Back

Mapping the Mutational Landscape for Streptokinase Binding to Plasminogen

Baid, S.; Holding, M. L.; Laurin, T.; Siemieniak, D.; Haynes, L. M.; Ginsburg, D.

2026-02-04 microbiology
10.64898/2026.02.04.703780 bioRxiv
Show abstract

Group A Streptococcus (GAS) expresses streptokinase (SK), a critical virulence factor that non-enzymatically activates the hosts plasminogen (PLG), to an active form, PLGSK, resulting in the degradation of fibrin clots and subsequent bacterial dissemination. PLGSK formation contrasts with the physiologic activation of PLG to the serine protease plasmin via a proteolytic mechanism. As a potent thrombolytic, SK has been used as a therapeutic to treat heart attacks and strokes. GAS SK is highly specific for human PLG, and sequence variation between SK from different GAS strains has been linked to differences in PLG binding and disease severity. We now report the application of deep mutational scanning (DMS) to map the effects of [~]71% of single amino acid substitutions within SK from Group C Streptococcus, which shares high sequence homology with GAS on its ability to bind human PLG. We first demonstrate that SK expressed as a fusion protein to the p3 coat protein of M13 filamentous phage retains its capacity to bind human PLG. Our subsequent DMS analysis using this phage system identifies regions of SK in which amino acid substitutions are likely to increase or decrease its affinity for PLG. Our findings suggest a complex protein-protein interaction in which long-range protein dynamics influence the conformational activation of PLG to PLGSK. These data lay the foundation for linking SK variation between GAS strains to differences in virulence, mapping the determinants of GAS SKs human specificity, and potentially contributing to the development of improved therapeutics for heart attack and stroke.

Matching journals

The top 3 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.