Synergistic antitumor effects of 9.2.27-PE38KDEL and ABT-737 in primary and metastatic brain tumors

Standard treatment, unfortunately, yields a poor prognosis for patients with primary or metastatic cancers in the central nervous system, indicating a necessity for novel therapeutic agents. Immunotoxins (ITs) are a class of promising therapeutic candidates produced by fusing antibody fragments with toxin moieties. In this study, we investigated if inherent resistance to IT cytotoxicity can be overcome by rational combination with pro-apoptotic enhancers. Therefore, we combined ITs (9.2.27-PE38KDEL or Mel-14-PE38KDEL) targeting chondroitin sulfate proteoglycan 4 (CSPG4) with a panel of Bcl-2 family inhibitors (ABT-737, ABT-263, ABT-199 [Venetoclax], A-1155463, and S63845) against patient-derived glioblastoma, melanoma, and breast cancer cells/cell lines. In vitro cytotoxicity assays demonstrated that the addition of the ABT compounds, specifically ABT-737, sensitized the different tumors to IT treatment, and improved the IC50 values of 9.2.27-PE38KDEL up to >1,000-fold. Mechanistic studies using 9.2.27-PE38KDEL and ABT-737 revealed that increased levels of intracellular IT, processed (active) exotoxin, and PARP cleavage correlated with the enhanced sensitivity to the combination treatment. Furthermore, we confirmed the synergistic effect of 9.2.27-PE38KDEL and ABT-737 combination therapy in orthotopic GBM xenograft and cerebral melanoma metastasis models in nude mice. Our study defines strategies for overcoming IT resistance and enhancing specific antitumor cytotoxicity in primary and metastatic brain tumors.


Introduction
Glioblastoma (GBM), arising from glial cells, is the most frequent and most malignant primary brain tumor in adults. The median survival (MS) for newly diagnosed GBM patients treated with the current standard of care, including surgery, radiation, and temozolomide PLOS ONE | https://doi.org/10.1371/journal.pone.0210608 January 9, 2019 1 / 17 a1111111111 a1111111111 a1111111111 a1111111111 a1111111111

Xenografts and cell lines
Human GBM xenografts (D-10-0021 MG, D-245 MG, and D-08-0326 MG) were established from patient tumors obtained through informed consent. The subcutaneous xenografts were harvested when the tumors reached an average volume of 500 mm 3 to 750 mm 3 ; tumor harvest date was between 30-45 days post-implantation. None of the subcutaneous tumors succumbed to ulceration or blistering. The maximum tumor volume threshold for the subcutaneous xenografts was set at 1500 mm 3 per our approved animal protocol A049-17-02. Human melanoma cell line (H350) were maintained in our laboratory. Human melanoma cell lines (DM440 and DM443) [31] were kindly provided by Dr. Douglas S. Tyler at Duke University. Human breast cancer cell lines (SUM159, SUM159-R113, and Hs 578T) [32] were kindly provided by Dr. Robin Bachelder at Duke University. All cells were cultured in an incubator at 37˚C, 5% CO 2 , and passaged when reached confluence with Accutase Cell Detachment Solution (BD Biosciences #561527). All GBM xenografts and the melanoma cell line H350 were maintained in 1x MEM ZINC Option media (Gibco #05-0009DJ) supplied with 10% fetal bovine serum (FBS). DM440 and DM443 cells were cultured in DMEM media (Gibco #11995) supplied with 10% FBS. SUM159 and SUM159-R113 cells were grown in Ham's F-12 base media (Gibco #11765) with the addition of 5% FBS, 1 μg/ml hydrocortisone, and 5 μg/ml human insulin. Hs 578T cells were maintained in DMEM media (Gibco #11995) with the addition of 10% FBS and 10 μg/ml human insulin.
Preparation of cell fractionations. Cells were harvested and resuspended in a cytosol extraction buffer [110 mM KOAc, 25 mM K-HEPES (pH 7.5), 25 μM MgCl 2 ] supplemented with 1 mM DDT, 150 μg/ml digitonin (Sigma #D141) and protease inhibitor cocktail. Cells were incubated on ice for 15 min, then the supernatant containing the cytosolic fraction was harvested by centrifugation at 10000 g for 10 min. The permeabilized cells were washed and centrifuged at 500 g for 2 min at 4˚C. The pellets were resuspended in an ER-solubilization buffer [200 mM KCl, 25 mM K-HEPES (pH 7.5), 10 mM MgCl2] supplemented with 1 mM DDT, 20 mg/ml n-dodecyl-b-D-maltoside (Sigma # D4641) and protease inhibitor cocktail. After a 30-min incubation on ice, the mixture was centrifuged at 10000 g for 10 min to remove nuclei, mitochondria, and incompletely solubilized cellular debris.
Global translation inhibition assay. Cells were treated with 5 μM puromycin (Tocris #4089) for 15 min, then washed twice and immediately lysed in polysomal lysis buffer supplemented with protease inhibitor cocktail.

Animal work approval
All animal work described in the manuscript have been reviewed and approved by the following animal research ethics committee of Duke University Medical Center: Duke Occupational and Environmental Safety Office, Duke Office of Animal Welfare Assurance, Duke Veterinary Committee, and finally the Duke Institute of Animal Care and Use Committee. The protocol number for the approved animal work is A049-17-02.

Anesthesia
Isoflurane anesthesia (Absolute Anesthesia Inc.) via inhalation was used during surgical procedures as it stabilizes the mice during surgery and allows for efficient post-surgical recovery. Euthanasia CO 2 , inhalation followed by decapitation was used for euthanasia. All animals were euthanized in accordance with the established humane endpoints listed in our approved animal protocol A049-17-02. For subcutaneous tumors, the humane endpoint is reached when tumor burden reaches 1500 mm 3 . For intracranial tumors, the humane endpoint is reached using neurological symptoms and body weight loss of 15% or the animals' inability/desire to move (the mouse does not move forward two steps when prompted gently).

Intracranial tumor models
All experiments were done in accordance with the Institutional Animal Care and Use Committee of Duke University Medical Center (A049-17-02). Animals were group-housed, maintained in a barrier facility, under pathogen-free conditions according to NIH guidelines. Nude mice (�22-30 g, 6-8 weeks, female:male = 1:1, Duke University, Division of Laboratory Animal Resources) were anesthetized by isoflurane inhalation and mounted onto the stereotactic frame (Stoelting Co.). The anterior cranial region was shaved, and an incision �1 cm in length was made in the skin over the skull. A 25-gauge needle attached to a 25-μl Hamilton syringe was used to pierce through the skull at coordinates 2.0 mm left lateral of the sagittal and 0.5 mm anterior to the bregma. The needle was inserted vertically to a depth of 2.5 mm from the dura mater. A total of 1x10 5 D-10-0021 MG GBM cells freshly dissociated from a subcutaneous mouse xenograft, or 1x10 5 DM440 melanoma cells harvested from culture, were injected in 5 μl of 1xPBS containing 2% methylcellulose (Sigma #M0512).
Five days post-tumor implantation, the animals were randomized into four treatment groups, with 9-10 mice per group: vehicle control, 9.2.27-PE38KDEL, ABT-737, and 9.2.27-PE38KDEL/ABT-737 combination therapy, randomized according to initial weight. For each mouse in the study, a brain infusion cannula (Alzet #0008851) attached to a subcutaneously implanted mini-osmotic pump (Alzet #1007D) was inserted directly into the intracranial tumor site for intratumoral delivery of the vehicle solution or the drugs, at a rate of 0.5 μl/ h for 72 hours. The vehicle control group received the PBS-based solution containing 5% Captisol and 2% mouse serum albumin. The IT monotherapy group received a total dose of 0.1 μg of 9.2.27-PE38KDEL diluted in the vehicle solution. The ABT-737 group received a total dose of 2.93 μg of ABT-737 (equivalent to 36 μl of the 100 μM of ABT-737 diluted in the vehicle control solution). The combination group received both 9.2.27-PE38KDEL and ABT-737 at the doses indicated above. During the surgery, animals in different groups were treated in a random order. Animals were observed twice daily for signs of distress. Survival endpoint is defined as the onset of neurologic symptoms (lethargy, seizure, repetitive circling, difficulty breathing, and hunched posture), greater than 15% loss of body weight, or death, whichever comes first. When endpoint was reached, mice were euthanized as described earlier. Kaplan-Meier survival curves were plotted and compared using log-rank test. Each study was repeated at least twice.

Expression of CSPG4 on tumor cells and in vitro cytotoxicity of IT, ABT, A-1155463, and S63845 monotherapies
Flow cytometry analysis (FACS) revealed the presence of CSPG4 on all GBM, breast cancer, and melanoma cell lines tested, albeit at varying levels ( Fig 1A-1C). Q-FACS analysis showed that the surface density of CSPG4 was lowest in SUM159-R113 cells (�100,000 molecules per cell) and highest in DM440 cells (�700,000 molecules per cell) ( Fig 1D). Despite abundant CSPG4 cell surface expression, in vitro cytotoxicity assays demonstrated that all GBM xenografts, all breast cancer cell lines, and the DM443 melanoma cell line were highly resistant to both Mel-14-PE38KDEL and 9.2.27-PE38KDEL ITs (IC 50 >100 ng/ml) (Tables 1 and 2). Only the H350 melanoma cells were susceptible to the cytotoxicity of 9.2.27-PE38KDEL (IC 50 of 11.67 ng/ml; Table 2). Both H350 (IC 50 = 42.50 ng/ml) and DM440 (IC 50 = 68.30 ng/ml) melanoma cells showed weak cytotoxicity upon Mel14-PE38KDEL treatment alone (Table 1).
These data suggest substantial inherent resistance to CSGP4-targeting ITs in the majority of cancer cells expressing the target. To explore the utility of pro-apoptotic enhancers in countering this resistance, we first evaluated the sensitivity of the cell lines in our panel to ABT-737, ABT-263, ABT-199, A-1155463, or S63845 at a concentration of 5-20 μM (see S1, S2A, S2B and S2C Figs for the IC 50 values). The absorbance of ABT/A-1155463/S63845-treated (Test) vs. vehicle-treated cells (0.5% DMSO in media) (Control) was used to calculate cell viability. A concentration of ABT/A-1155463/S63845 compound that yielded at least 70% cell viability (sub-therapeutic dose) in all three cell lines from GBM, melanoma, or breast cancer was selected for combination therapy. Thus, the concentration of ABT-737 chosen for combination with the ITs was 20 μM (GBM, melanoma), and 10 μM for breast cancer cell lines (S1A, S1B and S1C Fig). For ABT-263 combinations, we chose concentrations of 5 μM (GBMs, breast cancer), and 10 μM for the melanoma cell lines (S1D, S1E and S1F Fig). For ABT-199, the concentrations were 10 μM (GBMs, melanoma), and 5 μM for the breast cancer models (S1G, S1H and S1I Fig  stronger synergy with all three ABT compounds than Mel-14-PE38KDEL (Tables 1 and 2 Table). Moderate synergy (CI = 0.622 to 0.719) was observed in DM440 cells with the 9.2.27-PE38KDEL+ABT-737 combination at concentrations of �0.1 ng/ml (S1 Table). For the SUM159-R113 cells, synergy with the combination therapy was observed only at concentrations of �10 ng/ml (CI = 0.0172) (S1 Table).

Internalization of the 9.2.27-PE38KDEL alone or in combination with ABT-737
Next we investigated the role of CSPG4 internalization to account for the difference in the sensitivity of tumor cells to the 9.2.27-PE38KDEL, ABT combinations. Flow cytometry analysis revealed that D-10-0021 MG, DM440, and SUM159-R113 cells internalized 95.5%, 74%, and 46% of the surface-bound 9.2.27-PE38KDEL by 4 h, respectively (Fig 2A). 9.2.27-PE38KDEL internalization was the least efficient in Hs 578T cells, with only 11.5% of IT internalized at 4 h (Fig 2A). Internalization of 9.2.27-PE38KDEL was examined by western analysis of D-10-0021 MG cells treated with 9.2.27-PE38KDEL, ABT-737, or the combination at different intervals ( Fig 2B). One hour post-treatment, 41% and 61% of intact IT was internalized with the 9.2.27-PE38KDEL and 9.2.27-PE38KDEL+ABT-737 therapies, respectively. The maximal level of intact IT internalization with either monotherapy or the ABT-737 combination was reached at 4 h post-treatment (the relative band intensity of which was arbitrarily set as 100% for comparison). Notably, when compared to IT monotherapy, ABT-737 addition did not significantly affect the maximum level of intracellular, intact IT at 2-4 h post-treatment. An antibody against the ADP-ribosylating catalytic domain of Pseudomonas exotoxin was used to determine intact IT internalization and the rate of intracellular IT cleavage by furin. The 38 kDa cleaved IT fragment, was detected 1 h post-treatment with 9.2.27-PE38KDEL alone or in combination with ABT-737 (Fig 2B). A gradual increase in the accumulation of the cleaved exotoxin was observed up to 6 h of treatment. However, only 9-12% of the maximal level of intact, internalized IT was cleaved after 4 h of treatment (Fig 2B).
Since the maximal level of intact 9.2.27-PE38KDEL was detected intracellularly at 4 h posttreatment, we extended the time course study to 24 h for examining the accumulation of intact/cleaved IT (Fig 2C). The results confirmed that for D-10-0021 MG cells treated with either the IT monotherapy or the combination, the intracellular level of intact IT gradually declined after 4 h post treatment. Accumulation of cleaved exotoxin continued to increase till 8 h post-treatment, after which it began to decline, possibly due to proteasomal or/and caspase-dependent degradation (Fig 2C).

Inhibition of global translation and poly ADP ribose polymerase (PARP) cleavage by mono and combination therapies
Since IT-mediated cytotoxicity is through protein synthesis inhibition followed by apoptosis [11], puromycylation assays were conducted to assess the rate of global protein synthesis inhibition upon IT treatment in D-10-0021 MG, DM440, and SUM159-R113 cells. At 8 h post-treatment of D-10-0021 MG, the 9.2.27-PE38KDEL/ABT-737 combination inhibited global translation by 69%, compared to 36% or 13% with 9.2.27-PE38KDEL and ABT-737 monotherapies, respectively (Fig 3A). Similarly, DM440 (8 h post-treatment) and SUM159-R113 cells (16 h post-treatment), exhibited global protein synthesis reduction to a much greater extent in the combination therapy group compared to either of the monotherapies (Fig 3B and 3C).
Caspase-dependent PARP cleavage, which is often associated with apoptosis, was observed as early as 16 h post combination treatment in D-10-0021 MG, where the majority of intact PARP was cleaved (Fig 3A). On the other hand, neither monotherapy was able to induce detectable PARP cleavage, even at 24 h post-treatment ( Fig 3A). Likewise, in both DM440 ( Fig  3B) and SUM159-R113 cell lines (Fig 3C), a significantly greater extent of PARP cleavage was observed in the combination therapy compared to the monotherapies. However, near-complete cleavage of PARP (around 90%) was only observed 24 h post combination therapy in DM440 (Fig 3B), and 30 h post combination therapy in SUM159-R113 (Fig 3C). Complete PARP cleavage with the 9.2.27-PE38KDEL/ABT-737 combination at earlier time points in the D-10-0021 MG GBM cells compared to DM440 and SUM159-R113 cell lines, corresponds to elevated in vitro cytotoxicity noted with the combination therapy in this model ( Table 2). Time course analysis of global translation inhibition and total intact PARP levels (100% intact PARP levels were observed at �70-80% translation inhibition) in D-10-0021 MG, DM440, and SUM159-R113 cells confirmed that inhibition of protein synthesis by processed IT resulted in PARP cleavage and apoptosis (S5 Fig).

Relative amount of cleaved exotoxin following the combination treatment of 9.2.27-PE38KDEL and ABT-737
Upon receptor binding and internalization, ITs are activated by furin cleavage. To determine IT activation following 9.2.27-PE38KDEL (750 ng/ml) and ABT-737 (10 μM) combination, .5-fold more intact intracellular IT than DM440 and SUM159-R113, respectively (Fig 4A). Cleaved exotoxin in D-10-0021 MG cells was detected at 2 h post-treatment and reached maximum levels at 8 h post-treatment. In contrast, significantly lower levels (5-20%) or no cleaved exotoxin was observed in DM440 and SUM159-R113, respectively (Fig 4A). Again, at 16 h post combination therapy, near-complete PARP cleavage was detected in D-10-0021 MG cells, while there was minimal PARP cleavage in both DM440 and SUM159-R113 cell lines (Fig 4A). Bcl-xL levels declined at the 16 h time point in D-10-0021 MG cells, which could be due to caspase-dependent degradation (in correlation with PARP cleavage). No significant change in Bcl-xL levels were observed in DM440 and SUM159-R113 cells (Fig 4A). Total loss of Mcl-1 was noted at 16 h post combination therapy after the appearance of cleaved exotoxin in D-10-0021 MG cells (Fig 4A).
Since furin-processed exotoxin needs to translocate from the ER to cytosol to inhibit protein synthesis, the relative amounts of intact and cleaved exotoxin in the cytosol and ER fractions of D-10-0021 MG and DM440 cells were investigated. Cell lysates were fractionated into cytosolic, ER, and total fractions. Immunoblot analysis of marker protein distributions confirmed the presence of tubulin only in cytosolic fractions, and the ER-membrane protein Mcl-1 only in the ER fraction (Fig 4B), thus establishing proper separation of the different cellular fractions. At 4 h post-treatment for both D-10-0021 MG and DM440 cells, intact ITs were predominantly located in the ER fraction (Fig 4B). In D-10-0021 MG, cleaved exotoxins were detected in both ER and cytosolic fractions, whereas in DM440 cells, they were observed only in the ER fraction ( Fig 4B). Our data indicate that both internalization and furin-dependent cleavage of IT was ineffective for SUM159-R113 cell line (Fig 4A). In DM440 cell line, internalization of IT was efficient, but cleavage and translocation of cleaved exotoxin from ER to cytosol was significantly lower than that of D-10-0021 MG cells (Fig 4A). A minimal increase (�1.6 fold) observed in the cleaved exotoxin in cytosolic fractions of D-10-0021 MG cells following 9.2.27-PE38KDEL+ABT-737 treatment, might contribute to the higher sensitivity of the GBM cells to combination therapy. Thus, in the GBM PDX D-10-0021 MG, rapid internalization (1-2 h) of IT, efficient cleavage (2-8 h) and translocation of cleaved (active) exotoxin from ER to cytosol, complete PARP cleavage, and Bcl-xL and Mcl-1 degradation resulted in robust response to 9.2.27-PE38KDEL+ABT-737 therapy. Taken together the mechanistic studies show that the levels of intracellular IT, processed exotoxin, and PARP cleavage determine the sensitivity of tumor cells to the combination treatment. Our study provides a rationale for the difference in the in vitro efficacy of GBM, melanoma, and breast cancer cells to the 9.2.27-PE38KDEL+ABT-737 combination ( Table 2).

In vivo efficacy of 9.2.27-PE38KDEL and ABT-737 combination therapy
Orthotopic mouse model of primary brain tumors were established using GBM patientderived xenograft D-10-0021 MG cells. Histological analysis of brains five days post-tumor implantation demonstrated the presence of tumor mass in D-10-0021 MG model (Fig 5A). Thus, day five post-implantation was chosen for treatment initiation. Convection-enhanced delivery (CED), utilizing osmotic pumps, has been successfully used to bypass the blood-brain barrier and to deliver ITs directly into brain tumors [18]. The toxicity of 9.2.27-PE38KDEL was evaluated by CED at a total dose of 0.1 μg or 0.3 μg in mice bearing DM440 intracranial tumors. Toxicity events at an occurrence rate of 1-2 mice per every ten mice were observed only in the 0.3 μg 9.2.27-PE38KDEL group. A total dose of 100 μM (2.93 μg) of ABT-737 by CED was found to be safe for the mice without causing any toxicity. Therefore, a total dose of 0.1 μg of 9.2.27-PE38KDEL and 100 μM of ABT-737 were chosen for the in vivo combination studies.
In the D-10-0021 MG model, orthotopic delivery of 100 μM of ABT-737 did not improve MS compared to the vehicle control (p = 0.2871, Fig 5B). The 9.2.27-PE38KDEL (0.1 μg) monotherapy exhibited a modest 10.7% increase in MS compared to the control but failed to reach statistical significance (p = 0.0716). Compared to the vehicle control, the 9.2.27-PE38K-DEL+ABT-737 combination therapy prolonged the MS by 60.7% (p = 0.0012). The survival benefit of the combination was also statistically significant when compared to the 9.2.27-PE38KDEL (p = 0.0100) or ABT-737 (p = 0.0198) monotherapies. More importantly, 2/ 9 (22.2%) mice in the combination therapy group were tumor-free, as confirmed by the H&E staining of their brains (Fig 5C) at the termination of the study (Day 81).
Orthotopic mouse model of metastatic brain tumor was established using DM440 melanoma cells. Histological analysis of DM440 brains (Fig 6A) five days post-tumor implantation demonstrated the presence of tumor mass. In the DM440 model, neither 9.2.27-PE38KDEL (p = 0.1667) nor ABT-737 monotherapies improved MS (p = 0.1822) (Fig 6A). However, compared to the vehicle control, there was an 18.2% increase in MS for the combination group (p<0.0001). The survival benefit with the 9.2.27-PE38KDEL+ABT-737 combination was also statistically significant compared to the ABT-737 (p = 0.0004) or the 9.2.27-PE38KDEL (p = 0.0002) monotherapies.

Discussion
Brain tumors are composed of highly heterogeneous tumor cell populations that frequently harbor inherent resistance to therapeutic agents, resulting in incomplete eradication of tumor cells and tumor recurrence. Thus, sensitizing tumor cells is critical for improving the efficacy of IT-based therapies. By combining CSPG4 targeting ITs with ABT-737, targeting Bcl-2 family members, we were able to overcome widespread, inherent IT resistance in a panel of GBM, melanoma, and breast cancer cell lines. The IT+ABT-737 combination improved in vitro therapeutic efficacy up to >1,000-fold (Tables 1 and 2). The ability of ABT-737 to sensitize cancer cells to 9.2.27-PE38KDEL-mediated cytotoxicity was mechanistically confirmed through western analysis, where the combination therapy, but not the monotherapies, enhanced exotoxin cleavage, induced global translation inhibition, and PARP cleavage. Accordingly, only the combination therapy generated significant improvement in MS, and in some cases even cures, in mouse models of primary or metastatic brain tumors (Figs 5 and 6).
Our in vitro studies demonstrated variability in the sensitivity of individual cell lines/xenografts to 9.2.27-PE38KDEL+ABT-737 combination, i.e., cytotoxicity (IC 50 ) improvement of >1000-fold (D-10-0021 MG), >200-fold (DM440), and >17-fold (SUM159-R113) ( Table 2). Interestingly, while Hs 578T and D-10-0021 MG cells expressed similar levels of CSPG4 (Fig  1), Hs 578T failed to show improvement in cytotoxicity with all of the combinations tested (Tables 1 and 2), demonstrating that the surface CSPG4 density is not a critical determinant of tumor cell sensitivity to combination therapy. Examination of selected prosurvival and proapoptotic Bcl-2 family proteins at basal level and following 9. However, flow cytometry assay (Fig 2A) revealed that the internalization rate of 9.2.27-PE38K-DEL is a major determinant of tumor cell sensitivity to combination therapy.
Once internalized, ITs are processed by furin in the endosome and translocated from ER to the cytosol for ADP ribosylation of EF2 and protein synthesis inhibition. Western analysis revealed that at 8 h post combination treatment there was a 12-fold increase in cleaved exotoxin in D-10-0021 MG compared to DM440 (Fig 4A). Moreover, cell fractionation studies revealed that at 4 h post-treatment (9.2.27-PE38KDEL or 9.2.27-PE38KDEL+ABT-737), cleaved exotoxins were found in both cytosolic and ER fractions of D-10-0021 MG cells, while they were detected only in the ER fractions of DM440 ( Fig 4B). Thus, increase in cytosolic levels of cleaved exotoxin in D-10-0021 MG compared to DM440 cells corresponds to their enhanced sensitivity to 9.2.27-PE38KDEL and 9.2.27-PE38KDEL+ABT-737 therapies.
In D-10-0021 MG cells, compared to 9.2.27-PE38KDEL monotherapy there was �1.6 fold increase in cleaved exotoxin levels post 9.2.27-PE38KDEL+ABT-737 combination (Fig 4B). Thus, in contrast to previous studies [22], our results in the GBM xenograft D-10-0021 MG demonstrate that the addition of ABT-737 does not significantly alter the translocation of cleaved exotoxin from ER to cytosol. However, we observed PARP cleavage in all three tumor models (D-10-0021 MG, DM440, and SUM159-R113) post combination therapy, but not with 9.2.27-PE38KDEL or ABT-737 monotherapies (Fig 3). The efficiency of PARP cleavage correlated well with in vitro and in vivo efficacy of the combination therapy.
In the D-10-0021 MG model, compared to vehicle control, 9.2.27-PE38KDEL+ABT-737 combination therapy increased MS by 60.7% (p = 0.0012). The efficacy of combination therapy was highly significant when compared to ABT-737 (p = 0.0198) and 9.2.27-PE38KDEL monotherapies (p = 0.0100; Fig 5A). Furthermore, the combination therapy generated cures in 2/9 mice, as verified by histological staining (Fig 5C). Consistent with the lower sensitivity of DM440 cell line to the combination therapy in vitro, a modest increase in survival post combination was observed in vivo. Collectively, these data confirmed that the 9.2.27-PE38KDEL +ABT-737 combination therapy was able to overcome tumor cell resistance to IT monotherapy in vivo, delayed tumor growth, and, in some cases, generated cures.
While several groups have tried to gain mechanistic insights into various immunotoxin +Bcl-2 inhibitor anticancer therapies, only two studies [23,33] have shown modest antitumor efficacy (improvement in survival by several days without cures), in in vivo subcutaneous small cell lung cancer and melanoma tumor models. To achieve this modest improvement in survival, the authors utilized a total of eight doses of 50 mg/kg of ABT-737 + 0.4 mg/kg immunotoxin [23] or a total of two doses of 50 mg/kg of ABT-737 and 0.031 mg/kg immunotoxin [33]. Importantly, in our current intracranial study, we were able to generate 22% cures and improvement in survival utilizing a single dose of 0.147 mg/kg of ABT-737 and 0.005 mg/kg immunotoxin against the aggressive glioblastoma tumor. The ABT-737 and immunotoxin doses in the current study are �680-2,700-fold and �12-6,400-fold lower, respectively than the previous studies.
In conclusion, our in vitro studies, using a panel of GBM, melanoma, and breast cancer cell lines showed that Bcl-2 inhibitor ABT-737 reversed the resistance of tumor cells to IT treatment. We further confirmed the ability of 9.2.27-PE38KDEL+ABT-737 combination to overcome tumor resistance in orthotopic models of brain tumors. Mechanistic studies using 9.2.27-PE38KDEL and ABT-737 mono-and combination therapy revealed that that increased levels of intracellular IT, cleaved exotoxin, and PARP cleavage were determinants of enhanced sensitivity of tumor cells to the combination. Addition of ABT-737 had little effect on the rate of IT internalization but generated a small increase in the translocation of cleaved exotoxin from ER to cytosol. Our study provides insights into employing IT+ABT combinations for overcoming therapy resistance in primary and metastatic brain tumors.