.m..- . Alum uyswmmfin a. biufinifinhvu .«ubuhnflfi: a»? ‘ J} . nth: V . 7500.09.61. .b flux; u: 1:43 .. v. Bony-ho. - .. o u. ‘oav ‘.A . uuuuu ‘ .nfl‘o“)' nato.§hckcvhuc.i~v\‘ a v ‘1‘ r. AICJOA - 9s {#0 o v .A S i. a). . no" "ugws . . L-hvf—Q . i A c . u.cli '8 «In. . . p. m1.th ”1.4%. %§£ol%fluw. ”fififiwfihflflmflflu in .. M...“ . $1 SQ .3 I 1'1. . t! .o’c¢ v. v f . - gnaw- nun". :QHPbads . .394- ....lv.t....u..r.dur I . q ‘ . . i x v. kiln}; unfit/4...“ plrlik.‘ ¢ lulflfqilaufl... hbhboflflysflflfllisnddflP 1.u."l)|\bx a; 3": {‘1 £01931! ...... If: . < A . I. in . . bupqwun .1. l Vizun... “0...? “with... at. ‘ ‘ . 3 .. ‘. d... . .. ‘ h I‘ . I! ~ I. ’ U r. *' Ah"| "i‘Il.‘l-‘ In!!!” If. )1 ' "I . '. t .‘ H D m. l . 3.1'JHKL-r - E D .‘ a}; t I .‘fltfi‘ 431% ' if. H!‘ *3.- n1 ‘ I" I .* D" _‘| i. ' \ x.. A“. t. ink. . 1ND“?! ; I. .l_ 3;». * #141,: I” _ v ‘ ' . W £31 ta» 4"“: 13%. a ”'3' ' .. Him 'I 1,-~~' 3' '1 : _< '76; l . 1. \Uk O I I ‘5. I. I 2 a A l . ..o..l~i~..nvy\¢€.\v ‘nfii¢ I I ‘IO 0 '§ t..‘b ’ 59...? a -\.Iva bdlv’l‘fign - oIA . . . h. .u. A: A O . b. . .» \.w|‘v"0f . u yr 0. r:0vrv‘ f. 0‘. c I. u 2. . . . .. . lgzurhtafl. . ‘ I: .. .1. .7 .. Md. 1 .7 - . app 3 .f.’ o§I>.t-n‘1 . . . . .. . .i .2 . 7 .- to. s .7." luwioituuu. 1. vl.11:..i.41. a 1-.. . . 3.: . ... v. 21.1.1.1 . «I'I‘lu ilAJoI ’.§u..r v . . . - . . I ‘1 ~11 III . a. u \ ‘5. ‘l‘vil ‘ ‘4 l rOn'Q l n . ‘- 7 {n :‘l‘ a? I) . 1‘s.“ r Iflfifitlnbn...’ .40 .flncflllu . -- ..... . -’ svftliol‘o .. .V‘lcvau 1 . ....... -larx . ‘. .tl. . .- .. 13.....u.u511.l..!vk .-.Hfl..3m.h.u..£... . a u ‘ ...r.nh.1‘.‘.vc.b A .v . 2,. IV 7 .t' :o.‘ 0. all. "V"l\).u>l III! upturn. 2 z: , 1. . ‘ {$-13 ll 1 1 “IO.‘.. ”U“ "G‘ I £2. . . ) Wu.Al-\hk¢.uvf olvfilniullu‘... . .l IV II t . silk! 90.}! 51v... . . . . . shtulniidfluxihfi .I .a . .. . (mum‘fifldkuw ‘ x. - Uigxflu.3 Ificilplflqsllntixtiflh :Ii. . tn. . . .153?! . .c. o t. b mm“? tjbnfl31ufi ... ;.._. .é l\ lwll ‘ (r .. . . so nil .n .~4 U. ‘I . t V I ,.. I - .o-l 0. fi [:3 II o u |. .y; . - .DluotaL..I-I I . .‘ vl In tu N02- —> N20 —-> N2. \(NO?))' The first step is catalyzed by a molybdenum-containing nitrate reduc- tase similar to that found in plants and bacteria capable of nitrate assimilation (1), while the last step is catalyzed by an unusual copper-containing reductase (2). In contrast, the second step has remained controversial, with disagreement as to whether it is carried out by a single enzyme or by two enzymes, with NO as a free obligatory intermediate. Even among workers who favor a single enzyme, there is substantial controversy (3) regarding the mechanism by which two nitrite ions are converted to N20: (1) is the N-N bond of N20 formed by nucleophilic attack of a second NOZ' upon a metal-coordinated nitrosyl species (4) (Scheme 2); or (ii) is the nitrosyl intermediate first reduced to free nitroxyl (HNO), which spontaneously dimerlzes to N20 (5) (Scheme 3)? Although studies in a number of laboratories have been interpreted as favoring NO as a free obligatory intermediate (thus implying the existence of a separate NO reductase) (3,6,7), definitive evidence is still lacking. It has, however, been conclusively demonstrated by 13 Scheme 1. lh The pathway of denitrification, with identified or postulated intermediates indicated. 15 _:HZO II + E-Fen+ Nog== E-Fen-NOZ= E-Fe -NO 11w N20 + E-FeII +— *—-— E-Fen-(NZO3) Scheme 2. Formation of the N-N bond by nucleophilic attack of nitrite on an enzyme-bound nitrosyl. :HZO E-Fen + NO§=E-Fen NC,“ = E-Fenwo” 1°“ E-Fen +No‘ <-——E-Fen-NO- <—°— E-Fen-NO 22PU()-U"'_"_€’ [\hZC) Scheme 3. Formation of nitrous oxide by dimerization of nitroxyl anions. 16 82180 exchange and trapping experiments that nitrite reduction by whole cells (8) and the purified heme Edi-containing nitrite reductase (S) of Pseudomonas aerugigggg proceeds via an enzyme-bound nitrosyl inter- mediate (E-NO+). This nitrosyl intermediate arises from dehydration of coordinated nitrite (5,8), and is common to Schemes 2 and 3 (and possibly to NO formation from NOZ' as well (4)). Intermediates beyond the nitrosyl species remain nebulous, although evidence against the intermediacy of trans-hyponitrite (NZOZZ') (9) and oxyhyponitrite (N2032') (10) has been presented. The demonstration of positional isotopic equivalence of nitrogen atoms in product N20 (11) was in- terpreted as favoring HNO as an intermediate (Scheme 3), but in fact coordinated cis-hyponitrite is equally plausible if it undergoes a rapid intermolecular exchange of the coordinated and uncoordinated nitrogen atoms (12). Thus, with the possible exception of the oxyhyponitrite study (10), no evidence that would unambiguously distinguish between the three possible mechanisms has been reported. Competition between denitrification and nitrosation reactions has been demonstrated only using 15N0 and NHZOH, suggesting that an enzyme-nitrosyl intermediate is common to both reactions when NO, rather than N02- is the substrate (13); the relevance of this finding to nitrite reduction is perhaps open to question. The only experiments that have provided evidence, albeit indirect, for Scheme 3 vs. Scheme 2, are studies on the 15N isotope effect associated with denitrification of nitrite by E. stutzeri cells (14) and in soils (15), which found that the magnitude of the isotope effect increased with increasing nitrite concentration. The most obvious (but not unique) interpretation is that this result is 17 due to sequential addition of two nitrite ions to the enzyme prior to the first irreversible step. In a previous paper (13), we presented our first systematic study of dissimilatory nitrite reduction. With whole cells of g. stutzeri, it was shown that the extent of isotopic equilibration between H2180 and the product of denitrification, N20, decreased as the nitrite concentration increased, suggesting that H2180 and N02- compete for a common intermediate (i.e., favoring Scheme 2). In this work, we had to make the not unreasonable assumption that the 18O-enrichment of the free nitrite pool was negligible during the time of our measurements; this has now been confirmed experimentally by Shearer and Kohl (16). g. stutzeri nitrite reductase is thus a "sticky" enzyme, meaning that nitrite, once bound to the enzyme, is committed to react and does not readily dissociate. In this paper, we present the results of stable isotope studies relating to the pathway of nitrite reduction and to the effect of solubilization of the enzyme on the relative rates of the individual steps. We present evidence that strongly favors the pathway repre- sented in Scheme 2: N3", H2180, and N02- all compete for a common enzyme-bound nitrosyl intermediate. This implies that the N-N bond of N20 must be formed via nucleophilic attack of a second nitrite ion on a coordinated nitrosyl derived from the first nitrite. Our results are inconsistent with a pathway such as the one represented in Scheme 3, in which two equivalent nitroxyl anions combine to form nitrous oxide (5), and also eliminate NO as a free intermediate. 18 MATERIALS AND METHODS Cellggrowth and assay conditions. The bacterial strain used was g. stutzeri JM 300. Cultures were grown and harvested as described earlier (13), but cells were washed only once instead of three times. Cells were resuspended in a volume of 50 mM HEPES1 buffer pH 7.3, equal to 12 of the harvested volume. Cell disruption was by French press (3 passages, 12,000 psi) or by sonication with a Branson sonifier (2.5 min. on ice, 5x30 8. at 40% of maximum output with 30 8. intervals) as specified under Results and Discussion. Cell-free crude extracts were obtained by filtration through a 0.22 um filter. These extracts contained no whole cells, as they were unable to initiate growth when used as an inoculum into sterile medium, even after several weeks of monitoring for possible growth. Typical assay conditions are given here; variations are indicated in the text. Assays were carried out in 8 m1 serum vials sealed with butyl rubber stoppers and aluminum crimps. They contained 1 or 3 ml liquid phase. The final buffer concentration in assay mixtures was 25 mM HEPES (pH 7.3). The reducing system used was sodium succinate and the natural electron transport components present in the cell-free crude extracts. The reaction was initiated by anaerobic addition of nitrite. Labeled water (H2180) was added to the buffer. Addition of acetylene to inhibit nitrous oxide reductase proved unnecessary since this enzymatic activity virtually disappeared upon cell disruption. Vials were made anaerobic by flushing with argon. For each newly prepared extract, the specific activity was 1 The abbreviation used is: HEPES, N-2-hydroxyethy1piperazine- N'-2-ethanesulfonic acid. l9 .established by measurement of nitrous oxide production on the gas chromatograph. The incubation periods were chosen such that the extent of reaction, measured by GC and expressed as N20 produced as a fraction of the maximum N20 possible, would never exceed 201. This was crucial for the nitrosation experiments, where it was assumed that the nitrite and azide concentrations were roughly constant during the course of the experiment. Samples were analyzed by GC/MS immediately or stored frozen for later analysis. The rate of N20 production from 1 mM nitrite plus 50 mM azide was about 0.01 mM N/hour. Assays contained 0.1 to 0.3 ml crude cell free extract. Denitrification rates in crude extracts were two orders of magnitude slower than in whole cells. All experiments were reproduced at least twice. Reagents. All chemicals used were reagent grade. Azide stock solutions were prepared immediately before use. Stable isotopes were obtained from Monsanto (Mound, Ohio). 180- labelled water contained 15 atom: 180. Isotopic purity of 15N- labelled nitrite was better than 99.92. Controls. Sterile controls (autoclaved for 20 min. at 200°C) and controls without isotope were routinely performed for each experiment. Experiments were designed so that chemical N20 production was negligi- ble. GC/MS equipment and conditions. We used both an HP 5985 and an HP5995C GC/MS; the latter gave much better sensitivity. Source and mass analyzer temperatures were set at 150°C to prevent excessive decomposition of N20 to N2 and NO in the instrument. The electron multipliers were set in the range 200 to 1000 eV above autotune. RESULTS AND DISCUSSION Competition between denitrification and nitrosation. In our earlier work, we examined the competition between denitrification and nitrosation reactions using whole cells of g. stutzeri and NHZOH as a nucleophile to trap the nitrosyl intermediate. We found that hydro- xylamine was relatively inefficient at trapping the nitrosyl species, giving only a few percent nitrosation product at 10 to 320 mM NHZOH and 0.1 mM N02- (13). More importantly, however, we found that there was evidence for 180 exchange into an intermediate in the nitrosation reaction that is downstream from the enzyme-nitrosyl complex. Accordingly, we have examined azide as the trapping nucleophile. Since it cannot form an N202 intermediate during nitrosation, it de- creases the probability of 180 exchange into subsequent nitrosation intermediates. Azide is, however, completely ineffective as a nitrosyl trapping agent with whole cells under the conditions examined (1mM NalSNOZ, 1 to 10 mM N3', fresh Luria Broth medium as reductant); insignificant amounts of nitrosation products (§_IZ of total N20) were observed. Since azide is known to trap the nitrosyl intermediate in crude cell-free extracts (8) and in the purified heme £91 enzymes (5), this suggests that azide is not readily transported through the cell membrane against a charge gradient. Indeed, we find that cell free extracts prepared by French press treatment show up to 302 nitrosation with 50 mM NaN3 and 0.1 to 1.0 mM NalSNOZ. This finding is consistent with proposed location of nitrite reductase on the cytoplasmic side of the cell membrane (17). 20 21 If nitrite and a nucleophile such as azide are indeed competing for the nitrosyl intermediate common to Schemes 2 and 3, then it is possible to quantitate the flux through denitrification vs. nitrosation by measuring the relative amounts of N20-46 and N20-45 respectively (Scheme 4). Thus, the competition can be examined as a function of nitrite or azide concentration. Systematically varying the N3" concentration at constant [NOZ'] will not distinguish the pathways of Schemes 2 and 3, since in either case the amount of nitrosation is expected to increase with increasing [N3']. This type of experiment would simply confirm the existence of a trappable nitrosyl intermediate (5,8). In contrast, if the [NOZ'] is varied at constant [N3'], the predicted results differ for the two schemes. For the pathway in Scheme 2, the nitrosation: denitrification product ratio should decrease with increasing [NOZ'], since the second N02" will compete with N3’ for the nitrosyl intermediate. For a pathway such as that in Scheme 3, the nitrosation denitrification: ratio should be independent of [NOZ'], since nitrite does not appear in the scheme after formation of the nitrosyl intermediate. The data shown in Table 1 demonstrate that the N20-45/N20-46 ratio decreases with increasing [NOZ'], indicating that denitrification is indeed of a higher kinetic order with respect to nitrite than is nitrosation. At higher nitrite concentrations, a saturation effect is observed, as expected. Earlier attempts (5) to demonstrate this competition reaction failed, presumably because of the very high nitrite concentrations used (2 10 mM NOZ'). Effect of reductant on the nitrosation/denitrification ratio. In order to examine a more chemically defined system than LB medium, in 22 ‘5N15NO ‘5NO- denitrification ‘4N‘5No + N2 nitrosation Scheme 4. Nitrosation and denitrification products from 15N02 and N3". 23 Table 1 The ratio of nitrosation (N20-45) to denitrification (N20-46) products as a function of nitrite concentration. Numbers in parentheses are standard deviations. Conditions: [NaN3] = 50 mM, reductant = Luria broth; cell-free crude extracts of g. stutzeri prepared via French press; two or three replicates per nitrite concentration, all shown. Background N20 (m/z 45) was 1% of total N20 in control samples without azide. No significant amounts of N20 were formed in sterile controls. Isotope Abundance N20‘45 NaNOZ (mM) NZO-45 N20-46 ratio: --——-- N20-46 0.1 11440 46380 0.25 11080 48430 0.23 av. 0.24 z 0.01) 0.2 11650 48080 0.24 8865 41700 0.21 av. 0.22 (t 0.02) 0.5 3800 43820 0.087 6295 73710 0.085 av. 0.086 (t 0.001) 1.0 3565 60220 0.059 2320 49740 0.047 2680 54170 0.050 av. 0.052 (t 0.005) 24 which the actual reductant concentration is unclear, we carried out the same competition experiment as described above utilizing 50 mM sodium succinate as the reductant. The results are given in Table 2. The same general trend is observed as with Luria Broth medium, namely that the ratio of N20-45 (arising from nitrosation) to N20-46 (arising from denitrification) decreases with increasing concentration of nitrite, at least for [N02-].S 1 mM. The relative amount of nitrosation at a given nitrite concentration is, however, approximately a factor of 2-3 higher with succinate vs. LB medium, for reasons that are not clear. The fact that the trend in the N20-45/N20-46 ratio reversed between 1 and 10 mM nitrite suggested that reductant had become rate- 1imiting for denitrification at the highest nitrite concentrations. Accordingly, we examined the effect of varying the succinate concentra- tion at 0.05 and 10 mM NOZ' (Table 3). The data clearly show that the nitrosation:denitrification product ratio decreases with increasing succinate concentration at 10 mM N02". The variability between repeated experiments is relatively large for less than saturating levels of succinate, because the concentration of residual reductant from the growth medium and the internal reductant from the cells themselves vary, depending on how long the cells were starved for carbon prior to harvesting. Denitrification rates were independent of reductant at very low N02" concentrations, as expected. Oxygen-18 content of nitrous oxide from denitrification vs. nitro- sation. The major difference between the two proposed mechanisms of denitrification (Schemes 2 and 3) is the nature of the two nitrogen species that form the initial N-N bond. In Scheme 3, the N-N bond is 25 TABLE 2 Ratio of nitrosation (N20-45) to denitrification (N20-46) products as a function of nitrite concentration with succinate as reductant. Condi- tions as in Table 1 except for use of 50 mM sodium succinate in place of Luria Broth medium. Isotope Abundance N20-45 NaNOz (mM) N20~45 N20-46 ratio: N20-46 0.05 7110 9834 0.72 8710 10080 0.86 av. 0.79 (i 0.07) 0.10 9555 21500 0.44 10350 22900 0.45 av. 0.44 (i 0.01) 1.0 11313 56575 0.20 11075 72925 0.15 av. 0.17 (i 0.03) 10.0 411450 41500 0.28 8913 39750 0.25 av. 0.27 (t 0.03) 26 TABLE 3 Effect of reductant on ratio of nitrosation to denitrification pro- ducts. Conditions as in Table 2 except for indicated concentrations of succinate. Data from repeated experiments using different cell preparations are indicated by *. [succinate] (mM) N20-45 / N20-46 [NOZ'] = 10 mM [N02‘] = 0.05 mM 10 0.31 x 0.03 0.57 i 0.04 0.59 i 0.02* 50 0.19 z 0.02 0.49 t 0.01 0.25 t 0.05* 100 0.17 1 0.01 0.54 1 0.10 0.17 z 0.01* 27 formed by reaction of two equivalent nitroxyl anions (NO’). Therefore, if the reaction is carried out with ISNOZ' in the presence of H2180, the 180 content of the E-ISNO+ (nitrosyl) intermediate will be reflected in the 180 content of the denitrification product, 15N20, which can be measured directly. The 180 content of the nitrosyl intermediate can be measured by trapping it with 14N3' to produce 1[‘Nl-SNO; the 180 content of the nitrosation product will reflect that of the E-NO+ intermediate. In Scheme 2, however, the two nitrogen species are not equivalent at the point at which the N-N bond is formed by attack of free NOZ' on the E-NO+ intermediate. Only the latter will contain appreciable amounts of 180. Free nitrite will be virtually unlabelled, since the rate for dissociation of N02" from the enzyme is very low (16), and any 18O-labelled nitrite formed is diluted by the nitrite pool. The results of experiments in which the 180 content of N20 originating from denitrification (N20-48/(N20-46 + N20-48)) and from nitrosation with N3- (NZO-47/(N20-45 + N20-47)) was determined are shown in Table 4. We find that 15N20 from denitrification is 502-602 equilibrated with the H2180, while 14leNO from nitrosation is 80-852 equilibrated with the 32180. This is consistent only with Scheme 2, in which nitrite containing no 180 dilutes the 180 content of the E-NO+ intermediate. The ratios of the extent of equilibration are not exactly 1:2, as predicted by Scheme 2 and in more detail in Scheme 5 below, assuming that oxygen atoms are lost with equal probability from either nitrogen atom during conversion of the dinitrogen intermediates to N20. One possible explanation is the fact that, even at high [NOZ'] in the absence of N3', we always observe £3; 82 equilibration of N20 with 28 TABLE 4 18O isotopic enrichment of denitrification (N20-46 and N 0-48) and nitrosation (N20-45 and N20-47) products. f [NalsNOZ] 8 1.0 mM, reductant = 100 mM sodium succinate; cell-free extracts prepared by sonication. Isotope Abundance Isotope Abundance Conditions: NaN3] - 50 mM, N20‘46 N20-48 7. Equilib. N20-45 N20-47 Z Equilib. 11,807,400 620200 53.2 8,702,400 776,400 88.83 13,823,800 838500 51.2 9,364,600 760,500 81.3é av. 57.2 (2 4.0) av. 85.1 (i 3.8) Isotope Abundance Isotope Abundance N20-46 N20-48 z Equilib. N20-45 N20-47 z Equilib. 28,000 1235 55.0 3110 210 83.92 25,380 1085 53.3 2600 160 76.62 17,930 730 50.7 2810 180 79.72 16,200 670 51.6 2280 150 81.82 av. 52.7 (i 2.0) av. 80.5 (i 2.4) In: Data obtained on HP S995C using 9 atom 2 H2180. 2 Data obtained on HP 5985 using 7.3 atom 2 H2180. 29 Scheme 5. A detailed representation of the pathway by which N02" is reduced to N20 via sequential reaction of two N02- ions with the enzyme. I // —0‘ 0‘ /0‘ - + Feat-N ‘——___, 862* / :29' 2” 7e“ - N l O" l \\ '_H20 \\ + _ N — 0 /N — o 'O (121‘) (SZI) (I) +214+ +120 +2e" O I // Fe” - N" Fez" - N II I \ ~+ N+—O. u // O O ( 1211) (I!) A -N o .. 2 +N02 V 1': 2* _..‘"°3 Fe“ - No; ”“9“” Fe“ - Na 0* 1 “N02 +Hgo,-2H* (I) (II) (III) w +6- .9. h | . +9- ..e' Fez. - N | 0 (3:) +9.. ‘ -e‘ I 'NO‘ . 3’ 1 fi —— 3.- Fie +~o° “’ Fla N \0 (III) (II) 31 H2180 (13). This indicates either that N02” is unable to completely suppress 180 exchange into the E-NO+ intermediate or that some 180 exchange occurs via a denitrification intermediate containing an N-N bond, as we have observed for nitrosation with NHZOH (13). Mechanistic implications. The results presented above demonstrate that nitrite competes with azide for the E-NO+ intermediate that is formed by dehydration of nitrite. Our earlier work (13) and the isotope dilution experiments described above demonstrate that nitrite competes with H2180 (and with azide) for the same intermediate, presum- ably E-NO+. These results are consistent 2212 with a sequential mecha- nism for reaction of two nitrite ions to form N20 (e.g., a mechanism such as that outlined in Scheme 2), and eliminate the nitroxyl mechanism (Scheme 3) from further consideration. Because the mechanism of denitrification has been the object of substantial controversy, it is worthwhile to briefly consider the major lines of evidence adduced previously in favor of other mechanisms and against the sequential mechanism indicated by the present data. To facilitate this, a more detailed version of the mechanism shown in Scheme 2 is presented in Scheme 5; this is an expanded version of the hypothetical pathway presented in our original paper (4). This pathway (upper portion of Scheme 5) has several key features. Initially, nitrite binds to a ferrous heme (I) and is dehydrated to a reactive (18,19) ferrous-nitrosyl complex (III) via the ferrous-nitrite complex (II), as demonstrated by Hollocher and co-workers (5,8). NucleOphilic attack of a second nitrite on the coordinated NO+ of III produces IV, containing bound N203. Reduction by two electrons produces V, containing coordinated oxyhyponitrite, N2032'. Reduction 32 by a second two electrons and dehydration produces a species (VI) containing coordinated gig-hyponitrite, NZOZZ', which upon further dehydration yields the ferrous-N20 complex (VII). Loss of N20 from VII regenerates the ferrous heme (I). The lower portion of this scheme shows how evidence previously interpreted as favoring NO as a free intermediate may be accounted for and is discussed below. The major argument presented in favor of nitroxyl as an inter- mediate (8,11) has been the positional isotopic equivalence of nitrogen in IA’ISNZO produced by concomitant reduction of 15N02' and 14NO observed by Garber and Hollocher (11). Their data argue for a symmetrical intermediate in the reaction, which could be gighgg a free mononitrogen intermediate (e.g., HNO) 9; an effectively symmetrical dinitrogen intermediate. The latter is perfectly consistent with a sequential mechanism if the coordinated gig-hyponitrite intermediate (VI in Scheme 5) interconverts rapidly between the two isomers with different nitrogen atoms coordinated to iron (VI 2 VI'). Available chemical evidence suggests that this equilibration is likely to be very rapid. For example, variable temperature NMR studies have shown that 2 nitrogen atoms substituted pyridazines (which also contain two sp linked by a formal double bond) when bound to ruthenium porphyrins exchange nitrogen donor atoms in an intermolecular process at rates of 102-106 sec‘1 (12). Since substitution reactions of ruthenium com- plexes are generally much slower than for the corresponding iron complexes, one would expect such reactions at iron to be very rapid indeed, much faster than the overall enzymatic reaction. Thus, both the data on positional isotopic equivalence (11) and the 18O enrich- ments (8) reported by Garber and Hollocher are equally consistent with 33 either the nitroxyl or sequential mechanism. Scheme 5 postulates that oxyhyponitrite, N2032', is an enzyme- bound intermediate. Since Na2N203 is readily prepared (20), it is possible to examine whether N2032' is converted to N20 by the enzyme. Experiments with several denitrifying bacteria and whole cell extracts have been reported by Garber, Wehrli, and Hollocher as evidence that N2032’ "can be neither a free nor an enzyme-bound intermediate" in denitrification (10). This conclusion is open to question on two levels. First, the bacteria and extracts used showed very low denitri- fication activity (on the order of only 2-fold higher than controls with no cells, and in one case zero activity). The lability of HN2032', the decomposition of which to NO' and N02' is markedly catalyzed by metal ions (49a), is expected to lead to large background levels of gaseous products with whole cells or crude cell-free extracts. We have performed similar experiments with purified gd- type and Cu-containing nitrite reductases that have been extensively treated to minimize contamination by adventitious metal ions, and still find relatively high background levels of gaseous decomposition products (C. Hulse, E. Weeg-Aerssens, J. M. Tiedje, and B. A. Averill, unpublished results). Even if the data of Garber, Wehrli, and Hollocher (10) are accepted at face value, their interpretation is open to question. Examination of the enzymological literature reveals no general answer to the question of what one should expect when an enzyme is confronted with a putative intermediate that does not normally dissociate from the enzyme. There are, however, several specific cases in which this phenomenon has been examined. For example, oxaloacetate and NADPH are 34 postulated as nondissociable intermediates in the reaction of malic enzyme, yet the conversion of oxaloacetate and NADPH to L-malate and NADP+ is catalyzed by the enzyme at only 102 of the Vmax with NADPH, C02, and pyruvate (21). Similarly, formyl phosphate is an enzyme-bound intermediate in the formyltetrahydrofolate synthetase reaction, yet is turned over by the enzyme at ca. 32 of the rate of the normal sub- strates (MgATP, H4folate, and formate) (22). These results have been explained in terms of a sequential mechanism with a kinetically trapped intermediate (i.e., one with both a slow dissociation and a slow binding step) (22). Similar behavior for species V'in Scheme 5 is not unreasonable, and would render detection of enzymatic activity diffi- cult with a labile substrate such as N2032'. In the case of formyltetrahydrofolate synthetase, the lability of formyl phosphate prevented detection of catalytic activity with it as a substrate for over 25 years (23,24). The other major alternative mechanism for denitrification, proposed over a decade ago, postulates the existence of two enzymes, a nitrite reductase that produces NO as the sole product and a separate NO reductase that reduces 2 N0 molecules to N20 (25) (cf. Scheme 1). The evidence supporting the existence of two enzymes and NO as a free obligatory intermediate is: (i) the observation that most purified nitrite reductases produce only NO from NOZ', while at least small amounts of NO reductase activity are found in other fractions (6,7, 26- 29); (ii) denitrifying bacteria and cell-free suspensions produce and consume NO during nitrite reduction (30-35); (iii) nitrite reductase catalyzes exchange of N between isotopically labelled nitrite and a pool of added NO during reduction of N02" to N20 (36,37); and (iv) 35 formation of EPR signals due to ferric heme-NO complexes upon addition of nitrite to purified nitrite reductase (38-40). All of the above evidence, however, can be equally well explained in terms of a sequential mechanism catalyzed by a single enzyme (Scheme 5). Even though there is substantial evidence for the existence of two crude fractions in cell-free extracts of denitrifiers, this has not matured into proof, as a purified NO reductase has thus far eluded all investigations (7,25,34,41-44). The chemical reactivity of NO makes it reasonable to suggest that at least some conversion of NO to N20 may be due to secondary or non-physiological activities of other cellular components, as shown recently by Zumft for ferrous iron-ascorbate mixtures (45). Arguments (ii) and (iii) above can be readily interpreted in terms of the reactivity of the ferrous-nitrosyl inter- mediate III, as shown in the bottom portion of Scheme 5. Studies with synthetic heme nitrosyls (18,19) indicate that, in contrast to ferrous heme-NO complexes, the N0 of the one-electron oxidized species is labile (reaction III 2 IX), producing the ferric heme and free NO. This reaction would account for the production and consumption of NO by denitrifiers, for the small and relatively constant pool of NO observed during reduction of nitrite (30,35,37), for the exchange of labelled N between N02- and added NO (36,37), and for N20 production (25), cell growth (33), active transport (46), and proton translocation (7) with NO as sole electron acceptor, since the dehydration reaction (II 2 III) is known to be reversible (5,8,47). The traditional explanation for argument (iv) above has been the sequence N02- + ferrous heme°NO + NO. Since the dissociation of NO from ferrous heme-NO complexes is extraordinarily slow (even slower than CO dissociation (48)), this 36 sequence seems unlikely. Indeed, there is no evidence that the ferrous heme-NO complex forms or decays within the turnover time of the enzyme (i.e., that it is kinetically competent). As indicated in Scheme 5, this species (X) can form in a variety of ways and is irrelevant to the catalytic mechanism- Finally, an alternative sequential mechanism exists that is also consistent with the results reported herein. This involves a reaction of a coordinated nitroxyl anion (Fe-NO-) with N02“ to produce coordi- nated oxyhyponitrite (Fe-N2032'; species V'in Scheme 5) directly, and has as precedent the known reaction of HNO with N02“ to produce HN203' (49). Recent electrochemical studies by Kadish (19) and Fajer (50) have shown, however, that it is very difficult to reduce ferrous heme nitrosyl complexes (Fe2+-NO°). Reported reduction potentials for the Fe2+-NO'/Fe2+-NO' couple with porphyrin and related ligands are in the range of -0.8 to -0.9 V vs. Standard Hydrogen Electrode (19.50), and do not vary greatly with the nature of the macrocyclic ligand (porphyrin vs. chlorin vs. isobacteriochlorin (50)) or the axial ligand (19,50). Since the biological reductant is either ascorbate (E5 - + 60mV) or succinate (E5 = + 30mV) and E5 for the NOZ'INZO couple is +0.77V, it is difficult to accept the intermediacy of such a strongly reducing species as the coordinated nitroxyl anion in denitrification. (This argument is also relevant to the nitroxyl mechanism proposed earlier by Hollocher (5) and discussed above.) Effect of enzyme solubilization on the fate of the nitrosyl inter- mediate. If, as seems plausible, a common mechanism applies for at least all heme ggl-containing nitrate reductases, the finding of Garber and Hollocher (37) that different denitrifiers exhibit varying 37 degrees of 180 exchange into product N20 suggests that differences in active site environment may affect the partitioning of the nitrosyl intermediate among the three competing reactions shown in Scheme 5. A similar change in active site environment may also explain the apparent shift from N20 production in whole cells or cell-free extracts to NO production in purified heme £91 nitrite reductase. All that is required is that the relative rates of reactions III 9 IV and III 9 IX in Scheme 5 are reversed in the purified vs. membrane-bound enzyme. We have now obtained preliminary data using isotope labelling studies that suggest that the fate of the nitrosyl intermediate is indeed affected by the extent of solubilization of the enzyme. We have shown previously (13) with whole cells of g. stutzeri that the extent of 180 exchange between H2180 and product N20 is a function of the nitrite concentration. The extent of equilibration ranged from 352 at 90 uM nitrite to 7.92 at 9 mM nitrite. We have now repeated this experiment using cell-free crude extracts prepared by sonication. As shown in Table 5, the same trend toward increased 180 equilibration at lower nitrite concentrations is observed with the sonicated extracts, but all values are 3-4 fold higher than the results with whole cells. The competition experiments described in Tables 1-3 above employed cell-free crude extracts prepared by French press. For nitrite concentrations ranging from 50 uM to 1 mM and azide concentrations of 50 mM, we found that the fraction of total N20 formed via denitrifica- tion ranged from 56 to 832. Data obtained under the same conditions for cell-free extracts prepared by sonication are given in Table 6. Even though the general trend is again the same, with an increasing TABLE 5 18O isotopic enrichment of denitrification products as a function of nitrite concentration for sonicated crude cell-free extracts. Condit- iopg: reductant 8 100 mM sodium.succinate; medium contained 9 atom 2 H2 0. Isotope Abundance [N027] mM N20-44 N20-46 2 Equilibration 10 20,131,177 465,193 23.4 20,881,770 621,293 30.5 av. 27.0 (i 3.6) 1 21,923,078 1,254,134 59.2 28,666,146 1,692,224 61.1 av. 60.2 (i 1.1) 39 TABLE 6 Competition between nitrosation and denitrification in crude cell free extracts prepared by sonication. Conditions: as in Table 1 except for reductant . 100 mM sodium succinate; crude extracts prepared by sonication. Isotope Abundance N20-4S [NaNoz‘] (mM) N20-45 N20-46 Ratio: N20-46 0.05 20,020 164,100 0.122 21,380 168,200 0.127 av. 0.125 (a 0.003) 0.10 18,220 375,750 0.049 26,080 369,900 0.071 av. 0.060 (t 0.011) 1.0 21,270 405,950 0.052 18,990 420,780 0.045 av. 0.049 (1 0.004) 10.0 82,330 449,140 0.155 76,840 462,700 0.142 av. 0.149 (t 0.007) 40 fraction of N20 due to denitrification with increasing nitrite concen- tration, the actual figures are very different, with nitrosation account- ing for only 5 to 102 of total N20. (Once again reductant apparently becomes rate limiting at high [NOZ'], as noted previously in Table 2.) These observations, although preliminary in nature, suggest that isotope labelling studies can be used to probe alterations in relative rates of individual steps within the catalytic mechanism as the enzyme is purified. More detailed studies will be required to correlate observed changes in relative rates with the physical state of the enzyme (e.g., lipid content of the preparation, degree of aggregation) and are in pro- gress. Nonetheless, these results suggest that it is not unreasonable to look to perturbations in active site environment to explain the shift in product from N20 to NO as the enzyme is purified, rather than invoking the existence of separate N02' and NO reductases. Acknowledgments This research was supported by Grant CHE-8607681 from the N.S.F. Chemistry of Life Processes Program. We thank I. Toro-Suarez for assistance with the mass spectrometry and F. Dazzo for use of the Mass HP 5995C MS. Mass spectrometric data were obtained at the N.I.H./M.S.U. Spectrometry Facility, which is supported by Grant RR 0480 from the N.I.H. Division of Research Resources, and in the laboratory of F. Dazzo, Department of Microbiology and Public Health, M.S.U., where the purchase of the HP 59950 MS was made possible by Grant 151 RR 02704-01 from N.I.H. 10. 11. 12. LITERATURE CITED (a) Payne, W. J. Denitrification, Wiley-Interscience: New York, 1981, p. 68, (b) Adams, M. W. W., and Mortenson, L. E. in Molybdenum Enzymes, Spiro, T. G., ed., Wiley-Interscience: New York, 1985, pp. 519-593. (a) Zumft, W. G., and Matsubara, T. (1982) FEBS Lett. 148, 107-112, (b) Coyle, C. L., Zumft, W. G., Kroneck, P. M. H., Kdrner, H., and Jakob, W. (1985) Eur. J. Biochem. 153, 459-467, (c) Snyder, S. W., and Hollocher, T. C. (1987) J. Biol. Chem. 262, 6515-6525. Henry, Y., and Bessieres, P.(1984) Biochimie (Paris) 66, 259-289. Averill, B. A., and Tiedje, J. M. (1982) FEBS Lett. 138, 8-12. Kim, C.-H., and Hollocher, T. C. (1984) J. Biol. Chem. 259, 2092- 2099. Grant, M. A., Cronin, S. E., and Hochstein, L. I. (1984) A553; Microbiol. 140, 183-186. Shapleigh, J. P., and Payne, W. J. (1985) J. Bacteriol. 167, 837-840. Garber, E. A. E., and Hollocher, T. C. (1982) J. Biol. Chem. 257, 8091-8097. Hollocher, T. C., Garber, E., Cooper, A. J. L., and Reiman, R. E. (1980) J. 3191 Chem. 255, 5027-5030. Garber, E. A. E., Wehrli, S., and Hollocher, T. C. (1983) J. Biol. 9359; 258, 3587-3591. Garber, E. A. E., and Hollocher, T. C. (1982) J. Biol. Chem. 257, 4705-4708. Eaton, S. S., Eaton, G. R., and Holm, R. H. (1972) J. Organomet. Chem. 39, 179-195. 41 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 42 Aerssens, E., Tiedje, J. M., and Averill, B. A. (1986) J. Biol. Chem. 261, 9652-9656. Bryan, B. A., Shearer, G., Skeeters, J. L., and Kohl, D. H. (1983) g; Biol. Chem. 258, 8613-8617. Mariotti, A., Germon, J. C., and Leclerc, A. (1982) Can. J. Soil Sci. 62, 227-241. Shearer, G., and Kohl, D. H. submitted for publication. Saraste, M., and Kuronen, T. (1978) Biochim. Biophys. Acta 513, 117- 131. Wayland, B. B., and Olson, L. W. (1974) J. Am. Chem. Soc. 96, 6037- 6041. Olson, L. W., Schaeper, D., Lancon, D., and Kadish, K. M. (1982) g; Am. Chem. Soc. 104, 2042-2044. Addison, C. C., Cramlen, G. A., and Thompson, R. (1952) J. Chem. Soc. 338-345. Schimerlik, M. I., and Cleland, W. W. (1977) Biochemistry 16, 576- 583. Smithers, G. W., Johansouz, H., Kofron, J. L., Himes, R. H., and Reed, G. H. (1987) Biochemistry 26, 3943-3948. Jaenicke, L. v., and Brode, E. (1961) Biochem. Z. 334, 108-132. Sly, w. s., and Stadtman, E. R. (1963) J. Biol. 011% 238, 2639-2647. Payne, W. J. Denitrification, Wiley-Interscience: New York, 1981. Payne, W. J., in Denitrification, Nitrification, and Nitrous Oxide, Delwiche, C. C., ed., John Wiley and Sons: New York, 1981, pp. 85- 103. Payne, W. J., Riley, P. S., and Cox, C. D., Jr. (1971) J. Bacteriol 106, 356-361. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43 Cox, C. D., and Payne, W. J. (1973) Can. J. Microbiol. 19, 861-872. Matsubara, T., and Iwasaki, H. (1971) J. Biochem. 69, 859-868. Wharton, D. C., and Weintraub, S. T. (1980) Biochem. Biophys. Res. Commun. 97, 236-242. Bessieres, P., and Henry, Y. (1984) Biochimie 66, 313-318. Pichinoty, F., Garcia, J.-L., Mandel, M., Job, C., and Durand, M. (1978) C. R. Acad. Sci. (Paris) 286, 1403-1405. Pichinoty, F., Mandel, M., and Garcia, J.-L. (1979) J. Gen. Microbiol. 115, 419-430. Payne, W. J., Grant, M. A., Shapleigh, J., and Hoffman, P. (1982) J; Bacteriol. 152, 915-918. Betlach, M. R., and Tiedje, J. M. (1981) Appl. Environ. Microbiol. 42, 1074-1084. Firestone, M. K., Firestone, R. B., and Tiedje, J. M. (1979) Biochem. Biophys. Res. Commun. 91, 10-16. Garber, E. A. E., and Hollocher, T. C. (1981) J. Biol. Chem. 256, 5459-5465. Cox, C. D., Payne, W. J., and DerVartanian, D. V. (1971) Biochim. Biophys. Acta 253, 290-294. LeGall, J., Payne, W. J., Morgan, T. U., and DerVartanian, D. (1979) Biochem. Biophys. Res. Commun. 87, 355-362. Johnson, M. K., Thompson, A. J., Walsh, T. A., Barber, D., and Greenwood, C. (1980) Biochem. J. 189, 285-294. Knowles, R. (1982) Microbiol. Rev. 46, 43-70. Bryan, B. A., in Denitrification, Nitrification and Atmospheric Nitrous Oxide, Delwiche, C. C., ed., John Wiley and Sons: New York, 1981, pp. 676-684. 43. 4&0 45. 46. 47. 48. 49. 50. 44 Grant, M. A., Cronin, S. E., and Hochstein, L. I. (1984) Aggy; Microbiol. 140, 183-186. Hoglen, J., and Hollocher, T. C. (1986) Fed. Proc. 45, 1604A. Zumft, W. G., and Frunzke, K. (1982) Biochim. Biophys. Acta 681, 459-468. Garber, E. A. E., Castignetti, D., and Hollocher, T. C. (1982) Biochem. Biophys. Res. Commun. 107, 1504-1507. Swinehart, J. H. (1967) Coord. Chem. Rev. 2, 385-402. Chang, C. K. (1985) J. Biol. Chem. 260, 9520-9522. (a) Bazylinski, D. A., and Hollocher, T. C. (1985) Inorg; Chem. 24, 4285-4288, (b) Akhtar, M. J., Lutz, C. A., and Bonner, F. T. (1979) Inorg. Chem. 18, 2369-2375, (c) Hughes, M. N., and Wimbledon, P. E. (1977) J. Chem. Soc. Dalton Trans. 1650-1653, (d) Bonner, F. T., and Ravid, B. (1975) Inorg. Chem. 14, 558-563. Fujita, E., and Fajer, J. (1983) J. Am. Chem. Soc. 105, 6743-6745. CHAPTER FOUR PURIFICATION, MOLECULAR CHARACTERIZATION AND KINETIC PROPERTIES OF ESEUDOMONAS SIQIZERI CYTOCHROME £21 TYPE NITRITE REDUCTASE INTRODUCTION Two types of dissimilatory nitrite reductases are known to play a role in bacterial denitrification: one is more predominant among denitrifiers isolated to date and contains two kinds of hemes (c,g1). The other one is not a cytochrome, but contains copper and was first found in Alcaligeneg faecalig (formerly Egggggmggas denitrifigans; 1). Our work has been limited to nitrite reductases of the first type. The cytochrome gjgl type dissimilatory nitrite reductase (ferrocytochrome cznitrite oxidoreductase) has been purified in the past from several denitrifying bacteria. Original purification procedures are the following: W W (2.3.4.5.6.7.8). W115. en t (9.10). 5157;118:112; M111 (formerly We: gggigrifiiggng (5,11,12), Paragoccug dgnignifiigapg (formerly 8.191% W) (13.14.15) and 2217.29.99.92; WM (16). They are all dimers of two equivalent subunits with a total molecular weight in the 120 to 130 kDa range and with two hemes (l heme c and l heme d1) per subunit. Purification methods which gave high yields were the one for Ihiobacillus genigzifiiggpg (10), with 100 grams of cell paste yielding 45 mg pure enzyme and the Algaliggpgg faecalig nitrite reductase preparation (11), yielding 56 mg enzyme 45 46 from 4650 mg total protein. The most popular contemporary method gave a 25% recovery (7). A comprehensive review on denitrification and nitrite reduction is available (17). Over the last few years we have studied the pathway of dissimilatory nitrite reduction in Pseudomgpag stutzeri JM 300. In 1982, Averill and Tiedje (18) proposed a pathway in which nitrite is reduced by the enzyme to both nitric oxide and nitrous oxide, with nitric oxide postulated to be being an "abortive" product, i.e. in equilibrium with enzyme-bound nitric oxide (nitrosyl). We do not believe that nitric oxide is an obligatory intermediate in the conversion of nitrite to nitrous oxide by this enzyme. Isotope studies have provided conclusive evidence that, at least in this organism, the N-N bond in nitrite reduction is formed by nucleophilic addition of a second nitrite molecule to an enzyme bound intermediate (19,20; Ch. 3, this work). We believe this intermediate is most likely a heme- nitrosyl; there is also chemical precedence for N-N bond formation by nitrosation of metal stabilized nitrosyls, catalyzed by this enzyme (21). There has been much controversy in the past about the identity of the product(s) of nitrite reductases. Not only were there differences in behaviour between nitrite reductases from different bacterial species, but also between different preparations of the nitrite reductase from the same species. Most often, nitric oxide was reported as the only product of nitrite reduction (3,10,11,14,16,22,23). Occasionally, nitrous oxide was detected along with nitric oxide, but as a minor product (5,9,12,21,24,25,26). Nitrous oxide as major product has only been reported twice (27,28): both found transient formation of nitric oxide in the process of nitrite reduction to N20. 47 Nitrite reductase, as a consequence of its involvement in a respiratory pathway, is generally accepted to be membrane associated in vivo. A pool of soluble nitrite reductase is always present after cell disruption, but a fraction of the nitrite reductase pool is more tightly membrane associated (16,29,24). At least a significant portion of the enzyme is soluble and is most efficiently released from the membrane by sonication. There may be a chemical equilibrium in the cell between a membrane-bound and a soluble pool, or the soluble pool may be an artifact from cell breakage. In Eseudgmonas perfectgmarinug, 15% of the nitrite reductase activity was more tightly bound to the membrane than the rest, which led to the speculation that there were two "kinds" of nitrite reductases in the cell (24). The membrane bound nitrite reductase of Paracoccus halodenigzificaps was extracted with detergent, purified and shown to be identical to the cytoplasmic nitrite reductase (l6). Membranes from denitrifiers have been shown to possess nitric oxide reducing activity (24,26,30) and to reduce nitrite to nitrous oxide without detection of free nitric oxide (16,24,30). NO reduction generates energy through proton translocation (31) and thus it seems that nitric oxide reduction, like nitrite reduction, is part of a respiratory pathway. Nitrite reducing activity and nitric oxide reducing activity have been differentially extracted from membranes (30,32) and have been physically separated from each other. Reports on partially purified nitric oxide reductases are scarce and poorly characterized: an iron-flavoprotein in 2. agrugingga (33) and in 2. ggpgzggi (34); a cytochrome c in 2. pggfggpgmggigug (23,35). Nitrite reductases have been reported to reduce exogenous nitric oxide (27) and one purified "nitric oxide reductase" later 48 turned out to be the cytochrome gidl nitrite reductase (5). Nitric oxide reduction is typically much slower than nitrite conversion to nitrous oxide (21,24,36). Whole cells both produce and consume nitric oxide (5,11,33,36,37,38,39,40). This work is focused on the purification, molecular properties and reactivities with nitrite and nitric oxide of the purified nitrite reductase from Eggpggmgpag figugzggi JM300. Growth and partial purification of the nitrite reductase from Eggpggmgngs stutzeri (Van Niel strain) has been described by Kodama (41). We have obtained a highly purified nitrite reductase from Pseudomgnas stutzeri and used it address the following questions: (i) What are the substrates and products of the pure enzyme ? (ii) If significant N20 production from nitrite occurs, is this process kinetically competent (are Km values low and Vmax values high enough) to have potential ecological significance? (iii) Is there any effect of purification on the apparent kinetic parameters of the enzyme? (iv) How similar is this nitrite reductase to previously characterized nitrite reductases? 49 MATERIALS AND METHODS Qrgwtb gf 2, sputzeri. P. figggzggi JM300 was grown in the following medium: 30 g/l Tryptic Soy Broth (Sigma) supplemented with KNOB (5 g/l), NaHCO3 (2 g/l),10 uM CuSOA and 10 mg/l FeS04.7H20. Here we describe the purification of nitrite reductase from 15 1 of this medium, which yielded 65.3 g of wet cell paste. Cultures were routinely transferred as 0.5% inocula to 100 ml fresh medium, in 155 ml glass serum bottles with butyl rubber stoppers crimped with aluminum seals. The headspace above fresh medium was not made anaerobic, as the bacteria used the oxygen present initially and create their own anaerobic, denitrifying growth conditions. Larger quantities of bacterial cells were prepared by inoculating 100 ml of actively growing bacteria in late exponential phase (18 to 20 h after inoculation) into 15 l of medium in a large glass container closed with a rubber stopper and equipped with a sterile gas outlet: a 0.22 um filter connected to a hypodermic needle, inserted into the rubber stopper. Cultures in small serum bottles were grown while shaking at 100 rpm in a 37 C incubator. The 15 1 culture flasks were shaken at 50 rpm. The cells were harvested by centrifugation (12,000 g x 15 min) during late exponential growth when the gas (N2) formation rate has reached its maximum and has begun to decline: this was monitored by subsampling and measuring the rate of N2 evolution from a 1 ml subsample of cells assayed by gas chromatography in an argon-flushed 8 ml serum vial. SO QIgde_gx§zag§_prepazatign.' The cell paste was resuspended in 30 mM Hepes buffer, pH 7.3. The cells were lysed by sonication (Heat Systems - Ultrasonic W-225) at 40 % of maximum output for a total of 5 min in intervals of 30 s, followed by cooling in an ice bath. Longer sonication time did not result in any significant additional release of nitrite reductase activity in the supernatant. v as a . Enzyme activity was measured by gas evolution (NO + N20) with NADH/phenazine methosulfate (PMS) as the electron donating system and nitrite as substrate (42). The assay contained 2 umol NADH, 0.12 umol PMS and 5 umol NaN02 in a total volume of 1 ml. All stock solutions were made in Hepes buffer (50 mM, pH 7.3). NADH and nitrite were added anaerobically to argon flushed serum bottles (8 or 25 ml) which contained the PMS and the buffer. The reaction was started by anaerobic addition of an appropriate amount of enzyme. NADH stock solution was made fresh daily. Gas formation was monitored on a Perkin Elmer 910 gas chromatograph equipped with a 63Ni electron capture detector (ECD). The carrier gas was 85 % argon, 15 % methane. The oven temperature was 55 C and contained a 1.83 m stainless steel Porapak Q column. Carrier flow was adjusted so that the approximate retention times for nitric oxide and nitrous oxide were 1 min and 2.2 min, respectively. Under these conditions, the retention times of nitric oxide and oxygen are extremely close which makes it necessary to use a strictly anaerobic technique for sampling and injecting the gas phase of assay vials. We used gas-tight syringes equipped with a gas lock. For most experiments the enzyme activity was expressed as the initial rate of gas evolution; this is the most sensitive assay available. Recovery of activity during the enzyme purification was expressed as 51 the rate of NADH oxidation, measured as the rate of decrease in absorbence at the reduced NADH absorption maximum of 340 nm. The total volume of the assay was 0.6 ml, contained in 2 mm quartz anaerobic cuvettes (Precision Glass) stoppered with rubber Venoject stoppers. The amounts of assay reagents used (above) were adjusted accordingly. Purification 9f nitrite reductase. We modified the purification method as described by Parr et al.(7) for P. aeruginosa. The most important modification was the last column; Parr et a1. (7) used carboxymethyl cellulose. Since we found the isoelectric point of the P. stutzeri enzyme to be quite low (<6), we would have to expose the enzyme to a pH value below the physiological pH range (pH 6 to 8) in order to successfully use a cation exchange column. Instead, we used a hydroxyapatite column, since it had been successfully used by Newton (13) for Paraggggus denitrifiigags cytochrome 6,61. Ammonium sulfate precipitation: RNAase (0.1 mg) and 0.1 mg DNAse (both from Sigma) were added to the crude extract to reduce viscosity. Solid ammonium sulfate was added to bring the extract to 45% saturation. The pH was adjusted to 7.5. After centrifugation (12,000 g x 30 min) to remove cell debris, the supernatant was centrifuged for 2 h at 100,000 x g to remove small membrane fragments. The supernatant was dialyzed against 10 mM Tris pH 7. First anion exchange (DEAE-I): A DEAE-SZ column (Whatman, preswollen, 2.5 x 20 cm) was equilibrated with 10 mM Tris pH 7 and loaded. A near linear gradient of 0 to 400 mM KCl was created as follows: 200 ml of 10 mM Tris pH 7 and 200 ml of 400 mM KCl in 10 mM Tris pH 7 were connected by a salt bridge. The low salt solution was stirred and pumped to the top of the 52 column at a flow rate of 3 ml/min. Fractions of 3 ml were collected. This was a fast though effective clean-up before the sizing column. The green nitrite reductase eluted before a red heme containing fraction. All green fractions were combined (up to the first slightly pink fraction), concentrated to about 6 m1 and dialyzed against 50 mM Tris pH 7. Gel Filtration: A Sephacryl column (S-300, Pharmacia, 2.5 x 75 cm) was run in the upflow mode. Half of the sample (3 ml) was loaded at one time in order to optimize resolution and prevent column overload. Fractions of 2 to 3 ml were collected. All green fractions were combined for loading onto the next column, without dialysis. Second anion exchange (DEAR-II): After loading, the column (Whatman DE-52, 2.5 x 15 cm) was washed with 50 mM Tris pH 7 for about 2 h at 3 ml/min. An olive-green band (nitrite reductase) bound tightly to the top of the column. The column was eluted with a gradient of KCl in 50 mM Tris pH 7. The gradient was from 50 to 200 mM KCl (400 ml of each). Fractions of 3 ml were collected. Nitrite reductase eluted in two clearly distinguishable peaks, an early darker green fraction and a late lighter green fraction. The valley separating the two fractions in the gradient was around 100 mM KCl. The two green fractions were collected separately (termed FR.1 and FR.2) and each dialyzed against 10 mM potassium phosphate buffer, pH 7. Hydroxyapatite: The column (Biorad, Biogel HT, 1.5 x 20 cm) was equilibrated with 10 mM potassium phosphate buffer at pH 7 and loaded at 100 ml/h. Each fraction was loaded separately and washed with 150 mM potassium 53 phosphate buffer, pH 7 for l to 1.5 h. Fraction 1 nitrite reductase was eluted with 250 mM potassium phosphate buffer, pH 7. Fraction 2 bound more tightly to the column and was eluted with 500 mM potassium phosphate buffer, pH 7. For each of the two fractions that were loaded, the eluted green fractions were combined and dialyzed against 25 mM Hepes pH 7.3. net easurements. Kinetic parameters were estimated from initial rate measurements using N20 and NO production as the assay. Approximately 1 ug enzyme was used for N20 production and 0.1 ug enzyme for NO production to stay within the linear range of the detector and of the assay for at least 20 min. The data are replicates of three replicate measurements. Native molecular weight. The native molecular weight was estimated on a Superose 6 (1 cm x 30 cm) sizing column by FPLC (Pharmacia). The flow rate was 0.4 ml per min of 25 mM Hepes pH 7.3 and 0.1 M KCl. Molecular weight standards were from Biorad. Approximately 20 ug of each standard and of nitrite reductase were loaded onto the column. Subunit molecular weight. The subunit molecular weight was estimated by SDS PAGE using an SE 300 vertical slab unit (Biorad). Gels of 0.8 mm thickness (7.5 % T, 2.7 % C acrylamide separating gel and 4 % T, 2.7 % C acrylamide stacking gel) were prepared by the method of Laemmli (43). Electrophoresis was for 10 h at 35 mA for two gels. Gels were stained by silver staining (44,45). The molecular weight was estimated by the method of Weber & Osborn (46). Molecular weight standards were obtained from Sigma. lggglggt;ig_figggging. The apparatus was the same as above. The method used was by Gary Giulan, Dept. of Physiology, University of Wisconsin, Madison, WI as described in the Hoefer Scientific manual 54 (1987). Urea (6 M) was used in the gel and in the sample buffer; this was necessary to prevent precipitation in the wells. The pH gradient in the gel was 3.5 to 10 (ampholytes Pharmalyte, Sigma). Isoelectric focusing standards were from Sigma. Exogein nnngentnation. Protein was determined by the bicinchoninic (BCA) acid method (47) with bovine serum albumin (Sigma) as the standard; the reagents were obtained from Pierce. Spectgoscopy. Visible spectra were recorded on a Perkin-Elmer (model lambda 5) spectrophotometer. Sodium dithionite was added directly to the cuvette to obtain the reduced spectra. The enzyme is normally oxidized under air. 55 RESULTS Eugification 9f nitgige :eduntnge. Purification data are given in Table 1. The overall yield was high (86%); the amount of pure enzyme recovered was 37.5 mg, which indicated that roughly 0.5% of the total protein in the cells was nitrite reductase. Fig. 1 shows the different steps in the purification on an SDS-gel. Newton (13) used a hydroxyapatite column for Rngnnnnnnn ggnigzifinnnn nitrite reductase purification and obtained three fractions of nitrite reductase, which could be eluted by step gradients. Our findings are similar but not identical: a first red heme containing fraction eluted in 150 mM potassium phosphate buffer. Its visible spectrum was identical to the one from nitrite reductase apoprotein (devoid of heme g1, Fig. 2) and thus could be the apOprotein. The second and third heme containing fractions were two forms of active, highly purified nitrite reductase; both migrated as one band on SDS gels; the second one has a higher specific activity than the first one (Table 1). These two bands also were separated on ion exchange columns. From their visible reduced spectra, however, they showed no differences in apparent heme n1 content. The total yield of purified nitrite reductase decreased proportionally to the total purification time and to the amount of time the enzyme spent in diluted form. ect a r cte t . Reduced and oxidized visible spectra of 2. ngnnngni nitrite reductase are shown in Fig. 3. They are very similar to the ones shown for fingnnnnnnnn ngnnginnnn nitrite reductase (22). In the oxidized spectrum, one peak characteristic for 56 .huo>oueu aouou cumuno cu mosesm on efisosm use N.Mh can ~.mm you xfiouoummom poundsono euo3 meandoo ouuuoaohxouoa; ecu HH-mooem ”a HNH «mu em mo.- «.5 N .mm «m 4mm mmw mm.n m.Om H .mm euauoamxxouexz mm . awe new mo.m mm N .zm Cm #mo mew mm.~ mm H .mm HH-mooom mugs: >uu>duo< >Da>fiuo< eumfiooam :Hououm DCoEuaeey .ommuoseou euauuu: «wouusum mmcosoesomm .oEocom :oHuooamausm H mqm£emum0uoEou£o c0wmsaoxw Hem xn ommuospou ouwuuec “neuusum mmEOEOCSomm mo unmfio3 HoasooHOE o>wuoc map No :oHumcwEumuen .v .mflm MEDJO> ZOFZNPME m_>_.r<1_mm PF 0.. m w h w _ _ _ _ _ . NTm Z__>_<._._> Z_mO.._OO>_2 mmmOI 2.55m..<>0 Zm¥U_IO mm0m m.m (Mw) 901 64 meme 1 wow >3 ommuospmn ouwuufi: Hummusum mocOEOpsomm mo unmwws hoasooHOE Decansm ecu mo cofiumcflEhwuon .m .mfim ..mO 20 DMP<¢0=2 m02_.r<1_mm my VF or o N _ _ 1 _ _ _ _ _ _ _ 2.55m.._< 00m mmmOImmOIm mm<0_m0._.0<1_<0 m Z_mO>_>_ mwIZ< 0.20mm<0 l 0 2.55334 mz_>om_ ., m6, (MW) 501 m.m 65 Fig. 6. Determination of the isoelectric point of P. stutzeri nitrite reductase by isoelectric focusing. pl 9.5 8.5 7.5 6.5 5.5 4.5 3.5 66 1- LACTIC DEHYDROGENASE CARBONIC AN HYDRASE (HUMAN) CARBONIC AN HYDRASE NITRITE (BOVINE) ' REDUCTASE B LACTOGLOBULIN AMINOGLUCOSIDASE l l 1 l l L 1 l l 1 2 3 4 5 6 7 8 9 DISTANCE FROM ANODE (CM) 67 Fig. 7. Progress curves of nitric oxide and nitrous oxide production from nitrite by Eseudomonas stutzeni nitrite reductase. A: High initial nitrite concentration (1 mM) B: Low initial nitrite concentration (0.05 mM) 68 , [NO§]=1.0 NM 3000 -1 2400 ~ q 1800- 2 -° - E 1200- Q d 600 N20 0 I I l l l 0 40 80 120 160 200 time(min) 20005 [N021 = 0.05 mM 1600- Z —o 1200 E a. 8001 400- IFTIIr|II 020 40 60 80100 time (min) 69 The kinetic parameters for the two reactions were determined by measuring initial rates of NO production at low nitrite concentrations (1 to 25 uM), the second one by measuring initial rates of N20 production at higher nitrite concentrations (50 to 1000 uM). In this experiment the bottles were more exhaustively flushed to reduce the lag. Initial rates were calculated from the first three time points which were linear (approx. 3 to 20 min) The two Km values were different enough that measurement of one did not interfere with measurement of the other. In the nitrite concentration range where NO production rates were increasing, N20 production was still so slow that it did not interfere with the initial rate measurements. In the range where N20 production became apparent, NO production was at its Vmax. For NO production, the apparent Km was 1.35 uM N02‘ with an apparent Vmax of 1.6 umole N02'/mg.min (Fig. 8). For N20 production, the apparent Km was of 59 uM N02- with an apparent Vmax of 0.93 umole N02'/mg.min (Fig. 9). We compared these kinetic parameters with the ones for crude but ultracentrifuged extract, in order to determine whether the purification treatments between ultracentrifugation and pure enzyme from the hydroxyapatite column had any deleterious effect on the enzyme's kinetic properties. In the crude extract, the first Km (for NO production) was essentially the same as for the pure enzyme: 1.10 uM N02- (data not shown). The second Km (for N20 production, Fig. 10) was higher for the purified enzyme: 59 uM vs. 34 uM for the crude enzyme. While the factor of two difference is statistically significant, it is not clear whether this difference has any importance to understanding the behaviour of this enzyme. 7O 6N .omwposemh opwhpfic Heoupzpm mmcoeoczomm hp sewposooem mowxo veep“: mo meopoeoemm owpocwx .m .mE .23 moz. «a m. 3 o. e a _ _ . _ m _ _ _ _ _ EE x oE>Nco as." on> .. oz 6.65: o; . .21 mm; H .5. 1 SI- 00 CO V N (1-1“ - tutu - 5w) A/[EON] (D F 71 .ommposeoa opwepwc wemuuSpm mwcosocsmmm an coHposeoem mcwxo msoepflc mo mempoeoeme owuonx .m .mwh =23 Roz. COOP com com GOV CON q q q — _ _ _ _ q _ _ :_E x 053.5 9: 2-042 6.651 8.0 n 35> 1 5.3.8 u .5. 00? com coup -6w) A/[EON] Oll!lll ( L_11I 72 Fig. 10. Kinetic parameters of nitrous oxide production by crude, ultracentrifuged extract of 2. stutzegi cells, containing nitrite reductase. 73 l 1 l l 1 O C C O O O G (D V N F suun MMHOV ameuaa Eon] 120 " (burugw ~6w) 1 00 1 50 200 250 1 (MM) 50 2 [NO 74 Control experiments with autoclaved enzyme showed no chemical production of NO or N20 from nitrite, and no chemical N20 production from N0 (at NO concentrations representative for those seen in a typical progress curve). Also, the NADH/PMS reducing system did not sustain any significant chemical N20 production from mixtures of NO and N02' under our experimental conditions. Nitgig oxide as substzage. Neither the purified enzyme nor the crude ultracentrifuged extracts consumed nitric oxide when it was the only substrate. With the NADH/PMS reducing system, when the purified enzyme was given 1 mM 15N02' and 100 Pa NO in the headspace, the only product formed was 15'ISNZO: apparently, free NO is not reduced under these conditions, since neither 15’IZ‘NZO nor 14,14N20 were detected. The products were analyzed by mass spectrometry as described previousky (19). In earlier work with whole cells (19) and with non- ultracentrifuged crude extracts we did find nitric oxide reduction to nitrous oxide. Washed membrane fragments which were virtually devoid of all nitrite reducing activity still had much NO reducing activity (unpublished data). The NO reducing activity was apparently associated with the membrane fraction removed during the ultracentrifugation. 75 DISCUSSION Over the past few years we have obtained indirect evidence, from isotope studies in whole cells (19) and crude cell-free (but not ultracentrifuged) extracts (20; Ch. 3, this work) of P. stutzeri JM 300, for a pathway of nitrite reduction in this organism in which the first dinitrogen bond is made by nucleophilic addition of a nitrite anion to an enzyme bound nitrosyl intermediate. Earlier we investigated the effect of hydroxylamine on denitrification by whole cells with 15NOZ', resp. 15NO as substrates (19). With NO as a substrate, at hydroxylamine concentrations of 10 mM, all of the N20 formed was from nitrosation of hydroxylamine (Table 2), indicating effective interception of a mononitrogen intermediate in the NO reduction pathway by hydroxylamine. When ISNOZ' was a substrate, however, hydroxylamine at a concentration of 320 mM did not result in any significant formation of nitrosation product (14’15N20; Table 3), which indicates that hydroxylamine was too weak a nucleophile to intercept any mononitrogen intermediates in the nitrite reduction pathway in the whole cell. We believe this to mean that the main flux of nitrite nitrogen in the whole cell does not involve a free NO intermediate. In crude extracts, possible interference by an often postulated, but so far never purified, "nitric oxide reductase" may be a problem, therefore we purified the nitrite reductase and studied its kinetic properties. In a pathway in which there are two binding events, it should be possible to measure apparent Km values for each of the two nitrite binding events. We determined a Km of 1.4 uM for NO production 76 TABLE 2 Competition between nitrosation and denitrification when 15NO was substrate. NHZOH (mM) N20 from nitrosation (%) 0 0 5 38 10 100 25 100 TABLE 3 Absence of competition between nitrosation of hydroxylamine and denitrification when ISNOZ' was substrate. NHZOH (mM) N20 from nitrosation (%) 10 2.4 40 1.4 80 1.6 160 2.6 320 2.8 78 (first nitrite binding) and a Km of 59 uM for N20 production (second nitrite binding). Biphasic Lineweaver-Burk plots for Pseudononas neguginosn nitrite reductase have been presented by Saraste and Kuronen (29). These authors stated that: " The biphasic kinetics observed ... cannot be easily interpreted at present. Possible explanations are the presence of two isoenzymes with different affinities to nitrite,or an equilibrium between dimeric and monomeric forms, or interaction of the two nitrite binding sites in the dimeric enzyme ". We estimated two Km (NOZ') values from their biphasic plot: a low Km of approximately 6 uM, and a high Km of approximately 68 uM for the purified enzyme, values comparable to ours for P. stutzeri nitrite reductase. A Km for the second nitrite binding of this order of magnitude (around 60 uM) is low enough for this reaction to be potentially significant in nature. However, nitric oxide accumulates to a larger extent in the headspace above the pure enzyme or crude extracts (several 100 ppm) than it does above whole cells (10 ppm, for Elavobacterium; 51) or in soils (1 to 100 ppb). This could be attributed to physical factors: nitric oxide released inside the cell may quickly become bound again by an unknown component. Rupturing the cells may set it free instead (52). This is one possibility. The "component" may be a ”nitric oxide reductase". We find, however, there is also reason to believe that the membrane bound nitrite reductase "in vivo" may show a different kinetic behaviour than the purified enzyme. Saraste (53) found a 20-fold stimulation of the specific activity of the Esgnnnnnnnn nggnginnnn nitrite reductase cytochrome oxidase activity when the enzyme was incorporated into artificial 79 membrane vesicles. Unfortunately the nitrite reductase activity was not investigated. For solubilized but not purified Eseudonnnag stutzezi nitrite reductase we found a low Km for NO production of 1.10 uM N02°, close to the value for the pure enzyme, and a high Km for N20 production of 34 uM NOZ', about half of the value in the pure enzyme. A factor of about two was also found by Saraste and Kuronen (29) between the high Km values for nitrite in whole cells vs. pure enzyme. It seems that the Km for N20 production may be affected by solubilization and/or purification. At this time it is not possible to conclude whether, in the whole cell, the main nitric oxide sink is nitrite reductase or a putative nitric oxide reductase. In the membrane, nitrite reductase accepts electrons from its physiological electron donor - a process which may be many times faster than reduction of the solubilized enzyme with natural or physiological electron donors: typically, maximum rates of N20 production are two orders of magnitude slower in extracts than in whole cells (20). This is most probably due to a more efficient electron transfer process to membrane bound nitrite reductase as a terminal component in the respiratory chain. It is feasible that, in whole cells, nitrite reduction maximum rates are fast enough to keep NO accumulation low. Membrane associated nitric oxide reductases may may be hard to purify - several lines of evidence indicate that they certainly may exist: detection of membrane-bound enzymatic nitric oxide reducing capacity (26), separation of nitric oxide and nitrite reducing activities in different cellular fractions (30,32,35), purification of NO reductase in progress (54); we found differential extraction of nitrite reductase activity and nitric oxide reductase activity 80 possible in our own laboratory: virtually all of the nitrite reductase activity can be removed from membranes by repeated washes with buffers, but NO reductase activity remains membrane bound. Until a pure nitric oxide becomes available and the "in vivo" kinetics for both NO and nitrite reductases known, it will not be possible to say whether nitric oxide reductase or nitrite reductase is the main NO consumer in nature. Existence of nitrix oxide reducase (other than nitrite reductase) would not necessarily imply a role for it in denitrification. It may be an "NO scavenging" activity with a detoxifying function. The nitrite reductase of Eseudomnnns §§utze§i is apparently very "sticky": nitrite is virtually not released from the enzyme, once bound. Absence of isotopic exchange (ISN) between NO and nitrite (39) with whole cells, a very low isotopic exchange rate between a free unlabelled nitrite pool and 18O labelled water with whole cells (55) and the virtually zero rate of NO turnover by the pure enzyme (this work) all substantiate that statement. This explains why NO turnover by this enzyme can occur when nitrite is present (isotopic scrambling of 15N02' and 14NO nitrogen in product N20 in absence of isotopic exchange between the NO and N02' pools; 39) but not when nitrite is absent - the sticky enzyme does not release nitrite which is needed to drive the reaction forward according to the pathway (20). For probably the same reason, we have never seen stoichiometric turnover of nitrite to nitrous oxide with the pure enzyme. For Eggngnnnnns nggnginngn PAOl, the rate of N20 production with NO as only substrate was reported to be virtually zero (21). For the same organism (strain not reported; 27), measurable rates of NO reduction to N20 and stoichiometric conversion of nitrite to N20. The degree of isotopic 81 scrambling between the NO and N02' pools was very different for various denitrifiers (38,39). The pathway of dissimilatory nitrite reduction in denitrifiers is, at this point, quite well illustrated, at least in P. stutzggi JM300: evidence from previous isotope studies (19,20) taken together with the kinetic characterization of the enzyme in this work support the mechanism first proposed by Averill and Tiedje (18). It is likely that this pathway common to all nngl type nitrite reductases but that, for different denitrifier species and maybe even for different strains within the same species, the rate constants for individual reactions in the pathway may differ. 10. 11. 12. 13. 14. 15. 16. 17. 18. REFERENCES Miyata, M., and Mori, T. (1969) 1, 9f Biogngn, (10519) 66, 463-471 Horio T., Higashi, T., Yamanaka, T., Matsubara H., and Okunuki, K. (1961) J, Biol, Qnen, 236, 944-951 Walker, G. C., and Nicholas, D. J. D. (1961) Biochim. Biophys. Annn 49, 350-360 Kuronen, T., and Ellfolk, N. (1972) Biochin, Biophys. Acta, 275, 308-318 Matsubara, T., and Iwasaki, H. (1972) J, Biochgn, (Inkyo) 72, 57-64 Gudat J. G., Singh, J., and Wharton, D. C. (1973) Biochim. 3192111183451; 292 . 375-390 Parr, S. R. , Barber, D., Greenwood, C., Phillips, B. W., and Melling, J. (1976) Biochem. J, 157, 423-430 Silvestrini, M. C., Citro, G., Colosimo, A., Chersi, A., Zito, R., and Brunori, M. (1983) Anal, Biochem. 129, 318-325 Sawhney, V., and Nicholas, D. J. D. (1978) J, Gen, Migroniol, 106, 119-128 LeGall J., Payne, W. J., Morgan, T. V., and DerVartanian, D. (1979) Biochem. Biophys. Res, Commun, 87, 355-362 Radcliffe, B. C., and Nicholas, D. J. D. (1968) Bionhim. , Bionhys, Acnn 153, 545-554 Iwasaki, H., and Matsubara, T. (1971) J, Bingngn, (Tokyo) 69, 847-857 Newton, N. (1969) Binghim, Bionnys, ngta, 185, 316-331 Lam, Y., and Nicholas, D. J. D. (1969) fiioghim, Biophys, Acta, 180, 459-472 Timkovich, R., Dhesi, R., Martinkus, K. J., Robinson, M. K., and Rea, T. M. (1982) A190. Biochgn, Biophys. 215, 47-58 Mancinelli, R. L., Cronin, S., and Hochstein, L. I. (1986) Arch, Micgonioi, 145, 202-208 Henry, Y.,and Bessieres, P. (1984) Bioghimie, 66, 259-289 Averill B. A. and Tiedje, J. M. (1982) FEBS ngg, 138, 8-12 82 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 83 Aerssens, E., Tiedje, J. M., and Averill, B. A. (1986) J, Bigi, gnnn‘ 261, 9652-9656 Weeg-Aerssens, E., Tiedje, J. M., and Averill, B. A. (1987) J, An, Qngm, §on, (submitted). Kim, C. H., and Hollocher, T. C. (1984) J, Biol, Chen, 259, 2092-2099 Yamanaka, T., Ota, A., and Okonuki, K. (1961) Biochin, Biophys. Acta, 53, 294-308 Cox, C. D. and Payne, W. J. (1973) Can, J, Miggobioi, 19, 861-872 Zumft, W. G., and Vega, J. M. (1979) flinchim, Bignnys,nctn, 548, 484-499 Zumft, W. G., Sherr, B. F., and Payne, W. J. (1979). Biocnem, Biophys, Res, Connnn, 88, 1230-1236 Zumft, W. G., and Frunzke, K. (1982) Biochin. Bionhys,Acta, 681, 459-468 Wharton, D. C. and Weintraub, S. T. (1980) Bignnen,fiionhy§, Res, gonnun, 97, 236-242 Barbaree, J. M. and Payne, W. J. (1967) Mn;ing_§iningyi 1, 136-139 Saraste, M., and Kuronen, T. (1978) filgshimI_5122hI§I_A££éI.513. 117-131 Grant M. A., Cronin, S. E., and Hochstein, L. I. (1984) Arch. Microbini, 140, 183-186 Garber E. A. E., Castignetti, D., and Hollocher, T. C. (1982) Binghen, Biophys, Res, Connun, 107, 1504-1507 Shapleigh, J. P., Davies, K. J. P., and Payne, W. J. (1987) Biochim, Bionhyg, Anta, 911, 334-340 Fewson, C. A. and Nicholas, D. J. D. (1960) Naturg, 188, 794-796 Chung, C. W. and Najjar, V. A. (1956) J, Biol, gnen, 218, 627-632 Payne, W. J., Riley, P. S., and Cox, C. D. Jr. (1971) in Bactegioi, 106, 356-361 ‘ St. John, R. T., and Hollocher, T. C. (1977) J, Bioi, Chem, 252, 212-218 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 84 Renner, E. D., and Becker, G. E. (1970) Jl_Bnn§n;inli 101, 821-826 Firestone M. K., Firestone, R. B. and Tiedje, J. M. (1979) Bincnem, Bionhys, figs, gnnnnn, 9, 10-16 Garber, E. A. E. and Hollocher, T. C. (1981) 4, Biol, Chen, 256, 5459-5465 Garber E. A. E. and Hollocher, T. C. (1982) J, Biol, Chem, 257, 4705-4708 Kodama, T. (1970) Elnnt and Cell Bhyginl, 11, 231-239 Orii, Y., and Shimada, H. (1978) J, Bingnen, (19510) 84, 1543-1552 Laemmli, U. K. (1970) Nature (London). 277, 680-685 Wray, V., Baulikas, T., Wray, V., and Hancock, R. (1981) Anal, Binnhem, 118, 197-203 Eschenbruch, M. and Burk, R. R. (1982) Ann1i_Binnnnnn 125, 96-99 Weber, K., Pringle, J. R., and Osborn, M. (1969) Methods in Enzynol, XXVI, 3-27 Smith, P. K., Krohn, R. I., Hermanson, G. T., Mallia, A. K., Gartner, F. H., Provenzano, M. D., Fujimoto, E. K., Goeke, N. M., Olson, B. J., and Klenk, D. C. (1985) Anal, Biochem, 150, 76-85 Barrett, J. (1956) Biochem, J, 64, 626-639 Walsh, T. A., Johnson, M. K., Barber, D., Thomson, A. J., and Greenwood, C. (1980) J, Innzg, Biochem, 14, 15-31 Hill, K. E., and Wharton, D. C. (1978) J, Biol, Chen, 253, 489-495 Betlach, M. R. and Tiedje, J. M. (1981) Appl, an, flicnobiol, 42, 1074-1084 Payne, W. J., and Riley, P. S. (1969) Proc, Boc, Ban, Biol, Med. 132, 258-260 Saraste, M. (1978) Bionhin, Bionhyg, Antn 507, 17-25 Hoglen, J., and Hollocher, T. C. (1986) Red, 239;, 45, 1604A Shearer, 0., and Kohl, 0. (1987) J, Biol, Chen, (submitted). CHAPTER FIVE RECONSTITUTION OF THE DISSIMILATORY CYTOCHROME gfil TYPE NITRITE REDUCTASE FROM figfinngnnnns EIUIZBEI WITH THE SYNTHETIC HEME 21 INTRODUCTION Dissimilatory cytochrome 991 type nitrite reductases from denitrifying bacteria are typically dimers of 120,000 to 130,000 Da molecular weight. The subunits are structurally identical and contain one heme c and one heme n1 each. Heme c is covalently bound to the enzyme by means of two thioether linkages and noncovalently through coordination of its iron, presumably by two protein ligands (l). Heme n1 is not covalently bound and is linked to the enzyme by coordination of its iron to one ligand, probably an imidazole (1). In vivo, nitrite reductase is believed to be membrane associated: its physiological function is as a terminal oxidase (i.e. nitrite reductase) in the respiratory chain under anaerobic, denitrifying conditions. Its physiological electron donors are cytochrome c 551 and/or azurin in the case of Eseudomonas neruginosa (ferrocytochrome c-551zoxido- reductase, EC 1.9.3.2) (l). The enzyme from Ennnnnnnnnn nnnnngni may be a monomer (Ch. 4, this work). These dissimilatory nitrite reductases reduce nitrite to nitric oxide and nitrous oxide (Ch. 4, this work). Heme n1, thought to be the nitrite binding site (1), is easily removed from the enzyme by acidified acetone extraction (2-5). It was reported that 90 to 100% of 5S 86 the original activity was recovered when the apoenzyme from E. nggnginngn was reconstituted with its own native heme BI and that 5% of the activity could be recovered when it was reconstituted.with heme a from beef heart cytochrome oxidase. Activity assays, in this case, were done spectrophotometrically by the rate of reoxidation of cytochrome c-551 with 02 as oxidant (5). The structure of heme n1 was first suggested to be a chlorin type structure with two carboxylic acid groups on an unsaturated pyrrole and two methyl alcohol groups on a saturated pyrrole (6,a,b). More recent evidence shows the chromophore to be a dioxo-isobacteriochlorin with two ketone groups on adjacent pyrrole rings (Fig. l) (7,a,b). So far, this novel structure has been found only in bacterial nitrite reductases. The heme n1 has recently been synthesized (8). In this chapter we report on the biochemical activity and spectral properties of nitrite reductase apoprotein reconstituted with the synthetic heme n1. I acknowledge the collaboration of Weishih Wu who worked jointly with me in the experiments reported in this chapter. 87 mmmcwxo oeoucoou>o Hmfiuwuocn C N N m .Allli. C U mzmm oao: @ 050: .mawxccoflt .meosouflm m0 ousuosuum one m o z a :ofiuoswou oufluuec >uoucae5emmfle Ides mzmzmona £00 0.0: 2.00 o oao: .H .mae mmz All NOZ cofluostwu wufluuflc auoumaflEwmmm WEWZOmHm I~oo 88 MATERIALS AND METHODS Ih§_n;gnni§m. The nitrite reductase was purified from figgnnnnnnns finnnnggi strain JM300. The purification and some properties of this enzyme are described (Ch. 4, this work). Nitrite reductase had a spectroscopic absorbance ratio (on4ll/A280) of 1.1 to 1.2, as do purified preparations of Eseudonnnas ngnnginnnn nitrite reductase (9). The enpgactinn ann reconstitntion niggedung. The procedure used was a modification of the procedure of Hill and Wharton (5). A11 enzyme manipulations were carried out at 4 C. Typically, 2 mg purified nitrite reductase (on4lO/A280 > 1.1) in 1 ml of buffer (25 mM Hepes, pH 7.3) was added to 3 mg bovine serum albumin (Sigma); the latter was used as a "carrier” protein to aid in the precipitation. Cold acidified (0.024 N HCl) acetone (8 ml) was added; the protein precipitated and the heme n1 was extracted into the overlying acetone solution. The protein was separated from the acetone by centrifugation and extracted once more to ensure complete removal of heme n1. The protein pellet was washed once with phosphate buffer (0.25 M, pH 7) and redissolved in phosphate buffer with urea (0.25 M phosphate buffer pH 7, 6M urea). An excess amount of (10:1 molar ratio) heme n1 (native or synthetic) was added to the apoprotein solution and the mixture was incubated with gentle stirring for 30 min. The apoprotein + heme BI solution was dialyzed overnight against 50 mM Tris Buffer pH 7. The reconstituted enzyme was separated from the excess heme n1 by passing the solution over a short DEAE (diethyl aminoethyl cellulose) column (DE-23, Whatman, 0.5 x 5 cm): heme n1, which has a pKa of 4.5, bound 89 to the top of the column; the enzyme was eluted with 100 mM phosphate buffer, pH 7. Anniyi;y_n§§ny. Activity was measured by gas evolution (NO and N20) from nitrite with NADH/phenazine methosulfate as the electron donor system. The assay contained 6 umol NADH, 0.36 umol PMS and 3 umol NaN02 in a total volume of 3 ml. All stock solutions were made in Hepes buffer (50 mM, pH 7.3). An appropriate amount of enzyme solution was added, usually containing about 1 ug of enzyme. NADH and nitrite stock solutions were made oxygen free by repeated evacuating and filling with argon and were added anaerobically to a 25 ml serum bottle which contained the buffer and PMS and had been flushed with argon. The reaction was initiated by addition of enzyme. The nitrite concentration (0.5 mM) was saturating for both NO and N20 production by the original enzyme (Ch. 4, this work). Enzyme activity was expressed as the initial rate of gas evolution (NO or N20). Gas evolution was monitored on a Perkin Elmer 910 gas chromatograph equipped with a 13 Ni electron capture detector. The carrier gas was 85 % argon, 15 % methane. The column was a 6 foot stainless steel Porapak Q column operated at 55 C. Carrier flow rate was adjusted so that the approximate retention times for NO and N20 were 1 min and 2.2 min, respectively. Under these conditions the retention times of nitric oxide and oxygen were extremely close and it was necessary to use strict anaerobic techniques for sampling and injecting the gas phase of the assay vials. We used gas-tight syringes equipped with a gas lock. Visible spectra were recorded on a Perkin Elmer (model lambda 5) spectrophotometer. 9O .Eignngntinn_nf_§hg_nnng_g1. The native heme g1 was extracted from the enzyme by the procedure described above. The acetone solution which contained the heme was evaporated to near dryness in the dark at room temperature under a stream of argon. The residue was dissolved in phosphate buffer (0.25 M, pH 7) and centrifuged to remove any remaining protein precipitate. The solution of heme n1 was adjusted to pH 7 with NaOH and stored in the dark under argon. Extraction of native heme n1 by the method of Hill & Wharton (5) was tried but did not work in our hands: addition of 1 ml of 1.2 N NaOH to the 8 ml acidified acetone extract under air resulted in rapid decoloration (decomposition) of the heme n1 concentrated in the NaOH layer. Heme 91 is very labile in acetone solution (6,a) therefore it was important to minimize the time it spent in the acidified acetone. The synthetic heme n1 was prepared as described by Wu & Chang (8). Heme g, was prepared by insertion of iron into the free base tetramethylester of the tetrapyrrole (10). The hydrolysis of the methyl esters was done as follows: iron tetramethylester is dissolved in 40 m1 tetrahydrofuran (THF) followed by addition of 2 ml of l N KOH. The solution was stirred in the dark under nitrogen gas for 12 hours until the organic layer was almost colorless. The THF was evaporated and the pH of the basic aqueous solution (in ice bath) was adjusted to 7 with concentrated HCl. The enzyme concentration was estimated by the bicinchoninic acid method (11). For estimation of relative protein concentrations, the absorbance at 411 nm (oxidized enzyme) was used. The concentration of heme n1 was estimated by using the published extinction coefficient of 32,100 M'1.cm'1 for oxidized imidazole-ferriheme g, (4). 91 RESULTS Bngnnznl_nnn1nnggzi§n§inn. The visible (oxidized and reduced) spectra of Esgunomonng snnnzngi JM300 nitrite reductase are shown in Fig. 2. The spectral characteristics are very similar to the ones of Eseudononas angnginngn nitrite reductase (12). The Soret band at 417 nm in the reduced spectrum is attributed to to heme c and so are the split alpha band at 522 nm and the beta band at 555 nm. The heme n1 is responsible for the weak shoulder (Soret) at 460 nm and the broad absorbance in the 644 nm area. The visible spectrum of the apoprotein (Fig. 3) lost the spectral characteristics typical of heme n1. The spectra of the enzyme reconstituted with the synthetic and the native heme n1 are shown in Figs. 4 and 5, respectively. They regained the features typical of heme n1; the 460 nm shoulder is less prominent for spectra taken at lower protein concentrations such as the one in Fig. 5. The visible spectrum of the heme n1 extracted from 2. stutzeri nitrite reductase is shown in Fig. 6. It is identical to the visible spectrum of heme g1 from 2. nerugingsa (12). gecoveny of activity. The NO and N20 producing activities for both the apoprotein and the reconstituted enzyme preparations are summarized in Table 1. The apoprotein had no activity but when reconstituted with the native and synthetic heme n1 substantial activity was recovered. The data were corrected for loss of protein in the course of the reconstitution. Long term progress curves of nitrite reductase, apoprotein and apoprotein reconstituted with synthetic heme n1 are shown in Fig. 7. Apparently, N20 producing activity was never entirely recovered, whether the native or the synthetic heme g1 was used. The N20 producing activity of this enzyme was a more labile Q2 The absorption spectrum of nitrite reductase from Pseudomonas stutneri. The enzyme was dissolved in 0.25 M phosphate buffer pH 7.3; the enzyme concen- tration was 2 mg/ml. ___ oxidized, reduced with Na23204 I 93 cucv< Dov com /\ .30 :M: vm .o mo.d Norm 94 Fig. 3. The absorption spectrum of the apoprotein (-heme d1) of the nitrite reductase from Pseudomonas stutzeri 95 600 700 500 400 300 96 The absorption spectrum of nitrite reductase from Pseudomonas stutzeri reconstituted with synthetic heme d1. The enzyme was dissolved in 0.25 M phosphate buffer, pH 7.3 ————— , oxidized; reduced Wlth Na28204 I 97 Fig. 5. The absorption spectrum of nitrite reductase from Pseudomonas stutzeri reconstituted with native heme g . The enzyme was dissolved in 0.25 M phosphate buffer, pH 7.3. -—— — , oxidized; ——-— reduced with Na25204 99 Dow 00: OWN Till _ _ _ _ .4 / // Ill! £6 / I / mmm // mmm // 094/ // . / L 2.0 / / \.I \ \ I / \ , , \ , V , \ f m: , \ . I a f . . I I ... om o I I . . a . . . K ’ \le . I . x a a . . x _ \HH: — x r /r\ ‘ 3m . 1 3.0 S: r - .omd 100 Fig. 6. A: The absorption spectrum of heme d from Pseudomonas aerucinosa. The heme d was dissolved in acetone containing 0.12 N HCl. Yamanaka, T. and K. Okonuki. Biochim. Biophys. Acta. 67 (1963), p. 497. Fig. 6. B: The absorption spectrum of heme n1 from Pseudomonas stutzeri. The heme n was dissolved in acetone containing 0.24 N HCl. lOl Abs 400 500 600 700 (mil) 0 8 7 5 5 B 1 . I Bstutzeri home 51,. m " r 1 ’ .o / ( < ‘P/ L 4» II . \\ . Ud— 0.1,_ . . 0. , - - - 400 500 600 700 (nm) Mnm) 102 254-04 251-04 1 E+04 8000 4000 2E+04 1 E+O4 8000 pmole N 4000 2 E+04 1 E+O4 8000 4000 0 O IUIIIfiII Tl'l'l'l N20 ‘ N20 6 lllnllll IIIII [VII 60 120 180240 0 60 120180 240 300 Time(min) Fig. 7. Progress curves of nitric oxide and nitrous oxide production from 1 mM nitrite by A: B: C: 2, stutzeri nitrite reductase apoprotein (- heme d ) nitrite reductase, reconstituted with synthetic heme d1 103 activity than the NO producing activity. Possibly the latter was less affected by partial denaturation of the enzyme. 104 TABLE 1 Recovery of activity of nitrite reductase after reconstitution of the apoprotein with native and synthetic heme n1. Treatment of Enzymatic activity measured nitrite reductase - b . b N02 to NO N02 to N20 Intact,original enzyme 100 100 Apoproteina O 0 Reconstituted, native n1 50.9(6.5) 37.9(2.2) Reconstituted, synthetic n 81.1(4.4) 40.1(3.8) 1 a: Apoprotein remained soluble after dialysis to remove the urea and had no detectable activity during the time course of an initial rate experiment. Numbers between parentheses represent the deviation from the mean value of two initial rate determinations within the same experiment. b: Activities are expressed in relative units. The activity of the intact, original enzyme represents 100 %. 105 DISCUSSION The recovery of gas producing activity after reconstitution was substantial for both the synthetic and the native heme n1, but lower than the 90 to 100% recovery which was reported for the terminal oxidase activity after reconstitution of Bseudomonas aeruginnsn nitrite reductase with its own heme n1 (6). In our case, however, activity was determined by measuring the products of the physiological reaction of the enzyme. The NO/NZO product ratio of the enzyme reconstituted with synthetic heme n1 was somewhat lower than for the enzyme reconstituted with native heme g1; the latter ratio is more like the intact nitrite reductase. The synthetic heme a, solution we used was a mixture of two enantiomers. This may account for the difference in NO/NZO ratio if there was no distinction by the enzyme for the correct stereoisomer in the reassembly. As a consequence of the reconstitution procedure, whether synthetic or native heme n1 was used, the N20 producing activity was less effectively restored. We suspect this may be due to permanent denaturing effects caused by the precipitation - resolubilization, affecting the second nitrite binding to the enzyme. The second nitrite binding seems to generally be more sensitive to in vitro treatments (Ch. 4, this work). 10. 11. 12. REFERENCES . Henry, Y., and Bessieres, P. (1984) Binnhinig. 66, 259-289 . Barrett, J. (1956) Bigghem, J, 64, 626-639 . Horio, T., Higashi, T., Yamanaka, T., and Okonuki, K. (1961) J, Biol, gnen, 236, 944-951 . Walsh, T. A., Johnson, M. K., Barber, D., Thomson, A. J., and Greenwood, C. (1980) J, inogg, Biochem, 14, 15-31 . Hill, K. E., and Wharton, D. C. (1978) J, Biol, Chem, 253, 489-495 (a) Timkovich, R., Cork, M. S., and Taylor, P. V. (1984) J, Biol. gnem, 259, 1577-1585 (b) Timkovich, R., Cork, M. S., and Taylor, P. (1984) J, Biol, Chem, 259, 15089 - 15093 (a) Chang, C. K. (1985) J, Binl, gnen, 260, 9520-9522 (b) Chang, 0. K., and Wu, W. (1986) J, Biol, Chgn, 261, 8593-8596 . Wu, W., and Chang, C. K. (1987) J, Am, gnen, Bog, 109, 3149-3150 . Silvestrini, M. C., Colosimo, A., Brunori, M., Walsh, T. A., Barber, D., and Greenwood, C. (1979) Binnngnnni‘ 183, 701-709 Chang, C. K. (1980) IBQIEI_§IB£I 20, 147 Smith, P. K., Krohn, R. I., Hermanson, G. T., Mallia, A. K., Gartner, F. H., Provenzano, M. D., Fujimoto, E. K., Goeke, N. M., Olson, B. J., and Klenk, D. C. (1985) Annln_Binnngni 150, 76-85 Yamanaka, T., and Okonuki, K. (1963) Biochin, Bignnys, Acta. 67, 394-406 106 APPENDIX APPENDIX ERRATUM IN CHAPTER TWO Equation (1) in chapter two should be: instead of k3.k4/(k4 + (ks-k7-[N02-]/(k5 + k7))) Bow to derive the formula. Using the "analysis of partitioning" for calculating net rate constants for transfer of label, by Cleland, W. W. (1) First we need the net forward rate constants for the involved steps in the pathway; these are the rate constants which would produce the same flux through the step if this step were irreversible: N02' 1102' + E —z- E.N02' —a- 13.190+ A» E.N203 —’ N20 1' 3' 5' 7' Net rate constants: k7.- k7 k5.- (k5.k7.[N02'])/(k6 + k7) k3'- (k3.k5,)/(k2 + k3.k5.) We calculate the initial rate of transfer of label to product (in our case of 18O to N20). This is the rate of addition of the labelled 107 108 compound, multiplied by the net forward partitioning of the thus formed labelled enzyme bound intermediate. In other terms: k4.[E.N0+] x (k3.k5.)/(k2 + k3.k5,) (1) The rate of N20 production can be expressed as the product of any of the enzyme bound intermediates multiplied by its net forward rate constant; if we do this for E.NO+ we get: ([E.NO+].k5.k7.[N02’])/(k6 + k7) (2) The initial 180 content of N20 is expression (1) divided by expression (2); after some reorganizing we get: k3.k4/(k2 + (k3.k5.k7.[N02']/(k6 + k7))) REFERENCES l. Cleland, W. W. (1975) Biochemisnny 14, 3220-3224 "I11111111141111