Created by: Austin Cheng
Membrane phospholipid synthesis proceeds from the central intermediate phosphatidic acid (PA), which is formed from the two-step sequential acylation of glycerol-3-phosphate (G3P) (1-2). G3P is first acylated by G3P acyltransferase (GPAT) to form lysophosphatidic acid (LPA), which is then acylated by lysophosphatidic acid acyltransferase (LPAAT) to form PA (3-4). In bacteria, GPAT and LPAAT are called PlsB and PlsC, respectively (4). PlsC from the thermophilic bacterium Thermotoga maritima (TmPlsC, PDB ID: 5KYM) serves a vital role in the de novo synthesis of glycerophospholipids that make up the cell membrane (5).
TmPlsC belongs to a family of acylglycerophosphate acyltransferase (AGPAT) proteins, which is responsible for phospholipid synthesis and phospholipid remodeling (4,6). Members of the AGPAT family are similar to GPAT structures: they share four sequence motifs, as well as a HX4D active site motif. AGPAT proteins must simultaneously access substrates in the nonpolar membrane interior and aqueous membrane exterior. To determine the mechanism of this process, scientists determined the crystal structure of a model AGPAT protein, TmPlsC (7). The visualized structure reveals a largely hydrophobic N-terminal two-helix motif that anchors TmPlsC to one leaflet of the membrane, allowing its active site to simultaneously access membrane-bound LPA and the aqueous acyl donor and release the PA product directly into the membrane interior. Resolution of the N-terminal two-helix motif also reveals a dead-end hydrophobic tunnel whose length exactly selects acyl donor substrates of the same length, which explains the protein's acyl donor substrate selectivity. Determination of the structure and mechanism of TmPlsC reveals that other AGPAT proteins likely use similar methods for membrane association and substrate selectivity.
ExPASy is a bioinformatics resource portal which provides access to information such as the molecular weight and isoelectric point of proteins. The molecular weight of TmPlsC is approximately 28.7 kDa, and its isoelectric point is 10.02 (8).
PSI-BLAST (Position-Specific-Iterated Basic Local Alignment Search Tool) is a program used to find protein subjects with a similar primary structure to a protein query (9). Given a protein query, protein subjects are assigned a significant E value, which is calculated by comparing the sequences and assigning gaps. A gap is a segment of amino acids that exists in the subject's sequence, but not the queries. Gaps decrease total sequence homology, which means that the sequences are less similar. Subjects with few gaps are assigned low E values, which indicate high sequence homology. An E value of less than 0.05 is considered significant.
PSI-BLAST was run using TmPlsC as
the query against many protein subjects in the Protein Data Bank (PDB).
However, the most similar subject returned was assigned an E value of 1.6,
which is not significant. PSI-BLAST searches in the PDB database were unable to
find subjects with significant sequence homology to the query.
The Dali server is a tool used to
search for proteins with similar tertiary structure to a protein query (10). In
contrast to traditional methods that attempt to superimpose structures on each
other using a least-squares method, the Dali server uses a sum-of-pairs method
to compare intramolecular distances to produce a measure of similarity.
Proteins in the database are organized by fold so that searches are efficient.
Subject proteins are assigned a Z-score, and subjects with a Z-score above 2
are considered to have significant structural similarity.
TmPlsC was queried on the Dali server to find a protein named Glycerol-3-phosphate acyltransferase (GPAT, PDB ID: 1IUQ) from the squash plant Curcubita moschata (11-12). This protein has a Z-score of 15.6. Sequence alignment between TmPlsC and GPAT gave an E value of 0.11.
TmPlsC is a 247-residue intrinsic membrane protein containing a single subunit made up of two domains (7). Residues 1-61 form a N-terminal antiparallel two-α-helix motif, and the remaining residues 62-247 form an αβ-domain consisting of a seven-stranded β-sheet hydrophobic core surrounded by five α-helices and four 310 helices. Five of the seven β-strands form a parallel β-sheet that is antiparallel to the remaining two β-strands. The overall secondary structure is shown here.
The active site is contained in the αβ-domain and its key residues consist of His-84, Asp-89, Lys-105, and Arg-159. Asn-83 and Gln-85 form hydrogen bonds with the neighboring backbone amide nitrogen atoms of His-84 to fix it in place. Glu-156 also forms hydrogen bonds to His-84 to further hold the catalytic His-84 in place. Asp-89 maintains a lone pair of electrons on the Nε2 nitrogen of His-84, which then deprotonates the 2-hydroxyl of LPA. This prepares an oxyanion for nucleophilic attack of the ACP/CoA substrate to acylate the 2-hydroxyl position and form the PA product. The side-chain nitrogen atoms of Lys-105 and Arg-159 bind the 3'-phosphate of LPA through hydrogen bonding and electrostatic interactions. A segment of basic residues from 128-133 stabilizes the negatively charged acyl-ACP through electrostatic interactions.
Hydrophobic residues Tyr-20, Ile-21, Gly-25, Ile-49, and Phe-52 within the N-terminal two-helix motif form a hydrophobic tunnel that selects the 16:0 acyl substrate. Gly-25 sits at the end of the tunnel; when mutated to a larger residue, it blocks the end of the tunnel, changing the acyl substrate selectivity to 14:0. This confirms that the hydrophobic tunnel is responsible for the acyl substrate selectivity of TmPlsC.
Exposed hydrophobic residues Tyr-44, Trp-59, Phe-61, Trp-116, and Trp-200 on the surface of the N-terminal two-helix motif and nearby αβ-domain help TmPlsC associate with the phospholipid membrane through hydrophobic interactions. Exposed basic residues in the N-terminal two-helix motif help associate with the negatively-charged surface phosphates of the membrane.
There are three ligands listed in the PDB entry for TmPlsC: 1-heptadecanoyl-2-tridecanoyl-3-glycerol-phosphonyl choline, dodecyl-beta-D-maltoside, and nonane, though none of these are its substrates LPA and acyl-ACP or product PA (5, 7). The first and third ligands are molecules from the lipid membrane that crystallized along with TmPlsC, and the second ligand is a detergent used to separate TmPlsC from the membrane. The actual LPA substrate did not crystallize with TmPlsC, and is not shown here.
TmPlsC does not contain any prosthetic groups or associated metal ions (5). There are no alternate conformations for TmPlsC. TmPlsC is not known to be the target of any drugs.
TmPlsC and GPAT do not have significant similarity in primary structure due to an E value of 0.11, which is greater than 0.05 (9). However, they have high similarity in tertiary structure with a Z-score of 15.6, which is greater than 2 (10). The catalytic αβ-domains of TmPlsC and GPAT are nearly identical, with a β-sheet core surrounded by α-helices and 310 helices (11). The structures share a common active site as well. This similarity is not surprising, since both proteins are members of the AGPAT family of proteins, which share four sequence motifs in their αβ-acyltransferase domains, including a conserved HX4D active site motif (6, 12).
The structures differ in their N-terminal domains (11). Instead of an N-terminal two-helix domain, GPAT has four α-helices and one additional β-strand. However, the domains are functionally similar, allowing TmPlsC and GPAT to associate with the membrane. TmPlsC and GPAT share similar functions as acyltransferases along the two-step sequential acylation of G3P to LPA to PA (2-4). GPAT governs the first acylation, while TmPlsC governs the second acylation. The substrates are similar for each structure, which explains their tertiary structure similarity.