Glycobiology Advance Access originally published online on August 23, 2007
Glycobiology 2007 17(11):1150-1155; doi:10.1093/glycob/cwm089
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Celebrating the golden anniversary of the discovery of bacillosamine, the diamino sugar of a Bacillus*,,
2 Department of Biological Chemistry, The Weizmann Institute of Science, Rehovot 76100, Israel
1 To whom correspondence should be addressed: Tel: +972-8-934-3605; Fax: +972-8-9468256; e-mail: nathan.sharon{at}weizmann.ac.il
Received on August 1, 2007; revised on August 13, 2007; accepted on August 13, 2007
Abstract
Bacillosamine (2,4-diamino-2,4,6-trideoxy-D-glucose, Bac), a rare amino sugar, was discovered 50 years ago as a result of the follow-up of a chance observation made during studies of polypeptide synthesis by a Bacillus subtilis strain later renamed Bacillus licheniformis. In the following decades this amino sugar was almost completely ignored, although it was found in a number of bacterial polysaccharides and other metabolites. Recently, there has been a burst of interest in Bac when it was found to be a link glycan in eubacterial glycoproteins. In this retrospective, I review the chance discovery of Bac, its structural determination and its biosynthesis.
Key words: Bacillus licheniformis / Campylobacter jejuni / fucosamine / glycoproteins / polysaccharide
Bacillosamine (Bac), in the form of its di-N-acetyl derivative, has been identified in the carbohydrate–peptide linking group of N- and O-glycoproteins of a number of gram-negative eubacterial pathogens. The glycans attached to these linking groups were distinct from those present in any previously known glycoproteins. Thus, in Campylobacter jejuni many glycoproteins were found to carry the asparagine-linked heptasaccharide GalNAc
4-GalNAc4-(Glcβ3-)GalNAc4-GalNAc4-GalNAc3-BacAc2 (Young et al. 2002
; Liu et al. 2006). In Neisseria gonorrhoeae and Neisseria meningitidis a Bac-related diamino trideoxyhexose DATDH (the structure of which has yet to be fully identified) is a constituent of the trisaccharide Galβ4-Gal3-DATDH that is O-linked via serine to flagellar glycoproteins (Stimson et al. 1995; Aas et al. 2007). The discovery of these glycoproteins was among the major factors that have led to the overthrow of the dogmatic belief that eubacteria are not capable of synthesizing glycoproteins. Quite significantly N-glycosylation in C. jejuni is required for the immunogenicity of its multiple proteins, for its full capacity to attach and to invade in vitro cultured eukaryotic cells and to colonize the intestines of chickens and mice (Karlyshev et al. 2004; Vijayakumar et al. 2006).
Studies of the biosynthesis of the eubacterial glycoproteins revealed several unique features, some with wide ranging implications. Both the N-linked heptasaccharide of C. jejuni and the O-linked trisaccharide of N. gonorrhoeae are assembled by sequential transfer of their monosaccharide constituents from the corresponding sugar nucleotides to undecaprenyl-pyrophosphate prior to their en bloc transfer to protein (Aas et al. 2007). The case of the lipid-linked trisaccharide is most unusual, since it is destined to be O-linked. The oligosaccharyltransferase that functions in N-glycosylation of the eubacterial glycoproteins is not only simpler in structure than the eukaryotic enzyme, but also possesses a much broader specificity and is capable of attaching foreign N-glycans to a variety of proteins (Szymanski and Wren 2005;Weerapana and Imperiali 2006). Of major importance is the demonstration that the C. jejuni transferase was active when expressed in Escherichia coli (Wacker et al. 2002; Feldman et al. 2005), opening new possibilities for the production of human glycoproteins in bacteria and thereby forming the basis of a new era of glycoengineering.
In the present review, the chain of events that led to the first isolation of Bac is described, as well as how its structure was established by degradation and chemical synthesis. Also presented is the scheme of the biosynthesis of di-N-acetylbacillosamine that was postulated in the early 1960s and has recently been validated.
Stumbling on a Bacillus polysaccharide
As a new appointee to the Department of Biophysics of the fledgling Weizmann Institute, I was studying the formation of the simple polypeptide, poly-
-glutamic acid in a Bacillus subtilis strain later renamed Bacillus licheniformis. To learn more about protein biosynthesis, I traveled in 1956 to the Massachusetts General Hospital (MGH) and Harvard Medical School, Boston, to work as a fellow for a year with Fritz Lipmann, Nobel laureate for physiology and medicine, who was making seminal contributions in the area. Before leaving for Boston, I noted that the trichloroacetic acid-extracts of the B. subtilis strain prepared in search for precursors of the biosynthesis of poly--glutamic acid were highly viscous. Addition of an excess of alcohol to the extracts precipitated a material that was nondialyzable and gave color reactions of a polysaccharide. As no polysaccharides of B. subtilis were known at the time, I undertook its characterization. The acid hydrolyzate of the polysaccharide contained 60% of reducing sugars together with small quantities of amino acids, as well as close to 20% acetyl groups. Using the simple paper chromatographic and colorimetric methods for sugar analysis then available led to the conclusion that the polysaccharide was composed of galactose, N-acetylglucosamine and N-acetylgalactosamine in a molar ratio of 2:1:1 (Sharon 1957).
Isolation and preliminary characterization of a diamino sugar
On arrival at the MGH in the summer of 1956, I found out that close to Lipmann's Biochemical Research Laboratory there was a Carbohydrate Research Laboratory headed by Roger Jeanloz, well known for his structural studies of glycosaminoglycans. It was Jeanloz who encouraged me to submit the paper reporting on the B. subtilis polysaccharide to Nature, in spite of the fact that its analysis was far from complete (Sharon 1957). He also invited me to join his laboratory for a second year of postdoctoral studies, starting in the fall of 1957, with the aim of thoroughly characterizing the polysaccharide by chemical means. To determine the optical configuration of the hexosamine constituents, an acid hydrolyzate of several hundred milligrams of the polysaccharide was fractionated by ion exchange chromatography (Gardell 1953). Elution with 0.3 M HCl afforded two distinct peaks of glucosamine and galactosamine, respectively, which when crystallized were found to be of the D-configuration. That the galactose of the polysaccharide is also of D-configuration was established in separate experiments.
Regeneration of the ion exchange column called for washing with 2 M HCl. Rather than discarding the wash it was collected, the solvent evaporated and the residue was analyzed by paper chromatography. Surprisingly, the chromatogram revealed the presence in the residue of an amino sugar that was faster moving than any compound of this type known at the time (Figure 1).
|
A large quantity of crude polysaccharide (7.8 g) was prepared for further studies, and its acid hydrolyzate was fractionated by ion exchange chromatography as above, except that elution with 0.3 M HCl was continued until a third hexosamine peak emerged from the column (Figure 2). Eventually, 120 mg of crude material was obtained from this peak, which was purified by preparative paper chromatography and then crystallized and recrystallized to give 71 mg of pure material, designated Compound A.
|
Elemental analysis of Compound A gave the most unexpec- ted result, since it revealed the presence of two nitrogen atoms in each molecule. One of these could be ascribed to its 2-amino group, based on the Elson–Morgan reaction (Rondle and Morgan 1955)
and the IR spectrum of the compound that contained bands characteristic of a primary amine as well as of an amide. When compound A was N-acetylated the primary amine band disappeared, and only amide bands were seen (Compound B). Further chemical analysis showed that Compound A contained two C-methyls, one of which could be ascribed to an acetamide group, the other probably located at C-6 (identification of C-methyl groups can now be done easily by NMR). Compound A showed a mutarotation in solution, requiring the presence of a free hydroxyl at C-4 or C-5. One argument for the location of the acetamide group at position 4 of this compound was that the product of its N-acetylation (Compound B) failed to give a positive Morgan–Elson reaction; this reaction is characteristic of 2-acetamido sugars, but is inhibited by substitution at the C-4 position.
Treatment of Compound A with strong acid released in low yield Compound C that moved much more slowly on paper chromatography, consistent with its being more polar than the parent compound, as a result of the liberation of a primary amine by cleavage of the amide linkage. Degradation of Compound A by ninhydrin (Stoffyn and Jeanloz 1954) afforded the reducing Compound D which had a very high mobility on paper chromatograms, faster moving than pentoses (ribose or deoxyribose), methylpentoses (e.g., rhamnose) or methyltetroses, indicating its highly hydrophobic nature.
The above considerations indicated a structure of a 4-acetamido-2-amino-2,4,6-deoxyhexose for Compound A, and the derived structures for its N-acetyl derivatives (Compound B), as well as the diamino sugar obtained by strong acid treatment of Compound A (Compound C) and its ninhydrin degradation product (Compound D), all shown in Figure 3 (Sharon and Jeanloz 1959, 1960).
|
Enter bacillosamine
Establishment of the full structure of the diamino sugar by degradation and chemical synthesis was achieved in collaboration with several coworkers, most of them graduate students, and took over a decade. Compound A was thus shown to be 4-acetamido-2-amino-2,4,6-trideoxy-D-glucose, and it was named 4-N-acetylbacillosamine (Zehavi and Sharon 1973).
Oxidation of Compound A by a mixture of periodate and permanganate, followed by acid hydrolysis of the resultant product, yielded D-allothreonine (Figure 4). The D-allothreonine was identified by amino acid analysis, by paper chromatography, and by its susceptibility to oxidation with D-amino acid oxidase. This established that the diamino sugar belongs to the D series and that the substituents on C-4 and C-5 are cis. To identify the configuration at C-2, Compound A was reduced and dinitrophenylated to yield a 4-acetamido-2-(2,4-dinitrophenyl)amino-2,4,6-trideoxy-hexitol. Acid hydrolysis of the hexitol to unmask the 4-amino group followed by periodate-permanganate oxidation afforded N-(2,4-dinitrophenyl)-L-serine identified by paper chromatography and by optical rotatory dispersion. The amino group at C-2 is thus on the right in the Fischer projection of the molecule. Examination of the NMR spectrum of the fully acetylated derivative of Compound A revealed that it is in the C-1 conformation; that H-2, H-3, H-4, and H-5 are in trans diaxial positions; and that the acetoxy group at C-3 is equatorial. The above finding made it certain that Compound A belongs to the D-gluco series with a structure 4-acetamido-2-amino-2,4,6-trideoxy-D-glucose (4-N-acetylbacillosamine), as shown in Figure 5. The same figure shows 2,4-di-N-acetylbacillosamine, obtained in crystalline form, and fully characterized as well. We were not able, however, to obtain the parent compound, Bac, in good yield by acid treatment of 4-N-acetylbacillosamine, nor in any other way.
|
|
Final proof for the structure of Bac was achieved by chemical synthesis. Starting with glucosamine, 12 steps were required for its conversion into di-N-acetylbacillosamine (Liav et al. 1974
). Three other syntheses of this compound were subsequently reported (Bundle and Josephson 1980; Amin et al. 2006; Bedini et al. 2006). A by-product of our synthesis was N-acetyl-D-fucosamine, another rare amino sugar isolated earlier from polysaccharides of various Bacillus species (Sharon et al. 1964). In the course of these studies other 2,4-diamino-2,4,6-trideoxyhexoses, among them 2,4-diacetamido-2,4,6-trideoxy-D-galactose, were synthesized (Liav and Sharon 1973; Liav et al. 1976). The latter diamino sugar has been found in bacterial polysaccharides (Kenne et al. 1980; Lindberg et al. 1980) and in antibiotic complexes of bacterial origin (Tsuno et al. 1981).
As mentioned earlier, Bac was isolated from the acid hydrolyzate of the B. subtilis polysaccharide in the form of the 4-N-acetyl derivative (Sharon and Jeanloz 1960). It is also the form in which the diamino sugar occurs in the intact polysaccharide, as revealed by a high resolution NMR analysis at 800 MHz of the purified material and its derivatives, carried out in collaboration with Paul Messner and colleagues from the Agricultural University of Vienna, and Tali Reshef from the Weizmann Institute (Schaffer et al. 2001). The polysaccharide consists of a backbone of the repeating trisaccharide [3-GlcNAc3-Bac4NAcβ3-GalNAcβ–]. At the C-6 of both the N-acetylglucosamine and N-acetylgalactosamine residues are attached in a β-linkage galactose residues that are substituted at C-3 and C-4 by a pyruvoyl residue. Cyclic acetals of pyruvic acid of this or similar types are common in bacterial polysaccharides, but the repeating unit of the Bacillus polysaccharide is unique. It is noteworthy that the ratio of galactose, N-acetylglucosamine and N-acetylgalactosamine eventually found in the polysaccharide is 2:1:1, exactly as suggested in the first report on this substance (Sharon 1957).
Biosynthesis of di-N-acetylbacillosamine
With the isolation of D-fucosamine, some of its similarities with Bac became obvious, both being 2-amino-6-deoxy hexoses of the D-configuration produced by Bacillus species. A scheme was, therefore, proposed for the biosynthesis of these two compounds based on what was known at the time about the pathways of the formation of 6-deoxy and 4-amino-6-deoxy hexoses (Sharon et al. 1964) (Figure 6). In these pathways, the key intermediate is a nucleoside diphospho-6-deoxy-4-ketohexose, for example GDP-6-deoxy-4-keto-hexose that is formed from GDP-mannose in the biosynthesis of L-fucose. In the proposed scheme both N-acetylfucosamine and diacetylbacillosamine are formed from XDP-N-acetylglucosamine (X being a base such as uridine or guanosine) via the common key intermediate XDP-2-acetamido-6-deoxy-4-ketoglucose. Stereospecific reduction the above intermediate would afford XDP-N-acetylfucosamine, while reductive amination (also stereospecific) followed by N-acetylation, would give XDP-di-N-acetylbacillosamine. However, repeated attempts to demonstrate the formation of Bac from UDP-N-acetylglucosamine in extracts of B. subtilis were unsuccessful. Negative results were also obtained with GDP- and TDP-N-acetylglucosamine, synthesized specifically for that purpose (Harel et al. 1966). Thus, by the middle 1970s I left Bac completely, to concentrate all my efforts on the study of lectins (Sharon 2007).
|
Happily, the validity of the postulated scheme was essentially proven during the last 3 years by the molecular genetic identification and biochemical characterization in C. jejuni of the three enzymes that participate in the biosynthesis of UDP-di-N-acetylbacillosamine from UDP-N-acetyglucosamine. These are the UDP-N-acetyglucosamine C-6 dehydratase (Creuzenet 2004
), C-4 aminotransferase (Schoenhofen et al. 2006; Vijayakumar et al. 2006) and N-acetyltransferase (Olivier et al. 2007). Still missing is evidence for the enzyme required for the formation of UDP-N-acetyl-D-fucosamine from the key intermediate UDP-2-acetamido-6-deoxy-4-ketoglucose.
Since the Bac-containing glycoproteins of certain bacterial pathogens are required for the virulence of the bacteria (Vijayakumar et al. 2006) and this amino sugar is absent from higher animals, including humans, its biosynthesis may serve as a novel target for antibacterial drug action.
Conflict of interest statement
None declared.
Footnotes
* Dedicated to Roger W Jeanloz, mentor and friend, and glycobiology pioneer, in whose laboratory and under whose guidance I discovered bacillosamine, on the occasion of his 90th birthday, November 3, 2007. 
The editor is sad to convey that the dedicatee of this historical review, Dr. Roger Jeanloz, died September 2007, prior to its final publication. This marks the passing of a valued colleague who made significant contributions to the emergence of the discipline of Glycobiology.
Abbreviations
Bac, Bacillosamine; MGH, Massachusetts General Hospital
References
Aas FE, Vik A, Vedde J, Koomey M, Egge-Jacobsen W. Neisseria gonorrhoeae O-linked pilin glycosylation: Functional analyses define both the biosynthetic pathway and glycan structure. Mol Microbiol (2007) 65:607–624.[CrossRef][Web of Science][Medline]
Amin MN, Ishivata A, Ito Y. Synthesis of asparagine-linked bacillosamine. Carbohydr Res (2006) 341:1922–1929.[CrossRef][Web of Science][Medline]
Bedini E, Esposito D, Parrilli M. A versatile strategy for the synthesis of N-acetylbacillosamine-containing disaccharide building blocks related to bacterial O-antigens. SYNLETT (2006) 6:825–830.
Bundle DR, Josephson S. Chlorosulfation of amino-sugars—Synthesis of methyl 2-acetamido-4-amino-2,4,6-trideoxy-beta-D-glucopyranoside (bacillosamine) from a D-glucosamine derivative. Can J Chem (1980) 58:2679–2685.
Creuzenet C. Characterization of CJ1293, a new UDP-GlcNAc C6 dehydratase from Campylobacter jejuni. FEBS Lett (2004) 559:136–140.[CrossRef][Web of Science][Medline]
Feldman MF, Wacker M, Hernandez M, Hitchen PG, Marolda CL, et al. Engineering N-linked protein glycosylation with diverse O antigen lipopolysaccharide structures in Escherichia coli. Proc Natl Acad Sci USA (2005) 102:3016–3021.
Gardell S. Separation on Dowex 50 ion exchange resin of glucosamine and galactosamine and their quantitative determination. Acta Chimica Scand (1953) 7:207–215.
Harel E, Silbert JE, Sharon N. Chemical synthesis and characterization of thymidine diphosphate N-acetyl--D-glucosamine and guanosine diphosphate N-acetyl--D-glucosamine. Arch Biochem Biophys (1966) 117:296–300.[CrossRef][Web of Science]
Karlyshev AV, Everest P, Linton D, Cawthraw S, Newell DG, Wren BW. The Campylobacter jejuni general glycosylation system is important for attachment to human epithelial cells and in the colonization of chicks. Microbiology (2004) 150:1957–1964.
Kenne L, Lindberg B, Petersson K, Katzenellenbogen E, Ronamowska E. Structural studies of the O-specific side-chains of the Shigella sonnei phase I lipopolysaccharide. Carbohydr Res (1980) 78:119–126.[CrossRef][Web of Science]
Lindberg B, Lindqvist B, Lonngren J, Powell DA. Structural studies of the capsular polysaccharide from Streptococcus pneumoniae Type I. Carbohydr Res (1980) 78:111–117.[CrossRef][Web of Science][Medline]
Liav A, Hildesheim J, Zehavi U, Sharon N. Synthesis of 2-acetamido-2,6-dideoxy-D-glucose (N-acetyl-D-quinovosamine), 2-acetamido-2,6-dideoxy-D-galactose (N-acetyl-D-fucosamine), and 2,4-diacetamido-2,4,6-trideoxy-D-glucose from 2-acetamido-2-deoxy-D-glucose. Carbohydr Res (1974) 33:217–227.[CrossRef][Web of Science]
Liav A, Jacobson I, Sheinblatt M, Sharon N. Synthesis of 2,4-diacetamido-2,4,6-trideoxy-D-galactose. Carbohydr Res (1976) 66:95–101.
Liav A, Sharon N. Synthesis of 2,4-diacetamido-2,4,6-trideoxy-L-altrose, -L-idose, and -L-talose from benzyl 6-deoxy-3,4-O-isopropylidene-beta-L-galactopyranoside. Carbohydr Res (1973) 30:109–126.[CrossRef][Web of Science][Medline]
Liu X, McNally DJ, Nothaft H, Szymanski CM, Brisson JR, Li J. Mass spectrometry-based glycomics strategy for exploring N-linked glycosylation in eukaryotes and bacteria. Anal Chem (2006) 78:6081–6087.[Medline]
Olivier N, Chen M, Behr JR, Imperiali B. In vitro biosynthesis of UDP-N,N'-diacetylbacillosamine by enzymes of the Campylobacter jejuni general protein glycosylation pathway. Biochemistry (2007) 45:13659–13669.[CrossRef]
Rondle CJ, Morgan WT. The determination of glucosamine and galactosamine. Biochem J (1955) 61:586–589.[Web of Science][Medline]
Schäffer C, Scherf T, Christian R, Kosma P, Zayni S, Messner P, Sharon N. Purification and structure elucidation of the N-acetylbacillosamine-containing polysaccharide from Bacillus licheniformis ATCC 9945. Eur J Biochem (2001) 268:857–864.[Web of Science][Medline]
Schoenhofen IC, McNally DJ, Vinogradov E, Whitfield D, Young NM, et al. Functional characterization of dehydratase/aminotransferase pairs from Helicobacter and Campylobacter: Enzymes distinguishing the pseudaminic acid and bacillosamine biosynthetic pathways. J Biol Chem (2006) 281:723–732.
Sharon N. Polysaccharide containing amino sugar from Bacillus subtilis. Nature (1957) 179:919.[Medline]
Sharon N. Recollections. Lectins: Carbohydrate-specific reagents and biological recognition molecules. J Biol Chem (2007) 282:275–2764.
Sharon N, Jeanloz RW. The isolation of an aminohexose from Bacillus subtilis. Biochim. Biochim Acta (1959) 31:277–278.
Sharon N, Jeanloz RW. The diaminohexose component of a polysaccharide isolated from Bacillus subtilis. J Biol Chem (1960) 235:1–5.
Sharon N, Shiff I, Zehavi U. The isolation of D-fucosamine (2-amino-2,6-dideoxy-D-galactose) from polysaccharides of Bacillus spp. Biochem J (1964) 93:210–214.[Web of Science][Medline]
Stimson E, Virji M, Makepeace K, Dell A, Morris HR, et al. Meningococcal pilin: A glycoprotein substituted with digalactosyl 2,4-diacetamio-2,4,6-trideoxyhexose. Mol Microbiol (1995) 17:1201–1204.[CrossRef][Web of Science][Medline]
Stoffyn PJ, Jeanloz RW. Identification of amino sugars by paper chromatography. Arch Biochem (1954) 52:373–379.[CrossRef][Medline]
Szymanski CM, Wren BW. Protein glycosylation in bacterial mucosal pathogens. Nature Revs Microbiol (2005) 3:225–236.[CrossRef]
Tsuno T, Konishi K, Naito T, Kawaguchi H. Glysperin, a new antibiotic complex of bacterial origin. II. Structure of glysperins A, B, and C. J Antibiotics (Jpn) (1981) 20:390–398.
Vijayakumar S, Merkx-Jacques A, Ratnayake DB, Gryski I, et al. Cj1121c, a novel UDP-4-keto-6-deoxy-GlcNAc C-4 aminotransferase essential for protein glycosylation and virulence in Campylobacter jejuni. J Biol Chem (2006) 281:22733–22743.
Wacker M, Linton D, Hitchen PG, Nita-Lazar M, Haslam SM, et al. N-linked glycosylation in Campylobacter jejuni and its functional transfer into E. coli. Science (2002) 298:1790–1793.
Weerapana E, Imperiali B. Asparagine-linked protein glycosylation: from eukaryotic to prokaryotic systems. Glycobiology (2006) 16:91R–101R.
Young NM, Brisson JR, Kelly J, Watson DC, et al. Structure of the N-linked glycan present on multiple glycoproteins in the gram-negative bacterium, Campylobacter jejuni. J Bio Chem (2002) 277:42530–42539.
Zehavi U, Sharon N. Structural studies of 4-acetamido-2-amino-2,4,6,-trideoxy-D-glucose (N-acetylbacillosamine), the N-acetyl diamino sugar of Bacillus licheniformis. J Biol Chem (1973) 248:433–438.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||