Periodic Reporting for period 1 - CO-DEfENDENCE (Bacteriophages, phage-host interactions and evolution of defence mechanisms in beneficial bacterium Propionibacterium freudenreichii)
Reporting period: 2020-01-01 to 2021-12-31
Propionibacterium freudenreichii is a dairy-associated bacterium with multiple applications in the dairy industry, from adjunct cultures to bioprotective cultures. Many strains of this bacterium have been reported to have probiotic potential and all have the rare capability to produce active vitamin B12. At the same time, the bacterium is relatively poorly studied compared to the other dairy bacteria, especially in the context of their sensitivity to infection with bacterial viruses (bacteriophages). At the beginning of the project, very few bacteriophages infecting this bacterium were characterised, at the same time, the recently published genome sequences of the bacterial strains indicated a broader-than-average arsenal of defence systems protecting the bacteria from bacteriophage infection, despite overall low genetic diversity of the species. This, and the fact that P. freudenreichii is a non-pathogenic representative of the class Actinomycetes, which includes various pathogenic or opportunistic pathogenic bacteria such as Mycobacterium tuberculosis and other mycobacteria or Cutibacterium acnes, means that the results of this project can be translated to these other societally important bacteria as well.
The objectives of the project were to isolate and characterise new bacteriophages infecting P. freudenreichii from various samples and to sequence the genomes of the bacteriophages as well as the sensitive and insensitive bacterial hosts in order to better understand the dynamics of bacteriophage-bacterium interactions of the bacteriophage-bacterium pair.
As a result, we have characterised two new bacteriophages infecting P. freudenreichii as well as multiple new bacterial strains. We were able to uncover that the scarcity of bacteriophages known for this bacterium are not due to their absence, but due to the efficient defence systems encoded by the bacterium conferring their immunity and complicating phage detection. This also clarified as to why the bacteriophages found to date are comparatively more diverse than the ones known for related bacteria and that one of their strategies for overcoming bacterial defence systems is acquisition of diverse methylases, traditionally associated with Restriction-Modification systems working as protection against bacteriophage infection. One of the bacteriophages will be used in further studies to develop genetic tools for manipulation of Actinomycetes and also for possible phage display application.
The objectives of the project were to isolate and characterise new bacteriophages infecting P. freudenreichii from various samples and to sequence the genomes of the bacteriophages as well as the sensitive and insensitive bacterial hosts in order to better understand the dynamics of bacteriophage-bacterium interactions of the bacteriophage-bacterium pair.
As a result, we have characterised two new bacteriophages infecting P. freudenreichii as well as multiple new bacterial strains. We were able to uncover that the scarcity of bacteriophages known for this bacterium are not due to their absence, but due to the efficient defence systems encoded by the bacterium conferring their immunity and complicating phage detection. This also clarified as to why the bacteriophages found to date are comparatively more diverse than the ones known for related bacteria and that one of their strategies for overcoming bacterial defence systems is acquisition of diverse methylases, traditionally associated with Restriction-Modification systems working as protection against bacteriophage infection. One of the bacteriophages will be used in further studies to develop genetic tools for manipulation of Actinomycetes and also for possible phage display application.
Work performed within WP1 consisted of employing mutliple sequencing platforms to determine usability of each of them in generation of high quality whole genome sequences and to contribute to better understanding of P. freudenreichii as a species by generating multiple new genomes, especially those of bacteriophage-sensitive strains.
The results obtained in WP1 contributed to the conclusion that the strain TL110 is widely sensitive to phages due to the loss of CRISPR-Cas system, which appears to protect other strains previously exposed to similar phages (concluded on the basis of the composition of their CRISPR arrays) from re-infection. The number of sequenced strains had to be reduced due to the high cost of PacBio sequencing. However, time and resources spent on Nanopore sequencing turned out not to be wasted, as the platform has finally reached maturity and the article comparing the initial sequences with improved sequences against those generated by other sequencing platform within the project is soon due to be published.
Work performed in WP2 entailed testing of various samples, including dairy products, grassess, malts and silage for presence of bacteriophages infecting P. freudenreichii.
The main realisation of the WP2 was that isolation of bacteriophages from different materials in a traditional wayis impossible, as the strains easily acquire resistance to phages. Screening of different materials on unrelated strains yielded one, albeit very interesting, filamentous phage. The second phage resulted from the induction of previously discovered prophages, which could luckily be plaqued on two unrelated strains, which later turned out to have lost their CRISPR-Cas systems.
WP 3 focused on analysis of defence systems of newly sequenced bacterial strains
The sequenced strains were subjected to bioinformatics prediction of defence systems using dedicated tools such as DefenseFinder, PADLOC and CRISPR-finder. The activity of RM systems was confirmed by cross-infection of sensitive strains with a phage propagated on different hosts, determined to lead to a two log increase in phage titers. Infection with both phages allowed isolation of BIMs, which for tailed phage were determined to result from integration of the phage into the host’s chromosome as prophage, while the ones raised on filamentous phage await analysis. None of the BIMs displayed a weakened phenotype, which is typically associated with gained resistance to phages, therefore leading to the idea of vaccination of industrially relevant strains. While this idea is novel for P. freudenreichii, it has been successfully used before for CRISPR-Cas-containing strains of dairy-relevant Streptococcus thermophilus.
Within WP4, the presence and activity of antidefence systems present in bacteriophage genomes was explored
Analysis of sequences of the isolated phages led to discovery of RM-associated methylase on the genome of the tailed phage. Comparative analysis of the genomes of all the known phages infecting P. freudenreichii revealed that such methylases are widespread, found in 6 out of total 10 tailed phages known to infect this bacterial species. The methylase was predicted to be responsible for m5C type of methylations, activity of which is not possible to determine with PacBio sequencing. Therefore we used a new approach of RIMS-seq sequencing method based on restriction enzyme and Illumina sequencing conducted by one of our collaborators. This approach allowed us to determine the methylation motif sequence of the methylase and also to determine that it is active only during infection of a naïve strain (i.e. one not exposed to the phage before).
The results obtained in WP1 contributed to the conclusion that the strain TL110 is widely sensitive to phages due to the loss of CRISPR-Cas system, which appears to protect other strains previously exposed to similar phages (concluded on the basis of the composition of their CRISPR arrays) from re-infection. The number of sequenced strains had to be reduced due to the high cost of PacBio sequencing. However, time and resources spent on Nanopore sequencing turned out not to be wasted, as the platform has finally reached maturity and the article comparing the initial sequences with improved sequences against those generated by other sequencing platform within the project is soon due to be published.
Work performed in WP2 entailed testing of various samples, including dairy products, grassess, malts and silage for presence of bacteriophages infecting P. freudenreichii.
The main realisation of the WP2 was that isolation of bacteriophages from different materials in a traditional wayis impossible, as the strains easily acquire resistance to phages. Screening of different materials on unrelated strains yielded one, albeit very interesting, filamentous phage. The second phage resulted from the induction of previously discovered prophages, which could luckily be plaqued on two unrelated strains, which later turned out to have lost their CRISPR-Cas systems.
WP 3 focused on analysis of defence systems of newly sequenced bacterial strains
The sequenced strains were subjected to bioinformatics prediction of defence systems using dedicated tools such as DefenseFinder, PADLOC and CRISPR-finder. The activity of RM systems was confirmed by cross-infection of sensitive strains with a phage propagated on different hosts, determined to lead to a two log increase in phage titers. Infection with both phages allowed isolation of BIMs, which for tailed phage were determined to result from integration of the phage into the host’s chromosome as prophage, while the ones raised on filamentous phage await analysis. None of the BIMs displayed a weakened phenotype, which is typically associated with gained resistance to phages, therefore leading to the idea of vaccination of industrially relevant strains. While this idea is novel for P. freudenreichii, it has been successfully used before for CRISPR-Cas-containing strains of dairy-relevant Streptococcus thermophilus.
Within WP4, the presence and activity of antidefence systems present in bacteriophage genomes was explored
Analysis of sequences of the isolated phages led to discovery of RM-associated methylase on the genome of the tailed phage. Comparative analysis of the genomes of all the known phages infecting P. freudenreichii revealed that such methylases are widespread, found in 6 out of total 10 tailed phages known to infect this bacterial species. The methylase was predicted to be responsible for m5C type of methylations, activity of which is not possible to determine with PacBio sequencing. Therefore we used a new approach of RIMS-seq sequencing method based on restriction enzyme and Illumina sequencing conducted by one of our collaborators. This approach allowed us to determine the methylation motif sequence of the methylase and also to determine that it is active only during infection of a naïve strain (i.e. one not exposed to the phage before).
As a result of this project, we now know that P. freudenreichii is infected with multiple and rather diverse bacteriophages, however the efficient action of bacterial defence systems and resulting compensation in cell growth makes the detection of said phages complicated. At the same time, widespread presence of at least two different types of adaptive immune system CRISPR-Cas plays an important role in conferring the immunity. This knowledge is now used in development of bacteriophage instensitive strains through vaccination. Moreover, the very rare filamentous phage able to infect Gram-positive bacterium will be studies further with the aim to develop genetic tools, similar to those based on Escherichia coli M13 phage and the potential for application of the bacteriophage in phage display technology.