How CCTα puts a leash on phospholipid synthesis

The proportion of phosphatidylcholine (PC) in the membrane is controlled by CTP:phosphocholine cytidylyltransferase α (CCTα), which is known to be regulated by a dual auto-inhibitory and membrane-binding domain. However, the detailed mechanism by which this domain regulates CCTα activity is not clear. Ramezanpour et al. use a combined computational and biochemical approach to define new details of this mechanism, providing an elegant illustration of how the lipid-sensing domain of a phospholipid biosynthetic enzyme controls membrane homeostasis.

crystal structure of a catalytically silenced form of CCT␣, which revealed that domain M is composed of a disordered "leash" followed by the AI helix that docks against a linking helix (the ␣E helix) and at the active site to induce a nonproductive conformation (3). However, which if any of these contacts are functionally relevant, and thus the fundamental details of this regulatory mechanism, were not clear. In this issue, Cornell and colleagues utilize molecular dynamics (MD) simulations, mutagenesis, and cross-linking analysis to demonstrate that catalytic silencing by the AI helix involves constraints on catalytic residue Lys 122 and the ␣E helix at the C terminus of the catalytic domain (4). When the domain M leash associates with membranes, the AI helix is also displaced to the membrane surface, and CCT␣ inhibition is relieved.
The potential importance of the ␣E helix and Lys 122 in CCT␣ regulation was initially suggested by comparison with the structurally homologous non-lipid-regulated cytidylyltransferases. In contrast to those enzymes, CCT␣ has a glycine adjacent to Lys 122 , which should increase flexibility around the catalytic residue, and its ␣E helix is longer and predicted to be interrupted in the middle by a disordered flexible hinge. Thus, Ramezanpour et al. (4) suspected that interactions with the AI helix could produce a nonproductive conformation by restricting the movement of the Lys 122 loop and the ␣E helix. To test this, the authors initially show that substitution of alanine or proline for the glycine adjacent to Lys 122 inhibited activation by lipids, indicating that flexibility of the Lys 122 loop is essential. They then ran a series of forty 1-s MD simulations in which the protein was modeled alone or in combination with the CTP substrate and/or AI. Inclusion of the AI helix significantly constrained the N terminus and central hinge of the ␣E helix and changed the hydrogen-bonding frequency of the Lys 122 loop from interacting with CTP to interacting with other carbonyl groups, notably in the C terminus of the AI helix. As a result, the AI helix was able to steer Lys 122 away from a productive complex with its substrate CTP. The presence of Gly 123 was critical to allow close access of these groups, in agreement with the biochemical data.
During MD simulations carried out with the AI helix, the authors noticed that the 4-helix AI-␣E bundle of the CCT␣ dimer was stable, there was minimal backbone fluctuation of the ␣E helix, and its hinge region was constrained. However, when the AI helix was removed from the MD simulations, anticipating the conformational change that would occur when domain M binds membranes, there was a remarkable unwinding of the ␣E hinge into a splayed, bent configuration, and stable contacts formed between the Lys 122 loop and the C terminus of the ␣E helix. To test this experimentally, the authors created a CCT␣ construct containing Cys 217 in the C terminus of the ␣E helix that could be cross-linked to its dimeric counterpart. This cross-linking was reduced in the presence of phospholipid vesicles, indicating that enzyme binding to membranes did increase ␣E helix interchain distance. In a related experiment, constraining the ␣E helices in a disulfide-linked dimer of CCT␣-Cys 127 interfered with enzyme activation even though association with phospholipid vesicles was normal.
The studies from Cornell's group provides a clear picture of the auto-inhibitory state imposed by the AI helix in domain M, and how enzyme activation (or inhibitory relief) is achieved by association of the leash segment of domain M with membranes. Specifically, the two segments of domain M have opposing roles in CCT␣ regulation. On one hand, AI helices prevent productive interactions with the substrate CTP by forming stable bundling with ␣E helices and constraining the Lys 122 loop. On the other hand, the disordered domain M leash causes AI helix removal from the active site, allowing enzyme activation.
The model proposed by Cornell's group also points to important questions for future consideration. First, do lipids directly contact the active site and affect activity? This is suggested based on proximity of the enzyme active site to the membrane due to splaying of the ␣E helices, but it's not clear what specific role lipids might play in these AI/␣E-mediated interactions. Second, CCT␣ is known to associate with organelle membranes and monolayers of different composition and structure (Fig. 1). Does the activation mechanism described here apply at these different organelle surfaces, and what is the biological significance of CCT␣ activation at these different sites? One can envision the domain M leash scanning the surface of different organelle membranes until it identifies negative charge or surface imperfections that allow insertion, but further investigations are needed to know how this leash is being taken on a walk. Enriching membranes in fatty acids, diacylglycerol, or depleting PC results in translocation of CCT␣ to the nuclear envelope and/or export to the ER or, in insects, to lipid droplets (LD) (5). Activation of CCT␣ involves membrane sensing by the "leash" region of domain M (red loops and boxes), which removes the auto-inhibitory segment (yellow boxes) from the catalytic domain (blue ovals). CDP-choline produced in the cytoplasm is a substrate for choline/ethanolamine phosphotransferase (CEPT) to make PC in the ER. It is unclear how or whether CDP-choline generated in the nucleus is trafficked to CEPT (dashed arrow).