Tissue resident memory T cells in the human conjunctiva and immune signatures in human dry eye disease

Abstract

Non-recirculating resident memory (T_RM) and recirculating T cells mount vigorous immune responses to both self and foreign antigens in barrier tissues like the skin, lung and gastrointestinal tract. Using impression cytology followed by flow cytometry we identified two T_RM subsets and four recirculating T-subsets in the healthy human ocular surface. In dry eye disease, principal component analysis (PCA) revealed two clusters of patients with distinct T-cell signatures. Increased conjunctival central memory and naïve T cells characterized Cluster-1 patients, and increased CD8⁺ T_RMs and CD4⁺ recirculating memory T cells characterized Cluster-2 patients. Interestingly these T-cell signatures are associated with different clinical features: the first signature correlated with increased ocular redness, and the second with reduced tear break up times. These findings open the door to immune-based characterization of dry eye disease and T-subset specific immunotherapies to suppress T-subsets involved in disease. They may also help with patient stratification during clinical trials of immunomodulators.

Figure 1. T-cell surface markers in normal ocular surface.

(A) Cell surface markers for naïve, T_CM, T_EM and T_EMRA subsets used in our studies. (B) Flow cytometry showing proportion of naïve, T_CM, T_EM and T_EMRA subsets within CD4⁺ and CD8⁺ pools (left and right panels) in the nomal human conjunctiva. (C) Proportion of conjunctival naïve, T_CM, T_EM and T_EMRA subsets in CD4⁺ and CD8⁺ T cell pools expressed as a percentage of the total number of CD3⁺ T cells in healthy human conjunctiva. Each data point represents a separate individual; mean ± SEM shown.

Figure 2. Tissue resident memory T cells predominate in the human ocular surface.

(A) Cell surface markers for T_RM and T_RCM subsets used in our studies. (B) Flow cytometry showing proportion of CD69⁺CD103⁻T_RM, CD69⁺CD103⁺ T_RM and T_RCM in CD4⁺ T_EM, CD8⁺ T_EM and CD8⁺ T_EMRA subsets. (C) Proportion of conjunctival T_RMs (CD69⁺CD103⁻, CD69⁺CD103⁺), T_RCM (CD69⁻CD103⁺) subsets within T_EM and T_EMRA pools in healthy controls. The different T_RM populations in each T_EM/T_EMRA pool do not add up to 100% because we did not include CD69⁻CD103⁻ T cells in the figure. Each data point represents a separate individual; mean ± SEM shown.

Figure 3. Principal Component Analysis reveals two T cell signatures in dry eye disease.

(A) 2D Scatter diagram showing first 2 principal components. (B) Loading-plot showing the first principal component is largely determined by the ratio of T_CM to T_EM, a higher ratio will shift the point to the right of the horizontal axis, whereas a lower ratio will be represented on the left.

Figure 4. Two T cell signatures in dry eye disease.

(A) Conjunctival CD4⁺ and CD8⁺ naïve, T_CM, T_EM and T_EMRA subsets in controls (C) and Cluster-1 patients (1) and Cluster-2 patients (2). Each data point represents a separate individual; mean ± SEM shown. (B) Conjunctival T_RM, T_RCM subsets within CD4⁺T_EM and CD8⁺ T_EMRA pools in controls (C) and Cluster-2 patients (2). Each data point represents a separate individual; mean ± SEM shown.

Figure 5. Correlation of two T cell signatures with clinical parameters.

(A) Upper panel: ocular redness; Lower panel: Ocular redness (OR) scores in controls (C) Cluster-1 patients with T cell signature-1 (1) and Cluster-2 patients with T cell signature-2 (2). (B) OR scores show a positive correlation with CD4⁺ T_CM and CD8⁺ T_CM; controls (blue); Cluster-1 patients with T cell signature 1 (red); dotted red line indicates OR 1.9. Pearson’s correlation coefficient was calculated for 55 subjects. Two-tailed P < 0.0001 for both the correlations. (C) Upper panel: Picture of ocular surface obtained with an Oculus Keratograph 5 M; Lower panel: NI-TBUT in controls (C) and Cluster-1 patients with immune signature-1 (1) and Cluster-2 patients (2). (D) NI-TBUT correlates with CD4⁺CD69⁻CD103⁺T_EM-T_RCM and CD8⁺CD69⁺CD103⁺ T_EMRA-T_RM; controls (blue); Cluster-2 patients with T cell signature-2 (red); dotted red line indicates NI-TBUT = 10 s. Pearson’s correlation coefficient was calculated for 49 subjects. Two-tailed P < 0.0001.

Figure 6. Venn Diagram analysis.

(A) Proportion of DED patients (n = 52) with aqueous deficiency dry eye (Schirmer test <5 mm) or evaporative dry eye (NI-TBUT <6 s). (B,C) Proportion of Cluster-1 or Cluster-2 dry eye patients with aqueous deficiency dry eye or evaporative dry eye, respectively. The subject numbers analyzed for this diagram was 52. In all three panels, the numbers outside the circles are patients who did not qualify as aqueous, or evaporative or mixed. Venn diagram analysis was performed using online software at the following URL: http://bioinformatics.psb.ugent.be/webtools/Venn/.

Figure 7. T cell subset patterns in the ocular surface in controls and cluster-1 and cluster -2 patients.

(A) Schematic summarizing distribution of recirculating (naïve, T_CM, T_EM, T_EMRA, T_RCM) and non-recirculating (CD69⁺CD103⁻ T_RM, CD69⁺CD103⁺ T_RM) T-subsets in the healthy human ocular surface. Numbers indicate the proportion of each subset within the CD3⁺ T cell pool. (B) Schematic summarizing distribution of conjunctival T-subsets in Cluster-1 patients. Numbers indicate the proportion of each subset within the CD3⁺ T cell pool. (C) Schematic summarizing distribution of conjunctival T-subsets in Cluster-2 patients. Numbers indicate the proportion of each subset within the CD3⁺ T cell pool.