XOLURTLITY STUDIES O F THE SE(>RETIOY GBANULES O F THE GUINEA PIG PANCREAS 8. H. BENBLEY Hull Laboratory of Anatoniy, 1 A e University of Chicago Woerner ('38) has described, for the first time, deg 0 ciierative changes in the alpha cells of the island of Langcrlians, following continuous intravenous injection of dextrose i n guinea pigs. These changes occurred when moclerate aiuounts of sugar were injected and there was no resulting extensive hyperglycemia or glycosuria. However, these aiiirnals exhibited evidences of increased metabolic rate with extensive loss of weight aud there were coincident f a t t y changes in the mitochondria of t,he acinar cells of the pancreas. It occurred to us o n thc basis of these findings that there might be some relation between exliaustion of the alpha cells and increased fat metabolism under conditions of increased oxidation of sugar. In order to test this hypothesis it seemed desirable to obtain an extract of the alpha cells. Consequeiitly the following solubility studies were undertaken. 1. The pancreases froin three guinea pigs, whose pancreatic ducts had bccii ligated for 3 weelis, were frozen i n liquid isoperitaiie and dehydrated iii ~ a c u oa t -40°C. Pieces of this tissue were treated with twelve organic aiid ninc aqueous solrents and one piece TI-^ put directly into formol-Zenker for a control. The organic solvents used were: methyl alcohol, ethyl alcohol, butyl alcohol, iso-butyl alcohol, propyl alcohol, iso-propyl alcohol, iso-amyl alcohol, ally1 alcohol, ethylene dichloride, etliyleiie glycol, methyl cellosolve, and butyl cellosolvc. 131 THI A N A T O M I C A L XECORD, VOL. 72, NO. 2 OCTOBER, 1938 132 S. H. BENSLEY The tissue in the organic solvents was subjected lo negative pressure, t o extract tlie air from thc tissue, and allowed to remain in the solvcnt at room temperature for 24 hours. Thc pieces wcre then cleared in oil of bergamot, embedded in paraffin and sectoned at 5 to 8 p. One series of sections was mordanted with 1%osniic acid and stained with aniline acid fuchsin and methyl green ; another was mordarited with 2.5% aqueous solution of potassium hichromate and stained with neutral gentian. The aqueous solrents used were : distilled water, 0.85% sodium chloride, 10% sodium chloride, 0.1% acetic acid, 0.1% sodium hydroxide, 2.5% glycerine, 5% dextrose, 10% succharose and 1%urea. The pieces of tissuc were put into the aqueous solvents a t room temperature f o r 1 hour. They arid the control piecc wcrc then transferred to formol-Zenkcr, subjected to negative pressure, and then left in the fixative f o r 24 hours. Subsequently the tissues were washed, dehydrated, cleared, embedded i n paraffin, sectioned and stained by the two methods used, f o r the tissues treated with organic solvents. With freezing and drying in vacua, the restricted permeabilities of the cell membranes are abolished and solvents can readily penetrate the cell. Moreover, the solvents are not diluted by water in the cell. By this method, then, the action of the pure solvent on the intraeellular structures can be det ermined directly. The action of these solvents may be briefly summarized : 1. The pure organic solvents used, with the exception of etliylcne glycol, modified but did not remove the alpha, beta or zymogen granules. Ethylene glycol removed them all. 2. Of the aqueous solvents, distilled water, normal salt solution, 10c/b sodium chloride and 5% dextrose removcd the alpha granules ; 0.1% sodium hydroxide removed both alpha and beta granules; 1%urea modified both alpha and beta granules ; 5% dextrose nioclified the beta granules. Since the zymogen granules may almost completely disappear following ligation of the duct, the absence of zymogen SOLUBILITIES O F SECRETION G R A N U L E S 133 granules in the aqueous series could not be interpreted a s solution of them by the solvent. 11. IVhole fresh guinea pig pancreas was extracted for 24 hours at refrigeration temperature 1) with distilled water, 2 ) with normal salt solution and 3 ) first with an equal volume of 95% alcohol for 12 hours aiid then with distilled water f o r 24 hours. Pieces of each were then fixed f o r 24 hours in formol-Zenker, washed, debydrated, cleared, embedded i n paraffin, sectioned and stained with aniline acid fuchsin, methyl green and with neutral gentian. Examination of the sections revealed: A ) The alpha granules were removed with both water. and iiorinal salt solution. B) The beta, zymogen and hlankowski granules aiid ewii tlie mitochondria were resistant to both. C ) After ion fii-st with alcohol and then with water, only the Xaiikowski granules remained. F r o m this we may deduce that the alpha granules treated with alcohol, to wliich they a r e resistant, may be subsequently removed by water. 111. The reaction of pieces of fresh pancreas to various aqueous solvents was observed directly uiicler the oil immersion lens of the microscop. The identification of the zymogenic and Manko-cvski cells i n fresh tissue is easy with the dark-adapted eye, because of the size, refraction and distribution of the characteristic granules. The zymogen granules a r e the largest, most highly rcfractile and occiii’ grouped, for the most part, at thc free pole of the cell. The Mankowski granulcs a r e smaller, highly refractile and a r e distributed in the base or throughout the cell. The identification of the alpha and beta cells is somewhat more difficult in fresh tissue. A superficial island is located under low power by the slightly different color which the rich blood supply gives it. Under oil immersion, when the eye is dark adapted, the cells may be differentiated by the following characteristics. The beta cells which make up the majority of the cell content of the island a r e relatively small and well filled with tiny highly refractile granules. The 134 S. H. BENSLEY nuclei are f o r the most part spherical a i d (when the chromatin niay be visualized) have a considerable content of chromatin rather regularly arranged. The clear Golgi net may sometimes be visualized between the refractile granulcs. K h e n the beta granules are dissolved the cytoplasm appears optically homogeneous. The alpha cells are larger than the beta cells and tend to be elongated. The granules are larger, riot very refractile aiid appear watery. The nuclei are oval, pale and poor in chromatin, with a prominent nucleolus. When the alpha gmnulcs are dissolved the cytoplasm becomes highly vacuolatcd. The followiiig observations were recorded. 1. Distilled water and normal salt solution rapidly remove the alpha granules while the beta, zyinogen and Mankowski gimiules remain. On standing at room temperature there is a gradual dccrcnient in the beta and zynogeu granules, especially in distilled water. 2. Glycerine penetrates the cells slowly and the zymogen granules arc lost. No changes are observed in the alpha and beta granules. 3. Dilute sodium hydroxide rapidly and dramatically removes the alpha, beta and zymogen granules. The 3Iankow&i granules are somewhat modified. 4. Tt'ith dilute acetic acid, the alpha arid beta granules are preserved but with stronger acid the granules are riot distinguishable and the cells show a heavy precipitate. The zymogen granules are preserved in both cases. TESTS O F THE SOLUTION O F THE A1,PHA GRANIJLES On the basis of these findings fresh pancreas was ground up with sand and extracted with distilled water, the whole centrifuged aiid re-centrifuged at high speeds four time3 and the supernatant fat skimmed off. The sediment and llie suspensj on w3i-e examined by the stained smear method. Beta-like granules were observed in the suspension arid illcreasingly in the sediment. The filial suspension was filtered through a Berkefeld filter and collected into a sterile flask. SOLUBILITIES O F SECRETIOK GRANTTTlES 135 The total sediment, when extracted i n ‘70% alcohol, the extract evaporated and the residue taken up in Ringer’s solution and injected intravenously into a fasted rabbit, produced a €all in blood sugar equivalcrit to about that produced by 1unit of insulin. Preliminary chemical studies of the filtered extract showed : pII of about 7 ; a very fairit Millon reaction ; fine precipitation with ethyl alcohol, ether and acetic acid ; n o precipitation with salt solutions or dextrose. Intravenous injection of the filtered extract produced n o immediate toxic symptoms in the guinea pig, and a fall i n blood sugar only within the range of experimental error’. From these ohservations 1F-e may concludc that : 1. Alpha granules of the guinea pig pancreas a r e removed by water, normal salt, 10% salt, 5% dextrose mid tliliite sodium hydroxide solutions and a r e resistant to the pure organic solvents (m.ith the exception of ethylene glycol) a i d to dilute acetic acid. 2. Beta granules a r e rcmovcd by dilute sodium hydroxide, and ‘ioyb alcohol and a r e resistant to water, normal salt solution and dilute acetic acid. 3. Zymogen granules a r e removed by dilute sodium hydroxide, alcohol, glycerine and a r e resistant to water, normal salt solution and acetic acid. 4. Mankomski granules a r e resistant to all the aqueous solvents so f a r tested and to alcohol. These properties of the secretion granules correspond t o those described by Lane (’07) with the exception of the reaction of ilie alpha and beta granules t o acetic acid. Lalie believed that both types of granules were soluble in acetic acid. His negative findings may be due to a loss of staining capacity of tlie granules following fixation in acetic acid, or to chaiiges i n the cells produced by too great a coiicentratiori of the acid. Certainly, after fixation hy Bensley’s aceticosmic-bichromate fluid both granules may be well preserved and stained by the aniline acid fuchsin-methyl green method. The properties of the Mankowski granules correspond to those described by Bensley ( ’14PI5). 136 6. H. BENSLEY Because the alpha granules are so labile in water aiid salt solutions, it becomes apparent that no inethods of supravital staining o r perfusion with these fluids can be adequate to demonstrate these granules. And i n order t o study experimental changes in the granule content of the alpha cells, ercry effort must be made t o fix the tissue immediately. The author wishes t o express her gratitude to Prof. R. R. Rensley f o r his keen interest, his confirmation of observations and his technical and chemical assistance; to Dr. h’. TJ. Hoerr for his assistance with the freezing-drying proccdurc ; and to Dr. C. ,4.Woerner f o r blood sugar deteriniiiatioris. LITERATURE CITED BENSLEY,R. R. 1914-1915 Structure and relationship of thc islets of Langerhans. The Harvey Lectures, Series X, p. 373. LANE,hl. A. 1907 The cytological character of the arcas of Langerhans. Am. J. Anat., vol. 7, pp. 409-424. WOERNER,C’. A . 1938 Studies of the islands of Langrrhans after continuous i n t r n e n o u s injection of dextrose. Anat. Xec., 701. 71, pp. 33-57.