Figure 1.
Overview of synthetic minigene screening.
a) Libraries are synthesized on programmable microarrays, cleaved from the chip surface and provided as a single mixture of antisense oligonucleotide templates. b) Initial 96 well PCR reactions utilize individual sense, pool specific primers (green, purple, orange arrows) in combination with a common primer (red arrows) to amplify specific pools of antisense templates (multicolor regions) from the mixed oligo library. Synthesis of a complement for the common primer is dependent upon synthesis of the sense strand primed by a single, unique pool-specific primer in each well. This subdivides the library into ordered arrays of minigene pools, with 10 defined minigenes/well. c) A second PCR reaction sews a stop codon and a human beta-globin 3′ UTR (purple+gold boxes) onto each minigene using the common primer domain as an overlap. PCR is driven by the sense pool specific primer, and an antisense primer extending from a 130 base oligo dT tail through the 3′ end of the UTR. Inclusion of an oligo dT tail on the antisense strand encodes a polyA template on the end of each mature minigene. This template allows synthesis of poly-adenylated mRNA during in vitro transcription. d) Arrays of minigene pools are purified and subject to in vitro transcription in the presence of a cap analogue, producing an array of defined, fully translatable mRNA pools. e) IVT products are transfected into autologous CD40L expanded B cells for use as antigen presenting cells. Transfected APCs are used as stimulators and targets for in vitro stimulations and IFNγ ELISPOT assays.
Figure 2.
Screening and pool deconvolution ELISPOTs.
Screening: a) and b) are bar graphs of primary screening results from two newly diagnosed T1D subjects. Panel a) is subject ND2, panel b) is subject ND3. Each bar represents the IFNγ spot number for an individual cultured ELISPOT well stimulated with a single minigene pool. Red line indicates 5× standard deviations of wells stimulated with mock transfected autologous B cells. Red circles indicate three wells that scored positive in both screens. Deconvolution c) and d): cultured IFNγ ELISPOT assays testing ND2-derived CD8+ T cell responses against individual minigenes from c) pool 85 and d) pool 308. Individual minigenes were amplified using minigene specific primers. Subsequent PCR reactions added T7 and common sequences, and full length individual minigenes were rebuilt and tested as described in Fig. 1. Targeted minigenes encode 33 residue peptides displayed in red. Peptide epitope mapping e) and f): Direct IFNγ ELISPOT assays testing overlapping 15 residue peptides from GLIPR1 and EpCAM minigenes targeted in b) and c). Peptides were tested in triplicate using a direct 24 hour IFNγ ELISPOT assay with 1×105 CD8+ T cells/well. Purple residues indicate non-antigen derived sequences encoded by minigene flanking sequences. Red residues indicate 9 residue peptide epitope identified in subsequent epitope mapping experiments.
Table 1.
Responses to GLIPR1 4–12 and EpCAM 140–148 incases and controls, and HLA restriction studies.
Figure 3.
Responses to GLIPR1 4–12 and EpCAM 140–148 in cases and controls, and HLA restriction studies.
Responses to a) GLIPR1 (4–12) b) and EpCAM (140–148) in T1D, at risk and control subjects. In both panels, PBMCs from the indicated patient groups were stimulated with GLIPR1 or EpCAM peptides (filled bars) vs. control peptide (open bars) overnight. Anti-CD3/CD28 stimulation was used as a positive control and all responses to antibody stimulation were too numerous to count (data not shown). The graphs display the average raw number of spots from triplicate wells +/− the standard error. The (*) indicates for a given patient that the experimental peptide responses were significantly higher (FDR <0.05) than the response to the control peptide. The number of patients who significantly responded to the each experimental peptide is indicated by the fraction in the graph. PBMC from subject ND2 used for discovery of each epitope were not available and are not included in these graphs. c) HLA-A*0201 binding assay for GLIPR1 epitope TLATIAWMV on T2 cells. 1×106 T2 cells were incubated with 20 µg/ml of the indicated peptide in media for 4 hours at 37°C, stained for HLA-A*0201 and evaluated by flow cytomety. d). Presentation of GLIPR1 eptiope by HLA A*0201. T2 cells were loaded with 10 µg/ml GLIPR1 epitope TLATIAWMV or control epitope from pyruvate dehydrogenase for two hours at room temperature, washed and mixed with purified CD8+ T cells from one responding subject (T1D #2) in a direct IFNγ assay and was found to be statistically significant compared to the control peptide (FDR<0.05).