Na/K-ATPase as a target for anticancer drugs: studies with perillyl alcohol

Perillyl alcohol (POH) has the ability to induce apoptosis in cancer cells 37]-40] and has been used in the therapy of different tumors, including gliomas 31],32],41],42]. These effects of POH may be at least partly related to the inhibition of NKA activity
28],43].

As a first step in studying the ability of POH to inhibit glioma cell proliferation,
we evaluated the NKA activity based on the incorporation of Rb⁺ by tumor and non-tumor cells. POH inhibited the NKA activity in a dose-dependent
manner. The IC₅₀ values of U251 and U87 human GBM cell lines were similar and were very close to the
value previously found by our group using the A172 human GBM cell line 28]. In these three human GBM cell lines (U251, U87 and A172), POH completely inhibited
Rb⁺ uptake at 4 mM. In the non-tumor VERO cell line, the IC₅₀ was slightly higher and POH did not reach 100% inhibition at 4 mM. These IC₅₀ values were not significantly different, which was expected because kidney and GBM
cells express the same ?1 NKA isoform 23],44], against which POH has a higher degree of selectivity 28]. Due to the unavailability of normal human astrocyte cultures, we used mouse astrocytes.
In these cells, 2 mM POH inhibited completely the Rb⁺ uptake. It is interesting to note that mouse astrocytes mainly express the NKA ?2
and ?3 isoforms, which are more sensitive to the inhibitory effects of POH 28],44]. Furthermore, perillic acid (PA), a metabolite produced rapidly in the human body
after POH administration, was also tested. NKA activity was not affected by PA in
any of the cell lines used in this work. To assess whether the inhibition of Rb⁺ uptake by cells (U251, U87, VERO and mouse astrocytes) were caused by cell death,
cell viability experiments were carried out in the presence of POH, PA and ouabain
(OUA). The cell viability was significantly decreased by POH only when a high concentration
was used. OUA and PA did not affect the cell viability. Therefore, the decreased Rb⁺ uptake occurred through NKA inhibition and not due to cell death.

Low plasma levels of POH cannot be measured accurately, but the levels of their metabolites
are detectable due to the rapid degradation of POH 45]. Thus, POH administration is more advantageous than the use of cardiac glycosides
because its rapid metabolism decreases its undesirable adverse effects. Interestingly,
therapeutic doses of POH are far superior to those of cardiac glycosides 46]. The cytotoxic properties of compounds that may or may not be similar to cardiac
glycosides but can still affect NKA activity have already been tested 47]-52]. Though PA did not affect the cell viability of the studied cells, POH had a potential
cytotoxic effect on all cell lines. The cell viabilities of both tumor and non-tumor
cells were significantly reduced at lower dose and increased thereafter according
to the POH dose. The IC₅₀ value for POH was not significantly different between cell types (U251 and U87).
Similar results were obtained by Cho et al. (2012) 53] using glioma cell lines U87, U251 and LN229.

Due to its role in fundamental cellular functions such as proliferation, differentiation
and apoptosis, NKA has been considered as a target for drugs, especially those with
antitumor activities 17],25],52]. Several studies have reported the induction of apoptosis by cardiac glycosides in
various tumor cells 14],24],46],54]-56]. At this point, it is important to emphasize that POH induces apoptosis in various
tumor cells, including gliomas. However, the exact mechanism by which this drug induces
apoptosis is unclear 38],57]. It is known that POH preferentially inhibits the NKA ?1 isoform 28], which modulates apoptosis, cell migration and proliferation and is overexpressed
in the caveolae of GBM cells 26],27]. Considering these facts, we tested the activation of JNK, a main protein target
of the MAPK pathway controlling cell growth and/or death 15].

The human GBM cells (U87 and U251) and non-tumor cells (VERO cells and mouse astrocytes)
were treated with POH. POH increased JNK1/2 phosphorylation in the U87 and U251 cell
lines that was similar to mouse astrocytes but not VERO cells. Studies with fibroblasts
have shown that JNK activation and the pro-apoptotic protein Bax, a Bcl2 family member
58], are sufficient to cause the caspase-independent release of cytochrome c, which leads to apoptosis. Additionally, POH increases the Bax expression in non-small
cell lung cancers 39],40]. Although JNK phosphorylation by POH has not yet been described, Satomi et al. (1999) 59] showed that POH induces the increased expression and phosphorylation of c-Jun protein
in breast cancer cells, which is involved in cellular proliferation and apoptosis.
This phosphorylation occurred quickly at the N-terminal site of c-Jun, which is normally
phosphorylated by JNK. According to these researchers, c-Jun activation by POH in
these cells may represent a relevant early response to apoptosis. Thus, POH seems
to modulate the JNK signaling cascade 59]. In addition, the activation of p38, another protein of the MAPK family, was observed
in U87 cells, broadening the framework of intracellular signaling proteins involved
in POH-induced cell death. With respect to JNK and p38 activation in GBM cells, it
was recently reported that piperlongumine, an alkaloid with lipophilic properties
found in plants of the species Piper longum L., can activate JNK and p-38 MAPKs, leading to apoptosis of U87, LN229 and 8MG-BA
GBM cell lines through the accumulation of reactive oxygen species 60].

To correlate the POH induced-JNK phosphorylation with NKA, we evaluated JNK activation
in the presence of the Src kinase inhibitor dasatinib 61] and methyl ?-cyclodextrin, a molecule that extracts cholesterol from the plasma membrane,
disrupting lipid rafts 62] and thus blocking the MAPK pathway in the signalosome, which is a caveolae microdomain
in which the formation of the NKA-Src complex occurs. NKA mediated-cellular signaling
is initialized in the cholesterol and sphingomyelin-rich plasma membrane subfractions
of the caveolae, in which the NKA ?1 subunit interacts with various signaling proteins.
Src is a primary target of the NKA ?1 subunit and is responsible for communication
between NKA and other proteins. When the NKA-Src complex is activated, different signaling
pathways, including the MAPK pathways, are initiated in a specific manner depending
on the stimulus and cell type 7],63]. In U87 cells, the effect of POH on JNK1/2 activation was significantly reduced by
dasatinib and methyl ?-cyclodextrin pretreatment. Both pretreatments appeared to prevent
NKA activation in the signalosome. These results support our hypothesis that NKA is
directly involved in mechanisms involving POH-mediated JNK activation in human GBM
cells, triggering cell death through the activation of NKA in the signalosome.

Activated MAPKs are able to mediate the release of interleukin (IL) 64] and can be induced by cardiac glycosides through NKA in cytotrophoblast embryonic
cells 15]. However, POH suppressed the level of pro-inflammatory cytokines in ischemia-reperfusion
injury in rat brains 65]. In GBM, the presence of pro-inflammatory cytokines is associated with tumor growth
and hence with its malignancy. IL-6 induces the proliferation of tumor cells proliferation,
whereas IL-8 has angiogenic and chemotactic properties 66]. During the three incubation periods used (1, 6 and 24 hours), POH did not alter
the release of pro-inflammatory cytokines in either of the two human GBM cell lines.
On the other hand, an increase in the release of IL-8 after 24 hours of incubation
with POH was detected in the U251 cell line. Cho et al. (2012) 53] found a decrease in the release of IL-8 in U87 cells after 48 hours of incubation
using 0.6 mM POH. Since IL-8 enhances GBM invasion 67],68], this increase may be a strategy used by the cells in circumventing the POH-induced
effects on cell death.

POH showed a significant increase in early and late apoptosis or necrosis of U87 and
U251 cells, but when these cells were pretreated with a JNK inhibitor, a significant
decrease in cell death was found. Our results in the U251 and U87 human GBM cell lines
showed that the inhibition of JNK activation decreased the effects caused by POH,
demonstrating the possible involvement of JNK in the induction of apoptosis in GBM
cells. However, the reduction in the POH-induced cell death did not occur to the same
extent in both cell lines. Interestingly, in addition to this difference in the rates
of blocked cell death between these tumor cell lines, the effect of POH on the activation
of JNK was also lower in the U251 cells, as previously shown. Since the same intensity
of cell death was induced by POH in both GBM cell lines, part of this pathway might
be JNK-independent in U251 cells. It is possible that the activation of other MAPKs
or the involvement of the PI3K/Akt pathway are also linked to the NKA-Src complex
62],68].

Human GBM cells are driven to undergo apoptosis when treated with POH 38],57]. As shown previously, JNK activation appears to be involved in this phenomenon. To
confirm the involvement of apoptosis in POH-mediated cell death, U87 and U251 cells
were treated with POH or POH plus JNK inhibitor V and then were immunostained for
cleaved caspase-3. As expected, POH induced apoptosis in GBM cells, though interestingly,
JNK inhibition decreased the POH-induced apoptosis. NKA, especially the ?1 isoform,
plays an important role in GBM survival 26],27]. Inhibitors of this enzyme, particularly the cardiac glycoside UNBS1450 and the monoterpene
POH, cause death in these tumor cells by autophagy and apoptosis induction, respectively
26],38]. However, the mechanisms triggering apoptosis were not completely elucidated 38].