Mutations that alter the regulation of the chb operon of Escherichia coli allow utilization of cellobiose.

Abstract

Wild-type strains of Escherichia coli are normally unable to metabolize cellobiose. However, cellobiose-positive (Cel(+)) mutants arise upon prolonged incubation on media containing cellobiose as the sole carbon source. We show that the Cel(+) derivatives carry two classes of mutations that act concertedly to alter the regulation of the chb operon involved in the utilization of N,N'-diacetylchitobiose. These consist of mutations that abrogate negative regulation by the repressor NagC as well as single base-pair changes in the transcriptional regulator chbR that translate into single-amino-acid substitutions. Introduction of chbR from two Cel(+) mutants resulted in activation of transcription from the chb promoter at a higher level in the presence of cellobiose, in reporter strains carrying disruptions of the chromosomal chbR and nagC. These transformants also showed a Cel(+) phenotype on MacConkey cellobiose medium, suggesting that the wild-type permease and phospho-beta-glucosidase, upon induction, could recognize, transport and cleave cellobiose respectively. This was confirmed by expressing the wild-type genes encoding the permease and phospho-beta-glucosidase under a heterologous promoter. Biochemical characterization of one of the chbR mutants, chbRN238S, showed that the mutant regulator makes stronger contact with the target DNA sequence within the chb promoter and has enhanced recognition of cellobiose 6-phosphate as an inducer compared with the wild-type regulator.

Full text

Page 1

Mutations that alter the regulation of the chb operon of
Escherichia coli allow utilization of cellobiose
Aashiq H. Kachroo,1Aswani K. Kancherla,1,2
Nongmaithem S. Singh,3Umesh Varshney3and
Subramony Mahadevan1*
1Department of Molecular Reproduction, Development
and Genetics,2Molecular Biophysics Unit, and
3Department of Microbiology and Cell Biology, Indian
Institute of Science, Bangalore 560 012, India.
Summary
Wild-type strains of Escherichia coli are normally
unable to metabolize cellobiose. However, cellobiose-
positive (Cel+) mutants arise upon prolonged incuba-
tiononmediacontainingcellobioseasthesolecarbon
source. We show that the Cel+derivatives carry two
classes of mutations that act concertedly to alter the
regulation of the chb operon involved in the utilization
ofN,N?-diacetylchitobiose.Theseconsistofmutations
that abrogate negative regulation by the repressor
NagC as well as single base-pair changes in the tran-
scriptional regulator chbR that translate into single-
amino-acid substitutions. Introduction of chbR from
twoCel+mutantsresultedinactivationoftranscription
fromthechbpromoteratahigherlevelinthepresence
of cellobiose, in reporter strains carrying disruptions
of the chromosomal chbR and nagC. These transfor-
mants also showed a Cel+phenotype on MacConkey
cellobiose medium, suggesting that the wild-type per-
mease and phospho-b-glucosidase, upon induction,
could recognize, transport and cleave cellobiose
respectively. This was confirmed by expressing the
wild-typegenesencodingthepermeaseandphospho-
b-glucosidase under a heterologous promoter. Bio-
chemical characterization of one of the chbR mutants,
chbRN238S, showed that the mutant regulator makes
stronger contact with the target DNA sequence within
the chb promoter and has enhanced recognition of
cellobiose 6-phosphate as an inducer compared with
the wild-type regulator.
Introduction
Escherichia coli can respond to stress such as starvation
using a variety of strategies. Under conditions of starva-
tion wherein a novel substrate is provided as a sole nutri-
tional source, spontaneous mutants arise that are able to
utilize this novel compound. Many genetic systems, upon
mutational activation, have been shown to allow E. coli to
grow on novel substrates (Wright, 2004).
Most wild-type strains of E. coli are unable to utilize the
cellulose-derived disaccharide cellobiose as a carbon
source. Therefore, they cannot grow on minimal cello-
biose medium and form white colonies on MacConkey
cellobiose indicator medium. However, cellobiose-positive
(Cel+) mutants can be obtained after prolonged incubation
(~20–30 days) at room temperature (25–30°C) as papillae
on MacConkey cellobiose plates and as colonies on
minimal cellobiose plates (Kricker and Hall, 1984). The
Cel+phenotype was attributed to mutations in the genetic
system originally named the cel operon that was believed
to be cryptic (Parker and Hall, 1990a). The cel operon,
subsequently renamed the chb operon, is the inducible
genetic system involved in the catabolism of N,N′-
diacetylchitobiose (Keyhani and Roseman, 1997). The
chb operon comprises six open reading frames (ORFs)
(chbBCARFG) with a ~200 bp regulatory region (chbOP).
The chbBCA genes encode the IIB, IIC and IIAdomains of
the PTS-dependent permease, respectively (Keyhani
et al., 2000b), chbR encodes a dual function activator/
repressor (Plumbridgeand
encodes a phopho-b-glucosidase assigned to family 4 of
the glycosylhydrolase superfamily with wide substrate
specificity (Thompson et al., 1999) and chbG codes for a
protein of unknown function. The regulation of the chb
operon by chitobiose has recently been studied (Plum-
bridge and Pellegrini, 2004). It has been shown that the
three proteins ChbR, CRP and NagC regulate the expres-
sion of the chb operon. ChbR belongs to the AraC-like
dual function activator/repressor family of proteins. In the
absence of the substrate N,N′-diacetylchitobiose, ChbR
binds to the direct repeats present within chbOP and
represses transcription along with the negative regulator
NagC. In the presence of chitobiose, ChbR, along with
CRP, activates transcription from the chb promoter (Plum-
bridge and Pellegrini, 2004).
Activation of the chb operon allowing the utilization of
cellobiose was reported to occur either via insertion of
IS1, IS2 or IS5 within 72–180 bp upstream of an earlier
uncharacterized start site or by base substitutions in chbR
such that the putative repressor is able to recognize
Pellegrini,2004),chbF
Accepted 9 October, 2007. *For correspondence. E-mail mahi@
mrdg.iisc.ernet.in; Tel. (+91) 80 2293 2607; Fax (+91) 80 2360 0999.
Molecular Microbiology (2007) 66(6), 1382–1395 doi:10.1111/j.1365-2958.2007.05999.x
First published online 19 November 2007
© 2007 The Authors
Journal compilation © 2007 Blackwell Publishing Ltd

Page 2

cellobiose, salicin and arbutin as inducers (Parker and
Hall, 1990b). The need of prolonged incubation in order to
obtain these Cel+mutants and the complete absence of
constitutive Cel+mutants were, however, unexplained.
The timescale of ~30 days to acquire Cel+phenotype by
E. coli either by activating insertion elements or by muta-
tions within the chbR locus is long compared with activa-
tion by insertion mutations targeting the regulatory region
(bglR) of the cryptic bgl operon of E. coli, conferring a
salicin-positive (Sal+) phenotype within 24–48 h of incu-
bation on MacConkey salicin plates. This was partly
explained when the sequence of the chb operon reported
earlier (Parker and Hall, 1990a) was compared with the
wild-type E. coli K-12 genome (Blattner et al., 1997). The
comparison revealed many differences. These differ-
ences are clustered in the permease and in the putative
phospho-b-glucosidase. Based on this observation it was
hypothesized that additional mutations may be involved
in the activation of the chb operon allowing utilization of
cellobiose (Keyhani and Roseman, 1997).
The results presented in this communication show that
acquisition of a Cel+phenotype involves a minimum of two
types of mutations that primarily alter the regulation of the
chb genes. These two classes of mutations act in a con-
certed fashion and are necessary and sufficient to confer
a Cel+phenotype; mutations in the structural genes, pre-
dicted by earlier work, are not necessary. The biochemical
characterization of one of the chbR mutants, chbRN238S,
indicates that the enhanced activation seen in the mutant
is due to the combined action of enhanced binding to the
target sequences as well as improved recognition of cel-
lobiose as an inducer.
Results
An insertion within chbOP or a deletion of chbR do not
confer a Cel+phenotype
To examine whether the chb operon can be activated to
confer Cel+phenotype by a single mutational event of
insertion within the chbOP region as reported earlier
(Parker and Hall, 1990b), an artificial insertion of a ~1.1 kb
DNA fragment encoding chloramphenicol resistance was
made in the E. coli strain DY330 carrying a wild-type copy
of the chb genes. The target for the artificial insertion was
the same as that of an activating insertion element seen in
a subset of Cel+mutants (Fig. 1). The strain showed a
Cel–phenotype on MacConkey cellobiose medium (after
24–36 h of incubation at 37°C). Similar results were
obtained when the artificial insertion was moved from the
strain DY330 (chbOP::cat) into JF201 by P1 transduction
to rule out strain-specific differences. Interestingly, strains
carrying the artificial insertion within chbOP papillated
faster compared with the respective wild-type strains, sug-
gesting a positive role for the insertion in acquiring a Cel+
phenotype.
The possible effect of a null mutation in the putative
repressor gene chbR in conferring a Cel+phenotype
was also examined. A deletion of the chbR locus was
constructed as before using a targeted recombination
approach in the strain DY330 and subsequently moved to
JF201. The chbR deletion strains, when plated on MacCo-
nkey cellobiose medium, were Cel–and did not papillate to
yield Cel+mutants even after prolonged incubation of
30–40 days, suggesting an additional regulatory role for
chbR, consistent with its role in the activation of the operon
in the presence of N,N′-diacetylchitobiose (Plumbridge
and Pellegrini, 2004). Therefore, single mutations within
the chb operon are incapable of conferring a Cel+pheno-
type, contrary to earlier reports (Parker and Hall, 1990b).
Isolation and characterization of Cel+mutants from
different E. coli strains
In an attempt to understand the mechanism of activation
of the chb operon, large-scale isolation of ~120 Cel+
mutants from different strains of E. coli was carried out
(Table 1). Mutants were obtained as papillae on MacCo-
nkey cellobiose plates [MCP] or colonies on M9 minimal
cellobiose plates [M9]. Typically mutants appeared as
papillae in ~24 days on MCP and as colonies on M9
plates in ~14 days. The extent of cellobiose utilization by
different Cel+mutants was tested by plating on MacCon-
key cellobiose plates and spotting dilutions of different
Cel+mutants on M9 minimal cellobiose medium. The Cel+
phenotypes of 42 mutants tested are indicated in Table 3.
Variation in the cellobiose utilization phenotype of different
Cel+mutants suggested the presence of different types
of activating mutations. Contrary to earlier reports (Kricker
and Hall, 1987), none of the cellobiose-utilizing mutants
tested could utilize salicin and arbutin.
Fig. 1. The regulatory region (chbOP) of the chb operon showing
binding sites for NagC, ChbR and CAP proteins (based on
Plumbridge and Pellegrini, 2004). A newly characterized
transcription start site is also shown. Black arrows within the strong
NagCII binding site indicate insertion elements reported in Cel+
mutant strains. The region chosen for the artificial insertion
(chloramphenicol resistance marker) is indicated. The dashed
underlines indicate the -10 and -35 promoter elements and the
core CRP-cAMP binding site.
Activation of the chb operon of E. coli enabling utilization of cellobiose 1383
© 2007 The Authors
Journal compilation © 2007 Blackwell Publishing Ltd, Molecular Microbiology, 66, 1382–1395

Page 3

Mutations linked to the chb locus are involved in the
Cel+phenotype
The cellobiose utilization phenotype of the Cel+mutants,
at least in part, was shown to be linked to the chb operon
using P1 transduction. P1 lysate prepared using the Cel–
donor strain JF201 (chbR::cat), in which the chloram-
phenicol resistance gene is 100% linked to the chb
operon, was used to transduce the Cel+mutants. All
chloramphenicol-resistant transductants showed loss of
the Cel+phenotype, confirming linkage of at least one of
the associated mutations to the chb operon.
Many Cel+strains carry mutations in nagC
The NagC repressor that regulates the nag operon
involved in N-acetylglucosamine metabolism was also
shown to regulate the chb operon (Plumbridge and Pel-
legrini, 2004). The insertions seen in many Cel+mutants
disrupt the strong NagC binding site within the chb
regulatory region. However, only a small number of Cel+
mutants obtained carried insertion elements within the
regulatory region (chbOP) that disrupted the strong NagC
binding site. If repression by NagC has to be eliminated
for acquiring a Cel+phenotype, one possibility is that
these strains have mutations in the nagC locus itself. To
test this possibility, the wild-type nagC locus was trans-
duced into the Cel+mutants using the donor strain JF496,
which carries a Tn5 insertion within the asn locus that has
85% linkage to wild-type nagC. The transductants showed
a Cel–phenotype at a frequency of ~85%, indicating that
the Cel+strains had lesions within nagC that are associ-
ated with the Cel+phenotype, which is lost upon introduc-
tion of the wild-type nagC locus.
The MG1655 strain used for isolation of Cel+mutants
was inherently kanamycin resistant, pre-empting trans-
duction experiments using the kan marker. Alternatively
the wild-type nagC locus was introduced on a plasmid by
transformation. All ampicillin-resistant transformants were
Cel–, indicating that the chromosomal nagC carries re-
cessive mutations that are partly responsible for the Cel+
phenotype. These results confirm that mutations within
the chb operon and loss-of-function mutations within the
nagC locus play a critical role in conferring a Cel+
phenotype.
Characterization of the mutations linked to the chb
operon in Cel+mutants
The inability of chbR deletion strains to yield Cel+mutants
even after prolonged incubation (~30–40 days) suggested
an essential role for ChbR in conferring a Cel+phenotype.
To examine if the Cel+mutants harbour mutations within
chbR, the chbR locus (~900 bp) from different Cel+
mutants and the Cel–parent strains was amplified and
sequenced. The nucleotide sequence of the chbR locus
from the Cel–parent strains showed a 100% match with
Table 1. Bacterial strains used in this study.
StrainGenotype Source
JF201
DH5a
F-DlacX74 D(bgl-pho) 201 ara thi gyrA
F?/endA1 hsdR17 (rk–mk+) supE44 thi-1 recA1gyrA
(Nalr)relA1(DlacZYA argF)U169 deo FdlacD(lacZ)M15
(F-wt bgl° lam–rph-1) Kanr
JM101 [Dlac-pro, thi-1, supE, F′(traD36, proAB+, lacIq, lacZDM15)]
(chbR::cat) pchbOPB?–lacZ
JM101 [Dlac-pro, thi-1, supE, F′(traD36, proAB+, lacIq, lacZDM15)]
(chbR::cat; nagC::tet) pchbOPB′ lacZ
JM101 [Dlac-pro, thi-1, supE, F′(traD36, proAB+, lacIq, lacZDM15)]
(nagC::tet) pchbOPB′ lacZ
nagB2 asn50::Tn5
W3110 DlacU169 gal490 lcI857 D(cro-bioA)
DY330 (chbOP::cat)
DY330 (chbOP::cat) Cel+
DY330 (chbR::cat)
DY330 (chbR::kan)
JF201 (chbOP::cat)
JF201 (chbR::cat)
Cel+mutants of MG1655
Cel+mutants of MG1655
Cel+mutants of JF201
Cel+mutants of JF201
Cel+natural isolate
Cel+natural isolate
Cel+natural isolate
Cel+natural isolate
T7 polymerase-inducible expression strain
Reynolds et al. (1986)
Woodcock et al. (1989)
MG1655
JM-chb21
E. coli Genetic Stock Center
Plumbridge and Pellegrini (2004)
JM-chb22
Plumbridge and Pellegrini (2004)
JM-chb3Plumbridge and Pellegrini (2004)
JF496
DY330
DOPCam
AHK5
DRCam
DRKan
JOPCam
JRCam
MG-MCP-01-32
MG-M9-01-30
J-MCP-01-28
J-M9-01-25
AHK3 PI
MS201
NC 2.1
NC 7.1
Rosetta (DE3)
E. coli Genetic Stock Center
Yu et al. (2000)
This work
This work
This work
This work
This work
This work
This work
This work
This work
This work
This work
This work
This work
This work
Novagen
1384A. H. Kachroo et al.
© 2007 The Authors
Journal compilation © 2007 Blackwell Publishing Ltd, Molecular Microbiology, 66, 1382–1395

Page 4

the E. coli K12 sequence available in the NCBI database.
The sequences of the chbR gene obtained from 42 Cel+
mutants analysed showed a single-base mutation within
the coding region of chbR, resulting in a single-amino-acid
substitution at the protein level (Table 3). As indicated in
Fig. 2, the mutations are scattered across the different
domains of the ChbR protein. The role of these single-
amino-acid substitutions in conferring a cellobiose utiliza-
tion phenotype was analysed both at the phenotypic
level and with respect to their effect on transcription from
the chb promoter.
The effect of point mutations within the chbR locus on
chb promoter activity
To investigate the role of mutations within the chbR and
nagC loci in acquiring a Cel+phenotype, single-copy chro-
mosomal reporter strains that contain a chbOPB?–lacZ
transcriptional/translational fusion at the attB site (Plum-
bridge and Pellegrini, 2004) were used. Primarily three
strains were used in these studies. One of them,
JM-chb22, had disruptions of the chromosomal chbR and
nagC loci and the other, JM-chb21, had a chbR disruption,
but an intact nagC locus. The third strain, JM-chb3, had a
disruption of nagC, but an intact chbR locus. Two chbR
mutants that showed a strong Cel+phenotype were
chosen for the studies. The mutant chbR alleles were
introduced on plasmids and their effect on transcription in
trans was observed. The plasmid pChbRN238S carries
the chbRN238S allele which is a predominant mutation
seen in several Cel+mutants (Table 2). Another chbR
clone, pChbRY30C, carries the chbRY30C allele ampli-
fied from the Cel+natural isolate AHK3PI. The wild-type
and mutant chbR clones were introduced into different
reporter strains by transformation and b-galactosidase
assays were performed in the presence or absence of
10 mM cellobiose as inducer.
In the strain JM-chb22 (nagC::tet; chbR::cat), presence
of both chbR mutants resulted in a high basal level of
expression. More importantly, the transformant carrying
ChbRN238S showed an approximately threefold induc-
tion over the basal level in the presence of 10 mM cello-
biose whereas no induction was seen in the presence of
wild-type ChbR (Fig. 3). The overall activated level in the
Fig. 2. Different amino acid substitutions in the ChbR protein from various Cel+mutants due to point mutations at the chbR locus are shown.
The numbers in parenthesis are the number of strains showing the same type of amino acid substitution. Conserved domain database
analysis of ChbR showed that the N-terminal has similarity to the mannose 6-phosphate isomerase domain associated with carbohydrate
transport and metabolism. The C-terminal half of the protein showed partial similarity to the adenosine deaminase domain. The C-terminal end
of ChbR also carries an AraC-like helix–turn–helix (HTH) motif.
Table 2. List of plasmids used in this study.
Plasmid Genotype and description Source
pBR322
pChbR WT
pChbRN238S
pChbRY30C
pChbRL136S
pRA197T
pRNC2.1
pNWT
pACDH
placB-F
pOP-A
pChbF
pChbNC2.1
pDRIVE
pJES307
pJES (ChbR)
pJES (N238S)
Tetr, Ampr
chbR+in pBR322
chbR N238S in pBR322
chbR Y30C in pBR322
chbR L136S in pBR322
chbR A197T in pBR322
chbR A197T;Q172H in pBR322
nagC+in pBR322
plac followed by MCS, pACYC origin of replication
plac-chbB+C+A+chbR::kan chbF+in pACDH
chbOP+B+C+A+in pBR322
chbF+in pBR322
chbOPBCARF (amplified from NC2.1 Cel+) in pBR322
pUC origin TA cloning vector (Qiagen) plac–lacZ?
T7 polymerase expression vector for expressing specific native proteins
pJES301 chbR+(NdeI–BamHI)
pJES301 chbR N238S (NdeI–BamHI)
Boliver et al. (1977)
This work
This work
This work
This work
This work
This work
This work
Singh et al. (2005)
This work
This work
This work
This work
Qiagen
Tabor and Richardson (1985)
Plumbridge and Pellegrini (2004)
This work
Activation of the chb operon of E. coli enabling utilization of cellobiose1385
© 2007 The Authors
Journal compilation © 2007 Blackwell Publishing Ltd, Molecular Microbiology, 66, 1382–1395

Page 5

presence of the mutant was approximately sevenfold
higher as compared with that in the presence of the wild-
type ChbR. The presence of ChbRY30C resulted in an
approximately twofold induction by cellobiose over the
basal level. This induction by cellobiose was specific as
no induction was seen in the presence of similar concen-
trations of other substrates such as N-acetylglucosamine
or salicin. A modest repression was seen in the presence
of glucose (data not shown).
All Cel+strains harbouring mutations at the chbR locus
also contained loss-of-function mutations at the nagC
locus. This suggested that the phenotypic effect of the
mutations within chbR can be elicited only in the absence
of negative regulation by NagC. The b-galactosidase
assays performed in JM-chb21 strain containing the wild-
type copy of nagC indicated that neither of the chbR
mutants could activate transcription in the presence of
cellobiose (Fig. 3). The repression by NagC was epistatic
over the transcriptional activation by mutant ChbR in the
presence of cellobiose. Transformants carrying the wild-
type and mutant chbR clones in fact showed an approxi-
mately twofold reduction in promoter activity in the
presence of the wild-type nagC (Fig. 3). This reduction
could be attributed to the role of ChbR as a repressor in
the presence of NagC.
In order to see if the mutant ChbR proteins could acti-
vate transcription in the presence of a genomic copy of
wild-type chbR, b-galactosidase assays were carried out
in the strain JM-chb3 (nagC::tet; chbR+). ChbRN238S
could not activate transcription in the presence of 10 mM
cellobiose (Fig. 3), indicating that the wild-type chbR is
dominant over chbR N238S.
The role of mutations within the chbR and nagC loci in
conferring a Cel+phenotype
In an attempt to observe the effect of chbR mutations on
the cellobiose utilization phenotype, transformants of
JM-chb22 (nagC::tet; chbR::cat) carrying the wild-type
and mutant chbR clones described above were streaked
on MacConkey cellobiose plates. Transformants carrying
the mutant chbR clones conferred a Cel+phenotype to the
strain whereas those carrying the wild-type chbR clones
remained Cel–. Interestingly the ability of cellobiose
utilization could be correlated with the transcriptional
activation seen in the presence of cellobiose. The higher
the induction in the presence of cellobiose, as seen
in reporter assays, the faster was the Cel+phenotype
observed. The mutant ChbRN238S conferred a Cel+phe-
notype within 12 h of incubation at 37°C as compared with
the mutant ChbRY30C that conferred a Cel+phenotype
after 24 h of incubation at 37°C. However, mutant chbR
clones after transformation into a strain JM-chb21 carry-
ing the wild-type nagC locus could not confer cellobiose
utilization phenotype (Fig. S1).
These results suggested that the wild-type permease
(ChbBCA) and phospho-beta-glucosidase (ChbF), upon
induction could recognize, transport and cleave cello-
biose respectively. This was confirmed by cloning the
wild-type genes encoding the permease and phos-
pho-b-glucosidase under a heterologous promoter Plac.
Transformants of the Cel–strain DH5a carrying the
low-copy plasmid expressing the permease and the
phospho-b-glucosidase could utilize cellobiose efficiently
(Fig. S1). Transformants expressing either the chb per-
Fig. 3. b-Galactosidase assays using reporter
strains carrying the wild-type and mutant
chbR clones, in M9 medium containing 0.4%
glycerol and 0.4% casamino acids, with or
without 10 mM cellobiose. The reporter strains
JM-chb 21 (chbR::cat), JM-chb22 (nagC::tet;
chbR::cat) and JM-chb3 (nag::tet) carrying a
single-copy chbOPB?–lacZ
transcriptional/translational fusion, were used.
The results are the mean of three to six
independent measurements.
1386 A. H. Kachroo et al.
© 2007 The Authors
Journal compilation © 2007 Blackwell Publishing Ltd, Molecular Microbiology, 66, 1382–1395

Page 6

mease (chbBCA) or the phospho-b-glucosidase (chbF)
alone could not confer Cel+phenotype. These results
are consistent with the report that ChbF has a wide sub-
strate specificity that includes phospho-cellobiose and
phospho-salicin (Thompson et al., 1999). The results are
also consistent with the observation that cellobiose can
inhibit chitobiose transport by chbBCA (Keyhani et al.,
2000a).
The effect of mutations in chbR on the recognition of
chitobiose as an inducer
To investigate whether Cel+mutant strains retained
the ability to respond to chitobiose as an inducer,
b-galactosidase assays were carried out in the reporter
strain (JM-chb22) transformed with the wild-type and
the mutant chbR clones described above. These assays
showed that both mutants could activate transcription
approximately twofold in the presence of 2.4 mM N,N′-
diacetylchitobiose over and above the high basal level
(Fig. 4), indicating that they retain the ability to recognize
chitobiose. When the assays were performed in a
reporter strain JM-chb3 carrying the wild-type chbR at
the chromosomal locus and a disruption of nagC, the
activity in the presence of chitobiose showed an induc-
tion of approximately seven- to eightfold (Fig. 4), indi-
cating that the wild-type chbR at the genomic locus acti-
vates transcription better compared with chbR present
on the plasmid. This is likely to be related to the low
level of transcription from the endogenous plasmid pro-
moter and the lack of positive autoregulation of chbR
expression in the clones.
Variation in the amino acid sequence of ChbR across
different species and genera and the role of the chb
operon in Cel+natural isolates of E. coli
To examine whether Cel+natural isolates of E. coli also
carry mutations within the chb locus similar to those seen
in E. coli K12, several Cel+strains were obtained from
different sources and were further characterized. In addi-
tion to cellobiose utilization, some of the natural isolates
could also utilize salicin and arbutin although they carried
deletions of the bgl genes.
Sequence comparison at the chbR locus in all the Cel+
natural isolates of E. coli showed changes at the DNA
level resulting in the amino acid substitutions indicated in
Table 3. The single-amino-acid changeA197T seen in two
of the natural isolates was also seen in the pathogenic
strain O157::H7 (Sakai) and some strains of Shigella
sonnei. Another Cel+natural isolate NC2.1 showed an
additional change (Q172H) (Table 3). The chbR locus
from S. sonnei, E. coli O157:H7 (Sakai) and the Cel+
natural isolates (NC2.1, NC7.1) was amplified by PCR
and cloned in pBR322. The chbR clones, when trans-
formed into JM-chb22, could not confer a Cel+phenotype.
However, the chbR clones obtained from the Cel+natural
isolate MS201 (L136S) and AHK3 PI (Y30C) could confer
a Cel+phenotype to JM-chb22.
The entire chb operon, along with the chb promoter,
was amplified by PCR from the Cel+natural isolate NC2.1
and cloned in pBR322. The clone, when introduced into
the strain JM-chb22 carrying disruptions of the chromo-
somal nagC and chbR loci, could not confer a Cel+phe-
notype, suggesting the presence of additional genetic
Fig. 4. b-Galactosidase assays performed
using reporter strains carrying the wild-type
and mutant chbR clones, grown in M9
medium containing 0.4% glycerol and 0.4%
casamino acids, with or without 0.1%
N,N′-diacetylchitobiose (2.4 mM). The reporter
strains JM-chb22 (nagC::tet; chbR::cat) and
JM-chb3 (nagC::tet) carrying a single-copy
chbOPB?–lacZ transcriptional/translational
fusion were used. The results are the mean of
four independent measurements.
Activation of the chb operon of E. coli enabling utilization of cellobiose1387
© 2007 The Authors
Journal compilation © 2007 Blackwell Publishing Ltd, Molecular Microbiology, 66, 1382–1395

Page 7

systems for cellobiose utilization in this strain. Thus, many
Cel+natural isolates of E. coli may have activated genetic
systems other than the chb system such as the asc
operon (Parker and Hall, 1988) for cellobiose utilization.
Biochemical studies using purified ChbR
and ChbRN238S
In an attempt to investigate the mechanism of transcrip-
tional activation mediated by mutations in chbR, detailed
biochemical studies using purified ChbR and ChbRN238S
were carried out. Transformants carrying the chbR and
chbRN238S recombinant plasmids were induced with
IPTG. As in the earlier study (Plumbridge and Pellegrini,
2004) the wild-type protein could be purified by the use
of the three columns, heparin sepharose, mono-S and
hydroxyl-apatite (see Experimental procedures). The
same protocol, however, did not allow purification of
ChbRN238S as the mutant failed to bind heparin
sepharose, suggesting possible differences in the bio-
chemical properties of the mutant. ChbRN238S was
purified using a mono-S column (Fig. S2). The estimated
molecular mass of both the wild-type and the mutant
ChbRN238S was ~32 KDa on 12% SDS-PAGE, as
reported earlier (Plumbridge and Pellegrini, 2004).
Differential binding of ChbR and ChbRN238S to chbOP
TheintrinsicpropertyoftheChbRtobindtothespecificsite
within the chb regulatory region, chbOP, was exploited to
find possible differences between wild-type ChbR and
ChbRN238S, using the electrophoretic mobility shift assay
(EMSA). The 188 bp chbOP regulatory region used as the
target sequence in the assays is shown in (Fig. 5A).
The EMSAs were carried out using two concentrations
(5 nM and 10 nM) of the wild-type and the ChbRN238S
proteins (Fig. 5B). Wild-type ChbR formed two types of
DNA–protein complexes with the specific target DNA: one
that remained in the wells and could not be competed out
by excess of cold specific DNA (unlabelled chbOP) and
one that entered the gel and could be competed out by
excess of cold specific DNA (Fig. 5B, lanes 1–4 and 5–8).
On the other hand, ChbRN238S formed only a single
DNA–protein complex that could be competed out by
excess cold specific DNA (Fig. 5B, lanes 9–12 and
13–16). ChbRN238S was more efficient in forming the
discrete DNA–protein complex compared with ChbR, as
the complex can be detected even at the lower protein
concentration (Fig. 5B, lanes 1–4 versus lanes 9–12). In
addition, a lower concentration of cold specific DNA could
compete away the complex in the case of ChbR, indicat-
ing that the affinity of ChbRN238S for the target DNA is
higher. Both ChbR and ChbRN238S were unable to bind
a 120 bp DNAfragment generated by NruI digestion of the
188 bp template and a 60 bp synthetic oligonucleotide
dimer, carrying the direct repeats known to be the binding
site for ChbR (data not shown).
DNase I footprinting assays with wild-type ChbR
and ChbRN238S
To further ascertain the difference between ChbR and
ChbRN238S in their abilities to bind the specific site within
the chbOP regulatory region, DNase I footprinting studies
were carried out. These assays indicated that the ability of
ChbRN238S to bind a direct repeat within chbOP was
better than the wild-type counterpart (Fig. 6, compare
lanes 1 and 2; 3 and 4). Furthermore, ChbRN238S pro-
Table 3. Single-base-pair mutations in chbR resulting in single-amino-acid substitutions from different Cel+mutants and Cel+natural isolates.
Cel+strain (*)
GenBank
accession No.
Amino acid
change in ChbR
Presence of insertion
within chbOP
Cel+phenotypeaon MacConkey
cellobiose plates (after 24–48 h)
MG-MCP-02
MG-MCP-07
MG-MCP-12
MG-M9-19, 21, 28 (3)
J-MCP-01; J-M9-02-09, 11–14, 17,
18, 20, 21, 24 and 27; AHK5 (20)
J-MCP-03 and J-M9-16 (2)
J-MCP-17
J-M9-01
J-M9-10
J-M9-19 and 22 (2)
J-M9-23
MS201
AHK3 PI
NC2.1
NC7.1
MG-M9-25
MG-M9-26, 29 and MG-MCP-09 (3)
EF470556
EF470554
EF470558
EF470561
EF470559
K184T
I227L
N137K
S135A
N238S
–
–
–
–
–
+
++
+
+?
++
EF470551
EF470555
EF470558
EF470552
EF470550
EF470553
EF470557
EF470562
EF470560
EF470549
EF577379
EF577380
F138S
K184E
N137K
F138V
F138L
F146L
L136S
Y30C
Q172H; A197T
A197T
E164K
L263I
–
–
–
–
–
–
–
–
–
–
+
+
++
++
++
+
++
++
++
++
++
++
+
++
a. The phenotypes are indicated as (+) light pink, (+?) pink, and (++) red.
*The numbers in the parentheses indicate the number of Cel+strains carrying identical chbR mutations.
1388 A. H. Kachroo et al.
© 2007 The Authors
Journal compilation © 2007 Blackwell Publishing Ltd, Molecular Microbiology, 66, 1382–1395

Page 8

tected an additional 5–10 bp region downstream to the
direct repeats compared with the wild-type ChbR (Fig. 6,
lane 1 versus 2; 3 versus 4). The effect of longer incuba-
tion with DNase I on the protection of the specific binding
site within chbOP was minimal, suggesting tight DNA
binding. These results corroborate well with the DNA
binding studies performed with the two proteins wherein
the formation of a discrete DNA–protein complex was
better in the case of ChbRN238S compared with ChbR
(Fig. 5). The footprinting assays carried out with the wild-
type and the mutant ChbRN238S, in the presence and
absence of cellobiose 6-phosphate, did not show any
alteration in the protection of the specific binding site
within chbOP (Fig. 6, compare lanes 1–4 and 5–8). The
DNase I footprinting assay may not be sensitive enough
to pick up changes in the conformation of the mutant
protein. Circular Dichroism studies were carried out to
detect possible conformational change induced by the
effector.
Conformational studies with ChbR and ChbRN238S
Circular Dichroism spectroscopy was carried out in the far
UV range and the results are presented in Fig. 7. The
wild-type ChbR protein was predominantly a-helical with
the two troughs corresponding to two known wavelengths
(209 nm and 222 nm). ChbRN238S appeared to be much
more structured compared with the wild-type protein as
indicated by the higher negative values. The ability of a
single-amino-acid change resulting in such changes in the
conformation of the protein was noticeable. Upon incubat-
ing the proteins with the DNA corresponding to the chb
regulatory region, there was a distinct conformational
change in both the wild-type and the mutant proteins. No
such change was induced by DNA fragments that did
not show binding in EMSAs (data not shown). Upon the
addition of cellobiose 6-phosphate (100 mM) to the
DNA-bound wild-type ChbR and ChbRN238S, a distinct
change in the spectrum was observed only in the case of
the mutant protein (Fig. 7). This effect by cellobiose
6-phosphate was specific as there was no change in the
spectrum when glucose 6-phosphate was used (data not
shown). Negligible difference was observed in the spectra
of the wild-type ChbR bound to DNA in the presence
of cellobiose 6-phosphate or glucose 6-phosphate. These
results are also consistent with the b-galactosidase
assays carried out with the wild-type and mutant chbR
that showed induction by cellobiose.
The denaturation kinetics of the wild-type and the
mutant ChbR over a temperature range of 10–90°C at
222 nm indicated that the overall stabilities of the two
proteins are similar (data not shown).
Discussion
The experiments outlined in this study were carried out
to determine the genetic mechanism of activation of the
chb operon that enables the transport and catabolism
of cellobiose. Analysis of cellobiose-utilizing derivatives
obtained from a large-scale isolation of Cel+mutants,
Fig. 5. A. The 188 bp regulatory region used
for electrophoretic mobility shift assays. The
NruI-digested DNA fragment (restriction site
shown) was also used for the assays.
B. EMSA carried out with ChbR and
ChbRN238S using body-labelled 188 bp
chbOP (2.5 nM). The assay was carried out at
two different concentrations of the protein (5
and 10 nM). Unlabelled chbOP was used at
three different concentrations (10, 20 and
40 nM) to compete out the labelled DNA.
A
B
Activation of the chb operon of E. coli enabling utilization of cellobiose1389
© 2007 The Authors
Journal compilation © 2007 Blackwell Publishing Ltd, Molecular Microbiology, 66, 1382–1395

Page 9

carried out using different strains of E. coli, revealed the
presence of two classes of mutations that act concertedly.
These were mutations that resulted in the loss of NagC
repression and gain-of-function mutations in the chbR
locus. These two classes of mutations were necessary
and sufficient to confer a Cel+phenotype. Contrary to
earlier speculations, additional mutations in the structural
genes were not necessary.
Is there any sequential order in which the two mutations
occur? The chances of random mutations being loss-of-
function mutations are usually higher and therefore accu-
mulation of nagC mutations could be the first step in this
process. So is the case of insertional disruption of the
NagC binding site within chbOP. This may give a low-level
growth advantage to the mutants, leading to the accumu-
lation of gain-of-function mutation within the chbR locus.
The frequency of mutations in a transcriptionally activated
system is higher compared with loci that are silent; a
process that is termed transcription-associated mutations
or TAM (Beletskii and Bhagwat, 1996; Klapacz and
Bhagwat, 2005). The partial derepression of chb tran-
scription in nagC mutants could lead to a higher mutation
rate at the chbR locus, leading to the selection of the
second mutation in chbR. The sequential activation of
the chb genes to confer a Cel+phenotype is similar to the
activation of the bgl operon that carries an insertion within
the bglF locus by two classes of mutations: activation of
the silent promoter and excision of the insertion within
bglF (Hall, 1988). This observation of a relatively high
frequency of the activating mutations was among the
initial results that suggested the existence of the phenom-
enon of adaptive mutations in bacteria (Foster, 1993).
The role of mutations in chbR was analysed using
the two alleles that resulted in a strong Cel+phenotype:
chbRN238S, the most abundant mutation obtained and
the other, chbRY30C. Clones carrying the mutant chbR
could activate transcription in the presence of cellobiose
in a reporter strain carrying disruptions of both the chro-
mosomal nagC and chbR loci. This activation of transcrip-
tion was lost when the assays were performed in a strain
carrying the chromosomal nagC and a disruption of chbR,
Fig. 6. DNase I footprinting assay carried out with ChbR and
ChbRN238S (30 nM). The end-labelled chb regulatory region of
~340 bp (10 nM) was allowed to form the complex with the proteins
at room temperature for 15 min. The complex was incubated with
DNase I (5 ng ml-1) for 2 and 5 min at 37°C before terminating the
reaction. The protected regions of the 340 bp chb regulatory region
along with the labelled end are shown. The dotted line below the
sequence indicates the extra 5–10 bp protection by ChbRN238S.
Fig. 7. Molecular ellipticity values of Circular
Dichroism spectra (200–300 nm) of the
wild-type ChbR and the mutant ChbRN238S
at 1 mM concentration plotted on the y-axis.
The chbOP (188 bp) DNA was added at 1 mM
concentration. Cellobiose 6-phosphate [C6P]
(100 mM) was used as an effector molecule.
1390A. H. Kachroo et al.
© 2007 The Authors
Journal compilation © 2007 Blackwell Publishing Ltd, Molecular Microbiology, 66, 1382–1395

Page 10

underscoring the necessity of the loss of NagC repression
in acquiring a Cel+phenotype.
ChbRN238S could not activate transcription in a
reporter strain carrying the wild-type chbR at the chromo-
somal locus and a disruption of nagC in the presence of
cellobiose indicating the dominant nature of wild-type
chbR over chbRN238S. This result suggests that ChbR
may function as a dimer/multimer. Interestingly, the
mutant ChbR proteins retained the ability to induce
transcription in the presence of chitobiose.
ChbR belongs to the family of AraC-XylS like tran-
scription regulators (Gallegos et al., 1997; Tobes and
Ramos, 2002). Broadly two characteristics are associ-
ated with these transcription factors – their ability to bind
two different sites within the promoter region regulated
by the presence or absence of the effector molecule and
the induction of a conformational change upon binding
the effector molecule. EMSA and DNase I protection
assays confirmed that ChbRN238S has acquired the
ability to form a more efficient discrete DNA–protein
complex. ChbRN238S could additionally protect an extra
stretch of ~10 bp DNA adjacent to the direct repeat.
These results are consistent with the enhanced basal
transcriptional activity seen in the case of ChbRN238S
in reporter gene assays. Ellipticity measurements with
the ChbR proteins indicated that, in the presence of cel-
lobiose 6-phosphate and DNA, there is a conformational
change in the case of ChbRN238S, possibly towards an
activation state, which was absent in the case of the
wild-type protein. This is also reflected in the reporter
assays where a distinct induction over the high basal
level could be observed in the presence of cellobiose.
The enhanced transcriptional activity of ChbRN238S
leading to a Cel+phenotype is therefore due to a higher
basal level of expression that can be correlated to tighter
binding to DNA, which is enhanced by the effector cel-
lobiose 6-phosphate.
The mutant chbR alleles, when introduced into a strain
lacking the chromosomal nagC and chbR, conferred on it
a Cel+phenotype. This experiment indirectly suggested
that the wild-type permease (chbBCA) and the phosphor-
b-glucosidase (chbF) could transport and hydrolyse
cellobiose. Direct evidence was provided by cloning
the genes encoding the wild-type permease and b-
glucosidase under a heterologous promoter. Therefore
the acquisition of the Cel+phenotype is the result of alter-
ation in the regulation of the chb operon. This observation
is consistent with recent bioinformatic studies (Babu et al.,
2006; Lozada-Chávez et al., 2006) which suggest that
mutations in the transcription regulatory networks (TRNs)
are primarily responsible for the variation in bacterial
phenotypes. The results also underscore the flexibility of
structural genes in recognizing substrates with similar
chemical structures.
PreliminarystudieswerecarriedoutwiththeCel+natural
isolates of E. coli. These studies showed that although a
similar pathway of acquisition of a Cel+phenotype exists in
some strains, additional genetic systems in cellobiose
metabolism are also involved in other isolates.
The mechanism of activation of the chb operon to confer
the ability of cellobiose utilization is presented as a model
(Fig. 8). Wild-type E. coli is capable of generating an
inducing signal for derepression of the chb operon when
N,N′-diacetylchitobiose is used as an inducer. The
N-acetyl glucosamine 6-phosphate that is generated after
phosphorylation and subsequent hydrolysis of N,N′-
diacetylchitobiose acts as an inducing signal for NagC,
thereby relieving NagC repression. However, cellobiose is
not capable of generating that inducing signal. Wild type
E. coli can abrogate NagC binding to the strong NagCII
binding site within chbOP either by transposition of IS
elements within chbOP or by mutations in nagC. The
derepression of the chb operon due to the loss of NagC
regulation is still not sufficient to confer a Cel+phenotype.
To change the specificity of the operon for cellobiose,
gain-of-functionmutations
necessary.
The chbR locus emerges as a focal point for the evolu-
tion of cellobiose utilization in E. coli. The ChbR of E. coli
K12, when compared across different natural isolates of
E. coli and related genera, showed differences. The pre-
liminary studies with chbR isolated from S. sonnei, E. coli
O157:H7 and two Cel+natural isolates suggested the
existence of polymorphism at the chbR locus. The chbR
locus, when compared across unrelated genera, shows
many differences. Wild-type strains of some of these bac-
teria (Citrobacter, Yersinia pestis, Klebsiella) are known to
be Cel+and the chb operon is conserved in these bacteria.
Some of the changes seen at the amino acid level in ChbR
in Yersinia and Citrobacter were similar to the changes
observed in ChbR isolated from various Cel+mutants of
E. coli (Fig. S3). These observations suggest that the chb
systemhasundergonedivergentevolutionoveraperiodof
time wherein one group of organisms has evolved to utilize
primarilychitobioseandanothergrouptoutilizecellobiose.
However, mutations in the regulatory genes might allow
them to switch from one to the other depending on the
environment. Bacteria can use this as an efficient strategy
for rapid evolution under selective pressure.
atthe chbR locusare
Experimental procedures
Construction of strains carrying disruption of chbOP and
deletion of chbR
The strain carrying an artificial insertion of the chlorampheni-
col acetyl transferase gene within the chb regulatory region
(chbOP::cat) was constructed by targeted homologous
recombination in DY330 using the plasmid pKD3 (Yu et al.,
Activation of the chb operon of E. coli enabling utilization of cellobiose1391
© 2007 The Authors
Journal compilation © 2007 Blackwell Publishing Ltd, Molecular Microbiology, 66, 1382–1395

Page 11

2000).The oligonucleotide primers were chosen such that the
disruptionmimickedanaturalinsertionseeninCel+mutants.A
PCR fragment carrying the cat gene flanked by chbOP
sequences was introduced into DY330 after induction of the l
redrecombinationsystemandrecombinantswereselectedon
LB chloramphenicol medium. The strains DY330 (chbR::cat)
was generated using a similar strategy. A PCR fragment
coveringthecatgenewasamplifiedusingpKD3asatemplate
and 56 nucleotides long primers. The 5′ 36 nucleotides of the
forward and reverse primers corresponded to the N- and
C-terminals of the chbR gene respectively. A similar strain,
DY330(chbR::kan),wasconstructedusingtheplasmidpKD4.
The strain JF201 (chbOP::cat) was made by transducing the
(chbOP::cat) allele into JF201, an E. coli strain deleted for the
chromosomal bgl operon. Similarly, JF201 (chbR::cat) was
constructed by transducing the chbR::cat allele from the
parent DY330 (chbR::cat) into JF201.
Construction of plasmids
The plasmid carrying wild-type nagC was constructed by
PCR amplifying the nagC locus from JF201 using the primers
AHK7(5′-GCGAATTCATGACACCAGGCGGACAAGC-3′)
and AHK8 (5′-GCGGATCCTTAATTTTCCAGCAAATGC-3′).
Fig. 8. A model for the mutational activation
of the chb operon of E. coli allowing utilization
of cellobiose. See text for details.
1392 A. H. Kachroo et al.
© 2007 The Authors
Journal compilation © 2007 Blackwell Publishing Ltd, Molecular Microbiology, 66, 1382–1395

Page 12

The ~1.1 kb PCR fragment was initially cloned in pDRIVE
(Qiagen) and subcloned in pBR322.
The plasmid placB-F carrying the wild-type chb permease
(chbBCA) and the phosphor-b-glucosidase (chbF) without
the chb regulatory region was constructed by PCR amplifying
a ~5.1 kb DNA fragment from DY330 (chbR::kan; Cel–) using
theforward primerAHK9
GAAACACATT-3′) and the reverse primer AHK11 (5′-
GGAATTCCAGCCTCGGTTAATGTGC-3′). The PCR frag-
ment was digested using SacI and EcoRI restriction enzymes
and cloned at these sites within pACDH, a low-copy vector
derived from pACYC (Singh et al., 2005). The individual
clones containing the permease (chbF) and the phospho-b-
glucosidase (chbF) were constructed by PCR amplification of
the genes using appropriate sets of primers flanking the
ORFs. The complete chb operon from the Cel+natural isolate
NC2.1 was cloned by PCR amplifying a 4.5 kb DNA fragment
containing chbBCARF using the primer set ASK1 (5′-
GCGAATTCAACAACGGAAACC GGCC-3′) and AHK5 (5′-
CGGATCCTTAATCGCCGGATGCAAGG-3′) using genomic
DNA isolated from NC2.1. The EcoRI- and BamHI-digested
PCR product was cloned at these sites in pBR322.
The plasmids used in the overexpression of ChbR and
ChbRN238S were made as follows. The 850 bp chbR locus
from wild-type and the mutant strains was amplified by PCR
using Pfu DNApolymerase (MBI, Fermentas) and the forward
primerAHK10carrying an
ATGCAGCCAGTGATTAACGC-3′) and the reverse primer
AHK6 carryinga BamHI
TGAATTGTCAGGT-3′). The NdeI- and BamHI-digested DNA
fragments were cloned into an expression vector pJES307
(Tabor and Richardson, 1985; Plumbridge and Pellegrini,
2004). The clones were confirmed by sequencing.
(5′-CGAGCTCGATGGAAAA
NdeI site (5′-CCATATG
site(5′-CGGGATCCATATG
Method for large-scale isolation and characterization of
Cel+mutants
The E. coli strain JF201 was streaked on MacConkey cello-
biose plates and was spread on M9 cellobiose plates (without
casamino acids). Cellobiose used (SigmaAldrich) was free of
glucose. The plates were incubated for the first 24 h at 37°C
and subsequently incubated at room temperature (~25°C).
Cel+mutants appeared as papillae on existing colonies on the
MacConkey cellobiose plates and as distinctly growing colo-
nies on M9 cellobiose plates. The Cel+mutants started to
appear after 10 days on M9 cellobiose plates and after
20 days on MacConkey cellobiose plates. A total of 53 single
papillae or colonies were picked and streaked on fresh
MacConkey cellobiose plates. The papillae picked up from
MacConkey cellobiose plates are named JF-MCP-01 to
JF-MCP-28 and the Cel+mutants picked up from M9 cello-
biose plates are named J-M9-01 to J-M9-25.Atotal of 62 Cel+
mutants were similarly isolated from the strain MG1655. The
Cel+mutants picked up from MacConkey cellobiose plates
were named MG-MCP-01 to MG-MCP-32 and the Cel+
mutants picked up from M9 minimal cellobiose plates were
named MG-M9-01 to MG-M9-30.
DNA sequencing and analysis
DNAsequencing was carried out at Macrogen, Korea and the
in-house facility at the Indian Institute of Science, Bangalore,
India. Both strands were sequenced in all cases. Analysis of
the DNA and protein sequences was performed using the
Clone Manager-5 software, SciEd.; Vector NT1/Align soft-
ware, Invitrogen; CLUSTALW (http://www.ebi.ac.uk/clustalw/
http://www.ch.embnet.org/software/ClustalW.html);
NCBI BLAST (http://www.ncbi.nlm.nih.gov/BLAST/).
and
b-Galactosidase assay
Assays for b-galactosidase activity were carried out as
described by Miller (1972). Cells were grown in LB broth or in
M9 minimal medium supplemented with 0.4% glycerol and
0.4% casamino acids. Cellobiose (10 mM) or 0.1% (2.4 mM)
N,N′-diacetylchitobiose (Seikagaku Corporation, Japan) were
used as inducer when required.Average values of Miller units
of activity were computed based on at least three indepen-
dent measurements in each case.
Overexpression and purification of wild-type ChbR
and ChbRN238S
Wild-type ChbR was purified as per published protocol (Plum-
bridge and Pellegrini, 2004). The plasmid pJES ChbR was
introduced into the E. coli strain Rosetta (DE3) by transforma-
tion, selecting for resistance to ampicillin and chlorampheni-
col. Cultures (1000 ml) in 2¥ LB broth containing ampicillin
(100 mg ml-1) were grown at 37°C to 0.7 OD600. T7 RNA
polymerase synthesis was induced by the addition of 0.5 mM
IPTG. The cultures were harvested after 3 h, pelleted at
5000 r.p.m. for 5 min.The pellet was re-suspended in 25 ml of
buffer B (20 mM MES, pH 6.5, 1 mM EDTAand 10% glycerol)
containing 200 mM NaCl and a protease inhibitor cocktail
(Sigma Aldrich) at 4°C. All subsequent steps were carried
out at 4°C. The mixture was sonicated and centrifuged at
14 000 r.p.m. for 30 min. The supernatant was passed
through a 0.45 mm filter (Sartorius) and the filtrate was applied
ontoa1 mlheparinsepharosecolumn(USB)pre-equilibrated
with buffer B (with 200 mM NaCl). ChbR eluted at 600 mM
NaCl. The subsequent step of purification on 5 ml of mono-S
was similar to the published protocol (Plumbridge and Pelle-
grini, 2004). However, at the step of hydroxyl-apatite column
purification, a modification was made at the step (iii) of
washing carried out post the loading of the sample. Four steps
of equilibration and washing were used: (i) protein was loaded
ontothecolumnandwashedinbufferBwith400 mMNaCl,(ii)
washed with buffer B containing 3 M NaCl, (iii) washed with
bufferBcontaining200 mMpotassiumphosphate,adjustedto
pH 6.5,and(iv)elutedwithagradient200–800 mMphosphate
in buffer B. The solution containing pure wild-type ChbR
protein was dialysed against buffer B containing 0.3 M NaCl
and 50% glycerol and stored at -20°C.
ChbRN238S overexpression was performed by a proce-
dure similar to that used in the case of wild-type ChbR.
However, ChbRN238S could not be purified via heparin
sepharose column although the purification was repeated
many times with different salt concentrations (minimum of
150 mM NaCl below which the protein precipitates). The
mutant protein was then loaded on to a mono-S cation
exchange column and a linear gradient of 200 mM to 1 M
NaCl was applied. The protein eluted at 450 mM NaCl
concentration. The extractions containing partially purified
Activation of the chb operon of E. coli enabling utilization of cellobiose 1393
© 2007 The Authors
Journal compilation © 2007 Blackwell Publishing Ltd, Molecular Microbiology, 66, 1382–1395

Page 13

protein (~80%) were dialysed against buffer C (50 mM Tris
pH 6.5 containing 200 mM NaCl) and applied onto a mono-Q
anion exchange column (BIO-RAD). The protein eluted in
the flow-through. The purified protein fractions (~90%) were
dialysed independently with buffer B containing 50% glycerol
and 300 mM NaCl and stored at -20°C for further use.
Different dilutions of wild-type ChbR and ChbRN238S were
run on 12% SDS-PAGE for checking the purity and estimating
the amount of the proteins. The concentrations were also
checked using a Nanodrop spectrophotometer.
Labelling of DNA probes
For generating internally labelled probe, PCR reactions were
carried out in the presence of [a-32P]-dATP and the PCR
fragments were purified from 2% agarose gel using
Qiagen and Amersham gel purification kits. The eluate was
further purified by passing through a Sephadex G50 column.
The probe, after ethanol precipitation and wash, was
re-suspended in appropriate volumes of sterile water. The
purity of the DNAfragments was tested by electrophoresis on
12–15% polyacrylamide (19:1) or on 6% polyacrylamide gels
containing 8 M Urea. Oligonucleotides were end-labelled
using T4 polynucleotide kinase (MBI Fermentas) and [g-32P]-
ATP. The labelling was performed as per the protocol pro-
vided by the manufacturer. The labelled DNA was purified by
passing through Sephadex G50 mini column.
Electrophoretic mobility shift assay (EMSA)
The template used in EMSAwas a body-labelled 188 bp DNA
fragment carrying chbOP, PCR amplified using the primers
ASK1(5′-GCGAATTCAACAACGGAAACCGGCC-3′)
AHK14 (5′-CGGATCCGGGCTGAAAGGAGTATACG-3′). In
addition, a ~120 bp chbOP fragment generated by digestion
of 188 bp fragment with NruI and a 60 bp synthetic oligo-
nucleotide dimer containing the direct repeat known to be a
specific binding site for ChbR were also used. The DNA
concentration in binding reactions was 2.5 nM and protein
concentrations were 5 and 10 nM in a final volume of
10–15 ml. The assays were performed in 25 mM HEPES
buffer containing 100 mM K glutamate (pH 8.0). The mixture
was incubated at 4°C for 15–30 min. Competition with unla-
belled specific DNAwas carried out by incubating the mixture
with different concentrations of cold DNA (0–40 nM) for
another 15 min. The samples were directly loaded onto a 6%
native PAGE (1¥ TBE buffer system). The gels were run at
4°C and bands were visualized using a Bio-image Analyser
(Fuji film, Japan). The intensity of the bands was measured
using Fuji and Kodak ID imaging software.
and
DNase I footprinting assay
The protocol for the assay was the same as that described
by Plumbridge and Pellegrini (2004). The template used in
the assays was a ~350 bp DNA fragment obtained by PCR
using theend-labelledforward
GCGAATTCAACAACGGAAACCGGCC-3′) and the reverse
primer ASK2 (5′-CGGGATCCTGATACCAGTAAAGAGG-3′).
The proteins ChbR and ChbRN238S were used at a concen-
primer ASK1(5′-
tration of 30 nM. The DNase I concentration was 5 ng ml-1and
the reactions were incubated at 37°C for 2 and 5 min.Assays
were carried out in the presence or absence of 100 mM cel-
lobiose 6-phosphate. The reactions were stopped by the
addition of phenol. After ethanol precipitation of the aqueous
layer overnight at -20°C, the samples were analysed on 6%
denaturing polyacrylamide gels and quantified as above.
Circular Dichroism
Circular Dichroism spectroscopy was carried out using a
JASCO J-810 CD spectrometer. Measurements were made
using 400 ml volumes of proteins (1 mM) in quartz cuvettes
of 1 mm path length. All the assays were carried out in
buffer B (containing 200 mM NaCl). The 188 bp DNA frag-
ment carrying the chbOP region at 1 mM concentration was
used in the assays. The DNA–protein complex was allowed
to incubate for 15–30 min on ice after which the CD spec-
troscopy was carried out in the wavelength range of 200–
300 nm at 20°C. To one set of the DNA–protein solutions,
purified cellobiose 6-phosphate at a final concentration of
100 mM was added and incubated on ice for another 15 min
before the spectroscopic measurement. The CD values
were transformed into molecular ellipticity values using soft-
ware from JASCO.
Acknowledgements
We thank J. Plumbridge for many valuable suggestions and
for providing single-copy reporter strains and J. Thompson for
the gift of purified cellobiose 6-phosphate. We also thank
D.N. Rao, U.K. Madhusoodanan and the Department of
Biochemistry, Indian Institute of Science, for help with CD
spectroscopy. We are grateful to the reviewers of the earlier
version of the manuscript who suggested several im-
provements. This work was funded by a grant form the
Council of Scientific and Industrial Research (CSIR), Govern-
ment of India.
References
Blattner, F.R., Plunkett, G., 3rd, Bloch, C.A., Perna, N.T.,
Burland, V., et al. (1997) The complete genome sequence
of Escherichia coli K-12. Science 277: 1453–1474.
Babu, M.M., Teichmann, S.A., and Aravind, L. (2006) Evolu-
tionary dynamics of prokaryotic transcriptional regulatory
networks. J Mol Biol 358: 614–633.
Beletskii, A., and Bhagwat, A.S. (1996) Transcription-induced
mutations: increase in C to T mutations in the non-
transcribed strand during transcription in Escherichia coli.
Proc Natl Acad Sci USA 93: 13919–13924.
Boliver, F., Rodriguez, R.L., Greene, P.J., Betlach, M.C.,
Heyneker, H.L., and Boyer, H.W. (1977) Construction and
characterization of new cloning vehicles II. A multipurpose
cloning system. Gene 2: 95–113.
Foster, P. (1993) Adaptive mutation: the use of adversity.
Annu Rev Microbiol 47: 467–504.
Gallegos, M., Schleif, R., Bairoch, A., Hofmann, K., and
Ramos, J. (1997) AraC/XylS family of transcriptional
regulators. Microbiol Mol Biol Rev 61: 393–410.
1394 A. H. Kachroo et al.
© 2007 The Authors
Journal compilation © 2007 Blackwell Publishing Ltd, Molecular Microbiology, 66, 1382–1395

Page 14

Hall, B.G. (1988) Adaptive evolution that requires multiple
spontaneous mutations. I. Mutations involving an insertion
sequence. Genetics 120: 887–897.
Keyhani, N.O., and Roseman, S. (1997) Wild-type Escheri-
chiacoligrows onthe
diacetylchitobiose, by expressing the cel operon. Proc Natl
Acad Sci 94: 14367–14371.
Keyhani, N.O., Wang, L., Lee, Y.C., and Roseman, S.
(2000a) The chitin disaccharide N,N’-diacetylchitobiose
is catabolized by Escherichia coli and is transported/
phosphorylated by the phosphoenolpyruvate: glycose
phosphotransferase system. J Biol Chem 275: 33084–
33090.
Keyhani, N.O., Bacia, K., and Roseman, S. (2000b) The
transport/phosphorylation of
Escherichia coli. J Biol Chem 275: 33102–33109.
Klapacz, J., and Bhagwat, A.S. (2005) Transcription promotes
guanine to thymine mutations in the non-transcribed strand
of an Escherichia coli gene. DNA Repair 4: 806–813.
Kricker, M., and Hall, B.G. (1984) Directed evolution of cel-
lobiose utilization in Escherichia coli. Mol Biol Evol 1: 171–
182.
Kricker, M., and Hall, B.G. (1987) Biochemical genetics of the
cryptic gene system for cellobiose utilization in Escherichia
coli K-12. Genetics 115: 419–429.
Lozada-Chávez, I., Janga, S.C., and Collado-Vides, J. (2006)
Bacterial regulatory networks are extremely flexible in
evolution. Nucleic Acids Res 34: 3434–3445.
Miller, J. (1972) Experiments in Molecular Genetics. Cold
Spring Harbor, NY: Cold Spring Harbor Laboratory Press.
Parker, L.L., and Hall, B.G. (1988) A fourth Escherichia coli
gene system with the potential to evolve b-glucoside
utilization. Genetics 119: 485–490.
Parker, L.L., and Hall, B.G. (1990a) Characterization and
nucleotide sequence of the cryptic cel operon of Escheri-
chia coli K12. Genetics 124: 455–471.
Parker, L.L., and Hall, B.G. (1990b) Mechanisms of activation
of the cryptic cel operon of Escherichia coli K12. Genetics
124: 473–482.
Plumbridge, J., and Pellegrini, O. (2004) Expression of the
chitobiose operon of Escherichia coli is regulated by three
transcription factors: NagC, ChbR and CAP. Mol Microbiol
52: 437–449.
Reynolds, A.E., Mahadevan, S., Le Grice, S.F.J., and Wright,
chitin disaccharide,N,N?-
N,N’-diactylchitobiosein
A. (1986) Enhancement of bacterial gene expression by
insertion elements or by mutations in a CAP-cAMP binding
site. J Mol Biol 191: 85–95.
Singh, N.S., Das, G., Seshadri, A., Sangeetha, R., and Varsh-
ney, U. (2005) Evidence for a role of initiation factor 3 in
recycling of ribosomal complexes stalled on mRNAs in
Escherichia coli. Nucleic Acid Res 33: 5591–5601.
Tabor, S., and Richardson, C.C. (1985) A bacteriophage T7
RNA polymerase/promoter system for controlled exclusive
expression of specific genes. Proc Natl Acad Sci USA 82:
1074–1078.
Thompson, J., Ruvinov, S.B., Freedberg, D.I., and Hall, B.G.
(1999)Cellobiose-6-phosphate
Escherichia coli: characterization and assignment to the
unusual family 4 of glycosylhydrolases. J Bacteriol 181:
7339–7345.
Tobes, R., and Ramos, J.L. (2002) AraC-XylS database: a
family of positive transcriptional regulators in bacteria.
Nucleic Acid Res 30: 318–321.
Woodcock, D.M., Crowther, P.J., Doherty, J., Jefferson, S.,
DeCruz, E., Noyer-Weidner, M., et al. (1989) Quantitative
evaluation of Escherichia coli host strains for tolerance to
cytosine methylation in plasmid and phage recombinants.
Nucleic Acid Res 17: 2469–2478.
Wright, B.E. (2004) Stress-directed adaptive mutations and
evolution. Mol Microbiol 52: 643–650.
Yu, D., Elis, H.M., Lee, E.C., Jenkins, N.A., Copeland, N.G.,
and Court, D.L. (2000) An efficient recombination system
for chromosome engineering in Escherichia coli. Proc Natl
Acad Sci USA 97: 5978–5983.
hydrolase(CelF)of
Supplementary material
This material is available as part of the online article from:
http://www.blackwell-synergy.com/doi/abs/10.1111/
j.1365-2958.2007.05999x
(This link will take you to the article abstract).
Please note: Blackwell Publishing is not responsible for the
content or functionality of any supplementary materials sup-
plied by the authors. Any queries (other than missing mate-
rial) should be directed to the corresponding author for the
article.
Activation of the chb operon of E. coli enabling utilization of cellobiose 1395
© 2007 The Authors
Journal compilation © 2007 Blackwell Publishing Ltd, Molecular Microbiology, 66, 1382–1395