overexpression of hmgd causes the failure of pupariation in drosophila by affecting ecdysone...

Archives of Insect Biochemistry and Physiology 68:123–133 (2008)

© 2008 Wiley-Liss, Inc.DOI: 10.1002/arch.20237Published online in Wiley InterScience (www.interscience.wiley.com)

Overexpression of HmgD Causes the Failure ofPupariation in Drosophila by Affecting EcdysoneReceptor Pathway

Jing Chen, Hui Wang, and Yu-Feng Wang*

HmgD encodes Drosophila homologue of high mobility group proteins (HMGD), which are thought to have an architecturalfunction in chromatin organization. However, current opinions about the function of HMGD in Drosophila development arecontroversial. Our previous studies have shown that ubiquitous overexpression of HmgD caused the formation of melanotictumors in the Drosophila larvae by prematurely activating the Ras-MAPK pathway. Here we report that under maternal control,the viability of flies links with overexpression of HmgD, while under ubiquitous control, ActGal4, overexpressing HmgD ani-mals, which display prolonged larval stages around day 13, developmentally stagnate in the larva-white pupa transition.Ecdysone feeding did not rescue overexpressing HmgD animals. RT-PCR analyses show that overexpression of HmgD does notaffect the temporal expression pattern of ecdysone receptor gene EcR, whereas transcriptional patterns of some key regulatorygenes, such as E74A, E74B, E75A, E75B, βFTZ-F1, are changed greatly. These results suggest that ubiquitous overexpressionof HmgD results in the failure of pupariation neither by affecting the process of ecdysone synthesis and release nor byabnormal EcR transcription, but by causing expression of EcR regulatory nuclear receptors out of schedule. The results led us topostulate that overexpression of HMGD likely changes the signaling cascade of Drosophila metamorphosis by an interactionbetween HMGD and DNA strands, and subsequently by an error of DNA binding abilities and transcriptional activities of somenuclear receptor genes. Arch. Insect Biochem. Physiol. 68:123–133, 2008. © 2008 Wiley-Liss, Inc.

KEYWORDS: HmgD; overexpression; Drosophila; pupariation; ecdysone receptor pathway

Hubei Key Laboratory of Genetic Regulation and Integrative Biology, College of Life Sciences, Central China Normal University, Wuhan, P. R. China

Contract grant sponsor: National Natural Science Foundation of China; Contract grant number: 30300035; Contract grant sponsor: Scientific Research Foundationfor the Returned Overseas Chinese Scholars, State Education Ministry; Contract grant number: (2004) 527.

*Correspondence to: Yu-Feng Wang, College of Life Sciences, Central China Normal University, 152 Luoyu Avenue, Wuhan 430079 P. R. China.E-mail: [email protected]

Received 5 September 2007; Accepted 4 January 2008

INTRODUCTION

The gene HmgD encodes the Drosophila homo-logue of the high mobility group proteins, HMGD,which is closely related to the vertebrate HMG-Box(HMGB) proteins (formerly termed HMG-1/2).They are relatively abundant proteins that bind toDNA and bend DNA substantially, acting prima-rily as architectural facilitators in the assembly ofnuclear protein complexes (Thomas and Travers,2001). Circumstantial evidence suggests that HMGDreduces the compactness of chromatin packing dur-ing very early development of Drosophila when his-

tone H1 is absent. This loose structure could facili-tate the rapid condensation and de-condensationof chromatin required during the very short earlynuclear division cycles. As the maternal pool ofHMGD is titrated by rapid replication cycles andconcomitant chromatin assembly, and as histoneH1 undergoes synthesis, HMGD would function-ally be suppressed by histone H1 around the mid-blastula transition (MBT) (Ner and Travers, 1994).However, comparison of the DNA-binding capac-ity and specificity in recognition of DNA clearlyshows that HMGD does not share the same bind-ing properties with histone H1. Moreover, HMGD

124 Chen et al.

Archives of Insect Biochemistry and Physiology July 2008

is expressed not only in adult females and earlyembryos but also in later embryonic developmen-tal stages, suggesting that HMGD may exert otherfunctions than that of a specific early embryonicsubstitute of H1 (Renner et al., 2000).

In Xenopus, both linker histone B4 and HMGB1protein show the same properties as HMGD dur-ing early embryogenesis (Dimitrov et al., 1993) andin the assembly system in vitro (Nightingale et al.,1996). However, HMGB1 protein does not associ-ate with chromatin in cells during mitosis in earlymouse embryos, even when the level of histoneH1 is very low (Spada et al., 1998). In mice, asingle mutant in Hmgb1 or Hmgb2 did not resultin embryonic defects, except that a Hmgb1 mutantpup died within 24 h due to hypoglycemia andthat Hmgb2 knockout mice decreased male fertil-ity (Calogero et al., 1999; Ronfani et al., 2001).Hmgb3-deficient mice are viable but erythro-cythemic (Nemeth et al., 2003, 2005). Recently,Ragab et al. (2006) showed that homozygous mu-tant HmgD/Z is viable and exhibits only minor mor-phological defects. Taken together, the function ofHMGB in development still remains uncertain.

In order to further analyse the function of HMGDon the development of Drosophila, we generatedpUASpHmgD transgenic flies and overexpressedHmgD under both maternal control and ubiquitouscontrol (ActGal4), respectively. We found that undermaternal control some effects on the viability of em-bryos appeared to link with the overexpression ofHmgD. While overexpression of HmgD under ActGal4driver prematurely activated Ras-MAPK pathway andconsequently led to the formation of melanotic tu-mors in larvae (Chen et al., 2006). Here, we reportthat ubiquitous overexpression of HmgD led to dis-ruption of the temporal expression pattern of someecdysone receptor signaling pathway members, thusresulting in the failure of Drosophila pupariation.

MATERIALS AND METHODS

Fly Stocks and Their Propagation

Drosophila melanogaster used in our lab was wild-type strain Canton S, obtained from Central China

Normal University. All transgenic flies were con-structed in the Curie Institute (Paris, France). Lar-vae and flies were cultured on standard yeast/glucose medium.

Plasmid Construction

The coding region of HmgD gene (GeneBankAccession No. NM166480) was obtained by PCRamplification from a cDNA library of Drosophila0–4-h embryos (Clontech). The primers contain-ing Asp 718 and BamH I restriction sites wereHmgD5 (5′-ggggtacccatATGTCTGATAAGCCAAAA-CGCC-3′) and HmgD3 (5′-cgggatcccgCTACTCGCT-CTCATCATCGTC-3′). After digestion by enzymesAsp 718 and BamH I, PCR-amplified fragments wereinserted into pUASp plasmids (Rorth, 1998; Rorthet al., 1998), and sequenced from both sides toconfirm the transgenes (data not shown).

Generation of Transgenic Fly Strains and Genetics

Progeny embryos of W– fly (white eyes) werecollected at 25°C. Embryonic microinjections ofconstructs to generate transgenic lines were per-formed as described in Bellaiche et al. (1996).Progeny with red eyes were selected for balancingand mapping insertions. CyO and Tm2 are balanc-ers for chromosome II and chromosome III, re-spectively. Sp and Sb are markers of chromosomeII and chromosome III, respectively. nanos-Gal4,NGT40 (Bloomington Stock Center), and ActGal4/CyO-PscGFP (generated from ActGal4/CyO flies bya genetic method) females were used to cross withpUASp transgenic males for overexpressing HmgD.Transgenic flies were confirmed by PCR with prim-ers harboring two ends of the cloning sites of thepUASp vector (Chen et al., 2006).

Selection of Mutated Larvae

Virgin females of ActGal4/Cyo-PscGFP, whichoverexpress the target gene under the control of ac-tin ubiquitous promoter ActGal4 when UAS (Up-stream Activation Sequence) is present, were collectedand crossed with transgenic males pUASpHmgD A;

Chromatin Structure in Drosophila Development 125


pUASpHmgD E carrying four copies of pUASpHmgD.The eggs laid within 3–4 h were incubated at 25°Cfor 24 h until all larvae hatched out. The first instaroverexpressing HmgD (pUASpHmgD A/ActGal4;pUASpHmgD E/+) larvae were selected under the fluo-rescence stereomicroscope (Leica MZ16 F Germany)from their lack of GFP fluorescence. The wild-typelarvae (Canton S) were used as control.

Ecdysone Feeding Experiments

20-hydroxyecdysone (20E, Sigma, St. Louis,MO) was prepared to 12 mg/ml stock solution withethanol first. This stock solution (4.2 ml) was di-luted in 95.8 ml sterile water and added to 0.05 gdry yeast to make the yeast paste containing 0.5mg/ml 20E. As a negative control, 4.2 ml ethanolwas diluted in 95.8 ml sterile water and added to0.05 g dry yeast (Gates et al., 2004). One hundredmilliliters of sterile water containing 0.05 g dryyeast was used as normal yeast medium.

First instar larvae were allowed to develop for50 h after larvae hatching (ALH), a time point cor-responding to the early third instar in wild typeanimals. HmgD overexpressing larvae were dividedinto two groups. One group was transferred to the20E containing yeast paste, the other was trans-ferred to the ethanol containing paste as control.The same aged wild type larvae were cultured innormal yeast medium. As positive controls, em-bryos from the cross E75A81/ TM6B Tb Ubi-GFP×E75∆51/TM6B Tb Ubi-GFP (gift from C.S.Thummel) were collected for 3 to 4 h. The em-bryos were maintained at 25°C and allowed tohatch. E75A81/E75∆51 first instar larvae were selectedby their lack of GFP expression. The mid-secondinstar larvae were transferred to food with or with-out 20E for 6 h and then returned to regular yeastpaste.

These larvae were fed until all wild type larvaedeveloped to pupae. Developmental stages wereobserved and freshly corresponding yeast pasteswere provided three or four times a day. Aroundthe time of pupariation, the larvae were observedevery 15 min.

Staging of HmgD Overexpressing Larvae andWhite Pupae

The wandering third instar larvae were collectedaround 66 h ALH (Andres et al.,1993) and stagedat approximately 14–22 h relative to puparium for-mation (–16 h in Fig. 3, taking 2 h difference intoaccount during egg laying). Four hours later, thelarvae were sampled as –10 h in figures. Stationarylarvae were estimated to be –3 h. Newly formedwhite pupae were marked as 0 h relative to pu-parium formation and for future collection at vari-ous time points (Karim et al., 1993). The HmgDoverexpressing larvae were staged by synchroniz-ing wild-type animals at puparium formation. Tento fifteen samples were collected per time point.

RT-PCR Analysis for Ecdysone Signal Pathway

Total RNAs were isolated using the SV total RNAIsolation System (Promega) and were reverselytranscribed with M-MLV (Promega, Madison, WI)with oligo(dT) primers (Tiangen, Beijing, China ).rp49 was used to adjust the concentration of re-verse-transcribed first-strand cDNAs. Well-adjustedcDNAs were used as templates to amplify specificfragments with gene-specific primers (Table 1).Amplification conditions for PCR reactions were:94°C 30 s; 50–61 (based on various primers), 45 s;72°C, 1 min for 30 cycles.

TABLE 1. Primers Used for RT-PCRGenes Primer names SequencesEcR EcR-Fwd 5′-TTTCGGCATATAACCGAGATAAC-3′

EcR-Rev 5′-CGTATATGATCTATTATTCGCG-3′E74A E74A-Fwd 5′-GCCGGCGACTGTGCCACCAAGCTG-3′

E74A-Rev 5′-CGGAGTGAGCAGCGTATTTTGCTC-3′E74B E74B-Fwd 5′-GCGGACGTGGAGGCAGACGCCGAG-3′

E74B-Rev 5′-CGGAGTGAGCAGCGTATTTTGCTC-3′E75A E75A-Fwd 5′-GCCAACACGGGGAGGAACACGCCA-3′

E75A-Rev 5′-CAGAACGATGGCGCAGAACAGGCC-3′E75B E75B-Fwd 5′-ACAGTGCTGTGCCGCGTTTGCGGG-3′

E75B-Rev 5′-CAGAACGATGGCGCAGAACAGGCC-3′BR-C BRC-Fwd 5′-GCTCGAGCGGGCGCCGATGATG-3′

BRC-Rev 5′-AGGGGACACGGCGAGGTCACCG-3′Edg84A Edg84A-Fwd 5′-GATGATGTCCGTGGATTCAACGCCG-3′

Edg84A-Rev 5′-GGCGTGGTGCACGGTAGTCTTCAG-3′âFTZ-F1 FTZ-Fwd 5′-TACCAGGACACCACCTCCTCACAC-3′

FTZ-Rev 5′-AAGTCGATCACCTCGCCGCCGTG-3′E93 E93-Fwd 5′-AACGGTAGCAATGGCAACGGGC-3′

E93-Rev 5′-CTTGTTGTGCGCCAAATCGCTGT-3′

126 Chen et al.


Western Blot Analysis

Virgin females of ActGal4/Cyo-PscGFP werecrossed with two fly lines carrying two and four cop-ies of transgene, respectively. HmgD overexpressingfirst instar larvae were selected by their lack of GFPfluorescence. Wild-type larvae were collected as con-trols. The larvae were allowed to develop to mid-third-instar for extraction of proteins. The proteinswere separated by SDS-PAGE. HMGD levels weredetected by an antibody raised against the HMGBdomain of HMGD (rat anti-HMGD-100, 1:400).BCIP/NBT staining was performed subsequently.

Statistics Analysis

Assays were repeated three times. Data were ex-pressed as mean ± SE and the statistical signifi-cance of the results was assessed by one-wayanalysis of variance (ANOVA).

RESULTS

The Effect of Overexpressing HmgD UnderMaternal Control

Dominant phenotypic marker (CyO, curly wingslocated on chromosome II), were used to clearlycharacterize the effect of overexpressing HmgD inheterozygous transgenic flies carrying the insertionof HmgD on chromosome II. These heterozygoustransgenic flies were crossed with NGT40 femalesand the percentages of the flies with straight wings

(non Cyo, with Gal4) in the first generation fromthese crosses (F1) was investigated. Our results(Table 2) showed that percentages of straight wingflies in F1 were significantly lower than that of con-trols (P < 0.01) except D line flies, indicating thatoverexpressing HmgD decreases viability of flies.

Overexpression of HmgD Under the Ubiquitous Control(Act-Gal4) Caused Defects in Transition From ThirdInstar Larvae to White Pupae

Virgin females of ActGal4/Cyo-PscGFP, express-ing Gal4 under control of actin ubiquitous pro-moter, were crossed with transgenic males toinvestigate the effects of overexpressing HmgD onthe development of Drosophila. Western blottingshowed that compared with wild type, HMGD pro-tein was overexpressed under the control of ActGal4,and with the increase of transgenic pUASpHmgDcopies, the HMGD level also increased (Fig. 1).When males carrying four copies of pUASpHmgDwere crossed with females of ActGal4/Cyo-PscGFP,all progeny had curly (CyO) wings. This indicatesthat the presence of two copies of pUASpHmgDtransgenes and ActGal4 together in the same fly in-duced 100% lethality, whereas their siblings (curlywings), which did not overexpress HmgD due to alack of ActGal4, developed normally.

To investigate at which stage these flies died,virgin females of ActGal4/Cyo-PscGFP were crossedwith the transgenic males of pUASpHmgD A;pUASpHmgD E. The eggs laid within 4 h were in-cubated at 25°C for 24 h, while wild type eggs were

TABLE 2. Decreased Survival Rate Induced in Adults by HmgDExpression Under Maternal Control*

Maternal genotype: NGT40 (25°C)

Flies with straight Number ofPaternal genotype wings (%) ± SE flies analysed

[+; +]/Cyo-Tm9:Sb 69.7 ± 0.9 166pUASpHmgD A /Cyo; Sb/Tm2:ry 59.3 ± 1.2a 239[+;pUASpHmgD B]/Cyo-Tm9:Sb 58.0 ± 0.5a 105[+;pUASpHmgD C]/ Cyo-Tm9:Sb 62.4 ± 1.2a 156[+;pUASpHmgD D]/Cyo-Tm9:Sb 65.8 ± 1.5 149[Sp; pUASpHmgD E]/Cyo-Tm9:Sb 56.9 ± 0.6a 168[pUASpHmgD A; pUASpHmgD E]/Cyo-Tm9:Sb 58.0 ± 1.2a 238

*A, B, C, D, E (in roman, not italics): different transgenic lines. Data are expressedas mean ± SE and the statistical significance of the results was assessed using aone-way analysis of variance (ANOVA).aThe difference is significant, P < 0.01.

Fig. 1. Overexpression of HMGD was confirmed by West-ern blotting. A very weak signal was detectable in the 3rdinstar larvae of wild type (Lane 1). More protein was in-duced in the 3rd instar larvae by ActGal4 with transgenic 1copy of pUASpHmgD (Lane 2) and 2 copies of pUASpHmgD(Lane 3).



collected and incubated in the same way as thecontrol. The HmgD overexpressing first instar lar-vae were selected under the fluorescence stereomi-croscope by the lack of GFP fluorescence. All larvaewere allowed to develop in standard medium. Thenumber of living larvae and their pupariation sta-tus was assessed every day. We found that HmgDoverexpressing larvae developed in the same wayas wild type at the first and the second larval stages.However, when wild type third instar larvae beganto pupate, HmgD overexpressing larvae crawled tothe walls of vials, while none of them pupariatednormally. Most HmgD overexpressing animals didnot enter the normal pupal stage and died in afew days (Table 3). Very few malformed “pupae”

formed in 5 days after larval hatching. Some ofthese “pupae” were tanned a little, but most bodyparts resembled larvae (Fig. 2A). Abnormal bodiesdid not shorten like wild type animals at the be-ginning of pupariation (Fig. 2B and C). Althoughsome HmgD overexpressing larvae displayed nor-mally tanned pupae, the heads still connected withlarval spicules and did not evert at all, and the lar-val mouth hooks were not ejected (Fig. 2D–F).

Ecdysone Feeding Did Not Rescue HmgDOverexpressing Animals

The pupariation defects in HmgD overexpressinganimals could result from either a decrease in

TABLE 3. The Larval Development of Drosophila Overexpressing HmgD Under the Control of Act-Gal4 Driver*

Number of Survival rate of Survival rate of Percentage of Percentage of Survival rate ofFly line L1 (d1) L2 (d2) (%) L3 (d3) (%) pupae (d4) (%) pupae (d5) (%) adults (d10) (%)

Wild type 120 84.2 67.5 26.7 52.5 52.5pUASpHmgD A /ActGal4::::: pUASpHmgD E/+ 85 81.2 61.2 0 5.9a 0

*A, E (in roman): different transgenic lines.aThe percentage of malformed pupae. L1, L2, and L3: 1st, 2nd, and the 3rd instar larvae. Perhaps because these larvae were washed out from themedium every day for counting, the survival rates of the 2st and the 3rd instar larvae were a little lower than normal.

Fig. 2. Phenotypes of HmgD overexpressing animals.Most of the HmgD overexpressing animals died, showingold larval morphology (A). Few of them died just at thebeginning of metamorphosis as white pupae (B,C). Few

other HmgD overexpressiing animals could form tannedpuparia, but larvae metamorphosis failed (D,E). F: Wildtype pupa. [Color figure can be viewed in the online edi-tion which is available at www.interscience.wiley.com]

128 Chen et al.


ecdysone titer or a decrease in the ability of theecdysone signal to be transduced. To distinguishthese two possibilities, we examined the effects offeeding ecdysone to HmgD overexpressing larvae,which has shown to effectively rescue phenotypesassociated with ecdysone-deficient mutations inDrosophila (Venkatesh and Hasan, 1997; Freemanet al., 1999; Bialecki et al., 2002). Considering thatmost HmgD overexpressing larvae developed to thethird instar larval stage and their size and vitalitydid not show a significant difference from thoseof wild type, early third instar HmgD overexpressinglarvae were used to perform ecdysone feeding ex-periments. When all synchronizingly developedwild type larvae metamorphosed to pupae, no nor-mal HmgD overexpressing pupae were observed.All HmgD overexpressing larvae died in the sameway as described previously (Fig. 2). Feedingecdysone to E75A mutant second instar larvae,however, had a dramatic effect on their develop-ment, rescuing 50% (n = 20) of them to pupalstage, while almost all of E75A mutant secondinstar larvae that were maintained on food with-out 20E failed to molt and develop to later stages.This result is consistent with previous work(Bialecki et al., 2002). Our results demonstratethat ecdysone is not a limiting factor in HmgDoverexpressing animals.

Ubiquitous Overexpression of HmgD Disrupts theTemporal Expression Pattern of Ecdysone ReceptorSignaling Pathway

Since HmgD overexpressing animals displayedserious failure in pupariation and ecdysone feed-ing could not rescue HmgD overexpressing animals,it logically increased possibilities that there couldbe something wrong with the ecdysone signalingpathway (Koelle et al., 1991; Yao et al., 1992, 1993).The EcR-USP complex directly induces the transcrip-tion of primary-response genes, including earlygenes originally defined as ecdysone-inducible puffsin the larval salivary gland polytene chromosomes(Ashburner et al., 1974). Our RT-PCR analysesshowed that the EcR gene transcription pattern inHmgD overexpressing animals was virtually indis-

tinguishable from that in wild-type animals, whenrp49 was used as a quantitative control (Fig. 3).

Three well-characterized early genes Broad-Com-plex (BR-C), E74, and E75 encode transcription fac-tors that transduce and amplify hormonal signals(Dubrovsky, 2005). E74 encodes two isoforms ofan ETS domain transcription factor, designatedE74A and E74B (Burtis et al., 1990). E75 encodesthree orphan members of the nuclear receptor su-perfamily, designated E75A, E75B, and E75C(Segraves and Hogness, 1990). Their transcriptionpatterns around metamorphosis were tested inboth HmgD overexpressing animals and wild typeanimals. Although the timing and levels of earlygene BR-C expression in HmgD overexpressing ani-mals were similar to that of wild type animals (datanot shown), the temporal expression profiles ofearly genes E74 and E75 were altered significantlyin HmgD overexpressing animals compared withwild type (Fig. 3). In wild type animals, E74A andE75A were induced at the beginning of the whitepupa stage, and then repressed but still detectable,and subsequently accumulated at the white pupa–pupa transition around 8–10 h after pupariation(Fig. 3). In contrast, expression of these two geneswas variable in HmgD overexpressing animals.Neither E74A nor E75A transcription was detect-able at the time corresponding to wild-type larvapupariation. The expression patterns of these twogenes were compressed to only three time points,corresponding to 2, 6, and 10 h after wild-typepupariation (Fig. 3). In HmgD overexpressing ani-mals, E74B was prematurely transcribed to thetime of puparium formation of wild type, whereasE75B transcription was severely postponed frommid-third-instar larval stage to the time point cor-responding to white pupa and pupa stages ofwild-type (Fig. 3).

The early gene products activate a large groupof late genes, which are directly or indirectly in-volved in the metamorphic process (Thummel,2002). We also examined expression patterns oftwo late downstream genes Edg84A and βFTZ-F1,which encode key transcription factors in ecdys-one pathway cascades. Edg84A is expressed prima-rily in late white pupae and encodes pupa cuticle



Fig. 3. RT-PCR analysis of genes encoding key regula-tory factors in the ecdysone cascades during late larvaland white pupa development in wild type and HmgD over-

expressing animals, respectively. rp49 was used as a refer-ence gene.

130 Chen et al.


proteins that contribute to the synthesis of pupaepidermis (Fechtel et al., 1988, 1989). The over-expression of HmgD severely affected expressionof Edg84A. In wild type, Edg84A transcription wasfirst detected at the beginning of pupariation andthe level was gradually increased and then keptstable at the white pupa–pupa transition. However,in HmgD overexpressing animals, the mRNA levelof Edg84A was dramatically repressed until the timepoint relative to 10 h after puparium formation inwild type (Fig. 3). βFTZ-F1 also encodes an orphanmember of the nuclear hormone receptor superfam-ily, which is a critical regulator of white pupa–pupatransition in ecdysone cascades (Woodard et al.,1994; Broadus et al., 1999). In wild type, βFTZ-F1was expressed in a brief interval in the middle ofthe white pupa stage (6 h in Fig. 3) (Lam et al.,1999), while in HmgD overexpressing animals, itwas prematurely transcribed in the mid-third in-star larval stage; then the transcript could not bedetected at the time of the late third instar and theearly white pupa stage relative to wild type. Onlyat the time corresponding to the wild-type middlepupa stage (6 h) was a very weak expression de-tected (Fig. 3).

In addition, the stage-specific E93 early gene wasalso tested in our study. E93 is induced directly byecdysone in late white pupa salivary glands butshows no response to the signal several hours ear-lier, in a late-third-instar larva (Baehrecke andThummel, 1995). In HmgD overexpressing animals,E93 was expressed throughout the time relative tothe wild-type white pupa stage (data not shown),while the transcript was only detected at the latewhite pupa stage in wild type (Lam et al., 1999).

DISCUSSION

Ner and Travers (1994) showed that HMGD isassociated with condensed chromatin structuresduring the first six rapid nuclear cleavage cycles ofthe developing embryos. At that time, histone H1is absent from these structures. With the accumu-lation of H1 from the seventh nuclear division on-wards, the nuclei become more compact. Thiscompaction is paralleled by a reduction in the size

of mitotic chromatin, which implies that the con-densed state of chromatin induced by HMGD maybe less compact than the H1-containing standardchromatin fiber and that this state of chromatincould facilitate rapid nuclear cycles. An in vitrochromatin assembly system derived from pre-blas-toderm Drosophila embryos showed that incorpo-ration of purified HMGD into chromatin in amanner similar to histone H1 alters the nucleo-some repeat length, indicative of interaction withthe nucleosome linker DNA. Furthermore, histoneH1 can displace HMGD protein from chromatinin a concentration-dependent manner (Ner et al.,2001). Our results indicate that the overexpressionof the HmgD under maternal control influences theviability of the embryos. However, these effects areweak, since it does not result in 100% lethality.Overexpression of HmgD under the control of aubiquitously expressed Gal4 protein caused severelethality at the late larval stage, from which we in-fer that overexpression of HmgD influences the laterlarval stage rather than embryogenesis. Althoughin Xenopus B4 and HMGB1 could substitute for hi-stone H1 during early embryogenesis, single mu-tation in Hmgb1, Hmgb2, or Hmgb3 in mice did notresult in lethality during embryogenesis (Calogeroet al. 1999; Ronfani et al. 2001; Nemeth et al. 2003,2005). Recently, Ragab et al. (2006) showed thatthe homozygous mutant HmgD/Z is viable anddoes not exhibit severe defects in Drosophila. There-fore, HMGB protein may not be essential for theoverall organization of chromatin, but is criticalfor proper transcriptional control by specific tran-scription factors.

Ubiquitous overexpression of HmgD causedpupariation failure in Drosophila. Pupariation istightly controlled and initiated by 20E. The late-third-instar pulse of 20E is propagated through agenetic regulatory hierarchy, which defines the on-set of metamorphic transition. 20E first binds to aheterodimer receptor comprising the ecdysone re-ceptor, EcR, and the gene product of usp (Koelle etal., 1991), and subsequently activates a small groupof early genes (Dubrovsky, 2005). The products ofearly genes then activate transcription of more than100 “late genes” that control various aspects of



metamorphosis such as degeneration of larval tis-sues, differentiation of imaginal discs, and pupalcuticle production. Three well-characterized earlygenes are BR-C, E74 and E75, each of which en-codes a set of protein isoforms of DNA bindingtranscriptional regulators (Thummel, 2001). Ourfeeding experiments confirm that the defect ofpupariation in overexpression of HmgD mutantsis not due to the reduction of 20E levels. RT-PCRanalysis indicates that it does not result from de-fects of EcR transcription, either. Among earlygenes, the expression of both E74 and E75 is se-verely affected, while the transcription of BR-C doesnot change in comparison with wild type. BR-C isa key regulator in the initiation of, and progres-sion through, metamorphosis. BR-C null mutantsdie after a prolonged third instar. Misexpressionof BR-C during the second larval instar redirectsepidermal cells from larval to pupal cuticle pro-duction (Mugat et al., 2000; Gonzy et al., 2002;Zhou et al., 2004). A similar expression pattern ofBR-C in HmgD overexpressing animals and wildtype indicates that overexpression of HmgD doesnot affect the ecdysone signaling pathway at theBR-C transcription level. E74A plays a role in pu-parium formation (Fletcher and Thummel, 1995).E74B may play a role in prolonging larval musclesin the pupal stage until they are required for a suc-cessful head eversion and body shortening duringpuparium formation and in ecdysone induction ofimaginal disc evagination (Fletcher and Thummel,1995). In HmgD overexpressing animals, failure inpupariation, head eversion, and larval body short-ening were observed, which may be involved in dis-ruption of the expression pattern of E74 gene. E75gene encodes three protein isoforms designatedE75A, E75B, and E75C (Segraves and Hogness,1990). E75A mutation resulted in developmentaldelay and molting defects, suggesting that E75Amay function in amplifying or maintaining theecdysteroid titer during larval development so asto ensure proper temporal progression through thelife cycle (Bialecki et al., 2002). E75B may func-tion by targeting gene regulation (White et al.,1997). Interestingly, the transcription of E74A andE75A displayed similar temporal shifts in HmgD

overexpressing animals, suggesting that this muta-tion influences expression of early genes in theecdysone signaling pathway.

Edg84A, which encodes components of the pu-pal cuticle, was severely repressed in HmgD over-expressing animals. Our results show that expressionof Edg84A was very weak during the white pupastage, and was increased only at the time point cor-responding to 10 h after wild-type puparium for-mation. Actually, most overexpressing HmgDanimals died at very late third instar stage with noor little tanning. We infer that expression of thisgene must be repressed intensively. However, βFTZ-F1 was prematurely expressed in the mid-third in-star larval stage and reduced in the middle whitepupa stage in HmgD overexpressing larvae com-pared with wild type. Although temporal expres-sion patterns of both E75B and βFTZ-F1 arechanged in HmgD overexpressing animals, theirshifted expression still agrees with the point thatE75B can act as a repressor of the βFTZ-F1 compe-tence factor during metamorphosis (White et al.,1997). The altered temporal expression patterns ofearly and later genes during larval and white pupastages in HmgD overexpressing animals suggest thatoptimum expression of HmgD is required for theecdysone receptor pathway.

ACKNOWLEDGMENTS

We thank Dr. Nathalie Dostatni of the CurieInstitute for helping us to establish the transgenicflies and for some suggestions. We also thank Dr.Andrew A. Travers of the MRC Laboratory of Mo-lecular Biology, Cambridge, for kindly providingthe HMGD antibody, Dr. Carl S. Thummel of theUniversity of Utah School of Medicine, for kindlyproviding the flies of E75A81/TM6B Tb Ubi-GFPand E75∆51/TM6B Tb Ubi-GFP, and Dr. Inaki Iturbe-Ormaetxe of SIB, The University of Queensland,for a critical reading of the manuscript.

LITERATURE CITED

Andres AJ, Fletcher JC, Karim FD, Thummel CS. 1993. Mo-

lecular analysis of the initiation of insect metamorphosis:

132 Chen et al.


a comparative study of Drosophila ecdysteroid-regulated

transcription. Dev Biol 160: 388–404.

Ashburner M, Chihara C, Meltzer P, Richards G. 1974. Tem-

poral control of puffing activity in polytene chromosomes.

Cold Spring Harb Symp Quant Biol 38:655–662.

Baehrecke EH and Thummel CS. 1995. The Drosophila E93

gene from the 93F early puff displays stage- and tissue-

specific regulation by 20-hydroxyecdysone. Dev Biol

171:85–97.

Bellaiche Y, Bandyopadhyay R, Desplan C, Dostatni N. 1996.

Neither the homeodomain nor the activation domain of

Bicoid is specifically required for its down-regulation by

the Torso receptor tyrosine kinase cascade. Development

122:3499–3508.

Bialecki M, Shilton A, Fichtenberg C, Segraves WA, Thummel1

CS. 2002. Loss of the ecdysteroid-inducible E75A orphan

nuclear receptor uncouples molting from metamorphosis

in Drosophila. Dev Cell 3:209–220.

Broadus J, McCabe JR, Endrizzi B, Thummel CS, Woodard

CT. 1999. The Drosophila betaFTZ-F1 orphan nuclear re-

ceptor provides competence for stage-specific responses to

the steroid hormone ecdysone. Mol Cell 3:143–149.

Burtis KC, Thummel CS, Jones CW, Karim FD, Hogness DS.

1990. The Drosophila 74EF early puff contains E74, a com-

plex ecdysone-inducible gene that encodes two ets-related

proteins. Cell 61:85–99.

Calogero S, Grassi F, Aguzzi A, Voigtlander T, Ferrier P, Ferrari

S, Bianchi ME. 1999. The lack of chromosomal protein

Hmg1 does not disrupt cell growth but causes lethal

hypoglycaemia in newborn mice. Nat Genet 22:276–280.

Chen J, Gong YF, Hu Z, Wang YF. 2006. Effects of over-ex-

pression of high mobility group proteins gene on the de-

velopment of Drosophila. Acta Zool Sinica 52:335–341.

Dimitrov S, Almouzni G, Dasso M, Wolffe AP. 1993. Chro-

matin transitions during early Xenopus embryogenesis:

changes in histone H4 acetylation and in linker histone

type. Dev Biol 160:214–227.

Dubrovsky BE. 2005. Hormonal cross talk in insect develop-

ment. Trends Endocrinol Meta 16:6–11.

Fechtel K, Natzle JE, Brown EE, Fristrom JW. 1988. Prepu-

pal differentiation of Drosophila imaginal discs: identi-

fication of four genes whose transcripts accumulate in

response to a pulse of 20-hydroxyecdysone. Genetics

120:465–474.

Fechtel K, Fristrom DK, Fristrom JW. 1989. Prepupal differ-

entiation in Drosophila: distinct cell types elaborate a

shared structure, the pupal cuticle, but accumulate tran-

scripts in unique patterns. Development 106:649–656.

Fletcher J, Thummel CS. 1995. The Drosophila E74 gene is

required for the proper stage- and tissue-specific transcrip-

tion of ecdysone-regulated genes at the onset of metamor-

phosis. Development 121:1411–1421.

Freeman MR, Dobritsa A, Gaines P, Segraves WA, Carlson JR.

1999. The dare gene: steroid hormone production, olfac-

tory behavior, and neural degeneration in Drosophila. De-

velopment 126:4591–4602.

Gates J, Lam G, Ortiz JA, Losson R, Thummel CS. 2004. Rigor

mortis encodes a novel nuclear receptor interacting pro-

tein required for ecdysone signaling during Drosophila lar-

val development. Development 131:25–36.

Gonzy G, Pokholkova GV, Peronnet F, Mugat B, Demakova

OV, Kotlikova IV, Lepesant JA, Zhimulev IF. 2002. Isolation

and characterization of novel mutations of the Broad-Com-

plex, a key regulatory gene of ecdysone induction in Droso-

phila melanogaster. Insect Biochem Mol Biol 32:121–132.

Karim FD, Guild GM, Thummel CS. 1993. The Drosophila

Broad-Complex plays a key role in controlling ecdysone-

regulated gene expression at the onset of metamorphosis.

Development 118:977–988.

Koelle MR, Talbot WS, Segraves WA, Bender MT, Cherbas P,

Hogness DS. 1991. The Drosophila EcR gene encodes an

ecdysone receptor, a new member of the steriod receptor

superfamily. Cell 67:59–77.

Lam G, Hall BL, Bender M, Thummel CS. 1999. DHR3 is

required for the prepupal-pupal transition and differen-

tiation of adult structures during Drosophila metamorpho-

sis. Dev Biol 212:204–216.

Mugat B, Brodu V, Kejzlarova-Lepesant J, Antoniewski C, Bayer

CA, Fristrom JW, Lepesant JA. 2000. Dynamic expression

of Broad-Complex isoforms mediates temporal control of

an ecdysteroid target gene at the onset of Drosophila meta-

morphosis. Dev Biol 227:104–117.

Nemeth MJ, Curtis DJ, Kirby MR, Garrett-Beal LJ, Seidel NE,

Cline AP, Bodine DM. 2003. Hmgb3: an HMG-box family

member expressed in primitive hematopoietic cells that



inhibits myeloid and B-cell differentiation. Blood 102:

1298–1306.

Nemeth MJ, Cline AP, Anderson SM, Garrett-Beal LJ, Bodine

DM. 2005. Hmgb3 deficiency deregulates proliferation and

differentiation of common lymphoid and myeloid pro-

genitors. Blood 105:627–634.

Ner SS, Travers AA. 1994. HMG-D, the Drosophila melanogaster

homologue of HMG 1 protein, is associated with early

embryonic chromatin in the absence of histone H1. EMBO

J 13:1817–1822.

Ner SS, Blank T, Perez-Paralle ML, Grigliatti TA, Becker PB,

Travers AA. 2001. HMG-D and histone H1 interplay dur-

ing chromatin assembly and early embryogenesis. J Biol

Chem 276:37569–37576.

Nightingale K, Dimitrov S, Reeves R, Wolffe AP. 1996. Evi-

dence for a shared structural role for HMG1 and linker

histones B4 and H1 in organizing chromatin. EMBO J

15:548–561.

Ragab A, Thompson CE, Travers AA. 2006. High mobility

group proteins HMGD and HMGZ interact genetically with

the brahma chromatin remodelling complex in Drosophila.

Genetics 172:1069–1078.

Renner U, Ghidelli S, Schafer MA, Wisniewski JR. 2000. Al-

terations in titer and distribution of high mobility group

proteins during embryonic development of Drosophila

melanogaster. Biochim Biophys Acta 1475:99–108.

Ronfani L, Ferraguti M, Croci L, Ovitt CE, Scholer HR,

Consalez GG, Bianchi ME. 2001. Reduced fertility and sper-

matogenesis defects in mice lacking chromosomal protein

Hmgb2. Development 128:1265–1273.

Rorth P. 1998. Gal4 in the Drosophila female germline. Mech

Dev 78:113–118.

Rorth P, Szabo K, Bailey A, Laverty T, Rehm J, Rubin GM,

Weigmann K, Milán M, Benes V, Ansorge W, Cohen SM.

1998. Systematic gain-of-function genetics in Drosophila.

Development 125:1049–1057.

Segraves WA, Hogness DS. 1990. The E75 ecdysone-induc-

ible gene responsible for the 75B early puff in Drosophila

encodes two new members of the steroid receptor super-

family. Genes Dev 4:204–219.

Thomas HE, Stunnenberg HG, Stewart AF. 1993. Hetero-

dimerization of the Drosophila ecdysone receptor with ret-

inoid X receptor and ultraspiracle. Nature 362:471–475.

Thomas JO, Travers AA. 2001. HMG1 and 2, and related “ar-

chitectural” DNA-binding proteins. Trends Biochem Sci

26:167–174.

Thummel CS. 2001. Molecular mechanisms of developmen-

tal timing in C. elegans and Drosophila. Dev Cell 1:453–

465.

Thummel CS. 2002. Ecdysone-regulated puff genes 2000. In-

sect Biochem Mol Biol 32:113–120.

Venkatesh K, Hasan G. 1997. Disruption of the IP3 receptor

gene of Drosophila affects larval metamorphosis and ecdys-

one release. Curr Biol 7:500–509.

White KP, Hurban P, Watanabe T, Hogness DS. 1997. Coor-

dination of Drosophila metamorphosis by two ecdysone-

induced nuclear receptors. Science 276:114–117.

Woodard CT, Baehrecke EH, Thummel CS. 1994. A molecu-

lar mechanism for the stage-specificity of the Drosophila

prepupal genetic response to ecdysone. Cell 79:607–615.

Yao T, Segraves WA, Oro AE, McKeown M, Evans RM. 1992.

Drosophila ultraspiracle modulates ecdysone receptor func-

tion via heterodimer formation. Cell 71:63–72.

Yao TP, Forman BM, Jiang Z, Cherbas L, Chen JD, McKeown

M, Cherbas P, Evans RM. 1993. Functional ecdysone re-

ceptor is the product of EcR and Ultraspiracle genes. Na-

ture 366:476–479.

Zhou X, Zhou B, Truman JW, Riddiford LM. 2004. Over-

expression of broad: a new insight into its role in the

Drosophila prothoracic gland cells. J Exp Biol 207:1151-

1161.

overexpression of hmgd causes the failure of pupariation in drosophila by affecting ecdysone...

Documents