Modified protein

Abstract

The present disclosure a cohort of sialic acid binding molecules which comprise one or more modified carbohydrate binding modules (CBMs). The modified CBMs reduce the risk of adverse events related to the host immune response and/or the production of anti-drug antibodies (ADAs). The modified CBMs can be used in therapy or as medicaments and find specific application as molecules for the modulation of an immune response and/or cell growth. The modified CBMs may also be used as adjuvants, for example mucosal adjuvants and in the treatment and/or prevention of cancer, sepsis and/or diseases caused or contributed to by a pathogen that binds cell surface sialic acid-containing receptors.

Claims

1. A sialic acid binding molecule comprising the following sequence: TABLE-US-00023 (SEQ ID NO: 28) GAMVIEKEDVETNASNGQRVDLSSELDKLKKLENATVHMEFKPDPKAPA FYNLFSVSSATKKDEYFTMAVYNNTATLEGRGSDGKQFYNNYNDAPLK VKPGQWNSVTFTVEKPTAELPKGRARLYVNGGLSRTSLRSGNFIKDMPD VTHVQIGATKRANNTVWGSNLQIRNLTVYNRALTPEEVQKRSGGGSGVI EKEDVETNASNGQRVDLSSELDKLKKLENATVHMEFKPDPKAPAFYNLF SVSSATKKDEYFTMAVYNNTATLEGRGSDGKQFYNNYNDAPLKVKPGQ WNSVTFTVEKPTAELPKGRARLYVNGGLSRTSLRSGNFIKDMPDVTHVQI GATKRANNTVWGSNLQIRNLTVYNRALTPEEVQKRSGGSLGVPDFESDW FDVSSNSLYTLSHGLQRSPRRVVVEFARSSSPSTWNIVMPSYFNDGGHKG SGAQVEVGSLNIKLGTGAAVWGTGYFGGIDNSATTRFATGYYRVRAWI.

2. A pharmaceutical composition comprising the sialic acid binding molecule according to claim 1, formulated together with one or more excipients, carriers, adjuvants and/or buffers.

3. The pharmaceutical composition of claim 2, wherein the composition is formulated for oral, mucosal or parenteral administration.

4. The pharmaceutical composition of claim 3, wherein the composition is formulated for intranasal administration or inhalation.

5. A method of preparing the pharmaceutical composition of claim 2, comprising formulating the sialic acid binding molecule together with the one or more excipients, carriers, adjuvants and/or buffers.

Description

DETAILED DESCRIPTION

(1) The present invention will now be described in detail with reference to the following figures which show:

(2) FIG. 1: ProPred predictions of antigenic peptides. A. SpCBM sequence (SEQ ID NO. 2). B. PaTD sequence (SEQ ID NO. 14). Predicted binders are coloured blue, with the first residue of each binding region shown in red. Antigenic peptides predicted by Nordic Biopharma (green bars) and ProImmune (purple bars) are shown under the sequences.

(3) FIG. 2: Expression test of wild type and mutated domains. Lane 1, M12 standard; Lane 2, WT SpCBM; Lanes 3-11, Im15-Im23; Lanes 12-15, Im24-Im27; Lane 16, WT PaTD. A) Whole cell extracts, B) Soluble extracts.

(4) FIG. 3: Position of peptide 167-181 in the SpCBM structure

(5) FIG. 4: Expression and Ni-NTA pull-down of variants Im28 to Im34

(6) FIG. 5: Sites of the HEX17 mutations on the hexamer structure. The quarternary structure of HEX17 was modelled by assembling the crystal structures of the individual SpCBM (pdb code 4c1x) and PaTD (pdb code 2w38) into the hexamer (i.e. 6 copies of SpCBM and 3 copies of PaTD per molecule). The positions of the bound ligand (α2,3-sialyllactose) are shown in stick form (orange). The positions of the mutations are also shown: Blue, the sites of the A162P mutation; Cyan, the sites of the other two CBM mutations; Magenta, the sites of the TD mutations.

(7) FIG. 6: IL-8 stimulation. A549 cells were stimulated by the addition of 10 μg of biologic (Sp2CBMTD (aka SpOrig), HEX6 or HEX17). Cell supernatant was harvested at 24 h or 48 h timepoints and the IL-8 content was determined by ELISA. Statistical significance between control and treated cells was determined with one-way ANOVA using Tukey's multiple comparison test.

(8) FIG. 7: Multiplex analysis of inflammatory mediators. A549 cells were stimulated by the addition of 10 μg of biologic (Sp2CBMTD (aka SpOrig), HEX6 or HEX17). Cell supernatant was harvested at 6 h, 24 h or 48 h time-points and inflammatory mediators analysed using a Human Cytokine 12-plex Assay. Statistical significance between control and/or WT hexamer and hexamer variants was determined using a one-way ANOVA (Tukey's multiple comparison test).

(9) FIG. 8: Percentage survival of CBM-treated and untreated mice when lethally challenged with influenza strain PR8. CBM2, CBM3 and CBM4 represent HEX17, HEX6 and WT (SpOrig) respectively. Single CBM dosed animals were given 100 μg of CBM one day prior to lethal challenge with PR8; repeat dosed animal were given 2×0.1 μg of CBM at day-3 and day-1 prior to PR8 challenge.

(10) FIG. 9: Clinical scores of CBM-treated and untreated mice during PR8 infection. CBM2, CBM3 and CBM4 represent HEX17, HEX6 and WT (SpOrig) respectively. An ascending clinical score of 1 to 5 indicates no symptoms (1) to lethargy and death (5), respectively.

(11) FIG. 10: Percentage weight loss of CBM-treated and untreated mice during PR8 infection. CBM2, CBM3 and CBM4 represent HEX17, HEX6 and WT (SpOrig) respectively.

(12) FIG. 11: Anti-mCBM antibody analysis of Day 21 lung homogenates and sera tissue from a PR8-challenged mouse study. Statistical significance between WT hexamer and hexamer variants was determined using a one-way ANOVA (Tukey's multiple comparison test).

(13) FIG. 12a: Anti-CBM antibody levels from Day 35 BAL mouse samples.

(14) FIG. 12b: Anti-CBM antibody levels from Day 35 serum mouse samples

METHODS AND RESULTS

(15) Sp2CBMTD: Prediction of Immunogenic Regions

(16) Nordic Biopharma in Silico Screen

(17) The in silico T-cell epitope screening identified four significant and two borderline immunogenic clusters:

(18) Significant:

(19) TABLE-US-00013 Domain Residue range Sequence SpCBM 245 to 254 GVLSRTSLRS PaTD 340 to 349 WFSVSSNSLY PaTD 351 to 359 LSHGLQRSP PaTD 398 to 406 GSLNIRLGT

(20) Borderline:

(21) TABLE-US-00014 Domain Residue range Sequence SpCBM 167 to 178 FYNLFSVSSATK SpCBM 239 to 251 VRLYVNGVLSRTS

(22) ProImmune Human Donor T-Cell Proliferation Assay

(23) The ProImmune study highlighted two regions of high antigenicity and two regions of moderate antigenicity:

(24) High Antigenicity:

(25) TABLE-US-00015 Domain Residue range Sequence SpCBM 236 to 250 KGRVRLYVNGVLSRT PaTD 392 to 406 GAQVEVGSLNIRLGT

(26) Moderate Antigenicity:

(27) TABLE-US-00016 Domain Residue range Sequence SpCBM 167 to 181 FYNLFSVSSATKKDE PaTD 338 to 352 SDWFSVSSNSLYTLS

(28) ProPred in Silico Analysis

(29) A further in silico tool, the online ProPred server.sup.4, was also used. The output of the ProPred server is shown in FIG. 1. The relative positions of the Nordic Biopharma/ProImmune epitopes are also highlighted and indicate reasonable agreement between the three methods. In addition to the epitopes listed above, ProPred strongly predicted another immunogenic epitope in the SpCBM domain:

(30) TABLE-US-00017 Domain Residue range Sequence SpCBM 286 to 294 IRNLITYNR

(31) Mutations in the Individual CBM and TD Domains

(32) To guide the design of mutations that might reduce immunogenicity, ProPred was used to test the effect of changing each residue in these peptides to every alternative residue. Those that gave the greatest reduction in predicted number of allele binders were noted. As the crystal structure of both the SpCBM and TD domains are known, these mutations were also modelled to reduce the likelihood of introducing mutations that would obviously disrupt the protein structure.

(33) Initially, nine single mutations in SpCBM and four single mutations in PaTD were introduced and are listed below (‘Im’ is short for immunogenicity mutant):

(34) TABLE-US-00018 (SpCBM) variants Mutation (PaTD) variants Mutation WT Sp — WTTD — lm15 Y168W lm24 S342D lm16 L170A lm25 S345D lm17 L170T lm26 L348D lm18 V173G lm27 R403K lm19 V239A lm20 V239T lm21 V246G lm22 I286A lm23 Y292E

(35) Note: Im1 to Im14 (not shown) were introduced by mutagenesis into a non-codon optimized background, before the ProImmune data were available.

(36) Synthesis of WT and Mutated Constructs

(37) The genes encoding WT SpCBM, WT PaTD and the variants Im15 to Im27 were codon optimized for E. coli expression and synthesized by GeneArt. The genes were then cloned in-house into the pHISTEV vector for expression as 6His-tagged proteins.

(38) Expression and Biophysical Characterization

(39) An initial expression test was performed to assess solubility. The results show that all were expressed, but not all were soluble (FIG. 2). Note: solubility (or a lack thereof) is not necessarily a predictor of utility. One of skill will appreciate that when manufacturing or producing proteins, certain processes require the use of insoluble material as this is readily purified (from inclusion bodies and the like). Downstream protocols may then re-engineer proteins to modulate features such as solubility.

(40) Results of the expression test show that: Im16 (L170A) is insoluble or very poorly soluble Im25 (TD, S345D) is insoluble Im15 (Y168W) and Im17 (L170T) have reduced solubility Im18 (V173G) and Im22 (I286A) are slightly reduced. The remainder show soluble expression.

(41) The 13 soluble proteins were expressed in E. coli and purified by immobilized metal affinity chromatography (IMAC), followed by TEV digestion to remove the 6His-tag, then reverse IMAC and size exclusion chromatography (SEC).

(42) Ten purified domains (WT Sp, Im19, Im20, Im21, Im22, Im23, WTTD, Im24, Im26 and Im27) were further characterized by:

(43) (i) Thermofluor to measure melting temperature (Tm)

(44) (ii) Near UV circular dichroism (CD) to compare tertiary structures to WT

(45) (iii) Dynamic light scattering (DLS) to check oligomeric state in solution

(46) (iv) Surface plasmon resonance (SPR) to measure binding affinity to sialyllactose

(47) (v) Measurement of IL-8 cytokine stimulation

(48) The results are summarized in Table 1.

(49) TABLE-US-00019 TABLE 1 Qualitative summary of the biophysical characterizations of the WT domains and their variants. Colour coding is from green to red (including green′, orange and yellow), where green indicates that the variant closely resembles its WT counterpart for that particular characteristic and pale green (green′) or yellow indicate increasing degrees of differences. Red or orange indicate significant differences. Tm +/− NearUV Cytokine Name Mutation Solubility Purification 6SL CD DLS Biacore stimulation WTSp — Green Green Green Green Green Green Green lm15 Y168W Yellow Orange Red N/A N/A N/A N/A lm16 L170A Red N/A N/A N/A N/A N/A N/A lm17 L170T Yellow Orange N/A N/A N/A N/A N/A lm18 V173G Green′ Orange N/A N/A N/A N/A N/A lm19 V239A Green Green Green Green Green Green N/D lm20 V239T Green Green Green′ Green′ Green Green N/D lm21 V246G Green Green Green′ Green Green Green Green lm22 I286A Green′ Green Green′ Green Green Green Green lm23 Y292E Green Green Green′ Green′ Green Yellow Yellow lm24 S342D Green Green Green Green Green lm25 S345D Red N/A N/A N/A N/A lm26 L348D Green Green Green′ Green Green lm27 R403K Green Green Green Green Green WTTD — Green Green Green Green Green N/A: these characterizations were not performed due to poor solubility/purity of the protein. N/D: not determined.

(50) Sp Peptide 167-181:

(51) Im15, Im16, Im17, Im18 are all insoluble or poorly soluble (as stated, this does not necessarily impact on protein utility). These are in the ‘moderately’ antigenic region 167-181 (FYNLFSVSSATKKDE). This region is clearly very sensitive to change.

(52) Earlier results show that M156F, which sits adjacent to L170 (and I286), increases Tm by ˜4° C.

(53) This could therefore be combined with L170T. M156F does not increase predicted immunogenicity.

(54) M185I increases Tm by 5° C., and lies parallel to L170 (FIG. 3). This mutation could also be included. Note that, like M156F, M185I does not increase predicted immunogenicity but slightly reduces the number of predicted allele binders.

(55) Sp Peptide 236-250:

(56) Im19, Im20, Im21 all behave similarly to WT. These are in the ‘highly’ antigenic region 236-250 (KGRVRLYVNGVLSRT).

(57) Im19 (V239A) was chosen over the threonine mutation (Im20, V239T). There is no difference in predicted immunogenicity but Im19 is a closer match to WT Thermofluor Tm and Near UV spectrum. This would be combined with Im21 (V246G).

(58) Sp Peptide 286-294:

(59) Im22 (I286A) is broadly similar to WT while Im23 (Y292E) appears to exhibit reduced ligand affinity. This region, 286-294 IRNLTVYNR, was not flagged up by ProImmune but is strongly predicted by ProPred to be immunogenic.

(60) There is some indication that Im22 has lower Tm than WT. This residue is adjacent to M156 so may behave differently if M156F was included.

(61) TD peptide 338-352:

(62) Im24 (S342D) and Im26 (L348D) show similar characteristics to the WT trimerization domain, but with some suggestion of reduced Tm in Im26. These are in the ‘moderately’ antigenic region 338-352 SDWFSVSSNSLYTLS. The WT sequence was predicted to bind 9 alleles, while Im24 predicts 2 alleles and a Im24/Im26 double mutant predicts 1 allele.

(63) TD Peptide 392-406:

(64) Im27 (R403K) is similar to WT. It is part of the ‘highly’ antigenic region 392-406 GAQVEVGSLNIRLGT. Predicted alleles are reduced from 21 to 3 when this mutation is introduced.

(65) Synthesis of Multiple Mutation Combinations Im28-34

(66) The following mutations were introduced:

(67) i) M156F/L170T

(68) ii) M156F/L170T/M185I: In ProPred, alleles predicted for this region are reduced from 31 in the WT to 19 for this combination.

(69) iii) V239A/V246G: In ProPred, alleles for this region are reduced from 44 to 3.

(70) iv) I286A/Y292E: In ProPred, alleles are reduced from 41 to 1.

(71) v) V239A/V246G/I286A/Y292E combines the previous two doubles.

(72) vi) M156F/L170T/M185I/V239A/V246G/I286A/Y292E combines all the Sp mutations

(73) vii) TD: S342D/L348D/R403K: Predicted alleles are reduced from 9 to 1 for TD peptide 338-352 and alleles for peptide TD peptide 392-406 are reduced from 21 to 3. This triple mutant combines all the TD mutants. They are all surface exposed and distal to the N-terminal end of TD, so would not be expected to interfere with SpCBM in the hexamer form.

(74) The constructs are named Im28 to Im34:

(75) TABLE-US-00020 (SpCBM) variant Mutations lm28 M156F/L170T lm29 M156F/L170T/M185I lm30 V239A/V246G lm31 I286A/Y292E lm32 V239A/V246G/I286A/Y292E lm33 M156F/L170T/M185I/V239A/V246G/I286A/Y292E (PaTD) variant Mutation lm34 S342D/L348D/R403K

(76) 2.5 Expression and Biophysical Characterization of Im28-Im34

(77) As with the single mutations, the combinations Im28 to Im34 were synthesized by GeneArt and subcloned into pHISTEV for expression analysis. A nickel bead pull-down on the His-tagged soluble extract was also performed (FIG. 4).

(78) Hexameric Forms

(79) Design of Hexameric Constructs HEX1 to HEX17

(80) Genes encoding the hexameric forms (called Hex1 to Hex17) were synthesized by GeneArt:

(81) Sp2CBMTD

(82) TABLE-US-00021 variant Mutations HEX1 CBM1(L170T V239A V246G I286A Y292E)-CBM2(L170T V239A V246G I286A Y292E)-TD (S342D L348D R403K) HEX2 CBM1(V239A V246G I286A Y292E)-CBM2(V239A V246G I286A Y292E)- TD (S342D R403K) HEX3 CBM1(V239A V246G I286A)-CBM2(V239A V246G I286A)-TD (S342D R403K) HEX4 CBM1(V239A V246G)-CBM2(V239A V246G)-TD (S342D) HEX5 CBM1(V239A V246G)-CBM2(V239A V246G)-TD(R403K) HEX6 CBM1(V239A V246G)- CBM2(V239A V246G)-TD (S342D R403K) HEX17 CBM1(V239A V246G A162P)- CBM2(V239A V246G A162P)-TD (S342D R403K)

(83) The hexameric forms were synthesized in two parts to avoid problems associated with synthesising repeat sequences in the tandem CBM copies. The first gene covered the first CBM and the second part encompassed the second CBM plus the TD. These could then be simultaneously cloned into pHISTEV to create the Sp2CBMTD construct that trimerizes upon expression.

(84) The first hexamer, HEX1, contained the mutations L170T/V239A/V246G/I286A/Y292E in the CBMs and S342D/L348D/R403K in the TD.

(85) The solubility data of the individual domains indicated that HEX1 was unlikely to be soluble (again, not necessarily a reflection on the utility of the molecule); a further construct, HEX3, was synthesized. Note that HEX2 contained the same mutations as Hex3, but with the addition of Y292E.

(86) HEX3 was synthesized and subcloned into the pHISTEV vector. Expression was insoluble under all conditions tested (varying temperature, IPTG concentration, cell density at induction, with or without heat shock). The CBM-only domain containing the same three mutations (V239A V246G I286A) is soluble. A double mutant (V239A V246G) behaves very similarly to WT. Therefore, further variants (HEX4, HEX5 and HEX6) were designed and constructed by PCR/ligations, which exclude I286A and contain either one or both of the TD mutations.

(87) During the work on HEX6 a number of other versions were designed containing different combinations of the HEX6 mutations (numbered HEX7 to HEX16; not characterised).

(88) HEX17 contains the HEX6 mutations with an additional A162P mutation. This proline mutation has been shown to increase the single CBM Tm by 3-4° C. The proline mutation is not near the other mutations, the N- or C-termini or the ligand binding site.

(89) Characterization of the Hexameric Variants

(90) The expression, purification and characterization results are shown in Table 2. Based on these results, HEX6 and HEX17 were taken forward. The positions of the HEX17 mutations on the hexamer are shown in FIG. 5.

(91) TABLE-US-00022 TABLE 2 NearUV IL-8 Name Mutations Solubility Purification Thermostability CD Biacore assay Hex1 L170T/V239A/V246G/I286A/Y292E/ Red N/A N/A N/A N/A N/A S342D/L348D/R403K Hex2 V239A/V246G/I286A/Y292E/S342D/R403K (designed but not made) Hex3 V239A/V246G/I286A/S342D/R403K Red N/A N/A N/A N/A N/A Hex4 V239A/V246G/S342D Yellow Red N/A N/A N/A N/A Hex5 V239A/V246G/R403K Green Yellow Yellow N/A N/A N/A Hex6 V239A/V246G/S342D/R403K Green Green Yellow Green Green Yellow Hex7 Note: These constructs are different combinations of the Hex6 to 16 mutations and were designed as a back-up in case Hex6 failed Hex17 A162P/V239A/V246G/S342D/R403K Green Green Green′ Green Green reduced IL-8

(92) Table 2. Qualitative summary of the biophysical characterizations of the hexameric Sp2CBMTD variants. Colour coding is from green to red, where green indicates that the variant closely resembles its WT counterpart for that particular characteristic and pale green (green′) or yellow indicate increasing degrees of differences. Red or orange indicate significant differences. N/A: these characterizations were not performed due to poor solubility/purity of the protein. N/D: not determined.

Example 1: Inflammatory Mediators

(93) Aim: To measure the innate immune response of mCBM-treated human lung epithelial cells (A549) by analysing levels of inflammatory mediators over time.

(94) Administration of Sp2CBMTD to mammalian cells stimulated a pro-inflammatory response both in vitro and in vivo.sup.1,2. To determine whether this was still observed with modified hexameric sialic acid binding molecules, mammalian A549 cells were stimulated by the addition of 10 μg of biologic (Sp2CBMTD (aka SpOrig), HEX6 or HEX17 and cell culture medium was harvested at specific time-points post administration. The concentrations of inflammatory mediators were measured both by ELISA and a multiplex assay.

(95) Human IL-8 (benchmark cytokine for the study) response using a human 1× Mouse CXCL1/KC Quantikine ELISA Kit (R&D BioSystems). The concentration levels of IL-8 from stimulated A549 cells are shown in FIG. 6. It is evident that when A549 cells are stimulated with the modified hexamer HEX17, IL-8 levels are significantly lower than when compared to Sp2CBMTD (aka SpOrig), or Hex6-stimulated cells.

(96) Inflammatory mediator response using a Human Cytokine 12-plex Assay (Bio-Plex Pro™ Bio-Rad). FIG. 7 demonstrates the analysis of 12 inflammatory mediators from culture medium after A549 cell stimulation by Sp2CBMTD (WT, aka SpOrig), HEX6 and HEX17 (variants) at specific time points (6 h, 24 h, 48 h). Prior to analysis, samples were thawed and diluted 1:4 in PBS before using a human HS Cytokine-12 plex assay (R&D Systems). The data indicates that: HEX17 affects the levels of almost all the cytokines tested compared to SpOrig and HEX6. There is a significant reduction in observed concentration (pg/ml) with analytes IL-6, IL-8, GM-CSF and IFN-gamma at 48 h when compared to SpOrig and HEX6. When compared to control at 48 h, HEX17 appears to cause an increase in the level of all cytokines tested with the exception of IL-5, and VEGF (yet to be confirmed). HEX6 only showed reduced IL-6 stimulation compared to SpOrig at 48 h.

Example 2: In Vivo PR8 Mouse Data

(97) The objective of the study was to assess the efficacy of Sp2CBMTD (SpOrig) and its variants, in a mouse model of lethal influenza infection. Each of the candidate proteins were also administered in the absence of an influenza infection to assess whether they alone, caused any morbidity or mortality.

(98) Survival, Clinical Scores and Weight Loss.

(99) The results show that none of CBM2 (HEX17), CBM3 (HEX6) or CBM4 (WT, SpOrig) caused any overt morbidity or mortality alone. Administration of a single 100 μg dose of either CBM2 (HEX17), CBM3 (HEX6) and CBM4 (WT, SpOrig) one day prior to a lethal challenge with PR8 influenza virus elicited protection against PR8 infection, with greatest efficacy seen with HEX17 (100% survival), followed by SpOrig and then HEX6 (FIG. 8). Clinical scores were also lower with HEX17 compared to SpOrig and HEX6 (FIG. 9). Mice from single high dose treated groups that survived all lost weight at peak infection but soon recovered, in contrast to untreated, infected mice (FIG. 10).

(100) Anti-mCBM Antibody Analysis of Lung Homogenates and Sera Tissue from a PR8-Challenged Mouse Study.

(101) The objective of this study was to determine whether modified immunogenic epitopes of the modified variants of Sp2CBMTD (SpOrig) demonstrated reduced antibody levels in mice in a PR8-challenged study (it should be noted that epitopes were modified based on human MHC-class II binding information). For this, survived mice from PR8-challenged study were culled at Day 21 with lung and sera harvested and tested for anti-mCBM antibodies—IgG, IgA, IgE and IgM against coated antigen SpOrig (1 μg/well) in an ELISA format. The data shown in FIG. 11 indicated that: The modified protein HEX17 did show a significant (p<0.05) reduction in mouse lung IgM levels compared to SPORIG. Due to only one surviving mouse for HEX6 treatment, only HEX17 and SPORIG data was statistically analysed. There is some indication of a slight downward trend of antibody levels in mice (lung IgA and IgM) that were treated with the modified CBMs compared to SpOrig.

Example 3: The Effect of Repeat Intranasal Dosing of mCBMs, Sp2CBMTD and HEX17, in the Mouse

(102) The objective of this study was to assess the clinical effect and immune response of repeat dosing of both Sp2CBMTD (SpOrig) and HEX17 in mice over time.

(103) Experimental Procedures

(104) Intranasal Dosing.

(105) On Days 1, 15 and 29, cohorts of mice (10F, 10M BALB/c mice per agent) were dosed with 20 μg of either sterile PBS, Sp2CBMTD or HEX17 via the intranasal route, under recoverable gaseous anaesthesia (isoflurane/oxygen mix), at a fixed volume of 40 μL.

(106) Body Weights and Post-Dose Observations.

(107) All mice were weighed twice weekly from Day −1 until the end of the study (Day 35). Post-dose observations were recorded every 15 min for the first 2 h after dose administration and then every 30 min for the next 6 h.

(108) Analysis of Tissue Samples.

(109) Mouse tissue (serum and bronchoalveolar lavage (BAL)) were tested for anti-CBM antibody responses (IgG, IgA and IgM) against either Sp2CBMTD or HEX17 as coating antigens (1 μg/well) in an ELISA format. Absorbance readings at OD.sub.450nm (reference background OD.sub.620nm) were measured for each sample after reaction of an HRP-conjugated detection antibody with its chromogenic substrate 3,3′,5,5′-tetramethylbenzidine (TMB), to the different antibody types.

(110) Results:

(111) Clinical Scores.

(112) Further studies into the effect of repeat intranasal dosing of Sp2CBMTD and Hex17 in mice reveals that neither molecule had any significant effect on bodyweight or food consumption. Further, at later doses, HEX17 appears to be better tolerated, with some (resolving) clinical signs (including piloerection, hunched posture, underactivity, partially closed eyes and irregular breathing) being noted for a limited period after Sp2CBMTD administration.

(113) Anti-mCBM Antibody Analysis of Serum and BAL Tissue from Mice.

(114) The data shown in FIG. 12 indicated that after 35 days, where mice were given 3 doses of 20 μg CBM every two weeks between Days 1 and 29, an adaptive immune response was induced to both CBMs. HEX17 (an example of a modified or “de-immunized” CBM where epitopes are modified based on human MHC-class II binding information as described previously) demonstrated a significant (p<0.05) reduction of IgA in both BAL and serum tissues compared to Sp2CBMTD-treated mice. This was observed against both coated antigens. The difference in IgA response between the two candidates was also significant between male and female mice. The difference in IgG response was more evident in BAL samples than in sera. There was also a significant reduction of IgM levels from both BAL and serum samples when tested against Sp2CBMTD, but this was not significant in HEX17-coated plates.