Ageing studies in rat liver I. Complexity of RNA from two to ten months of age

Mechanisms of Ageing and Development, 15 (1981) 415-421

415

©Elsevier Sequoia S.A., Lausanne - Printed in the Netherlands

A G E I N G S T U D I E S IN R A T L I V E R . I. C O M P L E X I T Y O F R N A F R O M TWO T O T E N M O N T H S O F A G E

LEO J. GRADY and WAYNE P. CAMPBELL New York State Department o[ Health, Albany, New York 12201 (U.S.A.)

(Received June 16, 1980, in revised form November 13, 1980)

SUMMARY A highly sensitive nucleic acid hybridization assay was used to compare the extent of nonrepetitive DNA transcription in rat liver between the ages of two and ten months. The basic approach consisted of initially purifying the DNA expressed in liver at these ages and then using it in reactions with homologous and heterologtaus RNAs. Such experiments failed to reveal any differences in nonrepetitive DNA transcription as a function of age. The possibility was also explored that there might be an age-associated variation in the proportion of the total RNA complexity attributable to poly(A+) RNA. These experiments, too, were negative in that the poly(A+) RNA was found in all cases to account for approximately 50% of the total sequence diversity. Overall, these data strongly suggest that ageing is not accompanied by a steady, progressive change in the regions of the genome transcribed, either quantitatively or qualitatively.

INTRODUCTION The evidence suggesting that at the cellular level alterations in RNA transcription may be associated with the ageing process has been reviewed by Cutler [ 1]. In this regard it should be recalled that the genome of eukaryotic cells contains some DNA sequences that are repetitive and some which are nonrepetitive, and changes in transcription could occur in either or both categories. However, the fact that repetitive DNA consists of families of closely related, but not identical, sequences complicates the interpretation of R N A - D N A hybridization experiments [2] and for this reason we have focused our attention on the transcription of nonrepetitive DNA. In a series of experiments with mouse brain and liver, Cutler [3] obtained evidence that in each tissue the proportion of the nonrepetitive DNA transcribed as well as the actual RNA species synthesized varies with age. One difficulty with most earlier studies of this kind is that only a small fraction of the nonrepetitive DNA is transcribed (10% or less in liver) and the reproducibility of the experiments is at best + 1%. Consequently, it is

416 often hard to determine whether any small differences that occur are really significant. In the work described herein a 6-10-fold increase in sensitivity was achieved by isolating the DNA from RNA-DNA hybrids and then using it in reactions with homologous and heterologous RNAs. Using this approach, no differences in nonrepetitive DNA transcription could be detected over the first one.third of the life-span of the rat.

METHODS

Animals Male Fisher F344 rats were used exclusively and were obtained from the ageing colony of the National Institute on Ageing at the Charles River Breeding Laboratory. A strict cycle of 14 hours light, 10 hours dark was maintained, with food (Wayne Lab-Blox F6) and water available ad libitum. The animals were held for at least I0 days prior to use to allow adaption to this regimen. All surgical procedures were conducted between 9 and 10 a.m. to minimize diurnal variation. Isolation of total cell RNA Livers were removed from animals under ether anesthesia approximately two hours before the end of a light cycle. Immediately upon removal, livers were rapidly blotted with surgical gauze to remove excess blood, placed on a fiat aluminum plate which rested on a block of dry ice, and compressed into a thin wafer using a metal cylinder which had been cooled in liquid nitrogen. The quick-frozen liver was ground into a fine powder in a large mortar filled with liquid nitrogen. The nitrogen was then allowed to evaporate and the powdered, frozen liver was transferred to a beaker containing roughly ten times the liver volume of 0.01 M sodium acetate (pH 5.0), 0.1 M NaCI, and 125/lg/ml heparin at 4 °C. After thorough mixing, sodium dodecyl sulfate (SDS) was added to 0.5% and the RNA purified by hot phenol extraction as related elsewhere [4]. Purification of nuclear RNA When nuclei were to be prepared, the powdered liver was added to five times its own volume of 0.01 M Tris°HC1 (pH 8.4), 0.14 M NaCI, 0.0015 M MgC12, and 125 #g/ml heparin at 4 °C. An equal volume of the same buffer containing 0.1% Triton X-IO0 was rapidly stirred in and the mixture gently agitated for 5 min at 4 °C. The nuclei were collected by centrifuging for 5 min at 1000 g and 4 °C. The nuclei were gently resuspended in the above buffer and again agitated for 5 min in the presence of 0.05%Triton X.IO0. After collection by centrifugation, the nuclei were resuspended in the same buffer used for total RNA isolation. A small aliquot was removed, diluted, stained with trypan blue and counted in a hemacytometer to determine the yield of nuclei ( 5 - 7 × 10V/g of liver). The nuclei appeared clean with no evidence of cytoplasmic tabs. The remainder of the solution was made 0.5% with sodium dodecyl sulfate (SDS) and subjected to hot phenol extraction as above. The poly(A+) nuclear RNA was isolated on oligo(dT)-~llulose type 3 (Colloborative Research) using the procedure of Banfle et al, [5]. The final poly(A+)

417 and poly(A-) RNAs were precipitated with LiC1, collected by centrifugation, resuspended in 0.01 M Tris.HC1 (pH 7.3) and stored at - 2 0 °C until used. On a weight basis, nuclear RNA was equivalent to about 10% of the total cellular RNA. Likewise, about 10% of the nuclear RNA consisted of poly(A+) molecules. In vitro labeling of DNA The methods for isolating the nonrepetitive portion of rat fiver DNA and radiolabeling it in vitro with E. coli DNA polymerase I have been published [6]. Specific activity of the labeled DNA was between 1.5 and 2.0 × 107 tritium cpm//ag. Isolation of DNA sequences transcribed in liver The exact details of the procedure for isolating the DNA sequences expressed in liver (hereafter referred to as EDNA) have been given elsewhere [6]. Basically, the approach consists of isolating the DNA component of RNA-DNA hybrids after two successive rounds of hybridization. The final DNA obtained reacts to about 60% with homologous RNA. This DNA corresponds to the template from which the RNA species present in liver were transcribed under the experimental conditions which prevailed at the time of RNA extraction. Hybridization conditions Hybridization reactions were carried out and samples analyzed on hydroxylapatite as before [6]. The general reaction contained 1.2 mg of RNA and 0.001/~g of EDNA in 0.1 ml of 0.4 M phosphate buffer (composed of equimolar amounts of NaH2PO4 and Na2HPO4), 0.01 M EDTA, and 0.1% SDS. In experiments with nuclear poly(A+) RNA, the quantity of RNA employed was reduced to 24 pg.

RESULTS Before presenting the experimental results, we believe that a brief discussion of the sensitivity of the hybridization method used is in order. The haploid molecular weight of the rat genome is about 1.8 × 1012 [7], and of this 65-70% consists of nonrepetitive sequences [6, 7]. The nonrepetitive portion of the genome therefore represents 1.26 X 1012 daltons, or roughly 2.1 X 10 9 nucleotide pairs (NTP). We have previously shown [6] that under our conditions 6.0--6.7% of the nonrepetitive sequences are transcribed in rat liver, corresponding to 1.33 X 108 NTP. The methods employed in the present experiments increase the sensitivity of the measurements approximately ten-fold (i.e. the EDNA hybridizes to about 60% with homologous RNA). Thus, since we previously showed [6] that experimental differences in the neighborhood of 5% were readily detected, changes equivalent to as little as 0.3% of the nonrepetitive DNA could have been observed. Assuming asymmetric transcription, this means that shifts in RNA complexity equal to or greater than 1.3 X 107 nucleotides would be revealed. There are two final points to be noted. First, saturation values must be viewed as minimal estimates

418

of complexity because the possibility that sequences exist which occur at too low a frequency to be detected cannot be ruled out with certainty. Second, although the method employed in the present study is very sensitive to changes in sequence complexity, it is relatively refractory to shifts in the abundancies of the various RNAs. In the first series of experiments the fraction of the nonrepetitive DNA expressed (EDNA) in rat liver at the age of 2 months was isolated and then used in hybridization reactions with liver RNA prepared from animals that were 2, 3, 4 and 10 months old, As shown in Fig. 1, the EDNA hybridized to the same extent with each of the RNAs employed, thus indicating that all of the RNA species present in liver at 2 months of age are also present at the ages of 3, 4 and 10 months. ~

°

100

N

8o 60 < 4o

zc i

i

i

20

i

i

40

i

I

i

60

i

80

HOURS

i

I00 OF

L

i

i

120

i

140

i

i

160

L

i

L

IOO 200

INCUBATION

Fig. 1. Hybridization of 2-month EDNA with RNA from liver at various ages. Bars show the range of values for experiments using two independently isolated 2-month RNA preparations, one 3-month RNA preparation, one 4-month RNA preparation and two independently isolated 10-month RNA preparations. The data were normalized by setting the average saturation value (63%) of the 2-month EDNA vs. 2-month RNA reactions equal to 100%. I00 N

~,

eo

~ 6o ~

40

~

20 i 0

I 20

t

1 40

i

i 60

i

J 80

L

i i ~ L i i i i ! 10(3 120 140 160 180

200

HOURS OF INCUBATION

Fig. 2. Hybridization of 10-month EDNA with RNA from liver at 2, 3, 4 and 10 months of age. Bars show range of values for experiments using the same RNA preparations employed to obtain the data in Fig. I. The data were normalized by setting the average saturation value (65%) of the 10-month EDNA vs. 10-month RNA reactions equal to 100%.

In order to compare the portion of the nonrepetitive DNA transcribed in 10-monthold liver with that expressed at earlier ages, a 10-month EDNA was prepared and then hybridized with the same RNAs used in the initial experiments. It is clear from Fig. 2 that, as before, the age of the liver from which the RNA was extracted had no effect on the hybridization of the EDNA. These results show that all of the complex class of RNA molecules found in 10-month-old liver are present as well in the liver of younger animals.

419 The data obtained with 2-month EDNA and 10-month EDNA complement each other and lead to the conclusion that in rat liver, no changes of a magnitude of 1.3 X 107 nucleotides or greater occur in the regions of nonrepetitive DNA transcribed during the interval from 2 to I0 months of age. The fact that some RNA molecules are polyadenylated at their 3'-terminus is well established (for review, see refs. 8 and 9). Although the function of these poly(A) residues is not yet clear, there is some testimony that they influence RNA stability [9]. There is also increasing evidence that the poly(A)-containing RNA accounts for only a portion of the total RNA complexity [10-12]. On these premises it can be hypothesized that ageing may be accompanied by an alteration in the turnover rate of some RNA species and that this in turn might be reflected by a change in the number of different RNA molecules (proportion of total complexity) that are polyadenylated. This possibility has been explored by comparing the fraction of the total RNA diversity accounted for by poly(A+) molecules in the liver of animals at 3 and 10 months of age. The results of these experiments are presented in Fig. 3. When EDNA prepared from 3-month liver total RNA was reacted with nuclear poly(A+) and poly(A-) RNA from 3-month liver, the outcome was as shown in Fig. 3A. Only 50-55% of the EDNA hybridized with poly(A+) nuclear RNA, a situation similar to that reported by Bantle [12] for mouse liver. On the other hand, nuclear poly(A-) RNA saturated at the same level as total RNA. Since the addition of poly(A) is a posttranscriptional event [9], this latter result was not too surprising. Hybridization of the same 3-month EDNA with nuclear poly(A+) and poly(A-) RNA from lO-month liver

PO0 A

~ /

•

I

-

60 Q. uJ

40

~

20

~

0

I

i

i

,

i

i

f

i X~

40 ~

&

100 Fz

J

B

80

20 0

, J , i i t , , , J , i , i J i J ~ 40 60 80 IO0 t20 t40 160 180 HOURS OF INCUBATION

Fig. 3. (A) Hybridization of 3-month EDNA with.the following fractions of 3-month liver RNA: (e e), total cell; (m 1), nuclear poly (A-), (A A), nuclear poly(A+). Data were normalized by setting the m a x i m u m extent of hybridization obtained with total cell R N A (59%) equal to 100%. (B) Reaction of 3-month E D N A with the following fractionsof 10-month liverRNA:

(O. O), total cell; (I~ normalized as above.

l~), nuclear poly (A-); (4

A), nuclear poly(A+). Data were

420 gave results corresponding to those with 3-month RNA (Fig. 3B). Repetition of these experiments with EDNA isolated using 10-month liver total RNA (data not shown)also failed to reveal any differences either in the fraction of the total complexity residing in poly(A+) RNA or between 3-month and 10-month liver RNAs. We therefore conclude that during the first one-third of the life.span of the rat there is no substantial change in the proportion of the total liver RNA complexity contributed by polyadenylated molecules.

DISCUSSION In his original work with mouse liver, Cutler [3] found an inverse correlation between the age of the animal and the extent to which nonrepetitive DNA sequences were transcribed. The pattern observed was one of a steady decline in the amount of nonrepetitive DNA expressed with no apparent shoulder (or plateau) occurring in young animals. Using an improved hybridization methodology we have been unable to demonstrate a similar phenomenon in rat liver up to an age of 10 months. It is conceivable, however, that a reduction in the proportion of nonrepetititve DNA takes place at later ages and this eventuality is currently being explored. The discrepancy between our results and those of Cutler could also arise from important differences in the test systems employed. Nevertheless, the present data clearly show that a regular decrease in liver RNA complexity is not a universal accompaniment of ageing. RNA synthesis as a function of age has been studied in the nuclei of liver cells from F344 rats by Castle et al. [13]. They found an increase in messenger RNA and ribosomal RNA synthesis from 3 to 6 months of age, followed by a decrease over the period from 6 to 31 months. Since these investigators were concerned with the rate and extent of RNA transcription and not the actual diversity of the molecules produced, their data are neither incompatible with, nor directly comparable to, those presented herein. Finally, the results that we have obtained so far tend to argue against any model for ageing that calls for a steady progressive change in the regions of the genome transcribed, either quantitatively or qualitatively. Furthermore, our failure to observe any age-related variation in the proportion of the total RNA complexity represented by poly(A+) molecules suggests that, within the limits of detection, there is no intimate relationship between polyadenylation and ageing. After the present manuscript was completed, an independent report by Colman et al. [14] was brought to our attention in which it was shown that no detectable changes occur in the complexity of rat brain poly(A+) RNA between the ages of 2 and 32 months. The data presented herein serve not only to confirm their results in another tissue, albeit over a narrower age range, but also to extend them to include nonpolyadenylated RNAs as well. The finding of such similar results in two different tissues substantially strengthens the conclusions set forth above.

421

ACKNOWLEDGEMENTS We w o u l d like to t h a n k Mrs. Arlene B. N o r t h for her excellent technical assistance and Mrs. N a n c y Miller for typing the manuscript. This w o r k was s u p p o r t e d in part b y U.S. Public Health Service Grant No. AG 0 0 2 0 7 f r o m the National Institute on Aging.

REFERENCES 1 R. G. Cutler, Crossqinkage hypothesis of aging: DNA adducts in chromatin as a primary aging process. In K. C. Smith (ed.), Aging, Carcinogenesis, and Radiation Biology, Plenum Press, New York, 1976, pp. 4 4 3 - 4 9 2 . 2 B. J. McCarthy and R. B. Church, The specificity of molecular hybridization reactions. Annu. Rev. Biochem., 39 (1970) 131-150. 3 R. G. Cutler, Transcription of unique and reiterated DNA sequences in mouse liver and brain tissues as a function of age. Exp. GerontoL, 10 (1975) 37-60. 4 L. J. Grady and W. P: Campbell, Nonrepetitive DNA transcription in mouse cells grown in tissue culture. Nature, 243 (I 973) 195 -198. 5 J. A. Bantle, I. H. Maxwell and W. E. Hahn, Specificity of oligo(dT)-cellulose chromatography in the isolation of polyadenylated RNA. Anal. Biochem., 72 (1976) 413 - 4 2 7 . 6 L. J. Grady, W. P. Campbell and A. B. North, Nonrepetitive DNA transcription in normal and regenerating rat liver. Nucleic Acids Res., 7 (1979) 259-269. 7 D. S. Holmes and J. Bonner, Sequence composition of rat nuclear deoxyribonucleic acid and high molecular weight nuclear ribonucleic acid. Biochemistry, 13 (1974) 841-848. 8 B. Lewin, Gene Expression - 2. Eukaryotic Chromosomes, John Wiley and Sons Ltd., London, 1974, pp. 282-298. 9 M. Revel and Y. Groner, Post-transcriptional and translational controls of gene expression in eukaryotes. Annu. Rev. Biochem., 47 (1978) 1084-1085. 10 J. A. Bantle and W. E. Hahn, Complexity and characterization of polyadenylated RNA in mouse brain. Cell, 8 (1976) 139-150. 11 M. Jacquet and F. Gros, Expression of single copy DNA sequences in nuclear RNA from undifferentiated mouse embryonal carcinoma and differentiated muscle cell line. Nucleic Acids Res., 6 (1979) 1639-1655. 12 J. A. Bantle, Complexity of mouse liver nucleax RNA, poly(A) hnRNA, and poly(A) mRNA. J, CellBioL, 83 (1979) abstract No. 2352. 13 T. Castle, A. Katz and A. Richardson, Comparison of RNA synthesis by liver nuclei from rats of various ages.Mech. Ageing Dev., 8 (1978) 383-395. 14 P. D. Colman, B. B. Kaplan, H. H. Osterburg and C. E° Finch, Brain poly(A) RNA during aging: stability of yield and sequence complexity in two rat strains. J. Neurochem., 34 (1980) 335-345.