robinson J, Gibbon NE, Thatcher S FS. A mechanism of halophilism in Micrococcus halodenitrificans. J Bacteriol 1952; 64:69–77.
    32. Srinivasa S, Fitch KV, Lo J, Kadar H, Knight R, Wong K, et al. Plaque burden in HIV-infected patients is associated with serum intestinal microbiota-generated trimethylamine. AIDS (London, England) 2015; 29:443–452.
    33. Haissman JM, Knudsen A, Hoel H, Kj AA, Kristoffersen US, Berge RK, et al. Microbiota-dependent marker TMAO is elevated in silent ischemia but is not associated with first-time myocardial infarction in HIV infection. J Acquir Immune Defic Syndr 2016; 71:130–136.
    34. Miller PE, Haberlen SA, Brown TT, Margolick JB, DiDonato JA, Hazen SL, et al. Brief report: intestinal microbiota-produced trimethylamine-N-oxide and its association with coronary stenosis and HIV serostatus. J Acquir Immune Defic Syndr 2016; 72:114–118.
    35. Knudsen A, Christensen TE, Thorsteinsson K, Ghotbi AA, Hasbak P, Lebech AM, et al. Microbiota-dependent marker TMAO is not associated with decreased myocardial perfusion in well treated HIV-infected patients as assessed by 82Rubidium PET/CT. J Acquir Immune Defic Syndr 2016; 72:e83–e85.
    36. Chen ML, Yi L, Zhang Y, Zhou X, Ran L, Yang J, et al. Resveratrol attenuates trimethylamine-N-oxide (TMAO)-induced atherosclerosis by regulating TMAO synthesis and bile acid metabolism via remodeling of the gut microbiota. mBio 2016; 7:e02210–e02215.
    37. De Filippis F, Pellegrini N, Laghi L, Gobbetti M, Ercolini D. Unusual sub-genus associations of faecal Prevotella and Bacteroides with specific dietary patterns. Microbiome 2016; 4:57.
    38. Wu X, Zhang Q, Guo J, Jia Y, Zhang Z, Zhao M, et al. Metabolism of F18, a derivative of calanolide a, in human liver microsomes and cytosol. Front Pharmacol 2017; 8:479.
    39. Zhang J, Chaluvadi MR, Reddy R, Motika MS, Richardson TA, Cashman JR, et al. Hepatic flavin-containing monooxygenase gene regulation in different mouse inflammation models. Drug Metab Dispos 2009; 37:462–468.
    40. Wilson A, Teft WA, Morse BL, Choi YH, Woolsey S, DeGorter MK, et al. Trimethylamine-N-oxide: a novel biomarker for the identification of inflammatory bowel disease. Dig Dis Sci 2015; 60:3620–3630.
    41. Frank DN, St Amand AL, Feldman RA, Boedeker EC, Harpaz N, Pace NR. Molecular-phylogenetic characterization of microbial community imbalances in human inflammatory bowel diseases. Proc Natl Acad Sci U S A 2007; 104:13780–13785.
    42. Gevers D, Kugathasan S, Denson LA, Vazquez-Baeza Y, Van Treuren W, Ren B, et al. The treatment-naive microbiome in new-onset Crohn's disease. Cell Host Microbe 2014; 15:382–392.
    43. Wahlstrom A, Kovatcheva-Datchary P, Stahlman M, Backhed F, Marschall HU. Crosstalk between bile acids and gut microbiota and its impact on farnesoid X receptor signalling. Dig Dis 2017; 35:246–250.
    44. Li F, Jiang C, Krausz KW, Li Y, Albert I, Hao H, et al. Microbiome remodelling leads to inhibition of intestinal farnesoid X receptor signalling and decreased obesity. Nat Commun 2013; 4:2384.
    45. Vavassori P, Mencarelli A, Renga B, Distrutti E, Fiorucci S. The bile acid receptor FXR is a modulator of intestinal innate immunity. J Immunol 2009; 183:6251–6261.
    46. Warrier M, Shih DM, Burrows AC, Ferguson D, Gromovsky AD, Brown AL, et al. The TMAO-generating enzyme flavin monooxygenase 3 is a central regulator of cholesterol balance. Cell reports 2015; 10:326–338.
    47. Renga B, Francisci D, Carino A, Marchiano S, Cipriani S, Chiara Monti M, et al. The HIV matrix protein p17 induces hepatic lipid accumulation via modulation of nuclear receptor transcriptoma. Sci Rep 2015; 5:15403.
    48. Renga B, Francisci D, D’Amore C, Schiaroli E, Carino A, Baldelli F, et al. HIV-1 infection is associated with changes in nuclear receptor transcriptome, pro-inflammatory and lipid profile of monocytes. BMC Infect Dis 2012; 12:274.

    * Piotr Nowak and Peter Bergman shared last authorship, equal contribution.


    HIV-1; lipopolysaccharide; microbiota; trimethylamine-N-oxide

    Copyright © 2018 Wolters Kluwer Health, Inc.