Soluble Methane Monooxygenase Component Interactions Monitored by ¹⁹F-NMR

Abstract

Soluble methane monooxygenase is a multi-component metalloenzyme capable of catalyzing the fissure of the C-H bond of methane and insertion of one atom of oxygen from O₂ to yield methanol. Efficient multiple-turnover catalysis occurs only in the presence of all three sMMO protein components: hydroxylase (MMOH), reductase (MMOR), and regulatory protein (MMOB). The complex series of sMMO protein component interactions that regulate the formation and decay of sMMO reaction cycle intermediates is not fully understood. Here, the two tryptophan residues in MMOB and the single tryptophan residue in MMOR are converted to 5-fluorotryptophan (5FW) by expression in defined media containing 5-fluoroindole. In addition, the mechanistically significant N-terminal region of MMOB is ¹⁹F-labeled by reaction of the K15C variant with 3-bromo-1,1,1-trifluoroacetone (BTFA). The 5FW and BTFA modifications cause minimal structural perturbation, allowing detailed studies of the interactions with sMMOH using ¹⁹F-NMR. Resonances from the 275 kDa complexes of sMMOH with 5FW-MMOB and BTFA-K15C-5FW-MMOB are readily detected at 5 μM labeled protein concentration. This approach shows directly that MMOR and MMOB competitively bind to sMMOH with similar K_D values, independent of the oxidation state of the sMMOH diiron cluster. These findings suggest a new model for regulation in which dynamic equilibration of MMOR and MMOB with sMMOH allows transient formation of key reactive complexes that irreversibly pull the reaction cycle forward. The slow kinetics of exchange of the sMMOH:MMOB complex is proposed to prevent MMOR-mediated reductive quenching of the high-valent reaction cycle intermediate Q before it can react with methane.

Graphical Abstract

INTRODUCTION

Soluble methane monooxygenase (sMMO) is a multicomponent metalloenzyme capable of catalyzing the conversion of methane to methanol at ambient temperature and pressure.^{1, 2} The enzyme consists of three protein components: a 245 kDa (αβγ)₂ hydroxylase (sMMOH), a 38 kDa FAD and 2Fe-2S cluster-containing reductase (MMOR), and a 15 kDa cofactorless regulatory component (MMOB).^3-5 As shown in Scheme 1, the sMMOH active site contains a dinuclear iron cluster which serves to activate molecular oxygen for insertion into the C-H bond of methane (bond dissociation energy = 105 kcal/mol).^6-10

Scheme 1.

Reaction Cycle of sMMO and Structure of the sMMOH:MMOB Complex

The resting state of sMMOH contains a diferric cluster (Fe³⁺Fe³⁺_, sMMOH^ox) in which the irons are bridged by two solvent (OH⁻ or H₂O) molecules in addition to the carboxylate of Glu144. sMMOH^ox can form a complex with MMOR and receive two electrons to form the diferrous cluster (Fe²⁺Fe²⁺, sMMOH^red) in which Glu243 shifts to bridge the irons via one carboxylate oxygen, one bridging solvent is lost, and the bond to the second solvent is weakened. In this new configuration, the diiron cluster can bind O₂ between the irons upon dissociation of the weakly bound solvent. However, O₂ binding is observed to be very slow in the absence of the regulatory component MMOB (Structure of the complex shown in Scheme 1).^{11, 12} Binding of MMOB effects a 1,000-fold increase in the rate constant for the O₂ binding to the diiron cluster to form the first spectroscopically distinct intermediate of the reaction cycle, termed P*. ^{7, 11-14} Recent structural studies indicate that one cause of the decreased rate of O₂ binding in the sMMOH active site in the absence of MMOB is the near closure of the molecular tunnel that mediates the transit of O₂ from the solvent.¹² This bottleneck is relieved by conformational changes in both MMOB and sMMOH^red when the sMMOH^red:MMOB complex forms.^{10, 12} A second cause of the low reactivity of O₂ with sMMOH^red is the position of the Glu209 ligand to the diiron cluster, which blocks the approach to the open iron coordination site.¹⁵ An angle change of this residue in the sMMOH^red:MMOB complex exposes the site for O₂ binding.¹⁰ The formation of intermediate P* is followed by spontaneous formation of a peroxo-intermediate P, and finally, O-O bond cleavage to yield the reactive dinuclear Fe⁴⁺ intermediate Q.^{7, 16} Q can react directly with methane to form methanol with incorporation of one atom of oxygen sourced from O₂.

Intermediate Q is generated and stabilized by precisely coordinated sMMO protein component interactions.^{10, 12, 17-22} Although two electrons are required to generate Q, subsequent stabilization of this intermediate requires that further transfer of electrons from MMOR be blocked in order to prevent quenching of the highly electrophilic dinuclear Fe⁴⁺ cluster. Early chemical cross-linking studies of the sMMO components from Methylosinus trichosporium (Mt) OB3b showed that complexes between sMMOH and MMOB or MMOR can readily form.²³ Fluorescence quenching experiments utilizing either endogenous sMMOH tryptophan fluorescence or MMOB labeled with fluorophores demonstrated the formation of a sMMOH^ox:MMOB complex with a K_D value of ~68 nM.^{19, 23} However, measurement of the redox potential of sMMOH showed that a shift of −132 mV occurred upon complex formation with MMOB, indicating that MMOB binds much more weakly to diferrous sMMOH.^{11, 24, 25} While this finding might favor dissociation of MMOB upon reduction of sMMOH, subsequent studies based on site directed mutations in MMOB showed that this was not the case.¹⁸ Remarkably, mutations in different regions of MMOB caused the rate constants for different individual steps in the reaction cycle to significantly change. When two different MMOB variants were added in succession in a single turnover reaction of sMMOH^red, only the step affected by the first MMOB variant added was observed to change. Thus, the first MMOB variant cannot equilibrate with the second in the sMMOH complex during the time required for a single turnover. A non-dissociating MMOB might serve to physically block the ability of MMOR to transfer electrons to Q, provided MMOB and MMOR share a binding site on sMMOH. One conflict with this scenario is that MMOR must have access to resting sMMOH (sMMOH^ox) in order to transfer electrons at the start of the cycle, so MMOB cannot entirely block the reaction at this stage of the reaction cycle. Dissociation of MMOB from sMMOH^ox promoted by MMOR would resolve this problem, but this suggestion seems at odds with the high affinity of MMOB for sMMOH^ox.

Two current models for the regulation of electron transfer in the sMMO system are illustrated in Scheme 2. Model A was motivated by the observation in the chemical cross-linking study referenced above that MMOB and MMOR cross-link to different subunits of sMMOH.²³ In this model, MMOH and MMOB have independent binding sites, and electron transfer at the Q stage of the reaction cycle is blocked by a conformational change at either the MMOR-MMOH or MMOH alpha-beta subunit interface. Model B is based on studies of the protein components isolated from Methylococcus capsulatus (Mc) Bath.^{21, 22} An MMOR construct composed of just the domain containing the 2Fe-2S cluster, termed the Fd domain, was utilized to simplify the experiments. Results from hydrogen–deuterium exchange coupled to mass spectrometry (HDX-MS) and fluorescence anisotropy titration experiments using a 5-((((2-iodoacetyl) amino) ethyl) amino) naphthalene-1-sulfonic acid (IAEDANS)-labeled MMOB D36C variant suggested that the MMOB and Fd-MMOR binding sites overlap on sMMOH. The HDX-MS data showed that the Fd domain interacts with sMMOH surface residues located in the canyon region of the alpha subunit. This is the same location that X-ray crystal structures of the sMMOH:MMOB complex revealed to be the binding site for MMOB for both Mc Bath and Mt OB3b components.^{10, 12, 20} Fluorescence anisotropy titration in Mc Bath sMMO experiments showed that Fd-MMOR can displace MMOB from sMMOH. Furthermore, the Fd-MMOR was more effective at displacing a Δ (2-23) MMOB variant than full length MMOB. ^{21, 22} The researchers hypothesized that the N-terminal region of MMOB serves as an anchor which prevents MMOB from fully dissociating from sMMOH while MMOR is interacting with sMMOH. However, the shift in the MMOB position opens the binding site necessary for MMOR to transfer electrons to sMMOH. Finally, the affinity of MMOB for sMMOH was found to increase rather than decrease upon reduction of the diiron center, suggesting that MMOB might remain bound and block reduced MMOR from rebinding at later points in the reaction cycle.

Scheme 2. Models for Regulation of Electron Transfer to sMMOH.

H^ox = diferric sMMOH, H^red = diferrous sMMOH^red, B = MMOB, R^ox = fully oxidized MMOR, R^red = two electron reduced MMOR^red, H^Q = Intermediate Q of the reaction cycle. The lightning bolt graphic indicates electron transfer from MMOR^red to the diiron cluster of sMMOH.

The studies of the interaction of sMMO components from Mc Bath and Mt OB3b differ in the reported affinity of sMMOH^ox:MMOB (K_D^Mc = 550 nM;²² K_D^Mt = 68 nM^{19, 23}) and sMMOH^ox:MMOR (K_D^Mc = 900 nM;²⁶ K_D^Mt = 10 nM²⁷) complexes and the change in affinities for MMOB upon sMMOH reduction (K_D^Mc = 170 nM; ²² K_D^Mt = 2.2 μM¹⁹). It is possible that part of the disparity in these measurements stems from techniques used to measure the affinities, which often involved major structural modifications in the protein components from attaching large fluorophores. In this study, we have prepared Mt OB3b MMOB and MMOR in which the native tryptophan residues are conservatively changed to 5-fluorotryptophan (5FW).^28-30 Also, an MMOB variant in the N-terminal region is modified with a small ¹⁹F-containing label such that the modified residue mimics the size of the original lysine.³¹ These changes are found not to alter the overall structure or regulatory functions of MMOB or MMOR. They permit the use of ¹⁹F-NMR to study labeled-MMOB complex formation with sMMOH in the presence or absence of MMOR. The sensitivity of the NMR spectrum to changes in the environment and mobility of the tryptophans and N-terminal region fluorine label allows insight into the effects of component concentration and sMMOH oxidation state on complex formation in the absence of large external labels or truncated component variants. The studies support a new model for regulation in which progress through the reaction cycle is promoted by sMMO component exchange that is kinetically coupled to effectively irreversible downstream reactions.

EXPERIMENTAL PROCEDURES

Preparation of Hydroxylase and Reductase.

sMMOH was purified from Methylosinus trichosporium OB3b according to purification protocols described previously.³² A modification of the previous methods for MMOR purification was utilized and is described in Supporting Information.

MMOB Preparation and Biosynthetic Incorporation of 5-Fluorotryptophan.

Plasmids for wild type MMOB and W77F and K15C MMOB variants (See mutagenesis procedures and primer Table S1 in the Supporting Information) were transformed into E. coli BL21(DE3) chemically competent cells. The cells were grown to an optical density of 1.0 (600 nm) and then transferred from Luria Broth (LB) media to a defined medium which included 5-fluoroindole.^33-35 In the absence of the amino acid tryptophan, 5-fluoroindole is biosynthetically transformed into 5-fluorotryptophan by E. coli. IPTG (1 mM final concentration) was added to the defined medium and the BL21(DE3) E. coli cells were allowed to overexpress for 20 h at 20 °C. The cells were harvested by centrifugation. The cell pellet was washed with 25 mM MOPS pH 7 buffer, pooled and frozen at −80 °C. For protein purification, frozen cell pellets were resuspended in 25 mM MOPS pH 7 buffer (5 ml/gram cell paste). The cell suspension was sonicated for 5 min in 1-min intervals (1/2 inch disrupter horn, ~80% power output, ~80% duty cycle). The sonicated solution was centrifuged for 30 min at 46,000 % g. While the heterologous expression of unlabeled MMOB in LB media leads to the predominant presence of MMOB in inclusion bodies, the expression of ¹⁹F-labeled MMOB in a defined media leads to the predominant expression of MMOB in the soluble fraction. Hence, the purification protocol for MMOB was altered. The supernatant was collected and loaded onto a DEAE Q-Sepharose fast flow ion exchange column (3 x 12 cm) equilibrated with 25 mM Tris pH 8 at 4 °C. The column was developed using a 900 ml gradient of 0.05 to 0.25 M NaCl in 25 mM Tris pH 8 buffer at a linear flow rate of 33 cm/h. Fractions containing protein were identified using UV/Vis spectroscopy, and then assessed by SDS-PAGE gel to verify the presence of MMOB. The fractions of interest were pooled and then concentrated to ~25 ml using an Amicon stirred cell equipped with a 10 kDa filter under Ar pressure at 4 °C. The concentrated solution was loaded onto a G75 Sephadex size exclusion column (3 x 115 cm) equilibrated with 25 mM MOPS pH 7 at 4 °C. The protein was eluted from the column by the equilibration buffer flowing at a linear flow rate of 5 cm/h. Fractions were assessed by UV/Vis spectroscopy and SDS-PAGE analysis. Chosen fractions were pooled and concentrated to 25 ml using an Amicon stirred cell equipped with a 10 kDa filter and then further concentrated using a 10 kDa MWCO centrifugal filter. Purity of the final protein solution was assessed by SDS-PAGE. Electrospray mass spectrometry confirmed the stoichiometric incorporation of two fluorine atoms with no detectable residual wild type MMOB (Figure S1). The extinction coefficient used to determine the final concentration of the 5FW-MMOB is 23.2 mM⁻¹cm⁻¹ at 280 nm. Unmodified MMOB has an extinction of 20.8 mM⁻¹cm⁻¹ at 280 nm.

Biosynthetic Incorporation of 5FW into MMOR.

The following two vectors were transformed into BL21(DE3) E. coli cells: i.) an MMOR vector with an ampicillin resistance cassette, and ii.) a pACYC-isc vector with a chloramphenicol resistance cassette. The cells were grown to an optical density of 1.0 (600 nm) and then transferred from LB media to a defined medium which included 5-fluoroindole.³³ IPTG (1 mM final concentration) was added to the defined medium and the BL21(DE3) E. coli cells were allowed to continue to grow for 20 h at 20 °C. The rest of the steps in protein purification are identical to the protocol described in Supporting Information for unlabeled MMOR.

3-Bromo-1,1,1-trifluoroacetone (BTFA) Labeling of K15C-5FW-MMOB and MMOR.

BTFA was used to introduce fluorine into a cysteine residue based on established protocols.^{31, 36-39} 100 μM of purified K15C-5FW-MMOB, 1 mM tris(2-carboxyethyl)phosphine (TCEP), and 400 μM BTFA were added to 100 mM MOPS buffer pH 7, to a final volume of 2.5 ml. The solution was allowed to stir for at least 1 h at 4 °C. The solution was then passed through a PD-10 column to remove excess TCEP and BTFA. The PD-10 eluent containing BTFA-K15C-5FW-MMOB was concentrated using an Amicon Ultra-15 10K MWCO centrifugal filter device. Quantification of free sulfhydryl groups after BTFA modification using Ellman's reagent (5,5'-dithio-bis-[2-nitrobenzoic acid]) showed that >90% of the free cysteine was labeled. Wild type MMOR was modified using the same procedures.

Crystallization of the sMMOH:5FW-MMOB and sMMOH:BTFA-K15C-5FW-MMOB Complexes.

Procedures for crystallization of sMMOH:5FW-MMOB and sMMOH: BTFA-K15C-5FW-MMOB, X-ray crystallography, and crystal structure solution are described in Supporting Information.

Preparation of NMR samples.

All NMR experiments were prepared in 100 mM MOPS buffer, pH 7, containing 10% D₂O (v/v) and less than 0.01% TFA (v/v). TFA was used to calibrate the chemical shifts by setting the value of the TFA resonance to −76.55 ppm. Samples were first prepared in microfuge tubes and then transferred to an NMR tube. Reduced sMMOH samples were prepared in a Coy vinyl anaerobic chamber. The solutions needed for sMMOH reduction experiments were made anaerobic on a Schlenk line using argon gas. The procedure for reducing sMMOH is described elsewhere.³² The NMR tubes were capped with rubber septa and tape to maintain anaerobicity.

The dissociation constants for MMOB binding to diferric and diferrous sMMOH were determined by direct ¹⁹F-NMR monitored titration. Aliquots of diferric or diferrous (anaerobic) sMMOH were added to a constant concentration (5 μM) of ¹⁹F-labeled MMOB in 100 mM MOPS buffer pH 7 at 25 °C while maintaining a constant overall volume in an NMR tube. The intensities of resonances from free ¹⁹F-labeled MMOB decrease as sMMOH is added. The change in intensity was divided by the overall change in intensity at saturation to determine a fraction bound value. The sMMOH added contains a fraction (~30%) that cannot bind MMOB,¹³ but remains as a contribution to the free sMMOH during the titration. This fraction complicates the direct fitting of the data to a binding isotherm. Also, the low K_D of the complex (~80 nM, see Results) is 60-fold below the accessible concentration for NMR detection (~5 μM), further limiting the ability to directly fit the data. As a result, the observed titration is compared to computed curves for given K_D values using Equations 1-3 where free [H] is the total concentration of sMMOH sites not bound to MMOB, H’_Tot is the total concentration of sMMOH sites added that can potentially bind MMOB (~70% of the total sMMOH), H_Inactive is the total concentration of sMMOH that cannot bind MMOB, B_Tot is the total concentration of MMOB, [H’B] is the concentration of the sMMOH:MMOB complex and K_D is the trial dissociation constant for the sMMOH:MMOB complex. The equivalent formulas were used to simulate the titration of labeled MMOR with sMMOH.

$$[{\mathrm{H}}^{\prime }\mathrm{B}]=\frac{{\mathrm{H}}_{\text{Tot}}^{\prime }+{\mathrm{B}}_{\text{Tot}}+K_{\mathrm{D}}-\sqrt{({\mathrm{H}}_{\text{Tot}}^{\prime }+{\mathrm{B}}_{\text{Tot}}+K_{\mathrm{D}}{)}^2-4({\mathrm{H}}_{\text{Tot}}^{\prime }•{\mathrm{B}}_{\text{Tot}})}}{2}$$ Eq. 1

$$\text{Free}[\mathrm{H}]=({\mathrm{H}}_{\text{Tot}}^{\prime }-[{\mathrm{H}}^{\prime }\mathrm{B}])+{\mathrm{H}}_{\text{Inactive}}$$ Eq. 2

$$\text{Fraction bound}=\frac{[{\mathrm{H}}^{\prime }\mathrm{B}]}{{\mathrm{B}}_{\text{Tot}}}$$ Eq. 3

The fraction of active sMMOH that can bind MMOB was determined for each experiment based on the amount of intermediate Q formed in a single turnover stopped-flow experiment. The yield of intermediate Q to bound MMOB ratio has been established by Mössbauer spectroscopy.^{13, 40}

The K_D value for MMOR binding to sMMOH was determined by competition with MMOB. The approximately stoichiometric ¹⁹F-labeled MMOB (12 μM) complex with sMMOH (18 μM active sites of which ~ 12.6 μM can form the MMOB complex) was formed using either diferric or diferrous (anaerobic) sMMOH. Aliquots of MMOR were added and the increase in the NMR resonances from free ¹⁹F-labeled MMOB quantified. The fraction of labeled MMOB released was determined as the ratio of the increase in intensity divided by the difference between the intensities from the starting complex and free labeled MMOB. The resulting curve was fit by non-linear regression to a regular hyperbola. The half saturation MMOR concentration value and the K_D values for the diferric or diferrous sMMOH:¹⁹F-labeled-MMOB complexes were used with the online calculator BotDB to determine the approximate K_D for MMOR binding to sMMOH.^{41, 42}

¹⁹F-NMR spectroscopy.

Acquisition of spectra was performed using a Bruker 600-MHz Avance NEO equipped with a 5-mm triple resonance cryoprobe. Routine 1D-¹⁹F-NMR spectrum were acquired utilizing the zg pulse program (Bruker TOPSPIN Version 4.0.3). Acquisition parameters for the NMR experiments were: spectrometer frequency = 565.1 MHz, temperature = 298.0 K, number of scans = 512, acquisition = 0.2 s, relaxation delay = 2.0 s, pulse width 12 μs, and receiver gain = 101.0 unless otherwise noted. The raw data was processed using MestReNova version 14.1.2, released 03/23/2020. Processing consisted of the following steps: i.) import data, ii.) automatic phase correction followed by manual 0 and 1^st order corrections, iii.) baseline correction along F1 using either polynomial fit or splines methodology, and iv.) line broadening of 30 to 90 Hz using an exponential function.

¹⁹F chemical exchange saturation transfer (¹⁹F-CEST) experiments were conducted by first obtaining a 1D ¹⁹F-NMR spectrum of the sample. Next, the frequency values to be irradiated in the ¹⁹F-CEST experiment were chosen using the acquired 1D spectrum and saved as separate irradiation frequency lists containing three frequencies each. All irradiation frequency lists created contained at least one off-resonance frequency as a control. The stddiff pulse program was used to collect ¹⁹F-CEST data and the acquisition settings were: number of scans = 512, acquisition time = 0.2 s, Relaxation time = 4 s, saturation time = 4 s, Gaussian pulse width = 50 ms, F2 channel-shaped pulse for saturation = 50 dB, File name for SP9 = Gaus1.1000, number of irradiation frequencies = 3, and number of FIDs = 3.

Stopped-Flow Procedures.

Procedures for stopped-flow measurements of the rate constants for steps in a single-turnover of the reaction cycle are described in Supporting Information.

RESULTS

¹⁹F-Labeling of MMOB.

Past studies designed to characterize the sMMOH:MMOB complex in response to sMMOH oxidation state or the presence of MMOR have relied on changes in endogenous tryptophan fluorescence or spectral response from fluorescent or spin probes chemically added to cysteine residues introduced into MMOB.^{19, 21-23, 26, 27} However, the sensitivity and size of the introduced probes have limited their utility, especially when placed in the sMMOH:MMOB interface, where mutagenesis studies have shown precise interaction to be essential.^{18, 43} Inspection of the X-ray crystal structure of the Mt OB3b sMMOH:MMOB complex suggested an alternative approach, because the only two Trp residues of MMOB (W76 and W77) were found to be in or near the sMMOH:MMOB interface in the compactly folded core region of the protein.^{10, 12, 20} Replacement of these two Trp residues with 5-fluorotryptophan by growth of the MMOB overexpression strain on media containing 5-fluoroindole provide sensitive ¹⁹F-NMR probes, resulting in well-resolved resonances of a folded protein (Figure 1, inset).²⁸ The 5FW resonances were assigned to specific Trp residues by mutating W77 to Phe as shown in Figure 2.

Figure 1.

One dimensional ¹⁹F-NMR spectra of ¹⁹F-labeled MMOB. Main panel: Spectrum of BTFA-K15C-5FW-MMOB. The spectral regions for the resonances from the BTFA and 5FW labels are marked. Inset: Spectrum of 5FW-MMOB not containing the K15C mutation.

Figure 2.

5-Fluorotryptophan resonance assignment. Protein mutagenesis and overexpression of W77F-5FW-MMOB allowed a definitive assignment of the 5FW resonances.

A second objective of this study was to determine whether the N-terminal tail of MMOB responds differently than the core region to changes in sMMOH diiron cluster redox state or to the binding of MMOR to the sMMOH:MMOB complex. To address these questions, a BTFA fluorine probe was introduced in the N-terminal region by chemically modifying the K15C MMOB variant. This variant has been shown in previous studies not to cause large changes in sMMO catalysis.^{10, 12, 19} Post translational modification of the K15C-5FW-MMOB with BTFA was confirmed by 1D-¹⁹F-NMR (Figure 1). The BTFA resonance at −85.4 ppm is more intense than the resonances from the labeled Trp residues because of the three ¹⁹F atoms in the probe and its increased mobility, leading to a narrow linewidth. It is important to note that the 5FW resonances upfield of the BTFA-K15C resonance are unchanged (compare Figure 1 and Figure 1, inset, and see other examples below).

Impact of Incorporation of 5-Fluorotryptophan and BTFA on sMMO Steady State and Single-Turnover Kinetics.

Both 5FW-MMOB and BTFA-K15C-5FW-MMOB allowed efficient turnover of methane by the reconstituted sMMO system (Figure 3). Past observations have shown that the initial velocity of oxygen uptake or NADH utilization increases to a maximum when the MMOB and sMMOH active sites concentrations are approximately equal, followed by a decline in initial velocity.⁴ This result was interpreted to indicate that a complex of one MMOB with each α-subunit of sMMOH is necessary for rapid turnover. Moreover, this complex apparently forms nearly stoichiometrically even at sub-micromolar concentrations of the components used in the experiment. The same concentration dependence is observed when 5FW-MMOB is used in place of wild-type MMOB (Figure 3). However, when BTFA-K15C-5FW-MMOB is used, an approximate 2-fold excess over sMMOH active sites is required to maximize the initial velocity, suggesting a somewhat lower affinity. The same maximum initial velocity is reached using each of the MMOBs, suggesting that the fluorine labeling does not interfere with function or the rate limiting step in the reaction cycle.

Figure 3.

Steady state kinetics of sMMO turnover. An oxygen electrode was used to monitor the velocity of O₂ consumption for the reaction of 0.4 μM sMMOH (active sites), 1.2 μM MMOR and the indicated concentration ratio of either wild type MMOB (black) or 5FW-MMOB (red) or BTFA-K15C-5FW-MMOB (green). The other reaction components were 200 μM methane, 250 μM O₂, 400 μM NADH in 25 mM MOPS buffer pH 7.5 at 23 °C. Past studies have shown that the initial velocity maximizes at approximately stoichiometric sMMOH-MMOB complex formation and then declines.⁴

The potential for perturbation of the individual steps in the sMMO reaction cycle kinetics was evaluated by conducting single-turnover transient kinetics studies. Anaerobic, stoichiometrically reduced sMMOH was rapidly mixed with either MMOB, 5FW-MMOB, or BTFA-K15C-5FW-MMOB in a buffer containing a large excess of O₂ using a stopped-flow device (see Experimental Procedures). These pseudo-first order conditions allowed the rate constants for formation and decay of the intermediates to be obtained by multiple-summed exponential fitting of the time courses (Figure S2, Table 1).^{7, 13, 44} The rate constants for P* to P, P to Q and Q decay steps in the reaction cycles when using wild type MMOB and 5FW-MMOB showed either no or minor perturbation. Use of the BTFA-K15C-5FW-MMOB caused decreases in the rate constants for the P* to P and P to Q steps, while leaving the Q decay step unchanged. However, this ¹⁹F-labeled MMOB remained highly functional as reaction cycle regulator.

Table 1.

Rate Constants for Single Turnover Reactions

	MMOB species utilized in reaction
Rate constants (s−1)	MMOB	5FW-MMOB	BTFA-K15C-5FW-MMOB
kP formation	6.6 ± 0.6	5.9 ± 0.6	2.6 ± 0.8
kP decay/Q formation	2.4 ± 0.3	1.4 ± 0.3	0.5 ± 0.1
kQ decay	0.04 ± 0.01	0.04 ± 0.01	0.04 ± 0.01

Molecular Structures of the sMMOH:5FW-MMOB and sMMOH:BTFA-K15C-5FW-MMOB Complexes.

To assess whether incorporation of 5FW or BTFA into MMOB perturbs the structure of the sMMOH:MMOB complex, both the sMMOH:5FW-MMOB and sMMOH:BTFA-K15C-5FW-MMOB complexes were crystallized in the oxidized states of sMMOH and their structures solved at 2.08 Å (PDBID:7M8Q) and 2.22 Å (PDBID:7M8R), respectively (Table S2). In both crystals, the native (αβγ)₂ sMMOH and two bound MMOBs occupy the asymmetric unit. The structures show that the sidechain of 5FW77 points towards the sMMOH interface while the sidechain of 5FW76 points away (Figure 4A and B, see Figure S3 for a comparison of the positions of the tryptophans in free MMOB vs. sMMOH:5FW-MMOB complex). Additionally, all of the polar contacts that each 5FW makes are the same as those made by the respective tryptophans in the sMMOH complex with unlabeled MMOB. Structural alignment of the protein backbones of 7M8Q or 7M8R with that of the sMMOH:MMOB complex (PDBID: 6VK5) yields RMSD values of 0.097 Å and 0.149 Å, respectively, indicating that the structures are nearly identical. The overall structure of the sMMOH:MMOB complex and the conformational changes induced by MMOB have been discussed in detail elsewhere, and thus they are not described further here.^{10, 12} It is important to note that conservation of structure upon substitution of 5FW for tryptophan has been commonly observed in other proteins.^{30, 45-47}

Figure 4.

Structure of fluorine-labeled tryptophan residues in sMMOH:5FW-MMOB. A: Structure of the sMMOH:5FW-MMOB complex (PDB:7M8Q) showing the interface region containing 5FW76 and 5FW77. 5FW77 is buried in the interface whereas 5FW76 is partially solvent exposed. B: Detailed view of the interface region with electron density shown at 1σ.

To the best of our knowledge, the x-ray crystal structure of sMMOH:BTFA-K15C-5FW-MMOB is the first to show BTFA in a protein-protein complex. The BTFA-K15C probe is located on the N-terminal tail of MMOB and is solvent exposed (Figure 5 A and B) A polar contact between MMOB residues K15 and E23 is lost as a result of the mutation to cysteine and addition of the BTFA probe. Comparison of the two MMOBs in the asymmetric unit reveals different rotomeric conformations of the BTFA probe (Figure 5C). Also, the electron density of the three ¹⁹F atoms on BTFA is somewhat less than those of 5FW76 and 5FW77, likely due to the increased mobility of BTFA in the complex.

Figure 5.

A: Structure of the sMMOH:BTFA-K15C-5FW-MMOB complex (PDB:7M8R) interface region showing the relative position of the BTFA and 5FW ¹⁹F-labels. B: Detailed view of the BTFA probe with electron density shown at 1σ. C: Superimposed views of the BTFA probe in the two labeled MMOBs present in the asymmetric unit (Chains D and H) revealing different rotomeric conformations. D: Superposition of structures of sMMOH:MMOB (PDB:6VK5) and sMMOH:BTFA-K15C-5FW-MMOB showing relative size of BTFA-K15C and the native K15.

Complex of ¹⁹F-Labeled MMOB with diferric or diferrous sMMOH.

One-dimensional ¹⁹F-NMR experiments were performed to investigate the interaction between oxidized or reduced sMMOH and ¹⁹F-labeled MMOB. Upon addition of diferric sMMOH to¹⁹F-labeled MMOB, resonances from both 5FW and BTFA decrease sharply in intensity and do not shift or broaden. Importantly, two new resonances emerge downfield of the resonance from unbound BTFA-K15C-5FW-MMOB and one broad resonance emerges downfield of the two resonances from 5FW in unbound 5FW-MMOB or BTFA-K15C-5FW-MMOB (Figure 6A). Addition of diferrous sMMOH also elicits two resonances downfield of the resonance from unbound BTFA-K15C-5FW-MMOB, but they occur at different chemical shifts than observed after addition of diferric sMMOH (Figure 6B). The same broad resonance is observed downfield of the two resonances from 5FW in unbound 5FW-MMOB or BTFA-K15C-5FW-MMOB.

Figure 6.

One dimensional ¹⁹F-NMR spectra of sMMOH (150 μM active sites, ~105 μM available to bind MMOB) complexed with BTFA-K15C-5FW-MMOB (125 μM). Panel A: BTFA and 5FW spectral regions are shown on the left and right, respectively. New resonances at −84.5 ppm, −85.1 ppm and −121.5 ppm are labeled in addition to the resonances from residual unbound MMOB. Panel B: Effect of the oxidation state of sMMOH on the spectrum of the BTFA-label in the sMMOH:BTFA-K15C-5FW-MMOB complex.

Resonance Assignment of the 275 kDa sMMOH:BTFA-K15C-5FW-MMOB Complex.

The origin of the new resonances that arise from sMMOH:MMOB complex formation were probed using ¹⁹F-Chemical Exchange Saturation Transfer (¹⁹F-CEST).^48-51 In this experiment, weak selective Gaussian pulses are applied in small steps across the regions of interest in the ¹⁹F-NMR spectra of a sample containing sMMOH and BTFA-K15C-5FW-MMOB. This pulse is then followed by a hard non-selective 90° pulse. If the weak Gaussian pulse is applied to a region of the ¹⁹F-spectrum away from frequencies corresponding to either the bound or unbound state, no change in intensity will be observed in either set of resonances. However, if the weak Gaussian pulse is applied to a region near or directly on frequencies corresponding to the bound or unbound state, a decrease in intensity of resonances from both states will be observed following the strong pulse as a result of chemical exchange.

As shown in Figure 7A, when the weak Gaussian pulse is applied at the position of either of the resonances downfield of the intense resonance at −85.4 ppm originating from BTFA on the N-terminal tail of unbound MMOB, the intensity monitored at the −85.4 ppm resonance is decreased. A complementary result is found when monitoring using either of the downfield resonance intensities after the weak pulse is applied to the −85.4 ppm resonance (Figure 7B and C). These results show that the −85.4 ppm and downfield resonances are in chemical exchange and supports the conclusion that the downfield resonances arise from the sMMOH:BTFA-K15C-5FW-MMOB complex. Moreover, because both downfield resonances decrease upon application of the weak pulse at −85.4 ppm or when the weak pulse is applied to either downfield resonance, there are two separate interconverting conformational states of the complex in slow exchange.

Figure 7.

¹⁹F-NMR Chemical Exchange Saturation Transfer of sMMOH:BTFA-K15C-5FW-MMOB complex. A selective Gaussian pulse was applied to the ¹⁹F spectra at locations indicated by purple boxes. The black line indicates the loss of intensity of the resonance indicated by the labeled frequency and red arrow. If slow exchange is occurring between bound and free states, the black line should mirror the spectrum shown in red. A-C: BTFA spectral region. D-F: 5FW spectral region.

The spectral region near the resonances from 5FW in the MMOB core region was similarly probed (Figure 7, right). The results support the assignment of the broad −121.5 ppm resonance as arising from the sMMOH:BTFA-K15C-5FW-MMOB complex. However, when the weak pulse was applied at −121.5 ppm, a decrease in intensity was observed only when monitoring the 5FW77 signal at −126.6 ppm (Figure 7D and E). The complementary result was observed when the −126.6 ppm resonance was irradiated and the signal intensity monitored at −121.5 ppm (Figure 7F). Thus, it appears that the −121.5 ppm resonance reports on the chemical environment of only 5FW77 while bound to sMMOH, in accord with its placement in the sMMOH-MMOB interface revealed by the structure (Figures 4A and B and S3). This result is confirmed by the lack of a resonance at −121.5 ppm from the sMMOH:5FW-W77F-MMOB complex (Figure S4).

5FW-Labeled MMOR.

The primary structure of MMOR contains one tryptophan residue in the FAD domain. MMOR W107 was labeled with 5FW in the same manner as MMOB. The 1D ¹⁹F-NMR spectrum of heterologously expressed 5FW-MMOR is shown in Figure 8A. Two resonances of quite different intensity, but similar widths are observed in the spectrum. Treatment with TCEP failed to eliminate either resonance, so dimerization via a disulfide linkage is not responsible for the two environments observed for a single Trp. Another possibility is that there is a single form of the enzyme, but two conformations exist in slow exchange. To test this hypothesis, a ¹⁹F-CEST experiment was performed (Figure 9). It was observed that irradiation of either resonance resulted in very little decrease in intensity of the other resonance. Consequently, it seems likely that there are major and minor forms of the enzyme that do not interconvert on the NMR timescale.

Figure 8.

¹⁹F-NMR Spectra of 5FW-MMOR. A: Spectrum of 20 μM 5FW-MMOR in the absence of sMMOH (5000 scans). B: Spectrum of 20 μM 5FW-MMOR (red, 512 scans) after adding the ratio of sMMOH (active sites) shown. The minor −126.1 ppm feature is not resolved due to the decreased number of scans, but in other experiments was found to also disappear after addition of stoichiometric sMMOH.

Figure 9.

5FW-MMOR CEST. Panel A shows the 1D ¹⁹F-NMR CEST spectrum when a Gaussian pulse is applied to a region of the spectrum away from the 5FW-MMOR resonances. Panel B shows the effects when irradiated at −125.1 ppm. Panel C shows the effects when irradiated at −126.1 ppm. The yellow arrow indicates the frequency of the Gaussian pulse used.

Addition of stoichiometric sMMOH (active sites) resulted in elimination of the spectrum from MMOR (Figure 8B, blue). No new resonances are observed. It is likely that the resonances from 5FW are broadened into the baseline by formation of a tight sMMOH:MMOR complex. Titration of 5FW-MMOR with sMMOH gives a linear decrease in the −125.1 ppm resonance demonstrating high affinity (Figures 8B and S5). The concentration of 5FW-MMOR required for reasonable S/N (20 μM) is too high for accurate determination of the K_D value for this tight complex. Another approach to determination of this K_D is described below.

MMOR was also labeled with BTFA in the same manner as described for MMOB. However, the presence of nine cysteine residues led to incorporation of multiple labels and a ¹⁹F-NMR spectrum that was too complex to allow definitive studies.

Binding Affinity of sMMOH for ¹⁹F-Labeled MMOB.

The sharp and intense resonances from the relatively low molecular weight 5FW-MMOB and BTFA-K15C-5FW-MMOB allow quantification of a titration with sMMOH (examples are shown in Figures 10 and S6). The affinity for sMMOH is very high, and thus, the minimum concentration of ¹⁹F-labeled MMOB with acceptable S/N is higher than optimal for a precise K_D determination. Nevertheless, the titration plots allow boundaries to be set for K_D values for both diferric and diferrous sMMOH as shown in Figure 11. The dashed curves shown in Figure 11 are not fits of the data, but rather computed binding isotherms under the assumption that 70 % of the sMMOH is capable of binding to the ¹⁹F-labeled MMOB (see below and Experimental Procedures). The good match between the computed curve and the titration of 5FW-MMOB suggests a K_D value of roughly 80 nM for both diferric and diferrous sMMOH (Figure 11A). The relatively intense −85.4 ppm resonance of BTFA allows a more accurate titration of BTFA-K15C-5FW-MMOB by sMMOH. BTFA-K15C-5FW-MMOB show a high affinity (K_D ~ 80 nM) for diferric sMMOH, but a 10-fold lower (K_D ~ 800 nM) affinity for diferrous sMMOH (Figure 11B). Similar values are obtained from analysis of the titrations monitored using the resonances from the 5FW labels in BTFA-K15C-5FW-MMOB (Figures S7 and S8).

Figure 10.

Typical titration of BTFA-K15C-5FW-MMOB with sMMOH. BTFA-K15C-5FW-MMOB (5 μM, red spectrum) was progressively titrated with sMMOH (gray spectra) to a final concentration of 50 μM (active site concentration, blue spectrum). The BTFA spectral region is shown. A: diferric sMMOH, B: diferrous sMMOH.

Figure 11.

Plot of fraction of sMMOH bound to ¹⁹F-labeled MMOB versus concentration of free sMMOH. Data such as those shown in Figure 10 were normalized using an internal TFA standard and a fraction bound determined using the spectra of free ¹⁹F-labeled MMOB and that of nearly fully bound ¹⁹F-labeled MMOB after addition of 50 μM sMMOH (sites). The fraction bound times the concentration of ¹⁹F-labeled MMOB present subtracted from the concentration of sMMOH added yields the concentration of free sMMOH. The free concentration shown includes the portion of sMMOH which cannot bind MMOB. The dashed curves are computed for the K_D values shown assuming that 70% of the sMMOH can bind MMOB. Data represents the average of two titrations. A: Titration of 5FW-MMOB. B: Titration of BTFA-K15C-5FW-MMOB.

The high affinity of diferric sMMOH for BTFA-K15C-5FW-MMOB allows a check on the proposal that there is a fraction of sMMOH that cannot bind MMOB. As shown in Figure S9, the titration of ¹⁹F-labeled MMOB with sMMOH at an MMOB concentration over 1000-fold above the 80 nM K_D value results in the expected linear dependence on total sMMOH. However, complete saturation of the complex requires an excess of sMMOH consistent with the presence of a 30-40 % fraction that cannot bind the regulatory protein.

Binding Affinity of MMOR for the sMMOH:5FW-MMOB Complex.

As described in the introduction, a key question is whether MMOB and MMOR can form a ternary complex with sMMOH. The high affinity of 5FW-MMOB for diferric and diferrous sMMOH allow nearly complete complex formation when 12 μM of each (sites) are mixed in an NMR tube. Under these conditions, the −125.5 and −126.6 ppm resonances from the 5FW label of 5FW-MMOB are nearly extinguished. Incremental addition of MMOR is found to restore these resonances showing that the 5FW-MMOB is displaced by MMOR (Figure 12A). As shown in Figure 12B, a plot of the fraction of the original sMMOH:5FW-MMOB complex dissociated versus MMOR added is hyperbolic for both the diferric and diferrous forms of sMMOH:5FW-MMOB. In each case, the half dissociation values of 9.1 ± 2.3 and 13.7 ± 3.5 μM for the diferric and diferrous forms, respectively, are similar to the concentration of MMOB present (12 μM total). Computation^{41, 42} of the K_D values for the diferric and diferrous sMMOH:MMOR complex from this competitive titration gives values of K_D = 60 ± 16 nM and 90 ± 23 nM, respectively. Thus, MMOR and MMOB appear to have a similar K_D values for sMMOH in each oxidation state. It is interesting to note that for both diferric and diferrous sMMOH:5FW-MMOB titrations, ~30% of the starting complex either does not dissociate or the characteristic resonance is decreased in intensity by an unidentified interaction when MMOR is added. Simple addition of MMOR to 5FW-MMOB in the absence of sMMOH does not decrease the resonance intensity.

Figure 12.

Titration of the sMMOH:5FW-MMOB complex with MMOR. A: The ¹⁹F-NMR resonances of unbound 5FW-MMOB (green spectrum) appear as MMOR is added (grey spectra) to diferric sMMOH:5FW-MMOB (red spectrum, represents residual unbound labeled MMOB under the starting conditions). Saturation (blue spectrum) occurs without full restoration of the spectrum of unbound 5FW-MMOB. B: Plot of the fraction of sMMOH:5FW-MMOB remaining versus the total concentration of MMOR added. The fraction of dissociated sMMOH:5FW-MMOB complex was determined as the change in intensity of the resonance at −126.6 ppm relative to maximum observed change in the intensity of this resonance. The starting concentration of the sMMOH:5FW-MMOB complex was approximately 12 μM. The dashed lines are hyperbolic fits to the data. The half dissociation value is the amount of MMOR required for half of the observed dissociation reaction to be complete.

MMOR Causes Both the N-terminal Tail and the Core Region of MMOB to Dissociate from sMMOH.

The presence of fluorine labels in two regions of MMOB allows their dissociation from the sMMOH:BTFA-K15C-5FW-MMOB upon MMOR binding to be independently evaluated. The spectra in Figure 13A and B show that the resonances characteristic of free, ¹⁹F-labeled MMOB in the BTFA and 5FW spectral regions emerge during the titration of diferric sMMOH:BTFA-K15C-5FW-MMOB with MMOR. Thus, the binding of MMOR results in dissociation of both the core (5FW) and N-terminal (BTFA) regions of MMOB from sMMOH.

Figure 13.

¹⁹F-NMR monitored addition of MMOR to diferric sMMOH:BTFA-K15C-5FW-MMOB complex. MMOR was added to sMMOH:BTFA-K15C-5FW-MMOB complex (12 μM labeled MMOB, 18 μM sMMOH active sites, 70% of which can bind MMOB) leading to release of labeled MMOB. A: BTFA spectral region. B: 5FW spectral region. Red spectra: Spectrum under starting conditions listed above (mixture of residual unbound labeled MMOB and sMMOH-labeled MMOB complex, Green spectra: 12 μM MMOB alone, Blue spectra: after addition of 65 μM MMOR.

DISCUSSION

This study has shown that the incorporation of fluorine into the protein components of an enzyme, either by replacing the endogenous tryptophans with 5FW or by modifying a cysteine residue with BTFA, allows detailed studies of component interactions via ¹⁹F-NMR. These modifications cause minimal steric disruption, making them ideal as probes of protein-protein interfaces. In the current case, the combined mass of the dimeric sMMOH with 2 labeled-MMOBs bound is 275 kDa, a value often considered beyond the reach of NMR investigation. Nevertheless, this large particle gives easily distinguishable, well-resolved ¹⁹F-resonances, particularly in the case of BTFA labeling, that can be used for structural interrogation and quantification in the 5 μM protein concentration range. Large changes in the NMR spectra show that complex formation between sMMOH and ¹⁹F-labeled MMOB causes structural changes in both the compactly folded core region of MMOB and the N-terminal region as it transitions from disordered to ordered upon binding. Titration studies show that MMOB and MMOR both form tight complexes with MMOH, but they are competitive with each other for a binding site. The approximate component affinities that emerge from these studies suggest a new model for electron transfer and regulation in the sMMO system which is discussed here.

Structural Changes that Occur as MMOB binds to sMMOH.

The appearance of a new resonance at −121.5 ppm upon formation of the sMMOH:¹⁹F-labeled-MMOB complex is shown by mutagenesis and CEST measurements to arise from a large change in the environment of W77. The downfield shift relative to the resonance from W77 in free MMOB is consistent with movement of this Trp to a more hydrophobic environment.^{52, 53} No new resonance is observed from W76 as the complex forms. However, a slight broadening of the base of the resonance from W76 is consistently observed (compare Figures 2 and 4A, also see Figure S4), which may arise from the complex. This observation suggests that the resonance from the complex is not shifted, and thus, the environment of W76 is largely unchanged. These findings showcase the sensitivity of the ¹⁹F-NMR in probing these types of interactions, and they also correlate with the crystal structures of the complex reported here (Figures 4 and S3) and in other recent studies.^{10, 12} Importantly, 5FW-MMOB is found to be comparable to wild type MMOB both in steady state and single turnover kinetic studies, so 5FW substitution does not appear to alter either the structural or functional aspects of the interaction of sMMOH and MMOB.

When both 5FW and BTFA labels are incorporated into MMOB, the same shift in the resonance from W77 is observed, so the change in the N-terminal region does not affect the environment of the sMMOH interface residue in the compactly folded core region of MMOB. This finding is important because W77 is adjacent to the W308-tunnel that we propose is used in the transit of O₂ and probably also CH₄ to the active site of MMOH.¹² Accordingly, O₂ binding remains fast and non-rate-limiting, and the rate constant for Q decay is not greatly affected by the BTFA modification in the N-terminal region. However, the new resonances that appear in the ¹⁹F-NMR spectrum of the sMMOH: BTFA-K15C-5FW-MMOB complex show that the BTFA may have shifted to a more hydrophobic environment and that it occupies at least two distinct positions. Neither outcome is expected based on the crystal structure of the native sMMOH:MMOB complex where residue K15 is solvent exposed and occupies a single orientation, presumably due to the K15-E23 hydrogen bond which is lost upon mutation and BTFA labeling. The steady state kinetics using the doubly ¹⁹F-labeled MMOB are largely unchanged in rate limiting step (likely product release), but the single turnover kinetics for the P* to P and P to Q steps are slowed. Past studies have shown that the rate constants for these steps are individually slowed by two distinct His to Ala mutations in the N-terminal region.¹⁸ Consequently, it is possible that the more hydrophobic trifluoromethyl group of BTFA samples the more hydrophobic environment in the sMMOH-MMOB interface and perturbs this interaction slightly. However, the enzyme remains functional and all of the intermediates of the reaction cycle are formed in high yield. These observations highlight the precise interactions between the sMMOH and MMOB components that regulate every aspect of the sMMO reaction cycle.

Binding Affinity of Labeled-MMOB for sMMOH.

Past studies have used techniques such as fluorescence quenching, fluorescence anisotropy, and ITC to determine the binding affinity between components of the sMMO system. These reports are referenced and summarized in Table 2 for comparison with the current results.

Table 2.

Reported K_D Values for sMMO Component Complexes at pH 7

Method	Organism	sMMOHox- MMOB	sMMOHred- MMOB	sMMOHox- MMOR	sMMOHred- MMOR	Reference
		nM
ITC	Mc. Bath	3000		900		26
Fluorescence Anisotropy IEDANSb	Mc. Bath	550	170	8000c		21, 22
Fluorescence Tryptophan	Mt. OB3b	67a		10		23
Fluorescence BADANd	Mt. OB3b	68e	4500e			27
Fluorescence BADANd	Mt. OB3b	68 159e	2200 4500e			19
19F-NMR	Mt. OB3b	80f 80g	80f 800g	60c,f	90c,f	This work

^aFit to a thermodynamic cycle

^bIEDANS label on D35C-MMOB from Mc. Bath

^cMMOB present in solution but displaced by MMOR

^dBADAN label on A62C-MMOB from Mt. OB3b

^epH 7.4

^f5FW-MMOB

^g5FW-BTFA-K15C-MMOB

The wide range of values may derive from the types of probes employed, protein preparations, or the sensitivity of the techniques. It is unlikely that the differences reflect the bacterial origin of the sMMO components from Mc Bath or Mt OB3b because many studies have shown that these enzymes are remarkably similar structurally and functionally.^{1, 2, 10, 12, 14, 20, 54} The titration monitored by ¹⁹F-NMR described here combines the advantages of a sensitive, easily quantified readout of the incorporation of labeled-MMOB as the complex forms with minimal perturbation of the protein structure. The latter advantage allows the probe to be placed in or near the protein-protein interface where the most perturbation in the environment of the probe is likely to occur, particularly in the case of the 5FW probes.⁵⁵ The results show that the 5FW-MMOB and BTFA-K15C-5FW-MMOB complexes with sMMOH are both very strong, in near perfect agreement with previous measurements made using fluorescence techniques and the Mt. OB3b components (Table 2). In contrast, the results show the same strong affinity of 5FW-MMOB for diferrous sMMOH, whereas previous results using fluorescent probes showed a 30-65 fold decrease in affinity (or a 3-fold increase in affinity in the case of the Mc. Bath components²²). The similarity of the steady state and transient kinetic behavior of 5FW-MMOB and wild type MMOB suggests that the latter would also show little change in affinity for diferrous sMMOH. The BTFA-K15C-5FW-MMOB exhibits 10-fold weaker affinity for diferrous sMMOH. The higher apparent K_D value of roughly 800 nM is consistent with a doubling of the concentration required to reach full activity in the steady state assay under the assumption that high sMMO activity requires formation of a 1:1 sMMOH:MMOB complex (Figure 3). It may be significant that all of the studies reporting K_D values for diferrous sMMOH:MMOB complex except the current application of 5FW probes involve mutations and addition of fluorophores in or near the N-terminal region. Thus, the wide range of reported K_D values may reflect changes in binding affinity and/or alterations in the protein-protein interface similar to those reported here for the BTFA-K15C-MMOB probe.

One conflict between the past and current results is the observation that the redox potential of sMMOH decreases substantially when complexed with MMOB.^{24, 25} If the free energy of binding is coupled with the redox potential shift, as it should be for thermodynamic properties of states at equilibrium, the observed decrease of 132 mV in potential of a two electron reduction would indicate an increase in K_D from 80 nM to over 2 mM. This value is unreasonably high given the strong interactions between sMMOH and its regulatory protein demonstrated here and in past studies. It is possible that under the conditions of the redox potential measurements (multiple redox dyes and mediators) interactions beyond the binding of the two components occurred, despite concerted efforts to rule out this possibility. Alternatively, we have observed that, while microcrystals of sMMOH are reduced in seconds by chemical reductants, co-crystals of sMMOH:MMOB require up to 12 hours of incubation at room temperature.^{10, 12} Consequently, the 1-1.5 hour time at 4 °C allowed for equilibration in previous redox titrations may not have been sufficient. Additional studies are required to resolve this conflict. It is important to point out that one of the conclusions of this study is that MMOB is not bound to sMMOH when electron transfer occurs from MMOR (see below), so the redox potential of the sMMOH:MMOB complex is of decreased relevance.

Displacement of Labeled MMOB from the sMMOH Complex by MMOR.

Addition of MMOR to either the diferric or diferrous sMMOH:5FW-MMOB complex results in the release of MMOB as observed by the return of the ¹⁹F-NMR resonances of unbound ¹⁹F-labeled-MMOB. The amount of MMOR required to cause half of the 5FW-MMOB to dissociate in both titrations is approximately equal to the amount of 5FW-MMOB present, showing that 5FW-MMOB and MMOR have similar K_D values for formation of the complex with sMMOH. The similar kinetic behaviors of 5FW-MMOB and MMOB itself imply that MMOB and MMOR also have similar K_D values. This finding suggests that MMOB and MMOR compete equally for an sMMOH binding site both before and after electron transfer to initiate the catalytic cycle.

Mechanistic Significance.

As noted above, past studies indicate that once sMMOH is reduced, binds MMOB and then binds O₂ in the catalytic cycle, MMOB remains bound at least up to the product release step. Another indication that this is true is the failure of reduced MMOR to reduce intermediate Q during turnover, which would diminish or eliminate product yield.⁵⁶ The previously proposed models for regulation, product release, and electron transfer shown in Scheme 2 invoke transit through discrete, fully occupied states. In Model A, MMOB and MMOR bind fully in independent sites, while in Model B, MMOR completely dissociates the core region of MMOB while leaving the N-terminal region bound. Both of these models are made less likely by the current results which show that: (i) both the core and N-terminal regions of ¹⁹F-labeled MMOB are completely displaced by a large excess of MMOR, (ii) labeled-MMOB and MMOR have similar K_D values for MMOH implying an equilibrium state, and (iii) the K_D values for labeled MMOB and MMOR do not change appreciably with the oxidation state of sMMOH.

The results support the new model shown in Scheme 3 in which MMOR and MMOB compete continuously for the same binding site on sMMOH independent of oxidation state. The fractional occupancy of the binding site on sMMOH is biased by the relative concentrations of MMOB and MMOR in the cell, which favors MMOB by 20-fold in active sites ratio.⁴ Nevertheless, in the new model the reaction is pulled forward by effectively irreversible reactions that occur when a specific complex is formed (bold red arrows in Scheme 3). Dissociation of a small fraction of the sMMOH^ox:MMOB complex at the end of turnover would allow product release through the pore normally covered by bound MMOB.^{10, 12} Subsequent binding of reduced MMOR would lead to rapid, effectively irreversible, reduction (k = 96 s⁻¹ in the equivalent Mc Bath complex²⁶) and formation of sMMOH^red:MMOR^ox complex, thereby removing it from the equilibrium of diferric sMMOH and its component complexes. In the next step, a small fraction of MMOR would dissociate from the sMMOH^red:MMOR^ox complex, allowing MMOB to bind. The diferrous sMMOH:MMOB complex would very rapidly bind O₂,^{12, 19} removing this fraction from the equilibrium of diferrous sMMOH and its complexes, again pulling the reaction forward.

Scheme 3. New Model for Regulation of Electron Transfer and Substrate Binding.

The cycle shown begins (upper left) and ends (lower right in the product complex (intermediate T in Scheme 2). Bold red arrows represent irreversible steps that pull the reaction cycle forward. H^ox = diferric sMMOH, H^red = diferrous sMMOH^red, B = MMOB, R^ox = fully oxidized MMOR, R^red = two electron reduced MMOR^red, H^Q = Intermediate Q of the reaction cycle. The lightning bolt graphic indicates electron transfer from MMOR^red to the diiron cluster of sMMOH.

The current results do not give information about the affinities of MMOB and MMOR for the intermediates in the reaction cycle after O₂ binds. However, only the sMMOH^red:MMOB complex can bind O₂ rapidly, so it is the only complex that can proceed in the reaction cycle. Subsequent binding of MMOR may be prevented by an increase in affinity for MMOB by the diiron(III) and (IV) intermediates P, Q, and R. Perhaps more likely based on the current results, the kinetics of flux through the reaction cycle may not allow significant dissociation of MMOB from the intermediates. Transient kinetic measurements of 5 μM BADAN-A62C-MMOB binding to sMMOH revealed a pseudo first order association rate constant at 4 °C of 400 s⁻¹ and a dissociation rate constant of 2.4 s^-1.¹⁹ These values are in reasonable accord with the ¹⁹F-NMR spectrum presented here which shows a 5 ppm shift of the resonance from 5FW-MMOB as the complex with sMMOH forms. For a Larmor frequency of 565 MHz for ¹⁹F, these values, and the lack of resonance broadening during titration, require that the exchange rate constant of 5FW-MMOB in and out of the sMMOH:5FW-MMOB complex (k_on + k_off) be very slow relative to 17,750 s⁻¹ (5 ppm x 2π x 565 MHz).⁵⁷ In a single turnover with a typical assay concentration of 100 μM CH₄ present, the reaction cycle prior to product release is rate-limited by Q formation at 2.4 s⁻¹ (Table 1). All of the other intermediate steps are substantially faster, so MMOB would have a limited time window in which to exchange, and its 20:1 concentration advantage would help to protect Q from reduction by MMOR.

The regulatory principles presented in Model C may find application in other members of the bacterial multicomponent monooxygenase family. Perhaps the best studied of this group are the aromatic substrate monooxygenases.^58-60 For example, elegant structural studies of toluene 4-monooxygenase (T4MO) have demonstrated that the ferredoxin component (T4moC) that supplies the electrons required for catalysis and regulatory protein (T4moD) equivalent to MMOB share a binding surface near the diiron cluster in the active site. Thus, they cannot be co-resident throughout the reaction cycle.^{61, 62 ,63} Dynamic component exchange in the T4MO system may lead to efficient catalysis in a similar manner to that we describe here for sMMO.

Conclusion.

A pervasive theme in oxygenase mechanistic studies is the requirement to carefully control the hyper-reactive intermediates that are generated to enable transfer of an oxygen atom into a stable substrate bond.^1,3,43,58,64 The ultimate test of this principle is the reaction of the diiron(IV) intermediate Q of sMMO with the stable C-H bond of methane. The process of Q formation requires initial input of two electron to activate O₂, but the additional transfer of even a single additional electron after Q is formed would deactivate it sufficiently to obviate the attack on its only biologically relevant substrate. In sMMO, Nature has constructed a remarkable machine to control this process in which even minor changes in the interface between the catalytic sMMOH component and the MMOB regulatory component compromise rate and specificity. It is shown here that labeling with 5FW allows the investigation of the MMOB interaction with sMMOH using ¹⁹F-NMR in a manner that does not require perturbation of the interface. The frequent placement of Trp residues in hydrophobic protein-protein interfaces, the relative rarity of this amino acid, and ability to study micromolar reactant concentrations suggest that this approach will continue to find wide application. The sensitivity of the sMMOH-MMOB interface is highlighted by the effects of replacing a lysine with the structurally conservative Cys-BTFA label in the N-terminal of the regulatory protein (Figure 5D). Even this small change decreases the binding affinity ten-fold and alters the rate constants of several steps in the catalytic cycle. Nevertheless, the BTFA and 5FW probes together very effectively demonstrate the regulatory importance of both the core and N-terminal structural regions of MMOB. It is often convenient to think of enzymatic reactions as occurring in a linear fashion with one step completing before the next initiates. However, it is increasingly apparent that many biological processes instead occur by shifts in equilibrium of the relevant complexes.^{39, 64-67} The current results suggest that a form of this dynamic equilibrium concept is at the heart of sMMO regulation. Nature forms and protects the key oxidative intermediate by controlling the concentrations, lifetimes, and irreversibility of reactions of specific protein-protein complexes.