CRISPR (clustered regularly interspaced short palindromic repeats)–Cas (CRISPR-associated) systems are known to mediate bacterial defence against foreign nucleic acids. We recently demonstrated a non-canonical role for a CRISPR–Cas system in controlling endogenous gene expression, which had not previously been appreciated. In the present article, we describe the studies that led to this discovery, beginning with an unbiased genome-wide screen to identify virulence genes in the intracellular pathogen Francisella novicida. A gene annotated as encoding a hypothetical protein, but which we now know encodes the Cas protein Cas9, was identified as one of the most critical to the ability of F. novicida to replicate and survive during murine infection. Subsequent studies revealed a role for this protein in evasion of the host innate immune response. Specifically, Cas9 represses the expression of a BLP (bacterial lipoprotein) that could otherwise be recognized by TLR2 (Toll-like receptor 2), a host protein involved in initiating an antibacterial pro-inflammatory response. By repressing BLP levels, Cas9 mediates evasion of TLR2, promoting bacterial virulence. Finally, we described the molecular mechanism by which Cas9 functions in complex with two small RNAs to target the mRNA encoding the BLP for degradation. This work greatly broadened the paradigm for CRISPR–Cas function, highlighting a role in gene regulation that could be conserved in numerous bacteria, and elucidating its integral contribution to bacterial pathogenesis.
Linkage of a CRISPR–Cas system to bacterial gene regulation, host innate immune evasion and pathogenesis
CRISPR (clustered regularly interspaced short palindromic repeats)–Cas (CRISPR-associated) systems are well established to mediate bacterial defence against foreign nucleic acids  (described in detail elsewhere in this issue of Biochemical Society Transactions). We recently demonstrated an additional and novel role for a CRISPR–Cas system in regulating endogenous gene expression in the intracellular bacterial pathogen Francisella novicida . The unexpected discovery of a requirement for a CRISPR–Cas system in bacterial virulence stemmed from an unbiased interrogation of the genes required for F. novicida replication and survival during mammalian infection . This original work, and the subsequent research uncovering the mechanism by which the CRISPR–Cas system contributes to virulence and functions at the molecular level, is reviewed in the present article.
Identification of FTN_0757 as an essential F. novicida virulence gene
Francisella species are Gram-negative bacterial pathogens that share the ability to replicate robustly within mammalian cells . Critical to this process is their escape from the phagosome after entering host cells, such that they can reach their replicative niche in the host cytosol. Key to their ability to cause disease is their evasion of the host immune response while they replicate to high numbers. Until recently, extremely little was known about which genes Francisella spp. use to facilitate these virulence attributes. We set out to identify genes essential for bacterial replication and survival in mice, as a broad step towards unravelling how Francisella spp. cause disease. Work initiated in the laboratory of Dr Denise Monack employed an in vivo genome-wide screening approach called TraSH (transposon site hybridization), which was pioneered by other groups [5,6], to address this question . A library of transposon mutants was generated and used to infect mice, and the population surviving 48 h post-infection in the spleen (output) was compared with that in the inoculum (input). Mutants absent from the output harboured transposon insertions in genes critical for bacterial replication and/or survival during in vivo infection. Using this approach, we identified 164 genes, 44 of which encoded hypothetical proteins of unknown function . Subsequent testing of single deletion mutants validated a subset of the genes identified by the screen and revealed that FTN_0757, a gene encoding one such hypothetical protein, was essential for replication and/or survival (a mutant lacking FTN_0757 was 100000-fold attenuated). Although FTN_0757 was clearly critical for F. novicida virulence, these studies, as well as the lack of characterized FTN_0757 homologues, failed to reveal how this virulence factor contributed to pathogenesis.
FTN_0757 suppresses the host inflammatory response
In order to determine how FTN_0757 contributes to disease, we tested its role in several aspects of the interaction of F. novicida with host macrophages in vitro. A mutant lacking FTN_0757 was taken up by macrophages with similar kinetics as the wild-type strain U112 . In addition, the mutant escaped from the phagosome and replicated in the cytosol similarly to wild-type . However, transcriptional profiling of the macrophage response to the wild-type and mutant strains revealed that the mutant induced a much more robust inflammatory response . A broad range of mediators of the immune response (cytokines, chemokines and proteins involved in antibacterial defence) were induced by the mutant, many of which are known to be under the control of the transcription factor NF-κB (nuclear factor κB) .
During Francisella infection, early inflammatory signalling leading to NF-κB activation is dependent on TLR2 (Toll-like receptor 2) [8–11], a host innate immune receptor activated upon sensing of BLPs (bacterial lipoproteins) [12,13]. We confirmed that the inflammatory response induced by the FTN_0757 mutant was indeed TLR2-dependent . These results provided a first clue as to why the FTN_0757 mutant is so severely attenuated during in vivo infection, since it is unable to evade the inflammatory response. These data, however, still did not elucidate why the mutant activates TLR2 to a greater extent than wild-type. Since TLR2 is stimulated in response to BLP, we extracted total membrane proteins, enriched for BLPs and quantified their levels in the wild-type and mutant strains. Strikingly, we found that the FTN_0757 mutant contained roughly two times as much BLP as wild-type , providing a rationale for its increased signalling through TLR2, and revealing a novel mechanism of innate immune evasion via suppression of BLP levels.
We next investigated using SDS/PAGE and Coomassie Blue staining whether all BLPs were produced at higher levels in the FTN_0757 mutant, or whether increased production of a subset accounted for the higher levels in this strain. A single BLP, FTN_1103, was expressed far more robustly in the FTN_0757 mutant than in wild-type . Furthermore, deletion of FTN_1103 from the FTN_0757 mutant restored BLP levels to nearly those of wild-type, confirming that this BLP was almost exclusively responsible for the increased total BLP content in the mutant. Deletion of FTN_1103 in the FTN_0757 mutant similarly abrogated the increased inflammatory response induced by this strain, and almost completely restored its virulence . These data revealed that FTN_0757 was involved, directly or indirectly, in the repression of a specific BLP and that this facilitated evasion of TLR2 and the host pro-inflammatory response.
FTN_0757 (Cas9) is part of a CRISPR–Cas system
The molecular mechanism of FTN_0757 function in repressing BLP, however, remained unclear. Genomic analysis revealed that FTN_0757 was present upstream of a CRISPR region in the F. novicida genome (Figure 1), revealing an important clue . CRISPR arrays are regions of DNA containing palindromic repeats alternating with spacer sequences complementary to foreign nucleic acids . At the time, the role of CRISPR arrays in defence against foreign nucleic acids was well-established, and this function required the action of adjacently encoded Cas proteins . Bioinformatic analysis indicated that FTN_0757 has sequence similarly to the Cas protein Cas9, the defining member of Type II CRISPR–Cas systems, and that it was encoded directly upstream of genes encoding Cas1, Cas2 and Cas4 (Figure 1). To investigate whether FTN_0757/Cas9 was working in conjunction with other components of the CRISPR–Cas system, we generated deletion strains lacking each component and tested them for levels of BLP expression and in virulence assays. Deletion of the CRISPR array, cas1, cas2 or cas4 failed, however, to recapitulate the phenotype of the FTN_0757 mutant as far as BLP overexpression, the hyperinflammatory response during macrophage infection or attenuation during in vivo infection . These results suggested that the role of FTN_0757/Cas9 (hereinafter referred to as Cas9) in virulence and subversion of the inflammatory response was not due to a canonical function involving known components of the CRISPR–Cas system.
The CRISPR–Cas region in F. novicida
These studies were later greatly aided by the identification of another small RNA in the Type II CRISPR–Cas locus of Streptococcus pyogenes, called tracrRNA (transactivating CRISPR RNA), by Dr Emmanuelle Charpentier's group . tracrRNA was shown to associate with and be essential for the processing of the CRISPR array into individual crRNAs (CRISPR RNAs) consisting of one repeat and one spacer. Dr Alain Charbit's group demonstrated the presence of tracrRNA (named ftrA at the time) in Francisella holarctica strain LVS , and subsequent work indicated that it is also encoded in F. novicida [2,17]. Deletion of this small RNA from F. novicida recapitulated all of the phenotypes of the cas9 mutant (BLP overexpression, induction of a hyperinflammatory macrophage response, attenuation in vivo) . This suggested for the first time that Cas9′s function in regulating BLP levels and virulence might require other CRISPR–Cas components.
Since tracrRNA functions in conjunction with Cas9 and crRNA to target foreign nucleic acid, but the CRISPR array was not required for repression of BLP or virulence of F. novicida, we wondered whether another small RNA was involved. Bioinformatic analysis identified a novel small RNA not previously described in a CRISPR–Cas system, which we now term scaRNA (small CRISPR–Cas-associated RNA), downstream of the CRISPR array  (Figure 1). Deletion of the scaRNA revealed that it was also essential for BLP repression, evasion of the macrophage inflammatory response and virulence during in vivo infection . This meant that at least three components (the protein Cas9 and two small RNAs) were involved in BLP repression and virulence of F. novicida.
The Cas9 regulatory complex targets an endogenous mRNA
Since Cas9, tracrRNA and scaRNA were all required for the ability of F. novicida to repress BLP levels and cause disease, we hypothesized that they may form a complex, given the precedence of such a complex containing Cas9, tracrRNA and crRNA in the targeting of foreign nucleic acids. Immunoprecipitation of Cas9 from F. novicida and subsequent qPCR (quantitative real-time PCR) for tracrRNA and scaRNA indeed revealed that the three components formed a complex (termed the Cas9 regulatory complex) . Furthermore, mutation of a single residue in a putative RNA-binding domain (arginine-rich motif) of Cas9 abrogated complex formation, as did mutation of either tracrRNA or scaRNA regions predicted to mediate their interaction . A remaining question was how this Cas9 regulatory complex mediated repression of the FTN_1103 BLP.
Quantification of FTN_1103 mRNA in wild-type and mutant strains lacking Cas9 regulatory system components indicated that FTN_1103 was repressed at the mRNA level. This suggested that the Cas9 regulatory complex could be targeting FTN_1103 mRNA in much the same way that the canonical Cas9–tracrRNA–crRNA complex targets foreign nucleic acid for degradation. In fact, monitoring of mRNA levels after employing rifampin to block de novo transcription indicated that FTN_1103 mRNA was degraded and that all three Cas9 regulatory complex components were required for this activity . In addition, the FTN_1103 transcript was even detected co-immunoprecipitating with the Cas9 regulatory complex . The next question was how FTN_1103 mRNA was recruited to the Cas9 regulatory complex for degradation? In silico analysis predicted that the tracrRNA could base-pair with the FTN_1103 transcript (Figure 2). To test this prediction, we mutated residues in the FTN_1103-targeting region of tracrRNA, which indeed abolished the repression of the FTN_1103 transcript . Together, these data revealed a novel Cas9 complex involved in regulation of an endogenous transcript, which contributes to innate immune evasion and virulence.
Model of the Cas9 regulatory complex targeting an endogenous mRNA
The Cas9 regulatory complex is induced during infection to evade the host inflammatory response
It was unclear when the Cas9 regulatory complex was active during infection. We attempted to decipher this by studying the induction of the components of the system during macrophage infection in vitro. All three Cas9 regulatory complex components were strongly induced when F. novicida localized to the macrophage phagosome, and their levels began to decrease when the bacteria reached the cytosol . Interestingly, TLR2 is recruited to the macrophage phagosome during infection , but is not present in the cytosol. These data suggested that the Cas9 regulatory complex might function specifically when the bacteria are in the phagosome to facilitate the repression of BLP at the precise time that these proteins could be sensed by TLR2. Consistent with this hypothesis, quantification of FTN_1103 mRNA during infection demonstrated that this transcript was repressed during bacterial localization to the phagosome, and that this process was dependent on Cas9, tracrRNA and scaRNA . Therefore the Cas9 regulatory complex is tightly controlled to temporally repress BLP levels, facilitating innate immune evasion and virulence (Figure 3).
The role of the Cas9 regulatory system in F. novicida pathogenesis
The elucidation of a genetic regulatory function for a Type II CRISPR–Cas system significantly expands our understanding of how these machineries contribute to bacterial physiology. Furthermore, employment of this system in controlling a mechanism of innate immune evasion highlights an important role in bacterial virulence. Interestingly, Type II CRISPR–Cas systems are highly associated with pathogens and commensals, both types of bacteria that interact with eukaryotic hosts [17,19]. This suggests that the role of the Cas9 regulatory complex in controlling virulence traits may be conserved among numerous pathogens. Supporting this possibility, we showed that Cas9 is involved in the ability of Neisseria meningitidis to attach to, invade and replicate within epithelial cells . Furthermore, Louwen et al.  recently reported similar findings in Campylobacter jejuni. It is highly likely that Cas9 contributes to these virulence traits by regulating an endogenous gene(s); however, such a target has not yet been identified in either of these pathogens. It will be very interesting to determine whether Cas9 regulates BLP targets or different types of proteins in other bacteria, as well as whether Type I and Type III CRISPR–Cas systems regulate endogenous bacterial genes. Future work to answer these and many other questions will greatly enhance our understanding of CRISPR–Cas functionality, bacterial physiology, and mechanisms of immune evasion and virulence.
CRISPR Evolution, Mechanisms and Infection: A Biochemical Society Focused Meeting held at the University of St Andrews, U.K., 17–19 June 2013. Organized and Edited by Emmanuelle Charpentier (Laboratory for Molecular Infection Medicine Sweden, Sweden), John van der Oost (Wageningen University, The Netherlands) and Malcolm White (University of St Andrews, U.K.).
This work was supported by the National Institutes of Health (NIH) [grant numbers U54-AI057157 (from the Southeast Regional Center of Excellence for Biodefense and Emerging Infections) and R56-AI87673] to D.S.W., who is also supported by a Burroughs Wellcome Fund Investigator in the Pathogenesis of Infectious Disease award. T.R.S. was supported by a National Science Foundation Graduate Research Fellowship, as well as the Achievement Rewards for College Scientists (ARCS) Foundation.