Heme oxygenase-1-derived carbon monoxide induces the Mycobacterium tuberculosis dormancy regulon

The mechanisms that allow Mycobacterium tuberculosis (Mtb) to persist in human tissue for decades and to then abruptly cause disease are not clearly understood. Regulatory elements thought to assist Mtb to enter such a state include the heme two-component sensor kinases DosS and DosT and the cognate response regulator DosR. We have demonstrated previously that O(2), nitric oxide (NO), and carbon monoxide (CO) are regulatory ligands of DosS and DosT. Here, we show that in addition to O(2) and NO, CO induces the complete Mtb dormancy (Dos) regulon. Notably, we demonstrate that CO is primarily sensed through DosS to induce the Dos regulon, whereas DosT plays a less prominent role. We also show that Mtb infection of macrophage cells significantly increases the expression, protein levels, and enzymatic activity of heme oxygenase-1 (HO-1, the enzyme that produces CO), in an NO-independent manner. Furthermore, exploiting HO-1(+/+) and HO-1(-/-) bone marrow-derived macrophages, we demonstrate that physiologically relevant levels of CO induce the Dos regulon. Finally, we demonstrate that increased HO-1 mRNA and protein levels are produced in the lungs of Mtb-infected mice. Our data suggest that during infection, O(2), NO, and CO are being sensed concurrently rather than independently via DosS and DosT. We conclude that CO, a previously unrecognized host factor, is a physiologically relevant Mtb signal capable of inducing the Dos regulon, which introduces a new paradigm for understanding the molecular basis of Mtb persistence.     

Blockage of the channel to heme by the E87 side chain in the GAF domain of Mycobacterium tuberculosis DosS confers the unique sensitivity of DosS to oxygen

Two sensor kinases, DosS and DosT, are responsible for recognition of hypoxia in Mycobacterium tuberculosis. Both proteins are structurally similar to each other, but DosS is a redox sensor while DosT binds oxygen. The primary difference between the two proteins is the channel to the heme present in their GAF domains. DosS has a channel that is blocked by E87 while DosT has an open channel. Absorption spectra of DosS mutants with an open channel show that they bind oxygen as DosT does when they are exposed to air, while DosT G85E mutant is oxidized similarly to DosS without formation of an oxy-ferrous form. This suggests that oxygen accessibility to heme is the primary factor governing the oxygen-binding properties of these proteins.     

Protein-protein interactions between histidine kinases and response regulators of <Emphasis Type="Italic">Mycobacterium tuberculosis</Emphasis> H37Rv

Using yeast two-hybrid assay, we investigated protein-protein interactions between all orthologous histidine kinase (HK)/response regulator (RR) pairs of M. tuberculosis H37Rv and identified potential protein-protein interactions between a noncognate HK/RR pair, DosT/NarL. The protein interaction between DosT and NarL was verified by phosphotransfer reaction from DosT to NarL. Furthermore, we found that the DosT and DosS HKs, which share considerable sequence similarities to each other and form a two-component system with the DosR RR, have different cross-interaction capabilities with NarL: DosT interacted with NarL, while DosS did not. The dimerization domains of DosT and DosS were shown to be sufficient to confer specificity for DosR, and the different cross-interaction abilities of DosS and DosT with NarL were demonstrated to be attributable to variations in the amino acid sequences of the α2-helices of their dimerization domains.     

Ligand selectivity of soluble guanylyl cyclase: effect of the hydrogen-bonding tyrosine in the distal heme pocket on binding of oxygen, nitric oxide, and carbon monoxide

Although soluble guanylyl cyclase (sGC) functions in an environment in which O(2), NO, and CO are potential ligands for its heme moiety, the enzyme displays a high affinity for only its physiological ligand, NO, but has a limited affinity for CO and no affinity for O(2). Recent studies of a truncated version of the sGC beta(1)-subunit containing the heme-binding domain (Boon, E. M., Huang, S H., and Marletta, M. A. (2005) Nat. Chem. Biol., 1, 53-59) showed that introduction of the hydrogen-bonding tyrosine into the distal heme pocket changes the ligand specificity of the heme moiety and results in an oxygen-binding sGC. The hypothesis that the absence of hydrogen-bonding residues in the distal heme pocket is sufficient to provide oxygen discrimination by sGC was put forward. We tested this hypothesis in a context of a complete sGC heterodimer containing both the intact alpha(1)- and beta(1)-subunits. We found that the I145Y substitution in the full-length beta-subunit of the sGC heterodimer did not produce an oxygen-binding enzyme. However, this substitution impeded the association of NO and destabilized the NO.heme complex. The tyrosine in the distal heme pocket also impeded both the binding and dissociation of the CO ligand. We propose that the mechanism of oxygen exclusion by sGC not only involves the lack of hydrogen bonding in the distal heme pocket, but also depends on structural elements from other domains of sGC.     

Different Roles of DosS and DosT in the Hypoxic Adaptation of Mycobacteria

The DosS (DevS) and DosT histidine kinases form a two-component system together with the DosR (DevR) response regulator in Mycobacterium tuberculosis. DosS and DosT, which have high sequence similarity to each other over the length of their amino acid sequences, contain two GAF domains (GAF-A and GAF-B) in their N-terminal sensory domains. Complementation tests in conjunction with phylogenetic analysis showed that DevS of Mycobacterium smegmatis is more closely related to DosT than DosS. We also demonstrated in vivo that DosS and DosT of M. tuberculosis play a differential role in hypoxic adaptation. DosT responds to a decrease in oxygen tension more sensitively and strongly than DosS, which might be attributable to their different autooxidation rates. The different responsiveness of DosS and DosT to hypoxia is due to the difference in their GAF-A domains accommodating the hemes. Multiple alignment analysis of the GAF-A domains of mycobacterial DosS (DosT) homologs and subsequent site-directed mutagenesis revealed that just one substitution of E87, D90, H97, L118, or T169 of DosS with the corresponding residue of DosT is sufficient to convert DosS to DosT with regard to the responsiveness to changes in oxygen tension.Oxygen sensing is important for facultative anaerobes to adapt to changes in metabolic necessities during the transition between aerobic and anaerobic conditions. Although Mycobacterium tuberculosis (MTB) is an obligate aerobe, a gradual depletion of O₂ from its culture is known to lead to a drastic change in gene expression (8, 21, 24, 28, 34, 37, 39). Approximately 48 genes of M. tuberculosis were reported to be induced under early hypoxic conditions, which is mediated by the DosSR (DevSR) two-component system (16, 24, 34). The induction of the DosR regulon is important for survival of M. tuberculosis under hypoxic conditions and for it to enter the nonreplicating dormant state (2, 19). The DosSR two-component system consists of the DosS histidine kinase (HK) and its cognate DosR response regulator (RR) (24, 26, 29). The DosT HK, which shares high sequence similarity to DosS over the length of their primary structures, was also found to cross talk with DosR (26, 30). The N-terminal domains of DosS and DosT contain two tandem GAF domains (GAF-A and GAF-B from their N termini), and the three-dimensional structure of the GAF-A and GAF-B domains was determined (5, 25). A b-type heme is embedded in the GAF-A domain, composed of one five-stranded antiparallel β-sheet and four α-helices (5, 14, 25, 32). The heme is positioned nearly perpendicular to the β-sheet, and H149 and H147 of the polypeptides serve as the proximal axial ligands for DosS and DosT, respectively (5, 25). The ligand-binding state at the distal axial position of heme and the redox state of the heme iron modulate the autokinase activity of DosS and DosT. The O₂-bound (oxyferrous) and ferric forms of the HKs are inactive, whereas the unliganded ferrous (deoxyferrous) form as well as NO- and CO-bound forms are active (17, 36). The heme iron of DosT is stable against autooxidation of Fe²⁺ to Fe³⁺ in the presence of O₂, indicating that its conversion between deoxyferrous and oxyferrous forms is the mechanism by which DosT recognizes O₂ (17). However, the autooxidation property of oxyferrous DosS remains controversial. Kumar et al. (17) and Cho et al. (5) reported that DosS undergoes autooxidation on exposure to O₂, while other research groups demonstrated that the oxyferrous form of DosS is stable against autooxidation (13, 14, 36). Recently, different roles of DosS and DosT in O₂ sensing by M. tuberculosis were suggested. DosT plays a more important role in the early phase of hypoxic conditions than DosS when the growth of M. tuberculosis is transferred from aerobic to hypoxic conditions (11).Mycobacterium smegmatis possesses a single DevS HK that phosphorylates the DevR RR (20). The DevSR two-component system is also implemented in hypoxic adaptation of this bacterium (20). Like DosT of M. tuberculosis, the autokinase activity of M. smegmatis DevS was shown to be controlled by the ligand-binding state of its heme (18). Regarding the autooxidation property, DevS of M. smegmatis was suggested to be similar to DosT rather than DosS; i.e., the heme iron in DevS is resistant to autooxidation from an oxyferrous to a ferric state in the presence of O₂ (18).In this paper we report several lines of evidence for the functional difference between DosS and DosT in the hypoxic adaptation of mycobacteria and discuss the implications of these findings.     

DosS responds to a reduced electron transport system to induce the Mycobacterium tuberculosis DosR regulon

The DosR regulon in Mycobacterium tuberculosis is involved in respiration-limiting conditions, its induction is controlled by two histidine kinases, DosS and DosT, and recent experimental evidence indicates DosS senses either molecular oxygen or a redox change. Under aerobic conditions, induction of the DosR regulon by DosS, but not DosT, was observed after the addition of ascorbate, a powerful cytochrome c reductant, demonstrating that DosS responds to a redox signal even in the presence of high oxygen tension. During hypoxic conditions, regulon induction was attenuated by treatment with compounds that occluded electron flow into the menaquinone pool or decreased the size of the menaquinone pool itself. Increased regulon expression during hypoxia was observed when exogenous menaquinone was added, demonstrating that the menaquinone pool is a limiting factor in regulon induction. Taken together, these data demonstrate that a reduced menaquinone pool directly or indirectly triggers induction of the DosR regulon via DosS. Biochemical analysis of menaquinones upon entry into hypoxic/anaerobic conditions demonstrated the disappearance of the unsaturated species and low-level maintenance of the mono-saturated menaquinone. Relative to the unsaturated form, an analog of the saturated form is better able to induce signaling via DosS and rescue inhibition of menaquinone synthesis and is less toxic. The menaquinone pool is central to the electron transport system (ETS) and therefore provides a mechanistic link between the respiratory state of the bacilli and DosS signaling. Although this report demonstrates that DosS responds to a reduced ETS, it does not rule out a role for oxygen in silencing signaling.     

Role of distal arginine in early sensing intermediates in the heme domain of the oxygen sensor FixL

FixL is a bacterial heme-based oxygen sensor, in which release of oxygen from the sensing PAS domain leads to activation of an associated kinase domain. Static structural studies have suggested an important role of the conserved residue arginine 220 in signal transmission at the level of the heme domain. To assess the role of this residue in the dynamics and properties of the initial intermediates in ligand release, we have investigated the effects of R220X (X = I, Q, E, H, or A) mutations in the FixLH heme domain on the dynamics and spectral properties of the heme upon photolysis of O(2), NO, and CO using femtosecond transient absorption spectroscopy. Comparison of transient spectra for CO and NO dissociation with steady-state spectra indicated less strain on the heme in the ligand dissociation species for all mutants compared to the wild type (WT). For CO and NO, the kinetics were similar to those of the wild type, with the exception of (1) a relatively low yield of picosecond NO rebinding to R220A, presumably related to the increase in the free volume of the heme pocket, and (2) substantial pH-dependent picosecond to nanosecond rebinding of CO to R220H, related to formation of a hydrogen bond between CO and histidine 220. Upon excitation of the complex bound with the physiological sensor ligand O(2), a 5-8 ps decay phase and a nondecaying (>4 ns) phase were observed for WT and all mutants. The strong distortion of the spectrum associated with the decay phase in WT is substantially diminished in all mutant proteins, indicating an R220-induced role of the heme in the primary intermediate in signal transmission. Furthermore, the yield of dissociated oxygen after this phase ( approximately 10% in WT) is increased in all mutants, up to almost unity in R220A, indicating a key role of R220 in caging the oxygen near the heme through hydrogen bonding. Molecular dynamics simulations corroborate these findings and suggest motions of O(2) and arginine 220 away from the heme pocket as a second step in the signal pathway on the 50 ps time scale.     

A comparative resonance Raman analysis of heme-binding PAS domains: heme iron coordination structures of the BjFixL, AxPDEA1, EcDos, and MtDos proteins

The heme-PAS is a specialized domain with which a broad class of signal-transducing heme proteins detect physiological heme ligands. Such domains exhibit a wide range of ligand binding parameters, yet they are all expected to feature an alpha-beta heme binding fold and a predominantly hydrophobic heme distal pocket without a distal histidine. We have compared, for the first time, the resonance Raman spectra of several heme-PASs: the heme-binding domains of Bradyrhizobium japonicum FixL, Escherichia coli Dos, Acetobacter xylinum PDEA1, and Methanobacterium thermoautotrophicum Dos. In all cases, the nu(Fe)-(CO) and nu(C-O) values of the carbonmonoxy forms were consistent with coordination of the heme iron to histidine on the proximal side and binding of the CO without electrostatic interaction with the heme distal pocket. EcDos was unusual in having predominantly hexacoordinate heme iron in the deoxy and met forms. Despite an evident lack of CO interaction with the EcDos heme pocket, relatively low Fe-O(2) (562 cm(-1)) and N-O (1576 cm(-1)) stretching frequencies indicated that strong polar interactions with that heme distal pocket are possible for highly bent ligands such as O(2) or NO. None of the newly studied NO adducts exhibited evidence of the Fe-His rupture and pentacoordination previously noted for Sinorhizobium meliloti FixL. A low Fe-His stretching frequency, formerly interpreted as a strained Fe-His bond, and the slow association of O(2) with S. meliloti FixL failed to correlate with the newly studied proteins having low association rate or low equilibrium association constants for binding of O(2). We conclude that although heme-PASs share some features, they represent distinct signal transduction mechanisms.     

Interdomain interactions within the two-component heme-based sensor DevS from Mycobacterium tuberculosis

DevS is the sensor of the DevS-DevR two-component regulatory system of Mycobacterium tuberculosis. This system is thought to be responsible for initiating entrance of this bacterium into the nonreplicating persistent state in response to NO and anaerobiosis. DevS is modular in nature and consists of two N-terminal GAF domains and C-terminal histidine kinase and ATPase domains. The first GAF domain (GAF A) binds heme, and this cofactor is thought to be responsible for sensing environmental stimuli, but the function of the second GAF domain (GAF B) is unknown. Here we report the RR characterization of full-length DevS (FL DevS) as well as truncated proteins consisting of the single GAF A domain (GAF A DevS) and both GAF domains (GAF A/B) in both oxidation states and bound to the exogenous ligands CO, NO, and O2. The results indicate that the GAF B domain increases the specificity with which the distal heme pocket of the GAF A domain interacts with CO and NO as opposed to O2. Specifically, while two comparable populations of CO and NO adducts are observed in GAF A DevS, only one of these two conformers is present in significant concentration in the GAF A/B and FL DevS proteins. In contrast, hydrogen bond interactions at the bound oxygen in the oxy complexes are conserved in all DevS constructs. The comparison of the data obtained with the O2 complexes with those of the CO and NO complexes suggests a model for ligand discrimination which relies on a specific hydrogen-bonding network with bound O2. It also suggests that interactions between the two GAF domains are responsible for transduction of structural changes at the heme domain that accompany ligand binding/dissociation to modulate activity at the kinase domain.     

Mycobacterium tuberculosis senses host-derived carbon monoxide during macrophage infection

Mycobacterium tuberculosis (MTB) expresses a set of genes known as the dormancy regulon in vivo. These genes are expressed in vitro in response to nitric oxide (NO) or hypoxia, conditions used to model MTB persistence in latent infection. Although NO, a macrophage product that inhibits respiration, and hypoxia are likely triggers in vivo, additional cues could activate the dormancy regulon during infection. Here, we show that MTB infection stimulates expression of heme oxygenase (HO-1) by macrophages and that the gaseous product of this enzyme, carbon monoxide (CO), activates expression of the dormancy regulon. Deletion of macrophage HO-1 reduced expression of the dormancy regulon. Furthermore, we show that the MTB DosS/DosT/DosR two-component sensory relay system is required for the response to CO. Together, these findings demonstrate that MTB senses CO during macrophage infection. CO may represent a general cue used by pathogens to sense and adapt to the host environment.     

Ligand discrimination in soluble guanylate cyclase and the H-NOX family of heme sensor proteins

Soluble guanylate cyclases (s GC s) are eukaryotic heme sensor proteins that selectively bind NO in the presence of a large excess of the similar diatomic gas, O(2); this discrimination is essential for NO signaling. Recent discoveries place sGC in the H-NOX (heme nitric oxide and/or oxygen binding domain) family that includes bacterial proteins. The defining characteristic of this family is that some H-NOX proteins tightly bind O(2) whereas others, such as sGC, show no measurable affinity for O(2). A molecular basis for this ligand selectivity has now been established. A distal pocket tyrosine is requisite for O(2) binding and is used to kinetically distinguish between NO and O(2). In the absence of this tyrosine, the O(2) dissociation rate is so fast that the O(2) complex is never formed, whereas the rate of NO dissociation remains essentially unchanged, thus providing discrimination.     

The effect of ligand dynamics on heme electronic transition band III in myoglobin.

Band III is a near-infrared electronic transition at ~13,000 cm(-1) in heme proteins that has been studied extensively as a marker of protein conformational relaxation after photodissociation of the heme-bound ligand. To examine the influence of the heme pocket structure and ligand dynamics on band III, we have studied carbon monoxide recombination in a variety of myoglobin mutants after photolysis at 3 K using Fourier transform infrared temperature-derivative spectroscopy with monitoring in three spectral ranges, (1) band III, the mid-infrared region of (2) the heme-bound CO, and (3) the photodissociated CO. Here we present data on mutant myoglobins V68F and L29W, which both exhibit pronounced ligand movements at low temperature. From spectral and kinetic analyses in the mid-infrared, a small number of photoproduct populations can be distinguished, differing in their distal heme pocket conformations and/or CO locations. We have decomposed band III into its individual photoproduct contributions. Each photoproduct state exhibits a different "kinetic hole-burning" (KHB) effect, a coupling of the activation enthalpy for rebinding to the position of band III. The analysis reveals that the heme pocket structure and the photodissociated CO markedly affect the band III transition. A strong kinetic hole-burning effect results only when the CO ligand resides in the docking site on top of the heme group. Migration of CO away from the heme group leads to an overall blue shift of band III. Consequently, band III can be used as a sensitive tool to study ligand dynamics after photodissociation in heme proteins.     

Resonance Raman studies of cytochrome c' support the binding of NO and CO to opposite sides of the heme: implications for ligand discrimination in heme-based sensors

Resonance Raman (RR) studies have been conducted on Alcaligenes xylosoxidans cytochrome c', a mono-His ligated hemoprotein which reversibly binds NO and CO but not O(2). Recent crystallographic characterization of this protein has revealed the first example of a hemoprotein which can utilize both sides of its heme (distal and proximal) for binding exogenous ligands to its Fe center. The present RR investigation of the Fe coordination and heme pocket environments of ferrous, carbonyl, and nitrosyl forms of cytochrome c' in solution fully supports the structures determined by X-ray crystallography and offers insights into mechanisms of ligand discrimination in heme-based sensors. Ferrous cytochrome c' reacts with CO to form a six-coordinate heme-CO complex, whereas reaction with NO results in cleavage of the proximal linkage to give a five-coordinate heme-NO adduct, despite the relatively high stretching frequency (231 cm(-1)) of the ferrous Fe-N(His) bond. RR spectra of the six-coordinate CO adduct indicate that CO binds to the Fe in a nonpolar environment in line with its location in the hydrophobic distal heme pocket. On the other hand, RR data for the five-coordinate NO adduct suggest a positively polarized environment for the NO ligand, consistent with its binding close to Arg 124 on the opposite (proximal) side of the heme. Parallels between certain physicochemical properties of cytochrome c' and those of heme-based sensor proteins raise the possibility that the latter may also utilize both sides of their hemes to discriminate between NO and CO binding.     

Ligand binding channels reflected in the resonance Raman spectra of cryogenically trapped species of myoglobin

Variations in the v2 region of the Raman spectra of cryogenically trapped photoproducts of different liganded myoglobins as a function of ligand (CO, O2, and n-butyl isocyanide) and species (whale, tuna, elephant) are reported. These variations are attributed to differences in the population of "open" (ligand accessible) and "closed" (ligand inaccessible) conformations of the distal heme pocket. Based on these findings and those derived from other spectroscopies including x-ray crystallography, NMR, IR spectra, and ESR, a working model is presented which accounts for how the conformation of the distal heme pocket, the geometry of the bound ligand, the identity of the ligand, and the dynamics of the dissociated ligand are all interconnected.