Multidrug resistance is often the result of overexpression of membrane glycoproteins known as P-glycoprotein (P-gp) [1]. In humans, the drug transporter P-gp is encoded by the MDR1 gene while two encoding genes Mdr1a and Mdr1b, [2] are present in rodents. P-gp proteins belong to the super family of ATP-binding cassette transporters [3]. MDR proteins recognize a wide range of substrates of diverse chemical structure. This lack of substrate specificity explains the cross-resistance to several chemotherapeutic compounds, the characteristic clinical feature found in a multi-drug resistance phenotype. In addition to their overlapping substrate specificity, each transporter can handle a variety of unique compounds.

Pgp-MDR1 has been identified as an obstacle to both peripheral and CNS chemotherapy [4, 5]. Thus, drug refractory tumors present a clinical challenge which requires novel treatment modalities [6, 7]. In fact, despite recent advances in chemotherapy, the true potential of new agents has not been realised due to the multitude of resistance pathways that impair chemotherapeutic efficacy. Taken together, these findings demonstrate that: 1) MDR is often an insurmountable impediment to cancer chemotherapy; 2) In the case of brain tumors, MDR1 acts by a dual mechanism: a pharmacokinetic one mediated by endothelial cells at the blood-brain barrier and a pharmacodynamic one at the tumor cell plasma membrane where multidrug resistant transporters are widely expressed.

Limitations of current therapies are well known and still far from being addressed. For example, radiotherapy produces its biologic effect on cancerous tissue by ionization. Radiation therapy is carried out by megavolt energy radiation in the form of X-rays from a linear accelerator or gamma rays from a cobalt source. Radiation is highly penetrating, but in order to reach the region of the tumor being treated, the radiation beam must pass through regions containing healthy tissue and may therefore destroy these as well as the malignant tumor. Chemotherapy is limited by drug resistance [1, 7, 8] and by the fact that malignant and non-malignant cells are so similar that it is difficult to destroy cancerous cells without concurrently destroying healthy cells. The adverse effects of chemotherapy are notorious. Increasing the temperature of the tumor to a level at which cancerous cells are destroyed can be used to destroy malignant tumors. One method used for this purpose is to focus a beam of microwave energy, of the type generated in a microwave oven, onto the tumor. The drawback of this technique is that healthy tissues through which the beam must pass to reach the tumor have higher moisture content than the interior of the tumor and are therefore more reactive to microwave energy. Surgical approaches to excise a malignant tumor are limited by location of the tumor, as in the case of tumors in the brain, which are often inoperable. But even where the tumor is accessible, residual malignant cells are left behind, leading to recurrence.

Other approaches include the electrical stimulation in or around the tumor [9, 10], biological response monitors, or specific MDR1 inhibitors such us Tariquidar (XR9576) [11]. In an electrochemical procedure [10, 12], electrodes are implanted in or around the malignant tumor to be treated. The treatment lasts for several hours during one or more sessions and can be used either alone or in conjunction with other therapy such as chemotherapy or radiation therapy. Applied across these electrodes is a low DC voltage of about 6 ± 10 V with currents of 40 ± 100 mA, causing a current to flow between the platinum electrodes through the tumor. The exposure of living tissue to this electric field leads to a wide range of effects such as: 1) Water migrates from the anode to the cathode; 2) The "anodic site" in the tissue becomes strongly acidic and the cathodic site strongly alkaline; 3) Protein denaturation; 4) Bleaching of the local tissue and 5) Cell metabolism and its existing environment are disturbed severely by the electrochemical treatment, causing the destruction of both normal and tumor cells rapidly and completely [10, 12, 13].

More recently [see Additional file 1], very low-intensity (< 15 μA), low frequency AC (50 HZ) applied to dividing cells was shown to inhibit the cell cycle by affecting potassium channels [14] while another study has revealed that at low-intensity, intermediate-frequency (100 – 300 kHz) AC electrical fields may interfere with cytoskeletal mechanisms responsible for the formation of mitotic spindles [15], thus preventing tumor proliferation. In this context however, a cytostatic effect is not sufficient to decrease tumor burden if chemotherapy is ineffective. We now describe a new approach based on delivery of low frequency, low intensity AC stimuli directly to the tumor cells to achieve increased sensitivity to common chemotherapic agents in spite of robust MDR1 expression.

Cells undergoing stimulation protocols were exposed to 50 Hz AC (7.5 μA 32 cycles/pulse, 10 sec interval between pulses) for three days as described in the method section. We have previously shown that this stimulation paradigm is devoid of temperature-dependent effects, is not cytotoxic and specifically does not disrupt cell membrane integrity [14]. After 72 hours of continuous stimulation, cells were exposed for 3 hours to increasing concentrations (1, 2, 4, and 8 μM) of doxorubicin, an MDR1 substrate and potent chemotherapeutic drug to which these cell lines are exquisitely resistant (e.g., HL-60 cells) [18]. Cell viability was assessed by determination of adenylate kinase (AK) release [19] and calcein uptake [20]. Since calcein is a well known MDR1 substrate, this latter experiment was carried out at 4°C in order to inhibit MDR1 activity [17], allow for intracellular accumulation of calcein and assess for cell viability as described in the method section. The results shown in Figure 1a demonstrate that AC stimulation increased doxorubicin uptake (in red) leading to increased cell damage even at low concentrations as demonstrated by AK release measurements (Figure 2a). The sensitivity of this method is improved due to the equilibrium reaction being artificially driven to overproduce ATP by the addition of an excess of ADP. Moreover, unlike traditional ATP assays, AK levels are constant and do not fluctuate with metabolic state, providing a more reproducible, quantifiable signal which closely correlates with cell numbers. These features enable small changes in the cell population to be detected.

The effects of AC stimulation on cell viability were compared to those of the MDR1 inhibitor XR9576 (20 nM) [11, 21] concomitantly applied to doxorubicin. We found that AC stimulation affected cell survival more potently than pre-treatment with the MDR1 inhibitor. This was demonstrated by the analysis of cell cultures exposed to fluorescent doxorubicin followed by calcein uptake at low temperature. The effect of AC stimulation on cell viability is particularly evident at low doses of doxorubicin (1 and 2 μM, Figure 2b), since the differences were less obvious at higher concentrations of doxorubicin. The fact that 20 nM concentrations of XR exhibited an effect on cell viability capable of greatly enhancing the toxicity of doxorubicin suggest two things: 1) The intrinsic resistance to doxorubicin was almost entirely overridden by the specific MDR1 blocker XR9576 thus ruling out a significant contribution of MRP and 2) That the electrical stimulation applied had an effect much greater than that of XR9576 applied at EC50 level. Taken together these considerations imply that had we used much higher concentration of XR9576 we would have achieved the same effect as with electrical stimulation.

Immunohistochemical and fluorescence analysis of cells exposed to doxorubicin following AC stimulation revealed a significant decrease in MDR1 expression paralleled by an increased intracellular uptake of doxorubicin (Figure 3a). Figure 3b shows the relationship between MDR1 expression and the intracellular accumulation of doxorubicin. As expected in control cells the binding of Doxorubicin was nuclear, presumably involving DNA or other nuclear material. These bindings are not easily reversed, and therefore one expects persistence of the Doxorubicin signal even after manipulation of the cell membrane. We have used the same immunocytochemical approach to process cells prior and after stimulation. As can be clearly seen in figure 3a, following AC stimulation and under reduced MDR1 load increased Doxorubicin binding to the nuclear material was observed.

When the cellular distribution of MDR1 was investigated by computerized image analysis and immunohistochemistry, we discovered that AC stimulation not only decreased the expression of MDR1, but also altered its sub-cellular distribution. In non-stimulated cultures, MDR1 was found to be primarily expressed close to the plasma membrane [22], while in stimulated cultures we observed MDR1 expression predominantly in the cytosolic compartment (Figure 4a). The change in MDR1 distribution was indeed confirmed by Western blot analysis performed on membrane and cytosolic protein fractions (Figure 4a lower panel). This shift toward the cytosolic compartment may explain the loss of drug extrusion potency.

While one is tempted to presume that what happened is translocation from the membrane to the cytosol, our current hypothesis is that electrical stimulation achieved the opposite effect, for example impeding release of MDR1 from the cytosolic protein synthesis machinery and thus preventing correct insertion in the membrane. This seems to be so far the more appropriate explanation since the effects of electrical stimulation become evident only after several proliferative cycles (approximately 2 days). Furthermore, as shown in figure 4b the effects of AC stimulation are reversible. Cell cultures that have been exposed to the AC protocol for 3 day and then allowed to rest for 48 h demonstrate a mixed population of cells where the functional MDR1 distribution appears to be partially restored. It is also to be noted that AC stimulation concurrently decreases cell proliferation as demonstrated in our previous work [14]. This explains the apparently long exposure time to AC electric field required for the stimulation to achieve an effect (3 days) and the relatively long time for the cells to recover.

Ideally, a chemotherapeutic approach achieves high levels of target toxicity while sparing normal, healthy cells. In order to evaluate the specificity of AC for drug resistant cells, human brain astrocytes (non-MDR1 over-expressing) and drug-resistant, MDR1 over-expressing astrocytes isolated from drug refractory epileptic patients were exposed to doxorubicin (1 μM) for 3 hours following stimulation or pre-treatment with XR9576 at EC50 (20 nM) and over-saturating levels (1 μM) [21]. No significant differences of drug uptake were detected in normal astrocytes undergoing AC stimulation in comparison to control or XR9576-treated cultures (Figures 5a). However, MDR1 over-expressing astrocytic cultures exposed to the same experimental paradigms showed drug uptake patterns similar to that observed in tumor cell cultures. These data demonstrate that low frequency, low intensity AC stimulation is only effective when drug uptake is limited by MDR1 (Figure 5b).

In summary, we have shown that in addition to a cytostatic effect, AC stimulation also causes the accumulation of MDR1 protein into the cytosol (a non-functional placement) in disfavour of their functional localization on the plasma membrane. In addition, overall MDR1 expression was decreased in stimulated cells. These effects increased the cellular uptake of MDR1 drug substrates and ultimately boosted chemotherapic efficacy. The molecular mechanisms of this effect are yet unknown. There are several possible explanations including the potential involvement of a cytoskeletal mechanism in intracellular protein trafficking. The possibility that electrical stimulation inhibits the correct positioning of MDR1 in the cell membrane is supported by findings by others showing that electrical stimulation can affect microtubular structural assembly and spindle formation [15]. Based on this hypothesis, it is possible that MDR1 remains bound to either ribosomal proteins or other cytoplasmic compartments. Another significant finding of the study is the specificity of AC stimulation for cells expressing MDR1. Electrical stimulation did not increase the sensitivity of normal (non-drug resistant) cells to doxorubicin thus excluding the possibility that AC affects plasma membrane integrity. Furthermore, since several proliferative cycles are required in order for AC to be effective, cells which cell cycle is slow or almost inexistent (e.g. neurons) in comparison to tumors are practically unaffected by the exposure to the electric field. In addition to that, the effect of AC stimulation on the intrinsic drug resistance is reversible as is its anti-proliferative effect.

The main finding presented here is that very low intensity AC current delivered in the proximity of malignant tumors can reduce drug resistance by a mechanism implicating impaired translocation of MDR1 to its functional localization at the plasma membrane. In addition, as previously demonstrated [14] the same stimulation protocol promotes a cytostatic effect reducing tumor cells proliferation. These findings suggest a potential application of low intensity AC current in the treatment of tumor growth by synergistically reducing neoplastic cell division and intrinsic tumor drug resistance. In view of the widespread use of stimulators and stimulating electrodes for the treatment of a variety of other diseases (such as treatment-resistant depression, epilepsy and Parkinson) [23–26], it seems possible that coupling electrical stimulation to current chemotherapy protocols will improve the efficacy of our therapeutic approach to neoplasms. AC stimulation may also decrease the quantity of chemotherapic drugs required to reduce tumor burden. This should reduce chronic exposure to chemotherapic agents as well as the side effects typically associated to chemotherapy. It is also to be noted that both the cytostatic effect and the reduced drug resistance are fully reversible.