2.2. Cell lines and chemical compounds

The human HNSCC cell lines CAL33 (RRID: CVCL_1108), HSC3 (RRID: CVCL_YA14), HSC4 (RRID: CVCL_1289), FADU (RRID: CVCL_1218), and CAL27 (RRID: CVCL_1107) were purchased from ATCC. The 4NQO‐L cell line was developed in our laboratory by exposing mice to drinking water containing 50μg·mL⁻¹ 4‐nitroquinoline 1‐oxide (4NQO; N8141, Sigma, St. Louis, MO, USA) [25]. All cell lines were maintained at 37°C in a humidified atmosphere and 5% CO₂ in DMEM. Cells were routinely tested for mycoplasma infection, and treated with appropriate antibiotics as needed (De‐Plasma™, TOKU‐E, D022). Cells were tested for cell line authentication by the Technion Genomics Center, Technion Israel Institute of Technology, in March 2023. The test was performed using the Promega GenePrint 24 System to determine short tandem repeat (STR) profile of 23 loci plus Amelogenin for gender determination (X or XY). In addition, the male‐specific DYS391 locus is included to identify null Y allele results for Amelogenin. DNA sample from the kit (2800M Control DNA) was included in the analysis and served as positive control for the PCR step. No DNA template was also included as negative control. The results were analyzed using the 3500xl Genetic Analyzer (v.1 data collection software, Life Technologies, Carlsbad, CA, USA) and genemapper idx software (v1.3, Life Technologies). Allelic ladder was included in the run.

2.3. Antibodies and reagents

The following antibodies were purchased from Cell Signaling Technology: anti Phospho‐p44/42 MAPK (pErk 1/2) (#4370S), anti p44/42 MAPK (Erk1/2) (#4695) anti pAKT Ser473 (#4060), anti AKT (#4691), anti pS6 S240/244 (#5364), anti‐EGFR (#4267S), anti pEGFR (#2236L), and anti YAP1 (#14074). Anti Ki67 (#275R‐15) was purchased from Cell Marque, and Anti‐actin (691001), from MP Biomedicals, Santa Ana, CA, USA. Erlotinib (HY‐50896) was purchased from MedChemExpress, Monmouth Junction, NJ, USA. Novartis Pharma AG (Basil, Switzerland) provided trametinib (GSK1120212, JTP‐74057), BYL719, and GDC‐0941. All compounds were dissolved in dimethyl sulfoxide (DMSO) for in vitro studies and 5% carboxymethylcellulose (CMC) or corn oil with 4% DMSO for in vivo experiments.

2.4. IC₅₀ and synergy assay

Cells were seeded in 96‐well plates (3000–5000 cells per well), treated with increasing concentrations of the drug being tested, and allowed to proliferate for 4days. At the end point, cells were stained with crystal violet (1g·L⁻¹) for 10min at room temperature. The crystal violet was then dissolved with 10% acetic acid, and absorbance was measured at 570nm in a spectrophotometer (Epoch, Biotech, Agilent, Santa Clara, CA, USA). IC₅₀ values were calculated using graphpad software (v.7, GraphPad, San Diego, CA, USA). For the synergy assays, the proliferation of the cells in the different treatment groups was determined as the percentage of control (DMSO‐treated) cells; the colorimetric results were analyzed using synergyfinder Software (https://synergyfinder.fimm.fi/).

2.5. Phospho‐RTK array

CAL33 and HSC3 cells were seeded in 100‐mm well plates and allowed to adhere overnight. Cells were then treated with 20nm trametinib or DMSO. Total cell lysates of 300μg were analyzed using the Proteome Profiler™ Human Phospho‐RTK antibody arrays kit, catalog # ARY001B (R&D Systems, Burlington, ON, Canada), according to the manufacturer's protocol.

2.6. Western blotting

As described before [26], Cells were washed with cold phosphate‐buffered saline (PBS) and scraped into lysis buffer supplemented with phosphatase inhibitor cocktail (Stratech, B15001‐BIT, Birkenfeld, Germany). Lysates were cleared by centrifugation at 20 000 g for 10min at 4°C, and supernatants were collected and assayed for protein quantification using Bradford protein assay (Biorad, 5000006). Total lysate, 20μg, was resolved on NuPAGE 4–12% Bis‐Tris gels and transferred electrophoretically to PVDF membranes (Biorad, 1704159). Membranes were blocked for 1h in 5% bovine serum albumin (BSA) in Tris‐buffered saline (TBS)‐Tween and then hybridized using the primary antibodies in 5% BSA TBS‐Tween. Mouse and rabbit horseradish peroxidase‐conjugated secondary antibodies (1:20000, GE Healthcare, Chicago, IL, USA) were diluted in 5% BSA in TBS‐Tween. Protein–antibody complexes were detected by chemiluminescence with ECL (Westar Supernova, Cyanogen XLS3.0100, and Westar Nova 2.0 Cyanagen XLS071.0250, Bologna, Italy), and images were captured with a c300 Azure camera system (Dublin, CA, USA). Quantification was done using “image j.” Protein level was calculated relative to loading control (actin) and presented as fold change from the control sample.

2.7. Real‐time quantitative PCR

After 24h of treatment, total RNA was isolated from cells by RNA Mini Kit (Bioline – BIO‐52073, London, UK) according to the manufacturer's protocol. RNA, 1μg, was converted to cDNA using a qScript cDNA synthesis kit (Quanta Bioscience, 95047‐100) according to the manufacturer's protocol and as previously described [27]. Real‐time PCR was performed (Roche LightCycler® 480 II, Basel, Switzerland) using a prime‐time gene expression master mix (IDT, 1055770) with matching probes: human actin (Hs.PT.39a.22214847 from IDT) or human EGFR (Hs01076078_m1 from ThermoFisher Scientific).

2.8. Immunofluorescence

As previously described [25], cells were seeded on glass coverslips in 24‐well plate for 24h, then rinsed with cold PBS (4°C) and fixed in 4% PFA for 30min at room temperature. After that, cells were rinsed with PBS, followed by permeabilization on ice for 5min in PBS with 0.05% Triton X‐100 (Millipore Sigma, Burlington, MA, USA), re‐rinsed with PBS, and blocked in blocking solution for 30min at room temperature (5% BSA in Tween‐PBS). The cells were incubated with primary antibodies overnight at 4°C, rinsed with PBS, and incubated with Cy3‐conjugated anti‐rabbit secondary antibody (1:250; 111‐165‐144 Jackson ImmunoResearch, West Grove, PA, USA) at room temperature for 1h. Cells were then rinsed with PBS and mounted in DAPI Fluoromount‐G (0100‐20, Southern Biotech, Birmingham, AL, USA).

2.9. Staining of tumor tissue: immunohistochemistry and immunofluorescence

Immunohistochemistry – As described previously [28], following surgery, tumor tissue samples were obtained from patients with HNSCC at the Ear Nose and Throat Unit of Soroka Medical Center, Be'er Sheva, Israel. The collection of samples was conducted with the patients' understanding and written consent, and the study received approval from the local ethics committee (ethics code: 0421‐16‐SOR and 0103‐17‐SOR) in accordance with the Helsinki Declaration. Tumors were fixed in 4% paraformaldehyde solution overnight at room temperature and then maintained in 70% ethanol until embedding in paraffin. Paraffin‐embedded tumor blocks were sectioned into 5‐μm slices, loaded onto microscope slides, and deparaffinized at 60°C for 1h. After additional deparaffinization with a xylene substitute (3803672E, Leica, St. Gallen, Switzerland) and rehydration steps in a descending alcohol series, antigen retrieval was performed. The slides were incubated in 10mm citric acid buffer, pH 6.0 at 100°C for 15min, then cooled in buffer at room temperature and rinsed with doubly distilled water for 3min, 3′. Thereafter, endogenous peroxidases were inactivated with 0.3% hydrogen peroxide in methanol buffer for 30min. The slides were washed three times with PBS for 3min, and then blocked with 5% BSA in PBS‐T (0.1% TWEEN) for 1h at room temperature. Slides were then incubated overnight at 4°C with the primary antibodies, Ki67 (Sigma, AB9260, 1:200), pMAPK (CST, Danvers, MA, USA, 4370S, 1:200), pS6 (CST, 4857S, 1:800), and EGFR (CST, 4267S, 1:50) diluted in blocking solution. The following day, the slides were washed three times with PBS‐T. The ABC kit (VECTASTAIN®ABC, VE‐PK‐6200) was used for detection according to manufacturer's protocol, with 3,3′‐diaminobenzidine (DAB) (ACH500‐IFU; ScyTek Laboratories, Logan, UT, USA) as a substrate for color development. Slides were counterstained with hematoxylin, dehydrated, and mounted in mounting medium (3801740, Sub‐X, Leica, 3801740). All slides were digitalized using the Panoramic Scanner (3DHISTECH, Budapest, Hungary), and the analysis was performed with qupath‐0.2.3 software [29]. Fields of tumor tissue for analysis were chosen by a blinded investigator. Cells within the analysis field were detected by the qupath‐0.2.3 software and were defined as positive or negative to DAB staining according to a threshold set by two independent blinded investigators (Fig. S1H). Cell detection criteria and thresholds were strictly maintained between compared slides. The specificity of the staining and analysis threshold was verified by comparison with a matched negative control tissue, which was incubated with no primary antibody but subjected to all secondary antibody development processes.

Immunofluorescence – Tissue was processed as described above and then incubated with the primary antibody YAP (CST, 14074, 1:500) overnight at 4°C. The next day, cells were rinsed with PBS‐T and incubated with Alexa Fluor‐647 Anti‐Rabbit secondary antibody (Jackson ImmunoResearch, 111‐605‐144, 1:250) at room temperature for 1h. Finally, cells were washed with PBS‐T and mounted with DAPI Fluoromount‐G® (SouthernBiotech, Birmingham, CA, USA, 0100‐20).

2.10. RNA silencing

For the production of shRNA cell lines, we created lentiviruses by transfecting HEK293 cells with the viral plasmids psPAX2, pMD2.G, and PLKO with shRNAs scrambled sequence (designated shControl) or a sequence for the silencing of EGFR expression (designated shEGFR). Transfection was done by using PolyJet transfection reagent (SL100688, SignaGen, Frederick, MD, USA) according to the manufacturer's protocol. Viruses were collected after 48–72h and used for cell infection. Cells were seeded in 6‐well plates (150000 cells per well), infected with the lentiviruses in the presence of Polybrene (MilliporeSigma, 5G‐H9268) and selected with puromycin (Thermo Fisher Scientific, A11138‐03).

2.11. Cell proliferation assay

Cells were seeded in a 24‐well plate (10000–20000 cells per well) and treated the following day. At the endpoint, cells were stained with crystal violet (1g·L⁻¹). Quantification was performed by dissolving out the crystal violet (10% acetic acid) and reading the optical density at 570nm in a spectrophotometer (Epoch, Biotech).

2.12. In vivo experiments

Mice were maintained and treated in accordance with the Institutional Animal Care and Use Committee guidelines of Ben‐Gurion University of the Negev. Animal experiments were approved by the Institutional Animal Care and Use Committee (IL‐80‐12‐2015 and IL‐29‐05‐2018). Mice were housed in air‐filtered laminar flow cabinets with a 12‐h light/dark cycle and food and water ad libitum. At the end of the experiments, animals were euthanized with CO₂. To generate CAL33 tumor‐bearing mice, CAL33‐shControl and CAL33‐shEGFR‐3 cells (3×10⁶ and 5×10⁶ cells, respectively) were suspended in 100μL of PBS and injected subcutaneously into the right and left flanks of 6‐week‐old female NSG mice (NOD.Cg‐Prkdcscid Il2rgtm1Wjl/SzJ, Jackson Labs, Bar Harbor, ME, USA). Tumor‐bearing mice were then randomized into two groups of five to six mice, based on tumor volume (between 150 and 200mm³). To generate the syngeneic murine model, 4NQL cells (5×10⁶ cells) suspended in 50μL of PBS were injected into the lips of 6‐week‐old female C57BL/6J (WT) mice (Envigo, Huntingdon, UK, C57/BL/6). Tumor‐bearing mice were then randomized into four groups of five to six mice, based on tumor volume (between 150 and 200mm³). To generate HRAS mutant patient‐derived xenografts (PDXs), 3mm³ pieces of tumors were implanted into the right and left flanks of 6‐week‐old female NSG mice (NOD.Cg‐Prkdcscid Il2rgtm1Wjl/SzJ, Jackson Labs). Tumor‐bearing mice were then randomized into two groups of five to six mice, based on tumor volume (between 90 and 120mm³). In all in vivo experiments, trametinib (0.5mg·kg⁻¹) was dissolved in corn oil and administered daily by intraperitoneal (i.p.) injection, and BYL719 (25mg·kg⁻¹) was dissolved in 5% CMC and administered daily by oral gavage. Vehicle‐treated mice received both corn oil and 5% CMC. Tumors were measured twice a week with a digital caliper, and tumor volume was calculated according to the formula: V=(L W′ W′)π/2, where V is tumor volume, W is tumor width, and L is tumor length. Tumor volumes are plotted as means±SEM. In the toxicity study, C57BL/6 female mice were randomized into trametinib/BYL719‐treated and control groups (n=5). Trametinib (0.5mg·kg⁻¹) was dissolved in corn oil and administered by intraperitoneal (i.p.) injection, and BYL719 (25mg·kg⁻¹) was dissolved in 5% CMC and administered daily by oral gavage for seven consecutive days. Vehicle‐treated mice received corn oil and 5% CMC. 24h after the last treatment, blood was collected from the venous plexus of the eye. Serum was obtained by centrifugation (10min at 2000g) for analyte measurements (Biochemical Laboratory at Soroka Medical Center). Hepatic damage was assessed by serum levels of alanine aminotransferase (ALT) and alkaline phosphatase (ALK‐phos). Renal damage was indicated by the levels of blood urea and creatinine. Livers, lungs, kidneys, intestines, or spleens were analyzed for weight and morphology changes.

3.2. EGFR limits trametinib efficacy by hyperactivating the PI3K/AKT pathway

To investigate the potential role of EGFR in limiting trametinib efficacy, we reduced the expression of EGFR by shRNA in the CAL33 and HSC3 cell lines (Fig. S2A) and then evaluated the sensitivity of tumor cells to trametinib in vitro and in vivo. Proliferation assay showed that EGFR knockdown sensitized both cell lines to trametinib (Fig. 2A). To test the effect of EGFR expression on the efficacy of trametinib in mice, we injected shEGFR and shControl cells subcutaneously into NSG mice, and when tumors reached ~200mm³ in size, treatment with trametinib was initiated. Tumor growth kinetics showed that following exposure to trametinib shControl tumors exhibited a growth delay, but all tumors eventually progressed. In contrast, trametinib treatment of shEGFR tumor‐bearing mice resulted in stable disease, and tumor progression was not detected even after 22days of treatment (Fig. 2B).

Open in a separate window

Fig. 2

Epidermal growth factor receptor (EGFR) limits trametinib efficacy by hyperactivating the phosphatidylinositol‐4,5‐bisphosphate 3‐kinase/protein kinase B (PI3K/AKT) pathway. (A) 4‐day proliferation assay testing trametinib (50nm) efficacy in shControl and shEGFR CAL33 and HSC3 cell lines. Error bars indicate SEM. Data represent a representative experiment from three independent experiments. (B) (top) Tumor volume of the shControl or shEGFR CAL33 and HSC3 cell‐derived xenografts injected to NSG mice subsequently treated daily with trametinib (0.5mg·kg⁻¹i.p.) or vehicle (n=5). Error bars indicate SEM. (bottom) change in tumor volume from first to last day of treatment. (C) Western blot for the indicated proteins following 24h of 20nm trametinib treatment in shControl and shEGFR‐3 CAL33 and HSC3 cell lines. Numbers indicate the fold change in protein level normalized to Actin. Data represent a representative experiment from three independent experiments. (D) Western blot for the indicated proteins following 24h of 20nm trametinib treatment with or without 5μm EGFR inhibitor erlotinib, in CAL33 and HSC3 head and neck squamous cell carcinoma cell lines. Numbers indicate the fold change in protein level normalized to Actin. Data represent a representative experiment from three independent experiments. Statistical significance was calculated using the unpaired t‐test (**P<0.01, ***P<0.001, ****P<0.0001).

To obtain further insights into the downstream signaling of EGFR that may be responsible for limiting the efficacy of trametinib, we examined the phosphorylation status of AKT by western blot analysis of both shControl and shEGFR cell lines treated with trametinib or DMSO. In shControl cells, trametinib treatment induced EGFR overexpression and activation along with hyperactivation of the PI3K/AKT pathway, as indicated by an increase in pAKT. In contrast, in shEGFR cells, trametinib treatment did not induce pAKT upregulation (Fig. 2C). We confirmed the role of EGFR in activating the PI3K/AKT pathway by pharmaceutically inhibiting EGFR signaling with erlotinib. In that experiment, we evaluated the expression of EGFR, PI3K/AKT, and MAPK signaling in CAL33 and HSC3 cell lines treated with trametinib, erlotinib, or a combination of erlotinib and trametinib. EGFR blockage prevented trametinib‐induced PI3K/AKT hyperactivation in both CAL33 and HSC3 cell lines (Fig. 2D). Overall, these invivo and invitro results indicate that EGFR plays a role in limiting the efficacy of trametinib, putatively via activation of the PI3K/AKT pathway.

3.3. Preventing PI3K/AKT hyperactivation enhances trametinib efficacy

To explore whether hyperactivation of the PI3K/AKT pathway following trametinib treatment is a robust phenomenon, we determined pAKT in response to trametinib in three additional HNSCC cell lines (HSC4, FADU, and CAL27). Western blot analysis showed that four of the five cell lines exhibited an increase in pAKT within 24h of trametinib treatment (Fig. S2B). Therefore, we hypothesized that blockage of PI3K/AKT pathway with a PI3K inhibitor would enhance trametinib efficacy. To validate this premise, we investigated the biological and molecular response of CAL33 and HSC3 cells to treatment with trametinib with and without BYL719 (alpelisib), a p110a isoform‐specific inhibitor of PI3K, or GDC‐0941 (pictilisib), a pan PI3K inhibitor. A four‐day proliferation assay showed that combining trametinib with BYL719 or GDC‐0941 limited the proliferation of cells vs. single‐drug treatment (Fig. 3A, Fig. S3A). Protein expression analysis showed that a combined treatment with trametinib and either BYL719 or GDC‐0941 prevented the trametinib‐induced activation of the PI3K/AKT pathway in both tested cell lines (Fig. 3B).

Open in a separate window

Fig. 3

Dual treatment with trametinib and a phosphatidylinositol‐4,5‐bisphosphate 3‐kinase (PI3K) inhibitor has a synergistic effect in vitro. (A) Four‐day proliferation assay testing the efficacy of trametinib (20nm) and of BYL719 (2μm; the isoform‐specific inhibitor of PI3K), or GDC‐0941 (1μm; the inhibitor of α and δ PI3K), alone or in combination, in CAL33 and HSC3 head and neck squamous cell carcinoma (HNSCC) cell lines. Error bars indicate SD. Data represent a representative experiment from three independent experiments. (B) Western blot for the indicated proteins following 24h of 20nm trametinib treatment with or without BYL719 (2μm) or GDC‐0941 (1μm) in CAL33 and HSC3 HNSCC cell lines. Continued use of tumor cell lysates from Fig. 1B. Numbers indicate the fold change in protein level normalized to Actin. Data represent a representative experiment from three independent experiments. (C) Synergy scores and heatmap calculated by SynergyFinder for the combination of trametinib with BYL719 or GDC‐0941 in CAL33 and HSC3 HNSCC cell lines. Data represent a representative experiment from three independent experiments. Statistical significance was calculated using one‐way ANOVA (*P<0.05, **P<0.01, ***P<0.001, ****P<0.0001).

To explore whether the antiproliferative effect of trametinib in combination with BYL719 or GDC‐0941 is additive or synergistic, a synergy test was performed in HSC3 and CAL33 HNSCC cell lines. In this test, an average calculated synergy score above 1 indicates a synergistic antitumor effect between trametinib and the PI3K inhibitor. The synergy score of GDC‐0941/trametinib was higher than that for BYL719/trametinib in both models (Fig. 3C). To corroborate the findings, we performed similar experiments in other HNSCC cell line models, namely, HSC4, FADU, and CAL27. We found that BYL719/trametinib has a synergistic antitumor effect in CAL27, while GDC‐0941/trametinib has synergistic effects in HSC4 and CAL27 (Fig. S3B). Overall, we found that in most HNSCC models tested, trametinib treatment induced hyperactivation of the PI3K/AKT pathway and that blocking PI3K signaling has a synergistic antitumor effect when combined with trametinib.

3.4. Dual treatment with trametinib and BYL719 prolongs PDX growth arrest in mice

Since BYL719 was recently approved by the FDA [32] for the treatment of certain breast cancers and is in clinical development for HNSCC (NCT04997902), after promising results in a phase Ia clinical trial [33], we chose this agent for further investigation. To test the antitumor activity of trametinib, BYL719, and their combination in vivo, we used an HRAS‐mutated PDX model (designated here PDX3). Specifically, we implanted PDX3 into NSG mice, and after successful engraftment mice were randomized into four arms. The mice were treated daily with vehicle, trametinib (0.5mg·kg⁻¹), BYL719 (25mg·kg⁻¹), or a combination of both agents. Tumor growth kinetics showed that PDX3 was resistant to BYL719 treatment but responsive to trametinib monotherapy, as indicated by the delay in tumor progression. However, marked tumor growth arrest was observed in mice treated with the BYL719/trametinib combination, with no detectable increase in tumor volume, even after 40days of treatment (Fig. 4A, Fig. S4A). Importantly, the combination treatment with the two agents did not result in any observed toxicity, as the animals' weight remained stable during the efficacy experiment. Further biochemical tests of serum markers after 7days of treatment showed no major liver or kidney toxicity besides a reduction in the creatinine levels in mice treated with BYL719 and trametinib (Fig. S4C–E).

Open in a separate window

Fig. 4

Dual treatment with trametinib and BYL719 prolongs patient‐derived xenograft (PDX) growth arrest in mice. (A) Tumor volume of the HRAS‐mutated PDX3 model transplanted in NSG mice. Mice were randomized into four arms (n=5) and treated daily with trametinib (0.5mg·kg⁻¹i.p.), BYL719 (25mg·kg⁻¹ by oral gavage), a combination BYL719/trametinib treatment or vehicle. Error bars indicate SEM. (B) Representative immunohistochemical staining (X20, 100μm) and the quantification of cells positive for: phosphorylated mitogen‐activated protein kinase (pERK) staining (left) indicating the response to trametinib treatment (n=4 tumors and n=6 analysis fields), pS6 staining (middle) indicating activation of the phosphatidylinositol‐4,5‐bisphosphate 3‐kinase/protein kinase B/mammalian target of rapamycin (PI3K/AKT/mTOR) pathway (n=4 tumors and n=6 analysis fields), and Ki67 staining (right) indicating cell proliferation (n=4 tumors and n=8 analysis fields), after 4days of treatment. Error bars indicate SD. Statistical significance was calculated using one‐way ANOVA (*P<0.05, **P<0.01, ***P<0.001, ****P<0.0001).

To quantify the on‐target effect of the treatments on the activity of signaling pathways in tumor cells and on tumor cell proliferation, we used immunohistochemical staining to examine tumors after 4days of treatment. For evaluating activation of the MAPK and AKT pathways, we stained the tumors for pERK and pS6 as the downstream activator of AKT and mammalian target of the rapamycin (mTOR) pathway, respectively. As expected, tumors treated with trametinib monotherapy showed a reduction of pERK staining but simultaneously an increase in pS6 staining (an indicator for AKT/mTOR pathway activation) and EGFR expression (Fig. 4B, and Fig. S4B). Staining with the proliferation marker of Ki67 showed decreased tumor cell proliferation in tumors treated with either trametinib or BYL719 monotherapy. However, more significant inhibition of cell proliferation was detected in tumors treated with the BYL719/trametinib combination (Fig. 4B). These results indicate that the dual treatment inhibited the PI3K/AKT/mTOR and MAPK pathways, both which are needed for tumor cell proliferation.

We thank Prof Jennifer Grandis (University of California at San Francisco) for providing PDX‐3. We also thank Inez Mureinik for language editing. Graphical abstract was created with BioRender.com. This work was supported in part by the Cooperation Program in Cancer Research of the Deutsches Krebsforschungszentrum (DKFZ) and Israel's Ministry of Science, Technology and Space (MOST) (DKFZ‐MOST #001192) (to ME, JH, IK). Additional funding was received by the Israel Science Foundation (ISF, 302/21) (to ME), Natural Science Foundation of China and Israel Science Foundation (NSFC‐ISF #3409/20) (to ME, DK), and the United States – Israel Binational Science Foundation (BSF, #2021055 to) (to ME), NIH R01 (DE027738) (to LGTM), US Department of Defense (CA210784) (to LGTM) and the NIH/NCI Cancer Center Support Grant P30 (CA008748) (to LGTM).