Novel CD8+ cytotoxic T cell epitopes in bovine leukemia virus with cattle

Vaccine 33 (2015) 7194–7202

Contents lists available at ScienceDirect

Vaccine journal homepage: www.elsevier.com/locate/vaccine

Novel CD8+ cytotoxic T cell epitopes in bovine leukemia virus with cattle Lanlan Bai a,b , Shin-nosuke Takeshima a , Emiko Isogai b , Junko Kohara c , Yoko Aida a,∗ a

Viral Infectious Diseases Unit, RIKEN, Wako, Saitama 351-0198, Japan Laboratory of Animal Microbiology, Department of Microbial Biotechnology, Graduate School of Agricultural Science, Tohoku University, Sendai, Miyagi 981-8555, Japan c Animal Research Center, Hokkaido Research Organization, Shintoku, Hokkaido 081-0038, Japan b

a r t i c l e

i n f o

Article history: Received 28 August 2015 Received in revised form 27 October 2015 Accepted 29 October 2015 Available online 6 November 2015 Keywords: Bovine leukemia virus CD8+ CTL epitope mapping Experimentally infected calves Bovine MHC class I CD8+ T cell proliferation Cytotoxicity

a b s t r a c t Bovine leukemia virus (BLV) is associated with enzootic bovine leukosis and is closely related to human T cell leukemia virus (HTLV). The cytotoxic T lymphocyte (CTL) plays a key role in suppressing the progression of disease caused by BLV. T and B cell epitopes in BLV have been studied, but CD8+ CTL epitopes remain poorly understood. We used a library of 115 synthetic peptides covering the entirety of the Env proteins (gp51 and gp30), the Gag proteins (p15, p24, and p12), and the Tax protein of BLV to identify 11 novel CD8+ T cell epitopes (gp51N5, gp51N11, gp51N12, gp30N5, gp30N6, gp30N8, gp30N16, tax16, tax18, tax19, and tax20) in four calves experimentally infected with BLV. The number of CD8+ T cell epitopes that could be identified in each calf correlated with the BLV proviral load. Interestingly, among the 11 epitopes identified, only gp51N11 was capable of inducing CD8+ T cell-mediated cytotoxicity in all four calves, but it is not a suitable vaccine target because it shows a high degree of polymorphism according to the Wu–Kabat variability index. By contrast, no CTL epitopes were identified from the Gag structural protein. In addition, several epitopes were obtained from gp30 and Tax, indicating that cellular immunity against BLV is strongly targeted to these proteins. CD8+ CTL epitopes from gp30 and Tax were less polymorphic than epitopes from. Indeed, peptides tax16, tax18, tax19, and tax20 include a leucinerich activation domain that encompasses a transcriptional activation domain, and the gp30N16 peptide contains a proline-rich region that interacts with a protein tyrosine phosphatase SHP1 to regulate B cell activation. Moreover, at least one CD8+ CTL epitope derived from gp30 was identified in each of the four calves. These results indicate that BLV gp30 may be the best candidate for the development of a BLV vaccine. © 2015 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

1. Introduction Bovine leukemia virus (BLV) is a member of the family Retroviridae that the causative agent of enzootic bovine leukosis (EBL), which is the most common neoplastic disease of cattle and is closely related to human T cell leukemia virus (HTLV) [1,2]. BLV expresses the gag, pol, and env genes, which are both essential and common to retroviruses and are necessary for the production of infectious virions [3]. The gag gene of BLV is translated as the precursor Pr70 Gag and is processed into three mature proteins: p15 is the matrix protein, p24 is the most abundant capsid protein, and p12 is the nucleocapsid protein [3,4]. The BLV env gene encodes a mature surface glycoprotein (gp51) and a transmembrane protein (gp30) [3].

∗ Corresponding author. Tel.: +81 48 462 4408; fax: +81 48 462 4399. E-mail address: [email protected] (Y. Aida).

gp51 and gp30 are directly involved in infectivity events such as receptor recognition and membrane fusion [5]. Unlike other retroviruses, BLV encodes at least two regulatory proteins, Tax and Rex. Tax is a key contributor to the oncogenic potential, as well as a key protein involved in the replication of the virus [1,2,6,7]. The CD8+ cytotoxic T lymphocytes (CTLs) mediate cellular immunity by killing infected host cells and by producing cytokines that can inhibit viral replication [8–10]. The major histocompatibility complex (MHC) system in cattle is known as the bovine leukocyte antigen (BoLA) system, and is strongly involved in the subclinical progression of BLV infection [10,11]. MHC molecules are grouped into two classes (I and II) that are highly polymorphic. The ␣1 and ␣2 domains of MHC class I form a peptide binding groove that accommodates antigen-derived peptides and presents them to CD8+ T cells [10–12]. The binding of peptides to MHC class I proteins is the most selective event in antigen presentation to CD8+ T cells [13].

http://dx.doi.org/10.1016/j.vaccine.2015.10.128 0264-410X/© 2015 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4. 0/).

L. Bai et al. / Vaccine 33 (2015) 7194–7202

Cellular immunity is induced during natural BLV infection and plays a role in protecting hosts from BLV infection [9,14–16]. Previous studies postulated that cellular immunity against BLV antigens contributes to the suppression of BLV replication, thereby leading to the delay in disease progression [9,14–22]. However, in these previous studies, only a small number of peptides corresponding to selected regions of the BLV Env and Tax proteins were used to characterize CD4+ and CD8+ T cell epitopes for BLV. Here, we identified CD8+ T cell-restricted epitopes for BLV antigens as follows: (i) First, a library of 115 overlapping peptides covering the entirety of the Env proteins (gp51 and gp30), the Gag proteins (p15, p24, and p12), and the Tax protein was generated; (ii) Four calves with defined BoLA class I genotypes were experimentally infected with the amount of BLV taken from the same individual at the same time; (iii) Mapping of the CD8+ T cell epitopes was performed by the water-soluble tetrazolium salt (WST-8) assay using the synthetic peptide library and CD8+ T cells were took from the four BLV-infected calves; (iv) CD8+ T cell epitopes that induced

7195

proliferation in the initial screening were tested for their ability to induce CD8+ T cell cytotoxicity in a non-radioactive cytotoxicity assay; (v) The conservation of amino acids in CD8+ CTL epitopes was analyzed by the Wu–Kabat variability index; and (vi) the expression of these epitopes by cattle with different BLV proviral loads is described. These results may help to select the best candidates for synthetic peptide vaccine against BLV infection. 2. Materials and methods 2.1. Synthetic peptides A series of 20-mer peptides overlapping by 10 amino acids were designed based on the reported sequences of each mature peptide encoded and expressed by genes in env, gag, and tax. All of peptides were synthesized and purified to >70% purity (Sigma). These peptides are divided into 23 pools that were suspended in 80% dimethylsulfoxide (DMSO) (Fig. 1).

Fig. 1. Synthetic peptides of 20-mer overlapping peptides corresponding to the Gag (p15, p24, and p12), Env (gp51 and gp30) and Tax proteins. These overlapping peptides were made against each mature peptide encoded and expressed by genes in env, gag, and tax, which sequences of BLV Gag (GenBank accession no. LC057268), Env (GenBank accession no. EF600696), and Tax (GenBank accession no. EF600696) were reported. (A) Schematic representation of BLV genome organization and viral proteins. (B) The pool, name, sequence and amino acid positions of peptide are shown in column (left to right). The pool contains five individual peptides. a 15 amino acids in length; b 19 amino acids in length; and c 18 amino acids in length.

7196

L. Bai et al. / Vaccine 33 (2015) 7194–7202

2.2. Experimental animals Four BLV-negative Japanese Black calves (6–8-months-old) were experimentally challenged intravenously with blood containing a proviral load of 4 × 107 proviral copies per 105 cells, as determined by the BLV-coordination of common motifs-quantitative real-time polymerase chain reaction-2 (BLVCoCoMo-qPCR-2) assay from BLV-infected Holstein-Friesian cattle (Table 1). This study was approved by the Animal Ethical Committee and the Animal Care and Use Committee of the Animal Research Center, Hokkaido Research Organization (approval number 1302). 2.3. Amplification, cloning, and DNA sequencing of BoLA class I RNAs were extracted from peripheral blood mononuclear cells (PBMCs) obtained from each of experimantal calves and were reverse-transcribed into cDNA to obtain BoLA class I heavy chain cDNA clones. cDNAs were amplified by PCR using primers BoLA 1F (5 -ATGCGAGTYATGGGGCCGCGAGCCC-3 ) and BoLA 1R (5 -TCAMMCTTTAGGAACYGTGAGA-3 ). The PCR products were subcloned into the pGEM-T easy vector (Promega). The sequences were analyzed using softwares, Sequencher ver. 4.0.5 (Gene Codes), and Genetyx ver. 10, software (Genetyx). 2.4. Measurement of the BLV proviral load and anti-BLV antibodies The BLV proviral load was measured using the BLV-CoCoMoqPCR-2 method [23–26]. An anti-BLV antibody ELISA kit (JNC) was used according to the manufacturer’s instructions. 2.5. Preparation of PBMCs and CD8+ T lymphocytes PBMCs were separated according to the method of Miyasaka and Trunka [27]. CD8+ T cells were purified using the 7C2B monoclonal antibody (MAb) (mouse anti-bovine CD8; VMRD) and anti-mouse IgG MAb by the MACS System (Miltenyi Biotech). The CD8+ T lymphocyte-depleted PBMCs were used as the CD8− T cells in this study.

(Sigma) for 60 min. APCs (2 × 106 cells/ml) and CD8+ T cells (2 × 106 cells/ml) were mixed in the presence of either peptide pool containing individual peptides at 20 ␮M, 20 ␮M positive peptide, or 80% DMSO (negative control) at once, and were co-incubated at 37 ◦ C in a total volume of 110 ␮l of cell medium. After 109 h of incubation, 10 ␮l of WST-8 solution (Dojindo Molecular Technologies) was added to each well, and then the cells were incubated for 4 h. The microplate was read at an optical density (OD) of 450 nm. 2.7. CytoTox96 non-radioactive cytotoxicity assay The CD8+ T cells cytotoxicity was performed by non-radioactive cytotoxicity assay. It is a colorimetric alternative to 51 Cr release cytotoxicity assays. It quantitatively measures lactate dehydrogenase (LDH), a stable cytosolic enzyme that is released upon cell lysis. Released LDH in culture supernatant is measured with a 30 min coupled enzymatic assay, which results in the conversion of a tetrazolium salt (INT) into a red formazan product. The amount of color formed is proportional to the number of lysed cells. Effector cells (CD8+ T cells; 1 × 107 cells/50 ␮l or 2 × 107 cells/50 ␮l) were plated in the experimental wells to measure spontaneous LDH release by effector cells. Meanwhile, different concentrations of peptide were added to the experimental wells, and target cells (PBMCs; 1 × 106 cells/50 ␮l) were added to the experimental wells and the wells measuring minimum (spontaneous) and maximum LDH release and mixed at once. The microplate was incubated for 48 h that the effector cells were sufficiently activated by peptide at 37 ◦ C. Ten ␮l of Lysis solution was added to the maximum LDH release wells 45 min prior to the end of incubation period. After incubation, 50 ␮l of supernatant was transferred from each well into a fresh microplate for the enzymatic assay. Next, 50 ␮l of the reconstituted substrate mix was added to each well and incubated at room temperature for 30 min. Finally, 50 ␮l of stop solution was added and was recorded the absorbance at 492 nm after 45 min incubation. The percent cytotoxicity was calculated as follows: Percent cytotoxicity =

2.6. WST-8 assay

experimental − effector spontaneous − target spontaneous target maximum − target spontaneous × 100

The cell proliferation response against peptides was performed by WST-8 assay [44]. It is a determination of cell viability in cell proliferation that was reduced by dehydrogenase activities in cells to give a yellow-color formazan dye, which was soluble in the tissue culture media. The amount of the formazan dye, generated by the activities of dehydrogenases in cells, was directly proportional to the number of living cells. Antigen-presenting cells (APCs) were prepared by treating CD8− T cells with 50 ␮g/ml mitomycin C

2.8. Statistical analysis The conservation of epitopes was analyzed by Wu–Kabat variability index [28]. The function program in Microsoft Excel was used to conduct F-tests and t-tests for all results of cytotoxicity.

Table 1 Japanese black calves examined in the studya . Animal no.

Proviral loadb (CoCoMo-qPCR-2)

Breed

Age (month)

Sex

WBCsc (cells/␮l)

PBLsd (cells/␮l)

BoLA class 1e Allele

Antibodies to BLVf

A-JB B-JB C-JB D-JB

72 58,501 67,729 76,839

Japanese black Japanese black Japanese black Japanese black

6 7 6 8

♀ ♂ ♀ ♂

7480 13400 13500 14100

4520 8650 9250 8820

BoLA-N*02604 BoLA-N*02702 BoLA-N*02604 BoLA-N*02701

+ + + +

a

Blood samples (used for DNA and serum isolation) were collected for 10 weeks after the first inoculation. Proviral load (expressed as the number of copies of provirus per 105 peripheral blood mononuclear cells) was evaluated using CoCoMo-qPCR-2. c White blood cells (WBCs) were measured at a single time point in all calves during the experiment. d Peripheral blood lymphocytes (PBLs) were measured at a single time point in all calves during the experiment. e BoLA class1 alleles are according to the BoLA nomenclature committee of the IPD–MHC database, calves A-JB and C-JB carried allele (GenBank accession no. LC055984) which is identical to the previously reported allele, BoLA-N*02604, D-JB carried a novel allele, BoLA-N*02701 (GenBank accession no. LC055985), and B-JB carried allele (GenBank accession no. LC055986) which is identical to the previously reported allele, BoLA-N*02702. The amino acid sequences of BoLA class were determined from cDNA clone sequences using the pGEM® -T Easy Vector System. The allele was considered one of the predominantly expressed types in each experiment animal. f ELISA (enzyme-linked immunosorbent assay) to detect anti-BLV antibodies was performed using the BLV ELISA kit + BLV positive serum. b

L. Bai et al. / Vaccine 33 (2015) 7194–7202

3. Results 3.1. Preparation of synthetic peptides and BLV-infected experimental calves To characterize CD8+ T cell epitopes for BLV, previous studies [15,20,22,29] used a small number of peptides derived from portions of the Env and Tax proteins. To conduct a more extensive screen for CTL epitopes, we prepared a library of 115 synthetic peptides derived from the entire Env, Gag, and Tax proteins (Fig. 1). Mapping of CD8+ T cell epitopes was previously performed using lymphocytes from mice, sheep and cattle [15,20,22,29]. Here, four calves of the same age were experimentally infected with the same amount of BLV taken from a single individual infected cow (Table 1). In addition, the BoLA class I alleles of experimental calves were determined. According to the BoLA nomenclature committee of the IPD–MHC database, calves A-JB and C-JB carried allele, BoLAN*02604, B-JB carried allele, BoLA-N*02702, and D-JB carried a novel allele, BoLA-N*02701. The proviral loads of infected calves were estimated by BLV-CoCoMo-qPCR-2 [23–26]. These infected calves were proviral loads ranged from 72 to 76,839 copies per 105 cells (Table 1). 3.2. Proliferation of CD8+ T cells from BLV-infected calves CD8+ T cells were stimulated with the peptide pools that individual peptide at 20 ␮M and proliferations were measured by WST-8 assay. APCs were prepared by the CD8+ T cell-depleted PBMCs that treating with mitomycin C for 60 min to inhibit cells proliferation. Fig. 2 shows that peptide pools 9, 10, 11, 13, 14, 15, and 21 induced higher CD8+ T cell proliferation for calf A-JB; pools 8, 9, 11, 14, and 21 induced higher CD8+ T cell proliferation for calf B-JB; pools 9, 11, 14, 20, and 21 induced higher CD8+ T cell proliferation for calf C-JB; and pools 9, 13, 14, 15, and 21 induced higher CD8+ T cell proliferation for calf D-JB. To further map the CD8+ T cell epitopes recognized by experimental calves, CD8+ T cell proliferative responses to the each peptide within positive peptide pools were examined in the

7197

WST-8 assay (Fig. 3). On this second screening, only two peptides, gp30N16 and gp51N11, induced particularly high CD8+ T cell proliferation in calf A-JB, which had 72 copies BLV proviral load per 105 cells (Fig. 3A). By contrast, the three other calves with higher proviral loads than A-JB responded to a greater number of peptides. In calf C-JB, which had 67,729 copies BLV proviral load per 105 cells significant CD8+ T cell proliferation was observed in response to stimulated with five peptides (gp30N6, gp30N8, gp51N11, gp51N12, and tax19) (Fig. 3C). Interestingly, calf D-JB which proviral load was 76,839 copies per 105 cells, showed CD8+ T cell proliferative responses to the largest number (seven) of peptides (gp30N6, gp51N5, gp51N11, gp51N12, tax16, tax18, and tax20) (Fig. 3D). Peptides gp30N5, gp30N6, gp30N16, and gp51N11 increased the CD8+ T cell proliferation for calf B-JB, which proviral load was 58,501 copies per 105 cells (Fig. 3B). The number of CD8+ T cell epitopes was correlated with the proviral load in BLV-infected calves. In addition, all four calves responded to peptide gp51N11, indicating that this peptide was the strongest at inducing CD8+ T cell proliferation. Notably, no epitopes were found from Gag structural proteins. 3.3. Non-radioactive cytotoxicity assay To identify the immunodominant epitope, all 11 peptides that induced a CD8+ T cell proliferative response were tested for their ability to induce cytotoxicity. First, the optimal conditions for cytotoxicity assay were determined for using autologous PBMCs as supplemental APCs, and the optimal ratio of effector cells and target cells was determined as 10:1 for calf A-JB and 20:1 for the other three calves when using 5 × 104 PBMCs per well (data not shown). Next, non-radioactive cytotoxicity assays were carried out by incubating APCs with CD8+ T cells in the presence of different concentrations (0, 0.5, 5, or 20 ␮M) of peptides. Fig. 4A shows the results obtained from calf A-JB, gp30N16 and gp51N11 induced cytotoxic activity in a dose-dependent manner. By contrast, in calf C-JB, all five peptides significantly increased the CTL activity compared to the negative control, but none displayed dose-dependent

Fig. 2. CD8+ T cell proliferative responses to 23 peptide pools. CD8+ T cells were separated from PBMCs from four BLV-infected calves [A-JB (A), B-JB (B), C-JB (C), and D-JB (D)] and used as effector cells. CD8− T cells were treated with 50 ␮g/ml mitomycin C for 60 min at 37 ◦ C and used as APCs. CD8+ T cells (1 × 105 cells/50 ␮l) and CD8− T cells (1 × 105 cells/50 ␮l) were treated with 80% DMSO (negative control) or with 23 peptide pools, each containing five individual peptides at 20 ␮M of each peptide. After incubation for 109 h at 37 ◦ C, the CD8+ T cell proliferation responses were examined in a WST-8 assay. The bars represent the mean ± standard deviation (S.D.) of triplicate wells.

7198

L. Bai et al. / Vaccine 33 (2015) 7194–7202

Fig. 3. CD8+ T cell proliferative responses to individual peptides within positive peptide pools. CD8+ T cells (effector cells; 1 × 105 cells/50 ␮l) from four BLV-infected calves [A-JB (A), B-JB (B), C-JB (C), and D-JB (D)] and CD8− T cells (APCs; 1 × 105 cells/50 ␮l) were treated with mitomycin C and incubated with 80% DMSO (negative control) or 20 ␮M of each peptide from pools 9, 10, 11, 13, 14, 15, and 21 for calf A-JB, pools 8, 9, 11, 14, and 21 for calf B-JB, pools 9, 11, 14, 20, and 21 for calf C-JB, and pools 9, 13, 14, 15, and 21 for calf D-JB. The cells were treated with peptide for 109 h at 37 ◦ C and the CD8+ T cell proliferation responses were examined in a WST-8 assay. The bars represent the mean ± standard deviation (S.D.) of triplicate wells.

activity (Fig. 4C); a concentration of 0.5 ␮M, which was the lowest concentration tested for each peptide, induced near-maximal cytotoxicity. In contrast to A-JB and C-JB, D-JB showed differing patterns of CTL activity depending on the peptides (Fig. 4). However,

CTL lysed target cells in response to gp51N11 only at the highest concentration tested (20 ␮M). On the other hand, the tax16, tax18, and tax20 peptides induced cytotoxicity in a dose-dependent manner in D-JB, while gp51N12 had little effect. As shown in Fig. 4B,

L. Bai et al. / Vaccine 33 (2015) 7194–7202

7199

Fig. 4. Cytotoxic activity of CD8+ cytotoxic T cells against autologous PBMCs in response to peptides. These peptides had CD8+ T cell proliferative response in the WST-8 assay. CD8+ T cells (effector cells, EC) were separated from PBMCs from four BLV-infected calves [A-JB (A), B-JB (B), C-JB (C), and D-JB (D)] and incubated with autologous PBMCs (target cells, TC) with different concentrations of peptide (0, 0.5, 5, or 20 ␮M), and the cytotoxic activity of the CD8+ T cells was examined in a CytoTox96 non-radioactive cytotoxicity assay. A-JB was tested at an EC:TC ratio of 10:1, while B-JB, C-JB, and D-JB were used at an EC:TC ratio of 20:1. The bars represent the mean ± standard deviation (S.D.) of triplicate wells. Significant differences between the groups are indicated by asterisks (* p < 0.05, ** p < 0.01, and *** p < 0.001).

the gp30N16 peptide showed the cytotoxic activity in a dosedependent manner, reaching 80.9% and 77.9% cytotoxicity at 20 ␮M in calves B-JB and A-JB, respectively, while 0.5 ␮M of each of the other three peptides was sufficient to kill target cells in B-JB. 3.4. Estimation of the Wu–Kabat variability index for positive peptides The conservation of amino acids in the 11 CD8+ CTL epitopes was evaluated using the Wu–Kabat variability index. Fig. 5 shows the Wu–Kabat variability indexs of the 11 CD8+ CTL epitopes based on information on Env and Tax amino acid sequence variants obtained from the Genbank database. A value of 1 indicates that an amino acid site is monomorphic, while a value of >2 indicates a polymorphic site [28]. Interestingly, four residues of gp51N11 and three residues of gp51N12 had values exceeding 4 that the highest variability values (Fig. 5A). By contrast, the other epitopes of gp30 were slightly less polymorphic than the epitopes within the gp51 (Fig. 5B). Particularly, the gp30N16 contains a prolinerich region upstream of immunoreceptor tyrosine-based activation motifs (ITAMs) and immunoreceptor tyrosine-based inhibition motif (ITIMs) [30]. In addition, identified tax epitope region, which include a leucine-rich activation domain that encompasses a transcriptional activation region [31], are less polymorphic (Fig. 5C). Indeed, only one residue in each of epitopes gp30N5 and tax16 had values exceeding 4 (Fig. 5A and B). 4. Discussion Here, we determined 11 novel CD8+ CTL epitopes and produced four major conclusions. First, we found it interesting that this is

the first time T cell epitopes from the gp30 have been identified. In previous reports, some regions of gp51 and Tax were identified as T cell and B cell epitopes [15,20,22,29], but so far no T cell or B cell epitopes from gp30 have been reported. On the contrary, no epitopes were found in Gag proteins. Second, fewer CD8+ CTL epitopes were identified from calf A-JB with the low proviral load, and more CD8+ CTL epitopes were identified from higher proviral load than A-JB. This result strongly suggests that the number of CD8+ T cell epitopes correlates with the proviral load in BLVinfected cattle. Third, although several different CD8+ CTL epitopes were identified from individual calf, a common CD8+ CTL epitope, gp51N11, has capability of inducing cytotoxicity in all four calves. Thus, the predominant CD8+ CTL epitope may be located in the central portion of gp51. However, gp51N11 is not considered a suitable target for a vaccine, because it shows high variability according to the Wu–Kabat variability index. Fourth, Wu–Kabat variability index analysis clearly showed that CD8+ CTL epitopes corresponding to gp30 and Tax were less polymorphic than those of gp51. In addition, CD8+ CTL epitopes from all experimental calves were derived from the gp30 protein but not the Tax protein, indicating that the BLV gp30 region may be the best candidate for vaccine development. The number of CD8+ T cell epitopes appeared to correlate with the proviral load in BLV-infected cattle as evidence by: (i) The number of peptides that induced good CD8+ T cells proliferation increased with the BLV proviral load in the experimental calves; (ii) Four epitopes on the Tax protein were detected in only two cattle with viral loads of 67,729 and 76,839 copies per 105 cells. This suggests that Tax-specific CTL may be induced in animals with higher (>58,501 copies per 105 cells) proviral loads because the expression of Tax protein is very weak compared with the expression of the Env proteins; (iii) gp30N16-specific CTL was only found

7200

L. Bai et al. / Vaccine 33 (2015) 7194–7202

Fig. 5. Wu–Kabat variability index of the CD8+ CTL epitopes on all sequences of gp51, gp30, and Tax. The 517 sequences of gp51, 118 sequences of gp30, and 90 sequences of Tax were obtained from the GenBank database. The vertical axis indicates the Wu–Kabat index at each amino acid position and the horizontal axis shows the amino acid positions. (A) The Wu–Kabat index of the gp51 region (amino acids 1-301). Red, blue, and yellow blocks indicate the CD8+ CTL epitopes gp51N5, gp51N11, and gp51N12, respectively. (B) The Wu–Kabat index of the gp30 region (amino acids 302–515). The blue, brown, green, and pink blocks indicate the CD8+ CTL epitopes gp30N5, gp30N6, gp30N8, and gp30N16, respectively. (C) The Wu–Kabat index of Tax region (amino acids 1–309). The blue, red, green, and pink blocks indicate the CD8+ CTL epitopes tax16, tax18, tax19, and tax20, respectively. (D) The sequences and Wu–Kabat index at each position of the CD8+ CTL epitopes (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.).

L. Bai et al. / Vaccine 33 (2015) 7194–7202

in the calves A-JB and B-JB. gp30N16 encompasses a proline-rich region upstream of the ITAMs and ITIMs of the cytoplasmic tail of the envelope protein [30], which is known to be an SH3 binding site [32]. In addition, the down-regulatory phosphatase SHP-1 acts as a critical negative regulator of B-cell receptor signaling, and gp30 was thought be may act as a decoy to sequester SHP-1, Thus, gp30N16 was thought maybe relevant in the context of T and B mediated immunity [45]. Therefore, CTL targeting this domain may be efficient at eliminating BLV-infected cells by cytotoxic activity. In another previous study, synthetic peptides corresponding to the gp51 were used to test helper T cell epitopes, and peptide 98–117 was able to induce B cell neutralizing responses in mice and a particularly strong helper T cell response in BLV-infected sheep [15]. The 17 residues within gp51N11, which was identified in this study, correspond to peptide 98–117. Fig. 4 clearly illustrates that gp51N11 has cytotoxic activity on autologous BLVinfected cells from all four calves. The BoLA class I allele is highly polymorphic [10,11]. Indeed, the major BoLA class I allele was different in each of the four calves. The gp51N11 epitope may contain recognition site that binds to one or more BoLA class I alleles. Peptides 169–188 and 177–192 of gp51 induce good proliferation of bovine PBMCs depleted of B cells [14]. These peptides correspond to 12 residues within gp51N17, full length gp51N18, and 12 residues within gp51N19. Another peptide derived from Env, pep61, contains some sequences that induce T cell proliferation or mitogenicity in mice [21]. This peptide corresponds closely to gp51N7, which did not show CD8+ T cell proliferative activity in this study. Thus, the epitopes gp51N17–19 may be significant recognition sites in some cattle, but were not identified as such in the four experimental cattle in our study. We identified four tax CD8+ CTL epitopes that correspond to the region approximately halfway between amino acids 151 and 210 of the BLV Tax protein. Interestingly, this region of Tax contains a leucine-rich activation domain between residues 157 and 197 that contains 24% of the protein’s leucine residues and is possibly involved in heterologous protein interactions [31]. The leucine-rich activation domain also encompasses a transcriptional activating region that plays an important role in viral replication [31]. In addition, these CD8+ CTL epitopes on the Tax protein have few polymorphic sites (Fig. 5C). Various studies showed that the HTLV Tax protein contains several CTL epitopes [33–38] and induces strong CTL responses against HTLV-1-infected patients [39–41]. These results suggest that the tax CD8+ CTL epitope may be a potential candidate for an anti-BLV vaccine. By contrast, Sakakibara et al. reported that peptides of the BLV Tax protein at amino acids 111–130 and 131–150, and at 111–130 and 261–280, are T cell epitopes and B cell epitopes for mice, respectively [20]. These sequences of Tax corresponded to peptides tax11–13, tax13–15, and tax 26–28 in our study, which did not induce CD8+ T cell proliferation. We identified CD8+ CTL epitopes of Tax that induced high cytotoxic activity against autologous BLV-infected cells. The T cell epitopes that have been identified in mice by Sakakibara et al. may therefore not be CD8+ CTL epitopes in cattle, indicating that there are differences in CD8+ CTL epitopes between mice and cattle. The transmembrane protein gp30 is a glycosylated polypeptide [1,2] that can induce the fusion of viral and cellular membranes in infected cells [42]. However, it was unclear whether CD8+ CTL epitopes are present in the gp30. In this paper, four CD8+ CTL epitopes were identified in gp30 for the first time. In particular, gp30N16 showed strong cytotoxic activity in calves A-JB and B-JB, respectively. In addition, the cytotoxic activity at 0.5 ␮M of gp30N6 was 35.2–68.6% in calves B-JB, C-JB, and D-JB. Thus, CD8+ CTL epitopes from the gp30 were identified in all four BLV-infected calves, regardless of proviral load. Interestingly, four epitopes from the gp30 were less polymorphic than the epitopes from the gp51. These

7201

results indicate that the BLV gp30 region may be a useful target for vaccines development. Stability is an important factor for developing epitopedriven vaccines. In previous study [46], the peptide 121-145 of PPE68 (Rv3873) which was a major antigenic protein encoded by Mycobacteriun tuberculosis-specific genomic region of difference 1 was identified by antigen-specific T-cell lines from HLA-heteregeneous subjects, and conserved in all sequenced strains/species of M. tuberculosis and M. tuberculosis complex, and several other pathogenic mycobacterial species; the peptide aa 121–145 was thought to be exploited as a peptide-based vaccine candidate against tuberculosis and other mycobacterial diseases. Therefore, we analyzed the conservation of amino acids in the 11 CD8+ CTL epitopes using the Wu–Kabat variability index. All three epitopes from gp51 showed significantly higher variability than epitopes from gp30 and Tax. In particular, gp51N11 which induced responses in four all experimental calves, had the highest Wu–Kabat variability value. The epitope gp51N12 showed a similar pattern as gp51N11, in that it induced cytotoxic activity on autologous BLV-infected cells and had a high Wu–Kabat variability value. These data indicate that gp51N11 and gp51N12 are more variable at amino acid positions 141 and 155, respectively. The gp51 mediates binding of virions to cell surface receptors [42], and induces the expression of specific antibodies in infected animals [1]. Variability within the gp51N11 and gp51N12 epitopes allow for efficient modification of host immune responses. Although many vaccine trials have been undertaken [43], an effective vaccine has yet not been developed against BLV maybe because vaccine strategies have focused on the Env protein gp51. In contrast to gp51, epitopes derived from gp30 and Tax are less polymorphic. These results provide useful information for BLV vaccines development. 5. Conclusions Eleven novel CD8+ CTL epitopes were identified. However, no CD8+ CTL epitopes were found from the Gag proteins. Interestingly, several epitopes were obtained from Tax and gp30, indicating that cellular immunity against BLV is strongly targeted to these proteins. Tax protein has a few polymorphic sites, but tax epitopes just were identified in two calves. At least one epitope from gp30 region was identified in all four calves, and these epitopes have high conservation than epitopes of gp51. Although CTL epitopes were identified from the gp51, they have high variability and therefore not a suitable target for designing a vaccine. These results show that the gp30 region may be the best candidate for BLV vaccine development. Acknowledgments We thank Mr. Sunao Hirai, Mr Susumu Ogawa, Mr. Kenji Mizushiri, Mr. Naoya Takahashi, and Mr. Yoshihiro Okada of Animal Research Center, Hokkaido Research Organization, for help with drawing blood. We also thank Mis. Yuki Matsumoto, Mrs Mari Kikuya, Mis. Yuka Takahashi, Dr. Ayumu Ohno, and other members of the Viral Infectious Diseases Unit, RIKEN, for excellent technical assistance, kind help, and suggestions. We are grateful to the Support Unit for Biomaterial Analysis, RIKEN BSI Research Resources Center, for help with sequence analysis. This work was supported by a grant from the Program for the Promotion of Basic and Applied Research for Innovations in Bio-oriented Industry and by a grant from Integration Research for Agriculture and Interdisciplinary Fields in Japan. References [1] Gillet NA, Florins M, Boxus C, Burteau A, Nigro F, Vandermeers H, et al. Mechanisms of leukemogenesis induced by bovine leukemia virus: prospects for novel anti-retroviral therapies in human. Retrovirology 2007;4:18.

7202

L. Bai et al. / Vaccine 33 (2015) 7194–7202

[2] Aida Y, Murakami H, Takahashi M, Takeshima SN. Mechanisms of pathogenesis induced by bovine leukemia virus as a model for human T-cell leukemia virus. Front Microbiol 2013;4:328. [3] Sagata N, Yasunaga Y, Ohishi K, Tsuzuku-Kawamura J, Onuma M, Ikawa Y. Comparison of the entire genomes of bovine leukemia virus and human T-cell leukemia virus and characterization of their unidentified open reading frames. EMBO J 1984;3:3231–7. [4] Hamard-Peron E, Muriaux D. Retroviral matrix and lipids, the intimate interaction. Retrovirology 2011;8:15. [5] Voneche V, Portetelle D, Kettmann R, Willems L, Limbach K, Paoletti E, et al. Fusogenic segments of bovine leukemia virus and simian immunodeficiency virus are interchangeable and mediate fusion by means of oblique insertion in the lipid bilayer of their target cells. Proc Natl Acad Sci U. S. A 1992;89:3810–4. [6] Derse D. Bovine leukemia virus transcription is controlled by a virus encoded trans-acting factor and by cis-acting response elements. J Virol 1987;61:2462–71. [7] Willems L, Gegonne A, Chen G, Burny A, Kettmann R, Ghysdael J. The bovine leukemia virus p34 is a transactivator protein. EMBO J 1987;6:3385–9. [8] Haraguchi S, Good RA, Day NK. Immunosuppressive retroviral peptides: cAMP and cytokine patterns. Immunol Today 1995;16:595–603. [9] Elemans M, Florins A, Willems L, Asquith B. Rates of CTL killing in persistent viral infection in vivo. PoLS Comput Biol 2014:1003534, http://dx.doi.org/10.1371/journal.pcbi. [10] Yoko A, Takeshima SN, Cynthia LB, Azad KK. Bovine immunogenetics. In: Dorian JG, Anatoly R, editors. The genetics of cattle. London, UK: CBAI; 2015. p. 153–91. [11] Takeshima SN, Aida Y. Structure, function and disease susceptibility of the bovine major histocompatibility complex. Anim Sci J 2006;77:138–50. [12] Madden DR. The three-dimensional structure of peptide-MHC complexes. Ann Rev Immunol 1995;13:587–622. [13] Yewdell JW, Bennink JR. Immunodominance in major histocompatibility complex class I-restricted T lymphocyte responses. Ann Rev Immunol 1999;17:51–88. [14] Kabeya H, Ohashi K, Ohishit K, Sugimoto C, Amanumat H, Onuma M. An effective peptide vaccine to eliminate bovine leukaemia virus (BLV) infected cells in carrier sheep. Vaccine 1996;14:1118–22. [15] Callebaut I, Voneche V, Mager A, Fumiere O, Krchnak V, Merza M, et al. Mapping of B-neutralizing and T-helper cell epitopes on the bovine leukemia virus external glycoprotein gp51. J Virol 1993;67:5321–7. [16] Orlik O, Splitter GA. Progression to persistent lymphocytosis and tumor development in bovine leukemia virus (BLV)-infected cattle correlates with impaired proliferation of CD4+ T cells in response to gag- and env-encoded BLV proteins. J Virol 1996;70:7584–93. [17] Ohishi K, Suzuki H, Yamamoto T, Maruyama T, Mild K, Ikawa Y, et al. Protective immunity against bovine leukaemia virus (BLV) induced in carrier sheep by inoculation with a vaccinia virus-BLV env recombinant: association with cellmediated immunity. J Gen Virol 1991;72:1887–92. [18] Ohishi K, Suzuki H, Yasutomi Y, Onuma M, Okada K, Numakunai S, et al. Augmentation of bovine leukemia virus (BLV)-specific lymphocyte proliferation responses in ruminants by inoculation with BLV env-recombinant vaccina virus: their role in the suppression of BLV replication. Microbiol Immunol 1992;30:1317–23. [19] Onuma M, Hodatsu T, Yamamoto S, Higashihara M, Masu S, Mikami T, et al. Protection by vaccination against bovine leukemia virus infection in sheep. Am J Vet Res 1984;45:1212–5. [20] Sakakibara N, Kabeya H, Ohashi K, Sugimoto C, Onuma M. Epitope mapping of bovine leukemia virus transactivator protein Tax. J Vet Med 1998;60:599–605. [21] Kabeya H, Ohashi K, Sugimoto C, Onuma. M Bovine leukaemia virus envelope peptides cause immunomodulation in BALB/c mice. Vet Immunol Immunopathol 1999;68:39–48. [22] Gatei MH, Good MF, Daniel RC, Lavin MF. T-cell responses to highly conserved CD4 and CD8 epitopes on the outer membrane protein of bovine leukemia virus: relevance to vaccine development. J Virol 1993;67:1796–802. [23] Jimba M, Takeshima SN, Murakami H, Kohara J, Kobayashi N, Matsuhashi T, et al. BLV-CoCoMo-qPCR: a useful tool for evaluating bovine leukemia virus infection status. BMC Vet Res 2012;8:167–79. [24] Jimba M, Takeshima SN, Matoba K, Endoh D, Aida Y. BLV-CoCoMo-qPCR: quantitation of bovine leukemia virus proviral load using the CoCoMo algorithm. Retrovirology 2010;7:91–110. [25] Polat M, Ohno A, Takeshima SN, Kim J, Kikuya M, Matsumoto Y, et al. Detection and molecular characterization of bovine leukemia virus in Philippine cattle. Arch Virol 2015;160:285–96.

[26] Takeshima SN, Kitamura-Muramatsu Y, Yuan Y, Saito S, Aida Y. BLVCoCoMo-qPCR-2: improvements to the BLV-CoCoMo-qPCR by reducing primer degeneracy and constructing an optimal standard curve. Arch Virol 2015;160:1325–32. [27] Miyasaka M, Zdenek T. Sheep as an experimental model for immunology: immunological techniques in vitro and in vivo. Immunol Methods 1985;3:403–22. [28] Wu TT, Kabat EA. An analysis of the sequences of the variable regions of Bence Jones proteins and myeloma light chains and their implications for antibody complementarity. J Exp Med 1970;132:211–50. [29] Portetelle D, Dandoy C, Burny A, Zavada J, Siakkou H, Gras-Masse H, et al. Synthetic peptides approach to identification of epitopes on bovine leukemia virus envelope glycoprotein gp51. Virology 1989;169:34–41. [30] Hamilton VT, Stone DM, Pritchard SM, Cantor GH. Bovine leukemia virus gp30 transmembrane (TM) protein is not tyrosine phosphorylated: examining potential interactions with host tyrosine-mediated signaling. Virus Res 2002;90:155–69. [31] Willems L, Kettmann R, Burny A. The amino acid (157-197) peptide segment of bovine leukemia virus p34 tax encompass a leucine-rich globally neutral activation domain. Oncogene 1991;6:159–63. [32] Ren R, Mayer BJ, Cicchetti P, Baltimore D. Identification of a ten-amino acid proline-rich SH3 binding site. Science 1993;259:1157–61. [33] Hollsberg P, Weber WE, Dangond F, Batra V, Sette A, Hafler DA. Differential activation of proliferation and cytotoxicity in human T-cell lymphotropic virus type I Tax specific CD8T cells by an altered peptide ligand. Proc Natl Acad Sci USA 1995;92:4036–40. [34] Kannagi M, Shida H, Igarashi H, Kuruma K, Murai H, Aono Y, et al. Target epitope in the Tax protein of human T-cell leukemia virus type I recognized by class I major histocompatibility complex-restricted cytotoxic T cells. J Virol 1992;66:2928–33. [35] Koenig S, Woods RM, Brewah YA, Newell AJ, Jones GM, Boone E, 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;151:3874–83. [36] Niewiesk S, Daenke S, Parker CE, Taylor G, Weber J, Nightingale S, et al. Naturally occurring variants of human T-cell leukemia virus type I Tax protein impair its recognition by cytotoxic T lymphocytes and the transactivation function of Tax. J Virol 1995;69:2649–53. [37] Parker CE, Nightingale S, Taylor GP, Weber J, Bangham CR. 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;68:2860–8. [38] Pique C, Connan F, Levilain JP, Choppin J, Dokhelar MC. Among all human Tcell leukemia virus type I proteins, tax, polymerase, and envelope proteins are predicted as preferential targets for the HLA-A2-restricted cytotoxic T-cell response. J Virol 1996;70:4919–26. [39] Jacobson S, Shida H, McFarlin DE, Fauci AS, Koenig S. Circulating CD8+ cytotoxic T lymphocytes specific for HTLV-I pX in patients with HTLV-I associated neurological disease. Nature (Lond) 1990;348:245–8. [40] Kannagi M, Harada S, Maruyama I, Inoko H, Igarashi H, Kuwashima G, et al. Predominant recognition of human T cell leukemia virus type I (HTLV-I) pX gene products by human CD8+ cytotoxic T cells directed against HTLV-I-infected cells. Int Immunol 1991;3:761–7. [41] Parker CE, Daenke S, Nightingale S, Bangham CR, Activated. HTLV-Ispecific cytotoxic T-lymphocytes are found in healthy seropositives as well as in patients with tropical spastic paraparesis. Virology 1992;188: 628–36. [42] Johnston ER, Radke K. The S.U.T.M. envelope protein subunits of bovine leukemia virus are linked by disulfide bonds, both in cells and in virions. J Virol 2000;74:2930–5. [43] Kettmann R, Burny A, Callebaut I, Droogmans L, Mammerickx M, Willems L, et al. Bovine leukemia virus. Retrovir 1994;3:39–81. [44] Hojnik M, Dobovisek L, Knez Z, Ferk P. A synergistic interaction of 17beta-estradiol with specific cannabinoid receptor type 2 antagonist/inverse agonist on proliferation activity in primary human osteoblasts. Biomed Rep 2015;3:554–8. [45] Cantor GH, Pritchard SM, Orlik O, Splitter GA, Davis WC. Bovine leukemia virus transmembrane protein gp30 physically associates with the down-regulatory phosphatase SHP-1. Cell Immunol 1999;193:117–24. [46] Mustafa AS. Characterization of a cross-reactive, immunodominant and HLApromiscuous epitope of Mycobacterium tuberculosis-specific major antigenic protein PPE68. PloS One 2014;9:e103679.