Expression of Albumin Messenger RNA Detected by in Situ Hybridization in Preneoplastic and Neoplastic Lesions in Rat Liver1

(CANCER RESEARCH 46, 5903-5912, November 1986] Expression of Albumin Messenger RNA Detected by in Situ Hybridization in Preneoplastic and Neoplastic Lesions in Rat Liver1

Michael Schwarz,2 Gabriele Peres, David G. Beer, MaimÃ³nMaor, Albrecht Buchmann, Werner Kunz, and Henry C. Pitot German Cancer Research Center, Institut of Biochemistry, Im Neuenheimer Feld 280, D-6900 Heidelberg, Federal Republic of Germany [M. S., G. P., M. M., A. B., W. K.J, and McArdle Laboratory for Cancer Research, The Medical School, University of Wisconsin, Madison, Wisconsin 53706 Â¡D.G. B., H. C. P.]

ABSTRACT synthesis and excretion are frequently reduced in chemically induced primary hepatoma and transplantable hepatoma cells The process of chemical hepatocarcinogenesis is characterized by the in vivo and in vitro (2, 4, 7-10). For reasons which are not fully appearance of preneoplastic lesions showing changes in the expression of various marker enzymes. We have analyzed the phenotype of small understood the expression of albumin was found to vary exten preneoplastic foci and expansively growing nodules in liver sections sively in hepatoma and primary hepatocellular carcinoma (7). obtained from rats treated with various carcinogens. Changes within the We were therefore interested to know whether such variability lesions in canalicular adenosine triphosphatase, â€¢y-glutamyltranspepti- in albumin expression could already be detected in early pre dase, NADPH-(cytochrome P-450) reducÃase,cytochrome P-450 PB2, neoplastic and neoplastic lesions in rat liver. These focally epoxide hydrolase, and glycogen content were detected by means of growing lesions, which are generally regarded as tumor precur enzyme histochemical and immunohistochemical staining procedures. In sors, are characterized by changes in the expression of marker parallel sections the expression of albumin messenger RNA was investi enzymes such as canalicular ATPase, GGT, glucose-6-phospha- gated by in situ hybridization using a "S-labeled albumin specific com tase, NADPH-(cytochrome P-450) reducÃase, EH, and micro- plementary DNA probe. In general, small preneoplastic lesions showed somal monooxygenases (11-20). Both the multiplicity of en unchanged levels of albumin messenger RNA. In contrast, the expression of albumin messenger RNA was found to be reduced to varying degrees zyme changes which occur in distinct patterns within individual in large hepatic nodules. An expression of a-fetoprotein messenger RNA foci and nodules and the inducibility of certain enzymes point could not be detected in any of the nodules. No direct correlation between towards alterations in processes directly linked to cell differ the enzyme phenotype of the lesions and the degree in reduction of entiation. albumin messenger RNA could be established except that the reduction Albumin and AFP have been localized in livers of carcinogen was most pronounced in nodules which had lost their ability to store treated rats by means of immunohistochemical staining tech glycogen. Since the synthesis and excretion of albumin is a typical niques (5, 21-25). In the present investigation we have studied function of the differentiated hepatocyte in the adult animal, the observed the expression of albumin mRNA by using a powerful in situ decrease in albumin messenger RNA expression in large hepatic nodules hybridization technique in frozen liver sections containing pre is in accordance with the hypothesis of a gradual differentiation or retrodifferentiation of the cell population during carcinogenesis. Hyper- neoplastic and neoplastic lesions induced by treatment of rats plastic nodules produced by continuous treatment of rats with 4-dimeth- with various carcinogens. ylaminoazobenzene showed increased rather than decreased albumin levels. The analysis of albumin messenger RNA expression might there fore be used as a tool to discriminate between nodules of differing MATERIALS AND METHODS biological nature and fate. Treatment of Animals. Female Wistar rats (Zentralinstitut fÃ¼rVer suchstierkunde, Hannover, Federal Republic of Germany) were used INTRODUCTION throughout the experiments. Treatment with carcinogens was started at an age of 4 weeks. Five rats were given DEN in the drinking water The synthesis and excretion of albumin is one of the major at a dose of 100 mg/liter for 10 days. Following a recovery period of functions of the differentiated hepatocytes in mammalian liver. 10 days they were put on a diet containing 0.05% phÃ©nobarbital.The In the adult animal it is produced at high and constant levels time periods of treatment are given in Table 1. One rat was given 4- (1,2). The gene copy number has been shown to be in the range DAB (0.06%) in the diet for 85 days and was killed 2 days later. of one to three/haploid rat genome (1, 3). Variation in the The effects of partial hepatectomy and phÃ©nobarbitaltreatment on expression of the gene is not due to permanent changes in gene albumin mRNA expression in liver were investigated in 12-week-old number or structure but modulated by the steady state concen rats. Partial hepatectomy was performed according to the method of Higgins and Anderson (26). PhÃ©nobarbital-sodium (80 mg/kg body trations of the specific mRNA molecules (1,2, 4). Albumin is under developmental control. It is not or only minutely ex weight in 0.9% NaCl solution) was administered by i.p. injection. The pressed in the fetus where large quantities of Â«-fetoprotein are treatment schedules are given in the legend to Fig. 9. Preparation of Tissue Sections. Livers of rats were removed under produced instead, whereas during normal adult life only trace ether anesthesia and immediately frozen at â€”40"C.Microscope slides amounts of AFP3 are found (5, 6). The levels of albumin were rinsed in acetone prior to use, air dried, and briefly immersed in a 1% solution of serum albumin. Slides were then dried at 70'C and Received 1/28/86; revised 7/17/86; accepted 7/24/86. stored until use. For in situ hybridization, 10-/im frozen sections were The costs of publication of this article were defrayed in part by the payment prepared, mounted on the albumin-coated slides, and air dried (10-20 of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. min). For enzyme and immunohistochemistry freshly prepared cryostat 1This work was partly supported by the Deutsche Forschungsgemeinschaft, sections were used. ATPase activity was stained by the method of Sonderforschungsbereich 302 and by grants from the National Cancer Institute (CA-07575 and CA-22484) (H. C. P.) and is part of the doctoral thesis of G. P. Wachstein and Meisel (27) and GGT activity was stained according to The paper is dedicated to Professor Dr. E. Hecker on the occasion of his 60th the method of Rutenberg et al. (28). Glycogen content was demon birthday. strated by periodic acid-Schiff reaction (29). Slides were counterstained 2To whom requests for reprints should be addressed. with toluidine blue. Immunohistochemical staining for albumin, cyto 3The abbreviations used are: AFP, a-fetoprotein; EH, epoxide hydrolase; 4- chrome P-450 PB2, and EH and staining of NADPH-(cytochrome P- DAB, 4-dimethylaminoazobenzene; DEN, diethylnitrosamine; GGT, -y-glutamyl transpeptidase; cDNA, complementary DNA; PBS, phosphate buffered saline; 450) reducÃasewere performed as previously described (16). Rabbit SSC, standard saline citrate (0.15 M sodium chloride: 0.015 M sodium citrate, anti-albumin antiserum was from Bioscience products, Emmenbrucke. pH 7.4). Switzerland. Antisera against the other enzymes were a kind gift of 5903 Downloaded from cancerres.aacrjournals.org on September 30, 2021. © 1986 American Association for Cancer Research. ALBUMIN mRNA DETECTED BY IN SITU HYBRIDIZATION

Table 1 Expression of albumin and AFP mRNA, glycogen, and marker enzymes in neoplaslic nodules of rat lirer after start of NADPH Animal12 treatmentGGT186 (days) ATPase storageEHÂ± reducÃase P-450PB2+Â±Â± AFPÂ±Â± (10days)DEN +190 -' +Â± +

(a)" (10days)DEN +333 (b)3 +Â±' Â± - Â±Â±

(a) (10days)DEN +'333 ++ Â±' (b)4 Â±Â±Â£-Not Â± Â±Â±

(a) (10days)DEN Â±Â±285 -' -' Â± (b)56Carcinogen"DEN +Â±+â€¢ + Â±Â±'

(10days)4-DAB +87 +5+ Â±+

(continuously)PhÃ©nobarbital*YesYesYesYesYesNoKilling +Glycogen + Not Â± determinedCytochromedeterminedAlbumin " Carcinogens were given in the drinking water (DEN) or diet (4-DAB) for the time periods indicated. b PhÃ©nobarbitalwas given continuously in the diet and was withdrawn 10-14 days before killing except for rats 1, 2. and 5 which were kept on phÃ©nobarbitaluntil death. c Data are expressed in relation to surrounding tissue as decreased (- or ), unchanged (Â±),or increased (+). " More than one large neoplastic nodule is designated by (a) and (b). ' Nonhomogeneous staining. Drs. R. Wolf and F. Oesch, University of Mainz, Federal Republic of Germany. Hybridization Probes. Cloned rat albumin cDNA pRS-1 (30), rat a- fetoprotein cDNA pRAFP 87 (31), and human actin pseudogene cDNA phi 1 (32) were nick translated according to the method of Rigby et al. (33) using 32P-labeled (for RNA dot blots) or 35S-labeled (for in situ hybridization) deoxyribonucleotides (Amersham Buchler, Braun schweig, Federal Republic of Germany). The reaction was run for 1 (32P) to 3 (35S) h using the Bethesda Research Laboratories nick translation kit, 50 ^Ci [32P]dCTP or [35S]dCTP (Amersham Buchler) and 0.2 MgDNA. The resulting specific activities ranged from 2-5 x lO'cpm/Mgof DNA. In Situ Hybridization. Preparation of liver sections for hybridization was essentially according to the method of Hafen et al. (34). Air-dried slides were generally fixed at 4Â°C(15 min) in 4% freshly prepared paraformaldehyde (in PBS). Slides were then washed in PBS (3x5 min), 0.2 N HO (20 min), distilled water (5 min), 2 x SSC (30 min), distilled water (5 min), and 50 mivi Tris-HCI, pH 7.5, containing predigested pronase, 0.25 mg/ml (10 min), PBS containing glycine, 2 mg/ml (30 s), 2 x PBS (30 s), 4% paraformaldehyde in PBS (20 min), and PBS (3x5 min), acetylated by 0.25% acetic anhydride in 100 mivi methanolamine, pH 8 for 10 min (35); then washed in PBS (3x5 min), 2 x SSC (5 min), and dehydrated in 30, 60, 75, and 95% ethanol (2 min each). All washing steps were performed at room temperature except for the 30 min 2 x SSC treatment which was performed at 70Â°C (36). Slides were kept for up to 1 month at 4Â°Cbefore use for hybridi zation procedures. The hybridization procedure of Breborowicz and Tamaoki (35) was followed with slight modification. Sections were prehybridized with a mixture containing sonicated salmon sperm DNA (200 Mg/ml), polyadenylate (100 Mg/ml), 50% formamide, 10% dextran sulfate, and 2 x SSC at 37Â°Cfor 60 min. The prehybridization buffer was removed by immersing the slides into 2 x SSC. To 1 ml of the hybridization mixture 1-4.5 ug DNA (4.5-6.5 x 10' cpm) in 20 mivi dithiothreitol were added. DNA was made single stranded by heating the hybridization mixture for 10 min to 95Â°Cand 20 n\ were applied to each section. Slides were covered with siliconized coverslips (20 x 20 mm) and sealed with rubber cement/petrol ether (1/1). Hybridiza tion was performed at 37Â°Cfor 24 h in a humidified chamber. After hybridization slides were horizontally immersed in a solution of 2 x SSC containing 50% formamide and coverslips were carefully removed with surgical foreceps. Slides were then washed 5 times at room temperature in the same solution (25 ml each) and in 2 x SSC for 24 h with constant stirring. The buffer was changed twice during this period. The slides were then dehydrated in 75, 95, and 99% ethanol containing 0.3 M ammonium acetate for 5 min each and air dried. In general, slides were then arranged in an X-ray chamber and covered Fig. 1. Demonstration by in situ hybridization of albumin mRNA in frozen with Kodak XAR 5 X-ray film, which was exposed for 1-5 days at liver sections of a DEN/phenobarbital treated rat. The two photomicrographs room temperature (37). After development of the X-ray films slides show representative areas of large hepatic nodules (A) and surrounding normal were coated at 42Â°Cwith Kodak NTB2 emulsion diluted 1:1 with 0.6 liver tissues (5). Sections were hybridized with 3SS-labeled rat albumin pRS-1 cDNA at 37'C for 24 h; autoradiographic exposure of photoemulsion covered M ammonium acetate, pH 7.0 and exposed for 1-3 weeks at 4Â°Cin a slides was for 14 days. Note the strongly reduced signal intensity over hepatocytes dry chamber. After development in Kodak D-19 developer slides were within the nodule. Counterstained with H & E. x 740. 5904 Downloaded from cancerres.aacrjournals.org on September 30, 2021. © 1986 American Association for Cancer Research. ALBUMIN mRNA DETECTED BY IN SITU HYBRIDIZATION counterstained with hematoxylin/eosin. treated sections with pancreatic RNase (100 Mg/ml 2 x SSC, Northern Analysis of RNA and Dot Hybridization. Total cellular 37Â°C,for 30 min) prior to hybridization and observed a consid RNA was isolated from liver according to the method of Chirgwin et erably reduced signal (data not shown). In addition, since we al. (38). RNA was then separated in 1.2% agarose-formaldehyde gels nick translated the l.S-kilobase albumin cDNA while still according to the method of Maniatis et al. (39) and transferred to Gene linked to pBR322 the specificity of the signal was tested by Screen Plus filters (New England Nuclear) for Northern analysis. For nick translating pBR322 without albumin insert to the same dot blot analysis descending concentrations of RNA were blotted onto specific activity. No signal could be detected with this pBR322 a New England Nuclear Gene Screen Plus membrane. Filters were hybridized to nick translated 32P-labeled albumin cDNA in a buffer probe alone (data not shown). In order to exclude the possibility containing I M NaCl, 50% formamide, 1% sodium dodecyl sulfate, and that changes in film blackening observed were due to density 10% dextran sulfate for 12-24 h. Filters were washed under constant differences in normal and preneoplastic tissue we used a 35S- agitation in 2 x SSC (2x5 min at room temperature), 2 x SSC labeled actin cDNA probe as a control on parallel sections. The containing 1% sodium dodecyl sulfate (2 x 30 min at 6(1 (), and 0.1 x overall signal with this probe was weaker than that with the SSC (2 x 30 min at room temperature), and exposed to Kodak XAR 5 albumin probe and was uniformly distributed over normal and X-ray film with an intensifying screen at -70Â°Cfor various times. preneoplastic tissue (see Fig. 4). In order to obtain additional information on the localization of the bound cDNA probe the extent of silver grain formation was analyzed in photoemulsion RESULTS covered sections using light microscopy at high field magnifi Conditions of in Situ Hybridization and Specificity of Signal cation. Silver grains were localized predominantly over hepa- Detection. We found that the detection of albumin specific tocytes and the number of grains was strongly reduced over mRNA with a 35S-labeled cDNA probe was optimal under our neoplastic cells (see Fig. 1). Transections of larger blood vessels experimental conditions after an exposure of X-ray film for were essentially free of grains indicating a good signal to noise only 1-3 days by contact autoradiography without the use of ratio. In short, we observed a correspondence of macroscopi- intensifier screen. To test for nonspecific binding we first cally detected film blackening and microscopic grain count.

ATPa se GGT .-

* ReducÃase Albumin Fig. 2. Demonstration by in situ hybridization of a decrease in albumin mRNA in preneoplastic foci induced in rat liver by diethylnitrosamine. The expression of the marker enzymes ATPase, GGT, and NADPH-(cytochrome P-450) reducÃaseis demonstrated by enzyme histochemical staining in parallel sections. The conditions of in situ hybridization were as in the legend to Fig. 1. The signal from "S-labeled cDNA retained in the tissue section was monitored by directly overlaying the tissue section with X-ray film; autoradiographic exposure was for 2 days. Arrows, reduced albumin mRNA levels in two foci with transsection diameters of 1.2 and 0.45 mm. x 14.6. 5905 Downloaded from cancerres.aacrjournals.org on September 30, 2021. © 1986 American Association for Cancer Research. ALBUMIN mRNA DETECTED BY IN SITU HYBRIDIZATION The resolution of the X-ray film detection system used is very characteristic pattern) within the liver lobule to provide quite high, estimated to be in the range of only a few cell layers. better orientation (e.g., Fig. 3). Fig. 2 shows focal transections of small size exhibiting a reduced Expression of Marker Enzymes in Small Preneoplastic and hybridization signal adjacent to foci of similar diameter which Expansively Growing Nodules. Preneoplastic nodules were in show an unchanged hybridization signal. duced in rat liver by treating the animals with the carcinogens Albumin mRNA Expression within the Liver Lobule. The DEN or 4-DAB alone or by sequential treatment with DEN method of signal detection used in the present study facilitates and phÃ©nobarbital according to the two stage initiation-pro the analysis of the spatial distribution of albumin mRNA motion protocol. In the present investigation, liver lobes from expression within the liver lobule. Exposed X-ray film clearly seven rats were used which were kept frozen from previous showed a nonhomogeneous pattern indicating that albumin experiments. The available material was selected for hepatic mRNA is not uniformly distributed within the liver lobule (Fig. nodules with representative differences in their marker enzyme 3). Quite unexpectedly the expression levels were found to be expression rather than according to the protocol under which higher in the periportal than in the central areas of the liver these nodules were induced. acinus. This result confirms recent data on the immunohisto- As in our earlier investigations and for reasons previously chemical localization of albumin in the perfused mouse liver discussed (16) we took as markers of preneoplasia changes in (40) but is in contrast to data indicating that albumin is pro the activities of canalicular ATPase, GGT, and NADPH-(cy- duced by all hepatocytes within the liver lobule at similarly high tochrome P-450) reducÃaseand also the content of glycogen, levels (21, 22). PhÃ©nobarbital treatment did not change the cytochrome P-450 PB2, and EH. Data on the latter two markers distribution of albumin messenger. Serial sections were stained are given in Table 1. for additional markers which are differentially expressed (in a Small, presumably preneoplastic foci were generally charac-

ATPase ReducÃase Albumin . â€¢'-V B

ATPase; ReducÃase Albumin

Fig. 3. Demonstration by in situ hybridization of zonal differences in albumin mRNA expression within the liver lobule. The conditions of hybridization and signal detection were as in the legend to Figs. I and 2. The marker enzymes ATPase and NADPH-(cytochrome P-450) reducÃasestained in serial sections are also shown for orientation, c, centrilobular: p. periportal area. A, rat 2 of Table 1 treated with DEN followed by phÃ©nobarbital;B, control rat. x 14.6. 5906

AT Pa se

GGT

Glycogen

ReducÃase

Albumin

Act in

Fig. 4. Demonstration by in situ hybridization of albumin and actin mRNA expression in frozen liver sections obtained fron DEN/phenobarbital treated rats. Parallel serial sections are stained enzyme histochemically for the marker enzymes ATPase, GGT, glycogen content, and NADPH-(cytochrome P-450) reducÃase. Sections were hybridized with 35S-labeled albumin or actin cDNA; signal detection was by contact autoradiography with X-ray film; autoradiographic exposure was for 2 (albumin) or 5 (actin) days. A summary of the data on marker enzyme and albumin mRNA expression in the expansively growing nodules in the liver sections is given in table I along with the treatment protocol of their induction, a, b, more than one nodule of interest per section; InserÃ¬inliter section of animal 2 is magnified in Fig. 2. x 1.7.

5907

Downloaded from cancerres.aacrjournals.org on September 30, 2021. © 1986 American Association for Cancer Research. ALBUMIN mRNA DETECTED BY IN SITU HYBRIDIZATION terized by a decrease in ATPase activity and a concomitant from some of the large nodules showed decreased albumin increase in GGT activity. The majority of these foci also showed mRNA expression levels (Fig. 6). In agreement with marker an extensive glycogen storage (see Fig. 4) and increased EH enzyme expression, //; situ hybridization with an albumin cDNA levels, whereas NADPH-reductase activity and the amounts of demonstrated the presence of liver cell subpopulations within the cytochrome P-450 isoenzyme PB2 were found to vary large nodules as indicated by the differential hybridization considerably from decreased to unchanged or elevated levels. patterns (see Fig. 7). In general, those areas within large nodules In contrast to these smaller preneoplastic lesions with their characterized by changes in albumin mRNA expression also relatively uniform marker enzyme expression pattern, large showed alterations coincident with one or more of the marker hepatic nodules showed greatly varying combinations of marker enzymes. enzyme expression (Fig. 4). These data are summarized in Table No conclusive results were obtained in attempts to correlate 1. Nodules induced by continuous treatment of rats with 4- the degree of albumin mRNA reduction of individual foci with DAB show a marker enzyme pattern quite similar to those their marker enzyme expression. A possible correlation was observed with DEN/phenobarbital treatment. Staining inho- only suggested between the glycogen content and albumin mogeneity within larger nodules is an indication of the existence mRNA: nodules with unchanged or decreased glycogen levels of nodular subpopulations. exhibited the strongest reduction of albumin mRNA (Table 1). Expression of Albumin and a-Fetoprotein mRNA in Small and This is also exemplified in Fig. 7, where a subpopulation of Large Hepatic Nodules. All large nodules characterized by liver cells within a nodule showing increased glycogen storage marker enzyme changes showed decreased expression levels of also shows increased expression of albumin mRNA. albumin mRNA as compared to the surrounding tissue. Indi Most of the smaller foci observed showed unchanged or only vidual nodules in liver sections from one and the same animal slightly reduced albumin mRNA levels. Sections obtained from as well as from different animals showed variable degrees of animals treated with 4-DAB showed very characteristic hybrid reduction (see Fig. 4). A reduction in albumin level was also ization patterns. Albumin mRNA was not or only slightly demonstrated by immunolocalization of the protein (Fig. 5). In expressed in distinct zones around the hyperplastic nodules addition, Northern analysis of RNA which could be isolated induced in these livers. These zones of decreased albumin

Fig. 5. Immunohistochemical demonstration of reduced albumin levels in hepatic nodules. A, liver sections incubated with an albumin specific antibody and stained for immunoperoxidase activity using 3,3'-diaminobenzidine as a substrate, n, representative areas of hepatic nodules; s, surrounding liver tissue. B, albumin mRNA levels detected by in situ hybridization in parallel sections. Left series, magnification of a nodule of rat 1 shown in Fig. 4; right series, a liver section of rat 4. x 20. 5908

~ m x2.3Kb

Fig. 6. Hybridization of 3:P-labeled albumin-specific cDNA probe to cellular RNA isolated from hepatic nodules and control tissue. Ten *ig of total cellular Glycogen RNA were separated on a 1.2% agarose-formaldehyde gel. Lane I, RNA obtained from a large liver nodule of a rat treated with DEN (10 days; 100 ppm in the drinking water) followed by phÃ©nobarbital(0.1% in the diet). The animal was killed 12 months after start of treatment; Lanes 2-4, RNA isolated from the large hepatic nodules of rats I, 3, and 4 shown in Fig. 4; Lane C, RNA isolated from control liver. Heat-denatured XDNA restricted with ////Â»/IIIwas run on the same gel as size marker and hybridized to nick translated X DNA (Lane M). RNA transfer to filters, hybridizations, and washing procedures were as described in "Materials and Methods." Kh, kilobase. mRNA levels showed an increased hybridization reaction with Reductase the actin specific probe (see Fig. 8). The nodules themselves showed increased rather than decreased albumin mRNA expres sion. The effect of a high cell proliferation rate and of phÃ©nobar bital administration on albumin mRNA expression was also determined using dot blot hybridization analysis of RNA ob tained from regenerating liver at the time of maximal DNA synthesis, 24 h after a two-thirds partial hepatectomy, and of liver from rats given phÃ©nobarbital alone or in combination with a partial hepatectomy. Albumin mRNA levels in regener Fig. 7. Expression of albumin mRNA and marker enzymes in subpopulations of liver cells in hepatic nodules. Albumin mRNA levels are detected by in situ ating liver were slightly reduced but they were not affected by hybridization. Changes in marker enzyme expression are demonstrated by enzyme phÃ©nobarbitaladministration (see Fig. 9). histochemical staining in parallel serial sections. Left series, magnification of a No increase in in situ hybridization signal over background nodule of rat 3 shown in Fig. 4; right series is from rat 5. For further details see legend to Fig. 4 and Table I. Arrows, subpopulations of liver cells showing levels was seen in any of the nodules investigated using an AFP alterations in their albumin mRNA and marker enzyme expression, x 4.2. specific cDNA probe (Table 1). prepared and combined with an avidin- or streptavidin-peroxi- DISCUSSION dase detection system. Although this technique has proven very Conditions of Hybridization and Signal Detection. The expres powerful in a variety of experimental systems (41, 42), we were sion of albumin and its messenger RNA have been studied in not able to obtain any significant results when using liver slices isolated liver cells and liver tissue sections by means of immu- even after blocking any endogenous biotin in the liver section nohistochemical and in situ hybridization techniques respec prior to hybridization with a biotinylated probe. tively (21-25, 35). In the first part of the present investigation 35S-labeled probes both increased the sensitivity of the in situ we focused our interest on the improvement of the conditions hybridization method and made possible the use of X-ray film of hybridization and signal detection. While tritium-labeled contact autoradiography for macroscopic detection of lesions probes allow a good resolution and localization at the cellular due to the higher penetration distance of/3-particles emitted by and subcellular level, the requirements of comparative serial this radionuclide while at the same time displaying a sufficiently sections of relatively large foci and nodules necessitated the use good resolution for localization at the cellular level. of a method which would allow a macroscopic analysis of these Fixation of liver sections with paraformaldehyde has been lesions. In principle, biotinylated cDNA probes would have used in the present study. In agreement with other groups (34, been very suitable for this purpose. These probes can be easily 43) we find cross-linking fixatives to be better suited than 5909 Downloaded from cancerres.aacrjournals.org on September 30, 2021. © 1986 American Association for Cancer Research. ALBUMIN mRNA DETECTED BY IN SITU HYBRIDIZATION â€¢\,

ATPase GGT Glycogen

ReducÃase Albumin Actin Fig. 8. Demonstration by in situ hybridization of albumin and actin mRNA in frozen liver sections obtained from a 4-DAB treated rat (rat 6 in Table I). Parallel serial sections are stained enzyme histochemically for marker enzymes. For further details see legend to Fig. 4 and Table 1. x 3.6.

The number of albumin mRNA molecules per liver cell has been estimated to be in the range of about 40,000 (2). An increase in sensitivity of the in situ hybridization technique used >ug RNA is therefore necessary for the analysis of mRNA species with lower reiteration frequencies. This can be achieved by increasing the specific activities of the hybridization probes, especially by 10 â€¢â€¢â€¢â€¢using "S-labeled probes cloned into RNA expression vectors together with prolonged periods of X-ray film exposure times. 5 â€¢â€¢â€¢â€¢Changes in Albumin mRNA Expression in the Course of Hepatocarcinogenesis. The levels of albumin mRNA were found 2.5 to be unchanged or only very slightly reduced in small preneoplastic foci. It is recognized, however, that a minor population of the small foci observed could represent transections at the 1.25 â€¢â€¢â€¢ periphery of larger nodules. In accordance with previous results (20) we again find that 0.63 * â€¢ GGT is a predominant marker of hepatocarcinogenesis; almost all small foci identified by changes in other markers also showed Fig. 9. Effect of partial hepatectomy and phÃ©nobarbitaltreatment on the expression of albumin mRNA in rat liver. RNA was isolated from livers of rats increases in GGT activity. More recent data from Northern which were treated as follows: Lane 1, control; Lane 2, rat given i.p. injection of analyses of RNA from GGT positive and negative cells obtained 80 mg phenobarbital/kg body weight on 3 consecutive days and killed on day 4; Lane 3, animal subjected to two-thirds partial hepatectomy and killed 24 h by a separation method based on antibody binding of cell surface thereafter: Lane 4. same as Lane 3 except that the animal was pretrated with GGT from livers of rats treated sequentially with DEN and phÃ©nobarbital.Decreasing concentrations of total RNA were blotted onto New phÃ©nobarbital indicate that albumin mRNA levels in GGT England Nuclear Gene Screen and hybridized with 3!P-labeled albumin cDNA. positive cells are slightly increased compared to messenger levels in cells not expressing this enzyme (44). noncross-linking fixatives (data not shown). Perfusion of the In expansively growing nodules we find a reduction of albu rat liver with paraformaldehyde solution prior to organ removal min mRNA levels with some variation in the degree of the did not lead to a significant increase in hybridization signal in reduction. This is in accordance with results on the levels of spite of a better preservation of liver structure. Since the his- synthesis and secretion of albumin in transplantable hepatoma tochemical demonstration of marker enzyme activities was in vitro and in vivo where the gene was sometimes found to be made impossible by paraformadehyde perfusion this step was down regulated at the transcriptional level to almost zero levels thereafter omitted. (2-4, 6, 7, 9, 10). The observed lack of AFP expression could The liver blocks cut for the present investigation had been mean that the nodules investigated have not yet reached the kept frozen from earlier experiments at -70Â°C for over 1 year. state of malignancy where this protein might reappear. In This demonstrates the stability of at least the albumin mRNA principle, the technique of in situ hybridization used could be within these liver blocks and will potentially allow for the too insensitive to detect changes in AFP mRNA expression; rÃ©Ã©valuationofhistolÃ³gica! materials kept in storage for an however, we were also unable to detect AFP by immunohisto- extended period of time by in situ hybridization with other chemical staining of parallel liver sections in any of the nodules probes. investigated (data not shown). 5910 Downloaded from cancerres.aacrjournals.org on September 30, 2021. © 1986 American Association for Cancer Research. ALBUMIN mRNA DETECTED BY IN SITU HYBRIDIZATION The extent to which albumin mRNA was decreased was most 6- Sala-Trepat, J. M., Dever, J., Sargent, T. D., Thomas, K., Sell, S., and Bonner, J. Changes in expression of albumin and a-fetoprotein genes during pronounced in nodules no longer showing increased but rather rat liver development and neoplasia. Am. Chem. Soc., 18: 2167-2178, 1979. unchanged or decreased glycogen levels. The loss of the ability 7. Becker, F. F., Klein, K. M., Wolman, S. R., Asofsky, R., and Sell, S. Characterization of primary hepatocellular carcinomas and initial transplant to store glycogen has been demonstrated to be a useful discrim generations. Cancer Res., 33: 3330-3338, 1973. inator between preneoplastic and more advanced neoplastic g McLaughlin, C. A., and Pitot. H. C. The effect of various treatments in vitro nodules (15). Both glycogen storage and albumin synthesis are and in vivo on the binding of '"I-labeled anti-rat serum albumin to rat tissue polyribosomes. Biochemistry, 75:3541-3550, 1976. characteristic functions of the differentiated hepatocyte in the 9 Rotermund, H. M., Schreiber, G., Maeno, H., Weinssen, and Weigand, K. adult animal. The observed gradual decrease in glycogen storage The ratio of albumin synthesis to total protein synthesis in normal rat liver, and albumin mRNA expression during development of malig in host liver, and in Morris Hepatoma 9121. Cancer Res., 30: 2139-2146, nancy could therefore result from an alteration in the difieren- )0 1970. Schreiber, G., Rotermund, H. M., Maeno, H., Weigand, K., and Lesch, R. tiation state of the neoplastic cell population due to dediffer- The proportion of the incorporation of leucin into albumin to that into total entiation or retrodifferentiation or from a blocked development protein in rat liver and hepatoma Morris 5123tc. Eur. J. Biochem., 10: 355- 361, 1969. of stem cells (45, 46). This concept has been controversially Â¡Â¡Friedrich-Freksa, H., GÃ¶ssner, W., and Borner, P. Histochemische Unter discussed more recently following the observation that albumin suchungen der Cancerogenese in der Rattenleber nach Dauergaben von mRNA levels are increased in nodules produced by ethionine ]2 DiÃ¤thylnitrosamin. Z. Krebsforsch., 72: 226-239, 1969. Emmelot. P., and Scherer, E. The first relevant cell stage in rat liver in a choline deficient diet (47). Such increase was also observed carcinogenesis. A quantitative approach. Biochim. Biophys. Acta, 60S: 247- in the present study in nodules induced by continuous 4-DAB 304. 1980. 13. Kunz, W., Tennekes, H. A., Port. R. E., Schwarz, M., Lorke, D., and treatment. These lesions exhibit extremely high growth prop Schaude. G. Quantitative aspects of chemical carcinogenesis and tumor erties and are probably similar to the hyperplastic nodules promotion in liver. Environ. Health Perspect., SO: 113-122, 1983a. induced by the so-called Solt-Farber protocol (17, 48). Most 14-Schwarz, M., Pearson, D., Port, R., and Kunz, W. Promoting effect of 4- dimethylaminoazobenzene on enzyme altered foci induced in rat liver by .V marker enzyme changes in these hyperplastic nodules are iden nitrosodiethanolamine. Carcinogenesis (I uml.). 5: 725-730, 1984. tical to the ones observed in foci produced by DEN treatment 15. Bannasch, P. The cytoplasm of hepatocytes during carcinogenesis. Recent (see Fig. 8). The fact that most of the hyperplastic nodules Results Cancer Res., 19: 1-100, 1968. Buchmann, A., Kuhlmann, W., Schwarz, M., Kunz, W., Wolf, C. R., Moll, disappear within very short time periods after carcinogen with E., Friedberg, T., and Oesch, F. Regulation and expression of four cyto- drawal (17, 48, 49) suggests a difference in the nature of both chrome P-450 isoenzymes, NADPH, cytochrome P-450 reducÃase,the glu- tathione transferases B and C and microsomal epoxide hydrolase in preneo cell populations. Hyperplastic nodules may therefore represent plastic and neoplastic lesions in rat liver. Carcinogenesis (I uml.). 6: 513- focal lesions of reversible hyperfunction and may thus not be 521, 1985. suitable early markers of malignancy. It is interesting that the 17- Farber, E., and Cameron, R. The sequential analysis of cancer development. Adv. Cancer Res., 31: 125-226, 1980. expression of albumin mRNA can be used to discriminate Â¡gKunz, W.. Appel, K. E., Rickart, R., Schwarz, M., and StÃ¶ckle.G. Enhance between the different types of nodules. ment and inhibition of carcinogenic effectiveness of nitrosamines. In: H. Remmer, H. M. Bolt, P. Bannasch, and H. Popper (eds.). Primary Liver The reduction of albumin and increase in actin mRNAs Tumours, pp. 261-283. Lancaster, England: MTP Press Ltd., 1978. observed in distinct zones around nodules induced by 4-DAB 19. Pitot, H. C., Barsness, I... Goldsworthy, T., and Kitagawa, T. Biochemical could be explained by the fact that these areas consist of characterisation of stages of hepatocarcinogenesis after a single dose of diethylnitrosamine. Nature (Lond.), 271:456-458, 1978. connective tissue together with proliferating oval cells. The 2o Pitot, H. C., and Sirica, A. E. The stages of initiation and promotion in observed decrease in albumin mRNA expression in large he hepatocarcinogenesis. Biochim. Biophys. Acta, 605: 191-215, 1980. patic nodules could be partly explained by an increase in the 2I- Bernuau, D., Poliard, A., Tournier, I., Sala-Trepat, J., and Feldmann, G. All hepatocytes are involved in the expression of the albumin gene in the normal proliferation rate of the neoplastic cell population. Because the adult rat: a demonstration by in situ hybridization and immunoperoxidase hyperplastic nodules which were produced by continuous 4- techniques. Cell Biol. Rep., 9: 31-42, 1985. DAB treatment and which are characterized by extremely high 22' Yokota, S., and Fahimi, H. D. Immunocytochemical localization of albumin in the secretory apparatus of rat liver parenchyma! cells. Proc. Nati. Acad. proliferation rates (14) showed an increased rather than de Sci. USA, 78:4970-4974, 1981. creased albumin mRNA expression, this effect may contradict 23 Kuhlmann, W. D. Localisation of alphai-fetoprotein and DNA synthesis in liver cell populations during experimental hepatocarcinogenesis in rats. Int. the hypothesis of a link between cell proliferation and albumin J. Cancer, 21: 368-380, 1987. mRNA levels. 24. Sell, S. Distribution of a-fetoprotein- and albumin-containing cells in the livers of Fisher rats fed four cycles of A'-2-fluorenylacetamide. Cancer Res., JÂ«:3107-3113, 1978. 25. Feldmann, G., Maurice, M., Bernuau, D., Rogier, E. Morphological aspects ACKNOWLEDGMENTS of plasma protein synthesis and secretion by the hepatic cells. Int. Rev. Cytol., 96: 157-189, 1985. The authors acknowledge Rolf Schmitt for preparing and histochem- 26. Higgins, G. M., and Anderson, R. M. Experimental pathology of the liver: ically staining liver sections. We thank Karin Helm for typing the I. Restoration of the liver of the white rat following partial surgical removal. manuscript. Drs. Roland Wolf and Franz Oesch are acknowledged for Arch. Pathol., 12: 186-202, 1931. 27. Wachstein, M., and Meisel, E. Histochemistry of hepatic phosphatases at a making the antibodies used available to us. We also thank Dr. Leonard physiological pH. Am. J. Clin. Pathol., 27: 13-23, 1957. Grosso for helpful suggestions during the initial phases of this work. 28 Rutenberg, A. M., Kim, H., Fischbein, J. W., Hawkes, J. S., Wasserkrug, H. I .. and Seligman, A. M. Histochemical and ultrastructural demonstration of l-glutamyl transpeptidase activity. J. Histochem. Cytochem., 17: 517-526. REFERENCES 1969. 29. Hotchkiss, R. D. A microchemical reaction resulting in staining of polysac- 1. Sala-Trepat, J. M.. Sargent. T. D., Sell, S., and Bonner, J. a-Fetoprolein and charide structures in fixed tissue preparations. Arch. Biochem., 16:131-141, albumin genes of rats: no evidence for amplification-deletion or rearrange 1948. ment in rat liver carcinogenesis. Proc. Nati. Acad. Sci. USA, 76: 695-699, 30. Simmons, D. I... Lalley, P. A., and Kasper, C. B. Chromosomal assignments 1979. of genes coding for components of the mixed-function oxidase system in 2. Sell, S., Thomas, K., Michalson, M., Sala-Trepat, J., and Bonner, J. Control mice. J. Biol. Chem., 260: 515-521, 1985. of albumin and a-fetoprotein expression in rat liver and in some transplant- 31. Gal, A., Nahon, J.-L., Gomez-Garcia, M.. Trainer, I., and Sala-Trepat, J. able hepatocellular carcinomas. Biochim. Biophys. Acta, 564:173-178,1979. M. Organization of the albumin and a-fetoprotein genes in fetal and adult 3. Tse, T. P. H., Morris, H. P., and Taylor. J. M. Molecular basis of reduced rat tissues, and rat hepatomas. Differentiation, 29: 238-242, 1985. albumin synthesis in Morris Hepatoma 7777. Am. Chem. Soc., 17: 3121- 32. Moos, M., and Gall witz, D. Structure of two human ff-aclin-related processed 3128, 1978. genes one of which is located next to a simple repetitive sequence. EMBO 4. Cassio, D., Weiss. M. C, and Ott, M. O. Expression of the albumin gene in J., 2:757-761, 1983. rat hepatoma cells and their differentiated variants. Cell, 27: 351-358,1958. 33 Rigby, P. W. J., Dieckmann, M., Rhodes, C., and Berg, P. Labeling deoxy- 5. Abelev, G. I. a-Fetoprotein in ontogenesis and its association with malignant ribonucleic acid to high specific activity in vitro by nick translation with DNA tumors. Adv. Cancer Res.. 14: 295-355, 1971. polymerase I. J. Mol. Biol.. 113: 237-251, 1977. 5911 Downloaded from cancerres.aacrjournals.org on September 30, 2021. © 1986 American Association for Cancer Research. ALBUMIN mRNA DETECTED BY IN SITU HYBRIDIZATION

34. Hafen, E., Levine, M., Garber, R. L., and Gehring, W. J. An improved in cells and paraffin-embedded tissue sections using biotin-labeled hybridization situ hybridization method for the detection of cellular RNAs in Drosophila probes. Virology, 126: 32-50, 1983. tissue sections and its application for localizing transcripts of the homeotic 42. Singer, R. II., and Ward, D. C. Actin gene expression visualized in chicken antennapedia gene complex. EMBO J., 2:617-623, 1983. muscle tissue culture by using in situ hybridization with a biotinated nucleo- 35. Breborowicz, J., and Tamaoki, T. Detection of messenger RNAs of a- tide analog. Proc. Nati. Acad. Sci. USA, 79: 7331-7335, 1982. fetoprotein and albumin in a human hepatoma cell line by in situ hybridiza 43. Angerer, L. M., and Angerer, R. C. Detection of poly A* RNA in sea urchin tion. Cancer Res., 45: 1730-1736, 1985. eggs and embryos by quantitative in situ hybridization. NucÃ.Acids Res., 9: 36. Brahic, M., and Haase, A. T. Detection of viral sequences of low reiteration 2819-2840,1981. frequency by in situ hybridization. Proc. Nati. Acad. Sci. USA, 75: 6125- 44. Beer, D. G., M. Schwarz, N. Sawada, and H. C. Pilot. Expression of H-ras 6129, 1978. and c-myc protooncogenes in isolated y-glutamyl transpeptidase-positive rat 37. Haase, A. T., Gantz, D., Blum, H., Stowring, L., Ventura, P., Geballe, A., hepatocytes and in hepatocellular carcinomas induced by diethylnitrosamine. Moyer, I!,, and Brahic, M. Combined macroscopic and microscopic detection Cancer Res., 46: 2435-2441, 1986. of viral genes in tissues. Virology, 140: 201-206, 1985. 45. Potter, V. R. Phenotypic diversity in experimental hepatomas: the concept 38. Chirgwin, J. M., Przybyla, A. E., MacDonald, R. J., and Rutter, W. J. of partially blocked ontogeny. Br. J. Cancer, 38: 1-23, 1978. Isolation of biologically-active ribonucleic acid from sources enriched in 46. Uriel, J. Cancer, retrodifferentiation, and the myth of Faust. Cancer Res., ribonuclease. Biochemistry, IS: 5294-5299, 1979. 36:4269-4275, 1976. 39. Maniatis, T., Fritsch, E. F., and Sambrook, J. Molecular Cloning, A Labo 47. Petropoulous, C., Andrews, G., Tamaoki. T., and Fausto, N. a-Fetoprotein ratory Manual, pp. 202-206. Cold Spring Harbor, NY: Cold Spring Harbor and albumin mRNA levels in liver regeneration and carcinogenesis. J. Biol. Laboratory, 1982. Chem., 25Â«:4901-4906, 1983. 40. Tuczek, H. U., Grau, F. A., Mischlinsky, A., Wagner, T., and Wegner, G. 48. Soil, D., and Farber, E. New principle for analysis of chemical carcinogenesis. Distribution of albumin in normal and regenerating livers of mice. Histo- Nature (Lond.), 263: 701-703, 1976. chemistry, S3:165-169, 1985. 49. Moore, M. A., Hacker, H. J., and Bannasch, P. Phenotypic instability in 41. Brigati, D. J., Myerson, D., Leary, J. J., Spalimi/. B., Travis, S. Z., Fong, C. focal and nodular lesions induced in a short term system in the rat liver. K. Y., Hsiung, G. D., and Ward, D. C. Detection of viral genomes in cultured Carcinogenesis (Lond.), 4: 595-603, 1983.

5912

Downloaded from cancerres.aacrjournals.org on September 30, 2021. © 1986 American Association for Cancer Research. Expression of Albumin Messenger RNA Detected by in Situ Hybridization in Preneoplastic and Neoplastic Lesions in Rat Liver

Michael Schwarz, Gabriele Peres, David G. Beer, et al.

Cancer Res 1986;46:5903-5912.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/46/11/5903

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/46/11/5903. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.