Sequence and characterization of three genes within the hy drogenase gene cluster of Bradyrhizobium japonicum

Gene, 141 (1994) 47752 0 1994 Elsevier Science B.V. All rights reserved.

GENE

47

0378-l 119/94/$07.00

07727

Sequence and characterization of three genes within the hydrogenase gene cluster of Bradyrhizobium japonicum (Nickel enzyme; nitrogen

Changlin

fixation;

heterologous

expression;

recombinant

DNA; soybeans)

Fu and Robert J. Maier

Department of BiologJl, The Johns Hopkins University, Baltimore, MD 21218, USA Received by R.E. Yasbin: 15 September

1993; Revised/Accepted:

25 October/25

October

1993; Received at publishers:

25 November

1993

SUMMARY

A 2.0-kb DNA fragment downstream from the hydrogenase-encoding structural genes within the hydrogenase gene cluster of Bradyrhizobium japonicum was sequenced. Analysis of the nucleotide (nt) sequence revealed three open reading frames (ORFs), designated hupC, hupD and hupF, which encode polypeptides of 2821 and 10.7 kDa, respectively. Based on analysis of the nt sequence and physiological studies, hupSL (hydrogenase structural genes) and hupCDF are organized as a single transcriptional unit. Plasmid pRY12 carrying hupSL genes did not complement (restore) hydrogenase activity of the hupSL deletion mutant strain (JHCS2), whereas the activity of the mutant was considerably restored by pLD22 harboring the entire hydrogenase operon (hupSLCDF genes). Western blots revealed a very low level of hydrogenase protein in JHCS2 containing pRY12. The results suggest that the products of the hupCDF genes may be involved in either stabilizing the hydrogenase peptides (i.e., from degradation) or in post-translational regulation of hydrogenase production. The products of hupC and huiD were successfully expressed in Escherichia co/i by a phage T7 promoter system, although the apparent sizes of the gene products were slightly larger than those calculated from the deduced amino-acid sequences.

INTRODUCTION

totrophs,

thermophiles

Hydrogenases, which catalyze either the uptake or evolution of H,, are widely distributed in the bacteria and the archae. The physiological groups containing these HZ-metabolizing enzymes include aerobes, anaerobes, facultative anaerobes, autotrophs, heterotrophs, pho-

strate fermentations

The enzymes

hydrogenases N,

SSDI 0378-l

119(93)E0723-Q

wt, wild type;

[I,

consisting

ATP

bacteria the

of uptake

benefits biolog-

nitrogenase-evolved

chain, thereby making the nitrogenase reacenergy efficient. Hup is a membrane-bound

tally

Sp, spectinomycin;

by recycling

the presence

respiratory tion more

amino acid(s); B., Bradyrhizobium; BGal, Abbreviations: aa, B-galactosidase; bp, base pair(s); Hup, hydrogen-uptake hydrogenase(s); hup, gene encoding Hup; kb, kilobase or 1000 bp; MB, methylene

sulfate; state.

In addition,

sub-

in these di-

generating

protein,

SDS, sodium dodecyl denotes plasmid-carrier

bacteria.

roles in H, metabolism,

and energy conservation

(Hup) in N,-fixing

fixation

N,-fixing

hydrogen,

Correspondence to: Dr. R.J. Maier, Department of Biology, The Johns Hopkins University, Baltimore, MD 21218, USA, Tel. (l-410) 516-8276; Fax (l-410) 516-5213; e-mail: [email protected]

blue; nt, nucleotide(s); ORF, open reading frame; PAGE, polyacrylamide gel electrophoresis; RBS, ribosome-binding site; Rif, rifampicin;

play crucial

verse microorganisms. ical

and symbiotic

in a cytochrome-dependent

of 33-35

and 65kDa

1985; Stults et al., 1986a). The enzyme a component Maier,

(Arp,

subunits

contains

and has been well characterized 1985; Moshiri

and Maier,

(Arp,

nickel as enzymati-

1988; Ferber

and

1992).

The structural

genes (hupSL) encoding

B. japonicum

Hup have been sequenced by Sayavedra-Soto et al. (1988). They reported that the genes for the small and

48

large Hup subunits were transcribed unit (operon), and the transcription mediately

as one transcription was terminated im-

after the gene (hupL) encoding

by a termination Although

the

loop

the large subunit

(Sayavedra-Soto

sequence

of

a

third

et al., 1988). potential

following htlpL was not completely (Sayavedra-Soto et al., 1988), the predicted sequence

of the translated

(partial)

sequence

ORF

determined N-terminal

Sequences further downstream from hupF did not reveal the presence of any additional ORFs immediately following

hupF.

further

downstream

Interestingly

(5’-T GGCACN,TT

however,

GCT-3’)

approx.

hupF,

from

a

very closely

250 bp sequence

matching

the

consensus for sigma-54-type promoter sequence was found. It is possible that this region with the sigma-54

was later

consensus

sequence

could serve as a promoter

for a fur-

found to exhibit strong similarity to other gene products in many Hup-containing bacteria, such as HupM in

ther downstream region, including cessing (Fu and Maier, 1993a).

Rhodobacter

Gene hupC encodes a protein of 28 kDa (247 aa). A hydropathy plot (Kyte and Doolitle, 1982) of the deduced

cupsulutus

(Cauvin

et al., 1991), HoxZ

in

Azotobactev uinelandii (Menon et al., 1990a; 1992), HupC in Rhizobium leguminosurum (Hidalgo et al., 1992), HoxZ in

Alculigenes

HupC

eutrophus

in Escherichia

(Kortltike

coli (Menon

et al.,

1992)

and

et al., 1990b). In all

cases, these ‘third genes’ in the operon

are cotranscribed

with the hup genes; it was suggested for some of the above bacteria that these third genes are also somehow involved in Hup activity (Cauvin et al., 1991; Sayavedra-Soto and Arp, 1992). Based on molecular and genetic studies it is known that in addition to hydrogenase structural genes, there are several other genes located both downstream and upstream from the structural genes which are proposed to be involved in Hup activity, regulation, processing and/or metal incorporation into the enzyme. In this study, we report three additional genes (hupCDF) located downstream from the hup genes of B. juponicum. Based on the nt sequence and physiological studies, these genes are cotranscribed with the hup structural genes (hupSL), and seem to be involved in stabilizing the Hup proteins or in post-translational regulation of Hup.

RESULTS

AND DISCUSSION

(a) Nucleotide sequence and analysis A approx. 2.0-kb SacI-EcoRI fragment, located downstream from the Hup structural genes within a hup gene cluster in pSH22 (Horn et al., 1985), was sequenced. Analysis of this nt sequence revealed the presence of three ORFs designated hupC, hupD and hupF (Figs. 1 and 2) based on the sequence homology to other genes from different H,-metabolizing bacteria. hupC, hupD and hupF are all transcribed in the same direction as the Hup structural genes (hupSL), and are closely linked, with intergenic spaces of 16 bp between hupC and hupD and 26 bp between hupD and hupF. Gene hupC was separated by 9 bp from the stop codon (TAG) of hupL structural gene, encoding the large subunit. No transcriptional termination loops were observed between hupL and hupC. It is clear that the hupSL and hupCDF are organized in a single operon, although this is in contrast to a previous report (Sayavedra-Soto et al., 1988).

genes for Hup pro-

aa sequence of HupC (the product of the hupC gene) revealed that HupC was a membrane-bound protein with four transmembrane showed

58%

domains

aa identity

(data

not shown).

with HupM

HupC

of R. capsulutus

(Richaud et al., 1990; Cauvin et al., 1991), 63% with HoxZ of A. vinelundii (Menon et al., 1990a; 1992), 53% with HyaC of E. coli, (Menon et al., 1990b), 71% with HupC of R. leguminosurum (Hidalgo et al., 1992) and 54% with HoxZ of A. eutrophus (Kortltike et al., 1992). Alignment of these gene products shows a large number of conserved residues including six His (Fig. 3A). Mutant analysis studies suggest that the HupM protein of R. capsulatus may anchor the Hup protein into the cytoplasmic membrane, as mutations in the HupM-encoding gene prevented transfer of electrons from Hz to the respiratory chain (Cauvin et al., 1991), whereas HoxZ of A. vinelundii seems to be involved in activation of Hup (SayavedraSoto and Arp, 1992). In the case of the rumen bacterium Wolinellu succinogenes, the third ORF of the hup operon encodes a hydrophobic cytochrome b, which has been proposed as the direct electron acceptor from Hup (Dross et al., 1992). No specific roles have yet been proposed for a third gene product from the other hydrogenasecontaining bacteria. Gene hupD encodes a 21-kDa protein (192 aa). HupD showed 68% identity with HupD of R. leguminosarum (Hidalgo et al., 1993), 41% identity with HyaD of E. coli (Menon et al., 1990b), 41% identity with HoxM of A. eutrophus (Kortltike et al., 1992), 45% identity with HupD of R. cupsulutus (Colbeau et al., 1993) and 47% identity with HoxM of A. vinelandii (Menon et al., 1992 and Fig. 3B). Previous studies of a hoxM mutant of A. eutrophus suggested the HoxM protein is required for the attachment of the membrane-bound Hup to the cytoplasmic membrane (Kortliike et al., 1992), whereas the mutation in the homologous gene (hyuD) of the hup-I operon of E. coli seemed to affect the processing of the small and large subunits of Hup-1 (Menon et al., 1991). Gene hupF encodes a 10.7-kDa protein (98 aa). HupF showed 45% identity with the HupF of R. leguminosarum (Hidalgo et al., 1993), 32% identity with the HoxL of A.

49

------c

EeO I

’

EE II

m

RI I

I H

0

pSH22

PRY1 2

-

B 1 I H

B

B

m

E I I 1 HH

qBg

E I

E ,

II

BP

I

1 kb

s ‘:

pLDlO0

Fig. 1. Genetic and physical map of the hup gene cluster in pSH22 (Horn et. al., 1985) and its subclones. hupS and hupL (arrowed boxes) indicate the genes encoding the small and large Hup subunits, respectively. The small open box below the pSH22 map indicates the region involved in nickel metabolism (Fu and Maier, 1991). The hatched box indicates a region involved in Hup processing (Fu and Maier, 1993a). The dashed line above the pSH22 map indicates the region of DNA that was deleted in the chromosome of the mutant strain JHCSZ (Fu and Maier, 1993b). The arrowed line above the pSH22 map indicates the region of DNA (SacI-EcoRI fragment) sequenced in this study (genes hupCDF, see Fig. 2). pLD22 was generated by cloning a 8.0-kb Hind111 fragment of pSH22 into the Hind111 site of pVKlO1 broad-host-range vector (Knauf and Nester, 1982). E, EcoRI; B, BarnHI;

Bg, Bglll; H, HindIII;

S, SacI; not all restriction

sites on subclones

are indicated.

144 36

288

84 432

132

BPAACQVLAVF~LTR~LWAPV~N~~S~Q*~~2PV~~AQ~~~~VL~E~A CACCCTGG~

mTlrCCC’.ZT~G

576

TTCCGCTCTATKZ~~TC

180

WYAPLERKPXWYVGHNPLAQTAHPTGPTLPVAPNlVTGPALYSEGQGI P’GAmcGAACAKc

AGCACCTCUWCCTTCGGCATGXCGCGClGGTC

GTGTTCGTGATGGTGCACATCTACCCCGCGGTCCGTGMGACATC

228

DSWQEKLPCWVFAIWPNSQDVl3TWtl~LGM”ALVVFVt!VHIYAAVREDI A-~TCA~TGA~CGCAATTCCGCGAC

~TGACGATG@XACA A”@7HP

MSRQSIISSMISGBRQPRD’

861

‘CTGGTGCmTCDKmT

S

~mCCCOTCCG~C_~~~ffC~~~

S

QD

NR

IL

“L

G

TGTKCTCGTGAACTATC~

I

G

N

I

23

LWADE

CGACCGCCTTATCGTGTTCGA!XCCAT

1152

~~TGcGcWLcGRcwycn;cCTC~~C~~~~~~~~~~~C~C~C~TC~C~~CC~

119

DYGLEPGRLKLVRDDEVPRPTGAXAMSLflQTGFQEVISAADLLGRCPK GCAKTCGTGCTGACGGC lGCCMXC_GAKTCG

GccccGCTGAcoc

ATCGATCTCGCCTGCU\GGTCTTGGCCGAGTGGGCC GTCACCGTCAGCCGCCG 1296

CcxxGGTGCKGAC~w~

167

SLVLIGtQPLDLEDWGGPLTPPVRDQIAPS~DLACQVLASWG”TVSRR CAGCGCGcuTTGGcGGUXC

CGACCGan;CTfGCCAATW\W\TCGATUICGCCARTT~-~-C-C~~~~T~

WLTG-TCGGCTTGCCGATGACW\TTGTCGRGRCUiRT

SAPLAESBRLLANDIDBANYEnRPR’ GGCGTCTCGXGCTC

HupP

~~CU3GMT~~~~___C~~~C_~~

H

C

L

G

L

P

M

T

I

V

E

T

D

TCTATATC-CGCGATCCGGCTTCTCGA~GCGCCTGATT

GVSALCAPRGEQRRVSHLLLS”PPVGTf,VLVYIDTAI,-‘LLDSESARLI GGcGCcGcGATccRcGoK:

T~CCGCTCGACGG

=======

13 1584 61

98

GTGCCUTCICCAUJKXTATGGAATGCGTCTTTCTATCn

GGTU\RTGT-TA-~A~CTARCATAmTGGUC

1400

1728

TCGRCCGCWLGCC-~GCCRGCCCACCTGCGGTCCGR TGACGATTGCGACCGCTTCTTCGCCGACCTGA

GAAIDGLGAALDGEDCDRPPADLIDREPPLPAHLRSE’ GCCTcCG?xGCCGeA~

,008 71

GPGVRAVELFSRRYAVPDNVTILDGGTQGLYLVNYLESADRLIVFDAI CGACTA~CC

720

__

--

TATCTTATCATC=X%GGCT 0cOTCATCCC.C“XCAATAAAATATCTGATGTA

TU;CTTTCACTGAATA~ _H;mfIII CGAGCAATTCATAA‘XTT

-

1872

s _TCCAArnTTC

Fig. 2. Nucleotide sequence of the B. japonicum hupC,D,F genes, and deduced aa sequences. The 3’ end of hupL is shown RBS are underlined. The arrows indicate the inverted repeats. A possible sigma-54 consensus sequence is double-underlined.

1979

at the upper left. Potential Methods: For sequencing,

appropriate DNA fragments were cloned into the pBluescript II KS(+) and II SK(+) (Stratagene, La Jolla, CA, USA). The nested deletions were made by standard protocol using the exe/mung bean sequencing system kit from Stratagene. DNA was sequenced in both orientations. The nt sequence data were analyzed by using the PC/GENE software package (IntelliGenetics, Mountain View, CA, USA). The deduced aa sequences were compared with the Swiss-Prot protein sequence data bank (release 26). The sequence has been deposited in the GenBank with accession No. L24446.

vinelandii (Menon

et al., 1992) and 30% identity with the HupF of R. capsulatus (Colbeau et al., 1993 and Fig. 3C). A stretch of M-C-X-G-X-P at the N terminus of the predieted HupF was also found in HoxL of A. eutrophus (Kortltike et al., 1992). The functions of these homologous products are unknown.

(b) Physiological function of the hupCDF gene products Based on the nt sequence, hupCDF genes are cotranscribed with hupSL genes (i.e., a single transcriptional unit). To understand possible functions of hupCDF genes, complementation experiments were conducted with a Hup- mutant of B. japonicum (strain JHCS2 which was

1234

79 90

68 57 68 76

nST%J@ADRIADAlvrDEGAvASGREIKA~vn3FFIGSP-Pm YVEAPRBW

LluIPP

be.3

GSA

169 180 158 146 158 166

WVSMMWEUI~VPbbWT

YIRFAEPMQYVLN

YI~Ia~~Y~~~~~IO~~-~IT YIRPI8FAnGQrJJwFLILSPx~~ YIRFAliFVAAYIFNGl4U3R~~~SR~AKFFI~LT YIR SFA RWAFW F P

IIUmRJU[RPSADIGSUPLPQ~T 1

WXFL

P

QENP

XT

GFALYSlSGQGIlX3WJ~AIW---~DVElWE~~VV?VIWE

FI4 T

IYAAVR.SDIWRQSIISSJ4IB3SRQFRD GWALYWW--s ~FNIBIYAAVREDVbWWWS THISGSR~IS GFALYAEG~LF---*~ WPWVWYWVREDIVSRQSLI~D OT~Y~BBP~FA-~~~I~I~~D~I~ GFALYSSGAGWWWJYS~SIW---PkWDI BTIBBL(PWVILVFVMVEIYVAVREDIl4SRQWISSM~D GFAMY-PLL--~DVS~IWPVIVWUAE3D~TWWYRTFKD GAY,?. F V 8 BYAPXD 81 OR EBLQMW F

247 262 240 234 239 244

85 90 81 77 81 84

175 179 171 167 171 174

1234

kDa

200116 97.466.2-

45

-

31

-

RLLAWDIDSANYSMWA

192 PLLWDID?6FtYERRAE PAALK! 202 195 ~~WSp3 AVTVPELALDRYXAERP8 IGDERFblFQDL 202. P~LALGR~ABl!XAYRRGDIRFIXQPIZDD 207 DPSLMWAYERI~PSEDEACRI~SRFFP~ 210

--HupC

--HupD 21.50

88 QRRVSSLUSNPPIE'l!N~IGAAIffi~DRF--FADLIDRE K!IGIPSR"""GSSF IA@xREN8818ulL~ LLl%LGSAIRVLMDKA8AIDDALAGL&XAVEGRAFDML--FADLISPJ! 88 MCIGIPLR~GDEKXWnDlRLMEPP APaDIPLLIRLDMRSILMGRAMIRXALMWAVQAGDPAUAGLFADL-DlW 89 88 lbx&m~-

PQLPPBWAQLPPlUPT PaPPam

106

106

PLPEL Fig.3. Alignmentofthededuced aasequencesofthe HupC (A),HupD (B) and HupF (C) proteins from B. japonicum(Bj) with the homologous

products (see section a) from R. capsulatus (Rc), 4. uinelandii (Av), E. co/i (EC), R. Iegguminosarun~ (RI) and A. eutrophus (Ae). Consensus sequence

(Consen)

is shown

at the bottom.

created as described previously by Fu and Maier, 1993b). In this mutant strain a 3.2-kb BumHI fragment containing the promoter-regulatory region and hupSL genes has been replaced by a SpR cassette (Fu and Maier, 1993b, see Fig. 1). Plasmid PRY 12 (Novak and Maier, 1989) harboring all of hupSL, and pLD22 harboring the entire kupSLCDF genes (Fig. 1) were conjugated into the mutant strain JHCS2, and Hup activity was determined after derepression. Clearly, pRY12 did not complement methylene blue (MB)-dependent Hup activity of JHCS2, whereas pLD22 restored MB-dependent activity of the mutant (about 70% of the wt level) (Table I). The results

Fig. 4 (Top). Western blots of cell-free extracts from different strains probed with antibody towards the 65-kDa Hup subunit (Struts et al., 1986a). Lanes (10 uM protein/lane): 1,JH (wt); 2, JHCS2 (mutant); 3. JHCS2[pRYl2]; 4, JHCS2[pLD22]. Western blots were carried out (0.1% SDS-lo% PAGE) as described previously (Fu and Maier, 1993a). Fig. 5 (Bottom).

Expression

of hupCDF.

Autoradiogram

of [“‘S]-

methionine and [35S]cysteine (20 uCi, ICN Biomedicals, Irvine, CA, USA) labelled polypeptides separated by 0.1% SDS-15% PAGE from cell extracts of E. co/i BL21(DE3) (Studier and Moffatt, 1986) containing plasmid pBluescript II KS(+) or pLDlO0 harboring the hupCDF genes. Lanes: 1 and 3, pBluescript II KS(+); 2 and 4, pLD100. Cells treated with 200 ug Rif/ml (lanes 1 and 2) for 60 min or 90 min (lanes 3 and 4). The positions and molecular weights of standard size markers are shown in the left margin. The expressed products of HupC and HupD are indicated.

indicate that the kupCDF gene products are indeed transcribed from the same promoter as kupSL. The data here also indicate that the products of the kupCDF genes are required for Hup activity, stabilization or synthesis. This is further addressed below. Since MB receives electrons directly from Hup (Arp, 1985), one might expect MB-dependent Hup activity in

51 TABLE

I

Methylene

pBluescript blue (MB)-dependent

Hup activity

Strain [ plasmid]”

Hup activityb

JH (wt) JHCS2 @up-)

25.0

in cell-free extracts*

proteins

1976). Tetracycline mutant containing washed once and

in modified

Bergersen

medium

(Bishop

et al.,

(75 ug/ml) was included in the medium for the the plasmids. Cells were harvested at log phase, resuspended in carbon-free de-repression medium

(Stults et al., 1986b; Fu and Maier, 1993a). Cells were then incubated for 20 h in the presence of 5 uM NiCI, to create derepression conditions for Hup activity, and cell-free extracts were prepared as previously described (Fu and Maier, 1993a). bMB-dependent

hydrogenase

activity

was determined

This aligns the trangenes in the same

of pBluescript

II KS(+).

pLDlO0 (Fig. 1) was introduced (Studier and Moffatt, 1986). Two

of 28.5 and 25 kDa

BL21(DE3)

17.4

were grown

as the T7 promoter

The resulting plasmid into E. coli B2l(DE3)

products “Bacteria

(Stratagene).

of the hupCDF

direction

direction

0 0.4

JHCS2[pRY12] JHCS2[pLD22]

II KS(+)

scriptional

(Fig. 5)

were produced

corresponding

to

in E. coli

the

deduced

of hupC and hupD (28 and 21 kDa, respectively),

although the products are slightly larger than predicted. The product of hupF was not produced. Such specific lack of expression from other

noted

bacteria

1993) was probably

for other H,-metabolism

(Kortltike

genes

et al., 1992; Rey et al.,

due to the lack of a good RBS.

amperometri-

tally in 50 mM potassium phosphate/2.5 mM MgCI, buffer (pH 6.2) as described (Stults et al., 1986; Ferber and Maier, 1992). Data are the

ACKNOWLEDGEMENTS

average

We thank Dr. Tomas Ruiz-Argtieso for helpful suggestions on protein expression and Tom Ng for help with

of duplicates

and expressed

as nmol

H, oxidized/min

per mg

protein.

the mutant strain JHCS2 containing pRY12. However, cell-free extracts of JHCS2[pRY12] did not have significant Http activity (Table I). To establish whether this is due to the lack of the Hup protein per se or due to inactivity of the Hup apoprotein, Western blots were done on cell-free extracts of wt, mutant and the mutant containing plasmid pRY12 and pLD22 (Fig. 4). As expected, mutant strain JHCS2 did not produce any Hup while JHCS2 [ pLD22] produced a substantial amount of Hup protein (Fig. 4). Most notably, JHCS2[pRY12] produced a negligible amount of Hup protein (Fig. 4). These data indicate that the lack of Hup activity from JHCS2[pRY12] is due to lack of the hydrogenase protein. Previous BGal assays from a hup promoter-&Z transcriptional fusion plasmid in a similar hup- mutant strain JH47 (containing a Tn5 insertion in the hupS gene) indicated that the promoter for the hup operon is still active in the absence of the products of the hup operon (Kim and Maier, 1990). Therefore, the products (or one product) of hupCDF genes may play roles in stabilizing Hup proteins (so they are not degraded) or in posttranscriptional regulation of Hup. It is also possible that some of these gene products together with the small and large subunits of Hup form a complex. However, active Hup containing only two subunits can certainly be purified from the wt B. japonicum (Arp, 1985; Stults et al. 1986a; Ferber and Maier, 1992). In-frame deletions in each ORF will be necessary to understand the specific roles of each gene product. (c) Expression of hupCDF in E. coli A 1955-bp SacI-Hind111 fragment harboring hupCDF genes was cloned between the Sac1 and Hind111 sites of

computer analysis. This work number DE-FG02-89ER14011 Energy.

was supported by grant from the Department of

REFERENCES Arp,

D.J.: Rhizobium

japonicum

hydrogenase:

purification

to homo-

geneity from soybean nodules and molecular characterization. Biochem. Biophys. 237 (1985) 504-512. Bishop, P.E., Guevarra, J.G., Engelke, J.A. and Evans, Relationship between glutamine ties in the symbiotic associations

Arch. H.J.:

synthetase and nitrogenase activibetween Rhizobium juponicum and

GIycine max. Plant Physiol. 57 (1976) 542-546. Cauvin, B., Colbeau, A. and Vignais, P.M.: The hydrogenase

structural

operon in Rhodobacter capsulatus contains a third gene, hupM, necessary for the formation of a physiologically competent hydrogenase. Mol. Microbial. 5 (1991) 251992527. Colbeau, A., Richaud, P., Toussaint, B., Caballero, F.J., Elster, C., Delphin, C., Smith, R.L., Chabert, J. and Vignais, P.M.: Organization of the genes necessary for hydrogenase expression in Rhodobacter capsulatus. Sequence analysis and identification of two hyp regulatory mutants. Mol. Microbial. 8 (1993) 15-29. Dross, G., Geisler, V., Lenger, R., Theis, F., Krafft, T., Fahrenholz, F., Kojro, E., Duchene, A., Tripier, D., Juvenal, K. and Krtiger. A.: The quinone-reactive Ni/Fe-hydrogenase of Wolinella succinogenes. Eur. J. Biochem. 206 (1992) 93-102. Ferber, D.M. and Maier, R.J.: Incorporation of a bacterial membranebound hydrogenase into proteoliposomes. Anal. Biochem. 203 (1992) 2355244. Fu, C. and Maier, R.J.: Identification of a locus within the hydrogenase gene cluster involved in intracellular nickel metabolism in Bradyrhizobium japonicum. Appl. Environ. Microbial. 57 (1991) 350223510. Fu, C. and Maier, R.J.: A genetic region downstream from the hydrogenase structural genes of Bradyrhizobium japonicum that is required for hydrogenase processing. J. Bacterial. 175 (1993a) 295-298. Fu, C. and Maier, R.J.: Rapid and efficient selection of site-directed recombinant mutants from Bradyrhizobium japonicum by colony hybridization. FEMS Microbial. Lett. 19 (1993b) 33338.

52 Hidalgo, E., Palacios, J.M., Murillo, J. and Ruiz-Argtieso, T.: Nucleotide sequence and characterization of four additional genes in the hydro-

Moshiri, F. and Maier, R.J.: Conformational changes in the membranebound hydrogenase of Brudyrhizobium japonicum. J. Biol. Chem. 263

genase structural operon from Rhizobium leguminosrrrum bv. viciae. J. Bacterial. 174 (1992) 4130-4139. Horn, S.S.M., Graham, L.A. and Maier, R.J.: Isolation of genes (n$/flhup

(1988) 17809-17816. Novak, P. and Maier, R.J.: Identification

cosmids)

involved

Rhizobium

japonicum.

in hydrogenase and nitrogenase activities J. Bacterial. 161 (1985) 8822887.

in

Kim, H. and Maier, R.J.: Transcriptional regulation of hydrogenase synthesis by nickel in Bradyrhizobium juponicum. J. Biol. Chem. 265 (1990) 18729-18732. Knauf,

V.C. and Nester, E.W.: Wide host range vectors: a cosmid clone

bank of an Agrohacterium Ti plasmid. Plasmid 8 (1982) 45554. Kortltike, C., Horstmann, K., Schwartz, E., Rohde, M., Binsack, R. and Friedrich, B.: A gene complex coding for the membrane-bound hydrogenase of Alcaligenes eutrophus H 16. J. Bacterial. 174 ( 1992) 6277-6289. Kyte, J. and Doolitle, R.F.: A simple method for displaying the hydropathic Menon, and

character

of a protein.

A., Mortenson, genetic

J. Mol. Biol. 157 (1982) 1055132.

L.E. and Robson,

R.L.: Nucleotide

sequences

of hydrogen oxidation (ho.u) genes in J. Bacterial. 174 (1992) 454994557. Menon, A., Stults, L.W., Robson, R.L. and Mortenson, L.E.: Cloning. sequencing and characterization of the [NiFelhydrogenaseAzotobac~tw

analysis

uinelundii.

encoding

structural genes (ho?tK and hoxG) from Azotobucter Gene 96 (1990a) 67774. N.K., Robbins, J., Peck Jr., H.D., Chatelus, C.Y., Choi, E.S.

tkelandii.

Menon,

and Przybyla, A.E.: Cloning and sequencing of putative Escherichiu [NiFelhydrogenase-1 operon containing six open reading frames. J. Bacterial. I72 (1990b) 196991977. co/i

Menon, N.K.. Robbins, J., Wendt, J.C., Shanmugam, K.T. and Przybyla, A.E.: Mutational analysis and characterization of the Escherichiu co/i hJ’u operon, which encodes [NiFe] hydrogenase 1. J. Bacterial. 173 (1991) 4851-4861.

of a locus upstream

from the

hydrogenase structural genes that is involved in hydrogenase expression in Brudyrhizohium ,japonicum. Appl. Environ. Microbial. 55 (1989) 3051~3057. Rey, L., Murillo. J., Hernando, Y., Hidalgo, E., Cabrera, E., Imperial, J. and Ruiz-Argueso, T.: Molecular analysis of a microaerobically induced operon required for hydrogenase synthesis in Rhizohium leguminosarum biovar viciae. Mol. Microbial. 8 ( 1993) 47 I-48 1. Richaud, P., Vignais, P.M., Colbeau, A., Uffen, R.L. and Cauvin, B.: Molecular biology studies of the uptake hydrogenase of Rhodobacter and Rhodocyclus (1990) 413-418. cupsulutus

Sayavedra-Soto, rinelundii

FEMS

Microbial.

Rev. 87

L.A. and Arp, D.J.: The hoxZ gene of Azotobwtrr hydrogenase operon is required for activation of hydro-

genase. J. Bacterial. Sayavedra-Soto, Nucleotide

gelatinosus.

174

( 1992) 529555301.

L.A., Powell, G.K., Evans, H.J. and Morris, R.O.: sequence of the genetic loci encoding subunits of

japonicum uptake hydrogenase. Proc. Natl. Acad. Sci. USA 85 (1988) 8395-8399. Studier, F.W. and Moffatt, B.A.: Use of bacteriophage T7 RNA polymerase to direct selective high-level expression of cloned genes. J. Mol. Biol. 189 (1986) 113~130. St&s, L.W., Moshiri, F. and Maier. R.J.: Aerobic purification of hydroBradyrhizobium

genase from Rhizobium juponicum by affinity chromatography. J. Bacterial. 166 (1986a)7955800. Stults, L., Sray, W.A. and Maier, R.J.: Regulation of hydrogenase biosynthesis 280-283.

by nickel in B. jclporzieunr. Arch.

Microbial.

146 (1986b)