Ultrastructural Nuclear Changes Due to Tannic Acid

(CANCER RESEARCH 27, 1658-1671,September 1967]

Ultrastructural ANTONIO

Nuclear

RÕCELA, HAROLD

Changes

GRADY,

AND

Due to Tannic

DONALD

Acid

SVOBODA

Department of Pathology and Oncology, University of Kansas Medical Center, Kansas City, Kansas 66103

SUMMARY In rat liver, tannic acid causes an early and promptly revers ible form of nucleolar macrosegregation which differs from other forms of nucleolar segregation or capping. A small nucleolus with condensation of compact ribonucleoprotein granules into one or more dark zones is characteristic. These zones are surrounded by a light area made up of a mixture of ribonucleo protein granules and fibrils. In later stages the zones contain predominantly fibrils. The macrosegregation due to tannic acid is not related to significant decrease in RNA and proteins in the nuclear or cytoplasmic fractions. In this respect, the biochemi cal effects of tannic acid differ from certain other hepatocarcinogens, notably aflatoxin and lasiocarpine, which also cause early nucleolar abnormalities. In light of recent reports, nucleolar segregation or capping appears to be a morphologic expression of biochemical alterations in nuclear function, particularly the binding with DNA and inhibition of its template activity. The precise biochemical action of tannic acid on nucleoproteins is not known, and similarity of its action to that of other capping agents is not substantiated by the preliminary biochemical observations in the present study. INTRODUCTION Tannic acid, which has been implicated in human liver cancer in certain parts of the world (21), is present in certain natural foodstuffs and beverages and is hepatocarcinogenic in the rat (20). In this respect, it is similar to other naturally occurring hepatocarcinogens, such as pyrrolizidine alkaloids (9) and aflatoxin (23), which also may contaminate diets. Contamina tion by natural carcinogens may be of importance in under developed areas since it has been shown, for example, that the carcinogenicity of tannic acid is enhanced markedly by con comitant protein deficiency (22). Tannic acid, pyrrolizidine alkaloids, and aflatoxin are carcino genic in relatively small chronic doses or after brief exposure to large doses. In liver cells, aflatoxin (4, 52) and lasiocarpine (55) cause abnormalities in nuclear and nucleolar morphology which are associated temporally with decreased RNA and protein content. Although previous light (2, 7, 21) and electron micro scopic (1, 6) studies on tannic acid injury are available, there is relatively little information on ultrastructural changes in the nucleus and associated biochemical changes due to this agent. 1These studies were supported in part by USPHS Grants CA-5680 and CA-8055, and by the Kansas Division of the Amer ican Cancer Society. Received January 6, 1967; accepted May 12, 1967.

1658

In order to determine whether tannic acid, in common with other carcinogens, might produce similar cellular abnormalities, this study of early effects of tannic acid injury on the nucleus was undertaken. Considering the essential role of the nucleus in processes of basic importance to cell growth and division, it may be impor tant to renew attention to primary ultrastructural changes in the nucleus due to carcinogens. As emphasized by Stowell (50), studies on fine structure and at the gene level must be done before decisions about specific nuclear changes in carcinogenesis can be made. Subtle morphologic variations exist in the nucleus of liver cells injured by several agents and, accordingly, it is important to ascertain the possible significance of nuclear changes in the total cellular response to carcinogenic injury and to distinguish the essential biologic differences among several abnormalities in nuclear ultrastructure. MATERIALS AND METHODS Sixty female inbred rats (Fisher 344) were studied after a single subcutaneous dose of tannic acid (gallotannic acid, reagent grade) at varying intervals from 1 hour to 3 weeks as shown in Table 1. Forty rats were given a 700 mg/kg dose because, in preliminary studies, this dose caused negligible centrilobular necrosis in the liver. To assess the effect of hepatic necrosis or cytoplasmic degradation on the nuclear and nucleolar alterations, the animals receiving the higher dose (1200 mg/kg) were com pared with those receiving the lower dose. Blocks for electron microscopy were taken from the liver at all intervals of the ex periment while samples of pancreas, duodenum, and kidneys were blocked only at selected intervals (Table 1). At the time of sacrifice, complete autopsies (exclusive of the central nervous system) were done and tissues were prepared for light micros copy. Biopsies were done under light ether or Metofane anesthesia. Blocks for electron microscopy were fixed in osmium tetroxide buffered with s-collidine to pH 7.4, or in glutaraldehyde with postosmication. The blocks were dehydrated in a graded series of alcohols and embedded in a mixture of Epon 812 and Araldite. Sections were stained with lead hydroxide and/or uranyl acetate and examined in an RCA 3B or 3G electron microscope. For orientation, adjacent semithin sections (0.5-1.5 n) were stained with azure A in sodium bicarbonate. For cytochemical studies, samples were prepared for specific enzyme digestion by modifications of the methods of Swift (56) and Bernhard and Granboulan (5). Blocks of less than half a cubic millimeter volume were fixed in phosphate-buffered glu taraldehyde for 1 hour. These were incubated with the specific enzymes (ribonuclease, 1 mg/ml at pH 6.5; pepsin, 0.8% at pH CANCER RESEARCH VOL. 27

Ultrastructural

Nuclear Changes Due to Tannic Acid

1.5; and trypsin, 0.8% at pH 8) for 60 minutes, and digested nucleic acids were extracted with 5% trichloroacetic acid. The digested blocks were dehydrated and embedded as outlined above but without osmication. Biopsies for light microscopy were fixed in buffered formalin and embedded in paraffin. Sections were stained with hematoxylin and eosin, periodic acid-Schiff (PAS) with or without diastase digestion, methyl green pyronine, Feulgen's, and Perls'

marked ultrastructural changes in nucleoli, while the latter 3 intervals coincided with the morphologic recovery and postrecovery periods. Separation of the nuclear fraction was done by the method of Widnell and Tata (58). Nucleic acids were assayed by pentose analysis according to the method of Schnei der (41) and total proteins were determined by the Lowry method (26).

stains. Frozen sections were stained with oil red 0 for fat. Tissue from twelve control rats was processed similarly for light and electron microscopic study.

RESULTS

Biochemical

Studies

Liver from 16 female rats was assayed for DNA, RNA, and protein in both the nuclear fraction and in the homogenate as shown in Table 2. The 6-hour interval coincided with the most

Of the 60 rats injected with tannic acid, 52 were used for electron microscopic study since they survived the intended length of observation. Eight rats that died at approximately 48 hours (Table 1) were excluded from electron microscopic exam ination. Six of these received the higher dose (1200 mg/kg body weight), and at autopsy there was moderate centrilobular ne crosis of the liver. Liver

TABLE 1

Light Microscopy. After a single injection of tannic acid, liver cells showed nucleolar and cytoplasmic changes, each pre dominating at different intervals. As observed in semithin sec (mg/kg)700226»644°6"2222240120022228"2220Control10 tions, changes in the nucleoli were most marked during the first Interval1 24 hours while cytoplasmic changes predominated at 48 and 72 hours. hour3 By 6 hours, the nucleoli of cells throughout the lobule were hours6 small and transformed into two distinct components consisting hours12 of & light, homogeneous area containing one or more well-de hours18 marcated dense zones (Figs. 1, 2). Recovery was evident by 24 hours24 hours. These nucleolar changes were not apparent in paraffinhours48 embedded sections stained with hematoxylin and eosin, methyl hours72 green pyronine, or Feulgen stains. The positive staining with hours5 methyl green pyronine of the whole nucleolus suggested that days1 both the light and dark zones of the nucleoli, as observed in the week2 weeks3 Epon sections, contained RNA. The cytoplasm at this interval weeksTotal was generally intact (Fig. 3), though many centrilobular cells No. of ratsDosage contained fine fat droplets. From 48 to 72 hours, nucleolar abnormalities were not appar o Biopsies of the pancreas, kidney, and duodenal mucosa were ent. Cytoplasmic changes were more prominent than previously studied in addition to the liver which was blocked in all animals. 6Two rats given 700 mg/kg and 6 rats given 1200mg/kg died at but involved mainly the centrilobular and midzonal areas. Fatty change was marked and was associated with prominent cyto approximately 48 hours; these were excluded from ultrastructural plasmic degeneration manifest as accumulations of irregular study. Distribution

of Rats Given a Single Injection

of Tannic

Biochemical Estimations

Acid

TABLE 2 in Liver after 700 mg/kg of Tannic Acid fractionDNA

Interval(hours)Control6°2448168HomogenateDNA (mg/gm)2.6

(mg/mg DNA)3.3

DNA)74.3 (mg/mg

(mg/gm)1.5

DNA)0.17 (mg/mg

DNA)2.3 (mg/mg

0.8574.9 ± 0.081.6 ± 0.010.17 ± 0.102.6± 0.061.2 ± 0.010.23 ± ±1.0476.9 0.292.6 ± 0.161.0 ± 6.3961.9 ± 0.020.23 db 0.112.7 ± 2.42(P± 0.181.6 ± 0.030.20 ± 0.052.8 ± 0.005)661.9 = ± ±0.08RNA ±0.02Protein ± 0.19 ±0.20RNA ±0.04Protein 4.30(P = 0.05)Nuclear

0.052.3 ± 0.022.1 ± 0.202.4 ± 0.202.8 ±

0.043.6 ± 0.263.9 ± 0.273.7 ± 0.333.6 ±

" Four animals were used at each interval as well as in the control group. Tannic acid was given by subcutaneous injection. 6P values are given only where significant. SEPTEMBER

1967

1(55!)

Antonio Rácela, Harold Grady, and Donald Svoboda eosinophilic bodies, some of which were PAS-positive but did not contain glycogen. In a few animals, particularly those given the higher dose (1200 mg/kg), centrilobular hemorrhagic necro sis was seen. Hepatocellular regeneration was already evident. Mitoses were frequent and enlarged hyperchromatic nuclei were seen, particularly in the periportal areas. All changes were minimal by 5 days and, except for occasional cytoplasmic fat droplets, the liver appeared essentially normal at one week. Nuclear inclusions (Fig. 4), similar to those described by Korpassy (21), occurred at 6 hours but gradually disappeared between 1 and 2 weeks. These inclusions were eosinophilic and often surrounded by a clear halo. They stained positively for iron, were faintly pyroninophilic and were negative with the Feulgen stain. They were present irregularly in the lobule but were more numerous in centrilobular areas. Similar staining for iron was observed rarely in the cytoplasm of hepatocytes and von Kupffer cells. The nucleoli of the liver cells did not stain for iron. Electron Microscopy. The morphology of normal nucleoli of the rat liver was essentially similar to that described previ ously (5, 30, 31, 46). The terminology of nucleolar constituents viewed by electron microscopy is quite variable due to the use of terms such as vacuoles, pars amarpha, and nuckolonema in an attempt to relate ultrastructural features to light or phase con trast microscopic observation. In addition, even in the same animal, differences in morphology of nucleoli of different organs are present. The nucleoli of the liver, pancreas, renal proximal tubular epithelium, and duodenal mucosa differ in their size and configuration, however, as demonstrated by Marinozzi (28), the basic components are the same in most instances. In our labora tory, the following simplified terminology has been adopted for the nucleolus of the rat liver (Fig. 5) as well as the other organs compared in this study. Fibrillar component: generally disposed in moderately elec tron-dense, compact strands and composed of fibrils approxi mately 70 A in diameter. Often, the fibrils are dispersed, forming a light fibrillar area. Granular component; composed of aggregates of ribonucleoprotein (RNP) particles, 100-150 A in diameter, irregularly interposed throughout the fibrillar component. Protein matrix in which both the fibrillar and granular com ponents are embedded. The fibrillar and granular areas are closely associated, forming an. irregular branching meshwork. Enclosed by this skein or occupying irregular indentations at the nucleolar periphery, are relatively electron-lucent "pockets" which are composed either of chromatin or invaginations of the nucleoplasm. The nucleolusassociated chromatin partially or completely encircles the nucleolus (Fig. 5). Pathologic Changes. The morphologic changes in the liver at the two dose levels were qualitatively similar and differed only in degree of departure from the normal. At one hour, the nucleus was unaltered except for slight increase in interchromatin granules. By 3 to 6 hours, many nucleoli in all portions of the lobule were round and consisted of a large and comparatively light area of mixed granules and fibrils. Within the light area were one to five compact, dense granular zones (Figs. 6, 7). After a short period of incubation with ribonuclease, the RNP granules in the dense zone were digested (Fig. 8), while the H>(>0

fibrillar and some granular components in the light zone were still intact. With more complete digestion, both light and dark areas were indistinguishable and the nucleolus was faint and fairly homogenous. Although enzymatic digestion is often diffi cult to control, the results suggest that both zones of the altered nucleolus contain RNA. It is suggested that, in some nucleoli (Figs. 7, 9, 14), the RNP granules in the dark zones appear more distinct and smaller in diameter. Most nucleoli were small and compact (Figs. 7-11) and showed loss of the normal skein pat tern as well as obliteration of intranucleolar invaginations of nucleoplasm. In others, the progressive diminution in number of granules in the light zone contrasted to the persistence of gran ules in the darker zone (Figs. 9, 10). By 12 hours the light area consisted almost exclusively of fibrils (Fig. 11). In the fully developed alteration, termed macrosegregation, the nucleolus was shrunken, with one or several dark zones of compact RNP granules surrounded by a light area predominantly made up of fibrils. The nuclei were slightly enlarged and the nucleoplasm was finely dispersed with no increase in the interchromatin space. The interchromatin granules appeared to be further in creased in number. In addition, there were occasional inclusions consisting of particles that measured approximately 200-400 A in diameter (Fig. 6). These represented early forms of the iron inclusions observed by light microscopy since their density and composition were identical with those of larger bodies, positive with Perls' stain, present at later intervals (Figs. 12, 13). With complete digestion by ribonuclease, the granules in these inclu sions were removed; this finding, along with the slight pyroninophilia demonstrated by light microscopy, suggests that the inclusion bodies are complexes of iron and ribonucleic acid. By 12 hours, with the lower dose (700 mg/kg), some of the nucleoli recovered (Figs. 12, 13) while with the higher dose (1200 mg/kg) nucleolar shrinkage and macrosegregation were more pronounced than at the lower dose. The iron-containing inclusions aggregated into distinct round masses, often larger than the nucleolus (Fig. 12). In addition, defects or spaces in the nucleoplasm surrounding these masses (Fig. 13) corresponded to the halo around the inclusions observed by light microscopy. By 18 hours, the altered nucleoli were similar to those seen earlier but were observed less frequently. The remaining nucleoli were either normal or dispersed. An additional alteration ob served at this time involved only occasional nucleoli and con sisted of the formation of multiple, small but well-demarcated dense foci or plaques (Figs. 14, 15) which, in some instances, appeared to be fibrillar. These were seen in loosely structured dispersed nuclei with (Fig. 14) or without (Fig. 15) macrosegre gation. This type of alteration is similar to the early stage of nucleolar capping due to the pyrrolizidine alkaloids (Fig. 16). In other nucleoli (Fig. 17), the fibrillar component was less closely associated with the granular component than in normal nucleoli. At 24 hours, the nuclear size approached normal and virtually all nucleoli were recovered. As at 18 hours (Fig. 17), the inter chromatin granules were still increased. Iron-containing inclu sion bodies were likewise still prominent. At 48 to 72 hours, the nuclei in the immediate periportal cells contained enlarged nucleoli. In the remainder of the lobule, the nucleus was often irregular with invaginations and pseudoinclusions of cytoplasmic organelles. Nuclear fat vacuoles were often present. Nucleolar macrosegregation was not observed. CANCER

RESEARCH

VOL. 27

Ultrastructural Nuclear Changes Due to Tannic Acid From 5 days to 3 weeks, the nuclei were generally similar to control animals. The interchromatin granules appeared to return to normal amount by one week. Likewise, the iron-containing inclusions gradually became smaller and, by 3 weeks, were in conspicuous. There were no changes apparent in perichromatin granules, and the chromatin distribution was normal at all intervals. The apparent decrease in amount of chromatin in the first 24 hours was related to the nuclear enlargement rather than to an abso lute decrease. There were no changes in the nuclear pores or the nuclear envelope. Cytoplasmic Changes. The cytoplasmic changes were non specific. During the period of nucleolar abnormalities (first 24 hours), disorganization and dilation of the rough endoplasmic reticulum associated with ribosomal detachment and hyperplasia of the smooth endoplasmic reticulum were prominent. Foci of cytoplasmic degradation were minimal or absent in most cells. Recovery of the cytoplasmic alterations varied according to their location in the hepatic lobule. In the periportal regions, recovery was evident by 24 and 48 hours. In the centrilobular and midzones, however, the changes in ergastoplasm and focal cytoplasmic degradation were accentuated and fatty change was likewise increased at this interval. For the most part, cells with prominent cytoplasmic alterations showed no nucleolar macrosegregation. Instead, the nucleoli were normal or enlarged. Finely granular, electron-dense particles, possibly hemosiderin, were occasionally observed in the cytoplasm of the hepatocytes and of von Kupffer cells. By one week, the hepatic cells in all portions of the lobule were restored to normal. Other Organs At 6 hours, the nucleoli of the proximal renal tubular epithelial cells showed macrosegregation resembling that seen in hepatic cells. In contrast to the liver, distinct separation of granules and fibrils into relatively pure areas was observed often (Fig. 18). The light zone generally was made up exclusively of fibrils. In the kidney, nucleolar alteration was sustained at 24 hours and was not accompanied by significant cytoplasmic necrosis. Pan creatic acinar cells and the epithelial cells of the duodenal mucosa did not show nucleolar changes at 6 or 24 hours after injection of 700 mg of tannic acid. Biochemical Changes. The RNA and protein content of the nuclear fraction and liver homogenate (Table 2) were essentially unchanged at the interval of maximal nucleolar alteration (6 hours) or at recovery (24 hours). There was only a slight decrease in protein in the homogenate at 48 and 168 hours. Otherwise, tannic acid did not cause significant alterations in these selected biochemical parameters at the given intervals and dosage. The isolated nuclear pellet was examined in the electron microscope, and it showed good morphologic preservation of the nuclei and its components, including the altered nucleoli. There was only minimal contamination by cytoplasmic fragments or by ribosomes adherent to the external lamina of the nuclear enve lope. DISCUSSION Tannic acid in a single large dose causes prominent, early nucleolar alterations that are readily reversible. The formation SEPTEMBER

1967

of distinct zones of compact RXP granules in a dense protein matrix is the typical change due to tannic acid. The pattern therefore differs significantly from other forms of nucleolar segregation so far reported. It might be postulated that the process involved in the nucleolar alteration is a shrinkage of the nucleolus with spatial separation of normal constituents and eventual segregation into distinct zones, and/or the formation of a new type of nucleolar component that is segregated into distinct areas. Intranucleolar chromatin and invaginations of the nucleoplasm into the nucleolus were masked or obliterated in the process. The term "cap formation" of nucleoli was coined originally by Reynolds et al. (39), who described this change after administra tion of the carcinogen, 4-nitroquinoline-Ar-oxide. The term was applied to granular and fibrillar zonal transformation of the nucleolus with the formation of an additional light or dark area at the periphery simulating a "cap." The process referred to in this report as macrosegregation appears to resemble formation of nucleolar "caps" only superficially. Some of the renal changes (Fig. 18), where there is distinct separation between granular and fibrillar areas, resemble the nucleolar segregation as reported by Simard and Bernhard (45). With tannic acid, the changes in nucleoli of liver cells and of most renal cells differ, however, from the pattern previously designated as nucleolar caps and no transitional stages related to other known capping or segregating agents are seen. Slight variations in morphology in some in stances are probably a function of the dosage used, the degree of involvement, and interval of observations or the organ affected. Nucleolar segregation or the typical nucleolar "cap" occurs after the administration of aflatoxin BI (4, 52) or pyrrolizidine alkaloids (55) with decrease in the granular component, often termed nucleolar exhaustion, and segregation of nucleolar granules and fibrils into distinct zones. In contrast, with tannic acid, abundant and well-packed RNP granules persist in the dark zones, although there is progressive loss of granules in the light areas. Except for the complete segregation in some renal cell nuclei, the tannic acid changes do not form relatively pure segregation between granules and fibrils. With tannic acid and pyrrolizidine alkaloids, the nucleoli recover their normal mor phology within comparatively short intervals. The nucleolar changes with aflatoxin (52) and pyrrolizidines (55) bear a temporal association with decrease in RNA and protein in the nuclear fraction and homogenate of liver. With tannic acid, however, there are no significant alterations in these constituents. Although the explanation is not known at present, it is apparent that severe nucleolar alteration due to tannic acid, despite its resemblance to aflatoxin BI, actinomycin D, and other capping agents, is not necessarily associated with decreased RNA and protein content. It seems probable that these biochemical parameters are insufficiently sensitive or inappropriate to assess the primary functional defect due to tannic acid. In this context, the study of Gelboin et al. (13) demonstrating reversible inhibi tion of RNA polymerase due to aflatoxin BI is of special interest. The dose dependency of the nuclear changes was shown by the inefficacy of lower doses (100-200 mg/kg) used in preliminary studies. In chronic carcinogenesis experiments using lower doses, there were no nucleolar changes (54). Likewise, with aflatoxin and pyrrolizidine alkaloids, nucleolar changes were absent in the liver with chronic but carcinogenic doses (54). The nucleolar 1661

Antonio Rácela, Harold Grady, and Donald Svoboda changes probably represent acute toxic effects which are not necessarily related to the carcinogenic effect of tannic acid. This is also suggested by the presence of identical nucleolar abnor malities in the kidney, an organ not susceptible to tannic acid carcinogenesis. Moreover, as exemplified by actinomycin D (8), during the period of nucleolar capping regeneration was delayed until the nucleolus recovered normal morphology. It is interesting to note that nucleolar capping was not transmitted to successive generations in cultured cells (57). The nucleolar changes are not features of cellular necrosis. Even at the stage where the hepatic cells in all portions of the lobule were affected, the cytoplasm was generally intact, and focal cytoplasmic degradation was minimal or absent. Antitumor drugs such as mitomycin C (25) and anthramycin (unpublished observations) in addition to actinomycin D (16, 35, 37, 38, 42, 47) likewise cause nucleolar capping. Several antimetabolites reported by Simard and Bernhard (45) and UV irradiation of the nucleoplasm (but not of the nucleolus) reported by Montgomery et al. (34) produce indistinguishable nucleolar segregation. It has been maintained by Reynolds et al. (39) and more recently by Simard and Bernhard (45) that nucleolar segregation represents the morphologic expression of fairly specific biochem ical reactions. Of the variety of antimetabolites tested, Simard and Bernhard found that only those that bind with the DNA molecule and interfere with its template activity in DNAdirected RNA synthesis by RNA polymerase produce nucleolar segregation. Antimetabolites that act on nucleic acid precursors and polynucleotides or interfere with protein synthesis do not produce similar nucleolar alterations. This suggestion coincides with the known mechanism of action of capping agents such as actinomycin D (14), mitomycin C (15, 19), 4-nitroquinoline-Aroxide (36), aflatoxin (13, 48), pyrrolizidine alkaloids (10), and UV irradiation (40, 44). The precise mechanism of action of tannic acid on nucleic acids and their synthesis is not known although, on the basis of its effects on nucleolar morphology and in light of Bernhard's hypothesis, it is suggested that tannic acid interferes with transcription at the level of the DNA-RNA axis. The preliminary biochemical data, however, are inconclusive. A group of chemical hepatocarcinogens, diethylnitrosamine (32), dimethylnitrosamine (12), and ethionine (31), are known to alkylate the DNA molecule (28, 29, 49) but have not been shown to produce nucleolar macrosegregation. As previously indicated, the absence of significant acute nucleolar alterations due to diethylnitrosamine may be due to inadequate testing over a sufficiently broad dose range. In addition, alkylation alone may not be a general or sufficient cause of nucleolar macrosegregation but, in addition, interference with activity of RNA polymerase is required. Alterations in RNA polymerase activity, however, seem too nonspecific to provide an accurate explanation of the observed morphologic alterations. Recently, Jézéquel and Steiner observed nucleolar capping in aging tissue culture cells (18), however, the changes were associ ated with Mycoplasma infection (17). Viruses (3) such as those causing ectromelia (27) and molluscum contagiosum (11) cause nucleolar changes described as "spotted" nucleoli or nucleolar "plaques." The "spotted" nucleoli attributed to 5-fluorouracil (24) could be related to viruses in the cultured cells. "Spotted" nucleoli have likewise been observed in some variation resembling 1662

nucleolar segregation in nucleoli of liver cells of rats in protein deficiency (53) and in p-dimethyl-aminoazobenzene-induced liver tumors (51). In both instances, dark plaques or condensed fibrils segregated at the nucleolar periphery. Similar plaques were observed with ultraviolet irradiation of whole Chang liver cells or their nucleoli (34), total body irradiation of the rat (33), pyrrolizidine alkaloids (Fig. 16) and actinomycin D (38). Con sidering the variety of experimental manipulations which cause a similar nucleolar alteration, it was interesting to note that "plaques" or "spots" were observed during a stage of develop ment of macrosegregation due to tannic acid (Figs. 14, 15). It is possible that nucleolar caps or macrosegregation and nucleolar "spots" or plaques are related to the same biochemical defect in all instances though, in most cases, parallel biochemical study has not been pursued and such questions as (a) whether all the agents and conditions causing these alterations do so through a common mechanism and (ft) the identification of the biochemical and functional counterpart(s) of the abnormalities in nucleolar ultrastructure remain unanswered at the present time. In this context, and considering the scant information available about ultrastructure of the nucleus compared to the cytoplasm, it would appear desirable to establish the effects of carcinogenic agents, singly or in combination, on nuclear ultrastructure in order to establish a baseline of the morphologic responses in nuclei for further detailed investigation and assessment of the biologic significance of such responses. ACKNOWLEDGMENTS

We wish to acknowledge the technical assistance of Claire Nielson, Faye Brady, Kay Balle, and Lynne Schmutz. ADDENDUM

Since submission of this manuscript, it has become evident that certain of the nitrosamines, particularly dimethylnitrosomine, cause a variety of nucleolar abnormalities in acute studies. Also, Miyai and Steiner (Lab. Invest., 16: 677-692, 1967)and Shinozuka and Farber (Federation Proc., 26: 575, 1967)have demonstrated a form of nucleolar disorganization following acute ethionine in toxication. REFERENCES 1. Arhelger, R. B., Broom, J. S., and Boler, R. K. Ultrastructural Hepatic Alterations Following Tannic Acid Administration to Rabbits. Am. J. Pathol., 46:409-434,1965. 2. Baker, R. D., and Handler, P. Animal Experiments with Tannic Acid. Ann. Surg., 118:417-426,1943. 3. Bernhard, W. Some Problems of Fine Structure in Tumor Cells. Progr. Exptl. Tumor Res., S: 1-34, 1962. 4. Bernhard, W., Frayssinet, C., LaFarge, C., and le Breton, E. LésionsNucléolairesPrécocesProvoquéespar l'Afiatoxine dans les Cellules Hépatiquesdu Rat. Compt. Rend. Acad. Sci. Paris, ê61:1785-1788,1965. 5. Bernhard, W., and Granboulan, N. The Fine Structure of the Cancer Cell Nucleus. Exptl. Cell Res., Suppl. 9, pp. 19-53, 1963. 6. Boler, R. K., Broom, J. S., and Arhelger, R. B. Ultrastructural Renal Alterations Following Tannic Acid Administration to Rabbits. Am. J. Pathol., 4P: 15-32,1966. 7. Cameron, G. R., Milton, R. F., and Allen, J. W. Toxicity of CANCER RESEARCH VOL. 27

Ultrastructural Nuclear Changes Due to Tannic Acid Tarinic Acid. An Experimental Investigation. Lancet, 2: 179186, 1943. 8. Chiga, M., Kume, F., and Millar, R. C. Nucleolar Alteration Produced by Actinomycin D and the Delayed Onset of Hepatic Regeneration in Rats. Lab. Invest., IB: 1403-1408, 1966. 9. Cook, J. W., Duffy, E., and Shoental, R. Primary Liver Tu mours in Rats Following Feeding with Alkaloids of Senecio jacobea. Brit. J. Cancer, 4: 405-410, 1950. 10. Culvenor, C. C. J., Dann, A. T., and Dick, A. T. Alkylation as the Mechanism by Which the Hepatotoxic Pyrrolizidine Alka loids Act on Cell Nuclei. Nature, 195: 570-573,1962. 11. Dourmashkin, R., and Bernhard, W. A Study with the Elec tron Microscope of the Skin Tumour of Molluscum Contagiosum. J. Ultrastruct. Res., 3:11-38,1959. 12. Emmelot, P., and Benedetti, E. L. Changes in the Fine Struc ture of Rat Liver Cells Brought About by Dimethylnitrosamine. J. Biophys. Biochem. Cytol., 7: 393-396, 1960. 13. Gelboin, H. V., Wortham, J. S., and Wilson, R. G. Rapid and Marked Inhibition of Rat-liver RNA Polymerase by Aflatoxin B,. Science, 154:1205-1206, 1966. 14. Goldberg, I. H., and Reich, E. Actinomycin Inhibition of RNA Synthesis Directed by DNA. Federation Proc., 23: 958964, 1964. 15. Iyer, V. N., and Szybalski, W. A. A Molecular Mechanism of Mitomycin Action: Linking of Complementary DNA Strands. Proc. Nati. Acad. Sci. U.S., 50:355-362,1963. 16. Jézéquel, A. M., and Bernhard, W. Modifications Ultrastruc turales du Pancreas Exocrine de Rat sous PEffect de l'Actinomycine D. J. Microscop., 3: 279-296,1964. 17. Jézéquel, A. M., Shreeve, M. M., and Steiner, J. W. Segrega tion of Nucleolar Components in Mycoplasma-infected Cells. Lab. Invest., 16: 287-304, 1967. 18. Józéquel, A. M., and Steiner, J. W. Nucleolur Caps—A Sponta neous Form of Cellular Degeneration? Federation Proc., 25: 667, 1966. 19. Kersten, H. Action of Mitomycin C on Nucleic Acid Metabo lism in Tumor and Bacterial Cells. Biochim. Biophys. Acta, 55: 558-560, 1962. 20. Korpassy, B. The Hepatocarcinogenicity of Tannic Acid. Cancer Res., 19: 501-504,1959. 21. Korpassy, B. Tannins as Hepatic Carcinogens. Progr. Exptl. Tumor Res., 2: 245-290, 1961. 22. Korpassy, B., and Mosonyi, M. Influence of Dietetic Factors on Carcinogenic Acitivity of Tannic Acid. Lancet, /: 14-16, 1951. 23. Kraybill, H., and Shimkin, M. Carcinogenesis and Contami nated Foods. In: A. Haddow and S. Weinhouse (eds.), Ad vances in Cancer Research, Vol. 8, pp. 191-248. New York: Academic Press, 1964. 24. Lapis, K., and Benedeczky, I. Antimetabolite-Induced Changes in the Fine Structure of Tumour Cells. Acta Biol. Acad. Sci. Hung., 17: 199-215, 1966. 25. Lapis, K., and Bernhard, W. The Effect of Mitomycin-C on the Nucleolar Fine Structure of KB Cells in Cell Culture. Cancer Res., 25: 628-645, 1965. 26. Layne, E. Spectrophotometric and Turbidometric Methods for Measuring Proteins. In: S. P. Colowick and N. O. Kaplan (eds.), Methods in Enzymology, Vol. 3, pp. 448-450. New York and London: Academic Press, 1957. 27. Leduc, E. H., and Bernhard, W. Electron Microscope Study of Mouse Liver Infected by Ectromelia Virus. J. Ultrastruct. Res., 6: 466-488, 1962. 28. Magee, P. N. Cellular Injury and Chemical Carcinogenesis by N-Nitroso Compounds. In: R. W. Raven (ed.), Cancer Prog ress, pp. 56-66. London: Butterworth, 1963. SEPTEMBER

1967

29. Magee, P. N., and Farber, E. Methylation of Rat-Liver Nu cleic Acids by Dimethylnitrosamine in vivo. Biochem. J., S3: 114-124, 1962. 30. Marinozzi, V. Cytochimie Ultrastructurale du Nucléole—RNA et Protéines Intranucleolaires. J. Ultrastruct. Iles., 10: 433456, 1964. 31. Miyai, K., and Steiner, J. W. Fine Structure of Interphase Liver Cell Nuclei in Subacute Ethionine Intoxication. Exptl. Mol. Pathol., 4: 525-566, 1965. 32. Molbert, E., Hill, K., and Buchner, F. Die Kanzerisierung der Leberparenchymzelle durch Diaethylnitrosamine im elektronenmikroskopischen Bild. Beitr. Pathol. Anat. Allgem. Pathol., 126: 218-242, 1962. 33. Montgomery, P. O'B., Karney, D., Reynolds, R. C., and McOlendon, D. Cellular and Subcellular Effects of Ionizing Radiations. Am. J. Pathol., 44: 727-746,1964. 34. Montgomery, P. O'B., Reynolds, R. C., and Cook, J. E. Nu cleolar Caps Induced by Flying Spot Ultraviolet Nuclear Irradiation. Am. J. Pathol., 49:555-564,1966. 35. Oda, A., and Chiga, M. Effect of Actinomycin D on the Hepatic Cells of Partially Hepatectornized Rats. An Electron Microscopic Study. Lab. Invest., 14:1419-1427,1965. 36. Pa 1, J. S., Reynolds, R. C., and Montgomery, P. O'B. Studies on the Biochemical Effects of 4-Nitroquinoline-N-oxide. J. Cell Biol.,3i:83A, 1966. 37. Reynold, R. C., and Montgomery, P. O'B. Nucleolar and Cytoplasmic Alterations Produced by Actinomycin D and Other Metabolic Inhibitors. An Electron Microscopic and Time-lapse Study. Proc. Am. Assoc. Cancer Res., 6: 53, 1965. 38. Reynolds, R. C., Montgomery, P. O'B., and Hughes, B. Nu cleolar "Caps" Produced by Actinomycin D. Cancer Res., 24: 53, 1965. 39. Reynolds,

R. C., Montgomery,

P. O'B., and Karney,

D. H.

Nucleolar Caps, A Morphologic Entity Produced by the Car cinogen 4-Nitroquinoline-N-Oxide. Cancer Res., S3: 535-538, 1963. 40. Rupert, C. S. Questions Regarding the Presumed Role of Thymine Dimers in Photoreactivable Ultraviolet Damage to DNA. Photochem. Photobiol., 3: 399-103, 1964. 41. Schneider, W. Determination of Nucleic Acids in Tissues by Pentose Analysis. In: S. P. Colowick and N. O. Kaplan (eds.), Methods in Enzymology, Vol. 3, pp. 680-684. New York and London: Academic Press, 1957. 42. Schoefl, G. I. The Effect of Actinomycin D on the Fine Struc ture of the Nucleolus. J. Ultrastruct. Res., 10: 224-243, 1964. 43. Schwartz, H. S., Sternberg, S. S., and Philips, F. S. Pharmacol ogy of Mitomycin C. IV. Effects in vivo on Nucleic Acid Syn thesis. Comparison with Actinomycin D. Cancer Res., 23: 1125-1136, 1963. 44. Setlow, J. E. Effect of Ultraviolet on DNA: Correlations Among Biological Changes, Physical Changes and Repair Mechanism. Photochem. Photobiol., 3: 405-413, 1904. 45. Simard, R., and Bernhard, W. Le Phénomène de la Ségrégation Nucléolaire:Spécificité d'Action de Certains Antimétabolites. Intern. J. Cancer, /: 463-479,1966. 46. Smetana, K., Narayan, K. S., and Busch, H. Quantitative Analysis of Ultrastructural Components of Nucleoli of the Walker Tumor and Liver. Cancer Res., 26:786-796,1966. 47. Smuckler, E. A., and Benditt, E. P. The Early Effects of Actinomycin on Rat Liver. Lab. Invest., 14:1699-1709,1965. 48. Sporn, M., Dingman, C., Phelps, H., and Wogan, G. Aflatoxin Bi: Binding to DNA in vitro and Alteration of RNA Metabo lism in vivo. Science, 151:1539-1541,1966. 49. Stekol, J. A., Mody, U., and Perry, J.The Incorporation of the 16(0

Antonio Rocela, Harold Grady, and Donald Svoboda

50. 51.

52.

53.

Carbon of the Ethyl Group of Ethionine into Liver Nucleic Acids and the Effect of Ethionine Feeding on the Content of Nucleic Acids in Rat Liver. J. Biol. Chem., 2S5: PC59-PC60, 1960. Stoweil, R. E. Summary of Informal Discussion. Exptl. Cell Res., Suppl. 9. 107-110, 1963. Svoboda, D. J. Fine Structure of Hepatomas Induced in Rats with p-Dimethylaminoazobenzene. J- Nati. Cancer Inst., 33: 315-339, 1964. Svoboda, D. J., Grady, H. J., and Higginson, J. Aflatoxin Bi Injury in Rat and Monkey Liver. Am. J. Pathol., Ifi: 10231051, 1966. Svoboda, D., and Higginson, J. Ultrastructural Changes Produced by Protein and Related Deficiencies in the Rat Liver. Am. J. Pathol., 45:353-379,1964.

1664

54. Svoboda, D., Rácela, A., and Higginson, J. Variations in Ultrastructural Nuclear Changes in Hepatocarcinogenesis. Biochem. Pharmacol., 16: 651-657, 1967. 55. Svoboda, D. J., and Soga. J, Early Effects of Pyrrolizidine Alkaloids on the Fine Structure of Rat Liver Cells. Am. J. Pathol., 48: 347-373, 1966. 56. Swift, H. Cytochemical Studies on Nuclear Fine Structure. Exptl. Cell Res., Suppl. 9, pp. 54-67, 1963. 57. Tanaka, T., Tanaka, K. K., and Kinosita, R. Uridine-H Autoradiographic Study of Nucleolar Cap of 7,12-Dimethylbenzanthracene-Treated Cells. Ninth International Cancer Con gress, October 23-29, Abstracts of Papers, p. 103. Tokyo, Japan, 1966. 58. Widnell, C., and Tata, J. Isolation of Enzymatically Active Rat Liver Nuclei. Biochem. J., 92: 313-317, 1964.

CANCER RESEARCH VOL. 27

Ultrastructural

Nuclear Changes Due to Tannic Acid

FIGS. 1-18. Unless otherwise indicated, the sections were taken from liver of animals given a single subcutaneous injection of tannic acid at a dose of 700 mg/kg body weight. FIG. 1. Six hours. The nucleoli (arrows) are small, and transformed into fairly homogenous masses with one or more dark zones in their interior. This nucleolar alteration is seen in all portions of the hepatic lobule, s, sinusoid. Epon-embedded tissue. Azure A, X (MK). FIG. 2. Six hours. A high power view of a hepatocyte showing two nucleoli with nucleolar macrosegregation. The small dark zones in the interior of the nucleolus contrast with the remainder which is less dense and fairly homogenous. (Refer to Figs. 6, 7, 10, and 11 for ultrastructural comparison.) Epon-embedded tissue. Azure A, X 1480. FIG. 3. Six hours. The histology is generally intact with no areas of necrosis. The nucleolar alteration is not evident in formal in-fixed, paraffin-embedded preparations, cv, central vein; p, portal area. H & E, X 187. FIG. 4. Six hours. Occasional hepatocytes contain an intranuclear inclusion (I), surrounded by a halo. The inclusion body stains posi tively for iron, and is only very faintly pyroninophilic. n, nucleolus. Methyl green pyronine, X 1080. FIG. 5. Normal nucleolus, liver cell. The nucleolus is composed of granular (g) and fibrillar (/) components which are disposed in irreg ular, intertwining strands. The relatively "clear" areas in the nucleolus have the same features as the nucleoplasm. noe, nucleolusassociated chromatin; peg, perichromatin granule. Glutaraldehyde-osmium fixation, lead stain, X 77,000. FIG. 6. Liver cell, 6 hours. Dark zones (d) composed of densely packed ribonucleoprotein granules appear and contrast with the sur rounding light zone (!) which is composed of dispersed fibrils and granules. There are multiple, small inclusion bodies (/) composed of large granules (200-400 A in diameter) in the nucleoplasm. These are interpreted as precursors of the iron-containing bodies (see Figs. 4, 12, 13). Osmium fixation, lead stain, X 28,800. FIG. 7. Six hours. The liver cell nucleolus is small and compact with almost complete obliteration of the clear areas of nucleoplasm or intranueleolar chromatin. In the light zone (1), there is early segregation of granular (g) and fibrillar (/) components. Three well demarcated dark zones (d) of packed ribonucleoprotein granules are present. Osmium fixation, lead stain, X 28,800. Fig. 8. Six hours. In the early stage of KNase incubation, the dark zones (d) of densely packed ribonucleoprotein granules are digested leaving a fairly homogenous protein matrix. The light zone (/) of granules and fibrils is only partially digested, nao, nucleolus-associated chromatin; arrow, intranueleolar chromatin. Glutaraldehvde fixation, uranyl acetate-lead stain, X 72,000. Fig. 9. Six hours. The nucleolus is small and compact. In the light zone (l): granules (g) are abundant and generally dispers ed, d, dark zone of ribonucleoprotein granules. Osmium fixation, uranyl acetate stain, X (i(i,000. FIG. 10. Six hours. The granules in the light zone (/) are fewer than in Fig. 9. d, dark zone of ribonucleoprotein granules; nac, nucleo lus-associated chromatin. Glutaraldehyde-osmium fixation, lead stain, X 28,500. FIG. 11. Twelve hours (1200 mg/kg). The nucleolus is small and the light zone (I) is made up almost exclusively of fibrils, d, dark zone of ribonucleoprotein gramiles; icg, interchromatin granules. Osmium fixation, lead stain, X 28,800. FIG. 12. Twelve hours. In some nuclei, the nucleolus (n) has recovered its normal morphology. The inclusion body (7) composed of large granules corresponds to the inclusions seen by light microscopy (Fig. 4). Osmium fixation, lead stain, X 7,000. FIG. 13. Twelve hours. The iron-containing inclusion body (/) is surrounded by a space or a defect, probably due to enlargement of the interchromatin space (ics). This space is the ultrastructural equivalent of the halo around the inclusion body seen by light micros copy (Fig. 4). n, nucleolus. Osmium fixation, lead stain, X 8,300. FIG. 14. Eighteen hours. The nucleolus is dispersed and a dark zone of ribonucleoprotein granules (d) is still present. The aggregates of fibrillar (/) and granular components (g) resemble the normal pattern, though the granules often form a well-defined area. Small, multiple dense plaques (p) are irregularly distributed, giving the nucleolus a "spotted" appearance. Osmium fixation, lead stain, X 66,000. FIG. 15. Eighteen hours. The "spotted" appearance is more prominent and the plaques (p) appear to be fibrillar. As in the previous figure, the granular component (g) forms a large, well-defined area. The fibrillar component (/) is located mainly at the periphery. Osmium fixation, lead stain, X 22,000. FIG. 16. Lasiocarpine, 80 mg/kg, l hour. The nucleolus is dispersed with distinct granular (g) and fibrillar (/) areas. Dense plaques (p) similar to those due to tannic acid (Figs. 14, 15) are present. This alteration preceded the typical nucleolar capping produced by lasiocarpine. Osmium fixation, lead stain, X 17,000. FIG. 17. Eighteen hours. The nucleolus in some cells approximates the normal, although the granular (g) and fibrillar (/) components often form distinct areas. Interchromatin granules (icg) are increased. The iron-containing inclusion bodies (7) are still present. Osmium fixation, lead stain, X 11,400. FIG. 18. Renal tubular epithelium, 6 hours. Light (?) and dark (d) macrosegregation is observed in the kidney. In contrast to the liver cell, granular (d) and fibrillar areas (I) are segregated into relatively pure areas.

SEPTEMBER

1967

1;->

Antonio Rácela, Harold Grady, and Donald Svoboda

. . •; * v .•*'..•»•• **,--* O'<

1666

CANCER RESEARCH VOL. 27

-d .I '

SEPTEMBER

1967

1667

Antonio Rácela, Harold Gradi], and Donald Svoboda ;•

;* *

E

.nac

I V*.

''••'.' t». 166S

CANCER RESEARCH VOL. 27

Ultrastructural

Nuclear Changes Due to Tannic Acid

•. - •

l

' i

SEPTEMBER

1967

Antonio Rácela, Harold Grady, and Donald Svoboda

1670

CANCER RESEARCH VOL. 27

Ultrastructural

Nuclear Changes Due to Tannic Acid

•W*

;*

*S

• ¿t '

t ->v «£. , : •» ;-

*

• *"'*'J¿ •

W.'~ï

^fc*%^fer

••* ••>,./• v,--«^^ "iwV.rv T* "'%-

SEPTEMBER 1967

'wiiì

1671