Mechanism of silencing the catalytic domain by the regulatory membrane lipid binding domain of an amphitropic cytidylyltransferase

CTP: phosphocholine cytidylyltransferase (CT) is an auto-regulated, homodimeric enzyme that catalyzes the rate-limiting step in the synthesis of phosphatidylcholine (PC). CT is activated by binding to PC-depleted membranes using a lipid-induced amphipathic helix (domain M). This same M domain functions to silence catalysis when the enzyme is not membrane-engaged. While the structure of CT’s catalytic domain has recently been solved (PDB, 3HL4), if and how domain M makes contact with it to silence catalysis remains a mystery. To identify contact sites between domain M and other CT domains, I constructed single cysteine substitutions along domain M and conjugated each site to a sulfhydryl-reactive, biotinylated, benzophenone, BBP. After photo-cross-linking and trypsin digestion, the cross-linked peptides were affinity-purified and identified by mass spectrometry. Each domain M-conjugated site forged cross-links to the same set of catalytic domain peptides, which flank one side of the active site, and these contacts were broken upon membrane insertion of domain M. I then targeted the most conserved region in domain M for mutagenesis. Mutation of F289 and F293 to aspartates obliterated cross-links between the catalytic domain and the BBP-labeled site at neighboring Cys-301. However, loss of this contact did not relieve inhibition of catalysis, suggesting that this site is but one of several cooperating inhibitory segments in the M domain. Evidence for a partially disordered domain M which acquires order upon contact with the catalytic domain was obtained by fluorescence anisotropy, monitoring Oregon Green conjugated to sites in three distinct CT domains. All domain M sites had similar anisotropy values intermediate between the rigid catalytic domain and the very flexible C-terminal tail. Urea denaturation analysis suggested a weakly folded structure for domain M. Breaking a contact between M and C by the F → D mutation dramatically lowered the anisotropy of that sub-region and its resistance to unfolding by urea. These results support an emerging model in which silencing by domain M is achieved through multiple transient, alternating, contacts with the catalytic domain. Contacts with the catalytic domain may introduce ordering into an otherwise flexible domain M. These inhibitory contacts are replaced with protein-lipid interaction upon activation.