1 4Departamento de Química Física e Instituto de Biotecnología, Facultad de Ciencias ...

32 Pages·2010·2.46 MB·English

JBC Papers in Press. Published on May 24, 2010 as Manuscript M110.110403
The latest version is at http://www.jbc.org/cgi/doi/10.1074/jbc.M110.110403
IDENTIFICATION OF A CHEMORECEPTOR FOR TCA CYCLE INTERMEDIATES:
differential chemotactic response towards receptor ligands

Jesús Lacal1, Carlos Alfonso2, Xianxian Liu3§, Rebecca E. Parales3, Bertrand Morel4, Francisco
Conejero-Lara4, Germán Rivas2, Estrella Duque1, Juan L. Ramos1, and Tino Krell1

1Department of Environmental Protection, Estación Experimental del Zaidín, CSIC, 18008 Granada,
Spain; 2Centro de Investigaciones Biológicas, CSIC, 28040 Madrid, Spain; 3Department of
Microbiology, University of California, Davis, California 95616, USA; 4Departamento de Química
Física e Instituto de Biotecnología, Facultad de Ciencias, Universidad de Granada, 18071
Granada, Spain; §Present address: Department of Microbiology and Environmental Toxicology,
University of California, Santa Cruz, USA
Running head: chemoreceptor signal recognition and transduction
Address correspondence to: Tino Krell, Estación Experimental del Zaidín CSIC, Prof. Albareda 1, 18008
Granada, Spain. [phone] (ext. 294). Fax: [phone]; E-mail: [email protected]

We report the identification of McpS as cause a strong chemotactic response
D
the specific chemoreceptor for 6 TCA cycle (malate, succinate and fumarate) were o
w
intermediates and butyrate in found by Differential Scanning nlo
a
Pseudomonas putida. The analysis of the Calorimetry to increase significantly the d
e
d
bacterial mutant deficient in mcpS and midpoint of protein unfolding (Tm) and fro
m
complementation assays demonstrate that unfolding enthalpy (H) of McpS-LBD. h
McpS is the only chemoreceptor of TCA Equilibrium sedimentation studies show ttp://w
cycle intermediates in the strain under that malate, the chemoattractant which w
w
study. TCA cycle intermediates are causes the strongest chemotactic response, .jb
c
abundantly present in root exudates and stabilizes the dimeric state of McpS-LBD. .org
taxis towards these compounds is proposed In this respect clear parallels exist to the b/
y
to facilitate the access to carbon sources. Tar receptor and other eukaryotic gu
e
s
McpS has an unusually large ligand- receptors which are discussed. t o
n
binding domain (LBD) which is un- A
p
annotated in InterPro and which is Chemotaxis allows motile bacteria to migrate ril 3
predicted to contain 6 helices. The ligand towards or away from different environmental , 2
0
profile of McpS was determined by signals. The major components of the bacterial 19
chemotaxis apparatus include methyl-accepting
isothermal titration calorimetry of purified
chemotaxis receptor proteins (MCPs), the sensor
recombinant LBD (McpS-LBD). McpS
kinase CheA and the response regulator CheY.
recognizes TCA cycle intermediates but
Attractant binding to the chemoreceptor modulates
does not bind very close structural
CheA autophosphorylation activity and the
homologues and derivatives like maleate,
subsequent transphosphorylation of CheY, which
aspartate or tricarballylate. This implies interacts directly with the flagellar motor (1). The
that functional similarity of ligands, such specificity of a chemotactic response is determined
as being part of the same pathway, and not by the MCP, which is typically composed of a
structural similarity is the primary periplasmic ligand-binding domain (LBD) and a
element which has driven the evolution of cytosolic signaling domain. The molecular
receptor specificity. The magnitude of mechanism of chemotaxis has been studied
primarily in Escherichia coli which was shown to
chemotactic responses towards these 7
posses 4 MCPs (2).
chemoattractants, as determined by
The chemotactic behavior of soil and aquatic
qualitative and quantitative chemotaxis
microorganisms is poorly understood. Genome
assays, differed largely. Ligands which
1
Copyright 2010 by The American Society for Biochemistry and Molecular Biology, Inc.
analyses revealed that soil bacteria generally have and dimers is described. Interestingly, in the
a large number of MCPs (3). This is exemplified presence of saturating concentrations of aspartate
by the complete genomes of Pseudomonas and the K of the monomer-dimer equilibrium was at
D
Clostridium strains, which typically have more least 2 orders of magnitude lower, indicating a
than 20 MCPs. It is therefore likely that these significant stabilization of the LBD dimer by
bacteria show chemotactic behavior to an aspartate (23). Central to the understanding of
increased number of compounds, which might signal transduction processes is the question
reflect a major physiological importance. whether the affinity of a signal molecule for its
However, most of these MCPs are of unknown receptor determines the final response. There are
function. To compensate for nutritional shortages, several examples of one- and two-component
many soil microorganisms are able to use TCA regulator systems which show that ligand affinity
cycle intermediates present in plant root exudates for its receptor does not determine the final
or cell debris for growth (4,5). This, combined regulatory output (24,25). Ligand binding was
with the fact that TCA cycle intermediates are proposed to trigger differentially another
abundantly present in plant tissue and root molecular event which in turn determines the
exudates (4,6) might explain why many bacteria final regulatory output (24).
show a chemotactic response towards TCA cycle In this study we have identified a
intermediates (7-13). However, the molecular chemoreceptor, termed McpS that mediates
basis of this taxis remains poorly understood and chemotaxis specifically towards 6 TCA cycle D
o
so far only two receptors have been identified intermediates and butyrate. The LBD of McpS is w
n
which are Tcp of Salmonella typhimurium (9) unusually large and predicted to contain 6 helices. lo
a
d
mediating attraction to citrate and the malate- Using titration calorimetry (26) of purified ed
specific PA2652 of P. aeruginosa (7). protein, the precise thermodynamic profile for the fro
m
The inspection of the SMART database (14) binding of all ligands was determined. All ligands h
reveals that the large majority of MCP LBDs were found to compete in vivo and in vitro for the ttp://w
remains un-annotated. In the case of MCPs with same receptor binding site. The chemotactic w
w
annotated LBDs, these domains belong to the response triggered by these compounds varied .jb
families TarH (15), PAS, GAF (16), CACHE (17) largely. Compounds which triggered a strong c.o
and CHASE (18). However, there is a clear response were also most efficient in increasing brg/
research need to study MCPs with un-annotated the thermal stability of the recombinant LBD. y g
u
LBD. Data are presented which show that malate es
Ligand binding at the LBD causes an stabilizes the LBD dimer of McpS. This was t on
A
alteration of CheA activity bound at the other found to be a feature observed in other p
extreme of the chemoreceptor. MCP function is prokaryotic and eukaryotic receptors. ril 3
thus based on the transmission of the ligand , 20
1
9
mediated molecular stimulus over very large Experimental Procedures
distances (19). Some insight into the molecular
consequences of ligand binding on the Agarose plug assays were carried out as
chemoreceptor has been obtained. These studies previously described (10). Bacteria were grown in
have used primarily the Tar and Tsr receptors, M9 minimal medium supplemented with 15 mM
which have TarH type LBDs that form a 4-helix succinate. Plugs containing chemotaxis buffer
bundle structure (20). Murphy et al. (21) have (KH PO /K HPO , 0.05% (v/v) glycerol, 10 mM
2 4 2 4
studied the influence of serine binding to Tsr. The EDTA, pH 7.0) or 5 mM toluene were used as
authors have demonstrated that serine binding negative and positive controls, respectively.
causes a backbone shift by around 1 Å in the Qualitative capillary assays were carried out as
distance between 1 and 4 of the TarH LBD. described previously (10). Cultures of P. putida
This supports a molecular mechanism by which KT2440RmcpS::Tn5 (pRK415) and
ligand binding causes a piston shift of the final - KT2440RmcpS::Tn5 (pRK415-mcpS) were grown
helix of the LBD (22). in MSB medium containing 5 mM succinate and
Milliagan and Koshland (23) have studied the 20 g/ml tetracycline. Cells were harvested when
effect of aspartate binding to the recombinant the OD600 was between 0.15 and 0.35, washed once
LBD of Tar. In the absence of aspartate a in chemotaxis buffer and resuspended to an OD600
dynamic equilibrium between LBD monomers of 0.1. Capillaries contained 2% low-melting
2
temperature agarose in chemotaxis buffer with or equilibrated in buffer A. Elution was performed in
without added attractant (0.1% casamino acids as a single step with buffer A containing 300 mM
positive control; 5 mM or 50 mM organic acids). imidazol. Protein-containing fractions were pooled,
Quantitative capillary assays were carried out as concentrated to 5 ml, dialyzed against 50 mM
described previously (27). Cultures of P. putida Tris/HCl, 0.5 M NaCl, pH 8.0 and loaded onto a
KT2440R were grown in MSB medium containing HiPrepTM 26/60 SephacrylTM S200 gel filtration
5 mM succinate, harvested when the OD was column (GE Healthcare). Protein was eluted
600
between 0.25 and 0.35, washed once in chemotaxis isocratically (1 ml/min). Coomassie stained SDS-
buffer, and resuspended to an OD of 0.1. PAGE gels containing 30 µg of McpS-LBD did not
600
Bacterial mutants. The 12 mutants of P. putida show any additional bands.
KT2440 each deficient in one mcp gene were Buffer system for biophysical studies. All
obtained from the Granada mutant collection subsequent experiments were carried out in
(www.eez.csic.es/prec/). Mutants were generated polybuffer, pH 6.0. This buffer contains 5 mM
by random mutagenesis of P. putida with mini- Tris, 5 mM MES, 5 mM PIPES adjusted to pH 6.0
Tn5-Km as described (28). Further information with HCl. The study of the pH optimum for
on mutant generation and additional information binding was carried out in polybuffer adjusted to
of these 12 mcp genes are found in Table S1. pH 5.0-8.0.
Complementation of mcpS mutant. The mcpS gene Isothermal titration calorimetry (ITC).
was amplified from P. putida KT2440 genomic Measurements were done on a VP- D
o
DNA using primers 4520-BamH1For (5’- microcalorimeter (MicroCal, Amherst, MA, USA) w
n
GTAGTAGGATCCTGGAGAGCGTGCATGAA at 20ºC. McpS-LBD was dialyzed overnight lo
a
d
CAGC -3’) and 4520-SacIRev (5’- against polybuffer, pH 6. For binding studies, ed
GATGTAGAGCTCCCAGGTTCCAAAGGTCAG protein at 34-38 μM was introduced into the fro
m
ACG -3’), where the restriction sites used for sample cell and titrated with aliquots of ligand h
cloning into pRK415 (29) are underlined. The solution. The ligand solutions were prepared by ttp://w
cloned gene was sequenced to confirm that no PCR dissolving the compounds in dialysis buffer. For w
w
errors had occurred, and pRK415-mcpS was dissociation studies, McpS-LBD at 90 M was .jb
introduced into the mcpS mutant by mating from E. injected into buffer. Data analysis was carried out c.o
coli S17-1 (30). using the “One binding site” model of the brg/
Construction of expression plasmid for McpS-LBD. MicroCal version of ORIGIN, leaving all y g
u
The DNA fragment of mcpS encoding amino acids parameters floating. es
G47-S283 was amplified with primers that contain Analytical ultracentrifugation (AUC). An Optima t on
A
restriction sites for NdeI and BamHI. The resulting XL-I analytical ultracentrifuge (Beckman-Coulter, p
PCR product was then cloned into pET-28b Palo Alto, CA) was used. The detection was ril 3
(Novagen) using the same restriction enzymes that carried out using an UV-visible absorbance , 20
1
gave rise to pETMcpS. The resulting protein detection system. Experiments were conducted at 9
contains an N-terminal His-tag. 20ºC using an AnTi50 rotor and absorbance scans
Overexpression and purification of McpS-LBD. were taken at 230 or 280 nm. Sedimentation
E. coli BL21(DE3) containing pETMcpS was velocity was performed using epon-charcoal
grown at 30ºC until the culture reached an OD standard double sector centerpieces (12-mm
600
of 0.6. IPTG was then added to a final optical path) at a speed of 48,000 rpm.
concentration of 0.1 mM and the culture was Sedimentation coefficient distributions were
maintained at 16°C overnight. Cells were determined by direct linear least-squares boundary
harvested and frozen at -80ºC. Cells derived from a fitting of the sedimentation velocity profiles using
0.5 l culture were resuspended in 50 ml of buffer A SEDFIT software (31). SEDNTERP software (32)
(20 mM Tris/HCl, 0.2 M NaCl, 1 mM EDTA, 10 was used for the correction of s-values to standard
mM β-mercaptoethanol, 10 mM imidazol, 5% conditions (20°C and water). Short-column (85 μl)
glycerol, pH 8.0) containing CompleteTM protease equilibrium runs were carried out at multiple
inhibitor (Roche) and benzonase (Sigma-Aldrich). speeds (12,000, 14,000 and 18,000 rpm) and the
Cells were disrupted using French Press, which corresponding scans were measured at 230 and 280
was followed by centrifugation (45 min at nm. After the equilibrium scans, a high-speed
20,000 ×g). The supernatant was loaded onto a 5 centrifugation run (40,000 rpm) was done to
ml HisTrapHP column (GE Healthcare) previously estimate the corresponding baseline offsets.
3
Whole-cell apparent weight-average buoyant ring formation, indicative of a chemotactic
molecular weight of the protein samples were response, was observed for succinate and toluene,
determined by fitting a single species model to the which are both known to be chemoattractants for
experimental data using either a MATLAB P. putida (10). The screening of the 12 mutants
program based on the conservation of signal revealed that the strain with a knockout in the
algorithm (33) or the Hetero-Analysis program PP4658 open reading frame failed to respond to
(34). The corresponding protein molecular weights succinate (Fig. 1A), while maintaining
were calculated from the experimental buoyant chemotaxis towards toluene. The pp4658 gene is
mass, using 0.725 cm3/g as the protein partial well separated from flanking genes, which do not
specific volume. Several self-association models at encode proteins related to chemotaxis or motility
sedimentation equilibrium were globally fitted to (Fig. S1), and appears to form a single
multiple experimental data using the Hetero- transcription unit. PP4658 was named McpS
Analysis program (34). In parallel, a monomer/m- (succinate).
ner/n-mer self-association scheme was also fitted McpS contains an unusually large sensor
to secondary data (apparent weight-average domain. Analysis of the protein sequence of
molecular weight versus protein concentration) McpS by SMART (14) and DAS (35) allowed us
using a non-linear least-squares procedure with to identify a periplasmic ligand-binding domain
MATLAB scripts kindly provided by Dr. Allen (LBD), which spans 257 residues and is flanked
Minton (NIH). by two transmembrane regions (Fig. 1B). As D
o
Differential scanning calorimetry (DSC). Thermal such, the LBD of McpS is approximately 100 w
n
denaturation experiments were carried out with a amino acids larger than the LBD of the 4 MCPs lo
a
d
VP-DSC, capillary-cell microcalorimeter from of E. coli. Furthermore, these 4 LBDs all belong ed
MicroCal (Northampton, MA, USA) at a scan rate to the TarH family (InterPro signature fro
m
of 120°C/h. A scan rate of 30ºC/h was also used IPR003122) whereas the LBD of McpS remains h
to determine the effect of scan rate on thermal un-annotated in InterPro (36). The three- ttp://w
denaturation. McpS-LBD solutions were prepared dimensional structure of the TarH-LBD shows a w
by dialysis against polybuffer pH 6.0 and each 4-helix up-down-up-down bundle arrangement w.jb
ligand compound was subsequently added to the (20). Consensus secondary structure predictions c.o
protein samples and the buffer. Before each (37) clearly indicate that McpS-LBD consists of brg/
experiment several buffer-buffer baselines were six helices separated by short loops (Fig. 1C). y g
u
obtained to equilibrate the instrument. The Interestingly, of these 6 helices, there are two es
experimental thermograms were baseline- pairs of helices that span 22-23 amino acids, and t on
A
subtracted, corrected from the instrument’s one pair of helices that span approximately 60 p
response and normalized by the protein amino acids. Subsequent studies were aimed at ril 3
concentration. Reheating runs were systematically establishing to what extent this 6-helix domain is , 20
1
carried out to determine the calorimetric present in chemoreceptors of other bacterial 9
reversibility of the denaturation profiles. The species. We selected randomly 100
calorimetric enthalpies were obtained by chemoreceptor sequences from SMART using the
integration of the transition peaks. presence of a methylaccepting domain
(IPR004089) and two transmembrane regions as
RESULTS criteria. Using DAS, the sequence limits of the
LBD domains were identified and subsequently
PP4658 (McpS) is the chemoreceptor for the secondary structure was predicted as detailed
succinate. The chemotactic behavior of wild-type above. Interestingly, from these 100 sequences 11
P. putida KT2440 and 12 mutants deficient in a were found to have a LBD in the range of 240-
single MCP (Table S1) was analyzed using 280 amino acids and a secondary structure
agarose plug assays. This assay involves placing a consisting of 6 helices. Such sensing domains
solidified agarose plug containing were found in different species of the genus
chemoattractant in contact with a cell suspension, Pseudomonas (Uniprot: A4XQ69, Q9HUB1) but
and allowing the formation of rings that are also in Rhodopseudomonas (Q07HJ2),
indicative of a chemotactic response. No Aeromonas hydrophila (uniprot: A0KGQ7),
chemotaxis of P. putida KT2440 (Fig. 1A) was Marinobacter aquaeolei (A1U770), Deinococcus
observed for immobilized chemotaxis buffer, but radiodurans (Q9RYG4), Oceanobacillus
4
iheyensis (Q8EST3) or Bradyrhizobium isocitrate is that they contain a C4-dicarboxy
(A5EAD2). In all cases the LBD of these proteins moiety (including citrate and isocitrate, which are
was found to be un-annotated in SMART. It C2- substituted C4-dicarboxylates). In contrast, 2-
remains to be established whether this domain oxoglutarate, the only TCA cycle intermediate that
corresponds to a novel bacterial sensor domain. did not bind, is a C5-dicarboxylate (Fig. 3). Malate
McpS-LBD interacts directly with succinate. was the tightest binding ligand, followed by
To our knowledge no chemoreceptor that fumarate and oxaloacetate (Fig. 3).
mediates chemotaxis towards succinate has been In order to assess the specificity of McpS-LBD
characterized, and the mode of receptor for the recognition of TCA cycle intermediates,
interaction, either directly or via a binding maleate, itaconate and tricarballylate, close
protein, is unknown. To address this question structural homologues of TCA cycle intermediates
McpS-LBD was overexpressed, purified and (Fig. 3), were analyzed. None of these compounds
subjected to ITC studies (26). This technique showed binding, including maleate, which only
allows the determination of the thermodynamic differs from fumarate in the configuration of the
forces that drive the interaction, as well as double bond, suggesting that McpS-LBD shows
binding constants and stoichiometry. As shown in high binding specificity towards TCA cycle
Fig. 2A titration of buffer with 1 mM succinate intermediates. As shown in Fig. 2B, large heats of
resulted in small and uniform peaks representing binding were observed for fumarate, whereas
heats of dilution. In contrast, titration of McpS- titration signals for maleate were identical to the D
o
LBD with succinate (Fig. 2 A) gave rise to large heats of dilution. w
n
exothermic heat signals due to binding. Data In order to further consolidate this ligand lo
a
d
analysis revealed that binding was driven by profile, a series of physiologically relevant and ed
favorable enthalpy changes which were structurally less-related compounds were analyzed fro
m
counterbalanced by unfavorable entropy changes including various amino acids, sugars and organic h
(Fig. 3). To determine the optimum pH for this acids (Fig. S2). No binding was observed with any ttp://w
interaction, the experiment was repeated in of these compounds, leading to the conclusion that w
w
polybuffer over a pH range of 5 to 8. The highest the in vitro ligand profile of McpS consists of the 7 .jb
bμiMnd).i nTgh aef bfiinnidtyin wg acso nmsetaansutsr eadt paHt p 5H.0 6, .70. 0(K aDn d= 88.20 comApo uMncdps Sh-iLgBhlDig hdteimd eirn Friegc.o 3g.n izes one malate bc.org/
were 250, 330 and 570 M, respectively. All molecule. Subsequent studies were aimed at y g
u
subsequent in vitro experiments with McpS-LBD determining the binding stoichiometry between es
were carried out in polybuffer, at pH 6.0. McpS-LBD and its ligands. To this end it is t on
A
McpS-LBD recognizes most of the TCA cycle necessary to characterize first the oligomeric state p
intermediates and butyrate with high specificity. of the protein which was accomplished by ril 3
ITC was subsequently employed to determine the analytical ultracentrifugation experiments. , 20
1
ligand profile of McpS-LBD. All compounds that Equilibrium sedimentation experiments of McpS- 9
were used are shown in Fig. S2. Since succinate LBD were carried out at protein concentrations
was found to be a ligand, initial experiments were between 0.125 and 10 mg/ml (Fig. 4). Data
done with other dicarboxylic compounds. Titration analysis revealed that McpS-LBD exists as a
of McpS-LBD with 1 mM aspartate, oxalate (C2- mixture of monomers and dimers in solution, with
dicarboxylate), malonate (C3-dicarboxylate), a monomer-dimer association constant of 164 000
glutarate (C5-dicarboxylate) and pimelate (C7- M-1, corresponding to a K of 6.1 M. These
D
dicarboxylate) did not produce any binding heats. results are consistent with sedimentation velocity
It was found, however, that butyrate bound with an studies reported below.
affinity comparable to that of succinate (Fig. 3). ITC can also be used to determine the binding
Binding appeared to be specific to C4-carboxylates stoichiometry in cases were binding curves are
since lactate or valerate (C5-monocarboxylate) did sigmoid. However, curves that so far had been
not bind to McpS-LBD. obtained are hyperbolic (Fig. 2A, B). ITC curve
Next, the remaining TCA cycle intermediates shape depends on the c-value (38), which is a
were studied. Interestingly all TCA cycle function of affinity and ligand concentration. In
intermediates, except 2-oxoglutarate, were found order to generate a sigmoid binding curve,
to bind (Fig. 3). The common structural feature of experiments were repeated with the tightest
fumarate, malate, oxaloacetate, citrate and binding ligand, malate, at a high protein
5
concentration (100 M). In agreement with Chemotaxis toward toluene is mediated by a
equilibrium sedimentation studies, McpS-LBD at different MCP. However, the presence of an
this concentration was exclusively present as equimolar concentration of butyrate and isocitrate
dimer. The resulting sigmoid binding curve is did not alter chemotaxis towards toluene
shown in Fig. 2C. Curve fitting gave rise to an n- indicating that these compounds do not alter cell
value of 0.56, which is consistent with the binding motility. In addition microscopic inspection of
of a single malate molecule to the McpS-LBD cells containing citrate, isocitrate and butyrate did
dimer. not reveal any inhibition of cell motility.
Differential chemotactic response towards the Furthermore, P. putida KT2440 can grow in
7 ligands. The above studies show that McpS- minimal medium supplemented with any of these
LBD binds 7 ligands in vitro. Chemotaxis agarose 3 compounds.
plug assays were carried out with P. putida We hypothesized that there are two groups of
KT2440 and its mcpS mutant to evaluate the ligands: strong attractants (succinate, fumarate,
chemotactic behavior to these 7 ligands. The malate, oxaloacetate) and weak attractants
wild-type strain showed a strong chemotactic (citrate, isocitrate, butyrate), which differ
response towards succinate, malate and fumarate significantly in their chemotactic response. It has
(Fig. S3). This phenotype was characterized by to be noted that the in vitro binding affinity of the
the formation of well-defined rings, which started weak attractants butyrate and citrate for McpS-
to appear after 4 min. A chemotactic response to LBD is comparable to that of the strong attractant D
o
oxaloacetate was also observed. However, the succinate. w
n
intensity of ring formation was weaker and To precisely quantify the differences in the lo
a
d
delayed since ring formation started after around response towards both groups of attractants ed
6 minutes (Fig. S3). Using this technique no quantitative capillary assays were conducted (Fig. fro
m
response was observed for citrate, isocitrate and 6). Confirming the agarose plug assays malate, h
butyrate. fumarate and succinate showed the strongest ttp://w
Subsequently, chemotaxis towards these 7 response. Interestingly, their response was w
w
compounds was studied by qualitative and biphasic with maxima at attractant concentrations .jb
c
quantitative capillary assays which have a higher of 1-10 M and 10-100 mM. At all concentrations .o
rg
sensitivity than the agarose plug assay. The mcpS the response caused by the weak attractants was b/
y
mutant strain was complemented with the mcpS superior to the buffer control but significantly g
u
gene which was present in plasmid pRK415- inferior to the response towards strong attractants. es
t o
mcpS. The chemotactic behavior of this strain was Both groups of attractants compete in vitro n
A
compared to the mutant strain harboring the and in vivo. In order to identify the molecular p
empty plasmid pRK415. In contrast to results reason for the differential response of weak and ril 3
, 2
with the agarose plug assay, responses to malate, strong attractants we wanted to establish whether 0
1
9
succinate, fumarate, oxaloacetate, citrate, both groups of compounds share the same binding
isocitrate and butyrate were detected (Fig. 5). The site at McpS. To this end ITC competition
responses to citrate, butyrate and isocitrate were experiments were carried out which involved the
significantly weaker than those of the other four titration of McpS-strong attractant complexes with
attractants, as a response was only detected with a weak attractants and, vice versa. McpS-LBD
relatively high attractant concentration. The saturated with succinate or malate (2 mM) was
response to isocitrate was particularly weak. The titrated with butyrate and citrate and, vice versa. In
mutant strain carrying the empty vector responded all experiments previous saturation with a ligand
to the positive control and weakly to butyrate, but reduced significantly the binding of the secondary
not to the other tested organic acids. These results ligand. This is illustrated in Fig. 7 which
suggest that McpS is the only chemoreceptor for demonstrates that the presence of malate inhibits
all of the compounds except butyrate, which may citrate binding.
be detected by an additional MCP. Subsequently quantitative capillary assays
To determine whether the weak response were conducted to malate and succinate in the
towards citrate, isocitrate and butyrate is due to a absence and presence of citrate. The competing
nonspecific effect leading to cell paralysis, agent citrate was present in both, the capillary and
agarose plug assays to mixtures of toluene with the bacterial suspension, whereas the attractants
butyrate or isocitrate were carried out (Fig. S4). succinate and malate were only present in the
6
capillary. Bacteria migrate thus in response to a profile than when complexed to weak attractants.
malate and succinate gradient in the absence or The thermal denaturation of these 3 complexes is
presence of a constant, high concentration of characterized by a significant increase in H (87-
citrate. Under these experimental conditions the 106 kcal/mol) and Tm (37.7-42.8 ºC). In addition,
citrate concentration in the immediate there is a correlation between H values and the in
environment of the receptor LBD is likely to be vivo response: weak attractants have significantly
well above the malate or succinate concentration. lower enthalpy changes than strong attractants.
In both cases, the presence of citrate reduced This is further confirmed by oxaloacetate, which
significantly the chemotaxis towards succinate exhibits a weaker chemotactic response than
and malate, which is exemplified for the latter malate, fumarate and succinate and which has a
compound in Fig. 8. These in vitro and in vivo H inferior to these latter compounds but superior
data strongly suggest that weak and strong to the weak attractants.
attractants compete for the same binding site at Malate stabilizes the McpS-LBD dimer. We
McpS-LBD. then hypothesized that ligand binding might cause
Strong attractants increase the thermal a stabilization of the dimeric state of McpS-LBD,
stability of McpS-LBD. To explain the differential which might account for the increase in thermal
effect of strong and weak attractants in vivo, stability as observed by DSC. To address this
McpS-LBD in the absence and presence of the 7 question a series of biophysical experiments were
ligands were analyzed by Differential scanning conducted to investigate the complex between D
o
calorimetry (DSC). In these experiments McpS- McpS with malate. This latter compound caused w
n
LBD is heated at a constant rate and heat changes the strongest chemotactic response in vivo (Fig. 5, loa
d
caused by the thermal denaturation of the protein 6) and had the most pronounced effect on the ed
are recorded. Samples are scanned against a thermal stability of McpS-LBD in vitro (Fig. 9). fro
m
reference that contains all buffer components but ITC can also be used to study self-association h
lacks the protein, which implies that the heats by monitoring heat changes caused by the dilution ttp://w
measured can be exclusively attributed to protein of a concentrated oligomeric protein into buffer, w
w
unfolding. Ligands were added at a concentration where heat changes measured represent the .jb
which guaranteed the complete saturation of transition of a higher to a lower oligomeric state c.o
protein. This technique permits the determination (40). Fig. 10A shows the data for the injection of brg/
y
of the enthalpy change (H, energy necessary to 90 M McpS-LBD into dialysis buffer. According g
u
unfold one mole of protein) and the Tm to the above AUC studies, McpS-LBD at 90 M is est o
(temperature of the maximum of the thermal entirely present as a dimer. A series of exothermic n A
denaturation transition) of thermal unfolding (39). p
The resulting thermograms are shown in Fig. 9 and ptheeank sd iamrei noisbhs eirnv esdiz efo tro threea cinhi ttihael ilnejveeclt ioofn sd iwluhtiiochn ril 3, 2
the derived parameters are listed in Table I. heats. During the course of this experiment the 01
9
Rescans and analyses at different scan rates
protein concentration in the ITC cell is increased
revealed that the thermal denaturation is an
from 0 to 15.5 M. In agreement with the
irreversible process. Thermal denaturation of
equilibrium ultracentrifugation studies (Fig. 4),
protein in the absence of ligand is characterized by
which revealed a monomer-monomer dissociation
a Tm of 36.3ºC and a H of 66 kcal/mol. A group
constant of 6.1 M, the heat changes observed for
of 4 thermograms (Fig. 9) representing McpS-LBD
the injection of concentrated McpS-LBD into
in complex with citrate, isocitrate, butyrate, and
buffer represent the dissociation of protein dimers
oxaloacetate reveal Tm values similar to that of the
into monomers which is an exothermic process.
protein in the absence of ligand (36.1-37.1ºC),
This experiment was repeated in the presence of 2
while increased H values were observed ranging
mM malate, which was added to both, the protein
from 75 to 84 kcal/mol. These compounds are the
and buffer solutions. Under these conditions the
weak attractants, along with oxaloacetate, which
protein is a homogeneous sample entirely saturated
only triggered a modest chemotactic response.
with ligand. Under these conditions the heat
Within this group the latter compound was found
changes were very small and can be almost entirely
to have the highest H. On the other hand, Fig. 9 attributed to heats of dilution (Fig. 10A),
clearly shows that when complexed to the strong suggesting the absence of significant dimer
attractants malate, succinate or fumarate, McpS- dissociation.
LBD has a significantly different denaturation
7
To verify this hypothesis sedimentation velocity intermediates as their sole carbon and energy
ultracentrifugation studies were conducted. McpS- source. Bacteria possessing a complete TCA cycle
LBD at 1 mg/ml in the absence and presence of 2 require only an uptake system for the utilization of
mM of malate was submitted to analysis. The these compounds (41,42) and in addition, several
sedimentation coefficient distribution obtained for anaerobic routes have been described for the
McpS-LBD in the absence of ligand is shown in catabolization of TCA cycle intermediates (43).
Fig. 10B and reveals two species with standard s- This, combined with the fact that TCA cycle
values of 2.7 ± 0.1 S and 3.5 ± 0.2 S, which are intermediates are abundantly present in a variety of
compatible with a globular monomer (frictional natural habitats (4,6), explains why bacteria
ratios f/f = 1.2) and a dimer that slightly deviates frequently exhibit chemotactic behavior towards
0
from globular shape (f/f = 1.5). The relative these compounds (7-13). However, the molecular
0
abundance of the two species depends upon the mechanism of this type of chemotaxis is still
protein concentration in a manner that confirms the poorly understood as so far only a few receptors
existence of a monomer-dimer equilibrium (Fig. have been identified. Those that have been
4). In the presence of malate, a single peak was indentified are the Tcp receptor in S. typhimurium,
observed with an s-value of 3.7 ± 0.2 S. This peak which mediates positive and negative chemotaxis
is likely to present the protein dimer whereas the to citrate and phenol, respectively (9), and PA2652
absence of the peak corresponding to the protein of P. aeruginosa, which mediates chemotaxis to
monomer is observed. malate but not to any of the remaining TCA cycle D
o
To verify whether the binding of malate shifts the intermediates (7). Pseudomonas putida KT2440, w
n
protein monomer-dimer equilibrium towards the the model organism chosen for this study, has a lo
a
d
dimer, sedimentation equilibrium studies were saprophytic lifestyle, is able to efficiently colonize ed
conducted in analogy to data reported in Fig. 4 but roots and seeds (44) and was found to use organic fro
m
in the presence of malate. McpS-LBD at different acids present in root exudates as the primary h
concentrations was incubated with 1 mM malate. carbon source during rhizosphere colonization ttp://w
The analysis of the resulting data (Fig. 10C) (45). w
w
revealed that molecular weights for McpS-LBD By screening the chemotactic behaviour of .jb
did not vary in function of its concentration. At all bacterial mutants deficient in a single c.o
protein concentrations molecular weights were chemoreceptor we have identified the brg/
obtained which are close to the protein dimer. The chemoreceptor for succinate, which we term y g
u
average molecular weight derived from analyses of McpS (Fig. 1). In general, the establishment of es
different McpS-LBD concentrations was of 59 100 the complete chemoreceptor ligand profile by t on
A
± 2500 which is close to the expected size of the analysing single mutants can be misleading since p
protein dimer (57 226). At the lowest protein there are several examples which show that ril 3
concentration (0.25 mg/ml) a molecular weight of bacteria have multiple receptors for the same , 20
1
41 400 ± 200 was determined for McpS in the compound (3). Therefore, the LBD of McpS was 9
absence of malate, whereas a value of 59 700 ± produced as recombinant protein and its ligand
300 was measured in the presence of malate. profile was established by microcalorimetric
Taken together the calorimetric dilution studies, screening of compounds (Fig. 2 and 3). The
the sedimentation velocity experiments and the choice of the recombinant LBD for this series of
sedimentation equilibrium analyses it can be experiments was based on data available on Tar
concluded that the binding of malate stabilizes the which demonstrate that the molecular
dimeric form of McpS-LBD. determinants for the recognition of ligands by Tar
are exclusively present in the LDB. It was shown
DISCUSSION that: first, full-length Tar and its recombinant
LBD bind aspartate with a similar affinity of
A glimpse at a metabolic map of aerobic approximately 3 M (23,46,47); second, both
microorganisms reveals the central role of the TCA proteins recognize ligands with a stoichiometry of
cycle. Apart from its essential role in the 1 ligand per dimer (23,48); third, in both cases
generation of NADH, many biosynthetic routes this binding stoichiometry is caused by strong
branch off from it and many catabolic pathways negative cooperativity between the 2 ligand
funnel products into it. Due to its central metabolic binding sites per dimer (23,48); and fourth, in
role many bacteria are able to use TCA cycle
8
both cases ligand binding causes modest but show that the affinity is largely dependent on the
significant structural alterations (20,23,49). pH, with the optimum at pH 6. Of the 4 efficient
The ligand profile as determined by ITC based ligands that trigger chemotaxis in vivo, three
screening of McpS consists of succinate, fumarate, (malate, fumarate and oxaloacetate) were
malate, oxaloacetate, citrate, isocitrate and recognized by McpS with affinities ranging from 9
butyrate. Subsequently, the chemotactic behavior to 24 M (Fig. 3), which are thus comparable to
of the McpS mutant and its derivative that of serine binding to Tsr. Succinate was
complemented with the mcpS gene was analyzed. recognized with a lower affinity (82 M).
Data show that among the 26 chemoreceptors of P.
putida KT2440 McpS is the sole receptors for Our data reveal that the minimal structural
these 6 TCA cycle intermediates whereas an requirements of a ligand for binding to McpS are at
additional receptor for butyrate might exist (Fig. least 1 carboxylic group present at position 1 of a
5). The fact that there is a sole receptor C4-moiety (citrate and isocitrate can be considered
distinguishes taxis towards TCA cycle as C2-substituted C4-carboxylic acids, Fig. 3). The
intermediates well from other systems where only TCA cycle intermediate which is a C5-
multiple receptors have been identified for the dicarboxylic acid, oxoglutarate, was devoid of
same compound. This is exemplified by P. binding. However, most interestingly, isomers or
aeruginosa, which employs the CtpL and CtpH very close derivatives of the 7 ligands identified
receptors for taxis towards inorganic phosphate were devoid of binding. This is illustrated in Fig. 2 D
o
and the PctA, PctB and PctC chemoreceptors for fumarate, a tightly binding ligand, and its w
n
which have largely overlapping ligand profiles for isomer maleate, which did not bind. Other loa
d
the 20 L-amino acids (3). examples are the absence of binding of aspartate, ed
Apart from the determination of the ligand itaconate or tricarballylate (Fig. 3). Therefore, fro
m
profile, the ITC experiments with McpS-LBD have structural differences between ligands and non- h
established the following features of its action: ligands, for example between fumarate and ttp://w
[1] In general, two basic modes in the interaction maleate, are much smaller than structural w
w
between ligands and chemoreceptors can be differences between ligands, as exemplified by .jb
distinguished: either by direct binding of the fumarate and citrate. It can therefore be concluded c.o
chemoattractant to the receptor, or by the initial that functional similarities, i.e. belonging to the brg/
formation of a complex between the attractant with same metabolic route like the TCA cycle, has y g
u
an auxiliary protein which then binds to the dominated over structural similarities in the es
receptor, as shown for maltose, galactose, and evolution of the molecular recognition of McpS. t on
A
ribose chemotaxis in E. coli (2). Our ITC These findings provide interesting insight into p
experiments demonstrate the direct nature of the protein evolution and are in this case clearly ril 3
interaction of TCA intermediates with McpS, as is related to the evolutionary advantages which arise , 20
1
the case for citrate binding to Tcp (9), whereas the from chemotaxis towards TCA cycle 9
mode of interaction of PA2652 with malate intermediates. Different studies report that TCA
remains to be elucidated (7). cycle intermediates are present at high
[2] A binding stoichiometry of one ligand concentrations in root exudates and dry plant mass
molecule per receptor dimer has been observed for (4,6). P. putida is a saprophyte and was shown to
Tar (48) and Tsr (50). The currently reported ITC efficiently colonize plant roots. The physiological
studies show that the same stoichiometry is also relevance of McpS-mediated taxis is thus likely to
observed for the binding of malate to McpS (Fig. consist in accessing TCA cycle intermediates from
2C). This might indicate that the recognition of a debris of dead plants and from root exudates.
single ligand by a receptor dimer is a mode shared However, the magnitude of the chemotactic
by structurally different chemoreceptors. response towards the 7 attractants varied largely.
[3] The affinities of chemoattractants for receptors Malate, succinate, fumarate and oxaloacetate gave
have been determined for Tar and Tsr. Tar binds a significant response whereas citrate, isocitrate
aspartate with a K of around 3 M (46,47) and butyrate gave a response which was close to
D
whereas the K of serine for Tsr is in the range of the detection limit of the chemotaxis assays
D
10 to 27 M (51,50,52). These studies were all employed. Several studies have determined the
conducted at pH 7.0-7.5 and the pH dependence of organic acid composition of plant root tissue and
ligand binding has not been addressed. Here we plant root exudates (4,6,53-55). Citrate is one of
9
the most abundant organic acids in both samples determinant which defines the magnitude of the
(6,53,54) whereas no significant concentrations of final chemotactic response. The determination of
isocitrate were detected. In this context the very the detailed molecular consequences triggered by
weak chemotaxis towards 50 mM citrate and the binding of strong and weak attractants
isocitrate (Fig. 5) does not appear to be of involves the resolution of the corresponding co-
physiological relevance. Although butyrate is a crystal structures which is an ongoing activity in
product of fermentation of anaerobic bacteria such our laboratory.
as Clostridia (56) and although this compound is a In subsequent studies the impact of the
carbon source for growth of P. putida the weak binding of malate, the strongest chemoattractant,
taxis observed for butyrate does equally not point on the oligomeric state of McpS-LBD was
to a physiological relevance. In summary, amongst investigated (Fig. 10). Based on equilibrium
the 7 ligands of McpS, a physiological relevance in sedimentation ultracentrifugation studies of
the context of taxis towards root exudates and plant protein at different concentrations, a monomer-
debris can only be suggested for the 4 strong dimer equilibrium with a K of 6.1 M was
D
attractants succinate, malate, fumarate and described in the absence of bound ligand (Fig. 4).
oxaloacetate. When these experiments are repeated in the
presence of malate, protein was found to be
Many chemoreceptors are characterized by a exclusively present as dimers at all protein
relatively narrow attractant profile. Tcp of S. concentrations analyzed (Fig. 10C), indicating D
o
typhimurium mediates attraction to only citrate (9) that malate binding has shifted the monomer- w
n
and PA2652 of P. aeruginosa to only malate but dimer equilibrium towards the dimeric state. This lo
a
d
to no other TCA cycle intermediate (7). This conclusion is supported by microcalorimetric ed
contrasts McpS which recognizes 6 TCA cycle protein dilution studies as well as by fro
m
intermediates and butyrate. Tcp and PA2652 have sedimentation velocity ultracentrifugation h
LBDs of the TarH and Cache-2 (IPR013163) analyses (Fig. 10A,B). ttp://w
type, respectively, which are both of In this respect clear parallels exist to the Tar w
w
approximately 150 residues. McpS-LBD in chemoreceptor of E. coli. The LBDs of both .jb
contrast has a size of 257 amino acids, remains receptors differ in size (167 amino acids for Tar c.o
un-annotated in InterPro (36) and shows no as compared to 257 amino acids for McpS) and brg/
significant sequence similarities with the other share no significant sequence identity. The Tar- y g
u
two LBDs. Therefore, at least three different LBD was found to form a 4-helix-bundle es
domains are employed by MCPs to recognize structure (20). Milliagan and Koshland (23) have t on
A
TCA cycle intermediates: the TarH 4-helix studied the effect of aspartate binding on the p
bundle, the Cache domain and the 6-helix recombinant LBD of Tar. In the absence of ril 3
structure predicted for McpS. aspartate, a monomer-dimer equilibrium , 20
1
characterized by a K between 0.5–5M was 9
D
Protein samples in complex with the 7 ligands reported. Interestingly, in the presence of
were submitted to analyses by Differential saturating concentrations of aspartate the K was
D
Scanning Calorimetry. Interestingly, the 3 ligands at least 2 orders of magnitude lower, indicating a
which caused the strongest chemotactic response, significant stabilization of the LBD dimer (23).
malate, succinate and fumarate, induced a The K for the monomer–dimer equilibrium
D
significant increase in Tm, whereas the remaining determined for Tar is thus in agreement with the
ligands had only a marginal impact on the thermal corresponding value of 6.1 M as determined for
stability of McpS-LBD (Fig. 9, Table 1). McpS-LBD. In analogy to Milligan and Koshland
Although the affinity of oxaloacetate for McpS- (23), we establish that malate binding stabilizes
LBD is around three times higher than that for the LBD dimer (23). These observations are
succinate (Fig. 3), the former compound did not consistent with the notion that the stabilization of
significantly increase the protein Tm. The the LBD dimer is a molecular consequence of
capacity of ligands to increase the Tm (Table 1) chemoattractant binding common to receptor
appears to correlate better with the final LBDs which are unrelated at the sequence level.
chemotactic response (Fig. 6) than with the The stabilization of the receptor LBD dimer
binding affinity. This is consistent with the notion as a consequence of ligand binding is also a
that the binding affinity is not the only feature which has been observed for several
10