NK cell activation and CD4+ T cell α4β7 expression are associated with susceptibility to HIV-1

Abstract

We leveraged specimens from the RV217 prospective study that enrolled participants at high risk of HIV-1 acquisition to investigate how NK cells, conventional T cells, and unconventional T cells influence HIV-1 acquisition. We observed low levels of α4β7 expression on memory CD4+ T cells and invariant NK T (iNKT) cells, 2 cell types highly susceptible to HIV-1 infection, in highly exposed seronegative (HESN) compared with highly exposed seroconverter (HESC) participants. NK cells from HESN individuals had higher levels of α4β7 than did those from HESC individuals, presented a quiescent phenotype, and had a higher capacity to respond to opsonized target cells. We also measured translocated microbial products in plasma and found differences in phylum distribution between HESN and HESC participants that were associated with the immune phenotypes affecting the risk of HIV-1 acquisition. Finally, a logistic regression model combining features of NK cell activation, α4β7 expression on memory CD4+ T cells, and T-box expressed in T cells (Tbet) expression by iNKT cells achieved the highest accuracy in identifying HESN and HESC participants. This immune signature, consisting of increased α4β7 on cells susceptible to HIV infection combined with higher NK cell activation and lower gut-homing potential, could affect the efficacy of HIV-1 prevention strategies such as vaccines.

Keywords: AIDS/HIV; Infectious disease; NK cells; NKT cells; T cells.

Figure 1. Conventional and unconventional T cell phenotypes in HESN and HESC individuals.

(A) Representative flow cytometry plots. (B) Levels of α4β7⁺ and α4β7^hiCD45RO⁺CD4⁺ T cells. (C) Levels of α4β7⁺ iNKT cells, α4β7⁺CD4⁺ iNKT cells, α4β7⁺CD4^– iNKT cells, Tbet⁺ iNKT cells, and CD69⁺ iNKT cells in HESN and HESC individuals prior to HIV-1 acquisition. n = 25 HESC individuals; n = 74 HESN individuals. *P < 0.05, **P < 0.01, and ***P < 0.001, by Mann-Whitney U test. Horizontal lines represent the median with IQR.

Figure 2. NK cell phenotype in HESN and HESC individuals.

Violin plots showing expression levels of HLA-DR (A), Ki67 (B), ILT-2 (C), α4β7 (D), and CD16 (E) in NK cells from HESC individuals (n = 25) prior to HIV-1 acquisition and in HESN individuals (n = 74). *P < 0.05 and **P < 0.01, by Mann-Whitney U test. Horizontal lines represent the median with IQR.

Figure 3. NK cell degranulation ability in HESN and HESC individuals.

(A) Representative flow cytometry plots. (B) Levels of the degranulation marker CD107a in unstimulated or stimulated NK cells in coculture with gp160-expressing cells previously incubated in the presence or absence of PWH plasma. (C) Levels of induction of the degranulation marker CD107a after subtracting the value of CEM-gp160 without PWH plasma. Horizontal lines represent the median with IQR. n = 25 HESC individuals prior to HIV-1 acquisition; n = 74 HESN individuals. **P < 0.01, by Mann-Whitney U test.

Figure 4. Composition of the translocated microbial products in HESN and HESC individuals prior to seroconversion.

Violin plots showing the relative abundance as RPM of selected phyla that were significantly different between HESC individuals prior to HIV-1 acquisition (n = 25) and HESN individuals (n = 24). *P < 0.05, by 2-sided t test.

Figure 5. Associations between translocated microbial products and immune phenotype.

Heatmap showing the Spearman rho values for the association between translocated microbial products at the phylum level and immune phenotype for DNA (A) and RNA (B). P values below 0.05 (Spearman’s correlation test) are indicated by an asterisk. Phyla with a difference in relative abundance between HESC and HESN individuals are indicated by a yellow arrowhead. n = 49.

Figure 6. Statistical model estimating the ability of the immune phenotype to predict seroconversion.

(A) Logistic regression analysis estimating the statistical significance and level to which each marker might increase or decrease the odds of acquiring HIV. (B) Random forest analysis indicating the importance of each marker in predicting HESC and HESN status on the basis of the mean decrease in the Gini coefficient. The marker with a significant effect is indicated in red. Lines represent the 95% CI. (C) ROC curve of the optimized logistic regression model using 4 predictors: HLA-DR⁺ (percentage of NK cells), Tbet⁺ (percentage of iNKT cells), α4β7hi (percentage of CD45RO CD4⁺ T cells), and CD16⁺ (percentage of NK cells).