Abstract
CRISPR-Cas systems are widespread in bacterial and archaeal genomes, and in their canonical role in phage defence they confer a fitness advantage. However, CRISPR-Cas may also hinder the uptake of potentially beneficial genes. This is particularly true under antibiotic selection, where preventing the uptake of antibiotic resistance genes could be detrimental. Newly discovered features within these evolutionary dynamics are anti-CRISPR genes, which inhibit specific CRISPR-Cas systems. We hypothesized that selection for antibiotic resistance might have resulted in an accumulation of anti-CRISPR genes in genomes that harbour CRISPR-Cas systems and horizontally acquired antibiotic resistance genes. To assess that question, we analysed correlations between the CRISPR-Cas, anti-CRISPR and antibiotic resistance gene content of 104 947 reference genomes, including 5677 different species. In most species, the presence of CRISPR-Cas systems did not correlate with the presence of antibiotic resistance genes. However, in some clinically important species, we observed either a positive or negative correlation of CRISPR-Cas with antibiotic resistance genes. Anti-CRISPR genes were common enough in four species to be analysed. In Pseudomonas aeruginosa, the presence of anti-CRISPRs was associated with antibiotic resistance genes. This analysis indicates that the role of CRISPR-Cas and anti-CRISPRs in the spread of antibiotic resistance is likely to be very different in particular pathogenic species and clinical environments.
This article is part of a discussion meeting issue ‘The ecology and evolution of prokaryotic CRISPR-Cas adaptive immune systems’.
Keywords: CRISPR-Cas, acquired antibiotic resistance gene, anti-CRISPR
1. Introduction
Horizontal gene transfer (HGT) plays a crucial role in introducing genetic diversity into bacteria that can increase fitness. However, there can be a burden of increased genetic content when the encoded functions offer no benefit. A wide range of commensal and environmental bacteria, as well as mobile genetic elements (MGEs), form a reservoir of antibiotic-resistant genes (ARGs) and HGT permits the spread of ARGs between species [1]. Some of these resistance genes are integrated into the main bacterial chromosomes (figure 1). These elements encoding ARGs may be transferred by transformation, conjugation or transduction [2–4]. On the other hand, bacteria have evolved an arsenal of defence mechanisms that restrict both HGT and bacteriophage infection and provide genome protection (figure 1). These include established systems, e.g. restriction-modification systems and CRISPR-Cas, and a range of recently discovered systems [5,6]. However, the role of these systems and antibiotic resistance is poorly understood [5,7].
Figure 1.
Anti-CRISPR mediated survival model. (a) A bacterium with a chromosome encoding: CRISPR-Cas, an antibiotic target (AT), and an antibiotic resistance gene (ARG2) is shown. When a phage infects, the CRISPR-Cas system may prevent this. However, CRISPR-Cas may also hinder the acquisition of other DNA; e.g. antibiotic resistance genes (ARG1, shown here on a plasmid). (b) In the presence of an antibiotic for which there is an ARG gene (ARG3), four different possibilities are shown, with their fitness change in response to the antibiotic or phage pressure indicated by up or down arrows. (i) Die: death due to antibiotic sensitivity; (ii) target mutation: mutation of the AT will increase fitness under antibiotic selection; (iii) CRISPR failure: this could be by mutation/loss of the CRISPR-Cas system; this could facilitate ARG uptake, but likely reduce fitness under phage predation. Individuals that acquire spacers that target the ARG3-containing element would also be selected against. (iv) Anti-CRISPR acquisition: acr genes are gained from a MGE that will block/reduce the CRISPR-Cas activities and thus facilitate uptake of resistance genes (ARG3).
CRISPR-Cas systems are now well-known, sophisticated adaptive immune systems that provide heritable immunity against viruses [8], plasmids [9] and other MGEs [4]. These defence systems are abundant in both bacteria and archaea [10]. CRISPR-Cas systems hijack short segments of the intruders’ DNA and incorporate them into CRISPR loci to provide a molecular memory of past infections (figure 1) [11]. These segments of intruder DNA in the CRISPR loci are termed spacers and their complementary target sequences are known as protospacers. The CRISPR arrays are transcribed, processed into short CRISPR RNAs (crRNAs) and, through the formation of Cas ribonucleoprotein complexes and target recognition, lead to degradation of the invading element [12,13].
CRISPR-Cas systems in bacteria and archaea have spacers targeting phages, but also against other MGEs (figure 1) [14,15]. CRISPR-Cas-mediated immunity provides clear advantages during defence against virulent phages [14]. Sometimes, under high infection pressure, there is a greater advantage to alternative defence strategies, such as surface receptor mutation [16]. However, reducing potentially beneficial HGT increases the cost of maintaining CRISPR-Cas and thus might impose selective pressure to lose these systems [17,18], especially under a strong selection pressure to acquire an ARG (figure 1, ARG2). CRISPR-Cas may be lost under antibiotic selection pressure, when they may impede acquisition of ARGs [4,9,19,20]. Regain of CRISPR-Cas systems is possible though HGT [4,13]. Indeed, CRISPR-Cas systems can be found on MGE [21] and can also be transferred by generalized transduction [4]. However, some species have mechanisms to prevent this uptake of Cas proteins [22]. Phages may also be dependent on plasmid-encoded proteins, which further complicates the dynamics between phages, CRISPR-Cas systems and MGEs [23,24].
Protospacer sequences from non-viral MGEs are found integrated into CRISPR loci [25–28]. However, recent studies have failed to demonstrate a reduction by CRISPR-Cas of HGT over evolutionary periods, and have suggested balanced inhibitory effects of CRISPR-Cas on HGT [10,23,24,29]. In addition, it is likely that although functional CRISPR-Cas systems exist in some clades, e.g. Escherichia and Salmonella, they do not show much evidence of dynamically adapting [30–32] and different systems may have different roles in HGT and phage defence [13]. Surprisingly, CRISPR-Cas may increase HGT under competing selection pressures. For example, in Pectobacterium atrosepticum, CRISPR-Cas has recently been shown to enhance acquisition of chromosomal, genomic island and plasmid DNA by generalized transduction by providing protection against wild-type MGE infection [4].
To counteract CRISPR-Cas systems, viruses have evolved anti-CRISPR systems. At least 24 different families of anti-CRISPR (Acr) proteins, with the ability to block different CRISPR-Cas systems, have been identified in MGEs, viruses, bacterial and archaeal genomes [33–37]. These Acr proteins were first identified in the human pathogen Pseudomonas aeruginosa when prophages with an anti-CRISPR locus licensed a superinfecting phage to infect the lysogenized host [38].
The potential of CRISPR-Cas immunity to hinder the uptake of beneficial mobile elements (e.g. ARGs) has been widely discussed [9,13,39–41]. In a few species, there is evidence that CRISPR-Cas hinder HGT: Enterococcus faecalis [42], Streptococcus pneumoniae and Staphylococcus aureus [10]. In addition, Van Belkum et al. [7] analysed the CRISPR-Cas systems of 672 P. aeruginosa strains and correlated their presence or absence with antibiotic resistance. Although they observed an association for a few specific resistance genes (e.g. a negative correlation with sul1 for sulfonamide resistance) their data were more consistent with CRISPR-Cas systems acting to provide phage resistance rather than blocking the transfer of resistance genes directly.
We were motivated to investigate the association between CRISPR-Cas, acquired resistance genes and acr in clinically important bacterial species. For this purpose, we analysed the CRISPR-Cas, anti-CRISPRs and acquired ARGs of over 100 000 RefSeq bacterial genomes. We demonstrate that in most species there is little correlation between CRISPR-Cas and ARGs. However, in several bacterial species, the presence of CRISPR-Cas is inversely correlated with ARGs. Importantly, we report a positive association between anti-CRISPR genes and acquired antibiotic resistance in some species. In these cases, we propose that anti-CRISPR has aided HGT.
2. Methods
(a). Genomic data
Reference genomes of bacterial species were obtained from RefSeq86 (March 2018). These genomes are considered complete or draft by RefSeq but they are assembled to different levels, contig (50 819), scaffold (45 229), complete genome (7621) or chromosome (1278). For many species, there were multiple strains sequenced, with 39 having 250 strains, which are often different clinical isolates. RefSeq labels one strain from each species as a reference (138) or representative (5539). This is usually only one per species, but in a few cases there are more than one, particularly where several strains are frequently used as a reference (e.g. both P. aeruginosa PAO1 and PA14). The sequences from each of these bacterial strains contain contigs or scaffolds of the chromosomes and may also contain contigs or scaffolds from episomes.
(b). Identification of acquired ARGs and point mutations
The command line version of ResFinder (v 2.1) with default parameters was used: greater than 90% identity over 60% of the gene [43]. This version of ResFinder identifies acquired antimicrobial resistance genes where resistance is conferred by a complete gene. The ResFinder database is composed of 1649 acquired ARGs from 15 antibiotic drug classes. It does not identify resistance conferred by point mutations. As a control, in order to analyse the correlation between CRISPR-Cas and resistance mediated by point mutations, we analysed point mutations conferring in the polymyxin, fluoroquinolone and efflux pump genes. The command line version of Resistance Gene Identifier (RGI v 3.2.1) with default parameters was used to identify point mutations [44].
(c). Identification of anti-CRISPR genes
Twenty-four families of anti-CRISPR genes were analysed; 411 members of these families, from anti-CRISPRdb [45], were used to search the translated genomes by tblastn. All significant hits were collected and then filtered by using an e-value cutoff 10−10. Acr proteins are short (less than 120 amino acids), so this corresponded to a bit score cutoff of approximately 200. This stringent cutoff was chosen to avoid false positives, and identified 2821 matches in the genomes. Outside of the three species that were known to have Acr, and Ralstonia solanacearum, there were less than 20 (usually only one or two acr) found in each of 156 genomes. These small numbers were not used in the correlation analysis.
(d). Identification of CRISPR arrays
The corresponding CRISPR-Cas information was taken from precomputed CRISPR array files generated by CRISPRDetect [46] and NCBI RefSeq GFF files [47,48]. To determine the presence of cas genes, we used the RefSeq genome annotation, which uses the classification in Biswas et al. [49]. Then, we searched for the presence of CRISPR arrays using CRISPRDetect, with the default array-quality score cutoff of 4.0. Specific species were also analysed with CRISPRCasFinder [50].
(e). Identification of CRISPR spacer targets
To identify antibiotic resistance genes that are likely to be targeted by CRISPR-Cas systems, all CRISPR spacers from CRISPR arrays predicted by CRISPRDetect with an array-quality score above 4.0 were selected, and clustered using CD-Hit-EST with an identity-cutoff of 0.95, to produce a non-redundant FASTA file containing only the spacers at the cluster centres. The BLAST databases of DNA sequences of known classes of antibiotic resistance genes were from ResFinder's data-source. With these data, known ARGs that are targeted by spacers from a species of interest were detected using BLASTN (blast-short mode, word-size 7 and e-value cutoff 0.01). All hits with a bit score above 31 were selected. The targets were also analysed by using the default parameters CRISPRTarget [49].
(f). Statistical analysis
For each genome, the presence of a CRISPR-Cas system, an anti-CRISPR gene (acr) and an ARG of interest are all coded as binary variables. The RGI outputs were also transformed into binary variables.
For each species, the total number of genomes (N), and the number of genomes in which an anti-CRISPR gene and an ARG of interest are both observed (O), were counted. Then, the expected number of genomes (E) possessing both an anti-CRISPR gene and the ARG totally by chance were estimated (i.e. E = N×Pr(ARG) × Pr(acr), where Pr(.) means probability). The log frequency-ratios (log(O/E)) were computed to measure the association of anti-CRISPR and the ARG. To avoid dividing-by-zero or log of zero, a pseudocount of 1 × 10−6 was added to both the O and the E. A positive log frequency-ratio indicates a positive association, that the ARG tend to coexist with an anti-CRISPR gene. Otherwise, a negative association is observed, indicating that the presence of an anti-CRISPR gene tends to exclude the ARG. The 99.9% confidence intervals of the log frequency-ratios were computed by bootstrapping with 5000 replicates of N samplings with replacement. The association of the lack of CRISPR-Cas with the presence of an ARG, and antibiotic resistance related point mutations of interest, were also tested using this procedure. All of our data are available through github (https://github.com/davidchyou).
3. Results and discussion
(a). Anti-CRISPRs are present in genomes with or without CRISPR-Cas systems
To determine the association between CRISPR-Cas systems, anti-CRISPR (acr) genes and acquired ARGs, all 104 947 unique bacterial genomes from the NCBI RefSeq Genomes database were analysed. All genomes were searched for CRISPR arrays, cas genes, ARGs and acr genes (figure 2). This complete dataset includes many genomic sequences for some species; for example, there are more than 2000 P. aeruginosa genomes of different strains, both clinical and environmental isolates. Therefore, genomes of representative bacterial strains (5677), generally one from each species from the dataset, were also analysed. Both CRISPR arrays and cas genes were found in 34% (35 569) of all and 41% (2330) of the representative genomes (figure 2 and electronic supplementary material, table S1). This proportion is consistent with previous studies using our own or other methodologies [51–54]. Only genomes with both CRISPR arrays and cas genes were further considered; orphan arrays and genes were considered to be likely non-functional systems or false positives [46,55].
Figure 2.
Distribution of CRISPR, cas and anti-CRISPR in (a) the whole dataset (n = 104 947) and (b) representative bacterial genomes (n = 5677) from Refseq86 (58 960 of 104 947 and 2759 of 5677 contain neither CRISPR-Cas nor anti-CRISPRs). The proportions are provided in electronic supplementary material, table S1.
Of the 2330 representative bacterial genomes with CRISPR-Cas systems, anti-CRISPR (acr) genes were also identified in 92 of them (figure 2). In these bacteria, CRISPR-Cas systems are potentially impaired, which could result in an unstable state where the ineffective (or partially effective) CRISPR-Cas system would provide little advantage. Forty genomes had acr genes with neither CRISPR arrays nor cas genes (figure 2).
(b). Species with CRISPR-Cas vary in acquired antibiotic resistance genes
Bacteria become antibiotic resistant either by acquiring complete resistance genes and/or through mutations in their genome. Acquisition of resistance genes can be inhibited by CRISPR-Cas systems, whereas the emergence of resistance through mutation is expected to be unaffected. We tested whether there was a correlation between the presence of CRISPR-Cas and acquired ARGs. Only 15% (868 of 5677) of representative genomes and 19% (20 014 of 104 947) of RefSeq genomes contained ARGs that would confer resistance to at least one of the 15 major drug classes. To ensure adequate sample size for statistical analysis, only species for which there are over 250 genomes, and between 5% and 95% of genomes had CRISPR-Cas, were considered. The association of CRISPR-Cas with ARG classes was tested for 39 species. The log frequency-ratios (log(O/E)) are shown as a heatmap in figure 3, as is the percentage of strains with CRISPR-Cas; the raw data are in electronic supplementary material, table S2.
Figure 3.
Associations between the presence of CRISPR-Cas and acquired antibiotic resistance genes (ARGs) in genomes. The heatmap shows the correlation between the presence of CRISPR-Cas and acquired ARGs in the genomes of 39 species for 15 drug classes. The log frequency-ratio (log(Observed/Expected)) was calculated to identify possible association (Methods). Because the variable used is a binary indicator for the lack of CRISPR-Cas, a log(O/E) greater than 0 indicates a negative association between CRISPR-Cas and the ARGs (red shades), i.e. they are found together less often than expected. A value less than 0 indicates that the ARGs are more common in the presence of CRISPR-Cas (blue shades). Black borders indicate statistically significant associations, where 99.9% confidence intervals (CI) do not include zero (Methods), and a log(O/E) greater than 0.2 in magnitude. The final column indicates the percentage of strains with CRISPR-Cas.
For many tests of association between resistance class and CRISPR-Cas, the 99.9% confidence interval included zero and was not significant (grey or pale red and blue boxes without borders in figure 3). However, there were some significant negative associations, e.g. for E. faecalis CRISPR-Cas was negatively correlated with beta-lactam and glycopeptide resistance (red shades, figure 3). These results are consistent with the idea that CRISPR-Cas limits HGT in E. faecalis [42].
Conversely, a positive association between CRISPR-Cas and some drug classes was detected in some species, e.g. Neisseria meningitidis, and tetracycline resistance (blue shades, figure 3). These positive associations of CRISPR-Cas and ARGs might be due to circumstances under which CRISPR-Cas can enhance HGT via transduction [4].
(c). Species in which CRISPR-Cas may limit horizontal gene transfer
(i). Antibiotic-resistant Enterococcus faecalis strains have fewer CRISPR-Cas systems
Previous studies on 16 E. faecalis genomes have shown that there was a negative correlation between CRISPR-Cas and antibiotic resistance [42]. In agreement, our more extensive analysis demonstrated that about two-thirds of E. faecalis (338 of 514) genomes lacked CRISPR-Cas systems. These 338 genomes contained multiple ARGs, conferring resistance to multiple drug classes (electronic supplementary material, table S2). Palmer et al. found CRISPR-Cas loci were uniformly absent in those lineages with beta-lactam and vancomycin resistance. We also observed a strong correlation (figure 3) between the absence of CRISPR-Cas immunity and the presence of genes for beta-lactam and glycopeptide resistance (that are mainly vancomycin resistance genes; electronic supplementary material, table S5). Therefore, the absence of CRISPR-Cas in some strains of E. faecalis may increase fitness in acquiring ARGs and facilitating hospital adaptation as previously suggested [42,56]. The variable CRISPR-Cas content of E. faecalis strains may be due to events that have led to the loss of CRISPR-Cas (figure 1). Indeed, many of the genomes without CRISPR and cas had partial systems, either orphan arrays (136 of 338) or Cas proteins (32 of 338). This suggests that loss of the cas genes has preferentially occurred. In other systems the cas genes are the greater burden than the CRISPR arrays [26]. For another Enterococcus with many sequenced genomes, E. faecium, only 31 of 715 genomes had CRISPR-Cas; none of these have glycopeptide resistance, compared to 65% overall. This is consistent with the notion that these ARGs and CRISPR-Cas are also incompatible in E. faecium.
(ii). Pseudomonas aeruginosa strains with CRISPR-Cas have fewer antibiotic resistance genes
A previous study examined the relationship between CRISPR-Cas and ARGs in 672 clinical isolates of P. aeruginosa [7]. In our study, of the 2021 P. aeruginosa genomes, 95% (1922) contained one or more acquired ARGs, whereas 48% (961) had CRISPR-Cas (figure 3). The assemblies commonly contained resistance genes for aminoglycosides (1921), beta-lactams (1922), fosfomycins (1914), phenicols (1892) and sulfonamides (528). About 28% (148 of 528) of genomes with sulfonamide resistance genes (sul1, sul2 or sul3) had CRISPR-Cas (figure 3), compared with 48% overall. The lower proportion of sulfonamide resistance in strains harbouring CRISPR-Cas agrees with the earlier work on a clinical set [7]. Although other ARGs were detected, they were infrequent—notably colistin (1), macrolide (22), quinolone (9) and rifampicin (18) ARGs. These ARGs were also inversely correlated with CRISPR-Cas in P. aeruginosa (pale reds, figure 3), but due to smaller numbers, these were not statistically significant.
Resistance against other antibiotics had no association with CRISPR-Cas in P. aeruginosa. For example, the frequent aminoglycoside, beta-lactam and phenicol ARGs (listed above), were not associated with CRISPR-Cas. As a control we analysed point mutations conferring resistance to these antibiotic classes and, as expected, found no specific association between CRISPR-Cas and ARGs (Methods; electronic supplementary material, table S3). Overall, these data suggest that in P. aeruginosa, CRISPR-Cas may effectively target MGEs containing sul genes (Blastn results electronic supplementary material, table S4). These are commonly found in larger pathogenicity islands in P. aeruginosa [57,58].
It had previously been reported that there was no evidence of CRISPR spacers directly targeting ARG genes in P. aeruginosa. A search of the 564 non-redundant spacers from P. aeruginosa revealed 33 spacers with high-scoring matches to resistance genes in ResFinder's reference ARG (electronic supplementary material, tables S4 and S5). This is consistent with experimental laboratory studies that showed that spacers can be acquired from ARGs and impede their subsequent uptake [12,59]. To be functional for interference, the type I systems of P. aeruginosa require protospacer adjacent motifs (PAMs). PAMs were found for some targets within ARGs (e.g. blaBEL, blaLEN, blaOXA, aph, tet(V), etc.) associated with beta-lactam, aminoglycoside and tetracycline resistance (electronic supplementary material, figure S1 and table S6); these would be expected to functionally target ARGs. However, many targets lacked consensus PAMs, which indicates they are no longer functional for interference, but it is possible that due to sequence variation, the genes targeted in nature have PAMs at these positions (figure 1). Moreover, any other part of the MGE could be targeted by the CRISPR-Cas system.
(iii). Species in which CRISPR-Cas are rare but resistance is common
Some pathogens have few sequenced strains with CRISPR-Cas. For example, in this set of strains of Staphylococcus aureus, only 0.6% (45 of 7865) have CRISPR-Cas systems, and this rarity has been noted by others [60]. These S. aureus strains are mainly clinical isolates and it is expected that they will acquire multi-drug resistance through HGT mechanisms, such as acquiring plasmids containing ARGs [9,39]. In these strains, the low prevalence of CRISPR-Cas makes associations statistically insignificant (electronic supplementary material, table S2). However, 94% (7420 of 7865) are resistant to at least one class of drug and it is tempting to speculate that selection for antibiotic resistance has counter-selected CRISPR-Cas [19].
Similar to staphylococci, few (6 of 7389) Streptococcus pneumoniae strains have CRISPR-Cas. Indeed, to our knowledge, this is the first report of CRISPR-Cas systems in S. pneumoniae [61]. CRISPRCasFinder [50] also confirmed the presence of CRISPR-Cas systems in two genomes. These were type II-C based on the presence of cas1, cas2 and cas9 genes. These few S. pneumoniae strains containing CRISPR-Cas lacked ARGs. Other species also had less than 5% CRISPR-Cas, e.g. Enterococcus faecium (4%), Burkholderia cenocepacia (5%). Overall, in these bacterial species with few CRISPR-Cas systems, acquired ARGs are frequent.
Conversely, species in which almost all (greater than 95%) genomes contained CRISPR-Cas and thus almost always associated with ARGs could not be analysed for association. Salmonella enterica (97%), Campylobacter jejuni (96%), Yersinia pestis (96%), Streptococcus agalactiae (99%) and Shigella sonnei (98%) were in this category. However, it is clear that in these organisms CRISPR-Cas coexist with ARGs.
(d). Species in which CRISPR-Cas is positively associated with ARGs
Over three quarters of the E. coli genomes have CRISPR-Cas systems (76%; 5164 of 6800), and 2638 of 6800 genomes have ARGs for at least one drug class. We observed that genes associated with fosfomycin (244/259) and rifampicin (91/103) resistance genes are more common in those with CRISPR-Cas systems (blue, figure 3). This supports the idea that CRISPR-Cas systems in these organisms do not hinder the dissemination of resistance as reported previously [13,30–32].
By contrast, in E. coli 30% of genomes with resistance genes associated with beta-lactam, quinolone, macrolide and trimethoprim do not have CRISPR-Cas systems (1706 of 5661). This may indicate selection due to antibiotic exposure and the need to acquire ARGs through HGT, and loss of CRISPR-Cas in the ancestors of these strains [13] (figure 1; electronic supplementary material, table S2). Taken together this may further indicate that other factors may be contributing to the different antibiotic resistance and CRISPR-Cas distributions in these strains [13,30].
One quarter of Klebsiella pneumoniae genomes harbour CRISPR-Cas systems (26%; 650 of 2646). Strains of this species with CRISPR-Cas had higher than expected numbers of tetracycline resistance genes (figure 3), whereas a recent study on 176 K. pneumoniae strains has shown that isolates with subtype I-E CRISPR-Cas systems were more susceptible to aminoglycoside and beta-lactam, compared to CRISPR-negative counterparts, and also have less acquired ARGs in their genomes [62]. However, we did not observe any negative correlation between CRISPR-Cas and any ARGs (figure 3). Perhaps, the negative effect of CRISPR-Cas and these drugs is most significant in a clinically focused dataset, where strains were exposed to these antibiotics [62]. However, our data demonstrates the association may not be generalized for a large and diverse range of K. pneumoniae strains, where there may be a competing need of viral defence that benefits from CRISPR-Cas systems.
In Neisseria meningitidis and Vibrio cholerae, we also observed that some ARGs were positively associated with CRISPR-Cas (figure 3, blue squares). Tetracycline resistance occurred in most N. meningitidis strains with CRISPR-Cas (99%, 119/120) but only 40% overall. Beta-lactam resistance is found in about half of V. cholerae (26/59) but CRISPR-Cas in only 16%. It is possible that in these two cases the CRISPR-Cas systems may be facilitating the uptake of ARGs. Previously, natural transformation was shown to be blocked by the CRISPR-Cas system in N. meningitidis [4,63]. Resistance to beta-lactam drugs has been identified in Vibrio phage genomes [64,65] and a phage-encoded CRISPR-Cas system [66] has been found in Vibrio. Therefore, the gain of ARG and role of CRISPR-Cas through HGT is more complicated in Vibrio. However, it is also possible that the apparent positive effects of CRISPR-Cas on ARGs in these strains might be due to the increased acquisition of these ARGs via transduction [4,67,68].
(e). Anti-CRISPRs and antibiotic resistance
Recently, anti-CRISPR systems have been discovered. The acr genes encode a diverse range of small proteins that block specific CRISPR-Cas systems, and can be found in phages, prophages, plasmids and also in bacterial genomes [38]. The Acr proteins prevent CRISPR-Cas systems from functioning during interference. Therefore, Acrs can permit prophage integration in the presence of phage-targeting spacers, and also allow survival of strains that have acquired self-targeting spacers (figure 1). Thus, continued expression of Acr proteins becomes important for the host, because reduced Acr expression may result in unwanted genomic cleavage by the bacterial CRISPR-Cas system, known as ‘self-targeting’ [69]. We predicted that anti-CRISPR proteins would tend to coexist with ARGs in antibiotic-resistant bacteria by reducing the ability of CRISPR-Cas systems to prevent ARG acquisition. The association between CRISPR-Cas, ARGs and acr genes was tested on four species where a sufficient number of strains contained acr genes (Methods, electronic supplementary material, table S9; figure 4).
Figure 4.
Associations between anti-CRISPR and antibiotic resistance genes. The heatmap shows the correlation of anti-CRISPRs and acquired ARGs in four organisms for 15 drug classes. The log frequency-ratio log(O/E) was calculated to identify the association between anti-CRISPR and ARGs. A positive value of log(O/E) indicates strong association between them, whereas a negative value demonstrates that the ARGs are more common in the absence of anti-CRISPR. The coloured boxes with black borders indicate significant associations (as in figure 3).
(i). Anti-CRISPR homologues found in plant pathogen Ralstonia solanacearum
Interestingly, we commonly found AcrIIC2 homologues (e-value: <e−10, identity greater than 26% identity over 130 amino acids, and bit score greater than 38) in R. solanacearum (60/70), a plant pathogen causing Moko disease of banana and brown rot of potato [70]. Acr have been previously characterized in the plant pathogen Pectobacterium atrosepticum [71]. We did not observe any ARGs within this species, but 19 of 70 strains have CRISPR-Cas (electronic supplementary material, tables S2 and S7). However, all these systems were type I-E CRISPR-Cas (containing cas3, cas1, cas2, cas5, cas6, cas7, cse1 and cse2). In the absence of a detectable Type II-C CRISPR system, possibly those strains have acquired anti-CRISPRs horizontally as part of other elements and they have no target in R. solanacearum. Therefore, as expected in R. solanacearum there is no association between the presence of acr (AcrIIC2) and CRISPR-Cas (Type I-E).
(ii). Anti-CRISPR genes co-occur with acquired resistance in Pseudomonas aeruginosa
Previous studies have shown the presence of anti-CRISPRs in multi-drug resistant P. aeruginosa [7], but they did not test for any correlation with acr genes. Around half of the P. aeruginosa genome assemblies had acr genes (46%; 924 of 2021). Consistent with the idea that acr could facilitate ARG uptake (figure 1), there was a significant positive association between anti-CRISPRs and ARGs in P. aeruginosa for aminoglycoside, beta-lactam, fosfomycin and phenicol drug classes (figure 4), with anti-CRISPR, ARG and CRISPR-Cas co-occurring in greater than 460 strains (electronic supplementary material, table S8). The correlation between ARGs and anti-CRISPRs is consistent with the observation that many P. aeruginosa strains have CRISPR-Cas systems in their genome. Moreover, we observed some of the genes conferring beta-lactam or aminoglycoside resistance could be targeted by P. aeruginosa CRISPR spacers (see above). Taken together, our analyses indicate that inhibition of CRISPR-Cas systems by anti-CRISPRs might facilitate the uptake of ARGs in P. aeruginosa (figure 1). We did not observe any significant association for Listeria monocytogenes, but did see a significant negative correlation between acr and beta-lactam and tetracycline ARGs in N. meningitidis.
4. Conclusion
CRISPR-Cas enables bacteria to protect against invading viruses and other MGEs. However, in different species or environments the selective pressures on CRISPR-Cas will vary, and antibiotic usage may be a contributing factor. Some bacterial species completely lack CRISPR-Cas systems, some have it rarely, some retain these immune systems but inactivate them through anti-CRISPRs (figure 1)—we find evidence for all of these in different species (figure 3). Although anti-CRISPRs may have evolved to provide virulent and temperate phages with an advantage over CRISPR-Cas, they could act in multiple ways during bacterial evolution [72]. While anti-CRISPRs may reduce CRISPR-Cas defence against phages, this may enable the benefits of acquisition of ARGs. Previous studies on other MGE counter defence proteins, such as anti-restriction proteins, or ArdA from conjugative plasmids and transposons, also demonstrated a similar inhibition of the dissemination of antibiotic resistance in different species [73,74]. In P. aeruginosa, we observed a significant influence of anti-CRISPRs on the prevalence of antibiotic resistance (figure 4).
Under antibiotic selection, individual cells will die if they insert a spacer causing an MGE to be targeted. Therefore, acquisition of ARGs in active CRISPR-Cas containing strains is possible only in cells without MGE targeting spacers. A deeper analysis of CRISPR spacers versus specific ARG-disseminating elements could reveal why CRISPR-Cas, though a seemingly powerful defence against foreign DNA, does not seem to have a big effect on transfer of antibiotic resistance genes. Our analysis is currently limited to known anti-CRISPRs proteins. However, we are likely underestimating the effect of anti-CRISPR, because there are surely more ACR families awaiting discovery.
This study analysed multiple interacting systems across a broad range of species and large numbers of strains within them. It is perhaps not surprising that a simple rule of selection pressure does not apply to all bacterial strains. However, we find patterns emerging within particular species and antibiotic classes.
Supplementary Material
Supplementary Material
Supplementary Material
Supplementary Material
Supplementary Material
Supplementary Material
Supplementary Material
Supplementary Material
Supplementary Material
Supplementary Material
Supplementary Material
Supplementary Material
Data accessibility
Additional data is available at https://github.com/davidchyou.
Competing interests
We declare we have no competing interests.
Funding
This work was supported by a University of Otago Postgraduate Scholarship to S.S. and University of Otago Research Grant to C.M.B.
References
- 1.Baltrus DA. 2013. Exploring the costs of horizontal gene transfer. Trends Ecol. Evol. 28, 489–495. ( 10.1016/j.tree.2013.04.002) [DOI] [PubMed] [Google Scholar]
- 2.Fineran PC, Petty NK, Salmond GPC. 2009. Transduction: host DNA transfer by bacteriophages. In Encyclopedia of microbiology (ed. Schaechter M.), pp. 666–679. Oxford, UK: Elsevier. [Google Scholar]
- 3.Maniv I, Jiang W, Bikard D, Marraffini LA. 2016. Impact of different target sequences on Type III CRISPR-Cas immunity. J. Bacteriol. 198, 941–950. ( 10.1128/JB.00897-15) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Watson BNJ, Staals RHJ, Fineran PC. 2018. CRISPR-Cas-mediated phage resistance enhances horizontal gene transfer by transduction. MBio 9, 02406-17 ( 10.1128/mBio.02406-17) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Doron S, Melamed S, Ofir G, Leavitt A, Lopatina A, Keren M, Amitai G, Sorek R. 2018. Systematic discovery of antiphage defense systems in the microbial pangenome. Science 359, eaar4120 ( 10.1126/science.aar4120) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Ofir G, Melamed S, Sberro H, Mukamel Z, Silverman S, Yaakov G, Doron S, Sorek R. 2018. DISARM is a widespread bacterial defence system with broad anti-phage activities. Nat. Microbiol. 3, 90–98. ( 10.1038/s41564-017-0051-0) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.van Belkum A, et al. 2015. Phylogenetic distribution of CRISPR-Cas systems in antibiotic-resistant Pseudomonas aeruginosa. MBio 6, e01796-15 ( 10.1128/mBio.01796-15). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Barrangou R, Fremaux C, Deveau H, Richards M, Boyaval P, Moineau S, Romero DA, Horvath P. 2007. CRISPR provides acquired resistance against viruses in prokaryotes. Science 315, 1709–1712. ( 10.1126/science.1138140) [DOI] [PubMed] [Google Scholar]
- 9.Marraffini LA, Sontheimer EJ. 2008. CRISPR interference limits horizontal gene transfer in staphylococci by targeting DNA. Science 322, 1843–1845. ( 10.1126/science.1165771) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Gophna U, Kristensen DM, Wolf YI, Popa O, Drevet C, Koonin EV. 2015. No evidence of inhibition of horizontal gene transfer by CRISPR-Cas on evolutionary timescales. ISME J. 9, 2021–2027. ( 10.1038/ismej.2015.20) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Carter J, Hoffman C, Wiedenheft B. 2017. The interfaces of genetic conflict are hot spots for innovation. Cell 168, 9–11. ( 10.1016/j.cell.2016.12.007) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Staals RH, Jackson SA, Biswas A, Brouns SJ, Brown CM, Fineran PC. 2016. Interference-driven spacer acquisition is dominant over naive and primed adaptation in a native CRISPR-Cas system. Nat. Commun. 7, 12 853 ( 10.1038/ncomms12853) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Garcia-Martinez J, Maldonado RD, Guzman NM, Mojica FJM. 2018. The CRISPR conundrum: evolve and maybe die, or survive and risk stagnation. Microb. Cell 5, 262–268. ( 10.15698/mic2018.06.634) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Marraffini LA. 2013. CRISPR-Cas immunity against phages: its effects on the evolution and survival of bacterial pathogens. PLoS Pathog. 9, e1003765 ( 10.1371/journal.ppat.1003765) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Al-Attar S, Westra ER, van der Oost J, Brouns SJ. 2011. Clustered regularly interspaced short palindromic repeats (CRISPRs): the hallmark of an ingenious antiviral defense mechanism in prokaryotes. Biol. Chem. 392, 277–289. ( 10.1515/BC.2011.042) [DOI] [PubMed] [Google Scholar]
- 16.Chabas H, van Houte S, Hoyland-Kroghsbo NM, Buckling A, Westra ER. 2016. Immigration of susceptible hosts triggers the evolution of alternative parasite defence strategies. Proc. R. Soc. B 283, 20160721 ( 10.1098/rspb.2016.0721) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Kuo CH, Ochman H. 2009. Deletional bias across the three domains of life. Genome Biol. Evol. 1, 145–152. ( 10.1093/gbe/evp016) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Stern A, Keren L, Wurtzel O, Amitai G, Sorek R. 2010. Self-targeting by CRISPR: gene regulation or autoimmunity? Trends Genet. 26, 335–340. ( 10.1016/j.tig.2010.05.008) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Jiang W, Maniv I, Arain F, Wang Y, Levin BR, Marraffini LA. 2013. Dealing with the evolutionary downside of CRISPR immunity: bacteria and beneficial plasmids. PLoS Genet. 9, e1003844 ( 10.1371/journal.pgen.1003844) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Bikard D, Hatoum-Aslan A, Mucida D, Marraffini LA. 2012. CRISPR interference can prevent natural transformation and virulence acquisition during in vivo bacterial infection. Cell Host Microbe 12, 177–186. ( 10.1016/j.chom.2012.06.003) [DOI] [PubMed] [Google Scholar]
- 21.Koonin EV, Makarova KS. 2017. Mobile genetic elements and evolution of CRISPR-Cas systems: all the way there and back. Genome Biol. Evol. 9, 2812–2825. ( 10.1093/gbe/evx192) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Almendros C, Guzmán NM, García-Martínez J, Mojica FJ. 2016. Anti-cas spacers in orphan CRISPR4 arrays prevent uptake of active CRISPR–Cas-IF systems. Nat. Microbiol. 1, 16 081 ( 10.1038/nmicrobiol.2016.81) [DOI] [PubMed] [Google Scholar]
- 23.Gandon S, Vale PF. 2014. The evolution of resistance against good and bad infections. J. Evol. Biol. 27, 303–312. ( 10.1111/jeb.12291) [DOI] [PubMed] [Google Scholar]
- 24.Jalasvuori M, Friman VP, Nieminen A, Bamford JK, Buckling A. 2011. Bacteriophage selection against a plasmid-encoded sex apparatus leads to the loss of antibiotic-resistance plasmids. Biol. Lett. 7, 902–905. ( 10.1098/rsbl.2011.0384) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Wiedenheft B, Sternberg SH, Doudna JA. 2012. RNA-guided genetic silencing systems in bacteria and archaea. Nature 482, 331–338. ( 10.1038/nature10886) [DOI] [PubMed] [Google Scholar]
- 26.Westra ER, Buckling A, Fineran PC. 2014. CRISPR-Cas systems: beyond adaptive immunity. Nat. Rev. Microbiol. 12, 317–326. ( 10.1038/nrmicro3241) [DOI] [PubMed] [Google Scholar]
- 27.van Houte S., et al. 2016. The diversity-generating benefits of a prokaryotic adaptive immune system. Nature 532, 385–388. ( 10.1038/nature17436). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Shmakov SA, Sitnik V, Makarova KS, Wolf YI, Severinov KV, Koonin EV. 2017. The CRISPR spacer space is dominated by sequences from species-specific mobilomes. mBio 8, e01397-17 ( 10.1128/mBio.01397-17) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Vale PF, Lafforgue G, Gatchitch F, Gardan R, Moineau S, Gandon S. 2015. Costs of CRISPR-Cas-mediated resistance in Streptococcus thermophilus. Proc. R. Soc. B 282, 20151270 ( 10.1098/rspb.2015.1270) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Touchon M, Charpentier S, Pognard D, Picard B, Arlet G, Rocha EP, Denamur E, Branger C. 2012. Antibiotic resistance plasmids spread among natural isolates of Escherichia coli in spite of CRISPR elements. Microbiology 158, 2997–3004. ( 10.1099/mic.0.060814-0) [DOI] [PubMed] [Google Scholar]
- 31.Touchon M, Charpentier S, Clermont O, Rocha EP, Denamur E, Branger C. 2011. CRISPR distribution within the Escherichia coli species is not suggestive of immunity-associated diversifying selection. J. Bacteriol. 193, 2460–2467. ( 10.1128/JB.01307-10) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Shariat N, Timme RE, Pettengill JB, Barrangou R, Dudley EG. 2015. Characterization and evolution of Salmonella CRISPR-Cas systems. Microbiology 161, 374–386. ( 10.1099/mic.0.000005) [DOI] [PubMed] [Google Scholar]
- 33.Bondy-Denomy J. 2018. Protein inhibitors of CRISPR-Cas9. ACS Chem. Biol. 13, 417–423. ( 10.1021/acschembio.7b00831) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Hynes AP, et al. 2018. Widespread anti-CRISPR proteins in virulent bacteriophages inhibit a range of Cas9 proteins. Nat. Commun. 9, 2919 ( 10.1038/s41467-018-05092-w). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Zhu Y, Zhang F, Huang Z. 2018. Structural insights into the inactivation of CRISPR-Cas systems by diverse anti-CRISPR proteins. BMC Biol. 16, 32 ( 10.1186/s12915-018-0504-9) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Pawluk A, Davidson AR, Maxwell KL. 2018. Anti-CRISPR: discovery, mechanism and function. Nat. Rev. Microbiol. 16, 12–17. ( 10.1038/nrmicro.2017.120) [DOI] [PubMed] [Google Scholar]
- 37.Guo T, Han W, She Q. 2018. Tolerance of Sulfolobus SMV1 virus to the immunity of I-A and III-B CRISPR-Cas systems in Sulfolobus islandicus. RNA Biol. 1–8. ( 10.1080/15476286.2018.1460993) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Bondy-Denomy J, Pawluk A, Maxwell KL, Davidson AR. 2013. Bacteriophage genes that inactivate the CRISPR/Cas bacterial immune system. Nature 493, 429–432. ( 10.1038/nature11723) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Gholizadeh P, Aghazadeh M, Asgharzadeh M, Kafil HS. 2017. Suppressing the CRISPR/Cas adaptive immune system in bacterial infections. Eur. J. Clin. Microbiol. Infect. Dis. 36, 2043–2051. ( 10.1007/s10096-017-3036-2) [DOI] [PubMed] [Google Scholar]
- 40.Gudbergsdottir S, Deng L, Chen Z, Jensen JV, Jensen LR, She Q, Garrett RA. 2011. Dynamic properties of the Sulfolobus CRISPR/Cas and CRISPR/Cmr systems when challenged with vector-borne viral and plasmid genes and protospacers. Mol. Microbiol. 79, 35–49. ( 10.1111/j.1365-2958.2010.07452.x) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Sapranauskas R, Gasiunas G, Fremaux C, Barrangou R, Horvath P, Siksnys V. 2011. The Streptococcus thermophilus CRISPR/Cas system provides immunity in Escherichia coli. Nucleic Acids Res. 39, 9275–9282. ( 10.1093/nar/gkr606) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Palmer KL, Gilmore MS. 2010. Multidrug-resistant enterococci lack CRISPR-cas. MBio 1, e00227-10 ( 10.1128/mBio.00227-10) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Zankari E, Hasman H, Cosentino S, Vestergaard M, Rasmussen S, Lund O, Aarestrup FM, Larsen MV. 2012. Identification of acquired antimicrobial resistance genes. J. Antimicrob. Chemother. 67, 2640–2644. ( 10.1093/jac/dks261) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.McArthur A.G., et al. 2013. The comprehensive antibiotic resistance database. Antimicrob. Agents Chemother. 57, 3348–3357. ( 10.1128/AAC.00419-13) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Dong C, Hao GF, Hua HL, Liu S, Labena AA, Chai G, Huang J, Rao N, Guo FB. 2018. Anti-CRISPRdb: a comprehensive online resource for anti-CRISPR proteins. Nucleic Acids Res. 46, D393–D398. ( 10.1093/nar/gkx835) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.Biswas A, Staals RH, Morales SE, Fineran PC, Brown CM. 2016. CRISPRDetect: a flexible algorithm to define CRISPR arrays. BMC Genomics 17, 356 ( 10.1186/s12864-016-2627-0) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47.Chyou TY, Brown CM. 2018. Prediction and diversity of tracrRNAs from type II CRISPR-Cas systems. RNA Biol. 1–12. ( 10.1080/15476286.2018.1498281) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48.Tatusova T, Ciufo S, Federhen S, Fedorov B, McVeigh R, O'Neill K, Tolstoy I, Zaslavsky L. 2015. Update on RefSeq microbial genomes resources. Nucleic Acids Res. 43, D599–D605. ( 10.1093/nar/gku1062) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 49.Biswas A, Gagnon JN, Brouns SJ, Fineran PC, Brown CM. 2013. CRISPRTarget: bioinformatic prediction and analysis of crRNA targets. RNA Biol. 10, 817–827. ( 10.4161/rna.24046) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 50.Couvin D, et al. 2018. CRISPRCasFinder, an update of CRISRFinder, includes a portable version, enhanced performance and integrates search for Cas proteins. Nucleic Acids Res. 46, W246–W251. ( 10.1093/nar/gky425) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 51.Burstein D, Sun CL, Brown CT, Sharon I, Anantharaman K, Probst AJ, Thomas BC, Banfield JF. 2016. Major bacterial lineages are essentially devoid of CRISPR-Cas viral defence systems. Nat. Commun. 7, 10 613 ( 10.1038/ncomms10613) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 52.Koonin EV, Wolf YI. 2015. Evolution of the CRISPR-Cas adaptive immunity systems in prokaryotes: models and observations on virus-host coevolution. Mol. Biosyst. 11, 20–27. ( 10.1039/c4mb00438h) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 53.Koonin EV, Makarova KS, Zhang F. 2017. Diversity, classification and evolution of CRISPR-Cas systems. Curr. Opin. Microbiol. 37, 67–78. ( 10.1016/j.mib.2017.05.008) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 54.Sorek R, Kunin V, Hugenholtz P. 2008. CRISPR—a widespread system that provides acquired resistance against phages in bacteria and archaea. Nat. Rev. Microbiol. 6, 181–186 ( 10.1038/nrmicro1793) [DOI] [PubMed] [Google Scholar]
- 55.Zhang Q, Ye Y. 2017. Not all predicted CRISPR-Cas systems are equal: isolated cas genes and classes of CRISPR like elements. BMC Bioinf. 18, 92 ( 10.1186/s12859-017-1512-4) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 56.Varahan S, Hancock LE. 2016. To defend or not to defend: that's the question. mSphere 1, e00127-16 ( 10.1128/mSphere.00127-16) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 57.Gundogdu A, Long YB, Vollmerhausen TL, Katouli M. 2011. Antimicrobial resistance and distribution of sul genes and integron-associated intI genes among uropathogenic Escherichia coli in Queensland, Australia. J. Med. Microbiol. 60, 1633–1642. ( 10.1099/jmm.0.034140-0) [DOI] [PubMed] [Google Scholar]
- 58.Hsu JT, Chen CY, Young CW, Chao WL, Li MH, Liu YH, Lin CM, Ying C. 2014. Prevalence of sulfonamide-resistant bacteria, resistance genes and integron-associated horizontal gene transfer in natural water bodies and soils adjacent to a swine feedlot in northern Taiwan. J. Hazard. Mater. 277, 34–43. ( 10.1016/j.jhazmat.2014.02.016) [DOI] [PubMed] [Google Scholar]
- 59.Richter C, Dy RL, McKenzie RE, Watson BN, Taylor C, Chang JT, McNeil MB, Staals RH, Fineran PC. 2014. Priming in the Type I-F CRISPR-Cas system triggers strand-independent spacer acquisition, bi-directionally from the primed protospacer. Nucleic Acids Res. 42, 8516–8526. ( 10.1093/nar/gku527) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 60.Zhao X, Yu Z, Xu Z. 2018. Study the features of 57 confirmed CRISPR loci in 38 strains of Staphylococcus aureus. Front. Microbiol. 9, 1591 ( 10.3389/fmicb.2018.01591) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 61.Kilian M, Riley DR, Jensen A, Bruggemann H, Tettelin H. 2014. Parallel evolution of Streptococcus pneumoniae and Streptococcus mitis to pathogenic and mutualistic lifestyles. MBio 5, e01490-14 ( 10.1128/mBio.01490-14) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 62.Li HY, Kao CY, Lin WH, Zheng PX, Yan JJ, Wang MC, Teng CH, Tseng CC, Wu JJ. 2018. Characterization of CRISPR-Cas systems in clinical Klebsiella pneumoniae isolates uncovers its potential association with antibiotic susceptibility. Front. Microbiol. 9, 1595 ( 10.3389/fmicb.2018.01595) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 63.Zhang Y, Heidrich N, Ampattu BJ, Gunderson CW, Seifert HS, Schoen C, Vogel J, Sontheimer EJ. 2013. Processing-independent CRISPR RNAs limit natural transformation in Neisseria meningitidis. Mol. Cell 50, 488–503. ( 10.1016/j.molcel.2013.05.001) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 64.Muniesa M, Garcia A, Miro E, Mirelis B, Prats G, Jofre J, Navarro F. 2004. Bacteriophages and diffusion of beta-lactamase genes. Emerg. Infect. Dis. 10, 1134–1137. ( 10.3201/eid1006.030472) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 65.Balcazar JL. 2014. Bacteriophages as vehicles for antibiotic resistance genes in the environment. PLoS Pathog. 10, e1004219 ( 10.1371/journal.ppat.1004219) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 66.Naser IB, Hoque MM, Nahid MA, Tareq TM, Rocky MK, Faruque SM. 2017. Analysis of the CRISPR-Cas system in bacteriophages active on epidemic strains of Vibrio cholerae in Bangladesh. Sci. Rep. 7, 14 880 ( 10.1038/s41598-017-14839-2) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 67.Kim EJ, Yu HJ, Lee JH, Kim JO, Han SH, Yun CH, Chun J, Nair GB, Kim DW. 2017. Replication of Vibrio cholerae classical CTX phage. Proc. Natl Acad. Sci. USA 114, 2343–2348. ( 10.1073/pnas.1701335114) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 68.O'Shea YA, Boyd EF. 2002. Mobilization of the Vibrio pathogenicity island between Vibrio cholerae isolates mediated by CP-T1 generalized transduction. FEMS Microbiol. Lett. 214, 153–157. ( 10.1016/S0378-1097(02)00880-7) [DOI] [PubMed] [Google Scholar]
- 69.Pawluk A, Shah M, Mejdani M, Calmettes C, Moraes TF, Davidson AR, Maxwell KL. 2017. Disabling a Type I-E CRISPR-Cas nuclease with a bacteriophage-encoded anti-CRISPR protein. MBio 8, 01751-17 ( 10.1128/mBio.01751-17) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 70.Peeters N, Guidot A, Vailleau F, Valls M. 2013. Ralstonia solanacearum, a widespread bacterial plant pathogen in the post-genomic era. Mol. Plant Pathol. 14, 651–662. ( 10.1111/mpp.12038) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 71.Pawluk A, Staals RH, Taylor C, Watson BN, Saha S, Fineran PC, Maxwell KL, Davidson AR. 2016. Inactivation of CRISPR-Cas systems by anti-CRISPR proteins in diverse bacterial species. Nat. Microbiol. 1, 16 085 ( 10.1038/nmicrobiol.2016.85) [DOI] [PubMed] [Google Scholar]
- 72.Landsberger M, Gandon S, Meaden S, Rollie C, Chevallereau A, Chabas H, Buckling A, Westra ER, van Houte S. 2018. Anti-CRISPR phages cooperate to overcome CRISPR-Cas immunity. Cell 174, 908– 916.e912 ( 10.1016/j.cell.2018.05.058) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 73.McMahon SA, et al. 2009. Extensive DNA mimicry by the ArdA anti-restriction protein and its role in the spread of antibiotic resistance. Nucleic Acids Res. 37, 4887–4897. ( 10.1093/nar/gkp478) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 74.Chen K, Reuter M, Sanghvi B, Roberts GA, Cooper LP, Tilling M, Blakely GW, Dryden DT. 2014. ArdA proteins from different mobile genetic elements can bind to the EcoKI Type I DNA methyltransferase of E. coli K12. Biochim. Biophys. Acta 1844, 505–511. ( 10.1016/j.bbapap.2013.12.008) [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
Additional data is available at https://github.com/davidchyou.