The Purine Synthesis Gene Prat2 Is Required for ... - Semantic Scholar

3 downloads 0 Views 487KB Size Report
ABSTRACT. PRAT (phosphoribosylamidotransferase; E.C. 2.4.2.14) catalyzes the first reaction in de novo purine nucleotide biosynthesis. In Drosophila ...
Copyright Ó 2006 by the Genetics Society of America DOI: 10.1534/genetics.105.045641

The Purine Synthesis Gene Prat2 Is Required for Drosophila Metamorphosis, as Revealed by Inverted-Repeat-Mediated RNA Interference Yingbiao Ji and Denise V. Clark1 Department of Biology, University of New Brunswick, Fredericton, New Brunswick E3B 5A3, Canada Manuscript received May 13, 2005 Accepted for publication November 14, 2005 ABSTRACT PRAT (phosphoribosylamidotransferase; E.C. 2.4.2.14) catalyzes the first reaction in de novo purine nucleotide biosynthesis. In Drosophila melanogaster, the Prat and Prat2 genes are both highly conserved with PRAT sequences from prokaryotes and eukaryotes. However, Prat2 organization and expression during development is different from Prat. We used RNA interference (RNAi) to knock down expression of both Prat and Prat2 to investigate their functions. Using the GAL4–UAS system, Prat RNAi driven by Act5c–GAL4 or tubP–GAL4 causes variable pupal lethality (48–100%) and 50% female sterility, depending on the transgenic strains and drivers used. This observation agrees with the phenotype previously observed for Prat EMSinduced mutations. Prat2 RNAi driven by Act5C–GAL4 or tubP–GAL4 also results in variable pupal lethality (61–93%) with the different transgenic strains, showing that Prat2 is essential for fly development. However, Prat2 RNAi-induced arrested pupae have a head eversion defect reminiscent of the ‘‘cryptocephal’’ phenotype, whereas Prat RNAi-induced arrested pupae die later as pharate adults. We conclude that Prat2 is required during the prepupal stage while Prat is more important for the pupal stage. In addition, Prat and Prat2 double RNAi results in more severe pupal lethal phenotypes, suggesting that Prat and Prat2 have partially additive functions during Drosophila metamorphosis.

T

HERE are two pathways for synthesis of purine nucleotides: the salvage pathway and the de novo pathway. The salvage pathway recycles bases with phosphoribosylpyrophosphate. The de novo pathway includes 10 steps to synthesize inosine monophosphate (IMP), the common precursor of adenosine monophosphate and guanosine monophosphate. Disorders of purine synthesis can have severe effects in humans. For example, a mutation in adenylosuccinate lyase, which catalyzes the eighth step in the de novo synthesis pathway, results in the accumulation of SAICAR (5-amino-4-imidazole-Nsuccinocarboxamide ribotide) and is thought to be associated with mental retardation and autism (Stone et al. 1992; Kohler et al. 1999). Mutations in the bifunctional enzyme AICAR transformylase/IMP cycolohydrolase (ATIC), which catalyzes the final two steps in making IMP in the de novo synthesis pathway, cause dysmorphic features, severe neurological defects, and congenital blindness (Marie et al. 2004). Phosphoribosylamidotransferase (PRAT) catalyzes the first reaction in the de novo purine nucleotide biosynthesis pathway, which is a rate-limiting step (Zalkin and Dixon 1992). In Drosophila melanogaster, Prat is essential for development, where mutations cause pupal lethality or escaping adults with wing, leg, and eye 1 Corresponding author: Department of Biology, University of New Brunswick, 10 Bailey Dr., Fredericton, NB E3B 5A3, Canada. E-mail: [email protected]

Genetics 172: 1621–1631 (March 2006)

defects (Clark 1994) and sterile females with reduced life span (Malmanche and Clark 2004). Mutations in other de novo purine pathway genes (ade2, ade3, and ade5) have a similar phenotype to those in Prat, which has been called the purine syndrome phenotype (Tiong and Nash 1990, 1993; O’Donnell et al. 2000). These observations suggest that endogenous purine synthesis is required for metamorphosis of the adult fly during the prepupal and/or pupal stages, whereas maternal and dietary sources of purines are sufficient for embryonic and larval development, respectively. The Prat gene encodes a 546-amino-acid polypeptide with a predicted 53-amino-acid propeptide, which is processed to produce a mature enzyme (Clark 1994). PRAT has two functional domains: a glutaminase domain and a phosphoribosyltransferase domain (Zalkin and Dixon 1992). A second PRAT gene (Prat2) encodes a 547-amino-acid polypeptide with a predicted 57-aminoacid propeptide (Malmanche et al. 2003). The Prat2 polypeptide has 79% identity and all functional domains are conserved with Prat. However, Prat2 differs in gene organization and expression compared with Prat. In Prat2, two introns interrupt the open reading frame while the Prat open reading frame is uninterrupted by introns. In addition, since the two Prat genes reside on different arms of chromosome 3, Prat appears to be a retrotransposed copy of Prat2. Only Prat is maternally expressed, and zygotic expression of Prat2 begins, along with Prat, during embryogenesis (Tomancak et al. 2002;

1622

Y. Ji and D. V. Clark

Malmanche et al. 2003). Both genes are expressed throughout development, but at different levels at different times. Since gene organization and temporal expression patterns are conserved for the two genes in D. melanogaster and D. virilis (Malmanche et al. 2003), which shared a common ancestor 40 million years ago (Russo et al. 1995), the diversification of the two genes is likely to be functionally important. Since no mutation was available for Prat2, we chose to use RNA interference (RNAi) to investigate the role of Prat2 in fly development. Since the transitive RNAi pathway is absent in Drosophila (Roignant et al. 2003), we were able to silence the expression of each of the duplicated genes specifically using RNAi. We made upstream activating sequence Prat-inverted repeat (UAS–Prat-IR) and UAS–Prat2-IR transgenes expressing inverted-repeat (IR) RNA under the control of the GAL4–UAS binary expression system (Brand and Perrimon 1993). Our results show that Prat2 is also an essential gene that has a different role in pupal development in comparison to Prat.

MATERIALS AND METHODS Drosophila stocks and breeding conditions: Flies were cultured on standard cornmeal-molasses-agar media. Unless indicated, the culture temperature was 25°. The w1118 strain and the three GAL4 strains (y1 w*; P{Act5C–GAL4}25FO1/CyO, y1 and y1 w*; P{Act5C–GAL4}17bFO1/TM6B, Tb1 and y1 w*; P{tubP–GAL4}LL7/TM3, Sb1) were obtained from the Bloomington stock center. We also replaced the balancer chromosomes in the y1 w*; P{Act5C–GAL4}25FO1/CyO, y1 and y1 w*; P{tubP-GAL4}LL7/TM3, Sb1 strains with the GFP-bearing balancers from w*; In(2LR)noc4LScorv9R, b1/CyO, P{ActGFP}JMR1 and w*; Sb1/TM3, P{ActGFP}JMR2, Ser1 to make w*; P{Act5C– GAL4}25FO1/CyO, P{ActGFP}JMR1 and w*; P{tubP–GAL4} LL7/ TM3, P{ActGFP}JMR2 Ser. Likewise, balancers in two strains carrying Prat EMS-induced alleles [v; Prat12A19 e11/TM3, Sb e and v; Prat16A6 e11/TM3, Sb e (Clark 1994)] and a Prat region deficiency (v, Df(3R)dsx43 e11/TM3, Sb e) were replaced with the GFP-bearing balancer from w*; Sb/TM3, P{ActGFP}JMR2, Ser to get v; Prat12A19 e11/TM3, P{ActGFP}JMR2 Ser, v; Prat16A6 e11/ TM3, P{ActGFP}JMR2 Ser, and v, Df(3R)dsx43 e11/TM3, P{ActGFP}JMR2 Ser. The balancers of two ade2 mutants [ade23/SM5 and ade24/SM5 (Tiong et al. 1989)] were replaced by the GFP-bearing balancers from w*; In(2LR)noc4LScorv9R, b/ CyO, P{ActGFP}JMR1 to get ade23/CyO, P{ActGFP}JMR1 and ade24/CyO, P{ActGFP}JMR1 stocks. Construction of the Prat and Prat2 transgenes: The propeptide sequences of the Prat and Prat2 genes have no nucleotide sequence identity and thus DNA fragments corresponding to these regions plus the 59 untranslated sequence were used to construct Prat-IR and Prat2-IR transgenes housed in the pUAST Drosophila transformation vector (Brand and Perrimon 1993). For construction of Prat-IR, a 153-bp PCR product was amplified from a Prat cDNA (D. Clark, unpublished results), using the primers 59-TCCTTGCTGGCCGAGATGGATTCCGAT TCC-39 and 59-CAGCAGCAACATGTCAGC-39, and then cloned into the pGEM-T easy vector (Promega, Madison, WI) by TA cloning. At the same time, a 95-bp EcoRI/PstI spacer from the pOT2 plasmid (from the Berkeley Drosophila Genome Project, BDGP) was cloned into the pBluescript SKII vector (Stratagene, La Jolla, CA). Then a 125-bp EcoRI/NotI spacer was cut from the resulting pBluescript SKII plasmid and cloned into pUAST. Next

a 180-bp EcoRI fragment and a NotI fragment containing the 153-bp Prat propeptide fragment in the pGEM plasmid were cloned into pUAST 1 125-bp spacer vector in antisense-sense (AS) orientation or sense-antisense (SA) orientation to make the Prat-IR–AS and Prat-IR–SA constructs, respectively. For construction of Prat2-IR, two fragments were isolated from the Prat2 cDNA GH10034 (obtained from the BDGP). A 311-bp HindIII/PstI fragment (including 166 bp of the Prat2 59-UTR, 50 bp of the propeptide sequence, and 95 bp of the pOT2 vector sequence as a spacer) and a 216-bp EcoRI/ HindIII fragment (including 166 bp of the Prat2 59-UTR and 50 bp of the propeptide sequence) were cloned into pBluescript KSII vector (Stratagene), respectively. Then the 281-bp KpnI/XbaI fragment from the resulting pBluescript KSII 1 216-bp EcoRI/HindIII Prat2 fragment plasmid was cloned into the pVZ1 vector (Henikoff and Eghtedarzadeh 1987). A 291-bp XbaI and SacI fragment from the pVZ1 1 281-bp KpnI/ XbaI plasmid was then added to the pBluescript KSII 1 311-bp HindIII/PstI plasmid. Finally, the 637-bp KpnI fragment from this construct, which has the inverted-repeat Prat2 fragment in the antisense-sense orientation separated by a spacer, was cloned into pUAST to make the Prat2-IR–AS construct. P-element-mediated transformation: Prat-IR–AS, Prat-IR– SA, or Prat2-IR–AS plasmids were co-injected with the P helper plasmid P{p25.7wcD2-3}, a variant of P{p25.7wc} (Karess and Rubin 1984), into the host strain w1118 to produce transgenic flies using standard P-element-mediated transformation techniques (Spradling 1986). We obtained several independent Prat-IR–AS, Prat-IR–SA, and Prat2-IR–AS transgenic fly strains, which are homozygous viable. Southern blot analysis: To determine the copy numbers of Prat-IR and Prat2-IR in transformants, genomic DNA from the transgenic and w1118 strains was digested by HincII, run in a 0.7% agarose gel, and blotted to positively charged nylon membrane (Biodyne). The blots were hybridized with a 2.5-kb DraI and SacI fragment of the mini-white gene from the P-element vector, labeled using a random-primer fluorescein-dUTP kit (Perkin-Elmer, Norwalk, CT), and hybridized at 65° for 14 hr. All blots were detected using antifluorescein-alkaline phosphatase antibody (Perkin-Elmer) and dioxetane chemiluminescent substrate (New England Biolabs, Beverly, MA). The chemiluminescent signal was detected with Kodak Ultident X-ray film and the film was scanned with an Epson Expression 626 scanner with a transparent film adapter. Reverse transcriptase polymerase chain reaction: For all experiments, total RNA was extracted using Trizol (Invitrogen, San Diego) and reverse transcriptase polymerase chain reaction (RT–PCR) was performed using ready-to-go RT–PCR beads (Amersham, Buckinghamshire, UK), where cDNA synthesis was primed with an oligo(dT) primer. For semiquantitative RT–PCR analysis of the effect of Prat RNAi on endogenous expression of Prat and Prat2, non-CyO 0- to 4-hr-old female adults from the crosses between Act5C– GAL4/CyO and Prat-IR–AS or w1118 grown at 25° were collected. RNA from 0- to 4-hr-old female w1118 adults was used to establish amplification standard curves for Prat, Prat2, and RpL32 and to determine the exponential amplification range (Foley et al. 1993). On the basis of the curves, 23 cycles for Prat, 22 cycles for Prat2, and 16 cycles for RpL32 were chosen (data not shown). For semiquantitative analysis of the effect of Prat2 RNAi on endogenous expression of Prat and Prat2, non-GFP and GFP wandering third instar larvae from the progeny of UAS–Prat2IR–AS or w1118 crossed with Act5C–GAL4/CyO, ActGFP grown at 29° were sorted using an epifluorescent stereomicroscope (Zeiss). These larvae (and the wandering third instar larvae of w1118 grown at 29°) were then sorted by sex. RNA from male and female w1118 wandering third instar larvae was used to

Prat and Prat2 in Drosophila

1623 Figure 1.—Southern blot analysis of Prat-IR–AS and Prat2-IR–AS transformants. (A) Map for the P-element transgenes, showing the SacI/DraI restriction fragment from the mini-white gene used to probe the blots shown in B and C. This probe hybridizes to 0.9and 1.4-kb HinCII fragments from the white locus on the X chromosome, and the 0.9-kb HinCII fragment from the P element comigrates with the fragment of the same size. (B) HinCII digests of genomic DNA from Prat-IR–AS transformants. Lane 1, Prat-IR–AS-1; lane 2, Prat-IR–AS-2; lane 3, Prat-IR–AS-3; lane 4, Prat-IR–AS-4; lane 5, Prat-IR– AS-5; lane 6, w1118. (C) HinCII digests of genomic DNA from Prat2-IR–AS transformants. Lane 1, Prat2-IR–AS-1; lane 2, Prat2-IR–AS-2; lane 3, Prat2IR–AS-3; lane 4, Prat2-IR–AS-4; lane 5, Prat2-IR–AS-5; lane 6, w1118.

establish amplification standard curves for Prat, Prat2, and RpL32 and to determine the exponential amplification range. On the basis of the curves, 25 and 15 cycles were chosen to coamplify Prat and RpL32, respectively, and 22 and 16 cycles were chosen to coamplify Prat2 and RpL32, respectively. The primers for the amplification of the three genes and detection methods were described previously (Malmanche et al. 2003) except that the chemiluminescent signal was detected using the Bio-Rad (Hercules, CA) ChemiDoc system and measured by Bio-Rad Quantity-One software. All the experiments were performed twice, starting from the RNA extractions. Phenotypic analysis of Prat RNAi and Prat2 RNAi: Prat-IR– AS and Prat2-IR–AS transgenic strains and w1118 were crossed to Act5C–GAL4/CyO, ActGFP or tubP–GAL4/TM3, Ser, ActGFP and the parents were allowed to lay eggs for 8 hr. After 26 hr, the GFP and non-GFP first instar larvae were counted using an epifluorescent stereomicroscope to determine if embryonic lethality had occurred. In the same way, other samples of eggs were counted and transferred into bottles to grow until the wandering third instar stage, and then GFP and non-GFP larvae were counted to determine if larval lethality had occurred. Finally, pupal lethality was determined on the basis of the ratio between the Cy and non-Cy adults or Ser and non-Ser adults. For phenotypic analysis of the arrested pupae caused by Prat RNAi, Prat2 RNAi, double RNAi, and EMS-induced Prat and ade2 alleles, 100 GFP and non-GFP wandering third instar larvae were sorted, put into vials, and allowed to develop for 100 hr. The arrested pupae were staged and counted on the basis of morphological markers of development according to Bainbridge and Bownes (1981). Pupae were photographed using a Zeiss stereomicroscope equipped with a Pixera camera.

RESULTS

The BDGP expressed sequence tag (EST) project (Rubin et al. 2000) revealed two classes of Prat2 transcripts (a short Prat2–RA and a full-length Prat2–RB) (FlyBase 2003). Twenty ESTsequences from a variety of cDNA libraries match with Prat2–RB. Only a single EST (LD14516) from an embryonic cDNA library supports Prat2–RA, which has its first exon within the third intron

of Prat2–RB. In addition, using RT–PCR of embryos, we could not detect Prat2–RA whereas we could detect Prat2–RB (Y. Ji, unpublished data). These observations show that Prat2–RB is most likely the major transcript of Prat2 and that the EST of Prat2–RA likely reflects a rare, aberrantly spliced transcript. Thus, in designing a transgene for RNA-interference-mediated knockdown of Prat2, we used sequences derived from the 59-end of the Prat2–RB transcript and not Prat2–RA. An additional consideration is the fact that Prat and Prat2 are highly similar at the nucleotide level, save for the untranslated and propeptide regions. Thus, to ensure that the RNAi would be gene specific, we chose fragments containing 59-UTR and propeptide sequence for both Prat and Prat2 for construction of the RNAi transgenes. RNAi-mediated Prat loss of function causes pupal lethality and adult female sterility: The general organization of the Prat RNAi transgenes is shown in Figure 1A. Either the sense-antisense or antisense-sense orientation of inverted repeats has been used in different studies to silence a specific gene (Fortier and Belote 2000; Piccin et al. 2001). Therefore, we decided to test if the orientation of the inverted repeats in the RNAi transgene has an influence on silencing ability. We cloned fragments containing the propeptide sequence of Prat in both the AS and SA orientations. We obtained five Prat-IR–AS and four Prat-IR–SA transgenic strains where the insertions are all homozygous viable. Southern blot analysis showed that the transgenes are all intact (data not shown), that four Prat-IR–AS strains have a single copy of the transgene and one has two copies (Figure 1B), and that all Prat-IR–SA strains have a single copy of the transgene (data not shown). To induce expression of the transgenes, we crossed Prat-IR–AS and Prat-IR–SA transgenic strains with strains carrying the GAL4 drivers (Act5C–GAL4 and tubP– GAL4), which have ubiquitous expression of GAL4

1624

Y. Ji and D. V. Clark TABLE 1 Lethality of Prat-IR-AS and Prat-IR-SA strains expressing GAL4

Parental genotypes

Prat-IR/ GAL4

Percentage lethalitya

Prat-IR/ Balancer

Copies of IR transgene

$

#

$

#

$

#

0 2 1 1 1 2 1 1 1 1 0 1 1 0 2 1 1

407 147 68 6 29 22 95 94 115 107 98 57 206 165 65 28 39

395 52 20 0 6 2 8 94 106 104 93 26 102 215 25 8 9

362 460 203 262 278 314 182 84 101 113 42 120 257 116 174 226 214

358 398 221 239 249 257 195 83 104 102 50 87 216 127 217 208 220

0 68 67 98 89 93 48 0 0 0 0 80 65 0 74 91 87

0 86 91 100 98 99 96 0 0 0 0 82 72 0 93 98 98

1118

(#) Act5C-GAL4/CyO ($) 3 w Act5C-GAL4/CyO ($) 3 Prat-IR-AS-1 (#) Act5C-GAL4/CyO ($) 3 Prat-IR-AS-2 (#) Act5C-GAL4/CyO ($) 3 Prat-IR-AS-3 (#) Act5C-GAL4/CyO ($) 3 Prat-IR-AS-5 (#) Act5C-GAL4/CyO ($) 3 Prat-IR-AS-2; Prat-IR-AS-5 (#) Prat-IR-AS-4 ($) 3 Act5C-GAL4/CyO (#) Act5C-GAL4/CyO ($) 3 Prat-IR-SA-1 (#) Act5C-GAL4/CyO ($) 3 Prat-IR-SA-2 (#) Act5C-GAL4/CyO ($) 3 Prat-IR-SA-3 (#) Act5C-GAL4/TM6B ($) 3 w1118 (#) Act5C-GAL4/TM6B ($) 3 Prat-IR-AS-3 (#) Act5C-GAL4/TM6B ($) 3 Prat-IR-AS-5 (#) w1118 ($) 3 TubP-Gal4/TM3 (#) Prat-IR-AS-1 ($) 3 tubP-GAL4/TM3 (#) Prat-IR-AS-2 ($) 3 tubP-GAL4/TM3 (#) Prat-IR-AS-4 ($) 3 tubP-GAL4/TM3 (#) a

Percentage lethality ¼ ½(expected number of Prat-IR-AS/GAL4 progeny  observed number)/expected number 3 100%. Expected numbers were normalized to the progeny of parallel crosses between the GAL4 driver strain and the w1118 strain.

(Lee and Luo 1999; FlyBase 2003), and scored the adults (Table 1). The results show that flies carrying PratIR–AS and expressing GAL4 are reduced in number by 50% to almost 100% relative to their siblings. However, flies carrying Prat-IR–SA and expressing Act5C–GAL4 did not show any lethality (Table 1). Therefore, this observation suggests that the orientation of the invertedrepeat hairpin in the transgene may have some effect on silencing ability. We tested the fertility of the surviving adult flies carrying Prat-IR–AS and Act5C–GAL4 and found that 64% (n ¼ 100) of surviving females carrying Prat-IR–AS-5 and Act5C–GAL4 were sterile, 50% (n ¼ 107) of females carrying Prat-IR–AS-1 and Act5C–GAL4 were sterile, while 6% (n ¼ 100) of the wild-type control (Act5C–GAL4/1) were sterile. The sterility caused by Prat RNAi shows that female germline development requires de novo purine synthesis, as was previously observed for EMS-induced Prat loss-of-function alleles (Malmanche and Clark 2004). To determine the lethal phase for Prat RNAi, we crossed the Prat-IR–AS-4 strain to the Act5C–GAL4/CyO, Act–GFP strain. Taking advantage of the GFP marker on the CyO balancer chromosome, we could distinguish the progeny carrying the RNAi transgene and GAL4 (nonGFP) from its sibling not carrying GAL4 (GFP) in any stage of fly development. We found that the ratio of GFP and non-GFP progeny in first instar larvae and the wandering third instar larvae was 1:1 (Table 2). Therefore, there is no embryonic or larval lethality associated with Prat-IR–AS driven by Act5C–GAL4 and so the individuals

not surviving to adult are all dying within the prepupal and/or pupal stages. The percentage of pupal lethality of the progeny carrying Prat-IR–AS and GAL4 depends on the different transgenic strains, sex, drivers, and culture temperature. The progeny from five independent transgenic strains crossed with Act5C–GAL4/CyO had different levels of pupal lethality ranging from 48 to 100% (Table 1). This variation likely reflects position effects on transgene transcription. For example, one Prat-IR–AS strain (Prat-IR–AS-1), which has two copies of the transgene, did not show a strong RNAi effect as was seen for the other three Prat-IR–AS strains with one copy. The Prat-IR–AS-1 white gene marker has less expression, seen as eye pigment, than the other three Prat-IR–AS P elements, and so it is possible that one or both of the PratIR–AS-1 elements has been inserted into a transcriptionally repressive chromatin domain. Transgene copy number can also influence expression. After making a strain carrying both Prat-IR–AS-2 and Prat-IR–AS-5, we found increased pupal lethality compared to flies carrying either single-copy Prat-IR–AS (Table 1). In addition, we found that males carrying the Prat-IR–AS transgenes and GAL4 have higher pupal lethality than that of females (Table 1). In particular, one strain (Prat-IR–AS-4), with the transgene on the X chromosome, had 96% male lethality but 48% female lethality, suggesting that dosage compensation can influence the transcription of the transgene even under control of GAL4 protein. The use of a different ubiquitous GAL4 driver also has a profound effect on RNAi efficiency. We found that the

Prat and Prat2 in Drosophila

1625

TABLE 2 Embryonic and larval survival of flies carrying Prat-IR-AS or Prat2-IR-AS or both and GAL4 at 25°

Parental genotypes w1118 ($) 3 Act5C-GAL4 / CyO, GFP (#) Prat-IR-AS-4 ($) 3 Act5C-GAL4/ CyO, GFP (#) Prat2-IR-AS-4; Prat2-IR-AS-1 ($) 3 Act5C-GAL4/CyO, GFP Prat-IR-AS-4; Prat2-IRAS-4; Prat2-IRAS-1 ($) 3 Act5C-GAL4 /CyO, GFP

First instar larvae non-GFP:GFP

Third instar larvae non-GFP:GFP

366:361** 224:193** 229:241** 344:354**

143:158** 108:109** 162:164** 267:251**

** P $ 0.05 based on the chi-square test for goodness of fit to a 1:1 ratio.

tubP–GAL4 driver can induce a stronger RNAi effect than the Act5C–GAL4 driver, as reported by others (Giordano et al. 2002; Zhu and Stein 2004). In addition, we found that the Act5C–GAL4 driver on the second chromosome induces a stronger RNAi effect than the Act5C–GAL4 driver on the third chromosome. We also found that a higher culture temperature (29°) can increase the silencing ability of Prat RNAi (data not shown), as found in other reports (Fortier and Belote 2000; Negeri et al. 2002), presumably due to the increased GAL4 activity (Duffy 2002). These observations support the idea that the degree of lethality caused by Prat RNAi is dependent on the level of transgene expression, whether it is influenced by position effects, copy number, dosage compensation, or culture temperature. RNAi-mediated Prat2 loss of function also causes severe pupal lethality: We generated five Prat2-IR–AS transgenic strains that are homozygous viable. Southern blot analysis showed that the transgenes are intact (data not shown) and two of the five strains have multiple copies of the transgene (Figure 1C). To examine the Prat2 RNAi-induced phenotype, we crossed Prat2-IR transgenic strains with a Act5C–GAL4 or tubP–GAL4 strain at 25° and 29° (Table 3). We found that flies carrying four copies (Prat2-IR–AS-1) and expressing tubP– GAL4 showed 57% female lethality and 75% male lethality. However, in contrast to Prat-IR–AS, single copies of Prat2-IR–AS were unable to induce lethality in combination with either GAL4 driver at 25°. At 29°, we found that flies carrying more than one copy of Prat2IR–AS and Act5C–GAL4 had 82–100% pupal lethality, whereas flies carrying only one copy of Prat2-IR–AS and Act5C–GAL4 had less lethality, up to 60% (Table 3). However, we also found a high level of lethality in our control due to the nonspecific effects of Act5C–GAL4 at a higher temperature. Therefore, we wanted to determine if we could also increase the effect of Prat2 RNAi by increasing copy number further but culturing the flies at 25°. We constructed two strains carrying five copies (Prat2AS-4; Prat2AS-1 and Prat2AS-5; Prat2AS-1). Flies carrying five copies of Prat2-IR–AS driven by tubP–GAL4 had 90% pupal lethality, whereas those carrying Act5C– GAL4 had 70% pupal lethality (Table 3). The reciprocal crosses showed similar results (data not shown). These results for Prat2 RNAi again show that position effects,

the copy number of the transgene, and culturing temperature affect RNAi silencing ability. To further characterize the lethality caused by RNAimediated Prat2 loss of function, we collected eggs laid over 8 hr from crosses of Prat2-IR–AS-4; Prat2-IR–AS-1 with Act5C–GAL4/CyO, ActGFP and counted the ratio of the non-GFP and GFP in the first instar larval stage and the wandering third larval stage. We found no embryonic or larval lethality due to Prat2 RNAi (Table 2) and so we concluded that Prat2 RNAi causes pupal lethality. We examined the fertility of surviving adults carrying five copies of Prat2-IR–-AS and tubP–GAL4. We did not observe any female sterility (data not shown), as we expected, since Prat2 is not expressed maternally (Malmanche et al. 2003). RNAi is specific for the target gene: We used RT– PCR to determine if Prat RNAi specifically knocks down the expression of endogenous Prat mRNA but not Prat2 mRNA. Since 0- to 4-hr adult females have very strong Prat maternal expression (Malmanche et al. 2003), we chose this stage to measure Prat and Prat2 RNA levels in flies carrying Prat-IR–AS and Act5C–GAL4. The adult females carrying Prat-IR–AS-1 and Act5C–GAL4 had 31% of the Prat expression level relative to the control (Act5C–GAL4/1), while the adult female flies carrying Prat-IR–AS-5 transgene and Act5C–GAL4 had only 24% of the Prat expression level in the control (Table 4 and Figure 2A), consistent with the sterility that we observed for these females. Overall, the reduction in expression that we see is comparable to the results of Kennerdell et al. (2002). In contrast, Prat2 expression did not show a reduction but an increase (Table 4 and Figure 2A), demonstrating that Prat RNAi knocked down the expression of endogenous Prat mRNA specifically. Further investigation is needed to determine if the increased expression of Prat2 truly results from upregulation in the background of lower Prat expression. We also used RT–PCR to determine if Prat2 RNAi specifically knocks down the expression of endogenous Prat2 RNA but not Prat RNA. We chose to measure Prat2 and Prat expression using a culture temperature of 29° due to the stronger pupal lethality observed for flies carrying Prat2-IR–AS and Act5C–GAL4 at this temperature (Table 3). We examined third instar larvae for two independent transgenic lines (Prat2-IR–AS-1 and

1626

Y. Ji and D. V. Clark TABLE 3 Lethality of Prat2-IR-AS lines expressing GAL4

No. of IR transgenes

Parental genotypes 1118

Act5C-GAL4/CyO ($) 3 w

(#)

0

Act5C-GAL4/CyO ($) 3 Prat2-IR-AS-1 (#)

4

Act5C-GAL4/CyO ($) 3 Prat2-IR-AS-2(#)

2

Act5C-GAL4/CyO ($) 3 Prat2-IR-AS-3(#)

1

Act5C-GAL4/CyO ($) 3 Prat2-IR-AS-4(#)

1

Act5C-GAL4/CyO ($) 3 Prat2-IR-AS-5(#)

1

Act5C-GAL4/CyO ($) 3 Prat2-IR-AS-4; Prat2-IR-AS-1(#) Act5C-GAL4/CyO ($) 3 Prat2-IR-AS-5; Prat2-IR-AS-1(#) w1118 ($) 3 tubP-GAL4/TM3 (#) Prat2-IR-AS-1($) 3 tubP-GAL4/TM3 (#) Prat2-IR-AS-4; Prat2-IR-AS-1 ($) 3 tubP-GAL4/TM3 (#) Prat2-IR-AS-5; Prat2-IR-AS-1 ($) 3 tubP-GAL4/TM3 (#)

5 5 0 4 5 5

Prat2-IRAS/GAL4

Percentage lethalitya

Prat2-IR-AS/ Balancer

Temperature

$

#

$

#

$

#

25 29 25 29 25 29 25 29 25 29 25 29 25 25 25 25 25 25

407 150 252 0 118 47 166 92 226 224 180 230 86 80 165 87 34 11

395 29 282 0 101 2 150 26 212 75 160 66 42 51 215 72 41 9

362 227 275 205 74 254 126 229 218 377 162 407 279 259 116 144 198 70

358 168 302 118 77 207 105 100 180 263 170 291 187 212 127 169 225 79

0 34 0 100b 0 82b 0 60b 0 41 0 43 70 69 0 57 88 89

0 82 0 100b 0 99b 0 74 0 68 0 77 78 76 0 75 89 93

a Percentage lethality ¼ ½(expected number of Prat2-IR-AS/GAL4 progeny  observed number)/expected number 3 100%. Expected numbers were normalized to the progeny of parallel crosses between the GAL4 driver strain and the w1118 strain. b Values for lethality that are significantly different between Prat2-IR-AS/Act5C-GAL4 and the control (1/Act5C-GAL4) at 29° (P # 0.01 using a chi-square test).

Prat2-IR–AS-2). In this stage, Prat2 has a relatively higher expression than Prat (Malmanche et al. 2003). Two Prat2-IR–AS transgenic lines (Prat2-IR–AS-1 with four copies of the transgene and Prat2-IR–AS-2 with two copies of the transgene) and w1118 were crossed to Act5C– GAL4/CyO, Act–GFP at 29°. RT–PCR analysis showed that Prat2 RNAi reduced the endogenous Prat2 expression level in the third larval stage by 80%, with variation among the different strains and sex (Table 4 and Figure 2B). In contrast, Prat expression did not show a reduction but instead a small increase in some cases (Table 4 and Figure 2B), which is consistent with the situation for Prat2 expression during Prat RNAi.

Pupal lethal phenotypes induced by RNAi are different for Prat and Prat2: For Prat RNAi, we found that most of the lethal arrests occur just prior to eclosion except that a small proportion of flies arrest in the early pupal stage P5 (Table 5). The arrested pharate adults have all the external adult body structures except that the body has patches of necrotic tissues and/or missing abdominal tergites (Figure 3B). Although the female adults that do survive show sterility, characteristic of the EMS-induced alleles (Malmanche and Clark 2004), they do not show the same wing or leg defects. We further characterized the pupal lethal phenotype caused by two EMS-induced Prat alleles (Prat12A19 and Prat16A6) in

TABLE 4 Effect of Prat-IR-AS and Prat2-IR-AS expression on Prat2 and Prat RNA levels Prat2 Genotype Prat-IR-AS-1 /Act5C-GAL4 ($) Prat-IR-AS-5 /Act5C-GAL4 ($) Prat2-IR-AS-1 /Act5C-GAL4 (#) Prat2-IR-AS-1 /Act5C-GAL4 ($) Prat2-IR-AS-2 / Act5C-GAL4 (#) Prat2-IR-AS-2 / Act5C-GAL4($)

Prat

1

2

Average

1

2

Average

1.51 7.60 0.09 0.12 0.15 0.26

2.64 1.37 0.47 0.22 0.25 0.41

2.07 4.58 0.28 0.17 0.20 0.34

0.32 0.16 1.23 0.55 2.07 0.68

0.31 0.31 1.30 1.52 1.38 1.46

0.31 0.24 1.27 1.04 1.72 1.07

Numbers indicate expression level of Prat2 or Prat normalized with that of RpL32 and relative to the expression level in a control genotype lacking the RNAi transgene (see materials and methods). Columns 1 and 2 show the results from two replicates.

Prat and Prat2 in Drosophila

Figure 2.—Examples of RT–PCR analysis showing that RNAi is specific for the target gene. (A) Prat RNAi specifically reduces endogenous Prat RNA expression. RNA from 0- to 4-hr adult females was amplified from the following genotypes: lane 1, Prat-IR–AS-1/Act5C–GAL4; lane 2, Act5C–GAL4/1; lane 3, Prat-IR–AS-5/Act5C–GAL4. (B) Prat2 RNAi specifically reduces endogenous Prat2 RNA expression. RNA from third instar larvae was amplified from the following genotypes: lane 4, Prat2-IR—AS-1/Act5C–GAL4 (male); lane 5, Act5C–GAL4/1 (male); lane 6, Prat2-IR–AS-1/Act5C–GAL4 (female); lane 7, Act5C–GAL4/1 (female).

hemizygotes carrying one of these alleles and the Prat deficiency Df(3R)dsx43 (Clark 1994). Most of the Prat12A19/Df flies arrested in the early pupal stage P5 (Table 5, Figure 4H) and most of Prat16A6/Df flies arrested just prior to eclosion (Table 5). Thus, Prat RNAi behaves like the weak Prat allele (Prat16A6) with respect to pupal stage of arrest. Some of the Prat12A19/Df and Prat16A6/Df flies arrested at the P10 stage (red-eye stage) with patches of necrotic tissues on wing and legs (Figure 3, C and D, respectively) as was also seen in the bodies of arrested pharate adults caused by Prat RNAi (Figure 3B). For Prat2 RNAi, in flies carrying Prat2-IR-4; Prat2-IR-1 and tubP–GAL4, we found that 50% of arrested pupae had head eversion defects (Table 5 and Figure 3E), which is typical of the ‘‘crytocephal’’ phenotype seen

1627

for crc (crytocephal) (Hewes et al. 2000) and other ecdysone-responsive genes involved in metamorphosis (Fletcher et al. 1995). As for the common features of the crytocephal phenotype, the arrested pupa still kept its larval mouth parts attached to the imaginal head structure, did not fully elongate its legs, and its abdomen lacked cuticle and segmentation, although it developed eye pigments, wings, and thoracic bristles (Figure 3E). The translocation of the gas bubble in the prepupa is responsible for head eversion (Chadfield and Sparrow 1985) and we observed that the gas bubble failed to displace from the posterior to the anterior whenever we saw a failure in head eversion at 14 hr after pupariation. The cryptocephal phenotype was observed in animals carrying the Prat2-IR transgene and either the Act5C–GAL4 or the tubP–GAL4 driver. We also observed rarely that arrested pupae carrying Prat2-IR-5; Prat2-IR-1 and Act5C–GAL4 had a microcephalic head phenotype (Figure 3F) resulting from partial head eversion. This phenotype is also rare for crc and E74 mutations (Fletcher et al. 1995). The pupae with a microcephalic head phenotype retain their larval mouthparts and do not have a fully differentiated abdomen (Figure 3F). Since head eversion is a marker for the transition from the prepupal to pupal stage, we conclude that Prat2 is required during the prepupal stage while Prat is more important for the pupal stage. Double RNAi for Prat and Prat2 causes more severe phenotypes: To investigate the effect of Prat and Prat2 double RNAi on fly development, we constructed a strain with one copy of Prat-IR–AS on the X chromosome and five copies of Prat2-IR–AS on the second and third chromosomes (Prat-IR–AS-4; Prat2-IR–AS-4; Prat2IR–AS-1). This strain was crossed with either Act5C–GAL4 or tubP–GAL4 strains. The female progeny carrying Prat-IR–AS and Prat2-IR–AS transgenes and tubP–GAL4 had higher lethality with a significant difference compared to the female lethality caused by Prat or Prat2 RNAi alone using the parental lines (Table 6). This result suggests that Prat and Prat2 may have partially

TABLE 5 Prat and Prat2 RNAi causes mutant phenotypes with variable expressivity Pupal stagea Genotype v,e11/Df (3R)dsx43 e11 v, Prat12A19 e11/Df (3R)dsx43 e11 v, Prat16A6 e11/Df (3R)dsx43 e11 tubP–GAL4/1 Prat-IR-AS-4/tubP–Gal4 Prat2-IR–AS-4/1; Prat2-IR–AS-1/tubP–GAL4 Prat-IR–AS-4/1; Prat2-IR–AS-4/1; Prat2-IR–AS-1/tubP–GAL4 ade23/ade24 a

Stages according to Bainbridge and Bownes (1981).

Total 77 81 79 105 82 120 127 85

P2

P4(i)

P5

P10

63 4

11 9

crc

7 34 7

43

30 50

44 9

P15

Adult

3 5 32

74 2 34 105 8 30 5 5

67 46 6 23

1628

Y. Ji and D. V. Clark

Figure 3.—The pupal lethal phenotype caused by Prat RNAi, Prat EMS-induced mutations, and Prat2 RNAi. (Top) The dorsal side for each genotype. (Bottom) The corresponding ventral side of the same individual. (A) Stage P15 of the wild-type genotype (tubP–GAL4/1) at 96 hr after puparium formation. The following arrested pupae were photographed at 96 hr after puparium formation: (B) Prat-IR–AS-1/ tubP–GAL4; (C) Prat12A19 e11/ Df(3R)dsxR43 e11 (in pupa case); (D) Prat16A6 e11/Df(3R)dsxR43 e11 (in pupa case); (E) the cryptocephalic phenotype for Prat2-IR–AS4/1; Prat2-IR–AS-1/tubP–GAL4; (F) the microcephalic phenotype for Prat2-IR–AS-5/1; Prat2-IR–AS1/Act5C–GAL4. M, larval mouthhook; N, necrotic tissues; T, missing abdominal tergite.

additive functions. We also found that double Prat and Prat2 RNAi caused no embryonic or larval lethality (Table 2). We then characterized the pupal lethal phenotypes caused by Prat and Prat2 double RNAi. When the arrested pupae, induced by double Prat and Prat2 RNAi, were observed 96 hr after pupariation at 25°, we found that 27% of flies arrested at a very early prepupal stage

equal to the P2 stage of wild-type flies (Figure 4A and Table 5). Some of these arrested prepupae kept the thin and long larval shape, the anterior spiracle had failed to evert, and the gas bubble remained in the middle of the body, failing to move to the posterior (Figure 4B). Thirty-four percent of flies arrested at the late prepupal stage P4(i) and featured head eversion defects, the presence of the gas bubble in the posterior, and a

Figure 4.—The early pupal lethal phenotype caused by Prat/Prat2 double RNAi and ade2 mutations. The dorsal sides of the arrested pupae were photographed 96 hr after puparium formation. (A) P2 prepupa of the wild-type tubP–GAL4/ 1 photographed 6 hr after puparium formation. (B) Arrested P2 prepupa caused by Prat/Prat2 double RNAi. (C) Arrested P2 ade23/ade24 prepupa. (D) P5 pupa of the wild-type tubP–GAL4/1 photographed 14 hr after puparium formation. (E) Arrested P4(i) prepupa by Prat/Prat2 double RNAi. (F) Arrested P5 pupa by Prat/Prat2 double RNAi. (G) Arrested P5 ade23/ade24 pupa. (H) Arrested P5 Prat12A19 e11/Df(3R)dsxR43 e11 pupa.

Prat and Prat2 in Drosophila

1629

TABLE 6 Lethality of Prat-IR-AS plus Prat2-IR-AS lines expressing GAL4

Parental genotype w1118 ($) 3 Act5C-GAL4/CyO (#) Prat-IR-AS-4; Prat2-IR-AS-4; Prat2-IR-AS-1 ($) 3 Act5C-GAL4/CyO (#) w1118 ($) 3 tubP-GAL4/TM3 (#) Prat-IR-AS-4; Prat2-IR-AS-4; Prat2-IR-AS-1 ($) 3 tubP-GAL4/TM3 (#)

Prat-IR-AS 1 Prat2-IR-AS/ GAL4

Prat-IR-AS 1 Prat2-IR-AS/ Balancer

Percentage lethalitya

$

#

$

#

$

#

211 69 165 11

202 0 215 17

234 240 116 240

187 273 127 273

0 71 0 97b

0 100 0 96

a Percentage lethality ¼ ½(expected number of Prat-IR-AS 1 Prat2-IR-AS/GAL4 progeny  observed number)/expected number 3 100%. The expected progeny were normalized to the progeny of parallel crosses between the GAL4 driver strain and the w1118 strain. b Value for lethality of the Prat and Prat2 RNAi transgenic flies that is significantly different from the values found for both of the parental lines (P # 0.01 using a chi-square test).

shrunken abdomen (Table 5 and Figure 4E). This class did not have further development in eye color or wing and leg imaginal structures as seen with the cryptocephal phenotype induced by Prat2 RNAi alone. Twentyfour percent of flies had a successfully everted head that had partially collapsed, resulting in a big cavity in the anterior of the pupa (Table 5 and Figure 4F). This class also did not have further adult differentiation and in some cases the abdomen had shrunken greatly, suggesting that they had arrested at an early pupa stage (P5) (Figure 4D), similar to the phenotype for the strong Prat allele Prat12A19 (Figure 4H). Interestingly, this phenotype is also similar to the ‘‘head/abdomen-collapsed’’ phenotype seen for the 59 group of crc mutations (Hewes et al. 2000), although the authors indicated that many of these pupae had the features of further development such as eye pigmentation. In addition, 7% of arrested pupae had the cryptocephal phenotype as caused by Prat2 RNAi and 5% of pupae died as pharate adults (Table 5). Since the double Prat/Prat2 RNAi caused more severe phenotypes compared with single Prat or Prat2 RNAi, it appears that Prat and Prat2 have partially additive functions. To ask whether the lethal phenotype caused by double RNAi for Prat and Prat2 has a similar phenotype as that caused by the mutations of single-copy genes that act downstream in the de novo purine pathway, we characterized the pupal lethal phenotype of ade2 by crossing two EMS-induced ade2 lethal alleles (ade23 and ade24) (Tiong et al. 1989). We found that 8% of ade23/ade24 pupae arrested at the P2 stage with defects in eversion of the anterior spiracle and failure of gas bubble movement (Table 5 and Figure 4C), which is very similar to the earliest arrested prepupae caused by double RNAi for Prat and Prat2 (Figure 4B). Fifty-nine percent of the ade23/ade24 pupae arrested at the P5 stage (Table 5 and Figure 4G). This phenotype is very similar to a subset of the arrested pupae caused by the double RNAi for Prat and Prat2 (Figure 4F) and the strong Prat allele Prat12A19

(Figure 4H). Thus, Prat and Prat2 together may be required to supply the overall enzyme activity to fulfill the committed step of the de novo pathway. DISCUSSION

RNAi gene silencing has been used extensively to address gene function. There are two major differences in the RNAi mechanism between Drosophila and other systems such as plants and C. elegans. The first difference is that RNAi in Drosophila lacks the amplification process, called transitive RNAi (Roignant et al. 2003), which produces secondary siRNA and leads to further degradation of the mRNA sequence, upstream of the initial sequence in C. elegans and plants, and even downstream in plants (Vaistij et al. 2002). This difference makes it possible in Drosophila to silence a specific isoform or a paralog of a gene. In this report, we selected the 59-UTR and propeptide regions of Prat and Prat2, which have no sequence similarity, as the target region of RNAi and specifically knocked down each gene independently. The absence of the siRNA amplification step in Drosophila means that the silencing ability depends on the relative expression levels of the target gene and the transgene (Negeri et al. 2002). This view is supported by our finding that Prat RNAi resulted in higher levels of pupal lethality for males than for females, corresponding to the higher expression of Prat in females (Malmanche et al. 2003). We also observed that a single copy of the Prat-IR–AS transgene can cause lethality but five copies of the Prat2-IR–AS transgene are required to get a similar level of percentage of lethality. This difference may reflect a more abundant expression of Prat2 than Prat during the stages of fly metamorphosis. Although our results showed that the antisense-sense orientation of the UAS–Prat-IR transgene has greater silencing ability than the sense-antisense orientation, we did not similarly compare the effect of the orientation on Prat2

1630

Y. Ji and D. V. Clark

silencing and we are not aware of any other report on this issue. Thus, further comparison of this effect by knocking down many different genes by RNAi needs to be done to determine if we can make any generalization about this effect. The second difference between RNAi in Drosophila and in plants and C. elegans is the absence of system silencing so that siRNA cannot spread from one tissue to another (Roignant et al. 2003). Thus, in Drosophila, for genes whose functions are cell autonomous, RNAi can be used to knock down the gene in a tissue-specific way. However, for genes with non-cell-autonomous functions, it would be difficult to get a tissue-specific null phenotype. In our study, the adult escapers caused by Prat RNAi and Prat2 RNAi did not show the wing and leg defects as seen for even the weakest Prat EMSinduced alleles (Clark 1994). The reason is probably that the knockdown of Prat expression in these tissues is not very efficient or that downstream pathway components can be obtained from other tissues where the knockdown is not very efficient. This inefficiency could be due to many factors, including the level of Prat or Prat2 expression, the activity of the promoter driving GAL4, or the efficiency of GAL4 in particular tissues. The fact that the knockdown is sufficient to produce pupal lethality and female sterility, roughly comparable to that seen for the EMS-induced alleles (Clark 1994; Malmanche and Clark 2004), favors the idea that the knockdown does not occur with the same efficiency in all tissues. The Prat/Prat2 duplication appears to have resulted from a retrotransposition event that occurred prior to the divergence of D. melanogaster and D. virilis. As we previously reported (Malmanche et al. 2003), Prat and Prat2 have different expression patterns and one of the major differences between them is that only Prat has maternal expression, as do the other genes in the de novo purine synthesis pathway (Tomancak et al. 2002). This difference between the two genes is reinforced by our previous observation that Prat is essential for female germline development, as revealed by analysis of EMSinduced mutations of Prat (Malmanche and Clark 2004) and by the Prat RNAi results in this report. Thus, it appears that Prat (the retrogene) has obtained a cisregulatory element for its maternal expression during or after the retrotransposition event and that Prat2 (the founder gene) has lost its maternal expression. At the same time, the founder gene Prat2 has been maintained during evolution since it is required for the prepupal stage while Prat has become important in the pupal stage. Metamorphosis consists of the prepupal and pupal stages where, at the end of the wandering third larval stage, a pulse of ecdysone triggers puparium formation and the prepupal stage begins. During this stage, the larval midgut and anterior muscles undergo histolysis and the imaginal discs begin to evert. However, the

abdominal larval muscles persist and are responsible for the contraction that causes the gas bubble translocation (Fristrom and Fristrom 1993). The gas bubble first appears in the center of the prepupa and then moves posteriorly before reaching its final anterior destination. The prepupal stage lasts for 12 hr and another ecdysone pulse triggers the prepupal–pupal transition. The consequences of this transition are that the salivary glands undergo histolysis and the imaginal head everts from the inside sac of the thorax driven by contraction of the abdominal muscles. The head will occupy the anterior space after the gas bubble is expelled through the anterior spiracles. This stage in development is affected in several Drosophila mutants. The cryptocephal phenotype was first described for mutations of the crc gene (Hadorn and Gloor 1943), where one sees a head eversion defect, which includes failed gas bubble translocation, incomplete leg disc elongation, and incomplete or no segmentation and differentiation of the abdomen. The crc gene encodes a CRE-binding-proteinlike transcription factor (Hewes et al. 2000). Mutations in the ecdysone-response genes such as E74 (Fletcher et al. 1995), DHR3 (Lam et al. 1999), krupple-homolog (Pecasse et al. 2000), and crol (D’Avino and Thummel 1998) also have a cryptocephal phenotype. This phenotype is also caused by mutations of the Mmp1 gene encoding matrix metalloproteinase (Page-McCaw et al. 2003). Prat2 RNAi also caused a cryptocephal phenotype. However, Prat2 differs from most of these genes in that they are also required for embryonic and/or larval development. Prat2 is essential only during metamorphosis as found for krupple-homolog (Pecasse et al. 2000). Fletcher et al. (1995) proposed that the premature histolysis of the abdominal muscle group is the reason for the head eversion defects. It would be interesting to investigate if Prat2 RNAi caused premature death of the abdominal muscle or if perhaps simply a reduced ATP level interfered with abdominal muscle contraction. Our results showed that both Prat and Prat2 are essential for metamorphosis, even though they both encode proteins that are 79% identical. One possible explanation for their lack of redundancy is that PRAT forms hetero-dimers and tetramers with the two different subunits from Prat and Prat2. Studies in other systems have shown that active PRAT is a dimer, while inactivated PRAT, bound with AMP, for example, forms a tetramer (Yamaokaet al. 2001). Since Prat is exclusively expressed in the female germline and Prat2 has an important role in prepupal development, we do not think that it is necessary for PRAT enzyme to function in the fly by forming a hetero-oligomer. Since double RNAi causes more severe prepupal and pupal lethal phenotypes than single Prat and Prat2 RNAi, and since this phenotype is similar to the mutant phenotype of the single-copy gene in the pathway, ade2, it appears that Prat and Prat2 together provide the purine nucleotides essential for metamorphosis.

Prat and Prat2 in Drosophila We thank Nicolas Malmanche for advice on the RT–PCR and Jay Penney and the anonymous reviewers for their comments on the manuscript. This work was supported by a grant to D.V.C. from the Canadian Institutes of Health Research, in partnership with the New Brunswick Medical Research Fund and the New Brunswick Innovation Foundation.

LITERATURE CITED Bainbridge, S. P., and M. Bownes, 1981 Staging the metamorphosis of Drosophila melanogaster. J. Embryol. Exp. Morphol. 66: 57–80. Brand, A. H., and N. Perrimon, 1993 Targeted gene expression as a means of altering cell fates and generating dominant phenotypes. Development 118: 401–415. Chadfield, G. C., and C. J. Sparrow, 1985 Pupation in Drosophila melanogaster and the effect of the lethal cryptocephal mutation. Dev. Genet. 5: 103–114. Clark, D. V., 1994 Molecular and genetic analyses of Drosophila Prat, which encodes the first enzyme of de novo purine biosynthesis. Genetics 136: 547–557. D’Avino, P. P., and C. S. Thummel, 1998 Crooked legs encodes a family of zinc finger proteins required for leg morphogenesis and ecdysone-regulated gene expression during Drosophila metamorphosis. Development 125: 1733–1745. Duffy, J. B., 2002 GAL4 system in Drosophila: a fly geneticist’s Swiss army knife. Genesis 34: 1–15. Fletcher, J. C., K. C. Burtis, D. S. Hogness and C. S. Thummel, 1995 The Drosophila E74 gene is required for metamorphosis and plays a role in the polytene chromosome puffing response to ecdysone. Development 121: 1455–1465. FlyBase, 2003 The FlyBase database of the Drosophila genome projects and community literature. Nucleic Acids Res. 31: 172–175. Foley, K. P., M. W. Leonard and J. D. Engel, 1993 Quantitation of RNA using the polymerase chain reaction. Trends Genet. 9: 380–385. Fortier, E., and J. M. Belote, 2000 Temperature-dependent gene silencing by an expressed inverted repeat in Drosophila. Genesis 26: 240–244. Fristrom, D., and W. J. Fristrom, 1993 The metamorphic development of the adult epidermis, pp. 866–867 in The Development of Drosophila melanogaster, edited by M. Bate and A. M. Arias. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Giordano, E., R. Rendina, I. Peluso and M. Furia, 2002 RNAi triggered by symmetrically transcribed transgenes in Drosophila melanogaster. Genetics 160: 637–648. Hadorn, E., and H. Gloor, 1943 Cryptocephal ein spat wirkender Letalfaktor bei Drosophila melanogaster. Rev. Suisse Zool. 50: 256–261. Henikoff, S., and M. Eghtedarzadeh, 1987 Conserved arrangement of nested genes at the Drosophila Gart locus. Genetics 117: 711–725. Hewes, R. S., A. M. Schaefer and P. H. Taghert, 2000 The cryptocephal gene (ATF4) encodes multiple basic-leucine zipper proteins controlling molting and metamorphosis in Drosophila. Genetics 155: 1711–1723. Karess, R. E., and G. M. Rubin, 1984 Analysis of P transposable element functions in Drosophila. Cell 38: 135–146. Kennerdell, J. R., S. Yamaguchi and R. W. Carthew, 2002 RNAi is activated during Drosophila oocyte maturation in a manner dependent on aubergine and spindle-E. Genes Dev. 16: 1884–1889. Kohler, M., B. Assmann, C. Brautigam, W. Storm, S. Marie et al., 1999 Adenylosuccinase deficiency: possibly underdiagnosed encephalopathy with variable clinical features. Eur. J. Paediatr. Neurol. 3: 3–6. Lam, G., B. L. Hall, M. Bender and C. S. Thummel, 1999 DHR3 is required for the prepupal-pupal transition and differentiation of adult structures during Drosophila metamorphosis. Dev. Biol. 212: 204–216. Lee, T., and L. Luo, 1999 Mosaic analysis with a repressible cell marker for studies of gene function in neuronal morphogenesis. Neuron 22: 451–461. Malmanche, N., and D. V. Clark, 2003 Identification of transdominant modifiers of Prat expression in Drosophila melanogaster. Genetics 164: 1419–1433.

1631

Malmanche, N., and D. V. Clark, 2004 Drosophila melanogaster Prat, a purine de novo synthesis gene, has a pleiotropic maternal effect phenotype. Genetics 168: 2011–2023. Malmanche, N., D. Drapeau, P. Cafferty, Y. Ji and D. V. Clark, 2003 The PRAT purine synthesis gene duplication in Drosophila melanogaster and Drosophila virilis is associated with a retrotransposition event and diversification of expression patterns. J. Mol. Evol. 56: 630–642. Marie, S., B. Heron, P. Bitoun, T. Timmerman, G. Van Den Berghe et al., 2004 AICA-ribosiduria: a novel, neurologically devastating inborn error of purine biosynthesis caused by mutation of ATIC. Am. J. Hum. Genet. 74: 1276–1281. Negeri, D., H. Eggert, R. Gienapp and H. Saumweber, 2002 Inducible RNA interference uncovers the Drosophila protein Bx42 as an essential nuclear cofactor involved in Notch signal transduction. Mech. Dev. 117: 151–162. O’Donnell, A. F., S. Tiong, D. Nash and D. V. Clark, 2000 The Drosophila melanogaster ade5 gene encodes a bifunctional enzyme for two steps in the de novo purine synthesis pathway. Genetics 154: 1239–1253. Page-McCaw, A., J. Serano, J. M. Sante and G. M. Rubin, 2003 Drosophila matrix metalloproteinases are required for tissue remodeling, but not embryonic development. Dev. Cell 4: 95–106. Pecasse, F., Y. Beck, C. Ruiz and G. Richards, 2000 Kruppel-homolog, a stage-specific modulator of the prepupal ecdysone response, is essential for Drosophila metamorphosis. Dev. Biol. 221: 53–67. Piccin, A., A. Salameh, C. Benna, F. Sandrelli, G. Mazzotta et al., 2001 Efficient and heritable functional knock-out of an adult phenotype in Drosophila using a GAL4-driven hairpin RNA incorporating a heterologous spacer. Nucleic Acids Res. 29: E55. Roignant, J. Y., C. Carre, B. Mugat, D. Szymczak, J. A. Lepesant et al., 2003 Absence of transitive and systemic pathways allows cell-specific and isoform-specific RNAi in Drosophila. RNA 9: 299–308. Rubin, G. M., L. Hong, P. Brokstein, M. Evans-Holm, E. Frise et al., 2000 A Drosophila complementary DNA resource. Science 287: 2222–2224. Russo, C. A. M., N. Takezaki and M. Nei, 1995 Molecular phylogeny and divergence times of Drosophilid species. Mol. Biol. Evol. 12: 391–404. Spradling, A. C., 1986 P element-mediated transformation, pp. 175–197 in Drosophila: A Practical Approach, edited by D. B. Roberts. IRL Press, Oxford. Stone, R. L., J. Aimi, B. A. Barshop, J. Jaeken, G. Van den Berghe et al., 1992 A mutation in adenylosuccinate lyase associated with mental retardation and autistic features. Nat. Genet. 1: 59–63. Tiong, S. Y. K., and D. Nash, 1990 Genetic analysis of the adenosine3 (Gart) region of the second chromosome of Drosophila melanogaster. Genetics 124: 889–897. Tiong, S. Y. K., and D. Nash, 1993 The adenosine2 gene of Drosophila melanogaster encodes a formylglycinamide ribotide amidotransferase. Genome 36: 924–934. Tiong, S. Y. K., C. Keizer, D. Nash and D. Patterson, 1989 Drosophila purine auxotrophy: new alleles of adenosine2 exhibiting a complex visible phenotype. Biochem. Genet. 27: 333–348. Tomancak, P., A. Beaton, R. Weiszmann, E. Kwan, S. Shu et al., 2002 Systematic determination of patterns of gene expression during Drosophila embryogenesis. Genome Biol. 3: RESEARCH0088. Vaistij, F. E., L. Jones and D. C. Baulcombe, 2002 Spreading of RNA targeting and DNA methylation in RNA silencing requires transcription of the target gene and a putative RNA-dependent RNA polymerase. Plant Cell 14: 857–867. Yamaoka, T., M. Yano, M. Kondo, H. Sasaki, S. Hino et al., 2001 Feedback inhibition of amidophosphoribosyltransferase regulates the rate of cell growth via purine nucleotide, DNA, and protein syntheses. J. Biol. Chem. 276: 21285–21291. Zalkin, H. Z., and J. E. Dixon, 1992 De novo purine nucleotide biosynthesis. Prog. Nucleic Acid Res. Mol. Biol. 42: 259–287. Zhu, X., and D. Stein, 2004 RNAi-mediated inhibition of gene function in the follicle cell layer of the Drosophila ovary. Genesis 40: 101. Communicating editor: T. H. Eickbush