MP-Pt(IV): A MAOB-Sensitive Mitochondrial-Specific Prodrug for Treating Glioblastoma

The prognosis for patients with glioblastoma (GBM) is often grim, with limited chemotherapeutic options after surgical resection of tumors. The current standard of care is the Stupp protocol, which involves aggressive tumor resection, whole-brain/focal radiation with temozolomide, and then temozolomide chemotherapy (1, 2). Encouraging data from phase III clinical trials with drug combinations such as lomustine/temozolomide (median survival, 48.1 on lomustine/temozolomide vs. 31.4 months on temozolomide; ref. 3) or tumor treatment fields (median survival, 25.2 months; ref. 4), or immunotherapy (5) demonstrate significant recent therapeutic advances. Ultimately, resistance to temozolomide develops and patients undergo salvage therapy (chemotherapy or chemoradiotherapy), providing minimal efficacy at best, and significant toxicity.

GBM tumors tend to be hypoxic, and GBM cells are known to express active oxygen-sensing master gene controllers, HiF-1 and HiF-2 (6, 7). GBM tumors respond to hypoxia by metabolic remodeling, involving the upregulation of nonaerobic pathways, downregulation of mitochondrial respiration, and a reduction in mitochondrial mass (8, 9). The mitochondria of cancer cells generally have elevated ascorbate levels (10). These high levels result from the importation of dehydroascorbate into the cells by Glut transporters, then into mitochondria by Glut10, the SLC2A10 gene product, and subsequent reduction to ascorbate by glutathione S-transferase omega-1 or 2. Glut10 is highly expressed in GBM, consequently elevating mitochondrial ascorbate (11). GBM can acquire resistance to temozolomide via elevated mitochondrial ascorbate (12). Unlike nuclear DNA, mitochondrial DNA lack robust repair mechanisms (13) and are thus, sensitive to DNA-damaging chemotherapeutics. Mitochondria are thus very attractive chemotherapy targets and mitochondrial enzymes, such as monoamine oxidase B (MAOB; ref. 13), are being explored for drug development (14).

MAOB is located on the inner surface of the outer mitochondrial membrane and catabolizes monoamine substrates like dopamine. MAOB activity is elevated in GBM tumors (15) and expression is associated with poor patient outcomes (16). This overexpression of MAOB in GBM appears to be linked to the Warburg shift (17) and HiF-1 activation (18–20). We showed that cultured primary GBM cells have >3-fold higher MAOB activity than normal human astrocytes (NHA), and yet have 5- to 10-fold fewer mitochondria than NHAs (13).

We have previously reported the in vitro and in vivo efficacy of N,N-bis(2-chloroethyl)-2-(1-methyl-1,2,3,6-tetrahydropyridin-4-yl)propenamide (MP-MUS), an MAOB-activated prodrug that targets GBM mitochondria (21). MAOB oxidizes the uncharged methyl-tetrahydropyridine (MP-) group into its mitochondrially targeted cationic form, methyl-pyridinium (P⁺-). Lipophilic cations, like P⁺-, cross the inner mitochondrial membrane and extensively accumulate in the mitochondrial matrix, driven by the large (∼−180 mV) mitochondrial membrane potential. At 37°C, the mitochondria/cytosol accumulation ratio of [P⁺-] is approximately 1,000:1 (22, 23).

MP-MUS was shown to damage GBM mitochondria and kill GBM cells in culture. Unfortunately, the intrinsic reactivity of the nitrogen mustard group and low solubility of MP-MUS precluded clinical development. Our criteria for a suitable replacement for the nitrogen mustard group thus, included (i) small size allowing the prodrug to access the MAOB substrate channel, (ii) lipophilicity to cross the blood–brain barrier, (iii) low reactivity until inside the mitochondrial matrix, and (iv) ability to damage mitochondrial matrix contents.

MP-Pt(IV), our second-generation mitochondrial-targeting prodrug incorporates all our design criteria for selective delivery and activation of an inert prodrug into the mitochondrial matrix of GBM (Fig. 1). MP-Pt(IV) utilizes the same mitochondrial-targeting MP moiety (21, 24) as MP-MUS, but replaces the nitrogen mustard group of MP-MUS with an inert, asymmetric Pt(IV)-ethylene glycol chloride complex, -Pt(IV) (25). Unreactive Pt(IV) drugs can be reduced into their highly reactive Pt(II) daughter compounds by cellular ascorbate and thiols (26, 27). In silico modeling suggested that the small and hydrophobic -Pt(IV) group (28) would allow MP-Pt(IV) to circulate, cross the blood–brain barrier, and readily access the substrate pocket of MAOB, where it would be rapidly oxidized to give P⁺-Pt(IV).

We expected that P⁺-Pt(IV) would be rapidly converted into cisplatin in GBM mitochondria due to high ascorbate levels. Within GBM mitochondria, the regenerated cisplatin could deplete glutathione [GSH, thereby impairing detoxification of reactive oxygen species (ROS)] and induce cell death (29, 30).

The differential sensitivity of GBM toward MP-Pt(IV) emerges from three key factors: (i) GBM has much higher levels of MAOB than other tissues (15, 31), restricting the generation of mitochondrially targeted P⁺-Pt(IV), (ii) mitochondrial matrix P⁺-Pt(IV) levels will be highest in cells where (MAOB:mitochondria) ratios are highest, found in GBM (17), and (iii) high ascorbate levels in the GBM matrix drive the conversion of inert Pt(IV) into cisplatin (12).

Herein, we report the synthesis and results from in vitro and in vivo studies of MP-Pt(IV) in a primary xenograft mice GBM model. We postulated that the MAOB (32) and SLC2A10 (33) genes undergo simultaneous activation via HiF-induced changes in Sp1/Sp3 levels. To examine this hypothesis, we analyzed the overall survival (OS) outcomes of patients with GBM, with respect to MAOB and SLC2A10 mRNA levels, from the major GBM transcriptome databases, The Cancer Genome Atlas (TCGA)-GBM (RNA Seq V2 RSEM), Gravendeel, and Chinese Glioma Genome Atlas.

The Nernst equation can be used to calculate the accumulation of cations in the mitochondrial matrix, driven by the approximately −180 mV mitochondrial membrane potential (ΔΨ):

At 37°C, the mitochondria/cytosol accumulation ratio of [P⁺-] will be approximately 1,000:1 (22, 23).

We selected three genes to serve as proxies of mitochondrial numbers: COX15, TFAM, and MTPAP, each of which has been shown to correlate with steady-state levels of mitochondria in normal and cancer cells (34–39). Three databases were examined to determine the transcript levels of genes and patient OS (months). Optimal high/low transcript cutoffs for the generation of Kaplan–Meier survival curves were identified from the maximum difference of ranked-means and statistical significance was calculated using the log-rank χ² test. (28)

MP-Pt(IV) was obtained by HBTU-mediated coupling of Pt(IV)-glycol (25) and MP-carboxylic acid (21), followed by chromatographic purification, detailed in the Supplementary Data.

Mass spectra were determined on a Thermo-Scientific LCQ FLEET spectrometer in electron impact mode. Proton (¹H) NMR spectra were recorded at 600 MHz on a Bruker spectrometer using the solvent (DMSO-d6) as the internal standard. J values are given in hertz (Hz). TLCs were performed using Merck glass sheets precoated with Kieselgel 60 F₂₅₄ (0.2 mm) as the adsorbent and visualized with UV light (254 and 365 nm). Column chromatography was conducted under medium pressure on silica (Kieselgel 60, 240–400 mesh). Reagents were purchased from Sigma-Aldrich and used as received.

Amplex Red (150 μmol/L) and HRP (3 U/mL) were incubated with recombinant human (rh) MAOA or rhMAOB (1 U/mL) in PBS buffer, pH 7.4, at 37°C in 96-well format (200 μL volume/well). Because the small amounts of radicals generated by the photosensitization of resorufin (40) could reduce Pt(IV) into cisplatin, which in turn quenched the fluorescence of resorufin, we measured the formation of resorufin using the absorbance of 572–672 nm in a BioTek Synergy HT Spectrophotometer at t = zero and at 40 minutes.

Cells were fixed in ice-cold 4% paraformaldehyde for at least 1 hour and then washed in PBS (Thermo Fisher Scientific). MitoTracker, MitoSox Red, and H2-DCF assays were performed in nonpermeabilized cells. The activity of caspase-3 in fixed, 0.1% Triton-permeabilized cells was measured using the Molecular Probes R110-EnzChek Assay Kit (catalog no., E13184), incubating cells for 1 hour at 37°C.

Human glioma cells were harvested from patients during surgical excision and isolated within 10 minutes after removal. The primary cultures used here, GBM115 and GBM157, are highly aggressive and heterogeneous, with high MGMT expression and temozolomide IC₅₀ value (24 hours) >1 mmol/L. Tumors were homogenized with a pipette, and cells were grown for 2 weeks in DMEM supplemented with 20% FBS, GlutaMAX-I, sodium pyruvate, and penicillin/streptomycin. Glioma cells were grown to confluence 24 hours after treatment with an identical volume of the drug (in DMSO) or DMSO alone (maximum 0.04% v/v of DMSO/medium); cells were cultured in Costar 96-well Plates (Corning). After treatment, cells were grown for 24 hours in the presence or absence of all effectors (total volume of 250 μL). GBM115 and GBM157 were used at a low passage number in all in vitro experiments.

To modulate intracellular ascorbate, we preincubated primary GBM cell cultures overnight with unsupplemented control media and with media supplemented with 5 mmol/L dehydroascorbate or 5 mmol/L ascorbate. The media were replaced after 24 hours, and cells were incubated with MP-Pt(IV) for 24 hours, then given a viability dye, fixed, and counted. Additional control experiments were run after blocking MAOB activity by preincubation with 10 μmol/L selegiline followed by incubation with 14 μmol/L MP-Pt(IV) or incubation of GBM cultures with the precursor Pt(IV)-ethylene glycol at 140 μmol/L for 24 hours.

Study approval: tumors/resulting cultures obtained from GBM tissues had no identifiable patient information under institutional review board protocol 00014547. Animal research was conducted according to the Institutional Animal Care and Use Committee (IACUC) protocol AUP-0818-0043, approved by the IACUC of Methodist Hospital (Houston, TX). All animal care procedures conformed to the Guide for the Care and Use of Laboratory Animals (National Research Council, National Academy Press, Washington DC 1996).

We utilized a reported method for injecting human GBM cultures into the mouse brain via the postglenoid foramen (41). Primary GBM tumors were initially grown in the flanks of tumor donor mice, then excised, chopped, and homogenized in ice-cold PBS, 1:1, using a BeadBug benchtop homogenizer. The homogenized cells were diluted with a solution of Matrigel media (4:6), kept on ice, and were injected into the postglenoid foramen of 48 female athymic nude mice over 3 hours. On day 8 after xenografting, the 42 heaviest animals were selected and randomized into seven groups of 6 animals. Therapy began on day 11, and the groups underwent six different treatment regimens (Fig. 6). MP-Pt(IV) (15 mg/kg) was administered via tail vein injection in saline; temozolomide (5 mg/kg) was given by gavage, and whole-head radiation (2 Gy) was administered from a cesium source. Control group animals received vehicle by injection or gavage and were given sham radiation. All animals that received either radiation or sham radiation were first anesthetized with isoflurane inhalation, then by subcutaneous dexmedetomidine (0.5–1.0 mg/kg). Seven animals were placed in the thermostatically ballasted radiation shield then transferred into the Cs-safe, and then rod elevated (treatment) or not (sham). Postradiation/sham, animals were administered with the anesthesia antidote, atipamezole (1 mg/kg), subcutaneously.

The animals were monitored and weighed daily during the first 4 months, and then every 3–5 days thereafter. Animals were sacrificed when they reached an ethical endpoint and their brain/tumors were preserved in formaldehyde. The study was terminated on day 300 as the nude mice begin to suffer age-dependent maladies.

MP-Pt(IV) was designed to preferentially kill cells that have high MAOB, low levels of mitochondria, and high Glut10. Examination of GBM patient survival in the three databases as a function of MAOB levels shows that expression of MAOB is associated with poor patient outcome (Fig. 2A, i). Survival curves of tumors with high transcript levels/ratios are the top lines (circles) and low transcript levels/ratios are the bottom lines (triangles) in each graph.

Mitochondrial proxy transcripts for COX15, TFAM, and MTPAP are positively correlated with patient survival, with (Fig. 2A, ii) showing COX15 as an example. Supplementary Fig. S1A and S1B (i) shows similar plots with TFAM and MTPAP mRNA levels. The MAOB:mitochondria (COX15) ratio represents a risk factor for MP-Pt(IV) toxicity and patients with high MAOB:COX15 ratios have very poor survival (Fig. 2A, iii). High MAOB:mitochondria ratios, estimated using proxies TFAM and MTPAP, replicate this result (Supplementary Fig. S1A and S1B, ii).

Like MAOB, SLC2A10 levels are also highly correlated to poor patient outcomes (Fig. 2B, i) and high SLC2A10/COX15 ratios correlate with poor outcome (Fig. 2B, ii).

LDHA expression levels are known to be correlated with the Warburg shift in cancers, including GBM (42), and we observed that higher the Warburg shift, the poorer the GBM patient outcome (Supplementary Fig. S1C, i). A better measure of the Warburg shift is the LDHA:COX15 ratio, which should reflect the relative fluxes of anaerobic to aerobic respiration (Supplementary Fig. S1C, ii). As expected, the greater the Warburg shift, the poorer the patient outcome.

The aggressive GBM phenotype (high MAOB, high SLC2A10, and low mitochondria) is likely to be sensitive to MP-Pt(IV). In contrast, none of the other tissues in the body have this phenotype. The levels of MAOB and mitochondrial proxies in a subset of normal human tissues are presented in Supplementary Fig. S2. Data from the GTEx database (https://gtexportal.org/home/) show that MAOB levels tightly track mitochondrial proxies in normal tissues. As the MAOB:mitochondria ratio is >15 in GBM compared with NHA (17), neither liver nor brain should generate and accumulate P⁺-Pt(IV) better than GBM.

Enzymatic studies of MP-Pt(IV) with both rhMAOA and rhMAOB were examined using two methods, the Amplex Red peroxide assay and UV spectroscopy. MAOB kinetics as a function of [MP-Pt(IV)] are shown in Fig. 3A. The fitted Michaelis–Menten plot indicates that the K_m of MAOB for MP-Pt(IV) is 210 μmol/L and the Vmax was 13% of that observed using saturating benzylamine.

The enzyme catalyzed consumption of the MP-species and appearance of the corresponding dihydropyridinium and pyridinium species were studied using UV spectroscopy (21) to demonstrate the oxidation of MP-Pt(IV) to P⁺-Pt(IV) via the DP-Pt(IV) (Fig. 3B; ref. 43). We observed no activity of MAOA with MP-Pt(IV) in either assay system. To ensure that the rhMAOA we used was catalytically active, we successfully performed validation control studies using the oxidation of tyramine and benzylamine (21) as substrates. These studies prove that MP-Pt(IV) is selectively oxidized by MAOB. Additional control was performed using the MAOB-specific inhibitor, selegiline, which abolished MAOB catalyzed MP-Pt(IV) oxidation.

Experiments conducted to verify the reduction of MP-Pt(IV) by ascorbate into cisplatin are detailed in Supplementary Fig. S3. The products of the reduction of MP-Pt(IV) by ascorbate were examined using mass spectroscopy (Supplementary Fig. S3A). MP-Pt(IV) reduction occurs by a pair of one-electron reduction steps, the first step displacing the axial chloride and forming the Pt(III) species, and the second step releasing the axial MP-ethylene glycol and cisplatin (Supplementary Fig. S3B). Optical studies using 5 mmol/L ascorbate with 100 μmol/L MP-Pt(IV) indicate the reduction of MP-Pt(IV) was pseudo first-order (Supplementary Fig. S3C,i and ii).

The cytotoxic effect of MP-Pt(IV) is affected by the generation of cisplatin within the mitochondrial matrix, and this conversion should be accelerated by increasing the concentration of mitochondrial ascorbate. The rates at which different Pt(IV) compounds are reduced into their Pt(II) daughter drugs correlate with their cytotoxic effect against cancer cells (44), with ascorbate as the main biological reductant (45).

Figure 4 shows that MP-Pt(IV) is toxic toward both GBM115 and GBM157 cells with IC₅₀s of approximately 6 μmol/L and approximately 3 μmol/L, respectively. Dehydroascorbate sensitized cells toward MP-Pt(IV) toxicity, whereas ascorbate provided some protection toward MP-Pt(IV), with significance being found at and beyond 8 and 4 μmol/L, respectively (P < 0.05), compared with cells preincubated with unsupplemented media. Dehydroascorbate is imported into GBM cells via hexose transporters and elevates ascorbate. GBM cells do not typically import ascorbate, as they express low/no levels of the Na⁺/ascorbate cotransporters, SVCT1&2. Ascorbate added to the media of cultured cells typically acts as a prooxidant due to the generation of hydrogen peroxide resulting from ascorbate auto-oxidation (46), and so paradoxically externally added ascorbate can depress intracellular ascorbate levels.

Additional control experiments established that blocking MAOB activity with 10 μmol/L selegiline preincubation completely protected them from incubation with 14 μmol/L MP-Pt(IV). Growth experiments conducted with both GBM cultures incubated with the precursor Pt(IV)-ethylene glycol demonstrated no change in cell numbers.

To investigate the mechanism of MP-Pt(IV) toxicity, we incubated GBM157 cells with or without 4 μmol/L MP-Pt(IV) for 24 hours and examined markers of ROS and mitochondria function. The ROS probes, H₂-DCF-AM and MitoSox, indicate that MP-Pt(IV) causes an elevation in the steady-state levels of cellular peroxide and mitochondrial matrix superoxide, respectively (Fig. 5A and B).

Treatment with MP-Pt(IV) caused a modest increase in cellular peroxide levels, which was not statistically significant in any of the individual groups, but was statistically significant when the data from the three groups were pooled (P < 0.01). MP-Pt(IV) caused a 2-fold elevation of superoxide levels and such elevation typically lowers mitochondrial ΔΨ by inhibiting aconitase (47, 48) and also by activating phospholipase A2 (providing free fatty acids for uncoupling protein; ref. 49). In conditions where ΔΨ is low and ROS is high, mitochondria can undergo the permeability transition, releasing cytochrome c and Smac/DIABLO into the cytosol, thereby initiating the caspase-3 apoptotic cascade. Using a caspase-3–specific substrate, we found that MP-Pt(IV) causes the activation of this apoptotic cascade, and cells preincubated with dehydroascorbate had a much greater increase in caspase-3 activation (Fig. 5C).

In GBM157 cells that survived 4 μmol/L MP-Pt(IV), we found upregulation in mitochondrial numbers per surviving cell using the probe, MitoTracker (Fig. 5D). The degree of upregulation was affected by preincubation, dehydroascorbate > control > ascorbate. MAOB levels were also upregulated, again with dehydroascorbate > control > ascorbate (Fig. 5E). In Supplementary Fig. S4, we show representative images of GBM115 and GBM157 cells treated with increasing concentrations of MP-Pt(IV) labeled with MitoTracker and for MAOB, with nuclei labeled with DAPI. MAOB levels increased following incubation with MP-Pt(IV) and mitochondrial levels responded to this prodrug. Of note was the change in cellular morphology that resulted from treatment and the swollen, distended nuclei. This upregulation of mitochondria and MAOB, in response to MP-Pt(IV) treatment, was also previously seen with MP-MUS (21, 24).

We wanted to determine whether the addition of MP-Pt(IV) improved the efficacy of the Stupp protocol. Figure 6 shows the survival of mice with intracranial primary GBM xenografts treated with MP-Pt(IV) and/or conventional chemoradiotherapy. Forty-eight mice were given xenografts and, on day 8, the 42 heaviest mice were randomized into seven groups of 6 animals each.

Treatments began on day 11, and each animal was given six treatments in all, on days 11, 13, 15, 18, 20, and 22. The groups underwent combination/permutations of three treatments (Monday, Wednesday, and Friday):

• Tail vein injection of 15 mg/kg MP-Pt(IV) (≡8.2 mg/kg cisplatin) in saline or saline.

• Temozolomide 5 mg/kg, given by gavage or oral gavage solution.

• Whole-head radiation (2 Gy from cesium source) or sham radiation.

For comparison, the clinical doses for temozolomide, during chemoradiation, were 150 mg/m² (∼33.8 mg/kg dose in mice) and for cisplatin, 75 mg/m² (∼16.9 mg/kg dose in mice).

The animals were monitored daily for the length of the study and were sacrificed when they reached an ethical endpoint. The study was terminated on day 300, as beyond this point the nude mice began to suffer age-dependent maladies.

The animal weights (Supplementary Fig. S5) indicate that MP-Pt(IV) was well-tolerated and that animals gained weight after treatment.

Figure 6A shows animal survival using the traditional Kaplan–Meier survival curve format, with treatment groups that received sham radiation in the top plot and those that received six cycles of 2 Gy whole-head radiation shown in the bottom plot. Figure 6B shows a different presentation of the survival data, where we plotted the average mouse survival days post-xenograft versus study day. The curves are color coded and are shown in key form in the right panel.

In mice that were not irradiated (Fig. 6A, top), it was clear that temozolomide and MP-Pt(IV) potentiated each other's anticancer effect, but did not cause any drop in animal's weight (a metric of health). Combining temozolomide and MP-Pt(IV) provided immortalization for 10 months and GBM tumors could not be identified in brain tissue at the studies end. When given as a monotherapy, MP-Pt(IV) proved highly effective, with 50% of animals surviving 150 days postimplantation. The combined therapy of MP-Pt(IV) with Stupp was far more effective than Stupp alone (P < 0.0001), with only 1 animal reaching a tumor-related ethical endpoint and another suffering from an ulcer at an injection point. Detailed statistical analysis of the different treatment arms compared with control arm and cumulative mouse survival ratios of different treatment and control arms are shown in Supplementary Table S1.

Lipophilic cations such as triphenylphosphonium, which rapidly accumulate in the mitochondrial matrix, have been explored for mitochondrial-targeting of anticancer drug conjugates (23). Our approach exploits the overexpressed enzyme, MAOB, in the GBM mitochondria for an in situ generation of a lipophilic cation. Both MAOA and MAOB are recognized to mediate and regulate intracellular signaling pathways involved in neuronal growth (50). MAOA has been linked to developing neuronal architecture and synaptic activity, and with maladies like the onset of psychiatric disorders, depression, and aggressive impulsive behavior. In contrast, MAOB has been associated with diseases involving neuronal loss such as Alzheimer and Parkinson diseases. Our drug design strategy exploits the biochemical function of MAOB to enable mitochondrial targeting of a prodrug in GBM.

An ideal anticancer drug is highly toxic toward cancer cells, but is nontoxic to normal cells. In practice, however, most chemotherapeutics only show differential toxicity: highly toxic toward cancer cells, but only slightly less toxic toward the next most sensitive tissue(s). We examined the cancer cell type, especially the most aggressive phenotype, to identify enzyme system(s) that are highly upregulated, compared with the other normal cells, which could then be used to catalyze the maturation of a nontoxic prodrug into a toxic drug. GBMs, like many cancers, have low levels of mitochondria and yet simultaneously show high levels of the mitochondrial enzyme, MAOB, and the mitochondrial dehydroascorbate transporter, Glut10. The prodrug, MP-Pt(IV), was thus designed to exploit high mitochondrial MAOB and Glut10 to generate cisplatin within the mitochondria of GBM.

Within the mitochondria, the Pt(IV) group is activated by ascorbate/GSH reduction to give rise to cisplatin, within the mitochondrial matrix. Compared with our first-generation prodrug, MP-MUS (24), it is 10-fold more efficacious toward GBM in vitro and in vivo and can be used at much higher concentrations (15 compared with 8 mg/kg for MP-MUS) because of higher aqueous solubility (≈8 mg/mL in water compared with 1 mg/mL in 40% Kolliphor in saline for MP-MUS) and stability.

Data mining demonstrated that the most aggressive GBM shows a pronounced Warburg shift, which is linked to elevated MAOB and Glut10 levels, and low levels of mitochondria within these cancer cells. The mitochondrial damage inflicted by MP-Pt(IV) is a function of MAOB and Glut10 levels, both of which aid the generation of cisplatin from P⁺-Pt(IV) within the mitochondrial matrix.

GBM cells are acutely sensitive to MP-Pt(IV), with MP-Pt(IV)–treated cultured cells displaying elevated mitochondrial superoxide/peroxide production and the release cytochrome c, activating caspase-3, and priming the affected cells for apoptosis. GBM cells that are exposed to nonlethal MP-Pt(IV) treatment respond by upregulating mitochondrial levels, which also lead to an increase in the levels of MAOB, and thereby increasing their sensitivity to subsequent treatments with MP-Pt(IV).

MP-Pt(IV) monotherapy is highly effective in intracranial models of GBM, but when used in conjunction with temozolomide, it is curative. Animals treated with MP-Pt(IV) were healthy and gained weight during their treatment course. This potentiation of temozolomide therapy means that MP-Pt(IV) has a high potential for integration with the Stupp protocol. These studies demonstrate that adding MP-Pt(IV) to a chemoradiotherapy protocol could address a critical need in GBM treatment and makes MP-Pt(IV) an excellent candidate for preclinical development.

D.S. Baskin reports a patent submitted for this work (pending). No commercial activities have occurred or are anticipated any time soon. This work needs further development. No potential conflicts of interest were disclosed by the other authors.

S. Raghavan: Conceptualization, data curation, investigation, methodology, writing-original draft, writing-review and editing. D.S. Baskin: Conceptualization, resources, supervision, funding acquisition, methodology, project administration, writing-review and editing. M.A. Sharpe: Conceptualization, data curation, software, formal analysis, supervision, validation, investigation, visualization, methodology, writing-original draft, writing-review and editing.

This work was supported by Donna and Kenneth Peak, The Kenneth R. Peak Foundation, The John S. Dunn Foundation, The Taub Foundation, The Blanche Green Fund of the Pauline Sterne Wolff Memorial Foundation, The Kelly Kicking Cancer Foundation, The Gary and Marlee Schwarz Foundation, and The Methodist Hospital Foundation & The Veralan Foundation. The John S. Dunn Foundation also supports M.A. Sharpe's distinguished professorship. We are grateful to the many patients and their families who have participated in our studies, and who are dedicated to joining us in our fight against brain cancer. We thank Tanvi Kumar and Sophie Lopez for their technical support. The results shown here are in whole or partly based upon data generated by TCGA Research Network (https://www.cancer.gov/tcga. CCGA), and Gravendeel databases were accessed via GlioVis (http://gliovis.bioinfo.cnio.es/).

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.