Significance of MDR1 and multiple drug resistance in refractory human epileptic brain

Abstract

Background: The multiple drug resistance protein (MDR1/P-glycoprotein) is overexpressed in glia and blood-brain barrier (BBB) endothelium in drug refractory human epileptic tissue. Since various antiepileptic drugs (AEDs) can act as substrates for MDR1, the enhanced expression/function of this protein may increase their active extrusion from the brain, resulting in decreased responsiveness to AEDs.

Methods: Human drug resistant epileptic brain tissues were collected after surgical resection. Astrocyte cell cultures were established from these tissues, and commercially available normal human astrocytes were used as controls. Uptake of fluorescent doxorubicin and radioactive-labeled Phenytoin was measured in the two cell populations, and the effect of MDR1 blockers was evaluated. Frozen human epileptic brain tissue slices were double immunostained to locate MDR1 in neurons and glia. Other slices were exposed to toxic concentrations of Phenytoin to study cell viability in the presence or absence of a specific MDR1 blocker.

Results: MDR1 was overexpressed in blood vessels, astrocytes and neurons in human epileptic drug-resistant brain. In addition, MDR1-mediated cellular drug extrusion was increased in human 'epileptic' astrocytes compared to 'normal' ones. Concomitantly, cell viability in the presence of cytotoxic compounds was increased.

Conclusions: Overexpression of MDR1 in different cell types in drug-resistant epileptic human brain leads to functional alterations, not all of which are linked to drug pharmacokinetics. In particular, the modulation of glioneuronal MDR1 function in epileptic brain in the presence of toxic concentrations of xenobiotics may constitute a novel cytoprotective mechanism.

Figure 1

Immunohistochemical detection of MDR1 expression in human drug-refractory epileptic brain. Panels A-B show MDR1 expression at the BBB. Widespread GFAP immunoreactivity (green) co-localizes with MDR1 (red). Nuclei are stained blue with DAPI. Note that both parenchymal and perivascular astrocytes express MDR1 (arrowheads), as do endothelial cells of the brain capillaries (thin arrows). The box graph in B illustrates the percentage of GFAP-positive astrocytes (GFAP) in samples from 11 patients (Table 1, 1–11) that also expressed MDR1. Data points (circles), mean value (triangle), range (horizontal bars) and standard errors (° SE) are shown together with the 50^th percentile value. Panels C-D show the neuronal expression of MDR1. Double immunostaining with MDR1 and two neuronal markers (neurofilament, NF and NeuN) reveals that MDR1 is expressed in a subpopulation of epileptic neurons. Arrows point to MDR1 negative neurons, while arrowheads indicate the more frequently-occurring MDR1 positive neurons. Approximately 64% of cortical neurons (n = 264) were positive for MDR1. Quantitative analysis in D was obtained from 11 patients. Patient values (circles), mean (triangle), range (horizontal bars) and standard errors (± SE) are shown. The 50^th percentile is also indicated in the box graph.

Figure 2

Uptake of doxorubicin and phenytoin by astrocytes: effect of MDR1 blockade. Panel A shows 1 μM doxorubicin uptake in astrocytes from epileptic human brain (n = 3 patients) and controls. Note that uptake was decreased in "epileptic" astrocytes compared to normal astrocytes, and this difference was abolished when MDR1 inhibitors (1 μM XR9576 or 50 μM verapamil) were added. Micrographs depict red fluorescent doxorubicin uptake in the various experimental conditions as assessed by fluorescence microscopy. Panel B shows the time-course of intracellular accumulation of ¹⁴C-Phenytoin in "epileptic" astrocytes and controls. Note that ¹⁴C-Phenytoin uptake into epileptic astrocytes reached a plateau after 10 s incubation, while uptake in control astrocytes was greater at each time point and reached a plateau after 10 min incubation (p < 0.01 vs. control; n = 3). ¹⁴C-Phenytoin uptake in "epileptic" astrocytes returned to control level in the presence of 1 μM XR9576 (°p < 0.01 vs. control''; ^p < = 0.01 vs. "epileptic'' astrocytes+XR9576).

Figure 3

Glioneuronal MDR1 expression and resistance to Phenytoin induced cytotoxicity. Histograms in A show data (mean ± SE, n = 3) obtained using cortical slices from human epileptics (Table 1, ID# 9,10,11), or naïve rat cortical slices, treated for 5 h with 375 μM Phenytoin (PHE) ± 2 h preincubation with 3 μM XR9576. Note that PHE toxicity (as illustrated in the micrographs) was apparent in normal rat brain slices, and was greatly exacerbated in human epileptic tissue treated with the MDR1 blocker (p < 0.01). Micrographs depict immunohistochemical evidence of PHE-induced cytotoxicity in GFAP-positive astrocytes and NeuN-positive neurons as assessed by their co-localization with EthD-1, a marker of cell damage. All the cells (astrocytes and neurons) were assessed by DAPI staining. Note that the combination of PHE + XR9576 increased the percentage of injured cells (red) as compared to PHE alone. Panel B shows enlarged nuclei (identified by DAPI) in cells expressing (MDR1 positive) or not expressing (MDR1 negative) MDR1 protein. Note that small condensed nuclei (seen in MDR1 negative cells) reflect apoptosis or irreversible cell damage. In contrast, large nuclei with diffuse DNA staining (seen in MDR1 positive cells) are typical of healthy cells. The graph shows the positive correlation between cells with healthy nuclei and MDR1 expression (n = 34 independent values from slices obtained from 11 patients; p < 0.006).

Figure 4

Model of the proposed role of cell specific MDR1 expression in epileptic brain. Under physiological conditions, when the blood-brain barrier is intact, overexpression of MDR1 (and possibly other drug resistance proteins [5,9]) in endothelial cells causes active net extrusion of drugs from the brain into the vascular lumen (left panel, (1)). A fraction of the AED molecules bypassing the MDR1 barrier (see Fig. 1A) will diffuse into the lipophilic membranes of parenchymal neurons and glia. MDR1 expression in these cells will lead to diminished intracellular sequestration of drugs and may increase their interstitial levels [27]. However, since the total amount of AED in the epileptic tissue is reduced in the first instance by active BBB extrusion, free AED in the extracellular space may still remain below therapeutic concentrations (left panel, (2) and (4)). In addition, MDR1 overexpression in parenchymal astrocytes and neurons affords protection against toxic concentrations of xenobiotics (left panel (2) and (4)). During the transient opening of the BBB due to epileptic activity, AEDs may be back-fluxed into the blood stream by MDR1 expressed at the glial end-feet of perivascular astrocytes, constituting a "second defense barrier" in the brain (left panel, (3)). Right panels A and B schematically summarize these possible roles of MDR1 in epileptic human brain.