Degree Name

Doctor of Philosophy


Illawarra Health and Medical Research Institute


Streptococcus pyogenes (group A streptococcus, GAS) is a human specific pathogen responsible for a wide range of diseases. The majority of GAS infections give rise to uncomplicated disease, however, the migration of GAS from superficial to deep tissue sites can result in life-threatening invasive infections. GAS secrete streptokinase (SK), a potent plasminogen (Plg) activator which enables this bacterium to subvert the host Plg activation system to generate soluble and cell bound protease activity. Unlike the human Plg activators, u-PA and t-PA, which cleave the Plg activation bond (Arg561-VaI562) to generate the broad serine protease plasmin, SK lacks intrinsic protease activity. Instead, SK forms a stoichiometric complex with Plg and through non-proteolytic mechanisms, generates an active site in the bound Plg molecule to produce an activator complex with proteolytic activity. Among GAS isolates, SK gene (ska) sequences are polymorphic and can be grouped into two distinct sequence clusters (termed cluster type-1 and cluster type-2) with cluster type-2 being further divided into sub-clusters type-2a and type-2b. Allelic variants of SK produced by GAS isolates display unique Plg activation properties, however the biological significance of SK polymorphism among GAS isolates is yet to be elucidated.

This project describes the first comprehensive phenotypic study of GAS SK variants by characterising the structural, functional and biochemical differences displayed by SK molecules. Five distinct ska alleles representing examples from the three known phylogenetic sequence clusters, along with SKc from group C streptococcus were cloned and expressed as recombinant proteins. Structural analysis using far-UV circular dichroism spectroscopy indicated that all SK variants displayed similar secondary structure despite significant variation in amino acid sequence. Active site generation in Glu-Plg by SK variants was examined using the fluorescent active site titrant 4- methylumbelliferyl p-guanidinobenzoate. In these experiments, type-2b SK variants could not generate an active site in Glu-Plg through non-proteolytic mechanisms, while all other variants displayed this hallmark capacity of SK action. SKc, type-1 SK, and type-2a SK variants all bound human Glu-Plg with high affinity (KD ranging from 62 – 88 nM) when analysed by surface plasmon resonance experiments In comparison type- 2b SK variants displayed a 29-35 fold reduction in affinity for Glu-Plg. All SK variants had increased affinity (69 - 347 fold) for plasmin relative to Glu-Plg and could activate substrate Glu-Plg when a SK-plasmin activator complex was pre-formed.

Ligands that bind to Glu-Plg can affect the conformation of the protein which in turn can influence Glu-Plg activation. Therefore bacterial and human Plg binding ligands were investigated for their ability to influence the activation of Glu-Plg by SK variants. Despite type-2b SK not possessing the ability to generate an active site in native Glu- Plg, when active site experiments were repeated in the presence of fibrinogen, Plgbinding group A streptococcal M protein (PAM) (a Plg receptor constrained to type-2b expressing GAS strains), fibrinogen fragment D and fibrin, type-2b SK could generate an active site in Glu-Plg. Additionally, in contrast to the inhibition resistant SKc, type-1 and type-2a SK activator complexes, type-2b SK activator complexes (both those formed with Pln and those formed with Glu-Plg in the presence of ligands) were inhibited by the plasmin specific inhibitor, α2-antiplasmin (α2-AP). Interestingly, when a combination of PAM and fibrinogen was present type-2b SK activator complexes were resistant to α2-AP inhibition.

To determine if these observations could be translated into assays using more physiologically relevant conditions, activation assays were performed in pooled human plasma and in the presence of fibrin clots or whole cell GAS. As expected, in human plasma type-2b SK-Plg complexes were inhibited by the endogenous α2-AP unless exogeneous PAM was added to the assay. Additionally, when these assays were conducted in the presence of whole cell GAS expressing PAM, type-2b SK could activate Glu-Plg but the complex was only protected from inhibition by α2-AP when fibrinogen was also present. Similarly, when type -2b SK variants activated Glu-Plg that was bound to fibrin clots, these activator complexes were resistant to α2-AP inhibition. Taken together, this distinct Plg activation/inhibition mechanism displayed by type-2b SK variants would restrict plasmin activity to specific micro-environments within the host such as fibrin deposits or the bacterial cell surface through the action of α2-AP inhibition. As epidemiological studies have shown the type-2b streptokinase gene linage to be largely restricted to pam positive GAS strains which have a strong tendency to cause skin infections, we speculate that phenotypic SK variation functionally underpins a pathogenic mechanism whereby SK variants differentially focus Plg activation, leading to specific niche adaption within the host.

When the skatype-2a from the M1T1 GAS strain 5448 was exchanged with skatype-2b, the isogenic mutant displayed reduced virulence which was similar to that displayed by knockout mutant. Isogenic 5448 mutants that contained a skatype-1 allele displayed increased virulence. These data indicate that the phenotypic differences displayed by GAS SK variants do influence the invasive pathogenesis of this organism.

These findings suggest that SK variants produced by GAS isolates utilise distinct Plg activation pathways which, in turn, directly affects the pathogenesis of this organism. These findings have implications for the future elucidation of the molecular mechanism of SK mediated Plg activation, the generation of enhanced thrombolytic therapeutics and may highlight a potential role for the differential regulation of Plg activation in the pathogenesis of GAS (or the bacterial species that interact with the Plg activation system).