Serotonin and sensory signalling from the gastrointestinal lumen

It is interesting to look through the citation records of classic papers such as this – Bülbring & Lin (1958). It serves as a who's who in serotonin research and catalogues the development of our understanding of the role that 5-HT plays in sensory signalling from the gut lumen.

Bülbring had already shown in early publications that the peristaltic reflex was entirely mediated by neural mechanisms within the bowel wall since degenerative section of the extrinsic innervation had no effect on reflex activity (Bülbring et al. 1958). They had also demonstrated that the peristaltic reflex could be triggered by distension, but required a sensory mechanism within the mucosa since it was lost after removal of the mucous membrane or following topical application of local anaesthetic. It was also evident that there was a rich source of 5-HT in the gastrointestinal tract and that this was mainly within the mucosal epithelium (Feldberg & Toh, 1953). However, there was also something of a dilemma. Carcinoid syndrome, in which there is massive outpouring of 5-HT from mucosal enterochromaffin cells, is characterized by increased intestinal activity and diarrhoea. In contrast, when 5-HT was applied to isolated intestinal segments in an organ bath it inhibited or abolished the peristaltic reflex (Kosterlitz & Robinson, 1957). Bülbring and Lin therefore set out to ask the beautifully simple question – what happens if 5-HT is applied not to the serosa but to the mucosa? Their hypothesis was that this would mimic release of endogenous 5-HT which in turn would activate receptors associated with mucosal sensory mechanisms. This classic paper describes the methodological developments that were necessary to record the propulsion of luminal contents by pressure-evoked peristalis in the guinea-pig ileum (and rabbit jejunum) during application of 5-HT to the lumen, and gives a detailed account of experiments designed to unravel the role of 5-HT in sensory signalling.

First, a word or two about the available methodology. Reading again about kymographs and ‘home-made’ piston and float recorders to monitor pressure and volume contrasts with modern descriptions of solid-state electronics, digital devices and computer analysis. Limited pharmacological tools reflect the lack of appreciation at the time of the range of 5-HT receptor subtypes that are now known to be expressed in the gut mucosa and which influence sensory signalling. Assay for endogenous 5-HT was based on strips of rat stomach, uterus and colon rather than HPLC or amperometry. The paper is also very descriptive so that the reader can get a real feel for the whole study design, problems that were encountered and overcome, variability in responses and there's not a P value in sight. Yet, with cleverly designed protocols and enormous insight Bülbring and Lin established a concept for control that still holds 50 years later.

Their major findings can be summarized as follows. In the guinea-pig ileum peristalsis was triggered initially by a pressure rise of just 1–1.5 cmH₂O, rising to about 2 cmH₂O after an hour or so. Adding 5-HT to the lumen dose-dependently decreased threshold with a threshold concentration around 1 nm. Application of 5-HT to the bath invariably caused inhibition of peristalsis, even when this was already stimulated by intraluminal 5-HT. Removing the mucosa from the loop of intestine or luminal application of cocaine or procaine abolished the peristaltic reflex and under these conditions luminal 5-HT had no effect.

LSD and 2-bromo-d-LSD were used as 5-HT antagonists. Applied to the lumen they blocked the effect of co-administered 5-HT. However, when applied alone the antagonist raised the peristaltic threshold suggesting that it prevented the action of locally released 5-HT. However, peristalsis was not abolished indicating that 5-HT release is not an essential prerequisite for activation of the peristaltic reflex but is necessary to set the threshold.

5-HT overflow into the lumen increased in response to distension, an observation consistent with pressure-evoked release contributing to peristalsis. The increase was proportional to the degree of distension but declined rapidly over time. Again peristalsis was still evident even when 5-HT levels had fallen to very low levels and is therefore not essential for reflex activation to occur.

The amount of 5-HT released and the decline over time could be pharma-cologically manipulated. Inhibiting 5-HT breakdown with a luminal agent to inhibit amine oxidase (1 isonicotinyl-2-isopropylhydrazine) slowed the decline in 5-HT output over time. Addition of the 5-HT precursor, 5-HTP, increased the release of 5-HT and augmented peristalsis. Procaine, which abolished peristalsis, did not prevent 5-HT release in response to distension. Bath applied hexamethonium also abolished the peristaltic reflex but, again, 5-HT release into the lumen was preserved. In contrast luminal acetylcholine (at low concentrations) stimulated peristalsis.

Finally, with current interest in the role of 5-HT receptor subtypes in mucosal signalling, it is important to see that phenyldiguanide, which is now known to be a 5-HT3 receptor agonist, also stimulated peristalis when applied to the lumen.

These observations are interesting in light of subsequent studies that place Enterochromaffin cells in a pivotal role in sensory signalling both for enteric reflexes and in extrinsic afferent activation. Many studies since have demonstrated that 5-HT receptor blockade can reduce or prevent reflex activation by mechanical stimulation of the mucosa or following application of chemicals to the mucosal epithelium, particularly bacterial toxins (Gershon, 1999). 5-HT receptors are present on the terminals of both enteric and extrinsic sensory neurones and convey information related to the luminal mechanical and chemical environment to enteric reflex circuits and to reflex centres in the brainstem (Hillsley & Grundy, 1998; Hillsley et al. 1998; Bertrand et al. 2000). Such observations have enormous clinical relevance. The treatment of nausea and vomiting triggered by cancer chemo- and radiotherapy has been revolutionized by the discovery that the underlying mechanism involves EC cell activation and 5-HT acting on vagal mucosal afferents (Costall & Naylor, 2004). 5-HT ligands have also been shown to have clinical efficacy in treating patients with constipation predominant and diarrhoea predominant irritable bowel syndrome (IBS) (Camilleri et al. 2002). There is growing evidence that bioavailability of 5-HT is altered in some patients with IBS, particularly those whose symptom onset is related to an acute inflammatory insult. EC cell numbers are increased and the transport mechanisms that take up 5-HT following release are reported to be blunted although this is controversial (Dunlop et al. 2005). Bülbring and Lin's observations on the consequence of altering 5-HT metabolism are particularly relevant in this context. Current interest in altered 5-HT bioavailability in visceral hypersensitivity following changes in the serotonin transporter can also be traced back to their observations on the effects of increasing precursor availability and decreasing breakdown.

It is also now clear that there is an elaborate arrangement of receptors and ion channels on EC cells that regulate 5-HT release. 5-HT release is modulated by a variety of neuromodulators acting on adrenoreceptors, nicotinic and muscarinic cholinoreceptors and 5-HT3 receptors (Racke et al. 1996). Bülbring and Lin's observations with luminal acetylcholine and phenyldiguanide, which mimicked the effect of luminal 5-HT, may have arisen because of a common action on endogenous 5-HT release. There therefore exists a bidirectional communication between EC cells and sensory mechanisms on the one hand and enteric reflexes influencing EC cell function on the other. The work coming from Edith Bülbring's laboratory in the late 1950s was the foundation for this modern day concept.