Methylation and Rearrangement of Mouse Intracisternal A - Europe PMC

Vol. 3, No. 8

MOLECULAR AND CELLULAR BIOLOGY, Aug. 1983, p. 1371-1380 0270-7306/83/081371-10$02.00/0 Copyright © 1983, American Society for Microbiology

Methylation and Rearrangement of Mouse Intracisternal A Particle Genes in Development, Aging, and Myeloma LAURA L. MAYS-HOOPES,1t ANNE BROWN,2 AND RU CHIH C. HUANG2* Laboratory of Cellular and Molecular Biology, Gerontological Research Center, National Institute of Aging, Baltimore, Maryland 21224,1 and Department of Biology, The Johns Hopkins University, Baltimore, Maryland 212182

Received 6 December 1982/Accepted 5 May 1983

Sequences of DNA that hybridize on Southern blots with cloned intracisternal A-particle (IAP) sequences have been examined in genomic DNAs of neonatal mice, livers of adult mice (3, 6, 12, 18, 24, and 26 months old), and the solid myeloma tumor MOPC-315. The isoschizomers HpaII (CCGG or mCCGG) and MspI (CCGG or C'mCGG) were used to assess methylation. All the DNAs produced a major 0.5-kilobase MspI fragment that hybridizes with IAP probe. Only the myeloma DNA, and to a much lesser degree DNA from senescent mouse liver, produced this fragment in HpaII digest; the other DNAs all had IAP sequences resistant to HpaII digestion. These sequences thus become fully methylated to CmCGG early and remain so in adult life, except in the myeloma cells that are expressing the IAP genes. An increase in MspI-sensitive sites in IAP gene-containing DNA was observed in aging mice. The probe used to assess methylation, a 0.8-kilobase fragment produced by BamHI-HindIII double digestion, is common to several cloned IAP genes and is part of a region of DNA which is conserved in genomes of all mouse tissues. The probe hybridized to 1.5- and 1.4-kilobase doublet bands produced by BamHI, HindIII, and EcoRI triple digestions of neonatal DNA. These two bands were found in neonatal livers of Swiss Webster, BALB/c, and C57BL/6J mouse strains, showed less in adult liver, and were barely detectable in senescent livers from C57BL/6J mice. The mouse intracisternal A-particle (IAP) genes are a family of moderately repetitive sequences of about 1,000 copies per haploid genome (19-21, 25). These genes are expressed in early development up to the eight-cell stage, and then they are not typically transcribed further in normal tissues (3, 4, 44; K. Wujcik, dissertation, Johns Hopkins University, Baltimore, Md., 1982). Myeloma tumors have also been shown to express IAP genes (27). When the genes are transcribed, characteristic virus-like inclusions called A particles appear within the cisternae of the endoplasmic reticulum (21). The genes are a part of the genome of all known mouse strains, although there is some justification for viewing them as defective retroviruses (7, 19, 21, 33). Because they are expressed in early development and in a specific type cancer, the regulation of these genes is of particular interest. DNA methylation is correlated with transcriptional silence of several eucaryotic genes (12, 17, 29, 34, 39, 42), including immunoglobulin t Present address: Department of Biology, Occidental Col-

lege, Los Angeles, CA 90041.

1371

genes (8), ovalbumin genes (22), human globin genes (40), thymidine kinase genes (6), and genes of endogenous retroviruses (1, 12, 28, 39). Methylation appears to regulate gene transcription via controls over DNA-protein interactions in chromatin (11, 43). We tested the IAP genes to discover whether they were less methylated in cases where they are being expressed. The isoschizomers HpaII (CCGG or mCCGG) and MspI (CCGG or CmCGG) were used to assess methylation (24). Another mechanism that may control eucaryotic gene expression is altering the arrangement of sequences of DNA. Rearrangements can include the deletion of DNA sequences, as seen in immunoglobulin-producing cells between neonatal and adult life (10, 23). Rearrangements can also include the insertion of promoter sequences (26, 28) or of expression inhibitors as recently shown for IAP-related sequences (15). We examined the possibility of rearrangement, using restriction enzymes to produce known fragment sizes from adult DNAs that would hybridize with IAP probes. These sizes were compared with sizes obtained from neonatal, myeloma, senescent adult liver, and sperm DNAs.

1372

MAYS-HOOPES, BROWN, AND HUANG

MATERIALS AND METHODS Animals. Mice for sperm, liver, and neonatal DNA preparations were of BALB/c, Swiss Webster, and C57BL/6J cell lines, as noted in the figure legends. MOPC-315 solid myeloma tumors were propagated subcutaneously from a sample obtained through Litton Bionetics, Inc. (Rockville, Md.) from H. Eisen, MIT, Cambridge, Mass. Postpubertal mice (3 months old) were of the Swiss Webster strain. Adult mice (6, 12, 18, 24, and 26 months old) were C57BL/6J male mice from the colony at National Institute of Aging. These mice were tested for freedom from a wide variety of viruses and had measurable levels of antibody to only one: mouse pneumonia virus. No animals with gross pathological lesions were used. The mean lifespan of the male C57BL/6J mouse is 24 ± 2 months (41). Preparation of germ line DNA. Mouse sperm samples were isolated and purified according to a modification of the method of Calvin (5). Whole spermatozoa from six adult male Swiss Webster mice were squeezed from minced testis and epididymus into 10 mM Tris-hydrochloride (pH 7.5). The solution was filtered through 200-mesh nylon, and the filtrate was centrifuged for 10 min at 17,000 x g. The pellet was sonicated in 10 mM Tris to lyse any somatic cells and separate sperm heads from tails, brought to 1.8 M sucrose, and layered over 2 M sucrose in 10 mM Trishydrochloride (pH 7.5). The solution was centrifuged at 40,000 x g for 1 h. The pelleted sperm heads were then frozen before DNA extraction. The DNA was purified from the sperm heads by the following procedure. Sperm heads were resuspended in 10 ml of 1.1 M NaCI-0.1 M ,3-mercaptoethanol-0.15 M Tris-hydrochloride (pH 8.0)i.0% sodium dodecyl sulfate (SDS). The mixture was incubated at 37°C for 10 min, and no intact nuclei were visible by light microscopy. An equal volume of chloroform-isoamyl alcohol (25:1) was added, and the mixture was mixed at room temperature on a rotating wheel for 30 min. The aqueous layer was removed after centrifugation and to it was added 2.5 volumes of 95% ethanol. The DNA was then spooled out at the interface. DNA isolation and restriction. DNA was isolated from other samples by dry ice grinding of -70°C frozen livers or other tissue followed by extraction of the DNA (2). Pools of livers from six healthy male C57BL/6J animals at each age were used to prepare DNAs. In the cases of mice 3, 6, 18, and 24 months old, results were confirmed with a second such DNA preparation. The DNA was treated with proteinase K and with Worthington DNase-free RNase during isolation. Diphenylamine reactions agreed with the calculations of concentration based upon absorbance. Conditions for the restriction of DNA were specified separately as described in the figure legends. DNA restriction fragments were brought to 0.1% SDS and 4 mM disodium EDTA and heated at 52°C for 30 min to dissociate restriction nucleases. Electrophoresis and blotting. Samples were electrophoresed on 20-cm-long 1.5% agarose gels (Seakem or Bethesda Research Laboratories) in 40 mM Trisacetate buffer with 10 mM EDTA (pH 7.2) overnight at 25 V, 10 mA. Gels were stained with 1 ,ug of ethidium bromide per ml for 1 h and photographed. DNAs were denatured, and the gels were blotted onto

MOL. CELL. BIOL.

nitrocellulose sheets (Schleicher & Schuell Co.) according to Southern (38). In all cases, results were typical of at least two blots of separate endonuclease restrictions. Preparation of labeled probes. A 0.8-kilobase (kb) BamHI-HindIII DNA fragment (IAP sequences, region III) was isolated from the 81A plasmid inserts as previously described (2). The IAP probes were labeled via nick-translation with a-32P-nucleotides (30) to a specific activity of at least 107 cpm/,Lg and recovered via DEAE cellulose elution followed by overnight dialysis against 0.15 M NaCl. Hybridization and washing of blots. Blots were prehybridized for 2 h at 65°C in Denhardt's solution (9) plus 100 ,ug of yeast tRNA per ml and hybridized for 24 h or longer as specified in the same conditions but with 5 x 105 cpm/ml of the denatured, nick-translated probe. Blots were washed twice at 65°C with 3 x SSC (1 x SSC is 0.15 M NaCl plus 0.015 M sodium citrate)0.1% SDS, dried, and autoradiographed at -70°C with intensifying screens. The blot shown in Fig. 4 was washed a third time in 3x SSC-0.1% SDS and once with 0.1 x SSC-0.1% SDS at 65°C before autoradiography.

RESULTS Sequence homology of cloned IAP genes and specific features of IAP region III probe. There are approximately 1,000 IAP genes per haploid genome in Mus musculus. Although not exactly identical, sequences of these genes share extensive regional homology. We have isolated several of these IAP genes from a DNA bank from neonatal mice and have characterized seven of them by restriction enzyme analysis and by heteroduplex mapping (27). The unique features of the varied structural organizations of these genes have been illustrated diagrammatically, and restriction maps of IAP clone 81 are also included for clarification (Fig. 1). As indicated in Fig. 1, although region III sequences appear to be well conserved, region I and region II sequences and sequences at the junction of region II and region III are quite variable among members of the IAP gene family. For instance, the BamHI site at the region II-III boundary is common to clones 14, 19, 17, 81, and 71 but is missing in clones 62 and 20. In addition, most of the region II sequences in clones 62 and 20 are deleted. In all of the clones, however, the entire region III appears to be present. Because the variations in our IAP clones may indeed reflect genomic differences in mouse IAP genes, the 0.8-kb BamHI-HindIII DNA fragment of IAP 81 was selected as the probe for study. This fragment consists primarily of the conserved region III sequence and should therefore detect all classes of IAP genomic sequences. With this probe, IAP gene modification and rearrangements were detected in senescing mice and in mouse tumors. Partial demethylation of "C'CmGG" and

METHYLATION AND REARRANGEMENT OF MOUSE GENES

VOL. 3, 1983

1373

A I

+--

19

I +1

=90mmu i

14

m

H

0

LLLJ-..

:1

L

62 Eco RI

17 RI

-IMU

IA

t

I +1

I

ti

I

Eco R I

I

I

_.

20

M.... t

I

______J=

=

71

qIl

I

iL t

tl 4

r

I kb

B 81 (1) (2) A.

A

'4-

0

\s

I

a

T

T

:0 1

a

9

,j::

D

I

L

0.80

(3)

3'5'

FIG. 1. Restriction and heteroduplex analysis of cloned IAP genes. (A) Sequence organization of cloned mouse DNA fragments containing sequences complementary to IAP RNA. Homologous sequences were determined by electron microscopy and are indicated by areas of similar shading. Regions I, II, and III are indicated at the top of the diagram, as described in the text. Regions I and II differ among the clones, whereas region III is homologous to all members of the gene family. Boundaries differing less than the standard deviation of measurement from each pair of subclones were assumed to be at the same position. Symbols: -, sequences that are absent from the gene; a sequence that is not homologous to any of the other subclones; (at the end of each fraginent), the vector DNA (0) BamHI sites; t HindII sites. (See references 7 and 27 for detailed information.) The sequences to the left of the EcoRI sites in 17 and 81 exist in the corresponding lambda clones only. (B) Restriction map of lAP clone X81 (reference 7). Line 1 indicates the sizes in kb pairs of the EcoRI restriction fragments found in the mouse DNA contained in clone 81. Line 2 shows a more detailed restriction map of the mouse DNA in bacteriophage Charon 4A. Symbols: (at the ends of the map), the arms; I MspI; *, BamHI; 0, PstI; t, HindIII; [ ], SstI; |, EcoRI; *, the probe used in the experiments. Line 3 shows the complete IAP gene in this clone. LTR regions; I (in line 2), an MspI-HpaII site within the region from which the probe (underlined area) was derived. The other MspI-HpaII sites in the clone are not indicated on this figure. ---

-,

=

,

,

0],

"C31CGG" sequences within region III of IAP ini myeloma. A correlation between DNA methylation and transcriptional inactivity has been made in the case of several eucaryotic

genes

genes (17, genes are

29, 42). We reasoned that since IAP expressed in myeloma but not in tissues such as liver, kidney, and spleen (44), these genes might be methylated in normal tis-

1374


sue DNA and become partially demethylated in transformed cells. To approach this question, we first determined the MspI-HpaII sites within the 0.8-kb BamHI-HindIII fragment from the unmethylated plasmid DNA of clone 81. Two smaller fragments of 500 and 300 base pairs were generated from the 0.8-kb fragment upon MspI or HpaII digestion, and the mapped site in clone 81 is shown in Fig. 1B (downward arrow). If these MspI-HpaII sites exist in genomic IAP genes then they might be subject to modification by cellular methyltransferases. We wished, therefore, to determine the sensitivity of these sites in genomic DNA to MspI-HpaII digestion. Mouse myeloma or liver DNAs were predigested with BamHI, HindIII, and EcoRI enzymes and were further restricted either by MspI, HpaII, or no endonuclease. A Southern blot of digested DNA was hybridized with a 32p_ labeled 0.8-kb fragment of 81 (Fig. 1B). The result of this experiment is shown in Fig. 2. When liver and myeloma DNA were digested with BamHI, HindIII, and EcoRI, the restriction fragments which hybridized to the 0.8-kb IAP probe were identical (Fig. 2c). The heaviest band was found at 0.8 kb; this indicates that a substantial number of genomic IAP sequences contain the same BamHI-HindIII restriction sites found in several of our clones. In addition to the 0.8-kb band, the probe also hybridized to a 0.75-kb fragment from the mouse genome. The hybridization pattern produced by digestion with the three enzymes plus MspI is shown in Fig. 2b. In this case, the probe hybridized to a series of DNA fragments. In addition to the fragments of 0.8 and 0.75 kb, fragments of 3.5, 3.0, and 0.5 kb produced a signal. The pattern obtained with myeloma DNA is different from liver in that very little hybridization occurred at sizes greater than 2.0 kb, and there was an increase in the amount of 0.5-kb fragment compared with that found in the liver DNA. This increase may result from MspI digestion of 3.5and 3.0-kb sequences seen in liver if these sequences are present but undermethylated in myeloma DNA. It is also possible, however, that the 3.5- and 3.0-kb bands were not generated by MspI digestion of myeloma DNA, particularly if sequences containing IAP genes are rearranged in this tissue. In this case, increased amounts of the 0.5-kb fragment may have been generated from diverse DNA sequences, possibly those larger than 2.0 kb in myeloma DNA. A third source of 0.5-kb fragments in myeloma could be an increase in IAP genes containing region III sequences. The increase of some IAP genes in myeloma DNA has been recently reported (33). The digestion of liver and myeloma DNA with the three enzymes plus HpaII is shown in Fig.

MOL. CELL. BIOL. 1ii

i.

Is-4- 2 2 -

0.

5

I 0 544 K E? V

"v.

FIG. 2. Demethylation of IAP sequence in myeloma. BamHI, EcoRI, HindIII triple digestions were performed with 20 ,ug of DNA, 5 U of each nuclease, and a consensus buffer consisting of 60 mM Tris (pH 7.4), 7 mM MgCl2, 2 mM P-mercaptoethanol, and 80 mM NaCl. A sample was reserved, and the remainder was precipitated with an equal volume of 4 M ammonium acetate and 3 volumes of 95% ethanol. The pellet was collected, lyophilized, and resuspended in water. Three equal parts (5 ptg per part) were further treated in the following ways: one part was incubated in buffer without further digestion; the second part was further digested with MspI (5 U per 5 Fg of DNA). The third part was further digested with HpaII (5 U per 5 ,ug of DNA). Parallel digestions were performed with whole lambda bacteriophage DNA with the mouse DNA to check the completeness of digestion. All incubations were for 16 h at 37°C. Southern blot autoradiograph of 5 ,ug of genomic DNA from 6-month-old mouse liver (L) and myeloma MOPC-315 (My). Lane a, after digestion with BamHI, EcoRI, Hindlll, plus HpaII; lane b, after digestion with BamHI, EcoRI, HindIII, plus MspI; lane c, after digestion with BamHI, EcoRI, and Hindlll only. HindlIl digest of lambda DNA was included as size marker. DNA a size smaller than 0.5kb was run off the gel (1.5%). 32P-labeled 0.8-kb region III probe (108 cpm/Lg) from 81 was used as the probe. Hybridization conditions were as described in the text.

2a. In this case, as in Fig. 2c, the position of the major band is 0.8 kb. The 0.5-kb band, however, is present in the myeloma but not in the liver DNA. Therefore, in myeloma, region III sequences are undermethylated. The 0.5-kb fragment may come from sequences normally meth-

VOL. 3, 1983


ylated in liver DNA or may come from new IAP genes duplicated and not methylated in myeloma cells. A substantial, if not total, amount of genomic 0.8-kb fragment from liver and myeloma remained insensitive to both MspI and HpaII digestion. This insensitivity is probably not due to the lack of CCGG sequence in region III of the genomic IAP genes, since such a site was found in clone 81, but it is more likely that the CCGG sequence is present but fully modified to mCmCGG in many IAP genes of adult mouse DNA. Several IAP gene clones, such as 62, 20, and 14, retain region III homology but either lack the BamHI or HindIII sites common to clones 81, 17, 14, and 19 (Fig. 1) so that no 0.8-kb fragment can be generated by the triple digest regimen. To check whether region III sequences in these types of IAP genes are also methylated in mouse tissue and myeloma, we digested neonatal DNA, liver, and myeloma DNAs with either MspI or HpaII alone. Southern blots of these DNAs were hybridized to the 32P-labeled 0.8-kb fragment to probe for the extent of methylation of IAP genes in these DNAs (Fig. 3). We found that HpaII cut little, if any, IAP genes in neonatal or 6-month-old liver DNA, but digested IAP genes in myeloma DNA substantially. A series of myeloma DNA fragments including a 0.5-kb band were found to hybridize to the 0.8kb region III probe. Similar bands were detected, although in much greater intensities after MspI digests. The location of the two MspI sites flanking the 0.5-kb fragment has not been precisely mapped. It seems reasonable to suggest that one of these sites is within the region III since the cloned IAP 81 has retained an MspI site (Fig. 1B, downward arrow) in this region. The other MspI site may be situated at the junction of region II and region III, very close to the BamHI site found for IAP clone 81. In any case, it is likely that the 0.5-kb MspI fragment represents one of the highly reiterated arrangements of IAP genes since it hybridized to the IAP probe strongly. The probe also detected a 1.05-kb band which was generated by MspI (and HpaII) of myeloma DNA but not of neonatal or liver DNA. Demethylation of mCmCGG to CCGG or mCCGG at this site must have occurred during or after cell transformation. Decrease in methylation of UAP genes in liver DNA of aging mice. Blot hybridization (Fig. 3) indicates that the 0.5-kb MspI-HpaII fragment is specifically produced by HpaII digestion of myeloma DNA. This fragment is produced in all three tissues after MspI digestion. To test whether or not the production of this fragment or other IAP fragments is age dependent, Southern blots of MspI-HpaII-digested liver DNA from mice of different ages were probed by the

1375

B

A My

E

L

My

E

L

ai, kb -1 .5 -I. 2 -1.05 - .95

am

*1

-0.50

FIG. 3. Methylation of LAP genes in neonatal and adult mice. Southern blot autoradiograph of 5 ,ug of genomic DNA from mouse myeloma MOPC-315 (My), neonatal embryos (E), and 6-month-old mouse liver of BALB/c strain (L) after digestion (2 U per iLg of DNA, 16 h, 37°C) with MspI (A) and HpaII (B). A Hinfl digest of pBR322 DNA was included as a size marker. 32P-labeled 0.8-kb region III probe (108 cpm/4Lg) from IAP clone 81 was used as a probe. Hybridization conditions were as described in the text.

0.8-kb region III probe. Samples of DNA from each age group, equal to the samples applied to the gel, were quantitated by the diphenylamine assay (32). This was to assure that equivalent amounts of DNA from each age group were analyzed on the Southern blot. Details are in the legend to Fig. 4. Figure 4B shows a Southern blot autoradiograph of a series of DNAs from livers of mice 6 to 26 months old, representing fully adult to senescent samples. The autoradiograph shows that HpaII cut essentially little of the IAP DNA from young mice but that some HpaII sites are accessible to the enzyme in IAP genes of older mice (Fig. 4B, lane 10). In addition, it demonstrates that MspI generated a group of six discrete IAP-hybridizing fragments, including the 0.5-kb prominent band, and left a significant

1376

MAYS-HOOPES, BROWN. AND HUANG

MOL. CELL. BIOL. 2

3

4

c

(

b

FIG. 4. Increase of 0.5-kb MspI fragments of IAP genes during aging. Ethidium bromide staining (A) and Southern blot autoradiograph (B) of adult and senescent male C57BL/6J DNA digested with MspI or HpaII and hybridized to 0.8-kb region III probe. Each track contains 10 RI of mouse DNA plus 2 jig of bacteriophage lambda DNA. The content of mouse DNA in 10 RlI of sample (0.2 optical density at 260 nm) was determined separately by diphenylamine reaction (32). Close to identical amounts of restricted DNA from different age groups (10.9 ,ug, 6 months old; 10.4 pg, 12 months old; 11.4 p.g, 18 months old; 9.9 ,ug, 24 months old; 10.6 p.g, 26 months old) was applied to gel tracks for electrophoresis. Lanes 1, 6-month-old liver DNA digested with Mspl; lanes 2, 6-month-old liver DNA digested with HpaII; lanes 3, 12-month-old liver DNA digested with MspI; lanes 4, 12-month-old liver DNA digested with HpaII; lanes 5, 18-month-old liver DNA digested with MspI; lanes 6, 18-month-old liver DNA digested with HpaII; lanes 7, 24-month-old liver DNA digested with MspI; lanes 8, 24month-old liver DNA digested with HpaII; lanes 9, 26-month-old liver DNA digested with MspI; lanes 10, 26month-old liver DNA digested with HpaII. Digestions were performed for 16 h at 37°C with 2 U of enzyme per p.g of DNA.

amount of large-molecular-weight MspI-resistant IAP-containing DNA near the top of the gel

(Fig. 4B, odd-numbered lanes). To quantify the results more accurately, we made three separate densitometric scans of the MspI tracks on the autoradiogram shown in Fig. 4B. One of the tracings is shown in Fig. 5. The entire area under the tracing was totaled, as well as the areas of peak 1 and peak 7. In this case, the total area represents all the sequences in liver DNA from 6-, 12-, 18-, 24-, and 26-monthold mice (A, B, C, D, and E, respectively) which hybridize to the IAP probe. The area of peak 1 represents the fraction of IAP-hybridizing sequences which are not digested by Mspl. Peak 7 is the 0.5-kb MspI-HpaII fragment. When total hybridizable areas from all ages were quantified, we found that the variation was not more than 35% from the mean (18 and 26 months). It appears, therefore, that the total quantity of IAP sequences does not change drastically during the aging process. However, the 0.5-kb MspT fragment (peak 7) increased between four and seven times in aged mice (D and E). The ratio of areas of peak 7 to the total for 26-month-old mice is five times the ratio for 6-month-old mice. The quantities of several larger-molecular-

weight MspI fragments (peaks 2 through 6) also increased in 24- and 26-month-old mice. In addition, these increases were simultaneous with a decrease in peak 1, the high-molecular-weight MspT-resistant IAP-hybridizing material. The area ratio of peak 1 to the total is twofold higher in 6-month-old liver than the ratio in 26-monthold liver. This result suggests that many IAPcontaining DNA sequences in aging mice are more susceptible to digestion with MspT. The digestions shown in Fig. 4 and 5 were performed with whole undigested bacteriophage lambda DNA included with the mouse DNA. This DNA served as both size marker and as an indicator of the completeness of digestion. In each case, no partial digestion fragments were present when the incubations were terminated as judged by ethidium bromide staining (Fig. 4A), and the pattern of fragments was identical in the DNA of each gel track. In addition, none of the lanes in Fig. 4B gives an indication of partial digestion of IAP genes since each lane shows the same discrete bands in the autoradiogram. We therefore conclude that the Mspl resistance of the IAP sequences reflects the state of methylation of IAP genes and is not simply a result of incomplete digestion. Rearrangement of the IAP genes during devel-


VOL. 3, 1983

1377

cluding amplification) or modification of these genes has occurred during development. To exolore this idea, DNA from mice of different ages was digested with BamHI, HindIII, and EcoRI and probed by the 0.8-kb region III sequence. Figure 6A and B show an antoradiogram of a blot hybridization with region III probe of BamHI-, HindlIl-, and EcoRI-digested DNAs from livers of mice from neonates to 24 months. This regimen produced the major hybridizing 0.8-kb fragment equally in all samples. The same treatment produced doublet bands at around 1.4 and 1.5 kb and a band at 2.1, 0.75, and ca. 0.55 kb. Although the 0.8-, 2.1-, and 0.75-kb bands were of equal intensity in all samples, there was a decrease in intensity of the 1.5- and 1.4-kb doublet band in liver DNA isolated from aged mice. Although the doublet C was strongly detectable in neonatal mouse liver DNA (Fig. 6A, lane 1), it was only barely visible in liver DNA from 24-month-old mice (Fig. 6A and B, lane 5). The 0.55-kb band, also observed in the neonate (Fig. 6B, lane 1) was not observed in later stages. This 0.55-kb band, however, may be strain specific as it was not present in liver from Swiss Webster or BALB/c neonates (Fig. 6C, lanes 2 and 3). In any event, the 0.55-kb band appears to be stage specific in the C57BL/6J strain as does the 1.5- and 1.4-kb doublet. Unlike the 0.55-kb band, the doublet was observed in neonates of strain BALB/c and Swiss Webster mice, although the intensity of the doublet in both these strains relative to that of the 0.8-kb band is different from the C57BL/6J neonates (Fig. 6C). Whether the difference in intensity is due to strain differences is unknown at present. However, even in BALB/c and in Swiss Webster neonatal DNA, the 1.5- and 1.4kb band was evident; this was not so in adult liver DNA of these strains (data not shown). We were interested in examining the time at which the doublet appears and whether it is present in germline DNA or is established in some stage of embryonic development. We therefore probed BamHI-, HindIII-, and EcoRIN digested DNA from Swiss Webster neonates and FIG. 5. Densitometric trace of IAP sequences in sperm with the 0.8-kb fragment. The autoradiograph is shown in Fig. 6C. Lane 1 shows the mouse liver DNA after MspI digestion. MspI tracks of the autoradiogram shown in Fig. 4B were scanned to germline DNA in which intensity to the 1.5-kb band is clearly reduced or absent compared to compare the quantity of total hybridizable IAP sequences as well as the MspI fragments (peaks 2, 3, 4, the neonatal DNA. Each of the other bands, a 5, 6, and 7) in mouse liver DNA of different age complex series, is present in both samples in groups. (A) 6, (B) 12, (C) 18, (D) 24, and (E) 26 months comparable amounts. This includes the 1.4-kb old. Peak 1 is the high-molecular-weight IAP sequence band. which is MspI resistant. DISCUSSION The hypothesis that methylation of DNA is opment. The high reiteration frequency of the important in silencing gene transcription is sup0.5-kb Msp fragments of IAP genes in senescent ported by our data on IAP sequences. These mice suggests that perhaps rearrangment (in- sequences contain CCGG sequences that are

A

2 3

45 6

7

1378

08


MOL. CELL. BIOL.

B

A 2

3

4

5

2

3

4 5

2

3

2. 21-W: ..

I11

-| .4 -*1&

0. 8-bm UbI 0.75 kb o 55-

... *

VT

-0 8 kb

FIG. 6. Modification of IAP genes during development and aging. (A and B) Southern blot autoradiograph of genomic DNA (20 Fg) from livers of neonatal mice (lane 1) and 3-, 6-, 18-, and 24-month-old C57BL/6J mice (lanes 2, 3, 4, and 5, respectively). DNAs were digested with 150 U of BamHI, EcoRI, and Hindlil enzymes (overnight, 37°C) and hybridized with 0.8-kb region III probe. BstNI-digested pBR322 DNA was included during electrophoresis as a size marker. The X-ray film was exposed with the filter for 24 h (A). The filter shown in (A) was exposed for 120 h (B). (C) Southern blot autoradiograph of genomic DNA (5 ,ug) from Swiss Webster sperm (lane 1), Swiss Webster whole neonate (lane 2), and BALB/c whole neonate (lane 3). DNAs were digested with 30 U of BamHI, EcoRI, and HindIII enzymes (overnight, 37°C) and hybridized with 0.8-kb region III probe. A Hinfl digest of pBR322 was included during electrophoresis as a size marker.

fully methylated to CmCGG in all the non-IAP- during middle to late life in mice (16). Perhaps producing tissues tested, but a substantial num- IAP genes are also expressed at a higher level in ber of copies of IAP contain unmethylated normal tissues of aging mice. CCGG sequences in myeloma where these genes The physiological significance of IAP gene are expressed. The sequences do not exhibit rearrangement remains to be established. How100% demethylation, even in myeloma. The ever, it seems to occur throughout development methylation of some, but not all, copies of and aging. Relevant to this observation is the repetitive sequences has been previously ob- recent report that IAP sequences have been served (37). Presumably, the active copies are found in the intervening sequences of K-chain genes in mutant hybridomas producing aberrant demethylated. The functional significance of the observation K-chain gene products (15). Normally, the IAP that IAP sequences are hypomethylated in aging genes are not found in these regions. Although mouse liver is not clear at present, as IAPs only this event does not define a functional role for rarely have been observed in normal adult tis- the gene, it does illustrate the ability of IAP sue. However, the reported studies of IAPs genes to relocate in the genome. Thus, it is have not specifically focused on their occur- possible that the increase with senescence of the rence in aged tissue. Relevant to this point is the 0.5-kb MspI fragment may be due not simply to observation made by Huebner and Todaro of a decrease in methylation but to progressive IAP the activation of expression of C-type particles gene arrangements which either alter the posi-

VOL. 3, 1983


tions of the MspI-HpaII sites or produce more sequences from which the 0.5-kb fragment is derived.

The rearrangements detected in normal developing and senescing tissues may be important in transcriptional controls of the IAP genes and may be related to methylation changes. Recently, IAP long terminal repeats were shown to have the structure of transposable elements (18). If IAP genes transpose with age, then the transposition may be into a region of DNA with a methylation pattern differing from the pattern in the original location. The myeloma IAP gene organization appears to be similar to the normal adult arrangement; the only new fragments in myeloma are the 1.05kb MspI fragments which probably result from demethylation rather than from rearrangements. The rearrangement of sequences detected by the 0.8-kb IAP probe may reflect endogenous action of a particular set of recombination enzymes. One might speculate that such enzymes would be particularly active in neonatal life, to rearrange sequences such as the immunoglobulin genes, but would persist at a low level throughout life. Thus, repetitive sequences that were susceptible to rearrangement would undergo a burst of rearrangement early in life and then continue a slower process of rearrangement until death. This hypothesis implies that no particular function is served by the IAP rearrangement per se, which is viewed as a reflection of a more general process of development and aging. Of course, it is impossible to rule out some unknown function for the IAP sequence rearrangement via the data available at present. The possible bearing of rearrangement of the genome on senescence remains to be explored. It is interesting that this particular change is evidently progressive throughout life, rather than beginning abruptly at the time of senescence. A loss of reiterated sequences during in vitro aging of fibroblast cell lines has previously been described (35). Correlation of lifespan and DNA repair capacity has been noted for a variety of mammals (13,14). In addition, the degree of sister chromatid exchange observed after mitomycin C-induced DNA damage is a declining function of age in mice (25,31). These observations suggest that mechanisms of DNA breakage and rejoining may play a role in pacemaking for the aging process. Amplification of Alurepeat cluster has recently been observed during serial passage of human diploid fibroblasts in vitro and in lymphocytes from old donors (36). ACKNOWLEDGMENTS This work was supported by an IPA transfer awarded to L.L.M.H. and Gunther Eichhorn and by National Institutes of Health grants CA13953 and 5T32 AG00069 to R.C.C.H.

1379

We thank Gunther Eichhorn of the Gerontology Center, National Institute of Aging, for encouragement of this research.

LITERATURE CITED 1. Breznik, T., and J. C. Cohen. 1982. Altered methylation of endogenous viral promoter sequences during mammary carcinogenesis. Nature (London) 295:255-257. 2. Britten, R. J., D. E. Graham, and B. R. Neufeld. 1974. Analysis of repeated DNA sequences by reassociation. Methods Enzymol. 24:363-418. 3. Calarco, P., and L. Piko. 1973. Expression of A- and Ctype particles in early mouse embryos. J. Natl. Cancer Inst. 51:1971-1975. 4. Calarco, P., and D. SzollosL. 1973. Intercisternal A-particles in ova and preimplantation stages of mouse. Nat. New Biol. 243:91-93. 5. Calvin, H. I. 1976. Methods in cell biology, p. 85-104. David M. Prescott (ed.), Academic Press, Inc., New York. 6. Christy, B. A., and G. Scangos. 1982. Expression of transferred thymidine kinase genes is controlled by methylation. Proc. Natl. Acad. Sci. U.S.A. 79:6299-6303. 7. Cole, M., M. Ono, and R. C. C. Huang. 1981. Terninally redundant sequences in cellular intracisternal A-particle genes. J. Virol. 38:680-687. 8. Dackowski, W., and S. L. Morrison. 1981. Two heavy chain disease proteins with different genomic deletions demonstrate that nonexpressed heavy chain genes contain methylated bases. Proc. Nati. Acad. Sci. U.S.A. 78:70917095. 9. Denhardt, D. 1966. A membrane filter technique for the detection of complementary DNA. Biochem. Biophys. Res. Commun. 23:641-646. 10. Early, P., H. Huang, M. Davis, K. Calane, and L. Hood. 1980. An immunoglobulin heavy chain variable region gene is generated from three segments of DNA: VH, D and JH. Cell 19:981-982. 11. Groudine, M., R. Eisennan, and H. Welntraub. 1981. Chromatin structure of endogenous retroviral genes and activation by an inhibition of DNA methylation. Nature (London) 292:311-317. 12. Harbers, K., A. Schnieke, H. Stuhlmann, D. Jahner, and R. Jaenisch. 1981. DNA methylation and gene expression endogenous retroviral genome becomes infectious after molecular cloning. Proc. Natl. Acad. Sci. U.S.A. 787609-7613. 13. Hart, R. W., S. M. D'Ambrosio, K. J. Ng, and S. P. Modak. 1979. Longevity, stability, and DNA repair. Mech. Ageing Dev. 9:203-223. 14. Hart, R. W., and R. B. Setlow. 1974. Correlation between deoxyribonucleic acid excision repair and life-span in a number of mammalian species. Proc. Natl. Acad. Sci. U.S.A. 71:2169-2173. 15. Hawley, R., M. Shulman, H. Murialdo, D. Gibson, and N. Hozumi. 1982. Mutant immunoglobulin genes have repetitive DNA elements inserted into their intervening sequences. Proc. NatI. Acad. Sci. U.S.A. 79:7425-7429. 16. Huebner, R. J., and G. J. Todaro. 1969. Oncogenes of RNA tumor viruses as determinants of cancer. Proc. Natl. Acad. Sci. U.S.A. 64:1087-1093. 17. Jones, P., and S. Taylor. 1980. Cellular differentiation, cytidine analogs, and DNA methylation. Cell 80:85-93. 18. Kuff, E. L., A. Feenstra, K. Lueders, L. Smith, R. Hawley, N. Hozumi, and M. Shulman. 1983. Intracisternal A particles as movable elements in the mouse genome. Proc. NatI. Acad. Sci. U.S.A. 80:1992-1996. 19. Kuff, E. L., K. K. Lueders, and E. M. Skoilck. 1978. Nucleotide sequence relationship between intracisternal type A particles of Mus musculus and an endogenous retrovirus (M432) of Mus cervicolor. J. Virol. 28:66-74. 20. Lueders, K. K., and E. L. Kuff. 1977. Sequences associated with intracisternal A particles are reiterated in the mouse genome. Cell 12:963-972.

1380


21. Lueders, K. K., S. Segal, and E. L. Kuff. 1977. RNA sequences specifically associated with mouse intracisternal A particles. Cell 11:83-94. 22. Mandel, J. L., and P. Chambon. 1979. DNA methylation: organ specific variations in the methylation pattern within and around ovalbumin and other chicken genes. Nucleic Acids Res. 7:2081-2103. 23. Max, E. E., J. G. Seidnan, and P. Leder. 1979. Sequences of five potential recombination sites encoded close to an immunoglobulin K constant region gene. Proc. Natl. Acad. Sci. U.S.A. 76:3450-3454. 24. McClelland, M. 1981. The effect of sequence specific methylation on restriction endonuclease cleavage. Nucleic Acids Res. 9:5859-5865. 25. Nakanishi, U., R. Dein, and E. L. Schneider. 1980. Aging and sister chromatid exchange. Cytogenet. Cell Genet. 27:82-87. 26. Neel, B. G., W. S. Hayward, H. L. Robinson, J. Fang, and S. M. Astrin. 1981. Avian leukosis virus-induced tumors have common proviral integration sites and synthesize new RNAs: oncogenesis by promoter insertion. Cell 23:323-334. 27. Ono, M., M. D. Cole, A. T. White, and R. C. C. Huang. 1980. Sequence organization of cloned intracisternal A particle genes. Cell 21:465-473. 28. Payne, G. S., S. A. Cortneldge, L. B. Crittenden, A. M. Fadly, J. M. Bishop, and H. E. Vannus. 1981. Analysis of avian leukosis virus DNA and RNA in bursal tumors: viral gene expression is not required for maintenance of the tumor state. Cell 23:311-322. 29. Razln, A., and A. D. Riggs. 1980. DNA methylation and gene function. Science 210:604-610. 30. Rlgby, P. W. J., R. Dieckmann, C. Rhodes, and P. Berg. 1977. Labeling of deoxyribonucleic acid to high specific activity in vitro by nick translation with DNA polymerase I. J. Mol. Biol. 13:237-251. 31. Schneider, E. L., D. Kram, Y. Nakanishi, R. E. Monticone, R. R. Tice, B. A. Gilman, and M. L. Nieder. 1979. The effect of aging on sister chromatid exchange. Mech. Ageing Dev. 9:303-311. 32. Schneider, W. C. 1957. Determination of nucleic acids in tissue by pentose analysis. Methods Enzymol. 3:680-684.

MOL. CELL. BIOL. 33. Shen-Ong, G., and M. D. Cole. 1982. Differing populations of intracisternal A-particle genes in myeloma tumors and mouse subspecies. J. Virol. 42:411-421. 34. Singer, J., J. Roberts-Ems, F. W. Luthardt, and A. D. Riggs. 1979. Methylation of DNA in mouse early embryos, teratocarcinoma cells and adult tissues of mouse and rabbit. Nucleic Acids Res. 7:2369-2385. 35. Shmookler-Reis, R. J., and S. Goldstein. 1980. Loss of reiterated DNA sequences during serial passage of human diploid fibroblasts. Cell 21:739-749. 36. Shnookler-Reis, R. J., C. K. Lumpkin, Jr., J. R. McGill, K. T. Riabowal, and G. Goldstein. 1983. Extrachromosomal circular copies of an "inter-Alu" unstable sequence in human DNA are amplified during in vitro and in vivo aging. Nature (London) 301:394-398. 37. Sobieski, D. A., and F. C. Eden. 1981. Clustering and methylation of repeated DNA: persistence in avian development and evolution. Nucleic Acids Res. 9:6001-6015. 38. Southern, E. M. 1975. Detection of specific sequences among DNA fragments separated by gel electrophoresis. J. Mol. Biol. 98:503-517. 39. Stuhlmann, H., D. Jahner, and R. Jaenisch. 1981. Infectivity and methylation of retroviral genomes is correlated with expression in the animal. Cell 26:221-231. 40. Van der Poeg, L. H. T., and R. A. Flavell. 1980. DNA methylation in the human -y88-globin locus in erythroid and nonerythroid tissues. Cell 19:947-958. 41. Van Zwitten, M., G. Zurcher, H. Solleweld, and C. Hollander. 1981. Pathology, p. 1-36. In W. Adler and A. Nordin (ed.), Immunological techniques applied to aging research. CRC Press, Boca Raton, Florida. 42. Vardimon, L., R. Neumann, I. Kuhlmann, D. Sutter, and W. Doerfier. 1980. DNA methylation and viral gene expression in adenovirus-transformed and -infected cells. Nucleic Acids Res. 8:2461-2473. 43. Weintraub, H., A. Larsen, and M. Groudine. 1981. aglobin-gene switching during the development of chicken embryos: expression and chromosome structure. Cell 24:333-344. 44. Wivel, N. A., and G. H. Smith. 1971. Distribution of intracisternal A-particles in a variety of normal and neoplastic tissues. Int. J. Cancer 7:167-175.