112

Bioconjugate Chem. 2004, 15, 112−120

A Comparative Study of Different Presentation Strategies for an HIV Peptide Immunogen Luis J. Cruz,*,† Enrique Iglesias,‡ Julio C. Aguilar,‡ Luis J. Gonza´lez,‡ Osvaldo Reyes,‡ Fernando Albericio,†,§ and David Andreu| Barcelona Institute of Biomedical Research, Barcelona Science Park, 08028 Barcelona, Spain, Centro de Ingenierı´a Gene´tica y Biotecnologı´a, Cubanaca´n, Habana 10600, Cuba, Department of Organic Chemistry, University of Barcelona, 08028 Barcelona, Spain, and Department of Experimental and Health Sciences, Pompeu Fabra University, 08003 Barcelona, Spain. Received July 10, 2003; Revised Manuscript Received October 17, 2003

Different strategies have been used to increase the immunogenicity of an antigenic HIV peptide as a vaccine candidate. The selected B-cell epitope comprises 15 amino acids (317-331) of the V3 region of HIV-1, JY1 isolate (subtype D), in tandem with a T-helper epitope corresponding to the 830-844 region of tetanus toxoid. Several presentations, including oligomerization, multiple antigenic peptide dendrimers, and conjugation to dextran beads or to other macromolecular carriers, have been synthesized and evaluated. Murine sera from the different presentations of the V3 epitope have been compared with regard to antibody titers and cross-reactivity with heterologous HIV subtypes. The dendrimer version of the peptide conjugated to HBsAg protein was a better immunogen than the dendrimer alone and showed a higher immunogenicity than other multimeric presentations or than the peptide alone conjugated to dextran. The dendrimer version, either alone or conjugated to HBSAg, enhanced cross-reactivity toward heterologous V3 sequences relative to monomeric peptide. In addition, fine epitope mapping of the entire JY1 sequence by sera from the different immunization groups was performed by the spot synthesis technique. Results showed that the amino acids involved in molecular recognition were LXQXXY or LXQXLY, with particularly strong recognition of the C-terminal region LGQALY. However, cross-reactivity toward the heterologous sequences did not completely correlate with recognition of particular amino acids in the primary sequences. These results can find application in the development of HIV vaccine candidates.

INTRODUCTION

Although vaccines are widely regarded as the ideal solution for the control of infectious diseases, only a handful of them are currently available and a pressing demand is felt for improvements on existing ones, with particular emphasis on higher efficiency and undisputed safety profiles. In recent years, considerable attention has been focused on the development of peptide-based vaccines as an alternative to conventional formulations. Peptide vaccines have obvious advantages such as high safety, low cost, and easy handling and storage. In contrast to classical vaccines, they can be tailor-made against specific targets including antigenic sites not easily accessible to the immune system. However, they suffer also from limitations such as the usually low immunogenicity of small peptides and the vulnerability to proteolytic degradation. Many of these problems can be addressed by improved forms of presentation of the synthetic epitope to the immune system, including formulations such as multiple antigen peptide systems (MAPs1) (1), dendrimers (2), lipopeptides (3, 4), polymerization (5, 6), conjugation to carriers such as proteins * Corresponding author. David Andreu, Department of Experimental and Health Sciences, Pompeu Fabra University, Doctor Aiguader 80, 08003 Barcelona, Spain. david.andreu@ upf.edu. † Barcelona Institute of Biomedical Research. ‡ Centro de Ingenierı´a Gene ´ tica y Biotecnologı´a. § University of Barcelona. | Pompeu Fabra University.

or synthetic polymers (7, 8) or association to particulate systems such as liposomes (9), or immunostimulating complexes such as ISCOMs (10). Combined approaches such as coupling of MAPs to carrier proteins (11) have also been shown to enhance peptide immunogenicity and to induce protective or neutralizing antibody responses for parasitic and viral pathogens. In HIV-1 infection, the V3 loop is the primary target for neutralizing antibodies (12, 13) and is involved in other aspects of infectivity. Thus, sequence changes in V3 can affect chemokine receptor usage and therefore modulate which cell types are infected (14). This V3 region includes both T-helper and CTL epitopes (15). Antibodies against the principal neutralizing domain (PND) of V3 appear to exert some control over the viral load (16-18). For the HIV-1 MN (subtype B) isolate, a 1 Abbreviations: AAA, amino acid analysis; AcOH, acetic acid; Boc, tert-butyloxycarbonyl; Bzl, benzyl; BSA, bovine serum albumin; CFA, complete Freund’s adjuvant; DIPC, 1,3-diisopropylcarbodiimide; DTT, dithiothreitol; DCM, dichloromethane; DMF, N,N-dimethylformamide; DIPEA, diisopropylethylamine; EDTA, ethylenediaminetetraacetic acid; EDAC, 1-ethyl-3-(dimethylaminopropyl)carbodiimide; Fmoc, 9-fluorenylmethyloxycarbonyl; HBsAg, hepatitis B surface antigen; HF, hydrogen fluoride; HOBt, 1-hydroxybenzotriazole; HPLC, high-performance liquid chromatography; IFA, incomplete Freund’s adjuvant; MAP, multiple antigenic peptides; MALDI-TOF, matrixassisted laser desorption ionization with time-of-flight analysis; MS, mass spectrometry; PBS, phosphate-buffered saline; RP, reverse phase; TFA, trifluroacetic acid; TBTU, 2-(1H-benzotriazol-1-yl)-1,1,3,3-tetramethyluronuim tetrafluoroborate.

10.1021/bc034119j CCC: $27.50 © 2004 American Chemical Society Published on Web 12/11/2003

HIV Peptide Immunogen

Bioconjugate Chem., Vol. 15, No. 1, 2004 113

Figure 2. HPLC analysis of the polymerization reaction by gel filtration on TSK G 2000SW in 10% acetonitrile in water. Arrows a-d indicate, respectively, the elution volumes of bovine serum albumin (Mr 67 000), chymotrypsinogen A (Mr 25 000), lysozyme (Mr 14 000), and the monomer peptide (Mr 4000).

Figure 1. Reaction scheme for the polymerization of a peptide containing the JY1 epitope (filled box) in tandem with a T-epitope of tetanus toxoid (empty box) separated by a cathepsin cleavage site insert. Deprotection of both N- and C-terminal Cys residues gives a dithiol that polymerizes under basic conditions.

high titer against V3 has been correlated to slow progression rates in seropositive hemophilic Japanese patients (19). All these findings highlight the immunological relevance of the V3 loop, which has been included in a number of HIV-1 vaccine candidates (20). They also underscore the significance of studying how different ways of presentation of the synthetic peptide epitope affect the quality of the immune response. In the present work we have investigated several presentation strategies for the V3 epitope, with a view to enhance its immunogenicity. Our selected B-cell epitope comprises 15 residues (317-331) of the V3 region from HIV-1 JY1 isolate (21), linked through a Lys-Lys cathepsin-like cleavage site (6) to a helper T-cell epitope (residues 830-844) from tetanus toxoid (22). We have evaluated the anti-peptide antibody response of these presentations and their cross-reactivity toward other V3 sequences and have fine-mapped their recognition of the entire JY1 epitope sequence. EXPERIMENTAL PROCEDURES

Chemicals and Biologicals. Fmoc- and Boc- protected amino acids were purchased from Bachem (Switzerland). Solvents for peptide synthesis (dichloromethane, 2-propanol, N,N′-dimethylformamide) were obtained from Merck (Germany). Dextran beads cross-linked with epichlorhydrin (Sephadex G 150) were from Pharmacia Biotech (Sweden). Succinic anhydride and other chemicals were from Sigma (St. Louis, MO). Recombinant hepatitis B virus surface antigen (rHBsAg) and TAB9 protein, a multiepitope polypeptide (MEP) including the V3 region from six divergent HIV-1 isolates (LR150, JY1, RF, MN, BRVA, and IIIB, in this order) fused to the amino terminus (47 amino acids) of P64K protein from N. meningitidis, were obtained from CIGB (Cuba). Synthetic Peptide Antigens and Conjugates. Peptide Synthesis. The monomeric form of the immunogen [C(Acm)GRQSTPIGLGQALYTTKKQYIKANSKFIGITELGC(Acm)] (Figure 1) was synthesized by the Boc/Bzl strategy on 4-methylbenzhydryl-amine (MBHA) resin (100-200 mesh, 0.7 mmol/g, Fluka) using the tea bag method (23). Side chain protecting groups were acetamidomethyl (Acm) for Cys, benzyl ether for Ser and Thr, benzyl ester for Asp and Glu, tosyl for Arg, dichlorobenzyl for Tyr, and 2-chlorobenzyloxycarbonyl for Lys. Coupling

of the amino acids was achieved by activation with 1,3diisopropylcarbodiimide and was monitored by the Kaiser test (24). Side-chain deprotection [except for Cys (Acm)] and cleavage from the resin was performed by the lowhigh HF procedure, using HF-DMS-p-cresol (25:65:10) for 2 h at 0 °C followed by HF-DMS-anisole-thiocresol (79.8:10:10:0.2) for 1 h at 0 °C (25). The resin was subsequently washed with diethyl ether and dried off under vacuum. The crude peptide was extracted from the resin with 30% acetic acid, diluted with water, and lyophilized. Polymerization via S-Acm Deprotection/Oxidation with Iodine. The above peptide, with its Cys residues protected with Acm, was dissolved at 25 mmol/L in AcOH/H2O (3: 2, v/v), and 1.5 equiv of 80 mM aqueous HCl and 10 equiv of iodine in DMF were added. The polymerization reaction was stopped after 30 min by addition of L-ascorbic acid. The reaction mixture was purified on a PD10 Sephadex G-25 column (1.5 × 10 cm) equilibrated and eluted with 1% AcOH, to remove excess reagents and to prevent precipitation of polymers. After exhaustive dialysis against 1% AcOH, the solution was lyophilized. The polymer was characterized by HPLC gel filtration using a TSK G 2000SW column (8 × 300 mm) equilibrated and eluted with 10% MeCN (+0.05% TFA) in water (+0.1% TFA). Lysozyme, chymotrypsinogen A, and bovine serum albumin were used as molecular weight markers. MAP Synthesis. The MAP version of the V3 peptide (see Figure 3 for details) was assembled by stepwise Boc solid-phase methods (23), using the protection scheme described above, except that Cys residue at the Cterminus was protected with the 4-methoxybenzyl (Mob) group. The Lys residues responsible for branching (one in the tetravalent, three in the octavalent versions of the MAP, respectively) were incorporated as Boc-Lys(Boc)OH and coupled with TBTU/DIEA activation. Double couplings with 10-fold excess of Boc-amino acids and DIEA/HOBt were used for most residues of the V3 epitope. When required by the Kaiser test, capping steps were performed with 4% acetic anhydride and 1% DIEA in DMF. After low-high HF cleavage and deprotection as above, the MAPs were extracted into 0.1 M Tris-HCl, pH 8.0, containing 8 M urea, and dimerization of the MAP monomer by disulfide bond formation was allowed to proceed and monitored by the Ellman test (26). The homodimeric MAPs were dialyzed against 2 M urea in 0.1 M Tris-HCl buffer, pH 8.0 for 24 h, then dialyzed against 1 M AcOH, to remove all the urea and lyophilized. Reduction and S-Carboxymethylation. The lyophilized homodimeric MAPs were dissolved at a concentration of 5 mg/mL in 0.3 M Tris.HCl, 6 M guanidinium chloride,

114 Bioconjugate Chem., Vol. 15, No. 1, 2004

Cruz et al.

Figure 3. The JY1-MAP4 (A) and JY1-MAP8 (B) dendrimeric immunogens are disulfide dimers of lysine-branched constructs containing the JY1 and TT epitopes in a 2:1 ratio linked through a cathepsin cleavage site.

3 mM EDTA buffer, pH 8.5. A 20-fold molar excess of DTT over Cys residues was added, and the solution was flushed with nitrogen and incubated for 2 h at 37 °C. A 2-fold molar excess of iodoacetic acid over DTT was next added, and the solution was kept at room temperature for 1 h in the dark. Modified MAPs were characterized by RP-HPLC and mass spectrometry as described above. Assembly of Peptides on Dextran Beads. The dried dextran beads were chemically modified to introduce anchor functions suitable for peptide synthesis. Thus, 1 g of dextran beads was treated for 24 h with 4 mL of a 0.4 M Fmoc-β-alanine solution (activated with 0.48 M DIPC and 0.8 M NMI in DMF) to modify the hydroxyl groups into amine functions. Unreacted hydroxyl sites were acetylated by treatment with 20% acetic anhydride/ 10% DIPEA/10% NMI in DMF for 30 min. Peptide synthesis was subsequently carried out using Fmoc/tBu chemistry (27). Side chain protecting groups were tertbutyl for Glu, Thr, Ser, and Tyr; trityl for Gln and Asn; Boc for Lys and 2,2,5,7,8-pentamethylchroman-6-sulfonyl for Arg. Fmoc deprotection was done with 20% piperidine in DMF for 20 min, and the resulting Fmoc-piperidine adduct was quantitated spectrophotometrically to determine resin substitution. Couplings were carried out with 200 µmol each of Fmoc-amino acid, HOBt, and DIPC in DMF for 1 h and monitored by the bromophenol blue (28) and Kaiser ninhydrin tests (24). When required, recouplings were performed with 300 µmol each of Fmoc-amino acid, TBTU, and HOBt, in the presence of 900 µmol of DIEA, in DMF. Systematic capping after each coupling step was performed with 4% (v/v) acetic anhydride and 1% (v/v) DIEA in DMF. After the sequence assembly was completed, side chain protecting groups were removed with 50% (v/v) TFA in DCM containing 1% phenol, 2% ethanedithiol, 2% thiophenol, and 2% water (all v/v) for 2 h. Peptide-dextran beads were next washed (4× each) with DCM, DMF, and methanol. Finally, the peptide was cleaved from the dextran beads with 0.1 M NaOH for 15 min, neutralized, and analyzed by RP-HPLC. The main peak was collected and characterized by AAA and mass spectrometry. Conjugation of Peptides and MAP to Carrier Proteins. Peptides and MAP were coupled to HBsAg as described by Deen et al. (29). Briefly, 5 mg of HBsAg was dissolved in 5 mL of 0.2 M K2HPO4, pH 8.0, solid succinic anhydride (1 mg) was added, and the solution was stirred until all anhydride was dissolved, adding 3 M NaOH dropwise as required to keep the pH at 8-8.5. The activated protein was next exhaustively dialyzed against dilute HCl (pH 4.5), treated with 7 mg 1-ethyl-3-(dimethylaminopropyl) carbodiimide (EDAC), and stirred for 10 min at 25 °C. Then, 5 mg of peptide or MAP dissolved in 1 mL of 3 M guanidine hydrochloride in PBS was added and the mixture gently stirred for 3 h at 25 °C.

The conjugates were purified by gel filtration through Sephadex CL-4B (Pharmacia) eluted with PBS. The front-eluting fractions were pooled and quantitated by the Coomassie method (30). Chromatography. Soluble peptides and MAPs were analyzed by RP-HPLC on Vydac C18 (100 × 4.6 mm) and Baker C8 wide-pore (100 × 4.6 mm) columns, respectively, in a Pharmacia-LKB (Sweden) model 2150 dual pump chromatograph coupled to a Knauer (FRG) variable wavelength UV detector. Linear gradients of acetonitrile (+0.05% TFA) into water containing (+0.1% TFA) were used for the separation, at a 0.8 mL/min flow rate. Optical density was monitored at 226 nm. Data were processed with the BioCrom program (CIGB, Cuba). Amino Acid Analysis. Samples of S-carboxymethylated MAPs were hydrolyzed with 6 M HCl containing 0.1% phenol and 0.1% 2-mercaptoethanol for 24 h in vacuum-sealed ampules. After evaporation, amino acids in the hydrolyzate were determined in an Alpha Plus 4151 (Pharmacia-LKB, Sweden) analyzer using a sodium buffer system and o-phthalaldehyde derivatization for fluorescence detection. All samples were analyzed in triplicate. Mass Spectrometry. The FAB mass spectrum was recorded in a JMS HX-110HF double-focusing two-sector spectrometer (JEOL, Japan) equipped with a FAB gun, a collision cell, and a DMA-5000 data analysis system, as previously reported (31). Electrospray MS was performed in a hybrid quadrupole-time-of-flight (Q-TOF) instrument (Micromass, UK) fitted with a Z-spray nanoflow ion source (32). The mass spectrometer was operated with a source temperature of 80 °C and a drying gas flow rate of 50 L/h. Peptides, polymers, peptides cleaved off dextran beads, and MAP purified by RP-HPLC were dissolved in a 50% (v/v) water/ acetonitrile solution containing 1% acetic acid, to an approximate concentration of 5 pmol/µL. The peptide solution was infused at 3 µL/min into the mass spectrometer using a syringe pump (Harvard Apparatus, Holliston, MA). The capillary and cone voltage were set to 3 kV and 35 V, respectively. Data acquisition and processing were performed with the Micromass MassLynx package. Immunizations. Female Balb/c mice 6-8 weeks old (CENPALAB, Cuba) were inoculated subcutaneously on days 0, 14, 28, and 56 with 100 µL with 40 µg of antigen emulsionated in either CFA (day 0) or IFA (other days) (Sigma). Animals were bled 10 days after the fourth dose. A summary of immunization schedules is shown in Table 1. Enzyme-Linked Immunosorbent Assay against Synthetic Peptides. ELISA of anti-peptide sera was done as previously described (33). Briefly, 96-well polystyrene plates (High binding, Costar) were coated with

HIV Peptide Immunogen

Bioconjugate Chem., Vol. 15, No. 1, 2004 115 Table 2. V3 Peptides from Different HIV-1 Isolatesa

Table 1. Summary of Immunization Schedules groups

doses (µg)

adjuvant

mice

isolates

clades

amino acid sequences

JY1-MAP8 JY1-MAP4 Polymer peptide peptide-dextran HBsAg-JY1-peptide HBsAg-JY1-MAP4 HBsAg-JY1-MAP8 negative control

40 40 40 40 40 10 10 10 -

CFA-IFA CFA-IFA CFA-IFA CFA-IFA PBS CFA-IFA CFA-IFA CFA-IFA CFA-IFA

8 8 8 8 8 8 8 8 4

JY1 BRVA IIIB MN LR150 RF

D B B B B B

RQSTPIGLGQALYTT RKRITMGPGRVYYTT SIRIQRGPGRAFVTI RKRIHIGPGRAFYTT SRGIRIGPGRAILAT RKSITKGPGRVIYAT

100 µL of BSA-coupled V3 peptides (Table 2) at 10 µg/ mL and incubated overnight at 4 °C. After three washings with 0.05% Tween 20 in distilled water, plates were blocked with 2% milk in PBS (blocking solution) for 1 h at 37 °C. Samples were diluted with 0.05% Tween 20 and 5% sheep serum in blocking solution and incubated for 2 h at 37 °C. After four washings, an anti-mouse IgG peroxidase conjugate (Amersham, UK) was added and incubated for 1 h at 37 °C. The color reaction with o-phenylenediamine was allowed to develop for 10 min and then stopped with 3 M H2SO4 solution (50 µL), and the absorbance at 492 nm was determined in a SensIdent Scan (Merck, FRG). Results were expressed as log titer, determined by interpolation of OD values at fixed serum dilutions into a standard curve of a serum with known titer. Titer was defined as the log of the highest dilution giving twice the absorbance of negative control sera. Titers are given as the geometric mean ( SD of the titers of individual sera. Peptide Synthesis on Cellulose Membranes for Epitope Mapping. Fifteen 15-residue peptides corresponding to the Ala scan of the RQSTPIGLGQALYTT sequence were synthesized as spots on a cellulose membrane. Twenty-five other paper-bound peptides, seven and five residues long, were also synthesized for epitope mapping. The spot synthesis method of Frank et al. (34) was used. Functionalition of the paper (Whatman 540) was carried out by esterification of an Fmoc-β-Ala-OH residue, using DIPC and N-methylimidazole in DMF. The remaining residues were incorporated by paper synthesisadapted Fmoc/tBu chemistry (34). All peptides were acetylated at the N-terminus at the end of the synthetic program. Paper-Based Peptide Enzyme Immunoassay. Washing steps and incubations were performed at 25 °C with Tris-buffered saline (TBS) containing 0.05% Tween20 (T-TBS). Blocking of free binding sites was performed by overnight incubation with 5% skim milk in TBS. Incubation times were 3 h for sera and 2 h for alkaline phosphatase-conjugated anti-mouse IgG (Sigma). Detection of bound antibody was achieved by incubation with 5-bromo-4-chloro-3-indolyl phosphate (Sigma) in substrate buffer (0.1 M Tris, 0.1 M NaCl, 2 mM MgCl2, pH, 8.9). Washing with PBS stopped staining. The peptidebearing cellulose sheets were regenerated and prepared for the next run as described by Frank (34). Statistical Analysis. Statistical analyses were conducted using Microsoft Excel software. Averages, variance, and standard deviations (SDs) were determined using standard functions. All ELISA OD values were transformed to the natural logarithm to get a normal distribution in the immunized groups. An F-test was performed to assess variance homogeneity among groups and a Student’s t-test was done (p < 0.05 was considered statistically significant).

a

Sequence alignments using Clustal X software.

Table 3. Amino acid Compositions of Antigenic Peptidesa amino acid

peptide

polymer

JY1-MAP4

JY1-MAP8

peptide on dextran

Asx Thr Ser Glx Gly Ala Val Met Ile Leu Tyr Phe His Lys Arg L-Cys

1.3 (1) 3.6 (4) 1.7 (2) 4.5 (4) 5 (3) 1.8 (2) 4.1 (4) 2.8 (3) 2 (2) 1.1 (1) 4.5 (4) 0.8 (1) -

1.1 (1) 3.4 (4) 1.9 (2) 4.2 (4) 4.1 (3) 2 (2) 4.5 (4) 3.1 (3) 1.8 (2) 1.4 (1) 4.3 (4.0) 0.6 (1) 2.3 (2)

1.1 (1) 6.4 (7) 2.3 (3) 6.7 (6) 8.7 (8) 2.7 (3) 5.2 (5) 4.1 (4) 3.3 (3) 0.9 (1) 6.2 (6) 2.1 (2) 1.1 (1)

2.5 (2) 11.7 (14) 5.6 (6) 12.2 (12) 16.3 (19) 5.4 (6) 9.7 (10) 7.8 (8) 5.9 (6) 3.1 (2) 13.7 (11) 4.6 (4) 1.1 (1)

1.1 (1) 3.6 (4) 1.6 (2) 3.8 (4) 3.2 (3) 2.0 (2) 4.1 (4) 3.0 (3) 2.7 (2) 1.2 (1) 3.9 (4) 0.7 (1) -

a Data are expressed as number of residues per molecule. Experimental values represent the mean of three independent determinations. Theoretical values are given in parentheses. Asx: aspartic acid plus asparagines; Glx: glutamic acid plus glutamine. Proline was not detected.

RESULTS AND DISCUSSION

Synthesis of Immunogens. Free and Oligomeric Versions of the JY1 Peptide. The primary structure of the monomeric form of the immunogen is shown in Figure 1. It consisted of the JY1 epitope (RQSTPIGLGQALYTT) in tandem with the T-cell epitope of tetanus toxoid (TT, residues 830-844, QYIKANSKFIGITEL), with an intervening Lys pair designed as a cathepsin-like cleavage site and with flanking Cys-Gly and Gly-Cys residues at each end to promote oligomerization. The peptide was synthesized by stepwise solid-phase methods, with the two Cys residues protected with the HF-stable Acm group. After cleavage and characterization, the Acm-protected monomer was treated with iodine to remove the two Acm groups and oligomerization was performed in AcOH:H2O (3:2, v:v), where the monomer was highly soluble. The reaction proceeded quickly, and a high conversion to oligomer was observed by gel filtration HPLC (Figure 2). ESI-MS analysis of the main fractions of the oligomerization mixture allowed identification of the tetra-, tri-, and dimeric species. In addition, as expected, the monomer underwent intramolecular cyclization. Attempts to reduce the extent of this relatively undesired side reaction by varying the initial concentration of monomer were unsuccessful. Small levels of peptides with masses consistent with the cyclic disulfide forms of the oligomers were also detected. A summary of these findings is shown in Table 3. Multiple Antigenic Peptide System. Two multivalent (MAP-like) dendrimeric presentations of the JY1 peptide with defined composition and varying number of B- and T-cell epitopes were also designed and synthesized (Figure 3). The JY1-MAP4 immunogen, containing four B and two Th epitopes (Figure 3A), resulted from dimerization of a monomer with a single Lys branching

116 Bioconjugate Chem., Vol. 15, No. 1, 2004

Cruz et al. Table 4. Mass Spectrometry of V3 Antigenic Peptides antigen

theory (Da)

experimental (Da)

linear peptide cyclized monomer cyclized dimer cyclized trimer cyclized tetramer JY1-MAP2 JY1-MAP4 peptide on dextran beads

4031.5 3887.50 7774.01 11659.57 15546.06 5500.10 10898.55 3754.3

4031.9a 3887.02 7773.11 11659.67 15546.03 5498.84 10897.63 3754.4a

a Determined by fast atom bombardment mass spectrometry (FAB-MS).

Figure 4. Analytical characterization of JY1-MAP8. A: HPLC analysis after reduction and carboxymethylation; B. ES-MS spectrum of the main HPLC fraction (labeled with an arrow in panel A).

point, while the JY1-MAP8, with eight and four epitopes, respectively (Figure 3B), contained three Lys branching units in each monomer. The monomers were prepared by Boc/benzyl solid-phase chemistry and dimerized as above. To evaluate the quality of each synthesis, an aliquot of each final dimer was reduced and S-carboxymethylated and characterized by reverse phase HPLC (Figure 4). The divalent monomer (JY1-MAP2 precursor) was less clean than the tetravalent monomer (JY1-MAP4 precursor), probably due to the use of a resin with too high substitution. In both cases, however, the main fraction of each HPLC profile was identified by ESI-MS as the expected species (Table 4). In conclusion, the dendrimeric approach allowed the preparation of a high molecular weight immunogen of defined composition in a more reproducible form than oligomerization. Peptide on Dextran Beads. Cross-linked dextran beads were chemically modified with a substoichiometric amount of Fmoc-βAla to give an amine-functionalized support suitable for solid-phase peptide synthesis. The remaining hydroxyl functions of the polymer were blocked by acetylation to avoid the formation of deletion sequences and reduce the substitution to convenient levels. A β-Ala

loading of ca. 30 µmol/g of dry resin was determined by spectrophotometric titration of the Fmoc group. On this dextran support, the JY1-LysLys-TT sequence (as in Figure 1, but without the flanking Gly and Cys residues) was assembled by Fmoc chemistry, using HOBt- and DIPC-mediated couplings with relatively high excess of all components (10 equiv) in DMF. Most residues required double couplings with TBTU activation and prolonged reaction times (2 h) until a negative Kaiser test was observed. For additional safety, a capping step with acetic anhydride was performed after incorporation of each residue. The quality of the dextran-bound peptide was evaluated by RP-HPLC analysis of an aliquot cleaved from the support by alkaline hydrolysis (Figure 5). The main product of the chromatogram was satisfactorily characterized as the target peptide by ESI-MS (Table 4). Peptide and MAP Conjugates. To further enhance the immunogenicity of the anti-peptide response, both the monomeric and the two dendrimeric versions of the JY1TT peptide were conjugated to recombinant HBsAg protein. HBsAg is a polymeric protein of about 2000 kDa. The 25 kDa monomer contains 14 cysteines which form an uncertain number of intra- and intermolecular disulfide bridges, giving rise to particles of 22 nm average diameter, each containing approximately 100 monomer units (35). Our choice of HBsAg was based on its high molecular weight, proven stability, and high immunogenicity and because it is known to act as an effective and safe vaccine in humans. The conjugations of JY1-TT monomer and JY1-MAP4 and JY1-MAP8 dendrimers to HBsAg were done by means of succinic anhydride, a homobifunctional reagent that allows cross-linking of amino groups. The first step was succinylation of HBsAg under mildly alkaline conditions, followed by dialysis to remove excess succinic anhydride and thus avoid unwanted MAP-MAP and/or

Figure 5. HPLC analysis of crude (after base cleavage) RQSTPIGLGQALYTTKKQYIKANSK-FIGITEL-βAla2 synthesized on dextran beads. The peak labeled with an arrow corresponds to the correct sequence. Inset: FAB-MS of HPLC fractions corresponding to the arrow-labeled peak (see also Table 4).

HIV Peptide Immunogen

Figure 6. Antibody titers of mice after immunization with four doses of: A, JY1-MAP8, 40 µg; B, JY1-MAP4, 40 µg; C, polymer, 40 µg; D, peptide monomer, 40 µg; E, peptide on dextran beads, 40 µg; F, HBsAg-JY1-peptide, 10 µg; G, HBsAg-JY1-MAP4, 10 µg; H, HBsAg-JY1-MAP8, 10 µg; I, PBS negative control. Titer is defined as the log of the highest dilution giving twice the absorbance of negative control. Titers are the geometric mean ( SD corresponding to eight mice antisera.

peptide-peptide cross-linking. The carboxyl groups on the modified HBsAg were next activated with EDAC (a soluble carbodiimide) and allowed to react with the free amino groups of either linear monomer or JY1-MAP4 or JY1-MAP8. The conjugates were purified by gel filtration to avoid aggregation. For this purpose a Sephadex S300 matrix was selected. This matrix allowed a quick and efficient separation of those conjugates from byproducts such as carbodiimide and MAPs in excess. Comparison of the Antibody Response in Mice Immunized with Different Presentations of JY1 Peptide. Groups of eight Balb/c mice were inoculated subcutaneously with different presentations of the JY1 peptide antigen and their anti-peptide antibody responses after four doses of the epitope were evaluated by ELISA. Figure 6 shows the titers obtained for the different immunization groups. As can be seen, both HBsAg-JY1MAP4 and HBsAg-JY1-MAP8 conjugates elicited the highest titers against JY1. The superior immunogenicity of these conjugates may be attributed to the combination of large size and epitope multiplicity, plus the known T-cell response of the carrier protein. The importance of epitope multiplicity was evinced by the fact that the antipeptide titers elicited by monomeric JY1 conjugated to HBsAg (group F) were significantly lower (p < 0.01) than when presented as a MAP dendrimer (groups G and H). Epitope multiplicity, however, did not provide a simple, uniform pattern in the immune response against JY1. Thus, octameric JY1-MAP8 (group A) was significantly (p < 0.05) more immunogenic than monomeric JY1

Bioconjugate Chem., Vol. 15, No. 1, 2004 117

conjugated to HBSAg (group F), but a similar effect was not observed for tetrameric JY1-MAP4 (group B; p ) 0.893). On the other hand, the responses to HBsAg-JY1MAP4 and HBsAg-JY1-MAP8, while the highest among all immunization groups, did not vary significantly (p ) 0.107) despite the 2-fold difference in multiplicity, nor were there substantial differences between JY1-MAP8 and JY1-MAP4 immunization groups (p ) 0.055), in contrast to previous results (36). In earlier studies, the MAP system has been shown to be more immunogenic than the peptide conjugated to a carrier protein (37-39). However, the opposite result has also been observed (40), a phenomenon that can be related to the nature of the carrier molecule. The antibody responses generated by immunization with the linear polymer (group C) were significantly lower than those of the JY1-MAP8 dendrimer (p < 0.05). However, no significant difference was observed between the linear polymer and the JY1-MAP4 (p ) 0.468). As for the peptide-dextran conjugate, its anti-JY1 titers were not statistically different from those of the MAP versions or the linear polymer. Finally, as expected, antibody responses of any conjugated or multiple presentation of JY1 were significantly superior to those obtained with the monomeric version. Analysis of Cross-Reactivity. Pooled sera corresponding to the different immunization groups under study (Table 1) were diluted to an anti-JY1 titer of ca. 10-5 and compared with a panel of several V3 peptides from gp 120 HIV-1 (Table 2) for their ability to crossreact. In this experiment, the TAB9 protein, which integrates a number of V3 peptides, was used as a positive control. Results showed that at least three epitopes, MN, IIIB, and JY1, were readily accessible (36). As can be seen in Figure 7, the group immunized with the JY1-TT peptide did not exhibit cross-reactivity with other V3 epitopes even after four doses of immunogen. A similar result was found for the same peptide attached to dextran beads, and only marginal recognition was detected for the oligomerized versions of the epitope. On the contrary, when the JY1 sequence was presented as a MAP dendrimer, significant levels of cross-reactivity with the other epitopes were observed after four doses. Similar patterns of cross-reactivity were found for the most immunogenic conjugates HBsAg-JY1-MAP4 and HBsAg-JY1-MAP8. Thus, four 10-µg doses of. the former immunogen sufficed to achieve cross-reactivity with all the variant peptides. Since a substantial proportion of these 10 µg corresponds (in weight) to the HBSAg, not the JY1-MAP4, this means that the carrier has a

Figure 7. Cross-reactivity of pooled sera from different immunization groups after the fourth inoculation against heterologous V3 sequences. A, JY1-MAP8 B, JY1-MAP4; C, polymer; D, peptide monomer; E, peptide on dextran beads; F, HBsAg-JY1-peptide; G, HBsAg-JY1-MAP4; H, HBsAg-JY1-MAP8; I, negative control; J, positive control (TAB9 protein). The heterologous V3 sequences are shown in Table 1. Plates were coated with BSA-V3 conjugate as specified in the text.

118 Bioconjugate Chem., Vol. 15, No. 1, 2004

Cruz et al.

Figure 8. Alanine scan (first 15 sequences) and fine epitope mapping (20 sequences) of peptide JY1using the spot synthesis technique. Pooled sera from different immunization groups (after fourth dose) were diluted 1:100 and tested against the peptides. Columns A to J, as in Figure 7. Boxes are coded as follows: white, no recognition; light gray, weak recognition; dark gray, intermediate recognition; black, strong recognition.

significant role in the generation of broadly reactive antibodies. It is important to note that, in most cases, recognition of the heterologous peptides was more than 10 times lower than that of the homologous JY1 peptide. This means that only a minor subset of antibodies were able to react with at least another peptide. Those subsets of antibodies emerged only after repetitive immunizations (data not shown). To further analyze the cross-reactivity, an alignment analysis of all V3 peptide sequences (Table 2) was performed. The sequence closest to JY1 was MN, with 53.3% homology, followed by RF and BRVA, with 40% homology; LR150 with 33.3%; and the less related IIIB sequence with 26.7%. The ELISA data were in good agreement with this homology. Our results also agreed with previous findings that antibodies for a specific V3 variant provide cross-neutralization for different HIV-1 field isolates (41, 42). In conclusion, our experiment demonstrated that a certain potential for cross-reactivity is associated to the V3 peptide sequence, but that its intensity strongly depends on the presentation of the epitope to the immune system. Epitope Mapping. We also examined the effect of the different strategies of immunopeptide presentation on the

antibody response to individual residues of the linear JY1 epitope. Our goal was to determine which amino acids in that sequence were more directly involved in the antigen-antibody interaction. For this purpose, we performed fine epitope mapping and an alanine scan of the whole JY1-peptide using the spot synthesis technique. For epitope mapping, twenty seven- and five-residue peptides covering the JY1 sequence were prepared by simultaneous spot synthesis, as shown in Figure 8. For the alanine scan, fifteen 15-residue peptides with a systematic Ala replacement at each position of the JY1 sequence were similarly synthesized. The cellulose-bound peptides were then assayed by the above protocols against 1:100 dilutions of pooled sera from each immunization group. The alanine scan (Figure 8) showed that the LXQXXY or LXQXLY sequences were involved in the immune recognition by all sera. This was in good agreement with fine epitope mapping results, in which sera from all immunization groups strongly recognized the C-terminal region (LGQALY) of the JY1-sequence. In addition, mice immunized with either free peptide, peptide oligomers, or MAP also showed weak recognition of the N-terminal region of the JY1-epitope. This region was not recognized when the epitopes were presented as carrier protein

HIV Peptide Immunogen

conjugates. On the other hand, this analysis of the antibody response against the linear JY1 epitope did not provide any clues on the different cross-reactivity patterns observed in the above experiments. SUMMARY AND CONCLUSIONS

In this paper we have described different approaches to improve the immunogenicity of JY1 peptide epitope presentations. These include (i) the JY1-TT construct, where the JY1 epitope is presented in tandem with a well-known T-cell epitope, (ii) oligomeric, disulfide-linked versions of this chimeric sequence, with sizes between 8 and 16 kDa, iii) dendrimeric, MAP-type presentations of the B- and T-cell epitopes, with well-defined structures and masses in the 11-21 kDa range. These immunogens were administered either free or conjugated to HBsAg, an effective carrier protein. In addition, the JY-TT peptide was also administered immobilized on dextran beads, an inexpensive, biocompatible, and biodegradable natural polysaccharide which holds promise as an attractive adjuvant for peptide vaccination (43). This is, as far as we know, one of very few comparative studies on presentation of peptide epitopes to the immune system. We have observed the highest anti-peptide titers in animal groups immunized with MAP-carrier protein conjugates. All dendrimeric versions of the immunogens elicited antibodies that cross-reacted with other V3 HIV-1 sequences. Furthermore, we have shown that crossrecognition can be enhanced by conjugation to a carrier protein, with the advantage that less amount of immunogen is required to achieve a broadly reactive antibody response. This phenomenon of cross-reactivity toward variable regions such as the V3 loop is highly desirable in designing new vaccine candidates against VIH-1. In addition, this approach may be valuable to generate bivalent vaccines when the carrier protein is per se a vaccine immunogen like the HBsAg. ACKNOWLEDGMENT

This work was supported by the Center for Genetic Engineering and Biotechnology, Havana, Cuba. Work at Barcelona was supported by Spanish CICYT grants (PB97-0873 and BIO2002-04091-C03-01 to D.A., BQ20000235 to F.A.), by AECI (Spanish-Cuban cooperation, to D.A.), and by CERBA, Generalitat de Catalunya (to F.A. and D.A.). LITERATURE CITED (1) Tam, J. P. (1988) Synthetic peptide vaccine design: synthetic and properties of a high-density multiple antigenic peptide system. Proc. Natl. Acad. Sci. U.S.A. 85, 5409-5413. (2) Tomalia, D. A., Baker, H., Dewald, J., Hall, M., Kallos, G., and Martin, S (1985) A new class of polymers: Starburstdendritic macromolecules. Polym. J. 17, 117-132. (3) Volpina, O. M., Yarov, A. V., Zhmak, M. N., Kuprianova, M. A., Chepurkin, A. V., Toloknov, A. S., and Ivanov, V. T (1996) Synthetic vaccine against foot-and-mouth disease based on a palmitoyl derivate of the VP1 protein 135-159 fragment of the A22 virus strain. Vaccine 14, 1375-1380. (4) BenMohamed, L., Wechsler. S. L., and Nesburn, A. B. (2002) Lipopeptide vaccinessyesterday, today, and tomorrow. Lancet Infect Dis. 2, 425-431. (5) Lindner, W., and Robey, F. A. (1987) Automated synthesis and use of N- chloroacetyl- modified peptides for the preparation of synthetic peptide polymers and peptide-protein immunogens. Int J Pept. Protein Res. 30, 794-800. (6) Borras-Cuest, F., Fedon, Y., and Pellit-Camordan, A. (1988) Enhancement of peptide immunogenicity by linear polymerization. Eur J Immunol. 18, 199-202.

Bioconjugate Chem., Vol. 15, No. 1, 2004 119 (7) Defoort, J. P., Nardelli, B., Huang, W., and Tam, J. P. (1992) A rational designo f synthetic peptide vaccine wiht a built-in adjuvant. A modular approach for unambiguity. Int. J. Pept. Protein Res. 40, 214-221. (8) Singh, M., Li, X. M., Wang, H., McGee, J. P., Zamb, T., Koff, W., Wang, C. Y., and O′Hagan, D. T. (1998) Controlled release microparticles as a single dose diphtheria toxoid vaccine: immunogenicity in small animal models. Vaccine 16, 346352. (9) Gregoriadis, G. (1994) Liposomes as immunoadjuvants and vaccine carriers: antigen entrapment. Immunomethods 4, 210-216. (10) Lovgren, K., and Larsson, M. (1994) Conjugation of synthetic peptides to carrier iscoms: factors affecting the immunogenicity of the conjugate. J. Immunol. Methods 173, 237-243. (11) Cruz, L. J., Iglesias, E., Aguilar, J. C., Quintana, D., Garay, H. E., Duarte, C., and Reyes, O. (2001) Study of different coupling agents in the conjugation of a V3-based synthetic MAP to carrier proteins. J. Pept. Sci. 7, 511-518. (12) Moore, J. P., and Nara, P. L., (1991) The role of the V3 loop in HIV infection. AIDS 5, 521-533. (13) Nara, P. L., Garrity, R. R., and Goudsmit, J. (1991) Neutralization of HIV-1: a paradox of humoral properties. FASEB J. 5, 2437-2455. (14) Speck, R. F., Wehrly, K., Platt, E. J., Atchison, R. E., Charo, I. F., and Kabat, D. (1997) Selctive employment of chemokine receptors as human immunodeficiency virus type 1 coreceptors determined by individual amino acids within the envelope V3 loop. J. Virol. 71, 7136-7139. (15) Berzofsky, J. A. (1991) Development of artificial vaccines against HIV using defined epitopes. FASEB J. 5, 2412-2418. (16) Schreiber, M., Wachsmuth, C., Muller, H., Hagen, C., Schmitz, H., and Van-Lunzen, J. (1996) Loss of antibody reactivity directed against the V3 domain of certain human immunodeficiency virus type 1 variants during disease progression. J. Gen. Virol. 77, 2403-2414. (17) Rubinstein, A., Goldstein, H., Pettoello-Mantovani, M., Mizrachi, Y., Bloom, B. R., Furer, E., Althaus, B., Que, J. U., Hasler, T., and Cryz, S. J. (1995) Safety and immunogenicity of a V3 loop synthetic peptide conjugated to purified protein derivative in HIV-seronegative volunteers. AIDS 9, 243-251. (18) Letvin, N. L., Robinson, S., Rohne, D., Axthelm, M. K., Fanton, J. W., Bilska, M., Palker, T. J., Liao, H. X., Haynes, B. F., and Montefiori, D. C. (2001) Vaccine-elicited V3 loopspecific antibodies in rhesus monkeys and control of a simianhuman immunodeficiency virus expressing a primary patient human immunodeficiency virus type 1 isolate envelope. J Virol. 75, 4165-4175. (19) Yamanaka, T., Fujimura, Y., Ishimoto, S., Yoshioka, A., Konishi, M., Naritam, N., Mimaya, J., Meguro, T., Nakasone, T., Okamoto, Y., Yoshizaki, H., Yamada, K., and Honda, M. (1997) Correlation of titer of antibody to principal neutralizing domain of HIVMN strain with disease progression in Japanese hemophiliacs seropositive for HIV type 1. AIDS Res. Hum. Retroviruses 4, 317-326. (20) Esparza, J., Osmanov, S., and Heyward, W. I. (1995) HIV Preventive Vaccines. Progress to date. Drugs 50, 792-804. (21) Yourno, J., Josephs, S. F., Reitz, M., Zagury, D., F., WongStaal, R., and Gallo, C. (1988) Nucleotide sequences analysis of the env gene of a new Zairian isolate of HIV-1. AIDS Res. Hum. Retroviruses 4, 165-173. (22) Panina-Bordingnon, P., Tan, G., Termijtelen, A., Demotz, S., Corradin, G., and Lanzavecchia, A. (1989) Universally immunogenic T-cell epitopes: promiscuous binding to human MHC class II and promiscuous recognition by T-cells. Eur. J. Immunol. 19, 2237-2247. (23) Houghten, R. A., Graw, S. T., and Bray, M. K., (1986) Simultaneous multiple peptide synthesis: The rapid preparation of large numbers of discrete peptides for biological, immunological and methodological studies. BioTechniques 4, 522-526. (24) Kaiser, E., Colescott, R. L., Bossinger, C. D., and Cook, P. I. (1970) Color test for detection of free terminal amino groups in solid-phase synthesis of peptides. Anal.Biochem. 34, 595598.

120 Bioconjugate Chem., Vol. 15, No. 1, 2004 (25) Houghten, R. A., Bray, M. K., Degraw, S. T., and Kirby, C. J. (1986) Simplified procedure for carrying out simultaneous multiple hydrogen fluoride cleavages of peptides resins. Int. J. Pept. Protein Res. 27, 673-678. (26) Ellman, G. L. (1959) Tissue sulfhydryl groups.Arch. Biochem. Biophys. 82, 70-77. (27) Fields, G. B., and Noble, R., L. (1990) Solid-phase peptide synthesis utilizing 9- fluorenylmethoxycarbonyl amino acids. Int. J. Pept. Protein Res. 35, 161-214. (28) Krchnak, V., Maly, A., and Flegel, M. (1988) Solid-phase synthesis of a nonadecapeptide coded for by the v-myb oncogene. Int. J. Pept. Protein Res. 2, 239-244. (29) Deen, C., Claassen, E., Gerritse, K., Zegers, N. D., and Boersma, W. J. A. (1990) A novel carbodiimide coupling method for synthetic peptides: Enhanced antipeptide antibody responses. J. Immunol. Methods 129, 119-125. (30) Bradford, M. M. (1976). A rapid and sensitive method for the quantitation of microgram quantities of proteins utilizing the principle of protein-dye binding. Anal. Biochem. 72, 248254. (31) Gonzalez, J., Besada, V., Garay, H., Reyes, O., Padron, G., Tambara, Y., Takao, T., and Shimonishi, Y. (1996) Effect of the position of a basic amino acid on C-terminal rearrangement of protonated peptides upon collision-induced dissociation. J. Mass. Spectrom. 31, 150-158. (32) Mo, W., Sakamoto, H., Nishikawa, A., Kagi, N., Langridge, J. I., Shimonishi Y., and Takao, T. (1999) Structural characterization of chemically derivatized oligosaccharides by nanoflow electrospray ionization mass spectrometry. Anal. Chem. 71, 4100-4106. (33) Go´mez, C. E., Lo´pez-Campistrous, A. E., and Duarte, C. A. (1998) An immunoassay with bovine serum albumin coupled peptides for the improved detection of anti V3 antibodies in HIV-1 + Human sera. J. Virol. Methods 71, 7-16. (34) Frank, R. (1992) Spot-Synthesis: An easy technique for the positionally addressable, parllel chemical synthesis on a membrane support. Tetrahedron 48, 9217-9232. (35) Hardy, E., Martinez, E., Diago, D., Diaz, R., Gonzalez, D., and Herrera, L. (2000)Large-scale production of recombinant hepatitis B surface antigen from Pichia pastoris. J. Biotechnol. 77, 157-167.

Cruz et al. (36) Cruz, L. J., Quintana, D., Iglesias, E., Garcı´a, Y., Huerta, V., Garay, H. E., Duarte, C., and Reyes, O. (2000) Immunogenicity Comparison of a Multi-antigenic Peptide Bearing V3 Sequences of the Human Immunodeficiency Virus Type 1 with TAB9 Protein in Mice. J. Pept. Sci. 6, 217-224. (37) Chai, S. K., Clavijo, P., Tam, J. P., and Zavala, F. (1992) Immunogenic properties of multiple antigen peptide systems containing defined T and B epitopes. J. Immunol. 149, 23852390. (38) McLean, G. W., Owsianka, A. M., Subak-Sharpe, J. H., and Marsden, H. S. (1991) Generation of anti-peptide and antiprotein sera. Effect of peptide presentation on immunogenicity. J. Immunol. Methods 137, 149-157. (39) Kanda, P., Fritz, D. A., Gage, D. A., and Shuler, K. R. (1995) Dependence of the murine antibody response to an anti-CDR2 VH peptide on immunogen formulation. Mol. Immunol. 32, 1319-1328. (40) Helling, F., and Livingston, P. O. (1994) Ganglioside conjugate vaccines. Immunotherapy against tumors of neuroectodermal origin. Mol. Chem. Neuropathol. 21, 299-309. (41) Wang, C. Y, Looney, D. J, Li, M. L., Walfield, A. M., Ye, J., Hosein, B., Tam, J. P., and Wong-Staal, F. (1991) Longterm high-titer neutralizing activity induced by octameric synthetic HIV-1 antigen. Science 254, 285-288. (42) Zhang, P. F., Bouma, P., Park, E. J., Margolick, J. B., Robinson, J. E., Zolla-Pazner, S., Flora, M. N., and Quinnan, G. V. (2002)Peng Fei Zhang. A variable region 3 (V3) mutation determines a global neutralization phenotype and CD4independent infectivity of a human immunodeficiency virus type 1 envelope associated with a broadly cross-reactive, primary virus-neutralizing antibody response. J. Virol. 76, 644-655. (43) Diwan, M., Misra, A., Khar R. K., and Talwar, G. P. (1997) Long-term high immune response to diphtheria toxoid in rodents with diphtheria toxoid conjugated to dextran as a single contact point delivery system.Vaccine 15, 1867-1871.

BC034119J