New Insights on Homology-Dependent Silencing of I Factor Activity by Transgenes Containing ORF1 in Drosophila melanogaster

Corresponding author: Isabelle Busseau, IGH-CNRS, 141 rue de la Cardonille, 34396 Montpellier Cedex 05, France., busseau{at}igh.cnrs.fr (E-mail)

Communicating editor: J. A. BIRCHLER

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

I factors in Drosophila melanogaster are non-LTR retrotransposons that transpose at very high frequencies in the germ line of females resulting from crosses between reactive females (devoid of active I factors) and inducer males (containing active I factors). Constructs containing I factor ORF1 under the control of the hsp70 promoter repress I factor activity. This repressor effect is maternally transmitted and increases with the transgene copy number. It is irrespective of either frame integrity or transcriptional orientation of ORF1, suggesting the involvement of a homology-dependent trans-silencing mechanism. A promoterless transgene displays no repression. The effect of constructs in which ORF1 is controlled by the hsp70 promoter does not depend upon heat-shock treatments. No effect of ORF1 is detected when it is controlled by the I factor promoter. We discuss the relevance of the described regulation to the repression of I factors in I strains.

I factors in Drosophila melanogaster are non-long terminal repeat (LTR) retrotransposons of particular interest because high frequencies of transposition can be induced experimentally, resulting in the phenomenon of IR hybrid dysgenesis (BUCHETON 1990 ; BUSSEAU et al. 1994 ). All D. melanogaster strains fall into one of two categories: inducer (I) strains, which contain active I factors, and reactive (R) strains, which do not. I factors contain two open reading frames (ORFs): ORF1, which encodes an unknown function, and ORF2, which putatively encodes a polypeptide-containing endonuclease, reverse transcriptase, and ribonuclease H domains (FAWCETT et al. 1986 ; ABAD et al. 1989 ). I factors are stable in inducer strains but transpose very efficiently in the germ line of hybrid females resulting from crosses between the two classes of strains. The highest levels of retrotransposition are observed in females, denoted SF (for Stérilité Femelle), that are produced by crosses between females from a reactive strain and males from an inducer strain. Transposition also occurs in females, denoted RSF (for Réciproque Stérilité Femelle), that are produced by reciprocal crosses between females from an inducer strain and males from a reactive strain, but at frequencies that are lower than in SF females. Transposition is restricted to the female germ line. Accumulation of data reported by several authors (CHABOISSIER et al. 1990 ; LACHAUME et al. 1992 ; LACHAUME and PINON 1993 ; MCLEAN et al. 1993 ) indicate that I factor activity depends primarily on specific transcription initiated from an internal promoter localized entirely within the I factor 186-bp 5' untranslated region (UTR). This promoter is repressed in inducer strains but activated when introduced in the germ line of reactive strains, and it drives the synthesis of a full-length transcript that is believed to act as both the retrotransposition intermediate and the messenger for translation of the products of ORF1 and ORF2 (CHABOISSIER et al. 1990 ; BOUHIDEL et al. 1994 ). The abundance of transcripts at the time of transposition appears correlated with the transposition frequency (CHABOISSIER et al. 1990 ; MCLEAN et al. 1993 ).

In addition to high levels of transposition occurring in their germ line, SF females display a characteristic syndrome of sterility: a proportion of the eggs they lay fail to hatch and embryos die during early development. RSF females are normally fertile. Interestingly, the intensity of SF sterility, i.e., the proportion of eggs that do not hatch, appears correlated to some extent with the frequency of I factor retrotransposition, although the relationship between transposition and sterility is unclear (PICARD 1978 ). Therefore, the percentage of eggs that do not hatch is a rough estimate of the intensity of I factor activity in the germ line of SF females, within a range of observation between boundaries defined by normal fertility and total sterility corresponding to thresholds of detection and saturation, respectively.

I factors clearly self-regulate their activity since retrotransposition rarely occurs in inducer strains. When a single I factor is introduced into the genome of a reactive line, it transposes during a few generations until the number of copies reaches a certain point, estimated to be 10–15, above which the strain becomes inducer and transposition stops (PICARD 1978 ; PRITCHARD et al. 1988 ). This self-regulation has a maternal component, as is evidenced by the fact that the two reciprocal crosses are not equivalent: retrotransposition is less efficient in RSF females, having received I factors from their mother, than in SF females, having received I factors from their father.

Recent results have shed some light on the question of how I factors self-regulate. CHABOISSIER et al. 1998 showed that I factor activity decreases in the presence of several copies of the 5' UTR in the genome and that this effect correlates with a decrease of transcriptional activity of the I factor promoter. JENSEN et al. 1999A , JENSEN et al. 1999B showed that some other parts of the I factor (ORF1 and a small part of ORF2) have a repressor effect on I factor activity when they are under the control of the hsp70 promoter. This repressor effect is independent from the translation of a protein product, it increases with the copy number and over generations, and it is maternally transmitted. These observations suggest that self-regulation of the I factor is mediated by homology-dependent trans-silencing.

In this article we report independent results that confirm the results of JENSEN et al. 1999A , JENSEN et al. 1999B and bring some new insights on homology-dependent silencing of I factor activity by ORF1-containing transgenes.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Fly stocks:
All strains used in the experiments are M in the PM system of hybrid dysgenesis (ENGELS 1989 ). The genetic nomenclature follows (LINDSLEY and ZIMM 1992 ). w¹¹¹⁸ is an inducer strain. The strong reactive wK strain was obtained from LUNING 1981 . The strong reactive strain JA carries y and w mutations.

Plasmid constructions and transgenic lines:
Basic molecular biology techniques were adapted from SAMBROOK et al. 1989 . All cloned sequences from the I factor derive from pI407 (BUCHETON et al. 1984 ), which contains an active I factor (PELISSON 1981 ; PRITCHARD et al. 1988 ). Site positions are given according to the sequence published by FAWCETT et al. 1986 . Construction of hsORF2 was described previously and is referred to as phsORF2-HN in BUSSEAU et al. 1998 . hsORF1 and hsORF1as (Fig 1) were obtained by ligation of the HpaII-Klenow-treated HpaI fragment that contains all ORF1 (positions 161–1492) from pI407 into the HpaI site of the pCaSpeR-hs vector (THUMMEL and PIRROTTA 1991 ) and of the pCaSpeR-hs vector deleted of the BamHI fragment containing the hsp70 3' UTR, respectively. hsORF1fs is identical to hsORF1 except that a frameshift was introduced by insertion, using PCR with relevant oligonucleotides, of a T at position 191, just after the first ATG of ORF1, leading to the substitution of the sequence ATGACAGA ... by ATGATCAGA ... pI, ORF1, and pIORF1 were constructed in two steps.

• pI: the 5' UTR of the I factor was extracted from pI186 (MCLEAN et al. 1993 ) as a HindIII-BamHI fragment and ligated to the same sites in the polylinker of pBluescript-KS^-, producing the pBT186 plasmid; then the 5' UTR was extracted from pBTI186 as a KpnI-SpeI fragment that was ligated to the KpnI and SpeI sites of pCaSpeR-4 vector.

  
  

  View larger version (16K): In this window In a new window Download PPT slide   Figure 1. Constructs and transgenic lines. Drawings represent the I factor (top) and the different transgenes used in this study. ORF1 and ORF2 are indicated as open boxes while untranslated regions from the I factor (5'I and 3'I, 5' and 3' untranslated regions, respectively) are indicated as solid boxes. Light strippled boxes indicate the promoter (phs) and the terminator region (ths) of the hsp70 gene, while heavy strippled boxes indicate the 3' end of the P element (3'P). Arrows indicate the position of transcription initiation from the I factor or the hsp70 promoter. Parts of the transgenes that are irrelevant to the study (white sequences and the 5' end of the P element) are omitted. On the left of the drawings are the names of the different transgenes. On the right of the drawings are the names of the transgenic lines derived from the wK and JA strains, and the last column summarizes the repressor effects of the transgenes on I factor activity.

• ORF1: the I factor ORF1 sequences were extracted from pI407 as a MspI-EcoRV fragment and ligated to the same sites in the polylinker of pBluescript-KS^-, producing the pBTORF1 plasmid; then the ORF1 sequences were extracted from pBTORF1 as an EcoRI-HpaI fragment that was ligated to the EcoRI and StuI sites of the pCaSpeR-4 vector (THUMMEL and PIRROTTA 1991 ).

• pIORF1: the 5' UTR + ORF1 sequences of I were extracted from pI407 as a ClaI-HindIII fragment and ligated into the same sites in the polylinker of pBluescript-KS^-, producing the pBTIORF1 plasmid; then the 5' UTR + ORF1 sequences were extracted from pBTIORF1 as an EcoRI-HpaI fragment that was ligated to the EcoRI and StuI sites of the pCaSpeR-4 vector.

P-element-based transformations were essentially as described by SPRADLING and RUBIN 1982 , except that pUC hsp2-3 (Flybase ID no. FBmc0002087) was coinjected as the source of transposase. The recipient strains were wK or JA. Several independent homozygous transgenic lines were established for each construct by selecting dark orange-eyed flies, chromosome localizations of the transgenes were determined using balancer stocks, and integrity of the transgenes was checked by Southern blot analyses.

Crosses and measurements of female fertility:
All crosses were performed on standard fly medium (GANS et al. 1975 ) at 23° except where otherwise stated. Typically samples of 10–13 virgin females were mated to 8–10 males. They were transferred on fresh medium when necessary, until they were 8–10 days old, and then they were discarded. To determine the intensity of female sterility, 8–13 young SF females (<3 days old) were mated with the same number of males (brothers), transferred onto fresh medium the next day, and discarded the day after. Eggs were put to develop 24 hr at 25° and the percentage of hatched eggs was determined. Only measures of samples of >100 embryos were taken into account. Heat-shock treatments were applied during two successive generations by placing all developmental stages (from embryos to egg-laying adults) at 37° 1 hr once a day.

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

The repressor effect of ORF1 sequences is independent from the amount of ORF1 transcript:
Constructs in which ORF1 and ORF2 are controlled by the D. melanogaster hsp70 heat-inducible promoter (Fig 1) were introduced into the genome of the reactive strain wK. Homozygous transgenic lines HH and HN, containing the transgenes hsORF1 and hsORF2, respectively, (Fig 1) were established and maintained over 2 years (35 generations) at 20° before experiments. Each transgenic line contains one homozygous copy of the transgene, except HH9, which contains two. Each transgenic line was then divided into two sublines (including the recipient strain wK), and one of the sublines was submitted every day to a heat treatment of 1 hr at 37°. We verified that the heat treatment had no effect per se on the fertility of females from the wK strain (data not shown). We also checked by Northern blot analyses that they resulted in a large amount of ORF1 or ORF2 transcripts among the RNAs extracted from whole flies, whereas these transcripts are not detectable in the absence of heat treatments (Fig 2). In addition it has to be noted that HN transgenic lines were used in another study to show that hsORF2 transgenes can mobilize the defective I element IviP2 (BUSSEAU et al. 1998 ). Trans-complementation was more efficient upon heat induction, indicating that the hsp70 promoter is heat inducible at the time of I factor transposition.

View larger version (69K): In this window In a new window Download PPT slide   Figure 2. Analyses of ORF1-containing RNAs from HH and HN transgenic lines. (a) Northern blots of total RNAs extracted from transgenic lines HH9 (1) and HH17 (2) submitted (+) or not (-) to heat shock, hybridized with an RNA probe containing ORF1. Below is shown the result of hybridization of the same membrane with an RP49 probe as a control of RNA amounts. (b) Northern blots of total RNAs extracted from transgenic lines HN45 (1) and HN27 (2) submitted (+) or not (-) to heat shock, hybridized with an RNA probe containing ORF2. Below is shown the result of hybridizing the same membrane with an rDNA28S probe as a control of RNA amounts.

After two generations, the fertility of SF females issued from crosses between females from transgenic lines and males from the w¹¹¹⁸ inducer stock was measured (Fig 3) by determining the hatching percentage of their eggs. SF females issued from the strongly reactive wK control were, as expected, strongly sterile, with a percentage of hatching eggs close to 0% when they were <3 days old (Fig 3A) and increasing as they got older (6 and 12 days old, Fig 3B). This hatching percentage was not affected when the wK strain was submitted to a daily heat treatment for two generations before the dysgenic cross (Fig 3, a and b). This indicates that a daily heat shock at 37° for 1 hr in our experimental conditions has no observable effects on reactivity levels, contrary to long thermal treatments at 29° (BUCHETON 1978 , BUCHETON 1979A , BUCHETON 1979B ). By contrast, females of HH lines produced SF females that were less sterile, with the hatching percentages of the eggs laid by these SF females varying between 10 and >60% (Fig 3A). SF females issued from females of the sublines submitted to heat shocks every day did not display fertility levels different from those of SF females derived from females of the sublines that did not receive heat-shock treatments. This indicates that hsORF1 transgenes, containing the I factor ORF1 under the control of the hsp70 promoter, have a repressor effect on I factor activity and that this effect is independent from heat induction of the hsp70 promoter. By contrast, hsORF2 transgenes have no repressor effect on I factor activity: HN transgenic lines were also tested 100 (HN9 and HN27) and 140 (HN27) generations after transgenesis and still produced highly sterile SF females, whereas HH lines reproducibly produced SF females that were less sterile than SF females from the control cross (data not shown). All transgenic lines contain one homozygous copy of the transgene, except HH9, which contains two, and they produce the less sterile SF females.

View larger version (27K): In this window In a new window Download PPT slide   Figure 3. Repressor effect of hsORF1 and hsORF2 transgenes. (a) Hatching percentages of the eggs laid by SF females issued from crosses between females from the HH and HN transgenic lines obtained from the reactive stock wK, raised at 20°, and submitted or not every day to a heat treatment during two successive generations, 2 yr (35 generations, except HH9, for which the experiment was done three generations later) after transgenesis, and males from the w1118 inducer stock. Arrowheads point to transgenic line HH9 containing two homozygous transgenes. (b) Hatching percentages of the eggs laid by 6-day-old [wK(6)] and 12-day-old [wK(12)] SF females issued from crosses between females from the reactive stock wK submitted or not every day to a heat treatment and males from the w1118 inducer stock.

We tested the repressor effect of hsORF1 transgenes in another genetic background. Transgenic lines HOU, containing the same hsORF1 transgene as HH lines, and transgenic lines HOUf, containing the hsORF1fs transgene in which a frameshift just after the first ATG prevents expression of the ORF1 protein (Fig 1), were established by P-mediated transformation of the reactive strain JA. The establishment of each transgenic line was accompanied by the parallel establishment of a sibling nontransgenic line as a control. The transgenic and nontransgenic lines were derived, respectively, from orange- and white-eyed individuals issued from the same transformed fly. Dysgenic crosses of females of the lines with males from the w¹¹¹⁸ inducer stock were performed and the fertility of the resulting SF females was measured. Data are shown in Fig 4. HOU and HOUf transgenic lines produced SF females that were less sterile than nontransgenic females from the internal control. The fertility improvements appear weak a few generations after transgenesis (HOU at generation 6, HOUf at generation 8) but are more obvious at generations 40–42, except for line HOU57a, which shows no effect. It is noticeable that two transgenic lines, HOU33 and HOUf29.3, that produce the less sterile SF females, contain two homozygous copies of the transgenes, whereas other transgenic lines all have only one copy. These data indicate first that the repressor effect of hsORF1 transgenes is also effective in the JA background and second that hsORF1 transgenes can repress I factor activity independently from expression of an ORF1 protein. They also suggest that two copies of a transgene may have a stronger repressor effect than one copy.

View larger version (35K): In this window In a new window Download PPT slide   Figure 4. Repressor effect of various transgenes in the JA background. Hatching percentages of eggs laid by SF females issued from crosses between males from the w1118 inducer stock and females from transgenic lines (shaded bars) or females from sibling nontransgenic lines (open bars), except for HOU and HOUf lines at generation 40, for which the control was the JA stock. The numbers of generations after transgenesis are indicated in each graph. Arrowheads point to transgenic lines containing two homozygous transgenes.

The repressor effects of two hsORF1 transgenes are cumulative:
We tested whether the regulatory effect of hsORF1 is enhanced by increasing the number of copies of the transgene. Using lines HH13 and HH16 (transgenes on chromosomes II and III, respectively), we constructed various sublines as shown in Fig 5A. Reciprocal crosses between flies from lines HH13 and HH16 were made to obtain hybrid females heterozygous for both transgenes. These hybrid females were crossed to males of the wK strain. The progeny of these crosses were scored upon eye color intensity and mated to establish sublines. 1613-1, 1613-2, 1613-3, 1613-4, and 1613-5, containing both transgenes at the homozygous state, 1613-16H and 1613-13H, containing one homozygous transgene on chromosomes III and II, respectively, and 1613-wK, devoid of transgene, were obtained in the progeny of the crosses between HH16 females and HH13 males. 1316-7 and 1316-8, containing both transgenes at the homozygous state, and 1316-13H, containing one homozygous transgene on chromosome II, were obtained in the progeny of the reciprocal cross between HH13 females and HH16 males. The fertility of SF females obtained from crosses between females of these sublines and males of the w¹¹¹⁸ inducer stock was measured three generations after establishment of the sublines. As shown in Fig 5B, sublines 1613 and 1316, containing two homozygous transgenes, reproducibly produce less sterile SF females than sublines containing only one homozygous transgene, whereas the 1613-wK subline, devoid of transgene, produces strongly sterile SF females, as does the strongly reactive wK strain. Subsequent tests made 29 generations after establishment produced similar results (not shown). Therefore, the effect of hsORF1 transgenes is cumulative.

View larger version (39K): In this window In a new window Download PPT slide   Figure 5. Influence of the copy number of hsORF1 transgenes. (a) Scheme of crosses used to obtain 1613 and 1316 sublines containing two homozygous transgenes. For clarity only autosomes II and III are indicated. Flies of the various genotypes were recognized by their eye colors since the eyes of flies from the HH13 and HH16 lines are of slightly different orange tones, and the double homozygotes 1613 and 1316 have red eyes. (b) Hatching percentages of the eggs laid by SF females issued from crosses between females from the 1613 or 1316 sublines three generations after transgenesis and males from the w1118 inducer stock. Standard errors for three samples of sibling SF females are indicated as vertical bars.

The repressor effect of hsORF1 transgenes requires the association of the hsp70 promoter and ORF1 sequences:
We have tested the ability of the hsp70 promoter and of ORF1 sequences without any promoter to repress I factor activity. For this, transgenic lines H, containing the pCaSpeR-hs vector, and OU, containing the sequences of ORF1 (see Fig 1), were established by P-mediated transformation of the JA strain. They all contain one homozygous copy of the transgene. Nontransgenic sibling lines were also established as controls as described above and the regulatory effects of the transgenes were measured at generations 5, 8, 9, 12, and 13 after transgenesis in the case of OU lines and at generations 4 and 16 in the case of H lines. Typical data are shown in Fig 4: none of the transgenes exhibit a regulatory effect. This indicates that the regulatory effect of the hsORF1 transgenes requires the association of the sequences of the hsp70 promoter and of ORF1.

ORF1 under the control of the I factor promoter has no apparent regulatory effect:
The hsORF1 transgenes containing ORF1 under the control of a heterologous promoter are artificial and the regulatory effects obtained with these transgenes might be irrelevant to actual I factor regulation. We therefore decided to test whether ORF1 under the control of the I factor promoter displays a similar repressor activity. Transgenic lines pIOU (containing the pIORF1 transgene) and pI (containing the pI transgene alone, see Fig 1) were established by P-mediated transformation of the JA strain. They all contain one homozygous copy of the transgene, except pIOU17c and pI8f, which contain two. Nontransgenic sibling lines were also established as controls as described above and the regulatory effects were studied at generations 5, 8, 9, 12, and 13 after transgenesis. Typical data are shown in Fig 4. Females of pIOU and pI transgenic lines produced SF females displaying the same high level of sterility as females of the nontransgenic lines, indicating that ORF1 under the control of the I factor promoter has no repressor effect in these conditions. We verified by Northern blot analyses of total RNAs extracted from ovaries of pIOU transgenic lines that ORF1 transcripts are actually produced in these flies (data not shown).

The repressor effect of hsORF1 transgenes is independent from transcription orientation:
The transgene hsORF1as is similar to hsORF1, except that ORF1 is placed in antisense orientation under the control othe hsp70 promoter and that the hsp70 3' UTR is deleted (see Fig 1). Two lines, named HOUas49 and HOUas51, each containing two homozygous copies of this transgene, were obtained in the JA strain. Females of HOUas lines produced SF females that were more fertile than SF females produced by females of the JA strain. This repressor effect was not affected by heat-shock treatments applied every day during two successive generations according to the protocol described for HH lines (Fig 6). Therefore, transgenes that contain ORF1 under the control of the hsp70 promoter, whatever the orientation of ORF1, display similar regulatory properties on I factor activity.

View larger version (40K): In this window In a new window Download PPT slide   Figure 6. Repressor effect of hsORF1as transgenes. Hatching percentages of the eggs laid by SF females issued from crosses between females from the HOUas transgenic lines obtained from the JA reactive stock, raised at 23°, and submitted or not every day to a heat treatment during two successive generations, 40 generations after transgenesis, and males from the w1118 inducer stock.

Maternal transmission of the repressor effects of the transgenes:
Since I factor autoregulation has a maternal component, evidenced by the fact that I factor activity is lower in the germ line of RSF females (having received I factors from their mother) than in the germ line of SF females (having received I factors from their father), we tested whether the regulatory effects of the transgenes are transmitted maternally. For this, we performed the experiments described in Fig 7 with transgenic lines HOU33, HOUf29.3, and HOUas51r. Heterozygous females were produced by the two reciprocal crosses between flies from transgenic lines and flies from the JA stock (labeled and in Fig 7) and were crossed with males from the w¹¹¹⁸ inducer stock to produce SF females whose fertility was measured. The fertility of SF females issued from crosses of females from the homozygous transgenic lines (labeled in Fig 7) or females from the JA stock (labeled ø in Fig 7) with males from the w¹¹¹⁸ inducer stock was determined in parallel. In each case a clear maternal effect was observed: heterozygous females having received the transgene from their mothers produced SF females that were more fertile than heterozygous females having received the transgene from their fathers. A zygotic effect is also observed, since the progeny of heterozygous orange-eyed SF females (that received the transgene) were more fertile than that of white-eyed SF females (that did not receive the transgene).

View larger version (43K): In this window In a new window Download PPT slide   Figure 7. Maternal transmission of repressor effects. (a) Scheme of crosses used to test maternal transmission. Tg stands for transgene and represents hsORF1 (line HOU33), hsORF1fs (line HOUf29.3), or hsORF1as (line HOUas51). (b) Hatching percentages of the eggs from SF females issued from crosses between transgenic homozygous (), transgenic heterozygous ( and ), or nontransgenic (ø) females and males from the w1118 inducer stock. Standard errors for three samples of sibling SF females of the same genotype are indicated as vertical bars.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

In this article we test the ability of various transgenes containing ORF1 sequences to reduce I factor activity. Our experimental conditions were designed to detect repressor effects of transgenes on the intensity of the characteristic syndrome of sterility that is associated with high levels of I factor transposition. Not surprisingly, this methodology produces highly variable data. The first reason is inherent to any experiment that is based on transgenesis. The activity of a given transgene will obviously be modulated by surrounding regions, depending on its position in the genome. This is why we have systematically studied several transgenic lines for each construct. The second reason comes from the fact that I factor activity is modulated by the cellular state, called reactivity, that is prevalent in reactive stocks used to produce SF females. Reactivity is subject to variations due to environmental and physiological factors such as temperature and aging, and these changes are partially heritable for a few generations (BUCHETON and PICARD 1978 ; BUCHETON 1978 , BUCHETON 1979A , BUCHETON 1979B ; BUCHETON and BREGLIANO 1982 ). Even highly controlled laboratory conditions cannot ensure a strict homogeneity of fly culture parameters over many generations. For this reason, we have systematically tested the repressor effect of transgenes at several generations to ensure the reproduceability of the results.

Our data show that various transgenes containing ORF1 sequences under the control of the hsp70 promoter have a repressor effect on I factor activity. This repressor effect is observed with intact ORF1 sequences (transgenes hsORF1 in lines HH and HOU) as well as when a frameshift mutation prevents synthesis of the ORF1 protein (transgene hsORF1fs in lines HOUf) or when ORF1 is in an antisense orientation (transgene hsORF1as in lines HOUas). It is abolished when the hsp70 promoter is removed (transgene ORF1 in lines OU). These features mostly coincide with those reported by JENSEN et al. 1999A , JENSEN et al. 1999B . JENSEN et al. 1999B have suggested the involvement of a homology-dependent trans-silencing process depending upon the production of an RNA containing ORF1 sequences. Interestingly, recent reports indicate that mobility of Tc1 and of other DNA transposons in Caenorhabditis elegans can be regulated by a mechanism known as RNA interference (RNAi), which leads to the inhibition of gene function through double-stranded RNA interactions (KETTING et al. 1999 ; TABARA et al. 1999 ). Double-stranded RNA molecules have been shown to be capable of repressing homologous endogenous genes in D. melanogaster (KENNERDELL and CARTHEW 1998 ; MISQUITTA and PATERSON 1999 ; TUSCHL et al. 1999 ), suggesting that post-transcriptional mechanisms involving RNA intermediates may also occur in this species. It is tempting to speculate the involvment of RNAi in the regulation of transposable elements, including non-LTR retrotransposons in D. melanogaster.

However, two of our observations are apparently not in agreement with the hypothesis of RNA mediation. First, transgenes containing ORF1 under the control of the I factor promoter (transgene pIORF1 in lines pIOU) have no regulatory effect, although they are transcribed in the ovaries. Second, while the repressor effect of hsORF1 transgenes is enhanced by an increase of their copy number (sublines HH1316 and HH1613), we find that whatever the orientation of ORF1, it is not enhanced by heat induction of the hsp70 promoter, although this promoter is heat inducible in the female germ line at the time of I transposition (BUSSEAU et al. 1998 ). This indicates that, although regulation by hsORF1 constructs is dependent on the copy number of transgenes, it appears not to be correlated to the amounts of their transcripts. Thus, if an RNA intermediate is required for mediating repression by transgenes containing ORF1 sequences, its production is certainly not a sufficient condition. Maybe some specific characteristics of RNAs produced by hsORF1 transgenes could trigger trans-silencing. These RNAs differ at their 5' ends from those produced by pIORF1 transgenes due to distinct transcription initiation from the hsp70 and the I factor promoters (Fig 1) and might have different properties.

It is possible that a DNA-based recognition mechanism could also be invoked: indeed, our results show some similarities to cosuppression involving the Adh and white genes in D. melanogaster (PAL-BHADRA et al. 1997 , PAL-BHADRA et al. 1999 ): the presence of multiple w-Adh transgenes in the genome can repress the activity of the endogenous Adh gene in a copy-number-dependent manner (PAL-BHADRA et al. 1997 , PAL-BHADRA et al. 1999 ). This phenomenon is triggered by homology recognition at the DNA level of a nontranscribed segment of the Adh regulatory region. It is dependent on proteins of the Polycomb group, implying that changes in chromatin accessibility are involved. It is now important to determine whether the repressor effect of hsORF1 transgenes on I factor activity that we describe here involves proteins of the Polycomb group as well.

Some transgenes show no effect on I factor activity in our experimental conditions. However, these conditions would not have allowed us to detect low or slowly accumulating effects. JENSEN et al. 1999A , JENSEN et al. 1999B have shown that the repressor effect of hsORF1 transgenes increases over generations. We observe this phenomenon with HOU and HOUf transgenic lines, which produce SF females that are less sterile at generations 40–42 than at generations 6–8. We did not study OU, pIOU, pI, and H lines after generation 13; therefore, we would not have noticed a low repressor effect slowly accumulating over many generations. Besides, we used transgenic lines that contained only one or two homozygous transgenes, so we would not have detected repressor effects induced by multiple copies of a given transgene. For example, we report here that one or two copies of the 5' UTR have no detectable regulatory effect (lines pI), whereas CHABOISSIER et al. 1998 showed that transgenes containing two or three tandem repeats of the 5'UTR of the I factor repress I factor activity with efficiencies increasing with their copy number. Taking all this into consideration, we do not exclude the possibility that pIORF1 or ORF1 transgenes might have an effect on I factor activity that we would not have observed under our experimental conditions, but that could be detected after many generations or in the presence of multiple copies of the transgenes. Even in this case this would indicate that this effect is weaker than that observed with hsORF1 transgenes. The regulatory effects reported here and by JENSEN et al. 1999A , JENSEN et al. 1999B are associated with the presence of the hsp70 promoter. This suggests that the presence of the hsp70 promoter is important although it may not act simply through RNA production. In previous work, we have shown that an element defective in ORF2, IviP2, has no repressor effect on I factor activity (BUSSEAU et al. 1998 ; I. BUSSEAU, unpublished results). On the contrary, a similar defective element, IZ, displays a strong repressor effect on I factor activity (JENSEN et al. 1995 ). IZ contains a hsp70-LacZ reporter gene inserted within ORF2 (JENSEN et al. 1994 ), whereas IviP2 contains the D. melanogaster vermilion gene and is devoid of the hsp70 promoter (CHABOISSIER et al. 1995 ). It is conceivable that a particular sequence on the hsp70 promoter is used as a recognized motif for chromatin-binding proteins or is involved in chromatin remodeling.

This raises the question of the relevance of our results on I factor self-regulation: indeed, pIORF1 transgenes are more similar to I elements than hsORF1 transgenes are, since they possess the 5' UTR of the I factor instead of the hsp70 promoter. There is no evidence that one copy of the I factor in the absence of the hsp70 promoter would be sufficient to trigger repression. Maternal transmission is one characteristic of natural repression of I elements occurring in inducer strains. Our study shows that the repressor effect of hsORF1 transgenes on I factor activity is maternally transmitted, suggesting that our observations are relevant to I self-regulation. The regulation of the I factor is complex and still not understood. It might actually involve more than one mechanism. Whatever the case it is a very stimulating question that can be used to study epigenetic inheritance.

	ACKNOWLEDGMENTS

We thank Marie Balakireva and Maria del Carmen Seleme for their help in some of the experiments. This work was supported by grants from the Programme Génome of the CNRS and from the Association pour la Recherche sur le Cancer (ARC contract 5234). S.M. was supported by fellowships from the Ministère de la Recherche et de la Technologie and from the Association pour la Recherche contre le Cancer (ARC).

Manuscript received March 13, 2000; Accepted for publication July 10, 2000.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

ABAD, P., C. VAURY, A. PÉLISSON, M-C. CHABOISSIER, and I. BUSSEAU et al., 1989  A long interspersed repetitive element—the I factor of Drosophila teissieri—is able to transpose in different Drosophila species. Proc. Natl. Acad. Sci. USA 86:8887-8891[Abstract/Free Full Text].

BOUHIDEL, K., C. TERZIAN, and H. PINON, 1994  The full-length transcript of the I factor, a LINE element of Drosophila melanogaster, is a potential bicistronic RNA messenger. Nucleic Acids Res. 22:2370-2374[Abstract/Free Full Text].

BUCHETON, A., 1978  Non-Mendelian female sterility in Drosophila melanogaster: influence of ageing and thermic treatments. I. Evidence for a partly inheritable effect of these two factors. Heredity 41:357-369[Medline].

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

BUCHETON, A., 1979b  Non-mendelian female sterility in Drosophila melanogaster—influence of aging and thermic treatments: II. Action of thermic treatments on the sterility of SF females and on the reactivity of reactive females. Biol. Cell 34:43-50.

BUCHETON, A., 1990  I transposable elements and I-R hybrid dysgenesis in Drosophila. Trends Genet. 6:16-21[Medline].

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 levels. Heredity 40:207-223.

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

BUSSEAU, I., S. MALINSKY, M. BALAKIREVA, M. C. CHABOISSIER, and D. TENINGES et al., 1998  A genetically marked I element in Drosophila melanogaster can be mobilized when ORF2 is provided in trans. Genetics 148:267-275[Abstract/Free Full Text].

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].

CHABOISSIER, M. C., C. BORNECQUE, I. BUSSEAU, and A. BUCHETON, 1995  A genetically tagged, defective I element can be complemented by actively transposing I factors in the germline of I-R dysgenic females in Drosophila melanogaster.. Mol. Gen. Genet. 248:434-438[Medline].

CHABOISSIER, M. C., A. BUCHETON, and D. J. FINNEGAN, 1998  Copy number control of a transposable element, the I factor, a LINE- like element in Drosophila. Proc. Natl. Acad. Sci. USA 95:11781-11785[Abstract/Free Full Text].

ENGELS, W. R., 1989 P elements in Drosophila, pp. 437–484 in Mobile DNA, edited by D. E. BERG and M. M. HOWE. American Society for Microbiology, Washington, DC.

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

GANS, M., C. AUDIT, and M. MASSON, 1975  Isolation and characterization of sex-linked female-sterile mutants in Drosophila melanogaster.. Genetics 81:683-704[Abstract/Free Full Text].

JENSEN, S., L. CAVAREC, O. DHELLIN, and T. HEIDMANN, 1994  Retrotransposition of a marked Drosophila line-like I element in cells in culture. Nucleic Acids Res. 22:1484-1488[Abstract/Free Full Text].

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].

JENSEN, S., M. P. GASSAMA, and T. HEIDMANN, 1999a  Cosuppression of I transposon activity in Drosophila by I-containing sense and antisense transgenes. Genetics 153:1767-1774[Abstract/Free Full Text].

JENSEN, S., M. P. GASSAMA, and T. HEIDMANN, 1999b  Taming of transposable elements by homology-dependent gene silencing. Nat. Genet. 21:209-212[Medline].

KENNERDELL, J. R. and R. W. CARTHEW, 1998  Use of dsRNA-mediated genetic interference to demonstrate that frizzled and frizzled 2 act in the wingless pathway. Cell 23:1017-1026.

KETTING, R. F., T. H. HAVERKAMP, H. G. VAN LUENEN, and R. H. PLASTERK, 1999  Mut-7 of C. elegans, required for transposon silencing and RNA interference, is a homolog of Werner syndrome helicase and RNaseD. Cell 99:133-141[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].

LINDSLEY, D. L., and G. G. ZIMM, 1992 The genome of Drosophila melanogaster. Academic Press, New York.

LÜNING, K. G., 1981  Genetics of inbred Drosophila melanogaster: I. Induction of marker genes and preliminary recombination tests. Hereditas 95:181-188.

MISQUITTA, L. and B. M. PATERSON, 1999  Targeted disruption of gene function in Drosophila by RNA interference (RNA-i): a role for nautilus in embryonic somatic muscle formation. Proc. Natl. Acad. Sci. USA 96:1451-1456[Abstract/Free Full Text].

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].

PAL-BHADRA, M., U. BHADRA, and J. A. BIRCHLER, 1997  Cosuppression in Drosophila: gene silencing of Alcohol dehydrogenase by white-Adh transgenes is Polycomb dependent. Cell 90:479-490[Medline].

PAL-BHADRA, M., U. BHADRA, and J. A. BIRCHLER, 1999  Cosuppression of nonhomologous transgenes in Drosophila involves mutually related endogenous sequences. Cell 99:35-46[Medline].

PELISSON, A., 1981  The I-R system of hybrid dysgenesis in Drosophila melanogaster: are I factor insertions responsible for the mutator effect of the I-R interaction? Mol. Gen. Genet. 183:123-129[Medline].

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

PRITCHARD, M. A., J. M. DURA, A. PELISSON, A. BUCHETON, and D. J. FINNEGAN, 1988  A cloned I factor is fully functional in Drosophila melanogaster.. Mol. Gen. Genet. 214:533-540[Medline].

SAMBROOK, J., E. F. FRITSCH and T. MANIATIS, 1989 Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

SPRADLING, A. C. and G. M. RUBIN, 1982  Transposition of cloned P elements into Drosophila germ line chromosomes. Science 218:341-347[Abstract/Free Full Text].

TABARA, H., M. SARKISSIAN, W. G. KELLY, J. FLEENOR, and A. GRISHOK et al., 1999  The rde-1 gene, RNA interference, and transposon silencing in C. elegans.. Cell 99:123-132[Medline].

THUMMEL, C. S., and V. PIRROTTA, 1991 New PCaSpeR P element vectors. Dros. Inf. Newslett. 2.

TUSCHL, T. P., P. D. ZAMORE, R. LEHMANN, D. P. BARTEL, and P. A. SHARP, 1999  Targeted mRNA degradation by double stranded RNA in vitro.. Genes Dev. 13:3193-3197.

This article has been cited by other articles:

				J. Brennecke, C. D. Malone, A. A. Aravin, R. Sachidanandam, A. Stark, and G. J. Hannon An Epigenetic Role for Maternally Inherited piRNAs in Transposon Silencing Science, November 28, 2008; 322(5906): 1387 - 1392. [Abstract] [Full Text] [PDF]

				E. Sarot, G. Payen-Groschene, A. Bucheton, and A. Pelisson Evidence for a piwi-Dependent RNA Silencing of the gypsy Endogenous Retrovirus by the Drosophila melanogaster flamenco Gene Genetics, March 1, 2004; 166(3): 1313 - 1321. [Abstract] [Full Text] [PDF]

				S. Robin, S. Chambeyron, A. Bucheton, and I. Busseau Gene Silencing Triggered by Non-LTR Retrotransposons in the Female Germline of Drosophila melanogaster Genetics, June 1, 2003; 164(2): 521 - 531. [Abstract] [Full Text] [PDF]

				S. Jensen, M.-P. Gassama, X. Dramard, and T. Heidmann Regulation of I-Transposon Activity in Drosophila: Evidence for Cosuppression of Nonhomologous Transgenes and Possible Role of Ancestral I-Related Pericentromeric Elements Genetics, November 1, 2002; 162(3): 1197 - 1209. [Abstract] [Full Text] [PDF]

				S. Yang, S. Tutton, E. Pierce, and K. Yoon Specific Double-Stranded RNA Interference in Undifferentiated Mouse Embryonic Stem Cells Mol. Cell. Biol., November 15, 2001; 21(22): 7807 - 7816. [Abstract] [Full Text] [PDF]

				N. C. Lau, L. P. Lim, E. G. Weinstein, and D. P. Bartel An Abundant Class of Tiny RNAs with Probable Regulatory Roles in Caenorhabditis elegans Science, October 26, 2001; 294(5543): 858 - 862. [Abstract] [Full Text] [PDF]

				S. M. Elbashir, W. Lendeckel, and T. Tuschl RNA interference is mediated by 21- and 22-nucleotide RNAs Genes & Dev., January 15, 2001; 15(2): 188 - 200. [Abstract] [Full Text]