Aljannat, Mahab
(2017)
Bacterial moonlighting proteins of N. meningitidis: interaction with the host and role in pathogenesis.
PhD thesis, University of Nottingham.
Abstract
Neisseria meningitidis is a human-restricted pathogen that colonizes the nasopharynx, mostly without causing any disease. However, bacteria can invade into the bloodstream and cross the blood brain barrier resulting in life threatening sepsis and meningitis.
Protein moonlighting is a concept established to describe the additional task(s) that a protein exhibits alongside its canonical function. After translocation to the cell surface by an unknown mechanism, moonlighting proteins may interact with various host proteins, and exhibit virulence-associated functions.
Enolase (Eno), peroxiredoxin (Prx) and DnaK are meningococcal housekeeping proteins that are functionally described as a glycolytic enzyme, an anti-oxidant protein and a molecular chaperon, respectively. They have been identified on the surface of meningococci achieving non-housekeeping (moonlighting) functions related to interactions with the human protein plasminogen. This study sheds light on the moonlighting properties of these three proteins in the pathogenesis of meningococcal disease.
Meningococcal Eno, Prx and DnaK were cloned, overexpressed in E. coli cells, and the corresponding wild type recombinant proteins were affinity purified under non-denaturing conditions. The oligomerization status of each recombinant protein was determined by analytical gel filtration, in which rEno was identified as an octamer, rPrx as a hexamer and rDnaK was either in a dimeric or monomeric state. Polyclonal antiserum targeted against each recombinant protein was raised in rabbits.
A markerless pxn-knockout was generated in N. meningitidis MC58 rpsL-. The streptomycin resistance phenotype that the rpsL- allele confers was the basis behind the adopted mutation strategy. To delete NMB0946 (encoding meningococcal Prx), two mutagenic plasmids (pGUD2 and pGUD3) were constructed to allow two homologous recombination events. N. meningitidis rpsL- ∆pxn KanR rpsL+ was the resultant strain from the first event. The second homologous recombination event, facilitated by pGUD3, involved removing the selectable marker (KanR plus rpsL+ allele) to generate the markerless N. meningitidis rpsL- ∆pxn. Given that NMB0946 is part of a two-gene operon necessitated the construction of a complementation strain, in which a wild type copy of NMB0946 was reintroduced at an ectopic site of N. meningitidis rpsL- ∆pxn genome to complement that deletion.
Using equimolar amounts of the recombinant proteins in EIA assays, it was found that meningococcal rEno binds plasminogen (Plg) more strongly than rDnaK and rPrx. Plg binding was inhibited when the lysine analogue ϵ-aminocapronic acid was added suggesting the potential involvement of lysine residues. Substitution of the C-terminal lysine residues of rEno, rPrx and rDnaK with alanine residues, significantly yet not completely, reduced binding to Plg. For rEno, lysine residues at positions 297 and 355 were identified for the first time as additional sites for Plg. rPrx with mutation in the active cysteine site (rPrx185CA), which is known to inactivate the protein, was able to bind Plg to the same level as the wild type strain. Unlike the C-terminal lysine residues, substituting the sub-terminal lysines of rPrx and rDnaK (at positions 230 and 641, respectively) did not alter binding to Plg.
Employing whole-cell EIA, Prx and DnaK could be detected on the surface of wild-type encapsulated N. meningitidis MC58, while surface localisation of Eno was not detected under these conditions.
Under hydrogen peroxide-mediated oxidative stress, the N. meningitidis ∆pxn-mutant survived remarkably better than the parental strain. Using human whole blood as a model of meningococcal bacteraemia, it was found that N. meningitidis ∆pxn-mutant has a survival defect compared with the wild-type strain. Preliminary data suggest that the survival of the ∆pxn-mutant cells in the presence of polymorphonuclear leukocytes (PMNs) isolated from peripheral blood was similar to controls containing no PMNs. Moreover, the absence of Prx could not significantly reduce the ability of whole meningococcal cells to bind Plg.
In summary, the data suggest that Eno, Prx and DnaK bind plasminogen mainly via the C-terminal lysine residues. Lysine 297 and 355 of rEno are required for optimal Plg binding. The canonical and the moonlighting functions of meningococcal Prx are independent. This study also proposes that the absence of Prx does not impair overall Plg binding, has a positive impact on the ability of meningococci to tolerate exogenous hydrogen peroxide, and finally, has a very significant effect on the survival of meningococcal cells in human whole blood.
Actions (Archive Staff Only)
 |
Edit View |
Tools
Tools
![[img]](/style/images/fileicons/application_pdf.png)