Immunology and Inflammation

CD56^dimCD16^dim NK cells are the dominant effector cells against HIV-infected primary T-cells

William Howell
Courtney Branch
Jeffrey Ward
Zachary Davis
Emma Geatches
Edward Barker author has email address

¹Department of Microbial Pathogens and Immunity, Rush University Medical Center, Chicago, United States
Division of Oncology, Department of Medicine, Washington University in St. Louis, St Louis, United States
Department of Medicine, University of Minnesota, Minneapolis, United States

https://doi.org/10.7554/eLife.112322.1

Open access
Copyright information

Figures and data

Degranulation of NK cell subsets based on differential expression of CD56 and CD16 in response to purified autologous HIV productively infected T-cells
(A) Frequency of NK cell subsets [mean + standard deviation (SD)] within total peripheral blood NK cells of an HIV-uninfected donor [the donor used for (B)], based on differential CD56 and CD16 expression. NKp80 was used to identify CD56^negative NK cell subsets. (B) Frequency of NK cell subsets expressing CD107a following a 4-hour exposure at a 1:1 effector cell to target cell (E:T) ratio to purified autologous productively HIV-1_SHM-1 infected T-cells. Each circle represents a technical replicate; lines indicate mean values. All values are background-subtracted (target-exposed minus unexposed control). (C) Frequency of CD107a expression (mean + SD) on NK cell subsets across six different donors. The CD56^negative CD16^dim subset was excluded because its frequency in peripheral blood (<0.2% of NK cells; Figure 1-figure supplement 1B) yields event counts insufficient for reliable measurement. (D) Frequency of CD56^dim NK cell subsets [from (C)] that were also CD16^bright, CD16^dim, or CD16^negative across the same six donors. Panels A and B show data from one representative donor with three technical replicates; panels C and D show data from six donors analyzed in separate experiments, with each donor evaluated independently to avoid confounding from KIR and HLA polymorphism. Statistical analyses: one-way ANOVA with Dunnett’s post-hoc against CD56^dimCD16^dim for panel B; repeated-measures one-way ANOVA with donor matching and Dunnett’s post-hoc for panels C and D. Full ANOVA tables and post-hoc results are in Supplemental Tables 2 and 3. Significance: * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001.

Lysis of autologous HIV-productively infected T-cells by purified CD56^dimCD16^dim and CD56^dimCD16^bright NK cells
Purified NK cells from an uninfected individual were sorted into CD56^dimCD16^dim and CD56^dimCD16^bright NK cell subsets. Carboxyfluorescein succinimidyl ester (CFSE)-labeled purified autologous productively HIV-1_SHM-1 infected T-cells generated in vitro were added to the sorted NK cell subsets at various effector-to-target cell (E:T) ratios in triplicate. After 4 hours, cells were surface-stained for CD56 and CD3, then exposed to 7-amino-actinomycin D (7AAD). Mean percent specific lysis ± standard deviation (SD) was calculated as (target cell-exposed minus mean unexposed target cells) / (mean positive control minus mean unexposed target cells) × 100, based on 7AAD staining of CFSE-labeled CD3^positiveCD56^negative cells. The data shown are from one representative donor and are representative of two independent sort experiments performed with separate donors, each yielding similar results. Statistical analysis: two-way ANOVA (E:T ratio × subset) with Šidák’s multiple comparisons test between subsets at each E:T ratio. Full ANOVA table and post-hoc results are in Supplemental Table 4 for Figure 2. Significance: * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001.

Killing frequency of autologous HIV-productively infected T-cells by CD56^dimCD16^dim and CD56^dimCD16^bright NK cells as they naturally exist within the total NK cell population
(A) Percentage of specific lysis and percentage of CD107a⁺ CD56^dimCD16⁺ NK cells when exposed to purified autologous HIV-1_SHM-1 productively infected T-cells across increasing effector cell to target cell (E:T) ratios. Values represent mean ± standard deviation (SD) from three technical replicates. (B) Percentage of CD107a expressing CD56^dimCD16^bright and CD56^dimCD16^dim NK cells when exposed to autologous HIV-1_SHM-1 productively infected T-cells across increasing E:T ratios. (C) NK cell killing frequency (killing frequency = killed targets / active effectors, where killed targets = total targets × % specific lysis / 100 and active effectors = total effectors × % degranulation / 100) calculated from data in panels A and B. Statistical analyses: two-way ANOVA (E:T ratio × subset) with Šidák’s multiple comparisons test between subsets at each E:T ratio for panels B and C. Full ANOVA tables and post-hoc results are in Supplemental Tables 5 and 6 for Figure 3B and 3C. Significance: * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001.

The number of NK cells that degranulated when exposed to HIV-infected cells over a 60-minute period.
Purified NK cells exposed to autologous HIV-1_SHM-1 productively infected T-cells at an effector cell-to-target cell ratio of 1:1 were treated with a different fluorophore-labeled anti-CD107a antibody every 15 minutes for 60 minutes, blocked with excess unlabeled anti-CD107a, and washed between each labeling step. The number of fluorophores accumulated on CD56^dimCD16^bright and CD56^dimCD16^dim NK cells determined how many times each NK cell degranulated during the 60-minute exposure. SD = standard deviation. Statistical analysis: two-way ANOVA (subset × number of degranulation events) with Šidák’s multiple comparisons test between subsets at each event category. Full ANOVA tables and post-hoc results are in Supplemental Table 7 for Figure 4. Significance: * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001.

Role of NKG2D in the ability of NK cell subsets to degranulate in response to autologous HIV-productively infected T-cells
(A) CD3 and CD28-stimulated CD4⁺ T-cells remained uninfected or were infected with VSV-G pseudotyped DHIV3, which possessed (wild-type; WT) or lacked (ΔVpr) Vpr. Following infection, cells were stained with a recombinant human soluble NKG2D/IgG1 Fc chimera protein, followed by a fluorochrome-conjugated anti-human IgG1 Fc-specific antibody (Ab). The staining control consisted of cells stained only with the anti-human IgG1 Fc-specific Ab. Cells were then stained with fluorochrome-conjugated antibodies against surface CD4 and intracellular HIV-1 p24 capsid and analyzed by flow cytometry. (B) NK cells were isolated from healthy uninfected donors by negative selection from peripheral blood mononuclear cells and exposed for 4 hours to uninfected CD4^positive T-cells or purified HIV-infected cells possessing or lacking Vpr. Cells were stained with fluorochrome-conjugated antibodies against CD56, CD3, and CD107a and analyzed by flow cytometry. Dot plots show CD107a versus CD56 expression on CD3− gated cells, with quadrants set by FMO controls; the upper-right quadrant indicates degranulating (CD107a+) CD56+ NK cells, with the percentage shown. (C) NK cells in triplicate wells were treated with blocking antibodies to NKG2D or an isotype-specific control antibody, then exposed to purified autologous HIV-1_SHM-1 productively infected T-cells at various effector-to-target (E:T) ratios for 4 hours. Cells were stained with antibodies against CD107a, CD56, CD16, and CD3 and analyzed by flow cytometry. All values represent background-subtracted percentages (target-exposed minus mean unexposed controls). Statistical analysis: separate two-way ANOVAs (E:T ratio × treatment) for CD56^dimCD16^bright and CD56^dimCD16^dim subsets, with Šidák’s multiple comparisons test between isotype and anti-NKG2D treatment at each E:T ratio. Full ANOVA tables and post-hoc results are in Supplemental Table 17 for Figure 5C. Significance: * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001.

NKG2D and NKp46 surface expression and degranulation responses of CD56^dimCD16^bright and CD56^dimCD16^dim NK cell subsets following HIV-infected T-cell recognition
(A) Geometric mean fluorescence intensity (gMFI) of NKG2D surface expression on CD56^dimCD16^bright and CD56^dimCD16^dim NK cells following 4-hour exposure to purified autologous HIV-1_SHM-1 productively infected T-cells across different effector cell-to-target cell ratios. (B) gMFI of NKp46 surface expression on the NKp46-positive cells within the CD56^dimCD16^bright and CD56^dimCD16^dim NK cell subsets under the same conditions as (A). (C) CD107a degranulation responses of CD56^dimCD16^bright and CD56d^imCD16^dim NK cells from (A) across different effector cell-to-target cell ratios. (D) CD107a degranulation responses of CD56^dimCD16^bright and CD56^dimCD16^dim NK cells stratified by NKp46 expression (NKp46+ vs NKp46−) at a 1:1 effector-to-target cell ratio. Values represent mean ± standard deviation (SD). All CD107a values represent background-subtracted percentages (target-exposed minus mean unexposed controls). Statistical analyses: two-way ANOVA (effector cell-to-target cell ratio × subset) with Šidák’s multiple comparisons test between subsets at each effector cell-to-target cell ratio for panels A, B, and C; two-way ANOVA (subset × NKp46 status) with Šidák’s multiple comparisons test for panel D. Full ANOVA tables and post-hoc results are in Supplemental Table 18 for Figure 6. Significance: * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001.

Degranulation of NK cell subsets in response to anti-gp120 labeled autologous HIV productively infected T-cells in the presence or absence of ADAM17 inhibition.
(A) Mean percentage ± standard deviation (SD) of CD56dimCD16bright and CD56dimCD16dim NK cells degranulating following 4-hour exposure at an effector-to-target ratio of 1:1 to purified autologous HIV-1SHM-1 productively infected T-cells pretreated with varying concentrations of anti-gp120 broadly neutralizing antibody (VRC01). All values represent background-subtracted percentages (target-exposed minus mean unexposed controls). (B) CD107a degranulation of CD56dimCD16bright (upper sub-panel) and CD56dimCD16dim (lower sub-panel) NK cells following exposure to VRC01-coated autologous HIV-1SHM-1 productively infected T-cells in the presence of an ADAM17 inhibitor in DMSO or DMSO vehicle alone. p-values next to the symbols are comparisons of the vehicle versus an ADAM17 inhibitor for each VRC01 concentration. (C) Change in the percentage of each CD16 subset following 4-hour exposure to purified autologous HIV-1_SHM-1 productively infected T-cells across VRC01 concentrations, calculated as the percentage of the subset under target exposure minus the percentage of the same subset without target exposure, under either DMSO vehicle or ADAM17 inhibition, shown for CD56^dimCD16^bright (upper sub-panel) and CD56^dimCD16^dim (lower sub-panel) NK cells. p-values next to the symbols are comparisons between the vehicle and the ADAM17 inhibitor at each VRC01 concentration. Statistical analyses: two-way ANOVA (subset × VRC01 concentration) with Šidák’s multiple comparisons test between subsets at each VRC01 concentration for panel A; three-way ANOVA (subset × treatment × VRC01 concentration) with Šidák’s multiple comparisons test between DMSO and ADAM17 inhibition at each VRC01 concentration within each sub-panel for panel B; separate two-way ANOVAs (treatment × VRC01 concentration) for the CD56^dimCD16^bright and CD56^dimCD16^dim sub-panels, with Šidák’s multiple comparisons test between DMSO and ADAM17 inhibition at each VRC01 concentration, for panel C. Full ANOVA tables and post-hoc results are in Supplemental Tables 19, 21, and 22 for Figures 7A, 7B, and 7C. Significance: * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001.

Degranulation of NK cell subsets in response to anti-gp120 labeled autologous HIV productively infected T-cells in the presence or absence of NKG2D blocking antibodies and/or ADAM17 inhibition
(A) Mean percentage ± standard deviation (SD) of CD56^dimCD16^bright and CD56^dimCD16^dim NK cells degranulating following 4-hour exposure at an effector-to-target ratio of 1:1 to purified autologous HIV-1_SHM-1 productively infected T-cells pretreated with 1 µg/mL of anti-gp120 broadly neutralizing antibody (VRC01). NK cells were treated with DMSO vehicle, anti-NKG2D blocking antibody, ADAM17 inhibitor, or anti-NKG2D blocking antibody combined with ADAM17 inhibitor before being added to VRC01-coated HIV-infected T-cells. All values represent background-subtracted percentages (target-exposed minus mean unexposed controls). (B) Change in the frequency of CD56^dimCD16^bright and CD56^dimCD16^dim NK cells under each treatment condition, calculated as the percentage of the subset following 4-hour exposure to VRC01-coated HIV-infected T-cells minus the percentage of the same subset under matched treatment in the absence of target cells and antibody. Statistical analyses: two-way ANOVA (subset × treatment) with Tukey’s multiple comparisons test for panel A; two-way ANOVA (subset × treatment) with Šidák’s multiple comparisons test for panel B. Full ANOVA tables and post-hoc results are in Supplemental Tables 23 and 24 for Figures 8A and 8B. Significance: * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001

The number of NK cell subsets that degranulated when exposed to anti-gp120 Ab-coated HIV-infected cells over 60 minutes.
Purified autologous HIV-1_SHM-1 productively infected T-cells were coated with 1 µg/mL of the broadly neutralizing anti-gp120 antibody VRC01. NK cells, pretreated with DMSO vehicle or an ADAM17 inhibitor dissolved in DMSO, were exposed to the coated target cells at an effector-to-target cell ratio of 1:1 and treated with a different fluorophore-labeled anti-CD107a antibody every 15 minutes for 60 minutes, with each labeling step blocked by excess unlabeled anti-CD107a and washed between steps. The number of fluorophores accumulated on CD56^dimCD16^bright (upper panel) and CD56d^imCD16^dim (lower panel) NK cells determined how many times each cell degranulated during the 60-minute exposure. SD = standard deviation. Statistical analyses: separate two-way ANOVAs (treatment × number of degranulations) for the CD56^dimCD16^bright panel (with Šidák’s multiple comparisons test) and the CD56^dimCD16^dim panel (with Dunnett’s multiple comparisons test), comparing DMSO and ADAM17 inhibition at each degranulation count. Full ANOVA tables and post-hoc results are in Supplemental Tables 25 and 26 for Figure 9. Significance: * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001.

Sign up for email alerts