Vol. 265, No. 29, Issue of October 15, pp. 17753-17758,199O Prmted in U.S.A.

Mouse Satellite DNA Is Transcribed in Senescent Cardiac Muscle* (Received for publication, February 13, 1990, and in revised form, May 8, 1990)

James W. Gaubatz$$ and Richard G. Cutler7 From the iDepartment of Biochemistry, College of Medicine, University of South Alabama, Mobile, Alabama 36688 and the IlCerontology Research Center, National Institute on Aging, Baltimore, Maryland 21224

Mouse satellite DNA consists of highly repetitive tandem sequences located in the centromeric hetero- chromatin. It is generally assumed that these simple sequences are not transcribed. We have analyzed total cellular RNA preparations from mouse liver, kidney, brain, and heart tissues at different ages for satellite transcripts. Using recombinant probes containing the major mouse satellite sequence, satellite transcripts were detected only in the heart RNA samples. These transcripts were not detected in the heart muscle of young adult animals (2 and 6 months), but then ap- peared at the age of 12 months and continued to in- crease over 2-fold up to the age of 32 months. The transcripts were resistant to DNase I and sensitive to RNases and alkaline treatment. Northern hybridiza- tion experiments showed a large and heterogeneous size range of satellite transcripts. Control studies using short-interspersed (Bl and B2) and long-interspersed (Ll and IAP) repetitive DNA sequence probes did not show a similar age-related pattern of transcription. These results indicate that satellite transcription does occur in mice but is highly tissue- and age-specific. The unique occurrence of satellite transcription only in adult and senescent heart tissue indicates age to be an important determinant of gene activity. An under- standing of the regulatory mechanisms involved could lead to new insights in the biological role of satellite DNA, gene derepression of reiterated DNA sequences, and the aging processes of cardiac muscle.

Satellite DNA consists of many copies of simple sequences that are repeated in tandem. This fraction of eucaryotic chromosomes usually contains related, but not identical, members of a short sequence family (John and Gabor-Miklos, 1979). It is thought that such simple sequences are not tran- scribed or translated. There are several lines of evidence that support this opinion. Satellite DNA has a very limited coding capacity, and promoter and other regulatory sequences have not been described for satellite DNA per se. Most satellite DNA exists as heterochromatin. Constitutive heterochroma- tin is most frequently localized to the procentric and telomeric regions of chromosomes (Yunis and Yasmineh, 1971). The DNA associated with constitutive heterochromatin is en- riched in highly repeated satellite sequences (Jones, 1970; Jones and Corneo, 1971). The relationship between satellite sequences and chromosomal location appears to be clear in

* This research was supported by Grant AGO7860 from the Na- tional Institute on Aging. The costs of publication of this article were defrayed in part by the payment 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.

I To whom correspondence and reprint requests should be ad- dressed.

the domestic mouse, Mus musculus. There is one major sat- ellite component which makes up 7-11% of the mouse genome (Waring and Britten, 1966; Prashad and Cutler, 1976). The satellite DNA is concentrated near the centromeres of all except the Y chromosome (Pardue and Gall, 1970). These regions are constitutively heterochromatic and C-band posi- tive (Jones, 1970; Pardue and Gall, 1970). The DNA in this region is composed of a 234-bp’ variable repeat that is repre- sented approximately one million times in the genome (Horz and Altenburger, 1981). In somatic cells, mouse satellite se- quences are heavily methylated (Ponzetto-Zimmerman and Wolgemuth, 1984) and are in a nonactive chromatin confor- mation (Sperling et al., 1985).

The tightly packaged nature of heterochromatin suggests that satellite DNA sequences are not available for transcrip- tion. In support of this concept, several studies have failed to detect cellular RNA complementary to satellite sequences in mammalian species (Hsu, 1962; Flamm et al., 1969; Reeder, 1973). However, there is controversial evidence reporting satellite DNA transcription in mouse tissues (Hare1 et al., 1968) and in mouse L cells (Cohen et al., 1973). Some exam- ples of satellite DNA transcription have been reported for nonmammalian species. For example, transcription of satel- lite DNA was detected on lampbrush chromosomes of newt oocytes. The histone gene clusters of the newt Notophthulmus are separated by long tracts of satellite DNA (Stephenson et al., 1981), and both strands of this satellite DNA were tran- scribed on the lampbrush loops (Diaz et al., 1981). The evi- dence suggests that transcription begins at a histone gene promoter but fails to terminate at the end of the gene and continues into the flanking satellite region, giving rise to very long transcripts (Diaz et al., 1981). The transcribed satellite is located at two noncentromeric sites and is in an extended chromatin structure (Diaz et al., 1981). Such read-through transcription may be common in Amphibia (Wu et al., 1986).

In this report we describe experiments detecting satellite DNA transcripts in RNA samples purified from normal whole mouse tissues using recombinant DNA hybridization tech- niques. Of four tissues examined, satellite DNA transcription was detected only in heart tissue of animals of an age older than previously used in studies investigating satellite DNA transcription. These results could lead to new insights on the biological role of satellite DNA, mechanisms of gene derepres- sion of reiterated DNA sequences, and the aging processes of cardiac muscle.

EXPERIMENTAL PROCEDURES

Mice-Inbred C57BL/6J male mice were obtained at 4 weeks of age from The Jackson Laboratories, Bar Harbor, ME. Mice were subsequently maintained in the Gerontology Research Center animal colony until used in this study. Mice in this colony have a mean life

’ The abbreviations used are: bp, base pair(s); kb, kilobase( SDS, sodium dodecyl sulfate.

17753

by guest on June 14, 2018http://w

ww

.jbc.org/D

ownloaded from

17754 Mouse Satellite DNA Transcripts

span of about 25 months and a maximum life span of approximately 40 months (Kunstyr and Leuenberger, 1975). The C57BL/6J strain is unusually long-lived for an inbred mouse strain, displays a low spontaneous tumor incidence, and is widely used for gerontological studies.

RNA Isolation and Treatment-Mice of normal weight for their age with no observable pathological abnormalities were killed by cervical dislocation between 10:00 and 11:OO a.m. For a single prepa- ration, RNA was isolated from the pooled tissues of six animals of the same age using the CsCl centrifugation method of Glisin et al. (1974) as modified by Chirgwin et al. (1979). All RNA preparations had a 260 nm/280 nm spectral ratio greater than or equal to 2.1. The integrity of high molecular weight RNA was determined by gel electrophoresis; a minimum of 2:l mass ratio of 28 S:18 S rRNA was considered a sign that no degradation had occurred. RNA was stored in diethylpyrocarbonate-treated, autoclaved water at -80 “C until used. Where indicated, 10 pg of total cellular RNA was digested with 44 units of RNase-free DNase I (Boehringer Mannheim) in 100 ~1 of 0.1 M sodium acetate, pH 5.0, 5 mM MgCl, at room temperature for 30 min. RNase-treatment of samples consisted of digesting 10 +g of RNA with 5 units of RNase A (Sigma) and 6 units of RNase Tl (Sigma) in 100 ~1 of 10 mM Tris-HCl, pH 7.4, 1 mM EDTA for 30 min at 37 “C. To show alkaline sensitivity, samples of total cellular RNA were treated with 0.5 N NaOH at 65 “C for 30 min.

Slot Blots and Northern Blots-Three volumes of 15% formalde- hyde-10 x SSC were added to 100 ~1 of solution containing 10 rg of total cellular RNA, and the samples were denatured by heating at 65 “C for 15 min. Denatured samples were then blotted onto Nytran membranes (Schleicher & Schuell) using a Minifold II slot-blotter. Membranes were air-dried, baked for 1 h at 80 “C, then stored in a desiccator. In addition to blotting RNA preparations from mouse tissues, the following positive and negative controls were typically included on each membrane: 510, and 15 rg of mouse rRNA, 5, 10, and 15 pg of yeast tRNA; and four to eight different amounts of EcoRI-restricted PSAT, pIAP, or pMR196 probes (see below). For quantitative purposes, 10 pg of mouse rRNA was added to recombi- nant plasmids to correct for possible nonspecific blocking in hybrid- ization experiments. However, there was no detectable difference between probes alone and the mixtures of probe plus rRNA. For Northern blots, RNA samples (15 rg) were fractionated by electro- phoresis on 1% agarose (SeaKern, GT grade)-2.2 M formaldehyde gels at 100 V for approximately 3 h (Man&is et al., 1982). In some gels, an RNA ladder (0.24-9.5 kb. Bethesda Research Laboratories) was run in parallel to estimate sizes. Gels were washed to remove form- aldehyde, stained with ethidium bromide, UV-transilluminated, and photographed, and then the RNAs were transferred to Nytran mem- branes according to Maniatis et al. (1982).

DNA Probes-The recombinant plasmids used in this study were pMR225 (Bl), pMR142 (B2), pMR134 (Ll), pIAP (Intracisternal A Particle), PSAT (satellite), and pMR196 (satellite). These recombi- nants have been described previously (Flores et al., 1988; Bennett et al., 1984; Butner and Lo, 1986). The plasmid sequences of pMR134, pMR196, pIAP, and PSAT were pBR322. The vector sequences of recomhinants pMR225 and pMR142 were pSP64 (Bennett et al., 1984). Insert sizes ranged from 100 to 250 bp except for pIAP. pIAP is the EcoRI fragment-A of clone 81 (A-81)subcloned into pBR322. A-81 contains 6.8 kb of the 7.3-kb full IAP element (Cole et aZ.. 1982).

Labeling and Hybridization-DNA probes were labeled using a nick translation system from Bethesda Research Laboratories, following the manufacturer’s protocol. [a-32P]dCTP (800 Ci/mmol) was ob- tained from Amersham. Specific activities ranged from 2 to 10 X 10’ cpm/pg. Membranes were prehybridized overnight at 42 “C in 50% formamide, 5 x SSPE, 0.1% SDS, 5 x Denhardt’s solution, and 100 fig/ml denatured salmon sperm DNA (Maniatis et al., 1982). Dena- tured probes were added to the prehybridization solutions at input ratios of 2-5 x lo6 cpm/ml, and hybridizations were carried out at 42 “C for 24 to 48 h. After hybridization, membranes were washed once in 1 x SSC, 0.1% SDS for 10 min at room temperature, once in 0.1 x SSC, 0.1% SDS for 10 min at room temperature, once in 1 X SSC, 0.5% SDS for 10 min at 65 “C, and once in 0.1 X SSC, 0.5% SDS for 10 min at 65 “C (high stringency conditions). In some experiments, the last wash was omitted (moderate stringency condi- tions). Wet membranes were covered with plastic wrap and exposed to x-ray film (Kodak RB) for various times at -70 ‘C. Before reprob- ing membranes, they were first stripped by placing the wet membrane in 500 ml of deionized. distilled water at 90 “C and allowing the bath to cool to room temperature while shaking. To quantitate transcripts in total RNA preparations from mouse tissues, autoradiographs were

scanned with a soft laser microdensitometer (Biomed Instruments, Inc.) within a linear film range. Peak areas were determined by computer-assisted integration, and probe binding was normalized to correct for differences (if any) in hybridization efficiency as reported previously (Flores et al., 1988). To determine the amount of probe bound for experimental samples, mass units were extrapolated from a standard curve of homologous recombinant on the same membrane.

RESULTS

Satellite DNA Transcripts in Total Cellular RNA Samples Isolated from Heart Tissue-In the course of performing age- related transcriptional studies, we learned that RNA isolated from mouse tissues by guanidine isothiocyanate extraction/ CsCl ultracentrifugation is frequently contaminated with DNA. Therefore, a DNase I digestion step was incorporated into the isolation procedure. Nevertheless, we decided to routinely check RNA preparations for DNA contamination by probing for satellite sequences. The rationale for using satellite DNA in the control experiments was based on the fact that this repetitive sequence family comprises approxi- mately 10% of the C5’7BL/6J mouse genome (Prashad and Cutler, 1976) and the assumption that satellite DNA is not transcribed. The probe used was PSAT which contains a 235- bp repeat unit of the major mouse satellite. The sequence of PSAT agrees with the satellite consensus sequence except for eight nucleotide mismatches and one nucleotide deletion (Butner and Lo, 1986).

RNA slot-blots of total cellular RNA isolated from brain, heart, kidney, and liver tissues of mice ranging in age from newborn to 32 months were hybridized to 32P-labeled PSAT. We consistently observed only background levels of binding to all RNA samples except those isolated from heart tissue. As shown in Fig. IA, sequences complementary to satellite DNA were not detected in heart RNA samples from young animals of 2 and 6 months of age but were detected in 12- month heart tissue and appeared at even higher levels in 24- and 32-month-old animals. Each RNA preparation was ob- tained from six pooled tissues for each age group. Two inde- pendent preparations for heart, three independent prepara- tions for brain and liver, and one preparation for kidney (not shown) were analyzed for satellite transcripts at each age. Complementary sequences were detected in only six samples: the two sets of RNAs from 12-, 24-, and 32-month heart (Fig. IA; HI and Hz). We also tested the possibility that the hybridization results were a unique property of the probe PSAT. This was done by hybridizing the same blots with pMR196, an independent recombinant probe containing mouse satellite sequences (Pietras et al., 1983). The results were essentially the same as those observed for PSAT (Fig. lA; H3).

Although these samples were digested with DNase I, it is possible that the heart samples contained a DNase inhibitor that increased as a function of age, and thus the positive hybridization signals were due to undigested satellite DNA. To determine the biochemical structure of the complementary sequences hybridizing to the probe, an untreated 32-month heart sample was analyzed. The results are shown in Fig. 1B. Digestion of the sample with DNase I slightly reduced the hybridization intensity. Digestion with RNases A and Tl greatly reduced the amount of probe binding. However, a small fraction of the sequences complementary to the probe were RNase-resistant. These results suggest that the sample as isolated contained both satellite RNA and DNA but RNA in greater proportion than DNA. The finding that contami- nating DNA was present in the preparation was not surprising because, as mentioned previously, we had already encountered this problem, and this was the basis for treating samples with

by guest on June 14, 2018http://w

ww

.jbc.org/D

ownloaded from

Mouse Satellite DNA Transcripts

A B Tissues 32 mo.

Age Bl Ll HI B2 L2 H2 H3 Heart -

NB - o -Untreated

2 mo.- - -RNase

6 mo.-

12 mo.-

24 mo.-

32 mo.-

-

-

M

* -DNase

- 0 - DNase+RNase

-0 - DNase+NaOH

FIG. 1. Slot-blot analysis of mouse satellite sequences in total cellular RNAs isolated from brain, liver, and heart tissues as a function of age. A, RNAs (lo-rg samples) prepared from six pooled brain (B), liver (L), or heart (H) tissues at each age, ranging from newborn (NB) to 32 months, were treated with DNase I, denatured with formaldehyde, then blotted onto nylon membranes. The membranes were hybridized to “‘P-labeled PSAT which contains a 235-bp sequence of the major mouse satellite (Butner and Lo, 1986). Results from two experiments utilizing two independent preparations (sets 1 and 2) are shown. Hybridization above background (10 pg of mouse rRNA or yeast tRNA) was observed for 12-, 24-, and 32-month heart preparations. H3 is set H2 hybridized to radiolabeled pMR196, another major mouse satellite recombinant (Pietras et al., 1983). B, mouse satellite sequences present in a 32-month heart RNA preparation isolated by guanidine isothiocyanate extraction/ CsCl ultracentrifugation (Chirgwin et al., 1979) with no subsequent treatment, treated with RNases A and Tl, treated with DNase I, treated with a combination of DNase I and RNases A and Tl, or treated with a combination of DNase I and 0.5 N NaOH (65 “C, 30 min). RNA samples (10 pg each) were denatured in formaldehyde and annlied to nvlon membranes. The immobilized RNAs were hybridized to “lP-labeled PSAT, as described under “Experimental Procedures.”

DNase I prior to blotting. To confirm that the DNase I- resistant sequences were RNA, DNase I-treated samples were subsequently treated with RNase or 0.5 N NaOH. Either treatment completely eliminated hybridization to the satellite DNA probe (Fig. 1B). Similar results were obtained with a 24-month heart preparation. From these results we conclude that old mouse heart tissue contained satellite DNA tran- scripts.

Transcriptional Patterns of Other Repetitive Sequence Fam- ilies in Mouse Tissues as a Function of Age-The occurrence of satellite DNA transcripts in senescent mouse heart might reflect a generalized increase in all types of transcripts. To compare levels of transcripts, slot-blots of total cellular RNA were analyzed for sequences complementary to short-inter- spersed (Bl and BZ), long-interspersed (Ll), and endogenous retroviral-like (IAP) repetitive sequence families of the mouse genome. Hybridization results for Bl and B2 are shown in Fig. 2, A and B, respectively. It is apparent that short- interspersed transcripts were highest in brain and lowest in heart. The level of Bl and B2 transcripts in heart RNA preparations was highest in newborn tissue and did not in- crease during aging. Ll transcripts were exceedingly scarce in all samples (probe binding only slightly higher than back- ground) and did not increase with age (data not shown). The transcriptional pattern observed with Intracisternal A Parti- cle (IAP) genes was similar to that shown for Bl and B2. However, IAP transcripts were less abundant than Bl/B2 sequences. IAP gene products did not increase as a function of age in liver, brain, kidney, or heart (see below).* These results indicate that a general increase of repetitive DNA transcription has not occurred in heart tissue during aging.

Size of Satellite DNA Transcripts-To determine the size of satellite DNA transcripts, total cellular RNA preparations from mouse heart tissue were fractionated by electrophoresis on 1% agarose-2.2 M formaldehyde gels and transferred to a nylon membrane. The resultant Northern blot was hybridized to PSAT, and the results are shown in Fig. 3A. A clear age- related increase in complementary sequences binding to the satellite probe was observed again. The satellite RNA was

’ Gaubatz, J. W., and Cutler, R. G., (1990) Mech. Ageirzg Deu., in press.

A B B L H 0 L H

NE-- - - - 0

2 mo.- - - - 6 mo.- 0 0

12 mo.- I_ W

24 mo.- 0 -

32 mo.- -

FIG. 2. Slot-blots showing levels of transcripts complemen- tary to short-interspersed, repetitive DNA sequence families of the mouse genome. A, hybridization of Bl sequences to total cellular RNA isolated from mouse brain (B), liver (L), and heart (H) tissues at different ages. Ten pg of each sample were treated with DNase I, denatured by heating in formaldehyde, and applied to nylon membranes. The immobilized RNAs were hybridized with “‘P-labeled pMR225, a recombinant plasmid containing short-interspersed Bl sequences of the mouse genome (Bennett et al., 1984). These RNA samples correspond to set 2 in Fig. 1. B, hybridization of B2 sequences to total cellular RNA isolated from mouse brain (B), liver (L), and heart (H) tissues. The RNA samples and the experimental conditions were the same as in A above. The labeled probe was pMR142 which contains the short-interspersed B2 DNA sequences (Bennett et al., 1984) that are 97% homologous to the consensus sequence. Bl and B2 represent different moderately repetitive sequence families of the mouse genome and thus do not cross-hybridize with each other (Krayev et al., 1980).

high molecular weight. No distinct bands were evident; rather a heterogeneous size range was noted. The region of the gel showing the most binding of satellite probe was in the range of 6.0 to 12.0 kb. In addition, there was a lesser amount of complementary transcripts that was higher molecular weight and appeared not to be separated in this gel system.

To compare the satellite results with other repetitive se- quence families, the Northern blot was reprobed individually with nick-translated Bl, B2, and IAP sequences. The Bl, B2, and IAP DNA sequence families comprise approximately l.l%, 0.7%, and 0.3% of the mouse genome, respectively (Flores et al., 1988). Therefore, the relatively high copy num- ber of these elements should also make them sensitive probes for RNA transcripts. The results for IAP gene transcripts are shown in Fig. 3B. In agreement with slot-blot experiments,

by guest on June 14, 2018http://w

ww

.jbc.org/D

ownloaded from

17756 Mouse Satellite DNA Transcripts

A B NB 2 6 12 24 32 NB 2 6 12 24 32

+> t# ‘2 b -9.5-

-75-

-44- *n*.nau *

-24-

-1.4-

-024-

FIG. 3. Electrophoretic analysis of satellite- and IAP-spe- cific transcripts in total cellular RNAs from mouse heart tissues, ranging in age from newborn (NB) to 32 months. A, autoradiograph of Northern blot probed for mouse satellite sequences. RNA samples were treated with DNase I, denatured by warming in formaldehyde and formamide, electrophoresed in 1.2% agarose gels containing 2.2 M formaldehyde, and transferred to a nylon membrane. The membrane was hvbridized with “P-labeled USAT. Each lane contained 15 pg of RNA. B, autoradiograph of Northern blot hybrid- ized with an IAP probe. The membrane used for satellite hybridiza- tion shown in A was stripped and reprobed with [“‘P]pIAP as de- scribed in the text. The size makers shown were derived from parallel elect.rophoreses of an RNA ladder.

newborn heart tissue exhibited the greatest level of IAP transcripts. One major RNA species of 5.4 kb was present in all heart samples. This transcript has been described previ- ously in mouse tissues and tumor cells (Kuff and Fewell, 1985). There was no significant change in the level of the 5.4- kb RNA from 2 to 32 months of age. In addition, no high molecular weight smear corresponding to that shown for satellite transcripts was found. Bl and B2 sequences are transcribed as part of the large heterogeneous nuclear RNA (Krayev et al., 1980). Bl is also transcribed as a small llO- 130-nucleotide RNA, and B2 is synthesized as a MO-300-long nucleotide RNA (Kramerov et al., 1985). In addition to the small Bl/B2 RNAs, we observed a broad size range of heart transcripts hybridizing to Bl and B2 probes (data not shown). However, the predominant lengths of the transcripts contain- ing the Bl/B2 sequences were smaller than the large satellite transcripts. Thus, the spectrum of transcripts hybridizing to short-interspersed and long-interspersed (IAP) sequences is different in size from satellite RNA. Although the results do not exclude low levels of interspersed sequences from being associated with satellite DNA transcription, it seems reason- able to conclude that Bl/B2 and IAP sequences are not abundantly represented in satellite transcripts.

Quantitation of Satellite DNA Transcripts-In the foregoing analysis, the stringency of the hybridization conditions and the subsequent washes of the membranes had not been opti- mized for satellite sequences because in fact, we had not anticipated measuring such transcripts. The A + T content of mouse satellite is 63% (Horz and Altenburger, 1981), and the calculated T, of satellite DNA at a monovalent cation concentration of 0.2 M is 84 “C (Lewin, 1980), which is in agreement with measured melting transitions for native mouse satellite monomers and dimers (Southern, 1975). The last, high stringency wash used in preceding experiments was 0.1 X SSC, 0.5% SDS at 65 “C. This wash was indeed stringent since we estimate a T,, of 70 “C for native satellite DNA at this ionic strength. Such high stringency conditions should be all the more telling because reassociated satellite sequences will have a T,,, some 5 “C lower than native complexes due to mismatched base pairs (Southern, 1975). Therefore, to meas- ure more accurately the level of satellite transcripts in mouse

cells, the experiments were repeated using a final wash of 1 X SSC, 0.5% SDS, which is about 12 “C below the T, at this sodium concentration. Furthermore, hybridizations were car- ried out at 39 “C (instead of 42 “C) in 5 x SSC, 50% formamide to optimize the reassociation rate for these sequences (Birn- stiel et al., 1972).

Typical results obtained for the two washes (0.1 X SSC uersus 1 x SSC) are compared in Fig. 4A, and extrapolated mass values for satellite transcripts are plotted as a function of age in Fig. 4B. The level of satellite transcripts was 50- 60% higher measured under the less stringent conditions. At 32 months of age, mouse heart tissue contained approximately 35 pg of satellite RNA per 10 gg of total cellular RNA. If one assumes that l-2% of the total heart RNA is mRNA, satellite RNA would represent 0.02-0.04% of the mRNA fraction. Thus, satellite transcripts are rare. In contrast to satellite sequences, IAP gene transcription appears to be tightly reg- ulated in heart cells throughout the life span of this mouse strain. As shown in Fig. 4B, the level of IAP transcripts did not vary past 2 months of age, and, in the oldest age group, IAP transcripts were one-fifth as abundant as satellite tran- scripts.

DISCUSSION

Satellite DNA Transcription in Mouse Cells-The results from this study indicate that satellite DNA is transcribed in uiuo in a normal tissue of a mammalian species, Mus musculus. However, this transcription activity is highly tissue- and age- specific, being found only in heart tissue from old animals. Finding satellite DNA transcripts was unexpected because of the established view that centromeric and heterochromatic sequences are not transcribed in uiuo in mammalian cells (Lewin, 1987). Previous studies, however, have indicated con-

A B

ssc 1x 0.1x

NB+

2 ma*

6mo.+

12 ma+

24 mo.+

32 ma-

- --- --

6 12 18 24 30 Age (Months)

FIG. 4. Age-related quantitation of mouse heart satellite transcripts employing two stringency conditions and compar- ison to IAP transcripts. A, slot-blot analysis of total cellular RNA isolated from heart tissues at various ages. Treatment of samples was the same as that described in Fig. 1 and text, except that the final wash was either 1 X SSC, 0.1% SDS for 10 min at 65 “C or 0.1 X SSC, 0.5% SDS for 10 min at 65 “C. Each slot contained 10 pg of RNA. The recombinant pMR196 containing the major mouse satellite (Pietras et al., 1983) was used as the probe. The autoradiographic exposures were 24 h for both 1 x and 0.1 X SSC membranes. B, estimated amounts of satellite and IAP transcripts in mouse heart cells as a function of age. Data were obtained from microdensitometric scanning of slot blots as shown in A. Mass values were determined by extrapolation from standard curves of hybridization to various dilutions of homologous probes, PSAT, pMR196, or pIAP, on the same membrane. The results are expressed as picograms of transcript per 10 pg of total RNA (mean and S.E. of three determinations). Satellite transcripts determined with moderate stringency, 1 x SSC conditions, 0; satellite transcripts determined with the high strin- gency, 0.1 x SSC conditions, e, IAP transcripts determined with high stringency, 0.1 X SSC conditions, A.

by guest on June 14, 2018http://w

ww

.jbc.org/D

ownloaded from

Mouse Satellite DNA Transcripts 17757

tradictory results. Two studies have reported that mouse satellite sequences are inaccessible to Escherichiu coli RNA polymerase in vitro (Reeder, 1973; Gjerset and McCarthy, 1977), and Flamm et al. (1969) presented data that less than 1 in 60,000 RNA molecules from liver, spleen, and kidney cells was complementary to mouse satellite DNA. Alterna- tively, Hare1 et al. (1968) and Cohen et al. (1973) obtained results that suggested mouse satellite DNA, at least in mouse L cells, is transcribed in viuo. Some of the discrepancy might stem from the methods of analysis employed in these earlier studies. Detection of complementary sequences in these stud- ies were limited by the relatively low specific activities of the isotopically labeled nucleic acids, and, to compensate, vast excess of RNA was used in some cases. Thus, high sensitivity and specificity were reduced.

The probes used in this report, PSAT and pMR196, corre- spond closely to the consensus sequence of the major mouse satellite repeat (Horz and Altenburger, 1981; Pietras et al., 1983; Butner and Lo, 1986). It appears that the major satellite DNA is restricted to the centromeric regions of all mouse chromosomes except the Y chromosome (Pardue and Gall, 1970). These sequences have been characterized as hetero- chromatic DNA in metaphase and interphase mouse cells (Jones, 1970; Pardue and Gall, 1970). Pietras et al. (1983) have described a minor M. rn~~.~ulus satellite with a 130- nucleotide repeat length. This minor satellite has a repetitive frequency 510% of the major satellite, is located at the centromeres, and is physically closer to the kinetochore of mouse chromsomes than is the major satellite (Joseph et al., 1989). However, the minor satellite contains only one stretch of 23 out of 29 nucleotides that shows extensive sequence homology to the major mouse satellite (Pietras et al., 1983). It is doubtful that the RNA transcripts measured in the present experiments correspond to this minor satellite owing to the high stringency conditions used in the washes (Fig. 4 and “Results”). In the absence of data that show the major mouse satellite repeat is present in genes or gene clusters, it seems reasonable to conclude that the satellite transcripts observed by us have arisen from the centromeric, heterochro- matic satellite sequences. However, detailed sequence studies will need to be performed to clarify their origin.

The results presented in this report are consistent with the view that mouse satellite DNA is not widely or actively transcribed. We found no satellite DNA transcripts in mouse brain, liver, or kidney tissue at any age (Fig. 1 and data not shown for kidney) nor from embryonic brain and liver tissues. No satellite DNA transcription was detected in the young heart tissues, and the concentration of the satellite DNA transcripts detected in old heart tissue was not high (Fig. 4).

Satellite DNA Transcripts and Aging-Since satellite DNA sequences are expressed in older animals and do not appear to have promoters or to code for functional protein products, it is difficult to assign a functional role for the RNA tran- scripts. Thus, it appears most likely that satellite DNA tran- scripts represent a gene regulatory dysfunction, possibly con- tributing to the aging process of cardiac muscle cells (Cutler, 1985; Wilson et al., 1987). There is little information to support this concept, but previous results have shown that satellite transcripts are present in uiuo in mouse L cells (Hare1 et al., 1968; Cohen et al., 1973), which do represent an abnor- mal state of differentiation. In addition, the observation that a highly repetitive DNA sequence, located primarily at the centromeres and telomeres of rat chromosomes, is transcribed in hepatoma tissue culture cells (Sealy et al., 1981) might also be a case of improper transcription of heterochromatic simple sequences. The simple sequence transcripts were present in a

heterogeneous and high molecular weight population of hep- atoma tissue culture nuclear RNA molecules. Satellite tran- scripts found in old heart tissue are also heterogeneous and high molecular weight (Fig. 3). Large heterogeneous RNA may be a general property of satellite DNA transcription as indicated by the read-through transcription of such sequences in Amphibia (Diaz et al., 1981; Wu et al., 1986). However, satellite DNA transcription in Amphibia appears to be re- stricted to simple sequence DNA that is interspersed in gene clusters (Stephenson et al., 1981). An analogous situation is not known for mouse satellite DNA.

There is also some evidence of an accumulation of DNA damage with age in heart tissue. Earlier work has suggested that heart DNA in this mouse strain accumulates single strand breaks as a consequence of aging (Price et al., 1971). Other studies have indicated that the steady state level of alkyl and aromatic DNA adducts in heart cells increased dramatically, perhaps exponentially, in the latter part of the mouse’s life span (Gaubatz, 1986, 1989), and alkyl adducts can lead to single strand breaks. The highly supercoiled nature of nuclear chromatin depends upon the continuous integrity of the duplex DNA (Worcel and Benyajati, 1977). An age- dependent accumulation of single strand breaks in heart DNA could be responsible for localized uncoiling of chromatin, thereby making satellite sequences more accessible to RNA polymerase read-through. The age-related demethylation of mouse satellite DNA described by others (Howlett et al., 1989) might also facilitate transcription of these sequences. There are also other data indicating chromatin alterations and im- proper gene expression occurring with increasing age in mice (Cutler, 1985; Wilson et al., 1987; Semsei et al., 1989).

Genetic Regulation of Mouse Satellite DNA Transcription- The tightly coiled physical state of heterochromatin would appear to limit the synthetic activity of satellite DNA. In fact, genes that are brought into close proximity to heterochro- matic satellite sequences by translocation or inversion or experimental manipulation frequently show suppressed tran- scriptional activities (Baker, 1968, Talarico et al., 1988). Thus, any mechanism proposed to explain satellite transcript in senescent heart cells would have to address both the physi- cochemical structure of satellite DNA sequences as well as the tissue-specific and age-related occurrence of the tran- scripts. In the simple analysis, one could say that some process has induced chromatin uncoiling and satellite transcripts then arose through nonspecific initiation. The results that show short-interspersed (Bl and B2), long-interspersed (Ll), and IAP sequences did not increase during aging (Figs. 2 and 3) argue against a global derepression of heart DNA. It is possible that genomic rearrangements, such as nonhomologous recom- bination or transposition, have led to the linking of promoter sequences with satellite DNA. Satellite DNA has the property of being very unstable (Maresca et al., 1984), and mouse centromeric satellite DNA is associated with a high frequency of DNA rearrangements (Butner and Lo, 1986), but it is not clear why this instability should be uniquely associated with heart cells and increase with age.

We currently favor a mechanism based on satellite tran- scription in the newt (Diaz et al., 1981). In our proposal, mouse satellite DNA transcription is a result of an age- dependent alteration of a gene or gene cluster that flanks the centromeric DNA to terminate transcription. A higher rate of alteration of this gene cluster in heart tissue could perhaps be related to the higher metabolic rate and postmitotic nature of heart cells. Alteration in the flanking gene cluster would result in termination failure of adjacent non-satellite genes. The major satellite is indeed adjacent to the euchromatic long

by guest on June 14, 2018http://w

ww

.jbc.org/D

ownloaded from

17758 Mouse Satellite DNA Transcripts

arm of each mouse chromosome (Joseph et al., 1989). The mechanism proposed above makes several predictions,

among them: 1) additional non-satellite sequences will be associated with the transcribed satellite sequences, 2) the non-satellite sequences will be expressed in heart cells, but not other cell types, and 3) damaging agents that induce single strand nicks in DNA or other alterations will lead to an increased transcription of satellite sequences in mouse heart. There are, of course, other mechanisms which might explain mouse satellite transcription. Whatever mechanism is respon- sible, the important implication of the results reported here is that a tissue-specific genetic alteration has occurred only in old cells leading to the expression of a DNA sequence normally repressed. Even though the function of satellite DNA transcripts is not known, and indeed it may have no function, an age-dependent genetic alteration allowing satel- lite DNA transcription is clearly indicated from these studies.

Acknowledgments-We thank Edith Cutler (supported by the Paul Glenn Foundation for Medical Research) for the preparation of RNA samples, Sonia Flores for the preparation of repetitive sequence probes, and Brian Arcement for technical assistance. We also thank K. L. Bennett and N. D. Hastie for providing the cloned Bl, B2, and Ll repetitive sequences, R. C. Huang for supplying the IAP A-81 clone. and C. Lo for the gift of the DSAT recombinant. The critical comments of N. Flodin,-G. Roth,‘N. Holbrook, and G. Daniels, regarding the manuscript are appreciated. The Gerontology Research Center is fully accredited by the American Association for the Accreditation of Laboratory Animal Care.

REFERENCES Baker, W. K. (1968) Adu. Genet. 14,133-169 Bennett, K. L., Hill, R. E., Pietras, D. F., Woodworth-Gutai, M.,

Kane-Hass, C., Houston, J. M., Heath, J. K., and Hastie, N. D. (1984) Mel Cell. Biol. 4, 1561-1571

Birnsteil, M. L., Sells, B. H., and Purdom, I. F. (1972) J. Mol. Biol. 63,21-39

Butner, K. A., and Lo, C. W. (1986) J. Mol. Biol. 187, 547-556 Chirgwin, J. M., Przybyla, A. E., MacDonald, R. J., and Rutter, W.

J. (1979) Biochemistry 18, 5294-5299 Cohen, A. K.. Huh, T. Y., and Helleiner, C. W. (1973) Can. J.

Biochem. 5 i ,529:532 Cole. M. D.. Ono. M.. and Huana. R. C. (1982) J. Virol. 42. 123-130 Cutler, R. d. (1985) in Molecular%ology‘of Aging: Gene Stability and

Gene Expression (Sohal, R. S., Birnbaum, L., and Cutler, R. G., eds) pp. 307-340, Raven Press, New York

Diaz, M. O., Barsacchi-Pilone, G., Mahon, K. A., and Gall, J. G. (1981) Cell 24,649-659

Flamm, W. G., Walker, P. M. B., and McCallum, M. (1969) J. Mol. Biol. 40, 423-443

Flores, S. C., Sunnerhagen, P., Moore, T. K., and Gaubatz, J. W. (1988) Nucleic Acids Res. 16, 3889-3906

Gaubatz, J. W. (1986) J. Mol. Cell. Cardiol. 18, 1317-1320 Gaubatz, J. W. (1989) Arch. Gerontol. Geriatr. 8, 47-54

Gjerset, R. A., and McCarthy, B. J. (1977) Proc. Natl. Acad. Sci. U. S. A. 74, 4337-4340

Glisin, V., Crkvenjakov, R., and Byus, C. (1974) Biochemistry 13, 2633-2637

Harel, J., Hanania, N., Tapiero, H., and Harel, L. (1968) Biochem. Biophys. Res. Commun. 33,696-701

Horz. W.. and Altenbureer. W. (1981) Nucleic Acids Res. 9.683-696 HowIett, D., Dalrymple, S:, and MayaHoopes, L. L. (1989) Mutat.

Res. 219,101-106 Hsu, T. C. (1962) Exp. Cell Res. 27, 332-334 John, B., and Gabor-Miklos, G. L. (1979) Znt. Reu. Cytol. 58, l-114 Jones, K. W. (1970) Nature 225, 912-915 Jones, K. W., and Corneo, G. (1971) Nature New Biol. 233,268-271 Joseph, A., Mitchell, A. R., and Miller, 0. J. (1989) Exp. Cell Res.

183,494-500 Kramerov, D. A., Tillib, S. V., Lekakh, I. V., Ryskov, A. P., and

Georgiev, G. P. (1985) Biochim. Biophys. Acta 824,85-98 Krayev, A. S., Kramerov, D. A., Skryabin, K. G., Ryskov, A. P.,

Bayev, A. A., and Georgiev, G. P. (1980) Nucleic Acids Res. 8, 1201-1215

Kuff, E. L., and Fewell, J. W. (1985) Mol. Cell. Biol. 5, 474-483 Kunstyr, I., and Leuenberger, H. G. (1975) J. Gerontol. 30, 157-162 Lewin, B. (1980) Gene Expression 2: Eucaryotic Chromosomes, pp.

504-508,2nd Ed, John Wiley & Sons, New York Lewin, B. (1987) Genes ZZI, pp. 432-434, John Wiley & Sons, New

York Maniatis, T., Fritsch, E. F., and Sambrook, J. (1982) Molecular

Cloning: A Laboratory Manual, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY

Maresca, A., Singer, M. F., and Lee, T. N. H. (1984) J. Mol. Biol. 179,629-649

Pardue, M. L., and Gall, J. G. (1970) Science 168,1356-1358 Pietras, D. F., Bennet, K. L., Siracusa, L. D., Woodworth-Gutai, M.,

Chapman, V. M., Gross, K. W., Kane-Hass, C., and Hastie, N. D. (1983) Nucleic Acids Res. 11, 6965-6983

Ponzetto-Zimmerman. C., and Wolaemuth. D. J. (1984) Nucleic Acids Res. 12,2807-2822' -

Prashad, N., and Cutler, R. G. (1976) Biochim. Biophys. Acta 418, l-23

Price, G. B., Modak, S. P., and Makinodan, T. (1971) Science 171, 917-920

Reeder, R. H. (1973) J. Mol. Biol. 80, 229-241 Sealy, L., Hartley, J., Donelson, J., Chalkley, R., Hutchinson, N., and

Hamkalo, B. (1981) J. Mol. Biol. 145, 291-318 Semsei, I., Ma, S., and Cutler, R. G. (1989) Oncogene 4,165-470 Southern, E. M. (1975) J. Mol. Biol. 94,51-69 Sperling, ‘K., Kerem, B. S., Goitein, R., Kottusch, V., Cedar, H., and

Marcus. M. (1985) Chromosoma 93.38-42 Stephenson, E: C., Erba, H. P., and Gall, J. G. (1981) Cell 24, 639-

647 Talarico, D., Peverali, A. F., Ginelli, E., Meneveri, R., Mondello, C.,

and Valle, G. D. (1988) Mol. Cell. Biol. 8, 1336-1344 Waring, M., and Britten, R. J. (1966) Science 154, 791-794 Wilson, V. L., Smith, R. A., Ma, S., and Cutler, R. G. (1987) J. Biol.

Chem. 262,9948-9951 Worcel, A., and Benyajati, C. (1977) Cell 12,83-100 Wu, Z., Murphy, C., and Gall, J. G. (1986) Chromosoma 93, 291-297 Yunis, J. J., and Yasmineh, W. G. (1971) Science 174, 1200-1209

by guest on June 14, 2018http://w

ww

.jbc.org/D

ownloaded from

J W Gaubatz and R G CutlerMouse satellite DNA is transcribed in senescent cardiac muscle.

1990, 265:17753-17758.J. Biol. Chem. 

  http://www.jbc.org/content/265/29/17753Access the most updated version of this article at

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

to choose from all of JBC's e-mail alertsClick here

  http://www.jbc.org/content/265/29/17753.full.html#ref-list-1

This article cites 0 references, 0 of which can be accessed free at

by guest on June 14, 2018http://w

ww

.jbc.org/D

ownloaded from