A novel highly charged exopolysaccharide produced by two strains of Stenotrophomonas maltophilia recovered from patients with cystic fibrosis

Carbohydrate Research 346 (2011) 1916–1923

Contents lists available at ScienceDirect

Carbohydrate Research journal homepage: www.elsevier.com/locate/carres

A novel highly charged exopolysaccharide produced by two strains of Stenotrophomonas maltophilia recovered from patients with cystic fibrosis Paola Cescutti a,⇑, Bruno Cuzzi a, Gianfranco Liut a, Christine Segonds b, Giovanni Di Bonaventura c,d, Roberto Rizzo a a

Dipartimento di Scienze della Vita, Università di Trieste, via L. Giorgieri 1, Ed. C11, 34127 Trieste, Italy Laboratoire de Bactériologie-Hygiène, Institut Fédératif de Biologie, Hôpital Purpan, CHU Toulouse, TSA 40031, 31059 Toulouse Cedex 9, France c Dipartimento di Scienze Biomediche, Università ‘G. d’Annunzio’ di Chieti-Pescara, Via dei Vestini, 31, 66100 Chieti, Italy d Centro Scienze sull’Invecchiamento, Fondazione Università ‘G. d’Annunzio’ di Chieti-Pescara, Via Colle Dell’Ara, 66100 Chieti, Italy b

a r t i c l e

i n f o

Article history: Received 14 March 2011 Received in revised form 2 May 2011 Accepted 9 May 2011 Available online 13 May 2011 Keywords: Stenotrophomonas maltophilia Exopolysaccharide Primary structure NMR ESI-MS Cystic fibrosis

a b s t r a c t Stenotrophomonas maltophilia is a non-fermenting Gram-negative microorganism capable of causing chronic pulmonary infection in cystic fibrosis patients and its ability to form biofilms on polystyrene and glass surfaces, as well as on cystic fibrosis-derived bronchial epithelial IB3-I cells was recently demonstrated. The latter evidence might explain the power of S. maltophilia to produce persistent lung infections, despite intensive antibiotic treatment. In addition to being important components of the extracellular biofilm matrix, polysaccharides are involved in virulence, as they contribute to bacterial survival in a hostile environment. With the aim of contributing to the elucidation of S. maltophilia virulence factors, the exopolysaccharides produced by two mucoid clinical isolates of S. maltophilia obtained from two cystic fibrosis patients were completely characterised, mainly by means of ESI-MS and NMR spectroscopy. The results showed that, although the two isolates were recovered from two different patients living in different countries (Italy and France), the exopolysaccharides produced have an identical primary structure, with the following repeating unit:

Ac 2 [4)-β-D-Glcp-(1→4)-β-D-GalpA-(1→4)-β-D-GlcpA-(1→]n 3 ↑ 1 D-Lac-3-β-D-GalpA 4 Ac The exopolysaccharide is highly negatively charged for the presence of three uronic acids on four residues in the repeating unit. Moreover, an ether-linked D-lactate substituent is located on C-3 and one Oacetyl group on C-4 of the galacturonic acid side chain. Another O-acetyl group substitutes C-2 of the galacturonic acid in the backbone, making this primary structure unique. Ó 2011 Elsevier Ltd. All rights reserved.

1. Introduction

Abbreviations: Ac, acetyl; CF, cystic fibrosis; CFTR, cystic fibrosis transmembrane conductance regulator; deAc-SMA-EPS, S. maltophilia de-O-acetylated EPS; EPS, exopolysaccharide; ESI-MS, electrospray ionization mass spectrometry; FR-I, fraction I; FR-II, fraction II; FR-III, fraction III; FR-IV, fraction IV; FR-I.4, fraction I.4; Hex, hexose; HexA, hexuronic acid; Lac, lactate; SMA-EPS, S. maltophilia EPS; SMA-EPS-R, S. maltophilia carboxyl-reduced EPS; TMS, trimethylsilyl; YEM, yeast extract-mannitol. ⇑ Corresponding author. Tel.: +39 040 5583685; fax: +39 040 5583691. E-mail address: [email protected] (P. Cescutti). 0008-6215/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.carres.2011.05.007

Stenotrophomonas maltophilia is a gram-negative, rod-shaped, obligate aerobic bacterium belonging to the gamma subdivision of Proteobacteria. It is found in the home and hospital environments, particularly in water sources and has recently emerged as an important opportunistic human pathogen in cystic fibrosis (CF),1 debilitated and immuno-compromised2 patients. CF is a common inherited genetic disorder,3 caused by a mutation in the gene encoding the cystic fibrosis transmembrane conductance regulator (CFTR) protein, which is expressed in many different cells. In CF lungs, the mutated protein causes a defect in

P. Cescutti et al. / Carbohydrate Research 346 (2011) 1916–1923

the chloride ions transport which leads to impairment of mucociliary clearance, production of thick mucus and impaired innate immunity.4,5 The CF lung environment facilitates microbial colonisation, which results in chronic bacterial infections. Staphylococcus aureus, Haemophilus influenzae and Pseudomonas aeruginosa are the most commonly isolated organisms; chronic infection with P. aeruginosa results in a decline of pulmonary function and ultimate mortality. Several other non-fermenting Gram-negative multi-drug resistant pathogens have emerged in CF such as the Burkholderia cepacia complex, S. maltophilia, Achromobacter xylosoxidans, Pandoraea species, to name a few. The clinical impact of S. maltophilia, which colonises up to one third of the CF population, is not clearly established yet, but it was recently shown that CF patients with chronic S. maltophilia infection display a specific immune response, which is in turn associated with lower lung function.6 Furthermore, S. maltophilia invades and forms biofilm on the CF-derived bronchial epithelial IB3-I cell line, which supports the persistence of this opportunistic pathogen in CF patients.7 Lastly, it was determined8 that a strain of S. maltophilia was capable of forming biofilm on polystyrene and glass and that the lipopolysaccharide/ exopolysaccharide-coupled biosynthetic genes are necessary for biofilm formation, thus introducing the importance of exopolysaccharides (EPS) in host colonisation by S. maltophilia. In this article we present the characterisation of EPS produced by two different clinical isolates of S. maltophilia: strain SMHQ isolated from a 15 year old CF patient attending the CF Unit of the Children’s Hospital (CHU Toulouse, France) and responsible for chronic colonization since the age of 4 and strain SM248 obtained from a CF patient attending the Bambino Gesù Children’s Hospital (Rome, Italy).

2. Results and discussion 2.1. Exopolysaccharide composition The yield of S. maltophilia EPS (SMA-EPS) was about 3 mg of purified polymer/Petri dish. GLC analysis of the alditol acetates derivatives showed only glucose as neutral sugar, while GlcA and GalA were revealed after methanolysis of the EPS and derivatisation of the products to TMS methyl-glycosides. However, no definite molar ratios could be obtained from these data, because of the high resistance of SMA-EPS to acid hydrolysis. Methylation analysis of the native EPS was unsuccessful giving only peracetylated glucose. All these difficulties with derivatisation of the native EPS were overcome by performing composition, glycosidic linkages and absolute configuration analyses on the carboxyl reduced EPS. 2.2. Composition, glycosidic linkage and absolute configuration determination of carboxyl reduced SMA-EPS Because SMA-EPS was very resistant to acid hydrolysis, it was carboxyl reduced (SMA-EPS-R) and used for chemical derivatisations. SMA-EPS-R was hydrolysed and the products converted to alditol acetates before being subjected to GLC and GLC–MS. The chromatogram showed Gal, Glc and a slow eluting component (13 min after inositol used as internal standard) in a molar ratio 1.00:1.70:0.30. Electron-impact GLC–MS analysis identified the latter component as 3-O-hydroxyisopropylhexitol acetate. Finding this substituent after carboxyl reduction indicated the presence of a lactate group in the native EPS. Subsequently, SMA-EPS-R was permethylated, hydrolysed, the products derivatised to alditol acetates and analysed by GLC–MS. The chromatogram showed three peaks, which were identified as terminal 3-O-hydroxyisopropylhexose, 4-linked Glc and 3,4-linked Gal in the molar ratio 0.66:1.74:1.00. While the identity of the substituted hexose could

1917

not be established because of the lack of a standard, methylation analysis suggested 4-linked Glc, 4-linked GlcA and 3,4-linked GalA in equimolar amounts in the native polymer. The absolute configuration of the sugar residues was determined on SMA-EPS-R with and without previous ether cleavage. For the removal of the hydroxyisopropyl substituent, a sample of SMA-EPS-R was treated with boron tribromide. The samples were subjected to methanolysis, butanolysis, followed by derivatisation of the products to TMS O-butyl glycosides and analysis by GLC. The results obtained indicated that all the sugar residues have the D absolute configuration. The sample analysed before removal of lactate showed D-Glc and D-Gal in the molar ratio 2:1, while, after removal of the hydroxyisopropyl group, an equimolar ratio for D-Glc and D-Gal was found, thus indicating that lactate is linked to a D-GalA residue in the native polymer. 2.3. Determination of the absolute configuration of lactate The absolute configuration of lactate was determined via enzymatic conversion to pyruvate. Because of the resistance of SMAEPS to acid hydrolysis, the terminal non reducing residue carrying the lactate substituent was isolated resorting to Smith degradation of the polysaccharide, followed by purification on a Bio Gel P2 column. The lactate-substituted residue was then treated with alkali to achieve cleavage of the ether linkage through beta elimination, a reaction that maintained the original stereoisomer. After quenching the reaction with HCl, the presence of free lactate in the solution was revealed by TLC. The solution was then tested with two enzymatic kits containing L-lactate dehydrogenase (E.C.1.1.1.27) and D-lactate dehydrogenase (E.C.1.1.1.28). No reaction was detected using the former enzyme, while the latter resulted in the production of NADH, as evidenced by an increase in absorbance at 340 nm, thus revealing that the lactate substituent is in the D (R) absolute configuration. 2.4. ESI-MS characterisation of oligosaccharides produced by partial hydrolysis of SMA-EPS A sample of SMA-EPS (15 mg) was treated with 2 M TFA for 2 h at 100 °C and the product was then separated by gel permeation chromatography on a Bio Gel P2 column; the elution profile obtained is shown in Figure 1a. Fractions (FR) were named I–V, with fraction I corresponding to the void volume of the column. Fraction II was reduced with NaBH4, to mark the reducing end, and subsequently permethylated, before being subjected to ESI-MS and MS2 analyses. The MS2 spectrum of reduced, permethylated FR-II is reported in Figure 2, and the ions assignment in Table 1. The MS spectrum showed a pseudomolecular ion at 1015.4 m/z, corresponding to the [M+Na]+ adduct of a reduced permethylated tetrasaccharide containing three HexA, one Hex and one lactate substituent. Fragmentation of this ion resulted in formation of fragments starting both from the reducing and non-reducing ends, leading to the unambiguous location of the hexose at the reducing end and to the complete saccharidic sequence assignment. Because glucose was the only hexose determined by GLC analysis of the SMA-EPS, it was deduced that in FR-II it occupied the reducing end position. FR-III contained mainly the unsubstituted tetrasaccharide HexA-HexA-HexA-Hex; because the ether linkage is usually stable to treatment with 2 M TFA, the presence of small amounts of this oligosaccharide indicated that substitution with lactate might not be stoichiometrically complete. FR-I was separated on a Bio Gel P10 column in four different peaks (FR-I.1–I.4); the chromatogram obtained is shown in Figure 1b. Fraction I.4 was reduced with NaBH4 and subjected to ESI-MS: the pseudomolecular ion at 1567.4 m/z corresponded to the average mass of the [M+Na]+ adduct of a reduced octasaccharide

1918

P. Cescutti et al. / Carbohydrate Research 346 (2011) 1916–1923

a

300

Table 1 Assignment of ions generated upon fragmentation of reduced permethylated FR-II Observed m/z 345.2 493.3 563.3 711.4 781.3

FR-I

RI (mV)

200 a

100

FR-III FR-IV

10

FR-V

15

+

[Lac-HexA+Na] [HexA-Hexol+Na]+ [Lac-HexA-HexA+Na]+ [HexA-HexA-Hexol+Na]+ [Lac-HexA-HexA-HexA+Na]+

C Y C Y C

According to the fragmentation proposed by Domon and Costello.25

containing 2 Hex, 6 HexA and 2 lactate substituents. MS2 of the ion at 1567.4 m/z and MS3 of the ion at 447.1 m/z resulted in fragmentation of the oligosaccharide (Fig. 3 and Table 2) leading to hypothesise the following structure:

FR-II

0

Ion typea

Fragment ions

20

25

time (hr) 60

b

FR-I.1

RI (mV)

40

The data collected suggested that FR-II is the repeating unit of SMAEPS and FR-I.4 corresponds to an oligosaccharide having the size of two repeating units.

FR-I.2 FR-I.3

FR-I.4

2.5. NMR studies of fraction II 20

0

6

8

10

12

14

time (hr)

711.4

Figure 1. Elution profile of SMA-EPS hydrolysate separated on a Bio Gel P2 column (a) and elution profile of FR-I separated on a Bio Gel P10 column (b). Fractions referred in the text are indicated.

100 HexA 218

Lac-HexA 304

20

781.3

563.3 579.3

40

493.3

60

345.2 Lac-HexA

Relative Intensity (%)

80

HexA 218

HexA 218

Hex-ol 234

0 400

600

800

* 1000

m/z Figure 2. Fragmentation (MS2) spectrum of ion at 1015.5 m/z corresponding to the [M+Na]+ adduct of reduced, permethylated FR-II. The fragmentation pattern is shown. * Indicates parent ion at 1015.5 m/z; dashed lines show fragmentation starting from non-reducing end; solid lines show fragmentation starting from reducing end.

FR-II was extensively studied by NMR spectroscopy. The 1H NMR spectrum showed five anomeric signals in the region 5.3– 4.5 ppm, which were named Aa, Ab, B, C, D, with Ab partially overlapping with B, as reported in Figure 4. From composition analysis and MS analysis of FR-II it was determined that glucose is at the reducing end and therefore, the signals Aa and Ab were assigned to a Glc residue. The B–D anomeric protons had the b configuration. The resonance at 4.53 ppm between H-1’s of C and D was determined to be a ring proton through an HSQC experiment. The doublet at 1.40 ppm was in agreement with the methyl group of the lactate substituent. Integration of the anomeric signals gave the following results: Aa = 0.36; Ab + B = 1.51; C = 0.84; D = 1.00, while integration of the methyl group gave a value of 2.10. Two low intensity signals at 2.15 and 2.14 ppm were attributed to methyl functions of O-acetyl groups, which resisted the acidic treatment. COSY and TOCSY experiments led to the complete attribution of the proton spin systems for residues B–D, while for Aa and Ab only protons 1 and 2 were assigned. Interpretation of the HSQC plot gave the assignment for the carbon atoms (Table 3). After assigning the carbon atoms engaged in glycosidic linkages for residues B–D, C-4 of residue A, known to be the site of glycosidic linkage from methylation analysis of SMA-EPS-R (Section 2.2), was found at 79.63 ppm, with the corresponding proton at 3.63 ppm. ROESY experiments showed the following inter-residues cross-peaks: CH of lactate to B3; B1 to D3; D1 to C4; C1 to A4. Moreover, the chemical shifts of residues B and D were in very good agreement with those of galacturonic acid,9 while those of residue C with the spin system of glucuronic acid.10 In conclusion, NMR data together with composition and MS analysis established that FR-II has the following structure: Lac-3-b-GalA-(1?3)-bGalA-(1?4)-b-GlcA-(1?4)-Glc. 2.6. NMR studies of de-O-acetylated exopolysaccharide The 1H NMR spectrum of the de-O-acetylated SMA-EPS (deAcSMA-EPS) (Fig. 4) showed three signals in the anomeric region at 4.90, 4.68 and 4.50 ppm; their integration values were 1.00, 2.22 and 2.04, respectively, indicating overlapping of resonances. In fact, the HSQC plot revealed that the protons resonating at 4.68 ppm

1919

P. Cescutti et al. / Carbohydrate Research 346 (2011) 1916–1923

b

150

300

450

961.3 HexA 176

400

600

889.3

1000

1385.3 HexA 176

1319.4

1143.3

967.3

805.2

800

m/z

HexA 176

1137.3

0

Hex 162

Hex-ol 182

Lac-HexA 248

609.2

20

713.2

288.9

*

0

40

HexA 176

Hex 162

623.1

HexA 176

40

HexA 176

60

537.1

60

447.1

Relative Intensity (%)

80

180.9

Relative Intensity (%)

80

20

a

100

785.2

271.0

100

Lac-HexA 248

* 1200

1400

1600

m/z

Figure 3. Fragmentation (MS2) spectrum (a) of ion at 1567.4 m/z corresponding to the [M+Na]+ adduct of reduced FR-I.4 and MS3 spectrum (b) of fragment ion at 447.1 m/z. The fragmentation pattern is shown. * Indicates fragmented ions; dashed lines show fragmentation starting from non-reducing end; solid lines show fragmentation starting from reducing end.

Table 2 Assignment of ions generated upon fragmentation of reduced FR-I.4 Observed m/z 271.0 288.9 447.1 537.1 609.2 623.1 713.2 785.2 805.2 889.3 961.3 967.3 1137.3 1143.3 1319.4 1385.3 a b

Ion typea

Fragment Ions +

[Lac-HexA+Na] [Lac-HexA+Na]+ [Lac-HexA-HexA+Na]+ [HexA-HexA-Hex–H2O+Na]+ Hex-[Lac-HexA]-HexA–H2O+Na]+ [Lac-HexA-HexA-HexA+Na]+ [HexA-HexA-Hex-HexA–H2O+Na]+ [Lac-HexA-HexA-HexA-Hex+Na]+ [Lac-HexA-HexA-HexA-Hex-ol+Na]+ [HexA-HexA-Hex-HexA-HexA–H2O+Na]+ [Lac-HexA-HexA-HexA-Hex-HexA+Na]+ [Hex-[Lac-HexA]-HexA-HexA-Hex-ol+Na]+ [Lac-HexA-HexA-HexA-Hex-HexA-HexA+Na]+ [HexA-Hex-[Lac-HexA]-HexA-HexA-Hex-ol+Na]+ [HexA-HexA-Hex-[Lac-HexA]-HexA-HexA-Hex-ol+Na]+ [Lac-HexA-HexA-HexA-Hex-[Lac-HexA]-HexA-HexA+Na]+

B C B DFb DFb B DFb B Y DFb B Y B Y Y B

According to the fragmentation proposed by Domon and Costello.25 DF: double fragmentation.

connected with one anomeric and one ring carbon atoms, while the protons at 4.50 ppm connected with only one anomeric carbon atom, therefore, suggesting overlapping of two anomeric resonances. This finding was confirmed by examining the COSY plot. Therefore, four anomeric signals were present in the 1H NMR spectrum and were named from A to D, with signals C and D overlapping (Fig. 4). Their chemical shift values indicated that they all have the b configuration (Figure 4). As for FR-II, the resonance at 1.40 ppm was assigned to the –CH3 group of lactate. COSY and TOCSY experiments led to the identification of all the protons for each spin system, while inspection of the HSQC plot (Fig. 5) determined all the carbon atoms for every sugar residue (Table 4). The chemical shifts were all in good agreement with what was already established for FR-II and FR-I.4 (see supplementary content), confirming the assignment of signals A to glucose, C to glucuronic acid, B and D to galacturonic acid. The data collected confirmed that the glycosidic linkage of the glucose residue engaged C-4 and that residue D is branched, connecting consecutive repeating units through its C-4. HMBC experiment identified the carbonyl signals of each uronic acid residue. NOESY experiments showed the following inter-residues contacts: –CH of lactate to B3; B1 to D3; D1 to C4; C1 to A4; A1 to D4, thus establishing the structure for deAc-SMA-EPS:

2.7. NMR studies of native exopolysaccharide NMR investigation of SMA-EPS established the acetyl substitution pattern, as well as confirmed the data found for the de-O-acetylated sample. The 1H NMR spectrum of the native polymer was deeply different from the one of deAc-SMA-EPS (Figure 4). Two new signals related with acetylation were present at 5.65 and 5.13 ppm and from the HSQC plot it was shown that they were connected to carbons resonating at 72.15 and 70.19 ppm, respectively. Therefore, they were ascribed to ring protons geminal to the attached O-acetyl groups.11,12 Only H-1 of the Glc residue slightly changed its chemical shift in the native polymer, moving to 4.88 ppm, while all the other anomeric signals showed more

1920

P. Cescutti et al. / Carbohydrate Research 346 (2011) 1916–1923 B

with integration values of 3.12 and 2.87, respectively, indicating the presence of two acetyl groups per repeating unit. As for the other samples, COSY and TOCSY experiments led to the unravelling of the proton spin systems, while the HSQC plot revealed the carbon atom spin systems (Table 5). From the NMR data the identity of each spin system was then established, as reported in Table 5. Inspection of the COSY plot (Fig. 6) undoubtedly assigned the sites of acetylation, revealing that one O-acetyl group was linked to C-2 of the branched galacturonic acid residue and the other to C-4 of the terminal non-reducing galacturonic acid. The complete structure of the native EPS is depicted in Formula 3.

FR-II

C D Aβ Aα

deAc-SMA-EPS

B C,D A

3. Conclusion SMA-EPS B-4 D-2 A

6.0

5.5

D

B

5.0

C

4.5

3.5

4.0

3.0 ppm

Figure 4. 1H NMR spectra of FR-II, deAc-SMA-EPS and SMA-EPS. Anomeric signals of different spin system were named A–D; Aa and Ab indicate signals at the reducing end.

evident shifts with respect to the non acetylated sample. Three other anomeric signals were located at 4.65, 4.55 and 4.45 ppm and their integration gave the values 0.94, 1.03, 0.70, after setting equal to 1.00 the integration of the peak belonging to H-1 of Glc. The chemical shift of the methyl group of lactate moved from 1.40 to 1.25 ppm, suggesting the vicinity of an acetyl ester. Its integration value was equal to 3.57. Two signals attributable to methyl groups of acetyl esters resonated at 2.13 and 2.10 ppm,

The primary structure of SMA-EPS is a novel one among the bacterial polysaccharides: it has three uronic acid residues on a total of four sugars in the repeating unit and bears an additional negative charge due to the D-lactate substituent. This polysaccharide is also highly acetylated, having two O-acetyl groups per repeating unit, one of which renders the GalA of the backbone completely substituted. The abundance of uronic acids explains the high resistance of the polysaccharide to acid hydrolysis.13 The O-acetyl groups were also resistant to base treatment and they could be removed using 0.2 M NaOH, instead of 0.01 M normally used for other polymers in our laboratory. The abundance of negative charges is a common feature of the EPS produced by two other microorganisms infecting CF patients: alginate produced by P. aeruginosa and the completely pyruvylated b-(1?3)-glucan and pyruvylated a-(1?2)-mannan produced by Inquilinus limosus,14 thus strongly suggesting that negative charges somehow are a characteristic which constitutes an advantage for the microbes in the lung environment. It is also remarkable that the two strains examined in the present work were isolated from two different CF patients leaving in different countries, France and Italy. Therefore, this EPS produced by the two different isolates might be the

Table 3 1 H and 13C chemical shifts assignment of FR-II Residue

Chemical shift (ppm)a

Nucleus 1 1

H C

13 1

H C

13 1

H C

13 1

H C

13 1

H C

13 1

Lac a b

H C

13

2

3

4

5

5.22 92.62

3.57 72.03

3.63 79.63b

4.66 96.62

3.28 74.68

3.63b 79.63b

4.66 104.42

3.68 70.32

3.56 81.23

4.34 68.14

4.05 76.21

4.56 102.96

3.41 73.64

3.70 75.23

3.76 81.63

3.92 76.19

4.49 103.17

3.72 70.77

3.88 83.07

4.53 70.23

4.10 76.05

4.17 76.43

1.40 19.62

Chemical shifts are given relative to internal acetone (2.225 ppm for 1H and 31.07 ppm for Only one cross peak was found, suggesting that C-4 and H-4 of Aa and Ab are identical.

6

b

13

C).

3.81, 3.96 61.0

1921

P. Cescutti et al. / Carbohydrate Research 346 (2011) 1916–1923

A6

B4 B2

D2

C2 A3

D5 B5 C5

A5 A2

C3

D4

-CH Lac

A4 D3

C4

B3

D1 A1

B1

C1

ppm Figure 5. Expansion of the HSQC plot of deAc-SMA-EPS.

Table 4 1 H and 13C chemical shifts assignment of de-O-acetylated SMA-EPS Residue

Chemical shift (ppm)a

Nucleus 1

2

3

4

5

6

4.90 103.35

3.33 74.48

3.68 75.30

3.55 80.36

3.55 75.36

3.72, 3.92 61.47

4.68 104.33

3.68 70.57

3.53 81.25

4.35 68.34

4.00 76.37

175.32

1

H 13 C

4.50 103.39

3.41 73.65

3.68 75.30

3.75 81.86

3.89 76.23

175.42

1

4.50 103.39

3.83 71.24

4.00 81.69

4.68 78.94

4.05 75.96

174.50

182.12

4.15 76.60

1.40 19.63

1

H C

13 1

H C

13

H 13 C 1

Lac

H 13 C

a

Chemical shifts are given relative to internal acetone (2.225 ppm for 1H and 31.07 ppm for 13C). Carbonyl signals were detected in a 13C-NMR spectrum recorded at 70 °C and assigned to each uronic acid residue through an HMBC experiment.

one typical of S. maltophilia species, as alginate is for P. aeruginosa, a hypothesis which can only be confirmed after examining the polysaccharides biosynthesised by a large number of isolates.

4. Experimental 4.1. Bacterial growth and EPS purification A plate containing solid YEM medium (20 g mannitol, 2 g yeast extract, 15 g agar per litre) was inoculated with a S. maltophilia bacterial stock kept at 80 °C. This solid culture was then used to directly inoculate agar plates containing YEM medium, without growing a liquid culture, because the latter caused the appearance of non-mucoid phenotypes. The plates were incubated at 30 °C for 4 days, the cells were collected with a 0.9% NaCl solution and after addition of sodium azide were gently stirred at 4 °C for about 2 h.

The viscous bacterial cells suspension was centrifuged at 28,000g for 30 min at 15 °C. The supernatant was precipitated in 4 volumes of cold isopropanol, dissolved in water, dialysed first against 0.1 M NaCl, and then against water. After neutralising and filtering the polysaccharide solution, the absence of proteins and nucleic acids was verified by UV spectroscopy. The polysaccharide was recovered by lyophilisation and stored at 4 °C. The polysaccharide was further purified by precipitation with cetyl trimethylammonium bromide following the method reported by Scott.15 4.2. General procedures Analytical GLC was performed on a Perkin–Elmer Autosystem XL gas chromatograph equipped with a flame ionisation detector and an SP2330 capillary column (Supelco, 30 m), using He as carrier gas. The following temperature programmes were used: for alditol acetates, 200–245 °C at 4 °C/min; for partially methylated

1922

P. Cescutti et al. / Carbohydrate Research 346 (2011) 1916–1923

Table 5 1 H and 13C chemical shifts assignment of native SMA-EPS Residue

Chemical shift (ppm)a

Nucleus 1

2

3

4

5

6

H 13 C

4.88 103.33

3.34 74.46

3.64 75.07

3.54 80.25

3.54 75.51

3.72, 3.90 61.45

1

H C

4.65 101.29

5.13 72.15

4.19 80.00

4.72 78.92

4.09 76.04

H C

4.55 104.82

3.46 70.33

3.60 80.03

5.65 70.19

4.10 74.84

H C

4.45 103.61

3.35 73.91

3.64 75.47

3.72 81.97

3.71 77.60

4.10 76.78

1.25 19.51

1

13 1

13 1

13 1

a b

H C

13

Lac

Chemical shifts are given relative to internal acetone (2.225 ppm for 1H and 31.07 ppm for assignments can be interchanged.

13

b

b

C).

4.3. Production and purification of oligosaccharides

B2/3

B1

B1/2

D1 D3/4

D2/3

D2 D1/2

SMA-EPS was treated with 2 M TFA at 100 °C for 2 h, the sample was then evaporated to dryness several times and subsequently neutralised. The products were separated by gel permeation chromatography on a Bio Gel P2 column (1.6 cm i.d.  90 cm) using 0.05 M NaNO3 as eluent, and a flow rate of 6 mL/h. Fractions were collected at 20 min intervals and those belonging to the peak named FR-II ( Fig. 1a) were pooled together, dialysed using a 1000 Da cut-off membrane and lyophilised. The products which eluted in the void volume of the column (FR-I) were subsequently separated on a Bio Gel P10 column (1.6 cm i.d.  90 cm) using the same eluent and experimental settings, except for the collection time interval which was 15 min. Selected fractions belonging to the peak named FR-I.4 ( Fig. 1b) were pooled together, dialysed using a 1000 Da cut-off membrane and lyophilised. Elution was monitored using a refractive index detector (WGE Dr. Bures, LabService Analitica), which was connected to a paper recorder and interfaced with a computer via PicoLog software. 4.4. Reduction of the carboxylic groups in the SMA-EPS

B4 B4/5

B3/4

ppm Figure 6. Expansion of the COSY plot of SMA-EPS. Proton–proton correlations of residues bearing acetyl substituents are indicated.

alditol acetates, 150–250 °C at 4 °C/min. Separation of the trimethylsilylated (+)-2-butyl glycosides was obtained on a HP1 column (Hewlett–Packard, 50 m), using the following temperature programme: 135–240 °C at 1 °C/min. GLC–MS analyses were carried out on an Agilent Technologies 7890A gas chromatograph coupled to an Agilent Technologies 5975C VL MSD. Hydrolysis of the samples were performed in the following conditions; native EPS: 2 M TFA for 2 h at 125 °C; carboxyl reduced EPS: 2 M TFA for 1 h at 125 °C. Methanolysis was conducted for 16 h at 85 °C with 2 M HCl in methanol for the native EPS, and with 1 M HCl in methanol for the carboxyl reduced EPS. Alditol acetates were prepared as already described,16 permethylation of the carboxyl reduced EPS was achieved following the protocol by Harris,17 while oligosaccharides were permethylated using the protocol by Dell.18 Determination of the absolute configuration of the sugar residues was performed as described.19 De-O-acetylation was achieved treating a sample of SMA-EPS with 0.2 M NaOH at room temperature for 5 h.

A sample of SMA-EPS was treated with 0.5 M TFA at 100 °C for 1 h to reduce its molecular mass; 47 mg was dissolved in 22 mL of water and treated with 430 mg of carbodiimide, as previously described.20 The solution was neutralised by adding 50% aqueous acetic acid, and dialysed (cut-off 1000 Da) and 30 mg was recovered by lyophilisation (SMA-EPS-R). This sample was used to determine the monosaccharide composition, the glycosidic linkages and the absolute configuration of the sugar residues, before and after removal of the hydroxyisopropanol substituent, derived from reduction of lactate. To perform the de-eterification reaction according to Hough and Theobald,21 a sample of SMA-EPS-R was permethylated17 to solubilise it in dichloromethane, the solvent used for the reaction. 4.5. Determination of the lactate absolute configuration Determination of the lactate absolute configuration was achieved resorting to the specificity of enzymes, which required cleavage of the ether linkage without inversion of lactate configuration. Taking advantage of the lactate substitution in position 3 of the sugar ring, it was possible to perform a b-elimination reaction22 on the isolated substituted monosaccharide. Isolation of the lactate-substituted monosaccharide was achieved through Smith degradation23,24 of the native polysaccharide. SMA-EPS was subjected to complete oxidation with NaIO4, by dissolving 150 mg of polymer in 150 mL of 9.5 mM NaIO4 and incubating at

P. Cescutti et al. / Carbohydrate Research 346 (2011) 1916–1923

4 °C for 7 days. The reaction was stopped by addition of glycerol and the products were reduced with NaBH4. Addition of 50% CH3COOH after 16 h destroyed the excess of reducing reagent, the sample was dialysed and the solution volume subsequently reduced to 50 mL. TFA was added to have a final concentration of 2 M and the hydrolysis was conducted at 100 °C 18 h. After cooling, the sample was taken to dryness under reduced pressure. It was dissolved in water, its pH was adjusted to neutrality, and the sample was recovered by lyophilisation. It was separated on a Bio Gel P2 column (1.6 cm i.d.  90 cm), equilibrated and eluted with water. The flow rate was 6 mL/h and fractions were collected at 20 min intervals. Elution was monitored using a refractive index detector (WGE Dr. Bures, LabService Analitica), which was connected to a paper recorder and interfaced with a computer via picolog software. Following the separation trend, selected fractions were analysed by 1H NMR and ESI-MS (data not shown): the results evidenced a lactate-substituted disaccharide, a lactate-substituted monosaccharide and a mixture of low molecular mass products from the Smith degradation. Although the lactate-substituted monosaccharide was not expected to be a product of the Smith degradation, it was formed from the lactate-substituted disaccharide by hydrolysis, because the conditions used were quite strong. Fractions containing the lactate-substituted monosaccharide were pooled together, separated again on the same Bio Gel P2 column using the same experimental conditions, and 13 mg of sample were recovered by lyophilisation. Treatment of the lactatesubstituted monosaccharide (1 mg) with alkali,22 in particular 0.08 M phosphate buffer pH 12 at 37 °C for 18 h, resulted in b-elimination of the substituent and formation of an unsaturated monosaccharide. The solution was neutralised with HCl and the completeness of the reaction was checked resorting to TLC on silica gel on aluminium foil (Fluka) using CH3OH–CHCl3–H2O (20:20:7) as eluent and a 3% w/v solution of permangate in H2SO4 as development reagent. The reaction mixture was directly used as a source of substrate for the enzymatic tests: two different enzymatic kits were used, one containing L-lactate dehydrogenase (E.C.1.1.1.27) (Boehringer Mannheim) and the other D-lactate dehydrogenease (E.C.1.1.1.28) (Megazyme). The oxidation of lactate to pyruvate was monitored by measuring the production of the reduced coenzyme NADH, following the increase in absorbance at 340 nm. The reactions were set up following the instructions of the enzymatic kits. 4.6. NMR spectroscopy studies The molecular mass of EPS solutions (1 g/L) was decreased by treatment with a Branson sonifier equipped with a microtip at 2.8 Å. Samples were cooled in an ice bath and sonicated using 5 bursts of 1 min each, separated by 1 min intervals. They were subsequently exchanged three times with 99.9% D2O by lyophilisation and finally dissolved in 0.7 mL 99.96% D2O. For recording the experiments on FR-II and FR-I.4 2.6 mg of each sample were used, while 10 mg of native and de-acetylated EPS samples were used. Spectra were acquired at 25 °C for FR-II and at 70 °C for all the other samples on a VARIAN spectrometer operating at 500 MHz (1H). 2D experiments were performed using standard VARIAN pulse sequences and pulsed field gradients for coherence selection when appropriate. 13C NMR spectrum was acquired on deAc-SMA-EPS at 70 °C. HSQC spectra were recorded using 140 Hz one bond Jxh constant, while for HMBC spectrum 8 Hz multiple bond Jxh was

1923

applied. For each sample TOCSY spectra were acquired using three spin-lock times: 30, 80 and 140 ms. ROESY spectra were acquired with 200, 300 and 400 ms mixing time and 2.0 s relaxation time. NOESY experiments were recorded with 100 ms mixing time and 1.5 s relaxation time. Chemical shifts are expressed in ppm using acetone as internal reference (1H at d = 2.225 ppm, 13C at d = 31.07). NMR spectra were analysed using Mestre Nova software. 4.7. ESI MS analysis ESI mass spectra were recorded on a Bruker Esquire 4000 ion trap mass spectrometer connected to a syringe pump for the injection of the samples. The instrument was calibrated using a tune mixture provided by Bruker. Samples were dissolved in 50% aqueous methanol–11 mM NH4OAc at an appropriate concentration and injected at 180 lL/h. Detection was always performed in the positive ion mode. Acknowledgements The authors thank the Italian Cystic Fibrosis Foundation (Project #11/2006 with the contribution of CF delegation of Belluno, Italy), the Italian Ministry of University and Research (PRIN 2007) and the FVG regional project R3A2 within L.R.26/2005 for financial support. References 1. de Vrankrijker, A. M.; Wolfs, T. F.; van der Ent, C. K. Paediatr. Respir. Rev. 2010, 11, 246–254. 2. Chen, C. Y.; Tsay, W.; Tang, J. L.; Tien, H. F.; Chen, Y. C.; Chang, S. C.; Hsueh, P. R. Epidemiol. Infect. 2010, 138, 1044–1051. 3. Welsh, M. J.; Tsui, L.-C.; Boat, T. F.; Beaudet, A. L. Cystic Fibrosis. In The Metabolic and Molecular Basis of Inherited Disease; Scriver, C. R., Beaudet, A. L., Sly, W. S., Valle, D., Eds.; McGraw-Hill: New York, 1995; pp 3799–3876. 4. Saiman, L.; Siegel, J. Clin. Microbiol. Rev. 2004, 17, 57–71. 5. Yoon, S. S.; Hassett, D. J. Expert Rev. Anti-Infect. Ther. 2004, 2, 611–623. 6. Waters, V.; Yau, Y.; Prasad, S.; Lu, A.; Atenafu, E.; Crandall, I.; Tom, S.; Tullis, E.; Ratjen, F. Am. J. Respir. Crit. Care Med. 2010. doi:10.1164/ rccm.201009-1392OC. 7. Pompilio, A.; Crocetta, V.; Confalone, P.; Nicoletti, M.; Petrucca, A.; Guarnieri, S.; Fiscarelli, E.; Savini, V.; Piccolomini, R.; Di Bonaventura, G. B. M. C. Microbiology 2010, 10, 102–117. 8. Huang, T.; Somers, E. B.; Lee Wong, A. C. J. Bacteriol. 2006, 188, 3116–3120. 9. Arbatsky, N. P.; Shashkov, A. S.; Literacka, E.; Widmalm, G.; Kaca, W.; Knirel, Y. A. Carbohydr. Res. 2000, 323, 81–86. 10. Cescutti, P.; Toffanin, R.; Pollesello, P.; Sutherland, I. W. Carbohydr. Res. 1999, 315, 159–168. 11. Haverkamp, J.; van Halbeek, H.; Dorland, L.; Vliegenthatr, J. F. G. Eur. J. Biochem. 1982, 122, 305–311. 12. Cescutti, P.; Ravenscroft, N.; Ng, S.; Lam, Z.; Dutton, G. G. S. Carbohydr. Res. 1993, 244, 325–340. 13. Dutton, G. G. S. Adv. Carbohydr. Chem. Biochem. 1973, 28, 11–160. 14. Herasimenka, Y.; Cescutti, P.; Impallomeni, G.; Rizzo, R. Carbohydr. Res. 2007, 342, 2404–2415. 15. Scott, J. E. Methods Carbohydr. Chem. 1965, 5, 38–44. 16. Albersheim, P.; Nevins, D. J.; English, P. D.; Karr, A. Carbohydr. Res. 1967, 5, 340– 345. 17. Harris, P. J.; Henry, R. J.; Blakeney, A. B.; Stone, B. A. Carbohydr. Res. 1984, 127, 59–73. 18. Dell, A. Methods Enzymol. 1990, 193, 647–660. 19. Gerwig, G. J.; Kamerling, J. P.; Vliegenthart, J. F. G. Carbohydr. Res. 1978, 62, 349–357. 20. Taylor, R. L.; Conrad, H. E. Biochemistry 1972, 11, 1383–1388. 21. Hough, L.; Theobald, R. S. Methods Carbohydr. Chem. 1963, 2, 203–206. 22. Tipper, D. J. Biochemistry 1968, 7, 1441–1449. 23. Hay, G. W.; Lewis, B. A.; Smith, F. Methods Carbohydr. Chem. 1965, 5, 357–361. 24. Goldstein, I. J.; Hay, G. W.; Lewis, B. A.; Smith, F. Methods Carbohydr. Chem. 1965, 5, 361–370. 25. Domon, B.; Costello, C. E. Glycoconjugate J. 1988, 5, 397–409.