Structural analysis of a second acidic exopolysaccharide of Rhizobium meliloti that can function in alfalfa root nodule invasion

Carbohydrate

Research,

339

210 (1991) 339-347

Elsevier Science Publishers B.V., Amsterdam

Note

Structural analysis of a second acidic exopolysaccharide of Rhizohium meliloti that can function in alfalfa root nodule invasion B. Levery *’ , Hangjun

Steven

Hakomori*j** * The Bionwmhrane

Institute,

isrry’, Microbiologist.

Zhan *, Chi Chang

Lee*, John

201 Elliotr Ave W, Searfle,

und Purhohioloy,

**, Uniwrsirv

A: LeighI,

WA 98119 [U.S.A.)

_ of Washington,

and Sen-itiroh

of’Chem-

und Departments

Sr~artle, WA 98 195 (U.S.A.)

(Received May 23rd, 1990; accepted for publication in revised form September I st. 1990)

Symbiotic

bacteria

rides that appear produce

of the genus Rhizohium

to be essential

a calcofluor-binding

secrete a variety

of acidic polysaccha-

for root nodulation’-‘.

Normal strains of R. mefiloti (EPS), succinoglycan, a polymer hav-

exopolysaccharide

ing an octasaccharide /?-Glc-( 1-3)~/?-Glc-(

repeating unit with the sequence [4,6-O-( 1-carboxyethylidene)1-3)~/I?-Glc-( I -+6)+Glc-( l-6)]-4)-P-Glc-( 1~4)+Glc-( l-+4)-

&Glc-(

1 -+.Each repeating

1+3)-/?-Glc-(

one acetyl mutants,

and

one succinyl

which

glycan,

either

produce

form “empty”

they contain ride (EPSb),

results

describe

on alfalfa

are also

have been

the structural

monosaccharide

capable

of them

reported

Anul.ysis ojmonosucchuride acidic methanolysis,

Fig. 1. Peaks derivatives identifiable. Dudman pyranoside consistent acetal

methyl whose

0008-62 IS:‘91‘$03.50

g.l.c.-e.i.m.s.

analysis

times and e.i. mass spectra

spectra

were found

identification spectrum

@ 1991

on alfalfa,

succinoglycan’2.

Virtually

In this paper,

spectroscopy,

we

along with

composition. ~ When native EPSb was subjected to by per-O-trimethylsilylation of the resultant

methyl

gave the pattern were readily

shown

assignable

in as

by two major peaks not immediately to match

closely

those

observed

by

4,6-O-( l-carboxyethylidene)-o-galacto-

esters, while the g.1.c. peak ratios

in the I-D ‘H-n.m.r.

exopolysaccha-

nodules

and Walker”.

and

followed

and Lacey14 for the anomeric with the earlier

since

the isolation

by g.l.c.~m.s.

and glucose were followed

The e.i. mass methyl

of producing

exe

of the

fixation,

distinct

of EPSb by ‘H-n.m.r.

glycosides,

retention

of galactose

we reported

nitrogen-fixing

by Glazebrook

analysis

meliloti form

in nitrogen

a structurally

of forming

characterization

anhydrous

monosaccharide

secreting

with approximately

or a non-succinylated

Recently,

are incapable

and methylation

substituted

sites8’9. Rhizohium

that are ineffective

bacteroids’,‘“.“.

of R. meliloti mutants, with

a number

identical

unit is further

at undetermined

no succinoglycan

nodules

no intracellular

characterization although

group

of a resonance

appeared

similar.

for the methyl

of EPSb”.

Elsevier Science Publishers 9.V

This result was group of pyruvic

310

341

NOTE

The identification of the two peaks as arising from I-carboxyethylidene derivatives was confirmed by analysis using g.l.c.-c.i.m.s. with isobutane as reagent gas. The c.i. mass spectra of both derivatives were characterized by quasimolecular ions at m,/z 423 (M-H -) and 479 (MC,H,+), corresponding to the correct calculated nominal mass of 422 a.m.u. (see Fig. 2C). These peaks were accompanied by abundant fragments from loss of neutral MeOH (m/z 391) and methyl formate (mj: 363) (Scheme 1). This pattern can be contrasted with that typical for methyl per-O-trimethylsilyl hexopyranosides (see Fig. 2A), with quasimolecular ions at m/z 483 (M.H ‘) and 539 (MC&I;‘), accompanied by fragments for neutral losses of CH, (m/z 467), MeOH (m/z 451), and, in most cases, an additional loss of (CH~)~HSiOCH~) (I%/=361).

r

+

0

MH+ = 423 a.m.u OMe, H

391 a.m.u

Scheme 1

Coniponent sugar c u/u) 4,6-O-I-curbo~yethJ?[ihncGal (I) Native (21 Au. HOAc treated

51.6

12.1

36.3

54.8

45.2

not detected

-

‘1.

j

-‘.3.!

1.

.I

!.-l,

-,!

,?

_,

,,:_

343

NOTE

each of 3-proton singlets at typical chemical shifts for acetate (2.152 p.p.m.) and pyruvate (1.459 p.p.m.) were initially observed in the spectrum of intact EPSb, as reported previously”. On heating, the latter peak was seen to decay, being replaced by a singlet at 1.464 p.p.m. (free pyruvate), while the former peak was replaced by a singlet at 1.907 p.p.m. (free acetate). In a short time, the pyruvate groupwascompletely lost, even when I OOmMphosphate buffer at pD 9.8 was used as solvent. Therefore, all spectra for this study were acquired after complete hydrolysis of the pyruvic acetal. The downfield portion of a resolution-enhanced 1-D n.m.r. spectrum of depyruvated EPSb is reproduced in Fig. 3. Unambiguous assignments for all proton resonances in the ‘I-I-n.m.r. spectrum of depyruvated EPSb were obtained by successive applications of the Z-D DQF-PSCOSY, RELAY, TOCSY, and TQF-COSY sequences. Results from the latter two experiments are reproduced in Fig. 4A, and the assignments are summarized in Table II. Only two spin-systems were observed, which could be followed from anomeric resonances at 5.42 and 4.75 p.p.m.; analysis of splitting constants made it clear that these were from a-Gal and P-Glc residues, respectively. It was hoped that PS-NOESY data would give indications of linkage sites by displaying clear interglycosidicenhancements. The expected ok-diagonal peaks were indeed observed, one interglycosidic crossrelaxation for each anomeric proton {Fig. 4B). However, in each case the n.0.e. cross peak corresponded to a pair of degenerate resonances, to the cc-Gal H-2/3 from b-Glc H-l, and to the /?-Glc H-3/4 from a-Gal H-l. Thus, it was impossible to specify more precisely the inter-residue linkage sites. This was left to methylation analysis, as described next. Merhylurion analysis. -- Following successive permethylation, hydrolysis, reduction, and acetylation, intact EPSb yielded a simple pattern of partially methylated alditoi acetates (PMAAs): mainly 2,4,6-tri-0-Me-Glc and 2-mono-O-Me-Gal, with a small amount of 2,4,6-tri-O-Me-Gal. When EPSb was depyruvated by acid treatment prior to permethylation, only 2,4,6-tri-0-Me-Glc and a-Gal were obtained. This clearly indicates +3Glc and -+3Gal linkages exclusively, while confirming the 4,6-location of the I-carboxyethylidene groups. These results show that EPSb has a simple repeating-unit structure, -+3Galcc1-+3Glc/3l-+ 4, 5’ C CH/, >OIH confirming our previous proposal”. An acetate group is present at an as yet undetermined site. The same structure was also proposed by Glazebrook and Walker13, who further specified the site of acetylation as the &position of /3-Glc. Details of that structure elucidation appeared just as this paper was being submittedn’. It is interesting to consider how EPSb can perform the same function as the R. ~~~ilo~~succinoglycan, which has a considerably more complex octasaccharidc repeat-

Il-4: d I-1 . i

TQF.

! t

i 5.20

5.00

4.m

4.50

4.40

4.20

4.00

3

85

3.60

3

46

5. &x

p.p_m.

345

NOTE

ing-unit.

As pointed

feature

out by Glazebrook

in common,

might represent

and Walker’j,

the /I-Glc-(l-3)-Gal

a common

if the corresponding

recognition

OAc groups

occupy

the corresponding

pyruvated

as in EPSb. Thus, the proposed

ring (around

galactose

topology,

particularly is not a rather

only the unmodified of the glucosyl

part of the galactose residue.

might be questionable.

and strains

host-related

to which

process,

succinoglycan

structures

the extent

they suggest

would have to occupy

even within the family of related succinoglycan of Rhizohia,

will be required

produced

for clarification

Whether

this is

It is worth noting that, by different

specificity

shared US. unique features is an issue still open to debate15.“. experiments, along with a careful analysis of the three-dimensional of the polysaccharides,

have one

which

in both polysaccharides.

in R. meliloti

determinant

portion

specificity

invasion

the same position residue

including

HO-2), and a contiguous

sufficient to confer the required

linkage,

site in the nodule

However,

small area of molecular

the two structures

disaccharide

species

is conferred

by

Further biological structural features

of this question.

EXPERIMENTAL

strain

Isolution

ofEPSh.~ EPSb was isolated from culture supernatants

Rm7011

exoA (pMuc)

Dqqwvation.

-

as previously

Samples

of EPSb were treated

heated in a sealed tube for 20 hat 100”. Following at 37”, using EtOH

as co-distallant,

Monosaccharide

samples

analysis. -To

of R. meliloti

described”. with 10% v/v aq. HOAc (1 mL),

removal

of HOAc by flushing with N?

were dried in cucuo over NaOH

samples

for 3 h.

of EPSb, before and after depyruvation

(100 pg each), in 13 x 100 test tubes with Teflon-lined screw caps, were added M HCl in anhydrous MeOH (1 .O mL). These were sealed and heated for 16 h at 80”. The solutions were made neutral were centrifuged with MeOH under

and the MeOH

N?, and then dried were converted

selective detector column either

supernatant

into

with those

compounds’4. Hewlett-Packard

pertrimethylsilyl

program

by characteristic

of standard

Alternatively,

ethersi8.”

the precipitate

compounds,

system operating

once

in clean tubes

analyzed

methyl by g.l.c.-

to a 5970B mass-

phase fused silica capillary Monosaccharide

times and mass spectra compared

or with spectra

the analysis

and

interfaced

150-250“ at 4”jmin).

retention

was performed

58905 gas chromatograph

mass spectrometer/data

washing

were evaporated

using a 30-m DB-5 bonded

temperature

were identified

solutions

5890A gas chromatograph

(g.l.c.-MSD),

(splitless injection,

removed,

MeOH

in uucuo over P,O, for 2 h. The monosaccharide

on a Hewlett-Packard

derivatives

and Ac,O (50 ,uL) was added. After 6 h at 4”, the tubes

(1 .O mL). The combined

glycosides e.i.m.s.

with Ag$O,,

interfaced

published

on a system

for pyruvated consisting

of a

to a Jeol HX-1 lo/DA-5000

in the chemical-ionization

(isobutane)

mode.

Fig. 4. Sections of 2-D ‘H-n.m.r. spectra of depyruvated EPSb: A, upper left-hand sector, TQF-COSY; lower right-hand sector, TOCSY; B, NOESY. J, in panel B, refers to a cross-peak corresponding mainly to J coupling between P-Glc H-l and H-2 which was not suppressed; int. refers to interglycosidic dipolar correlations.

NOTE

347

REFERENCES I 2 3 4 5

J. A. Leigh, E. R. Signer, and G. C. Walker, Proc. Nutl. Acad. SC;. USA, 82 (1985) 6231-6235. J. A. Leigh, J. W. Reed, J. F. Hanks, A. M. Hirsch, and G. C. Walker, CelI, 51 (1987) 579-587. A. K. Chakravorty, W. Zurkowski, J. Shine, and B. G. Rolfe. J. Mol. A,&. Genet., I (1982) 5855596. H. Chen, M. Batley, J. Redmond, and B. G. Rolfe, .I. PIant Ph~md., 120 (1985) 331 --349. D. Borthakur, C. E. Barber, J. W. Lamb, M. J. Daniel% J. A. Downie. and A. W. B. Johnston, Mol. Gen. Genel., 203 (1986) 32(t323. 6 S. P. Djordjevic. H. Chen, M. Batley, J. W. Redmond, and B. G. Rolfe, J. Eucleriol., 169 (1987) 53-60. 7 J. A. Downie, and A. W. B. Johnston, Ceil, 47 (1986) 153 154. 8 P. Aman, M. McNeil, L. Franzen. A. G. Darvill, and I’. Albersheim, Curhohydr. Res., 95 (1981)2633282. 9 P.-E. Jansson, L. Kenne, B. Lindberg, H. Ljunggren, J. Liinngren, U. Ruden, and S. Svensson. 1. Am. Chem. Sot., 99 (1977) 3812~ 3815. IO T. M. Finan. A. M. Hirsch. J. A. Leigh, E. Johansen. G. A. Kuldau, S. Deggan, G. C. Walker, and E. R. Signer, Cell. 40 (1985) X69-877. 11 J. A. Leigh, and C. C. Lee, J. Bucteriol., 170 (198X) 332773332. 12 H. Zhan, S. B. Levery, C. C. Lee, and J. A. Leigh, Proc. Null. Aad. Sri. USA, 86 (1989) 3055-3059. 13 J. Glazebrook, and G. C. Walker, Cell, 56 (1989) 661-672. I4 W. F. Dudman, and M. J. Lacey, Cnrhohydr. Res., 145 (1986) 175191. 15 M. McNeil, J. Darvill, A. G. Darvill, P. Albersheim, R. Vanveen. P. Hooykaas. R. Schilperoort, and A. Dell, Curhohqdr. Res., 146 (1986) 307-326. I6 G.-R. Her, J. Glazebrook, G. C. Walker, and V. N. Reinhold, Curhohydr. Res., 198 (1990) 305-312. 17 S. Philip-Hollingsworth, R. 1. Hollingsworth, and F. B. Dazzo, J. Biol. Chem., 264 (1989) 1461-1466. I8 C. C. Sweeley, R. Bentley, M. Makita, and W. W Wells, J. Am. Chem. Sot.. 85 (1963) 249772507. 19 R. A. Lame. W. J. Esselman, and C. C. Sweeley, Mcth. Enzymol. 28 (1963) 159- 167. 20 S. Hakomori, J. Biochem. /Tokyo), 55 (1964) 205 208. 21 S. B. Levery, and S. Hakomori. Method.7 Enzymol., I38 (1987) 13-25 22 H. Clausen. S. B. Levery, E. D. Nudelman, M. R. Stroud, M. E. K. Salyan, and S. Hakomori, J. Biol. Chem., 262 (1987) 14228 14234. 23 H. Bjiirndal, C. G. Hellerqvist, B. Lindberg, and S. Svensson, Angevv. Chrm. Inrl. Ed. Eng., 9 (1970) 61@619. 24 P.-E. Jansson, L. Kenne, H. Liedgren, B. Lindberg. and J. Liinngren, Chem. Commun. Urzic. Stockholm, 8 (1976) l-75. 25 D. Marion, and K. Wiithrich, Biochem. Biophys. Rrr. Commun.. I 13 (1983)967 -974. 26 U. Piantini. 0. W. Sorensen, and R. R. Ernst, J. Am. Chem. Sot.. 104 (1982) 68OC-6801. 27 M. Rance, 0. W. Sorensen. G. Bodenhausen, G. Wagner, R. R. Ernst, and K. Wiithrich, B&hem. Biophys. Rex. Commun., I 17 (I 983) 479 485. 28 G. Eich, G. Bodenhausen, and R. R. Ernst, J. Am. Chem. .%x. 104 (1982) 3731--3732. 29 A. Bax, and G. Drobny, J. Mugn. Reson., 61 (1985) 306-320. 30 L. Braunschweiler, and R. R. Ernst, J. Magn. Re.con., 53 (1983) 521 528. 31 A. Bax, and D. G. Davis, J. Magn. Reson., 65 (1985) 355%360. 32 A. J. Shaka, and R. Freeman. J. Magn. Resort., 51 (1983) 1699173. 33 G. Bodenhausen. H. Kogler, and R. R. Ernst, J. Mqn. Resan.. 58 (1984) 37&388.