Transposable Elements as a Factor in the Aging of Drosophila melanogaster CHRISTOPHER J. I. DRIVER" AND STEPHEN W. McKECHNIEb aDepartment of Science Deakin University-Rusden Clayton, Victoria 3168, Australia bDepartment of Genetics and Development Biology Monash University Clayton, Victoria 3168, Australia
INTRODUCTION A model of aging based on the possibility that somatic DNA may accumulate mutations has been discussed for several decades. The model has significant explanatory power, but many authors have pointed out substantial problems,'*2 and hence various modifications to the model have been discussed.l*2Since the elucidation of the nature of transposable elements, it has become apparent that most spontaneous mutations in the germ line occur as a result of transposable element a ~ t i o n .It~ thus seems reasonable to consider a role for transposable elements in somatic mutation and aging. This possibility has been raised by other authors. Kirkwood* suggested a role for transposable elements but presented no data. Murray4 discusses a theoretical model for transposable elements and aging of cells in culture. Recently P elements that were made somatically active have been shown to shorten life span. This life span shortening was proportional to the number of copies of the element in the g e n ~ m eFinally, .~ a series of studies on a senescence-related phenomenon in the filamentous fungus Podospora indicates that aging was associated with activity of a mobile intron that was thought to cause damage by reinsertion. This plasmid behaved as if it were a transposable element.6 Various observations suggest that some transposable elements are active in somatic tissues. Thus extrachromasomal circular DNA corresponding to transposable elements and viruslike particles containing transposable-element nucleic acid have been observed in somatic tissues of many species.'** Phenotypic changes in somatic tissues such as would be observed from the excision of an element have been observed in both D r ~ s o p h i l a ' - and ~ ~ in the nematode Caenorhabditis elegans.I3 Several cancers have been shown to be caused by the insertion of an transposable element in the region of an 0nc0gene.l~However, the frequency of somatic activity of transposable elements and the mechanisms involved in triggering activity remain to be established. The species best suited to test this hypothesis is Drosophila rnelanogaster, since most of its transposable elements have probably now been found.) In addition its age phenotype responds to dietary changes in the same way that that of other species does.Is For example, substantial restriction of dietary intake by dilution of the medium produced a decrease in life span.15 However, with small dilutions of the medium, a modest increase in life span occurred." Furthermore, aging can be accelerated by a high-fat diet in this species.15 83
84
ANNALS NEW YORK ACADEMY OF SCIENCES
Chromosome number appears to have no effect on aging in many species, including Drosophila. 16~17Male Drosophila contain a single X-chromosome, which accounts for about 23% of the haploid genome; whereas females contain two active X-chromosomes. Consequently, males are very much more sensitive to loss-offunction X-linked mutations. Both X-chromosomesare active in the adult female. ” Nevertheless, females and males have very similar life spans.I9 It is of interest to determine whether or not both sexes respond in the same degree to somatic activity of transposable elements. In this report we first examine two transposable elements for somatic activity. Second, we artifically activate somatic transposition of a P element and measure the effects on life span. This effect is compared between the sexes. Finally, we make use of a strain harboring P element DNA that shows eye color mosaicism and reversion to wild phenotype (red eyes) as a result of somatic and germ line transposition. In this strain, we test the effect of a high-fat diet that accelerates aging on the frequency of transposition, as measured by eye color mosaicism and red eyes.
METHODS Strains of Drosophila melanogaster. The wild-type flies used were Canton-S and the F1 of two inbred strains, OD4 and CC3, derived from females captured at Chateau Tahbilk, in Victoria, Australia.’O This second set of flies was abbreviated OD4/CC3. Birm 2 : ry506(abbreviated Birm-2) contains the second chromosome of the Birmingham strain, with 17 nonautonomous P elements.” PA2,3(99B) contains a P element with an in-phase deletion between the second and third intron in the gene for the transposase. The transposase is somatically active and fails to act on its own element.21To obtain viable flies that were the first-generation cross of PA2,3(99B) and Birm-2 strains, development was at 18”C, so that the P element would not be somatically active.*’ For comparison the parent strains were also raised at 18°C. Mosaic strains were constructed from the CaSpeR 3 1.4 strain by crossing with PA2,3(99B) and breeding from mosaic-eyed females. The strain CaSpeR 3 1.4 contains the CaSpeR element on the third chromosome.22This element contains the wild-type white gene and is bounded by P element termini. In this strain the color produced by the activity of the gene is reduced by a position effect. Full expression is only possible when the gene is transposed into a region where suppression of gene activity is less. When a somatically active transposase is provided in trans, flies are produced in which patches of deep color appear on the pale background. In addition white patches are produced. Flies showing wildtype eye color and also white eyes are observed.” Plasmids Used. cDm2055 for copia,’ cDm2042 for 412,23and sAF2 for Adh.24 Diets. Control and high-fat media were as previously de~cribed.’~ Aging Protocol. Newly eclosed adults were kept in batches of 20 in vials of fresh medium and transferred weekly. Deaths were scored at the time of transfer. Deaths were considered to occur midway between times of counting. DNA Preparation and Southern Blotting. Genomic DNA was extracted using the method of McGinnis et a1.,26from batches of 150 files after snap freezing in liquid nitrogen and grinding to a fine powder in a mortar and pestle. DNA from agarose electrophoresis gels was Southern blotted onto nylon membranes using standard procedure^.^' Probe DNA was prepared from purified plasmids DNA and w3’P ATP via a random hexanucleotide primers method (Oligo-labeling kit from
85
DRIVER & McKECHNIE: TRANSPOSABLE ELEMENTS
Bresatec), and probing was by standard conditions of hybridization and X-ray film e x p o s ~ r e . ~Autoradiographic ’ gels were scanned and quantified with a Hoeffer GS3000 densitometer associated with Apple software.
RESULTS Somatic Activity of Transposable Elements Using the two types of wild-type flies, Southern blots were carried out on genomic DNA, undigested with restriction enzymes, from young (3-day-old) and old (45-day-old) males and probed with plasmids containing either 412 or copia
“r I MEAN PERCENT INCREASE IN BAND INTENSITY IN ADULT DNA ,oo
412
COPIA
FIGURE 1. Increase in transposable element band intensity of unrestricted DNA, determined with males of two wild-type strains (hatched column, Canton-S; open columns, OD4/ CC3). Paired adjacent DNA spots for young and old flies from the same cohort were compared on the same membrane and the ratio of intensities determined. To correct for variation in the amount of DNA on the membrane, the spots were reprobed with Adh. The number of such pairs of old and young DNA spots scanned was 4-6. Error bars on the graph are SEM. The horizontal line through the column marks the expected ratio for no change with age.
sequences. DNA from corresponding young and old flies were run in adjacent wells for pairwise comparison. To correct for variation in amount of DNA bound, each membrane was reprobed with a plasmid containing the Adh sequence. The autoradiographic intensity paralleled the ethidium bromide fluorescence, and no additional band that might be extrachromosomal circular DNA was visible. The mean ratios of band intensities for DNA from old and young flies are shown in FIGURE1. Element 412 provided evidence of an increase in the band intensity with age in the Canton-S strains. The OD4/CC3 flies did not show evidence of an increase for 412. However, in the OD4/CC3 flies, for copia, a significant increase with age was obtained. When the DNA was digested with Eco R1 or Hind 111, electrophoresed in agarose gel, Southern blotted, and probed with 412 or copia, no new bands appeared. Such a result would be expected if new DNA sequences
ANNALS NEW YORK ACADEMY OF SCIENCES
86
FEMALES MOSAICS DOUBLE MOSAICS RED EYES WHITE EYES
MOSAICS DOUBLE MOSAICS RED EYES WHITE EYES
0
5
1
0
1
5
2
0
2
5
r
)
3
5
4
0
4
5
5
0
PERCENTAGE
FIGURE 2. The effect of diet on the percentage of eye phenotypes in the mosaic eye strain. The numbers of red-eyed flies, white-eyed flies, and mosaics are expressed as percent of total flies. The frequency of flies mosaic in both eyes (double mosaics) is expressed as percent of flies mosaic in at least one eye. The number in each group was between 150-300. Significant differences in ratios between dietary treatments are indicated by asterisks ( p < 0.01, difference between population proportions28).No errors are quoted, as this would be inappropriate for this statistical test. Open bars, low-fat diet; shaded bars, high-fat diet.
arising from insertion of the transposable elements is dispersed throughout the genome. With the mosaic-eyed strain, the effect of diet on the appearance of phenotypes expected to result from transposition is shown in FIGURE 2. The high-fat diet significantly elevated the incidence of mosaics in both sexes. Since the effect of diet on this frequency could be the result of a differential effect on the survival of genotypes, a second measure was also used: the ratio of mosaicism in the second eye of flies known to be mosaic in one eye. The ratio of double-eye mosaics to total mosaics was also higher on the high-fat than on the low-fat diet in both sexes. It was also noted that the frequency of occurrence of nonmosaic red eyes was increased in the high-fat diet, whereas the frequency of appearance of all-white eyes was decreased on high-fat diets (FIG.2).
P Element Transposition and Life Span To test whether a reduction in life span in the F1 strain containing the fully active element occurs, the life span of the strain was compared with the life spans of the parent strains. When two inbred strains are crossed, one would expect the
DRIVER & McKECHNIE: TRANSPOSABLE ELEMENTS
87
life span of the FI would be equal to, or greater than, the man of the parental strain, depending on whether heterosis O C C U ~ S . ' ~ Life spans were measured at 18°C for the strain carrying the PA2,3 element, the strain carrying the nonautonomous elements (Birm-2), and the F1 strain carrying both. The life span of the F1 males was indistinguishable from that of the mean of the parental strains, and that of the F1 females was greater than the mean of the parental strains (FIG.3), suggesting some heterosis in females. These measurements were repeated at 25°C. A significant reduction of life spans with respect to the mean of the parental strains was observed for both sexes (FIG.3).
DISCUSSION Is there a complement of transposable elements that are somatically active in normal adult tissues and that contribute to the aging processes? As a first step towards answering this question, we have measured an increase in DNA band intensity for the elements 412 and copia, this latter element being measured in the strain only. The detected increase suggests that genomic tissue from old flies contains more copies of these elements than the equivalent DNA from young flies.
UFESPANS AT 25' C P dona 2.3 Birmingham2 MEAN F1 ESPANS AT lE0C P dona 2,3
Birmingham-? MEAN F1 0
20
40
80
80
100
laD
UFESPAN IN DAYS
FIGURE 3. Top: Comparison of mean life spans at 25°C of the Birmingham-2 strain, the PA2,3 strain, and flies from their FI. Numbers of flies in each group: PA2,3: 161 males, 249 females; Birmingham-2: 129 males, 191 females; 241 FI males, 245 F1 females. Bottom: Comparison of mean life spans of the same strains measured at 18°C. Numbers of flies in each group: PA2.3: 181 males, 216 females; Birmingham-2: 55 males, 86 females; 154 FI males, 206 FI females. Open bars. females; harchedbars, males. Error bars, SEM. For each sex. significant differences between mean FI life span and the means of the parental strains are indicated by asterisks ( p > 0.01; Student's r-test); the difference that is not significant is indicated NS.
88
ANNALS NEW YORK ACADEMY OF SCIENCES
An appearance of transposable elements at new sites in the genome has not been demonstrated. To d o so would be technically difficult, since each reinsertion would be likely to be in a different position and produce a unique DNA sequence. We have examined only two transposable elements in one strain and one in a second. For each element we found that one strain showed an increase in intensity of the band detected by the DNA probe for that element. We have interpreted this as an increase in the relative number of that transposable element in the DNA extracted from the old flies. Since there are more than 40 types of transposable element^,^ it would be reasonable to expect that other elements will be found to transpose in this species during aging. However, several factors should be considered. Previous reports on somatically active transposable elements with mariner in Drosophilu9-" and the Tcl element in C. elegansI2 indicate that there is strain-specific variation in activity. The data presented in this paper suggests the same is true for 412. Such variation is likely to be a general property of such elements. The elements that we chose to investigate are of a type such that transposition in postmitotic tissue would be expected to produce an increase in copy number.) Although most transposable elements in Drosophilu would be expected to have this p r ~ p e r t y it , ~ is not likely to be universal. Accordingly, different techniques would be needed to investigate the somatic activity of the full complement of elements. Although it has been known for some time that diet effects aging in many species,l**the exact cause of this effect has been elusive. Data presented here suggest that germ line or somatic transposition resulting in reinsertion of a P element is promoted by a high fat diet. However data also suggests that the rate of excision, leading to white eyes, was decreased by the high-fat diet. Excision of a P element is unlikely to be a deleterious event unless imprecise excision causes damage to a gene that was functional when the P element was present. Insertion has a higher probability of inactivating a gene.3 These results suggest that potentially damaging transposition events are promoted by a high-fat diet. Our data is consistent with the hypothesis that transposition of a P element is increased by a diet known to increase the rate of aging. This effect may be common to many elements if diet modified a common factor, such as the state of the chromosome. In that event, changes in damage induced by transposable elements may contribute to the effect of diet on the age phenotype. At 25" in these constructs Pelements are reported to be active in transposition.21 We were able to verify this. When the FI flies carrying a somatically active P element were raised at 25°C only a few adults were produced. These were of low viability and had a high frequency of deformities, as reported by Robertson et Accordingly, to obtain viable nondeformed adults we raised the flies at 18°C. The life spans of these flies, subsequently measured at 25"C,showed a significant reduction, with respect to the mean of the adult strains, in both sexes and to about the same degree. No such reduction was observed when the life spans were measured at the lower temperature, at which transposition does not occur. The life span of male Fls was indistinguishable from that of the mean of the parental strains, and that of the female was somewhat greater than the mean of the parental life spans, indicating some heterosis. Thus the life shortening effect of somatic transposition of P elements was not less in females than in males. One of the major obstacles to the ready acceptance of the somatic mutation model of aging is that in situations where the chromosome number differs, the aging rates are similar. The lack of a shorter life span in male Drosophilu, compared to females, is one such example. The observation that somatically active elements also do not produce more damage in males may provide an explanation for this
DRIVER & McKECHNIE: TRANSPOSABLE ELEMENTS
89
paradox. Alternative theoretical reasons for the absence of an effect on transposable element damage are provided by the exponential growth model of Murray4and the dominant/codominant model of Morley .I7 Additional alternative explanations should also be considered. For example, the element may cause somatic mutation in a limited number of “hot spots,” which as a result of natural selection are not on the more sensitive X chromosome. We are currently investigating this further. New insights provided by this idea are that in higher organisms, including humans, somatic activity of transposable elements may cause aspects of the age phenotype. Transposition may be responsive to control by identifying and changing environmental triggers and suppressors of this activity. Pharmacological support may be possible for tissues that are deficient in the product of genes that are identified as hot spots for transposable element mutagenesis. New approaches to the genetic analysis of age-associated pathologies may be possible. Finally, a possible role for transposable elements may be considered in the expression of such rare adult-onset genetic conditions as Huntington’s chorea.
CONCLUSION We have presented data indicating that transposable elements may be somatically active in adult tissues. The somatic mutation thereby produced may contribute to the aging phenotype. Somatic mutation by transposable elements may be able to explain the chromosome number paradox and appears to be able to contribute to the link between diet and aging. This hypothesis deserves further consideration. It can reasonably be expected that insights may arise that will be of value theoretically and clinically. The model also has the advantage of being readily testable in other organisms using current tools.
SUMMARY We have considered the hypothesis that transposable elements may contribute to the aging process through somatic mutation. We have presented evidence to suggest that at least two elements, Copia and 412, are capable of somatic activity in adult Drosophila tissue. A strain harboring a third transposable element, P, was produced that showed eye color mosaicism and reversion to wild phenotype (red eyes) as a result of somatic and germ line transposition. A high-fat diet, known to accelerate aging, increased the frequency of eye color mosaicism and red eyes. We induced life span shortening by artificially activating somatic transposition of P elements, and the extent of reduction in life span was similar in both sexes. These data are consistent with the notion that some aspects of the age phenotype may be caused by mutational activity of transposable elements in somatic tissues. The hypothesis is readily tested in other organisms, including humans. It offers new dimensions in the understanding and management of age-associated changes.
ACKNOWLEDGMENTS The Birm-2 and the Casper strains were a gift from Dr. John Sved, and the PA2,3(99B) strain was a gift from Dr. John Oakeshott. We would like to thank both gentlemen for these strains.
90
ANNALS NEW YORK ACADEMY OF SCIENCES REFERENCES
I . HAYFLICK, L. 1985. Theories of biological aging. Exp. Gerontol. 20: 145-159. T. B. L. 1989. DNA, mutations and aging. Mutat. Res. 219 1-7. 2. KIRKWOOD, D. J. & D. H. FAWCETT. 1986. Transposable Elements in Drosophila 3. FINNEGAN, melanogaster. Oxford Surveys on Eukaryotic Genes, Vol 3. Academic Press. London. 4. MURREY, V. 1990. Hypothesis: Are transposons a cause of ageing?’ Mutat. Res. 237: 59-63. 5 . WOODRUFF,R. 1991. Mutation and evolution: Lifespan reduction by transposons and premeiotic clusters. 38th Meeting of the Genetics Society of Australia, Monash University. Abstr. 52. H. D. 1990. Molecular analysis of ageing processes in fungi. Mutat. Res. 6. OSIEWACS, 237: 1-8. J. W. 1990. Extrachromosomal circular DNA and genomic plasticity in 7. GAUBATZ, eukaryotic cells. Mutat. Res. 237: 271-292. 8. EMORI,Y., T. SHIBA,S. KANAYA,S. INOUYE,S. YUKI& K. SAIGO.1985. The nucleotide sequence of Copia and Copia-related RNA in Drosophilu virus-like particles. Nature 315: 773-776. A. KOGA& D. HARTL.1991. Introductionofthe transposable 9. GARZA,D., M. MEDHORA, element mariner into the germline of Drosophila melanogaster. Genetics 128(2): 303-310. 10. MEDHORA,M., K. MARUYAMA & D. HARTL.1991. Molecular and functional analysis of the mariner mutator element Mos I in Drosophila. Genetics 128(2): 311-318. K. & D. HARTL.1991. Evolution of the mariner mutator element in I I . MARUYAMA, Drosophika species. Genetics 128(2): 3 19-329. 1991. Somatic mutation in the wings of Drosophilu 12. GETZ, C. & N. VAN SCHAIK. rnelanogaster females dysgenic due to P elements when reared at 29°C. Mutat. Res. 248: 187-194. Y. & L. YESNER.1987. High frequency excision of transposable elements 13. EMMONS, Tcl in the nematode Caenorhabditis elegans is limited to somatic cells. Cell 3 6 599-605. E., 0. DREAZEN, A. KLAR.1. G. RECHAV,D. RAM,J. B. COHEN& D. 14. CANAANI, GIVOL.1983. Activation of the c-mos oncogene by insertion of an endogenous intracisternal A-particle genome. Proc. Natl. Acad. Sci. USA 8 0 71 18-7122. 15. DRIVER,C. J. I., G. COSOPODIOTIS, R. WALLIS& G. ETTERSHANK. 1986. Is a fat metabolite the major diet dependant accelerator of ageing? Exp. Gerontol. 21: 497-507. 16. LAMB,M. J. 1965. The Effects of X-irradiation on the longevity of triploid and diploid females Drosophila rnelanogaster. Exp. Gerontol. 1: 181-184. 17. MORLEY,A. A. 1982. Is ageing the result of dominant and codominant mutations? J. Theoret. Biol. 9 8 469-474. 18. LUCCESI,J. C. & J. E. MANNING.1987. Gene dosage compensation in Drosophika melanogaster. Adv. Genet. 24: 371-429. & G. MOKRYNSKI. 1985. Aging in Drosophila. In 19. BAKER,G. T. 111, M. JACOBSON Handbook of Cell Biology and Aging. V. Cristofalo, Ed.: 51 1-578. CRC Publishers. Boca Raton, FL. P. W. HEINSTSTRA & M. J. PYKA.1991. Heritable 20. GEER,B. W.. S. W. MCKECHNIE, variation in ethanol tolerance and its association with biochemical traits in Drosophila melanogaster. Evolution 45: 1107-1 109. H. M., C. R. PRESTON,R. W. PHILLIS,D. M. JOHNSTON-SCLITZ, W. K. 21. ROBERTSON, BENZ& W. R. ENGELLS.1988. A stable genomic source of P element transposase in Drosophila rnelanogaster. Genetics 118 461-470. 22. PIROTTA,V. 1988. Vectors for P element transformation in Drosophika. In Vectors: A Survey of Molecular Cloning Vectors and Their Uses. R. L. Rodriguiz & D. T. Denhart, Eds.: 437-456. Butterworths. Boston, MA. S. ISHIMARU & K. SAIGI. 1986. Nucleotide sequencecharacterisa23. YAKI,S., S. INOUYI, tion of a Drosophila retrotransposon, 412. Eur. J. Biochem. 158 403-410.
DRIVER & McKECHNIE: TRANSPOSABLE ELEMENTS
91
24. GOLDBERG, D. A. 1980. Isolation and partial characterisation of the Drosophila melanoguster Adh gene. Proc. Natl. Acad. Sci. USA. 77(10): 5794-5798. C. J. I. & G. COSOPODIOTIS. 1978. The effect of a high fat diet on longevity 25. DRIVER, of Drosophilu mclanogastcr. Exp. Gerontol. 14: 95- 100. 26. MCGINNES, W . , A. W. SHERMOEN & S. K. BECKENDORF. 1983. A transposable element insertedjust 5’ to a Drosophila glue protein gene alters gene expression and chromatin structure. cell 34: 75-84. 27. SAMBROOK, J . , E. F. FRITSCH& T. MANIATIS. 1989. Molecular cloning: A laboratory manual, 2nd edit. Cold Spring Harbor Laboratory Press. Cold Spring Harbor, NY. W. 1983. Biostatistics: A foundation for analysis in the health sciences, 3rd 28. DANIEL, edit.: 192-193. Wiley and Sons. New York, NY.
INTRODUCTION A model of aging based on the possibility that somatic DNA may accumulate mutations has been discussed for several decades. The model has significant explanatory power, but many authors have pointed out substantial problems,'*2 and hence various modifications to the model have been discussed.l*2Since the elucidation of the nature of transposable elements, it has become apparent that most spontaneous mutations in the germ line occur as a result of transposable element a ~ t i o n .It~ thus seems reasonable to consider a role for transposable elements in somatic mutation and aging. This possibility has been raised by other authors. Kirkwood* suggested a role for transposable elements but presented no data. Murray4 discusses a theoretical model for transposable elements and aging of cells in culture. Recently P elements that were made somatically active have been shown to shorten life span. This life span shortening was proportional to the number of copies of the element in the g e n ~ m eFinally, .~ a series of studies on a senescence-related phenomenon in the filamentous fungus Podospora indicates that aging was associated with activity of a mobile intron that was thought to cause damage by reinsertion. This plasmid behaved as if it were a transposable element.6 Various observations suggest that some transposable elements are active in somatic tissues. Thus extrachromasomal circular DNA corresponding to transposable elements and viruslike particles containing transposable-element nucleic acid have been observed in somatic tissues of many species.'** Phenotypic changes in somatic tissues such as would be observed from the excision of an element have been observed in both D r ~ s o p h i l a ' - and ~ ~ in the nematode Caenorhabditis elegans.I3 Several cancers have been shown to be caused by the insertion of an transposable element in the region of an 0nc0gene.l~However, the frequency of somatic activity of transposable elements and the mechanisms involved in triggering activity remain to be established. The species best suited to test this hypothesis is Drosophila rnelanogaster, since most of its transposable elements have probably now been found.) In addition its age phenotype responds to dietary changes in the same way that that of other species does.Is For example, substantial restriction of dietary intake by dilution of the medium produced a decrease in life span.15 However, with small dilutions of the medium, a modest increase in life span occurred." Furthermore, aging can be accelerated by a high-fat diet in this species.15 83
84
ANNALS NEW YORK ACADEMY OF SCIENCES
Chromosome number appears to have no effect on aging in many species, including Drosophila. 16~17Male Drosophila contain a single X-chromosome, which accounts for about 23% of the haploid genome; whereas females contain two active X-chromosomes. Consequently, males are very much more sensitive to loss-offunction X-linked mutations. Both X-chromosomesare active in the adult female. ” Nevertheless, females and males have very similar life spans.I9 It is of interest to determine whether or not both sexes respond in the same degree to somatic activity of transposable elements. In this report we first examine two transposable elements for somatic activity. Second, we artifically activate somatic transposition of a P element and measure the effects on life span. This effect is compared between the sexes. Finally, we make use of a strain harboring P element DNA that shows eye color mosaicism and reversion to wild phenotype (red eyes) as a result of somatic and germ line transposition. In this strain, we test the effect of a high-fat diet that accelerates aging on the frequency of transposition, as measured by eye color mosaicism and red eyes.
METHODS Strains of Drosophila melanogaster. The wild-type flies used were Canton-S and the F1 of two inbred strains, OD4 and CC3, derived from females captured at Chateau Tahbilk, in Victoria, Australia.’O This second set of flies was abbreviated OD4/CC3. Birm 2 : ry506(abbreviated Birm-2) contains the second chromosome of the Birmingham strain, with 17 nonautonomous P elements.” PA2,3(99B) contains a P element with an in-phase deletion between the second and third intron in the gene for the transposase. The transposase is somatically active and fails to act on its own element.21To obtain viable flies that were the first-generation cross of PA2,3(99B) and Birm-2 strains, development was at 18”C, so that the P element would not be somatically active.*’ For comparison the parent strains were also raised at 18°C. Mosaic strains were constructed from the CaSpeR 3 1.4 strain by crossing with PA2,3(99B) and breeding from mosaic-eyed females. The strain CaSpeR 3 1.4 contains the CaSpeR element on the third chromosome.22This element contains the wild-type white gene and is bounded by P element termini. In this strain the color produced by the activity of the gene is reduced by a position effect. Full expression is only possible when the gene is transposed into a region where suppression of gene activity is less. When a somatically active transposase is provided in trans, flies are produced in which patches of deep color appear on the pale background. In addition white patches are produced. Flies showing wildtype eye color and also white eyes are observed.” Plasmids Used. cDm2055 for copia,’ cDm2042 for 412,23and sAF2 for Adh.24 Diets. Control and high-fat media were as previously de~cribed.’~ Aging Protocol. Newly eclosed adults were kept in batches of 20 in vials of fresh medium and transferred weekly. Deaths were scored at the time of transfer. Deaths were considered to occur midway between times of counting. DNA Preparation and Southern Blotting. Genomic DNA was extracted using the method of McGinnis et a1.,26from batches of 150 files after snap freezing in liquid nitrogen and grinding to a fine powder in a mortar and pestle. DNA from agarose electrophoresis gels was Southern blotted onto nylon membranes using standard procedure^.^' Probe DNA was prepared from purified plasmids DNA and w3’P ATP via a random hexanucleotide primers method (Oligo-labeling kit from
85
DRIVER & McKECHNIE: TRANSPOSABLE ELEMENTS
Bresatec), and probing was by standard conditions of hybridization and X-ray film e x p o s ~ r e . ~Autoradiographic ’ gels were scanned and quantified with a Hoeffer GS3000 densitometer associated with Apple software.
RESULTS Somatic Activity of Transposable Elements Using the two types of wild-type flies, Southern blots were carried out on genomic DNA, undigested with restriction enzymes, from young (3-day-old) and old (45-day-old) males and probed with plasmids containing either 412 or copia
“r I MEAN PERCENT INCREASE IN BAND INTENSITY IN ADULT DNA ,oo
412
COPIA
FIGURE 1. Increase in transposable element band intensity of unrestricted DNA, determined with males of two wild-type strains (hatched column, Canton-S; open columns, OD4/ CC3). Paired adjacent DNA spots for young and old flies from the same cohort were compared on the same membrane and the ratio of intensities determined. To correct for variation in the amount of DNA on the membrane, the spots were reprobed with Adh. The number of such pairs of old and young DNA spots scanned was 4-6. Error bars on the graph are SEM. The horizontal line through the column marks the expected ratio for no change with age.
sequences. DNA from corresponding young and old flies were run in adjacent wells for pairwise comparison. To correct for variation in amount of DNA bound, each membrane was reprobed with a plasmid containing the Adh sequence. The autoradiographic intensity paralleled the ethidium bromide fluorescence, and no additional band that might be extrachromosomal circular DNA was visible. The mean ratios of band intensities for DNA from old and young flies are shown in FIGURE1. Element 412 provided evidence of an increase in the band intensity with age in the Canton-S strains. The OD4/CC3 flies did not show evidence of an increase for 412. However, in the OD4/CC3 flies, for copia, a significant increase with age was obtained. When the DNA was digested with Eco R1 or Hind 111, electrophoresed in agarose gel, Southern blotted, and probed with 412 or copia, no new bands appeared. Such a result would be expected if new DNA sequences
ANNALS NEW YORK ACADEMY OF SCIENCES
86
FEMALES MOSAICS DOUBLE MOSAICS RED EYES WHITE EYES
MOSAICS DOUBLE MOSAICS RED EYES WHITE EYES
0
5
1
0
1
5
2
0
2
5
r
)
3
5
4
0
4
5
5
0
PERCENTAGE
FIGURE 2. The effect of diet on the percentage of eye phenotypes in the mosaic eye strain. The numbers of red-eyed flies, white-eyed flies, and mosaics are expressed as percent of total flies. The frequency of flies mosaic in both eyes (double mosaics) is expressed as percent of flies mosaic in at least one eye. The number in each group was between 150-300. Significant differences in ratios between dietary treatments are indicated by asterisks ( p < 0.01, difference between population proportions28).No errors are quoted, as this would be inappropriate for this statistical test. Open bars, low-fat diet; shaded bars, high-fat diet.
arising from insertion of the transposable elements is dispersed throughout the genome. With the mosaic-eyed strain, the effect of diet on the appearance of phenotypes expected to result from transposition is shown in FIGURE 2. The high-fat diet significantly elevated the incidence of mosaics in both sexes. Since the effect of diet on this frequency could be the result of a differential effect on the survival of genotypes, a second measure was also used: the ratio of mosaicism in the second eye of flies known to be mosaic in one eye. The ratio of double-eye mosaics to total mosaics was also higher on the high-fat than on the low-fat diet in both sexes. It was also noted that the frequency of occurrence of nonmosaic red eyes was increased in the high-fat diet, whereas the frequency of appearance of all-white eyes was decreased on high-fat diets (FIG.2).
P Element Transposition and Life Span To test whether a reduction in life span in the F1 strain containing the fully active element occurs, the life span of the strain was compared with the life spans of the parent strains. When two inbred strains are crossed, one would expect the
DRIVER & McKECHNIE: TRANSPOSABLE ELEMENTS
87
life span of the FI would be equal to, or greater than, the man of the parental strain, depending on whether heterosis O C C U ~ S . ' ~ Life spans were measured at 18°C for the strain carrying the PA2,3 element, the strain carrying the nonautonomous elements (Birm-2), and the F1 strain carrying both. The life span of the F1 males was indistinguishable from that of the mean of the parental strains, and that of the F1 females was greater than the mean of the parental strains (FIG.3), suggesting some heterosis in females. These measurements were repeated at 25°C. A significant reduction of life spans with respect to the mean of the parental strains was observed for both sexes (FIG.3).
DISCUSSION Is there a complement of transposable elements that are somatically active in normal adult tissues and that contribute to the aging processes? As a first step towards answering this question, we have measured an increase in DNA band intensity for the elements 412 and copia, this latter element being measured in the strain only. The detected increase suggests that genomic tissue from old flies contains more copies of these elements than the equivalent DNA from young flies.
UFESPANS AT 25' C P dona 2.3 Birmingham2 MEAN F1 ESPANS AT lE0C P dona 2,3
Birmingham-? MEAN F1 0
20
40
80
80
100
laD
UFESPAN IN DAYS
FIGURE 3. Top: Comparison of mean life spans at 25°C of the Birmingham-2 strain, the PA2,3 strain, and flies from their FI. Numbers of flies in each group: PA2,3: 161 males, 249 females; Birmingham-2: 129 males, 191 females; 241 FI males, 245 F1 females. Bottom: Comparison of mean life spans of the same strains measured at 18°C. Numbers of flies in each group: PA2.3: 181 males, 216 females; Birmingham-2: 55 males, 86 females; 154 FI males, 206 FI females. Open bars. females; harchedbars, males. Error bars, SEM. For each sex. significant differences between mean FI life span and the means of the parental strains are indicated by asterisks ( p > 0.01; Student's r-test); the difference that is not significant is indicated NS.
88
ANNALS NEW YORK ACADEMY OF SCIENCES
An appearance of transposable elements at new sites in the genome has not been demonstrated. To d o so would be technically difficult, since each reinsertion would be likely to be in a different position and produce a unique DNA sequence. We have examined only two transposable elements in one strain and one in a second. For each element we found that one strain showed an increase in intensity of the band detected by the DNA probe for that element. We have interpreted this as an increase in the relative number of that transposable element in the DNA extracted from the old flies. Since there are more than 40 types of transposable element^,^ it would be reasonable to expect that other elements will be found to transpose in this species during aging. However, several factors should be considered. Previous reports on somatically active transposable elements with mariner in Drosophilu9-" and the Tcl element in C. elegansI2 indicate that there is strain-specific variation in activity. The data presented in this paper suggests the same is true for 412. Such variation is likely to be a general property of such elements. The elements that we chose to investigate are of a type such that transposition in postmitotic tissue would be expected to produce an increase in copy number.) Although most transposable elements in Drosophilu would be expected to have this p r ~ p e r t y it , ~ is not likely to be universal. Accordingly, different techniques would be needed to investigate the somatic activity of the full complement of elements. Although it has been known for some time that diet effects aging in many species,l**the exact cause of this effect has been elusive. Data presented here suggest that germ line or somatic transposition resulting in reinsertion of a P element is promoted by a high fat diet. However data also suggests that the rate of excision, leading to white eyes, was decreased by the high-fat diet. Excision of a P element is unlikely to be a deleterious event unless imprecise excision causes damage to a gene that was functional when the P element was present. Insertion has a higher probability of inactivating a gene.3 These results suggest that potentially damaging transposition events are promoted by a high-fat diet. Our data is consistent with the hypothesis that transposition of a P element is increased by a diet known to increase the rate of aging. This effect may be common to many elements if diet modified a common factor, such as the state of the chromosome. In that event, changes in damage induced by transposable elements may contribute to the effect of diet on the age phenotype. At 25" in these constructs Pelements are reported to be active in transposition.21 We were able to verify this. When the FI flies carrying a somatically active P element were raised at 25°C only a few adults were produced. These were of low viability and had a high frequency of deformities, as reported by Robertson et Accordingly, to obtain viable nondeformed adults we raised the flies at 18°C. The life spans of these flies, subsequently measured at 25"C,showed a significant reduction, with respect to the mean of the adult strains, in both sexes and to about the same degree. No such reduction was observed when the life spans were measured at the lower temperature, at which transposition does not occur. The life span of male Fls was indistinguishable from that of the mean of the parental strains, and that of the female was somewhat greater than the mean of the parental life spans, indicating some heterosis. Thus the life shortening effect of somatic transposition of P elements was not less in females than in males. One of the major obstacles to the ready acceptance of the somatic mutation model of aging is that in situations where the chromosome number differs, the aging rates are similar. The lack of a shorter life span in male Drosophilu, compared to females, is one such example. The observation that somatically active elements also do not produce more damage in males may provide an explanation for this
DRIVER & McKECHNIE: TRANSPOSABLE ELEMENTS
89
paradox. Alternative theoretical reasons for the absence of an effect on transposable element damage are provided by the exponential growth model of Murray4and the dominant/codominant model of Morley .I7 Additional alternative explanations should also be considered. For example, the element may cause somatic mutation in a limited number of “hot spots,” which as a result of natural selection are not on the more sensitive X chromosome. We are currently investigating this further. New insights provided by this idea are that in higher organisms, including humans, somatic activity of transposable elements may cause aspects of the age phenotype. Transposition may be responsive to control by identifying and changing environmental triggers and suppressors of this activity. Pharmacological support may be possible for tissues that are deficient in the product of genes that are identified as hot spots for transposable element mutagenesis. New approaches to the genetic analysis of age-associated pathologies may be possible. Finally, a possible role for transposable elements may be considered in the expression of such rare adult-onset genetic conditions as Huntington’s chorea.
CONCLUSION We have presented data indicating that transposable elements may be somatically active in adult tissues. The somatic mutation thereby produced may contribute to the aging phenotype. Somatic mutation by transposable elements may be able to explain the chromosome number paradox and appears to be able to contribute to the link between diet and aging. This hypothesis deserves further consideration. It can reasonably be expected that insights may arise that will be of value theoretically and clinically. The model also has the advantage of being readily testable in other organisms using current tools.
SUMMARY We have considered the hypothesis that transposable elements may contribute to the aging process through somatic mutation. We have presented evidence to suggest that at least two elements, Copia and 412, are capable of somatic activity in adult Drosophila tissue. A strain harboring a third transposable element, P, was produced that showed eye color mosaicism and reversion to wild phenotype (red eyes) as a result of somatic and germ line transposition. A high-fat diet, known to accelerate aging, increased the frequency of eye color mosaicism and red eyes. We induced life span shortening by artificially activating somatic transposition of P elements, and the extent of reduction in life span was similar in both sexes. These data are consistent with the notion that some aspects of the age phenotype may be caused by mutational activity of transposable elements in somatic tissues. The hypothesis is readily tested in other organisms, including humans. It offers new dimensions in the understanding and management of age-associated changes.
ACKNOWLEDGMENTS The Birm-2 and the Casper strains were a gift from Dr. John Sved, and the PA2,3(99B) strain was a gift from Dr. John Oakeshott. We would like to thank both gentlemen for these strains.
90
ANNALS NEW YORK ACADEMY OF SCIENCES REFERENCES
I . HAYFLICK, L. 1985. Theories of biological aging. Exp. Gerontol. 20: 145-159. T. B. L. 1989. DNA, mutations and aging. Mutat. Res. 219 1-7. 2. KIRKWOOD, D. J. & D. H. FAWCETT. 1986. Transposable Elements in Drosophila 3. FINNEGAN, melanogaster. Oxford Surveys on Eukaryotic Genes, Vol 3. Academic Press. London. 4. MURREY, V. 1990. Hypothesis: Are transposons a cause of ageing?’ Mutat. Res. 237: 59-63. 5 . WOODRUFF,R. 1991. Mutation and evolution: Lifespan reduction by transposons and premeiotic clusters. 38th Meeting of the Genetics Society of Australia, Monash University. Abstr. 52. H. D. 1990. Molecular analysis of ageing processes in fungi. Mutat. Res. 6. OSIEWACS, 237: 1-8. J. W. 1990. Extrachromosomal circular DNA and genomic plasticity in 7. GAUBATZ, eukaryotic cells. Mutat. Res. 237: 271-292. 8. EMORI,Y., T. SHIBA,S. KANAYA,S. INOUYE,S. YUKI& K. SAIGO.1985. The nucleotide sequence of Copia and Copia-related RNA in Drosophilu virus-like particles. Nature 315: 773-776. A. KOGA& D. HARTL.1991. Introductionofthe transposable 9. GARZA,D., M. MEDHORA, element mariner into the germline of Drosophila melanogaster. Genetics 128(2): 303-310. 10. MEDHORA,M., K. MARUYAMA & D. HARTL.1991. Molecular and functional analysis of the mariner mutator element Mos I in Drosophila. Genetics 128(2): 311-318. K. & D. HARTL.1991. Evolution of the mariner mutator element in I I . MARUYAMA, Drosophika species. Genetics 128(2): 3 19-329. 1991. Somatic mutation in the wings of Drosophilu 12. GETZ, C. & N. VAN SCHAIK. rnelanogaster females dysgenic due to P elements when reared at 29°C. Mutat. Res. 248: 187-194. Y. & L. YESNER.1987. High frequency excision of transposable elements 13. EMMONS, Tcl in the nematode Caenorhabditis elegans is limited to somatic cells. Cell 3 6 599-605. E., 0. DREAZEN, A. KLAR.1. G. RECHAV,D. RAM,J. B. COHEN& D. 14. CANAANI, GIVOL.1983. Activation of the c-mos oncogene by insertion of an endogenous intracisternal A-particle genome. Proc. Natl. Acad. Sci. USA 8 0 71 18-7122. 15. DRIVER,C. J. I., G. COSOPODIOTIS, R. WALLIS& G. ETTERSHANK. 1986. Is a fat metabolite the major diet dependant accelerator of ageing? Exp. Gerontol. 21: 497-507. 16. LAMB,M. J. 1965. The Effects of X-irradiation on the longevity of triploid and diploid females Drosophila rnelanogaster. Exp. Gerontol. 1: 181-184. 17. MORLEY,A. A. 1982. Is ageing the result of dominant and codominant mutations? J. Theoret. Biol. 9 8 469-474. 18. LUCCESI,J. C. & J. E. MANNING.1987. Gene dosage compensation in Drosophika melanogaster. Adv. Genet. 24: 371-429. & G. MOKRYNSKI. 1985. Aging in Drosophila. In 19. BAKER,G. T. 111, M. JACOBSON Handbook of Cell Biology and Aging. V. Cristofalo, Ed.: 51 1-578. CRC Publishers. Boca Raton, FL. P. W. HEINSTSTRA & M. J. PYKA.1991. Heritable 20. GEER,B. W.. S. W. MCKECHNIE, variation in ethanol tolerance and its association with biochemical traits in Drosophila melanogaster. Evolution 45: 1107-1 109. H. M., C. R. PRESTON,R. W. PHILLIS,D. M. JOHNSTON-SCLITZ, W. K. 21. ROBERTSON, BENZ& W. R. ENGELLS.1988. A stable genomic source of P element transposase in Drosophila rnelanogaster. Genetics 118 461-470. 22. PIROTTA,V. 1988. Vectors for P element transformation in Drosophika. In Vectors: A Survey of Molecular Cloning Vectors and Their Uses. R. L. Rodriguiz & D. T. Denhart, Eds.: 437-456. Butterworths. Boston, MA. S. ISHIMARU & K. SAIGI. 1986. Nucleotide sequencecharacterisa23. YAKI,S., S. INOUYI, tion of a Drosophila retrotransposon, 412. Eur. J. Biochem. 158 403-410.
DRIVER & McKECHNIE: TRANSPOSABLE ELEMENTS
91
24. GOLDBERG, D. A. 1980. Isolation and partial characterisation of the Drosophila melanoguster Adh gene. Proc. Natl. Acad. Sci. USA. 77(10): 5794-5798. C. J. I. & G. COSOPODIOTIS. 1978. The effect of a high fat diet on longevity 25. DRIVER, of Drosophilu mclanogastcr. Exp. Gerontol. 14: 95- 100. 26. MCGINNES, W . , A. W. SHERMOEN & S. K. BECKENDORF. 1983. A transposable element insertedjust 5’ to a Drosophila glue protein gene alters gene expression and chromatin structure. cell 34: 75-84. 27. SAMBROOK, J . , E. F. FRITSCH& T. MANIATIS. 1989. Molecular cloning: A laboratory manual, 2nd edit. Cold Spring Harbor Laboratory Press. Cold Spring Harbor, NY. W. 1983. Biostatistics: A foundation for analysis in the health sciences, 3rd 28. DANIEL, edit.: 192-193. Wiley and Sons. New York, NY.
