Mahony, Douwe van Sinderen. 8. 9. 10. Supplementary Information. 11. 12. 13 ... UCC2003, and B. breve 017W4-39 cultures, grown for 8 h, and treated with a ...
1 2 3
Characterization and induction of prophages in human gut-associated Bifidobacterium hosts
4 5 6
Travis N. Mavrich, Eoghan Casey, Joana Oliveira, Francesca Bottacini, Kieran James, Charles
7
M.A.P. Franz, Gabriele A. Lugli, Horst Neve, Marco Ventura, Graham F. Hatfull, Jennifer
8
Mahony, Douwe van Sinderen
9 10 11 12 13 14
Supplementary Information
15
Supplementary Materials and Methods
16 17
Optimization of mitomycin C induction. Mitomycin C concentration was optimized using 96-
18
well microtiter plates. Wells with 500 µl RCM were inoculated from B. breve JCM 7017, B. breve
19
UCC2003, and B. breve 017W4-39 cultures, grown for 8 h, and treated with a 10-fold serial
20
titration of mitomycin C (ranging from 0.0003 µg/ml to 3 µg/ml) for 14 h. Growth inhibition was
21
observed for concentrations at and above 0.03 µg/ml. Similar inhibitory profiles were observed
22
with 0.03 µg/ml mitomycin C for 2.5 ml, but not 50 ml, cultures. Therefore, for 50 ml cultures, 0.3
23
µg/ml mitomycin C was used.
24 25
Induction verification using PCR. Prophage induction was confirmed by PCR amplification
26
across the attP site that forms after excision and circularization. One pair of primers was
27
designed to test dnaJ2-integrated phages (5’-TCCGTAAAAACAGGTTAAAAACCG-3’, 5’-
28
AAACGTTGGAATCACGCCATTCC-3’). Separate primer pairs were designed for Bb447phi1 (5’-
29
GTCACACCCACCAGAATCATGAATC-3’,5’-GTTAAGAAGACTTGCTGATGGAGTTG-3’) and
30
Bb423phi2 (5’-CGAACCACTGTGTCATCATCTC-3’, 5’-AGCGAGATAACTTGGACGATCAAC-
31
3’). Induction of previously described prophages in strains B. choerinum LMG 10510 and B.
32
moukalabense DSM 27321 was confirmed using previously described primers19. Amplification
33
proceeded in 25 µl reactions containing 1 µl filtered supernatant with Taq polymerase according
34
to manufacturer’s instructions, using a thermocycler protocol of 25-30 cycles of denaturation at
35
94oC for 30 s, annealing at 55oC for 30 s, and extension at 72oC for 1 min.
36 37
Plaque assays. Plaque generation was attempted using a variety of phage samples, indicator
38
strains, and growth media. To test for spontaneous phage release, filtered supernatants of
39
saturated cultures were used. To test for mitomycin C-induced phage release, filtered
40
supernatants or PEG precipitated samples from mitomycin C-treated cultures (as generated for
2
41
flow cytometry) were used. Confluent lawns were prepared by mixing 4.5 ml Reinforced
42
Clostridial Top Agar (30 ml Reinforced Clostridial Agar + 60 ml RCM, with or without 2 mM
43
CaCl2) with 200-300 µl saturated bifidobacterial culture (grown overnight directly from a freezer
44
stock) and allowed to solidify on RCA plates. For each phage sample, 3-5 µl was spotted onto
45
the overlay and allowed to dry, and plates were incubated at 37oC in an anaerobic chamber for
46
24-48 h. Phage samples were generated from several lysogens (B. choerinum LMG 10510 and
47
B. moukalabense DSM 27321) and predicted lysogens (B. breve 082W4-8, B. breve 180W8-3,
48
B. breve 139W4-23, B. breve 017W4-39, and B. breve 215W4-47a), and they were tested
49
against all of the originating lysogens and predicted lysogens as well as several non-lysogens
50
(B. breve JCM 7017, B. breve NCIMB 702258, and B. breve UCC2003). As an alternative to
51
spotting, some saturated cultures were directly mixed with phage samples in a 1.5 ml tube and
52
aerobically incubated on the bench at room temperature for 10-15 min prior to being mixed with
53
top agar and poured as an overlay. Additionally, TOS media (Sigma) was used as an alternative
54
to RCM.
55 56
Rin shufflon inversion analysis in uninduced Bb423phi1. Genomic inversions within the Rin
57
shufflon of the uninduced Bb423phi1 prophage were identified in previously reported B. breve
58
139W4-23 raw whole genome sequencing reads18. Pacbio long reads (average read length >10
59
kb) that map across the Rin shufflon locus (coordinates 1,307,900 to 1,310,888) with at least
60
80% sequence identity were selected using BLAT aligner v36x2. Variant shufflon orientations in
61
this subset of reads were identified using dotplot alignments in mummer v3.0. Sequence
62
coverage of each variant was computed using the identified long reads as reference sequences
63
and performing a mapping assembly using the RS_Resequencing.1 protocol implemented in
64
SMRT Analysis portal v2.3. The resulting assembled reads were inspected using the Next
65
Generation Sequencing (NGS) visualization tool Tablet (https://ics.hutton.ac.uk/).
66
3
67
Gene content flux analysis. Changes in gene content and nucleotide sequence similarity were
68
computed as previously described53. Briefly, for each bifidophage genome, pairwise nucleotide
69
distances to all other actinobacteriophages were computed using Mash, and gene content
70
dissimilarities were computed using pham designations in Phamerator. Each phage’s
71
evolutionary mode is predicted by assessing the distribution of pairwise genomic similarities
72
using previously determined mode boundaries.
73 74
4
75
Supplementary Figure Legends
76 77
Supplementary Figure S1. Mitomycin C treatment increases bifidoprophage excision.
78
Excision and circularization of the predicted prophages were examined by PCR. a, Primers
79
(arrows) were designed in all predicted prophages so that they are divergent in the integrated
80
genome orientation but convergent in the excised and circularized genome to amplify across the
81
attP. b, Prophage induction was tested in several strains by PCR amplification of filtered culture
82
supernatants (F.S.) treated (+) or not treated (-) with mitomycin C. Three to four replicates were
83
tested per strain. For each prophage of interest, a no template control (NTC) and several
84
unfiltered saturated cultures (S.C.) were included as negative and positive controls. The full
85
length of each lane from the loading well to leading edge is displayed. A star (*) indicates the
86
expected band size corresponding to attP amplification.
87 88
Supplementary Figure S2. Mitomycin C treatment increases dnaJ2-integrated
89
bifidoprophage copy number. DNA from mitomycin C-treated culture supernatants was
90
sequenced for several B. breve strains and reads were mapped to the lysogen genome (black
91
line). Enlarged view of the integrated prophage (white box) locus in each strain highlights the
92
increased sequencing coverage of the prophage relative to the host genome.
93 94
Supplementary Figure S3. Flow cytometry calibration and gating strategy. a, FACSCalibur
95
settings were calibrated using mitomycin C-treated L. lactis non-lysogen (strain UC509.9) and
96
lysogen (strain NZ9000 with TP901-1 prophage) samples. Scatterplots comparing forward
97
scatter (FSC-H) to (bottom) side scatter (SSC-H) and (top) Syto9 fluorescence (FL1-H) were
98
adjusted to reproduce previously described results41. b, Flow cytometry of several negative
99
controls, plotted as in panel a, to identify different types of events to gate. Samples include flow
100
sample buffer (¼ strength Ringer’s solution), flow sample buffer with reference microsphere
5
101
beads (¼ strength Ringer’s solution + Beads), and growth medium processed with the entire
102
protocol (RCM). c, Flow cytometry of common bifidobacterial growth media (MRS, MMRS +
103
Glucose, TOS), plotted as in panel b. d, Boxplots of individual parameters (FSC-H, SSC-H,
104
FL1-H) from flow cytometry results for several strain-free controls are used to define boundaries
105
of each parameter for debris (beige) and bead (blue) events. Some samples have been treated
106
(+) or not treated (-) with mitomycin C (MmC), Syto9 stain, and beads. e, Boundaries defined in
107
panel d were used to create three-dimensional debris and bead gates. The gating strategy for
108
all flow cytometry samples utilizes these two gates for removal of debris events followed by
109
removal of bead events. All non-debris and non-bead “gated” events are used for downstream
110
analysis to assess levels of prophage induction.
111 112
Supplementary Figure S4. Flow cytometric analysis of mitomycin C-treated
113
bifidobacterial samples. Gated events from one representative replicate of each sample type
114
are plotted. Scatterplots of FSC-H and SSC-H (left) and density plots of FL1-H (right) comparing
115
events either between mitomycin C-treated L. lactis non-lysogenic UC509.9 (blue) and
116
lysogenic NZ9000(TP901-1) (red) strain samples or between mitomycin C-treated (red) and
117
untreated (blue) bifidobacterial growth medium (RCM) and strain samples.
118 119
Supplementary Figure S5. Mitomycin C induced changes in supernatant composition. a,
120
For all gated events from each replicate set of paired mitomycin C treated (red) and untreated
121
(blue) samples (from Supplementary Fig. S4), (top) barplot of the proportion of total events and
122
(bottom) boxplots of event fluorescence (FL1-H) highlight changes in supernatant composition.
123
Replicate sets are numbered. Individual strain names are indicated along with whether they are
124
non-lysogens, lysogens, or predicted lysogens. RCM = growth medium with no cell culture. L.
125
lactis untreated = mitomycin C-treated non-lysogen (UC509.9); L. lactis treated = mitomycin C-
6
126
treated lysogen (NZ9000 with TP901-1 prophage). b, Boxplots display the fold changes in the
127
(top) abundance and (bottom) median fluorescence of events from paired samples in panel a.
128 129
Supplementary Figure S6. Bifidoprophages contain phase variation systems. a, Enlarged
130
view of the left arm genes of dnaJ2-integrated prophages from Fig. 1a highlights the genomic
131
context of the Rin shufflon. Genes are colored according to pham designation, and any putative
132
functions are listed (TMP = tape measure protein; DIT = distal tail protein; RBP = receptor
133
binding protein). The color spectrum between genomes is the same as in Fig. 1a. b, Enlarged
134
view of the left arm genes of tRNAMet-integrated phages from Fig. 1c highlights a putative phase
135
variation system in these genomes.
136 137
Supplementary Figure S7. Bb423phi1 induced phage genomes harbor multiple Rin
138
shufflon variants. a, Three contigs (numbered by size) representing the entire phage genome
139
are assembled by Newbler, but a 100% consensus of the complete genome is not achieved.
140
Contigs can be connected in multiple arrangements due to reads mapping across more than
141
one contig, and these discrepant reads occur near or within the RBP locus (colors and gene
142
numbering as in Fig. 5a). b, One possible contig orientation involves reads that straddle the
143
three contigs (dashed lines) with approximately equal coverage. c, Other contig orientations are
144
possible, but they are represented by much lower read coverage and they do not obviously
145
assemble into a single alternative genome. d, Two sequential inversion events (double arrows)
146
at rix sites result in three shufflon variants that sufficiently account for all hybrid sequence reads.
147 148
Supplementary Figure S8. Bb423phi1 uninduced prophage genomes harbor multiple Rin
149
shufflon variants. Analysis of the previously reported B. breve 139W4-23 genome sequencing
150
reads18 reveals three variant orientations of the Bb423phi1 prophage Rin shufflon. (Left) The
151
variant nucleotide sequence orientations were assembled and all reads in the sample were
7
152
mapped to each variant. The genome map below the histogram and the coordinates above the
153
histogram reflect the predominant variant in the published genome. The points of inversion in
154
each variant are indicated below (double arrows and coordinates). Average coverage across
155
each variant is indicated by the dotted line, and the percentage of all reads in the sample that
156
map to the variant orientation is indicated. (Right) Dotplot sequence comparison of each variant
157
to the published prophage locus orientation highlights the points of inversion.
158 159
Supplementary Figure S9. dnaJ2-integrated prophages exhibit high gene content flux.
160
Pairwise comparisons (black circles) of nucleotide sequence and gene content between dnaJ2-
161
integrated phages and all other actinobacteriophages as previously described53 to highlight
162
gene content flux patterns, with high (HGCF, blue) and low (LGCF, green) gene content flux
163
regions indicated.
164 165 166
8
a
b
Supplementary Figure S1
230
139W4-23
0 80
Bb423phi1
082W4-8
0
60
Bb48phi1
180W8-3
0 Bb83phi1
40
017W4-39
0 30
Bb439phi1
215W4-47a
0 Bb447phi1
30
0
139W4-23
Bb423 phi2
20kb Supplementary Figure S2
a UC509.9
NZ9000 (TP901-1)
FSC-H
FSC-H
d 104 103 102 101 1 104 103 102 101 1 104 103 102 101 1 Syto9 MmC Beads -
b Ringer’s
Ringer’s + Beads
RCM
Sample
-
-
H2O
+ -
+
+ +
Ringer’s
e FSC-H
FSC-H
All
FSC-H
debris gate
c MRS
MMRS+G
TOS
non-Debris
Debris
bead gate non-Beads (“gated”)
Beads
downstream analysis FSC-H
FSC-H
FSC-H
Supplementary Figure S3
L. lactis
LMG 10510
RCM
DSM 27321
JCM 7017
082W4-8
NCIMB 702258
180W8-3
UCC2003
139W4-23
FSC-H
FL1-H
017W4-39
215W4-47a
FSC-H
FL1-H
Supplementary Figure S4
a 0.25 0.20 0.15 0.10 0.05 0 104 103 102 101
treated untreated
1 1 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 4 1 2 34 1 2 3 4 1 2 3 1 2 3 4
b
Supplementary Figure S5
a
2 kb Binf-1
14
15
1135 1134
lysin
RBP DIT
1136
TMP
Bb48phi1
14
1
999 998
lysin
RBP DIT
10
TMP
Bb83phi1
14
999
1 1000
lysin holin
RBP DIT
TMP
Bb439phi1
14
1014
1 1015
lysin holin
RBP DIT
TMP
14
Bb423phi1
15
1126 1125
holin lysin
112
TMP
14
Bl30698phi1
1159
115 1158
holin lysin
TMP
11
Bl157phi1 Rin
TMP
2 kb 689b-1 int
Bb423phi2
106 1069
holin lysin
b
12
1070
lysin
tail protein
portal
capsid
lysin
tail protein
portal
capsid
int inversion locus
Supplementary Figure S6
a contig 1 (38,499 bp)
Rc
rin
1117 nterm.
1112
Rv2 contig 2 1116 (1,255 bp) contig 3 (326 bp)
Rv3 1115
Rv4
Rv5
1114
1113
Rv1 1117
b
160x
1112
207x
1117
156x
1116
1115
1114
1113
1117 nterm.
1114
1113
1117 nterm.
7x
c
6x
1112
1117
1116
1115
12x 12x
d Variant 1
1
1112
1113
1114
1115
1116
1117
1117 nterm.
Variant 2
2
1112
1117 nterm.
1112
1117 nterm.
Variant 3
3
Supplementary Figure S7
1112
1113
1114
1115
1116
1117
1117 nterm.
1112
1113
1114
1115
1116
1117
1117 nterm.
1112
1113
1114
1115
1116
1117
1117 nterm.
Supplementary Figure S8
1.0 0.8 HGCF 0.6 0.4 0.2 0.0
LGCF 0.0 0.1 0.2 0.3 0.4 0.5 nucleotide distance
Supplementary Figure S9