* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Download Control of DNA excision efficiency in Paramecium
Document related concepts
4654–4662 Nucleic Acids Research, 2001, Vol. 29, No. 22 © 2001 Oxford University Press Control of DNA excision efficiency in Paramecium Karine Dubrana1 and Laurence Amar1,2,* 1Laboratoire de Génétique Moléculaire, Ecole Normale Supérieure, 46 Rue d’Ulm, 75230 Paris Cedex 05, France and 2UPRES-A 8080, IBAIC, Bat 444, 91405 Orsay Cedex, France Received July 17, 2001; Revised and Accepted September 26, 2001 ABSTRACT Programmed excision of internal eliminated sequences (IESs) occurs at thousands of sites in ciliate genomes. How this is controlled is largely unknown. Here, we report the characterization of the non-efficiently excised 156ψG-11 IES from Paramecium primaurelia strain 156 and that of the efficiently excised 168ψG-11 IES, an allelic variant from strain 168. Then, we report a genetic and molecular analysis of IES excision efficiency in F1 progeny derived from interstrain crosses and in F2 homozygous progeny derived from F1 autogamy. IES 168ψG-11 excision efficiency was ∼100% in all cases. IES 156ψG-11 excision efficiency was 19 ± 13% in F1 progeny and 0.6 ± 1.1% in F2 progeny. No transexcision event between IESs 156ψG-11 and 168ψG11 was detected within the F1 progeny. These data demonstrate that the excision efficiency of this IES is variable and controlled by a cis-acting element. This should encompass positions 8 and/or 9 of the right IES end, which display allele differences. Finally, the 30-fold stimulation of IES 156ψG-11 excision efficiency within F1 progeny relative to F2 progeny demonstrates that Paramecium IES excision efficiency is sensitive either to a conjugation-specific trans-acting factor provided by the zygotic genome, or to homologous chromosome cross-talk. INTRODUCTION Programmed DNA rearrangements occur in a wide range of genomes. However, in no genome do they occur as frequently as in ciliate genomes. These are rearranged every 100–1000 bp on average, thereby offering great opportunities to study some of the regulatory systems and mechanisms affecting eukaryotic genome plasticity. Ciliates are unicellular eukaryotes that harbor two types of nuclei, macronuclei and micronuclei. These carry out different functions within the cell. Macronuclei are transcriptionally active and carry out vegetative functions. Micronuclei are DDBJ/EMBL/GenBank accession nos AF384101, AF384102 transcriptionally silent and provide genetic continuity between sexual generations. In the course of sexual reproduction the micronuclei undergo meiosis, whereas the macronuclei degenerate. The fusion of two gametic nuclei produces a zygotic nucleus. This nucleus divides twice and the daughter nuclei then differentiate into a micronucleus or a macronucleus. In the second case, the whole genome is processed through chromosome fragmentation, telomere addition to the broken ends and excision of internal eliminated sequences (IESs) that are eventually eliminated. Massive DNA amplification also occurs; this accounts for the large size, and hence the name, of the macronuclei (1,2). In Paramecium, IESs are excised from the micronuclear genome every 1000 bp on average. Since their first identification in 1994, some 80 Paramecium IESs have been characterized. IESs are usually short (26–882 bp), AT-rich (80%) DNA segments. They are invariably surrounded by 5′-TA-3′ repeats, one copy of which is retained at the excision junction. Two IESs have been shown to form extrachromosomal circles in the course of macronucleus development, the IES ends being joined by one 5′-TA-3′ copy (3). How Paramecium IESs are excised and how Paramecium IES excision is controlled are still largely unknown. IES excision can be negatively regulated by a trans nuclear epigenetic mechanism. Injection of one IES into the parental macronucleus results in inhibition of excision of the endogenous IES within the developing macronucleus (4). However, in nine other cases injection of IESs in similar amounts results in either partial inhibition of excision of the endogenous IES or does not result in any excision inhibition (5). Functional analysis of natural and mutagenesis-derived IES alleles has demonstrated that the 5′-TA-3′ dinucleotide invariably flanking IES ends is strictly required for IES excision. A mutation in the 5′-TA-3′ dinucleotide is associated with either defective excision of the mutated IES (6) or a shift of the IES end towards the next two 5′-TA-3′ dinucleotides, located 1 and 5 bp downstream (7). Functional analysis has also shown that a mutation of nucleotide position 5 from one IES (numbering of IES nucleotide positions starts at the 5′-TA-3′ dinucleotide) totally impairs its excision (8). Yet, a previous comparison of IESs demonstrated nucleotide preferences for IES nucleotide positions 3, 4, 5, 6 and 8 (9). These preferences point to a role of these nucleotides in IES recognition and/or excision (9). *To whom correspondence should be addressed at: UPRES-A 8080, IBAIC, Bat 444, 91405 Orsay Cedex, France. Tel: +33 1 69 15 66 38; Fax: +33 1 69 15 68 03; Email: [email protected] Present address: Karine Dubrana, Département de Biologie Moléculaire, Université de Genève, 30 quai Ernest Ansermet, 1205 Genève, Switzerland Nucleic Acids Research, 2001, Vol. 29, No. 22 4655 Figure 1. IESs from the G gene and ψG pseudogene of P.primaurelia. (A) Maps of the micronuclear version of the G gene and ψG pseudogene. IESs are shown in red. Large insertion/deletion differences between the G gene and ψG pseudogene loci are indicated in gray. Probes BS (left) and XH (right) are drawn in blue. (B–D) Sequences of the homologous 156ψG-11, 168ψG-11 and 156G-11 IESs. IESs are flanked by 5′-TA-3 ′ dinucleotides, one copy of which is retained at the macronuclear excision junction. Nucleotides differing between IESs 168ψG-11 or 156G-11 and IES 156ψG-11 are underlined. Note that IES 156G-11 is 1 nt longer than IESs 156ψG-11 and 168ψG-11. The extra nucleotide is shown as an excluded nucleotide to preserve IES sequence alignment. Recently, a determinant(s) lying external to one 28 bp IES, 31–72 bp from its end, has been characterized as essential for its excision (10). In a previous study, we characterized the micronuclear version of the G gene encoding the G surface antigen in Paramecium primaurelia strain 156 and that of a ψG pseudogene (11). By comparing the micronuclear version of the G gene to one of its macronuclear copies, we identified six IESs (Fig. 1A). All of them had paralogs in the ψG pseudogene, i.e. counterparts inserted at homologous positions. However, a comparison of the micronuclear version of the ψG pseudogene with one of its macronuclear copies showed that one of the paralogous IESs, IES 156ψG-11, failed to be excised. These data suggest that IES 156ψG-11 is defective for excision. Here we provide an en masse analysis of IES 156ψG-11 excision indicating that IES 156ψG-11 is excised at low frequency. We show that non-efficient IES 156ψG-11 excision is due to a defective cis-acting element(s). Our data strongly suggest that this element(s) encompasses nucleotide positions 8 and/or 9 of the IES right end, thereby providing the first functional demonstration of the importance of these positions for IES excision. In addition, our data demonstrate that a transacting factor encoded by the zygotic genome participates in the control of IES 156ψG-11 excision efficiency. Together, these data help us understand how IES excision is regulated in Paramecium. MATERIALS AND METHODS Cell lines and culture Paramecium primaurelia strains 156 and 168 are wellcharacterized strains which have been used extensively in genetic studies of the surface antigen system. The karyonides, the daughter cells from conjugant cells, harbor macronuclei produced from single developmental events. All the F1 and F2 progeny cultures, as well as cultures 156/1–2, 156/16–17 and 168/1–6, were monokaryonidal cultures. Cultures 156/3–15 and 156/18 were polykaryonidal. DNA extraction Cultures of exponentially growing cells at 1000 cells/ml were centrifuged. After washing in Dryl’s medium, the pellet was resuspended in 1 vol of the same buffer and added to 4 vol of lysis solution [0.44 M EDTA, pH 9.0, 1% SDS, 0.5% N-laurylsarcosine (Sigma) and 1 mg/ml proteinase K (Merck)] at 55°C. The lysate was incubated at 55°C for at least 5 h, gently extracted with phenol and dialyzed twice against TE (10 mM Tris–HCl and 1 mM EDTA, pH 8.0). DNA analysis and sequencing DNA restriction and electrophoresis were carried out according to standard methods. DNA was blotted onto Hybond N+ membrane (Amersham) in 0.4 N NaOH after depurination in 0.25 N HCl. Hybridization was carried out overnight in 7% SDS, 0.5 M sodium phosphate and 1% bovine serum albumin (BSA) at 60°C for probes BS and HX, at 35°C for oligonucleotide A. Probes BS and HX were labeled using a random priming kit (Boehringer-Mannheim, Mannheim, Germany). Oligonucleotide A was labeled in a kinase reaction. Membranes hybridized with probes BS and HX were washed at 60°C in 0.1% SDS and 0.2× SSC (1× SSC is 0.15 M NaCl and 0.015 M sodium citrate, pH 7.0). Membranes hybridized with oligonucleotide A were washed at 50–58°C in 0.1% SDS and 2× SSC. Hybridization signals were revealed and quantified 4656 Nucleic Acids Research, 2001, Vol. 29, No. 22 using a Fuji phosphorimager. PCRs were performed in volumes of 25 µl comprising 1× PCR buffer supplied by the manufacturer (Promega), 1.25 mM MgSO 4, 50 µM of each dNTP, 1 µM oligonucleotides and 0.8 U Tfl thermostable DNA polymerase. Amplification comprised 32 cycles of 92°C for 1 min, 63°C for 1 min and 72°C for 2 min, with a final extension at 72°C for 3 min. PCR was performed in a Perkin Elmer Cetus thermocycler in capped 0.5 ml polypropylene microcentrifuge tubes (Sigma). DNA cloning and library construction To clone an insert encompassing IES 168ψG-11, PCR was performed using an oligonucleotide that annealed within the macronuclear-destined sequence located upstream from IES 168ψG-11 and an oligonucleotide that annealed within the downstream IES 168ψG1416. Inserts encompassing the IES 168ψG-11 excision site were cloned from a partial DNA library. This library was constructed from a DNA restricted with BglII and size fractionated on a 0.8% agarose gel. The DNA was eluted from the fractions using the GeneClean procedure (Bio 101 Inc.) and screened by PCR for presence of the ψG pseudogene. The >17 kb BglII-digested DNA fraction was selected and cloned into the λ EMBL4 vector. Sequence analysis Sequences were analyzed with the Gap program from the Wisconsin Package v.9.1 (Genetics Computer Group, Madison, WI). Nucleotide sequence accession numbers The GenBank accession no. for IES 156ψG-11 and its flanking sequence is AF384101; for IES 168ψG-11 and its flanking sequence, AF384102. RESULTS IES 156ψG-11 is non-efficiently excised from macronuclear genomes In a previous study we identified IES 156ψG-11 from P.primaurelia strain 156 as putatively defective for excision (11). To accurately define the IES 156ψG-11 excision pattern, we have analyzed macronuclear DNA molecules from 15 cell cultures. PCR was performed using oligonucleotide primers A and B that specifically anneal 60 bp downstream and 225 bp upstream from IES 156ψG-11, respectively (Fig. 2A). Control PCRs were performed on cloned inserts that harbored either the micronuclear version of the ψG pseudogene or an IES-deleted macronuclear copy (see below for a description of this copy). PCR products were then run on an agarose gel and hybridized with oligonucleotide A. PCR products of 285 and 329 bp were expected for IES 156ψG-11-deleted and IES 156ψG-11-harboring fragments, respectively. However, in order to analyze the PCR data correctly, we first characterized the ψG pseudogene macronuclear content of the 15 DNAs. Usually, the PCR products generated from whole-cell DNA only reflect the macronuclear DNA content of Paramecium cells as a consequence of the low cell micronuclear to macronuclear DNA ratio (each cell harbors two diploid micronuclei and a 800–1700n macronucleus). The ψG pseudogene has been shown to be under-amplified within the polyploid Figure 2. IES 156ψG-11 non-efficient excision. (A) Diagram of the PCR and hybridization reactions designed to analyze the excision pattern of IESs 156ψG-11 and 168ψG-11. PCRs were performed using oligonucleotides A and B. The PCR products were electrophoresed and then hybridized with oligonucleotidic probe A. (B) PCR products generated from strain 156 and 168 cultures. (C) PCRs products generated from F2 progeny. The upper PCR products identified IES-harboring fragments; the lower PCR products identified IES-deleted fragments. Control PCRs were performed on DNA inserts that encompassed the ψG pseudogene micronuclear version from strain 156 (lane Mic ψG) or a ψG pseudogene macronuclear copy from strain 168 (lane Mac ψG). Note that no PCR products were generated from inserts harboring the G gene micronuclear version (lane Mic G) or a G gene macronuclear copy (lane Mac G). macronuclear genome (11). Therefore, it was important to check that the 329 bp PCR product always originated from macronuclear DNA templates. The ψG pseudogene macronuclear content was analyzed relative to the G gene macronuclear content (this content was constant amongst all DNAs). DNAs 156/1–15 were restricted with BglII, electrophoresed and probed with fragment BS. This fragment, derived from the DNA located >3.0 kb upstream from the 5′-end of the ψG pseudogene, hybridized to DNA from both the G and ψG loci (Fig. 1A). Signals of 2.8, 5.1–6.7, 8.0–9.7, 11.0–>12.2 and >>12.2 kb were identified (Fig. 3A). No significant levels of alternative forms of DNA rearrangement could be detected over a >200 kb region that encompassed the G gene (12). Therefore, the 2.8 kb signal, which was the only signal to be revealed in all DNAs, must identify a fragment from the G locus, and the four signals of 5.1–6.7, 8.0–9.7, 11.0–>12.2 and >>12.2 kb, which were present in only some DNAs, must identify four fragment families from the ψG locus. Families of fragments differing by 1–2 kb have been exclusively associated with DNA fragmentation in Paramecium (programmed DNA fragmentation occurs every 300 kb on Nucleic Acids Research, 2001, Vol. 29, No. 22 4657 Figure 4. Macronuclear DNA molecules arising from the ψG locus. Hatched boxes defined the nucleotide positions at which telomeric repeats are added. From top to bottom, macronuclear DNA molecules end 1.0–2.6 kb upstream from the ψG pseudogene start and 0.3–2.0, 3.3–4.3 and >15 kb downstream from it. IES 156ψG-11, which is retained on most molecules, is shown in red. Probes BS (left) and XH (right) are drawn in blue. Figure 3. Analysis of the ψG pseudogene macronuclear content from (A) cultures of strain 156 and (B) F1 progeny arising from crosses between strains 156 and 168. Genomic DNA was restricted with BglII, separated by electrophoresis on a 0.8% agarose gel and probed with probe BS derived from the 5′-side of the ψG pseudogene (see Fig. 1). The 2.8 kb signal identified a fragment from the G locus. The signals of 5.1–6.7, 8.0–9.7, 11.0–>12.2, >11.2 and >>12.2 kb identified fragment families from the ψG locus. Probe BS faintly cross-hybridized with two bands indicated with dots. F1 progeny 1–4 were raised from the 156/18 × 168/5 mating, F1 progeny 5–8 from the 156/18 × 168/6 mating, F1 progeny 9–12 from the 156/18 × 168/7 mating and F1 progeny 13–16 from the 156/18 × 168/8 mating. Macronuclear genomes from F1 progeny 1–2, 5–6, 9–10 and 13–14 developed within 168 parents, those from F1 progeny 3–4, 7–8, 11–12 and 15–16 developed within 156 parents. average within the developing macronuclei). Heterogeneity of 1–2 kb reflects heterogeneity in the nucleotide position to which telomeric repeats are added, and heterogeneity in the length of the added telomeric repeats. Therefore, signals of 5.1–6.7, 8.0–9.7 and 11.0–>12.2 kb should identify families of telomeric fragments ending 1.0–2.6 kb upstream of the ψG pseudogene start and 0.3–2.0 and 3.3–4.3 kb downstream of it (Fig. 4). The band >>12.2 kb should identify the full-size BglII restriction fragment derived from DNA molecules ending >12.0 kb from the ψG pseudogene start. PCRs performed with a genomic oligonucleotide and an oligonucleotide corresponding to Paramecium telomeric repeats confirmed that macronuclear molecules were telomerized on both sides of the ψG pseudogene start (data not shown). Only the 11.0–>12.2 and >>12.2 kb signals identified fragments long enough to encompass IES 156ψG-11 with any certainty. The 5.1–6.7 kb signal identified fragments ending far upstream of the ψG pseudogene. The 8.0–9.7 kb signal identified fragments ending close to the ψG pseudogene start. However, size estimation from the Southern blot analysis was not precise enough to ensure that some of these fragments did not end immediately upstream of the ψG pseudogene start, within IES 156ψG-11 itself. No 11.0–>12.2 and >>12.2 kb signals were detected within DNAs 156/8–14, indicating that these DNAs lack a significant level of ψG pseudogene within their macronuclear genome. In contrast, DNAs 156/1–7 and 156/15 displayed 11.0–>12.2 and >>12.2 kb signals. In these DNAs, ratios of the 2.8 kb signal to the 11.0–>12.2 and >>12.2 kb signals were >1:0.10 (Table 1, A). Further analysis showed that a ratio of 1:0.10 defined a ψG pseudogene macronuclear content >80n (data not shown). Therefore, at least 8 of the 15 DNAs (DNAs 156/1–7 and 156/15) harbored a ψG pseudogene macronuclear content large enough to analyze IES 156ψG-11 excision. No ∼300 bp PCR product was detected for DNAs 156/1–7, 156/9, 156/11–12 and 156/15 (Fig. 2B); instead, a PCR product of ∼330 bp was detected. In the case of DNAs 156/9 and 156/11–12, the lack of IES-deleted fragments could be accounted for by a lack of ψG pseudogene macronuclear copies. However, this cannot be the case for DNAs 156/1–7 and 156/15, since these DNAs harbored a ψG pseudogene macronuclear content >80n. In contrast to the PCRs performed on DNAs 156/1–7, 156/9, 156/11–12 and 156/15, the PCRs performed on DNAs 156/8, 156/10 and 156/13–14 generated both ∼300 and ∼330 bp PCR products. These were quantified. The ∼300 bp PCR product that identified IES 156ψG-11deleted fragments accounted for 2–15% of the total PCR products (Table 1, A). These fragments harbored a canonical excision junction that corresponded to one of the IES flanking 5′-TA-3′ dinucleotides (Fig. 1B). Although the more sophisticated techniques available for quantitative PCR could have provided a better estimation of the relative amount of IES-deleted and IES-retaining fragments, the classical PCR technique used here clearly established that IES 156ψG-11 was sporadically excised. Together, these data demonstrate that IES 156ψG-11 is a partially defective IES, characterized by a low excision efficiency. IES 156ψG-11 has an efficiently excised allelic counterpart To identify the factor(s) responsible for non-efficient IES 156ψG-11 excision, we searched for an IES allelic variant 4658 Nucleic Acids Research, 2001, Vol. 29, No. 22 Table 1. IES 156ψ-11 excision efficiency from micronuclear genomes and ψG pseudogene amplification level within macronuclear genomes of (A) cultures from P.primaurelia strain 156, (B) F1 progeny arising from matings between strains 156 and 168 and (C) F2 progeny arising from F1 autogamy (only F2 progeny harboring the ψG pseudogene 156 sequence are shown) A B C Culture ψG pseudogene content in macronuclear genome IES 156ψG-11 excision efficiency 156/3 0.27 0 156/4 0.40 0 156/5 0.43 0 156/6 0.20 0 156/7 0.31 0 156/8 0 0.02a 156/9 0 0 156/10 0 0.06a 156/11 0 0 156/12 0 0 156/13 0 0.06a 156/14 0 0.15a 156/15 0.10 0 156/1 0.25 0 156/2 0.24 0 156/18 0.05 0.05 156/19 0.11 0.13 F1 /3 0.22 0.10 F1 /4 0.30 0.22 F1 /7 0.13 0.23 F1 /8 0.18 0.09 F1 /11 0.12 0.33 F1 /12 0.10 0.36 F1 /15 0.18 0 F1 /16 0.11 0 Mean 0.17 0.17 Standard deviation 0.07 0.14 F1 /1 0.45 0.10 F1 /2 0.33 0.16 F1 /5 0.45 0.24 F1 /6 0.39 0.38 F1 /9 0.37 0.15 F1 /10 0.36 0.11 F1 /13 0.32 0.43 F1 /14 0.26 0.14 Mean 0.35 0.21 Standard deviation 0.06 0.13 F1 mean 0.19 F1 standard deviation 0.13 F2 /510 0.37 0.02 F2 /513 0.66 0 F2 /514 0.34 0 F2 /515 0.40 0 F2 /520 0.45 0 F2 /79 0.33 0 F2 /228 0.28 0.03 F2 /23 0.46 0 F2 /28 0.32 0.01 Mean 0.40 0.01 Standard deviation 0.12 0.01 displaying high excision efficiency. We cloned and sequenced IES 168ψG-11 from P.primaurelia strain 168 (Fig. 1C) and analyzed the DNA from four cell cultures. PCRs were performed and analyzed as above using oligonucleotides A and B, which are common to the 156 and 168 alleles (Fig. 2A). A unique signal of ∼300 bp was revealed in all PCRs (Fig. 2B). No ∼330 bp fragment, which would have been produced from IES 168ψG-11-harboring macronuclear molecules, was identified. This demonstrates that IES 168ψG-11 is excised from ∼100% of the macronuclear DNA molecules. IESs 156ψG-11 and 168ψG-11 display allele-specific excision efficiencies The excision efficiencies of IESs 156ψG-11 and 168ψG-11 were characterized in F 1 progeny arising from crosses between strains 156 and 168. Conjugation in ciliates is a reciprocal process which results in the formation of a genetically identical zygotic nucleus in each of the two parental cells. Each parent then divides into two karyonidal cells, each of which harbors an independently developed macronuclear genome. Of the four progeny arising from an interstrain cross, two harbor macronuclear genomes formed in the context of the 156 parent and two harbor macronuclear genomes formed in the context of the 168 parent. In the case where a trans-acting factor(s) encoded by the 156 parental genome determines non-efficient IES 156ψG-11 excision, IESs 156ψG-11 and 168ψG-11 should be non-efficiently excised in progeny developed within parent 156 and efficiently excised in progeny developed within parent 168. Where a trans-acting factor(s) encoded by the zygotic genome determines non-efficient IES 156ψG-11 excision, the excision of both IESs 156ψG-11 and 168ψG-11 should be efficient (or both non-efficient, depending on which factor-encoding allele is recessive), whatever the parent in which the progeny genomes have developed. In contrast, in the case where a cis-acting factor(s) determines non-efficient IES 156ψG-11 excision, this non-efficient excision should be allele specific and occur in progeny from both parents. We studied the DNA from 16 progeny obtained from four interstrain matings. In a preliminary step, we determined the ψG pseudogene to G gene macronuclear content as above. Probing BglII-restricted DNAs from progeny F1/1–16 and from parents 156 and 168 with fragment BS revealed the 2.8 kb signal that is specific for the G locus and the 5.1–7.1, 7.1–10.2 and >11.2 kb signals that are specific for the ψG locus (Fig. 3B). The >11.2 kb signal identified macronuclear molecules ending >3.3 kb downstream of the ψG pseudogene start (see Fig. 4). The ratio of the 2.8 to >11.2 kb signal ranged within the interval 1:0.10–1:0.45 (Table 1B). This indicated that DNAs from F1 progeny displayed a ψG pseudogene macronuclear IES 156 ψ-11 excision efficiency is given in percent. It was established by quantifying the IES 156 ψ-11-deleted PCR products relative to whole PCR products. The ψG pseudogene macronuclear content is quantified relative to the G gene macronuclear content. A ratio of 1:0.10 defines a ψG pseudogene macronuclear content >80n. aNote that in the case of DNAs 156/8, 156/10 and 156/13–14, quantification of the IES 156 ψ-11-deleted PCR products did not allow calculation of IES 156 ψ-11 excision efficiency because the ratio of the micronuclear to macronuclear DNA was <<80n. Nucleic Acids Research, 2001, Vol. 29, No. 22 4659 Figure 5. Analysis of the ψG pseudogene 156 and 168 sequences in F1 progeny. Genomic DNAs were restricted with EcoRI and HindIII, separated by electrophoresis on a 0.8% agarose gel and hybridized with probe HX derived from the 5′-side of the ψG pseudogene (see Figs 1A and 4). content >80n. In addition, we determined the levels of the 156 and 168 ψG pseudogene sequences in the F1 macronuclear genomes. DNAs were restricted with HindIII and EcoRI, electrophoresed and probed with fragment HX derived from the IES 156ψG-11 nearby upstream region (see Fig. 1A). The sequence of fragment HX was common to the 156 and 168 sequences with the exception of one nucleotide position. Signals of 0.5 and 0.6 kb were revealed, which identified the 156 and 168 ψG pseudogene sequences, respectively (Fig. 5). Both signals were quantified (data not shown). Although sequence 168 displayed a short amplification advantage over sequence 156, both were nearly equal within macronuclear genomes (41 ± 7 versus 59 ± 9%). Finally, the analysis of F2 progeny showed that the 156 and 168 sequences were similarly fragmented on both sides of the ψG pseudogene (see below). These data indicate that DNA from all the F1 progeny harbored a macronuclear content of the 156 and 168 ψG pseudogene sequences large enough to allow analysis of IES 156ψG-11 and 168ψG-11 excision. To determine the excision efficiency of IESs 156ψG-11 and 168ψG-11 in the F1 progeny, PCRs were done using oligonucleotides C and D, which are common to the 156 and 168 alleles. These oligonucleotides generated PCR products that encompassed the HindIII and FokI restriction site polymorphisms of the 156 and 168 sequences. Oligonucleotide C annealed 519 and 520 bp upstream of IESs 168ψG-11 and 156ψG-11, respectively; oligonucleotide D annealed 440 bp downstream of them. Control PCRs were performed on the same cloned inserts as above (the ψG pseudogene micronuclear version derived from strain 156 and the ψG pseudogene macronuclear copy derived from strain 168). The PCR products were then restricted with HindIII and FokI, electrophoresed and hybridized with the internal oligonucleotide A. PCR products of 698 and 742 bp were expected for the 156ψG11-deleted and 156ψG-11-harboring fragments, respectively, and PCR products of 961 and 1005 bp for the 168ψG-11deleted and 168ψG-11-harboring fragments, respectively. PCR products of ∼700 and ∼750 bp were detected in all F1 DNAs except for DNAs F1/15–16, which lacked PCR products of ∼700 bp (Fig. 6). A PCR product of ∼1 kb that could identify IES 168ψG-11-deleted or IES 168ψG-11-harboring DNA fragments was detected in all DNAs. An analysis of the PCR products generated in other PCRs indicated that the PCR products of ∼1 kb identified IES 168ψG-11-deleted fragments (data not shown). No signal of ∼800 or ∼850 bp, which would have attested to recombination between the FokI and HindIII restriction sites and thereby recombination between the flanking sequences of IESs 156ψG-11 and 168ψG-11, was detected, except perhaps in DNA F1/7. The size difference (44 bp) between IES 156ψG-11-deleted (962 bp) and IES 156ψG-11-retaining (1006 bp) templates is <5%. In our hands, PCRs performed with two mixed cloned inserts of ∼1 kb that have an ∼5% size difference generate PCR products with similar efficiency (7). Therefore, quantification of the restriction fragments of ∼700 and ∼750 bp was performed and the relative proportion of IES 156ψG-11deleted templates was calculated for each DNA (Table 1B). IES 156ψG-11 excision efficiency was similar between macronuclear genomes developed within 156 (17 ± 14%) and 168 parents (21 ± 13%). In summary, IESs 156ψG-11 and 168ψG-11 retained their parental pattern of excision in F1 progeny, IES 156ψG-11 excision efficiency remaining far lower (19 ± 13%) than IES 168ψG-11 excision efficiency (∼100%). Nevertheless, IES 156ψG-11 excision efficiency from the F1 progeny was significantly stimulated, compared to that from cultures of the 156 parental strain (2 ± 4%). Stimulation of IES 156ψG-11 excision efficiency did not exhibit any parent dependence (17 ± 14 versus 21 ± 13%), thereby excluding the possibility that it had been caused by a strainspecific trans-acting maternal factor(s). Altogether, these data demonstrate that a cis-acting determinant(s) is responsible for non-efficient IES 156ψG-11 excision and that a trans-acting factor encoded by the zygotic genome participates in the control of IES 156ψG-11 excision efficiency. Figure 6. Excision efficiency from IESs 156ψG-11 and 168 ψG-11 in F 1 progeny. The PCR products were restricted with FokI and HindIII, which identified restriction site polymorphisms between the 156 and 168 sequences, separated on a 1.5% agarose gel and hybridized with oligonucleotide A. The upper PCR products identified IES 168 ψG-11-deleted fragments. The intermediate PCR products identified IES 156ψG-11-harboring fragments. The lower PCR products identified IES 156 ψG-11-deleted fragments. The nature of the minor PCR product indicated by the black dot is not known. Control PCRs were as in Figure 2. 4660 Nucleic Acids Research, 2001, Vol. 29, No. 22 IES 156ψG-11 displays different excision efficiencies between F1 and F2 progeny To characterize the zygotic trans-acting factor(s) that participates in the control of IES excision, we have investigated the genetic linkage of the locus encoding this factor(s) to the ψG-11 site. This was done by characterizing the excision efficiency from IESs 156ψG-11 and 168ψG-11 in F2 progeny derived from F1 autogamy. The zygotic nuclear genome that forms during autogamy results from fusion of two daughter nuclei arising from mitosis of one haploid nucleus. As a consequence, the genome of the F2 autogamous progeny is entirely homozygous. We studied the DNA of 30 progeny arising from autogamy of four F1 progeny. The DNAs were first typed for the ψG pseudogene allele by looking for the HindIII restriction site, which maps ∼350 bp upstream of IES 156ψG-11 (data not shown). Nine progeny harbored the 156 allele and 21 progeny the 168 allele. The DNAs were then analyzed for ψG pseudogene macronuclear content as above (data not shown). All DNAs exhibited signals identifying G and ψG sequences, the ratios of which were >1:0.10 (ratios from progeny harboring the ψG pseudogene 156 allele are shown in Table 1C). Therefore, all DNAs harbored a ψG pseudogene macronuclear content large enough to analyze IES 156ψG-11 and 168ψG-11 excision. The DNAs were then used to perform PCRs with oligonucleotides A and B (Fig. 2C). An ∼300 bp PCR product was generated from the DNA of all 21 progeny harboring the 168 allele. No ∼330 bp PCR product was identified. This indicated an ∼100% excision efficiency for IES 168ψG-11. In contrast, almost no PCR product of ∼300 bp was produced from the DNA of the nine progeny harboring the 156 allele. In these DNAs a PCR product of ∼330 bp was mostly or exclusively generated. Both PCR products were quantified and an IES 156ψG-11 excision efficiency of 0.6 ± 1.1% was established (Table 1, C). These data confirm that a cis-acting determinant(s) is responsible for non-efficient IES 156ψG-11 excision. Furthermore, they indicate that if another locus than the ψG locus constantly participates in the control of IES 156ψG-11 excision, this locus is closely linked to the ψG locus (see Discussion). DISCUSSION Non-efficient IES excision Macronucleus differentiation in ciliates involves the excision of IESs that are eventually eliminated. In a previous study we characterized six IESs from the G gene of P.primaurelia strain 156 and their six paralogs from a ψG pseudogene. Comparison of the micronuclear version of the G gene and ψG pseudogene to one of their macronuclear copies showed that one of the 12 IESs, IES 156ψG-11, was defective for excision. Here we have provided an en masse analysis of IES 156ψG-11 excision efficiency. By analyzing 15 independent cultures from strain 156, we have shown that IES 156ψG-11 excision is non-efficient rather than null. Having identified an efficiently excised IES 156ψG-11 allelic variant, we have performed a genetic analysis of the determinant(s) responsible for nonefficient IES 156ψG-11 excision. We have established that non-efficient excision of IES 156ψG-11 does not result from epigenetic control involving maternal factors. Indeed, efficient IES 156ψG-11 excision never occurs within F1 progeny, even for those that developed within 168 parents. Conversely, efficient IES 168ψG-11 excision occurs in all F1 progeny, even those that developed within 156 parents. Furthermore, only efficient IES 168ψG-11 excision and non-efficient IES 156ψG-11 excision occur within F2 progeny. Together, these data demonstrate that a defective cis-acting determinant(s) is responsible for non-efficient IES 156ψG-11 excision in strain 156. Until now, Paramecium IESs have been assumed to display an all-or-nothing excision pattern, although most of them were characterized by comparison of a micronuclear DNA version with a unique macronuclear copy. En masse analysis of Paramecium IES excision has only shown sporadically retained IESs (13). Further, the view of an all-or-nothing excision pattern was strengthened by characterization of mutant IESs that were fully defective for excision rather than partially defective. Our data are the first demonstration that IES excision may be a non-efficient process and that IES excision efficiency is controlled. Since IES insertion is frequent within genes, the control of IES excision efficiency constitutes a way of regulating the number of transcription templates in the highly polyploid macronuclear genomes of ciliates. The cis-determinant(s) of IES excision efficiency A recent analysis of a series of molecules injected into developing macronuclei of P.tetraurelia has shown that DNA flanking the 28 bp 51A2591 IES is necessary for excision of this IES (10). Sequences flanking allelic IESs 156ψG-11 and 168ψG-11 are >99% identical; the first nucleotide differences outside these IESs map 165 bp upstream and 335 bp downstream of their left and right boundaries, respectively. In the case where non-efficient excision of IES 156ψG-11 is caused by an external cis-acting determinant(s), the above nucleotide positions would be candidates for defining them. However, given the distances that separate these nucleotide positions and the IES ends, it seems more plausible that differences in IESs 156ψG-11 and 168ψG-11 excision efficiency result from differences in the IES 156ψG-11 and 168ψG-11 sequences. IESs 156ψG-11 and 168ψG-11 are identical except at right end positions 8 and 9 (see Fig. 1B and C). A difference in right end position 8 by itself could be responsible for the difference in IES excision efficiency since IES position 8 displays a nucleotide preference, an indication of its importance for IES excision (9). Only left position 8 of IES 168ψG-11 fits the preferred nucleotide. Alternatively, mutations at both left positions 8 and 9 could account for non-efficient IES 156ψG-11 excision. This hypothesis would agree with the observation that the paralogous IES 156G-11, the right end of which is identical to that of IES 168ψG-11, is also efficiently excised (Fig. 1D). If differences in IESs 156ψG-11 and 168ψG-11 excision efficiency result from differences in the IES 156ψG-11 and 168ψG-11 sequences, our data provide the first functional demonstration of the importance of IES nucleotide positions 8 and 9 for IES excision. The zygotic determinant(s) of IES excision F1 and F2 progeny display a 30-fold difference in IES 156ψG-11 excision efficiency. This is true whether F1 progeny develop from 156 or 168 parents. A cut at one end of IES 168ψG-11 Nucleic Acids Research, 2001, Vol. 29, No. 22 4661 could generate a 3′-OH terminus able to trans-attack one end of IES 156ψG-11 and thereby trigger an enhancement of IES 156ψG-11 excision efficiency. However, the joining of IESs 156ψG-11 and 168ψG-11 flanking sequences should have produced 156 and 168 chimeric macronuclear DNA molecules. No such chimeric molecules were detected, indicating that enhancement of IES 156ψG-11 excision efficiency did not result from IES 156ψG-11 and 168ψG-11 trans-excision. On the other hand, difference in IES 156ψG-11 excision efficiency between F1 and F2 progeny cannot be the consequence of a difference in the amplification of IES 156ψG-11-deleted and IES 156ψG-11-retaining molecules: IES 156ψG-11-retaining and IES 168ψG-11-deleted allelic molecules were similarly amplified within the macronuclear genome of the F2 progeny. Rather, the difference in IES 156ψG-11 excision efficiency between F1 and F2 progeny must be a consequence of a transacting factor differently provided by the zygotic genome in autogamy and conjugation. A conjugation-specific factor(s) could be responsible for the enhanced excision efficiency of IES 156ψG-11 within F1 progeny (the F2 progeny arose from autogamy). Alternatively, IES 156ψG-11 excision efficiency enhancement in F1 progeny may result from the action of a factor encoded by the 168 allele of a locus closely linked to the ψG locus. Although this could occur by chance, it seems more plausible that IES 156ψG-11 excision efficiency enhancement in F1 progeny results from cross-talk between the 156 and 168 ψG pseudogene sequences. Cross-talk between alleles is known to modulate many different mechanisms, including gene expression in Drosophila, chromatin structure and meiotic double-strand break frequency in Saccharomyces cerevisiae (14,15) and transposon excision in plants (16). Homologous chromosome cross-talk may involve indirect physical interactions via soluble factors or direct physical interactions. Somatic pairing of the homologous chromosomes regularly occurs in the developing macronuclei of some hypotrichous ciliates. Although no such pairing has ever been characterized in the holotrichous ciliates Paramecium and Tetrahymena, the characterization of recombinant DNA molecules within the developing macronuclear genome of Tetrahymena attests to physical interactions between homologous chromosomes in their genomes (17). IES excision and chromosome fragmentation In Paramecium, programmed chromosome fragmentation usually occurs within DNA domains spaced by a few kilobases. We have shown here that four fragmentation domains map over the ψG locus: one domain maps 1.0–2.6 kb upstream of the ψG pseudogene start, two domains within the ψG pseudogene and the last downstream of its 3′-end. A large difference in the use of these four DNA domains between karyonides must account for the previously reported large variation in the ψG pseudogene macronuclear content (11). Determinants of DNA fragmentation are unknown in Paramecium. Although the G and ψG paralogous loci have arisen from duplication of an ancestral genomic locus, their patterns of fragmentation are highly distinct. None of the four fragmentation DNA domains located around the ψG pseudogene has a paralog around the G gene, except that lying at the ψG pseudogene 3′-end, the paralog of which is non-efficiently used (13). Macronuclear molecules that end within this paralog retain a 67 bp IES at their ends, or close to them. This has led to the proposal that chromosome fragmentation and IES excision could be alternative DNA rearrangements: the retained IESs would act as a cis-acting determinant for fragmentation by either providing accurate sites to cut or inducing cut sites in their flanking sequences. In that hypothesis, the non-efficiently excised IES 156ψG-11 could act as a cis-acting determinant for DNA fragmentation. However, IES 156ψG-11 is retained not only on macronuclear molecules ending close to it, but also on those ending much further downstream. Furthermore, the 168 sequence displays the same DNA fragmentation pattern as the 156 sequence, although IES 168ψG-11 is fully excised. Therefore, IES 156ψG-11 does not appear to cis-determine any fragmentation event. ACKNOWLEDGEMENTS We thank Rick Willet and Anne Baroin for correcting the manuscript. This work was supported by the Centre National de la Recherche Scientifique (grant no. 97N63/0016), the Ministère de l’Education Nationale, de la Recherche et de la Technologie (Programme de Recherche Fondamentale en Microbiologie et Maladies Infectieuses et Parasitaires), the Groupement de Recherches et d’Etudes sur les Génomes (grant no. 22/95) and the Association pour la Recherche sur le Cancer (grant no. 1374). K.D. was the recipient of a doctoral fellowship from the Association pour la Recherche sur le Cancer. REFERENCES 1. Prescott,D.M. (2000) Genome gymnastics: unique modes of DNA evolution and processing in ciliates. Nature Rev. Genet., 1, 191–198. 2. Klobutcher,L.A. and Herrick,G. (1997) Developmental genome reorganization in ciliated protozoa: the transposon link. Prog. Nucleic Acid Res. Mol. Biol., 56, 1–62. 3. Betermier,M., Duharcourt,S., Seitz,H. and Meyer,E. (2000) Timing of developmentally programmed excision and circularization of Paramecium internal eliminated sequences. Mol. Cell. Biol., 20, 1553–1561. 4. Duharcourt,S., Butler,A. and Meyer,E. (1995) Epigenetic self-regulation of developmental excision of an internal eliminated sequence on Paramecium tetraurelia. Genes Dev., 9, 2065–2077. 5. Duharcourt,S., Keller,A.M. and Meyer,E. (1998) Homology-dependent maternal inhibition of developmental excision of internal eliminated sequences in Paramecium tetraurelia. Mol. Cell. Biol., 18, 7075–7085. 6. Mayer,K.M. and Forney,J.D. (1999) A mutation in the flanking 5 ′-TA-3′ dinucleotide prevents excision of an internal eliminated sequence from the Paramecium tetraurelia genome. Genetics, 151, 597–604. 7. Dubrana,K., Le Mouel,A. and Amar,L. (1997) Deletion endpoint allelespecificity in the developmentally regulated elimination of an internal sequence (IES) in Paramecium. Nucleic Acids Res., 25, 2448–2454. 8. Mayer,K.M., Mikami,K. and Forney,J.D. (1998) A mutation in Paramecium tetraurelia reveals functional and structural features of developmentally excised DNA elements. Genetics, 148, 139–149. 9. Klobutcher,L.A. and Herrick,G. (1995) Consensus inverted terminal repeat sequence of Paramecium IESs: resemblance to termini of Tc1related and Euplotes Tec transposons. Nucleic Acids Res., 23, 2006–2013. 10. Ku,M., Mayer,K. and Forney,J.D. (2000) Developmentally regulated excision of a 28-base-pair sequence from the Paramecium genome requires flanking DNA. Mol. Cell. Biol., 20, 8390–8396. 11. Dubrana,K. and Amar,L. (2000) Programmed DNA under-amplification in Paramecium primaurelia. Chromosoma, 109, 460–466. 12. Caron,F. (1992) A high degree of macronuclear chromosome polymorphism is generated by variable DNA rearrangements in Paramecium primaurelia during macronuclear differentiation. J. Mol. Biol., 225, 661–678. 13. Amar,L. (1994) Chromosome end formation and internal sequence elimination as alternative genomic rearrangements in the ciliate Paramecium. J. Mol. Biol., 236, 421–426. [Erratum, 1997, J. Mol. Biol., 265, 465.] 4662 Nucleic Acids Research, 2001, Vol. 29, No. 22 14. Xu,L. and Kleckner,N. (1995) Sequence non-specific double-strand breaks and interhomolog interactions prior to double-strand break formation at a meiotic recombination hot spot in yeast. EMBO J., 14, 5115–5128. 15. Rocco,V. and Nicolas,A. (1996) Sensing of DNA non-homology lowers the initiation of meiotic recombination in yeast. Genes Cells, 1, 645–661. 16. van Houwelingen,A., Souer,E., Mol,J. and Koes,R. (1999) Epigenetic interactions among three dTph1 transposons in two homologous chromosomes activate a new excision-repair mechanism in petunia. Plant Cell, 11, 1319–1336. 17. Deak,J.C. and Doerder,F.P. (1998) High frequency intragenic recombination during macronuclear development in Tetrahymena thermophila restores the wild-type SerH1 gene. Genetics, 148, 1109–1115.