CHEMICAL BIOLOGY / BIOLOGICAL CHEMISTRY 334 CHIMIA2001, 55, No.4 Chimia 55 (2001) 334-339 ©. Schweizerische Chemische Gesellschatl ISSN 0009-4293 Applications of Protein Epitope Mimetics in Vaccine Design. A New Supersecondary Structure in the Circumsporozoite Protein of Plasmodium fa/ciparum? Bernhard Pfeiffera, Rafael Moreno b, Kerstin Moehle a, Gerd Pluschke b*, and John A. Robinson a* Rinaldo Zurbriggen C, Reinhard GIOckc, A b s tra c t: An approach to synthetic vaccine design is illustrated, focusing on the immunodominant (NPNA)n repeat region of the circumsporozoite (CS) protein of the malaria parasite P la s m o d iu m fa lc ip a ru m . Modelling suggests that the NPNAN motif may adopt a helical ~-turn, which is tandemly repeated in the CS protein to generate a novel supersecondary structure. Cyclic peptidomimetics of this NPNAN motif were synthesized and shown by NMR to adopt helical turns in aqueous solution. When incorporated into Immunopotentiating Reconstituted Influenza Virosomes (IRIVs), humoral immune responses were generated in mice that crossreact with native CS protein on sporozoites. IRIVs are a human-compatible delivery system that appear generally suitable for inducing antibody responses against conformational epitopes using constrained peptidomimetics. This approach may offer great potential for the design of molecularly defined synthetic vaccines, including those targeted against multiple antigens and development stages of P . fa /c ip a ru m , or against other infectious agents. Keywords: Peptide· Peptidomimetic . Protein folding' Introduction Malaria is one of the world's most debilitating diseases, with more than 2 billion people currently at risk worldwide [1], and a toll of several hundred million illnesses, and 1.5-2.7 million deaths annually (WHO, World Health Report, 1998). There is presently no effective vaccine against the parasite, and older established drugs like chloroquine are rapidly losing their effectivenesss due to resistance. On the brighter side, ongoing efforts to sequence the 14 chromosomes and around 7000 genes in the malaria parasite [2] Prof. J.A. Robinson" "Institute of Organic Chemistry University of Zurich Winterthurerstrasse 190 CH-8057 Zurich Tel.: +41 1 6354242 Fax: + 41 1 6356833 E-Mail: robinson@oci.unizh.ch 'C o rre s p o n d e n c e : h ttp ://w w w .u n iz h .c h /o c i/ bSwiss Tropical Institute cSwiss Serum and Vaccine Institute, Bern Secondary structure' will no doubt provide many exciting opportunities for the discovery of new drug targets [3][4], as well as highlighting new antigens as potential malaria vaccine candidates [5][6]. A vaccine against the extracellular stages of the parasite should function by stimulating the immune system to produce antibodies that recognize the intact parasite. This can, in principle, be achieved with either an attenuated or irradiated parasite, or a subunit such as a recombinant surface protein, or a specific surface epitope in the form of a synthetic peptide. One of the many difficulties in designing an effective malaria vaccine is the complex life cycle of the parasite [7][8]. When an infected A n o p h e l e s s p . mosquito bites a human host, thousands of threadlike sporozoites enter the bloodstream. Within a matter of minutes, the sporozoites can invade liver cells, where they are hidden from the antibody-arm of the immune system. Within the liver cell, the sporozoite develops into a schizont containing more than 10 000 uninucleat- Synthetic vaccine ed merozoites. The growing parasite causes the hepatocyte to rupture, releasing the merozoites into the blood stream. The merozoites can invade blood cells, and undergo further multiplication. Alternatively, merozoites can develop into a sexual stage known as gametocytes, which can reinfect mosquitoes. In the mosquito gut, an oocyst is formed, out of which new sporozoites emerge, thereby completing this remarkable life cycle. Not surprisingly, the surface proteins on sporozoites that stimulate an immune response are not the same as those found on merozoites, or indeed on other developmental stages of P . j a l c i p a r u m . Hence an effective malaria vaccine will most likely need to be composed of several immunogens targeting multiple developmental stages of the parasite. It has been known for some time that sporozoites attenuated by X-irradiation can induce a protective, immune response against malaria challenge [9]. The dominant antibody target on these attenuated sporozoites is the major surface pro- CHEMICAL BIOLOGY / BIOLOGICAL 335 CHEMISTRY CHIMIA 2 0 0 1 ,5 5 , No.4 tein, the circumsporozoite (CS) protein [10]. The central portion of the CS protein (M r <= 44 kDa [11][12]) contains 41 tandem repeats of a tetrapeptide, 37 of which are Asn-Ala-Asn-Pro (NANP) and four of which are Asn-Val-Asp-Pro (NVDP). In fact, sequencing studies have shown that many proteins and genes of malaria parasites contain extensive arrays of tandemly repeated amino acid motifs [13]. It was shown that linear, tandemly repeated NANP peptides can elicit antibodies in mice and rabbits that recognize the native CS protein and block sporozoite invasion of hepatocytes [12][ 14-16]. These results were a prelude to vaccination studies in humans with synthetic (NANP)3 peptides conjugated to tetanus toxin, which proved that an anti-sporozoite immune response could be generated in this way, but the efficacy was not good enough for use as a malaria vaccine [17]. Subsequently, a number of studies were initiated to optimise the immune response to (NANP)n peptides (e .g . [1 8 21]). It is noteworthy in all these efforts, that the conformation(s) of the NANP repeats in the CS protein was not known, and so could not be taken into account in the design process. It seemed most likely that short linear (NANPh peptides would be largely unstructured in aqueous solution, and be susceptible to rapid proteo- lytic degradation in serum. A later study also suggested that a significant part of tural unit NPNAN, commonly quoted the immune response against a linear (NANPh peptide is directed against the chain terminii [22], which of course are not present in the native CS protein. Conformation plays a key role in the ability of peptides to elicit antibody responses against folded proteins. Linear peptides often elicit antibodies that bind well to denatured proteins, but less frequently recognize conformational epitopes in native protein structures. Unfortunately, the structure of the (NANP)-repeat region in the CS protein is still unknown. It is likely, however, to adopt a stable and repititious structure. Early theoretical studies led to models involving various helical-like structures [23-25]. It was also suggested that the NPNA repeat unit (1) might adopt a stable type-I ~-tum (Fig. 1) [24]. This is important, since it implies that the repeating structural motif is formed by the turn-forming (NPNA)n cadence rather than by (NANP)n' Experimental support for this idea was obtained later from NMR studies of peptides in aqueous solution with both the NANP and NPNA cadences [26]. These studies provided evidence that linear (NPNA)n peptides exist in a dynamic equilibrium between unfolded and ~-tum-like hydrogen-bonded, folded conformations, with the folded forms encompassing the struc- NPNA. rather than the more four residue ~-tum Stabilization of ~-tums in the NPNA motif has been achieved both by C(a)backbone methylation of proline (2) [27] [28] and by incorporation of NPNA motifs into template-bound cylic peptidomimetics (3) [29] and (4) [30], without abolishing the ability of these analogues to elicit sporozoite cross-reactive antibodies in mice. Interestingly, the mimetic 4 exhibits in NOESY spectra strong d N N (i,i+ 1) connectivities between Asn5 6 6 and Ala as well as Ala and Asn7, indicative of a helical ~-tum within the NPNAN motif, of the type suggested earlier by Dyson and coworkers [26]. A molecular model of 4, consistent with the NMR data, predicts a helical tum for the NPNAN unit comprising a type-I ~-turn, with the Asn3 CO in H-bonding distance of the Ala6 HN, and in addition the possibility of an i (Asn3 CO) to i+ 4 (Asn 7 HN) hydrogen bond, i . e . with Ala6 in the aregion of 1j>/'l1 space. We are exploring an approach to synthetic vaccine design, which involves using protein epitope mimetics (PEMs), in the form of cyclic peptidomimetics, coupled to a human compatible adjuvant, for the induction of antibody responses against conformational epitopes. Allied with the use of combinatorial chemistry Pro4-AsD5 Ja AJD3 Ala2 , 6-AsD7 "- Pros I9 AiD! \ HN 1 R=H 2 R=Me ASD / y«:-<:m 10 o } HOOC 4 AcHN-(NPNAh-CONH2 5 Fig. 1. Structures ofthe NPNA motif and related peptidomimetics. The arrow on structure 5 indicates the position forthe crosslink (see text) CHEMICAL BIOLOGY / BIOLOGICAL 336 CHEMISTRY CHIMIA methods, this approach may have great potential for the identification and optimization of molecularly defined synthetic vaccine candidates, in a form directly suitable for human clinical trials. We describe below some results of these studies, focusing on the NPNA-repeats of the malaria parasite, which lead to a new proposal as to how this region might fold in the native CS protein. This information may be of value in attempts to design more effective anti-sporozoite vaccines. on sporozoites. Examination of molecular models suggested that a suitable cross-link could be formed by introducing an amino group at the ~-position of Pr06 and amide coupling to the spatially adjacent side chain carboxyl of Glu as a replacement for Alal6, i . e . as indicated in Fig. 1-3. A model of this cross-linked peptidomimetic was constructed, and the model also remained in the expected conformation during MD simulations in water (Fig. 3B). 2001, 55, NO.4 by the peptide backbone. Finally, the 20-mer 12 was acetylated to give 13 for conformational studies by NMR, and 12 was also coupled v i a a succinate linker to a regioisomer of phosphatidyl ethanolamine (PE', I -palmitoyl-3-0Ieoyl-phosphatidylethanolamine) to afford the conjugate 14 ready for incorporation into an IRIV (see below). A Molecular Modelling Synthesis of the Peptidomimetic Using the backbone <1>/\1' angles for Asn3-Asn7 taken from earlier models of 4 [30], a linear peptide was built with the sequence Ac-(NPNA)s-NH2 (5) wherein the helical turn conformation (with the appropriate backbone <I>/'l' angles) was also tandemly repeated. The resulting model of 5 (Fig. 2) was stable in MD simulations in water solvent, and adopted the expected repititious supersecondary structure shown in Fig. 3A. Conceivably, this supersecondary structure may be close to the preferred conformation of the NPNA-repeat region in the native CS protein. To explore this idea further, we set out here to stabilize this supersecondary structure by appropriate cross-linking of the peptide backbone in 5, and by examining the ability of the resulting crosslinked peptidomimetic to elicit antibodies that recognize the native CS protein The required orthogonally protected (2S,3R)-3-arninoproline (9) was prepared from the known ~-lactam 6 [31] as shown in Scheme 1. The chemistry is straightforward, and the synthesis proceeds in good yields. The required cross-linked peptidomimetic was prepared by solidphase peptide synthesis methods, as outlined in Scheme 2. The 20-mer peptide 10 was assembled using Fmoc-chemistry. Cleavage from the resin and removal of side-chain protecting groups proceeded in one step to afford 11. The key backbone coupling of the Apr06 and Glu16 side chains was then achieved in a remarkably clean and high yielding cyclization in DMF with HATD. Monitoring the reaction by HPLC showed essentially quantitative cyclization of the precursor (data not shown). This high efficiency probably reflects the fact that the required conformation is strongly preferred Fig. 2. A computer model of 5 in a stable conformation in which each NPNAN motif adopts a helical ~-turn (see text). The extended (NPNA)nsequence in the CS protein may adopt a repeated helical ~-turn supersecondary structure related to that shown here. The arrow indicates the position for the cross-link (see text). Colour coding; Asn = pink, Pro = cyan, Asn = coral, Ala = yellow. The C(a) atoms are marked with a ball Fig. 3. A, Representation of superimposed structures taken from an MD simulation in water of peptide 5 residues Asn 5-Ala 16 with the repeated helical ~-turn supersecondary structure (see Fig. 2). The sausage was calculated using the average displacement of the C(a) atoms, which is represented by the spline radius. The same colour coding used in Fig. 2 applies here. B, As in A, except the molecule is a model of peptidomimetic 13. The Apr0 6 residue is in cyan and Glu 16 is in green. C, As in B, except the molecule is the average NM R structure of 13 deduced in water by NMR and dynamic simulated annealing. The sausage now represents the average displacement of the C(a) atoms of four NMR structures. The cross-linked residues are shown in ball-andstick. The N- and C-terminal NPNA motifs are omitted for clarity. This Fig. was prepared using MOLMOL [34] CHEMICAL BIOLOGY / BIOLOGICAL 337 CHEMISTRY CHIMIA Conformational Studies HO -; The preferred solution conformation of 13 was studied by NMR and MD methods in aqueous solution at pH 5 and 293K, in analogy to previous studies of 3 [29]. The 1D IH NMR spectra indicated the presence of a major conformer and two minor ones (ratio 80: 14:6), the latter two most likely arising due to c i s - t r a n s isomerism at Asn-Pro peptide bonds, in analogy to earlier studies [26][27][29]. The minor forms were not considered further. A full assigment of the IH spectrum of the major form was complicated by chemical shi~t overlap, particularly of the Asn H-C(I3) resonances. However, the peptide backbone HN, H-C(a) resonances could be assigned unambiguously. 2D NOESY spectra showed strong d N N (i,i+ 1) connectivities between the peptide NH groups of Asn 7 and Ala8 as well as Ala8 and Asn9 in the first helical turn, Asn II and Ala 12 as well as Ala 12 and Asn 13 in the next helical turn, and between Asn 15 and GIu 16 as well as Glu 16 and Asn 17 in the last helical tum. These together with the observation of long range NOEs between Pro H-C(a) (i+1) and Ala HN (i+3), provide evidence for three relatively stable helical turns form- Mtt Mtt I I Mtt i, ii <;.r.(0 •• Z Z iii, iv II 7 6 cf..HBOC v -v ii N = Fmoc-Apro(Boc)-OH Scheme 1. i, (TfhO, CH 2 CI2 , pyridine (98%); ii, NaBH 4 , THF/DMF (58%); iii, K 2 S 2 0 S' Na2HP04, MeCN/H 2 0 (80%); i v , (BocbO, CH 2 CI, DMF, Et3N (76%); v , LiOH, THF, H 2 0 (99%); v i , Pd-C, MeOH, H 2 (93%); v ii, Fmoc-Osucc, iPr2NEt, CH 2CI (68%) ed by the residues Asn5-Asn9, Asn9Asn13, and AsnI3-Asn17. Average solution structures for 13 were calculated using NOE-derived distance restraints by dynamic simulated annealing and MD simulations, using methods described earlier [29]. The resulting average structures reveal a common core comprising the anticipated three helical turns from Asn5-AsnI7, with higher flexibility in the regions of the N- and C-termini (Fig. 3C). The backbone con- Mtt I I Mtt Mtt I I Mtt Mtt I I formation of the central region, however, corresponds closely to the expected supersecondary structure deduced for models of 5 (Fig. 3A and 3B). We conclude, therefore, that although the mimetic is not rigid, it can adopt a supersecondary structure comprising three interlinked helical turns, each based on the (NPN AN) motif. In future work we will seek to strengthen this conclusion, through the synthesis and conformational analysis of other cross-linked peptidomimetics. Mtt I ~ 10 Fmoc-Asn-Pro-Asn-Ala-Asn-AP.ro-Asn-A1a-A~m-Pro-Asn-Aia-ASn-Pro-Asn-Glu-ASn-Pro-Asn-Aia-Co-NHI I Boc 11 illu t i, TFA, iPr3SiH, H20 (95:2.5:2.5) Fmoc-Asn-Pro-Asn-Ala-Asn-Apro-Asn-Ala-Asn-Pro-Asn-Ala-Asn-Pro-Asn-Glu-Asn-Pro-Asn-A1a-CONH 2 tI ii, HATU iii, Piperidine, DMF RHN-Asn-Pro-Asn-Ala-Asn-Apro-Asn-Ala-Asn-Pro-Asn-Ala-Asn-Pro-Asn-Glu-Asn-Pro-Asn-Ala-CONH 2 1 1 12 R=H I t 13 R=Ac 1 14 5 010 10 iv , Succinic anhydride,DMF, DMAP v, PE', HATU 15 20 HN- ~n-Pro-Asn-A1a-~n-Apro-Asn-Ala-Asn-P~o-Asn-A1a-Asn-Pro-A;n-Glu-Asn-Pro-Asn-A~-CONH2 HN I ~ \./'0-,-0 o I -(oeo oeo Scheme 2. Synthesis of the peptidomimetics 9 COOH I Fmoc Mtt I DMB DMB <;.r.(o 2001. 55. NO.4 13 and 14 CHEMICAL BIOLOGY / BIOLOGICAL 338 CHEMISTRY CHIMIA 2001.55, NO.4 Immunological Studies For immunizations, the constrained peptidomimetic was coupled to PE' (14), and then incorporated into an IRIV. IRIVs are spherical, unilammelar vesicles, prepared from a mixture of phospholipids and influenza virus surface glycoproteins [32]. The hemagglutinin membrane glycoprotein of influenza virus plays a key role in the mode of action of IRIV s. This major antigen of influenza virus is a fusion-inducing component, which facilitates antigen delivery to immunocompetent cells. The IRIV technology has been licensed already for human clinical use. In the case of the hepatitis A vaccine Epaxal-Berna™ [33], the first IRIV-based non-influenza vaccine to receive a product license for human use from a national authority, the hepatitis A antigen spontaneously binds to the IRIY. In this study, we have coupled the peptidomimetic covalently to PE' (14) and incorporated the mimetic-PE' conjugates into the IRIV. Since IRIVs have been licensed already for human use, the peptidomimetics can be tested in a format that directly allows human clinical studies. Antibody responses elicited by IRIV cross-reaction was demonstrated by a competition experiment. Incubation of the antiserum with the sporozoites in the presence of 13 completely abolished immunostaining (Fig. SB). The IRIV formulations thus elicited a significant proportion of parasite binding antibodies among the total anti-mimetic immune response. Future studies will focus on more detailed analyses of cross-reactivity of monoclonal anti-14-IRIV antibodies and structural studies to determine the conformations of bound antigen. Synthetic linear peptides are often compromised as vaccine candidates due to their inherent flexibility and susceptibility to proteolysis. Linear peptides often elicit antibodies that bind well to denatured proteins, but less frequently recognize conformational epitopes in native protein structures. A further problem is the weak immune responses elicited by linear peptides, even conjugated to carrier proteins, when administered in alum, 3.5 .0 ....• 00 s::1 3 2.5 Q) 2 ..-Q 1.5 0. 1 C'd u ....• ..•..• 0 0.5 0 1.E-Ol 1.E-02 1.E-03 I.E-04 1.E-05 1.E-06 Dilution 3.5 loaded with 14 were studied in BALB/c mice. After a pre-immunization with the influenza vaccine Inflexal Berna™ (Berna-Products, Bern, Switzerland), and three doses of IRIV-14, the sera of all immunized mice contained mimeticspecific antibodies, as demonstrated by ELISA with 14 coated on ELISA plates (Fig. 4A). The cross-reactivity of these anti-sera with the template-bound peptidomimetics 3 and 4 was also analysed by ELISA. The sera from three of four mice immunized with 14-IRIV cross-reacted with the mimetic 4 (Fig. 4B), but none reacted with 3 (Fig. 4C). It is interesting to note that a helical NPNAN tum is possible in 4 but not in 3 [30]. That a significant part of the antibody response to 14IRIV cross-reacts with 4 also means that these cross-reacting antibodies should not recognize alone the ends of the peptide chain in 14, but rather the novel helical-tum supersecondary structure in the central part of the molecule. However, structural studies with monoclonal antibodies will be necessary to confirm this conclusion. The binding of anti-14-IRIV antisera to the CS-protein was analysed by an indirect immunofluorescence assay using P . fa lc ip a r u m sporozoite preparations. In all immunized animals a significant anti-sporozoite antibody response was detected (Fig. SA). The specificity of the Significance C ....• 3 s::1 2.5 00 ................ ................ Q) ..-Q C'd 2 u ....• ..•..• 1.5 0. 1 0 0.5 0 LE-O1 1.E-02 1.E-03 1.E-04 Dilution 3.5 .0 ....• ~ c 3 2.5 Q) ..-Q ~ ....• ..•..• 0. o 2 1.5 1 0.5 o 1.E-OI 1.E-02 I.E-03 1.E-04 Dilution Fig. 4. Serum IgG titres of BALBlc mice immunized three times with 14-IRIV. ELISA was performed in ELISA microtiter plates coated either with 14 (A), a PE-conjugate of 4 (B) or with 3 conjugated to a multiple antigen peptide for coating on the ELISA plates (C) and incubated with serial dilutions of the sera of individual mice. Bound IgG was detected using alkaline phosphatase-conjugated antibodies specific for mouse gamma heavy chains CHEMICAL BIOLOGY / BIOLOGICAL 339 CHEMISTRY CHIMIA 2001,55. NO.4 A cross-react with the native CS protein on P . fa lc ip a r u m sporozoites. Further stud- ies are now planned with a small library of related mimetics, to provide further B support for the biological relevance of this new supersecondary structure, and to further optimise the immunological response with a view to application as a potential anti-sporozoite vaccine candidate. In general, this approach appears to offer great potential for the design of molecularly defined combined synthetic vaccines, including those targeted against multiple antigens and development stages of P . f a l c i p a r u m , and against other infectious agents. W.R. Ballou, R.A. Wirtz, I.H. Trosper, R. L. Beaudoin, M.R. Hollingdale, L.H. Miller, c.L. Diggs, M. Rosenberg, ence [16] F. Zavala, J.P. Tam, M.R. Hollingdale, A.H. Cochrane, I. Quakyi, R.S. Nussenzweig, V. Nussenzweig, S c i e n c e 1 9 8 5 , 228, 1436-1440. [17] D.A. Herrington, D.F. Clyde, G. Losonsky, M. Cortesi a, 1.R. Murphy, J. Davis, S. Baqar, A.M. Felix, E.P. Heimer, D. Gillessen, E. Nardin, R.S. Nussenzweig, V. Nussenzweig, M.R. Hollingdale, M.M. Levine, N a t u r e 1987, 328, 257-259. [18] H.M. Etlinger, L. Renia, H. Matile, M. Manneberg, D. Mazier, A. Trzeciak, D. Gillessen, E u r . J . I m m u n o L . 1991, 21, 1505-1511. [19] J.P. Tam, P. Clavijo, Y.-A. Lu, V. Nussenzweig, R. Nussenzweig, F. Zavala, J . E x p . M ed. A c k n o w le d g e m e n ts The authors are grateful to the Swiss National Science Foundation and the Swiss Commission for Technology and Innovation for supporting this work and Dr. Pflieger (Roche, Basel) for a generous gift of compound 6. 1990,171,299-306. [20] A. Pessi, D. Valmori, P. Migliorini, C. Tougne, E. Bianchi, P.-H. Lambert, G. Corradin, G.D. Giudice, E u r . 1. I m m u n o l . 1991,21,2273-2276. [21] G.A. deOliveira, P. Clavijo, R.S. Nussenzweig, E.H. Nardin, V a c c i n e 1994, 12, 1012-1017. Received: February 20, 2001 [22] H.M. Etlinger, A. Trzeciak, R oy. Soc. Lond. B Fig. 5. Immunofluorescence staining of P. fa lc ip a ru m sporozoites by mouse anti-14-IRIV antiserum, uSing a File-labelled secondary anti-mouse IgG antibody (A). Incubation of the primary antibody with the parasites in the presence of the mimetic 13 (50 J.lg/ml) abolished staining of the sporozoites (8) [1] E. Marshall, S c ie n c e [2] M.J. Gradner, 2000, 2 9 0 , C urro O p in . 4 2 8 -4 3 0 . G e n e t. 438. P a r a s ito l. Today 2000, 16, 434-438. the commonly used human compatible adjuvant. In this work, an approach to synthetic vaccine design is followed, in which cyclic peptidomimetics are presented to the immune system in multiple copies on the surface of Immunopotentiating Reconstituted Influenza Virosomes (IRIVs). These virosome particles contain also influenza virus proteins that facilitate uptake of the virosome by immunocompetent cells. IRIVs have been licensed already for human use, so the peptidomimetics can be tested in a format that directly allows human clinical studies. Here, peptidomimetics of the central (NPNA)n repeat region of the circumsporozoite (CS) protein of the malaria parasite P l a s m o d i u m f a l c i p a r u m have been studied. Previous NMR and modelling studies suggests that NPNAN units in this region adopt a helical p-tum, which may be tandemly repeated to form a novel supersecondary structure. To test this proposal, a cyclic mimetic was prepared, and shown by NMR methods to adopt a preferred conformation having three tandemly repeated helical turns. Antibodies raised against the mimetic were shown to lm m u n o l. A.V.S. Hill, C urro O p in . 2000, 12, 437-441. [6] R.P. Anders, A. Saul, P a r a s ito l. [7] C.E. Chitnis, M.J. Blackman, Today P a r a s ito l. 2000,16,411-415. [8] L.H. Bannister, J.M. Hopkins, R.E. Fowler, S. Krishna, G.H. Mitchell, P a r a s i t o L . Today P roc. N a tl. A cad. Sci. U SA 1987, 84, 4470-4474. K.D. Gibson, H.A. Scheraga, A cad. Sci. U SA P roc. N a tL . 1986, 83, 5649-5653. [26] H.J. Dyson, A.C. Satterthwait, ner, P.E. Wright, B i o c h e m i s t r y R.A. Ler- 1990, 29, 7828-7837. 2000,16,444--447. Today T rans. 6 9 -7 2 . 391-419. [25] [5] M. Plebanski, P h il. [24] B.R. Brooks, R.W. Pastor, F.W. Carson, [3] I. Macreadie, H. Ginsburg, W. Sirawarapam, L. Tilley, P a r a s i t o L . T o d a y 2000,16, [4] D.J. Carucci, 1993, 3 4 0 , [23] I.K. Roterman, K.D. Gibson, H.A. Scheraga, J . B i o m o l . S t r u c t . D y n a m . 1989, 7, D ev. 1999,9,704--708. S ci- 1985, 228, 958-962. 2000, 16,427-433. [9] L.H. Miller, R.J. Howard, R. Carter, M.P. Good, V. Nussenzweig, R.S. Nussenzweig, S c i e n c e 1986, 234, 1349-1356. [10] P. Potocnjak, N. Yoshida, R.S. Nussenzweig, V. Nussenzweig, J . E x p . M e d . 1980,151,1504--1513. [11] G.N. Godson, J. Ellis, P. Svec, D.H. Schlesinger, V. Nussenzweig, N a t u r e 1 9 8 3 , 3 0 5 ,2 9 - 3 3 . [12] J.B. Dame, J.L. Williams, T.F. McCutchan, J.L. Weber, R.A. Wirtz, W.T. Hockmeyer, W.L. Maloy, J.D. Haynes, I. Schneider, D. Roberts, G . S . Sanders, E.P. Reddy, c.L. Diggs, L.H. Miller, S c i e n c e 1984,225,593-599. [13] K. Berzins, R.F. Anders, in 'Malaria: molecular and clinical aspects' Ed.: Mats Wahlgren, Peter Perlmann, Harwood Academic Publ, 1999, p. 181-216, [14] W.R. Ballou, 1. Rothbard, R.A. Wirtz, D.M. Gordon, J.S. Willimas, R.W. Gore, r. Schneider, M.R. Hollingdale, R.L. Beaudoin, W.L. Maloy, L.H. Miller, W.T. Hockmeyer, S c i e n c e 1985, 228, 996-999. [15] J.F. Young, W.T. Hockmeyer, M. Gross, [27] C. Bisang, C. Weber, J. Inglis, C.A. Schiffer, W.P. van Gunsteren, I. Jelesarov, H.R. Bosshard, 1.A. Robinson, J . A m . C h e m . Soc. 1995,117,7904-7915. [28] A.P. Nanzer, A.E. Torda, C. Bisang, C. Weber, J.A. Robinson, W.F. van Gunsteren, J . M o L . B i o l . 1997,267, 1012-1025. [29] C. Bisang, L. Jiang, E. Freund, F. Emery, C. Bauch, H. Matile, G. Pluschke, J.A. Robinson, J . A m . C h e r n . S o c . 1998, 1 2 0 , 7439-7449. [30] R. Moreno, L. Jiang, K. Moehle, R. Zurbriggen, R. GlUck, J.A. Robinson, G. Pluschke, unpublished. [31] I. Heinze-Krauss, P. Angehrn, R.L. Charnas, K. Gubernator, E.-M. Gutknecht, C. Hubschwerlen, M. Kania, C. Oefner, M.G.P. Page, S. Sogabe, J.-L. Specklin, F. Winkler, J . M e d . C h e m . 1998,41, 3961- 3971. [32] R. Gliick, V a c c in e 1999, 17, 1782-1787. [33] F. Ambrosch, G. Wiedermann, S. lonas, B. Althaus, B. Finkel, R. Gli.ick, C. Herzog, V a c c i n e 1997, 15,1209-1213. [34] R. Koradi, M. Billeter, K. Wiithrich, J . M oL . G raph. 1996,14,51-55.