The enteric nervous system (ENS) is a network, embedded within the wall of the gut, which takes part in the control of various gastrointestinal functions. To ensure appropriate regulation of the distinct functions, the ENS has to adapt to environmental changes (Giaroni et al. 1999). One such example is to adapt transmission within the ENS by up- or down-regulation of receptor expression and/or transmitter release (Giaroni et al. 1999). In this respect, various studies have investigated changes in gastrointestinal innervation in ageing individuals. In rats, for example, the number of nitrergic myenteric neurons within the small and large intestine decreases with age, whereas the number of cholinergic neurons remains stable (Johnson et al. 1998; Takahashi et al. 2000). Studies in humans have revealed a decreased number of myenteric neurons in older subjects (De Souza et al. 1993; Gomes et al. 1997). By contrast, few data are available for variations in the neurochemical code of enteric cell bodies that are associated with the development of an animal. The forestomach of the sheep is an ideal model for studying plasticity in the neurochemical code, owing to the drastic changes in feeding behaviour of sheep from birth to adulthood. In ruminating sheep, the forestomach serves as a huge fermentation chamber ensuring effective breakdown of herbal food by symbiotic micro-organisms. The rumen is the largest compartment of the forestomach. It expresses specific motility patterns that are distinct from those observed in other parts of the gastrointestinal tract. Ruminal motility primarily serves for mixing of ingesta and does not show peristaltic movements (Ruckebusch, 1989). Forestomach motility is strongly influenced extrinsically by the sympathetic and parasympathetic nervous system as well as intrinsically by the ENS (Ruckebusch, 1989). Enteric neurons influence muscle activity by releasing excitatory and inhibitory neurotransmitters such as acetylcholine, substance P (SP), nitric oxide and/or vasoactive intestinal peptide (VIP) (Vassileva et al. 1978; Denac et al. 1987; Schneider & Eades, 1998). Recently, it has been demonstrated that intrinsic myenteric neurons that innervate the smooth muscle layers of the rumen express a specific neurochemical code (Pfannkuche et al. 2002).

In contrast to adult ruminants, newborn ruminants do not use the rumen as a fermentation chamber. The ingested milk flows directly into the abomasum via the reticular groove (Ruckebusch et al. 1983). Consequently, the rumen remains empty during suckling (Ruckebusch et al. 1983). Motoric and metabolic functions only develop when the ruminant begins to ingest solid food (Ruckebusch et al. 1983).

We hypothesized that these different stages are associated with changes in the innervation pattern of the ruminal wall. Therefore, we studied the neurochemical coding of the myenteric neurons in the rumen during development from the suckling lamb up to the adult sheep. In particular, changes within the proportion of the different neurochemically defined populations containing different excitatory and inhibitory neurotransmitters were quantified. For this purpose, we obtained quadruple immunohistochemical labelling against choline-acetyl transferase (ChAT), nitric oxide synthase (NOS), VIP and SP in the myenteric plexus of the rumen of suckling lambs, fattened lambs and adult sheep. We were able to detect age- and/or nutrient-specific changes in some neurochemically defined populations of myenteric neurons.

To enhance immunoreactivity of the neuropeptides SP and VIP in myenteric neurons, tissues were incubated with colchicine. For that purpose, tissues were pinned out in a Sylgard-covered Petri dish and the mucosa and submucosa were completely dissected off. After several washes in sterile Krebs-Ringer solution, specimens were transferred to the tissue culture Petri dish and cultured for 24 h at 37 °C in culture medium. Culture medium (Dulbecco's modified eagle medium F12) was supplemented with 40 µm colchicine, 10% heat-inactivated fetal calf serum, 100 IU mL⁻¹ penicillin, 100 µg mL⁻¹ streptomycin, 1.24 µg mL⁻¹ amphotericin B, 20 µg mL⁻¹ gentamicin, 2.1 mg mL⁻¹ NaHCO₃ and 1 µm nifedipine, pH 7.4 (all chemicals from CC-Pro, Neustadt, Germany or Sigma, Deisenhofen, Germany). During the culture period, the tissue was continuously agitated using a rocking tray. After organotypic culture, specimens were fixed for 24 h at 4 °C in 0.1 m phosphate buffer containing 4% paraformaldehyde and 15% picric acid. During fixation, the specimens were pinned in a maximally stretched manner. The fixed tissues were washed in 0.1 m phosphate buffer and stored in PBS (0.1 m) containing 0.1% NaN₃.

The tissues were then pre-incubated and permeabilized in PBS containing 4% horse serum and 0.5% Triton X-100. The primary antibodies were diluted in the same solution. The tissues were incubated for 40 h at room temperature in the solution containing primary antibodies. The following antisera were used: rabbit anti-choline acetyltransferase (ChAT, 1 : 1000, P3YEB; Schemann et al. 1993), mouse anti-nitric oxide synthase (NOS, 1 : 40, N31020, Transduction Laboratories, USA), rat anti-substance P (SP, 1 : 1000, 10-S015, Fitzgerald, USA) guinea-pig anti-vasoactive intestinal polypeptide (VIP, 1 : 1000, GHC7161, Peninsula, USA) and rabbit anti-neuron-specific enolase (NSE, 1 : 3000, 16625, Polyscience, USA).

After incubation with the primary antibodies, the specimens were washed three times and incubated for 5 h in buffer solution containing the secondary antibodies. Affinity-purified secondary anti-rabbit, anti-mouse, anti-rat and anti-guinea-pig antibodies raised in donkeys, conjugated to carbocyanine (Cy2), indocarbocyanine (Cy3), biotin or indodicarbocyanine (Cy5) were used (all purchased from Dianova, Hamburg, Germany). The final dilutions of secondary antibodies were 1 : 200 (Cy2 conjugates), 1 : 500 (Cy3 and Cy5 conjugates) and 1 : 50 (biotin conjugates). Biotin-conjugated secondary antibodies were visualized using streptavidin conjugated with aminomethylcoumarin acetate (AMCA) at a dilution of 1 : 50. Finally, the specimens were washed in PBS, mounted on poly-l-lysine-covered slides and coverslipped with a solution of NaHCO₃/Na₂CO₃ (0.5 m, pH 7.0) containing 0.1 NaN₃ and 80% glycerol.

Although ChAT-positive and NOS-positive neurons represent the total number of myenteric neurons in the rumen of adult sheep (Pfannkuche et al. 2002), it is not known whether this is also true for earlier stages of development. Therefore, some ChAT- and NOS-labelled specimens were restained to visualize NSE. The primary rabbit–anti-NSE antibodies were labelled with secondary antibodies conjugated with biotin, which was further visualized using streptavidin–AMCA conjugates.

The preparations were examined using an epifluorescence microscope (IX70, Olympus, Japan). Appropriate filters were applied to visualize the fluorophores separately (Pfannkuche et al. 1998). Pictures were acquired with a black-and-white video camera (Mod. 4910, Cohu, San Diego, USA) connected to a Macintosh computer and controlled by IPLab Spectrum 3.0 software (Signal Analytics, Vienna, USA). Frame integration and contrast enhancement were employed for image processing.

Differences were considered statistically significant at P < 0.05.

Fig. 1 Immunohistochemical coding of myenteric neurons within the rumen of suckling lambs. Each panel shows the same part of the ganglion stained against ChAT (A), SP (B), NOS (C) and VIP (D). Myenteric neurons were either immunoreactive for ChAT/– (white (more ...)

Fig. 2 Immunohistochemical coding of myenteric neurons within the rumen of the ruminating sheep, i.e. fattened lambs (A–D) and adult sheep (E–H). The ganglion was stained against ChAT (A,E), SP (B,F), NOS (C,G) and VIP (D,H). Myenteric neurons (more ...)

The number of ganglia per cm² decreased significantly with age up to the ruminating stages where it remained stable. It was 140 ± 10 (n = 8, N = 5) for suckling lambs, 10 ± 3 cm⁻² (n = 7, N = 7) for fattened lambs and 7 ± 2 cm⁻² (n = 3, N = 3) for adult sheep.

ChAT-positive neurons were either only positive for ChAT but not for another of the antigens tested (further referred to as ChAT/–), or the neurons where additionally immunoreactive for SP (further referred to as ChAT/SP). All of the NOS-positive neurons were immuno-reactive for VIP (further referred to as NOS/VIP).

Triple labelling against ChAT, NOS and NSE was used to address the question of whether ChAT and NOS represent the total entity of myenteric neuronal cell bodies. In all experimental groups, the NSE-positive myenteric neurons (500 neurons counted in each preparation) were either immunoreactive for ChAT or for NOS.

Ruminal myenteric ganglia of suckling lambs mainly consisted of ChAT/SP immunoreactive neurons (Tables 1 and 2; Fig. 1). All other myenteric neurons belonged equally to the populations ChAT/– or NOS/VIP (Tables 1 and 2; Fig. 1).

Table 1 Absolute number of neurochemically defined myenteric neurons per ganglion

Table 2 Relative number of neurochemically defined myenteric neurons per ganglion

Within the group of fattened lambs and of adult sheep the relative size of all three subpopulations differed significantly (ChAT/SP > NOS/VIP > ChAT/–) (Tables 1 and 2; Fig. 2).

Compared with the suckling lambs, in fattened lambs and adult sheep, the absolute number of neurons in both cholinergic populations was lower (Table 1). Additionally, ganglia of the adult sheep contained a significantly smaller proportion of ChAT/– neurons than ganglia of suckling or fattened lambs (Table 2). In contrast to the ChAT/– encoded population, the relative number of ChAT/SP neurons did not change from the suckling lamb up to the adult sheep (Table 2).

Although the absolute number of NOS/VIP neurons remained stable from suckling lambs up to the adult sheep (Table 1), the relative number of nitrergic neurons was higher in the ruminating sheep (i.e. fattened lambs and adult sheep) than in the suckling lambs (Table 2).

This study revealed differences in the neurochemical code of myenteric neurons within the ovine rumen. The differences were strongly associated with the development of the rumen. Although the same three neurochemically defined subpopulations of myenteric neurons could qualitatively be detected in all experimental groups, quantitative differences between ruminating and non-ruminating sheep became obvious. Apart from quantitative changes in the neurochemical code of the myenteric neurons, notable changes in the number of ganglia per cm² and in the number of neurons per ganglion were revealed. The ganglion density decreased from the suckling lamb up to the adult sheep. A correlation between body size and number of ganglia per cm² has already been shown between different species (Gabella, 1971). Age and innervation density might also correlate, as shown for humans (Wester et al. 1999).

Comparing myenteric ganglia from suckling lambs with both groups of ruminating sheep, a reduction in the number of myenteric neurons could be detected (Table 1). This observation is in accordance with studies in cattle by Kitamura et al. (1986), showing that nerves were more abundant within the reticulorumen of calves than in the reticulorumen of adult cattle. Findings obtained from other species and localizations in the gastrointestinal tract indicate an age-dependent loss of myenteric neurons (Gabella, 1989; Giaroni et al. 1999; Wester et al. 1999). In our study, however, age-dependent alterations did not occur consistently in all neuronal subpopulations examined. A decrease in neuronal number per ganglion between pre-ruminating and ruminating sheep was found only in cholinergic populations. But the number of neurons did not decrease equally. The dissimilarity in the change of absolute number led to alterations in the relative size of the neurochemically defined subpopulations. The relative size of the population expressing ChAT/– immunoreactivity decreased, whereas the proportion of the ChAT/SP neurons remained stable and the proportion of the NOS/VIP neurons increased. The relative increase in the number of NOS/VIP immunoreactive neurons from the pre-ruminating to the post-ruminating stage is based on the fact that the size of this population remained equal as far as the absolute number of neurons per ganglion was concerned. This finding is in accordance with studies of the gastrointestinal tract of rats from the suckling period up to adulthood (Matini et al. 1997; Timmermans et al. 1999). Matini et al. (1997) found a stable number of myenteric nitrergic neurons per mm² of the myenteric plexus in all ages examined. Additionally, a relative increase in the number of nitregic myenteric neurons from suckling to adulthood was detected. The increasing number of NOS/VIP immunoreactive neurons in ruminating sheep suggests that the rumen of ruminating sheep is more nitrergic and VIPergic controlled than the rumen of pre-ruminating animals. Generally, nitric oxide as well as VIP are known inhibitory neutrotransmitters acting on the ruminal smooth muscle layers (Denac et al. 1987; Schneider & Eades, 1998). In the rumen of adult sheep, neurons immunoractive for NOS/VIP project to both circular and longitudinal muscle (Pfannkuche et al. 2002). Providing the rumen with an increasing inhibitory innervation at the onset of rumination seems to be meaningful from a teleological point of view because the rumen has then to adapt to large amounts of ingesta and to increase its volume by growth as well as by dilatation (Ruckebusch, 1989).

In contrast to the NOS/VIP population, the size of the ChAT/– population decreased absolutely as well as relatively from suckling lambs to ruminating sheep. This decrease may be due to a specific loss of purely cholinergic neurons. ChAT/– immunoreactive neurons do not project directly to the smooth muscle in the rumen of adult sheep (Pfannkuche et al. 2002). It has been suggested that they have projections to the epithelium or to other myenteric ganglia (Pfannkuche et al. 2002). However, the meaning of a possible loss of this putative projection still seems unclear in the rumen as this organ serves its main functions in the ruminating sheep (Ruckebusch, 1989).

In summary, this study shows, for the first time, age- and development-associated plasticity of ruminal myenteric neurons. The observed changes in populations with a different neurochemical code may be due to the adaptation of the rumen to drastic changes in feeding habit during development.

The skilful technical assistance of Petra Philipp is gratefully acknowledged. This study was supported by Deutsche Forschungsgemeinschaft Pf 403 1-1 and Pf 403 1-2 and Akademie für Tiergesundheit (AfT).