Reflections on More Than 30 Years Association with Hanns

Norman G. Bowery1    Department of Pharmacology, University of Birmingham Medical School, Edgbaston, United Kingdom
1 Corresponding author: email address: n.g.bowery@bham.ac.uk

1 Introduction

It is a great honor for me to contribute to this volume in recognition of the work of one of the most, if not the most, prolific researcher in the field of benzodiazepines, Hanns Möhler. He has been foremost in the discovery of the basis for their actions and for facilitating the introduction of different benzodiazepines into clinical medicine and there is no doubt that his contribution has been of paramount importance.

I first met Hanns in Spatind, Norway, where we were attending one of the most influential symposia in the field of amino acid neurotransmission. It was organized by Frode Fonnum under the auspices of NATO. One focus of the presentations was the establishment of binding sites for GABA in mammalian brain tissue as exemplified by Enna, Beaumont, and Yamamura (1978), Lloyd and Dreksler (1978), and Olsen, Greenlee, Van Ness, and Ticku (1978). In addition, there were many “firsts” at this meeting. An example of which was the possibility that endogenous “inhibitors” of GABA receptor binding such as phospholipids (Johnston & Kennedy, 1978) are present in mammalian brain tissue. It was 1977 and nothing was known at that time about the structure of the GABA receptor or the exact distribution of receptor binding sites in the mammalian brain. But Curtis, Duggan, Felix, and Johnston (1970) in Australia had described the first competitive antagonist of the GABA receptor. Recognizing this, Hanns and his colleague, Okada, were able to use radiolabeled bicuculline to demonstrate binding sites on synaptic membranes for the first time at the meeting (Möhler, 1979; Mohler & Okada, 1978).

I had been working on peripheral nervous tissue, namely, the rat superior cervical ganglion, at this time and had found that GABA receptors were present on neurones in this tissue and when activated produced neuronal depolarization (Bowery & Brown, 1974). This appeared to be analogous to the depolarization of primary afferent fibers produced by GABA in mammalian spinal cord (Curtis, 1978). In fact, evidence had shown that this depolarization of nerve terminals (primary afferent depolarization) was responsible for physiological inhibition of dorsal roots. Thus, activation of GABAergic interneurones within the spinal cord can reduce sensory input.

Thus far, the response to GABA in superior cervical ganglia was detected by electrophysiological surface recording from intact isolated tissue. The advent of GABA receptor binding techniques, which were adequately described at this meeting, prompted Hanns and me to consider the possibility of detecting the presence of binding sites in homogenates of ganglia. As a consequence, we arranged to do a series of experiments in his laboratory in Basel. These experiments were performed over a period of about 6 weeks during which we obtained preliminary evidence for the existence of saturable binding sites on this tissue. This culminated in studies conducted by David Hill, in my laboratory in London (Bowery, Hill, & Möhler, 1979), showing the nature of these binding sites in bovine superior cervical ganglia.

While in Basel, Hanns and I discussed in great detail about the possibility of a novel receptor for GABA existing on neurones within the brain. This was prompted by findings that my colleague, Alan Hudson, and I had obtained in isolated atria of the rat (Bowery & Hudson, 1979). There was no evidence at that time for any other GABA receptor with distinct pharmacological properties being present within the brain or elsewhere. Binding sites with different affinities had been described by, for example, Johnston and Kennedy (1978), Olsen et al. (1981), and Guidotti, Gale, Suria, and Toffano (1979), but there was no pharmacological distinction between them. In fact, evidence indicated that any separation might be due to the removal of endogenous inhibitors when neuronal membranes were extensively washed. Washing appeared to serially uncover binding sites with higher affinity for GABA.

Our studies suggested the presence of a novel receptor for GABA in synaptic membranes, the pharmacology of which is quite distinct from the classical chloride-dependent GABA receptor. The observations that led to this discovery emanated from experiments using rat-isolated atria. We hypothesized that if GABA receptor activation on neurones of superior cervical ganglia could produce the same effect on the nerve terminals of these ganglionic neurones, this would produce terminal depolarization analogous to that occurring at primary afferent terminals in the spinal cord (Curtis, 1978).

Of course, we could not examine any depolarization produced by GABA directly but instead decided to study the effect of GABA on the release of noradrenaline from atrial tissue evoked by transmural stimulation. For this purpose, we chose to look at the release of radiolabeled noradrenaline as it had been recently established that 3H-noradrenaline taken up by isolated atria was released from transmitter stores in nerve terminals within the heart tissue in response to nerve stimulation (Iversen, 1974). The results of our experiments showed that, as predicted, GABA reduced the evoked release of 3H-noradrenaline (Bowery et al., 1981; Bowery & Hudson, 1979). This was most evident in the presence of an α1 adrenoceptor antagonist to suppress feedback inhibition of noradrenaline (Bowery et al., 1981; Kalsner, 1973). We initially assumed that this inhibition by GABA was due to nerve terminal depolarization. However, when we began to examine the pharmacology of this effect, we found that the recognized GABA receptor antagonist, bicuculline, would not prevent the action of GABA (Bowery et al., 1981), and, moreover, the GABA analogue, β-chlorophenyl GABA (baclofen) acted as an agonist mimicking the action of GABA. This compound has no effect at bicuculline-sensitive receptors (see Bowery, 1993).

I remember discussing these results in detail with Hanns, and he strongly encouraged me to pursue these findings. We subsequently published our initial findings, which prompted us to discover whether this action of GABA could be detected in brain tissue. It soon became evident that we were looking at a novel receptor for GABA, which was not chloride-dependent and had a distinct pharmacological profile (Bowery, 1993). We wanted to emulate the binding studies for GABA that by now had become firmly established but could not see a way of detecting this novel site in the presence of the established GABA receptor. Using the recognized 3H-GABA binding technique in sodium-free medium and in the presence of bicuculline, no residual saturable binding was observed. The addition of cations such as calcium and nickel had no effect on this binding (Enna & Snyder, 1977). We decided that the only way forward was to obtain some radiolabeled baclofen and to examine for any saturable binding. For this, I returned to Basel not to Hanns’ laboratory at Hoffman LaRoche but to that of Helmut Bittiger at the then CIBA-Geigy laboratories.

Baclofen was first discovered by this group in an attempt to produce a GABA mimetic that might be used as a centrally active sedative/muscle relaxant. It was initially designed as a GABA analog, which, unlike GABA, would cross the blood–brain barrier (Bein, 1972; Keberle & Faigle, 1972). However, there was never any evidence to show that it acted at the chloride-dependent GABA receptor even though it produced muscle relaxation in humans. It was first marketed for the treatment of muscle rigidity in 1972 and remains the drug of choice in such conditions.

Fortunately, the CIBA-Geigy group had produced tritiated baclofen and was kind enough to provide us with a sample. They also provided us with their raw data from experiments in which they had attempted, but failed, to obtain evidence for the presence of binding sites for 3H-baclofen on synaptic membranes. All of their experiments had been conducted in Na+- and Ca2 +-free media as had been employed for 3H-GABA binding. So David Hill and I decided that we would use the same physiological medium that we had used for our release studies in isolated atria and brain slices...