of 12
S1
Supporting Information:
DNAmediated signaling by proteins
with 4Fe4S clusters is necessary for
genomic integrity
Michael A. Grodick,
1
Helen M. Segal,
1
Theodore J. Zwang,
1
and
Jacqueline K. Barton
1,*
1
Division of Chemistry and Chemical Engineering,
California Institute of Technology, Pasadena CA 911
25
* To whom correspondence should be addressed, Emai
l: jkbarton@caltech.edu.
S2
SUPPLEMENTARY ANALYSIS
Atomic force microscopy assay.
An alternative way to calculate the binding density
ratio is to treat each individual image as a sample
, to plot them as a histogram (Figure S3) and to
find the mean of the normal distribution. Using thi
s methodology, instead of treating each
surface as an individual experiment, a binding dens
ity ratio of 1.61 ± 0.08 is found for DinG with
mismatched DNA. A binding densitry ratio of 1.40 ±
0.05 is found for DinG mixed with EndoIII
and mismatched DNA. Finally a binding density ratio
of 0.91 ± 0.04 is found for DinG mixed
with EndoIII Y82A.
Because the AFM images are snapshots of the system
near equilibrium, the number of
bound proteins on strands reflects the apparent non
specific binding affinity of the proteins for
DNA. As such, another way to analyze the data would
be to qualitatively analyze the apparent
binding affinity of each protein in the different e
xperiments by visualizing the occupancy of the
DNA i.e. the percent of DNA with a minimum of 1, 2,
3, or 4 proteins on a strand of DNA. Since
DNAbinding proteins with 4Fe4S clusters have a lo
wer DNA binding affinity when the cluster
in the reduced vs. oxidized (19), in a collection o
f proteins, some reduced and some oxidized, CT
between proteins should affect the overall effectiv
e binding affinity for DNA. Conversely, if
only one protein is bound to DNA, its affinity woul
d not be affected by DNA CT. This was
observed for mixtures of DNA and DinG (Figure S4),
which also indicates that DinG does not
preferentially bind a mismatch. The same was observ
ed for mixtures of DinG with WT EndoIII,
or with EndoIII Y82A; the percent of long, wellmat
ched strands with one protein bound is the
same for DinG/EndoIII and DinG/Y82A mixtures (Figur
e S5). If more than one protein is bound
to a given strand, however, each subsequent protein
that binds has a probability to transport
charge through the DNA and promote another protein’
s dissociation, in a way that parallels anti
S3
cooperative binding. If a mismatch is present, whic
h attenuates CT, however, this anti
cooperative effect would be lessened as it would if
EndoIII Y82A is substituted for EndoIII,
decreasing the concentration of signaling partners.
This is precisely what is observed (Figures
S4 and S5). The percentage of long strands with a m
inimum of 2, 3, or 4 bound DinG proteins is
lower for DinG when there is no mismatch, then when
there is a mismatch. When DinG is mixed
with EndoIII compared to EndoIII Y82A, an anticoop
erative effect is once again observed.
There is a decreased percentage of long strands of
DNA bound with a minimum of 2, 3, or 4
proteins for DinG and EndoIII compared to DinG and
EndoIII Y82A. This observation,
especially when considered in conjunction with thei
r similar population of singleboundprotein
DNA complexes, further supports the ability of DinG
and EndoIII to cooperate using DNA
mediated CT.
S4
Figure S1. SDSPAGE gel for purification of DinG. L
anes are referred to as 1 to 8 from left to
right. The lanes contain an SDSPAGE weight standa
rdlow range (Biorad) (Lane 1), the
supernatant from the cell lysate (lane 2), the pell
et from the cell lysate (lane 3), the filtered cell
lysate (lane 4), the the Histrap HP column flowthr
ough (lane 5), the collected fractions from the
Histrap HP column (lane 6), the collected fractions
from the Superdex 200 column (lane 7), and
a 10x dilution of the stored protein after thawing
(lane 8). Corresponding molecular weights for
each of the six bands in the weight standard lane a
re designated to the left of the image.
kDa
103.0
80.7
49.5
36.5
28.8
19.6
S5
Figure S2. Statistical data for the AFM experiments
.
Histograms showing the distribution of
binding density ratios within the population of sam
ple images. The value on the xaxis is the
upper range of a .25 unit wide separation, meaning
that the column plotted above “1” is the
number of images with a binding density ratio betwe
en 0.75 and 1.
S6
Figure S3. Effect of
nth
deletion in InvA (InvA F
nth
). Cultures of MM (M9 + 0.2% glucose) or
MM + ampicillin (100 Ig/mL) were inoculated with si
ngle colonies of the InvA parent strain
(blue), InvA F
nth
p(RNaseH) (green), InvA F
nth
p(EndoIII
D138A) (orange), InvA F
nth
p(WT
EndoIII) (black), InvA F
nth
p(EndoIII
Y82A) (brown), InvA F
nth
p(empty) (purple), or InvA
F
nth
(red). Growth was monitored over time. ± SEM
S7
Figure S4. Occupancy of DinG on long strands of DNA
. The percent of long strands of either
wellmatched (green) or mismatched (red) DNA is plo
tted against the minimum number of
proteins bound to a strand.
Figure S5. Occupancy of either EndoIII and DinG or
EndoIII Y82A and DinG on long strands of
wellmatched DNA. The percent of wellmatched long
strands of either EndoIII and DinG
(green) or EndoIII Y82A (red) DNA is plotted agains
t the minimum number of proteins bound to
a strand.
S8
Supplementary Table 1. Oligonucleotides for electro
chemistry substrates, gene
replacements, sequencing or colony PCR, and sitedi
rected mutagenesis
Note:
All oligonucleotides except for the modified strand
for electrochemistry were purchased
from IDT
Designation and use
Sequence (5’ to 3’)
(bases highlighted in red yield
the point mutation after SDM)
Primer for cloning
dinG
forward
GGTTTTCCCATGGCATTAACCGCC
Primer for cloning
dinG
reverse
CATCATTAAAGCTTCCGACGGCGT
pET28b
dinG
insert sequencing forward 1 T7 promoter primer
pET28b
dinG
insert sequencing forward 2
(within gene)
ACTGACGCCGAACAATCAGGA
pET28b
dinG
insert sequencing reverse 1 T7 terminator primer
pET28b
dinG
insert sequencing reverse 2
(within gene)
TTCGGCAAATGACTGTAAGCCCAC
Substrate for electrochemistry – 20mer
thiolated modified strand
HSC6GTGCTGCAACGTGTCTGCGC (annealed
with either the wellmatched complement or abasic
complement to yield the substrate used in
experiments)
Substrate for electrochemistry – 35 mer
complementary strand for wellmatched
substrate
AGACTGCAGACGAGAGCGCAGACACGTTGC
AGCAC
Substrate for electrochemistry – 35 mer
complementary strand for abasic substrate
AGACTGCAGACGAGAGCGCAGACACGTTGC
A_CAC (“_” represents an abasic site)
AFM substrates
3.8 kb long strands and 1.6 or 2.2 k
b shorts strands
were prepared as described previously (22, 35)
F
dinG
::cm
R
, forward primer
CCGAAAAATGCCACAATATTGGCTGTTTAT
ACAGTATTTCAGGTTTTCTCGTGTAGGCTGG
AGCTGCTTC
F
dinG
::cm
R
, reverse primer
CCGAAAAATGCCACAATATTGGCTGTTTAT
ACAGTATTTCAGGTTTTCTCGTGTAGGCTGG
AGCTGCTTC
F
dinG
::cm
R
, sequencing forward
GATGGTGTCTTGCATGACGTG
F
dinG
::cm
R
, sequencing reverse
TCAATACGCCGCCCAACTCA
SDM reverse primer for generation of
pBBR1MCS4
nth
Y82A
CGATTGGGCTTGCTAACAGCAAAGCAGAAA
ATATCATCAAAACCTGC
SDM forward primer for generation of
pBBR1MCS4
nth
D138A
CTATTCGTGTCGCCACGCACATTTTCCGCGT
TTGTAATC
SDM reverse primer for generation of
pBBR1MCS4
nth
D138A
CGGAAAATGTGCGTGGCGACAGCAATAGTC
GGCCAGC
S9
pBBR1MCS4
nth
sequencing forward
GGTGCTGATGCCGCTGGCGATTCAG
pBBR1MCS4
nth
sequencing reverse
TGTGCTGCAAGGCGATTAAGTTGG
Genomic
nth
region check forward 1
GAGATCCGCATTCCCATTTA
Genomic
nth
region check reverse 1
GGCTTAACGGCGATATGTTC
InvA check 1 from (33)*
CCAGTCATTTGGCGAAAG
InvA check 2 from (33)*
GGCGTAATAGCGAAGAGG
*
Used together as primers in PCR to amplify a ~4250 b
p product if
rrnA
is inverted
S10
Supplementary Table 2: Plasmids used in this study
Plasmid Designation
Description
Source, reference,
or method to
construct
pBBR1MCS4 or
p(empty)
pBBR1MCS4, a vector for the
constitutive expression of genes
placed in the MCS (multiple
cloning site)
(22)
pBBR1MCS4
nth
or
p(WT EndoIII)
pBBR1MCS4 carrying the
nth
gene in the MCS, constitutively
expresses WT EndoIII
(22)
pBBR1MCS4
nth
D138A or p(EndoIII
D138A)
pBBR1MCS4 carrying the
nth
D138A gene in the MCS,
constitutively expresses EndoIII
D138A
Sitedirected mutagenesis of
pBBR1MCS4
nth
using SDM
primers listed in Table S3.
pBBR1MCS4
nth
Y82A or p(EndoIII
Y82A)
pBBR322MCS4 carrying the
nth
Y82A gene in the MCS,
constitutively expresses EndoIII
Y82A
Sitedirected mutagenesis of
pBBR1MCS4
nth
using Y82A
forward and reverse primers listed in
Table S2
pET28b
dinG
Overexpresses DinG in
presence of IPTG
Insertion of
dinG
amplicon into pET
28b(+) (Novagen) as described
above (30)
pEMAp
R
or
p(RNaseH)
pACYC184 derived vector
carrying the
rnh
+
gene and an
Ap
R
gene, overexpresses
RNaseH
Gift from Dr. Bénédicte Michel (Le
centre de la Recherche Scientifique,
GifsurYvette, France) (33)
S11
S12
Supplementary Table 3. Strains used in this study
Strain
Genotype designation
Source, reference, or
method to
construct
Lac
+
forward reversion assay, a GC:TA transversion assa
y to probe MutY activity
CC104
F128(CSH104)
lacI373
,
lacZ574
,
ara
600
,
(gptlac)5
,
λ

,
relA1
,
spoT1
,
thiE1
(22)
CC104 F
dinG
CC104 F
dinG
::cm
R
Inactivation of
dinG
by replacement
with cm
R
using pKD3 and pKD46
(CGSC & 7)
InvA >
nth
Rloop assay to probe DinG activity
JW16251 F,
(araDaraB)567
,
lacZ4787
(::rrnB3),
λ

,
nth782::kan
,
rph1
,
(rhaDrhaB)568
,
hsdR514
Obtained from Coli Genetic Stock
Center (Yale University, New
Haven, CT)
BW16847 F,
(codBlacI)3
,
phoA532
,
pdxH15
(Am),
purR106::Tn10
,
(phnP
phnD)3330(phnC?)
Obtained from Coli Genetic Stock
Center (Yale University, New
Haven, CT)
InvA
MG1655 FlacZ FattB::spcR Inv
(attL15cm
R
attR75kan
R
)
Gift from Dr. Bénédicte Michel at
the Centre de Génétique
Moléculaire (CNRS) (33)
MG001
BW16847 F
nth
::kan
R
BW16847 * P1 JW16251
InvA F
nth
InvA F
nth
::kan
R
,
pdxH15
(Am),
purR106::Tn10
MG001 * P1 InvA and colony PCR
to find colonies with the F
nth
::
kan
R
locus
InvA F
purR
InvA
purR106::Tn10
MG001 * P1 InvA and colony PCR
to find colonies without the F
nth
::
kan
R
locus