Report

An Fe2IVO2 Diamond Core Structure for the Key Intermediate Q of Methane Monooxygenase

See allHide authors and affiliations

Science  24 Jan 1997:
Vol. 275, Issue 5299, pp. 515-518
DOI: 10.1126/science.275.5299.515

Abstract

A new paradigm for oxygen activation is required for enzymes such as methane monooxygenase (MMO), for which catalysis depends on a nonheme diiron center instead of the more familiar Fe-porphyrin cofactor. On the basis of precedents from synthetic diiron complexes, a high-valent Fe2(μ-O)2 diamond core has been proposed as the key oxidizing species for MMO and other nonheme diiron enzymes such as ribonucleotide reductase and fatty acid desaturase. The presence of a single short Fe-O bond (1.77 angstroms) per Fe atom and an Fe-Fe distance of 2.46 angstroms in MMO reaction intermediate Q, obtained from extended x-ray absorption fine structure and Mössbauer analysis, provides spectroscopic evidence that the diiron center in Q has an Fe2IVO2 diamond core.

The MMO enzyme system found in methanotrophic bacteria initiates the oxidation of methane (1, 2), thereby preventing the atmospheric egress of nearly 1 billion tons of this greenhouse gas annually. MMO catalyzes the difficult oxidation of methane (CH4) to methanol (CH3OH) with incorporation of one oxygen atom from O2. The soluble MMO system consists of three separate protein components termed the hydroxylase (MMOH), reductase (MMOR), and component B (MMOB) (3, 4). The crystal structures of MMOH have revealed a nonheme diiron active site (5, 6, 7) where oxygen activation and substrate oxidation occur (4). Transient kinetic analysis of a single-turnover reaction has revealed at least five and probably six intermediates in the catalytic cycle, among which intermediate Q is the key oxidizing species (8, 9, 10, 11). The Mössbauer properties of Q indicate an exchange-coupled high-valent FeIVFeIV cluster. The FeIV oxidation state has been assigned on the basis of the large decrease in isomer shift from δ = 0.50 mm s−1 for FeIIIFeIII MMOH to δ = 0.17 mm s−1 for Q (12); the latter value is comparable to the δ values for well-characterized FeIV complexes (13, 14). To date, no high-valent intermediate of any metallooxygenase has been structurally characterized. Here, we report extended x-ray absorption fine structure (EXAFS) studies of Methylosinus trichosporium OB3b MMOH intermediate Q that provide spectroscopic evidence that an enzyme uses an Fe2(μ-O)2 diamond core for alkane oxidation (15).

A rapid freeze-quench technique allowed us to trap Q in the optimal time domain after mixing FeIIFeII MMOH with 100% O2-saturated buffer in the presence of two equivalents of MMOB (8, 9, 12). The samples were analyzed by Mössbauer spectroscopy to provide an independent quantitation of the reaction cycle intermediates in the samples before and after the x-irradiation inherent in the EXAFS experiment. Figure 1 shows a 4.2 K Mössbauer spectrum of Q sample 1 and the corresponding features that make up this spectrum. The progressive decrease in isomer shift δ upon passage through successive reaction cycle intermediates FeIIFeII MMOH, P (16), and Q indicates the increasing oxidation state of the two Fe sites. The Mössbauer-determined compositions of the two freeze-quenched samples 1 and 2 are listed in Table 1. As indicated, intermediate Q represented substantial fractions of each sample, 61% and 44%, respectively; these percentages were the same before and after irradiation.

Fig. 1.

Mössbauer spectra of MMOH from Methylosinus trichosporium OB3b recorded at 4.2 K. Shown are representative spectra of FeIIIFeIII MMOH (δ = 0.50 mm s−1), FeIIFeII MMOH (δ = 1.30 mm s−1), and transient intermediates P (FeIIIFeIII, δ = 0.67 mm s−1) and Q (FeIVFeIV, δ = 0.17 mm s−1), as well as the spectrum of Q sample 1 used for EXAFS studies. Quadrupole doublets are indicated by brackets; isomer shifts δ are marked by filled triangles. The solid line drawn through the spectrum of sample 1 is a superposition of computed spectra for the FeIIFeII and FeIIIFeIII forms, P, and Q using fractions listed in Table 1.

Table 1.

Compositions of freeze-quenched EXAFS samples of MMOH intermediate Q determined by Mössbauer spectroscopy. An optimal time window from 100 to 320 ms, determined by stopped-flow spectroscopy at 17°C, was used to quench a single-turnover reaction and trap intermediate Q according to the experimental procedure reported previously (8, 9, 12). Sample 1 was trapped at 150 ms and sample 2 was trapped at 300 ms, allowing us to study two samples with different concentrations of Q for comparison of their EXAFS feature intensities.

View this table:

The R-space EXAFS spectra of four MMOH samples (17) have features that correspond to the distances from each Fe site to surrounding atoms (Fig. 2). For example, the FeIIFeII MMOH sample shows one prominent feature at R ∼ 2.1 Å (Fig. 2A) that represents the first coordination sphere around the iron sites. This peak is best fit by a shell of 0.5 O/N atom per Fe at 2.02 Å and 4 O/N atoms at 2.20 Å (Table 2); this result is consistent with the presence of four carboxylates, two histidines, and two solvent molecules ligated to the diiron center, as revealed by the crystal structure of the FeIIFeII MMOH from Methylococcus capsulatus (Bath) (6). The feature at R ∼ 3.4 Å corresponds to an Fe-Fe distance of 3.43 Å (Table 2), which is slightly longer than the 3.28 Å distance deduced from x-ray crystallography (6). This Fe-Fe interaction could not be discerned in the EXAFS analysis of FeIIFeII MMOH from M. capsulatus (Bath) (18).

Fig. 2.

Fourier-transformed (FT) EXAFS experimental data (dotted line) and fits (solid line; see Table 2) of (A) FeIIFeII MMOH, (B) Q sample 1 (61% Q, trapped at 150 ms), (C) Q sample 2 (44% Q, trapped at 300 ms), and (D) decayed Q. The values of R on the x axis have been arbitrarily incremented by 0.35 Å, a typical phase shift, to provide a more realistic estimate of the metal-scatterer distances. Only shells with R < 3 Å were considered for the fits for Q samples 1 and 2. The vertical dashed line drawn through all the spectra highlights the presence of the 2.5 Å feature characteristic of intermediate Q.

Table 2.

X-ray absorption spectroscopic data analysis for samples in the catalytic pathway of MMOH. The diiron cluster concentration in Q samples 1 and 2 was ∼0.5 mM, whereas that for the FeIIFeII and decayed Q samples was ∼1.5 mM with 25% (v/v) glycerol. The latter sample was bubbled with O2 to ensure complete reaction. The buffer for all experiments was 100 mM MOPS (pH 7.7). The ranges for EXAFS data fitting and Fourier transform were as follows: 1.35 to 4.05 Å and 2 to 12 Å−1 (FeIIFeII MMOH); 1.25 to 3.15 Å and 2 to 11 Å−1 (Q samples 1 and 2); and 1.35 to 3.55 Å and 2 to 14 Å−1 (decayed Q). Values in parentheses represent σ2, the Debye-Waller factors, which reflect the extent of disorder (static and dynamic) within the shells. The negative values for the short Fe-O shells are typical of short iron-oxo bonds and merely reflect the weaker and longer bonds of the Fe(acac)3 standard (acac, acetylacetonate) (40).

View this table:

Figure 2D shows the R-space spectrum of a sample labeled “decayed Q” in which intermediate Q had been formed and then allowed to decay completely by standing at room temperature for 10 min, as indicated by the loss of its characteristic yellow color. The R-space spectrum resembles that previously reported for FeIIIFeIII MMOH (19) and can be simulated with similar parameters (Table 2); in particular, there are two different Fe-Fe distances, 3.03 and 3.32 Å, that correspond to the presence of two populations of diiron center with different core structures (19).

In contrast to the FeIIFeII MMOH and decayed Q samples, the two Q samples show three prominent peaks at R ∼ 1.5, 2, and 2.5 Å in their R-space spectra (Fig. 2, B and C). The 2 Å feature corresponds to a shell of 4 O/N atoms at ∼2.05 Å (Table 2), and its large Debye-Waller factor arises from the wide range of Fe-ligand bond lengths expected for a sample consisting of several different species. The new features at R ∼ 1.5 and 2.5 Å originate from an O/N atom at 1.77 Å and an Fe atom at 2.46 Å (Table 2). The intensities of these new features in the two Q samples and the partial occupancies derived from the fits are in agreement with the amount of Q determined by Mössbauer spectroscopy.

The pre-edge region of an x-ray absorption spectrum provides information that complements the EXAFS analysis, as the integrated area of the pre-edge 1s → 3d transition is sensitive to the centrosymmetric nature and the coordination number of the metal site (20, 21). The FeIIFeII MMOH and decayed Q samples have pre-edge areas of 10 and 14 units, respectively (Table 2); these values fall in the reported range for five-coordinate iron sites (20, 21). In contrast, each FeIV site of intermediate Q in samples 1 and 2 has a pre-edge area of 28 units (22). This large value implies that the FeIV centers in Q have a highly distorted geometry and are likely to have a coordination number no greater than 5. The highly distorted geometry and the consequent large pre-edge area observed here result from the presence of a short Fe-O bond (1.77 Å), analogous to observations for diiron complexes having an oxo bridge (20).

Our EXAFS analysis of intermediate Q thus shows a species with two structural features not found in either FeIIFeII or FeIIIFeIII MMOH (18, 19): a single 1.77 Å Fe-O bond per Fe atom and a 2.46 Å separation between the two FeIV ions. A length of 1.77 Å is too long to be associated with a terminal FeIV =O bond, as found in high-valent iron-oxo porphyrins and heme peroxidase compounds I and II (1.60 to 1.66 Å) (23), but is consistent with an Fe-μ-O bond of a (μ-oxo)diiron unit (1.74 to 1.82 Å) (24). A short Fe-Fe separation of ∼2.5 Å has not yet been observed for any diiron-oxo protein, although Fe-Fe distances of 2.7 Å are common for Fe-S proteins (25). Such short distances can only be enforced by the presence of two single-atom bridges forming a diamond-shaped core. Precedents of such a M2(μ-O)2 (M = metal) diamond core have been found for synthetic complexes (26, 27, 28, 29, 30) as well as for the O2-evolving complex of photosystem II (31, 32), where the metal-metal distances range from 2.5 to 2.9 Å.

The sole crystallographically characterized example of a synthetic Fe2(μ-O)2 diamond core, [FeIII2 (μ-O)2(6-Me3-TPA)](ClO4)2 [6-Me3-TPA = tris(6-methylpyridyl-2-methyl)amine], has a centrosymmetric rhomb with Fe-μ-O distances of 1.84 and 1.92 Å and an Fe-Fe distance of 2.7 Å (27). Given the EXAFS analysis and the Mössbauer evidence for nearly identical FeIV sites (12), we propose that MMOH intermediate Q has an analogous FeIV2 O2 diamond core structure (Scheme 1). Consistent with its higher valent state, the short Fe-O bonds (1.77 Å) for Q are shorter than those of the synthetic compound with an FeIII2(μ-O)2 core (27) but are comparable to those of a complex with an FeIIIFeIV(μ-O)2 core (28). The 2.46 Å Fe-Fe distance for Q is, however, shorter than those found in these model complexes (2.7 to 2.9 Å). In synthetic Mn2O2 complexes, the Mn-Mn distance can be shortened by ∼0.1 Å with the introduction of an additional carboxylate bridge (33). This observation prompts us to propose the presence of such a bridge for Q. Indeed, a carboxylate bridge (Glu144) is a common feature in all of the FeIIFeII and FeIIIFeIII MMOH crystal structures (5, 6, 7). This carboxylate bridge may serve to hold the diiron unit together as it proceeds through the catalytic cycle (1).

Scheme 1.

Diiron core structures of key species in the catalytic cycle of MMOH.

Scheme 1 illustrates the proposed structural changes in the diiron unit during catalysis. The Fe2(μ-O)2 diamond core has been proposed to be the key high-valent species in the oxidation of substrate in nonheme diiron enzymes on the basis of structural, spectroscopic, and reactivity studies of synthetic diiron complexes (15). Our combined Mössbauer-EXAFS investigation provides experimental evidence that such a diamond core participates in the MMO reaction cycle. This core can be readily attained by homolysis of the O-O bond in its precursor P (34), which, from a comparison with model complexes (35, 36, 37), is likely to be a (μ-1,2-peroxo)diiron(III) species (Scheme 1).

The FeIV2 (μ-O)2 diamond core addresses how nonheme diiron active sites store the oxidizing equivalents. In cytochrome P-450 (38), one oxidizing equivalent is proposed to be stored on the iron center and another on the porphyrin ligand, forming a short-lived FeIV=O porphyrin radical species, which is equivalent to heme peroxidase compound I (39). For MMO, the role of the porphyrin is assumed by the second iron, and each iron site serves to store one oxidizing equivalent by forming an FeIVFeIV species (12). The EXAFS-derived dimensions of the FeIV2(μ-O)2 core indicate that each iron has one short (∼1.77 Å) and one long (∼2.05 Å) Fe-μ-O bond, which suggests that this structure may be viewed as a head-to-tail dimer of FeIV=O units (Scheme 2). This dimerization may serve to stabilize the high-valent iron-oxo moiety, thereby allowing its observation and characterization in MMO. Although MMOH intermediate Q appears to be functionally equivalent to the reactive species of cytochrome P-450, the structural divergence of the key reactive species revealed here offers the possibility that a novel manifestation of the chemistry of highly activated oxygen in biology will emerge from continuing studies.

Scheme 2.

The spectroscopic studies described here provide direct experimental evidence for the participation of an Fe2(μ-O)2 diamond core in the catalytic cycle of MMO. This precedent lends credence to the hypothesis that such Fe2(μ-O)2 cores participate in the redox cycles of related nonheme diiron enzymes such as ribonucleotide reductase, fatty acid desaturases, and membrane alkane hydroxylases (1, 15).

REFERENCES AND NOTES

  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.
  31. 31.
  32. 32.
  33. 33.
  34. 34.
  35. 35.
  36. 36.
  37. 37.
  38. 38.
  39. 39.
  40. 40.
  41. 41.
View Abstract

Navigate This Article