Studies on the molecular-genetic basis of replicative senescence in Werner syndrome and normal fibroblasts

Experimental Gerontology, Vol. 24, pp. 461--468, 1989

0531-5565/89 $3.00 + .00 Cop)right © 1989 Pergamon Press plc

Printed in the USA. All fights reserved.

STUDIES ON THE MOLECULAR-GENETIC BASIS OF REPLICATIVE SENESCENCE IN WERNER SYNDROME AND NORMAL FIBROBLASTS

S. GOLDSTEIN, S. MORA~O, H. BENES, E.J. MOER~Ar~, R.A. JONES, R. THWEATr, R.J. SHMOOKLERREIS and B.H. HOWARD1 Departments of Medicine and Biochemistry and Molecular Biology, University of Arkansas for Medical Sciences, and Geriatric Research Education and Clinical Center, John L. McClellan Memorial Veterans Hospital, Little Rock, Arkansas

- - Based on evidence that human diploid fibroblasts (HDF) from the Wemer syndrome (WS) of premature aging might overexpress an inhibitor of DNA synthesis (IDS), we prepared a eukaryotic eDNA expression library from WS mRNA and tested it for IDS activity in a transient assay. Two of six WS eDNA pools tested gave IDS activity, then on plus/minus screening revealed several differentially expressed eDNA clones. By slot blot and Northern analysis, one eDNA clone was found to be overexpressed in WS and normal senescent HDF, but not in quiescent normal HDF, indicating that it is senescence-specific. Further studies are needed to clarify: a) whether this eDNA truly acts as an IDS, b) if so, whether it acts alone or in concert with other cDNAs, and c) whether it is involved in the degenerative and malignant sequelae of WS and normal aging. Abstract

Key Words: Werner syndrome, DNA synthesis inhibitors, overexpression, senescence, quiescence, eDNA expression library

INTRODUCTION THE SIGNALreport of Hayflick and Moorehead (1961), that human diploid fibroblasts (HDF) have a finite replicative life span, launched the modem era of aging research: studies were now possible at the level of relatively homogeneous cell populations and their biochemical and molecular components. Copious studies have sought mechanisms for HDF senescence, many pursuing the error catastrophe hypothesis of Orgel (1963). However, it has become increasingly likely that HDF senescence is a genetically programmed, quasi-differentiative process. This conceptual transition has come about for several reasons (Goldstein and Shmookler Reis, 1984; Norwood and Smith, 1985). In brief, no convincing evidence has been presented that error

Correspondence to: S. Goldstein, John L. McClellan Memorial Veterans Hospital, 151 Research, 4300 W. 7th Street, Little Rock, AR 72205. ~Laboratory of Molecular Biology, National Cancer Institute, National Institutes of Health, Bethesda, MD 20892.

461

462

s. GOLDSTEIN et al.

frequencies increase in protein or DNA synthesis during the HDF life span. More directly, the division potential of HDF appears to depend on the number of cell generations undergone, up to a critical limit (see Harley and Goldstein, 1978), and not upon metabolic or calendar time, a result which is best reconciled with a genetic program. Cell fusion studies also provide powerful mechanistic insights by demonstrating that in heterokaryons (hybrid cells with nuclei from different cell sources in a single cytoplasm), the senescent nucleus has a dominant inhibitory effect on nuclear DNA synthesis of early-passage, vigorously growing (young) HDF. Moreover, this inhibition can be abolished by drugs which block protein and RNA synthesis. Taken together with the recent demonstration that microinjected mRNA from senescent cells can inhibit DNA synthesis in young HDF (Lumpkin et al., 1986), the data strongly suggest that senescent HDF elaborate a protein inhibitor of DNA synthesis (IDS). SENESCENCE VERSUS QUIESCENCE What remains unclear is whether the IDS is unique to aging cells, or whether senescent cells recruit the same genes which lead to quiescent growth arrest (Stein et al., 1985; Lumpkin et al., 1986). Recently, Schneider et al. (1988) described six cDNA clones preferentially expressed in a cDNA library of mouse 3T3 cells growth arrested by serum depletion and high density. The cDNAs varied in abundance from 0.0002 to 2% of the library and in size from 0.8 to 10 kilobase pairs, and the corresponding mRNAs identified by Northern analysis were downregulated with different kinetics upon re-induction of growth by serum repletion. One of these cDNAs was expressed whether growth arrest was induced by density-dependent inhibition or serum starvation. While these studies provide important insights into expression of growth-arrestspecific genes in 3T3 cells, the molecular connections between quiescence and senescence of HDF remain unclear. To pursue the IDS, we initially attempted to prepare senescence-specific subtracted cDNA libraries (e.g. see Sargent and Dawid, 1984; Duguid et al., 1988), screened with subtracted senescent vs. young probes. For this procedure, cDNA preparations from normal senescent HDF were hybridized to a six-fold excess of young cell mRNA. The single-stranded cDNA remaining after removal of cDNA:RNA duplexes and free mRNA by hydroxylapatite chromatography was then used to generate a subtracted library, or as template for synthesis of 32p-radiolabelled subtracted probe. These attempts were unsuccessful (H. Benes et al., unpublished results), perhaps due to the admixture of senescent cells in ostensibly young cultures (Harley and Goldstein, 1978; Sherwood et al., 1988). We elected, therefore, to focus on HDF from Werner Syndrome (WS). WERNER SYNDROME WS is a rare autosomal recessive disorder featuring short stature, a readily discernible phenotype of premature aging, and in particular the early appearance of a wide variety of age-related pathology including atherosclerosis, malignancy, osteoporosis, insulin-resistant diabetes mellitus, lenticular cataracts, and severe skin atrophy and ulceration (Epstein et al., 1966). A universal finding is the reduced capacity for growth of HDF derived from subjects with WS compared to age-matched controls. WS cells grow more slowly, develop the senescent morphology early, and demonstrate a pronounced reduction in the replicative life span (see Salk, 1982). We chose WS HDF, therefore, with the expectation that overexpression of

MOLECULAR BASIS OF REI~ICATIVE SENESCENCE

463

senescence-specific, growth-inhibitory genes in these mutant cells might allow differential screening of cDNA libraries without prior subtraction. PREPARATION OF cDNA LIBRARY We have utilized a eukaryotic cDNA expression library (Okayama et al., 1987) to screen for the IDS by a novel technique (Padmanabhan et al., 1988; and see below), which allows relatively rapid assay for IDS following transfection of such WS cDNAs into young HDF. We applied an additional strategy to prepare the cells, in analogy with recent approaches to positive growth regulation using serum depletion/repletion protocols to isolate immediate early genes (Lau and Nathans, 1987; Almendral et al., 1988). We reasoned that levels of the putatively overexpressed IDS in WS cells might be further augmented by serum depletion/repletion, in homeostatic response to positive growth signals following induction of immediate early genes. By the same token, the serum depletion/repletion protocol, when used on young control cells, would further reduce their levels, if any, of IDS, so that differences between WS and control cDNAs would stand out in bolder relief. We now report early results in our attempts to identify a senescent IDS. A single strain of HDF, derived from skin of a 47-year-old Japanese subject with classical WS, the product of a consanguineous marriage, was utilized throughout these studies. WS cells, approximately halfway through their abbreviated replicative life span of - 1 6 mean population doublings, were grown to half confluence in Eagle's medium supplemented with 15% fetal bovine serum, rinsed with phosphate-buffered saline, and incubated at 37 °C for 5 days in medium containing 1% serum, the minimum concentration required to maintain these cells as intact monolayers. Cells were then refed with medium containing 20% fetal bovine serum, and harvested 24 h later. RNA was prepared, passed over oligo dT columns to isolate poly A + mRNA, and then inserted as cDNA into a eukaryotic expression vector as described (Okayama et al., 1987). This recombinant vector uses the SV40 early promoter to drive constitutive expression of the downstream cDNA, and an SV40 intron just 5' to and a polyadenylation site 3' to the insert, which are believed to enhance cytoplasmic export and stability of the mRNA, respectively. To determine the quality of this cDNA library, we assessed recovery of mRNA encoding human beta actin, which is rather abundant in animal cells, and has a coding region containing 61% G + C (Ponte et al., 1984). The efficiency in recovering a full-length cDNA of high GC content and near-modal mRNA size thus provides a stringent test of the cloning procedure. We screened 7200 colonies with the human beta actin probe, and found 44 (0.61%) contained inserts corresponding to beta actin. This frequency lies in the middle of the reported range of 0.2 to 1.0 percent for beta actin mRNA abundancy in a variety of cells and tissues. Seven of these 44 actin clones (16%), following excision with BamHI endonuclease, yielded inserts of - 1 . 9 kbp, the full length of the mRNA. To ensure representation of translatable mRNA sequences among these clones, we synthesized an 18-mer oligonucleotide primer homologous to the sense strand of SV40 DNA in the plasmid vector, just upstream from the cDNA, and determined the sequence of about 200 bp spanning the 5' untranslated regions of two 1.9 kbp beta actin cDNAs by the dideoxy-chain termination method. We demonstrated in one clone that the coding sequence extended to the 5' AUG initiator codon for protein synthesis and would thus allow proper initiation and full synthesis of beta actin, and by inference, the IDS protein. In the second clone, which was thought to be full length because of its BamHI-excised 1.9 kbp insert, we found deletion of the upstream region including the AUG site. Therefore, some instability occurs during preparation and manipulation of these libraries, but we can conclude

464

s. GOLDSTEIN et al.

that roughly 8% of beta actin clones are potentially expressible. Based on the 0.8% abundancv of antiproliferative mRNAs in senescent normal HDF estimated by Lumpkin et al. (1986), we might anticipate on the order of 0.06% full-length, expressible IDS clones in a cDNA library' from those cells. WS cells, presumed to be IDS overproducers, may be expected to have a substantially higher representation of IDS mRNA. For example, if IDS were 10-fold more abundant in WS, and if 8% of these cDNAs are full length and expressible, then perhaps - 0 . 6 % of WS cDNAs could express the IDS, and in pools of about 200 cDNA clones, there could be 1 IDS cDNA which might be detectable by our assay. ASSAY SYSTEM FOR INHIBITORS OF DNA SYNTHESIS We have employed a transient assay system which utilizes a highly efficient, yet gentle electroporation protocol to introduce the gene for the 55 kd ( " T A C " ) subunit of the interleukin-2 receptor (IL2R) into recipient cells. The particular value of this technique for our purpose is that it enables most HDF to continue cycling (Goldstein et al., 1989). Once TAC is expressed as a neoantigen on the cell surface, magnetic affinity cell sorting (MACS) is carried out to select the transfected subpopulation (Padmanabhan et al., 1988). In brief, a pRSV-IL2R recombinant plasmid containing the IL2R cDNA, driven by a promoter within the Rous sarcoma virus 3' long-terminal repeat (C.M. Fordis et al., manuscript in preparation), is co-transfected by electroporation into young HDF synchronized at G2/M phase, along with pools of - 2 0 0 randomly selected cDNA clones from the WS cDNA expression library. Following overnight incubation in 10 mM sodium butyrate and 40 h of expression time, the last 14 to 16 h of which include labeling with 3H-thymidine, cells are harvested and incubated in suspension in a non-aggregating medium with magnetic beads coupled to a monoclonal antibody specific for TAC. Cells are then sorted in a magnetic field, by placing the cell/bead suspension between magnetic plates. Not all cells will have taken up the pRSV-IL2R, but in that fraction that does (usually 10-30%) there is -> 90 percent co-uptake and co-expression of the second plasmid (Padmanabhan et al., 1988), in this case the WS cDNAs. After resuspending the cells and repeating the magnetic procedure twice, an enriched cell population is obtained and prepared for liquid scintillation counting of 3H-thymidine incorporated into DNA. INHIBITION OF DNA SYNTHESIS BY WERNER SYNDROME cDNAS Results of one such experiment are shown in Table 1. Six randomly chosen pools of the WS cDNA library, each containing - 2 0 0 cDNA clones, were assayed for IDS expression. Controls were the same recombinant vector containing individual pure species of cDNA (Table 1 legend), but the various WS pools also serve as controls for each other. Two pools, WS-2/3 and WS-6, showed moderate inhibitory activity, which we interpret to indicate that they may contain, in contrast to the other four pools, one or more full-length cDNA clones encoding (and expressing) the IDS protein. Alternatively, if two or more such proteins are involved in IDS activity, then pools WS-2/3 and WS-6 may contain the multiple species needed to act in concert. WS-2/3 and WS-6, plus pool WS-5, the latter shown not to have IDS activity (Table 1), were re-assayed as in Table 2. It should be noted that in this experiment we doubled the amount of DNA electroporated into cells and also assayed for IDS at two cell densities. The data show a more pronounced inhibitory effect of WS-2/3 and WS-6, particularly after plating cells at high density. Although the basis of this phenomenon is unclear, we speculate that the onset of

465

MOLECULARBASIS OF REPLICATIVESENESCENCE TABLE1. ASSAYFORINHIBITIONOF 3H-THYMIDINEINCORPORATIONIN YOUNGNORMALFIBROBLASTS raTERTRAr~SF~'rlONWITHcDNA lOOkSFROMWS FIBROBLASTS Single cDNA or Pool

Control

Werner Syndrome

A2 P4

WS- 1 WS-2/3 WS-4 WS-5 WS-6

Percentage of 3H-TdR Sorted

Percentage of Control

14.7 11.1 x= 12.9

--100

11.2 9.4 12.0 12.1 9.1

87 73 93 94 71

w s cDNA pools were grown on nitrocellulose filters containing ~200 colonies each and plasmid DNA purified from randomly chosen filters, except in the case of WS-2/3 (where colonies from two filters were combined). Controls were individual species of cDNA: A2 contained an -600 bp eDNA insert of unknown identity, while P4 contained a full-length expressible eDNA for HPRT (p4aA8, Jolly et al., 1983). Vigorously dividing human fibroblasts (fetal lung strain HSC172, 1.5 x 106 ceils in 0.5 ml total volume) were electroporated at 140 V and 960 ~F capacitance as described (Goldstein et al., 1989) with 20 I~g of the sorting vector pRSV-IL2R, 20 I~g of one of the WS eDNA pools or control plasmids, and 20 i~g salmon sperm DNA as carder, then seeded at 5 × 105 cells/T25 flask, into growth medium containing 10 mM Na butyrate. After 14 h medium was removed, adherent cells were rinsed with phosphate buffered saline, and refed with growth medium minus butyrate. Twenty-six hours later, that is, following a 40-h expression time, with 3H-thymidine labeling over the last 16 h, cells were harvested, incubated with monoclonal antibody to the TAC subunit of IL2R coupled to magnetic beads and sorted magnetically. Incorporation of 3H-TdR into DNA was determined by trichloroacetic acid precipitation and 3H counting in a liquid scintillation counter. A presorted aliquot of cells was removed for determination of initial (total) counts and results are expressed as percent 3H sorted (100 × 3H DPM in sorted cells divided by 31-1DPM in total cells).

density-dependent inhibition occurring at the time of 3H labeling in these H D F cultures m a y accentuate the IDS activity within W S c D N A pools.

ISOLATION OF DIFFERENTIALLY EXPRESSED CDNA CLONES W e than c a r d e d out + / - screening of the WS-2/3 and W S - 6 c D N A pools, utilizing two 32p-labeled c D N A probes prepared from total m R N A s , the first derived from W S H D F and the second from H D F of a n o r m a l 52-year-old male. W e have to date identified five clones with differential expression, confirmed by slot-blot and Northern analysis to have -> 5 × the hybridization signal to W S R N A versus normal R N A . O n e clone has proven to be h o m o l o g o u s to fibronectin based on probing at high stringency with a 2.3 kb fragment o f the 7.9 kb full-length fibronectin c D N A (Kornblihtt et al., 1984). Three other clones are h o m o l o g o u s to each other, and on Northern analysis each hybridizes to two R N A species at - 5 . 8 and 4.8 kb. It is noteworthy that the steady-state R N A levels determined on Northern blots are higher in W S than in n o r m a l early-passage H D F b y -> 10-fold for the upper band but only b y two- to three-fold for the lower band. D N A sequence determination of - 2 0 0 nucleotides of one o f these

s. GOLDSTEINet al.

466

TABLE 2. ASSAYFOR INHmrnoNOF 3H-THYMIDINEINCORPORATIONIN YOUNG NORMALFIBROBLASTSAFrERTRANS~ON WITHcDNA POOLSFROMWS FIBROBLASTS:EFFECTOF CELLPLATINGDENSITY

Cell Density

Single eDNA or Pool

Percentage of 3H-TdR Sorted

Percentage of Control

High

P4 WS-2/3 WS-5 WS-6

29.3 13.5 22.6 12.6

100 46 77 43

Low

P4 WS-2/3 WS-5 WS-6

23.7 15.6 21.7 14.5

100 67 92 61

Procedural details were as in Table 1, with the following exceptions. young cells were eleetroporated with 40 p,g of each WS eDNA pool plus 20 p,g of pRSV-IL2R; salmon sperm DNA carrier was omitted to keep total DNA at 60 p,g in 0.5 ml total volume. After electroporation, flasks were seeded at 5 x 105 cells (high density) or 2 × l0 s cells (low density).

clones indicates a 96% homology with the human proalpha 1 chain of type I collagen, which is believed to yield two polymorphic RNA species from a single gene (Chu et al., 1985). IDENTIFICATION OF A SENESCENCE-SPECIFIC eDNA One clone (6-5) has proved extremely interesting. Northern analysis reveals a single RNA species at - 2 . 5 kb; both Northern and slot-blot quantitations indicate -> 10-fold overexpression in both early-passage WS cells and senescent normal cells, compared to young normal cells. Moreover, expression is very low in young normal cells whether they are in exponential growth or at the very high density that exists when they are maintained, with refeeding, for up to two weeks as postconfluent cultures. Studies are now underway to ascertain expression of clone 6--5 in HDF rendered quiescent by serum depletion, and in a variety of cells and tissues in vitro and in vivo taken from rodent and human donors of various ages. Functional assays by MACS and fluorescence activated cell sorting and other systems should also directly reveal whether clone 6-5 encodes a true senescence-specific IDS, and if so, whether it acts alone or requires the concerted action of additional species. Although fibronectin and procollagen have not hitherto been known to have IDS activity, they could in their locations as extracellular or pericellular proteins, and when overexpressed, as normal or perhaps as variant species, serve to mask receptors for growth-promoting stimuli. They would then act synergistically, as cofactors, with a true IDS working within the cell. In the context of growth arrest during senescence and quiescence, we were eager to examine the potential role of the only tumor suppressor eDNA so far purified, the retinoblastoma susceptibility (RB) gene (Friend et al., 1986), which in its normal configuration has now been shown to selectively inhibit growth in vitro and tumorigenesis in vivo of RB, osteosarcoma and certain breast cancer lines (Huang et al., 1988). We found on probing Northern blots with RB eDNA, that contrary to expectation, RB gene expression was actually highest in young normal HDF during logarithmic growth, but fell off markedly in young cells at confluence, and also in old normal HDF and WS cells when confluent or subconfluent. These preliminary observations

468

s GOLDSTEINet al

human hypoxanthine phosphoribosyltransferase. Proc. Natl. Acad. Sci. USA 80, 477~J~81. 1983. KORNBLIHTT, A.R., VIBE-PEDERSEN, K., and BARALLE, F.E. Human fibronectin: Molecular cloning evidence for two mRNA species differing by an internal segment coding for a structural domain. EMBO J. 3, 221-226, 1984. LAU, L.F. and NATHANS, D. Expression of a set of growth-related immediate early genes in BALB/c 3"13 cells: Coordinate regulation with c-fos or c-myc. Proc. Natl. Acad. Sci. USA 84, 1182-1186. 1987. LUMPKIN, C.K., JR., McCLUNG, J.K., PEREIRA-SMITH. O.M., and SMITH, J.R. Existence of high abundance antiproliferative mRNAs in senescent human diploid fibroblasts. Science 232, 393-395. 1986. NORWOOD, T.H. and SMITH, J.R. The cultured fibroblast-like cell as a model for the study of aging. In: Handbook of the Biology of Aging, Finch, C.E. and Schneider, E.L. (Editors), pp. 291-321, Van Nostrand Reinhold Co., New York, 1985. OKAYAMA, H., KAWAICHI, M., BROWNSTEIN, M., LEE, F., YOKOTA, T., and ARAI, K. High-efficiency cloning of full-length cDNA: Construction and screening of cDNA expression libraries for mammalian cells. Meth. Enzymol. 154, 3-28, 1987. ORGEL, L.E. The maintenance of the accuracy of protein synthesis and its relevance to ageing. Proc. Natl. Acad. Sci. USA 49, 517-521, 1963. PADMANABHAN, R., CORSICO, C.D., HOWARD, T.H., et al. Purification of transiently transfected cells by magnetic affinity cell sorting. Anal. Biochem. 170, 341-348, 1988. PONTE, P., NG, S.Y., ENGEL, J., GUNNING, P., and KEDES, L. Evolutionary conservation in the untranslated regions of actin mRNAs: DNA sequence of a human beta-actin cDNA. Nucleic Acids Res. 12, 1687-1696, 1984. SALK, D. Werner's syndrome: A review of recent research with an analysis of connective tissue metabolism, growth control of cultured cells, and chromosomal aberrations. Hum. Genet. 62, 1-15, 1982. SARGENT, T.D. and DAWID, I.B. Differential gene expression in the gastrula of Xenopus laevis. Science 222, 135-139, 1983. SCHNEIDER, C., KING, R.M., and PHILIPSON, L. Genes specifically expressed at growth arrest of mammalian cells. Cell 54, 787-793, 1988. SHERWOOD, S.W., RUSH, D., ELLSWORTH, J.L., and SCHIMKE, R.T. Defining cellular senescence in IMR-90 cells: A flow cytometric analysis. Proc. Natl. Acad. Sci. USA 85, 9086-9090, 1988. STEIN, G.H., NAMBA, M., and CORSARO, C.M. Relationship of finite proliferative lifespan, senescence, and quiescence in human cells. J. Cell. Physiol. 122, 343-349, 1985.

MOLECULARBASISOFREPLICATIVESENESCENCE

467

imply, therefore, that RB gene expression does not play a role in growth arrest during senescence or quiescence, at least of HDF,

CONCLUSION In views of these early results, further studies should provide insights into defining: a) the genetic distinctions between senescent and quiescent states; b) whether expression of the senescence-specific clone 6-5 correlates with, and indeed contributes to replicative arrest in vitro and in vivo; and c) whether the cognate gene, or mutations therein, is involved in the origin of degenerative and malignant diseases, respectively, in WS and in normal aging persons. Finally, while WS cells may overproduce an IDS, it is difficult to reconcile the autosomalrecessive nature of this disorder with the incompletely dominant interactions of cell hybrids between WS and normal young HDF. We have proposed (Goldstein e t a l . , in press) that WS is caused by a primary mutation in a t r a n s - a c t i n g repressor gene) allowing derepression of a secondary locus, putatively encoding a protein for IDS. Related studies in our laboratory may also provide insight into this possibility. Acknowledgments -- We thank Hiroto Okayama for his generous assistance in constructionand handling of the WS

eDNA library,James Smithand ThaddeusDryja for supplyingfibronectinand RB probes, respectively,MichaelFordis for helpful discussion,Daphne Chien and Tazuko Howard for technical assistance, and Diane Earnest for typing the manuscript. Supportedby grants from the NationalInstituteon Aging, the Veterans Administration,and fellowshipsto SM from the SandozFoundationfor GerontologicalResearch, and to SG from the John SimonGuggenheimFoundation and the Fogarty InternationalCenter of the NIH.

REFERENCES ALMENDRAL, J.M., SOMMER, D., MACDONALD-BRAVO, H., BURCHKARDT, J., PERERA, J., and BRAVO, R. Complexityof the early geneticresponse to growth factors in mouse fibroblasts.Mol. Cell. Biol. 8, 2141)-2148, 1988. CHU, M.L,, DE WET, W., BERNARD, M., and RAMIREZ, F. Fine structural analysis of the human pro-et-l(I) collagengene. Promoter structureAlu I repeats, and polymorphictranscripts.J. Biol. Chem. 2611, 2315-2320, 1985. DUGUID, J.R., ROHWER, R.G., and SEED, B. Isolation of cDNAs of scrapie-modulatedRNAs by subtractive hybridizationof a eDNA library. Proc. Natl. Acad. Sci. USA 85) 5738-5742, 1988. EPSTEIN, C.J., MARTIN, G.M., SCHULTZ, A.L., and MOTULSKY, A.G. Werner's syndrome:A review of its symptomatology, natural history, pathologic features, genetics and relationship to the natural aging process. Medicine 45) 177-221, 1966. FRIEND, S.H., BERNARDS, R., ROGELJ, S., et al. A human DNA segment with properties of the gene that predisposes to retinoblastomaand osteosarcoma. Nature 323, 643-646, 1986. GOLDSTEIN, S. and SHMOOKLERREIS, R.J. Geneticmodificationsduringcellularaging. Mol. Cell. Biochem. 64, 15-30, 1984. GOLDSTEIN, S., FORDIS, C.M., and HOWARD, B.H. Enhancedtransfectionefficiencyand improvedcell survival after electroporation of G2/M-synchronizedcells and treatment with sodium butyrate. Nucleic Acids Res. 17) 3959-3971, 1989, GOLDSTEIN, S., MURANO, S., and SHMOOKLERREIS, R.J. Wernersyndrome:A moleculargenetic hypothesis. J. Gerontol., in press. HARLEY, C.B. and GOLDSTEIN, S. Culturedhumanfibroblasts:Distributionof cell generationsand a criticallimit. J. Cell. Physiol. 97, 509-516, 1978. HAYFLICK, L. and MOORHEAD, P.S. The serial cultivationof human diploid cell strains. Exp. Cell Res. 25, 585-621, 1961. HUANG, H.J.S., YEE, J.K., SHEW, J.Y., et al. Suppressionof the neoplasticphenotype by replacementof the RB gene in human cancer cells. Science 242, 1563-1566, 1988. JOLLY, D.J., OKAYAMA,H., BERG, P., et al. Isolationand characterizationof a full-lengthexpressibleeDNA for