Conjugation of Latent Membrane Protein (LMP)-2 Epitope to Gold

Publication Date (Web): December 22, 2008 ... In the presence of a CALNN capping peptide, the AuNP−peptide conjugates are stable in solution without...
0 downloads 0 Views 279KB Size
Download PDF
24

Bioconjugate Chem. 2009, 20, 24–31

ARTICLES Conjugation of Latent Membrane Protein (LMP)-2 Epitope to Gold Nanoparticles as Highly Immunogenic Multiple Antigenic Peptides for Induction of Epstein-Barr Virus-Specific Cytotoxic T-Lymphocyte Responses in Vitro Wai-Hung Cheung,‡,§ Vera Sau-Fong Chan,§,| Heung-Wing Pang,‡ Man-Kin Wong,‡ Zhi-Hong Guo,⊥ Paul Kwong-Hang Tam,§ Chi-Ming Che,‡ Chen-Lung Lin,*,§,| and Wing-Yiu Yu*,†,‡ Department of Applied Biology and Chemical Technology, The Hong Kong Polytechnic University, Hung Hom, Kowloon, Hong Kong, Department of Chemistry and Open Laboratory of Chemical Biology of the Institute of Molecular Technology for Drug Discovery and Synthesis, The University of Hong Kong, Pokfulam Road, Hong Kong, Department of Surgery, L9-30, Laboratory Block, 21 Sassoon Road, Faculty of Medicine Building, The University of Hong Kong, Hong Kong, Division of Surgery, Oncology, Reproductive Biology and Anaesthetic, Faculty of Medicine, Imperial College, U.K., and Department of Chemistry, Hong Kong University of Science and Technology, Clear Water Bay, New Territories, Hong Kong. Received April 22, 2008; Revised Manuscript Received November 3, 2008

Nasopharyngeal carcinoma is a neoplasm with a high incidence in Southeast Asia, and it is strongly associated with Epstein-Barr virus (EBV) activation involving the expression of a weakly immunogenic protein, namely, latent membrane protein (LMP)-2. Previous immunological studies already identified the human leukocyte antigen (HLA)-A11 restricted peptide epitope (SSCSSCPLSK) in the LMP-2 antigen. In this work, we prepared gold nanoparticle (AuNP)-peptide conjugate 1 by treating the nanoparticles with the N-cysteinated LMP-2 epitope. The AuNP-peptide conjugates have been characterized by TEM (15-24 nm in diameter) and UV-vis spectroscopy (surface plasmon resonance absorption band at λmax ) 520 nm). In the presence of a CALNN capping peptide, the AuNP-peptide conjugates are stable in solution without aggregation at room temperature for at least 48 h. By ELIspot studies, AuNP-peptide conjugate 1 was found to elicit a significantly stronger INF-γ response [number of spot forming cells (SPC) ) 727 ( 198] from peripheral blood mononuclear cells of healthy HLA-A11 donors when compared to that induced by the unconjugated LMP-2 peptides (SFC ) 73 ( 28). Further studies showed that dendritic cells treated with conjugate 1 can effect CD8+ T-cell activation leading to epitope-specific cytotoxic T lymphocyte killing responses in vitro.

INTRODUCTION Nasopharyngeal carcinoma (NPC) is a prevalent malignancy in Southeast Asia. Conventional treatment is ineffective for advanced NPC patients and is often accompanied with severe long-term side-effects. In this regard, immunotherapy is now attracting current interest in developing more effective cancer treatment. Epstein-Barr virus (EBV) is a γ-herpesvirus that resides in over 95% of the general population as a life-long asymptomatic harmless infection (1, 2). However, poorly differentiated and undifferentiated NPC is strongly associated with EBV activation. The EBV latent membrane proteins (LMP-1 and -2) expressed in the tumor tissues have become the target antigens for cancer immunotherapy. A recent clinical * (W.-Y.Y.) E-mail: [email protected]. (C.-L.L.) E-mail: [email protected]. ‡ Department of Chemistry and Open Laboratory of Chemical Biology of the Institute of Molecular Technology for Drug Discovery and Synthesis, The University of Hong Kong. § Department of Surgery, The University of Hong Kong. | Imperial College. ⊥ Hong Kong University of Science and Technology. † The Hong Kong Polytechnic University.

study by Straathof and co-workers showed that the adoptive transfer of LMP- and LMP-2 specific cytotoxic T-lymphocytes (CTLs), which were generated by stimulating the peripheral blood mononuclear cells (PBMCs) with autologous EBVtransformed lymphoma cell lines (LCL), could induce antiviral response in vivo and conferred tumor regression in some advanced NPC patients (3). Apart from the LMP proteins, the LCLs are also known to produce six other nuclear antigens (EBNAs 1, 2, 3A, 3B and 3C, and EBNA-LP) (4) and some of which are oncogenic (5-7). Despite the encouraging clinical results, the use of EBV-transformed LCLs for CTL priming may jeopardize the safety of patients. Apparently, the development of synthetic peptide antigens that can induce EBV-specific CTL activation is clearly advantageous over whole-protein antigen mixtures in terms of safety and ease for production. Dendritic cells (DC) are professional antigen-presenting cells for activation of antigen-specific CTL, which underlies effective cancer immunotherapy. Dendritic cells are responsible for processing and presenting human leukocyte antigen (HLA)restricted peptide epitopes to CD8+ T cells through the major histocompatibility complex (MHC)-I pathway (8-11). Recently, one of us reported that intranodal injection of dendritic cells ex vivo pulsed with a HLA-A11 restricted LMP-2 epitope (SSC-

10.1021/bc800167q CCC: $40.75  2009 American Chemical Society Published on Web 12/22/2008

Conjugation of Latent Membrane Protein (LMP)-2

SSCPLSK) induced EBV specific CTL in vivo, and regression of metastatic tumor in selected NPC patients has been observed (12). Yet, the antitumor effect of this DC-based immunotherapy for NPC was not seen in all tested individuals, and the poor immunogenicity of the LMP-2 epitope is probably one of the reasons. Conventionally, keyhole limpet heamocycnin (KLH) and bovine serum albumin (BSA) have been used as carriers to conjugate with peptides for inducing antibody responses. Recently, KLH has been shown to have an adjuvant effect that primes dendritic cells for enhancing peptide specific CD8+ T cell responses (13-15); however, the potential side effects (16) from the use of this potent adjuvant have been a cause for concern in human settings. As a result, several alternative approaches have been exploited to enhance the immunogenicity of peptide antigens for human uses. A notable example is the multiple antigenic peptide (MAP) system based on the conjugation of multiple peptide epitopes onto a benign carrier such as liposomes and branched/dendrimeric poly lysine core (17-22). While arduous synthetic processes are often required, the purification and characterization of the MAP peptides are also technically demanding. In addition, the number of different epitopes incorporated into the construct is also limited, rendering development of a broad scope peptide vaccine difficult. The use of nanoparticles as delivery vehicles for therapeutic agents is receiving current attention (23-27). For example, nanoparticle based biodegradable polymers such as poly(γglutamic acid) and poly(DL-lactic-coglycolic acid) have been used to deliver peptide/protein antigens to murine and human dendritic cells for effective induction of antigen-specific CTL responses (28). However, the nanoparticles were prepared by self-assembly of their polymers, and this process often operates in a small scale and at high dilution. By virtue of high surface area and chemical stability, biomolecule functionalized gold nanoparticles (AuNP) have found important applications in many areas, including catalysis, biosensing, and diagnostics (29-36). In search of highly immunogenic peptides for potential application to NPC immunotherapy, here we describe that bioconjugation of the LMP-2 epitope (SSCSSCPLSK) to gold nanoparticle (AuNP) is a simple and effective approach for assembling solution-stable immunogenic multivalent peptides. Immunoassay studies showed that the AuNP-peptide conjugate is a potent immunogen in inducing interferon (IFN)-γ production by peripheral blood mononuclear cells. Also, treatment of human dendritic cells with the nanoparticle-peptide conjugate resulted in EBV-specific CTL priming in vitro.

MATERIALS AND METHODS The Fmoc-protected amino acids and the Wang resins were purchased from EMD Biosciences, Inc. (Novabiochem, 10394 Pacific Center Court, San Diego, CA 92121); the Wang resins precoated with amino acids were purchased from Advanced ChemTech Inc. (Louisville, KY), and the other chemicals were purchased from Aldrich Chemical Co. (P.O. Box 206, Milwaukee, WI 53201). All of the solvents were of HPLC or peptide synthetic grade and were used as received. The nanopure water was collected from a Barnsted water purification system (nanopure diamond). Buffy coats were obtained from the Hong Kong Red Cross for research purposes; anti-HLA antibodies were purchased from US Biology Corp. (USA); antihuman CD8 and the isotype control antibodies were obtained from PharMingen (USA). The cytokines, GM-CSF (granulocyte macrophage colony stimulating factor) and Interleukin (IL)-4, were obtained from Peprotech Inc. (USA). MultiScreenHTS-IP PVDF filter Elispot plates (MSIPS4510) were purchased from Millipore Corp (USA); the ELIspot kit (SEL285) and ELIspot Blue Color Module (SEL002) were purchased from R&D System (USA). The 51Cr-sodium

Bioconjugate Chem., Vol. 20, No. 1, 2009 25

chromate isotope for the chromium release assays was purchased from GE Healthcare Medical Co. Ltd., (CJS1-1mCi, USA). The peptides were prepared by a Chemspeed 500 parallel synthesizer. All of the samples were lyophilized by a Savant system (SC250DDA speedvac Plus). The peptides were analyzed and purified on a Waters HPLC system (Waters Alliance 2690 equipped with a 996 PDA UV-vis detector or Waters delta 500 preparative HPLC). The masses of the samples were determined by either matrix-assisted laser desorption ionization time-of-flight (MALDI-TOF, Amersham Biosciences, Voyager DE-STR MALDI TOF Mass spectrometry) or fast atom bombardment ionization (FAB, Finnigan MAT 95). The gold nanoparticles (AuNP) were analyzed by a transmission electron microscope [TEM, Philips (Techian 20)]. For the uptake of the AuNP-peptide conjugates by dendritic cells, a transmission electron microscope of Philips EM208s model was employed. Solid Phase Peptide Synthesis. All of the peptides including the multiple antigenic peptides (MAP) were synthesized by solid phase peptide synthesis using the Fmoc-protected amino acid strategy (37). The crude peptides were purified by preparative HPLC, and >95% purity of the peptides were confirmed by analytical HPLC equipped with a UV-vis detector (monitoring wavelength λ ) 214 nm). The purified peptides were lyophilized to give an amorphous white solid. The peptides were dissolved in 10 µL of DMSO and diluted with sterile phosphate buffered saline (PBS) to 2 mM. The solutions were kept at -70 °C for long-term storage (6 months) and -20 °C for temporary storage (2 weeks). The mass spectrometry data of the synthetic peptides are available in Supporting Information. The multiple antigenic peptide (MAP) 7 was synthesized from the Wang resin precoated with ([Fmoc-Lys(Fmoc)]2-Lys-Cys(Acm)-β-Ala-resin (SM5105, Advanced ChemTech, Inc., Louisville, KY) using the Fmoc protected amino acid strategy as described earlier. The crude products were dialyzed with a dialysis tubing (1200 dalton cutoff; Sigma, D7884-10FT) in nanopure water containing urea (10%) and acetic acid (10%) for 48 h at room temperature. The purified products were lyophilized and analyzed by ESI-MS. The animo acid sequence of 7 was confirmed by the amino acid sequencing technique (Department of Zoology, The University of Hong Kong). MAP 8 was synthesized in a similar manner using the precoated Wang resin ([Fmoc-Lys(Fmoc)]4-Lys2-LysCys(Acm)-β-Ala-Wang resin, SM5102). Synthesis of Gold Nanoparticles (AuNPs) (38). A sodium citrate solution (0.5 g in 10 mL of water) was added to a boiling KAuCl4 solution (36 mg in 90 mL of water). The color of the reaction mixture changed gradually from pale yellow to wine red over 30 min. The gold nanoparticles obtained in this manner were found to have particle sizes of 15-24 nm based on transmission electron microscopic analysis. The gold nanoparticle solution was used for the subsequent peptide conjugation without purification. Synthesis of AuNP-Peptide Conjugate 1. To the gold nanoparticle solution prepared above (9 mL) was added the LMP-2′ peptide (C-SSCSSCPLSK; 0.1 mL of 2 mM in water) and the capping peptide CALNN (Cys-Ala-Leu-Asn-Asn) (0.9 mL of 2 mM in water). The reaction solution was stirred overnight at room temperature to afford AuNP-peptide conjugate 1 as a wine red solution. On the basis of HPLC analysis of the reaction mixture, unreacted/unbound LMP-2′ and CALNN peptides were not detected. The conjugate was purified by Sephadex column chromatography and lyophilized to obtain a red powder (3.0 mg). The red powder was resuspended in PBS for TEM, UV-vis spectroscopic analysis. For the preparation of AuNP-peptide conjugates 2-6, a similar protocol was employed by using different ratios of LMP2′ and CALNN peptides (see Supporting Information).

26 Bioconjugate Chem., Vol. 20, No. 1, 2009

HPLC Analysis of Peptides. Ten microliters of a sample was injected to a reverse phase column (Waters, Xterra column, 4.6 × 150 mm) using a gradient of solvent mixtures [solvent A (H2O with 0.1% trifluoroacetic acid) and solvent B (acetonitrile with 0.1% trifluoroacetic acid); gradient program: 0-3 min A (100%), 20 min A (50%), 25 min A (100%), 25-30 min A (100%)] using an Alliance 2690 HPLC system with a photo diode array PDA (996) detector (monitoring wavelength λ ) 214 nm). The % amount of the peptides was determined by comparing the authentic samples. Peripheral Blood Mononuclear Cells (PBMCs) (39). The PBMCs were harvested from the buffy coats (Red Cross Hong Kong) by Ficoll-Hypaque (Pharmacia Biotech, Uppsala, Sweden) density gradient centrifugation. The PBMCs were washed with PBS three times and resuspended in culture medium for immediate use or frozen for later experiments by keeping in liquid nitrogen. Human Leucocyte Antigen (HLA) Typing. The buffy coat (1 mL) was diluted with PBS (1 mL) and incubated with ACK buffer (2 mL; 150 mmol/L NH4Cl, 61mmol/L KHCO3, and 1 mmol/L Na2EDTA; pH 7.4) for 5 min to lyse the red blood cells. The white blood cells were washed with PBS three times and resuspended in PBS (1 million cells/mL; 50 µL). To identify the HLA types of each PBMC, HLA-A2, HLA-A11, HLA-A2/ A24, and HLA-A11/A24 antibodies (US Biology, H6098-28A and H6098-07A) were employed for immunostaining (at 4 °C for 20 min), followed by incubation with FITC-conjugated secondary antibody and subsequent flow cytometric analysis (BD, FACS Calibur). In this work, only the HLA-A11+/A2+ donors were used for the experiments. The HLA A2/A24 and HLA A11/A24 antibodies (US Biology, H6098-31, H6098-11) were also employed to discriminate any HLA A24+ donors. ELIspot Assay. The ELIspot assay was performed using the ELIspot kit (SEL285) and ELIspot Blue Color Module (SEL002) by following the manufacturer’s instruction. Briefly, 2 µL of the AuNP-peptides as prepared above was added to 1 mL of PBMCs (4 × 106 cell number) in complete medium (10 unit/ mL of penicillin, 10 µg/mL of Streptomycin, and 10% fetal bovine serum in RPMI1640) and IL-2 (10 unit/mL) at 37 °C with 5% CO2 for 18 h. The treated PBMCs were placed in triplicate onto the microtiter plate (96 wells) precoated with capture antibody. The ELIspot plate was kept at 37 °C with 5% CO2 overnight and then washed with washing buffer (PBS with 0.05% of tween) three times. A detection antibody was added to the microtiter plate, and the mixture was kept at 2-8 °C for 16 h. The microtiter plate was then washed with washing buffer (three times) and the INFγ-releasing cells were revealed as color spots by adding a color developing agent. The blue spots were counted with a dissection microscope. Generation of Dendritic Cells (40). PBMCs (1 - 2 × 107 cells/well) were plated on a 6-well plate at 37 °C with humidified 5% CO2 for 2 h in complete medium. The nonadherent cells were then removed, and fresh complete medium with GM-SCF (100 ng/mL) and IL-4 (100 ng/mL) was added to the semiadherent monocytes. The medium and the cytokines were changed every two to three days. On day 7, the immature dendritic cells were harvested. The purity of the harvested dendritic cells was about 80% as determined by immunostaining with antihuman CD11c antibody, followed by flow cytometry analysis. Uptake of AuNP-Peptide Conjugate by Dendritic Cells (31). Briefly, 2 µL of 1 was added to 5 × 105 immature dendritic cells in 1 mL medium at 37 °C in humidified 5% CO2 for 24 h. The treated cells were washed twice and fixed with 2.5% glutaraldehyde in cacodylate buffer at 4 °C for 16 h. Afterward, the dendritic cells were then washed and supported in agar gel. The agar gel was further dehydrated by shaking with stepwise increase in ethanol concentration from 50% to 100%. The

Cheung et al.

dehydrated agar was embedded with epoxy resin by polymerization overnight at 60 °C. The resin was sliced for electron microscopic analysis. The dendritic cells were examined by a Philips EM208s transmission electron microscope (Electron Microscope Unit, Department of Pathology, Queen Mary Hospital). Generation of Peptide-Specific Cytotoxic T-lymphocytes. The HLA A11+ CD8+ T cells were isolated from HLA-typed PBMCs by negative depletion (Dynal Biotech ASA, Dynabeads Goat antimouse IgG, 110.33) according to the manufacturer’s instructions. The CD8+ T cells were obtained with about 80% purity based on flow cytometry analysis. On day 7 (see Scheme 1), 1 (2 µL) was added to the immature dendritic cells (iDCs, 2 × 105 cells irradiated with 3000 rad) in 0.1 mL medium generated earlier, and the mixture was incubated for 1 h. The iDCs were then washed and mixed with CD8+ T cells (1 × 106 cells) in 1 mL of complete medium with β-mercapthanol (55 µM) and IL-2 (50 unit/mL). On day 10, 50 unit/mL of IL-2 was added to the medium. On day 14, CD8+ T cells were restimulated with the autologous dendritic cells pulsed with the peptide-AuNP conjugate as described above, and the restimulation procedure was repeated on day 21. The CD8+ T cells were harvested on day 28, and the specific killing efficiency of CTL was assessed by the chromium (51Cr) release assay. In addition to 1, MAP 7 (4 µL of 2 mM in water) and LMP-2 (native epitope ) SSCSSCPLSK, 50 µL of 2 mM in water), 6 (2 µL of working solution), MAP-A24 (4 µL of 2 mM in water; see Figure 3 for structure) and HLA mismatched peptide (TYGPVFMSL, 50 µL of 2 mM in water) were used for comparison and as irrelevant controls. Chromium (51Cr) Release Assay (41). To prepare the target cells, autologous or HLA-matched PBMCs (4 × 106 cells) were cultured with phytohemagglutinin (PHA, 0.2 mg/mL) and IL-2 (400 unit/mL) in complete medium for 2 days to obtain PHA blast cells (42). The blast cells (43) (2 × 106) were washed three times and treated with 100 µL (0.1 mCi) of Na2CrO4 (1 mCi, Amersham, CJS1-1mci) for 1 h at 37 °C with 5% CO2 to obtain chromium labeled target cells. The targets were split into two equal portions and pulsed with LMP-2 peptide (SSCSSCPLSK, 100 µM) and irrelevant peptide (YLSGANLNL, 100 µM) for 1 h. The peptide-pulsed PHA blasts were resuspended to 105 cell/mL and aliquot (100 µL, 104 cells/ well) into the U-bottom 96-well plate in triplicate. The specific CTLs (which were stimulated with 1, 6, 7, MAPA24, and HLA mismatched peptide) were harvested and used as effector cells for the cytotoxicity assay. Effector cells in triplicate (6.0 × 105, 2.0 × 105, 6 × 104, and 2 × 104 cells per well; cell number would vary according to donors) were placed on a 96 U bottom plate and cocultured with target cells (1 × 104 cells per well) for 4 h. The maximum release was obtained by lysing the target cells with 5% Triton X-100 in PBS. The spontaneous release was obtained from the supernatant of the target cells in medium only. The released chromium was counted by a scintillation counter (LKB Wallac, 1470 Wizard). % Cell lysis ) (experimental release - spontaneous release)/(maximum release - spontaneous release) × 100%.

RESULTS AND DISCUSSION Gold nanoparticles (AuNP) were prepared by reduction of KAuCl4 (1 mmol) with sodium citrate according to a reported method (38). The nanoparticles showed a characteristic surface plasmon resonance absorption band at λ ) 520 nm, and a TEM study revealed a size distribution of 15-24 nm (Figure 1). By standard solid phase peptide synthesis, the LMP-2 epitope (SSCSSCPLSK) (44, 45) was prepared with purity >95% as confirmed by HPLC analysis. In this work, a cysteine residue

Conjugation of Latent Membrane Protein (LMP)-2

Bioconjugate Chem., Vol. 20, No. 1, 2009 27

Scheme 1. Flow Chart Showing the Processes of the Cytotoxic T-Lymphocyte Generation and the

51

Cr Release Assaya

a Reagents and conditions: (i) PBMC (2 × 107 /well, 6 well), 37 °C, 5% CO2, 2 h in complete medium (10 unit/mL of penicillin, 10 mg/mL of Streptomycin, 10% fetal bovine serum in RPMI 1640); (ii) removed non-adherent cells and washed semi-adherent monocytes by complete medium; (iii) replenished monocytes with complete medium containing GMCSF (100 ng/mL) and IL-4 (100 ng/mL) and changed medium every 3 days; (iv) 10% DMSO, 50% fetal bovine serum in RPMI 1640, in liquid nitrogen; (v) replenished with completed medium, 37 °C, 5% CO2; (vi) mouse anti-human CD4, CD14, CD16, and CD56 (10 µg/mL/106 cell), 4 °C, 20 min, washed 3 times; (vii) Dynabeads (goat anti-mouse, IgG 110.33, 100 µL/mL), 4 °C, 30 min; (viii) magnetic beads depletion; (ix) immature DC (0.2 × 106, irradiated with 3000 rad), 1 (2 µL), 1 h; (x) CD8+ T (1 × 106, 1 mL) in complete medium with β-mercapthanol (55 µM) and IL-2 (50 unit/mL); (xi) replenish fresh medium with β-mercapthanol (55 µM) and IL-2 (50 unit/mL) every 3 days; (xii) PBMC (4 × 106), Phytohemagglutinin (PHA, 0.2 mg/mL) and IL-2 (400 unit/mL) in complete medium, in 2 days to become blast cells; (xiii) blast cells (2 × 106), Na2CrO4 (100 µL of 1 mCi/mL), 1 h, 37 °C, 5% CO2, washed 3 times; (xiv) divided PHA blast cells into 2 equal parts and pulsed with LMP-2 peptide (SSCSSCPLSK, 50 µL of 2 mM) and irrelevant (YLSGANLNL, 50 µL of 2 mM), 37 °C, 5% CO2, 1 h, washed 3 times; (xv) co-cultivated different E:T ratio of CTLs (100 µL) to PHA blast cells 104 cell/100 uL/well, 37 °C, 5% CO2, 4 h. (Note: E:T ratio varied from donor to donor and depended on CTL harvested.)

was appended to the N-terminus of the LMP-2 epitope (i.e., C-SSCSSCPLSK ) LMP-2′) to enable peptide ligation to the Au surface. Conjugation of LMP-2 Peptide Epitope to Gold Nanoparticles. To the AuNP solutions was added the LMP-2′ peptide and the CALNN (Cys-Ala-Leu-Asn-Asn) peptide (mole ratio of LMP-2′/CALNN ) 1:9), and a red solution of AuNP-peptide conjugate 1 was obtained. HPLC analysis of the red solution revealed that all of the LMP-2′ and the CALNN peptides were completely consumed for ligation to the nanoparticles. Conjugate 1 can be purified by Sephadex G-10 column chromatography and lyophilized to afford a red powder (3.0 mg). It is noteworthy that any species corresponding to the inter/intramolecular disulphide bond formation were not observed by LC-MS analysis (46). By employing different quantities of the LMP-2′ and CALNN peptides for the bioconjugation, AuNP-peptide conjugates 2-6 were prepared (see Supporting Information for details). In all cases, the LMP-2′ peptides were completely

consumed for the bioconjugation, and thus, the equivalent peptide concentrations for the AuNP-peptide conjugates in solution should be 20 (1), 40 (2), 60 (3), 140 (4), and 200 µM (5). These solutions will be utilized as the working solutions for the biological assays. Treating the AuNPs with CALNN peptide alone (i.e., without LMP-2′ peptide) gave conjugate 6, which will be used as the negative control in the immunoassays. As noted earlier, conjugate 1 can be isolated as a red powder after Sephadex column purification followed by lyophilization. Conjugate 1 can be dissolved readily in PBS, and the UV-vis spectrum is characterized by a distinct UV-vis absorption band at λmax ) 520 nm. Similar spectral features were observed for 2-6 as well as the peptide-free gold nanoparticles (Figure 1b). For all of the AuNP-peptide conjugates, no significant changes in the particle diameters were observed after peptide conjugation according to TEM analyses. As reported by Levy and co-workers (47), the CALNN sequence is important for maintaining solution stability of the

28 Bioconjugate Chem., Vol. 20, No. 1, 2009

Cheung et al.

Figure 1. (a) TEM image of AuNP-peptide conjugate 1 (diameter ) 15-24 nm) against PBS (shaded area); (b) UV-vis spectra of 1 (solid line) and 6 (dotted line) in PBS showing a characteristic surface plasmon band at 520 nm. Inset: schematic depiction of the AuNP-peptide conjugates.

Figure 2. ELIspot results depicting the INF-γ release from the PBMCs of a HLA-A11 donor 1 upon stimulation by the peptides/conjugates; HLA mismatched A24 peptide ) TYGPVFMSL.

AuNP-peptide conjugate without aggregation. By monitoring the surface plasmon resonance band (520 nm) of AuNP-peptide conjugates 1-4 (in PBS buffer) containing the CALNN sequence, no significant spectral changes were observed for those samples. The solutions of 1-4 were found to be stable at room temperature for at least 48 h without apparent precipitation. However, for the PBS solution of 5 that lacks the CALNN sequence, significant precipitation occurred over 48 h at room temperature. As expected, 6 containing only the CALNN peptide alone was stable in PBS solution under identical conditions. INF-γ ELIspot Assay for the Peptide-Conjugated Gold Nanoparticles. Gamma interferon (INF-γ) is an important cytokine in provoking antiviral immunity (48). The immunogenicity of the AuNP-peptide conjugates 1-6 was evaluated by their activity to activate INF-γ producing cells in peripheral blood mononuclear cells (PBMCs) using the enzyme linked immunospot (ELIspot) assay. PBMCs from four healthy HLAA11 donors were employed for this study. Briefly, PBMCs were treated with the AuNP-peptide conjugate 1 (equivalent epitope concentration in culture medium ) 0.04 µM) in a 24-well plate for 16 h at 37 °C. The peptide-treated PBMCs (2 × 105 cells) were then incubated in triplicate wells for 24 h with INF-γ capture antibody precoated on a PVDF membrane. After washing, a color developing agent was added to the plate, and the INF-γ-releasing cells were identified as colored spots. The number of spot-forming cells (SFC) was counted with a dissection microscope. Figure 2 shows the ELIspot results from donor 1. As expected, stimulation with nonpeptide conjugated AuNPs, conjugate 6 with CALNN peptide-alone, and the HLA-mismatched (TYGPVFMSL) peptides could not induce any significant IFN-γ response in the PBMCs. Interestingly, the unconjugated LMP-2

Figure 3. Schematic depiction of the multiple antigenic peptides 7, 8, and MAP A24.

and LMP-2′ peptides (10 µM) also failed to induce a significant IFN-γ response [spot-forming cells (SFC) ) 73 ( 28 vs 117 ( 24 of no peptide control) suggesting that this peptide per se was a poor immunogen and was unable to stimulate significant cytokine response from the PBMCs of healthy individuals by a simple single stimulation procedure. These results were in agreement with previous findings that this HLA-A11 restricted LMP-2 epitope (SSCSSCPLSK) is weakly immunogenic and that the CTL precursor frequency for this epitope in healthy individuals is very low (e.g., < 1/100,000 PBMCs) (49). However, when this weakly immunogenic epitope was administered in the form of gold nanoparticles, i.e., conjugate 1, a strong and significant INFγ response was observed (SFC ) 727 ( 198 vs 117 ( 24 in no peptide control, P ) 0.007). It is noteworthy that in these experiments, the equivalent epitope concentration of conjugate 1 (0.04 µM) was much less than that of the unconjugated peptide (10 µM) under our experimental conditions, and yet, the observed INFγ response with conjugate 1 was 10-fold stronger than of the native peptide. Apart from the AuNP-peptide conjugates, the immunogenicity of the multiple antigenic peptides (MAP) derived from a branched lysine core was also examined in parallel by the ELIspot assays. In this work, antigen 7 conjugated with four LMP-2 epitopes (Figure 3) was also prepared by conventional

Conjugation of Latent Membrane Protein (LMP)-2

Figure 4. ELIspot results depicting the INF-γ release from the PBMCs of a HLA-A11 donor 4 upon stimulation by AuNP-peptide conjugates. Conjugates 1 with SFC of 1325 ( 218 (P < 0.001) when compared with the untreated group.

solid phase peptide synthesis. We found that stimulating the PBMCs of donor 1 by antigen 7 (8 µM) produced a sizable INF-γ response (SFC ) 237 ( 33, P ) 0.008 when compared with that of the no peptide control). Yet, this response was only about one-third of that induced by the AuNP-peptide 1 (SFC ) 727 ( 198). However, an octavalent derivative 8 (i.e., eight LMP-2 epitopes covalently linked to a lysine core) failed to elicit a comparable INFγ response (SFC ) 88 ( 4; data not shown). The ELIspot results for two other HLA-A11 healthy donors-2 and -3 are given in Supporting Information. Irrespective of the variation among the donors, 1 induced the strongest INF-γ responses in the PBMCs of all the individuals versus other peptides tested herein. Other AuNP-peptide conjugates with different LMP-2′ content were also tested for the INFγ release; the ELIspot assay was performed by employing a fixed quantity (2 µL) of the working solutions of 1-5. As shown in Figure 4, stimulation of the PBMCs from a healthy A11 donor-4 with 1 (equivalent epitope concentration ) 0.04 µM) induced a strong INFγ response (SFC ) 1325 ( 218; P ) 0.0005 compared with that of the no peptide control). Comparable responses were obtained for 2 (0.08 µM), 3 (0.12 µM), and 4 (0.28 µM; equivalent epitope concentrations in the culture medium given in brackets). However, conjugate 5 (equivalent epitope concentration ) 0.4 µM without the CALNN capping peptide) induced a minimal INFγ response (SFC ) 145 ( 27). It is noteworthy that this observed induction is only comparable to that induced by unconjugated LMP-2 peptide (10 µM; SFC ) 155 ( 19). These results indicate that the stabilization effect offered by the CALNN capping peptide is essential for the optimal induction of immune response by these AuNP-peptide conjugates. Dendritic cells (DC) are important for the efficient activation of antigen-specific CTL response, and we therefore examined the uptake of the AuNP-peptide conjugate by human dendritic cells. Immature DCs from a healthy HLA-A11 donor-5 were cocultured with 1 for 24 h. After washing, the stimulated DCs were processed for transmission electron microscopy (TEM) analysis. As shown in Figure 5, the gold nanoparticles can be identified as dark spots in the cell cytoplasm, suggesting that internalization of the antigen into the DCs may enable antigen processing by the MHC class I pathway for CD8+ T lymphocyte priming. Next, we turned to test the ability of the AuNP-peptide conjugates to induce antigen specific CTL response. The immature dendritic cells derived from the PBMCs of three healthy HLA A11+ donor-6, -7, and -8 were stimulated with either conjugate 1 or other control peptides for 1 h, and were subsequently cocultured with autologous CD8+ T cells (80% purity) and restimulated weekly over 21 days (Scheme 1). The

Bioconjugate Chem., Vol. 20, No. 1, 2009 29

Figure 5. (a) TEM image of a dendritic cell showing the uptake of 1 into the cell (N denotes the nucleus); (b) magnification of the marked area shows 1 indicated by arrows.

Figure 6. Results of % specific killings by CTLs reactivated from three HLA A11+ donors 6, 7, and 8 with various antigens, in response to autologous PHA blast cells pulsed with LMP-2 (SSCSSCPLSK). HLA mismatched peptide YLSGANLNL as irrelevant peptide. Inset indicates the effector-to-target (E:T) ratio.

cytotoxic CD8+ T-cells were subsequently harvested and tested for their ability to lyse autologous PHA blast cells presenting the EBV LMP-2 epitope SSCSSCPLSK by 51Cr release assay (41). PHA blasts pulsed with an irrelevant peptide YLSGANLNL was also included as a nonspecific target for determining antigen-specific CTL killing. The antigen specific CTL killing was determined by the percent cell lysis observed with the LMP-2 target versus the one observed with an irrelevant target. As depicted in Figure 6, stimulation of the DCs with 1 produced effective CTL responses for all three donor-6, -7, and -8 as manifested by the significant lysis of peptide-specific target cells of 15-25%. In contrast,