Although ART has achieved long-term viral suppression, incomplete immune reconstitution and chronic non-AIDS-related illnesses have always been concerns in PLWH. Since accompanied metabolic changes occurred during HIV infection, whether they could recover after ART treatment and their relationship with immune reconstitution needs further exploration. In this study, we detected 938 metabolites using metabolomics analysis, and further analyzed 770 of them which are endogenous metabolites in PLWH. The –omics approach provided an unbiased review of the overall metabolic dynamics. Although significantly different metabolite profiles were revealed between HIV seronegative healthy controls and PLWH with years of ART, metabolic profiles were not markedly affected by different immune reconstitution levels. Collective analysis of metabolomics and inflammatory profiling results further indicated potential biomarkers identified in PLWH related to immunological factors and cytokines linked to cardiovascular events.
Chronic HIV infection, the resultant inflammation, and ART toxicities could all contribute to activation of immune systems and metabolic imbalance in the host. In line with other publications, our metabolomics results indicated that HIV patients’ metabolic profiles could be clearly separated from HIV seronegative controls (9, 10, 18). The plasma of untreated HIV patients had decreased levels of glutathione and tryptophan pathway metabolites, and increased levels of fatty acids, sphingomyelin, phospholipid and lysophospholipids comparing to HCs (9, 11, 18). Purine and pyrimidine metabolisms in untreated HIV patients were also markedly different from the healthy controls, which were seldom mentioned in previous studies (12, 19, 20). The changed purine and pyrimidine metabolism could be coming from active HIV virion replication that requires high rates of nucleotide turnover. Elevated levels of many acyl cholines in our study might indicate a higher level of β-oxidative stress in PLWH (15, 21). After ART, several metabolites in pyrimidine metabolism including 2-deoxyuridine and 3-ureidopropionate were restored to the healthy control levels, which other metabolites such as 1-stearoyl-2-oleoyl-GPS (18:0/18:1), maltotetraose, and 5-hydroxyhexanoate were still different from their levels in healthy controls. Levels of UMP, IMP, ADP and ADP-ribose were not restored to the healthy control levels. We hypothesized that the underlying biological mechanisms of these differences might be related to ART medication and suppressed HIV virion replication, which needs to be further investigated in future studies.
Previous studies have demonstrated that metabolic changes by HIV are associated with host immune profiles and biomarkers of chronic inflammation (5, 9, 22). For instance, N. Chantal Peltenburg et al. reported that sphingomyelins and phospholipids were negatively correlated to IP-10 and sIL-2R, and triglycerides were linked to MCP-1 (9). In our study, we found that IL-10, D-dimer, VCAM-1, TNF-RII and ICAM-1 were positively correlated with most of the significantly changed metabolites from HIV infection, but negatively correlated with changed metabolites in nucleotide metabolism. After ART, most of the changed lipids were positively correlated with changes of IL-8 and IL-7, but most of the changed amino acid metabolites were negatively connected with IL-8 and IL-7. It has been reported that accumulation of ICAM-1, TNF-RII, VCAM-1 and D-dimer in vessel wall is a hallmark of atherosclerosis and acute coronary syndrome, and is mediated by the interaction between adhesion molecules on endothelial and circulating cells (23, 24). Proinflammatory cytokine IL-8and inductive cytokine IL-7 were elevated in HIV infection and promote virus replication. The immune-metabolic networks revealed that HIV suppression through ART attenuates the pro-inflammatory status and risk for cardiovascular disease but might contribute to dysregulation of lipid and amino acid metabolisms in the meantime.
The current study has several limitations. Firstly, untargeted metabolomics analysis only provided the relative abundance of detected metabolites, so further quantitative studies are required to confirm the changes of the identified metabolites in the metabolomics study. Secondly, the study was retrospective, so the sample collection procedures and ART treatment might not be well controlled, which may result in nonspecific variations in the metabolomics study results. Other factors such as the duration of HIV infection, smoking habits, timing of sampling, co-infection status or comorbidity were not fully considered in the study, either.
Overall, combining metabolomics analysis and cytokines measurement, we described the plasma metabolic-immune network of HIV patients before and after treatment in comparison with the healthy controls. The results indicated the central role of lipid and nucleotide metabolism disturbance in HIV infection, and found that these metabolic changes could not be fully restored by viral suppressive ART. This was an explosive study and the underlying biological mechanisms of the metabolites described in the study require further validation.
Table 1
Clinical characteristics of patients
Characteristics | INRs N = 25 | IRs N = 25 | HCs N = 25 | p-value |
Age, mean years (SD) | 45.8 ± 11.7 | 43.3 ± 10.3 | 44.7 ± 12.4 | 0.554 |
Male, n (%) | 24 (96.0) | 24 (96.0) | 23 (92.0) | 0.705 |
Route of transmission, n (%) Sexual Blood Unclear/others | 20 (80) 2 (8.0) 3 (12.0) | 19 (76.0) 2 (8.0) 4 (16.0) | - | 0.919 |
Initial ART regimen, (%) 2 NRTIs + NNRTI 2NRTIs + PI 2 NRTIs + INSTI Other | 20 (80.0) 1 (4.0) 4 (16.0) 0 (0.0) | 21 (84.0) 2 (8.0) 1 (4.0) 1 (4.0) | - | 0.368 |
Switching ART regimen, n (%) 2 NRTIs + NNRTI 2NRTIs + PI 2 NRTIs + INSTI Other | 18 (72.0) 4 (16.0) 3 (12.0) 0 (0.0) | 15 (60.0) 4 (16.0) 4 (16.0) 2 (8.0) | - | 0.491 |
Viral load (lg copies/mL) | 5.07 ± 0.73 | 5.13 ± 0.61 | - | 0.752 |
Weight(kg) | 59.8 ± 10.7 | 66.0 ± 10.0 | - | 0.042 |
Baseline CD3 + CD4 + T counts (cells/ul) | 25 (7, 77) | 44 (13, 53) | - | 0.676 |
Peak CD3 + CD4 + T counts (cells/ul) | 244 (201, 277) | 610 (555, 820) | | < 0.001 |
Notes: INR: Immune non-responder; IR: Immune responder; HC: healthy control; ImmoSD: standard deviation NRTIs: Nucleoside reverse-transcriptase inhibitors; NNRTI: Non-Nucleoside reverse-transcriptase inhibitors; PI: Proteinase inhibitor; INSTI: Integrase inhibitor.