Cloning, purification and characterisation of Staphylococcus warneri lipase 2.

A gene encoding an extracellular lipase was identified in Staphylococcus warneri 863. The deduced lipase is organised as a prepro-protein and has significant similarity to other staphylococcal lipases. The mature part of the lipase was expressed with an N-terminal histidine tag in Escherichia coli, purified and biochemically characterised. The results show that the purified lipase (named SWL2) combines the properties of the staphylococcal lipases characterised so far. It has both a high preference for short chain substrates and surprisingly, it also displays phospholipase activity. Homology alignment was used to analyse sequence-function relationships of the staphylococcal lipase family with the aim to identify the structural basis underlying the different properties of the staphylococcal lipases.

Modifying the substrate specificity of staphylococcal lipases.

The lipase from Staphylococcus hyicus (SHL) displays a high phospholipase activity whereas the homologous S. aureus lipase (SAL) is not active or hardly active on phospholipid substrates. Previously, it has been shown that elements within the region comprising residues 254-358 are essential for the recognition of phospholipids by SHL. To specifically identify the important residues, nine small clusters of SHL were individually replaced by the corresponding SAL sequence within region 254-358. For cloning convenience, a synthetic gene fragment of SHL was assembled, thereby introducing restriction sites into the SHL gene and optimizing the codon usage. All nine chimeras were well-expressed as active enzymes. Eight chimeras showed lipase and phospholipase activities within a factor of 2 comparable to WT-SHL in standard activity assays. Exchange of the polar SHL region 293-300 by the more hydrophobic SAL region resulted in a 32-fold increased k(cat)/K(m) value for lipase activity and a concomitant 68-fold decrease in k(cat)/K(m) for phospholipase activity. Both changes are due to effects on catalytic turnover as well as on substrate affinity. Subsequently, six point mutants were generated; G293N, E295F, T297P, K298F, I299V, and L300I. Residue E295 appeared to play a minor role whereas K298 was the major determinant for phospholipase activity. The mutation K298F caused a 60-fold decrease in k(cat)/K(m) on the phospholipid substrate due to changes in both k(cat) and K(m). Substitution of F298 by a lysine in SAL resulted in a 4-fold increase in phospholipase activity. Two additional hydrophobic to polar substitutions further increased the phospholipase activity 23-fold compared to WT-SAL.

Catalytic role of the active site histidine of porcine pancreatic phospholipase A2 probed by the variants H48Q, H48N and H48K.

The catalytic contribution of His48 in the active site of porcine pancreatic phospholipase A2 was examined using site-directed mutagenesis. Replacement of His48 by lysine (H48K) gives rise to a protein having a distorted lipid binding pocket. Activity of this variant drops below the detection limit which is 10(7)-fold lower than that of the wild-type enzyme. On the other hand, the presence of glutamine (H48Q) or asparagine (H48N) at this position does not affect the structural integrity of the enzyme as can be derived from the preserved lipid binding properties of these variants. However, the substitutions H48Q and H48N strongly reduce the turnover number, i.e. by a factor of 10(5). Residual activity is totally lost after addition of a competitive inhibitor. We conclude that proper lipid binding on its own accelerates ester bond hydrolysis by a factor of 10(2). With the selected variants, we were also able to dissect the contribution of the hydrogen bond between Asp99 and His48 on conformational stability, being 5.2 kJ/mol. Another hydrogen bond with His48 is formed when the competitive inhibitor (R)-2-dodecanoylamino-hexanol-1-phosphoglycol interacts with the enzyme. Its contribution to binding of the inhibitor in the presence of an interface was found to be 5.7 kJ/mol.

Identification of a calcium binding site in Staphylococcus hyicus lipase: generation of calcium-independent variants.

In this study we have identified the presence of a high-affinity binding site for calcium in the lipase from Staphylococcus hyicus. By means of isothermal titration calorimetry we showed that the enzyme binds one calcium per molecule of enzyme with a dissociation constant of 55 microM. The residual activity of the apoenzyme compared to the activity in the presence of calcium ions varies from 65% at 10 degreesC to nearly zero at 40 degreesC. On the basis of primary sequence alignment with other staphylococcal lipases and the lipases from Bacillus thermocatenulatus and from Pseudomonas glumae in combination with site-directed mutagenesis, aspartates 354 and 357 could be identified as calcium ligands. Kinetic measurements with the D357E variant showed that replacement of Asp357 by a glutamate decreased the affinity for calcium ions 30-fold. Introduction of a lysine, an asparagine, or an alanine at position 357 and of a lysine or an asparagine at position 354 resulted in calcium-independent variants. Isothermal titration calorimetry confirmed the loss of calcium binding. Although the D357K, D357N, and D357A variants did not bind calcium, at room temperature they were nearly as active as wild-type lipase in the presence of calcium, but at elevated temperatures these calcium-independent lipases showed a reduced activity. Over the whole temperature range the activities of the D354K and D354N variants are significantly lower than wild-type enzyme in the presence of calcium and are comparable to the activity of the wild-type apoenzyme. Our results show that binding of calcium is important for the structural stabilization of staphylococcal lipases (and possibly other lipases) and that it is possible to engineer calcium-independent variants on the basis of limited structural homology with another lipase.

The phospholipase activity of Staphylococcus hyicus lipase strongly depends on a single Ser to Val mutation.

Site-directed mutagenesis and domain exchange were used to investigate the role of the C-terminal domains of Staphylococcus hyicus lipase (SHL) and S. aureus lipase (SAL) in substrate selectivity. The introduction of a single point mutation coding for the substitution of Val for Ser356 in SHL yields an enzyme which has retained full lipase activity, but with more than 12-fold lower phospholipase activity. Starting with this S356V variant of SHL the C-terminal 40 amino acids were replaced by the corresponding SAL sequence. Although 23 amino acid changes were introduced simultaneously the impact on the phospholipase/lipase activity ratio was only 4-fold. We therefore conclude that in the C-terminal domain it is Ser356 which mainly determines phospholipase activity. The introduction of a Val357 to Ser substitution in SAL did not turn SAL into a phospholipase, showing that residues from other domains contribute to this activity as well. The results are discussed in view of the sequence homology of lipases and (lyso)phospholipases.

Cloning, purification and characterisation of the lipase from Staphylococcus epidermidis--comparison of the substrate selectivity with those of other microbial lipases.

On the chromosome of Staphylococcus epidermidis RP62A the lipase gene (gehSE1) is immediately flanked by the icaAA'BC operon, which is involved in biofilm formation. Since lipase production might play a role in staphylococcal skin colonisation as well, we studied the biochemical properties of the staphylococcal lipases more closely. The DNA sequence and the deduced protein sequence revealed that gehSE1 is very similar to the lipase sequence of S. epidermidis strain 9. Like other staphylococcal lipases, gehSE1 is organised as a preproenzyme. The part of gehSE1 coding for the mature lipase was cloned and overexpressed as a fusion protein with an N-terminal histidine tag in Escherichia coli. The lipase was purified to homogeneity using a combination of precipitation techniques, metal-affinity chromatography and gel filtration. Biochemical characterisation showed that this lipase is closely related to the lipase from Staphylococcus aurelis NCTC8530. Both enzymes have a pH optimum around 6, are very stable at low pH, and need calcium as a cofactor for catalytic activity. The preferred substrates are small triacylglycerols, with a maximum activity toward tributyrylglycerol. Comparison of the substrate selectivity with those of other microbial lipases showed that phospholipids are generally poor substrates for lipases. An exception is the lipase from Staphylococcus hyicus, which prefers phospholipids as a substrate, distinguishing this staphylococcal lipase from other microbial lipases. These results are discussed in view of the structure/function relationships of staphylococcal lipases, and the possible involvement of these enzymes in biological processes such as skin colonisation and pathogenesis.

Substrate specificity of Staphylococcus hyicus lipase and Staphylococcus aureus lipase as studied by in vivo chimeragenesis.

Staphylococcus hyicus lipase (SHL) and Staphylococcus aureus lipase (SAL) are highly homologous enzymes, yet they show remarkable differences in their biochemical characteristics. SHL displays a high phospholipase activity, hydrolyses neutral lipids, and has no chain length preference, whereas SAL only degrades short-chain fatty acid esters. To identify the regions in the primary sequence of SHL responsible for phospholipase activity and chain length selectivity, a set of histidine-tagged SAL/SHL chimeras was generated by in vivo recombination in Escherichia coli. Several classes of chimeric enzymes were identified on the basis of restriction site analysis. All chimeras were well-expressed as active enzymes. They were characterized for their specific activities on both phospholipids and p-nitrophenyl esters of various chain lengths. Phospholipase activity appeared to be determined by three regions, all located in the C-terminal domain of SHL. Testing of the enzymatic activity of the chimeras toward p-nitrophenyl esters showed that chain length selectivity is defined by elements within the region of residues 180-253. Moreover, also residues along the stretch 275-358 contribute to the binding of acyl chains. Interestingly, several chimeras were even more active than the parent enzymes on long-chain p-nitrophenyl esters.