Content uploaded by Jeffrey Cole
Author content
All content in this area was uploaded by Jeffrey Cole on Nov 10, 2015
Content may be subject to copyright.
Journal
of
General Microbiology
(1
978),
106,
1-12.
Printed in Great Britain
1
The Chromosomal Location and Pleiotropic Effects
of
Mutations
of
the
nirA+
Gene
of
Escherichia
coli
~12:
The
Essential Role
of
nirA+
in Nitrite Reduction
and in Other Anaerobic- Redox Reactions
By B. MARY NEWMAN
AND
J.
A. COLE
Department
of
Biochemistry, University
of
Birmingham,
P.O.
Box
363,
Birmingham
B15
2TT
(Received
9
December
1977)
Cytochrome
cjS2,
which has been implicated as an electron carrier for nitrite reduction
by
Escherichia coli,
has been separated from NADH-nitrite oxidoreductase activity. The
cytochrome is therefore not required for the reduction of nitrite by NADH
in vitro.
Never-
theless, some mutants which were selected by their inability to use nitrite as a nitrogen
source during anaerobic growth synthesize neither NADH-nitrite oxidoreductase nor
cytochrome
cSS2.
The defects
in
these mutants are due to mutations in a single gene,
nirA,
which is located at about minute 29 on the recalibrated linkage map. Experiments with an
F’
plasmid which carries a
nirA+
allele established that
nirA+
is dominant
to
the defective
allele. Other mutants, defective in nitrate reductase activity because of mutations in the
chlA
or
chZB
genes, synthesized nitrite reductase and cytochrome
cSj2
in the absence of
nitrate or nitrite.
A mutant with a defective
fnr
gene was also NirA- and, conversely,
nirA
mutants were
Fnr-. In a series of transduction experiments, attempts to separate the
nirA
and
fnr
defects
were unsuccessful. Furthermore, no complementation was observed when an F’ plasmid
carrying a defective
nirA
allele was transferred into the
fnr
strain. It is concluded that the
fir
gene described by Lambden
&
Guest (1976)
is
identical to the
nirA
gene and that its
product affects the synthesis or assembly of a variety of anaerobic redox enzymes which
include nitrite reductase, cytochrome
cjS2,
nitrate reductase, fumarate reductase and formate
hydrogen1 yase.
INTRODUCTION
Nitrate and nitrite are reduced rapidly by anaerobic cultures of
Escherichia coli
K12, and
nitrite can serve as the sole nitrogen source to support growth in a continuous culture
(Cole
et al.,
1974). Presumably, therefore, nitrite is reduced to ammonia. The reduction of
nitrate to nitrite is catalysed by a membrane-bound enzyme complex. This reaction is
coupled to oxidative phosphorylation and therefore provides energy for anaerobic
metabolism (Payne, 1973
;
Haddock
&
Kendall-Tobias, 1975
;
Kroger, 1977). Many species
of bacteria reduce nitrite to molecular nitrogen, a process which also generates ATP.
Several enzymes which catalyse steps in these reactions are associated with c-type cyto-
chromes and with the cytoplasmic membrane (Payne, Riley
&
Cox, 1971; Matsubara,
1971; Yamanaka
&
Okuwuki, 1974). In contrast, nitrite reduction by
E. coli
is catalysed
by the soluble enzyme NADH-nitrite oxidoreductase (EC 1.6.6.4) and there is as yet no
evidence that this reaction is coupled to ATP synthesis. A soluble cytochrome,
cSS2,
has
~ ~ ~~
Vol.
105,
No.
2
was
issued
17
April
1978
1-2
2
B.
M. NEWMAN AND
J.
A. COLE
been implicated in nitrite reduction by this organism because it is synthesized only during
anaerobic growth and highest concentrations are found in extracts of bacteria in which
nitrite reductase is most active (Fujita, 1966; Fujita
&
Sato, 1966, 1967; Cole
&
Wimpenny,
1968). Furthermore, both the activity of the enzyme and the cytochrome concentration
were high in extracts of a mutant unable to reduce nitrate because of a defect in the
chZA
gene (O'Hara
et
al.,
1967; Venables, Wimpenny
&
Cole, 1968; Kavanagh
&
Cole, 1976;
Newman
&
Cole, 1977). Other mutants with a NirA- phenotype were isolated on the basis
of their inability to assimilate nitrite during anaerobic growth: these are deficient in both
nitrite reductase activity and cytochrome
c552
(Cole
&
Ward, 1973).
All
of these results
could be accounted for if cytochrome
c552
were an integral part
of
the NADH-nitrite
oxidoreductase complex, or if the synthesis or activity of the enzyme and the cytochrome
were controlled by a common mechanism that is defective in
chZA
and NirA- strains. A third
possible explanation is that NirA- strains are defective in more than one gene.
To
distinguish
between these possibilities, we have investigated whether cytochrome
c552
can be separated
from nitrite reductase activity without
loss
of enzyme activity. The biochemical properties
of NirA+ revertants and
nirA/nirA+
merodiploids have been determined, and the
nirA
gene
has been mapped with reference to the
hemA,
cysB,
pyrF
and
fnr
genes (Bachmann, Low
&
Taylor, 1976; Lambden
&
Guest, 1976).
METHODS
Organisms.
The origins of the various strains of
Escherichia coli
~12 are listed in Table 1. Bacteria were
stored at 4 "C as stab cultures in nutrient agar deeps (Oxoid CM1).
Media and growth conditions.
Nitrogen-free (NF) salts contained (per
1
distilled water):
10.5
g K2HP04;
4.5
g
KH,PO,; 1 g trisodium citrate; 125 mg Na2S04; and
5
ml sulphur-free trace metals (Cole
et
al.,
1974).
The carbon source was glucose at 27 mM and nitrogen sources were as indicated in the text; these were
added aseptically before inoculation from sterile stock solutions of 40
%
(w/v) glucose,
20
%
(w/v) casein
hydrolysate (Oxoid L41), 2 M-NH,C~ and 1 M-N~NO,. In some experiments Lennox (1955) broth at the
equivalent of single strength replaced the casein hydrolysate. Inocula for batch cultures were aerated at 37
"C
for 6 to 8 h in 20 ml Lennox broth, and were transferred to 200 ml of prewarmed NF salts in
a
250 ml conical
flask. After 16 h at 37 "C without aeration, this culture was transferred to 1.8 1
NF
salts in a 2
1
conical flask.
Escherichia coli
HfrC was grown in
a
continuous culture in minimal medium in which
5
mM-NO,- was
the nitrogen source. At this concentration, nitrite limited growth and the dilution rate was 0.029 h-l (Cole
et ul.,
1974).
Preparation of cell-free extracts.
Bacteria were harvested and washed as described previously (Cole
et al.,
1974), and resuspended in 2 to
5
vol. TEA or TEM buffer, pH8.0 (50m~-Tris/HC1, 5m-EDTA and,
respectively,
5
m-ascorbate or 15 m-mercaptoethanol). Bacteria were broken by passage through a French
pressure cell at 4 "C and 69 MPa
(10000
lbf in-,) or, where indicated, by three 15
s
periods of sonication
with an MSE sonicator (probe diameter, 1.0 cm) operating at 20000
Hz.
High speed supernatant (HSS) and
cell membrane (CM) extracts were prepared as described previously (Cole
&
Rittenberg, 1971).
Enzyme assays.
NADH-nitrite oxidoreductase
(EC
1 .6.6.4) activity was assayed as described by Cole
et al.
(1974) except that open cuvettes at 30 "C were used for some assays. Activities were lower in HSS extracts
prepared by sonication and assayed anaerobically at room temperature than in those prepared with the
French press and assayed at 30 "C. For this reason, enzyme activities should only be compared within indi-
vidual experiments. Duplicated assays were reproducible to within 1
%
with each HSS extract and to within
20
%
with extracts of independent cultures of the same strain.
Formate-nitrate oxidoreductase activity was determined with bacteria which had been grown for 16 h
without aeration in 800 ml minimal medium (Newman
&
Cole, 1977) supplemented with 27 mM-glucose and
6.5 g nutrient broth
1-1
(Oxoid CM1) (Lambden
&
Guest, 1976). Cell membranes resuspended in
50
mM-
Na+/K+ phosphate buffer pH 7.4 were incubated at 37 "C with
40
mM-Na+ formate and
40
m-KNO, under
a
stream of N,. The protein concentration in the assay tube was in the range
0.5
to
5.0
mg ml-l, depending
on the strain used. Samples
(50~1)
were transferred at 1 or 2 min intervals for 10 to 20 min into 2-7 ml
1
%
(w/v) sulphanilamide in 1 M-HCl. At the end of each set of assays
0.3
mlO.02
%
(w/v) N-1-naphthylethylene-
diamine dihydrochloride was added. The absorbance at 540 nm, determined after 30min at room
temperature, was proportional to the quantity of nitrite in the sample in the range
0
to 100 nmol. Nitrate
reductase activities are expressed as nmol nitrite accumulated min-l (mg CM protein)-l. The rate of nitrite
accumulation was constant during the incubation, and was proportional to the concentration of CM protein
Strain
~l004a
c38
c~40
c~64
cs65
CB75
c~98
CB201
~~203
~~207
CB210
CB211
CB212
~~213
cs217
CB222
CB222
/F'23
~~223
~~242
~~247
~~248
cs253
~~254
~~255
0~56
0~75
RK~
chlA16
RK~
chlCl9
RK7
RK7
~hlB36
RK7
chlE13
RK7
chlE44
TKG49
~1485~
Characterization of nirA mutants
of
E. coli
Table
1.
Escherichia coli strains and their origins
3
Phenotype or genotype*
ilv met hemA
chlC
nirAI
cysB trp
cysB trp thyA
pyrF nirAl
nirAI cysB+ pro trp his str
nirB chl
Cys-Hfr
NirB
-
NirA- Cys-
nirA2
Cys-
nirA3
unknown growth requirement
NirA-
NirB- Cys-
NirB
-
nirAl pro trp recA
nirAl pro trp recA/F' trp+ nirA+
nirAI pro trp recA
trp nirAl
PYrF
fnr
+trp
Hfr
met
gal trpA trpR fnr
pyrD thi his trp recA mtl xyl ma1
gal str/F' trpf
Prototrophic Hfr
Prototrophic Hfr
Prototroph
chlA gal
chlB gal
chlC gal
chlE gal
chlE gal
thr
lea
ilvA his arg thipyrF
Prototroph, lysogenic for Plkc
thyA lac qme
Source or method of isolation
B. Haddock
W.
A.
Venables
See Cole
&
Ward (1973)
M. Jones-Mortimer
Derivative of ~~64 spontaneously resistant to 20 ,ug trimethoprim
Transduce ~~242 with P1 propagated on cs247. Select Trp+;
See Newman
&
Cole (1977)
N-me thyl-N'-ni tro-N-ni trosoguanidine
(N
G) mu tagenesis
of
NG mutagenesis of 0~56
NG mutagenesis of 0~75
NG mutagenesis of 0~75
NG mutagenesis of 0~75
NG mutagenesis of 0~75
Ethyl methanesulphonate (EMS) mutagenesis of
0~75
EMS mutagenesis of 0~75
See Newman
&
Cole (1977)
Conjugate ~~F23/KL181 with ~~222. Select Trp+ clone
Grow cs222/F'23 with 75
pg
acridine orange ml-l. Plate
survivors on to nutrient agar and screen for Trp- clone
Transduce cs64 with P1 propagated on ~~40. Select
Cys+;
screen for Nir-
Transduce c~64 with P1 propagated on
TKG~~.
Select
Cysf;
screen for Pyr-
Transduce c~64 with P1 propagated on
TKG~~.
Select Cys+;
screen for Pyr-Trp-
Grow cs242 in minimal media with
5
mM-NO,- and
0.5
m-
NH4+
as nitrogen source. After anaerobic growth, inoculate
minimal agar plates containing
5
mwnitrate as the nitrogen
source. Select large colonies after
5
d of anaerobic growth
Transduce JRG861a to
Fnr+
with P1 propagated on ~~248,
selecting for anaerobic growth on lactate/fumarate plates
Transduce ~~248
to
Pyr+ with JRG861a-Pl. Screen for formate
hydrogenlyase-negative clone
NCLB, Torry Research Station, Aberdeen
J.
R. Guest
B. Bachmann
ml-'
screen for Pyr-Nir-
0~56
R.
Curtiss
111
R.
Curtiss
I11
MacGregor (1975)
MacGregor (1975)
MacGregor (1 975)
MacGregor (1975)
MacGregor (1975)
MacGregor (1 975)
J.
R.
Guest
J.
R.
Guest
*
The system of genetic nomenclature
is
based upon the recommendations of Demerec
et al.
(1966).
4
B.
M.
NEWMAN AND
J.
A. COLE
Table
2.
Separation of nitrite reductase activity from cytochrome c,,,
Enzyme activities were determined at 30 "C in open cuvettes, and are expressed as nmol NADH
oxidized min-l (mg protein)-l.
Sample
Cytochrome
c552
Nitrite reductase
Total
7-
protein nmol Recovery Activity Recovery
(mg)
(
%)
(
%)
HSS
extract 1120 1010 100 42 100
(NH4),S04 supernatant 714 775 77
0
0
Desalted (NH,),SO, pellet 429
188
19 53 45
DE52 pool
45
0 0
169 16
in the range
0
to 1.0 mg ml-l for wild-type strains and in the range 1 to
5
mg ml-l for Fnr-, NirA- and
chlorate-resistant mutants, The specific activities of independently prepared samples of membranes from
wild-type strains were reproducible to within 20
%,
but far greater variations were found for mutant strains:
this was probably caused by inactivation of the nitrate reductase complex during the preparative procedures
(see also Lambden
&
Guest, 1976). Nevertheless, the differences between the activities of wild-type and mutant
strains were sufficiently great for these assays to yield reliable data.
Formate hydrogenlyase activity was determined qualitatively by growing bacteria in test-tubes containing
5ml
Lennox broth, 27m-glucose and an inverted Durham tube. A clearly visible bubble of
H2
had
accumulated in wild-type cultures after 18 to 24 h at 37 "C. No gas bubble was visible for at least 48 h when
the
fnr
mutant JRG861a was tested.
Strains were tested for their ability to reduce nitrite as described by Cole
&
Ward (1973), but with the
following modifications. Bacteria from single colonies were inoculated into 1
ml
Lennox broth. After 16 h
at 37 "C, 4 ml
NF
salts supplemented with 27 mM-glucose, 14 ~M-NH,+,
5
mM-NaNO, and required amino
acids at 20 pg ml-l was added. After 3 h at 37 "C and at subsequent hourly intervals, drops from each culture
were tested for the presence of nitrite. Wild-type bacteria had usually reduced the nitrite after 3 to
5
h,
but
nitrite remained in Nir- cultures after 6 h.
Cytochrome
c552
was assayed as described previously (Cole
et
al.,
1974). Difference spectra of cytochromes
at
-
196 "C were obtained by immersing cuvettes containing the oxidized and reduced samples in liquid
nitrogen and recording the spectrum with a Perkin-Elmer 356 double-beam spectrophotometer equipped
with a cryostat attachment.
Separation of nitrite reductase from cytochrome
~552.
Proteins precipitated from an
HSS
extract
of
E.
coIi
HfrC by 40
%
saturated
(NH,),SO,
in TEA were collected by centrifugation for 10 min at 10000
g
and
redissolved in TEM pH
8-0.
After passage through a column
of
Sephadex G-25 (25.0
x
4.0 cm diam.), the
desalted proteins were applied to a column
of
Whatman DE52 cellulose (17.5
x
2.8 cm diam.) equilibrated
with TEM pH
8.0.
Nitrite reductase was most active in fractions 72 to
80
(see Fig. 1)
:
these fractions were
therefore combined to give the DE52 pool.
Genetic methods.
Lennox (1955) broth was used for conjugation and transduction experiments, as
described by Newman
&
Cole (1977). Unless otherwise stated, a virulent derivative of bacteriophage Plkc
was used to prepare lysates for transduction experiments.
Mutants unable to reduce nitrite were isolated as described by Cole
&
Ward (1973).
To isolate homogenotized
nirA
merodiploid strains, a98
pro his trp nirA/F'nirA+trp+
was grown
aerobically for 16 h in minimal medium supplemented with proline and histidine and then
dilutions were plated on to a similar medium containing 1.25
%
(w/v) agar to obtain single colonies. Twenty-
five Nir- isolates were detected amongst 700 colonies that were tested, and five
of
these were retained for
complement a tion experiments.
to
RESULTS
Separation
of
cytochrome c552 from nitrite reductase activity
Escherichia coli
strain HfrC was grown in anaerobic continuous culture with nitrite as
the nitrogen source. Bacteria suspended in TEA were broken with the French pressure cell
and
HSS
extracts were prepared. Nitrite reductase activity was completely precipitated by
40
%
saturated
(NH,),SO,,
but 80
%
of
the cytochrome
c552
remained soluble (Table 2).
The
(NHA,SO4
pellet was resuspended in TEM and desalted by passage through a Sephadex
G-25
column equilibrated with TEM. Active fractions were then applied to an anion
Characterization
of
nirA mutants
of
E.
coli
5
0.4
M-KCl
30
r
/
I
1.0
5
!h
ug
2
0.8
-3
0.6
2
0.4
$
22
-1
XD
0.2
v,
-T
10
20
30
40
50
60
70
80
90
100
Fraction
no.
Fig.
1.
Separation of nitrite reductase activity
(0)
from cytochrome c552
(@)
by Whatman
DE52
anion exchange chromatography. For experimental details, see Methods.
Table
3.
Nitrite reductase activity and cytochrome
c552
concentration
of
wild-type and
mutant strains
of
E.
coli
Escherichiu
coli
strains were grown without aeration in NF salts supplemented with
Lennox
(1955)
broth,
20
m-NH,+ and
2
m~-No,-.
HSS
extracts were prepared after
cells
had been broken by
sonication. Nitrite reductase activities were assayed in anaerobic cuvettes at room temperature.
Activities are expressed as nmol NADH oxidized min-l (mg protein)-l, and cytochrome c552
concentrations as pmol (mg protein)-'.
All
results represent the mean of at least two experiments.
Strain
cs203
~~207
CB20 1
CB210
CB211
CB212
~~213
~~217
0~75
0~56
Nitrite
reductase
2
5
5
6
8
30
0
0
200
90
Cytochrome
c552
35
55
Broad peak from
550
to
560
nm
0
0
0
40
13
33
35
Phenotype
NirB-
NirB-
NirA-
NirA-
NirA-
NirA-
NirB-
NirB-
Nir+
Nir+
exchange column and eluted with a linear gradient of KCl in TEM. The cytochrome was
eluted as
a
well-defined peak by
0.16
M-KCl, before the peak of enzyme activity (Fig.
1).
The fractions containing most
of
the enzyme activity and the highest concentrations
of
cytochrome
c552
were pooled and assayed alone and together for both proteins. Partially
purified nitrite reductase was uncontaminated with cytochrome
c552
and its activity was
neither inhibited nor enhanced by the cytochrome fraction. The specific activity of nitrite
reductase had increased fourfold during the separation procedures, and
16
"/o
of the original
activity was recovered. It is concluded that cytochrome
c,,,
is not a component of the
NADH-nitrite oxidoreductase complex.
Properties
of
mutants unable
to
reduce nitrite
Eight newly isolated mutants defective in their ability to reduce nitrite were grown
anaerobically with nitrite. The activity of nitrite reductase and the concentration
of
cyto-
chrome
c552
in
HSS
extracts of these strains are shown in Table
3.
Nitrite reductase activity
6
B.
M.
NEWMAN AND
J.
A.
COLE
was low or absent in seven of these mutants, but extracts of CB212 retained some activity
which was less than 50
%
of
the lowest activity detected in wild-type bacteria. Four of the
mutants, ~~207, ~~210, ~~211 and ~~212, were similar to the original NirA- isolates in
that cytochrome
c552
could not be detected. These strains are therefore designated NirA-.
Cytochrome
c552
was detected in strains cs201, ~~203, ~~213 and ~~217. These isolates
represent a new class
of
mutant and are designated NirB-. Their biochemical and genetic
properties will be reported elsewhere.
Chromosomal location
of
the nirA gene
Bacteriophage
P1
was propagated on strain ~~40
cysB+nirAI
and the lysate was used to
transduce strain c~64 to Cys+. In one experiment, 196 Cys+ transductants were tested for
their ability to reduce nitrite: 4 were Nir- and 192 were Nir+. This low cotransduction
frequency
(2
%)
suggested that the
nirAI
defect is located about 1-5 min away from
cysB.
The frequency of cotransduction of
nirA
with
hemA
and
pyrF
genes located on either side
of
cysB
was therefore investigated. The
hemA
mutant ~1004a was used as the recipient in
a transduction with strain c~98
nirAI
as the donor, and 200 Hem+ transductants were
tested for their ability to reduce nitrite. None were Nir-. In duplicate experiments with the
cysB+trp+pyrF
strain ~~247 as the recipient and ~~242
nirAI trp
as the donor, 17 out of
297 and 5 out
of
80 Pyr+ recombinants were Nir-. The cotransduction frequency between
pyrF+
and
nirAI
is therefore about 6
%.
If it is assumed that the Wu (1966) mapping
function is applicable to these low cotransduction frequencies, !it can be calculated that the
nirA
gene is located approximately 1.2 min away from the
pyrF
gene. Only
1
of
the 22
Pyr+Nir- transductants had inherited the donor Trp- phenotype. The
nirA
gene is therefore
on the opposite side
of
pyrF
to the
cysB
gene and the tryptophan operon and is located at
approximately minute 29.2 on the linkage map
of
Bachmann
et
al.
(1976).
Two other NirA- mutants, CB210 and ~~211, were used as donors to transduce ~~247 to
Pyr+.
In
each experiment at least 2 out of 100 PyrF+ recombinants were Nir-. Furthermore,
the Nir- defect in ~~210 did not cotransduce with
hemA+.
The mutations in ~~210 and
~~211 are therefore located in the same 2 min segment
of
the chromosome as the
nirAI
mutation in strains ~~40, c~98 and ~~242.
Pleiotropic eflects
of
a
nirA mutation
Four
cysB+nirAI
recombinants were purified from the transduction with c~40 as the
donor and c~64 as the recipient. Cell-free extracts of these strains prepared after anaerobic
growth with nitrite were deficient in both nitrite reductase activity and cytochrome
c552
(Table
4).
Similar data were obtained with Pyr+Nir- recombinants from transductions with
strains ~~210, CB211 and ~~242 as the donor and ~~247 as the recipient. In contrast,
Pyr+Nir+ transductants synthesized both proteins (Table 4). It is apparent, therefore, that
the
nirA
gene affects the synthesis of both nitrite reductase and cytochrome
c552.
To exclude the possibility that all three NirA- strains carry mutations in two or more
closely linked genes or that all of the strains have deletions covering more than one gene,
a spontaneous Nir+ revertant
of
~~242
nirA
was isolated. The nitrite reductase activity of
this revertant, ~~253, was similar to that of the original wild-type. Furthermore, the low
temperature cytochrome spectrum of an
HSS
extract
of
~~253 was identical to that
of
the wild-type with an absorbance maximum at 550 nm as well as at 556 and 559 nm (Fig. 2).
In contrast, no absorbance maximum at
550
nm was observed in spectra of
nirA
strains
(Fig. 2). Thus cytochrome
c552
synthesis was restored to a revertant that had been selected
on the basis of its ability to reduce nitrite.
PyrF+ transductants were selected from two experiments with the revertant ~~253 as the
donor and either
a
pyrF nirAI
or a
pyrF nirAf
strain as the recipient. When strain ~~247
pyrFnirA+
was the recipient, 100 out of 100 Pyr+ transductants were Nir+. The revertant
Characterization
of
nirA mutants
of
E.
coli
7
Table
4.
Nitrite reductase activity and cytochrome c552 concentration in nirA+ and
nirA
transductan ts
Escherichia
coIi
strains were grown without aeration in minimal medium containing
1
g casein
hydrolysate l-l,
20
mM-NH,+ and
2.5
m~-N0,-. The activity of nitrite reductase was
assayed at
30
"C
with open cuvettes. Nitrite reductase activities are expressed as nmol
NADH
oxidized min-l (mg protein)-l, and cytochrome c552 concentrations as pmol (mg protein)-l.
Strains
~~64, ~~75, -242, ~~247, ~~248
and
~~253
share a common genetic background.
Donor
c~40
Nir-
~~242
nirAl
~~211
nirA3
c~210
nirA2
~~253
nir+
JRG8 6 1 a
fnr
Control:
~~G861a
Control
:
-248
I
Recombinant
Recipient phenotype
c~64
Nir+ Cys+Nir-
(1)
(2)
(3)
(4)
~~247
nir+
Pyr+Nir-
(1)
(2)
~~247
nir
+
Pyr+Nir-
(1)
(2)
Pyr+Nir+
~~247
nir+
Pyr+Nir-
(1)
(2)
Pyr+Nir+
CB75
nirAl
Pyr+Nir+
(1)
(2)
(3)
~~248
nip+
Pyr+Fnr-
(1)
Nitrite
reductase
34
28
31
22
0
9
0
34
555
0
0
292
716
993
799
0
C
y
t
ochrome
<
10
<
10
<
10
<
10
<
10
<
10
<
10
<
10
206
<
10
<
10
172
82
183
92
<
10
c552
(2)
0
<
10
(3)
3
<
10
-
5
<
10
-
125 124
Mutant
strain
~~242
ievertan t
rl\
~~253.
520
540 560
580
600 620
640
Wavelength (nm)
Fig.
2.
Difference spectra
of
reduced minus oxidized
HSS
extracts of the
nirAl
strain
~~242
and the
revertant
~~253.
Samples were frozen in liquid nitrogen and spectra were recorded with a Perkin-
Elmer
356
double-beam spectrophotometer. The vertical bar corresponds to an absorbance incre-
ment
of
0.1
for the revertant sample, and
0.003
for the mutant sample (protein concentrations were
12.5
and
4.3
mg ml-l, respectively).
8
B. M.
NEWMAN AND
J.
A.
COLE
Table
5.
Nitrite reductase activity and cytochrome
c5,,
concentration in
chlorate-resistant mutants
Bacteria were grown in NF salts with
1
g casein hydrolysate
1-'
and, where indicated,
10
mM-KNO,
or
5
mM-NaNO,. Cell-free extracts were prepared with the French pressure cell, and nitrite reductase
activities were determined at
30
"C
with open cuvettes. Each line of data is derived from a single
independent culture. Nitrite reductase activities are expressed as nmol NADH oxidized min-l
(mg
protein)-l, and cytochrome
c552
concatrations as pmol (mg protein)-l.
Additional Nitrite Cytochrome
Strain N source reductase
CSSZ
RK~
(wild-type) None
62
<
10
54
<
10
RK~
chIAI6
None
536 103
478 97
RK7
~hlB36
None
625 118
438 49
RK7
chICI9
None
71
<
10
75
<
10
RK7
~hlE13
None
45
<
10
15
<
10
RK~
chlE44
None
52
<
10
43
<
10
RK~
chlAl6
Nitrate
589
<
10
RK~
chIB36
Nitrate
500
<
10
RK~
chICI9
Nitrate
293 35
RK~
chIEI3
Nitrate
383
<
10
RK~
chIE44
Nitrate
359 37
RK~
chlAl6
Nitrite
446
<
10
RK~
chIB36
Nitrite
361
<
10
c38
chIC
None
151
<
10
113 60
c38
chlC
Nitrate
348
<
10
~~253
therefore no longer carries a
nirA
defect that can readily be segregated from the
reversion site. Conversely, when
CB75
pyrFnirA1
was the recipient,
94
out of
100
Pyr+
transductants were still Nir-: none of the six that were Nir+ had inherited the Trp- pheno-
type of the donor. Three of these Pyr+Nir+ transductants were assayed for nitrite reductase
activity and cytochrome
c552
synthesis: all were similar to the wild-type in both respects
(Table
4).
The cotransduction frequency of the Nir+ reversion site with
pyrF+
is therefore
6
%,
and the mutation must lie on the opposite side of
pyrF
with respect to the tryptophan
operon. The reversion site is therefore in, or extremely close to the
nirA
gene.
Dominance
of
the nirA+ allele
The preceding experiments established that defects in the
nirA
gene result in the loss
of
at least two proteins. It is possible, therefore, that
nirAf
is a regulatory gene coding for
a positive control protein or a repressor. If the second alternative is correct, it would be
necessary to suggest that
nirA
mutants synthesize
a
superrzpressor similar to that specified
by the
lacis
allele (Jacob
&
Monod,
1961).
Dominance tests were used to distinguish between
these two possibilities.
The
nirA+trp+
plasmid
F'23
was introduced into
CB222,
a
red
derivative
of
c~98
nirA trp,
and the nitrite reductase activity and cytochrome
~552
concentration in the merodiploid, the
original mutant and a cured derivative of the merodiploid were determined. Only the
merodiploid strain reduced nitrite during growth, and nitrite reductase activity and cyto-
chrome
~552
were detected only in
HSS
extracts of this strain.
At the end of growth,
25
%
of
the single colony isolates from the
c~222/F'23
culture were
Trp+ and were able to donate this phenotype to
c~64
trp
by conjugation. Thus
a
minority
Characterization of nirA mutants of
E.
coli
9
Table
6.
Nitrite reduction and forrnate hydrogenlyase activity during growth, and
formate-nitrate oxidoreductase activity of various nirA, fnr and wild-type strains
Formate hydrogenlyase activities were recorded after
30
h at 37
“C;
each culture was then sup-
plemented with 27 mM-glucose and incubated for
a
further 50 h at 37
“C.
Data for formate-nitrate
oxidoreductase activities [expressed as nmol
NOz-
formed min-l (mg protein)-l] are for single
experiments: activities of 2.0 and 22.5 nmol NO,- formed min-l (mg protein)-l were found for
independent cultures of strains JRG861a and ~~248, respectively.
Gas production by
Nitrite formate
duction
-
nitrate
during After After oxido-
re- hydrogenlyase Formate-
Strain Parent Relevant genotype growth 30
h
80h reductase
fnr
fnr
+
nirA
+
fnr
+
nirAl
nirAl /F’nirA*
nirA+
revertant of
nirAl
nirA2
nirA3
fnr
-
+
10.8
+
+
+++
125.0
+
+
++
ND*
3.7
+ +
+++
30.0
+
+
+++
21.3
+
+
+++
18.3
-
2.2
-
- - -
-
-
ND
ND
-
-
-
-
-
-
*
ND,
Not determined.
of the bacteria had retained the plasmid. For this reason, the nitrite reductase activity of
this culture was lower than that of a haploid
nirA+
strain. The
nit-A+
plasmid
F’23
therefore
restored nitrite reductase activity and cytochrome
c552
synthesis to a
nirA
strain, and the
nirA+
allele is dominant to
nirAl.
Nitrite reductase and cytochrome c552
in
chlorate-resistant mutants
Previous reports that nitrite reductase and cytochrome
c552
are derepressed
in
a
chlA
mutant suggest that there is a common component which regulates the synthesis, assembly
or activity of both of these proteins as well as the nitrate reductase complex (O’Hara
et
al.,
1967;
Venables
et al.,
1968)
The effects of mutations in other
chl
genes were therefore
investigated.
Nitrite reductase and cytochrome
c552
were apparently derepressed during growth
of
chlA
and
chlB
mutants without nitrate or nitrite, but both oxidizing agents repressed the synthesis
of cytochrome
c552
(Table
5).
The regulation of the synthesis of both proteins in
chlC
and
chlE
strains was similar to that of the wild-type.
No
nitrite accumulated in the growth
medium when strains
chlC38, chlE13
and
chlE44
were grown with
10
mM-KNO,.
The de-
repression of nitrite reductase in these strains is therefore probably caused by nitrate rather
than its reduction product, nitrite.
Relative positions
of
the nit-A and fnr genes
At least three genes,
nirA, chlA
and
chlB,
affect the synthesis
of
both nitrite reductase and
cytochrome
cSs2.
Lambden
&
Guest
(1976)
have described other mutants with defects in the
fnr
gene which are also defective in the synthesis or activity of several proteins that catalyse
anaerobic oxidation-reduction reactions. Furthermore, the
fnr
gene, like the
nirA
gene,
is
5
to
9
%
cotransducible with
pyrF+
and is located at approximately minute
29
on the
recalibrated
E. coli
linkage map (Lambden
&
Guest,
1976).
It was therefore of interest
to determine the relative positions of the
nirA
and
fnr
genes to investigate whether they
10
B.
M.
NEWMAN
AND
J.
A.
COLE
might be sufficiently close to be part of an operon concerned with anaerobic metabolism
or
its regulation.
Preliminary experiments established that the
fnr
strain ~~~861a was unable to reduce
nitrite and, conversely, that the
nirA
strain ~~242 was Fnr- (Table 6). Furthermore, a variety
of other
nirA
strains were Fnr-, but the Nir+ revertant ~~253 and the
nirA/nirA+
merodi-
ploid C~242/F’23 were Fnr+ (Table 6). Strain ~~248 was transduced with phage PI which
had been propagated on ~~G861a, and
200
Pyr+ transductants from each of two independent
experiments were scored for their Nir and Fnr phenotypes. In the first experiment, 197 Pyr+
transductants were both Fnr+ and Nirf, and 3 were both Fnr- and Nir- (Table 4). In the
second experiment, 192 Pyrf transductants were also Fnr+Nir+, and
8
were Fnr-Nir-. For
the reciprocal cross, phage propagated on strain ~~248 was used to transduce ~~G861a to
Fnr+, and transductants were selected by their ability to grow anaerobically with lactate
plus fumarate as the source of carbon and energy. Two hundred Fnr+ transductants were
all Nir+. It was possible, therefore, that the NirA- and Fnr- phenotypes result from defects
in
a
single gene which has previously been designated either
nirA
or
fnr.
Lack
of
complementation between nirA and fnr strains
Two approaches were used in attempts to demonstrate complementation between
nirA
and
fnr
mutant strains. If the NirA- and Fnr- phenotypes result from defects in separate
genes, one would predict that abortive transductants would be generated when an
fnr
strain
was transduced with phage
P1
that had been propagated on
a
nirA
strain. Similarly, wild-
type transconjugants should be generated following the transfer of an
F’
plasmid carrying
a
defective
nirA
gene into an
fnr
strain.
A lysate of avirulent bacteriophage PI kc, generated by growing strain ~1485~ aerobically
for 16 h in Lennox broth, was propagated on strain ~~242
nirA gal+
and used at a multi-
plicity of infection of 2 to transduce JRG861a
fnr gal
to Fnr+ or Gal+. Fnr+ transductants
were selected on glycerol/fumarate plates (Lambden
&
Guest, 1976). Colonies of three
different sizes had formed after 4 d of anaerobic growth. The frequency of occurrence
of
the
largest colony type was 3.9
to
8.0
per
los
phage particles compared with a frequency of
40 to 190 Gal+ transductants per
los
phage from the same experiments. Although far more
of the small and very small colonies were seen, similar numbers formed on control plates
spread with ~~~861a which had not been infected with Plkc. The large colonies from three
such experiments (60 isolates) were purified and tested for their ability to reduce nitrite and
to catalyse the formate hydrogenlyase reaction. All were Nir+Fnr+. Similarly, small and
very small colonies from both transduced and control samples of ~~~861a were all Nir-Fnr-.
Although it is impossible to state that abortive transductants were not seen in these experi-
ments, the very close proximity of the Nir- and Fnr- defects carried by the donor and
recipient strains has been confirmed.
The merodiploid ~~98
nirA trp/F’nirA+trp+
and five homogenotized Nir- derivatives
were incubated for 30 min in Lennox broth with strain J~G861a, and Trp+ transconjugants
were selected. The Nir and Fnr phenotypes of 100 purified transconjugants were determined
after their fertility had been established by transferring the
cysB+trp+
alleles into strain
c~65
cysB trp thyA.
Thirty transconjugants which had acquired the
F’nirAftrpf
plasmid
were wild-type for formate hydrogenlyase and nitrite reductase activities and formed
only large colonies during anaerobic growth on glycerol/fumarate plates. In contrast,
70
transconjugants which had inherited the
F’nirA trp+
plasmids were Nir-: formate
hydrogenlyase was either less active or absent and only a few large colonies formed on
glycerol/fumarate plates. These large colonies were subsequently shown to be Nir+Fnr+
recombinant derivatives of the original Nir-Fnr- transconjugants
:
their occurrence is
positive evidence that the F’ plasmids carried a defective
nirA
gene rather than a small
internal deletion.
Characterization
of
nirA mutants
of
E.
coli
11
DISCUSSION
Although it has previously been suggested that cytochrome
c552
is involved in nitrite
reduction by
E.
coli,
the separation of these two proteins without a concomitant decrease
in the specific activity of the enzyme has now established that the cytochrome is not required
for the reduction of nitrite by NADH
in vitro
(see Cole
&
Wimpenny, 1968). However,
chlA
and
chlB
mutants are derepressed for both proteins in the absence of nitrate or nitrite,
and
nirA
mutants which carry a defect in a single gene lack both the enzyme and the
cytochrome. Common components therefore regulate the synthesis or assembly of both
proteins. The
nirA
gene is located at approximately minute 29, distal to
cysB
rather than
adjacent to
hemA
as shown on the current linkage map (Bachmann
et
al.,
1976). The low
frequency of recombination and lack of complementation between a NirA- strain and an
Fnr- strain indicate that the
nirA
and
fnr
genes are probably identical.
Several possible explanations for the pleiotropic phenotypes of
nirA
mutants have been
considered. One is that all
nirA
strains carry highly polar mutations in an operator proximal
gene of an operon which codes for nitrite reductase and cytochrome
~552.
This is unlikely,
however, because the rates of synthesis of these two proteins are not always coordinately
regulated as would be expected if they were under the control
of
a common promoter,
operator or attenuator (see Table
5
and Cole
et al.,
1974). Furthermore, preliminary data
not presented here indicate that the probable structural gene for nitrite reductase,
nirB,
is
unlinked to
nirA.
Any model postulated to explain the regulation of nitrite reductase and cytochrome
c552
synthesis must be consistent with the observations that these proteins are induced during
anaerobic growth but repressed when cultures are aerated; that they are induced by low
concentrations of nitrate or nitrite, but not by high concentrations of these oxidants; and
that neither nitrate nor nitrite are required for them to be synthesized by
chlA
and
chlB
mutants (Cole
&
Wimpenny, 1968; Cole
et
al.,
1974; Table
5).
Furthermore, the model
should be compatible with the defects in fumarate reductase, nitrate reductase (EC 1 .6.6. l),
hydrogenase (EC 1 .12.7.1) and formate hydrogenlyase that were described by Lambden
&
Guest
(1976).
One possibility is that the
nirA+
gene codes for a positive regulator protein
that is essential for the synthesis of a variety of enzymes involved in anaerobic redox
reactions. It is further suggested that the activation of genes for nitrite reductase and cyto-
chrome
cjj2
by the
nirA+
product is prevented by another factor, B, which is involved in
nitrate reduction and is missing or defective in
chlA
and
chlB
strains. It is pertinent that
Pateman
&
Cove (1967) have postulated that the synthesis of nitrite reductase in
Aspergillus
nidulaiqs
is
also regulated by a component of nitrate reductase.
Alternative explanations of the data cannot be excluded because the dominance of a
nirA+
to a
nirA
allele is equally consistent with a post-translational function for the
nirA+
gene
product (protein A). For example, although the multiple defects that arise from an
fnr
or
a
nirA
mutation exclude the possibility that
nirA
is a structural gene for the affected enzymes,
they are consistent with the suggestion that
A
is an enzyme which is involved in the synthesis
of a prosthetic group
-
possibly a haem group
-
common to each of these enzymes. Until
the postulated proteins
A
and
B
have been isolated, it will be impossible to determine
how
effector molecules such as oxygen, nitrate and nitrite regulate anaerobic oxidation-reduction
processes.
We are grateful to Professors P.
H.
Clarke and
H.
L. Kornberg for suggesting possible
cytoplasmic roles for the
nirA
gene product; to Dr
J.
R. Guest for extensive discussions and
for providing the
fnr
mutant strain, and to Miss
P.
Branson for excellent technical assistance.
B.
M.
N.
was supported by a Research Studentship from the Science Research Council.
12
B.
M.
NEWMAN AND
J.
A.
COLE
REFER
BACHMANN, B.
J.,
Low, K. B.
&
TAYLOR, A. L.
(1976).
Recalibrated linkage map of
Escherichia
coli
K12.
Bacteriological Reviews
40, 116-167.
COLE,
J.A.
&
RITTENBERG,
S.
C.
(1971).
A com-
parison of respiratory processes in
Spirillum
volutans, Spirillum itersonii
and
Spirillum serpens.
Journal of General Microbiology
69, 373-383.
COLE,
J.
A.
&
WARD,
F.
B.
(1973).
Nitrite reductase-
deficient mutants of
Escherichia coli
~12.
Journal
of General Microbiology
76, 21-29.
COLE,
J.
A.
&
WIMPENNY,
J.
W.
T.
(1968).
Metabolic
pathways for nitrate reduction in
Escherichia coli.
Biochimica et biophysica acta
162, 39-48.
COLE,
J.
A.,
COLEMAN, K.
J.,
COMPTON, B. E.,
KAVANAGH,
B. M.
&
KEEVIL,
C. W.
(1974).
Nitrite and ammonia assimilation by anaerobic
continuous cultures of
Escherichia coli. Journal
of
General Microbiology
85, 11-22.
DEMEREC,
M.,
ADELBERG,
E.
A.,
CLARK, A.
J.
&
€€ARTMAN,
P. E.
(1966).
A
proposal for a uniform
nomenclature in bacterial genetics.
Genetics
54,
FUJITA,
T.
(1966).
Studies on soluble cytochromes in
Enterobacteriaceae. I. Detection, purification and
properties of cytochrome c552 in anaerobically
grown cells.
Journal of Biochemistry
60, 204-215.
FUJITA,
T.
&
SATO,
R.
(1966).
Studies on soluble
cy
t
o
chromes in En tero bac
t
er iaceae
.
IV
.
Possible
involvement of cytochrome c552 in anaerobic
nitrite metabolism.
Journal of Biochemistry
60,
FUJITA,
T.
&
SATO, R.
(1967).
Studies on soluble
cytochromes in Enterobacteriaceae.
V.
Nitrite-
dependent gas evolution in cells containing cyto-
chrome c552.
Journal of Biochemistry
62, 230-238.
HADDOCK,
B.
A.
&
KENDALL-TOBIAS,
M.
W.
(1975).
Functional anaerobic electron transport linked to
the reduction of nitrate and fumarate in mem-
branes from
Escherichia coli
as demonstrated by
quenching
of
atebrin fluorescence.
Biochemical
Journal
152, 655-659.
JACOB, F.
&
MONOD,
J.
(1961).
Genetic regulatory
mechanisms in the synthesis of proteins.
Journal
of Molecular Biology
3, 318-356.
KAVANAGH,
B.
M.
&
COLE,
J.
A.
(1976).
The regu-
lation of nitrogen metabolism in batch and con-
tinuous cultures of
Escherichia coli:
facts and
artefacts. In
Continuous Culture
:
Applications and
New Fields,
Proceedings of the 6th International
Symposium on Continuous Culture of Micro-
organisms, pp.
184-194.
Edited by
A.
C.
R.
Dean,
61-76.
691-700.
ENCES
D.
C. Ellwood, C.
G.
J.
Evans and
J.
Melling.
Chichester
:
Ellis Horwood.
KROGER, A.
(1977).
Phosphorylative electron trans-
port with fumarate and nitrate
as
terminal
hydrogen acceptors.
Symposia of the Society for
General Microbiology
27, 61-93.
LAMBDEN, P.
R.
&
GUEST,
J.
R.
(1976).
Mutants
of
Escherichia coli
~12
unable to use fumarate as an
anaerobic electron acceptor.
Journal of General
Microbiology
97, 145-1 60.
LENNOX,
E.
S.
(1955).
Transduction
of
linked genetic
characters
of
the host by bacteriophage
P1.
Virology
1, 190-206.
MACGREGOR, C.
H.
(1975).
Synthesis of nitrate
reductase components in chlorate-resistant mu-
tants of
Escherichia coli. Journal of Bacteriology
MATSUBARA, T.
(1971).
Studies on denitrification.
XIII.
Some properties of the N,O-anaerobically
grown cell.
Journal of Biochemistry
69, 991-1001.
NEWMAN, B. M.
&
COLE, J.A.
(1977).
Lack of
a regulatory function for glutamine synthetase
protein in the synthesis of glutamate dehydro-
genase and nitrite reductase
in
Escherichia coli
K12.
Journal
of
General Microbiology
98,369-377.
O’HARA,
J.,
GRAY, C., PUIG,
J.
&
PICHINOTY,
F.
(1967).
Defects in formate hydrogenlyase in
nitrate-negative mutants of
Escherichia coli. Bio-
chemical and Biophysical Research Communications
PATEMAN,
J.
A.
&
COVE,
D.
J.
(1967).
Regulation of
nitrate reduction in
Aspergillus nidulans. Xature,
London
215, 1234-1237.
PAYNE, W.
J.
(1973).
Reduction of nitrogenous
oxides by microorganisms.
Bacteriological Reviews
PAYNE, W.
J.,
RILEY, P.
S.
&
Cox, C. D.
(1971).
Separate nitrite, nitric oxide and nitrous oxide
reducing fractions from
Pseudomonas perfecto-
marinus. Journal of Bacteriology
106, 356-361.
VENABLES,
W.
A.,
WIMPENNY,
J.
W.
T.
&
COLE,
J.
A.
(1968).
Enzymic properties of a mutant of
Escherichia coli
~12
lacking nitrate reductase.
Archiv fur Mikrobiologie
63, 117-121.
Wu, T. T.
(1966).
A model for three-point analysis
of random general transduction.
Genetics
54,
YAMANAKA,
T.
&
OKUWUKI, K.
(1974).
Cyto-
chromes. In
Microbial Iron Metabolism,
-4
Com-
prehensive Treatise,
pp.
349-400.
Edited by
J.
B.
Neilands. New
York
:
Academic Press.
121, 1117-1121.
28,951-957.
409-452.
405-41
0.
