ARTICLES https://doi.org/10.1038/s41564-018-0159-x © 2018 Macmillan Publishers Limited, part of Springer Nature. All rights reserved. 1 Department of Medical Microbiology, University Medical Center Utrecht, Utrecht, The Netherlands. 2 Department of Microbiology, University of Chicago, Chicago, IL, USA. 3 CIRI, Centre International de Recherche en Infectiologie, Inserm, U1111, Université Claude Bernard Lyon 1, CNRS UMR5308, Ecole Normale Supérieure de Lyon, Université Lyon, Hospices Civils de Lyon, Lyon, France. 4 Department of Microbiology and Immunology, UCSF Diabetes Center, Keck Center for Noncoding RNA, University of California, San Francisco, San Francisco, CA, USA. 5 Institute for Glycomics, Griffith University, Gold Coast, Queensland, Australia. 6 St. Giles Laboratory of Human Genetics of Infectious Diseases, Rockefeller Branch, The Rockefeller University, New York, NY, USA. 7 These authors contributed equally: Angelino T. Tromp, Michiel Van Gent. 8 These authors jointly supervised this work: Thomas Henry, András N. Spaan. *e-mail: thomas.henry@inserm.fr; a.n.spaan@umcutrecht.nl S taphylococcus aureus is a major bacterial pathogen in humans and is responsible for a diverse disease spectrum, ranging from superficial skin and soft tissue infections to severe invasive disease. Severe infections with S. aureus have a poor prognosis 1 . Treatment is further complicated by the emergence of methicillin- resistant S. aureus (MRSA) strains 2 and by a lack of advancements in vaccine development 3 . A better understanding of the host–pathogen interaction during infection with S. aureus is essential to develop new therapeutic approaches. Phagocytes have a pivotal role in the containment of S. aureus early after infection 4 . To counteract elimination by phagocytes, S. aureus secretes an arsenal of virulence factors. Among these are the leukocidins, a family of bi-component pore-forming toxins that target and kill phagocytes 5,6 . Human S. aureus isolates secrete up to five different leukocidins 7 : Panton–Valentine leukocidin (PVL; also known as LukSF-PV), γ-haemolysin AB and CB (HlgAB and HlgCB, respectively), leukocidin ED (LukED) and leukocidin AB (LukAB; also known as LukGH). Chromatography elution profiles differentiate the leukocidin protein components into S-migrating (slow) and F-migrating (fast) components, which are, with the exception of LukAB, secreted as inactive monomers 5 . Each canoni- cal leukocidin combination consists of S- and F-components that hetero-oligomerize into an octameric membrane-spanning pore 5,7 . The leukocidins show structural and functional resemblance to the single-component pore-forming toxin of S. aureus, α-toxin (haemolysin-α (Hla)) 8 , but the biological rationale for a bi-compo- nent system remains unresolved. Functional interactions by forma- tion of non-canonical combinations of S- and F-components that are active (as for PVL and HlgCB 9–11 ) or inactive (as for PVL and LukED 12 ) suggest that the contribution of leukocidins to pathogen- esis differs when expressed simultaneously. However, the contribu- tion of the leukocidins to infection is incompletely understood 5,7 . Specificity for human phagocytes and resistance of murine phagocytes to the majority of leukocidins hinder investigation dur- ing infection 7 . Recently, proteinaceous receptors have been iden- tified for all leukocidins 7,13–20 . These receptors are targeted by the S-components in a species-specific manner. For the S-components of PVL and HlgCB, LukS-PV and HlgC, respectively, the human complement C5a receptor 1 (hC5aR1) was identified as the major receptor 14,16 . The identification of hC5aR1 as a shared receptor for LukS-PV and HlgC explains the specificity for human phagocytes as both toxins are incompatible with the murine C5aR1 ortho- logue 14,16 . Although differences exist in the interaction between LukS-PV and HlgC with hC5aR1 (refs 21,22 ), the necessity for S. aureus to secrete apparently redundant toxins is incompletely appreciated 7 . Even though receptors have been identified for all leukocidin S-components, it remains to be established whether the F-components also have host receptors. A recent study on the Human CD45 is an F-component-specific receptor for the staphylococcal toxin Panton–Valentine leukocidin Angelino T. Tromp 1,7 , Michiel Van Gent  1,7 , Pauline Abrial 3 , Amandine Martin 3 , Joris P. Jansen 1 , Carla J. C. De Haas 1 , Kok P. M. Van Kessel 1 , Bart W. Bardoel  1 , Elisabeth Kruse 1 , Emilie Bourdonnay 3 , Michael Boettcher 4 , Michael T. McManus 4 , Christopher J. Day 5 , Michael P. Jennings  5 , Gérard Lina 3 , François Vandenesch  3 , Jos A. G. Van Strijp 1 , Robert Jan Lebbink 1 , Pieter-Jan A. Haas 1 , Thomas Henry  3,8 * and András N. Spaan  1,6,8 * The staphylococcal bi-component leukocidins Panton–Valentine leukocidin (PVL) and γ-haemolysin CB (HlgCB) target human phagocytes. Binding of the toxins’ S-components to human complement C5a receptor 1 (C5aR1) contributes to cellular tropism and human specificity of PVL and HlgCB. To investigate the role of both leukocidins during infection, we developed a human C5aR1 knock-in (hC5aR1 KI ) mouse model. HlgCB, but unexpectedly not PVL, contributed to increased bacterial loads in tissues of hC5aR1 KI mice. Compared to humans, murine hC5aR1 KI neutrophils showed a reduced sensitivity to PVL, which was mediated by the toxin’s F-component LukF-PV. By performing a genome-wide CRISPR–Cas9 screen, we identified CD45 as a receptor for LukF-PV. The human-specific interaction between LukF-PV and CD45 provides a molecular explanation for resistance of hC5aR1 KI mouse neutrophils to PVL and probably contributes to the lack of a PVL-mediated phenotype during infection in these mice. This study demonstrates an unsuspected role of the F-component in driving the sensitivity of human phagocytes to PVL. Corrected: Publisher Correction NATURE MICROBIOLOGY | VOL 3 | JUNE 2018 | 708–717 | www.nature.com/naturemicrobiology 708