PEPTIDES DERIVED FROM A MALARIA PARASITE SURFACE ANTIGEN EFFECT OF ADJUVANTS AND NUMBER OF IMMUNIZATIONS ON THE TOTAL ANTIBODY RESPONSE IN MICE

: Peptides corresponding to known B cell epitopes in a 4SkDa merozoite surface glycoprotein of the human malaria parasite Plasmodium falciparum were conjugated via 6-maleimido caproic acyl N-succinimide ester to bovine serum albumin. The conjugates were used to immunize Balblc mice in saline and with the following adjuvants: Freunds adjuvant, the muramyl dipeptide derivatives murametide and murabutide, and aluminium hydroxide. Three injections of antigen were given intra-muscularly at 21d intervals and sera obtained after each injection. Antibody levels against peptides were measured by ELISA. Freunds adjuvant produced the highest titre of antibodies (up to lo-' after three injections). Antibody levels obtained with alum absorbed antigen reached titres of lo5 - lo4 while antigen administered in salirce alone yielded titres of 10.'. The soluble MDP derivatives did not produce higher antibody levels than antigen administered in saline. The results suggest that three injections of peptide-camer conjugates in alum orpline may be an acceptable procedure for raising sufficiently high titre anti-peptide antibodies in animals and man. Immunolluorcsccncc studies with acctonclmcthanol Lxcd asexual blood stages of P.falcipanun showed that anti-P102 but not anti-P103 sera reacted with the merozoite surface (unpublished observations). This indicates that- the peptide sequence and its location within GYMSSA may be important in eliciting antibodies reactive with native GYMSSA.


INTRODUCTION
Antibodies produced against peptides are used in biological research for a variety of purposes e.g. as probes for determining protein structure, in immunodiagnosis,, in identifying protein corresponding to open reading frames in nucleic acid sequences etc. Sinc'e anti-peptide antibodies are able to react specifically with the corresponding native protein in many immunization with chemically synthesized pcptidcs is also being dcvclopcd as a mclhod of vaccinating against infectious disease^.^-^ 'Peptides are often poorly immunogenic and need to be conjugated to carrier proteins or otherwise suitably modified in order to elicit antibodies! An adjuvant is also regarded as being ,needed to augm'ent the immune response to peptide-carrier conjugates.S The combination of an emulsion of mineral oil and killed mycobacteria known as Freunds complete adjuvant (FCA) is one of the most potent adjuvants known for stimulating cellular immunity and antibody production. However, it is toxic and produces lesions at the site of injection and is.therefore considered to be unsuitable for human use and is not preferred for use with animals. The active component for adjuvant action in mycobacteria has been .characterized as N-acetylmuramyl-L-alanyl-D-isoglutamine or muramyl dipeptide '(MDP)? MDP possesses pyrogenic activity. However, from screening a large, number of MDP derivatives, two compounds, murametide (N-acetylmuramyl-L-Ala-D-Gln-OCH3) MUM and murabutide (N-acetylmuramyl-L-Ala-D-Gln-n-butyl ester) MBU that lack pyrogenic activity, and therefore are candidate adjuvants for human use, were identified.' Aluminium salts (alum) have been widely used in traditional human and veterinary vaccines as adjuvant^.^ Alum has also been used with a recombinant surface protein to induce immunity in. man against hepatitis B .~ Peptide-based vaccines from protein antigens of human malaria parasites are reported to induce partial protection in man when used with alum as an a d j u~a n t .~* '~ In these studies with malaria peptides, the antibody levels developed were reported to be modest (titres of 1120-11400) and cell-mediated responses relatively poor. Conditions for the development of optimal immune responses to peptide antigens therefore require further investigation. A 45 kDa surface membrane glycoprotein on merozoites of the human malaria parasite Plasrnodiltrn falcipantnl'1'12 is one of the candidate molecules for developing a malaria vaccine, since antibodies against 'defined epitopes of the antigen inhibit parasite growth and reinvasion in c~l t u r e s .~~l~~ We have used two peptides derived from the sequence of this antigen in the FC 27 strain of P. f~lcipancm'~ to investigate the effects of FCA, alum, MUM and MBU and the number of immunizations on antibody production in mice.

METHODS AND MATERIALS
Peptides p he sequences of synthetic peptides derived from the malaria parasite surface glycoprotein,11i'2 in the single letter code for amino acids, were as follows :

Peptide
Residues Sequence PI02 122-138 TAADTPTATESISPSPPC P103 207-226 RNNHPQNTSDSQKECTDGNK P102 contains a B cell epitope that is present in one of the allelic forms of the 45kDa antigen and is the target of an inhibitory monoclonal antibody.14 PI03 is a relatively hydrophillic sequence that on independent structural analysis was shown to contain predicted B and T cell epitopes. Both P102 and P103 are recognized by antibodies present in up to 60% of the sera of persons living in a P. falcipanlnt endemic area of Sri Lanka (submitted for publication).
A cysteine was incorporated into the C-terminus of PI02 for coupling, while P103 contained a naturally occurring cysteine in position 221. Peptides were synthesized by standard t-Boc solid phase chemistry as modified for multiple synthesis.15 They were analysed and purified by reverse phase high performance liquid chromatography (HPLC) using a C18 column and a Shimadzu 8A liquid chromatograph (Shimadzu, Tokyo). A gradient of acetonitrile containing 0.1% trifluoroacetic acid was used for elution. The eluted peptides were lyophilised and stored desiccated at -2 0 '~ until use. The HPLC profiles of peptide 102 and peptide 103 and the fraction used in ;he experiments are shown in Figure 1 and Figure 2 respectively. The amino acid composition, determined after complete hydrolysis of the peptides, using the Waters Picotag system (Waters, Bedford, MA), was consistent with the expected sequence of the peptides.

Preparation of peptide carrier conjugates
The peptides were covalently linked to bovine serum albumin using .6-maleimido eaproic acyl N-succinimide ester ( M a ) essentially as described previously.'6 Briefly, to a solution of 5mg bovine serum albumin (BSA) in 0.45ml of 0.01M sodium phosphate buffer, pH 7.5,0.05ml of 50mM MCS in dry dimethylformamide was added

Abs 215nm
Time ( m i d and the mixture gently shaken for 30min at 26'~. 0.05ml of the MCS solution was again added and the incubation continued for a further 30min. The maleimidated protein was then dialysed against 0.1M sodium phosphate pH 6.5.5mg of peptide was dissolved in 0.5ml of 0.1M sodium phosphate, pH 6.5, 1mM EDTA and added to 0.5ml of the maleimidated serum albumin. The mixture was equilibrated with nitrogen, sealed in a nitrogen atmosphere and incubated for 18h at 2 6 '~. The peptide-carrier conjugate was then dialysed against several changes of 0.01M phosphate burrered saline pH 7.2 (PBS).
The conjugation ratio determined by comparison of the M, of the peptide-carrier conjugate with maleimidated carrier by sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDS-PAGE).

Production .of antisera
Groups of 4 'nine month old Balbtc mice were injected, for primary and boosting immunizations, with peptidecarrier conjugate equivalent to 50pg of peptide per mouse intra-muscularly at two sites. Female mice were used for immunization with PI02 and PBS control and male mice for immunization with P103. (a) For FCA immunization, the first injection was given in FCA followed by incomplete Freunds adjuvant for the boosting injections. (b) For immunization with the two MDP derivatives, 100,iig of MBU and MUM in PBS were mixed with the conjugates for the primary and boosting injections. (c) For immunization with alum, aluminium hydroxide precipitate was washed in PBS and the conjugate absorbed onto it for injection. Both primary and boosting injections contained alum. (d) For immunization without adjuvant, mice were injected with the conjugate in PBS.
Boosting injections were given on day 21 and day 42 and the mice tail bled for sera 10d after each injection. After the third injection the mice were sacrificed and bled out. Pre-immune sera were also collected before commencing the immunizations.

Quantitation of antibodies
Anti-peptide antibody levels were determined by enzyme linked immunosorbent -assay (ELISA) using peptides as antigens. Costar 96 well microtitre plates (Biorad, Richmond, CA) were coated with 0.05 ml of a O.Olmg ml-I solution of peptide in PBS overnight at 37'~. Unreacted absorption sites in the wells were blocked with a 5% wt vol-' solution of non-fat-milk powder (Blotto) for 2h at 26'~. The wells were washed and then incubated with 0.05ml of tenfold serial dilutions of individual mouse sera for 2h at 26'~. After washing four time in PBS, the wells were treated with 0.05ml of a 1:2000 dilution of sheep anti-mouse IgGhorse radish peroxidase conjugate with H and L chain specificity (Silenius, Melbourne, Australia) in Blotto for lh. This reagent detects all murine immunoglobulin isotypes. The wells were then washed and the colour reaction developed with 2,2' -azino -bis (3-ethyl benz thiazoline -6sulphonic acid) and H202 for 30min. Optical density at 405'nm was then measured in a Multiskan Plus 11' ELISA reader (Flow labs, Irvine, Scotland). .

RESULTS
Molecular weight determination by SDS-PAGE analysis demonstrated that 12 mols of P102 and 16 of PI03 were present per mol of BSA in the two conjugates (data not shown). The individual mice within a group yielded antisera with similar reactivity and specificity in the ELISA. Hence the mean absorption at 405nm obtained with sera within a group are presented below.
Anti-PI03 antibodies raised with alum as adjuvant specifically reacted with PI03 compared to P102 (Figure 3 when sera were tested against peptides conjugated to an unrelated carrier protein, kcyhole limpet haemocyanin, (KLH), through glutaraldehyde.
Highest titres of anti-peptide antibodies were obtained with Freunds adjuvant when the sera obtained after three injections with P102-BSA ( Figure 4) and P103-BSA ( Figure 5) were tested by ELISA. While antibody titres of 10"-10~~ were observed with Freunds adjuvant, titres of 1 0 -~-1 0~ were seen when alum was -used as an adjuvant and after immunization without adjuvant. MBU and MUM did not significantly increase antibody production when compared to injections given without adjuvant.
Three injections were generally necessary .to obtain significant antibody production with P102-BSA and P103-BSA when Freunds adjuvant was not used since the antibody responses after two injections in such cases were weak (Figures 6 & 7). However, appreciable antibody response was detected with PI03 after two injections when mice were immunized with P103-BSA in PBS only (Figure 7). A comparative analysis of the pre-immune sera and sera obtained after each injection of P~O~-B S A with alum as adjuvant clearly showed that three injections were needed for significant antibody production (Figure 8). Similar results were obtained with the corresponding anti-PI03 sera.

DISCUSSION
The results show that immunization with Freunds complete adjuvant followed by Freunds incomplete adjuvant for subsequent injections gave the best result with peptide-BSA conjugates for anti-peptide antibody production. However, after three injections, the antibody response to injections of antigen 'without adjuvant and with alum as adjuvant were sufficiently strong to yield antibody titres of l o 3lo". Similar results were obtained with two additional peptides conjugated to BSA (unpublished observations). Thus immunization without adjuvant or with alum as an adjuvant for peptide-carrier conjugates may be sufficient for manypurposes and has the advantage that it avoids the use of an adjuvant that is toxic and produces locally severe lesions.
MBU and MUM did not show significant advantage as adjuvants in the present study. This may be because MBU and MUM, as is the case with MDP are clearedrapidly from blood. Hence delivery in an aqueous medium1' may be less effective than when MDP or its derivatives are given as an emulsion with mineral oil and antigen6 The covalent linkage of MBU and MUM to the peptide-carrier conjugate may be an alternative way of enhancing adjuvant action since this has been shown to be effective with MDP.'~ Immunolluorcsccncc studies with acctonclmcthanol Lxcd asexual blood stages of P.falcipanun showed that anti-P102 but not anti-P103 sera reacted with the merozoite surface (unpublished observations). This indicates that-the peptide sequence and its location within GYMSSA may be important in eliciting antibodies reactive with native GYMSSA.
In general, the antibody responses to P102 and PI03 induced by the different immunization schedules were similar except that PI03 induced a greater antibody response when injected in PBS than in alum. The reason for this discrepancy is not clear, but it may be related to the presence of a T-cell activating determinant in ~1 0 3 '~ but not P102. Cellular as well as humoral immunity has been shown to act against asexual blood stage malaria parasites in man2' No attempt was made to measure T cell responsivenesss in the present investigation. Immunization with alum is generally considered to induCe poor T cell immunity.7 However, significant T-cell responsiveness was observed in mice immunized in alum with a herpes simplex virus glycoprotein derived peptide conjugated to o~a l b u m i n .~~ Different immunoglobulin isotypes that are reported to be selectively induced by different adjuvants: and which may have different effector functions in immunity, were also not measured in the present investigation. Immunization with peptide vaccines derived from the sporozoite9 and asexual blood stage1' P. falcipanun proteins, in alum as an adjuvant, produced imrnunofluorescence and ELISA antibody titres of 1120-11400 in human sera. The poor antibody response to the sporozoite protein was attributed to previous exposure to the tetanus toxoid carrier inhibiting the antibody response,9 a process which may be overcome with further antigen injection^.^^ Antibody titres observed in our study are comparable to the serum antibody levels against a 185-200 kDa merozoite surface protein that completely protect owl monkeys against P. falcipanrrn malaria (imrnunofluorescence titres of 10-~-10").~~ Higher ELISA antibody titres (lom5) against sporozoite antigens have been produced in mice using alum absorbed liposomes containing lipid A and peptide-carrier. 24 The toxic effects of lipid A and liposomes however need to be further investigated before human use.' Our results show that useful antibody levels against peptides can be produced in mice, and potentially therefore in humans, after three injections of peptide-carrier conjugates in .PBS or aluminium hydroxide, an adjuvant commonly used in man and domestic animals.