In vitro electrical activity of the equine pelvic flexure




Reasons for performing study: The generation and maintenance of intestinal motility patterns involve the complex interactions of several components including the gastrointestinal pacemaker cells (interstitial cells of Cajal, ICC). Central to ICC function is the generation of rhythmic pacemaker currents, namely slow waves, which represent the rate limiting step for intestinal smooth muscle contractions. Currently, intracellular slow wave activity has not been demonstrated in the equine colon.

Objectives: To characterise the in vitro myoelectrical activity of the equine pelvic flexure using intracellular recording techniques.

Methods: Intestinal samples were collected immediately following euthanasia from 14 normal horses. One millimetre thick tissue sections were pinned and superfused with warmed, oxygenated Krebs solution. Intracellular recordings were made from smooth muscle cells close to the submucosal border of the circular muscle layer. The L-type Ca2+ channel blocker nifedipine was added to the superfusion fluid in 9 experiments while the Na+ channel blocker tetrodotoxin was added to the superfusion fluid in 4 experiments. The data were recorded and stored using an acquisition system and a software package used to analyse the recordings.

Results: In 10 of the 14 horses, electrical events consistent with slow wave patterns were recorded from individual smooth muscle cells. Surprisingly, adding nifedipine to the superfusion fluid abolished all electrical activity. In contrast, tetrodotoxin had no apparent effect on the electrical activity.

Conclusions: Assuming that the electrical events were slow waves, the blockade by nifedipine suggests significant and potentially important differences in the ionic mechanisms responsible for slow waves in the different regions of the equine intestinal tract, which deserve further evaluation.


The generation and maintenance of intestinal motility patterns is a complex process involving the close interaction of several components including smooth muscle cells, neurons, pacemaker cells (interstitial cells of Cajal, ICC) as well as local chemical factors (Huizinga and Lammers 2008). Recently, the ICC have received much attention and, although there is some disagreement as to some of their proposed functions, there is general agreement that they generate rhythmic membrane depolarisations known as slow waves (Ward et al. 1994; Huizinga et al. 1995; Torihashi et al. 1995). These typically consist of a rapid upstroke depolarisation, partial repolarisation and then a sustained, ‘plateau’ potential that can last several seconds (Szurszewski 1987).

The ICC currents are conducted to the smooth muscle cells and evoke rhythmic slow wave depolarisations, which may activate voltage-dependent, dihydropyridine-sensitive (L-type) Ca2+ channels resulting in calcium influx and contraction (Horowitz et al. 1999). Neural inputs can condition the smooth muscle cell response to the ICC currents to either facilitate or reduce the possibility of the electrical threshold for Ca2+ channel activation being reached (Horowitz et al. 1999). Facilitating the electrical threshold may result in full depolarisation of the muscle membrane, indicated by an action potential resulting in a muscle contraction.

Hudson et al. (2001) were able to demonstrate slow wave activity in the equine ileum using in vitro intracellular recording techniques. It was subsequently documented that these persisted in horses with equine dysautonomia (grass sickness), although the temporal characteristics were altered (Hudson et al. 2002). Currently, slow wave activity has not been recorded in the equine large colon using this recording technique although it has been described in other mammalian species (Smith et al. 1987; Rae et al. 1998; Plujàet al. 2001; Yoneda et al. 2002). These studies have demonstrated a complex electrical pattern that, in contrast to the small intestine, may involve separate pacemaker centres, one of which may originate from the submucosal border of the circular muscle layer (Smith et al. 1987; Rae et al. 1998; Plujàet al. 2001). It is reasonable to speculate that equally complex electrical activity also exists in the equine colon.

The aim of this study was to characterise the in vitro intracellular myoelectrical events of the pelvic flexure in the normal horse. The pelvic flexure was chosen as this anatomical area is thought to be an important motility control centre in the horse and is also important clinically as many equine motility disorders, such as impactions, involve this region (Lowe et al. 1980; Sellers et al. 1982, 1984; Lopes and Pfeiffer 2000).

It is intended that the current study could form a basis for future clinical studies investigating the effect of different disease processes on slow wave activity such as impactions and paralytic ileus as well as evaluating the ICC as a possible therapeutic target.

Materials and methods

Samples of pelvic flexure apex were collected immediately following euthanasia by i.v. administration of quinalbarbitone/cinchocaine (Somulose)1 from 14 mature horses (Table 1). Tissue samples were immediately placed in modified oxygenated Krebs solution (118 mmol/l NaCl, 4.75 mmol/l KCl, 2.54 mmol/l CaCl2, 1.2 mmol/l MgSO4, 1.15 mmol/l NaH2PO4, 25 mmol/l NaHCO3 and 11.1 mmol/l glucose). The median age of the horses was 16.0 years (range 4–30 years) and comprised 10 geldings and 4 mares. All horses were subjected to euthanasia for reasons not relating to the intestinal tract and all samples were collected with the owners' written consent.

Table 1. Details of horses included in study
HorseAge (years)SexBreedReason for euthanasia
  1. G: gelding, F: female, TBx: Thoroughbred cross, WB: Warmblood.

114GIrish DraughtMaxillary tumour
224GShetland ponyOronasal fistula
417FTBxPharynx tumour
619FTBxMandibular tumour
810GIrish Draught-xLaminitis

A double-bladed scalpel was used to cut 1 mm thick sections from the pelvic flexure, in an orientation parallel to the circular muscle. The section was pinned on Sylgard in a tissue bath and superfused with warmed, oxygenated Krebs solution to ensure that the temperature in the tissue bath was kept between 36.5–37.5°C. The tissue sections were allowed to equilibrate for at least 60 min prior to commencing recordings.

Intracellular recordings were made from smooth muscle cells using glass microelectrodes filled with filtered 2 mol/l KCl and with resistances ranging from 20–45 MΩ. A calibrating graticule in the eyepiece of the dissecting microscope was used to determine the exact position of impaled smooth muscle cells relative to the thickness of the muscle layer. All recordings were made close to the submucosal border of the circular muscle layer. Cellular recordings were considered acceptable if there was a sharp initial drop in voltage when impalement was made and if the resting membrane potential (RMP) remained stable thereafter. The RMP was measured at the most negative reading for each impalement.

The L-type Ca2+ channel blocker nifedipine2 (1 µmol/l) was added to the superfusion fluid in 9 experiments (i.e. 9 horses). In 2 of these, nifedipine was added prior to commencing recording as strong muscle contractions precluded stable cellular impalements.

The sodium channel blocker tetrodotoxin3 (TTX) (1 µmol/l) was added to the superfusion fluid in 4 experiments (i.e. 4 horses).

The data were recorded and stored using an acquisition system (Power Lab 8SP)4 interfaced to a Power Macintosh G4 computer. The software package Chart5 was used to analyse the RMP, the amplitude, frequency and duration of membrane potential oscillations and other waveforms.


Electrical activity consistent with slow waves was recorded from the circular smooth muscle layer of pelvic flexure from 10 horses. This activity was characterised by a fast depolarising upstroke, a variable plateau phase and a slower repolarising phase (Fig 1). As shown in Figure 1, action potentials were invariably triggered at the peaks of each slow wave depolarisation. Whenever action potentials were recorded, a concomitant contraction of the smooth muscle in the tissue section was observed.

Figure 1.

Intracellular recording of membrane potential from circular smooth muscle cells in the pelvic flexure region of the equine large intestine. The segments of recording are from cells (A and B) in different experiments using tissue from Horses 1 and 9, respectively (seeTable 1). Each waveform is characterised by a fast depolarising upstroke, rapidly followed by a spike potential (arrow), then a variable plateau phase and a slower repolarising phase.

A degree of variation was found in the values for RMP and amplitude and frequency of the slow wave-like activity. This is typified by the 2 recordings shown in Figure 1. Although the waveforms recorded from Cells A and B have a similar appearance, Cell A had a more negative RMP, a larger amplitude slow wave and a slightly higher slow wave frequency than cell B. Overall, RMP ranged from -14.0 to -62.0 mV (mean -39.5 mV, median -41.0 mV). The range of amplitude for the slow wave-like activity was 2–11 mV (mean 4.0 mV, median 4.0 mV). The frequency ranged from 12–60/min (mean 20.7/min, median 22.0/min) between, but also within, particular animals.

In 7 different experiments (horses), nifedipine (1 µmol/l) abolished all ongoing electrical activity when added to the superfusion fluid during the experiment. In 4 of these, the electrical activity resumed following removal of the nifedipine and a 60–90 minute wash-out period. In 2 other horses, nifedipine was added at the onset of recording because of marked spontaneous tissue contractions making stable impalements very difficult. No electrical activity was recorded in these 2 experiments. Successful recording of slow wave-like activity was made in 3 further horses in which no nifedipine was applied. All attempts to make recordings failed in the 2 remaining horses, the first due to technical difficulties and the other for unknown reasons.

In 4 separate experiments, the addition of TTX (1 µmol/l) to the superfusion fluid was shown to have no effect on the ongoing electrical activity.


To the authors' knowledge, this is the first study to demonstrate in vitro intracellular electrical patterns consistent with slow wave activity in the equine colon.

Electrical recordings were made specifically from smooth muscle cells close to the submucosal border. This was because studies in other species have indicated that colonic slow wave activity originates from ICC located in this region (Smith et al. 1987; Rae et al. 1998; Plujàet al. 2001). The consistent recording of electrical activity in the current study suggests that pacemaker activity may also originate in this region in the horse. However, a previous immunohistochemical study failed to demonstrate distinct bands of ICC in the submucosal region of the pelvic flexure, although the density throughout the circular muscle appeared greater than in other parts of the equine colon (Hudson et al. 1999). Further investigations, including ablation studies, are required to determine the anatomical location and organisation of ICC driving this activity.

There was a large degree of variation in the characteristics of the slow waves recorded. Similar variations have also been observed in preparations involving the equine small intestine (Hudson et al. 2001). This may be due to the different behaviour of individual smooth muscle cells or indeed smooth muscle syncytia. However, it may also be the result of separate populations of pacemaker cells discharging at slightly different frequencies as suggested by Bortoff (1965) who noticed waxing and waning of slow wave amplitude during in vitro studies of feline jejunum. It is also possible there is variation in the degree of influence that the enteric nervous system exerts in these in vitro tissue sections; this is an area that warrants further investigation.

It was interesting to observe that the addition of nifedipine at a concentration of 1 µmol/l completely abolished all electrical activity. This differs from previous findings in the equine ileum where nifedipine administration at this concentration only resulted in a reduction in frequency and increase in duration of slow waves (Hudson et al. 2001). However, although the current findings differed from those observed in the equine ileum, they are consistent with previous investigations of colonic electrical activity in other species (including human, rat and mouse) where 1 µmol/l nifedipine abolished all activity to indicate a dependence on activation of voltage-gated L-type Ca2+ channels (Rae et al. 1998; Plujàet al. 2001; Yoneda et al. 2002, 2003). In contrast, slow waves were not abolished by nifedipine in the canine colon (Smith et al. 1987; Huizinga et al. 1991). Thus there would appear to be marked species variability with respect to the involvement of L-type calcium channels in the generation of slow wave activity in this region. It is also possible that in the horse, L-type voltage-gated calcium channels may be responsible for slow wave generation, whereas the canine colon relies on different mechanisms. Regardless of the mechanism, the current findings indicate significant and potentially important differences in the ionic mechanisms responsible for slow waves in the different regions of the equine intestinal tract. It remains to be seen whether these have clinical relevance.

The precise ionic mechanisms involved in the generation of slow waves have not yet been established but it has been proposed that pacemaker currents are initiated by voltage-independent, Ca2+-inhibited, nonselective cation conductances in the ICC (Koh et al. 2002; Goto et al. 2004; Sanders et al. 2004). Regardless of the initial mechanism, the opening of L-type Ca2+ channels is closely linked to smooth muscle cell contraction (Sanders 1996; Farrugia 1999) as also demonstrated in the equine ileum through inhibition of spike potentials when adding L-type Ca2+ channel blockers such as nifedipine (Hudson et al. 2001).

Based on the shapes and patterns of the electrical activity recorded it seems reasonable to assume that the activity recorded in the current study did indeed originate from the ICC. This argument is strengthened by the observation that this activity was not affected by the addition of the nerve-blocking drug TTX. However, this argument could be further strengthened by carrying out ablation studies as well as adding modulators of intracellular calcium stores such as ryanodine or indeed of IP3 receptor inhibitors such as 2-aminoethoxydiphenyl borate to the superfusion fluid, which prevents calcium release from the endoplasmic reticulum, a mechanism thought to be instrumental in slow wave generation (Takaki 2003; Sanders et al. 2004).

Although the current study has made some interesting observations, further work is required to achieve a better understanding of the mechanisms involved in generating and maintaining equine intestinal motility patterns.

Conflicts of interest

No conflicts of interest have been declared.

Sources of funding

The Norwegian School of Veterinary Science funded the study.


We are grateful to all clinicians and staff at the Royal (Dick) School of Veterinary Studies who helped with the collection of tissue samples.

Manufacturers' addresses

1 Dechra Pharmaceuticals PLC, Stoke-on-Trent, Staffordshire, UK.

2 Sigma, Watford, Hertfordshire, UK.

3 Sigma, St Louis, Missouri, USA.

4 AD Instruments, Oxford, UK.

5 Chart, AD Instruments, Oxford, UK.