[excerpt]
In 1980, John and I published the first modern review on the enteric nervous system (Furness and Costa 1980). Based on years of actual dissection, I drew a segment of guinea pig intestine with the different layers partially separated.
We started preparing a list of likely functional classes of enteric neurons, based on immunohistochemical data, and the effects of electrical stimulation of enteric pathways by pharmacological agents. In that review, we commented in a foot note, “It is of interest that this appears to be of the same order as the number of nerve cells in the spinal cord”. This perhaps triggered the myth of a separate small ‘brain’ in the gut.
Until the early 1980s, the main way to reveal enteric neural circuits had been based on the distribution of nerve cells and their terminals revealed by histochemical or immunohistochemical methods. Histochemistry utilises different dyes to visualise chemicals in cells, ie via chemical reactions an applied reagent, whereas immunohistochemistry used antibodies to localise specific molecules in the tissue.
What was needed was a way to map enteric neural pathways. The classic method applied in the CNS used the degeneration of nerve fibres when they were disconnected from their cell bodies. To do this in the intestine required selective interruption of enteric pathways via micro-lesions. We succeeded by interrupting surgically in vivo the continuity of the myenteric plexus (myotomy) with micro-scissors or by removing sections of myenteric plexus (myectomy). Using this approach, we could investigate the ‘projections’ of several histochemical classes of enteric neurons for the first time.
Not surprisingly, we applied this strategy first to somatostatin-containing neurons; this was the first neuropeptide studied in the intestine (Costa et al 1980c). This established that somatostatin neurons in the guinea pig small intestine are descending interneurons which form long descending chains with other cells of the same class.
Similarly, we also established the projections of substance-P neurons (Costa et al 1981). Together with equivalent results obtained for the projections of 5HT enteric neurons, we were able to address a crucial question about the nature of synaptic transmission in enteric ganglia.