Host population structure and species resolution reveal prophage transmission dynamics

Abstract

Much knowledge about bacteriophages has been obtained via genomics and metagenomics over the last decades. However, most studies dealing with prophage diversity have rarely conducted phage species delimitation (aspect 1) and have hardly integrated the population structure of the host (aspect 2). Yet, these two aspects are essential in assessing phage diversity. Here, we implemented an operational definition of phage species (clustering at 95% identity, 90% coverage) and integrated the host's population structure to understand prophage diversity better. Gathering the most extensive data set of Acinetobacter baumannii phages (4,152 prophages + 122 virulent phages, distributed in 46 countries in the world), we show that 91% (875 out of 963) of the prophage species have four or fewer prophages per species, and just five prophage species have more than 100 prophages. Most prophage species have a narrow host range and are geographically restricted; yet, very few have a broad host range being well spread in distant lineages of A. baumannii. These few broad host range prophage species are not only cosmopolitan but also the most abundant species. We also noted that polylysogens had very divergent prophages, belonging to different prophage species, and prophages can easily be gained and lost within the bacterial lineages. Finally, even with this extensive data set, the prophage diversity has not been fully grasped. Our study highlights how integrating the host population structure and a solid operational definition of phage species allows us to better appreciate phage diversity and its transmission dynamics.

Importance: Much knowledge about bacteriophages has been obtained via genomics and metagenomics over the last decades. However, most studies dealing with prophage diversity have rarely conducted phage species delimitation (aspect 1) and have hardly integrated the population structure of the host (aspect 2). Yet, these two aspects are essential in assessing phage diversity. Here, we implemented an operational definition of phage species (clustering at 95% identity, 90% coverage) and integrated the host's population structure to understand prophage diversity better. Gathering the most extensive data set of Acinetobacter baumannii phages, we show that most prophage species have four or fewer prophages per species, and just five prophage species have more than 100 prophages. Most prophage species have a narrow host range and are geographically restricted; yet, very few have a broad host range being well spread in distant lineages of A. baumannii. These few broad host range prophage species are cosmopolitan and the most abundant species. Prophages in the same bacterial genome are very divergent, and prophages can easily be gained and lost within the bacterial lineages. Finally, even with this extensive data set, the prophage diversity has not been fully grasped. This study shows how integrating the host population structure and clustering at the species level allows us to better appreciate phage diversity and its transmission dynamics.

Keywords: ANI; Acinetobacter baumannii; bacteriophage genetics; phage species; population genomics; prophages; species definition.

Fig 1

Frequency of prophages. (A) Number of prophages in bacterial isolates from different hosts: animal (pink), humans (orange), and plants (green). The numbers on the top of the violin plots indicate the bacterial genomes considered in each case. (B) Number of prophages in bacterial isolates belonging to different STs as per Pasteur MLST. Only the STs with 10 or more bacterial isolates are shown. Each color refers to a different ST.

Fig 2

Number of prophage species. (A) Frequency of prophage species with different numbers of prophages. The size of the circle (and the y-axis) denotes the frequency of prophage species with a given number of sequences (x-axis). Blue denotes cases where prophage species were only found in one ST; orange indicates that some prophage species were found in one ST but other species in two STs, whereas dark red shows that prophages were found in two or more STs. (B) Network showing the Ab prophage species with two or more sequences per cluster; the nodes are color-coded as per the host (see color key), and shapes in the case of animal hosts show different types of animals. The dashed squares highlight hybrid Ab_PS where prophages came from bacterial isolates from different hosts.

Fig 3

Diversity of prophage species. (A) Heat map showing those prophage species that were found in bacterial genomes from two or more STs. To the right, there are the Ab prophage species, and at the bottom, there are the different STs. Blue denotes presence, and cream denotes absence. (B) Ab prophage species accumulation curve. The y-axis gives the number of prophages as a function of the bacterial genomes used to infer the prophages. The gray area denotes the confidence intervals.

Fig 4

Prophage geographic expansion and polylysogeny. (A) Prophage species with X number of prophages (x-axis) found in different countries (y-axis). The color of the dot shows how many STs the prophages are found. (B) Similarity of prophages within the same bacterial genome. ANI values between the prophages in individual bacterial genomes. Color code as follows: animal (pink), humans (orange), and plants (green). The black dashed line denotes an ANI value of 95%, whereas the blue dashed line shows the mean of all the ANI values (~4%).