Our research program has focused on identifying critical cellular interactions during the unfolded response during protein overexpression in the yeast Saccharomyces cerevisiae. Our laboratory seeks to use a systematic approach that goes beyond the empirical observations and optimizations common to more traditional studies, in order to establish a fundamental understanding of the key interactions in these systems.

We are studying overexpression of single-chain antibody in S. cerevisiae. Since the advent of hybridoma technology, enabling researchers to make antibodies in a deliberate and defined way, there has been a great deal of interest in the use of antibodies in therapeutic and diagnostic applications. By 2008, engineered antibodies are expected to account for >30% of all revenues in the biotechnology market, yet antibodies remain difficult to produce in recombinant systems.

Studies from our laboratory show that the single-chain antibody (scFv) expressed in yeast is retained in the endoplasmic reticulum due to slow folding, and the scFv is bound by two cellular proteins, yeast BiP (Kar2) and PDI. The accumulation of unfolded scFv protein activates the unfolded protein response and subsequently shuts down expression. Using the known interactions with cellular proteins, we developed a mechanistic model that predicted that BiP overexpression would lower cell stress and improve scFv secretion. Although antibody expression was higher when BiP levels were increased experimentally, the stress response was only marginally ameliorated. Using chemical stressors, we also found that BiP overexpression had very modest effects on cell stress. These data suggest that BiP has only a minor role in the stress response pathway activation, in contrast with traditional models of BiP-regulated dimerization of the Ire1p receptor.

Recently, we have investigated the effect of BiP interactions on scFv folding and secretion by making specific changes to the scFv itself via site-directed mutagenesis, with the goal of identifying sites in the protein that would alter the interactions with BiP. Sites were identified either via a rational approach based on a BiP binding algorithm, or using a combinatorial approach combined with a surface display screen for enhanced scFv secretion. We determined that in vivo BiP binding had little effect on protein secretion levels, as determined by measuring levels of interaction by co-immune precipitation and comparing the results of secretion and ER stress. These results are consistent with a new model developed in collaboration with the Doyle group (UCSB) for endoplasmic reticulum interactions between BiP, Ire1p, and unfolded scFv proteins, in which Ire1p is regulated directly by unfolded proteins, and are also supported by recent molecular studies of the Ire1p receptor from the Walter and Kohno/Kimata labs.

Extended Abstract Status: Not Uploaded