Luciano A. Marraffini

Modifying bacterial genomes with phage recombinases, a process termed recombineering, has enabled precise genome editing, but its reach is limited by host specificity: each new organism generally requires its own recombinase to be identified and optimized. Using known phage recombination genes as synteny anchors, we mined Staphylococcus genomes and recovered a diverse collection of phage recombination cassettes, spanning the RAD52-like annealase superfamily (Sak, Erf) and the RAD51/RecA-like Sak4 family. Functional testing of representatives in S. aureus identified several highly efficient systems that support recombineering. We engineered the most efficient of these into a single-plasmid tool that yields markerless editing in S. aureus in four days. Notably, where other recombineering systems rely on host-specific factors for nuclease evasion and ssDNA loading, this cassette encodes both activities itself: a predicted broad-spectrum nuclease inhibitor and a dedicated single-stranded DNA-binding protein. We therefore tested the cassette across phylogenetically distant hosts, and observed recombination in Streptococcus pyogenes, Enterococcus faecium, Francisella novicida , and Escherichia coli . These results identify a self-sufficient class of phage recombination systems and provide a portable tool for genetic manipulation in bacteria that lack established recombineering methods.

Bacteria and archaea often harbor multiple CRISPR-Cas loci to defend against mobile genetic elements. Little is known, however, about whether and how different CRISPR-Cas systems are differentially regulated, in many instances due to the impossibility of studying CRISPR immunity in native hosts. Here we investigated the regulation of the endogenous type II-B and type V-A CRISPR-Cas systems present in the opportunistic human pathogen Francisella novicida U112. We found that while the type II-B system is constitutively expressed, the type V-A system is differentially expressed at stationary phase and high cell density. We identified MtvS1 and MtvS2 as factors required for this regulation, as well as for the modulation of many additional genes in stationary phase, some of which are required for Francisella virulence. Both Francisella MtvS proteins bind to RNA polymerase. MtvS1 is predicted to interact with the β’ subunit of the RNA polymerase, and MtvS2 with multiple RNA polymerase subunits as well as MtvS1. We propose that MtvS1 and MtvS2 constitute noncanonical alternative sigma factors involved in the regulation of the expression of the type V-A CRISPR locus and other genes in Francisella . Last, we show that the MtvS1 homolog YgfB is required for expression of the type I-E CRISPR-Cas system in E. coli , a result that suggests a broader role in gene regulation for these alternative sigma factors.

Prokaryotic Type III CRISPR-Cas systems synthesize cyclic oligoadenylate (cOA) second messengers that activate CRISPR-associated Rossmann fold (CARF) immune effectors during viral infection. Here we characterize Chp1, a membrane-embedded CARF effector that assembles as a tetrameric pore harboring autoinhibitory lipid fenestrations that occlude the pore in the inactive state. cOA binding triggers a structural rearrangement that (i) closes the fenestrations, (ii) eliminates the lipid obstruction of the pore to open it, and (iii) remodels the pore entrance from hydrophobic to polar to allow ion permeation. This conformational switch triggers depolarization of the membrane of the infected cell, which enters growth arrest and becomes inhospitable for viral replication. Our results uncover a lipid-gated mechanism that expands the repertoire of CRISPR-mediated defense strategies.

SUMMARY: The most abundant prokaryotic mechanisms of defense against phage predation involve the recognition and destruction of the infecting DNA. One method of counter-defense is the incorporation of modified nucleobases into the phage genome to avoid interaction with enzymes that target the viral DNA. T-even coliphages replace cytosine with 5-hydroxymethylcytosine (5hmC) that in some cases are further decorated with glucosyl groups. To explore the diversity of immunity genes that recognize 5hmC, we infected a library of metagenomic DNA inserts from uncultured, non-sequenced soil bacteria with a mutant T4 phage that harbored only non-glucosylated 5hmC on its genome. Bacteria that resisted infection carried a DNA glycosylase, Brig3, that specifically excises 5hmC nucleobases to generate abasic sites in the phage genome and prevent viral proliferation. The crystal structure of Brig3 bound to its substrate revealed a catalytic mechanism in which the 5hmC nucleobase is flipped out of the DNA into the active site and replaced by an asparagine residue that inserts into the double helix to contact the complementary guanosine. Brig3 is encoded within an operon that also encodes BapA, a hydrolase that removes glucosyl groups from glucosyl-5hmC present in the genome of otherwise Brig3-resistant T-even phages carrying this hypermodified base. Our results uncover a defense strategy in which the combined action of BapA and Brig3 widens the immune response to restrict the infection of T-even phages with genomes that are either partially or completely glucosylated.

Clustered regularly interspaced short palindromic repeats (CRISPR)–Cas (CRISPR-associated) systems provide adaptive immunity against phage infection in prokaryotes using an RNA-guided complex that recognizes complementary foreign nucleic acids. Different types of CRISPR–Cas systems have been identified that differ in their mechanism of defense. Upon infection, type III CRISPR–Cas systems employ the Cas10 complex to find phage transcripts and synthesize cyclic oligo-adenylate (cOA) messengers. These ligands bind and activate CARF immune effectors that cause cell toxicity to prevent the completion of the viral lytic cycle. Here, we investigated two proteins containing an N-terminal haloacid dehalogenase (HAD) phosphatase domain followed by four predicted transmembrane helices and a C-terminal CARF domain. We named these proteins Chp for CRISPR-associated HAD phosphatase. We show that, in vivo, Chp localizes to the bacterial membrane and that its activation induces a growth arrest, leads to a depletion of ATP and IMP, and prevents phage propagation during the type III CRISPR–Cas response. In vitro, the CARF domain of Chp binds cyclic tetra-adenylates and the HAD phosphatase domain dephosphorylates dATP, ATP, and IMP. Our findings extend the range of molecular mechanisms employed by CARF effectors to defend prokaryotes against phage infection.

Rapid molecular assays guiding treatment of methicillin-resistant Staphylococcus aureus detect SCCmec (Xpert) or the SCCmec–orfX junction (BCID2). Sequence variation in this region can disrupt primer binding, yielding false-negative results. Investigation of a missed bloodstream infection linked escape to a CRISPR-Cas–associated SCCmec variant, leading to identification of 64 variants from 45 patients—2% of 2432 screened. Misdiagnosis was restricted to clonal complex 5, a hospital-associated lineage; 11 of 40 SCCmec/junctions evaded detection by BCID2 or Xpert. Variants had mecA instability and circulated in healthcare settings. Our findings reveal a unique escape mechanism and underscore a threat to diagnostic accuracy.

Thoeris systems use proteins with Toll/interleukin-1 receptor domains to protect prokaryotes from phage infection through the synthesis of a cyclic adenosine diphosphate ribose signalling molecule, which activates an effector that depletes the host of the essential metabolite NAD ⁺ to limit viral propagation. How infection is recognized during Thoeris immunity is not known. Here we investigate the staphylococcal Thoeris defense system, ThsA–B1–B2, and found that, upon infection, the major capsid proteins of Siphoviridae phages from serogroup B, but not A, form a complex with ThsB1 and ThsB2 to activate Thoeris defense. Thoeris cyclases from Streptococcus also recognize major capsid proteins. Our results suggest that the accumulation of capsid mutations that enable avoidance of Thoeris recognition may be an important evolutionary force behind the structural diversity of prokaryotic viruses. More broadly, given that some mammalian immune pathways contain Toll/interleukin-1 receptor domains that recognize viral structures, our findings highlight a conserved mechanism of innate antiviral immunity.

A remarkable feature of CRISPR-Cas systems is their ability to acquire short sequences from invading viruses to create a molecular record of infection. These sequences, called spacers, are inserted into the CRISPR locus and mediate sequence-specific immunity in prokaryotes. In type II-A CRISPR systems, Cas1, Cas2 and Csn2 form a supercomplex with Cas9 to integrate viral sequences. While the structure of the integrase complex has been described, a detailed functional analysis of the spacer acquisition machinery is lacking. We developed a genetic system that combines deep mutational scanning (DMS) of Streptococcus pyogenes cas genes with a method to select bacteria that acquire new spacers. Here, we show that this procedure reveals key interactions at the Cas1-Cas2 interface critical for spacer integration, identifies Cas variants with enhanced spacer acquisition and immunity against phage infection, and provides insights into the molecular determinants of spacer acquisition, offering a platform to improve CRISPR-Cas-based applications.

Type III CRISPR-Cas systems defend against viral infection in prokaryotes by using an RNA-guided complex that recognizes foreign transcripts and synthesizes cyclic oligoadenylate (cOA) messengers to activate CRISPR-associated Rossmann-fold (CARF) immune effectors. In this study, we investigated a protein containing a CARF domain–fused Toll/interleukin-1 receptor (TIR) domain, Cat1. We found that Cat1 provides immunity by cleaving and depleting oxidized nicotinamide adenine dinucleotide (NAD ⁺ ) molecules from the infected host, inducing a growth arrest that prevents viral propagation. Cat1 forms dimers that stack upon each other to generate long filaments that are maintained by bound cOA ligands, with stacked TIR domains forming the NAD ⁺ cleavage catalytic sites. Furthermore, Cat1 filaments assemble into distinct trigonal and pentagonal networks that enhance NAD ⁺ degradation. Cat1 presents an unprecedented chemistry and higher-order protein assembly for the CRISPR-Cas response.