A potent erythropoietin-mimicking human antibody interacts through a novel binding site.

Recombinant human erythropoietin (rHu-EPO) is used to treat anemia by activating the erythropoietin receptor (EPOR) in erythroid progenitor cells, leading to proliferation and differentiation into mature red blood cells. To allow less frequent dosing, a hyperglycosylated version of EPO has been developed with a longer half-life. In principle, an agonistic antibody targeting EPOR would offer an even longer half-life, support robust monthly dosing, and, unlike EPO products, reduce the risk of pure red cell aplasia. The efficiency of signaling and corresponding potency of previously reported antibody mimics are generally suboptimal compared with EPO and not suitable for clinical use. Here we describe a potent, fully human, agonistic antibody (ABT007) targeting EPOR that supports potent, more sustained, and less pulsatile elevation of hematocrit in a human EPOR-expressing transgenic mouse model compared with standard doses of rHu-EPO while requiring less frequent dosing. Resolution of the crystal structure of the EPOR extracellular domain (ECD) complexed to the ABT007 Fab fragment, determined at 0.32 nm, identifies a binding site that is consistent with a novel mechanism of receptor activation based on a unique antibody-imposed conformational change. These results demonstrate that a symmetric molecule can serve as a potent activator of the EPOR.

Divergence of Genbank and human tumor Bcl-2 sequences and implications for binding affinity to key apoptotic proteins.

Heterodimerization of antiapoptotic and pro-apoptotic Bcl-2 family of proteins provides an important mechanism for apoptosis regulation. Knowledge about key amino acids in the binding groove of native Bcl-2 contributing to this interaction will greatly facilitate the design of Bcl-2-specific inhibitors. There are two different Bcl-2 sequences, M13994 and M14745, in Genbank. Chimeric proteins Bcl-2(1) and Bcl-2(2) derived from the above sequences, although similar in structure, showed different binding affinities to Bak and Bad BH3 peptides (Petros et al., 2001). In this study, we show that the Bcl-2(1) sequence in normal and tumor human tissue samples differs from M13994 and M14745, and contains P59, T96, R110, S117 and G237. The actual sequence in the binding pocket matches the Bcl-2-Ig fusion sequence X06487, originally identified in a t(14:18) translocation of the Bcl-2 gene, associated with follicular lymphoma. The possible effects of the observed amino acid differences compared to M13994 and M14745 were investigated by combining structural data with fluorescence anisotropy. G110R substitution confers on Bcl-2(1) substantially increased binding affinity to Bak, Bad and Bax BH3 peptides, demonstrating that R110 is a key contributor to the BH3 binding affinity of Bcl-2. Although NMR structure did not predict R110 involvement in binding to these BH3 peptides, fluorescence anisotropy data clearly points to a critical role for this residue in binding to pro-apoptotic Bcl-2 family members.

Chkl binds and phosphorylates BAD protein.

Chk1 (checkpoint kinase 1) is a serine-threonine kinase that is critical for G2/M arrest in response to DNA damage. Chk1 phosphorylates Cdc25C at serine-216, a major regulatory site, in response to DNA damage. Furthermore, Chk1 also phosphorylates Cdc25A on serine 123 which accelerates its degradation through the ubiquitin-proteasome pathway and arrests cells in late G2-phase after DNA damage. In the present study, we demonstrated that Chk1 phosphorylates pro-apoptotic protein BAD (Bcl-2/Bcl-XL-Antagonist, causing cell Death) in vitro. In vitro phosphorylation analysis with various mouse BAD peptides has revealed two phosphorylation sites for Chk1 at serine-155 and serine-170. When wild-type and mutant BAD (S155A) constructs were transfected into 293T cells, an association between BAD and Chk1 was observed by co-immunoprecipitation. In addition, there was an increase in the phosphorylation of serine-155 following DNA damage by adriamycin treatment. Our results suggest that Chk1 associates with BAD and phosphorylates the BAD protein at serine-155. Taken together, our results suggest that Chk1 may inactivate BAD by associating with and phosphorylating residues critical for BAD function in response to DNA damage.