Ion Channel Drugs Suppress Cancer Phenotype in NG108-15 and U87 Cells: Toward Novel Electroceuticals for Glioblastoma

Pantoprazole alone has been reported to increase the alkalinity of the lysosomes by inhibiting V-ATPase channels [86]. However, we saw no alkalization of the lysosomes and the lysosomal signal was increased. Our results agree with a recent study showing that pantoprazole at neutral pH does not inhibit V-ATPase channels and that it instead increases lysosomal biogenesis [120]. Interestingly, that same study showed that pantoprazole interfered with proteasome function, which might explain why it worked so well as a combinatorial treatment. Recent studies have shown that proteasome dysfunction along with an increase in p27Kip1 leads to senescence and endoreplication, resulting in cells with large nuclei [162]. In addition, the loss of PTEN in U87 (ATCC) cells results in cells that preferentially senesce in response to stress [163]. The presence of endoreplication in strongly senescent cells was evidenced by the size of the nuclei in both cell lines and may explain why the BrdU incorporation results were not as low as the cell count data would suggest they should be. The treatment with pantoprazole alone showed less differentiation and senescence when compared to pantoprazole in combination with retigabine, NS1643, TMZ, or rapamycin in U87 cells. Pantoprazole has been shown to be effective against glioblastoma in vitro [83], but to our knowledge, this is the first study that has tested its efficacy in combination with NS1643, lamotrigine, retigabine, rapamycin, or temozolomide.

The leading GBM treatment, TMZ, was not significantly effective at reducing proliferation in NG108-15 cells. This has also been observed for some GBM cases as well. However, our combination of TMZ with pantoprazole in NG108-15 cells did significantly lower their proliferation as compared to pantoprazole alone, but did not terminally differentiate them, as evidenced by the recovery assay. However, we do note that TMZ by itself was not effective in NG108-15 cells but was significantly effective in U87 cells. This effectiveness in U87 cells might explain why the combination of pantoprazole or NS1643 with TMZ was so effective at reducing proliferation and increasing differentiation markers in these cells and not in the NG108-15 line. It should be noted that these combinations also dramatically increased senescence markers in these cells and caused a significant decrease of the YAP nuclear to cytoplasmic ratio.

The drug concentrations used in these studies were chosen to be close to the Cmax reported for the highest dosages that did not result in unacceptable toxicity. Retigabine, FDA-approved for epilepsy, was used at 10 µM, which is close to the reported upper Cmax values obtained with a 1200 mg/day dose of about 2250 ng/mL or about 7.4 µM [164]. Blood-brain barrier penetrance is good for retigabine, with free plasma concentrations being about the same as free brain concentrations [165]. Our choice of using a high serum to test our drugs and drug combinations also helps in strengthening the clinical relevance of our treatments, since retigabine also has a high plasma protein binding affinity of about 80% [165]. We do note, however, that the binding of a compound in fetal bovine serum can be different from the binding in human serum [166].

Rapamycin, an FDA-approved drug for immunosuppression, was used at a dose of 100 nM for our immunological studies. This concentration is under the highest dosage of 40 mg/day of a nano-amorphous oral formulation in a fasted state, with a Cmax of 219 ng/mL or about 239 nM, and toxicity at this level was deemed manageable [167]. Furthermore, rapamycin has been used in a phase 1 clinical trial for glioblastoma and was found to cross the blood-brain barrier effectively [168].

Pantoprazole, an FDA-approved proton pump inhibitor, was used at a dose of 100 µM. The highest dosage of pantoprazole, given for Zollinger-Ellison syndrome is 240 mg/day and results in a Cmax of 42 mg/L or about 110 µM if the Cmax is proportional to that given for a 30 mg/day dose (as it is for dosages up to 80 mg/day) [169]. However, pantoprazole has poor blood-brain barrier penetrance of only 2% [170]. Therefore, its use in glioblastoma therapy will have to rely on novel methods for delivery across the blood-brain barrier, of which there are many new strategies being developed [171]. It should be noted, however, that all tests performed on pantoprazole in this study were conducted at a neutral pH, so pantoprazole’s efficacy may increase in an in vivo setting where the tumor microenvironment is more acidic [172].

NS1643, a hERG activator and potentiator of KCNQ2, KCNQ4, and KCNQ2/3 channels, is not currently approved by the FDA. However, it has recently been used in a breast cancer xenograft model in immunodeficient mice and did not show any overt toxicity on the heart or on normal breast epithelial cells [131].

The electrophysiological analysis in NG108 cells showed what immediate changes the compounds had on membrane potential and did not show the changes in potential over time. Pantoprazole did not have a significant effect on the membrane potential of the NG108-15 cells alone; however, the effect of pantoprazole on the negation of the hyperpolarization seen with retigabine, lamotrigine, NS1643, and rapamycin is extremely interesting considering how quickly it occurred. If pantoprazole was merely blocking the function of those other compounds, then it is curious as to why the combination of the compounds would be more effective than pantoprazole alone. More study into the mechanism of this block is needed. In addition, the lack of hyperpolarization in the NG108-15 cells with the combination treatments points to the fact that the hyperpolarization of the membrane potential is not necessarily needed for ion modulating drugs to have an effect. The depolarization of NG108-15 cells seen in the combination of pantoprazole with retigabine or rapamycin could have effects on the ability of the cell cycle to proceed if the needed level of hyperpolarization to proceed through the S phase is not reached, an effect seen specifically in OPCs [173,174,175]. However, this is not the case for the combination of pantoprazole with NS1643 or lamotrigine, which did not have any significant change in resting membrane potential as compared to control in NG108-15 cells.

A more detailed electrophysiological analysis also needs to be undertaken on the U87 cells. In this study, we were only able to find a DiBAC4(3) based analysis of resting membrane potential and may have had some interference with the dye caused by the compounds themselves. The differences in the U87 cell’s membrane potential as compared to the NG108-15 cells after immediate and long-term treatment should also be investigated.

As for the differentiation analysis, it would be interesting to do a more thorough study where multiple markers are stained for in the same sample to see whether markers for more than one lineage are co-expressed in U87 or NG108-15 cells.

The use of the NG108-15 cell line in high serum conditions for the initial screening of many compounds and combinations gave us a subset of drugs that showed a significant decrease in proliferation in the human U87 glioblastoma cell line treated for six days and showed depressed growth up to four days after treatment was removed. These results suggest that treatment could be given in an intermittent manner. The top combinations were capable of terminally differentiating and causing senescence in NG108-15 cells under normally prohibitive conditions, giving us robust candidates that showed terminal differentiation and senescence in U87 cells. Future studies will test these same compounds on primary human glioblastoma cell lines, patient tumor slices, and in vivo mouse models.