Protection from drug-induced hepatocellular changes by pretreatment with conjugating enzyme inhibitors in rats.

The present paper describes the role of conjugating enzymes in the development of hepatotoxicity after administration of repeated doses of a novel monoamine oxidase type-A (MAO-A) inhibitor, (5R)-3-[2-(( 1S)-3-cyano-1-hydroxypropyl)benzothiazol-6-yl]-5-methoxymethyl-2-oxazolidinone (E2011). The effects of pretreatment with three kinds of conjugating enzyme inhibitors on hepatic lesions induced by E2011 were evaluated in female Sprague-Dawley rats. The inhibitors used were 2,6-dichloro-4-nitrophenol (DCNP; inhibitor of sulfotransferase (ST)), pentachlorophenol (PCP; inhibitor of both ST and acetyltransferase (AT)) or ranitidine (inhibitor of UDP-glucuronosyltransferase (UDP-GT)). Two weeks treatment of E2011 alone at an oral dosage of 150 mg/kg induced hepatocellular changes characterized by nuclear enlargement. Daily pretreatment with DCNP (10 mg/kg, i.p.) enhanced the E2011-induced hepatocellular changes accompanied by single cell necrosis. On the other hand, the hepatotoxicity was clearly diminished by PCP (5 mg/kg, i.p.). Ranitidine pretreatment had no effect. Protection by PCP was attributed to the inhibitory effects of AT in addition to ST; it was considered that the hepatocellular changes caused by E2011 were largely dependent on the formation of acetyl conjugate(s).

Assessment of potential mutagenic activities of a novel benzothiazole MAO-A inhibitor E2011 using Salmonella typhimurium YG1029.

The potential initiation activities of a novel monoamine oxidase type-A (MAO-A) inhibitor E2011, which induced preneoplastic foci in the rat liver, were investigated by comparing the mutagenic activity of E2011, 6-aminobenzothiazole (6-ABT, a structural scaffold of E2011) and its derivatives, which are suggested primary reactive metabolites for E2011-induced hepatotoxicity in the rats in vivo, in the Ames assay system employing two Salmonella tester strains, TA100 and YG1029, a bacterial O-acetyltransferase-overproducing strain of TA100. E2011, a tertiary amine, showed no mutagenic activity both in the Salmonella typhimurium TA100 and YG1029 with and without S9 mix. On the other hand, a secondary aromatic amine ER-174238-00, a typical decarbonated metabolite of E2011, showed weak but significant mutagenicity in YG1029 in the presence of S9 mix, and a primary aromatic amine ER-174237-00, an N-dealkylated derivative of ER-174238-00, exhibited S9-dependent potent mutagenicity in YG1029. Thus, it appears that primary and secondary amino moieties of benzothiazole derivatives at C(6)-position are the specific structures contributing to their mutagenic activity. In addition, the alkyl group at C(2)-position of E2011, ER-174237-00 and ER-174238-00 is suggested to intensify the mutagenic activity, since the mutagenicity of ER-174237-00 is approximately two-fold higher than that of 6-ABT, which has hydrogen at C(2)-position in the place of the alkyl group. These results strongly suggest that E2011 has potential initiation activities in the rat liver in vivo after undergoing decarbonation, one of the metabolic pathways, at the carbonyl moiety of oxazolidinone ring to form mutagenic amine(s).

Comparison of the mutational spectra of the lacZ transgene in four organs of the MutaMouse treated with benzo[a]pyrene: target organ specificity.

We recently demonstrated that not all organs with a high rate of induction of mutation in the lacZ transgene develop tumors in the lambdalacZ transgenic mice (MutaMouse) used for a long-term carcinogenicity study with benzo[a]pyrene (BP). To better understand the role of chemical-induced in vivo mutations in carcinogenesis, we compared the mutational spectra of the lacZ transgene in four organs of the MutaMouse obtained 2 weeks after five daily consecutive oral treatments with 125 mg/kg/day BP. lacZ transgenes were analyzed in two target organs (forestomach and spleen) and two non-target organs (colon and glandular stomach) for BP-induced carcinogenesis in MutaMouse, and all of these organs were highly mutated in the lacZ transgene. The sequence data showed similar mutational spectra of the lacZ transgene between the two target organs; the predominant mutations were G:C-->T:A transversions (55% and 50% for forestomach and spleen, respectively), followed by deletions (20% and 21% for forestomach and spleen, respectively) mainly at G:C site. The frequent G:C-->T:A transversions are consistent with reports of the mutational spectra produced in the p53 gene in tumors generated in rats and mice exposed to BP. In contrast, the mutational spectra of the lacZ transgene in the two non-target organs are different from those in the target organs, and are also suggested to differ from one another. These findings suggest an organ/tissue-specific mechanism of mutagenesis.

Toxicological response of rats to a novel monoamine oxidase type-A inhibitor, (5R)-3-[2-((1S)-3-cyano-1-hydroxypropyl)benzothiazol-6-yl]-5- methoxymethyl-2-oxazolidinone (E2011), orally administered for 13 weeks.

(5R)-3-[2-((1S)-3-cyano-1-hydroxypropyl)benzothiazol-6-yl]-5- methoxymethyl-2-oxazolidinone (E2011) is a novel monoamine oxidase type-A (MAO-A) inhibitor. In order to assess toxicological profiles of E2011, doses of 0 (as controls), 30, 100 mg/kg of E2011 were administered to male and female Sprague-Dawley rats once a day for 13 weeks orally by gavage. No mortality or any toxic signs except salivation occurred due to E2011 treatment. Decreased body weight gain and food consumption, increases of alkaline phosphatase and increases of liver weight were the major treatment-related findings observed predominantly in the 100 mg/kg group. Histological examination revealed nuclear enlargement of hepatocytes with appearance of altered cell foci in some cases, and acinar atrophy in Harderian glands in the 100 mg/kg group. Since the histopathological findings in the liver were indicative of an ongoing carcinogenic process, glutathione S-transferase placental form (GST-P) positive hepatic foci were identified immunohistochemically and examined morphometrically. Although GST-P positive hepatic foci were detected in all groups including controls, the number and area of GST-P positive hepatic foci were significantly higher in female rats treated with 100 mg/kg than those in controls. In this paper, possible mechanisms of specific lesions in the liver and Harderian glands will be discussed.

Multiple organ mutation in the lacZ transgenic mouse (Muta mouse) 6 months after oral treatment (5 days) with benzo[a]pyrene.

We have recently demonstrated that not all organs with high rates of mutation in the lacZ transgene develop tumors using the Muta Mouse. To better understand the role of in vivo mutation in carcinogenesis, we examined the mutant frequencies (MF) of the lacZ transgene in tumor-bearing and non tumor-bearing organs. MF, recovered after 2 weeks (the data taken from our previous study) and after 26 weeks following oral doses of 125 mg kg-1 day-1 benzo[a]pyrene (BP) for five days were compared. The organs examined included the target organs (forestomach, spleen, and lung) and non-target organs (colon, glandular stomach, and liver) for BP carcinogenesis. The data indicated that lacZ MF were markedly increased over spontaneous frequencies in the organs examined and that the organ which showed the highest MF was the colon, followed by the forestomach>spleen>glandular stomach, liver, and lung in that order. These findings indicate that the MF of the lacZ transgene in each organ, even 26 weeks after the start of the treatment does not fully correlate with the known target organs of BP. Furthermore, the lacZ MF in a non-papilloma region of a forestomach with a papilloma was equivalent to the two highest MF observed in the healthy colon (non-target organ) of mice at 26 weeks. These observations also indicate that the generation of tumors requires the induction of mutations as well as other factor(s) specific to the target organs. These results clearly suggest that highly mutated organs do not always progress to tumors in the transgenic mouse.

Mechanism of weight gain suppressing effect of ER-40133, an angiotensin I converting enzyme inhibitor, in growing rats.

Effects of ER-40133, an inhibitor of angiotensin converting enzyme (ACE), on weight gain and sodium and potassium balance were studied in growing SD male rats. Thirty-two animals (seven weeks of age) were divided into two groups; one received a standard diet containing 0.227% sodium and the other a low (0.065%) sodium diet. They were divided into four subgroups; one control group and three treated groups receiving 3, 10 or 30 mg/kg of ER-40133, by gavage, once a day for five consecutive days. Body weight gain (average of the standard and low sodium diet groups) was -32% in the 3 mg/kg group,-74% in 10 mg/kg group and -99% in 30 mg/kg group, when compared with the control group. There was a highly linear correlation between suppression of body weight gain and reduction in sodium and potassium retention for both groups of animals given the standard and low sodium diet. The reduced sodium retention, the primary effect of ACE inhibitors, accounted for about 30% of suppressed weight gain, and the reduced potassium retention, the secondary effect of sodium deficiency, could account for the rest about 70% of weight suppression by ER-40133.

Toxicity profile of benzo[a]pyrene in the male LacZ transgenic mouse (MutaMouse) following oral administration for 5 consecutive days.

The toxicity profile of benzo[a]pyrene (BP) was examined in the MutaMouse. The transgenic mouse integrated with lambda gt10 lacZ vectors is used worldwide as an experimental animal in in vivo mutagenesis testing systems. There are few toxicity studies including carcinogenicity in the MutaMouse, and so far only a few carcinogenicity studies of BP accompanied with hematological and plasma biochemical examinations have been conducted even in generic mice. Accordingly, male mice were orally administered BP at doses of 75 and 125 mg/kg/day for 5 consecutive days, and complete autopsy was conducted together with pathological, hematological, and plasma biochemical examinations and measurement of organ weights 41 weeks after the last treatment. Squamous cell papilloma and hyperplasia in the forestomach were induced at incidences of 25 and 50%, respectively and were induced 26 weeks after the final treatment without any significant alterations in t he hematological and plasma biochemical parameters in mice of the 125 mg/kg/day BP-treated satellite group. Fourty-one weeks after the final treatments, 75 and 125 mg/kg/day BP induced squamous cell carcinoma, papilloma, and hyperplasia in the forestomach at incidences of 18 and 18%, 36 and 45%, and 91 and 91%, respectively, and anemia possibly due to continuous hemorrhage from tumors in the forestomach. BP (125 mg/kg/day) also produced malignant lymphoma with an incidence of 18%, accompanied by a marked increase in leukocyte count and decrease in erythrocyte count and by a remarkable decrease in body weights 26 and 39 weeks after the last treatment. Moreover, administration of 75 and 125 mg/kg/day BP induced bronchiolar-alveolar hyperplasia in the lung at incidences of 18 and 9%, respectively. Slight increases were also observed in the weight of the liver and in the levels of urea nitrogen, creatinine, and potassium ion in the plasma biochemical examinations, although no significant pathological alterations were found in the liver and kidney. This study provides new information about BP toxicity including carcinogenicity in the MutaMouse developed for in vivo mutational analysis.

Microtubule antagonists activate programmed cell death (apoptosis) in cultured rat hepatocytes.

We investigated the mechanism of lethal injury following the disruption of microtubules in cultured hepatocytes treated with vinblastine (VBL) or colchicine (COL). These agents kill hepatocytes by a process readily distinguished from two well-known pathways that lead to a loss of viability, namely, oxidative stress and inhibition of mitochondrial electron transport. Cell killing with VBL and COL was accompanied by fragmentation of DNA. Both the loss of viability and the fragmentation of DNA were prevented by the inhibition of protein synthesis within 6 hours following exposure to VBL or COL. Cell death and the fragmentation of DNA were also prevented when Ca2+ was removed from the culture medium. By contrast, the inhibition of protein kinase C prevented cell killing by VBL or COL, but did not alter the extent of DNA fragmentation. The requirements here for protein synthesis, extracellular Ca2+, and protein kinase C activity define a model of apoptosis, or programmed cell death, that seems to involve mechanisms that can be dissociated from the fragmentation of DNA.

Differing effects of the inhibition of poly(ADP-ribose) polymerase on the course of oxidative cell injury in hepatocytes and fibroblasts.

The effects of the two inhibitors of poly(ADP-ribose) polymerase, 3-aminobenzamide (ABA) and benzamide (BA), on the oxidative killing of L929 mouse fibroblasts and primary cultures of rat hepatocytes were studied. The killing of L929 cells by tert-butyl hydroperoxide (TBHP) occurred by two mechanisms, one sensitive and the other insensitive to the antioxidant N,N'-diphenylphenylene diamine (DPPD). Cell killing by either mechanism was prevented by the ferric iron chelator deferoxamine. ABA and BA prevented the killing of L929 cells that occurred in the presence, but not in the absence, of DPPD. ABA and BA inhibited the activity of poly(ADP-ribose) polymerase by 85%. Protection was accompanied by the sparing of the depletion of both NAD and ATP, but there was no effect of either ABA or BA on the iron-dependent appearance of single-strand breaks in DNA. Depletion of ATP by treating the fibroblasts with 2-deoxyglucose and sodium azide did not result in any loss of viability. H2O2 similarly killed the L929 cells by a mechanism that depended on a source of ferric iron. However, DPPD had no effect on the cell killing, and ABA and BA completely protected the cells in the presence or absence of DPPD. H2O2 caused the appearance of single-strand breaks that were prevented by deferoxamine, but again not by ABA or BA. ABA and deferoxamine reduced, but did not prevent, the depletion of both NAD and ATP occurring with H2O2. With the cultured hepatocytes, ABA and BA inhibited poly(ADP-ribose) polymerase at concentrations that were without effect on either the extent of cell killing or the depletion of NAD occurring with either TBHP, H2O2, or menadione. These data indicate that the relationship between oxidative DNA damage and the genesis of lethal injury is very different in the two types of cells. In the fibroblasts, the appearance of single strand breaks in DNA was accompanied by depletion of NAD and ATP and subsequently by the death of the cells. These events were mediated by the activity of poly(ADP-ribose) polymerase, as inhibition of the enzyme prevented their development. In the hepatocytes, inhibition of poly(ADP-ribose) polymerase was without effect on the oxidative death of the cells.

Effect of a novel 5-lipoxygenase inhibitor, E6080 on bronchospasm, airway cellular infiltration and leukotriene production in guinea pigs.

6-Hydroxy-2-(4-sulfamoylbenzylamino)-4,5,7-trimethylbenzothiazo le hydrochloride (E6080), an orally active and selective 5-lipoxygenase inhibitor, dose-dependently inhibited the bronchospasm induced by antigen (ovalbumin) inhalation in sensitized conscious guinea pigs. The inhibitory effect of E6080 was more potent than that of a typical 5-lipoxygenase inhibitor, AA861, but less than that of a leukotriene (LT) antagonist, LY171883. When airway infiltration of neutrophils and eosinophils was measured in bronchoalveolar lavage fluid (BALF) at 6 h after antigen inhalation by passively sensitized guinea pigs, the inhibitory effect of E6080 on neutrophil infiltration was more marked than that on eosinophil infiltration. The inhibitory effect of E6080 on bronchoalveolar cellular infiltration and bronchoepithelial damage was confirmed by examination of photomicrographs of the lung. In addition to the above pharmacological effects, E6080 inhibited the increase in BALF levels of both i-LTC4 and i-LTB4. These results suggest that E6080 may prove to be effective for the treatment of asthma, in which large amounts of leukotrienes (LTs) are elaborated.

Involvement of tumor necrosis factor-alpha in development of hepatic injury in galactosamine-sensitized mice.

Intravenous injection of lipopolysaccharide and D-galactosamine, at doses of 0.2 micrograms/kg and 800 mg/kg, respectively, elicited massive hepatic necrosis within 24 hr in C3H/HeN mice. The plasma L-alanine aminotransferase (ALT, E.C. 2.6.1.2) or L-aspartate aminotransferase (AST, E.C. 2.6.1.1) activities at this point reached more than 2,000 IU/L. However, overt hepatic injury as evaluated by the plasma aminotransferase activities did not develop in mice in which only lipopolysaccharide or only D-galactosamine was injected. No tumor necrosis factor-like activities could be detected in the plasma of galactosamine- and lipopolysaccharide-injected mice as determined by the assay of cytotoxicity to highly tumor necrosis factor-sensitive L-P3 cells through the experimental period of 24 hr. However, passive immunization against mouse tumor necrosis factor-alpha with polyvalent rabbit anti-mouse tumor necrosis factor-alpha antiserum, which was able to neutralize the cytotoxic effects of recombinant mouse tumor necrosis factor-alpha on L-P3 cells, could protect the mice from the development of hepatic injury in a dose-dependent manner. Simultaneous injection of recombinant human tumor necrosis factor-alpha, instead of lipopolysaccharide, with 800 mg/kg of D-galactosamine in lipopolysaccharide-resistant C3H/HeJ mice sensitized the animals more than one thousand-fold to the development of hepatic injury. The livers appeared to be morphologically similar to those of galactosamine- and lipopolysaccharide-injected C3H/HeN mice.

Involvement of tumor necrosis factor-alpha in the pathogenesis of activated macrophage-mediated hepatitis in mice.

The possible involvement of tumor necrosis factor-alpha in the pathogenesis of an experimentally induced hepatitis was investigated. Balb/c mice were primed with Propionibacterium acnes to induce the infiltration of mononuclear cells into the liver. Immunohistochemical study showed that most of the accumulated mononuclear cells at 7 days were Mac-2 positive, suggesting that they were activated macrophages. An injection of lipopolysaccharide resulted in massive hepatic necrosis and high mortality in the mice within 24 hours. Plasma tumor necrosis factor-alpha activity initially rose sharply and then declined over 3 hours. The increase in plasma aminotransferase activity correlated well with the elevation of plasma tumor necrosis factor-alpha activity. Pretreatment with dexamethasone or 16,16-dimethyl-prostaglandin E2 attenuated not only the elevation of plasma tumor necrosis factor-alpha activity but also the increase in plasma aminotransferase activity and improved the survival rate. Passive immunization against tumor necrosis factor-alpha showed protective effects. These findings suggest that tumor necrosis factor-alpha released from activated macrophages may play a crucial role in the pathogenesis of this murine hepatitis.

Prophylactic activity of dihydroheptaprenol, a synthetic polyprenol derivative, against Sendai virus infection in mice.

The effect of a chemically synthesized polyprenol derivative, dihydroheptaprenol (DHP), on the non-specific resistance of mice to Sendai virus infection was investigated. The mice that received 200 micrograms of DHP intranasally twice, at 3 days and 1 day before the infection, showed a significant protection against Sendai virus infection. Treatment of mice twice even with as much as 2000 micrograms of DHP through the subcutaneous route, however, had no protective effect against infection. Excess interferon and tumour necrosis factor production in intranasally DHP-treated mice was seen 1 day after the infection when compared with Sendai virus alone controls or with DHP alone controls. Variance analysis of these findings indicates a prophylactic activity of DHP in pulmonary viral infections.

Induction and immunohistochemical localization of cytochrome P-450 PCN by non-steroidal compound, in rat liver microsomes.

Induction of cytochrome P-450 PCN (P-450 PCN) by non-steroidal compound, M79193 (cyclohexane spiro-2,6-chloro-1'-(3-dimethyl-aminopropyl) spiro(chroman-4,4'-imidazolidine)-2',5'-dione], was investigated in both sexes of rats. The immunohistochemical localization of P-450 PCN was also studied in liver lobules of untreated and M79193-treated male rats. Immunoblot analysis of rat liver microsomes with anti-P-450 PCN indicated that the amount of P-450 PCN increased 5- to 7-fold by the treatment with M79193, but no increase was observed in P-450 PB-1 (P450IIB1) or P-450 MC-1 (P-450IA1). P-450 PCN was uniformly distributed in liver lobule of untreated rats, and was significantly increased by the treatment with M79193 in both the periportal and pericentral regions of the lobules.

Effect of aztreonam on immunohistochemical localizations of cytochrome P-450 (M-1) in male rats.

The localization of sex-specific cytochrome P-450, P-450 (M-1), was investigated immunohistochemically in the liver of untreated- and Aztreonam-treated male rats. P-450 (M-1) was uniformly distributed in the liver lobule of untreated rats. By treatment with Aztreonam, P-450 (M-1) was decreased more in the periportal region rather than the pericentral region of the lobule. In addition, oxidation of testosterone and 1 alpha-hydroxycholecalciferol catalyzed by P-450 (M-1) was also decreased by Aztreonam administration.

Induction of cytochrome P-450 isozymes by chromanamine derivatives in rat liver.

1. Pretreatment of rats with 6-(3-picolyl)amino-2,2,5,8-tetramethylchromane (PATC) for 7 days resulted in a significant increase in the activities of benzphetamine N-demethylase, p-nitroanisole O-demethylase and aniline hydroxylase in liver microsomes prepared 24 h after the last treatment. 2. Analysis by Western blot showed that PATC induces cytochrome P-450 b, P-450 c and P-450 d, which are the major forms of cytochrome P-450 in liver microsomes of rats when pretreated with phenobarbital and 3-methylcholanthrene. 3. Exposure of liver sections to the antibodies to cytochrome P-450 b and P-450 c resulted in intense immunostaining within the centrilobular regions, but produced staining of considerably weaker intensity in the perilobular region. Semiquantitative immunochemical analysis, by image analyser, of cytochrome P-450 b and P-450 c showed that centrilobular hepatocytes were stained more intensively than perilobular hepatocytes. 4. These results indicate that PATC induces cytochromes P-450 b and P-450 c, in the centrilobular hepatocytes to a greater degree than those in the perilobular hepatocytes. 5. Co-administration of PATC with pentobarbital caused a significant increase in pentobarbital sleeping time. Furthermore, PATC was found to cause a decrease in the activity of benzphetamine N-demethylase in liver microsomes prepared 30 min after treatment with the drug.

Immunochemical studies for the presence of P-450 HFLa, a form of cytochrome P-450 in human fetal livers in human hepatocellular carcinoma cells.

Immunohistochemical localization of P-450 HFLa, a form of cytochrome P-450 in human fetal livers was investigated in human hepatocellular carcinoma. The cytoplasm of carcinoma cells positively reacted with anti-P-450 HFLa antibodies. It was found that the carcinoma cells showing a pseudoglandular pattern or poorly differentiated appearance exhibited a weaker reactivity with anti-P-450 HFLa antibodies than did relatively differentiated carcinoma cells. In the case of hepatoblastoma, the polygonal or round-shaped tumor cells which differentiated into epithelial structure exhibited a positive reaction with anti-P-450 HFLa antibodies, whereas the spindle-shaped tumor cells which showed a sarcomatous appearance did not react with the antibodies.

Dog liver glutathione S-transferase and its strong immunoreactivity with rat transferase-P(7-7).

Dog liver cytosolic glutathione S-transferases (GSTs) were investigated to characterize their properties in comparison with rat liver transferases. Dog liver GSTs after the glutathione affinity column chromatography showed three subunit bands on SDS-polyacrylamide gel electrophoresis. These three subunits, designated as Yd1 (mol.wt 26,000), Yd2 (mol.wt 27,000) and Yd3 (mol.wt 28,000), were distinctly different from rat liver GST subunits, i.e. Ya(1) (mol.wt 26,500), Yb1(3)/Yb2(4) (mol.wt 27,500) and Yc(2) (mol.wt 28,500). Western blot analysis revealed that Yd1, Yd2 and Yd3 were immunoreacted with anti-rat GST 7-7, 1-1 and 3-3 antibodies, respectively. Four transferase activity fractions, I (pH greater than 7.63), II (pH 6.92), III (pH 5.80) and IV (pH 5.65), were obtained from affinity purified GSTs by chromatofocusing. Each fraction exhibited a characteristic substrate specificity. GST-II, III and IV were all strongly immunoreacted with anti-rat GST 7-7 antibody by immunoblotting, thus suggesting the occurrence of the heterogeneity of transferases immunologically related to rat GST subunit 7 in dog liver. Immunohistochemical examination showed that transferases immunoreacted with anti-GST 7-7 antibody have diffusely distributed throughout the lobule, while enzymes related to subunit 3 have been localized in a narrow range of cells around the central vein. These data suggest that GSTs immunologically associated with rat transferase subunit 7 may be major forms in dog liver.

The occurrence of acute lymphoblastic leukemia shortly after the cessation of human growth hormone therapy.

We encountered a case of a 21-year-old man who had developed acute lymphoblastic leukemia (ALL) shortly after the cessation of approximately six years' human growth hormone treatment for pituitary dwarfism. The patient had also been treated with thyroid preparations and methylandrostanolone, a kind of anabolic steroid. Factors relating to the onset of ALL in this case are discussed.

Immunohistochemical studies of DMBA-induced rat mammary tumors.

In order to clarify the biological characteristics of rat mammary tumors induced by 7,12-dimethylbenz-[a]-anthracene (DMBA), histochemical and immunohistochemical studies were performed. Two types of luminal spaces were observed within the tumor. In one type, the lumen was surrounded by eosinophilic columnar cells which were strongly reactive for soybean agglutinin (SBA) but weakly stained with keratin antibodies. In the luminal spaces, substances positive for PAS, dialyzed iron ferrocyanide or alcian blue and resistant to mucopolysaccharidase were occasionally observed. Ultrastructurally, the luminal surface was characterized by the presence of microvilli and tight junctions. In the other type, the lumen was often found in highly cellular foci and surrounded by pale, polygonal or elongated cells which were weakly stained with keratin antibodies but not SBA. The luminal spaces presented a peculiar structure filled mainly with mucoid substances sensitive to hyaluronidase, chondroitinase ABC and heparitinase, and the inner surface of the spaces was surrounded by basement membrane components: laminin, fibronectin and type IV collagen. The results of the present study therefore showed that DMBA-induced mammary tumor consists, partly, of a structure resembling human adenoid cystic carcinoma.