Comparison of the Pseudomonas aeruginosa and Escherichia coli PhoQ sensor domains: evidence for distinct mechanisms of signal detection.

The PhoP-PhoQ two-component system is present in a number of Gram-negative bacteria where it has roles in Mg(2+) homeostasis and virulence. PhoQ is a transmembrane histidine kinase that activates PhoP-mediated regulation of a set of genes when the extracellular concentration of divalent cations is low. Divalent cations are thought to interact directly with the periplasmic PhoQ sensor domain. The PhoP-PhoQ systems of Escherichia coli and Pseudomonas aeruginosa are similar in their biological response to extracellular divalent cations; however, their sensor domains display little sequence identity. Here we have begun to explore the consequences of this sequence divergence by comparing the biophysical properties of the P. aeruginosa PhoQ sensor domain with the corresponding E. coli sensor domain. Unlike the E. coli protein, the P. aeruginosa PhoQ sensor domain undergoes changes in the circular dichroism and fluorescence spectra as well as destabilization of its dimeric form in response to divalent cations. These results suggest that distinct mechanisms of signal detection are utilized by these proteins. A hybrid protein in which the E. coli sensor domain has been substituted with the corresponding P. aeruginosa sensor domain responds normally to the presence of extracellular divalent cations in vivo in E. coli. Thus, despite apparent differences in the structural response to its stimulus, the P. aeruginosa sensor domain transduces signals to the E. coli PhoQ cytoplasmic kinase domain in a manner that mimics normal E. coli PhoQ function.

The PhoP-PhoQ two-component signal transduction system is required for pathogenesis of Salmonella typhimurium, Salmonella typhi, Yersinia pestis, and Erwinia caratova (1)(2)(3)(4). Its role in S. typhimurium virulence is due at least in part to its function as a regulator of genes responsible for lipopolysaccharide modification that increase bacterial survival in the host organism (5,6). Similar lipopolysaccharide modification also enhances resistance to polymixin by P. aeruginosa (7)(8)(9) and occurs in isolates recovered from cystic fibrosis patients (8).
Additionally, the PhoP-PhoQ system plays a role in the survival of S. typhimurium by limiting Mg 2ϩ concentrations through its regulation of magnesium transporter genes and likely plays a similar role in nonpathogenic Escherichia coli (10).
PhoP-PhoQ possesses the domain structure of the archetypal two-component system (11). The transmembrane PhoQ histidine kinase consists of an extracellular sensor domain coupled to an intracellular kinase domain, and the cytoplasmic PhoP response-regulator is composed of a receiver domain and a transcriptional regulator domain. Extracellular divalent cations such as Mg 2ϩ and Ca 2ϩ have been shown to repress PhoPmediated transcriptional regulation in E. coli, S. typhimurium, and P. aeruginosa (7,9,10,12). Studies in S. typhimurium and E. coli support the idea that these cations are signaling ligands that exert their effects through direct interaction with the PhoQ sensor domain (10,(12)(13). Studies of the S. typhimurium system suggest that ligand binding at the sensor domain activates a phosphatase activity in the intracellular kinase domain that dephosphorylates phospho-PhoP (14). Alignment of the PhoP and PhoQ proteins from the three organisms reveals a significant degree of identity except for the sensor domains, where only 15 of 129 residues are identical ( Fig. 1). This sequence divergence suggests a potential difference in the mechanisms of signal detection. In this paper we show that the P. aeruginosa PhoQ sensor domain responds to divalent cations in a manner that is distinct from that of the E. coli PhoQ sensor domain and is characterized by changes in its oligomeric state and secondary and tertiary structures. Despite this difference, substitution of the E. coli sensor domain with the P. aeruginosa sensor results in a hybrid protein that responds normally to extracellular divalent cations in vivo. Thus, the P. aeruginosa PhoQ sensor domain can regulate the enzymatic output of the E. coli PhoQ kinase domain in a manner that mimics the normal function of the E. coli protein.

EXPERIMENTAL PROCEDURES
Plasmids and Bacterial Strains-PEMQ2a (7) is a pUCP20-derived plasmid that contains the phoP-phoQ operon from P. aeruginosa. pLPQ2 (12) is a pSC101-derived plasmid in which expression of the E. coli phoP-phoQ operon is driven by the lacUV5 promoter. pNL3 (12) is a pBR322-derived reporter plasmid for assaying PhoP-mediated transcriptional activation that contains the phoN promoter fused to lacZ. pNL2 is a low copy number reporter plasmid in which the phoN-lacZ fusion has been subcloned into pGB2 (15). pLPQ*2 is a derivative of pLPQ2 in which DNA encoding the sensor domain of PhoQ has been replaced with the corresponding region of the P. aeruginosa phoQ gene. The hybrid gene consists of codons 1-44 and 189 -486 of E. coli phoQ, and the intervening DNA consists of codons 37-165 of P. aeruginosa phoQ. pAED4PaQ is a derivative of pAED4 (16) used to express the P. aeruginosa PhoQ sensor domain (residues 37-165) for purification. Details of plasmid constructions are available upon request. E. coli strain CSH26⌬Q (12), in which the chromosomal phoQ gene has been inactivated * This work was supported by National Institutes of Health Grant AI41566 (to C. D. W.) with additional funding from the Irma T. Hirschl Trust and an institutional award from the Howard Hughes Medical Institute to the College of Physicians and Surgeons. Acquisition of the analytical ultracentrifuge was made possible by a shared instrumentation grant (RR12848) from the National Institutes of Health. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. by deletion, was used for in vivo activity assays as described (12).
Protein Purification-The P. aeruginosa PhoQ sensor domain (residues 37-165 plus an initiating methionine) was purified from E. coli strain BL21 transformed with plasmid pAED4PaQ. Expression of the PhoQ sensor domain was induced with isopropyl-1-thio-␤-D-galactopyranoside, and the cells were harvested and lysed by French press. Cellular debris was sedimented by centrifugation, and the pellet was resuspended in 20 mM Tris⅐Cl (pH 8.0), 0.2 M NaCl, and 7 M urea. The debris was removed by centrifugation and the supernatant dialyzed against 20 mM Tris⅐Cl (pH 8.0). This solution was loaded onto a MonoQ (Amersham Pharmacia Biotech) anion-exchange column and eluted with a linear gradient from 0 to 1.0 M NaCl in 20 mM Tris⅐Cl (pH8.0). Fractions containing the sensor domain (as assayed by gel electrophore-sis and Coomassie Blue staining) were pooled, dialyzed against sample buffer (20 mM Tris⅐Cl (pH 8.0) ϩ 0.2 M NaCl), and concentrated. This solution was loaded onto a Superdex 75 gel filtration column and eluted with sample buffer. Fractions containing the sensor domain were pooled, concentrated, and dialyzed against sample buffer. The sensor domain was judged to be more than 95% pure by Coomassie Blue staining of a sample following SDS-polyacrylamide gel electrophoresis. Protein concentration was calculated using an extinction coefficient of 11,544 M Ϫ1 cm Ϫ1 at 280 nm based on the presence of 1 tryptophan and 5 tyrosines in the sequence (17). N-terminal sequencing was performed by the Columbia University Protein Facility. The E. coli PhoQ sensor domain was expressed and purified as described (12).
Measurement of Oligomeric State-Sedimentation equilibrium centrifugation experiments were performed at 20°C using a Beckman XLI analytical Ultracentrifuge at rotor speeds of 22,000, 27,000, and 33,000 rpm. For these experiments, sample buffer Ϯ 10 mM CaCl 2 or 10 mM MgCl 2 was used. Data for the sensor domain were collected at protein concentrations of 45, 40, and 18 M in the absence of divalent cations and 150 and 25 M in the presence of divalent cations. The program Match (D. Yphantis, National Analytical Ultracentrifugation Center, Storrs, CT) was used to confirm that equilibrium had been attained, and data for individual channels were extracted and edited using RE-EDIT (Jeff Lary, National Analytical Ultracentrifugation Center, Storrs CT). The data were nonlinear least squares-fitted to single or multiple species models using NONLIN (19). Data fit to a single species model used the function where C (r) and C (ro) are the protein concentrations at a given radial position and a reference position o, respectively, M is the molecular weight of the protein, is the partial specific volume calculated from the amino acid sequence, is the solvent density calculated using the program SEDNTERP (20), is the angular velocity of the rotor, R is the gas constant, T is the absolute temperature, and ␦C is a correction term for the base line. For the single species model, M was allowed to vary  a Average molecular weight (ϮS.D.) calculated from equilibrium centrifugation data recorded at 22,000, 27,000, and 33,000 r.p.m. and globally fit using a single species function (see "Experimental Procedures"). b M a divided by M r (14,770), the monomer molecular weight calculated from the amino acid sequence. and was calculated from the best fit. Data fit to a multiple species model used the function i is the monomer concentration at some reference position to the i th power (i refers to the degree of association), K a1 is the equilibrium constant for the oligomerization of the monomer to an i-mer, is the reduced molecular weight for the i-mer, and is equivalent to (r o 2 )/2 as described above for the single species function.
Measurement of Secondary Structure and Tryptophan Burial-Circular dichroism (CD) measurements were performed in sample buffer with or without divalent salts at 22°C using a Jasco J600 instrument. Divalent cation binding was monitored by changes in CD ellipticity at 222 nm. For these assays, stocks of protein (4 M) in 0 and 10 mM divalent salts were mixed to achieve the proper divalent cation concentration. The apparent equilibrium constant for dissociation of the bound divalent cation was calculated using the equation where f u is the fraction of unbound protein as measured by the fractional change in ellipticity. Fluorescent measurements were performed in sample buffer at 22°C using an Aminco-Bowman AB2 fluorescence spectrophotometer. The sample was excited at 280 nm, and spectra were recorded at 1-nm intervals from 310 to 400 nm. Divalent cation binding experiments were performed as described above, and binding was monitored by changes in the center of spectral mass ( p ), which is defined as ⌺ (f i ⅐ I )/⌺ F i , where f i is fluorescence intensity at wavelength and I is1/ (18).

RESULTS
The PhoP and PhoQ proteins of E. coli and S. typhimurium are 93 and 86% identical, respectively (21). Alignment of these sequences with the PhoP and PhoQ proteins of P. aeruginosa ( Fig. 1) reveals that the three PhoP proteins share 118 identical residues (52.7%), and the PhoQ proteins share 143 identical residues (31.9%). The high degree of identity between the PhoP proteins is seen in both the receiver domain, which is conserved among all members of the two-component family, and the nonconserved regulator domain. Comparison of the PhoQ sequences reveals that the conserved kinase domains possess a high degree of identity (42.9%), whereas the nonconserved sensor domains share only 15 identical residues (11.6%). This observation suggests that the way the P. aeruginosa PhoQ protein senses and responds to extracellular signals may differ from that of the two enteric bacteria. To test this idea, we cloned, expressed, and purified a fragment of P. aeruginosa PhoQ corresponding to the extracellular sensor domain (residues 37-165) and compared its biophysical characteristics to the corresponding E. coli PhoQ sensor domain.
Divalent Cations Destabilize the Dimeric Form of the P. aeruginosa Sensor Domain-The oligomeric form of the P. aeruginosa sensor domain was assayed by sedimentation equilibrium and size-exclusion chromatography. In the absence of divalent cations, the protein sedimented with an average molecular mass that increased from 18.1 kDa at 18 M protein to 22.7 kDa at 45 M (Table I), suggesting that the domain exists as an equilibrium of monomer and dimer species in this concentration range. The data were well fit by a monomer-dimer model but poorly fit by a monomer model (Fig. 2, top panels). The equilibrium constant calculated from a global fit of the data to the monomer-dimer model gave a value of 99 M. Experiments could not be performed at higher protein concentrations because the protein began to undergo nonspecific ag-gregation under these conditions. Because of the obvious aggregation at protein concentrations above 45 M, we were concerned that the increase observed in the average molecular mass as protein concentrations were elevated in the 18 -45 M range was also due to nonspecific aggregation. However, as shown in Fig. 2 (middle panels), when the data were fit to a monomer-N-mer model, the best fit gave an N-value of 2.07, and the data were poorly fit by a monomer-tetramer model. These results argue strongly that the increase in average molecular mass as protein concentration is elevated is primarily because of the formation of a specific dimeric species rather than the formation of nonspecific higher ordered species. In the presence of 10 mM MgCl 2 or 10 mM CaCl 2 the protein sedimented with an average molecular mass that was close to that expected for a monomer (Table I) and were well fit by a monomer model (Fig. 2, bottom panels) at protein concentrations as high as 150 M. In size-exclusion chromatography, the sensor domain elutes as a species with a calculated molecular mass of 36.5 kDa in the absence of divalent cations and 20.7 and 18.8 kDa, respectively, when 10 mM MgCl 2 and 10 mM CaCl 2 were added to the elution buffer (Fig. 3). Although the increase in molecular mass calculated from the elution profile in the absence of divalent cations could be due in part to an increase in the hydrodynamic radius of the monomer, the protein migrated as a species with a lower calculated average molecular mass at lower protein concentrations (31.3 kDa compared with 36.5 kDa), consistent with a concentration-dependent oligomerization event (data not shown). In contrast, the average molecular mass was unchanged at lower protein concentrations when MgCl 2 or CaCl 2 were added to the buffer. The molecular masses calculated from size-exclusion chromatography are somewhat higher than that calculated from sedimentation equilibrium analysis and from that expected from the amino acid sequence, indicating that the sensor domain is likely to possess an extended or elongated structure in both the bound and unbound states.
Secondary Structure and Tryptophan Burial Are Altered in Response to Divalent Cations-The addition of MgCl 2 to the buffer dramatically affects the CD spectrum of the P. aeruginosa sensor domain. There is a significant increase in negative ellipticity in the wavelength range between 208 and 240 nm (Fig. 4a, upper panel). The increase in elliptical intensity at 222 nm predicts the acquisition of roughly 30% ␣-helical secondary structure (22). In contrast, the CD spectrum of the E. coli sensor domain does not change significantly in the presence of MgCl 2 (Fig. 4a, lower panel). Similar results have been described for the S. typhimurium PhoQ sensor domain (13).
The fluorescence spectrum of the P. aeruginosa sensor domain shows a significant increase in intensity and a blue shift of max from 338 to 331 nm (Fig. 4b, upper panel) when MgCl 2 is present. These results indicate that the single tryptophan residue (Trp 123 ) undergoes burial to an environment that is less exposed to solvent (23). Similar changes in the fluorescence spectra are observed when Ca 2ϩ , Zn 2ϩ , Mn 2ϩ , or Ba 2ϩ are added but not salts with monovalent cations (data not shown). The E. coli sensor domain shows little change in its fluorescence spectrum in the presence of MgCl 2 (Fig. 4b, lower panel), but comparisons are difficult to interpret because of the presence of three additional tryptophan residues in the E. coli sensor domain. The fluorescence and CD results suggest that the sensor domain of P. aeruginosa undergoes a ligand-mediated conformational change characterized by a significant increase in secondary structure and tryptophan burial. Alternatively, the isolated domain of the P. aeruginosa protein may exist as a population of folded and unfolded species, and divalent cation binding shifts the equilibrium toward the folded state. A population of E. coli PhoQ sensor domains that have been partially unfolded by urea can be stabilized by the addition of divalent cations, but the stabilization is dependent on a cluster of acidic residues that is not present in the P. aeruginosa protein ( Fig. 1; Ref. 12). In either case the P. aeruginosa PhoQ sensor domain appears to detect divalent cation signals in a manner distinct from that utilized by E. coli PhoQ. Titration experiments were performed in which the increase in CD signal and the center of fluorescence spectral mass were measured, respectively, as a function of added MgCl 2 . As shown in Fig. 5, the fractional changes in ellipticity and center of spectral mass in response to added Mg 2ϩ are monophasic and nearly superimposable, indicating that the increase in secondary structure and tryptophan burial are concurrent events. Moreover, these results indicate the absence of populated intermediates during the transition. If binding and the conformational change are concerted events, these curves are a measure of the affinity of the protein for Mg 2ϩ giving an apparent K d of roughly 300 M [Mg 2ϩ ](f u )/(1 Ϫ f u ).
Pseudomonas PhoQ and a Hybrid P. aeruginosa-E. coli PhoQ Are Functional in E. coli-To examine the ability of the P. aeruginosa PhoQ protein to function in E. coli, a plasmid containing the phoP-phoQ operon from P. aeruginosa was introduced into a strain of E. coli in which the phoQ gene was deleted. PhoQ activity was assayed by expression of a PhoPmediated reporter gene. As shown in Fig. 6a, reporter expression is dependent on the P. aeruginosa phoP-phoQ operon and is regulated in a Mg 2ϩ -dependent fashion, indicating that the P. aeruginosa PhoQ protein does not require a Pseudomonasspecific protein or its native membrane environment for function. The concentration of extracellular Mg 2ϩ at which our reporter is half-maximally repressed (roughly 30 M) is an order of magnitude lower than that calculated above for halfmaximal Mg 2ϩ binding. The local concentration of Mg 2ϩ in the periplasm, where the sensor domain resides, may be significantly higher than outside the cell because of the Donnan potential across the outer membrane (24). Although the results of our biophysical studies indicate significant differences in the structural response of the P. aeruginosa and E. coli PhoQ sensor domains to divalent cations, we were interested to see whether the sensor domain of one protein could substitute for the other. To test this idea, a hybrid gene was constructed in which the resulting protein consists of E. coli PhoQ for all cytoplasmic and transmembrane sequences and P. aeruginosa PhoQ for the sensor domain. As shown in Fig. 6b, the hybrid behaves in a manner similar to that of the wild type E. coli PhoQ protein in that it activates expression of a reporter gene in vivo when the bacteria are grown in Mg 2ϩ -limited media and reporter expression is repressed by extracellular MgCl 2 . These results indicate that the conformational change that the P. aeruginosa sensor domain undergoes in the presence of divalent cations is signaled to the E. coli PhoQ transmembrane and cytoplasmic PhoQ domains to influence the enzymatic output in a manner that mimics normal E. coli PhoQ function. Expression in the presence of the hybrid is roughly 3-fold higher than in the presence of the E. coli protein under repressive conditions. Apparently the conformation of the ligand-bound P. aeruginosa sensor domain is unable to promote formation of the fully repressed form of the kinase domain. DISCUSSION The results presented here show that the P. aeruginosa PhoQ sensor domain differs from that of E. coli and S. typhi-murium in its structural response to divalent cations. The P. aeruginosa PhoQ sensor domain displays an increase in ␣-helical secondary structure, tryptophan burial, and destabilization of its dimeric state. In contrast, the E. coli and S. typhimurium PhoQ sensor domains show no difference in secondary structure or tryptophan burial, and their oligomeric states are monomeric in both the presence and absence of divalent cations (12,13). 1 These differences suggest that distinct mechanisms of signal detection are utilized by the two types of PhoQ proteins and support the idea that the evolution of these different mechanisms is the consequence of extensive sequence divergence of the sensor domain regions, although the reason for these differences is unclear. Perhaps the PhoQ protein of P. aeruginosa adapted to sense additional signals that are required for its existence in a different spectrum of environments. Interestingly, the P. aeruginosa sensor domain does not possess a cluster of acidic residues present in either the E. coli or the S. typhimurium sensor domains (Fig. 1). Conservative substitution of this cluster with noncharged isosteres in the E. coli protein results in a phenotype in which the repressed state of the protein is favored under normally activating Mg 2ϩ -limiting conditions (12). The crystal structure of the E. coli sensor domain reveals that the acidic cluster is in a lobe that is positioned away from the main body of the protein. 2 Apparently this lobe is not present in the P. aeruginosa protein because a gap is introduced at this region by the CLUSTALW sequence alignment program (Fig. 1). The acidic cluster may perform a function in the E. coli and S. typhimurium proteins that required the evolution of a different mechanism of signal detection. The acidic cluster is also absent in PhoQ homologs in Providencia stuartii and Erwinia carotovora. It will be interesting to see whether these proteins respond to their signaling ligands in a manner that resembles that of the P. aeruginosa PhoQ protein.
The modular nature of the histidine kinases suggests that the divergence of the sensor domains could have arisen as the result of a DNA recombination event. Alternatively, the divergence could simply have resulted from successive mutations. We favor the latter possibility based on the following observations. (i) The secondary structure of the P. aeruginosa sensor domain predicted by the program NNSP (25) agrees reasonably well with the secondary structure in the E. coli PhoQ sensor domain crystal structure (Fig. 1). (ii) Gaps and insertions in the P. aeruginosa PhoQ sequence generated in an alignment by the CLUSTALW program are in regions of the E. coli protein that form loops or turns in the crystal structure. (iii) A cluster of 10 amino acids (residues 84 -93 in the P. aeruginosa protein) contains 7 identical residues among the three proteins (Fig. 1), indicative of a conserved region of functional and/or structural importance. The adaptability of two-component systems has been widely exploited in prokaryotes to sense and respond to numerous unrelated signals. Apparently this adaptability can extend to homologous systems with similar functions in different organisms.
What is the nature of the conformational change induced by ligand binding in the P. aeruginosa PhoQ protein, and how is it related to signaling? CD and fluorescence spectroscopy suggest that ligand binding may produce a significant change in the secondary and tertiary structures. The change detected by these methods is similar when assayed at low protein concentrations where the monomeric form predominates and at higher protein concentrations where a significant population of dimers is present (data not shown), indicating that the confor-FIG. 6. In vivo activity of PhoQ proteins. a, E. coli strain CSH26⌬Q carrying the pNL2 reporter plasmid and pEMQ2a (expressing P. aeruginosa PhoQ, filled circles) or pBR322 (negative control, open circles) were assayed for ␤-galactosidase activity following growth in N-medium supplemented with varying amounts of MgCl 2 . b, E. coli strain CSH26⌬Q/FЈlacI Q kan carrying the pNL3 reporter plasmid and pLPQ2 (expressing the wild type E. coli PhoQ protein), pLPQ*2 (expressing the hybrid P. aeruginosa/E. coli hybrid PhoQ protein), or pGB2 (negative control) were assayed for ␤-galactosidase activity following growth in N-medium supplemented with either 10 M MgCl 2 or 10 mM MgCl 2 . mational effects of ligand binding do not require a dimer to monomer transition. Alternatively, the changes in CD and fluorescence spectra elicited by ligand binding may be the result of stabilizing the folded state of a partially unfolded population of sensor domain molecules. Additional studies will be necessary to determine what effect ligand binding has on the spectral properties of the sensor domain in the context of an intact membrane-bound protein.
Presently, we cannot say whether the function of ligand binding is to produce a conformational change that destabilizes the dimeric interface of the sensor domain or whether some other aspect of ligand binding regulates signaling. The His-Asp phospho-relay requires initial autophosphorylation of the sensor kinase followed by transfer of the phosphoryl group to the response regulator. Histidine kinases function as dimers and the autophosphorylation reaction occurs by trans-phosphorylation of one subunit by the other subunit (11). This arrangement suggests a model in which signal transduction occurs via oligomerization of the molecule to form an active dimer. Our results are consistent with a model in which extracellular divalent cations bind to the P. aeruginosa PhoQ sensor domain and decrease the net phosphorylation of PhoP by favoring the formation of inactive PhoQ monomers. Although such a model could explain signaling in the P. aeruginosa PhoP-PhoQ system, it is unlikely to serve as a general mechanism. In some systems, stimuli affect the autokinase activity; however, in other systems the stimulus does not appear to affect the transphosphorylation reaction but instead affects an activity resident in the histidine kinase that de-phosphorylates its cognate response regulator. The latter mechanism has been proposed for the S. typhimurium PhoP-PhoQ system where extracellular Mg 2ϩ has been shown to regulate a PhoQ phosphatase activity (14). Perhaps signal-mediated dimerization plays a role in some systems but not others. Alternatively, destabilization of the dimer interface in the P. aeruginosa PhoQ protein may not correspond to a dimer-to-monomer transition of the intact protein but rather influences the packing of the transmembrane segments to regulate the activity of the cytoplasmic kinase domain within a stable dimer. A third possibility is that the intermolecular interactions that are disrupted to destabilize the sensor domain dimers in these experiments are replaced by other interactions in the ligand-bound intact protein that result in altered interface packing but are not strong enough to hold the dimer of the isolated domain together. Clearly, additional experiments will be necessary to determine how our results with the isolated domain apply to the function of the full-length PhoQ protein. Our results with the hybrid PhoQ protein suggest that information is transduced through the transmembrane segments in a similar fashion for the PhoQ proteins of E. coli and P. aeruginosa. In such a case, it may be that the biophysical differences observed for the response to cations is peripheral to an underlying similar mechanism of signal transduction for both proteins (e.g. ligand-mediated alteration of the dimer interface). However, we cannot rule out the possibility that signaling occurs within a stable dimer for the E. coli protein, but ligand-mediated disruption of intermolecular interactions between sensor domain monomers in the hybrid destabilizes the dimer to favor formation of an inactive monomer. These questions will also require additional biochemical and biophysical experiments with the intact proteins.