- Email: [email protected]
Structure of the glycocalyx polysaccharide of Pseudomonas fragi ATCC 4973
Carbohydrate Research, 216 (1991) 495.-504 Elsevier Science Publishers B.V., Amsterdam
495
Structure of the glycocalyx polysaccharide fragi ATCC 4973” Lesley A. S. Parolis, Haralambos School of Pharmaceutical
of Pseudomonas
Parolis+,
Sciences, Rhodes University, Grahamstown 6140 (South Ajkica)
Guy G. S. Dutton, Chemistry Department,
University of British Columbia, Vancouver V6T I Y6 (Canada)
Phillip Lee Wing, and Brent J. Skura Department of Food Science, (Received
November
20th,
University of British Columbia, Vancouver (Canada) 1990; accepted
for publication
February
1lth. 1991)
ABSTRACT
The structure bacterium
implicated
and ID- and 2D-n.m.r.
of the exocellular in the spoilage spectroscopy.
glycocalyx
of meat,
The polysaccharide,
other and to the meat tissue, has the regular jr-o-ManpNAc-(1-4)~/?-D-Glcp-(I
polysaccharide
has been determined
repeating
+. Random
partial
of Pseudomonas fragi ATCC
4973, a
using hydrolysis,
analysis
methylation
which aids in the adhesion
of the cells to each
unit +4)-3-O-[(R)-l-carboxyethyll-1-D-Glcp-(I U-acetylation
occurred
in some preparations
-3). of the
polysaccharide.
INTRODUCTION
Pseudomonasfragi, which are motile, aerobic, rod-shaped, Gram-negative bacteria, are members of the group of bacteria’ responsible for the spoilage of meat in moist, aerobic conditions at temperatures of - 20”. The production of slime by the bacteria is implicated in the spoilage process. The attachment of bacterial cells to surfaces involves reversible sorption associated with hydrophobic and van der Waals forces2, and then permanent adhesion which involves the formation of a glycocalyx between the bacteria and the substrate3. Costerton et ~1.~demonstrated that the micro-organisms adhere to substrates by means of a mass of tangled fibres ofpolysaccharide with the formation of a “felt-like glycocalyx”. P.fragiexudes such a polysaccharide’ that aids in its adhesion to the meat surface and to its neighbours. We now report the structure of the polysaccharide obtained from P. fragi ATCC 4973. RESULTS
AND DISCUSSION
Isolation, composition, andn.m.r. spectra of thepolysaccharide.
4973 bacteria were grown on trypticase-soy * Dedicated to Professor Grant ’ Author for correspondence. 0008-62
I S/91/$03.50
Buchanan
on the occasion
@ 1991 ~ Elsevier
-P.
fragi ATCC
agar, harvested, and suspended in aqueous of his 65th birthday.
Science Publishers
B.V
CH Ok 1 L
101;
-?‘9
CH
‘\ ‘\
:
y \
j
HC+O;CH L__-_.__ ‘L'
Ac!lCH2 \
101;
CHDOAc
375;
LvlAc --__-_-_ rCHOAc
i AcOCH?
~_1_1______--__24 .. _j ;CHDAc
I 1 ____!CHUAc 320 / CH+lMe
1
CH$Ac
i I ' CHOAc it-__-____
2
_-' -h 1 9
AcOCD, / + ; HC?O& ,j j '_k----____
_-'
CHJ
/
103'
FHOAc ;HOAc CH$kC :i
POLYSACCHARIDE
FROM
Pseudomonasfragi
497
HOD
/
J_JJ&, , b-r/
1
50
4.0
45
,--
/i
p p m.
\
.J
,
\, 2.0
3.5
1.5
Fig. 1. 500-MHz ‘H-n.m.r. spectrum of a solution of the P. fragi polysaccharide
aqueous
48% hydrogen
product function
showed that cleavage to glucose resided in the ether substituent. ‘H-N.m.r.
the assignments
bromide’”
and 13C-n.m.r. shown
with respect
the chemical
shifts
configuration.
of the ‘H resonances
experiments,
case (3-O-lactylglucose, polysaccharide
In order
derived
downfield
and
the acidic
The
may now be assigned
to determine
sugar
‘H signals
methylation,
at 6 1.41 and
configuration
were prepared
reveal, as expected,
downfield
and confirmed
by the
respectively,
of the
can be assigned the R and S
of the 3LacGlc,
and R-( +)-2-chloropro-
from S-( -)-
shifts for the C-3 resonances,
3-substitution
are similar
and no major differences
differences
include
H-6 resonance (0.891 p.p,m.
the slightly
by the lactyl group.
and the downfield
chemical
for cc-S and 1.384 p.p,m,
data do not allow the configuration
compared
(see Exin Table II,
with the data for
The two sets of n.m.r.
the R and S isomers
between
lower chemical
and
4.35 given
pionic acidi2, respectively, and 1,2:5,6-di-O-isopropylidene-D-glucofuranose perimental). The n.m.r. data for the sodium salts of the two sugars, shown glucose,
of
whereas the gluco
as 3-O-( l-carboxyethyl)glu-
to the CH, and CH groups,
the absolute
the
shift of the C-3 resonances
the signal at 19.5 1 p.p.m. in the ‘C spectrum
of 3-O-lactyl-D-glucose
from
that the acidic
confirmed the 3-substitution, the J values accorded with
with data from the hydrolysis,
identified
3LacGlc).
lactyl substituent; likewise, to the CH, group. isomers
together
acetate
and established
(1D and 2D) of the acidic sugar allowed
I. The distinct
to those of glucose”,
These results,
de-etherification
had occurred,
spectroscopy
in Table
the anomers,
and g.1.c. of the alditol
(acid form) in D,O at 85”.
shift (0.063 p.p.m.
are apparent. downfield)
shifts of the C-2 resonances
for/&S’). The differences
data Minor
of the a-R
of the S isomer
in the two sets of n.m.r.
of the lactyl ether to be assigned
unambiguously.
However, the [a], values (+ 69” for R and - 1.7” for S) of the two synthetic sugars and the retention times of their derived alditol acetates in g.1.c. were quite different (2.57 min
B
a
4.642 7.8 96.918
H ‘J
C
c
5.252 3.2 92.603
4.642 7.8 96.675
3.6 93.103
3.611
3.302 9.5 73.988
3.599 9.5 71.671
3.319 9.2 75.372
9.5 72.562
2 3.560
3.382 8.9 85.121
3.569 9.5 82.412
3.384 9.0 85.159
9.1 82.460
3
4 3.478
5 3.819
6 3.964
6’
3.5017 9.9 70.688
3.502 8.4 70.688
3.480 9.1 69.618
9.6 69.662
76.642
3.425
72.422
3.820
16.145
3.463
72.492
5.7
3.713 2.1,12.3 61.427
3.757 1.6,12.5 61.257
3.883 5.6
3.828 5.4
3.878 1.9,12.3 5.5 61.618
3.701
1.4,12.5 61.448
3.826
4.306 7.0 79.551
4.283 7.0 79.551
4.248’ 7.0 79.435’
7.0 79.510’
4.289h
CH
5.13
I
jH ‘J
C
H ‘J
‘J (Hz) c
w
Lactyl
Proton or carbon
19.590
I.411
19.590
1.411
19.581
1.410
19.581
1.410
CH,
182.863
182.863
182.856d
I 82.956d
COOH
” Chemical shift in p_p.m. downfield from acetone (6 2.23 for ‘H and 31.07 for “C). Spectra were recorded for the sodium salt form in D,O at 30” and 300 MHz. ’ ’ Values may be interchanged.
R
a
s
B
Anomer
Isomer
N.m.r data for 3-@(R)- and 3-O-[(S)-I-carboxyethyl]-~-glucose
TABLE II
5 Q-.
B
z
k!
3
E
Ti
a
POLYSACCHARIDE
FROM Pseudomonasfragi
for R and 2.45 min for S, relative indicated
501
to that
the acidic sugar from P. fiagi
ZD-N.m.r.
spectroscopy
the polysaccharide experiments, in order
from a high-temperature
enhanced.
The data indicated
The /I-ManNAc the ‘H-n.m.r.
assignment spectrum
new gluco-type” C-4 of 3LacGlc
was confirmed
of ax
were
that had been resolution-
by the absence
to be /I.
of the signal at 6 4.89 from
polysaccharide
and by the appearance
of a
analysis. spectra,
were complicated
which were all acquired
by the twinning
of the twinning
polysaccharide,
acquisition
and with different but, on lowering
to partial
with NaBD,
and conversion
then revealed
the derivative
This result
also explains
the small
with different was detectable.
by reducing
of the products
3 and the characteristic
proportion
samples
of the
At 85”, there was This effect was
group of the lactyl moiety
was demonstrated
G.l.c.--m.s.
of some of the signals, e.g., H- 1 of
varied
twinning
of the carboxylic
formation
on the acid form of
times and temperatures.
the temperature,
lactonisation
Lactone
polysaccharide
at the lower
a hydrolysate
into the alditol fragment
of the acetates.
with m/z 103.
of “hydroxypropylhexitol”
derivative
above.
Sugar sequence. determined
by 2D-n.0.e.
residue n.O.e.‘s
configurations
observed
between
the sequence of a; however,
~
The sequence
of the sugar residues
spectroscopy
(NOESY”),
are listed in Table
assigned
TABLE
spectrum
are labelled
The J values
to be CX,and the Glc and ManNAc
a and H-2 of b and c. The extent
described
resonances.
spectra
and HETCOR”
signal for H-lb at 6 4.5 (jJ 8 Hz). The glycosylation shifts observed for and Glc and C-3 of ManNAc accorded with the substitution patterns
the polysaccharide,
temperatures.
COSY’J,
in Table III. The sugar residues
of the deaminated
from the methylation
ascribed
RELAY
1D ‘H-n.m.r.
and
1-carboxyethyl]-D-glucose.
shifts of their H-l
the 3LacGlc
The 1D- and 2D-‘H-n.m.r.
no twinning
on DB-225)
~ The ‘H- and ‘C-n.m.r.
COSY’3,
and the data are presented
measured
deduced
using
chemical
hexa-acetate
to be 3-0-[(R)-
of thepolysaccharide.
were assigned
of the decreasing
of glucitol
IV. The intra-residue
and conformations
N.O.e.‘s were also observed
because
between
N.0.e. data for the P. fru~i polvsaccharide
al -t4)-a-LacGlc
3.95(b3), 3.60 (a2)
bl -3)./I-ManNAc
4.58(b2).3.95(b3),3.66(c4), 3.48(b5)
Cl
+4)-/)‘-Glc + CH of Lx
3.65(c3), 3.51(c5), 3.60(a3,a2)
unit was and intra-
The inter-residue
with the n.O.e.‘s
H-l of b and H-4 of c, indicated the lactyl CHand
the lactyl CH and H-l of c resonated
IV
inter-
n.O.e.‘s are consistent
of the sugars.
H-l of a and H-3 of b, and between
a-kc.
in the repeating
and the observed
H-2 and H-3
so closely together
in the
POLYSACCHARIDEFROM
Pseudomonasfragi
503
h and then ultracentrifuged (60 0009, 3.5 h), and the polysaccharide was precipitated from the supernatant solution with ethanol. A solution of the precipitate in water was centrifuged and freeze-dried. A solution of the residue in 50mM phosphate buffer (pH 7) was treated with trypsin (37”, 16 h). After destruction of the enzyme by heat (1 h, steam bath), the solution was dialysed against tap water (3 days), decationised, and freezedried. Gel-permeation chromatography yielded purified polysaccharide (1.6 g) that had a fairly broad distribution of molecular weights. Prior to n.m.r. studies, the polysaccharide was O-deacetylated with 0.1 M NaOH for 4 h at 40”. Isolation and characterisation of 3-O-r(R)-I-carboxyethyll-D-glucose. - The polysaccharide (50 mg) was hydrolysed (4M trifluoroacetic acid, 125”, 1 h), the acid was removed under vacuum, and the hydrolysate was applied to a column (1 x 7 cm) of Amberlite IRA-400 (AcO-) resin. The neutral and amino sugars were eluted with water, and the acidic sugar with M formic acid. The acid was removed under vacuum and a solution of the residue in water was freeze-dried to yield the title sugar (12 mg), [a], + 8 lo (sodium salt, c 0.64, water). A portion (2 mg) was treated with aqueous 48% hydrogen bromide (0.5 mL) for 6 min at loo”, the acid was neutralised (Ag2C0,), the salts were removed by ultracentrifugation, and the supernatant solution was concentrated. The product was converted into the alditol acetate and analysed by g.1.c. The acidic sugar was further analysed by lD- and 2D-n.m.r. spectroscopy (COSY, HETCOR), and the results are listed in Table I. Preparation of 3-O-[(R)-l-carboxyethylj-D-glucose and its S isomer. - 1,2:5,6Di-O-isopropylidene-D-glucofuranose (100 mg) was treated” separately with R-( + )and S-( -)-2-chloropropionic acid (0.3 mL) and NaH in dry tetrahydrofuran. The products were purified by chromatography (1: 1 dichloroethane-ether) on silica gel and hydrolysed (2~ trifluoroacetic acid, loo”, 2 h). Gel-permeation chromatography on Bio-Gel P-2 (H20) of the products yielded 3-0-[(R)- 1-carboxyethyl]-D-glucose (25 mg), [(xl, + 69” (sodium salt, c 1.25, water) and 3-O-[(51- 1-carboxyethyll-D-glucose (30 mg), [a], - 1.7” (sodium salt, c 1.35, water). A portion (2 mg) of each sugar was methanolysed, carboxyl-reduced, hydrolysed, and analysed by g.1.c. on DB-225 (220”) as the alditol acetate. 1D-N.m.r. and ZD-n.m.r. spectroscopy (COSY, HETCOR) of the two isomers (as sodium salts) gave the data in Table II. N.m.r. spectroscopy. - Samples were deuterium-exchanged by freeze-drying solutions in D,O, and then dissolved in 99.99% D,O (0.45 mL) that contained a trace of acetone as internal reference (S 2.23 for ‘H, and 31.07 p.p.m. for “C). Spectra were recorded at 30”, 45”, and 85” with a Bruker WM-500 spectrometer equipped with an Aspect 2000 computer, and/or Bruker AM-400 and AM-300 spectrometers equipped with Aspect 3000 computers. ‘H-Homonuclear shift-correlated experiments (COSY13 and l-step RELAYCOSY14) and homonuclear dipolar-correlated experiments (NOESY”) were performed at 30”, using spectral widths sufficient to include the areas of interest. Data matrices of 512 x 2048 data points were collected for 32 transients for each t, delay. The matrices were zero-filled in the t, dimension and transformed in the magnitude mode by use of a
Financiai support was provided by the Foundation f-or-Research Dc~~elopment (Pretoria) (to H.P.). and the Natural Sciences and Engineering Rcscarch C‘i)ttncil oI Canada (to G.G.S D. ).
495
Structure of the glycocalyx polysaccharide fragi ATCC 4973” Lesley A. S. Parolis, Haralambos School of Pharmaceutical
of Pseudomonas
Parolis+,
Sciences, Rhodes University, Grahamstown 6140 (South Ajkica)
Guy G. S. Dutton, Chemistry Department,
University of British Columbia, Vancouver V6T I Y6 (Canada)
Phillip Lee Wing, and Brent J. Skura Department of Food Science, (Received
November
20th,
University of British Columbia, Vancouver (Canada) 1990; accepted
for publication
February
1lth. 1991)
ABSTRACT
The structure bacterium
implicated
and ID- and 2D-n.m.r.
of the exocellular in the spoilage spectroscopy.
glycocalyx
of meat,
The polysaccharide,
other and to the meat tissue, has the regular jr-o-ManpNAc-(1-4)~/?-D-Glcp-(I
polysaccharide
has been determined
repeating
+. Random
partial
of Pseudomonas fragi ATCC
4973, a
using hydrolysis,
analysis
methylation
which aids in the adhesion
of the cells to each
unit +4)-3-O-[(R)-l-carboxyethyll-1-D-Glcp-(I U-acetylation
occurred
in some preparations
-3). of the
polysaccharide.
INTRODUCTION
Pseudomonasfragi, which are motile, aerobic, rod-shaped, Gram-negative bacteria, are members of the group of bacteria’ responsible for the spoilage of meat in moist, aerobic conditions at temperatures of - 20”. The production of slime by the bacteria is implicated in the spoilage process. The attachment of bacterial cells to surfaces involves reversible sorption associated with hydrophobic and van der Waals forces2, and then permanent adhesion which involves the formation of a glycocalyx between the bacteria and the substrate3. Costerton et ~1.~demonstrated that the micro-organisms adhere to substrates by means of a mass of tangled fibres ofpolysaccharide with the formation of a “felt-like glycocalyx”. P.fragiexudes such a polysaccharide’ that aids in its adhesion to the meat surface and to its neighbours. We now report the structure of the polysaccharide obtained from P. fragi ATCC 4973. RESULTS
AND DISCUSSION
Isolation, composition, andn.m.r. spectra of thepolysaccharide.
4973 bacteria were grown on trypticase-soy * Dedicated to Professor Grant ’ Author for correspondence. 0008-62
I S/91/$03.50
Buchanan
on the occasion
@ 1991 ~ Elsevier
-P.
fragi ATCC
agar, harvested, and suspended in aqueous of his 65th birthday.
Science Publishers
B.V
CH Ok 1 L
101;
-?‘9
CH
‘\ ‘\
:
y \
j
HC+O;CH L__-_.__ ‘L'
Ac!lCH2 \
101;
CHDOAc
375;
LvlAc --__-_-_ rCHOAc
i AcOCH?
~_1_1______--__24 .. _j ;CHDAc
I 1 ____!CHUAc 320 / CH+lMe
1
CH$Ac
i I ' CHOAc it-__-____
2
_-' -h 1 9
AcOCD, / + ; HC?O& ,j j '_k----____
_-'
CHJ
/
103'
FHOAc ;HOAc CH$kC :i
POLYSACCHARIDE
FROM
Pseudomonasfragi
497
HOD
/
J_JJ&, , b-r/
1
50
4.0
45
,--
/i
p p m.
\
.J
,
\, 2.0
3.5
1.5
Fig. 1. 500-MHz ‘H-n.m.r. spectrum of a solution of the P. fragi polysaccharide
aqueous
48% hydrogen
product function
showed that cleavage to glucose resided in the ether substituent. ‘H-N.m.r.
the assignments
bromide’”
and 13C-n.m.r. shown
with respect
the chemical
shifts
configuration.
of the ‘H resonances
experiments,
case (3-O-lactylglucose, polysaccharide
In order
derived
downfield
and
the acidic
The
may now be assigned
to determine
sugar
‘H signals
methylation,
at 6 1.41 and
configuration
were prepared
reveal, as expected,
downfield
and confirmed
by the
respectively,
of the
can be assigned the R and S
of the 3LacGlc,
and R-( +)-2-chloropro-
from S-( -)-
shifts for the C-3 resonances,
3-substitution
are similar
and no major differences
differences
include
H-6 resonance (0.891 p.p,m.
the slightly
by the lactyl group.
and the downfield
chemical
for cc-S and 1.384 p.p,m,
data do not allow the configuration
compared
(see Exin Table II,
with the data for
The two sets of n.m.r.
the R and S isomers
between
lower chemical
and
4.35 given
pionic acidi2, respectively, and 1,2:5,6-di-O-isopropylidene-D-glucofuranose perimental). The n.m.r. data for the sodium salts of the two sugars, shown glucose,
of
whereas the gluco
as 3-O-( l-carboxyethyl)glu-
to the CH, and CH groups,
the absolute
the
shift of the C-3 resonances
the signal at 19.5 1 p.p.m. in the ‘C spectrum
of 3-O-lactyl-D-glucose
from
that the acidic
confirmed the 3-substitution, the J values accorded with
with data from the hydrolysis,
identified
3LacGlc).
lactyl substituent; likewise, to the CH, group. isomers
together
acetate
and established
(1D and 2D) of the acidic sugar allowed
I. The distinct
to those of glucose”,
These results,
de-etherification
had occurred,
spectroscopy
in Table
the anomers,
and g.1.c. of the alditol
(acid form) in D,O at 85”.
shift (0.063 p.p.m.
are apparent. downfield)
shifts of the C-2 resonances
for/&S’). The differences
data Minor
of the a-R
of the S isomer
in the two sets of n.m.r.
of the lactyl ether to be assigned
unambiguously.
However, the [a], values (+ 69” for R and - 1.7” for S) of the two synthetic sugars and the retention times of their derived alditol acetates in g.1.c. were quite different (2.57 min
B
a
4.642 7.8 96.918
H ‘J
C
c
5.252 3.2 92.603
4.642 7.8 96.675
3.6 93.103
3.611
3.302 9.5 73.988
3.599 9.5 71.671
3.319 9.2 75.372
9.5 72.562
2 3.560
3.382 8.9 85.121
3.569 9.5 82.412
3.384 9.0 85.159
9.1 82.460
3
4 3.478
5 3.819
6 3.964
6’
3.5017 9.9 70.688
3.502 8.4 70.688
3.480 9.1 69.618
9.6 69.662
76.642
3.425
72.422
3.820
16.145
3.463
72.492
5.7
3.713 2.1,12.3 61.427
3.757 1.6,12.5 61.257
3.883 5.6
3.828 5.4
3.878 1.9,12.3 5.5 61.618
3.701
1.4,12.5 61.448
3.826
4.306 7.0 79.551
4.283 7.0 79.551
4.248’ 7.0 79.435’
7.0 79.510’
4.289h
CH
5.13
I
jH ‘J
C
H ‘J
‘J (Hz) c
w
Lactyl
Proton or carbon
19.590
I.411
19.590
1.411
19.581
1.410
19.581
1.410
CH,
182.863
182.863
182.856d
I 82.956d
COOH
” Chemical shift in p_p.m. downfield from acetone (6 2.23 for ‘H and 31.07 for “C). Spectra were recorded for the sodium salt form in D,O at 30” and 300 MHz. ’ ’ Values may be interchanged.
R
a
s
B
Anomer
Isomer
N.m.r data for 3-@(R)- and 3-O-[(S)-I-carboxyethyl]-~-glucose
TABLE II
5 Q-.
B
z
k!
3
E
Ti
a
POLYSACCHARIDE
FROM Pseudomonasfragi
for R and 2.45 min for S, relative indicated
501
to that
the acidic sugar from P. fiagi
ZD-N.m.r.
spectroscopy
the polysaccharide experiments, in order
from a high-temperature
enhanced.
The data indicated
The /I-ManNAc the ‘H-n.m.r.
assignment spectrum
new gluco-type” C-4 of 3LacGlc
was confirmed
of ax
were
that had been resolution-
by the absence
to be /I.
of the signal at 6 4.89 from
polysaccharide
and by the appearance
of a
analysis. spectra,
were complicated
which were all acquired
by the twinning
of the twinning
polysaccharide,
acquisition
and with different but, on lowering
to partial
with NaBD,
and conversion
then revealed
the derivative
This result
also explains
the small
with different was detectable.
by reducing
of the products
3 and the characteristic
proportion
samples
of the
At 85”, there was This effect was
group of the lactyl moiety
was demonstrated
G.l.c.--m.s.
of some of the signals, e.g., H- 1 of
varied
twinning
of the carboxylic
formation
on the acid form of
times and temperatures.
the temperature,
lactonisation
Lactone
polysaccharide
at the lower
a hydrolysate
into the alditol fragment
of the acetates.
with m/z 103.
of “hydroxypropylhexitol”
derivative
above.
Sugar sequence. determined
by 2D-n.0.e.
residue n.O.e.‘s
configurations
observed
between
the sequence of a; however,
~
The sequence
of the sugar residues
spectroscopy
(NOESY”),
are listed in Table
assigned
TABLE
spectrum
are labelled
The J values
to be CX,and the Glc and ManNAc
a and H-2 of b and c. The extent
described
resonances.
spectra
and HETCOR”
signal for H-lb at 6 4.5 (jJ 8 Hz). The glycosylation shifts observed for and Glc and C-3 of ManNAc accorded with the substitution patterns
the polysaccharide,
temperatures.
COSY’J,
in Table III. The sugar residues
of the deaminated
from the methylation
ascribed
RELAY
1D ‘H-n.m.r.
and
1-carboxyethyl]-D-glucose.
shifts of their H-l
the 3LacGlc
The 1D- and 2D-‘H-n.m.r.
no twinning
on DB-225)
~ The ‘H- and ‘C-n.m.r.
COSY’3,
and the data are presented
measured
deduced
using
chemical
hexa-acetate
to be 3-0-[(R)-
of thepolysaccharide.
were assigned
of the decreasing
of glucitol
IV. The intra-residue
and conformations
N.O.e.‘s were also observed
because
between
N.0.e. data for the P. fru~i polvsaccharide
al -t4)-a-LacGlc
3.95(b3), 3.60 (a2)
bl -3)./I-ManNAc
4.58(b2).3.95(b3),3.66(c4), 3.48(b5)
Cl
+4)-/)‘-Glc + CH of Lx
3.65(c3), 3.51(c5), 3.60(a3,a2)
unit was and intra-
The inter-residue
with the n.O.e.‘s
H-l of b and H-4 of c, indicated the lactyl CHand
the lactyl CH and H-l of c resonated
IV
inter-
n.O.e.‘s are consistent
of the sugars.
H-l of a and H-3 of b, and between
a-kc.
in the repeating
and the observed
H-2 and H-3
so closely together
in the
POLYSACCHARIDEFROM
Pseudomonasfragi
503
h and then ultracentrifuged (60 0009, 3.5 h), and the polysaccharide was precipitated from the supernatant solution with ethanol. A solution of the precipitate in water was centrifuged and freeze-dried. A solution of the residue in 50mM phosphate buffer (pH 7) was treated with trypsin (37”, 16 h). After destruction of the enzyme by heat (1 h, steam bath), the solution was dialysed against tap water (3 days), decationised, and freezedried. Gel-permeation chromatography yielded purified polysaccharide (1.6 g) that had a fairly broad distribution of molecular weights. Prior to n.m.r. studies, the polysaccharide was O-deacetylated with 0.1 M NaOH for 4 h at 40”. Isolation and characterisation of 3-O-r(R)-I-carboxyethyll-D-glucose. - The polysaccharide (50 mg) was hydrolysed (4M trifluoroacetic acid, 125”, 1 h), the acid was removed under vacuum, and the hydrolysate was applied to a column (1 x 7 cm) of Amberlite IRA-400 (AcO-) resin. The neutral and amino sugars were eluted with water, and the acidic sugar with M formic acid. The acid was removed under vacuum and a solution of the residue in water was freeze-dried to yield the title sugar (12 mg), [a], + 8 lo (sodium salt, c 0.64, water). A portion (2 mg) was treated with aqueous 48% hydrogen bromide (0.5 mL) for 6 min at loo”, the acid was neutralised (Ag2C0,), the salts were removed by ultracentrifugation, and the supernatant solution was concentrated. The product was converted into the alditol acetate and analysed by g.1.c. The acidic sugar was further analysed by lD- and 2D-n.m.r. spectroscopy (COSY, HETCOR), and the results are listed in Table I. Preparation of 3-O-[(R)-l-carboxyethylj-D-glucose and its S isomer. - 1,2:5,6Di-O-isopropylidene-D-glucofuranose (100 mg) was treated” separately with R-( + )and S-( -)-2-chloropropionic acid (0.3 mL) and NaH in dry tetrahydrofuran. The products were purified by chromatography (1: 1 dichloroethane-ether) on silica gel and hydrolysed (2~ trifluoroacetic acid, loo”, 2 h). Gel-permeation chromatography on Bio-Gel P-2 (H20) of the products yielded 3-0-[(R)- 1-carboxyethyl]-D-glucose (25 mg), [(xl, + 69” (sodium salt, c 1.25, water) and 3-O-[(51- 1-carboxyethyll-D-glucose (30 mg), [a], - 1.7” (sodium salt, c 1.35, water). A portion (2 mg) of each sugar was methanolysed, carboxyl-reduced, hydrolysed, and analysed by g.1.c. on DB-225 (220”) as the alditol acetate. 1D-N.m.r. and ZD-n.m.r. spectroscopy (COSY, HETCOR) of the two isomers (as sodium salts) gave the data in Table II. N.m.r. spectroscopy. - Samples were deuterium-exchanged by freeze-drying solutions in D,O, and then dissolved in 99.99% D,O (0.45 mL) that contained a trace of acetone as internal reference (S 2.23 for ‘H, and 31.07 p.p.m. for “C). Spectra were recorded at 30”, 45”, and 85” with a Bruker WM-500 spectrometer equipped with an Aspect 2000 computer, and/or Bruker AM-400 and AM-300 spectrometers equipped with Aspect 3000 computers. ‘H-Homonuclear shift-correlated experiments (COSY13 and l-step RELAYCOSY14) and homonuclear dipolar-correlated experiments (NOESY”) were performed at 30”, using spectral widths sufficient to include the areas of interest. Data matrices of 512 x 2048 data points were collected for 32 transients for each t, delay. The matrices were zero-filled in the t, dimension and transformed in the magnitude mode by use of a
Financiai support was provided by the Foundation f-or-Research Dc~~elopment (Pretoria) (to H.P.). and the Natural Sciences and Engineering Rcscarch C‘i)ttncil oI Canada (to G.G.S D. ).