Covalent Binding of Lipid to Protein

Abstract

A new structure of a lipid and its covalent linkage to a protein (murein-lipoprotein of the Escherichia coli outer membrane) is described. Glycerylcysteine (S-(propane-2',3'-diol)-3-thio-2-aminopropanic acid) a t the N-terminal end of the polypeptide chain is the attachment site of two ester-bound fatty acids. An additional fatty acid is bound as amide to the N-terminal group.The diglyceride residue on the cysteine could be derived from the phospholipid pathway since the fatty acid composition is very similar to that of the phospholipids from the same cells. I n contrast, 65O/, of the amide-linked fatty acid is palmitate. The main fatty acids are palmitic acid (53O/,), cis-vaccenic acid (20.7 O/,), 9,lO-methylene-hexadecanoic acid (10.6O/,) and palmitoleic acid The cell wall of bacteria with the exception of some very halophilic species [l] contains a so-called rigid layer which is made up of polysaccharide chains cross-linked by short peptide bridges [2,5,50]. It is called murein, peptidoglycan, muco-or glycopeptide. I n Escherichia coli and closely related Enterobacteriaceae [6,7] about lo6 lipoprotein molecules are covalently bound to tbe polysaccharide-peptide macromolecule. The total giant macromolecule is henceforth called murein-lipoprotein complex. The complete determination of the amino acid sequence of the lipoprotein [8,9] showed that the protein molecules are bound on average to every 10 th-12 t h diaminopimelate residue of the murein by the Cterminal lysine and that the N-terminal end is probably blocked by a covalently bound lipid. All amino acid positions except the termini could be determined by conventional sequence analysis which made binding of lipid to amino-acid side-chains a t internal sites of the polypeptide chain unlikely. I n a short communication [lo] we reported that a peptide containing the total lipid present in the lipoprotein could be isolated from a digest of the lipoprotein with proteases. This pointed to a single attachment site of the lipid. About three fatty acid residues were determined per mole of lipoprotein. No glycerol or phosphate could be found which ruled out an ordinary phospholipid. Efforts were therefore directed towards identifying a compound which could link three fatty acid residues to the polypeptide chain. This paper describes the identification of this compound, the structure of the lipid and its attachment site to the protein. MATERIALS AND METHODSGrowth and Labeling of Bacteria E. coli B/r were grown by E. Merck (Darmstadt) in a glucose-salt medium enriched with lo/, yeast extract to the late logarithmic growth phase and kept frozen a t -60 "C until used. E. coli W7 (dap-, Zys-), a derivative of E. coli W173-25 [ll] is a double mutant unable to synthesize and to decarboxylate diamino-

The presence of cysteine in the cytoplasmic domain of the vesicular stomatitis virus glycoprotein is required for palmitate addition.

“…However, a mutant that retained the normal transmembrane domain and had a new cytoplasmic domain of 12 amino acids specified by simian virus 40 sequences was not labeled with palmitate. This result suggested the possibility that the palmitate was added within the first 14 residues of the cytoplasmic domain. These 14 residues do not contain serine, but they do contain a single cysteine residue.…”

“…There was, however, a single cysteine residue located seven amino acids from the transmembrane domain. A cysteine residue is a candidate for a thioester linkage to palmitate, or for a thioether linkage to a more complex structure such as the glyceryl palmitate which has been identified in the E. coli murein lipoprotein (14). Mutagenesis of the Cysteine Codon.…”

Cell Envelope and Shape of Escherichia coli K12. Crosslinking with Dimethyl Imidoesters of the Whole Cell Wall

“…It was not clear whether or not some phospholipid remains in ghosts extracted with Na dodecyl sulfate. A quantitative determination would be made somewhat difficult by the fact that most of the phospholipid components occur also in the lipoprotein (our protein IV) described by Braun et al (10,11). Since ghosts do not lose shape upon complete removal of phospholipid (2), we asked whether ghosts without phospholipid can be crosslinked the same way as normal ghosts.…”

“…Therefore, the 5-10% water found represent minimum values. Considering, in addition, the presence of known (11) and possibly unknown non-amino-acid substituents of the ghost proteins, as well as the fact that in normal ghosts there is no other major constituent besides protein, phospholipid, and lipopolysaccharide (2), it is rather safe to assume that ghosts such as those shown in Fig. 1 (right) are practically pure protein sacs.…”

Existence of several homologous sequences in the Escherichia coli chromosome to the gene for the major outer membrane lipoprotein

“…It has a peculiar amino-terminal structure, a glycerylcysteine residue to which 3 fatty acid residues are covalently linked [3]. It is produced from a secretory precursor, prolipoprotein, with a signal peptide consisting of 20 amino acid residues [4].…”