GABAergic neurones are involved in a vast variety of brain functions, including central control of arterial blood pressure. Visualisation of these cells is very difficult because of the lack of reliable antibodies. Moreover, none of the available stains are suitable for high quality fluorescence imaging of the fine structure of these cells. Visualisation of these neurones in live preparations has been attempted in a transgenic mouse where a short piece of the promoter of the GABA-synthesising enzyme GAD67 was used to drive EGFP expression in a subset of GABAergic cells in the hippocampus (Oliva, Jr. et al. 2000). Since the expression in other parts of the brain has not been well characterised and these animals are not readily accessible, we report here an alternative strategy for visualising living GABA interneurones within the brainstem.

We have generated an adenoviral vector that incorporates 3.7kb of the GAD67 promoter and the first exon-intron sequence to drive the expression of EGFP and tested the expression profile in brainstem and hippocampus. Male Wistar rats (75-150g) were anaesthetised (ketamine, 60 mg kg^-1 and medetomidine, 250 µg/kg, ip) and bilateral injections of adenoviral vectors were made using fine glass capillaries. The rats were allowed to recover for 5-7 days and were then deeply anaesthetised (pentobarbitone 100 mg kg^-1, ip), perfused and brainstem sections processed for GABA using primary anti-GABA polyclonal antibodies and secondary rhodamine-coupled antibodies. In the hippocampus and nucleus tractus soliitarii (NTS) this vector resulted in medium-intensity EGFP expression in numerous neurones (soma size ~ 15-20 Ám). In the hippocampus their location was consistent with the distribution of inter-neuronal layers. The absolute majority of EGFP-positive cells (> 90 % in NTS) were immunopositive for GABA. In addition, we have never observed EGFP expression in glial cells ( < 10 Ám) although some of these cells brightly stained for GABA presumably reflecting GABA uptake from the extracellular space. This vector has also been used to transfect putative GABAergic neurones in organotypic slice cultures of hippocampus and NTS. Fine details of these neurones were visualised using live cell confocal imaging. In conclusion, our novel vector is selective for GABAergic neurones, at least in certain brain areas. This approach will greatly facilitate imaging and electrophysiological studies centred on the role of GABAergic inhibition in physiological and pathological states.

Financial support: Royal Society (23697), BBSRC (7/JE616459), BHF (RG/02/011), WT (AL/069061).