INTERNATIONAL JOURNAL OF SYSTEMATIC BACTERIOLOGY, J a n . 1987, p. 43-51
0020-7713/87/0143-Q9$02.00/0
Copyright 0 1987, International Union of Microbiological Societies
Vol. 37, No. 1
Azospirillurn halopraeferens s p . nov. a Nitrogen-Fixing Organism
Associated with Roots of Kallar Grass (Leptochloa fusca (L.)
Kunth)
BARBARA REINHOLD,l* THOMAS HUREK,l ISTVAN FENDRIK,l BRUNO POT,2 MONIQUE GILLIS,2
KAREL KERSTERS,2 STEFAAN THIELEIvIANS,~AND JOZEF DE LEY2
Institute of Biophysics, University of Hannover, 0-3000 Hannover 21, Federal Republic of Germany, and Laboratorium
voor Microbiologie en Microbiele Genetica, Rijksuniversiteit, B-9000 Gent, Belgium2
Among the nitrogen-fixing bacteria associated with the roots of Leptochloa fusca (L.) Kunth in salt-affected
soils in the Punjab region of Pakistan, we found a homogeneous group of eight diazotrophs. Cells are vibrioid
to S shaped, are motile by one polar flagellum, and produce granules of poly-P-hydroxybutyrate. They have
a respiratory type of metabolism, show microaerophilic growth when fixing nitrogen, grow well on salts of
organic acids, and can also use fructose and mannitol. On nitrogen-free semisolid media, they require biotin,
utilize mannitol, but not glucose or sucrose, and cannot acidify glucose aerobically or anaerobically. Optimal
growth occurs at 0.25 % NaCl and 41OC. Deoxyribonucleic acid (DNA)-ribosomal ribonucleic acid (rRNA)
hybridizations show that the organisms belong to the Azospirilhm rRNA branch, where they cluster together
with Azospirilhm amazonense. They form a phenotypically and protein electrophoretically homogeneous group
of bacteria, clearly distinct from Azospirillum amazonense, Azospirillum lipoferum, and Azospirillum brasilense.
As no DNA-DNA binding was found with any of the three Azospirillum species, we propose a fourth
Azospirilhm species for this group of isolates. Because of better growth at increased NaCl concentrations, we
named the new species Azospirillum halopraeferens, Strain Au 4 (= LMG 7108) is the type strain, which has
been deposited at the Deutsche Sammlung von Mikroorganismen, Gottingen, Federal Republic of Germany, as
DSM 3675.
Several genera of nitrogen-fixing bacteria are associated
with the roots of various plants (17,24). In the rhizosphere of
tropical and subtropical grasses, bacteria of the genus
Azospirillum are widely distributed (16). It is generally
accepted that these bacteria can enhance the growth of the
plant, and it has been suggested that nitrogen fixed by the
bacteria can be transferred to the plant (15, 35, 36).
With the increasing occurrence of saline soils or saline
irrigation water, there might arise interest in plant-associated
diazotrophs showing salt tolerance or even enhanced growth
in higher saline conditions. Salt tolerance has been reported
for the N2-fixing Campylobacter nitrofigilis (34) and for some
rhizobia (42).
Leptochloa fusca (L.) Kunth (Kallar grass) is a salttolerant grass used as a pioneer plant in Pakistan on saltaffected low-fertility soils. Three to four cuttings a year are
taken without the addition of any nitrogenous fertilizer (31).
During a study of the population of diazotrophs associated
with Kallar grass grown in the Punjab region of Pakistan, we
found high numbers of nitrogen-fixing bacteria in and on the
roots. Some of these isolates resembled Azospirillum spp. in
morphology and microaerophilic N2-dependent growth, but
could not be assigned to one of the known species (38).
The genus Azospirillum was initially described with two
species, Azospirillum lipoferum and Azospirillum brasilense
(43). Deoxyribonucleic acid (DNA)-ribosomal ribonucleic
acid (rRNA) hybridizations showed that both species constitute a tight cluster on a separate rRNA branch in rRNA
superfamily IV sensu De Ley (5, 10). In this technique, Tm(el
(the temperature [in “C] at which 50% of the DNA-rRNA
hybrid is denatured) is the decisive taxonomic parameter for
* Corresponding author.
assessing taxonomic relatedness at generic and suprageneric
levels (12, 13, 25).
Recently, a new species, Azospirillum amazonense, was
proposed (19, 30) mainly on the basis of phenotypic data.
The results of rRNA competition experiments (20) confirmed
that A . amazonense is a close relative of A . brasilense and A .
lipoferum. The proposals of a fourth Azospirillum species
(‘ ‘Azospirillum seropedicae”) (1) was not retained because
analogous DNA-rRNA homology studies indicated that
these organisms could not belong in the Azospirillum genus.
They were subsequently placed in a new genus, Herbaspirillum (2).
It is evident that the DNA-rRNA hybridization technique
is the obvious means to reveal the taxonomic status of our
nitrogen-fixing bacteria isolated from Kallar grass. In this
work, we measured their Tm(e)values versus [14C]rRNA
from A . brasilense ATCC 29145T. Since no Tm(e)
values are
yet known for A . amazonense and Herbaspirillum
seropedicae, representative strains of these species were
also included. Phenotypic analysis, comparative gel electrophoresis of sodium dodecyl sulfate (SDS)-solubilized proteins, and DNA-DNA hybridizations revealed the relationships within our group of Kallar grass isolates and with the
other Azospirillum species. Our studies led us to propose a
fourth Azospirillum species, Azospirillum halopraeferens,
which we shall describe here.
MATERIALS AND METHODS
Isolation. Roots of Leptochloa fusca were collected from
fields on a saline-sodic soil in the area of Punjab, Pakistan.
Rhizoplane dilutions were used to inoculate tubes of two
different semisolid N-free malate media. For enrichment and
isolation, synthetic malate (SM) medium (37) without NH&l
43
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Mon, 31 Jul 2017 16:00:23
44
REINHOLD ET AL.
INT.J. SYST.BACTERIOL.
TABLE 1. Strains useda
Species
Strains assigned by us to
Azospirillurn halopraeferens
Reference and other strains:
Azospirillum lipoferum
Azospirillurn brasilense
Azospirillum arnazonense
Herbaspirillurn seropedicae
Strain designation:
original or as received
Other strain designations
Au 4T
Au 2
Au 5
Au 7
Au 9
Au 10
Au 11
Au 12
LMG 7108T, DSM 3675T
LMG 7107
LMG 7109
LMG 7110
LMG 7111
LMG 7112
LMG 7113
LMG 7114
DSM 1691T
SpBr 17
DSM 1690T
ATCC 29145T
Y lT(Hannover)"
YIT (Gent)'
DSM 2787T
Y13
Y9
Z67T
278
SP 59bT, ATCC 29707T
LMG 1264, ATCC 29709
LMG 1263T, ATCC 29145=, SP 7T
see DSM 1690r
LMG 7115T
LMG 65OST, Am 14T,J ATCC 35119T
LMG 7117T
LMG 6510, Am27d
LMG 6511, Am23d
ATCC 35892T
ATCC 35893
Received fromb:
Our isolate
Our isolate
Our isolate
Our isolate
Our isolate
Our isolate
Our isolate
Our isolate
DSM
J. Dobereiner
DSM
ATCC
J. Dobereiner
J. Dobereiner
DSM
J, Dobereiner
J, Dobereiner
J. Dobereiner
J. Dobereiner
ATCC, American Type Culture Collection, Rockville, Md. ; DSM, Deutsche Sammlung von Mikroorganismen, Gottingen, Federal Republic of Germany;
LMG, Culture Collection Laboratorium Microbiologie, Gent, Belgium; J. Dobereiner, EMBRAPA-PNPBS, CEP 23.851, Seropedica Km 47, Rio de Janeiro,
Brazil.
Our isolates were obtained by B. Reinhold from the root surface of Kallar grass Leptochloafusca (L.j Kunth, grown in the Shakot area, Punjab, Pakistan.
A subculture of Y1 was obtained from J. Dobereiner independently in Gent and Hannover. From strain Y1 (Gent) we obtained a dry (tl) and a moist (t2j colony
type after freeze-drying.
Strain numbers from reference 19.
and yeast extract was supplemented with 10 ml of a vitamin
solution sterilized by filtration per liter. The vitamin solution
contained (per liter of distilled water) the following: 200 mg
of p-biotin, 40 mg of calcium pantothenate, 200 mg of
rnyo-inositol, 40 mg of niacinamide, 20 mg of p aminobenzoic acid, 40 mg of pyridoxine hydrochloride, 20
mg of riboflavin, and 4 rng of thiamine hydrochloride. The
second synthetic malate medium was adapted to the salt
concentrations commonly found in the soil saturation extract
of the Kallar grass site and contained (in 1 liter of distilled
water) the following: 5.0 g of DL-malic acid, 4.8 g of KOH,
1.2 g of NaCl, 2.4 g of Na2S04, 0.5 g of NaHC03, 0.22 g of
CaC12, 0.25 g of MgS04 - 7H20, 0.17 g of K2SO4, 0.09 g of
Na2C03, 0.077 g of Fe(II1)-ethylenediaminetetraacetate
(EDTA), 0.13 g of K2HP04, 0.1 mg of biotin, 0.2 mg of
MnC12 - 4 H ~ 0 ~ 0mg
. 2 of H$03,0.15 mg of ZnC12, 0.02 mg
of CuC12 - 2H20, 2 mg of Na2Mo04 2H20, and 2.0 g of
agar. The final pH of the medium was 8.5; it was not further
adjusted. Malic acid, KOH, and agar were dissolved in
one-half of the total volume and autoclaved; the remaining
salts were sterilized by filtration after dissolving them in
one-half of the total volume and discarding the precipitate
after centrifugation.
Cultures for enrichment and isolation were incubated at
30°C. After growth had occurred as a subsurface pellicle,
cultures were assayed for nitrogenase activity by the acetylene reduction test. Tubes were incubated with 10% acetylene for 2 h, and the ethylene formed was measured in a gas
chromatograph with a flame ionization detector (Erba Science, Milano, Italy), fitted with a Porapak N column (0.8 m
by 2-mm inner diameter) at 90°C. For isolation, cultures
carrying out acetylene reduction were transferred to fresh
N-free semisolid medium. After 48 to 72 h of incubation,
parts of the pellicles were diluted in N-free SM medium, and
portions were used to seed plates of molten malate medium,
which deviated from the medium for enrichment by containing 0.8% agar instead of 0.2%, and af 20 mg of yeast extract
per liter. After incubation of the agar plates for 1 week,
single colonies were transferred to N-free semisolid medium.
After growth, pellicles were again purified by embedding in
the same way. Subsurface pellicles obtained from single
colonies were tested for nitrogenase activity and checked for
purity on potato infusion (BMS) agar (14) containing
bromothymol blue (0.5% [wthol] in 3 ml of ethanol per liter),
congo red agar (3), and tryptic soy agar (41). Pure cultures
were maintained at 41°C by biweekly transfer in semisolid
N-free SM medium supplemented with 0.25% NaCl and
adjusted to pH 7.2. The strains were also preserved in liquid
nitrogen as described previously (37).
Bacterial strains. All strains used are listed in Table 1. In
one culture of Azospirillum arnazonense YIT, two stable
colony types designated as t l (dry, irregular, and wrinkled)
and t2 (moist, circular, and convex) appeared on sucrose
medium plates (SM medium containing 1%sucrose instead
of malate and 1.5% agar, adjusted to pH 6.3).
Morphology and physiological tests. Cell dimensions were
determined by phase-contrast microscopy. Gram staining
was carried out by the method of Drews (18). The type of
flagellation was determined by electron microscopy by using
preparations negatively stained with uranyl acetate.
Unless indicated otherwise, the physiological tests were
carried out at 41°C for Azospirillurn halopraeferens and at
35°C for the other Azospirillurn species. The effects of
incubation temperature, pH, and NaCl concentration on
growth were evaluated by measuring nitrogenase activity of
cultures in semisolid N-free SM medium. Precultures were
grown in liquid SM medium with reciprocal shaking (100
rpm). Tubes (15 ml) containing 5 ml of medium were
inoculated with 30 ~1 of a washed-cell suspension with an
optical density of 0.1 at 578 nm. After 24 to 48 h of growth at
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Mon, 31 Jul 2017 16:00:23
VOL.37, 1987
AZOSPIRILLUM HALOPRAEFERENS SP. NOV.
the indicated temperature, pH, or NaCl concentration, cultures which were in the logarithmic growth phase were
incubated with 10% acetylene for 1.5 h, and the ethylene that
formed was determined by gas chromatograph. The values
(see Fig. 5 ) represent the means of four replicates. For
testing the effect of the pH on growth, the buffer concentration of the SM medium was doubled. Cultures not growing
within 48 h were checked for growth after prolonged incubation (up to 14 days). In all tests carried out with Azospirillum amazonense DSM 2787T in SM medium, malate was
replaced by 1%sucrose, and the pH was adjusted to 6.3. The
effect of NaCl on grawth was also studied in liquid SM
medium containing 0.5 g of NH4C1per liter and 0.1 g of yeast
extract per liter at 36°C with reciprocal shaking (100 rpm).
Growth in medium with and without the addition of 2.5 g of
NaCl was compared for the type strains of each species in
two replicates. Inocula were grown in the same liquid
medium as the test cultures. Growth was monitored by
means of optical density at 578 nm, and the doubling time, td,
was calculated.
For the following physiological tests we used a log-phase
inoculum, grown with reciprocal shaking (100 rpm) on liquid
SM medium. Cells were washed twice in SM medium free of
malate, N , and biotin, and were adjusted to an optical
density of 0.12 at 578 nm. Unless indicated otherwise, test
media were inoculated with 30 pl of the suspension. Growth
in the presence of 3% NaCl was tested by using N-free
semisolid SM medium. Growth in the form of a subsurface
pellicle was measured after 7 days of incubation. The biotin
requirement was tested on N-free biotin-free semisolid SM
medium with a biotin-containing medium as a control.
Growth was checked after 2 to 3 days. The sole carbon
sources for growth in N-free semisolid media were tested as
described by Krieg and Dobereiner (28) without the addition
of bromothymol blue. Glucose, mannitol, or sucrose was
added after sterilization by filtration. Growth responses were
measured after 3, 5 , and 7 days.
The sole carbon sources (see Table 3) for growth on media
containing ammonium sulfate as the nitrogen source were
tested by an auxanographic method following a procedure
described by Krieg and Dobereiner (28). A 10-ml volume of
a cell suspension with an optical density of 0.12 was used per
100 ml of molten medium. Disks (diameter, 7 mm) were
punched out from Whatman filters (GF/C), sterilized, dipped
into filter-sterilized aqueous solutions (5%) of the carbon
sources adjusted to pH 7.0, and placed on the surface of
solidified agar in plates. After 4.5 days of incubation, growth
as a visible zone of turbidity around the disks was checked.
Aerobic acidification of peptone-based glucose broth was
carried out as described by Krieg and Dobereiner (28). For
A. amazonense, the pH was adjusted to 6.3, and bromopheno1 blue was replaced by 15 mg of chlorophenol red per liter.
The development of a yellow color after 4 days of incubation
was considered to be a positive reaction.
The anaerobic acidification of glucose or fructose was
tested by the method of Tarrand et al. (43) by using microtiter plate broth. For A . amazonense, the medium was
modified as described above. Cultures were checked for the
development of a yellow color after 2 weeks of incubation.
Acidification of carbohydrate-containing media (see Table 3)
was tested by using a modification of the method described
by Tarrand et al. (43). In microtiter plate broth, bromothymol blue was replaced by 50 mg of chlorophenol red per
liter. The carbohydrate-containing media were dispensed in
sterile 15-ml tubes (1 ml per tube). After inoculating each
tube with 30 pl of washed-cell suspension, acidification was
45
considered to be positive if within 4.5 days of incubation the
pH had dropped to 5.6 as indicated by chlorophenol red.
Tests for nitrate reductase and denitrification were carried
out as described by Smibert and Krieg (41). The basal
medium was SM medium supplemented with 0.2% agar.
Nitrite was determined by a quantitative colorimetric assay
(41); production of gas was taken as evidence for denitrification. Cultures were monitored daily for the occurrence of
nitrite and gas production. Tetramethyl-p-phenylenediamine
was used to test oxidase (41). Urease was tested on
Christensen urea agar (41). The presence of poly-phydroxybutyrate was determined by the method of Smibert
and Krieg (41).
PAGE of soluble cell proteins. Bacteria were grown in one
or two Roux flasks on a medium containing (in 1 liter of
0.00625 M phosphate buffer) 3.0 g of papaic digest of
soymeal (E. Merck AG, Darmstadt, Federal Republic of
Germany), 17 g of pancreatic digest of casein (Oxoid Ltd.,
London, England), 5.0 g of NaC1, and 1.0 g of sucrose (final
pH 7.0). The same medium with final pH 6.3 was used for A .
amazonense. Protein samples were prepared from whole
cells by SDS treatment as described earlier (27). SDSpdyacrylamide gel electrophoresis (PAGE) was performed
by the method of Laemmli (29), with small modifications
described previously (27).
Preparation of high-molecular-weight DNA. DNA was prepared by the method of Marmur (32) modified as follows. (i)
TES buffer (150 mM NaC1, 20 mM Tris [tris(hydroxymethyl)aminomethane], 50 mM EDTA) was used as the lysis
buffer. (ii) A proteinase K treatment at 50 pg/ml for 2 to 3 h
at 37°C was carried out to replace the NaC104 treatment. (iii)
We used a mixture of chloroform-phenol (1:l [vol/vol]) for
deproteinization. The phenol was freshly distilled, saturated
with 100 mM Tris buffer, and contained 0.15% (wt/vol)
8-hydroxyquinoline. The DNA for DNA-rRNA hybridizations was further purified on a CsCl gradient.
DNA-DNA hybridizations. The degree of binding, a quantitative measurement of the homology between DNAs of a
pair of strains, was determined spectrophotometrically from
the initial renaturation rates by using the method of De Ley
et al. (6). The DNA concentration was ca. 50 pg/ml, and the
optimal renaturation temperature (TOR) in 2~ SSC ( I x SSC
is 0.15 M NaCl plus 0.015 M sodium citrate) was 82°C. A
Gilford 2600 spectrophotometer, equipped with a
thermostatted cuvet chamber and a Hewlett-Packard plotter
7225 A, was used to measure the renaturation rates. The
molecular complexity of the DNAs was calculated from the
renaturation reaction constants with the genome size of
Escherichia coli B as a reference (21, 23).
DNA-r@NA hybridizations. For the fixation of singlestranded high-molecular-weight DNA on membrane filters,
we used the fixation procedure described by De Ley and
Tytgat (8) and Sartorius filters type SM 11358. After simulation of the hybridization step as described by De Smedt
and De Ley ( l l ) , each concentration of DNA on the filter
was determined chemically by the method of Richards (39).
The filters with DNA were preserved at 4°C in vacuo (7).
DNA-rRNA hybridizations were carried out as described
previously (10, 13, 25). Two parameters were measured: (i)
T,(,,, which is the temperature at which 50% of the hybrid is
denatured and (ii) the percentage of labeled rRNA duplexed
(in micrograms per 100 pg of DNA fixed on the filter after
ribonuclease treatment). Both parameters are derived from
the melting curve of the hybrids.
DNA base composition, The average guanine-plus-cytosine
(G + C) content of the DNA from the strains investigated
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Mon, 31 Jul 2017 16:00:23
INT. J. SYST. BACTERIOL.
REINHOLD ET AL.
46
TABLE 2. Results from DNA-rRNA and DNA-DNA hybridizations with [14]rRNAfrom Azospirillum brasilense ATCC 29145T
% DNA binding with DNA from:
rRNA
binding (%)
Strains used
Azospirillum
halopraeferens Au 4T
67.5
69.0
66.6
66.4
66.7
70.4
69.6
64.5
64.7
Azospirillum brasilense DSM 1690T
Azospirillum lipoferum SpBr 17
Azospirillum amazonense Y 1 (Gent)
Azospirillum amazonense Y 13
Azospirillum amazonense Y9
Azospirillum halopraeferens Au 4T
Azospirillum halopraeferens Au 5
Herbaspirillum seropedicae Z67=
Herbaspirillum seropedicae 278
<25"
ND
<25"
<25"
<25"
<25"
<25"
100
106
ND
ND
0.18
ND
0.03
0.07
0.06
0.12
0.07
0.03
0.03
81.6
ND'
73.4
75.2
75.3
74.8
74.0
61.O
59.2
Azospirillum
amazonense
YIT (Gent)
100
63
64
c25
<25
ND
ND
25% is the lower reliability border of the method used.
ND. Not determined.
was measured by thermal denaturation (9) and was calculated by the equation of Marmur and Doty (33) as modified
by De Ley (4).
RWULTS AND DISCUSSION
A new group of nitrogen-fixing spirillumlike bacteria was
isolated from Kallar grass roots. Their negative gram reaction, the vibrioid to S-shaped cells, the cell diameter of about
1 pm, the motility by means of one polar flagellum, the
presence of poly-fbhydroxybutyrate granules, the posses-
sion of a respiratory type of metabolism, the ability of
denitrification, the preference for salts of organic acids as
carbon sources, and the ability to utilize and produce acid
from some sugars, indicated that they could be related to the
Azospirillum genus. Furthermore, they were able to fix NZ,
as shown unequivocally by 15N2incorporation with strain Au
4T (38). Nz-dependent growth was microaerophilic.
Proof for the assignment of the rhizosphere organisms
from Kallar grass to the genus Azospirillum was given by the
T,(e) values of the hybrids formed between their DNA and
14C-labeledrRNA from A . brpsilense ATCC 29145=.
z
c
3
4
'I,
E
cn
3
a
4
Y
.-C
3
i!
Y
3
v)
a
I
I
i
72!
76.8
L
74.4
71.8
1
6d
I
66.!
HERBASPIRILLUM SEROF'EDICAE *
ITHER rRNA- SUPERFAMILIES
60
FIG. 1. SimplifiedrRNA Cistron similarity dendrogram of part of rRNA superfamily IV, based on Tm(,)sof DNA-rRNA hybrids (25). The
vertical lines represent labeled rRNAs from reference strains (dashed lines were used for partly unpublished results from B. Pot, M. Gillis,
and J. De Ley). The thick bars indicate the range which we observed within a given species or a small group. Abbreviations: FLAV.,
Flwobacterium; R . , Rhizobium; C1, cluster; RHPS., Rhodopsepdomonas; A., Azospirillum.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Mon, 31 Jul 2017 16:00:23
VOL.
37, 1987
AZOSPIRILLUM HALOPRAEFERENS SP. NOV.
47
A. HALOPRAEFERENS
A. AMAZONENSE
A. LIPUFERUM
A. BRASILENSE
FIG. 2. Normalized SDS-PAGE patterns of eight Azospirillum halopraeferens strains, six Azospirillum amazonense strains, the type strain
of Azospirillum brasilense, and a representative strain of Azospirillum lipoferum.
DNA-rRNA hybridizations. To detect the generic relationship of Azospirillum halopraeferens and to confirm the
generic status of Azospirillum amazonense, DNAs of five
representative strains were hybridized with labeled rRNA
from Azospirillum brasilense ATCC 29145T. The results of
the hybridizations [expressed as Tm(e)and the percentage of
rRNA binding] are compiled in Table 2. Figure 1 represents
a dendrogram based on Tm(el.Our results concern part of
rRNA superfamily IV (10, 11, 22, 25). Both strains of
Azospirillum halopraeferens, Au 4T and Au 5 , as well as the
strains of Azospirillum amazonense, YIT (Gent), Y9, and
Y13, are located at the same level on the Azospirillum rRNA
branch ( l l ) , which forms a trinity together with the rRNA
branches of Rhodospirillum rubrum and some Aquaspirillum
species (10). Each of these branches deserves at least a
separate generic rank. With Tm(e)values ranging from 73.4 to
75.3"C, both Azospirillum halopraeferens strains are quite
distinct from the Azospirillum brasilense-Azospirillum
lipoferum cluster [T,(,) values of 80.0 to 82.5"Cl which were
located by De Smedt et al. (10) and which are their genotypically closest relatives. No other organisms were found on
this rRNA branch. The Tm(e) value for Azospirillum
brasilense DSM 1690T (81.6"C) was determined here as a
control on the rRNA from Azospirillum brasilense ATCC
29145T and was quite comparable to the previously reported
value of 82.5"C (10). The two representative strains of
Herbaspirillum seropedicae Z67T and 278 did not belong in
rRNA superfamily IV [T,,,, values of 59.2 and 61.O0C], but
in rRNA superfamily I11 (B. Pot, M. Gillis, and J. De Ley,
unpublished data).
Comparison of protein electrophoregrams. Figure 2 shows
the SDS-PAGE patterns of all strains investigated. The eight
Azospirillum halopraeferens strains have almost indistinguishable protein electrophoregrams, indicating that they
constitute a homogeneous group of bacteria. It is our experience (26) that bacteria with almost identical protein patterns possess a high genome similarity. The electrophoregrams of Azospirillum halopraeferens are clearly dif-
ferent from those of Azospirillum lipoferum, Azospirillum
brasilense, and Azospirillum amazonense.
Because phenotypic and genotypic data of the type strain
of Azospirillum amazonense Y 1 were obtained with subcultures which were independently obtained from J. Dobereiner
and because strain YIT (Gent) showed two different and
stable colony types after lyophilization, all of our subcultures of Azospirillum amazonense YIT were compared in
SDS-PAGE with the reference strain DSM 2787T. They
displayed nearly identical protein electrophoregrams (Fig.
2), indicating that both colony types of strain YIT (Gent)
were only colony morphology variants. The SDS-PAGE
patterns of Azospirillum amazonense Y9 and Y13 were very
similar to the pattern of the type strain.
DNA-DNA hybridizations. Because of the protein electrophoretic homogeneity of the Azospirillum halopraeferens
strains, it was sufficient to investigate only two representative strains of this cluster. As expected, we found 100%
DNA-DNA binding between strains Au 4T and Au 5 (Table
2). No meaningful percentage of DNA-DNA binding could
be detected between DNA of Azospirillum halopraeferens
Au 4T and DNA from representative strains of the three
other species of the genus Azospirillum (Table 2). The other
results in Table 2 were included as controls. As expected
(19), no meaningful percentage of DNA-DNA binding could
be detected between Azospirillurn brasilense ATCC 29145T
and Azospirillurn lipoferum SpBr 17. Between Azospirillum
amazonense YIT (Gent) and Y13 we found 63% DNA-DNA
binding, which is in very good agreement with the value of
67% as obtained with the S1 endonuclease procedure (19).
As could be expected from the electrophoregrams, we found
a similar value of 64% binding between Azospirillum
amazonense Y9 and YIT (Gent).
The molecular complexities of the DNA of Azospirillum
halopraeferens Au 4T and Au 5 were not significantly different; the mean value is (5.4 5 0.4) x lo9 (average of 16
measurements). This is in the range of the values we found
for the molecular complexities of Azospirillum brasilense
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Mon, 31 Jul 2017 16:00:23
48
INT. J. SYST.BACTERIOL.
REINHOLD ET AL.
FIG. 3. Electron micrograph of a negatively stained Azospirillum
halopraeferens Au 4T cell derived from a log-phase culture in liquid
SM medium supplemented with 0.25% NaCl; pH adjusted to 7.2.
DSM 1690T (5.6 x 109 and of Azospirillum amazonense
(mean value of 5.4 x lo9for strain YIT [Gent], Y13, and Y9).
Compared with other values (21), these molecular complexities are very high. Whether this is due to chromosomal
DNA only or to the additional large plasmid DNA, as
described previously (44), cannot be unequivocally revealed
by this method which measures the total amount of unique
sequences.
The arguments that the Kallar grass root strains constitute
a new species in the genus Azospirillum are summarized as
follows. (i) The results of DNA-DNA homology determinations showed that the new Azospirillum organisms are different from Azospirillum amazonense since we did not
detect meaningful DNA-DNA binding values between both
groups. (ii) The protein gel electrophoregrams of the Kallar
grass root organisms are (almost) identical and quite different from those of Azospirillum amazonense. (iii) Compared with all known Azospirillum species, the isolates from
Kallar grass roots possess two unique properties: a high
optimal temperature for growth (41°C) and a preference for
higher NaCl concentrations.
Therefore, the bacteria we isolated from the roots of
Kallar grass should be included in a new and fourth species
of Azospirillum. Slightly halophilic bacteria grow best at 2 to
5% NaCl (40). Our isolates have a lower optimal salt
concentration (0.25% NaCl) and thus cannot be called
halophiles. We propose for this new species the name
Azospirillum halopraeferens sp. nov. A detailed description
of the type strain and seven other representative strains is as
follows.
Azospirillum halopraeferens sp. nov. Azospirillum
halopraeferens (ha. lo' prae. fe. rens. Gr. n. hals, halos, salt,
the sea; L. v. praeferre, to prefer; N. L. part. adj. salt
preferring). The cells are gram negative, vibrioid to S
shaped; only few helical cells occur in alkaline media. Motile
in liquid medium with a rapid cork screwlike motion by one
polar flagellum (Fig. 3).
Fluorescent pigments are not produced. No growth on
plates of BMS medium or congo red medium. Good growth
on tryptic soy agar, forming cream-colored, circular, flat
colonies with an entire margin. When culivated on N-free
semisolid medium at 30 or 35"C, young cultures which have
just developed a fine pellicle contain cells of small cell width,
averaging from 0.7 to 0.9 pm. When growing at 41°C in
medium supplemented with 0.25% NaCl at pH 7.2, the width
of young cells is 1to 1.2 pm. In both growth conditions, cells
tend to become slightly wider, longer, and S shaped in older,
more alkaline cultures. Such cells seldom exceed a length of
5 km. Only very few long, helical cells are observed. The
occurrence of S-shaped cells is also pronounced in the late
log phase on liquid SM medium, containing 0.25% NaCl
(Fig. 4).
FIG. 4. Aspect of Azospirillum halopraeferens Au 4T by phasecontrast microscopy. Cells from late log-phase culture in liquid SM
medium supplemented with 0.25% NaCl; pH adjusted to 7.2.
Cell size: 0.7 to 1.4 pm in diameter and 2.4 to 5 pm in
length. On semisolid N-free malate medium they form veillike pellicles which first develop several millimeters below the medium surface and later move to the surface.
They produce intracellular granules of poly-P-hydroxybutyrate.
Chemoorganoheterotrophic. Respiratory type of metabolism with oxygen as the terminal electron acceptor. All
strains carried out acetylene reduction. Nitrate can be used
for denitrification; N2 can be fixed in microaerobic conditions. Oxidase and urease positive. Grow well on salts of
organic acids such as L-malate, succinate, fumarate, pyruvate, DL-lactate, P-hydroxybutyrate, and gluconate. Grow
on D-mannitol, glycerol, and D-ribose. Use and acidify
D-fructose.
The characteristics which differ among the eight strains
are the utilization of carbon sources and acidification of
carbohydrate-containing media (Table 3). On semisolid
TABLE 3. Utilization of carbon sources and acidification of
carbohydrate-containing media by Azospirillum halopraeferens
strains
% of strains
Feature
Sole carbon sources for growth
(by auxanographic method):
L-Malate, fumarate, succinate,
P-hydroxybutyrate, DL-lactate,
pyruvate, gluconate, D-mannitol,
glycerol, D-fructose, D-ribose
Citrate
L-Arabinose
Sorbitol, D-galactose, sucrose
Acid produced from:
D-Fructose
D-Ribose
L-Arabinose
D-x ylose
D-Galactose
L-Rhamnose
Dulcitol
Maltose
Sorbitol
myo-Inositol
D-Lactose
D-Mannitol, starch
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Mon, 31 Jul 2017 16:00:23
giving
positive
reactions
Strain(s) giving the
less common result
100
88
63
0
100
88
38
38
25
25
25
12
12
12
12
0
Au 12
Au 5, Au 9, Au 12
Au 11
Au 2, Au 7, Au 11
Au 2, Au 4T, Au 5
Au 7, Au 11
A u ~ Au
~ ,10
Au 7, Au 11
Au 4=
Au 4T
Au 4T
Au 12
VOL. 37, 1987
AZOSPIRILLUM HALOPRAEFERENS SP. NOV.
c
I
=
X
I
I
al
al
5
W
NaCl added ( g / l )
FIG. 5. Effect of NaCl concentration on N2 fixation and growth
of Azospirillum halopraeferens Au
(a),Azospirillum lipoferurn
DSM 1691T (A), Azospirillurn brusilense DSM 1690T (0),and
Azospirillum arnazonense DSM 2787T (0).
NaCl was added to
N-free semisolid SM medium containing 0.01% NaCl.
TABLE 4. Characteristics differentiating Azospirillum species
from each other"
Characteristic
Cell width (pm)
Cell length exceeding
5 pm dominating in
alkaline N-free
semisolid medium
Optimum temp for growth
("C)
Growth at:
pH 6.0
pH >6.8
Growth in presence of 3%
NaCl
Biotin requirement
Sole carbon sources for
growth in N-free
semisolid medium:
D-Glucose
D-Mannitol
Sucrose
Acidification of peptonebased glucose broth
Acid from glucose or
fructose broth
anaerobically
Nitrate reductase
Denitrification
DNA base composition
(mol% G + C)
A . lipoferum
A . bra- amazeA'
1.0-1.7
1.0-1.2 0.9-1.0
+
-
silense
nense
A . halopraeferens
0.7-1.4
-
-
37
37
35
41
+
+
+orw
+
-
(-Ib
d(-)
+ (+I
+
-
(-1
W
- (-)
-
(-1
-
+
+
+
-
+ (+I
+ (+I
WC
- (-1
+ (+I
-
+ or v (+)
-
+ (+I
d (+I
69-70
-
49
N-free medium, growth is produced on mannitol but not on
glucose or sucrose. Not capable of aerobic acid production
on peptone-based glucose medium. No acid production on
fructose or glucose media anaerobically. Biotin required for
growth. Better growth at 41 than at 35"C, and no growth at
44°C. They are able to grow at 20°C.
The effect of different NaCl concentrations on the acetylene reduction activity of isolate Au 4T is shown in Fig. 5.
Acetylene reduction activity and growth is optimal in N-free
SM medium supplemented with 0.25% NaC1. All eight
isolates grew better in 0.25% NaCl than with 0.01% or 0.75%
NaCl. After 7 days of incubation, growth as a subsurface
pellicle is observed in 3% NaC1, although after 48 h of
incubation acetylene reduction activity can not yet be detected. The addition of 0.25% NaCl to liquid SM medium
containing bound nitrogen also increases the growth rate.
The addition of 0.25% NaCl decreases the f d from 9.7 to 4.7
h for strain Au 4T. (Azospirillum amazonense DSM 2787T did
not respond to 0.25% NaCl [ f d = 3.3 h with and without
addition]. With Azospirillum brasilense DSM 1690T and
Azospirillum lipoferum DSM 1691T the f d increased slightly
by the addition of 0.25% NaCl, from 1.9 to 2.5 h and from 2.3
to 2.5 h, respectively.) All strains grow well in media at a pH
between 6.8 and 8.0 with an optimum at pH 7.2. The mol%
of G + C of DNA is 70. Source: isolated from root surface of
Leptochloa fusca (L.) Kunth grown on saline-sodic soils in
Shakot area, Punjab, Pakistan. The type strain is Au 4 (=
DSM 3675 = LMG 7108).
Description of the type strain. The description of the type
strain is the same as given above for the species, with the
following additions: grows on citrate and D-arabinose. Produces acid from D-ribose, D-xylose, L-rhamnose, maltose,
sorbitol, and myo-inositol. The DNA base composition is 70
mol%G + C.
Differentiation from other Azospirillum species. Phenotypically, the Kallar grass root organisms form a very homogeneous group which can be clearly differentiated from the
known Azospirillum species (Table 4). They differ from
Azospirillum lipoferum by their inability to acidify glucose or
fructose broth anaerobically or peptone-based glucose broth
aerobically, to grow on glucose as the C source in N-free
semisolid medium, and by the lack of typical pleomorphism.
They can be differentiated from Azospirillum brasilense by
the requirement of biotin and by the ability to use mannitol,
on which they grew only weakly within 3 days of incubation
in N-free semisolid medium, but showed better growth after
a prolonged incubation period of 5 to 7 days. They differ
from Azospirillum amazonense by their requirement for
biotin, the utilization of mannitol but not glucose and sucrose, the capability to denitrify, and the alkali tolerance.
The reference strains we used gave results which are in good
agreement with those reported in the literature.
ACKNOWLEDGMENTS
+(+I
d
d(+)
70-71 67-68
+
+
69-70
a Data from this study and from references 19, 30, and 42. + , Positive in
more than 90% of the strains; d, positive in 11 to 89% of the strains; -,
negative in more than 90% of the strains; w, scant growth; v, strain instability.
Data in parentheses were determined by us for reference strains:
Azospirillum lipoferum DSM 1691T,Azospirillum brasilense DSM 1690T,and
Azospirillum amazonense DSM 2787=.
Growth after prolonged incubation (5 to 7 days).
We are indebted to L. Luciano, Abteilung fur Zellbiologie und
Elektronenmikroskopie, Medizinische Hochschule Hannover, for
the preparation of the electron micrographs, and to J. Dobereiner,
EMBRAPA-PNPBS, CEP 23.851, Seropedica Km 47, Rio de
Janeiro, Brazil, for the gift of strains.
J.D.L. is indebted to the Fonds voor Geneeskundig Wetenschappelijk Onderzoek, Belgium for personnel grants, B.P. to the
Instituut tot Aanmoediging van het Wetenschappelijk Onderzoek in
Nijverheid en Landbouw, for a scholarship, and B.R. to the
Studienstiftung des deutschen Volkes, Federal Republic of Germany, for a scholarship.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Mon, 31 Jul 2017 16:00:23
50
INT. J. SYST.BACTERIOL.
REINHOLD ET AL.
LITERATURE CITED
1. Baldani, J. I., V. L. D. Baldani, M. J. A. M. Sampaio, and J.
Dobereiner. 1984. A fourth Azospirillum species from cereal
roots. An. Acad. Bras. Cienc. 56:365.
2. Baldani, J. I., V. L. D. Baldani, L. Seldin, and J. Dobereiner.
1986. Characterization of Herbaspirillum seropedicae gen.
nov., sp. nov., a root-associated nitrogen-fixing bacterium. Int.
J. Syst. Bacteriol. 36:8&93.
3. CBcares, E. A. R. 1982, Improved medium for isolation of
Azospirillum spp. Appl. Environ. Microbiol. 44:99&991.
4. De Ley, J. 1970. Reexamination of the association between
melting point, buoyant density, and chemical base composition
of deoxyribonucleic acid. J. Bacteriol. 101:738-754.
5. De Ley, J. 1978. Modern molecular methods in bacterial taxonomy: evaluation, application, prospects, p. 347-357. In Proceedings of the 4th International Conference on Plant Pathogenic Bacteria, vol. 1. Gilbert-Clarey, Tours, France.
6. De Ley, J., H. Cattoir, and A. Reynaerts. 1970. The quantitative
measurement of DNA hybridization from renaturation rates.
Eur. J. Biochem. 12:133-142.
7. De Ley, J., and J. De Smedt. 1975. Improvements of the
membrane filter method for DNA:rRNA hybridization. Antonie
van Leeuwenhoek J. Microbiol. Serol. 41:287-307.
8. De Ley, J., and R. Tytgat. 1970. Evaluation of membrane filter
methods for DNA-DNA hybridization. Antonie van
Leeuwenhoek J. Microbiol. Serol. 36:193-207.
9. De Ley, J., and J. Van Muylem. 1963. Some applications of
deoxyribonucleic acid base composition in bacterial taxonomy.
Antonie van Leeuwenhoek J. Microbiol. Serol. 29:344-358.
10. De Smedt, J., M. Bauwens, R. Tytgat, and J. De Ley. 1980. Intraand intergeneric similarities of ribosomal ribonucleic acid
cistrons of free-living, nitrogen-fixing bacteria. Int. J. Syst.
Bacteriol. 30:10&122.
11. De Smedt, J., and J. De Ley. 1977. Intra- and intergeneric
similarities of Agrobacterium ribosomal ribonucleic acid
cistrons. Int. J. Syst. Bacteriol. 27:222-240.
12. De Vos, P., M. Goor, M. Gillis, and J. De Ley. 1985. Ribosomal
ribonucleic acid cistron similarities of phytopathogenic Pseudomonas species. Int. J. Syst. Bacteriol. 35169-184.
13. De Vos, P., K. Kersters, E. Falsen, B. Pot, M. Gillis, P. Segers,
and J. De Ley. 1985. Comamonas Davis and Park 1962 gen.
nov., nom. rev. emend., and Comamonas terrigena Hugh 1962
sp. nov., nom. rev. Int. J. Syst. Bacteriol. 35:443453.
14. Dobereiner, J. 1980. Forage grasses and grain crops, p. 535-556.
In F. J. Bergersen (ed.), Methods for evaluating biological
nitrogen fixation. John Wiley & Sons, Ltd., Chichester, United
Kingdom.
15. Diibereiner, J. 1983. Ten years Azospirillum, p. 9-23. In W.
Klingmiiller (ed.), Azospirillum 11: genetics, physiology, ecology. Birkhauser Verlag, Basel.
16. Dobereiner, J., and J. M. Day. 1976. Associative symbioses in
tropical grasses: characterization of microorganisms and
dinitrogen fixing sites, p. 518-538. In W. E. Newton and C. J.
Nyman (ed.), Proceedings of the 1st International Symposium
on N2 Fixation, Washington State University Press, Pullman.
17. Dobereiner, J., J. M. Day, and P. J. Dart. 1972. Nitrogenase
activity and oxygen sensitivity of the Paspalum notatumAzotobacter paspali association. J. Gen. Microbiol. 71:103-116.
18. Drews, G. 1976. Mikrobiologisches Praktikum, 3rd ed. SpringerVerlag, Berlin.
19. Falk, E. C., J. Dobereiner, J. L. Johnson, andN. R. Krieg. 1985.
Deoxyribonucleic acid homology of Azospirillum amazonense
Magalhiies et al. 1984 and emendation of the description of the
genus Azospirillum. Int. J. Syst. Bacteriol. 35:117-118.
20. Falk, E. C., J. L. Johnson, V. L. D. Baldani, J. Diibereiner, and
N. R. Krieg. 1986. Deoxyribonucleic and ribonucleic acid homology studies of the genera Azospirillum and Conglomeromonas. Int. J. Syst. Bacteriol. 36:80-85.
21. Gillis, M., and J. De Ley. 1975. Determination of the molecular
complexity of double-stranded phage genome DNA from initial
renaturation rates. The effect of DNA base composition. J. Mol.
Biol, 98:447464.
22. Gillis, M., and J. De Ley. 1980. Intra- and intergeneric similarities of the ribosomal ribonucleic acid cistrons of Acetobacter
and Gluconobacter. Int. J. Syst. Bacteriol. 30:7-27.
23. Gillis, M., J. De Ley, and M. De Cleene. 1970. The determination
of molecular weight of bacterial genome DNA from renaturation
rates. Eur. J. Biochem. 12:143-153.
24. Haahtela, K., T. Wartiovaara, V. Sundman, and J. SkujinS.
1981. Root-associated N2 fixation (acetylene reduction) by Enterobacteriaceae and Azospirillum strains in cold-climate
spodosols. Appl. Environ. Microbiol. 41:203-206.
25. Jarvis, B. D. W., M. Gillis, and J. De Ley. 1986. Intra- and
intergeneric similarities between the ribosomal ribonucleic acid
cistrons of Rhizobium and Bradyrhizobium species and some
related bacteria. Int. J. Syst. Bacteriol. 36:129-138.
26. Kersters, K. 1985. Numerical methods in the classification of
bacteria by protein electrophoresis, p. 337-368. In M.
Goodfellow, D. Jones, and F. G. Priest (ed.), Computerassisted bacterial systematics. Acadmic Press, Inc. (London),
Ltd., London.
27. Kiredjian, M., B. Holmes, K. Kersters, I. Guilvout, and J.
De Ley. 1986. Alcaligenes piechaudii, a new species from
human clinical specimens and the environment. Int. J. Syst.
Bacteriol . 36:282-287.
28. Krieg, N. R., and J. Dobereiner. 1984. Genus Azospirillum
Tarrand, Krieg and Dobereiner 1979, 79*= (effective publication: Tarrand, Krieg and Dobereiner 1978, 967), p. 94-104. In
N. R. Krieg and J. G. Holt (ed.), Bergey’s manual of systematic
bacteriology, vol. 1. The Williams & Wilkins Co., Baltimore.
29. Laemmli, U. K. 1970. Cleavage of structural proteins during the
assembly of the head of bacteriophage T4. Nature (London)
227:680485.
30. Magalhges, F. M., J. I. Baldani, S. M. Souto, J. R. Kuykendall,
and J. Dobereiner. 1983. A new acid tolerant Azospirillum
species. An. Acad. Bras. Cienc. 55417430.
31. Malik, K. A., Y. Zafar, and A. Hussain. 1980. Nitrogenase
activity in the rhizosphere of Kallar grass (Diplachne fusca
(Linn.) Beauv.). Biologia 26:107-112.
32. Marmur, J. 1961. A procedure for the isolation of deoxyribonucleic acid from micro-organisms. J. Mol. Biol. 3:208-218.
33. Marmur, J., and P. Doty. 1962. Determination of the base
composition of deoxyribonucleic acid from its thermal denaturation temperature. J. Mol. Biol. 5109-118.
34. McClung, C. R., D. G. Patriquin, and R. E. Davis. 1983.
Campylobacter nitrofigilis sp. nov., a nitrogen-fixing bacterium
associated with roots of Spartina alterniJlora Loisel. Int. J.
Syst. Bacteriol. 33:605-612.
35. Okon, Y. 1985. Azospirillum as a potential inoculant for agriculture. Trends Biotechnol. 3:223-228.
36. Postgate, J. R. 1982. The fundamentals of nitrogen fixation.
Cambridge University Press, Cambridge.
37. Reinhold, B., T. Hurek, and I. Fendrik. 1985. Strain-specific
chemotaxis of Azospirillum spp. J. Bacteriol. 162:190-195.
38. Reinhold, B., T. Hurek, E.-G. Niemann, and I. Fendrik. 1986.
Close association of Azospirillum and diazotrophic rods with
different root zones of Kallar grass. Appl. Environ. Microbiol.
52520-526.
39. Richards, G. M. 1974. Modification of the diphenylamine reaction giving increased sensitivity and simplicity in the estimation
of DNA. Anal. Biochem. 57:369-376.
40. Schlegel, H. G., and H. W. Jannasch. 1981. Prokaryotes and
their habitats, p. 43-82. In M. P. Stan-, H. Stolp, H. G. Triiper,
A. Balows, and H. G. Schlegel (ed.), The prokaryotes. Springer-Verlag, Berlin.
41. Smibert, R. M., and N. R. Krieg. 1981. General characterization, p. 409-443. In P. Gerhardt, R. G. E. Murray, R. N.
Costilow, E. W. Nester, W. A. Wood, N. R. Krieg, and G. B.
Phillips (ed.), Manual of methods for general bacteriology.
American Society for Microbiology, Washington, D.C.
42. Sprent, J. I. 1984. Effects of drought and salinity on
heterotrophic nitrogen fixing bacteria and on infection of legumes by rhizobia, p. 295-302. In C. Veeger and W. E. Newton
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Mon, 31 Jul 2017 16:00:23
VOL.37, 1987
AZOSPIRILLUM HALOPRAEFERENS SP. NOV.
(ed.), Advances in nitrogen fixation research. Dr. W. Junk
Publishers, The Hague, The Netherlands.
43. Tarrand, J. J., N. R. Krieg, and J. Dobereiner. 1978. A
taxonomic study of the Spirillum lipoferum group, with descriptions of a new genus, Azospirillum gen. nov., and two species,
Azospirillum lipoferum (Beijerinck) comb. nov. and Azospiril-
51
lum brasilense sp. nov. Can. J. Microbiol. 24:967-980.
44. Wood, A. G., E. M. Menezes, C. Dijkstra, and D. E. Duggan.
1982. Methods to demonstrate the megaplasmids (or
minichromosomes) in Azospirillum, p. 18-34. In W. Klingmuller
(ed.), Azospirillum genetics, physiology, ecology. Birkhauser
Verlag, Basel.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Mon, 31 Jul 2017 16:00:23