Building Life Block by Block

Abstract

This thesis uses X-ray crystallography and cryo-EM to show how structure dictates specificity in two unrelated systems: the lipid signalling enzyme Autotaxin and the transcriptional coregulator Zincore, from a single enzyme pocket to a multisubunit complex controlling gene expression. Part A: Autotaxin and lipid signalling Chapter A1 introduces Autotaxin (ATX), which produces the signalling lipid LPA from LPC in the bloodstream. Its tripartite active site, made of an orthosteric pocket, a hydrophobic pocket and an allosteric tunnel, underlies the classification of ATX inhibitors into five types. Chapter A2 reviews the seven ENPP family members and explains why only ATX (ENPP2) is secreted and has lysophospholipase D activity. Chapter A3 uses cell assays with ATX mutants and inhibitors, together with an in vivo model of pulmonary fibrosis, to show that the allosteric tunnel drives receptor selective signalling, favouring P2Y type over EDG type LPA receptors independently of catalysis, which informs which inhibitors may work best clinically. Chapter A4 designs Type VI inhibitors, built by extending the Type IV inhibitor ziritaxestat to occupy all three regions of the tripartite site, and characterises them structurally, in cell assays and in rat plasma. Chapter A5 discusses Part A, arguing that ATX acts as both an enzyme and a lipid chaperone, which should guide future ATX and LPA receptor targeted therapies. Part B: Structural basis of Zincore mediated regulation of zinc finger transcription factors Chapter B1 introduces C2H2 zinc finger proteins (ZNFs), the largest transcription factor family in humans. Many ZNFs activate transcription without a known activation domain or cofactor, unlike KRAB repressors. Chapter B2 shows that the linkers between zinc fingers actively encode domain orientation and carry phosphorylation sites that can regulate DNA binding. Chapter B3 identifies Zincore, formed by SEPHS1 and QRICH1, which recognises the conserved C2H2 fold through an arginine clamp on SEPHS1 rather than a DNA sequence, so it can act across many ZNFs by locking them onto DNA to enable activation. Chapter B4 shows that the SEPHS1 homodimer lets one Zincore complex engage two DNA bound ZNFs, suggesting a role in chromatin looping independent of CTCF and cohesin. Chapter B5 uses cryo-EM to resolve nine Zincore assembly states: QRICH1 recruitment depends on SEPHS1 already engaging a ZNF and DNA, and the SEPHS1 dimer shifts between open and closed forms linked to QRICH1 occupancy and ATP binding. Chapter B6 documents the cryo-EM methods used in Chapters B3 to B5. Chapter B7 discusses Part B, framing Zincore as a regulator of ZNF occupancy on DNA rather than ZNF activity, and asks whether ZMYM family proteins can replace QRICH1, which would turn Zincore from an activator into a repressor.