JOURNAL OF BACTERIOLOGY, Dec. 1985, p. 1064-1070 Vol. 164, No. 3 0021-9193/85/121064-07$02.00/0 Copyright © 1985, American Society for Microbiology Isolation and Characterization of Paracoccus denitrificans Mutants with Defects in the Metabolism of One-Carbon Compounds NELLIE HARMS,'* GERT E. DE VRIES,2 KICK MAURER,' EDUARD VELTKAMP,2 AND ADRIAAN H. STOUTHAMER' Department of Microbiology' and Department of Genetics,2 Biological Laboratory, Free University, 1007 MC Amsterdam, The Netherlands Received 11 June 1985/Accepted 12 September 1985 Mutants deficient in the metabolism of one-carbon compounds have been obtained by treating Paracoccus denitrificans with the mutagen N-methyl-N'-nitro-N-nitrosoguanidine. Mutants were selected without enrich- ment procedures by newly developed plate screening tests. The obtained mutants were characterized by their growth responses, cytochrome composition, ehizyme activities, and immunogenic reaction with antisera against dehydrogenase. By these criteria five mttant classes could be distinguished. Class I mutants are involved in the expression of methaniol dehydrogenase. Three mutants of this class have a defect in the structural gene. A double mutant was found with defects in the expression of both methanol dehydrogenase and hydrogenase. Class II mutants have a defect in a regulatory gene involved in the regulation of both methanol dehydrogenase and methylamine dehydrogenase. Class III mutants are deficient in metabolism. A defect may exist in the expression of a second non-NAD-linked formaldehyde dehydrogenase which was postulated to be involved in C1 metabolism. Class IV mutants are deficient in cytochrome c. Mutants of class V have a defect in synthesis of the molybdenum essential for the function of formate dehydrogenase.

Paracoccus denitrificans is a gram-negative bacterium and UV radiation of M. organophilum that were unable to capable of autotrophic growth with methanol, methylamine, grow with methanol (11, 16, 26). Because these organisms do or formate as an energy source. In addition, a variety of not grow autotrophically, specific selection of mutations in substrates can be used heterotrophically (24). Methanol arid methanol dissimilation was not possible, and the majority of methylamine are oxidized to formaldehyde by methanol the mutants isolated were found to have defects in the serine dehydrogenase (MDH) and methylamine dehydrogenase, pathway. However, mutants with defects in MDH or respectively (4, 10, 20). Electrons from methanol flow from cytochrome c from Pseudomonas sp. AMl; M. organo- the MDH cofactor (PQQ) via a philum, and also Hyphomicrobium sp. X have been reported postulated CO-binding c-type cytochrome, cytochromes c (11, 16, 21, 28, 31). Whether these mutants have defects in and aa3 to (36). In P. denitrificans all formaldehyde the structural MDH gene, in regulatory genes, or in ancillary generated is thought to be oxidized via formate to carbon functions is unclear. Linkage studies in Pseudomonas sp. dioxide by NAD-dependent dehydrogenases. For synthesis AM1 were performed with the broad-host-range plasmid of biomass, carbon dioxide is fixed by ribulose bisphosphate R68.45, which was shown to mobilize the chromosome in carboxylase, the key of the Calvin cycle (4, 10). this organism (31, 32). Linkage was determined between In methylotrophs using the ribulose bisphosphate cycle, genes involved in the proper expression of MDH, methyl- no data are available on the genetics of methanol oxidation at amine dehydrogenase, and cytochrome c and in the genes this time. The isolation of mutants in other methylotrophs determining resistance to streptomycin, phosphonomycin, has proveti to be difficult. Williams and co-workers were and cycloserine. Linkage relationships in M. organophilum unable to induce mutations by UV irradiation in have been investigated by a gene transformation system Methylomonas albus, Methylococcus capsulatus, and developed by O'Connor et al. (25, 27, 28). The genes coding Methylosinus trichosporium (OB4) and succeeded only with for the of formaldehyde assimilation and genes the mutagen N-methyl-N'-nitro-N-nitrosoguanidine (NTG) involved in MDH expression were reported to be grouped on (37, 38). They suggested that the lack of error-prone SOS a single operon. However, recent results from Allen and DNA repair mechanisms was responsible for this phenome- metabolism are non. Genetic studies have been performed with the obligate Hanson (1) show that the genes for methanol methylotroph Methylophilus methylotrophus, where carbon not closely linked. Genes involved in methanol oxidation fixation follows the ribulose monophosphate pathway. Tem- have been cloned from Pseudomonas sp. AM1 (13, 14) and perature-sensitive, auxotrophic, and glutamate synthase- M. organophilum (1) after identification by complementation deficient mutants have been isolated with NTG (8, 41). of a variety of mutants. However, no mutants with defects in methanol utilization The efficiency and energetics of methanol metabolism have yet been reported. Studies have focused primarily on have been subject of previous studies on P. denitrificans in the facultative methylotrophs Pseudomonas sp. AM1 and our laboratory. Here we report the isolation and character- Methylobacterium organophilum, where carbon assimilation ization of mutants with defects in the metabolism of one- follows the Icl- variant of the serine pathway. Mutants have carbon compounds. In conjunction with the studies on the been isolated by NTG treatment of Pseudomonas sp. AM1 physiological regulation of MDH induction (G. E. de Vries, N. Harms, and K. Maurer, manuscript in preparation), this study forms the basis for future cloning of P. denitrificans * Corresponding author. genes involved in methanol oxidation. 1064 VOL. 164, 1985 P. DENITRIFICANS MUTANTS IN METHANOL OXIDATION 1065

TABLE 1. Bacterial strainsa triphenyltetrazolium chloride per ml. Among red-staining colonies reducing triphenyltetrazolium chloride to the insol- Strain Relevant phenotype' reference uble formazan, white colonies were visible that were picked from the original glycerol plates and tested further for PD1001 Wild type NCIB 8944 growth on methanol. Controls, with the and (DSM 413) wild type the PD11O1 PD1001, Rif This study previously isolated cytochrome c-deficient mutant PD1201 PD1103 P O1101, Spcr This study (35), indicated that incubation on plates either with or PD1201 PD1001, cyt c- (35) without methanol for periods shorter than 24 h always PD1205 PD1101, MeOH-, allyl OH' This study resulted in red colonies. Beyond this period only the wild PD1210 PD1001, Stri, cyt aa3- NS-3 (39) type retained the capacity to reduce triphenyltetrazolium PD1214 through PD1101, MeOH- This study chloride, demonstrating utilization of methanol. With this PD1221 screening procedure 25,000 colonies were tested, and 15 a The rmutants PD1214 through PD1221 were derived from PD1101, except mutants were found that were unable to grow on methanol. the strains PD1215 and PD1220, which are derivatives of strain PD11Q3. In method B, mutagenlied cells were seeded on 9-cm b Rif, rifampin resistance; Spc, sjpectinomycin resistance; Str, streptomycin methanol-agar plates that, were covered with 4 ml of 0.1% resistance; cyt c -, cytochrome c negative; cyt aa'-, cytochrome aa3 nega- ethanol-MM agar. After incubation for 72 h, pinpoint colo- tive; MeOH-, no growth on methanol; allyl OHr, resistance to allyl alcohol; NS-3, definition. nies were detected among otherwise normal-appearing col- onies. In addition a triphenyltetrazolium chloride overlay as described above was performed, and large nonstaining col- onies were found (method C). Both colony types were MATERIALS AND METHODS purified by restreaking and tested for growth on methanol; Among 21,000 colonies tested, 3 mutants that were unable to Bacterial strains and culture conditions. The P. denitrificans grow on methanol were selected via method B, and 5 bacterial strains used in this study are listed in Table 1. Strains mutants were selected via method C. In method D, mutagen- were maintained on Luria,broth (LB) agar plates or stored in ized cells were plated on glycerol-ethanol-MM agar contain- 50% glycerol at -20°C. The bacteria were grown in batch ing. 0.01% allyl alcohol. A single mutant, PD1205, unable to cultures at 32°C either on LB or on mineral salts medium use methanol as the sole source of energy was found among (MM) containing the following (per liter): KH2PO4, 1.7 g; allyl alcohol-resistant colonies. K2HPO4, 8 g; NH4Cl, 1.0 g; Na2MoO4- 2H20, 145 mg; Revertants. Depending on the existing growth deficiencies, MnSO4, 2 mg; EDTA, 37 mg; Fe citrate, 5 ml of a solution a few drops, of a fresh culture of each mutant were spread containing 4 mM FeSO4, 5 mM citric acid, and 18 mM H2SO4; onto one of the following substrates: methanol, methyl- NaHCO3, 0.5 g; CaCl2, 73 mg; MgCl2, 0.1 g (pH 7.5). amine, choline, or glycerol plus nitrate. In the center of the Filter-sterilized carbon sources were added at 0.5% (wt/vol), plate, 10 ,ul of ethyl methanesulfonate or NTG (1 mg/ml) was except formate, which was added at 0.1% (wt/vol). For placed. After 2 to 5 days colonies appeared which were anaerobic growth, MM was used with KNO3 (25 mM) as the purified by restreaking and tested for the original mutant terminal electron acceptor. For autotrophic growth on phenotype. molecular hydrogen as the sole source of energy, the Enzyme activities. All spectral studies were performed composition of the gas phase was 90% N2, 5% C02, 4% H2, with an Aminco DW-2a spectrophotometer (American In- 1% 02. Chemostat cultures were performed in a Bioflo strument Co.). Enzyme activities were determined by fol- fermentor with a working volume of 250 ml under carbon lowing chaniges in dual-wavelength absorption measure- limitation on MM supplemented with 0.01% Difco yeast ments at 25°C (dichlorophenolindophenol, 612, reference 580 extract, 5 mM glucose, and 7.5 mM methanol at a dilution rate nm; NADH, 340, reference 375 nm). For dye-linked enzyme of 0.01 h-1 at 32°C (pH 7.5). PQQ, used in some of the assays, cells were permeabilized with 0.5% toluene for 15 experiments, was a gift of J. A. Duine, Delft, The min at 0°C, centrifuged, and suspended in 1 mM Tris (pH Netherlands. 7.0). Methanol dehydrogenase was assayed by the method of Selection of mutants. Sodium citrate (100 mM, pH 5.5) and Anthony and Zatman (2). Dye-linked formaldehyde dehy- NTG (50 p.g/ml) were added to P. denitrificans log-phase drogenase (FyDH) was assayed by the method of Johnson MM-glycerol cultures, and incubation was continued at and Quayle (17). NAD-dependent FyDH and formate dehy- 350C. At different time intervals the cells were washed twice drogenase activities were determined in whole cells with the with MM, and viable counts were performed. Mutagenized assay mixtures of Johnson and Quayle (17). Cetyl- cells were grown overnight on LB supplemented with 40 Kg tnmethylammonium bromide was included at 0.01% as de- of rifampin per ml. The increase in mutation frequencies was scribed previously for the assay of the soluble hydrogenase determined by titration of cells on LB and on LB supple- in Alcaligenes eutrophus (12). For induction of hydrogenase, mented with either streptomycin (50 ,ug/ml) or cells were grown on glycerol in the presence of hydrogen. spectinomycin (100 pLg/ml). Mutagenized cultures were The capacity for hydrogen uptake was assayed mano- stored in 50% glycerol at -200C. Mutants unable to grow on metrically in 60 mM potassium phosphate buffer (pH 6.8) methanol, but able to grow on. glycerol, were selected via with a Warburg microrespirometer (29) with methylene blue four methods. as the electron acceptor. Protein was measured by the Method A is a variant of the method of Bochner and Lowry et al. method with bovine serum albumin as the Savageau (6). NTG-treated cells were plated on glycerol- standard (19). MM agar plates in a dilution that would result in 100 to 200 Cytochrome spectra and oxidase test-. Dithionite-reduced colonies per plate. Colonies were transferred to a disk of minus air-oxidized spectra of succinate-grown cells were filter paper (Schleicher & Schull Co.; 595, 9 cm) which was measured at 77 K. The oxidase test (NADI reaction, a- placed, bacteria faced up, on an MM plate containing naphthol + N,N-dimethyl-p-phenylene diamine + 02 methanol. After incubation at 320C for 24 h, the filter paper indophenol blue + H20) was performed after growth on was overlaid with 0.7% agar containing 25 ,ug of 2,3,5- succinate plates by the method of Willison et al. (40). 1066 HARMS ET AL. J. BACTERIOL.

NAD-AIDH NADH - Fp ethanol ADH acetoaldehyde acetateIa Fe-S1 4

H2 H!y-UQ10 methanol min dye-linked AIDH I MOH cyt b dye-linked FaDH I TV 11, cyt c ' cyt ccO - POLO formaldehyde formate MoCo carbon dioxide

cyt aa3 / NAD-FyDH NH3 NAD-FaDH RuBP methylamine formamide j cycle

choline cat glycine biomass FIG. 1. Methanol and related catabolic pathways in P. denitrificans. The postulated lesions in the different classes of mutants found are indicated with Roman numerals. Abbreviations: Hy, hydrogenase; Fp, flavoprotein; Fe-S, iron-sulfur center; UQ1o, ubiquinone-10; ADH, ethanol dehydrogenase; A1DH, aldehyde dehydrogenase; MaDH, methylamine dehydrogenase; PQQ, pyrroloquinoline quinone; FaDH, formate dehydrogenase; MoCo, molybdenum cofactor; RuBP, ribulose bisphosphate.

Immunoassays. The enzyme-linked immunosorbent assay utilization by the direct screening techniques described in was performed as described by Mooi et al. (22), except that Materials and Methods. the reactions were arrested by the addition of 50 ,ul of 2.5 M Characterization of mutants. For estimating enzyme activ- H2SO4 and quantitated on a microtiter plate reader ities in the obtained mutants, derepressing conditions for the (Organon, Holland). Titers were calculated, using the half- different enzymes were investigated. MDH induction was maximal optical density value, by comparison with standard detected after growth in glucose-methanol carbon-limited amounts of purified methanol dehydrogenase. For Western continuous culture, in batch cultures on choline or methyl- blotting a 2-mm 8% polyacrylamide gel, prepared by the amine, and in cells grown autotrophically. No MDH activity method of Laemmli (18), was loaded with a sonified cell could be measured after growth on a nonlimiting supply of extract (50 ,ug of protein). After gels were run, the protein carbohydrates, organic acids, or alcohols other than metha- bands were tranferred to nitrocellulose (Schleicher & Schull; nol. Similar results were obtained after batch growth on BA, 85 to 0.45 ,um) as described by Towbin et al. (33). The mixed substrates containing a non-C1 plus metha- nitrocellulose was washed in buffer (10 mM Tris hydrochlo- nol. FyDH and formate dehydrogenase activities could be ride [pH 7.4], 0.5% Tween, 0.9% NaCl) after blotting over- measured under the same derepressing conditions used to night at 100 mA and incubated with immunoglobulin G, detect MDH activities. Hydrogenase was induced after directed against MDH, that was coupled to horseradish mixotrophic growth on glycerol and hydrogen. In contrast to peroxidase. MDH protein bands were visualized with o- earlier reports on the strain employed (23), no hydrogenase dianisidin (Fluka, Switzerland) (15). was detectable under heterotrophic conditions. The 9 mutants presented in this study were selected from RESULTS the 24 mutants obtained based on their distinct characteris- Induction and selection of mutants. The effects of various tics and could be distinguished in five different classes. mutagenic treatments were tested on P. denitrificans. UV Figure 1 shows the route of methanol catabolism leading irradiation or treatment with hydroxylamine, sodium nitrite, to carbon dioxide and includes all known enzymes that may or sodium sulfite did not result in dramatic increases of have a role in this process. Several other substrates with frequencies at which resistance was aquired for streptomy- related catabolism have been included. The characterization cin or spectinomycin (10-8). The presence of 50 p.g of NTG of mutants was based on this scheme, and the postulated per ml for 10 min in a vigorously growing culture resulted in lesions in the different classes of mutants found are indicated 99% killing, and frequencies of drug resistance 5,000 times with Roman numerals. above spontaneous levels were obtained. Class I mutants with a structural defect in the methanol Conventional enrichmnent techniques with a penicillin-type dehydrogenase gene would be expected to be impaired in antibiotic during the nonpermissive growth phase proved methanol utilization only. Four representatives with distin- unsuccessful for the isolation of mutants with defects in guishable characteristics have been studied in detail (Tables methanol utilization. P. denitrificans suffers from variable 2 and 3). Strain PD1214 showed neither enzymatic MDH lag periods before the onset of growth on methanol, resulting activity nor immunogenic reaction with antisera against in a significant percentage of wild-type cells that escape the MDH as determined by the enzyme-linked immunosorbent killing action of the drug. In addition, starvation for energy assay. In the mutants PD1205, PD1215, and PD1220 no resulted in a rapid decrease in the number of viable cells enzymatically active MDH was detectable, but different (data not shown), which negatively influiences the efficiency levels of MDH were detected by the enzyme-linked immuno- of the enrichment procedure. We decided, therefore, to sorbent assay. Strain PD1205 consistently displayed inter- select P. denitrificans mutants with defects in methanol mediary levels of MDH present, whereas PD1215 and VOL. 164, 1985 P. DENITRIFICANS MUTANTS IN METHANOL OXIDATION 1067

TABLE 2. Growth responses of P. denitrificans mutants deficient in the utilization of methanol Growth responseb on the following: Oxidase Class Strain Isolation Methanol Methylamine Choline Formate N03' Ethanol test I PD1214 A6, B1, C5, D1 - + + + + + + PD1205 - + + + + + + PD1215 - + + + + + + PD1220 - + + + + + + II PD1218 A1,B1 - - + + + + + III PD1216 A5 - - - + + + + PD1217 - - + + - + IV PD1219 A2 - - + + + + V PD1221 A1, B1 - - + - - + + Wild type PD1103 + + + + + + + a The number of mutants isolated in each class is preceded by a capital letter indicating the isolation method used (see the text). b Growth responses were determined on solid media. c Nitrate was used either as a nitrogen source or as an electron acceptor.

PD1220 induced MDH as in the wild type. Gels (8% sodium Class II mutants, represented by strain PD1218, were dodecyl sulfate-polyacrylamide) were run with cell extracts impaired in growth on either methanol or methylamine. The from the wild type and these four mutants grown on choline. amounts of MDH induced during glucose- and methanol- Western blots showed that the mutants possessed MDHs limited growth conditions were slightly lower than values with molecular weights equal to that of the wild type. No obtained with wild-type strain PD1103 under glucose-limited conditions were found under which MDH could be detected growth conditions (1 to 4 ,ug of MDH per mg of protein). No in extracts from PD1214 on Western blots (data not shown). MDH was detected in these mutants by enzyme-linked With the exception of PD1220, for all mutants in this class immunosorbent assay when grown in batch cultures that the activities of other enzymes tested were similar to those in would highly stimulate MDH synthesis in the wild type. In the wild type. PD1220 expressed low levels of hydrogenase Table 3 this is shown for choline, but it was observed in compared with the wild type (data not shown), and in autotrophically grown cells as well. The addition of PQQ in revertants obtained on methanol plates this phenotype was the concentration range of 1 F.M to 1 mM did not restore the not restored. This indicates that in this strain a double defect in these mutants under any condition tested. In mutation may have occurred. In contrast to PD1205 and revertants of PD1218, selected for growth on either methanol PD1215, no revertants could be obtained from mutant or methylamine, the wild phenotype was restored. PD1214. Class III mutants, exemplified by the strains PD1216 and PD1217, were unable to grow on methanol, methylamine, or choline. PD1217 could be distinguished from PD1216 by the fact that growth was also impaired with substrates that TABLE 3. Enzyme activities of P. denitrificans mutants deficient would yield aldehydes as catabolic intermediates. The in- in the utilization of methanola creased sensitivity to aldehydes of mutant PD1217 was FyDH FaDH apparent by the inhibition of growth by the mere presence of MDH by ELISA MDHc (NAD)d (NAD)d choline or ethanol in glycerol batch cultures. Significantly raised activities of NAD-dependent FyDH were observed in A B A B A B A B strain PD1217 and shown to be constitutive, since similar PD1103 28 17 3.3 0.8 2.6 1.1 7.1 43.5 levels were found in glycerol batch cultures. A dye-linked 0.7e 5.7e FyDH, hypothesized to act as a general aldehyde dehydro- PD1214 <0.01 <0.01 <0.1 <0.1 2.5 4.9 5.4 56.1 genase in Hyphomicrobium sp. X (20), was detected with PD1205 ND 6 ND <0.1 ND 3.3 ND 61.7 PD1215 ND 17 ND <0.1 ND 5.0 ND 61.6 only low activities in both the mutants and the wild type (0.2 PD1220 1 16 <0.1 <0.1 3.0 0.8 4.2 39.3 nmol of dichlorophenolindophenol per min per mg of pro- PD1218 0.3 <0.01 <0.1 <0.1 1.8 1.4 6.4 40.5 tein). Revertants from strain PD1216 and PD1217 that re- PD1216 0.2 NG <0.1 NG 5.2 0.9e 4.0 5.6e gained the ability to grow on choline would simultaneously PD1217 3 NG <0.1 NG 57.1 21.7e 2.7 5.2e revert with respect to methanol and methylamine utilization. PD1219 0.4 38 <0.1 0.6 2.5 0.3 4.2 49.5 This demonstrates the pleiotropic character of these muta- PD1221 35 20 2.5 2.2 1.1 1.2 <0.1 1.5 tions. However, three out of five revertants of strain PD1217 aNG, no growth; ND, not determined; A, glucose and methanol-limited that were tested retained the observed constitutive synthesis chemostat cultures; B, choline batch cultures. of NAD-dependent FyDH. b Micrograms of MDH per milligram of protein. During the oxidation of methanol or methylamine, elec- c Nanomoles of dichlorophenolindophenol per minute per milligram of trons are donated to the electron transport chain at the level protein. d Nanomoles of NADH per minute per milligram of protein. of cytochrome c (36) (Fig. 1). Therefore, the inability of class eEnzyme activities determined after batch growth in glycerol plus IV mutants to grow on methanol or methylamine may be due hydrogen. to a defect in the respiratory chain at or beyond cytochrome 1068 HARMS ET AL. J.. BACTERIOL.

This may, of course, have been due to the limited number of mutagenized colonies screened. Nevertheless, some aspects of the missing classes will be included in the discussion. Class I. Mutants that are unable to grow on methanol, but do grow on methylamine and choline, most likely have a defect in the proper expression of MDH. The availability of antibodies against MDH offered the possibility to detect mutations in the structural gene for MDH. Mutant PD1214, which does not express MDH at all and is not revertable, may, for instance, suffer a deletion in a regulatory gene or in

I I ...... the structural gene such that protein cannot be synthesized. C300 530 560 590 620 The mutations in PD1205 and PD1215, which are both revertable and therefore are single point mutations, may also WAVELENGTH (nm) be located in the structural gene, since expression of inactive FIG. 2. Dithionite-reduced minus air-oxidized difference MDH was found. Lack of detectable enzymatic activity in cytochrome spectra of succinate-grown P. denitrificans cells. Sym- vitro may be caused by a defect in the or in the bols: (-) wild type; (.) mutant PD1219. interaction with PQQ. Classes II and IV. Mutants that do not grow on methanol or on methylamine may have a defect in the regulation of the respective dehydrogenases. Alternatively a mutation may c. Electron flow to oxygen is catalyzed by cytochrome c and have occurred in the biosynthesis of common components cytochrome aa3 and may be analyzed by the NADI reaction involved in electron transport, e.g., PQQ, cytochrome c, or (40). Mutant PD1219 was negative in this test (Table 1). In cytochrome aa3. The role of cytochrome aa3 is probably not dithionite-reduced minus air-oxidized difference spectra it essential, since mutants with defects in this cytochrome was shown that cytochrome c is virtually absent in this were found to grow normally on methanol. A postulated mutant (Fig. 2). The alpha-absorption maxima, characteris- alternative, oxidase a1 (34), may function under these con- tic of a cytochrome c (547 nm) and a cytochrome cl (550 nm), ditions. In mutants with a defect in the biosynthesis of PQQ, were not detectable, whereas cytochrome aa3 was present as no direct influence on the synthesis of apo-MDH is ex- in the wild type, as indicated by the presence of an absorp- pected. Although cotranslational activation of MDH and tion maximum at about 604 nm. Anaerobic growth of PD1219 PQQ incorporation remains a possibility, the absence of on nitrate as the electron acceptor was poor, whereas nitrite significant MDH levels under inducing conditions as shown was accumulated and less amounts of nitrogen were evolved for PD1218 (class II) argues for a defect in a regulatory gene than by the wild type. This suggests a block in electron that probably influences both MDH and methylamine dehy- transfer to nitrite and nitrous oxide reductases that are drogenase synthesis. Since mutants defective in the synthe- known to receive electrons originating from cytochrome c sis of PQQ were not found, PQQ either may fullfil a still (30). In revertants obtained on either methanol or methyl- distinct and essential role in P. denitrificans metabolism or amine plates the wild phenotype was restored. A may share its biosynthetic route with that of other compo- cytochrome aa3 mutant isolated by Willison et al. (39) as well nents essential for growth. A carbon monoxide-binding as cytochrome aa3 mutants isolated in our laboratory (F. van cytochrome c has been postulated to be involved in electron Leeuwen, unpublished results) were able to grow on meth- transfer from PQQ to cytochrome c in P. denitrificans (36). anol. This component possibly has a role analogous to that of Class V mutants, represented by strain PD1221, were cytochome CL in Methylophilus methylotrophus and Pseu- impaired in growth on either methanol, methylamine, for- domonas sp. AMI or blue copper proteins in the electron mate, or formamide. This mutant induced MDH and FyDH transfer originating from methylamine dehydrogenase in at levels similar to those of the wild type, but low levels of Pseudomonas sp. AML. If this component was essential, formate dehydrogenase activity were detected. With an Mo mutants with a defect in cytochrome cc, would have been cofactor complementation test (9), this mutant was not able detected by the absence of growth on methanol, but other- to synthesize a functional Mo cofactor. Colonies of this wise unchanged MDH regulation or in vitro activity. Re- mutant grown on choline agar plates acidified the medium as markably, all cytochrome c mutants obtained here (mutant demonstrated by the inhibition of triphenyltetrazolium chlo- PD1219 [class IV]) or reported previously (40) appear to lack ride reduction. On methanol plates, mutant PD1221 was able most, or all, spectrally detectable cytochrome c. Since P. to cross-feed all other mutants isolated. These observations denitrificans possesses three distinct c-type cytochromes seem to warrant the conclusion that formate was excreted by (5), this would imply either a common regulation or a similar this mutant under these conditions. Revertants obtained on mode of heme incorporation. Lesions leading to the loss of a nitrate would simultaneously revert with respect to metha- single c-type cytochrome may be difficult to distinguish nol, methylamine, and formate utilization. phenotypically due to functional cross-complementation. Class III. Mutants that grow on neither methanol, me- DISCUSSION thylamine, nor choline were postulated to have a defect in formaldehyde catabolism. In P. denitrificans NAD- The P. denitrificans mutants with defects in the metabo- dependent and dye-linked enzyme activities have been dem- lism of one-carbon compounds were found to belong to five onstrated (20). Since MDH was shown to oxidize formalde- different classes and to give useful information on the hyde in vitro (4), three enzymes are potentially involved in metabolism of methanol. Considering the many compounds formaldehyde catabolism. Indeed when growth of P. known to be involved in methanol oxidation (Fig. 1), one denitrificans occurred on methylamine or choline as the sole might have expected a number of additional specific mutant source of carbon, considerable amounts of MDH were classes that were not isolated with the method employed. induced with high activities (data not shown). A significant VOL. 164, 1985 P. DENITRIFICANS MUTANTS IN METHANOL OXIDATION 1069 role for MDH in formaldehyde oxidation may be discounted, studies, the mutants obtained may be used as tools in the however, since mutant PD1214 (class I), which completely isolation of genes involved in C1 metabolism. lacks MDH, possessed wild-type growth rates on methyl- amine or choline. Since in many organisms in vitro- ACKNOWLEDGMENT measured activities were low and no specific increase in enzyme activities was noted during growth on methanol (20), These investigations were supported by the Netherlands Technol- the role of the dye-linked FyDH in methanol metabolism is ogy Foundation. unclear. Marison and Attwood have postulated that the dye-linked FyDH in methylotrophs is a general aldehyde LITERATURE CITED dehydrogenase not directly involved in the dissimilation of 1. Allen, L. N., and R. S. Hanson. 1985. Construction of broad host C1 compounds (20). In contrast, it has been reported for range cosmid cloning vectors: identification of genes necessary Pseudomonas aminovorans that the dye-linked enzyme is for growth of Methylobacterium organophilum on methanol. J. induced during growth on C1 compounds (3). In the mutants Bacteriol. 161:955-962. PD1216 and PD1217 the activity of MDH and NAD- 2. Anthony, C., and L. J. Zatman. 1964. The microbial oxidation of dependent FyDH, theoretically sufficient for formaldehyde methanol. The methanol oxidizing enzyme of Pseudomonas sp. is On the basis of the M27. Biochem. J. 92:614-621. catabolism, unimpaired. characteristics 3. Bamforth, C. W., and M. L. O'Connor. 1979. The isolation of of the mutants PD1216 and PD1217, we have concluded that pleiotropic mutants of Pseudomonas aminovorans deficient in in P. denitrificans a general aldehyde dehydrogenase is the ability to grow on methylamine and an examination of their involved in the oxidation of formaldehyde. The reason for enzymatic constitution. J. Gen. Microbiol. 110:143-149. the low in vitro activity of the postulated enzyme reported in 4. Bamforth, C. W., and J. R. Quayle. 1978. Aerobic and anaerobic this study and by others (3, 17, 20) is not known. Lability of growth of Paracoccus denitrificans on methanol. Arch. Micro- the enzyme and the use of nonphysiological electron accep- biol. 119:91-97. tors (3, 20) are possible explanations. Significant differences 5. Berry, E. A., and B. L. Trumpower. 1985. Isolation of ubiquinol exist between PD1216 and PD1217. The lesion in mutant oxidase from Paracoccus denitrificans and resolution into PD1216 a cytochrome bc, and cytochrome c-aa3 complexes. J. Biol. may result, therefore, in similar but less stringent Chem. 260:2458-2467. phenotype compared with that of PD1217. It is possible, 6. Bochner, B. R., and M. A. Savageau. 1977. Generalized indica- however, that a defect in a yet-unknown component exists. tor plate for genetic, metabolic, and taxonomic studies with The NAD-dependent FyDH is constitutively expressed both microorganisms. Appl. Environ. Microbiol. 33:434-444. in the mutant PD1217 and in 60% of its revertants. Loss of 7. Burke, K. A., K. Calder, and J. LasceOles. 1980. Effects of the general aldehyde dehydrogenase may be compensated molybdenum and tungsten on induction of nitrate reductase and by an additional mutation that resulted in the constitutive formate dehydrogenase in wild type and mutant Paracoccus expression of formaldehyde dehydrogenase. Alternatively a denitrificans. Arch. Microbiol. 126:155-159. single mutation in PD1217 resulted in both phenotypes. 8. Byrom, D. 1984. Host-vector systems for Methylophilus Since among PD1217 revertants derivatives were found that methylotrophus, p. 221-223. In R. L. Crawford and R. S. Hanson (ed.), Microbial growth on C, compounds. Proceedings either retained or lost the constitutive expression of NAD- of the Fourth International Symposium. American society for dependent FyDH, both loci, if separate, must be closely Microbiology, Washington, D.C. linked. 9. Claassen, V. P., L. F. Oltmann, S. Bus, J. van 'tRiet, and A. H. Class V. Mutants were obtained that were defective for Stouthamer. 1981. The influence of growth conditions on the growth on methanol or formate as the sole source of energy. synthesis of molybdenum cofactor in Proteus mirabilis. Arch. In P. denitrificans, formate is known to be oxidized by an Microbiol. 130:44 49. NAD-dependent formate dehydrogenase, which was re- 10. Cox, R. B., and J. R. Quayle. 1975. The autotrophic growth of ported to contain an Mo cofactor (7). The identity of mutant Micrococcus denitrificans on methanol. Biochem. J. PD1221 with the to a functional Mo 150:569-571. inability synthesize 11. Dunstan, P. M., C. Anthony, and W. T. Drabble. 1972. Micro- cofactor was confirmed by the finding that these mutants bial metabolism of C, and C2 compounds. The involvement of were also impaired in growth on nitrate, whether nitrate was glycollate in the metabolism of ethanol and of acetate by used as the nitrogen source or as the terminal electron Pseudomonas AML. Biochem. J. 128:99-106. acceptor. 12. Friedrich, B. E. Heine, A. Finck, and C. G. Friedrich. 1981. No mutants were found in either formaldehyde or formate Nickel requirement for active hydrogenase formation in NAD-linked dehydrogenases. For both substrates dye- Alcaligenes eutrophus. J. Bacteriol. 145:1144-1149. linked dehydrogenase activity could be demonstrated. A 13. Fulton, G. L., D. N. Nunn, and M. L. Lidstrom. 1984. Molecular tetrazolium nonreducing response, in the selection proce- cloning of a malyl coenzym A from Pseudomonas sp. dures to strain AM1, a facultative methylotroph. J. Bacteriol. followed, may require both enzymes be inactivated 160:718-723. or alternatively the production of toxic metabolites leading 14. Gautier, F., and R. Bonewald. 1980. The use of plasmid R1162 to cell death. The first situation indeed occurred in class V and derivatives for gene cloning in the methanol-utilizing Pseu- mutants, which were devoid of all formate dehydrogenase domonas AML. Mol. Gen. Genet. 178:375-380. activity, due to the absence of a functional Mo cofactor. The 15. Hawkes, R., E. Niday, and J. Gordon. 1982. A dot-immuno- isolation of class III mutants probably was due to the lethal binding assay for monoclonal and other antibodies. Anal. character of the lesion when formaldehyde is produced Biochem. 119:142-147. during catabolism of methanol or choline. Even enhanced 16. Heptinstall, J., and J. R. Quayle. 1970. Pathways leading to and and constitutive levels of the NAD-dependent formaldehyde from serine during growth of Pseudomonas AM1 on C, com- dehydrogenase were not sufficient to relieve these mutants pounds or succinate. Biochem. J. 117:563-572. from 17. Johnson, P. A., and J. R. Quayle. 1964. Microbial growth on C1 this defect. Significant amounts of formaldehyde may compounds. 6. Oxidation of methanol, formaldehyde and for- therefore be produced at a specific site in the environment of mate-grown Pseudomonas AML. Biochem. J. 93:281-289. the membrane. 18. Laemmli, U. K. 1970. Cleavage of structural proteins during the Attempts to obtain the missing mutants are continuing. It assembly of the head of bacteriophage T4. Nature (London) is obvious that, apart from their usefulness in physiological 227:680-685. 1070 HARMS ET AL. J. BACTERIOL.

19. Lowry, 0. H., N. J. Rosebrough, A. L. Farr, and R. J. Randall. domonas AML. J. Gen. Microbiol. 129:262-2632. 1951. Protein measurement with the Folin phenol reagent. J. 32. Tatra, P. K., and P. M. Goodwin. 1984. R-factor-mediated Biol. Chem. 193:265-275. chromosome mobilization in the facultative methylotroph Pseu- 20. Marison, I. W., and M. M. Attwood. 1980. Partial purification domonas sp. strain AM1, p. 224-227. In R. L. Crawford and and characterization of a dye-linked formaldehydedehydro- R. S. Hanson (ed.), Microbiol growth on Cl compounds. Pro- genase from Hyphomicrobium X. J. Gen. Microbiol. ceedings of the Fourth International Symposium. American 117:305-313. Society for Microbiology, Washington, D.C. 21. Marison, I. W., and M. M. Attwood. 1982. A possible alternative 33. Towbin, H., T. Staehelin, and J. Gordon. 1979. Electrophoretic mechanism for the oxidation of formaldehyde to formate. J. transfer of proteins from acrylamide gels to nitrocellulose Gen. Microbiol. 128:1441-1446. sheets: procedure and some applications. Proc. Natl. Acad. Sci. 22. Mooi, F. R., F. K. de Graaf, and J. D. A. van Embden. 1979. USA 119:142-147. Cloning, mapping and expression of the genetic determinant 34. van Verseveld, H. W., M. Braster, F. C. Boogerd, C. Chance, that encodes for the K88ab antigen. Nucleic Acids Res. and A. H. Stouthamer. 1983. Energetic aspects of growth of 6:849-865. Paracoccus denitrificans: oxygen limitation and shift from 23. Nokhal, T.-H., and H. G. Schlegel. 1980. The regulation of anaerobic nitrate-limitation to aerobic succinate-limitation. hydrogenase formation as a differentiating character of strains Arch. Microbiol. 135:29-236. of Paracoccus denitrificans. Antonie van Leeuwenhoek J. Mi- 35. van Verseveld, H. W., K. Krab, and A. H. Stouthamer. 1981. crobiol. Serol. 46:143-155. Proton pump coupled to cytochrome c oxidase in Paracoccus 24. Nokhal, T.-H., and H. G. Schlegel. 1983. Taxonomic studies of denitrificans. Biochim. Biophys. Acta 635:525-534. Paracoccus denitrificans. Int. J. Syst. Bacteriol. 33:26-37. 36. van Verseveld, H. W., and A. H. Stouthamer. 1978. Electron- 25. O'Connor, M. L. 1981. Extension of the model concerning transport chain and coupled oxidative phosphorylation in meth- linkage of genes coding for Cl-related functions in anol-grown Paracoccus denitrificans. Arch. Microbiol. Methylobacterium organophilum. Appl. Environ. Microbiol. 118:13-20. 41:437-441. 37. Williams, E., and M. A. Shimmin. 1978. Radiation-induced 26. O'Connor, M. L., and R. S. Hanson. 1977. Enzyme regulation in filamentation in obligate methylotrophs. FEMS Microbiol. Lett. Methylobacterium organophilum. J. Gen. Microbiol. 4:137-141. 101:327-332. 38. Williams, E., M. A. Shimmin, and B. W. Bainbridge. 1977. 27. O'Connor, M. L., and R. S. Hanson. 1978. Linkage relationships Mutations in the obligate methylotrophs Methylococcus between mutants of Methylobacterium organophilum impaired capsulatus and Methylomonas albus. FEMS Microbiol. Lett. in their ability to grow on one-carbon compounds. J. Gen. 2:293. Microbiol. 104:105-111. 39. Willison, J. C., B. Haddock, and D. H. Boxer. 1981. A mutant of 28. O'Connor, M. L., A. Wopat, and R. S. Hanson. 1977. Genetic Paracoccus denitrificans deficient in cytochrome aa3. FEMS transformation in Methylobacterium organophilum. J. Gen. Microbiol. Lett. 10:249-255. Microbiol. 98:265-272. 40. Willison, J. C., and P. John. 1979. Mutants of Paracoccus 29. Peck, H. D., Jr., and H. Gest. 1956. A new procedure for assay denitrificans deficient in c-type cytochromes. J. Gen. Microbiol. of bacterial hydrogenases. J. Bacteriol. 73:70-80. 115:443-450. 30. Stouthamer, A. H., F. C. Boogerd, and H. W. van Verseveld. 41. Windass, J. D., M. J. Worsey, E. M. Pioli, D. Pioli, P. T. Barth, 1982. The bioenergetics of denitrification. Antonie van K. T. Atherton, E. C. Dart, D. Byrom, K. Powell, and P. J. Leeuwenhoek J. Microbiol. Serol. 48:545-553. Senior. 1980. Improved conversion of methanol to single cell 31. Tatra, P. K., and P. M. Goodwin. 1983. R-plasmid-mediated protein by Methylophilus methylotrophus. Nature (London) chromosome mobilization in the facultative methylotroph Pseu- 287:396-401.