EMVI sponsored Meeting on Long Synthetic Peptides in Malaria Vaccine Development, Institut Pasteur, Paris, 7 October 1999

Chairperson: Dr. Søren Jepsen

1. Introduction to EMVI: Dr. Søren Jepsen

Dr. Jepsen introduced the formation of the European Malaria Vaccine Initiative (EMVI) following discussions within VINCOMAL and African Malaria Vaccine Testing Network (AMVTN). EMVI is endorsed by 26 Institutions world-wide, with its headquarters in the Centre for International Health, Bergen, Norway. Dr. Jepsen is the Manager of the Secretariat and communicates on behalf of EMVI with the Commission. EMVI has a Board containing representatives of the donor EU Member States (currently France, Italy, Sweden, The Netherlands, Denmark), DG12, INCO, Key Action 2, DG8, AMVTN and Bergen University. Each Board Member nominates 3 Members to the Scientific Advisory Committee (SAC). The Board approves the release of funding.

EMVI has the specific aim of funding the limited GMP production of candidate malaria vaccines and their Phase I testing in human clinical trials.

The first call for proposals was issued November 5, 1998 which received 16 letters of intent. Two proposals were to establish a European and an African Centre for Clinical Trials, and were evaluated separately. The SAC judged 5 research proposals as promising and invited full applications. Two proposals were recommended to be combined, and the 4 full proposals were judged qualified for support at the Board meeting of September 1, 1999. Three proposals focused on the use of Long Synthetic Peptides (LSP), and this meeting was called to advise the Board on the scientific rationale of using LSP’s as malaria vaccines.

The Board had suggested that an advantage of using LSP’s would allow similar adjuvants to be used for the different vaccines, and were cheaper and quicker to use than recombinant proteins. The Board had therefore recommended that one proposal to use recombinant proteins should be amended to use LSP’s.

The three proposals were to use LSP’s for Plasmodium falciparum LSA3, MSP3 and GLURP antigens, either separately, and for a combination of GLURP and MSP3. Therefore the following discussions focused on two specific questions. Firstly the rationale for developing LSP’s and their advantages and disadvantages compared to recombinant proteins, including Quality Assurance (QA) and Quality Control (QC) issues. Secondly whether the antigen sequences chosen for synthesis represented those regions of the proteins that have been shown to be associated with protection to malaria.

There was a general introductory discussion between Dr. Ahmed Bouzidi and Dr. Pierre Druilhe on the advantages of LSP’s as a research tool for Proof of Concept validity as vaccines candidates, based on advantages of speed and cost. It was noted that a small batch could be produced for immunisation of volunteers but issues of scale up were critical. Dr. Bouzidi emphasised that EMEA compliance was essential.

2. Long Synthetic Peptides as vaccines: Dr. Giampietro Corradin

Dr. Corradin described the background to LSP’s, starting in the early 1990’s when new protecting groups and catalysts were introduced in the synthesis process. From an initial 50-60 amino acid peptides, LSP’s can now be synthesised up to at least 100 amino acids using an ABI Synthesiser. By capping at each step, deletions (missing amino acids) can be minimised. Dr. Corradin emphasised the speed and purity of synthesis of LSP’s. He said he could produce 1g unpurified peptide in 3 weeks, and, using a 1 step purification method (histidine-tagged polypeptide, Ni column) produce about 350 mg pure protein.
There was considerable discussion by all participants on the questions of integrity of the peptide, deletions, and impurities. Dr. Corradin emphasised that deletions were minimal, and impurities represented shorter sequences and chemical modifications (usually terbutylation) of the final polypeptide, unlike contaminants in recombinant proteins which would include non-antigen polypeptides.

• It was the view of the Panel that LSP’s represented a significant advance with acceptable purity and integrity.

Where conformation was critical, Dr. Corradin described synthesis of a LSP derived from the P. falciparum Circumsporozoite protein (CSP) representing most of the C-terminal non-repeat region and known to contain important T cell epitopes. He argued that the LSP was in correct conformation as there were no free cysteine-associated SH groups, making chemical cleavage difficult. He also cited chemical synthesis of a model chemokine containing 80 amino acids and two disulphide bridges that was as active as native protein.

• The Panel agreed that LSP’s could be expected to mimic correct conformation and might have the biological activity of native protein.

Dr. Corradin described human clinical trials with his P. falciparum CSP LSP in Switzerland, based on earlier mouse trials using the murine analogue P. berghei CSP LSP which elicited complete CD8 T cell protection in mice. In preliminary studies the P. falciparum CSP LSP also induced CD8 T cell responses in mice. In humans, the CSP LSP mixed with montanide elicited high antibody titers by ELISA to the peptide that recognised the surface of P. falciparum sporozoites, and good CD4 and CD8 T cell responses. These responses were boostable and lasted at least 5 months. Dr. Corradin emphasised that choice of adjuvants with LSP’s was important, as alum gave a poorer response. Based on these encouraging results, Dr. Corradin was now planning a new trial followed by challenge to determine efficacy.

• It was the view of the Panel that these findings were highly relevant to the LS3, MSP3, and GLURP proposals.

More discussion focused on whether the material synthesised by Dr. Corradin was GLP or GMP grade, and from that arose the question of regulatory controls on LSP’s. Dr. Corradin said that his material was GLP, and was approved by the Swiss authorities for limited human use.

3. Manufacturing and Regulatory requirements: Dr. Ahmed Bouzidi

During a thorough presentation, Dr. Bouzidi described to the Panel the various manufacturing steps required for GMP. There was discussion during and after as to whether all these steps were required, and Dr. Corradin pointed out that he was able to proceed to clinical trials without fulfilling all these steps. However, Dr. Bouzidi and others argued that it was in the manufacturers interests to fulfil all GMP requirements as early as possible as any major variations in manufacture could require the regulatory process to start over.

• The Panel later asked that these steps be clearly identified in the proposals.

The overheads used by Dr. Bouzidi are attached to this report in the Appendix. Briefly, there are 5 critical steps: GMP manufacture, final product controls, stability studies, pharmacotoxicological studies, and clinical study protocols. All this GMP process is regulated by the EMEA.

• The Panel agreed that these steps should be followed for manufacturing the malaria LSP’s.

It is required that all GMP manufacturing is done within an approved facility. Dr. Bouzidi described a new GMP facility being completed in Lausanne, Switzerland, completion date March 31, 2000.

• The Panel agreed that the proposed LSP’s could be manufactured in this facility for clinical trial, and that a Vaccine Master Plan (VMP) should be developed.

There was considerable discussion about each of the steps. Of particular concern was the question of acceptable limits of contamination. Dr. Bouzidi indicated that up to 10% contamination, reproducibility has to be assessed in three successive batches, but it was important that the reproducibility of up to 10 batches was determined. It was pointed out that during purification, mass spectroscopy of each fraction of the major product HPLC peak should be performed to determine purity. Various critical points in these steps were discussed, including the Drug Master Files (DMF) of all reagents and relevant SOP’s. The final product control leads to the QC Certificate, based on purity, integrity, toxicity and expected efficacy. The Panel discussed expected efficacy, which is often difficult to establish, and agreed that there should be immunological markers of the product in mice studies, including antibody and T cell responses. The use of monkeys to assay these responses was discouraged.

There was further discussion concerning final product controls. Dr. Bouzidi strongly advocated the series of standards, which Dr. Corradin suggested were not all required at least for initial human trials.

• This was a concern of the Panel, which asked the proposers in their final application to include a GANT chart of these various steps.

The Panel asked for full discussion on the question of the final formulation (LSP and adjuvant) and how this would be regulated. While alum is the only adjuvant licensed for human use, preliminary studies have indicated that oil-based emulsions are far more effective. There was considerable discussion concerning QC of the final emulsion.

• The Panel agreed that a SOP for preparation of the emulsion may be required.

It is likely that the LSP will be supplied as a lyophilized powder that is reconstituted in injectable water, and then mixed (by vortex) with the oil-based emulsion directly before use. The Panel agreed that the reproducibility of the emulsion, as well as its possible toxicology, was vital.

• The Panel agreed that there was considerable confidence that the LSP approach was feasible for these malaria vaccine candidates, and that a list of deliverables was required from the proposers. These include the Vaccine Master Plan, validation reports of the LSP’s production processes, LSP’s lot files, QC certificates, intermediary reports of stability studies, final reports of LSP pharmacotoxicological studies, and clinical study protocols.

One candidate malaria antigen, GLURP, was originally proposed to be expressed as a recombinant protein, and there was a lengthy discussion as to the merits of LSP’s and recombinants. The general consensus was that LSP’s could be manufactured faster than recombinants, were cheaper, and presented fewer regulatory concerns, such as formation of Master Seed Banks for recombinants. However, the Panel was also aware that the problems of contamination of LSP’s might differ from that of recombinants. Since there was a general understanding that eventually the proposed LSP vaccines, after human experimental validation, may be used as recombinants, the Panel was concerned that the LSP’s are as fully characterised as possible to establish the basis for any immunogenicity and efficacy in humans.

• The Panel indicated that it was satisfied that the various manufacturing steps, and potential problems, had been identified.

The Panel next discussed the specifics of each antigen. Since the LSP’s only cover part of the native protein, the Panel wanted assurance that the sequences chosen for synthesis represent the important immunogenic domains.

4. Malaria Vaccine Candidate Antigens LSA3 and MSP3: Dr. Pierre Druilhe

Dr. Druilhe again pointed out that use of LSP’s allows selection of important domains, and contaminants represent smaller amounts of incomplete sequence, with probable antigenic similarity to the full length LSP, whereas the major contaminants of recombinants are other proteins.

Dr. Druilhe stated that 17 over-lapping LSP’s covering LSA3, and other LSP’s representing MSP3, had already been synthesised by Dr. Corradin. He then described the evidence suggesting that LSA3 and MSP3 were important candidates for further testing.

LSA3

LSA3 is an approx. 1500 amino acid protein that he discovered by differential screening of a genomic P. falciparum library. Using recombinant constructs, Dr. Druilhe had identified the DG7-29 region as important, in antigenicity studies with human volunteers immunised with irradiated P. falciparum sporozoites, and in highly convincing mouse, Aotus monkey and chimpanzee studies. This region will elicit CD4 Th1 responses and importantly CD4 g-interferon as surrogate markers of protection. He has used different constructs including lipopeptides, microparticles containing DG7-29, recombinant LSA3 or DNA vaccines. Of particular importance was the findings that 1 of two chimpanzees were protected by a DNA vaccine or recombinant LSA3. Recently, Dr. Herrera in Cali, Colombia has developed a strain of P. falciparum adapted to Aotus monkeys. Dr. Druilhe has now shown that a LSA3 LSP (GP1 representing 120 amino acids of DG7-29) protected 4 of 6 Aotus monkeys. Comparison of montanide and SBS-2 adjuvants showed that SBS-2 gave better responses.

• The Panel fully discussed these results and agreed that the GP1 LSP vaccine should be further developed for human clinical trials.

MSP3

Dr. Druilhe has previously demonstrated that IgG1 and IgG3 antibodies to the P. falciparum merozoite surface protein MSP3 mediated ACDI-dependent protection. The main MSP3 immunogenic region was a conserved globular loop adjacent to helical structures. Human affinity purified antibodies to this domain protected SCID mice in the presence of human monocytes. A LSP, spanning MSP3 amino acids 181-276, has been synthesised by Dr. Corradin and will be tested for immunogenicity using montanide and SBS-2 adjuvants as for LSA3.

• The Panel agreed that this LSP represented the major immunogenic area of interest within MSP3.

5. Malaria Vaccine Candidate Antigen GLURP: Dr. Michael Theisen

Dr. Theisen described immunogenicity studies with the large (>200 kDa) asexual stage antigen GLURP that is detected on the surface of merozoites. There are three domains of interest: R0 (non-repeat), R1 (repeat) and R2 (repeat). IgG1 antibodies to R0 and IgG3 antibodies to R2 mediate ACDI. B cell epitopes have been identified in R0, including three that contain a common amino acid motif KHKD. GLURP is widely recognised by human antibodies and T cells, and antibodies were found to be associated with absence of disease or low parasitemia in studies in Senegal and Ghana.

His original proposal was to use GLURP expressed as secreted recombinant proteins in Lactococcus lactis. However, the SAC had advised that the LSP approach should be used. The Panel discussed the merits of using the L. lactis expressions system, which allows relatively easy purification of the GLURP polypeptide by ultrafiltration and ion exchange chromatography.

• The Panel agreed that the other problems associated with recombinant expression (protein, DNA or chemical contaminants, needs for Master Cell Banks etc.) were a convincing argument to use LSP’s.
• Dr. Theisen agreed that LSP’s were a valid approach for development of a GLURP vaccine.

Discussion then turned to the need to identify T cell epitopes within the GLURP R0 domain. This was recognised as important, and they could be identified in relatively straightforward studies in humans exposed to malaria in Senegal.

• Since the R0 domain is approx. 292 amino acids, it was agreed by the Panel that Dr. Corradin should synthesise 3 long over-lapping LSP’s. The Panel recommended that EMVI should be authorised to provide 6,000 Euro to Dr. Theisen for Dr. Corradin to synthesise the 3 LSP’s (at 2,000 Euro each) as the LSP approach would enable comparison of the three vaccine constructs. Based on these results, Dr. Theisen will design the final LSP.

6. Conclusions

After a concluding discussion, the Panel recommended that:

LSA3, MSP3 and GLURP LSP’s should be synthesised in the Lausanne facility under GMP conditions for human clinical trials.

The proposers (Drs. Druilhe and Theisen) should provide EMVI with a timetable of the Deliverables, and a GANT chart showing how each step will be organised and controlled.

The meeting concluded at 4 pm with agreement from all present on these decisions.

MH

PARTICIPANTS

Claude Auriault
UMR-CNRS 8527 - SEDAC-Therapeutics, Institut de Biologie de Lille, 1 rue du Professeur Calmette, F-59021 Lille cedex, France
Tel.: +33 3 20 87 12 38
Fax: +33 3 20 87 12 33
e-mail: claude.auriault@pasteur-lille.fr

Ahmed Bouzidi
SEDAC-Therapeutics, Institut de Biologie de Lille, 1 rue du Professeur Calmette, F-59021 Lille cedex, France
Tel.: +33 3 20 87 11 04
Mobile: +33 6 87 60 36 61
Fax: +33 3 20 87 12 33
e-mail: ahmed.bouzidi@wanadoo.fr

Joe Cohen (absent)
SmithKline Beecham Biologicals, Research and Development, 89 rue de L’Institut, Rixensart B-1330, Belgium
Tel.: +32 2 656 8331
Fax: +32 2 656 8113
e-mail: joe.cohen@sbbio.be

Giampietro Corradin
University of Lausanne, CH-1066 Epalinges, Switzerland
Tel.: +41 21 692 5731
Fax: +41 21 692 5705
e-mail: giampietro.corradin@ib.unil.ch

Pierre Druilhe
Institut Pasteur, 25 Rue du Docteur Roux, F-75724 Paris Cédex 15, France
Tel.: +33 1 45 68 85 78
Fax: +33 1 45 68 86 40
e-mail: druilhe@pasteur.fr

Hélène Gras-Masse (absent)
UMR8525 and SEDAC-Therapeutics- Institute of Biology of Lille, BP447, F-59021 Lille Cedex, France
Tel.: +33 3 20 87 12 14
Fax: +33 3 20 87 12 33
e-mail: helene.gras@pasteur-lille.fr

Michael Hollingdale, reporter
University of Leeds, Leeds LS2 9JT, UK
Tel.: +44 113 233 2897
Fax: +44 113 233 2882
e-mail: bgymrh@leeds.ac.uk

Arne Holm (absent)
Den Kgl. Veterinær- og Landbohøjskole,Thorvaldsensvej 40, DK-1871 Frederiksberg C, Denmark
Tel.: +45 35 28 24 29
Fax: +45 35 28 23 98
e-mail: ah@pepsyn.chem.kvl.dk

Dr. Søren Jepsen, Chairman
EMVI, Artillerivej 5, DK-2300 Copenhagen S, Denmark
Tel.: +45 32 68 31 88
Fax: +45 32 68 32 28
e-mail: sje@ssi.dk

Henk Stunnenberg
Department of Molecular Biology, Toernooiveld 1, NL-6525 ED Nijmegen,The Netherlands
Tel.: +31 24 3653431
Fax: +31 24 3652938
e-mail: stunnenb@sci.kun.nl

Charles de Taisne
Pasteur Mérieux Connaught Group, 58 Avenue Leclerc, F-69007 Lyon, France
Tel.: +33 4 37 37 71 18
Fax: +33 4 37 37 71 19
e-mail: cdetaisne@fr.pmc-vacc.com

Michael Theisen
SSI, Artillerivej 5, DK-2300 Copenhagen S, Denmark
Tel.: +45 32 68 37 79
Fax: +45 32 68 32 28
e-mail: mth@ssi.dk