Am. J. Hum. Genet. 52:661-667, 1993

Loss of Telomeric DNA during Aging of Normal and Trisomy 21 Human Lymphocytes Homayoun Vaziri,* Franpois Schachter,4 Irene Uchida,t Lan Weit Xiaoming Zhu,§ Rita Effros,§ Daniel Cohen,4 and Calvin B. Harley* Departments of Biochemistry and tPediatrics, McMaster University, Hamilton, Ontario; tCentre d'Etude du Polymorphisme Humain, Paris; and §Department of Pathology, University of California School of Medicine, Los Angeles

Summary The telomere hypothesis of cellular aging proposes that loss of telomeric DNA (TTAGGG) from human chromosomes may ultimately cause cell-cycle exit during replicative senescence. Since lymphocytes have a limited replicative capacity and since blood cells were previously shown to lose telomeric DNA during aging in vivo, we wished to determine (a) whether accelerated telomere loss is associated with the premature immunosenescence of lymphocytes in individuals with Down syndrome (DS) and (b) whether telomeric DNA is also lost during aging of lymphocytes in vitro. To investigate the effects of aging and trisomy 21 on telomere loss in vivo, genomic DNA was isolated from peripheral blood lymphocytes of 140 individuals (age 0-107 years), including 21 DS patients (age 0-45 years). Digestion with restriction enzymes Hinfl and RsaI generated terminal restriction fragments (TRFs), which were detected by Southern analysis using a telomere-specific probe (32P(C3TA2)3). The rate of telomere loss was calculated from the decrease in mean TRF length, as a function of donor age. DS patients showed a significantly higher rate of telomere loss with donor age (133 ± 15 bp/year) compared with age-matched controls (41 ± 7.7 bp/year) (P < .0005), suggesting that accelerated telomere loss is a biomarker of premature immunosenescence of DS patients and that it may play a role in this process. Telomere loss during aging in vitro was calculated for lymphocytes from four normal individuals, grown in culture for 10-30 population doublings. The rate of telomere loss was -120 bp/cell doubling, comparable to that seen in other somatic cells. Moreover, telomere lengths of lymphocytes from centenarians and from older DS patients were similar to those of senescent lymphocytes in culture, which suggests that replicative senescence could partially account for aging of the immune system in DS patients and in elderly individuals.

Introduction The dependence of DNA polymerases on primers and the unidirectional 5'-to-3' direction of synthesis poses a problem for complete replication of the ends of eukaryotic chromosomes (Olovnikov 1971, 1973; Watson 1972). This problem is circumvented in eukaryotes by telomeres, the specialized nucleoprotein structures containing highly conserved repeats (reviewed in Blackburn and Szostak 1984). In vertebrates, telomeric DNA Received September 8, 1992; revision received December 3, 1992. Address for correspondence and reprints: Dr. Calvin B. Harley, Department of Biochemistry, McMaster University, Hamilton, Ontario L8N 3Z5, Canada. ©) 1993 by The American Society of Human Genetics. All rights reserved. 0002-9297/93/5204-0002$02.00

terminates in TTAGGG repeats (Moyzis et al. 1988; Allshire et al. 1989). Telomeric repeats are synthesized by telomerase, a ribonucleoprotein that is capable of elongating telomeres de novo and hence of overcoming loss of telomeric DNA due to incomplete replication or degradation of ends. Telomerase activity was first detected in Tetrabymena (Greider and Blackburn 1985) and later in extracts from immortalized human cells (Morin 1989; Counter et al. 1992). Several lines of evidence suggest that telomere shortening plays a causal role in cellular aging. We have previously shown that, in human fibroblasts, telomeres shorten as a function of cell doublings in vitro and in vivo and that initial telomere length predicts the replicative capacity of these cells (Harley et al. 1990; Allsopp et al. 1992). Loss of telomeric DNA during aging in vivo 661

662

has also been observed in peripheral blood cells and colon mucosa epithelia (Hastie et al. 1990). Telomere shortening in normal (nonimmortalized) human cell strains is associated with the inability to detect telomerase in extracts from these cells (Counter et al. 1992). In contrast, immortal cells express telomerase, and their telomeres do not progressively shorten (Counter et al. 1992). A causal link between telomere loss and cell-cycle exit has not been proved, but the sudden increase in the number of dicentric chromosomes in senescing fibroblasts (Saksela and Moorhead 1963; Benn 1976; Sherwood et al. 1989) and the significant age-related increase of these abnormalities in human peripheral blood lymphocytes (Bender et al. 1989) suggest that the integrity of chromosome ends is compromised during replicative aging. The above observations are the basis of the telomere hypothesis of cellular aging and immortalization (Harley 1991). Aging of the immune system could account for some of the morbidity of elderly individuals and DS patients. If this were true and if telomere shortening plays a role in cellular aging, then one might predict critically shortened telomeres in the lymphocytes of these individuals. To investigate this, we measured the length of terminal restriction fragments (TRFs) that contain the terminal TTAGGG repeats in peripheral blood lymphocytes (PBLs) of 140 individuals (age 0-107 years), including 18 centenarians (age >99 years) and 21 Down syndrome (DS) patients. Here we report that mean TRF length shortens as a function of age in these cells in vitro and in vivo and that cells from DS patients have a higher rate of telomere loss with age, compared with age-matched controls.

Subjects, Material, and Methods All blood samples were obtained with informed consent. Normal subjects (age 0-90 years) were primarily

healthy volunteers from the Hamilton (southwestern Ontario) region who were ostensibly free from any blood disorder. DNA from all centenarians (age >99 years) was from the Centre d'Etude du Polymorphisme Humain (CEPH) collection, Paris. DS individuals were all trisomy 21 by cytogenetic analysis and were also primarily from southwestern Ontario. Culture of Human Peripheral Blood T Lymphocytes Adult peripheral blood samples were collected, and mononuclear cells were isolated by Ficoll-Hypaque gradient centrifugation and then were cryopreserved in liquid nitrogen. Cultures were initiated by mixing 106

Vaziri et al.

mononuclear cells with 106 irradiated (8,000 Rad) lymphoblastoid cells (Epstein-Barr virus-transformed B cells) or by mixing 106 mononuclear cells with 10 jg phytohemagglutinin (PHA-P; Difco)/ml in each well of a 48-well cluster plate (Costar). After 8-11 d, cells were washed and plated in 2-ml wells of 24-well cluster plates at a concentration of 2-4 X 105 cells/ml. Cultures were passaged every 3-4 d or whenever viable cell concentration (determined by trypan blue exclusion) reached >8 x 105 cells/ml. Cultures were terminated when they showed no proliferative response to irradiated lymphoblastoid cells or when there were no viable cells present in the entire visual field of the hemocytometer. Once transferred to the 2-ml wells, cells were continuously exposed to 25 U of recombinant interleukin-2 (Amgen)/ml in RPMI (Irvine Scientific) supplemented with 10%-20% FCS, 2 mM glutamine, and 1 mM Hepes or in AIM V1', a DMEM/F-12 basal medium containing purified human albumin, transferrin, and recombinant insulin (Gibco), supplemented with 25% Ex-cyte (an aqueous mixture of lipoprotein, cholesterol, phospholipids, and fatty acids [Miles Diagnostics]). At each cell passage, the number of population doublings (PDs) was calculated according to the following formula: PD = In(final no. of viable cells/initial no. of cells)/ln 2. Isolation of Cells PBLs (including 15% monocytes) were isolated using Ficoll-Hypaque gradient centrifugation (Boyum 1968). Purification of human B and T cells (Gutierrez et al. 1979) and neutrophils (Dooley et al. 1982) was performed on a discontinuous percoll gradient. Densities were monitored using Density Marker Beads (Pharmacia). T-cell identification was confirmed by erythrocyte rosette testing, essentially as described by Gutierrez et al. (1979), and by Wright and Giemsa staining. All cell populations consisted of >98% viable cells, as judged by trypan blue exclusion. Isolation of DNA Cells were washed three times in PBS, and the pellet was resuspended at a density of 106_107 cells/ml in proteinase K digestion buffer (100 mM NaCl, 10 mM Tris pH 8,5 mM EDTA, and 0.5% SDS) containing 0.1 mg proteinase K/ml. The lysate was incubated at 48°C overnight and then was extracted twice with phenol/ chloroform/isoamyl alcohol (25:24:1 [v/v/v]) and once with chloroform. Nucleic acid was precipitated with 95% ethanol and was dissolved in TE (10 mM Tris

Telomere Loss in Aging and Trisomy 21 and 1 mM EDTA, pH 8). In some experiments, RNA was first degraded with pancreatic RNase A, but this had no effect on measurements of telomeric DNA. Analysis of Telomeric DNA DNA (10 jig) was digested with Hinfl and RsaI (BRL)

(20 U each), reextracted as described above, precipitated with 95% ethanol, washed with 70% ethanol, dissolved in 50 pJ TE, and quantified by fluorometry. One microgram of digested DNA was resolved by electrophoresis in 0.5% [w/v] agarose gels poured on Gel Bound (FMC Bioproducts) for -700 V-h. Gels were dried at 60'C for 30 min, denatured, neutralized, and probed with 5' end-labeled 32P-(C3TA2)3, as described by Allsopp et al. (1992). Autoradiograms exposed within the linear range of signal response were scanned with a Hoefer densitometer. The signal was digitized and subdivided into 1-kbp intervals between 2 kbp to 21 kbp, for calculation of the mean TRF length (L) by using the formula L = X(ODi Li)/XODi, where ODi = integrated signal in interval i and Li = TRF length at the midpoint of interval i. Results

When measured as a function of donor age, mean TRF length in PBLs of 119 unrelated normal individuals (age 0-107 years) declined at a rate of 41 ± 2.6 bp/year (P < .00005; r = .83) (fig. 1). This rate of TRF loss for PBLs is close to that previously found for peripheral blood cells by Hastie et al. (1990). When our data were separated according to gender, it was noticed that males lost telomeric DNA at a rate slightly faster than that of females (50 ± 4.2 vs. 40 ± 3.6 bp/year, respectively), but this difference did not reach statistical significance (P = .1). The 18 centenarians (age 99107 years) among our population of normal individuals had a mean TRF length of 5.28 ± 0.4 kbp (fig. 1). TRF length in these long-lived individuals is predicted by extrapolation of the line for individuals 0-80 years of age, suggesting that centenarians did not have an unusually long initial telomere length or an unusually slow rate of telomere loss. Interestingly, the SD of mean TRF values for the centenarians (0.4 kbp) was much smaller than that for other age groups (- 1 kbp). It is possible that this represents selection of a more homogeneous population of cells with age or that there exists some mechanism for homogenization of telomere length with time. However, it is also possible that the group of centenarians was less genetically diverse than the younger populations in our study.

663 Mean TRF length was also analyzed in PBLs of 21 DS individuals (age 2-45 years), and the rate of loss was compared with that of 68 age-matched controls (age 0-43 years). Cells from DS patients showed a significantly greater rate of telomere loss (133 ± 15 bp/year vs. 41 ± 7.7 bp/year, respectively; P < .0005) (fig. 2). The rate of telomere loss in DS individuals <24 years old was similar to that for all DS individuals and was still significantly increased, compared with that in agematched normals (P < .001). To test whether TRF length in the major leukocyte cell populations differed significantly, we isolated B and T lymphocytes and neutrophils from young and old normal donors and from young and old DS patients. All subpopulations of cells had similar TRF distributions and similar decreases in TRF size with donor age (fig. 3), suggesting that shifts in the relative frequency of these different major cell types with age does not account for the loss of telomeric DNA. We have not analyzed subpopulations of cells within these lineages. To determine the rate of telomere loss as a function of cell doublings, normal lymphocytes from four individuals were cultured in vitro, and mean TRF lengths were measured at several PD levels (fig. 4). Mean TRF length decreased 120 ± 35 bp/PD in these strains, which is within the range observed for other human somatic cell types (Harley et al. 1990; Allsopp et al. 1992; Counter et al. 1992). The relatively large variation in initial TRF reflected in the y-intercept (8.6 ± 2.0 kbp) was indicative of the variation seen in TRF of PBLs from unrelated donors (fig. 1, right). Mean TRF length at senescence for four of the cell strains shown here, which achieved >20 PDs in vitro, was 5.5 ± 0.8 kbp (fig. 4, right). The observed TRF values in vivo for PBLs of centenarians (5.3 ± 0.4 kbp) and old DS patients (4.9 ± 0.6 kbp) were not significantly different from this value, suggesting that a fraction of the cells from these individuals were at the limit of their replicative capacity.

Discussion

Peripheral blood lymphocytes, like fibroblasts and other replicating somatic cells, have a limited division capacity (Effros et al. 1990). Thus, our results showing that telomeres in PBLs from normal individuals shorten during aging in vivo and in vitro (a) extend similar observations on human fibroblasts (Harley et al. 1990; Allsopp et al. 1992) and (b) support the hypothesis that telomere loss may be involved in replicative senescence.

664 laneil: AGE:

Vaziri et al. 1 S

2

3

a 214

4 5

6 7 8 9 10 11 12 13 14 IS 16 17 18 19 20

M9

a i9 ia im iN

-O

ios isa 1

99 io

isi go jo lM

i}

0s m

21_*p _ l 12 _

8 61_ 5

I Hi

z

4.

LJ

z

wL

r= 0.83 m- -41 bp/yr

1

40

60

80

104

DONOR AGE (YEARS' Figure I Loss of telomeric DNA as a function of donor age in PBLs. Seven milliliters of peripheral blood was obtained from 119 donors (age 0-107 years), and genomic DNA from PBLs was isolated and digested with restriction enzymes for Southern analysis, as described in Subjects, Material, and Methods. Left, Autoradiogram showing a subset of the normal individuals, including most of the centenarians. Donor age and size markers (in kbp) are indicated. In this experiment, sperm DNA (lane S), which has long TRFs, was included as a control. The bands appearing at '-2, 1.6, and 1 kbp are nontelomeric repetitive sequences that hybridize to the probe, even at high stringency. They serve as a convenient control for DNA integrity and quality of the gel. Right, Mean TRF length from quantitative analysis of autoradiograms, as described in Subjects, Material, and Methods, plotted as a function of donor age. The slope (-41 ± 2.6 bp/year) of the linear regression line is significantly different from 0 (P < .00005). -

We also found that, in DS, the rate of telomere loss in PBLs as a function of donor age was three times higher than that in age-matched normal donors. Since DS is characterized by immune dysfunction, including thymus abnormalities, derangements of both lymphoid and myeloid cell compartments (reviewed by Ugazio et al. 1990), and premature T-cell aging (Rabinowe et al. 1989), the accelerated loss of telomeres in PBLs could reflect a generalized early senescence of immune cells in these individuals. It is possible that the decrease in telomere length with age in vivo reflects, in part, changes in the subpopulations of cells, which are known to occur in both normal individuals (reviewed in Thoman and Weigle 1989) and DS patients (Cossarizza et al. 1991). However, the absolute and relative changes for the major T-cell subpopulations are relatively small in normal individuals (Murasko and Goonewardene 1990). In addition, our measurements in B and T cells, as well as neutrophils, showed similar TRF-length distributions between the subpopulations in young and old normal and DS individuals and confirmed the loss of telomeric

DNA observed in total PBLs (fig. 3). These data indicate that aging of the lymphoid and myeloid lineages is characterized by similar rates of telomeric DNA loss. Although we have not yet been able to further subdivide these subpopulations for telomere analysis, it seems unlikely that differences in the major subpopulations within each lineage (e.g., CD4 vs. CD8 lymphocytes) would be greater than the differences in the major leukocyte subdivisions and that these differences could account for the systematic and parallel loss that we see with age in both myeloid and lymphoid lineages. Finally, within cultured fibroblasts, variation in intitial TRF length exists between clones from a single mass culture, but all clones lose telomeric DNA during replicative aging (R. C. Allsopp and C. B. Harley, unpublished data). Thus, telomere loss due to cell division with age is the simplest interpretation of our data. If we accept that telomere length is a biomarker of the replicative history of normal somatic cells (Harley et al. 1990; Hastie et al. 1990; Harley 1991; Levy et al. 1992) and of their future replicative potential (Allsopp et al. 1992), then the increased rate of telomere loss in

Telomere Loss in Aging and Trisomy 21

665

15-

aCL

m Y

I

1

5Li inrDrinmseldom 21 (n=21) r -0.89, m= -133 bp/y (n=6), r= -0.56, m= -43 bp/y mu-i NormaI 0

0

10

20

30

40

50

DONOR AGE (YEARS) Figure 2 Accelerated telomere loss in DS patients. Genomic DNA isolated from PBLs of DS patients was analyzed according to the method described in fig. 1 and Subjects, Material, and Methods. Mean TRF length is shown as a function of donor age, for DS patients (U) and age-matched controls (U). The slopes of the linear regression lines (-133 ± 15 bp/year for trisomy 21, vs. -43 ± 7.7 bp/year for normal) are significantly different (P < .0005). Two DS samples, at ages 24 and 47 years, (E), were analyzed a second time, 2 years after the first analysis. These points indicate the reproducibility of data from separate experiments but were not included in the statistical analysis.

Since lymphocytes are continuously renewed, telomere shortening during aging in vivo suggests that telomerase is either absent or present in low levels in progenitor blast cells and primitive stem cells. If there were no telomerase activity, then the rate of telomere loss with age in circulating cells (-40 bp/year) (see fig. 1 and Hastie et al. 1990) might reflect the average rate of telomere loss in the bone marrow. If we further assume that the measured rate of telomere loss in lymphocytes in vitro (- 100 bp/cell doubling) (fig. 4) applies to stem cells as well, we could estimate that the stem-cell population undergoes -0.4 doublings/year, on average. This is derived from the ratio (40 bp/year)/(100 bp/ doubling). The comparable values for fibroblasts are (15 bp/year)/(75 bp/cell doubling) 0.2 cell doublings/year (Allsopp et al. 1992). The estimate of 0.4 cell doublings/year for hematopoietic stem cells may be low, since partial telomerase activity or other factors could reduce the rate of telomere loss in vivo in stem cells. Direct measurements of telomere length and telomerase activity in cells at various points in the hematopoietic lineage, coupled with independent knowledge of cell kinetics, will allow us to assess the utility of telomere length as a biomarker of cell aging in vivo. , -- Normal -- r- Trisomic --26y 60y

2y

45y

P B T N P B T NI P B T N P B T N PBTNPBTN1iPBTNPBTNff. t.

21 -

PBLs from DS patients could reflect a higher turnover rate of these cells in vivo. There are several possible explanations for this. For example, immune cells in DS individuals could have reduced viability or abnormalities in maturation due to trisomy 21, which lead directly to increased cell turnover. However, it is also possible that the amount of telomere loss in PBLs from DS patients is greater per cell doubling than that in normal individuals. In yeast, altered expression of several genes leads to abnormal regulation of telomere length (e.g., see Lundblad and Szostak 1989; Conrad et al. 1990). If the expression of genes involved in telomere length regulation is altered due to trisomy 21, then the rate of telomere loss could increase in DS. Further work on telomere length and cell turnover and function both in vivo and in vitro in DS cells is required to determine both the mechanism that accounts for the higher rate of telomere loss in these individuals and whether telomere loss might play a causal role in immune system failure.

129-

76-

4-

2-

Loss of telomeric DNA in subpopulations of blood Figure 3 cells in DS patients and normal individuals. Cells were isolated according to the protocol described in Subjects, Material, and Methods, and genomic DNA was analyzed for TRF lengths, as described in fig. 1. P = PBLs; B = B cells; T = T cells; and N = neutrophils.

666 PDL:

10

15

2i 12-llAl l l 1I \

1- -l| ~PDL

22 25 28

1

ii 10--

9J

9~~~ 7~~~

-J

4

Vaziri et al.

7-

0

10M

30

Figure 4 Analysis of telomere loss during aging of lymphocytes in vitro. T lymphocytes were cultured for various lengths of time in vitro, and PDs were determined according to the protocol described in Subjects, Material, and Methods. Left, Autoradiogram of terminal restriction fragments from genomic DNA isolated and prepared for Southern analysis, as described in fig. 1. PDs of a typical T-lymphocyte culture are indicated above each lane; size markers (in kbp) are shown at the left. Right, Decrease in mean TRF length, as a function of PDs, for DNA from four normal individuals. Donor ages for these cells were not available. Mean TRF length at terminal passage from a fifth donor, for which multiple passages were not available, is also shown (5).

Premature senescence of the immune system in DS is possibly a major factor in the similarity of DS pathology and normal aging (Martin 1978). In support of this idea, lymphocytes of old DS patients and old normal individuals share several characteristics, including diminished response of T cells to activate and proliferate in response to antigen, low replicative capacity, and reduced B- and T-cell counts (Cossarizza et al. 1991; Franceschi et al. 1991). Our findings that telomere length in PBLs decreased faster in DS patients than in normal individuals and that the mean TRF length in centenarians and old DS patients in vivo was similar to that of senescent lymphocytes in vitro (-5 kbp) support and extend those observations. Moreover, our data suggest that replicative senescence within the lymphoid and myeloid lineages in vivo might contribute to the compromised immune system of both elderly individuals and DS patients.

Acknowledgments We thank the families of the DS patients for their cooperation and J. Waye for providing fetal blood samples. We also thank Alexy Olovnikov, Rich Allsopp, Edwin Chang, Silvia

Bacchetti, and Carol Greider for helpful discussions and Charles Epstein for critical comments. This work was supported by the Medical Research Council of Canada (support to C.B.H.) and by U. S. National Institutes of Health grants AG09383A (to C.B.H.) and AG05309 (to R.E.).

References Allshire RC, Dempster M, Hastie ND (1989) Human telomeres contain at least three types of G-rich repeats distributed non-randomly. Nucleic Acids Res 17:4611-4627 Allsopp RC, Vaziri H, Patterson C, Goldstein S, Futcher AB, Greider CW, Harley CB (1992) Telomere length predicts replicative capacity of human fibroblasts. Proc Natl Acad Sci USA 89:10114-10118 Bender MA, Preston RJ, Leonard RC, Pyatt BE, Gooch PC (1989) Chromosomal aberration and sister-chromatid exchange frequencies in peripheral blood lymphocytes of a large human population sample. Mutat Res 212:149-154 Benn PA (1976) Specific chromosome aberrations in senescent fibroblast cell lines derived from human embryos. Am J Hum Genet 28:465-473 Blackburn EH, Szostak JW (1984) The molecular structure of centromeres and telomeres. Annu Rev Biochem 53:163194 Boyum A (1968) Isolation of mononuclear cells and granulo-

Telomere Loss in Aging and Trisomy 21 cytes from human blood. Scan J Clin Lab Invest 21 Suppl 97:77-89 Conrad MN, Wright JH, Wolf AJ, Zakian VA (1990) RAP1 protein interacts with yeast telomeres in vivo: overproduction alters telomere structure and decreases chromsome stability. Cell 63:739-750 Cossarizza A, Ortolani C, Forti E, Montagnani G, Paganelli R, Zannotti M, Marini M, et al (1991) Age-related expansion of functionally inefficient cells with markers of natural killer activity in Down's syndrome. Blood 77:12631270 Counter CM, Avilion AA, LeFeuvre CE, Stewart NG, Greider CW, Harley CB, Bacchetti S (1992) Telomere shortening associated with chromosome instability is arrested in immortal cells which express telomerase activity. EMBO J 11:1921-1929 Dooley DC, Simpson JF, Meryman HT (1982) Isolation of large numbers of fully viable human neutrophils: a preparative technique using percoll density gradient centrifugation. Exp Hematol 10:591-599 Effros RB, Perillo NL, Bhuta S, Walford RL (1990) In vitro studies of human T lymphocyte senescence. In: Finch CE, Johnson TE (eds) Molecular biology of aging. Alan R Liss, New York, pp 265-279 Franceschi C, Monti D, Cossarizza A, Fagnoni F, Passeri G, Sansoni P (1991) Aging, longevity, and cancer: studies in Down's syndrome and centenarians. Ann N Y Acad Sci 621:428-440 Greider CW, Blackburn EH (1985) Identification of a specific telomere terminal transferase activity in Tetrahymena extracts. Cell 43:405-413 Gutierrez C, Bernabe RR, Vega J, Kreisler M (1979) Purification of human T and B cells by a discontinuous density gradient of percoll. J Immunol Methods 29:57-63 Harley CB (1991) Telomere loss: mitotic clock or genetic time bomb? Mutat Res 256:271-282 Harley CB, Futcher AB, Greider CW (1990) Telomeres shorten during ageing of human fibroblasts. Nature 345:458-460 Hastie ND, Dempster M, Dunlop MG, Thompson AM, Green DK, Allshire RC (1990) Telomere reduction in human colorectal carcinoma and with ageing. Nature 346:866-868

667 Levy MZ, Allsopp RC, Futcher AB, Greider CW, Harley CB (1992) Telomere end-replication problem and cell aging. J Mol Biol 225:951-960 Lundblad V, Szostak JW (1989) A mutant with a defect in telomere elongation leads to senescence in yeast. Cell 57:633-643 Martin GM (1978) Genetic syndromes in man with potential relevance to the pathobiology of aging. Birth Defects 14:539 Morin GB (1989) The human telomere terminal transferase enzyme is a ribonucleoprotein that synthesizes TTAGGG repeats. Cell 59:521-529 Moyzis RK, Buckingham JM, Cram LS, Dani M, Deaven LL, Jones MD, Meyne J, et al (1988) A highly conserved repetitive DNA sequence, (TTAGGG)n, present at the telomeres of human chromosomes. Proc Natl Acad Sci USA 85:6622-6626 Murasko DM, Goonewardene IM (1990) T-cell function in aging: mechanisms of decline. Annu Rev Gerontol Geriatr 10:71-96 Olovnikov AM (1971) Principles of marginotomy in template synthesis of polynucleotides. Dokl Akad Nauk SSSR 201:1496-1499 (1973) A theory of marginotomy. J Theor Biol 41:181-190 Rabinowe SL, Rubin IL, George KL, Adri MN, Eisenbarth GS (1989) Trisomy 21 (Down's syndrome): autoimmunity, aging and monoclonal antibody-defined T-cell abnormalities. J Autoimmun 2:25-30 Saksela E, Moorhead PS (1963) Aneuploidy in the degenerative phase of serial cultivation of human cell strains. Proc Natl Acad Sci USA 50:390-395 Sherwood SW, Rush D, Ellsworth JL, Schimke RT (1989) Defining cellular senescence in IMR-90 cells: a flow cytometric analysis. Proc Natl Acad Sci USA 85:9086-9090 Thoman ML, Weigle WO (1989) The cellular and subcellular bases of immunosenescence. Adv Immunol 46:221-261 Ugazio AG, Maccario R, Notarangelo LD, Burgio GR (1990) Immunology of Down syndrome: a review. Am J Med Genet Suppl 7:204-212 Watson JD (1972) Origin of concatameric T4 DNA. Nature N Biol 239:197-201