Roe deer and cattle (including calf) belong to Cervidae and Bovidae families respectively, both are members of the Ruminantia suborder. One aim of this work was to evaluate in Roe deer the adaptation to diet in relation with animal domestication. Thus we have characterized the pancreas of the Roe deer as well as its content in pancreatic enzymes to compare data obtained in Roe deer with those collected in calf. When data are expressed in relation to BW, the pancreas and pancreatic protein weights are higher in Roe deer than in calf as well as most enzyme activities measured. Moreover, both receptor subtypes of the CCK/gastrin family peptides are present in the Roe deer’s pancreas like in the calf’s. The consequences for digestive processes and their regulation are suggested in the following discussion.
GIT morphology and physiology
Pancreas weight was higher in ruminant animals than in weaning or milk-fed animals, as previously shown in calves and sheep . This difference however disappeared when the organ weight was expressed as a percentage of BW in calves but not in Roe deer. In milk-fed and in weaned calves aged from 138 to 156 d, the pancreas weight expressed as percentage of BW is similar to the value obtained in adult (0.78 g/kg BW, ). In contrast, for the young Roe deer, this value was higher than in adult. To explain these differences, it would be interesting to have histometrical data (number of cells per acinus, cell area, etc.) to determine whether these variations come from pancreas structure and/or higher secretory capacities.
Compared with ruminant calves, the GIT of young (weaning) and adult (ruminant) Roe deer is characterized by several particularities. Roe deer have a selective feeding behaviour with a high proportion of soluble plant cell contents in their diet [10, 11]. Moreover, a number of morphological and physiological particularities of their GIT were underlined; they exhibit a relatively small forestomach with large orifices between the different sections, a short retention time of ingesta, relatively large salivary glands producing high volumes of saliva, and a highly developed reticular groove which is retained in the adults [4, 22, 23]. Considering these digestive particularities, we can speculate that Roe deer have incomplete fermentation in their rumen, a ruminal escape of digesta components (starch, fats, etc.) and/or rumen bypass of rapidly digestible components (soluble sugars, etc.) from the oesophagus to the abomasum down to the reticular groove. These anatomical and physiological observations suggest that a part of their diet components can reach the abomasum and then the intestinal lumen and be digested according to the mechanisms described in monogastric animals. These observations justify that milk-fed calves be introduced in our experiments since in the GIT of this animal, milk bypasses the rumen thanks to the reticular groove closure, as it is the case for a part of the Roe deer’s diet. In such cases, enzymatic digestion plays a much more important role than in ruminant calves.
Another morphological and physiological particularity of the GIT compared with calf is that Roe deer have a shorter intestine (relative to BW) and a smaller ratio between small and large intestine (70:30 vs 82:18) . In these conditions, even the digestion is important in the small intestine in both animals, that of the Roe deer could be less effective, at least for starch and sugars, due to less intestinal enzymes and a shorter transit time. But the enzyme activities measured in the pancreas were much greater in Roe deer than in calf and this could compensate this intestinal deficiency in enzymatic digestion.
In Roe deer, pancreatic enzyme activities were seldomly studied and references on this subject are very scarce. In this study, we have compared Roe deer and calf digestive enzymes also because of available data on domestic ruminant animals [5, 6]. In Roe deer, Rowell-Schäfer et al  reported, we believe, the only data available on pancreatic α-amylase in comparison with those measured in sheep. In agreement with our data, they showed that α-amylase activity was about twice that found in the domestic ruminant animal. They also investigated intestinal disaccharidase activities and found them higher in Roe deer than in ruminant sheep and goats. In the current experiment, with regards to lipase in calf, it seems that there was an adaptation to lipid level since lipid level in diet ingested by Milk-fed animals was higher than in solid diet consumed by weaned calves. By contrast there was no adaptation in the Roe deer species. Also, this relation did not stand true for proteolytic enzyme activities and protein levels in the diets.
Thus, proteolytic enzyme activities were largely higher in Roe deer than in calves, mainly for chymotrypsin and elastase II. This observation could be in relation to particular digestive products observed in Roe deer. Among these products, Roe deer secrete about 9–14 times more saliva (6.7-13.5 vs 0.5-1.5 ml/10 min/kg BW) than sheep . Saliva usually contains high amounts of proline-rich proteins (PRPs) whereas their levels are relatively low in calf and sheep saliva. Furthermore, pancreatic chymotrypsin contents in adult Roe deer are 150-fold higher than in ruminant calf. Of interest is the fact that chymotrypsin cleaves the amino-acid bonds in proteins where proline is present. Moreover, elastase II specifically hydrolyses peptide bonds involving highly hydrophobic amino acids  which are abundant (30 %) in the first part of the primary structure of PRPs . These phenomena could allow the Roe deer to reuse its PRPs.
Interestingly, the Roe deer prefer tannin-rich plants which provides a nutritional benefit for the animal [11, 27]. When present in diets of ruminants, tannins generally reduce growth performance and increase the excretion of nutrients in ruminants. Moreover, in human diets and in those of non-ruminant animal species, tannins can reduce the digestibility of proteins, carbohydrates and minerals; they can also result in lower digestive enzyme activities and cause damage to the mucosa of the GIT . By contrast, the presence of tannins in the diet at a low dose could result in the formation of tannin-protein complexes which can enhance protein assimilation by avoiding microbial breakdown of protein in the rumen. Because PRPs bind tannins, they thus reduce their deleterious effects on protein digestion in Roe deer for which the presence of tannins in the diet is considered to have benefit effects.
In bovine species, the specific enzymatic activities of trypsin, elastase 1 and amylase were higher in ruminant than in milk-fed calves; these changes illustrate the modifications in their diet composition . In contrast, there was no modification in the corresponding values for chymotrypsin, elastase II and lipase. However, in milk-fed calves, lipid intake declines after weaning. In Roe deer, few differences were observed between young and adult animals in their pancreatic enzymes with only chymotrypsin specific activity being much lower in young Roe deer than in adults. Even if these young animals may have been at least partially ruminants, this phenomenon could be linked to a lesser ingestion of tannin-rich plants and consequently to a lesser production of PRPs in saliva in the young.
CCK/gastrin family receptors
Until now and to the best of our knowledge, no data have been reported on the gut regulatory peptides and more particularly those on the CCK/gastrin family, as well as their receptors in Roe deer pancreas . In this study, the immunohistochemical findings illustrate, for the first time, the presence of both CCK-1 and CCK-2 receptor subtypes in the Roe deer pancreas. These results are well supported by the western blot presented which also clearly indicates the presence of both subtypes in tissue homogenates. The molecular weights of both subtypes, around 90 kDa for the CCKR-1 and 75–80 kDa for the CCKR-2 agrees with those previously found in rat pancreatic cell membranes (CCKR-1) and in human and rat pancreatic homogenates (CCKR-2) [19, 30]. The differences in molecular weight between species may depend on the glycolysation status of each protein. These two receptors have different affinities for CCK and gastrin. The CCKR-1 subtype exhibits a high affinity for sulphated CCK and a 1,000-fold lower affinity for gastrin, whereas the CCKR-2 subtype interacts with gastrin and CCK with almost the same affinity and poorly discriminates sulphated and non-sulfated peptides . Even though we cannot compare hormone data between species in this study, we can ascertain the presence in the pancreas of both types of CCK receptors. This suggests that the corresponding regulatory peptides could act in the regulation of the development and production of the Roe deer pancreas as well as their secretory response to hormone stimulation.
The CCKR-1 has been localized in the pancreas β- and α-cells of several species, like the mouse, rat, human and pig  and the calf . The presence of the CCKR-1 in islet cells of Roe deer pancreas suggests a direct action of CCK on the regulation of islet cell secretion. This effect has been suggested by others scientists [30, 32] who localized CCKR-1 in islet cells. In other species, CCKR-1 was also localised in pancreatic endocrine cells out of islet. In the rat pancreas, immunofluorescence confocal microscopy studies and in situ hybridization studies localised CCKR-1 in both insulin- and glucagon-producing cells. In pig and human pancreas, immunofluorescence and confocal microscopy studies support the expression of CCKR-1 in glucagon secreting cells only . In Roe deer, we have observed CCKR-1 in endocrine cells but not in acinar cells; therefore, we can suggest that in this species CCK may also play an indirect role on the exocrine pancreas, like in pig and rat. For instance, some authors [34, 35] have suggested that in the pig, CCK could indirectly stimulate the exocrine pancreas by a mechanism mediated locally in the small intestine. Furthermore, the cytoplasmic processes of endocrine cells found positive for CCKR-1 were in close contact with acinar cells. This particular morphological feature suggests that these endocrine cells may control the functions of other cells via paracrine secretion consequently to CCK stimulation.
Morisset et al.  have demonstrated the localization of CCKR-2 protein in the islet’s δ-cells which secrete somatostatin and its unique presence cells of the calf and cow. However, Saillan-Barreau et al  reported the major site of CCKR-2 expression in glucagon cells in adult and fetal human pancreas. This does not occur in pancreas of Roe deer, since CCKR-2 were only detected in the tunica intimae of several arterioles, suggesting a possible important role of CCK upon the regulation of blood flow in pancreas. Indeed, in Sprague–Dawley rats, CCK mediated vasodilatation of the mesenteric vascular bed via its pre-synaptic CCKR-2 release . Similarly caerulein, a CCK analogue, was shown to increase pancreatic blood flow and vascular conductance in rat via activation of CCKR-2 receptors .