Insect Biochemistry and Molecular Biology 32 (2002) 247–253www.elsevier.com/locate/ibmb

Rapid communication

3×P3-EGFP marker facilitates screening for transgenic silkwormBombyx mori L. from the embryonic stage onwards

J.-L. Thomas*, M. Da Rocha, A. Besse, B. Mauchamp, G. ChavancyUnite Nationale Sericicole, INRA, 25 quai J.J. Rousseau, 69350, La Mulatiere, France

Received 15 August 2001; received in revised form 1 October 2001; accepted 5 October 2001

Abstract

Transgenesis was recently achieved inBombyx mori L., but it has proved difficult and time-consuming to screen the numerousprogeny to identify the transgenic individuals. As the 3×P3-EGFP marker has been shown to be a suitable universal marker fortransgenic insects (Nature 402 (1999) 370), we evaluated its use for embryonic-stage screening forB. mori L. germline transform-ation. Using thepiggyBac-derived vector pBac{3×P3-EGFPaf}, we were able to isolate four transgenic individuals from about120,000 embryos (560 broods). The screening was straightforward due to EGFP production in the G1 embryonic stemmata, whichwas visible through the translucent egg chorion. EGFP was produced in the stemmata and central and peripheral nervous systemsfrom the fifth day of embryonic development. It persisted at high levels in the stemmata throughout the larval stage, and was alsopresent in the compound eyes and nervous tissues of the pupae and the compound eyes of the moths. 2002 Elsevier ScienceLtd. All rights reserved.

Keywords: Bombyx mori; Transgenesis;piggyBac; Transposon; pBac{3×P3-EGFPaf}; Marker expression

1. Introduction

Germline transgenesis was recently achieved inBom-byx mori L. (Tamura et al., 2000) after many attemptsover a number of years (Ninaki et al., 1985; Tamura etal., 1990; Coulon-Bublex et al., 1993; Nagaraju et al.,1996). It was made possible by the use of a combinationof the EGFP marker gene cloned in thepiggyBac trans-poson under control of the Bm-Actin3 promoter (Caryet al., 1989; Fraser et al., 1995). The resultingpiggyBac-derived vector, pPIGA3GFP, was useful for the screen-ing of G1 individuals at the larval stage but necessitatedthe rearing of numerous silkworms. This time-consum-ing work is the main limitation of this promoter–markercombination. With the Bm-Actin3GFP marker,expression can be viewed in mosaic G0 larvae and inthe transgenic larvae of subsequent generations. Theexpression of this marker is visible in the vitellophagesof the G0 eggs, but not in G0 embryonic tissues or inthe eggs of the later generations. It is therefore not poss-

* Corresponding author.E-mail address: thomas@pop.univ-lyon1.fr (J.-L. Thomas).

0965-1748/02/$ - see front matter 2002 Elsevier Science Ltd. All rights reserved.PII: S0965-1748 (01)00150-3

ible, with this promoter, to screen transgenic individualsbefore the G1 larval stage.

With a view to reducing the larval rearing effortrequired, we investigated the universal 3×P3-EGFPmarker developed in Wimmer’s lab (Berghammer et al.,1999; Horn et al., 2000; Horn and Wimmer, 2000). Theartificial 3×P3 promoter drivesEGFP expression notonly in the ocelli and omatidia of adultTribolium cas-taneum and Drosophila melanogaster, but also in thestemmata of beetle larvae and in the Bolwig organs andnervous tissues of dipteran larvae, which start to fluor-escence before hatching, at embryonic stages (Horn etal., 2000; Hediger et al., 2001).

These data suggested that the 3×P3-EGFP marker wasvery likely to be expressed in the stemmata ofB. moriL. embryos. InB. mori L. embryos, the stemmata are incontact with the chorion, which is almost transparent.This enabled us to use the 3×P3-EGFP marker to selecttransgenic individuals as embryos, overcoming thenecessity to rear thousands of G1 larvae.EGFPexpression was detectable from the fifth day of embry-onic development until the imago stage. This made itpossible to determine the precise expression pattern ofthe 3×P3-EGFP marker inB. mori.

248 J.-L. Thomas et al. / Insect Biochemistry and Molecular Biology 32 (2002) 247–253

2. Materials and methods

2.1. B. mori L. strain

The Indian polyvoltin strain Nistari was obtained froma silkworm collection maintained at UNS/INRA(France). Silkworms were reared at 25°C and fed withmulberry leaves from spring to autumn and on an arti-ficial diet during winter. After hatching from microin-jected eggs, first instar larvae were fed an artificial dietand reared in groups under standard conditions. G0adults were mated together and G1 eggs were screenedfor EGFP expression from the fifth day of incubationonwards. The positive G1 eggs were isolated, hatchedand the larvae reared separately. Positive adult G1were backcrossed.

2.2. Egg preparation for microinjection and screening

For egg collection, male and female moths wereallowed to mate overnight by artificial light. In the morn-ing, the females were placed on a plastic sheet andallowed to lay in the dark for one hour. The eggs weredisinfected with 4% formaldehyde solution for 5 min,rinsed with distilled water and finally dried with absoluteethanol. The piece of the plastic sheet on which the eggswere laid was glued with cyanocrylate glue onto a55 mm Petri dish.

Glass needles were pulled from 20 µl precalibratedpipettes (Vitrex, ref 1264) and were sharpened to anangle of 30°, using a Narishige EG4 grinder. Eggs weremicroinjected as soon as possible, and in all cases wereinjected no later than four hours after oviposition. Unlessotherwise specified, the injection was made into the dor-sal posterior third of the egg, avoiding the ventral side,where the germ band develops. We injected 5–10 nl ofa 1:1 mixture of vector and helper plasmids (0.5 µg/µltotal DNA concentration) in deionised water into theeggs. The injection hole was then sealed with a smalldrop of cyanocrylate glue. Throughout egg incubation,a saturated atmosphere was maintained by placing twowet pieces of GF/B glass filter paper (Whatman) on thecoverslip of the Petri dish. The Petri dish was placedupside-down in a closed box kept at 25°C.

To make it easier to see the marker in the embryos,we immersed the eggs in water or 90% ethanol. Thischanged the refractive index of the surface of the chor-ion, rendering it almost transparent, making it easier tosee the fluorescence in the stemmata.

2.3. DNA constructs

The transposon-encoding plasmid, pHA3pig (6.2 kb),is described elsewhere (Tamura et al., 2000). The pig-gyBac-derived vector pBac{3×P3-EGFPaf} (7.3 kb) wasgenerously provided by E.A. Wimmer and is described

in Horn and Wimmer (2000). Both DNA constructs wereamplified using the Hybaid Maxi Flow DNA preparationkit. Their final concentration was adjusted to 1 µg/µl inwater and they were aliquoted and stored at �20°C. TheDNA vector and helper mixture, or the vector alone,were injected at a concentration of 0.5 µg/µl.

2.4. Inverse PCR (iPCR) and junction sequences

Genomic DNA was extracted from G1 moths, G1embryonic larvae or from the G2 silk gland. DNA waspurified by standard SDS lysis-phenol extraction treat-ment after incubation with proteinase K. The DNA wasfurther treated with RNAse and purified as described byHediger et al. (2001). DNA was digested with Hae IIIand circularised by ligation for 3 h at 18°C. PCR wasperformed on the circularised fragments, using primersequences, in opposite orientations, corresponding tosequences between the restriction site and the endsequence of the piggyBac vector. For the 3� junction, theforward primer (PRF) 5�-CCTCGATATACAGACCGATAAAACACATGC-3� and reverse primer (PRR) 5�-AGTCAGTCAGAAACAACTTTGGCACATATC-3�were used. For the 5� junction, the forward primer (PLF)5�-CTTGCACTTGCCACAGAGGACTATTAGAGG-3�and reverse primer (PLR) 5�-CAGTGACACTTACCGCATTGACAAGCACGC-3� were used. PCR fragmentswere separated by electrophoresis in a 0.8% agarose geland plugs of single bands were reamplified and purified(PCR purification kit, Qiagen, D). The purified frag-ments were directly sequenced (GENOME Express;Meylan, France) with the PLR primer for the left (5�)boundary and PRF for the right (3�) boundary of the vec-tor.

3. Results

3.1. Pattern of 3×P3-EGFP marker expression in B.mori G0 embryos

We checked that the 3×P3-EGFP marker was func-tional in B. mori embryos by co-injecting the piggyBacvector and the helper plasmid pHA3pig into eggs at a1:1 (w/w) ratio (1.17:1 H/V; molar ratio). We injectedDNA into the eggs at the anterior pole to favour itslocation in the cephalic area, where the stemmata differ-entiate. This may have a detrimental effect on hatching,but this was not considered important in this preliminaryevaluation of the usefulness of the marker. Once thestemmata had differentiated, we observed expression ofthe marker through the egg shell (Fig. 1). On subsequentdays, and until the larval pre-hatched stage, the numberof embryos producing EGFP steadily increased (Table2). The 3×P3-EGFP marker was expressed not only inthe stemmata, but also in other tissues, some of which

249J.-L. Thomas et al. / Insect Biochemistry and Molecular Biology 32 (2002) 247–253

Fig. 1. Expression in stemmata of a G0 embryo (arrow). The brown-ish dot corresponds to the vitellogenic clot. The second egg did notexpress the marker although the vector was injected.

were identified as nervous system tissues. Expressionwas observed in the thorax and abdomen as well as inthe head (data not shown). It was clear that the markerwas expressed in nervous system tissues in all thesebody parts. In many cases it was necessary to dissectembryos to view or to confirm EGFP expression becausethe melanin of the vitellogenic clot covering the injectionhole was opaque. Expression of the 3×P3-EGFP markerwas clearly seen in G0 hatched larvae (Fig. 2) and wouldclearly have been detected in G1 eggs with a trans-parent chorion.

3.2. Conditional pBAC{3×P3-EGFPAF} expression inG0 embryos

Unlike the pPIGA3GFP vector carrying the Bm-Actin3 promoter, the pBac{3×P3-EGFPaf} vectormediated fluorescence in the embryos, which was visiblethrough the egg chorion and rarely in vitellophages ofthe Nistari strain.

This made it possible to test injection parameters tofacilitate the detection of expression from the fourth dayof embryonic development onwards. However, in mostcases, expression became detectable between 6 and 10days of development (Table 2).

Although the Bm-Actin3 promoter gave strong, earlyexpression in the vitellophages of the eggs (48 h afterinjection), it concealed any potential expression in theembryonic tissues. Expression in eggs driven by this pro-moter is poorly sensitive to injection location (anterioror posterior) and to the presence or absence of the helperplasmid. We used the pBac{3×P3-EGFPaf} vector to testthe effect of injection location on the frequency ofexpression in the target tissue. Injections into the anteriorpart of the egg gave a frequency of expression 3.5 times

Fig. 2. Expression in the five stemmata on the left side of a hatchedG0 larva. Expression was visible on both sides of the head; h: head,t: thorax. The outline of the larva is shown by the dashed line.

higher than that obtained when the DNA solution wasinjected into the posterior part of the egg. Moreover,expression in stemmata was observed only for injectionsinto the anterior part of the egg. If DNA was injectedinto the posterior part of the egg, only thoracic andabdominal nervous tissues expressed the 3×P3-EGFPmarker. This difference demonstrates clearly the positioneffect of the injections. However, injections into the pos-terior part gave a hatching frequency 5.6 times higherthan that for injections into the anterior part of the egg(Table 1). We avoided ventral injection because itresulted in too low a frequency of embryonic develop-ment (data not shown).

We also tested the effect of the helper plasmid on thefrequency of expression from the pBac{3×P3-EGFPaf}vector. We compared the co-injection of the helper plas-mid and vector (in two helper/vector weight ratios: 1:1and 1:10) with injection of the vector alone (Table 2).Surprisingly, expression was observed in the absence ofthe helper plasmid, even late in embryonic development,probably due to the persistence of episomal plasmid cop-ies. Such an expression was never observed in our handswith common plasmid vectors. Among diverse plasmidvectors used, only one carrying a recombinant densoviralvector (Jourdan et al., 1990; Giraud et al., 1992) wasable to give expression in embryonic somatic tissues at2, 4 and 10 days of development (J.-L. Thomas unpub-lished results). We found that the helper plasmidincreased the frequency of the marker expression.Indeed, the frequency of expression with the 1:1 weightratio (1.17:1, H/V molar ratio) was 5–13 times higherthan that in the absence of the helper plasmid (Table 3).In this case, we estimate that about 80% (6.69–1.36/6.69) to 93% (7.99–0.59/7.99) of expression isaccounted for by somatic integrated vector.

We also tested a 1:10 ratio of helper plasmid to vector.In this case, the frequency of vector-mediated fluor-escence was lower, by a factor of 1.2–1.6, than that forthe 1:1 weight ratio of helper plasmid to vector.

3.3. The 3×P3-EGFP marker facilitates efficientscreening for piggyBac germline transgenesis

Based on our preliminary experiments, we chose a 1:1(w/w, 0.5 µg/µl) mixture of vector helper plasmid into

Table 1Comparison of the frequency of EGFP expression following co-injec-tion of the pB3×P3-EGFPaf vector and the pHA3 helper (w/w ratioof 1:1) into the anterior and posterior poles of the eggs

Injection No. injected No. GEP+ No. hatchedlocation

Anterior 320 11 (3.4%) 8 (2.5%)Posterior 420 4 (0.95%) 59 (14%)

250 J.-L. Thomas et al. / Insect Biochemistry and Molecular Biology 32 (2002) 247–253

Table 2Comparison of various proportions of helper plasmid to vector on the frequency of vector expression in the embryonic stemmata. Numbers ofpositive embryos for each recorded day correspond to the total number of positive embryos on that day; ND: not determined

Exp. Conditions No. of eggs No. of GFP-positive embryos at n days of development (%) Ratioinjected

4 6 7 8 11

1 H/V (1:1) 131 ND ND ND ND 2 (1.53) 1:1/1:10H/V (1:10) 218 ND ND ND ND 2 (0.92) 1.65H/V (1:1) 338 1 3 17 27 (7.99) ND ND

2 H/V (1:10) 167 0 0 3 11 (6.59) ND 1:1/1:101.21

V 170 0 0 1 1 (0.59) ND 1:1/0:113.54

3 H/V (1:1) 524 ND ND ND 35 (6.69) ND 1:1/0:1V 808 ND ND ND 11 (1.36) ND 4.92

the posterior third of the egg. The DNA solution wasinjected into the dorsal side, towards the ventral side ofthe egg, where the germ band develops. We injected6335 eggs, from which 2190 larvae hatched (34.6%). Ofthe 1210 mated G0 moths, 1100 were mated in singlepair matings and the remaining 110 G0, all of the samesex, were backcrossed with uninjected moths of the othersex. We obtained 560 broods, including one single pairmating brood with four positive individuals. Thus thepercentage of G0 moths with transgenic progeny was0.08% (1/1210), as it is likely that only one of the twoparents was responsible for the transformation events.Despite this very low frequency, the transgenic individ-uals were easy to identify among the 120,000 eggsscreened (one brood corresponds to about 200–250 eggs)when they were immersed under water or ethanol (Figs.3(b) and 6(b)).

3.4. Pattern of expression during successivedevelopmental stages in the transgenic silkworm

The four G1 moths, two males and two females, werebackcrossed with their wild-type counterpart and G2progeny eggs were screened for GFP. From the fifth dayof embryonic development onwards, we detected GFPnot only in the differentiating stemmata, but also in thenervous system comprising the cerebrum and the ventral

Fig. 3. Expression in a four-day old G2 egg. (a) Bright field, (b) GFPfluorescence system.

ganglionic chain (Figs. 3(b), 4(b) and 5(b)). On dissec-tion, we observed that some peripheral nervous tissuesexpressed the 3×P3-EGFP marker (Figs. 4(b) and 5(b)).Later in development, 3×P3-EGFP marker expressionwas observed in the differentiated stemmata of seven-day old embryo (Fig. 6) and in the differentiating com-pound eyes of the pupae (Fig. 7) and moths (Fig. 8).From the four back-crosses we obtained four G2 batcheswith different Mendelian proportions of GFP positiveembryos. Brood TJL1 had 75% positive embryos, TJL2and TJL3 had 50% and TJL4 had 65% positive embryos.This is consistent with broods TJL2 and TJL3 carryinga single integration and brood TJL1 carrying two inser-tions on two different chromosomes. The situation forbrood TJL4 must be more complex, because the pro-portion (65%) was intermediate between 50 and 75%(see below).

3.5. Evidence for germinal transgenesis by apiggyBac-specific transposition process

Inverse PCR experiments were performed on G2 pro-geny of TJL1, TJL3 and TJL4 G1 parents (Table 3).None of the TJL2 G2 eggs hatched. We identified sixdifferent genomic junction sequences flanking the 5� and3� piggyBac ITR (inverted terminal repeat) sequences inbroods TJL1, TJL3 and TJL4. We observed two vectorinsertions in the progeny of TJL1, one in the progenyof TJL3 and four in the progeny of TJL4. ITRs werebordered by the characteristic TTAA sequence, theknown target site of piggyBac (Wang and Fraser, 1993).Except for TJL4.6, in which the 3� junction sequencewas the original baculovirus DNA sequence cloned intothe p3E1.2 vector all other junction sequences representnovel sequences that are most probably derived from theB. mori genome. PCR analysis showed that TJL1.1 andTJL1.4 carried the same insertion (A), which differed

251J.-L. Thomas et al. / Insect Biochemistry and Molecular Biology 32 (2002) 247–253

Fig. 4. Expression throughout the body of a five-day old G2 embryo (photomontage). The arrows show stemmata. The orange arrowheads showpositive peripheral nervous system. (a) Bright field, (b) GFP fluorescence system.

Fig. 5. Expression in the head and thorax of a five-day old embryo. The arrow shows the stemmata. The orange arrowheads show positiveperipheral nervous system. (a) Bright field, (b) GFP fluorescence system. The brown dot in (a) is an artefact and not the pigmented stemmata.

Fig. 6. Expression in the stemmata of a seven-day old G2 embryo.At the bottom, a neighbouring, negative embryo is shown. (a) Brightfield, (b) GFP fluorescence system.

Fig. 7. Expression in differentiating compound eyes in seven-day oldnymphae. (a) Bright field, (b) GFP fluorescence system.

from that carried by TJL1.3 (B) (Table 2). This confirmsthat the TJL1 transgenic parent carried two insertions.TJL3.2 carried a single insertion identical to that ofTJL4.1 whereas TJL4.2, TJL4.3 and TJL4.6 carried dif-

Fig. 8. Expression in the compound eyes of one of the four G1 mothsobtained. (a) Bright field, (b) GFP fluorescence system.

ferent insertions. Finally, four different insertions werecarried by the TJL4 G1 parent. We found six insertionsin the eight G2 individuals analysed. This means that thetwo G0 parents carried at least six piggyBac integrationsand that it is very likely that only one of the two parentscarried these six integrations.

4. Discussion

In this paper, we demonstrate the potential value ofthe 3×P3-EGFP marker both for the screening of trans-genic B. mori L. individuals and for evaluating improve-ments in the frequency of transgenesis.

In the first published experiments describing the suc-cessful use of a piggyBac transposon for B. mori L.germline transgenesis, the Bm-Actin3EGFP marker was

252 J.-L. Thomas et al. / Insect Biochemistry and Molecular Biology 32 (2002) 247–253

Table 3Identification by iPCR of the genomic insertion sites of the pBac{3×P3-EGFPaf} vector. Six different integration events were identified in eightindividuals from 3 G2 transgenic broods. Two insertions were carried by the TJL1 G1 parent. The TJL3 parent carried one insertion identical toone of the four discovered in the G2 progeny of the TJL4 parent; ND: not determined

InsertionG2 individuals 5� Genomic sequences 3� Genomic sequences

identification

TJL 1.1 and TJL 1.4 A …TTAACGACATTACTACGTGGCTTTTAApiggyBac piggyBacTTAAGGGTGTTTCCATTTATT…TJL 1.3 B ND piggyBacTTAATAAAACTACATTCAATA…TJL 3.2 and TJL 4.1 C …CCGGG (Hae III site) TTAApiggyBac piggyBacTTAAGGANCGNCTNCATTTNT…TJL 4.2 D …CCANACGCTCNACGCAGCCAAATTAApiggyBac piggyBacTTAACGATCATCAAACCGCG…TJL 4.3 E ND piggyBacTTAATGAACGCTAATTTTCAC…TJL 4.6 F …CCGG (Hae III site) TTAApiggyBac piggyBacTTAAATAATAGTTTCTAATTT…

found to be useful for the screening of G1 transgenicindividuals, but such screening was possible only at thelarval stage (Tamura et al., 2000). However, the fre-quency of transgenesis (0.7–3.9%) made it necessary toexamine a large number of G1 larvae several times fromthe first to the third instar. This screening was difficultbecause it was necessary to check moving larvae fed onan artificial diet. When the diet was cut to present a levelplane, the larvae rapidly made excavations like craters,resulting in a large number of different focus levels.Moreover, the rearing of thousands of G1 larvae is verytedious and time-consuming. Thus, it is clear that if amarker such as 3×P3-EGFP could be expressed inembryonic stemmata, the screening process would besimplified. Stemmata differentiate from the fifth day ofdevelopment and are immediately in contact with thetranslucent chorion. Furthermore, eggs are laid in aplane, at a single level, which should also facilitatescreening. In our first experiment, we investigatedwhether all these considerations were useful. Expressionof the 3×P3-GFP marker was visible through the chorionin G0 embryos, although a larger proportion of positiveembryos were detected after dechorionation.

We also observed EGFP production in other tissues.The artificial 3×P3 promoter, containing three optimalbinding sites for Pax 6 homodimers, drives the tissue-specific expression of the GFP gene in the stemmata ofembryos. Expression was also observed in nervous sys-tem tissues including the brain, ventral ganglionic chainand peripheral nervous tissues. These results are consist-ent with those obtained in D. melanogaster by Horn etal. (2000) and Musca domestica by Hediger et al. (2001).Weak fluorescence was only rarely observed in the vitel-lophages of the Nistari strain, which is an advantage inscreening for embryonic somatic expression. We alsothought that, unlike the Bm-Actin3GFP marker, the3×P3-EGFP marker, which was expressed at discretesites in G0 embryos and was specific for embryonictissues, might be useful for evaluation of the conditionsrequired to target poorly represented tissues such asstemmata or nerve precursor cells. This could beextrapolated to other piggyBac vectors carrying pro-

moter of such a restricted tissue specificity and drivingexpression of vital marker gene like GFP gene. It is dif-ficult, if not impossible, with Bm-Actin3GFP, due to itsloose specificity to evaluate the effect of the presence orabsence of the helper plasmid on the frequency ofsomatic embryonic expression or the position effect ofthe site of injection as well. In both cases, expressionspreads in many vitellophages to all parts of the egg.Such evaluation is possible at the larval stage, but onlyafter the death of numerous embryos and the loss of asmany potential interesting results helping the statisticview. We assessed the potential value of the 3×P3-EGFPmarker for evaluation from early stages of embryonicdevelopment, by comparing pBac{3×P3-EGFPaf} vectorinjection with and without injection of the helper plas-mid. Expression was detected from the fourth day ofembryonic development in some cases, but mostly afterseven or eight days of embryonic development (12 daysof incubation are required for the hatching of micro-injected eggs; 10 days are normally required for thehatching of wild-type control eggs), and its frequencywas favoured by the presence of the helper plasmid.Injection of the helper plasmid with the vector, in anequal weight ratio (1.17 molar ratio in favour of thehelper) gave a frequency of expression 5–13 timeshigher than that in the absence of the helper plasmid.The helper plasmid seems to be efficient in mediatingsomatic transgenesis, probably stabilising the vector bymeans of integration. It has to be clear that this con-clusion may be drawn only because 3×P3-EGFP markeris specifically expressed in embryonic tissues. Methodsimproving the efficiency of helper plasmid functionshould be straightforward to assess with the pBac{3×P3-EGFPaf} vector. Our results suggest that to obtain germ-line transgenesis, with the method in our own hands, itwould probably be more relevant to inject the vector intothe posterior part of the egg or into the ventral part wherethe germ cells will differentiate. We injected the vectorinto the posterior third of the egg, on the dorsal side,opposite to the site of germ band differentiation, andpushed the needle inside the vitellus toward the ventralside, where the germ cells will appear in the germ band

253J.-L. Thomas et al. / Insect Biochemistry and Molecular Biology 32 (2002) 247–253

(Miya, 1958; Nardi, 1993; Nakao, 1999). We did notinject into the anterior part, targeting the pronuclei orthe zygote, nor into the ventral part, targeting the germcells, because injections with our system, in these areasof the egg were deleterious for normal embryonic devel-opment and hatching, as expected from Myohara’swork (1994).

We found that all the known qualities of the 3×P3-EGFP marker in species such as D. melanogaster, T.castaneum (Berghammer et al., 1999; Horn et al., 2000)and M. domestica (Hediger et al., 2001) were also valu-able in B. mori L. transgenesis as demonstrated in theIndian polyvoltin Nistari strain and recently in the whiteeye mutant W1-pnd (results not shown). The mostimportant feature of this marker is that it makes it poss-ible to screen large numbers of G1 broods, with ease, atan early embryonic stage. This somatic tissue-specificmarker should also be of value for evaluating methodsto increase the frequency of germline transgenesis.

Acknowledgements

We would like to thank B. Declerieux, B. Perret andL. Fontaine for rearing silkworms; C. Royer and A.-M.Grenier for helpful discussion. We would also like tothank E.A. Wimmer for critical reading of the manu-script, for his kindness and for sharing his data andreagents. In addition, we thank his collaborators: A. Ber-ghammer and C. Horn.

References

Berghammer, A.J., Klinger, M., Wimmer, E.A., 1999. A universalmarker for transgenic insects. Nature 402, 370–371.

Cary, L., Goebel, M., Corsaro, H.H., Wang, H.H., Rosen, E., Fraser,M.J., 1989. Transposon mutagenesis of baculoviruses: analysis ofTrichoplusia ni transposon IFP2 insertions within the FP-locus ofnuclear polyhedrosis viruses. Virology 161, 8–17.

Coulon-Bublex, M., Mounier, N., Couble, P., Prudhomme, J.C., 1993.Cytoplasmic actin A3 gene promoter injected as supercoiled plas-mid is transiently active in Bombyx mori embryonic vitellophages.Roux’s Arch. Dev. Biol. 202, 123–127.

Fraser, M.J., Cary, L., Boonvisudhi, K., Wang, H.H., 1995. Assay formovement of lepidopteran transposon IFP2 in insect cells using abaculovirus genome as a target DNA. Virology 211, 397–407.

Giraud, C., Devauchelle, G., Bergoin, M., 1992. The densovirus ofJunonia ceonia (Jc DNV) as an insect cell expression vector.Virology 186, 207–218.

Hediger, M., Niessen, M., Wimmer, E.A., Dubendorfer, A., Bopp, D.,2001. Genetic transformation of the housefly Musca domestica withthe lepidopteran-derived transposon piggyBac. Insect Mol. Biol. 10,113–119.

Horn, C., Jaunich, B., Wimmer, E.A., 2000. Highly sensitive, fluor-escent transformation marker for Drosophila transgenesis. Dev.Genes Evol. 210, 623–629.

Horn, C., Wimmer, E.A., 2000. A versatile vector set for animal trans-genesis. Dev. Genes Evol. 210, 630–637.

Jourdan, M., Jousset, F.X., Gervais, M., Skory, S., Bergoin, M.,Dumas, B., 1990. Cloning of the genome of a densovirus and res-cue of infectious virus from recombinant plasmid in the insect hostSpodoptera littoralis. Virology 179, 403–409.

Miya, K., 1958. Studies on the embryonic development of the gonadin the silkworm, Bombyx mori L. J. Faculty Agric. Iwate University3, 436–467.

Myohara, M., 1994. Fate mapping of the silkworm, Bombyx mori,using localized UV irradiation of the egg at fertilization. Develop-ment 120 (10), 2869–2877.

Nagaraju, J., Kanda, T., Yukuhiro, K., Chavancy, G., Tamura, T., Cou-ble, P., 1996. Attempt at transgenesis of the silkworm (Bombyxmori L.) by egg-injection of foreign DNA. Appl. Entomol. Zool.31, 587–596.

Nakao, H., 1999. Isolation and characterization of a Bombyx Vasa-likegene. Dev. Genes Evol. 209, 312–316.

Nardi, J.B., 1993. Modulated expression of a surface epitope onmigrating germ cells of Manduca sexta embryos. Development118, 967–975.

Ninaki, O., Maekawa, H., Gamo, T., Koga, K., Sakaguchi, B., 1985.Hatchability of silkworm eggs injected with DNA at early embry-onic stages. J. Seric. Sci. Jpn. 54, 428–434.

Tamura, T., Kanda, T., Takiya, S., Okano, K., Maekawa, H., 1990.Transient expression of chimeric CAT genes injected into earlyembryos of the domesticated silkworm Bombyx mori. Jpn. J. Genet.65, 401–410.

Tamura, T., Thibert, C., Royer, C., Kanda, T., Abraham, E., Kamba,M., Natuo, K., Thomas, J.-L., Mauchamp, B., Chavancy, G., Shirk,P., Fraser, M., Prudhomme, J.-P., Couble, P., 2000. Germline trans-formation of the silkworm Bombyx mori L. using a piggyBac trans-poson-derived vector. Nat. Biotechnol. 18, 81–84.

Wang, H.H., Fraser, M., 1993. TTAA serves as the target site for theTFP3 lepidopteran transposon insertions in both nuclearpolyhedrosis virus and Trichoplusia ni genomes. Insect Mol. Biol.1, 109–116.