GATA-3 is a proto-oncogene in T-cell lymphoproliferative neoplasms

Neoplasms originating from thymic T-cell progenitors and post-thymic mature T-cell subsets account for a minority of lymphoproliferative neoplasms. These T-cell derived neoplasms, while molecularly and genetically heterogeneous, exploit transcription factors and signaling pathways that are critically important in normal T-cell biology, including those implicated in antigen-, costimulatory-, and cytokine-receptor signaling. The transcription factor GATA-3 regulates the growth and proliferation of both immature and mature T cells and has recently been implicated in T-cell neoplasms, including the most common mature T-cell lymphoma observed in much of the Western world. Here we show that GATA-3 is a proto-oncogene across the spectrum of T-cell neoplasms, including those derived from T-cell progenitors and their mature progeny, and further define the transcriptional programs that are GATA-3 dependent, which include therapeutically targetable gene products. The discovery that p300-dependent acetylation regulates GATA-3 mediated transcription by attenuating DNA binding has novel therapeutic implications. As most patients afflicted with GATA-3 driven T-cell neoplasms will succumb to their disease within a few years of diagnosis, these findings suggest opportunities to improve outcomes for these patients.

#421403) then stained with appropriate antibodies. Cells were washed before flow analysis. Flow cytometry data was acquired using CyAn™ ADP Analyzer (Beckman Coulter) and the data were analyzed by Summit 5.3 or FlowJo. Mean fluorescence value was used for protein expression in cells. All antibodies are listed in Supplemental Data Table 4.

Immunohistochemistry
Lymphoma tissue sections were deparaffinized and stained on BOND RX automated stainer (Leica Biosystems) using Bond Polymer Refine Detection kit (Leica #DS9800). Sections were subjected to EDTA-based antigen retrieval (BOND ER2 solution #AR9961) for 40 min at 100°C. Non-specific binding was blocked by incubating sections with Background Sniper protein block (Biocare Medical #BS966) for 15 minutes. Sections were then incubated with anti-GATA-3 rabbit monoclonal antibody (Cell Signaling #5852) at 1:100 dilutions for 30 minutes followed by signal detection with polymer-HRP-labeled secondary antibody and DAB. Sections were counterstained with hematoxylin for 10 min, dehydrated and mounted with Micromount media (Leica Microsystems #3801730). For semi-quantitative analysis, the intensity of staining (no staining=0, dim=1, moderate=2, high=3) and percentage of cells staining (quantified in 10% increments) were multiplied to generate an H-score (range: 0-300). Cases with an h-score ≥100 were scored as GATA-3 positive.

Mouse models and in vivo experiments
Mouse studies were approved by the University Committee on Care and Use of Animals (UCUCA) and performed in accordance with guidelines established by the Unit for Laboratory Animal Medicine (ULAM) at University of Michigan. Mice were housed under specific-pathogen free conditions. Treatment allocation was randomized, and all animals in given experiments were included for analysis. P53 floxed, SNF5 floxed, and CD4-Cre mice were obtained from Jackson Laboratory. GATA-3 floxed mice were kindly provided by Dr. Doug Engel.(1) SNF5 mice were provided on a mixed background, but were backcrossed for at least 10 generations onto a B6 background. All F1 mice were genotyped using tail DNA (Invitrogen Animal Tissue Direct PCR Kit). Mice were monitored two times weekly for event-free survival (EFS), where an event was defined as the development of easily palpable hepatosplenomegaly, bulky lymphadenopathy (>5mm), or any other condition requiring anesthesia. For adoptive transfer experiment, 5×10 6 bulk splenocytes obtained from lymphoma-bearing donor mice were retroorbitally injected into 12-16 weeks old recipient C57BL/6 mice (Jackson Laboratory). We assumed that for recipients from a common donor, event times are normally distributed with a mean log(4) and standard deviation 0.3, thus assuring power at least 80% to detect a 60% relative change in EFS with the number of biologic and technical replicates included in the experiments performed. Any mice that developed periorbital tumors following adoptive transfer, likely due to technical inefficiency, and prior to study treatment, were excluded. Otherwise, mice were randomized (by cage assignment) to treatment, and investigators unblinded to treatment allocation. Where indicated, mice were treated with combined vincristine (0.5 mg/kg, i.p) and cyclophosphamide (40 mg/kg, i.p) weekly (two weeks after adoptive transfer) for 2-3 consecutive weeks, and mice followed for EFS. For the cell-line-and patient-derived xenografts studies, 2x10 6 (cell lines) or 5x10 6 (PDX) cells were injected subcutaneously on the shaved flanks of 12-16 weeks old female NSG mice (Jackson lab, strain #005557). The longest and shortest tumor dimension was measured once to twice weekly, and tumor volumes calculated using a modified ellipsoidal formula [volume = 0.5(length x width 2 )]. For the PDX model, tumor-bearing mice were randomized and treated with A-485 (100 mg/kg) or vehicle control intraperitoneally for four consecutive days, and then euthanized on day 5. The tumors and spleens were weighed, and cell lysates generated from tumors for immunoblotting.

Genotyping
Genomic DNA is isolated from the tails of mice using proteinase K (Qiagen, #19131) and cell lysis solutions as per the manufacturer's recommendations. After vortexing, tails are incubated overnight at 55 o C in a bead bath. The next day, protein precipitation solution (Promega, #A1120) is added, samples vortexed, and then incubated on ice for at least 5 minutes. After centrifugation, DNA is precipitated from supernatants by 100% isopropanol, and then washed with 70% ethanol. After drying in a heating block, the DNA is suspended in DNase free water for PCR using Taq DNA polymerase, primers, and PCR master mix (Invitrogen, #18038042). The primer sequences utilized are shown in the Extended Data Table 3.
(2) HEK293T cells were cultured in DMEM supplemented with 10% FBS. Other cells were cultured in RPMI1640 supplemented with 10% FBS, as previously described (3). CCRF-CEM and THP6 cells were maintained in the medium supplemented with 50 μM βmercaptoethanol (Sigma-Aldrich #M3148). Inducible expression in each engineered cell line was performed by addition of doxycycline (10 ng/ml) to the growth medium. All cell lines were mycoplasma free and independently authenticated by short tandem repeat (STR) profiling, performed by ATCC (data not shown), and immunophenotyping (data not shown). The status of p53 in/del/SNV for each cell line was determined using both CCLE and COSMIC databases. Nutlin-3 was purchased from Selleckchem (S1061). Cells were treated with 5 μM nutlin-3 for 16 hours at 37 °C. A-485 and MG132 were purchased from Selleckchem (S8740 and S2619). Cells were treated with 2 μM A485 and/or 10 μM MG132 for 6 hours at 37 °C. Chetomin was purchased from APExBIO (A4505). dCBP-1 was purchased from MedChemExpress (HY-134582). The cells were treated with 1 μM Chetomin for 6 hours or 1 μM dCBP-1 for 5 hours at 37 °C. For cell selection, puromycin (Takara #631306) was utilized at concentrations (0.25-1 μg /ml) appropriate for each cell line. For ex vivo studies using primary T-cell lymphoma specimens (Extended Data Table 2), malignant T cells were sorted using CD3 or CD4 microbeads (Miltenyi), as previously described. (3,4) Sorted cells were cultured with anti-CD3/CD28 Dynabeads (Thermo Fisher Scientific #11132D) at a 1:1 ratio, where indicated. CPI-818 was provided by Corvus Pharmaceuticals under a material transfer agreement. IL-10 sandwich ELISA was performed using cell-free supernatants (BioLegend).

BIO-ID Assay
GATA-3 cDNA was fused with biotin protein ligase harboring an R118G mutation (BirA) in the plVX-Tet-ON vector (Fig. S9) (5). A nuclear localization sequence (NLS) cDNA was similarly fused with biotin protein ligase (BirA-NLS). BirA-GATA-3 or BirA-NLS were ectopically expressed in Jurkat cell lines. BirA fusions were induced with doxycycline for 72 h, biotinylation was induced by adding 50 μM of Biotin for 24 h. Cell lysates were generated, and biotinylated proteins were precipitated utilizing streptavidin-sepharose beads (GE Healthcare #17-5113-01) before submission to MS/MS. GATA-3 binding partners were defined as those for which the mean ratio of BirA-GATA-3/BirA-NLS biotinylated peptides was >5 and the BirA-GATA-3/BirA-NLS ratio was >5 in at least 2 out of 3 technical replicates.

GATA-3 DNA-binding assay
HEK293T cells were transfected with indicated vectors. Nuclear extract was prepared freshly from HEK293T cells according to manufacturer's instruction (Thermo Fisher Scientific #78835). The nuclear extract was pre-cleared with ~7 μl Dynabeads for 1 hour at 4°C and then quantified by Pierce BCA protein assay. Single stranded DNA oligos probe containing a palindromic GATA site (AGAATGTAGCCCTGGACTTCTCCCGCTCGCTATCAGATAAGGCCTTATUUCGATAAGGCC TTATCTGATAGCGAGCGGGAGAAGTCCAGGGCTACATTCT) or a mutant GATA-3 site (AGAATGTAGCCCTGGACTTCTCCCGCTCGCAAAAATTTTAGGCCTTATUUCGATAAGGCCT AAAATTTTTGCGAGCGGGAGAAGTCCAGGGCTACATTCT) were synthesized by Integrated DNA Technologies (Coralville, IA). For the binding reaction, 10 nM DNA oligo probe, 100 μg of nuclear extract were added to the PBS buffer containing proteinase inhibitor in a final assay volume of 200 μl. The binding assay was carried out at 4°C for 1 hour, and then 1 μg GFP antibody (Santa Cruz #sc-9996) was added for overnight. Following day, 8 μl Dynabeads was added to the assay mixture for 2 hours at 4°C. Non-specific bindings were washed twice with 200 μl PBS-T (0.05% Tween-20 ) at 4°C. GATA-3 protein and probe complex were eluted in 35 μl buffer (0.1M Glycine, pH 2.8) and heated at 70°C for 10 min. GATA-3 DNA binding abilities were measured using real-time PCR, which was performed on the BIO-RAD C1000 Thermal Cycler PCR instrument with 2× SYBR Green (Alkali Scientific Inc. #QS2020). Primer sequences used in this study were as follows: 5'-AGAATGTAGCCCTGGACTTC-3' and 5'-TTCTCCCGCTCGCTATCA-3'. Ct value was normalized to Ct value from the sample transfected with pLVX-AcGFP GATA-3 without probe. The relative GATA-3 DNA binding abilities were expressed as a fold change of pLVX-AcGFP GATA-3 vs. the pLVX-AcGFP. Three independent experiments were performed.

RNA isolation, sequencing and quantitative RT-PCR
RNAs were extracted using RNeasy Mini Kit (Qiagen #74106), and on column DNase I digestion (Qiagen, RNase-free DNase set #79254) was performed during extraction to avoid DNA contamination. RNA concentration was determined by NanoDrop TM One/Onec Spectrophotometer (Thermo Fisher Scientific). RNA was assessed for quality using the TapeStation (Agilent, Santa Clara, CA). Samples were prepared using the NEBNext Ultra II Directional RNA Library Prep Kit for Illumina (NEB #E7760L), Ribo depletion Module NEBNext rRNA Human/Mouse/Rat (NEB #E6310X) and NEBNext Multiplex Oligos for Illumina Unique dual (NEB #E6440L), where 100 ng of total RNA was ribosomal depleted using the rRNA Depletion module. The rRNA-depleted RNA is then fragmented 7 minutes determined by RIN (RNA Integrity Number) of input RNA as per protocol, and copied into first strand cDNA using reverse transcriptase and dUTP mix. Samples undergo end repair and dA-Tailing step followed by ligation of adapters. The products are purified and enriched by PCR to create the final cDNA library. Final libraries were checked for quality and quantity by TapeStation (Agilent) and Qubit (Thermo Fisher Scientific). This pool was subjected to 151bp paired-end sequencing according to the manufacturer's protocol (Illumina NovaSeq). Bcl2fastq2 Conversion Software (Illumina) was used to generate de-multiplexed Fastq files. For quantitative RT-PCR, complementary DNA (cDNA) was prepared using the QuantiTect Reverse Transcription Kit (Qiagen #205311). qPCR was performed using the Radiant SYBR Green H-ROX qPCR Kits (Alkali Scientific Inc. #QS2020) and analyzed using the ΔΔCt method with control samples set as 1. Primers used in this study were in the Extended Data Table 3.

Nanostring and immune gene expression profiling
Nanostring technology was performed as previously described. Briefly, RNA were extracted using RNeasy FFPE kit (Qiagen #73504) and measured using a Nanodrop 1000 instrument (ThermoFisher), 150 ng total RNA from each patient was analyzed using the Nanostring platform and the nCounter PanCancer Immune Profiling Panel (NS_CANCERIMMUNE_C2929) with downstream normalization and gene quantification performed using nSolver (v4.0).

Chromatin immunoprecipitation (ChIP)
ChIP was performed using SimpleChIP® Plus Enzymatic Chromatin IP Kit (Cell Signaling #9004) with minor modifications. Briefly, 5-10 million cells per IP were fixed using formaldehyde (1% final concentration) for 10 min at room temperature and were stopped with 125 mM glycine addition for 5 min at room temperature. After washing, cells were lysed in prepared Buffer A for 10 min at room temperature. After lysing, nuclei were collected (500 × g, 4°C, 3 min) and digested with prepared Buffer B and Micrococcal Nuclease for 20 min at 37 °C. After adding 50mM EDTA to stop digestion, nuclei were collected (16,000× g, 4°C, 1 min) and resuspend in prepared ChIP buffer. Nuclei were then sonicated with two pulses of 40 seconds each and 30 s incubation on wet ice at setting 4 using an ultrasonic cell disruptor (Microson) with a 18-inch probe. Chromatin was clarified (16,000× g, 4°C, 10 min) and incubated with GATA-3 and IgG antibody overnight. Chromatin was then incubated with 30 µL Dynabeads Protein G (Invitrogen #10004D) for 4 hours and washed with low-salt washing buffer three times and high-salt washing buffer once. Elution and de-crosslinking were performed overnight in 150 elution buffer (300 mM NaCl, 5 mM DTT and 0.1% SDS in TE buffer, pH8.0) at 65 °C, and RNA and proteins were digested by adding RNase A and proteinase K, respectively. Eluted samples were purified by QIAquick TM PCR Purification Kit (Qiagen #28104) and were ready for high-throughput sequencing. For ChIP-seq library preparation, NEBNext ChIP-Seq Library Prep Master Mix Set for Illumina (NEB #E6240) was used according to the manual's instructions. Library quality was determined by TapeStation (Aglient) prior to being sequenced on the Illumina NovaSeq-6000 platform. For ChIP-qPCR, same amount of purified DNA was tested using the Radiant SYBR Green H-ROX qPCR Kits (Alkali Scientific Inc. #QS2020) and analyzed using the ΔΔCt method with control samples set as 1. Primers used in this study are in the Extended Data Table 3. Three independent experiments were performed.

Cell viability
Cell viability was analyzed by RealTime-Glo TM MT Cell Viability Assay (Promega #G9711) according to manufacturer's instructions. Briefly, 5,000 cells were mixed with substrate and NanoLuc TM enzyme and seeded to 96-well plate in triplicate or quadruplicate. Luminescence was measure from day 0 to day 7. Data shown was performed in at least 3 independent experiments.

Flow Cytometry
For cell surface staining, cells were stained with fluorochrome-conjugated antibodies and washed before flow analysis. For intracellular staining, cells were fixed and permed using FOXP3 Fix/Perm Buffer Set (BioLegend #421403) then stained with appropriate antibodies. Cells were washed before flow analysis. Flow cytometry data was acquired using CyAn™ ADP Analyzer (Beckman Coulter) and the data were analyzed by Summit 5.3 or FlowJo. Mean fluorescence value was used for protein expression in cells. All antibodies are listed in Extended Data Table 4.

Fluorescence in-situ hybridization (FISH)
Evaluation for TP53 deletions was performed on formalin-fixed, paraffin-embedded sections using FISH with a probe set that includes a TP53 locus-specific identifier (17p13.1) and a chromosome 17 chromosome enumeration probe (Abbott, Abbott Park, IL). All cases were evaluated for adequacy by a pathologist board-certified in hematopathology (N.B.). 100 cells from each case were required for evaluation with a ratio of TP53:CEP17 < 0.8 required for evidence of deletion.

Protein Identification by LC-Tandem MS
In-gel digestion: HEK293T cells were transfected with GFP-tagged GATA-3. 2 days post-transfection, cells were collected for IP with GFP antibodies. The samples were separated by SDS-PAGE and stained with SimplyBlue SafeStain (Thermo Fisher Scientific #LC6060). The protein samples were processed and analyzed at the Mass Spectrometry Facility of the Department of Pathology at the University of Michigan. Gel slice corresponding to 76 MW (GFP was 26 MW and GATA-3 was 49 MW) was destained with 30% methanol for 4 h. Upon reduction (10 mM DTT) and alklylation (65 mM 2-Chloroacetamide) of the cysteines, proteins were digested overnight with 500 ng of sequencing grade, modified trypsin (Promega) at 37°C. Peptides were extracted by incubating the gel with 150 µl of 50% acetonitrile/0.1% TFA for 30 min at room temperature. A second extraction with 150 µl of 100% acetonitrile/0.1% TFA was also performed. Both extracts were combined and dried in a vacufuge (Eppendorf).

Database Search:
Proteins were identified by searching the data against Homo sapiens protein database (20353 entries; reviewed; downloaded on 06/02/2019) using Proteome Discoverer (v2.3, Thermo Scientific). Search parameters included MS1 mass tolerance of 10 ppm and fragment tolerance of 0.2 Da; two missed cleavages were allowed; carbamidimethylation of cysteine was considered as fixed modification. Oxidation of methionine, deamidation of asparagine and glutamine, acetylation on lysine, arginine, serine, threonine, and tyrosine were considered as potential modifications. FixedPSM validator of the Proteome Discoverer was used to retain only higher quality PSMs (ΔCn ≥ 0.05). All PSMs corresponding to the modified peptides were verified manually.

Snapshot metabolomics
Metabolites were extracted from cells by adding cold 80% methanol, incubating at -80°C for 10 minutes, followed by centrifugation at 10,000xg for 10 minutes at 4°C. The resulting metabolite supernatant were collected. Metabolite extracts were normalized to cell number, and the normalized fraction was transferred to a fresh 1.5mL tube and lyophilized by speedvac. Dried metabolite pellets were resuspended in a 50:50 mixture of MeOH and water.
Liquid chromatography-based targeted tandem mass spectrometry (LC-MS/MS)-based metabolomics were performed and the data analyzed as previously described. (7,8) In brief, samples were run on an Agilent 1290 Infinity II LC -6470 Triple Quadrupole (QqQ) tandem mass spectrometer system consisting of the 1290 Infinity II LC Flexible Pump (Quaternary Pump), the 1290 Infinity II Multisampler, the 1290 Infinity II Multicolumn Thermostat with 6 port valve and the 6470 triple quad mass spectrometer. Agilent Masshunter Workstation Software LC/MS Data Acquisition for 6400 Series Triple Quadrupole MS with Version B.08.02 was used for compound optimization, calibration, and data acquisition.

Molecular dynamics simulations
X-ray crystal structure atomic coordinates of the C-terminal zinc finger of mouse GATA-3 bound to DNA were downloaded from the Protein Data Bank (PDB ID: 3DFV). A model of GATA-3 bound in the extended c-terminal conformation was created by taking the coordinates of DNA and protein chain D, with coordinates for residues 355 to 366 taken from conformation B. A model with acetylated Lys358 was also constructed by manually building the acetyl group using the Build tool in Maestro (Schrodinger Release 2020-3). The models were processed using Protein Preparation Wizard (Schrodinger Release 2020-3). Hydrogen atoms were added, bond orders were assigned, zero-order bonds to metals were assigned, and protein termini were capped with amide and acetyl groups. Protonation states were determined using PROPKA, the hydrogen bonding network was optimized, and the models were energy minimized with heavy atoms constrained to within 0.3 Å of initial position. Each model was placed in a volumeminimized orthorhombic TIP3P water box with distance buffers of 10 Å • 10 Å • 10 Å with periodic boundaries. In each case, the system was neutralized with K+ ions and salt concentration was set to 0.15 M KCl, and the OPLS3e force field was assigned. 500 ns of molecular dynamics simulation were carried out using DesmondGPU (Schrodinger Release 2020-3) for the Lys358 acetylated and non-acetylated models after treatment with the default equilibration protocol. Production phase of each simulation was conducted at 300 K in an NPT ensemble at a pressure of 1.01325 bar using the Langevin barostat with a 1.0 ps relaxation time and the Nose-Hoover thermostat with a 2 ps relaxation time with coupling style set to isotropic. The RESPA integrator was used with the time step set to 2 fs with initial randomized velocities set using the default seed (2007). Coulombic interactions were calculated using the useries method and the short-range cut-off was set to 9.0 Å. The recording interval of the trajectory was set to 100 ps.
GATA-3 were imaged continuously with a 488nm laser in TIRF on a Nikon Eclipse Ti2-E inverted microscope (TIRF 100x, 1.49 NA objective lens) with 1.5x external magnification. The images were collected by a Photometrics Prime 95B sCMOS camera. GATA-3 images of fully immobilized Karpas 299 cells were acquired for 500-1000 frames at a power density of 133 Wcm -2 laser power and 200 ms exposure.
GATA-3-EGFP molecules that immobilized for at least 1 second (5 frames) were used for analysis. Dwell times were determined by calculating the number of frames individual GATA-3 stayed in frame and immobilized in the selected ROI, followed by calculating the time length in seconds. An unpaired t-test was used to determine significance.

Integrated analysis
Overlap analysis between differentially expressed genes identified by DESeq2 and annotated genes identified by GREAT was firstly made. Then bed files were analyzed with whole gene expression by BETA (v1.0.7) to predict direct target genes and GATA's activating or repressive function. (23) For gene dependency analysis, GATA-3 dependency score from Depmap website (https://depmap.org/portal/) were collected and their rank in individual cell line were calculated using in-house shell scripts. Dot plots and bar plots were generated by ggplot2 (v3.3.5).