Recent outbreaks of Meningitis and chicken flu in honk Kong, as well as the growth of diseases such as AIDs, are constant reminders of the need for new ways to combat infectious diseases, say Frank Reichel & Geert-Jan Boons.

Vaccination is one of the most powerful approaches for protecting humankind against infectious diseases. The word "vaccine" is derived from the Latin vacca, which means cow, and can be traced back to Jenner's 1783 discovery1 that inoculating humans with the cowpox virus protected them against smallpox. The cowpox virus is much less virulent than the human equivalent, but the viruses are sufficiently similar in that the immune response elicited by the cowpox virus is also directed against the human pox pathogen. It took another century until scientists firmly established that inoculation with attenuated or inactivated microorganisms can introduce a protective immunological response. Meanwhile, the current widespread application of whole-cell vaccines has resulted in the control of many diseases, such as Diphteria, Polio, Measles, rabies, Tetanus and Whooping Cough.
However, the use of whole-cell vaccines is not useful or safe in all cases. (1,2,3) Killed or attenuated cell vaccines may contain too little of the immunising antigen, and may also introduce an array of non-immunising agents - some of which may be hazardous. Sub-unit vaccines which only contains the immunising agents - for example a protein or polysaccharide - offer an attractive alternative to whole-cell vaccines. In the past, scientists manufactured protein-based sub-unit vaccines using biological procedures to grow the pathogen and physical or chemical techniques to isolate the desired fragments. Nowadays we can over-express proteins in bacteria or yeast and conveniently obtain them in large quantities.(1)
Bacterial capsular polysaccharides (CPS) are another well-established group of sub-unit vaccines and several are currently licensed for use in preventing pneumococcal pneumonia and meningitis caused by Haemophilus influenzae type b, Neisseria meningitidis and Strptococcus pneumoniae.(4) Purified CPS-based vaccines are non-toxic and we can chemically and physically define their identity and homogeneity. In addition, carbohydrate structures are generally highly conserved in microorganisms, which guarantees that we can apply a particular polysaccharide vaccine over a long period of time. On the other hand, proteins of pathogenic microorganisms may undergo mutations resulting in a resistance against antibodies raised by a vaccine. For example, it is possible to detect several mutants of a membrane protein of the human immunodeficiency virus (HIV) in infected but healthy patients.(5)

Non-cooperation

Until recently, a number of problems restricted researchers from further developing polysaccharides as vaccines. The most important of these is the inability of polysaccharides to induce immunological memory(4,6) and their poor immunogenicity in children younger than two years. These problems are related to the so-called T-cell independence of pure polysaccharides. For most immunogens, including polysaccharides, antibody production is based on the cooperative interactions of two

Table 1. Possible Conjugate Vaccines for Pneumonia, Influenza and Meningitis

BSA: Bovine serum albumin; CRM 197 : nontoxic mutant diptheria toxin; DTd: diphtheria toxoid; OMP: outer membrane protein; TTd: tetanus toxoid.

types of lymphocytes (white blood cells), B-cells and T-cells.(2) Pure polysaccharides cannot activate T-cells and therefore have a restricted immunogenicity.

However, the T-cell independence of polysaccharides can be effectively overcome by covalent coupling to proteins. Researchers have shown that the resulting polysaccharide-protein conjugates can produce high levels of polysaccharide-specific antibodies in infants and give a booster reaction after re-exposure. In 1987, the first conjugate vaccine was approved for preventing invasive diseases caused by Haemophilus influenzae type b in children aged between 18 months and 6 years.(4) Researchers are also investigating other conjugate vaccines that target Streptococcus pneumoniae and Neisseria meningitidis (Table 1)

Make or Break

An important step in designing a conjugate vaccine is developing a chemical method to link a saccharide to a protein covalently. The major problem is minimising structural changes within the specific immunological binding sites (epitopes) of both the saccharide and protein. Moreover, desired coupling procedures should avoid cleavages of acid-, base- or oxidation-sensitive glycosidic linkages and side-chain functional groups within the protein. Furthermore, saccharide loading needs to be controlled: too little saccharide will not give an effective immunological reaction, while too much can mask protein T-cell epitopes. A widely used

conjugation method is the reductive amination of a reducing end of a saccharide chain with a free amino functionality of a lysine side chain (Scheme 1). The reducing end of a cyclic saccharide (1) is in equilibrium with the open chain aldehyde configuration (2). Reacting a primary amine of a protein (3) with (2) produces an imine (Schiff base) (4), which can be reduced with cyanoborohydride to yield a stable conjugate (5).(4)

Usually Researchers isolate bacterial polysaccharides from a biological source. However, recent developments in chemically synthesising oligosaccharides makes it possible to prepare in relatively large quantities pure and well-defined oligosaccharides containing the minimal structural requirements for immunogenity.(7) It is also possible to equip a synthetic oligosaccharide with an artificial linker moiety that has a unique reactivity and allows selective coupling with a carrier protein. This allows us to overcome problems such as low recovery of pure material and the loss of vital carbohydrate fragments during the coupling of natural polysaccharides to proteins. A synthetic approach is particularly valuable when a saccharide is relatively small or when the natural oligosaccharide contains undesirable fragments. At the University of Birmingham, we have recently proposed a synthetic approach to develop a lipopolysaccharide-based vaccine to protect against Neisseria meningitidis - one of the several bacteria responsible for bacteria meningitis. Lipopolysaccharides (LPS) are important constituents of the outer membrane of Gram-negative bacteria, and are composed of a carbohydrate (saccharide) and a lipid moiety. Our work follows on from earlier studies by Harold J. Jennings and co-workers at the National Research Council of Canada, who demonstrated that conjugates of the carbohydrate moiety of meningococcal LPS elicit a relevant antibody response.(8) This moiety is composed of an inner-core region that contains the unsual higher carbon sugar L-glycero-D-manno-heptose (LD-Hepp, (6)) and 3-deoxy-D-manno-octulosonic acid (KDO); this innercore region is responsible for stimulating important antibodies. Further, it contains an undesirable lacto-N-neotetraose carbohydrate, which is structurally very similar to saccharides that are attached to human glycoproteins and may induce an autoimmune reaction.

By Organic synthesis, we were able to make the relevant inner-core epitopes without the lacto-N-neotetraose saccharide moiety.(9) This in turn has enabled us to synthesise a range of LD-Hepp sugars including disaccharides such as (7) and (8), and trisaccharides (9), containing an aminopropyl spacer for selective coupling with the carrier protein tetanus toxoid (Fig.1). Before proceeding to describe the results of our work with these "vaccines", it is important to note that not all N-meningitidis bacteria against which they are to be directed are exactly the same; different strains produce different types of saccharides. These different types of carbohydrates have different immunological properties, ie antibody elicited against one type may not recognise another type, resulting in different LPS immunotypes L1-L9.

Immune Responses

All synthetic saccharide-tetanus toxoid conjugates elicited in rabits an immune response against L2 immunotypes, but only the branched trisaccharide tetanus toxoid conjugate was able to elicit an immune response against L1 and L3, 7, 9 LPS. Thus, the branched saccharide structure is tought to be the minimal structure required for L1 and L3, 7, 9-specific immune response and to be a part of the cross-reactive epitope of these LPS immunotypes.(10)

The most popular carrier proteins are tetanus toxoid and diphteria toxoid, whic are also constituents of the existing DPT (diphteria, polio, tetanus) vaccine. However, frequently applying these proteins may result in immuno-tolerance. Furthermore, T-epitopes - particular peptide sequences of 12-16 amino acids that activate T-cells - on a carrier protein may differ from T-epitopes on outer membranes proteins of Neisseria meningitidis. Thus, the heterologous T-epitopes of a carrier protein may not be effectively activated during infection.
To solve the latter problem, we have developed(11) a new generation of vaccines that are fully synthetic and composed of a peptide-carbohydrate-lipopeptide conjugate

(fig.2 (10)).We synthesise such compounds by a highly efficient route combining solid- and solution-phase methodologies. The LD-Hepp saccharide part of (10) (blue), against which anti-bodies will be raised, is derived from the inner-core region of meningococcal LPS. The peptide sequence (red) is part of an outer membrane of Neisseria meningitidis and has been identified as a potent T-epitope;(12) this component induces immunological memory. The third bulding block, the lipopeptide N-palmitoyl-S-[2,3-bis(palmitoyloxy)-(2RS)- propyl]-(R)- cystein (Pam3Cys) (green) possesses13 the built-in ability of self-assembly. This leads to the formation of aggregates (eg Liposomes) large enough to be recognised by the immune system.(10) These aggregates are efficiently absorbed onto the surface of cells known as macrophages from where they are subsequently presented to T-cells.(14) At Birmingham, we are investigating the self-assembly abilities of Pam3Cys and related compounds,(15) and have already shown that the chirality of these compounds as well as the conjugation of peptides and carbohydrate are important factors for self-assembly into different superstructures such as vesicles, stacked bilayers and tubes.

Anti-cancer jabs

Several recent studies indicate that a vaccination strategy may be useful in treating cancer, and carbohydrates also play important roles in this new class of vaccines.(14,16) Many cell-surface proteins and lipids are glycosylated and these carbohydrate attachments are important for manycrucial biological processes.The saccharide moiety of a glycoprotein may be important for protein folding and stabilising the three-dimensional srtucture of a protein, it can mask protein epitopes on the peptide backbone, or it can act as a ligand for other proteins. Cancer cells exhibit incomplete and unusual glycosylations patterns, which lead to saccharide

structures (Fig.3) that are absent in normal cells.(14,16)

These tumour-related saccharides are one of the most consistent changes of particular cancer cells and form the basis for research into the immunotherapy of cancer. Scientists hope that an immune response elicited against tumour-related antigens will selectively destroy cancer cells that exhibitthese saccharide structures or at least inhibit metastases. Clinical experiments using vaccine preparations based on whole tumour cells walls have met with very little success and one possible explanation is that the vaccines are not sufficiently immunogenic. However, conjugates of synthetic tumour-related antigens such as Tn (11), Thomson-Friedenreich (12) and sialyl-Tn (sTn) (13), with an antigenic carrier protein, such as keyhole limpet haemocyanin (KLH), are much more immunogenic and have been successfully applied in active immunotherapy of tumour-bearing hosts.

In clinical trials carried out by B. Michael Longenecker at the University of Alberta, Canada, patients with metastatic breast cancer underwent immunotherapy with KLH-sTn vaccine. They received low pre-treatment doses of the immunomodulator cyclophosphamide, which enhances the poor immune

response of cancer patients. Patients given the vaccine developed an anti-carbohydrate immunoresponse and had significantly higher survival rates than patients who did not receive treatment. Furthermore, levels of sTn antibodies showed an inverse correlation with growth in measurable tumours.(14)
Only small amounts of tumour-related antigens can be isolated from natural sources and they are therefore very difficult to conjugate with a carrier protein without inducing undesirable structural changes. Fortunately, organic synthesis has provided substantial quantities of important oligosaccharides or glycopeptides in pure form. Researchers have described several synthetic routes for synthesising cancer-related antigens;16 Scheme 2 shows one example(17) for synthesising the sTn antigen. Glycosylation of the N-acetyl neuraminyl donor (14) with acceptor (15) proceeds with high regioselectivity and we can isolate compound (16) in a high yield exclusively as the a-anomer. The anomeric selectivity arises from neighbouring participating group of the thiophenyl moiety. We can remove the thiophenyl group in a later stage in the synthesis by radical mediated reduction to produce compound (19) or (20). The properly protected sTn disaccharide (16) is a precursor to be used in solid-phase glycopeptide synthesis.
Synthetic Scope
In summary then, saccharide-protein conjugate vaccines, prepared by mild conjugation methods, offer many advantages over classical polysaccharide vaccines. A carbohydrate epitope can be obtained by synthetic organic procedures and preliminary studies demonstrate that the next generation of vaccines may be fully synthetic and composed of a carbohydrate, peptide and lipid. This "minimal" approach towards vaccination offers the prospect of future tailormade vaccines.

Synthetic Scope

In summary then, saccharide-protein conjugate vaccines, prepared by mild conjugation methods, offer many advantages over classical polysaccharide vaccines. A carbohydrate epitope can be obtained by synthetic organic procedures and preliminary studies demonstrate that the next generation of vaccines may be fully synthetic and composed of a carbohydrate, peptide and lipid. This "minimal" approach towards vaccination offers the prospect of future tailormade vaccines.

References

1. M. Mackett and J.D. Williamson, Human vaccines and vaccination, Oxford: Bios Scientific, 1995 and F. Brown et al, Vaccine design. Chichester: Wiley, 1993
2. S.J. Cruz, Immunotherapy and vaccines, p.7. Weinheim:VCH, 1997
3. C.A. Mims, The pathogenesis of infectious disease. London: Academic, 1976; and R.A. Lerner, Sci. Am., 1983, 248 (2), 48.
4. H.J. Jennings and R.K. Snood, in Neoglycoconjugates, preparation and application, p.325, Y.C. Lee and R.T. Lee (eds). San Diego, Academic, 1994.
5. S. Wain-Hobson, Nature (London), 1995, 373, 102; and I. Najéra et al, J. Virol., 1995, 69,23.
6. D.A.A Ala'Aldeen and K.A.K Cartwright, J. Infect., 1996, 33, 153.
7. G.J. Boons, Contemp. Org. Synth., 1996, 173.
8. H.J. Jennings et al, Infect. Immunol., 1984, 43, 407.
9. G.J.P.H. Boons et al, Recl. Trav. Chim., Pays-Bas, 1989, 108, 339; G.J.P.H. Boons et al, Tetrahedron Lett., 1989, 30, 229; G.J.P.H. Boons et al, Angew. Chem. Int. Ed. Engl., 1989, 28, 1504; G.J.P.H. Boons et al, J. Carbohydr. Chem., 1991, 10, 995.
10. G.J.P.H. Boons, PhD thesis. Leiden: University of Leiden, 1991; and A.F.M Verheul et al, Infect. Immunol., 1991, 59, 3556.
11. F. Reichel et al, Chem. Comm, 1997, 2087.
12. E.J.H.J. Wiertz et al, J. Exp. Med., 1992, 176, 79.
13. K.H. Wiesmüller et al, Int. J. Pept. Protein Res., 1992, 40, 255.
14. R.R. Koganty et al, DDT, 1996, 1, 190
15. F. Reichel et al, submitted for publication; F. Reichel et al, in preparation.
16. T. Toyokuni and A.K. Singhal, Chem. Soc. Rev., 1995, 231.
17. Y. Nakahara et al, Carbohydr. Res., 1991, 216, 211.