by Mustakallio (1), Cormane and Kalsbeek (2), and Bradshaw et al. (3) in the

Transcription

1 TILE JOURNAL OF TNVISP1GAT1VE DERMATOLOGY Copyright 1965 by The Willia1us & Wilkins Co. Vol. 44, No. 4 Printed in U.S.A. ELECTRON MICROSCOPIC STUDY OF "ADENOSINE TRIPHOSPHATASE" ACTIVITY IX MALIGNANT MELANOMA CELLS OF THE SYRIAN GOLDEN HAMSTER* In spite of the fact that tyrosinase activity iii melanocytes under various conditions has been extensively studied by means of lusto chemical and biochemical methods during the last two decades, other enzymes appear to have received relatively little attention in the study of figment cell growth. Recently the presence of adenosine triphosphatase (ATPase) activity in melanocytes was histochemically demonstrated by Mustakallio (1), Cormane and Kalsbeek (2), and Bradshaw et al. (3) in the human epidermis at the lightmicroscopy level. In the present report the localization of extra mitoehondrial adenosine triphosphatase activity was histochemically studied in the malignant melanoma cells of the Syrian golden hamster under the light and electron microscope. MATE1tTALS AND METHODS A pigmented type of the mlignant melanoma (code No. M. Mel 3) discovered by Fortner et al. (4) was transferred to and maintained in adult male Syrian golden hamsters of the colony of this laboratory by routine subcutaneous transplantation. Morphological and biological characteristics of this tumor were described in great detail (4), and they were remarkably similar to the features 01 the corresponding tumor in man. About 10 days after transplantation, the tumors actively growing tip to approximately 1 cm in diameter were surgically removed and cut into small blocks (2 3 mm thickness). Most of them were rapidly immersed in 6.5 per cent glutaraldehyde (5) buffered to nh 7.2 with 0.1 M cacodylate. Fixation was carried out in the cold (0 4 C) for 2 hours. Some fresh blocks were further sliced into smaller pieces (1 mm in diameter) and fixed in Palade's 1 per cent osmium tetroxide fixative for 1 hour in the cold. After fixation, osmiumfixed materials were em This investigation was supported by grants C 2893 and C5717 from the National Cancei Institute, National Institutes of Health, U.S. Public Health Service. Received for publication May 28, * From the Division of Oncology, The Chicago Medical School, Institute for Medical Research, Chicago, Illinois. t Present address: 11 Kamifujimaecho, Komagome, Bunkyoku, Tokyo, Japan. TAKASHI NAKAI, M.D., PH.l).t bedded in Epon according to Luft's method (6). The glutaraldehydefixed tissues were washed in cold 0.05 M cacodylate buffer containing 7.5 per rent sucrose at ph 7.4 for several changes (2 hours total). They were then stored in the same solution at 0 4 C overnight. After storage, the fixed tissues were quickly frozen in dry ice, and thin (15,z) and thick (40 frozen sections were cut in a cryostat and collected in the buffered sucrose plus 0.05 M MgCL at 0 4 C. They were transferred into the incubation mixture modified by OteroYilardebó (t al. (7) from the standard ATPase medium of Waehstein and Meisel (8). The mixture used was eomposed of 8.3 >< 10 M adenosine triphosphate (ATP) disodium salt, 2.5 X 10 M Pb(No:)2 and 0.05 M MgC12 in Trismaleate buffer at ph 7.2. The sections were incubated at 37 C for periods of from 30 to 60 minutes. At various times thin sections were picked up from the medium, rinsed in distilled water, treated for less than 1 minute in M acetic acid and transferred to ammoniurn sulfide, (NH4)25, for developing a dark precipitate of lead sulfide (PbS). When a satisfactory result was observed in the thin sections under the light microscope, the thick sections were taken out from the incubation mixture and rinsed in distilled water followed by brief treatment with M acetic acid. These thick sections were again washed in several changes of distilled water. Most of them were then postfixed in Palade's osmium tetroxide fixative for 1 hour at 0 4 C, and embedded in Epon according to Luft's proce (lure (6). Some sections were embedded in Epon without the osmium refixation. In order to evaluate the specificity of the ATPase reaction, control experiments were carried out at the light and electron microscopic level. Control tissues were incubated for 60 minutes in the substratefree medium or in the medium where 3 >< 10 M sodium flglycerophosphate was substituted for ATP at the same ph. Ultrathin sections were cut on an LKB Ultrotome with glass knives, mounted on formvarcoated grids, and stained in a saturated solution of uranyl acetate in distilled water for 30 to 60 minutes. They were examined in an RCA EMIl 3F electron microscope operated at 50 KY. 264 RESULTS Light Microscopy, Incubation of melanoma tissues in the ATPase medium resulted in the formation of a visible reaction product in the blood vessels and in many tumor cells (Fig. 1). The reaction product appeared to be accumu

2 E. M. STUDY OF "ATPASE" IN MELANOMA CELLS 265 lated in tile periphery of the tumor eehs, and dendritie processes became discernibie in some celis. ATPase activity seemed to vary from one ceh to the other. Some celis were stained much more heavily, and their ceh borders and cyto piasm showed dense deposits. No reaction product deveioped when sections of the identical tumor were incubated in the control medium without ATP or in the medium where sodium flgiycerophosphate was substituted for ATP at the same ph. Electron Microscopy. An electron micrograph of a portion of melanoma cehs fixed in Paiade's osmium tetroxide fixative is shown in Fig. 2. These cehs contained an indented nucieus, round or ovai mitochondria, numerous pigment granuies and weil deveioped Goigi zones. Roughsurfaced endopiasmic reticuium and free ribosomes, the iatter often grouped into rosettes, Were comparatively numerous. Pigment granuies were round or poiygonal in profile and measured 100 to 450 mi in diameter. They contained various amounts of eiec tron opaque materiai interpreted as meianin. Some irreguiar substructure was observed in the pooriy fihed granuies which usuaily had a singie membrane. Pinocytic vesicies were rareiy found aiong the cell borders. 1 In the giutaraidehydefixed materiais incubated with ATP, the final product of the enzymatic reaction was iocahzed to the ceh membranes and Goigi zones of meianoma celis (Figs. 3 and 4). Sites of ATPase iocahzation were characterized as intenseiy eiectron opaque, granuiar deposits of lead phosphate, PbPO4, under the eiectron microscope. Most of the tumor cells had the dense deposits on their membrane, although in some instances the reaction product was discontinuousiy deposited and faiied to form a continuous layer (Figs. 3 and 4). Lead phosphate was quite frequentiy deposited in either a spotty or linear fashion on the Goigi zones (Figs. 3 and 4). Mitochondrial ATPase activity appeared to be stih retained even after the glutaraidehyde fixation, although greatiy diminished, and numerous precipitates of fine electron opaque particies were seen virtuahy in ail mitochondria (Figs. 4,5 and 6). When the materiais were fixed in giutaraidehyde and then postfixed in osmium tetroxide, some pigment granuies having a heavy accumuiation of melanin showed a fairly high electron opacity, which often made an exact evaluation of the enzymatic reaction in these granules difficuit. However, the majority of the pigment It Fio. 1. Light photoniicrograph of a frozen section of the hamster melanoma incubated for 60 minutes in ATPase medium. Dark deposits of reaction product in the cell borders can be seen in many tumor cells. Some cells are stained darker, and dendritic processes are often discernible. Note small blood vessels shown in lower and right upper portions of the illustration. X 450.

3 266 THE JOURNAL OF INVESTIGATIVE DERMATOLOGY 1:ti _t.,...j._t. : L:y, 1: HTh.' c :L ''. d 1. '..A;.:: J 6 / fl. ' FIG. 2. Electron micrograph of a portion of melanoma cells containing many pigment granules, welldeveloped Golgi zones (G) and mitochondria (M). A centriole (C) and a portion of an indented nucleus (N) are also illustrated. X 18,000.

4 L 4 ry M :'t, t : L.V : a I' $4 S4.?t 4 0 '0 4. K. * 4. 'C 'F t: :: : a: ":' \ ',: a :.'.4 FIG. 3. Tumor cells of malignant melanoma fixed in cold buffered glutaraldehyde and incubated for 1 hour with ATPase medium. The tissue is refixed in Palade's osmium tetroxide fixative. The dense reaction product, lead phosphate, covers around the cell membrane in almost continuous fashion. Note that Golgi zones also exhibit dense deposits of ATPase reaction product. Pigment granules show a lower electron opacity than the lead phosphate, thus indicating the possible absence of ATPase reaction product from these granules (see also Figs. 4 7). X 10,

5 268 THE JOURNAL OF INVESTIGATIVE DERMATOLOGY 7% a' 4 * ft. '4 :c.*twi,,:.. S I.. ft C. M 7.. S. " p t $ S m wc. t. tr41.4.,r $. I 4t' 1* ' ' 4 '4 t : rk.1 I? N 4! : FIG. 4. Similar material as in Fig. 3. The reaction product deposited around the cell membrane (CM) is somewhat discontinuous in the lower part of the illustration. The final dense product appears in association with several Golgi zones, often outlining the parallel lamellae (arrows). Note the fine lead phosphate precipitates in the rnitochondria. X 16,500. SI.

6 E. M. STUDY OF "ATPASE" IN MELANOMA CELLS 269 granules except occasional insignificant precipitates of a few lead particles. The difficulty of distinguishing lead deposits from the pigment granules with a high electron opacity was overcome by the examination of nonosmificated materials. Although preservation of the fine structure was not as good as the osmiumr1 t ' 1.: I t1 t I F I ' N FIGS. 5 AND (1. Portion of melanoma cells. The dense reaction product is localized in the cell membrane and Golgi zone. Mitochondiia exhibit fine lead phosphate precipitates. Note that no significant reaction product is seen in pigment granules. A scattering of some fine nonspecific deposits is visible. X 22,000 arid 36,500. FIG. 7. Similar material as in Fig. 3, but the refixation in Palade's osmium tetroxide fixative is omitted. Portion of a melanoma cell is shown with the dense reaction product localized in the cell membrane. Note the absence of lead precipitates in pigment granules which show a relatively low electron opacity because of the onussion of osmium fixation. X 31,500. granules, particularly those with incomplete melanization, had a lower electron opacity than lead deposits, so that the latter could be (listinguislieci from the former if present. Careful and close inspection of many electron micrographs including Figs. 3 6 revealed that practically no reaction product occurred in the pigment

7 270 THE JOURNAL OF INVESTIGATIVE DERMATOLOGY refixed tissues, pigment granules seen in the nonosmificated tissues had a much lower electron opacity (Fig. 7). Under this condition almost no accumulation of the reaction product was detected in the pigment granules (Fig. 7), while the cell membrane, Golgi zone and, to a much lesser extent, mitochondria showed the enzyme activity. In the control tissues incubated in the substratefree medium or in the mixture where sodium /3glycerophosphate was substituted for ATP at the same ph, no lead deposits were observed in any of the organdies of melanoma cells. DISCUSSION Recently it was described by Sabatini et al. (5) that glutaraldehydefixed tissues showed an excellent preservation of ultrastructural detail and an adequate amount of various enzyme activities survived for histochemical demonstration. The combination of this fixation and histochemical incubation in the ATPase medium has made it possible to correlate the localization of "ATPase" with the well preserved fine structure of melanoma cells. The electron micrographs revealed that the electron opaque reaction product, lead phosphate, produced by the hydrolysis of ATP was clearly localized in the cell membranes and Golgi zones of hamster melanoma cells. A light accumulation of the fine lead precipitates also occurred in the mitochondria. It is our interpretation, however, that no significant ATPase reaction product was present on the pigment granules. Several recent papers (1 3) have dealt with the presence of ATPase activity in normal epidermal melanocytes observed under the light microscope. In an attempt to demonstrate ncural elements of human epidermis by means of histochernical reactions, Mustakallio (1) found that the dendritic network of Langerhans' cells and probably of melanocytes as well as the papillary nerves and their intraepidermal extensions showed a positive ATPase activity. Subsequently, Cormane and Kalsbeek (2) also reported the presence of ATPase activity in dendritic cells, presumably Langerhans' cells, of the normal human skin. They interpreted that the cell membrane is a possible site of the enzyme activity (2). Bradshaw et al. revealed in their lightmicroscopic analysis of epidermal melanocytes of normal, human skin incubated in the ATPase medium of Wachstein et at. (9) that the cell membranes, mitochondria and possibly melanosomes were the sites of the enzyme activity. In the hamster melanoma cells, however, deposits of the reaction product were not found in the melanosomes, while the Golgi zones in addition to the cell membranes showed almost constantly a positive "ATPase" activity. It is now of considerable importance to evaluate the specificity of this reaction. According to Bradshaw et al. (3), melanin granules in Malpighian cells show nonspecific staining when ATPase is demonstrated by methods which utilize the sulfide conversion of sulfurmetal compounds. However, this did not apply to the present experiment, since the reaction product was not deposited in the melanin granules of our material. Moreover, the treatment with ammonium sulfide was omitted for electron microscopic cytochemistry. There is also the possibility that the hydrolysis of ATP is due to the action of the nonspecific alkaline phosphatase. However, the extramitochondrial as well as mitochondrial lead deposits were not present in the melanoma tissues incubated in the medium without ATP or in the medium where sodium flglycerophosphate was substituted for ATP. Novikoff et al. (10) strongly suggested, according to their results of histochemical and cell fractionation studies of ATPase activity in the liver, that the enzyme activity shown by the Wachstein and Meisel method (8) with ATP is a specific ATPase rather than a nonspecific alkaline phosphatase. Novikoff et al. (10) and Novikoff (11) emphasized that phactivity curves are of considerable value for distinguishing substratespecific phosphatases in tissues, particularly in the liver and spleen, where there is relatively low alkaline phosphatase activity present. They felt that the specific ATPase can be histochemically demonstrable at ph 7.2, which is used in the Wach.steinMeisel method, and that the activity of nonspecific alkaline phosphatase can be neglected. Moreover, this view has received further supports from a serics of electron microscopic studies of various tissues done by Novikoff and his coworkers (12 16). In sections of kidney, liver and hepatomas as well as some other tissues incubated with various nucleosidephos

8 E. M. STUDY OF "ATPASE" IN MELANOMA CELLS 271 phates, the reaction product is always accumulated to specific membranes (12 16). Since the cytoplasm of melanoma cells is known to possess a relatively low level of alkaline phosphatase activity (17, 18), the enzyme activity demonstrated under our experimental conditions is most likely the specific magnesiumactivated "ATPase". It is of great interest that the "ATPase" activity is localized in the malignant melanoma cells which show a rapid multiplication, high invasiveness and active melanin synthesis, since ATPase is one of the nucleosidephosphatases which are known to mediate the high level energy necessary for a variety of synthetic and functional activities of the cell (19). It is also believed that ATPase appears to play a role in the transport of inorganic phosphate (20, 21) and Na and K (22, 23) across cell membranes. The ultrastructural localization of "ATPase" demonstrated in the Golgi zone and cell membrane of the melanoma cells appears to be relevant to the functions of these organdies. Possible roles and functional significance of nucleosidephosphatases in the membrane system of cells are well documented and discussed in great detail by Novikoff et al. (14). They suggested that the split of ATP or other energyrich nucleosidephosphate might lead to alteration of a contractile protein in the cytomembrane, as in the mechanism of cell transport. In our material, however, more studies on various nucleosidephosphatases will be required to evaluate the functional significance in our present observation. Cytochemical studies of ATPase as well as other phosphatases in the melanocyte system of Syrian hamsters under various experimental conditions are in progress. Although it was not the purpose of this experiment to demonstrate the mitochondrial ATPase activity, the light deposits of reaction product in mitochondria appeared to indicate that a small amount of the enzyme activity still remained. It was reported by many investigators (5, 13, 24 29) that significant deposits due to ATPase in the mitochondria cannot be demonstrated in the electron microscope when tissues are subjected to prolonged treatment in chemical fixatives including aldehydes preceding the incubation with ATP. Recently, however, OteroVilardebó et al. (7) observed an extensive deposition of reaction product within mitochondria of rat colonic mucosa fixed overnight in formalin when frozen sections, prepared from the tissue which was once chilled at 70 C, were incubated with ATP. The mitochondria of rat kidney treated in the identical manner failed to give a positive reaction (7). Our finding that some degree of the "ATPase" activity survived under our experimental conditions may add another example to that of OteroVilardebó et al. (7). SUMMARY Malignant melanoma of Syrian golden hamster was fixed in cold 6.5 per cent glutaraldehyde. Frozen sections were incubated in a modified WachsteinMeisel ATPase medium and examined under the light and electronmicroscope. The magnesiumactivated "ATPase" was localized in the cell membrane and Golgi zone of the melanoma cells. Mitochondria appeared to retain some degree of the enzyme activity after the aldehydè fixatipn. However, no such activity was histochemically demonstrable in the melanosomes under the experimental conditions employed. REFERENCES 1. MTJSTAKALLIO, K.: Adenosine triphosphatase activity in neural elements of human epidermis. Exp. Cell Res., 28: 449, CORMANE, H. H. AND KALSBEEK, 0. L.: ATPhydrolyzing enzymes in normal human skin. Dermatologica, 127: 381, BRADSUAw, B., WACHSTEIN, M., SPENCE, J. AND ELIAS, J. M.: Adenosine triphosphatase activity in melanocytes and epidermal cells of human skin. J. Histochem. Cytochem., 11: 465, FORTNER, J. G., MAHY, A. G. AND SCHRODT, G. R.: Transplantable tumors of the Syrian (golden) hamster. Part 1: Tumors of the alimentary tract, endocrine glands and melanomas. Cancer lies., 21 (Suppl.): 161, Sn'rnci, D. 1)., BENSCH, K. AND BARRNETT, R. J.: Cytochemistry and electron microscopy. The preservation of cellular ultrastructure and enzymatic activity by aldehyde fixation. J. Cell Biol., 17: 19, LIJFT, J. H.: Improvements in epoxy resin embedding methods. J. Biophys. Biochem. Cytol., 9: 409, OTEROVILARDEBó, L. R., LANE, N. AND GODMAN, G. C.: Demonstration of nfitochondrial ATPase activity in formalinfixed colonic epithelial cells. J. Cell Biol., 19: 647, WACHSTEIN, M. AN!) MEISEL, E.: Histochemistry of hepatic phosphatases at a physiologic ph with special references to the demonstration of bile canaliculi. Amer. J. Clin. Path., 27: 13, 1957.

9 272 THE JOURNAL OF INVESTIGATIVE DERMATOLOGY 9. WACHSTEIN, M., BRADSHAw, M. AN!) ORTIZ, J. M.: ilistochemical demonstration of mitochondrial adenosine triphosphatase activity in tissue sections. J. Histochem. Cytochem., 10: 65, Noviicon, A. B., HAUSMAN, 1). II. AND POD BER, E.: The localization of adenosine tnphosphatase in the liver: in situ staining and cell fractionation studies. J. ilistochem. Cytochern., 6: 61, Noviico, A. B.: Distinguishing substratespecific phosphatases. J. Histochem. Cytochem., 6: 251, NovIKoFF, A. B.: Biochemical and staining reactions of cytoplasmic constituents. In: Developing Cell Systems and Their Control, p ). Rudnick, ed. New York, Ronald Press Co., ESSNER, E., N0vIKOFF, A. B. ANT) MASEK, B.: Adenosine triphosphatase and 5 'nucleotidase activities in the plasma membrane of liver cells as revealed by electron microscopy. J. Biophys. Biochem. Cvtol., 4: 711, No\TIK0FF, A. B., ESSNER, E., GOLDEISCriER, S. AND MARGARETA, H.: Nucleosidephosphatase activities of cvtomernbranes. Symp. Internat. Soc. Cell Biol., 1: 149, ESSNER, E. AND NovIxoFF, A. B.: Cytological studies on two functional hepatomas. Interrelations on endoplasmic reticulum, Golgi apparatus, and lysosornes. J. Cell Biol., 15: 289, NOvIKOFF, A. B. AND ESSNER, E.: Pathological changes in cytoplasmic organelles. Fed. Proc., 21: 1130, GREENBERG, S. S., KOPAC, M. J. AND GORDON, M.: Cytology and cytochernistry of melanoma cells. Ann. N.Y. Acad. Sci., 67: 55, RosE, 0. 0.: Observations on human melanoma cells in tissue culture. Ann. N.Y. Acad. Sci., 100: 92, LINDBERG, 0. AND ERNSTEH, L.: Chemistry and physiology of mitochondria and microsomes. In: Protoplasmatologia Handbuch der Protoplasrnaforschung, p L. V. lleilbrunn and F. Weber, eds. Wien, Springer Verlag, GOURLEY, 1). R. H.: The role of adenosine tnphosphate in the transport of phosphate in the human erythrocyte. Arch. Biochern. 4: 711, PRANKERD, T. A. J. AND ALTMANN, K. I.: A study of the metabolism of phosphorus in mammalian red cells. Biochem. J., 58: 622, KEYNES, II. 1).: The energy source for active transport in nerve and muscle. In: Membrane Transport and Metabolism, p A. Kleinzeller and A. Kotyk, eds. New York, Academic Press, Inc., POsT, R. L. ANt) ALBRIGIIT, C. P.: Membrane adenosine triphosphatase system as a part of a system for active sodium and potassium transport. In: Membrane Transport and Metabolism, p A. JKleinzeller and A. Kotylc, eds. New York, Academic Press, Inc Asiiwowni, C. T., LU1BEL, F. J. AND STEWART, S. C.: The fine structural localization of adenosine triphosphatase in the small intestine, kidney, and liver of the rat. J. Cell Biol., 17: 1, DEI3EYER, J. M., DEMAN, J. C. H. AND PER S1JN, J. P.: ATPase activity on the intercalated disc and C hands of mouse heart muscle. 1. Cell J3iol., 13: 452, KAPLAN, S. E. AND NOVKOFF, A. B.: The localization of adenosine trihosphatase activity in rat kidney: Electron microscopic examination of reaction product in formalincalcium fixed frozen sections. J. Histochem. Cytochem., 7: 295, LAZARUS, S. S. AND BARDEN, H.: Histochernistry and electron microscopy of mitochondrial aclenosinetriphosphatase. j. Histochern. Cytochern., 10: 285, SCARPELLI, 1). (1. AND CRAIG, E. L.: The fine localization of nucleoside triphosphatase activity in the retina of frog. J. Cell Biol., 17: 279, TICE, L. AND BARRNETT, II. J.: Fine structural localization of adenosinetriphosphatase activity in heart muscle myofibrils. J. Cell Biol., 15: 401, 1962.