A high-resolution cryo-EM structure of a bacterial M-protein reveals a compact structure that diverges from related M-proteins

The surface of Streptococcus pyogenes (GAS) is studded with virulence determinants, with the most abundant being the characteristic M-protein used to serotype various strains of the bacterium. There are >250 strains of GAS serotypically distinguished by their M-proteins. Major pathogenic mechanisms of GAS require that this microorganism hijacks host components for survival, many of which are involved in hemostasis. One of these processes involves the binding of human host plasminogen (hPg) to an abundant GAS M-protein receptor (PAM). When bound to PAM, hPg is readily activated to the serine protease plasmin (hPm) by bacterial and host hPg activators, and cell-bound hPm is protected from inactivation by its natural inhibitors. This stabilizes a potent protease on GAS cells which aids in their survival and dissemination. Highly evolutionary domain-related M-proteins are assumed to form long alpha-helical projections, without tertiary structure, although no M-protein complete structure has been determined. Here, we employed cryogenic electron microscopy to solve such a structure anchored to a lentivirus particle membrane. Contrary to the belief in this field that M-proteins are extended long tropomyosin-like coils, we show that PAM folds through intra- and inter-domain interactions to a much more globular form on the cell surface. The nature of the folding and the many interactions involved in forming the PAM tertiary structure are summarized herein. SignificanceWe provide a unique approach to solve high-resolution structures of Streptococcus pyogenes (GAS) M-proteins, abundant virulence determinants on the GAS surface. Because of their unusual nature, no full high-resolution structure of any M-protein has been determined, especially when membrane-bound. Herein, we provide a unique general methodology for solving these structures by engineering a M-protein to be anchored to a lentivirus particle membrane for effective use in cryo-EM. Using this approach, we provide the first structure of a complete bacterial M-protein and show, that this M-protein is a monomeric globular structure on the cell surface, and not a dimeric coiled-coil, as generally believed. Thus, individual M-proteins may adopt structures that have evolved to accommodate their major host binding partner.