Enteroendocrine cells of the cyprinid fish, Barbus conchonius

Information on the endocrine regulation of digestion in fish is scarce especially on stomachless cyprinids. In the present study (chapters I, III) 3 distinct enteroendocrine cell types will be described for the intestinal epithelium of Barbus conchonius. With the light microscope, enteroendocrine cells only stained moderately after some argyrophil reactions; therefore, the distinction is mainly based on the size of the basally located secretory granules. Cell type I (small granules) is distributed throughout the intestine, but with the highest frequency in the third segment; cell type II (intermediate granules) is mainly present in the first segment, whereas most of the cells of type III (large granules) are found in the intestinal bulb.Most if not all enteroendocrine cells are of the "open" type, which has a dendrite-like process to the intestinal lumen. The ultrastructure of the apical part (pinocytotic vesicles, cilium, microtubules) indicates that these cells probably have a chemo- receptory function. In contrast to the strongly innervated pancreatic endocrine cells, nerve endings are not found in the vicinity of the enteroendocrine cells. Consequently, the gut endocrine cells seem to be self-sufficing in their functioning; they receive adequate stimuli at their apical end, that activate or inhibit the basal granule release. However, a paracrine impact from one cell to another cannot be excluded.Comparison of enteroendocrine cells with pancreatic endocrine cells (chapter II) revealed only one common cell type for gut and pancreas, viz. cell type III resembles the pancreatic A 2r cell. Therefore, cell type III is probably involved in the secretion of a pancreatic hormone, possibly a glucagon-like immunoreactive peptide (GLI, formerly enteroglucagon) or, as discussed in chapter III, pancreatic polypeptide (PP). In contrast with mammals, the D (= A 1 ) cells of fish, which are assumed to produce somatostatin, are abundantly present in the pancreatic islets but not in the intestinal epithelium.The possible functions of the two other cell types is discussed in chapter III. Cell type II is probably involved in CCK-PZ secretion. As this type contains intermediate granules the indication "I" cell, used for mammals, may be maintained for this cell type. The location throughout the intestine suggests that cell type I produces a hormone that controls gut motility. As serotonin cannot be demonstrated in the intestinal epithelium of fish, the hormones motilin, vasoactive intestinal peptide (VIP), and neurotensin may be considered for this function.As a consequence of the absence of peptic digestion and multicellular intestinal glands, the digestive tract of cyprinids contains a relatively simple endocrine regulatory system, in which only 3 distinct enteroendocrine cell types can be recognized. Therefore, hormones directly or indirectly related to the presence of a stomach cannot be expected, such as gastrin, histamin, gastric inhibitory peptide (GIP) and secretin. Moreover, the evolution of the endocrine system in fish may be less developed; consequently some "primitive hormones" resembling two or more mammalian hormones might be expected.Neither serotonin nor catecholamines can be demonstrated in the enteroendocrine cells of adult fish, even not with amine precursors (5- HTP or L-DOPA); consequently, the APUD characteristic ( A mine P recursor U ptake and D ecarboxylation) is absent in these cells of adult specimen (chapter I). On the other hand, APUD cells appear to be present in the intestinal epithelium from day 3 untill day 6 of development (chapter IV). This short-time APUD facility has probably to be considered as a rudiment of a neural origin of the enteroendocrine cells. The present study (chapter V) shows that a neural crest origin is hardly possible, and presumptive enteroendocrine cells are supposed to migrate in the early developmental stages from the neurectoderm. It is not certain whether a similar origin can be suggested for all enteroendocrine cell types. Particularly for type III (= A 2r ) cells, a neurectodermal origin is hardly conceivable, as their granules are found in intermediate cells of the pancreas, which contain both exocrine and endocrine granules (chapter II).The absence of the APUD characteristics as from the larval stage onwards is in contrast with the presence of this facility in gastro-enteric endocrine cells of birds and mammals. This may be explained by assuming that granulecontaining enteroendocrine cells are able to proliferate (chapter VI). Hence differentiation from stem cell to enteroendocrine cell may occur only during embryonic development, and the APUD facility must possibly be considered as a differentiation characteristic. Whether such mature-type enteroendocrine cells proliferate in or outside the epithelium is not yet known.In chapter VI it is shown that the turn-over time of the enteroendocrine cells is considerably longer than that of absorptive cells. Thus, the moving upward into the folds must be much slower than of other enterocytes. 'This might be attributed to a stronger adhesion of the enteroendocrine cells to the basement membrane.The present study can only give assumptions with respect to function, origin and renewal of the enteroendocrine cells of a cyprinid species. Additional experiments remain to be done to provide further information.