Circulating CCR6+ILC proportions are lower in multiple sclerosis patients

Abstract

Objectives: The role of innate lymphoid cells (ILC), particularly helper ILC, in the pathogenesis of multiple sclerosis (MS) is not well understood. Here, we present a comprehensive analysis of peripheral ILC subsets in MS patients prior and after alemtuzumab administration using mass cytometry.

Methods: Circulating ILC were analysed by mass cytometry in MS patients before and after alemtuzumab. These were compared with non-MS controls. MS-related shifts among ILC immunophenotypes were further elucidated by fast interpolation-based t-SNE (Flt-SNE) dimensionality reduction.

Results: Neither natural killer (NK) cells nor helper ILC (ILC1, ILC2 and ILC3) levels were altered following alemtuzumab treatment. However, CD56^bright NK cell expansions were observed in relapsing patients. MS patients prior to alemtuzumab further displayed proportional shifts from ILC1 to ILC2, with MS-associated decreases in CCR6⁺ helper ILC proportions.

Conclusion: CD56^bright NK cells during relapse indicate an immediate response to disease reactivation, while CCR6-related shifts among helper ILC suggest altered ILC migration to the CNS during MS.

Keywords: CCR6; ILC; MS; NK cells; alemtuzumab.

Figure 1

Gating strategy for ILC. (a) Biaxial gating for Lin⁻ (CD3⁻CD19⁻CD14⁻CD11c⁻CD123⁻CD34⁻FcεRIα⁻TCRab⁻) live CD45⁺ PBMC (peripheral blood mononuclear cells). (b) Natural killer (NK) cells were defined as Lin⁻CD56⁺CD94⁺ and subdivided into CD56^dim and CD56^bright NK cells. (c) Helper ILC were defined as Lin⁻CD56⁻CD94⁻CD127⁺ and subdivided into innate lymphoid cell (ILC) type 1 (CD294⁻CD117⁻), ILC2 (CD294⁺CD117^+/−) and ILC3 (CD294⁻CD117⁺).

Figure 2

Alterations in CD56^bright and CD56^dim NK cells follow relapses but are not a long‐term effect of alemtuzumab treatment. (a) Peripheral blood mononuclear cell (PBMC) counts (10⁶ cells mL⁻¹) across patients. Cell counts (cells mL⁻¹) and percentages (as proportion of PBMC) of (b) CD56^bright NK cells and (c) CD56^dim NK cells. NK cells were defined as live CD45⁺Lin⁻ (CD3⁻CD19⁻CD14⁻CD11c⁻CD123⁻CD34⁻FcεRIα⁻TCRab⁻) CD56⁺CD94⁺. For comparisons of cell counts and percentages between all five groups (non‐MS (multiple sclerosis) controls (n = 9), untreated MS patients (prior, n = 11) and MS patients post1 (< 12 months after alemtuzumab dose, n = 8/9 for cells mL⁻¹ and n = 9/9 for proportion of PBMC), post2 (> 12 months, n = 10) alemtuzumab and relapse (n = 3)), a PERMANOVA was done followed by pairwise comparisons with Holm's correction. Prior, post2 and relapse groups were compared with non‐MS controls (for three comparisons). A linear mixed‐effects model was calculated when comparing between MS patients before and after treatment. 4999 permutations were then run to calculate P‐values. Five multiple comparisons were made (prior to post1, post2 and relapse, and post1 to post2, and post2 to relapse) using a further 4999 permutations with Holm's correction. Mean is shown in non‐MS controls, P‐values < 0.1 are shown.

Figure 3

Helper ILC and subsets (cells mL⁻¹ and %PBMC) following alemtuzumab. Cell counts (cells mL⁻¹) and percentages (as a proportion of PBMC) of (a) Lin⁻CD56⁻CD94⁻CD127⁺ helper ILC and subsets (b) ILC1 (CD294⁻CD117⁻), (c) ILC2 (CD294⁺CD117^+/−) and (d) ILC3 (CD294⁻CD117⁺). Helper ILC were defined as live CD45⁺Lin⁻CD127⁺ (CD3⁻CD19⁻CD14⁻CD11c⁻CD123⁻CD34⁻FcεRIα⁻TCRab⁻) CD56⁻CD94⁻. For comparisons of cell counts and percentages between all five groups (non‐MS (multiple sclerosis) controls (n = 9), untreated MS patients (prior, n = 11) and MS patients post1 (< 12 months after alemtuzumab dose, n = 8/9 for cells mL⁻¹ and n = 9/9 for proportion of PBMC), post2 (> 12 months, n = 10) alemtuzumab and relapse (n = 3)), a PERMANOVA was done followed by pairwise comparisons with Holm's correction. Prior, post2 and relapse groups were compared with non‐MS controls (for three comparisons). A linear mixed‐effects model was calculated when comparing between MS patients before and after treatment. 4999 permutations were then run to calculate P‐values. Five multiple comparisons were made (prior to post1, post2 and relapse, and post1 to post2, and post2 to relapse) using a further 4999 permutations with Holm's correction. Mean is shown in non‐MS controls, P‐values < 0.1 are shown.

Figure 4

ILC subsets shift within the Lin⁻CD127⁺ helper ILC compartment in MS patients. (a) Levels of ILC1 (CD294⁻CD117⁻) and (b) ILC2 (CD294⁺CD117^+/−) as proportions of helper ILC across timepoints (%Helper ILC). Helper ILC were defined as live CD45⁺Lin⁻CD127⁺ (CD3⁻CD19⁻CD14⁻CD11c⁻CD123⁻CD34⁻FcεRIα⁻TCRab⁻) CD56⁻CD94⁻. For comparisons of cell percentages across timelines (non‐MS (multiple sclerosis) controls (n = 9), untreated MS patients (prior, n = 11) and MS patients post1 (< 12 months after alemtuzumab dose, n = 9), post2 (> 12 months, n = 10) alemtuzumab and relapse (n = 3)), a PERMANOVA was done followed by pairwise comparisons with Holm's correction. Prior, post2 and relapse groups were compared with non‐MS controls (for three comparisons). Five multiple comparisons were made (prior to post1, post2 and relapse, and post1 to post2, and post2 to relapse) using a further 4999 permutations with Holm's correction. 4999 permutations were run to calculate P‐values. Mean is shown across groups, P‐values < 0.1 are shown.

Figure 5

CCR6⁺ ILC1 and ILC2 are reduced in MS patients. (a) FIt‐SNE plots generated on helper ILC (live CD45⁺Lin⁻CD127⁺ (CD3⁻CD19⁻CD14⁻CD11c⁻CD123⁻CD34⁻FcεRIα⁻TCRab⁻CD56⁻CD94⁻)). Dimensionality reduction was done on all Lin⁻CD127⁺ helper ILC (from all patients). Dimensionality reduction plots were calculated using markers in Supplementary table 1. Expression levels of selected markers are shown, with helper ILC subsets ILC1 (CD294⁻CD117⁻), ILC2 (CD294⁺CD117^+/−) and ILC3 (CD294⁻CD117⁺), as annotated. Identification of CCR6^+/− helper ILC. (b) For comparisons of cell percentages between all five groups (non‐MS controls (n = 9), untreated MS patients (prior, n = 10) and MS patients post1 (< 12 months after alemtuzumab dose, n = 8/9), post2 (> 12 months, n = 10) alemtuzumab and relapse (n = 3)), a PERMANOVA was done followed by pairwise comparisons with Holm's correction. Prior, post2 and relapse groups were compared with non‐MS controls (for three comparisons). A linear mixed‐effects model was calculated when comparing between MS patients before and after treatment. 4999 permutations were then run to calculate P‐values. Five multiple comparisons were made (prior to post1, post2 and relapse, and post1 to post2, and post2 to relapse) using a further 4999 permutations with Holm's correction. Mean is shown in non‐MS controls, P‐values < 0.1 are shown.