Karolinska Institute scientists have found a way to attack glioblastoma multiforme-one of the most deadly forms of brain cancer-by inducing an unconventional cell death pathway that triggers catastrophic vacuolization.1 The team identified a small molecule that prolonged survival in a mouse model of glioblastoma without affecting normal brain tissue, but it will likely need to combine the compound with a conventional anticancer agent to translate it for clinical use.

The Karolinska team looked for a nonconventional therapy for glioblastoma multiforme (GBM) because compounds that target tumorigenic pathways have barely made a dent in the disease.

Standard GBM treatment is surgical resection coupled with chemotherapy and radiotherapy, but only 3%-5% of patients survive longer than 3 years before the disease recurs.2

Although genetic studies have revealed many potential targets in the disease, the pathways involved are complex and diverse, and many GBM tumors have multiple mutations.3,4 That has made it difficult to develop agents that can block tumor progression because inhibiting any one pathway has little impact.

Patrik Ernfors and colleagues at Karolinska started with the assumption that GBM cells might accumulate genetic alterations and gain functions not necessarily involved in cancer that could affect cellular processes.

Ernfors is head of office at the Laboratory of Molecular Neurobiology and a professor of tissue biology and molecular and developmental biology at Karolinska.

The team set up a phenotypic screen to find compounds that would affect visible morphological changes in two patient-derived glioblastoma cell lines but not in mouse embryonic stem cells or human fibroblasts.

The initial screen against a 1,364-compound library yielded 63 compounds, which were confirmed as active in 7 other patient-derived glioblastoma cell lines. A panel of cell-based assays-including cytotoxicity, cell viability and apoptosis-combined with in silico ADME analyses reduced the list to 17 hits.

Using zebrafish embryo toxicity and cardiotoxicity screens, the team identified three compounds that lacked toxicity and also decreased tumor size compared with vehicle in a zebrafish model of GBM.

The top hit, containing a quinoline-alcohol scaffold, was named vacquinol-1 and showed an IC50 value of 2.36 mM in cell viability assays. Vacquinol-1 was about 60-fold more potent than the glioma drug temozolomide, which had an IC50 value of 139 mM.

Cancer Research Technology Ltd. and Merck & Co. Inc.'s Temodal temozolomide and Reliance Life Sciences' TemoRel temozolomide are marketed to treat brain cancer.

Vacquinol-1 was ineffective on bladder, prostate, breast and neuroblastoma cancer cell lines, suggesting that the compound might work by targeting cellular vulnerability found in the glioblastoma cell lines.

Next, the researchers looked for vaquinol-1's mechanism of action.

In live cell imaging, vacquinol-1-treated glioblastoma cells showed dose-dependent cell rounding, membrane ruffle formation and rupture of the plasma membrane, which led to necrotic-like cell death. The cell rupture was attributed to the production of large vacuoles induced by massive macropinocytosis.

The team then performed an unbiased shRNA screen with vacquinol-1 in glioblastoma cells and found that those expressing mitogen-activated protein kinase kinase 4 (MAP2K4; MKK4) shRNA were more resistant than cells expressing other shRNAs. The researchers concluded that MKK4 acts as a critical node in vacquinol-1-induced catastrophic vacuolization.

Finally, the team used a mouse model of GBM to evaluate the efficacy of vacquinol-1 on tumors that had been allowed to develop for six or seven weeks.

Despite the advanced stage of cancer, vacquinol-1 administered intracranially or orally abolished tumors or decreased tumor size without affecting normal brain tissue. The mouse glioblastoma cells showed necrosis and accumulation of macropinocytic vacuoles similar to those seen in vitro. By contrast, vehicle-treated mice had large tumors, with brain tissue showing necrosis and hemorrhage.

Vaquinol-1 prolonged survival in GBM mice that received the oral compound once daily for five days starting two weeks after engraftment. In the treated mice, 6 out of 8 survived for at least 80 days, whereas all the vehicle-treated mice died and had a median survival of 31.5 days.

Results were published in Cell.

The team's next steps are to identify a dosing regimen and perform toxicity studies to move into clinical trials.

Opening the therapeutic window

Ernfors told SciBX that he believes vacquinol-1 would be a good addition to the GBM armamentarium because the compound can act on cells with different proliferation rates, whereas temozolomide and radiotherapy act primarily on rapidly dividing cells.

"The compound's macropinocytosis action works independently of cell proliferation and should also be able to target relatively quiescent stem cell-like, tumor-initiating cells or drug-resistant GBM," he said.

However, Paul Dent, a professor of biochemistry and molecular biology at Virginia Commonwealth University, thinks much more data on preclinical toxicology and tissue distribution need to be obtained to support an IND filing.

Dent's laboratory is working on IL-24 (MDA7) as a gene therapy to sensitize malignant glioma to ionizing radiation.

"Ideally, a more efficacious form of the agent needs to be developed. It would then be a logical next step to perform Phase I studies in recurrent GBM patients," he said. He noted that the team has developed an agent with an IC50 value of about 0.39 mM, but they report that unrelated toxicity begins at 15 mM. That therapeutic window might not be enough for translation to the clinic. 

Ernfors' team is looking for better compounds. "Using chemical synthesis and SAR studies, we were able to show that vacquinol-1 is critically dependent on a conserved 4-(piperidin-2-ylmethanol)-quinoline scaffold and produced a series of structural analogs with increased potency," he said.

"We are interested in optimizing the compounds and also pursuing related compounds that modulate other targets within the vacuolization pathway," he added.

Dent wanted to see more details in the mechanisms underlying the MKK4-induced catastrophic vacuolization. "The identification of MKK4 can only be considered a preliminary step in developing an understanding of the mechanism of agent action," he said. "Thus, an unbiased approach examining multiple signaling pathways at the level of activity-phosphorylation and acetylation-will be required to define key target pathways."

Kevin Roth, a professor and chair of pathology at The University of Alabama at Birmingham School of Medicine, thought that inducing GBM macropinocytosis and cell death by targeting MKK4 might be complicated by the fact that MKK4 has been reported to have both pro- and antitumorigenic properties. Other targets of macropinocytosis should be investigated.

Roth's laboratory research focuses on the regulation of neuronal apoptosis and development of autophagy-targeted therapies for brain cancers.

"Macropinocytosis is a topic of great interest recently," said Roth. "Normal cellular processes such as macropinocytosis or autophagy can promote the tumor cells' survival or progression but at the same time make the tumor cells vulnerable to agents that are relatively harmless to normal cells," he said.

Previous studies have shown examples of chemical agents that induce neuronal or cancer cell death through stimulation of macropinocytosis.5-8 The nonapoptotic cell death pathway described in those papers "appears similar if not identical to what is described by the Karolinska team," noted Roth.

"To more fully understand the mechanism of action of vacquinol-1, Ernfors' team might want to interrogate the entire macrocytic pathway, from endosomal maturation through lysosome fusion and degradation," said Roth.

According to Dent, translation to the clinic might require finding an effective combination of this compound with an established anticancer drug.

"Empiric combinations of established drugs with this agent could possibly yield interesting results," Dent told SciBX. "From a translational perspective, it will be important to determine the effect of the agent in combination with standard-of-care drugs in GBM such as BiCNU [1,3-bis(chloroethyl)-1-nitrosourea] and temozolomide, as well as with ionizing radiation."

BiCNU carmustine is marketed by Bristol-Myers Squibb Co. and Emcure Pharmaceuticals Ltd. for brain cancer, Hodgkin's disease, non-Hodgkin's lymphoma (NHL) and multiple myeloma (MM).

"Only by fully understanding the mechanisms of action of vacquinol-1 can progress be made to permit a more rational combination with other clinically relevant agents," said Dent.

A patent application has been filed in the U.S. and Sweden and is available for licensing from Karolinska Institute Innovations AB.

Baas, T. SciBX 7(14); doi:10.1038/scibx.2014.391
Published online April 10, 1014


1.   Kitambi, S.S. et al. Cell; published online March 18, 2014; doi:10.1016/j.cell.2014.02.021
Contact: Patrik Ernfors, Karolinska Institute, Stockholm, Sweden
e-mail: patrik.ernfors@ki.se

2.   Dolecek, T.A. et al. Neuro-oncol. 14, v1-v49 (2012)

3.   Dent, P. et al. Cancer Biol. Ther. 7, 1335-1340 (2008)

4.   Polivka, J. Jr. et al. Anticancer Res. 32, 2935-2946 (2012)

5.   Kaul, A. et al. Cell Signal. 19, 1034-1043 (2007)

6.   Overmeyer, J.H. et al. Mol. Cancer Res. 6, 965-977 (2008)

7.   Robinson, M.W. et al. J. Med. Chem. 55, 1940-1956 (2012)

8.   Trabbic, C.J. et al. ACS Med. Chem. Lett. 5, 73-77 (2014)


Bristol-Myers Squibb Co. (NYSE:BMY), New York, N.Y.

Cancer Research Technology Ltd., London, U.K.

Emcure Pharmaceuticals Ltd., Pune, India

Karolinska Institute, Stockholm, Sweden

Karolinska Institute Innovations AB, Stockholm, Sweden

Merck & Co. Inc. (NYSE:MRK), Whitehouse Station, N.J.

Reliance Life Sciences, Rabale, India

The University of Alabama at Birmingham School of Medicine, Birmingham, Ala.

Virginia Commonwealth University, Richmond, Va.