In their recent publication in The Journal of Immunology, post-doctoral researcher Dr. Douglas Chung from Dr. Pamela Ohashi’s lab revealed that TGF-β1 and IL-15 can induce NK cells to exhibit immunosuppressive properties in vitro.

​​

Previous studies have demonstrated that NK cells can counteract T cell activation and enhance the recruitment of regulatory T cells (Tregs) into tumours. Induced Tregs (iTregs) can arise through the differentiation of conventional effector CD4+ T cells via TGF-β1. Cytokines that promote different CD4+ T cell types can also affect CD8+ cytolytic T cells and innate lymphoid cell (ILC) subsets. As T cells and ILCs require similar cytokines, Chung et al hypothesised that, like iTregs, cytokines like TGF-β1 can give rise to inhibitory ILCs.​ To test this, CD56+ NK cells isolated from healthy peripheral blood (PB) were used to determine if TGF-β1 induced the generation of CD103+ NK-like regulatory cells. They discovered that stimulation with combined TGF-β1 and IL-15 led to the highest expression of CD103 on NK cells compared to TGF-β1 or IL-15 alone. Additionally, these NK cells also exhibited other tissue-resident markers such as CD49a and CD69.

​

Ovarian ascites, which is a fluid that accumulates in the peritoneal cavity of ovarian cancer patients, contains cancer cells, immune cells, and soluble factors that promote tumour growth and immune evasion, including TGF-β. Culturing healthy PB NK cells in 50% ascites supernatant combined with IL-15 demonstrated expression of CD103, CD69 and CD49a at levels similar to NK cells derived from the ovarian ascites itself. Furthermore, CD103 expression was abolished by culturing with anti-TGF-β neutralizing antibodies. Next, they used flow cytometry to compare the phenotype of inhibitory CD103+CD56+ ILCs, found in ovarian tumours, and TGF-β1/IL-15-induced NK-like cells. By day 3 of stimulation with TGF-β1/IL-15, CD103+ NK-like cells began to appear and had expanded by day 7. In addition to CD49a and CD69, the CD103+CD56+ ILCs and CD103+ NK-like cells also expressed CD101, known to be highly expressed on immunoregulatory cells such as Tregs.

​

Finally, the authors evaluated if TGF-β1/IL-15-induced NK-like cells can suppress T cells via co-culture with anti-CD3/anti-CD28 stimulated autologous PB CD4+ T cells. They found that TGF-β1/IL-15-induced CD103+ NK-like cells significantly reduced the absolute numbers of CD4+ T cells in culture. To uncover if the CD103+ NK-like cells were reducing absolute number of T cells through killing, they measured the perforin levels. Indeed, they discovered that soluble perforin levels were substantially higher in the supernatant when autologous CD4+ T cells were co-cultured with TGF-β1/IL-15-induced CD103+ NK-like cells.

​

To summarise, Chung et al highlight how TGF-β may induce an inhibitory phenotype in NK cells in the solid tumour microenvironment. With growing interest in NK cell-based therapies for solid tumors, this study emphasizes the critical need to understand how the tumor microenvironment may hinder these treatments. Future work should investigate methods to evade these alterations and improve antitumor immunity.

​​