1 Introduction

The development of vaccines is a great success in the history of medicine, being important in both prevention and eradication of infectious disease in humans. The principle of vaccination was successfully introduced in the late eighteenth century when Edward Jenner used cowpox to immunise humans against smallpox (Riedel 2005). A global vaccination effort, spearheaded by the WHO, led to the eradication of smallpox in 1980 (Fenner 1993). Since then, effective vaccines against diphtheria, tetanus, pertussis, measles, polio, and hepatitis A and B have been developed, which form part of vaccination schedules worldwide (WHO 2015). Although the majority of vaccines bind to protein epitopes within a pathogen, attention is now being turned to targeting glycan structures coating the surface of several major human pathogens. These glycans play roles in structural stability, shielding the pathogen surface from immunodetection, and are important for attachment to host receptors to gain entry into cells.

In 1923, Heidelberger and Avery showed that the sugars coating different Streptococcus pneumonia strains were not only diverse in structure but were also recognised by the human immune system and could be used to elicit a long-lasting antibody-mediated protective immune response (Heidelberger and Avery 1923; Heidelberger et al. 1950). Since then, the FDA has licensed carbohydrate-based vaccines in humans against the bacteria Haemophilus influenzae type b, S. pneumonia and Neisseria meningitidis. These vaccines contain pathogen-specific carbohydrates coupled to a carrier protein that induces a glycan-binding antibody response specific for the carbohydrate structures on the pathogen’s surface. Although there are currently no licensed carbohydrate-based vaccines against viruses, several antibodies directed against viral glycan structures have been isolated. In particular, glycan-binding broadly neutralizing antibodies (bnAbs) against human immunodeficiency virus 1 (HIV-1) have been shown to protect against SHIV (simian/human immunodeficiency virus chimera) infection in macaque models and highlight the potential for targeting viral glycans for vaccine design (Moldt et al. 2012; Shingai et al. 2014). This chapter will discuss the challenges and prospects for development of carbohydrate-based vaccines against major human pathogens and summarise current research in this area.

2 Glycans on Pathogens as Targets for Vaccine Development

Glycosylation is a post-translational modification, occurring on both proteins (N- and O-linked glycosylation) and lipids. Many pathogens display carbohydrate structures on their surfaces, and depending on how a pathogen acquires its surface carbohydrates, the host will see these carbohydrates as ‘self’ or ‘non-self’. Pathogens that use their own glycosylation machinery tend to display non-self glycan structures rendering them an attractive epitope for vaccine design. Pathogens that hijack the host cell glycosylation machinery mainly display self-glycans, which poses challenges for the development of vaccines (see Sect. 2.5). However, there are examples where glycosylation can diverge from the host cell leading to self-glycans being displayed in an arrangement that is seen as non-self by the host, e.g. the mannose patch on HIV Envelope. Here, we briefly describe the carbohydrate structures coating the major human pathogens; viruses, bacteria, fungi and parasites (Fig. 1).

Fig. 1
figure 1

Representative glycan structures present on bacteria, viruses, parasites and fungi

Full size image

2.1 Viruses

Viruses are intracellular parasites that use the host cell to replicate and produce new viral particles to sustain and spread infection. They hijack the host-cell secretion and glycosylation machinery to produce glycoproteins bearing both N- and O-linked glycans and are therefore mainly recognised as self by the host. N-linked glycosylation occurs at Asparagine (Asn) residues within the consensus sequence Asn-X-Threonine (Thr)/Serine (Ser) (where X can be any amino acid except Proline (Pro)). It follows a strict ordered assembly beginning in the ER with addition of Glc3Man9GlcNAc2 to Asn within the consensus motif (Fig. 2). This glycan is then trimmed to Man5GlcNAc2 by the ER- and Golgi-resident mannosidase enzymes. Diversification to complex-type glycans begins with addition of a β1,2-linked GlcNAc residue to Man5GlcNAc2 in the medial Golgi apparatus. Further trimming and processing in the medial and late Golgi apparatus lead to a wide array of hybrid-type and complex-type glycans and these structures are often dependent on the producer cell type. As this process is not genetically determined, viral glycoproteins tend to consist of a number of glycoforms displaying different glycan structures. Interestingly, HIV-1 displays host-derived high-mannose glycans in clusters. As high-mannose glycans are generally intermediates in the glycosylation pathway and not typically found in mammalian cells, they are recognised as non-self by the immune system (Bonomelli et al. 2011; Doores et al. 2010b) and this cluster forms the epitope of a number of HIV broadly neutralizing antibodies (Figs. 1 and 3) (Walker et al. 2011; Sok et al. 2014; Sanders et al. 2002a; Scanlan et al. 2002). O-linked glycosylation occurs on Ser and Thr residues but as there is no precise consensus sequence, it is harder to predict sites of O-linked glycosylation. Although most O-linked glycosylation occurs in Ser/Thr-rich regions, isolated residues can also be glycosylated. The initial glycan used for attachment to Ser/Thr is predominantly GalNAc and this monosaccharide is then modified by host-cell glycosyltransferases. Viral surface glycoproteins are typically important for cell tropism and infectivity and often hide the antigenic protein from the immune recognition (Vigerust and Shepherd 2007).

Fig. 2
figure 2

Schematic of the N-linked complex glycan biosynthesis pathway. Glycans are assembled as a Glc3Man9GlcNAc2 precursor on the membrane anchor dolichol phosphate on the outside of the ER. The oligosaccharide is flipped inside the ER lumen and transferred onto asparagine residues in the sequon Asn-X-Ser/Thr of the nascent protein chain. Once bound to the protein, glucose is removed and the protein is passed into the Golgi apparatus where the oligosaccharides get further trimmed to Man5GlcNAc2. Further modifications include the addition of GlcNAc and Gal and/or sialic acid and core fucose to form different variants of complex-type glycans

Full size image
Fig. 3
figure 3

(Reproduced and adapted from Behrens et al. Cell Reports 2016)

Location of N-linked glycans on HIV-1 Env important for glycan-binding HIV-1 bnAbs. Model shows location of glycans and the processing state. Red arrows indicate bnAbs targeting highlighted glycan sites

Full size image

2.2 Bacteria

Bacteria, both gram-positive and gram-negative, are covered by cross-linked polymer peptidoglycans, which form the major structural component of the periplasm. The peptidoglycan consists of a polysaccharide (composed of N-acetylglucosamine and N-acetylmuramic acid (MurNAc) in β1-4-linkage) that is covalently cross-linked by short peptides (Brown et al. 2015; Royet and Dziarski 2007). In addition, bacteria produce a polysaccharide capsule that forms the outer surface and mediates adhesion and has a role in virulence (Moxon and Kroll 1990). Antibodies against these polysaccharides are responsible for protection. The structure of the polysaccharide capsule is a unique feature between species and specific serotypes and is determined by the bacterial genome. For example, 90 different polysaccharides have been described for S. pneumoniae (Henrichsen 1995). In general, bacterial polysaccharides are made up of a combination of hexuronic acids, N-aceylated hexosamines and neutral or amino sugars such as rhamnose. However, some bacteria have polysaccharides containing self-structures, e.g. meningococcus Group B polysaccharide contains sialic acid (Rubens et al. 1987; Finne et al. 1983; Wyle et al. 1972).

2.3 Fungi

The cell wall of fungi consists of complex polysaccharides that are important for maintaining shape and integrity. Fungi possess their own secretory pathway and synthesise their own carbohydrates and glycoconjugates. The fungal wall is highly cross-linked and composed of polymers of galactose, glucose, N-acetylglucosamine, mannose, and/or rhamnose, and is specific for the fungal species. The mannose residues in mannans are usually α1–6 linked, whereas the glucose residues in glucans are mostly β1-3 linked (Cummings and Doering 2009). The fungal cell wall also contains glycosylated and/or glycosylphosphatidylinositol (GPI)-anchored proteins and some unusual glycolipids. These sugars are virulence determinants and are highly antigenic.

2.4 Parasites

Parasites pass through multiple developmental life stages during their life cycles often involving different hosts that can impact glycosylation. The glycoconjugates that cover their surface are comprised partly of glycan types typically absent in mammals such as schistosomal Lewis X (LeX, Galβ1-4(Fucalpha1-3)GlcNAc), LacDiNac (LDN, GalNAcβ1-4GlcNAc) or fucosylated LDN (LDN-F, GalNAcβ1-4(Fucα1-3)GlcNAc) (Fig. 1) and are rich in GPI-anchored proteins, polysaccharides and polysaccharide-binding lectins. These glycoconjugates make up the glycocalyx that is crucial for parasitic virulence (Rodrigues et al. 2015).

2.5 Challenges for the Development of Carbohydrate-Based Vaccines

A number of challenges associated with the generation of antibodies against carbohydrates have meant that the potential of carbohydrate-based vaccines has not yet been fully realised (Scanlan et al. 2007). Firstly, due to the T-cell-independent nature of anti-carbohydrate immune responses, antibody responses against carbohydrates tend to be less robust, short-lived and consist mostly of IgM. Secondly, glycan–protein interactions are inherently weak and therefore binding affinities must be enhanced through multivalent binding and avidity effects. For example, lectins overcome this weak binding by using multiple carbohydrate-binding domains to interact with an array of carbohydrate ligands. Some anti-HIV glycan-binding antibodies bind multiple glycans within one antigen-binding site (Crispin and Doores 2015; Sok et al. 2014; Walker et al. 2011; Sanders et al. 2002a; Scanlan et al. 2002). Thirdly, glycoproteins typically exist as a number of different glycoforms. As protein glycosylation is not directly template driven like protein synthesis, the same protein backbone can be glycosylated with different glycan structures. Further, these glycan structures adopt multiple conformations. These factors increase the microheterogeneity of glycosylation that in turn weakens the antigenic response to carbohydrates. Carbohydrate-binding antibodies need to accommodate this glycan microheterogeneity to avoid further weakening of the antigenic response (Rudd and Dwek 1997). Fourthly, as glycosylation is present in all cells the host may recognise the carbohydrates as self and be tolerated by immune cells.

These combined effects make glycans poorly immunogenic. One of the main concerns and limitations of eliciting antibodies against self-carbohydrates is their potential auto-reactivity and consequent negative selection during B cell development in vivo. A further challenge arises from pathogens that are transmitted between different species. Arthropod-borne pathogens such as Dengue virus (DENV) replicate first in mosquitos (surface proteins dominated by oligo- and paucimannose structures) before they infect humans (proteins dominated by complex-type sugars) (Dejnirattisai et al. 2011; Hacker et al. 2009; Crispin and Doores 2015), and therefore, the carbohydrate epitopes to be targeted by vaccines are different.

2.6 Currently Licensed Carbohydrate-Based Vaccines

Despite the highlighted challenges, several bacterial carbohydrate-based vaccines have been developed that are widely used. In 1977 the first carbohydrate vaccine, PneumoVax (Merck and Co.), was licensed consisting of a mix of 14 unconjugated capsular polysaccharides isolated from 14 different S. pneumonia strains. However, this vaccine was not very efficacious in the elderly and immunocompromised and therefore was replaced by the 23-valent vaccine, Pneumovax 23, consisting of 23 polysaccharides identified to be most pathogenic and covering 90% of strains causing serious disease (van Dam et al. 1990).

To overcome the challenges described above, induction of a T cell-dependent immune response against carbohydrates is favourable. To induce a long-lasting memory B cell response, precursor B cells are recruited to and activated in germinal centres formed in secondary lymphoid organs. CD4+ T cells, in particular follicular T-helper cells TFH, assist in affinity maturation, class switching and plasma B cell generation (MacLennan 1994; Sadanand et al. 2016; de Bree and Lynch 2016). The T-cell-independent nature of the carbohydrates can be overcome by chemically conjugating carbohydrates to carrier proteins to recruit CD4+ T cells. Suitable coupling proteins are tetanus toxoid (TT), diphtheria toxoid (DT), a mutated version of DT (CRM197) or the outer-membrane protein (OMP) of Neisseria meningitidis (Goldblatt 2000; Ada and Isaacs 2003). These carriers also provide several multiple loci for conjugation to increase multivalency. The first protein-conjugated carbohydrate vaccine was licensed in 1985 and protects against Haemophilus influenzae type b (Hib), the most common cause for meningitis and pneumonia. An improved version was approved 2 years later. It contains the main component of the Hib capsule called polyribosyl ribitol phosphate (PRP) conjugated to either tetanus toxoid (PRP-T, Hiberix, Merck an Co.), diphtheria toxoid (Pentacel, Sanofi Pasteur) or the meningococcal OMP (PedvaxHIB, Merk and Co.). In 2000, a conjugated S. pneumonia vaccine called PVC 7 was approved (Prevnar, Pfizer). The vaccine contains seven pneumococcal polysaccharides individually conjugated to the diphtheria CRM197 protein. In 2010, six more polysaccharides were added to constitute the PVC 13 vaccine (Prevnar 13, Pfizer). In case of the vaccine against Neisseria meningitides, purified capsular polysaccharides from serogroups A, C, W135 and Y (Men ACWY) have been used as vaccines in infants. The meningococcal polysaccharides of Men ACWY consist of homopolymers of either N‐acetyl mannosamine‐1‐phosphate in case of serogroup A or sialic acid residues in case of C, W135 and Y and are poorly immunogenic. However, chemical coupling of the polysaccharides to DT (Menactra, Sanofi Pasteur) or to CRM97 (Menveo, Novartis Vaccines and Diagnostics) enhances immunogenicity greatly (US Food and Drug Administration 2015). Vaccines against other glycosylated pathogens are under development and will be discussed in the following sections.

3 Vaccines Under Development Against Major Human Pathogens

3.1 Viruses

Thus far, no carbohydrate-based vaccine against viral glycans has been developed. Unlike some other pathogens, enveloped viruses use the host-cell glycosylation machinery to attach host-derived glycans to their viral surface and are therefore typically thought to be poorly immunogenic. However, recent advances in the isolation of antibodies from infected patients have led to the isolation of neutralizing antibodies (nAbs) that bind to carbohydrates on viral surface glycoproteins.

3.1.1 HIV-1 Envelope Glycans Are a Target for HIV-1 Broadly Neutralizing Antibodies

The main example of glycan-binding anti-viral neutralizing antibodies is HIV-1 broadly neutralizing antibodies (bnAbs). These are effective against a large number of circulating strains within each virus group and are able to prevent infection when passively administered to macaques in challenge studies (Moldt et al. 2012). Characterisation of their epitopes and their longitudinal development in vivo can give valuable information for vaccine design. HIV-1 infects immune cells (CD4+ T cells, dendritic cells and macrophages) and integrates into the host genome. There is currently no cure or vaccine and anti-retroviral drugs are the only effective treatment.

HIV-1 Envelope (Env) is the sole target for HIV bnAbs. Env consists of a trimer of gp41/gp120 heterodimers. It is first expressed as the precursor gp160, which is glycosylated in the ER and further processed in the Golgi and cleaved via Furin into gp120 and gp41. HIV Env is highly glycosylated with up to 50% of its mass consisting of host-derived N-linked glycans making it one of the most heavily glycosylated proteins known. In each monomeric gp120, there are approximately 25–30 potential glycosylations sites (PNGS) (Coss et al. 2016). It is thought that these glycans are important for transmission through interaction with lectins such as DC-SIGN but the glycans also shield conserved regions of the protein from the immune system. Therefore, the glycans on Env are often referred to as the glycan shield or silent face. Although the glycans are attached via the host-cell glycosylation machinery, analysis of the glycans present on Env shows an unusually high abundance of oligomannose-type glycans (Fig. 1). It has been shown that the dense clustering of N-linked glycosylation sites restricts access by the ER α-mannosidase I enzyme, thus preventing cleavage of Man8–9GlcNAc2 glycans resulting in a patch of underprocessed non-self high-mannose glycans (Bonomelli et al. 2011; Crispin and Doores 2015; Doores et al. 2010a; Go et al. 2013; Zhou et al. 2002; Behrens et al. 2016). This cluster is often referred to as the intrinsic mannose patch and is conserved across geographical HIV clades. The additional restrictions associated with trimerization results in a further elevation in oligomannose glycans leading to the trimer-associated mannose patch (Bonomelli et al. 2011; Crispin and Doores 2015; Doores et al. 2010a; Alexandre et al. 2010; Eggink et al. 2010).

Despite the challenges associated with eliciting glycan-binding antibodies described above, some HIV-infected individuals, after 2–3 years of infection, elicit a neutralizing antibody response that is targeted against the glycan structures on HIV Env and can neutralise up to 80% of circulating HIV strains (Goo et al. 2012; Sok et al. 2014). These bnAbs target three distinct glycan regions on Env. The most potent bnAb families target an area, which centres on glycan N332 on the outer domain of gp120 and their epitopes include protein residues in the V3 loop in addition to glycans at N332/N334, N301 and/or N295 (e.g. PGT121, PGT128, 10-1074) (Fig. 3) (Walker et al. 2011; Kong et al. 2013; Mouquet et al. 2012). The bnAb 2G12 targets the same region but only binds to glycans. 2G12 uses a unique domain-exchange structure to form a multivalent binding surface that enhances binding through increased avidity (Calarese et al. 2003). The other two glycan epitopes centre, firstly, on N160 and protein residues in the V1/V2 loop (PG9, PG16, PGT145) (Julien et al. 2013b; Pancera et al. 2013) and, secondly, on glycans at the interface of gp120 and gp41, including N88, N611 and N637 and surrounding protein residues (e.g. PGT151, 35O22, 8ANC195) (Falkowska et al. 2014; Huang et al. 2014; Scharf et al. 2015) (Fig. 3). Generally, these HIV bnAbs overcome the weak glycan–protein interactions by binding to multiple glycans and protein residues within one Fab.

Passive administration of HIV bnAbs to macaques is able to protect from SHIV infection in challenge studies (Moldt et al. 2012). These studies suggest that if these bnAbs could be elicited through vaccination then they would be able to prevent infection in humans (Moldt et al. 2012; Shingai et al. 2014). However, there are a number of challenges associated with eliciting such bnAbs through vaccination. They are highly mutated (30–40% compared to 10–15% found in other neutralizing antibodies) and often contain multiple insertions and deletions as well as a long complementarity determining regions H3 (CDRH3) that penetrates through the glycan shield to reach the protein surface beneath (e.g. PG9, PGT128) (Scheid et al. 2009; Walker et al. 2011; McLellan et al. 2011). Further, it can take 2–3 years of chronic infection before bnAbs can be detected (Kwong et al. 2013). Neutralisation escape usually occurs through addition or deletion of N-linked glycan sites. However, several bnAbs can accommodate viral escape by binding promiscuously to other glycans in close proximity (Doores et al. 2015; Krumm et al. 2016; Sok et al. 2014).

3.1.2 Vaccine Strategies for Eliciting Glycan-Binding HIV bnAbs

Research in the field of HIV vaccines is increasingly highlighting the HIV glycan shield as a good target for HIV vaccine design. BnAbs targeting the glycans on HIV-1 Env are highly desirable due to their high potency and breadth, and their ability to bind alternative arrangements of glycans within the mannose patch. Several strategies have been and are currently being investigated.

Initial design strategies focused on chemical synthesis of clusters of high-mannose glycans. These consisted of precise oligodendron scaffolds or multivalent arrays of glycans on viral particles such as bacteriophage Qβ (Astronomo et al. 2010; Doores et al. 2010b). Although these immunogens were able to elicit mannose-binding antibodies, they were unable to neutralise HIV-1. Similarly, using antigenic mimicry of yeast carbohydrates to induce 2G12 like antibodies was also unsuccessful (Agrawal-Gamse et al. 2011; Wang 2013). Immunization with monomeric gp120 or gp140 has been attempted in a number of studies but non-neutralising antibodies were mainly produced (Flynn et al. 2005; Pitisuttithum et al. 2006). A major breakthrough occurred with the development of a stabilised native-like trimeric Env (BG505 SOSIP.664). The trimer was stabilised through the introduction of a disulphide bond to covalently link the gp41 and gp120 subunits, and the addition of trimer-stabilizing mutations between gp41 subunits (Julien et al. 2013a; Sanders et al. 2002b, 2013; Khayat et al. 2013; Binley et al. 2000). The quaternary structure of this protein was shown to be important for correct glycosylation of Env (Pritchard et al. 2015). Immunization of SOSIP trimers derived from a viral strain with moderate sensitivity to neutralizing antibody (classified as tier 2) (Seaman et al. 2010) in small and non-human primate animal models produced homologous tier 2 neutralizing antibodies; however, heterologous neutralizing antibodies against other tier 2 viruses was not detected (Sanders et al. 2015). It is not known whether the elicited antibodies directly contact glycan structures on Env.

Current immunogen design strategies are now focused on recapitulating the bnAb affinity maturation process occurring in HIV-infected individuals. This involves characterising the bnAb lineage from germline and determining Env sequences that drive development of this lineage (Mouquet 2015; Jardine et al. 2013; Yang et al. 2015). It is thought that stepwise administration of rationally designed immunogens to activate B cell precursors might drive antibody maturation. Although current strategies are focused on eliciting CD4 binding site bnAbs (Jardine et al. 2015; Dosenovic et al. 2015; Jardine et al. 2016), these approaches have also been explored for glycan-binding HIV bnAbs (Steichen et al. 2016).

3.1.3 Targeting Glycans on Other Viral Pathogens for Vaccine Design

Although not as well studied as HIV, there are other viral targets for which glycan-binding bnAbs have been isolated.

3.1.3.1 Hepatitis C

Hepatitis C (HCV) is a major human pathogen with more than 170 million people suffering from chronic HCV infection leading to liver cirrhosis or liver cancer. Although progress has been made in the development of treatment and a cure, these are very expensive and a vaccine is desperately needed. The HCV glycoprotein consists of a heterodimer of E1/E2 and is the target of neutralizing antibodies. E1 and E2 have up to 6 and 11 PNGS, respectively, and these glycans are predominantly high mannose due to budding of viral particles from the ER.

A number of bnAbs have been isolated that potently inhibit HCV infection in vitro (Srinivasan et al. 2016). Antibodies AR3A-D target discontinuous protein and glycan residues within E2 (regions 396–424, 436–447 and 523–554) with Ser424, Gly523, Pro525, Gly530, Asp535, Val538 and Asn540 constituting conserved contact residues preventing receptor engagement (Law et al. 2008). AR4A is a potent neutralizing antibody but only recognises protein residues (Drummer and Poumbourios 2004). When the three most potent antibodies AR3A, AR3B and AR4A were generated via adeno-associated virus vectors in mice, all three antibodies potently inhibited HCV infection in vitro (de Jong et al. 2014; Law et al. 2008). Humanised mice that express the three AAV-derived antibodies were protected from infection when challenged with HCV. Furthermore, AR3A, AR3B and AR4A can abrogate an established HCV infection in vitro in primary human foetal liver cultures (de Jong et al. 2014). The therapeutic effect of these antibodies has also been demonstrated in vivo. When passively administered to mice with an established HCV infection, a single dose of the bnAbs reduced the genome copy number to detection limit. Essentially, these animals can be regarded as cured, probably by protecting healthy liver cells from infection and allowing for clearance of HCV in already infected hepatocytes (de Jong et al. 2014). Passive immunization using these bnAbs could be used as a feasible therapeutic, particularly in patients unresponsive to other anti-viral therapies, although antibody doses need to be determined in humans. Vaccine development targeting the conserved AR regions with immunogen-induced antibodies is a promising strategy.

3.1.3.2 Dengue Virus

Dengue virus (DENV), a member of the Flaviviridae family, is a mosquito-borne virus infecting about 400 million people yearly, predominantly in South America and the Pacific Islands, and there is currently no treatment. Neutralizing antibodies have been identified from infected individuals that target the surface glycoprotein, E. At the initial infection, the glycan profile of the surface protein is dominated by high- or paucimannose N-glycans consistent with production in insect cells while after the virus replicates in the human host, the glycans switch to complex type (Dejnirattisai et al. 2011; Hacker et al. 2009). Antibodies 747(4) A11, 747 B7, 752(2) C8 and 753(3) C10 contact the highly conserved glycans at positions N67 and N153. Antibodies A11 and B7 belong to the E-dimer-dependent epitope 2 group (EDE2) and are strictly dependent on the glycan at position N153, while C8 and C10 belong to the E-dimer-dependent epitope 1 group (EDE1) and can bind without N153 being glycosylated. The different glycan dependencies arise due to different mechanisms of binding to the protein component of the epitope. EDE2 and EDE1 both bind to the valley formed by the β strand on the domain II side (residues 67–74, 97–106 and 246–249). EDE2 interacts with the ‘150 loop’ of domain I (residues 148–159) containing N153 while EDE1 targets domains I and III and induces disorder of the 150 loop (Rouvinski et al. 2015). In a similar way to HIV, high levels of somatic mutation have been shown to be important for binding. Current vaccine design strategies use a similar approach as for HIV-1 vaccines by using a stabilised pre-fusion E dimer to stimulate B cells that give rise to DENV bnAbs. Recently, a recombinant, live, attenuated, tetravalent dengue vaccine has just been approved in Brazil, the Philippines and Mexico only. Dengvaxia (Sanofi Pasteur) is safe and effective against serotypes 1–4 yet it is not known if this vaccine elicits antibodies targeting the N-linked glycans on E.

3.1.3.3 Arenaviruses

Finally, and in contrast to the viruses described above, arenaviruses do not appear to induce any effective glycan-binding neutralizing antibody response (Sommerstein et al. 2015). Lassa virus, the most prominent member of the Old World arenavirus family, is endemic in West Africa causing haemorrhagic fever in severe cases and ultimately leading to death in up to 10% of the cases. Other members are of the New World arenaviruses causing Argentine, Venezuelan, Bolivian and Brazilian haemorrhagic fever are Junin, Guanarito, Machupo and Sabia viruses, respectively (Charrel and de Lamballerie 2003). Glycans present on the surface proteins GP do not prevent specific antibody induction per se but counteract neutralization by reducing the antibody on-rate and occupancy and prevent antibody-mediated virus control. This may explain why antibody immunity induced by either vaccination or the natural infection tends to be low. In this case, carbohydrate vaccines would have the opposite effects and are therefore not a good target for vaccine design against arenaviruses (Sommerstein et al. 2015).

3.2 Bacteria

As discussed above, several conjugate and polysaccharide-based vaccines have been developed against bacterial pathogens. However, there are many clinically relevant bacteria for which vaccine development is still ongoing. Bacterial vaccines are particularly challenging due to the heterogeneity and complexity of capsular polysaccharides (Astronomo and Burton 2010). Further, bacterial glycans can be very similar to human glycans, which can result in immune tolerance and poor immunogenicity. Bacterial glycans can be chemically modified to enhance immunogenicity. However, these modification needs to be finely balanced to maintain the glycan integrity for the immune system to still be able to recognise the pathogen but be distinguishable from the human version (Astronomo and Burton 2010). Current glycoconjugate vaccines are produced by purification of the polysaccharide from the native pathogen followed by chemical coupling to a carrier protein. This process can lead to impurities and batch-to-batch variation, and therefore, novel approaches using synthetic carbohydrates are currently being investigated. Synthetic vaccines allow reproducible production of well-defined conjugates free from contaminants. This system can be adapted for use with many serotypes and multiple serotype combinations are possible. Synthetic carbohydrate vaccines offer the possibility of varying the sugar chain length and coupling density, as well as chemical modification of the glycan, to induce the optimal antibody response (Pozsgay 2008). Overall, synthetic carbohydrate vaccines help to understand the relationship between chain length, saccharide composition and secondary structured by creating a flexible platform to improve immunogenicity (Astronomo and Burton 2010).

3.2.1 Shigella

Shigella causes approximately 1.1 million deaths each year with 60% being children under 5 years of age (Kotloff et al. 1999). There are approximately 50 serotypes grouped into 5 serogroups. S. sonnei is the most prevalent in developed countries, while S. flexneri is prevalent in developing countries. The O-linked polysaccharides (O-PS) are a major component in the bacterial cell wall and determine the serotype. To protect against the major circulating strains, a multivalent vaccine needs to be developed. A vaccine containing S. dysenteriae type 1, S. sonnei, S. flexneri 2a, 3, and 6 would protect against 75% of Shigella outbreaks (Levine et al. 2007). The inclusion of the three S. flexneri O-PS would cross-protect against all 11 serotypes as they share the same polysaccharide antigen (Noriega et al. 1999). Conventional conjugation of ShigellaO-polysaccharide covalently linked to a carrier protein has given good efficacy, however not in young children (Passwell et al. 2010). Therefore, new vaccine approaches are being pursued. Current research is focused on synthetic production of more immunogenic conjugate vaccines. For example, a synthetic pentasaccharide made of three repeating units of the O-PS of S. flexneri 2a is recognised by antibodies in serum from infected individuals and can convey protection in mice (Phalipon et al. 2009). When administered as a conjugate linked to TT in the presence of the adjuvant alum, it induced an increased anti-S flexneri 2a antibody response that can be sustained for more than one year (van der Put et al. 2016). Furthermore, potential vaccine candidates against S. dysenteriae include multiple repeats of the tetrasaccharide of the O-PS. Different densities of tetra-, octa-, dodeca- and hexadecasaccharides were tested, but only the octa-, dodeca- and hexadeca-conjugates were able to induce an immune response (Pozsgay et al. 1999).

3.2.2 Protein-Glycan Coupling Technology to Develop Vaccines Against Shigella Dysenteriae and Francisella Tularensis

An experimental vaccine candidate produced with a new technique called protein glycan coupling technology (PGCT) is under investigation against Shigella dysenteriae and Francisella tularensis (Fig. 4). PGCT can be applied to numerous combinations of recombinant protein–glycan structures and replaces in vitro conjugation of polysaccharide to proteins. This method uses a coupling enzyme oligosaccharyltransferase (OST, for example, PglB isolated from Campylobacter jejuni) that transfers the reducing end of a sugar to a consensus recognition sequence in the carrier protein (D/EYNXS/T). All three components are cloned into bacterial expression plasmids and transformed into the appropriate E. coli strain to express a recombinant glycan–protein conjugate that can easily be produced in large scale and purified in a single step (Terra et al. 2012). Additional advantages are the maintenance of structural integrity due to a quick turnover and the usage of E. coli over more pathogenic bacteria. The carrier proteins might not contain the glycosylation consensus sequon; however, they can be engineered onto the target protein (Fisher et al. 2011). The limiting factor for this method is the substrate specificity of the OST PglB. It only transfers glycans with a reducing end sugar containing an acetamido group in the C2 position (Wacker et al. 2006). Other PglBs with broader specificity are under investigation (Terra et al. 2012). A glycan-conjugate vaccine against S. dysenteria using PCGT has been developed by GlycoVaxyn AG. The O-PS was coupled to the carrier protein exotoxin A of Pseudomonas aeruginosa. A Phase I trail demonstrated the vaccine candidate was not only safe but elicited a good antibody response against Shigella O-PS at a range of doses (Hatz et al. 2015).

Fig. 4
figure 4

(adapted from Cuccui et al. 2013 and Terra et al. 2012)

Schematic of the protein glycan coupling technology. E. coli cells are transformed with three bacterial expression plasmids encoding for the carrier protein harbouring the glycan acceptor sequon D/EXNXS/T, the polysaccharide locus and the oligosaccharyltransferase (OST) PglB. The OST transfers the glycan containing the reducing end sugar (orange star) onto the carrier protein to produce the desired glycoprotein in large amounts

Full size image

The bacterium F. tularensis, a class A bioterrorism agent, causes tularaemia and is very infectious at low doses, has a high fatality rate and is easily transmitted via aerosol. Attempts to use a live vaccine failed as it caused disease in the mouse model (Fortier et al. 1991). Purification of the O-antigen from the pathogen [a tetrasaccharide with the structure 4-α-d-GalNAcAN-(1–4)-α-d-GalNAcAN-(1–3)-β-d-QuiNAc-(1–2)-β-d-Qui4NFm-(1-)] is challenging due to the difficulties associated with handling of highly infectious agents. Therefore in the current strategy, the O-antigen is produced in E. coli and then coupled to carrier protein exotoxin A from P. aeruginosa with the help of the OST PglB (Prior et al. 2003). This glycoconjugate vaccine showed promising efficacy in a mouse model, demonstrating that the PGCT method might be a valuable strategy to produce glycoconjugate vaccines against challenging to work with bacterial pathogens (Cuccui et al. 2013).

3.3 Fungi

Fungi can cause severe infections, particularly in hospital settings when a patient’s immune system is compromised. Despite there being thousands of fungal species, only a select few cause human disease including several species of Candida as well as other mycoses like Aspergillum and Cryptococcus, and dimorphic fungi such as Histoplasma capsulatum or Blastomyces dermatitidis. The protective cell wall of fungi is generally covered by antigenic polysaccharides such as β-1,3-glucan, β-1,6-glucan, α-mannan, β-mannan and chitin (a linear homopolymer of β-1,4-linked N-acetylglucosamine). Together, they form the scaffold for predominantly GP1-anchored glycoproteins that are themselves extensively modified with N-linked and O-linked oligosaccharides (Klis et al. 2011; Plaine et al. 2008; Bowman and Free 2006). The major obstacle for broad vaccine development is the divergence in cell wall structure between fungal strains. Most vaccine strategies have been univalent consisting of one antigen rendering them strain specific. High treatment costs highlight the need for an effective vaccine in both healthy and immunocompromised patients that induces long-lasting immunological memory. It is thought that this can be achieved by combining mechanisms of innate and adaptive immune response (Roy and Klein 2012; Dockrell et al. 1999; Leung et al. 2004; Levin et al. 2001; Madhi et al. 2005; Nordoy et al. 2002; Tedaldi et al. 2004).

3.3.1 Candida Albicans

The most prevalent fungal infection worldwide is caused by Candida albicans, a component of the commensal gut microbiota. This polymorphic organism exists in two life forms: unicellular (yeast) or multicellular (hyphae) and it is the hyphal form that causes the disease. Breaches in the tissue barrier (most common cause, triggered mainly by surgery, catheters, fasting, extended hospital stay) altered gut flora (through use of broad-spectrum antibiotics) or immune suppression can lead to the development of mucosal candidiasis. In the most severe cases, the fungus circulates in the bloodstream causing a highly lethal candidemia and can infect organs (Cassone 2013; Diekema et al. 2012; Pappas 2006).

C. albicans displays a wide variety of antigenic factors such as members of the heavily glycosylated agglutinin-like substance (Als), superoxide dismutase (Sod) or secreted aspartyl proteases (Sap) families (Cassone 2013). Thus far, two vaccines targeting protein antigens have successfully completed Phase I clinical trials. The first consists of rAls3p-N (recombinant N-terminus of Als3p, NDV-3, NovaDigm Therapeutics) (Schmidt et al. 2012) and the second (PEV-7, Pevion Biotech) relies on amino acids 77–400 of the secreted Sap2p (De Bernardis et al. 2012, 2015). The Als3-containing vaccine includes both a TH1 and TH17 cell-based immune response as well as an Als3-specific antibodies response. Even though these approaches induce a promising immune response, they fail to induce a persistent protection due to univalency of the strategy (Cassone 2013). Glycoconjugate immunogens could be a promising alternative for a multivalent vaccine strategy.

Glycoconjugate vaccines consisting of β-glucan and β-mannan polysaccharides conjugated to tetanus or diphtheria toxoid (Torosantucci et al. 2005; Xin et al. 2008) have been shown to elicit the required T cell response as well as induce antibodies that cross-protect against major fungal pathogens including Asperillus spp. and Cryptococcus spp. (Gaffen et al. 2011; Kaufmann 2007; Romani 2011; Spellberg et al. 2008). These vaccines could be delivered via virosomes or other nanoparticles. Those carriers have been used successfully in other vaccine candidates (e.g. against influenza) and been shown to induce a robust T cell response (Cusi 2006; Moser et al. 2013; Radosevic et al. 2008; Zhao et al. 2014). Immunization with the β-1,3-glucan polysaccharide laminaran, isolated from algae and conjugated to CRM197 (Lam-CRM197), in combination with the adjuvant MF59 was found to reduce mortality when mice were challenged with a lethal dose of C. albicans (Bromuro et al. 2010; Pietrella et al. 2010). This vaccine was also effective against Aspergillus, and Cryptococcus suggesting cross-protection against fungi is possible by targeting the glucans in their cell wall (Bromuro et al. 2010; Torosantucci et al. 2005).

Since the fungus is also part of the commensal microbiota, the ideal vaccine would induce antibodies specific for the disease causing hyphal form of Candida (Cassone 2013). Attenuated or inactivated full organism versions, while able to induce long-lasting immunity for viral or bacterial pathogens (Minor 2015), have to be different enough to prevent cross-reactivity with commensal yeast strains (Saville et al. 2009). Therefore, a more specific approach being investigated focuses on coupling the polysaccharides to hyphal associated proteins to increase specific anti-fungal efficacy. Two such proteins, expressed as virulence factors in the cell wall during the yeast-to-hypha transition, are the hyphally regulated protein (Hyr1) or the hyphal wall protein (Hwp1) (Cassone et al. 2010; Luo et al. 2010; Xin et al. 2008). A β-mannan trisaccharide coupled to Hwp1 [β-(Man)3-Hwp1]-induced significant protection (80–100%) in mice studies (Xin et al. 2008), rendering this glycoconjugate vaccine approach highly promising. Of note, such antibodies could also be administered safely to treat fungal infections passively in infants and immunocompromised patients.

3.3.2 Aspergillus Fumigatus

The second most common fungal infection is caused by Aspergillus fumigatus leading to allergic bronchopulmonary or invasive aspergillosis and other fungal diseases and has a very high mortality rate particularly in immunocompromised individuals. Vaccination against A. fumigatus with crude antigen preparations has been shown to be effective in immunocompromised mice (Ito and Lyons 2002). The immunodominant antigen in the crude extract has been determined: Asp f3, a peroxisomal protein in the hyphae (Diaz-Arevalo et al. 2011), and when the protein was administered subcutaneously, it protected mice from a lethal dose of A. fumigatus (Ito et al. 2006). With regard to glycans, galactofuranose-containing glycolipids (glycosylinositolphosphoceramid and GPI-anchored lipophosphogalactomannan) as well as polysaccharide (α1-3glucan, β1–3glucan, and galactomannan) induced a robust immune response in mice, activating IL-17 or IFN-γ, IL-17-, and IL-10-producing T cells, respectively (Bozza et al. 2009). Antibodies against β1–3glucan partially contribute to the immune response (Torosantucci et al. 2005).

3.3.3 Cryptococcus Neoformans

Among the Crytococcus spp., Cryptococcus neoformans is the third most common cause of invasive fungal diseases. The capsule of C. neoformans consists predominantly of the polysaccharides α-1,3-d-mannopyranose units with single residues of ß-d-xylopyranosyl and ß-d-glucuronopyranosyl attached called glucuronoxylomannan (GXM) and two minor components, galactoxylomannan (GalXM) and mannoprotein rendering it the prime target for vaccine design (Cherniak and Sundstrom 1994). This dominant virulence factor GXM causes an immunodysregulatory effect and thereby suppresses the host inflammatory response and prevents opsonophagocytosis (Shoham et al. 2001; Spellberg 2011). However, when GXM coupled to a carrier such as TT is administered to mice, it induces a protective antibody response (Mukherjee et al. 1993). The same is true for GalXM when the polysaccharide its conjugated to bovine serum albumin (BSA) or a protective antigen of Bacillus anthracis (Chow and Casadevall 2011). Even passively administered GXM-specific antibodies protect mice from cryptococcosis (Casadevall et al. 1998; Mukherjee et al. 1993). However, GXM vaccines also elicit non-specific or deleterious antibodies that bind to other sites in GXM reducing the protective efficacy of the vaccine (Devi et al. 1991; Mukherjee et al. 1995). Additionally, GXM exerts immunomodulatory effects that interfere with leukocyte migration (Ellerbroek et al. 2004; Vecchiarelli 2000). To assess the side effects of these deleterious antibodies in humans, therapeutic efficacy of passively administered antibodies was assessed in a phase I dose-escalation study with the goal to establish the safety of antibody administration in patients with cryptococcal antigenemia. HIV-1 infected patients with a history of culture-proven and treated monococcal meningitis were administered with dfferent doses of murineanticryptococcal monoclonal antibody 18B7 and the study found that the antibody is safe to be administered at doses up to 1.0 mg/kg without evidence of toxicity as suggested by in vitro or animal studies (Larsen et al. 2005).

Alternative approaches under investigation include a peptide mimotope (P13) of the polysaccharide GXM conjugated to either BSA or tetanus toxoid. These vaccines were shown to prolong survival of mice infected with a lethal dose of C. neoformans (Fleuridor et al. 2001). As mentioned above, the laminaran (β-1,3-glucan polysaccharide) vaccine also cross-protects against C. neoformans. Other examples of protective antigens constituting the capsule include the less abundant mannoproteins. Yeast proteins contain generally terminal mannosylated glycans, which are distinct from host glycans. The role of terminal mannosylation, more the O- mannosylation than the N- mannosylation, has been found to be crucial for mediating a potent T cell-mediated immune response, CD8+-T cell function and proliferation as well as secretion of pro-inflammatory cytokines such as TNF and IL-12 (Specht et al. 2007; Lam et al. 2005; Luong et al. 2007) and should be taken into consideration for inclusion in future vaccine design strategies to stimulate the immune system additionally and hence increase efficacy of any other vaccine (Levitz et al. 2015).

Multivalent, broad-spectrum anti-fungal vaccines are the ultimate goal to achieve. A multivalent vaccine would provide broad, additive and synergistic protection reducing the chances of immune evasion as has been achieved for S. pneumonia vaccination. Heat-killed Saccharomyces cerevisiae (heat-killed yeast, HKY) vaccination has been shown to protect mice against systemic aspergilliosis (Liu et al. 2011), coccidioidomycosis (Capilla et al. 2009), candidiasis (Liu et al. 2012) and cryptococcosis (Majumder et al. 2014) suggesting a broad vaccine could be possible if the right immunogens were identified. Furthermore, a combination of common cell wall polysaccharide components, e.g. β-glucan and β-mannan or mannoproteins, conjugated to fungal proteins, such as Candida proteins Als3, Hwp1 or Hyr1 alone or in combination, maybe a promising approach to enhance not only efficacy but also breadth of the immune response (Casadevall and Pirofski 2006; Johnson and Bundle 2013). However, efficacy, cross-reactivity, and sustainability of the immune response need to be determined in animal models, as well as in humans.

3.4 Parasites

Parasites are complex organisms both genetically and biologically and cause diseases such as schistosomiasis, malaria, toxoplasmosis and leishmaniasis, and are caused by protozoan parasites. Several peptide-, protein- and DNA-based vaccines have been tested, but with limited success (Graves and Gelband 2003). Glycan and peptide antigens are very complex, not always present on all parasite life stages and hence are relatively poorly understood. It is only recently that glycans have been identified as being a dominant antigen on certain parasites (Hokke and Deelder 2001; Thomas and Harn 2004). However, vaccine development has been slow as it is difficult to obtain enough glycan material due to challenges in culturing of parasites or complex synthetic protocols. This section focuses on the trematode Schistosome mansoni and the protozoan parasite Plasmodium falciparum.

3.4.1 Schistosomes

Schistosomiasis is caused by trematodes (helminthic parasites) such as Schistosoma mansoni, S. intercalatum, S. haematobium, S. japonicum and S. mekongi. The life cycle involves both human and snail hosts and the different life stages (larval (cercaria), schistosomula, adult worms and eggs) present different glycan structures that also vary in their abundance. It is the egg stage that typically causes severe disease. Eggs spread disease through shedding into the urine or faeces. When the eggs get trapped in the human tissue they cause fibrosis, calcification, hepatosplenomegaly and potentially fatal bleeding from oesophageal varices (Richter et al. 1998). During schistosome infection, a robust antibody response is induced, however this response is not very protective and immunity can take many years to develop (Eberl et al. 2001b; Hagan and Sharaf 2003). Praziquantel (PZQ) is the only approved drug for treatment of schistosome infection. Many parasite immunogenic proteins have been investigated in vaccine studies, but with limited efficacy (Hewitson and Maizels 2014; Hagan and Sharaf 2003). Glycans rather than proteins are the immunodominant antigens with antibodies mainly targeting the glycan structures on parasitic glycoproteins (Nyame et al. 2004). The development of glycan microarrays displaying parasite glycans has allowed the identification of antigenic motifs (van Diepen et al. 2012a, 2015). One such array is a shotgun microarray where glycans isolated directly from the parasite are printed on glass slides allowing unique and unusual parasite-specific glycans to be included that are not available through chemical synthesis (van Diepen et al. 2012b, 2015; de Boer et al. 2008). Antibodies elicited during infection target specific N- and O-linked antigenic structures containing the motifs LacDiNAc (LDN, GalNAcβ1-4GlcNAc), LDN-F (GalNAcβ1-4(Fucα1-3)GlcNAc), Lewis X (LeX, Galβ1-4(Fucα1-3)GlcNAc) and core α3-fucose/β2-xylose (Fig. 1). These motifs are present in all parasite life stages but vary in surface expression levels and exposure. Terminal β-linked GalNAc is an example of a non-self parasitic motif not found in vertebra. LDN is continuously present while LeX and LDN-F are mainly synthesised after transformation into schistosomula (Smit et al. 2015). Importantly, LeX targeting IgM, IgG as well as IgA have been isolated, while LDN and LDN-F targeting IgG and core α3-fucose/β2-xylose targeting IgE have been found (Van Roon et al. 2004).

A vaccine producing antibodies against glycans and glycoproteins present on the different parasite life stages is likely required for full protection in humans. A vaccine consisting of soluble egg antigens (SEA) conjugated to cholera toxin B subunit has been trailed in mice and shown to induce IgE, enhance the Th2 response, and reduce liver granuloma and mortality (Sun et al. 2001; Okano et al. 1999). Radiation-attenuated schistosomes have been successfully used to induce a IgM and IgG response (Delgado and McLaren 1990; Hewitson et al. 2005; Eberl et al. 2001a). The protective antigen has not been identified yet and the application of these vaccine approaches in humans is questionable. Parasite glycans are particularly of interest for vaccine design due to their high density and multivalent display on the parasite surface. However, it is imperative to identify the correct glycan structures to achieve the most specific and potent immune response. This is an area still under investigation (Luyai et al. 2014; Mandalasi et al. 2013; Prasanphanich et al. 2014) and the development of glycan microarrays will strongly inform these studies.

3.4.2 Plasmodium

Another very well-studied protozoan parasite of the genus Plasmodium causes Malaria in humans. P. falciparum is the most pathogenic and is transmitted as sporozoites via mosquitos. It infects hepatocytes in the liver where it matures to merozoites which can then infect red blood cells. The parasite causes red blood cell lysis and release, and it spreads to other red blood cells causing severe tissue pathology. Current vaccine strategies target several different stages of the parasite life cycle, pre-erythrocytic, erythrocytic and transmission blocking. The malaria vaccine MosquirixTM (RTS,S by GSK) is the most advanced pre-erythrocytic vaccine candidate and contains portions of the circumsporozoite protein. It is about 30% effective against the most prevalent parasite in Africa, P. falciparum, but is not effective against P. vivax, which is prevalent outside of Africa. The requirement of four doses further complicates the application of this vaccine (Todryk and Hill 2007).

Novel vaccine strategies are focused on GPI anchors, which are dominant toxins responsible for the pathology. They are important in signal transduction processes by activating macrophages and vascular endothelial cells leading to production of pro-inflammatory mediators such as nitrous oxide, tumour necrosis factor α and intercellular adhesion molecule-1 (Schofield et al. 1996; Tachado et al. 1996; Nyame et al. 2004). GPIs are produced synthetically and contain multiple mannose residues, glucosamine and 6-myoinositol-1,2-cyclic phosphate (NH2-CH2-CH2-PO4-(Manα1-2)6Manα1-2Manα1-6Manα1-4GlcNH2α1-6myo-inositol-1,2-cyclic phosphate) conjugated to the carrier keyhole limpet haemocyanin (KLH) (Fig. 1) (Hewitt et al. 2002; Liu et al. 2005; Schofield et al. 2002). In vivo studies showed good IgG titre induction against parasitic GPI, but importantly not mammalian GPI. Immunization of mice with the GPI conjugate resulted in a 75% reduction in death rate after Plasmodium challenge and showed cross-reactivity with other Plasmodium species, also to those only infecting mice. This phenotype demonstrates a conserved GPI structure. However, the GPI conjugate vaccine did not prevent infection per se nor did it cause parasite death but was found instead to neutralise toxicity associated with infection (Nyame et al. 2004).

The carbohydrate epitope Galα1-3Galβ1-4GlcNAc-R (α-gal) is a structure that was deleted from the human glycan repertoire during evolution. α-gal is therefore a xeno-antigen and is expressed by bacterial components of the microbiota. 1–5% of circulating IgG and IgM in humans recognises this glycan structure (Macher and Galili 2008). Interestingly, Plasmodium sporozoites have α-gal on the surface, whether α-gal is produced by the parasite itself of by the mosquito is not yet clear (Galili et al. 1998; Warburg et al. 2007). It has been shown that α-gal antibodies elicited against the gut microbiota protect from P. falciparum infections in humans and transmissions in mice experiments (Yilmaz et al. 2014). However, during natural infection in humans, α-gal antibodies levels are too low to be protective. Strategies to boost this promising α-gal immune response, e.g. via addition of adjuvants or by coupling the α-gal to Plasmodium-specific antigens, are under investigation as a vaccine candidate (Benatuil et al. 2005; Yilmaz et al. 2014). This strategy could potentially protect against other vector-borne protozoan parasites expressing α-gal, such as Leishmania spp. and Trypanosoma spp.

4 Discussion

There is a lot of activity in the field of carbohydrate-based vaccine research with some promising candidates in the pipeline. Although there are currently only carbohydrate-based vaccines against bacterial pathogens, the potential for their development against other pathogens is high. This is particularly true for viral pathogens considering the isolation of many glycan-binding neutralizing antibodies. However, the approaches used are quite different. For viral pathogens, a reverse vaccinology methodology is being utilised where functional antibodies are isolated from natural infection and then characterisation of their binding epitopes is informing immunogen design in an attempt to re-elicit these antibodies in vivo. For other glycosylated pathogens the approach is to generate immunogens that contain surface glycans/polysaccharides coupled to protein carriers. With the development of high-throughput glycan microarray methodologies and antibody isolation techniques, it would be possible to use this reverse vaccinology approach with the other glycosylated pathogens described above. In this way, additional information on how glycan-binding antibodies generated from natural infection differ from those generated from vaccination and may inform vaccine design against many different relevant pathogens.