Targeting TGF-beta activation in tumors - Targeting TGF-b activation in tumors
Abstract
Transforming growth factor-ß (TGF-ß) drives immune dysfunction in the tumor microenvironment by inducing
regulatory T cells (Tregs) and inhibiting cytolytic CD8+ T cells and helper Th1 cells. TGF-b is ubiquitously
expressed in mammals as isoforms TGF-b1, -b2, and -b3, but is maintained in an inactive form by non-covalent
interaction with its propeptide, the latency associated domain of TGF-b (LAP). The integrin avb8 binds to the
LAP of TGF-b1 and TGF-b3 and mediates their activation. Germline or conditional genetic deletion studies have
revealed that integrin avb8-mediated activation of TGF-b is essential for the in vivo activation of TGF-b, and thus
avb8 acts as a key modulator of TGF-b function. In general, integrins are adhesion molecules and mediate the
attachment of cells to extracellular matrix proteins. Integrin avb8 recognizes an Arg-Gly-Asp (RGD) motif and
interacts with fibronectin, vitronectin, and latent TGF-ß isoforms, although it binds considerably more strongly to
latent TGF-ß than to other RGD-containing proteins (Ozawa, 2016). Despite the clear association of TGF-b
signaling and T cell function, few therapies that target TGF-b have been successful, largely due to pan-inhibitor
toxicity. To address this therapeutic challenge, we have identified a mouse monoclonal antibody (AMHA-11) that
selectively blocks the interaction of the human integrin avb8 with its ligand, latent transforming growth factor-b
(TGF-b). The AMHA-11 antibody is unique in that it selectively perturbs the avb8-mediated activation of TGF-b
isoforms 1 and 3 and does not inhibit TGF-b2, which lacks an integrin-binding RGD motif. Additionally, because
of redundant activities of other av integrins, cell adhesion is not perturbed by AMHA-11. This affords a higher
degree of selectivity in perturbing only integrin avb8-mediated activation of TGF-b activation and not the residual
cell adhesion properties, which may be undesirable to inhibit. In addition, global inactivation of TGF-b is likely to
have undesirable side effects since TGF-b is an essential homeostatic epithelial and immune effector. In Phase
1 we will identify high-affinity humAbs that inhibit avb8-mediated activation of TGF-b, then rank-order them in
vitro for their ability to inhibit TGF-b activation and to modulate T cell activity. We will also evaluate them in a
small animal model as monotherapy and in combination with anti-PD-1. We are confident that a novel therapy
will result from this more selective approach.
