Microbiology (1 999),145,743-753
Printed in Great Britain
Organization and expression of nitrogen=
fixation genes in the aerobic nitrogen-fixing
uniceIIuIar cyanobacter ium Synechococcus sp.
strain RF=l
Tan-Chi Huang,' Rong-Fong Lint1Mu-Kuei Chul and Horng-Ming Chen2
Author for correspondence: Tan-Chi Huang. Tel.
e-mail : [email protected]
1
Institute of Botany,
Academia Sinica, Nankang,
Taipei, Taiwan 115,
Republic of China
National Narcotics Bureau,
Department of Health,
Executive Yuan, Taiwan
115, Republic of China
+ 886 2 2782 9590. Fax: + 886 2 2782 7954.
Sixteen nif and 'nif-associated genes (expressed only under conditions of
nitrogen fixation) in Synechococcus sp. strain RF-1 have been cloned and
sequenced. All of the nif and nif-associated genes identified in Synechococcus
RF-1 were arranged in a continuous cluster spanning approximately 18 kb and
containing seven operons. The nifH operon (nifH-nifD-nifK)has been reported
previously. nifB, fdXN, nifS8 nifU and nifP were found to be located upstream
of the nifH operon. nifB-WxN-nifS-nifU were expressed as an operon. A nifPlike gene was found t o be located just upstream of nifB. nifE, nifN, nifX, nifW
and the nif-associated hesA, hesB and 'fdx' were found to be located
downstream from nifK. The genes located downstream from nifK are arranged
nifE-nifN-nifX-orf-nifW-hesA-hesB-'fdx' and span approximately 7 kb. The
function of the ORF situated between nifX and niNV is not known. However, it
was identified as a counterpart of ORF-2 in Anabaena sp. strain PCC 7120 based
on the deduced amino acid sequence. Northern hybridization and primer
extension analysis indicated that the nif and nif-associated genes are
organized in nifE-nifN, nifx-orf, niW-hesA-hesBand fdx'-containing operons,
respectively. According to the results of this study and previous reports, the
genes are expressed in a rhythmic pattern with peaks during the dark phase
when the culture is grown in a 12 h IighVl2 h dark regimen. The rhythm
persisted after the culture was transferred to continuous illumination.
Keywords : cyanobacteria, nitrogen fixation, gene organization, rhythmic expression,
Synechococcus RF-1
INTRODUCTION
Cyanobacteria may contribute significant nitrogen into
the global nitrogen cycle since they are widely distributed in different environments, especially in paddy
rice fields and marine environments. Morphologically,
cyanobacteria can be divided into unicellular, nonheterocystous filamentous and heterocyst-forming
types. Nitrogen fixation is carried out by all of the
heterocyst-forming type, while only a few species of
unicellular and non-heterocystous filamentous types can
fix nitrogen aerobically. T o protect the nitrogenase from
oxygen inhibition, about 5-10% of cells of the heterocyst-forming type differentiate into specialized nitrogenThe GenBank accession numbers for the sequences reported in this paper
are AF001780 (upstream of nifH) and AF003700 (downstream from nifK)
0002-2703 0 1999 SGM
fixing cells. The correlation between heterocyst differentiation and structure of nif genes has been intensively
studied in Anabaena sp. strain PCC 7120 (Haselkorn &
Buikema, 1992). Species of the non-heterocystous filamentous type usually fix nitrogen only under low
ambient oxygenic pressure (Fay, 1992). The unicellular
nitrogen-fixing cyanobacteria can fix nitrogen and
carbon dioxide within the same cell. These microorganisms have developed a unique strategy to protect
nitrogenase from oxygen inactivation by temporal
separation of nitrogen fixation and photosynthesis.
Structural analysis of the nif genes is important for
understanding nitrogen fixation at the biochemical level
and for revealing the genetic relationships among
different diazotrophs (Postgate & Eady, 1988). About 20
genes identified in Klebsiella pneumoniae were found to
be involved in nitrogen-fixing activity. Characterization
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sun, 18 Jun 2017 01:13:53
743
T.-C. H U A N G a n d OTHERS
of the structural genes ( n i p , n i p and niflo for
nitrogenase has been performed in various diazotrophs.
However, the structure and organization of nitrogenfixation-related genes has been documented only in
some diazotrophs such as Klebsiella pneumonia,.
(Arnold et al., 1988), Azotobacter uinelandii (Jacobson
et al., 1989),Bradyrhixobium japonicurn (Ebeling et al.,
1987; Fischer et al., 1988; Noti et al., 1986),
Rhodobacter capsulatus (Klipp, 1990), Clostridium
pasteurianurn (Chen et al., 1986; Wang et al., 1989) and
Anabaena PCC 7120 (Haselkorn & Buikema, 1992).
The ability of nitrogen-fixing unicellular cyanobacteria
to protect nitrogenase by temporal separation is a unique
character among the diazotrophs. However, information concerning the structure of nif genes in
unicellular cyanobacteria is very limited (Kallas et al..,
1983, 1985).
The nitrogen-fixing unicellular cyanobacteria can be
grouped into sheath and sheathless forms (Waterbury &
Rippka, 1989). Studies on nitrogen-fixing activity in the
sheath form have been carried out in Gloeocapsa sp.
strain LB795 and Gloeothece sp. strain ATCC 27152
(Gallon et al., 1988), while those in the sheathless form
have been investigated in Synechococcus sp. strain RF-1
(Huang & Grobbelaar, 1995), Synechococcus BG43511
(Mitsui et al., 1986) or Cyanothece sp. strain ATCC
51142 (Colon-Lopez et al., 1997). In the unicellular
cyanobacterium Synechococcus sp. strain RF-1, the
nitrogen-fixing activity in a diurnal light/dark-entrained
culture was found to exhibit a circadian rhythm (Huang
& Grobbelaar, 1995). To understand the control mechanism of rhythmic nitrogen fixation, it is important to
characterize the structure of the nifgenes. Therefore, the
nif genes in Synechococcus sp. strain RF-1 were cloned
and sequenced. The nucleotide sequence of the nifHDK
operon has been reported previously (Chen et al., 1996).
In this report, the nucleotide sequence and expression of
n i p , n i p , fdxN, nifS, n i p , nifE, n i m , nifX, n i p and
some ' nif-associated ' genes (expressed only under conditions of nitrogen fixation) are presented. In addition,
the organization of all nif and nif-associated genes
cloned from Synechococcus sp. strain RF-1 is reviewed
and discussed.
METHODS
Organisms and cultivation. Axenic culture of Synechococcus
sp. strain RF-1 (PCC 8801) was grown in BG-11 or BG-110
medium under the conditions described previously (Huang &
Chow, 1988). Cultures were grown in light (about 35 pmol
m-2s-l) from white fluorescent tubes at 28 0-5 "C. They were
incubated either in continuous light (L/L) or in a diurnal
light/dark (L/D) regimen, depending on the requirement of
the experiment.
DNA isolation and construction of a genomic library.
Genomic DNA in Synechococcus sp. strain RF-1 was extracted
as described previously (Chen et al., 1996). The genomic DNA
was partially digested with Sau3A and DNA fragments with
sizes ranging from 10 to 23 kb were isolated from lowmelting-point agarose gels (Sambrook et al., 1989). The sizefractionated fragments were ligated into BamHI-digested
744
EMBL4, according to the manufacturer's procedures
(Stratagene).
Cloning of nif genes located upstreamand downstream from
nifHDK operon. The genomic library of Synechococcus sp.
strain RF-1 prepared in EMBL4 was used for screening the nif
genes by the standard plaque hybridization technique
(Sambrook et al., 1989). It is known that nifU in Synechococcus sp. strain RF-1 is located upstream of nifH, based
on our previous report (Chen et al., 1996). Therefore, a
synthetic oligonucleotide complementary to part of nifU was
used as a probe for screening the nifgenes located upstream of
the nifHDK operon. From our previous results (Chen et al.,
1996), one subclone, NF-3, is known to contain nifK and some
other genes. Subsequent sequencing revealed that nifE was
located immediately downstream from nifrc. Therefore, a
synthetic oligonucleotide complementary to part of nifE was
used as a probe to screen DNA fragments from the genomic
library. The probe was end-labelled with digoxigenin (DIG)
and plaque hybridization was carried out using the DIG
Luminescent Detection Kit according to the manufacturer's
instructions (Boehringer Mannheim). Positive plaques were
selected and subcloned into pGEM-7Zf( - ) for sequencing.
Nucleotide sequencing and comparison. The nucleotide
sequence of both strands of the plasmid DNA was determined
by the dideoxy chain-termination method on an automated
DNA sequencer (Pharmacia),according to the manufacturer's
procedures. Sequencing of dsDNA was first carried out with
universal sequencing primers using the Auto Read Sequencing
Kit (Pharmacia), following the supplier's instructions. Other
oligonucleotide primers were synthesized and used to complete sequencing of both strands. The extent of identity of the
deduced amino acid sequences of nif genes in Synechococcus
sp. strain RF-1 with the products encoded by corresponding
genes from other diazotrophs was searched for by using the
PCGENE program.
Isolation of total RNA. Thirty millilitres of exponentially
growing cells (about 2 x lo7 cells ml-l) of Synechococcus sp.
strain RF-1 were collected by centrifugation, washed twice
with distilled water and then collected in a 1.5 ml screw-top
vial. Teflon-coated sea-sand (0.2 g, 0-1-0.3 mm diameter;
Merck), 0.5 ml solution D (4 M guanidine thiocyanate, 25 mM
sodium citrate, 0.5 '/o laurylsarcosine and 0.72 '/o P-mercaptoethanol), 0.5 ml phenol/chloroform (4: 1, v/v) and 0.1 ml
2 M sodium acetate (pH 4.0) were then added. The cells were
broken by vibration with a Mini-Beadbeater (Biospec) set at
5000 r.p.m. for 120 s. The broken cells were centrifuged at
13000 g at 4 "C for 5 min. The supernatant was re-extracted
twice with phenol/chloroform (1:1).An equal volume of icecold absolute alcohol was added to the supernatant. Total
RNA was collected after freezing at -70 "C for more than 1 h
and then resuspended in an appropriate volume of diethyl
pyrocarbonate (DEPC)-pretreated water.
Primer extension analysis. The primer extension analysis was
performed as described previously (Chen et al., 1996). About
30pg total RNA prepared from Synechococcus sp. strain
RF-1 was used for each primer extension analysis.
Northern hybridization. About 5 pg total RNA was added
into 20 pl RNA loading buffer (50% formamide, 6.5%
formaldehyde, 2.5 pg ethidium bromide ml-', 1 x MOPS),
denatured at 65 "C for 10 min, cooled in ice and then
fractionated by 1.2 '/o agarose gel electrophoresis (containing
3.7% formaldehyde). The RNA gel was immersed in 1 x
phosphate buffer (50 mM NaH,PO,, 50 mM NaOH, 5 mM
EDTA, pH 6.5) for 30 min and then blotted onto Genescreen
Plus membrane (NEN)with 20 x SSPE overnight (1x SSPE is
0.15 M NaCl, 0-01 M NaH,PO,. H,O, 1 mM EDTA, pH 7.4).
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sun, 18 Jun 2017 01:13:53
Organization and expression of nitrogen-fixation genes
The membrane was prehybridized in 15 m1/10 x 10 cm' hybridization solution [0*2x SSPE, 0.5o/' laurylsarcosine, 1o/'
SDS, 1 o/' blocking reagent (Boehringer Mannheim)] at 65 "C
for a t least 3 h. Hybridization was carried out at 65 "C
overnight in 5 ml hybridization solution by adding about
25 ng denatured DIG-labelled DNA probe. The membrane
was washed twice with solution containing 0.1 x SSPE and
0.1 '/" SDS at 65 "C for 10 min. Detection of the DIGchemiluminescent reaction was carried out as described by the
manufacturer (Boehringer Mannheim).
of nifu. Ten positive clones exhibiting strong hybridization signals were isolated. The phage DNA was
restricted with EcoRI, HindIII, or ApaLI and then
subcloned into pGEM-7Zf ( - ). Four independent subclones, NF-4, NF-5, NF-6 and NF-7, were selected (Fig.
1) and both strands of DNA from each subclone were
sequenced. Five ORFs in a 6.6 kb fragment were
identified after searching for all three possible reading
frames with ATG as the initiation codon.
RESULTS
Identification of the nifB operon by the deduced amino acid
sequence. Among the five ORFs, the reading direction of
the first is the reverse of the other four. Therefore, it is
characterized in another section below. The second
ORF beginning from position 1961 consists of 1677 bp.
It encodes a polypeptide of 558 aa. The predicted amino
acid sequence exhibits a high degree of identity (73.89 %)
to that of the polypeptide encoded by nip from
Anabaena PCC 7120 (Mulligan & Haselkorn, 1989).
The genes located upstream of the nifH operon
Cloning and sequencing. About 1 x lo4 plaques from the
Synechococcus RF- 1 genomic library were screened
using a synthetic oligonucleotide complementary to part
Hindlll EcoRl EcoRl
EcoRl
1
Hindlll ApaLl
Hindlll
The third ORF (position 3652-4014) is present 15 bp
downstream from the termination codon of n i p . It
encodes a polypeptide of 120 aa. The predicted amino
acid sequence shows 43.1% identity to the product of
fdxN from Anabaena PCC 7120. Although the similarity
of the amino acid sequence is not very high, it contains
two cysteine motifs, C-X,-C-X,-C-X,-CP at the N
terminus and C-X,-C-X,-C-X,-CP at the C terminus,
which were found in the nif-associated ferredoxins (Fig.
2) (Mulligan et al., 1988).
1
Fig. 1. Schematic representation of a 6.6 kb fragment of the
Synechococcus RF- 1 chromosome containing the nifB- fdxN-nifSnifU operon and nifP located upstream of nifB. The location of
the DNA fragments from the subcloned recombinants (NF-4, NF5, NF-6 and NF-7) used for sequence analysis is shown by white
boxes. The restriction sites used for cloning are indicated by
arrows.
The fourth ORF (position 4119-5321) is present 104 bp
downstream from the termination codon of fdxN. The
deduced amino acid sequence has 74-25% identity with
the polypeptide encoded by nifS from Anabaena PCC
~
* * *
**
*
* *
**
RF- 1 MSYTITSDCIACDRCRVQCPTNAIRI-ESGVLLIDPTQCNHCVGSYGT-PQCAAMCPTNF 58
I I I I I I II I I
I l l II I I I
II
I
II
I Ill I Ill
7120 MAYTITSQCfSCKLCSmCPTGAIKIAENGQHWfDSELCTNCVDTWTVPQCKAGCPTCD 60
RF- 1 GCLPDNSCFQEVDLQLTDYIRYQSLVl'GLKTKGHHKYWQH"DWSQK----LA
114
II
psDywEGwFA"RvI~~TKK-QDYWERWFNCYSQKFS
Ill II I
1 II
II I I I I I I
I
107
7120
-----------
120
RF- 1 TLFSCS
7120 KXQGEILGV
116
__
.....................................................................................................
Fig. 2. Comparison of the deduced amino
acid sequence of fdxN in Synechococcus RF-1
with that of Anabaena PCC 7120. The
cysteine residues probably involved in
C4Fe-451clusters are indicated by asterisks.
Table I . Comparison of nif genes in Synechococcus RF-1 with their corresponding
counterparts in other diazotrophs
nijR
nifS
nifU
nifE
nifN
nifX
nifW
I
Amino acid identity (%)
Gene
Anabaena PCC 7120
Klebsiella pneumoniae
Azotobactev vinelandii
73.9
74-3
64.0
75.6
59.5
59.7
59.0
44.2
56.0
45.4
53.6
36.7
20.7
16.3
5 1.3
55.4
49.2
55.4
44.8
41.5
34.2
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sun, 18 Jun 2017 01:13:53
T.-C. H U A N G a n d OTHERS
Fig. 3. Primer extension analysis of nifB (a) and nifP (b). Lanes
1, product of primer extension when an oligonucleotide was
annealed with total RNA and extended as described in
Methods.
Fig. 5. Expression of the nifP gene examined by Northern
hybridization. (a) Rhythmic expression of the nifP gene.
Samples for RNA isolation were taken at 4 h intervals (as
indicated above each lane) after the 12 h U12 h D-entrained
culture was transferred t o UL. An equal amount of RNA (about
5pg) was loaded on each lane. (b) Transcription of the nifP
gene with (BG-11 medium) or without (BG-110 medium)
combined nitrogen.
.
., ..,.,.,.........,............. .., ,,,
..........,,,,............................................................................................
Fig. 4. Expression of the nifB operon examined by Northern
analysis. Samples were taken a t 3 h intervals (as indicated
above each lane) after the 12 h U12 h D-entrained Synechococcus RF-1 culture was transferred t o UL. An equal amount of
RNA (10 pg) was loaded on each lane. (a) Pattern of RNA-DNA
hybridization with the nifB probe. (b) Level of rRNAs (165 and
235) from the same experiment (used as internal standard).
7120 (Mulligan & Haselkorn, 1989).The deduced amino
acid sequence of the fifth ORF (position 5486-6370) has
63.95% identity with the polypeptide encoded by nifU
from Anabaena PCC 7120 (Mulligan & Haselkorn,
1989). Therefore, we conclude that the four ORFs [2,3,
4 and 5 in AF001780 (GenBank)] in Synechococcus RF1 correspond to n i p , fdxN, nifS and nifU in Anabaena
PCC 7120, respectively. A comparison of n i p , nifS and
n i p with the corresponding nif genes in other diazotrophs is shown in Table 1.
Transcription start site and the expression of the nifB
operon. Primer extension analysis was used to locate the
start site of transcription. As shown in Fig. 3(a), a single
transcript for n i p was detected. The transcription start
746
site is nucleotide ‘A’ which lies 206 bp upstream of the
start codon of the n i p gene. Expression of the n i p
operon was examined by Northern hybridization. As
shown in Fig. 4, four bands with sizes of 1-8,2.2,3.5 and
4.6 kb were detected. Based on molecular size, the four
bands correspond to the mRNAs of n i p , nip-fdxN,
nip-fdxN-nifS and nip-fdxN-nifS-nifU, respectively.
The mRNA level of the n i p operon changed in a 24 h
rhythmic pattern after the 12 h L/12 h D-entrained
culture was transferred to L/L, with the active synthesis
period at about 14-17 h. Since the time periods from 0 to
12 h and from 12 to 24 h correspond to the L and D
phases, respectively, before the transfer to constant
illumination, the expression of n i p peaked early in the
dark phase and was very weak during the light phase
under the diurnal L/D regimen. The expression of the
nifHDK operon in Synechococcus RF-1 also exhibits a
circadian rhythm, with the mRNA occurring exclusively
during the dark phase when the culture grows in a
diurnal L/D regimen (Huang & Chow, 1990). Results in
Fig. 4 indicated that the expression phase of the n i p
operon was the same as that of the nifHDK operon.
The gene located upstream of nifB. A complete ORF with
a reverse reading direction to the n i p operon was found
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sun, 18 Jun 2017 01:13:53
Organization and expression of nitrogen-fixation genes
~-
.......................................................
-70
-60
-20
-40
-1
TTAATAAA~CAATAAA~CTA~C~~GATAC.AAACn;TAATAGTCAAAGACGACTGCTATAAATCTTA..7120
IIIIIIIIIIIIIIII
I
Ill
I
I
I I
II
I l l 1 I
IIIIIIII
I
GTAATAAATCCAATCAATCTOAACTCAGOCAOCCTCA~CA~C~TCATCAACCC~AGA~~TAT~CTTGTCAT
RF-1
-i o
-60
-40
-20
-1
upstream of nifB. The inverted ORF (position 832-95)
encodes a polypeptide of 245 aa. Based on the similarity
to reported protein sequences, the deduced amino acid
sequence shows 31.43% identity with the product of
cysE from Escherichia coli, which encodes a serine
acetyltransferase (Denk & Rock, 1987). The ORF was
temporarily assigned as a cysE-like gene (‘cysE’). The
results of primer extension analysis indicated that the
transcription start site of ‘cysE’ is nucleotide ‘A ’ located
30 bp upstream of the start codon (Fig. 3).
An ORF located immediately downstream from n i p in
Azotobacter chroococcum has been identified and designated n i p (Evans et al., 1991). It is 41 % identical to the
product of cysE from E. coli. The n i p gene has been
suggested to encode a nif-specific serine acetyltransferase
required for optimizing expression of nitrogenase activity. The inverted ‘cysE’ in Synechococcus RF-1 is
39.5 % identical to n i p of Azotobacter chroococcum.
According to Northern hybridization (Fig. 5),the ‘ cysE’
gene shows expression only under nitrogen-fixing conditions. Based on its similarity to the Azotobacter gene
and its expression limited to nitrogen-fixing conditions,
the ‘cysE’ gene was designated n i p . Fig. 5 also shows
that the transcription activity of nifP exhibits the same
circadian rhythm as nifB (Fig. 4) in Synechococcus
RF-I.
Comparison of nif promoter sequences. The nucleotide
sequences of the presumptive nifB and n i p promoters
in Synechococcus RF-1 were compared with the promoter regions of the counterparts in Anabaena PCC
7120 reported by Mulligan & Haselkorn (1989). N o
significant similarity was found for n i p . However, two
regions containing a number of identical base pairs were
found in nifB (Fig. 6). One is located about 10 bp
upstream of the transcription start site and the other is
located from position -61 to -73.
The genes located downstream from nifK
Cloning and sequencing. Synthetic oligonucleotides com-
plementary to parts of nifK and nifE were used to screen
DNA fragments containing nif genes from the genomic
library cloned in EMBL4. The positive clones were
isolated and then subcloned into pGEM-7Zf( -). As
shown in Fig. 7, four independent subclones, NF-3, NF8, NF-9 and NF-10, were selected for DNA analysis. The
nucleotide sequence covering a region of about 7 kb
located downstream from nifK was revealed by subsequent sequencing of the four subclones. Nine ORFs
Xbal
..............................................
Fig- 6. Comparison of nifB promoter
sequences in Anabaena PCC 7120 and in
Synechococcus RF-1. Identities are marked
with solid bars between sequences. A gap
was introduced in the nifB sequence of
71
to improve the
alignment.
Hindlll Xbal
BspHl Hindlll Dral BspHl
Nk-9
Dral
NF-10
Fig. 7. Schematic representation of a 7 kb fragment of the
Synechococcus RF-1 chromosome containing the nif and nifassociated genes downstream from nifK. The ORF regions are
shown as black boxes. The location of the DNA fragments from
the subcloned recombinants (NF-3, NF-8, NF-9 and NF-10) used
for sequence analysis is shown by white boxes. The restriction
sites used for cloning are indicated by arrows.
reading in the same direction were found based on
searching for all three possible translation frames with
ATG as the initiation codon.
Identification of nif genes by the deduced amino acid
sequences. The nucleotide sequences for the nine possible
ORFs along with their 5’ flanking regions were identified. The first ORF (position 328-1737) encodes a
polypeptide of 469 aa. The deduced amino acid sequence
exhibits a high degree of identity (75.6%) to the
polypeptide encoded by nifE from Anabaena PCC 7120.
The second ORF (position 1806-3191) encodes a polypeptide of 461 aa. The deduced amino acid sequence
shows a 59.5% identity with the product of n i p from
Anabaena PCC 7120. The third ORF (position
3314-3721) encodes a polypeptide of 135 aa. The
deduced amino acid sequence shows a 59.7% identity
with the product of n i p from Anabaena PCC 7120.
These results indicate that the three ORFs in Synechococcus RF-1 are nifE-, n i p - and nifX-like, respectively.
Based on data submitted recently to GenBank by
Buikema, Scapino & Haselkorn (U47055), two functionally unidentified genes, designated ORF-2 and ORF-1,
are located downstream from nifX in Anabaena PCC
7120. Both ORF-2 and ORF-1 in Anabaena PCC 7120
encode polypeptides of 158 and 71 aa, respectively. The
amino acid sequences deduced from the fourth (position
38214291) and fifth (position 4404-4634) ORFs downstream from nifK in Synechococcus RF-1 are 57.6 and
67.2 YO identical to the products of ORF-2 and ORF-1 in
Anabaena PCC 7120, respectively. Therefore, the fourth
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sun, 18 Jun 2017 01:13:53
747
T.-C. H U A N G a n d OTHERS
and fifth ORFs in SynechococcusRF-1 were identified as
the counterparts of ORF-2 and ORF-1, respectively, in
Anabaena PCC 7120.
The amino acid sequence deduced from the sixth ORF
(position 4636-4971) has 59 Yo identity with the product
of nifw from Anabaena PCC 7120. The amino acid
sequences deduced from the seventh (position
4980-5546) and eighth (position 5875-6234) ORFs show
75.3 and 58.4% identity with the products of hesA
(U4705.5) and hesB (Borthakur et al., 1990) from
Anabaena PCC 7120, respectively.
The ninth ORF (position 6273-6617) encodes a polypeptide of 114 aa. This protein is identified as a
ferredoxin and the gene was designated ‘fdx ’ because of
the presence of the characteristic cysteine cluster :
CX,CX,C. The deduced amino acid sequence is 47%
identical to ferredoxin 1 (PetF) from the vegetative cells
of Anabaena PCC 7120 (Alam et al., 1986). It is also
47 % identical to the nitrogen-fixation-specific ferredoxin (FdxH) from Anabuena PCC 7120 (Bohme &
Haselkorn, 1988). Further studies are needed to identify
a specific role for the product of ‘fdx ’ in Synechococcus
RF-1.
A comparison of nif genes in Synechococcus RF-1 with
the corresponding counterparts of Anabaena, Klebsiella
and Azotobacter is shown in Table 1.
Gene expression examined by Northern hybridization. DIG-
labelled DNA probes specific to nifE, n i m , nifX, orf2,
orfl, nifw, hesA, hesB and ‘ f d x ’ , prepared by PCR,
were used to investigate gene expression. Samples for
total RNA extraction were taken at 4 h intervals for a
period of 46 h after the Synechococcus RF-1 culture
grown in a 12 h L/12 h D diurnal regimen was transferred to continuous light. The results of Northern
hybridization indicated that nifE, n i m , nifX, orf2, orfl,
nifw, hesA, hesB and ‘fdx’ were all expressed in a
circadian rhythmic pattern (Fig. 8). The transcripts of
these genes were detected mainly during the dark periods
when grown in a diurnal L/D regimen and the rhythms
persisted after the culture was transferred to continuous
illumination.
As shown in Fig. 8(a) and (b),the signals detected by the
DNA probe specific to nifE or n i m were smeared and
the sizes were smaller than predicted. This may have
been caused by the degradation of mRNA during
isolation or poor specificity of the DNA probe. However, defined mRNA bands were detected when the same
RNA preparation was characterized with other DNA
probes (Fig. 8)’ so the smeared signals were not
necessarily due to the breakdown of mRNA during
isolation. They are also unlikely to be caused by the
poor specificity of the DNA probe, since no RNA-DNA
hybridization band was detected during the light periods
(Fig. 8). Therefore, we speculate that the transcripts of
nifE and n i m are relatively unstable and can be
degraded immediately after translation. However, two
major bands with sizes of 0.5 and 1.1kb were detected
by the nifX probe (Fig. 8c). The lengths of these two
748
bands correlate well with those predicted from the
transcript of nifX and that extending from nifX to orf2,
respectively. As shown in Fig. 8(d), the band of 1.1kb
was present as a major band when a DNA probe specific
to orf2 was used. The results of Fig. 8(a-d) indicate that
nifE, n i m , nifX and orf2 are unlikely to be transcribed
as a single operon. The observation of a major band of
0.5 kb correlating to nifX suggests that nifX is expressed
as a unit distinct from nifE. The possibility is further
supported by the results of primer extension analysis
described below.
When the DNA probes specific to orfl and nifw were
used, the hybridization patterns were the same. As
shown in Fig. 8(e) and (f), two major bands with sizes of
0.6 and 2.1 kb were detected. The presence of a major
transcript of 0.6 kb suggests that orfl and nifw are
expressed as a distinct unit from nifX and orf2. The
length of 0.6 kb correlates closely with the predicted
length of the transcript extending from orfl to nifw. It
appeared that orfl, instead of encoding a polypeptide,
was just an untranslated leader region of nifw after
using oligonucleotides specific to orfl and nifw for
primer extension analysis (described below). When total
RNA was hybridized by hesA and hesB probes, respectively, a major band of 2.1 kb was detected (Fig. 8g,
h). Since no transcription initiation site was detected for
hesA or hesB (see below), n i p , hesA and hesB are
probably organized as a single operon. The size of 2.1 kb
correlates closely to the predicted length of the transcript
extending from nifw to hesB or from nifw to ‘fdx’.
However, the latter possibility was excluded because
‘fdx’ seemed to be transcribed in a unit different from
nifw, according to Northern analysis. As shown in Fig.
1O(d), two bands with sizes of 0.4 and 1.4 kb were
detected by an ‘fdx’ probe. The size of the major band
(0.4 kb) correlated well with the predicted length of the
mRNA transcribed from ‘ f d x ’ .It is suggested that ‘fdx’
is organized in a separate operon from nifw.
A clear band of 4.4 kb was detected when using nifX or
orf2 as probe (Fig. 8c, d). It is possible that the band is
a transcript extending past the ‘fdx’ gene. If this is the
case, then we would expect to detect the 4.4 kb band
when orfl, n i p , hesA, hesB or ‘fdx’ are used for
Northern hybridization, as revealed by the results of the
study of the nifllDK operon in Anabaena PCC 7120
(Haselkorn et al., 1986) and Synechococcus RF-1
(Huang & Chow, 1990). However, the 4.4 kb band was
not detected when orfl, nifw, hesA, hesB or ‘fdx’ were
used.
Identification of the start site by primer extension analysis.
To define the transcriptional unit more precisely, primer
extension analysis using synthetic oligonucleotides specific to the coding regions of nifE, n i m , nifX, orf2, orfZ,
n i p , hesA and hesB, respectively, was performed. A
significant extended product was detected when the
oligonucleotide specific to nifE, nifX, orfl or nifw was
assayed. However, no reasonable result was found when
n i p , orf2, hesA or hesB was examined. When a
synthetic oligonucleotide complementary to the coding
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sun, 18 Jun 2017 01:13:53
Organization and expression of nitrogen-fixation genes
~-
Fig. 8. Rhythmic expression of nifE (a), nifN (b), nifX (c), off2 (d), off1 (e), nifW (0, hesA (9) and hesB (h) in Synechococcus
RF-1 examined by Northern hybridization. Samples for RNA isolation were taken a t 4 h intervals (as indicated above each
lane) after the 12 h U12 h D-entrained culture was transferred t o UL. An equal amount of RNA (about 5 pg) was loaded
on each lane. DIG-labelled DNA probe specific t o each gene was prepared by PCR. The estimated sizes of major
hybridizing bands are shown by arrowheads. (The expression pattern of ‘fdx’ was similar t o the nif genes; however, t o
reduce the size of the figure, the data for ’fdx‘ are not included.)
sequence from position 56 to 33 of n i p was used, the
‘ G ’ at position 222 upstream of the start codon of nifE
was found to be the transcription start site (Fig. 9a).
When an oligonucleotide complementary to the coding
sequence from position 58 to 29 of nifX was used, the
‘ G ’ at position 45 upstream of the start codon of nifX
was shown to be the start site (Fig. 9b). These results
indicate that nip-nifN and nifX-orf2 are organized in
separate operons.
When a synthetic oligonucleotide complementary to the
coding sequence from position 61 to 38 of nifw was
used, as shown in Fig. 9(d), the same ‘G’ identified in
orfl was also found to be the start site of nifw. Thus, it
appears that the region of ‘orfl ’ is possibly an untranslated leader region of nifw rather than a gene. The
primer extension data support the suggestion that n i p nifN, nifX-orf2 and nifw-hesA-hesB, respectively, are
organized in a transcriptional unit.
When an oligonucleotide complementary to the coding
region from position 198 to 178 of orfl was used, the
extension site was found to be the ‘G’ situated at
position 7 within the coding region of orfl (Fig. 9c).
hesA, hesB and ‘fdx’ genes are expressed only under
conditions of nitrogen fixation. Fig. 8(g)and (h) show that
the hesA-like and hesB-like genes in Synechococcus RF-1
are expressed in a rhythmic pattern similar to the nif
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sun, 18 Jun 2017 01:13:53
749
T. C. H U A N G a n d OTHERS
................................................................................................................................................................................................................................................................
Fig. 9. Primer extension analysis using the niff- (a), nifX- (b), o r f l - (c) and nifW-specific (d) primers. Lanes 1, product of
primer extension (arrow) when a synthetic oligonucleotide was annealed with total RNA and extended as described in
Methods. The synthetic oligonucleotides specific to nifE, nifX, o r f l and nifW are described in Methods.
genes when the culture is grown in continuous illumination after a diurnal L/D regimen. The ‘fdx’ gene
also exhibits the same rhythmic pattern (data not
shown). To examine whether hesA, hesB and ‘fdx’ are
concerned with nitrogen fixation, total RNA was
extracted from cultures grown in media with sodium
nitrate (BG-11) and without sodium nitrate (BG-llo),
respectively. Northern hybridization showed that hesA,
hesB and ‘fdx’ were only expressed in the culture
without a combined nitrogen source (Fig. 10).
DISCUSSION
Repeating sequences in the non-coding regions
The occurrence of repeating sequences in the noncoding regions of nif genes in heterocyst-forming cyanobacteria has been documented in Anabaena PCC 7120
(Mulligan & Haselkorn, 1989), Anabaena azollae
(Jackman & Mulligan, 1995) and Calothrix sp. PCC
7611 (Mazel et al., 1990). The function of the repeats is
unknown; however, it has been suggested that they may
be important for the regulation of transcription
(Mulligan & Haselkorn, 1989), chromosome maintenance (Mazel et al., 1990) or responsible for mRNA
stability or processing (Jackman & Mulligan, 1995).
.................................................................................................................................................
Fig- 10. Comparison of the expression patterns of hesA (a), hesB
The repeating sequences that occur in nif operons of
heterocyst-forming cyanobacteria were not found in
Synechococcus RF-1. However, an unusual DNA sequence, (CCCCCACTCTCCCACT), (the underlined
‘ T ’ is replaced by ‘ C ’ in the fifth copy), was found to be
situated in the non-coding region between nifB and n i p
(transcribed in opposite directions). The function of the
repeating sequence is not known. However, in view of
the high G C ratio and the location at the middle of the
non-coding region between nifB and n i p , it might
(b) and ‘fdx’ (d) between cultures grown in nitrate-containing
(BG-11) and nitrate-free (BG-1 lo) media. The transcript levels of
hesA, hesB and ’fdx’ were assayed by Northern hybridization.
Samples for RNA isolation were taken a t 18 and 22 h after the
12 h U12 h D-entrained cultures were transferred to UL.
Northern hybridization of nifX (c) under the same conditions
was performed as control.
Downloaded from www.microbiologyresearch.org by
750
IP: 88.99.165.207
On: Sun, 18 Jun 2017 01:13:53
+
Organization and expression of nitrogen-fixation genes
.....................................................................................................
P
-
B
fdxNS
U
H
D
K
E
N
Orf
X i
hesA hesB ?dx’
Wl
/
*
1 kb
function as a motif to separate the two genes transcribed
in opposite directions.
Another unusual feature is situated in the untranslated
leader of nifE. It contains two copies of TCAGGAG-
TTAGGAGTCAGAAGGCAAAAGGCAAAAGTT
(the underlined ‘ C ’ is replaced by ‘ T ’ in the second
copy), three copies of CCCCACCCT and a reverse
repeat of CCCTCCCCACTC. The products of nifE and
nifr’r, required for Mo/Fe cofactor biosynthesis, have
significant degrees of sequence identity when compared
to the Mo/Fe protein a- and p-subunits (Brigle et al.,
1987). Based on the results of Northern hybridization
(Fig. 8), the transcripts of nifE and nifr’r seem to be
degraded more rapidly than other nif genes. We speculate that the degradation may occur immediately following the translation and the introduction of such an
unusual feature in the untranslated leader may have a
role in mRNA processing. O n the other hand, since the
transcripts of nifE and nifN are relatively unstable, it is
also possible that the unusual feature may play a role in
increasing the translation efficiency.
Organization of nif and nif-associated genes
Comparison of the physical organization of nif and nifassociated genes between Klebsiella pneumoniae and
Anabaena PCC 7120 (Dean & Jacobson, 1992) or among
Klebsiella pneumoniae, Axotobacter vinelandii, Bradyrhizobium japonicum, Rhodobacter capsulatus and
Clostridium pasteurianum (Haselkorn & Buikema,
1992; Dean & Jacobson, 1992) has been reviewed.
Characterization of 16 nif and nif-associated genes in
Synechococcus RF-1 has been studied in this paper and
our previous report (Chen et al., 1996). These genes are
clustered in a continuous arrangement spanning approximately 18 kb with seven transcriptional units (Fig. 11).
They all have the same reading direction except n i p .
The structural genes for nitrogenase (nifH, n i p and
nifK) are arranged as a nifHDK operon, an arrangement
common among diazotrophs. The n i p operon, containing n i p , fdxN, nifS and n i p , is located upstream of
niftl. The organization of the n i p operon (nip-fdxNnifs-nifu) in Synechococcus RF-1 is the same as in
Anabaena PCC 7120 and other cyanobacteria tested, but
different from the other diazotrophs. The nif and nif~
~-
Fig. 77. Physical organization of the nif and
nif-associatedgenes from Synechococcus RF1. Arrows indicate the position and direction
for the transcription initiation sites revealed
by Northern hybridization and primer
extension analysis. nifP, hesA, hesB and fdx’
were assigned as nif-associated genes
because they were expressed only under
nitrogen-fixing
conditions. The
ORF
downstream of nifX was identified as a
counterpart of ORF-2 in Anabaena PCC
7120.
Information
concerning
the
organization is presented in this paper and
in our previous publication (Chen et a/.,
1996).
associated genes located downstream from nifK in
Synechococcus RF-1 were found to be arranged nifEnifN-nifx-orf-nifW-hesA-hesB-‘
fdx ’. The organization
is nearly the same as in Anabaena PCC 7120 except that
ORF-3 located between nifK and nifE in Anabaena PCC
7120 (U47055) was not found in Synechococcus RF-1.
The operon containing n i p , located upstream of n i p ,
and that containing ‘fdx’, located downstream from
hesB, in Synechococcus RF-1 have not been fully
characterized. Recently, a nif] gene encoding an oxidoreductase in Anabaena has been identified (Bauer et
al., 1993; Schmitz et al., 1993). Therefore, it is likely that
more nif and nif-associated genes in Synechococcus RF1 will be revealed after further investigation.
Rhythmic expression of nif and nif-associated genes
Based on the results of this study and our previous
report (Huang & Chow, 1990), all nif and nif-associated
genes examined in Synechococcus RF-1 were expressed
in a rhythmic pattern with transcription activity occurring mainly within the dark phase when the culture
was grown in a diurnal L/D regimen. It has been
suggested that the rhythmic nitrogenase activity is
controlled by a ‘circadian oscillator’ (Huang et al.,
1990). Several possibilities for an ‘ oscillator ’ to regulate
the circadian rhythm of nif gene expression have been
put forward. They may be directly regulated by the
product of the ‘oscillator’, or they may be controlled
stepwise. The non-coding regions of n i p , nifH and nifE
of Synechococcus RF-1 were examined and analysed ;
however, no significant structural similarity among
them was found. The results suggest that the rhythmic
expression of nif genes is unlikely to be regulated directly
by the product of an ‘oscillator’. It is known that the
expression of nitrogen-fixation genes is affected by
oxygen concentration. There is a close association
between the rhythmic nitrogenase activity and the dark
respiration rate in Synechococcus RF-1 (Grobbelaar &
Huang, 1991). The dark respiration rate increased
considerably a few hours before the increase in nitrogenase activity when Synechococcus RF-1 was grown in
a diurnal L/D regimen. Therefore, the rhythmic expression of nif genes can be controlled by the rhythmic
change of oxygen concentration which in turn is
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sun, 18 Jun 2017 01:13:53
751
f:C.
H U A N G a n d OTHERS
regulated directly or indirectly by a ‘circadian oscillator ’.
ACKNOWLEDGEMENTS
This work was funded by the Academia Sinica and the
National Science Council of the Republic of China.
REFERENCES
Alam, 1.. Whitaker, R. A., Krogmann, D. W. & Curtis, S. E. (1986).
Isolarioii and sequence of the gene for ferredoxin 1 from the
cyanobacterium Anabaena sp. strain PCC 7120. J Bacteriol 168,
126.5- 1271 ,
Arnold, W., Rump, A,, Klipp, W., Priefer, U. B. & Puhler, A. (1988).
Nucleotide sequence of a 24206-base-pair DNA fragment
carrying the entire nitrogen fixation gene cluster of Klebsiel1,z
pneumoniae. J Mol Biol 203, 715-738.
Bauer, C. C.. Scappino, L. A. & Haselkorn, R. (1993). Growth of the
cyanobacterium Anabaena on molecular nitrogen: nifJ is required
when iron is limited. Proc Nut1 Acad Sci USA 90, 8812-8816.
Blihme, H. & Haselkorn, R. (1988). Molecular cloning and
nucleotide sequence analysis of the gene coding for heterocyst
ferredoxin from the cyanobacterium Anabaena sp. strain PCC
7120. Mol Gen Genet 214, 278-285.
Borthakur, D., Basche, M., Buikema, W. 1.. Borthakur, P. 6.
Haselkorn, R. (1990). Expression, nucleotide sequence and mutational analysis of two open reading frames in the nifgene region
of Anabaena sp. strain PCC 7120. Mol Cen Genet 221, 227-234.
Brigle, K. E., Weiss, M. C., Newton, W. E. & Dean, D. R. (1987).
Products of the iron-molybdenum cofactor-specific biosyntheti’:
genes, nifE and nifN, are structurally homologous to the products
of the nitrogenase molybdenum-iron protein genes, n i m and
nifK. J Bacteriol 169, 1547-1553.
Chen, H.-M., Huang, T.4. & Chien, C.-Y. (1996). Nucleotide
sequence of the nifHDK operon in the aerobic nitrogen-fixing
unicellular Synechococcus RF-1. Bot Bull Acad Sin (Taipei) 37,
99-105,
Chen, K. C. K., Chen, J. 5. & Johnson, J. L. (1986). Structural
features of multiple nifH-like sequences and very biased codon
usage in nitrogenase genes of Clostridium pasteurianum. ,I
Bacteriol 166, 162-172.
Colon-Lopez, M. 5.. Sherman, D. M. & Sherman, L. A. (1997).
Transcriptional and translational regulation of nitrogenase in
light-dark- and continuous-light-grown cultures of the unicellular cyanobacterium Cyanothece sp. strain ATCC 51142. 1
Bacteriol 179, 43 194327.
Dean, D. R. & Jacobson, M. R. (1992). Biochemical genetics of
nitrogenase. In Biological Nitrogen Fixation, pp. 763-834. Edited
by G. Stacey, R . H. Burris & H. J. Evans. New York: Chapman
LC: Hall.
Denk, D. & Rock, A. (1987). L-Cysteine biosynthesis in Escherichia
ioli: nucleotide sequence and expression of the serine acetyltransferase (cysE) gene from the wild-type and a cysteineexcreting mutant. Gen Microbiol 133, 515-525.
Ebeling, 5.. Hahn, M., Fischer, H. M. & Hennecke, H. (1987).
Identification of nifE-nifN-nifS-like genes i n Bradyrhizobium
japonicunz. Mol Gen Genet 207, 503-508.
Evans, D. J., Jones, R., Woodley, P. R., Wilborn, J. R. & Robson,
R. L. (1991). Nucleotide sequence and genetic analysis of thc
Azotobacter chroococcum nifUSVWZM gene cluster, including L:
new gene (nip)which encodes a serine acetyltransferase. ,I
Racteriol 173, 5457-5469.
752
Fay, P. (1992). Oxygen relations of nitrogen fixation in cyanobacteria. Microbiol Rev 56, 340-373.
Fischer, H. M., Bruderer, T. & Hennecke, H. (1988). Essential and
non-essential domains in the Bradyrhizobium japonicum NifA
protein : identification of indispensable cysteine residues potentially involved in redox activity and/or metal binding. Nucleic:
Acids Res 16, 2207-2224.
Gallon, J. R., Perry, 5. M., Rajab, T. M. A., Flayeh, K. A. M., Yunes,
J. S. & Chaplin, A. E. (1988). iMetabolic changes associated with
the diurnal pattern of nitrogen fixation in Gloeothece. J Gen
Microbiol 134, 3079-3087.
Grobbelaar, N. & Huang, T.-C. (1991). Relationship between the
nitrogenase activity and dark respiration rate of Synechococcus
RF-1. FEMS Microbiol Lett 83, 99-102.
Haselkorn, R. & Buikema, W. J. (1992). Nitrogen fixation in
cyanobacteria. In Biological Nitrogen Fixation, pp. 166-190.
Edited by G. Stacey, R. H. Burris & H. J. Evans. New York:
Chapman & Hall.
Haselkorn, R., Golden, 1. W., Lammers, P. J. & Mulligan, M. E.
(1986). Developmental rearrangement of cyanobacterial nitrogenfixation genes. Trends Genet 2,255-259.
Huang, T.-C. & Chow, T.-J. (1988). Comparative studies of some
nitrogen-fixing unlcellular cyanobacteria isolated from rice fields.
J Gen Microbiol 134, 3089-3097.
Huang, T.-C. & Chow, T.-J. (1990). Characterization of the
rhythmic nitrogen-fixing activity of Synechococcus sp. RF-1 at the
transcription level. Curr Microbiol 20, 23--26.
Huang, T.-C. & Grobbelaar, N. (1995). The circadian clock in the
prokaryote Synechococcus RF-I. Microbiology 141, 535-540.
Huang, T.-C., Tu, J., Chow, T.-J. & Chen, T.-H. (1990). Circadian
rhythm of the prokaryote Synechococcus sp. RF-1. Plant Physiot
92, 531-533.
Jackman, D. M. & Mulligan, M. E. (1995). Characterization of a
nitrogen-fixation (nif) gene cluster from Anabaena azollae l a
shows that closely related cyanobacteria have highly variable but
structured intergenic regions. Microbiology 141, 2235-2244.
Jacobson, M. R., Brigle, K. E., Bennett, L. T., Setterquist, R. A.,
Wilson, M. 5.. Cash, V. L., Beynon, J., Newton, W. E. & Dean,
D. R. (1989). Physical and genetic map of the major nifgene cluster
from Azotobacter vinelandii. J Bacteriol 171, 1017-1027.
Kallas, T., Rebiere, M. C., Rippka, R. & Marsac, N. T. (1983). The
structural nif genes of the cyanobacteria Gloeothece sp. and
Calothrix sp. share homology with those of Anabaena sp., but the
Gloeothece genes have a different arrangement. J Bacteriol 155,
42743 1.
Kallas, T., Coursin, T. & Rippka, R. (1985). Different organization
of nifgenes in nonheterocystous and heterocystous cyanobacteria.
Plant Mol Biol 5, 321-329.
Klipp, W. (1990). Organization and regulation of nitrogen fixation
genes in Rhodobacter capsufatus. In Nitrogen Fixation : Achievements and Objectives, pp. 467474. Edited by P. G. M. Gresshoff,
L. E. Roth, G. Stacey & W. E. Newton. New York: Chapman &
Hall.
Mazel, D., Houmaid, J., Castets, A. M. & Tandeau de Marsac, N.
(1990). Highly repetitive DNA-sequence i n cyanobacterial genomes. J Bacteriol 172, 2755-2761.
Mitsui, A., Kumazawa, S., Takahashi, A., Ikemoto, H.. Cao, 5. &
Arai, T. (1986). Strategy by which nitrogen-fixing unicellular
cyanobacteria grow photoautotrophically. Nature 323, 720-722.
Mulligan. M. E. & Haselkorn. R. (1989). Nitrogen fixation (nifl
genes of the cyanobacterium Anabaena species strain PCC 7120.
J Biol Chem 264, 19200-19207.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sun, 18 Jun 2017 01:13:53
Organization and expression of nitrogen-fixation genes
Mulligan, M. E., Buikema, W. 1. & Haselkorn, R. (1988). Bacterialtype ferredoxin genes in the nitrogen fixation regions of the
cyanobacterium Anabaena sp. strain PCC 7120 and Rhizobium
meliloti. J Bacteriol 170, 4406-4410.
Noti, 1. D., Folkerts, O., Turken, A. & Szalay, A. (1986). Organization and characterization of genes essential for symbiotic
nitrogen fixation from Bradyrhizobium japonicum 1110. J Bacteriol 167, 774-783.
Postgate, 1. R. & Eady, R. R. (1988). The evolution of biological
nitrogen fixation. In Nitrogen Fixation : Hundred Years After, pp.
3140. Edited by H. Bothe, F. J. de Bruijn & W. E. Newton. New
York : Proceedings of the 7th International Congress on Nitrogen
Fixation.
Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989). Molecular
Cloning: a Laboratory Manual, 2nd edn. Cold Spring Harbor,
NY: Cold Spring Harbor Laboratory.
Schmitz, O., Kentemich, T., Zimmer, W., Hundeshagen, B. &
Bothe, H. (1993). Identification of the nifJ gene coding for
pyruvate :ferredoxin oxidoreductase in dinitrogen-fixing cyanobacteria. Arch Microbiol 160, 62-67.
Wang, S.-Z., Chen, J . 4 . & Johnson, 1. L. (1989). Nucleotide and
deduced amino acid sequence of nifE from Clostridium pasteurianum. Nucleic Acids Res 17, 3299.
Waterbury, J. B. & Rippka, R. (1989). Order Chroococcales
Wettstein 1924, emend. Rippka et al., 1979. In Bergey’s Manual of
Systematic Bacteriology, vol. 3, pp. 1728-1746. Edited by J. T.
Staley, M. P. Bryant, N. Pfenning & J. G. Holt. Baltimore:
Williams & Wilkins.
Received 27 May 1998; revised 22 September 1998; accepted
11 November 1998.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sun, 18 Jun 2017 01:13:53
753