Elevated glycolysis imparts functional ability to CD8+ T cells in HIV infection

Abstract

The mechanisms inducing exhaustion of HIV-specific CD8⁺ T cells are not fully understood. Metabolic programming directly influences T-cell differentiation, effector function, and memory. We evaluated metabolic profiles of ex vivo CD8⁺ T cells in HIV-infected individuals. The baseline oxygen consumption rate of CD8⁺ T cells was elevated in all infected individuals and CD8⁺ T cells were working at maximal respiratory capacity. The baseline glycolysis rate was enhanced only during early untreated HIV and in viral controllers, but glycolytic capacity was conserved at all stages of infection. CD8⁺ T-cell mTOR activity was found to be reduced. Enhanced glycolysis was crucial for HIV-specific killing of CD8⁺ T cells. CD8⁺ T-cell cytoplasmic GAPDH content was reduced in HIV, but less in early infection and viral controllers. Thus, CD8⁺ T-cell exhaustion in HIV is characterized by reduced glycolytic activity, enhanced OXPHOS demands, dysregulated mTOR, and reduced cytoplasmic GAPDH. These data provide potential metabolic strategies to reverse CD8⁺ T-cell dysfunction in HIV.

Figure S1.. Schema of Seahorse extracellular flux assay and time courses with metabolic inhibitors.

(A, B) illustrate schematic diagrams of oxygen consumption rate (OCR) and ECAR time courses of isolated CD8⁺ T cells under basal conditions and following perturbation of the mitochondrial respiration with oligomycin, FCCP, and antimycin A/Rotenone. These profiles were generated by Seahorse XFe96 extracellular flux analyzer. Non-mitochondrial OCR = [Antimycin A/Rotenone OCR]; Basal Respiration OCR = [Basal OCR] − [Non-mitochondrial OCR]; ATP linked OCR = [Basal OCR] − [Oligomycin OCR]; Spare Reserve Capacity OCR = [FCCP Peak OCR] − [Basal OCR]; Baseline Glycolytic ECAR = [Basal ECAR]; Maximum Glycolytic ECAR = [Antimycin A/Rotenone ECAR]; Glycolytic Reserve ECAR = [Maximum Glycolytic ECAR] − [Baseline Glycolytic ECAR].

Figure 1.. Oxygen consumption rate (OCR) and ECAR metabolic profiles of ex vivo overnight rested or activated CD8⁺ T cells from HIV early–infected cART–naïve and cART–treated donors.

(A) illustrates a representative OCR profile of overnight-rested CD8⁺ T cells taken ex vivo from HIV-uninfected (black line), HIV-early cART–naïve (red line), and HIV early–cART–treated (orange line) donors. These are prospectively matched samples from the same donors before and after being on cART treatment for at least 1 yr. (B, C, D) Hollow bars represent baseline, ATP-linked, and spare capacity OCRs of overnight-rested CD8⁺ T cells from HIV-uninfected (black line, n = 5), HIV-early cART–naïve (red line, n = 6), and HIV-early cART–treated (orange line, n = 5) donors. (E) illustrates a representative ECAR profile of ex vivo overnight-rested CD8⁺ T cells from HIV-uninfected (black line), HIV-early cART–naïve (red line), and HIV-early cART–treated (orange line) donors. (F, G, H) Hollow bars represent baseline, maximal, and glycolytic capacity ECARs of overnight-rested CD8⁺ T cells from HIV-uninfected (black line, n = 5), HIV-early cART–naïve (red line, n = 6), and HIV-early cART–treated (orange line, n = 5) donors. Data are representative of at least three technical repeats. * indicates P < 0.05, **P < 0.01, and ***P < 0.001 by unpaired two tailed nonparametric Mann–Whitney test. Error bars are mean ± SEM.

Figure 2.. Oxygen consumption rate (OCR) and ECAR metabolic profiles of ex vivo overnight-rested or activated CD8⁺ T cells from HIV-chronic cART–naïve and chronic CART–treated donors.

(A) illustrates a representative OCR profile of overnight-rested CD8⁺ T cells from HIV-uninfected (black line), HIV-chronic cART–naïve (blue line), and HIV-chronic cART–treated (purple line) donors. (B, C, D) Hollow bars represent baseline, ATP-linked, and spare capacity OCRs of overnight-rested CD8⁺ T cells from HIV-uninfected (black line, n = 5), HIV-chronic cART–naïve (blue line, n = 7), and HIV-chronic cART–treated (purple line, n = 6) donors. (E) illustrates a representative ECAR profile of overnight-rested CD8⁺ T cells from HIV-uninfected (black line), HIV-chronic cART–naïve (blue line), and HIV-chronic cART–treated (purple line) donors. (F, G, H) Hollow bars represent baseline, maximal, and glycolytic capacity ECARs of overnight-rested CD8⁺ T cells from HIV-uninfected (black line, n = 5), HIV-chronic CART–naïve (blue line, n = 7), and HIV-chronic cART–treated (purple line, n = 6) donors. Data are representative of at least three technical repeats. * indicates P < 0.05, **P < 0.01, and ***P < 0.001 by unpaired two-tailed nonparametric Mann–Whitney test. Error bars are mean ± SEM.

Figure 3.. Oxygen consumption rate (OCR) and ECAR metabolic profiles of overnight-rested or activated CD8⁺ T cells from viral controllers (VCs).

(A) illustrates a representative OCR profile of overnight rested CD8⁺ T cells from HIV-uninfected (black line) and VC (green line) donors. (B, C, D) Hollow bars represent baseline, ATP-linked, and spare capacity OCRs of overnight-rested CD8⁺ T cells from HIV-uninfected (black line, n = 5), and VC (green line, n = 6) donors. (E) illustrates a representative ECAR profile of overnight rested CD8⁺ T cells from HIV-uninfected (black line) and VC (green line) donors. (F, G, H) Hollow bars represent baseline, maximal, and glycolytic capacity ECARs of overnight-rested CD8⁺ T cells from HIV-uninfected (black line, n = 5) and VC (green line, n = 6) donors. Data are representative of at least three technical repeats. * indicates P < 0.05, **P < 0.01, and ***P < 0.001 by unpaired two-tailed nonparametric Mann–Whitney test. Error bars are mean ± SEM.

Figure 4.. Summary of metabolic profile data. A color-coded schematic diagram is provided to summarize the metabolic data generated with Seahorse extracellular flux analyzer.

(A) summarizes the data of overnight rested (unstimulated) CD8⁺ T cells from all the different HIV cohorts compared with overnight-rested CD8⁺ T cells from HIV-uninfected donors. (B) summarizes the data of overnight anti-CD3/28 stimulated CD8⁺ T cells from all the different HIV cohorts against overnight anti-CD3/28–stimulated CD8⁺ T cells from HIV-uninfected donors. (C) summarizes the data of overnight-activated CD8⁺ T cells from all the different cohorts against overnight-rested autologous CD8⁺ T cells. The legend for the color code is on the top right corner of the figure.

Figure S2.

Oxygen consumption rate (OCR) metabolic profiles of overnight rested or activated CD8⁺ T cells. (A) illustrates a representative OCR profiles of overnight rested CD8⁺ T cells from HIV-uninfected (black line), HIV-early cART–naïve (red line), HIV-early cART–treated (orange line), HIV-chronic cART–naïve (blue), HIV-chronic cART–treated (purple), and viral controllers (VCs) (green). (B) illustrates representative OCR profiles of overnight-activated CD8⁺ T cells from HIV-uninfected (black line), HIV-early cART–naïve (red line), HIV-early cART–treated (orange line), HIV-chronic cART–naïve (blue), HIV-chronic cART–treated (purple), and VC (green) donors. (C, D, E) represents summary data of overnight rested (hollow bars) and overnight-activated (dotted bars) baseline, ATP-linked, and spare capacity OCRs of HIV-uninfected (black line), HIV-early cART–naïve (red line), HIV-early cART–treated (orange line), HIV-chronic cART–naïve (blue), HIV-chronic cART–treated (purple), and VC (green) donors. Data are representative of at least three technical repeats and compiled from Figs 1–3. * indicates P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001 by unpaired two-tailed nonparametric Mann–Whitney test. Error bars are mean ± SEM.

Figure S3.

ECAR metabolic profiles of overnight-rested or activated CD8⁺ T cells. (A) illustrates a representative ECAR profile of overnight-rested CD8⁺ T cells from HIV-uninfected (black line), HIV-early cART–naïve (red line), HIV-early cART–treated (orange line), HIV-chronic cART–naïve (blue), HIV-chronic cART–treated (purple), and viral controllers (VCs) (green). (B) illustrates a representative ECAR profile of overnight-activated CD8⁺ T cells from HIV-uninfected (black line), HIV-early cART–naïve (red line), HIV-early cART–treated (orange line), HIV-chronic CART–naïve (blue), HIV-chronic CART–treated (purple), and VC (green) donors. (C, D, E) represents overnight-rested (hollow bars) and overnight-activated (dotted bars) baseline, maximal, and glycolytic capacity oxygen consumption rates of HIV-uninfected (black line), HIV-early cART–naïve (red line), HIV-early cART–treated (orange line), HIV-chronic cART–naïve (blue), HIV-chronic cART–treated (purple), and VC (green) donors. Data are representative of at least three technical repeats and compiled from Figs 1–3. * indicates P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001 by unpaired two-tailed nonparametric Mann–Whitney test. Error bars are mean ± SEM.

Figure 5.. mTORC1 signaling and surface transport channels in CD8⁺ T cells.

(A) The basal mTORC1 activity in CD8⁺ T cells after 5 h of rest or activation with plate bound anti-human CD3 and anti-human CD28 was determined by PhosFlow in HIV-uninfected (black bar, n = 3), HIV-early cART–naïve (red bar, n = 3), and HIV-chronic cART–treated (purple bar, n = 4) donors by determining the frequency of pS6 in CD8⁺ T cells. (B) represents the frequency of CD98 expression on the surface of HIV-uninfected (black bar, n = 7), HIV-early cART (cTx)–naïve (red bar, n = 20), and HIV-chronic cART–treated (purple bar, n = 10) CD8⁺ T cells. (C) represents the median fluorescent index of CD98 expression on the surface of HIV-uninfected (black bar, n = 7), HIV-early cART–naïve (red bar, n = 20), and HIV-chronic cART–treated (purple bar, n = 10) CD8⁺ T cells. (D) represents the frequency of CD71 expression of the surface of HIV-uninfected (black bar, n = 7), HIV-early cART–naïve (red bar, n = 20), and HIV-chronic cART–treated (purple bar, n = 10) CD8⁺ T cells. (E) represents the median fluorescent index of CD71 expression on the surface of HIV-uninfected (black bar, n = 7), HIV-early cART–naïve (red bar, n = 20), and HIV-chronic cART–treated (purple bar, n = 10) CD8⁺ T cells. * indicates P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001 by unpaired two-tailed nonparametric Mann–Whitney test. Error bars are mean ± SEM.

Figure 6.. mTORC1 inhibition does not rescue exhausted HIV-specific CD8⁺ T-cell function.

(A) flowchart of the experimental design. (B) illustrates a representative data set from an HIV-chronic cART–naïve donor. The CD8⁺ T-cell function was plotted with CFSE on the x-axis (proliferation) against intracellular IFN-γ plus TNF-α on the y-axis (response to antigen or mitogen). Columns represent rapamycin treatment and rows represent stimulation conditions. (C) represents the antigen specific functional output of CD8⁺ T cells from HIV-chronic cART–naïve donors pretreated with 50 nM of rapamycin. (D) represents the antigen-specific functional output of CD8⁺ T cells from HIV-chronic cART–naïve donors treated with 12.5, 25, and 50 nM of rapamycin during the 6 d culture. (E) represents the antigen specific functional output of CD8⁺ T cells from HIV-chronic cART–treated donors pretreated with 50 nM of rapamycin. (F) represents the antigen-specific functional output of CD8⁺ T cells from HIV-chronic cART–treated donors treated with 12.5, 25, and 50 nM of rapamycin during the 6 d culture. HIV-chronic cART–naïve n = 6, and HIV-chronic cART–treated n = 6. Error bars are mean ± SEM.

Figure 7.. Effect of metabolic inhibitors on HIV-specific killing capacity of CD8⁺ T cells.

(A) Schematic of killing assay experimental setup. Table indicates proposed effect of inhibitors on metabolism. (B, C, D, E, F, G) The killing capacity of CD8⁺ T cells from overnight rested (n = 11) donors were compared against CD8⁺ T cells that were either: (B) activated overnight (n = 11) or (C) had 20 mM of 2-DG (n = 10) or (D) 10 mM of DCA (n = 8) or (E) 1 μM of oligomycin (n = 11) during killing assay, or (F) were pretreated with 1 μM of oligomycin (n = 8) or (G) 10 mM of DCA (n = 5) and washed before setting up the killing assay. * indicates P < 0.05 by unpaired two tailed nonparametric Mann–Whitney test.

Figure S4.

ImageStream figures of GAPDH in CD8⁺ T Cells. After gating on focused singlet CD8⁺ T cells, a double-positive gate on GAPDH and nuclear dye was set within the Dextramer negative (Dextramer−) or Dextramer positive (Dextramer+) populations of CD8⁺ T cells to determine the nuclear localization of GAPDH using similarity feature. The negative gate on nuclear localization was set as the gate for cytoplasmic GADPH. (A) represents the GAPDH localization within the CD8⁺ T cells outside the CMV dextramer− population in an HIV-uninfected donor. (B) represents the GAPDH localization within the CD8⁺ T cells within the CMV dextramer+ population in an HIV-uninfected donor. (C) represents the GAPDH localization within the CD8⁺ T cells within the CMV dextramer− population in an HIV-early cART–naïve donor. (D) represents the GAPDH localization within the CD8⁺ T cells within the CMV dextramer+ population in an HIV-early cART–naïve donor. (E) represents the GAPDH localization within the CD8⁺ T cells within the HIV dextramer− population in an HIV-early cART–naïve donor. (F) represents the GAPDH localization within the CD8⁺ T cells within the HIV dextramer+ population in an HIV-early cART–naïve donor. The number underneath the cGAPDH gate represents the proportion of CD8⁺ T cells with proportionally more GAPDH present in the cytoplasm.

Figure 8.. GAPDH localization in ex vivo CD8⁺ T cells from HIV-uninfected and HIV-infected individuals.

(A) represents the frequency of CMV nonspecific (dextramer−) and CMV-specific (dextramer+) CD8⁺ T cells with cGAPDH from HIV-uninfected (n = 3) donors. (B) represents the frequency of dextramer− and CMV specific (dextramer+) CD8⁺ T cells with cGAPDH from HIV-early cART–naïve (n = 4) donors. (C) represents the frequency of dextramer− and CMV-specific (dextramer+) CD8⁺ T cells with cGAPDH from HIV-early cART–treated (n = 3) donors. (D) represents the frequency of dextramer− and CMV-specific (dextramer+) CD8⁺ T cells with cGAPDH from HIV-chronic cART–treated (n = 3) donors. (E) represents the frequency of dextramer− and CMV-specific (dextramer+) CD8⁺ T cells with cGAPDH from viral controller (VC) (n = 3) donors. (F) represents the median fluorescent index (MFI) of cytoplasmic GAPDH intensity of CMV nonspecific (dextramer−) and CMV-specific (dextramer+) CD8⁺ T cells from HIV-uninfected (n = 3) donors. (G) represents the MFI of cytoplasmic GAPDH intensity of dextramer− and CMV-specific (dextramer+) CD8⁺ T cells from HIV-early cART–naïve (n = 4) donors. (H) represents the MFI of cytoplasmic GAPDH intensity of dextramer− and CMV-specific (dextramer+) CD8⁺ T cells from HIV-early cART–treated (n = 3) donors. (I) represents the MFI of cytoplasmic GAPDH intensity of dextramer− and CMV-specific (dextramer+) CD8⁺ T cells from HIV-chronic cART–treated (n = 3) donors. (J) represents the MFI of cytoplasmic GAPDH intensity of dextramer− and CMV-specific (dextramer+) CD8⁺ T cells from VC (n = 4) donors. (K) represents the frequency of dextramer− and HIV-specific (dextramer+) CD8⁺ T cells with cGAPDH from HIV-early cART–naïve (n = 3) donors. (L) represents the frequency of dextramer− and HIV-specific (dextramer+) CD8⁺ T cells with cGAPDH from HIV-early cART–treated (n = 3) donors. (M) represents the frequency of dextramer− and CMV-specific (dextramer+) CD8⁺ T cells with cGAPDH from HIV-chronic cART–treated (n = 3) donors. (N) represents the frequency of dextramer− and HIV-specific (dextramer+) CD8⁺ T cells with cGAPDH from VC (n = 4) donors. (O) represents the MFI of cytoplasmic GAPDH intensity of dextramer− and HIV-specific (dextramer+) CD8⁺ T cells from HIV-early cART–naïve (n = 3) donors. (P) represents the MFI of cytoplasmic GAPDH intensity of dextramer− and HIV-specific (dextramer+) CD8⁺ T cells from HIV-early cART–treated (n = 3) donors. (Q) represents the MFI of cytoplasmic GAPDH intensity of dextramer− and HIV-specific (dextramer+) CD8⁺ T cells from HIV-chronic cART–treated (n = 3) donors. (R) represents the MFI of cytoplasmic GAPDH intensity of dextramer− and HIV-specific (dextramer+) CD8⁺ T cells from VC (n = 4) donors. The dextramer− population was combined within each of the different HIV-infected cohorts for analysis purposes. Error bars are mean ± SEM.

Figure 9.. PD-1 expression frequency on CD8⁺ T cells with oxygen consumption rate (OCR) versus ECAR plot for CD8⁺ T cells taken from HIV-uninfected and HIV-infected individuals.

(A) represents the frequency of PD-1 expression on the surface of CD8⁺ T cells from HIV-uninfected (n = 3), HIV-early cART–naïve (n = 4), HIV-chronic cART–naïve (n = 5), HIV-chronic cART–treated (n = 5), and viral controllers (VCs) (n = 5). (B) represents the baseline OCR versus baseline ECAR plot of HIV-uninfected (black), HIV-early cART–naïve (red), HIV-early cART–treated (orange), HIV-chronic cART–naïve (blue), HIV-chronic cART–treated (purple), and VCs (green) of ex vivo overnight-resting (unstimulated) CD8⁺ T cells. (C) represents the baseline OCR versus baseline ECAR plot of HIV-uninfected (black), HIV-early cART–naïve (red), HIV-early cART–treated (orange), HIV-chronic cART–naïve (blue), HIV-chronic cART–treated (purple), and VCs (green) after overnight CD3/28 activation of ex vivo CD8⁺ T cells. These data were compiled from the data set from Figs 1–3. Dotted lines represent the points on the graph where ECAR = OCR.