Selective modulation of epileptic tissue by an adenosine A3 receptor-activating drug

Abstract

Background and purpose: Adenosine, through the A₁ receptor (A₁R), is an endogenous anticonvulsant. The development of adenosine receptor agonists as antiseizure medications has been hampered by their cardiac side effects. A moderately A₁R-selective agonist, MRS5474, has been reported to suppress seizures without considerable cardiac action. Hypothesizing that this drug could act through other than A₁R and/or through a disease-specific mechanism, we assessed the effect of MRS5474 on the hippocampus.

Experimental approach: Excitatory synaptic currents, field potentials, spontaneous activity, [³H]GABA uptake and GABAergic currents were recorded from rodent or human hippocampal tissue. Alterations in adenosine A₃ receptor (A₃R) density in human tissue were assessed by Western blot.

Key results: MRS5474 (50-500 nM) was devoid of effect upon rodent excitatory synaptic signals in hippocampal slices, except when hyperexcitability was previously induced in vivo or ex vivo. MRS5474 inhibited GABA transporter type 1 (GAT-1)-mediated γ-aminobutyric acid (GABA) uptake, an action not blocked by an A₁R antagonist but blocked by an A₃R antagonist and mimicked by an A₃R agonist. A₃R was overexpressed in human hippocampal tissue samples from patients with epilepsy that had focal resection from surgery. MRS5474 induced a concentration-dependent potentiation of GABA-evoked currents in oocytes micro-transplanted with human hippocampal membranes prepared from epileptic hippocampal tissue but not from non-epileptic tissue, an action blocked by an A₃R antagonist.

Conclusion and implications: We identified a drug that activates A₃R and has selective actions on epileptic hippocampal tissue. This underscores A₃R as a promising target for the development of antiseizure medications.

Keywords: GABAergic transmission; adenosine A1 receptor; adenosine A3 receptor; epilepsy; hippocampus; neuronal excitability.

FIGURE 1

MRS5474 in a virus antibody-free rodent facility, receiving food (autoclaved diet pellets)and water (sterile water treated by reverse osmosis) RS5474 did not affect excitatory postsynaptic currents (EPSCs) in CA1 pyramidal cells in rat hippocampal slices, in clear contrast with the A₁ receptor agonist, CCPA. Time-course of changes of EPSC peak amplitude are shown in (a) n = 6 cells from five rats, (b) n = 5 cells from four rats (QX-314 in the intracellular solution to inhibit sodium channels and prevent action potential spiking independently of input strength) and (c) n = 5 cells from four rats. EPSC peak amplitude was normalized in each experiment, taking as 100% baseline the values recorded for 10 min before drug application (1); drug effects were assessed by comparing baseline values with values recorded at 30–40 min after drug application (2). Values are as mean ± SEM. The horizontal lines below drug names indicate drug presence in the perfusion solution. Insets in each panel: representative averaged superimposed EPSC traces recorded in the same cell before (1) and by the end (2) of drug application. In (a) at right is shown the paired EPSC amplitude (pA) of all cells included in the data shown in (a) at left before MRS5474 (baseline) and in the last 10 min of MRS5474 perfusion; ns: P > 0.05 (two-tailed paired t test). In all experiments, the fast component of inhibitory GABAergic transmission was blocked by adding picrotoxin (50 μM) to the aCSF. Note that the selective A₁ receptor agonist, CCPA (c), caused a marked inhibition of EPSCs in clear contrast to the absence of effect of MRS5474 (a, b).

FIGURE 2

MRS5474 did not affect field excitatory postsynaptic potentials (fEPSPs) in mice hippocampal slices, in clear contrast to the A₁ receptor agonist, CPA. Recordings were performed in the CA1 area of acute hippocampal slices upon stimulation of the Schaffer collaterals. In (a) and (b), normalized averaged time-course changes in fEPSP slopes (%, mean ± SEM) taking as baseline (100%) the values recorded for 10 min before drug application, are shown. The horizontal lines below drug names indicate drug presence in the perfusion solution, and the arrows indicate the time of starting perfusion of each drug concentration. Insets: representative superimposed averaged fEPSP traces recorded from the same slice before (control) and by the end of (a) MRS5474 (250 nM) or (b) CPA (30 nM, an A₁ receptor agonist) application. (a) n = 7 slices from six mice. (b) n = 4 slices from four mice. As n=4 for these experiments, statistical analysis was not carried out, and results should be regarded as preliminary, though fully consistent with what is known from the literature (e.g. Sebastião et al., 1990). Note also that the fEPSP inhibition caused by the selective A₁ receptor agonist, CPA (b)clearly contrasts with the absence of effect of MRS5474 (a).

FIGURE 3

MRS5474 inhibited spontaneous activity in rat organotypic rhinal–hippocampal slices under a depolarizing condition. In (a) and (b) are shown representative field potential recordings of spontaneous activity from the CA3 area of a control slice (no drug [a]) and of a slice to which MRS5474 (250 nM) was added to the perfusion solution as indicated by the arrow (b). The upper panel in (a) and (b) shows representative recordings during the entire experimental time, while the lower panel shows representative ictal-like discharges in basal conditions in the same slices as in the corresponding upper panel; note that the timescale is different in both the upper and lower panels. In (c) and (d) are shown the mean frequency of spikes within the bursts at the times indicated below each bar; data obtained at 10–20 min were in the absence of MRS5474 (basal ictal discharge) and were normalized to 1, whereas data at 50–60 or 80–90 min were either in the absence (c) or in the presence (d) of MRS5474 (250 nM). Data are shown as mean with the dots representing individual data points. Dots from the same experiment are connected by a straight line. Dots corresponding to the experiments represented in (a) and (b) are highlighted in blue. In all experiments, the concentration of KCl in the perfusion solution was increased to 8.5 mM (depolarizing conditions). *P < 0.05. Friedman test followed by Dunnett’s multiple comparisons test.

FIGURE 4

MRS5474 decreased fEPSPs in hippocampal slices taken from rats with established epilepsy (EE). Recordings were performed from the CA1 area of acute hippocampal slices upon stimulation of the Schaffer collaterals. In (a) recordings were from slices taken from a rat with EE, and in (b) from a control rat. In the left panels of both (a) and (b) are shown representative time-course changes in fEPSP slopes taking as baseline (100%) the values recorded for 10 min before drug application. The horizontal lines below drug names indicate drug presence in the perfusion solution. Insets: representative superimposed averaged fEPSPs traces recorded from the same slices as in the panel where they are inserted and at the times indicated by the numbers. In the right panels of both (a) and (b) are shown paired EPSP slope values (mV·ms⁻¹) from all experiments included in the corresponding left panel, in baseline (Period 1 indicated on the left panel) and under MRS5474 (Period 2 indicated on the left). Dots corresponding to the experiments represented in (a) and (b) are highlighted in red. *P < 0.05; ns: P > 0.05, two-tailed paired t test.

FIGURE 5

MRS5474 inhibited GAT-1-mediated GABA uptake, an action mimicked by an adenosine A₃ receptor agonist and antagonized by an A₃ receptor antagonist but not by an A₁ receptor antagonist. GABA uptake was performed in acute hippocampal slices from rats. In both (a) and (b) are shown the absolute values (fmol·mg⁻¹ protein) of GAT-1-mediated GABA uptake in control conditions (no drug, ●) or in the presence of MRS5474 (50 nM) (a, ▲) or MRS5698 (100 nM) (A₃R agonist, b, ▼). Data from the same experiment are connected by a straight line. In (c) are shown the pooled data (fmol·mg⁻¹ protein) in (from left to right) control conditions, in the presence of MRS5474 (50 nM), and in the presence of the A₃ receptor agonist MRS5698 (100 nM). In (d) are shown the pooled data as % change in GAT-1-mediated GABA uptake caused by (from left to right) MRS5474 (50 nM), MRS5474 (50 nM) in the presence of MRS1523 (10 μM, an A₃ receptor antagonist), MRS5474 (50 nM) in the presence of DPCPX (50 nM, an A₁ receptor antagonist) and MRS5698 (100 nM, an A₃ receptor agonist). In (e) and (f) are shown the pooled data in control conditions (no drug, left column in each panel) or in the presence of MRS1523 (10 μM, A₃ receptor antagonist) (e, right column) or DPCPX (50 nM, A₁ receptor antagonist) (f, right column). In (c–f), data are shown as mean ± SEM with the dots representing individual data points in each experiment; *P < 0.05; ns: P > 0.05, two-tailed paired t test (a, b), one-way ANOVA followed by multiple comparisons tests (c: Tukey’s test, to compare each column with all other columns; d: Dunnett’s test, to compare the first column with all other columns) or two-tailed unpaired Student’s t test (e, f).

FIGURE 6

A₃ Receptor immunoreactivity was enhanced in hippocampal samples from patients with refractory epilepsy. A representative Western blot of A₃ receptor immunoreactivity in epileptic resected tissue from patients with refractory epilepsy and in post-mortem tissue from control patients (CTRL) is shown in (a). GAPDH was used as a control. (b) Pooled data from densitometric analysis of the 35 kDa band (A₃ receptor, as indicated by the antibody supplier) corrected for GAPDH (37 kDa) in the same samples. Values were normalized taking as 1 the values obtained in control samples. CTRL and epileptic samples were always analysed on the same membrane. Data in (b) are shown as mean ± SEM with the dots representing individual data points (sample details in Table 2); each data point corresponds to a sample from a different patient. *P < 0.05, two-tailed unpaired t test with Welch’s correction for unequal variances.

FIGURE 7

MRS5474, through A₃ receptor activation, selectively enhances GABAergic currents (I-GABA) in epileptic human hippocampal samples. (a) Representative currents evoked by GABA (500 μM, applied as indicated by the white horizontal bars) recorded before (black trace) and after (red trace) incubation with MRS5474 (5 μM, for 2 h and 30 min, first and second rows) or with MRS5474 under similar conditions but in the presence of MRS1523 (10 μM), an A₃ receptor antagonist (third row). Recordings were from oocytes injected with membranes prepared from epileptic patient tissue samples (TLE, first and third rows) or control (CTRL, second row); each row shows sample recordings from the same oocyte before and by the end of incubation with MRS5474. (b) Concentration–response curve obtained in a non-cumulative manner; thus, consecutive concentrations were tested in different oocytes. The ordinates represent the percentage change of the amplitude of GABA-evoked responses after incubation with different concentrations of MRS5474, as indicated in the abscissae; 100% represents the amplitude of currents recorded before exposure to MRS5474 in the same oocytes. The raw amplitude values before and after incubation for each concentration are the following: [0.05 μM] from 43 ± 10 nA before incubation to 47 ± 14 nA after incubation, n = 8, #1 and #2; [0.5 μM] from 68 ± 10 nA before incubation to 79 ± 16 nA after incubation, n = 8, #2 and #3; [1.5 μM] from 41 ± 4.6 nA before incubation to 53 ± 6.0 nA after incubation, n = 8, #1 and #2; [5 μM] from 47 ± 8.9 nA before incubation to 71 ± 17 nA after incubation, n = 15, #1–3; and [10 μM] from 44 ± 6.5 nA before incubation to 62 ± 6.3 nA after incubation, n = 8, #1–3. (c) Comparison (in another set of experiments) between the effect (as % change in the ordinates) in current amplitude caused by a submaximal concentration of MRS5474 (1.5 μM) in the absence (n = 8 oocytes) and in the presence (n = 11) of the A₃ receptor antagonist, MRS1523 (10 μM). (d) Comparison between the effects of a near maximal concentration of MRS5474 (5 μM) in oocytes injected with membrane samples from epileptic patients (same data as in [b], n = 15) and from oocytes injected with membranes from controls (n = 10 oocytes from different frogs, #4). In (b–d), values are expressed as mean ± SEM. The statistical analysis shown in (b) represents the comparison between current amplitude before and after MRS5474 (two-tailed paired Student’s t test) and in (c) and (d) represents the comparison between the pooled data in the two conditions shown in each panel (two-tailed unpaired Student’s t test). For all panels: *P < 0.05.