A major obstacle to antitumor immunity in the tug-of-war between the immune system and cancer is T cell exhaustion, the hierarchical loss of T cell effector activities associated with the co-expression of T cell inhibitory receptors, especially PD-1 and TIM-3, in CD8 T cells. This is a result of the differentiation of PD-1^loTIM-3^– precursor exhausted T cells into PD-1^hiTIM-3⁺ terminally exhausted T cells. Clinical inhibition of PD-1 has achieved unprecedented clinical success; however, the majority of cancer patients still do not benefit.

Current research posits a canonical PD-1 signaling pathway in which ligand engagement triggers tyrosine phosphorylation of the ITIM and ITSM motifs in the cytoplasmic domain of PD-1, which then recruit and activate the SH2 domain-containing protein tyrosine phosphatase SHP2 to dephosphorylate key proteins in TCR and co-stimulation signaling, thereby dampening T cell activity. Whether and how this pathway interacts with other key regulators of T cell function is less clear. We set out to answer this question by identifying proteins that bind PD-1, through a combination of co-immunoprecipitation and mass spectrometry.

It caught my eye immediately when the mass spectrometry result came back, with galectin-9 (Gal-9) sitting near the top of the candidate list. From my previous experience with galectin structure and function and the knowledge of Gal-9 as a TIM-3 ligand that induces T cell death, it did not take long for me to picture a model in which Gal-9 uses its two carbohydrate-recognition domains (CRDs) to crosslink PD-1 and TIM-3. Such a configuration of physical interaction may "distract" Gal-9 from TIM-3 and attenuate Gal-9/TIM-3-induced T cell death, and explain why PD-1⁺TIM-3⁺ T cells can survive, and in many cases even dominate the CD8 T cell population, in the TME, where Gal-9 is often abundant.

The project began to take shape after stimulating discussions with Dr. Hung and Linlin, co-first author of this paper. As expected, subsequent protein binding and functional assays with purified proteins and cells provided strong evidence that PD-1 competes with TIM-3 for the C-CRD of Gal-9 to form (TIM-3/Gal-9/PD-1)_n lattices at the expense of (TIM-3/Gal-9/TIM-3)_n lattices, and thereby reduces Gal-9/TIM-3-induced death of PD-1⁺TIM-3⁺ T cells. Persistence of these PD-1⁺TIM-3⁺ T cells is important for tumor control; at the early stages of differentiation, these "transitory" T cells retain at least some effector functions and respond to immunotherapy to exert antitumor activity.

We postulated that rescuing these TIM-3⁺ T cells from Gal-9-induced death by inhibiting Gal-9 could promote antitumor T cell response. Surprisingly, when we treated mice bearing syngeneic tumors with an anti-Gal-9 antibody, we did not observe durable tumor regression. Could this be due to impaired T cell co-stimulation after Gal-9 inhibition, as Gal-9 has been shown to be important for the signaling of 4-1BB, a T cell co-stimulatory receptor of the tumor necrosis factor receptor superfamily (TNFRSF)? We tested this possibility by combining anti-Gal-9 with an agonistic antibody to GITR, another member of the TNFRSF family of co-stimulatory receptors, to promote T cell co-stimulation. As expected, such combination synergistically suppressed tumor growth and prolonged overall survival.

Single cell analysis of tumor-infiltrating leukocytes from these mice by CyTOF, however, suggests regulatory T (T_reg) cell count as a key factor in such synergy. In turned out that anti-Gal-9 not only expanded a subset of CD8 T cells displaying phenotypes of transitory CD8 T cells, it did so to T_reg cells as well. It is conceivable that these T_reg cells counteract CD8 T cells to reduce the efficacy of anti-Gal-9 therapy. Surprisingly, combination of anti-Gal-9 with GITR agonism led to a near complete depletion of T_reg cells and a further expansion of the TIM-3⁺ subset of CD8 T cells, resulting in a high CD8 T cell/T_reg cell ratio that likely explains the synergistic effects on tumor regression.

As an immune checkpoint that induces T cell death, Gal-9/TIM-3, like other immune checkpoints, needs to be tightly controlled to make sure that the immune response is commensurate with the level of the immunological insult to the organism. We found that similar to the PD-1 ligand PD-L1, Gal-9 expression is upregulated by interferon β and γ, cytokines that are produced in growing tumors, suggesting that tumors can use similar cytokine cues to hijack multiple immune checkpoints to aid immune escape. Indeed Gal-9 is overexpressed in many cancers, and its expression correlates with poor prognosis. In addition, tumors refractory to anti-PD-1/anti-CTLA4 treatment express high levels of Gal-9.

In summary, our work suggests that Gal-9-induced death contributes to reduced persistence of T cells in the TME, and this is regulated by Gal-9 interactions with PD-1 and TIM-3. Inhibition of Gal-9-induced T cell death could be a viable strategy when combined with other therapeutic modalities, especially those that diminish T_regcells. Limited T cell persistence is a major challenge for current cancer immunotherapy, including immune checkpoint blockade and adoptive T cell/CAR-T cell transfer.  We anticipate that our findings will inform next generations of cancer immunotherapy to target a broad spectrum of malignancies, including those with primary or adaptive resistance to current immunotherapies.

Revising the manuscript with COVID-19 running rampant was a race against both the virus and time. We worked in shifts in the lab to reduce the chance of infection, meaning that I had to go to work early in order to finish before 1 pm those experiments that normally take a full day and leave work before 1:30 pm. There is no complaint about it though. For me getting up early is now a habit and I know how to better manage my time to work more efficiently. Being able to fully address the reviewer's concerns with my co-authors before the deadline under such special circumstances is really fulfilling. A publication is a significant mark in your life if years later you still want to come back to read it with a sense of pride. It is a snapshot of a bittersweet journey in science.