Youth Who Control HIV on Antiretroviral Therapy Display Unique Plasma Biomarkers and Cellular Transcriptome Profiles Including DNA Repair and RNA Processing

Abstract

Combination antiretroviral therapy (ART) suppresses detectible HIV-1 replication, but latent reservoirs and persistent immune activation contribute to residual viral-associated morbidities and potential viral reactivation. youth with HIV (YWH) virally suppressed on ART early in infection before CD4 T cell decline with fewer comorbidities compared to adults represent a critical population for identifying markers associated with viral control and predictors of viral breakthrough. This study employed a multi-omics approach to evaluate plasma biomarkers and cellular gene expression profiles in 52 participants, including 27 YWH on ART for 144 weeks and 25 youth with no infection (NI) (ages 18-24). Among the 27 YWH, 19 were virally suppressed (VS; <50 RNA copies/mL), while eight were non-suppressed (VNS; >50 RNA copies/mL). VS YWH displayed unique bioprofiles distinct from either VNS or NI. Early viral suppression mitigates inflammatory pathways and normalizes key biomarkers associated with HIV-related comorbidities. Genes upregulated in pathways linked to cellular homeostasis such as DNA repair, RNA processing, and transcription regulation may diminish viral breakthrough and maintain sustained HIV control on ART. Candidate markers and putative molecular mechanisms were identified, offering potential therapeutic targets to limit viral persistence, enhance HIV treatment strategies, and pave the way for improved clinical outcomes.

Trial registration: ClinicalTrials.gov NCT00491556 NCT00683579.

Keywords: DNA repair; biomarkers; bioprofiles; cellular homeostasis; comorbidities; inflammation; transcriptome; viral suppression; youth with HIV.

Figure 1

Plasma HIV-1 RNA (viral load) and CD4 T cell number for YWH. (a) Viral load distribution at entry and end of study for VS and VNS YWH; (b) CD4 T cell numbers at entry and end of study for VS and VNS YWH.

Figure 2

Unsupervised PCA based on 23 plasma biomarker data. To visualize the variance among the study groups, biomarker data scaled by calculating the z-score was applied to the PCA. The ellipses were generated with 95% confidence intervals for each group. Symbols: dot, participants ellipses, participant clusters. Grey dot: NI; Blue dot: VS; Pink dot: VNS. Grey ellipses: 22 NI with 3 VS and 1 VNS; Blue ellipses: 14 VS with 3 NI and 4 VNS; Pink ellipses: 3 VNS and 2 VS.

Figure 3

Random forest classification model. Beeswarm plot of RF model showing the top fifteen biomarkers that are more predictive of VS compared with VNS (a) or with NI (b). Biomarkers are ranked by their SHAP value in descending order of their relative contribution to the classification model. Each point on the plot represents the biomarker measurement from a single participant. Larger SHAP values correspond to a larger contribution of a biomarker in classification of a participant. The classification model for VS compared with VNS showed an AUC-ROC of 1.0, while VS or VNS compared with NI each showed an AUC-ROC of 0.9.

Figure 4

Differential expression of genes (DEGs) in VS compared with VNS and NI. (a) DEG analysis was performed to compare VS with VNS and NI. Differentially expressed genes (DEGs) showing an absolute FC ≥ 1.3 and FDR ≤ 0.05 were considered significantly altered. Red: upregulated DEGs; blue: downregulated DEGs. (b) DEGs from the two comparisons were plotted as a Venn diagram to study the gene overlap between the two VS comparisons. (c) Fold change (FC) of 13 DEGs shared across all the comparison was plotted as a heatmap and clustered based on FC.

Figure 5

Functional enrichment and network analysis. (a) Functional enrichment analysis was performed to characterize the DEGs in VS compared with VNS or NI using a p value cut off ≤ 0.001. Green bars: Pathways perturbed by DEGs when VS compared with VNS; yellow bars: pathways perturbed by DEGs when VS compared with NI. (b) Network analysis of significant pathways shows DEGs connecting the pathways for VS compared with VNS (left panel) or VS compared with NI (right panel). Nodes: Pathways; edges: DEGs connecting the pathways. Green circles: VS compared with VNS; yellow circles: VS compared with NI; red: upregulated DEGs; blue: downregulated DEGs within each pathway; Star: candidate DEGs associated with viral suppression in both VS comparisons.

Figure 6

Candidate DEGs/genes associated with viral suppression. Genes playing critical roles in regulating molecular mechanisms that contribute to the HIV-1 control at different levels were identified in VS compared with VNS. Genes regulating apoptosis, transcription, pathways involved in post-translation modification (PTM) (such as protein stabilization, phosphorylation, regulation of serine/threonine kinase activity), as well as inflammation (platelet activation) were identified in VS compared with NI. Symbols: Dots—DEGs; grey dots—total DEGs in both VS comparisons; green dots—candidate DEGs identified in VS vs. VNS; yellow dots—candidate DEGs identified in VS vs. NI; candidate DEGs labeled by gene symbol—red: upregulated; blue: downregulated.