How the enzyme diacylglycerol kinase can form membrane anchors and an active site from so few amino-acid residues has long been a mystery. Crystal structures reveal that it gets by with a little help from its friends. See Letter p.521
The bacterial enzyme diacylglycerol kinase has always been a bit of a rebel. As an apparent evolutionary 'orphan', this integral membrane protein is functionally and structurally distinct from other kinases identified so far. Furthermore, its size (just 121 amino-acid residues) makes its modus operandi challenging to understand: how does so small an enzyme manage to span the cell membrane while still forming a functional active site that must accommodate not only a bulky, fatty substrate but also a water-soluble ATP molecule? On page 521 of this issue, Li et al.1 report crystal structures of diacylglycerol kinase from the bacterium Escherichia coli. The structures reveal how the enzyme exploits the milieu in which it resides, greatly increasing our understanding of this protein and providing insight into integral membrane enzymes in generalFootnote 1.
Diacylglycerol kinase (DgkA) catalyses a key step in the synthesis of oligosaccharide biomolecules2, converting diacylglycerol to phosphatidic acid by transferring a phosphate group from ATP. Most ATP-dependent kinases have the same general structure: an evolutionarily conserved, two-lobed core that has all the elements required for catalysis, including catalytic residues that are absolutely conserved; characteristic ATP-binding motifs such as P-loops and Walker motifs; and a hydrophobic spine connecting the lobes3. DgkA, however, lacks these features. As one of the first integral membrane enzymes to be studied, it has served as an intriguing model not only of catalysis in such enzymes, but also of the structure, stability, assembly and folding of membrane proteins4.
DgkA is the smallest kinase known, weighing in at about half the size of a typical kinase. It therefore has to use most of its residues to build helices that integrate it into the cell membrane, while still crafting an active site that accommodates both ATP and diacylglycerol — which contains a hydrophilic 'head' group and bulky hydrophobic 'tails'. Rather than building a complicated and energetically costly single protein molecule in the membrane, evolution has forged a trimer from three DgkA molecules. On the basis of their crystal structures of the monomers, Li and colleagues propose that each monomer 'borrows' a component from its neighbour to create a composite active site. The borrowed component is an amphiphilic (both hydrophobic and hydrophilic) amino-terminal helix. This helix is strategically located at the interface between the membrane and the cytoplasm, and provides a crucial catalytic residue.
Confirmation of the putative active site will require further experimental support from an ATP-bound structure of DgkA, or, even better, from a structure of the enzyme in complex with both ATP and a substrate, to verify the locations of the ATP- and substrate-binding pockets and other fine structural features. Nevertheless, the proposed active site is fully consistent with earlier work2 in which residues that are essential for catalysis were identified. The suggested active site also makes intuitive sense, because water-soluble ATP should bind to the cytoplasmic portion of DgkA with its transferable phosphate group (the γ-phosphate) oriented towards the membrane–cytoplasm interface, in the optimal position for attachment to the substrate (Fig. 1).
Although appropriate ATP binding is necessary for a functional active site, the substrate also needs to be stabilized and correctly oriented to allow efficient phosphate transfer from ATP. The complex visualized in the crystal structure described by Li and co-workers contains detergent molecules, which the authors used to aid crystallization of the protein and which happen to resemble the substrate. Serendipitously, two of these molecules are found in the structure at a location that primes them for phosphate transfer: the tails are embedded in and stabilized by the membrane, whereas the head groups reach into cytoplasmic space, coming close to the proposed location of ATP's γ-phosphate.
So, the membrane is not only the milieu in which DgkA resides — it also forms an integral part of the active site, dictating the orientation of the substrate to allow efficient phosphate transfer. It would be difficult for DgkA to stabilize and correctly orient the bulky, amphiphilic substrate without the membrane because, as a small kinase, it does not have the resources to create an adequate binding pocket. This suggests that DgkA has evolved together with the cell membrane5 to create an optimal, composite active site.
Interestingly, Li and colleagues' crystal structure is at odds with a previous 'backbone-only' structure6 of DgkA that was obtained using nuclear magnetic resonance (NMR). Although the overall homotrimeric architecture of DgkA is consistent between the two structures, the active sites differ hugely — an unprecedented observation for NMR and crystallographic structures. The dissimilarity is caused by domain swapping: in the NMR structure, one of the three transmembrane helices from a DgkA molecule is switched with that of another molecule.
The different structures might reflect the different conditions used in the two analyses, or the challenges involved in obtaining the NMR structure. However, given the extreme thermal stability of the membrane-embedded DgkA homotrimer (it can survive at temperatures as high as 100 °C)4, it is difficult to imagine a seamless switch between 'swapped' and 'unswapped' homotrimers. In the light of the high-resolution crystal structure described by Li and co-workers, the NMR results might need to be revisited.
The kinases are one of the largest and most extensively studied protein families, but DgkA shows that there are still surprises to be found. Its remarkable structure and composite active site add an intriguing twist to the kinome — the set of known kinases in an organism. The enzyme's ingenious design results in an active site that is much simpler than those of other kinases. Indeed, DgkA provides a new perspective for understanding the architecture and associated catalytic mechanisms of membrane enzymes. As well as structural determination of ATP-bound and substrate-bound DgkA complexes, future work should include detailed investigation of the enzyme's catalytic mechanism and regulation, and resolution of the discrepancy between the NMR and crystal structures.
*This article and the paper under discussion1 were published online on 15 May 2013.
Li, D. et al. Nature 497, 521–524 (2013).
Van Horn, W. D. & Sanders, C. R. Annu. Rev. Biophys. 41, 81–101 (2012).
Taylor, S. S., Keshwani, M. M., Steichen, J. M. & Kornev, A. P. Phil. Trans. R. Soc. B 367, 2517–2528 (2012).
Pieringer, R. A. & Kunnes, R. S. J. Biol. Chem. 240, 2833–2838 (1965).
Badola, P. & Sanders, C. R. J. Biol. Chem. 272, 24176–24182 (1997).
Van Horn, W. D. et al. Science 324, 1726–1729 (2009).
About this article
Phosphoryl transfer reaction catalyzed by membrane diacylglycerol kinase: a theoretical mechanism study
Physical Chemistry Chemical Physics (2015)
Resonance assignments of a membrane protein in phospholipid bilayers by combining multiple strategies of oriented sample solid-state NMR
Journal of Biomolecular NMR (2014)
Journal of Magnetic Resonance (2014)
Biophysical Journal (2014)
Quarterly Reviews of Biophysics (2014)