SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. References
  7. Acknowledgements

In a previous study, we established that leptin controls food intake and immune responses by acting on intestinal vagal chemosensitive mechanoreceptors via a functional link with interleukin-1β (Il-1β). Since the control of intestinal motility is one of the main roles of the vagal afferent fibres, we investigated the effects of leptin on intestinal electromyographic (EMG) activity which reflects intestinal motility. For this purpose, the effects of locally injected leptin on small intestine spontaneous EMG activity were studied in 23 anaesthetised cats. The EMG activity was recorded using bipolar electrodes implanted in the proximal small intestine. Leptin and Il-1β (0.1, 1 and 10 μg), administered through the artery irrigating the upper part of the intestine 20 min after cholecystokinin (CCK, 10 μg, i.a.), had significant (P < 0.001) excitatory effects on intestinal EMG activity. The effects of both substances were blocked by the endogenous interleukin-1β receptor antagonist (Il-1ra, 250 μg, i.a.), by atropine (250 μg, i.a.) and by vagotomy. In the absence of CCK, leptin and Il-1β had no effect on intestinal electrical activity. It can therefore be concluded that: (1) leptin is effective only after the previous intervention of CCK, (2) the enhancement of the electrical activity induced by leptin involves Il-1β receptors and the cholinergic excitatory pathway, (3) the modes whereby the leptin-induced enhancement of EMG activity occurs strongly suggest that these effects are due to a long-loop reflex involving intestinal vagal afferent fibres and the parasympathetic nervous system.

Leptin is known to be secreted mainly by adipose tissue in order to regulate food intake and body weight homeostasis by acting on the hypothalamus (Zhang et al. 1994; Pelleymounter et al. 1995). However, leptin is not released by adipocytes alone, it is also synthesised by cells of other types, namely epithelial cells in both rat and human stomach (Bado et al. 1998; Sobhani et al. 2000). In addition, leptin-specific receptors have been found to exist in peripheral tissues, especially in the intestinal intramural plexuses and the interstitial cells of Cajal in guinea-pigs (Liu et al. 1999), and in the sympathetic prevertebral ganglionic neurones in mice and rats (Miller et al. 1999). These findings suggest that leptin may also play an extra-hypothalamic role, particularly during food intake.

The results of previous studies have indicated that in addition to the direct effects observed on brain targets via hormonal pathways, leptin can act on gastric (Wang et al. 1997, 2000) and intestinal (Gaigéet al. 2002) vagal afferent nerve fibres sending rapid signals to the central nervous system via neuronal pathways. It is worth noting that the presence of CCK has been found to modify the responses to leptin of both organs (Wang et al. 1997, 2000; Barrachina et al. 1997b; Gaigéet al. 2002) and that at least in the case of the intestine, these effects of leptin are strongly dependent on the release of Il-1β (Gaigéet al. 2002). Moreover, these studies showed the presence of several types of leptin-sensitive gastric and intestinal vagal afferent nerve fibres (Wang et al. 1997; Gaigéet al. 2002). However, the exact functional role of these fibres is still a matter of debate.

It has by now been clearly established that ingestion is associated with numerous gastrointestinal motor processes and that strong functional interactions occur between the various parts of the digestive tract, such as the gastrointestinal reflexes controlling gastric emptying and intestinal motility (Gonella et al. 1987; Raybould & Lloyd, 1994; Yuan et al. 2001). In addition, the main role of the gastrointestinal afferent fibres is that of controlling gastrointestinal motility (Tansy & Kendall, 1977; Gonella et al. 1987). The results of previous studies have shown that peripheral application of leptin does not affect gastric emptying (Barrachina et al. 1997a; Martinez et al. 1999). Although the vagal influence decreases towards the distal part of the small intestine (Kewenter, 1965; Hall et al. 1982), the proximal part of the intestine, namely the duodenum, seems to possess a considerable amount of vagal innervation (Berthoud et al. 1991, 1997). For these reasons, it was decided in the present study to examine the effects of leptin and Il-1β on duodenal electromyographic (EMG) activity with a view to elucidating the roles of the various leptin-sensitive vagal afferent fibres.

Methods

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. References
  7. Acknowledgements

The experimental procedures described here were carried out in keeping with the European guidelines for the care and use of laboratory animals (Council Directive 86/6009/EEC). These experimental methods were practically the same as those described by Bouvier et al. (1990) and Gaigéet al. (2002). Twenty-three adult European cats of either sex weighing 2.8 to 4.2 kg (mean value: 3.4 ± 0.2 kg) were used in these acute experiments. The animals had fasted for 24 h before undergoing surgery; they were given water ad libitum.

Surgical preparation

Anaesthesia.

Anaesthesia was induced by administering a gaseous mixture (2 % halothane, 98 % air; Bélamont, Neuilly-sur-Seine, France). The trachea was cannulated and a catheter was inserted into a radial vein, through which the main anaesthetic (chloralose, 80 mg kg−1, i.v., Sigma, Saint Quentin-Fallavier, France) was then administered intravenously. The depth of anaesthesia was monitored by checking that the palpebral reflex was absent and the pupils constricted. In addition, the blood pressure and heart rate were systematically monitored throughout the experiment to detect any changes in the level of anaesthesia. The appropriate depth of anaesthesia was characterised by constricted pupils, stable blood pressure and heart rate, and by the absence of the palpebral reflex. Whenever any of these parameters changed, showing that the anaesthesia was weakening, a flush of chloralose (8 mg kg−1, i.v.) was administered. Body temperature was maintained at 37 ± 1 °C by means of a heating blanket.

In each experiment, the surgical procedure began when the parameters characterising the appropriate depth of anaesthesia were reached (after approximately 40-60 min). The stabilisation time allowed to elapse between the end of the surgery and the start of the tests with the compounds was 2 h.

At the end of each experiment, the cats were killed by administering an overdose of barbiturates (sodium thiopentone, 70 mg kg−1, i.v., Sanofi, Libourne, France).

Administration of pharmacological substances.

Pharmacological drugs were administered locally through a catheter introduced into the left femoral artery so that its tip was located at the level of the coeliac artery irrigating the upper part of the small intestine. Its position was checked post mortem by performing a 1 % methylene blue injection (Sigma, Saint Quentin-Fallavier, France) into the catheter.

Dissection of the vagus nerves.

Both vagus nerves were dissected along 1 cm in the neck so that they could be cut during the experiment.

Stimulation techniques

Chemical stimulation.

All the substances were administered within 5 s. Each injection was followed by a flush with 1 ml of saline solution. To prevent the occurrence of any tachyphylaxic disorders, the time elapsing between successive injections was set at 60 min. In the tests performed to check whether CCK, substance P, or phenylbiguanide (PBG, a serotoninergic agonist) modified the effects of leptin and Il-1β, this interval was reduced, however, to 20 min, since this was the time at which the maximum combined CCK/leptin effects on cat intestinal vagal afferent fibres were observed (Gaigéet al. 2002).

Drugs.

The following drugs were used: murine leptin lyophilised from a sterile filtered solution in phosphate-buffered saline (PBS) (AMGEN Thousand Oaks, USA), recombinant human interleukin-1β (Il-1β, AMGEN), recombinant human interleukin-1β receptor antagonist (Il-1ra, AMGEN), cholecystokinin-8 sulfated (CCK, Sigma, Saint Quentin-Fallavier, France), substance P (Sigma), phenylbiguanide (PBG, Sigma) and atropine sulfate salt (Sigma).

All the substances were diluted in 1 ml of physiological saline. Leptin and Il-1β were injected at the following doses: 0.1, 1 and 10 μg (6.25, 62.50, 625.00 pmol and 99.50, 995.02, 9950.25 pmol, respectively). CCK, substance P and PBG were injected in a dose of 10 μg (8.75, 7.42 and 56.43 nmol, respectively). All these doses were similar to those previously used by Gaigéet al. (2002) to activate vagal afferent fibres. Il-1ra and atropine were administered in a dose of 250 μg (14.71 and 369.39 nmol, respectively; Bouvier & Gonella, 1981; Gaigéet al. 2002). The weight of the animals was not taken into account, since the drugs were injected locally into the intestine.

Electromyographic recordings.

Bipolar recordings of proximal small intestine electrical activity were performed using the method described by Basmadjian & Stecko (1962). The electrodes, which consisted of 0.1 mm insulated wire of NiCr alloy (Cr 21 %), were implanted into the smooth muscle. The electrodes were connected to a polygraph ink pen recorder (alternating current amplifiers, band pass 0-120 Hz, time constant 2.5 s). In each experiment, bipolar electrical recordings (with the two electrodes 5 mm apart) were performed at several points (minimum: 3) 4 cm apart to investigate the duodenum. The first electrode was implanted 2 cm below the bile duct. All the effects recorded after drug injection were similar with all the derivations used. However, only one recording per animal was used in the statistical analysis. The recordings started 5 min before the injections were administered and lasted for 20 min. The frequencies of the spike potentials (SP min−1) and slow waves (SW min−1) were automatically calculated by means of a Macintosh computer (Apple) using MacLab System hardware and software programs (Analog Digital Instruments) to assess any changes in small intestine motility.

Statistical analysis

An index to the small intestine motility was estimated, based on the mean spike potential frequency (SP min−1) recorded for 4 min prior to drug injection (control) and for 4 min after the onset of the effect. However, whenever the drugs had no effect, the second value was estimated, assuming the latency to be that observed when this drug actually had an effect, namely when it was administered after CCK. For each injection, we calculated the percentage of variation of the mean discharge frequencies (%). Data are expressed as means ±s.e.m. The values given in parentheses are the mean discharge frequency before the injection ±s.e.m.vs. the mean discharge frequency after the injection ±s.e.m. Statistical analysis was performed using Student's paired t test to assess the significance of the results. An analysis of variance followed by Scheffépost hoc test were also carried out to compare the effects of multiple drug injections. The significance level was taken to be P < 0.05.

Results

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. References
  7. Acknowledgements

The spontaneous EMG activity of the small intestine was studied in 23 fasted cats under chloralose anaesthesia. The electrical activity of the small intestine consists mainly of slow waves, i.e. slow variations in the resting membrane potential of the smooth muscle cells, with occasional superimposed spike potentials inducing muscle contractions. The frequencies of the slow waves were calculated before and after each injection and always remained constant, ranging from 11.50 to 17.25 SW min−1 (mean value: 14.50 ± 0.06 SW min−1). The effect of the vehicle administered was investigated in each cat, and no significant changes in the spontaneous spike potential or slow wave activity were observed after intra-arterial local administration (n= 23).

Effects of leptin on small intestine spontaneous EMG activity

Effects of leptin.

Leptin administered at three increasing doses, 0.1 μg (n= 7), 1 μg (n= 7) and 10 μg (n= 23), never affected the mean frequency of the spike potentials (3.75 ± 1.17 vs. 3.50 ± 1.05, 4.68 ± 1.03 vs. 4.68 ± 1.05 and 3.18 ± 0.43 vs. 3.18 ± 0.42 SP min−1) (Figs 1Ab, 1Bb and 2).

image

Figure 1. Effects of leptin (10 μg, i.a.), CCK (10 μg, i.a.) and leptin (10 μg, i.a.) after CCK on the spontaneous electrical activity of one cat small intestine

A, effects of leptin, CCK and leptin after CCK on the EMG activity of one cat small intestine. Aa, (control), the basal activity consisted of slow waves with occasional superimposed spike potentials inducing intestinal contractions. Ab, leptin administration (arrow) had no effect on the EMG activity. Ac, CCK administration (arrow) induced an increase in the mean frequency of the spike potentials (c1, c2, continuous traces). Ad, leptin administered (arrow) 20 min after CCK enhanced the mean frequency of the spike potentials (d1, d2, continuous traces). B, histograms comparing the effects of intra-arterially administered leptin (10 μg, i.a.), those of CCK (10 μg, i.a.) and those of leptin after CCK on the EMG activity of the same small intestine. Ba, control. Bb, leptin administration (arrow). Bc, CCK administration (arrow). Bd, leptin administration (arrow) 20 min after CCK. Leptin increased the small intestine electrical activity only after CCK. (SP min−1: spike potentials per minute).

Download figure to PowerPoint

image

Figure 2. Dose-response curve giving the effects of leptin and Il-1β (0.1, 1 and 10 μg; i.a.) on the spontaneous electrical activity of cat small intestine

At each dose, leptin and Il-1β alone had no effects on the EMG activity of the small intestine. However, leptin significantly enhanced this activity in all three doses tested after CCK administration (10 μg, i.a.). Il-1β significantly enhanced small intestine EMG activity in doses of 1 and 10 μg after CCK administration (10 μg, i.a.). ns: not significant; **P < 0.001; ***P < 0.0001; (n= 7).

Download figure to PowerPoint

Effects of leptin after CCK.

CCK, at a dose of 10 μg, significantly increased (P < 0.0001) the mean frequency of the spike potentials by 962.13 ± 58.61 % (3.58 ± 0.26 vs. 36.74 ± 1.42 SP min−1) with a latency of 10.6 ± 0.6 s and a duration of 351.5 ± 7.2 s in all the cats tested (n= 23; Fig. 1Ac and 1Bc). Leptin administered 20 min after CCK injection enhanced the mean frequency of the spike potentials (Figs 1Ad, 1Bd, 2 and 3A). The responses of seven cats were investigated at three increasing doses of leptin: 0.1, 1 and 10 μg. At all these doses, leptin significantly increased (P < 0.001, P < 0.001 and P < 0.0001) the mean frequency of the spike potentials by 205.83 ± 35.60 % (3.39 ± 0.33 vs. 10.21 ± 1.24 SP min−1), 495.02 ± 26.21 % (3.79 ± 0.64 vs. 22.64 ± 2.70 SP min−1) and 842.98 ± 29.15 % (3.64 ± 0.39 vs. 33.89 ± 3.21 SP min−1), respectively (Fig. 2). These effects had the following latencies: 24.7 ± 2.1, 23.4 ± 2.8 and 20.0 ± 3.1 s, respectively, and the following durations: 304.1 ± 18.8, 309.7 ± 18.3 and 328.6 ± 16.0 s, respectively. The dose-response curve indicates that the dose of 10 μg can be routinely employed to stimulate the small intestine (Fig. 2). In all the cats tested (n= 23), leptin (10 μg) injected 20 min after CCK administration significantly increased (P < 0.0001) the mean frequency of the spike potentials by 802.44 ± 25.22 % (3.92 ± 0.43 vs. 35.25 ± 3.09 SP min−1), with a latency of 20.7 ± 3.5 s and a duration of 338.3 ± 11.1 s (Fig. 3A). Two injections of 10 μg leptin performed at an interval of 20 min had no effect on the mean frequency of the spike potentials (n= 4). At all the doses tested, leptin injections affected neither the arterial blood pressure (101.89 ± 11.63 vs. 104.32 ± 12.79 mmHg) nor the cardiac frequency (71.46 ± 0.87 vs. 73.42 ± 0.74 beats min−1).

image

Figure 3. Statistical analysis of the effects of leptin (Lep, 10 μg, i.a.) on the spontaneous electrical activity of cat small intestine before and after CCK, substance P and PBG

Columns labelled C give the mean values of the control responses expressed as 100%. The vertical bars at the top of the columns give the s.e.m.A, statistical analysis of the effects of leptin alone and 20 min after CCK (10 μg, i.a.) and those of CCK on the EMG activity (n= 23). Leptin alone had no effect on the small intestine EMG activity, whereas leptin administration 20 min after CCK increased the EMG activity. Note that CCK also enhanced the EMG activity of the small intestine. B, statistical analysis of the effects of leptin alone and 20 min after substance P (subP, 10 μg, i.a.) and those of substance P on the EMG activity (n= 4). Substance P increased the small intestine electrical activity, but leptin alone and 20 min after substance P had no effect on the EMG activity. C, statistical analysis of the effects of leptin alone and 20 min after PBG (10 μg, i.a.) and those of PBG on the EMG activity (n= 4). Neither PBG alone nor leptin before and 20 min after PBG had any effect on the EMG activity of the small intestine. *P < 0.05, ***P < 0.0001: Student's paired t test refers to the effects of the drug as compared to the control. †P < 0.05, †††P < 0.0001: Scheffépost hoc test refers to the comparison of the effects of the various drugs tested. ns: not significant (whatever the test).

Download figure to PowerPoint

Effects of leptin after substance P.

Substance P given at a dose of 10 μg significantly increased (P < 0.05) the mean frequency of the spike potentials by 504.46 ± 110.56 % (2.63 ± 0.30 vs. 16.13 ± 3.61 SP min−1) with a latency of 9.8 ± 1.9 s and a duration of 317.8 ± 63.6 s (n= 4). In these four cats, leptin (10 μg) administered either alone or 20 min after substance P did not affect the mean frequency of the spike potentials (Fig. 3B).

Effects of leptin after PBG.

PBG administered at a dose of 10 μg had no effect on the mean frequency of the spike potentials (n= 4). In these four cats, leptin (10 μg) administered either alone or 20 min after PBG did not affect the mean frequency of the spike potentials (Fig. 3C).

Effects of Il-1ra on small intestine activation by leptin after CCK.

In the four cats tested with Il-1ra, CCK and leptin, injecting CCK alone significantly increased (P < 0.001) the mean frequency of the spike potentials by 943.06 ± 48.40 % (3.42 ± 1.24 vs. 35.37 ± 1.98 SP min−1) with a latency of 11.0 ± 1.1 s and a duration of 338.7 ± 18.7 s. CCK administered to those four cats 5 min after Il-1ra (250 μg) also induced a significant (P < 0.001) increase of 967.53 ± 54.32 % (3.56 ± 1.41 vs. 37.22 ± 2.00 SP min−1) with a latency of 10.0 ± 1.1 s and a duration of 357.9 ± 14.3 s. The excitatory effects of CCK observed before Il-1ra administration did not differ significantly from those observed after Il-1ra. Leptin administered at a dose of 10 μg, either alone or 5 min after Il-1ra, never affected the mean frequency of the spike potentials (n= 4). In those four cats, leptin (10 μg) injected 20 min after CCK significantly increased (P < 0.001) the mean frequency of the spike potentials by 831.24 ± 37.41 % (3.48 ± 0.92 vs. 36.17 ± 3.00 SP min−1), with a latency of 19.9 ± 3.5 s and a duration of 316.6 ± 21.5 s. However, the same dose of leptin administered after CCK pre-treatment and 5 min after Il-1ra had no further significant effects on the mean frequency of the spike potentials. The excitatory effects of leptin were therefore significantly inhibited (P < 0.001) by Il-1ra (Fig. 4A and B).

image

Figure 4. Effects of leptin (10 μg, i.a.) after CCK (10 μg, i.a.), and leptin after CCK and Il-1ra (250 μg, i.a.) on the spontaneous electrical activity of cat small intestine

A, effects of leptin (10 μg, i.a.) after CCK (10 μg, i.a.), and those of leptin after CCK and Il-1ra (250 μg, i.a.) on the EMG activity of one cat small intestine. Aa (control), the basal activity consisted of slow waves with occasional superimposed spike potentials inducing intestinal contractions. Ab, leptin administered 20 min after CCK enhanced the mean frequency of the spike potentials (trace: 1 min after injecting leptin). Ac, leptin administered 20 min after CCK and 5 min after Il-1ra had no further effect on the EMG activity (trace: 1 min after leptin injection). B, statistical analysis of the effects of leptin administered 20 min after CCK, and those of leptin administered 20 min after CCK and 5 min after Il-1ra (n= 4). Column labelled C gives the mean value of the control responses expressed as 100%. The vertical bars at the top of the columns give the s.e.m. Il-1ra administration inhibited the effects of leptin after CCK on the small intestine EMG activity. ns: not significant, **P < 0.001: Student’ s paired t test refers to the effects of the drug as compared to the control. ††P < 0.001: Scheffépost hoc test refers to the comparison of the effects of the various drugs tested.

Download figure to PowerPoint

Effects of atropine on small intestine activation by leptin after CCK.

In the four cats tested, injection of atropine (250 μg) alone, a muscarinic receptor blocking agent, completely abolished the spontaneous spike potentials (3.61 ± 0.33 vs. 00.00 ± 0.00 SP min−1) with a latency of 9.8 ± 0.9 s. When administered after injecting atropine, leptin (10 μg), CCK and leptin 20 min after CCK had no effect on the activity of the small intestine. In all the cats tested (n= 4), the effects of CCK and leptin after CCK were therefore inhibited by atropine administration.

Effects of vagotomy on small intestine activation by leptin after CCK.

In the four cats tested, sections of the two vagal trunks, performed 1 cm above the nodose ganglia, significantly decreased (P < 0.05) the mean frequency of the spike potentials by 21.21 ± 8.38 % (4.19 ± 0.33 vs. 3.38 ± 0.55 SP min−1). After vagotomy, CCK at a dose of 10 μg significantly increased (P < 0.005) the mean frequency of the spike potentials by 462.03 ± 41.37 % (3.50 ± 0.54 vs. 19.19 ± 2.55 SP min−1) with a latency of 10.5 ± 1.1 s and a duration of 319.0 ± 12.7 s (n= 4). Vagotomy was found to significantly reduce (P < 0.05) the intensity and the duration, but not the latency of the effects of CCK. After vagotomy, injecting leptin (10 μg) either alone or 20 min after CCK had no effect on the mean frequency of the spike potentials. In all the cats tested (n= 4), vagotomy significantly inhibited (P < 0.001) the excitatory effects of leptin observed 20 min after CCK administration (Fig. 5).

image

Figure 5. Statistical analysis of the effects of vagotomy on the small intestine EMG activation induced by leptin (10 μg, i.a.) 20 min after CCK (10 μg, i.a.)

For each cat tested (n= 4), leptin (or Il-1β) was applied after CCK pre-treatment and tested before and after vagotomy. Sectioning the two vagal trunks significantly (P < 0.05) decreased the mean frequency of the spike potentials by 21.21 ± 8.38 %. Column labelled C gives the mean value of the control responses (before and after vagotomy), expressed as 100%. Vagotomy inhibited the effects of leptin after CCK on the small intestine EMG activity. The vertical bars at the top of the columns give the s.e.m. ns: not significant, **P < 0.001: Student's paired t test refers to the effects of the drug as compared to the control. ††P < 0.001: Scheffépost hoc test refers to the comparison of the effects of the various drugs tested.

Download figure to PowerPoint

Effects of Il-1β on small intestine spontaneous electrical activity

Effects of Il-1β.

Il-1β administered at three increasing doses - 0.1 μg (n= 7), 1 μg (n= 19) and 10 μg (n= 7) - never affected the mean frequency of the spike potentials (2.96 ± 0.48 vs. 3.07 ± 0.55, 3.29 ± 0.42 vs. 3.43 ± 0.39 and 3.96 ± 0.77 vs. 4.29 ± 0.84 SP min−1) (Figs 2, 6Ab, 6Bb and 6C).

image

Figure 6. Effects of Il-1β (1 μg; i.a.), Il-1β after CCK (10 μg, i.a.), and those of Il-1β after CCK and Il-1ra (250 μg; i.a.) on the spontaneous electrical activity of cat small intestine

A, effects of Il-1β, Il-1β after CCK, and those of Il-1β after CCK and Il-1ra on the EMG activity of one small intestine. Aa (control), the basal activity consisted of slow waves with occasional superimposed spike potentials inducing intestinal contractions. Ab, Il-1β administration (arrow) had no effect on EMG activity. Ac, Il-1β administered (arrow) 20 min after CCK induced an increase in the mean frequency of the spike potentials (c1, c2 continuous traces). Ad, Il-1β administered (arrow) 20 min after CCK and 5 min after Il-1ra enhanced the mean frequency of the spike potentials. B, histograms comparing the effects of intra-arterially administered Il-1β (1 μg, i.a.), those of Il-1β after CCK (10 μg, i.a.) and those of Il-1β after CCK and Il-1ra (250 μg; i.a.) on the EMG activity of the same small intestine. Ba, control. Bb, Il-1β administration (arrow). Bc, Il-1β administered (arrow) 20 min after CCK. Bd, Il-1β administered (arrow) 20 min after CCK and 5 min after Il-1ra. (SP min−1 : spike potentials per minute). C, statistical analysis of the effects of Il-1β (1 μg, i.a.) alone, 20 min after CCK (10 μg, i.a.), and 20 min after CCK and 5 min after Il-1ra (n= 4). Column C gives the mean value of the control responses expressed as 100%. The vertical bars at the top of the columns give the s.e.m. Il-1β increased the EMG activity only 20 min after CCK. This effect was blocked by previously administering Il-1ra. ns: not significant, **P < 0.001: Student's paired t test refers to the effects of the drug as compared to the control. ††P < 0.001: Scheffépost hoc test refers to the comparison of the effects of the various drugs tested.

Download figure to PowerPoint

Effects of Il-1β after CCK.

Il-1β administered 20 min after CCK injection enhanced the mean frequency of the spike potentials in all the cases studied (Fig. 6Ac, 6Bc and 6C). The responses of seven cats were investigated at three increasing doses of Il-1β: 0.1, 1 and 10 μg. At 0.1 μg, no effect of Il-1β was observed. However, at 1 and 10 μg, Il-1β significantly increased (P < 0.001 and P= 0.0001) the mean frequency of the spike potentials by 857.32 ± 61.74 % (3.14 ± 0.67 vs. 29.00 ± 3.88 SP min−1) and 999.31 ± 51.32 % (3.43 ± 0.48 vs. 36.32 ± 4.19 SP min−1), respectively (Fig. 2). These effects had the following latencies: 23.4 ± 1.9 and 20.0 ± 3.0 s, respectively, and the following durations: 312.6 ± 17.8 and 319.0 ± 6.9 s, respectively. The dose-response curve indicates that the dose of 1 μg can be routinely employed to stimulate the small intestine (Fig. 2). In all the cats tested (n= 19), Il-1β (1 μg) given 20 min after CCK administration significantly increased (P < 0.0001) the mean frequency of the spike potentials by 803.61 ± 41.31 % (3.70 ± 0.39 vs. 33.37 ± 3.01 SP min−1), with a latency of 22.7 ± 1.8 s and a duration of 333.1 ± 11.8 s. Neither of two injections of the same dose of Il-1β, performed at an interval of 20 min, had any effect on the mean frequency of the spike potentials (n= 4).

Effects of Il-1ra on small intestine activation by Il-1β after CCK.

In the four cats tested with Il-1ra, CCK and Il-1β, injecting CCK alone (10 μg) significantly increased (P < 0.001) the mean frequency of the spike potentials by 948.86 ± 78.36 % (3.44 ± 1.02 vs. 36.25 ± 1.11 SP min−1) with a latency of 10.5 ± 2.9 s and a duration of 317.5 ± 26.4 s. CCK administered 5 min after Il-1ra (250 μg) also induced a significant (P < 0.001) increase of 936.62 ± 61.39 % (3.68 ± 1.27 vs. 37.89 ± 1.77 SP min−1) with a latency of 10.2 ± 2.0 s and a duration of 329.8 ± 21.7 s (n= 4). The excitatory effects of CCK observed before Il-1ra did not differ significantly from those observed after Il-1ra. Il-1β administered at a dose of 1 μg either alone (Fig. 6Ab and 6Bb) or 5 min after Il-1ra never affected the mean frequency of the spike potentials (n= 4). Il-1β (1 μg) injection 20 min after CCK significantly increased (P < 0.001) the mean frequency of the spike potentials by 822.43 ± 39.51 % (3.48 ± 0.52 vs. 35.36 ± 3.00 SP min−1), with a latency of 21.0 ± 1.8 s and a duration of 308.6 ± 19.8 s (Fig. 6Ac and 6Bc). By contrast, Il-1β administration in the same dose after CCK pre-treatment and 5 min after Il-1ra had no further significant effect on the mean frequency of the spike potentials (Fig. 6Ad, 6Bd and 6C). The excitatory effects of Il-1β were therefore significantly inhibited (P < 0.001) by Il-1ra (n= 4).

Effects of atropine on small intestine activation by Il-1β after CCK.

In the four cats tested, injecting atropine (250 μg) alone completely abolished the spontaneous spike potentials with a latency of 8.5 ± 0.9 s. After injecting atropine, Il-1β (1 μg), CCK and Il-1β administered 20 min after CCK had no effect on the activity of the small intestine (n= 4).

Effects of vagotomy on the activation of the small intestine by Il-1β after CCK.

In the same four cats tested with leptin, after vagotomy, injecting Il-1β (1 μg) before and 20 min after CCK had no effect on the mean frequency of the spike potentials (Fig. 7). Vagotomy significantly inhibited (P < 0.001) the effects of Il-1β (1 μg), 20 min after CCK administration (n= 4).

image

Figure 7. Statistical analysis of the effects of vagotomy on the small intestine electrical activation induced by Il-1β (1 μg, i.a.) 20 min after CCK (10 μg, i.a.)

For each cat tested (n= 4), leptin (or Il-1β) was applied after CCK pre-treatment and tested before and after vagotomy. Sectioning the two vagal trunks significantly (P < 0.05) decreased the mean frequency of the spike potentials by 21.21 ± 8.38 %. Column C gives the mean value of the control responses (before and after vagotomy), expressed as 100%. Vagotomy inhibited the effects of Il-1β after CCK on the small intestine EMG activity. The vertical bars at the top of the columns give the s.e.m. (n= 4). ns: not significant, **P < 0.001: Student's paired t test refers to the effects of the drug as compared to the control. ††P < 0.001: Scheffépost hoc test refers to the comparison of the effects of the various drugs tested.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. References
  7. Acknowledgements

The results obtained in the present study clearly show that the local intra-arterial injection of leptin 20 min after CCK administration stimulates the small intestine EMG activity in cats, whereas in the absence of CCK, leptin has no effect on this activity. The leptin-evoked spike potential activity was found to increase in a dose-dependent manner. This effect was not due to the arterial distension associated with the drug injection (1 ml within 5 s), since the frequency of the spike potentials remained unchanged when the vehicle was injected alone.

CCK/leptin interactions are known to be involved in the control of various functions such as satiety signals and body weight loss (Barrachina et al. 1997b; Wang et al. 1997, 1998; Matson et al. 2000; Buyse et al. 2001; Gaigéet al. 2002), but no information is available so far about the mechanisms underlying these interactions. Our results demonstrate at least that the mechanism whereby CCK induces the effects of leptin cannot be due to the well-known excitatory effect of CCK on intestinal motility (Vizi et al. 1973; Milenov et al. 1998), since leptin was injected 20 min after CCK, the excitatory effects of which last for only about 6 min. The spike potential frequency recorded before injecting leptin alone was therefore similar to that observed before injecting leptin 20 min after CCK. In addition, it is noteworthy that substance P, which increases the EMG activity (Holzer & Holzer-Petsche, 1997), does not induce this excitatory effect of leptin. In the experiments focusing on CCK/leptin interactions, leptin was systematically injected twice: (1) before CCK in order to check the absence of any effect on EMG activity, (2) after CCK. The effects observed were not due to a tachyphylaxic process, since two injections of leptin given 20 min apart had no effect on EMG activity.

Intestinal EMG activation induced by leptin via vagal afferent nerve fibres

The leptin-evoked intestinal motor activation was not due to the peripheral activation of the gastrointestinal tract, but involved the CNS, since it was abolished by vagotomy. The effects of leptin therefore resulted neither from the activation of the leptin-specific receptors present in the intestinal intramural plexuses, nor from the involvement of leptin-sensitive intramural neurones (Liu et al. 1999). The present intestinal EMG activation occurs via a long-loop reflex involving vagus nerves. Moreover several lines of evidence support the idea that a vagal afferent nerve fibre-mediated reflex was involved.

Previous in vitro (Wang et al. 1997) and in vivo (Barrachina et al. 1997b; Gaigéet al. 2002) data have indicated that leptin and CCK are able to act separately on vagal afferent fibres but that synergistic interactions between these two drugs also occur which modulate signals to the brain. Barrachina et al. (1997b) and Wang et al. (1997) have shown the existence of two groups of leptin-responsive gastric vagal afferent terminals in rats. Those in the first group produce an activatory response to leptin, and CCK pre-treatment does not affect their sensitivity to leptin. No interactions therefore seem to occur between leptin and CCK in this group of fibres. However, the afferent terminals in the second group either produce an inhibitory response to leptin or do not respond at all, whereas after CCK pre-treatment, the sensitivity of this group of fibres changes and they produce an activatory response to leptin, which shows that interactions have occurred between these two drugs. In a previous study on anaesthetised cats, we established that leptin has excitatory effects on some intestinal vagal afferent fibres (type 1 units) and inhibitory effects on the others (type 2 units), which were potentiated and abolished, respectively, when leptin was administered 20 min after CCK (Gaigéet al. 2002). In cat small intestine, both types of vagal afferent nerve fibres are affected by the strong interactions occurring between leptin and CCK, two drugs involved in controlling satiety signals and food intake.

In addition, we have previously reported that the effects of leptin/CCK interactions on leptin-activated intestinal vagal afferent fibres are actually specific ones, since they are triggered in response to substance P and PBG, whereas these two substances have no effect on the excitatory effects of leptin (Gaigéet al. 2002). Among the substances which were found to be able to stimulate vagal afferent fibres, CCK was the only one which enhanced the effects of leptin on these fibres (Gaigéet al. 2002). A similar finding was obtained here as far as the intestinal EMG activity was concerned. As a matter of fact the present data indicate that intra-arterial leptin injection administered 20 min after substance P or PBG has no effect on small intestine spontaneous EMG activity in cats. This result supports the above hypothesis that leptin-evoked intestinal motor activation may be mediated by vagal afferent nerve fibres.

In addition, we previously established that injecting Il-1ra, an endogenous peptide which is a specific Il-1β receptor antagonist, inhibits the excitatory response to leptin of vagal afferent fibres (Gaigéet al. 2002). We therefore tested the effects of Il-1ra on the activation of small intestine EMG activity induced by leptin after CCK, in order to check whether the vagal afferent nerve fibres may be involved in this reflex. No significant changes in the mean spike potentials were observed when leptin after CCK pre-treatment was injected 5 min after Il-1ra. This result confirms the involvement of vagal afferent fibres in the small intestine contractions induced by leptin. It also shows that Il-1β receptors are involved in mediating the effects of leptin. This is consistent with many data on both humans and animals showing that any inflammation, whatever its origin, induces a release of Il-1β and an increase in the plasmatic leptin levels (Barbier et al. 1998; Faggioni et al. 1998; Arnalich et al. 1999; Francis et al. 1999). Moreover, data obtained on rats by Luheshi et al. (1999) have shown that the effects of leptin on food intake and body temperature are mediated by Il-1β. Lastly, Lostao et al. (1998) have described the presence of leptin receptors in the plasma membrane of immune cells located in the lamina propria of the small intestine in rats. These results also confirm the existence of a functional link between leptin and Il-1β.

Lastly, this hypothesis is confirmed by the fact that Il-1β enhances small intestine EMG activity via similar pathways to those taken by leptin. Indeed, the present study clearly shows that intra-arterially administered Il-1β in the absence of CCK has no effect on small intestine spontaneous EMG activity in cats, just like leptin. This finding is consistent with those obtained in several previous studies. Data obtained on rats in vitro have shown that increasing the Il-1β concentrations did not affect the basal jejunal, ileal, or colonic contractility (Aubéet al. 1996), and that the in vivo effects of Il-1β on intestinal motility can be mainly ascribed to a central action (Fargeas et al. 1993). It is noteworthy that Il-1β, intra-arterially injected 20 min after CCK administration, enhances the small intestine spontaneous EMG activity in exactly the same way as leptin. These results indicate that the effects of leptin involve Il-1β, thus confirming the existence of a functional link between leptin and Il-1β, and strongly suggest that Il-1β induces a short-term activation of the vagal afferent nerve fibres.

Intestinal EMG activation induced by leptin via the parasympathetic cholinergic excitatory pathway

Atropine, which is a muscarinic receptor-blocking agent generally used to inhibit motor activity, suppresses the excitatory effects of leptin on intestinal EMG activity when administered locally. This result indicates that the leptin-evoked stimulation takes the parasympathetic cholinergic excitatory pathway. The activation induced by leptin is therefore due neither to the inhibition of the parasympathetic non-adrenergic non-cholinergic pathway, nor to the inhibition of the sympathetic nervous system, which is known to be activated by leptin (Haynes et al. 1997) and to mediate a tonic inhibitory effect on intestinal motility (Gonella et al. 1987). Atropine is also known to block the excitatory effects of CCK on intestinal motility (Milenov et al. 1998). One might therefore hypothesise that the inhibitory effects of atropine on leptin might be due to the lack of effect of CCK. There exist at least two arguments which rule out this hypothesis: (1) as discussed above, the mechanism whereby CCK induces the effects of leptin cannot involve the well-known excitatory effects of CCK on intestinal motility; (2) some excitatory effects of CCK on intestinal motility persisted after vagotomy (Vizi et al. 1973; Milenov et al. 1998) and Il-1ra administration, whereas the effects of leptin were abolished.

The fact that the excitatory effects of CCK on the intestinal EMG activity persisted after vagotomy confirms the presence of peripheral CCK receptors. These receptors have been described in guinea-pigs, where CCK has been reported to activate the intestinal motility in vitro (Vizi et al. 1973; Milenov et al. 1998).

A reflex effect generated by leptin has been described in the literature (Niijima, 1998, 1999). However, the latter author reported that leptin injected into the white adipose tissue of the epididymis evoked reflex activation of the sympathetic nerve activity and abolished the vagus nerve activity, whereas what we observed here was a reflex activation of the vagus nerve activity. Moreover, the latencies of the effects described by Niijima (1998, 1999) were about 1 h, which is far too long to be in keeping with the latencies observed in the present study.

In addition, atropine and vagotomy also abolish the excitatory effects of Il-1β administered after CCK on intestinal EMG activity, which indicates that Il-1β, like leptin, acts via the parasympathetic cholinergic excitatory pathway. This intestinal motor activation does not involve the Il-1β receptors described in the myenteric plexus in rats (Hahn et al. 1998), since the activation of these receptors was found to reduce the release of acetylcholine (Main et al. 1993; Aubéet al. 1996). In addition, these effects involve long latencies (90-150 min; Aubéet al. 1996), which strongly suggests that local Il-1β administration has no short-term effects on small intestine motility in anaesthetised cats.

Functional implications of the leptin-evoked enhancement of intestinal EMG activity

The present findings strongly suggest that the mechanisms whereby leptin combined with CCK enhances small intestine EMG activity involve the vago-vagal reflex circuit. The latency of the effects of leptin is consistent with the presence of a reflex mechanism involving the CNS. It has been suggested that a peripheral integration may take place in the sensory vagal afferent fibres, where a message reflecting the peripheral conditions is developed (Barracchina et al. 1997b; Wang et al. 1997, 2000; Gaigéet al. 2002). Our previous (Gaigéet al. 2002) and present studies show for the first time that the peripheral integration of this sensory information induces reflex effects. Actually, when not associated with CCK, leptin smoothly activates type 1 intestinal vagal afferent fibres and inhibits type 2 fibres, and probably thus informs the CNS about body weight (Gaigéet al. 2002). However, the fact that a post-prandial release of CCK also occurs (Meyer, 1975; Go, 1978; Grider, 1994; Raybould & Lloyd, 1994) indicates that, when associated with CCK, leptin is more directly related to food intake. Under these conditions, type 1 units are in fact strongly potentiated, whereas the inhibition of type 2 units is completely blocked (Gaigéet al. 2002), which increases the intestinal EMG activity. It can thus be concluded that a single substance, namely leptin, can mediate various items of information through a single nerve fibre, in a chemical environment mimicking physiological conditions, i.e. in the presence or absence of CCK. However, these findings do not allow us to determine whether type 1 units alone, type 2 units alone, or both, mediate the effects of leptin on intestinal EMG activity. In any case, the information mediated by these two types of unit is integrated into the CNS. Our results also show that the leptin-specific receptors found to exist in the intestinal intramural plexus and on the interstitial cells of Cajal (Liu et al. 1999) do not seem to be involved in the control of motility.

References

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. References
  7. Acknowledgements
  • Arnalich F, Lopez J, Codoceo R, Jimenez M, Madero R & Montiel C (1999). Relationship of plasma leptin to plasma cytokines and human survival in sepsis and septic shock. J Infect Dis 180, 908911.
  • Aubé AC, Blottiére HM, Scarpignato C, Cherbut C, Rozé C & Galmiche JP (1996). Inhibition of acethylcholin induced intestinal motility by interleukin 1 β in the rat. Gut 39, 470474.
  • Bado A, Levasseur S, Attoub S, Kermorgant S, Laigneau JP, Bortoluzzi MN, Moizo L, Lehy T, Guerre-Millo M, Le Marchand-Brustel Y & Lewin MJ (1998). The stomach is a source of leptin. Nature 394, 790793.
  • Barbier M, Cherbut C, Aube AC, Blottiere HM & Galmiche JP (1998). Elevated plasma leptin concentrations in early stages of experimental intestinal inflammation in rats. Gut 43, 783790.
  • Barrachina MD, Martinez V, Wang L, Wei JY & Tache Y (1997a). Synergistic interaction between leptin and cholecystokinin to reduce short-term food intake in lean mice. Proc Natl Acad Sci U S A 94, 1045510460.
  • Barrachina MD, Martinez V, Wei JY & Tache Y (1997b). Leptin-induced decrease in food intake is not associated with changes in gastric emptying in lean mice. Am J Physiol 272, R10071011.
  • Basmadjian JV, & Stecko G (1962). A new bipolar electrode for electromyography. J Appl Physiol 17, 849852.
  • Berthoud HR, Carlson NR & Powley TL (1991). Topography of efferent vagal innervation of the rat gastrointestinal tract. Am J Physiol 260, R200207.
  • Berthoud HR, Patterson LM, Willing AE, Mueller K & Neuhuber WL (1997). Capsaicin-resistant vagal afferent fibers in the rat gastrointestinal tract: anatomical identification and functional integrity. Brain Res 746, 195206.
  • Bouvier M, & Gonella J (1981). Nervous control of the internal anal sphincter of the cat. J Physiol 310, 457469.
  • Bouvier M, Grimaud JC & Abysique A (1990). Effects of stimulation of vesical afferents on colonic motility in cats. Gastroenterology 98, 11481154.
  • Buyse M, Ovesjo ML, Goiot H, Guilmeau S, Peranzi G, Moizo L, Walker F, Lewin MJ, Meister B & Bado A (2001). Expression and regulation of leptin receptor proteins in afferent and efferent neurons of the vagus nerve. Eur J Neurosci 14, 6472.
  • Faggioni R, Fantuzzi G, Fuller J, Dinarello CA, Feingold KR & Grunfeld C (1998). Il-1 beta mediates leptin induction during inflammation. Am J Physiol 274, R204208.
  • Fargeas M-J, Fioramonti J & Bueno L (1993). Central action of interleukin 1 β on intestinal motility in rats: mediation by two mechanisms. Gastroenterology 104, 377383.
  • Francis J, Mohankumar PS, Mohankumar SM & Quadri SK (1999). Systemic administration of lipopolysaccharide increases plasma leptin levels: blockade by soluble interleukin-1 receptor. Endocrine 10, 291295.
  • Gaigé S, Abysique A & Bouvier M (2002). Effects of leptin on cat intestinal vagal mechanoreceptors. J Physiol 543, 679689.
  • Go VLW, (1978). The physiology of cholecystokinin. In Gut Hormones, ed. BloomSR, pp. 203207. Churchill Livingstone, Edinburgh
  • Gonella J, Bouvier M & Blanquet F (1987). Extrinsic nervous control of motility of small and large intestine and related sphincters. Phys Rev 67, 902961.
  • Grider JR, (1994). Role of cholecystokinin in the regulation of gastrointestinal motility. J Nutr 124, 13341339S.
  • Hahn A, Huber A, Neumayer N & Allescher H-D (1998). Effect of interleukin-1 β on the ascending and descending reflex in rat small intestine. Eur J Pharmacol 359, 201209.
  • Hall KE, El-Sharkawy TY & Diamant NE (1982). Vagal control of migrating motor complex in the dog. Am J Physiol 243, G276284.
  • Haynes WG, Morgan DA, Walsh SA, Mark AL & Sivitz WI (1997). Receptor-mediated regional sympathetic nerve activation by leptin. J Clin Invest 100, 270278.
  • Holzer P, & Holzer-Petsche U (1997). T achykinins in the gut. Part I. Expression, release and motor function. Pharmacol Therapeut 73, 173217.
  • Kewenter J, (1965). The vagal control of the jejunal and ileal motility and blood flow. Acta Physiol Scand 251, 168.
  • Liu M, Seino S & Kirchgessner AL (1999). Identification and characterization of glucoresponsive neurons in the enteric nervous system. J Neurosci 19, 1030510317.
  • Lostao MP, Urdaneta E, Martinez-Anso E, Barber A & Martinez JA (1998). Presence of leptin receptors in rat small intestine and leptin effect on sugar absorption. FEBS Lett 423, 302306.
  • Luheshi GN, Gardner JD, Rushforth DA, Loudon AS & Rothwell NJ (1999). Leptin actions on food intake and body temperature are mediated by IL-1. Proc Natl Acad Sci U S A 96, 70477052.
  • Main C, Blennerhassett P & Collins SM (1993). H uman recombinant interleukin-1β suppresses acetylcholine release from rat myenteric plexuses. Gastroenterology 104, 16481654.
  • Martinez V, Barrachina MD, Wang L & Tache Y (1999). Intracerebroventricular leptin inhibits gastric emptying of a solid nutrient meal in rats. Neuroreport 10, 32173221.
  • Matson CA, Reid DF, Cannon TA & Ritter RC (2000). Cholecystokinin and leptin act synergistically to reduce body weight. Am J Physiol 278, R882890.
  • Meyer JH, (1975). Release of secretin and cholecystokinin. In: Gastrointestinal Hormones, ed. ThompsonJC, pp. 475489. University of Texas Press, Austin
  • Milenov K, Kalfin R, Todorov S & Raichev P (1998). Neuropeptides of the cholecystokinin group: effects and mechanisms of action on the gastro-intestinal and gall bladder motility. Physiol Pharmacol Bulgarica 23, 8591.
  • Miller SM, Schmalz PF, Benarroch EE & Szurszewski JH (1999). Leptin receptor immunoreactivity in sympathetic prevertebral ganglion neurons of mouse and rat. Neurosci Lett 265, 7578.
  • Niijima A, (1998). Afferent signals from leptin sensors in the white adipose tissue of the epididymis, and their reflex effect in the rat. J Autonom Nerv Syst 73, 1925.
  • Niijima A, (1999). Reflex effects from leptin sensors in the white adipose tissue of the epididymis to the efferent activity of the sympathetic and vagus nerve in the rat. Neurosci Lett 262, 125128.
  • Pelleymounter MA, Cullen MJ, Baker MB, Hecht R, Winters D, Boone T & Collins F (1995). Effects of the obese gene product on body weight regulation in ob/ob mice. Science 269, 540543.
  • Raybould HE, & Llyod KC (1994). Integration of postprandial function in the proximal gastrointestinal tract. Role of CCK and sensory pathways. Annals NY Acad Sci 713, 143146.
  • Sobhani I, Bado A, Vissuzaine C, Buyse M, Kermorgant S, Laigneau JP, Attoub S, Lehy T, Henin D, Mignon M & Lewin MJ (2000). Leptin secretion and leptin receptor in the human stomach. Gut 47, 178183.
  • Tansy MF, & Kendall FM (1977). Systemic effects of visceral afferent fiber stimulation. In: Nerves and the Gut, ed. BrooksFP & EversPW, pp. 154196. NJ Slack, Thorofare
  • Vizi SE, Betraccini G, Impicciatore M & Knoll J (1973). Evidence that acetylcholine released by gastrin and polypeptides contributes to their effect on gastrointestinal motility. Gastroenterology 64, 268277.
  • Wang L, Barachina MD, Martinez V, Wei JY & Tache Y (2000). Synergistic interaction between CCK and leptin to regulate food intake. Regul Peptides 92, 7985.
  • Wang L, Martinez V, Barrachina MD & Tache Y (1998). Fos expression in the brain induced by peripheral injection of CCK or leptin plus CCK in fasted lean mice. Brain Res 791, 157166.
  • Wang YH, Tache Y, Sheibel AB, Go VL & Wei JY (1997). Two types of leptin-responsive gastric vagal afferent terminals: an in vitro single-unit study in rats. Am J Physiol 273, R833837.
  • Yuan SY, Costa M & Brookes SJ (2001). Neuronal control of the pyloric sphincter of the guinea-pig. Neurogastroenterol Motil 13, 187198.
  • Zhang Y, Proenca R, Maffei M, Barone M, Leopold L & Friedman JM (1994). Positional cloning of the mouse obese gene and its human homologue. Nature 372, 425432.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. References
  7. Acknowledgements

The authors would like to thank AMGEN (Thousand Oaks, USA) for their generous gifts of leptin, Il-1β and Il-1ra. We appreciate the helpful assistance of Mrs F. Farnarier in the preparation of this manuscript.