TGF-β1, β2 and β3 bind a common receptor to exert vastly diverse effects in cancer, supporting either tumor progression by favoring metastases and inhibiting anti-tumor immunity, or tumor suppression by inhibiting malignant cell proliferation. Global TGF-β inhibition thus bears the risk of undesired tumor-promoting effects. We show that selective blockade of TGF-β1 production by Tregs with antibodies against GARP:TGF-β1 complexes induces regressions of mouse tumors otherwise resistant to anti-PD-1 immunotherapy. Effects of combined GARP:TGF-β1/PD-1 blockade are immune-mediated, do not require FcγR-dependent functions and increase effector functions of anti-tumor CD8+ T cells without augmenting immune cell infiltration or depleting Tregs within tumors. We find GARP-expressing Tregs and evidence that they produce TGF-β1 in one third of human melanoma metastases. Our results suggest that anti-GARP:TGF-β1 mAbs, by selectively blocking a single TGF-β isoform emanating from a restricted cellular source exerting tumor-promoting activity, may overcome resistance to PD-1/PD-L1 blockade in patients with cancer.

Adamts16 encodes a disintegrin-like and metalloproteinase with thrombospondin motifs, 16, a member of a family of multi-domain, zinc-binding proteinases. ADAMTS-16 is implicated in a number of pathological conditions, including hypertension, cancer and osteoarthritis. A large number of observations, including a recent report of human ADAMTS16 variants in cases of 46,XY disorders/differences of sex development (DSD), also implicate this gene in human testis determination. We used CRISPR/Cas9 genome editing to generate a loss-of-function allele in the mouse in order to examine whether ADAMTS-16 functions in mouse testis determination or testicular function. Male mice lacking Adamts16 on the C57BL/6N background undergo normal testis determination in the fetal period. However, adult homozygotes have an average testis weight that is around 10% lower than age-matched controls. Cohorts of mutant males tested at 3-months and 6-months of age were fertile. We conclude that ADAMTS-16 is not required for testis determination or male fertility in mice. We discuss these phenotypic data and their significance for our understanding of ADAMTS-16 function.

Sex determination in mammals is controlled by the dominance of either pro-testis (SRY-SOX9-FGF9) or pro-ovary (RSPO1-WNT4-FOXL2) genetic pathways during early gonad development in XY and XX embryos, respectively. We have previously shown that early, robust expression of mouse Sry is dependent on the nuclear protein GADD45γ. In the absence of GADD45γ, XY gonadal sex reversal occurs, associated with a major reduction of Sry levels at 11.5 dpc. Here, we probe the relationship between Gadd45g and Sry further, using gain- and loss-of-function genetics. First, we show that transgenic Gadd45g over-expression can elevate Sry expression levels at 11.5 dpc in the B6.YPOS model of sex reversal, resulting in phenotypic rescue. We then show that the zygosity of pro-ovarian Rspo1 is critical for the degree of gonadal sex reversal observed in both B6.YPOS and Gadd45g-deficient XY gonads, in contrast to that of Foxl2. Phenotypic rescue of sex reversal is observed in XY gonads lacking both Gadd45g and Rspo1, but this is not associated with rescue of Sry expression levels at 11.5 dpc. Instead, Sox9 levels are rescued by around 12.5 dpc. We conclude that Gadd45g is absolutely required for timely expression of Sry in XY gonads, independently of RSPO1-mediated WNT signalling, and discuss these data in light of our understanding of antagonistic interactions between the pro-testis and pro-ovary pathways.