EA040853B1 - OLIGOSACCHARIDES WITH A CEP STRUCTURE CONTAINING SEVEN OR MORE UNITS FROM 4,6-DIDEOXY-4-ACYLAMIDO-ALPHA-PYRANOSE AND THEIR CONJUGATES AS VACCINES DIRECTED AGAINST INFECTIONS CAUSED BY BRUCELLA ORGANISMS - Google Patents

Description

Technical field

The present invention relates to a polysaccharide molecule which is useful as a vaccine component for vaccinating animals against infection with Brucella organisms, such that a DIVA test (differentiation of infected from vaccinated animals) is possible. The present invention also relates to novel diagnostic methods for detecting infection with a Brucella organism.

Background of the invention

Brucellosis is one of the most serious zoonotic diseases in the world and is caused by bacteria of the genus Brucella. They are non-spore-forming coccobacterial rods with a cell wall characteristic of Gram-negative bacteria, which includes peptidoglycans, outer membrane proteins, and lipopolysaccharide (LPS). The species B. abortus, B. melitensis and B. suis have the greatest impact on human and animal health. The LPS of field strains of these species have an O-polysaccharide (OPS) that protrudes from the cell wall, rises above the surface and changes colony morphology, which is why they are described as smooth molecules and as strains that contain smooth (s) LPS. Strains that do not contain OPS on their surface are described as rough and contain rough (r) LPS. The main symptom of the disease in farm animals is the inability to reproduce, which is most often expressed in the form of miscarriage and male infertility. This leads to large losses in animal husbandry and to the spread of the disease among people. Otherwise, many animals will look healthy on the outside.

The impact of this disease is felt in most countries of the globe. Even in the absence of disease, it is important to conduct an examination to maintain a healthy status. In most countries, the disease continues to persist or may reemerge, and large natural reservoirs exist.

The fight against endemic brucellosis is possible only through mass vaccination. Testing and culling programs at high levels of disease are unaffordable, unwise and unpleasant. After a period of mass vaccination, a further downward trend in prevalence can be achieved by vaccinating breeding animals. In practice, however, it can be difficult to provide effective vaccination with sufficient coverage. Current vaccines provide insufficient protection to eliminate the disease. Therefore, a testing and culling program based on serology results is required to achieve disease eradication. Serodiagnosis is a key tool in maintaining health and facilitating the livestock trade, as well as for epidemiological research. Other methods of disease detection, such as direct culture of Brucella and PCR to detect Brucella-specific DNA, are not sensitive enough, are more expensive, and if cultured, the cultures carry significant health risks. Immunodiagnostics using cellular reactions is also not effective enough due to low diagnostic sensitivity and reactions resulting from vaccination. (Pouillot et al (1997) Vet Res 28:365-374).

The currently recognized brucellosis vaccines are B. abortus S19 and B. abortus RB51 for use in cattle, and B. melitensis Rev1 for use in sheep and goats. All have significant and well-documented deficiencies (Blasco et al (2015) Veterinary Vaccines for Developing Countries. FOA, Rome). For example, they are all live (in this case, the use of a cold chain is necessary for distribution), have residual virulence in farm animals, and are pathogenic to humans. Both B. abortus RB51 and B. melitensis Rev1 are resistant to antibiotics that are important for the treatment of this disease in humans. S19 B. abortus and Rev1 B. melitensis may be unable to synthesize OPS, resulting in a loss of protective efficacy. This is especially problematic with Rev1 B. melitensis (Mancilla et al (2010) Journal of Bacteriology 192: 6346-6351).

B. abortus S19 and B. melitensis Rev1 are smooth strains and therefore their use results in the induction of antibodies that are positive in standard serological tests for Brucella infection. This is because the serodiagnostic antigens used in these tests are based on the presence of OPS, which gives them excellent sensitivity, while antigens without OPS are ineffective diagnostic tool (McGiven (2013) Rev Sci Tech 32:163-176). In many cases, this makes it impossible to distinguish between infected and vaccinated animals. This creates a significant barrier to effective control, as it makes it extremely difficult to effectively implement vaccination along with a serodiagnostic-based test and exclusion strategy.

A rough strain of B. abortus RB51 has been bred to alleviate serological reactions resulting from vaccination with S19 of B. abortus in cattle. This rough strain was bred by passage on rifampicin-soaked culture medium (Schurig et al (1991) Vet Microbiol 28:171-188). Of the many mutations spontaneously created by an antibiotic, some affect the synthesis of OPS. The strain can neither synthesize OPS nor, if the synthesis is supplemented by introducing a gene, export it to the cell surface (Vemulapalli et al (2000)

Infect Immun 68). Therefore, the use of such a vaccine does not result in the production of antibodies to OPS and has limited interference with routine serology.

Despite this advantage of RB51 B. abortus, its acceptance has been far from universal. Its protective efficacy and safety in cattle compared to B. abortus S19 is also controversial (Moriyon et al (2004) Veterinary Research 35:1-38). On mouse models of protection, i.e. in a primary animal model, it has been shown to be significantly less protective than B. abortus S19 (Monreal et al (2003) Infection and Immunity 71:3261-3271).

Attempts to develop B. abortus RB51 as a vaccine against brucellosis in sheep and goats have not been successful in terms of protection (Idrissi et al (2001) Rev Sci Tech 20:741-747). Rough strains of B. melitensis have also been evaluated as vaccines for sheep, but have not shown sufficient protective efficacy (Barrio et al (2009) Vaccine 27:1741-1749). B. melitensis rough strain B115 has been used in sheep but has been shown to be insufficiently safe, with a high miscarriage rate reported (Perez-Sancho et al (2014) Vaccine 32:18771881). Previous studies in mice have shown that B115 B. melitensis induces the production of antibodies to OPS, which is due to the presence of cytoplasmic OPS (Cloeckaert et al (1992) J Gen Microbiol 138:1211-1219). A comprehensive comprehensive study of the protective ability of rough B. melitensis mutants to protect against B. melitensis infection in mice showed that rough variants that did not synthesize OPS had significantly less protective properties than their smooth counterparts (Gonzalez et al (2008) PLoS ONE 3: e2760).

Anti-OPS antibodies have been shown to be protective in a mouse model of brucellosis (Montaraz et al (1986) Infection and Immunity 51:961-963). The combination of humoral and cellular protection has been shown to be synergistic (Grill0 et al (2006) Vaccine 24:2910-2916) and provide optimal protection. Although the data from the natural host is less strong, the evidence based on the relative properties of smooth and rough strains described above is strong, especially for small ruminants.

The search for improved vaccines against Brucella is a recognized need, but despite decades of effort, no new vaccines have been introduced for animal use since the introduction of RB51 B. abortus. Research into vaccines based on recombinant protein subunits, Brucella protein expression vectors, or DNA vaccines has not yet made any impact in the art. In the case of attenuated smooth mutants, Brucella encountered the same antibody induction problems as with B. abortus S19 and B. melitensis 16M, and sufficient additional advantages in other respects, such as safety and efficacy, were not demonstrated to be was to be adopted outside the research laboratory.

As discussed, new effective developments have been seriously hampered by the persistent and seemingly intractable problem that OPS appears to be required for both optimal vaccine protection in livestock and effective serodiagnosis. This is a long-standing barrier to the development of a vaccine with optimal protective properties that can be used in a vaccination and testing scheme for DIVA (Differentiating Infected from Vaccinated Animals).

The main building block in OPS Brucella is a homopolymer of 4,6-dideoxy-4-formamidomanopyranosyl (D-Rha4NFo) units, which are linked in various ways: α(1 -> 2) and α(1 -> 3) (Meikle et al (1989) Infect Immun 57:2820-2828). The proportion of each bond type seems to vary between strains of Brucella, with a frequency of 0 to 20% being characteristic of the α(1-3) bond type and the remainder being the α(1-2) bond type. It should be noted that only the B. suis biovar type 2 strain was found to lack α(1–3) linkages (Zaccheus et al (2013) PLoS One 8:e53941).

As detailed in WO 2014/170681 (the contents of which are incorporated herein by reference in their entirety), the relative proportions and distribution of α(1-2) and α(1-3) bonds in Brucella homopolymer OPS and Y. enterocolitica O:9 allow to create different, but not necessarily fully described, binding antibody epitopes. Brucella has three distinct antigenic epitopes that can be found in OPS for which there is strong evidence for their structure (Bundle et al (1989) Infect Immun 57:2829-2836), which are summarized in Table 1. 1.

- 2 040853

Table 1. OPS epitopes

Epitope name Number of perosamines Characteristics In which OPS present C/Y 3-4 N-formylperosamines are linked exclusively by a(1->2) bonds All smooth strains of Brucella as well as 0:9 Y. enterocolitica A 5 or more N-formylperosamines combine a(1->2) connections Predominantly in all A-dominant strains of Brucella, as well as in 0:9 Y. enterocolitica M 2-6 Present by at least one a(1->3) bond without a(1->2) links, with one or two adjacent a(1->2) links; the location of the a(1->3) bond in the epitope is not defined Predominantly in OPS-dominant Brucella strains, but also, to a lesser extent degree, in A- dominant strains. Not found in 0:9 Y. enterocolitica or biovar 2 B. suis

Nielsen et al (Nielsen et al (1989) Am J Vet Res 50:5-9) suggested the presence of an additional epitope at the non-reducing end region of the OPS. These authors put forward the idea that during vaccination with S19 B. abortus, antibodies were produced against this theoretical epitope.

OPS Brucella is formed as a block copolymer of D-Rha4NFo (Kubler-Kielb & Vinogradov (2013) Carbohydrate Research 378:144-147), with two polymer units combined into one molecule along with three non-D-Rha4NFo sugars at the reducing end , which form adapter and seed sites (Kubler-Kielb & Vinogradov (2013) Carbohydrate Research 366:33-37). The first D-Rha4NFo polymeric element found at the reducing end is a sequence of D-Rha4NFo units, all of which are α(1->2) linked. This sequence is linked to a second polymeric element consisting of one or more D-Rha4NFo tetrasaccharide units containing a central α(1-3) bond, while otherwise the bonds are α(1-2) bonds. As mentioned above, the presence of an α(1-3) bond is a characteristic feature of the M-epitope. OPS of M-dominant strains of Brucella have several times more of these tetrasaccharide units associated with the polymer with α(1–2) bonds. OPS of A-dominant strains contain one or two of these terminal tetrasaccharides linked to a longer polymer with α(1-2) bonds. Therefore, an α(1-3) bond is present near the top of every OPS molecule, whether it comes from an A- or M-dominant strain of Brucella.

The significance of this binding detail is that it virtually changes the shape of the OPS and affects antibody binding. This has been shown in numerous studies using monoclonal antibodies (mAbs) (Bundle et al (1989) Infect Immun 57:2829-2836) and absorbed monospecific polyclonal sera used in the classic Brucella biotyping scheme to classify strains as any of the A , M, or mixed A and M serotypes (Alton et al (1994) INRA Editions).

Infection with other Gram-negative bacteria that share similar OPS structures can induce the production of antibodies that cross-react with Brucella OPS (Corbel (1985) Vet. Bull. 55: 927-942), leading to false positive serological reactions (FPSRs). The most frequently mentioned of these is Yersinia enterocolitica O:9, as it contains a homopolymer that consists of D-Rha4NFo units with exclusively α(1-2) linkages (Caroff et al (1984) Eur J Biochem 139:195-200).

The invention disclosed in WO 2014/170681 relates to various synthetic oligosaccharide structures based on the Brucella OPS structure. All structures contain at least two 4,6-dideoxy-4-formamido-D-mannopyranose units and contain at least one α(1-3) bond in pairs of units. Depending on the number of units included in the oligosaccharide, the structure can be used as a universal antigen capable of detecting antibodies raised against any Brucella OPS or Y. enterocolitica, or as an M-specific antigen capable of detecting only antibodies raised against Brucella OPS containing α(1-3) bonds (McGiven et al (2015) Journal of Clinical Microbiology). This allowed the user to distinguish for the first time an animal infected with Brucella from an animal that could be infected with either Brucella or O:9 Y. enterocolitica.

Brief description of the invention

The inventors of the present invention have recently identified and characterized another important structural feature of OPS Brucella, which for the first time allows the creation of a vaccine that can be used in the DIVA testing system, which is described in this document.

As further discussed below, the initial work described in this paper is also

- 3 040853 led to the evaluation of the serodiagnostic properties of some synthetic oligosaccharide structures, which would otherwise not be considered. Some of them unexpectedly showed excellent properties in serodiagnostic tests for brucellosis.

According to the first aspect of the present invention, there is provided a molecule containing a chain of seven or more adjacent units of 4,6-dideoxy-4-acylamido-a-mannopyranose, the adjacent units being connected by a C1-C2 or C1-C3 bond, the chain having a reducing end , while the pyranose ring in the chain unit furthest from the reducing end is linked to a cap structure, where the cap structure is not 4,6-dideoxy-4-acylamido-α-D-pyranose linked to the rest of the molecule through its C1, and where the cap structure consists of formula 4,

where R ₄ represents acylamido or its deacylated variant, OH, C1-C ₅ -alkoxy, C1-C5 alkyl or hydroxylated C1-C ₅ -alkyl;

R5 is C1- _C5 alkyl or hydroxylated C1- _C5 alkyl; and R ₆ and R ₇ are independently selected from -H, -CH3 or -CHO.

The second aspect of the present invention relates to a vaccine composition for use as a vaccine against infection of an animal with a Brucella organism, comprising a molecule according to the first aspect of the present invention and a vaccine-carrying entity.

The third aspect of the present invention relates to a method of vaccinating an animal against infection with a Brucella organism, comprising administering to the animal a protective amount of a molecule according to the first aspect of the present invention, or a vaccine composition according to the second aspect of the present invention.

A fourth aspect of the present invention relates to a method for detecting an animal infected with a Brucella organism from a population of animals known to include individuals that have been vaccinated according to the method of the third aspect of the present invention, the method comprising bringing a biological sample obtained from the animal , in contact with the antigen for DIVA, while detection of binding of the antibody to the antigen for DIVA indicates that the sample was obtained from an animal infected with Brucella.

In this aspect, the antigen for DIVA comprises:

structure XII

and/or structure XI

- 4 040853 and/or structure II

The fifth aspect of the present invention relates to a method for detecting antibodies to Brucella in a sample, including bringing the sample into contact with a diagnostic conjugate containing a molecule according to the first aspect of the present invention.

Brief description of graphic materials

Hereinafter, embodiments of the present invention will be described, by way of example only, with reference to FIGS. 1-16 where:

in fig. 1 shows the antibody binding profile, shown as iELISA endpoint titer, of sera from mice vaccinated with TT-dsg-1,2-reKca (structure I) compared to various synthetic oligosaccharide-BSA conjugates (1-2 hexasaccharide=structure IX; 1-3 hexasaccharide=structure VIII; tetrasaccharide=structure VI; trisaccharide=structure V; trisaccharide=structure IV; disaccharide=structure III; monosaccharide=structure II), as well as in comparison with various sLPS antigens (S99 B. abortus; sLPS 16M B. melitensis sLPS 0:9 Y. enterocolitica);

in fig. 2 shows iELISA serological titers of bovine sera using mannose-modified and equivalent unmodified oligoperosamine-BSA conjugates (mono=structure II; C-mono=structure XIII; tri=structure V; C-tri=structure XIV, hexa=structure IX ; C-penta=structure XVIII; positive denotes sera from a known infected animal and negative denotes sera from a known uninfected animal);

in fig. 3 shows the antibody binding profile, shown as iELISA endpoint titer, of sera from mice vaccinated with TT-dsg-1,2-renTa( _He reducing) (structure XXI) compared to various synthetic oligosaccharide conjugates (heptasaccharide=structure XXII; tetrasaccharide=structure VI; disaccharide=structure III) and various sLPS antigens (B. abortus S99; B. melitensis sLPS 16M; Y. enterocolitica sLPS 0:9);

in fig. 4 shows antibody binding, shown as endpoint titer (y-axis) obtained by iELISA, of sera (48 days post-immunization) from two groups of 8 CD1 mice immunized with two types of Brucella TT-OPS conjugate evaluated against different antigens (tetanus toxoid; B. abortus S99 whole cells [A-dominant OPS]; B. melitensis 16M whole cells [M-dominant OPS]; B. suis biovar 2 whole cells [OPS with exclusively a(1->2) linkages sLPS S99 B. abortus [A-dominant OPS] sLPS 16M B. melitensis [M-dominant OPS] 1,2-hexasaccharide=structure IX 1,3-hexasaccharide=structure VIII tetrasaccharide=structure VI trisaccharide=structure XII, disaccharide=structure III); horizontal bars show the median titer, the range of tested titers started from logio2.0;

in fig. 5 shows iELISA results using B. abortus sLPS S99 (x-axis) and conjugated (c) B. abortus S99 OPS antigens (y-axis) (capped conjugation process on OPS antigen) for 20 serum samples collected from a large cattle 45 days after vaccination with S19 B. abortus (open triangles), and 60 samples from cattle from herds infected with field strains of B. abortus (filled diamonds);

in fig. 6 shows mean (solid line) and individual (marker) iELISA results (y-axis) for 12 sera from 12 B. abortus infected cattle and mean (dashed line) ELISA results for 4 sera from 4 uninfected cattle. cattle, according to the results of three different ELISA analyzes (x-axis): one with a hexasaccharide with exclusively 1,2-bonds (hex 1,2=structure IX), one with a trisaccharide with exclusively 1,2-bonds (three 1,2= structure XII) and one with a monosaccharide (mono=structure II);

in fig. 7 shows the mean iELISA result for serum samples from 4 animals experimentally infected with B. abortus strain 544 (A-dominant) at each of the selected time points (3, 7, 16, 24 and 53 weeks post-infection, x-axis), by the results of two ELISA assays: one with a hexasaccharide with exclusively 1,2 bonds (structure IX) and one with a trisaccharide with exclusively 1,2 bonds (structure XII);

in fig. 8 shows iELISA results for serum samples from 17 pigs infected with B. suis biovar 2 (A-dominant, OPS has exclusively 1,2-linkages) designated as infected and 12 randomly selected uninfected pigs designated as random. non-infectious. ELISA analyzes were performed using a trisaccharide exclusively 1,2-linked BSA conjugate (structure XII) (at a coating concentration of 2.5 μg/ml) and using an equal weight mixture (coating concentration of each antigen 1.25 μg/ml). ml, total concentration=2.5 μg/ml) exclusively 1,2-linked trisaccharide-BSA conjugate (structure XII) and M-tetrasaccharide-BSA conjugate (structure VI);

-5 040853 in FIG. 9 shows iELISA results (y-axis) using antigen in the form of B. abortus S99 sLPS for 20 sera from 4 cows experimentally infected with 544 B. abortus (solid lines), each of which was sampled at 3, 7, 16, 24 and 53 weeks after infection (axis

x), and for 20 sera from 4 cows experimentally infected with O:9 Y. enterocolitica, which were also sampled at 3, 7, 16, 24 and 53 weeks post-infection;

in fig. 10 shows iELISA results (y-axis) using an exclusively 1,2-linked hexasaccharide antigen (structure IX) for 20 sera from 4 cows experimentally infected with 544 B. abortus (solid lines), each of which was sampled 3, 7, 16, 24, and 53 weeks post-infection (x-axis), and for 20 sera from 4 cows experimentally infected with O:9 Y. enterocolitica, which were also sampled at 3, 7, 16, 24 and 53 weeks after infection;

in fig. 11 shows iELISA results (y-axis) using an exclusively 1,2-linked trisaccharide antigen (structure XII) for 20 sera from 4 cows experimentally infected with 544 B. abortus (solid lines), each of which was sampled 3, 7, 16, 24, and 53 weeks post-infection (x-axis), and for 20 sera from 4 cows experimentally infected with O:9 Y. enterocolitica, which were also sampled at 3, 7, 16, 24 and 53 weeks after infection;

in fig. 12 shows iELISA results (y-axis) using monosaccharide antigen (structure II) for 20 sera from 4 cows experimentally infected with 544 B. abortus (solid lines), each of which was sampled at 3, 7, 16, 24 and 53 weeks post-infection (x-axis), and for 20 sera from 4 cows experimentally infected with O:9 Y. enterocolitica, also sampled at 3, 7, 16, 24 and 53 weeks post-infection;

fig. 13 is a scatterplot showing iELISA results using B. abortus sLPS S99 (x-axis) compared to iELISA results using an exclusively 1,2-linked trisaccharide conjugate antigen and BSA (structure XII) (y-axis). ), with data points showing results for 29 sera from 29 B. abortus-infected bovines (infected, filled diamonds), 31 sera from 31 Brucella-uninfected bovines, which were false positive in standard serodiagnostic tests for brucellosis (FPSR, open circles), and 20 serum samples from 20 randomly selected uninfected cattle (random non-infectious, crosses);

fig. 14 is a scatterplot showing iELISA results using B. abortus sLPS S99 (x-axis) compared to iELISA results using an equal weight mixture (coating concentration of each antigen 1.25 μg/mL, total concentration=2, 5 µg/mL) 1,2-linked-only trisaccharide-BSA conjugate (structure XII) and M-tetrasaccharide-BSA conjugate (structure VI) (y-axis), with data points showing results for 29 sera samples from 29 infected B abortus bovines (infected, filled diamonds), 31 sera from 31 Brucella-uninfected bovines that were false-positive on standard brucellosis serodiagnostic assays (FPSR, open circles), and 20 sera from 20 randomly selected non-infected bovine animals (case non-infectious, crosses); and fig. 15 is a scatterplot showing the results of iELISA using a trisaccharide exclusively 1,2-bonded BSA conjugate (structure XII) (x-axis) compared to the results of iELISA using an equal weight mixture (coating concentration of each antigen 1, 25 µg/mL, total concentration=2.5 µg/mL) exclusively 1,2-linked trisaccharide-BSA conjugate (structure XII) and M-tetrasaccharide-BSA conjugate (structure VI) (y-axis) with data points show results for 29 sera from 29 B. abortus-infected bovines (infected, filled diamonds), 31 sera from 31 Brucella-uninfected bovines that were false positive in the standard brucellosis serodiagnostic assay (FPSR, open circles) ), and 20 serum samples from 20 randomly selected non-infected cattle (random non-infectious, crosses); and in FIG. 16 shows the results of the consumption of an oxidative reagent (sodium metaperiodate [SMP]) when applied to the RBT antigen; the figure shows a standard curve of known SMP concentration (x-axis) versus optical density (OD) at 405 nm (y-axis), with individual data points shown as black crosses, and OD values of oxidizing reagents obtained at different points from the start oxidation process shown to the right of the x-axis (cell incubation).

Examples

Example 1. Initial work on the development of a possible vaccine candidate.

According to the work disclosed in WO 2014/170681 and in (Ganesh et al (2014) Journal of American Chemical Society 136: 16260-16269) and (McGiven et al (2015) Journal of Clinical Microbiology 53: 1204-1210), it was it has been suggested that it may be possible to develop a chain-formed vaccine

- 6 040853 and 4,6-dideoxy-4-formamido-α-D-mannopyranose units, which are linked exclusively by C1-C2 bonds. This is because, according to the results of observations, the shorter oligosaccharides described in these publications (such as di- or tetrasaccharides), which contain one C ₁ -C ₃ bond and a limited number of C ₁ -C ₂ bonds, with less are likely to bind to antibodies whose production is induced by polysaccharides that are associated exclusively with C ₁ -C ₂ bonds. It has been suggested that then vaccination with a polysaccharide with exclusively C ₁ -C ₂ bonds will distinguish an animal infected with an OPS organism where C 1 -C ₃ bonds are present.

Therefore, initial experiments were carried out in which mice were immunized with a hexasaccharide with exclusively C1- _C2 bonds conjugated to tetanus toxoid via a disuccinimidylglutarate (DSG) linker (structure I). Structure I is called TT-dsg-1,2-hexa.

These constructs were expected to induce antibodies only to the A and C/Y epitopes, but not to the M epitopes, due to the lack of a C1- _C3 bond.

After immunizing mice with TT-sq-1,2-hexa and TT-dsg-1,2-hexa, the sera of the animals were tested for BSA-conjugated 1,2-hexasaccharide (structure IX) and, as expected, showed good answer. Serum was also tested against native bacterial antigens in the form of lipopolysaccharides (LPS) from Brucella abortus, Brucella melitensis and O:9 Yersinia enterocolitica and again good responses were observed.

The sera were then tested with various antigens in the form of synthetic oligosaccharide conjugates, as described previously and shown as structures II, III, IV, V, VI, VII and IX in the following table. 2 (in which, in the third column, S is the 4,6-dideoxy-4-formamido-α-D-mannopyranose unit, S2S is the adjacent units linked by a C1- _C2 bond, and S3S is the adjacent units linked by a C1- _C3 bond ).

Surprisingly, the immunized sera were found to recognize antigens comprising a C1- _C3 glycosidic linkage, although the immunizing antigen lacked a C1- _C3 linkage (FIG. 1).

The unexpected antibody response even to a monosaccharide antigen (structure II) led the inventors of the present invention to the conclusion that the development of the originally proposed vaccine for DIVA would not be successful, since any oligosaccharide, regardless of the presence or absence of C1- _C3 bonds, could induce at least such a response.

- 7 040853

Table 2. Synthetic conjugates of oligosaccharides and BSA

Structure number Scheme sugars / bonds Structure Monosaccharide II S He OHCHN-vV O “V 9 n BSA H cgh> Disaccharide (linked by Ci-C ₃ bond) III S3S He OHCHN— buthh he | onsnm-H | o o-A__CHH. n 1____ Trisaccharide IV S2S3S he ONSNK-UCH |o noChH O OHCHN-4-IO but HH, he |OHCHN-4d0 ⁰ Trisaccharide V S3S2S He OHCHN—H-lq BUT hh, he YUNSNY-CH O o —_CHCH OHCHN-CH O ”4 Ϊ” o---V Oh Oh tetrasaccharide VI S2S3S2 S He OHCHN-Ά O noX-M ABOUT OHCHN-Av-O noHA he 1 OHCHnA^ \o----VX-XD O OHCHN—A|O noH-C n J___1 Oh Oh Pentasaccharide VII S2S3S2 S2S \ ^{is he} OHCHN-tM? no-HH _h he OHCHN-l-4-Ιθ o-HH O ohchn-4-IQ but-HH o OHCHN-H- O noChH Ί ⁰ n ^{BSA n} p Ο^Ό Hexasaccharide VIII S2S3S2 S2S2S . OH ΟΗΟΗνΆ|0 noX^L . o OHCHN-vM ⁰ , HO-XM _h o* ^OHCHN-^^ O ΟΗΟΗΝ-Ά|9 noXA Q OHCHN-uch|0 noX-'M OHCHN-AlQ Ch « . oh oh

- 8 040853

Hexasaccharide IX S2S2S2 S2S2S OHCHN-v4°> NoA"-kD . ABOUT OHCHN-tA-o noA-M O OHOHN'vVQ O OHCHN-γν O BUT-A-KD O OHCHN—γ%|0 butA-KD ABOUT OHCHN-V^lq but A-*) oh oh Trisaccharide (associated with using DSG) X S2S3S he ΟΗΟΗΝ'-γΑΙΟ ho-X-Vl o OHCHN— noL—Χή he OHCHN-I—uX|o oA—*3 Η ⁿ o o Disaccharide (linked by Ci-C ₂ bond) XI S2S he OHCHN-vX-Jo ΗΟ-Α-ΑΔ O OHCHN- θ but-V-Hell 9 Η __/ ^{BSA H} JI 0 0 Trisaccharide (bound exclusively by C1- _C2 bonds) XII S2S2S . OH OHCHN-vX-IO but A-Ad o OHCHN-V-Vlo HoX-Αή O OIIGHN^XO ^H0 ^ 4 Η Η Μ Ο 0

Methods that were used for example 1.

Animal: Female CD1 mice (Charles River, Canada) aged 6-8 weeks were used to study the immune response. All procedures and experiments involving animals were performed using a protocol approved by the Animal Care Committee of the Department of Biological Sciences of the University of Alberta. The protocol was approved in accordance with Canadian Animal Care Council (CCAC) guidelines.

Antigen: All synthetic oligosaccharide antigens were prepared as previously described (Ganesh et al (2014) Journal of the American Chemical Society 136:16260-16269) or in the Appendix below. For animal experiments, a hexasaccharide of six perosamine units, all linked via 1,2-glycosidic bonds, was conjugated to tetanus toxoid (TT) using a dsg linker (disuccinimidylglutarate) to give a molecule with structure I described above (also referred to as TT- DSG-1,2-reKca). The hexasaccharide was synthesized with an amino-terminated linker at the reducing end (Ganesh et al (2014) Journal of the American Chemical Society 136:16260-16269). A mixture of hexasaccharide and DSG (15 eq.) in DMF and 0.1 Μ PBS buffer (4:1, 0.5 ml) was stirred at room temperature for 6 h. The reaction mixture was concentrated in vacuo and the residue was washed with EtOAc 10 times to remove excess DSG. The resulting solid was dried under vacuum for 1 hour to give the DSG-activated oligosaccharide. Activated hexasaccharide (0.518 µmol) was added to a solution of tetanus toxoid (0.025 µmol) in 0.5 M borate buffer pH 9 and stirred slowly at 21°C for 3 days. The reaction mixture was then washed with PBS buffer, filtered through an ultrafine filtration tube (10000 MWCO, 4x10 ml) and the resulting tetanus toxoid conjugate was stored in PBS buffer. Mass spectrometric analysis of MALDI-TOF showed that the conjugate had an average of 11.7 hexasaccharides per molecule of tetanus toxoid.

For immune response screening by ELISA, the same hexasaccharide was conjugated to another carrier protein, namely bovine serum albumin (BSA), using square chemistry (Ganesh et al (2014) Journal of the American Chemical Society 136:16260-16269) as previously described (eg WO 2014/170681) to give structure IX. In addition, immune responses were also screened for 9040853 using various synthetic oligosaccharides (structures II-VI and structures

VIII from table. 1). Various native sLPS from Brucella abortus, Brucella melitensis and

Yersinia enterocolitica.

Vaccine formulation: At the very beginning of the immunization, fresh alum meal was prepared according to the published protocol (Lipinski et al (2012) Vaccine 30:6263-6269). Briefly, solutions of 0.2 molar KAl(SO ₄ ) ₂ ,12H ₂ O and 1.0 molar NaHCO ₃ were prepared separately and sterilized by filtration. Then 10 ml of the second solution (bicarbonate solution) was quickly added to 20 ml of the first solution with vigorous shaking. To avoid loss of material due to foaming from violent outgassing, the mixing step was carried out in a 200 ml beaker. The resulting precipitate of alum flour was washed with PBS (which had previously been filtered and sterilized in an autoclave) and precipitated by centrifugation at 4000 g for 7 min. This wash cycle was continued until the pH of the supernatant was identical to PBS (pH 7.3). Finally, alum flour was suspended in PBS at a concentration of 50 mg/ml, thimerosal (0.01% w/v) was added and the mixture was stored at 4°C.

The alum flour was mixed with the TT conjugate in a 5:1 weight ratio and the mixture was left to shake overnight before administration to the animals.

Immunization: Animals were immunized three times with an interval of 21 days. A total volume of 250 μl including 12 μg of TT-dsg-1,2-hexa (equivalent to 1 μg of 1,2-hexasaccharide) was injected into each mouse, of which 150 μl was injected intraperitoneally and the remaining 100 μl was injected subcutaneously. Before the start of immunization, a preliminary blood sampling was performed. Animals were sacrificed on the 10th day after the last injection and the final blood sampling was performed.

Serum treatment: After collection, mouse blood was incubated at 37° C. for one hour, then centrifuged at 1500 g for 10 minutes. The clarified serum from the upper fraction was collected and stored at -20°C until use.

Immunoassays: Antibody levels in mouse serum were examined by enzyme-linked immunosorbent assay (ELISA). The published protocol (Bundle et al. (2014) Bioconjugate Chemistry 25:685-697) was followed with minor modification. Briefly, polystyrene microtiter plates were incubated with coated antigen (1 μg/ml, 100 μl/well) at 4°C overnight, then washed (5x) with PBST (0.05% Tween-20 in phosphate buffered saline, PBS). Then serially ^10-fold dilutions of mouse serum (100 μl/well) were added to the coated well. The initial dilution for sera was 1:100. After incubation at room temperature for 2 hours, the plates were washed (5x) with PBST. The plate was then incubated with 100 µl/well diluted 1:5000 goat anti-mouse IgG labeled with HRPO (KPL, 1.0 mg/ml stock solution) for 30 min at room temperature, then washed (5x) with PBST. The peroxidase substrate 3,3',5,5'-tetramethylbenzidine (TMB) was added with H ₂ O ₂ . After 15 minutes, the reaction was quenched by the addition of phosphoric acid (1 M, 100 μl/well). Plates were read at 450 nm and data was processed using Origin software. All sera were diluted with 0.1% BSA in PBST. The end point dilution (x ₀ ) was recorded as the serum dilution giving an absorbance of 0.2 above background, and the serum titer was calculated as the reciprocal of x ₀ . All data were processed using Origin 9 software and GraphPad Prism.

Example 2 Study of a possible epitope at the non-reducing end of the 4,6-dideoxy-4-formamido-a-O-mannopyranose chain.

Since the present inventors have observed binding of antibodies raised against structure I even to a monosaccharide antigen (structure II), it has been suggested that the terminal sugar provided an epitope for antibody binding (herein referred to as the terminal epitope). To investigate whether the binding potential of perosamine alone (4,6-dideoxy-4-formamido-a-O-mannopyranose) depends on specific structural features that only terminal perosamine possessed (for example, hydroxyls located on both C2 and C3) produced various synthetic oligosaccharides containing a cap structure at the non-reducing end, which are shown in Table 1 below. 3. Some of the oligosaccharides were coupled to BSA by square chemistry and some by DSG as described above.

- 10 040853

Table 3. Additional synthetic oligosaccharide-BSA conjugates providing cap structures on terminal perosamine

Structure number Scheme sugars / bonds Structure Mannose-related monosaccharide XIII S He nodtso noh-ch he OHCHN 440 O He Mannose-related trisaccharide Modified methoxy disaccharide Disaccharide modified with two methoxy groups Modified methoxy group trisaccharide Mannose-related pentasaccharide XIV XV XVI XVII XVIII S3S2S S3S S3S S2S3S S2S2S2 S2S he / he not-M ⁰ noH- o ΟΗΟΗΝ-ΐΧ|θ noX-Xi he OHCHN GTX O O-XXX o OHCHN-D|O ΗθΧ-Χή _o n J, I O O O Me OHCHN—νΧ|θ IOL— Xi OH ONSNM-UyU1°y 1 9 Η ⁿ O O ome OHCHN—D-10 MvoCLH. he OHCHnJ—uChd o o OM OHCHN—iAJ o noX-M o OHCHN—A-|o but UHA he OHCHN o o he 1 he noDcho noX-Χή o ohchn-vM o. noX-Xή o OHCHN Ύ- θ, but J- o OHCHN-yX O noX-Xd. about OHCHN-dX|°. noX-Xi. o OHCHN-41°, noKh-KhL Ί 9 η llA ld air II Ο ^Ζ Ό

Sera from field-infected cattle were tested with a selection of the above structures (FIG. 2). This demonstrated that modification of the non-reducing end of the sugar chain did affect the serological reactivity of the oligosaccharide antigen. This effect was stronger for shorter oligosaccharides, suggesting that as the length of the linear epitope increased, the effect of antibody binding with the loss of the terminal epitope proportionally decreased.

The positive data points in FIG. 2 is the average of 6 sera from infected animals that were positive for conventional serological assays. Negative data points represent the mean of 2 serum samples that were negative in these assays. In positive samples, the average result for monosaccharide antigen (structure II) at 1/100 dilution of serum was approximately 60%. The equivalent result for the modified monosaccharide (structure XIII) was about 5%. The titer reduction was at least 64-fold and there was no difference between negative and positive sera using the modified monoperosamine (structure XIII).

- 11 040853

In contrast, when using unmodified monoperosamine (structure II), a difference was observed. Thus, it appears that none of the antibodies that bind to structure II could bind the modified version, structure XIII. The present inventors concluded that the antibody repertoire against OPS Brucella in these samples included antibodies that bound to a single 4,6-dideoxy-4-formamido-α-D-mannopyranose unit, but this binding was only possible when sugar link was in the terminal position. The ability of these antibodies to bind was severely limited once 4,6-dideoxy-4-formamido-α-D-mannopyranose was no longer available as a terminal sugar unit, but was only available within the linear structure.

A similar pattern occurred when assessing trisaccharides, although the contrast between the modified (structure XIV) and unmodified (structure V) antigens was not so pronounced (4-8-fold decrease in titer). Presumably, a less pronounced contrast reflected an increased ability of the trisaccharide in structure XIV to act as a linear antigen. This pattern was also observed with 1-2-hexasaccharide (structure IX) and modified 1-2-pentasaccharide (structure XVIII).

Based on these data, the present inventors concluded that there was a significant subset of anti-OPS antibodies in which antigen binding to short oligosaccharides was significantly affected by the presence or absence of a terminal 4,6-dideoxy-4formamido-α-D-mannopyranose unit.

Similar experiments were performed using the same sera from B. abortus infected cattle using different oligosaccharide antigens modified by replacing the C2-OH (hydroxyl) group on the terminal perosamine with an -OMe group. The overall results are shown below in table. 4.

In the table below. 5 shows more serological data obtained by inoculating synthetic antigens into sera from B. abortus-infected cattle (n=20), infected samples, and from sera from uninfected cattle (n=20), uninfected samples. Mannose-linked monosaccharide (structure XIII) and mannose-linked trisaccharide (structure XIV) antigens have poor diagnostic properties (low AUC values [area under the dose-response curve]) because they do not effectively distinguish between infected and non-infected specimens. Structure XIII has particularly poor properties when compared to the exceptional and completely unexpected diagnostic properties of an unmodified (ie, capless) monosaccharide (structure II).

Table 4. Results of 6 serum samples from cattle infected with B. abortus

Antigen Strong position Weak position Negative monosaccharide squarate (structure II) 6 0 0 1-3-disaccharide square (structure III) 6 0 0 tl-2-trisaccharide square (structure IV) 6 0 0 tl-2-trisaccharide dsg (structure X) 6 0 0 11-3-trisaccharide square (structure V) 6 0 0 exclusively 1-2- hexasaccharide square (structure IX) 6 0 0 associated with mannose exclusively 1-2- pentasaccharide squarate (structure XVIII) 5 1 0 mannose-related tl-3 trisaccharide squarate (structure XIV) 2 2 2 OME modified tl-2 trisaccharide dsg (structure XVII) 2 2 2 OME-modified 1-3-disaccharide dsg (structure XV) 1 2 3 mannose-bound monosaccharide squarate (structure XIV) 0 1 (very weak) 5

Based on the results shown in table. 5, the present inventors concluded that even the inclusion of one OMe group in the C2 terminal monosaccharide was sufficient to

- 12 040853 cancel most of the response with the production of antibodies. This supported the idea that the terminal 4,6-dideoxy-4-formamido-α-D-mannopyranose was a specific structure distinct, in the context of antibody recognition, from the linear polymer of 4,6-dideoxy-4-formamido-α- D-mannopyranose units.

Table 5. Diagnostic characteristics (YImax with DSn, DSp and AUC) for specimens from animals that are positive in culture for B. abortus compared to specimens from random animals not infected in the field

Antigen Yim ah DSn Dsp AUC Mannose-related monosaccharide (structure XIII) 26.16 75 51.16 0.5558 Mannose-related trisaccharide (structure XIV) 53.02 60 93.02 0.7733 Mannose-related pentasaccharide (structure XVIII) 78.72 95 83.72 0.9605 Monosaccharide (structure II) 83.02 90 93.02 0.9663 Hexasaccharide (structure IX) 95 95 100 0.9942 Pentasaccharide (structure VII) 100 100 100 1.00 Nonasaccharide (structure XIX, presented below) 100 100 100 1.00 Disaccharide (structure III) 100 100 100 1.00 Tetrasaccharide (structure VI) 100 100 100 1.00

(DSn=diagnostic sensitivity (%);

DSP=diagnostic specificity (%);

AUC=area under the curve (ROC);

ROC=plot of sensitivity versus false positive rate;

YI=Youden index (DSn+DSp-100);

YI _max = maximum value of YI that can be reached with +/- limit value variation.)

Therefore, the inventors hypothesized that the response to modified oligosaccharides from serum from infected animals may be similar to the response to unmodified oligosaccharides from serum from animals immunized with antigens that do not have an apical epitope (i.e., there is no terminal 4,6-dideoxy- 4-formamido-α-D-mannopyranose). In the first case, only antibodies to linear structures will bind and the response will be low (very low in the case of short oligosaccharides). In the second case, there will be no antibodies available to bind to the apex, and therefore the only response observed will be associated with antibodies to the linear structure. The response with the production of these antibodies to short oligosaccharides will also be low.

Methods that were used for example 2.

Antigen: Perosamine oligosaccharides were conjugated to tetanus toxoid (TT) using a dsg linker (disuccinimidylglutarate) or using square chemistry as described above.

Bovine serology studies: Bovine serum antibody levels were studied by enzyme-linked immunosorbent assay (ELISA) as previously described (McGiven et al (2015) Journal of Clinical Microbiology 53:1204-1210).

- 13 040853

Example 3 Oxidation of the sugar at the terminal end of the OPS to disrupt the terminal epitope.

The present inventors have adapted the disclosure of Stefanetti et al. (Stefanetti et al. (2014) Vaccine 32: 6122-6129) to disrupt the terminal sugar structure in OPS Brucella. OPS Salmonella typhimurium was moderately oxidized with sodium metaperiodate by these collaborators. This opens the rhamnose ring to form aldehydes, which can then be conjugated to amines on CRM ₁₉₇ (genetically detoxified diphtherotoxin) via reductive amination. The rhamnose sugar in this method is an internal sugar and not a terminal sugar. Therefore, oxidation is possible because the polymer is attached via C1 and _C4 so that the cis-adjacent hydroxyl groups on _C2 and _C3 are available for oxidation. In the case of Brucella, the perosamines have a D-rhamnose backbone (like D-mannose, but without the OH on C6), but since each non-terminal rhamnose in OPS Brucella is linked to its neighbor at the terminal end via either C2 or C3, the terminal group is the only rhamnose with cis-neighboring hydroxyl groups on C2 and C3.

Therefore, a similar method can be used on OPS Brucella to create the terminal structure shown below (XX structure).

The methods that were used for example 3.

OPS Brucella (from B. abortus S99 and B. suis biovar 2 [strain Thomsen]) was purified by hot phenol extraction (Westphal et al (1952) Uber die Extraction von Bakterien mit Phenol/Wasser. Z. Naturforsch. 7:148-155) followed by moderate acid hydrolysis and size exclusion chromatography (Meikle et al. (1989) Infect Immun 57:2820-2828). OPS for TT conjugation was oxidized at 2 mg/ml in 10 mM sodium metaperiodate (SMP) and 50 mM sodium acetate buffer (pH 5.5) for 1 hour in the dark. This was enough to oxidize the neighboring hydroxyl groups of the diol on the 2nd and 3rd carbon atoms of the terminal sugar. Residual SMP was removed by desalting with a PD10 column (Sephadex-G25 column) according to the manufacturer's instructions (GE Healthcare). An appropriate volume of elution buffer allowed the OPS to be eluted.

Example 4 Vaccination with capped hexasaccharide.

In view of the apparent importance of the apical epitope in antibody production, a heptasaccharide linked to TT via a non-reducing terminal end of the sugar chain (TT-dsg-1,2-hepta( _{non-reducing infusion )} ) was prepared (structure XXI). This conjugation method disrupts the apical epitope as there is no available terminal 4,6-dideoxy-4-formamido-α-D-mannopyranose. The structure is structure XXI shown below.

- 14 040853

Therefore, it is a capped hexasaccharide that is not 4,6dideoxy-4-formamido-D-mannopyranose.

Mice were immunized with this conjugate and sera were assessed by iELISA for various antigens. In FIG. 3 shows the results. In FIG. 1 shows comparative results from example 1 for animals vaccinated with TT-dsg-1,2-hexa (structure I). This indicated that vaccination with structure XXI with a disrupted apical epitope induced the production of antibodies with significantly reduced binding affinity for the proposed diagnostic conjugate antigens (di- and tetrasaccharides, structures III and VI, respectively). However, there was still a reaction against disaccharide and tetrasaccharide antigens, indicating that TT-dsg-1,2-hepta( _non- reducing) (structure XXI) was also not suitable for use as a vaccine in the test system. DIVA.

The methods that were used for example 4.

Animals; vaccine formulation; immunization; whey processing; immunoassay: all as described above for example 1.

Antigen: 4,6-dideoxy-4-formamido-a-P-mannopyranose hexasaccharide was prepared according to previously described methods (Eis & Ganem (1988) Carbohydrate Research 176:316-323).

For immune response screening by ELISA, the same heptasaccharide was conjugated to bovine serum albumin (BSA) as previously described (eg WO 2014/170681). The resulting heptasaccharide (actually a capped hexasaccharide) had the following structure:

Additional synthesis and conjugation methods used to prepare TT-DSG-1,2hep(non-reducing) (structures XXI) and BSA-DSG-1,2-ΓΓΠTa( _non- reducing) (structures XXII) can be found in the appendix below.

- 15 040853

In addition, immune responses were also screened using various synthetic oligosaccharides (structures III, VI and IX) as well as various natural bacterial cell wall antigens from Brucella abortus and O:9 Yersinia enterocolitica.

Example 5 Vaccination with Conjugated Top Polysaccharide.

The present inventors then attempted vaccination using a much longer polysaccharide conjugated to a carrier protein through a non-reducing apical end in an additional attempt to obtain a vaccine that could be used in the DIVA testing system. The aim was to obtain a vaccine molecule that would lead to the production of antibodies that would not bind to the proposed diagnostic conjugate antigens (di- and tetrasaccharides, structures III and VI, respectively). As described in WO 2014/170681, these antigens are already useful for distinguishing between animals infected with Brucella and animals that are not infected or infected with O:9 Yer sinia enterocolitica (or strains of Brucella that have OPS without α1,3-glycosidic bonds ).

Mice were immunized as follows with OPS from B. abortus S99 (which has approximately 2% α1,3-linkages) and OPS from Thomsen biovar 2 strain of B. suis (a polysaccharide strain with exclusively α1,2-linkages), both of which are conjugated to TT via a terminal sugar. Therefore, the structure derived from S99 B. abortus was structure XXIII below, in which conjugation to TT is through C ₃ , or a related structure in which conjugation to TT is through C ₂ or both C ₂ and C ₃ .

The structure obtained from B. suis biovar 2 was structure XXIV, which is shown below. Again, structure XXIV shows TT conjugation via C ₃ , but the structure derived from B. suis biovar 2 may be a related structure in which TT conjugation is via C ₂ or both C ₂ and From ₃ .

- 16 040853

The final serum of the animals was then tested for bacterial lipopolysaccharide (LPS) antigens from B. abortus S99 and Brucella melitensis strain 16M (approximately 20% α1,3 ties), whole cell antigens from B. abortus S99, B. melitensis 16M and biovar 2B. suis, as well as for tetanus toxoid. The results are shown in FIG. 4. Serums were also tested against synthetic antigens having structure IX (1,2-hexasaccharide), structure VIII (1,3 hexasaccharide), structure VI (tetrasaccharide), structure XII (trisaccharide with only 1,2 bonds) and structure III (disaccharide ).

A trisaccharide (structure XII) was included in the analysis so that, together with an exclusively 1,2-linked hexasaccharide (structure IX), it could be assessed how the length of an exclusively 1,2-linked oligosaccharide affects the binding of antibodies induced by glycoconjugate immunogens. with structures XXIII and XXIV. When testing serum at a dilution of 1/100, there was no response to the trisaccharide (structure XII). The results showed at least a tenfold difference in mean titer between hexasaccharide (structure IX) and trisaccharide (structure XIII). The magnitude of this difference was greater than expected given the fact that the trisaccharide antigen with only 1,2-linkages is considered the epitope for the anti-C/Y antibody (Table 1) and that these antibodies were thought likely to be responsible. for the observed cross-reactions between A-dominant (eg S99 B. abortus) and M-dominant (eg 16M B. melitensis) Brucella antigen serotypes.

To demonstrate that the antigens are capable of detecting antibodies induced by Brucella infection, these trisaccharide (structure XIII) and hexasaccharide antigens (structure IX) were evaluated against sera from naturally or experimentally infected animals. In FIG. 6 shows the ELISA results for both antigens when tested against sera from 12 B. abortus naturally infected cattle and 4 B. abortus non-infected cattle. They show that both antigens are able to detect all sera from infected animals without reacting to sera from non-infected animals, indicating that they are suitable as antigens for DIVA. In addition, they show that the difference in results between the two antigens was very small; the average results were 138.6% for the hexasaccharide (structure IX) compared to 125.9% for the trisaccharide (structure XIII). The figure also shows the results for the monosaccharide (112.2%) (structure II). The magnitude of these results was unexpected.

The results obtained from immunizations of mice with structures XXIII and XXIV suggest that antibodies to the apical epitope, rather than antibodies to the linear epitope, are the main types that bind to short antigens with exclusively 1,2-linkages.

- 17 040853

Prior to this assessment, it was believed that the lack of binding of antibodies induced by these structures to shorter oligosaccharides containing a 1,3-bond was predominantly due to the fact that the 1,3-bond prevents antibodies from binding to linear sequences of 1,2-bonds; it was assumed that the apical epitope plays an important but lesser role. The data that have now been obtained using a trisaccharide with only 1,2-linkages demonstrate that the apical epitope plays a more important role in serodiagnosis than previously thought.

Two exclusively 1,2-linked antigens (structures VIII and XII, hexasaccharide and trisaccharide, respectively) were also tested on sera from four bovine animals experimentally infected with Brucella abortus strain 544 (A-dominant strain); samples were taken at 3, 7, 16, 24 and 53 weeks after infection. The mean titers from these samples are shown in FIG. 7. For four of the five sampling dates, the average results for the trisaccharide were higher than for the hexasaccharide, and the results on the other date were very close. These results demonstrate that antigens in the form of trisaccharides and hexasaccharides with only 1,2-linkages have very similar and very good sensitivities when applied to sera from Brucella-infected cattle under experimental and natural conditions. Therefore, the at least 10-fold difference observed between these two antigens when introduced into sera from mice immunized with structure XXIII is most likely due to the nature of the antibodies induced by this immunization, and not to any inherent differences in diagnostic sensitivity. these two antigens (because they are equal). Incorporation of the cap structure into the OPS through the modification process prevented the formation of antibodies to the apical epitope of the generated OPS. Consequently, antibodies were produced only against linear epitopes; the longer length of the hexasaccharide antigen allowed more of these antibodies to bind, while the shorter length of the trisaccharide did not. These results support the conclusion that much of the sensitivity of an antigen in the form of a trisaccharide with only 1,2-linkages depends on the detection of antibodies to the apical epitope produced at the time of infection. These differences make an exclusively 1,2-linked trisaccharide (structure XII) an effective diagnostic antigen for DIVA.

For the same reasons, an antigen in the form of a disaccharide with exclusively 1,2-linkages (structure XI) is also an effective diagnostic tool for DIVA. Clearly, it will not bind antibodies induced by a molecule containing a cap structure as described herein, such as structures XXIII and XXIV. This is supported by the diagnostic data shown for the monosaccharide (structure II) in FIG. 6 and in table. 5 (DSn=90%, DSP=93.02%). It can be seen from them that even this small antigen has an unexpectedly high diagnostic sensitivity and specificity.

To further demonstrate its usefulness as an antigen for DIVA, the trisaccharide antigen with exclusively 1,2-linkages (structure XII) was used to detect antibodies to OPS Brucella in sera from 17 B. suis biovar 2-infected pigs and in sera from 12 pigs that have not been infected with Brucella. These samples were also tested with an equal mixture (by weight of conjugate diagnostic antigen) of a tetrasaccharide acting as a specific M antigen (structure VI) and a trisaccharide conjugate with exclusively 1,2-bonds (structure XII). These results are presented in the scatter plot in FIG. 8. It demonstrates that trisaccharide (structure XII) detected all samples from 17 infected pigs and did not react to any of the samples from 12 uninfected pigs. The mixed antigen preparation showed almost identical results. Biovar 2 OPS does not contain 1,3-bonds (Zacchus et al. (2015) PLoS One 8, e53941), so it would not be expected that the antibodies produced against it would bind so well to oligosaccharide conjugates that act as specific M -antigens such as tetrasaccharide (structure VI) and disaccharide (structure III). However, these antibodies bound well to the trisaccharide with exclusively 1,2-bonds (structure XII); this was the case when it was used alone or in combination with a tetrasaccharide (structure VI). These results demonstrate that the DIVA vaccine and diagnostic concept described herein can also be applied against infection with B. suis, including infection with B. suis biovar 2.

From here it can be seen that vaccination with OPS conjugated to a carrier protein through a non-reducing terminal end caused the production of antibodies capable of binding to bacterial sLPS antigens to whole Brucella cells, with the antigen in the form of a hexasaccharide with exclusively a1,2-linkages (structure IX ), as well as with a universal antigen (eg as described in WO 2014/170681; structure VIII). However, binding to shorter antigens (structures III, IV, V, VI and XII) was not observed.

Comparing the results of this example with those of example 4, the authors of the present invention concluded that a longer polymer of at least seven 4,6-dideoxy-4-formamido-α-D-mannopyranose units is required to create a vaccine for DIVA. Antibodies that are raised against such a polysaccharide and do not have a terminal apex epitope (disturbed by conjugation with the vaccine-carrying protein through the non-reducing end of the polymer) were not detected by short antigenic structures having structures III, IV, V, VI and XII, so that these structures can be used as DIVA agents to distinguish an animal that has been vaccinated with a modified polysaccharide from an animal that has been infected with Brucella.

The methods that were used for example 5.

Obtaining oxidized and TT-conjugated OPS:

OPS Brucella was purified by hot phenol extraction (Westphal et al (1952) Z. Naturforsch. 7:148-155) followed by mild acid hydrolysis and size exclusion chromatography (Meikle et al. (1989) Infect Immun 57:2820-2828). OPS for TT conjugation was oxidized at 2 mg/ml in 10 mM sodium metaperiodate (SMP) and 50 mM sodium acetate buffer (pH 5.5) for 1 hour in the dark. This was enough to oxidize the neighboring hydroxyl groups of the diol on the 2nd and 3rd carbon atoms of the terminal sugar. Residual SMP was removed by desalting with a PD10 column (Sephadex-G25 column) according to the manufacturer's instructions (GE Healthcare). An appropriate volume of elution buffer allowed the OPS to be eluted.

The oxidized OPS was then subjected to reductive amination. Before removing salts into water using a Sephadex G-25 column, oxidized OPS was incubated in PBS at final concentrations of 5 mg/ml OPS and 0.5 M ammonium chloride, as well as 0.1 M sodium cyanoborohydride at 37°C for 24 h, and then lyophilized.

OPS was then incubated at 5 mg/ml with 5 mg/ml DSG in PBS for 45 minutes on a rotary shaker before salts were removed with a Zeba 40 kDa column in fresh PBS. The OPS-DSG samples were then incubated with tetanus toxoid (TT) at final concentrations of approximately 2.5 and 0.5 mg/mL, respectively. Incubation was carried out for 2 h at room temperature (in the dark) on a rotary shaker. Glycine was then added at a final concentration of 2 mg/mL and incubated for an additional 15 min. The samples were then subjected to fractionation using SEC-HPLC to separate the glycoconjugate from the non-included OPS. Binding of glycoconjugates to anti-Brucella antibodies was confirmed by SDS-PAGE with silver staining and Western blotting.

Animals and Immunization: Three groups of 8 female CD1 mice were used at 7 weeks of age at the time of pre-bleeding. A preliminary blood sampling (100 μl) was taken from each mouse (from the tail vein) to obtain serum, from which the initial antibody titer to native and putative antigens for DIVA was determined. The antibody titer was assessed using indirect ELISA analyses.

Two days later, mice were immunized with 5 μg of the designated glycoconjugate antigen suspended in physiological PBS without adjuvant. The dose was administered subcutaneously in a volume of 100 μl. 19 days after immunization, a blood sample of 100 μl of blood was taken from each mouse via the tail vein. After another 2 days (21 days after the 1st immunization), the mice were immunized for the 2nd time with the same antigen, composition, dose, volume and in the same way as in the 1st immunization. 33 days after the 1st immunization, a 100 μl blood sample was taken from the mice via the tail vein. After another 2 days (35 days after the 1st immunization), mice were immunized for the 3rd time with the same antigen, composition, dose, volume and in the same way as in the 1st immunization. Two weeks later (49 days after the first immunization), the mice were painlessly sacrificed, then dissected to extract blood from the thoracic cavity after cutting the aorta.

Immunoassays: Smooth LPS antigens from S99 B. abortus and 16M B. melitensis were diluted to 0.6 μg/ml and TT was diluted to 2.5 μg/ml in carbonate buffer (Sigma). B. abortus S99 whole cell antigens, B. melitensis 16M and B. suis biovar 2 (Thomsen) were diluted to 15.6 μg/ml in carbonate buffer (Sigma). Synthetic antigens (structures III, VI, VII, VIII) were diluted to 2.5 μg/ml in carbonate buffer (Sigma).

100 μl per well of each antigen was loaded into standard ELISA binding plates (Nunc). The plates were incubated overnight at 4-8°C, then washed four times with PBS-Tween at 200 μl per well and blotted on filter paper.

Mouse sera were diluted in logarithmic dilutions in the amount of 1/100, 1/316.22, 1/1000, 1/3162.27, 1/10000, 1/31622.7, 1/100000, 1/316227, 1/1000000 and 1/3162270 in casein buffer and 100 μl per well were added to antigen coated plates. As a control, the BM40 monoclonal antibody was diluted to 5 μg/ml in casein buffer (Sigma) and plated at 100 μl per well. Also included as controls were 100 μl per well of positive control sera, mouse sera from mice immunized with hexasaccharide structure I, and negative control sera from normal (non-immunized) mice.

The plates were incubated for 30 min at room temperature on a stirrer at 120 rpm, then washed four times with PBS-Tween at 200 μl per well and blotted on filter paper. HRP-conjugated anti-mouse immunoglobulins (Dako) were diluted 1:1000 in casein buffer (Sigma) and plated at 100 μl/well. Plates were incubated for 60 min for synthetic antigens and for 30 min for sLPS and whole cell antigens at room temperature on a 19 040853 stirrer at 120 rpm, then washed four times with PBS-Tween at 200 µl per well and blotted on filter paper. Substrate buffer (pH 4.0) (Fluka) with diammonium salt of 2,2'-azino-bis(3-ethylbenzothiazoline-6-sulfonic acid) (ABTS) (Sigma) and 3% hydrogen peroxide (Sigma) was added to the plates according to 100 µl per well and incubated at room temperature for 20 min. The reaction was retarded with 0.1 M sodium azide, 100 μl per well, and the plates were read in 405 nm absorbance mode. Data were calculated as the mean minus background for duplicate wells as a percentage of BM40 control wells tested with disaccharide structure III as it was added to each plate tested.

Optical densities (OD) for each sample and dilution were corrected for background by subtracting the OD of control wells that were not supplemented with serum but otherwise subjected to the same treatment as described above. Quantitative data for the samples were then normalized by expressing the OD as a percentage of the positive control. End titers were calculated (using GraphPad Prism 6) as the dilution at which the signal (expressed as a percentage of the positive control) was equal to the positive/negative threshold. This threshold was calculated as the mean of the pre-bleed samples plus 1.96 times the standard deviation of the pre-bleed samples.

iELISA assays in bovine and porcine sera: for ELISA, oligosaccharide-BSA conjugates (structures II, VI, IX, XII) were passively immobilized on the surface of standard polystyrene ELISA plates by incubation overnight in carbonate buffer at 4°C in concentration 2.5 µg/ml (1.25 µg each for mixed antigen coating), 100 µl/well. The plates were washed 4 times with 200 μl/well PBST (PBS containing 0.05% (v/v) Tween 20), blotted dry. The sera were diluted 1/50 in buffer (in duplicate) and added 100 μl per well. The plates were incubated for 30 minutes at room temperature at 160 rpm, after which they were washed and blotted dry as described above. In the case of bovine sera, an HRP-conjugated murine anti-bovine IgG1 antibody conjugate was used. In the case of porcine sera, HRP-conjugated recombinant protein A/G was used. The conjugates were diluted to working concentration in buffer and the plates were incubated, washed and blotted dry as in the case of the serum incubation step. Plates were then developed with ABTS (diammonium salt of 2,2'-azino-bis(3-ethylbenzothiazoline-6-sulfonic acid)) and hydrogen peroxide as substrate for 10-15 min, stopped with 0.4 mM sodium azide and read at a wavelength of 405 nm. Optical density for duplicates were averaged and background OD subtracted (buffer only instead of sera). This value was then expressed as a percentage of a standard positive control serum from a B. abortus-infected cow (when testing bovine serum) or a positive control serum from a B. suis-infected pig (when testing porcine samples). In each case, a negative control sample was always run to ensure data quality.

The same ELISA method was used to test using the sLPS antigen. sLPS was diluted to working concentration and passively applied to standard polystyrene ELISA plates as described for oligosaccharide conjugates. The rest of the procedure was carried out as described for oligosaccharide conjugates.

Several groups of field sera were evaluated by iELISA using the antigens described above. Specific sample numbers are described above. All specimens classified as infected were obtained from animals that were confirmed to be infected with bacteriological culture of B. abortus (in the case of cattle) or B. suis (in the case of pigs), from tissues obtained from the animals themselves (most samples of cattle bovine and all porcine samples), or the animals were serologically positive for brucellosis (as determined by conventional serology) and were members of a herd that was confirmed by bacterial culture to be infected with B. abortus. Randomly collected samples from animals (cattle and pigs) that were not infected with Brucella have been collected in the UK since 2007.

Oligosaccharide-BSA conjugates (structures II, VI, IX and XII), B. abortus sLPS and modified OPS (i.e. capped) B. abortus were evaluated by ELISA in a panel of sera from cattle experimentally infected with either B. abortus or or O:9 Y. enterocolitica. Two groups of four hybrid Holstein/Fresian cattle were independently infected with either B. abortus strain 544 (109 colony forming units) via the ocular route or O:9 Y. enterocolitica (1012 colony forming units) orally 4 times on different days. The two groups of animals were then kept separately to prevent cross-infection. All cattle were confirmed free of Yersinia and Brucella prior to experimental infection, and microbiological studies confirmed that subsequent infection had occurred. Serum from each animal was tested at 3, 7, 16, 24 and 53 weeks post-infection. All animal procedures were performed in accordance with the United Kingdom Animal (Scientific Procedures) Act,

- 20 040853 adopted in the UK in 1986.

Example 6 Use of OPS containing a cap structure as a diagnostic antigen.

A bovine study was conducted to evaluate the potential of novel OPS-based antigens to discriminate between antibodies induced by B. abortus field strains or B. abortus S19 vaccine.

Two antigens were assessed based on OPS using standard iELISA methods. They were a standard preparation of smooth lipopolysaccharide (sLPS) from B. abortus S99 and purified OPS from B. abortus S99 that had been modified and conjugated with a carrier to facilitate attachment to the surface of an ELISA plate (cOPS). This modification and conjugation created a cap structure in OPS, i.e. violated the terminal epitope.

These antigens were evaluated against the following panel of sera (no multiple animal sampling): 20 samples taken 45 days post-vaccination, 60 samples taken from herds that were confirmed by culture to be infected with field strains of B. abortus. The results are shown in FIG. 5. In addition, 7 negative and 7 positive controls were added. Vaccination was carried out by the conjunctival route using a dose of 5-10x109 CFU S19 B. abortus.

The sLPS antigen was most effective in differentiating samples from infected and non-infected animals. As might be expected, it was also most susceptible to reaction with sera from vaccinated animals. The cOPS antigen also detected sera from infected herds. The main finding was that the cOPS antigen was less sensitive in detecting antibodies induced by the vaccine, while remaining sensitive to antibodies induced in the field due to true infection. This was evidenced by the AUC values (distinguishing sera from infected herds from vaccinated animals), which were 0.8817 for the cOPS antigen and 0.6800 for the sLPS antigen. There was a significant difference between these two AUC values (P<0.01). The results of this study showed that the capped OPS antigen is an excellent serological tool in areas where B. abortus S19 vaccination is carried out.

The methods that were used for example 6.

Preparation of antigens: Brucella sLPS obtained from B. abortus S99 were purified by hot phenol extraction (Westphal et al (1952) Uber die Extraction von Bakterien mit Phenol/Wasser. Z. Naturforsch. 7:148155). OPS was obtained from it by moderate acid hydrolysis and size exclusion chromatography (Meikle et al. (1989) Infect Immun 57:2820-2828). OPS was oxidized at 2 mg/ml in 10 mM sodium metaperiodate (SMP) and 50 mM sodium acetate buffer (pH 5.5) for 1 hour in the dark. This was enough to oxidize the neighboring hydroxyl groups of the diol on the 2nd and 3rd carbon atoms of the terminal sugar. Residual SMP was removed by desalting with a PD-10 column (Sephadex-G25 column) according to the manufacturer's instructions (GE Healthcare). An appropriate volume of elution buffer allowed the OPS to be eluted.

The oxidized OPS was then subjected to reductive amination. Before removing salts into water using a Sephadex G-25 column, oxidized OPS was incubated in PBS at final concentrations of 5 mg/ml OPS and 0.5 M ammonium chloride, as well as 0.1 M sodium cyanoborohydride at 37°C for 24 h, and then lyophilized. OPS was then activated by incubation at 5 mg/ml with 5 mg/ml DSG in PBS for 45 minutes on a rotary shaker before desalting with a Zeba 40 kDa column in fresh PBS. Palmitic acid hydrazide (PAH) was dissolved to 10 mg/mL in DMSO and 1 part was added to 9 parts of OPS in PBS for a final dilution of 4.5 mg/mL OPS and 1 mg/mL RAH. Samples were reacted for 2 h at room temperature on a rotary shaker, after which excess PAH was removed by desalting in _H2O using a 40 kDa MWCO Zeba column. Then OPS conjugated with PAH was lyophilized.

Immunoassays: sLPS and cOPS were diluted to 0.5 and 5 μg/ml, respectively, in carbonate buffer (pH 10). 100 μl per well of each antigen was added to standard ELISA binding plates. The plates were incubated overnight at 4-8° C., then washed 5 times with wash solution (0.0014% w/v sodium monohydrogen orthophosphate and 0.1% Tween-20 in H ₂ O) and blotted dry.

Bovine sera were diluted 1/50 in PBS containing 0.1% Tween-20 and 100 μl per well were added to antigen coated plates. The plates were incubated for 1 hour at room temperature on a rotary shaker and then washed and blotted dry as described above. Protein A/G HRP conjugate was diluted to 0.05 μg/ml in PBS containing 0.1% Tween 20 and 100 μl was added to each well. The plates were then incubated, washed and dried as above for serum incubation. The substrate buffer was dibasic sodium phosphate with citric acid pH 5.5. One 10 mg tablet of OPD (o-phenylenediamine dihydrochloride) and 100 µl of 3% H2O2 was added to 25 ml of substrate buffer and 100 µl of the resulting solution was added to each well. The plates were developed for 15-30 minutes and then the optical density (OD) values were read at 450 nm. OD values for samples and controls were corrected for background by subtracting the OD of the well to which only buffer was added (no sera). The background-corrected OD value for each sample was expressed as a percentage of the background-corrected OD value of the total positive control.

Vaccination studies: The protective efficacy of the vaccine formulation was tested in accordance with the OIE (World Organization for Animal Health) requirements for testing the immunogenicity of B. abortus S19 and B. melitensis Rev1 vaccines (as described in chapters 2009 on bovine brucellosis (chapter 2.4.3) and brucellosis of goats and ovine (chapter 2.7.2) in the OIE Manual of Diagnostic Tests and Vaccines for Terrestrial Animals Mice were immunized as previously described for Example 5, except that on day 49 they were given a challenge dose of 100 µl delivered by ip and containing 2x105 cfu of B. abortus strain 544 (or B. melitensis strain 16M) Mice were sacrificed after 15 days.

Reference batches of B. abortus S19 and B. melitensis Rev1 vaccines and a negative (PBS only) control were evaluated simultaneously to demonstrate that the procedure had been performed correctly and to provide control points against which to evaluate the protective efficacy of the new vaccine. The procedure for quantifying protective efficacy by determining spleen weight and bacterial load is described below.

Each spleen was dissected under aseptic conditions, fat was removed, and the spleen was weighed and homogenized. Alternatively, spleens could be frozen and stored at -20°C for 24 hours to 7 weeks. Each spleen was homogenized under aseptic conditions using a glass grinder (or in appropriate sterile bags in a Stomacher homogenizer) in nine times its weight in PBS, pH 6.8, and three serial tenfold dilutions (1/10, 1/ 100 and 1/1000) of each homogenate obtained. 0.2 ml of each dilution was dispensed in quadruplicate on agar plates; two plates were incubated in a 10% CO ₂ atmosphere (allowing the growth of both vaccine and challenge strains) and the other two plates were incubated in air (inhibited the growth of CO ₂ -dependent B. abortus challenge strain 544), both at 37°C for 5 days.

Brucella colonies were counted on the dilutions corresponding to the cups, which were observed less than 300 cfu. If no colonies were observed in the plates corresponding to the 1/10 dilution, the spleen was considered to be infected with five bacteria. These amounts of Brucella per spleen were first recorded as X and expressed as Y after the following conversion: Y=log(X/logX). Then the mean and standard deviation were calculated, which are the response of each group of six mice.

The conditions of the control experiment were satisfactory if: i) the response of unvaccinated mice (mean of Y) was at least 4.5; ii) the response of mice vaccinated with the S19 reference vaccine was below 2.5; and iii) the standard deviation calculated for each batch of six mice was below 0.8.

Example 7. Intact whole cell diagnostic antigen (for the rose bengal test) containing a cap structure.

If the OPS is also attached to other molecules, the process of removing the top from the OPS can be carried out. Through these attachments, the OPS can form part of a larger entity, including the entire bacterial cell from which it normally protrudes.

The terminal perosamine in the OPS chain can be degraded by mild oxidation, thereby creating a cap structure at the remote end of the OPS chain, as described herein. This reaction, if maintained under appropriate conditions, is very specific for the chemical groups that are part of the terminal perosamine in OPS. Therefore, it is possible that disruption (creation of a cap structure) can be performed on an OPS if present in a more complex combination of molecules and components, without any significant or detrimental effect on non-OPS components. As a result, diagnostic benefits can be obtained from the capped OPS as described in Example 6 above, even if the OPS is in an unpurified state.

This approach was evaluated using a diagnostic whole cell agglutination assay known as the rose bengal test (RBT). This test is commonly used as a screening test for the serodiagnosis of brucellosis and is described by the OIE (World Organization for Animal Health) as suitable for this purpose. The diagnostic antigen consists of intact whole cells of B. abortus (strains S99 or S1119-3, both biovar 1 and A-dominant) that have been stained pink with rose bengal dye and then suspended in pH 3.65 buffer ( ±0.05). This dye makes it very easy to visualize the agglutination that occurs when the antigen is mixed with the test serum, which contains antibodies to Brucella. As with all traditional diagnostic tests used to serodiagnose brucellosis caused by smooth strains of Brucella (cases caused by B. abortus, B. melitensis and B. suis species), the main diagnostic molecule in the antigen is OPS, since it is the molecule to which most of the antibodies induced upon infection are produced (Ducrotoy et al. (2016) Veterinary Immunology and Immunopathology 171:81-102).

To create a cap structure (i.e. destroy the top) of the OPS in the RBT antigen that exists

- 22 040853 on the cell surface, the antigen was separated from the assay buffer by centrifugation and suspended in cold (4°C) oxidizing reagents (10 mm sodium metaperiodate in 0.1 M sodium acetate buffer, pH 5.5). Cells were incubated with these reagents in the dark at 4°C until the moderate oxidation reaction was completed. This was confirmed by measuring the consumption of sodium metaperiodate over time and reaching a stage where replenishment of sodium metaperiodate back to 10 mM resulted in no further consumption. At this stage, it was believed that all OPS on the cell surface were capped (i.e., their top was destroyed). Cells were centrifuged to separate them from the oxidative buffer and resuspended in test buffer for serological evaluation.

The analytical sensitivity of oxidized (capped) and non-oxidized (uncapped) antigens was compared using a dilution series (in negative bovine serum) of a known positive bovine sample from an animal infected with B. abortus: pure, 1/2 1/4 , 1/8, 1/16, 1/32, 1/64 and 1/128. The negative serum used for dilution was also tested. The results showed that the two antigens (before and after oxidation) had the same effect, agglutination was observed up to a dilution of 1/64. A 1/128 dilution of positive serum and pure negative serum was negative with both antigens.

To assess diagnostic sensitivity, oxidized (with a cap) and non-oxidized (without a cap) antigens were added to 17 sera from cattle not infected with Brucella and 17 sera from cattle infected with B. abortus. The results showed that all samples from cattle infected with Brucella agglutinated with both antigens, and all samples from uninfected cattle did not agglutinate with either antigen. The positive and negative controls used in this evaluation also gave correct results for both antigens.

The results of this study demonstrated that relatively crude antigens that contained OPS can be fully oxidized under moderate conditions and remain effective serodiagnostic antigens. It has already been shown that this provided the formation of a cap structure in OPS. In the results obtained during this study, the ability of the oxidized antigen to distinguish between samples of animals infected with Brucella from non-infected animals did not change. This is consistent with the data presented in Example 6, where the diagnostic performance of oxidized (capped/untopped) OPS for these sample types was excellent. In this example, capped OPS showed reduced sensitivity to samples from animals vaccinated with S19 B. abortus. An oxidized (i.e. capped) RBT antigen also has this property and will therefore be superior to a non-oxidized antigen in differentiating between sera from animals infected with Brucella smooth strains and sera from animals vaccinated with Brucella smooth strain vaccines (such as S19 B. abortus and Rev1 B. melitensis).

This method of capping the OPS can also be applied to other Brucella OPS contained in diagnostic antigens, such as sLPS, which are used for ELISA (where the OPS is attached to backbone sugars, which in turn are attached to a lipid A), other antigens (whole cells, cell lysates or fractions) used in agglutination assays (such as serum agglutination assay, O-PSantigen agar gel immunodiffusion reaction, and complement fixation assay).

The methods that were used for example 7.

RBT was carried out as described in the OIE guidelines (OIE 2016, Brucellosis chapter 2.1.4. Guidelines for diagnostic tests and vaccines for terrestrial animals. OIE, Paris). Test (and control) sera were added to the test antigen side of the RBT, 30 µl each, on a smooth white surface. They were then both mixed together to form an oval or circle approximately 2 cm in diameter. This mixture was gently shaken at room temperature (18-26°C) for 4 minutes. After this time, the mixture was examined and any visible agglutination was considered positive. All samples that were tested as described above were analyzed using this method. For each test, an assay was performed with positive and negative control sera.

To oxidize the RBT antigen, the volume of the antigen at the working concentration was centrifuged at 3000 g to pellet the cells. The supernatant, i.e. the test buffer was removed and the cells were resuspended in a volume equal to that removed, cold (+4° C.) oxidative reagent (10 mM sodium metaperiodate in 0.1 M sodium acetate buffer, pH 5.5). Cells were incubated for 30 min in the dark and then centrifuged as before. The supernatant was removed and replaced with fresh oxidizing reagent. This process was repeated four more times so that the antigen was subjected to a total of 6 times 30-minute oxidation with an hourly replenishment of the oxidizing reagent. After the fifth hour, the cells were centrifuged as described above, the supernatant was removed and replaced with the original volume of fresh test buffer. The replenishment process allowed the reaction to proceed without restriction in terms of sodium metaperiodate consumption and without deviating from moderate oxidation conditions that facilitate specific reaction with neighboring cis-diols on the OPS terminal perosamine.

The content of sodium metaperiodate in the extracted supernatant was measured by dosing

- 23 040853 100 µl of oxidizing reagent into individual wells of a 96-well ELISA plate. Then, 100 μl of 0.5 mg/ml ABTS (2,2-azinobis-3-ethylbenzthiazoline-6-sulfonic acid) in pH 4.0 buffer was added to each test well. In the presence of sodium metaperiodate, this resulted in a color change that was proportional to the concentration of that oxidizing agent. Color change after 15 minutes was measured with an ELISA plate reader set to measure absorbance at 405 nm. The molarity of sodium metaperiodate in the test sample was calculated by comparison with a standard curve that was generated using control wells containing a known concentration of sodium metaperiodate. The results of the consumption of the oxidizing agent are shown in Fig. 16, where A is the flow rate after the first 30 minute period, B is the flow rate after two 30 minute periods, and so on. Thus, the figure shows a standard curve of the known concentration of sodium metaperiodate as a function of optical density (O.D. at 405 nm) and OD values of oxidizing reagents extracted at different time points from the start of the oxidation process (on the right side). As can be seen, after 30 minutes most of the sodium metaperiodate was consumed. The first replenishment (2x30 min) was not in the same amount as was consumed, but only slightly more than half of the spent sodium metaperiodate. The second replenishment (3x30 min) contained more than half (about 7 mm) of the remaining sodium metaperiodate. The third replenishment (4x30 min) contained approximately 8 mM concentration of sodium metaperiodate, and the fourth (5x30 min) and fifth (6x30 min) replenishments contained approximately 9 mM of the remaining sodium metaperiodate.

It is clear from these data that sodium metaperiodate was consumed, and that consumption slowed down and then virtually stopped when cells that had already been exposed to sodium metaperiodate were injected. The graph in this figure shows that after five oxidation cycles there was no more significant reagent consumption. From this it was concluded that the antigen was completely oxidized, all molecules capable of being oxidized by this moderate process were completely oxidized. After this oxidation process, the cells were centrifuged as described previously, the supernatant was removed and they were resuspended in test buffer. These cells were then evaluated for diagnostic performance by incorporating them into the test sera described above. The oxidized RBT antigen was analyzed in parallel with the original RBT antigen that had not been oxidized.

Example 8 Use of antigens in the form of trisaccharides (structures XII) and disaccharides (structures XI) with exclusively 1,2-bonds as serodiagnostic antigens in brucellosis.

The properties of antigens in the form of trisaccharides (structure XII) and disaccharides (structure XI) with exclusively 1,2-linkages as diagnostic tools for DIVA were described above in Example 5. The effectiveness of these antigens was shown by demonstrating their good diagnostic sensitivity for detecting infection with B. abortus in cattle and infection (Biovar 2) with B. suis in pigs. Thus, the value of these antigens in the context of DIVA has been well established.

Conducted an additional study to assess the suitability and benefits of these antigens for type serology in the presence or absence of vaccination. Since the role of the apical epitope was demonstrated in the course of the work described here as an important aspect of the diagnostic sensitivity of these antigens, it was expected that the presence of an identical epitope on OPS O:9 Y. enterocolitica (4,6-dideoxy-4-formamido-mannopyranosyl polymer with exclusively 1,2 bonds) will lead to cross-reactions and false positives, which may be excessive. This expectation was reinforced by the strong responses shown by sera from pigs infected with B. suis biovar 2 (because the long OPS repeat element from this biovar is identical to Y. enterocolitica O:9) to an antigen in the form of a trisaccharide with exclusively 1,2-linkages (structures XII).

To assess the extent of this cross-reactivity, sera from four bovine animals experimentally infected with Brucella abortus strain 544 and four bovine animals experimentally infected with O:9 Y. enterocolitica were evaluated (using samples taken at 3, 7, 16, 24 and 53 weeks after infection). To measure responses to other diagnostic antigens, serological tests were performed using several different antigens, and the results of these tests are presented as line graphs: B. abortus sLPS S99 (Fig. 9), a hexasaccharide with exclusively 1,2-bonds (structure IX) ( Fig. 10), a trisaccharide with exclusively 1,2-bonds (structure XII) (Fig. 11) and a monosaccharide (structure II) (Fig. 12). Table 6 shows the number of samples from cattle infected with O:9 Y. enterocolitica with serological results higher than the lowest serological result obtained from samples from cattle infected with 544 B. abortus. This gave an indication of the number of false positives that occurred in this set of samples using the method of the present invention if the sensitivity criterion was set to 100% (since all animals were infected with Brucella). 100% assay sensitivity is essential in a Brucella testing system because failure to detect a Brucella-positive animal can be devastating.

- 24 040853

Table 6. Number of specimens from cattle infected with O:9 Y. enterocolitica with serological results greater than the lowest serological result obtained from specimens from cattle infected with 544 B. abortus

Antigen Presence/absence of the apex No. positive for 0:9 U. enterocolitica specimens sLPS S99 B. abortus Availability 15 A hexasaccharide conjugate with exclusively 1,2 bonds and BSA (structure IX) Availability eleven Trisaccharide conjugate with exclusively by 1,2-bonds and BSA (structure XII) Availability 7 Monosaccharide-BSA conjugate (structure II) Availability 4

The results demonstrate that if the diagnostic antigens had an apical epitope, the number of Y. enterocolitica O:9 positive specimens decreased, and hence the specificity increased as the antigen length became shorter: sLPS>hexasaccharide>trisaccharide>monosaccharide. This was even though the terminal perosamine in the OPS of B. abortus, B. melitensis, B. suis, and O:9 of Y. enterocolticia was exactly the same. The shorter the oligosaccharide was (with only 1,2 bonds), the more the specificity increased; this could not be predicted.

Detailed evaluation of serological results showed that although the results for samples from Y. enterocolitica O:9-infected animals were initially high compared to the exclusively 1,2-linked trisaccharide (structure XII), they then dropped rapidly in all four animals. By week 16, the results in all four animals remained lower than the lowest result in the Brucella-infected animals. The results in Brucella-infected animals did not start to decrease significantly from the period after the 16th week, and even then two animals had consistently high results. Serological profiles obtained using B. abortus sLPS S99 antigen (the antigen recommended by the OIE for iELISA on Brucella) showed a drop in results obtained from animals infected with O:9 Y. enterocolitica immediately in 3 out of 4 cases (but not as abruptly as was the case with a trisaccharide with only 1,2-bonds), with one sample from Brucella-infected animal number 2 starting to give a relatively low result at week 53. The results in O:9-infected Y. enterocolitica animal number 2 increased until week 7 and remained high until week 24.

The results from the 1,2-linked exclusively hexasaccharide antigen (structures IX) demonstrated the properties of both the trisaccharide (structure XII) antigen and the sLPS antigens from s99 as corresponding to its intermediate length. Although the results in samples obtained from animals infected with O:9 Y. enterocolitica decreased relatively quickly, the result in animal 2 increased from the 2nd to the 7th week. The results for monosaccharides showed a good difference between types of infection, although some of the results for samples from Brucella-infected animals were quite low, reflecting the more limited sensitivity of this antigen.

Trisaccharide with exclusively 1,2-bonds (structure XII), sLPS S99 B. abortus and a mixture (50/50 by weight) of antigens in the form of trisaccharides with exclusively 1,2-bonds (structures XII) and tetrasaccharides with exclusively 1,2- ties (structure VI) were tested on 29 sera from bovines infected in the field with B. abortus, 20 sera from randomly selected, non-Brucella-infected bovines, and 31 sera from bovines that were false positive for conventional serodiagnostic tests for Brucella. Data are presented in 3 scatter plots: B. abortus sLPS S99 vs. exclusively 1,2-linked trisaccharide (structures XII) (Fig. 13), B. abortus sLPS S99 vs. 50/50 mixture of exclusively 1,2-linked trisaccharide ( structure XII) and a specific Mantigenic tetrasaccharide (structure VI) (Fig. 14) and a trisaccharide with exclusively 1,2-bonds (structure XII) versus a 50/50 mixture of a trisaccharide with exclusively 1,2-bonds (structures XII) and specific M- antigenic tetrasaccharide (structure VI) (Fig. 15).

The scatter plots (FIGS. 13-15) show that all three antigenic preparations completely distinguished samples from infected animals from samples from randomly selected non-infected animals. However, significant overlap was observed in all cases with samples from the FPSR population. In table. 7 shows the ability of antigens to distinguish samples from infected ones by means of

- 25 040853

Brucella animals and from animals from FPSR populations where sensitivity is high. The need for high sensitivity was reflected both in sample populations and in testing requirements. The results showed that with a diagnostic sensitivity of 100 and 96.6%, both a trisaccharide with exclusively 1,2-linkages (structure XII) and a 50/50 mixture of a trisaccharide with exclusively 1,2-linkages (structures XII) and M-antigenic tetrasaccharide (structures VI) outperformed native B. abortus S99 sLPS (the current standard antigen recommended by the OIE).

Table 7. Specificity for the FPSR population when the positive/negative threshold of the test is adapted to different antigen sensitivities (number of positive samples in parentheses)

Diagnostic sensitivity sLPS S99 B. abortus Trisaccharide _____c exclusively by 1,2 bonds Mixture 100.0% 0.0% (0) 12.9% (4) 9.7% (3) 96.6% 16.1% (5) 38.7% (12) 35.5% (11) 93.1% 58.1% (18) 45.2% (14) 38.7% (12)

The results and conclusions obtained on field sera and experimentally infected sera were consistent. At the highest levels of diagnostic sensitivity, the specificity obtained using an exclusively 1,2-linked trisaccharide (structures XII) was superior to that obtained with the natural sLPS antigen (the current standard antigen). Reducing the sensitivity requirement gave excellent performance in sLPS, although the reduction in sensitivity was undesirable. Data from experimental infections suggested that the specificity of different antigens depends on how close to the time of infection with a cross-reacting organism, such as O:9 Y. enterocolitica, the sample is taken.

Therefore, a trisaccharide (structure XII) with an exclusively 1,2-linkage, a disaccharide (structure XI) with an exclusively 1,2-linkage, and a monosaccharide (structure II) were not only unexpectedly highly sensitive diagnostic sensitive antigens for brucellosis, but they also unexpectedly had high diagnostic specificity.

In the present example, serology was performed on samples taken from infected animals, while the infectious Brucella biovar was an A-dominant strain. It would be expected that such infections would cause an increase in the proportion of antibodies that react with perosamine sequences with exclusively 1,2-bonds (4,6-dideoxy-4-formamidomanopyranosyl), but not antibodies against sequences containing perosamines with 1,2 3-links. The ability of an exclusively 1,2-linked trisaccharide (structure XII) and antigens in the form of a specific M antigenic tetrasaccharide (structure VI) to detect antibodies to OPS induced by such infections has been shown above or previously (WO 2014/170681). The antigens may be useful when they are used alone, or, as shown above, they give a good effect when used together.

If infections with M-dominant strains take place, then the antibodies induced are likely to be shifted towards a higher proportion of antibodies to perosamine sequences containing 1,3-bonds. Under such circumstances, a specific M antigenic tetrasaccharide (structure VI) will be a more sensitive diagnostic tool. The use of two antigens in combination (structures XII or XI in combination with structure VI, e.g. used as a mixture) gave optimal sensitivity in both scenarios, namely when infected with Adominant or M-dominant strains of B. abortus, B. melitensis and B. suis.

In summary, the work described herein provides an antigen combination that is a versatile antigen that is sensitive, compatible with DIVA, more specific than native antigens such as the OIE recommended B. abortus sLPS S99 antigen, and cheaper to manufacture and use. than a longer synthetic universal antigen.

The methods that were used for example 8.

Used serological methods and samples were the same as described for example 5, with the addition of false positive serological reaction samples (FPSR). These sera were collected in the UK between 1996 and 1999, more than 10 years since its official brucellosis-free status was announced. These sera were positive in at least one of the four conventional bovine brucellosis serodiagnostic assays: CFT, SAT, cELISA, or iELISA, which are approved by the OIE. Other than serology, there was no culture or epidemiological evidence of disease.

Application: methods for the synthesis of polysaccharides.

Synthesis of heptasaccharide.

The synthesis involved the use of three key building blocks of the known S12 methyl glycoside with a protective group. Two glycosyl donors 11 and 13 (described below) were used to extend

- 26 040853 oligosaccharide with 1,2-bonds, wherein the donor 11 and the cap-bearing residue 13 carried a connecting element for conjugation to the protein. Compound 11 allowed chain extension one residue at a time, and the temporary acetate protecting group on O-2 could be easily removed, exposing the hydroxyl group for further chain extension. The preformed capping residue with attached linker 13 was prepared from the known methyl 2,3-Oisopropylidene-6-deoxy-α-D-mannopyranoside S5 and the protected linker 12, which in turn was derived from commercially available benzyl (5-hydroxypentyl)carbamate by two-step conversion to S13 and then to 12 (Scheme 4S). A series of transformations made it possible to carry out the reaction of S5 with 12, and then additional reactions gave thioglycoside 13.

The detailed construction of these intermediates proceeded as described below.

Synthesis of thioglycoside donors 11 Methyl-4-azido-3-O-benzyl-4,6-dideoxy-α-D-mannopyranoside (S11).

The analytical data for the title compound were essentially the same as previously described (Bundle et al (1988) Carbohydr Res 174:239-251).

1,2-di-O-acetyl-4-azido-4,6-dideoxy-α-D-mannopyranose (S12).

A solution of S11 (5 g, 17.05 mmol) in acetic anhydride/acetic acid/sulfuric acid (50:20:0.5, 50 ml) was stirred at 21° C. for 3 h and then poured into an ice-cold 1 M solution K2CO3 (80 ml). The mixture was then diluted with CH ₂ Cl ₂ (~100 ml) and washed with water (2x30 ml), saturated aqueous NaHCO ₃ (35 ml) and brine (15 ml). The organic phase was separated, dried over MgSO ₄ and concentrated under vacuum. The residue was purified by silica gel column chromatography (gradient elution with ethyl acetate-hexane) to give the title compound S12 (5.6 g, 91%) as a sticky liquid. Analytical data for S12: Rf=0.35 (ethyl acetate/hexane, 1/4, v/v); [a]D ²¹ =+30.71 (c=1.51, CHC13); ¹³ C NMR (176 MHz, CDCI3): δ: 169.8, 168.3, 136.9, 128.5, 128.3, 128.1, 91.0, 75.7, 71.8, 69, 3, 66.3, 63.5, 20.8 (x2), 18.5 ppm; HRMS (ESI): m/z, calculated for C^NjOiNa [M+Na]+: 386.1323, revealed: 386.1322.

n-Tolyl-2-O-acetyl-4-azido-3-O-benzyl-4,6-dideoxy-1-thio-α-D-mannopyranoside (11).

_{BF 3 _} -Et ₂ O (0.32 ml, 2.57 mmol). When TLC showed that the reaction was complete, the mixture was diluted with CH ₂ Cl ₂ (~50 ml) and washed with water (2x10 ml), saturated aqueous NaHCO ₃ (15 ml) and brine (10 ml). The organic phase was separated, dried over MgSO ₄ and concentrated under vacuum. The residue was purified by flash column chromatography (gradient elution with ethyl acetate-hexane) to give 11 as a sticky liquid (0.854 g, 92.9%). Analytical data for 11: Rf=0.7 (ethyl acetate/hexane, 1/3, v/v); [a] _D ²¹ =+135.5 (c=2.25, CHCI3); ¹³ C NMR (176 MHz, CDCI3): δ: 170.0, 138.1, 137.0, 132.4, 132.3, 129.9, 129.8, 129.6, 128.5, 128, 5, 128.4, 128.1, 86.4, 76.4, 71.7, 69.1, 68.2, 64.2, 21.1, 21.0, 18.4 ppm; HRMS (ESI): m/z, calculated for C ₂₂ H ₂₅ N ₃ O ₄ SNa [M+Na]+: 450.1458, revealed: 450.1465.

Synthesis of linker bromoalkane 12 5-(N-benzyl((benzyloxy)carbonyl)amino)pentanolbenzoate (S13).

Benzoyl chloride (0.88 ml, 7.59 mmol) was added dropwise to a stirred solution of benzyl(5hydroxypentyl)carbamate (commercially available) (1.5 g, 6.32 mmol) in aqueous CH ₂ Cl ₂ (15 ml), containing Et ₃ N (1.76 ml, 1.26 mmol), at 0°C. After 1 min, DMAP (1.7 g, 13.9 mmol) in aqueous CH ₂ Cl ₂ (10 ml) was added dropwise to the reaction mixture and stirred at room temperature overnight. The resulting mixture was diluted with CH ₂ Cl ₂ (~30 ml) and washed with aqueous HCl (1 M, 1x10 ml), water (60 ml), aqueous NaHCO ₃ (30 ml) and brine (30 ml). The organic phase was separated, dried over MgSO ₄ and concentrated under vacuum. The residue was quickly filtered off on silica gel (gradient elution with ethyl acetate-hexane) to give an almost pure compound as an oil. This crude material was directly used for benzylation.

To a solution of a protected benzoyl compound (0.9 g, 2.63 mmol) dissolved in anhydrous DMF (10 mL) was added NaH (0.12 g, 2.89 mmol) at 0°C. The mixture was stirred at 0° C. for 45 minutes and then BnBr (0.37 ml, 3.16 mmol) was added. After stirring for a further 12 hours, when TLC showed that the reaction was complete, it was quenched with H ₂ O at 0°C and the mixture was diluted with EtOAc. The aqueous layer was extracted with EtOAc (5x25 ml) and the organic phases were combined and dried over Na2SO4. The desired product S13 (1.093 g, 96.1%) was obtained after flash column chromatography (gradient elution with ethyl acetate - hexane) of the condensed product. Analytical data for S13: Rf=0.6 (ethyl acetate/hexane, 1/3.5, v/v ¹³ C NMR (176 MHz, CDCI3): δ: 166.6, 156.7, 156.2, 137.9, 136.8, 132.8, 130.4, 129.5, 128.5, 128.4, 128.3, 127.8, 127.3, 127.2, 67.2, 64, 8, 64.7, 50.5, 50.2, 47.0, 46.0, 28.4, 27.8, 27.4, 23.3 ppm HRMS (ESI): m/z, calculated for C ₂₇ H ₂₉ NO ₄ Na [M+Na]+: 454.1989, identified: 454.1986.

Benzyl-N-benzyl(5-bromopentanyl)carbamate (12).

Sodium methoxide (~0.8 ml, 0.5 M solution) was added to a solution of S13 (1.0 g, 2.32 mmol) in CH3OH (15

- 27 040853 ml) until reaching pH ~9 and the resulting mixture was stirred in an argon atmosphere for 6 h at 21°C.

Thereafter, the reaction mixture was neutralized with Amberlite IR 120 (H+) ion exchange resin, the resin was filtered off and washed successively with CH ₃ OH. The combined filtrate was concentrated in vacuo and this crude material was used directly for bromination.

To a solution of the deprotected compound (0.96 g, 2.92 mmol) dissolved in anhydrous CH ₂ Cl ₂ (15 ml) was added CBr ₄ (1.85 g, 5.55 mmol) and PPh ₃ (1 .54 g, 5.86 mmol) at 0°C. The reaction mixture was allowed to warm to room temperature and stirred for another 3 hours. When the reaction was seen to be complete by TLC, it was quenched with H ₂ O at 0°C, then the mixture was diluted with CH ₂ Cl ₂ (~50 ml) and washed with water (2x10 ml), saturated aqueous NaHCO ₃ (15 ml) and brine (15 ml). The organic phase was separated, dried over MgSO ₄ and concentrated under vacuum. The residue was purified by silica gel column chromatography (gradient elution with ethyl acetate-hexane) to give the title compound 12 (1.085 g, 94.8%) as a liquid. Analytical data for 12: Rf=0.85 (ethyl acetate/hexane, 1/4, v/v); ¹³ C NMR (176 MHz, CDCl3): δ: 156.7, 156.2, 137.8, 136.7, 128.5 (x2), 128.4, 128.0, 127.9, 127.4 , 127.3, 127.2, 67.3, 67.2, 50.6, 50.3, 46.9, 46.0, 33.6, 33.4, 32.3 (x2), 27, 2, 26.8, 25.3 ppm; HRMS (ESI): m/z, calculated for C20H24NO2BrNa [M+Na]+: 412.0883, revealed: 412.0878.

Synthesis of p-tolylthioglycoside donor 13.

Scheme 4S Conditions: a) CbzBnN(CH ₂ ) ₅ Br, NaH, DMF, 0°C to room temperature, 18 h; b) TFA/H2O (9:1), CH ₂ Cl ₂ , room temperature, 10 min; c) BzCl, DMAP, Et ₃ N, CH ₂ Cl ₂ , 0°C to room temperature, 12 h; d) Ac ₂ O, AcOH, H2SO4, at room temperature, 4 h; e) BF ₃ :Et ₂ O, p-toluenethiol, CH ₂ Cl ₂ , 0°C to room temperature, 10 h.

Methyl 2,3-O-isopropylidene-6-deoxy-α-D-mannopyranoside (S5).

The analytical data for the title compound were in fact the same as previously described (Eis & Ganem (1988) Carbohydrate Research 176:316-323).

Methyl 4-O-(5'-N-benzyl-5'-N-carboxybenzylpentanyl)-2,3-O-isopropylidene-6-deoxy-α-D-mannopyranoside (S14).

To a solution of S5 (2.0 g, 9.17 mmol) dissolved in anhydrous DMF (15 ml) was added NaH (0.4 g, 10.08 mmol) at 0°C. The mixture was stirred at 0° C. for 45 minutes and then CbzBnN(CH ₂ ) ₅ Br (4.5 g, 11.01 mmol) was added. After stirring for a further 12 hours, when TLC showed that the reaction was complete, it was quenched with H ₂ O at 0°C and the mixture was diluted with EtOAc. The aqueous layer was extracted with EtOAc (5x25 ml) and the organic phases were combined and dried over Na ₂ SO ₄ . The desired product S14 (3.26 g, 73.2%) along with removed alkene and a small amount of unreacted starting material S5 (0.16 g) was obtained after flash column chromatography (gradient elution with ethyl acetate - hexane) of the condensed product. Analytical data for S14: Rf=0.6 (ethyl acetate/hexane, 1/4, v/v); [a] _D = +20.48 (c=2.11, CHC13); ¹³ C NMR (176 MHz, CDO3): δ: 156.7, 156.1, 137.9, 136.9, 136.8, 128.5, 128.4, 127.9, 127.8, 127, 3, 127.2, 109.0, 98.0, 82.0, 78.5, 75.9, 71.3, 67.1, 64.5, 54.7, 50.4, 50.1, 47.1, 46.1, 29.8, 28.0, 27.9, 27.5, 26.3, 23.4, 17.7 ppm; HRMS (ESI): m/z, calculated for С3^DOw [M+Na]+: 550.2775, revealed: 550.2785.

Methyl 4-O-(5'-N-benzyl-5'-N-carboxybenzylpentanyl)-6-deoxy-α-D-mannopyranoside (S15).

A solution of S14 (1.0 g, 1.89 mmol) in TFA:H ₂ O (9:1, 10 mL) was stirred at 21° C. for 30 min and then poured into an ice-cold 1 M K ₂ CO ₃ solution (50 ml). The mixture was then diluted with CH ₂ Cl ₂ (~50 ml) and washed with water (2x30 ml), saturated aqueous NaHCO ₃ (25 ml) and brine (15 ml). The organic phase was separated, dried over MgSO ₄ and concentrated under vacuum. The residue was purified by silica gel column chromatography (gradient elution with ethyl acetate-hexane) to give the title compound S15 (0.742 g, 80.3%) as an oil. Analytical data for S15: Rf=0.4 (ethyl acetate/hexane, 1/1, v/v); [a] _D ²¹ =+38.31 (c=1.27, CHCl3); ¹³ C NMR (176 MHz, CDCl3): δ: 156.7, 156.3, 137.8, 136.6, 129.6, 128.5, 128.4, 127.9, 127.8, 127, 3, 127.2, 100.3, 81.7, 71.4, 71.3, 71.2, 67.2, 67.1, 54.8, 50.5, 50.3, 47.1, 46.1, 30.0, 29.8, 27.9, 27.2, 23.2, 17.9 ppm; HRMS (ESI): m/z, calculated for C27H ₃₇ NO ₇ Na [M+Na]+: 510.2462, detected: 510.2462.

Methyl 2,3-di-O-benzoyl-4-O-(5'-N-benzyl-5'-N-carboxybenzylpentanyl)-6-deoxy-α-D-mannopyranoside (S16).

- 28 040853

Benzoyl chloride (0.23 ml, 1.97 mmol) was added dropwise to a stirred solution of S15 (0.4 g, 0.82 mmol) in anhydrous CH ₂ Cl ₂ (10 ml) containing Et ₃ N (0.46 ml , 3.28 mmol), at 0°C. After 2 minutes, DMAP (0.451 g, 3.69 mmol) in aqueous CH ₂ Cl ₂ (5 ml) was added dropwise to the reaction mixture and stirred at room temperature overnight. The resulting mixture was diluted with CH ₂ Cl ₂ (~20 ml) and washed with aqueous HCl (2 M, 1x5 ml), water (20 ml), aqueous NaHCO ₃ (10 ml) and brine (10 ml). The organic phase was separated, dried over MgSO ₄ and concentrated under vacuum. The residue was purified by silica gel column chromatography (gradient elution with ethyl acetate-hexane) to give the title compound S16 (0.513 g, 90%) as an oil. Analytical data for S16: Rf=0.7 (ethyl acetate/hexane, 1/3.5, v/v); [a]D ²¹ = -70.58 (c=1.71, CHC13); ¹³ C NMR (176 MHz, CDCl,): δ: 165.5, 165.2, 156.7, 156.3, 137.9, 133.3, 133.0, 129.9, 129.8, 129 .8, 129.6, 128.5, 128.4, 128.3, 127.9, 127.8, 127.1, 98.5, 79.5, 73.1, 72.9, 72.1 , 71.1, 67.6, 67.1, 55.0, 50.4, 50.1, 47.0, 46.0, 29.9, 27.8, 27.4, 23.3, 18 .0ppm; HRMS (ESI): m/z, calculated for C^^Na [M+Na]+: 718.2987, revealed: 718.298.

1-O-acetyl-2,3-di-O-benzoyl-4-O-(5'-N-benzyl-5'-N-carboxybenzylpentanyl)-6-deoxy-α-D-mannopyranose (S17).

A solution of S16 (0.5 g, 0.716 mmol) in acetic anhydride/acetic acid/sulfuric acid (50:20:0.5, 10 mL) was stirred at 21°C for 3 h and then poured into an ice-cold 1 M solution K2CO3 (50 ml). The mixture was then diluted with CH ₂ Cl ₂ (~20 ml) and washed with water (2x30 ml), saturated aqueous NaHCO3 (15 ml) and brine (10 ml). The organic phase was separated, dried over MgSO4 and concentrated in vacuo. The residue was purified by silica gel column chromatography (gradient elution with ethyl acetate-hexane) to give the title compound S17 (0.485 g, 92.2%) as a liquid. Analytical data for S17: Rf=0.55 (ethyl acetate/hexane, 1/4, v/v); [a] _D ²¹ = -48.86 (c=1.51, CHCl); ¹³ C NMR (176 MHz, CDCl3): δ: 168.6, 165.4, 165.2, 137.8, 136.7, 133.5, 133.2, 129.8, 129.6, 129 .4, 129.0, 128.5, 128.5, 128.4, 128.2, 127.9, 127.8, 127.2, 127.1, 90.8, 79.1, 73.4 , 71.8, 70.1, 69.9, 67.1, 50.4, 50.1, 46.9, 45.9, 29.9, 27.8, 27.4, 23.3, 21 .0, 18.1, ppm; HRMS (ESI): m/z, calculated for C ₄₂ H ₄₅ NO ₁₀ Na [M+Na]+: 746.2936, revealed: 746.2931.

n-Tolyl-2,3-di-O-benzoyl-4-O-(5'-N-benzyl-5'-N-carboxybenzylpentanyl)-6-deoxy-1-thio-α-D-mannopyranoside (13).

_{BF 3 -Et 2} O (0.25 ml, 1.99 mmol). When TLC showed that the reaction was complete, the mixture was diluted with CH ₂ Cl ₂ (~30 ml) and washed with water (2x10 ml), saturated aqueous NaHCO ₃ (10 ml) and brine (20 ml). The organic phase was separated, dried over MgSO4 and concentrated in vacuo. The residue was purified by flash column chromatography (gradient elution with ethyl acetate-hexane) to give 13 as a white solid (1.18 g, 90.7%). Analytical data for 13: Rf=0.65 (ethyl acetate/hexane, 1/4, v/v); [a]D ²¹ = -1.02 (c=0.9, CHCl); ¹³ C NMR (176 MHz, CDCl3): δ: 165.4, 165.3, 156.7, 156.1, 138.1, 137.9, 136.9, 136.8, 133.4, 133 .2, 132.7, 132.3, 130.0, 129.9 (x2), 129.8, 129.7 (x2), 129.6, 128.5 (x2), 128.4 (x2) , 127.9, 127.8, 127.3 (x2), 127.2, 86.2, 79.7, 73.3, 73.1, 72.6, 72.4 (x2), 69.2 , 67.1, 50.5, 50.2, 47.0, 46.0, 30.0, 27.9, 27.4, 23.3, 21.2, 18.0 ppm; HRMS (ESI): m/z, calculated for (/LHA/^Xu [M+Na]+: 810.3071, revealed: 810.3069.

Methyl 4-azido-3-O-benzyl-4,6-dideoxy-α-D-mannopyranoside

The analytical data for the title compound were in fact the same as previously described (Bundle et al (1988) Carbohydr Res 174:239-251), (Eis & Ganem (1988) Carbohydrate Research 176:316-323).

Assembly of the heptasaccharide.

Glycosylation of the methyl glycoside with activated thioglycoside 11 gave the disaccharide 14. This was subjected to an interesterification reaction to remove the ethyl acetate group, exposing the hydroxyl group for a repeated glycosylation and interesterification sequence. This was repeated 4 more times, which in turn gave trisaccharides 16 and 17, tetrasaccharides 18 and 19, pentasaccharides 21 and 22, and hexasaccharides 22 and 23. Then, in the final chain extension reaction, a cap-bearing moiety was attached with a connecting element by reacting 13 with 23 to give the heptasaccharide 24, and after removing the benzoate ester, the partially deprotected alcohol 25. The deprotection was carried out in a series of steps including reduction of the azido groups to an amine followed by their N-formylation followed by a hydrogenolysis step to remove the benzyl moieties with ester groups and the amino protecting groups (Ganesh et al (2014) Journal of the American Chemical Society 136:16260-16269). Compound 8 was then conjugated to protein by selective tetra-amino activation with bis-succinimide ester (DSG) or dibutyl squarate to give activated intermediates S26 and S27. S26 was reacted with tetanus toxoid to produce vaccine glycojugate 9 and S27 was reacted with BSA to produce screening antigen 10.

Methyl-4-azido-2-O-acetyl-3-O-benzyl-4,6-dideoxy-α-D-mannopyranosyl (1^2) 4-azido-3-Obenzyl-4,6-dideoxy-α- D-mannopyranoside (14).

- 29 040853

The glycosyl acceptor compound S11 (1.42 g, 4.84 mmol) and the glycosyl donor compound 11 (2.27 g, 5.33 mmol) were combined, azeotropically distilled twice with anhydrous toluene (5 ml) and placed under atmosphere high vacuum for 2 h. The mixture was then dissolved in CH ₂ Cl ₂ (20 ml), treated with freshly activated 4 A molecular sieves (1.5 g), stirred under Ar at room temperature for 1 h. NIS (2 .4 g, 9.71 mmol). After cooling to -10°C, TMSOTf (0.19 ml, 0.971 mmol) was added and the reaction mixture was allowed to warm to room temperature. When TLC showed that the reaction was complete, saturated aqueous NaHCO ₃ (15 ml) and CH2C12 was added and the resulting mixture was passed through celite to remove molecular sieves. The combined filtrates were washed with aqueous Na ₂ S ₂ O ₃ (20%) and water. After extraction of the aqueous layer with CH ₂ Cl ₂ (3x15), the combined organic phase was dried over Na ₂ SO ₄ , concentrated in vacuo and purified by silica gel column chromatography (gradient elution with ethyl acetate/hexane) to give the disaccharide 14 (2.66 g, 92.1%) as a sticky liquid. Analytical data for 14: Rf=0.5 (ethyl acetate/hexane 1:4, v/v); [a]D ²¹ =+36.24° (c=1.92, CHCI3); ¹³ C NMR (176 MHz, CDCI3): δ: 169.7, 137.6, 137.1, 128.5 (x2), 128.4, 128.0, 127.9, 127.8, 99.7 , 99.4, 77.7, 75.4, 73.7, 72.0, 71.6, 67.6, 67.2, 66.9, 64.1, 63.8, 54.9, 20 .9, 18.5(x2)ppm; HRMS (ESI): m/z, calculated for C ₂₉ H ₃₆ N ₆ O ₈ Na [M+Na]+: 619.2487, detected: 619.2481.

Methyl 4-azido-3-O-benzyl-4,6-dideoxy-α-D-mannopyranosyl (1^2) 4-azido-3-O-benzyl-4,6dideoxy-α-D-mannopyranoside (15) .

Sodium methoxide (~1.2 ml, 0.5 M solution) was added to a solution of 14 (2.6 g, 4.36 mmol) in CH ₃ OH:THF [4:2] (20 ml) to reach a pH of ~9 and the resulting mixture was stirred in an argon atmosphere for 6 h at 21°C. Thereafter, the reaction mixture was neutralized with Amberlite IR 120 (H+) ion exchange resin, the resin was filtered off and washed successively with CH ₃ OH. The combined filtrate was concentrated in vacuo and purified by silica gel column chromatography (gradient elution with ethyl acetate-hexane) to give the deprotected disaccharide compound 15 (2.3 g, 95.4%) as a white foam. Analytical data for 15: Rf=0.4 (ethyl acetate/hexane 1:4.5, v/v); [a] _D ²¹ =+28.71 (c=1.56, CHCI3); ¹³ C NMR (176 MHz, CDCI3): δ: 137.5, 137.1, 128.6, 128.5, 128.3, 128.2 (x2), 128.0, 100.8, 99.9 , 77.8, 77.6, 73.6, 72.1 (x2), 67.3, 67.2, 66.9, 64.3, 63.8, 54.9, 18.6, 18, 4ppm; HRMS (ESI): m/z, calculated for C ₂₇ H ₃₄ N ₆ O ₇ Na [M+Na]+: 577.2381, detected: 577.2381.

Methyl-2-O-acetyl-4-azido-3-O-benzyl-4,6-dideoxy-α-D-mannopyranosyl (1^2) 4-azido-3-Obenzyl-4,6-dideoxy-α- D-mannopyranosyl (1^2) 4-azido-3-O-benzyl-4,6-dideoxy-α-D-mannopyranoside (16).

The glycosyl acceptor compound 15 (2.25 g, 4.06 mmol) and the glycosyl donor compound 11 (1.90 g, 4.46 mmol) were combined, azeotropically distilled twice with anhydrous toluene (5 mL) and placed under atmosphere high vacuum for 2 h. The mixture was then dissolved in CH2Cl2 (25 ml), treated with freshly activated 4 A molecular sieves (1.6 g), stirred under Ar at room temperature for 1 h. NIS (1.83 g , 8.11 mmol). After cooling to 10°C, TMSOTf (0.16 ml, 0.893 mmol) was added and the reaction mixture was allowed to warm to room temperature. When TLC indicated that the reaction was complete, saturated aqueous NaHCO3 (15 mL) and CH2Cl2 were added and the resulting mixture was passed through celite to remove molecular sieves. The combined filtrates were washed with an aqueous solution of Na ₂ S ₂ O ₃ (20%) [30 ml] and water (20 ml). After extracting the aqueous layer with CH ₂ Cl ₂ (3x15), the combined organic phase was dried over Na ₂ SO ₄ , concentrated in vacuo and purified by silica gel column chromatography (gradient elution with ethyl acetate/hexane) to give trisaccharide 16 (3.09 g, 88.9%) as a sticky liquid. Analytical data for 16: Rf=0.65 (ethyl acetate/hexane 1:5, v/v); ¹³ C NMR (176 MHz, CDCI3): δ: 169.7, 137.4, 137.3, 137.1, 128.5 (x2), 128.4, 128.1, 128.0 (x3), 100.3, 99.8, 99.1, 77.5, 76.8, 75.4, 73.5, 72.1, 72.0, 71.5, 67.8, 67.6, 67, 1, 67.0, 64.4, 64.0, 63.8, 54.9, 21.0, 18.6 (x2), 18.3 ppm; HRMS (ESI): m/z, calculated for C ₄₂ H ₅₁ N ₉ O ₁₁ Na [M+Na]+: 880.36, detected: 880.3607.

Methyl 4-azido-3-O-benzyl-4,6-dideoxy-α-D-mannopyranosyl (1^2) 4-azido-3-O-benzyl-4,6dideoxy-α-D-mannopyranosyl (1^ 2) 4-azido-3-O-benzyl-4,6-dideoxy-α-D-mannopyranoside (17).

Sodium methoxide (~1.5 ml, 0.5 M solution) was added to a solution of 16 (3.0 g, 3.5 mmol) in CH ₃ OH:THF [4:2] (20 ml) until pH ~9 and the resulting mixture was stirred in an argon atmosphere for 6 h at 21°C. The reaction mixture was then neutralized with Amberlite IR 120 (H+) ion exchange resin, the resin was filtered off and washed successively with CH3OH. The combined filtrate was concentrated in vacuo and purified by silica gel column chromatography (ethyl acetate-hexane gradient elution) to give the deprotected trisaccharide compound 17 (2.6 g, 91.2%) as a white solid foam.

Analytical data for 17: Rf=0.45 (ethyl acetate/hexane 1:5, v/v); ¹³ C NMR (176 MHz, CDC13): δ: 137.3 (x2), 137.2, 128.6 (x3), 128.5, 128.3, 128.3, 128.2 (x2), 128 .1 (x2), 128.0, 100.5, 100.4, 99.8, 77.6, 77.5, 76.8, 73.6, 73.3, 72.2, 72.1 ( x2), 67.8, 67.3, 67.1, 67.0, 64.4, 64.2, 63.8, 54.9, 18.6 (x2), 18.3 ppm; HRMS (ESI): m/z, calculated for C ₄₀ H ₄₉ N ₉ O ₁₀ Na [M+Na]+: 383.3495, detected: 838.3501.

- 30 040853

Methyl-2-O-acetyl-4-azido-3-O-benzyl-4,6-dideoxy-a-O-mannopyranosyl (1^2) 4-azido-3-Obenzyl-4,6-dideoxy-a- O-mannopyranosyl (1^2) 4-azido-3-O-benzyl-4,6-dideoxy-a-O-mannopyranosyl (1^2) 4-azido-3-O-benzyl-4,6-dideoxy- a-O-mannopyranoside (18).

The glycosyl acceptor compound 17 (2.05 g, 2.51 mmol) and the glycosyl donor compound 11 (1.18 g, 2.76 mmol) were combined, azeotropically distilled twice with anhydrous toluene (5 ml) and placed under atmosphere high vacuum for 2 h. The mixture was then dissolved in CH ₂ Cl ₂ (20 ml), treated with freshly activated 4 A molecular sieves (1.2 g), stirred under Ar at room temperature for 1 h. NIS (1 .13 g, 5.02 mmol). After cooling to 10°C, TMSOTf (0.1 ml, 0.553 mmol) was added and the reaction mixture was allowed to warm to room temperature. When the reaction was seen to be complete by TLC, saturated aqueous NaHCO ₃ (10 ml) and CH ₂ Cl ₂ were added and the resulting mixture was passed through celite to remove molecular sieves. The combined filtrates were washed with aqueous Na ₂ S ₂ O ₃ (20%) and water. After extraction of the aqueous layer with CH ₂ Cl ₂ (3x10), the combined organic phase was dried over Na ₂ SO ₄ , concentrated in vacuo and purified by silica gel column chromatography (gradient elution with ethyl acetate/hexane) to give tetrasaccharide 18 (2.49 g, 87.8%) as a syrup. Analytical data for 18: Rf=0.5 (ethyl acetate/hexane 1:4, v/v); ¹³ C NMR (176 MHz, CDCl ,): δ: 169.8, 137.4, 137.3, 137.1 (x2), 128.6 (x2), 128.5 (x2), 128.4 , 128.3, 128.2, 128.1, 128.0 (x3), 100.3, 100.1, 99.7, 99.1, 77.4, 76.6, 75.4, 73, 6, 73.4 (x2), 72.2, 72.1, 72.0, 71.5, 67.8, 67.6, 67.1, 66.9, 64.3, 64.2, 64 .0, 63.8, 54.9, 21.0, 18.6 (x2), 18.5, 18.4 ppm; HRMS (ESI): m/z, calculated for C ₅₅ H ₆₆ N ₁ 2O ₁ 4Na [M+Na]+: 1141.4714, detected: 1141.473.

Methyl-4-azido-3-O-benzyl-4,6-dideoxy-a-O-mannopyranosyl (1^2) 4-azido-3-O-benzyl-4,6dideoxy-a-O-mannopyranosyl (1^ 2) 4-azido-3-O-benzyl-4,6-dideoxy-α-D-mannopyranosyl (1^2) 4-azido-3-O-benzyl-4,6-dideoxy-α-D-mannopyranoside (19) .

Sodium methoxide (~1.2 ml, 0.5 M solution) was added to a solution of 18 (2.2 g, 1.95 mmol) in CH ₃ OH:THF [4:2] (15 ml) until pH ~9 and the resulting mixture was stirred in an argon atmosphere for 6 h at 21°C. The reaction mixture was then neutralized with Amberlite IR 120 (H+) ion exchange resin, the resin was filtered off and washed successively with CH3OH. The combined filtrate was concentrated in vacuo and purified by silica gel column chromatography (gradient elution with ethyl acetate-hexane) to give the title compound 19 (1.86 g, 88.7%) as a white solid. Analytical data for 19: Rf=0.4 (ethyl acetate/hexane 1:4, v/v); ¹³ C NMR (176 MHz, CDC13): δ: 137.3 (x2), 137.1, 128.6 (x2), 128.5, 128.4, 128.3 (x2), 128.2 (x3 ), 128.1, 128.0, 100.4, 100.3, 100.2, 99.7, 77.7, 77.4, 76.6, 73.6, 73.5, 73.2, 72.2, 72.1 (x3), 67.8, 67.3, 67.1, 66.9, 64.3, 64.2 (x2), 63.8, 54.9, 18.6 ( x2), 18.5, 18.3 ppm; HRMS (ESI): m/z, calculated for C ₅₃ H ₆ 4N ₁₂ O ₁₃ Na [M+Na]+: 1099.4608, detected: 1099.4625.

Methyl-2-O-acetyl-4-azido-3-O-benzyl-4,6-dideoxy-a-O-mannopyranosyl (1^2) 4-azido-3-Obenzyl-4,6-dideoxy-a- O-mannopyranosyl (1^2) 4-azido-3-O-benzyl-4,6-dideoxy-a-O-mannopyranosyl (1^2) 4-azido-3-O-benzyl-4,6-dideoxy- a-O-mannopyranosyl (1^2) 4-azido-3-O-benzyl-4,6-dideoxy-aD-mannopyranoside (20).

The glycosyl acceptor compound 19 (1.63 g, 1.51 mmol) and the glycosyl donor compound 11 (0.712 g, 1.66 mmol) were combined, azeotropically distilled twice with anhydrous toluene (5 ml) and placed under high vacuum for 2 h. Then the mixture was dissolved in CH ₂ Cl ₂ (15 ml), treated with freshly activated 4 A molecular sieves (1 g), stirred under Ar atmosphere at room temperature for 1 h. NIS (0.681 g, 3. 03 mmol). After cooling to -10°C, TMSOTf (0.06 ml, 0.33 mmol) was added and the reaction mixture was allowed to warm to room temperature. When the reaction was seen to be complete by TLC, saturated aqueous NaHCO3 (10 ml) and CH2Cl2 were added and the resulting mixture was passed through celite to remove molecular sieves. The combined filtrates were washed with aqueous Na2S2O3 (20%) [15 ml] and water (15 ml). After extraction of the aqueous layer with CH ₂ Cl ₂ (3x10), the combined organic phase was dried over Na2SO4, concentrated in vacuo and purified by silica gel column chromatography (gradient elution with ethyl acetate/hexane) to give pentasaccharide 20 (1.92 g, 91.9 %) in the form of a sticky liquid. Analytical data for 20: Rf=0.7 (ethyl acetate/hexane 1:4, v/v); ¹³ C NMR (176 MHz, CDCI3): δ: 169.8, 137.4, 137.3, 137.2, 137.1, 128.6 (x3), 128.5 (x2), 128.4, 128.3 (x2), 128.2, 128.1 (x2), 128.0 (x3), 100.3, 100.2, 100.0, 99.7, 99.1, 77.4, 76 .6, 76.5, 75.4, 73.7, 73.6, 73.4, 73.3, 72.2 (x2), 72.1, 72.0, 71.5, 67.8 ( x2), 67.6, 67.1, 66.9, 64.3, 64.2 (x2), 64.1, 63.8, 54.9, 21.0, 18.6 (x2), 18 .5 (x2), 18.4 ppm; HRMS (ESI): m/z, calculated for C ₆₈ H ₈₁ N ₁₅ O ₁₇ Na [M+Na]+: 1402.5827, detected: 1402.5856.

- 31 040853

Methyl-4-azido-3-O-benzyl-4,6-dideoxy-a-O-mannopyranosyl (1^2) 4-azido-3-O-benzyl-4,6dideoxy-a-O-mannopyranosyl (1^ 2) 4-azido-3-O-benzyl-4,6-dideoxy-a-O-mannopyranosyl (1^2) 4azido-3-O-benzyl-4,6-dideoxy-a-O-mannopyranosyl (1^ 2) 4-azido-3-O-benzyl-4,6-dideoxy-a-omannopyranoside (21).

Sodium methoxide (~1.2 ml, 0.5 M solution) was added to a solution of 20 (1.8 g, 1.31 mmol) in CH ₃ OH:THF [4:2] (15 ml) until pH ~9 and the resulting mixture was stirred in an argon atmosphere for 6 h at 21°C. The reaction mixture was then neutralized with Amberlite IR 120 (H+) ion exchange resin, the resin was filtered off and washed successively with CH3OH. The combined filtrate was concentrated in vacuo and purified by silica gel column chromatography (gradient elution with ethyl acetate-hexane) to give the title compound 21 (1.57 g, 89.8%) as a white foam. Analytical data for 21: Rf=0.55 (ethyl acetate/hexane 1:4, v/v); ¹³ C NMR (176 MHz, CDCI3): δ: 137.3 (x2), 137.2, 129.0, 128.6 (x4), 128.4, 128.3 (x4), 128.2 (x2 ), 128.1, 128.0, 100.5, 100.3, 100.2 (χ2), 99.7, 77.7, 77.4, 77.0, 76.6, 76.5, 73 .7, 73.6, 73.4, 73.2, 72.2, 72.1 (x3), 67.8 (x2), 67.3, 67.1, 66.9, 64.4, 64 .2, 63.8, 54.9, 18.6 (x2), 18.5 (x2), 18.3 ppm; HRMS (ESI): m/z, calculated for C ₆₆ H ₇₉ N ₁₅ O ₁₆ Na [M+Na]+: 1360.5721, detected: 1360.5749.

Methyl-2-O-acetyl-4-azido-3-O-benzyl-4,6-dideoxy-a-O-mannopyranosyl (1^2) 4-azido-3-Obenzyl-4,6-dideoxy-a- O-mannopyranosyl (1^2) 4-azido-3-O-benzyl-4,6-dideoxy-a-O-mannopyranosyl (1^2) 4-azido-3-O-benzyl-4,6-dideoxy- a-O-mannopyranosyl (1^2) 4-azido-3-O-benzyl-4,6-dideoxy-aD-mannopyranosyl (1^2) 4-azido-3-O-benzyl-4,6-dideoxy- a-O-mannopyranoside (22).

The glycosyl acceptor compound 21 (1.45 g, 1.08 mmol) and the glycosyl donor compound 11 (0.556 g, 1.3 mmol) were combined, azeotropically distilled twice with anhydrous toluene (5 ml) and placed under high vacuum for 2 h. The mixture was then dissolved in CH ₂ Cl ₂ (15 ml), treated with freshly activated 4 A molecular sieves (1 g), stirred under Ar at room temperature for 1 h. NIS (0.488 g, 2. 16 mmol). After cooling to -10°C, TMSOTf (43 μl, 0.24 mmol) was added and the reaction mixture was allowed to warm to room temperature. When the reaction was seen to be complete by TLC, saturated aqueous NaHCO ₃ (10 ml) and CH ₂ Cl ₂ were added and the resulting mixture was passed through celite to remove molecular sieves. The combined filtrates were washed with an aqueous solution of Na ₂ S ₂ O ₃ (20%) [10 ml] and water (15 ml). After extracting the aqueous layer with CH ₂ Cl ₂ (3×10), the combined organic phase was dried over Na ₂ SO ₄ , concentrated in vacuo and purified by silica gel column chromatography (gradient elution with ethyl acetate/hexane) to give hexasaccharide 22 (1.601 g, 90. 1%) as a sticky liquid. Analytical data for 22: Rf=0.65 (ethyl acetate/hexane 1:4, v/v); ¹³ C NMR (176 MHz, CDCI3): δ: 169.8, 137.4, 137.3, 137.2, 137.1 (x3), 128.6 (x4), 128.5 (x2), 128 .4, 128.3 (x2), 128.2, 128.1 (x2), 128.0 (x3), 100.3, 100.1 (x2), 100.0, 99.7, 99.1 , 77.4, 76.7 (x2), 76.5, 75.4, 73.6 (x2), 73.5, 73.4, 73.3, 72.2, 72.1, 72.0 , 71.5, 67.8 (x4), 67.6, 67.1, 66.9, 64.3 (x2), 64.2 (x2), 64.1, 63.8, 54.9, 21.0, 18.6 (x2), 18.5 (x3), 18.4 ppm; HRMS (ESI): m/z, calculated for C ₈₁ H ₉₆ N ₁₈ O ₂₀ Na [M+Na]+: 1663.694, detected: 1663.6982.

Methyl-4-azido-3-O-benzyl-4,6-dideoxy-a-O-mannopyranosyl (1^2) 4-azido-3-O-benzyl-4,6dideoxy-a-O-mannopyranosyl (1^ 2) 4-azido-3-O-benzyl-4,6-dideoxy-a-O-mannopyranosyl (1^2) 4azido-3-O-benzyl-4,6-dideoxy-a-O-mannopyranosyl (1^ 2) 4-azido-3-O-benzyl-4,6-dideoxy-a-Omannopyranosyl (1^2) 4-azido-3-O-benzyl-4,6-dideoxy-a-O-mannopyranoside (23) .

Sodium methoxide (~1.0 ml, 0.5 M solution) was added to a solution of 22 (1.3 g, 0.792 mmol) in CH ₃ OH:THF [4:2] (15 ml) until a pH of ~9 was reached and the resulting the mixture was stirred in an argon atmosphere for 6 h at 21°C. Thereafter, the reaction mixture was neutralized with Amberlite IR 120 (H+) ion exchange resin, the resin was filtered off and washed successively with CH ₃ OH. The combined filtrate was concentrated in vacuo and purified by silica gel column chromatography (ethyl acetate-hexane gradient elution) to give the title compound 23 (1.17 g, 92.3%) as an oil. Analytical data for 23: Rf=0.5 (ethyl acetate/hexane 1:4, v/v ¹³ C NMR (126 MHz, CDC13): δ: 137.3, 137.2 (x2), 128.7 ( x3), 128.6 (x2), 128.4 (x3), 128.3 (x2), 128.2, 128.1 (x3), 100.5, 100.3, 100.2 (x2), 100.1, 99.8, 77.7, 77.5, 76.6 (x2), 73.7, 73.6, 73.5 (x2), 73.3, 72.2 (x2), 72 .1, 67.9, 67.8, 67.4, 67.2, 67.0, 64.4, 64.2, 63.9, 54.9, 18.7, 18.6 (x2), 18.5 (x2), 18.3 ppm HRMS (ESI): m/z, calculated for C ₇₉ H ₉₄ N ₁₈ O ₁₉ Na [M+Na]+: 1621.6835, detected: 1621.688.

Methyl-2,3-di-O-benzoyl-4-O-(5'-N-benzyl-5'-N-carboxybenzylpentanyl)-6-deoxy-α-Dmannopyranosyl (1^2) 4-azido-3- O-benzyl-4,6-dideoxy-a-O-mannopyranosyl (1^2) 4-azido-3-O-benzyl4,6-dideoxy-a-O-mannopyranosyl (1^2) 4-azido-3- O-benzyl-4,6-dideoxy-a-O-mannopyranosyl (1^2) 4-azido-3-O-benzyl-4,6-dideoxy-a-O-mannopyranosyl (1^2) 4-azido- 3-O-benzyl-4,6-dideoxy-a-Omannopyranosyl (1^2) 4-azido-3-O-benzyl-4,6-dideoxy-a-O-mannopyranoside (24).

The glycosyl acceptor compound 23 (0.270 g, 0.169 mmol) and the glycosyl donor compound 13 (0.146 g, 0.186 mmol) were combined, azeotropically distilled twice with anhydrous toluene (5 mL), and placed under high vacuum for 2 h. The mixture was then dissolved in CH ₂ Cl ₂ (10 ml), treated with freshly activated 4 A molecular sieves (0.3 g), stirred under Ar atmosphere at room temperature for 1 h. NIS (0.076 g, 0.337 mmol ). After cooling to -10°C, TMSOTf (6.4 μl, 0.037 mmol) was added and the reaction mixture was allowed to warm to room temperature. When TLC showed that the reaction was complete, saturated aqueous NaHCO ₃ (5 ml) and CH2Cl2 were added and the resulting mixture was passed through celite to remove molecular sieves. The combined filtrates were washed with aqueous Na2S2O ₃ (20%) [10 ml] and water (10 ml). After extracting the aqueous layer with CH ₂ Cl ₂ (3x5), the combined organic phase was dried over Na ₂ SO ₄ , concentrated in vacuo and purified by silica gel column chromatography (gradient elution with ethyl acetate/hexane) to give heptasaccharide 24 (0.334 g, 87. 4%) as a sticky liquid. Analytical data for 24: Rf=0.65 (ethyl acetate/hexane 1:4, v/v); [a] _D ²¹ = -6.71° (c=1.23, CHC13); ¹³ C NMR (126 MHz, CDCL3): δ: 165.3, 165.1, 156.6, 156.1, 137.9, 137.5, 137.4, 137.3, 137.2 (x2) , 136.8, 133.3, 133.0 (x2), 129.9, 129.8, 129.6, 129.1, 128.7 (x2), 128.6 (x2), 128.5 ( x2), 128.4 (x2), 128.3 (x2), 128.2, 128.1, 127.9, 127.8, 127.3, 125.3, 100.4 (x3), 100, 2 (x2), 99.8, 99.1, 79.8, 77.5, 76.7, 76.6, 73.7, 73.6, 73.5, 73.3, 72.2 (x2 ), 72.1 (x2), 71.9, 70.9, 68.5, 68.1, 67.9, 67.8, 67.1, 67.0, 64.4, 64.2, 63 .9, 54.9, 50.5, 50.2, 47.0, 46.1, 29.7, 29.4, 27.8, 27.4, 23.3, 18.7, 18.6 (x3), 18.5 (x2), 18.0 ppm; HRMS (ESI): m/z, calculated for C ₁₁₉ H ₁₃₅ N ₁₉ O ₂₇ Na [M+Na]+: 2284.9667, detected: 2284.9732.

Methyl 4-O-(5'-N-benzyl-5'-N-carboxybenzylpentanyl)-6-deoxy-α-D-mannopyranosyl (1^2) 4azido-3-O-benzyl-4,6-dideoxy- α-D-mannopyranosyl (1^2) 4-azido-3-O-benzyl-4,6-dideoxy-a-Omannopyranosyl (1^2) 4-azido-3-O-benzyl-4,6-dideoxy- α-D-mannopyranosyl (1^2) 4-azido-3-O-benzyl4,6-dideoxy-a-O-mannopyranosyl (1^2) 4-azido-3-O-benzyl-4,6-dideoxy- α-D-mannopyranosyl (1^2) 4-azido-3-O-benzyl-4,6-dideoxy-α-D-mannopyranoside (25).

Sodium methoxide (~0.2 ml, 0.5 M solution) was added to a solution of 24 (0.26 g, 0.115 mmol) in CH ₃ OH:THF [2:3] (10 ml) until a pH of ~9 was reached and the resulting the mixture was stirred under argon for 6 hours at 21°C. The reaction mixture was then neutralized with Amberlite IR 120 (H+) ion exchange resin, the resin was filtered off and washed successively with CH ₃ OH. The combined filtrate was concentrated in vacuo and purified by silica gel column chromatography (ethyl acetate-hexane gradient elution) to give the title compound 25 (0.215 g, 91.2%) as an oil. Analytical data for 25: Rf=0.25 (ethyl acetate/hexane 1:3.3, v/v); [a] _D ²¹ =+79.2 (c=2.21, CHCl·,); ¹³ C NMR (176 MHz, CDCI3): δ: 156.7, 156.3, 137.8, 137.4, 137.3, 137.2 (x2), 137.1 (x2), 136.7, 129.0, 128.6 (x2), 128.5, 128.4, 128.3 (x2), 128.2, 128.1, 128.0, 127.9, 127.8, 127.3, 100.8, 100.4, 100.3, 100.2, 100.1 (x2), 99.7, 81.6, 77.4, 76.5, 73.6, 73.6, 73.5 , 73.5, 72.9, 72.2, 72.1, 72.1, 72.0, 71.7, 71.1, 68.2, 67.8, 67.7, 67.2, 66 .9, 64.3, 64.2, 54.9, 50.5, 50.3, 47.1, 46.1, 29.7, 29.4, 27.9, 27.2, 23.3 , 18.6 (x2), 18.5 (x4), 17.9 ppm; HRMS (ESI): m/z, calculated for C ₁₀ 5H ₁₃₁ N ₂₀ O ₂₅ [M+NH ₄ ]+: 2071.9589, found: 2071.9639.

Methyl 4-O-(5'-aminopentanyl)-6-deoxy-α-D-mannopyranosyl (1^2) 4,6-dideoxy-4-formamidoa-D-mannopyranosyl (1^2) 4,6-dideoxy -4-formamido-α-D-mannopyranosyl (1^2) 4,6-dideoxy-4formamido-a-O-mannopyranosyl (1^2) 4,6-dideoxy-4-formamido-α-D-mannopyranosyl (1 ^2) 4,6dideoxy-4-formamido-a-O-mannopyranosyl (1^2) 4,6-dideoxy-4-formamido-a-O-mannopyranoside (8).

A stirred solution of 25 (0.11 g, 0.054 mmol) in a mixture of pyridine (5 ml) and water (2 ml) was sparged with H2S for 0.5 h at 40°C and continued stirring for 16 h. After that, argon was sparged within 10 min, the solvents were removed under vacuum and the residue was co-evaporated with toluene (3x10 ml) and dried. Mass spectrometric analysis showed completion of the reaction to give the corresponding amine compound and no other products resulting from incomplete reduction were found. HRMS (ESI): m/z, calculated for C105H140N7O25 [M+H]+: 1898.9893, detected: 1898.99.

This crude material was directly used for formylation. Freshly prepared formic anhydride (5 ml, ether solution) was added to the amine compound in CH3OH (5 ml) at -20°C and stirred for 3 h, then allowed to warm slowly to 21°C. Thereafter, the solvents were evaporated and the residue was subjected to silica gel column chromatography (gradient elution with methanol-dichloromethane) to obtain a heptasaccharide. High resolution mass spectrometric analysis showed completion of the formylation reaction. HRMS (ESI): m/z, calculated for C ₁₁₁ H ₁₃₉ N ₇ O ₃₁ Na [M+Na]+: 2088.9408, detected: 2088.9405.

The formylated compound was dissolved in CH3OH/H2O (2:1, 10 ml), Pd(OH) ₂ on charcoal (20%, 0.060 g) was added. It was then stirred under hydrogen gas pressure at 21° C. for 16 hours. After filtration through a Celite pad, it was washed with CH 3 OH (3 x 10 ml) and the solvents were removed in vacuo. The residue was purified by reverse phase HPLC on a C18 column in a water-acetonitrile gradient and lyophilized to give the title compound 8 (0.0427 g, 61.2%, over 3 steps) as a white foam. Analytical data for 8: [a] _D ²¹ =+42.44 (c=1.02, H ₂ O); ¹³ C NMR (126 MHz, CDCl ,): δ: 168.8 (x2), 168.6, 165.9 (x4), 165.7, 103.2, 103.1 (x4), 102.7 , 102.5, 101.5 (x2), 100.4, 100.3, 81.8, 78.2, 78.1 (x3), 78.0 (x2), 77.9, 73.5 ( x2), 71.3, 70.8 (x2), 69.2, 68.7 (x2), 68.5, 68.4, 67.8,

- 33 040853

57.9, 56.4, 55.9, 55.8 (χ2), 52.9 (χ2), 52.7 (χ2), 40.4, 29.7, 27.5, 23.2, 17 .9 (χ2), 17.8 (χ2), 17.7, 17.6 (χ4) ppm;

HRMS (ESI): m/z, calculated for C ₅₄ H ₉₂ N ₇ O ₂₉ [M+H]+: 1302.5934, detected: 1302.5928.

Methyl 4-O-(5'-[M-succinimidyl]glutarylamidopentanyl)-6-deoxy-a-O-mannopyranosyl (1^2)

4,6-dideoxy-4-formamido-a-O-mannopyranosyl (1^2) 4,6-dideoxy-4-formamido-a-omannopyranosyl (1^2) 4,6-dideoxy-4-formamido-a- O-mannopyranosyl (1^2) 4,6-dideoxy-4-formamido-a-O-mannopyranosyl (1^2) 4,6-dide ezoxy-4-formamido-a-O-mannopyranosyl (1^2) 4,6dideoxy -4-formamido-a-O-mannopyranoside (S26).

A mixture of heptasaccharide 8 (9 mg) and disuccinimide glutarate (15 eq.) in DMF and 0.1 M PBS buffer (4:1, 1.5 ml) was stirred at room temperature for 6 h. The reaction mixture was concentrated in vacuo and the residue was washed with EtOAc 10 times to remove excess disuccinimide glutarate. The resulting solid was dried under vacuum for 1 hour to give activated oligosaccharide S26, which was directly used for conjugation with BSA and tetanus toxoid. MALDI TOF MS (in positive ion mode): calculated for C ₆₃ H ₁₀₀ N ₈ O ₃₄ Na [M+Na]+ m/z, 1535.6342; identified, 1535.9996.

1-[(2'-Aminoethylamido)carbonylpentyl)-6-deoxy-a-O-mannopyranosyl (1^2) 4,6-dideoxy-4formamido-a-O-mannopyranosyl (1^2) 4,6-dideoxy- 4-formamido-a-O-mannopyranosyl (1^2) 4,6dideoxy-4-formamido-a-O-mannopyranosyl (1^2) 4,6-dide ezoxy-4-formamido-a-O-mannopyranosyl (1 ^2) 4,6-dideoxy-4-formamido-a-O-mannopyranosyl (1^2) 4,6-dideoxy-4-formamido-a-omannopyranoside] 2-butoxycyclobutene-3,4-dione (S27).

To a stirred solution of heptasaccharide 8 (0.006 g, 0.005 mmol) in water (0.5 ml) and EtOH (0.5 ml) was added a solution of 3,4-dibutoxy-3-cyclobutene-1,2-dione (20% in ethanol , 50 μl) and the pH was adjusted to 8 by careful addition of an aqueous solution of NaHCO ₃ (1%). After 1 hour, mass spectrometry showed that the reaction was complete; the reaction mixture was neutralized with CH3COOH (10%) and concentrated in vacuo. The residue was purified by reverse phase HPLC on a C18 column in a water-acetonitrile gradient and lyophilized to give the title compound S27 (0.005 g, 72.6%) as a white foam. Analytical data for S27: ¹³ C NMR (126 MHz, CDC13): δ: 190.3, 184.2, 183.8, 178.4, 178.0, 174.3, 168.8, 168.7, 168 .6, 165.9, 165.7, 103.2, 103.2 (χ4), 102.8, 102.6, 101.6, 100.4, 100.3, 81.9, 78.2, 78.1 (χ2), 78.0 (χ2), 77.9, 73.4, 71.3, 70.7, 69.8, 69.2, 69.1, 68.8, 68.6, 68.5, 68.4, 57.9, 56.4, 55.9, 55.8, 52.9 (χ2), 52.7, 40.4, 32.4, 30.7, 30.5 , 29.7, 27.5, 23.3, 23.2, 19.2, 19.0, 17.9, 17.8 (χ2), 17.7 (χ4), 17.6, 13.9 ppm HRMS (ESI): m/z, calculated for C ₆₂ H ₉₉ N ₇ O ₃₂ Na [M+Na]+: 1476.6335, detected: 1476.6406.

Conjugation of an oligosaccharide with a protein.

Preparation of tetanus toxoid conjugate 9 (structure XVI): activated heptasaccharide S26 (0.8 mg, 0.518 μmol) was added to a solution of tetanus toxoid (4 mg, 0.026 μmol) in 0.5 M borate buffer pH 9 (1 ml) and slowly stirred at 21°C for 3 days. The reaction mixture was then washed with PBS buffer, filtered through an ultrafine filtration tube (10,000 MWCO, 4 x 10 ml) and the resulting tetanus toxoid conjugate 9 was stored in PBS buffer. Mass spectrometric analysis of MALDI-TOF showed that conjugate 9 had an average of 10.02 heptasaccharides per tetanus toxoid molecule.

Preparation of conjugate 10 with BSA (structure XVII): BSA (10 mg) and activated heptasaccharide S27 (4.5 mg) were dissolved in 0.1 M PBS buffer, pH 9 (1.2 ml), and slowly stirred at 21° C within 3 days. The reaction mixture was then diluted with Mili-Q water, filtered through an ultrafine filtration tube (10,000 MWCO, 4 x 10 mL), lyophilized to give BSA conjugate 10 as a white foam (12.2 mg). Mass spectrometric analysis of MALDI-TOF showed that conjugate 10 had an average of 10.27 heptasaccharides per BSA molecule.

Synthesis of a trisaccharide with exclusively 1,2 bonds.

Ethyl-4-azido-2,3-di-O-benzoyl-4,6-dideoxy-a-O-mannopyranosyl (1^2) 4-azido-3-O-benzyl-4,6dideoxy-1-thio- a-O-mannopyranoside (S18).

The analytical data for the title compound were in fact the same as previously described (Bundle et al (1988) Carbohydr. Res. 174, 239-251).

5'-Methoxycarbonylpentyl-4-azido-3-O-benzyl-4,6-dideoxy-a-O-mannopyranoside (S19).

The analytical data for the title compound were in fact the same as previously described (Ganesh et al (2014) J. Amer. Chem. Soc. 136, 16260-16269.

5'-Methoxycarbonylpentyl-4-azido-2,3-O-benzoyl-4,6-dideoxy-a-O-mannopyranosyl (1^2) 4azido-3-O-benzyl-4,6-dideoxy-a-O -mannopyranosyl (1 ^2) 4-azido-3-O-benzyl-4,6-dideoxy-a-omannopyranoside (S20).

The glycosyl acceptor compound S19 (0.2 g, 0.491 mmol) and the glycosyl donor compound 11 (0.414 g, 0.589 mmol) were combined, azeotropically distilled twice with anhydrous toluene (5 ml), and placed under high vacuum for 2 h. The mixture was then dissolved in CH ₂ Cl ₂ (15 ml), treated with freshly activated 4 A molecular sieves (1 g), stirred under Ar at

- 34 040853 room temperature for 1 hour. NIS (0.221 g, 0.982 mmol) was added to the mixture. After cooling to -10°C, TMSOTf (19.5 μl, 0.108 mmol) was added and the reaction mixture was allowed to warm to room temperature. When TLC showed that the reaction was complete, a saturated aqueous solution of NaHCO ₃ (5 ml) and CH ₂ Cl ₂ was added and the resulting mixture was passed through celite to remove molecular sieves. The combined filtrates were washed with aqueous Na ₂ S ₂ O ₃ (20%) and water. After extracting the aqueous layer with CH ₂ Cl ₂ (3x5), the combined organic phase was dried over Na ₂ SO ₄ , concentrated in vacuo and purified by silica gel column chromatography (gradient elution with ethyl acetate-hexane) to give the disaccharide S20 (0.418 g, 81. 3%) as a sticky liquid. Analytical data for S20: Rf=0.7 (ethyl acetate/hexane 1:4.5, v/v); [a] _D ²¹ = -14.49° (c=1.79, CHC13); ¹ H NMR (500 MHz, CDCl,): δ 8.02-8.05 (m, 2H, ArH), 7.95-7.97 (m, 2H, ArH), 7.64-7.68 ( m, 1H, ArH), 7.50-7.57 (m, 3H, ArH), 7.33-7.41 (m, 8H, ArH), 7.22-7.26 (m, 3H, ArH ), 7.13-7.17 (m, 1H, ArH), 5.71 (dd, J=3.3, 1.5 Hz, 1H, H-2s), 5.59 (dd, J=10 .3, 3.3 Hz, 1H, H-3c), 5.06 (d, J=1.8 Hz, 1H, H-1c), 5.02 (d, J=1.8 Hz, 1H, H-1c), 4.76 (d, J=11.7 Hz, 1H, CHPh), 4.62-4.69 (m, 4H, 3 CHPh, H-1a), 3.95 (dd, J =2.2, 0.7 Hz, 1H, H-2c), 3.90 (dd, J=2.2, 0.7 Hz, 1H, H-2a), 3.76-3.81 (m , 2H, H-Sv, H-5c), 3.74 (dd, J=9.9, 2.9 Hz, 1H, H-3a), 3.71 (s, 3H), 3.69 (t , J=9.9 Hz, 1H, H-4s), 3.55-3.65 (m, 3H, H-4v, H-5s, -O-CHh), 3.43-3.49 (m , 1H, H-5a), 3.38 (dt, J=9.7, 6.4 Hz, 1H, -O-CH2a), 3.27 (t, J=9.9 Hz, 1H, H- 4a), 2.33-2.39 (m, 2H, CH2f), 1.64-1.72 (m, 2H, -CH2e), 1.56-1.64 (m, 2H, -CH2c), 1.35-1.42 (m, 2H, -CH2d), 1.38 (d, J=5.6 Hz, 3H, H6c), 1.32 (d, J=5.9 Hz, 3H, H -6c), 1.29 (d, J=5.9 Hz, 3H, H-6a); ¹³ C NMR (126 MHz, CDCl,): δ 174.0, 165.2, 164.9, 137.5, 137.3, 133.4, 133.3, 129.8 (x2), 129.6 , 129.3, 128.5 (x2), 128.4, 128.2, 128.1 (x2), 128.0, 100.3, 99.0, 98.8, 77.9, 73.9 , 73.5, 72.3 (x2), 70.9, 69.5, 68.0, 67.5, 67.2, 64.5, 63.9, 63.5, 51.5, 34, 0, 29.1, 25.7, 24.7, 18.6(x2), 18.4ppm; HRMS (ESI): m/z, calculated for C53H ₆₁ N ₉ O ₁₄ Na [M+Na]+: 1070.423, detected: 1070.4248.

5'-Methoxycarbonylpentyl 4-azido-4,6-dideoxy-α-D-mannopyranosyl (1^2) 4-azido-3-O-benzyl4,6-dideoxy-α-D-mannopyranosyl (1^2) 4- azido-3-O-benzyl-4,6-dideoxy-α-D-mannopyranoside (S21).

Sodium methoxide (~0.3 ml, 0.5 M solution) was added to a solution of S20 (0.39 g, 0.372 mmol) in CH ₃ OH:THF [4:2] (12 ml) until a pH of ~9 was reached and the resulting the mixture was stirred under argon for 6 hours at 21°C. The reaction mixture was then neutralized with Amberlite IR 120 (H+) ion exchange resin, the resin was filtered off and washed successively with CH ₃ OH. The combined filtrate was concentrated in vacuo and purified by silica gel column chromatography (gradient elution with ethyl acetate-hexane) to give the title compound S21 (0.299 g, 95.6%) as a white solid. Analytical data for S21: Rf=0.3 (ethyl acetate/hexane 1:1.5, v/v); [a] _D ²¹ =+84.18 (c=1.55, CHCl3); 1H NMR (500 MHz, CDCl3): δ 7.30-7.44 (m, 10H, ArH), 5.00 (d, J=1.8 Hz, 1H, H-1c), 4.90 (d , J=1.5 Hz, 1H, H-1s), 4.72 (d, J=11.4 Hz, 1H, CHPh), 4.61-4.67 (m, 4H, 3 CHPh, H- 1a), 3.93-3.97 (m, 2H, H-2c, H-2c), 3.81-3.87 (m, 2H, H-2a, H-3a), 3.76 (dd , J=9.9, 2.9Hz, 1H, H-3b), 3.73 (dd, J=10.0, 2.9Hz, 1H, H-3c), 3.70 (s, 3H ), 3.51-3.64 (m, 3H, H-5b, H-5c, -O-CH2b), 3.43-3.49 (m, 1H, H-5a), 3.40 (t , J=9.9 Hz, 1H, H-4s), 3.36 (dt, J=9.7, 6.4 Hz, 1H, -O-CH2a), 3.27 (t, J=9, 9 Hz, 1H, H-4c), 3.40 (t, J=10.2 Hz, 1H, H-4a), 2.49 (d, J=6.9 Hz, 1 OHg), 2.34 (t, J=7.4 Hz, 2H, -CH2f), 2.18 (d, J=3.9 Hz, 1 OH2c, 1.63-1.70 (m, 2H, -CH2e), 1, 54-1.61 (m, 2H, -CH2c), 1.33-1.40 (m, 2H, -CH2d), 1.30 (d, J=6.2 Hz, 6H, H-6c, H -6c ⁾ , 1.20 (d, J=6.2 Hz, 3H, H-6a); (x2), 128.3, 128.2 (x2), 128.1, 100.7, 100.4, 98.7, 77.7, 77.2, 77.2, 73.8, 73.2 , 72.3, 72.2, 70.2, 69.9, 67.8, 67.5, 67.4, 67.1, 65.8, 64.4, 64.2, 51.6, 33 .9, 29.1, 25.7, 24.7, 18.6(x2), 18.2pp m; HRMS (ESI): m/z, calculated for C39H ₅₃ N ₉ O ₁₂ Na [M+Na]+: 862.3706, detected: 862.3705.

5'-Methoxycarbonylpentyl-4,6-dideoxy-4-formamido-α-D-mannopyranosyl (1^2) 4,6-dideoxy4-formamido-α-D-mannopyranosyl (1^2) 4,6-dideoxy-4 -formamido-α-D-mannopyranoside (4).

A stirred solution of S21 (0.2 g, 0.239 mmol) in a mixture of pyridine (5 ml) and water (2 ml) was sparged with H 2 S for 0.5 h at 40°C and continued stirring for 16 h. After that, argon was sparged within 10 min, the solvents were removed under vacuum and the residue was co-evaporated with toluene (3x10 ml) and dried. Mass spectrometric analysis showed completion of the reaction to give the corresponding amine compound and no other products resulting from incomplete reduction were found.

This crude material was directly used for formylation. Freshly prepared formic anhydride (5 ml, ether solution) was added to the amine compound in CH ₃ OH (5 ml) at -20°C and stirred for 3 h, then slowly allowed to warm to 21°C. Thereafter, the solvents were evaporated and the residue was subjected to silica gel column chromatography (gradient elution with methanol-dichloromethane) to give the trisaccharide. High resolution mass spectrometric analysis showed completion of the formylation reaction. HRMS (ESI): m/z, calculated for C ₄₂ H ₅ 9N3O1 ₅ Na [M+Na]+: 868.3838, detected: 868.3837.

The formylated compound was dissolved in CH ₃ OH/H ₂ O (2:1, 15 ml), Pd(OH) ₂ on charcoal (20%, 0.090 g) was added. It _was then stirred under hydrogen gas pressure at 21° C. for 16 h.

- 35 040853 creators under vacuum. The residue was purified by reverse phase HPLC on a C18 column in a water-acetonitrile gradient and lyophilized to give the title compound 4 (0.094 g, 59.3%, over 3 steps) as a white foam. Analytical data for 4: [a] _D ²¹ =+31.58 (c=1.16, H2O); 1H NMR (700 MHz, D2O): δ 8.20-8.24 (Z) and 8.03-8.0607 (E) (m, 3H, NCHO), 5.16-5.22 (m, 1H, H-1B), 5.05-5.08 (m, 1H, H1s), 4.89-4.93 (m, 1H, H-1a), 4.13-4.19 (m, 1H , H-2c), 4.06-4.13 (m, 2H, H-2s, H-3s), 3.92-4.03 (m, 6H, H-2a, H-3A, H-3b , H-4s, H-4B, H-4a), 3.87-3.92 (m, 2H, H-5A, H-5C), 3.80-3.84 (m, 1H, H-5B ), 3.71-3.75 (m, 1H, -OCH2b), 3.71 (s, 3H), 3.56 (dt, J=9.9, 5.9 Hz, 1H, -O- CH2a), 2.42 (t, J=7.4 Hz, 2H, -CH2f), 1.60-1.68 (m, 4H, CH2e, -CH2c), 1.40 (dq, J=14, 8, 7.3 Hz, 2H, -CH2d), 1.20-1.30 (m, 9H, 3χΗ-6); ¹³ C NMR (176 MHz, D2O): δ 178.4, 168.6 (χ2), 165.7, 165.7 (χ2), 102.9, 102.8, 101.5, 99.1, 78 .5, 78.4, 78.2, 78.1, 78.0, 69.8, 69.1, 68.8, 68.7 (χ2), 68.6, 68.5 (χ2), 68 .3 (χ2), 67.9, 57.8, 52.9, 52.8, 52.7 (χ2), 52.5, 34.4 (χ2), 28.9, 25.7, 24, 8, 17.8 (χ2), 17.7 (χ2), 17.6, 17.5 (χ2) ppm. HRMS (ESI): m/z, calculated for C ₂₈ H4 ₇ N ₃ O ₁₅ Na [M+Na]+: 688.2899, detected: 688.2908.

(2'-Aminoethylamido)carbonylpentyl-4,6-dideoxy-4-formamido-α-D-mannopyranosyl (1^2) 4,6dideoxy-4-formamido-α-D-mannopyranosyl (1^2) 4,6- dideoxy-4-formamido-α-D-mannopyranoside (S24).

A solution of 4 (0.06 g, 0.09 mmol) in freshly distilled 1,2-diaminoethane (3.0 ml) was stirred at 65°C for 48 h. After that, the excess of the reagent was removed under vacuum, and the residue was co-evaporated with CH3OH (3x 10 ml) and dried. The residue was purified by reverse phase HPLC on a C18 column in a water-acetonitrile gradient and lyophilized to give the title compound S24 (0.052 g, 83.15%) as a white foam. Analytical data for S24: [a] _D ²¹ =+37.05 (c=1.14, H2O); 1H NMR (500 MHz, D2O): δ 8.24-8.33 (Z) and 8.05-8.12 (E) (m, 3H, NCHO), 5.23-5.26 (m, 1H , H-1c), 5.12 (s, 1H, H-1c), 4.934.97 (m, 1H, H-1A), 4.19-4.24 (m, 1H, H-2B), 4 .10-4.18 (m, 2H, H-2C, H-3C), 3.96-4.08 (m, 6H, H-2A, H-3A, H3B, H-4C, H-4B, H-4a), 3.91-3.96 (m, 2H, H-5A, H-5C), 3.84-3.89 (m, 1H, H-5B), 3.77 (dt, J =9.7, 6.8 Hz, 1H, -OCH2b), 3.57-3.63 (m, 1H, -O-CHa), 3.33 (t, J=6.2 Hz, 2H, - CH2g), 2.82 (t, J=6.2 Hz, 2H, -CH2h), 2.33 (t, J=7.4 Hz, 2H, -CH2f), 1.64-1.74 (m , 4H, -CH2e, -CH2c), 1.39-1.49 (m, 2H, -CH2d), 1.25-1.35 (m, 9H, 3χH-6); ¹³ C NMR (126 MHz, D2O): δ 178.3, 168.8 (χ2), 165.8 (χ2), 103.0, 102.9, 101.6, 99.3, 78.6, 78 .3, 78.2, 78.1, 69.9, 69.2, 69.0, 68.9, 68.8 (χ2), 68.6 (χ2), 68.5, 68.4, 57 .7, 53.0, 52.8 (χ2), 52.7, 42.1, 42.1, 40.7, 36.7, 29.1, 26.0, 25.9, 17.9 ( χ2), 17.8 (χ2), 17.7 (χ2), 17.6 ppm; HRMS (ESI): m/z, calculated for C29H5 ₁ N ₅ O ₁₄ Na [M+Na]+: 716.3325, detected: 716.333.

1-[(2'-Aminoethylamido)carbonylpentyl-4,6-dideoxy-4-formamido-a-O-mannopyranosyl (1^2) 4,6dideoxy-4-formamido-a-O-mannopyranosyl (1^2) 4 ,6-dideoxy-4-formamido-a-O-mannopyranoside]-2butoxycyclobutene-3,4-dione (S25).

To a stirred solution of S24 (0.015 g, 0.022 mmol) in water (0.5 ml) and EtOH (0.5 ml) was added a solution of 3,4-dibutoxy-3-cyclobutene-1,2-dione (20% in ethanol, 70 µl) and the pH was adjusted to 8 by careful addition of aqueous NaHCO3 (1%). After 1 hour, mass spectrometry showed that the reaction was complete; the reaction mixture was neutralized with CH3COOH (10%) and concentrated in vacuo. The residue was purified by reverse phase HPLC on a C18 column in a water-acetonitrile gradient and lyophilized to give the title compound S25 (0.0133 g, 73.2%) as a white foam. Analytical data for S25: 1H NMR (700 MHz, D ₂ O): δ 8.21-8.23 (Z) and 8.05 (E) (m, 3H, NCHO), 5.19 (s, 1H, H-1c), 5.07 (s, 1H, H-1c), 4.90-4.92 (m, 1H, H-1a), 4.68-4.75 (m, 2H, -CH ₂₁ ), 4.14-4.19 (m, 1H, H-2c), 4.07-4.13 (m, 2H, H-2s, H-3s), 3.92-4.02 (m, 6H, H-2a, H-3a, H-3b, H-4c, H-4c, H-4a), 3.89 (m, 2H, H- _5a , H- _5c ), 3.79- 3.85 (m, 1H, H-5 _V ), 3.73 (t, J=5.0 Hz, 1H, -CH _2g ), 3.65-3.71 (m, 1H, -O-CH _2b ), 3.62 (t, J=5.0 Hz, 1H, -CH _2g ), 3.51 (dd, J=9.6, 6.5 Hz, 1H, -O-CH _2a ), 3 .40-3.45 (m, 2H, -CH _2h ), 2.19-2.27 (m, 2H, CH _2f ), 1.77-1.84 (m, 2H, -CH _2j ), 1 .51-1.64 (m, 4H, -CH _2e , -CH _2c ), 1.46 (dt, J=15.5, 7.9 Hz, 2H, -CH _2k ), 1.30-1, 36 (m, 2H, -CH ₂ d), 1.20-1.30 (m, 9H, 3χH-6), 0.94-0.98 (m, 3H, -CH2l); ¹³ C NMR (176 MHz, D ₂ O): δ 189.7, 184.1, 178.4, 177.8, 174.5, 168.6, 165.7, 165.7, 102.8, 101 .5, 99.1, 98.9, 78.4, 78.1, 75.2, 75.1, 69.8, 69.1, 68.8, 68.7, 68.6, 68.4 , 68.3 (χ2), 57.8, 52.9, 52.7, 52.5, 45.0, 44.9, 40.2, 40.0, 36.6, 32.3, 29, 1, 26.0, 25.9, 25.8, 25.7, 19.0, 18.9, 17.8 (χ2), 17.7 (χ2), 17.6, 17.5, 13, 8ppm; HRMS (ESI): m/z, calculated for C ₃₇ H ₅₉ N ₅ O ₁₇ Na [M+Na]+: 868.3798, detected: 868.3808.

Preparation of BSA conjugate 5: BSA (15 mg) and trisaccharide squarate S25 (3.8 mg, 6.77 µmol) were dissolved in 0.1 M PBS buffer, pH 9 (600 µl) and stirred slowly at 21°C within 3 days. The reaction mixture was then diluted with Mili-Q water, filtered through an ultrafine filtration tube (10,000 MWCO, 4 x 10 mL), lyophilized to give BSA conjugate 5 as a white foam (17.6 mg). Mass spectrometric analysis of MALDI-TOF showed that conjugate 5 had an average of 16.2 disaccharides per BSA molecule.

Claims (6)

CLAIM 1. A molecule containing a chain of seven or more adjacent units of 4,6-dideoxy-4-acylamidoα-mannopyranose, and adjacent units are connected by a C1-C ₂ - or C1-C ₃ bond, the chain has a reducing end, while the pyranose ring in the chain link farthest from the reducing - 36 040853 end, is linked to a cap structure, where the cap structure is not 4,6-dideoxy-4-acylamido-a-Vpyranose linked to the rest of the molecule through its C1, and where the cap structure consists of the formula 4, where R4 is acylamido or its deacylated variant, OH, C1- _C5 alkoxy, C1-C5 alkyl or hydroxylated C1- _C5 alkyl; R5 is C1- _C5 alkyl or hydroxylated C1- _C5 alkyl; and R6 and R ₇ are independently selected from -H, -CH3 or -CHO. 2. Vaccine composition for use as a vaccine against infection of an animal with a Brucella organism, containing a molecule according to claim 1 and an object carrying the vaccine. 3. A method of vaccinating an animal against infection with a Brucella organism, comprising administering to the animal a protective amount of a molecule according to claim 1 or a vaccine composition according to claim 2. 4. A method for detecting an animal infected with Brucella from a population of animals known to include individuals that have been vaccinated by the method of claim 3, the method comprising contacting a biological sample obtained from the animal with an antigen to DIVA, wherein detection of binding of the antibody to the antigen for DIVA indicates that the sample was obtained from an animal infected with Brucella. 5. A method for detecting antibodies to Brucella in a sample, which includes bringing the sample into contact with a diagnostic conjugate containing a molecule according to claim 1. 6. The method according to claim 4, characterized in that the antigen for DIVA contains: structure XII and/or Structure XI and/or Structure II