The diversity and host interactions of Propionibacterium acnes bacteriophages on human skin

16 Pages·2015·1.48 MB·English

by  Jared Liu

TheISMEJournal(2015)9,2078–2093
©2015InternationalSocietyforMicrobialEcology Allrightsreserved 1751-7362/15
www.nature.com/ismej
ORIGINAL ARTICLE
The diversity and host interactions of
Propionibacterium acnes
bacteriophages
on human skin
This paper has been corrected and a corrigendum also appears in this issue
Jared Liu1, Riceley Yan1, Qiao Zhong1,2, Sam Ngo1, Nathanael J Bangayan1, Lin Nguyen1,
Timothy Lui1, Minghsun Liu3, Marie C Erfe4, Noah Craft4, Shuta Tomida1,6 and Huiying Li1,5
1Department of Molecular and Medical Pharmacology, Crump Institute for Molecular Imaging, David Geffen
School of Medicine, UCLA, Los Angeles, CA, USA; 2Department of Laboratory Medicine, Suzhou Municipal
Hospital, Suzhou Hospital Affiliated to Nanjing Medical University, Suzhou, China; 3Department of
Microbiology, Immunology and Molecular Genetics, David Geffen School of Medicine, UCLA, Los Angeles,
CA,USA;4LosAngelesBiomedicalResearchInstituteatHarbor-UCLAMedicalCenter,LosAngeles,CA,USA
and 5UCLA-DOE Institute for Genomics and Proteomics, Los Angeles, CA, USA
Theviralpopulation,includingbacteriophages,isanimportantcomponentofthehumanmicrobiota,
yetispoorlyunderstood.Weaimtodeterminewhetherbacteriophagesmodulatethecompositionof
the bacterial populations, thus potentially playing a role in health or disease. We investigated the
diversity and host interactions of the bacteriophages of Propionibacterium acnes, a major human
skin commensal implicated in acnepathogenesis. Bysequencing 48P.acnes phagesisolated from
acnepatientsandhealthyindividualsandbyanalyzingtheP.acnesphagepopulationsinhealthyskin
metagenomes, we revealed that P. acnes phage populations in the skin microbial community are
oftendominatedbyonestrain.Wealsofoundphagestrainssharedamongbothrelatedandunrelated
individuals, suggesting that a pool of common phages exists in the human population and that
transmission of phages mayoccur between individuals. To better understand the bacterium–phage
interactions in the skin microbiota, we determined the outcomes of 74 genetically defined
Propionibacteriumstrainschallengedby15sequencedphages.DependingonthePropionibacterium
lineage,phageinfectioncanresultinlysis,pseudolysogeny,orresistance.IntypeIIP.acnesstrains,
wefoundthatencodingmatchingclusteredregularlyinterspacedshortpalindromicrepeatspacersis
insufficient to confer phage resistance. Overall, our findings suggest that the prey–predator
relationship between bacteria and phages may have a role in modulating the composition of the
microbiota. Our study also suggests that the microbiome structure of an individual may be an
important factorin thedesign ofphage-basedtherapy.
TheISME Journal (2015) 9, 2078–2093; doi:10.1038/ismej.2015.47; published online7April 2015
Introduction (Rohwer and Thurber, 2009) and regulate both the
abundance and diversity of their bacterial hosts by
Thehumanskinisinhabitedbyhundredsofmicrobial
predation (Suttle et al., 1990; Waterbury and Valois,
species, including bacteria, fungi and viruses. The
1993; Rohwer, 2003; Rodriguez-Valera et al., 2009).
homeostasis of this ecosystem is important to its
Although the skin bacterial community has been
functionasabarrieragainstinfectionandcolonization
studied by several groups (Gao et al., 2007; Costello
of pathogens on the skin surface. Bacteriophages
et al., 2009; Grice et al., 2009; Kong et al., 2012;
are important components of the human microbiota.
The Human Microbiome Project Consortium, 2012;
They are a reservoir of diversity-generating elements
Blaseretal.,2013;Fitz-Gibbonetal.,2013;Nakatsuji
et al., 2013), relatively few studies have character-
ized the skin viral community (Foulongne et al.,
Correspondence: H Li, Department of Molecular and Medical
2012;Maetal.,2014;Wylieetal.,2014).Inparticular,
Pharmacology, Crump Institute for Molecular Imaging, David
Geffen School of Medicine, UCLA, 4339 CNSI, 570 Westwood thecompositionanddynamicsofbacteriophagesand
Plaza,Building114,LosAngeles,CA90095-1770,USA. theirinteractionswithbacterialhostsontheskinare
E-mail:[email protected] not well understood.
6Currentaddress:DepartmentofGenomeBiology,KinkiUniversity,
The microbial community in the skin pilosebac-
Osaka,Japan.
eous unit is dominated by Propionibacterium acnes,
Received 9 October 2014; revised 12 February 2015; accepted
26February2015;publishedonline7April2015 which accounts for nearly 90% of the microbiota
P.acnesphagepopulationintheskinmicrobiota
JLiuetal
2079
(Fitz-Gibbon et al., 2013). Although P. acnes is a On the other hand, bacterial hosts can influence
major skin commensal, it has been considered a phage populations through antiviral mechanisms,
pathogenic factor for acne vulgaris (Leyden, 2001; such as the restriction modification mechanism and
Bojar and Holland, 2004), one of the most common the bacterial adaptive immune system utilizing
skin diseases affecting over 80% of adolescents clustered regularly interspaced short palindromic
and young adults (White, 1998; Bergler-Czop and repeat (CRISPR) sequence arrays (Horvath and
Brzezin´ska-Wcisło, 2013). Our previous 16S ribosomal Barrangou, 2010). In our effort to characterize the
RNAmetagenomicstudydemonstratedthatP.acnes straindiversityofP.acnesinthepilosebaceousunit,
strain population structure in pilosebaceous units we discovered that all sequenced type II P. acnes
differs significantly between acne patients and strains harbor Type I-E CRISPR and CRISPR-
healthy individuals (Fitz-Gibbon et al., 2013). associated (Cas) proteins (Fitz-Gibbon et al., 2013;
Certainstrainsarehighlyassociatedwiththedisease Tomida et al., 2013). Marinelli et al. (2012)
(Lomholt and Kilian, 2010; McDowell et al., 2012; suggested that the CRISPR mechanism explains the
Fitz-Gibbon et al., 2013; Tomida et al., 2013), while resistance of certain P. acnes strains to phage
some strains are enriched in healthy skin (Fitz- infection, yet noting that some of their observations
Gibbon et al., 2013). were inconsistent with this theory.
In parallel, P. acnes phages are dominant bacter- Tobetterunderstandhowbacteriophagesmodulate
iophages in the pilosebaceous unit (Fitz-Gibbon the bacterial composition of the skin microbiota and
et al., 2013). It has been known for over 50 years their potential roles in skin health and disease, in
that P.acnesphages exist onthe humanskin(Brzin, this study, we determined the diversity of P. acnes
1964). They have the morphology of siphoviruses, phages and their interactions with bacterial hosts in
consisting of a ~50nm icosahedral head and a theskinofacnepatientsandhealthyindividuals.We
~150nm flexible tail (Farrar et al., 2007). Zierdt sequenced the genomes of 48P. acnes phages
et al., (1968) isolated phage 174 from spontaneous isolated from 37 individuals and investigated
plaques of a P. acnes isolate. Phage 174 was able to whether certain phage strains dominate the skin
lyse nearly all P. acnes strains tested in the study. microbiota. By analyzing the skin metagenome data
Subsequently, more P. acnes phages were isolated, from the Human Microbiome Project (HMP), we
which exhibited lytic as well as pseudolysogenic further verified our conclusions from analyzing
behavior (Lood and Collin, 2011). However, in the sequenced phage isolates. We also challenged a
past decades, the study of P. acnes phages had been panel of 74 genetically defined Propionibacterium
limited to the development of phage typing systems strains against 15 of the sequenced phages to
to distinguish different serotypes of P. acnes determine the outcome and mechanisms of their
(Jong et al., 1975; Webster and Cummins, 1978). interactions.
Recent sequencing of 14P. acnes phages (Farrar
et al., 2007; Lood and Collin, 2011; Marinelli et al.,
2012) suggested that they have limited genetic Materials and methods
diversity with over 85% nucleotide identity in the
genome. All sequenced genomes are similar in size Phage isolation and DNA extraction
and structure with 45–47 genes encoded in two Skinfolliclesampleswerepreviouslycollectedfrom
oppositely transcribed regions named the left arm the nose of acne patients and individuals with
and right arm (Farrar et al., 2007; Lood and Collin, healthy skin as reported in the study by Fitz-
2011; Marinelli et al., 2012). Gibbon et al., (2013). To best represent the diversity
Much is to be learned about whether bacterio- of populations and history of medical care, the
phages drive the diversity and dynamics of the skin subjects were recruited from private practice, man-
bacterial community. The ratio between P. acnes aged care and public hospital settings, as well as
phages and P. acnes was ~1:20 in pilosebaceous outsideofdermatologyclinicsinSouthernCalifornia.
units,basedonapooledhealthyskinsamplethatwe Written informed consent was provided by all study
analyzed previously (Fitz-Gibbon et al., 2013), but subjects.
canvaryinalargerangeamongindividualsandover Thefolliclecontentscollectedonthesurfaceofthe
time. P. acnes phages do not encode integrases in nose strip were mashed using a sterile loop (Fish-
their genomes (Farrar et al., 2007; Lood and Collin, erbrand, Pittsburgh, PA, USA), and plated onto a
2011), suggesting their inability to stably integrate blood agar plate (Teknova Brucella Agar Plate with
into the host chromosome. They can kill the host HeminandVitaminK,Teknova,Hollister,CA,USA).
bacteriathroughcelllysisorcanenterapseudolyso- The sample plates were incubated at 37°C for
genicstateinthehoststrain(LoodandCollin,2011), 5–7daysanaerobicallyusingtheAnaeroPackSystem
in which the phage DNA persists in infected cells (Mitsubishi Gas Chemical Company, Tokyo, Japan)
withoutlysingthehostorintegratingintoitsgenome. (Fitz-Gibbon et al., 2013).
Whether P. acnes phages modulate the relative Phageplaquesobservedonthecultureplateswere
abundancesofdifferentP.acnesstrainsbyselectively isolatedby puncturingthe agarwith asterilepipette
killingspecificstrainsofP.acnesandthusplayarole tip and resuspending each tip in 50μl SM buffer
in skin health and disease is unknown. (0.1M sodium chloride, 8mM magnesium sulfate
TheISMEJournal
P.acnesphagepopulationintheskinmicrobiota
JLiuetal
2080
heptahydrate, 1M Tris-HCl, pH 7.5, 2% gelatin and determiningtheoverlappingintervalsbetweenallof
1mM calcium chloride). Each phage resuspension the 61 coordinate sets. The core region sequences
was spread onto A-media plates (12gl−1 pancreatic were concatenated for the subsequent multiple
digest of casein, 12gl−1 yeast extract, 22.2mM sequencealignments.Singlenucleotidepolymorphisms
D-glucose,29.4mMgl−1potassiumphosphatemono- (SNPs) in the core regions were identified by using
basic, 8mM magnesium sulfate heptahydrate and 20 the ‘show-snps’ option of Nucmer with the default
gl−1 agar) with top agar containing P. acnes strain setting.Inaddition,thesetofnon-synonymousSNPs
ATCC6919. After incubation at 37°C for 2 days, was obtained by masking the third codon positions
single plaques were selected and propagated using in the coding regions. Using MEGA5 (Tamura et al.,
the same host strain, medium and incubation 2011), phylogenetic trees were constructed by the
conditions. Suspensions of each phage isolate were Neighbor Joining method from p-distances based on
prepared by eluting plates with 8ml SM buffer at all SNP sites in the core regions or only the non-
roomtemperature,filteringwitha0.22-μmPESfilter synonymousSNPs.Bootstrapping wasperformedon
(Millipore, Billerica, MA, USA) to remove bacterial 1000 replicates.
cells, and storing at 4°C. Phage titers were deter-
mined by plaque assay.
Analysis of nucleotide polymorphism within single
Phage DNA extraction was performed using the
phage strains
Lambda Mini Kit (Qiagen, Valencia, CA, USA) with
The genetic variation within a phage strain was
the following modifications. Phage particles were
measured by the number of SNPs found in the
precipitatedinBufferL2bycentrifugationat20000g
sequencingdatafromaclonalphagepopulation.The
at 4°C for 1h. Extracted DNA was precipitated SNPs were identified as sites in a strain’s genome
overnight at −20°C before centrifugation. assemblythatwerecoveredby⩾30readswithPhred
quality score ⩾30 and with ⩾10% of these reads
Phage electron microscopy differing from the consensus sequence.
Copper grids (400 mesh formvar per carbon film)
(TedPella,Redding,CA,USA)wereglowdischarged.
Analysis of metagenomic shotgun sequencing data of
Phagecultureswereapplied,followedbyawashwith
0.22-μm-filtered water. The samples were stained the skin microbiota
Phage diversity was analyzed in the metagenomic
with 1% uranyl acetate and examined under a JEOL
shotgun sequencing data from 27 retroauricular
JEM-1200EX electron microscope (JEOL, Peabody,
crease samples collected in the HMP (The Human
MA, USA) with an accelerating voltage of 80kV.
Microbiome Project Consortium, 2012). The SRA
accessions of these samples are SRS013261,
Phage genome sequencing, assembly and annotation
SRS024598, SRS013258, SRS024596, SRS019016,
Phage genomes were sequenced in multiplex using
SRS019015, SRS019033, SRS019063, SRS019064,
the Roche GS FLX Titanium (Roche, Branford, CT,
SRS019081, SRS024655, SRS024620, SRS020263,
USA) or the Illumina MiSeq (Illumina, San Diego,
SRS020261, SRS017851, SRS017849, SRS057083,
CA, USA) platform. Sequence reads were initially
SRS024482, SRS045606, SRS058221, SRS018978,
assembled using MIRA 3.2.1 (Chevreux et al., 1999),
SRS058182, SRS016944, SRS046688, SRS015381,
and the resulting contigs were manually finished in
SRS052988andSRS019116.Accesstothephenotype
Consed 23.0 (Gordon et al., 1998). Some phage
data for this study (phs000228.v3.p1) was obtained
genomes required additional PCRs and amplicon
from dbGaP. MIRA (Chevreux et al., 1999) was used
sequencing to fill the gaps between contigs. Fully
to identify phage reads in each data set by mapping
assembled phage genomes were annotated using
against the P. acnes phage PA6 genome. Parameters
Genemark.hmm (Lukashin and Borodovsky, 1998) similar to thedefaultwere used: ⩾20 ntoverlapand
and Glimmer v3.02 (Delcher et al., 1999) with ⩾60% identity. To estimate the number of P. acnes
manual corrections. All phage genome sequences
phage strains in each data set, we first performed
have been deposited in GenBank under BioProject
a de novo assembly using the extracted phage reads
PRJNA173665 with accession numbers JX570702-
from the metagenomic data. We then aligned all the
JX570714, KJ578758-KJ578792.
phage reads to the resulting contigs to identify SNPs
in the core genome regions. The assemblies and
alignments were manually inspected using Consed
Genome analysis and phylogenetic tree construction
(Gordon et al., 1998). The same criteria used for
Sequences present in all 62 phage genomes were
nucleotide polymorphism identification in single
defined as core regions of the phage genome. To
phage strains were applied as described above.
identify these core regions, we first generated
alignments between the PA6 genome and each of
the other 61 phage genomes using Nucmer (Kurtz Analysis of phage genes under diversification
et al., 2004). This yielded 61 sets of starting and Multiple sequence alignments of Group VI and
ending coordinates describing intervals within the Group VIII phages and their related phages
PA6 genome that align with a given phage genome. (PHL037M02 and PHL073M02) were generated
Wethencalculatedthecoreregionsforallphagesby using MAFFT (Katoh et al., 2002). The positions of
TheISMEJournal
P.acnesphagepopulationintheskinmicrobiota
JLiuetal
2081
all mismatches and gaps were recorded. Sites The PCR was run under the following conditions:
of discrepancy were plotted in Artemis (Rutherford initial denaturation at 94°C for 5min, 35 cycles of
et al., 2000). denaturation at 94°C for 45s, annealing at 53°C for
35s and extension at 72°C for 1min, with a final
extension at 72°C for 10min.
Propionibacterium culture P.acnesATCC6919cultures,whichwerere-grown
P.acnes,P.humerusii,P.granulosumandP.avidum
afterlyticinfectionwithspecificphages,weretested
strains were cultured under anaerobic conditions in forsuperinfectionimmunitybypassagingsequentially
Clostridial Reinforced medium (Oxoid, Thermo
two to four times without further phage infection.
Fisher Scientific, Waltham, MA, USA) at 37°C for
Phage resistance was assayed using the same cross-
4–6 days. Propionibacterium cultures were used to streak method described above. The presence of
prepare top agar overlays for phage culture on
phage DNA in re-grown cultures was determined by
A-media plates.
PCR using the primers targeting the phage gp11
gene (Forward 5′-GGCTGGAACACGTAAAGCG-3′,
Reverse 5′-CACGATCGATCAACTCAACC-3′). The
Phage resistance test
PCR was run under the following conditions: initial
The resistance/susceptibility of Propionibacterium
denaturation at 94°C for 5min, 35 cycles of
strains against phages was determined using
denaturation at 95°C for 45s, annealing at 58°C for
a modified cross-streak assay. Fifteen of the 48
35s and extension 72°C for 1min, with a final
newly sequenced phages were randomly chosen for
extension at 72°C for 10min.
theanalysis.Twosetsofphagesthateachbelongsto
the same group, PHL010M04 and PHL066M04 in
Group VIII and PHL115M02, PHL085N00, and
PHL085M01 in Group VI, were included. The CRISPR analysis
bacterial strains were streaked across in A-media CRISPR spacer sequences were previously identified
plates, along with ATCC6919 on the same plate as inP.acnesgenomes(Fitz-Gibbonetal.,2013;Tomida
acontrol.Approximately 5μlof106PFUml−1phage et al., 2013). The spacer sequences were aligned
suspension was spotted onto each bacterial streak. against all phage genomes using BLASTn. Protospa-
The plates were incubated at 37°C anaerobically for cers with up to two mismatches were identified.
2 days. At least five replicates of each cross-streak
experiment were performed to determine whether
the strains were susceptible or resistant. Results
For the strains that showed resistance in the
Phage isolation and genome features
modified cross-streak experiment, we further
In an effort to determine the diversity of the skin
quantitatively determined the resistance by assaying
microbiota,wepreviouslycollected203skinsamples
the efficiency of plaquing of the phages relative to
from 179 individuals, including 94 samples from
P.acnesstrainATCC6919,calculatedasthefollowing:
acnepatientsand109samplesfromhealthyindividuals
1 with clear skin (Li, 2010; Fitz-Gibbon et al., 2013).
¼
Resistance Twenty-four individuals were sampled twice over a
Efficiencyofplaquing 4–6 month period. When we cultured the skin
¼ TiterofphagestrainXonATCC6919 samples for bacteria, we observed phage plaques in
TiterofphagestrainXonbacterialstrainY 49 of these samples: 14 from acne patients and
35 from healthy individuals. P. acnes phages were
We considered a 100-fold or greater increase in
efficiency of plaquing to be evidence of resistance. found more frequently in samples from healthy
individuals than from acne patients. Our rate of
The plaques on cross-streak plates were visually
phage detection (24% of investigated samples) is
inspected by one person and scored for turbidity
similartothosepreviouslyreported,rangingfrom26
based on the re-growth of the bacteria after plaque
formationusingthefollowingscale:0=clear,1=little to 30% (Marples et al., 1973; Puhvel and Amirian,
1979). The phages that we isolated have the
to no re-growth, 2=mild re-growth, 3=moderate
morphology of siphoviruses as previously described
re-growth and 4=heavy re-growth. The average
plaqueturbidityscoreofallthestrainsfromthesame (Farrar et al., 2007). A representative electron micro-
graph is shown in Supplementary Figure S1. Among
P. acnes clade was calculated for each of the tested
the phage isolates obtained from these samples, we
phages and was compared among different clades.
selected 21 phages from acne patients and 27 from
healthy individuals for whole genome sequencing
Pseudolysogeny characterization using 454 or MiSeq platforms (Supplementary
PCR was performed on phage suspensions using the Table S1). In some samples, multiple phage plaques
primers annealing to the ends of the phage genomes wereisolatedandselectedforsequencing.Allphage
(Forward 5′-CCGAAGCCGACCACATCACACC-3′, genomes were assembled, completed and annotated
Reverse 5′-TCATCCAACACCTGCTGCTGCC-3′) to (Supplementary Figure S2). A representative phage
determine whether phage genomes are circularized. genome is shown in Figure 1.
TheISMEJournal
P.acnesphagepopulationintheskinmicrobiota
JLiuetal
2082
The P. acnes phage genomes are highly similar to not used in the genome assembly process. We
eachother(SupplementaryFigureS2).The48phage identified the nucleotide positions with a minor
genomes have comparable sizes (29.0–29.8Kb) and allele frequency X10% and covered by at least 30
GC contents (53.7–54.5%) (Supplementary Table reads with Phred quality ⩾30. We found that each
S1). The sequence identity between any pair of phage genome contains 0–11 polymorphic sites
genomes ranges from 85.2 to 100%. On average 45 (Supplementary Table S1). The number of poly-
openreadingframeswerepredictedineachgenome. morphisms in each assembled genome did not
Consistent with previous reports (Farrar et al., 2007; increase beyond 11 sites despite the large numbers
Lood and Collin, 2011; Marinelli et al., 2012), these of sequencing reads obtained (up to 9120× genome
open reading frames were arranged compactly coverage). From this analysis, we conclude that the
within the left and right arm regions of the genome background level of genetic polymorphism within a
(Figure1,SupplementaryFigureS2).Ouranalysisof clonal P. acnes phage isolate is ~11bp. Thus, phage
the 48 new phage genomes supports the annotation isolateswithasimilarorsmallernumberofnucleotide
of the gp22/gp23 locus as a single open reading differences throughout the entire genome can be
frame (495 to 522bp) on the minus strand considered as belonging to the same phage strain.
(Supplementary Figure S2). This is different from Sincethephageswithineachgroup(I–IX)differbyat
previous annotations based on a small number of most 14bp (Supplementary Table S2), they likely
genomes(Farraretal.,2007;LoodandCollin,2011). represent clones of the same phage strain.
Phylogenetic relationships among the phage genomes Diversity of P. acnes phages in the human skin
To determine the genome diversity of P. acnes The relationships among the phages within each
phages, we compared 62 sequenced phage genomes, group (I–IX) provide insights on the diversity of
including our 48 phage genomes and 14 previously P.acnesphagesintheskinmicrobiota.Weobserved
published genomes (Farrar et al., 2007; Lood and three types of relationships among the highly
Collin, 2011; Marinelli et al., 2012). Similar to their similar phages within the groups (Figure 2 and
bacterialhost,P.acnesphageshavelimitedgenomic Supplementary Table S2). First, phages within the
diversity.All62phagesarehighlysimilaringenome same group were isolated from the same sample of
sequence. The core regions, which are shared by all the same individual. These include PHL067M01,
sequencedgenomes,consistof22348bp(76%ofthe PHL067M09andPHL067M10inGroupI;PHL082M00,
averagegenomelength)andcontain7232SNPs.The PHL082M02,PHL082M03andPHL082M04inGroupII;
average distance among the phages was 0.257 PHL064M01 and PHL064M02 in Group V and
(substitution rate at the SNP sites). A phylogenetic PHL116M00 and PHL116M10 in Group IX. Only one
tree constructed based on these SNPs (Figure 2), or other pair of phages, PHL117M00 and PHL117M01,
only the non-synonymous SNPs (Supplementary which were isolated from the same sample, was not
Figure S3), shows that no particular phylogenetic the same strain. Our data suggest that while an
clades were found among the phages. individual microbiota can harbor multiple strains of
Despite the lack of phylogenetic lineages among phages,itislikelymorecommonthatonestrainofP.
the P. acnes phage genomes, we observed several acnes phage dominates the phage population. Sec-
groups (I–IX), each of which consists of nearly ond, phages within the same group were isolated
identical phages (Figure 2). The phages within each fromdifferentsamplesofthesameindividualsovera
group (I–IX) differ by no greater than 14bp within periodof14to21weeks.TheseincludePHL085M01
the entire 29kb genome (Supplementary Table S2). and PHL085N00 in Group VI; PHL114L00 and
As a comparison, the average pairwise difference PHL114N00 in Group VII and PHL151M00 and
among all 62 phages is 3176bp. To determine PHL151N00 in Group VIII. All paired phages from
whether the nearly identical phages within each our longitudinal samples were nearly identical to
group represent clones of the same phage strain, we each other. This suggests that the same phage strain
estimated the frequency of genetic polymorphism in can persist in an individual skin microbiota. Third,
each of our phage isolates. We mapped all available some phages within the same group were isolated
sequence reads of each phage to its assembled from different individuals, such as the phages in
consensus genome, including the sequence reads GroupsIII,IV,V,VIandVIII.Ofthe43uniquephage
PHL009M11
Figure 1 A representative genome of the newly sequenced P. acnes phages. The annotated genome of PHL009M11 is shown as a
representativeofthe48newlysequencedP.acnesphagegenomes.Onaverage45openreadingframesareencodedineachphagegenome.
TheISMEJournal
P.acnesphagepopulationintheskinmicrobiota
JLiuetal
2083
I
II
IX
VIII
*
* III
*
*
* IV
VII
From acne patient
From subject with clear skin
VI
From subject with unknown acne status V
* From siblings
Figure2 P.acnesphagesarehighlysimilartoeachotherwithnosignificantphylogeneticlineagesobserved.Aphylogenetictreeofthe
62 currently sequenced phage genomes was constructed based on the 7232 SNPs in the core regions. No significant lineages were
observed.Ninegroupsofnearlyidenticalphagesareindicated,highlightingthatthephageswithineachgroupbelongtothesamestrain.
Brancheswithbootstrapvalueso80(basedon1000resamplings)werecollapsed.
strains represented by all 62 isolates sequenced to metagenomic shotgun sequencing data collected
date, 5 strains from our study currently show from healthy individuals in the HMP (The Human
evidence of inhabiting more than one individual. Microbiome Project Consortium, 2012). Although
This suggests that a pool of common P. acnes phage P.acnesphageswerepreviouslyfoundinmetagenomic
strainsexistsinthehumanpopulation.Interestingly, sequencing data, their diversity was not analyzed.
among the five shared strains, two inhabited related This is the first time that the population diversity of
individuals. In Group IV, the two nearly identical P.acnesphagesintheskinmicrobiomeischaracter-
phages,PHL150M00andPHL308M00,wereisolated ized. Among the available 27 HMP skin samples,
from two brothers. Two of the four phages in Group 9 were collected from the left retroauricular crease
VIII,PHL010M04andPHL066M04,andtheirclosely and 18 were collected from the right retroauricular
related phage strain PHL073M02 were isolated from crease (Supplementary Table S3). Some samples
three siblings (Figure 2 and Supplementary Table were collected from the same individuals. We first
S2). This suggests that transmissions of skin bacter- extractedtheP.acnesphagereadsfromeachsample.
iophages, either directly or via the transmission of The number of phage reads was from 24–612512
phage-carrying bacterial hosts, are likely to occur (0–2060× coverage on PA6 genome, Supplementary
between related individuals. Table S3), independent of the sequencing depth of
each sample, suggesting that the relative abundance
ofP.acnesphagesintheskinmicrobiotacanvaryin
P. acnes phage populations in the skin microbiome a large range among individuals.
To validate that our above findings on the phage To determine P. acnes phage populations in
diversity in the human skin microbiota were not individual skin microbiota, we next assembled
biased due to isolated phages, we analyzed the skin P. acnes phage genomes in each metagenomic
TheISMEJournal
P.acnesphagepopulationintheskinmicrobiota
JLiuetal
2084
dataset.Sevensamples(25.9%oftotalsamples)had The assembled phage genomes in the remaining
a modest to high sequencing coverage of P. acnes threesamples,HMP03,HMP08andHMP24,contained
phages (17×–2060×) (Supplementary Table S3), o20 SNPs, which are in the range of the SNPs found
which allowed metagenomic assembly. We per- in a single phage strain. Due to their lower phage
formed the SNP analysis on the assembled phage coverages, it is yet inconclusive whether single or
genomes using thesame criteriaasin the analysis of multipledominantphagestrainswerepresentinthese
complete phage genomes described above. To eval- communities.
uate the effect of sequencing depth on the detection Our analysis of the phage diversity from the HMP
rate of SNPs, for samples with 430× phage cover- metagenomic data showed that among the skin
age, we repeated the phage genome assembly and communities that had detectable P. acnes phages,
SNP analysis using only portions of phage reads at three harbored only one dominant strain while one
various coverages (20×,100×,250×,500×,1000× harbored two different strains. These findings are
and 1500×), as applicable. One of the samples, consistent with our conclusions based on the
HMP20, had 41000 SNPs in the core regions of the isolated phages from our study cohort.
assembled phage genome, which leveled off to 1353
sites when all the sequence reads (538×) were used
intheassembly(Figure3,SupplementaryTableS4). Phage genes under diversification
ThissuggeststhattheP.acnesphagepoolinHMP20 Phylogenetically related strains can reflect phage
wasadequatelysampledandthatitlikelyconsistsof diversification under selection. Two phages,
two dominant phage strains based on phage genome PHL037M02 and PHL073M02, are highly related to
comparison (Supplementary Materials). the members of Groups VI and VIII, respectively
On the other hand, three other samples, HMP04, (Figure 2). However, they contain many more
HMP09 and HMP15, each had six or fewer SNPs in nucleotide variations than 11bp, and thus are
thecoregenomeregions,despitehavingsubstantially considered separate strains based on our criterion.
higher sequencing coverages than HMP20. This Nonetheless, their high degree of similarity to the
suggests that these samples each harbor only one membersofthesegroupsmayreflectrecentselective
dominantP.acnesphagestrain.Thisresultsupports pressuresdrivingphagediversification.Weidentified
our earlier conclusion that an individual skin 160 sites of nucleotide variations between
microbiotaisoftendominatedbyoneP.acnesphage PHL037M02andtheGroupVImembers,50ofwhich
strain. We were able to assemble high-quality draft are non-synonymous. All but one of these genetic
genomes of P. acnes phages from these three differences is located in a region encoding Gp16,
metagenomic data sets, which cover 96.4–96.7% of Gp17andGp18,asannotatedinthegenomeofphage
the core genome regions. They are typical of the PA6(SupplementaryFigureS5A).Theexactfunctions
P.acnesphagestrains,withhighsimilaritiestothe62 of these genes are unknown, but their location near
sequenced genomes. A phylogenetic tree including the 3’ end of the left arm between structural protein
these three new genomes with the 62 isolated phage genes and lysis protein genes suggests that they
genomes is shown in Supplementary Figure S4. could encode late-acting proteins. Based on the
SamplesHMP04andHMP09werecollectedfromthe McDonald–Kreitman test (Egea et al., 2008), gene
left and right retroauricular crease of the same gp17 is under selection (P=0.011).
individual, and the phage genomes assembled from We identified 81 sites of nucleotide variations
these two samples are highly similar, potentially between PHL073M02 and the Group VIII members.
originating from the same strain. The sequence differences lie primarily within the
region encoding an endolysin and a putative type II
holin (Gp20 and Gp21, Supplementary Figure S5B).
These lytic cycle proteins permeabilize the cell
e core enome mlayemerbratnoeanredledaesgeradenethweexptrhaacgeellulapraprteipcltiedsoglfyrcoamn
NPs detected in thssembled phage g 11,,04600200000 HHHHHMMMMMPPPPP0120095043 tPPtthhhHHereeLLye00lb71iska30iecbMMltlyei00nr24oigarsliagnlaiindnvhaidontPsegtHd.tiLwfnr0Aoo6tmsh6MetGhp0srera4oem,suvapeiwmoeuhersoeVlauynIsIicsIeoehmsloatmelrtdaenel.dtmipTobfhnhreaoeurgmdsse,,,
Sa
of of 10 HMP08 strain. This suggests that the endolysin and holin
Number regions 0 0 500 1,000 1,500 2,000 2,500 HMP24 gafiernendeisun,ngw,diehnricahrlalapsrieedqueseesvneocnleutditaiplohnfoa.rgeCpsoh,nawsgieestfmeonuutnltdiwpfliritechqatuiteohnnist,
Sequencing coverage
amino acid variations in these two gene products.
Figure3 P.acnesphagepopulationintheskinmicrobiotaisoften This suggests that mechanisms determining host
dominatedbya singlestrain. Thenumbers of SNPs identified in
bacteriumlysisspecificityandkineticsmaybeunder
thecoreregionsofP.acnesphagegenomesareshownatdifferent
selection in these phages. The sequence variation
sequencing coverages. The phage genomes were assembled from
theHMPmetagenomicshotgunsequencingdata. sites in these two genes among the phages may be
TheISMEJournal
P.acnesphagepopulationintheskinmicrobiota
JLiuetal
2085
potentialtargetsforphageengineeringtomanipulate IB-3showthatthesestrainsencodecomponentsofa
their lytic activities against bacterial hosts. restriction modification system (genes PPA1611 and
The two highly similar phage genomes assembled PPA1612 in KPA171202). This may explain their
from the HMP samples, HMP04 and HMP09, which resistance to phages. Among the nine type II strains,
werecollectedfromthesameindividual,differedby two strains, HL001PA1 and HL042PA3, were highly
222nucleotides.Mostofthesevariationsarecentered resistant to some of the phages. This is consistent
at the 5′ end of the right arm of the genome, which with previous observations that strains of this type
encodes putative regulatory or DNA-binding proteins were more frequently resistant to phages (Webster
of largely unknown functions. and Cummins, 1978). The resistance to phages
observed in type II strains could be partially
attributed to the CRISPR mechanism encoded in
Range and specificity of Propionibacteria–phage their genomes, which is addressed below. The only
interactions type III strain tested, HL201PA1, was resistant to all
To determine whether P. acnes phages modulate the 15 phages. Since type III strains are not commonly
relativeabundancesofdifferentP.acnesstrainsinthe foundontheskinoftheface,itispossiblethatthese
skinmicrobiotabyselective killing, wecharacterized P. acnes phages isolated from the face have not yet
the host range and specificity of P. acnes phages.We evolved a mechanism to infect type III strains.
tested 15 of the 48 sequenced phages against a panel To determine whether P. acnes phages modulate
of74Propionibacteriumstrains,including67P.acnes the abundance and diversity of other species in
strains,3P.humerusiistrains,1P.granulosumstrain addition to P. acnes in the skin microbiota, we
and 3 P. avidum strains. Except for the P. acnes investigated the host range of P. acnes phages in
strains KPA171202, ATCC11828, HL201PA1 and related species. Strains of other human skin-
HL202PA1, all of these Propionibacterium strains associated Propionibacteria, including 3 strains of
were isolated from the same cohort of subjects P.humerusii,1strainofP.granulosumand3strains
sampled for phages. The genomes of all 67P. acnes of P. avidum, were tested against the 15 phages
strains and 3 P. humerusii strains have been (Figure 4b). P. humerusii is a newly defined species
sequenced (Fitz-Gibbon et al., 2013; Tomida et al., (Butler-Wu et al., 2011). In our previous study,
2013). Our bacterial collection included all major P. humerusii was one of the major species found on
lineages of P. acnes found on the human skin, with the skin with a relative abundance of 1.9% in the
multiplestrainsrepresentingeachofthemajorclades, pilosebaceous unit based on 16S ribosomal RNA
IA-1, IA-2, IB-1, IB-2, IB-3 and II, as well as one type analysis (Fitz-Gibbon et al., 2013). It is closely
III strain. We constructed a phylogenetic tree of the related to P. acnes with 498% identity in the 16S
67P. acnes strains based on the SNPs in their core ribosomal RNA gene sequence. P. granulosum and
genomic regions (Figure 4a) (Tomida et al., 2013). P.avidumarecommonskincommensals(Cummins,
Usingamodifiedcross-streakmethod,wedetermined 1976; Ördögh and Hunyadkürti, 2013). While all
theresistance/susceptibilityofeachofthe74bacterial tested P. granulosum and P. avidum strains showed
strains against the 15 phages. In total, 1110 bacter- strong resistance to all the phages, two P. humerusii
ium–phage interactions were measured. Each experi- strains, HL037PA2andHL037PA3,were susceptible
ment was repeated a minimum of five times. For the to all the phages tested. The third P. humerusii
bacterialstrainsthatshowedresistancetophages,we strain, HL044PA1, was susceptible to 10 of the 15
determinedthefoldincreaseinresistancebymeasuring phagestested.Ourresultsshowthatthehostrangeof
efficiency of plaquing (EOP) relative to the P. acnes P. acnes phages is not limited to P. acnes but also
strain ATCC6919, which is known to be susceptible includes a closely related Propionibacterium species,
to all tested phages. suggesting that P. acnes phages may also be able to
We found that the outcome of the bacterium– modulate P. humerusii populations in the skin
phageinteractionsisP.acneslineagedependent.All microbiota.
type I P. acnes strains (clades IA-1, IA-2, IB-1 and
IB-2) except clade IB-3 were susceptible to all tested
phages (Figure 4a). Among them, the phages often P. acnes phages can adopt a pseudolysogenic state
formed turbid plaques on P. acnes strains of clade depending on the host P. acnes strains
IA-1, but clear plaques on strains of clades IB-1 and IthasbeensuggestedthatP.acnesphagesmayenter
IB-2, as summarized in Figure 5. This suggests that pseudolysogeny as an alternative to the lytic cycle
thesephagesengageintwodifferentstatesdepending (Farrar et al., 2007; Lood and Collin, 2011). As
on the host strains: a pseudolysogenic response in describedabove,wediscoveredthatP.acnesphages
cladeIA-1strains,andalyticcycleincladeIB-1and often adopt a pseudolysogenic state in clade IA-1
IB-2 strains. strains, but rarely in clade IB-1 and IB-2 strains,
CertainP.acnesstrainsofcladesIB-3,IIandIIIare suggesting that their pseudolysogeny is dependent
highly resistant to multiple phages. Two strains of on the host P. acnes strains (Figure 5). In support of
clade IB-3 (KPA171202 and HL030PA1) were highly the existence of pseudolysogeny in P. acnes phages
resistant to most of the tested phages with a ⩾100- and consistent with the result by Marinelli et al.,
fold increase in resistance. The genomes of clade (2012), our genome sequencing data revealed
TheISMEJournal
P.acnesphagepopulationintheskinmicrobiota
JLiuetal
2086
Phage Strains
Group VI Group VIII
Proapcionnesib aSctrtaeirnium Clade Ribotype CRISPR/Cas PHL071N05 PHL113M01 PHL111M01 PHL082M00 PHL060L00 PHL067M10 PHL112N00 PHL037Z02 PHL115M02 PHL085N00 PHL085M01 PHL114L00 PHL073M02 PHL010M04 PHL066M04
HL036PA1 532 - S S S S S S S S S S S S S S S
HL036PA2 532 - S S S S S S S S S S S S S S S
HL036PA3 1 - S S S S S S S S S S S S S S S
HL005PA3 1 - S S S S S S S S S S S S S S S
HL005PA2 1 - S S S S S S S S S S S S S S S
HL020PA1 1 - S S S S S S S S S S S S S S S
HL027PA2 1 - S S S S S S S S S S S S S S S
HHLL100103PPAA12 IA-1 11 -- SS SS SS SS SS SS SS SS SS SS SS SS SS SS SS
HL087PA2 1 - S S S S S S S S S S S S S S S
HL063PA1 1 - S S S S S S S S S S S S S S S
HL072PA2 5 - S S S S S S S S S S S S S S S
HL072PA1 5 - S S S S S S S S S S S S S S S
HL046PA2 1 - S S S S S S S S S S S S S S S
HL002PA2 1 - S S S S S S S S S S S S S S S
HL002PA3 1 - S S S S S S S S S S S S S S S
HL078PA1 1 - S S S S S S S S S S S S S S S
HL106PA2 1 - S S S S S S S S S S S S S S S
HL099PA1 4 - S S S S S S S S S S S S S S S
HL083PA1 1 - S S S S S S S S S S S S S S S
HL038PA1 4 - S S S S S S S S S S S S S S S
HL074PA1 4 - S S S S S S S S S S S S S S S
HHLL000455PPAA11 IA-2 44 -- SS SS SS SS SS SS SS SS SS SS SS SS SS SS SS
HL007PA1 4 - S S S S S S S S S S S S S S S
HL096PA1 5 - S S S S S S S S S S S S S S S
HL043PA1 5 - S S S S S S S S S S S S S S S
HL043PA2 5 - S S S S S S S S S S S S S S S
HL053PA1 4 - S S S S S S S S S S S S S S S
HL056PA1 4 - S S S S S S S S S S S S S S S
HL025PA1 1 - S S S S S S S S S S S S S S S
HL086PA1 8 - S S S S S S S S S S S S S S S
HL082PA1 8 - S S S S S S S S S S S S S S S
HHLL101503PPAA22 IB-1 88 -- SS SS SS SS SS SS SS SS SS SS SS SS SS SS SS
HL092PA1 8 - S S S S S S S S S S S S S S S
HL110PA1 8 - S S S S S S S S S S S S S S S
HL030PA2 3 - S S S S S S S S S S S S S S S
HL063PA2 3 - S S S S S S S S S S S S S S S
HL037PA1 3 - S S S S S S S S S S S S S S S
HL059PA1 16 - S S S S S S S S S S S S S S S
HL059PA2 16 - S S S S S S S S S S S S S S S
HL025PA2 3 - S S S S S S S S S S S S S S S
HL005PA4 3 - S S S S S S S S S S S S S S S
HL067PA1 3 - S S S S S S S S S S S S S S S
HL002PA1 IB-2 3 - S S S S S S S S S S S S S S S
HL027PA1 3 - S S S S S S S S S S S S S S S
HL046PA1 3 - S S S S S S S S S S S S S S S
HL083PA2 3 - S S S S S S S S S S S S S S S
HL013PA1 3 - S S S S S S S S S S S S S S S
HL050PA1 3 - S S S S S S S S S S S S S S S
HL050PA3 3 - S S S S S S S S S S S S S S S
HL087PA1 3 - S S S S S S S S S S S S S S S
HL087PA3 3 - S S S S S S S S S S S S S S S
KHPLA013701P2A012 IB-3 11 -- SS > S10 > S10 SS > S10 >> 1100 >> 1100 >> 1100 >> 1100 >> 1100 >> 1100 >> 1100 >> 1100 >> 1100 >> 1100
HL050PA2 1 - S S S S S > 10 S S S S S S S S S
HL060PA1 2 + S S S S S S S S S S S S S S S
HL103PA1 2 + S S S S S S S S S S S S S S S
HL082PA2 2 + S S S S S S S S S S S S S S S
AHTLC0C011P18A218 II 22 ++ SS SS > S10 SS SS > S10 SS SS SS SS SS > S10 SS SS SS
HL110PA3 6 + S S S S S S S S S S S S S S S
HL110PA4 6 + S S S S S S S S S S S S S S S
HL042PA3 6 + > 10 S > 10 > 10 S > 10 S > 10 > 10 > 10 > 10 > 10 > 10 > 10 > 10
HL202PA1 6 + S S S S S S S S S S S S S S S
HL201PA1 III N/A - > 10 > 10 > 10 > 10 > 10 > 10 > 10 > 10 > 10 > 10 > 10 > 10 > 10 > 10 > 10
S susceptible
10 fold increase in resistance RT1 RT2 RT3 RT4 RT5 RT6 RT8 RT16 RT532
Phage Strains
Group VI Group VIII
NStarmaine PropiSonpiebcaiectserium PHL071N05 PHL113M01 PHL111M01 PHL082M00 PHL060L00 PHL067M10 PHL112N00 PHL037M02 PHL115M02 PHL085N00 PHL085M01 PHL114L00 PHL073M02 PHL010M04 PHL066M04
HL037PA2 S S S S S S S S S S S S S S S
HL037PA3 P. humerusii S S S S S S S S S S S S S S S
HL044PA1 S S S S > 10 S > 10 >10 > 10 S S S S S > 10
HL078PG1 P. granulosum > 10 > 10 > 10 > 10 > 10 > 10 > 10 > 10 > 10 > 10 > 10 > 10 > 10 > 10 > 10
HL063PV1 > 10 > 10 > 10 > 10 > 10 > 10 > 10 > 10 > 10 > 10 > 10 > 10 > 10 > 10 > 10
HL083PV1 P. avidum R R R R R R R R R R R R R R R
HL307PV1 R R R R R R R R R R R R R R R
S susceptible
R resistant
10 fold increase in resistance
Figure4 HostrangeandspecificityofP.acnesphages.Resistance/susceptibilityofPropionibacteriumstrains(inrows)against15ofthe
48sequencedphages(incolumns)isshown.(a)Phageinfectionoutcomesof67P.acnesstrains.AllP.acnesstrainsincladesIA-1,IA-2,
IB-1andIB-2weresusceptibletothetestedphages,whilephageresistancewasfoundinstrainsofcladesIB-3,IIandIII,coloredinpink.
ThedendrogramtotheleftshowsthephylogeneticcladesofP.acnesstrains(Fitz-Gibbonetal.,2013).Onlytopologyisshown.(b)Phage
infectionoutcomesofthreeP.humerusiistrains,oneP.granulosumstrainandthreeP.avidumstrainsshowthatP.acnesphagescould
infectandlyseP.humerusiistrains,whileP.granulosumandP.avidumwereresistanttoalltestedphages.
TheISMEJournal
P.acnesphagepopulationintheskinmicrobiota
JLiuetal
2087
PHL111M01
3.5
PHL071N05
3 PHL060L00
core 2.5 PPHHLL011627NM0100
y s 2 PHL115M02
dit 1.5 PHL113M01
bi PHL085M01
Tur 1 PHL037M02
0.5 PHL114N00
PHL010M04
0 PHL066M04
IA-1 IA-2 IB-1 IB-2 II PHL073M02
(n=16) (n=14) (n=6) (n=17) (n=7)
P. acnes clade
Figure5 ThefrequencyofphagepseudolysogenyvariesamongdifferentP.acnesstrains.Theturbidityofthephageplaquesformedon
P.acnescultureplateswasexamined.Datafrom60P.acnesstrainschallengedby13phageswererecordedandsummarized.The60
P.acnesstrainsbelongtofiveclades:IA-1(n=16),IA-2(n=14),IB-1(n=6),IB-2(n=17)andII(n=7).Theaverageturbidityscoreofthe
bacterialstrainsineachcladechallengedbyeachphage(incolors)isshown.Independentofthephagestested,theplaqueturbidityscore
variedamongdifferentP.acnesstrains:itishighincladeIA-1andlowincladesIB-1andIB-2.Thissuggeststhatthephagestendtoentera
pseudolysogenicstateincladeIA-1strainsandalyticcycleincladesIB-1andIB-2strains.
the ends of the phage genomes to be flanked by 11- (Farrar et al., 2007; Lood and Collin, 2011), but
nucleotide single-stranded overhangs. Previous exists as an extrachromosomal element.
reports suggested that these overhangs may be Insummary,allaboveresultssuggestthatP.acnes
involvedinthecircularizationofthephagegenomes phages can adopt a pseudolysogenic state in clade
(Farrar et al., 2007; Lood and Collin, 2011). To test IA-1 strains.
thishypothesis,wedesignedPCRprimersannealing
to the ends of the phage genomes. A PCR product
spanning the two ends with the predicted size Resistancetobacteriophagesdoesnotcorrelatewiththe
(~735bp) and sequence was amplified from all the presenceofmatchingCRISPRspacersintypeIIP. acnes
phages tested, suggesting that the phage DNA can strains
existinacircularformmediatedbytheoverhangsat Since certain type II P. acnes strains are resistant to
the ends (Supplementary Figure S6A). phages, we next investigated whether the CRISPR/
To demonstrate the pseudolysogenic properties of Cas mechanism could explain the resistance of the
P. acnes phages, we infected P. acnes strain typeIIstrainsagainstphages.Amongthe67P.acnes
ATCC6919, a clade IA-1 strain (Liu et al., 2014), strains, 9 belong to type II and encode CRISPR/Cas
withfivedifferentphages(PHL060L00,PHL112N00, elements. Each of the type II strains has one to nine
PHL037M02,PHL073M02andPHL114L00).Follow- 33-bp spacers in their CRISPR arrays (Tomida et al.,
ing the formation of plaques by each phage, re- 2013). In total, they encode 36 spacers, 20 of which
growth of the bacteria was observed starting in the are unique. We identified 34 unique protospacers in
center of the plaque regions. The re-grown bacteria the15testedphagegenomesthatmatchanyofthe20
showed no evident lysis when challenged by 13 unique spacer sequences in the 9 type II P. acnes
phages that previously could lyse this P. acnes strains. Because the CRISPR/Cas system has been
strain, including the phages of the initial exposure shown to tolerate a limited number of mutations in
(Supplementary Figure S6B). This suggests that the protospacer targets (Semenova et al., 2011; Manica
bacterial host gained superinfection immunity after et al., 2013), in our analysis we allowed up to two
infection by these phages. mismatches for a sequence to be considered a
To further determine whether the phage DNA recognizable protospacer. Similar results were
exists as an episome in the bacterial host after obtainedwhenonlyperfectlymatchingprotospacers
infection, we tested the presence of phage DNA in were considered. We found that all identified
four distinct colonies of the re-grown ATCC6919 protospacers are located primarily on the left arm
culture that was initially infected with PHL060L00 ofthephagegenomes,whichismoreconservedthan
andsubsequentlypassaged.Twoofthefourcolonies therightarm(SupplementaryFigureS7).Inaddition,
produced an expected 437bp amplicon in a PCR the locations of the protospacers are generally
targetingthegp11gene(SupplementaryFigureS6C), conserved among all phage genomes that harbor
supportingthepresenceofphageinasubpopulationof the same protospacer sequences. These suggest that
there-grownP.acnesasanepisome.Inaddition,we the CRISPR/Cas system tends to target the more
sequenced the genomic DNA extracted from two conserved regions of the phage genomes.
ATCC6919 cultures, each of which was passaged In contrast to a prior report (Marinelli et al., 2012),
fromare-growncultureafterphageinfection.Inboth wefoundthattheresistance/susceptibilityofthenine
cases, the phage reads obtained from whole genome type II P. acnes strains against phages did not
sequencing were assembled into complete genomes correlate with the presence/absence of at least one
that are separate from host P. acnes contigs, phage-matching CRISPR spacer (r=0.39, Figure 6).
supporting earlier evidence that P. acnes phage There were multiple observations that even though
DNA does not integrate into the host genome the strain encodes a matching CRISPR spacer, it was
TheISMEJournal