Ya Ha, PhD

In our laboratory we are primarily interested in: (i) the biochemical mechanisms of integral and peripheral membrane proteins, and (ii) the design of small-molecule chemical probes that can selectively target them within the complex environment of a eukaryotic cell. Our earlier X-ray crystallographic work has revealed the first atomic-resolution structures for amyloid precursor protein (APP-E2), rhomboid intramembrane protease (GlpG), GxGD intramembrane protease (FlaK), phosphatidylinositol 4-phosphate 5-kinase (PIP5K), and cis-prenyltransferase (NgBR/DHDDS). Furthermore, in collaboration with our colleagues in the chemistry department, we are currently developing lipid kinase inhibitors that can modulate cell metabolism or regulate membrane trafficking. These novel probes not only shed light on the kinase’s function and mechanism, but also provide a platform for further development into new drugs to treat diabetes, cancer, or viral infection.

(1) Phosphatidylinositol phosphate kinase

There are three types of phosphatidylinositol phosphate (PIP) kinases that are homologous in amino acid sequence within their catalytic domains. The type I kinase PIP5K is responsible for the synthesis of the bulk of cellular PI(4,5)P₂ from lipid substrate PI(4)P. The related type II kinase PIP4K and type III kinase PIKfyve differ in substrate binding specificity, recognizing PI(5)P and PI(3)P, respectively. Our recent work revealed two sequence segments within the catalytic domain that contribute to the kinase’s ability to distinguish these structurally similar lipids. Ongoing crystallographic work aims to provide the structural basis for this hypothesis.

High-throughput screening and structure-based design have yielded highly potent and selective inhibitors for the alpha and beta isoforms of type II kinase PIP4K. We currently use these chemical probes to unravel the complex functions of PIP4K, especially in energy metabolism. Recent genetic studies have revealed a unique dependence of p53-null tumor cells on PIP4Kalpha and PIP4Kbeta. Based on this, we are actively investigating the potential of using PIP4K inhibitors to kill tumor cells that harbor loss-of-function mutations in tumor suppressor p53.

(2) cis-prenyltransferase

Eukaryotic cis-prenyltransferase catalyzes the rate-limiting step in the synthesis of dolichol, glycosyl carrier lipids required for protein glycosylation. Different from the ancestral bacterial enzymes, it is composed of two protein subunits, a membrane-bound Nogo-B receptor (NgBR) subunit and a soluble dehydrodolichyl diphosphate synthase (DHDDS) subunit, and generates a range of long-chain lipid products. Mutations in either NgBR or DHDDS subunits have been found to cause human disease. A recent crystal structure of the NgBR/DHDDS heterodimeric complex has shed light on the composition of the enzyme’s active site and suggested how interactions with the membrane regulates the enzyme’s activity and influences isoprene product chain length, which we plan to further test with biochemical and biophysical experiments. Another interesting question under investigation is whether some disease-causing mutations, which usually reduce activity, could be rescued by small-molecule allosteric activators.

(3) Intramembrane protease

Intramembrane proteases are involved in many important pathways responsible for metabolic regulation and cell signaling. The X-ray structures of rhomboid protease and GxGD protease, both solved first in our laboratory, have revealed general architectural principles for these two membrane protein families, enabling us to ask specific questions about their unique biochemical mechanisms. One question concerns how the protease changes conformation during catalysis. Since the active site of the protease is filled with water, it needs to be closed initially to minimize unfavorable contact with lipid. How does transmembrane substrate, whose diffusion is restricted to the membrane plane, gain access to the active site? The crystal structures showed that the proteases have narrow transmembrane domains, suggesting that the lipid bilayer is constricted around the protein. Can this affect the presentation of buried cleavage sites to the protease? Finally, how does the protease achieve specificity? To study these questions, we apply a range of biochemical and biophysical techniques to the two protease systems described above. The knowledge generated from these studies has both conceptual and practical significance because many membrane proteases are potential targets for pharmacological intervention.