CD226 opposes TIGIT to disrupt Tregs in melanoma

Tregs and acts primarily in Tregs to dampen antitumor immunity (27).In addition, CD226 + TIGIT -Tregs exhibit reduced demethylation of the Foxp3 gene locus at the Treg-specific demethylated region (TSDR) ex vivo and lower suppressive capacities following in vitro expansion (28).However, the role of CD226 in modulating Treg functions remains unknown.Here, we report that CD25 hi Foxp3 + Tregs in HDs and MPs expressed high-level TIGIT and low-level CD226 as compared with CD4 + effector T cells (Teffs).Tregs in tumors further upregulated TIGIT and downregulated CD226, resulting in a higher TIGIT/CD226 ratio.TIGIT and CD226 exerted opposite effects in Tregs upon PVR binding.TIGIT acted in Tregs to augment their suppression, while CD226 disrupted their suppression and stability.A high TIGIT/CD226 ratio in Tregs associated with higher Treg frequencies in tumors and poor clinical outcome upon immune checkpoint blockade (ICB). ResultsTregs exhibit a high TIGIT/CD226 expression ratio in the periphery and at tumor sites of MPs.We first evaluated TIGIT expression by Tregs and non-Tregs ex vivo in peripheral blood mononuclear cells (PBMCs) of HDs and MPs and MM TILs.CD25 hi Foxp3 + Treg frequencies were higher in MM TILs (mean frequency, 7.4% ± SD 5.9%) than in PBMCs of MPs (1.6% ± 1.2%) and HDs (1.7% ± 0.8%) (Figure 1A and Supplemental Figure 1A; supplemental material available online with this article; https://doi.org/10.1172/jci.insight.121157DS1).CD25 hi Foxp3 + Tregs in PBMCs of HDs and MPs expressed higher TIGIT (HD: mean frequency, 77.1% ± 7.4%; mean MFI, 1,510 ± 500; MP: mean frequency, 83.7% ± 14.1%; mean MFI, 1,903 ± 744, respectively) than CD25 + Foxp3 -(HD: mean frequency, 21.6% ± 5.8%; mean MFI, 313 ± 120; MP: mean frequency, 26.6% ± 12.4%; mean MFI, 410 ± 214, respectively) and CD25 -Foxp3 -(HD: mean frequency, 15% ± 6.1%; mean MFI, 233 ± 115; MP: mean frequency, 18.3% ± 6.5%; mean MFI, 286 ± 150, respectively) CD4 + Teffs as well as total CD4 + T cells (HD: mean frequency, 18.8% ± 5.3%; mean MFI, 293 ± 107; MP: mean frequency, 21.4% ± 6.9%; mean MFI, 383 ± 121, respectively) (Figure 1, A andB).In MM, CD25 hi Foxp3 + Tregs expressed higher TIGIT (mean frequency, 99.1% ± 1.6% and MFI, 6,119 ± 4,481) than CD25 + Foxp3 -(mean frequency, 65% ± 16.6% and MFI, 1,766 ± 1,292), CD25 -Foxp3 -(mean frequency, 35.5% ± 11.2% and MFI, 491 ± 372), and total (mean frequency, 53.2% ± 12.6% and MFI, 1,568 ± 1,447) CD4 + TILs (Figure 1, A andB).Tregs exhibited higher TIGIT expression in MM TILs than in PBMCs of MPs and HDs (Figure 1, A and B).We also observed higher TIGIT expression by Tregs than CD8 + TILs in MM (Supplemental Figure 1B).Because TIGIT competes with CD226 for binding to the same ligands, we next wanted to assess CD226 expression by Tregs of HDs and MPs.Circulating CD25 hi Foxp3 + Tregs in PBMCs of HDs and MPs exhibited lower CD226 expression ex vivo both in terms of frequency (mean ± SD, 71.1% ± 15.2% and 74.1% ± 11.4% for HDs and MPs, respectively) and MFI (1,450 ± 743 and 1,569 ± 406 for HDs and MPs, respectively) than CD25 -Foxp3 -(mean, 89.1% ± 7.7% and MFI, 1,957 ± 990 for HDs; mean, 94.6% ± 2.1% and MFI, 3,268 ± 1,464 for MPs), CD25 + Foxp3 -(mean, 89% ± 8% and MFI, 4,141 ± 2,080 for HDs; mean, 90.8% ± 5.2% and MFI, 4,477 ± 1,573 for MPs) CD4 + Teffs as well as total CD4 + T cells (mean, 90.4% ± 5.4% and MFI 2,669 ± 1,401 for HDs; mean, 92.6% ± 3.4% and MFI, 3,505 ± 1,318 for MPs) (Figure 1, C andD).As a result, the ratio of TIGIT to CD226 expression was higher on CD25 hi Foxp3 + Tregs than on CD25 -or CD25 + Foxp3 -Teffs in PBMCs of HDs and MPs (Figure 1E).In MM, CD25 hi Foxp3 + Tregs exhibited lower CD226 expression (mean, 57.6% ± 14.3% and MFI, 1,007 ± 440) than CD25 -Foxp3 -(mean, 84.7% ± 9.9% and MFI, 2,450 ± 1,067), CD25 + Foxp3 -(mean, 81% ± 15.2% and MFI, 2,276 ± 1,249), and total CD4 + TILs (mean, 80.7% ± 12.2% and MFI, 2,278 ± 828), as well as circulating Tregs in HDs and MPs (Figure 1, C and D), resulting in a higher TIGIT/CD226 ratio at tumor sites (mean ratio of the percentages of TIGIT + / CD226 + cells: 1.7 ± 0.6, 1.2 ± 0.3, and 1.3 ± 0.2, and mean ratio MFI of TIGIT/CD226, 13.9 ± 11.5, 2.7 ± 1.1, and 2.7 ± 1.8, for CD4 + Tregs in MM TILs and PBMCs of MPs and HDs, respectively; Figure 1E).Collectively, as compared with Teffs, Tregs in the periphery of HDs and MPs and at tumor sites displayed higher and lower expression of TIGIT and CD226, respectively, resulting in an increased TIGIT/ CD226 ratio that is predominant in the tumor microenvironment.Tregs fail to upregulate CD226 upon TCR activation.To investigate whether a high TIGIT/CD226 ratio in Tregs is reversed by TCR activation, Tregs were isolated from PBMCs and MM of MPs prior to in vitro stimulation (IVS) and evaluation of CD226 and TIGIT expression by flow cytometry and realtime PCR.In sharp contrast to CD25 -Teffs, circulating and Ti CD25 hi CD127 -Tregs failed to upregulate CD226 upon TCR activation while both Tregs and Teffs upregulated TIGIT (Figure 2, A-C).Upon TCR