Search

The New Delhi metallo-β-lactamase (NDM-1) mediates resistance to β-lactam antibiotics. NDM-1 was likely formed as the result of a gene fusion between sequences encoding the first six amino acids of cytoplasm-localised aminoglycosidase, AphA6, and a periplasmic metallo-β -lactamase. We show that NDM-1 has an atypical signal peptide and is inefficiently secreted. Two new blaNDM-1 alleles that have polymorphisms in the signal peptide; NDM-1(P9R), a proline to arginine substitution, and NDM-2, a proline to alanine substitution (P28A) were studied. Here, we show that both the P9R and P28A substitutions improve secretion compared to NDM-1 and display higher resistance to some β-lactam antibiotics. Mass spectrometry analysis of these purified NDM proteins showed that the P28A mutation in NDM-2 creates new signal peptide cleavage sites at positions 27 and 28. For NDM-1, we detected a signal peptide cleavage site between L21/M22 of the precursor protein. We find no evidence that NDM-1 is a lipoprotein, as has been reported elsewhere. In addition, expression of NDM-2 improves the fitness of E. coli, compared to NDM-1, in the absence of antibiotic selection. This study shows how optimization of the secretion efficiency of NDM-1 leads to increased resistance and increased fitness.

Secreted proteins contain an N-terminal signal peptide to guide them through the secretion pathway. Once the protein is translocated, the signal peptide is removed by a signal peptidase, such as signal peptidase I. The signal peptide has been extensively studied and reviewed; however, the mature region has not been the focus of review. Here we cover the experimental evidence that highlights the important role of the mature region amino acid residues in both the efficiency and the ability of secreted proteins to be successfully exported via secretion pathways and cleaved by signal peptidase I.

Bacillus subtilis has five type I signal peptidases, one of these, SipW, is an archaeal-like peptidase. SipW is expressed in an operon (tapA-sipW-tasA) and is responsible for removing the signal peptide from two proteins: TapA and TasA. It is unclear from the signal peptide sequence of TasA and TapA, why an archaeal-like signal peptidase is required for their processing. Bioinformatic analysis of TasA and TapA indicates that both contain highly similar signal peptide cleavage sites, both predicted to be cleaved by Escherichia coli signal peptidase I, LepB. We show that expressing full length TasA in E. coli is toxic and leads to cell death. To determine if this phenotype is due to the inability of the E. coli LepB to process the TasA signal peptide, we fused the TasA signal peptide and two amino acids of mature TasA (up to P2′) to both maltose binding protein (MBP) and β-lactamase (Bla). We observed a defect in secretion, indicated by an abundance of unprocessed protein with both TasA-MBP and TasA-Bla fusions. A series of mutations in both TasA-MBP and TasA-Bla were made around the junction of the TasA signal peptide and the fusion protein. Both of these studies indicate that residues around the predicted TasA signal sequence cleavage site, particularly the sequence from P3 to P2′, inhibit processing by LepB. The cell death observed when TasA and TasA signal sequence fusion proteins are expressed is likely due to the TasA signal peptide blocking LepB and thereby the general secretion pathway.

Signal peptides direct proteins from the cytoplasm to the periplasm. These N-terminal peptides are cleaved upon entry to the periplasm by either signal peptidase I, or signal peptidase II for lipoproteins. Signal peptidase I is a serine protease that has either a serine-lysine or serine-histidine catalytic dyad present in the active site. The recognition site for signal peptide cleavage by signal peptidase I has been defined primarily by an Ala-X-Ala motif at the C-terminal end of the signal peptide, one amino acid away from the cleavage site. We used a verified set of signal peptidase I cleaved proteins from E. coli to look for novel conserved features, focusing on the N-terminus of the mature protein. We observed a striking bias for the presence of acidic residues at second position of the mature protein (P2'), and a complete absence of aromatic amino acids at the same position. Whole genome analysis of the predicted set of all E. coli and B. subtilis secreted proteins confirmed the same strong bias for acidic residues at P2' of the mature protein, and against aromatic amino acids at the same position. When these studies were extended to archaeal genomes (M. voltae and S. tokodaii) and the yeast genome from S. cerevisiae, this bias was not observed. E. coli and B. subtilis primarily express a signal peptidase I contains a serine-lysine catalytic dyad, whilst those of archaeal and eukaryotic origin generally have a serine-histidine catalytic dyad. This difference may explain the differential bias for acidic residues and against aromatic residues at P2'. These observations suggest additional key residues that may favor or prevent signal sequence recognition or cleavage by signal peptidase I, and thereby facilitate more accurate in silico prediction of signal peptidase I cleavage sites.

Exported proteins require an N-terminal signal peptide to direct them from the cytoplasm to the periplasm. Once the protein has been translocated across the cytoplasmic membrane, the signal peptide is cleaved by a signal peptidase, allowing the remainder of the protein to fold into its mature state in the periplasm. Signal peptidase I (LepB) cleaves non-lipoproteins and recognises the sequence Ala-X-Ala. Amino acids present at the N-terminus of mature, exported proteins have been shown to affect the efficiency at which the protein is exported. Here we investigated a bias against aromatic amino acids at the second position in the mature protein (P2′). Maltose binding protein (MBP) was mutated to introduce aromatic amino acids (tryptophan, tyrosine and phenylalanine) at P2′. All mutants with aromatic amino acids at P2′ were exported less efficiently as indicated by a slight increase in precursor protein in vivo. Binding of LepB to peptides that encompass the MBP cleavage site were analysed using surface plasmon resonance. These studies showed peptides with an aromatic amino acid at P2′ had a slower off rate, due to a significantly higher binding affinity for LepB. These data are consistent with the accumulation of small amounts of preMBP in purified protein samples. Hence, the reason for the lack of aromatic amino acids at P2′ in E. coli is likely due to interference with efficient LepB activity. These data and previous bioinformatics strongly suggest that aromatic amino acids are not preferred at P2′ and this should be incorporated into signal peptide prediction algorithms.

Non-optimal codons are defined by low usage and low abundance of corresponding tRNA, and have an established role in translational pausing to allow the correct folding of proteins. Our previous work reported a striking abundance of non-optimal codons in the signal sequences of secretory proteins exported via the sec-dependent pathway in Escherichia coli. In the current study the signal sequence of maltose-binding protein (MBP) was altered so that non-optimal codons were substituted with the most optimal codon from their synonymous codon family. The expression of MBP from the optimized allele (malE-opt) was significantly less than wild-type malE. Expression of MBP from malE-opt was partially restored in a range of cytoplasmic and periplasmic protease deficient strains, confirming that reduced expression of MBP in malE-opt was due to its preferential degradation by cytoplasmic and periplasmic proteases. These data confirm a novel role for non-optimal codon usage in secretion by slowing the rate of translation across the N-terminal signal sequence to facilitate proper folding of the secreted protein.

Non-optimal codons are generally characterised by a low concentration of isoaccepting tRNA and a slower translation rate compared to optimal codons. In a previous study, we reported a 20-fold reduction in maltose binding protein (MBP) level when the non-optimal codons in the signal sequence were optimised. In this study, we report that the 20-fold reduction is rescued when MBP is expressed at 28 degrees C instead of 37 degrees C, suggesting that the signal sequence optimised MBP protein (MBP-opt) may be misfolded, and is being degraded at 37 degrees C. Consistent with this idea, transient induction of the heat shock proteases prior to MBP expression at 28 degrees C restores the 20-fold difference, demonstrating that the difference in production levels is due to post-translational degradation of MBP-opt by the heat-shock proteases. Analysis of the structure of purified MBP-wt and MBP-opt grown at 28 degrees C showed that although they have similar secondary structure content, MBP-opt is more resistant to thermal unfolding than is MBP-wt. The two proteins also exhibit different tryptic fragment profiles, further confirming that they are folded into conformationally different states. This is the first study to demonstrate that signal sequence non-optimal codons can influence the folding of the mature exported protein.

Signal sequence non-optimal codons have been shown to be important for the folding and efficient export of maltose binding protein (MBP), a SecB dependent protein. In this study, we analysed the importance of signal sequence non-optimal codons of TolB, a signal recognition particle (SRP) dependent exported protein. The protein production levels of wild type TolB (TolB-wt) and a mutant allele of TolB in which all signal sequence non-optimal codons were changed to a synonymous optimal codon (TolB-opt), revealed that TolB-opt production was 12-fold lower than TolB-wt. This difference could not be explained by changes in mRNA levels, or plasmid copy number, which was the same in both strains. A directed evolution genetic screen was used to select for mutants in the TolB-opt signal sequence that resulted in higher levels of TolB production. Analysis of the 46 independent TolB mutants that reverted to wild type levels of expression revealed that at least four signal sequence non-optimal codons were required. These results suggest that non-optimal codons may be required for the folding and efficient export of all proteins exported via the Sec system, regardless of whether they are dependent on SecB or SRP for delivery to the inner membrane.