Galectin 9 Levels as a Potential Predictor of Intact HIV Reservoir Decay

Abstract

Background: During antiretroviral therapy (ART), the HIV reservoir shows variability, with cells carrying intact genomes decaying faster than those with defective genomes, particularly in the first years. The host factors influencing this decay remain unclear.

Methods: Observational study of 74 PWH on ART, 70 (94.6%) of whom were male. Intact proviruses were measured using the intact proviral DNA assay, and 32 inflammatory cytokines were quantified using Luminex immunoassay. Linear spline models assessed the impact of baseline cytokine levels and their trajectories on intact HIV kinetics over seven years.

Results: Baseline Gal-9 was the strongest predictor, with lower levels predicting faster decay. A 10-fold decrease in baseline Gal-9 correlated with a 45% (95% CI, 14%-84%) greater annual decay of intact HIV genomes. Higher baseline interferon-inducible T-cell α chemoattractant (ITAC), interleukin 17 (IL-17), and macrophage inflammatory protein 1α (MIP-1α) levels also predicted faster decay. Longitudinal increases in MIP-3α and decreases in IL-6 were linked to a 9.5% and 10% faster decay, respectively.

Conclusions: The association between lower baseline Gal-9 and faster intact HIV decay suggests targeting Gal-9 could enhance reservoir reduction. The involvement of MIP-3α and IL-6 highlights a broader cytokine regulatory network, suggesting potential multi-targeted interventions.

Keywords: HIV persistence; HIV reservoir; cytokines; galectin 9; inflammation.

Figure 1.

Network plot illustrating baseline correlations (Spearman's ρ coefficients) between the frequency of intact provirus (IntactFreq) and 3′ (Defect3Freq) and 5′ (Defect5Freq) defective proviruses and baseline cytokine levels. Only those correlations with ρ > 0.3 are represented. Data were obtained from 74 samples, and analysis was run in duplicate.

Figure 2.

HIV genome trajectories: intact (left panel) and defective (right panel). Data were obtained from 74 samples.

Figure 3.

Effect of 10-fold higher baseline cytokine levels on genome trajectories: A, intact; B, defective. We fitted separate mixed models for each cytokine. The predictors were the log₁₀-transformed cytokine concentrations at the time of first available sample during viral suppression. The outcomes were the intact or defective HIV DNA copies per million cells (log_e transformed). The x-axis represents the estimated effect on rate of change of intact proviruses per year, after applying the formula 100 × (exp[coefficient] − 1). A, For each galectin 9 (Gal-9) 10-fold increase at baseline, there was a mean 45% (95% CI, 14%–84%) greater increase of intact HIV genomes per year. This corresponds to 45% faster decay for each 10-fold lower baseline Gal-9. B, For each 10-fold increase of Gal-9 and macrophage inflammatory protein 2 (MIP-2) over time, we observed a mean 9.4% (95% CI, −1.4% to 21.3%) and 2.7% (95% CI, −5.4% to 5.9%) faster increase of defective HIV genomes per year. This corresponds to 9.4% and 2.7% faster decay for each 10-fold lower baseline Gal-9 and MIP-3α. In the opposite direction, for each 10-fold higher CD4/CD8 ratio, there was 6.9% (95% CI, .4–12.9) faster decay of defective genomes per gear. Data were obtained from 74 individuals providing 196 observations. Cytokine measurements were run in duplicate.

Figure 4.

Effects of cytokine changes on genome trajectories: A, intact; B, defective. We fitted separate mixed models for each cytokine. The predictors were the log₁₀-transformed cytokine concentrations longitudinally measured during viral suppression. The outcome was the intact or defective HIV DNA copies per million cells (log_e transformed). The x-axis represents the change in rate of change of intact proviruses per year, after applying the formula 100 × (exp[coefficient] − 1). A, For each 10-fold increase of interleukin 6 (IL-6) over time, there was a mean 10.0% (95% CI, .3%–20.6%) faster increase of intact HIV genomes. This corresponds to a 10% faster decrease of intact HIV per each 10-fold decrease of IL-6 over time. In the opposite direction, for each 10-fold decrease of macrophage inflammatory protein 3α (MIP-3α) over time, intact HIV decreased 9.5% (95% CI, 1.5%–16.9%) faster per year. B, For each 10-fold decrease of the following variables, the mean faster decays of defective HIV were 7.9% (95% CI, .8%–14.5%) for CD4/CD8 ratio, 3.4% (95% CI, .1%–7.1%) for CD4+ T-cell counts, 2.1% (95% CI, .1%–4.1%) for IL-17a, 2.2% (95% CI, .4%–3.9%) for MIP-1a, and 2.2% (95% CI, .4%–4.0) for granulocyte-macrophage colony-stimulating factor. Data were obtained from 74 individuals providing 196 observations. Cytokine measurements were run in duplicate.