Single-cell profiling identifies a CD8bright CD244bright Natural Killer cell subset that reflects disease activity in HLA-A29-positive birdshot chorioretinopathy

Abstract

Birdshot chorioretinopathy is an inflammatory eye condition strongly associated with MHC-I allele HLA-A29. The striking association with MHC-I suggests involvement of T cells, whereas natural killer (NK) cell involvement remains largely unstudied. Here we show that HLA-A29-positive birdshot chorioretinopathy patients have a skewed NK cell pool containing expanded CD16 positive NK cells which produce more proinflammatory cytokines. These NK cells contain populations that express CD8A which is involved in MHC-I recognition on target cells, display gene signatures indicative of high cytotoxic activity (GZMB, PRF1 and ISG15), and signaling through NK cell receptor CD244 (SH2D1B). Long-term monitoring of a cohort of birdshot chorioretinopathy patients with active disease identifies a population of CD8^bright CD244^bright NK cells, which rapidly declines to normal levels upon clinical remission following successful treatment. Collectively, these studies implicate CD8^bright CD244^bright NK cells in birdshot chorioretinopathy.

Fig. 1. Flow-cytometry profiling revealed altered Natural killer (NK) cell composition in peripheral blood of birdshot uveitis.

A Schematic representation of flow cytometry of major immune cell lineages in fresh peripheral blood of 139 uveitis patients (UV) and 80 age-, sex-matched healthy controls (HC). Figure is created with BioRender.com (www.biorender.com), released under a Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International license. B Flow-cytometry quantification of the percentage of NK cells in the peripheral blood of uveitis patients (Uveitis) compared to healthy controls (Healthy). Healthy n = 80; Uveitis n = 139. ***P = 0.0006. Data are presented as mean values ± SEM. Source data are provided as a Source Data file. C Flow-cytometry analysis of NK cells (CD3^-CD19^-CD56⁺) in the fresh blood of different uveitis subgroups indicate the significant expansion of NK cells were restricted to birdshot, definite sarcoidosis and serpiginous sub-groups of uveitis cohort. Values are presented in the form of box and whiskers plot and represent medians with ranges (Whiskers: two-sided Tukey test). Healthy n = 80; Uveitis n = 139; WDS n = 18; Birdshot n = 18; VKH n = 15; Definite Sarcoidosis n = 8; Presumed Sarcoidosis n = 8; Serpiginous n = 7; Undifferentiated n = 65. P values are from non-parametric Mann–Whitney U test, ****P < 0.0001, ***P = 0.0006, **P = 0.004, *P = 0.02. WDS, White Dot Syndromes; VKH, Vogt–Koyanagi–Harada disease. Source data are provided as a Source Data file. D The flow-cytometry gating strategy for CD56^bright and CD56^dim subsets of NK cells using CD56 and CD16 in peripheral blood. E NK cell and CD56^bright and CD56^dim subset quantification in peripheral blood of birdshot uveitis (BCR-UV) and age-matched healthy controls (HC). HC n = 15; BCR-UV n = 18. P values are from two-sided unpaired t test, ***P = 0.0004, **P = 0.002, *P = 0.01. Data are presented as mean values ± SEM. Source data are provided as a Source Data file. F The fluorescence intensity of TNF and IFN-γ produced by all NK cells, as determined by flow cytometry analysis upon stimulation with Lymphocyte Activation Cocktail (BD Biosciences). Analysis was conducted by stimulating PBMCs and flow-gating on NK cells from BCR-UV (red) and healthy controls (HC, blue). HC n = 8; BCR-UV n = 8. Statistical analysis was done using one-way ANOVA- Tukey’s multiple comparisons test; P values are indicated above the histograms. Source data are provided as a Source Data file.

Fig. 2. Single-cell RNAseq analysis identifies transcriptional NK cell subsets.

A UMAP plot of 306,425 PBMCs from 12 birdshot uveitis patients (BCR-UV) and 12 healthy controls (HC). The identified lineage clusters are highlighted (color-coded). B Dot plot showing the expression profile of NK surface marker-encoding genes across the annotated cell clusters identified in A. C UMAP plots of NK cells gated from the C4_Natural Killer cells cluster in A comparing NK subclusters in BCR-UV and HC. The 12 transcriptional NK cell subclusters are color-coded. D Dot plot showing the expression profile of NK surface marker encoding genes in each of the subclusters as identified in C.

Fig. 3. Single-cell RNAseq analysis delineates NK cell subsets associated with birdshot uveitis.

A A total of 306,425 PBMCs from 12 birdshot uveitis patients (BCR-UV) and 12 healthy controls (HC) are analyzed by single-cell RNAseq. Dot plot shows two-sided proportion test analysis of NK cell clusters gated from the C4_Natural Killer cells cluster in Fig. 2A and comparing between BCR-UV and HC. Significant differences at a false discovery rate of 5% [FDR < 0.05] and log2 fold difference = [Log2(FD)], are indicated in red. The data points represent log2 fold difference, with error bars showing the estimated 95% confidence interval, based on the distribution of bootstrapped log2 fold differences. B Heatmap of the gene expression profile of highly expressed genes (n = 10 unique genes) in clusters 1, 2, 6, and 10 of NK cells as identified in A.

Fig. 4. CD244^bright CD8^bright NK cell subset is significantly expanded in peripheral blood of birdshot uveitis patients.

A t-SNE plot of flow cytometry data from Healthy control (HC) n = 11, birdshot uveitis (BCR-UV) n = 18. Clusters identified by FlowSOM analysis (n = 12 clusters) are shown. B Comparison of the frequency of NK clusters between BCR-UV and Healthy controls. The -log10 (nominal P values) are from a two-sided likelihood ratio test. Significantly different clusters 0, 4, and 5 are highlighted. Source data are provided as a Source Data file. C Scatter plot of the frequency of cells of cluster 0 in peripheral blood in patients versus controls. HC n = 11, BCR-UV n = 18. ***P = 0.002. Data are presented as mean values ± SEM. Source data are provided as a Source Data file. D Heatmap showing the mean fluorescent intensity (MFI) of surface markers in NK clusters. Source data are provided as a Source Data file. E Principal component analysis (PCA) of NK clusters based on expression of surface markers. Source data are provided as a Source Data file. F Representative dot plots and corresponding frequency of CD8a⁺CD244⁺ NK cells in the peripheral blood of Healthy controls (HC, blue) and birdshot uveitis patients (BCR-UV, red). HC, n = 10; BCR-UV, n = 15. Statistical comparison is done using two-sided unpaired t test. ***P = 0.0003. Data are presented as mean values ± SEM. Source data are provided as a Source Data file.

Fig. 5. IL-15/IL-18-mediated stimulation increased CD244 expression in NK cells and CD8 + NK cells birdshot uveitis patients show elevated activation profile.

A Histograms of the surface expression of CD69 and CD244, after 48 hrs stimulation of isolated peripheral blood NK cells from healthy donors. with recombinant human IL-15, IL-18 and CXCL-16 proteins. n = 3 B Frequencies of TNF and IFN-γ producing CD8⁺ NK cells, as determined by flow cytometry analysis upon stimulation with Lymphocyte Activation Cocktail (BD Biosciences). Analysis was conducted using PBMCs from BCR-UV (red) and NK cells from healthy controls (HC, blue). HC n = 5; BCR-UV n = 5. Statistical analysis was done using one-way ANOVA- Tukey’s multiple comparisons test; ****P < 0.0001, ***P = 0.0003. Data are presented as mean values ± SEM. Source data are provided as a Source Data file. C CD56⁺ CD8⁺CD244⁺ NK cell population (representative gating strategy is shown on the left) from age-matched BCR-UV patients and healthy controls (HC) are tested for the surface expression of NKG2D (CD314) and NKp30 (CD337) proteins (bar plots on the right). n = 6. **P = 0.001. Data are presented as mean values ± SEM. Source data are provided as a Source Data file.D CD244⁺CD8A⁺ cells from C4_Natural Killer Cells cluster in the single-cell RNAseq analysis (subcluster shown as in Fig. 2A) were analyzed for the expression of transcripts of the indicated inflammation-associated genes in BCR-UV and HC. The double positive population expressing at least one transcript copy of CD244 and CD8A were more in BCR-UV and expressed more inflammation-associated genes compared to the HC.

Fig. 6. Response to systemic immunomodulatory therapy is accompanied by normalization of circulating CD8⁺ CD244⁺ NK cells in patients.

A Representative Fluorescein angiography (FA) image the left and right eye of a birdshot uveitis (BCR-UV) patient at baseline and after 1 year of systemic immunomodulatory therapy. B Histograms are showing the expression of CD8a and CD244 proteins in circulating CD56^dim CD16⁺ NK cell surface of healthy control (HC), BCR-UV patient at the baseline (Base) and on month 6 (M06) and year 1 (Y01). Values within the box indicate the mean fluorescent intensity (MFI) of CD8a and CD244 proteins. C Representative flow plots show staining of CD8 and CD244 proteins within CD56^dim CD16⁺ NK cells in the blood of healthy control (HC), BCR-UV patient at baseline with no treatment (Base), and longitudinally at follow-up visits at month 6 (M06) and year 1 (Y01) after treatments. D Frequencies of CD244 and CD8 double positive cells within CD56^dim CD16⁺ NK population in age-matched healthy control (HC) and in BCR-UV at baseline (Base) and treatment follow up visits. HC, n = 10; Baseline, n = 15; Month 03, n = 4; Month 06, n = 4 and Year 01, n = 15. Statistical comparison is done using one-way ANOVA- Tukey’s multiple comparisons test. *P = 0.01. Data are presented as mean values ± SEM. Source data are provided as a Source Data file.