Elsevier

Plasmid

Volume 104, July 2019, 102418
Plasmid

Review
A phylogenetic test of the role of CRISPR-Cas in limiting plasmid acquisition and prophage integration in bacteria

https://doi.org/10.1016/j.plasmid.2019.102418Get rights and content

Highlights

  • Bacterial strains with CRISPR-Cas had on average less than half the plasmid counts of related CRISPR-Cas absent strains.

  • Reversals to this rule are driven by higher stochasticity when total plasmid counts between the two are low.

  • No CRISPR-Cas effects were seen on intact prophage counts or the total amount of genomic DNA found in the bacteria.

Abstract

CRISPR-Cas is a prokaryotic defense system capable of protecting the cell from damaging foreign genetic elements. However, some genetic elements can be beneficial, which suggests the hypothesis that bacteria with CRISPR-Cas incur a cost of reduced intake of mutualistic plasmids and prophage. Here we present the first robust test of this hypothesis that controls for phylogenic and ecological biases in the distribution of CRISPR-Cas. We filtered the available genomic data (~7000 strains from ~2100 species) by first selecting all pairs of conspecific strains, one with and one without CRISPR-Cas (controlling ecological bias), and second by retaining only one such pair per bacterial family (controlling phylogenetic bias), resulting in pairs representing 38 bacterial families. Analysis of these pairs of bacterial strains showed that on average the CRISPR-Cas strain of each pair contained significantly fewer plasmids than its CRISPR-Cas negative partner (0.86 vs. 1.86). It also showed that the CRISPR-Cas positive strains had 31% fewer intact prophage (1.17 vs. 1.75), but the effect was highly variable and not significant. These results support the hypothesis that CRISPR-Cas reduces the rate of plasmid-mediated HGT and, given the abundant evidence of beneficial genes carried by plasmids, provide a clear example of a cost associated with the CRISPR-Cas system.

Introduction

The prokaryotic defense system, CRISPR-Cas, has become well known for its ability to provide immunity for its bacterial host against foreign genetic elements (Barrangou et al., 2007; Marraffini and Sontheimer, 2008). The system is composed of a CRISPR Cassette region, made up of a series of 25–50 bp long DNA sequences known as “spacers”, each flanked by similarly sized “repeat” sequences, and the Cas (CRISPR associated) genes (Barrangou et al., 2007). The spacers are derived from exogenous sources of genetic material and are able to target these sources (through sequence matching) for degradation carried out by the Cas genes (Sorek et al., 2008). The primary benefit of the CRISPR-Cas system comes from its ability to protect bacteria from pathogenic elements such as bacteriophage. However, it has been proposed that the immunity provided by the CRISPR-Cas system may come at a cost to the host cell, a view supported by the observation of both inter- and intra-specific variation in the presence of CRISPR-Cas (Jiang et al., 2013; Vale et al., 2015). These proposed costs include metabolic costs (Weinberger and Gilmore, 2012), risks of self-targeting autoimmunity (Stern et al., 2010), and the opportunity cost of reducing the ability of the bacteria to uptake genes from its environment (Marraffini and Sontheimer, 2008; Jiang et al., 2013; Palmer and Gilmore, 2010).

We focus here on the last of these, the proposed opportunity cost of reduced acquisition of novel DNA. CRISPR-Cas systems can prevent DNA uptake via plasmid conjugation (Marraffini and Sontheimer, 2008), transformation with naked DNA (Bikard et al., 2012), and viral transduction (Edgar and Qimron, 2010). Experimental studies supporting the negative effect of reducing plasmid uptake have found CRISPR-Cas to act as a barrier to the uptake of drug resistant plasmids in both Staphylococci (Marraffini and Sontheimer, 2008) and Enterococci (Palmer and Gilmore, 2010), although this was not seen in Escherichia. coli (Touchon et al., 2012). Similarly, if the intake of new beneficial genes through transduction is limited by CRISPR-Cas, then this may also impose a cost on bacteria carrying it. There is experimental evidence of CRISPR-Cas reducing the chance of lytic phage integration in E. coli (Edgar and Qimron, 2010). Furthermore, it has been observed that in the genus Bifidobacterium that species with CRISPR-Cas often carry spacers matching prophage sequence found in other species lacking CRISPR-Cas (Briner et al., 2015), and across Streptococcus pyogenes strains it was found that prophage counts were lower in strains with CRISPR-Cas (Nozawa et al., 2011). Natural selection favors lysogenic phage that bring with them genes that increase host fitness (Obeng et al., 2016), since their persistence depends primarily upon vertical transmission rather than infectious horizontal transfer. For example, it has been proposed that elevated levels of antibiotic resistance within Staphylococcus aureus are spread through transduction of lysogenic phage (Haaber et al., 2017).

Thus the hypothesis that CRISPR-Cas can impose an opportunity cost by limiting DNA acquisition has some experimental and observational support in a few bacterial taxa. However, it has yet to be established as a phenomenon affecting bacteria in general that potentially influences the distribution of CRISPR-Cas across the domain.

One previous study attempted to test the generality of this hypothesis by comparing the available prokaryote genomes with CRISPR-Cas (532 from bacteria; 96 from archaea) with those lacking CRISPR-Cas (705 from bacteria; 18 from archaea) (Gophna et al., 2015). They failed to find support for the hypothesis and based on this result concluded that any effect of CRISPR–Cas on HGT was occurring at a short time scale and could not be detected at the level of species evolution. However, these results should be treated with caution due to the lack of controls for phylogenetic or ecological bias. It has been recognized since the 1980s (Felsenstein, 1985) that a failure to control for phylogenetic biases in the distribution of a trait (in this case, the presence or absence of CRISPR-Cas) can lead to spurious results, a recognition that led to the development of strict guidelines in the application of the “comparative method” (O'Meara, 2012). In particular, the genomes available for the study were heavily biased towards species of medical importance (often with multiple strains per species) and biased taxonomic sampling of this type can be a source of serious bias in comparative studies. In addition, the possibility that the presence or absence of CRISPR-Cas may be associated with unknown environmental variables creates a second layer of potential bias; indeed it was found that in the study that growth temperature was an important confounding factor, with species carrying CRISPR-Cas having more spacers if they were adapted to high temperature (Gophna et al., 2015).

To eliminate these potential sources of bias, and to incorporate the possibility that comparisons on a relatively short time scale (i.e. within species) were most appropriate, we analyzed data from across the eubacteria kingdom comparing genomes with and without CRISPR-Cas within the same species. Furthermore to avoid any phylogenetic bias, we used only one species to represent each bacterial family. Using this approach, we could rigorously test the hypothesis that CRISPR-Cas presence has the general property of reducing the acquisition of new genetic material across a wide array of bacteria, all from different families. The two primary hypotheses being tested were whether CRISPR-Cas presence reduced the number of plasmids acquired and whether it reduced the level of prophage integration. We also examined whether there was any general effect of CRISPR-Cas in limiting genome size, after plasmids and prophage were excluded. A secondary pair of questions addresses whether CRISPR-Cas presence influenced the size of plasmids or of prophage that are taken up and retained by bacteria. By identifying whether CRISPR-Cas acts as a functional barrier to accumulation of these diverse types of DNA via HGT, we may better understand the costs and benefits involved in maintaining CRISPR-Cas.

Section snippets

Identifying CRISPR-Cas positive and CRISPR-Cas negative strains for pairwise analysis

Initial screening was carried out using CRISPRdb (Grissa et al., 2007a), a database representing close to 7000 bacterial strains (regardless of CRISPR presence/absence) representing over 2100 bacterial species (as of 4/01/2018). Bacterial genomes entered in this database are searched for CRISPR Cassettes using the program, CRISPRFinder (Grissa et al., 2007b). Using this information, bacterial species were considered for inclusion in the analysis if they had at least one completely sequenced

Identification of CRISPR-Cas variable species

A total of 132 species from 45 different families with variable CRISPR-Cas presence (C+)/absence (C-) were identified from CRISPRdb (Supplemental File 1). Analysis of these CRISPR-Cas variable species identified 38 families represented by at least one species with a fully sequenced C+ strain (as determined by having at least 4 nearby Cas genes) and a fully sequenced C– strain. These 38 families spanned over 5 bacterial phyla with 4 coming from Actinobacteria, 2 from Bacteroidetes, 1 from

Discussion

CRISPR-Cas is not present in all species of bacteria, and in species where it is present it is not always found in all strains. These observations suggest that CRISPR-Cas is not universally beneficial. To understand the distribution of CRISPR-Cas among bacteria, it is important to determine if CRISPR-Cas serves as a barrier against the acquisition of new, potentially beneficial genes. To this end, we looked at content differences in three types of DNA between paired strains from the same

Acknowledgements

We thank Joel Sachs and Mark Springer for their helpful comments on the manuscript.

Author contributions

Derek O'Meara contributed to the writing of this manuscript, the collection of data, and the statistical design/analysis. Leonard Nunney contributed to the writing of this manuscript and the statistical design/analysis.

Competing interests

The authors declare that there are no conflicts of interest regarding the work performed in this paper.

References (39)

  • D. Burstein

    Major bacterial lineages are essentially devoid of CRISPR-Cas viral defence systems

    Nat. Commun.

    (2016)
  • D.A. Cooksey

    Plasmid-determined copper resistance in Pseudomonas syringae from impatiens

    Appl. Environ. Microbiol.

    (1990)
  • R. Edgar et al.

    The Escherichia coli CRISPR system protects from lysogenization, lysogens, and prophage induction

    J. Bacteriol.

    (2010)
  • J. Felsenstein

    Phylogenies and the comparative method

    Am. Nat.

    (1985)
  • J.S. Godde et al.

    The repetitive DNA elements called CRISPRs and their associated genes: evidence of horizontal transfer among prokaryotes

    J. Mol. Evol.

    (2006)
  • G.W. Goldberg et al.

    Resistance and tolerance to foreign elements by prokaryotic immune systems — curating the genome

    Nat. Rev. Immunol.

    (2015)
  • U. Gophna

    No evidence of inhibition of horizontal gene transfer by CRISPR–Cas on evolutionary timescales

    ISME J.

    (2015)
  • I. Grissa et al.

    The CRISPRdb database and tools to display CRISPRs and to generate dictionaries of spacers and repeats

    BMC Bioinforma.

    (2007)
  • I. Grissa et al.

    CRISPRFinder: a web tool to identify clustered regularly interspaced short palindromic repeats

    Nucleic Acids Res.

    (2007)
  • Cited by (9)

    • CRISPR/Cas12a-based technology: A powerful tool for biosensing in food safety

      2022, Trends in Food Science and Technology
      Citation Excerpt :

      The clustered regularly interspaced short palindromic repeats (CRISPR) system originated from the adaptive immune system of the bacteria and archaea, acting as a “gene weapon” against foreign genetic material, such as phages and plasmids (Koonin et al., 2017; Watson et al., 2021; O'Meara & Nunney, 2019).

    • Evolutionary Ecology and Interplay of Prokaryotic Innate and Adaptive Immune Systems

      2020, Current Biology
      Citation Excerpt :

      The role of type IV RM systems may be more limited, but this will depend on the epigenetic status of MGEs in the environment, which has not been sufficiently examined. Conclusions vary about the effect of CRISPR–Cas systems on long-term HGT [58–60]. Some lineages with strong signatures of HGT are depleted in CRISPR–Cas systems [61], but across lineages there are no clear correlations between HGT and CRISPR–Cas activity [59].

    • Marinomonas mediterranea synthesizes an R-Type bacteriocin

      2024, Applied and Environmental Microbiology
    • CRISPR-Cas is associated with fewer antibiotic resistance genes in bacterial pathogens

      2022, Philosophical Transactions of the Royal Society B: Biological Sciences
    View all citing articles on Scopus
    View full text