The Strange World of LPXTGase

Courtesy of Vincent A. Fischetti  ENIGMATIC ENZYME: Computer-generated ribbon model of the C-terminal end of the M protein sequence containing the conserved LPXTG motif (red). This region is also found in all C-terminal-anchored surface proteins from gram-positive bacteria. Imagine an enzyme assembled from multiple peptides, each the product of a different gene. Imagine that of the 60 amino acids in the sequence, only 40 are identifiable in the standard repertoire; the others are novel i

Mar 24, 2003
Nicole Johnston
Courtesy of Vincent A. Fischetti
 ENIGMATIC ENZYME: Computer-generated ribbon model of the C-terminal end of the M protein sequence containing the conserved LPXTG motif (red). This region is also found in all C-terminal-anchored surface proteins from gram-positive bacteria.

Imagine an enzyme assembled from multiple peptides, each the product of a different gene. Imagine that of the 60 amino acids in the sequence, only 40 are identifiable in the standard repertoire; the others are novel in structure and are likely not assembled on the ribosome but by vast, cumbersome megaprotein complexes. Imagine that the completed enzyme is essential to the life of the bacterium and that an identified inhibitor may pave the way for a powerful new class of broad-spectrum antibiotics. Welcome to the strange world of LPXTGase.

Most surface proteins of Gram-positive bacteria possess a terminal LPXTG sequence that gets cleaved as the proteins journey across the cell septum for attachment to the cell surface. The LPXTG sequence, found in most cell surface proteins, suggests a common anchoring mechanism. To identify enzymes involved in the process, Rockefeller University team members Vincent Fischetti, Sung Lee, and Vijaykumar Pancholi screened streptococcal cell extracts to identify an enzyme that could cleave the LPXTG motif. After detecting cleavage activity, they discovered the enzyme in extracts from both streptococci and staphylococci, and named it LPXTGase. The researchers believe that it is the first enzyme found that is made by nonribosomal peptide synthesis (see 5-Prime), and that it could be the product of several genes.1 Says Biochemist Hieronim Jakubowski, University of Medicine and Dentistry of New Jersey: "An enzyme of such unusual nature doesn't have any precedence in biology."

Let's not jump to conclusions, say some researchers. "Genes encoding peptides [derived by nonribosomal peptide synthesis] are highly conserved and have core sequences that can be used to generate primers to amplify parts of such genes," explains biochemist Mohamed Marahiel, Philipps University, Marburg, Germany. "Until we have this information, one should be careful about [assuming] the origin of such unusual peptides/proteins."

Nonribosomal peptide synthesis, or NRPS (and similarly related polyketide synthesis) occurs mainly in actinomycetes, bacilli and filamentous fungi. NRPS synthetases consist of a series of active sites called modules. Each module carries out catalysis and modification of the tethered growing peptide chain.2,3 The small (2-20) amino acid peptides produced include: the antibiotics penicillin, erythromycin, and vancomycin, the immunosuppressant cyclosporin, and anticancer agents,2,3 all of which have unusual structures. They are assembled in the cytoplasm by large megaproteins called nonribosomal peptide synthetases.

HALLMARKS OF NRPS Is LPXTGase produced by NRPS? If so, it would be a radical increase in complexity. First described in the 1960s, NRPS has been known to produce only short peptides. The LPXTGase is heavily glycosylated, which is required for its activity, and has a highly unusual amino acid sequence with no aromatic amino acids. The enzyme is rich in alanine and 30% of its amino acids are uncommon, meaning that they fall outside the usual realm of amino acids that make peptides and enzymes. The presence of such amino acids are hallmarks of NRPS, suggesting that this is the first enzyme discovered that is not produced on the ribosome, and it may be common among Gram-positive bacteria.

"This is an important enzyme for the cell that is not synthesized by the ribosome," explains Fischetti. "[But] this is probably not a typical NRPS system. Molecules [produced by known NRPS systems] are a maximum of 20 amino acids. This enzyme is about 60 amino acids, so [these particular bacteria] are doing it in a new way. Before this paper, only short-peptide antibiotics have been identified that are produced by NRPS."

Says Jakubowski, "An enzyme of such a limited set of [regular] amino acids--they found just 11--and some unusual amino acids is kind of mind-boggling. If we take their interpretation of unusual amino acids, then [NRPS] is likely, because there's no way for unusual amino acids to get into a peptide by the ribosomal route."

Microbiologist Olaf Schneewind, University of Chicago, agrees that the group's findings are significant. "I think [they're] right in claiming that this is the first example that NRPS could lead to an enzyme," he says. "Finding mutations to abrogate synthesis of this enzyme will provide an avenue for studying its synthesis."

But not everyone agrees. "The work presented is interesting in that the anchor endopeptidase seems to represent a unique protein with so many unusual modifications and 'unnatural' amino acid residues, so that one would think that it could be made nonribosomally," says Marahiel. "All nonribosomal peptides known to date are in the range of 2 to 48 residues, and no such peptide was shown to have catalytic activity and the size shown for the LPXTG endopeptidase."

At the 3rd European Conference on Marine Natural Products held last fall in Munich, researchers introduced a peptide called polytheonamide and said that NRPS produced it, Marahiel continues. The peptide is highly modified with unusual amino acid residues that are methylated and N-terminal blocked, as well as D-isomers. However, the encoding gene for this putative NRPS peptide is still unknown.

Furthermore, Marahiel says, there are several examples of ribosomally synthesized peptides that are strongly modified posttranslationally. Examples include the lantibiotics such as epidermin, nisin, and cinnamycin, which can reach 47 residues. Marahiel further raises the possibility that such peculiar products may even be the outcome of some creative engineering on the part of some organisms. "One may think about a possible combination of ribosomal and nonribosomal machines to generate such diversity of structure and function," he suggests.

Biochemist Hans von Doehren, Institute for Biochemistry and Molecular Biology, Technical University, Berlin, also isn't entirely convinced that the endopeptidase is the product of NRPS. He believes that further studies using sophisticated mass spectrometry methods may provide more conclusive evidence. "To make a sequence-defined peptide of 50 or more residues, the cost of DNA [required physical space] is very high," says von Doehren. For instance, the gene encoding the enzyme that produces peptaibol, an antibiotic of 18 amino acids, is 80 kb long.4 Fitting in an enzyme that has been extrapolated to 60 residues would not be easy. "Very unlikely, but not impossible," he says.

Courtesy of Vincent A. Fischetti
 Vincent A. Fischetti and Sung Lee

MULTISTEP ASSEMBLY Such an unlikely scenario raises another issue that this enzyme is not the product of one gene, but rather, several. "We cannot only look at it in terms of one gene, one enzyme," explains Fischetti. He believes that LXPTGase production requires at least one of these complex enzymes to assemble it, with others being called into action that are responsible for the enzyme's modifications. The end result is a multistep assembly line. "It is not a single step from ribosome to enzyme. In this case, it is ribosome to enzyme[s] and that enzyme[s] then assembles the final polypeptide," he says.

However, neither conserved NRPS genes nor the unusual NRPS peptides have been found in streptococci so far. "When you look at the streptococcal genome, you don't find them," Fischetti adds. This lack of an obvious genomic NRPS signature and the uncharacteristically large size of the streptococcal enzyme further implicate multiple genes and enzymes.

"It's fairly obvious that the only way [this enzyme] could be produced is by being formed by a more complex mechanism," Fischetti continues. "You don't see these very large NRPS proteins which can be 1 mb in size. They're huge. To allow for several amino acids to be placed in sequence, you need different modules and domains with amino acid transfer being performed in these monster molecules. To produce the LPXTGase, such a protein would have to be three times the size [of known synthetases]. That's too huge and you don't see anything [in streptococci] to that effect."

But von Doehren says that large NRPS peptides arising from a single synthetase are entirely possible. "Peptide synthetases making repeated sequences may make peptides with more than 1000 residues, such as the gamma-D-glutamyl capsule of some bacilli," he explains. But in the case of LPXTGase, such repeated sequences were not found.

The Rockefeller group also discovered an inhibitor of the enzyme following ultrafiltration of the cell extract, which had been masking the presence of the enzyme by interfering with its activity. This as yet unnamed inhibitory molecule that can abrogate LPXTGase activity points to a potentially promising new class of antibiotics. Says Fischetti: "It kills all the Gram-positive bacteria that we've tested, including Bacillus anthracis, and kills most of the Gram-negative bacteria as well."

If Fischetti's proposal is proven correct, it is conceivable that other similarly made enzymes exist, which could help researchers unravel what is unknown about the genome.

Nicole Johnston ( is a freelance writer in Hamilton, Ontario.

1. S.G. Lee et al., "Characterization of a unique glycosylated anchor endopeptidase that cleaves the LPXTG sequence motif of cell surface proteins of Gram-positive bacteria," J Biol Chem, 277:46912-22, 2002.

2. D.E. Cane et al., "Harnessing the biosynthetic code: combinations, permutations, and mutations," Science, 282:63-8, 1998.

3. D. Schwarzer, M.A. Marahiel, "Multimodular biocatalysts for natural product assembly," Naturwissenschaften, 88:93-101, 2001.

4. A. Wiest et al., "Identification of peptaibols from Trichoderma virens and cloning of a peptaibol synthetase," J Biol Chem, 277:20862-8, 2002.