(Received for publication, February 12, 1997, and in revised form, April 22, 1997)
From the Department of Molecular Biology, The Hebrew University-Hadassah Medical School, P. O. Box 12272, Jerusalem 91120, Israel
BglG, the response regulator of the bgl sensory system, was recently shown to be phosphorylated on a histidine residue. We report here the localization of the phosphorylation site to histidine 208. Localization of the phosphorylated histidine was carried out in two steps. We first engineered BglG derivatives with a specific protease (factor Xa) cleavage site that allowed asymmetric splitting of each prephosphorylated protein to well defined peptides, of which only one was labeled by radioactive phosphate. This allowed the localization of the phosphorylation site to the last 111 residues. Subsequently, we identified the phosphorylated histidine by mutating each of the three histidines located in this region to an arginine and following the ability of the resulting mutants to be in vivo regulated and in vitro phosphorylated by BglF, the bgl system sensor. Histidine 208 was the only histidine which failed both tests. The use of simple techniques to map the phosphorylation site should make this protocol applicable for the localization of phosphorylation sites in other proteins. The bgl system represents a new family of sensory systems. Thus, the mapping reported here is an important step toward the definition of the functional domains involved in the transduction of a signal by the components that constitute systems of this novel family.
The bgl operon in Escherichia coli, induced
by environmental signal (-glucosides), is regulated by a novel
sensory system which consists of a membrane-bound sensor, BglF, and a
response regulator, BglG (1). The response regulator is an RNA-binding protein which controls operon expression by transcriptional
antitermination (2). The sensor is a phosphotransferase system
transport protein which controls the activity of the response regulator
by reversible phosphorylation according to
-glucoside availability
(3-5). Reversible phosphorylation of the response regulator by the
sensor was shown to modulate its dimeric state (6).The bgl
system is not a member of the known family of two-component regulatory systems involved in signal transduction (reviewed in Refs. 7-10). The
bgl proteins, BglF and BglG, share no homology with the
sensors and regulators of the two-component systems, respectively.
Moreover, it was recently shown that BglG, the response regulator of
the bgl system, is phosphorylated on a histidine
residue,1 unlike response regulators of the
two-component systems which are phosphorylated on an aspartate. Hence,
the bgl system represents a novel family of sensory systems.
Other systems in different organisms were suggested to affiliate to
this new family (11-19). To understand the rules of recognition and
interaction between the components which constitute systems of the new
family, it is important to define the functional domains involved in
the transduction of a signal by these components. Localization and characterization of the dimerization and the phosphorylation sites on
BglG is also important for the elucidation of the mechanism by which
phosphorylation affects the dimeric state of a protein.
In this paper we report that identification of the phosphorylated residue on BglG, the response regulator of the bgl system, as histidine 208. Mapping of the phosphorylated histidine was carried out in two steps: the first step defined the protein region to which the phosphorylation site maps; the second step defined the specific histidine which is phosphorylated. In the first step we introduced factor Xa cleavage sites to BglG. Proteolysis after in vitro phosphorylation with radioactive phosphate allowed the localization of the phosphorylation site to the last 111 residues of the protein. In the second step we mutated the three histidines located in this region, each at a time, to arginines and followed the ability of the mutants to be in vivo regulated and in vitro phosphorylated by BglF. The ability of all the BglG mutant derivatives used in this study to fold correctly was deduced from various assays that monitored their ability to function as transcriptional antiterminators. Our protocol for the localization of the BglG phosphorylation site can be adopted and applied to localize phosphorylation sites of other proteins.
The following E. coli K12 strains were
used: K38 (HfrC thi +); AE304-7 and AE304-9,
both carry a defective bglG gene and an activating mutation
in bglR (20); MA152 and MA200, both carry a
bgl
-lacZ fusion on their chromosome (
bglR7 bglG
lacZ+ lacY+), but while the
first is
bgl, the second is bgl+
(20). Salmonella typhimurium strain LJ144
(cpd-401, cysA1150/F
198), contains the
pts operon on an E. coli plasmid, F
198, and thus produces increased levels of Enzyme I, HPr, and Enzyme
IIAglc (21).
Plasmids pT712 and pT713, containing the phage T7
late promoter, and plasmid pGP1-2, carrying the T7 RNA polymerase gene
under control of the CI857 repressor, were obtained from Life
Technologies, Inc. Plasmid pT7OAC-F carries the entire bglF
gene cloned downstream of the T7 promoter in pT712; plasmid pT7FH-G
carries the entire bglG gene cloned downstream of the T7
promoter in pT713 (3). Plasmid pMN25 carries the entire bglG
gene cloned in pBR322 (20).
The following plasmids were constructed by introducing insertion or base substitution mutations (see the details of the mutagenesis below) into the bglG gene in plasmid pT7FH-G. pCQ-G1 and pCQ-G2 encode for BglG with an insertion of factor Xa cleavage site, Ile-Glu-Gly-Arg, between amino acids 113 and 114 and amino acids 167 and 168, respectively. pCQ-G4, pCQ-G5 and pCQ-G6 encode for BglG with His-208, His-219, and His-278 mutated to Arg, respectively.
MediaEnriched media, M9 salts, and M63 salts minimal media were prepared essentially as described by Miller (22). The minimal medium used for [35S]methionine labeling was the same as that used by Tabor and Richardson (23) with 0.4% succinate as the carbon source. The minimal medium used for [3H]tryptophan labeling was the same as that used for [35S]methionine labeling, except that it lacked tryptophan rather than lacking methionine and cysteine. Ampicillin (200 µg/ml) and kanamycin (30 µg/ml) were included in the media when growing strains which contain plasmids that confer resistance to either one of these antibiotics. MacConkey arbutin plates were prepared as described previously (24). MacConkey lactose plates were prepared from lactose MacConkey agar (Difco).
ChemicalsFactor Xa was obtained from New England Biolabs.
[-32P]ATP (3000 Ci/mmol) was obtained from Rotem
Industries Ltd. (Beer-Sheva, Israel). [35S]Methionine
(1200 Ci/mmol) was obtained from DuPont. [3H]Tryptophan
(33 Ci/mmol) was obtained from American Radiolabeled Chemicals Inc.
PEP,2 pyruvic acid, and pyruvate kinase
were obtained from Sigma. [32P]PEP was prepared and
separated from [32P]ATP as described before (3).
All manipulations with recombinant DNA were carried out by standard procedures (25). Restriction enzymes and other enzymes used in recombinant DNA experiments were purchased commercially and were used according to the specifications of the manufacturers.
Site-directed MutagenesisSite-directed mutagenesis was
carried out by overlap extension with polymerase chain reaction as
described by Ho et al. (26). To mutate the bglG
gene in plasmid pT7FH-G to its alleles that encode for BglG derivatives
containing the factor Xa cleavage site insertion (Ile-Glu-Gly-Arg), the
following primers were used: 5ATTGAAGGGCGCCTGCTGCCCAACCC3
and
5
GGCGCCCTTCAATCACGTTTTGCTGAAAG3
to create pCQ-G1 with the
insertion introduced between codons 113 and 114 of the bglG
gene; 5
ATAGAGGGCCGGGGAAATATGGAGGATG3
and
5
CCGGCCCTCTATGCTCATTTGGGCACT3
to create pCQ-G2 with the insertion introduced between codons 167 and 168 of the bglG
gene.
To mutate the bglG gene in plasmid pT7FH-G to its alleles
that encode for BglG derivatives containing an arginine instead of a
certain histidine, the following primers were used:
5GACTGGTTACGCGTCTGAAG3
and its complement to create pCQ-G4 encoding
BglG whose His-208 is replaced by Arg; 5
CGTATTCTTGAACGGGCCTCA3
and
its complement to create pCQ-G5 encoding BglG whose His-219 is replaced
by Arg; 5
CAAAGAACGTTAACACCTGCAG3
and its complement to create pCQ-G6 encoding BglG whose His-278 is replaced by Arg.
The mutations introduced new sites for restriction enzymes which were useful during the screening for the mutant plasmids. The mutations were confirmed by DNA sequence determinations.
Measurements ofAssays for
-galactosidase activity were carried out as described by Miller
(22). Cells were grown in minimal medium which was supplied with 0.4%
succinate as a carbon source.
Cell extracts enriched for various BglG and membrane fractions enriched for BglF were prepared as described before (3). The proteins were expressed from their respective genes cloned under T7 promoter control in plasmids pT7FH-G, pCQ-G1, pCQ-G2, pCQ-G4, pCQ-G5, pCQ-G6, and pT7OAC-F. Expression of T7 RNA polymerase, specified by plasmid pGP1-2 which is compatible with the above plasmids, was induced thermally. The E. coli K38 strain was used as a host. A soluble fraction from S. typhimurium LJ144, which overproduces Enzyme I and HPr, was prepared as described by Begley et al. (27).
In Vitro Phosphorylation of BglG and Its Mutant DerivativesPhosphorylation of BglF, present in membrane fractions, in the presence of [32P]PEP and an S. typhimurium LJ144 cytoplasmic fraction (used as the source of Enzyme I and HPr), was carried out as described by Amster-Choder et al. (3). Phosphorylation of the various BglG derivatives was carried out by incubating cellular extracts enriched for these proteins with mixtures containing prephosphorylated BglF, as described by Amster-Choder et al. (3).
[35S]Methionine- or [3H]Tryptophan Labeling of BglG DerivativesCells containing plasmids carrying different bglG alleles (encoding the various BglG derivatives) under the control of the phage T7 promoter were induced and labeled with either [35S]methionine or [3H]tryptophan in the presence of rifampicin as described by Tabor and Richardson (23). Labeling times were 2 min or 60 min when labeling with [35S]methionine or [3H]tryptophan, respectively.
Purification of Isotope-labeled BglG Derivatives and Their Proteolysis with Factor Xa[32P]BglG was separated from [32P]BglF, a membrane protein, by centrifugation at 150,000 × g for 1 h at 4 °C. The supernatant was fractionated on 12.5% SDS-polyacrylamide gel. After electrophoresis, the wet gel was exposed to an x-ray film at 4 °C. A gel slice containing the 32P-labeled BglG band was excised, and the protein was extracted from it by incubation in factor Xa buffer (20 mM Tris-HCl, 100 mM NaCl, and 2 mM CaCl2) for 2 h at 37 °C. Purified [35S]BglG was obtained by the same procedure. To purify [3H]BglG, [35S]BglG was used as a marker for the [3H]BglG position on gel, due to the inability to detect 3H-labeled proteins after autoradiography without prior enhancement. This procedure was applied for the purification of all isotope-labeled BglG derivatives. Proteolysis with factor Xa was carried out by incubating 50-100 µg of the purified radioactively labeled proteins with 15 µg/ml factor Xa and 1.5 mg/ml bovine serum albumin in factor Xa buffer at 23 °C for 6 h. Reactions were terminated by 1% EDTA.
Electrophoresis and AutoradiographyProteins were incubated
for 30 min at 30 °C in electrophoresis sample buffer containing 62.5 mM Tris-HCl (pH 6.8), 2% SDS, 5% -mercaptoethanol,
10% glycerol, and 0.01% bromphenol blue. Electrophoresis of proteins
was carried out on 10% SDS-polyacrylamide gels as described by Laemmli
(28). Alternatively, Tricine-SDS-polyacrylamide gels with 4% stacking
gel, 10% spacer gel, and 16.5% separating gel, prepared as described
by Schagger and von Jagow (29), were used when indicated. Samples were
fractionated alongside prestained low or mid-range protein molecular
weight markers (Amersham). After electrophoresis, gels with
3H-labeled proteins were fixed by 50% methanol and 10%
acetic acid for 30 min, followed by H2O wash for 30 min,
and then enhanced with 0.5 M salicylic acid, 0.5 M NaOH, and 1% glycerol for 30 min (30). All gels were
dried and exposed to Kodak XAR-5 x-ray film at
70 °C.
Previous studies have demonstrated the
existence of two forms of BglG in vivo, phosphorylated and
non-phosphorylated, implying that there is a single phosphorylated
species of BglG (4). The phosphorylated residue on BglG was identified
as a histidine.1 As a first step toward the localization of
the phosphorylated histidine on BglG, we attempted to identify the
region on BglG that contains this histidine. To do that, we took
advantage of the idea that a cleavage site for a specific protease can
be inserted into proteins by in vitro manipulations (31).
Our plan was to engineer BglG mutants, each containing a factor Xa site
located asymmetrically in the protein, to enable cleavage of the mutant proteins to two fragments of different sizes. Cleavage of these mutant
BglG derivatives following their in vitro phosphorylation is
expected to give only one 32P-labeled fragment in each
case, due to the phosphorylation of BglG on a single residue. The
ability of the engineered BglG derivatives to fold correctly, dimerize,
bind RNA, and antiterminate bgl transcription can be
verified by testing their ability to complement a chromosomal mutation
in the bglG gene and enable -glucoside utilization by the
mutant strains, when expressed from a plasmid.
We first introduced a factor Xa site between amino acids 113 and 114 of
BglG by site-directed mutagenesis (see "Experimental Procedures")
to generate MG1. As shown in the experimental design scheme, presented
in Fig. 1, the site was introduced asymmetrically to
enable cleavage of the mutant protein to two polypeptides, "a" and
"b," consisting of 117 amino acids and 165 amino acids, respectively. Indeed, incubation of 35S-labeled MG1 with
factor Xa resulted in the appearance of two labeled protein fragments
with molecular masses corresponding to those expected for peptides a
and b, respectively (Fig. 2, lane 2). The
stronger intensity of b relative to a is due to the higher methionine
content in the former. The two fragments did not appear when wild-type
BglG was incubated with factor Xa (Fig. 2, lane 1). In
addition, incubation of both wild-type BglG and MG1 with factor Xa
resulted in nonspecific cleavage that generated slightly shortened
derivatives of these proteins (Fig. 2, lanes 1 and
2). However, this did not interfere with the identification of the specific cleavage products, a and b. In vitro
phosphorylation of MG1 prior to proteolysis with factor Xa resulted in
the appearance of one 32P-labeled fragment, the b fragment
(Fig. 2, lane 3). To indisputably identify the slower
migrating fragment as polypeptide b, the C-terminal fragment, we took
advantage of the fact that all tryptophans in BglG are clustered in
fragment b. Thus, cleavage of BglG, which was labeled with
[3H]tryptophan, by factor Xa is expected to generate only
one labeled fragment, the b fragment. The results presented in Fig.
3 demonstrate that indeed
[3H]tryptophan-labeled MG1 generates only one labeled
fragment, the slower migrating fragment, after incubation with factor
Xa (Fig. 3, lane 2). This is in contrast to the two labeled
fragments produced by factor Xa proteolysis of [35S]MG1
(Fig. 3, lane 1). Thus, this experiment verifies the
identification of the slower migrating fragment as polypeptide b. The
results obtained with MG1 indicate that the phosphorylation site
resides in the last 165 amino acids of BglG. The possibility that MG1 is misfolded and thus phosphorylated differently from wild type was
ruled out by the demonstrated ability of MG1 to complement bglG strains the same as wild type (Table
I).
|
We next constructed MG2 that contains a factor Xa site in a position different from MG1. This was accomplished by introducing a factor Xa site between amino acids 167 and 168. Cleavage of MG2 with factor Xa is expected to generate two fragments, "c" and "d," consisting of 171 amino acids and 111 amino acids, respectively (see Fig. 1). Incubation of 35S-labeled MG2 with factor Xa resulted indeed in the appearance of two labeled fragments with molecular masses corresponding to those expected for peptides c and d (Fig. 2, lane 4). Proteolysis of in vitro phosphorylated MG2 with factor Xa yielded one 32P-labeled fragment, the d fragment (Fig. 2, lane 5). MG2, like MG1, strongly complemented bglG strains (Table I). Based on the results with MG1 and MG2, we could conclude that the phosphorylation site resides in the last 111 amino acids of BglG.
Mapping the Phosphorylated Amino Acid on BglGWe subsequently
constructed three mutants of BglG, each with one of the three
C-terminal histidines of BglG (His-208, His-219, and His-278, marked as
asterisks in Fig. 1) mutated to an arginine, to generate MG4
(H208R), MG5 (H219R), and MG6 (H278R). These mutants were checked for
their ability to complement bglG strains (Table I). The
plasmid-encoded MG4, MG5, and MG6 mutants strongly complemented the
chromosomal mutation in the bglG gene (the same as wild-type BglG) and enabled growth of the mutant strains on -glucosides. The
results of this analysis suggest that these mutants can form dimers,
bind to BglG target site on the RNA, and lead to antitermination of
bgl transcription.
To directly demonstrate the ability of the three His to Arg mutants to
antiterminate transcription, we made use of strain MA152 which is
deleted for the bgl operon and carries a chromosomal bgl-lacZ fusion (20). The lacZ gene is not
expressed in this strain, because transcription terminates at the
bgl terminator, located upstream of the lacZ
gene. Expression of plasmid-encoded BglG renders the lacZ
expression in this strain constitutive, due to this protein's ability
to prevent transcription termination. The ability of plasmid-encoded
MG4, MG5, and MG6 to antiterminate transcription and enable
lacZ expression in MA152 was tested by observing the color
of the colonies containing these plasmids on MacConkey lactose plates
and by measuring the
-galactosidase levels produced by the cells
expressing them. As shown in Table II, all three mutants
behaved like wild type in their ability to enable lacZ
expression. These results indicate that all three mutants perform like
wild type and antiterminate transcription at the bgl-lacZ
fusion.
|
After verifying that the three His to Arg mutant proteins do not differ
from wild-type BglG in their activity, we aimed at studying their
ability to be phosphorylated by BglF. Phosphorylation of BglG by BglF
was shown before to be the reason for the negative effect that BglF
exerts on BglG activity (3). Thus, as one approach to determine which
histidine on BglG is phosphorylated by BglF, we tested the effect of
the three missense mutations in BglG on its ability to be negatively
regulated by BglF in vivo. To address this question we made
use of strain MA200 which carries the same chromosomal
bgl-lacZ fusion as MA152, but is also Bgl+
(20). Expression of lacZ in this strain is inducible,
i.e.
-galactosidase is produced only upon the relief of
BglF inhibition by the addition of
-glucosides to the growth medium.
Expression of a plasmid-encoded BglG mutant that cannot be
phosphorylated by BglF renders lacZ expression in this
strain constitutive, i.e. independent of
-glucoside
addition. We tested the effect of plasmid-encoded MG4, MG5, and MG6 on
lacZ expression in MA200 by the same methods employed to
study their effect in MA152, i.e. colonies color on MacConkey lactose plates and
-galactosidase activity measurements. The results of these tests are summarized in Table III.
MG6 behaved like wild-type BglG and did not enable lacZ
expression in the absence of
-glucosides, while MG4 and MG5 led to
constitutive expression of lacZ, independent of
-glucoside addition to the medium. Hence, MG4 and MG5 are not
subjected to negative regulation by BglF in vivo. This
observation suggested to us that the phosphorylated residue on BglG is
either His-208 (mutated in MG4) or His-219 (mutated in MG5) .
|
Another approach we applied to determine which of the histidines in
BglG is phosphorylated by BglF, was to test the ability of the His to
Arg mutants to be phosphorylated by BglF in vitro. We have
shown before that BglF, phosphorylated in vitro in the presence of [32P]PEP, Enzyme I, and HPr, can transfer a
phosphoryl group to BglG (3). We added extracts of cells overproducing
the different BglG derivatives, wild type and mutants, to mixtures
containing prephosphorylated BglF, and further incubated them. Aliquots
removed after 1, 5, and 15 min were analyzed by SDS-polyacrylamide gel electrophoresis followed by autoradiography, and the results are presented in Fig. 4. MG6, mutated in histidine 278, was
phosphorylated in vitro at the same pattern as wild-type
BglG (compare lanes 3-5 with lanes 12-14 in
Fig. 4). No phosphorylation occurred in the case of MG4, which is
mutated in histidine 208 (Fig. 4, lanes 6-8). MG5, mutated
in histidine 219, was phosphorylated by BglF, albeit, at a slightly
reduced rate than the wild-type protein (Fig. 4, lanes
9-11). A control with BglG incubated in a phosphorylation system
lacking BglF (lanes 1 and 2) rules out the
possibility of a direct phosphorylation of BglG by the
phosphotransferase system general proteins. Hence, MG4 is the only His
to Arg mutant that fails to be both in vivo regulated and
in vitro phosphorylated by BglF. These results indicate that
histidine 208 is the amino acid which is phosphorylated on BglG.
The phosphorylated amino acid on the response regulator of the bgl system, BglG, was recently identified as a histidine residue.1 In this study we used a combination of biochemical and genetic approaches to map the phosphorylated histidine on BglG. We first defined the protein region to which the phosphorylation site maps by introducing a protease cleavage site to two locations in the protein, enabling its proteolysis to well defined fragments. Phosphorylation of a protein containing an engineered cleavage site with radioactive phosphate followed by its proteolysis results in the appearance of one radioactively labeled fragment, the one on which the phosphorylated residue resides. Hence, by constructing two mutant BglG proteins (MG1 and MG2), each containing a factor Xa cleavage site at a different location, we could locate the phosphorylated histidine to last 111 residues. This part of the protein contains three histidines, which we mutated, each at a time, to arginine, the residue which is most similar to histidine. To determine which histidine is the phosphorylated one, we followed the mutants' ability to be negatively regulated by BglF in vivo on one hand and to be phosphorylated by BglF in vitro on the other hand. A mutation in histidine 208 prevented both in vivo regulation and in vitro phosphorylation of the mutant protein by BglF. We therefore concluded that histidine 208 constitutes the phosphorylation site on BglG. A mutation in histidine 219, located near the phosphorylation site, enabled phosphorylation of the mutant protein in vitro but affected its ability to be negatively regulated in vivo. The close proximity of this mutation to the phosphorylation site seems to provide a satisfying explanation for this effect.
Ample precautions were taken to ensure that the histidine to arginine
mutants are otherwise identical to the wild-type protein. These
proteins were shown to strongly complement a mutation in the
bglG gene and enable -glucoside utilization. Their
transcriptional antitermination activity was also demonstrated directly
by their ability to express a lacZ gene fused downstream to
the bgl transcriptional terminator. As part of the
precautions, this ability was assayed in two different ways. Passing
all these hurdles requires that the mutant proteins fold, dimerize,
bind RNA, and antiterminate transcription the same as the wild-type
protein. These strict demands for the performance of the mutant
proteins validate the conclusions regarding the phosphorylation
behavior of these mutants.
The method used by us for the localization of the phosphorylated residue on BglG can be applied to map phosphorylation sites on other proteins. Insertion mutagenesis, to introduce a protease cleavage site, or base substitutions, to change residues which are candidates for phosphorylation sites, can be accomplished quickly and efficiently by polymerase chain reaction nowadays. In case the protein of interest is available as a purified polypeptide, mutagenesis of individual suspected residues is not required. Rather, the introduction of a proteolytic site into the protein, which enables sequencing by Edman degradation from internal positions (31), should enable the identification of the phosphorylated residue.
The bgl system represents a new family of bacterial regulatory systems involved in signal transduction (1). Transduction of a signal by this system involves reversible phosphorylation of the BglG response regulator on a histidine residue by the BglF membrane-bound sensor (3, 4).1 Other systems in various organisms were suggested to affiliate to this new family (11-19). The phosphorylation events involved in signal transmission by these systems are not well defined yet. Mapping of the phosphorylation site in BglG is an important step toward the definition and the characterization of the functional domains involved in the communication between sensors and regulators of this novel family of sensory systems. Comparison of the sequence around the BglG phosphorylation site with the respective sequences of its known homologues, e.g. SacY, SacT, and LicT in Bacillus subtilis and BglR in Lactobacillus lactis, reveals that the phosphorylated histidine in BglG, His-208, and the residues flanking it are conserved in all these proteins. It is worth mentioning that the other two histidines that we mutated in BglG, His-219 and His-278, are not conserved in these homologues.
The bgl system was the first example of a reversible phosphorylation modulating a specific and well characterized conformational change in a transcription factor, in this case a change in the oligomeric state of BglG which controls its activity. Whereas non-phosphorylated BglG is a dimer that can bind RNA and antiterminate transcription, phosphorylated BglG is an inactive monomer (6). Mapping and characterizing the nature of both the dimerization and the phosphorylation sites on BglG is important for the elucidation of the mechanism by which phosphorylation affects the dimeric state of a protein. Studies of the BglG dimerization domain are underway.
Mutant derivatives of BglG have been described that do not respond to negative regulation by BglF. These mutants, BglG33 and BglG4, led to constitutive expression of the bgl operon in vivo (20, 32) and showed little or no phosphorylation by BglF in vitro (3). The mutations have been shown to map to Asp-100 and His-160, respectively.3 The identification of histidine 208 as BglG phosphorylation site suggests that these mutations are in other sites involved in BglG-negative regulation and phosphorylation. Whether they are part of the site which is recognized by the kinase and interacts with it or they are part of a phosphorylation module composed of residues remote from each other in the primary sequence but brought to close vicinity during folding awaits future structural studies.
We thank A. Wright for the gift of AE304-7 and AE304-9 strains.