Skip to main content
PMC home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2019 Nov 7.
Published in final edited form as: Immunity. 2016 Feb 2;44(2):391–405. doi: 10.1016/j.immuni.2016.01.006

Innate Lymphoid Cells are Depleted Irreversibly During Acute HIV-1 Infection in the Absence of Viral Suppression

Henrik N Kløverpris 1,2, Samuel W Kazer 3,4,5,6, Jenny Mjösberg 7, Jenniffer M Mabuka 1, Amanda Wellmann 1, Zaza Ndhlovu 8, Marisa C Yadon 1, Shepherd Nhamoyebonde 1, Maximilian Muenchhoff 8,9, Yannick Simoni 10, Frank Andersson 11, Warren Kuhn 12, Nigel Garrett 13,14, Wendy A Burgers 15, Philomena Kamya 3,8, Karyn Pretorius 8, Krista Dong 3, Amber Moodley 3, Evan W Newell 10, Victoria Kasprowicz 1, Salim S Abdool Karim 14,16, Philip Goulder 8,9, Alex K Shalek 3,4,5,6,17, Bruce D Walker 3,8,19, Thumbi Ndung’u 1,3,8,18, Alasdair Leslie 1,3
PMCID: PMC6836297  NIHMSID: NIHMS1056054  PMID: 26850658

SUMMARY

Innate lymphoid cells (ILCs) play a central role in the response to infection by secreting cytokines crucial for immune regulation, tissue homeostasis, and repair. Although dysregulation of these systems is central to pathology, the impact of HIV-1 on ILCs remains unknown. We found that human blood ILCs were severely depleted during acute viremic HIV-1 infection, and that ILC numbers did not recover following resolution of peak viremia. ILC numbers were preserved by antiretroviral therapy (ART), but only if initiated during acute infection. Transcriptional profiling during the acute phase revealed up-regulation of genes associated with cell death, temporally linked with a strong IFN acute-phase response and evidence of gut barrier breakdown. We found no evidence of tissue redistribution in chronic disease and remaining circulating ILCs were activated but not apoptotic. These data provide a potential mechanistic link between acute HIV-1 infection, lymphoid tissue breakdown and persistent immune dysfunction.

Keywords: Innate Lymphoid Cells, HIV-1 Infection

INTRODUCTION

Hallmarks of HIV-1 pathology include immunodeficiency, lymphoid tissue destruction, gut barrier breakdown, and systemic immune activation(Veazey et al., 1998). These features are only partially reversed by fully suppressive long-term antiretroviral therapy (ART)(Sanchez et al., 2014; Zeng et al., 2012). The underlying mechanisms remain unclear, presenting a serious barrier to the development of novel interventions to improve immune reconstitution in HIV-1 infected individuals. Recent studies suggest that the rapid depletion of interleukin 17 (IL-17) and IL-22 producing CD4+ T cells within gut associated lymphocyte tissue (GALT) during acute infection(Schuetz et al., 2014) lead to gut epithelial break down, as IL-17 and IL-22 are the key cytokines in mucosal homeostasis. In fact, the most rapid and prolific producers of IL-17 and IL-22 in tissue are innate lymphoid cells (ILCs) rather than conventional Th17 and Th22 cells(Cella et al., 2008; Cupedo et al., 2009) prompting the question: what happens to this important immune subset after HIV-1 infection?

ILCs are not antigen specific and lack rearranged B- and T-cell receptors. ILCs are grouped into ILC1, ILC2 and ILC3, that share functional characteristics with Th1, Th2 and Th17 cells, respectively(Spits et al., 2013). ILCs respond rapidly to damage, prior to B and T cell expansion, and are therefore crucial for tissue homeostasis and repair during acute and chronic disease(McKenzie et al., 2014). In particular, ILCs are important in mucosal barrier maintenance through tissue repair, wound healing, and regulation of the immune response to commensals (Tait Wojno and Artis, 2012). As a result, ILCs are emerging as key players in many infectious and non-infectious diseases, where they may aid or impair proper immune response. Illustratively, group 2 ILCs (ILC2s) accumulate in the lung following influenza infection and restore epithelial integrity(Monticelli et al., 2011), and loss of gut ILC3s precipitates inflammatory bowel conditions through the IL-22 axis (Sonnenberg et al., 2012; Tait Wojno and Artis, 2012). In contrast, untreated multiple sclerosis, allergic asthma, and psoriasis are associated with expansions of ILCs in peripheral blood that may drive pathology(Bartemes et al., 2014; Perry et al., 2012; Teunissen et al., 2014).

The impact of HIV-1 on ILC populations in circulation and at mucosal barrier sites remains unclear. Given the central role of ILCs in gut epithelial integrity, immune regulation and other systems dysregulated in HIV-1 disease, this represents a significant gap in our understanding of HIV-1 pathology.

In this study, we found that circulating ILCs were depleted 7–14 days after infection, in conjunction with an acute phase immune response. ILC numbers did not recover, in chronic infection, even with long term fully suppressive ART. In contrast, we found that ILC populations were maintained if ART was started during early acute infection before peak viremia. We show that remaining ILCs circulating in chronic disease display an activated phenotype, but no direct evidence of apoptosis or migration to tissue sites. RNA sequencing of ILCs during early acute infection, implicates apoptosis and cell death during ILC depletion. These gene signatures were diminished by early ART. Together these data suggest that depletion of circulating ILCs was mediated by cell death driven by high VL during acute HIV-1 infection and associated with markers of acute viral response.

RESULTS

ILCs are depleted during chronic HIV-1 infection and inversely correlate with viral load setpoint

ILCs are defined as lymphocytes that are negative for B and T cell lineage markers and conventional natural killer (NK) cell markers (CD16 and CD94), but positive for CD127 and CD161. Using flow cytometry of peripheral blood samples from HIV-1 infected and uninfected human donors, we adopted a traditional gating strategy and identified three phenotypically distinct ILC populations as described(Spits et al., 2013): CRTH2-CD117-CD56-CD25−/+ (ILC1), CRTH2+CD117−/+CD56-CD25−/+ (ILC2) and CRTH2−CD117+CD56−/+CD25−/+ (ILC3) (Figure 1A and Figure S1A). To verify the identity of these three ILC populations, we turned to an unbiased data analysis tool, t-distributed stochastic neighbour embedding (tSNE), which simultaneously analyses all flow-measured parameters, rather than sequentially gating. tSNE, like a principal component analysis (PCA), clusters cells that share similar expression patterns together while accounting for potential non-linear relationships between markers(Becher et al., 2014). Using this approach, we identified 5 clusters within the lineage negative (Lin−)CD127+ population: ILC1, ILC2, ILC3, NK and ‘non ILCs’ that corresponded to the phenotypes of human ILCs (Figure 1B)(Spits et al., 2013). To confirm the identify of these ILC subsets functionally, we performed intracellular cytokine staining, using CD4+ T cells as a control, and, as expected, found mutually exclusive interferon γ (IFNγ) and IL-13 production from ILC1 and ILC2 subsets, respectively (Figure 1C). The ILC3 subset in blood produced IL-2 and tumor necrosis factor α (TNFα), but did not express NKp44 (not shown) and therefore did not secrete IL-22(Teunissen et al., 2014). However, a high frequency of NKp44+ ILC3 cells isolated from tonsil produced IL-22 (Figure 1C). In addition, ILC subsets expressed the transcription factors T-bet, GATA-3 and RORγt in the expected patterns for the ILC1, ILC2 and ILC3 subsets, respectively and relative to controls (conventional NK (T-bet) and CD4-Th1 (T-bet) and Th2 (GATA-3) cells) (Figure 1D)(Spits et al., 2013). Finally, we performed RNA sequencing of sorted ILC and CD4+ T cells from the blood of 9 healthy donors and found that the ILC populations displayed an ILC transcriptional signature compared to sample matched CD4+ T-cells (Figure 1E) (Table S1-S3). The ILC subsets themselves were closely related but transcriptionally distinct (Table S1-S3), typified by expression of canonical ILC lineage genes, such as CD117 (cKit; ILC3), IL1R (ILC3) and KLRG1 (ILC2, Figure S1B). Together, these data confirm the precise identification of the main human ILC subsets described.

Figure 1. Identification of human ILCs from blood shows HIV specific depletion of circulating ILCs.

Figure 1.

Open in a new tab

(A) Representative conventional flow cytometry plots from an HIV-1 uninfected donor showing the hierarchical phenotype gating strategy from singlet lymphocytes to the ILC1 (orange), ILC2 (red) and ILC3 (blue) populations indicated by arrows and color-coded gates. Lineage (Lin) gate contains anti- CD3, CD4, CD11c, CD14, CD19, CD34, BDCA2, FcER1, TCRαβ, TCRγδ antibodies. (B) tSNE clustering of human PBMCs pre-gated for lymphocytes/singlets/live/CD45+/CD3−/Lin−/CD127+ that shows two-dimensional representation of high dimensional space based on phenotype markers CD161, CD117, CD56, CRTH2, CD4, NKp44, CD25, CD62L, CD69, CCR6, CD94 with gates indicating 5 identified clusters with heat-map showing the relative expression intensity for each marker within the 5 identified clusters (NK, ILC3, ILC1, ‘non-ILC’, ILC2). (C) Cytokine production after media or PMA/Ionomycin stimulation from blood CD4+ T cells, ILC1s, ILC2s and ILC3s for 6 cytokines IL2, IL13, IFNγ, TNFα, IL17A and IL22 with tonsil derived ILC3 shown as NKp44 vs IL22. (D) Transcription factor expression within CD4+ Th2 cells (CRTH2 gated), CD4+ Th1 cells (CD56 gated), NK cells (CD3−CD94+CD56+CD16+) ILC1, ILC2 and ILC3 for T-bet, GATA-3 and RORγt. (E) Innate lymphoid cell score based on RNAseq generated gene expression within CD4+ T-cell, ILC2 and ILC3 sorted populations from blood of n=9 HIV uninfected individuals based on recent published gene transcripts of ILCs(Robinette et al., 2015). (F) Absolute ILC counts for HIV-1 uninfected (n=136), HIV-1 infected with undetectable plasma virus (VL<50) (n=16) and HIV-1 infected individuals with detectable viremia (VL>50) (n=91). (G) ILC frequency expressed as % of CD45+ lymphotyces for HIV-1 uninfected (n=122), HIV-1 infected with undetectable plasma virus (VL<50) (n=14) and HIV-1 infected individuals with detectable viremia (VL>50) (n=115). P - values by Dunn’s test for multiple comparisons. (H) Grouping of cells gated from lymphs/singlets/live/CD45+/CD3−/Lin−/CD127+ according to automatic (unbiased) cluster designation for accumulated data from 18 HIV-1 uninfected (left) and 21 HIV-1 infected subjects with each distinct cluster named by its unique number inside gates. (I) Bar graph showing the mean percentage contribution from each cluster (x-axis), corresponding to the tSNE plots in (h), to the overall Lin−CD127+ population with P<0.02 indicated by * and calculated by t-test comparing HIV-1 uninfected (n=18) and HIV-1 infected individuals (n=21). P-values by student t-test and Sidak-Bonferroni method for multiple comparisons. See also Figure S1 and S2.

Next, we compared the absolute number of blood ILC in a total of 223 samples from HIV-1 uninfected, ART-naive viremic (HIV-1 RNA>50 copies/ml plasma) and aviremic (<50 copies/ml) individuals with chronic HIV-1 infection (Figure 1F). In viremic subjects, we observed depletion of all three ILC populations (P < 0.0001). However, ILC1 and ILC2, but not ILC3s (P = 0.009), were preserved in aviremic subjects (Figure 1F). ILC frequencies expressed as percentage of total CD45+ lymphocytes confirmed their depletion during chronic HIV-1 (Figure 1G), and showed that changes in frequency of other hematopoietic subsets did not impact ILC measurements during chronic infection. In addition, we observed a significant negative correlation between HIV-1 RNA viral load (VL) setpoint and ILC frequency (P = 0.07 to 0.007, R = −0.18 to −0.32) (Figure S2A,B), similar to the well described correlation between VL and absolute CD4+ T-cell counts (P < 0.001, R = −0.47) (Figure S2C). Thus, ILCs are severely depleted in chronic viremic infection with a direct negative correlation to VL setpoint.

To examine shifts in ILC subsets driven by chronic HIV-1 infection, we used the tSNE algorithm(Becher et al., 2014) to obtain an unbiased analysis of ILC distribution. By gating on Lin−CD127+, we identified 38 distinct clusters with shared surface marker expression characteristics (Figure 1H). Ten of these clusters were significantly enriched or depleted in HIV-1 infected subjects (Figure 1I), predominantly from within the ILC2 and ILC3 populations. Of these, cluster 9 (ILC2) and cluster 20 (ILC3) remained significant after controlling for multiple comparisons (P = 4.5×10−5 and P = 1.7×10−11), suggesting that the circulating ILCs that remained during chronic HIV infection were phenotypically altered.

ILCs are depleted during early acute HIV-1 infection

To further investigate the dynamics of ILC depletion, we turned to a unique acute HIV-1 infection cohort(Ndhlovu et al., 2015). Women in this cohort were tested for the presence of HIV-1 nucleic acid in plasma twice a week, and therefore HIV-1 infections were identified within a maximum of 4 days from their last negative test and approximately 5–14 days after transmission; corresponding to Feibig stage I(McMichael et al., 2010). We tracked ILC1s, ILC2s and ILC3s in seven individuals throughout the course of peak viremia and into chronic infection (Figure 2). We found normal ILC frequencies at the first time-points that were sampled before peak viremia, but observed a rapid ILC depletion that coincided with the peak VL (day 7–14) (Figure 2B,C) and persisted without rebound into chronic infection (P < 0.016) (Figure 2D). This observation remained significant when we analyzed the absolute ILC counts (Figure 2E) (Figure S3A-C), demonstrating that ILC depletion is not a result of changes in the frequency of other subsets during this disease phase. In contrast, and as expected, the characteristic early nadir of absolute CD4+ T-cell count, temporally associated with peak viremia, rebounded rapidly, although to suboptimal levels (Figure 2F). Thus, these data show that the ILC depletion observed during chronic viremic infection occured very early in the acute phase of infection and that, unlike CD4 T cells, ILCs failed to recover following the resolution of acute viremia to setpoint VL.

Figure 2. All ILC populations are depleted during early acute HIV-1 infection.

Figure 2.

Open in a new tab

(A) Data from the acutely HIV-1 infected PID 0398–271 subject followed longitudinally over 10 time-points from day 1 to day 249 from the day of first HIV+ RNA test, with ILC1, ILC2 and ILC3s shown as a percentage of the total CD45+ lymphocytes (left y-axis, coloured line) and the plasma VLs shown as HIV+ RNA copies/ml plasma (right y-axis, black line). (B) Data as shown in (a), but for the entire cohort (n=7). (C) Data as shown in (b), but cumulative data presented as mean values for the entire cohort (n=7) with error bars showing s.e.m. (D) Percentage ILC1, ILC2 and ILC3 cells of total CD45+ lymphocytes shown for week 0 and week 6 after day of first HIV-1 positive test and compared to the HIV uninfected subjects (116). (E) ILC frequencies expressed as absolute ILC counts shown for week 0 and week 6 after day of first HIV-1 positive test and compared to the HIV uninfected subjects (n=116). (F) Data from the acutely HIV-1 infected PID 0398–271 subject as shown in (A) but for absolute CD4+ T-cell counts (pink line, right y-axis).

P - values by Dunn’s test for multiple comparisons. All samples are from ‘FRESH’ cohort(Ndhlovu et al., 2015). See also Figures S2-S4.

Early depletion of ILCs coincides with spikes in epithelial gut breakdown

ILCs are required to maintain an effective gut barrier and to regulate the immune response to commensal microbiota(Sonnenberg et al., 2012). We therefore next sought to define the kinetics of ILC decline in early acute HIV-1 infection in relation to the damage to gut associated lymphoid tissue that occurs during primary infection in non-human primates(Veazey et al., 1998). Changes to gut integrity during acute HIV-1 infection were assessed indirectly by measuring the levels of intestinal fatty acid binding protein 1 (I-FABP), a plasma marker previously associated with gut barrier breakdown(Hunt et al., 2014). In one subject (PID 0398–271), we found a peak in I-FABP levels one week after ILC depletion and 2 weeks after HIV-1 plasma RNA detection (Figure S4A). When we compared the relative levels of I-FABP for the entire cohort, we consistently found maximum levels occurring 2 weeks after HIV-1 RNA detection that coincided with ILC depletion (Figure S4B). Although I-FABP levels return to baseline once VL reaches setpoint, the association between ILC levels and I-FABP was significant at 2 and 3 weeks post HIV-1 detection (P < 0.008) (Figure S4C), corresponding to one week after peak viral replication (Figure S4D). This data is consistent with the hypothesis that massive viral replication during acute infection leads to profound damage to the gut epithelial barrier and precipitates the well described association between microbial translocation, immune activation and disease progression. As ILCs are central to gut epithelial repair, the 4-fold increase in I-FABP levels and coincident loss of ILC from circulation suggests a possible link, although this may be circumstantial and additional data is required to test this correlation.

ILC depletion is irreversible despite successful suppression of viremia by antiretroviral treatment in chronic infection

We next sought to investigate the relationship between ILC depletion and immune reconstitution following ART initiation. In a longitudinal treatment cohort(Abdool Karim et al., 2010), we measured ILC frequencies in chronically infected individuals at the last time-point before ART initiation (median 213 weeks after infection) and 2 years into successful treatment (median 308 weeks after infection) (Figure 3A,B), as indicated by partial CD4+ T cell reconstitution (Figure 3B) and reduced immune activation (Figure 3C). Unexpectedly, all three ILC populations failed to return to normal frequencies, and they remained significantly lower than in uninfected individuals despite undetectable VLs (P < 0.02) (Figure 3D-G). Only blood ILC3s significantly rebounded with ART (P < 0.0001), but never to the frequencies observed in uninfected individuals (median 70 vs 430 ILC3 cells). There was no correlation between the recovery of ILCs and CD4+ T cells in the same subjects (P > 0.7) (data not shown). Thus, the ILC depletion observed during both acute and chronic HIV-1 infection was irreversible when treatment was initiated in the chronic phase.

Figure 3. ILCs are not reconstituted after successful treatment initiated in chronic infection.

Figure 3.

Open in a new tab

Twenty-four individuals were tested 3-monthly for presence of early detectable HIV-1 specific p24 antibodies and followed over 9 years with average treatment initiation starting at median values of 213 weeks after infection and sampled again 2 years later. (A) Median HIV-1 RNA copies/mL plasma before and after treatment with the absolute CD4+ T-cell counts shown in (B). (C) Percentage of HLA-DR+CD38+ expression gated on CD4+ T cells and CD8+ T cells before and after ART treatment. (D) Absolute ILC counts comparing HIV-1 uninfected individuals (n=81) to unmatched HIV-1 infected individuals (n=22) with successful viral suppression after 2 yrs of treatment. (E) Absolute ILC1, ILC2 and ILC3 counts from matched HIV-1 infected individuals before and after ART start (n=22) with horizontal dotted line representing median absolute ILC counts for HIV-1 uninfected individuals. (F) ILC frequencies expressed as % of CD45+ lymphocytes comparing HIV-1 uninfected individuals (n=84) to unmatched HIV-1 infected individuals with successful viral suppression after 2 yrs of treatment (n=22). (G) Data as in (e) but expressed as % ILCs of total CD45+ lymphocytes. P – values by the Wilcoxon matched-paired signed rank test, Mann-Whitney U test and with correlation coefficients shown as spearman rank r – and P – values.

ILC depletion is prevented by antiretroviral treatment initiated during early acute HIV-1 infection

To investigate the potential capacity of ART initiated in early infection to reverse the negative impact of HIV-1 on circulating ILCs, we analyzed a unique subset of individuals in whom ART was initiated at the earliest possible time-point: on the first day of HIV-1 RNA detection and within 5–14 days after HIV-1 transmission. From one subject (PID 0444–312), we observed reduced peak viremia, preserved CD4+ T cells (871 before vs 745 cells/ul 4 weeks after infection) and found no depletion of blood ILCs (Figure 4A). The preservation of ILCs relative to untreated subjects was consistent in all seven individuals receiving ART during early acute infection (Figure 4B) in contrast to acute infected individuals that did not receive treatment (Figure 4C). ILC frequencies in these subjects fluctuated to some extent during the course of acute infection but no clear patterns emerged. Whether this represents heterogeneity in the response of ILCs in these individuals to their ART, or natural variation is unclear. However, at >6 weeks into infection, all three ILC subsets were present at significantly higher frequencies than in untreated subjects (P < 0.018) (Figure 4D). Thus, ILC depletion during HIV-1 infection can be prevented by early ART.

Figure 4. ILCs are preserved by treatment initiation during early acute HIV-1 infection.

Figure 4.

Open in a new tab

(A) Data from one acute infected subject (FRESH cohort) that was treated one day after detection of HIV-1 RNA in plasma (VL=2,900) as indicated by the black arrow and with ILC frequency shown as % of CD45 lymphocytes tracked throughout 400 days after HIV-1 infection (left y-axis) and with HIV-1 RNA copies/mL plasma (right y-axis). ILC1 orange, ILC2 red, ILC3 blue and viral load indicated by black lines (B) Longitudinal data from 7 acutely HIV-1 infected subjects initiated on ART within one day of plasma HIV-1 detection and shown as % ILC of CD45 lymphocytes throughout acute infection. (C) Longitudinal mean values comparing 7 acutely infected individuals initiated on ART one day after HIV RNA detection (colored lines) to 7 acutely infected individuals not receiving ART (grey lines) throughout 11 weeks after HIV detection with error bars representing s.e.m. (D) ILC1, ILC2 and ILC3 frequencies for six acutely infected individuals receiving ART by day 1 of HIV detection ‘ART+’ individuals at >6 weeks after HIV detection compared to 9 treatment naïve acutely infected individuals from the same cohort 6 weeks into infection ‘no ART (6w)’. P - values by Dunn’s test for multiple comparisons.

ILC depletion associated with signatures of activation and Fas up-regulation

ILCs are unlikely to be directly infected by HIV-1 as they do not express the CD4 co-receptor and, in vitro, we were unable to infect ILCs with HIV-1 using high titers of X4 and R5 virus (data not shown). In order to investigate alternative mechanisms of ILC depletion, we first measured the anti-apoptotic factor Bcl-2 and activated caspase-3, which are both involved in apoptosis. Although we were able to increase intensities of caspase-3 in all ILC subsets by incubation with the pro-apoptotic molecule camptothecin, we did not detect any significant changes in absolute (Figure 5A) or relative intensity (Figure 5B) of Bcl-2 or caspase-3 expression when comparing ILCs from HIV-1 uninfected and chronic infected subjects. As expected(Petrovas et al., 2004), we did detect elevated signatures of apoptosis in CD8+ T cells from chronically infected individuals using this assay (data not shown).

Figure 5. ILCs from peripheral blood of chronic HIV-1 infected individuals are not apoptotic but display an activated phenotype and upregulation of Fas (CD95).

Figure 5.

Open in a new tab

(A) Activated caspase-3 measured by median fluorescence intensity (MFI) after log fold titration of camptothecin at 0, 8, 80 and 800 μM shown for ILC1, ILC2 and ILC3 with cumulative data fro MFI values for caspase-3 comparing 20 HIV negative and 18 chronic HIV positive individuals for ILC1, ILC2 and ILC3 subsets. (B) One representative example of percentage Bcl-2 and caspase-3 negative ILC1, ILC2, ILC3 and NK (CD3−CD56+CD94+CD16+) cell subsets with cumulative data shown on the right as the percentage of Bcl-2−/casp-3− cells of ILC1, ILC2 and ILC3 subsets. (C) MFI expression for CD69 on ILC1, ILC2 and ILC3 cells comparing HIV negative (18) and chronic HIV positive subjects (n=21). (D) Fas (CD95) expression on ILC1, ILC2 and ILC3 gated cells overlayed by HIV-1 uninfected (grey) and HIV-1 infected (color) comparing 20 uninfected and 20 infected subjects (right). (E) Plasma IP-10 mean levels (left y-axis) shown for 14 acutely infected treatment naïve subjects sampled before infection (7–60 days prior to HIV detection plotted as day −7) floowed 30 days into infection (left) and with 4 subjects treated 1 day after HIV-1 detection (right) and with plasma HIV RNA copies/ml levels shown on the right y-axis. (F) Same as for (e) but with data for IFNα2. Error bars represent s.e.m. values. P - values by Mann-Whitney U test and paired t-test. See also Figure S5.

We next examined the activation status of the remaining circulating ILCs after HIV-1 infection, as HIV-1 induced immune activation is implicated in CD4 T-cell depletion(Brenchley et al., 2004). We found a significant increased surface expression of the lymphocyte activation marker CD69 on ILC2 and ILC3 (P < 0.0001) (Figure 5C) and on T cells (data not shown) in HIV-1 positive subjects, demonstrating activation of ILCs in response to infection. However, when we measured CD38, a marker of general immune activation on T cells, we found low expression levels on ILC1, ILC2 and ILC3s compared to CD4+ and CD8+ T cells with no correlation to ILC frequencies in HIV-1 infected individuals (data not shown). In addition, we found no difference in CD38 expression on ILC1, ILC2 and ILC3s comparing infected and uninfected individuals in contrast to significant differences observed for both CD4+ and CD8+ T cells (P < 0.01) (data not shown) suggesting that different activation markers exist for ILCs and T cells in HIV-1 infection. We observed significant up-regulation of the mucosal tissue homing receptors α4β7, but only in the ILC3 subset (Figure S5A) and not in their T-cell counterparts or in ILC1 and ILC2 subsets (data not shown).

Plasma IL-7 is known to be elevated in chronic HIV-1 infection(Hodge et al., 2011) and it is the ligand for CD127, a receptor that is expressed on all ILC subsets and is critical for their generation and maintenance (Spencer et al., 2014). Therefore, we next measured plasma levels of IL-7 by ELISA in matched samples and found a weak positive associations with the frequency of ILC1 and ILC2 and plasma IL-7 levels in HIV-1 infected individuals, but no association with ILC3s (Figure S5B). These data suggest that IL-7 does not play a role in persistent ILC depletion in chronic HIV-1 infection.

A recent study suggests that cell death protein Fas-FasL interactions are involved in ILC3 apoptosis in a humanized mouse model of HIV-1(Zhang et al., 2015). Therefore, we measured the relative intensity of CD95 (Fas) expression in chronic HIV-1 and found significant up-regulation on ILC2 and ILC3 subsets measured by % CD95 expression (Figure 5D), and for the ILC1 subset by relative MFI (P = 0.004) (data not shown). The Fas mediated apoptosis reported by Zhang et al. was driven by IFN-α, and we therefore looked to see whether the ILC depletion observed in our acute HIV-1 cohort correlated with plasma quantities of this cytokine, reported to be induced in this early phase of infection(Stacey et al., 2009). Plasma concentrations of IFN-α and the IFN-induced protein IP-10 were measured in samples from 2–12 weeks before infection (plotted as day −7) followed by 3, 7, 14 and 30 days after HIV-1 RNA detection. IP-10 was induced by primary viraemia (66%), confirming immune activation and the existence of a strong IFN signature associated with acute infection (P = 0.038) (Figure 5E). In addition, we observed a modest increase (11%) in IFN-α over baseline prior to peak viremia (P = 0.077) Figure 5F.). In the limited samples available (n=4), we found that IP-10 concentrations were greatly reduced by immediate ART, consistent with a blunted IFN response, but the impact on IFN-α was not apparent (Figure 5E,F). These data are consistent with the hypothesis that ILC depletion is driven by immune activation during acute viraemia, and prevented by early treatment, although the role of IFN-α specifically is not clear.

RNA-Seq analysis reveals down-regulation of genes associated with cell viability and proliferation in ILCs immediately following HIV-1 infection

To gain a deeper understanding of ILC depletion in HIV-1 infection, we performed bulk RNA-Seq on samples from early acute infection. RNA-Seq is a sensitive and powerful method for determining changes in cell populations and behaviors(Rapaport et al., 2013). mRNA from CD4+ T-cells, ILC2, and ILC3 cells isolated at various time points were sequenced from two untreated patients and two patients who started ART immediately after viral RNA detection. Transcriptional comparisons were made between both HIV detection and peak viremia, and peak viremia and 6 weeks after detection for each patient in order to generate lists of significantly up- or down-regulated genes (Figure 6A, Figure S6A).

Figure 6. ILCs show up-regulation of genes associated with apoptosis and cell death in early acute HIV-1 infection.

Figure 6.

Open in a new tab

RNA-Seq was performed on samples from two untreated and two early ART treated subjects (FRESH cohort) during the course of early acute HIV infection. (A) Sampling points and associated VLs for each patient. NB Patient 0444–312 did not have a sample collected prior to peak VL. (B) Heat map of activation z-scores for functionally enriched gene sets differentially expressed between initial viral detection and peak viremia. (C) Similar plot comparing enrichments between peak viremia and approximately 6 weeks after detection. Z-score was calculated using log fold change in expression values (see Methods). See also Figure S6.

First, we compared viral detection and peak viremia to understand the transcriptional changes in response to the antigenemia and cytokine response associated with this phase(McMichael et al., 2010). Both ILC2s and ILC3s in untreated patients showed statistically significant (P < 0.01) down-regulation of genes associated with cell proliferation and viability, and up-regulation of those linked to apoptosis and cell death (Figure 6B, Figure S6B, Table S4). These changes were less apparent in an early acute ART treated patient (PID 0629–453), demonstrating a mitigated response to HIV infection. In addition, several key upstream immune regulators were found to change in ILC2s and ILC3s immediately following infection (Figure S6D). Although ILCs lack T-cell receptor expression at the protein level, there were significant changes in genes associated with CD3, also reported in transcriptional profiling of ILCs in murine models(Robinette et al., 2015). However, we found no evidence of TCR gene modules and the role of these genes in ILCs is unclear.

We also compared transcriptional profiles of ILCs at peak viremia and approximately 6 weeks after infection to probe the state of the ILCs that survive acute infection. In contrast to early changes, we found that ILC2s from untreated patients displayed up-regulated cell proliferation and cell viability, and down-regulated cell death and apoptosis (Figure 6C, Figure S6C, Table S5). This was consistent with our flow cytometry data showing that ILCs persisting in chronic infection show no evidence of increased apoptosis or cell death. Additionally, although phenotypically skewed – (see Figure 1H,I), the remaining ILCs, show signs of typical cellular function and immune response (Figure S6E). The early ART treated patients, on the other hand, show mixed responses after peak viremia. Thus, ILC2s and ILC3s, in the absence of early antiretroviral treatment, undergo cell death by apoptosis during peak viraemia.

Tonsil and gut resident ILCs are not enriched or depleted in chronic HIV-1 infection

Current studies of the involvement of ILCs in disease have focused primarily their effector function in lymphoid and non-lymphoid tissue such as lung, skin and intestine(McKenzie et al., 2014). Therefore, to investigate how changes in circulating ILCs relate to lymphoid tissue resident cells, we next turned to tissue samples from HIV-1 infected and uninfected subjects.

We first examined surgically removed tonsils, lymphoid organs that support HIV-1 replication and undergo profound tissue remodeling during progressive disease(Doitsh et al., 2014; Sanchez et al., 2014). We identified tissue resident ILCs and confirmed these subsets by transcription factor staining in comparison to NK and CD4+ Th2 cells (Figure 7A). We found higher frequencies of ILC1s and ILC3s, but not ILC2s, compared to the blood (Figure S7A), with a distinct CD69+ and CD62L− phenotype (Figure 7B) that is characteristic of tissue resident T-lymphocytes. No significant depletion of CD4+ T cells from the HIV-1 infected tonsils, known to be directly infected by the virus, was observed (Figure 7C)(Doitsh et al., 2014). Interestingly, no significant effect of HIV-1 on ILC frequency, subtype or phenotype in tonsil tissues was observed (Figure 7D,E), nor any differences of the apoptotic markers active caspase-3 and Bcl-2 (data not shown), that might explain their loss from circulation. We next examined ILC frequencies in the gut mucosa, which is a major site of HIV-1 replication and CD4+ T-cell depletion(Mattapallil et al., 2005), and where ILC3s play a crucial role in homeostasis and barrier function. Using gut biopsies from subjects undergoing colonoscopy, we identified a distinct lineage negative CD127+ ILC population that was dominated by cKit positive ILC3 cells expressing CD56, NKp44, CCR6, CD69 and transcription factors RORγt, Helios and AHR (Figure 7F). In 46 subjects we found increased levels of ILC1 and ILC3, but not ILC2, compared to tonsil and blood (P < 0.0001) (Figure 7G). However, we observed no differences in gut resident ILC frequencies between HIV-1 infected (n=9) and uninfected (n=37) individuals, nor do we detect any difference in NKp44 and CD56 ILC3 phenotype (Figure 7G). Together these data provide no evidence that ILC are either recruited to or depleted from lymph nodes or the gut mucosal barrier.

Figure 7. Magnitude and phenotype of tissue resident ILCs identified using transcription factors within tonsil and gut tissue.

Figure 7.

Open in a new tab

(A) Gating of live CD45+CD3− ILC1 (Lin− CD127+CD161+CRTH2−CD117+CD56−/+), ILC2 (Lin−CD127+CD161+CRTH2+), ILC3 (Lin− CD127+CD161+CRTH2−CD117+CD56−/+), NK (CD3-CD94+CD56+) and CD4+-Th2 (CD3+CD4+CRTH2+) cells and overlayed for CD69, CD56 and NKp44 expression and stained for T-bet, Eomes, GATA-3, Helios, AHR and RORγt transcription factors. (B) Median fluorescence intensity (MFI) of CD69 and CD62L expression on ILC1, ILC2 and ILC3 subsets in tonsil and blood resident T cells with NKp44+/− ILC3 subsets in tonsils (NKp44 not expressed in blood cells). (C) Frequency of CD4+ T cells shown as % of CD45+ cells within tonsil cells obtained from 12 HIV uninfected and 10 HIV infected subjects. (D) Comparing ILC1, ILC2 and ILC3 subsets expressed as % of CD45+ lymphocytes in HIV infected (n=12) and uninfected individuals (n=12). (E) ILC3 phenotype distribution of CD56 and NKp44 positive and negative subsets with FACS plot showing one representative example and pie charts showing data for a total of 21 subjects. (F) Gut tissue resident live CD45+CD3−CD4− lymphocytes gated as in (A) with ILC3 cells overlayed on Lin+ (CD45+CD3−Lin+) shown for CCR6 and CD69 and shown for RORγt, Helios and AHR transcription factors for Gut Lin+, CD4+ T, NK and ILC3 cells. (G) Cumulative data for ILC1, ILC2 and ILC3 frequencies expressed as % of CD45+ lymphocytes from gut, tonsil and blood (left) with ILC1, ILC2 and ILC3 frequencies shown for n=39 HIV-1 uninfected and n=9 HIV infected subjects (middle) and comparing ILC3 NKp44/CD56 phenotype expression from HIV-1 uninfected and infected subjects (right). (H) Intracellular cytokine staining after media or PMA/Ionomycin stimulation for 5 hrs shown for IL2, TNFα, IL22 and GM-CSF production in NKp44+ ILC3 gated cells and with cumulative data shown in (I) obtained from HIV-1 infected (n=9) and uninfected individuals (n=14). P – values by the Wilcoxon matched-paired signed rank test and Mann-Whitney U test with horizontal bars representing median values. See also Figure S7.

Finally, to assess the impact of HIV-1 on ILC function, we measured cytokine production in tonsil ILCs from infected and uninfected individuals, stimulated non-specifically with PMA and ionomycin. Cytokine production in these cells was restricted to NKp44+ILC3s (Figure 7H, Figure S7B; P < 0.0001), and was dominated by IL-2, TNFα and GM-CSF and, to a lesser extent IL-22. No IFN-γ or IL-13 was detected from NKp44+ RORγt+ ILC3 cells (data not shown). Although a consistent trend was observed for decreased cytokine production in tonsil resident ILC3s from HIV-1 infected individuals, these differences were not significant in the sample size obtained here. Taken together, ILC frequencies in tonsil and gut tissue from HIV infected subjects did not support ILC redistribution from circulation.

DISCUSSION

The role of ILCs during chronic viral infection in humans remains unclear(Diefenbach, 2013), despite compelling evidence highlighting the importance of ILCs in immune regulation and mucosal barrier maintenance(Sonnenberg et al., 2012; Tait Wojno and Artis, 2012). Here, we report a rapid and irreversible depletion of blood ILCs during acute HIV-1 infection that persists in chronic infection in proportion to VL. During the chronic phase, ILC depletion is associated with altered subset composition and increased expression of activation (CD69), tissue homing (α4β7) markers, and the Fas death receptor (CD95). Up-regulation of FAS makes ILC3s more susceptible to anti-CD95 antibody induced apoptosis in vitro(Zhang et al., 2015), however, we detect no apoptotic ILCs ex-vivo in chronic human HIV-1 infection measured either by caspase 3 activation or loss of bcl-2. In contrast, in early acute infection we find ILCs up-regulate genes associated with cell death and apoptosis, potentially explaining their disappearance in the absence of early ART. This coincides with a strong IFN response induced by peak viral replication, demonstrated by the rapid elevation in plasma IP-10 and to a lesser extent IFNα. We find no evidence of enrichment of tissue resident ILCs in either tonsil or gut samples, suggesting that ILC depletion from circulation is explained by apoptosis rather than tissue redistribution.

Despite the correlation between VL and ILC levels in chronic infection, removal of viral burden and reduction of immune activation during this phase with fully suppressive ART does not restore circulating ILC populations. There is a partial rebound of ILC3s, but they remain well below the levels found in healthy donors. This suggests that sustained viremia leads to a fundamental impairment of the ILC arm of the immune system, which could have far reaching immunological consequences(Marchetti et al., 2013). The role of acute viremia in ILC depletion is supported by data from patients in whom ART was initiated prior to peak viremia. In these individuals we observe no sustained depletion of ILCs and, in contrast to untreated individuals or those treated later, ILCs remain at the levels observed in uninfected individuals. This is associated with an absence of the transcriptional signature of apoptosis, and the strong plasma IP-10 response observed in the untreated individuals. What role this early IFN response has in ILC depletion is not clear from our data, although IFNα in particular has been implicated in ILC3 depletion in the humanized mouse model(Zhang et al., 2015). Importantly, depletion of ILCs from blood is not a general acute phase response to infection, as filarial infection in humans is associated with expansion of blood ILCs(Boyd et al., 2014), suggesting the specific relevance of this phenomena to HIV-1. Indeed, more than 90% of individuals sampled from the same populations investigated here were infected with CMV and EBV (unpublished data), yet displayed normal blood ILC levels compared to individuals from areas without endemic viral infections(Munneke et al., 2014). In addition, direct infection of ILCs by HIV-1 is highly unlikely as they lack viral entry receptors consistent with resistance to high tittered in vitro infection of either X5 or R4 virus.

Those ILCs that do survive peak viremia and persist in chronic infection show no evidence of apoptosis by protein expression (Bcl-2 and active caspase-3) and exhibit comparatively down-regulation of apoptosis and cell death transcription signatures by RNA-seq. In fact, genes associated with viral infection and immune response are up-regulated in these populations, consistent with the measured increase in CD69 expression. Whether a certain subset of ILCs never initiates apoptosis in early acute infection or a population is consistently renewed at lower levels in the blood during chronic infection remains unclear.

In addition to apoptosis during acute infection, a potential mechanism for the depletion of circulating ILCs is that they home to major sites of HIV-1 replication and tissue damage. The fact that we detect no significant increase in ILCs within tonsil or gut tissue may be due to sensitivity. However, significant increases in ILC subsets are detectable in the livers of fibrotic mice(Mchedlidze et al., 2013) and skin of human psoriasis suffers(Teunissen et al., 2014). Furthermore, significant enrichment of NKp44+ lineage negative cells, which likely represent the ILC3 NKp44+ subset, are observed in the tonsils of SIV infected macaques(Reeves et al., 2011). Subsequent studies observed depletion of the ILC3 NKp44+ subset from the GALT and mesenteric lymph nodes of acutely and chronically SIV infected Macaques(Li et al., 2014; Xu et al., 2015). Why an enrichment of ILC NKp44+ cells occurs in the tonsils of SIV challenged monkeys and not in naturally infected human subjects remains unclear, but these data imply that either enrichment or depletion of ILCs in the context of retroviral infection would be possible to detect.

The fact that Li et al observed a depletion of an ILC3-like subset from the GALT of acutely SIV infected Macaques(Li et al., 2014) is consistent with the depletion of circulating ILC3s we observe in acute HIV-1 infection and concomitant spike in I-FABP levels. The limited recovery of blood ILC3s observed after successful drug treatment suggests that this subset remains impaired, and that continued immune activation might relate to the functional inability of gut resident ILC3s in HIV-1 infected individuals(Zhang et al., 2015) to restore gut barrier integrity and prevent microbial translocation(Brenchley et al., 2004). Whether ILCs play a direct role in the HIV-1 pathology is difficult to study in human samples and therefore animal studies are warranted to elucidate the mechanistic details relating to the direct consequence of ILC depletion in HIV/SIV pathology(Reeves et al., 2011; Zhang et al., 2015) However, we believe that ILCs are likely to play an important role in HIV-1 pathology given the existence of an IL-22 producing NKp44+ ILC3 subset that is required to both maintain the gut mucosa and limit the response to gut microbial contents(Cella et al., 2008; Sonnenberg et al., 2012). Irreversibly depleted ILC3s by HIV-1 infection would suggest a clear mechanism behind the continued immune activation observed even in individuals with successful long term viral suppression by ART(Sanchez et al., 2014; Zeng et al., 2012), which is the strongest predictor for the onset of AIDS(Hunt et al., 2014).

In summary, we demonstrated that the persistent and irreversible ILC depletion that occurs immediately after HIV-1 acquisition correlates with disease stage and is not restored by long term fully suppressive ART, but can be blocked by early treatment. This provides a potential mechanistic link between HIV-1 infection, lymphoid tissue breakdown and persistent immune dysfunction that merits further exploration, and suggests the importance of early ART administration in maintaining normal immune system composition and functionality.

MATERIALS AND METHODS

Subjects

We used samples from a total of 122 HIV-1 uninfected subjects and 137 HIV-1 infected subjects. All participants were women with sub-Saharan Zulu/Xhosa ancestry from four independently collected cohorts within or in the greater area of Durban, KwaZulu-Natal, South Africa. See Supplementary material for further information.

Clinical parameters

VLs were obtained using the Roche Amplicor 1.5 assay (iThimba and CAPRISA002 cohorts) or the BioMerieux Nuclisens v2.0 (FRESH and GATEWAY cohorts) at Global Clinical and Viral Laboratories, Durban, South Africa. CD4+ T-cell counts and total lymphocyte counts were determined as previously described(Abdool Karim et al., 2010).

Flow cytometry

We used different antibody panels for phenotype and transcription factor staining. See supplementary materials for specific antibodies used throughout the study.

All samples were surface stained at room temperature for minimum 20 mins and intracellularly stained at room temperature for minimum 20 mins. All samples were fixed in 2% Paraformaldehyde before acquisition on a 4 laser, 17 parameter BD Fortessa flow cytometer within 24 hours of staining. Data were analysed using FlowJo v. 9.7.2 (TreeStar).

ELISA

Intestinal Fatty Acid Binding Protein (I-FABP) was measured using the ELISA kit human FABP2 DuoSet, R&D systems and plasma IL-7 levels were measured using the recombinant human IL-7 kit from R&D systems (cat#207-iL). IFNα and IP-10 were measured using the Milliplex kit (Millipore) and completed according to the manufacturers protocol.

tSNE analysis of flow cytometry data

Unbiased representations of multi-parameter flow cytometry data were generated using the t-distributed stochastic neighbor embedding (tSNE) algorithm(van der Maaten and Hinton, 2008). tSNE is a non-linear dimensionality reduction method that optimally places cells with similar expression levels near to each other and cells with dissimilar expression levels further apart. See supplementary methods for how tSNE analysis were executed.

RNA-Seq

CD4+ T cells, ILC2s, and ILC3s (100,000 – 50 cells) were sorted from PBMCs as described above into 300 μL of RLT Lysis Buffer (Qiagen) supplemented with 1% v/v 2-mercaptoethanol, briefly vortexed, spun down, and snap-frozen on dry ice. Cellular mRNA was then isolated and processed for RNA-Seq as described previously(Trombetta et al., 2014). See supplementary methods for details.

Sequencing libraries were then prepared from WTA product using Nextera XT (Illumina). After library construction, one final AMPure XP SPRI clean-up (0.8 volumes) was conducted. Library concentration and size were measured with the KAPA Library Quantification kit (KAPA Biosystems) and a TapeStation (Agilent Technologies), respectively. Finally, samples were sequenced on a NextSeq500 (30 bp paired-end reads) to an average depth of 5 million reads. See Supplementary methods for details on gene expression data analysis.

Statistical analyses

We used the Mann-Whitney U-test for comparison of median values between two groups only and the Dunn’s multi comparisons test to compare median values of more than 2 groups. The Wilcoxon matched-pairs signed rank test was used for paired testing of median values before and after antiretroviral treatment for matched samples. We used the Spearman rank correlation test to compare correlation between two parameters and reported r-values and P-values. Statistical analyses were performed using GraphPad Prism version 6.0c (GraphPad software, Inc).

Supplementary Material

Supplement
Table S1
Table S2
Table S3
Table S4
Table S5

ACKNOWLEDGEMENTS

The Danish Agency for Science, Technology and Innovation (grant #12–132295), Lundbeck Foundation (grant #R151–2013-14624) and MAERSK Foundation (H.N.K). The Collaboration for AIDS Vaccine Discovery of the Bill and Melinda Gates Foundation, and NIH grant AI067073 (BDW). Partial support from the Bill and Melinda Gates Foundation, the International AIDS Vaccine Initiative (IAVI) (UKZNRSA1001), and the NIAID (R37AI067073). The South African Research Chairs Initiative, an International Early Career Scientist Award from the Howard Hughes Medical Institute and the Victor Daitz Foundation (T.N). The National Science Foundation Graduate Research Fellowship Program (NSF GRFP) (S.W.K). The Searle Scholars Program (A.K.S).

Dr Hollis Shen for extensive support during cell sorting. All ‘iThimba’, ‘Gateway’, ‘CAPRISA002/004’ and ‘FRESH’ Acute Infection Study participants. The supportive role of the CAPRISA002/004 studies. Carly G. K. Ziegler and Travis K. Hughes (Shalek Lab) for help on analysis of the RNA-Seq data.

NIH AIDS Reagent Program, Division of AIDS, NIAID, NIH: α4-β7 monoclonal antibody (cat#11718) from Dr. A. A. Ansari.

Footnotes

The authors have declared that no competing interests exist.

REFERENCES

  1. Abdool Karim Q, Abdool Karim SS, Frohlich JA, Grobler AC, Baxter C, Mansoor LE, Kharsany AB, Sibeko S, Mlisana KP, Omar Z, et al. (2010). Effectiveness and safety of tenofovir gel, an antiretroviral microbicide, for the prevention of HIV infection in women. Science 329, 1168–1174. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Bartemes KR, Kephart GM, Fox SJ, and Kita H. (2014). Enhanced innate type 2 immune response in peripheral blood from patients with asthma. The Journal of allergy and clinical immunology 134, 671–678 e674. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Becher B, Schlitzer A, Chen J, Mair F, Sumatoh HR, Teng KW, Low D, Ruedl C, Riccardi-Castagnoli P, Poidinger M, et al. (2014). High-dimensional analysis of the murine myeloid cell system. Nature immunology 15, 1181–1189. [DOI] [PubMed] [Google Scholar]
  4. Boyd A, Ribeiro JM, and Nutman TB (2014). Human CD117 (cKit)+ Innate Lymphoid Cells Have a Discrete Transcriptional Profile at Homeostasis and Are Expanded during Filarial Infection. PloS one 9, e108649. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Brenchley JM, Schacker TW, Ruff LE, Price DA, Taylor JH, Beilman GJ,Nguyen PL, Khoruts A, Larson M, Haase AT, and Douek DC (2004). CD4+ T cell depletion during all stages of HIV disease occurs predominantly in the gastrointestinal tract. The Journal of experimental medicine 200, 749–759. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Cella M, Fuchs A, Vermi W, Facchetti F, Otero K, Lennerz JKM, Doherty JM, Mills JC, and Colonna M (2008). A human natural killer cell subset provides an innate source of IL-22 for mucosal immunity. In Nature, pp. 722–725. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Cupedo T, Crellin NK, Papazian N, Rombouts EJ, Weijer K, Grogan JL, Fibbe WE, Cornelissen JJ, and Spits H (2009). Human fetal lymphoid tissue–inducer cells are interleukin 17–producing precursors to RORC+ CD127+ natural killer–like cells. In Nature immunology, pp. 66–74. [DOI] [PubMed] [Google Scholar]
  8. Diefenbach A (2013). Innate lymphoid cells in the defense against infections. In European Journal of Microbiology and Immunology, pp. 143–151. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Doitsh G, Galloway NL, Geng X, Yang Z, Monroe KM, Zepeda O, Hunt PW, Hatano H, Sowinski S, Munoz-Arias I, and Greene WC (2014). Cell death by pyroptosis drives CD4 T-cell depletion in HIV-1 infection. Nature 505, 509–514. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Glatzer T, Killig M, Meisig J, Ommert I, Luetke-Eversloh M, Babic M, Paclik D,Bluthgen N, Seidl R, Seifarth C, et al. (2013). ROR&gamma;t+ Innate Lymphoid Cells Acquire a Proinflammatory Program upon Engagement of the Activating Receptor NKp44. In Immunity, pp. 1223–1235. [DOI] [PubMed] [Google Scholar]
  11. Hodge JN, Srinivasula S, Hu Z, Read SW, Porter BO, Kim I, Mican JM, Paik C, Degrange P, Di Mascio M, and Sereti I (2011). Decreases in IL-7 levels during antiretroviral treatment of HIV infection suggest a primary mechanism of receptor-mediated clearance. Blood 118, 3244–3253. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Hu Y, Wang K, He X, Chiang DY, Prins JF, and Liu J (2010). A probabilistic framework for aligning paired-end RNA-seq data. Bioinformatics 26, 1950–1957. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Hunt PW, Sinclair E, Rodriguez B, Shive C, Clagett B, Funderburg N, Robinson J, Huang Y, Epling L, Martin JN, et al. (2014). Gut epithelial barrier dysfunction and innate immune activation predict mortality in treated HIV infection. The Journal of infectious diseases 210, 1228–1238. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Kurashima Y, Goto Y, and Kiyono H (2013). Mucosal innate immune cells regulate both gut homeostasis and intestinal inflammation. In Eur. J. Immunol., pp. 3108–3115. [DOI] [PubMed] [Google Scholar]
  15. Li H, Richert-Spuhler LE, Evans TI, Gillis J, Connole M, Estes JD, Keele BF, Klatt NR, and Reeves RK (2014). Hypercytotoxicity and Rapid Loss of NKp44+ Innate Lymphoid Cells during Acute SIV Infection. PLoS pathogens 10, e1004551. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Marchetti G, Tincati C, and Silvestri G (2013). Microbial translocation in the pathogenesis of HIV infection and AIDS. Clinical microbiology reviews 26, 2–18. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Mattapallil JJ, Douek DC, Hill B, Nishimura Y, Martin M, and Roederer M (2005). Massive infection and loss of memory CD4+ T cells in multiple tissues during acute SIV infection. Nature 434, 1093–1097. [DOI] [PubMed] [Google Scholar]
  18. Mchedlidze T, Waldner M, Zopf S, Walker J, Rankin, Andrew L, Schuchmann M, Voehringer D, Mckenzie, Andrew NJ, Neurath, Markus F, Pflanz S, and Wirtz S (2013). Interleukin-33-Dependent Innate Lymphoid Cells Mediate Hepatic Fibrosis. In Immunity, pp. 357–371. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. McKenzie AN, Spits H, and Eberl G (2014). Innate lymphoid cells in inflammation and immunity. Immunity 41, 366–374. [DOI] [PubMed] [Google Scholar]
  20. McMichael AJ, Borrow P, Tomaras GD, Goonetilleke N, and Haynes BF (2010). The immune response during acute HIV-1 infection: clues for vaccine development. Nature reviews. Immunology 10, 11–23. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Monticelli LA, Sonnenberg GF, Abt MC, Alenghat T, Ziegler CGK, Doering TA, Angelosanto JM, Laidlaw BJ, Yang CY, Sathaliyawala T, et al. (2011). Innate lymphoid cells promote lung-tissue homeostasis after infection with influenza virus. In Nature immunology, pp. 1045–1054. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Munneke JM, Bjorklund AT, Mjosberg JM, Garming-Legert K, Bernink JH, Blom B, Huisman C, van Oers MH, Spits H, Malmberg KJ, and Hazenberg MD (2014). Activated innate lymphoid cells are associated with a reduced susceptibility to graft-versus-host disease. Blood 124, 812–821. [DOI] [PubMed] [Google Scholar]
  23. Ndhlovu ZM, Kamya P, Mewalal N, Kloverpris HN, Nkosi T, Pretorius K, Laher F, Ogunshola F, Chopera D, Shekhar K, et al. (2015). Magnitude and Kinetics of CD8(+) T Cell Activation during Hyperacute HIV Infection Impact Viral Set Point. Immunity 43, 591–604. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Perry JS, Han S, Xu Q, Herman ML, Kennedy LB, Csako G, and Bielekova B (2012). Inhibition of LTi cell development by CD25 blockade is associated with decreased intrathecal inflammation in multiple sclerosis. Science translational medicine 4, 145ra106. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Petrovas C, Mueller YM, Dimitriou ID, Bojczuk PM, Mounzer KC, Witek J, Altman JD, and Katsikis PD (2004). HIV-specific CD8+ T cells exhibit markedly reduced levels of Bcl-2 and Bcl-xL. Journal of immunology 172, 4444–4453. [DOI] [PubMed] [Google Scholar]
  26. Rapaport F, Khanin R, Liang Y, Pirun M, Krek A, Zumbo P, Mason CE, Socci ND, and Betel D (2013). Comprehensive evaluation of differential gene expression analysis methods for RNA-seq data. Genome biology 14, R95. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Reeves RK, Rajakumar PA, Evans TI, Connole M, Gillis J, Wong FE, Kuzmichev YV, Carville A, and Johnson RP (2011). Gut inflammation and indoleamine deoxygenase inhibit IL-17 production and promote cytotoxic potential in NKp44+ mucosal NK cells during SIV infection. Blood 118, 3321–3330. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Robinette ML, Fuchs A, Cortez VS, Lee JS, Wang Y, Durum SK, Gilfillan S, Colonna M, and Immunological Genome C (2015). Transcriptional programs define molecular characteristics of innate lymphoid cell classes and subsets. Nature immunology 16, 306–317. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Sanchez JL, Hunt PW, Reilly CS, Hatano H, Beilman GJ, Khoruts A, Jasurda JS, Somsouk M, Thorkelson A, Russ S, et al. (2014). Lymphoid Fibrosis Occurs in Long-Term Nonprogressors and Persists With Antiretroviral Therapy but May Be Reversible With Curative Interventions. The Journal of infectious diseases. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Satoh-Takayama N, Vosshenrich CA, Lesjean-Pottier S, Sawa S, Lochner M, Rattis F, Mention JJ, Thiam K, Cerf-Bensussan N, Mandelboim O, et al. (2008). Microbialflora drives interleukin 22 production in intestinal NKp46+ cells that provide innate mucosal immune defense. Immunity 29, 958–970. [DOI] [PubMed] [Google Scholar]
  31. Schuetz A, Deleage C, Sereti I, Rerknimitr R, Phanuphak N, Phuang-Ngern Y, Estes JD, Sandler NG, Sukhumvittaya S, Marovich M, et al. (2014). Initiation of ART during Early Acute HIV Infection Preserves Mucosal Th17 Function and Reverses HIV-Related Immune Activation. PLoS pathogens 10, e1004543. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Sonnenberg GF, Monticelli LA, Alenghat T, Fung TC, Hutnick NA, Kunisawa J, Shibata N, Grunberg S, Sinha R, Zahm AM, et al. (2012). Innate lymphoid cells promote anatomical containment of lymphoid-resident commensal bacteria. Science 336, 1321–1325. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Spencer SP, Wilhelm C, Yang Q, Hall JA, Bouladoux N, Boyd A, Nutman TB, Urban JF, Wang J, Ramalingam TR, et al. (2014). Adaptation of Innate Lymphoid Cells to a Micronutrient Deficiency Promotes Type 2 Barrier Immunity. In Science, pp. 432–437. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Spits H, Artis D, Colonna M, Diefenbach A, Di Santo JP, Eberl G, Koyasu S, Locksley RM, McKenzie AN, Mebius RE, et al. (2013). Innate lymphoid cells--a proposal for uniform nomenclature. Nature reviews. Immunology 13, 145–149. [DOI] [PubMed] [Google Scholar]
  35. Stacey AR, Norris PJ, Qin L, Haygreen EA, Taylor E, Heitman J, Lebedeva M, DeCamp A, Li D, Grove D, et al. (2009). Induction of a striking systemic cytokine cascade prior to peak viremia in acute human immunodeficiency virus type 1 infection, in contrast to more modest and delayed responses in acute hepatitis B and C virus infections. Journal of virology 83, 3719–3733. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Tait Wojno ED, and Artis D (2012). Innate lymphoid cells: balancing immunity, inflammation, and tissue repair in the intestine. In Cell host & microbe, pp. 445–457. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Teunissen MB, Munneke JM, Bernink JH, Spuls PI, Res PC, Te Velde A, Cheuk S, Brouwer MW, Menting SP, Eidsmo L, et al. (2014). Composition of innate lymphoid cell subsets in the human skin: enrichment of NCR(+) ILC3 in lesional skin and blood of psoriasis patients. The Journal of investigative dermatology 134, 2351–2360. [DOI] [PubMed] [Google Scholar]
  38. Trombetta JJ, Gennert D, Lu D, Satija R, Shalek AK, and Regev A (2014). Preparation of Single-Cell RNA-Seq Libraries for Next Generation Sequencing. Current protocols in molecular biology / edited by Frederick M. Ausubel … [et al.] 107, 4 22 21–24 22 17. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. van der Maaten L, and Hinton G (2008). Visualizing data using t-SNE. Journal of Machine Learning Research 9, 2579–2605. [Google Scholar]
  40. Veazey RS, DeMaria M, Chalifoux LV, Shvetz DE, Pauley DR, Knight HL, Rosenzweig M, Johnson RP, Desrosiers RC, and Lackner AA (1998). Gastrointestinal tract as a major site of CD4+ T cell depletion and viral replication in SIV infection. Science 280, 427–431. [DOI] [PubMed] [Google Scholar]
  41. Xu H, Wang X, Lackner AA, and Veazey RS (2015). Type 3 innate lymphoid cell depletion is mediated by TLRs in lymphoid tissues of simian immunodeficiency virus-infected macaques. FASEB J. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Zeng M, Southern PJ, Reilly CS, Beilman GJ, Chipman JG, Schacker TW, and Haase AT (2012). Lymphoid tissue damage in HIV-1 infection depletes naive T cells and limits T cell reconstitution after antiretroviral therapy. In PLoS pathogens, p. e1002437. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Zhang Z, Cheng L, Zhao J, Li G, Zhang L, Chen W, Nie W, Reszka-Blanco NJ, Wang FS, and Su L (2015). Plasmacytoid dendritic cells promote HIV-1-induced group 3 innate lymphoid cell depletion. J Clin Invest 125, 3692–3703. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Zheng Y, Valdez PA, Danilenko DM, Hu Y, Sa SM, Gong Q, Abbas AR, Modrusan Z, Ghilardi N, de Sauvage FJ, and Ouyang W (2008). Interleukin-22 mediates early host defense against attaching and effacing bacterial pathogens. Nature medicine 14, 282–289. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplement
NIHMS1056054-supplement-Supplement.pdf (1.6MB, pdf)
Table S1
NIHMS1056054-supplement-Table_S1.xlsx (111.8KB, xlsx)
Table S2
NIHMS1056054-supplement-Table_S2.xlsx (88.1KB, xlsx)
Table S3
NIHMS1056054-supplement-Table_S3.xlsx (88.1KB, xlsx)
Table S4
NIHMS1056054-supplement-Table_S4.xls (2.1MB, xls)
Table S5
NIHMS1056054-supplement-Table_S5.xls (2.8MB, xls)

RESOURCES