All the mechanoreceptors investigated in the present study were in series muscle mechanoreceptors, as confirmed by the fact that they responded to a moderate distension of the intestine (Grundy, 1992). They were probably tension receptors located in the intraganglionic laminar ending (Phillips & Powley, 2000).
Leptin and food intake
The present results show that leptin can affect the spontaneous discharge patterns of numerous (71%) vagal chemosensitive mechanoreceptors in the intestine in two ways: by significantly increasing the discharge frequencies of 68 % of the vagal afferent nerve fibres (type 1 units) and significantly decreasing those of the remaining 32 % (type 2 units). These results are consistent with the in vitro data published in rats, showing that the effects of leptin on the vagal afferent nerve fibres innervating the stomach are also of two kinds (Wang et al. 1997). Based on the effects of leptin, three different types of intestinal vagal mechanoreceptors were identified (type 1, type 2 and leptin-insensitive units). Units of all these types were present in the same animal. The three types of neurones involved cannot be differentiated in terms of the basal discharge frequencies, which were all identical. However, the differences between their responses to mechanical stimulation as well as to CCK administration indicate that three different populations of mechanoreceptors were involved. Both type 2 neurones and leptin-insensitive neurones are strongly sensitive to distension, but the former are CCK sensitive as well. Also strongly CCK sensitive are type 1 neurones which, however, are less distension sensitive than the two other categories. The presence of two populations of mechano- and chemosensitive neurones was confirmed by the fact that SP and PBG activate type 1 but not type 2 units. Leptin and CCK insensitive units might be purely mechanoreceptors, whereas both type 1 and type 2 units are mechanosensitive receptors, which belong to two different populations having different functions.
Although CCK always induced an increase in the discharge frequency of both type 1 and type 2 units, these effects differed in their intensity and their duration between these two populations of units. This finding supports the existence of two different populations. However, the latency of the effects of CCK was similar with both types of units, which indicates that CCK affects vagal sensory neurones via a similar mechanism (namely, a direct activation via CCK receptors).
The fact that all the mechanoreceptors in our study were connected to type C fibres indicates that all the neurones studied have small cell bodies (about 25 μm in diameter) (Harper & Lawson, 1985) and they cannot therefore be differentiated in terms of their morphology.
The two types of afferent nerve fibres cannot be differentiated in terms of their basal discharge frequencies, which are identical. However, since the type 1 and type 2 units responded in significantly different ways to CCK, they seem to belong to two different populations. This assumption was confirmed by the fact that only the type 1 units were activated by both SP and PBG. Previously published data have indicated that the percentage of vagal chemosensitive mechanoreceptors responding to one or several chemicals varies (Mei et al. 1996).
The short latency of the effects of leptin strongly suggests that this substance acts directly on the sensory endings. It is worth noting that this latency is similar to that recorded in the case of CCK, with which a direct effect may also occur (Mei, 1978; Mei et al. 1996).
The methods employed here were identical to those previously described by other authors for studying the discharge frequencies of vagal afferent units (Mei et al. 1996). The effects observed here were not due to arterial distension associated with drug injection (1 ml within 5 s), since the discharge frequencies were not affected when vehicle alone was injected. Although some substances are able to activate the cell bodies of vagal afferent neurones directly (Dun et al. 1991), the effects of leptin observed here were definitely of peripheral origin. As a matter of fact, these effects were completely abolished upon sectioning the ipsilateral vagus nerve caudally to the ganglion. The peripheral action of leptin has been confirmed by the lack of any systemic effects of the drug on cardiac frequency and arterial pressure (Haynes, 2000). In addition, the fact that leptin injection resulted in no changes in the intestinal intraluminal pressure indicates that these effects are probably not due to motor activation. Moreover, atropine, which is a muscarinic receptor blocking agent generally used to inhibit motor activity, failed to modify the effects of leptin. Our results are in keeping with those obtained by Mei et al. (1996), showing that most of the vagal mechanoreceptors in the intestine (71 %) are chemosensitive. But the fact that some of them (29 %) are not affected by the substances tested indicates that these mechanoreceptors either are not sensitive to endogenous substances, or are not chemosensitive at all. All the mechanoreceptors responding to leptin are spontaneously active. Identical data have been published on the stomach vagal afferent neurones (Wang et al. 1997). We have observed that leptin injection never results in activation of previously silent units, indicating that the mechanoreceptors activated by leptin are only low-threshold muscle receptors (Grundy, 1992). Moreover the disappearance of spontaneous activity after sectioning of vagus nerve caudally to nodose ganglion confirms the peripheral origin of the activity recorded in our experiments.
These results showing that CCK can activate intestinal vagal mechanoceptors are in agreement with previously published data (Blackshaw & Grundy, 1990; Mei & Lucchini, 1992). As established above, the present effects are peripheral sensory effects, since the activation of afferent neurones always occurred prior to the motor effects, as described by Mei et al. (1996). These authors have also reported that the sensory effects of CCK persist under atropine. Although the activatory responses of type 1 units to CCK are significantly larger in amplitude and duration to those of type 2 units, the similarity between their latencies strongly suggests that a similar mechanism of activation is involved. In 15 experiments, leptin was administered both before CCK (first injection) and 20 min after CCK (second injection). The effects induced by the second injection differed significantly from those observed after the first one. In type 1 units, the effects of leptin are enhanced by CCK, whereas in type 2 units, they are inhibited by CCK. This is not a tachyphylaxic phenomenon, because two injections of leptin given 20 min apart have identical effects on the vagal mechanoreceptors. The enhancement of leptin by CCK was not due to motor activation, which had recovered its basal level before the leptin injection was performed. Concerning the mechanism whereby CCK enhanced the effects of leptin, this enhancement cannot have been due to the effect of CCK on the basal activity, because leptin was injected 20 min after CCK, the effects of which last approximately 8 min. The discharge frequency of the mechanoreceptors was therefore similar before injection of leptin alone and before injection of leptin 20 min after CCK. Indeed, although gastric leptin has been found to be rapidly released from the fundic mucosa in response to feeding, data in the literature indicate that, in rats in vivo, gastric leptin is released into the blood compartment 15 min after perfusion of the stomach with CCK (Bado et al. 1998). This time is consistent with the effects we have observed. These effects might result from an increase in the concentration of circulating leptin. This finding supports the idea that plasma leptin may affect the vagal intestinal afferents, but we cannot rule out the possibility that gastric leptin secreted within the fundic mucosa may diffuse into the intestinal mucosa in the vicinity of vagal afferent nerve endings (Wang et al. 2000). Changes in the effects of leptin on vagal mechanoreceptors occurring in response to CCK have also been described in the stomach (Wang et al. 1997), but these changes involve different modes of interaction. CCK does not affect the leptin-induced activation of the vagal mechanoreceptors in the stomach, but those which are not initially affected by leptin become leptin sensitive after CCK. This difference does not seem to be due simply to differences in the doses injected or the times elapsing between CCK and leptin injections. It probably reflects the fact that the patterns of innervation differ between these organs. SP and PBG, which are routinely employed for activating sensory visceral receptors (Paintal, 1973; Mei et al. 1996), activate type 1 units but do not alter the activatory effects of leptin, which indicates that leptin/CCK interactions are actually specific ones. Previous authors have described the relationships between leptin and CCK, and also, in the case of some effects, the need for the simultaneous occurrence of the two substances (see Wang et al. 2000; Lewin & Bado, 2001). In mice, intraperitoneal injections of these two substances induce an early satiety signal. But when delivered separately, the same doses have no effect (Barrachina et al. 1997). In mice, the effects of leptin on the loss of body weight are also enhanced by CCK (Matson et al. 2000). Upon studying the expression of c-fos protein in the hypothalamus in the same species, Wang et al. (1998) observed similar enhancing effects. These authors reported that simultaneous injections of CCK and leptin increased the number of neurones activated in the paraventricular nucleus by about 50 % in comparison with CCK alone.
As postulated by Wang et al. (1997, 1998), and Buyse et al. (2001), post-prandially released CCK and circulating leptin might have combined effects on vagal afferent fibres and thus generate a satiety signal to the central nervous system. Depending on its plasmatic level, leptin may therefore modulate the combination of messages arising from the stomach and intestine and thus optimize the satiety signal.
Leptin and inflammation
Our results show that injecting Il-1ra, an endogenous peptide which is a specific Il-1β receptor antagonist, inhibits the excitatory response to leptin. This is consistent with data obtained on rats by Luheshi et al. (1999), showing that the effects of leptin on food intake and body temperature involve Il-1β. As a matter of fact, intracerebroventricular injection of Il-1ra inhibits the cessation of food intake and the increase in the body temperature induced by leptin. In addition, in mice deprived of Il-1β, no decrease in food intake occurred after leptin administration (Faggioni et al. 1998). Moreover some studies on rodents with genetic abnormalities as regards leptin production or leptin receptor synthesis have shown that a deficit occurs in the level of the animals proinflammatory cytokine expression (Loffreda et al. 1998). On the other hand, numerous data in the literature point to the existence of a correlation between the release of interleukin-1β and the plasmatic leptin levels. In humans as well as in animals, most of the studies available so far have shown that any inflammation, whatever its origin, induces a release of Il-1β that results in an increase in the plasmatic leptin level (Barbier et al. 1998; Faggioni et al. 1998; Arnarlich et al. 1999; Francis et al. 1999). 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.
Our results also confirm that leptin is involved, via intestinal vagal afferent fibres, in the mechanisms responsible for inflammation, and probably for a process with which it is generally associated, namely anorexia (Sarraf et al. 1997; Gualillo et al. 2000).
One of our noteworthy results involves the effects of leptin, which were found to depend strongly on Il-1β receptors. The possible existence of a functional link between leptin and Il-1β should therefore be taken into account when dealing not only with the control of immune responses but also with the control of food intake.