Download Cloning, Expression in Escherichia coli, and Characterization of

Vol. 175, No. 21
JO(URNAL OF BACTERIOLOGY, Nov. 1993, P. 7056-7065
0021-9193/93/217056-1 0$02.00/0
Copyright © 1993, American Society for Microbiology
Cloning, Expression in Escherichia coli, and Characterization
of Cellulolytic Enzymes of Azoarcus sp.,
a Root-Invading Diazotroph
BARBARA REINHOLD-HUREK,' t* THOMAS HUREK,' t MARC CLAEYSSENS,2
AND MARC VAN MONTAGU'
Laboratorium
voor
Genetika' and Laboratorium voor Biochemie,2 Universiteit Gent,
B-9000 Ghent, Belgium
Received 25 May 1993/Accepted 2 September 1993
We screened members of a new genus of grass-associated diazotrophs (Azoarcus spp.) for the presence of
cellulolytic enzymes. Out of five Azoarcus strains representing different species, only in the endorhizosphere
isolate BH72, which is also capable of invading grass roots, was significant endoglucanase activity, in addition
to 13-glucosidase and cellobiohydrolase activity, present. Reducing sugars were readily released from
medium-viscosity carboxymethylcellulose (CMC), but neither CMC, cellulose filter strips, Avicel, cellobiose,
nor D-glucose served as the sole carbon source for growth of Azoarcus spp. Clones from a plasmid library of
strain BH72 expressed all three enzymes in Escherichia coli, apparently not from their own promoter. According
to restriction endonuclease mapping and subclone analysis, ,-glucosidase and cellobiohydrolase activities were
localized on a single 2.6-kb fragment not physically linked to a 1.45-kb fragment from which endoglucanase
(egl) was expressed. Two isoenzymes of endoglucanase probably resulting from proteolytic cleavage had pl
values of 6.4 and 6.1 and an apparent molecular mass of approximately 36 kDa. Cellobiohydrolase and
,I-glucosidase activity were conferred by one enzyme 41 kDa in size with a pl of 5.4, which we classified as an
unspecific exoglycanase (exg) according to substrate utilization and specificity mapping; hydrolysis of various
oligomeric substrates differentiated it from endoglucanase, which degraded substituted soluble cellulose
derivatives but not microcrystalline cellulose. Both enzymes were not excreted but were associated with the
surface of Azoarcus cells. Both activities were only slightly influenced by the presence of CMC or D-glucose in
the growth medium but were enhanced by ethanol. egl was located on a large transcript 15 kb in size, which
was detectable only in cells grown under microaerobic conditions on N2. Surface-bound exo- and endoglucanases with some unusual regulatory features, detected in this study in a strain which is unable to metabolize
cellulose or sugars, might assist Azoarcus sp. strain BH72 in infection of grass roots.
Cellulose is a constituent of lignocellulosic materials of plant
origin and is composed of D-glucopyranosyl units linked by
3-1,4 bonds. For its microbial conversion complex, batteries of
several enzymes are required (52). At least three different
classes of hydrolytic enzymes are thought to participate: endo1,4-p-glucanase (EC 3.2.1.4), exo- l,4-cellobiohydrolase (EC
3.2.1.91), and 1-glucosidase (EC 3.2.1.21).
Cellulose degradation has been reported to occur concomitantly with nitrogen fixation in aerobic, shipworm-associated
bacteria (57) and in anaerobic freshwater mud and soil isolates
(34). Among plant-associated nitrogen-fixing bacteria, cellulolytic activity is present in the actinomycete Franckia sp. (49)
and in Rhizobium sp. (38, 43). In the legume-Rhizobium
symbiosis, ultrastructural studies provided evidence that rhizobial cellulases may be involved in various steps of the infection
process (8, 9). The grass-associated diazotroph Azospirillulm
brasilense, although able to colonize the interior of grass roots
(35, 51), is pectinolytic (55), but so far no cellulolytic activity
has been reported to be present. Cellulose only supported
growth of Azospirillum cells in coculture with cellulolytic
bacteria (22).
Corresponding author.
t Present address: Max-Planck-Institut fur Terrestrische Mikrobiologie, Karl-von-Frisch-Str., D-35043 Marburg, Germany.
t Present address: Pfarracker 5, D-35043 Bauerbach, Germany.
We were interested in the cellulolytic features of Azoarcus
spp., a new genus of gram-negative, nitrogen-fixing bacteria
colonizing grass roots. Taxonomic studies located them in the
beta subclass of the Proteobacteria, in which they cluster in five
groups differing at the species level (47). Two of them were
proposed as new named species, Azoarcus indigensT and
Azoarclis commlunis (47). Members of the genusAzoarcus were
found to be in root-zone-specific association with Kallar grass
(Leptochloa fiusca (L.) Kunth), in which they were isolated
from the root interior (46). Kallar grass is grown as a pioneer
plant on salt-affected, frequently flooded, low-fertility soils in
the Punjab of Pakistan (50). Indirect evidence for colonization
of the root interior obtained by isolation procedures was first
supported by immunofluorescence with antibodies which are
Azoarcus genus specific (44). Light and electron microscopic
immunogold studies of better resolution revealed that these
bacteria colonize the aerenchymatic air spaces but also deeply
penetrate roots into stele and stem bases (27, 28). In gnotobiotic culture, strain BH72 was capable of systemic infection of
Kallar grass and rice seedlings, invading the cortex inter- and
intracellularly and even penetrating into xylem vessels. Anticipating that plant polymer-degrading enzymes of bacterial
origin were involved, we wanted to screen members of Azoarcus spp. for occurrence of cellulases. Here, we report cloning
and characterization of two classes of cellulolytic enzymes,
which share an unusual combination of features indicating they
might be involved in the infection process.
7056
V()L. 175, 19)93
CELLULASES OF AZOARCUS SP.
7057
TABLE 1. Bacterial strains and plasmids"
Dcscriptioin
Strain or plasmid
Strains
A. indigenis VB32'
A. comnhunis SWub3'
Azoarcclus sp. strain BH72
Azoarcus sp. strain S5b2
Azoarcus sp. strain 6a3
E. coli DH5st F'
Wild type
Wild type
Wild type
Wild typc
Wild type
F' ecdA1 hsdRJ7 (rK m-iK+) supE44 thi-i recAI gyrA re/lAl 80/a(cZAM15
Sourcc or reference
A(/acZYA-argF)1,,,X,
47
47
47
47
47
23
Plasmids
pUC I9
pUCl8
Ap' ColEl replicon
Apr ColEl replicon
pBGV5.2
Ap'-, 5.3-kb KpnI fragment of pBGV5 cloned in pUC18
pBGVI 1.58
Ap', 1.45-kb SstI-KpsnI fragment of pBGV5 cloned in pUCI8
pBGV8.l
Apr, 4.8-kb SmaI-XbalI fragment of pBGV8 cloned in pUC18
Plasmitds not shown in this talbIc arc described in Fig. 2.
59
59
This study
This study
This study
"
MATERIALS AND METHODS
Bacterial strains and plasmids. The strains and plasmids
used in this study are listed in Table 1. Azoarcius strains
originated from roots of Kallar grass, Leptochloa fusca (L.)
Kunth, grown in Pakistan, and represented five groups of
Azoarclus spp. differing at the species level (47). Physical maps
of plasmids are illustrated in Fig. 2.
Media and growth conditions. Escherichia coli cells were
grown at 37°C in Luria broth (LB) (2) or on LB agar
supplemented with carbenicillin (150 pg/ml) to maintain plasmids and were grown, when indicated, with isopropyl-p3-Dthiogalactopyranoside (IPTG; 24 p.g/ml).
Azoarculs spp. were grown on VM medium (47) with rotary
shaking at 28°C as an inoculum for cellulase tests and rotary
shaking at 37°C for cell mass production. To obtain microaerobic growth, Azoarcus sp. strain BH72 was grown at 33°C in 100
ml of liquid SM medium (46), with fourfold-strength phosphate buffer and 20 mM L-proline added, in a gas-tight sealed
1 liter Erlenmeyer vial with reciprocal shaking (70 strokes per
min). Cultures were harvested when the initial 02 concentration of 1.7% in the headspace had decreased to 0.8%; atmospheric 02 was measured with a GC1 1 gas chromatograph with
a thermal conductivity detector (Delsi Instruments, Suresnes,
France) by using a Molecular Sieve DS column (2 m, 1/8-mm
DS, 80 to 100 mesh; Alltech). To measure the expression of
cellulase by Azoarcus spp., a modification of a medium proposed by Kim and Wimpenny (KW medium [29]) was used,
which contained 0.4% (NH4)2S04, 0.01% NaCl, 0.01%
MgSO4, 0.01% CaCl2, 0.05% yeast extract, 33 mg of Fe(III)EDTA per liter, 0.05 M potassium phosphate buffer (pH 7.0),
and 0.04% medium-viscosity carboxymethylcellulose (CMC;
no. C 4888; Sigma, St. Louis, Mo.). Addition of 0.5% filtersterilized D-glucose or 1.5% agar was optional.
When tested for utilization of carbon sources, Azoarcus cells
were grown on KW plates for 4 days, scraped off, and washed
twice in saline. They were used as an inoculum (at 6 [ig of
protein per ml) for test cultures in liquid KW medium without
yeast extract (25 ml in 125-ml Erlenmeyer flasks) and were
incubated at 37°C with rotary shaking (150 rpm) with three
replicates each. Complex carbohydrates were autoclaved, and
D-glucose and cellobiose were sterilized by filtration and added
at 1% (wt/vol). Nitrogen-fixing growth on the respective carbon sources was tested with nitrogen-free semisolid malate
medium (46), with potassium malate being replaced by the
respective carbohydrate (1% [wt/vol]). Culture tubes were
inoculated with cells washed as described above (at -1 [ig of
protein per culture), and nitrogen fixation was determined by
the acetylene reduction assay (47).
Azoarctus cells for protein extractions were grown in 150 ml
of liquid KW medium supplemented with 3 ml of filtersterilized ethanol per liter in screw-capped 500-ml Erlenmeyer
flasks at 28°C with rotary shaking (60 rpm); after 48 h, ethanol
(3 ml/liter) was again added and the cells were grown for a
further 12 h.
Screening of bacterial colonies for B-glucosidase, cellobiohydrolase, and endoglucanase activity. Azoarcus strains growing exponentially on VM medium were streaked on KW plates
with or without D-glucose, incubatcd for 4 days at 37°C, and
then incubated at 37°C with the appropriate overlay to screen
for cellulase activities. Overlays consisted of 8 ml of 0.05 M
potassium phosphate (pH 7.0)-buffered agarose (0.7%) containing 0.5 mg of either 4-methylumbelliferyl-3-D-glucoside
(MUG; Sigma) or 4-methylumbelliferyl-3-cellobioside (MUC;
Sigma) per ml for 3-glucosidase or cellobiohydrolase detection
or 0.2% CMC for endoglucanase detection. After 4 to 10 h,
plates were exposed to 302 nm of UV light on a transilluminator, and the active colonies were identified by the appearance of blue fluorescence. Endoglucanase activity was detected
after overnight incubation by staining with Congo red (Merck,
Darmstadt, Germany); the appearance of a clear-yellowish
halo on a red background indicated endoglucanase activity
(53). When necessary, plates were incubated for 5 min with 10
ml of 1 M HCI to enhance contrast (49).
To detect 3-glucosidase and cellobiohydrolase activity in E.
coli clones, cells were grown overnight on LB plates with IPTG
and then were incubated with a MUG or MUC overlay for I to
4 h at 37°C. To visualize expression of endoglucanase, LB
plates were overlaid with CMC-agarose supplemented with
carbenicillin and IPTG and clones were inoculated deeply into
the agar. After incubation at 37°C overnight, plates were
stained with Congo red as described above.
General cloning and DNA-analysis techniques. Plasmids
from E. coli were prepared by the alkaline lysis procedure of
Birnboim and Doly (7), as modified by Ausubel and Frederick
(2), for both small-scale and large-scale preparations. Restriction endonuclease digestions, ligations, analysis and recovery
of DNA fragments on agarose gels, and autoradiography were
carried out according to standard protocols (2). DNA fragments were labeled with 32p with a random primed labeling kit
(Boehringer Mannheim). T4 DNA ligase and calf intestine
alkaline phosphatase were purchased from Bethesda Research
Laboratories and Promega, respectively. E. coli competent
7058
REINHOLD-HUREK ET AL.
cells were prepared and transformed according to the method
of Kushner (30).
Construction and screening of a representative genomic
library. High-molecular-weight genomic DNA was prepared
from Azoarcus sp. strain BH72 by the method of Marmur (37),
modified as described previously (45), except that further
purification by CsCl gradient was omitted. DNA (240 pLg) was
partially digested with Sau3AI and was size fractionated with a
sucrose gradient (2). Fragments 5 to 15 kb in size were used for
construction of a library in pUC19 which was digested to
completion with BamHI and dephosphorylated. Vector and
insert DNA in equimolar amounts, as well as vector DNA
alone as a control, were ligated and subsequently used for
transformation of E. coli. A total of 2,880 clones positive in
blue-white selection with 5-bromo-4-chloro-3-indolyl-,3-D-galactopyranoside (20 ,ug/ml of LB agar) were transferred to
microtiter plates and replica plated. Bacterial colonies were
screened for expression of 3-glucosidase, cellobiohydrolase,
and endoglucanase. Positive clones were restreaked and tested
again to confirm activity.
Preparation of whole-cell extracts and cell fractionation.
For enzymatic quantitation, Azoarcus sp. strain BH72 cells
were grown on liquid KW medium supplemented with ethanol
unless stated otherwise. Cells from 15 to 25 ml of culture were
processed at 4°C; they were pelleted, washed once in 10 ml of
10 mM Tris-HCI (pH 8.0), and suspended in I ml of the same
buffer to give an optical density (at 578 nm) of -60. To obtain
total cellular extracts, cells were disrupted by ultrasonication
with a B-10 cell disrupter (Branson Sonic Power Corp.) on ice
at a 40-W output for three cycles of 30 s each. Cell debris was
pelleted by centrifugation at 13,000 x g, and supernatants
were used for enzyme assays. For extraction of proteins with
Triton X-100, washed cells were shaken for 10 min at 28°C and
60 rpm in Tris-HCl buffer containing 0.1% Triton X-100 and
10 mM Na2-EDTA. After centrifugation at 9,000 x g for 10
min, the supernatant was used.
The first step of cell fractionation was performed by centrifugation at 10,000 x g for 10 min. To obtain the extracellular
fraction, the culture supernatant was further centrifuged at
20,000 x g for 25 min and concentrated 30-fold by ultrafiltration through an Amicon YM10 membrane. Pelleted cells were
subjected to fractionation by the osmotic shock method as
described by Cornelis et al. (14), except that the 25% sucrose
solution was made up in 10 mM Tris-HCl (pH 7.0). Residual
cells from which the periplasmic fraction had been prepared
were ultrasonicated as described above to obtain the cytoplasmic fraction.
The total cellular protein of E. coli was prepared by ultrasonication as described above or by the freeze-thaw method.
Cells suspended in Tris-HCI buffer were rapidly frozen and
thawed (dry ice-cooled ethanol, 30°C).
Determination of protein concentrations. The protein content of cell extracts was determined by the Bio-Rad protein
assay (Bio-Rad Laboratories Ltd.). Cellular protein was estimated by the micro-Goa method (4).
IEF, zymograms, and SDS-PAGE. For isoelectric focusing
(IEF) LKB Ampholine PAGplates in the pH range 3.5 to 9.5
(no. 1804-101, LKB) were used with a Multiphor II electrophoresis chamber (LKB). Electrophoresis and Coomassie brilliant blue staining were carried out according to the manufacturer's instructions, except that experiments were run at lower
voltage (400 V) for 2.5 h. For zymograms, we followed the
outlines given by Coughlan (15). Agarose overlays for IEF gels
contained 0.8% agarose in 0.05 M potassium phosphate buffer
(pH 7.0) and the respective substrate at 0.01% (MUG or
MUC) or 0.1% (CMC). After incubation at 37°C, overlays
J. BA(CF[-'RIOL.
proceeded as described above, except that CMC overlays were
washed for 10 min in phosphate buffer prior to Congo red
staining.
Proteins from extracts obtained by ultrasonication were
separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) on 12% polyacrylamide gels as described by Laemmli (31).
Enzyme assays. ,B-Glucosidase and cellobiohydrolase activities were quantified with the chromophoric substrates 2chloro-4-nitrophenyl-r3-D-glucoside or 2-chloro-4-nitrophenyl3-cellobioside (11) at 2 mM in 200 mM potassium phosphate
buffer (pH 6.8) with 20 [tg of cell protein extract added (heat
inactivated at 95°C as a negative control); the increase in A405
was monitored spectrophotometrically upon incubation at
37°C. One unit of specific activity is defined as 1 ,umol of
chromophore released min'-'' mg protein', with the extinction coefficient at 405 nm equalling 9,000 M-' cm '. Qualitative assays with other substrates with the same chromophore
were similarly performed. Enzymic hydrolysis of fluorogenic
substrates was qualitatively monitored by UV illumination.
Endoglucanase activity was assayed with the chromogenic
substrate Remazol brilliant blue R-CMC according to the
method of Cleary (13) with modifications. Twenty micrograms
of cell protein extract was added to 0.2% Remazol brilliant
blue R-CMC dissolved in 50 mM Tris-HCl buffer (pH 6.8), and
samples were incubated at 37°C for 10 h. The specific increase
in A5910 was calculated, taking into account a negative control
(heat-inactivated extract), as AA599 h -' mg of protein .
Qualitative tests for degradation of Ostazin brilliant redhydroxyethylcellulose (Sigma) were carried out as described by
Biely et al. (6) with the buffer described above. Enzymatic
hydrolysis of Remazol brilliant blue R-Avicell (0.2%) was
tested with 160 mM Tris-HCI (pH 6.8) and calculated from the
increase in A5,0 of the supernatant after centrifugation.
Reducing sugars were quantified by a modification of the
Somogyi-Nelson method as described by Marais et al. (36).
Northern analysis. RNA was isolated from bacterial cells
from 30 ml of an exponentially growing culture by the hotphenol method (1). Northern (RNA) blotting and hybridization were carried out according to standard protocols (2).
After separation of nucleic acids on a 0.9% agarose gel, they
were blotted onto a nylon membrane (Hybond N; Amersham),
fixed to the membrane by baking at 80°C, and finally were
hybridized at 65°C.
RESULTS
Occurrence of cellulolytic activities in Azoarcus strains.
Representative members of all five Azoarcus spp. were
screened for the presence of hydrolytic enzymes known to be
involved in cellulose degradation. Screening of colonies was
carried out with fluorophoric substrates to detect exo-1,4-,Bglucanases (,-glucosidase, EC 3.2.1.21; exo-1,4-cellobiohydrolase, EC 3.2.1.91), and a CMC-plate-clearing test was used to
detect endo-1,4-p-D-glucanase (EC 3.2.1.4). All but one strain,
6a3, were positive or weakly positive for f-glucosidase,
whereas cellobiohydrolase activity was detected only in A.
communis SWub3T. Endoglucanase activity was only detectable in strain BH72. Whether D-glucose was added to KW
medium or not, no significant difference was observed. However, when ethanol was added to glucose-free KW agar to
promote colony growth, detection levels were considerably
enhanced. Strong 3-glucosidase activity could be found in
strains BH72 and SWub3T (Fig. IA). Although cellobiohydrolase activity was generally weaker, it could be detected in all
four strains (Fig. IB). In contrast, considerable endoglucanase
VOL. 175, 19933
CELLULASES OF AZOARCUS SP.
7059
FIG. 1. Cellulolytic activities of Azoarcus sp. strains in colony assays. Bacteria werc grown on KW agar plates supplemented with ethnilol (6
ml/liter), overlaid with the respective substrate and viewed under UV illumination (A and B) or evaluated by Congo red staining (C). Overlays
contained fluorophoric substrates to detect P-glucosidase (MUG [A]) or cellobiohydrolase (MUC [B]) activity or CMC to detect endoglucanasc
activity (C) and were incubated for 4 h, 10 h, or overnight, respectively. Numbers on plates: 1, Azowrcus sp. strain BH72; 2, A. indigens VB32"'; 3,
A. communis SWub3?, 4, Azoarcius sp. strain S5b2.
activity was only seen in strain BH72; A. commutnis SWub3T
showed only very weak activity (Fig. IC). Also, with ethanol as
a carbon source, none of these enzymic activities were found in
strain 6a3.
All Azoarculs strains were tested for degradation of cellulose
filter strips (9 by 1 cm; Whatman 40) in culture solutions
consisting of KW medium without CMC, KW medium supplemented with either D-glucose, potassium malate, or ethanol, or
KW medium in the absence of any additional carbon source.
Even within 1 month of incubation at 37°C, no structural
disintegration of the filter strips was visible, indicating that
there was no significant cellulolytic activity on this substrate.
However, cells of strain BH72 readily released reducing
sugars from soluble cellulose CMC. When cells were grown on
liquid KW medium with 0.02% CMC and 6 ml of ethanol per
liter for 3 days, culture supernatants contained 12 ± 4 pg of
reducing sugar per ml more than controls, corresponding to 29
± 10 p.g per mg of protein. Controls consisted of the sum of
the reducing sugar contents of uninoculated medium and
inoculated, CMC-free cultures. All assays were performed in
triplicate.
Utilization of cellulose and other carbohydrates. Utilization
by Azoarculs sp. strain BH72 of the following cellulose types,
derivatives, and hydrolysis products was tested: medium-viscosity CMC, cx-cellulose fiber (C 8002; Sigma), microcrystalline
Avicel (Merck), cellobiose, and D-glucose. Growth was monitored by measurement of optical density (at 578 nm) or, in the
case of insoluble cellulose, by microscopic examination after 2,
7, and 12 days of incubation, and at the end cellular protein
was determined. There was no evidence for growth on any of
the carbohydrates. The protein contents found with D-glucose
(134 ± 6 pg), cellobiose (132 ± 6 kg), and CMC (129 ± 9 pg)
coincided with those found in control cultures (131 ± 8 p.g).
The same compounds also did not support nitrogen-fixing
growth or nitrogen fixation. Within 2 weeks of incubation,
there was no formation of a subsurface pellicle, which is typical
for nitrogen-fixing growth of Azoarclus spp., and no acetylene
reduction could be detected.
Cloning of cellulases and expression in E. coli. E. coli clones
of a total genomic library from Azoarcuts sp. strain BH72
constructed in pUC19 were screened by the overlay technique
for all three cellulase activities found in this strain: P-glucosidase, exo-1,4-cellobiohydrolase, and endo-1,4-3-glucanase. Of
2,880 colonies tested, seven clones showed P-glucosidase activity within 1 to 2 h. The same clones according to their
positions in the microtiter plates also showed a positive, albeit
weaker, reaction for cellobiohydrolase, which was well detectable within 4 h of incubation. As deduced from different
positions in the microtiter plates, five different clones were
positive for endoglucanase activity. To allow an initial grouping
of the clones, PstI restriction patterns of plasmid DNA from
individual clones were compared. Clones pBGV7 and pBGV6,
as well as pBGV9 and pBGV8, had almost identical patterns,
so that only one of each was further analyzed.
P-Glucosidase and endoglucanase clones were subjected to
endonuclease restriction mapping (Fig. 2). Restriction maps of
clones pBGV10, pBGV12, pBGV6, and pBGV8 overlapped in
a central genomic region which was likely to carry the 3-glucosidase (g,Ic) and cellobiohydrolase (cbh) genes (Fig. 2A). To
confirm this assumption and to localize the genes more accurately, fragments were subcloned into pUCI9 in the same
orientation as in clone pBGV6. A 2.6-kb subclone (pBGV6.6)
expressed both activities strongly, whereas smaller subclones
had no or only weak 3-glucosidase activity (Fig. 2A), indicating
that parts of the gene necessary for function might be located
upstream from the HindlIl site. Also the endoglucanase
activity (egi) seemed to be located as one gene in an overlapping region of pBGV2, pBGV3, pBGVl 1, pBGVI, and
pBGV5. Subcloning could localize the activity on a 1.45-kb
SstI-KpnI fragment (Fig. 2B, pBGV11.5). There was no evidence for a second endoglucanase gene located within 5 kb
upstream or 12.5 kb downstream of egl in this genomic region
(pBGV2.1 and pBGV2.2, respectively).
Restriction maps of genomic regions harboring 3-glucosidase or endoglucanase gave no evidence for physical linkage.
Because all clones of one type which we had obtained from
screening of the library were in the same orientation with
respect to the IacZ promoter of pUC 19, we anticipated that
cellulase genes were not under control of their own promoter
when expressed in E. coli. When the orientation of egl-active
fragments (pBGV1 1.5 or pBGV5) was reversed (pBGV1 1.58
or pBGV5.2, respectively), no endoglucanase activity could be
detected. Similarly, reversion of a glc- and cbh-active fragment
(pBGV8 to pBGV8. 1) resulted in loss of 3-glucosidasc or
cellobiohydrolase activity, indicating that the respective Azo-
7060
REINHOLD-HUREK ET AL.
J. BA(TFI-RIOL..
½½~~~~~~~~~--
1
½
2
3
2
3
2
3
4,
Phenotype
glc cbh
glc/cbh
7.0 -
pBGV10
pBGV12
6.5 -
pBGV6
pBGV8
pBGV6.6
6.0 -
pBGV6.6 K
pBGV6.6H
1kb
II
egl
pBGV2
pBGV3
pBGVll
pBGV1
pBGVS
pBGV21
pBGV2.2
pBGV5.1
pBGV11.5
lkb
FIG. 2. Physical map of genomic regions carrying putative ,-glucosidase and cellobiohydrolase (glc/cbh) genes corresponding to an
exoglycanase (exg) and the endoglucanase gene (egi) of Azoarcus sp.
strain BH72 and plasmid inserts in pUCI9 transcribed from the 1acZ
promoter (right side of figure). For clones pBGV2 and pBGV3, the
position of the second Sall site from the right is one of two possible
locations, the second being 0.7 kb to the left. The enzymatic activities
of P-glucosidase (glc), cellobiohydrolase (cbh), and endoglucanase
(egl) determined by colony assays are given to the right: +, positive; w,
weak activity; -, no activity detectable.
arcus promoters either were not located on the fragments or
were not active in E. coli.
In order to compare cloned and wild-type cellulases, enzyme
activities were located on IEF gels by zymograms. Sonicated
cell extracts of Azoarcus sp. strain BH72 were not separated
into distinct patterns in IEF-PAG plates but gave smears,
probably due to the presence of lipids. We overcame this
problem by extracting samples once with chloroform, a treat-
ment compatible with endoglucanase but detrimental to ,B-glucosidase activity. The latter was resistant to Triton X-100 and
EDTA, which we used to release cell surface proteins, including ,-glucosidase (see below). ,B-Glucosidase and cellobiohydrolase activities of the E. coli clone were focused at the same
pH (pl 5.4); wild-type enzymes behaved the same way (Fig. 3).
In contrast, endoglucanase activity obtained from a clone
carrying a small 1.45-kb insert exhibited two distinct bands with
pIs of 6.4 and 6.0, respectively (Fig. 4). These two isoenzymes
also appeared in Azoarcus spp., although with different intensity ratios.
The apparent molecular masses of the cellulases expressed
by E. coli were estimated by SDS-PAGE (Fig. 5). A new,
41-kDa band appeared in E. coli cells expressing 3-glucosidase
and cellobiohydrolase from a 2.7-kb fragment when induced
with IPTG, which was not seen in uninduced cells or induced
cells carrying insert-free plasmid (Fig. SA). When endoglucanase was induced from a 1.45-kb fragment, newly formed
bands were not clearly visible (not shown). However, with a
different method of protein extraction (SDS-soluble proteins
[45]) a new, 36-kDa band was clearly visible (Fig. SB).
Enzymatic properties of the cloned cellulases. Substrate
5.1 -
C
A
B
FIG. 3. IEF and zymograms of cloned (lanes 2) and wild-type
(lanes 3) exoglucanase. Protein extracts were obtained by the freezethaw method from E. coli cells carrying plasmid pUC19 as a negative
control (lane 1) or pBGV6.6 (lanes 2) and from cell surface proteins
prepared by Triton X- 100 washes from Azoarcus sp. strain BH72 (lanes
3). Gels were stained with Coomassie brilliant blue (A) and for
13-glucosidase (B) or cellobiohydrolase (C) activity with overlays
incubated for 15 or 45 min, respectively. Twenty and 8 ,utg of protein
(lanes Al and A2 and lane A3, respectively) were loaded for Coomassie staining; 10 and 12 ,ug of protein (lanes B2 and C2 and lanes B3
and C3, respectively) were loaded for zymograms. Numbers to the left
correspond to pls of protein standards. Enzymatically activc bands at
pl 5.1 are found in both cloned (lanes 2) and wild-type (lanes 3)
preparations.
utilization was determined for Azoarcus sp. strain BH72 cells
and for individual cloned cellulases expressed in E. coli and
tested with cell extracts (Table 2). E. coli cells carrying
insert-free pUC19 were negative for all substrates tested (not
shown). The substrate specificity of cloned ,B-glucosidase and
cellobiohydrolase (exoglucanase) was clearly different from
that of endoglucanase. The activity of the cellobiohydrolase
and ,B-glucosidase is incompatible with substituted, soluble
cellulose derivatives, but a variety of oligomeric sugars are
substrates, indicating that the presumed glc/cbh gene codes for
an exoglucanase (exg). The activity of the endocellulase is
clearly in contrast with this behavior. Microcrystalline cellulose
(Avicel) was not attacked by any of the enzymes or the
wild-type Azoarcus strain. This clearcut difference in substrate
specificity allowed us to differentiate activities in cell extracts
containing both enzymes.
To elucidate the mode of action and substrate specificity of
cellulases, cleavage sites of 3-glucosidase and cellobiohydrolase and endoglucanase were determined with cell extracts
from E. coli clones. 1,4-,-Cellooligosaccharides carrying a
fluorophoric group (MUGj) were tested as substrates for the
respective enzymes (4 h of incubation), and the products were
separated by thin-layer chromatography. Inspection of the
products obtained (Fig. 6) indicated that endoglucanase preferably attacked oligosaccharides larger than cellobiose and
released large oligomers when sufficiently long substrates were
used. 3-Glucosidase attacked dimers as expected (Table 2), but
it also attacked larger substrates, preferably cleaving the bond
at the fluorophoric group (nonreducing end). Apparently,
more-highly oligomeric substrates were also cleaved at internal
bonds.
Temperature and pH profiles of exoglucanase and endoglucanase activities were obtained from cell extracts of Azoarcus
sp. strain BH72 (not shown). Profiles for exoglucanase measured either as 3-glucosidase or as cellobiohydrolase activity
correlated satisfactorily. The specific activity was 3- to 10-fold
VOL. 1 75, 19'93
CELLULASES OF AZOARCUS SP.
1
2
2
3
TABLE 2. Substrate specificity of cloned cellulases and wholc cells
of Azoarcus sp. strain BH72
3
8.1 -
Substratc specificity of:
Exoglu- Endoglu- Whole
canaIsc" canase" cells"
Substrate
7.06.5-
MUG
MUC
2-Chloro-4-nitrophenyl-f3-D-glucoside
+"
-
+
+
-
+
-
+
+
2-Chloro-4-nitrophenyl-3-cellobioside
+
+
+
+
-
-
+
2-Chloro-4-nitrophenyl-l3-lactoside
p-Nitrophenyl-P-D-xyloside
4-Methylumbelliferyl-13-D-xyloside
4-Methylumbelliferyl-l3-xylobioside
CMC'
Remazol brilliant blue R-CMC
Remazol brilliant blue-Avicel
Ostazin brilliant red-hydroxyethylcellulose
6.05.1-
A
b
FIG. 4. IEF and zymograms of cloned (lanes 2) and wild-type (lanes
3) endoglucanase. Protein extracts were obtained by the freeze-thaw
method from E. coli cells carrying plasmid pUC19 as a negative control
(lane 1) or pBGV1 1.5 (lanes 2) and from whole-cell preparations from
Azoarculs sp. strain BH72 extracted with chloroform (lanes 3). Gels were
stained with Coomassie brilliant blue (A) or for endoglucanase activity
(B) with overlays incubated for 3.5 h. Fifteen, 7, and 12 xg of protein
were loaded on lanes Al, A2, and A3, respectively, for Coomassie
staining; 15 and 24 ,ug of protein were loaded on lanes B2 and B3,
respectively, for zymograms. Numbers to the left correspond to pIs of
protein standards. Two isoenzymes with pIs of 6.4 and 6.0, respectively,
appear from cloned (lanes 2) and wild-type (lanes 3) preparations.
higher for 3-glucosidase than for cellobiohydrolase. The pH
optima were 6.4 and 6.0 for exoglucanase and endoglucanase,
respectively. The temperature optima coincided at 46°C.
Cellular location of cellulases in Azoarcus sp. Determination
A
1
2 3 4
---
B
97.4
66.2
42.7
31.2
21.5
5
97.4
-
66.2
a
42.7
_
6
7061
7
-
+
+
+
ND"
ND
ND
+
+
+
+
-
-
" Cell extracts of E. coli cells carrying plasmid pBGV6.6.
"'Cell extracts of E. coli cells carrying plasmid pBGVI 1.5.
'Azoarcus sp. strain BH72 cells grown on KW medium, with complex
substrates added during growth.
d Cleavage was estimated by appearance of the released chromophore, seen as
fluorescence or color.
' ND, not determined.
/'Cleavage was estimated by appearance of reducing sugars.
of specific activities after cell fractionation revealed that both
cellulases were mainly surface associated (Table 3). Exoglucanase quantified as ,B-glucosidase activity was not detectable
in the culture supernatant, and only low specific activities were
found in the cytoplasmic fraction. To enhance limits of detection, proteins in the culture supernatant were concentrated by
ultrafiltration. Even after 30-fold concentration, no 3-glucosidase activity was detected. To evaluate whether ultrafiltration
might reduce endo- and exoglucanase activities, we concentrated an aliquot of the cell surface fraction; specific activities
were not affected (data not shown). For endoglucanase activity, we obtained a similar pattern of distribution. Endoglucanase activity could be detected neither in the culture supernatant nor in the cytoplasmic fraction but was detected only in
the cell surface and periplasmic fractions.
Regulation of cellulases and transcription of endoglucanase. To test the effect of carbon sources on the expression
"t"
_
"f
.
MUG, + endoglucanase
MUG, + exoglucanase
El-31.2
21.5
_
_
El-Ei---
4
El-El-EI-.
-
FIG. 5. Apparent molecular weights of exoglucanase (A) and endoglucanase (B) expressed in E. coli. Whole-cell extracts prepared by
ultrasonication (A) or SDS extraction (B) from E. coli cells carrying
plasmid pUC19 (lanes 2 and 6), pBGV6.6 (lanes 3 and 4), or
pBGVI 1.5 (lane 7) were separated by SDS-PAGE and stained with
Coomassie brilliant blue. Approximately 40 p.g of protein extracts per
lane was loaded. For lane 3, cellulases were repressed by addition of
1% glucose and omission of IPTG. For lanes 4 and 7, cellulases were
induced by omission of glucose and addition of IPTG. Lanes 1 and 5
show molecular mass markers (sizes are given in kilodaltons). Additional protein bands appearing in cells expressing cellulases are
marked with arrowheads and have apparent molecular masses of 41
kDa (A) and 36 kDa (B).
4
El4El4
ElE
lEl-
lE4lE4
FIG. 6. Specificity mapping of cloned endoglucanase and exoglucanase with MUGn derived from the cellooligosaccharides (n = 1 to 6).
Protein extracts from E. coli cells carrying plasmid pBGBI 1.5 or
pBGV6.6 were incubated with MUG,,, and products were separated by
thin-layer chromatography (12). Solid circles indicate the positions of
the fluorophoric group. Arrows mark cleavage sites estimated from the
mobility of fluorescing bands.
70)62
REINHOLD-HUREK ET AL.
J. BACT11RIOL.
TABLE 3. Cellular location of cellulase activity in Azoarcus sp.
strain BH72
kb
Endoglucanase
of
-tity(Exoglucmngasc
acprteprotcin
ivit (U/mg
± SD)
Fraction
Friiction
Not detectable
Not detectable
Culture supernatant
Concentrated culture
supernatant"
Cell surface fraction
Periplasmic fraction
Cytoplasmic fraction
Total cells
107
92
5
30
activity
(AA590
h mg
of
protein
--
SD)"
Not detectable
Not detectable
2.15 ± 0.30
2.50 ± 0.04
Not detcctable
0.71 ± 0.05
± 7
± 1
± (0
± 0
14.5 -*
9.5 7.54.4 -
2.4 -
1.4 -
SpCcific activity of -glUcosidase was determined with 2-chloro-4-nitrophenyl
as substrate.
P-i)-glucoside
/
Specific increase in absorption was determined with Remazol brilliant blue
R-CMC as a substrate.
High-molccukar-wcight fraction was concentrated 30-fold by ultratfiltration.
"
0.24 of cellulases by Azoarcuis sp. strain BH72, cells were grown for
48 h on the respective carbon sources, and specific enzyme
activities in cell extracts were determined. Both exoglucanase
and endoglucanase showed similar patterns of activity (Table
4) and were expressed under all conditions tested. A slight
induction by the substrate CMC was seen only for exoglucanase and not for endoglucanase. There was no strong
repression by the end product D-glucose. Other carbon sources
which well support growth of Azoarcus cells acted differently:
DL-malate hardly affected or even reduced exo- and endoglucanase activity, whereas ethanol had the strongest inducing
effect, increasing both activities 4- and 1.6-fold, respectively.
Oxygen also had a pronounced effect on expression of
endoglucanase. In Northern analysis with the 1.45-kb SstI-Pstl
fragment as a probe, we found a large, -15-kb transcript on
which egl was located (Fig. 7). It was only detected in cells
grown microaerobically on N2 and was not detected when cells
were grown aerobically on combined nitrogen (Fig. 7). As a
control for the absence of genomic contamination in the RNA
preparation of Fig. 7A, we visualized nucleic acids by ethidium
bromide staining after gel separation (not shown). Additionally, we hybridized Northern blots from the same preparation
with genomic probes for other Azoarcus genes, namely nifH
and nifDK; hybridizing bands were obtained only at lower
molecular sizes (-6.2 kb), indicating therc was no genomic
contamination at 15 kb (not shown).
TABLE 4. Effect of carbon sources on cellulase activities of
Azoarcus sp. strain BH72"
Cairbon source added'
Yeast
Yeast
Yeast
Yeast
Yeast
extract
extract
extract
extract
extract
+
+
+
+
CMC
CMC + glucose
CMC + malate
CMC + ethanol
P-Glucosidase
Endoglucanasc
activity (AA590
protein +
SD)
protcin ±
SD)
8.56 ± 0.01
10.67 ± 0.t)8
9.55 ± (1.11
12.22 ± 0.88
41.11 ± 1.0()
0.030 ± 0.001
0.028 ± 0.001
0.026 ± 0.002
0.019 ± 0.0(1
0.044 ± 0.000
activity
(U/mg ot
h
mg of
" Specific aictivitics of cnzymes in cell extracts were determined after ultriasonication of total cells.
' Carbon sources in mineral salts of KW medium were as follows: yeast extract,
11.5 g/liter; CMC, 11.4 g/liter; 1)-glucose, 1) g/liter; il.-malic acid, 2.5 g/liter
(neutralized with KOII); ethalnol, 4 ml/litcr.
++
02 +
FIG. 7. Northern blot analysis of endoglucanase expression. RNA
was isolated from Azoar-cus sp. strain BH72 grown on malate microaerobically on N, (A) or aerobically on combined nitrogen (B).
Twelve micrograms of RNA in each lane was separated by agarose gel
electrophoresis, blotted onto nylon membrane, and hybridized to a
329P-labeled endoglucanase probe consisting of the 1.45-kb insert of
pBGVI 1.5. The arrow marks labeled transcript. Symbols: + +, high 02
concentration (aerobic); +, low 0O concentration (microaerobic).
DISCUSSION
Although cnzymes known to be involved in cellulase breakdown were present in four of five Azoarcus spp., we focused on
Azoarcus sp. strain BH72 for further studies because it had the
widest spectrum of cellulolytic enzymes. From this root-invading diazotroph, we detected and cloned two types, an exoglucanase and an endoglucanase.
Although strain BH72 showed 3-glucosidase as well as
cellobiohydrolase activities, we obtained evidence that both
are conferred by only one enzyme, as follows. (i) All clones
from a representative genomic library expressing one activity
also express the other; however, with 2.7 kb, the smallest fully
positive subclone was still large enough to harbor two open
reading frames. (ii) Wild-type and cloned enzymes of both
activities have the same pl (5.4). (iii) Activity profiles of pH
and temperature dependence coincide. (iv) From the smallest
fully active subclone, only one major protein is expressed (41
kDa). (v) The cloned enzyme not only cleaves cellobiose or
releases it but it also hydrolyzes larger oligomeric sugars,
releasing, e.g., cellotriose.
Furthermore, since aryl-3-D-xylosides are cleaved, the enzymc can best be characterized as an unspecific exoglycanase.
According to its degradation pattern of chromophoric cellooligosaccharides (Fig. 6), the enzyme can tentatively be classified as a family F cellulase (12, 24). To this transition group of
cellulases and xylanases belong an exoglucanase (Cex) from
Cellulomonas fimi (19) and a xylanase (XynZ) from Clostridium thermocellum (20).
Although evidence for two endoglucanase isoenzymes was
obtained from zymograms, both forms are probably derived
from one gene product. The 36-kDa protein would require an
open reading frame -1 kb in size, which would not allow
expression of a second open reading frame of this size from the
1.45-kb fragment if they were not overlapping. Two bands in
Voi.- 1 75? 1 993
IEF presumably result from proteolytic cleavage. Cleavage
sites appear to be identical in Azoarcuts sp. strain BH72 and E.
coli, because the cloned and wild-type enzymes have the same
pIs. The appearance of two active bands 55 and 53 kDa in size
was probably due to proteolytic cleavage when cloned C.
thermocellum celB was expressed (3). Also, cellulases of C. fimni
(CenB [41]) and Clostridium cell/ulolytictum (EGCCA [17])
were reported to be subjected to proteolytic attack. As in the
case of Pseudomonias solanacearuim (25, 26), the second product might also represent a mature form after removal of a
signal sequence. Specificity mapping (Fig. 6) allows a preliminary classification of the endocellulase in family D (12),
although sequence analysis should be decisive in this matter.
It is remarkable that Azoarcus sp. strain BH72, although
possessing a cellulolytic system, cannot grow on cellulose or on
its breakdown products. In addition to filter paper, CMC,
cellobiose, and D-glucose, for which results are presented here,
Azoarcus spp. grew on none of 49 carbohydrates which we
tested for taxonomic purposes (47). There are cellulase-positive eubacteria which are regarded as not truly cellulolytic
because they show no or weak growth on celluloses; in Erwinia
chrysanit/lemiii (10) or P. solanaceanlini (48), cellulases might
assist in releasing nutrients from the host plant; also a cellulase
was detected in Franckia sp., which is the symbiont of many
actinorhizal plants, but growth on filter paper was not obvious,
although growth on sugars occurs (49). Bacteroides ruminicola
does not use celluloses for growth, but breakdown products
such as cellodextrines and cellobiose are used (39). In contrast,
extreme nutritional constraints which do not allow consumption of any of the breakdown products as in Azoarcuts spp. are
unique among cellulose-degrading bacteria.
This raises the intriguing question of which physiological or
ecological function such an enzyme system might have for
these bacteria. Bearing in mind that they are capable of
infecting grass roots where they can penetrate into the stele
and that they were even found in the stem base of Kallar grass
grown in situ (27, 28), one might anticipate that cellulases are
involved in the infection process, allowing Azoarcus cells to
reach a certain microhabitat. Also, an unusual combination of
some further features of this cellulolytic system point in this
direction.
In some gram-positive cellulolytic bacteria, cellulases are
surface associated. In C. thermocellum, cellulases are organized in a complex, the cellulosome, which is cell surfaceassociated at some stages of the process of cellulose degradation (32). Also, in Ruminococcus albus, most of the cellulase
activity was found to be cell bound (58). In other gram-positive
bacteria, cellulases are present in the culture supernatant, e.g.,
in C. fimi (33), Streptomyces lividans (54), and Franckia sp. (49);
many gram-negative bacteria also release cellulases, e.g., Fibrobacter sluccinogenes (40) and Prevotella rluminicola (39). In
our context, gram-negative phytopathogens are of special
interest; P. solanacearlum (26) and E. chrysanthemi (10) excrete
endoglucanases into the culture supernatant. In contrast, exoand endoglucanase of the gram-negative Azoarcuts spp. are cell
surface associated and might therefore mediate a more localized digestion of plant polymers in comparison with phytopathogens, causing less damage to the host.
Most bacterial cellulases seem to be regulated by inductionrepression mechanisms depending on the carbon source. In
accordance with such observations, it has been suggested that,
e.g., several cel genes of C. therrnocell/tm are regulated by a
mechanism analogous to catabolite repression (42). However,
in some rumen bacteria, such as F. succinogenes (40) and P.
ruiminicola (39), there are more constitutively expressed cellulases which are not repressed by D-glucose or cellobiose. Both
CELLULASES OF AZOARCUS SP.
7063
cellulases of Azoarcus sp. strain BH72 also do not appear to be
subject to strong induction by CMC or repression by D-glucose.
Although genes are not physically linked, they seem to be
under coordinate regulation. Taking into account the fact that
this strain is restricted in use of carbohydrates for growth, end
product inhibition or catabolite repression would be of no
obvious physiological advantage. In contrast to Franckia sp.,
carbon sources supporting growth well (49) do not necessarily
enhance cellulase activity, as exemplified by malic acid. The
only substrate we found to significantly enhance activity is
ethanol, a substance which might be of importance in the
Azoarcuts habitat. When aeration of roots is reduced by flooding, ethanol concentrations in roots cain rise immediately (16),
a situation which occurs in rice and Kallar grass culture; in root
tips, ethanol can always be detected, even under atmospheric
oxygen pressure (5). Induction of endoglucanase by ethanol
and microaerobic conditions in strain BH72 might thus support
penetration of roots under these conditions.
An unusual feature of the Azoarcus endoglucanase is its
location on a large transcript -15 kb in size. Evidence for a
large transcript also comes from the lack of transcription of
endoglucanase from its own promoter in E. coli. In contrast,
most bacterial cellulases studied so far are transcribed monocistronically. Also, E. chrysanthemi (21) and P. solanacearlum
(48) endoglucanases appear to be transcribed from their own
promoters. To our knowledge, the only reported evidence for
polycistronic transcription comes from primer extension experiments with P. ruminicola endoglucanase (56). So far, we have
no data concerning genes being cotranscribed with eg/.
In several bacteria, multiple cellulases are present, such as in
C. fimni, C thermocellum, R. albus (e.g., see reference 18), and
E. chrysanthemi (10). InAzoarcius sp. strain BH72, no evidence
was obtained for such a multiplicity either from the results of
library screening or from low-stringency hybridization with the
homologous egl probe (data not shown). Therefore, mutational
analysis might give clear answers about the putative role of this
enzyme in the infection process. Biochemical evidence for an
unusual type of regulation of cellulases will be further investigated genetically by transcriptional fusions with reporter
genes.
ACKNOWLEDGMENTS
B.R.-H. and T.H. were supported by postdoctoral training grants
awarded by the EC.
We thank Marcelle Holsters for hosting us in her laboratory and Jan
Desomer for valuable suggestions. We are grateful to Karel Spruyt,
Vera Vermaercke, and Claudia Nickel-Reuter (Marburg) for help with
the artwork.
REFERENCES
1. Aiba, et al. 1981. Evidence for two functional gal/ promoters in
intact Eschericlhia coli cells. J. Biol. Chem. 256:11905-1 19 t1).
2. Ausubel, I., and M. Frederick. 1987. Current protocols in molecular biology. John Wiley & Sons, New York.
3. Beguin, P., P. Cornet, and J. Millet. 1983. Identification of the
endoglucanase encoded by the celB gene of Clostuidiiui,i tlilermoce/llum. Biochimie 65:495-5(t(.
4. Bergersen, F. J. 198t). Measuremenlt of nitrogen fixation by direct
means, p. 65-110. In F. J. Bergersen (ed.), Methods for evaluating
biological nitrogen fixation. John Wiley and Sons Ltd., Chichester,
United Kingdom.
5. Betz, A. 1957. Zur Atmung wachsender Wurzelspitzen. Planta
50:122-143.
6. Biely, P., D. Mislovicova, and R. Toman. 1985. Soluble chromogenic substrates for the assay of endo-1,4-3-xylanases and endo1,4-p-glucanases. Anal. Biochem. 144:142-146.
7. Birnboim, H. C., and J. Doly. 1979. A rapid alkaline extraction
7064
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
J. BAC1 ECRIOL.
REINHOLD-HUREK ET AL.
procedure for screening recombinant plasmid DNA. Nucleic
Acids Res. 7:1513-1523.
Callaham, D., and J. Torrey. 1981. The structural basis for
infection of root hairs of Trifolium repens by Rhizobium. Can. J.
Microbiol. 59:1647-1664.
Chalifour, F., and A. Benhamou. 1989. Indirect evidence for
cellulase production by Rhizobium in pea root nodules during
bacteroid differentiation: cytochemical aspects of cellulose breakdown in rhizobial droplets. Can. J. Microbiol. 35:821-829.
Chippaux, M. 1988. Genetics of cellulases in Erwinia chrysanthemi,
p. 219-234. In J.-P. Aubert, P. Beguin, and J. Millet (ed.),
Biochemistry and genetics of cellulose degradation. Academic
Press, London.
Claeyssens, M. 1988. The use of chromophoric substrates and
specific assays in the study of structure-activity relationships of
cellulolytic enzymes, p. 393-397. In J. P. Aubert, P. Beguin, and J.
Millet (ed.), Biochemistry and genetics of cellulose degradation.
Academic Press, London.
Claeyssens, M., and B. Henrissat. 1992. Specificity mapping of
cellulolytic enzymes: classification into families of structurally
related proteins confirmed by biochemical analysis. Protein Sci.
1:1293-1297.
Cleary, B. V. 1988. Soluble, dye-labeled polysaccharides for the
assay of endohydrolases. Methods Enzymol. 160:74-86.
Cornelis, P., C. Digneffe, and K. Willemont. 1982. Cloning and
expression of a Bacillus coagulans amylase gene in Escherichia coli.
Mol. Gen. Genet. 186:507-511.
Coughlan, M. P. 1988. Staining techniques for the detection of the
individual components of cellulolytic enzyme systems. Methods
Enzymol. 160:135-144.
Crawford, R. M. M., and M. Baines. 1977. Tolerance of anoxia and
ethanol metabolism in tree roots. New Phytol. 79:519-526.
Fierobe, H.-P., C. Gaudin, A. Belaich, M. Loutfi, E. Faure, C.
Bagnara, D. Baty, and J.-P. Belaich. 1991. Characterization of
endoglucanase A from Clostridium cellulolyticum. J. Bacteriol.
173:7956-7962.
Gilkes, N. R., B. Henrissat, D. G. Kilburn, R. C. Miller, Jr., and
R. A. J. Warren. 1991. Domains in microbial P3-1,4-glycanases:
sequence conservation, function, and enzyme families. Microbiol.
Rev. 55:303-315.
Gilkes, N. R., M. L. Langsford, D. G. Kilburn, R. C. Miller, Jr.,
and R. A. Warren. 1984. Mode of action and substrate specificities
of cellulases from cloned bacterial genes. J. Biol. Chem. 259:
10455-10459.
Grepinet, O., M.-C. Chebrou, and P. Beguin. 1988. Nucleotide
sequence and deletion analysis of the xylanase gene (xynZ) of
Clostridium thermocellum. J. Bacteriol. 170:4582-4588.
Guiseppi, A., B. Cami, J.-L. Aymeric, G. Ball, and N. Creuzet.
1988. Homology between endoglucanase Z of Erwinia chrysanthemi and endoglucanase of Bacillus subtilis and alkalopholic
Bacillus. Mol. Microbiol. 2:159-164.
Halsall, D. M., and A. H. Gibson. 1986. Comparison of two
Cellulomonas strains and their interaction with Azospirillum
brasilense in degradation of wheat straw and associated nitrogen
fixation. Appl. Environ. Microbiol. 51:855-861.
Hanahan, D. 1983. Studies on transformation of Escherichia coli
with plasmids. J. Mol. Biol. 166:557-580.
Henrissat, B., M. Claeyssens, P. Tomme, L. Lemesle, and J.-P.
Mornon. 1989. Cellulase families revealed by hydrophobic cluster
analysis. Gene 81:83-95.
Huang, J., and M. A. Schell. 1992. Role of the two-component
leader sequence and mature amino acid sequences in extracellular
export of endoglucanase EGL from Pseudomonas solanacearum. J.
Bacteriol. 174:1314-1323.
Huang, J., M. Sukordhaman, and M. A. Schell. 1989. Excretion of
the egl gene product of Pseudomonas solanacearum. J. Bacteriol.
171:3767-3774.
Hurek, T., B. Reinhold-Hurek, M. van Montagu, and E. Kellenberger. Submitted for publication.
Hurek, T., B. Reinhold-Hurek, M. van Montagu, and E. Kellenberger. 1991. Infection of intact roots of Kallar grass and rice
seedlings by "Azoarcus." Dev. Plant Soil Sci. 48:235-242.
Kim, B. H., and J. W. T. Wimpenny. 1981. Growth and cellulolytic
activity of Cellulomonas flavigena. Can. J. Microbiol. 27:12601266.
30. Kushner, S. R. 1978. An improved method for transformation of
Escherichia coli with ColEI derived plasmids, p. 17-23. In H. W.
Boyer and S. Nicosia (ed.), Genetic engineering. Elsevier/North
Holland, Amsterdam.
31. Laemmli, U. K. 1970. Cleavage of structural proteins during
assembly of the head of bacteriophage T4. Nature (London)
227:680-685.
32. Lamed, R., and E. A. Bayer. 1988. The cellulosome concept:
exocellular/extracellular enzyme reactor centers for efficient binding and cellulolysis, p. 1t)1-1 16. In J. P. Aubert, P. Beguin, and J.
Millet (ed.), Biochemistry and genetics of cellulose degradation.
Academic Press, London.
33. Langsford, M. L., N. R. Gilkes, N. R. Warkarchuk, W. W. Kilburn,
D. G. Miller, Jr., and R. A. J. Warren. 1984. The cellulase system
of Cellulomonas fimi. J. Gen. Microbiol. 130:1367-1376.
34. Leschine, S. B., K. Holwell, and E. Canale-Parola. 1988. Nitrogen
fixation by anaerobic cellulolytic bacteria. Science 242:1157-1159.
35. Levanony, H., Y. Bashan, B. Romano, and E. Klein. 1989. Ultrastructural localization and identification of Azospirillum brasilense
Cd on and within wheat root by immuno-gold labeling. Plant Soil
117:207-218.
36. Marais, J. P., J. L. De Wit, and G. V. Quicke. 1966. A critical
examination of the Nelson-Somogyi method for the determination
of reducing sugars. Anal. Biochem. 15:373-381.
37. Marmur, J. 1961. A procedure for the isolation of deoxyribonucleic acid from microorganisms. J. Mol. Biol. 3:208-218.
38. Mateos, P. F., J. I. Jiminez-Zurdo, J. Chen, A. S. Squartini, S. K.
Haack, E. Martinez-Molina, D. H. Hubbell, and F. B. Dazzo. 1992.
Cell-associated pectinolytic and cellulolytic enzymes in Rhizobium
leguminosarum biovar Trifolii. Appl. Environ. Microbiol. 58:18161822.
39. Matsushita, O., J. B. Russell, and D. A. Wilson. 1990. Cloning and
sequencing of Bacteroides ruminicola B,4 endoglucanase gene. J.
Bacteriol. 172:3620-3630.
40. McGavin, M., J. Lam, and C. W. Forsberg. 1990. Regulation and
distribution of Fibrobacter succinogenes subsp. succinogenes S85
endoglucanases. AppI. Environ. Microbiol. 56:1235-1244.
41. Meinke, A., N. R. Gilkes, D. G. Kilburn, R. C. Miller, Jr., and
R. A. J. Warren. 1991. Multiple domains in endoglucanase B
(CenB) from Cellulomonas fimi: functions and relatedness to
domains in other polypeptides. J. Bacteriol. 173:7126-7135.
42. Mishra, S., P. Beguin, and J.-P. Aubert. 1991. Transcription of
Clostridium thermocellum endoglucanase genes celF and celD. J.
Bacteriol. 173:80-85.
43. Morales, V., E. Martinez-Molina, and D. Hubbell. 1984. Cellulase
production by Rhizobium. Plant Soil 80:407-415.
44. Reinhold, B., T. Hurek, and I. Fendrik. 1987. Cross-reaction of
predominant nitrogen-fixing bacteria with enveloped, round bodies in the root interior of Kallar grass. Appl. Environ. Microbiol.
53:889-891.
45. Reinhold, B., T. Hurek, I. Fendrik, B. Pot, M. Gillis, K. Kersters,
S. Thielemans, and J. De Ley. 1987. Azospirillum halopraeferens sp.
nov., a nitrogen-fixing organism associated with the roots of Kallar
grass (Leptochloa fusca (L.) Kunth). Int. J. Syst. Bacteriol. 37:4351.
46. 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.
52:520-526.
47. Reinhold-Hurek, B., T. Hurek, M. Gillis, B. Hoste, M. Vancanneyt,
K. Kersters, and J. De Ley. 1993. Azoarcus gen. nov., nitrogenfixing Proteobacteria associated with roots of Kallar grass (Leptochloa fusca (L.) Kunth), and description of two species, Azoarcus
indigens sp. nov. and Azoarcus communis sp. nov. Int. J. Syst.
Bacteriol. 43:574-584.
48. Roberts, D. P., T. P. Denny, and M. A. Schell. 1988. Cloning of the
egl gene of Pseudomonas solanacearum and analysis of its role in
phytopathogenicity. J. Bacteriol. 170:1445-145 1.
49. Safo-Sampah, S., and J. G. Torrey. 1988. Polysaccharide-hydrolyzing enzymes of Franckia (Actinomycetales). Plant Soil 112:8997.
VOL. 175, 1993
50. Sandhu, G. R., and K. A. Malik. 1975. Plant succession-a key to
the utilization of saline soils. Nucleus 12:35-38.
51. Schank, S. C., R. L. Smith, G. C. Weiser, D. A. Zuberer, J. H.
Bouton, K. H. Quesenberry, M. E. Tyler, J. R. Milam, and R. C.
Littell. 1979. Fluorescent antibody technique to identify Azospirillim brasilense associated with roots of grasses. Soil Biol. Biochem.
11:287-295.
52. Sharrock, K. R. 1988. Cellulase assay methods: a review. J.
Biochem. Biophys. Methods 17:81-106.
53. Teather, R. M., and P. J. Wood. 1982. Use of Congo redpolysaccharide interactions in enumeration and characterization
of cellulolytic bacteria from the bovine rumen. Appl. Environ.
Microbiol. 43:777-780.
54. Theberge, M., P. Lacaze, F. Shareck, R. Morosoli, and D. Kluepfel.
1992. Purification and characterization of an endoglucanase from
Streptomyces lividans 66 and DNA sequence of the gene. Appl.
Environ. Microbiol. 58:815-820).
CELLULASES OF AZOARCUS SP.
7065
55. Tien, T. M., H. G. Diem, M. H. Gaskins, and D. H. Hubbell. 1981.
Polygalacturonic acid transeliminase production by Azospirilllin
species. Can. J. Microbiol. 27:426-431.
56. Vercoe, P. E., and K. Gregg. 1992. DNA sequence and transcription of an endoglucanase gene from Prevotella (Bacwroides)
ruminicola AR20. Mol. Gen. Genet. 233:284-292.
57. Waterbury, J. B., C. B. Calloway, and R. D. Turner. 1983. A
cellulolytic nitrogen-fixing bacterium cultured from the gland of
deshayes in shipworms (Bivalvia: Teredinidae). Science 221:14011403.
58. Wood, T. M., C. A. Wilson, and C. S. Stewart. 1982. Preparation of
the cellulase from the cellulolytic anaerobic bacterium Ruininococcus albus and its release from the bacterial cell wall. Biochem.
J. 205:129-137.
59. Yanisch-Perron, C., J. Vieira, and J. Messing. 1985. Improved
M13 phage cloning vectors and host strains: nucleotide sequences
of the M13mpl8 and pUC19 vectors. Gene 33:10)3-119.