Evidence for a Multistep Control in Transposition of I Factor in Drosophila melanogaster

Corresponding author: Christophe de La Roche Saint André, Institut de Biologie du Développement de Marseille, Marseille-Luminy, Case 907, 13288 Marseille Cedex 9, France, laroche{at}ibdm.univ-mrs.fr (E-mail).

Communicating editor: R. S. HAWLEY

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Drosophila melanogaster strains belong to one of two interactive categories, inducer (I) or reactive (R), with respect to the I-R system of hybrid dysgenesis. The dysgenic interaction results from the presence of several transposition-competent copies of a LINE-like element, the I factor, only in the genome of I strains. When a cross is performed between I males and R females, I factor transposes at high frequency in the germ line of F₁ daughters, known as SF females. This transposition burst results in the sterility of SF females. I factor transposes by reverse transcription of a full-length transcript. Specific RT-PCR experiments were done to compare the amount of I factor transcript in samples corresponding to various transposition frequencies. The sensitivity of the method allowed the ready detection of the I factor RNA in every tissue and genetic background examined. Comparison of amplification signals suggests that I factor activity in ovaries is regulated at different levels. First, the amount of I factor RNA subjected to negative and positive regulation. Whereas the negative control, which limits transposition in nonpermissive contexts, may be exerted by an I factor encoded repressor function, the positive control is linked to reactivity level, a cellular state maternally inherited from R mothers. Additionally, negative regulation is also exerted downstream of I factor RNA. This differs notably from previous conclusions in which transcription was envisaged as the main level of regulation of the I factor transposition.

I-R hybrid dysgenesis was described in the early 1970s (PICARD and L'HERITIER 1971 ). In this system, Drosophila melanogaster strains fall into two categories denoted "inducer" (I) and "reactive" (R). The dysgenesis is observed in F₁ females, resulting from crosses between reactive females and inducer males. These females, called SF females, are partly sterile. Even though the number of eggs laid by these females is normal, the hatching percentage is abnormally low due to blockade of early embryonic development (LAVIGE 1986 ). Moreover, the surviving progeny of SF females exhibits a high level of mutations, chromosomal rearrangements, and nondisjunctions (PICARD et al. 1978 ). Reciprocal crosses produce normally fertile females, known as RSF females, as do crosses within both categories of strains. All crosses give rise to normal male progeny.

I factor, a non-LTR retrotransposon similar to mammalian LINEs, was shown to be responsible for the I-R hybrid dysgenesis phenomenon (FAWCETT et al. 1986 ). Inducer strains carry 5–15 copies of I elements per haploid genome, dispersed on chromosome arms (PELISSON and BREGLIANO 1987 ). Among these copies, some of them are competent for transposition (FAWCETT et al. 1986 ). The reactive strains do not have any I element on their chromosome arms. Both categories of strains bear a similar pattern of defective I elements in the pericentromeric ß-heterochromatin (BUCHETON et al. 1984 ).

The term "chromosomal contamination" was used to describe the acquisition of active I factors by chromosomes of reactive origin. The efficiency of contamination was used as an estimation of I factor transposition. Using this criterion, it was shown that: (1) while being stable in the inducer context, the I factor transposes at high frequency in the germ line of SF females and to a lesser degree in the germ line of RSF females (PICARD 1976 ); (2) the frequency of transposition of I factors in dysgenic crosses is regulated by the reactivity level of the R mothers (PICARD 1978 ). Different experimental data (PROUST and PRUDHOMMEAU 1982 ; PROUST et al. 1992 ) support the view that the fertility defects of SF females are a direct consequence of I factor transposition. Therefore, the quantification of the SF sterility level, measured by the percentage of nonhatching eggs laid by the SF females, can be used as an indirect estimation of I factor transposition frequency in SF females germ line.

The severity of SF sterility in a dysgenic cross may vary within a wide range, depending on the reactive parental stock used in the cross (BUCHETON et al. 1976 ). According to this criterion, one can distinguish reactive strains with a weak, medium or strong reactivity level. Significant characteristics of the reactivity levels are the following: (1) reactivity levels are determined by a cellular state present in the mature oocytes laid by an R female (BUCHETON and BREGLIANO 1982 ), which is inherited mainly maternally from one generation to the next, the chromosomes being the major determinants only after several generations (BUCHETON and PICARD 1978 ); (2) nongenetic factors, such as aging or thermic treatments, can modify the reactivity levels, and these modifications are heritable, cumulative, and reversible (BUCHETON 1979 ); (3) reactivity levels are enhanced by DNA-damaging agents (BREGLIANO et al. 1995 ) and correlate with repair and recombination activities (LAURENCON and BREGLIANO 1995 ; LAURENCON et al. 1997 ), suggesting that reactivity may correspond to a more general biological function beyond the action on I factor transposition.

Transposition of the I factor has been shown to require an RNA intermediate ( JENSEN and HEIDMANN 1991 ; PELISSON et al. 1991 ). The potential molecular intermediate of transposition, an RNA encompassing the full-length I factor, was detected in Northern blot hybridization experiments (CHABOISSIER et al. 1990 ). This RNA was specifically detected in the female germ line and shown to be at higher amounts in SF than in RSF flies. It was concluded from these data that the level of transcription or steady-state I factor RNA is a major determinant in the control of I factor transposition (CHABOISSIER et al. 1990 ). However, this study was limited to the context of a single dysgenic cross, and the influence of the reactivity level has not been investigated. Therefore, we wanted to extend the observations to situations where the genetic background and the reactivity level are different. Because of the paucity of I factor transcript, direct quantitative comparisons between various RNA samples was not possible (CHABOISSIER et al. 1990 ). Anticipating such a technical problem, we have developed a sensitive method based on RT-PCR to quantitate and compare the amounts of I factor RNA more accurately.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Drosophila stocks and culture conditions:
The following strains were employed: est 28 and Cha are strong reactive lines; Paris is a reactive stock (this stock was weakly reactive for years and then became a mix of weak and strong reactive flies); P1m and PF9b are sublines, weakly and strongly reactive, respectively, derived from this stock; Canton-S is a standard inducer stock; Cha-(RC ⁺) is an inducer stock, resulting from transposition of I factors from the genome of an inducer stock to the chromosomes of the Cha reactive strain.

Measurement of sterility level:
Young daughters from each cross were put on fresh food with sib males and allowed to lay eggs during 24 hr. Then they were discarded, and 2 days later the percentage of nonhatching eggs was scored. In the case where flies are SF females, i.e., daughters of reactive females mated with standard inducer males, the percentage of nonhatching eggs measures the reactivity level of the reactive mothers.

RNA extractions:
Ovaries were dissected 4 days after eclosion (except when specified) and stored at -80°, as well as the corresponding male and female carcasses. Total RNA was prepared by a phenol/guanidinium isothiocyanate procedure using TRIzol Reagent (GIBCO-BRL, Grand Island, NY ), according to the manufacturer's instructions. Two hundred fifty µl TRIzol was used for 20 ovaries and 500 µl TRIzol was used for 20 male or 20 female carcasses. After ethanol precipitation, RNA were resolubilized in water. RNA quantity and quality were estimated by UV absorption at OD₂₆₀ nm and by visual inspection after migration on agarose gel and EtBr staining.

Poly(A)⁺ purification:
Poly(A)⁺ RNA was isolated from total RNA preparations using the Dynabeads mRNA purification kit (Dynal). Ten µg total RNA was mixed with 50 µl Dynabeads Oligo(dT)25 for the first round of purification, and half of the eluted RNA was mixed with 50 µl Dynabeads Oligo(dT)25 for the second round of purification.

Reverse transcription and cDNA amplification:
Random-primed cDNA was prepared from 5 µg total RNA using the Stratascript kit (Stratagene, La Jolla, CA) in a final volume of 50 µl. Amplification was carried out in a final volume of 30 µl in the presence of 1 µl of the reverse-transcribed RNAs, 30 µM dNTPs, 3 mM MgCl₂, 0.2 µM primers, and 1 unit Taq polymerase (Eurobio, Les Ulis, France). Samples were subjected (except when specified) to 30 cycles PCR in a Crocodile III thermal cycler (Appligene, Illkirch, France). For I cDNA 5' fragments amplification, each cycle included (except when specified) 30 sec denaturation at 92°, 15 sec annealing at 68°, and 30 sec elongation at 72°. After amplification, 10 µl aliquots were analyzed by agarose gel electrophoresis and BET staining. The sequences of the primers used were: AGA GATAAGTCGTGCCTCTC 22/41-GTACTCGGACTGTTTCG TAC 721/702 (I factor 5' part); CCGACCCATCTCCCTCAACT 3059/3078-TCGCAAGGTCGGCTTTAAGG 4812/4793 (I factor 3' part); AGTCGCCTACAATGGTCTGC 39/58-GTTCGAATC GTTGCTAACGG 1119/1100 (glucose-6-phosphate dehydrogenase); ATGTCAAAGGGGGGCGAGCACAAGCGCGTC 1391/1420-CAACTGTGGCGCACTGGA 2046/2029 (yem-alpha). Positions refer to the sequences reported by FAWCETT et al. 1986 for the I factor, FOUTS et al. 1988 for the glucose-6-phosphate dehydrogenase gene, and AIT-AHMED et al. 1992 for the yem-alpha gene.

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Designing PCR primers specific for active I factors:
Using the RT-PCR method, we aimed at amplification of cDNA fragments specific for the full-length I factor transcript, i.e., the potential intermediate of transposition (CHABOISSIER et al. 1990 ). One major problem to overcome was the presence in the genome of D. melanogaster of multiple copies of I elements, which belong to different classes (BUSSEAU et al. 1994 ). Functional I elements, called I factors, are present only in inducer strains at various sites on chromosome arms. In addition, two kinds of defective I elements can be distinguished: (1) 5'-truncated elements, also present on the chromosome arms of inducer strains, which correspond to recent insertion events. They have lost their 5' end upon transposition, probably because of an early arrest of reverse transcription ( JENSEN and HEIDMANN 1991 ); and (2) I elements that are confined within the pericentromeric ß-heterochromatin. This latter category is present in the genome of all strains, inducer as well as reactive, and probably represents very old components of the genome (BUCHETON et al. 1984 ).

Transcription of the I factor initiates from an internal RNA polymerase II promoter, located at the 5' end (MCLEAN et al. 1993 ). This internal promoter is always deleted in the 5'-truncated euchromatic elements. It is therefore assumed that these elements are unable to be transcribed and thus are inactive, as shown for the longest truncated element described so far (BUSSEAU et al. 1989 ). However, the illegitimate transcription of a truncated sequence inserted near a genomic promoter is possible. To avoid the amplification of such read-through transcripts, we have chosen to amplify a fragment located in the 5' part of the I element sequence, with a 5' primer close to the transcription start site (Figure 1A).

View larger version (32K): In this window In a new window Download PPT slide   Figure 1. —Specific amplification of active I factor cDNA fragments. (a) The structure of the I factor is shown at the top. White boxes represent the two long open reading frames, ORF1 and ORF2. The location of primers used for amplification are indicated. Below are some sequence pieces of the I factor: sequences of the primers (Ig5.4 and Ig3.0) and the RsaI restriction site are underlined. Nucleotides that are different in the defective heterochromatic elements are indicated below the coding strand (Ig5.4 and RsaI) and above the noncoding strand (Ig3.0). Nucleotide numbers refer to the sequence reported by FAWCETT et al. 1986 . (b) RT-PCR analysis of various reactive (R) and inducer (I) strain RNA samples. All amplifications were performed with the same Ig5.4/Ig3.0 pair of primers. Lanes 1, 2, and 3: amplification with est28, Cha, and Paris cDNA samples. Lanes 4, 5, 6, and 7: amplification products obtained with Canton-S and Cha-(RC +) cDNA samples before (lanes 4 and 6) and after (lanes 5 and 7) digestion with RsaI. cDNAs were subjected to 28-cycle amplification, except for Cha-(RC +) cDNAs which were subjected to 26-cycle amplification only. Positions of full-length amplification product (700 bp) and of RsaI restricted fragments (345 and 355 bp) are indicated. (c) Left: RT-PCR with Canton-S RNA with (+) or without (-) reverse transcriptase during the reverse transcription reaction. Right: G6PD amplification made with Canton-S genomic DNA (g) or with reverse-transcribed Canton-S RNA (c).

The pericentromeric defective elements are transcribed in all categories of flies (CHABOISSIER et al. 1990 ). Compared with complete I factors, they are affected by various rearrangements and many point mutations (CROZATIER et al. 1988 ). No complete ORF1 or ORF2 could be found in any of them. The sequenced 5' regions of several heterochromatic elements show similar point mutations, which are not present in the I factor sequence (VAURY et al. 1990 ). We have designed specific primers ending at nucleotides corresponding to these distinctive point mutations in order to specifically amplify active I factor cDNA sequences (Figure 1A). Moreover, an RsaI restriction site located in the middle of the I factor sequence bounded by the primers is absent in the defective elements' sequences (Figure 1A). Therefore, RsaI digestion could be used to confirm the nature of the amplified products.

Specific amplification of active I factors cDNAs fragments:
RNAs prepared from various reactive and inducer strains were reverse transcribed and amplified. As expected, the discrimination between active and defective elements is not absolute and relies greatly on the annealing temperature. At low temperatures (below 60°), in addition to the expected amplification product, various RsaI-resistant fragments are amplified in all cDNAs tested, which likely correspond to defective I elements (not shown). Increase of annealing temperature (above 65°) allows the specific amplification of the RsaI-sensitive product only in cDNA samples corresponding to inducer flies (Figure 1B). Therefore, such a product corresponds to active I factor sequences.

Because no direct distinction could be made between amplified products corresponding to genomic or complementary I DNA sequences, we have checked for the absence of contribution of contaminating genomic DNA to the observed signals. Two types of controls were made (Figure 1C). First, the amplification was conducted on control cDNA reactions in which reverse transcriptase was omitted. The absence of signal (Figure 1C, left) showed that amplification was dependent on the reverse transcription step. Second, the same samples were amplified with primers located on both sides of an intron of the glucose-6-phosphate dehydrogenase (G6PD) gene (Figure 1C, right). The size of the amplified product was the one expected for a cDNA relieved of the intron. Therefore, the amplification products obtained with our I factor-specific primers correspond to cDNAs.

Amount of I factor RNA and transposition frequency:
All possible combinations of crosses were carried out with reactive est28 and inducer Canton-S flies. Crossing of est28 females with Canton-S males produced SF flies, and the reciprocal est28/Canton-S cross gave RSF flies. The nonhatching percentage of the eggs laid by the 3-day-old SF and RSF females was 98 and 2%, respectively. Canton-S x Canton-S and est28 x est28 crosses gave the corresponding inducer and reactive flies. In each case, total RNAs were isolated from young female and male adult progeny. After reverse transcription, the 5' part of I factors cDNAs was specifically amplified (Figure 2A), as demonstrated by the absence of signal with est28 cDNAs. The amount of product amplified with SF or RSF ovary cDNAs was similar and slightly less than with Canton-S ovary cDNAs. Comparable results were obtained following independent RT reactions. G6PD served as a control to confirm equal efficiency of reverse transcription of each of the RNAs. This result was surprising, because it contradicts previous studies (CHABOISSIER et al. 1990 ) describing a very reduced amount of I factor transcript in the ovaries of RSF females compared to those of SF females, and even smaller amounts of I factor RNA in the inducer context.

View larger version (48K): In this window In a new window Download PPT slide   Figure 2. —RT-PCR analysis of RNAs from est28, Canton-S, and the corresponding SF and RSF flies. (a) Amplification of I factor (28 cycles) or G6PD (25 cycles) cDNA fragments from ovaries (O), carcasses (C), and males (M) with reverse-transcribed RNA samples. (b) Amplification kinetics analysis (28, 29, and 30 cycles) with SF and RSF ovary samples. (c) RT-PCR of SF RNAs alone and after two-, three-, and sixfold dilution in est28 RNAs (28-cycle amplification). (d) Amplification of yem-alpha cDNA fragments from SF and RSF samples. O: ovaries; C: carcasses; M: males (27-cycle amplification).

We have tested our RT-PCR reactions for a possible saturation, which could impede any real quantitative comparison. First, SF and RSF signals were compared after different numbers of amplification cycles (Figure 2B). The amount of PCR product increases regularly with the cycle number in a similar fashion for the SF and the RSF samples. Second, RNAs from SF flies were diluted in those of est28 flies, which are devoid of I factor transcript, to obtain qualitatively equivalent RNA samples containing decreasing amounts of I factor RNA. A decrease of the signal is observed from the first dilution (Figure 2C). Moreover, the regular decrease of the signal strength along the dilution range attests to the reliability of the RT step. Third, reducing fourfold the amount of RNA for reverse transcription diminishes the signal strength similarly for the SF and RSF samples (not shown). Therefore, neither the reverse transcription nor the amplification step appear to be saturated in our experimental RT-PCR conditions.

According to the reconstruction experiment (Figure 2C), the amount of I factor RNA present in Canton-S ovaries is about twofold that of the RSF sample. A small difference was observed between the SF and RSF samples in independent RT experiments, the SF signal being slightly more intense than the RSF one. However, this difference was not clearly above the background variation inherent to our experimental conditions (not shown). In addition, quantitative discrimination of a factor of two was effective (Figure 2C) and reproducible. From these observations, we conclude that the difference in the amount of I factor RNA between SF and RSF samples is certainly less than twofold, i.e., far from the difference previously described (CHABOISSIER et al. 1990 ).

The variation in signals obtained with the carcasses or male cDNA samples reproduced those seen with the ovary cDNA samples (Figure 2A). Although any direct comparison between samples belonging to different categories, i.e., ovaries, carcasses, and males has no significance, this is also in contradiction to previous results (CHABOISSIER et al. 1990 ), which show I factor transcripts only in ovary. To confirm our results, amplifications were done with primers corresponding to an ovary-specific gene, yem-alpha. Expression of this gene in the adults has been detected by RNA dot-blot exclusively in the female germ line (AIT-AHMED et al. 1990 ). RT-PCR signals obtained with carcasses and male cDNAs are greatly reduced compared to ovary cDNAs (Figure 2D). Thus, compared to yem-alpha, the expression of I factor is not ovary-specific.

Effect of varying methodological parameters:
The fact that our results are clearly different from those presented earlier (CHABOISSIER et al. 1990 ) could be explained by differences in experimental approaches. We first wondered if the difference was due to the fact that the fragment we amplified concerns only the 5' part of I factor sequence, while previously the transposition intermediate was identified as a full-length transcript (CHABOISSIER et al. 1990 ). In the RSF sample, the amplified 5' fragment could actually correspond to truncated RNAs. Using primers targeting the 3' part of the I factor RNA (up to nt 4812, see MATERIALS AND METHODS) in conditions where no amplification signal is seen in samples corresponding to reactive flies (not shown), the amount of amplified product was similar in SF and RSF cDNAs (Figure 3A). Coamplification of the 5' and 3' fragments shows a similar ratio of the signals whatever the sample analyzed (Figure 3B). If incomplete, the transcripts present in RSF are at least 4.8 kb long and would have been detected in Northern blots (CHABOISSIER et al. 1990 ). Therefore, RNAs detected in our RT-PCR certainly correspond to full-length transcripts.

View larger version (55K): In this window In a new window Download PPT slide   Figure 3. —est28/Canton-S SF and RSF samples: 3' amplification and poly(A)+ purification. (a) Amplification analysis (29 and 31 cycles) of 3' I factor cDNA fragments from ovary reverse-transcribed RNA samples. (b) Coamplification (29 cycles) of 5' (700 bp) and 3' (1753 bp) I factor cDNA fragments from ovary (O) and carcass (C) reverse-transcribed RNA samples. (c) Effect of poly(A)+ purification on ovary RNA samples. Top panel: gel migration analysis of RNAs used for reverse transcription before (lane 0) and after one (lane 1) or two (lane 2) rounds of poly(A)+ purification. Positions for the ribosomal RNAs are indicated (arrowheads). Middle panel: amplification with the I factor primers (29-cycle amplification) of the corresponding reverse-transcribed RNA samples. Bottom panel: amplification with the yem-alpha primers (26-cycle amplification).

Another difference was the nature of RNA samples used. We amplified cDNAs prepared with total RNA, while poly(A)⁺ was used in Northern blot hybridization and S1 mapping experiments (CHABOISSIER et al. 1990 ). We did experiments to check for a possible differential behavior of SF and RSF I factor RNAs during the poly(A)⁺ purification procedure (Figure 3C). Ovary RNAs were reverse-transcribed before and after one or two rounds of poly(A)⁺ purification and then amplified with either I or yem-alpha specific primers. The yem-alpha transcript is polyadenylated (AIT-AHMED et al. 1990 ) and is used as an internal control. The similarity of the yem-alpha signal in all the fractions demonstrates the efficient concentration of poly(A)⁺ during preparation. On the contrary, there is a clear decline of the I signal from the first step of purification, showing that the I factor RNA copurifies only partially with the poly(A)⁺ fraction. No difference is observed in the behavior of the SF and the RSF I factor RNAs. This strongly argues against a bias due to the nature of the RNA fraction used as the main reason for our divergent results.

Effect of the genetic background on I factor RNA level:
The last experimental difference between our work and previous work (CHABOISSIER et al. 1990 ) is biological and related to the genetic context of the strains used. The reactive and inducer flies used in the dysgenic cross were Cha and Cha-(RC ⁺) (CHABOISSIER et al. 1990 ). Cha is a strong reactive line, and the corresponding inducer Cha-(RC ⁺) line was obtained after transposition of I factors from the genome of an inducer stock to the chromosomes of the Cha strain (PELISSON and BREGLIANO 1987 ). We prepared total RNAs from Cha and Cha-(RC ⁺) flies, and the corresponding SF and RSF flies. The nonhatching percentages of eggs laid by the SF and RSF 3-day-old females were 100 and 10%, respectively. Notably, the average amount of I factor RNA detected by RT-PCR was greater than in the corresponding samples of the est28/Canton-S cross, i.e., fewer amplification cycles are required for analysis to avoid saturation. Otherwise, the results are qualitatively similar to those obtained with the est28/Canton-S pair (Figure 4A), except for a clear excess seen with the SF ovary sample. This excess is specific for ovaries, i.e., is not seen with the corresponding carcass or male samples (data not shown). To estimate the magnitude of this difference, RT-PCR was made after serial dilution of SF ovary RNAs into those of Cha, which are devoid of any I factor transcript (Figure 4B). A signal equivalent to that obtained for RSF was obtained between the four- and sixfold dilution, indicating that the ratio in the amount of I factor RNA between SF and RSF ovaries is around fivefold.

View larger version (49K): In this window In a new window Download PPT slide   Figure 4. —RT-PCR analysis of RNAs from Cha, Cha-(RC +), and the corresponding SF and RSF flies. (a) Amplification of I factor cDNA fragments from ovary (O) and carcass (C) reverse-transcribed RNA samples (26-cycle amplification). (b) Comparative RT-PCR of RSF RNAs and serially diluted SF RNAs (26-cycle amplification). SF RNAs were diluted two-, four-, and sixfold in Cha RNAs. 0: no dilution. (c) Amplification of I factor cDNA fragments from ovaries of young (y; 4-day-old) and old (o; 60-day-old) SF and RSF flies.

The sterility level of SF females decreases as they get older (PICARD 1971 ), and the I factor RNA was detected only in the RNAs from young dysgenic flies (CHABOISSIER et al. 1990 ). We have compared the RT-PCR signals obtained with the RNA samples of ovaries from young (4-day-old) and old (60-day-old) SF females and those of the corresponding RSF females. The nonhatching percentage of the eggs laid by old SF females was about 50%. A decrease in signal with aging is observed only for SF flies; no difference is detected between the samples of old SF and RSF flies, whether young or old (Figure 4C).

Ratio of amount of I factor RNA present in SF and RSF ovaries is correlated with the kinetics of SF sterility decline:
The differences observed between the est28/Canton-S and the Cha/Cha-(RC ⁺) crosses underline the significance of the genetic background. In fact, neither the reactive nor the inducer strains are the same in the two crosses. To decipher which parameter is the determinant for the amount of I factor RNA, we made additional crosses involving the same strains but in different combination. Ovary RNA from the SF and RSF progeny of the four alternative crosses was compared (Figure 5A). Different observations can be made. First, the average amount of I factor RNA detected in the ovaries of progeny flies depends on I strain used in the cross, with more RNA detected when Cha-(RC ⁺) is used as the inducer strain. Second, it is only when the parental inducer strain is Cha-(RC ⁺) that a clear increase in the amount of I factor RNA in the SF samples can be seen relative to RSF. The magnitude of this increase depends on the reactive strain, around fivefold with Cha or two- to threefold with est28, as estimated by reconstruction experiments (see above).

View larger version (31K): In this window In a new window Download PPT slide   Figure 5. —Correlation between I factor RNA amount SF:RSF ratio and kinetics of SF sterility decrease. (a) Amplification of I factor cDNA fragments from ovary reverse-transcribed RNA samples of SF and RSF flies produced by crossing est28, Cha and Canton-S, Cha-(RC +) (in all combinations, as indicated). 28 and 26: 28 and 26-cycle amplification. (b) Evolution of sterility of SF flies. Abscissa: age of SF females in days. Ordinate: SF sterility (nonhatching %). Symbols (empty and dark circles and squares) which identify SF from the different crosses correspond to those indicated below the amplification signals in (a).

Despite their similar sterility (98–100%) when they are young, the evolution of the sterility with aging differs significantly between the different SF flies (Figure 5B). While decreasing quickly when the est28/Canton-S or the Cha/Canton-S SF females get older, the sterility of Cha/Cha-(RC ⁺) SF females begins to decline only after 15 days. The kinetics of the decrease in sterility is intermediate in the case of the est28/Cha-(RC ⁺) SF flies. Thus, the amplitude of SF/RSF ratio in the amount of I factor RNA is correlated to the kinetics of SF sterility decline.

Influence of the reactivity level:
The difference between Cha/Cha-(RC ⁺) and est28/Cha-(RC ⁺) SF flies can be attributed to the reactive strain. To test if reactivity level is the primary cause, comparisons were made in a more isogenic context. Two sublines of the same reactive Paris stock called P1m and PF9b, weakly and strongly reactive, respectively, were used for dysgenic crosses. This provides females with very different reactivity levels in a very close genetic background. Inducer males were from the Cha-(RC ⁺) stock. The nonhatching percentages of the eggs laid by the P1m/Cha-(RC ⁺) and the PF9b/Cha-(RC ⁺) SF females were 3 and 100%, respectively. The evolution of sterility of PF9b/Cha-(RC ⁺) SF females with aging is similar to that of Cha/Cha-(RC ⁺) SF females, which is indicative of a very strong I-R interaction (not shown). RNAs extracted from the SF flies, and the corresponding RSF flies, were compared by RT-PCR (Figure 6A). Equivalent amplification signals were obtained with all the ovary samples, except for the PF9b/Cha-(RC ⁺) SF sample, whose signal was much stronger. This difference is ovary-specific, i.e., not seen in the carcass samples, and is associated with a great difference in reactivity levels of the P1m and PF9b mothers. The amount of I factor RNA detected in ovaries of 60-day-old PF9b/Cha-(RC ⁺) SF females is similar to that of young or old P1m/Cha-(RC ⁺) SF flies (Figure 6B). Thus, as in the case of the Cha/Cha-(RC ⁺) cross, the amount of I factor RNA follows the decline of sterility observed in PF9b/Cha-(RC ⁺) SF flies.

View larger version (39K): In this window In a new window Download PPT slide   Figure 6. —Effect of the reactivity level. (a) Amplification of I factor cDNA fragments from ovary (O) and carcass (C) reverse-transcribed RNA samples of SF and RSF flies (28-cycle amplification). Mothers belong to highly (PF9b) and low (P1m) reactive Paris sublines. (b) Comparison of ovary amount of I factor RNA in young (y; 4-day-old) and old (o; 60-day-old) SF flies. (c) Effect of the generation breeding pattern. Amplification of I factor cDNA fragments from ovary (O) and carcass (C) reverse-transcribed RNA samples of SF flies. Reactive mothers belong to Paris sublines (PF9b orP1m) bred with short (GC) or long (GL) generations (see text).

When a strong reactive stock is bred with short generations, i.e., each generation being recovered from young mothers, the reactivity level remains high. In contrast, if the same stock is bred with long generations (old mothers), reactivity progressively decreases and reaches a low value, which is stable as long as the long generation pattern is maintained (BUCHETON 1979 ). Isogenic flies with different reactivity levels can be compared on the basis of this cumulative effect of maternal aging. Dysgenic crosses were made involving the same P1m and PF9b stocks maintained independently during 34 long generations, P1m GL34 and PF9b GL34. In this case, each new generation was issued from 30–45-day-old females. The nonhatching percentages of the eggs laid by the 3-day-old SF females, which measure the reactivity level of the mothers, were 100% (PF9b), 10% (PF9b GL34), 7% (P1m), and 2% (P1m GL34), respectively. RNAs extracted from the ovaries of the four categories of SF flies were compared by RT-PCR (Figure 6C). The amount of I factor cDNA amplified in SF ovaries when mothers are PF9b GL34 is clearly less than when mothers are PF9b, and similar to the amount detected when mothers are P1m whatever the breeding conditions. No change is detected in the carcass samples.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Different levels of control in the regulation of I factor transposition:
Using specific RT-PCR conditions, we have compared different RNAs prepared from biological material differing in the frequency of I factor transposition. I factor RNA was readily detectable in every biological sample analyzed. These include those, like carcasses and males, where no RNA was detected before (CHABOISSIER et al. 1990 ). Sensitivity of the method has allowed quantitative comparisons, which have brought new insights: (1) a clear effect of the genetic background could be demonstrated, in particular on the SF:RSF ratio in the amount of I factor RNA; (2) the I factor RNA amount does not appear as the limiting factor for transposition in every case, and the existence in ovary of a negative control downstream of I factor RNA must be assumed; (3) in SF ovaries, the positive effect of the maternal component associated with the reactivity level on the amount of I factor RNA is demonstrated for the first time. Thus, at least in ovary, multiple control levels of I factor transposition are evidenced by our studies.

Comparison of our data with published results:
The reasons for the discrepancies between the results of CHABOISSIER et al. 1990 and ours appear to be both methodological and biological. The strains used are partly concerned, as shown by differences observed between the Canton-S x est28 and the Cha/Cha-(RC ⁺) crosses. Considering the Cha/Cha-(RC ⁺) cross, we agree that I factor RNA abundance is maximum in SF ovaries, but I factor RNA is also readily detected in ovaries of inducer flies and somatic tissues (female carcasses and males). Because there is little full-length I factor RNA, the sensitivity threshold of the method is critical. The use of RT-PCR has allowed us to detect easily the I factor RNA in all the samples analyzed, making quantitative comparisons possible.

Other studies used transgenic flies transformed with constructs containing a reporter gene fused to various 5' parts of the I factor sequence (LACHAUME et al. 1992 ; LACHAUME and PINON 1993 ; MCLEAN et al. 1993 ; UDOMKIT et al. 1996 ). Expression of the transgenes was found to be limited to germline cells of dysgenic and reactive females (LACHAUME et al. 1992 ) or at least greatly enhanced in the ovaries of reactive females (MCLEAN et al. 1993 ). The data were interpreted in terms of transcriptional regulation. However, because I factor is transcribed from an internal promoter (MCLEAN et al. 1993 ), use of transgenes giving chimeric I factor-reporter gene RNAs was necessary, and I factor sequences present in these transcripts could affect other steps in expression beyond transcription. This limitation particularly applies to one construct (LACHAUME et al. 1992 ; LACHAUME and PINON 1993 ) where, in addition to the I factor 5'UTR, different parts of the ORF1 sequence are fused with the lacZ gene. In this case, histological detection of ß-galactosidase activity could reflect only partly the behavior of the transcript, and alternative interpretations exist for the data. In other transgenic experiments (MCLEAN et al. 1993 ), the constructs contain only parts of the 5' UTR of the I factor joined to the chloramphenicol acetyltransferase (CAT) gene, and recent data (UDOMKIT et al. 1996 ) indicate that the reported effects on CAT expression may be related to transcriptional regulation. In agreement with our results, a significant CAT activity was detected in all the samples examined.

Regulation in ovary: a complex blend of mechanisms:
In the ovary tissue, the amount of I factor RNA correlates, to a certain extent, with the transposition frequency. Because no variation is detected in the corresponding carcass and male samples, this mode of regulation appears very likely to be germline specific. Depending on the samples chosen for comparison, several independent regulation levels are evidenced.

First, in all crosses examined, the signal detected in SF is always superior to that seen in RSF, albeit with a variable magnitude. This indicates a negative regulation of the amount of I factor RNA amount in RSF that must be related to additional information present in oocytes of inducer origin. Such a regulation is very likely achieved through the action of an I factor encoded repressor. The ORF1 product could have such an activity as previously proposed (BUSSEAU et al. 1994 ). This is supported by a recent work describing reactive lines transformed with an ORF1 transgene and having a progressive reduction in their capacity to mobilize I factor in a dysgenic cross ( JENSEN et al. 1995 ). If the ORF1 product is really a repressor, the presence of I factor RNA not associated with transposition is not surprising. The RNA detected in nonpermissive contexts for transposition may be required for the constitutive synthesis of ORF1 repressor.

Second, the difference in the amount of RNA detected in samples according to the inducer origin, i.e., Canton-S or Cha-(RC ⁺), is probably related to a difference in I factor copy number and/or transcriptional activity of individual copies present in the two inducer strains. Accordingly, the amount of I factor RNA is greater in RSF issued from Cha-(RC ⁺), where transposition frequency is low, than in SF issued from Canton-S, where transposition frequency is high (Figure 5A). This demonstrates that the absolute amount of I factor RNA does not reflect the transposition frequency, and indicates the existence of a negative regulation downstream of I factor RNA.

Third, when one concentrates on the SF and RSF ovary signals, a correlation appears between the magnitude of the SF:RSF ratio in the amount of I factor RNA and the corresponding kinetics of SF sterility decline (Figure 5A and Figure B). The higher this ratio, the slower the decrease in sterility. Therefore, this ratio could be viewed as a measure of the strength of the I-R interaction.

Fourth, a positive influence of the reactivity level on the amount of I factor RNA is demonstrable (Figure 6A). Moreover, the cumulative effect of aging, which affects similarly the reactivity level and the amount of I factor RNA (Figure 6C), indicates that it is the maternally inherited component of reactivity, which is involved in this control.

Finally, considering the effect of aging, the decrease in the amount of I factor RNA follows the decrease of transposition frequency (Figure 4C and Figure 6C). Strikingly, the signals obtained with old SF flies issued from highly reactive mothers are similar to those of young RSF flies (Figure 4C) or young SF flies issued from low reactive mothers (Figure 6C). The existence of a basal level of expression, perhaps not influenced by any negative regulation, can be invoked.

Is there any regulation in nonovarian tissues?
The detection of I factor RNA in carcasses does not support the notion of a transcriptional restriction in somatic cells (CHABOISSIER et al. 1990 ), but is consistent with the detection of significant CAT activity in carcass extracts (MCLEAN et al. 1993 ). No significant difference between SF and RSF samples is demonstrable whatever the genetic background (Figure 3); nor is there an effect of the long generation breeding pattern (Figure 6C). This shows that I factor RNA in carcasses and ovaries is differently regulated, a fact corroborated by transgenic studies (MCLEAN et al. 1993 ). The fact that I factor RNA is not ovary-specific argues against transcriptional regulation as the main mechanism to explain the tissue-specificity of I transposition. One simplest view is the absence of any regulation in carcasses. In this case, the I factor RNA detected is only a useless, e.g., nontranslated, product. Alternatively, absence of transposition in carcasses could result from the constitutive action of a host or I factor encoded repressor activity.

Differential effect of reactivity and I factor encoded repressor:
Recently, it was shown that the reactivity levels might be one manifestation of an inducible repair- recombination system, whose biological role might be comparable to that of the SOS response in bacteria (BREGLIANO et al. 1995 ; LAURENCON and BREGLIANO 1995 ). With respect to this view, we want to propose a refined model of I factor transposition regulation, which takes into account the new results presented in this study (Table 1). In this model, I factor transposition in ovaries is under both positive and negative control. In SF females, the positive control is exerted by the reactivity level and determines the amount of I factor RNA. In this case, the transposition frequency depends directly on this amount of RNA. The negative control refers to an I factor encoded regulation, which prevents transposition in RSF and inducer flies. This is exerted at the level of, and downstream of, the amount of I factor RNA. As was already mentioned, the reactivity level is closely related to repair-recombination activities (BREGLIANO et al. 1995 ; LAURENCON and BREGLIANO 1995 ; LAURENCON et al. 1997 ). Therefore, it depends on a host function and must also exist in inducer flies, as suggested by recent data (LAURENCON et al. 1997 ). That is why we assume that in inducer or RSF oocytes, there is a certain reactivity level, but its effect must be obscured by the presence of sufficient amount of I factor encoded repressor.

As with previous models (BUSSEAU et al. 1994 ), the existence of a low level of transposition in RSF context is not easy to explain, especially the difference between the RSF and inducer female germlines. In fact, RSF is theoretically equivalent to inducer, with respect to the maternally inherited repressor molecule with only half the number of I factor copies. The relationship between the amount of repressor and the copy number is probably not a linear one, but in the absence of more available data no satisfactory explanation is evident.

	ACKNOWLEDGMENTS

We thank FRANCOISE GAY for excellent technical assistance, ANNE LAURENCON for her contribution in the settling of the different Paris sublines, MARIA BALAKIREVA for providing the 3' I factor primer sequences and OUNISSA AÏT-AHMED for the gift of the yem-alpha primers. OUNISSA AÏT-AHMED, NATHALIE DOSTATNI and SACHA KALLENBACH are greatly acknowledged for their careful reading of the manuscript. This work was supported by the Centre National de la Recherche Scientifique, the Université de la Méditerranée and the Ministère de la Recherche, de l'Enseignement et de la Technologie.

Manuscript received August 18, 1997; Accepted for publication December 22, 1997.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

AÏT-AHMED, O., M. THOMAS-CAVALLIN, C. JOBLET, and M. CAPRI, 1990  Expression in the central nervous system of a subset of the yema maternally acting genes during Drosophila embryogenesis. Post-embryonic expression extends to imaginal discs and spermatocytes. Cell Differ. Dev. 31:53-65[Medline].

AÏT-AHMED, O., B. BELLON, M. CAPRI, C. JOBLET, and M. THOMAS-DELAAGE, 1992  The yemanuclein-: a new Drosophila DNA binding protein specific for the oocyte nucleus. Mech. Dev. 37:69-80[Medline].

BREGLIANO, J.-C., A. LAURENÇON, and F. DEGROOTE, 1995  Evidence for an inducible repair-recombination system in the female germ line of Drosophila melanogaster: I. Induction by inhibitors of nucleotide synthesis and by gamma rays. Genetics 141:571-578[Abstract].

BUCHETON, A., 1979  Non-Mendelian female sterility in Drosophila melanogaster. Influence of ageing and thermic treatments. III. Cumulative effects induced by these factors. Genetics 93:131-142[Abstract/Free Full Text].

BUCHETON, A. and J.-C. BREGLIANO, 1982  The I-R system of hybrid dysgenesis in Drosophila melanogaster: heredity of the reactive condition. Biol. Cell 46:123-132.

BUCHETON, A. and G. PICARD, 1978  Non-Mendelian female sterility in Drosophila melanogaster: hereditary transmission of reactivity level. Heredity 40:207-223.

BUCHETON, A., G. LAVIGE, G. PICARD, and P. L'HERITIER, 1976  Non-mendelian female sterility in Drosophila melanogaster: quantitative variation in the efficiency of inducer and reactive strains. Heredity 36:205-314[Medline].

BUCHETON, A., R. PARO, H. M. SANG, A. PELISSON, and D. J. FINNEGAN, 1984  The molecular basis of I-R hybrid dysgenesis in Drosophila melanogaster: identification, cloning and properties of the I factor. Cell 38:153-163[Medline].

BUSSEAU, I., A. PELISSON, and A. BUCHETON, 1989  Characterization of 5' truncated copies of the I factor in Drosophila melanogaster.. Nucleic Acids Res. 17:6939-6945[Abstract/Free Full Text].

BUSSEAU, I., M. C. CHABOISSIER, A. PELISSON, and A. BUCHETON, 1994  I factors in Drosophila melanogaster : transposition under control. Genetica 93:101-116[Medline].

CHABOISSIER, M. C., I. BUSSEAU, J. PROSSER, D. J. FINNEGAN, and A. BUCHETON, 1990  Identification of a potential RNA intermediate for transposition of the LINE-like element I factor in Drosophila melanogaster.. EMBO J. 9:3557-3563[Medline].

CROZATIER, M., C. VAURY, I. BUSSEAU, A. PELISSON, and A. BUCHETON, 1988  Structure and genomic organization of I elements involved in I-R hybrid dysgenesis in Drosophila melanogaster.. Nucleic Acids Res. 16:9199-9213[Abstract/Free Full Text].

FAWCETT, D. H., C. K. LISTE, E. KELLET, and D. J. FINNEGAN, 1986  Transposable elements controlling I-R hybrid dysgenesis in Drosophila melanogaster are similar to mammalian LINEs. Cell 47:1007-1015[Medline].

FOUTS, D., R. GANGULY, A. G. GUTIERREZ, J. C. LUCCHESI, and J. E. MANNING, 1988  Nucleotide sequence of the Drosophila glucose-6-phosphate dehydrogenase gene and comparison with the homologous human gene. Gene 63:261-275[Medline].

JENSEN, S. and T. HEIDMANN, 1991  An indicator gene for detection of germline retrotransposition in transgenic Drosophila demonstrates RNA-mediated transposition of the LINE I elements. EMBO J. 10:1927-1937[Medline].

JENSEN, S., L. CAVAREC, M. P. GASSAMA, and T. HEIDMANN, 1995  Defective I elements introduced into Drosophila as transgenes can regulate reactivity and prevent I-R hybrid dysgenesis. Mol. Gen. Genet. 248:381-390[Medline].

LACHAUME, P. and H. PINON, 1993  Germ line expression of the I factor, a functional LINE from the fruit fly Drosophila melanogaster, is positively regulated by reactivity, a peculiar cellular state. Mol. Gen. Genet. 240:277-285[Medline].

LACHAUME, P., K. BOUHIDEL, M. MESURE, and H. PINON, 1992  Spatial and temporal expression of the I factor during oogenesis in Drosophila melanogaster.. Development 115:729-735[Abstract].

LAURENÇON, A. and J.-C. BREGLIANO, 1995  Evidence for an inducible repair-recombination system in the female germ line of Drosophila melanogaster: II. Differential sensitivity to gamma rays. Genetics 141:579-585[Abstract].

LAURENÇON, A., F. GAY, J. DUCAU, and J.-C. BREGLIANO, 1997  Evidence for an inducible repair-recombination system in the female germ line of Drosophila melanogaster : III. Correlation between reactivity levels, crossovers frequency and repair efficiency. Genetics 146:1333-1344[Abstract].

LAVIGE, J. M., 1986  I-R system of hybrid dysgenesis in Drosophila melanogaster: further data on the arrest of development of the embryos from SF females. Biol. Cell 56:207-216.

MCLEAN, C., A. BUCHETON, and D. J. FINNEGAN, 1993  The 5' untranslated region of the I factor, a long interspersed nuclear element-like retrotransposon of Drosophila melanogaster, contains an internal promoter and sequences that regulate expression. Mol. Cell. Biol. 13:1042-1050[Abstract/Free Full Text].

PELISSON, A. and J.-C. BREGLIANO, 1987  Evidence for rapid limitation of the I element copy number in a genome submitted to several generations of I-R hybrid dysgenesis in Drosophila melanogaster.. Mol. Gen. Genet. 207:306-313.

PELISSON, A., D. J. FINNEGAN, and A. BUCHETON, 1991  Evidence for the retrotransposition of the I factor, a LINE element of Drosophila melanogaster.. Proc. Natl. Acad. Sci. USA 88:4907-4910[Abstract/Free Full Text].

PICARD, G., 1971  Un cas de stérilité femelle chez Drosophila melanogaster lié à un agent transmis maternellement. C.R. Acad. Sci. Paris 272:2482-2487.

PICARD, G., 1976  Non-Mendelian female sterility in Drosophila melanogaster: heredity and transmission of I factor. Genetics 83:107-123[Abstract/Free Full Text].

PICARD, G., 1978  Non-Mendelian female sterility in Drosophila melanogaster, further data on chromosomal contamination. Mol. Gen. Genet. 164:235-247.

PICARD, G. and P. L'HERITIER, 1971  A maternally inherited factor inducing sterility in Drosophila melanogaster.. Dros. Inf. Serv. 46:54.

PICARD, G., J.-C. BREGLIANO, A. BUCHETON, J. M. LAVIGE, and A. PELISSON et al., 1978  Non-Mendelian female sterility and hybrid dysgenesis in Drosophila melanogaster.. Genet. Res. 32:275-287[Medline].

PROUST, J. and C. PRUDHOMMEAU, 1982  Hybrid dysgenesis in Drosophila melanogaster. I. Further evidence for and characterization of the mutator effect of the inducer-reactive interaction. Mutat. Res. 95:225-235[Medline].

PROUST, J., C. PRUDHOMMEAU, V. LADEVEZE, M. GOTTELAND, and M. C. FONTYNE-BLANCHARD, 1992  I-R hybrid dysgenesis in Drosophila melanogaster: use of in situ hybridization to show the association of I factor DNA with induced sex-linked recessive lethals. Mutat. Res. 268:265-285[Medline].

UDOMKIT, A., S. FORBES, C. MCLEAN, I. ARKHIPOVA, and D. J. FINNEGAN, 1996  Control of expression of the I factor, a LINE-like transposable element in Drosophila melanogaster.. EMBO J. 15:3174-3181[Medline].

VAURY, C., P. ABAD, A. PELISSON, A. LENOIR, and A. BUCHETON, 1990  Molecular characteristics of the heterochromatic I elements from a reactive strain of Drosophila melanogaster.. J. Mol. Evol. 31:424-431[Medline].

This article has been cited by other articles:

				J. F. Morales, E. T. Snow, and J. P. Murnane Environmental factors affecting transcription of the human L1 retrotransposon. II. Stressors Mutagenesis, March 1, 2003; 18(2): 151 - 158. [Abstract] [Full Text] [PDF]

				S. Chambeyron, A. Bucheton, and I. Busseau Tandem UAA Repeats at the 3'-End of the Transcript Are Essential for the Precise Initiation of Reverse Transcription of the I Factor in Drosophila melanogaster J. Biol. Chem., May 10, 2002; 277(20): 17877 - 17882. [Abstract] [Full Text] [PDF]

				J. F. Morales, E. T. Snow, and J. P. Murnane Environmental factors affecting transcription of the human L1 retrotransposon. I. Steroid hormone-like agents Mutagenesis, May 1, 2002; 17(3): 193 - 200. [Abstract] [Full Text] [PDF]

				E. Gauthier, C. Tatout, and H. Pinon Artificial and Epigenetic Regulation of the I Factor, a Nonviral Retrotransposon of Drosophila melanogaster Genetics, December 1, 2000; 156(4): 1867 - 1878. [Abstract] [Full Text]