Innate DNA sensing is impaired in HIV patients and IFI16 expression correlates with chronic immune activation

Abstract

The innate immune system has been recognized to play a role in the pathogenesis of HIV infection, both by stimulating protective activities and through a contribution to chronic immune activation, the development of immunodeficiency and progression to AIDS. A role for DNA sensors in HIV recognition has been suggested recently, and the aim of the present study was to describe the influence of HIV infection on expression and function of intracellular DNA sensing. Here we demonstrate impaired expression of interferon-stimulated genes in responses to DNA in peripheral blood monuclear cells from HIV-positive individuals, irrespective of whether patients receive anti-retroviral treatment. Furthermore, we show that expression levels of the DNA sensors interferon-inducible protein 16 (IFI16) and cyclic guanosine monophosphate-adenosine monophosphate synthase were increased in treatment-naive patients, and for IFI16 expression was correlated with high viral load and low CD4 cell count. Finally, our data demonstrate a correlation between IFI16 and CD38 expression, a marker of immune activation, in CD4(+) central and effector memory T cells, which may indicate that IFI16-mediated DNA sensing and signalling contributes to chronic immune activation. Altogether, the present study demonstrates abnormal expression and function of cytosolic DNA sensors in HIV patients, which may have implications for control of opportunistic infections, chronic immune activation and T cell death.

Keywords: DNA sensing; HIV; IFI16; cGAS; immune activation; interferon-stimulated genes.

Fig. 1

Interferon-stimulated gene (ISG) expression in response to transfected DNA are impaired in HIV patients [irrespective of highly active anti-retroviral therapy (HAART)]. Peripheral blood monuclear cells (PBMCs) were transfected with ssDNA, poly(dA:dT) (both 4 μg/ml) or mock (lipofectamine control) for 6 h. Total mRNA was isolated and expression of C-X-C motif chemokine 10 (CXCL10) mRNA (a,c) and tumour necrosis factor (TNF)-α (b,d) were measured by reverse transcription–quantitative polymerase chain reaction (RT–qPCR). Data are presented as fold induction of cytokine by ssDNA or poly(dA:dT) relative to lipofectamine control after normalization to glyceraldehyde 3-phosphate dehydrogenase (GAPDH). The results were analysed statistically with a multiple one-way analysis of variance (anova) test comparing each group of HIV-infected patients with controls. NR = normalized ratio.

Fig. 2

Correlations between different DNA stimuli and different cytokines/interferon-stimulated genes (ISGs). Pearson's correlation of poly(dA:dT) (x-axis) and HIV-ssDNA (y-axis)-induced C-X-C motif chemokine 10 (CXCL10) mRNA: R = 0·5108, P (two-tailed) < 0·0001 (a); ISG56 mRNA: R = 0·7615 P (two-tailed) < 0·0001 (b) and tumour necrosis factor (TNF)-α: R = 0·2364 P (two-tailed) < 0·0001 (c). Correlations of mRNA expression levels after poly(dA:dT) and ssDNA transfections in (d) highly active anti-retroviral therapy (HAART) naive, (e) immunological non-responders (INRs), (f) responders and (g) controls; x-axis: CXCL10; y-axis: TNF-α and ISG56, all normalized to glyceraldehyde 3-phosphate dehydrogenase (GAPDH) levels. Two-tailed P-values were: naive: P < 0·0001; INRs: P = 0·0002; responders: P = 0·0001; controls: P < 0·0001; all groups merged: P < 0·0001.

Fig. 3

Expression of DNA sensors and downstream signal molecules in HIV patients and controls. Total RNA was isolated from unstimulated peripheral blood mononuclear cells (PBMCs) from the different HIV patient subgroups and controls. mRNA levels of (a) interferon-inducible protein 16 (IFI16), (b) cyclic guanosine monophosphate–adenosine monophosphate synthase (cGAS), (d) DDX41, (e) STING, and (f) IRF3 were determined by reverse transcription–quantitative polymerase chain reaction (RT–qPCR) and each HIV patient subgroup was compared to controls by a multiple one-way analysis of variance (anova) test. (c) Pearson's correlation of cGAS and IFI16 mRNA expression in highly active anti-retroviral therapy (HAART) naive; n.s. = not significant.

Fig. 4

Distribution of CD4⁺ T cells, CD4^– T cells and monocytes in HIV subgroups and controls. (a) CD4⁺ cell count in serum at date of inclusion in the study. (b) Pearson's correlation of CD4⁺ cell count and C-X-C motif chemokine 10 (CXCL10) mRNA expression after ssDNA transfection (4 μg/ml with lipofectamine) in the entire cohort. P (two-tailed) = 0·0948, R = 0·05480. Different CD4⁺ (c) and CD4^– (d) T cell subsets are shown as determined by flow cytometry as follows: naive T cells (CD45RA⁺CD27⁺CCR7⁺), central memory T cells (CD45RA^–CD27⁺CCR7⁺), effector memory T cells (CD45RA^–CD27⁺CCR7^–) and terminally differentiated T cells (CD45RA⁺CD27^–CCR7^–). (e) Estimation of fraction of monocytes (as percentage) in peripheral blood mononuclear cells (PBMCs). Pearson's correlations of (f) monocyte fraction versus log (HIV-RNA copies/ml): R = 0·2292, P = 0·0381, (g) monocyte fraction versus interferon-inducible protein 16 (IFI16) mRNA expression and (h) monocyte fraction versus C-X-C motif chemokine 10 (CXCL10) mRNA expression; n.s. = not significant.

Fig. 5

Correlations in HAART-naive patients of interferon-inducible protein 16 (IFI16), cyclic guanosine monophosphate–adenosine monophosphate synthase (cGAS), C-X-C motif chemokine 10 (CXCL10), and viral load. Pearson's correlations of (a) IFI16 mRNA expression versus log (HIV-RNA copies/ml): R = 0·2292, P = 0·0381, (b) cGAS mRNA expression versus log (HIV-RNA copies/ml), (c) IFI16 versus C-X-C motif chemokine 10 (CXCL10) mRNA expression, (d) cGAS versus CXCL10 mRNA expression: R = 0·1239, P = 0·0030 and (e) CXCL10 expression versus log (HIV-RNA copies/ml); n.s. = not significant.