Acetylation of the O-specific polysaccharide isolated from Shigella flexneri serotype 2a

Abstract

Shigella flexneri causes diarrheal diseases especially in infants and children in developing countries. Modifications of lipopolysaccharide (LPS) molecule, like bacteriophage-mediated glucosylation and acetylation of the O-specific chain (O-SP), are important for the LPS antigenicity and consequently for the immunogenicity of the polysaccharide-based vaccines against shigellosis. Here we report the degree of O-acetylation and the localisation of O-acetyl groups and side-chain glucose substitution in the O-SP (scheme) in different preparations of S. flexneri type 2a LPS.

Surface polysaccharides of pathogenic bacteria, including capsular polysaccharides (CPS) or the O-specific polysaccharide (O-SP) of lipopolysaccharides (LPS), serve both as essential virulence factors and as protective antigens. Covalent binding of CPS or of O-SP to proteins both increases their immunogenicity and confers T-cell dependence to these saccharides, making them suitable vaccines for infants and children ^1,2. Shigella flexneri causes dysentery mostly in developing countries with more fatalities then any other Shigella species ³. The disease can be prevented by vaccination using polysaccharide part of the LPS as an immunogen ⁴. S. flexneri is divided into 13 serotypes, which with the exception of serotype 6 share identical linear backbone. This basic structure is known as serotype Y and the additions of α-glucose and O-acetates at various positions define serological identity of the particular strains ^5,6.

S. flexneri serotype 2a is the most prevalent strain among those causing endemic shigellosis ⁷. The structure of its O-specific polysaccharide was determined and O-acetylation has been noticed ⁵, however the position of the O-acetyl groups was not analysed. The presence and the position of O-acetyl groups in the polysaccharide chain may influence the immune response ⁸. The importance of the O-acetylation for polysaccharide immunogenicity was shown for Neisseria meningitides, group B streptococci, Staphylococcus aureus, Salmonella typhi or Cryptococcus neoformans polysaccharides ^9-14. Another modification of Shigella LPS is a bacteriophage-encoded side-chain glucosylation, which was proposed to promote bacterial invasion of host ells ¹⁵. The degree of glucosylation in S. flexneri 2a is an important characteristic of its O-SP. Here we present the localization of O-acetyl groups in the O-SP, and quantitation of the glucosylation of the O-SP in different preparations of the polymer.

S. flexneri 2a LPS was isolated using phenol-water extraction. It was found that LPS distributes between phenol and water phases (phenol-LPS and water-LPS); most of the LPS was recovered from phenol phase. SDS PAGE analysis of these products (Fig. 1) showed that they have similar degree of polymerization.

Figure 1.

Silver stained SDS-PAGE separation of the LPS from S. flexneri 2a. 1-marker, 2, 5- phenol-LPS at 4 and 1 μg per well, respectively; 3, 6 water-LPS, isolate from 1999 at 4 and 1 μg per well, respectively; 4, 7 - water-LPS, isolate from 2006 at 4 and 1 μg per well, respectively; 8- water-LPS after O-deacetylation at 4 μg per well.

O-Specific polysaccharides were prepared from both phenol-LPS and water-LPS (phenol-PS and water-PS) by acetic acid hydrolysis. Several preparations of the polysaccharides, isolated at various times from different strains of the serotype 2a at NIH and NRC were analyzed. They had similar NMR spectra (Fig. 2) differing by the relative signal intensities. Water-PSs contained more core fragments than the phenol-PS as demonstrated by stronger signals of the core monosaccharides, particularly well visible signals at 5.53 and 5.90 ppm, belonging to H-1 of α-Glc b and α-Gal c from outer core fragment:

Figure 2.

Integration of the ¹H NMR spectrum of the highly acetylated water-PS sample with ∼60% acetylation at A3 (top) and of the least acetylated water-PS sample with ∼30% acetylation at A3 (bottom). Core signals are of high intensity in both samples.

A set of the NMR spectra - COSY, TOCSY, NOESY, HSQC, and HMBC - were recorded for the water and phenol extracted polysaccharides and all major signals were assigned to the following structure (Fig. 3, Table 1):

where O-acetyl groups are non-stoichiometric.

Figure 3.

Fragment of COSY (green), TOCSY (red), and NOESY (black) spectra of the S. flexneri 2a polysaccharide. Residues marked with letters A-E belong to the repeating units without O-acetate at A3; residues marked with letters A′-D′ (residue E is weakly sensitive and not specially marked) belong to the repeating units with O-acetate at A3. Residue D″ is O-acetylated at O-6 and belong to the repeating unit without O-acetate at A3; its variant from the repeating unit with O-acetate at A3 is not visible because of low intensity. Some minor signals of the unidentified origin are present.

Table 1.

NMR data for native and O-deacylated polysaccharide (δ, ppm). N-Ac at D3 with no O-Ac: 2.06/23.8; N-Ac at D3 with O-Ac on A3: 2.11/23.8; OAc at A3: 2.16/21.8; OAc at D6: 2.21/22.0 ppm (H/C).

Residue	Nucleus	1	2	3	4	5	6
α-Rha A	H	5.13	4.14	3.87	3.32	3.74	1.29
	C	102.7	80.3	71.4	74.0	70.9	18.1
α-Rha A 3- O-Ac	H	5.17	4.25	5.07	3.51	3.84	1.29
	C	102.7	78.2	74.1	71.4	70.9	18.1
α-Rha B	H	5.02	4.08	3.87	3.47	3.81	1.29
	C	102.8	80.8	71.4	73.8	70.8	18.1
α-Rha B, Ac on A3	H	5.07	4.13	3.87	3.47	3.81	1.29
	C	102.8	80.8	71.4	73.8	70.8	18.1
α-Rha C	H	4.84	3.94	3.93	3.78	4.15	1.34
	C	102.4	72.2	80.6	76.5	70.6	19.3
α-Rha C, Ac on A3	H	4.88	3.94	3.93	3.78	4.15	1.34
	C	102.0	72.2	80.6	76.5	70.6	19.3
β-GlcNAc D	H	4.72	3.83	3.64	3.54	3.44	3.76/3.90
	C	103.7	57.2	82.9	69.9	77.4	62.4
β-GlcNAc D, Ac on A3	H	4.53	3.83	3.64	3.62	3.44	3.76/3.90
	C	102.7	57.2	82.9	69.7	77.4	62.4
β-GlcNAc D 6-O-Ac	H	4.75	3.86	3.64	3.62	3.64	4.32/4.41
	C	4.53	3.83	82.9	69.7	74.7	64.8
α-Glc E	H	5.18	3.54	3.71	3.42	3.94	3.79/3.87
	C	99.1	73.0	74.3	71.4	73.4	62.3

Relative configurations of the constituent monosaccharides were identified on the basis of vicinal proton coupling constants and ¹³C NMR chemical shifts. They were in agreement with the standard values for each monosaccharide. Anomeric configurations were deduced from the J_1,2 coupling constants and chemical shifts of H-1, C-1, and C-5 signals, as well as by observation of intraresidual NOE connectivities (H-1/H-3, H-1/H-5) characteristic for the β pyranosides.

Several sets of signals were found for all monosaccharide residues due to partial acetylation. Side-chain Glc E signals were fuzzy and therefore not assigned to acetylated or non-acetylated repeating units. Position of the O-acetyl groups (O-Ac) was determined on the basis of strong downfield shift of the proton attached to a carbon atom bearing O-Ac group. Two such signals were found, H-3 of the Rha A (shifted from 3.87 ppm in non-acetylated variant to 5.07 ppm in 3-O-acetylated one) and H-6/H-6′ of the GlcNAc D (shifted from 3.76/3.90 ppm in non-acetylated variant to 4.32/4.41 ppm in 6-O-acetylated one). Partial acetylation of the O-3 of Rha A affected strongly positions of H-1 signals of all monosaccharides of the repeating unit, while the presence of the O-acetate at O-6 of GlcNAc D influenced significantly only the signals of this residue itself.

The degree of acetylation of Rha A at A3 was relatively easy to determine because the signals of H-1 of the residue Rha C were separated due to this acetylation and were not overlapped with other signals. Integration of these signals gave 30 to 50 % acetylation of the A3 hydroxyl group in various preparations of the O-SP (Fig. 2).

For the elucidation of the degree of acetylation of O-6 of GlcNAc D integral intensity of D-6b of the 6-O-acetylated residue D was compared with the sum of integral intensity of both H-1 signals of the Rha residue C. This gave 30 to 60 % of acetylation in different preparations.

The potential relation of acetylation occurring at two places within the repeating unit, in example that two acetates are simultaneously present or absent, was studied. Tracing NOE connectivities between the monosaccharides showed that acetylation is random; all combinations of acetylated and non-acetylated residues were found.

For the analysis of the amount of glucosylation the spectra of O-deacetylated polysaccharide from both water and phenol phase were studied. No signals were found that can belong to non-glucosylated repeating units. Integral intensities of the anomeric signals were close to 1:1:1:1:1 (Fig. 4). Thus side-chain glucose is present in all repeating units.

Figure 4.

¹H spectrum of the O-deacylated polysaccharide from phenol phase. Integrals are normalized for Glc E H-1 signal. Signal intensity shows 100% presence of this monosaccharide. Core signals integral is 1:25 to repeating unit.

Removal of O-acetyls with ammonium hydroxide resulted in the partial loss of antigenicity, as judged by the intensity of precipitin line in immunodiffusion assay performed with the serum raised against the whole killed bacteria and compared to the native O-acetylated LPS. Chemical O-deacetylation however removes also ester bound fatty acids from Lipid A thus changing the molecule size and conformation and therefore more detailed study on the O-acetyls as a part of epitope of the S. dysenteriae 2a LPS are planned as the next step. Currently several synthetic vaccine candidate are under development¹⁶ and also different preparation of bacterial O-SP in conjugate vaccines are being tested in clinical trials ^2,4, therefore it is important to characterize these O-SP preparations in respect to the presence and localization of the O-acetyl and glucose moieties for the consistency of the final vaccine formulation.

1. Experimental

1.1. Growth of bacteria and isolation of LPS

Shigella flexneri type 2a strain 2457T10¹⁷ was grown in ultrafiltered Triptic Soy Broth (Difco Laboratories) with 5 g of glucose and 5 mM magnesium sulphate per liter, for 20 h at 20 °C with stirring and aeration; the pH was maintained at ∼7.5 by addition of ammonium hydroxide. The identity of bacteria was confirmed by culture, Gram staining and agglutination with typing antisera. LPS was extracted by hot phenol method¹⁸ and after dialysis recovered from each phase.

1.2. Mild hydrolysis of the LPS

The LPSs (20-80 mg) were treated with 2 % acetic acid at 100 °C for 3 h, precipitate removed by centrifugation, soluble products separated by gel chromatography on Sephadex G-50 column to give polysaccharide and core fractions. Polysaccharide was O-deacylated by heating (60 °C) in 12 % ammonia for 3 h.

1.3. NMR spectroscopy

NMR spectra were recorded at 35 °C in D₂O on a Varian UNITY INOVA 500, instrument, using acetone as reference for proton (2.225 ppm) and carbon (31.5 ppm) spectra. Varian standard programs COSY, NOESY (mixing time of 400ms), TOCSY (spinlock time 120 ms), HSQC, and gHMBC (long-range transfer delay 100 ms) were used.

1.4. Serologic methods

Immunodiffusion was performed in 1 % agarose in PBS against hyperimmune sera obtained by multiple injections of whole killed bacterial cells as described ¹⁵. Removal of acetyls for serological studies was perform with 5 % ammonium hydroxide, 15 h, 23 °C with stirring and the O-deacylated LPS or PS was desalted on Sephadex G-50 column (1 × 50 cm) eluted with water.

Table 2.

The degree of O-acetylation in different preparations of S. flexneri type 2a O-specific polysaccharides (O-SP).

Preparation and date	Acetylation at A3 (%)	Acetylation at A6 (%)
Strain 2457T10; 1992; water LPS	50	50
Strain 2457T10; 1995; water LPS	50	50
Strain 2457T10; 1997; water LPS	35	40
Strain 2457T10; 1999; water LPS	30	40
Strain 2457T10; 2005; water LPS	50	60
Strain 2457T10; 2005; phenol LPS	30	30