Immunology and Cell Biology (2002) 80, 178–187

Research Article

Enhancing the immunogenicity and modulating the fine epitope recognition of antisera to a helical group A streptococcal peptide vaccine candidate from the M protein using lipid-core peptide technology

W E N D Y A H A Y M A N , 1 I S T V A N T O T H , 2 N I C H O L A S F L I N N , 3 M A R T I N S C A N L O N 2 a n d M I C H A E L F G O O D 1

1The Cooperative Research Centre for Vaccine Technology and The Australian Centre for International and Tropical Health and Nutrition, The Queensland Institute of Medical Research, Royal Brisbane Hospital, Brisbane, 2The School of Pharmacy, University of Queensland, St Lucia, Queensland, Australia and 3The School of Pharmacy, University of London, London, United Kingdom

Summary A conserved helical peptide vaccine candidate from the M protein of group A streptococci, p145, hasbeen described. Minimal epitopes within p145 have been defined and an epitope recognized by protectiveantibodies, but not by autoreactive T cells, has been identified. When administered to mice, p145 has lowimmunogenicity. Many boosts of peptide are required to achieve a high antibody titre (> 12 800). To attempt toovercome this low immunogenicity, lipid-core peptide technology was employed. Lipid-core peptides (LCP) consistof an oligomeric polylysine core, with multiple copies of the peptide of choice, conjugated to a series of lipoaminoacids, which acts as an anchor for the antigen. Seven different LCP constructs based on the p145 peptide sequencewere synthesized (LCP1→LCP7) and the immunogenicity of the compounds examined. The most immunogenicconstructs contained the longest alkyl side-chains. The number of lipoamino acids in the constructs affected theimmunogenicity and spacing between the alkyl side-chains increased immunogenicity. An increase in immuno-genicity (enzyme-linked immunosorbent assay (ELISA) titres) of up to 100-fold was demonstrated using thistechnology and some constructs without adjuvant were more immunogenic than p145 administered with completeFreund’s adjuvant (CFA). The fine specificity of the induced antibody response differed for the different constructsbut one construct, LCP4, induced antibodies of identical fine specificity to those found in endemic human serum.Opsonic activity of LCP4 antisera was more than double that of p145 antisera. These data show the potential forLCP technology to both enhance immunogenicity of complex peptides and to focus the immune response towardsor away from critical epitopes.

Key words: M protein, rheumatic fever, Streptococcus pyogenes.

Introduction

Group A streptococci (GAS) are among the most commonand widespread human pathogens. They cause a wide rangeof infections, of which the most prevalent are acute pharyn-gitis and impetigo. The organism has, more significantly, longbeen the focus of intense clinical and investigative interestbecause of its association with the non-suppurative sequelaeof acute rheumatic fever (ARF), rheumatic heart disease (RHD)and acute glomerulonephritis (AGN). Protection against GASinfection is believed to be mediated predominantly by opsonicantibodies directed against the surface M protein, the majorvirulence factor of GAS.1 An epitope has been identified fromthe conserved carboxyl terminus of the M protein (p145: a

20-mer with the sequence LRRDLDASREAKKQVEKALE),which is recognized by antibodies in the sera of most adultsliving in areas of high streptococcal exposure.2–4 Both affinity-purified p145-specific human antibodies as well as antibodiesinduced in mice following immunization with p145 arecapable of opsonizing multiple serotypes of GAS, suggestingthat this peptide might be an ideal vaccine candidate.2–6

The elucidation of protective epitopes within the Mprotein of GAS has proceeded rapidly, thus permitting thedevelopment of peptide vaccines that could elicit protectionwhilst avoiding autoimmunity. Unfortunately, the prospectsof a useful vaccine have lagged, in part due to the lack of asuitable means of inducing high titre anti-M protein responseswhen immunizing with synthetic peptides. In the case ofpeptide 145, only modest anti-p145 murine titres have beenattainable, often after long and involved immunization pro-tocols employing CFA. With the addition of an effectiveadjuvanting system, it may be possible to elicit an earlier,more potent and more prolonged response.

Correspondence: Prof. M Good, The Queensland Institute ofMedical Research, 300 Herston Road, PO Royal Brisbane Hospital,Qld 4029, Australia. Email: michaelG@qimr.edu.au

Received 11 October 2001; accepted 29 November 2001.

Fine specificity of streptococcal vaccine 179

Several experimental adjuvants have been developed, butthe only adjuvants licensed for human use are aluminium salts(alum). However, despite the wide use and acceptance ofalum, there are problems associated with its use. First, not allproteins and peptides are equally adsorbed.7 Second, there arestrict conditions for the storage of vaccines absorbed ontoalum and, third, aluminium-associated pathology andvaccine-specific IgE production have been reported.8

To induce a high antibody response to a synthetic peptide,one successful approach has been to polymerize peptides byincorporation into a polylysine core, to form a multipleantigenic peptide (MAP).9 The synthesis of immunogens inthis manner allowed the conformation and quantity of peptideto be accurately determined and a higher degree of immuno-genicity was generated compared to that of the monomericpeptide.10–12

To attain an even higher antibody response, the antigenscan be synthesized with lipoamino acids, which facilitateanchoring to the cell membrane.13,14 Lipoamino acids are α-amino acids with long hydrocarbon side-chains. Lipoaminoacids and their oligomers, lipopeptides, represent a class ofcompounds that combine the structural features of lipids withthose of amino acids.13

A subsequent improvement in this technology has been thedevelopment of the lipid-core peptide (LCP) system, whichincorporated the lipoamino acid-based methodology into thepolylysine system to enhance the membrane-binding abilityof complex synthetic peptides.15,16 This system has the poten-tial not only to be used as a combined adjuvant carrier–vaccine, but also as a drug delivery system.17–19

A study by Zhong et al. has shown that the incorporationof a monomeric peptide into a single LCP structure enhancedimmunogenicity significantly (up to 3200-fold).20 In thepresent study, we investigated the effect of the LCP system onthe adjuvanticity of p145 peptide. We also examined howimmunogenicity is affected by changing various characteris-tics of the LCP constructs. A total of seven LCP constructswere synthesized, which differed in the number of p145epitopes attached to the polylysine core, the length of thehydrocarbon side-chains, the number of α-amino acid units inthe lipidic backbone and their spacing.

Materials and Methods

Mice

The following strains of H-2 congenic mice were used: B10.BR(H-2k), B10.D2 (H-2d), B10 (H-2b). All mice were obtained from theAnimal Resources Centre (Willetton, WA, Australia).

Peptides

The sequences of p145, the chimeric peptides (J1–J9 and J14), theshorter J ‘inserts’ (JI1–JI9) and Jcon, which have been describedelsewhere,5–21 are given in Table 1. The peptides were synthesizedusing the tea-bag method20 and assessed for purity by HPLC. TheN-terminal peptide of the M protein of M6 GAS, peptide 161, wasalso included in the study (Table 1).

The synthesis of the LCPs was accomplished automatically by astepwise solid phase procedure on a MBHA Novabiochem resin(substitution 0.48 mmol/g resin) as described elsewhere.15,16,20

Briefly, the lipoamino acids were coupled to the resin peptides

manually.13 Synthesis of the LCP construction was achieved usinga fourfold excess of preformed symmetrical anhydride of N-Bocamino acids in dichloromethane/N-methylpyrrolidone. The protectinggroups were Boc groups for the α-amino termini. For all residues,the first coupling was done with the preformed symmetrical anhyd-ride in CH2Cl2, and a second coupling done in dichloromethane/N-methylpyrrolidone. When necessary, further couplings were per-formed until the coupling yield reached a minimum of 99.8%.Deprotection of the N-termini was performed in 65% trifluoroaceticacid (TFA) in dichloromethane (20 mL for 1 min, then another 20 mLfor 10 min). The deprotected resin LCP was neutralized with 10%di-isopropylethylamine in dichloromethane. The LCP was removedfrom the resin support using the high HF method (1.5 mL cresol,1.5 mL thiocresol, 20 mL HF) to yield the crude LCP, which wasprecipitated with ethyl acetate and redissolved in 6 mol/L guanidineHCl–0.1 mol/L Tris solution (20 mL).

Lipid-core peptide purification was done on a Vydac C18 5 RACcolumn (Vydac, Hesperia, CA, USA). Acetonitrile of HPLC grade(Aldrich, Milwauke, WI, USA) and water were filtered through amembrane filter and degassed with helium prior to use. All HPLCsolvents contained 0.1% trifluoroacetic acid. Analytical separationwas achieved with a solvent gradient beginning with 0% acetonitrile,increasing constantly to 60% acetonitrile at 30 min, staying at thisconcentration for 20 min and decreasing steadily to 0% acetonitrilefor 10 min at a constant flow rate of 0.7 mL/min. For preparativeseparation, a preparative C18 column was used. Separation wasachieved with a solvent gradient beginning with 0% acetonitrile,increasing constantly to 60% acetonitrile at 180 min, staying at thisconcentration for 60 min and decreasing steadily to 0% acetonitrilefor 30 min at a constant flow rate of 7 mL/min. The gradient wascontrolled by two microprocessor-controlled Gilson 302 single-pistonpumps (Gilson, Middleton, WI, USA). Compounds were detectedusing a Holochrome UV-VIS detector (Gilson) at 214 nm (analytical)and 230 nm (preparative). Chromatographs were recorded by an LKB2210 single-channel chart recorder. The lipoamino acids wereracemic, hence the LCP constructs were diastereomeric mixtures. TheHPLC fractions of the major product, whose intensities were greater

Table 1 List of the synthetic chimeric J peptides and the shorterinsert peptides of p145

Peptide name Peptide amino acid sequence

p145 LRRDLDASREAKKQVEKALEJcon DKVKQAEDKVKQLEDKVEELQDKVKQLEJ1 QLEDKVKQLRRDLDASREAKEELQDKVKJ2 LEDKVKQARRDLDASREAKKELQDKVKQJ3 EDKVKQAERDLDASREAKKQLQDKVKQLJ4 DKVKQAEDDLDASREAKKQVQDKVKQLEJ5 KVKQAEDKLDASREAKKQVEDKVKQLEDJ6 VKQAEDKVDASREAKKQVEKKVKQLEDKJ7 KQAEDKVKASREAKKQVEKAVKQLEDKVJ8 QAEDKVKQSREAKKQVEKALKQLEDKVQJ9 AEDKVKQLREAKKQVEKALEQLEDKVQLJI1 LRRDLDASREAKJI2 RRDLDASREAKKJI3 RDLDASREAKKQJI4 DLDASREAKKQVJI5 LDASREAKKQVEJI6 DASREAKKQVEKJI7 ASREAKKQVEKAJI8 SREAKKQVEKALJI9 REAKKQVEKALEp161 RVFPRGTVENPDKARELLNK

Bold residues are those contained within p145.

180 WA Hayman et al.

than half of the maximum absorption, were collected. The productsshowed single bands on SDS gels at the expected molecular weights.The structure of each LCP is shown schematically in Fig. 1. StockLCPs were dissolved in sterile PBS at a concentration of 10 mg/mLand stored at –20°C.

Antibody production

Five to six-week-old female mice were immunized subcutaneously atthe tail base with 30 µg of peptide or LCP that had been eitheremulsified in CFA or dissolved in PBS (a total of 50 µL wasadministered). Peptide was prepared at a concentration of 60 µg/50 µL and then emulsified with an equal volume of CFA. The precisevolume was administered using a graduated syringe. Three subse-quent booster injections of 3 µg of peptide or LCP dissolved in PBSwere given at weekly intervals. Before each boost, mice were bled viathe tail artery and sera prepared and stored at –20°C until used.

ELISA

Peptides and LCPs were diluted to 0.5 µg/mL, in carbonate-coatingbuffer, pH 9.6, and coated onto polyvinyl plates (ICN BiochemicalsAustralia, Seven Hills, NSW, Australia) in a volume of 100 µL/wellovernight at 4°C. The peptide or LCP was removed and the wellsblocked with 150 µL of 5% skim milk PBS–Tween 20 for 2 h at 37°C(or overnight at 4°C). The plates were then washed three times withPBS-Tween 20 buffer. Mouse antisera were serially diluted in 0.5%skim milk PBS–Tween 20 buffer, starting at an initial concentrationof 1:100 to a final dilution of 1:12 800. Each sample was diluted to afinal volume of 100 µL and incubated for 1.5 h at 37°C. The plateswere washed five times and peroxidase-conjugated goat antimouseIgG (Tago Inc., Burmingham, CA, USA) added at a dilution of1:3000 in 0.5% skim milk PBS–Tween 20 for 1.5 h at 37°C. Afterwashing, 100 µL of o-phenylenediamine dihydrochloride (OPD)substrate (Sigma Chemical Co., St Louis, MO, USA) was addedaccording to the manufacturer’s instructions and incubated at roomtemperature for 30 min in the dark. The absorbance was measured at450 nm in an ELISA reader (Multiscan; Flow Laboratories, McLean,VA, USA). The titre was defined as the lowest dilution that gave anabsorbance of more than three standard deviations above the meanabsorbance of the control wells (which contained normal mouseserum).

Bactericidal assay

Murine antibodies were assayed for opsonic activity against station-ary phase M6 GAS as described elsewhere.2,3,5,6,23 Briefly, M6 GASwere cultured overnight (a minimum of 16 h) in 5 µL of Todd–Hewittbroth at 37°C. Before the assay, the bacteria were diluted to 10–5 innormal sterile saline (inoculum size = 75). A volume of 50 µL of thebacterial dilution was then added to 50 µL of heat-inactivated(10 min at 60°C) normal mouse serum or heat-inactivated immuneserum. Heparinized non-opsonic human blood (400 µL) was added tothe mixture and the tubes incubated for 3 h at 37°C with gentlerocking. The human blood was screened for lack of direct bactericidalactivity and was not pooled. Fifty microlitres of the 3 h culture wasmixed with 20 µL of horse blood agar and poured onto bacteriaplates. The number of colonies on each plate was counted afterincubation for 24 h at 37°C. The percentage of bacteria killed wasdetermined by counting the number of colonies growing after incuba-tion with immune sera and comparing this to the number of coloniesgrowing after incubation in control mouse sera.

Nuclear magnetic resonance spectroscopy

Samples for nuclear magnetic resonance (NMR) spectroscopy con-tained between approximately 0.2 mmol/L and 3.0 mmol/L peptide orLCP in 90% H2O/10% 2H2O. Additional spectra were recorded forthe free peptide and LCP4 following the addition of approximately25% trifluoroethanol (TFE). Spectra were recorded on a Bruker ARX500 spectrometer (Bruker Bruker Analytik Rheinstetten Baden-Wurttensberg, Germany) with a sample temperature of 293K. For theone-dimensional (1D) experiments, water suppression was achievedusing low polar irradiation during the relaxation delay of 2 s. Thefollowing 2D experiments were recorded: total correlation spectro-scopy (TOCSY) spectra24 with a MLEV1725 isotropic mixing periodof 80 ms and nuclear overhauser enhancement spectroscopy(NOESY) spectra26 with a mixing period of 200 ms. In all 2Dexperiments the carrier was set at the centre of the spectrum on thewater resonance frequency and quadrature detection was used forboth dimensions. Spectra were recorded in phase-sensitive modeusing time proportional phase incrementation (TPPI).27 For theTOCSY and NOESY experiments a water suppression by gradienttailored excitation (WATERGATE)28 scheme was used employinggradient pulses of approximately 6 Gcm–1 either side of a 10 kHz 3-9-19 binomial pulse.

Results

The design of the lipid-core peptides (LCP1→LCP7)

Previous studies have shown that the monomeric peptide(p145) can generate an immune response in B10.BR mice.5–6

Because of the low immunogenicity of this peptide whenadministered as a monomer with conventional adjuvants,repeated boosting was required to attain a high titre(> 12 800) antibody response. In an attempt to overcome thisproblem, LCP technology was employed.

To design an optimized LCP system, seven LCP con-structs based on the p145 sequence were synthesized(LCP1→LCP7). LCP1 contained eight p145 epitopes on thepolylysine core attached to a lipoamino acid trimer with 10carbon long side-chains (Fig. 1). To determine how immuno-genicity was affected by the number of p145 epitopes and bythe number of α-amino acid units, LCP2 and LCP3 weresynthesized. LCP2 contained only four p145 epitopes on thepolylysine core attached to a lipoamino acid dimer with 10carbon long side-chains, whilst LCP3 contained four p145epitopes on the polylysine core (as seen in LCP2) and threeα-amino acid units with alkyl side-chains 10 hydrocarbons inlength (as seen in LCP1; Fig. 1).

A third variable examined was the effect of the length ofthe alkyl side-chain on the immunogenicity of the p145peptide. Four different constructs were synthesized (LCP4,LCP5, LCP6 and LCP7), all of which had four p145 epitopeson the polylysine core and a lipidic backbone of five α-aminoacid units with three lipidic residues. The length of three alkylside-chains, however, ranged from eight, 10, 12 and 18hydrocarbons, respectively (Fig. 1). A glycine spacer was alsobuilt into the lipoamino acid-anchoring moiety in this series.

The immunogenicity of p145 monomer and the lipid-core peptide constructs (1–7) in B10.BR mice

The immunogenicities of the LCP constructs were initiallycompared with the p145 monomer in B10.BR mice. Each

Fine specificity of streptococcal vaccine 181

Figure 1 A schematic representation of the seven lipid-core peptides (LCP) (LCP1→LCP7). LCP1 contained eight p145 sequences on anoligomeric polylysine core attached to three α-amino acid units, with alkyl side-chains that were 10 hydrocarbons in length. LCP2contained only four p145 sequences on the polylysine backbone and only two α-amino acid units. LCP3 contained three α-amino acid unitsand four p145 epitopes on the oligomeric polylysine core. Peptides LCP4, LCP5, LCP6 and LCP7 contained four p145 epitopes and fiveα-amino acid units, in which the length of the alkyl side-chains increased from eight hydrocarbons to 10, 12 and 18 hydrocarbons,respectively.

182 WA Hayman et al.

mouse was immunized with 30 µg of peptide or LCP in CFAand bled at weekly intervals to check the titre of antipeptideantibodies in the serum. Boosts of peptide were given on days21, 29 and 37. As seen in Fig. 2, all LCP constructs(LCP1→LCP7) were more immunogenic than the monomericp145 peptide. Following immunization with p145 monomer,antibodies were not detected by ELISA until day 36. A lowtitre (1280) was detected following the second boost. Follow-ing the third boost, the titre had only doubled to 2560 (Fig. 2).All mice immunized with the LCP constructs had detectableantibody levels by day 20 or 28 (following the first boost),with four constructs giving titres equal to or greater than12 800 (LCP2, LCP5–LCP7) (Fig. 2).

Effect of varying the number of antigenic units on immunogenicity

The titres of p145-specific antibodies in the sera of miceimmunized with LCP1 and LCP3 were compared (LCP1contained eight p145 epitopes, whereas LCP3 had only four).The titre of p145-specific antibody in the sera of miceimmunized with LCP1 was consistently higher than the titreof antibody in mice immunized with LCP3 (Fig. 2). After thethird boost, the average titre of p145 antibodies in the sera ofmice immunized with LCP1 was 15 360, whereas the averagetitre of antipeptide antibodies in mice immunized with LCP3was 6400 (Fig. 2), suggesting that increasing the number ofepitopes within the LCP system enhances immunogenicity.

Figure 2 The immunogenicity of p145 monomer and lipid-core peptides (LCP) 1–7 in B10.BR mice. The peptide and LCPs wereadministered in CFA. Each group consisted of five mice; (a) Day 20; (b) Day 28; (c) Day 36; (d) Day 44.

Fine specificity of streptococcal vaccine 183

Effect of varying the number of α-amino acid units on immunogenicity

LCP2, which has an anchor containing two lipoamino acids,induced a higher titre of p145-specific antibodies than LCP3,which possesses a lipidic anchor with three lipoamino acids.Following the third boost of peptide (day 44) the average titreof the LCP2 antisera was 20 480, whereas the titre of LCP3antisera was only 6400 (Fig. 2). Increasing the number oflipidic α-amino acid units had a negative effect on immuno-genicity in this series.

Effect of alkyl side-chain length and spacing on immunogenicity

As shown in Fig. 2, increasing the length of the alkyl side-chains attached to the α-amino acid units did increase theimmunogenicity of the LCP constructs. LCP4, which hasalkyl side-chains eight hydrocarbons in length, was lessimmunogenic in B10.BR mice than LCP5, LCP6 and LCP7,which have alkyl side-chains 10, 12 and 18 hydrocarbons inlength, respectively. There was a direct relationship betweenlength of side-chains and immunogenicity at each bleed.

Increasing the distance between the alkyl side-chains bythe addition of a glycine spacer between the lipoamino acid-based backbone was also seen to have a positive effect onimmunogenicity. This was observed when comparing theimmunogencity of LCP3 and LCP5, whereby the only structuraldifference was a glycine spacer between the alkyl side-chains.The titre of p145 antibodies in mice following immunizationwith LCP5 was higher (18 133) than those induced in miceimmunized with LCP3 (6400) (Fig. 2).

Immunogenicity of the lipid-core peptides in mice following administration without adjuvant

As discussed elsewhere,17–19 the LCP system enhances thelipophilicity and membrane-binding effect of synthetic peptidesand, as such, may be useful as a combined adjuvant–vaccine.Therefore, it was of interest to determine if the LCP constructs

could induce production of p145-specific antibodies in micewithout the need for any conventional adjuvants.

The immunogenicities of LCP4, LCP5, LCP6 and thep145 monomer were compared when coadministered withCFA and when administered without the use of any adjuvant.As shown in Fig. 3, mice immunized with each of the threeLCP constructs (when coadministered with either PBS orCFA) had produced antibodies to p145. However, the anti-body titre in all cases was found to be higher in the serum ofmice immunized with antigen emulsified in CFA rather thanPBS. Peptide 145 administered without adjuvant did notinduce any antibody response even after repeated boosting.However, when the LCP constructs LCP4 and LCP6 wereadministered without any adjuvant, they proved to be moreimmunogenic than the p145 monomer coadministered withCFA (Fig. 3).

B-cell epitope specificity of antisera derived from mice immunized with either p145 or the lipid-core peptide series

Given the serotypical diversity of GAS in nature,1 the highlyconserved C-terminal of the M protein of GAS is a currentfocus of vaccine development. Of major concern with thedevelopment of a vaccine based on this region, however, isthe potential for immunological cross-reactivity between thisregion of the M protein and host tissue antigens such ascardiac myosin.29–32

It has been shown previously that the immunodominantepitope displayed by p145 is conformational, occurring as anα-helical coil.21 To map the minimal epitopes within p145, apanel of chimeric peptides were synthesized that placed short,12 amino-acid windows of the p145 sequence within theframework of a larger α-helical coiled peptide, Jcon(Table 1).21 Using these peptides, the epitope on p145 thatwas recognized by sera from mice immunized with p145was mapped to residues 7–20 (amino acid sequence:ASREAKKQVEKALE) (J7, J8, J9).5 In the present study, itwas of interest to see if this p145 epitope was recognizedfollowing immunization with the different LCP constructs ina panel of B10.BR mice (four per group). As shown

Figure 3 The immunogenicityof p145 monomer and lipid-corepeptides 4–6 in B10.BR mice fol-lowing administration in eitherPBS or CFA.

184 WA Hayman et al.

elsewhere5 and repeated here (Table 2), p145 antisera boundpredominantly to peptides J7, J8 and J9 (confirming that thedominant epitope on p145 lies within residues 7–20) with onemouse also recognizing J2. However, mice immunized withLCP4, LCP5 or LCP6 recognized fewer epitopes on p145 (asdemonstrated by lack of recognition of J8 and J9), whereasmice immunized with LCP1, LCP2, LCP3 or LCP7 recog-nized multiple epitopes on p145 (Table 2).

We were curious as to whether the antigenic epitopesdisplayed in the LCPs adapted a different structure, whichmight explain the different patterns of B-cell epitope recogni-tion. Thus, we examined the structural characteristics of theLCPs using NMR.

One-dimensional NMR spectra for the free peptide andeach of the seven LCP constructs in aqueous solution were allsimilar, suggesting that the peptide adopted a similar confor-mation in all cases. Two-dimensional spectra were recordedfor the free peptide and several of the constructs. For thespectra acquired in aqueous solution, analysis of the chemicalshifts33 provided no indications of regular elements of second-ary structure in either the free peptide or the constructs LCP3,LCP4, LCP5 or LCP6. All of the cross-peaks observed ineach of the NOESY spectra resulted from intra-residue orsequential interactions, confirming the lack of regular second-ary structure and suggesting that, in the case of the constructs,the individual peptide units do not interact with one another.With spectra acquired for the free peptide and the construct

LCP4, the addition of 25–30% TFE confirmed the previouscircular dichroism (CD) findings that the peptide adopts ahelical confirmation in the presence of TFE.19

Opsonic activity of antisera derived from B10.BR mice immunized with p145 monomer or LCP1→LCP7

Antisera from mice immunized with p145 monomer or LCP1to LCP7 were tested for opsonic activity against M6 GAS ina bactericidal assay. The mean level of opsonic activitydisplayed by p145 antisera was 28%. By comparing individ-ual sera, there was seen to be a correlation between opsonicactivity and titre (Table 3).

The titre of antisera from mice immunized with the sevenLCPs ranged from 12 800 to 1 638 400 against the immuniz-ing peptide (Table 3). The mean bactericidal activity for eachLCP peptide was 55% for LCP2, 34% for LCP3, 63% forLCP4, 39% for LCP5, 27% for LCP6 and 37% for LCP7(Table 3). Pooled LCP1 antisera had an opsonic activity of47%. It was of interest to note that LCP4 antisera, while notexhibiting the highest ELISA titre, showed the highestopsonic activity. The immune response to LCP4 sera was themost focused (Table 2) and recognized J2 and J7 – the samerecognition pattern as that shown for sera from AustralianAboriginals living in streptococcal-endemic areas.3 LCP6antisera, while showing a similar pattern of recognition, hadless opsonic activity and a lower ELISA titre. LCP2 and

Table 2 Binding specificity of p145 and lipid-core peptides (LCP) (1–7) sera to the chimeric J peptides

Peptide on ELISA plate Mouse serum sample

p145 J1 J2 J3 J4 J5 J6 J7 J8 J9 Jcon

M1 145 +++ – – – – – – + +++ + –M2 145 +++ – – – – – – +++ +++ +++ –M3 145 +++ – +++ – – – – +++ +++ – –M4 145 +++ – – – – – – +++ +++ +++ –M1 LCP1 +++ – + – ++ – +++ +++ – – –M2 LCP1 +++ – +++ +++ +++ +++ +++ +++ +++ – –M3 LCP1 ++ – – ++ +++ ++ +++ ++ + – –M4 LCP1 +++ – +++ ++ +++ +++ +++ +++ +++ +++M1 LCP2 +++ – +++ +++ ++ +++ +++ ++ ++ – –M2 LCP2 +++ – +++ +++ +++ ++ +++ + – – –M3 LCP2 +++ + +++ + +++ + ++ + – – –M4 LCP2 +++ +++ +++ +++ +++ +++ +++ +++ +++ ++ –M1 LCP3 +++ – – – + – – – – – –M2 LCP3 +++ – + +++ +++ +++ +++ + – – –M3 LCP3 +++ + ++ – + + + + – – –M4 LCP3 +++ – + +++ +++ ++ +++ +++ + + –M1 LCP4 +++ + +++ – – – – +++ – – –M2 LCP4 +++ – – – – – – +++ – – –M3 LCP4 +++ – +++ – – – – +++ – – –M4 LCP4 +++ – +++ – – – – +++ – – –M1 LCP5 +++ – – – – – – +++ – – –M2 LCP5 + – – – – – – – – – –M3 LCP5 +++ – + – – – – + – – –M4 LCP5 +++ – – – – – – +++ – – –M1 LCP6 +++ – – – – – – +++ – – –M2 LCP6 +++ – +++ – – – – +++M3 LCP6 +++ – +++ – – – – +++ +++ – –M4 LCP6 +++ – +++ – – – – +++ + – –M1 LCP7 +++ – +++ +++ +++ +++ +++ +++ – – –M2 LCP7 +++ – ++ +++ – – ++ +++ – – –M3 LCP7 +++ +++ +++ +++ ++ +++ – +++ +++ – –M4 LCP7 +++ + +++ +++ +++ +++ + +++ + – –

(+++) Titre ≥ 12 800; (++) 12800 > titre ≥ 6400; (+) 6400 > titre ≥ 3200; (–) titre ≤ 1600.

Fine specificity of streptococcal vaccine 185

LCP3 antisera showed broader specificity and less opsonicactivity than LCP4 antisera despite the fact that the ELISAtitre of LCP2 antisera was much greater than the LCP4 sera.

Antibodies to the amino-terminus of the M protein canopsonize the organism completely.1 Peptide 161 representedthe N-terminal epitope from M6 GAS and was used as apositive control in all bactericidals for the present study(Table 1). As shown in Table 3, p161 antisera exhibited apotent bactericidal activity against M6 GAS with a 96%reduction in colonies at a titre of 1600.

Immune response to lipid-core peptides is restricted by H-2 genes

The immune response to p145 in mice has been shownpreviously to be genetically restricted to H-2k.5 In the presentstudy, three congenic strains of mice which varied only at H-2(B10.BR (H-2k), B10.D2/n2 (H-2d) and B10 (H-2b)) wereimmunized with peptides 145, LCP1, LCP2 and LCP3. Ascan be seen in Fig. 4, B10.BR mice were the only strain ableto generate a significant immune response to any of thepeptides. However, two individual B10.D2 mice were alsoable to produce antibodies to LCP1 and LCP3, but themaximum titre achieved was very low (200 and 100, respec-tively) (Fig. 4).

Discussion

A total of seven LCP constructs (LCP1→LCP7) weredesigned and synthesized. Each LCP differed from the otherby either the number of p145 sequences attached to theoligomeric polylysine core, the number of lipoamino acids,the length of the alkyl side-chains or the spacing between thelipoamino acid units. The effect of these variables on theimmunogenicity of the LCP was examined in B10.BR mice.Increasing the number of lipoamino acid units in the back-bone decreased the immunogenicity of the constructs. How-ever, increasing the number of p145 peptide sequences within

Figure 4 The immunogenicityof p145 and lipid-core peptides(LCP) 1–3 in B10.BR, B10.D2and C57BL/10 mice. The peptideand LCPs were administered inCFA. Each group consisted of fivemice.

Table 3 The ELISA titre and opsonic activity of p145 and lipid-corepeptide (LCP) (1–7) antisera against M6 group A streptococci

Mouse serum sample

ELISA titre to immunizing

peptide

Opsonic activity (%)

Individual serum sample

Average for group

N MS – – –LCP1 (pool) 409 600 47 47M1 LCP2 204 800 54 55M2 LCP2 819 200 72M3 LCP2 819 200 40M1 LCP3 3200 42 34M2 LCP3 1600 49M3 LCP3 102 400 14M4 LCP3 102 400 29M1 LCP4 204 800 44 63M2 LCP4 204 800 50M3 LCP4 204 800 65M4 LCP4 204 800 92M1 LCP5 12 800 34 39M2 LCP5 12 800 53M3 LCP5 204 800 35M4 LCP5 102 400 32M1 LCP6 51 200 15 27M2 LCP6 12 800 31M3 LCP6 51 200 43M4 LCP6 51 200 21M1 LCP7 1638 400 32 37M2 LCP7 819 200 44M3 LCP7 819 200 36M4 LCP7 409 600 37M1 145 200 0 28M2 145 6400 51M3 145 800 31M4 145 400 28p161 (pool) 1600 – 96

Percent reduction in CFU = [1 – (mean CFU from immune sera/mean CFU from the non-immune sera)] × 100.

M6 group A streptoccoci inoculum size = 75; 1900 colonies grewin the blood + bacteria control (no antisera).

186 WA Hayman et al.

the LCP system, placing spacer units between the lipoaminoacids and increasing the length of the alkyl side-chains didenhance the immunogenicity of the LCP constructs.

Some LCPs were found to be more immunogenic evenwhen administered without the use of any conventionaladjuvant when compared with the p145 monomer coadminis-tered with adjuvant. Furthermore, high titres of p145 anti-bodies were detected as early as day 20 in B10.BR mice whenimmunized with the LCP constructs, whereas antipeptideantibodies in mice immunized with p145 (coadministeredwith adjuvants) were not detected until day 36 (after thesecond boost). Seven days after the third and final boost withpeptide, the titre of antipeptide antibody in mice immunizedwith the p145 monomer was 2560, whereas the titre ofantibody in the mice immunized with LCP7 was 288 000.This represents an approximate 100-fold increase in the titreof antibody. There is evidence that the lipid side-chains dohave adjuvanting properties (Fig. 3). The side-chains mayfacilitate binding to and uptake by APCs. The overallenhanced immunogenicity after boosting (compared withp145) may relate to ongoing adjuvanticity of the LCPs, butnot p145 monomer, following each boost. Nuclear magneticresonance data were able to show that differences in immuno-genicity did not relate to gross structural differences betweenthe constructs.

Even though the LCPs were able to increase the immuno-genicity of the p145 vaccine candidate, they were unable tobreak the genetic restrictions imposed by murine MHC. Thepeptides LCP1 to LCP3 were tested in three strains of H-2congenic mice, but only B10.BR mice were found to give anantibody response. These data are consistent with the lipidside-chains acting primarily in an adjuvanting role, facilitat-ing binding to and uptake by APCs, but not affectingMHC–peptide binding.

The location of the immunogenic B-cell epitope(s) withinthe p145 sequence, once incorporated into the LCP construct,was also determined. Rheumatic heart disease is thought to bean autoimmune illness and heart cross-reactive B-cellepitopes have been localized in the conserved region of the Mprotein.2,34,35 Peptide 145 is also found within the conservedregion of the M protein. Although there is no associationbetween human antibodies to p145 and disease,2 the humanantibody response to p145 is restricted to two fine epitopes, J2and J7.3 It would be unwise to induce an antibody responsethat targeted a broader range of fine epitopes. The murineB-cell epitope for p145 has been shown previously to belocated within residues 7–20 (J7, J8, J9) (sequence:ASREAKKQVEKALE).5 Therefore, it was of interest todetermine the location of the B-cell epitope(s) within thep145 sequence when it was incorporated into the LCP frame-work. We were able to demonstrate that epitope recognitionof the different antisera differed. Of interest, antibodies withan epitope repertoire identical to that of endemic humanantisera (J2, J7) were induced by one construct, LCP4, andwere more opsonic. Opsonic activity was also related to titre.The ability to modulate fine epitope specificity does haveimportant applications for vaccine design.

Curiously, some individual mice (e.g. M2, LCP5) did notrecognize any of the epitopes defined by J1→J9, but stillopsonized, suggesting that other undefined minimal epitopesalso exist within p145.

Lipid-core peptide technology has the ability to alter thequantity and quality of an immune response and may be ofparticular advantage for vaccine design in which pathogenicimmune responses must be avoided or when critical epitopesmust be targeted.

Acknowledgements

We acknowledge Ms D’Arne Squire and Ms Peta Alexanderfor their expert technical assistance. This work was supportedby The Prince Charles Hospital Foundation, The NationalHeart Foundation of Australia, the Cooperative ResearchCentre for Vaccine Technology, the National Health andMedical Research Council (Australia), and the AustralianCentre for International and Tropical Health and Nutrition.The synthesis of the LCP constructs was supported by Well-come Trust Grant 045936/Z/95.

References

1 Lancefield RC. Current knowledge of type-specific M antigensof group A streptococci. J. Exp. Med. 1962; 89: 307.

2 Pruksakorn S, Currie B, Brandt E et al. Towards a vaccine forrheumatic fever: Identification of a conserved target epitope onM protein of group A streptococci. Lancet 1994; 344: 639–42.

3 Brandt ER, Hayman WA, Currie B et al. Opsonic human anti-bodies from an endemic population specific for a conservedepitope on the M protein of group A streptococci. Immunology1996; 89: 331–7.

4 Brandt ER, Hayman WA, Currie B, Pruksakorn S, Good MF.Human antibodies to the conserved region of the M protein:Opsonization of heterologous strains of group A streptococci.Vaccine 1997; 15: 1805–12.

5 Hayman WA, Brandt ER, Relf WA, Cooper JA, Saul A, GoodMF. Mapping the minimal murine T cell and B cell epitopeswithin a peptide vaccine candidate from the conserved region ofthe M protein of group A streptococcus. Int. Immunol. 1997; 9:1723–33.

6 Pruksakorn S, Galbraith A, Houghten RA, Good MF. ConservedT and B cell epitopes on the M protein of group A streptococci:Induction of bactericidal antibodies. J. Immunol. 1992; 149:2729–35.

7 Seeber SJ, White JL, Hem SI. Predicting the adsorption ofproteins by aluminium-containing adjuvants. Vaccine 1991; 9:201–3.

8 Gupta RK, Relyveld EH. Adverse reactions after injections ofadsorbed diphtheria–pertussis–tetanus (DPT) vaccines are notonly due to pertussis organisms or pertussis components in thevaccine. Vaccine 1991; 9: 699–702.

9 Tam JP. Synthetic peptide vaccine design: Synthesis and proper-ties of a high-density multiple antigenic peptide system. Proc.Natl Acad. Sci. USA 1988; 85: 5409–13.

10 del Guidice G, Tougne C, Louis JA, et al. A multiple antigenpeptide from the repetitive sequence of Plasmodium malariaeCS protein induces a specific antibody response in mice ofvarious H-2 haplotypes. Eur. J. Immunol. 1990; 20: 1619–22.

11 Tam JP, Clavijo P, Lu YA, Nussenzweig V, Nussenzweig R,Zavala P. Incorporation of T and B epitopes of the circumsporo-zoite protein in a chemically defined synthetic vaccine againstmalaria. J. Exp. Med. 1990; 171: 299–306.

12 Zavala F, Chai S. Protective anti-sporozoite antibodies inducedby a chemically defined synthetic vaccine. Immunol. Lett. 1990;25: 271–4.

Fine specificity of streptococcal vaccine 187

13 Gibbons WA, Hughes RA, Szeto A et al. Synthesis resolutionand structural elucidation of fatty amino acids and their homo-and hetero-oligomers. Liebigs Ann. Chem. 1990; 1990: 1175–83.

14 Toth I. A novel chemical approach to drug delivery: Lipidicamino acid conjugates. J. Drug Target. 1994; 2: 217–39.

15 Toth I, Danton M, Flinn N, Gibbons WA. A combined adjuvantand carrier system for enhancing synthetic peptides immuno-genicity utilizing lipidic amino acids. Tetrahedron Lett. 1993;34: 3925–8.

16 Toth I, Gibbons WA. 1992. Synthetic peptide vaccines. BritishPat. Appl. no. 9215780.9 (24 July 1992); EU Pat. Appl. no.93917902.4.

17 Hughes RA, Toth I, Ward P, Ireland SJ, Gibbons WA. Lipidicpeptides. III: Lipidic amino acid and oligomer conjugates ofmorphine. J. Pharm. Sci. 1991; 80: 1103–5.

18 Hughes RA, Toth I, Ward P et al. Lipidic peptides. V: Penicillinand cephalosporinacid conjugates with increased lipophilic char-acter. J. Pharm. Sci. 1992; 81: 845–8.

19 Toth I, Hughes RA, Ward P et al. Fatty peptides. VI: Penicillinand cephalosporin esters with increased lipophilic character. Int.J. Pharm. 1991; 77: 13–20.

20 Zhong G, Toth I, Reid R, Brunham RC. Immunogenicity evalua-tion of a lipidic amino-acid based synthetic peptide vaccine forChlamydia trachomatis. J. Immunol. 1993; 151: 3728–36.

21 Relf WA, Cooper J, Brandt ER et al. Mapping a conserved con-formational epitope from the M protein of group A streptococci.Peptide Res. 1996; 9: 12–20.

22 Houghten RA. General method for the rapid solid-phase synthe-sis of large numbers of peptides: Specificity of antigen–antibodyinteraction at the level of individual amino acids. Proc. NatlAcad. Sci. USA 1985; 82: 5131–5.

23 Lancefield RC. Differentiation of group A streptococci with acommon R antigen into three serological types, with specialreference to the bactericidal test. J. Exp. Med. 1957; 106:525–44.

24 Braunschweiler L, Ernst RR. Coherence transfer by isotropicmixing: Application to proton correlation spectroscopy. J. Magn.Reson. 1983; 53: 521–8.

25 Bax A, Davis DG. MLEV-17 based two dimensional homo-nuclear magnetisation transfer spectroscopy. J. Magn. Reson.1985; 65: 355–60.

26 Kumar A, Ernst RR, Wuthrich K. A two-dimensional nuclearOverhauser enhancement (2D NOE) experiment for the elucida-tion of complete proton–proton cross-relaxation networks in bio-logical macromolecules. Biochem. Biophys. Res. Commun. 1980;95: 1–6.

27 Marion D, Wuthrich K. Application of phase sensitive two-dimensional correlated spectroscopy (COSY) for measurementsof 1H–1H spin-spin coupling constants in proteins. Biochem.Biophys. Res. Commun. 1983; 113: 967–74.

28 Piotto M, Saudek V, Sklenar V. Gradient-tailored excitation forsingle-quantum NMR spectroscopy of aqeous solutions. J.Biomol. NMR 1992; 2: 661–5.

29 Dale JB, Beachey EH. Protective antigenic determinant of strep-tococcal M protein shared with sarcolemmal membranes ofhuman heart. J. Exp. Med. 1982; 156: 1165–76.

30 Dale JB, Beachey EH. Multiple heart-crossreactive epitopes ofstreptococcal M proteins. J. Exp. Med. 1985; 161: 113–22.

31 Kraus W, Dale JB, Beachey EH. Identification of an epitope oftype I streptococcal M protein that is shared with a 43-kDaprotein of human myocardium and renal glomeruli. J. Immunol.1990; 111: 4089–93.

32 Bronze MS, Dale JB. Epitopes of streptococcal M protein thatevoke antibodies that crossreact with human brain. J. Immunol.1993; 151: 2820–8.

33 Wishart DS, Sykes BD, Richards FM. Relationship betweennuclear magnetic resonance chemical shift and protein secondarystructure. J. Mol. Biol. 1991; 222: 311–33.

34 Vashishtha A, Fischetti VA. Surface-exposed conserved regionof the streptococcal M protein induces antibodies cross-reactivewith denatured forms of myosin. J. Immunol. 1993; 150:4693–701.

35 Cunningham MW, Antone SM, Gulizia JM, McManus BA,Gauntt CJ. α-Helical coiled-coil molecules: A role in auto-immunity against the heart. Clin. Immunol. Immunopathol. 1993;68: 118–23.