Localization and Crystal Structure
Streptococcal a-enolase is essential for bacteria, since the protein plays an important role as an enzyme in the EMP pathway. It is therefore considered impossible to construct a conventional knockout mutant with the eno gene encoding a-enolase (Bergmann et al. 2001; Feng et al. 2009; Boleij et al. 2011), which hampers studies of its possible role in streptococcal virulence. The amino acid sequences in streptococcal species share a 90-100% identity. Streptococcal a-enolase localizes in the cytoplasm as well as mammalian cells. In addition, a-enolase is expressed on the surface of bacterial cells and released into culture supernatant, although it lacks a signal sequence and the cell-wall-anchoring motif LPXTG (Pancholi and Fischetti 1998; Bergmann et al. 2001; Hughes et al. 2002; Ge et al. 2004; Mori et al. 2012). A binding experiment with radio-labeled recombinant a-enolase and S. pneumoniae clarified that secreted a-enolase can be re-associated onto the pneumococcal cell surface (Bergmann et al. 2001). Furthermore, ELISA findings obtained using an autolysin-deficient mutant strain showed that approximately 24% of pneumococcal a-enolase is released into supernatant in an autolysin-dependent manner (Mori et al. 2012). However, the secretion system and cell-surface localization mechanism require further study.
The crystal structures of a-enolase have been identified in many eukaryotic and prokaryotic organisms and shown to form dimers or octamers (Pancholi 2001; Ehinger et al. 2004). Biological multimeric structures are necessary for these enzyme activities. The monomers of yeast enolase formed by hydrostatic pressure are enzymatically inactive (Kornblatt et al. 1998). Although most organisms including the Gram-negative bacterium Escherichia coli express a-enolase as a dimeric structure, at least some Gram-positive organisms such as Bacillus subtilis and several Streptococcus species express octameric a-enolase (Brown et al. 1998; Ehinger et al. 2004). Human and pneumococcal multimeric structures are shown in Figure 14.1. There is a possibility that ancestral enolase was an octamer and the evolutionary process has independently altered enolase multiple times to form a dimeric structure (Fig. 14.2; Brown et al. 1998). It is interesting to note that there is a conformational difference between the a-enolase of the y-proteobacterium such as E. coli and firmicutes such as B. subtilis and Streptococcus species. The divergence point for these groups of bacteria is estimated to have occurred 2.8-3.6 billion years ago (Battistuzzi et al. 2004). The dimeric form of GAS a-enolase is unstable regardless of the fact that enolases from most organisms are dimers. No significant dimer or multimer amounts were formed in dissociation experiments using mutations such as F137L and E363G to destabilize the octameric structure (Karbassi et al. 2010). Cork et al. demonstrated that a lysine residue at position 344 in GAS a-enolase plays an important role in maintaining the integrity of the octameric structure using site- directed mutagenesis of surface-located lysine residues. In addition, findings from a surface plasmon resonance (SPR) assay with immobilized plasminogen showed that the mutation K344E lacked enzyme activity and monomers of yeast enolase, and also showed increased plasminogen-binding ability. On the other hand, K304A and K334A mutations did have altered plasminogen-binding ability as compared to the wild type, whereas other lysine residue mutations in
Figure 14.1 Multimeric crystal structure of S. pneumoniae and human a-enolase.
S. pneumoniae a-enolase (PDBID: 1w6t) forms an octamer and the human version (PDBID: 3B97) is a dimer. This figure was produced using Yorodumi (run by PDBj) with Jmol, available at http://pdbj.org/emnavi/viewtop.php.
Figure 14.2 Phylogenetic tree of Streptococcus a-enolase. The tree is calculated and drawn by "Average Distance using BLOSUM62" on Jalview 2.8.2. The amino acid sequences were obtained from each genome referential strains.
K252 + 255A, K435L, and A434-435 resulted in greatly diminished binding ability (Cork et al. 2009). These results indicate that the presence of lower multimeric forms may interact more efficiently with plasminogen. Kornblatt et al. reported that the native structure of octameric a-enolase of GAS is not able to interact with human plasminogen based on findings obtained with multiprolonged methods, including fluorescence polarization, isothermal titration calorimetry, dynamic light scattering, analytical ultracentrifugation, pull-down reactions, and SPR. Interestingly, the non-native structure of monomeric or multimeric enolase is capable of binding human plasminogen, while the nonnative structure of plasminogen can also bind native enolase (Kornblatt et al. 2011). Recently, another group that included Kornblatt used a different set of multi-prolonged methods, including dual polarization interferometry, atomic force microscopy observations, isothermal titration calorimetry, dynamic light scattering, and fluorescence resonance energy transfer, and showed that native GAS a-enolase and plasminogen can bind with each other in the presence of a sticky surface (Balhara et al. 2014).