3D Analysis of the TCR/pMHCII Complex Formation in Monkeys Vaccinated with the First Peptide Inducing Sterilizing Immunity against Human Malaria

T-cell receptor gene rearrangements were studied in Aotus monkeys developing high antibody titers and sterilizing immunity against the Plasmodium falciparum malaria parasite upon vaccination with the modified synthetic peptide 24112, which was identified in the Merozoite Surface Protein 2 (MSP-2) and is known to bind to HLA-DRβ1*0403 molecules with high capacity. Spectratyping analysis showed a preferential usage of Vβ12 and Vβ6 TCR gene families in 67% of HLA-DRβ1*0403-like genotyped monkeys. Docking of peptide 24112 into the HLA-DRβ1*0401–HA peptide–HA1.7TCR complex containing the VDJ rearrangements identified in fully protected monkeys showed a different structural signature compared to nonprotected monkeys. These striking results show the exquisite specificity of the TCR/pMHCII complex formation needed for inducing sterilizing immunity and provide important hints for a logical and rational methodology to develop multiepitopic, minimal subunit-based synthetic vaccines against infectious diseases, among them malaria.


Introduction
The appropriate fit of antigenic and immunogenic peptides inside the groove or peptide binding region (PBR) of class II major histocompatibility complex molecules (MHCII) is a crucial event for the formation of an appropriate T-cell receptor-(TCR)peptide-MHCII complex (TCR/pMHCII) and the subsequent activation of an antibody-mediated immune response [1]. Fine antigen recognition and binding specificity is conferred by the interaction of hypervariable amino acid sequences of a and b TCR chains, named Complementarity Determining Regions 1, 2 and 3 (CDRs), with structural features of the pMHCII complex, being most diversity concentrated in the b-chain CDR3 [2,3,4,5].
Malaria disease, in particular the one caused by the Plasmodium falciparum parasite, remains a serious public health problem worldwide, causing more than 500 million cases and killing 3 million of them per year [6]. To develop a fully effective antimalarial vaccine, so desperately needed, it is therefore essential to understand, at the deepest level, the formation of the TCR/ pMHCII complex capable of conferring sterilizing immunity against this deadly disease.
To develop a logical and rational methodology for designing minimal subunit-based, multiepitopic, multistage, chemically synthesized vaccines, capable of inducing sterilizing immunity against this threatening scourge and some others, we have iden-tified functionally-relevant, short (15-20-mer-long synthetic peptides or minimal subunits), conserved High Activity Binding Peptides (HABPs) derived from P. falciparum proteins involved in invasion to host cells as promising malaria vaccine targets [7].
However, conserved HABPs were found to be neither antigenic nor immunogenic or protection inducers when tested in Aotus monkeys, a non-human primate model highly susceptible to human malarias [8,9] and whose immune system molecules share a high degree of similarity with their human counterparts, specially with those involved in antigen presentation such as a/b TCRs [10,11] and MHCII-HLA-DRb1*-like molecules (88% to 100% similarity is reported for the Peptide binding region or PBR of these molecules) [12].
To solve this absence of antigenicity and immunogenicity, hundreds of trials were carried out with native and modified HABPs in large numbers of Aotus monkeys, finding that conserved HABPs could be rendered immunogenic and sterilizing immunity inducers by replacing their critical host cell binding residues by others having similar mass but opposite polarity. Specific replacement rules were defined [13] such that F must replace R and viceversa (F«R); W«Y; L«H; I«N; P«D; M«K; A«S; C«T or V; Q«E; and G has special physicochemical properties [13,14].
Bearing in mind these principles, we studied the Merozoite Surface Protein 2 (MSP-2), a 48-69 kDa glycosylphosphatidyli-nositol (GPI)-anchored merozoite surface molecule considered as promising antimalarial vaccine candidate due to its surface localization and immunological properties [15]. The screening of 20-mer-long peptides spanning the entire sequence of MSP-2 and overlapping with their neighbors by 5 residues led to the identification of the conserved N-terminal HABP 4044 ( 21 KNES-KYSNTFINNAYNMSIR 40 ), binding with high affinity to red blood cells (RBCs) [16] but same as reported for other conserved HABPs, 4044 was neither immunogenic nor induced protection against experimental challenge with the highly virulent Aotusadapted P. falciparum FVO strain when its polymerized form was used as immunogen. Therefore, HABP 4044 peptide analogues modified in critical RBC binding residues identified by glycine analogue screening (highlighted in bold types throughout the manuscript), were synthesized according to the rules mentioned above [17]. Immunization and challenge trials in monkeys lead to the identification of peptide 24112 ( 24 SKYSNTFNINAYNM-VIRRSM 43 ; modified residues are shown underlined) as the most promising subunit vaccine component since it induced high antibody titers that recognized a P. falciparum merozoite surface protein by IFA and a ,60 kDa molecule in P. falciparum schizont lysate, as assessed by Western blot analysis [17], and induced sterilizing immunity in 2 out of the 10 monkeys being immunized in a first trial and 1 of the 8 monkeys immunized in a second trial [18]. This protection was associated with high specific binding activity to purified HLA-DRb1*0401 molecules in vitro [18].
Being aware of the genetic control of the immune response, specially by MHCII molecules, and the exquisite specificity of the TCR-pMHCII complex formation in the induction of the appropriate immune response, modified HABP 24112 was inoculated in this study into in a new group of MHCII genotyped Aotus monkeys to analyze, at the structural level, the precise molecular mechanism of TCR/pMHCII complex formation resulting in induction of sterilizing immunity. For this purpose, spectratyping assays [19] were carried out to analyze the preferential usage of TCR b-chain variable region (Vb) CDR3 families in those Aotus monkeys developing high antibody titers and sterilizing immunity against the P. falciparum parasite upon vaccination with the modified synthetic peptide 24112, as well as in those that, despite developing lower antibody titers or not producing antibodies, were not protected.
The TCR-24112-MHCII complex associated with inducing sterilizing immunity against P. falciparum was studied by performing docking and energy minimization studies of this modified HABP with the previously reported X-ray crystallographic structure of the Hemagglutinin (HA) peptide, cocrystallized inside the HLA-DRb1*0401-HA peptide-TCR complex [20,21]. Care was taken to modify this reported structure according to the HLA-DRb1*04 and TCR Vb CDR3 amino acid sequence variations identified in protected and non-protected Aotus monkeys to determine spontaneous H bond formation, van der Waals (vdW) interactions, residue orientation and intermolecular distance differences between interacting atoms, to analyze this difference at the 3D structural level.
The results demonstrate, at the atomic level, that this first sterilizing-immunity inducing antimalarial peptide fits inside the specifically modified HLA-DRb1*0403-24112-HA1.7 TCR complex, displaying a completely different pattern from that of monkeys carrying a different HLA-DRb1*-like genotype that were not protected, thereby demonstrating that immunogenicity and induction of sterilizing immunity against transmissible diseases, like malaria, requires for specifically modified HABPs to fit appropriately inside the PBR of MHCII molecules so as to be specifically recognized by the TCR. Furthermore, the data here provided can be extrapolated to other molecules and pathogens as part of a logical and rational vaccine development methodology against transmissible diseases, malaria one of them.
Monkeys in Groups A, B and C developed two strikingly different immune responses after being vaccinated with peptide 24112 emulsified in Freund's adjuvant ( Figure 1). Indirect immunofluorescence assays (IFA) detected very high antibody titers ($1:1,280) in four out of the six (,67%) HLA-DRb1*0403like genotyped monkeys in Group A (Ao191, Ao250, Ao149 and Ao259) after the second and third immunizations, as indicated by the recognition of a merozoite membrane immunofluorescence pattern in late schizonts, which is in complete agreement with the surface localization of MSP-2 ( Figure 2A). Furthermore, Western blot analyses of P. falciparum-schizont lysates using the same hyperimmune sera showed strong reactivity with protein bands of 63, 54, 51 and 48 kDa, which are close to the molecular weight of MSP-2 (,69kDa) and its cleavage fragments. The remaining two HLA-DRb1*0403-like monkeys in Group A (Ao239 and Ao277) developed lower antibody titers (1:320) and showed a weaker Western blot recognition of MSP-2 and its cleavage fragments at the same serum dilution (1:200) ( Figure 2B).
While nonprotected control monkey developed very high parasitemias patent by day 5, reaching $5% by days 8-11 and requiring immediate treatment, the same four HLA-DRb1*0403like monkeys that developed high antibody titers ($1:1,280) were fully protected against experimental challenge. Protection being understood as the complete absence of blood parasites in the whole slide (screening under the fluorescence microscopy and with Acridine Orange staining of ,1,000,000 RBCs per monkey per day) during the 15 days that the experiment lasted (Figure 1 and 2C). This indicated that 24112 confers sterilizing immunity to two thirds (66.6%) of the HLA-DRb1*0403-like genotyped Aotus monkeys (Ao191, Ao250, Ao149 and Ao259). On the contrary, the other two monkeys (Ao239 and Ao277) producing lower antibody titers (1:320) developed high parasitemia levels on day 4-5 postchallenge that were comparable to the levels shown by monkeys in the control group ( Figure 2C).
Only three out of the five HLA-DRb1*0422-like monkeys in Group B (Ao142, Ao224 and Ao159) developed low antibody titers (1:160) after the third immunization, whereas no antibodies were detected in the other two Group B monkeys (Ao148 and Ao208). No anti-P. falciparum antibodies were detected in sera from Group C or Group D monkeys, as determined by IFA titers ( Figure 1) and Western blot assays nor were any of the monkeys in Groups B, C or D protected against experimental challenge with P. falciparum ( Figure 2C).

Docking of peptide 24112 into the MHCII of protected and non-protected monkeys
Based on the strong association between the HLA-DRb1*0403 genetic characteristic with 24112's sterile immunity induction ability, the 3D structure of this modified HABP determined by us [17] and the 3D structure of the HLA-DRb1*0401 molecule determined by Hennecke [20,21], docking studies of 24112 were performed on this MHCII molecule.
In our previous 1 H NMR studies, peptide 24112 displayed a distorted type III9 b-turn spanning from residue Y3 to T6 and a classical type III9 b-turn structure between residues A11 and M14 [17], whereas the rest of the molecule was unstructured. Similar to the HA peptide's crystallographic structure complexed with the HLA-DRb1*0401 molecule [20], 24112 peptide's residues fitting into Pocket 1 (Y12 fuchsia) and Pocket 9 (M20 green) were downwardly-orientated and deeply embedded into these pockets ( Figure 3A), whereas residues fitting into Pocket 4 (V15 dark blue) and Pocket 6 (R17 brown) were more shallowly embedded than the above-mentioned residues [22,23,24].
Several other probable binding registers of 24112 were docked on HLA-DRb1*0401 but none of them completely displayed the characteristic binding motives of this alleles [25] and none allowed a perfect fit inside the PBR of this MHCII molecule, therefore suggesting that the only probable structure was YNMVIRRSM (underlined residues fit into pockets 1, 4, 6 and 9, respectively).
By the same token, docking of 24112 into the modified HLA-DRb1*0422 structure of Ao148 (non-protected, non-antibody producer monkey) showed the spontaneous formation of SIX of the 10 canonical H bonds between atoms of the peptide's backbone and MHCII lateral chain residues ( Figure 3E,F). Binding of Y12, N13 and M20 involved the same residues mediating binding of 24112 to HLA-DRb1*0403 in Ao191, but displayed a totally different H-bonding pattern in the 24112-HLA-DRb1*0422 complex between V15 with Pocket 4 Glna9 (in dark blue, 2.27 Å ), and R17 with Pocket 6 Lysb71 (in brown, 1.71 Å ). It can be clearly seen that no H bonds are established between 24112's R18 backbone and lateral chains of HLA-DRb1*0422 At this point, it is worth to remember that H bonds are not determined by their interactions distances but also by their torsion angles.
Additional differences were observed between these two pMHCII complexes (highlighted in pale gray in Figure 3C,F). In the HLA-DRb1*0403 molecule, vdW interactions were preferentially established between lateral chain atoms of N13 with Thrb77, R17 with Glua21, Glua11 and Aspa66 (the latter two confer a negatively charge character to pocket 6 that strongly stabilizes binding of positively charged residues like R, K or H [23]), a polarity further accentuated by the Aspb70Gln difference in this allelic variant. H bonds are also established between lateral chain atoms of R18 with Aspb66 and Aspb70. Meanwhile in HLA-DRb1*0422 peptide 24112, vdW interactions ( Figure 3F, pale gray) are specifically established between lateral chain atoms of N13 with Asnb77, R17 with Glua11, R18 with Trpb61 and Tyrb67, and S19 with Tyrb67; displaying a strikingly different H bonding and vdW interaction pattern in Pocket 6, that results in a different orientation of R18 ( Figure 3D). These H-bonding differences approximate R18 to Trpb61 in the HLA-DRb1*0403-like-24112 complex preventing an appropriate H bonding between backbone atoms of 24112 with Asna69 and Tyrb30 ( Figure 3A) residues of the HLA-DRb1*0422 molecule, as situation clearly visible in Figure 3D, were these two H bonds that are critical for the anchoring and stabilization of the peptide inside the PBR are not observed in Figure 3E nor measured in Figure 3F. It is well know that H bonds give high stability and specificity to the pMHCII complex formation due to the exquisite energetic dependence on the stereochemistry and geometry of the bond.

Preferential usage of Vb families in immunized monkeys
The spectratyping analysis of the TCR repertoire provides a global picture based on a reduced number of cells, and helps determining the complexity and stability of a T cell repertoire in response to an antigenic stimulus, but it only provides a qualitative description of the T-cell mediated immune response raised against a particular antigen. Nevertheless, the methodology here reported is useful, specific and sensitive enough for detecting the different TCR CDR3b sequences that expand in response to a specific antigen.
Based on the Aotus Vb sequences reported to date [10,11], the spectratyping analysis of the CDR3 amplicons showed a polyclonal Gaussian-like distribution for the majority of TCR Vb families in the pre-immune (P0) samples ( Figure S1a) typical of unstimulated T cells. The CDR3 size distribution pattern ( Figure S1b) changed strikingly to skewed oligoclonal expansions of one, two or three dominant Vb families in Ao191, Ao259, Ao149, Ao250, Ao239, Ao277, Ao142, Ao224 and Ao148 upon immunization with peptide 24112; however, such skewed patterns were only associated with high antibody production in Ao191, Ao259, Ao149, Ao250, Ao239 and Ao277. CDR3 Vb distribution of control monkeys receiving only Freund's adjuvant (Group D) showed Gaussian-like patterns similar to the ones found in P0 samples.
Post-immunization lymphocyte samples of the fully-protected HLA-DRb1*0403-like genotyped monkeys (Group A) developing high antibody titers after the second (II 20 ) and third (III 20 ) doses showed a selectively higher usage of the TCR families Vb12 (Ao191, Ao149 and Ao250) and Vb6 (Ao259) in T-cell clones expanding in response to immunization with 24112 (Table S1). Striking differences were found in the TCR family usage of the non-protected monkeys Ao239 and Ao277 developing lower antibody titers against P. falciparum, which preferentially expressed Vb5 and Vb19 TCR families ( Table 1).
HLA-DRb1*0422-like non-protected monkeys (Group B) developing low or no detectable antibody titers ( Figure 1) showed a preferential usage of Vb9 and Vb5 in Ao142; Vb7, Vb10 and Vb28 in Ao224; and Vb7, Vb15 and V28 in Ao148 (Table S1), but none of them displayed a preferential usage of Vb12 and Vb6 families, therefore suggesting an exclusion mechanism at the TCR level between protected and nonprotected monkeys in the TCR Vb usage, in spite that some of them can produce antibodies (although at a lower level) when being immunized with the same epitope. Therefore these later antibodies must have different , R18 (dark gray in P7), S19 (yellow in P8) and M20 (green in P9). Top view panels B and E display the H bonds (shown as doted lines) established between backbone atoms of peptide 24112 (represented as sticks) and side-chain atoms of residues from the MHCII a and b chains (depicted as pink and blue ribbons, respectively) in protected (group B monkeys) as well as non-protected (group E monkeys). The nitrogen and oxygen atoms are shown as blue and red balls, respectively. Black segments in the b-chain correspond to the residues that were modified according to the MHCII sequence of Ao191 (HLA-DRb1*0403) and Ao148 (HLA-DRb1*0422). (C and F) H bonds and vdW interactions, measured in Amstrongs (Å ), between 24112 with HLA-DRb1*0403 and HLA-DRb1*0422 molecules. Interactions involving different atoms are highlighted in pale gray, while interactions involving common residues are not shadowed. The color code for those residues establishing such H bonds is the same used in Figure 1. doi:10.1371/journal.pone.0009771.g003 affinities and/or structural characteristics, a subject currently under study at our institute.

Sequencing of TCR-Vb CDR3 segments
The Vb CDR3 sequences of expanded T-cell clones indicated a random usage of J segments in each of the expanded Vb families. For example, Vb12 CDR3 expanded sequences of the antibodyproducer protected Ao191 and Ao149 monkeys included almost all possible J segments (J1.1, J1.2, J1.4, J2.1, J2.2 and J2.3; Table  S1). This shows that different T lymphocyte subpopulations with similar CDR3b lengths are involved in the sterilizing immune response induced by peptide 24112 and suggests a clear polyclonal response to a molecularly defined antigen in individuals with specific and determined genetic backgrounds, clearly visible in the reactivity of these sera in Western blot analyses ( Figure 2B).
The amino acid sequence variations found in the Vb12 D region of the fully protected Ao191 ( Table 1; clone 3) as well as the differences detected in the Vb15 D region of the nonprotected, non-antibody producer Ao148 ( Table 1; clone 5) were used to modify the Vb3S1 sequence (belonging to the same Vb12 family) of the HLA-DRb1*0401-HA-HA1.7 TCR crystal structure [20,21], as described in detail below.

Structural analysis of the MHCII-24112-TCR complex
The previously reported HLA-DRb1*0401-HA-HA1.7 TCR crystal structure [20,21] was modified to determine the HLA-DRb1*0403-like-24112-TCR structure of the fully protected Ao191 and the HLA-DRb1*0422-like-24112-TCR structure of the non-protected Ao148. These molecules were analyzed and compared, at the structural level, to examine H bond formation and distance differences. For the HLA-DRb1*0403-like-24112-TCR complex, the TCR Vb3S1 sequence was modified in the CDR3b region according to the variations found in the Vb12 D region of clone 3 from protected Ao191 as follows (TCR residues are written in three-letter code throughout this manuscript): Phe3b96Ser, Leu3b97Thr, Glu3b98Gly, Gly3b99Leu, Gly3b100Pro and As-p3b105Gly in the J region ( Table 1 and Table S1), which are numbered according to Hennecke's system [21]. For the HLA-DRb1*0422-like-24112-TCR complex, the TCR Vb3S1 was replaced by the Vb15 family sequence found in clone 5 from the non-protected, non-antibody producer Ao148 as follows: Arg3b96Ser, Asp3b97Thr; Glu3b98Gly; Glu3b99Leu; As-p3b100Pro and Asp3b105Gly in the J region.
Based on the above mentioned information and the fact that TCR Va displays limited polymorphism (therefore their CDRs were not cloned in this study), the 3D structure of the HLA-DRb1*0401-HA-HA1.7 TCR complex, used as template, showed in docking analyses that binding of such modified HA1.7 TCR to both HLA-DRb1*04-24112 complexes is mediated by the interaction of the TCR Va region with the peptide's N-terminal portion, while the C-terminal fragment established contact with Vb regions ( Figure 4A,D), as previously described elsewhere [20,21,26].
In the analysis of the TCR contacting residues of HLA-DRb1*0422-24112 complexed with the CDR3b Vb15 region of the non-protected Ao148 (Figure 4 D,E,F) only ONE H bond ( Figure 4F, dark gray indicated by an asterisk) is spontaneously formed between S19 and Glub99 (1.52 Å ), while TWO vdW interactions are observed between N13 with Aspa28 (3.91 Å ) from CDR1a (in white ribbon), and R18 with Glub98 (3.61 Å ) from CDR3b (in red ribbon), perhaps due to the horizontal localization of this peptide's residue R18 ( Figure 4E). This complex displays therefore fewer and weaker electrostatic interactions with the TCR. TCR footprint on the HLA-DRb1*04 complexed molecules Figure 5 shows the H bonds established between the HLA-DRb1*0403-modified molecule and the Vb12-clone-3-modified TCR (left) of the antibody-producer, protected Ao191 with their corresponding interatomic distances ( Figure 5, A,B and C left panels), as well as the 3D structure and interatomic distances between the HLA-DRb1*0422-modified molecule and the Vb15clone-5-derived TCR of the non-antibody producer, non-protected Ao148 ( Figure 5D, E, and F right panels).
The distances and electrostatic properties also show that only five (5) salt bridges ( Figure 5C,F; dark gray) are established between the residues conforming the non-protective HLA-DRb1*0422-Vb15 TCR complex, whereas in the protectionassociated HLA-DRb1*0403-Vb12 TCR complex, eight (8) salt bridges, are determined suggesting a stronger and more stable interaction of this pMHCII complex; a condition needed to stabilize the complex and properly activate the immune system toward a sterilizing immunity response.
It has been clearly shown that the strongest interactions are established between lateral chain atoms of MHCII molecules and atoms from the peptide's backbone (Figure 3), whereas in the peptide/TCR interaction the strongest bonds are established with the peptide's lateral chain atoms (Figure 4), among which salt bridges are the most important electrostatic forces.
In essence, this study clearly shows that important differences exist between residues' orientation and electrostatic forces established in these TCR/pMHCII complexes, which are associated with two different immunological outcomes induced by the same peptide in slightly different variants of the same HLA-DRb1*04 allele.

Discussion
The structural data obtained in this study at the 3-dimensional level provides strong evidence that the TCR/pMHCII complex formed with the specifically modified HABP 24112 derived from the P. falciparum MSP-2 protein, inductor of high anti-parasite antibody titers and sterile immunity in 67% of HLA-DRb1*0403-like genotyped Aotus monkeys (Figure 1), has a strikingly different 3D structure conformation from the one observed in non-protected non-antibody producer HLA-DRb1*0422-like genotyped monkeys immunized with the same peptide (Figures 3-5). This is clearly demonstrated by a large set of differences in the salt bridges, H bonds and vdW interactions, interatomic distances and contacting residues' orientation between both complexes.
In this study, we show that Aotus monkeys developing high antibody titers and being protected against experimental challenge with the lethal P. falciparum malaria FVO strain carry the HLA-DRb1*0403-like genetic marker and have a preferential usage of Vb12 and Vb6 TCR families ( Table 1), whereas the same molecule was poorly or not immunogenic nor protection-inducing in Aotus monkeys carrying a slightly different HLA-DRb1*04 allelic variant like HLA-DRb1*0422 and that such weaker and/or nonprotective immune response was associated with the preferential usage of TCR Vb 9, 5, 7, 10, 15 and 28 families; therefore confirming the exquisite specificity of the TCR/pMHCII formation in sterilizing immunity induction against malaria.
While Group C monkeys immunized with 24112 and carrying some other non HLA-DRb1*04-like alleles did not develop any detectable antibody titers nor were protected against experimental challenge, same as monkeys in control Group D (inoculated with Freund's adjuvant only), these data show that peptide 24112 is neither recognized nor presented by these MHCII alleles, a fact corroborated by the inability of 24112 to bind experimentally to other purified HLA-DRb1 * molecules (Group C) [17,18]. Therefore the immune response observed in Group A HLA-DRb1*0403 monkeys is HLA-DRb1*0403-specific and not the result of an unspecific stimulation induced by the Freund's adjuvant, as show by Group D results (here only one HLA-DRb1*0422 monkey was included due to the low frequency of this allelic variant).
Docking studies clearly demonstrate that the TCR/pMHCII complex has to be properly assembled in order to induce sterilizing immunity against this deadly disease, expanding and confirming the elegant structural studies on antigenicity with the Influenza virus Hemagglutinin A (HA) peptide [22], immunogenicity with several immunodominant epitopes from hen egg lisozyme (HEL) [32] and Conalbumin (CA) [33], autoimmunity induced by the Col II peptide in rheumatoid arthritis [24] and many more. However it should be highlighted that the present study is hitherto the first immunogenicity and sterilizing immunity study on against any microbe thoroughly analyzed up to the structural level.
Bearing in mind that the 24112 modified peptide induced high immunogenicity and sterilizing immunity in 67% of HLA-DRb1*0403-like genotyped monkeys, these data clearly suggest that additional modified HABPs should be included in a vaccine formulation in order to activate more TCR/pMHCII complexes and to develop complete protection against malaria in HLA-DRb1*0403 individuals. Furthermore, a substantially larger number of modified HABPs has to be included in a fully effective, multiepitopic, multistage, minimal subunit-based, chemically synthesized antimalarial vaccine capable of conferring protection to ALL different HLA-DRb1* alleles, and variants, even more considering that the parasite employs multiple molecules and invasion mechanisms [34], the majority of which display a large number of exquisite genetic variations to evade the host's immune system pressure that are able to distract the complete immune system b y displaying just one amino acid variation [35]. Therefore all parasite pathways and strategies for invading RBCs and liver cells (or at least the most relevant ones), must be destroyed or at least blocked to prevent parasite's entry into target cells. This notion is further supported by the recent transcriptome analyses demonstrating that ,58-90 of all P. falciparum proteins are involved directly in RBC invasion [36] and probably a similar number are involved in sporozoite invasion to the liver cells [37].
In our docked structures of the protection-associated HLA-DRb1*0403-24112-TCR complex, the most variable regions of the TCR (CDR3a (yellow) and CDR3 b (red)) are located in the central portion of the binding interface with 24112, while the most conserved CDR1 and CDR2 regions make contact with the upper surface of the MHCII helices that are surrounding CDR3 like a gasket [30,31]. Although residues M14 in P3 (pale blue) and I16 in P5 (pink) are not establishing H bonds or vdW interactions with the TCR, they fit perfectly well inside the hydrophobic grooves.
Meanwhile, the CDR3a does not make contact with any 24112 contacting residues (e.g. P3 (M14), P5 (I16), P7 (R18)) in the 24112-HLA-DRb1*0422 complex, and fewer H bonds and vdW forces are established in this complex, which suggests little probability of TCR inspection and weaker interaction of this pMHCII complex with the TCR in this non-protective complex.
The striking variations found between the Vb15 clone-5 TCR-24112-HLA-DRb1*0422 and the Vb12-clone-3 TCR-24112-HLA-DRb1*0403 complexes in the CDRs usage of evolutionary conserved residue contacts [27,28,29,30,31], specifically regarding salt bridge formation (5 in the former complex and 8 in the later complex) as well as in the orientation of contacting residues in the TCR-MHCII footprint, suggests a different recognition pattern and a different docking footprint on the TCR/pMHCII interaction between these 2 complexes associated with two different immune responses: non-antibody production and non-protection versus antibody production and protection induction, respectively.
The conservation of the critical residues controlling the ''TCR/ pMHCII interaction'' suggests that the vast peptide's recognition specificity displayed in sterilizing immunity against modified peptides is almost entirely driven by the appropriate configuration of the HLA-DRb1*0403-24112 peptide complex, which displays a completely different array of H bonds and vdW networks compared to the non-immunogenic, non-protective HLA-DRb1*0422-24112 peptide complex structure [27,28,29,30,31].
Studies by Kersh et al. [38] on the binding of Hemoglobin (Hb 64-76) peptide to I-E K molecules have elegantly shown that subtle modifications such as replacing the D residue in the E73D sequence of the Hb peptide fitting inside Pocket 6 modifies the orientations and interatomic distances of P5, P7 and P8 with their TCR contacts, which is in turn reflected in a 1,000-fold reduction in the potency of the altered peptide 24112 to induce antibody production. Our situation is different since the same peptide is anchored to both TCR/pMHCII complexes but the slight differences within the genetic background of the Aotus HLA-DRb1*04-like molecules lead to different H bonding and vdW patterns, specially between the peptide's backbone atoms and lateral chains residues of the pMHCII molecule. Furthermore, conserved residues controlling the ''TCR/pMHCII interaction'' display a different hydrogen bonding and salt bridge formation pattern.
Since in previous studies in which we had characterized the Aotus MHC-DRb exon 2 [12,39] in a large number of monkeys show that the amino acids that define the PBR pockets share a mean similarity of 89-94% with HLA-DRb1*0403 (with a minimum similarity of 82-88% and a maximum of 94-100%), a mean similarity of 91% with HLA-DRb1*0422 (a minimum similarity of 88% and maximum of 94%), and quite similar similarity values with the other alleles analyzed in this study. Furthermore since most substitutions are functionally and evolutionary conserved [40,41], we can conclude that the results observed in the Aotus monkey model mimic in a very high degree the TCR-pMHCII complex formation in humans, thus highlighting the extreme importance of this nonhuman primate for human vaccine development.
Given that Glna9, Asna62, Asna69, Trpb61 and Asnb82 residues establishing H bonds with the peptide's backbone are conserved among humans and Aotus (unpublished results), these data suggest that the modified HABP 24112 properly fitting into HLA-DRb1*0403 molecules could be used almost immediately for human vaccination in individuals carrying this allelic variant and highlights this peptide as the first component of a sterilizing-immunity-inducing, multiepitopic, multistage, minimal subunit-based, chemically synthesized antimalarial vaccine for human use.
Furthermore, the data clearly show that developing fully effective vaccines is a far more elaborated and complex process than merely vaccinating humans with recombinant fragments (either individually or as mixtures), viral vector, DNA-based or unmodified short or long synthetic peptides [42], which have lead to «disappointing» and frustrating results [42,43,44], some of them with deleterious effects. The evidence also indicates that the molecular characteristics of the microbe and the host play a fundamental role in the conformation of the appropriate TCR/ pMHCII complex to induce sterilizing immunity.
Finally, it is worth noting that all the evidence shown here was obtained by using the most stringent system to asses sterilizing immunity (since a 100% infective dose of an Aotus-adapted P. falciparum strain was intravenously inoculated into monkeys), which allow us to conclude that multiantigenic, multistage, minimal subunit-based, specifically-modified, chemically-synthesized vaccines against any transmissible disease like malaria, must be appropriately designed and modified to fit perfectly well insidethe TCR/pMHCII complex so as to induce sterilizing immune responses.

Materials and Methods
Peptides 24112 monomer and polymers were synthesized by using t-boc chemistry [45]. Polymerization was allowed by adding CG residues to the N and C terminal ends of peptide 24112 ( 24 SKYSNTFNINAYNMVIRRSM 43 ), with a carefully standardized oxidation procedure that guarantees the formation of high molecular weight polymers (8-24kDa) for immunization purposes.

Aotus monkeys
Forty wild-caught Aotus monkeys were kept in stainless-steel cages at FIDIC's primate station in Leticia, Amazonas, Colombia, and maintained in strict accordance with the NIH guidelines for animal care and the Colombian Ministry of Health (Law 84/ 1989), under the weekly supervision of CORPOAMAZONIA officials and a primatologist. Monkey sera (1:20 dilution) were screened by IFA to determine previous exposure to Plasmodium parasites. Monkeys testing positive were returned to the jungle without further manipulation [8]. Monkeys' parasitemias were assessed daily by reading under fluorescence microscope by screening 1.000.000 Acridine Orange stained RBCs and they were immediately treated whenever P. falciparum infected RBCs were $5%, or before if the monkey's health condition had deteriorated. Treatment consisted of orally administered pediatric doses of Chloroquine (10 mg/kg on the first day and 7.5 mg/kg per day until day five). Once assuring total clearance of parasites from blood and excellent health condition, monkeys were released back into their natural habitat close to the site where they had been captured with the supervision of CORPOAMAZONIA officials. All procedures were approved and supervised by FIDIC's Ethics Committee in Health Research (Resolution No. 008430 of 1993, Colombian Ministry of Health) and by FIDIC's Primate Station Ethics Committee.

MHCII-DRB genotyping
Genomic DNA of each monkey was isolated from peripheral blood lymphocytes to amplify the MHCII-DRb exon 2 segment using high-fidelity PCR (ACCUZYME DNA Polymerase), as described elsewhere [12]. Sequences were obtained by cloning amplicons into the pCRHBlunt vector (Invitrogen TM ) and allele types were assigned by comparing them to previously reported Aotus alleles [12].

Immunization and challenge
Based on their similarity to human HLA-DRb1* alleles determined by genotyping [12], eighteen monkeys were selected for this study and classified into different immunization groups ( Figure 1). Groups A (6 monkeys), B (5 monkeys) and C (3 monkeys) were immunized on days 0, 20 and 40 with 125 mg of peptide 24112 emulsified in Freund's complete adjuvant for the first dose, and in Freund's incomplete adjuvant for the second and third doses. Group D (4 monkeys) received only Freund's adjuvant on the same days. One milliliter of peripheral blood was collected on day 0 (pre-immune or P0) and 20 days after the second and third immunizations (II 20 and III 20 , respectively) to obtain lymphocytes and sera for immunological studies. On day 60, all monkeys were challenged by intravenous inoculation of a 100% infective dose (100,000 infected erythrocytes) of the P. falciparum FVO Aotus-adapted strain, freshly obtained from a previously infected monkey. Blood parasitemia levels were monitored daily for 15 days by Acridine Orange staining. Sterilizing immunity was defined as the complete absence of parasites in the blood of protected monkeys during the 15 days that the experiment lasted, as assessed by reading the whole slide (,1.000.000 RBCs were screened). Controls and non-protected monkeys developed patent parasitemias by day 5 and high parasitemias ($5%) by days 8-11 after the challenge, therefore requiring immediate treatment ( Figure 1C). Monkeys were kept in quarantine and released back into the jungle in company of CORPOAMAZONIA officials.

IFA antibody titers
Air-dried slides containing P. falciparum late-schizonts (FCB-2 strain) obtained from a synchronized continuous culture were used for determining antibody titers, using twofold serial dilutions of monkey sera (initial dilution: 1:40). Slides were washed 6 times with PBS, incubated with the appropriate dilution of affinity purified FITC-labeled goat anti-Aotus F(ab')2 IgG fragment, washed and read by fluorescence microscopy.

Western blot analysis
Briefly, whole schizonts' lysate of the same FCB-2 synchronous culture was separated in a discontinuous PAGE system, electrotransfered to nitrocellulose membranes and incubated with appropriate monkeys' sera dilutions and with goat anti-Aotus IgG F(ab')2 fragment conjugated to Alkaline phosphatase, to asses immunoreactivity.
No attempts were made to determine cellular immune responses due to the limited amount of drowned blood and the priority given to the T cell cloning and spectratyping analyses.

Characterization of the TCR repertoire
The cDNA was synthesized with the SuperScript TM III kit (Invitrogen, CA USA) using RNA isolated from peripheral blood lymphocytes as template. Nineteen specific forward primers and a common primer annealing in the b-chain constant (Cb) region were designed to amplify all Aotus Vb families reported to date [11]. Additionally, a set of reverse and forward primers amplifying a 419-bp fragment of the Aotus TCR a-chain constant (Ca) segment was used as amplification control ( Table  S2).
The Ca segments were coamplified together with each of the 19 CDR3 Vb segments using 0.25 mM of Ca forward and reverse primers, and 0.75 mM of Vb and Cb primers. Amplification products were subjected to 6 run-off reaction cycles using Ca and Cb reverse primers labeled with 6-Carboxyfluorescein (6-FAM) on the 59 end. Concentrations were: 0.5 mM of fluorolabeled reverse primers and 1 mL of each Vb coamplification product.

Spectratypes and CDR3 TCR Vb segments of immunized Aotus monkeys
The bp-length and signal intensity of fluorolabeled amplicons were measured in an ABI PRISM 310 Genetic Analyzer using Rox 500 molecular weight markers (both from Applied Biosystems, CA, USA). Spectratypes were analyzed using GeneScanH Analysis Software, assuming that each peak corresponded to a particular TCR CDR3 rearrangement. Spectratype patterns (Gaussian or skewed) were visually evaluated and T-cell oligoclonal expansions were calculated in relative fluorescence intensity units as described elsewhere [46]. Clonally expanded CDR3 segments were amplified by high-fidelity PCR as described above, sequenced and analyzed using Clustal W software.

Molecular modeling
3D structure models of the Aotus ab TCR-24112-HLA-DRb1*0403 and Aotus ab TCR-24112-HLA-DRb1*0422 molecular complexes was generated by using the crystallographic structure of the human ab TCR HA1.7 HA peptide-HLA-DR4 (DRA*0101 and DRb1*0401) as molecular template (PDB code: 1J8H) [20]. Replacements were made on these molecules based on the differences found in the PBR of the high-antibody producer, fully-protected Ao191 monkey on its HLA-DRb1*0403-like bchain amino acid sequence (Phe37Tyr, Ser57Asp, Lys60Tyr, Leu61Trp, Ile67Leu, Asp70Gln, Ser74Glu and Gly86Val) as well as in the non-antibody producer, non-protected Ao148 monkey genotyped as HLA-DRb1*0422-like (Tyr26Phe, Leu67Tyr). Another structure was generated by replacing the Vb12 family of the HA1.7 TCR CDR3 D region for the 96 Phe-Leu-Glu-Gly-Gly 100 sequence found in the D region amino acid sequence of Vb12 clone 3 derived from the protected Ao191 monkey; Gly105 of the J region was also replaced by Asp. By the same token, the HA 1.7 TCR CDR3 96 Arg-Asp-Glu-Glu-Asp 100 D region was replaced by the 96 Ser-Thr-Gly-Leu-Pro 100 sequence found in the Vb15-clone 5 sequence of the non-antibody producer and nonprotected Ao148 monkey.
A conjugate gradient algorithm was applied to minimize energies and build a more energetically favorable model of the complex's position. To obtain the most appropriate model, 6 to 8 simulations with 10,000 iterations were performed for each structure, with each of the Vb sequences in the complete template (PDB code: 1J8H). Insight II (2000) Biopolymer module software (Accelrys Software Inc., USA), run on an Indigo 2 Station (Silicon Graphics), was used for superimposing the backbones of the original template and the obtained model without further refinements. The rootmean-square deviation between two molecule conformations was then determined. Figure S1 Spectratype analysis of the TCR Vb repertoire. (A) Representative Gaussian-like profile shown by most TCR Vb families of Aotus 191 before being immunized with peptide 24112. Only Vb families 2, 19, 24 and 29 showed a skewed profile in the pre-immune sera. (B) Comparison between TCR Vb repertoire before (P0 above) and after (PIII below) immunization with peptide 24112. The most notably expanded families were: Vb12 (in Ao191); Vb6 (Ao259); Vb12 (Ao149); Vb5 (Ao239); Vb19 (Ao277); Vb9 (Ao142); Vb7 (Ao224) and Vb15 (Ao148), which displayed a Gaussian-like distribution pattern in P0 samples and a skewed pattern in PIII samples from the same monkeys.