University of Southern California Dissertations and Theses

Interactions of diazene homologues with Azotobacter vinelandii nitrogenase enzyme and model systems

(USC Thesis Other)

Interactions of diazene homologues with Azotobacter vinelandii nitrogenase enzyme and model systems

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content INTERACTIONS OF DIAZENE HOMOLOGUES WITH AZOTOBACTER VINELANDIINITROGENASE ENZYME AND MODEL SYSTEMS by Anton M. Sitneonov A Dissertation Presented to the FACULTY OF THE GRADUATE SCHOOL UNIVERSITY OF SOUTHERN CALIFORNIA In Partial Fulfillment of the Requirements for the Degree DOCTOR OF PHILOSOPHY (Chemistry) May 1998 © 1998 Anton M. Simeonov Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. UNIVERSITY OF SOUTHERN CALIFORNIA THE GRADUATE SCHOOL UNIVERSITY PARK LOS ANGELES. CALIFORNIA 9000? This dissertation, written by Anton Momtchilov Simeonov under the direction of Dissertation Committee, and approved by all its members, has been presented to and accepted by The Graduate School, in partial fulfillment of re quirements for the degree of DOCTOR OF PHILOSOPHY -BeaH of Graduate Studies Date .. .I...}.. DISSERTATION COMMITTEE Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Acknowledgements I would like to thank Prof. Charles McKenna for his research advisement His constant devotion to the projects and ability to suggest ideas and solutions certainly made a difference. From him I learned not only a number of experimental techniques but also how to conduct independent research and present my ideas and achievements. He made himself available for questions and discussions at virtually every time during my work. Special thanks go to all present and past group members as well as USC Chemistry staff who were all part of the creative and nurturing environment that helped me grow professionally and feel supported at every moment The financial support by USDA- CREES, the USC-Oakley Merit Fellowship, and the Department of Chemistry arc greatfiilly acknowledged. Finally, I cannot be thankful enough to my family for their endurance and encouragement throughout these years. The material presented in Chapters 1,3, 5, 6 and 7 has been the work of the author only. The majority of the diazirine work (Chapter 2) has been performed by former group members (Dr. H. Eran and M. Bravo) and is described as a part of a published paper. The cubane model work (Chapter 4) was done in collaboration with Prof. D. Coucouvanis and his group (University of Michigan) and is also presented as a whole. The cluster synthesis, X-ray and EPR analyses, and assay execution was done at the University of Michigan, whereas the preparation and analysis of cw-dimethyldiazene, as well as the HPLC analyses of the reaction products was done by the author at USC. The large-scale Fe protein purification from frozen cells described in Chapter 7 was performed by Dr. W. Gutheil, Mr. S. Krause, and the author. ii Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table of Contents Page Acknowledgements n List of Figures v Abstract viii Chapter 1: A Convenient Phase-Transfer Method for Preparation of Pure cis- Dimethyldiazene (m-Azome thane) in Aqueous Solution. Proton and Carbon NMR Studies of trans- and cw-Dimethyldiazene. 1 Abstract 2 Introduction 2 Experimental 3 Results and Discussion 6 References 11 Chapter 2: Reduction of Cyclic and Acyclic Diazene Derivatives by Azotobacter Vinelandii Nitrogenase: Diazirine and /ra/w-Dimethyldiazene. 24 Abstract 25 Introduction 26 Experimental 29 Results 40 Discussion 44 References 56 Chapter 3: Nitrogenase-Catalyzed Reduction of m-Dimethyldiazene: A New Chemical Probe of the Enzyme Mechanism. 72 Introduction 73 Experimental 73 Results 75 Discussion 80 References 87 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chapter 4: The Catalytic Reduction of os-Dimethyldiazene by the [MoFe3S J 3 * Clusters. The Four-Electron Reduction of a N=N Bond by a Nitrogenase- Relevant Cluster and Implications for the Function of Nitrogenase. 113 Abstract 114 Introduction 114 Experimental 117 Results and Discussion 121 References 128 Chapter 5 :1. Modified Protocol for Preparation of Iron-Molybdenum Protein and Cofactor, n. Interactions of m-Dimethyldiazene with Azotobacter vinelandii Cultures. 146 Introduction 147 Experimental 148 Results and Discussion 135 References 162 Chapter 6 : Interaction of Mono-Methyldiazene with Azotobacter vinelandii Nitrogenase. 170 Introduction 171 Experimental 172 Results and Discussion 179 References 191 Chapter 7: Positional Isotope Exchange and Washout Studies of Azotobacter vinelandii Nitrogenase. 220 Introduction 221 Experimental 223 Results and Discussion 231 References 233 iv Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure List of Figures 1.1 Apparatus for preparation of aqueous solutions of 2. 1.2 Results from a typical irradiation process in the designed apparatus. 1.3 ‘H NMR of 1 in CDC13 showing the expanded satellite quartets. 1.4 ‘H NMR of the acetonitrile (ACN) layer from the irradiated apparatus. 1.5 ‘H NMR of the dimethylformamide (DMF) layer from the irradiated apparatus. 2.1 Absence of HD formation for reduction of diazirine in 50 % D2 atmosphere. 2.2. Lineweaver-Burk plot of kinetic data for reduction of trans- dimethyldiazene 2.3 Inhibition of H2 evolution by fra/ts-dimethyldiazene 2.4 Inhibition of /ra/w-dimethyldiazene reduction by CO and acetylene 2.5 Inhibition of acetylene reduction by /ra/ir-dimethyldiazene 2.6 Electron allocation to alternative products of //wis-dimethyldiazene reduction as a function of the Av2:Avl ratio. 3.1 Time course data for reduction of m-dimethyldiazene. 3.2 Plots of kinetic data for reduction of m-dimethyldiazene to methylamine, methane and ammonia. 3.3 Electron allocation to alternative products of m-dimethyldiazene reduction as a function of the Av2:Avl ratio. 3.4 Inhibition of CjH2 reduction by m-dimethyldiazene. 3.5 Inhibition of H2 evolution by m-dimethyldiazene. 3.6 Inhibition of m-dimethyldiazene reduction to ammonia by CjH2 . 3.7 Inhibition of methylamine formation by CO. 3.8 Inhibition of m-dimethyldiazene reduction by N2 . 3.9 Absence of HD formation in the reduction of m-dimethyldiazene in 50 % D2 atmosphere. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 3.10 ‘H and 2 H NMR analysis of the gas phase of m-dimethyldiazene reduction in Dz O. 109 3.11 Electron isodensity surfaces for N2 , N2H2 isomers and disubstituted diazenes. 111 4.1 The FeMo-cofactor of nitrogenase. 130 4.2 An ORTEP plot of the anion of I. 132 4.3 Time course of methylamine production from the reduction of cis- dimethyldiazene. 134 4.4 Product distributions after 3 h of reaction for complete and control systems. 136 4.5 Inhibition of m-dimethyldiazene reduction by PE^. 138 4.6 Reaction velocity versus initial cfr-dimethyldiazene concentration. 140 4.7 Percent conversion of m-dimethyldiazene to methylamine vs. initial m - dimethyldiazene concentration. 142 4.8 Possible reaction pathways for the reduction of m-dimethyldiazene by the synthetic cuboidal cluster. 144 5.1 Arrangement of the Amicon 2000 kit outside the glove box. 164 5.2 Flowchart showing the FeMo protein purification and analysis. 166 5.3 Survival and further growth of A. vinelandii culture on N-free medium after addition of methylamine and dilution. 168 6 .1 lH NMR of reaction mixture showing the emergence of the mono- methyldiazene peak. 193 6 .2 ‘H and l3 C NMR of mono-methyldiazene. 195 6.3 UV monitoring of mono-methyldiazene generation. 197 6.4 Time-course plot of mono-methyldiazene formation. 199 6.5 Proton and carbon NMR spectra of formaldehyde hydrazone. 201 6 .6 Formation of a hemochrome from the reaction between reduced human hemoglobin and mono-methyldiazene. 203 6.7 Inhibition of H2 evolution by mono-methyldiazene. 206 6 .8 Inhibition of H2 evolution by mono-methyldiazene in 22 ml vials. 208 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 6.9 Maximum N2 pressure as a function of mono-methyldiazene concentration. 210 6.10 Inhibition of QHj reduction by mono-methyldiazene. 212 6.11 Inhibition of QHj reduction by mono-methyldiazene in 22 ml vials. 214 6.12 HPLC analysis of the liquid phases from the Hj inhibition experiment 216 6.13 Methylamine formation in the methane detection experiment 218 7.1 Typical ‘H NMR spectrum of recovered inorganic phosphate from a washout experiment 235 7.2 Trimethylphosphate calibration curve for 360 MHz instrument 237 7.3 3 1 P NMR and mass spectra of a washout sample showing very little exchange. 239 7.4 3lP NMR and mass spectra of a washout sample showing significant exchange. 241 Table 1.1 trans- and cu-Dimethyldiazene: ‘H and I3 C NMR spectral data. 13 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Abstract The chemical mechanism of N2 reduction to NH3 is not yet understood in terms of specific enzyme moieties and well-defined intermediates. Diazene (diimide, HN=NH) is a proposed intermediate in biological nitrogen fixation but utilization of N2 H2 as a binding- site probe for die nitrogenase enzyme has been precluded by its extreme instability under the conditions of enzyme turnover. In the present work, interactions of nitrogenase with diazene homologues, trans- and m-dimethyldiazene, were investigated In addition, a preliminary study of monomethyldiazene was undertaken. A simple method for preparation of concentrated aqueous solutions of pure cis- dimethyldiazene was devised, obviating low-temperature vacuum line manipulation of this compound. Photoisomerization of /ra/u-dimethyldiazene to its cis isomer using a biphasic hexane-water system with irradiation of the organic layer alone facilely provides >0.10 M cis-isomer in the aqueous phase, from which contaminating fra/ts-dimethyldiazene is easily removed by hexane extraction. A complete set of ‘H and 1 3 C NMR chemical shift and coupling constant data in D2O was obtained for both isomers. High-resolution ‘H NMR of the fra/is-isomer in CDC13 and D2O reveals the presence of a long-range 12CH3<— > 13CH3 proton-proton coupling (5 /h h = 1-59 Hz). The structure and stereocompactness of cis- dimethyldiazene makes it an interesting candidate for study as a metal ligand and source for methyl radicals. The synthetic procedure presented here will facilitate its use in these or other applications. Successful preparation of ar-dimethyldiazene in acetonitrile and dimethylformamide confirms that the method can be generalized to any other solvent that possesses hexane-immiscibility and a relatively high dielectric constant viii Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Both trans- and m-dimethyldiazene were shown to be catalytically reduced by nitrogenase, thus demonstrating the first nitrogenase substrates containing an unstrained -N=N- moiety. Cu-dimethyldiazene and nitrogen both inhibit reduction competitively but are themselves inhibited by it noncompetitively. Unlike N2, m-dimethyldiazene does not support D2 -dependent HD formation and is insensitive to H2 inhibition. Deuterium labeling of the m-dimethyldiazene reduction products under fixing conditions indicates that no scrambling of hydrogen atoms takes place in the methyl group during the enzymatic reduction. The results are consistent with transfer of a single hydrogen species to this group in concert with C-N bond scission. The utility of m-dimethyldiazene as a mechanistic probe was further tested with the single cubane MoFe3S4 iron-molybdenum cofactor model of Coucouvanis et al. The model cluster also binds and catalytically reduces m-dimethyldiazene, but only at the N=N bond, leading to exclusive formation of methylamine. This underlines the important role of the amino acid residues surrounding the cofactor in the wild-type FeMo protein. The Mo-center of the model system has been shown to be involved in substrate reduction. Lastly, a preliminary investigation was conducted on /ra/u-monomethyldiazene as the closest mono-alkyl homologue of diazene. Compared to dimethyldiazenes, this molecule has less steric bulk and lacks symmetry around the N=N bond; it also has an azo- nitrogen-bound hydrogen atom, as does diazene itself. The instability of mono- methyldiazene has severely limited its use in experimental research in general but the present work demonstrates that a considerable stabilization of the compound can be achieved by handling it in Dz O solutions, thus permitting short-timescale nitrogenase ix Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. assays to be performed in its presence. Monomethyldiazene inhibits nitrogenase hydrogen evolution, and acetylene reduction by the enzyme. Methylamine appears to be formed in nitrogenase-catalyzed mono-methyldiazene reduction. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER 1 A CONVENIENT PHASE-TRANSFER METHOD FOR PREPARATION OF PURE C/S-DIMETHYLDIAZENE (C7S- AZOMETHANE) IN AQUEOUS SOLUTION. PROTON AND CARBON NMR STUDIES OF TRANS- AND CIS- DIMETHYLDIAZENE Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ABSTRACT Photoisomerization of fra/w-dimethyldiazene (rm/is-azomethane) 1 to its cis isomer 2 using a biphasic hexane-water system with irradiation of the organic layer alone facilely provides > 0.10 M 2 in the aqueous phase, from which contaminating 1 is easily removed by hexane extraction. Proton and carbon-13 NMR studies yielded a complete set of chemical shift and coupling constant data for both isomers in D2O. The carbon-13 chemical shift of 1 is downfield from 2, contradicting a previously reported theoretical (IGLO) prediction. High-resolution proton NMR of 1 in CDC13 and P 2O reveals the presence of a long-range 12CH3<— >I3CH3 proton-proton coupling (5 /h h = 1-59 Hz). Preparation of cis- dimethyldiazene in acetonitrile and dimethylformamide is also described. INTRODUCTION The preparation and chemistry of cw-dimethyldiazene (c/s-azomethane) is of interest for several reasons, including its structural relationship to cw-diazene, a possible intermediate in nitrogen fixation. 1 7ra/w-azomethane 1 can be prepared by several methods,2-4 and is stable in pure form at room temperature. The cis isomer 2, however, rapidly isomerizes to formaldehyde methylhydrazone at room temperature, and must be manipulated in organic solutions at temperatures below -SO °C using vacuum line techniques to avoid decomposition.^-? Vibrational and photoelectronic spectroscopy of both 1 and 2 have been reported,3,8-13 but the NMR of 2 has not been studied. Seeking a simple, direct route to aqueous solutions of pure 2, we have investigated the photoisomerization of 1 in a mixed-phase solvent system. Photoconversion of 1 to 2 was previously reported to proceed at room temperature in D2O with an equilibrium constant Keq = [21/11] of 0 .0 9 ,5 and 2 was noted to be stable in aqueous so lu tio n ,5*6 however, 2 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. solutions of pure 2 were prepared by photoiiradiation of 1 as a glass at -196 °C, followed by trap-to-trap distillation and co-condensation into a vessel with D2 O. We report here a phase-transfer photoisomerization procedure for direct preparation of concentrated aqueous solutions of 2 at room temperature. A complete set of and NMR data was obtained for both azomethane isomers in D2O. As an extension of the scope of the above method, 2 was subsequently prepared in two non-aqueous solvents, acetonitrile and dimethylformamide. EXPERIMENTAL SECTION All NMR samples were prepared in CDCI3 , D2O, CD,CN or dimethylformamide-dT (DMF) as noted. NMR spectra were recorded at 250.13 or 360.13 MHz and NMR spectra at 62.89 MHz on Bruker AC-250 or Bruker AM-360 spectrometers. Chemical shifts were recorded in ppm relative to internal CDCI3 (8 = 7.24, *H; 5 = 77.0, 13C), internal D2O (8 = 4.63, 1H), internal CD,CN (8 = 1.93, 1H) and internal acetone (8 = 2.04, !H; 8 = 29.8, ^ C ). Coupling constants ( ,J) are reported in Hz. Hexane (Mallinckrodt) was analytical grade, D2O (Cambridge Isotope Laboratories) was 99.9 % D, and H2O (Aldrich) was HPLC grade. Ultraviolet and IR spectra were recorded on Shimadzu UV-260 spectrophotometer and Perkin-Elmer 281 spectrophotometer respectively. The extinction coefficients for 1 (eW 3 = 25) and 2 (e35 3 = 2 4 0 )5 in water were used to measure the concentrations in aqueous solutions. The literature values were verified by *H NMR spectra of samples to which known amounts of a standardized methylamine solution were added, followed by peak integration (data not shown). Photoisomerization 3 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. was performed in a Rayonet photochemical reactor equipped with eight GE G8TS germicidal lamps. Tranr-azom ethane (1). Sym-dimethylhydrazine dihydrochloride (10 g, 75 mmol) (Aldrich) was oxidized by HgO in water and purified by passage in a N2 stream through -80 °C and -196 *C traps.2 Yields were 55-67 % and purity was satisfactory (IR**, NMR^, UV^). The product was either stored at -196 °C, or else used immediately for preparation of its solutions in hexane or water by condensation into a flask filled with the degassed solvent and cooled to -80 °C (dry ice/t-PrOH). The same approach was adopted for preparation of NMR samples of newly synthesized I. For IR spectroscopy, 1 was expanded into a pie- evacuated gas cell with NaCl windows (74 mm path length, 30 ml capacity) at partial pressures of 0.066 to 0.4 atm. Irradiation of I in D2 O Solution and Extractive Purification of 2. In a Beckman quartz cuvet (Ar) was placed 3.5 ml 0.5 M 1 in D2O. The mixture was irradiated in the photochemical reactor for 5 h, alter which a 0.5 ml sample was removed for NMR analysis (Hamilton gas-tight syringe). The spectrum showed that photochemical equilibrium (1:2 = 10:1) had been established. To the sample in an NMR tube (5 mm, capped with rubber septum) was added an equal volume of degassed hexane, the tube was inverted repeatedly, the settled hexane layer was removed, and the spectrum was recorded. The process was repeated until the proton signal for 1 was no longer detectable. Loss of 2 was negligible. Preparation of Aqueous Solutions of 2 in the Designed Apparatus. The apparatus described in Fig. 1, containing 5 ml D2O, was assembled and secured by a clamp. A robber septum stopper was emplaced and the apparatus connected to 4 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. an At manifold via a 18G needle. Three cycles of pumping and argonation were performed, after which a 20-30 ml hexane solution of 1 (0.1-1 M) was injected. The lower portion of the apparatus was wrapped with A 1 foil to a height S mm above the top of the aqueous layer. The apparatus was then removed from the manifold and fitted above a small magnetic stirrer inside the photochemical reactor placed in a cold room (4 °C). Stirring was continuous (ca. 120 rpm.) and irradiation was done either continuously or discontinuously (see Results and Discussion). The progress of the reaction was monitored by removal of the apparatus from the reactor and withdrawal of NMR samples from die selected layer by the use of a 20G 25 cm stainless steel needle (Aldrich). When the proton signal of the cis isomer in D2O samples ceased to grow, irradiation was stopped, the hexane layer was removed and residual 1 was removed from the aqueous layer by repeated extraction with degassed hexane in the same apparatus. During the extractions, all liquids were kept cold by periodically immersing the apparatus in liquid nitrogen. The resulting colorless aqueous solution of 2 was free of detectable impurities (NMR, UV). When 2 was prepared in H2O, the progress of the reaction was monitored by removal of an aqueous sample, extraction with hexane, and recording the UV spectrum. Preparation of cis-dimethyldiazene in acetonitrile and dimethylformamide. c/s-Dimethykliazene was obtained in acetonitrile by irradiation in the same apparatus as described above. For stability studies, samples at ca. 60 mM concentration were aliquoted in several NMR tubes and stored at different temperatures. Periodically, the tubes were allowed to warm up to room temperature and ‘H NMR was measured to determine the degree of decomposition. The later was expressed as a percentage of the cis- dimethyldiazene peak height; integration proved to be unreliable in this particular case. For preparation of 2 in DMF, hexane solution of /ra/u-azomethane (ca. 1.3 M, ca. 3 ml) was mixed with 0.7 ml DMF-d, in an Ar-purged quartz cell fitted with rubber septum 5 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. stopper. The lower half of the cell was wrapped with A 1 foil The cell also contained a small (4-5 mm) teflon spin bar. The mixture was irradiated with stirring directly into the UV reactor for 2.5 hours, and then a sample for NMR was withdrawn. Subsequent to these initial tests, 2 was prepared in DMF by irradiation of the two-phase system in a 10 mm quartz test tube fitted with a rubber septum stopper. RESULTS AND DISCUSSION Trans-azomethane I was prepared by the method of Renaud and Leitch^ from 1,2- dimethylhydrazine and purified from traces of water and chloromethane (by-product) by standard vacuum line techniques. This isomer is moderately soluble in water (Henry's Law constant of 1.50 M/atm at 3 0°C ),^ permitting its aqueous (H2O or D2O) solutions to be prepared up to -0.5 M. Irradiation of die above solutions in a photochemical reactor (Experimental Section) yielded an equilibrium mixture of 1 and 2 in the ratio of 10:1 as previously reported.^ Extraction of the equilibrium mixture with degassed hexane resulted in complete removal of 1 (<0.5 %, NMR) horn the aqueous phase. Proton NMR of the hexane layer (CDCI3) showed 2 to be virtually absent (<1% of 1) in the organic phase, consistent with the polar nature of the cis isomer. The much greater solubility of 2 in water compared to hexane prompted us to consider photoisomerization of 1 in a two-phase system, in which only the organic layer is irradiated. Under these conditions, newly formed 2 was expected to transfer to the aqueous phase and accumulate there as the irradiation proceeded, thereby enriching the dark phase and correspondingly depleting the irradiated phase in 2 , allowing its net formation to continue. 6 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ^ ► ^ -N 2 M 0 2 1 I V Hexane 1 0 0 m u > D I 1 H 2 O or DjO 2 D ark To test this idea, the apparatus shown in Fig. 1 was constructed. The lower two- thirds of the irradiation vessel portion was made of quartz. The standard taper joint and stopcock in the upper, removable portion permits attachment to a gas manifold or a vacuum line, if necessary. Details of charging the vessel and masking the aqueous phase zone are given in the Experimental Section. Other designs were subsequently tested, including a simplified version consisting of a quartz tube converted into a large test tube which was capped with a proper size rubber septum stopper. Figure 2 shows the result of a typical irradiation experiment For dilute (<0.05 M) starting solutions of 1, the concentration of 2 reached constant values after ca. 12 h of irradiation (data not shown). When concentrated solutions of 1 were used, longer irradiation times were necessary in order to achieve the maximum yield of 2. In our unthermostated photoreactor, however, despite operation in a 4°C cold room, long continuous irradiations caused excessive warming of the mixture and formation of byproducts (not studied further). This complication was eliminated by applying the UV radiation discontinuously in ca. 1 h intervals followed by 30 min. radiationless stirring. A reaction vessel equipped with a cooling jacket would doubtless permit uninterrupted irradiation. It was possible to prepare highly concentrated (up to 1 M) solutions of 1 by condensing it into degassed hexane at -80 °C; they were stored safely in a -20 °C freezer for prolonged periods of time. By charging the apparatus with these solutions at 4 °C and performing the irradiation, it was possible to obtain 2 in 7 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. concentrations of up to 0.30 M. At the end of the process, residual 1 was quantitatively removed from the aqueous layer by extraction with degassed hexane. Aqueous solutions of purified 2 were stable for at least 4 months (NMR, UV) at -20 °C. Storage at room temperature for more than two weeks resulted in tautomerizadon of 2 as evidenced by the appearance of NMR peaks corresponding to the formaldehyde methylhydrazone.5’1 '7 A fact worth noting is that the conversion of 1 into 2 , although about three times higher than obtained with monoaqueous irradiation, is still less than predicted by Keq and our estimates of the respective partition coefficients (defined as molarity in aqueous phase/molarity in organic phase): Kp( (0.3) and Kpc (100). There are two possible explanations for this result. Firstly, in our simple apparatus, stray UV light may enter the aqueous phase, depleting cis isomer by conversion to trans isomer. Secondly, some photodecomposition occurs, as evidenced by a moderate pressure build-up over the course of the reaction. It is clear, however, that the remaining hexane solution of 1 can be reused by simply withdrawing the aqueous layer and recharging the apparatus with a new portion of degassed water, and, if necessary, with an additional amount of concentrated hexane solution of 1. With concentrated D20 solutions of 1 and 2 at hand, we decided to obtain a set of *H and 13c NMR data for both compounds using relatively high-field (250, 360 MHz) spectrometers, as the literature data were obtained on early NMR instruments, the NMR data for 2 in D2O used HDO as the reference, and the NMR of 2 was not reported. The results of our studies are summarized in Table 1. The proton chemical shifts for both isomers are in good agreement with the published values obtained at 100 MHz or b elo w .5 .1 5 The satellites in the proton spectrum of 1 appear as quartets (5/hH = 1-59 Hz), which we attribute to long-range 5-bond proton-proton coupling in naturally abundant 8 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. /ra/w-H3^C-N=N-12cH3 (Figure 3). 3/HH coupling has been reported for 1 only in a complex formed with W(CO)5.7 hi this case, the methyl groups were chemically non equivalent (only one N coordinated to the metal). The isomer 1 thus demonstrates a homoallylic coupling system, wherein the trans configuration of the double bond particularly favors the distal spin-spin coupling interaction. 18 The corresponding/ value in 2, expected to be <1 Hz,?>18 was not discemable from our spectra. A l^C NMR chemical shift of SS.6S ppm was previously reported for 1 in CDCI3 I9 whereas we obtain 56.70 ppm. The l^C chemical shifts of both 1 and 2 have been predicted on the basis of IGLO calculations.20 Our 13c chemical shift values (D2O) are comparable in magnitude to the calculated chemical shifts, but occur in reverse order (1>2). This disagreement with the predicted l^C chemical shift order is also seen for the lH chemical shifts (ref. 6; Table 1) and points to an anomalous behavior for 1 and 2 relative to larger homologues.21 Preparaton of cis-dimethyldiazene in acetonitrile and dimethylformamide. Figure 4 shows a typical ‘H NMR spectrum obtained from the acetonitrile layer of the irradiated system in the cold room (5-8 °C). It is important to note that the relative orientation of the proton peaks of the two isomers remained the same as in D2 0. In addition, similar 5-bond proton-proton coupling was observed upon a closer examination of the NMR spectra (data not shown). The relative distribution of 1 in the two layers as well as the final yield of 2 were not determined. However, the concentration of the routinely prepared batches of 2 in acetonitrile was 0.1 -0.2 M. The samples of 2 stored at room temperature and 4 °C showed significant decomposition after 3 days, whereas the 9 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ones kept at -20, -80 and -196 °C were stable for at least 14 days as determined by ‘H NMR. A very precise determination of the long-term (>10-11 days) cu-dimethyldiazene stability at these lower three temperatures was rendered fairly difficult because of the relatively large number of thaw-refreeze cycles to which the NMR tubes had to be subjected. For die purpose of reactivity and kinetic evaluation of the cis- dimethyldiazene/FeMo-cofactor system, it was necessary to explore the possibility of preparing this compound into a cofactor-compatible solvent Due to the lack of readily available deuterated N-methylformamide, the traditional solvent for the iron-molybdenum cofactor handling, perdeuterated dimethylformamide was employed. The proton spectrum of a typical irradiation mixture (cold room, 5-8 °C) (Figure 5) revealed the very large alkyl signals from the hexane partially dissolved into the DMF layer (0.9 and 1.3 ppm), the very small signals (quintets) from the DMF-d, residual methyl protons (2.74 and 2.91 ppm), and the sharp peaks of trans- (3.68 ppm) and cu-dimethyldiazene (3.58 ppm). Once again, apart from a small downfield shift of ca. 0.1 ppm, the trans-fcis- pair of peaks appeared at the same chemical shift, and the frequency difference between the two isomers themselves (0.10 ppm) was virtually unchanged from that observed in the other two solvents. This fact, therefore, may allow one to "find" these peaks in more complicated spectra, should such situation arise. Apart from some impurities peaks known to come from die hexane solvent, there were no significant breakdown products observed in this and other NMR. The yield of 2 in DMF as well as the stability of its preparations in this solvent were not investigated. In summary, we have developed a simple room temperature method for preparation of concentrated aqueous solutions of pure cir-azomethane, obviating vacuum line manipulation of this compound. The structures and stereocompactness of I and 2 have 1 0 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. made them interesting candidates as metal l i g a n d s , 7 * 19 ^ p ^ o . ^ thermal lability make them valuable sources for methyl r a d i c a l s . 2 2 - 2 4 The procedure presented here will facilitate the use of 2 in these or other applications. The successful preparation of cis- dimethyldiazene in the three different solvents supports the basic idea behind the above method, thus making it possible to achieve the compound’s preparation in any other solvent that possesses the two main properties shared by water, acetonitrile and dimethylformamide, namely hexane-immiscibility and relatively high dielectric constant. REFERENCES (1) McKenna, C. E. Chemical Aspects of Nitrogenase; Pergamon Press:, 1980. (2) Renaud, R.; Leitch, L. C. Can. J. Chem. 1954,52, 545-549. (3) Craig, N. C.; Ackermann, M. N.; MacPhail, R. A. /. Chem. Phys. 1978, 68, 236-46. (4) Thiele, J. Chem. Ber. 1909,42, 2575-2580. (5) Hutton, R. F.; Steel, C. J. Am. Chem. Soc. 1964,86, 745-6. (6) Ackermann, M. N.; Craig, N. C.; Isberg, R. R.; Lauter, D. M.; MacPhail, R. A.; Young, W. G. J. Am. Chem. Soc. 1977, 99, 1661-3. (7) Ackermann, M. N.; Dobmeyer, D. J.; Hardy, L. C. J. Organomet. Chem. 1979,182, 561-79. (8) West, W.; Killingsworth, R. B. J. Chem. Phys. 1938, 6, 1-8. (9) Stevens, J. F., Jr.; Curl, R. F., Jr.; Engel, P. S. / . Phys. Chem. 1979, 83, 1432-8. (10) Stevens, J. F., Jr. Diss. Abstr. Int. B 1977,38, 1238. (11) Mosher, O. A.; Foster, M. S.; Flicker, W. M.; Beauchamp, J. L.; Kuppermann, A. J. Chem. Phys. 1975,62, 3424-30. (12) Haselbach, E.; Schmelzer, A. Helv. Chim. Acta 1971,54, 1575-80. (13) Ackermann, M. N.; Craig, N. C.; Isberg, R. R.; Lauter, D. M.; Tacy, E. P ./. Phys. Chem. 1979,83, 1190-200. 1 1 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. (14) Dung, J. R.; Pale, C. B.; Harris, W. C. J. Chem. Phys. 1972, 56, 5652- 5662. (15) Freeman, J. P. /. Org. Chem. 1963,28, 2508-2511. (16) McKenna, C. E.; Simeonov, A. M.; Eran, H.; Bravo-Leerabhandh, M. Biochemistry 1996,35, 4502-14. (17) Lemal, D. M.; Menger, F.; Coats, E. J. Am. Chem. Soc. 1964, 86, 2395- 2401. (18) Jackman, L. M.; Stemhell, S. Applications o f NMR Spectroscopy in Organic Chemistry, Fergamon: Oxford, 1969. (19) Van Baar, J. F.; Vrieze, K.; Stufkens, D. J. J. Organomet. Chem. 1974, 81, 247-59. (20) Schindler, M. J. Am. Chem. Soc. 1987,109, 5950-5. (21) Engel, P. S.; Bishop, D. J. J. Am. Chem. Soc. 1975, 97, 6754. (22) Weldon, M. K.; Friend, C. M. Surf. Sci. 1994,310, 95-102. (23) North, S. W.; Longfellow, C. A.; Lee, Y. T. J. Chem. Phys. 1993, 99, 4423-9. (24) Squillacote, M.; De Felippis, J. J. Org. Chem. 1994,59, 3564-71. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 1. trans- and cis-Dimethyldiazene: !H and 1 3 C NMR Spectral Data. Dimethyldiazene Isomer NMR Parameter trans (1) cis (2 ) lH d (ppm) 3.70 (CDC13)* 3.54 (D20)b 3.44 (D2O y > >HV h h (Hz) 1.59 (CDCI3) 1.56 (D2O) 1 3 C d (ppm) 56.70 (CDC13)C 55.12 (D2O) 46.89 (D2O) ,3C ^ ( H z ) 136.3 (CDCI3) 137.7 (D20) 143.6 (D2O) * A value of 3.67 ppm was reported in ref. S. b Hutton and Steel (ref. S) reported 1 and 2 as having resonances at 0.984 and 1.08 ppm upfield from HDO. c Van Baar et al. (ref. 19) report 55.65 ppm. 13 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 1. Apparatus for preparation of aqueous solutions of 2. The lower portion was made from a quartz tube to which a standard taper 24/40 joint was annealed. The top portion consisted of a 24/40 joint, glass stopcock, and 14/20 joint annealed together. 14 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 14/20 24/40 cm Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 2. Results from a typical irradiation process performed in the designed apparatus. The apparatus in Fig. 1 was charged with 5 ml D2O and 25 ml 0.5 M hexane solution of 1. a. lH NMR spectrum of the aqueous layer after 3 h; 1:2 ratio = 1.3:1; b. Same, after 5 h; 1:2 ratio = 1:2.0; c. The aqueous layer from b. after extraction with five 10 ml portions of hexane. Note: the resonances due to the hydrazone rearrangement product from decomposition of 2, if present, would have been observed near 2.5 and 6.5 ppm. 16 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. H D O __J__ b _L r 8 t-------1 ----- 1 ------ 1 -----r 6 5 4 3 2 ppm 17 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 3. ‘H NMR of 1 in CDCL, showing the expanded satellite quartets. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. K |T T )'l | I I I I | I I I I | I I I I | I I I I | I I I I | I I I I | I I I I | I I I I | I I I T [ I I I I | 7.0 6.S 6 .0 9 .9 9.0 4 .5 4 .0 3.9 3 .0 2 .9 PPM Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 4. lH NMR (250 MHz, 8 scans) of the acetonitrQe (ACN) layer from the irradiated apparatus after ca. 50 hours of discontinuous irradiation in the cold room. 20 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. trans Hexane CIS ACN i— i i |— i— i— i i | i i i i—| i i i —i | i i i i— |— i— i— i— i— | i i i i |— i— r 7 .0 0 6.00 5.00 4.00 3 .0 0 2 .0 0 1.00 21 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 5. ‘H NMR (250 MHz, 8 scans) of the dimethylfonnainide (DMF) layer from the irradiated apparatus after ca. 2.5 hours of discontinuous irradiation in the cold room. 22 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Hexane trans cw D M F ^ — i — r i i— i— i— i—|—i i i — i | i i i — i j — » i i r | i — i r i"i— i — i — i — r 7.00 0.00 5.00 4.00 9.00 2.00 1.00 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER 2 REDUCTION OF CYCLIC AND ACYCLIC DIAZENE DERIVATIVES BY A ZOTOBACTER VINELANMI NITOOGENASE: DIAZIRINE AND TRANS- DIM ETHYLDIAZENE Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ABSTRACT Nitrogenase reduces N2 to NH3, but the mechanistic details are unclear. Diazene (N2H2), a proposed 2e‘/2H+ intermediate on the reduction pathway, is labile under typical enzyme assay conditions, and no evidence is available on whether or not it can be reduced by or inhibit nitrogenase. In this paper, we compare the interactions of A.vinelandii (Av) nitrogenase with two diazene analogues: the previously-studied diazirine, a photolabile diazene containing the azo (-N=N-) group in a strained, three-membered ring, and trans- dimethyldiazene, a diazene containing an unstrained /rans-disubstituted N=N bond. Earlier studies in this group have established the main kinetic parameters of the diazirine interaction with nitrogenase. Diazirine synthesis, characterization and its employment in diazirine/D2/H2 0 assay is presented here as a part of verification of an earlier result The experimental Henry's Law constant (1.50 M/atm) determined for rra/w-dimethyldiazene in H2O shows that it has about 20-fold higher solubility than diazirine in water at 30 °C. /ra/u-Dimethyldiazene is reduced by nitrogenase to ammonia, methane and methylamine in a ratio of 1 :1 .2 :1.3 with Km values for the three products of 0.51-0.58 M. The product ratio does not change significantly when the component ratio (Av2:Avl) is varied over 2.06— 13.62. /ra/tf-Dimethyldiazene reduction is inhibited non-competitively by CO and C2H2 with Ki values of ca. 0.0008 and 0.006 atm respectively. The results are discussed with respect to the stereoelectronic differences between the two compounds. A "random- edge" reduction scheme is proposed for the diazirine reduction, in which a random cleavage of the C—N and N=N bonds results in formation of bound methyldiazene and diaziridine /diaminomethane intermediates respectively, which are subsequently reduced to the final products. For /ra/u-dimethyldiazene, the initial C—N cleavage yields CH4 and bound methyldiazene, which is then reduced to CH3NH2 and NH3. 25 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. INTRODUCTION Nitrogenase ( EC 1.18.6.1) is the enzyme responsible for biological nitrogen fixation, which can be represented by the following equation: W N2 + 8H+ + 8e - ---------- > 2NH3 + H2 (1) In the absence of exogenous reducible substrates, nitrogenase functions as a hydrogenase. 2 Besides the enzyme, sources of energy (ATP) , 1 electrons (dithionite), and protons (H3 0 + in aqueous solutions) are required for in vitro activity. Standard nitrogenases from different types o f azotrophs all consist of two metalloproteins: an iron (Fe) protein (dinitrogenase reductase), and a molybdenum-iron (MoFe) protein (dinitrogenase). In die aerobe Azotobacter vinelandii. the Fe protein (Av2) is a y2 dimer (-60 kDa) 3 and die MoFe protein (A vl) is a a 2 p2 tetramer (-240 kDa). 4 The X-ray structures of both proteins (Av2 , 2.9 A and A vl, 2.7 A) are now available. 3’5 Although both nitrogenase components are necessary for enzyme activity, their functions differ. As implied by the nomenclature given above, 1 the Fe protein transfers electrons to the MoFe protein which is believed to contain the active site of N2 reduction. This active site is thought to include the FeMo-cofactor, 6 a unique cluster of one Mo atom, 6 -8 Fe atoms, 8- 9 S atoms and one homocitrate ligand. 7 Models of the FeMo-cofactor have been proposed based on X-ray crystallographic electron densities observed at the location of the cluster which is - 10 A below the protein surface. In one model, 5 the cofactor is shown to contain two clusters of composition 4Fe-3S and lMo-3Fe-3S that are bridged by 3 putative non-protein S2* and a low density 26 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ligand, possibly a N or an O. Besides its protein ligands (a Cys (cx275), two His (al95 and a442) and a Gin (a l9 1 )), the cofactor is coordinated through its Mo atom to homocitrate in a bidentate manner. The proposed intracluster cavity in the cofactor model is too small by 0.5 A to accommodate N2, but it has been suggested that the cofactor metal cage expands on reduction to admit the substrate. 5 Several modes of binding between the cofactor (Mo or Fe site) and N2 have been proposed (e.g., end-on or side-on) 1 0 ’U but none have been unequivocally verified. The chemical mechanism of N2 reduction to NH3 (e.g. through three 2H+/2e' steps, a 4H+/4e- step followed by a 2H+/2e' step, or else a single 6 H+/6e- step) 10 is not yet well understood in terms of specific enzyme moieties and well-defined intermediates. The possibility of an enzyme-bound form of diazene (HN=NH), a 2e'/2H+ intermediate, remains a subject of active inquiry by model chemists, 12 and hydrazine (H2N-NH2), a 4e" /4H+ intermediate, has been detected in acid- or base-quenched nitrogenase systems. 13 A metal-bound diazene tautomer (e.g. M=N-NH2) has also been postulated as a 2e'/2H+ intermediate on the pathway of N2 reduction by nitrogenase. 14 Alternative substrates and inhibitors are of value as probes for elucidating the mechanism of nitrogen fixation. 2 The majority of these molecules contain a reducible triple bond (HCN, MeNC, HN3/N 3*, C2H2; and CO, an inhibitor which is not reduced) (some of the substrates have side-reactions with nitrogenase, e .g.. CN- is a potential modifier of the enzyme binding site 15-17 and nitrite inactivates the Fe protein 18’1 9 ). Chemical probes containing a -N=N- double bond analogous to that in diazene are clearly of interest The cyclic azo compound diazirine was proposed as a possible substrate and active site probe 20 27 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. shortly after the discovery that the structurally related cyclopropene 21 is a nitrogenase substrate with unique properties, although unstrained acyclic alkenes such as ethylene are not bound by wild-type nitrogenases. Diazirine contains the -N=N- (double) bond in a candidate as a photolabel for the nitrogenase active site. Shortly thereafter, diazirine was briefly reported to be an inhibitor 20 of nitrogenase, and also a substrate reduced to methane, ammonia 24 and methylamine. 25 More detailed studies26 revealed that diazirine is reduced by nitrogenase to methane, methylamine, and ammonia in a ratio of 1 : 2.4 : 4.3 with a Km value for all three products similar (0.05—0.07 mM) to that of dinitrogen (0.06—0.12 mM). The Km value of diazirine does not depend on the ratio of nitrogenase Fe protein (Av2) to nitrogenase MoFe protein (Avl) at Av2:Avl ratios of 0.71 and 14.91. Diazirine potently and competitively inhibits acetylene reduction by Av nitrogenase with Ki = 0.03 mM, and is predicted to inhibit H2 evolution completely at pressures above ca. 0.003 atm. It is not clear whether the special structural properties conferred by the small-ring geometry of diazirine are essential to its interaction with nitrogenase, as is the case with cyclopropene vs. ethylene or propene: no acyclic azo compound has been demonstrated to be reduced by nitrogenase. The recent availability of the X-ray crystallographic structures of Avl and the subsequently developed FeMo-co model stimulates further examination of alternative strained, three-membered ring. In addition, it is photolabile 22 and therefore is a potential : n = n : . n = n . • • D in itrogen D iazirin e fra/T S-D im ethyldiazene 28 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. substrates as probes for nitrogenase reduction mechanisms. To obtain a better understanding of the role of steric and electronic effects in these interactions, we have also investigated a new diazene analogue as a nitrogenase substrate, /ra/u-dimethyldiazene. In contrast with diazirine, /ra/u-dimethyldiazene is an unstrained, acyclic structure which however has a less compact steric profile, with its two methyl groups stretching in opposite directions at a 111.9° angle from the -N=N- bond axis, 27 within a common molecular plane. It also differs electronically, in terms of the orientation of its lone pair electrons, their coordination aptitude, and in the absence o f Walsh-like C—N orbitals. By analogy with diazirine, /ra/is-dimethyldiazene reduction could lead to methylamine, ammonia, methane, and/or other products. Presented here is an account of the synthesis and A. vinelandii nitrogenase interactions of diazirine in die presence of D2 as well as documentation that rra/u-dimethyldiazene is a new reduction substrate for the enzyme. EXPERIM ENTAL R eag ents. Chemicals and biochemical reagents used in nitrogenase assays were obtained from Sigma/Aldrich Co. in the purest grade available. CO, H2, N2, and acetylene were obtained in 99.9% purity from MG Industries or Gilmore Liquid Air Co. Methane (99.0%) and ethylene (99.5%) gas chromatography (GC) standards were from MG Industries. Ethane (99.99%) was purchased from Matheson. 1,2-Dimethylhydrazine dihydrochloride (99+%) and dansyl chloride (98%) were purchased from Aldrich. Mercuric oxide (reagent grade), magnesium chloride and sodium tetraborate decahydrate (both analytical grade) were procured from Mallinckrodt. For diazirine synthesis, f-butyl ether (Aldrich) was purified by fractional distillation at 75 °C and 80 mm Hg. Sodium dithionite (Sigma) used in the trans- 29 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. dimethyldiazene assays was purified by twofold recrystallization using a method described elsewhere. 28 Analytical Methods. IR and UV spectra were measured on Peridn-Elmer 281 IR and Beckman Acta VI or Shimadzu UV-260 UV-visible spectrophotometers, respectively. *H NMR spectra were recorded at 250.13 or 360.13 MHz and 1 3 C NMR spectra at 62.89 MHz on Broker AM- 250 or Broker AM-360 spectrometers, unless noted otherwise. Identification of CH4 as a reduction product was performed by mass-spectrometry on an LKB 9000 mass spectrometer at 70 eV for diazirine reduction and on a Hewlett-Packard 5989A GCVMS coupled to an HP 5965B IRD for dimethyldiazene reduction. Gas chromatography (GC) was performed on a Van an 2400 or Shimadzu GC14A equipped with dual flame ionization (FI) detectors or a Varian 3700 gc equipped with both FI and thermal conductivity (TC) detectors. GC peaks were integrated using a Varian 485 or Hewlett-Packard 3390A recording integrator. Chromatograms were also recorded on a Varian A-5 recorder or Macintosh Centris computer via the Mac Integrator package from Rainin Instruments. The HPLC system used to determine ammonia and methylamine has been previously described. 29 Calculations were performed with an IBM personal computer using a Lotus 1-2-3 spreadsheet (Lotus) or on Apple Macintosh computers using the Microsoft Excel 4.0 spreadsheet Synthesis of Diazirine and Its Precursors: General Comments. Diazirine was synthesized from dichloroamine (prepared in situ from sodium hypochlorite and ammonium chloride in sodium formate buffer) and r-octylazomethine by an adaptation of the method of Graham. 30 We find that diazirine gas is stable in the dark 30 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. over saturated aqueous Na2S C >4 (see below), however it must be handled on the vacuum line with die utmost caution, and close attention must be given to reproducing exactly die procedure described (see also important safety note in section "Synthesis and Purification of Diazirine" below). Synthesis of t-Octylazomethine. 31 Aqueous formaldehyde (0.68 mol, 37% solution) was added drop wise to an equimolar amount of t-octylamine 32 at 10 XI with vigorous stirring over 2 h. The upper layer was allowed to stand at 4°C over KOH overnight The crude product (stench) was purified by distillation at 54°C (30 mm Hg): lit, b.p. 50-52°C (13 mm Hg) 33 and identified by IR and NMR (250 MHz, CDC13 ): 8 0.88 (s, 9H), 1.15 (s, 6 H ), 1.58 (s, 2H), 7.35, 7.23 (2d, 2 J m = 16.0 Hz); the 1 3 C NMR (63 MHz) spectrum of this compound was also determined: 5 28.85, 29.93, 31.58, 32.39, 55.67, 147.49. Synthesis and Purification of Diazirine The synthetic apparatus (assembled in a well-ventilated hood) consisted o f a three necked 1 L round-bottomed flask (thermometer, stir bar) fitted with a 500 mL addition funnel, connected through a series of four U-tube traps to a vacuum pump. Each trap could be individually isolated by teflon-sealing glass vacuum stopcocks (1 0 mm bore) and were cooled as follows: a) -35°C (/-propanol slush bath, minimal CO2); b) -80°C (/- propano]/C0 2 ); c) -142°C (methylcyclopentane/liquid N2 slush bath; Vigreux-type inner tube); and d) -196°C (liquid N2). The bath temperatures were verified with a Cu-constantan thermocouple connected to a Varian A-5 recorder. The reference temperatures were ice water (0°Q and liquid N2 (*196°C). Im portant safety note: hazardous procedure. We stress that the vacuum line employed used only teflon-sealing type glass stopcocks. Because of 31 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. die explosive nature of neat diazirine in condensed phase, all glassware was covered with protective wire mesh reinforced heavily with plumber's tape and the entire apparatus was placed behind a protective shield. A face shield and gloves should always be worn by die operator. Extreme caution should be exercised, and we recommend that the reaction scale should not be increased beyond that specified here. Under no circumstances should additional liquid N2 be added to the methylcyclopentane slush bath once diazirine generation or purification is underway, as explosion in this trap, possibly caused by sudden crystallization of solid diazirine below die normal slush bath temperature, may ignite the slush solvent The sodium formate buffer (6.75%, 75 mL), ammonium chloride solution (21.4% , 75 mL) and f-butyl ether (75 mL) were mixed, and added with stirring at 5 °C (ice-NaCl bath) to a three-necked 1 L round-bottomed flask. The sodium hypochlorite solution (0.4 N, 150 mL) was placed in the addition funnel. The entire apparatus including the traps was flushed briefly with Ar. f-Octylazome thine (3.53 g) was then added to the reaction vessel and the system was opened to vacuum (< 1 mm Hg). The entire sodium hypochlorite solution was added to the reaction mixture over 6 min with magnetic stirring. The reaction mixture foamed and its temperature rose to 14°C Five minutes after the reaction subsided (-15 min.), the reaction vessel was isolated from die traps. After evacuation (10 min), the traps were individually isolated and disconnected. The trap at -142°C was connected to the vacuum line for further purification of diazirine through the carbon tetrachloride/dry ice and methylcyclopentane / liquid N2 traps. The purified diazirine was expanded in a 500 mL volume on the vacuum line, which has been described in detail previously (Fig. 2 of ref. 33). 34 Usually, a pressure of 40 mm Hg diazirine was observed (0.05 atm, 1.13 mmol of diazirine). Diazirine is a photolabile gas which decomposes when irradiated at 313 nm with 32 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. a Hg lamp. 22,35 Diazirine was therefore handled under minimal light and stored in the dark. Characterization and Quantitation of Diazirine Diazirine was characterized in die gas phase by IR and UV spectroscopy. A GC method for analysis of diazirine and its possible decomposition products was established. a) IR spectrum o f rfiariring- on die vacuum line, diazirine was expanded (approximately 0.04 atm) into a pre-evacuated cylindrical gas IR cell with NaCl windows (5 cm path length, 20 mL capacity). The IR spectrum (data not shown) was in good agreement with those previously published. 36,37 In particular, characteristic C-H absorptions at 3000-3200 cm*1 and multiple bands at 1610-1660 cm-1 (tentatively assigned to N=N stretching vibrations) were observed. Unlike the reference spectrum, 37 our spectrum showed no CO2 impurity peak at 2300 cm*1 (strong asymmetric stretching vibration). 38 Earlier studies in this group had found no evidence of photodecomposition in the samples when exposed to subdued room lighting over several hours. 39 b) U V spectrum of diazirine- an anaerobic, saturated sodium sulfate solution was displaced from a septum-sealed cylindrical quartz cell (1 cm path, 3 mL capacity) with diazirine from the dual-bulb storage device (0.04 atm in Ar) using a 3 mL plastic syringe (Becton Dickinson). A similar, empty cell was used as reference. The U V spectrum was in good agreement with a published spectrum. 36 In particular, several sharp and regularly spaced peaks (288, 294, 301, 308.5, 313, 316.5 and 322.5 nm) were observed between 282 and 324 nm. The molar absorptivity (e) of 176 liter mole*1 cm*1 at 308.5 nm 36 was used to calculate diazirine concentrations in the gas phase. 33 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. c) A C quantitation of diaririne and its reduction products: diazirine was separated by GC from methane (one of the nitrogenase-catalyzed reduction products), ethylene and acetylene on a 183 x 0.32 cm stainless steel column of Porapak N (Analabs) at 40°C and detected by FI. The He carrier gas flow rate was 60 mL/min. A fast flow rate and low oven temperature were preferred to minimize the possibility of thermal decomposition o f the diazirine samples. Typical retention times for methane, ethylene, acetylene, and diazirine were 26,46,112, and 270 sec, respectively. d) niagjrine puritv: To quantitate the levels of H2, 0 2 , or N2 in diazirine, a 183 x 0.32 cm copper GC column filled with S A molecular sieve was used with TC detection. GC peaks corresponding to H2, 0 2 , and N2 appeared at 40, 60, and 110 sec, respectively (diazirine was not detected). The Ar carrier gas flow was 25 mL/min, the oven temperature 40°C, and the TC filament current 107 mA. It was established by this method that less than 0.4% of O2 or N2 was present in the samples. Synthesis, Purification and Handling o f /rans-Dimethyldiazene. trans-Dimethyldiazene was prepared from 1,2-dimethylhydrazine by the method of Renaud. 40 The product was stored in gas tubes (Pyrex, J 14/20 ground glass joint) kept in a liquid nitrogen refrigerator. All gas handling was done on the vacuum line referred to above. Typical yields of crude product were between 65 and 70%, as determined by the pressure and volume of the expanded gas product at room temperature. For IR spectroscopy, tra/u-dimethyldiazene was expanded into a pre-evacuated gas cell with NaCl windows (74 mm path length, 30 ml capacity) at partial pressures of 0.066 to 0.4 atm. IR spectra obtained from several samples were fully congruent with the published spectrum,41 displaying a prominent ensemble of three strong doublets centered at 2926 cm-1 (symmetric stretching mode of the methyl groups), and singlets at 1440 and 1450 cm-1 34 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. (methyl group bending vibrations). NMR samples (ca. 1 M) were prepared by condensing 1 liter of gas at P-0.02 atm into a 5 mm NMR tube containing 0.6 ml degassed CDQ3 and cooled in a diethyl malonate siush bath (-51.5 °Q . All product spectra showed the expected singlet at 3.67 ppm 42 with impurities at 2.90-2.95 ppm totaling 1%. Crude trans- dimethyldiazene was 99.2 % pure by GC analysis. Purification by vacuum distillation at - 74 °C and 75-100 mm (56% yield) in an apparatus described previously 34 increased the purity to 99.6 % (GC). /ftvu-Dimethyldiazene was quantitated by GC on a Chromosorb P column (80-100 mesh, 150 cm x 3.2 mm) at 0 °C, with a helium flow rate of 40 ml/min; typical retention times were 5.48-5.65 min. rra/u-Dimethyldiazene was used in assays either in an undiluted form or diluted with Ar. It was also delivered as concentrated aqueous solutions with molarities determined by UV just prior to use. trans- Dimethyldiazene at 1 atm was stable by GC and IR for the duration of the experiments (5-6 h) under indoor fluorescent lights. /ra#is-Dimethyldiazene Solubility in Water The water solubility of /ra/u-dimethyldiazene was determined by adding varying amounts of the gas to vented vials (5 ml) containing Ar and degassed water (1.0 ml) which were subsequently incubated at 30 °C. Gas (10 pL) and liquid (1 pL) injections into the gc were used to determine the partitioning of /ra/is-dimethyldiazene between the two phases. The molar amount of /ro/u-dimethyldiazene in each sample and phase was determined using a standard curve. A "reverse" experiment was also performed to verify the solubility results in which a sample of the liquid phase from an equilibrated water/dimethyldiazene system was transferred into an empty, Ar-filled vial, allowed to reach a new equilibrium, and re-analyzed for fra/u-dimethyldiazene partitioning. When stock solutions of higher concentrations (up to 0.91 M) were prepared, the molarity of tra/u-dimethyldiazene was 35 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. determined from the UV spectrum of die solution using an £343 of 25. 43 trans- Dimethyldiazene stability (P = 0.434 atm in argon) in the presence of assay components (see below) was verified by GC analysis. Nitrogenase Proteins. The Avl and Av2 proteins were purified according to a previously published procedure. 44 The protein specific activities were 1000-1600 and 1800-2500 for Av2 and A vl, respectively. The molecular weights used were 60 kDa and 240 kDa, respectively. 3<4 The protein component ratio used in this work is defined as the molar ratio of Av2 to Avl. Acetylene Reduction Assays. This assay is commonly used to measure nitrogenase activity. 45 The method described here and the ones to follow have several elements in common: a) vaccine bottles (21 and 5 mL, Wheaton) sealed with flanged rubber septa (Sigma/Aldrich); b) an ATP generating solution containing adenosine triphosphate (5 |imol, Na2ATP*2 H2 0 , Sigma), creatine phosphate (25 fimol, disodium salt, Pierce), creatine phosphokinase (8 units, Sigma), MgCl2 (5 |im ol, Mallinckrodt), and Hepes buffer (25 pmol, titrated to pH 7.3 with NaOH, Calbiochem); c) a degassing protocol in which vaccine bottles were attached to a vacuum manifold through 22G needles. The bottles were evacuated three times to replace air with Ar, d) the reductant, consisting o f 0.25 mL of 0.08 M dithionite prepared anaerobically; e) an incubation protocol (in this case, 10 min at 30 °C in a shaker bath at 100-200 rpm); 0 a procedure for teiminating the assay: in this case, by injection of 0.25 mL of a 10% TCA solution. The following protocol was used: the ATP-generating solution and buffer (used to normalize all the assay volumes to 1 mL) were added to the vaccine 36 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. bottles which were then sealed and degassed on a manifold. Dithionite was then added. Acetylene (1 mL) purified through an i-propanol/C0 2 trap was added. The reaction was started by addition of dinitrogenase reductase. The reaction mixture was then incubated at 30 °C for 10 min in a shaker bath. A 10% TCA solution (0.23 mL) was added to stop the reaction. For die NH3 assays, the reaction was terminated with 0.1 mL H Q in saturated KIO3.29 Analysis of acetylene and its reduction product, ethylene, was performed using FI detection, with the GC operated at 43 or 30 °C, on a 50 x 0.32 cm stainless steel column of Porapak N eluted with He at flow rate of 30 mL/min. The nitrogenase specific activity was defined as unit-(mg of protein)*1 . A unit of activity was defined as one nmol of ethylene formed per min. Acetylene was used as the internal standard. 21 The ratio of the relative sensitivities of the GC to ethylene compared to acetylene (response ratio) was 1.1 - 1.2. The specific activity was calculated as described previously. 21 Diazirine and /rarts-Dimethyldiazene Reduction Assays. A protocol similar to that used for acetylene reduction assays (see above) was followed. The diazirine partial pressure inside die dual-bulb storage device was usually 0.25 atm in Ar. Volumes of 0.2-1.0 mL were withdrawn with an Ar-flushed disposable plastic syringe and injected into the assay bottles. Similarly, stock solutions of pure trans- dimethyldiazene in H2O were injected into the vials. Reaction times varied between 5-30 min. Reactions were carried out under reduced lighting to ensure the absence of substrate photodecomposition and were terminated as described above. When several reduction products (gaseous or dissolved) were analyzed, the gaseous components (e.g., diazirine, /ra/tt-dimethyldiazene, acetylene, ethylene, methane, or H2) were quantitated first by GC methods. Typically, a gas aliquot (0.05-0.2 mL) was withdrawn from assay bottles and 37 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. simultaneously replaced by an equivalent volume of Ar. The soluble components (e.g., methylamine or ammonia) were analyzed last using HPLC methods. Quantitation of Ammonia and Methylamine Ammonia was detected in early experiments by an indophenol method which was found to be at least four times more sensitive than the Nessler assay for NH3.48 In later experiments HPLC analysis was preferred using a dansyl chloride precolumn derivatization. 29 The HPLC method afforded higher sensitivity with a lowo- limit of detection (O.S nmol NH3/1 T 1 L), while simultaneously providing analysis of any methylamine formed. The same method was very sensitive for methylamine Gower limit of detection of 0.025 nmol/mL). The assay components of the nitrogenase mixture did not cause any serious interference in the HPLC measurements but did contribute a background of 0.1-0.15 nmol NH3 per assay. To ensure reliability, calibration curves were run periodically. Quantitation of H2 and CH 4 H2 evolution was measured by GC as described above (see section on diazirine purity). Standard samples contained 0-3 pmol of H2 gas (Airco), purified through a Deoxo cartridge (Engelhard) to remove traces of O2. The standard bottles also contained 1 mL of nitrogenase assay mixture without the enzyme. After incubation for 10 min at 30 °C, the assay bottles received an injection of 0.1 mL of a 10% (w/v) trichloroacetic solution (TCA) to stop the enzymatic reaction. An aliquot of the gas phase (0.2 mL) was injected into the GC. The lower limit of detection by this method was 10 nmol H2. Quantitation of methane was done on the Porapak N GC system for quantitation of ethylene and acetylene described earlier. A standard line was constructed prior to each experiment by assaying gas aliquots 38 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. from reaction vials incubated with the complete assay system without the substrate, and containing 0-400 nmol CH4. Detection of HD Formation in the Presence of Diazirine. This experiment was designed to test for nitrogenase-catalyzed HD evolution in presence of diazirine and D2 in an H2O-based reduction assay reduction system, following the studies o f Burris. 49 A method had been developed in this group 39 for direct GC separation and detection of D2 and HD based on a modification of the method of Ohkoshi et a l. 30 Varian 3700 GC, equipped with a TC detector, and a 183x0.32 cm copper molecular sieve (S A, 60-80 mesh) column immersed in liquid N2 (-196 °Q was used with H2 as a carrier gas at a flow rate of 25 ml/min. The GC parameters were as follows: injector T 40 °C, detector T 60 °C, column oven temperature off, filament T 150 °C, filament current 266 mA. The HD standard was prepared by slow addition of D2O to UAIH4 in an evacuated flask just prior to use. Determination of D2/HD was done immediately after the nitrogenase reaction was terminated. Assays with gas phase mixtures of 50 % D2 (Airco or Aldrich 99.8 %)/40 % N2/IO % Ar and 50 % D2/5O % Ar served as positive and negative controls respectively. The HD formation experiment had been performed earlier in this group, and the purpose of the present repeat was to verify the previously obtained negative result by utilizing the enhanced integration and data examination capabilities afforded by the Maclntegrator software. Controls and Standards for Diazirine and trans-Dimethyldiazene Reduction Assays Several sets of controls and standards were included in each reduction experiment Bottles contained the complete assay mixture w ithout nitrogenase but with one of the 39 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. following components: a) Ar, b) diazirine gas or /raw-dimethyldiazene, to monitor any non-enzymatic reduction; c) varying amounts of ammonium chloride, to generate an NH3 standard curve with 0-40 ng of NH3; d) methane gas, to generate a CH4 standard curve with 0-400 nmol of CH4; e) H2 gas, to construct a H2 standard curve with 0-32 nmol of H2. In addition, two sets of controls containing die complete assay mixture with the enzyme were used to verify: a) H2 evolution in the absence of exogenous substrates; b) acetylene reduction. RESULTS Aqueous Solubility of Diazirine and Traiis-dimethyldiazene. The aqueous solubility of diazirine was earlier found to be 0.078 M -atnr1.39 Diazirine thus displays sparing solubility in water, and the percent mole fraction in the liquid phase of the total gas originally introduced into an assay vial is small under our conditions (1.5 % in a 21 ml assay vial/1 ml liquid, 8 % in a 5 ml assay vial/1 ml liquid). As a result, at most only a minor adjustment to the initial value of the gas partial pressure was required. The solubility of /ra/u-dimethyldiazene in water was 1.50 M.atm*1 , showing that fra/u-dimethyldiazene has a much higher water solubility than diazirine, nitrogen (0 .0 0 2 M.atm*1 ) or acetylene (0.0035 M.atm*1 ). 51 A plot of molar concentrations of trans- dimethyldiazene versus its partial pressures at 30 °C gave a regression line which passed through the origin with a good linear correlation (r2 = 0.998) over the entire data range (up 40 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. to 0.169 atm frans-dimethyldiazene); distribution of /ra/w-dimethyldiazene between the gas and liquid phases reached equilibrium within <10 min (GC, data not shown). Due to the relatively high aqueous solubility of the rro/u-dimethyldiazene molecule, a substantial percent mole fraction would be transferred to the liquid phase in our assays, lowering the partial pressure significantly. We therefore used the Henry's Law constant we obtained for /rans-dimethyldiazene to calculate its final liquid phase molar concentration (C ) and its final gas phase partial pressure (P , atm) in assay bottles after equilibrium was established. Let V/ = volume (Liters) of rra/u-dimethyldiazene gas at atmospheric pressure P (atm) injected into an assay vial with gas and liquid phase volumes W g and V/ (Liters) respectively, and ng and nt be the actual amounts (moles) of iftz/u-dimethyldiazene present at any moment in the gas and liquid phases respectively. Then at given moment, A *= — n‘ (1), and P is = (2). The initial estimate of n/ using P = P will be too high, since P falls as the solute dissolves. The relative pressure change corresponding to a new estimate is <3). where P' is the new partial pressure obtained from the Henry's Law and the first estimate of n /: using the known value of the Henry's Law constant K for rra/u-dimethyldiazene. Since Henry's Law relates molarity of a gas in a solvent with its momentary partial pressure, to determine P and n/ after dissolution has reached equilibrium one must use 41 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. iterative methods. The value of n/ was allowed to vary (Goal Seek function, Excel 4.0 spreadsheet) until the condition P/P 1% was fulfilled. The values C and P in the assay bottle were then calculated using (4) and setting P = P . No HD Formation in Diazirine Reduction When the nitrogenase-catalyzed diazirine reduction was carried out in the presence of D2 as described in Methods, no HD formation comparable to that observed with 40 % N2 (Sx background HD in 90 min assay) was detected (Figure 1). Triplicate assay vials were used in both N2 and diazirine-containing assays. Nitrogenase-Catalyzed Reduction of /ra/ss-Dimethyldiazene. fra/u-Dimethyldiazene is reduced in the presence of nitrogenase to methylamine, methane and ammonia. Omission of any required component from the enzyme assay (Avl and/or Av2, ATP generator, reducing agent) resulted in detection of background levels of the three products similar to those obtained with diazirine. Time course experiments with /rans-dimethyldiazene concentrations in the liquid phase ranging from 0.17 to 0.318 M showed that the evolution of the three products was linear for up to 35 min with no lag phase observed, therefore the steady-state experiments used a 30 min incubation time to allow for accumulation of the highest possible amounts of products (see below). The Km and Vm values obtained for the three products (Fig. 2) are 0.51 ± 0.0354 M and 21.01 ± 2.04 nmol/mg-min, 0.58 ± 0.0451 M and 19.16 + 1.971 nmol/mg-min, and 0.53 + 0.0477 M and 15.72 + 2.11 nmol/mg-min for methylamine, methane and ammonia respectively. 42 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Inhibition of Hydrogen Evolution and Electron Balance for trans- Dimethyldiazene Reductions. In an attempt to determine the effect of rra/is-dimethyldiazene on nitrogenase- catalyzed proton reduction, /nms-dimethyldiazene reduction products were quantitated together with H2 in a series of assays with variable substrate concentration. The results (Fig. 3) show that /ranr-dimethyldiazene causes a weak inhibition of H2 evolution and that the decrease in die amount of H2 produced corresponds to the amounts of methane, methylamine and ammonia produced within the error of the experiment Effects of Acetylene, Carbon Monoxide and Hydrogen on trans- Dimethyldiazene Reduction. Reduction of /ro/tr-dimethyldiazene is inhibited non-compeddvely by acetylene and carbon monoxide. Preliminary experiments showed that both gases at low pressures (~0.008 atm for C2H2 and -0.001 atm for CO) inhibited the reduction of trans- dimethyldiazene at 0.318 M sufficiendy to impair quantitation of the three products. The estimated inhibition constant of CO for all three products is 0.0008 atm (Fig. 4A). A similar plot (Fig. 4B) was obtained when the reduction was measured in the absence and the presence (0.005 atm) of acetylene, from which an estimated K; of 0.006 atm for all three products was obtained. When, in separate assays, /ra/w-dimethyldiazene (0.17 M) reduction was performed in the absence and the presence of 0 .1 0 and 0.25 atm H2, the changes in the amounts of products evolved were <1.9% (data not shown) which places the lower Ki limit at 3.85 atm. 43 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Weak Inhibition of Acetylene Reduction by /rans-Dimethyldiazene fra/iy-Dimethyldiazene was evaluated as an inhibitor of acetylene reduction. The Lineweaver-Burk plot of the data, shown in Fig. 5, demonstrates that the inhibition mode is non-competitive with a Ki value of 0.093 ± 0.0061 M. No curvature was observed when the slopes and y-intercepts from Fig. S were replotted against the inhibitor concentration (not shown). Both N2 and diazirine are competitive inhibitors of C2H2, and it is possible that the acyclic azo compound has a second, non-specific binding interaction with the enzyme; it may also affect C2H2 solubility. Effect of the Protein Component Ratio on the trans-Dimethyldlazene Reduction Product Distribution Protein component titration curves for the three products were obtained by carrying out the reduction assays at different Av2 to Avl ratios (Fig. 6 ). As evidenced by these results, die ratio methylamine: methane: ammonia depends little on the protein component ratio in the range 3.1-13.62. Attempts to determine the product distribution at Av2 limiting conditions were unsuccessful because of the dramatic drop in the activity below a protein ratio of 2 (Fig. 6). DISCUSSION Experimental Aspects of Diazirine and /rans-Dimethyldiazene as Nitrogenase Probes. Once prepared, purified, and stored over saturated aqueous Na2S0 4 , diazirine is readily and safely manipulated for nitrogenase assays, in which it had been shown to be substrate of the enzyme. Because its molar Km, and Kj values, are quite low and its water solubility as measured in our experiments is ca. 100-fold greater than N2, it is convenient 44 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. to use at low partial pressures in He or Ar, and its major reduction products are readily determined by GC (CH4) or HPLC (CH3NH2, NH3). Nevertheless, its aqueous solubility is not so high as to require more than a modest correction in the final partial pressure in standard assays; the partial pressure is readily verified by GC analysis of a gas phase aliquot. Diazirine is stable to die assay reagents (minus enzyme) at 30 °C With trans- dimethykiiazene, the aqueous solubility increases another 2 0 -fold, and a substantial fraction of total substrate is found in the liquid phase, requiring iterative solutions to find the final partial pressure. It is fortunate that molar concentrations of the substrate can be attained in assay mixtures, as its Km is quite high (see below). Both diazirine and trans- dimethyldiazene are sensitive to UV light, but are handled readily under ambient lighting, it being recommended to minimize exposure in the case of diazirine. Both compounds are chemically stable on prolonged storage in contact with H2O. Relevant Structural Features of Diazirine and /rans-Dimethyldiazene. Before considering the behavior of the two azo compounds as nitrogenase substrates, it is useful to summarize certain of their molecular properties that may be pertinent to their differing interactions with the presumptive active metal center in nitrogenase. A key feature of diazirine is its unique structure as a diazene having the -N=N- group confined within a highly strained, three-membered ring. Sterically, this results in a compact molecule similar to N2 in its NN edge dimension, with the nitrogen atom unshared electron pairs nominally in orbitals oriented cis at an angle of ca. 120° to the NN bond axis, similarly to those in wr-N2H2. The NN distance of 1.228 A in diazirine is close to that expected for a double bond, and the CN distance of 1.482 A is similar to analogous values for acyclic systems. 52 Bonding in diazirine has unusual features, and has been well characterized experimentally 45 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. and theoretically. 35,52-56 The highest occupied MO (HOMO) in diazirine combines the NN lone pair orbitals with a somewhat Walsh-like orbital which is CN bonding and NN antibonding. Electron donation via ligation would thus strengthen the NN bond, and weaken the CN bonds, tending to enhance ring-opening chemistry. The HOMO corresponding to the asymmetric combination o f die two lone-pair orbitals has ca. 30 % CN and ca. 70 % lone-pair character in a 6 -membered diazene, but 6 8 % and 32 % respectively in diazirine. 52 This would be expected to modify the donor ligand properties o f diazirine in its interactions with the unoccupied, acceptor orbitals of site metal center(s), although it evidendy does not preclude formation of 3,3-disubstituted diazirine complexes with Fe2(CO)9, for example, in which the diazirine bridges two Fe atoms, with preservation of the azo —N=N— bond. 53 in other Fe carbonyl complexes and in Ru carbonyl complexes with diazirine the N=N bond is cleaved. 53 In contrast, rro/u-dimethyldiazene is a conventional acyclic azo compound, but the /ra/u-geometry orients the two lone pair orbitals in opposite directions within the molecular plane, thus allowing only one syn interaction with a metal site. Weaker metal coordination by trans vs. cis azo derivatives has been reported. 53 Although diazirine is a polar compound with a dipole moment of 1.59 D, 56 its solubility in H2O is considerably lower than that of /ra/ir-dimethyldiazene, which has zero dipole moment One explanation for the observed difference might be the greater basicity of the more localized two lone pairs of electrons in the acyclic diazene, resulting in enhanced hydrogen-bonding to the protic solvent Differences and similarities in the steric profiles and ligand electron donor properties are summarized in the Scheme below (differences in the LUMOs are probably also relevant but are not represented for simplicity). 46 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. x. S > *ISF=N. / f X # X D iazirine fcans-Dimethyldiazene Comparison o f steric profiles and electron donor properties of diazirine and trans- dimethyldiazene. Solid arrows depict < r (or o-like) electron donation sites. Unfilled arrows symbolize availability of it* electron acceptor sites for filled metal d-orbital backbonding (geometry not specified). The two substrates should also present quite different steric profiles to the enzyme active site. As pointed out already, diazirine is similar in size to N2 for an NN edge-on approach to the site, and will be only slightly more hindered in a CN edge-on approach, owing to the presence of the two methylene hydrogen atoms at one end of the edge. Edge- approach by traras-dimethyldiazene will be made awkward by the protruding methyl groups both geminal and vicinal to either coordinating N. End-on (or perpendicular) approaches to the site could also yield significantly different steric interactions with site moieties neighboring the approach path required for binding and reduction of each molecule. Substrate Reduction Pathways. Azo substrate interactions with nitrogenase can be characterized on several bases: a) the products formed, the maximal velocities for formation of each, and their ratios, corresponding to absolute and relative rates of cleavage for C —N and N=N bonds; b) the Km values for the substrates; c) inhibition by or o f other compounds bound by Avl; d) effects on H2 formation and electron flux. 47 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Beginning with the diazirine reduction, we note that formally this strained-ring diazene could be reductively cleaved in several different ways: 1) exclusive cleavage of both C—N bonds, to generate a (metal-bound) carbene, plus N2, followed by separate reduction of the two fragments. The initial step of this process is analogous to the photochemical dissociation of diazirine. 55 If the N2 were co bound (implying >1 metal coordination site), the reduction products would be CH4 (2e~, 2H+) and 2NH3 (6e-, 6H+). N 2 ase .N = N . 2* < C - N > Nz + :C H 2 6e*/6H + 2e'/2H + 2 N H < C H 4 Diazirine reduction: path 1. Square brackets [ ] enclose bound species. The idea that N2, formed in situ from a precursor substrate, can be retained and reduced at the active site rather than released in free form has precedent from studies on azide reduction. 57 However, this mechanism is not supported for diazirine reduction: it does not account for CH3NH2 production and predicts the wrong NH3:CH4 ratio (2:1, vs. 4:1 observed). Alternatively, it could be considered as one component of a multi-path binding and reduction scheme, wherein a second process accounts for the CH3NH2 and modifies the product ratio to that observed experimentally. However, one might still expect to observe features considered unique to N2, such as inhibition by H2 58 and enzyme- dependent HD formation in D2/H2O assays, 49 some of which were observed for azide reduction to N2. 57 We were unable to detect either phenomenon in diazirine reduction by 48 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. nitrogenase. A further argument against two completely separate and independent reduction processes is the observation that formation of all three products from diazirine exhibits the same Km value, within the experimental error. This argument could of course fail in die face of two processes which coincidentally had very similar kinetic parameters (see further discussion below). W e also have no evidence ( 0«H » • 14*44 4M K O . M O •t.OOB N . m 0.0 M l * CMMI M iN B l T O m T > p * 1 1».7«0 *N- T M M H >H W »V ) • 1« • > • 44U* 100.000 • 44M2 100.000 •0.1 0.0 1 1 « .« 1 * Tfta H ilgM Q iV ) A rM (|iV -M e) A imK -*7 «1>H1 l oo.oao • imoi ioo.o o o 61 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. F igure 2 . Lineweaver-Burk plot of kinetic data for reduction of trans- dimethyldiazene to methylamine (▲), methane ( • ) and ammonia (■ ). The reactions were run for 30 min in the presence o f 0.10 mg A vl at a protein ratio (Av2:Avl) of 6 .6 . 62 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. tf * > 0.225 0.15 0.075 0 6 2 4 0 2 l/jrans-dimethyldiazene, M -l Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 3 . Inhibition of H2 evolution by /rans-dimethyldiazene. The assay conditions were the same as in Fig. 2. The total nanomoles of electron pairs allocated to product formation ("nmol corr.") were calculated based on the relationships 2e~/H2, 3e~ /NH3, 2 e~/CH3 NH2 and le'/CH*. Plots show electron pairs allocated to total product formation from the diazene substrate (0 ), H2 evolution (□ ), and the sum of both (O). 64 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1000 0 » 1 -v Y ^ L- 0 0.1 0.2 0.3 fra/u-Dimethyldiazene, M Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 4 . Inhibition of /raru-dimethyldiazene reduction by CO (A) and acetylene (B). The assay conditions and the plot symbols for the three products were the same as in Fig. 2. Open symbols: control (no CO/C2H2), filled symbols: 0.0006 atm CO (A)/0.005 atm C2H2 (B). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 0 3 0.2 0.3 0.2 0.1 - 2 - 1 0 1 2 3 4 1/rra/tr-dimethyldiazene, M' 1 67 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 5 . Inhibition of acetylene reduction by /ftznr-dimethyldiazene. The assay conditions were the same as in Fig. 2 with exception of the reaction time which was 10 min. The trans-dimethyldiazene concentrations were 0 (# ), 0.05 M (A), 0.08 M (O), 0.09 M (0) and 0.17 M (Q ). 68 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 0.007 0.004 > 0.002 -200 -100 0 100 200 l/zram-dimethyldiazene, atm*1 69 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. F ig u re 6 . Electron allocation to alternative products of rrans-dimethyldiazene reduction as a function of the Av2 :Avl ratio. Reactions were carried out in presence of 0.10 mg A vl and variable amounts o f Av2. The incubation time was 30 min, and the trans- dimethyldiazene concentration was 0 3 1 8 M. Plot symbols for the three products are the same as in Fig. 2. Inset shows the dependence of the product ratios CH4:NH3 (O) and CH3NH2:NH3 (A) on the ratio Av2:Avl. 70 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 30 20 10 1.5 Av2/Avl 0 5 10 0 A v 2 / A v l 7 1 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER 3 NITROGENASE-CATALYZED REDUCTION O F CIS - DIM ETH YLDIAZENE: A NEW CHEMICAL PROBE OF THE ENZYM E MECHANISM Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. INTRODUCTION Diazene (N2H2, diimide) has been postulated to be a bound 2e~ intermediate along the nitrogenase-catalyzed dinitrogen reduction pathway. 1 The observation that acid- or base-quenched nitrogenases fixing N2 release N2H4 2 demonstrates that some form of enzyme-bound reduced intermediate retaining a nitrogen-nitrogen bond has a significant lifetime. To date, there has been no evidence that nitrogenase can reduce an unstrained, cis- azo (-N=N-) group. Diazene itself can be readily generated in situ at neutral pH, but rapidly decomposes to N2, H2 and N2H4 3 and early efforts to detect an interaction with nitrogenase were unsuccessful. 4 Recently, we showed that the simplest cfr-dialkyldiazene, cis-dimethyldiazene 2 , can be readily prepared in pure form as a stable aqueous solution up to 0.5 M by phase-specific photochemical isomerization of the trans isomer 1 under phase- transfer conditions. 5 We have now evaluated 2 as a potential substrate and inhibitor of nitrogenase from A. vinelandii. EXPERIM ENTAL Preparation, characterization and handling of cfr-dimethyldiazene have been described. 5 A. vinelandii nitrogenase proteins were purified and characterized according to published procedures. 6 As measured by the acetylene reduction assay, Avl and Av2 had specific activities of 1800-2200 and 1700-1880 nmol C2H4/min.mg respectively. Nitrogenase-catalyzed reduction assays were performed as described elsewhere. Sodium dithionite (Sigma) was used after twofold recrystallization from 0.1 M NaOH - methanol under anaerobic conditions. 7 Unless otherwise specified, the assay mixture (final liquid 73 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. volume of 1.0 ml in a 5 ml vial, pH 7.3) contained 0.10 mg Avl and 0.16S mg Av2. Analyses for reduction products were performed as described previously.8* 9 CH4 as a reduction product was identified by mass spectrometry on a Hewlett-Packard S989A MS coupled to an HP 5965B GC. Assays at variable Av2 : Avl ratio were done at constant Avl concentration. Reductions of m-dimethyldiazene in the presence of D2 were performed using an assay cocktail volume of 2 ml containing 0.40 mg Avl and 0.624 mg Av2 at a ratio of 6.2. Dinitrogen reduction assays performed in the presence of D2 (99.8 atom % D, Aldrich), in which the initial partial pressure of N2 was varied over 0 - 0.40 atm (balance: 0.50 atm D2 and Ar, 1 atm = 101,325 Pa), served as positive controls. Negative controls included assay vials without N2 or cu-dimethyldiazene, or without enzyme. After 30 min incubation, the entire head space gas sample (3 ml) was removed via displacement with saturated Na2SC >4 solution and injected into the GC for HD/D2 analysis.10 Data were collected and processed on an Apple Centris 610 computer using the Maclntegrator I hardware/software package (Rainin Instruments). Reagents for reductions of cu-dimethyldiazene in D2O were procured or prepared as described elsewhere. 11 Assays were performed in 10 ml vials with a liquid phase volume of 3 ml containing 12 mg Avl and 19.8 mg Av2 at a ratio of 6 .6 . After 100 min, reductions were stopped by injection of 100 pL 20 % DC1 in D2O (Aldrich), and a 5 ml head space sample was immediately transferred (CCI4 displacement) into a vigorously shaken 10 mm NMR tube containing 3 ml degassed CCI4 pre-cooled in ice water. The resulting sample was analyzed for methane isotopomers by (270.129 MHz) and 2H (41.467 MHz) NMR on a Bruker WP-270SY instrument. Calibration samples of CH4 (99.95 %, Aldrich) and CD4 (99%, Cambridge Isotope Laboratories) were obtained 74 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. commercially, whereas CH3D, CH2D2, and CHD3 were prepared by Zn-Cu contact couple-reduction of CH3I, CH2Br2 , and CHBr3 , respectively following a published method. 12 After passage through a 0.2 pm filter and an anion exchange resin (Dowex 1X8, 2 g) to remove precipitated proteins and the bulk of ATP, ADP and creatine phosphate, the liquid phase from the reduction assay was analyzed by NMR on a Broker AM-360 FT NMR instrument at 360.135 MHz. The same sample was subsequently concentrated in vacuo to remove ca. 95 % of D2O, and diluted with H2O for analysis by 2H NMR. Calibration samples included 50 mM CH3NH2 (40 % aq. solution, Aldrich) in the complete D2 0 -exchanged assay mixture minus the nitrogenase components, and 100 mM CD3ND2.DCI (98 %, Aldrich) in H2O . Analysis of the isotope labeling pattern by GC-MS as described above for the methane identification was unsuccessful. RESULTS Reduction of cis-Dimethyldiazene: General Features. cu-Dimethyldiazene was reduced by A. vinelandii nitrogenase under standard assay conditions ([Av2] > [Avl]) to methylamine, methane and ammonia. The gas phase GC peak was first identified by its retention time compared to a standard of CH4 (99.95 %) incubated in a vial containing the complete assay system except the reducible substrate. In a separate experiment, 0.15 M cu-dimethyldiazene was incubated for 2 hr in an assay mixture containing increased concentrations of enzyme and ATP generator in order to maximize product A headspace gas aliquot (3 ml) was then injected into a GC-MS spectrometer (Materials and Methods) and the eluted single gas product peak was identified as methane by its m/z 16,15 and 14 peaks. No additional products were detected in the gas phase by either conventional GC or GC-MS. The two liquid phase reduction products, 75 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. NH3 and CH3NH2 , were identified by matching their HPLC retention times with internal or external standards. 8 When authentic CH3NH2 was incubated with the complete nitrogenase assay system in the absence of cu-dimethyldiazene, no conversion of CH3NH2 to CH4 and/or NH3 was detected. The reduction of c/r-dimethyldiazene was shown to be ATP- and DT- dependent by running the respective controls lacking these reagents. Incubation (30 min) of 0.1 M substrate with die assay system in the absence of ATP generator or DT resulted in production of < 1 nmol of each product Similar results were obtained when Avl, Av2, or the substrate itself were omitted. Kinetic Patterns of the Reduction Two preliminary sets of experiments were performed in order to establish optimal assay parameters for Michaelis-Menten kinetics. In the first set monitoring product evolution as a function of time revealed that over the 0.045-0.15 M substrate concentration range used, time courses for formation of CH4, NH3 and CH3NH2 were linear during at least 35 min with no apparent lag period (Figure 1). Therefore, steady-state assays were run for 30 min to ensure maximum product formation. In a second set of experiments, die assays were performed in the presence of varying amounts of enzyme at a constant Av2 : Avl ratio of 6 .6 . As the previously noted dilution effect 13 was observed at <0.25 pM Avl, subsequent kinetic studies used an Avl concentration of 0.42 pM (corresponding to 0.1 mg Avl per assay vial). Data from assays with variable initial substrate concentrations (0.015 -0.18 M) were then plotted in Lineweaver-Burk format (Fig. 2a) to yield the basic kinetic parameters of the dimethyldiazene reduction. Data were also plotted in the Hanes- Woolf (Fig. 2b) and Eadie-Hofstee formats to verify linearity (Fig. 2c). Under Avl- limiting conditions (Av2 : Avl = 6 .6), the Km values obtained for the three products were 76 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 59.5 mM, 60.1 mM and 60.6 mM for CH3NH2, CH4 and NH3 respectively. The corresponding Vmax values were 76.1 nmol CH3NH2/min.mg Avl, 47.8 nmol CKj/min.mg Avl and 46.8 nmol NH3/min.mg Avl. From the latter three values, the following product ratios were calculated: 1.6 CH3NH2 :1.0 CH4 : 1.0 NH3. Product ratio dependence on the total electron flux was studied by performing the assay in the presence of different protein component ratios (Av2: Avl in range of 0.5 -20.1). At Av2 : Avl > 5, the relative rates for the three products remained virtually unchanged (Fig. 3). Reliable measurements of Vmax at low Av2 : Avl ratios (<2) were prevented by the low rate of product formation under these conditions. Effect of cis-Dimethyldiazene on Acetylene Reduction and Hydrogen Evolution. Acetylene (0.005-0.02 atm) reduction assays were carried out in the presence of varying amounts of cu-dimethyldiazene (0-0.16 M). A Lineweaver-Burk plot of the data revealed competitive mode of inhibition (Fig. 4). Replots of the respective slopes and KmW for the three datasets vs. the inhibitor concentration showed no curvature and yielded a Ki value of 126 mM m-dimethyldiazer.e. In another experiment, nitrogenase- catalyzed ar-dimethyldiazene (0-0.28 M) reduction products were quantitated together with the hydrogen evolved. Within the error of the experiment, the sum of all reduction products (including H2) normalized per 2H+/2e* was equal to the amount of hydrogen produced in the absence of m-dimethyldiazene (Fig. 5). The achievement of complete electron balance indicates that no major reduction products have been unaccounted for, and the absence of substrate inhibition of total electron flux. It would be premature however, to assign any electron allocation coefficient to this substrate because of the lack of data beyond the S = 4Km point 77 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Inhibition of cis-Dimethyldiazene Reduction by C2 H 2 , CO and N2* Acetylene and carbon monoxide are frequently employed chemical probes for the nitrogenase active site. We therefore carried out a series of experiments in which the effects of C2H2 (0-0.00875 atm) and CO (0-0.0015 atm) on the enzymatic cfr-dimethyldiazene (0.045-0.18 M) reduction were studied. Both compounds were shown to be non competitive inhibitors of cu-dimethyldiazene with Ki values for C2H2 of 0.0067 atm (NH3) (Fig- 6), 0.0069 atm (CH4) and 0.0070 atm (CH3NH2); the corresponding values for CO were 0.00097 atm (NH3), 0.00099 atm (CH4) and 0.0011 atm (CH3NH2) (Fig. 7). Similar Ki values have been reported for the CO inhibition of other nitrogenase- catalyzed reductions. l4>15 cu-Dimethyldiazene reduction was competitively inhibited by N2 with Kj values of 0.117 atm (CH4) and 0.119 atm (CH3NH2) (Fig. 8). Absence of HD Evolution from Reduction Assays in D2 Atmosphere and Insensitivity to H 2. We previously reported9 that, like all other known alternative nitrogenase substrates, the strained-ring azo compound diazirine did not cause HD formation in an aqueous fixing system under D2. When triplicate assays were performed with cis- dimethyldiazene concentrations of 0.169 M and 0.084 M as described in Methods, no HD formation (<20 nmol) was detected (Fig. 9b). A simultaneous GC assay for methane indicated that the enzymatic reduction had proceeded to the extent expected from the Km/Vm values and the substrate concentration used. Positive controls with N2 (Fig. 9a) showed HD (ca. 2000 nmol) at levels consistent with previously published results. 16 It was therefore concluded that, within the limits of detection (ca. 20 nmol), no HD formation is observable during the nitrogenase-catalyzed reduction of cfr-dimethyldiazene. Substrate 78 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. reductions performed under 100 % H2 (99.99+ %, Aldrich) atmosphere failed to reveal any detectable inhibition caused by hydrogen. In (unreported) preliminary experiments, H2 inhibition was observed, but this result could not be reproduced when the H2 source was changed, and was apparently due to N2 contamination in the H2 originally supplied to us. Deuterium Incorporation into the cis-Dimethyldiazene Reduction Products. Deuterium labeling of the as-dimethyldiazene reduction products produced in D2O- fixing systems should provide some insight into the mechanism of C-N bond breaking and formation of CH4 and CH3NH2. Attempts to analyze the methane product labeling pattern by GC-MS at 20 eV were unfruitful, but the different methane isotopomers were readily distinguished by FT and 2H NMR using a reference mixture of CH4 , CH3D, CH2D2, and CHD3 obtained as described in Methods (Fig. 10A, spectrum a). The availability of a *H NMR spectral window for methylamine detection (peak at 5 1.9 - 2.3 ppm, depending on pH) in the liquid assay phase was established by analysis of a mixture containing all assay components except the proteins. Reduction of 0.22 M c/s-dimethyldiazene in D2O followed by analysis of the gas phase by NMR revealed CH3D as the only reduction product within the limits of detection (<5-8% of other isotopomers). The identity of die product was verified by comparing the chemical shift and multiplicity of the peaks in both the *H and 2H spectra (Fig. 10). The triplet in the proton spectrum correlated with the quartet in the 2H, proton-coupled NMR, both yielding coupling constants (Vhd) of ca. 1.9 Hz. In the liquid phase, CH3ND2 was the only methylamine isotopomer found (<5 % other products). These results indicate that no scrambling of hydrogen atoms takes place in the 79 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. substrate methyl group during the enzymatic reduction, pointing to a concerted C-N bond scission step. DISCUSSIO N Demonstration of els -N=N- Binding by Nitrogenase. The ability of nitrogenase to bind an unstrained cis-N=N- bond, as in cis-H-N=N- H, has not been previously demonstrated. Stabilization of H-N=N-H by metal complexadon has been argued as a critical event in nitrogenase catalysis, 17 as N2H2 represents an energy maximum in the thermodynamics of a 2e~/2H+ reduction sequence: N2 — > [N2H2] -* [N2H4] — » 2NH3, where [ ] surround species not in equilibrium with the external environment Our results show that m-dimethyldiazene is bound by the enzyme, and reduced by it at both the -N=N- bond (to CH3NH2) and CH3-N bond (to CH4 + Pairing of the CH4/NH3-forming processes separately from the CH3NH2-forming process is suggested by the similarity in the Vm values for methane and ammonia and by the independent variation of CH4/NH3 vs. CH3NH2 at low [Av2]:[Avl] ratios. An equivalent process for site-bound C&-N2H2 would give: NH3). N=|:N, 4H+/4e' 2CH3NH2 H3C / \ c h 3 h3c c h 3 + c h4 80 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. / N=|=N^ 4H;74e ► 2NH3 H H However, for the reduction of oms-dimethyldiazene, we considered a different pathway: At low electron flux, this would suggest that CH4 would be favored over the other two products. If the process leading to a bound N2H2 were to occur and if that process was coupled to H2 binding, we might expect to detect its presence by H2 inhibition and HD formation; we did not. The sensitivity of our experiment vs. N2 as substrate can be estimated based on the values for the reduction efficiencies (see later) of cis- dimethyldiazene (11 %) and N2 (75 %), and the fact that during the HD formation experiments, both substrates were present at levels near 2Km. At these conditions, the reduction of c/s-dimethyldiazene could cause formation of HD in the amounts ca. 7-fold less than the ones obtained from N2, or ca. 300 nmol (see Results). Since our negative results have an upper limit of 20 nmol HD, the overall sensitivity of this experiment is ca. Kinetic Equivalence o f cis-Dimethyldiazene and N2 Given the formal process of converting N2 (sp-hybridized N) to NH3 (sp3- hybridized N) via a diazene species (sp2-hybridized N), 2, which, like diazene itself, has h3c CH 4 + [CHg-N=NH] CH3NH2 + NH3 15: 1. 81 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. sp2-hybridized N atoms, might be expected to bind similarly to the active site metal cluster. N2 competitively inhibits reduction of 2 , like it does acetylene, consistent with their binding to the same state of the enzyme. Further supporting this similarity of the two substrates is the observation that 2 , like N2 itself, is non-competitively inhibited by C2H2. Other Insights into Binding: Electronic and Steric Effects. It is instructive to compare 2 with its trans-isomer 3, with diazirine 4 and with C2H2 as nitrogenase substrates. C2H2 is reduced by wild-type enzyme only to the alkene product, ethylene. 18 Why is no ethane, or CH4, formed? Our results for 2 vs. • % C2H2 indicate a decisive role for the lone pair electrons of xN=Nv (and by extension, • • • • H2N—NH2 ) in binding to the active site, as against the Jt bonding elections in the substrates. Secondly, the Km value of 2 (ca. 60 mM) falls between the Km of N2 (0.1 mM) or the sterically compact 4 (ca. 0.08 mM) and that of the sterically awkward trans isomer 3 (ca. 550 mM). This clearly points to a strong steric influence on Km, suggesting that despite the kinetic complexity of nitrogenase, Km is an estimate of Ks, the substrate dissociation constant. The reduction efficiency of cis-dimethyldiazene can be measured from the observed product Vm values. Using the assumed stoichiometries of 4e'/4H+ for reduction to two methyl amine product molecules, and 8e~ /8 H+ for the reduction to (2CH4 + 2 NH3), for [Av2] > [Avl] a maximum of 246 nmol (2e'/2H+)/min can be diverted to reduction of this substrate, i.e. efficiency is 11 %, compared to N2 (75 %) and C2H2 (ca. 100 %). Interestingly, this value lies between the reduction efficiencies of /ram-dimethyldiazene {sS %) and diazirine (14 %). The similarity of the corrected Vm's of the unstrained and 82 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. strained-ring cis-azo substrates, compared to their Km value differences which represents a factor of 102, shows that the reduction efficiency reflects a different phenomenon than binding differences due to steric effects, and is favored by a same-site accessibility of the substrate N atoms. Conversely, the Vmar values for 2-4 are rather similar (in the 15-80 nmol/mg.min range), and all are lower than Vmax for N2, notwithstanding the tremendous difference in thermodynamic reactivities between the inert N2 molecule, the relatively reactive unstrained azo isomers, and the highly energetic diazirine. This reinforces the idea that kcat for nitrogenase is relatively indifferent to substrate energy, provided it fits into the active site. Having both vicinal N lone pairs oriented on the same site of the plane perpendicular to the molecular plane is important for distinguishing 2 from 3, but what about 2 vs. 4, where the lone pair symmetry is comparable? Here it can be noted that the lone pairs of 4 are substantially delocalized with the Walsh orbitals involved in forming the "bent" C—N single bonds, and the simple Lewis structures are misleading. In Fig. 11, we compare the electron potential isodensity surfaces of N2 with those of cis- and trans- diazene, the corresponding dimethyl diazene isomers, and, standing in lieu of an alternative hydrazido [2 ] intermediates N2H2 tautomer (a [4*] hydrazido intermediate would be comparable in overall dimensions). The graphics include information about the polarity of each molecule, unperturbed by its interaction with a binding site. It can be seen that the increase in cavity dimension along the N-N bond from N2 to the N2H2 isomers is not large and could be accommodated by a 4.5 A cavity. Diazirine entering the cavity N-N edge-on is similarly accommodated. However, the methyl groups of 2 and 3 increase the crosswise dimension of these molecules by nearly 2 A. In going from diazirine to 2 to 3, the Km increases by 103 and then by 9, representing +4 and +1.3 kcal/mol respectively in the binding energy (we assume Km % Ks). If this result is compared with the series: 83 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. C2H2/cyclopropene/l-propyne/2 -butyne, then the Km results are 0.25 mM, 0.1 mM, 30 mM, and ca. 100 mM. By contrast, cis and trans 2-butene which are isosteres of 2 and 3 are not detectably bound. Thus the site has a degree of flexibility (as if it were completely rigid, much larger steric effects would be predicted), but is still limited in this respect - it is a jaw able to be opened only so wide without intolerable strain. If the site were indeed optimized to stabilize N2H2, significant binding of N2H2 analogs might be expected. Such binding might be offset by the unfavorable steric interactions discussed above. A characteristic reaction of cis-dimethyldiazene in environments of low dielectric Held strength, such as weakly polar or nonpolar solvents, is its spontaneous tautomerization to formaldehyde N-methylhydrazone. If formed in or near the nitrogenase active site, for example in a region where solvent H2O is excluded, such a compound might be expected to undergo 2e~/2H+ reduction to methane and methyl hydrazine, as shown. N— H2C=N— —-/4€- c h 44 + n h2 n h c h3 H3C CH3 (site) \ 4 2 3 This process requires incorporation of two D atoms into the methane produced by enzymatic reduction in P 2O, but only 1 D was found. We conclude that no such process is involved in the reduction of c/s-dimethyldiazene, and therefore no evidence is provided for a nonpolar binding region at the active site. In crystalline MoFe proteins, the FeMo-co clusters are 1 0 A or deeper below the surface but in a hydrophilic environment, with 12 H2O molecules clustered about the homocitrate end. 19 By contrast, P clusters are in a hydrophobic environment providing evidence against a direct reductive role for these clusters. 84 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. A nitrogenase-bound N2H2 reduction intermediate corresponding to a diazene isomer has been considered for many years, 4,20,21 but unlike hydrazine (shown to be a substrate, and also released from acid- or base-quenched N2-fixing nitrogenase), there is no evidence available on whether N2H2 is bound as an intermediate of N2 reduction, or can be bound as a substrate. Past efforts to demonstrate the latter yielded negative results. 4 Diazene is readily generated in situ in aqueous solutions at room temperature, but undergoes disproportionation reactions 2N2H2 -► N2 + N2H4 N2H2 -► N2 + H2 forming respectively two substrates, and a substrate and a reduction (H2O) product; the formed substrates all give the same product expected from direct N2H2 reduction, namely NH3. Our demonstration that unstrained as well as strained alkyl diazenes of greater size than the N2H2 parent are reduced by nitrogenase indicates that diazene itself should interact with the enzyme. In reducing N2, the catalytic site faces a progressively changing set of challenges: a) the basicity of N2 is extremely low, but can be enhanced by oxidative addition to a low- valent metal atom. In addition, the substrate is nonpolar. Two conceivable strategies for activation are: 1) polarize N2 by interaction with an asymmetric site leading to "Chatt" type asymmetric reduction via M=N + NH3. This is supported by the recent ZINDO calculation analysis of an interior FeMo-co binding site by Zerner. 22 2) Use a (semi) symmetrical site, leading to a diazene-like intermediate stabilized by a surrounding matrix of bonding metal orbitals and/or hydrogen bonds, as suggested for diazene-transition metal complexes. 23 Possible 2e" N2H2 intermediates are highly unstable in the absence of metal atom 85 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. coordination. In diazene, the lone pair elections have reoriented by 60° from the N—N bond axis, and their basicity, although still low, has increased. One jt* orbital remains to accept electrons from a metal atom d orbital of matching symmetry. At a 4e~/4H+ level of reduction (H2N—NH2), the intermediate is quite basic and presumably could be in an H2N—NH3+ state. All it* orbitals are fully occupied. At the 6 e~/6 H+ level of reduction, NH3 would undoubtedly be protonated (pKa of NH4+ = 9.5), leading to less affinity for the site (no interaction other than electrostatic now available) and is expelled. Considering that CH3NH2 is a severalfold-stronger base than NH3, it should form CH3NH3+ in the site. Evidently, this species is no longer bound, so further reduction is prevented and the methyl ammonium product released. However, the unique cleavage of the C-N bond in cis- dimethyldiazene, occurring with a transfer of a single H" (or H+) to the methyl group evidently generates an intermediate capable of further reduction to NH3. Proposals for the N2 binding site have included the interior FeMo-cavity encompassing the entire substrate molecule, interactions of a partly penetrant N2 with interior cavity Fe atoms, cluster surface Fe atoms, and the Mo atom. 22,24-26 Arguments from studies with mutant nitrogenases have suggested multiple binding sites (or different modes of binding to the same site), following similar suggestions of structurally congruent sites differing in oxidation state, with N2 uniquely able to react solely with the most reduced site. 27 Various model chemistries have been adduced as evidence for a number of these conjectures. In particular, with respect to bound intermediates on ^-reduction pathways, Mo (and related W) chemistry has been developed to demonstrate the feasibility of an M(-H)-N-NH2 - like intermediate.28 Alternatively, the objection to N2H2 as an intermediate on the basis of its particularly great instability has been refuted most recently 86 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. by Sellman who has demonstrated significant metal complex - stabilization of this species 2 3 . In conclusion, c/s-dimethyldiazene has been shown to be catalytically reduced by A. vinalandii nitrogenase with bond cleavage taking place at both N=N and C-N sites, thereby making it the first nitrogenase substrate containing unstrained cis-azo bond. Although displaying some kinetic similarities with N2, such as reciprocal inhibitory relationship with C2H2, cu-dimethyldiazene still differs significantly from the natural substrate in a number of ways, including the inability to support HD formation and insensitivity towards H2. The value of this compound as a mechanistic tool was further explored in the next chapter. REFERENCES (1) Newton, W. E.; Bulen, W. A.; Hadfield, K. L.; Stiefel, E. I.; Watt, G. D. In Recent Developments in Nitrogen Fixation; Newton, W. E., Postgate, J. R., Rodrigues-Bamieco, C., Eds.; Academic Press: New York, 1978. (2) Thomeley, R. N. F.; Eady, R. R.; Lowe, D. J. Nature 1978, 2 7 2 , 557-558. (3) Hunig, S.; Muler, H. R.; Thier, W. Angew. Chem. Intl. Ed. Engl. 1965, 4, 271- 280. (4) Burris, R. H.; Winter, H. C.; Munson, T. O.; Garcia-Rivera, J. In Non-Heme Iron Proteins: Role in Energy Conversion; San Pietro, A., Ed.; Antioch Press: Yellow Springs, OH, 1965, p 315-321. (5) Simeonov, A. M.; McKenna, C. E. J. Org. Chem. 1995, 60, 1897-9. (6 ) McKenna, C. E.; Nguyen, H. T.; Huang, C. W.; McKenna, M. C.; Jones, J. B.; Stephens, P. J. In From Cyclotrons to Cytochromes (Si. D. Kamen Symposium); Kaplan, N. O., Robinson, A. B., Eds.; Academic Press: New York, 1982, p 397- 416. (7) McKenna, C. E.; Gutheil, W. G.; Song, W. Biockem. Biophys. Acta 1991, 1075, 109-117. 87 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. (8 ) Bravo, M.; Eran, H.; Zhang, F. X.; McKenna, C. E. Anal. Biochem. 1988, 175, 482-491. (9) McKenna, C. E.; Simeonov, A. M.; Eran, H.; Bravo-Leerabhandh, M. Biochemistry 1996, 35, 4502-14. (10) Ohkoshi, S.; Fujita, Y.; Kwan, T. Bull. Chem. Soc. J. 1958, 31, 770-1. (11) McKenna, C. E.; McKenna, M.-C.; Huang, C. W. Proc. Natl. Acad. Sci. USA 1979, 76, 4773-7. (12) Anet, F. A. L.; O'Leary, D. J. Tet. Lett. 1989, 30, 2755-8. (13) Thomeley, R. N. F.; Lowe, D. J. Biochem. J. 1984, 224, 903-9. (14) Hwang, J. L.; Chen, C. H.; Burris, R. H. Biochim. Biophys. Acta. 1973, 292, 256-270. (15) Burris, R. H. In A Treatise on Dinitrogen Fixation', Hardy, R. W. F., Ed.; John Wiley & Sons: New York, 1979, p 569-604. (16) Li, J.; Burris, R. H. Biochemistry 1983, 22, 4472-80. (17) Sellmann, D.; Soglowek, W.; Knoch, F.; Moll, M. Angew. Chem. Int. Ed. Engl. 1989, 28, 1271-1272. (18) Hardy, R. W. F. In A Treatise on Dinitrogen Fixation-, Hardy, R. W. F., Bottomley, F., Burns, R. C., Eds.; Wiley-Inlerscience: New York, 1979; Vol. I and II, p 515-568. (19) Bolin, J. T.; Ronco, A. E.; Morgan, T. V.; Mortenson, L. E.; Xuong, N.-h. Proc. Nat. Acad. Sci. USA 1993, 90, 1078-1082. (20) Yates, M. G. In Biological Nitrogen Fixation; Stacey, G., Burris, R. H., Evan, H. J., Eds.; Chapman & Hall: New York, 1992, p 685-735. (21) Burgess, B. K. In Advances in nitrogen fixation research; Veeger, C., Newton, W. E., Eds.; M. Nijhoff/W. Junk: Pudoc, Wageningen, 1984, p 103-114. (22) Stavrev, K. K.; Zemer, M. C. Chem. Eur. J. 1996, 2, 83-7. (23) Sellmann, D. In Molybdenum Enzymes, Cofactors, and Model Systems; Stiefel, E. I., Coucouvanis, D., Newton, W. E., Eds.; American Chemical Society: Washington, DC, 1993, p 332-345. (24) Chan, M. K.; Kim, J.; Rees, D. C. Science 1993, 260, 792-794. (25) Henderson, R. A. J. Chem. Soc. Dalton Trans. 1995, 503-11. (26) Leigh, G. J. Eur. J. Biochem. 1995, 229, 14-20. 88 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. <27) S eW 1 S n W n E p R ?be^ K' : “ S r C sh “ - J ' Cantwell, J. S.; Thrasher, K L S r f S ' o I P U ?* Science and biotechnology in Agriculture• pWmnerfield, R. J., Ed.; Kluwer Academic Publishers; Dordrecht, 1995; Vol. 27,’ (28) ^ ^ L ML T l ^ m . m i r 7 P- K : * * * * “ • C - E : Coucouvanis, D. 89 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 1 . Time course data for reduction of cis-dimethyldiazene to methylamine, methane and ammonia. The reactions were run in the presence of 0.10 mg Avl at a protein ratio (Av2:Avl) of 6 .6 . 90 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 40 3 0 n m o l C H 4 n m o l M e-N H 2 n m o l N H 3 0 5 1 0 1 5 2 0 2 5 3 0 3 5 4 0 4 5 5 0 5 5 6 0 6 5 7 0 T, min 91 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 2. Lineweaver-Burk (a), Hanes-Wolf (b), and Eadie-Hofstee (c) plots of kinetic data for reduction of cis-dimethyldiazene to methylamine (▲), methane ( • ) and ammonia (■). The reactions were run for 30 min in the presence of 0.10 mg Avl at a protein ratio (Av2tAvl) of 6 .6 . Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. a) 0.2 0 . 1 5 0.1 0 . 0 5 0 -20 1/S, M -1 0 2 0 4 0 6 0 8 0 93 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. b) 0 . 0 0 4 0 .0 0 3 3 0.002 0.001 0 . 0 2 5 0 . 0 5 0 .0 7 5 - 0 . 0 5 - 0 .0 2 5 0 S c) 6 0 5 0 4 0 3 0 20 10 0 2 5 0 7 5 0 1000 5 0 0 0 V/S 94 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 3. Electron allocation to alternative products of c/s-dimethyldiazene reduction as a function of the Av2:Avl ratio. Plot symbols for the three products are the same as in Fig. 2. 95 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 80 A « ■ 1 « l 2 4 0 - 3 i 2 0 - 0 5 10 1 5 20 A v 2 : A v l C o m p o n e n t R a t io 96 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. F ig u re 4 . Inhibition of Q R , reduction by cv-dimethyldiazene. Reactions run for 10 min. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. l / S A . - I 1 /S A , 0 . 0 3 M I 0 . 0 0 3 I / S A , 0 . 0 6 M I s < M 1 /S A . 0 . 1 3 M I o 9 a J 1 /S A , 0 . 1 6 M I 0.002 0.001 2 5 0 - 5 0 1 5 0 200 - 1 5 0 -100 100 -200 0 5 0 1 /atm C2H2 96 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 5 . Inhibition of H2 evolution by m-dimethyldiazene. The assay conditions were the same as in Fig. 2. Hots show electron pairs allocated to total product formation from the diazene substrate, H2 evolution, and the sum of both. 99 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1 5 0 0 H 2 , n m o l e - / m g - m i € > 3 N H 3 + C H 4 f 2 C H 3 N H 2 2 H 2 + 3 N H 3 + C H 4 + 2 C H 3 N H 2 n 1 5 0 0 0 .3 0.2 0 0.1 cfr-dimethyldiazene, M 100 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. F ig u re 6 . Inhibition o f cis-dimethyldiazene reduction to ammonia by C,H2. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 0 . 1 5 - C 2 H 2 + 0 .0 0 3 7 5 a im C 2 H 2 + 0 .0 0 5 a t m C 2 H 2 + 0 .0 0 7 a tm C 2 H 2 0.1 + 0 . 0 0 8 7 5 a tm C 2 H 2 0 . 0 5 3 0 -20 -10 10 20 0 1/c/s-dimethyldiazene, M'* 102 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. F ig u re 7 . Inhibition o f methylamine formation by CO. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. C O X Z C O S £ -C O + 0 .0 0 0 5 a tm C O + 0 . 0 0 1 a tm C O 0 . 0 6 + 0 .0 0 1 2 5 a tm C O + 0 .0 0 1 5 a tm C O 0 . 0 4 o.o: -20 -10 10 20 3 0 l/cti-dimethyldiazene, M -1 104 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. F ig u re 8 . Inhibition of cz's-dimethyldiazene reduction to methane (A) and methylamine (B) by N2. Partial pressures of N2 were 0 (O ), 0.05 (♦ ), 0.10 ( • ) , 0.13 (A), and 0.16 atm (■ ) respectively. 105 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. m 0 . 0 4 5 0 0 . 0 1 5 1/S, M*1 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 9. Absence of HD formation in the reduction of 0.169 M cis- dimethyldiazene in 50 % D fc atmosphere (b). Shown for comparison are the formation of HD in a positive-control assay with 0.2 atm N2 (a) and the background level of HD obtained from an injection of 3 ml P 2 used in the assays (c). 107 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. HD T T T T T T ▼ 9.0 19.0 min 20.0 9.0 T T ▼ ▼ 24.0 9.0 106 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. F igure 1 0 . *H (A) and 2H (B) NMR analysis o f the gas phase of cis- dimethyldiazene reduction in P 2O. The proton spectra show the standard equimolar mixture of methane-do-3 (a, 360.135 MHz, 640 acquisitions) and the assay sample (b, 270.129 MHz, 40 acquisitions). 2H spectra show a standard mixture of ca. 1:2 CH3D and CD4 (a, proton-decoupled, 41.467 MHz, 40 acquisitions), and the assay gas sample (b, proton- decoupled, c, proton-coupled; 41.467 MHz, 9000 acquisitions). Small differences in the chemical shift values are due to the samples being run unlocked except for the L H standard (Aa), which was run locked in a CCI4 solvent containing ca. 5 % CD2CI2' 109 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CH« CH3D ch 2D2, CHD3 * / / (B) a o o o 040 O JO 030 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. F ig u re 1 1 . Electron isodensity surfaces (at 0.002 electrons/au3) for N2 (1), N2H2 isomers (5-7) and disubstituted diazenes (2-4). This surface profile is similar to a space filling (CPK) model in that it reflects the space occupied by the molecule when placed in a solid lattice or associated in a liquid. The colors map the electrostatic potential on the surfaces, red indicating a region of negative potential (potential H+ - binding site) and blue a region of positive potential. Km values (mM) for substrates are from Hwang etal. 1 4 (1), McKenna eta l.9 (3, 4) and this work. The electrostatic potential surfaces were generated on the same scale on a Macintosh Power PC computer by ab initio Hartree-Fock 3-21 G* calculation using MacSpartan software from Wavefunction, Inc. (Irvine, CA). I l l Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Km ,m M $ § o Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER 4 THE CATALYTIC REDUCTION OF C /S-D IM ETH YLD IAZEN E BY THE [MoFe3 S4 ]3 + CLUSTERS. THE FOUR-ELECTRON REDUCTION OF A N=N BOND BY A NITROGENASE- RELEVANT CLUSTER AND IMPLICATIONS FOR THE FUNCTION OF NITROGENASE. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ABSTRACT The catalytic reduction of cis-dimethyldiazene by the (Et4 N)2 [(CI4 -cat) (CH3 CN)MoFe3 S4 Cl3 ] cluster (Cl4 -cat = tetrachlorocatecholate) is reported.1 '3 Unlike the reduction of cfr-dimethyldiazene by the Fe/Mc/S center of nitrogenase, which yields methylamine, ammonia and methane (die latter from die reduction of the C-N bond), the reduction of cis-dimethyldiazene by the synthetic cluster yields exclusively methylamine. In separate experiments, it was shown that the C-N bond of methylamine is not reduced by the [M oF e^J3 * core, perhaps accounting for the differences observed between the biological and abiological systems. 1,2-Dimethylhydrazine, a possible partially-reduced intermediate in the reduction of cis-dimethyldiazene, was also shown to be reduced to methylamine. Interaction of methylamine with the Mo atom of the cubane was confirmed through the synthesis and structural characterization of (Et4 N)2 [(CI4 - cat)(CH3 NH2 )MoFe3 S4 CI3 ]. Phosphine inhibition studies strongly suggest that the Mo atom of the [M oFejSJ3 * core, which has a Mo coordination environment very similar to that in nitrogenase, is responsible for the binding and activation of cis-dimethyldiazene. The reduction of a N=N bond exclusively at the heterometal site of a nitrogenase-relevant synthetic compound may have implications regarding the function of the nitrogenase Fe/Mo/S center, particularly in the latter stages of dinitrogen reduction. INTRODUCTION Nitrogenase is capable of reducing many small, unsaturated molecules and ions, the most important of which is atmospheric N2 . The enzymatic synthesis of ammonia by reduction of atmospheric N2 under ambient conditions supplies nitrogen in the form needed in the biosynthesis of proteins and nucleic acids. It is not surprising therefore that much research has been undertaken in order to achieve a better understanding of this remarkable 114 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. reduction process.4 In recent X-ray structure determinations o f nitrogenase the structure of die Fe/Mo/S active site has been obtained at near-atomic resolution.3 This MoFe,S9 cluster of nitrogenase, referred to as the FeMo-cofactor, is comprised of MoFe3S3 and Fe4 S3 site- voided cubanes linked by three bridging sulfides (Figure 1). The Mo atom of the FeMo- cofactor appears to be coordinatively saturated with a terminal imidazole from a histidine residue and a bidentate R-homocitrate ligand completing the coordination sphere. The synthesis of exact analogs for the FeMo-cofactor has not been accomplished. Nevertheless, partial analogs exist in synthetic clusters.6 Outstanding among these clusters are simple Fe/Mo/S cubanes that contain the [MoFe3 S J cores7 and a Mo atom with first and second coordination spheres nearly identical to those found for the Mo atom in nitrogenase. In addition to the structural and electronic similarities,7 it has recently been shown that these [MFejSJ"* clusters (M = Mo, n = 3 or M = V, n = 2, the latter of which are relevent to the alternative V-nitrogenase8 ) are also partial functional models for the enzyme. Substrate reductions that have been investigated with these clusters include the reduction of hydrazine to ammonia9 and acetylene to ethylene and ethane.1 0 The most significant results from these studies include the observation that the heterometal atom (Mo or V) in these clusters is either exclusively9 or principally1 0 involved in the activation of the substrate toward reduction. The synthetic clusters most efficient in their ability to act as catalysts have been the (Et4 N)2 [(L)(L')MoFe3 S4 C lJ single cubanes, with either a labile solvent molecule weakly coordinated to the Mo atom (L = bidentate tetrachlorocatecholate, L’ = CH3 CN7 ) or, paradoxically, a tridentate polycarboxylate ligand (L, L' = polycarboxylato ligand1 1 ) that results in a coordinatively saturated Mo atom. When the structure of the Fe/Mo/S center in nitrogenase revealed that the Mo atom of the cluster was coordinatively saturated, the direct involvement of the Mo atom in substrate binding appeared unlikely. The ability of the polycarboxylate-ligated [M oFe^J3 * model clusters to serve as catalysts suggested a special 115 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. role for die coordinated carboxylate ligands. Indeed, die observation that carboxylate- bound clusters with coordinatively saturated Mo atoms are also catalytically active in substrate reduction implies their ability to generate coordination sites for the substrate by displacing one of the "arms" of the carboxylate ligand through protonation.9 ’ 1 1 In some cases,9 these polycarboxylate-ligated clusters are actually better catalysts than die catecholate precursors, presumably due to die protonated arm of the ligand that may serve as a "shuttle”, returning a proton to the reduced substrate. While N2 is not reduced by the synthetic [MFe3 S Jn + clusters, the catalytic reactivity of the latter suggests the possibility of partially reduced substrates interacting and being reduced at the heterometal atom of die Fe/M/S center in nitrogenase. With regard to dinitrogen reduction, it has been suggested that dinitrogen is activated in the Fe3 (p-S)3 Fe3 "cage" created by the six three-coordinate Fe atoms in the central part of the cofactor.1 2 The mechanism of dinitrogen reduction is believed to proceed through diazene-Iike intermediates,1 3 although diazene has not yet been demonstrated to interact with the nitrogenase cofactor.1 4 At present, it is not known where the reduction of dinitrogen to ammonia occurs. It could take place entirely at the six-Fe "cage" or it may undergo the initial two- or four-electron reduction at the six-Fe "cage" and the final two- to four-electron reduction and cleavage o f the N-N bond at the Mo atom. Recently, it has been reported that both cis- and /ra/is-dimethyldiazene are substrates for nitrogenase, and as such represent the first example of reduction of an N=N bond by the Fe/Mo/S centra1 of nitrogenase.1 3 Products detected included methylamine, methane, and ammonia in ratios that were dependent on both the specific isomer used as substrate and (as demonstrated for the cis isomer) the Fe:FeMo protein ratio. The strained- ring diazene, diazirine, is also reduced by nitrogenase to the same products.1 5 * In order to investigate what role, if any, the Mo atom may play in reduction of a N=N bond, cis- dimethyldiazene was investigated as a potential substrate for the [MoFe3 S J ^ cubanes. 116 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Herein, we report on the catalytic reduction of cis-dimethyldiazene by the (Et4 N)2 [(Cl4 - cat)(CH3 CN)MoFe3 S4 C lJ cubane. Results show that the only detectable product from the reduction of cis-dimethyldiazene by the synthetic cluster is methylamine. Additionally, it has been demonstrated through inhibition studies that activation and reduction of cis dimethyldiazene occurs exclusively at the Mo site. The implications of these observations on the possible function of the cofactor of nitrogenase are also discussed. EXPERIM ENTAL General Considerations. All manipulations were performed under an inert atmosphere using standard glove box and Schlenk techniques. Solvents were distilled under N2 from the appropriate drying agents (diethylether and THF from sodium/benzophenone, CH3 CN from B2 03 ) or stored over 3 A molecular sieves (absolute ethanol) and thoroughly degassed with N2 or Ar prior to use. Reagent grade chemicals were purchased from Aldrich Chemical Company (cobaltocene (CoCp^, 99% 2,6-lutidine (Lut), anhydrous DMF, methylamine, ethylamine, dimethylhydrazine dihydrochloride, NaBPh4 , 1.0 M ethereal HC1, PEt,) and used without further purification. Freshly prepared solutions of c&-dimethyldiazene1 6 in distilled, degassed CH3 CN (typically 0.1 to 0.2 M as determined by UV spectroscopy, e3 6 7 = 266) were shipped to University of Michigan on dry ice and stored at -196 °C until immediately prior to use. Physical Measurements. Infrared spectra (Csl disks) were obtained using a Nicolet 740 FT-IR spectrometer (far-IR: 500 - 150 cm 1 ) or a 5DXB FT-IR spectrometer (mid-IR: 4000 - 400 cm'1 ). Quantification of methylamine and ammonia was performed using an HPLC technique previously described." An HP 5890 Series II gas chromatograph equipped with either a porapak N column (Supelco) or a 4% carbowax 20m column (Supelco) was used in order 117 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. to detect methane or EtNH2 , respectively. EPR studies and elemental analyses were performed by the Biophysics Research Division and the Analytical Services Division, respectively, at the University of Michigan. Analysis samples were routinely kept under dynamic vacuum for 12 hours before submission. Preparation of Compounds Analytically pure 2,6-lutidinium hydrochloride (Lut HQ) was prepared from the reaction between lutidine and ethereal HC1. Lut-HBPh4 was prepared from the metathesis reaction between Lut HQ and NaBPh4 in ethanol. (Et4 N)2 [(Cl4 - cat)(CH3 CN)MoFe3 S4 Cl.j]u and (Ph4 P)2 [Fe4S4 C lJ 1 9 were obtained by procedures similar to those previously reported. (Et4N)2[(CI4-cat)(RNH 2)MoFe3S 4CI,l (R = Me or Et) An amount of RNHj (0.10 mL of a 2 M EtOH solution) was added to an CH3 CN solution (30 mL) of (Et4 N)2 [(Cl4 -cat)(CH3 CN)MoFe3 S4 Cl3 ] (0.21 g, 0.20 mmol) in one portion. After approximately one hour of stirring, the solution was filtered and ether (150 mL) was layered on the filtrate. After overnight standing, a near quantitative yield of brown crystals was isolated by filtration and washed well with ether. Elemental and IR analyses were satisfactory. X-ray crystallography was performed at the University of Michigan. Reduction of cis-dimethyldiazene. A) Time-Course Studies. A 125 mL flask was charged with 0.03 g (159 pmol) CoCp2 and 0.03 g (209 pmol) Lut HC1. Acetonitrile was then added (35.5 mL) to form a slurry. An aliquot of a 4.8 mM CH3 CN solution of (Et4 N)2 [(Q 4 -cat)(CH3 CN)MoFe3 S4 Cl3 ] (0.45 mL, 2.2 pmol) was then added, followed immediately by an aliquot of the cis-dimethyldiazene solution (32 pmol). The addition of the substrate marked t = 0 hr. At t = 0.5, 1.0, 2.0, 3.5 and 5 hr, a 100 pL sample of head space gas was obtained from the reaction flask and analyzed for 118 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CH4 . In addition, 3.0 mL samples were removed from the reaction flask and placed in 5 mL septum-capped vials. These aliquots were immediately quenched with 0.2 mL o f a 1.0 M aqueous HC1 solution. The vials were analyzed for methylamine, ethylamine, and ammonia. Reactions were performed under both N2 and Ar. Also, "blanks" were run under the same conditions but only sampled at 3.0 hr. B) Phosphine-Inhibition Studies. In a 20 mL vial fitted with a septum, 8.3 mL of a 7.2 mM CoCp2 solution (CH3 CN, 60 pmol), 8.3 mL of a 10.9 mM L utH C l solution (CH3 CN, 90 pmol) and an appropriate amount of a 9.5 mM PEt, solution (CH3 CN, from 0-10 equivalents based on cubane concentration) were combined. A 0.2 mL aliquot of a 4.8 mM cluster solution (CH3 CN, 0.96 pmol) was then added, followed immediately by a 0.08 mL aliquot of die substrate solution (CH3 CN, 15 pmol). Addition of the substrate marked t = 0 hr. A 2.0 mL aliquot of the reaction solution was obtained at t = 1 hr and immediately quenched with 0.2 mL of a 1.0 M aqueous HQ solution in a 5 mL septum-capped vial. Vials were subsequently analyzed for methylamine and ammonia. In some reactions, (Ph4 P)2 [Fe4 S4 C IJ was used in place of the (Et4 N)2 [(Cl4 -cat)(CH3 CN)MoFe3 S4 CI3 ] cluster. C) Kinetic Data. Reactions were prepared in 125 mL flasks. The total solution volume was brought up to 40 mL with acetonitrile. Aliquots of acetonitrile solutions of CoCp2 (15.0 mL, 108 pmol), L utH Q (10.0 mL, 109 pmol) and cluster (0.15 mL, 0.72 pmol) analogous to those described in section B above were placed in a flask containing predetermined amounts of CH3 CN. With the concentrations of these reagents held constant, the appropriate amount of a cis-dimethyldiazene solution (3 to 167 equivalents based on cluster concentration) was added to the flask, marking t = 0 hr. At t = 1 hr, 3.0 mL aliquots of die reaction solution were obtained and worked-up as described above. 119 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. D) Attempted Reductions of 1,2-dimethylhydrazine, methylamine and ethylamine. To a 125 mL reaction flask filled with an appropriate amount of CH3 CN (total solution volume was 40 mL) was added 4.2 mL of a 7.2 mM CoCp2 solution (30 pmol), 0.03 g of Lut.HBPh4 (70 pmol) and 0.2 mL of a 4.8 mM (Et4 N)2 [(Cl4 - cat)(CH3 CN)MoFe3 S4 Cl3] solution. An excess of the appropriate substrate (MeNH2 , EtNHj or 1,2-dimethylhydrazine2 1 as CH3 CN solutions, 10-15 equivalents based on cubane concentration) was then added, marking t = 0 hr. At t = 1 hr, aliquots were obtained and quenched with HC1 as described and analyzed for products. In the case o f the amines, the headspace gas in the reaction solutions was analyzed for CH4 and the solution analyzed for ammonia by the indophenol method as previously described.9 In the case of 1,2- dimethylhydrazine, the vials were analyzed for methylamine by HPLC. The reactions described in Sections A, B and D were performed in duplicate, while those in Section C were performed in triplicate. Additionally, each reaction aliquot was analyzed for products at least twice. While repeat measurements on each vial typically did not vary by more than 10%, absolute product yield varied slightly between identical reactions. Regardless, the trends observed for the time course of the reaction, the phosphine inhibition and the v vs. [ro-dimethyldiazene] curves were consistent between sets of experiments and are reported as such. E) Recovery and Identification of the [MoFe}SJ3 * Catalyst. A 125 mL flask was charged with 23 mL of a 7.2 mM CoCpj solution, 23 mL of a 10.9 mM LutHCl solution, and 3.0 mL of a 4.8 mM solution of (Et4 N)2 [(Cl4 - cat)(CH3 CN)MoFe3 S4 Cl3 ]. Addition of 0.42 mL of a 0.1 M cu-dimethyldiazene solution marked t = 0 minutes. After stirring for 90 minutes, the solution was taken to dryness. The resulting residue was washed well with THF to remove any unreacted CoCp2 , and 120 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. subsequently dissolved in 10.0 mL DMF. An aliquot of this DMF solution (1.3 mM cubane) was then obtained and subjected to a quantitative EPR analysis. RESULTS AND DISCUSSION Synthesis and C rystallographic R esults. After some 20 years of very thorough studies, the synthesis, ligand-substitution properties and reactivity of clusters with the [M oFejSJ3 * core arc well established.7 It also has been demonstrated that a) the catecholate moiety on the Mo atom of the cubane can be removed through protonation by acidic catechols2 2 and polycarboxylic acids1 1 and b) the single labile solvent molecule on the Mo atom in (Et4 N)2[(CI4 -cat)(CH3 CN)MoFe3 S4 Cl3 ] and related clusters can be readily replaced by any number of tr-bases, but generally not by rc-acids except in the case of reduced cores.2 3 It is not unexpected, therefore, that amines will readily coordinate to the Mo atom of the cluster through the N-lone pair, replacing the relatively poor CH3 CN ligand. The sole purpose in verifying this ligand substitution synthetically and crystallographically is to establish unambiguously the Mo-amine interaction in light of the result that neither MeNHj nor EtNH2 are reduced by die [MoFe3 S J3 + core (vide infra). The methylamine single cubane I, shown in Figure 2, crystallizes with two distinct anions within the asymmetric unit, one in which all atoms are located on general positions (A) and the second which is bisected by a crystallographic mirror plane through Mol, Fe2, S2 and S4 (B). The Mo-N distance in both anions is very nearly the same at 2.28(2) and 2.35(4) A. The average M o-Fe (2.737(6) A), Mo-S (2.346(9) A), Mo-O (2.09(2) A), Fe- S (2.26(1) A), Fe-Cl (2.22(1) A) and Fe--Fe (2.727(7) A) distances are unexceptional. 121 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Given the abundance of Mo-cubane structures in the literature,7 no further description of this structure is warranted. Substrate Reductions. As a reference point, catalytic reductions were generally performed with the substrate in approximately 15-fold excess relative to the catalyst, with externally added sources of electrons (CoCp2 , £4 eq) and protons (LutHCl, ^6 eq). As shown in Figure 3, cis-dimethyldiazene is catalytically reduced to methylamine under these conditions. There is an initial burst to approximately 50% conversion (based on 2 moles of methylamine for every mole of cis-dimethyldiazene) in the first hour of reaction time for the 15:1 substrate/catalyst ratio, with conversion slowly leveling off. This behavior may be due to a slow precipitation of the catalyst as cations are generated in solution (i.e., CoCp2 *, MeNH3 *), a behavior typically observed in the reduction systems previously investigated.9 In order to verify that 1) the substrate was being activated and reduced by the [MoFe3 S J3 * core exclusively and 2) methylamine was obtained only from reduction of cis- dimethyldiazene, a number of "blanks" were performed. These blanks included reaction systems with all reagents present except a) the substrate, b) the [MoFe3 S J3 + catalyst, c) proton and electron sources and d) with 10 equivalents of PEt, added to inhibit the Mo site. As shown in Figure 4, while methylamine is present in all systems, amounts of methylamine from these blanks is typically 10% of the amounts detected from reactions where all reagents are present and the [MoFe3 S4 ]3 + cluster is present as a catalyst These results taken together strongly suggest that cis-dimethyldiazene is catalytically reduced by the [MoFejSJ3 * cuboidal core in the presence of externally added protons and electrons. The integrity of the (Et4 N)2 [(Cl4 -cat)(CH3CN)MoFe3 S4 Cl3] catalyst at the end of the reaction was verified through EPR spectroscopy. After removal of excess CoCp2 (S = 1/2) from the reaction mixture (THF washings were essentially colorless), an EPR spectrum of 122 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. the reaction solution was obtained. The characteristic S = 3/2 spectrum for the [MoFe3 S J3 + core was observed, and the signal integrated to 1.1 ± 0.3 mM, indicating no cluster decomposition during the first 90 minutes of reaction time. Previous studies on the identification of the recovered (Et4N)2[(Cl4 -cat)(CH3 CN)MoFe3S4 Cl3] cluster after catalytic cycles included elemental analysis, electronic and infrared spectroscopy and demonstration of further catalysis by the recovered cubane.9 , 1 0 These results taken together demonstrate the robust nature of (E^N^KC^-catXCHjCbOM oFe^Clj] under reaction conditions and identifies the cubane as the active catalyst for the reduction o f a variety of substrates. Ammonia yields for the reactions shown in Figure 4 were 1) exceedingly small, only reaching approximately 1% of the total yield of methylamine in the catalytic systems and 2) essentially equivalent for all systems, catalytic reactions and blanks alike. It was therefore concluded that the reduction of cis-dimethyldiazene with these synthetic clusters led to formation of methylamine but not ammonia, unlike what is observed in the enzyme system. No change in product distribution was observed when the reaction was performed under Ar instead of N2 . Given the inability for N2 to serve as a ligand for the Mo-cluster in either the 3+ or 2+ state, this result was expected. The lack of reduction products derived from C-N bond reduction in cis- dimethyldiazene (the only route leading to NH3 ) was verified in two individual experiments. First, GC analysis of the headspace gases in the catalytic systems routinely showed no CH4 up through a 5 hour period. Methane would necessarily have to be present in a 1:1 ratio with any ammonia that was produced in the system. Second, in experiments using methylamine as a "substrate" for the [MoFe3 S J ^ cluster, no ammonia or methane were observed at any time as a result of potential reduction of the C-N bond. Given that methylamine is a better ligand for the [MoFe3 S J 3 * cluster than CH3 CN based on synthetic and crystallographic results (vide supra), the lack of reactivity observed with methylamine is due to its inability to be reduced by the [MoFe3 S4 ]3 + core and not on its inability to 123 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. interact with the catalyst The inability of the model systems to reduce methylamine may be due to the lack of a free lone-pair needed for protonation once the substrate is coordinated to the catalyst In model systems employing hydrazine as a substrate, protonation of the bound hydrazine was shown to be a necessary first step in making the reduction potential of the cluster accessible.9 No turnover of hydrazine to ammonia was observed in systems which lacked sufficiently acidic protons to protonate hydrazine. The ability of the [MFe3 S Jn + (M = Mo, n = 3; M = V, n = 2) cores to catalytically reduce hydrazine to ammonia has been established previously.9 As a matter of principle, however, it was important to verify that 1,2-dimethylhydrazine (a possible reduced intermediate in the reduction of cis-dimethyldiazene) reacted in a similar manner. Under catalytic conditions (15-fold excess of 1,2-dimethylhydrazine relative to catalyst), approximately 80% of the substrate was reduced to methylamine within the first hour, a yield essentially identical to that reported for the unsubstituted hydrazine.9 6 A "blank" reaction which contained no catalyst showed the typical background levels of methylamine (6%) established previously. In these reactions, Lut HBPh4 was used instead of Lut HCI to insure that a m a jo rity of the substrate stayed soluble initially, given that the dihydrochloride salt of 1,2-dimethylhydrazine is quite insoluble in CH3 CN. In addition to methylamine, it was established by HPLC and GC that ethylamine is also a product in these systems (Figure 4). Yields of ethylamine a) are largest when there is no as-dimethyldiazene in the system, b) are essentially nonexistent in the absence of catalyst or protons/electrons, c) are present in lower amounts when cu-dimethyldiazene is present and d) are not observed when 10 equivalents of phosphine (per cluster) are present in the system. These results suggest that the [M oFejSJ3 * core may reduce the solvent, CHjCN, to ethylamine at a very slow rate. When "better" ligands are present (phosphine, hydrazine, cw-dimethyldiazene), the yields are either decreased or non-existent In separate 124 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. experiments, it was verified that while ethylamine binds to the Mo-atom of the cluster (II), it is not reduced to ammonia and methane. Role of the Mo Atom in Substrate Reduction. It has been well established in the reduction of other nitrogenase substrates by the synthetic [MFe3 S JB * clusters that the heterometal (either Mo or V) is either exclusively (in the case o f hydrazine)9 or predominantly (in the case of C2H2)1 0 involved in the activation of the substrate toward reduction. The role of the heterometal in the cfr-dimethyldiazene system was investigated by 1) using the [Fe4 S4 C IJ2 * as a potential catalyst in place of the [M oFejSJ3 * core and 2) observing the effect of the addition of phosphine to the catalytic system. Phosphine is known to bind strongly to the Mo atom2 1 and presumably dramatically effects the ability of other, weaker a-bases to coordinate. Using the all-Fe cluster as a potential catalyst, amounts of methylamine observed after one hour were well within background limits established with the blanks, suggesting that the Fe atoms in the [MoFe3 S J 3 * cluster are not involved in catalysis. Additionally, as shown in Figure 5, it is clear that as phosphine is added in greater amounts to the reaction system (from 0 to 10 equivalents of phosphine), the turnover of substrate to products drops dramatically. At 1 equivalent of phosphine, the amount of product in one hour drops to about half of the amount obtained when [PEtJ = 0. Between 5 and 6 equivalents of phosphine, the amount of methylamine present in the reaction solutions was determined to be well within background limits, indicating complete inhibition. These results taken with those results obtained previously with other substrates demonstrates the importance of die Mo atom in substrate binding and activation toward reduction. Investigation of Kinetics. From initial time-course studies, it was determined that the conversion of substrate to product with time was essentially linear within the first 60 minutes. Therefore, catalytic 125 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. systems were performed over a number of cis-dimethyldiazene concentrations and analyzed at t = 1 hr. As shown in Figure 6, it is clear that typical saturation kinetics observed for enzyme systems are not applicable in this model system. This observed trend may be due in part to insolubility o f the catalyst as cations are generated in solution (CoCp2 \ MeNH3 + ). At lower substrate :catalyst ratios, the catalyst remains in solution to effect near quantitative reduction within the first hour. At higher concentrations o f substrate, the reaction velocity drops to a minimum, followed by a gradual increase. As shown in Figure 7, however, the percent conversion of cis-dimethyldiazene to methylamine (based on 2 eq. of product per mole of substrate) is essentially constant at high [m-dimethy Idiazene]. A Comparison Between the Synthetic and the Enzymatic Systems. It has been reported1 3 that as a substrate for die A. vinelandii nitrogenase, cis- dimethyldiazene is reduced to NH3 , CH4 and CH3 NH2 in a ratio of 1.0:1.0:1.6 ± 0.1 (Fe:FeMo protein ratio of 6.6), in contrast to the synthetic system employing [MoFe3 S J3 + as catalyst, where methylamine is produced exclusively. In the synthetic system, methylamine is not reduced and hence no methane or ammonia are observable as products. It seems likely that the reaction proceeds by one of two pathways (assuming mononuclear activation), as depicted in Figure 8. The coordinated as-dimethyldiazene may be reduced in a "symmetric" way, forming dimethylhydrazine as an intermediate, which has been shown to yield methylamine upon protonation and reduction. Alternatively, the substrate may be reduced in an “unsymmetric" manner, forming methylamine and a Mo-bound imine after the first two-electron, two-proton step. Either way, it is clear in the model system that methylamine is the only product and that the Mo atom seems to be of primary importance. In the enzymatic system, it seems likely that either a) the N=N bond and the C-N bond are is some way reduced concomitantly or b) methylamine is an initial product, which is activated in some way by the nitrogenase cofactor. If the latter, the methylamine would 126 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. have to bind to a center that was capable of removing significant electron density from die C-N bond given that only one lone pair of electrons are available and, when bound to a metal center, there would be no place for protonation to occur on methylamine. Either way, it is clear that a different mechanism is occurring in the biological and abiological systems, specifically with regard to C-N bond cleavage. In conclusion, the nitrogenase enzyme has recendy been shown to catalyze the reduction of cis-dimethyldiazene to methane, ammonia, and methylamine, and the nitrogenase model compound (Et4 N)2 [(CI4 -cat)(CH3 CN)MoFe3S4 C l3] also reduces die N=N bond in m-dimethyldiazene, but methylamine is the only observable product The inability of the model compound to reduce methylamine, even though it has been unambiguously demonstrated that methylamine binds to the Mo-atom in the cluster, may account for the lack of methane and ammonia formation. Insofar as the cuboidal clusters can be considered reactivity models for the cofactor of nitrogenase, the ability of the [M oFejSJ3 * cores to reduce both N=N and N-N exclusively at the Mo site may suggest the Mo-atom in the cofactor is, in fact, not "innocent" as suggested in the past,1 2 but rather plays a vital role in the later stages of dinitrogen reduction. We have previously considered the possibility of initial activation of dinitrogen at the unique, 3-coordinate Fe atoms of the cofactor, followed by "migration" of a partially-reduced species (hydrazine) to the peripheral Mo atom where reduction to ammonia proceeds.9 * This work not only supports this possibility, but also suggests that the Mo atom may be involved in the reduction process at the diazene level. 127 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. REFERENCES (1) Department of Chemistry, The University of Michigan, Ann Arbor MI 48109-1055 (2) Department of Chemistry, The University of Southern California, Los Angeles CA 90089-0744 (3) Present Address: Department of Macromolecular Science, Case Western Reserve University, Cleveland OH 44106 7202 (4) (a) Molybdenum Enzymes, Cofactors and Model Systems; Stiefel, E.I,; Coucouvanis, D.; Newton, W.E., eds. ACS Symposium Series 535; American Chemical Society: Washington, D.C., 1993; Chapters 10-23. (b) Eady, R.R.; Leigh, G.J. J. Chem. Soc. Dalton Trans. 1994, 2739. (c) Kim, J.; Rees, D.C. Biochemistry 1 994,33, 389. (d) Evans, D.J.; Henderson, R.A.; Smith, B.E., in Bioinorganic Catalysis, J. Reedjik, ed., Marcel Decker, Inc., New York: 1993, Chapter 5. (5) (a) Kim, J., Rees, D.C., Science, 1992, 257, 1677. (b) Kim, J., Rees, D.C., Nature, 1992,360, 553. (c) Chan, M.K., Kim, J., Rees, D.C., Science, 1993, 260, 792. (d) Bolin, J.T., Campobasso, N., Muchmore, S.W., Minor, W ., Morgan, T.V., Mortenson, L.E. in New Horizons in Nitrogen Fixation, Palacios, R., Mora, J., Newton, W.E., eds., Kluwer Academic Publishers: Dordrecht, 1993, pp. 89-94. (e) Bolin, J.T., Campobasso, N., Muchmore, S.W., Morgan, T.V., Mortenson, L.E., in Molybdenum Enzymes, Cofactors and Model Systems, ACS Symposium Series No. 535, E.I. Stiefel, D. Coucouvanis, W.E. Newton, eds., American Chemical Society, Washington, D.C.: 1993, pp 186-195. (6) Coucouvanis, D. Acc. Chem. Res. 1991,24, 1. (7) (a) Holm, R.H. Adv. Inorg. Chem. 1992,5S,1, and references therein, (b) Holm, R.H.; Simhon, E.D. in "Molybdenum Enzymes", Spiro, T.G., ed.; John Wiley & Sons, Inc.: New York, 1985, p. 1-87, and references therein. (8) (a) Ciurli, S.; Holm, R.H. Inorg. Chem. 1989, 28, 1685. (b) Kovacs, J.A.; Holm, R.H. Inorg. Chem. 1987,26, 702. (9) (a) Demadis, K.D.; Malinak, S.M.; Coucouvanis, D. Inorg. Chem. 1996, 35, 4038. (b) Coucouvanis, D.; Demadis, K.D.; Malinak, S.M.; Mosier, P.E.; Tyson, M.A.; Laughlin, L.J. J. Mol. Cat. A: Chemical 1996, 107, 123.(c) Malinak, S.M.; Demadis, K.D.; Coucouvanis, D. J. Am. Chem. Soc. 1995,117, 3126. (d) Demadis, K.D.; Coucouvanis, D. Inorg. Chem. 1995, 34, 3658. (e) Coucouvanis, D.; Mosier, P.E.; Demadis, K.D.; Patton, S.; Malinak, S.M.; Kim, C.G.; Tyson, M.A. J. Am. Chem. Soc. 1 9 9 3 ,115, 12193. (10) Laughlin, L.J.; Coucouvanis, D. J. Am. Chem. Soc. 1 995,117, 3118. 128 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. (11) (a) Demadis, K.D.; Coucouvanis, D. Inorg. Chem. 1995,34 ,436, and references therein, (b) Demadis, K.D.; Coucouvanis, D. Inorg. Chem. 1995, 34, 3658, and references therein. (12) (a) Stavrev, K.K.; Zemer, M.C. Chem. Eur. J. 1996, 2, 83. (b) Dance, I.G . Aust. J. Chem. 1994, 47, 919. (c) Deng, H.; Hoffman, R. Angew. Chem. Int. Ed. Engl. 1993,32, 1062. (d) Chan, M.K.; Kim, J.; Rees, D.C. Science 1993, 260, 792. (13) Thomeley, R.N.F.; Lowe, D J. Biochem. J. 1984,224, 887. (14) Burris, R.H.; Winter, H.C.; Munson, T.O.; Garcia-Rivera, J. in Intermediates and Cofactors in Nitrogen Fixation, A. San Pietro, ed. Antioch Press, Yellow Springs, OH: 1965, p. 315. (15) (a) McKenna, C.E.; Simeonov, A.M ., Eran, H.; Bravo-Leerabhandh, M. Biochemistry 1996,35, 4502. (b) McKenna, C.E.; Simeonov, A.M. In Nitrogen Fixation: Fundamentals and Applications, I.A. Tikhonovoch, N.A. Provorov, I. Romanov, W.E. Newton, eds., Kluwer Academic Publishers, Dordrecht: 1995, p. 158.(c) Simeonov, A.M.; McKenna, C.E., submitted. (16) Simeonov, A.M.; McKenna, C.E. J. Org. Chem. 1995, 60, 1897. (17) Bravo, M.; Eran, H.; Zhang, F.X.; McKenna, C.E. Anal. Biochem. 1988, 175, 482. (18) Hearshen, D. O.; Hagen, W. R.; Sands, R. H.; Grande, H. J.; Dunham, W. R. J. Magn. Reson. 1986,69, 440 and references therein. (19) Palermo, R.E.; Holm, R.H. J. Am. Chem. Soc. 1983,105, 4310. (20) Wong, G.B.; Bobrik, M.A.; Holm, R.H. Inorg. Chem. 1978,17, 578. (21) In order to prepare a standard solution of 1,2-dimethylhydrazine in CH3 CN, a slight excess of EtjN was used in order to neutralize and hence solubiUze the dihydrochloride salt. (22) Palermo, R.E.; Sinh, R.; Bashkin, J.K.; Holm, R.H. J. Am. Chem. Soc. 1984, 106, 2600. (23) Mascharik,P.K.; Armstrong, W.H.; Mizobe, Y.; Holm, R.H. J. Am. Chem. Soc. 1 9 8 3 ,105, 475. 129 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 1. The FeMo-cofactor of nitrogenase. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. T \ \ / K * 1 T h C O Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 2. An ORTEP plot (30% probability ellipsoids) showing the anion of (Et4N)2[(Cl4 - catXCHgNH^M oFejS^] (I) 132 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 133 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. C K 7 ) Figure 3 . Time-course of methylamine production from the reduction of cis- dimethyldiazene. [(Et4N)2[(Cl4 - catXCHJNH 2)MoFe3S4a 3]] = 54 pM , [c/s - dimethyldiazene]0 = 800 pM . Conversion at 5 hrs. is approximately 90%. 134 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 25 15 ■o 10 5 0 2 1 3 4 0 5 Time (hr) Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 4. Product distributions after 3 hours for reaction systems as indicated. See Experimental Section for reaction conditions. 136 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission. Peak Area, jiV-sec N » © I o tn I § o\ 1 U l No substrate • E No cubane No protons, no electrons i H ° 3 2 C O 2 UJ o Complete assay - V/////////7A K > Comple,' assay Y //// / / / / / / / / ^ W A . Lw.v.y.y.v.v.v.v.v.v.-.v.y.v.v.v.v.v.vXv.v.v.v.v.v.v.v.v.v.v.v.v.v.v.v.w.l U > Complete + 1 eq. PEt3 - V/////////777?, Complete + 10 eq. PEt3 - * *' * ....................................... 7.0E+05 F igure 5. Inhibition of cw-dimethyldiazene reduction by PEtj. See Experimental Section for reaction conditions. 138 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CH3NH2 p e a k area 8 0 0 0 0 0 y * 2.85E-08X + 1.04E-06 Q r2 - 9.83E-01 > 1 .0 E - 0 5 6 0 0 0 0 0 5.0E-06 4 0 0 0 0 0 O .O E + O O 4 0 0 500 100 200000 5 0 0 6 0 0 3 0 0 4 0 0 100 200 pMPEt3 139 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 6 . Reaction velocity (v) vs. [c 6 150 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. volumes) NH4 -formate buffer (+2% DT) were placed into the exchanging buffer chamber. The diafiltration was done at ca. 60 psi and was complete within ca. 9 hours. All three solutions remained reducing throughout the experiment Iron-molybdenum cofactor preparation. Iron-molybdenum cofactor (FeMo-co) was isolated by using a modified version of the original method by Shah and Brill.6 The FeMo protein was exchanged into ammonium formate buffer whose components are removable by pumping, and the protein acidification step was done with the volatile trifluoroacetic instead of citric acid. During this procedure,7 all protein and cofactor manipulations were done in a Vacuum Atmospheres MO-40 glove box, whose 0 2 -scrubbing catalyst was regenerated twice within two subsequent nights just prior to use. Every attempt was made to minimize the time a sample spent outside the box during the centrifugation steps, including pre-taring the centrifuge tube by eye prior to removal from the antechamber. Dimethylformamide (DMF), N-methylformamide (NMF) and trifluoroacetic acid (TFA) were all obtained from Aldrich and purified by drying and low-pressure distillations as described elsewhere.7 Approximately 1.2 ml of "FeMoX diafiltered" (ca. 16.8 mg) were subjected to water dilution to 6 ml, followed by pH adjustment with 0.58 M TFA to about 2.2 (stirring). Upon acidification, the brown protein solution turned grayish and became very turbid. After three-minute stirring, the pH was brought up to ca. 5.5 by the addition of 0.35 M ammonia, and the mixture was left to stand for 25 min. Subsequent centrifugation (11 min, 2000 rpm, JA 20 rotor) led to the formation of brownish-gray protein pellet and clear supernatant (discarded). The pellet was briefly stirred with 4 ml DMF and then subjected to another 11 min centrifugation at 3000 rpm. The DMF supernatant was carefully removed, and the pellet was resuspended into ca. 2 ml NMF. The mixture was 151 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. vigorously vortexed for ca. 7 min, and then centrifuged for 11 min at 5000 rpm. A grayish pellet was obtained along with brownish supernatant A second extraction yielded similarly, although even more lightly colored supernatant A total of 3.0 ml of cofactor solution were obtained. Reconstitution assay of the iron-molybdenum cofactor. Since the cofactor preparation appeared very dilute, it was decided to use 15 pi of preparation per reconstitution vial, instead of the 5 pi originally planned. Incubation of this constant amount with varying volumes (0-400 pi) of DJ42 cell-free extract for ca. 20 min was followed by a transfer of the mixture into 5 ml assay vials and performing the standard acetylene assay.8 Revival and initial derepression tests of DJ42 mutant strain. The A. vinelandii mutant DJ42 is one of the several variants that contain deletions which cause die strains, upon derepression (incubation in N-free medium which turns on the biosynthesis of nitrogen fixation proteins), to produce essentially FeMo cofactor-less FeMo protein.9 The nitrogenase activity of crude extracts of such bacteria can be reconstituted by a brief incubation with preparation of FeMoco; this assay serves, among other purposes, as a convenient way to assay the activity (and therefore verify the integrity) of different cofactor preparations. The starter culture was a kind gift from Prof. D. Dean, Virginia Polytechnical Institute. The strain was initially revived by growth on nitrogen- containing plate (2 passes), followed by a transfer to a liquid medium. The liquid culture started growing rapidly, and a liquid sample taken at 6 8 Klett revealed very high motility. At ca. 85 Klett, a portion of the growing culture was transferred into another side-arm flask containing N-free medium, and the incubation was continued. After initial rapid growth, 152 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. the depletion of N supply in the medium led to a gradual slowdown and ultimately, to a flat growth curve. It was, therefore, demonstrated that the revived culture behaves normally in that it grows vigorously in N-containing medium but is unable to fix N2 and stops growing when transferred to N-free solution. DJ42 growth and derepression in the Microferm fermentor. From an overnight culture (Fembach flasks), an inoculation was performed into 21 L of sterile medium in the Microferm fermentor. The temperature was maintained at 29-31 °C throughout the incubation, whereas the agitation rate and the air flow rates were ca. 250- 270 rpm and ca. 25 L/min, respectively. The culture reached exponential growth and was harvested at ca. 131 Klett. The cells were then resuspended into a nitrogen-free medium, and the incubation continued for another ca. 4.5 hours. Growth on residual nitrogen initially continued for ca. 2 hr, after which the cell density remained relatively flat The bacteria were then harvested and frozen in liquid nitrogen. Cell breaking was done after resuspension into 1.5 volumes of 25 mM Tris, pH 8.0. Two cycles of breaking were performed using the 40 ml French Pressure Cell to yield dark brown lysate. Centrifugation for 100 min at 20,000 rpm in JA-20 rotor followed by volume measurement (ca. 120.5 ml), pH adjustment and DT addition yielded ca. 124 ml of extract which was subsequently frozen in liquid nitrogen. The extract was later found to have a protein concentration of ca. 19.8 mg/ml. Reconstitution assays of this new extract with old FeMo-cofactor samples yielded a satisfactory activity of 58.6 nmol C2H4/min.mg compared to a specific activity of 50.0 for a sample from Virginia Polytechnical Institute, thereby making the extract usable for evaluations of new cofactor preparations. 153 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. In vivo experiments: A. vinelandii culturing. The preparation and handling of aqueous solutions of cfr-dimethyldiazene has been described. 10 The modified Burk's medium used for/4, vinelandii growth contained in 1 L: 20 g sucrose, 1 mmol Mg2+ (as MgSfXi), 0.2S mmol Ca2+ (as C a d i) > 10 pmol Mo (as Na2Mo0 4 > , 100 pmol Fe3 4 1 (as Fe(m)-EDTA complex made from mixture of FeS0 4 , KOH, and EDTA through which air was bubbled until a stable burgundy color was obtained), 6 mmol phosphate buffer (pH 7.4). The sucrose, dissolved in the phosphate buffer, was autoclaved separately from the metal components to avoid the formation of precipitate. The cultures were maintained generally by regular subculturing into 25 ml fresh N-free medium in 125 ml side-arm flasks as well as growth on Petri plate with the same medium (2 % purified agar, Becton Dickinson), and subsequent selection of single colonies. Liquid cultures were regularly examined on microscope (hanging drop slides) to verify the cell shape, motility and absence of foreign species. Maintenance growth as well as the in vivo experiments (except where noted otherwise) were performed in a warm room at 30 deg C on rotary shaker at 200-250 rpm. Growth curves were constructed by turbidimetry on a Klett-Summerson Model 800-3 photoelectric colorimeter (red filter) with distilled water used as a zero-Klett calibration standard. In vivo experiments: assays and analyses. Except otherwise noted, the in vivo reduction assays were carried out in 21.5 ml vials fitted with septum stoppers. The bacterial culture was added first, followed by 3-4 cycles of pumping and argonation, after which the substrate stock solution was injected, and the mixture was pumped and argonated three more times. The assay was initiated shortly thereafter by injection of oxygen to ca. 0.1 atm. Gas aliquots (100 pi) were withdrawn at designated time points and analysis for methane/ethane/ethylene/acetylene 154 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. was performed on Shimadzu GC-14A gas chromatograph (assay vials were being incubated right next to the instrument) equipped with Porapak N column and flame ionization detector as described elsewhere.8 Analysis for methylamine and/or ammonia was performed by withdrawal of liquid aliquots (assay vials were incubated in the warm room), removal of cells and other particulates by a 10 min centrifugation on an Eppendorf microfuge (14,000 rpm), and derivadzation of a 0.4 ml portion with dansyl chloride for HPLC detection as described previously.8* 11 RESULTS A N D DISCUSSION Analysis of the FeMo protein before and after the modified desalting step. Biuret analyses performed showed that the total amounts of protein contained in die permeate (5.11 g) and the diafiltered FeMo ("FeMol diaf.", 17.74 g) matched the total amount of starting protein (23.98 g) very well (ca. 96 %) as seen in Figure 2. This indicated that very little, if any, retention of protein on the membranes had taken place. The convenience and utility of the Amicon system as applied to this purification step consists not only of significant time savings (2 hr/one operator versus overnight) but also of increased operational simplicity (outside versus glove box) and additional purifying power, the latter due to the fact that the larger MW-cutoff filter (100 kDa vs. 50 or 30) used leads to the removal of more contaminating proteins along with the desalting. In addition, the shorter time spent by the protein at room temperature minimizes its degradation. To our knowledge, such a step for processing of the FeMo protein of Azotobacter vinelandii nitrogenase has not been previously detailed. Further improvements to the existing FeMo purification part of the nitrogenase purification protocol1 * 2 could include the use of another 155 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. recently developed Amicon product, the Mini Plate Bioconcentrator, which would be ideal for processing of protein samples in the 50-200 ml volume range. Material and Activity Balance from the Purification. As seen on Fig. 1, a total of ca. 24 g of protein with SA of 488 nmol/min/mg and total number of units of 11.7 million was purified to five fractions of FeMoII protein containing a total of 10.13 million units. Out of these five fractions, two were taken and subjected to the crystallization protocol. From a total of 7.65 million units (FeMoIL 1 & 2), a total of 5.83 million (ca. 77 %) was obtained mainly in the form of FeMoS (4.94 million). The relatively low yield of FeMoX (0.82 million) can be explained by the fact that the protein for crystallization was not concentrated sufficiently to allow a maximum yield. With regards to the specific activities of all the analyzed fractions, it must be pointed out that since ten samples were assayed simultaneously, the complete Fe/FeMo titration curves were not obtained and some specific activity results may have been underestimated. Analyses of the formate-exchanged FeMo protein and FeMo-cofactor preparation. The FeMoX obtained into ammonium formate buffer was assayed with purified Fe protein to verify that its integrity was maintained after the diafiltration step. A value of almost exactly 2000 for the specific activity was obtained, thus indicating that the FeMo protein has undergone all the treatments up to the cofactor preparation without losing its activity. The DJ42-cofactor reconstitution titration placed the cofactor activity ca. 3 units per pi, which is within the values obtained previously.7 The total number of units in the whole preparation would be 3,000 pi x 3 units/pl = 9,000, or ca. 27 % of the starting FeMo protein units (1.2 ml x 14 mg/ml x 2000 units/mg = 33,600). The visibly low concentration 156 with permission of the copyright owner. Further reproduction prohibited without permission. of the cofactor preparation has very likely led to a distortion in the assay as too high an NMF concentration interferes with proper reconstitution, and may have led to a significant underestimation of the true activity. Preferably, more concentrated cofactor samples should be used in these assays in order to avoid the above NMF effect In vivo experiments: methylamine tolerance tests. In the first test Azotobacter vinelandii culture (31 ml), grown in 125 ml side-arm flasks, was allowed to reach 75 Klett (doubling time of 2.6 hr), and at this point 36 pi of 0.1 M of methylamine solution was added to a final concentration of 0.12 mM. The incubation was resumed although a minimal perturbation of ca. 3 Klett was recorded upon methylamine addition, the culture growth resumed at the same rate (doubling time of ca. 2.6 hr), which was maintained until ca. 110 Klett. When the same culture reached ca. 120 Klett, a dilution was performed by removal of 10 ml culture and addition of 10 ml fresh sterile N-free medium, followed by the addition of 6 pi 11.34 M methylamine stock solution to a final concentration of 2.2 mM. Upon this dilution, the cell density went down to 74 Klett, and the growth rate reached the previous levels within ca. 30 min. An examination of a hanging drop slide at this point revealed some cell clumping but did not show any major loss of motility. The exponential grow was maintained at least until 100 Klett as seen on Figure 3. A conclusion was made that methylamine does not interfere with the cell growth within the general limits of 0 -1 mM. At the scale of this experiment, such final molarity would correspond to a total of ca. 31 pmol of methylamine product formed. Several additional tests were carried out to confirm the above tolerance levels. In one of them, addition of methylamine to 0.47 mM again allowed the cells to continue growing, whereas in a separate experiment, the addition of methylamine to ca. 3.3 mM led to a complete abolition of the cells' growth as evidenced by die flat Klett reading for next 3 hr 157 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. of observation. In addition, hanging drop sample taken ca. 2 hr after the mixing showed an extensive clumping and widespread absence of motile cells. A final conclusion was made that formation (at least, not a gradual one) of methylamine at levels above ca. 1.5 mM may have a negative effect on the cell growth, or at least distort the results of any other kinetic experiment However, at the molarity levels of or-dimethyldiazene used as well as the expected in vivo enzymatic activity of nitrogenase, the maximum molarity of any methylamine formed was not expected to be above 0.1 mM. Some preparations of cu-dimethyldiazene are done in D2O in order to better evaluate the presence of impurities and/or breakdown products in the substrate stock solution. Tests were done to determine the effect (if any) of added D2O to the growing A . vinelandii cultures. Upon addition of an equal volume of D2O, the only effect observed was the net Klett dilution: the growth immediately resumed at the pre-dilution rate and repeated examinations (hanging drop) showed no apparent changes in the cells' shape and motility. Additions to growing cultures of both methylamine (0.47 mM) and D2O (1:1) also did not show adverse effects. Attempted Detection of Methane and/or Other C-H Products. A preliminary experiment utilizing an irradiation mixture of trans- (ca. 0.1 M) and cw-dimethyldiazene (ca. 0.03 M) (with traces of hexane and other impurities) revealed a presence of methane and ethane in the head space, the amounts of which did not change with time over the period of 0 - 32 min. Additional controls with culture and water, culture and QHj (in vivo acetylene assay), and substrates with water showed that the origin of the gases in the head space was from the low quality trans- / cu-dimethyldiazene preparation, and that the AvOP culture itself (70 and 92 Klett respectively for the substrate and control assays) was not contaminated. All subsequent experiments were done with completely 158 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. frans-dimethyldiazene-free preparations of the cis isomer, which were additionally pumped and argonated 3-5 times to ensure complete removal of background methane. Vials with substrate incubated with sterile AvOP medium were routinely included as controls. In another set of assays, bacteria (1 ml, 89 Klett) were incubated in 5 ml vials with equal volume of cu-dimethyldiazene (assay concentration ca. 59 mM) for up to 43 min. GC analysis of aliquots taken at 2, 19 and 43 min showed an extremely small methane peak, the area of which (corresponding to ca. 0 .1 nmol in the total head space) did not change over time. A simultaneous assay of 1 ml culture with 150 pi C2H2 in 5 ml vial, however, revealed that the cells' specific activity was only ca. 0 .0 1 1 nmol C2H4/min.ml.Klett Experiments with AvOP in the past have shown that actively growing cells generally display SA values in the order of 0.1 - 0.2 nmol/min.ml.Klett Therefore, it was still impossible to conclude anything definitive regarding the reducing ability of AvOP with respect to cw-dimethyldiazene. The experiments were generally rendered difficult because of the need to synchronize the production of highly concentrated cis- dimethyldiazene solutions free of impurities and the availability of actively growing cells displaying sufficiently high specific activity (subject to verification by a separate C2H2 assay). An actively growing culture (2 ml, 81 Klett, doubling time of ca. 2.7 hr) was used in two simultaneous assays with 2 ml cts-dimethyldiazene (60 mM assay concentration) and 500 pi C2H2, in separate 10 ml assay vials. The vial size and the culture volume were increased in attempt to provide larger surface area for gas-liquid contact as well as to hopefully generate a larger amount of product A total of three gas samples were taken from each vial for the 41 min duration of the assays. The m-dimethyldiazene assays showed the presence of < 1 nmol of methane (GC peak area of ca. 40 pV-sec), which again did not change as a function of time, whereas the C2H2 assay yielded a specific activity (based on 159 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. the total 41 min of incubation) of 0.11 nmol C2H4/min.ml.Klett. The amount of ethylene formed in that assay was ca. 700 nmol. If one takes into consideration the concentrations of C2H2 (»K m ) and cfr-dimethyldiazene (=Km) in these assay vials relative to their nitrogenase Km values, as well as the relative Vmax values for these two substrates (SA of ca. 2000 for QHj and ca. 58 for CH4 formation from cu-dimethyldiazene), then assuming similar behavior of m-dimethyldiazene in the in vivo system, the expected amount of methane produced should be around 10 nmol • a value clearly not detected in any of our experiments. Since the minute methane peak as well as another, yet smaller in size, signal were persistently showing in all samples, additional tests were done in attempt to trace their origin. Injection of 100 pi 0 2 taken directly from the supply cylinder resulted in appearance of exactly the same peaks. Moreover, in this oxygen sample they were represented at virtually the same levels (accounting for injection errors) as in the bacterial assay head space samples, after taking into consideration the partial pressure of O2 in those. It was concluded that no detectable methane evolution results from the co-incubation of A. vinelandii cultures with cfr-dimethyldiazene. Testing the Liquid Phase for Methylamine. Previous experiments with mutant strains have shown that the supernatants obtained even from high cell density cultures (>100 Klett) do not contain any substances that interfere with the HPLC derivatization and detection of methylamine and ammonia. Just prior to our current in vivo assays, control tests were done again. Aliquots (0.4 ml, duplicates) of HPLC grade water, sterile medium, and supernatant from AvOP actively growing culture (79 Klett, doubling time 2.9 hr) were derivatized and analyzed by HPLC. All samples showed no detectable methylamine, placing the upper limit at around 0.1 nmol. 160 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. To test for methylamine formation, AvOP cultures (2 ml, 114 Klett) were incubated for 30 and 45 min with equal volumes of m-dimethyldiazene solution (58 mM in assay) into two separate 21.5 ml vials. As controls, two other vials were include (at 30 and 45 min) which contained sterile medium and the substrate in the same amounts. HPLC analysis showed large methylamine peaks in all four vials (in millions of pv-sec): 2.18 (medium control at 30 min), 1.81 (AvOP at 30 min), 2.10 (medium at 45 min), and 1.99 (AvOP at 45 min). These peak areas corresponded to ca. 1.7 pmol of methylamine in die assay vials and were at least a hundred fold higher than the amounts reasonably expected. To verify these results, additional experiments were performed with a different, newly synthesized batch of c/s-dimethyldiazene, which showed clean lH NMR and UV spectra. Four (two duplicates) 21.5 ml vials containing 1.5 ml each c»*dimethyldiazene (55 mM) and AvOP (83 Klett, doubling time 2.4 hr) or medium were incubated for 35 min. The medium used as blank was obtained by removal of the bacteria (microfuge, followed by passage through 0.2 pm filter) from the same 83 Klett culture used for the positive assays. HPLC analysis of aliquots showed essentially the same result: large (2.3-2.7 million pV-sec) methylamine peaks were present in all four samples. It was concluded that the high methylamine background is created by a rapid ris-dimethyldiazene breakdown in the presence of the medium components. Additional tests involving test tube incubations (30 min) of either HPLC grade water or cw-dimethyldiazene solution with water (1:1) (2 sets, triplicates) followed by HPLC analysis showed that, except for one ds-dimethyldiazene - containing vial which showed a very small (ca. 11,000 pV-sec or 1 nmol) methylamine peak, no appreciable amounts of methylamine are contained within or formed upon the incubation of the cis- dimethyldiazene stock solution at these conditions. The in vivo tests were repeated for a third time using 1 ml each m-dimethyldiazene and AvOP culture (107 Klett, doubling time 161 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 2.3 hr) or medium (in duplicates). Again, after 35 min incubation, the same large methylamine peaks (1.73-1.90 million pV-sec) were observed in all four vials. It was concluded that at the conditions employed no methylamine formation as caused by the action of the actively dividing and fixing A. vinelandii bacteria on cu-dimethyldiazene can be detected. In addition, in all these assays, with exception of the derivadzation system and the methylamine background peaks, no ammonia or other product peaks were observed. One possible reason for the above observations may be that a compound in the bacterial growth medium catalyses the tautomerization of m-dimethyldiazene to the isomeric formaldehyde methylhydrazone, which in turn hydrolyzes in the presence of sucrose and/or other reducing agent to methylamine. The tautomerization of diazene would essentially lead to its depletion and thus could explain the failure to detect methane as a possible reduction product To circumvent the problem, it should be possible to conduct growth and assay experiments involving selective replacements of the bacterial medium components, thereby eliminating the agent causing the unwanted tautomerization. REFERENCES (1) McKenna, C. E.; Nguyen, H. T.; Huang, C. W.; McKenna, M. C.; Jones, J. B.; Stephens, P. J. In From Cyclotrons to Cytochromes (M. D. Kamen Symposium); Kaplan, N. O., Robinson, A. B., Eds.; Academic Press: New York, 1982, p 397- 416. (2) Gemoets, J. P. Ph.D. Thesis Thesis, U. of Southern California, (1990). (3) Simeonov, A. M.; McKenna, C. E. 1997, in preparation. (4) Malinak, S. M.; Simeonov, A.; Mosier, P. E.; McKenna, C. E.; Coucouvanis, D. J. Am. Chem. Soc. 1997, 7 /9 , 1662-7. (5) McKenna, C. E.; Huang, C. W. Nature 1979,280, 609. (6) Shah, V. K.; Brill, W. Proc. Natl. Acad. Sci. U.SA. 1977, 74, 2768-71. (7) Bravo-Leerabhandth, M. Ph.D. Thesis Thesis, U. of Southern California, (1990). (8) McKenna, C. E.; Simeonov, A. M.; Eran, H.; Bravo-Leerabhandh, M. Biochemistry 1996,35, 4502-14. 162 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. (9) Brigle, K.; Weiss, M.; Newton, W. E.; Dean, D. R. J. Bacteriol. 1987, 169, 1547-53. (10) Simeonov, A. M.; McKenna, C E. J. Org. Chem. 1995,60, 1897-9. (11) Bravo, M.; Eran, H.; Zhang, F. X.; McKenna, C. E. Anal. Biochem. 1988, 175, 482-491. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 1. Arrangement of the Amicon 2000 ldt outside the glove box. Shown are the 2 L reservoir (a), spiral-wound cartridge (b), masterflex pump (c) with speed controller (d), 3 L flask for diluting buffer (e), and effluent collection flask (f). 164 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. F igure 2. Flowchart showing the FeMo protein purification and analysis. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. FeMol comb. 2200 mL, 23.98 g, 11.7e6 units, 100% SA =488 Permeate 3650 ml, 5.11 g Diafiltration/concentration in Amicon spiral-wound carttridge system; MW cutoff of 100,000 t FeMol diaf. 1570 ml, 17.741 g, 10.09e6 units, 863 % SA =569 DEAE Sepharose Fast Flow; K100/45 column; 33 cm height; 0.15 and 0.25 M NaCl/Tris step gradient FeMoII early: 370 ml, 1.517 g, SA 1016,1.54e6 units, 13.2 % FeMoII.l FeMoII.2 FeMoII3 FeMoII.4 600 ml, 4.200 g, SA 1266,532e6 units, 45.5 % 540 ml, 2.133 g, SA 1092,233e6 units, 19.9 % 410 ml, 1.066 g, SA 697,0.74e6 units, 6 3 % 290 ml, 0.452 g, SA 438,0.20e6 units, 1.7 % Desalting/concentration in Amicon 2000 and 404 cells to <0.04 M salt; 1 hr incubation at 38 deg C; centrifugation, washing with 25 mM Tris, FeMo protein crystalls solubilization with 0.25 M NaCl/Tris FeMoS: 250 ml, 4.238 g, SA 1167,4.94e6 units, 64.5 % FeMoX: 57 ml, 0.456 g, SA 1791,0.82e6 units, 10.7 % FeMoS': 43 ml, 0.056 g, SA 1202,0.07e6 units, 0.9 % Repeat of the crystallization protocol FeMoX2: 30 ml, 1.26 g, SA 1944,2.45e6 units Diafiltration into formate buffer I jd: * FeMoX diafiltered: SA 2000,33,600 units FeMo-cofactor isolation protocol FeMo-co: 3 ml, 3 units//*!, 9000 units, 27 % of FeMoX diafiltered Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 3. Survival and further growth of A. vinelandii culture on N-free medium after addition of methylamine and dilution. See Experimental for details. 168 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. - 1 0 1 2 3 4 Time, hr Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER 6 INTERACTIONS OF MONO-M ETHYLDIAZENE WITH AZOTOBACTER VINELANDII NITROGENASE Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. INTRODUCTION Our work with two unstrained azo homoiogues of diazene, trans- and cis- dimethyldiazene, has demonstrated their utility as mechanistic probes for nitrogenase mechanism.1* 3 In an attempt to further expand the scope of these probes, we turned our attention to monomethyldiazene. The mechanistic importance of mono-methyldiazene can be summarized as follows. Simply on the basis of its size, it brings us one step closer to a putative diazene intermediate, as one of the methyl groups present on the previously studied dimethyldiazene substrates is absent here, thereby removing part of the steric bulk. More importantly however, a hydrogen atom is attached to one of the azo nitrogens which exactly mimics one “half’ of the diazene intermediate. Unlike the previously studied diazirine and dimethyldiazenes1 ’ 2, there is no longer a symmetry here with respect to the N=N bond, a fact which should allow in principle the azo group’s approach to the active site to be studied. If one nitrogen atom of the azo group is partially bound inside the FeMo cofactor cavity (end-on), and the other one remains outside, that “outside” nitrogen would be expected to be preferentially protonated during subsequent steps. In such a case, mono methyldiazene should bind inside the cofactor cavity with its NH end, and upon reduction, should form methane together with ammonia, and possibly methylamine. If, however, a side-on binding takes place, one might expect preferential reduction at the azo bond leading to formation of methylamine and ammonia, but little or no methane. Although mono-methyldiazene might appear to be an obvious candidate for nitrogenase substrate reduction mechanistic studies, the only research on this unstable compound related to nitrogen fixation mechanisms involved iron-sulfur model cluster system in which mono-methyldiazene was not added to the model compound in free form 171 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. but was rather generated by oxidation of already coordinated methylhydrazine. A look in the literature shows that mono-methyldiazene has not been very popular in general as a subject of study. A search by registry number in the Chemical Abstracts On-line database yielded only 14 references for the trans- and 9 for the cis-isomer. Mono-methyldiazene is reputedly readily generated as its trans-isomer by the reaction of N-methylhydroxylamine and hydroxylamino-O-sulfonic acid in the presence of base.4 H 2NOSO3* » » [:NH1 --------------- ► [CH 3N(OH)NH2] „ trans-CH3 N=NH O H * CH3NHOH *h 2 0 Most of the time, direct manipulations of the compound have been precluded by its instability and the related need for use of low temperature vacuum line manipulations4* 8. Thus, it has been reported that condensed rra/u-mono-methyldiazene decomposes within a few minutes to Nj/CH4 when allowed to warm up to room temperature in a vacuum line.4* 6 In addition, cfr-mono-methyldiazene has been virtually unstudied because it reportedly decomposes even at temperatures slightly above -100 °C.8 However, it had been noted that the N-deuterated mono-methyldiazene (CH3 -N=N-D) is significantly more stable.3 We reasoned that, in analogy with cu-dimethyldiazene,9 this compound might be stabilized sufficiently in D2 0 to permit some short-timescale nitrogenase assay experiments. The spontaneous formation of N2 , a substrate of nitrogenase reduced to NH3 , a possible diazene product, and of CH4, another possible diazene product, also create serious experimental challenges. EXPERIMENTAL SECTION General. Reagents for the preparation of mono-methyldiazene, N-methylhydroxylamine hydrochloride (98 %) and hydroxylamino-O-sulfonic acid (97 %), were purchased from 172 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Aldrich Chemical Company and stored in Drierite-filled jars. In addition, the acid was kept in a refrigerator in accordance with the manufacturer’s recommendations. During die course of this work, it was noticed that stock solutions of hydroxylamino-O-sulfonic acid (HOS) generally decomposed after several days when kept at room temperature (UV, data not shown). It was decided, therefore, to prepare HOS solutions only up to 1 hr before the experiments. Human hemoglobin (MW 64,500) was obtained from Sigma and stored in a desiccated jar at -20 °C Stock solutions (2.5xl0's M, 1.0x10“ * M in heme units) were prepared in Tris buffer (0.5 M, pH 8.1) and stored at the same temperature. lH NMR spectra were recorded at 250.13 or 360.13 MHz and 13c NMR spectra at 62.89 or 90.566 MHz on Bruker AC-250 or Bruker AM-360 spectrometers, respectively. Chemical shifts were recorded in ppm relative to internal CDC13 (5 = 7.24, H; 8 = 77.0, 1 3 C), internal D2 0 (8 = 4.63, H), internal CD3 CN (8 = 1.93, 'H; 8 = 1.3, * 3 C) and internal methanol (8 = 3.2, 'H; 8 = 49.0, l3 C). Coupling constants (J) are reported in Hz. Ultraviolet (UV) spectra were recorded on Shimadzu UV-260 spectrophotometer. The extinction coefficient of mono-methyldiazene in water (e = 24)4 was used to measure its concentrations in aqueous solutions. The literature values were verified by NMR spectra of samples to which known amounts of HPLC-grade methanol were added, followed by peak integration. Generation of mono-methyldiazene in NMR tubes. The reaction was done on a volume scale of 1.0 ml. NMR tube (5 mm) was fitted with a rubber septum stopper and flushed with argon for ca. 5 min. Separately, stock solutions of the three reagents were prepared as follows. In 300 pi degassed D2 0 were 173 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. dissolved 22.6 mg HOS (200 pmol), and in 660 pi degassed D2 0 were dissolved 16.7 mg N-methylhydroxylamine hydrochloride (Me-HA, 200 pmol); 40 pi 30 % NaOD (99+ % atom D, Aldrich) were added subsequently to the Me-HA solution. The basified Me-HA solution was transferred to the NMR tube and the initial proton NMR spectrum was recorded. In some experiments, small volumes (5-10 pi) of degassed HPLC-grade methanol were injected prior to recording the spectrum. The NMR tube was then removed from the instrument, the HOS solution was injected, the tube was inverted gently 5-10 times and then returned for recording additional NMR spectra. A timer was started approximately 3 seconds after the injection of the HOS solution. Generation of mono-methyldiazene in UV cells. In a procedure similar to the one described for the NMR experiments, stock solutions (200 pmol scale) were prepared in degassed D2 0 . A quartz cuvet fitted with a rubber septum stopper was flushed with Ar and filled with the basified Me-HA solution. Although an instrument baseline was generally run against degassed water, subsequent tests revealed that the Me-HA solution does not shift that baseline significantly (data not shown). The HOS stock solution was brought to the instrument room inside a Sample-Lok syringe and injected into the cuvet when needed. For time-course measurements, a timer was started approximately 3 seconds after the HOS injection. When the interactions of mono-methyldiazene with hemoglobin were studied, separate sets of cuvets (for recording the initial spectra of the two reactants) were used and base lines were run prior to each titration. 174 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Preparation of mono-methyldiazene on a vacuum line and its subsequent dissolution and storage. The compound was prepared by the method of Ackermann7 with only small modifications. Generally, a scale of 12 mmol (up from 3.3 mmol in Ackermann’s protocol) was utilized for which 1.0 g Me-HA (12 mmol) and 1.7 g HOS (IS mmol) were used. In the original method, HOS was the limiting reagent but during our NMR studies it was found that some of the excess Me-HA had co-distilled with the final product (see Results). It was subsequently decided to change the reagents’ ratio. Immediately after weighing, the 1.7 g HOS were transferred to an argonated 300 ml three-neck round-bottom flask which also contained a magnetic spin bar; thus, the solid HOS was kept under argon until the flask was attached to the vacuum line. In a 250 ml round-bottom flask containing argonated 120 ml 1.00 M NaOH was dissolved 1.0 g Me- HA. The three-neck round-bottom flask was attached to the vacuum line at its middle neck and was supported with a magnetic stirrer. To one of the two side necks, the 250 ml flask was attached via a standard taper arrangement The connection was L-shaped and permitted simultaneous pumping of the two flasks as well as the rotation of the solution-containing flask for the addition of the basified Me-HA to the solid HOS. After the flasks were connected and the magnetic stirring tested, the whole system was pumped very cautiously (gassing) for several minutes to remove air from the system. The lighting in the laboratory was reduced as a precaution, and the side-arm flask rotated to release its content; magnetic stirring was started. The solid HOS dissolved within seconds after the addition of the Me-HA/NaOH solution, and vigorous gas evolution was persistent for at least 5-7 min. During that period, the mono-methyldiazene being formed was allowed to condense in a U-shaped trap cooled by liquid nitrogen. After the reaction had subsided, the trap was isolated, and the reaction system was vented to atmospheric pressure by opening the vacuum line against a Dewar full of liquid nitrogen, thus allowing pressure 175 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. equalization but not an immediate contact of oxygen with the reaction mixture. This step was included as a precaution because early reports pointed out the dangers in exposing mono-methyldiazene gas to oxygen (explosion with flame).5’ 6 The set of both flasks was then quickly detached from the vacuum line and the liquid contents diluted with a large amount of tap water. The vacuum line was pumped once again with a gas collection tube attached next to the mono-methyldiazene trap. Quick warming (heat gun) was then applied to the U-shaped trap along with liquid nitrogen cooling of the gas collection tube. This step allowed the mono-methyldiazene product to quickly distill with minimal decomposition to N2 and CH4 (small amounts of non-condensable material were evident, however) and separate from the small amounts of water also present in the trap. The yellow solid mono- methyldiazene was then dissolved in ca. 35 ml degassed Dz O by allowing the joint of the gas collection tube to warm up, fitting it with a rubber septum stopper, flushing with argon and injecting the water along with the removal of the rest of the tube from the liquid nitrogen. Upon water injection (intense shaking), the tube warmed up gradually and a solution was formed within ca. 1 min of shaking; some pressure developed during the process due to further diazene decomposition. The tube was then vented by attachment to a gas manifold (Ar) and the stock solution of mono-methyldiazene aliquoted into several sealed and argonated 5 ml assay vials (2-3 ml per vial). The vials were then frozen and stored in liquid nitrogen containers. For subsequent NMR and UV analyses, the mono- methyldiazene stock solutions were injected into septum-sealed and argon-flushed NMR tubes or UV cuvets as necessary. Nitrogenase assays in the presence of mono-methyldiazene: effects on hydrogen evolution and acetylene reduction. Nitrogenase assays were generally performed as described previously.1 The assay vials used were of 5 ml volume with the liquid phase usually kept at 1.0 ml. Due to the diazene instability, great care was taken with respect to planing the additions of reagents 176 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. and especially the timing of mono-methyldiazene stock solution thawing and use. In general, the preparation and dispensing of ATP-generator and dithionite solutions were completed first, and then the vial with diazene stock solution was removed from the liquid nitrogen container and allowed to warm up. During the diazene thawing, the necessary Fe and FeMo proteins were taken from the liquid nitrogen container and also allowed to warm up. In all assays, the mono-methyldiazene solution molarity was measured immediately after the completion of its thawing by recording the UV spectrum of an aliquot Immediately after the UV analysis, the actual assay was initiated by the rapid addition of enzyme and diazene. In the case of acetylene assay, the reaction was terminated by the injection of 0.25 ml 25 % trichloroacetic acid and the vials were analyzed for ethylene/acetylene content by GC separation of 50 pi gas aliquot When hydrogen evolution was being studied, the assays were terminated by the injection of 0.1 ml 1 M HCl/sat’d K I03 and immediately afterwards, 150 pi gas aliquots were analyzed for hydrogen on a Varian 3700 GC equipped with thermal conductivity detector.1 A set of assay vials containing 1.0 ml buffer as a blank and variable amounts of hydrogen (400- 1200 nmol per vial) was similarly incubated and treated with terminating reagent for generation of calibration curve. In more recent experiments performed to confirm the H2 and C2 H2 inhibition patterns observed initially, 22 ml assay vials were used. These were processed in an identical manner with exception of the head space aliquots used for the GC analyses which were 300 pi (thermal conductivity detection of Hj/N2 ) and 200 pi (flame- ionization detection of O tyQ ItyQ H ^ respectively. Hydrogen evolution assay in the presence of mono-methyldiazene: analysis of the liquid phase for diazene reduction products. After the completion of the hydrogen GC analyses above, the seals of the assay vials were removed to allow dithionite decomposition. The whole liquid phase of each vial 177 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. was transferred to an Eppendorf tube and centrifuged briefly (S min at ca. 10,000 rpm) to pellet die denatured proteins. The liquid samples were then stored at -20°C until needed. HPLC analysis of 50-400 pi aliquots by dansyl chloride pre-column derivadzadon and fluorescence detection was performed as described before.1 * 10 Specially-timed assays in search of methane as a reduction product. After a number of preliminary experiments, the following optimized assay protocol was used. Before the assay, ethane/methane calibration tests were performed on the GC using the standard conditions for ethylene/acetylene detection. A total of 16 vials were utilized which contained different ratios of QH^/CH^ Vials were initially loaded with 0.80 ml buffer, sealed and argonated. Just prior to the tests, the vials were vented, ethane (10-40 pi) and methane (10-25 pi) were added in different combinations, and the five-minute incubation was initiated after the addition of the remaining 0.20 ml buffer, thus imitating mono-methyldiazene stock solution addition. After 5 min of shaking at 30 °C, 100 pi head space aliquot was injected into the GC. Calculated molar (same as volume) ratios were compared to the obtained ethane/methane peak area ratios. A mean ratio [area CH^area C2 H6 ]/[nmol Cftynmol Q H J of 0.5185 (s.d. 0.0297) was obtained. The modified assay mixtures were prepared as follows. The ATP generator solution contained 100 pmol creatine phosphate (CP) per assay vial, up from the 25 pmol normally used; dithionite amount per assay was also boosted from 20 to 25 pmol. Both components were delivered via the same volume injections (0.40 and 0.25 ml per 1.00 ml assay liquid phase) but the initial stock concentrations were increased accordingly. The proteins and amounts used were FeGG38B, 100 pi (1.6 mg) and FeMoX2, 20 pi (0.84 mg). The assay was conducted with the help of Mr. S. Krause who was in charge of recording the exact time of the different manipulations. A total of nine vials were used, and these were divided 178 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. in the following manner four "positives" (“+”) containing pre-mixed nitrogenase, four "negatives" containing Tris buffer instead of enzyme, and one vial containing the same amount of enzyme and buffer (up to 1.0 ml liquid volume) but containing no mono- methyldiazene, and receiving C2 H2 injection for activity testing at the end of the experiment All vials were pre-incubated for ca. S min at 30 °C. Just prior to the start of the assay, the mono-methyldiazene stock solution was determined to be 32 mM (UV). The nitrogenase Fe and FeMo proteins were pre-mixed in a vial (Ar) about IS min before the beginning of the assay. Right before the 30 °C pre-incubation, the vented vials (still without enzyme and diazene) received 40 pi C2H6 injection as an internal GC standard (except for vial 9). The eight vials were processed in pairs, and were, from #1 to 8, lim e points recorded included addition of protein, mono-methyldiazene, removal of the vial from the shaker bath, withdrawal of the two 100 pi gas aliquots (Sample-Lok, followed by a second syringe whose content was injected into the GC instrument first), and the liquid phase reaction termination by the injection of 0.25 ml HCl/KIOj.CHyCjHg GC separation was monitored and recorder on a Macintosh Centris computer by using the Maclntegrator hardware/software package by Rainin Instruments. Vial 9 was processed immediately after the last pair of diazene vials, and there, after the S min shaking and one SO pi gas sampling, the vial was left on the bench top without shaking and sampled for ethylene formation two more times. Satisfactory specific activity of ca. 1000 units/mg was obtained from that vial. RESULTS AND DISCUSSION Formation of mono-methyldiazene: NMR observations. Immediately upon mixing of HOS and basified Me-HA solutions, vigorous gas evolution was observed and the NMR tube was essentially filled with bubbles. As mono- 179 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. methyldiazene is a gas at room temperature, it can be expected that initially the bubbles consist mostly of that compound but as the reaction progresses, there would be more methane and nitrogen in the gas phase as a result of the rapid mono-methyldiazene decomposition. Close examinations of lH NMR spectra did reveal the presence of a methane peak near 0 ppm (data not shown). The intense gas bubble formation in all mono- methyldiazene-containing solutions presented frequent problems especially when the solutions were relatively more concentrated (50-100 mM) or when longer spectral acquisition times were necessary. In the case of mono-methyldiazene generations in the NMR tube, the proton peaks were very wide and it was an usual occurrence to lose die instrument lock. Figure 1 shows the typical outcome of a reaction of HOS and basified Me- HA in an NMR tube. Since in all of the previous studies, the mono-methyldiazene chemical shift at around 3.8 ppm had been reported for acetonitrile solutions at ca. -40 °C,6 it was important to not just reproduce the diazene formation reaction but to establish the room temperature chemical shift of that compound in water. That later parameter would serve as one of the criteria for the mono-methyldiazene’s integrity and stability in such solutions. As evident from Figure 1, the starting Me-HA is represented by its expected singlet at ca. 2.4 ppm which goes down sharply in size a short time after the addition of HOS. A new peak at 3.7-3.8 ppm appears, and grows within 1-3 min to reach a relatively constant size. As this peak appeared reproducibly in all tests, and because of its very close proximity to the mono-methyldiazene acetonitrile chemical shifts reported in the literature, it was concluded that in our reaction the title compound was indeed made. The successful formation of mono-methyldiazene was confirmed by UV spectroscopy (see next section) and by NMR studies of D2 0 solutions of the compound prepared on a vacuum line. When these neutral solutions were left in the NMR tubes for 15-20 min without agitation, the bubble formation subsided, and subsequent gentle tapping 180 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. yielded relatively bubble-free solutions, thereby permitting some longer NMR studies. In some cases, the fresh D2 0 solutions were diluted two- to three-fold with degassed Dz O to further minimize the gas formation. The proton spectra of such solutions revealed a singlet at 3.72 ppm. Prolonged standing (ca. 14 hr) of such solutions at room temperature but without any agitation did result in ca. 50-60 % decomposition of the diazene but with no observable formation of hydrazone (peaks were not discernible in the spectrum) (Figure 2). Such absence of tautomerization indicates that enzyme assays can be performed with mono- methyldiazene without die risk of the later quickly converting to another potentially nitrogenase-reacdve form. NMR studies of mono-methyldiazene stock solutions added to nitrogenase assay system containing Hepes buffer, ATP generator and dithionite also showed that the compound only decomposes to methane and nitrogen and does not form observable quantities of hydrazone during at least 5-8 min. With relatively stable solutions of mono-methyldiazene at hand, an attempt was made to record its 1 3 C NMR spectrum. As the available stock solutions were too dilute for quick NMR analysis, a sample for an overnight run was prepared by allowing an NMR tube filled with 0.5 ml solution to “stabilize” by standing for 30 min with periodical gentle tapping. In addition, 20 pi degassed HPLC-grade acetonitrile was added as both molarity and chemical shift internal reference. Based on our previous studies of cis- and trans- dimethyldiazenes,9 we expected the mono-methyldiazene to have a I3 C chemical shift in the general range of 40-60 ppm; therefore, acetonitrile (119 and 1.3 ppm) provided a convenient referencing window. After an overnight run, the spectrum in Figure 2 was obtained. Along with the acetonitrile peaks and a singlet at 35.7 ppm (later assigned to contaminating Me-HA), the only other peak present was at 59.1 ppm. An expansion of the peak revealed that it was a 1:1:1 triplet with a coupling constant of ca. 3.9 Hz (Figure 2) which was apparently due to the coupling of the methyl carbon with the deuterium attached 181 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. to die second azo-nitrogen. The Jraras-configuration of the double bond particularly facilitates such long-range coupling.9 Proton spectra of the solution taken before and after the 1 3 C acquisition confirmed the integrity of the compound. Rate of mono-methyldiazene formation: UV studies. A very convenient way of studying the mono-methyldiazene formation and especially its rate is by performing die reaction in a quartz cuvet and recording the UV spectrum. In addition to just verifying the reaction rate, die UV measurements would be extremely helpful in establishing molarity values for the mono-methyldiazene stock solutions to be used in enzyme assays as well as the decomposition rate of such stock solutions for the duration of the assays. Figure 3 shows a typical set of UV spectra obtained from frequent scanning of a reaction between HOS and basified Me-HA mixed in the cuvet as described in the Experimental Section. It must be pointed out that just before the start of each scan, the cuvet had to be removed from the holder and tapped onto the wooden bench surface to remove the rapidly forming bubbles which would otherwise seriously distort the collected spectrum. As reported back in the 1970’s, the mono- methyldiazene presents a peak at 350 nm with £ = 24.4 In our experiments, the peak was very close to that value, at 346-347 nm, when the instrument default slit width of 2 nm was used; lowering die slit width to 0.2 nm resulted in reporting the X m a x as 350.1 nm. The value of the extinction coefficient was verified approximately (within 20 %) by obtaining a solution molarity by UV, and then, within 10 min, running a 'H NMR on the same sample with MeOH as an internal standard (data not shown). Figure 4 shows a time-course plot of mono-methyldiazene formation. It is evident that the reaction proceeds very fast and the mono-methyldiazene formation generally subsides after 5-6 min. 182 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Mono-methyldiazene tautomerization in basic medium. NMR spectroscopy of formaldehyde hydrazone. As mentioned earlier, die rate of the mono-methyldiazene formation reaction depends not only on the molarities of HOS and Me-HA, but also on the concentration of the free strong base.4 To accelerate the reaction, and thereby minimize the chances of mono-methyldiazene decomposition, one would ideally set up the reactants’ ratios in such a way that the base would be in a great excess. At high pH, however, mono-methyldiazene is known to tautomerize quickly, giving rise to the respective formaldehyde hydrazone. Indeed, when in the ‘H NMR reactions described above NaOD was used in excess, the newly-formed diazene peak soon began to diminish in size and the previously reported doublet of doublets for the hydrazone became observable (Figure S). The two doublets, typical of a pair of chemically non-equivalent protons with a short-range coupling (two- bond, “H-C-H”) of 11 Hz, appear at ca. 6.2 and 6.7 ppm. Although mono-methyldiazene quickly forms the hydrazone at high pH, that reaction is not instantaneous (data not shown). Therefore, in order to prepare formaldehyde hydrazone for additional NMR studies, mono-methyldiazene was generated on 200 pmol scale in a sealed 5 ml vial in the presence of ca. 4-fold excess of NaOD, and the vial was allowed to stay at room temperature overnight to ensure complete tutomerization. HPLC- grade methanol, serving as an internal standard, was added to the solution prior to NMR studies. Figure S shows the ‘H NMR spectrum of the hydrazone followed by a I3 C spectrum. On the latter, along with the methanol peak at 49 ppm, only one other peak is observable at 137.6 ppm. The above NMR data were consistent with previously reported values. H’12 In addition, a coupled I3 C NMR spectrum confirmed the assignment of the 137.6 ppm peak by revealing its splitting by the two non-equivalent protons. The above experiments establish a facile route to formaldehyde hydrazone and confirm its identity. Although it has been shown (see first section above) that mono- 183 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. methyldiazene does not form the hydrazone at the pH, ionic strength and time scale conditions of the nitrogenase assay, it appears that a separate study of the interactions of that compound with nitrogenase would be of importance as previous studies have shown that cyanide,1 3 cyanamide,14 and cyanate15 do interact with the enzyme. It is also worth mentioning that the carcinogenic effect of hydrazine has been explained by its reaction with endogenous formaldehyde to form formaldehyde hydrazone in the liver with the latter compound producing DNA-methyladng species upon activation by catalase.1 2 ’1 6 Formation of a complex between mono-methyldiazene and hemoglobin. It has been shown in the past that diazene itself is capable of forming complexes with a number of heme-containing compounds such as hemin, hemoglobin, myoglobin and legoglobin.1 7 To further verify the presence of mono-methyldiazene in our preparations, we attemped to detect complex formation between mono-methyldiazene and a heme- containing compound. Human hemoglobin was chosen for the experiments not only due to its availability, but also because of the fact that a demonstration of mono-methyldiazene binding to the protein-buried heme moiety in hemoglobin would most likely indicate that complex formation is possible with smaller heme-containing compounds such as hemin. As shown in Figure 6, the initial ferri-hemoglobin is converted into ferrous form by addition of dithionite. That spectrum, characterized by a peak at 558 nm, changes dramatically when a portion of mono-methyldiazene solution is added. Two peaks, a tall one at 555 nm, and a smaller at 526 nm, are readily observable, and their sizes grow further with the addition of more diazene. Two very similar peaks have been observed in the past, termed a (555 nm) and (3 (526 nm) bands, respectively, for the diazene (N2 H2 ) hemochrome.1 7 '18 The formation of a mono-methyldiazene hemochrome was verified by a “reverse” titration in which a pre-reduced ferrous hemoglobin was added to a dilute 184 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. solution of the diazene (Figure 6). The extinction coefficient for this complex at 526 nm was estimated to be about 18,000 (data not shown) and was within the range of values reported previously for other hemochrome s.18 Although detailed stability tests were not performed, observations of the above complex stored in the UV cuvet showed that it was stable for at least 2-3 hr suggesting stabilization of mono-methyldiazene upon binding to the iron center as was previously observed for N2 H2 -Hb and related hemochromes. M ono-methyldiazene inhibits hydrogen evolution catalyzed by A . vinelandii nitrogenase. When nitrogenase enzyme is supplied with electrons (reducing equivalents in form of dithionite in the in vitro system) and ATP, in the absence of any exogenous substrate, it would simply reduce the protons from the aqueous solvent to form hydrogen. Compound added to such a system generally has either no effect on the hydrogen evolution or acts as an inhibitor of this reaction. In the latter case, the compound can either act only as an inhibitor and not be transformed (reduced) by the enzyme, carbon monoxide being the most prominent example, or it can also be reduced by the enzyme, thereby forming its own reduction products. 13,19-22 Therefore, virtually every study of an alternative nitrogenase substrate begins with such hydrogen inhibition experiment Standard nitrogenase assays were performed in the presence of variable amounts of 33 mM stock solution of mono-methyldiazene. After 10 min of incubation at 30 °C, the reactions were stopped and the gas phases assayed for hydrogen by GC as detailed in the Experimental Section. Included among die vials was a set of H2 calibration standards. As evident from the gas chromatogram in Figure 7, vials containing larger concentrations of mono-methyldiazene had smaller amounts of hydrogen produced. A plot of the amounts of H2 formed as a function of mono-methyldiazene concentration shows a gradual decrease down to ca. 38 % of Jie amount formed in the absence of diazene, or a net deficit of ca. 500 nmol H2 (Figure 7b). Although data points at higher mono-methyldiazene 185 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. concentrations were not available, a preliminary calculation showed that it is not likely that the diazene would be able to completely suppress H2 evolution at infinite concentrations.2'22 The liquid phases of the above assay were saved and later analyzed for the presence of reduction products, such as methylamine and ammonia. As shown in a later section, methylamine was present in all assay vials, and its quantity increased as a function of the concentration of mono-methyldiazene. To further examine and confirm the effect of mono-methyldiazene on die nitrogenase-catalyzed H2 evolution, the above assay was repeated successfully for a total number of 5 times. In the last two experiments, the assay vial size was increased to 22 ml (21 ml head space volume) in order to minimize the effect of the formed N2 by lowering its partial pressure. The results from the last inhibition experiment are shown on Figure 8. Once again, examination of the curve indicates that 50 % inhibition of H2 evolution is effected by ca. 8 mM mono-methyldiazene. To assess the possible effect of the nitrogen formed from the mono-methyldiazene decomposition, the curve on Figure 9 was constructed by assuming that a complete decomposition of the diazene takes place instantaneously to form N2 + CH4. As evident from the curve, even in the assay vials containing the highest mono-methyldiazene concentrations, the maximum partial pressure of N2 is ca. 0.025 atm, or 25 % of its Km value. Thus, it can be concluded that the major inhibition effect comes from mono-methyldiazene. Additional support for this conclusion was obtained by examination of the size of the small broad peaks around 2.5 min (due to the co-elution of N2 and CHJ on the GC traces (TCD analysis) and comparing these with N2 peaks from air injections. Effect of mono-methyldiazene on acetylene reduction kinetics. Acetylene reduction to ethylene is one of the earliest established properties of nitrogenase and it serves as a convenient method for assaying the enzyme activity as well as 186 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. mutual inhibition studies to establish the relationship between QH j and the other alternative substrates. Acetylene itself is a rather peculiar inhibitor in that it inhibits N2 reduction non- competitively, but is itself inhibited competitively by dinitrogen.13’ 23 To study the effect of mono-methyldiazene on die nitrogenase-catalyzed acetylene reduction, a set of 16 vials was utilized containing C2H2 at 4 different initial partial pressures (0.006-0.042 atm, or 0-7 times Km), and mono-methyldiazene at 4 different concentrations (0-11.9 mM). The results are shown in Figure 10 where die amounts of ethylene produced are presented in a double-reciprocal (Lineweaver-Burk) plot. Although several data points were lost due to handling problems during the assay, the resulting straight line fits showed satisfactory correlation (i2 > 0.9). The mode of inhibition deduced from the data was likely to be either competitive or a mixed type; a pure non-competitive mode appeared much less likely. From an examination of the plot for a straight line with an x-intercept of about a half of the -1/Km value, an approximate value for the Ki of 6-8 mM can be deduced. It is noteworthy that a similar mono-methyldiazene concentration range can be deduced from an examination of the hydrogen inhibition curve for diazene levels effecting a 50 % decrease in H2 evolution (Figure 7). Thus, approximately the same concentrations, 6-8 mM, of compound are sufficient to effect a 50 % decrease in both acetylene and proton reduction and may be indicative of the Km value for mono-methyldiazene nitrogenase reduction. The above experiment was repeated by using 22 ml assay vials to minimize the possible effect of N2 (see Hydrogen Inhibition Results and Discussion). The Lineweaver-Burk plot shown on Figure 11 essentially confirms the results from the previous experiment The inhibition mode observed is of mixed-type. Methane formation from mono-methyldiazene decomposition and possible nitrogenase-catalyzed reduction. As mentioned in the Introduction, methane is one of die possible products resulting from a C-N bond reduction by nitrogenase. In order to differentiate between the methane 187 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. formed during the continuous mono-methyldiazene decomposition and the possible enzyme-catalyzed reduction, a number of preliminary experiments were performed with the goal to establish assay conditions that would allow a catalytic reduction to be delected against background CH4. Among the parameters changed as shown in the Experimental Section were the amount of total enzyme, which was boosted almost tenfold to 0.84 mg FeMo protein, the amount of creatine phosphate necessary to support nitrogenase turnover, and the amount of reductant Among the preliminary tests conducted were acetylene reduction time-course assays with different amounts of protein which were done to verify the existence of a linear phase in the ethylene evolution curve during at least the initial 5-7 min of the assay under these “extreme” conditions. The assay presented here was performed as described in the Experimental Section. From the possible 16 (8 vials x 2 GC samples per vial) methane data points, only one was lost due to a late start of the Maclntegrator acquisition. In terms of absolute peak areas, die values obtained from the duplicate samplings did not differ significantly from one another, thereby proving the reliability of the double-syringe sampling method. However, when the amounts of methane (nmol) were compared between positives and negatives, a surprising lack of correlation was observed in that among the four pairs of +/-, two were showing the positives to have produced larger amounts of methane, whereas the other two were showing exactly the opposite (also, to the same degree). A closer examination of the data revealed that within each pair the presence or absence of enzyme did not apparently matter, but that the larger amount of methane was detected in the second vial within the pair. The only way the two vials differed from one another was that the first vial received its diazene injection first from a syringe that contained stock solution sufficient for both vials, and then the second vial received the remaining half. It was noticed that even upon a short stay inside the syringe, the diazene stock solution quickly developed bubbles. It appears therefore, that the diazene delivery method was chiefly responsible for the observed 188 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. differences. In a future experiment, one might be able to eliminate this problem by delivering the diazene stock solution via single-portion injections for each vial. Methylamine as a possible mono-methyldiazene nitrogenase-catalyzed reduction product. If nitrogenase catalyzes the N=N bond reduction of mono-methyldiazene as described in the Introduction, methylamine, along with ammonia, should be detectable in the liquid phase. The HCl/KI03 -quenched assay reaction mixtures from the hydrogen evolution and methane formation experiments were analyzed by HPLC for the presence of methylamine in the assay liquid phases. The results are presented in Figures 9 and 10 respectively. As presented before (Figure 7), higher mono-methyldiazene concentrations caused a decrease in the amounts of hydrogen produced in a nitrogenase assay. An HPLC analysis of these same assay vials showed that methylamine levels are also dependent on the increase of the diazene concentration (Figure 12): although the first two vials (complete nitrogenase system without mono-methyldiazene) had significant methylamine background, the amounts in the later vials, when corrected for these background levels, also showed a smooth increase as mono-methyldiazene was increased. Further supporting the above finding was the HPLC analysis of the 8 vials from the methane detection experiment presented earlier. Even if methane formation could not be conclusively established or rejected in that experiment, the possible presence of methylamine in the liquid phase could be established unequivocally as the background mono-methyldiazene decomposition results only in formation of methane and nitrogen (which could only be nitrogenase-reduced to ammonia) but not methylamine (Fig. 2). Figure 10 shows that with the exception of vial 6, all other data points support methylamine formation in excess of background in all “positive” vials (see Experimental). As one half of the vials in this second experiment were no-enzyme mono-methyldiazene controls, the results from Figure 13 complement the ones presented in Figure 12 (variable diazene, plus 189 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. no-diazene control) and establish that the methylamine curve in Figure 12 is most likely a manifestation of its formation as a result of nitrogenase action and not an artifact of a background methylamine coming from the mono-methyldiazene stock solutions. In conclusion, mono-methyldiazene was successfully prepared and stored in D2 0 . Its stability at the conditions of a typical nitrogenase assay for at least S min was verified and the only products detected were the expected methane and nitrogen. The compound was shown to interact with the iron centers of reduced human hemoglobin to form a typical hemochrome complex. Although mono-methyldiazene has been considered for a long time extremely unstable and thus might appear impossible to evaluate as a nitrogenase substrate or inhibitor, the present preliminary experiments successfully demonstrate that, with appropriate modifications and inclusion of relevant controls, information can be gathered regarding its interactions with the enzyme. Although the numerous preliminary assays have so far failed to answer the question about methane as a reduction product, further improvements, that are currently under consideration, might help in conducting a conclusive experiment According to the preliminary results obtained thus far, methylamine appears to be a product of mono-methyldiazene reduction; further experiments are being designed with the goal of establishment of Km and Vm values for methylamine formation. The inhibition properties of mono-methyldiazene against acetylene and proton reduction appear to correlate with one another as represented by the similar Ki ranges deduced. This reinforces the conclusion that the compound does indeed interact with the active site(s) of nitrogenase and that further experiments with this new chemical probe should prove fruitful. 190 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. REFERENCES (1) McKenna, C. E.; Simeonov, A. M.; Eran, H.; Bravo-Leerabhandh, M. Biochemistry 1996,35, 4502-14. (2) Simeonov, A. M.; McKenna, C. E. 1997, in preparation. (3) Malinak, S. M.; Simeonov, A.; Mosier, P. E.; McKenna, C. E.; Coucouvanis, D. J. Am. Chem. Soc. 1997,119, 1662-7. (4) Ackermann, M. N.; Ellenson, J. L.; Robison, D. H. J. Amer. Chem. Soc. 1968, 90, 7173-4. (5) Ackermann, M. N.; Hallmark, M. R.; Hammond, S. K.; Roe, N. A. Inorg. Chem. 1972,11, 3076-82. (6) Tsuji, T.; Kosower, E. M. / . Am. Chem. Soc. 1971, 93, 1992-9. (7) Ackermann, M. N. Inorg. Chem. 1 971,10, 272-6. (8) Craig, N. C.; Kliewer, M. A.; Shih, N. C. J. Am. Chem. Soc. 1979, 101, 2480- 2. (9) Simeonov, A. M.; McKenna, C. E. J. Org. Chem. 1995,60, 1897-9. (10) Bravo, M.; Eran, H.; Zhang, F. X.; McKenna, C. E. Anal. Biochem. 1988, 175, 482-491. (11) Heaton, B. T.; Jacob, C.; Monks, G. L.; Hursthouse, M. B.; Chatak, I.; Somerville, R. G.; Heggie, W.; Page, P.; Villax, I. J. Chem. Soc., Dalton Trans. 1996, 61-7. (12) Lambert, C. E.; Shank, R. C. Carcinogenesis 1988, 9, 65-70. (13) Burris, R. H. In A Treatise on Dinitrogen Fixation; Hardy, R. W. F., Ed.; John Wiley & Sons: New York, 1979, p 569-604. (14) Miller, R. W.; Eady, R. R. Biochim. Biophys. Acta 1988, 952, 290-6. (15) Rasche, M. E.; Seefeldt, L. C. Biochemistry 1997,36, 8574-85. (16) Fitzgerald, B. E.; Shank, R. C. Carcinogenesis 1996,17, 2703-2709. (17) Hanstein, W. G.; Lett, J. B.; McKenna, C. E.; Traylor, T. G. Proc. Natl. Acad. Sci. U.SA. 1 9 6 7 ,58, 1314-6. (18) Falk, J. E. Porphyrins and Metalloporphyrins', Elsevier: New York, 1975. (19) Burgess, B. K. 1985, 543-549. (20) Burgess, B. K.; Lowe, D. J. Chem. Rev. 1996, 96, 2983-3011. 191 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. (21) McKenna, C. E. Chemical Aspects o f Nitrogenase; Pergamon Press:, 1980. (22) Rivera-Ortiz, J. M.; Burris, R. H. / . Bacteriol. 1975,123, 537-545. (23) Hwang, J. L.; Chen, C. H.; Burris, R. H. Biochim. Biophys. Acta. 1973, 292, 256-270. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 1. *H NMR of (a) basified Me-HA and the emergence of the mono- methyldiazene peak (after 1.0 min (b) and 1.9 min (c)) upon injection of HOS solution. 193 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1—f" 0.00 ^ 1 I 1 4 .N “I r_ I.M v— 1 T I . N 7.00 0.00 194 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. F igure 2 . lH NMR of mono-methyldiazene sample before (a) and after (c) an overnight 1 3 C NMR (b) acquisition. 195 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. m i -1 - - x - a o o i 1 « “ T" 8 0 7 196 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. F ig u re 3 . UV monitoring of mono-methyldiazene generation as result of the reaction between HOS and basified Me-HA mixed in the 1 cm cuvet at room temperature as described in the Experimental Section. Spectra were taken approximately every 1 min. 197 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. I & 1 < t= 1 min 0.000 300 500 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. F ig u re 4 . Time-course plot of mono-methyldiazene formation from a 200 /<mol reaction scale. Details are in the Experimental Section. 199 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 0 .1 5 □ □ □ 0 .1 2 5 0.1 < 0 .0 7 5 0 .0 5 0 .0 2 5 0 5 0 1 0 0 1 5 0 2 0 0 2 5 0 3 0 0 3 5 0 4 0 0 4 5 0 5 0 0 5 5 0 6 0 0 6 5 0 7 0 0 Time, sec 200 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. F ig u re 5. Reaction between Me-HA and HOS (200 ftmol scale, as described in the experimental Section) in the presence of excess NaOD and MeOH as a standard (a). Proton-decoupled (b) and coupled (c) 1 3 C NMR o f formaldehyde hydrazone are also shown. 201 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. S u O u I ft i b 390 ' ' ‘ ' 200 too too 00 0 pom 202 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. r ^ \ F ig u re 6. Formation of a hemochrome from the reaction of reduced hemoglobin (2.5 x lfr5 M) (b) (by reaction o f oxidized Hb (a) with dithionite) with mono- methyldiazene (30 mM) (c and d). Essentially the same spectrum results from the addition of reduced hemoglobin to dilute mono-methyldiazene (e-g). 203 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 4 H M 3.113 a o o b o o o e o c r u u r \ > 0 1 o o © o o o w > 1 V I o e o o o 204 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. e < a g S o e e o f M - 4 ta M • 4 < a e g N ta w « a o o o o o Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. F ig u re 7 . Inhibition of H, evolution by mono-methyldiazene. The GC trace (a) includes first a set of 5 H2 calibration injections, and then the assay samples followed by two air injections. The plot in (b) shows the decrease in nmol Hj produced as a function of the diazene molarity. The assay was run for 10 min in the presence of 0.048 mg FeMoX- 37 and 0.80 mg FeGG38B. 206 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. nm ol H2 formed a cdSbntkm J U ^ X - L r v ,X v aI/v IaIa^ / alr^JcctfcM M u l . n i n v ii 1111111 1111 11111 ii m u i T r r i i 11111 0.0 1 1 1 1 1 1 1 1 1 1 ■ ii 1 1 1 ■■ ii 1 1 1 1 1 67.8 ndn 800 b 600- 400- 2 0 0- 0 6 2 4 8 10 12 mono-MeDz, mM after addition 207 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. F ig u re 8 . Inhibition of H2 evolution by mono-methyldiazene in 22 ml assay vials. The assay vial 13 had initial mono-methyldiazene concentration of 10.7 mM but was pre incubated for 10 min prior to the addition o f the nitrogenase proteins to allow for extensive diazene decomposition. The assay was run for 5 min in the presence of 24 ftl FeMoX-37 (0.192 mg) and 20 ftl FeGG38B (0 3 2 mg). 206 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1 6 0 0 1 1 4 0 0 - 120 0 - I 1000- & ™ 8 0 0 - K 1 6 0 0 “ 4 0 0 - 200 - 0 - - t 2 i 4 T 6 vial 13 T 8 T " 10 T “ 12 T " 1 4 1 6 mono-methyldiazene, mM 209 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. F ig u re 9 . Maximum N2 partial pressure in 21 ml head space as a function of mono-methyldiazene initial concentration. It is assumed that a complete decomposition of the diazene to N2 and CH4 takes place. 210 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 0 .0 3 I £ 0 . 0 2 5 - 0.0 2 - 0 . 0 1 5 - 3 S 0.01 - 0 . 0 0 5 - o Q l T 2 r 4 ~r 6 8 T " 10 "T " 12 T " 1 4 1 6 Mono-methyldiazene Initial Concentration, mM 211 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. F ig u re 10 . Inhibition of Q H , reduction by mono-methyldiazene. The assay was run for 5 min in the presence of 0.048 mgFeMoX-37 and 0.80 mg FeGG38B. 212 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1/nmol C2H4 0.02 0 . 0 1 5 - 0.01 - 0 . 0 0 5 - y = 7.10E-05x + 4.29E-03 r2 = 1.00E+00 y = 6.43E-05x + 3.34E-03 r2 = 9.96E-01 y = 3.48E-05x + 4.24E-03 r2 = 9.46E-01 y = 1.59E-05x + 3.51E-G3 r2 = 9.83E-01 1 /n m o l C 2 H 4 , - I 1 /n m o l C 2 H 4 . + 0 . 1 5 m l I 1 /n m o l C 2 H 4 , + 0 . 2 5 m l I 1 /n m o l C 2 H 4 , + 0 . 3 5 m l I - 5 0 0 1/PC2H2, atm-1 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. F ig u re 11. Inhibition of C,H2 reduction by mono-methyldiazene. The repeat experiment was performed in 22 ml vials as described in the Experimental Section. Each vial assay vial contained 6 pi FeMoX-37 (0.048 mg) and 5 pi FeGG38B (0.080 mg). 214 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1/nmol C2H4 0.025 0.02 0 .0 1 5 0.01 0 .0 0 5 y = 6.05E-05x + 6.85E-Q3 i2 = 6.1 IE-01 y = 4.13E-05x + 5.87E-Q3 i2 = 6.93E-01 y = 3.21E-05x + 5 .10E-03 r2 = 9.38E-01 y = 2.79E-05x + 5.06E-03 r2 = 9.70E-01 y = 1.90E-05x + 4.42E-03 i2 = 8.88E-01 □ l/ n m o l C 2 H 4 , - I 1 /n m o l C 2 H 4 , + 3 .9 m M I l/ n m o l C 2 H 4 , + 6 .5 m M I l/ n m o l C 2 H 4 , + 9 .1 m M I l/ n m o l C 2 H 4 , + 1 1 6 m M I 2 4 0 - 2 0 0 200 l/p C2H2’ atm 1 215 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. F ig u re 1 2 . HPLC analysis of the liquid phases from the first H2 inhibition experiment as described in the Experimental Section. The methylamine peaks are indicated by arrows. 216 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 93.2 0.0 217 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. F ig u re 13. Methylamine formation in the methane detection experiment The methylamine peaks are indicated by arrows. 218 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. I III n i l III III I III M il III I IIH 'IH i i i i i i i i i m i 111 m i i n i i i i i 0.0 103.5 219 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER 7 POSITIONAL ISOTOPE EXCHANGE AND WASHOUT STUDIES OF AZOTOBACTER V tN E L A N D U NITROGENASE Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. INTRODUCTION The chemical mechanism of ATP hydrolysis is not yet fully understood. Work from our laboratory previously demonstrated that A. vinelandii nitrogenase acts as a “normal” ATPase by catalyzing nucleophilic attack on the ^phosphorus atom of ATP in P,y-[180 ]- directly involved in ATP hydrolysis, possibly by enhancing the nucleophilicity of an adjacent water molecule. Our laboratory previously reported briefly on the application of positional isotope exchange (PIX) to nitrogenase in an attempt to obtain information about the nature of the ATP mechanism, finding no evidence for a significant positional isotope conditions at 30 °C. Recently, Mr. Stephan Krause in our laboratory in collaboration with Prof. G. L. Kenyon (UCSF) and coworkers began a further investigation of possible exchange processes catalyzed by nitrogenase proteins, using y-[I80]-ATP The process of PIX in this form of labeled ATP is illustrated below where the labeled oxygen atoms are shown in bold. A TP. ^ The X-ray structure of the Fe protein suggests that a caiboxylaie residue may be exchange? of p,y-[180]-A TP (31P NMR spectroscopy) by Av nitrogenase under turnover O O O I I I I I I HAE || || \ d O— P— O + O— P - 0 ) —AMP 0 — P— O— P— O IB ia O o o t o o B N OH O H 221 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. When y-[180 4]-ATP is cleaved by nucleophilic attack at the y-phosphorus atom, the result is ADP with the labeled oxygen on the ^-phosphorus atom and the y-phosphorus atom covalently attached to the nucleophile. Free rotation about the O-Pp bond in die enzyme-bound ADP intermediate, followed by reversal of die nucleophilic cleavage reaction would result in reformation of labeled ATP but with the label distributed 2/3 in nonbridging (N) positions and 1/3 in the original bridging (B) position.^ The redistribution can be detected by 31P NMR spectroscopy, relying upon the slightly different upfield 31P chemical shifts produced by adjacent bridging versus non-bridging 180 , relative to 160 . In addition, initial and assay recovered y-[180 4 ]-ATP can be compared by mass spectrometry since in a reversible process, reformed y-[180 4]-ATP would show the loss of at least one 180-label to a water molecule. Finally, Thomeley et al.9 have reported MgADP-induced enhancement o f a [1 6 0 ] exchange with [l8OJ-phosphate (“inorganic phosphate washout “). Dale and Hackney eliminated myosin-catalyzed PIX previously observed by others, 10 using highly purified myosin. 11 The original claim of myosin-catalyzed PIX was explained by the presence of contaminating adenylate kinase. 11 They did detect washout in the cleaved Pt, leading them to conclude that ATP hydrolysis is reversible but free rotation about the O-Pp bond is not allowed. Hasset and coworkersl^ also eliminated PIX catalysis from pyruvate kinase by further purification of the protein. 13 It is thus extremely important in studies involving apparent enzyme-catalyzed PIX to eliminate possible contributions made by contaminating proteins. 222 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. EXPERIM ENTAL SECTION Synthesis of [180 4 ]*Phosphate. [U 0 4 ]-Phosphate was prepared using a modification of the method of R ay. 14 Phosphorus pentachloride (753.3 mg, 3.62 mmol, 1 equiv.) was added to a 25 ml round bottomed flask and slurried with 5 ml of 1,4-dioxane (Mallinckrodt) in a dri-box. To this 1 8 was added in a dropwise manner over two minutes, a solution of H2 O (521 pi, 28.7 mmol, 1.98 equiv. of 95 - 98 % total enrichment in "O) in 3 ml of 1,4-dioxane. Upon addition, the slurry dissolved, indicative of conversion to phosphoric acid. The solution was then stirred at room temperature for an additional 37 min. Diethylamine (748 pL, 7.23 mmol) was added slowly to the reaction mixture and stirring continued at room temperature for another 50 min. The reaction mixture was diluted with water (2 ml) and then concentrated in vacuo to approximately 3 ml. The solution was then diluted to approximately 20 ml and passed through a Dowex 1 x 8 column (3 x 17 cm), HCO3' form. The column was first eluted with water, followed by a linear gradient of 0-0.5 M triethylammonium bicarbonate buffer (800 ml). Column fractions were assayed for inorganic phosphate using a modified version of the Ames molybdate-ascorbic acid a s s a y . 15 Phosphate-containing fractions were then combined and tributylamine was added to form the tributylammonium salt The combined fractions were then concentrated in vacuo to dryness, yielding a white residue. The residue was dissolved in methanol followed by drying in vacuo. Methanol treatment was repeated three times. A portion of the inorganic phosphate was derivatized to the permethylated derivative for analysis by GC/MS as described later. 223 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Synthesis o f y-[18C > 4]- ATP. 16 ADP tributylammonium salt was prepared by passing the sodium salt of AW (704.8 mg, 1.44 mmol, Sigma Chemical) through a Dowex SO x 8 column (1.5 x 50 cm, H+ form) followed by the addition of 1 equivalent of tributylamine to the aqueous solution (345 pi). The water was removed in vacuo to yield a clear oil. The oil was then redissolved in methanol followed by solvent removal under reduced pressure a total of four times to give a white semisolid. The tributylammonium salt was then transferred to a 100 ml round bottomed flask, concentrated, and placed under vacuum for 1 h. The salt was then blanketed with argon, slurried with 10 ml of DMF and cooled to 0 °C. The ADP salt was then activated by the addition of carbonyl diimidazole (1.17 g, 7.2 mmol) and stirring at 0°C for 15 min followed by equilibration to room temperature and stirring for 13.5 h. The reaction was quenched with methanol (230 pi, 5.7 mmol) to remove excess carbonyl diimidazole and the mixture stirred at room temperature for an additional 25 min. In a separate flask, [lgO]-labeled inorganic phosphate tributylammonium salt (438.3 mg, 1.51 mmol) was slurried in DMF (3 ml) and then added to the activated ADP mixture. This reaction was stirred at room temperature and monitored by HPLC. Heat was gently applied to the heterogeneous mixture after 6 h to make the reaction homogeneous. This reaction was then allowed to stir for 2 days at room temperature and then quenched with 0.1 N NaOH (60 ml), and extracted with ethyl acetate (2 x, 50 ml each). The aqueous layer was concentrated in vacuo to about 20 ml then applied to a DEAE A-25 Sephadex column (2.7 x 55 cm) and washed with water. The samples were then eluted using a linear gradient of 0.1-0.5 M triethylammonium-bicarbonate (pH 8.5, 3 L) followed by elution of 1 L of 0.6 M triethylammonium-HCC>3', and finally 1 L o f 0.75 M triethylammonium- HCO3 '. Fractions were then analyzed by UV spectrophotometry at 259 nm. Appropriate fractions of ATP, identified by TLC, were combined followed by the addition of 224 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. tributylamine. The aqueous mixture was then concentrated in vacuo to dryness yielding a white semisolid. The latter was dissolved in methanol and then concentrated down to a white semisolid. This methanol treatment was repeated three more times and then the flask was placed under vacuum to give a clear oil. Mass Spectrometry. Isotopic distribution of the methylated inorganic phosphate samples was analyzed by high resolution mass spectrometry of the GC-separated trimethyl phosphate on a Hewlett-Packard S989A MS coupled to an HP 5890 GC at the Mass Spectrometry Facility, UC Riverside, CA. Trimethyl phosphate samples in ether or CDCI3 were separated on a DB-5 capillary column (30 m x 0.25 pm i.d.) with helium carrier and detected by CH4 chemical ionization MS (7 scans/sec). The GC oven temperature program was: 50 °C x 1 min, temperature gradient o f 50 -150 °C at 10 ° C /min, temperature gradient of 150 - 250 °C at 30 ‘ C / min, and 250 C C x 1 min. Typically, 2-5 pi aliquots (ca. 5-15 nmol of trimethyl phosphate) were injected. Prior to determinations, a sample of commercially unlabeled trimethyl phosphate ( 1 0 0-fold dilution in ether) served as a control to verify retention time and instrument response. Methylation of the reisolated inorganic phosphate was verified by *H (360.13 MHz) and 31P (145.786 MHz) NMR on Bruker AM-360 instrument An approximate calibration curve was constructed by running NMR (64 acquisitions) of standard solutions (0.4 ml) of (MeObPO in CDG3 (0.5 - 8 mM) and integration of the respective proton peaks. In all cases, the trimethylphosphate proton doublet was positively identified not only by its chemical shift (ca. 3.7 ppm, with variations due to the presence of ether, water and/or methanol) but also by its coupling constant of 10.9-11.1 Hz. Such verification was made necessary by the relatively small peak size in the case of the washout samples 225 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. (see Results). In the cases where excess methanol obscured the trimethylphosphate proton peaks, methylation and sample recovery was verified by 31P NMR. Since the samples were not referenced individually, but instead a predetermined SR value was set, some minor (± 0 .1 ppm) chemical shift differences appeared; however, the peaks always remained around 2.9 ppm. Whenever sufficient instrument time was available, sample identity was also verified by measuring the proton-coupled 31P NMR spectrum which revealed splitting by the 9 methyl protons. Isotopic distribution of the methylated inorganic phosphate samples was analyzed by high resolution mass spectrometry of the GC-separated trimethylphosphate on a Hewlett-Packard 5989A MS coupled to an HP 5890 GC at the Mass Spectrometry Facility, UC Riverside, CA. Trimethylphosphate samples in ether or CDCI3 were separated on a DB-5 capillary column (30 m x 0.25 pm i.d.) with helium carrier and detected by CH4 chemical ionization MS (7 scans/sec). The GC oven temperature program was: 50 °C - hold for 1 min, ramp 50 -150 °C at 10 °C /min, ramp 150 - 250 °C at 30 °C /min, 250 °C - hold for 1 min. Typically, 2-5 pL sample aliquots (ca. 5-15 nmol of trimethylphosphate) were injected. At the beginning of the determinations, a sample of commercial unlabeled trimethylphosphate (1 0 0-fold dilution in ether) was run in order to establish the retention time and verify the instrument response. Inorganic phosphate work-up. Washout assay reaction mixtures were worked up for isolation of inorganic phosphate in the following manner. The reaction mixtures (ca. 400 pi) were thawed and immediately subjected to microcentrifugation in Microcon-10 devices (Amicon) for protein removal. The resultant mixtures were passed through pre-conditioned (10 ml wash with HPLC grade methanol followed by 10 ml wash with HPLC grade water) Sep-Pak Gassic 226 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. C-18 cartridges (Waters) and inorganic phosphate was collected as die first fraction by elution with HPLC-grade water. Control runs, in which component elution was followed by phosphate analysis^ and UV, showed that the collected inorganic phosphate contained ca. 2 % ADP at the conditions of the experiment To convert the inorganic phosphate into its acidic form, solutions of the former were passed through Dowex 50WX8-200 strongly acidic cation exchange resin (Aldrich) and the collected fractions, typically 7 ml, were subjected to rotary evaporation at T 40 °C to minimize sample loss as well as the possibility o f oxygen scrambling at these low-pH conditions. Recovered inorganic phosphate (as a tributylammonium salt) from the PIX experiments was dissolved in a minimum amount of HPLC grade water and passed through the cation exchange resin in a similar manner. Due to the possibility of sample contamination with environmental l 6 0 phosphate, various measures were taken to preserve the samples' integrity. These included using brand new glassware when possible, using a brand new C-18 cartridge for every individual sample, including the triplicate controls, and scrupulous washing of the reusable glassware which was done with phosphate-free Dove detergent and included rinses with hot tap water, distilled water, acetone, ethanol, and HPLC grade water. All glassware, including brand new NMR tubes, was oven-baked for at least 2 hr prior to use. Since prolonged rotary evaporation led to significant losses of the phosphoric acid residue, pumping was stopped as soon as the sample size was reduced to a small drop (ca. 20-40 pi). The residue was immediately taken up in MeOH and Et2 0 and subjected to extensive diazomethane treatment in the Aldrich MNNG-Diazomethane Apparatus by charging the apparatus twice with the recommended 147 mg (1 mmol) l-methyl-3-nitro-l- nitrosoguanidine (MNNG) reagent (Aldrich Technical Information Bulletin AL-180). The latter step was necessitated by the presence in the final sample of some water and especially 227 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. strong acids carried through all workup steps from the assay buffer (mainly chloride ion from MgCl2) along with Pi, which initially led to very low methyladon yields. After the second methyladon, the solvents (MeOH and Et2 0 ) were removed by careful application of an At stream, and die resulting residue was dissolved in ca. 400 pi CDCI3 for NMR analysis. At this step, for reasons already mentioned at the rotary evaporation step, the Ar flush was stopped early, without making an attempt to remove all of the remaining MeOH. Separate experiments showed that at c a 1 pmol levels of trimethylphosphate, leaving an Ar stream against a thin film of the compound for even a short time causes a significant loss of material, which, given the scale of die enzyme experiment (2 pmol o f starting labeled phosphate), proved critical for obtaining a sufficient amount of GC-MS sample. Before transportation to the HRMS facility, the NMR tubes containing the samples were cut off to minimize dead volume and facilitate syringe needle access. The liquid volume was then reduced to 120-150 pi by a very brief Ar flush and the resulting truncated tubes (ca. 5 cm in height) were capped and sealed with parafilm. Samples were stored in a refrigerator. The work-up of each sample was performed without interruption during the course of a single day. Enzyme purification. A) Large-scale nitrogenase preparation from frozen cell paste. The general protocol developed in this group was followed. 17,18 4 vinelandii frozen cell paste (1,647 g) were thawed and homogenized with 25 mM Tris buffer (pH 7.4) to a total volume of 3.4 L. The suspension was washed once by a brief centrifugation and resuspension in cold Tris. The cells were lysed on an Aminco French Press equipped with a 40 ml rapid-filling cell at 16,000-20,000 psi. The dark brown lysate (3,160 ml) was pH adjusted to 7.4 by quick titration with 1 M NaOH, degassed by careful pumping and 228 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. argonation, and then treated with deoxyribonuclease and ribonuclease for 1 h at room temperature (Ar) to remove the nucleic acids. Sodium dithionite (DT) stock solution was added to 3 % (v/v). The crude extract was subjected to heat treatment for 5 min at 56 * C followed by a rapid cooling in ice bath in an apparatus consisting of two glass coils as described previously. 17,18 The heat-treated extract was collected directly inside the glove box and periodically tested for reducing power with methyl viologen. Overnight centrifugation at 1 0 ,0 0 0 rpm resulted in the separation of the heat-sensitive proteins (pellet, discarded) from the very dark nitrogenase-containing supernatant Pharmacia K 100/45 column was filled with ca. 1600 ml settled DEAE-sepharose fast-flow gel and equilibrated with 25 mM Tris/DT. The heat-treated extract was loaded directly from the glove box and formed a large dark brown band. A step gradient was applied which consisted of 0.1 M NaCl/Tris, followed by 0.25 M NaCl to elute the FeMo protein (FeMo-I fractions), and finally, 0.5 M NaCl which removes the Fe protein (Fe-I stage). An Isco V4 detector was used to monitor the eluent’s absorption at 440 nm throughout these chromatography steps. The crude FeMo protein obtained from this purification was later used in a separate concentration and purification sequence as described in Chapter 5. B) Further purification of Fe protein. Fe-I protein obtained from the first DEAE column was further purified by a linear NaCl gradient (0.25-0.50 M) on a second DEAE-sepharose anion-exchange column, Pharmacia K 50/30, filled with ca. 450 ml settled gel. The gradient was prepared inside the glove box in a 1 L American Scientific gradient mixer and fed outside by a Masterflex pump. This second column allows Fe protein (Fe-II, first band) to be separated from die presumed flavodoxins (green-gray, second band). The Fe-II fraction was first concentrated about tenfold in Amicon 2000 and 402 stirred ultrafiltration cells inside the glove box, and 229 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. then separated onto a 5 x 80 cm column packed with Sephaciyl S-200 gel filtration medium. An additional step in die purification procedure was introduced which involved passage through a (80x2.5 cm) Sephacryl S-200 (Pharmacia) gel filtration chromatography. To eliminate any traces of FeMo present in the Fe protein preparations which were previously associated with the observation of exchange in experiments described below, another S-200 column chromatography step was performed on the Fe protein. Columns were carefully packed to ensure maximum resolution. The column was eluted with 25 mM Tris, 0.5 M NaCl, 0.1 mg mL* 1 dithiothreitol, 2.0 mM DT, at pH 7.4. A high salt concentration was used to reduce band broadening due to secondary Donan effects. In the second additional purification step performed by Mr. S. Krause, Superdex 75HR 10/30 column (Pharmacia) was flushed overnight with the eluting buffer (25 mM TRIS, 2 mM MgCl2, 3 % DT, and 0.1 mg/ml DTT). All solvents were tested for their reducing power using methyl viologen. The Fe protein was first concentrated in a dri-box using Microcon-30 (Amicon) concentrators, and then filtered with Ultrafree MC 0.22 urn centrifuge filters (Millipore). The filtrate was injected into the HPLC system. The flow rate of the system was kept at 0.5 ml/min for the first 3 minutes (loading) and then lowered to 0.3 ml/min for the remaining elution. All collected fractions were analyzed for their protein concentrations and purity by the biuret method and SDS-PAGE. The pooled highly purified FeGG38B’ protein showed no detectable ATPase activity (< 1 pmol ATP hydr. /min./mg) or background C2H2 reduction (< 1 pmol C2H4 /minVmg) when analyzed by Mr. S. Krause. 230 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. RESULTS AND DISCUSSIO N Inorganic phosphate derivatization protocol: general observations. During initial tests, a number of phosphate samples failed to give a reliably detectable signal for MS analysis. After performing a number of tests, it was concluded that the most likely reason for sample loss was the presence in the final methylation mixture of strong mineral acids, mainly HC1 derived from the chloride ion in the 5 mM MgCl2 present in the enzyme assay. As the presence of such strong acids may lead to preferential consumption of diazomethane, at the initial experiments die methylating reagent may not have been present in sufficient excess to counter that effect The inclusion of a second methylation step and allowing the reaction itself to proceed for a longer time appeared to have solved the sample loss problem. In typical washout experiments, which contained 2 pmol of starting 180 -labeled inorganic phosphate, the final yields of trimethylphosphate were all in the range of 0.2 - 0.4 pmol with the main losses occurring after the cation exchange step (data not shown). Our experience shows that a final yield of ca. 80 • 100 nmol is necessary to obtain a reliable mass spectral pattern under the conditions used, therefore the scale of the washout experiment was not readily decreased below that used here. Inorganic phosphate derivatization protocol: NMR results. Figure 1 shows two typical ‘H NMR spectra of recovered and methylated inorganic phosphate samples. As seen in the figure, depending on the exact amounts of MeOH/EtjO used to redissolve the H+ -exchanged phosphate and on the amount of residual water in the small droplet of H3P 0 4 , the spectra are dominated by any combination of MeOH, water, and EtjO. Calibration experiments (Figure 2) following the relative area of the 231 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. trimethylphosphate doublet versus the residual proton peak from the chloroform solvent provided reliable estimates of the actual amount of derivatized phosphate present in the NMR tube, at the end of all isolation and derivatization steps. Such an estimate was later found useful in deciding what fraction of that sample to inject into the GC/MS system for analysis. During the NMR tests of calibration and real washout inorganic phosphate samples, a number of 3 IP NMR spectra of high quality were collected, by generally running 2-4 hr acquisitions on the 360 MHz instrument A close examination of the phosphorus peaks revealed die presence of small peaks, usually on either side of the main singlet but never on both, forming a kind of ladder. Although the resolution o f these peaks was obviously (Figure 3 and 4) close to the instrument limits, die same, ca. 0.02 ppm, chemical shift differences were found to be reproducible among all the samples studied. After the samples of the newly synthesized tetralabeled inorganic phosphate as well as washout recovered material were analyzed by GC/MS, it became evident that the 3 IP NMR ladders were almost perfectly matched in their height proportions by the mass-spectral pattern obtained for the differently labeled trimethylphosphate species. Figure 3 and 4 show the comparison between outputs o f die two analytical methods for two washout samples which show essentially no washout (Figure 3, major peak in both spectra is 1 8 0 4-Pi) and a significant washout/l6 0 exchange (Figure 4, major peak in both spectra is 1 60 4-Pi), respectively. It was, therefore, possible to assess, at least in a crude way, the isotopic distribution of these samples along with the NMR verification of their successful derivatization. Recent tests by Mr. S. Krause on a higher-field, 500 MHz, instrument confirmed the above findings, and provided an increased peak resolution. When the GC/MS analysis is performed, only a small time slice (usually chosen arbitrarily by the operator) o f the trimethylphosphate GC peak is analyzed and reported by MS, which, due to the very small sample size, may allow for some sampling error. In contrast to this method, the NMR analysis described above 232 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. may offer not only die advantage of being easily performed at an in-house instrument but also the added reliability of obtaining a “bulk” picture of the whole sample mass contained in the NMR tube. Protein purification results. The large-scale purification produced initially 3.16 L lysate which contained approximately 3.8 x 106 units o f Fe protein activity and 9.6 x 106 units of FeMo activity. The first DEAE column yielded approximately 11.3 x 106 units of FeMo activity, and 5.6 x 106 units, or 73 %, of Fe activity. Subsequent Fe protein purification steps resulted in a significant loss of total activity (ca. 50 % loss after the second Sephacryl S-200 column separation) due to technical problems but the final protein had a very high specific activity of ca. 1900 and was very pure by gel electrophoresis criteria. REFERENCES (1) Mortenson, L. E. Proc. Natl. Acad. Sci 1967,52, 272-279. (2) Ashby, G. A.; Thomeley, R. N. F. Biochem. J. 1987,246, 455-465. (3) Stiefel, E. I.; Cramer, S. P. In Molybdenum Enzymes; Spiro, T. G., Ed.; Wiley Interscience: New York, 1985, p 117-159. (4) Watt, G. D.; Wang, Z.-C. Biochemistry 1986,25, 5196-202. (5) Hageman, R. V.; Orme-Johnson, W. H.; Burris, R. H. Biochemistry 1 9 8 0 ,19, 2333-2342. (6) McKenna, C. E.; Gutheil, W. G.; Kenyon, G. L.; O., M. T. Bioorganic Chemistry 1989,17, 377-384. (7) McKenna, C. E.; Gutheil, W. G.; Kenyon, G. L.; Matsunaga, T. O. In Nitrogen Fixation Research Progress', Evans, H. J., Bottomley, P., Newton, W. E., Eds.; Martinus Nijhoff: Dordrecht, 1985, p 635. (8) Midlefort, C. F.; Rose, I. A. J. Biol. Chem. 1976,251, 5881-5887. 233 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. (9) Thomeley, R. N. F.; Ashby, G. A.; Julius, C.; Hunter, J. L.; Webb, M. R. Biochem. J. 1991,277, 735-741. (10) Geeves, M. A.; Webb, M. R.; Midelfort, C. F.; Trentham, D. R. Biochemistry 1980,19, 4748-4754. (11) Dale, M. P.; Hackney, D. D. Biochemistry 1987,19, 8365-8372. (12) Hassett, A.; Bleattler, W.; Knowles, J. R. Biochemistry 1982, 21, 6336- 40. (13) Lowe, G. D.; Sprout, B. S. J. Chem. Soc., Chem. Comm. 1978, 783-5. (14) Ray, W. J. J. Labelled Comp. Radiopharm. 1992,31, 637-9. (15) Ames, B. N. Methods Enzymol. 1966,8, 115. (16) Reynolds, M. A.; Oppenheimer, N. J.; Kenyon, G. L. J. Am. Chem. Soc. 1983, 705, 6663-6667. (17) Gemoets, J. P. Ph.D. Thesis, U. of Southern California, 1990. (18) McKenna, C. E.; Nguyen, H. T.; Huang, C. W.; McKenna, M. C.; Jones, J. B.; Stephens, P. J. In From Cyclotrons to Cytochromes (M. D. Kamen Symposium); Kaplan, N. O., Robinson, A. B., Eds.; Academic Press: New York, 1982, p 397-416. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 1. Typical lH NMR spectrum of recovered derivatized inorganic phosphate from a washout experiment showing the trimethylphosphate doublet (arrow) in die environment of dominating water/ether (a) and methanol (b). 235 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 3 1 7 6 2 5 4 236 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. F ig u re 2 . A calibration curve for 360 Mhz instrument obtained by 'H NMR peak integration of trimethylphosphate standards dissolved in CDC13 whose residual peak was set to area = 1 0 0 . 237 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 600- 55 500- □ C O ■ s & s s c o 9 s 400- 300- 200 • 1 0 0 - tn s tn u n e n t - N U T S P ro g ra m T 4 T 6 (MeO)3PO, mM in 0.4 ml sample 238 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. F ig u re 3 . 3 1 P NMR (top) and mass (bottom) spectra of a washout sample (“SK.G”) showing veiy little exchange. 239 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ze Z7 Z8 3.0 3.1 ppm Abundance 100 147 00 145 70 141 3 0 190 ISO Mass/Charge Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. F ig u re 4 . 3IP NMR (top) and mass (bottom) spectra of a sample (“PIX.2”) showing considerable exchange. 241 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 2.7 2.8 3.0 ppm Abundance 100-3 141 145 70- 143 ISC 190 140 190 Mass/Charge 242 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission IMAGE EVALUATION TEST TARGET (Q A -3 ) A A 1 . 0 l.l 1.25 |X2 [3 6 IM 23 2.2 1 .4 l U i 1 .6 150mm IIW 4 G E . In c 1653 East Main Street Rochester, N Y 14609 USA Phone: 716/482-0300 Fax: 716/288-5989 0 1993. Applied Image. Inc.. All Rights Reserved Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. INFORMATION TO USERS This manuscript has been reproduced from the microfilm master. UMI films the text directly from the original or copy submitted. Thus, some thesis and dissertation copies are in typewriter face, while others may be from any type o f computer printer. The quality o f this reproduction is dependent upon the quality o f the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleedthrough, substandard margins, and improper alignment can adversely affect reproduction. In the unlikely event that the author did not send UMI a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion. Oversize materials (e.g., maps, drawings, charts) are reproduced by sectioning the original, beginning at the upper left-hand comer and continuing from left to right in equal sections with small overlaps. Each original is also photographed in one exposure and is included in reduced form at the back of the book. Photographs included in the original manuscript have been reproduced xerographically in this copy. Higher quality 6” x 9” black and white photographic prints are available for any photographs or illustrations appearing in this copy for an additional charge. Contact UMI directly to order. UMI A Bell & Howell Information Company 300 North Zed) Road, Ann Arbor MI 48106-1346 USA 313/761-4700 800/321-0600 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. UMI Number: 9902866 UMI Microform 9902866 Copyright 1998, by UMI Company. All rights reserved. This microform edition is protected against unauthorized copying under Title 17, United States Code. UMI 300 North Zeeb Road Ann Arbor, MI 48103 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Isolation and characterization of mouse GRIP1, a novel transcriptional coactivator of steroid receptors

PDF

The reaction of sulfenyl halides with cis-trans olefins

PDF

The metabolism in vitro of 4-C¹⁴-cytidine

PDF

Addition compounds of boron fluoride with certain sulfoxy-amines

PDF

Protein-protein interactions in blood serum

PDF

Calculations of electrostatic interactions in proteins

PDF

The biochemistry of glycoproteins from human prostate

PDF

Novel phenylpolyene-bridged second-order nonlinear optical chromophores and new thermally stable polyurethanes for electro-optic applications

PDF

Chiroptical spectroscopy of nitrogenase

PDF

Computer modeling and free energy calculatons of biological processes

PDF

The isolation and characterization of rat plasma albumins

PDF

The genes for the glucose-regulated protein GRP94 and GRP78 are co-ordinately regulated by common trans-acting factors

PDF

Moloney murine leukemia virus (MoMuLV) envelope protein: Interaction with receptor and engineering for targeted gene transfer

PDF

The effects of irradiation upon lipid and carbohydrate metabolism in the rat

PDF

The incorporation of radioactive carbons of glucose into the protein-bound amino acids of one-day-old mouse brain

PDF

Novel substrates of nitrogenase

PDF

Semantic heterogeneity resolution in federated databases by meta-data implantation and stepwise evolution

PDF

The enzymatic and nonenzymatic hydrolysis reactions of s-aryl phosphorothioates

PDF

Synthesis of binucleating ligands with relevance to hemocyanin

PDF

Excited states of nitrogen heterocycles and their interaction with specific anions

Asset Metadata

Creator Simeonov, Anton Momtchilov (author)

Core Title Interactions of diazene homologues with Azotobacter vinelandii nitrogenase enzyme and model systems

School Graduate School

Degree Doctor of Philosophy

Degree Program Chemistry

Degree Conferral Date 1998-05

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag chemistry, biochemistry,OAI-PMH Harvest

Language English

Contributor Digitized by ProQuest (provenance)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c17-358856

Unique identifier UC11350737

Identifier 9902866.pdf (filename),usctheses-c17-358856 (legacy record id)

Legacy Identifier 9902866.pdf

Dmrecord 358856

Document Type Dissertation

Rights Simeonov, Anton Momtchilov

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA