Histochemistry (1986) 85 :441-452 Histochemistry 9 Springer-Verlag1986 Chromogranin A (CGA) in the gastro-entero-pancreatic (GEP) endocrine system I. CGA in the mammalian endocrine pancreas D. Grube 1 *, D. Aunis 2, F. Bader 2, Y. Cetin 1, A. Jfirns 1, and S. Yoshie 3 1 Abteilung Anatomic I der Medizinischen Hochschule, Postfach 6101 80, D-3000 Hannover, Federal Republic of Germany 2 INSERM, Unit6 de Neurochimie, Strasbourg, France, 3 Department of Anatomy, Nippon Dental University, Niigata, Japan Accepted July 19, 1986 Summary. Chromogranin A (CGA), a protein at first de- tected in the adrenal medulla, has recently been found also in other organs, e.g. the endocrine pancreas. However, im- munohistochemical findings concerning the cellular source of pancreatic CGA were controversial. Therefore, the endo- crine pancreas of 10 mammalian species (man, tupaia, mole, cat, dog, pig, guinea pig, rabbit, rat) was investigated im- munohistochemically for CGA-like immunoreactivities on serial semithin plastic sections using a high-titer polyclonal antiserum against bovine CGA. The results show that basi- cally all pancreatic endocrine cell types are CGA-immuno- reactive; however, every species has its own pattern of CGA-immunoreactive cell types. Other findings of the pres- ent studies indicate that the physiological function of CGA in pancreatic endocrine cells is related to the storage mecha- nisms of peptide hormones. Finally, a methodological ap- proach is given to obtain not only qualitative but also semi- quantitative data during immunohistochemical investiga- tions. Introduction The endocrine cells of the pancreas - like other members of the" APUD cell series" or the "paraneurons" - in addi- tion to peptides or amines contain a variety of substances which in part also serve as cellular markers of these cellular communities (ATP and other nucleotides, enzymes like neu- ron specific enolase etc. ; for reviews see: Pearse 1969, 1980; Fujita 1977, 1983; Grube 1982). Recently, an additional substance obviously common to the majority of the diffuse neuroendocrine system has been identified, chromogran- in A. This protein is a large acidic protein which originally has been detected in the adrenal medulla (Blaschko et al. 1967) where it is co-stored and co-secreted with the corre- sponding amines and peptides. Now, this protein has been found immunochemically and immunohistochemically in many locations outside the adrenal medulla: in nervous elements (Somogyi et al. 1984; Nolan et al. 1985), in lym- phatic organs (Hogue Angeletti and Hickey 1985), and in various endocrine organs and cells including the endocrine pancreas (O'Connor 1983; O'Connor et al. 1983 a, b; Lloyd and Wilson 1983; Cohn et al. 1984; DeStephano et al. 1984; Lloyd etal. 1984, 1985; Faser etal. 1985; Nolan etal. * To whom offprint requests should be sent 1985). Concerning the endocrine pancreas, previous immu- nohistochemical findings related to the cellular source of CGA have been controversial. Only certain endocrine cell types or all pancreatic endocrine cells have been found to exhibit CGA-like immunoreactivities. Moreover, the data given for the same species differ from one author to the other. These discrepancies may be caused by several factors: types and specificities of the antisera used (at present 8 dif- ferent antisera against CGA are applied in immunohisto- chemical studies of CGA); processing of the tissues and immunohistochemical techniques employed; interspecies variations of CGA-like immunoreactivities. The present investigation, therefore is designed to eluci- date the cellular localization of CGA in the endocrine pan- creas, taking into account the factors mentioned above. During our studies the endocrine pancreas of 10 mamma- lian species has been investigated by light-microscopical im- munohistochemistry performed on serial semithin plastic sections using a high-titer specific antiserum against bovine CGA. Our findings demonstrate that basically all endocrine cell types are CGA-immunoreactive even though in the same species not all endocrine cells show CGA immuno- reactivities. In addition, another important result is related to the "optimal" dilutions of the antisera used as first layer in the immunohistochemical staining protocol. It will be shown that a submaximal dilution of the antisera is appro- priate to obtain not only qualitative but also semiquantita- rive data in immunohistochemistry. Material and methods Tissues and tissue peparation During the present study pancreata of 10 mammalian species were investigated including man (n =4), tupaia (n = 3), pig (G6ttingen minipig, see Beglinger et al. 1975; n : 1), dog (n = 2), cat (n = I), mole (n=2), cattle (n =4), guinea pig (n = 2), rabbit (n = 2), and rat (n=2). Small tissue specimens were quenched in melting Freon 22 precooled with liquid nitrogen, freeze-dried, and fixed by vapor-phase p-formaldehyde. Specimens from the guinea pig pancreas were also fixed by vapor-phase di-ethylpyrocarbonate (= DEPC, see Pearse and Polak 1975). Subsequently, the specimens were immersed in propylene oxide and embedded in epoxy resin (araldite). Adrenals from man, cattle and rat, and specimens from the gastrointestinal mucous membrane of man, dog, mole, guinea pig and rat served as reference organs and were treated in the same way. In addition, tissue samples from the human pancreas