Cathepsin G Is Broadly Expressed in Acute Myeloid Leukemia and Is an Effective

Cathepsin G is broadly expressed in acute myeloid leukemia and is an effective immunotherapeutic target

Alatrash et al.

Supplementary Information

Supplementary Methods
Supplementary Figure legends
Supplementary References
Supplementary Figures S1-S6 (in separately attached file)

Supplementary Methods

Patient samples and cell lines

Patient and healthy donor (HD) samples were obtained after appropriate informed consent through institutional review board approved protocols (Lab99-062, Lab01-473 and Lab05-0654) at the MD Anderson Cancer Center (MDACC). U937 (myelomonoblastic leukemia), HL-60 (promyelocytic leukemia) and T2 (B-cell/T-cell hybridoma) cell lines were obtained from American Type Culture Collection (Manassas, VA).Cell lines were cultured in RPMI-1640 (HyClone, Logan, Utah) supplemented with 10% FBS (Sigma-Aldrich, St. Louis, MO), Penicillin (100U/ml)/Streptomycin (100 g/ml; HyClone), and were kept in 5% CO2 at 37C. HLA-A2*0201 transfection was performed as described previously using lentiviral vectors.1Briefly, the cell lines were transduced with lentiviral vectors encoding HLA-A*02:01 (HLA-A2) in XVivo-15 serum free media (Cambrex, East Rutherford, NJ) at MOI 5 in polybrene. HLA-A2 expression was verified by flow cytometry prior to using the cell lines in assays.Cell lines were authenticated by DNA fingerprinting at MDACC within 6 months of use in experiments.

The AML patient samples used for peptide elution were selected based on their AML subtype, percent blasts, sample viability, and human leukocyte antigen (HLA) status. Healthy individual and patient PB mononuclear cells (PBMC) and BM were enriched using standard Histopaque 1077 (Sigma-Aldrich) gradient centrifugation.

The AML patient samples used in the reverse-phase protein array (RPPA) analysis have been described previously.2, 3 Peripheral blood (PB) and bone marrow (BM) specimens were collected from 511 patients with newly diagnosed AML, 21 patients with newly diagnosed APL, and healthy donors evaluated at MDACC from September 1999 to March 2007. A paired relapse sample was also collected when available. The patient population included a high percentage of patients with unfavorable cytogenetics (49%) and/or an antecedent hematologic disorder (40%). Of the 511 patients, 415 were treated at MDACC and were evaluable. Among the treated patients, 277 received regimens that contained high-dose cytarabine (Ara-C), 35 received standard-dose Ara-C, and 8 received low-dose Ara-C. Most of the Ara-C-treated patients also received other treatments and a variety of regimens were given to the other 95 patients.

Expansion of peptide specific cytotoxic T lymphocytes (CTL)

CG1-specific (CG1-CTL) and HIV-specific (HIV-CTL) were generated using a modified dendritic cell (DC)-based CTL expansion method.4Briefly, HLA-A2+ healthy donor PBMC were adhered on 6-well plates in Macrophage Serum Free Medium (Gibco, Carlsbad, CA) at 37oC. Cells that remained in suspension following the adherence step (i.e. lymphocytes) were removed after 4 hours. These cells were then cultured in RPMI 10% FBS for 5 daysand stimulated with IL-7 (10 ng/mL) (BioLegends, San Diego, CA) and IL-2 (10 ng/mL) (R&D, Minneapolis, MN). Adherent cells from the initial step were matured into dendritic cells (DC) by addition of IL-4 (50 ng/mL) (BioLegends), GM-CSF (100 ng/mL) (Sanofi, Bridgewater, NJ) and TNF- α (25 ng/mL) (R&D) for 5 days, and were then detached and pulsed with CG1 or HIV-peptides (40µg/mL) for 4-hours at 37oC. The lymphocytes and DC were then co-cultured in RPMI 10% FBS, and re-stimulated with IL-2 (25 ng/mL) and IL-7 (10 ng/mL) for 1 week to allow for antigen specific CTL proliferation. On day 14, cells were harvested, specificity was confirmed using calcein-AM (BD Biosciences, San Jose, CA) cytotoxicity assays and cells were then used in the in vivo animal studies. For the murine models, the same healthy donor PBMC were used to expand both CG1-CTL and HIV-CTL to control for background killing caused by allogeneic effects.

Cell-mediated cytotoxicity assay

A standard cytotoxicity assay was used to determine specificity of the expanded CTL for their targets, as described previously.5, 6Briefly, 1000 target T2 cells were pulsed with CG1 peptide or irrelevant PR1 peptide (VLQELNVTV) resuspended in 10 µl (1.0 x105 cells/ml) of RPMI-1640 and stained with calcein-AM (BD Biosciences) for 90 minutes at 37 C. Cells were then washed 3 times with RPMI-1640 and then co-incubated with 10 µl of peptide-specific CTLs at varying effector to target (E:T) ratios.Target and effector co-cultures were incubated for 4-hours at 37˚C in 5% CO2. Trypan blue was then added to each well and fluorescence was measured using an automated FLx800 Fluorescence Reader (Bio-Tek, Winooski, VT). Percent specific cytotoxicity was calculated as follows: (1-(Fluorescence Target+Effector – Fluorescencemedia)/(FluorescenceTarget alone – Fluorescence media)) x 100.

In vivo Mouse Methods

NOD/SCID gamma (NSG) mice7(4-6 week old females) were purchased from Jackson Laboratory (Bar Harbor, ME). HLA-A*0201 primary AML samples with a high leukemia burden and U937-A2 cells were administered via tail vein at a dose of 1 x 106 to 1 x 107 cells to sublethally irradiated (250 cGy) mice. After confirming engraftment (1-5% PB HLA-A2+/human (h) CD45+ cells)8,CG1- or HIV-CTL (0.5 x 106) were administered to mice via tail vein. Mice were monitored for clinical GVHD and AML engraftment 3 times/week.9 Mice were sacrificed at approximately 7 weeks following treatment or when mice became moribund. Bone marrow (BM) was processed using standard methodology and analyzed for residual human AML by flow cytometry as previously described.8, 9 The following flow cytometry antibodies were used to identify the leukemia: CD13, CD33 (BD Biosciences, Sparks, MD), CD3, hCD45 (BioLegend, San Diego, CA), mouse (m)CD45 (eBioscience, San Diego, CA) and GFP.

Peptide elution and mass spectrometry analysis

For peptide elution, more than 5 x 107 cells from patient samples and leukemia cell lines were lysed in standard lysis buffer. Cell lysate was subjected to ultra-centrifugation to remove cellular debris. The pan-HLA class I antibody (clone W6/32, Novus Biologicals, Littleton, CO), which specifically binds HLA class I molecules10was added to the cell lysate and was incubated overnight. Following incubation with antibody, Protein A/G resin was added to the lysate to precipitate bound HLA molecules. The bound protein was transferred to AminoLink Coupling Resin and Immobilization columns and eluted with 0.1% aqueous acetic acid solution and the concentration of acetic acid in pooled fractions was adjusted to 10% to separate the HLA molecule from β2-microglobulin(β2m)and the peptides. The solution was filtered using a 3 kDa molecular weight cut-off (MWCO) membrane and the filtrate was subjected to liquid chromatography (LC) tandem mass spectrometry (MS/MS).

For discovery phase tandem MS/MS, eluted HLA class I-bound peptides were injected onto a high-sensitivity high performance liquid chromatography (HPLC) system (Dionex 3000 RSLC, Thermo Fisher Scientific, Grand Island, NJ), separated by reversed-phase chromatography in 0.1% formic acid water-acetonitrile on 2.6 micron XB-C18 (Kinetex, Phenomenex, Torrance, CA) and analyzed on an Orbitrap Elite mass spectrometer (Thermo Scientific) using data-dependent acquisition. The Mascot algorithm searched acquired MS/MS spectra against the SwissProt complete human protein database. Search results were cross-referenced with the appropriate HLA-binding specificities using NetMHC 3.4 ( Peptides eluted were compared to synthetic peptides by obtaining retention-time windows for the synthetic peptide then targeted methods were constructed using mass windows of 1 Da around each m/z.

Colony forming unit (CFU) assays

We used CFU assays to determine the effects of CG1-CTL on healthy donor BM (HDBM).11 Briefly, HDBM was cultured overnight prior to co-culturing with CTLs. On day 1, 2 x 104HDBM cells were co-cultured with 1 x 105CTLs for 4 hours in a 48 well plate. After co-culture, the HDBM and CTLs were re-suspended in MethoCultTMH4034 Optimum Human Medium (StemCell Technologies, Vancouver, BC) and IMDM medium supplemented with 2% FBS. Cells were then plated on low adhesion 6 well plates (SmartDishTM6-well plates, StemCell Technologies) in triplicate. Cells were incubated for 1 week to allow colonies to form. On days 7 and 14, colonies were counted using an inverted light microscope.

Sample preparation and cathepsin G reverse-phase protein array (RPPA)

Samples were collected for RPPA and enriched for leukemic cells using Ficoll-Hypaque (Sigma-Aldrich) density-gradient separation to yield a mononuclear fraction, followed by CD3/CD19 depletion using magnetic antibody-conjugated cell separation (Miltenyi Biotec, San Diego, CA) to remove T and B cells if they were calculated to be more than 5% on the basis of the differential. After this protocol, the blast purity reached approximately 98% as determined in approximately 20 samples by flow cytometry. The samples were normalized to a concentration of 1 X 104 cells/L and a whole-cell lysate was prepared. RPPA was carried out following the methodology and validation of the technique described fully in previous publications.12, 13 Briefly, patient samples were printed in 5 serial dilutions onto slides along with normalization and expression controls. Slides were probed with 232 strictly validated primary antibodies, including one against cathepsin G (Abcam ab8816, Cambridge, MA). The stained slides were analyzed using MicroVigene software (Vigene Tech, Carlisle, MA) to produce quantified data.

CG1 Tetramer staining

To confirm the frequency and avidity of expanded CG1-CTL, the following fluorescently conjugated CG1/HLA-A2 tetramer (Baylor MHC Tetramer Core, Baylor College of Medicine) and antibodies were added to each sample: PE-CG1/HLA-A*0201 tetramer; APC/Cy7-anti-CD8; APC-anti CD3 and lineage (Lin) markers including pacific blue-anti-CD4, CD14, CD16 (BD) and CD19 (Biolegend). The cells were washed, fixed and analyzed using a LSRII Fortessa flow cytometer (BD). Live, Lin-, CD3+, CD8+, tetramer+ cells were then enumerated. Fluorescence minus one (FMO) controls were performed for each sample.

Statistical Analysis

For the RPPA analysis, supercurve algorithms were used to generate a single value from the 5 serial dilutions.14 Loading control and topographic normalization procedures accounted for protein concentration and background staining variations.15 Patients were divided into thirds based on CG expression levels. Normalization procedures were performed using the R software program. Comparison of protein levels between paired samples was done using the paired t-test. Associations between CG expression and categorical clinical variables were assessed in R using standard t-tests, linear regression, or mixed-effects linear model. The Kaplan-Meier method was used to generate survival curves. A Cox proportional hazards regression model was used to investigate the association between CG level and OS as categorized variables. Outcome analyses were carried out using the Statistica software V12 (StatSoft, Tulsa, OK).

For the in vitro and in vivo studies, results were expressed as means +/- standard deviation. ANOVA and t-tests were used to analyze the significance between treatment groups and controls. P values of less than 0.05 were considered to be statistically significant.

Supplementary Figure Legends

Supplementary Figure 1. CG1 is an effective target in a U937-A2 xenograft model and CG1 CTLs demonstrate antigen specificity. (a)Irradiated NSG mice were engrafted with a CG-expressing human HLA-A2+/GFP+ U937 AML cell line (U937-A2) (0.5 x 106 cells) on day 0. On day 1, mice were treated with bulk-expanded negative control HIV-CTL (0.5 x 106), CG1-CTL (0.5 x 106) or were left untreated. HIV-CTL and CG1-CTL were expanded from the same HLA-A2+ normal donor for each set of mice. On day 14, mice were sacrificed and bone marrow was harvested and analyzed for leukemia by flow cytometry. Results are expressed as the percentage CD13+/GFP+ cells from the viable human (h) CD45+/ mouse (m) CD45- population. Results reflect 4 independent experiments; *P<0.04.(b)T2 cells were cultured overnight in media containing 40 μg/mL of CG1 or PR1 (negative control), labeled with calcein-AM, and then co-cultured with CG1-CTLs for 4 h. Cytotoxicity was determined by measuring released calcein-AM. Data are means +/-SEM from duplicate wells from a representative experiment.(c)Representative flow cytometry plot demonstrating CG1-HLA-A2 tetramer staining of expanded normal donor CG1-CTL that were used in the experiments. Expanded CG1-CTL (live/CD3+/CD8+/CG1-HLA-A2 tetramer+/dump-) demonstrate a broad affinity for CG1-HLA-A2, as shown by the MFI distribution of the CG1-HLA-A2 tetramer stain. Doublet cells based on forward scatter area (FSC-A) and FSC height (FSC-H) were excluded from the gating.

Supplementary Figure 2. Gating strategy used to identify AML in bone marrow from leukemia-bearing NSG mice.After obtaining a single cell suspension, bone marrow was stained with live aqua dead stain, human (h) antibodies against CD45 (hCD45), CD3, CD13 (for U937), and CD33 (for UPN#1), as well as mouse (m)CD45. U937-A2 cell line was transduced with GFP. (a)Leukemia from UPN#1 was identified as live, hCD45+/mCD45-/CD3-/CD33+. Doublet cells based on forward scatter area (FSC-A) and FSC height (FSC-H) were excluded from the gating. (b)U937-A2 was identified as hCD45+/mCD45-/CD13+/GFP+

Supplementary Figure 3. CG1 peptide is presented on primary HLA A2+ AML blasts and the HLA-A2+ cell line HL60-A2.Lysate from primary patient leukemia blasts (n = 7) and cell lines (n=4) (Table 1) was subjected to ultracentrifugation and affinity purification using the anti-pan HLA antibody W6/32. Bound protein was eluted with 0.2 M aqueous acetic acid and standard acid elution methodology was used to separate the HLA molecule from B2-microglobulin and bound peptides. The solution was then filtered and the low molecular weight peptide-containing fraction was analyzed using liquid chromatography (LC) tandem mass spectrometry (MS/MS). Peptide identification was performed using Mascot software and the non-identical protein database maintained by NCBI. (a) The MS for CG1 is shown for AML patient UPN#1 that uniquely identifies the peptide as being singly charged with a MH+ of 918.49 daltons (Da) and eluting at a retention time of 33.28 minutes. (b) CG1 peptide was eluted from HL60-A2 cell line, also singly charged with a MH+ of 918.49 Da and eluting at a retention time of 33.21 minutes.(c)Synthetic CG1 (FLLPTGAEA) peptide was analyzed using liquid chromatography (LC) tandem mass spectrometry (MS/MS). The mass spectrum for CG1 is shown and identifies the peptide as being singly charged with a MH+ of 918.49 daltons (Da) and eluting at a retention time of 33.29 minutes.

Supplementary Figure 4. Cathepsin G is broadly expressed at the transcript level across AML cases. Relative cathepsin G transcript expression is shown across multiple malignancies and normal tissues, including in AML (LAML). The RNAseq data were downloaded from the public databases of The Cancer Genome Atlas (TCGA) and The Genotype-Tissue Expression (GTEx) project and are reported as transcripts per kilobase million (TPM).

Supplementary Figure 5. High cathepsin G level is associated with shorter overall survival (OS) in AML. (a) Kaplan-Meier plots showing OS in AML patients (n=415) comparing patients above median CG protein expression by RPPA to patients below median CG expression. Results are significant by Cox univariate model testing (P= 0.04).(b) Kaplan-Meier plots showing OS in AML patients with intermediate cytogenetics and mutated FLT3 (n = 76) comparing patients with high CG protein expression (upper 2/3) and low CG expression (lowest 1/3). Results are significant by Cox univariate model testing (P = 0.019).

Supplementary Figure 6. CG expression correlates with the expression of known leukemia associated proteins.Pearson correlation coefficient was performed and demonstrates a correlation between CG and a number of proteins that are known to play a role in leukemogenesis and leukemia progression. The selected proteins were identified based on a Pearson correlation coefficient threshold of 0.2.

Supplementary References

1.Rodriguez-Cruz TG, Liu S, Khalili JS, Whittington M, Zhang M, Overwijk W, et al. Natural splice variant of MHC class I cytoplasmic tail enhances dendritic cell-induced CD8+ T-cell responses and boosts anti-tumor immunity. PloS one 2011; 6(8): e22939.

2.Carter BZ, Qiu Y, Huang X, Diao L, Zhang N, Coombes KR, et al. Survivin is highly expressed in CD34(+)38(-) leukemic stem/progenitor cells and predicts poor clinical outcomes in AML. Blood 2012; 120(1): 173-180.

3.Carter BZ, Qiu YH, Zhang N, Coombes KR, Mak DH, Thomas DA, et al. Expression of ARC (apoptosis repressor with caspase recruitment domain), an antiapoptotic protein, is strongly prognostic in AML. Blood 2011; 117(3): 780-787.

4.Porcelli S, Morita CT, Brenner MB. CD1b restricts the response of human CD4-8- T lymphocytes to a microbial antigen. Nature 1992; 360(6404): 593-597.

5.Alatrash G, Mittendorf EA, Sergeeva A, Sukhumalchandra P, Qiao N, Zhang M, et al. Broad cross-presentation of the hematopoietically derived PR1 antigen on solid tumors leads to susceptibility to PR1-targeted immunotherapy. Journal of immunology 2012; 189(11): 5476-5484.

6.Zhang M, Sukhumalchandra P, Enyenihi AA, St John LS, Hunsucker SA, Mittendorf EA, et al. A Novel HLA-A*0201 Restricted Peptide Derived from Cathepsin G Is an Effective Immunotherapeutic Target in Acute Myeloid Leukemia. Clinical cancer research : an official journal of the American Association for Cancer Research 2013; 19(1): 247-257. Epub 2012/11/14. eng.

7.Shultz LD, Schweitzer PA, Christianson SW, Gott B, Schweitzer IB, Tennent B, et al. Multiple defects in innate and adaptive immunologic function in NOD/LtSz-scid mice. Journal of immunology 1995; 154(1): 180-191.

8.Sergeeva A, He H, Ruisaard K, St John L, Alatrash G, Clise-Dwyer K, et al. Activity of 8F4, a T-cell receptor-like anti-PR1/HLA-A2 antibody, against primary human AML in vivo. Leukemia 2016.

9.Provasi E, Genovese P, Lombardo A, Magnani Z, Liu PQ, Reik A, et al. Editing T cell specificity towards leukemia by zinc finger nucleases and lentiviral gene transfer. Nature medicine 2012; 18(5): 807-815.

10.Walz S, Stickel JS, Kowalewski DJ, Schuster H, Weisel K, Backert L, et al. The antigenic landscape of multiple myeloma: mass spectrometry (re)defines targets for T-cell-based immunotherapy. Blood 2015; 126(10): 1203-1213.

11.Sergeeva A, Alatrash G, He H, Ruisaard K, Lu S, Wygant J, et al. An anti-PR1/HLA-A2 T-cell receptor-like antibody mediates complement-dependent cytotoxicity against acute myeloid leukemia progenitor cells. Blood 2011; 117(16): 4262-4272.

12.Kornblau SM, Tibes R, Qiu YH, Chen W, Kantarjian HM, Andreeff M, et al. Functional proteomic profiling of AML predicts response and survival. Blood 2009; 113(1): 154-164.

13.Tibes R, Qiu Y, Lu Y, Hennessy B, Andreeff M, Mills GB, et al. Reverse phase protein array: validation of a novel proteomic technology and utility for analysis of primary leukemia specimens and hematopoietic stem cells. Molecular cancer therapeutics 2006; 5(10): 2512-2521.

14.Hu J, He X, Baggerly KA, Coombes KR, Hennessy BT, Mills GB. Non-parametric quantification of protein lysate arrays. Bioinformatics 2007; 23(15): 1986-1994.

15.Neeley ES, Kornblau SM, Coombes KR, Baggerly KA. Variable slope normalization of reverse phase protein arrays. Bioinformatics 2009; 25(11): 1384-1389.