Activin A is a member of the TGF-β superfamily, a family of structurally related growth factors that have been implicated in embryogenesis, homeostasis, and cancer. All members consist of pro- and mature domains which are post translationally cleaved. Upon signalling the pro-domain dissociates from the mature growth factor. Signalling in this family requires a type I and a type II receptor. Though there are over thirty growth factors, there are only seven type I and five type II receptors resulting in promiscuity between ligands and receptors. In order to fully elucidate the roles of these proteins and develop therapeutics, specific inhibitors must be generated.

In this thesis, I targeted activin A to achieve specificity. I first targeted mature activin A through the use of XChem, conducting a fragment screen against mature activin A crystals. I describe the identification of twelve fragment hits that bind in the putative type I receptor binding site and the validation and optimisation of one of these fragments.

I then used the structure of the pro-mature complex of activin A to design and generate a series of GB1- fused peptides based on the α1-helix-loop-α2-helix motif of the pro-domain. I determined the α1-helix-loop epitope to be key for interactions with the mature domain, before investigating the effect of truncation and mutation on this epitope. I then determined the potency of the fusion GB1-H1LH2 as an inhibitor of activin A signalling (IC50 = 4.4 μM). I further optimised GB1-H1LH2 through dimerization resulting in a large increase in inhibitor potency (IC50 = 71.5 nM). Lastly, I screened this dimeric inhibitor against several TGF-β family members to determine it specifically inhibits both activin A and activin B.

To conclude, I discuss the findings of this thesis before considering how this methodology could be applied to other TGF-β family members.