Phosphorylation
Phosphorylation often occurs on serine, threonine, and tyrosine residues of target proteins with the addition of phosphate by kinases and its removal by phosphatases. As a highly efficient modification, phosphorylation modulates protein functions by altering conformation, promoting ubiquitination and degradation, influencing protein-protein interactions, or regulating enzyme activities (Huttlin et al., 2010). Due to long-standing co-evolution with mammalian host cells, many bacterial pathogens have evolved effector proteins harboring kinase or phosphatase activities to directly engage with host phosphorylation signaling (Grishin et al., 2015). Yersinia pseudotuberculosisYpkA is arguably the first reported prokaryotic virulence factor possessing Ser/Thr protein kinase activity (Galyov et al., 1993) and its activation is dependent on a host factor, actin (Barz et al., 2000, Dukuzumuremyi et al., 2000, Juris et al., 2000). YpkA phosphorylates the heterotrimeric G protein, Gαq, on a critical serine residue, thereby inhibiting its GTP binding (Navarro et al., 2007). In addition to its kinase activity, YpkA also possesses the guanidine nucleotide dissociation inhibitor (GDI) domain, and together they work synergistically to disrupt host actin cytoskeleton (Prehna et al., 2006). Later, vasodilator-stimulated phosphoprotein (VASP) was reported to be another enzymatic substrate of YpkA and its modification led to the disruption of host actin dynamics (Ke et al., 2015). S.Typhimurium effector SteC harbors kinase activity that is required for the formation of the F-actin meshwork associated withSalmonella -containing vacuoles (SCVs). Several enzymatic substrates of SteC have been identified by various approaches, including Hsp27 (Imaimi et al., 2013), formin family FMNL proteins (Walch et al., 2021; Poh et al., 2008) and the MAP kinase MEK (Odendall et al., 2012). Together, SteC-mediated phosphorylation of these target proteins coordinately modulates the host F-actin cytoskeleton (Heggie et al., 2021).
L. pneumophila exploits phosphorylation signaling pathways by delivering an array of kinase effectors into host cells including LegK1-K4, LegK7 and Lem28/Lpg2603 (Hervet et al., 2011; Lee et al., 2018; Sreelatha et al., 2020). LegK1 was first identified in a screen ofL. pneumophila effectors that activate the NF-κB pathway (Ge et al., 2009). Mechanistically, LegK1 directly phosphorylates the IκB family of inhibitors. LegK2 targets the ARP2/3 complex whose phosphorylation triggers actin cytoskeleton remodeling in host cells (Michard et al., 2015). LegK7 hijacks the host Hippo pathway by phosphorylating the scaffolding protein MOB1, leading to the degradation of the transcriptional regulators TAZ and YAP1 (Lee et al., 2018). The host targets of LegK4 were identified to be the Hsp70 chaperone family by using a chemical genetic screen (Moss et al., 2019). LegK4-dependent phosphorylation of cytosolic Hsp70 inhibits its ATPase activity and hence protein folding capacity, resulting in global translation arrest in host cells. Thus far, host targets of other L. pneumophilakinase effectors are still elusive and their functions in bacterial infection remain to be investigated.
Other than kinases, bacterial pathogens also encode and secret effector proteins harboring phosphatase activities. Y. pseudotuberculosisYopH is perhaps the first reported bacterial effector possessing potent tyrosine phosphatase activity (Bliska et al., 1991). Later, two independent reports found that YopH targets focal adhesions in host cells by dephosphorylating p130Cas and FAK, thereby disrupting peripheral focal complexes and inhibiting bacterial uptake into HeLa cells (Persson et al., 1997; Black et al., 1997). In T cells, YopH targets the tyrosine kinases Lck and ZAP-70 together with the adaptors SLP-76 and LAT, and their dephosphorylation alters T cell-mediated immune responses (Alonso et al., 2004; Gerke et al., 2005). Overlapping with the previously reported targets in different host cells, the PRAM-1/SKAP-HOM and SLP-76 signaling pathways are also dephosphorylated in a YopH-dependent manner to impair neutrophil responses in an animal infection model (Rolán et al., 2013). In many of these studies, a catalytic mutant of YopH was often employed to precipitate the interacting proteins for identifying potential phosphatase substrates. Furthermore, Salmonella effector SptP has a C-terminal protein tyrosine phosphatase domain fused to its N-terminal GTPase activating protein (GAP) domain (Stebbins & Galán, 2000). It dephosphorylates the host AAA+ ATPase VCP to regulate the biogenesis ofSalmonella -containing vacuoles (SCVs) (Humphreys et al., 2009). Again, the substrate VCP was identified from the pull-down samples by using a phosphatase inactive mutant of SptP. Furthermore, SptP targets a tyrosine kinase Syk and a vesicle fusion protein, N-ethylmalemide-sensitive factor to suppress immune responses in mast cells through its phosphatase activity (Choi et al., 2013). Additional phosphatase effectors include WipA, WipB, Ceg4 and Lem4 from L. pneumophila (Pinotsis & Waksman, 2017; Prevost et al., 2017; Quaile et al., 2018; Beyrakhova et al., 2018) and Coxiella burnetiieffector CinF that dephosphorylates IκBα to inhibit host NF-κB signaling (Zhang et al., 2022).
Nevertheless, our understanding of bacterial manipulation of host phosphorylation signaling pathways remains rather limited, largely because of technical challenges in sensitively and reliably measuring phosphorylation events of proteins (especially on a global level). Therefore, many studies on bacterial kinases or phosphatases, in particular some of the early work as we discussed above, were carried out without the assistance of the latest MS tools on more definitive measurements of protein modifications as well as precise determination of phosphorylation sites. In addition, the potential impact(s) of a bacterial kinase/phosphatase was not probed on systems-level during bacterial infection. Such holistic views are particularly desired when a single effector (e.g., YpkA, SteC, YopH and SptP) may target multiple host pathways. Fortunately, the arrival and maturation of MS-based phosphoproteomics in the last two decades renders large-scale analyses of cellular phosphorylation network possible.
Indeed, phosphopeptides can be efficiently enriched (e.g., by titanium dioxide (TiO2)) from digested total cell lysates at relatively low cost, allowing high-throughput phosphoproteome profiling by LC-MS (Olsen et al., 2010; Villén et al., 2007). Global phosphoproteomic analyses of host cells infected by bacterial pathogens were also reported (Rogers et al., 2011; Imami et al., 2013; Schmutz et al., 2013). Such phosphoproteome measurements, if designed with proper controls, can be exploited as a universal and powerful approach to identify host enzymatic substrates of a given effector protein with either kinase or phosphatase activities. For example, enteropathogenicE. coli (EPEC) effectors NleH1/2 are Ser/Thr protein kinases and they have been shown by a number of studies to attenuate NF-κB activation (Gao et al., 2009; Royan et al., 2010; Wan et al., 2011) and inhibit host apoptotic pathway during infection (Hemrajani et al., 2010). Nevertheless, no phosphorylated substrates of NleH1/2 were reported until a recent study in which a large-scale phosphoproteomic screen successfully identified host microvillus protein Eps8 as the kinase target (Pollock et al., 2022). Phosphorylation of Eps8 on Ser775 inhibits its bundling activity, leading to its dispersal from the attaching and effacing (AE) lesion during EPEC infection. This work exemplifies the utility of phosphoproteomics in identifying enzymatic substrates of bacterial kinases.