Selective expansion of polyfunctional pathogen-specific CD4(+) T cells in HIV-1-infected patients with immune reconstitution inflammatory syndrome

Abstract

Since the introduction of highly active antiretroviral therapies (ART), the prognosis for HIV-1 patients has improved immensely. However, approximately 25% of patients can experience a variety of inflammatory symptoms that are collectively known as immune reconstitution inflammatory syndrome (IRIS). Studying the etiology and immunopathology of IRIS has been hampered by the fact that the symptoms and associated opportunistic infections are highly varied. We hypothesized that there is a common mechanism underlying IRIS pathogenesis and investigated a patient group with IRIS related to different pathogens. Functional and phenotypic characterization of PBMC samples was performed by polychromatic flow cytometry after in vitro stimulation with relevant antigenic preparations. In most patients, IRIS events were characterized by the robust increase of preexisting polyfunctional, highly differentiated effector CD4(+) T-cell responses that specifically targeted the antigens of the underlying co-infection. T-cell responses to HIV-1 or other underlying infections were not affected and did not differ between IRIS and non-IRIS patients. These data suggest that patients with IRIS do not have a generalized T-cell dysfunction; instead, IRIS represents a dysregulated CD4(+) T-cell response against residual opportunistic infection antigen. These studies were registered at www.clinical-trials.gov as NCT00557570 and NCT00286767.

Figures

Figure 1

Longitudinal analysis of total CD4+ and CD8+ T-cell phenotypes of IRIS and non-IRIS patients during ART. Characteristics of total CD4+ (A,C) and CD8+ T cells (B,D) were analyzed in PBMC samples from IRIS (red) and non-IRIS patients (blue) before as well as after 1, 3, 6, and 12 months of ART. (A-B) Activation phenotype and presence of recent thymic emigrants (RTE; in CD4+ T cells only). (C-D) Representation of T-cell differentiation states. T-cell differentiation subsets were defined by expression of CD45RO (“RO”), CCR7 (“R7”), and CD27 (“27”). TNV indicates naive; TCM, central memory; TTM, transitional memory; TEM, effector memory; and TTE, terminal effector. TCM*, TTM*, and TTE* represent phenotypically defined populations that are not described in the literature but that arise by this gating scheme; their activation phenotype and cytokine potential most closely resemble that of TCM, TTM, and TTE, respectively; hence their nomenclature. Graphs represent interquartile ranges, median bars, as well as individual data points. Gray boxes indicate the first time point within 3 months of clinical manifestation of IRIS. Dashed lines separate pre-ART from on-ART samples. All time points were compared between patient groups (results indicated in black above bars/pies), as well as with corresponding pre-ART measurements within each patient group (results are color-coded and indicated below bars/pies). *P < .01; **P < .001; ***P < .0001.

Figure 2

The magnitude of HIV-1 Gag-specific T cells does not differ significantly between IRIS and non-IRIS patients. The total response magnitude, measured by production of IFN-γ and/or IL-2 and/or TNF, generated by HIV-1 Gag reactive CD4+ (A) and CD8+ T cells (B) of IRIS and non-IRIS patients were compared at the 5 analysis time points. Graphs represent interquartile ranges, median bars, as well as individual data points. Dashed lines separate pre-ART from on-ART samples. Gray boxes indicate samples from IRIS patients within 3 months of clinical IRIS onset. All time points were compared between patient groups (no statistically significant differences found) and to corresponding pre-ART measurements within each patient group: red asterisks for IRIS; and blue asterisks, non-IRIS (indicated below graphs). *P < .01; **P < .001.

Figure 3

Cytokine pattern and phenotype of HIV-1 Gag-specific T cells do not differ significantly between patient groups. HIV-1 Gag-reactive CD4+ (A-B) and CD8+ T cells (C-D) of IRIS and non-IRIS patients were compared at the 5 analysis time points. (A,C) Cytokine pattern. Relative proportion of total HIV-1 Gag-reactive cells producing each possible combination of the cytokines measured. (B,D) Differentiation state. Dashed lines separate pre-ART from on-ART samples. Gray boxes indicate samples from IRIS patients within 3 months of clinical IRIS onset. All time points were compared between patient groups and with corresponding pre-ART measurements within each patient group (no statistically significant differences found). Individual pie segments were also compared between time points within each patient group (indicated within relevant pie segments): *P < .01; **P < .001. i.d. indicates insufficient data/(not enough samples met inclusion criteria).

Figure 4

Elevated CD4+ T-cell responses to relevant Ags in patients with mycobacterial- or fungal-associated IRIS events. CD4+ T-cell responses to TB, MAC, C neoformans, and H capsulatum were analyzed. For this purpose, patients were divided into 3 groups: those with IRIS events to mycobacterial- or fungal-associated OI and stimulated with the relevant Ag (mycobacterial/fungal-associated IRIS; red), those with IRIS events to other Ags, which can include mycobacterial/fungal Ags, but stimulated here with IRIS-irrelevant Ags (other IRIS; pink), and non-IRIS patients (non-IRIS; blue). Each data point represents stimulation with one Ag only. Shown are the response magnitudes by Ag: TB (A), MAC (B), C neoformans (C), and H capsulatum (D), as well as grouped for all 4 Ags (E). (A-D) Longitudinal data points are only connected for IRIS patients with IRIS manifestations associated with the given Ag. Cytokine pattern (F), T-cell subset distribution (G), and activation phenotype (H; tp3 only) of CD4+ T cells reactive to TB, MAC, C neoformans, or H capsulatum were also determined. Gray boxes indicate the first time point within 3 months of clinical manifestation of IRIS. Dashed lines separate pre-ART from on-ART samples. All groups were compared within time points with the mycobacterial/fungal-associated IRIS group (color-coded asterisks above bars or pies), as well as with corresponding pre-ART measurements within groups: red for IRIS; and blue asterisks, non-IRIS (indicated below graphs or pies). (G) The number of samples per pie are indicated (see “Data analysis”; these numbers also apply to data in panel H. i.d. indicates insufficient data. *P < .05; **P < .01; ***P < .001).

Figure 5

During ART, only frequencies of CD4+ cytokine-producing T cells specific to IRIS-associated Ags increase dramatically. PBMC samples from 5 IRIS patients (A) and 5 non-IRIS patients (B) were stimulated with MAC, H capsulatum, C neoformans, TB, CMV, JCV, or HIV-1. The IRIS-associated opportunistic infection and other known infecting pathogens tested for in the present assays are indicated for each patient. The number of stimulations performed with each PBMC sample was determined by the number of cells available, and priorities were given to those Ags to which a given patient was known to have been exposed. Gray boxes indicate the first time point within 3 months of clinical manifestation of IRIS. Dashed lines separate pre-ART from on-ART samples. Bold lines indicate T-cell responses to IRIS-associated Ags. IRIS patients were selected for illustration if stimulation data were available for at least 2 Ags and at least 4 time points. If 2 Ags fulfilled these criteria, all other stimulations with at least 2 data points were shown for that patient. Non-IRIS patients were selected according to the aforementioned criteria, as well as having known to be exposed to at least one of the Ags being tested: 1 indicates tp1; 2, tp2; 3, tp3; 4, tp4; and 5, tp5.