Secondary Logo

Journal Logo

Basic Science

Protective Efficacy of Multiepitope Human Leukocyte Antigen–A*0201 Restricted Cytotoxic T-Lymphocyte Peptide Construct Against Challenge With Human T-Cell Lymphotropic Virus Type 1 Tax Recombinant Vaccinia Virus

Sundaram, Roshni PHD*†; Lynch, Marcus P BS*†; Rawale, Sharad PHD*; Dakappagari, Naveen PHD*; Young, Donn PHD; Walker, Christopher M PHD§∥; Lemonnier, Francois PHD; Jacobson, Steven PHD#; Kaumaya, Pravin T. P PHD*†**††‡‡§§

Author Information
JAIDS Journal of Acquired Immune Deficiency Syndromes: November 1st, 2004 - Volume 37 - Issue 3 - p 1329-1339


Human T-cell lymphotropic virus type 1 (HTLV-1) is the causative agent of several inflammatory disorders and is known to have infected an estimated 10 million people worldwide. The most prominent disorder is HTLV-1–associated myelopathy or tropical spastic paraparesis, a chronic inflammation of the central nervous system. HTLV-1 associated myelopathy/tropical spastic paraparesis (HAM/TSP) also causes an aggressive T-cell malignancy called adult T-cell leukemia (ATL).1–3 HTLV-1 is associated with a multitude of naturally processed cytotoxic T-lymphocyte (CTL) epitopes, giving rise to a pronounced CTL response in infected individuals. Therefore, strategies aimed at generating a broadly protective CTL response that targets multiple antigens and major histocompatibility complex (MHC) alleles may be useful against infection. Although a large number of studies have implicated the importance of CTLs in the clearance of viral infections, the delivery of peptides into the class I pathway that can induce antigen-specific CTL responses in vivo is limited by a number of factors such as antigen processing and transport of epitopes to the endoplastmic reticulum (ER) for association with MHC class I. Most MHC class I ligands are generated by the activity of the 20s proteasomes,4 which are the main cellular proteases that play a key role in the elimination of proteins marked for degradation. The proteasomes display a marked selectivity in the peptides that are liberated and strongly influenced by the residues that flank the C-terminal putative cleavage site.5

Another crucial step in the development of vaccine candidates for clinical trials is the demonstration of protection in animal models of infectious disease or cancer. There are several animal models of HTLV-1 infection, including small animal and nonhuman primates.6–12 These models have been useful in the study of the pathogenesis and molecular mechanisms of infection as well as in vaccine studies using recombinant and protein-based vaccine candidates. Some of the studies included the demonstration of immune responses (humoral, cell mediated, or both) involved in protection against infection, these immune responses were all species restricted. However, and, to date, there is no animal model that can be used to evaluate human determinants involved in protection because of the problem of MHC restriction. Hence, it is not possible to evaluate MHC-restricted cell-mediated responses in the context of human infection and immunity. Furthermore, HTLV-1 does not replicate well in murine cells. Recombinant vaccinia virus expressing the antigen of interest can be used to infect immunized animals and to evaluate the efficiency of vaccination as a feasible alternative. Indeed, several reports have described the use of recombinant vaccinia viruses expressing the desired target antigen from different viruses that can infect mice.13–15 In the absence of an appropriate animal model, this surrogate model of using recombinant HTLV-1 vaccinia virus and human leukocyte antigen (HLA)-A*0201 transgenic mice is an attractive alternative for preclinical protection studies using human MHC-restricted vaccine candidates.

The design and delivery of a multiepitope CTL construct with appropriate spacers was recently reported by us and others to be efficient in the generation of responses against individual epitopes.16,17 We showed that a peptide construct composed of 3 CTL epitopes with intervening arginine spacers was efficiently processed in vitro and in vivo to liberate the individual epitopes and that high cellular responses were observed in HLA-A*0201 transgenic mice. We wished to evaluate whether the positioning of these epitopes relative to each other in such multiepitope constructs would affect the rate or pattern of generation of the individual epitopes and whether this would affect the immunogenicity of each corresponding epitope. This is an important consideration when designing multiepitope vaccines, because each construct is seen by the proteasome as a different substrate. We designed a new multiepitope immunogen incorporating Tax11–19, Tax178–186, and Tax306–315 and separated by double-arginine residues. Four different variants of this construct were designed in which the 3 epitopes were placed in different orientations relative to each other, with intervening double arginine residues.

We show here that the positioning of the epitopes within the construct affects the rate of peptide processing, which resulted in the 4 constructs displaying varying degrees of immunogenicity against each intended epitope. Furthermore, we show that the construct consisting of 3 immunogenic CTL epitopes from the Tax protein of HTLV-1 positioned in the most optimum orientation was effective in reducing viral replication in immunized animals that were challenged with Tax recombinant vaccinia virus. To our knowledge, this is the first demonstration of protection using human MHC-restricted epitopes derived from HTLV-1.


Peptide Synthesis

The 4 triple-epitope constructs were synthesized on a Milligen/Biosearch 9600 peptide synthesizer (PE Biosystems, Foster City, CA). The peptides were synthesized using a 4-methylbenzhydrylamine resin as the solid support (substitution of 0.54 mmol/g) as described previously.18 Briefly, the Fmoc/t-butyl synthetic method was used, employing 4-(hydroxymethyl) phenoxyacetic acid as the linker. After the final deprotection step, protecting groups and the peptide resin bond were cleaved with 90% trifluoroacetic acid, 5% anisole, 3% thioanisole, and 2% ethanedithiol. The crude peptides were purified by semipreparative reverse-phase high-performance liquid chromatography (RP-HPLC) using a Vydac C4 or C18 column (10 mm × 25 cm) and were >95% pure before immunization. The identity of the peptides was confirmed by matrix-assisted laser desorption ionization time-of-flight mass spectrometry (MS; m/z = 3899 for all 4 constructs) at the Mass Spectrometry and Proteonomics Facility of Campus Chemical Instrument Center (CCIC, Columbus, OH).

In Vitro Proteasomal Cleavage Analysis

Twenty micrograms of each of the 4 peptides were incubated separately with 2 μg of 20s immunoproteasomes purified from LCL 721 cells containing the interferon-γ (IFN-γ)–inducible subunits LMP2, LMP7, and MECL-1 (Immatics Biotechnologies, Germany) at 37°C for indicated time points in a total volume of 300 μL of assay buffer (1 mM of N-[2-hydroxyethyl]piperazine-N′-[2-ethesulfonic acid] (HEPES), pH 7.8; 2 mM of MgAc2; and 1 mM of dithiothreitol) as previously described.19,20 The digestion was stopped using 30 μL of trifluoroacetic acid (TFA) (Pierce), and the digested samples were stored at −20°C until analyzed.

Mass Spectrometric Analysis

The digested peptide samples were analyzed by liquid chromatography–nanospray tandem MS (nano-LC-MS). Briefly, capillary nano-LC-MS was performed on a separate Q-TOF2 system equipped with an orthogonal nanospray source from New Objective (Woburn, MA), which was operated in positive ion mode. The LC system was a Waters Alliance 2690 Separation Module (Waters, Milford, MA). The solvent A was water containing 50 mM of acetic acid, and the solvent B was acetonitrile. Ten microliters of each sample was first injected onto the trapping column and then washed with 50 mM of acetic acid. The injector port was switched to inject, and the peptides were eluted from the trap onto the column. A 10-cm, 50 mM ID BioBasic C18 column packed directly in the nanospray tip was used for chromatographic separations.Peptides were eluted directly off the column into the Q-TOF2 system using a gradient of 3% to 80% solvent B over 35 minutes, with a flow rate of 280 μL/min and a precolumn split of 500 nL/min. The total run time was 70 minutes. The nanospray capillary voltage was set at 2.8 kV, and the cone was set at 55 V. The source temperature was maintained at 100°C. Mass spectra were recorded using MassLynx 3.5 automatic switching functions. Mass spectra were acquired from mass of 300 to 2000 d every 1 second, with a resolution of 8000 formula weight half mass (FWHM). The intensities of the peaks in the mass spectra were used to estimate the relative amounts of peptides generated after digestion and expressed as a percentage of the total amount of peptide digested at a given time point.

Animals and Immunizations

Human leukocyte antigen–A*0201 monochain transgenic H-2Db β-2m double-knockout (HHD) mice, as described elsewhere,21 were used for these studies. To determine the immunogenicity of the various multiepitope constructs, groups of 4 to 5 HHD mice were immunized once with 100 μg of multiepitope peptide or with a mixture of 33 μg of each of the 3 individual epitopes combined with 140 μg of TT3, a promiscuous T-helper epitope from tetanus toxoid (residues 947–967) and 100 μg of adjuvant N-acetyl-glucosamine-3-acetyl-l-alanyl-d-isoglutamine (nor-MDP; Peninsula Laboratories, Belmont, CA)22 emulsified 50:50 in 4:1 squalene/Arlacel A (Sigma, St. Louis, MO). The emulsions were injected subcutaneously at the base of the tail. Eleven days after immunization, the mice were killed and the spleens were removed. For the challenge experiments, HHD mice were immunized twice, 3 weeks apart, before challenge.

Cell Culture

EL4/HHD cells and HeLa-HHD cells were kindly provided by Francois Lemonnier, Pasteur Institute, France, and have been described earlier.21 All cells were maintained in RPMI, 10% fetal calf serum (FCS), and 1% pen/strep (p/s). HHD-transfected cells were maintained in selection medium (RPMI, 10% FCS, 1% p/s, and 0.5% G418). Hu143TK cells are osteosarcoma cells and were maintained in Dulbecco modified Eagle medium (DMEM), 10% FCS, and 1% p/s. BSC-1 (monkey kidney cell line) also was maintained in DMEM, 10% FCS, and 1% p/s. Vaccinia virus constructs expressing the Tax protein or the influenza hemagglutinin protein have been described elsewhere.23

51Cr Release Assay

Eleven days after immunization, spleens were harvested from immunized mice and pooled. Single cell suspensions were prepared, and the splenocytes were stimulated in vitro with 10 μM of relevant nonamer peptide (Tax11–19, Tax178–186, or Tax233–241) or an irrelevant peptide derived from the proto-oncogene Her-2/neu (aa 402–410 [TLEEITGYL]) at 37°C for 1 week. Cells were cultured in RPMI 1640 with 10% FCS, 1% p/s, and 25 mM of HEPES buffer supplemented with 5% Rat-T-Stim without concanavalin A (BD Labware, Bedford, MA). One milliliter of culture medium was replaced every 48 hours. On day 6, the cells were harvested for use as effectors in a standard 4-hour 51Cr release assay using EL4-HHD cells as targets.

Cytolytic properties of the splenocytes from immunized mice were examined by the standard 51Cr release assay. For the vaccinia-infected targets, target cells were infected with the relevant vaccinia construct at an multiplicity of infection (MOI) of 50:1 overnight for 18 hours before the 51Cr release assay. Briefly, target cells were radiolabeled with Na251CrO4 (Perkin Elmer, Boston, MA) at 100 μCi per 106 cells for 1 hour. The target cells were also sensitized with 10 μΜ of relevant peptide or irrelevant peptide at the time of radiolabeling. The cells were washed 3 times (RPMI 1640, 10% FCS, and 1% p/s), and 5000 target cells were plated onto triplicate wells of 96-well round-bottom tissue culture plates and used at different effector-to-target (E/T) ratios as indicated. After 4 hours, the supernatants were collected and counted on a γ-counter (Beckman model 5500). Spontaneous and maximum (100%) release was determined from wells containing medium alone or 5% sodium dodecyl sulfate (SDS). Specific lysis was calculated in triplicate as follows: (Experimental Release − Spontaneous Release)/(100% − spontaneous) × 100.

Interferon-γ Release Assay

Splenocytes from immunized mice were stimulated in vitro for 6 days with splenocytes pulsed with 10 μM of relevant peptide. After 6 days, the splenocytes were harvested and further incubated with various target cells as indicated for 24 hours in RPMI, 10% FCS, and 1% p/s in a total volume of 200 μL. After 24 hours, the cultures were centrifuged and the supernatants were collected and stored at −20°C. The amount of IFN-γ released into the supernatant was estimated using a sandwich enzyme-linked immunosorbent assay (ELISA) per the manufacturer’s protocols (Pharmingen, CA). Spontaneous or background IFN-γ secretion in wells containing splenocytes and peptide unpulsed targets (media only) was subtracted from values obtained with specific peptide stimulation.

Plaque Assay for Viral Titers

BSC-1 cells were used as indicator cells. Briefly, serial 10-fold dilutions of the virus preparation were made in DMEM, 2% FCS, and 1% p/s. Two hundred microliters of the various dilutions was plated on confluent BSC-1 indicator cells in 6-well tissue culture plates. After 48 hours of incubation, the medium was aspirated out and the cells were stained with 0.1% crystal violet in 20% ethanol for 10 minutes. The plates were washed once with phosphate-buffered saline (PBS) and dried, and the plaques were counted to estimate the viral titer as plaque-forming units (pfu).

Challenge Studies

Ten days after the last immunization, HHD mice were challenged intraperitoneally with 5 × 106 pfu of recombinant vaccinia virus expressing the Tax protein of HTLV-1 (p40-VV) or the control recombinant vaccinia virus expressing the hemagglutinin protein from influenza virus (HA-VV). In some experiments, the mice were depleted of CD8+ T-cells by intraperitoneal injection of 100 μg of anti-CD8 antibody (YTS on days −4, −3, −1, 0, and +1, with day 0 being the day of challenge with vaccinia virus. Five days after challenge, the mice were killed and the ovaries were removed, freeze-thawed twice, homogenized, sonicated, and assayed for vaccinia virus titers by plating serial 10-fold dilutions on BSC-1 indicator cells as previously described.15 The log10 decrease in pfu in the ovary (pfu/ovary) was calculated as follows: the log10 of the ratio of the (viral titers of each of the 3 controls [mock-immunized mice, naive mice, or HA-VV–challenged mice immunized with the 236 peptide])/(the viral titers of mice immunized with the 236 peptide, immunized with the 236 peptide and depleted of CD8+ T-cells before challenge with p40-VV, or immunized with a mixture of the 3 individual epitopes of which the 236 peptide is composed). Each treatment group provided data representative of at least 8 mice.

Statistical Analysis

Statistical analyses compared the effect of various peptide vaccines on vaccinia titer (pfu/ovary). Because titer data were not normally distributed, nonparametric Kruskal-Wallis and Mann-Whitney U tests to compare groups with Bonferroni adjustment was used. Based on the number of tests performed a 2-sided level of significance of 0.005 was defined as statistically significant.


Design, Synthesis, and Characterization of the Multiepitope Cytotoxic T-Lymphocyte Peptide Constructs

The CTL epitopes Tax11–19 and Tax178–186 were previously evaluated for immunogenicity.16 We have identified another CTL epitope, Tax306–315, which elicited high cytolytic responses (unpublished observations) when immunized as a single epitope mixed with TT3.25 These epitopes were incorporated in novel multiepitope constructs as previously reported by us, in which the individual epitopes were synthesized colinearly with double-arginine spacers.16 To determine if the positioning of the individual epitopes in the multiepitope construct may cause differential processing by the proteasome and consequently affect the epitope immunogenicity, different variants of the multiepitope peptide construct were also synthesized. The orientation of the individual epitopes and their designations are depicted in Table 1.

Table 1
Table 1:
Variants of Multiepitope Constructs Synthesized to Determine the Effect of the Relative Positioning of Individual Epitopes on the Rate of Processing of the Construct by Immunoproteasomes

Immunoproteasomal Liberation of Individual Peptides from Multiepitope Constructs

Each of the 4 multiepitope CTL peptide constructs were digested using purified immunoproteasome. Three different time points of 12, 24, and 48 hours were used for the digestion. At the earliest time point of 12 hours of digestion, the 236 peptide orientation was successfully digested and all 3 individual epitopes could be detected by MS (Fig. 1A). Furthermore, the undigested peptide could not be detected at the earliest 12-hour time point, indicating that there was complete substrate turnover (data not shown). The relative abundance of each epitope calculated as described in the Methods section increased marginally at the 24-hour time point. By 48 hours, the relative abundance of all 3 individual epitopes in the digest had decreased substantially and a weak signal was detected. For peptide 362, the Tax11–19 epitope was observed in decreasing abundance at the 12-, 24-, and 48-hour time points. Tax178–186 epitope was observed only at 24 hours, and Tax306–315 was observed at 24 and 48 hours, although the signal was extremely weak (see Fig. 1B). Likewise, for peptide 326, Tax11–19 and Tax178–186 were observed in marginally increasing abundance at all time points, but Tax306–315 could only be detected at 24 and 48 hours, again at extremely low levels (see Fig. 1C). The peptide 632 construct had the slowest rate of digestion, and none of the epitopes could be detected initially at 12 hours (see Fig. 1D), with low to negligible substrate turnover (data not shown). At 24 and 48 hours, however, all 3 epitopes (Tax11–19, Tax178–186, and Tax306–315) could be detected, with complete substrate turnover comparable with the peptide 236 construct. Overall, in all 4 constructs, the relative abundance of Tax11–19 was the highest and that of Tax306–315 was the lowest. The peptide 236 construct had the highest rate of digestion.

A–D, Immunoproteasomal cleavage analysis of variants of multiepitope constructs. The 4 multiepitope constructs were digested and analyzed by capillary liquid chromatography–nanospray tandem mass spectrometry. The relative abundance of each epitope detected in the sample is expressed as a percentage of the total summed mass peak intensities of the digested peptide substrate at the indicated incubation times.

Immunogenicity Studies of Multiepitope Variants in Human Leukocyte Antigen–A2.1 Monochain Transgenic H-2Db β-2m Double-Knockout Transgenic Mice

The 4 variants of the multiepitope peptide construct were tested individually for their ability to elicit a CTL response in HHD transgenic mice. The cytolytic responses against relevant peptide-pulsed EL4/HHD target cells are shown in Figure 2. The Tax11–19 epitope was immunogenic as part of all 4 constructs, with the highest lysis seen in the peptide 236 and 632 constructs. The Tax178–186 epitope also induced cytolytic responses as a multiepitope construct, and the highest responses were again obtained in the peptide 236 construct, followed by the peptide 632 and 326 constructs. We found that although the Tax306–315 epitope was highly immunogenic in HHD mice when administered as a single epitope, the cytolytic responses observed when it was administered as part of a multiepitope construct were extremely weak to negligible. This was the case in all 4 multiepitope variants. Splenocytes stimulated in vitro from mice immunized with each construct failed to show any lysis when tested against target cells pulsed with an irrelevant epitope derived from the proto-oncogene Her-2/neu epitope (data not shown). Also, splenocytes from mice immunized with each of the multiepitope constructs expanded in vitro with the same irrelevant peptide failed to show any lysis against target cells pulsed with the irrelevant peptide, demonstrating antigen-specific priming of CTL. In summary, the 236 construct gave the most optimum responses against all individual epitopes as compared with the other multiepitope variants.

Cytolytic responses of 4 variants of multiepitope constructs. Splenocytes from groups of 4 to 5 human leukocyte antigen–A2.1 monochain transgenic H-2Db β-2m double-knockout (HHD) mice immunized once separately with each multiepitope construct or splenocytes from naive control mice were stimulated in vitro for 6 days with the relevant individual epitope as described in the Methods section. Cytolysis was tested in a standard 51Cr release assay against EL4/HHD cells pulsed with 10 μM of the indicated epitope. Data represent mean responses with standard error from 2 independent experiments.

The multiepitope constructs were also evaluated for their ability to induce the release of IFN-γ by stimulation of splenocytes. Freshly harvested splenocytes from HHD mice immunized with the various multiepitope constructs were stimulated for 6 days with the relevant epitope and then tested against epitope-pulsed EL4/HHD target cells for their ability to be activated to release IFN-γ. Splenocytes from the 236 construct–immunized mice were the most efficient in releasing IFN-γ, followed by those from the 632 construct (Fig. 3). In contrast, the 362 and 326 constructs elicited extremely low levels of IFN-γ against all individual epitopes. Unexpectedly, with the 236 construct, the levels of IFN-γ induced by splenocytes stimulated with the Tax306–315 epitope were also high (8000 pg/mL). This was in contrast to the other multiepitope constructs, which did not elicit a high response against this epitope. Thus, overall, the position of the different epitopes in the 236 construct was the most optimum in inducing immune responses in HHD mice. The data correlated with the peptide digestion studies, because cytolytic responses and IFN-γ secretion were best observed with the 236 peptide, followed by the 632 peptide; the 632 peptide was digested at the slowest rate, but the relative abundance of the epitopes generated was comparable with the 236 construct.

Cytokine release by multiepitope variants. Splenocytes from groups of 4 to 5 human leukocyte antigen–A*0201 monochain transgenic H-2Db β-2m double-knockout (HHD) mice immunized once separately with each multiepitope construct and stimulated in vitro for 6 days with each epitope were tested against relevant epitope-pulsed EL4/HHD target cells in a 24-hour assay. Interferon-γ (IFN-γ) released into the supernatant was quantitated using a sandwich enzyme-linked immunosorbent assay. Data represent mean values from 2 independent experiments after subtraction of background IFN-γ secretion obtained by stimulation with EL4/HHD cells without peptide.

Multiepitope Tax-Derived Peptide Immunization Leads to a Reduction of Virus Titers When Challenged With a Recombinant Vaccinia Virus Expressing the Tax Antigen and Is Dependent on CD8+ T-Cells

The 4 variants of the multiepitope vaccines were differentially immunogenic in the HHD mice as determined by functional cytolytic and cytokine release assays. Based on these results, we examined the protective efficacy of the multiepitope T-cell vaccine candidate against viral infection. The 236 multiepitope construct was used for all the challenge studies, because the highest cytolytic responses and cytokine release were observed by immunization with this construct. Because HTLV-1 does not infect mice, we used a surrogate model of p40-VV to challenge vaccinated HHD mice. HHD mice immunized with the 236 peptide construct were challenged with p40-VV. Immunized mice showed a greater than 3-log reduction in viral load as compared with mock-immunized (TT3 only) or naive control mice, which was statistically significant (P < 0.001; Table 2). Additionally, mice immunized with a mixture of the 3 individual peptides that comprise the 236 construct were only capable of reducing viral titers by 1.84 log when compared with mock-immunized p40-VV challenged mice, which was not statistically significant (P = 0.007). Furthermore, the reduction in viral titers from immunization with the 236 construct was specific to the Tax-expressing vaccinia construct. This was demonstrated by groups of 236 peptide construct–immunized mice that did not show any reduction in viral replication when challenged with the HA-VV (see Table 2).

Table 2
Table 2:
Effects of Immunization of HHD Mice With the 236 Peptide and Their Reliance on CD8+ T-Cells for the Reduction of Viral Titers Against Challenge With Recombinant Vaccinia Constructs p40-VV and HA-VV

To demonstrate the role of CD8+ T-cells in mediating the reduction in viral titers in the 236 peptide construct–vaccinated mice, we immunized HHD mice with the 236 peptide construct. Before challenge with p40-VV, the mice were injected with anti-CD8 antibodies to deplete the CD8+ T-cell population. As shown in Table 2, there was a less than 1 log reduction in viral replication in CD8+ T-cell–depleted mice as compared with the mock-immunized or naive groups of control mice. These data demonstrate the generation of CD8+ T-cell immunity in response to the 236 peptide construct.

Splenocytes from Multiepitope Peptide–Immunized Mice Recognize Vaccinia Virus–Infected Target Cells

To demonstrate antiviral mechanisms induced by 236 peptide immunization, splenocytes from 236 immunized mice were stimulated in vitro with the relevant epitopes and tested against target cells that were infected overnight with p40-VV or HA-VV. Tax11–19- and Tax178–186-stimulated splenocytes selectively killed p40-VV–infected targets with minimal effect on HA-VV–infected targets. Tax306–315-stimulated splenocytes did not show a cytolytic response (Fig. 4A). In addition, we examined the release of IFN-γ in the same splenocytes stimulated by p40-VV–infected target cells. In vitro Tax178–186 stimulated splenocytes secreted the highest levels of IFN-γ (670 pg/mL), followed by Tax11–19-stimulated splenocytes (298 pg/mL) and then Tax306–315-stimulated splenocytes (147 pg/mL) (see Fig. 4B).

Cytolysis of recombinant vaccinia virus expressing the Tax protein (p40-VV)–infected target cells. A, Splenocytes from 236 multiepitope immunized human leukocyte antigen–A*0201 monochain transgenic H-2Db β-2m double-knockout (HHD) mice expanded in vitro with individual epitopes for 6 days were tested for cytolysis against p40-VV–infected HeLa-HHD target cells at a 100:1 effector-to-target (E/T) ratio. Background lysis of irrelevant recombinant vaccinia virus expressing the hemagglutinin protein from influenza virus (HA-VV)–infected HeLa-HHD target cells is also shown for each group. B, Splenocytes from 236 immunized HHD mice expanded in vitro with individual epitopes were tested for IFNγ release against HeLa-HHD cells infected with p40-VV or irrelevant HA-VV control virus at a 100:1 E/T ratio. The level of IFN-γ released was quantitated as described in the Methods section. Data represent mean values from triplicate samples with standard error.


The development of prophylactic and therapeutic vaccine strategies to combat viral, bacterial, and parasitic infections and cancer is urgently required. Given that our understanding of the correlates of protection against HTLV-1 is still limited and not clearly defined, effective control of HTLV-1 infection requires not only a concerted approach involving multiple defense mechanisms (antibodies, cytotoxic T cells, and helper T cells) but antigens from different gene products (eg, Env, Tax, Rex, Gag). We have developed multivalent strategies for eliciting high-affinity antibodies to the env gp46 and gp21 proteins26,27 as well as multivalent strategies for eliciting CD8+ T-cell immunity16 to Tax gene products, in which we aimed at the delivery of multiple CTL epitopes simultaneously for priming multispecific CTL responses in a manner that closely mimics natural viral infection.

Recombinant vaccines such as vaccinia virus vectors expressing HTLV-1 proteins or whole-protein vaccines have shown various levels of protection in several HTLV-1 animal models.7,9–11 To date, however, there is no effective vaccine against HTLV-1 infection. Furthermore, prior immunity (because of natural exposure or other vaccinations) to the viral vector itself may limit its ability to serve as an effective vector for the delivery of recombinant genes from different infectious agents or cancer antigens28 or may interfere with the development of immunity to the recombinant antigen. Whole-protein vaccinations are largely undefined, and there is the disadvantage of epitope suppression. Epitope-based vaccines have the advantage of being defined and amenable to changes that may enhance the protective efficacy of vaccine formulations, because epitopes that elicit weak or deleterious immune responses may be selectively avoided. The potential of a DNA minigene/string of beads approach of displaying multiepitope peptides from a single or several different pathogens or different HLA has been demonstrated in different animal models of infection (small animal models).29–32 The formulation of multiepitope vaccines may be complicated by the phenomenon of immunodominance, however, in which only some epitopes within the construct are immunogenic. Although the reasons behind immunodominance are still a topic of intense investigation, multiple factors such as epitope processing and transport, binding affinity of the particular epitope for MHC, the available T-cell repertoire, and flanking residues may play an important role in dictating the response to the different epitopes.33–35 Hence, after selection of epitopes that bind with sufficiently high affinity for MHC class I, it is necessary to design multiepitope constructs that will be optimally processed such that the correct structural environment is present to induce strong CD8+ T-cell responses.

The multicatalytic proteasomes that are responsible for the generation of the correct C-terminal end of the CTL epitope are largely influenced by the flanking residues of the putative epitope.36 The proteasomal degradation system in concert with the Tap transporter system has been shown to limit the repertoire of peptides that bind to MHC class I because of inefficient processing. Conversely, the versatility of the proteasome proteolysis may also contribute to the multitude of overlapping peptide substrates.

The orientations of the same set of epitopes within a multiepitope construct would result in different peptide substrates that differ with respect to the flanking residues for each epitope. Therefore, we hypothesized that such constructs would be presented as novel peptide substrates to the 20s proteasome, which may result in differential processing. The resulting epitope turnover would influence the immunogenicity of the individual epitopes. In support of our hypothesis, we observed that indeed there were differences in the processing rates. Although all 4 constructs were immunogenic, they varied in their degree of immunogenicity.

This effect was especially pronounced in the activation of IFN-γ release by splenocytes from immunized mice. Only the 236 orientation construct successfully activated the release of IFN-γ against target cells pulsed with each of the 3 epitopes and, more importantly, against p40-VV–infected vaccinia virus. We speculate that because the 236 peptide was cleaved at a much higher rate than the other constructs, a relatively higher number of peptide-MHC complexes may have been expressed on the cell surface, which could be responsible for higher activation levels of antigen-specific T-cells. Additionally, these results show that there are different thresholds or requirements for lytic activity and the secretion of antiviral cytokines and that it is important to use multiple read-out assays to characterize cellular responses completely against potential vaccine candidates.37

Our results indicate that the requirements for proteolytic cleavage seem to be complex; therefore, it is currently difficult to make any generic conclusions for the design of the most optimum construct for a given set of epitopes. Hence, the choice of the orientation of epitopes in a multivalent construct must be empiric.

IFN-γ secretion is an important antiviral mechanism that may enhance antiviral potential while avoiding the destructive effects of lytic activity of induced CTLs against infected target cells. Epitopes that only induce IFN-γ secretion but not lysis of specific target cells have been previously reported. Such epitopes may be particularly useful for infections such as HTLV-1, where the primary cells infected are the CD4+ T-cells that are instrumental during initial priming of immune responses and for maintenance of memory.38,39 In this regard, we found that the levels of IFN-γ secreted by splenocytes from HHD mice that were immunized with the 236 construct were much higher than those obtained with splenocytes from HHD mice immunized with each individual epitope (data not shown).

This observation correlated with the challenge experiments in mice immunized with a mixture of individual epitopes, which resulted in a reduction of viral titers that was not statistically significant (see Table 2). It is possible that functionally different CTL clones are stimulated against the same epitope, which is dependent on the method of priming of T- cells by antigen-presenting cells.

The Tax protein is the viral transcriptional transactivator and is expressed early during HTLV-1 infection.40,41 It is also the immunodominant target antigen for the cellular immune response.42–44 In general, it has been observed that high levels of HTLV-1–specific CTLs are found in HAM/TSP patients and asymptomatic carriers. In contrast, the levels of CTLs are much lower in ATL patients.45–49 Furthermore, the prognosis of ATL patients is also poor because of the high resistance to chemotherapeutic interventions. These observations implicate the role of host immunity against HTLV-1 in the development of disease. In individuals who mount a brisk CTL response during the initial stages of natural infection, a large number of HTLV-1–infected cells are killed by Tax-specific CTLs before they can complete the replication of the mature HTLV-1 virions. Some cells do survive and proliferate, however. The CTLs in these individuals thus play a role in maintaining a low proviral load, and many of these individuals remain asymptomatic HTLV-1 carriers. In low CTL responders, the resultant high proviral load caused overstimulation of the CTL response, which may be responsible for inflammatory disorders such as HAM/TSP.50

The challenge model that we used in this study allowed us to evaluate the efficacy of human epitopes that have been previously identified in patients infected with HTLV-1. The Tax11–19 epitope has been widely studied with respect to its binding affinity and its possible role in the pathogenesis of HAM/TSP.51–53 However, there are other conflicting reports demonstrating that this epitope is protective and that HLA-A*0201–positive HTLV-1–infected individuals are less likely to develop disease.50,54 Likewise, immune responses against the Tax178–186 epitope were detected in peripheral blood of HTLV-1–infected individuals several years ago, and the epitope was also classified as a strong MHC class I binder.44,49 Furthermore, protection studies in athymic rats using recombinant Tax protein also demonstrate the applicability of Tax-based epitopes for therapeutic purposes.6,8,12 Like many other persistent viruses, HTLV-1 has evolved mechanisms for survival in the face of an active immune response. The protection studies described in this report confirm that the epitopes identified during natural infection are also immunogenic in an active immunization setting. Hence, these epitopes may find potential in a vaccination or therapeutic setting to amplify the existent cell-mediated response, which may aid in more efficient clearance of the virus especially during the early stages of disease. The use of multiepitope constructs would also help against the emergence of CTL escape mutants through selective pressures of single immunodominant epitopes.55 Such multiepitope vaccines may be used to incorporate several CTL epitopes from different gene products such as Gag to induce high-magnitude multispecific CTL responses.


The authors are grateful to Mirdad Kazanji for useful suggestions and critical reading of the manuscript.


1. Uchiyama T, Yodoi J, Sagawa K, et al. Adult T-cell leukemia: clinical and hematologic features of 16 cases. Blood. 1977;50:481–492.
2. Poiesz BJ, Ruscetti FW, Gazdar AF, et al. Detection and isolation of type C retrovirus particles from fresh and cultured lymphocytes of a patient with cutaneous T-cell lymphoma. Proc Natl Acad Sci USA. 1980;77:7415–7419.
3. Osame M, Usuku K, Izumo S, et al. HTLV-I associated myelopathy, a new clinical entity. Lancet. 1986;1:1031–1032.
4. Rock KL, Goldberg AL. Degradation of cell proteins and the generation of MHC class I-presented peptides. Annu Rev Immunol. 1999;17:739–779.
5. Mo AX, van Lelyveld SF, Craiu A, et al. Sequences that flank subdominant and cryptic epitopes influence the proteolytic generation of MHC class I-presented peptides. J Immunol. 2000;164:4003–4010.
6. Hanabuchi S, Ohashi T, Koya Y, et al. Development of human T-cell leukemia virus type 1-transformed tumors in rats following suppression of T-cell immunity by CD80 and CD86 blockade. J Virol. 2000;74:428–435.
7. Hakoda E, Machida H, Tanaka Y, et al. Vaccination of rabbits with recombinant vaccinia virus carrying the envelope gene of human T-cell lymphotropic virus type I. Int J Cancer. 1995;60:567–570.
8. Hanabuchi S, Ohashi T, Koya Y, et al. Regression of human T-cell leukemia virus type I (HTLV-I)-associated lymphomas in a rat model: peptide-induced T-cell immunity. J Natl Cancer Inst. 2001;93:1775–1783.
9. Ibuki K, Funahashi SI, Yamamoto H, et al. Long-term persistence of protective immunity in cynomolgus monkeys immunized with a recombinant vaccinia virus expressing the human T cell leukaemia virus type I envelope gene. J Gen Virol. 1997;78:147–152.
10. Nakamura H, Hayami M, Ohta Y, et al. Protection of cynomolgus monkeys against infection by human T-cell leukemia virus type-I by immunization with viral env gene products produced in Escherichia coli. Int J Cancer. 1987;40:403–407.
11. Shida H, Tochikura T, Sato T, et al. Effect of the recombinant vaccinia viruses that express HTLV-I envelope gene on HTLV-I infection. EMBO J. 1987;6:3379–3384.
12. Ohashi T, Hanabuchi S, Kato H, et al. Prevention of adult T-cell leukemia-like lymphoproliferative disease in rats by adoptively transferred T cells from a donor immunized with human T-cell leukemia virus type 1 Tax-coding DNA vaccine. J Virol. 2000;74:9610–9616.
13. Kiszka I, Kmieciak D, Gzyl J, et al. Effect of the V3 loop deletion of envelope glycoprotein on cellular responses and protection against challenge with recombinant vaccinia virus expressing gp160 of primary human immunodeficiency virus type 1 isolates. J Virol. 2002;76:4222–4232.
14. Arribillaga L, de Cerio AL, Sarobe P, et al. Vaccination with an adenoviral vector encoding hepatitis C virus (HCV) NS3 protein protects against infection with HCV-recombinant vaccinia virus. Vaccine. 2002;21:202–210.
15. Ahlers JD, Belyakov IM, Matsui S, et al. Mechanisms of cytokine synergy essential for vaccine protection against viral challenge. Int Immunol. 2001;13:897–908.
16. Sundaram R, Sun Y, Walker CM, et al. A novel multivalent human CTL peptide construct elicits robust cellular immune responses in HLA-A*0201 transgenic mice: implications for HTLV-1 vaccine design. Vaccine. 2003;21:2767–2781.
17. Livingston B, Crimi C, Newman M, et al. A rational strategy to design multiepitope immunogens based on multiple Th lymphocyte epitopes. J Immunol. 2002;168:5499–5506.
18. Dakappagari NK, Pyles J, Parihar R, et al. A chimeric multi-human epidermal growth factor receptor-2 B cell epitope peptide vaccine mediates superior antitumor responses. J Immunol. 2003;170:4242–4253.
19. Theobald M, Ruppert T, Kuckelkorn U, et al. The sequence alteration associated with a mutational hotspot in p53 protects cells from lysis by cytotoxic T lymphocytes specific for a flanking peptide epitope. J Exp Med. 1998;188:1017–1028.
20. Kessler JH, Beekman NJ, Bres-Vloemans SA, et al. Efficient identification of novel HLA-A(*)0201-presented cytotoxic T lymphocyte epitopes in the widely expressed tumor antigen PRAME by proteasome-mediated digestion analysis. J Exp Med. 2001;193:73–88.
21. Pascolo S, Bervas N, Ure JM, et al. HLA-A2.1-restricted education and cytolytic activity of CD8(+) T lymphocytes from beta2 microglobulin (beta2m) HLA-A2.1 monochain transgenic H-2Db beta2m double knockout mice. J Exp Med. 1997;185:2043–2051.
22. Jones WR, Bradley J, Judd SJ, et al. Phase I clinical trial of a World Health Organisation birth control vaccine. Lancet. 1988;1:1295–1298.
23. Siomi H, Shida H, Nam SH, et al. Sequence requirements for nucleolar localization of human T cell leukemia virus type I pX protein, which regulates viral RNA processing. Cell. 1988;55:197–209.
24. Cobbold SP, Jayasuriya A, Nash A, et al. Therapy with monoclonal antibodies by elimination of T-cell subsets in vivo. Nature. 1984;312:548–551.
25. Ho PC, Mutch DA, Winkel KD, et al. Identification of two promiscuous T cell epitopes from tetanus toxin. Eur J Immunol. 1990;20:477–483.
26. Frangione-Beebe M, Albrecht B, Dakappagari N, et al. Enhanced immunogenicity of a conformational epitope of human T-lymphotropic virus type 1 using a novel chimeric peptide. Vaccine. 2000;19:1068–1081.
27. Lairmore MD, DiGeorge AM, Conrad SF, et al. Human T-lymphotropic virus type 1 peptides in chimeric and multivalent constructs with promiscuous T-cell epitopes enhance immunogenicity and overcome genetic restriction. J Virol. 1995;69:6077–6089.
28. Yang ZY, Wyatt LS, Kong WP, et al. Overcoming immunity to a viral vaccine by DNA priming before vector boosting. J Virol. 2003;77:799–803.
29. Thomson SA, Khanna R, Gardner J, et al. Minimal epitopes expressed in a recombinant polyepitope protein are processed and presented to CD8+ cytotoxic T cells: implications for vaccine design. Proc Natl Acad Sci USA. 1995;92:5845–5849.
30. Hanke T, Schneider J, Gilbert SC, et al. DNA multi-CTL epitope vaccines for HIV and Plasmodium falciparum: immunogenicity in mice. Vaccine. 1998;16:426–435.
31. Hanke T, Neumann VC, Blanchard TJ, et al. Effective induction of HIV-specific CTL by multi-epitope using gene gun in a combined vaccination regime. Vaccine. 1999;17:589–596.
32. Velders MP, Weijzen S, Eiben GL, et al. Defined flanking spacers and enhanced proteolysis is essential for eradication of established tumors by an epitope string DNA vaccine. J Immunol. 2001;166:5366–5373.
33. Yewdell JW, Bennink JR. Cut and trim: generating MHC class I peptide ligands. Curr Opin Immunol. 2001;13:13–18.
34. Pamer E, Cresswell P. Mechanisms of MHC class I–restricted antigen processing. Annu Rev Immunol. 1998;16:323–358.
35. Chen W, Norbury CC, Cho Y, et al. Immunoproteasomes shape immunodominance hierarchies of antiviral CD8(+) T cells at the levels of T cell repertoire and presentation of viral antigens. J Exp Med. 2001;193:1319–1326.
36. York IA, Goldberg AL, Mo XY, et al. Proteolysis and class I major histocompatibility complex antigen presentation. Immunol Rev. 1999;172:49–66.
37. Alexander J, Del Guercio MF, Fikes JD, et al. Recognition of a novel naturally processed, A2 restricted, HCV-NS4 epitope triggers IFN-gamma release in absence of detectable cytopathicity. Hum Immunol. 1998;59:776–782.
38. Borrow P, Tough DF, Eto D, et al. CD40 ligand-mediated interactions are involved in the generation of memory CD8(+) cytotoxic T lymphocytes (CTL) but are not required for the maintenance of CTL memory following virus infection. J Virol. 1998;72:7440–7449.
39. Altfeld M, Rosenberg ES. The role of CD4(+) T helper cells in the cytotoxic T lymphocyte response to HIV-1. Curr Opin Immunol. 2000;12:375–380.
40. Sun SC, Elwood J, Beraud C, et al. Human T-cell leukemia virus type I Tax activation of NF-kappa B/Rel involves phosphorylation and degradation of I kappa B alpha and RelA (p65)-mediated induction of the c-rel gene. Mol Cell Biol. 1994;14:7377–7384.
41. McKinsey TA, Brockman JA, Scherer DC, et al. Inactivation of IkappaBbeta by the tax protein of human T-cell leukemia virus type 1: a potential mechanism for constitutive induction of NF-kappaB. Mol Cell Biol. 1996;16:2083–2090.
42. Parker CE, Nightingale S, Taylor GP, et al. Circulating anti-Tax cytotoxic T lymphocytes from human T-cell leukemia virus type I-infected people, with and without tropical spastic paraparesis, recognize multiple epitopes simultaneously. J Virol. 1994;68:2860–2868.
43. Koenig S, Woods RM, Brewah YA, et al. Characterization of MHC class I restricted cytotoxic T cell responses to tax in HTLV-1 infected patients with neurologic disease. J Immunol. 1993;151:3874–3883.
44. Pique C, Connan F, Levilain JP, et al. Among all human T-cell leukemia virus type 1 proteins, tax, polymerase, and envelope proteins are predicted as preferential targets for the HLA-A2-restricted cytotoxic T-cell response. J Virol. 1996;70:4919–4926.
45. Kannagi M, Matsushita S, Harada S. Expression of the target antigen for cytotoxic T lymphocytes on adult T-cell-leukemia cells. Int J Cancer. 1993;54:582–588.
46. Kannagi M, Matsushita S, Shida H, et al. Cytotoxic T cell response and expression of the target antigen in HTLV-I infection. Leukemia. 1994;8(Suppl 1):S54–S59.
47. Kannagi M, Sugamura K, Sato H, et al. Establishment of human cytotoxic T cell lines specific for human adult T cell leukemia virus-bearing cells. J Immunol. 1983;130:2942–2946.
48. Jacobson S, Shida H, McFarlin DE, et al. Circulating CD8+ cytotoxic T lymphocytes specific for HTLV-I pX in patients with HTLV-I associated neurological disease. Nature. 1990;348:245–248.
49. Parker CE, Daenke S, Nightingale S, et al. Activated, HTLV-1-specific cytotoxic T-lymphocytes are found in healthy seropositives as well as in patients with tropical spastic paraparesis. Virology. 1992;188:628–636.
50. Bangham CR. The immune response to HTLV-I. Curr Opin Immunol. 2000;12:397–402.
51. Jacobson S. Cellular immune responses to HTLV-I: immunopathogenic role in HTLV-I–associated neurologic disease. J Acquir Immune Defic Syndr Hum Retrovirol. 1996;13(Suppl 1):S100–S106.
52. Jacobson S, Krichavsky M, Flerlage N, et al. Immunopathogenesis of HTLV-I associated neurologic disease: massive latent HTLV-I infection in bone marrow of HAM/TSP patients. Leukemia. 1997;11(Suppl 3):73–75.
53. Jacobson S, McFarlin DE, Robinson S, et al. HTLV-I-specific cytotoxic T lymphocytes in the cerebrospinal fluid of patients with HTLV-I-associated neurological disease. Ann Neurol. 1992;32:651–657.
54. Jeffery KJ, Siddiqui AA, Bunce M, et al. The influence of HLA class I alleles and heterozygosity on the outcome of human T cell lymphotropic virus type I infection. J Immunol. 2000;165:7278–7284.
55. Furukawa Y, Kubota R, Tara M, et al. Existence of escape mutant in HTLV-I tax during the development of adult T-cell leukemia. Blood. 2001;97:987–993.

human T-lymphotropic virus type 1 cytotoxic T-lymphocyte epitopes; multivalent peptide; antigen presentation/processing; recombinant Tax vaccinia virus

© 2004 Lippincott Williams & Wilkins, Inc.