Research ArticleHIV

A vaccine-induced gene expression signature correlates with protection against SIV and HIV in multiple trials

See allHide authors and affiliations

Science Translational Medicine  28 Aug 2019:
Vol. 11, Issue 507, eaaw4236
DOI: 10.1126/scitranslmed.aaw4236
  • Fig. 1 A gene expression signature seen in B cells is associated with protection in multiple SIV/HIV vaccine studies.

    (A) The same B cell–based gene signature was significantly enriched in the Ad26/gp140 arms of the 09-11 (n = 10) and 13-19 (n = 11) NHP studies in the uninfected compared with infected monkeys, (B) in the human RV144 vaccine study (n = 172), (C and D) and in two independent RV144-like pox-protein vaccine regimens in NHP (n = 39).

  • Fig. 2 A subset of genes from the B cell gene signature predicts protection and is associated with reduced SIV/SHIV infection.

    (A) Fifty-three enriched genes from the 09-11 study B cell signature were used to generate a composite GES consisting of the average of standardized normalized gene expression. The receiver operator characteristic (ROC) curve illustrates the discriminating ability of the GES of enriched genes from the Ad26/gp140 arms of the 09-11 training dataset (AUC = 0.84; 95% CI, 0.52 to 1) and the 13-19 test dataset (AUC = 0.86; 95% CI, 0.56 to 1) to predict protection from acquisition of SIV (n = 10) and SHIV (n = 11), respectively. (B) The model from the 09-11 study was also able to predict protection from acquisition of SHIV in the Ad26/Ad26 + gp140 validation arm of the 13-19 study (n = 12) by logistic regression (AUC = 0.75; 95% CI, 0.41 to 1). (C) Effect of the GES as a continuous variable by Cox proportional hazards ratio (HR) is shown for both NHP studies.

  • Fig. 3 The B cell gene signature and GES associated with immune correlates of protection after vaccination.

    (A and B) ADCP is the only functional immune response that associates with the protective enriched B cell gene signature in both the 09-11 (n = 10) and combined 13-19 (n = 23) studies. (C and D) Relative ranking of immune responses (blue) and GES (red) based on time to infection in the 09-11 and 13-19 studies were generated by averaging the VIP scores across 1000 repeated feature selection processes. Standard error bars of the average of the VIP scores are shown for each feature. *week 96, #week 64.

  • Fig. 4 Functional relationships of the 53 enriched genes from the B cell gene signature.

    (A) Shown is a gene interaction network derived from the protective B cell gene signature. The network edges represent coexpression (purple) and genetic interactions (green). Red nodes represent the 53 enriched genes, and cyan squares indicate interacting network genes. (B) Effect of expression of the 53 genes individually on time to infection as a continuous variable in the 09-11 (n = 10) and 13-19 (n = 23) studies. The most protective genes with HR ≤ 0.5 are shown in red, and of these the overlapping genes in the 09-11 and 13-19 studies are boxed. *P < 0.05.

Supplementary Materials

  • stm.sciencemag.org/cgi/content/full/11/507/eaaw4236/DC1

    Fig. S1. Gating strategy used for flow sorting lymphocyte populations in rhesus monkeys.

    Fig. S2. Logistic regression prediction model in the Ad26/MVA + gp140 arm.

    Fig. S3. qPCR validates higher GES in B cells as protective.

    Fig. S4. Correlation of gene expression with magnitude of immune responses in the 09-11 and 13-19 studies.

    Table S1. Total number of enriched genes in the B cell gene signature for the different trials in this study.

    Table S2. Time points for different assays in the 09-11 and 13-19 studies.

    Table S3. B cell gene signature enrichment associated with higher magnitude of immune responses in the 09-11 and 13-19 studies.

    Table S4. Univariate analysis of immune responses and GES on time to infection in the 13-19 study.

    Table S5. Univariate analysis of immune responses and GES on time to infection in the 09-11 study.

    Data file S1. Enriched GSEA gene sets (NES ≥ 1.4, P < 0.001) in different vaccine regimens in the 09-11 and 13-19 studies.

    Data file S2. Functional categorization of the 53 genes in the B cell gene signature by network analysis.

  • The PDF file includes:

    • Fig. S1. Gating strategy used for flow sorting lymphocyte populations in rhesus monkeys.
    • Fig. S2. Logistic regression prediction model in the Ad26/MVA + gp140 arm.
    • Fig. S3. qPCR validates higher GES in B cells as protective.
    • Fig. S4. Correlation of gene expression with magnitude of immune responses in the 09-11 and 13-19 studies.
    • Table S1. Total number of enriched genes in the B cell gene signature for the different trials in this study.
    • Table S2. Time points for different assays in the 09-11 and 13-19 studies.
    • Table S3. B cell gene signature enrichment associated with higher magnitude of immune responses in the 09-11 and 13-19 studies.
    • Table S4. Univariate analysis of immune responses and GES on time to infection in the 13-19 study.
    • Table S5. Univariate analysis of immune responses and GES on time to infection in the 09-11 study.
    • Legends for Data files S1 and S2

    [Download PDF]

    Other Supplementary Material for this manuscript includes the following:

    • Data file S1 (Microsoft Excel format). Enriched GSEA gene sets (NES ≥ 1.4, P < 0.001) in different vaccine regimens in the 09-11 and 13-19 studies.
    • Data file S2 (Microsoft Excel format). Functional categorization of the 53 genes in the B cell gene signature by network analysis.

Stay Connected to Science Translational Medicine

Navigate This Article