Biochimica et Biophysica A cta, 654 (1981) 31-41
31
Elsevier/North-HollandBiomedicalPress BBA 99861 LIVER AND KIDNEY NUCLEAR RNA SYNTHESIS AND MODIFICATIONS IN DIMETHYLNITROSAMINE-TREATED RATS ILGAWINICOV * Fels Research Institute and Department o f Biochemistry, Temple University, School of Medicine, Philadelphia, PA 19140 (U.S.A.)
(Received November 20th, 1980)
Key words: RNA synthesis; Dimethylnitrosamine; RNA modification; (Rat cell nucleus)
RNA synthesis was measured in nuclei isolated from rat liver and kidney 22 h post injection of 30 mg dimethylnitrosamine/kg body weight. The nuclear preparations were shown by electron microscopy to consist of clean hepatocytes and the liver nuclei showed no apparent necrosis at that time. In vitro RNA synthesis and methylation were proportional to time and nuclear concentration, as well as dependent on exogenous nucleoside triphosphates and S-adenosylmethionine. 60-70% of the in vitro synthesis was inhibited by 1/.tg/ml c~-amanitin. Total liver nuclear RNA synthesis was increased after dimethylnitrosamine exposure, but, unlike RNA synthesis in nuclei after partial hepatectomy, both ~,-amanitin-sensitive and -resistant synthesis were increased. Differences were found between dimethylnitrosamine-treated liver and kidney nuclear RNA synthesis which was sensitive to inhibition by 1-10/zg/ml ~,-amanitin, presumably a product of RNA polymerase Ill. Nuclear RNA methylation with S-adenosylmethionine, which was dependent on new RNA synthesis, differed between dimethylnitrosaminetreated rat liver and kidney nuclei. The endogenous RNA methyl substituents labeled in vitro showed differences in levels of methylation of bases, the 2'-0 position of ribose and caps in comparisons between control and dimethylnitrosamine-treated nuclei from both liver and kidney. Significant differences were obtained in both nuclear RNA transcription and methylation in vitro between the two tissues in response to pretreatment of the rat in vivo dimethylnitrosamine.
Introduction
The methylating carcinogen, dimethylnitrosamine, covalently interacts with DNA, RNA and protein [1,2]. In the rat, this carcinogen is metabolized in both the liver and the kidney; however, kidney tumors are induced only with doses of 30 mg/kg [3,4]. The liver becomes necrotic in response to this dose of dimethylnitrosamine, but it eventually regenerates and no tumors of the liver can be detected. It is widely believed that the carcinogen interactions result in alterations of gene expression,
* Present address: Department of Biochemistry, University of Nevada, Reno, NV 89557, U.S.A.
which represents the mechanism of action whereby chemical carcinogens are able to transform cells. Although several attempts have been made to measure the alteration of gene expression in transformed cells by hybridization methods [5-10], very little is known about the specific effects of dimethylnitrosamine on RNA transcription in the nucleus as pertaining to different classes of RNA. Therefore, carcinogen induced alterations in the expression of cellular genetic information were measured by nuclear RNA transcription in control and dimethylnitrosamine-treated rat liver and kidney nuclei. At the time for this measurement no major histological changes had occurred, but the early comparison of RNA synthetic activity possibly may reveal specific responses to dimethylnitrosamine, signaling events which will
0 005-2787/81/0000-0000/$02.50 © Elsevier/North-HollandBiomedicalPress
32 ultimately lead the two tissues to react in such a different manner. At 22 h post injection of dimethylnitrosamine the DNA of cells still contains a sufficient number of modifications by the carcinogen as detected by alkaline sucrose gradient sedimentation; however, no circulating dimethylnitrosamine can be detected in the animal [ 11 ]. At this same time, liver DNA synthesis and cell division has not yet begun [11] and gross liver necrosis is not detectable. But kidney cells isolated from as early as 4 h after exposure to a single dose of dimethylnitrosamine, have shown upon culture in vitro a 'transformed' appearance [ 12,13]. The nuclear system (described in this paper) from both rat liver and kidney is able to synthesize RNA in vitro from exogenously added nucleoside triphosphates and also carry out several of the methylation processing reactions observed in vivo. Since these nuclei represent their transcriptional state at the time of isolation, measurements utilizing added substrates can effectively monitor different aspects of gene expression in vitro and avoid the in vivo problems of labeling RNA such as pool size, transport and rate of phosphorylation of exogenously injected substrates. To differentiate between the activities of the three classes of eucaryotic RNA polymerases [14], the in vitro synthesized RNA products of liver and kidney nuclei from control and dimethylnitrosamine-treated animals were characterized according to their sensitivity to the inhibitor a-amanitin. These measurements, together with analysis of the endogenous methyl substituents of the different RNA classes, have revealed a differential response of the liver and kidney nuclear RNA synthetic capacity at 22 h after the initial exposure to dimethylnitrosamine. Experimental procedures
Animals Male Sprague Dawley rats weighing 150-200 g were maintained on Purina rat chow and water ad libitum. Dimethylnitrosamine was dissolved in 0.9% NaC1 and administered to the animals at 30 mg/kg intraperitoneally. Control animals received an equivalent amount of 0.9% NaC1 by the same route. Partial hepatectomy refers to removal of about 70% of the liver.
Chemicals Dimethylnitrosamine was purchased from East. man Organic Chemicals, Rochester, NY; 3H-labeled GTP and UTP, [a-32p]GTP and S-adenosyl[methyl3H]methionine were purchased from New England Nuclear. The a-amanitin was obtained from Sigma. Preparation of nuclei Liver and kidney nuclei were prepared by a modification of the procedure of Lynch et al. [3]. All steps were carried out at 0°C. Liver samples or kidneys were excised, quickly chilled, rinsed in 0.9% NaC1 and 0.25 M sucrose/5.5 mM CaC12. The tissue was then minced with scissors in 0.25 M sucrose/5.5 mM CaCI2, suspended in 10 vol./g tissue of the same medium and homogenized with 25 strokes of a loosefitting dounce homogenizer, followed by five strokes with a tight-fitting pestle. The homogenate was filtered through four layers of gauze and centrifuged in an IEC centrifuge for 3 min at 2000 rev./min. The supernatant fraction was discarded and the pellet was dispersed in 2 M sucrose/1 mM CaC12 and layered over a 10 ml cushion of 2 M sucrose/1 mM CaC12. The nuclei were sedimented through sucrose by centrifugation in a Beckman SW 27 rotor at 105 000 Xg for 45 min at 2°C. The nuclear pellet was drained and found to be free of contamination by whole cells as determined by staining with UNA (methyl-green pyronine vital stain). Nuclei were suspended in 0.I vol./g tissue of 0.3 M sucrose/1 mM MgC12/2 mM dithiothreitol and used immediately in assays as described. RNA synthesis measurements The conditions of RNA synthesis were essentially as described previously [15] in presence of 3 # M S-adenosyl [3H]methionine (10 Ci/mmol), 1 mM ATP, 0.5 mM CTP, UTP and GTP. In experiments using labeled UTP (spec. act. 400 mCi/mmol) or GTP (spec. act. 200 mCi/mmol) the UTP concentration was lowered to 0.1 raM, or GTP concentration to 0.25 raM. Incubations were carried out at 25°C for the indicated time periods and showed negligible incorporation of radioactivity from UTP and GTP in absence of added all four nucleoside triphosphates. Concentrations of a-amanitin were used as indicated for individual experiments. Total acid-insoluble incorporation was measured
33 by precipitation of an aliquot of the assay mixture with ice-cold 5% trichloroacetic acid/30 mM pyrophosphate. The precipitate was collected on Whatman GF/C f'dters and dried and the radioactivity was measured by liquid scintillation counting. The RNA synthesis was expressed as incorporation of radioactive precursor per mg DNA in the assay system. RNA extraction and analysis The nuclei were lysed in high salt buffer and treated with DNAase at room temperature [16] and the RNA was extracted with 0.5% sodium dodecyl sulfate/phenol/chloroform [17]. The extracted RNA was concentrated by ethanol precipitation and separated from mononucleotides and acid soluble oligonucleotides on a Sephadex G-50 column as described previously [ 15]. Extensive enzymatic digestion of the RNA was carried out in two steps. RNA was first digested with RNAase Ta [18] at pH 4.5, followed by digestion with ribonuclease A and T1 at pH 7.4 as described by Schibler et al. [19]. The mononucleotides, oligonucleotides and 'caps' were separated by DEAESephadex column chromatography in 7 M urea [20]. A partial digest of yeast RNA was used as absorbance marker to calibrate the column elution of the various charge peaks. Determination of DNA DNA sedimentation in alkaline sucrose gradients was carried out as described by Zubroff and Sarma [21]. The DNA in each fraction as well as the concentration of DNA in each nuclear preparation was determined fluorimetrically [21] in triplicate.
The period of 20-22 h post injection was, therefore, tested for the above criteria, since DNA synthesis commences shortly thereafter in the liver [ 11 ]. Carcinogens induce DNA-strand damage in vivo, which can be monitored by sedimentation analysis of DNA in alkaline sucrose gradients. To measure the presence of modified DNA in the two tissues, nuclei from liver and kidney were lysed directly on an alkaline sucrose gradient and sedimented as described by Zubroff and Sarma [21]. The gradient profiles of DNA from control and dimethylnitrosarnine treated animals are shown in Fig. 1. It is apparent that 22 h after exposure to dimethyinitrosamine the DNA of both liver and kidney still contains a sufficient number of modifications which can be detected as nicks by alkaline sucrose gradient sedimentation and shift the peak of DNA to lower molecular weight. The purity of the nuclear preparations which were to be used in RNA synthesis were checked by electron microscopy and a typical field is presented in Fig. 2A. The majority of the identifiable structures are clean hepatocyte nuclei with well-defined nucleoli. In comparison, the nuclear preparation from a hemorrhagic liver (48 h after intraperitoneal injection of 50 mg/kg dirnethylnitrosamine) contains poorly defined hepatic nuclei, as well as a large number of other cell types, including inflamatory cells (Fig. 2B), which can be readily distinguished from the control tissue. Nuclei, prepared from rat liver 22 h after exposure to dimethylnitrosamine at 30 mg/kg, (the time period utilized for measuring nuclear RNA transcription in this paper), appeared under the same
u)
Results
z
LIVER -i- DMN
KIDNEY -I- DMN
50
bJ Z
To examine nuclear RNA synthesis in rat liver and kidney after dimethyinitrosamine treatment, it was necessary to establish a time at which nuclei could be isolated from each tissue capable of transcribing RNA efficiently, contain an altered DNA template and yet not show any substantial change in nuclear popula. tion caused by liver necrosis. Numerous studies have been presented in the literature which have dealt with turnover of alkyl groups in DNA as well as other macromolecules after dimethylnitrosamine exposure.
~ 2o l
to,~o
4
8
12
16
4
8
12
I-6
SEDIMENTATION
Fig. 1. Effect of administration of dimethylnitrosamineon the sedimentation of DNA from liver and kidney 22 h post injection. Alkalinesuczosegradient sedimentation:o o~ control rat, • --, DMN injected rat.
34
/
Fig. 2. Electronmicrographs (X3630) of a section of nuclear pellet from rat liver after sedimentation through 2 M sucrose. Nucleaz pellet from: A, control rat liver; B, hemorrhagic rat liver, 48 h after intraperitoneal injection of 50 mg]kg dimethylnitrosamine and C, rat liver, 22 h after intraperitoneal injection of 30 mg/kg dimethylnitrosamine.
35 TABLE I DISTRIBUTION OF NUCLEI FROM LIVER IN PREPARATIONS FROM CONTROL AND DIMETHYLNITROSAMINETREATED RATS Pictures of nuclear spreads were taken at a magnification of ×6050. A random selection of fields (as in Fig. 2) was made and all structures were counted in each field. The following numbers of fields were counted for each experiment: Expt. I, control = six fields, dirnethyinitrommine 50 mg/kg = seven fields; Expt. II, control = ten fields, dimethylnitrosamine 30 rrtg/kg = ten fields. 'Other' covers any structures not identified either as hepatoeyte nuclei or inflamatory cells. Includes cross-sections of small portions of hepatocyte nuclei. All values are numbers of nuclei counted. Hepatocytes
Inflamatory cells
Number Experiment I Control Dimethyinitrosamine (50 mg/kg, 48 h) Experiment II Control Dimethylnitrosamine (30 mg/kg, 22 h)
%
Number
52
77
14
7
30
15
69
23
70
23
159
53
148
70
9
4
56
26
159
68
10
4
65
28
conditions indistinguishable (Fig. 2C) from control nuclei. The three types o f nuclear preparation were analyzed quantitatively as shown in Table I. As noted qualitatively in Fig. 2, the control and 30 mg/kg dimethylnitrosamine-treated nuclear preparations are indistinguishable in the distribution o f nuclear structures per field, with the great majority of structures being unambiguously hepatocyte nuclei. The nuclear preparation from a hemorhagic liver can be easily identified by the low percentage o f hepatocyte
121 LIVER
_o Q.
R
/
j~ ~o
3b Time
19 f ~--0-
(rain)
,'o
--
0
2'o
~o
Fig. 3. Time course of [~-32P]UTP incorporation in acid insoluble material by rat liver and kidney nuclei at 25° C with or without ~g/ml c~-amanitin: • •, control nuclei, o
o, with 1 pg/ml a-amanitin.
%
Number
%
nuclei, and the increased number o f various other cell types present. Electronmicrographic appearance of the nuclear preparations, therefore, suggests that the nuclear R N A synthesis experiments, after dimethylnitrosamine exposure, are actually carried out using comparable preparations o f liver nuclei. The ability o f b o t h liver and kidney nuclei to support R N A synthesis for a prolonged period o f time was investigated. Fig. 3 shows the time course of incorporation of radioactivity from UTP in acid
KIDNEY
LIVER
ib
Other
KIDNEY
/ Nucleor coficentrotion
Fig. 4. Incorporation o f [~-32p]uTP ill acid insoluble material by increasing concentrations of rat fiver and kidney nuclei after incubation at 25°C for 15 rain. Nuclear concentration is marked in arbitrary units where 1 = 470/Ag DNA/ ml for liver and 1 = 317/zg/ml for kidney nuclei.
36 insoluble product in control nuclei in presence and absence of 1 #g/ml c~ amanitin. Virtually identical data were obtained with nuclei from dimethylnitrosamine-treated animals. Continued incorporation can be seen in nuclei from both tissues with almost linear incorporation kinetics for 30 min by liver nuclei. Inhibition by 1 ~zg/ml a amanitin shows that 50 to nearly 70% of the RNA synthetic activity is due to RNA polymerase II. The proportionality of this incorporation to added nuclei over a range of concentrations is shown in Fig. 4. To compare the RNA synthetic capacity of nuclei from control and dimethylnitrosamine-treated animals, liver and kidney nuclei were incubated at 25°C in vitro and acid insoluble incorporation of [a-32P] UTP or [3H]GTP was measured. Summarizing the data of 4 - 6 experiments: (a) liver nuclei from control animals incorporated 104 -+ 14 pmol nucleoside triphosphate per mg DNA in a 15 min incubation, while liver nuclei from dimethylnitrosamine-treated animals (30 mg]kg) incorporated 153 -+35 pmol per mg DNA; (b) kidney nuclear incorporation under the same conditions was 100 _+26 pmol/mg DNA from control animals and 102 + 11 pmol/mg DNA from dimethylnitrosamine-treated animals. Nuclear recoveries were comparable from tissues of control and dimethytnitrosamine-treated animals as measured by recovery of DNA/g tissue, with liver yielding 1.54 -+0.02 mg DNA/g tissue and kidney 1.10 + 0.04 mg DNA/g tissue in a typical experiment. Experiments measuring incorporation by acid-precipitable or RNA-extraction methods gave comparable results.
Thus, no major difference in the total RNA synthetic capacity was observed between kidney nuclei of control and dimethylnitrosamine-treated rats, but the liver nuclei of dimethylnitrosamine treated animals showed a 50% increase in RNA synthetic capacity. Eucaryotic RNA polymerases have distinct susceptabilities to inhibition by c~-amanitin [22,23]. While most of RNA polymerase II activity is inhibited at concentrations of 1 gg/ml and RNA polymerase III is susceptible to concentrations of 10-100 gg/ml, RNA polymerase I, which is responsible for rRNA synthesis, is resistant to inhibition by a-amanitin even at 100/.tg/ml. Therefore, further analysis of the type of RNA synthesized in nuclei from control, dimethylnitrosamine treated and hepatectomized animals was carried out using a-amanitin inhibition of RNA synthesis in vitro. Table II summarizes the data obtained from several experimental animals. Only a low percentage of the in vitro synthesized RNA can be clearly attributed to RNA polymerase III, as is indicated by the differential sensitivity to inhibition by a-amanitin from 1 to 10/.tg/ml. This is shown in Table II only qualitatively, since 100 #g/ml is probably required for complete inhibition by this polymerase. Still, most of the changes observed in nuclei after dimethylnitrosamine treatment of the animal or after hepatectomy have to depend on the relative ratios of the products of RNA polymerases I and II. The data in Table II show that the most dramatic shift is encountered in the nuclei of hepatectomized animals where ribosomal RNA synthesis by polymerase I becomes predominant in both
TABLE II INHIBITIONOF NUCLEARRNA SYNTHESISIN VITRO BY a-AMANITIN Nuclei were prepared 22 h post-operative or after intraperitoneal injection of 30 mg/kg dimethylnitrosamine. [a-a2P]UTP incorporation in acid-insoluble material was measured in triplicate after 15 min at 25°C with or without a-amanitin. Nuclear activity in absence of a-amanitin is expressed as 100% for each nuclear preparation and is equivalent to values as stated in the text. Values are calculated from data as pmol UTP incorporated/mg DNA + S.D. Liver nuclei a-Amanitin: 0 Control Dimethylnitrosamine treated Hepatectomized *
100 100 100
* This represents an average from two animals.
Kidney nuclei
1 ~g/ml
10 ug/ml
0
1 ~g/ml
10/~g/ml
58_+11 52 _+10 70
53_+ 1 53 _+10 68
100 100 100
31_+ 3 40 _+10 62
28±5 37 + 8 61
37
liver and kidney nuclei as shown by its proportional resistance to a amanitin. In contrast, the proportions of 1/.tg/ml et-amanitin sensitive and resistant RNA synthesis do not appear to differ greatly between the dimethylnitrosamine-treated and control animals. This finding indicates that changes in the RNA synthetic pattern observed in the liver nuclei from dimethylnitrosamine-treated animals are not primarily due to the induction of cell cycle events as in the case after partial hepatectomy. To characterize the increased RNA synthesis products in dimethylnitrosamine-treated animals and to determine if rat liver and kidney nuclei respond differentially to dimethylnitrosamine in RNA synthesis and modification reactions, a double label experiment was carried out in which we measured the in vitro RNA synthesis by incorporation of [a-32P]UTP (in presence of 1 and 10 /zg/ml a-amanitin), and measured the products of endogenous RNA methylation. It can be shown that exogenously added S-adenosyl[methyl-3H]methionine is utilized by isolated liver and kidney nuclei in endogenous methylation reactions leading to methylation of nuclear RNA. Ribo. somal RNA precursors, Hn RNA, tRNA and most of the small nuclear RNA species are methylated. Since some RNA species are predominantly methylated in the 2.OH position of ribose (such as rRNA and some of the small nuclear RNAs), some are base methylated (such as tRNA) and for some others (Hn RNA and some of the small nuclear species) the 'cap' structures are characteristic of the 5' terminus, analysis of endogenous methylation products gives additional information about carcinogen induced changes in gene expression. Fig. 5 shows the column elution prof'fle of an RNA digest of in vitro labeled nuclear RNA from liver nuclei incubated with S-adenosyl[methyl-3H]methionine. A substantial number of counts elute with the mononucleotide fraction (base methylation) with a - 2 charge, a large peak of dinucleotide (2-O-ribose methyl) elutes with a charge of - 3 and a distinct peak of labeled material elutes in the 'cap region' with a charge of - 5 to - 6 . Since all three categories of modification are formed in vitro, it appeared possible to monitor them in conjunction with ct amanitin sensitivity and to evaluate any differences in RNA synthetic capacity of the rat liver kidney nuclei after dimethylnitrosamine treatment. Table III summarizes the data from such an experi-
-2
-3
-4
-5
-6
50 I
0m I(
=E
30
(,.) -r
IO
:r
2'0
J 60
\ I00
140
(FRACTION NO.)
Fig. 5. Analysis of the methylated derivatives of in vitro labeled rat liver nuclear RNA. The nuclei were labeled with S'adenosyl[methyl-3H]methionine under conditions of RNA synthesis as described in Methods. The RNA was extracted, hydrolyzed sequentially with RNAases T2, T1 and A, and the products were separated on a DEAE-Sephadex A-25 column in 0.02 M Tris-HC1 (pH 7.4)/7 M urea with a 0.1-0.5 M NaC1 gradient in the same buffer. Arrows indicate charge positions of elution of absorbance (260 rim) markers from a yeast RNA hydrolysate.
ment for liver and kidney nuclei from control and dimethylnitrosamine-treated rats, in which RNA synthesis and methylation were monitored as a function of ,v-amanitin sensitivity. A number of important differences can be discerned from the nuclear synthetic capacity of the two tissues. Kidney nuclei again show a slightly greater percentage of the total RNA product as sensitive to 1/.tg/ml a amanitin, namely transcript by RNA polymerase II. Both liver and kidney control nuclei showed a specific 7-8% transcript sensitivity to 10 gg/ml et amanitin, the product prim. arily of RNA polymerase III. However, this percentage was lower in nuclei from both tissues of the dirnethylnitrosamine.treated animal, with the kidney showing a 4% inhibition between 1-10 /.tg/ml
38 TABLE 11I a-AMANITIN INHIBITION OF IN VITRO NUCLEAR RNA SYNTHESIS AND METHYLAT1ON Nuclear RNA was synthesized in vitro in presence of [a-32p]UTP and S-adenosyl [methyl-3H]methionine at the indicated concentration of a-amanitin. The RNA was isolated and extensively digested with T2, T1 and Pancreatic A ribonucleases and the digests analyzed by column chromatography as described in Methods. All values are cpm. All counts were normalized for RNA recovery. Liver preparations = cpm/2.2 mg DNA; kidney preparations = cpm/0.8 mg DNA. Liver nuclei a-Amanitin: 0 Control [a-a2P]UTP 3 H-labeled: Mononucleotide Dinucleotide 'Cap region' Dimethylnitrosamine [a.32p] UTP a H-labeled: Mononucleotide Dinucleotide 'Cap region' Dimethylnitrosamine/control [a-a2P]UTP 3H-labeled: Mononucleotide Dinucleotide 'Cap region'
Kidney nuclei 1 ug/ml
10 ~g/ml
0
1 ug/ml
471 112
301 418
33 891
139932
97 467
11 210
4026 3732 292
178 30 94
818 702 111
852 776 122
255 13
12 207 n.a.
577 995
373 767
10 390
191 708
133 780
8678
5995 5929 446
1115 645 195
212 679 20
1148 932 136
189 0
83 305 n.a.
n.a.
n.a.
10 pg/ml
1.25
1.25
0.31
1.37
1.28
0.77
1.49 1.59 1.53
6.26 21.5 2.07
0.26 0.97 0.18
1.34 1.20 1.15
0.74 0 -
6.92 1.47 -
a-amanitin and the liver a 2% inhibition. The largest differences between the two tissues are, however, observed by analyzing the in vitro methylation of nuclear RNA and their response t o inhibition of RNA synthesis by a amanitin. We have shown previously that methylation of RNA in isolated L-cell nuclei is closely linked to RNA transcription, particularly as measured by a-amanitin sensitivity of the RNA polymerase II product [15]. As shown in Table III, the levels of endogenous RNA methylation were significantly higher for nuclei isolated from dimethylnitrosamine-treated animals, with the highest levels of methylation observed for all modification types from liver nuclei. Notable differences were also seen between the liver and kidney nuclear RNA modifications after the carcinogen treatment. Thus, while the ratios of methylation were greater than the observed increase in RNA synthesis in the liver nuclei from dimethylnitrosamine-treated animals, the increased rate of methylation was below
the proportion expected by increased RNA synthesis of the kidney nuclei. Several other differences can be observed for RNA modification between the two tissues and their response to dimethylnitrosamine. The mononucleotide methylation is more sensitive to inhibition by 1 #g/ml a-amanitin in dimethylnitrosamine-treated liver nuclei (20%) than in control nuclei (4%). Control kidney nuclei, on the other hand, show a greater sensitivity to inhibition by 1 #g/ml a-amanitin (30%), than the carcinogen-treated kidney nuclei (16%). Moreover, while liver control and dimethylnitrosamine-treated nuclei show approximately the same degree of inhibition of dinucleotide formation by 10 /.tg/ml o¢-amanitin (19 and 12%, respectively), kidney nuclei (control and dimethylnitrosamine-treated) show a higher level of inhibition by the same concentration o f a-amanitin (27 and 33%, respectively). The 'cap' region sensitivity to a-amanitin inhibition could not be measured accurately in kidney nuclei, but
39 most of the counts in the region are inhibited by a-amanitin, showing that the formation of these structures is not dependent on RNA transcription by RNA polymerase I, and therefore, are certainly true cap structures. Differences in response were noted (Table III) in a-amanitin inhibition of 'cap' formation between control and dimethylnitrosamine treated liver nuclei with a large reduction in the sensitivity between 1 and 10#g a-amanitin in the treated liver nuclei, indicating a change in the proportion of 'capped' molecules between those that are transcrined by RNA polymerase II, and those that are transcribed by RNA polymerase III. Discussion
The differential effects of a single dose of 30 mg dimethylnitrosamine/kg on rat liver and kidney have been well documented in terms of liver necrosis and subsequent kidney tumor production [4,1,24,12]. The interaction of this carcinogen with cellular maeromolecules has also been extensively investigated. However, the subsequent differential gene expression in these tissues leading to such widely divergent pathways as necrosis and tumorigenesis has been difficult to approach. Some measurements have been made on the level of biosynthesis of nuclei acid and protein in the whole animal [11]. Close correlation has also been established for the persistence of DNA methylation by dirnethylnitrosamine in rat liver and kidney and the events in tumor induction [25-27]. Since O6-methylguanine can lead to misincorporation of UMP or AMP in vitro by RNA polymerase [28], it is possible that such altered DNA templates would also alter gene expression in vivo. This report presents data which indicate that comparable nuclear preparations from liver and kidney, 22 h after exposure to dimethylnitrosamine indeed show differences in gene expression. These differences are apparent not only between nuclei of the two tissues, but also between the dimethylnitrosamine-exposed nuclei and those from control animals. Analysis of the accumulation of abundant frequency cytoplasmic RNAs between hepatoma and regenerating liver by Reiners and Busch [34] has led to the suggestion that the differences in gene expression in this case are due to regulation at the transcriptional or post-transcriptional level.
To evaluate these results, several other aspects of dimethylnitrosamine action on tissues must be considered. Time after exposure to dimethylnitrosamine is important in these experiments, since (1) extensive necrosis of the liver would change the composition of the nuclear preparation as shown in Fig. 2; and (2) alkylation of protein and RNA during a time when the active species of the carcinogen is present in the tissues may obscure results of experiments measuring transcription and processing. Thurberville and Cra. dock [33], have reported maximal alkylation of liver histones by dimethylnitrosamine about 5 h after exposure with a decline between 5 and 24 h. Finally, the question of inhibition of any of the RNA polymerases needs to be considered, since it has been shown [29] that dimethylnitrosamine inhibits rat liver nuclear RNA synthesis 4-12 h after exposure to a dose of 40 mg]kg body weight and that this inhibition can be correlated with inactivation of RNA polymerase II. Such inactivation was not observed in the present studies (with 30 mg/kg dose) at 22 h post injection, since 1 btg/ml a-amanitin-inhibitable RNA synthesis represents the same percentage of total RNA synthesis in both liver and kidney after dimethylnitrosamine treatment. Presumably, the RNA polymerase would show higher levels of protein alkylation at 4-12 h after exposure to dimethylnitrosamine at 40 mg/kg than at 22 h at 30 mg/kg body weight, and thus explain the apparant discrepancy in results. It is therefore possible that the RNA polymerase II inactivation [29] is overcome at 22 h post injection in absence of circulating dimethylnitrosamine. Temporary inactivation of RNA poly. merase II population by alkalytion of the enzyme ( 4 - 1 2 h after dimethylnitrosamine administration) seems likely in view of the results reported by the same authors [29] which showed in vitro inactivation of the polymerase by methylmethanesulfonate. Eliminating these alternative explanations of the data, the present observations show a differential response to dimethylnitrosamine in both transcription and methylation patterns of the two tissues. Differences in total RNA synthesis were observed between liver and kidney with the liver showing an increased rate of synthesis after dirnethylnitrosamine exposure. The increased rate appeared to be due to activity by both RNA polymerase I and RNA poly. merase II, in contrast to liver nuclei from a hepatec-
40 tomized animal, where the predominant RNA synthetic activity was represented by RNA polymerase I. The results from endogenous in vitro RNA synthesis using liver nuclei from hepatectomized animals agree with those reported by Yu [35], with similar levels of in vitro inhibition by a-amanitin. Differences in the level of RNA polymerase II activity were also noted between dimethylnitrosamine-exposed liver and kidney nuclei. Since RNA polymerase III varies widely in its susceptibility to a-amanitin inhibition, the concentration (10 #g/ml) used in these experiments was most likely insufficient to effect complete inhibition of its activity. Therefore, the effects of dimethylnitrosamine on polymerase III activity can be compared only in qualitative and not quantitative terms for the two tissues. However, since inhibition is proportional to the log of concentration of a-amanitin [36] the maximum difference in RNA polymerase III products could be a factor of 2. Since both liver and kidney contain several cell types, one must necessarily remain cautious on the type of differential response to dimethylnitrosamine encountered in these experiments. It has been shown by Pegg and Hui [26,27] that alkylation of liver DNA after injection of dimethylnitrosamine is about 10times as efficient as that for kidney DNA, while the removal of O6-guanine proceeds more slowly from kidney than liver DNA. These two processes in vivo may be expected to affect the transcription measured in vitro. Similarily, since only a subpopulation of kidney cells may give rise to tumors subsequent to dimethylnitrosamine injection [13], the differential transcriptional and processing response produce d in vitro may not represent accurately the actual event leading to tumorigenesis. However, the ability to detect changes in gene expression after exposure to dimethylnitrosamine at this level, provides a system for further analysis using limiting cell populations. Most eucaryotic RNA species have to undergo a series of post-transcriptional processing steps for the production of functional RNA. One of these steps is specific methylation of the RNA. High tRNA methyltransferase activity has been reported for malignant tumors in comparison of their corresponding normal tissue [30,31] and altered activities of methyltransferases have been described in neoplastic cells [32]. It is interesting to note that analysis of methylated products in nuclear RNA between liver and kidney
after dimethylnitrosamine treatment also show significant differences in these patterns and their change in response to this compound. The nuclear system at this point has not been able to differentiate between that fraction of RNA methylation reactions involving ribosomal RNA precursors and a pool of undermethylated RNA molecules in the nuclei. However, the differences in methylation patterns which depend on new RNA synthesis indicate that the two tissues indeed respond differently to dimethylnitrosamine not only in levels of transcription by the eucaryotic polymerases, but also in subsequent processing of the transcribed product. The processing differences may reflect either different transcription products, or alterations which have occurred in the processing steps themselves. Experiments are being directed to differentiate between these possibilities. Acknowledgements The author thanks Ms. Anne Gunn for her excellent technical assistance, Ms. Julia Zubroff for hepatectomies and all the DNA measurements in the course of this work, and Mr. Joe Martin and Dr. Gordon Hard for various aspects of the electron microscopy. This work was supported in part by the NIH Biomed GSR Grant, Temple University, School of Medicine. References 1 Magee, P.N., Pegg, A.E. and Swann, P.F. (1975) In Handbuch der AUgemeinen Pathologic (Grundman, E., ed.), pp. 329-420, Springer-Verlag,Berlin 2 Lawley, P.D. (1976) in Chemical Carcinogens (Searle, C.E., ed.) ACS Monograph Series No. 173, pp. 83-244 American Chemical Society, Washington, DC 3 Lynch, W.E., Brown, R.F., Umeda, T., Langreth, S.G. and Lieberman, I.J. (1970) Biol. Chem. 245, 3911-3916 4 Magee, P.N. and Barnes, J.M. (1962) J. Path. Bacteriol. 84, 19-31 5 Drews, J., Brawerman, G. and Morris, H.P. (1968) Eur. J. Biochem. 3,284-292 6 Shearer, R.W. and Smuckler, E.A. (1971)Cancer Res. 31, 2104-2109 7 Garrett, C.T., Moore, R.E., Katz, C. and Pitot, H.C. (1973) Cancer Res. 33, 2469-2475 8 Williams, J.G., Hoffman, R. and Penman, S. (1977) Cell 11,901-907 9 Getz, M.J., Reiman, Jr. H.M., Siegal, G.P., Quinlan, T.J., Proper, J., Elder, P.K. and Moses, H.L. (1977) Cell 11, 909-921
41 10 Hixsch, F.W., Nail, K.N., Hayes, L.C., Raju, K.S., Spohn, W.H. and Busch, H. (1977) Cancer Res. 37, 3694-3700 11 Stewart, B.W. and Magee, P.N. (1971) Biochem. J. 125, 943-952 12 Borland, R. and Hard, G.C. (1974) Eur. J. Cancer 10, 177-184 13 Hard, G.C., King, H., Borland, R., Stewart, B.W. and Dobrostanski, B. (1977) Oncology 34, 16-19 14 Roeder, R.G. and Rutter, W.J. (1969) Nature 224, 234 15 Winicov, I. and Perry, R.P. (1976) Biochemistry 15, 5039-5046 16 Penman, S. (1966) J. Mol. Biol. 17,117-130 17 Perry, R.P., LaTorre, J., Kelley, D.E. and Greenberg, J.R. (1972) Biochim. Biophys. Acta 262,220-226 18 Winicov, I. (1977) Biochemistry 16, 4233-4237 19 Sehibler, U., Kelley, D.E. and Perry, R.P. (1977) J. Mol. Biol. 115,695-714 20 Perry, R.P., Kelley, D.E., Frederici, K.H. and Rottman, F. (1975) Cell 4,387-394 21 Zubroff, J. and Sarma, D.S.R. (1976) Anal. Biochem. 70, 387-396 22 Balatti, S.P., Ingles, C.J., Lindell, T.J., Morris, P.W., Weaver, R.F., Weinberg, F. and Rutter, W.J. (1970) Cold Spring Harb. Syrup. Quant. Biol. 35,649-657 23 Weinman, R. and Roeder, R.G. (1974) Proc. Natl. Acad. Sci. U.S.A. 71, 1790-1794
24 Magee, P.N. and Barnes, J.M. (1967) Adv. Cancer Res. 10,163-246 25 Nicoll, J.W., Swann, P.F. and Pegg, A.E. (1975) Nature 256,261-262 26 Pegg, A.E. and Hui, G. (1978) Biochem. J. 173,739-748 27 Pegg, A.E. and Hui, G. (1978) Cancer Res. 38, 20112017 28 Gerchman, L.L. and Ludlum, D.B. (1973) Biochim. Biophys. Acta 308,310-316 29 Herzog, J. and Farber, J.L. (1976) Cancer Res. 36, 1761-1770 30 Waalkes, T.P., Adamson, R.H., O'Gara, R.W. and GaUo, R.C. (1971) Cancer Res. 31, 1069-1073 31 Lakings, D.B., Waaikes, T.P., Borek, E., Gehrke, C.W., Mrochek, J.E., Langmore, J. and Adamson, R.H. (1977) Cancer Res. 37,285-292 32 Liau, M.C., Hunt, M.E. and Hurlbert, R.B. (1976) Biochemistry 15, 3158-3164 33 Thurberville, C. and Craddock, V.M. (1971) Biochem. J. 124,725-739 34 Reiners, J.J., Jr. and Busch, H. (1980) Biochemistry 19, 833-841 35 Yu, F.L. (1975) Biochim. Biophys. Acta 395,329-336 36 Hossenlopp, P., Wells, D. and Chambon, P. (1975) Eur. J. Bioehem. 58,237-251