High-frequency excision of transposable element Tc1 in the nematode caenorhabditis elegans is limited to somatic cells

High-frequency excision of transposable element Tc1 in the nematode caenorhabditis elegans is limited to somatic cells

Cell, Vol. 36, 599-605. March 1964, Copyright(D 1964 by MIT C0926674/64/030599-07 $O~.CCI/O High-Frequency Excision of Transposable Element Tel...

2MB Sizes 0 Downloads 5 Views

Cell, Vol. 36,

599-605. March


Copyright(D 1964

by MIT



High-Frequency Excision of Transposable Element Tel in the Nematode Caenorhabditis elegans Is Limited to Somatic Cells Scott W. Emmons and Lewis Yesner Department of Molecular Biology Division of Biological Sciences Albert Einstein College of Medicine Bronx, New York 10461

Summary Tel transposable elements in the nematode Caenorhabdiis elegans undergo excision at high frequency. We show here that this excision occurs primarily or entirely in the somatic tissues of the organism. Absence of germ-line excision is demonstrated by showing that Tel elements are genetically stable; elements at particular genomic sites, as well as the overall number of elements in the genome, were stably maintained during a year of continuous, nonselective propagation. Somatic excision is demonstrated by showing that empty Tel sites arise during a single generation of growth of a synchronous population and are not inherited by the next generation. These results suggest that excision of Tel elements is under the control of tissue-specific factors.

cation during transposition, sequencing data are ambiguous (Rosenzweig et al., 1983b): Tel may cause a duplication of two bases of target DNA, or it may not duplicate any target sequences. We have shown previously, using one of the two highcopy-number strains, Bergerac, that Tel elements in this strain undergo excision from their sites of insertion at high frequency (Emmons et al., 1983). Excision was observed in Southern hybridizations of genomic DNA using sequences flanking Tel insertion sites as probes. At each of the insertion sites we examined, the DNA was polymorphic. It contained not only the expected restriction fragment carrying the Tel element but also a small amount of a fragment carrying an unoccupied insertion site. These unoccupied sites were shown to be present at the level of a few percent in the DNA of a population of worms after a small number of generations of growth from a single, selffertilizing individual, indicating that the polymorphism was the result of excision events occurring at high frequency during growth. These events result in loss of the Tel element and re-ligation of the target DNA in an apparently precise (& 50 bp) fashion. Such excision, if it occurred in the germ line, would result in a high degree of genetic instability of Tel elements. We show that, on the contrary, Tel elements are genetically stable; excision occurs primarily or entirely in the somatic tissues of the organism.

Introduction Results One of the intriguing properties of transposable elements in eucaryotes is their ability to respond to developmental signals within a tissue or organism. Experiments in plants, most notably in maize, have shown that such properties as the frequency of transposition of transposable elements, the frequency of excision of transposable elements, and expression of genes where transposable elements reside may differ in different plant tissues or locations within a tissue, or at different times during development (McClintock, 1951; Fedoroff, 1983). In Drosophila, activation of P element transposition in certain interstrain crosses is limited primarily to germinal tissue (Engels, 1979; Bregliano and Kidwell, 1983). Here we describe another example from an animal of tissue-specific regulation, in which spontaneous excision of a transposable element is confined to somatic tissues. The genome of the nematode Caenorhabditis elegans contains a 1.6 kb repetitive DNA sequence, denoted Tel , that has the properties of a transposable element (Emmons et al., 1983; Liao et al., 1983). Conserved copies of Tel are found at approximately 30 chromosomal sites in most C. elegans strains, while in two strains several hundred dispersed copies of Tel are present. These high-copynumber strains presumably have undergone a process of amplification during which additional copies of Tel have inserted at many new genomic sites. The Tel sequence is bounded by 54 bp perfect inverted repeats and contains an open reading frame of 819 bases (Rosenzweig et al., 1983a). With regard to the creation of a target site dupli-

Tel Elements Are Stably Maintained during LongTerm Propagation Despite the evidence for a high rate of excision, Tel elements at particular genomic sites are genetically stable. This provides evidence for a general absence of germ-line excision events. One method we used to assess the genetic stability of Tel elements was to study a population of worms before and after continuous, long-term propagation, to determine whether Tel elements had been lost from particular sites. Such a loss should have occurred if germ-line excision took place, provided excision was not balanced by reinsertion into the same site and did not result in a growth disadvantage. Analyzing a population of worms before and after its propagation through many generations effectively amplified the probability of observing germ-line excision events by allowing the products of such events to accumulate. We demonstrated the genetic stability of Tel elements inserted at three independent chromosomal sites in Bergerac. Each Tel element was previously discovered as a polymorphism of restriction fragment length between Bergerac, a high-copy-number strain, and Bristol, a low-copynumber strain. At each site Bergerac has a single Tel element inserted into sequences that are otherwise identical with those of Bristol. The resulting polymorphism has been mapped in each case to a particular linkage group or region of a linkage group. The cloned DNA fragments that were used as probes to study excision were the Bristol

fragments lacking Tel that span the Tel insertion site in Bergerac. These probes were as follows: plasmid pCe1001, carrying a 7.0 kb Barn HI fragment that has been mapped to linkage group V (Emmons et al., 1983; S. Carr and D. Hirsh, personal communication); plasmid pCe(Br)Tl, carrying a 3.5 kb Eco RI fragment that lies adjacent to C. elegans actin genes and has been mapped within a 2% recombination interval in the center of linkage group V (Liao et al., 1983; Files et al., 1983); plasmid pCe1003, carrying a 2.65 kb Hind Ill fragment, which was subcloned from a recombinant phage carrying yolk protein genes, and which maps to linkage group X (Blumenthal et al., 1984; this work). To study the amount of excision of these elements after long-term propagation, we cultivated Bergerac worms on petri plates at 20°C for 72 generations. First, a population of 10,000 was grown in two generations from a single worm. Thereafter, approximately 200 worms were transferred to a fresh plate every 4 to 7 days, or at approximately one-generation intervals, for a period of 13 months. Periodically during the course of the experiment, worms were frozen and stored in liquid nitrogen so that they could be subsequently revived for comparison of their properties (Brenner, 1974). In Figure 1 an analysis is presented of DNA from generations 8, 34, and 72 for the amount of excision at the Tel site that hybridizes to pCe1003. Since the amount of excision depends on the age of the worms, as shown below, DNA was isolated from uniform populations of Ll larvae. pCe1003 hybridizes to a genomic fragment of 4.25 kb, carrying a Tel element (upper arrows in the figure), and to a small amount of a 2.65 kb fragment that results from excision (lower arrows in the figure). (Other fragments that cross-hybridize are due to repetitive yolk protein genes; Blumenthal et al., 1984). No significant increase in the amount of the 2.65 kb fragment can be detected in the DNA of the later generations. The amounts of excision in the three generations are roughly the same and the excision has occurred at a few percent of the sites. Identical results were obtained at the sites homologous to pCe(Br)Tl and pCe1001 (data not shown). A dramatic increase in the percentage of empty Tel sites would have occurred in this experiment had the frequency of germ-line excision per generation been at least 0.01. This was determined by using a computer to model and analyze the propagation experiment. Computer modeling was necessary because of the finite size of the population passed at each generation. In an infinitely large population a steady increase in the fraction of empty Tel sites would occur under the assumptions of irreversible germ-line excision and selective neutrality. In a finite population, however, a stochastic element is introduced when animals are randomly selected to enter each new generation. We simulated this stochastic process in a computer program and compiled the results of many trials carried out at various values of the germ-line excision probability per generation, P, (Figure 2). The simulated experiments showed that if excision occurred with a probability per


Figure 1. Tel Elements Are Not Lost during Long-Term


DNA from Ll larvae of generations 8.34, and 72 of Dergerac wom-rs was analyzed with pCe1003 after digestion with Hind III. The worms were propagated as a population starting from a single worm in the first generation The upper arrows indicate the fragment carrying Tel in Bergerac (4.25 kb). The lower arrows indicate the corresponding fragment lacking Tel in Bristol. A small and constant amount of this fragment is also present in Bergerac as a result of excision of the Tel element. The cloned 2.65 kb fragment from Bristol was used as the probe. Lanes labeled a contained one tenth the amount of DNA in lanes labeled b, which was about 3 pg. This dilution was included to allow accurate comparisons of the amounts of DNA loaded from each of the three generations.

generation per site of 0.01, this would result in an overall loss in the population of 50% per site, with a standard deviation of lo%, after propagation for 72 generations. These results are the average and standard deviation for 40 simulated trials. The experimental result clearly shows no such loss, implying that germ-line excision events, if they occur at all, must occur at a frequency of much less than 1% per generation.

Evidence for Stable Inheritance of Tel Elements from Analysis of Single Worms We confirmed the lack of accumulation of empty Tel sites during long-term propagation by analyzing single worms from the 72nd generation and showing that they carried Tel elements at the three chromosomal locations studied in this work. No worm was found that had inherited an empty Tel site in either DNA complement. This analysis was carried out on 41 worms cloned from the 72nd generation. Each worm was allowed to give rise to a population on plates, and DNA from each population was

Somatic Excision of a Transposable 601





digestion with Hae III, which together include nearly all of the genomic sequences homologous to Tel (Emmons et al., 1983). DNA from Bergerac worms of generations 8, 34, and 72 had the same amount of hybridization to these fragments within the probable accuracy of the measurement (&20%; data not shown). From this observation we can extend our conclusion regarding the general absence of germ-line excision, drawn from studies of Tel elements at three chromosomal locations, to most or all of the other Tel elements in the genome as well, unless these undergo germ-line excision events that are precisely balanced by insertions at new chromosomal sites.

Somatic Excision Demonstrated in Synchronous Populations






Simulation of Propagation


The predicted fraction of DNA strands in a population of worms that retain Tel at a particular chromosomal site is given as a function of the number of generations of propagation of the population, assuming various probabilities of germ-line loss per generation (P.). The prediction is based on the results of a computer simulation of the actual experiment carried out with nematodes. The computer program used a random-number generator to determine for each gamete passed on to the next generation whether it carrfed a Tel element at the site in question, The probability of its doing so was determined by the amount of excision that had already occurred in the population in previous generations as well as the probability of loss in the current generation, P.. Each round of selection in the program corresponded to one of the transfers in the experiment and selected the same number of worms that were actually transferred. The curves are the averaged results of 40 trials. Sars showing one standard deviation around the mean value are given for the curve for 1% excision per generation.

analyzed by Southern hybridization using one or more of the three probes. Twelve clones were analyzed with pCe1001, 29 clones with pCe(Br)Tl , and 38 clones with pCe1003. Since two chromosomes were probed for each clone, this gave a total of 158 Tel sites that were analyzed. In every case the populations showed only a small amount of Tel excision, indicating that the worm from which the population had been derived was homozygous for Tel insertion at the probed site (data not shown). Thus even after examination of single worms from the 72nd generation we find no evidence of germ-line excision.

The Overall Number of Tel Elements in the Genome Is Stable No drastic decline in the overall number of Tel elements in the Bergerac genome occurred during continuous propagation. We measured the overall number of Tel elements in the genome by comparing the amount of hybridization of a Tel-specific probe to genomic restriction fragments homologous to Tel in Southern hybridization experiments. The fragments we analyzed were those produced by

The above evidence for genetic stability of Tel elements implied that the excision that was occurring at high frequency must be confined to somatic lineages. Therefore, like the somatic cells themselves, empty Tel sites must arise anew in each generation. We demonstrated this process occurring at each of the three sites studied here. We analyzed the amount of excision that had occurred as a function of age in a synchronously developing population of worms and in their progeny. The amount of excision was assessed from the amount of the genomic fragment that arises from the insertion site by loss of the Tel element. The amount of this fragment was found to increase during development and to fall again at the begin ning of the next generation, as predicted. Figure 3 shows this result at the site that hybridizes to pCe1001. Staged populations of Bergerac worms were obtained by isolating early embryos from young, gravid hermaphrodites and allowing them to develop synchronously as described under Experimental Procedures. At intervals during the first and second generations a portion of the population was withdrawn for analysis of the amount of Tel excision. The probe hybridized to a major Barn HI fragment of 8.7 kb, representing the filled Tel site, to a 7.0 kb Barn HI fragment arising by excision, and to several other genomic fragments (Emmons et al., 1983). The fragments other than the 8.7 kb and 7.0 kb fragments hybridize to the probe because of one or more shared repetitive sequences. Short repetitive sequences are known to be interspersed in C. elegans DNA (Emmons et al., 1979, 1980). The amount of the 7.0 kb fragment varies with the stage of the worms. It is least in DNA of embryos and increases as the worms develop through larval stages. In embryos of the next generation the amount falls again to the lowest level, clearly showing that most or all empty Tel sites are not inherited. A similar result obtained with pCe1003 is shown in Figure 4. A quantitative analysis of results with the three probes is presented in Figure 5. The percentage of empty sites increased from as little as 0.3% in embryos to as much as 10% in late larval stages. In adults the percentage falls somewhat. This is perhaps to be expected from the proliferation of germ cells in the gonad and from the develop-

cell 602






10 2-c





a 6 4

Figure 3. Analysis of Excision

in Staged Populations

Using pCe1001

DNA isolated from wonts withdrawn at the stages shown from a synchronously devefoping population of Bergerac worms (EMS. Embryo; Ll , Ll larvae; L3. L3 larvae; Fl :EMB, embryo of the subsequent generation; Fl :Ll , Li larvae of the subsequent generation) was digested with Barn HI and hybridized afler agarose gel electrophoresis to pCe1001. Wons of the Fl generation were obtained by treating the population after it matured with sodium hypochlodte as described under Experimental Procedures. The genomic restriction fragment carrying Tel (8.7 kb) and the fragment that arises from it by excision of the Tel element (7.0 kb) are indicated by arrows. The other fragments that hybridize do so because they share repetitive sequences with the probe. For each sample a series of dilutions was run so that the amount of the fragment arising from excision could be accurately compared with the amount of the fragment carrying the filled site. Lanes labeled a contain about 3.0 Pg of genomic DNA. Lanes labeled b and c contain, respectively, %I and ‘/SOthe amount of DNA in lanes a.

5;i :: Ly

2 i.


Figure 4. Analysis of Excision

in Staged Populations

Using pCe1003

The same DNAs shown in Figure 1 were digested with Hind Ill and hybridized to pCe1003. The lanes labeled Adult are from a second synchronous population. The upper arrows indicate the filled site (4.25 kb) and the lower arrows the empty site (2.65 kb). Lanes labeled a have about 2.0 pg of DNA and those labeled b have one tenth that amount.

ment of embryos internally. Adult hermaphrodites of C. elegans contain 954 somatic cells, 2600 germ cells, and about 12 developing embryos (Sulston and Horvitz, 1977; Hirsh et al., 1976; Kimble and Hirsh, 1979). The germ cells develop postembryonically from two progenitor cells in the hatchee, and hence the proportion of cells that are somatic continuously falls during development (Kimble and Hirsh, 1979). Accordingly, the fraction of empty Tel sites in






EMB LI L3 ADULTFI:EMBFI:LI e-n--a b a b a b a b a b a b Bristol







EMB LI L3 FI:EWB VLI Figure 5. Amount of Excision as a Function of Stage Results are shown of a quantitative analysis of the gels shown in Figures 3 and 4 and of two other experiments, one with the same DNA samples but probed with pCe(Br)Tl, and one of an entirely independent experiment analyzed with pCelCO1 (hatched bars). The gels were analyzed by tracing them with a Joyce-Lobe1 densitometer and detenining the areas of the peaks, Only bands well below the saturation level of the film were traced. The percentage of excision was determined by comparing the amount of the band resulting from excision with the amount of the band representing a filled site in one of the diluted samples. Asterisk indicates no excision band was detectable (less than 1% excision).

somatic DNA is higher than the fraction in DNA from whole worms. In DNA of embryos and Ll larvae of the next generation the percentage of empty Tel sites falls again to the level in the same stages of the parental generation. We do not know whether excision occurs with equal frequency in all somatic cells, or is more frequent in some tissues than in others. If it occurs equally in all somatic tissues, then the probability of somatic excision is around 10% and hence occurs more than IO times as frequently as germ-line excision. If excision is confined to only certain somatic tissues, then the probability of excision in these tissues is even higher.

High-Frequency Somatic Excision Is Probably a General Property of Tel Elements in Bergerac The Bergerac genome contains approximately 200 dispersed Tel elements (Emmons et al., 1963) four of which have been separately studied. Three are shown in this report to undergo excision at high frequency in somatic cells. A fourth element has been isolated and mapped to

Somatic Excision of a Transposable 603


linkage group I and shown to undergo frequent excision (Rose et al., 1983). We argue that excision of this element is likely to be somatic also, since germ-line excision at the observed frequency would have rapidly resulted in its loss, as we show here. A fifth element, studied previously, was observed to undergo excision but at a much lower frequency (Emmons et al., 1983). We have now found that this is not a Tel element. It is a sequence similar in size to Tel , present in Bergerac and absent in Bristol. It has been studied using the Bristol empty site cloned in plasmid pCel4a. We studied the restriction sites carried by this element by genomic Southern hybridization of Bergerac DNA and found that it lacks all of the expected Tel restriction sites for the restriction endonucleases Sal I, Xho I, Eco RV, and Cla l-six restriction sites in all (Emmons et al., 1983; Rosenzweig et al., 1983a; data not shown). This element is therefore either unrelated to Tel or highly diverged from it. Thus, without exception, known Tel elements undergo high-frequency excision. This excision is shown here to be confined to somatic tissues for three elements, and is likely to be for the fourth as well, and hence we conclude that somatic excision is probably a general property of Tel elements.

Discussion Despite the evidence for high-frequency excision, Tel elements are genetically stable. This makes it possible to use them as genetic markers, and they have proven useful for correlating the physical and genetic maps of C. elegans. Files et al. (1983) have located C. elegans actin genes in linkage group V, and Blumenthal et al. (1984) have located yolk protein genes in linkage group X, both using DNA polymorphisms resulting from the presence of Tel elements near these genes in Bergerac. The excision events that Tel elements undergo appear to be confined to somatic cells, Our conclusions apply only to a particular type of excision event-namely, one that results in re-ligation of the target DNA in a reasonably faithful manner. We have not yet determined whether the excision is precise, restoring the exact target sequence, or instead creates small insertions or deletions. Events producing large aberrations that did not restore a targetsite fragment of nearly the original size would not have been detected. Events resulting in a chromosome break could have been detected provided the broken ends were stable. Such ends would have resulted in the appearance of two specific smaller fragments that when summed together equalled in size the re-ligated fragment we observe. No such fragments have been detected. If germ-line excision occurs, it must be at a frequency of much less than 10e2 per gamete per generation. Thus there is at least an order of magnitude difference in the excision probability between germinal and somatic tissue. Conceivably, germ-line excision could occur at a higher frequency, but by a different pathway-for example, one

that produced a large aberration or a lethal event such as a chromosome break. We consider this unlikely. We observed no decline in the overall number of elements during long-term propagation (although germ-line excision could be balanced by reinsertion), and if germ-line excision occurred at high frequency but was accompanied by a lethal event, the strain would be sterile as a result of the cumulative effects of the excision of the approximately 200 elements in the genome. Precise excision of transposable elements is rare for procaryotic elements generally and for certain eucaryotic elements, such as Drosophila copia elements (Kleckner, 1981; Rubin et al., 1982). For other eucaryotic elements it is more common. P elements inserted at the white locus in Drosophila melanogaster undergo precise excision at a frequency greater than 10e3 during hybrid dysgenesis (Rubin et al., 1982; O’Hare and Rubin, 1983). An FB element also inserted at the white locus in Drosophila undergoes precise excision at a similar frequency in any genetic background (Collins and Rubin, 1983; Green, 1967). These excision events are detected initially by their genetic consequences, and indeed the most thoroughly studied examples of excision of eucaryotic transposable elements come from genetic studies in maize, most notably those of McClintock (McClintock, 1951; Fedoroff, 1983). Excision of Tel elements differs from that of the Drosophila elements by its extremely high frequency and restriction to somatic lineages. In this regard it more nearly resembles events in maize and other plants, where transposable elements can undergo frequent spontaneous excision in somatic tissues, resulting in variegated mutant phenotype (Fincham and Sastry, 1974; Fedoroff, 1983). Little is known about the mechanism(s) of transposition of transposable elements in eucaryotes. It could be that in eucaryotes transposition is not duplicative, leaving behind an unaltered copy of the transposon, as it is in procaryotes, but that instead excision of the element with re-ligation of the target sequences occurs as the first step of the transposition pathway. A number of experiments in maize, particularly those of Brink and coworkers, support such a transposition mechanism for maize controlling elements (Greenblatt and Brink, 1962). If excision does precede transposition, somatic excision would be evidence for constant activation of the transposition pathway in somatic cells. We have detected extrachromosomal copies of Tel that may be products of excision and/or represent transposition intermediates (Ruan and Emmons, submitted). Why is excision of Tel elements confined to somatic cells? Since the somatic and germinal lineages of C. elegans are of nearly the same length, undergoing approximately the same number of rounds of cell division, and giving rise to roughly the same number of cells (Hirsh et al., 1976; Kimble and Hirsh, 1979; Sulston and Horvitz, 1977; Sulston et al., 1983) the difference in the observed frequency of excision of Tel elements in the two lineages cannot simply be due to a greater number of rounds of cell division in one lineage over the other and a correlation

cell 604

of excision with DNA replication or another cell-cycleassociated event. It might be that excision is developmentally regulated by local chromosomal factors such as flanking DNA sequences or local transcriptional activity. In this case the three elements we have studied would happen to be located at sites where excision is activated in one or more somatic tissues. Since any element that excises at high frequency in germ tissue is likely to be lost, we would expect most elements in the genome to be activated in somatic tissues only. A hypothesis of this sort predicts a precise developmental specificity for excision of each element, which would differ from that of other elements at different chromosomal sites. From the data presented here it is not possible to decide whether or not this is the case. Although there appears from our data to be a continuous increase in the number of empty Tel sites during development, this could be due either to continuous excision or to excision at a precise time during the development of a particular lineage, followed by relatively greater subsequent proliferation of this lineage relative to other lineages. An alternative explanation for somatic excision is that excision of Tel elements is controlled by factors that distinguish somatic cells as a whole from germ cells. A fundamental difference between the somatic and germ lineages was postulated to exist by Weismann and received early support from the work of Boveri on the nematode Ascaris, where chromatin diminution occurs only in blastomeres destined to give rise to somatic cells (Wilson, 1925). This restriction of chromatin diminution to the somatic line was shown by Boveri to be due to the inheritance by the germ cells of certain granules from the egg cytoplasm, and such germ cell determinants have been demonstrated in a number of other animals, including C. elegans (Wilson, 1925; Eddy, 1975; Strome and Wood, 1982). Whether excision of Tel elements is governed by factors specific to single tissues or by factors present in all germ cells or all somatic cells, Tel elements should provide an opportunity for analyzing the nature of the developmental mechanisms involved.

Nematcdes Bergerac warns used in this work were strain EM1002. This line derives from a single worm selected in this laboratory in November 1961, from a populatii of Bergemc (BD) worms obtained from the Caenorhehdifis Genetics Center, in December 1980, and maintained frozen. Bristol worms were N2 from the laboratory of D. Hirsh. Culture conditions were those of Brenner (1974) wfth certain modkicatkons. Solii medium was RNGM, identical with NG agar of Brennar, except that it contained 0.05 g of Bactopeptone (Diico) per liier instead of 2.5 g. The purpose in reducing the amount of peptone is to prevent the overgrowth of cultures by contaminating bacteria. On RNGM the standard bacterial food source, E. coli strain DF50, grows very poorly. Worms were grown instead with bacterial strain SE57, which was isolated in this laboratory and identifii as a species of Klebsielh. Alternatively, worms were fed by addii pelleted E. cdi to plates. Liquid culture condiiins were those given by S&ton and Brenner (1974) except up to 200 g wet weight of E. coii was added per liter.

Growth of Synchronaur Po~ubtbns Synchronous pcgulations were initiated by isolating embryos from young gravfd hermaphrodttes. Embryos were isdated from worms matured on plates by dissolving worms and bacterial debris in 1% NaDCl (Fisher, 4% 6%), 0.25 M KDH. at 27OC. As soon as the carcasses were disscived (about 5 min) eggs were collected by pelleting and washed with M-9 buffer (Brenner, 1974). They were allowed to hatch swfrfing in M-9 buffer overnight. After hatching, Ll larvae were collected by pefleting and allowed to develop synchronously on plates. They were fed at intervals as required with pelleted E. coti. Progress of growth was monitored by rneasurfrg the fength of the worms using a Wild MSA dissecting microscope fftted with a measurfng eyepiece with a 12mm: 120 scale. Ths stags of the M)rms was determined frcm the length following values given by Byeriy et al. (1976). Long-Tm Continuous v EM1002 worms were grown continuously at 20°C from December 1961. through January 1963. Ths population was maintained on a 9 cm petri plate with SE57. Every 4 to 7 days (the generation time at 2PC is approximately 4 days) a new population was initiated by transferring a small number of worms to a fresh plate with a pipette and M-9 buffer. The number transferred, determined exactly each time by counting, ranged between 56 and 674, with a mean value of 191. After 4 to 7 days of growth the piate was crowded wfth approximately 10,ooO worms. At the end of the experiment the brocxd size of the worms was ccmpared with that of representatives of the starting population, which had been maintained frozen (Brenner, 1974) and was found to be unchanged. Measured average brood sizes at 20% were EM1002 (6th generatii), 117; EM1002 (72nd generation), 113; Bristol (N2), 212. EM1002 worms of both generations were sterile at 25OC, indicating the temperaturesensitive phenotype characteristic of the Bergerac strain had not reverted (Wood et al., 1960). Two other distinguishing characteristics of Efergerac. uncoordinated movement and production of approximately 1% deformed animafs. also had not reverted. Hence, by all the criteria we have examined, Bergerac behaves as a stable strain. DNA IWorn-s were separated from bacteria by pelleting in M-9 buffer (Brenner, 1974) at low speed or, n necessary, by floating on 35% w/v sucrose (Sulston and Brenner, 1974). They were then resuspended at a ccncentratii of up to 10 worms per milliliter in a solution containing 1% SDS, 0.05 M EDTA. 0.1 M Tris buffer (pfi 6.5) 0.2 M NaCl. 1% &nerceptcethanol, and 106 &ml of proteinase K (EM Biochemicals). and incubated at 65OC. After 15-20 min, the carcasses dissolved and the suspension cleared and became highly viscous. II was then extracted wfth phenol and with chloro form:isoamyf alcohol (241) at room temperatie and dialyzed extensively against lx SET (0.15 M NaCI, 0.05 M Trfs, 0.001 M EDTA. pfi 7.9) after addition of RNAase A to 20 pg/ml. DNA prepared in this manner without ethanol precipitation was usually at a high enough concentration to analyze directiy by restriction endonuclease diiestion and agarose gel efectrophoresis. lf the concentration was too low, DNA was cleaved by treatment with a restriction endonuclease and then precipitated with ethand before get electrophoresis. Labeling of hybridization probes by nick translation. agarose gel electrophoresis (0.7% agarose), and Southem hybridizations (Southem. 1975) were carried out as described previously (Emmons et at., 1979). lsdation of pCe1003 and VeMcation That the Associated Polywsm Is a Tel Etament Plasmid pCelOO3 was constructed by subcloning into pBR322 a 2.65 kb Hind III fragment from phage X3-6 following standard methdos (Maniatis et ai., 1962). Phage X3-8 carries a segment of the Bristol genome including yolk protein genes (Blumenthal et al., 1994). The 2.65 kb Hind HI fragment hybridizes to a 4.25 kb Hind Ill fragment in Bergerac, and this polymorphism was used to map the segment to the X chromosome (Bfumenthal et al.. 1964). We have shown that the inserted DNA in Bergerac is a Tel element by showing in a Southem hybridizatan of Bergerac DNA, using pCe1Ca.X as a probe, that the 4.25 kb fragment carries appropriately located Tel restriction Sites. We have veriied the presence of Sal I, Xho I, Cla I, and Eco RV sites in locations expected for a Tel element (Emmons et al., 1983; Rosenzweig et al.. 1983a; data not shown),


Excision of a Transposable



element in Caenorhabditis 3569.

We thank Drs. S. Hawley, S. Henry, and L. Shapiro and members of our laboratory for their critical reading of the manuscript and Dr. T. Blumenthal for the gift of h3-6. We are grateful to M. Matone and S. Smith for careful preparation of media and to J. Stein for computer programming. This work was canted out under a grant from the National lnstiiutes of Heatth to S. E., who was also the recipient of a Faculty Research Award from the American Cancer Society Nematodes used in this wcrk were provided by the Caenorhabditis Genetics Center, which is supported by a contract between the NIH and the Curators of the University of Missouri. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “adverfisement” in accordance with 16 U.S.C. Section 1734 solely to indicate this fact.

Maniatis, T., Fritsch, E. F., and Sambrook, Laboratory Manual. (Cold Spring Harbor, Laboratory).



18, 1963; revised



Proc. Nat. Acad. Sci.. USA 80, 3565J. (1962). Molecular Cloning, A New York: Cdd Spring Harbor

McClintcck, B. (1951). Chromosome organization Cold Spring Harbor Symp. &ant. Biol. 16. 13-47.

and genie expression.

O’Hare, K., and Rubin, G. M. (1963). Structures of P transposable elements and their sites of insertion and excision in the Drosophila melanogster genome. Cell 34, 25-35. Rose, A. M., Baille. D. L., Candido, E. P. M., Beckenbach, K. A., and Nelson, D. (1962). The linkage mapping of cloned restriction fragment length differences in Caenofha&ditis elegans. Md. Gen. Genet. 788, 286291. Rosenzweig. B.. Liao. L. W., and Hirsh, D. (1963a). Sequence of the C. e/egans transposable element Tel Nucl. Acids Res. 7 I.4201 -4209.

28, 1963

Rosenzweig, B., Liao, L. W., and Hirsh, D. (1983b). Target sequences for the C. elegans transposable element Tel. Nucl. Acids Res. 7 7, 71377140. Blumenthal, T., Squire, M.. Kirttand, S., Cane, J., Donegan, M., Spieth. J.. and Sharrock, W. (1964). Cloning of a yolk protein gene family from Caenorhabdiris elegans. J. Mol. Biol.. in press.

Rubin. G. M., Kidwetl, M. G.. and Bingham, P. M. (1962). The molecular basis of P-M hybrid dysgenesis: the nature of induced mutations. Cell 29, 987-994.

Bregliano, J., and Kidwell, M. (1983). Hybrid dysgenesis determinants. In Mobile Genetic Elements, J. A. Shapiro, ed. (New York: Academic Press), pp. 363-410.

Southern, fragments

Brenner. S. (1974). The genetics 71-94.

of CaenorhaWifis




Byerly, L.. Cassada. R. C., and Russell, Ft. L. (1976). The life cycle of the nematode Caenorhabditis e/egans. I. Wild-type growth and reproduction. Dev. Biol. 57, 23-33. Collrns, M.. and Rubin, G. M. (1963). High-frequency precise excision the Drosophila foldback transposable element. Nature 303, 259-260. Eddy, E. M. (1975). Germ plasm and the differentiation line. Int. Rev. Cytol. 43, 229-260. Emmons. constancy nematode 1337.


of the germ cell

S. W., Klass, M. R., and Hirsh, D. (1979). Analysis of the of DNA sequences during development and evolution of the Caenorhabcfifis elegans. Proc. Nat. Acad. Sci. USA 76, 1333-

Emmons, S. W.. Rosenzweig, B., and Hirsh, D. (1960). Arrangement of repeated sequences in the DNA of the nematode CaenorhaWitis elegans. J. Mol. Biol. 744, 461-500. Emmons, S. W., Yesner, L., Ruan, K.-S., and Katzenberg, D. (1963). Evidence for a transposon in Caenorhabditis elegans. Cell 32, 55-65. Engels, W. R. (1979). Extrachromosomd control of mutability in Drosophila melanogaster. Proc. Nat. Acad. Sci. USA 76, 401 l-4015. Fedoroff, N. (1983). Controlling elements in maize. In Mobile Genetic Elements, J. A. Shapiro, ed. (New York: Academic Press), pp. I-63. Files, J. G., Carr, S., and Hirsh. D. (1963). Actin gene family of Csenorhabditis elegans. J. Mol. Biol. 764, 355-375. Fincham, J. R. S., and Sastry, G. R. K. (1974). maize. Ann. Rev. Genet. 8, 15-50.




Green, M. M. (1967). The genetics of a mutable gene at the white locus of Drosophila melanogesrer. Genetics 56, 467-482. Greenblatt, I. M.. and Brink, R. A. (1962). Twin mutations variegated pericarp in maize. Genetics 47, 469-501.

in medium

Hirsh, D., Cppenheim, D., and Klass, M. (1976). Development of the reproductive system of Csenorhabditis elegans. Dev. Biol. 49, 200-219. Kimble, J., and Hirsh, D. (1979). The post-embryonic cell lineages of the hermaphrodite and male gonads in Csenorhabditis elegans. Dev. Biol. 70, 396417. Kleckner, N. (1961). Genet. 75, 341-404. Liao, L. W., Rosenzweig,



in prokaryotes.

Ann. Rev.

B.. and Hirsh, D. (1963). Analysis of a transposable

E. M (1975). Detection of specific sequences among DNA separated by gel electrcphoresis. J. Mol. Biol. 98, 503-517.

Strome. S., and Wood, W. B. (1982). lmmunofluorescence visualization of germ line-specific cytoplasmic granules in embryos, larvae, and adutts of Caenorhabdiiis eiegans. Proc. Nat. Acad. Sci. USA 79. 1556-1562. Sulston. J. E., and Brenner. S. (1974). The DNA of Caenorhabditis Genetics 77, 94-104.


Sulston, J. E., and Horvitz, H. R. (1977). Post-embryonic cell lineages of the nematode Caenorhabditis elegans. Dev. Biol. 56, 110-156. Sulston, J. E., Schierenberg, E., White, J. G., and Thomson, J. N. (1963). The embryonic cell lineage of the nematode Caenorhakditis e/egans. Dev. Biol. 700, 64-l 19. Wilson, E. B. (1925). The Cell in Development Macmillan).

and Heredity.

(New York:

Wood, W. B., Hecht, R., Carr, S., Vanderslice, R., Wolf, N., and Hirsh, D. (1980). Parental effects and phenotypic characterization of mutations that affect early development in CaenorhafHitis elegans. Dev. Biol. 74, 446469.