Supplemental Figure 1. Immune phenotype of mcd19 targeted CAR T and dose titration of in vivo

Transcription

1 Supplemental Materials Supplemental Methods Supplemental Figure 1. Immune phenotype of mcd19 targeted CAR T and dose titration of in vivo efficacy. Supplemental Figure 2. Gene expression of fluorescent-protein tagged CAR T cells. Supplemental Figure 3. Fluorescent protein tagged CAR T cells function similarly to non-tagged counterparts. Supplemental Figure 4. Transduction efficiency and immune phenotype of mcd19 targeted CAR T cells for survival study (Figure 2D). Supplemental Figure 5. Transduction efficiency and immune phenotype of CAR T cells used in irradiated CAR T study (Fig. 3B-C). Supplemental Figure 6. Differential gene expression of CD4+ m19-humbbz CAR T cells. Supplemental Figure 7. CAR expression and CD4/CD8 subsets of human CD19 targeted CAR T cells for Figure 5E-G. Supplemental Figure 8. Transduction efficiency and immune phenotype of mcd19 targeted wild type (WT) and TRAF1 -/- CAR T cells used for in vivo study (Figure 6D). Supplemental Figure 9. Mutated CAR T cells have increased NF-κB signaling, improved cytokine production, anti-apoptosis, and in vivo function. Supplemental Figure 1. TRAF and CAR co-expression in human CD19-targeted CAR T cells. Supplemental Figure 11. TRAF2 over-expressed h19bbz CAR T cells show similar in vivo efficacy to h19bbz CAR T cells in an aggressive leukemia model. S1

2 Supplemental Table 1. Probesets increased in m19z and vs CAR T cells. Supplemental Table 2. Probesets increased in vs m19z and CAR T cells. Supplemental Table 3. Probesets differentially expressed in m19z vs CAR T cells. Supplemental Table 4. Probesets differentially expressed in vs CAR T cells. S2

3 Supplemental Methods Gene Expression Microarray. For m19z, and comparison, three million CAR T cells were incubated with T3-mCD19 cells overnight. The next day live CAR T cells were sorted into Trizol (Thermo Fisher Scientific, Waltham, MA). RNA was isolated according to manufacturer s instructions and run on a MOE 43A 2. array Mouse Genechip (Affymetrix, Santa Clara, CA) at the Genomics Core Facility. Gene expression analyses and graphic representations were performed with the Partek Genomics Suite Software. RMA normalization was performed and values generated for each probeset for all samples. Differentially expressed genes were detected by ANOVA and probesets of statistical significance were defined by a-fold change > 2 and a FDR.5. RNA-SEQ. For m19z, and m19-humbbz comparison, three million CAR T cells were incubated with T3-mCD19 cells for 48 hr. Live CD4+ CAR T cells were sorted into Trizol. RNA was isolated according to manufacturer s instructions and evaluated for quality. The Genomic Core performed mrna enrichment and cdna library preparation using the Illumina Tru-seq stranded mrna sample prep kit. Final RNA-seq libraries were reviewed for size and quality on the Agilent TapeStation, followed by quantitative PCR-based quantitation with the Kapa Library Quantification Kit. The libraries sequenced on two NextSeq high-output 2x75 paired-end sequencing runs in order to generate approximately 4 million pairs of reads per sample. Sequence reads were aligned to the human reference genome in a splice-aware fashion using Tophat2 (1), allowing for accurate alignments of sequences across introns. Aligned reads were quantitated at the gene level using HTseq (2). Normalization, expression modeling, and difference testing were performed using DESeq (3). Quality control measures included custom scripts and RSeqC (4) to examine read count metrics, alignment fraction, chromosomal S3

4 alignment counts, expression distribution measures, and principle components analysis and hierarchical clustering. Differentially expressed genes were detected by ANOVA and probesets of statistical significance were defined by a-fold change > 4 and a FDR.1. Gene set enrichment analysis was performed (on the gene expression values) to analyze the enrichment of the gene sets using GSEA software ( C5 collection version v6. from the Molecular Signature Database (MSigDB v6. C5), which contains the expert-curated gene ontology (GO) gene sets, were used in the analysis. We used vertebrate homology resource ( to convert between homologues human and mouse genes. For all comparisons, data was collapsed to gene symbols. 1 permutations based on gene sets were performed. Gene sets were ranked according to false discovery rate (FDR) q-value. At the default FDR q-value cut-off within GSEA of.25, we identified 3 gene sets that are upregulated in m19z and 68 gene sets upregulated in. Microarray and RNA-SEQ data have been submitted to GEO (Gene Expression Omnibus) with the accession number GSE CD33-targeted CARs Anti-CD33 antibodies were developed at the Vanderbilt Antibody and Protein Resource using standard methods (5). Briefly, after completing a series of immunizations splenocytes of immunized mice were isolated and fused to a non-ig secreting myeloma cell line and grown in a semi-solid plate. Antibodysecreting clusters were identified in semi-solid plates and selected for clonal expansion in 96 well plates. S4

5 During expansion supernatant was collected and assayed for CD33 binding by ELISA as well as flow cytometry. Based on this screening hybridomas were selected for expansion and isolation of RNA, which was used to amplify IgH and IgL rearrangements. Based on the IgH and IgL rearrangements scfv were designed and cloned into the NcoI/NotI sites of our human CD19-targeted CAR in the SFG retroviral cassette. This allowed replacement of the anti-human CD19 scfv with anti-human CD33 scfv. These constructs were then used to produce gammaretroviral supernatant as described in Methods. In vivo NALM6 animal model of CD19-targeted CAR T cells The NALM6 leukemia mouse model has been described (6). Briefly, NALM6-GL cells were i.v. injected to NSG mice at 5x1 5 dose. Four days later, mice were treated with 3x1 5-1x1 6 human CD19 targeted CAR T cells. Human CD19 targeted CAR T cells with excess TRAF2 were made by CAR and mouse TRAF2 co-transduction or transduction with a bicistronic construct combining CAR and human TRAF2. Blood samples were collected weekly for flow cytometry. Leukemia burden was evaluated weekly using bioluminescence imaging on an IVIS system. Survival was monitored. Mice were sacrificed when they develop signs of progressive leukemia. S5

6 A m19 z CD4 CD44 CD8 CD62L B B cells/μl blood B6 m19 z (1E6) m19z (3E5) m19z (1E6) (3E5) (1E6) (3E5) (1E6) Thy1.1 T cells/μl blood B6 m19 z (1E6) m19z (3E5) m19z (1E6) (3E5) (1E6) (3E5) (1E6) Supplemental Figure 1. Immune phenotype of mcd19 targeted CAR T and dose titration of in vivo efficacy. (A) Immune phenotype of transduced T cells used in 5x1 6 dose in vivo study (Figure 1C&D). Cells were pre-gated on single live cells. (B) In vivo B cell killing and T cell persistence with T cell dose titration. After Eµ-ALL injection, different doses of T cells were given to 3 mg/kg CTX preconditioned mice. B (B22+CD19+) and T (CD3+Thy1.1+) cells in peripheral blood were quantitated 3 weeks after CAR T injection. Each dot indicates one mouse. Data are from one single experiment (n=34 total). S6

7 A C + 5 LTR GFP/ V H G/S V L EC TM G/S 3 LTR mcherry SD SA + 5 LTR GFP/ V H G/S V L EC TM CD3 G/S 3 LTR mcherry SD SA + GFP/ 5 LTR V H G/S V L EC TM CD28 CD3 G/S 3 LTR mcherry SD SA + 5 LTR V H G/S V EC TM m4-1bb CD3 G/S GFP/ L 3 LTR mcherry SD SA CD8 Leader CD8 B D CAR% 8% 6% 4% 2% % mcherry Protein L mcherry Protein L mcherry Protein L mcherry Protein L m19 z m19z E m19z m19-mbbz Supplemental Figure 2. Gene expression of fluorescent-protein tagged CAR T cells. (A) Schematic of genetic constructs for mcd19 targeted CARs. Shown are the long terminal repeats (LTR), packaging signal Ψ, splice donor (SD), splice acceptor (SA), VH and VL regions of the scfv (single-chain variable fragment), the extracellular hinge (EC), transmembrane (TM), and intracellular regions of the retroviral construct. G/S, (Gly4Ser1)3 linker sequence. (B) Comparison of fluorescence protein and Protein L as a method to evaluate CAR expression. One million T cells transduced with mcherry-tagged CARs were incubated with 1 µg Biotin-Protein L and then fluochrome-conjugated streptavidin. Cells were subjected to flow cytometry. Data are representative of 4 independent experiments. (C) Principal component S7

8 analysis (PCA) of mcd19-targeted CAR T cells stimulated with antigen. (D) Venn Diagram demonstrating the number of genes differentially expressed (n=25) in CAR T cells compared to both m19z and CAR T cells. (E) A heatmap of the 25 differentially expressed genes. The list of 25 genes is included in Supplemental Tables 1-4. For (C)-(E), CAR T cells with the m19z,, or CAR tagged to the fluorescent protein GFP were incubated with 3T3- mcd19 AAPC at 1:1 E:T ratio overnight, FACS-sorted, and lysed to isolate RNA. Each group of CAR T cells was transduced, stimulated, and sorted independently in triplicates. S8

9 A B C GFP Survival IFN (pg/ml) CD8 CD44 D % 5% CD3 CD4 CD62L % B cells/μl blood8 13 Days after leukemia WK 1 m19 z m19z **** **** WK 2 WK 4 *** ** **** *** **** WK 8 WK 12 TNF (pg/ml) m19 z m19z ** m19 z m19z m19δz m19z CAR T/μl blood * WK 1 **** ns B cells/femur6 15 WK 2 WK 4 WK 8 m19 z WK 12 m19z * Supplemental Figure 3. Fluorescent protein tagged CAR T cells function similarly to non-tagged counterparts. (A) Cytokines released by fluorescent protein tagged CAR T cells upon antigen stimulation. Day 4 CAR T cells were co-cultured with 3T3-mCD19 at 1:1 ratio for 24 hr. Supernatant was subjected to luminex assay for IFNγ and TNFα. (B) Immune phenotype of CAR T cells with a fluorescent protein tag. Day 4 CAR T cells were harvested, beads removed and subjected to flow cytometry. Cells were pre-gated on single live cells (top) or CD3+CAR+ cells (middle & bottom). (C) Survival (n=5), in vivo B cell killing and CAR T persistence in mice treated by CAR T cells with a fluorescent protein tag. Seven days after i.v. injection with 1X1 6 Eµ-ALL cells, mice were i.p. injected with CTX at 25 mg/kg and then one day later i.v. injected with 1X1 6 CAR T cells. Survival was monitored. BM was isolated 11 days after CAR T injection and subjected to flow cytometry. B (B22+CD19+) and CAR T (CD3+- CAR+) cells were counted using CountBright beads. Each dot indicates one mouse (n=3 per group). Data are from one single experiment. (D) B and CAR T cell counts over time in the blood after CAR T treatment. B6 mice were injected with CTX (25 mg/kg) and CAR T cells (3X1 5 ). Blood samples were collected over time and B and CAR T cell numbers were measured by flow cytometry (n=1 per group). Survival curve, logrank test; all other data, unpaired t test. *p<.5; **p<.1; ***p<.1; ****p<.1; ns, not significant. CAR T/femur m19 z m19z * S9

10 m19δz m19z m19-humbbz CD8 GFP CD44 CD4 CD62L Supplementary figure 4. Transduction efficiency and immune phenotype of mcd19 targeted CAR T cells for survival study (Figure 2D). Data are representative of four independent productions used to generate CAR T cells for the survival experiments of Figure 2D. For Supplemental Figure 4. Transduction efficiency and immune phenotype of mcd19 targeted CAR transduction efficiency (top panel), cells were pre-gated on single live cells. For immune Tphenotype cells for survival study (Figure 2D). Data representative four independent (middle and bottom panels), cells are were pre-gated onofsingle live CAR T productions used to generate CAR T cells for the survival experiments of Figure 2D. For transduction efficiency (top panel), (CD3+CAR+) cells. cells were pre-gated on single live cells. For immune phenotype (middle and bottom panels), cells were pre-gated on single live CAR T (CD3+CAR+) cells. S1

11 m19 z m19-humbbz CAR CD8 CD4 CD44 CD62L Supplemental Figure 5. Transduction efficiency and immune phenotype of CAR T cells used in irradiated CAR T study (Figure 3B-3C). Day 4 transduced cells were harvested, beads removed, stained with antibodies and subjected to flow cytometry. For transduction efficiency (Top panels), cells were pre-gated on single live cells. For immune phenotype (middle and bottom panels), cells were pregated on CD3+CAR+ cells. S11

12 Supplemental Figure 6. Differential gene expression of CD4+ m19-humbbz CAR T cells. T cells with the m19z,, or m19-humbbz CAR were incubated with 3T3-mCD19 AAPC at 1:1 E:T ratio, FACS-sorted, and lysed to isolate RNA. Each group of CAR T cells was transduced, stimulated, and sorted independently in biologic triplicates. (A) PCA of mouse CD19-targeted CAR T cells stimulated with antigen. (B) Venn Diagram demonstrating the number of genes differentially expressed in m19-humbbz CAR T cells compared to m19z and CAR T cells. (C) GSEA demonstrates gene sets correlating to NF-κ B regulatory pathways are differentially expressed in m19-humbbz CAR T cells versus m19z or CAR T cells. S12

13 Untrans h19bbz mut1 mut2 mut3 mut4 CD3 CD8 CAR CD4 Supplemental Figure 7. CAR expression and CD4/CD8 subsets of human CD19 targeted CAR T cells for Figure 5E-G. For CAR expression (top), cells were pre-gated on single live cells. For immune phenotype (bottom), cells were pre-gated on CD3+CAR+ cells. S13

14 m19 z WT WT TRAF1 -/- m19-humbbz WT m19-humbbz TRAF1-/- CD3 CD8 CD4 CD44 GFP CD62L Supplemental Figure 8. Transduction efficiency and immune phenotype of mcd19 targeted wild type (WT) and Traf1 -/- CAR T cells used for in vivo study (Figure 6D). Day 4 transduced cells were harvested, beads removed, stained with antibodies and subjected to flow cytometry. For transduction efficiency (top panel), cells were pre-gated on single live cells. For CD4/CD8 subsets (middle panel) and memory subsets (bottom panel) cells were pre-gated on single live CAR T (CD3+CAR+) cells. S14

15 A B C Luminescence **** **** m19 z m19-humbbz mut1 CD8+IFN + CAR T% 5% 4% 3% 2% 1% % m19 z ** * m19-humbbz mut1 CD8+Bclxl+ CAR T(%) 4% 3% 2% 1% % m19 z m19z m19-humbbz D m19-humbbz mut1 ns ** * mut1 CAR E F G 2 * 4 * ns ns 15 * ** 3 *** **** 1 2 B cell/μl blood CD3 5 m19-humbbz mut1 CAR T cell/μl blood 1 m19-humbbz mut1 Donor T cell/μl blood ns m19-humbbz ns ** *** mut1 Supplemental Figure 9. Mutated CAR T cells have increased NF-k B signaling, improved cytokine production, anti-apoptosis, and in vivo function. (A) NF-k B signaling in mcd19 targeted CAR T cells. CAR T cells derived from NF-k B-RE-luc transgenic mice were cocultured with 3T3-mCD19 for 4 hr. Cell lysates were prepared and subjected to a luciferase assay. Bioluminescence was measured and correlates to NF-k B signaling. Data are representative of three independent experiments. (B) Intracellular IFNg and (C) BCL-XL expression in CD8+ CAR T cells stimulated with 3T3-mCD19. Data are representative of two independent experiments done in triplicate. (D) CAR expression in T cells used for in vivo study below. Cells were pre-gated on single live cells. (E) B cells (CD19+B22+) in the blood 1 week after CAR T cells injection. (F) CAR T cells (CD3+GFP+) in the blood 1 week after CAR T cells injection. (G) Donor T cells (CD3+Thy1.1+) in the blood 1 week after CAR T cells injection. (E-G) Each dot indicates one mouse (n=1 per group). Bar graph shown as mean ± SD. All data, unpaired t test. *p<.5; **p<.1; ***p<.1; ****p<.1; ns, not significant. S15

16 A Untrans h19bbz +TRAF1 +TRAF2 +TRAF3 TRAF TRAF CAR B Untrans h19bbz +TRAF1 +TRAF2 +TRAF3 C IFN (pg/ml) CAR h19bbz +TF1 ns +TF2 +TF3 TNF pg/ml h19bbz +TF1 +TF2 +TF3 IL2 pg/ml h19bbz +TF1 +TF2 +TF3 IL6 pg/ml h19bbz +TF1 +TF2 +TF3 Supplemental Figure 1. TRAF and CAR co-expression in human CD19-targeted CAR T cells. (A) CAR and TRAF expression in hcd19 targeted CAR T cells before antigen stimulation for viability, proliferation, and cytotoxicity assays (Figure 7B-D). Cells were pre-gated on single live cells. Data are one representative of 3 healthy donors. (B) CAR and TRAF expression in hcd19 CAR T cells 3 days after stimulation with 3T3-hCD19. Cells were pre-gated on single live cells. Data are from one experiment in triplicate. Numbers indicate percentages of gated cells. (C) Cytokine production. CAR T cells were activated on 3T3-hCD19 at 1:1 ratio. After 24 hr supernatant were harvested and cytokines were measured by ELLA. Bar graphs shown as mean ± SD. Data are one representative of 3 different healthy donors in triplicate. All data, unpaired t test. ns, not significant. S16

17 A B C 6 1% Survival 5% % Days after NALM6-GL Untrans h19bbz +TF2 * ** ns 2 NALM6/μl blood1 Untrans Untrans h19bbz D E F 1% Untrans h19bbz 8 +TF2 6 5% 4 ns Survival % Days after NALM6-GL ** *** ns CD19+ cells/μl blood 4 2 h19bbz ns +TF2 +TF2 Total counts Total counts Untrans Untrans h19bbz h19bbz ns +TF2 * +TF2 Supplemental Figure 11. TRAF2 over-expressed h19bbz CAR T cells show similar in vivo efficacy to h19bbz CAR T cells in an aggressive leukemia model. NSG mice were implanted with leukemia by i.v. injecting 5x1 5 NALM6-GL cells. Four days later, mice were i.v. injected with 3x1 5-1x1 6 CAR T cells. Blood samples were collected weekly. Leukemia burden was evaluated weekly using bioluminescence imaging. Survival was monitored. For (A-C), TRAF2 was co-transduced to make CAR T cells, n=15 total, and data are from one experiment. For (D-F), a bicistronic construct expressing CAR and TRAF2 was transduced to make CAR T cells. Survival data are pooled from two independent experiments (n=26), and counts are from one experiment. Each dot indicates a mouse. Survival curve, logrank test; all other data, unpaired t test. *p<.5; **p<.1; ns, not significant. S17

18 Supplemental Table 1. Probesets increased in m19z and vs. CAR T cells Probeset ID Gene Symbol Fold- Change Probeset ID Gene Symbol Fold- Change _at Gzmf _at Serpinb _at Serpinb9b _at Plek _at Ccl _a_at Klrk _at Il _at Paqr _at Gstm _a_at P2ry _at Klra _a_at Ctbp _a_at Dhrs _a_at Metrnl _at Rgs _at Gpd _at Pdcd _at Ccr _at Gng _at Tmem _at Gzmg _at Gzmd _at Myo _at Sh3bp _at Irf _s_at S1a _at Serpine _at Swap _at Rgs _at Ccr _at Irf _a_at Bcat _at Nrgn _x_at Gzmd _at Hhex _at Prss _at Myo _at Spin _at Serpinb1b _a_at Ehd _at Ifng _at Spry _at Cd _at Upp _a_at Prf _at Tmem _at Klrg _at Sypl _at Havcr _s_at Pkp _at Gpx _at Sypl _a_at Lat _a_at Cdkn2a _at Serpinb6b _at Cd _at Klrb1c _at Adssl _a_at Nr4a _at Pkd _at Plek _at Myo1e _at Mcpt _s_at Sypl _at Spp _at Galnt _a_at Gem _x_at Myb _a_at Tnfrsf _at Cyp _at Scel _a_at Myb _at Sult2b _a_at Zbtb _x_at Gzme _a_at Myb _at Lhx _a_at Anxa2 2.3 S18

19 _at Gzmc _at Gipc _a_at Klf _at Ctbp _at Cd _at Nek _at _a_at Bcl2l _at Slc17a _at Rai _a_at Tjp _a_at Uxs _at LOC _at Kcna _a_at Ccl _at Exo _at Gzmd _at Tmem _a_at Rgs _a_at Bmp _at Epas _a_at Txk _at Slc17a _at Litaf _at Plek _at Litaf _at Fcer1g _a_at Dusp _at Oit _at Gm _at Tnfrsf _a_at Bspry _at Srgap _a_at Lyn _at Ccl _at Eomes _at Rgs _a_at Ypel _a_at Cd _s_at Ypel _at Lgals _at Cables _a_at Ier _a_at Rnf _a_at Slc37a _x_at Klra _at Tubb _s_at Klhdc _at Xcl _s_at Mt _at Bmp _a_at Gm _at Cd _a_at Abcb _at Dsc _at Pdcd1lg _at Tmem2 2.9 S19

20 Supplemental Table 2. Probesets increased in vs. m19z and CAR T cells Probeset ID Gene Symbol Fold- Change Probeset ID Gene Symbol Fold- Change 14231_at Fos _at Mbd _at Dusp _at Tlr _s_at Igfbp _at St8sia _at Cox7c _a_at Snord _at Btf _a_at Morc _a_at Zfp _at Cacnb _s_at Cox7c _at Gm _at Ubl _at Zfp _s_at Ighm _at Fam11b _at Klf _x_at Gm _at Actn _at Ppp1r15a _at Tcf _at Ndufa _a_at Igfbp _at Gm _at Nsg _at St8sia _at LOC _at Eif2s _a_at Slamf _at Itgb _a_at Ighm _at Rbm _at Cyp2s _at Actn _x_at H2-Aa _at Nfkbia _at Tcf _at Abca _at Fosb _at Cmah _at Klf _a_at Ssbp _at Itgae _at Itga _at Gm _at Gm _at Jun _at Adcy _at C _s_at Dbp _at Tdrp _at Srsf _a_at Ifi _x_at Stx1a _s_at Trib _at Psma _s_at Nfkbia _at Sfswap _at Ifit _at Tmem _at Itga _x_at Gm _at Nfkbia _x_at Snhg _at Slamf _at Cd _at Rbm _a_at Fbxl S2

21 Supplemental Table 3. Probesets differentially expressed in m19z vs. CAR T cells Probeset ID Gene Symbol Fold- Change Probeset ID Gene Symbol Fold- Change _x_at Sdc _at Crtam _a_at Ugt1a _at Smpdl3b _x_at LOC _x_at _x_at Ceacam _a_at Endod _at Tfpi _at Fdps _at Sdc _at Zeb _at Penk _at Pglyrp _s_at Klrb1b _at Tfrc _s_at Ugt1a _at Pdia _at Antxr _at Tmbim _at Gstt _s_at Ceacam _at Ccl _at Cyp _at Ccl _at Klhdc _a_at Klrc _s_at Ceacam _at Mycn _at Sema7a _at Khdc1a _at Phactr _at Coro2a _at Lrrk _at Vcl _at Napsa _a_at Rgs _at Endod _at Itih _at Car _a_at Chn _at Mtm _a_at Rgs _at Sdf2l _s_at Ceacam _x_at Vat _at Nrp _at Cd _x_at Ceacam _a_at Dusp _s_at Ltbp _at Plk _x_at Klra _at Cbx _at Pkp _x_at Endod _at Klrc _at Myo _a_at Tfrc _at Lag _at Nrp _a_at LOC _at Fam11c _a_at Hmgcs _at _at Mt _a_at Ugt1a _a_at Cd _at Scd _at Ndrg _x_at Ceacam _at Shmt _at Tg _at Uhrf1 2. X686_5_at _at Hbegf _a_at Tacc _at Cdkn2b 2. S21

22 _at Chn _at Ndc _at Klrb1b _at Hist3h2ba _at Lad _at Cd _at Alcam _at Taf _at Zfp _at Rapgef _at Slc25a _at Myl _at Ltbp _at 18132O8Rik _at Nrp _at Enah -2.3 X57349_5_at Tfrc _at Nfkbiz _at Klrb1a _at Pdlim _at Calcb 2.3 S22

23 Supplemental Table 4. Probesets differentially expressed in vs. CAR T cells Probeset ID Gene Symbol Fold- Change Probeset ID Gene Symbol Fold- Change _at Esm _a_at Shisa _s_at Eps _a_at Elmo _a_at Atp1b _at Ltb4r _at Ctsg _a_at Tnfsf _at Eps _at Alpk _at Sepn _a_at Pdzrn _a_at Lmna _at Tirap _at Clnk _a_at Cysltr _at Gzmk _a_at Naip _at Basp _at Pilrb _at Gcnt _a_at Eps _at E2Ri _at Tiam _at Stk32c _at Osbpl1a _at Tdgf _a_at Slc37a _a_at Rhoq _at Zak _a_at Tcf _x_at Cnn _at Tcf _a_at Cmpk _at Tspan _at Nfkbia _at Tyrobp _x_at Cnn _at Col6a _at Ldlrap _x_at Arhgap _a_at Capn _a_at Fndc _at Oas _a_at Fhl _a_at Rbm _at Shisa _at Sh3bp _at Marcks _at Ltb _s_at Gcnt _at Srsf _at Piwil _at Frat _x_at Marcks _a_at Hist1h1c _at Serpinb1a _at Thra _a_at Cma _at Mphosph _at Gcnt _at Ifit _at Tnfsf _at Klf _at Chpt _at Pou6f _at Cpa _s_at H2-Ab _a_at Ptger _x_at Tgm _at Hunk _at Tlr _at Marcks _x_at Nfix _a_at Frmd4b _at Ptgir _at Eomes _at S23

24 142364_at Gpr _x_at Tgm _x_at Tcf _a_at Enpp _at Aoah _a_at Sfswap _at Plxdc _at Abca _at Acvr _at Ncf _s_at Gcnt _a_at Serinc _at Dusp _at Setd1a _at Eps8l _at Slamf _at Ptgr _at Rabgap1l _at Prnp _at Gdap _at Tcf _at Sh3bp _at P2ry _at _a_at H2-Q _at Fas _at Chpt _a_at Cd _s_at Fam84a _a_at Slamf _at Dab2ip _at Dbp _at Tcf _at Ssbp _at S1a _at Il6ra _at Sox _at Thra _x_at Hba-a _a_at Idh _at Errfi _a_at Tspan _a_at Chpt _a_at Tnfrsf _at Plxdc _at Ccl _at Ell _at Ccr _x_at Marcks _at Cd _at Marcks _a_at Nfix _at Ccr _at Cd S24

25 SUPPLEMENTAL REFERENCES 1. Trapnell C, Pachter L, and Salzberg SL. TopHat: discovering splice junctions with RNA-Seq. Bioinformatics. 29;25(9): Anders S, Pyl PT, and Huber W. HTSeq--a Python framework to work with high-throughput sequencing data. Bioinformatics. 215;31(2): Anders S, and Huber W. Differential expression analysis for sequence count data. Genome Biol. 21;11(1):R Wang L, Wang S, and Li W. RSeQC: quality control of RNA-seq experiments. Bioinformatics. 212;28(16): Markham NO, Cooper T, Goff M, Gribben EM, Carnahan RH, and Reynolds AB. Monoclonal antibodies to DIPA: a novel binding partner of p12-catenin isoform 1. Hybridoma (Larchmt). 212;31(4): Zhao Z, Condomines M, van der Stegen SJC, Perna F, Kloss CC, Gunset G, Plotkin J, and Sadelain M. Structural Design of Engineered Costimulation Determines Tumor Rejection Kinetics and Persistence of CAR T Cells. Cancer Cell. 215;28(4): S25