Fassò et al. 10.1073/pnas.0712269105. |
Fig. 7. Generation of TRAMP-reactive CD8+ T cell lines and hybridomas. (A) Confirmation of specificity of the T cell hybridomas to TRAMP cells: 0.85 ´105 BTZs were incubated with 3 ´ 106 TRAMP-C2, MC38 or B16BL6 cells overnight and the T cell response was measured by LacZ production. Standard deviations of triplicate cultures are shown. (B) MHC restriction of the BTZs: 2 ´ 104 TRAMP-C2 cells were incubated in triplicate cultures for 1 h with an anti-H-2Kb (Y3, ATCC) or anti-H-2Db antibody (B22.249.RI, Cedar Lane, CA) before addition of BTZs (1 ´ 105/well). Cells were then incubated overnight and the T cell response measured. Standard deviations are shown.
Fig. 8. SPAS-1 sequence. Sequence alignment of the full-length SPAS-1 protein with its human ortholog SH3GLB2. Identical amino acids are shaded in dark gray, similar amino acids in light gray. The predicted coiled-coiled and SH3 domains, as well as the mouse T cell epitope defined below, are indicated. Predicted domains were identified by using the algorithm SMART (Simple Modular Architecture Research Tool) at EMBL.
Fig. 9. Identification of the antigenic peptide. (A) Titrated doses of the predicted H-2 Db-binding peptides Db-1 to Db-8 (HMLDLQKQL, LALLADELI, THVNHLHCL, ISSTHVNHL, YSLPGMDPD, PVVPTGAVL, YAQCYRHML and TSPTTTAATM) were pulsed onto H-2Db-expressing L-cells and tested for activation of the BTZ1.4 T cell hybridoma. (B) Deletion mutants were generated around the regions coding for the putative H-2Db-1, H-2Db-2, and H-2Db-3 peptides and were tested along the originally cloned SPAS-1 insert containing the 3' UTR. Plasmid DNA of SPAS-1 cDNA deletion constructs (in titrated amounts) were transiently cotransfected into LMtk- cells together with the relevant H-2Db cDNA and B7-2 cDNA (included to augment the activation). Two days later, 8.5 ´ 104 BTZ1.4 T cell hybridomas were added to the wells in triplicate cultures for overnight stimulation and assayed for LacZ activity. Standard deviations are shown. (C) Minigenes (at 0.5 mg/ml) encoding the putative SPAS-1 T cell epitope were transiently co-transfected into LMtk- cells together with the relevant H-2Db cDNA and B7-2 cDNA. Two days later, 8.5 ´ 104 BTZ1.4 T cell hybridomas were added to the wells in triplicate cultures and assayed for LacZ activity the next day. Standard deviations are shown. (D) The reactivity of the BTZ1.4 T cell hybridoma for the synthetic peptide corresponding to the newly defined SPAS-1 T cell epitope STHVNHLHC (SNC9-H8) is compared to predicted Db-1 and Db-3 synthetic peptides. Standard deviations are shown for triplicate cultures. (E) Naturally processed tumor peptides (NPTPs) from IFN-g-treated TRAMP-C2 cells were extracted and fractionated through reverse-phase HPLC (open circles). As a comparison, the synthetic peptide SNC9-H8 was run under identical conditions (solid circles). The fractions were tested for presence of antigenic activity by using the BTZ1.4 T cell hybridoma.
Fig. 10. Prediction of HLA-A2-binding peptides in human SPAS-1/SH3GLB2. To confirm binding to HLA-A2, the predicted SPAS-1/SH3GLB2 synthetic peptides P1-P5 [P1: LLADELITV (LV-9); P2: YMADAASEL (YL-9); P3: LLLEGISST (LT-9); P4: FLTPLRNFL (FL-9); and P5: ILSASASAL (IL-9)] were tested in a T2 binding assay. Four out of the five peptides bound to HLA-A2 as demonstrated by stabilization of surface HLA-A2 expression. Positive controls: Flu peptide and CEA Cap1 peptide; negative control: DMSO.
Fig. 11. Over-expression of huSPAS-1 protein in THP-1 cells infected with the pMGlyt2-SPAS-1/SH3GLB2 virus. Over-expression of human SPAS-1 protein in the human monocytic leukemia-derived cell line THP-1 infected with the pMGlyt2-SPAS-1/SH3GLB2 virus was confirmed by Western blot analysis under reducing conditions. Lane 1: lysate from 50,000 untransduced THP-1 cells. Lane 2: lysate from 50,000 THP-1 cells transduced with the pMGlyt2 IRES CD8 empty vector. Lane 3: lysate from 50,000 THP-1 cells transduced with pMGlyt2 SPAS-1/SH3GLB2 IRES CD8 vector. A polyclonal rabbit anti-SPAS-1 antibody raised to aa188-202 was used for detection of the 45-50kDa SPAS-1 protein (Sigma Genosys). As control for loading amounts, the same membrane was blotted with an anti-b-actin antibody (SC Biotech, CA).
SI Material and Methods
Animals.
All mice experiments were performed under the approved mouse protocol #R015-0604BRC at the University of California, Berkeley.cDNA library, expression screens and DNA constructs and peptides.
Both the construction of the unidirectional cDNA library, as well as the screen for isolating T cell-stimulating antigen clones, have been described (1). Briefly, using poly(A)+ mRNA from IFN-g-treated TRAMP-C2 tumor cells, a unidirectional cDNA library was constructed (Superscript Choice System, GIBCO-BRL) in the BstXI and NotI sites of the mammalian expression vector pcDNA1 (Invitrogen). The cDNAs were screened by transforming competent bacteria with recombinant plasmids and culturing in pools of 30-100 cfu in 96-well U-bottom plates. Miniscale preparation of the bacterial plasmid DNA was performed directly in the 96-well plates and subsequently plasmids were transfected into 3 ´ 104 LMtk- cells cotransfected with H-2Db and B7-2 cDNA. Two days later, 8.5 ´ 104 BTZ5.65 were added per well and cocultured overnight. The T cell response was measured as the LacZ activity by the conversion of the substrate chlorophenol red b-pyranoside (CPRG) at 595 and 655 nm as reference. Cultures with above background absorbance were identified as positive pools. Repeating the screen with individual colonies obtained from the positive cDNA pool identified the plasmid encoding the antigenic activity.Deletion constructs of the SPAS-1 cDNA were prepared by standard PCR techniques. The minigene constructs encoding MSTHVNHLHC, MSTHVNHLHCL, MSSTHVNHLHC, and MTHVNHLHCL were prepared by using complementary oligonucleotides corresponding to the indicated sequences. The putative H-2 Db-binding peptides H-2Db-1 to H-Db-8 (HMLDLQKQL, LALLADELI, THVNHLHCL, ISSTHVNHL, YSLPGMDPD, PVVPTGAVL, YAQCYRHML and TSPTTTAATM) were predicted using the computer algorithm of the BioInformatics and Molecular Analysis Section (BIMAS) at http://bimas.dcrt.nih.gov/molbio/hla_bind/. The peptides were synthesized commercially (Genemed Synthesis).
Immunizations and generation of T cell lines and hybridomas.
CD8+ T cell lines were established from the spleens and lymph nodes of C57BL/6 male mice immunized with GM-CSF-producing TRAMP-C2 cells (TRAMPC2-GM) and anti-CTLA-4 antibody. Briefly, 5-week-old C57BL/6 mice were injected s.c. with 1.5 ´ 106 irradiated (12 krads) GM-TRAMP-C2 cells, followed by five further boosts 3 weeks apart with this identical treatment. The same mice also received three i.p. injections of 100 mg of anti-CTLA-4 antibody [clone 9H10 (2)] on days 1, 4, and 7 after the first vaccination. Six days after the last boost, the spleens and lymph nodes were harvested, pooled, and a single cell suspension was prepared in ice-cooled T cell media (RPMI-1640 supplemented with 2 mM L-glutamine, penicillin, streptomycin, nonessential amino acids, sodium pyruvate, 10 mM Hepes Buffer (GIBCO BRL), 5 mM 2-ME, and 10% FCS (Sigma-Aldrich). Lymphocytes were plated at a density of 20 ´ 106 cells per well in a six-well plate. B7-TRAMP-C2 cells were used as stimulators after being treated at 106 cell/ml with 80 mg/ml Mitomycin C (Sigma-Aldrich) for 1.5 h at 37°C and subsequently washed five times in PBS. The treated stimulators were then added to the lymphocytes at 1.0 ´ 106 cells per well in a final 5% final concentration of RAT T-STIM with ConA (Becton Dickinson). On day 7, the lymphocytes were purified on a Ficoll gradient and restimulated with 1.0 ´ 106 Mitomycin C-treated B7-TRAMP-C2 cells in 5% ConA supernatant. Flow cytometric analysis of the T cell line showed the cells were uniformly CD8+. Generation of T cell hybridomas has been described in detail (3).T cell hybridoma-specific responses were measured by the production of b-galactosidase activity as described previously (3). Briefly, 8-10 ´ 104 T cell hybridomas were cocultured overnight in flat bottom 96-well plates with either 2-5 ´ 104 tumor cells or with L-cells expressing the restricting H-2Db MHC molecule pulsed with the appropriate peptide antigen. To increase the amount of endogenous MHC class I expression on the tumor cells, they were incubated at 37°C in 5% CO2 for 48 h in 1 ng/ml IFN-g (Biosource) before being mixed with the T cell hybridomas. The peptide/MHC-induced T cell response was assayed as LacZ activity by using the substrate chlorophenol red b-D-galactopyrannoside (CPRG; Calbiochem). The conversion of CPRG to chlorophenol red was measured at 595 and 655 nm as a reference wavelength with a 96-well microplate reader (Bio-Rad).
Mouse T cell activation assays.
Specific T cell responses against peptide/MHC were measured by the production of IFN-g by either ELISA, ELISPOT, intracellular cytokine staining (ICS), or by cytolytic activity in a JAM assay. T cell activation assays were performed 5 days after the last restimulation of the T cell line by incubating 1.0 ´ 105 Ficoll-purified T cells with 0.3 ´ 105 TRAMP-C2, MC38, B16BL6, or EL4 cells. IFN-g production in the culture supernatants was measured 24 h later with a sandwich ELISA (BD Biosciences). For ex vivo assays, the frequency of IFN-g secreting CD8+ T lymphocytes specific for the mutated and the wild-type SPAS-1 H-2Db-restricted epitopes SNC9-H8 (STHVNHLHC) and SNC9-R8 (STHVNHLRC) or for the irrelevant H-2Kb-restricted epitope SL8 (SIINFEKL) from ovalbumin were determined by ELISPOT using a BD Biosciences kit (BD Biosciences) 6 days after last vaccination. The cytokine response was expressed as the number of IFN-g spot-forming cells (SFCs) per 106 splenocytes. The frequency of IFN-g secreting CD8+ T lymphocytes specific for SPAS-1-SNC9-H8 and SPAS-1-SNC9-R8 was also determined by intracellular cytokine staining using a BD Bioscences kit. Briefly, 106 splenocytes from naïve or vaccinated mice were incubated at specific peptide concentrations for 5 h in the presence of Brefeldin A. After incubation, the samples were prepared according to the BD Biosciences protocol. For detection of cytolytic activity, 3H-thymidine-labeled tumor cells were used as targets in a JAM assay as described (4). Ficoll-purified effector T cells were incubated at varied effector to target (E:T) ratios with 10,000 3H-thymidine labeled target cells for 5 h. Percent specific lysis was calculated as the amount of 3H-thymidine recovered from target cells that were not killed as follows: [(SL-E)/SL ´ 100], where SL = spontaneous lysis (no effectors), E = experimental value.Real-time RT-PCR.
Total RNA from tissues and cells was isolated by using TRIzol reagent (Sigma-Aldrich). cDNA was synthesized by using oligo (dT) as the first primer and the Superscript RNase H- Reverse Transcriptase kit from GIBCO/BRL. Real-time PCR for SPAS-1 and the reference b-actin was performed in an ABI Gene Amp 5700 machine with the SYBR green mastermix kit (Applied Biosystems). A standard curve established using known quantities of PCR template was used for both SPAS-1 and b-actin detection. CT values were set manually above a threshold that reflected linear amplification and converted into arbitrary units by using the known standard curve. SPAS-1 expression data were then normalized relative to b-actin. Primer sequences for SPAS-1 were 5'-GAA GCC AAA GCC ACG ACG GTG-3' and 5'-CAT CGT TCC AAA GCG CGG AGG-3'; for b-actin, 5'-TGG AAT CCT GTG GCA TCC AT-3' and 5'-TTT ACG GAT GTC AAC GTC ACA CT-3'.Extraction and HPLC analysis of naturally processed tumor peptides (NPTPs).
The total acid-soluble peptide pool from TRAMP-C2 cells was extracted and the NPTPs were fractionated by reverse-phase HPLC, as described previously (5). Briefly, 1.0 ´ 109 TRAMP-C2 cells were washed with PBS and extracted with 1 ml of 1% trifluoroacetic acid (TFA) in boiling water for 5 min. The irrelevant 15-mer peptide Hen Egg Lysozyme 74-88 was added at 1 mM concentration to the extract as a carrier to prevent nonspecific losses. Cellular debris was removed by centrifugation and the extract was fractionated by HPLC after filtration through a 10 kDa Millipore filter to remove large molecular weight species. The fractions were collected in flat-bottom 96-well plates, dried in a vacuum centrifuge, and resuspended in 30 ml of PBS + 10% DMSO. The stimulatory activity of each fraction was assayed by adding to each well 0.85 ´ 105 BTZ1.4 T cell hybridomas and 0.3 ´ 105 H2-Db-expressing L cells in a total volume of 200 ml and culturing overnight at 37°C in 5% CO2. Mock injections with sample buffer alone were performed before each extract sample, using the same column and identical run conditions, to demonstrate absence of cross-contamination between samples.Cloning of human SPAS-1/SH3GLB2.
RT-PCR was performed on cDNA from the human prostate cancer cell line DU-145 (ATCC) by using the primers GCC GCC ACC ATG GAG TTC AAC ATG AAG AAG CTG GCG and CTA GCT AGA CAG TTC CAA GTA GGT GAC AG. The PCR product was shuttled through the TOPO TA cloning vector (Invitrogen) into the NotI and SurfI cloning sites of the pMG-lyt2 retroviral vector (gift of Dr. Michael Curran, Memorial Sloan-Kettering Cancer Center). Retroviruses containing the empty PMGlyt2 vector or the SPAS-1/SH3GLB2-encoding PMGlyt2 vector were produced by transient transfection of the Phoenix-E packaging line. To increase virus production, the cells were additionally transfected with plasmids encoding ecotropic envelope and gag-pol proteins (gift of Dr. Michael Curran). Infection of THP-1 cells with the retroviruses was performed by centrifugation at 1,200 ´ g by using filtered retroviral supernatants, supplemented with 4 mg/ml polybrene (Sigma-Aldrich) and 25 mM Hepes Buffer (GIBCO/BRL). Infected THP-1 cells were expanded and sorted for equally high CD8 expression (mouse CD8 is in the retroviral vector following an IRES element).HLA-A2-binding peptide prediction and T2-binding assay.
The computer algorithms SYFPEITHI, BIMAS, and nHLApred (http://www.syfpeithi.de/, http://www-bimas.cit.nih.gov/molbio/hla_bind/, http://www.imtech.res.in/raghava/nhlapred/neural.html) were used to predict HLA-A2-binding epitopes in the SPAS-1/SH3GLB2 protein. Five peptides P1 to P5 (9 mers) displaying the highest-binding scores according to all of the three algorithms were synthesized [P1: LLADELITV (LV-9); P2: YMADAASEL (YL-9); P3: LLLEGISST (LT-9); P4: FLTPLRNFL (FL-9); and P5: ILSASASAL (IL-9)] (Peptidogenic Research).A T2-binding assay was used to assay the ability of the peptides to bind and stabilize HLA-A2 molecules on T2 cells as described (6). Briefly, 1 ´ 105 T2 cells were incubated with each one of the five peptides at a final concentration of 100 mM plus 5 nM b2 microglobulin for 18 h at 37°C in 5% CO2 in triplicate wells. As controls for HLA-A2 binding, we used the influenza A matrix protein (MP) 58-66 peptide. The level of stabilized HLA-A2 was detected by flow cytometry by using an anti-HLA-A2 antibody (BD PharMingen).
1. Mendoza LM, Malarkannan S, Shastri N (2001) Identification of CD8+ T-cell-stimulating antigen genes in cDNA libraries. Methods Mol Biol 156:255-263.
2. Krummel MF, Allison JP (1995) CD28 and CTLA-4 have opposing effects on the response of T cells to stimulation. J Exp Med 182:459-465.
3. Malarkannan S, Mendoza LM, Shastri N (2001) Generation of antigen-specific, lacZ-inducible T cell hybrids. Methods Mol Biol 156:265-272.
4. Matzinger P (1991) The JAM test. A simple assay for DNA fragmentation and cell death. J Immunol Methods 145:185-192.
5. Malarkannan S, Goth S, Buchholz DR, Shastri N (1995) The role of MHC class I molecules in the generation of endogenous peptide/MHC complexes. J Immunol 154:585-598.
6. Cerundolo V, et al. (1990) Presentation of viral antigen controlled by a gene in the major histocompatibility complex. Nature 345:449-452.