TIM-3 does not act as a receptor for galectin-9

T cell immunoglobulin and mucin protein 3 (TIM-3) is a type I cell surface protein that was originally identified as a marker for murine T helper type 1 cells. TIM-3 was found to negatively regulate murine T cell responses and galectin-9 was described as a binding partner that mediates T cell inhibitory effects of TIM-3. Moreover, it was reported that like PD-1 the classical exhaustion marker, TIM-3 is up-regulated in exhausted murine and human T cells and TIM-3 blockade was described to restore the function of these T cells. Here we show that the activation of human T cells is not affected by the presence of galectin-9 or antibodies to TIM-3. Furthermore, extensive studies on the interaction of galectin-9 with human and murine TIM-3 did not yield evidence for specific binding between these molecules. Moreover, profound differences were observed when analysing the expression of TIM-3 and PD-1 on T cells of HIV-1-infected individuals: TIM-3 was expressed on fewer cells and also at much lower levels. Furthermore, whereas PD-1 was preferentially expressed on CD45RA−CD8 T cells, the majority of TIM-3-expressing CD8 T cells were CD45RA+. Importantly, we found that TIM-3 antibodies were ineffective in increasing anti-HIV-1 T cell responses in vitro, whereas PD-L antibodies potently reverted the dysfunctional state of exhausted CD8 T cells. Taken together, our results are not in support of an interaction between TIM-3 and galectin-9 and yield no evidence for a functional role of TIM-3 in human T cell activation. Moreover, our data indicate that PD-1, but not TIM-3, is a promising target to ameliorate T cell exhaustion.
Author Summary Inhibitory costimulatory receptors are a hallmark of exhausted T cells, which accumulate during chronic infection with viruses like HIV-1. Recently, TIM-3 was described as functional receptor on exhausted human T cells. Galectin-9 was reported as an inhibitory ligand for TIM-3 on murine T cells, but it was not known whether galectin-9 has a role in human T cell activation processes. We have found that the activation of human T cells is not affected by the presence of galectin-9 or antibodies to TIM-3. Furthermore, we demonstrate that galectin-9 does not serve as a ligand of human or murine TIM-3. Analysis of T cells of HIV-infected individuals regarding the expression of TIM-3 and PD-1 demonstrates that TIM-3 is expressed on fewer cells and also at much lower levels. In fact, TIM-3 expression characterizes a T cell population that is distinct from the PD-1 expressing exhausted T cells. Our results indicate that PD-1, but not TIM-3, is a promising target to ameliorate T cell exhaustion.