NMDA receptors are present not only in the CNS but also in the periphery, and there is growing evidence for their involvement in visceral nociceptive processing. For example, the glycine_B selective antagonists MRZ 2/576 (brain permeant) and MRZ 2/596 (brain impermeant (Danysz & Parsons, 1998) were both anti-allodynic in a rat model of Irritable Bowel Syndrome (IBS); the delayed rectal allodynia (10 h, 0.4 ml distension) induced by LPS (1 mg/kg i.p.) was abrogated by MRZ 2/576 and MRZ 2/596 at 0.1 mg/kg by 79.7 and 52.6%, respectively (Sladek et al. 2002). In the current work, any preparative surgery was done under sodium pentobarbitone anaesthesia (60 mg/kg i.p); at the end of the experiments animals were humanely killed, as were all donor animals. Brain penetration was examined in an in vitro model of the blood-brain barrier (BBB) using bovine brain capillary endothelial cells co-cultured with rat cortical astrocytes (Danysz et al. 2005). MRZ 2/596 showed a permeability coefficient similar to that of the BBB-impermeant marker sucrose, 1.11 ± 0.05 (SD) x 10^-5 cm/s and 0.77 ± 0.02 x 10^-5 cm/s, respectively. On the other hand, MRZ 2/576 showed a medium BBB permeability in vitro (pe-coefficient 34.56 ± 4.42 x 10^-5 cm/s). These in vitro data are consistent with brain microdialysis studies in awake rats (5 days after probe implantation), where MRZ 2/596 and MRZ 2/576 at the very high dose of 30 mg/kg i.p. generated maximal concentrations in brain of 0.11 ± 0.05 μM (S.E.M.) (n=3) and 1.34 ± 0.43 μM (n=5), respectively. Central and peripheral NMDA receptor properties have also been compared in patch-clamp experiments. The IC₅₀ rank order of steady-state inward current responses of cultured rat hippocampal neurons (Danysz et al. 2005) to NMDA (200 μM with glycine 10 μM) were memantine (NMDA receptor channel blocker) > MRZ 2/576 > MRZ 2/596 (n = 4-8). In contrast, in cultures of dorsal root ganglia (DRG) from adult rats, NMDA activated currents were antagonized by memantine, MRZ 2/576 and MRZ 2/596 with similar IC₅₀s of 1 μM (n=7-11). Visceral DRG neurons were identified by retrograde labelling with Fast DiI-oil (Molecular Probes) injected into the muscle wall of the small intestine 10-14 days before cell preparation. The NR1 and all NR2 subunits were expressed as shown by Western blotting, immunostaining, and RT-PCR of both DRG cultures and cryosections (cf. Marvizon et al. 2002). However, for all NR1/NR2 subunits examined, no difference in the expression pattern was detected that might explain the different glycine-related pharmacology in DRG cells in comparison to the CNS.