Radiation-induced activation of the phosphatidyl inositol-3 kinase/Akt signal transduction pathway requires Akt binding to phosphatidyl-inositol phosphates (PIP) on the cell membrane. The tyrosine kinase bone marrow X kinase (Bmx) binds to membrane-associated PIPs in a manner similar to Akt. Because Bmx is involved in cell growth and survival pathways, it could contribute to the radiation response within the vascular endothelium. We therefore studied Bmx signaling within the vascular endothelium. Bmx was activated rapidly in response to clinically relevant doses of ionizing radiation. Bmx inhibition enhanced the efficacy of radiotherapy in endothelial cells as well as tumor vascular endothelium in lung cancer tumors in mice. Retroviral shRNA knockdown of Bmx protein enhanced human umbilical vascular endothelial cell (HUVEC) radiosensitization. Furthermore, pretreatment of HUVEC with a pharmacologic inhibitor of Bmx, LFM-A13, produced significant radiosensitization of endothelial cells as measured by clonogenic survival analysis and apoptosis as well as functional assays including cell migration and tubule formation. In vivo, LFM-A13, when combined with radiation, resulted in significant tumor microvascular destruction as well as enhanced tumor growth delay. Bmx therefore represents a molecular target for the development of novel radiosensitizing agents. [Cancer Res 2008;68(8):2861–9]
The microvasculature is a major component of cancer and supports tumor growth. We and others have studied the inherent resistance of tumor vascular endothelium to cytotoxic effects of ionizing radiation. Ionizing radiation activates signal transduction through the phosphatidyl inositol-3 kinase (PI3K)/Akt pathway, which enhances endothelial cell viability (1–3). Akt activity is critical for this process because dominant-negative Akt mutant (Thr308Ala and Ser473Ala) overexpression in endothelial cells abrogates radiation-induced cell survival response (4). Moreover, we have previously shown that ionizing radiation–induced Akt activation is eliminated by overexpression of mutant p85 component of PI3K (5). Mutant p85 functions as a dominant negative by preventing activation of p110 catalytic subunit of PI3K. This inhibition prevents the production of phosphatidylinositol phosphates (PIP) that activate Akt, resulting in enhanced radiation effect. Therefore, inhibition of the PI3K signal transduction pathway can abrogate the endothelial cell survival signaling mediated by Akt.
Although the PI3K/Akt pathway is a major contributor to radiation resistance seen in tumor microvasculature, other pathways activated shortly after ionizing radiation are also being investigated. Indeed, the activation of Akt has been shown to be critically dependent on binding of the pleckstrin homology domain of Akt to specific PIPs, phosphatidylinositol(3,4,5)-triphosphate, in particular, which allows the colocalization of Akt with upstream activators (6). Bone marrow X kinase (Bmx), also known as epithelial and endothelial tyrosine kinase, contains a pleckstrin homology domain as well as Src homology (SH)2 and SH3 domains capable of interacting with several types of second messengers and adaptor proteins that are present in human umbilical vein endothelial cells (HUVEC; refs. 7–13). Bmx is the ubiquitously expressed member of the Tec family of nonreceptor tyrosine kinases with high expression in lung, prostate, and the heart (10, 11). In addition, salivary epithelium, granulomonocytic cells, endothelial cells, and epithelial cells express this protein in relatively high amounts (10, 11, 14–16). Bmx seems to act both upstream and downstream of PI3K (7, 10, 11, 17–21). Bmx also interacts with G-proteins (10, 22–24), integrins/NRTKs (7, 25), tumor necrosis factor receptors (9), as well as various protein tyrosine phosphatases (26) and lipid phosphatases (27).
Because Bmx seems to potentiate proliferative and cell survival signaling in many cell types, we investigated Bmx signaling transduction during the radiation response in vascular endothelium. In the present study, radiation-induced Bmx activation in vascular endothelium was investigated. We show that inhibition of Bmx activity enhances the effectiveness of radiation on vasculature. These findings support the hypothesis that Bmx promotes a cell survival pathway and is a molecular target for drug development in the treatment of cancer.
Materials and Methods
Cell culture. HUVECs were purchased from Clonetics and maintained in EBM-2 medium supplemented with EBM-2 singlequots (Cambrex). HUVECs were limited to passages 3 to 6. Lewis Lung carcinoma (LLC) cells were purchased from American Type Tissue Culture and maintained in DMEM supplemented with 10% fetal bovine serum (FBS) and 1% penicillin-streptomycin. Cell lines were incubated at 37°C in a 5% CO2 incubator.
LFM-A13 (30 μmol/L in vitro or 50 mg/kg, i.p. in vivo) and DMSO was obtained from Sigma. Drug was administered to cells 30 to 60 min before irradiation. A Mark-1 Irradiator 137Cs (JL Shepard and Assoc.) was used to irradiate HUVEC cultures at a dose rate of 1.897 Gy/min.
Retrovirus production and HUVEC infection. Negative control and Bmx shRNA retroviral constructs were purchased from OriGene, Inc. A total of five different constructs (labeled Bmx A–E) were tested for Bmx knockdown in HUVEC. The retroviruses were produced according to manufacturer's protocol with some modifications. LiNX packaging cell line, purchased from Open Biosystems, was grown on 10-cm tissue culture plates to 30% to 40% confluency in medium containing DMEM with 10% FBS [complete growth medium (CGM)] with hygromycin, penicillin, and streptomycin supplementation. These cells were then transfected with retroviral plasmid DNA by incubation with 5 mL transfection mix for 4 to 6 h. The transfection mix contained 12 μg of shRNA retroviral vector DNA within 600 μL serum-free DMEM without antibiotics, and 240 μL of room temperature Arrest-In transfection reagent (Open Biosystems) within 4.4 mL serum-free DMEM without antibiotics, which was prepared and kept at room temperature for 45 min before transfection to allow for transfection complexes to be formed. After initial 4- to 6-h transfection, 5 mL of CGM were added and incubated overnight. The medium was changed and the cells were incubated for at least an additional 24 h. Supernatant was collected and filtered through a 45-μm filter to produce the viral stock.
For HUVEC infection, HUVECs were grown on tissue culture plates to 50% confluency. On the day of infection, the HUVECs were incubated in medium containing 5 μg/mL polybrene for 4 h before infection. Medium was then removed and 1.5 mL of virus supernatant supplemented with 5 μg/mL polybrene were added directly to the cells and allowed to adsorb for 40 to 60 min, and then 7 mL of HUVEC medium containing 5 μg/mL of polybrene were added, and the cells were incubated for 24 h. The medium was changed to regular HUVEC growth medium, and the cells were incubated for an additional 2 d to allow for Bmx knockdown.
Cell lysis and immunoblot analysis. HUVECs of passage 3 to 6 were treated with or without Bmx inhibition (LFM-A13 for 60 min or shRNA retrovirus infection 48 h before) followed by irradiation and then harvested at the indicated times. Cells were processed and immunoblotted as described previously (28). Antibodies were as follows: PY20HRP (BD Biosciences), actin (Sigma-Aldrich), Bmx and PY40 Bmx (Cell Signaling Technology), phospho-Akt (S473), Akt (Cell Signaling Technology), as well as horseradish peroxidase–labeled mouse anti-rabbit secondary antibodies (Sigma-Aldrich) except for PY20HRP, which was prelabeled. Films were scanned into Adobe Photoshop with subsequent densitometry analysis using NIH Image J 1.37v. Experiments were performed at least thrice.
Immunoprecipitation and in vitro kinase assay. HUVECs of passage 3 to 6 were grown to 70% to 80% confluency and then serum starved for 5 h. The cells were then treated with sham or 3 Gy irradiation. After treatment, the cells were incubated at 37°C with 5% CO2 for the indicated times. For inhibition studies, inhibitor was added at a 1:1,000 dilution 60 min before irradiation. After incubation, the tissue culture dishes were placed on ice and washed twice with ice cold 1× PBS followed by lysis using 400 μL M-PER containing protease and phosphatase inhibitors (Sigma) for 5 min. The cells were scraped and transferred to Eppendorf tubes, vortexed for 20 sec, and incubated on ice for 30 min. After clearing the lysate by centrifugation at 15,000 g for 15 min at 4°C, the supernatant was quantified for protein concentration using bicinchoninic acid method before immunoprecipitation. Immunoprecipitation was performed using the Catch and Release v 2.0 system (Upstate) according to manufacturer's protocol with some modification. Briefly, spin columns containing binding resin were prepared by washing twice with 1× wash buffer (2,000 g/30 sec). After column preparation, 500 μg of cell lysate were combined with 2 μg of anti-Bmx antibody (H-220; Santa Cruz Biotechnology, Inc.), 10 μL of affinity ligand, and enough 1× wash buffer to have a 500 μL final volume. This was added to the capped spin column that was rotated end-over-end overnight at 4°C. The spin column was washed thrice with 1× wash buffer followed by elution using 70 μL of 1× nondenaturing elution (native protein elution) buffer. For in vitro kinase assay, 35 μL of eluate were then combined with 25 μL of kinase buffer containing 25 mmol/L Tris (pH 7.5), 5 mmol/L β-Glycerophosphate, 2 mmol/L DTT, 0.1 mmol/L Na3VO4, and 10 mmol/L MgCl2 with 1 μL of 10 mmol/L ATP, and this was incubated for 20 min at 37°C. The reaction was stopped by adding 20 μL of 4× SDS sample buffer and boiled for 5 min. Samples were then run on SDS-PAGE and anti-PY20HRP (BD Biosciences) and anti-Bmx (Cell Signaling Technology). Western blotting was performed to identify phosphorylated protein bands at 75 to 80 kDa. Densitometric analysis was performed using NIH Image J.
WST-1 assay. This assay is a modification of an 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay and was performed per manufacturer's protocol (Calbiochem). Briefly, HUVEC were infected with shRNA retrovirus, and after 48 h of incubation, cells were lifted by trypsinization, were counted, and were plated at 10,000 cells per well of 96-well microtiter plates in duplicate. These plates were then subjected to 0 or 2 Gy and incubated for 24 h at 37°C with 5% CO2. After incubation, 10 μL of WST-1 labeling mixture was added to each well and mixed gently before returning to the incubator for 2 h. Absorbance at 450 nm was measured on a microplate reader and results were plotted using Microsoft Excel Software.
Clonogenic survival. LLC or passage 3 to 6 HUVECs were grown to 70% to 80% confluency. Cells were washed with 1× PBS, trypsin suspended, and were counted and adjusted to specific densities for each condition. The cells were then plated on tissue culture plates and allowed to attach for 4 h. LFM-A13 or DMSO was added at a 1:1,000; dilution followed 60 min later by 0, 2, 4, or 6 Gy. Medium was changed after irradiation. Ten to fourteen days after irradiation, the plates were fixed with 70% ethanol and stained with 1% methylene blue. Colonies were then counted using a dissection microscope with positive colonies containing at least 50 cells. Surviving fraction was calculated by the equation (number of colonies formed/number of cells plated)/(number of colonies for sham irradiated group/number of cells plated). Dose-enhancing ratios were calculated by dividing the dose (Gy) for radiation alone by the dose for radiation plus treatment (normalized for plating efficiency of treatment) for which a surviving fraction of 0.2 is achieved. These results were then plotted in a semilogarithmic format using Microsoft Excel software.
Endothelial cell tubule formation assay. HUVEC of passage 3 to 6 were grown to 70% to 80% confluency. Cells were then washed with 1× PBS and suspended by trypsinization. Cells were counted and adjusted to 2.5 to 5 × 104 cells per mL in medium. Seventy-five microliters of Matrigel (BD Biosciences) were plated into each well of a 96-well plate and allowed to polymerize at 37°C. Cell suspensions (200 μL; 8–12 × 103 cells) were added to each well. After 30 min, DMSO control or LFM-A13 was added. Thirty minutes later, dishes were treated with sham or 3 Gy and were then incubated until tubules had formed in control plates (4–6 h). Digital photographs were taken of individual wells and tubules were counted by an observer blinded to the treatment conditions. The mean and SE were calculated (n = 3).
Endothelial closure assay. HUVEC of passage 3 to 6 were grown on glass slides and were subjected to gap formation as described previously (28). Cells were treated with LFM-A13 or DMSO control before 0 or 3 Gy and incubated for the indicated times. Photographs of cell defect and surrounding cells were taken, and relative cell density within the defect was calculated as follows: (number of cells/original cell defect area)/(number of cells/surrounding area).
Apoptosis assays. HUVEC were grown to 70% to 80% confluency before treatment with LFM-A13 or DMSO control. Cells were then irradiated with 0 or 3 Gy and harvested 24 h after irradiation. Annexin V–FITC Apoptosis Detection kit (BD PharMingen) was used to stain cells [propidium iodide (50 ng) and Annexin V–FITC (5 ng) were added to 105 cells] for flow cytometry according to manufacturer’s protocol. For each treatment, the percentage of cells undergoing apoptosis (±SE) was calculated. Camptothecin (5 μmol/L) treatment for 6 h served as positive control.
HUVEC were also assayed for apoptosis using a 4',6-diamidino-2-phenylindole (DAPI) staining of the nuclei to identify cells undergoing morphologic changes. Seventy to eighty percent confluent HUVECs were treated with or without LFM-A13, incubated for 1 h, and then irradiated at 3 Gy. Cells were returned to the incubator for an additional 24 h before DAPI staining. Multiple high powered fields (at least seven) were examined by an observer who was blinded to the experimental conditions for each of the cultures. The percentage of cells demonstrating apoptotic nuclei were quantified. The mean and SE were calculated for each treatment group.
Tumor vascular window model. We have previously described the tumor vascular window model technique (29). Briefly, three mice for each group had LLC tumors grown within a vascular window such that tumor vasculature could be visualized within window frames containing a coordinate system for serial photography. Animals were treated with LFM-A13 by i.p. injection 60 min before 2 Gy irradiation using an 80 kVp superficial X-ray machine (Pantak X-ray Generator). Serial color photographs were taken to document blood vessel appearance on days 0 to 7. Photographs were scanned and processed using Adobe Photoshop software (Adobe) to mark the center of vessels, verified by an observer blinded to treatment groups. Vascular length density (VLD) was quantified for each microscopic field using ImagePro Plus v. 5.1 software (Media Cybernetics, Inc.). The mean and SE of VLD in each treatment group were calculated and plotted. All animals used were cared for according to Vanderbilt University's Institutional Animal Care and Use Committees guidelines.
Tumor immunohistochemistry. LLCs were s.c. injected into the hind limb of C57/BL6 mice to form syngeneic homografts. When tumors grew to 200 mm3, (∼7 d) the mice were treated with five consecutive daily treatments of i.p. LFM-A13 followed 45 min later by 3 Gy irradiation using an 80-kVp superficial X-ray generator. Twelve hours after the last radiation treatment, mice were sacrificed and tumors were harvested, fixed in paraffin, and sectioned by the Vanderbilt University Immunohistochemistry Core Facility as we have described previously (28). Immunostaining was with goat anti-CD34 (Santa Cruz Biotechnology), and microvascular photos were analyzed using ImagePro software with pixel number quantified.
Tumor growth delay. LLCs were s.c. injected into the hind limb of C57/BL6 mice to form syngeneic homografts. When tumors grew to 200 mm3, (∼7 d) the mice were treated with five consecutive daily treatments of i.p. LFM-A13 followed 45 min later by 3 Gy irradiation using an 80 kVp superficial X-ray generator. Serial measurements of tumor dimensions were taken by caliper, and tumor volume was calculated using the modified ellipsoid formula (length × width × depth)/2. The mean and SE were plotted using Microsoft Excel software.
Statistical analysis. The mean and SEs for all assays were calculated using Microsoft Excel software. Student's t test was performed to determine P values between treatment groups. P values of ≤0.05 were considered statistically significant.
Bmx is activated in endothelium upon irradiation. We examined primary culture vascular endothelial cells (HUVEC) to determine whether Bmx was activated by ionizing radiation, because of its similarities in structure and signaling with that of Akt, and was possibly contributing to radiation resistance. Figure 1A shows a time course of Bmx activation upon irradiation with a clinically relevant dose of 2 Gy. Tyr40, present in the pleckstrin homology domain of Bmx, becomes phosphorylated during its activation (7). Bmx is phosphorylated at 60 minutes after 2 Gy of irradiation. To confirm this finding, we used an in vitro kinase assay in which Bmx was immunoprecipitated from irradiated or sham-irradiated endothelial cells and then incubated with ATP in a kinase reaction. These samples were then run on SDS-PAGE and probed for antiphosphotyrosine to analyze autophosphorylation of Bmx. As shown in Fig. 1B, Bmx was activated after irradiation. Examination of total Bmx revealed no change in Bmx levels at any of the time points that were assayed. Densitometric quantitation (mean and SE) from four separate experiments is shown as well. Interestingly, Bmx showed significant kinase activity immediately after irradiation and then has a second peak of lesser activity at 1 hour.
Bmx knockdown enhances radiation efficacy in endothelium. Because we were able to detect a clear activation of Bmx after a clinically relevant dose of ionizing radiation, we wanted to determine whether Bmx activation protects the endothelial cells from cytoxic damage. Because primary culture endothelial cells, such as HUVEC, have low transfection yields, we used a retroviral shRNA system to knockdown Bmx levels before irradiation. Figure 2A shows five different retroviral constructs (A through E) for Bmx as well as a negative control construct (Neg) that were used to infect HUVEC. After 48 hours, infected cells were harvested and lysates were prepared for total Bmx Western blotting. As can be seen, construct A (shBmxA) provided ∼90% knockdown of Bmx protein levels compared with the negative control shRNA vector. Bmx knockdown experiments were performed with or without irradiation. Figure 2B shows MTT-based (WST-1) survival assay with HUVEC infected with either shBmxA or negative control vectors. After 48 hours, cells were counted and plated at 10,000 cells per well in duplicate within 96-well dishes. The cells were treated with either sham (0 Gy) or 2 Gy irradiation and incubated for 24 hours. After this incubation, WST-1 labeling mixture was added to each well and analyzed at absorbance at 450 nm to determine mitochondrial viability. Normalized values for absorbance at 450 nm are shown as mean and SE. Combined Bmx knockdown with irradiation decreases HUVEC survival.
Pharmacologic inhibition of Bmx. Having established that Bmx knockdown can enhance radiation efficacy in endothelial cells, we wanted to determine whether or not pharmacologic inhibition of Bmx would show the same effect. Bmx-specific inhibitors (30–33) have been described, particularly LFM-A13, which targets the Tec family. Because Bmx is the only Tec family member expressed in endothelium, we studied this drug in HUVEC. The drug, LFM-A13, has been shown to block vascular endothelial growth factor (VEGF)-induced signaling through Bmx inhibition in HUVEC at a dose of 25 μmol/L. Therefore, we used 30 μmol/L LFM-A13 for in vitro studies. Figure 3A shows 3 μmol/L (subtherapeutic) versus 30 μmol/L LFM-A13 preincubation on radiation-induced Bmx activation in the in vitro kinase assay at the time points with highest Bmx activation. As can be seen, 30 μmol/L but not 3 μmol/L LFM-A13 attenuates the activation of Bmx in response to 3 Gy.
Bmx inhibition attenuates endothelial cell viability. To determine whether LFM-A13 produces a radiosensitization effect in HUVEC, we studied clonogenic survival assays in HUVEC with LFM-A13 preincubation (Fig. 3B). HUVEC were pretreated with DMSO vehicle control or 30 μmol/L LFM-A13 45 minutes before irradiation with 0, 2, 4, or 6 Gy. Colonies were allowed to form >10 days, which were then counted, and the surviving fraction was calculated for each radiation dose. These studies indicated that 30 μmol/L LFM-A13 can radiosensitize HUVEC compared with the control as evidenced by the statistically significant downward survival curve shift. The dose enhancing ratio was 1.47.
Apoptosis was studied to determine whether this is a mechanism of enhanced cytotoxicity. Figure 3C illustrates the effect of LFM-A13 on apoptosis within these cells. HUVEC treated with 30 μmol/L LFM-A13 or DMSO control were subjected to sham or 3 Gy irradiation and then incubated for 24 hours before trypsinization and flow cytometric analysis. Annexin V–propidium iodide staining revealed that drug or 3 Gy alone was not capable of shifting cells into either early (Q4-1) or late (Q2-1) apoptosis but that the combination of LFM-A13 and 3 Gy caused a statistically significant (P < 0.001 versus LFM-A13 or 3 Gy alone) increase in apoptotis. To confirm these findings, HUVEC were treated with either 30 μmol/L LFM-A13 or DMSO control with or without 3 or 6 Gy irradiation and incubated for 24 hours. These cells were fixed and stained with DAPI, and the percentage of apoptotic cells was quantified. As shown in Fig. 3D, the combination of LFM-A13 and irradiation resulted in enhancement of apoptosis (*, P < 0.05 versus DMSO control; **, P < 0.001 versus LFM alone).
Bmx inhibition attenuates endothelial cell function. Functional assays of endothelial cells include cell migration and capillary-like tubule formation. Figure 4A illustrates the effect of LFM-A13 and irradiation on endothelial migration across a gap (endothelial cell closure assay) at 12 and 24 hours. HUVEC were plated on glass slides and grown to 80% confluency. A gap region, free of cells, was then created using a 200-μL pipette tip. These slides were then treated with 30 μmol/L LFM-A13 or DMSO control for 45 minutes before either 0 or 3 Gy. Cells were fixed and stained at 12 or 24 hours, and photographs were taken of the gap region and the surrounding cells to determine the ability of the HUVEC to migrate across and fill the gap. Relative cell density was calculated for each condition to control for the cytotoxic effects of treatment as shown in Fig. 4B. By 24 hours, control cells effectively migrated across the gap. Thirty micromoles of LFM-A13 or 3 Gy alone had minimal effect on attenuating endothelial cell closure at both 12 and 24 hours compared with vehicle-treated control. However, the combination induced a greater than additive effect that was statistically significant (*, P < 0.05 versus control; **, P < 0.01 versus LFM-A13). Figure 4C and D show capillary tubule formation assay. HUVEC plated onto Matrigel were treated with 30 μmol/L LFM-A13 or DMSO with or without 3 Gy irradiation and allowed to form tubules. The cells were then fixed and stained. The number of tubules were quantified and plotted. Representative photographs are shown in Fig. 4C and quantified in Fig. 4D. Cells that were treated with both LFM-A13 and 3 Gy showed a significant reduction (P < 0.005) in tubules formed compared with cells treated with either treatment alone.
Bmx inhibition attenuates tumor vasculature. To determine whether Bmx inhibition enhances radiation-induced destruction of tumor vasculature in vivo, we used i.p. injection of LFM-A13 before irradiation. A tumor vascular window chamber was placed on the dorsum of the C57BL6 mice, and LLC cells were implanted within the dorsal skin fold to allow for visualization of intravital tumor vasculature. Serial photographs were taken of the same region of the tumor, allowing for assessment of blood vessel formation and destruction. Figure 5A shows the effect of a single 50 mg/kg i.p. administration of LFM-A13 before 2 Gy irradiation. Representative photographs show that combination treatment results in dramatic reduction in tumor blood vessels. These results were quantified for each treatment condition as mean VLD with SE (Fig. 5B; P < 0.0014 versus LFM-A13 or 2 Gy alone). To confirm these findings, we used a hind limb syngeneic homograft model for determining vascular density within tumor sections. LLCs were implanted into the hind limbs of C57BL6 mice, and after tumors were formed, they were subjected to either daily LFM-A13 (50 mg/kg i.p. injection) or DMSO, followed 45 minutes later by 3 Gy or sham irradiation for a total of five treatments. The tumors were then harvested and prepared for immunohistochemistry analysis. Vessels were stained by anti-CD34 as shown in Fig. 5C, and these were quantified as shown in Fig. 5D. As can be seen, combination treatment was most effective at attenuating blood vessel formation (P = 0.043 versus ionizing radiation; P = 0.0001 versus LFM-A13 or vehicle control).
Bmx inhibition did not affect radiation sensitivity of LLC. Although we have shown the enhancement of radiation by Bmx inhibition within vascular endothelium, we determined whether Bmx inhibition could also affect radiation sensitization in the lung cancer cell line (LLC). As shown in Fig. 6A, LFM-A13 showed no difference in clonogenic survival compared with DMSO control in LLCs. This suggests that LFM-A13 enhancement of radiation was limited to its antivascular effect in this tumor model.
Bmx inhibition enhances radiation efficacy in tumor growth delay. Although LFM-A13 did not affect the radiosensitivity of the LLCs, LFM-A13 could still enhance radiation effects in vivo as an antivascular treatment. To determine whether treatment with LFM-A13 could enhance tumor growth delay in irradiated tumors, mice bearing LLC hind limb tumors were treated, as in Fig. 5C, with i.p. injection of 50 mg/kg LFM-A13 or DMSO 45 minutes before 3 Gy or sham irradiation for 5 consecutive days. The mean tumor volume and SE are plotted for each treatment group in Fig. 6B. Whereas LFM-A13 or radiation treatment alone resulted in a small growth delay, combination treatment showed a statistically significant enhancement of growth delay (P = 0.027). The fold increase in tumor volume at day 13 is also shown as mean and SE. These data suggest that LFM-A13 can enhance the efficacy of therapeutic radiation.
The purpose of this study was to determine the role of Bmx in the radiation response of tumor vasculature. We and others have shown that radiation induces Akt phosphorylation (4, 29, 34–36). Because the pleckstrin homology domains of Bmx and Akt bind the same phosphatidyl-inositol (3, 4, 5) phosphate, we hypothesized that Bmx might have a similar activation profile in response to ionizing radiation. Activation of Bmx occurred at clinically relevant doses of radiation: 2 to 3 Gy. Interestingly, we did note two waves of activation in the in vitro kinase assay. The early and more pronounced activation occurred immediately after irradiation, whereas the later wave occurred after 30 minutes and was less pronounced. It is possible that the second wave represents a “maturation event” such as phosphorylation of Tyr40 within the pleckstrin homology domain. Phosphorylation of this site has been shown to correlate with FAK activation in other cell types (7). shRNA knockdown of Bmx resulted in radiation sensitization in HUVEC, suggesting that Bmx inhibition may be a promising pharmacologic target for radiation enhancement. Although LMF-A13 is clinically used as a Btk inhibitor, many groups have used LFM-A13 as a Bmx inhibitor due to the high homology between Bmx and Btk. Because Btk is only found in bone marrow–derived cells, we felt that LFM-A13 could be used effectively for Bmx inhibition in our model systems. As we show in the in vitro kinase assay, LFM-A13 effectively attenuates Bmx activation in response to radiation. To confirm that this drug was not blocking all pleckstrin homology domain–containing kinases, we performed a time course and dose response of LFM-A13 with irradiation. Even at doses as high as 100 μmol/L, we saw no attenuation of radiation-induced Akt phosphorylation indicative of its activation (data not shown). This drug that specifically targets Bmx not only enhanced the cytotoxic effects of irradiation on HUVEC but also inhibited the function of these cultured endothelial cells. Apoptosis and clonogenic studies revealed that LFM-A13 was capable of inducing radiosensitization in HUVEC. Moreover, LFM-A13 in combination with radiation resulted in dramatic effects on endothelial cell migration as evidenced by the endothelial closure assay and tubule formation assay.
The vascular effects were more pronounced in the in vivo tumor models. Tumor vascular window blood vessels were minimally inhibited by 2 Gy or LFM-A13 alone. However, in combination, LFM-A13 and 2 Gy substantially disrupted tumor blood vessel formation. This antivascular effect was confirmed in hind limb tumor models that showed that daily LFM-A13 and 3 Gy significantly affected the tumor microvasculature. Tumor growth delay was displayed in the combination arm that was more than additive.
Tec family kinase inhibition has been garnering attention, although mainly in relation to anti-inflammation. Interestingly, ImClone has developed a Bmx single chain intrabody system that can partially attenuate Src transformation potential (37). Recently, Pan et al. (38) has discovered a number of selective irreversible Btk inhibitors aimed at treating rheumatoid arthritis. Moreover, CGI Pharmaceuticals, Inc. has been developing their own novel Btk inhibitors, cgi1316 and cgi1746, for use in inflammatory diseases. However, LFM-A13, a rationally designed inhibitor developed by Parker Hughes Institute, has been extensively tested in preclinical models. The pharmacokinetics and toxicity data have been previously published (31), which provided the basis for the present study. The drug seems to be well-tolerated based on these murine studies. We have also tested another commercially available Tec family inhibitor, terreic acid (33), which seemed to be at least as effective as LFM-A13 in our in vitro studies (data not shown). However, because very little is known about in vivo toxicity and pharmacokinetics for terreic acid, we focused our study on LFM-A13.
Several lines of evidence point to Bmx as a critical player in angiogenesis, cell survival, and proliferation, particularly in response to cancer-promoting factors such as VEGF, epidermal growth factor, androgens, and Src transformation (16, 18, 25, 39–45). Bmx is activated during cellular stress such as hypoxia (17, 41, 46, 47) and plays a cytoprotective role (39–45). Bmx activation up-regulates VEGF expression in endothelial cells, and VEGF can further activate Bmx in a “feed forward” manner (17). In addition, several cancers, including prostate (48, 49) and breast (7, 50), express Bmx that is also active. Other studies show that Bmx interacts with p53 after DNA damage from chemotherapeutics in a bidirectional inhibitory manner (48). Because a predominant effect of ionizing radiation is DNA damage, this interaction may play a role in radiation sensitization by Bmx inhibitors.
Bmx provides an alternative cell survival pathway to that of PI3K-Akt signal transduction. It is possible that treatments that have targeted PI3K-Akt signaling might be deriving some of their efficacy from concomitant Bmx inhibition. Further studies are necessary to determine whether Bmx inhibition in combination with PI3K-Akt blockade will provide additional benefit. Nevertheless, Bmx inhibition remains an attractive potential target for radiation enhancement because Bmx activation occurs rapidly and transiently after radiation, such that prolonged Bmx inhibition is probably not necessary for radiation sensitization to occur. This hypothesis is consistent with our cell culture assays in which the drug was removed shortly after irradiation. Therefore, it is possible that short-acting drug formulations may be effective with radiation with less systemic effects that typically occur with long-term administration.
In summary, Bmx is a new molecular target for radiation sensitization based on in vitro and in vivo experimentation in vascular endothelium. Ongoing studies using various cancer cell lines will help us to determine whether certain cancers are more susceptible to Bmx inhibition when treated with radiation. Ultimately, clinical evaluation of Bmx inhibitors with radiation will be critical during the development of Bmx as a biological target for therapy.
Grant support: Immunohistochemistry was performed with the following grant support U24 DK59637 and P30 CA68485-08. Additional grant support from RSNA HPSD0505, RSNA HPSD0605, DOD PC06015, R01-CA112385, 2R01-CA89674, R01-CA125757, P30-CA68485, Vanderbilt-Ingram Cancer Center, and Ingram Charitable Fund.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.