June 19, 2014
What if scientists could design a completely new protein that is precision-tuned to bind and inhibit cancer-causing proteins in the body? Collaborating scientists at the UW Institute for Protein Design (IPD) and Molecular Engineering and Sciences Institute (MolES) have made this idea a reality with the designed protein BINDI. BINDI (BHRF1-INhibiting Design acting Intracellularly) is a completely novel protein, based on a new protein scaffold not found in nature, and designed to bind BHRF1, a protein encoded by the Epstein-Barr virus (EBV) which is responsible for disregulating cell growth towards a cancerous state.
EBV is implicated in multiple cancers, including Burkitt’s lymphoma and Hodgkin’s lymphoma. BHRF1 is a homologue of the prosurvival human Bcl-2 proteins and interacts with ‘executioner’ proteins to prevent apoptosis (cell death) and maximize virus production. The activity of Bcl-2 proteins is counteracted by a set of proapoptotic proteins that share a 26-residue Bcl-2 homology 3 (BH3) helical motif that binds a hydrophobic groove on the Bcl-2 protein. Senior fellow Dr. Erik Procko of the IPD and Stayton lab graduate student Geoffrey Berguig in the UW Department of Bioengineering, along with collaborators at the Fred Hutchinson Cancer Research Center, sought to design a protein that could bind the BH3 groove of BHRF1 and inhibit its cancer-causing activity in vivo.
This work is described in a Cell paper entitled A computationally designed inhibitor of an Epstein-Barr viral Bcl-2 protein induces apoptosis in infected cells.
Computational design of de novo proteins
BHRF1-binding proteins were created by grafting side chains from the a BH3 peptide onto a larger and more rigid de novo helical scaffold to allow for greater affinity and specificity of interaction with BHRF1, beyond just the BH3 motif (Figure 1). The designs were solubly expressed and tested by yeast surface display for binding to BHRF1. Candidate designs were further optimized via rounds of error-prone PCR mutagenesis and site-specific saturation mutagenesis followed by fluorescence activated cell sorting (FACS) to obtain binders optimized for affinity, stability and specificity; a binder is desired that targets BHRF1 over other closely related Bcl-2 proteins. One design, BINDI, bound BHRF1 with a Kd of 220 pM (very tight binding) and displayed significantly increased E. coli expression and improved specificity.
The crystal structure of BINDI was shown to be in very close agreement with the computationally designed model (Figure 2). When introduced into EBV-infected cancer cell lines, BINDI effectively induced apoptosis. To test BINDI in an EBV-position B cell lymphoma mouse model, a novel antibody-micelle carrier was used to overcome the challenge of in vivo intracellular delivery of proteins. When treated intravenously with BINDI coupled to the micelle carrier, these mice experienced extended lifespans and slowed tumor progression. This data is the first demonstration that a de novo computationally designed protein can reduce tumor growth and prolong survival in a preclinical model.
At question is whether this technology can be applied beyond cancer treatment to other disease areas. To this end, designer proteins, such as BINDI, that selectively kill target cells provide an advantage over the toxic compounds used in currently developed antibody-drug conjugates. The ability to design functional proteins using de novo scaffolds suggests that is possible to design such proteins to bind any target of interest. Work is ongoing at the IPD and in the Stayton lab to optimize dosing, targeting and delivery of BINDI to increase its therapeutic efficacy.
Additional Press Coverage
UW Health Sciences NewsBeat ran a story entitled Computer-designed protein causes cancer cells’ death. The Molecular Engineering and Sciences Institute posted a story about the multi-lab collaboration that led to this paper entitled MolES research lab collaboration leads to cancer fighting therapy. See also news at Neomatica and comments at Reddit.
Related references include: