Copyright � 2009 by the Genetics Society of AmericaDOI: 10.1534/genetics.109.102053

Drosophila ISWI Regulates the Association of Histone H1 WithInterphase Chromosomes in Vivo

Giorgia Siriaco,* Renate Deuring,* Mariacristina Chioda,† Peter B. Becker†

and John W. Tamkun*,1

*Department of Molecular, Cell and Developmental Biology, University of California, Santa Cruz, California 95064 and†Adolf-Butenandt-Institute, Molecular Biology, Munich Center of Integrated Protein Science,

Ludwig-Maximilians-University, 80336 Munich, Germany

Manuscript received February 25, 2009Accepted for publication April 8, 2009

ABSTRACT

Although tremendous progress has been made toward identifying factors that regulate nucleosomestructure and positioning, the mechanisms that regulate higher-order chromatin structure remain poorlyunderstood. Recent studies suggest that the ISWI chromatin-remodeling factor plays a key role in thisprocess by promoting the assembly of chromatin containing histone H1. To test this hypothesis, weinvestigated the function of H1 in Drosophila. The association of H1 with salivary gland polytenechromosomes is regulated by a dynamic, ATP-dependent process. Reducing cellular ATP levels triggersthe dissociation of H1 from polytene chromosomes and causes chromosome defects similar to thoseresulting from the loss of ISWI function. H1 knockdown causes even more severe defects in chromosomestructure and a reduction in nucleosome repeat length, presumably due to the failure to incorporate H1during replication-dependent chromatin assembly. Our findings suggest that ISWI regulates higher-orderchromatin structure by modulating the interaction of H1 with interphase chromosomes.

THE packaging of DNA into chromatin is critical forthe organization and regulation of eukaryotic genes.Thebasicunitofchromatinstructure—thenucleosome—-can be packaged in 30-nm fibers and increasinglycompact structures. Higher-order chromatin structureinfluences many aspects of gene expression, includingtranscription factor binding, enhancer–promoter in-teractions, and the organization of chromatin intofunctional domains. Histone H1 and related linkerhistones are important determinants of higher-orderchromatin structure. These abundant, basic proteinsshare a common structure consisting of a globularwinged helix DNA-binding domain flanked by a shortN-terminal segment and a C-terminal domain of �100amino acids (Brown 2003). The winged helix domainof H1 binds the nucleosome near the site of DNA entryand exit; the flanking domains interact with core andlinker DNA to promote the formation and packaging of30-nm fibers in vitro (Robinson and Rhodes 2006;Maier et al. 2008).

In vitro studies suggest that nucleosomal arrays havean intrinsic propensity to fold into 30-nm fibers that arestabilized by association of H1 (Carruthers et al. 1998).However, the function of H1 in vivo is not well un-

derstood. In lower eukaryotes, proteins related to H1play surprisingly subtle roles in chromosome organiza-tion and gene expression (Godde and Ura 2008). Inhigher eukaryotes, the study of H1 function has beencomplicated by the presence of multiple, functionallyredundant H1 subtypes (Khochbin 2001). H1 expres-sion has been partially reduced in nematodes, frogs, andmice (Godde and Ura 2008). A partial reduction in H1levels has limited effects on gene expression in mice, butleads to the formation of nucleosome arrays that are lesscompact than normal (Fan et al. 2005). The immuno-depletion of H1 from Xenopus extracts results in theassembly of elongated metaphase chromosomes thatfail to align and segregate properly (Maresca et al.2005). These findings suggest that H1 plays an impor-tant role in chromosome organization. Since it has notbeen possible to completely eliminate H1 in any highereukaryote, its function in vivo remains a topic ofconsiderable debate.

The association of H1 with chromatin is highlydynamic. In both Tetrahymena and mammals, H1 israpidly exchanged between chromatin fibers (Lever et al.2000; Misteli et al. 2000; Dou et al. 2002; Catez et al.2006). The dissociation of H1 from chromatin is thoughtto disrupt 30-nm fibers and provide an opportunity fortranscription factors or other regulatory proteins toaccess DNA. The association of H1 with chromatin isinfluenced by H1 phosphorylation, core histone acetyla-tion, and competition with other chromatin-binding

Supporting information is available online at http://www.genetics.org/cgi/content/full/genetics.109.102053/DC1.

1Corresponding author: 350 Sinsheimer Labs, Department of Molecular,Cell and Developmental Biology, University of California, Santa Cruz, CA95064. E-mail: tamkun@biology.ucsc.edu

Genetics 182: 661–669 ( July 2009)

http://www.genetics.org/cgi/content/full/genetics.109.102053/DC1http://www.genetics.org/cgi/content/full/genetics.109.102053/DC1http://www.genetics.org/cgi/content/full/genetics.109.102053/DC1http://www.genetics.org/cgi/content/full/genetics.109.102053/DC1http://www.genetics.org/cgi/content/full/genetics.109.102053/DC1

proteins (Catez et al. 2006). However, little is knownabout either the mechanism of H1 exchange or how thisprocess is regulated in vivo.

One of the best candidates for a factor that regulates H1assembly is Drosophila ISWI. ISWI is the ATPase subunitof multiple chromatin-remodeling complexes—includingCHRAC, NURF, and ACF—that slide nucleosomes andalter the spacing of nucleosome arrays (Bouazoune andBrehm 2006). ACF also promotes the assembly of chro-matin containing H1 in vitro (Lusser et al. 2005).Although ISWI is not required for H1 expression in vivo,the loss of ISWI function leads to the decondensation ofmitotic and polytene chromosomes accompanied by theloss of H1 (Corona et al. 2007). On the basis of theseobservations, we proposed that ISWI regulates chromo-some structure by promoting H1 assembly (Corona et al.2007). To test this hypothesis and clarify the function ofhistone H1 in vivo, we investigated phenotypes resultingfrom the loss of H1 in Drosophila.

MATERIALS AND METHODS

Drosophila stocks and crosses: Flies were raised on corn-meal, agar, yeast, and molasses medium, supplemented withmethyl paraben and propionic acid. The GAL4 system (Brandet al. 1994) was used to drive the expression of His1-RNAi andISWIK159R. da-GAL4 is expressed broadly at all stages of de-velopment (Gerber et al. 2004). For viability studies, UAS-His1-dsRNA males were crossed to da-GAL4 or Df(1)w67c2 femalesand the progeny were scored for survival to adulthood. Allcrosses were carried out at 29� unless otherwise indicated.

Generation of transgenic strains bearing UAS-His1-dsRNAtransgenes: The Drosophila His1 coding region was amplifiedfrom Canton-S genomic DNA by PCR using the primers 59-CGAATTCGACAGTTGAGAAGAAAGTGGTCC-39 and 59-GGGTGGCCATCTTGGCCGTAGTCTTCGCT-39 or 59-CCGCTCGAGACAGTTGAGAAGAAAGTGG-39 and 59-GGGTGGCCTAGATGGCCGTAGTCTTCGCTT-39. The resulting PCR products weredigested with SfiI and ligated to form an inverted repeat flankedby EcoRI and XhoI sites. The inverted repeats were cleaved withEcoRI and XhoI and subcloned into pUAST. BLAST searchesrevealed that the His1 fragment in this construct is not sufficientlyrelated to other regions of the Drosophila genome to generateoff-target effects. Transformants were generated by P-element-mediated transformation using the Df(1)w67c2 strain. Ho-mozygous viable transformants used in the study includeUAS-His1-dsRNA-8-4 and UAS-His1-dsRNA-13-1 on the X chro-mosome and UAS-His1-dsRNA-10-3 on chromosome 3.

Generation of H1-Flag-CFP transgenic strains: The codingsequence for Drosophila His1 was amplified by PCR from acDNA clone using the primers 59-GCTATGCTATGCGGCCGCATGTCTGATTCTGCAGTT-39 and 59-CATACCGGTCTTGTCGTCGTCGTCCTTGTAGTCCTTTTTGGCAGCCGTAG-39.The sequence of CFP was amplified by PCR using the primers59- GCTATGCTATGCGGCCGCACCGGTATGGTGAGCAAGGGCGA-39 and 59-CACTAGTTACTTGTACAGCTCGTCCATG-39.The PCR products were cloned in the pCR2.1-TA Topo vector(Invitrogen). The H1 insert was digested with SpeI and NotIand subcloned into pBS-SK. The CFP fragment was digestedwith AgeI and SpeI and cloned into pBS-dH1 using the samerestriction sites. The H1–flag–CFP fusion was digested withNotI and SpeI and subcloned downstream of a constitutivelyexpressed a-tubulin promoter in pCaSpeR4 (generously pro-

vided by Konrad Basler). The construct was sequenced and usedto generate a y w strain bearing a homozygous viable insertion onthe third chromosome by P-element-mediated transformation.

Analysis of polytene chromosome structure: Salivary glandsof third instar larvae were dissected in 0.7% NaCl and fixed in1.85% formaldehyde/45% acetic acid as previously described(Corona et al. 2007). To analyze the effect of H1 knockdownon chromosome structure, da-GAL4 females were mated toP[w1, UAS-His1-dsRNA-8-4]/Y males. To analyze the effect ofISWIK159R expression on chromosome structure, H2AvD-GFPfemales were mated to w; P[w1, eyGAL4], P[w1, UAS-ISWIK159R-HA-6His]11-4/TM3 males at 18�. Chromosome preparationswere analyzed using a Zeiss Axioskop 2 plus fluorescentmicroscope equipped with an Axioplan HRm CCD cameraand Axiovision 4.2 software (Zeiss). For DNA quantification,images were captured using identical exposure times. Chro-mosome boundaries were identified and the sum pixel in-tensity within chromosomes was calculated using Volocitysoftware (Release 4.2.1; http://www.improvision.com). Anti-bodies used in this study are affinity-purifed rabbit anti-ISWI(Tsukiyama et al. 1995), rabbit anti-H3K9me3 (Abcam,ab8898), and rabbit anti-H4Ac(tetra) (Active Motif, 39179).

Electrophoresis and protein blotting: To analyze the effectof H1 knockdown on nucleosome repeat length, da-GAL4females were mated to P[w1, UAS-His1-dsRNA-8-4]/Y males.Salivary gland protein extracts were prepared from third-instarlarvae and analyzed by protein blotting as described previously(Corona et al. 2007) using affinity-purified rabbit antibodiesagainst ISWI (Tsukiyama et al. 1995) and rabbit polyclonalantibodies against H1(Ner and Travers 1994) and H3(Abcam, ab1791).

Analysis of salivary gland chromatin by micrococcalnuclease digestion: Salivary gland chromatin was extractedfrom P[w1, UAS-His1-dsRNA-8-4]/1; da-GAL4/1 or control da-GAL4/1 third instar larvae and partially digested with micro-coccal nuclease as described previously (Corona et al. 2007).Images were obtained using a GelDoc camera and QuantityOne software (Bio-Rad Laboratories). Separate experimentswere carried out at least three times and gave highly re-producible results.

Confocal microscopy and FRAP analysis: For live analysis ofpolytene chromosome phenotypes resulting from the loss ofISWI or H1 function, or ATP depletion, one or two represen-tative nuclei were chosen to be analyzed per gland. In general,the appearance of nuclei within a single salivary gland was veryreproducible. Thus, the imaged nuclei are representative of amuch larger number of nuclei observed in several glands.Nuclei at the tip of the gland were analyzed whenever possibleto ensure consistent results. To analyze the effect of H1knockdown on chromosome structure in living cells, da-GAL4, H2AvD-GFP females were mated to P[w1, UAS-His1-dsRNA-8-4]/Y males. To analyze the effect of ISWIK159R

expression on chromosome structure in living cells, H2AvD-GFP females were mated to P[w1, eyGAL4], P[w1, UAS-ISWIK159R-HA-6His]11-4/TM3 males at 18�. Live polytene chromosomenuclei were imaged using an inverted microscope (DM IRB,Leica Microsystems) equipped with a laser confocal imagingsystem (TCS SP2, Leica Microsystems). Three dimensionalreconstruction and volume calculations of 0.5-mm sections ofpolytene nuclei were performed by Volocity software (release4.2.1, http://www.improvision.com). The change in chromatincompaction was established by calculating the ratio of volumeto DNA. The ratio of control samples was normalized to one.FRAP analysis of salivary gland nuclei was carried out using aninverted microscope (DM IRB, Leica Microsystems) equippedwith a laser confocal imaging system (TCS SP2, Leica Micro-systems). Images were acquired and analyzed using the FRAPapplication of the Leica Microsystems confocal software

662 G. Siriaco et al.

http://www.improvision.comhttp://www.improvision.comhttp://www.improvision.comhttp://www.improvision.com

version 2.61. Salivary glands were dissected from third instarlarvae and incubated in Schneider’s insect medium (Sigma)containing 50 mg/ml aphidicolin (Sigma), or the equivalentvolume of DMSO, for 4 hr. Glands were then transferred to acoverslip and covered in mineral oil for FRAP analysis. Foreach experiment, 10 single imaging scans were acquiredfollowed by 15 bleach pulses of 600 ms within a square regionof interest (ROI) measuring 6 3 6 mm. Images were thencollected every 0.6 sec (10 images), every 3 sec (10 images),and every 5 sec (30 images). For imaging, the laser power wasattenuated to 16% of the bleach intensity. A second ROImeasuring 6 3 6 mm within the same polytene nucleus wasused to normalize fluorescence values against background.FRAP recovery curves were generated and analyzed usingMicrosoft Excel. The recovery curves described represent theaverage values of eight or more experiments.

To investigate ATP dependence of H1 exchange, salivaryglands were treated with agents that block oxidative phos-phorylation. Salivary glands were dissected from third instarlarvae expressing H2AvD-GFP or H1-CFP, and incubated for1 hr in 13 PBS containing 100 mm sodium azide (Sigma), 100 mmantimycin A (Sigma), or 2 hr in 13 PBS containing 250 mmrotenone (MP Biomedicals, LLC). Control glands were in-cubated for 1 hr in 13 PBS. Glands were then transferred to acoverslip and covered in mineral oil for live analysis. Sectionsof 1 mm whole salivary gland nuclei were acquired. Seventeenazide-treated H1-CFP and 10 azide-treated H2AvD-GFP nucleiwere analyzed. For untreated control experiments, 11 H1-CFPand 6 H2AvD-GFP nuclei were analyzed. Similar treatmentshave been shown to reduce ATP levels in Drosophila salivaryglands by two- to fourfold within 2 hr (Leenders et al. 1974).

RESULTS

H1 is essential for Drosophila development: Unlikeother higher eukaryotes, Drosophila contains only oneH1 subtype (His1) that is highly related to mammalianH1. Classical genetic approaches cannot be used tostudy His1 since .100 copies of this gene are present inthe Drosophila genome. We therefore used RNA in-terference (RNAi) to study His1 function. Strainsbearing a transgene encoding a His1 hairpin-loopRNA under the control of a GAL4-inducible promoter(UAS-His1-dsRNA) were generated by P-element-medi-ated transformation. To induce the expression of this

transgene, transformants were crossed to strains bear-ing a daughterless-GAL4 (da-GAL4) transgene that isubiquitously expressed at high levels. The expressionof the His1 hairpin RNA under the control of the da-GAL4 driver resulted in death during late larval or earlypupal stages, indicating that H1 is essential for de-velopment (Table 1). We were unable to completelyeliminate His1 expression in imaginal discs or larvalneuroblasts, but occasionally observed interphase nu-clei with highly disorganized chromatin in these tissues,accompanied by a severe reduction in the number ofmetaphase chromosomes (data not shown). These datasuggest that H1 is essential for progression throughmitosis.

Histone H1 is a major determinant of chromosomestructure in vivo: We next analyzed phenotypes result-ing from H1 knockdown in the larval salivary gland.Repeated rounds of DNA replication in the absence ofcytokinesis in this tissue leads to the formation ofpolytene chromosomes that serve as a useful modelfor interphase chromosomes. The expression of His1dsRNA in the salivary gland led to a significant re-duction in H1 levels (Figure 1A) accompanied by highlypenetrant changes in chromosome structure (Figure 1,C, D, and F). Similar phenotypes resulted from theexpression of three independent insertions of the UAS-His1-dsRNA transgene, but were never observed inlarvae bearing only the da-GAL4 driver or UAS-His1-dsRNA transgene (Figure 1B and data not shown). Themost common phenotype resulting from His1 knock-down was the broadening of chromosome arms withoutan obvious disruption of their banding pattern (Figure1, C and D). The increase in chromosome size was notdue to extra rounds of replication, since the DNAcontent of chromosomes of control larvae and larvaeexpressing His1 dsRNA were similar (Figure 1E).Decondensed regions of chromatin and ectopic con-tacts between chromosome arms were occasionallyobserved (data not shown). In extreme cases, thebanding pattern was completely disrupted and individ-ual arms were no longer distinguishable (Figure 1F).

TABLE 1

His1 is essential for development

Survival to adulthood

Cross Females Males

da-GAL4 3 P[w1, UAS-His1-dsRNA-8-4]/Y 0 44Df(1)w 3 P[w1, UAS-His1-dsRNA-8-4]/Y 78 85da-GAL4 3 P[w1, UAS-His1-dsRNA-13-1]/Y 0 127Df(1)w 3 P[w1, UAS-His1-dsRNA-13-1]/Y 94 121da-GAL4 3 P[w1; UAS-His1-dsRNA-10-3] 0 0Df(1)w 3 P[w1; UAS-His1-dsRNA-10-3] ND ND

Homozygous da-GAL4 or Df(1)w virgin females were mated to males bearing UAS-His1-dsRNA transgenes onthe X (8-4, 13-1) or third chromosome (10-3) at 29� and scored for survival to adulthood. In all cases, lethalityoccurred at the late larval or early pupal stages. ND, not determined.

Drosophila ISWI Regulates H1 Assembly 663

The chromosome defects resulting from His1 knock-down were not limited to euchromatin, as evidenced bythe dispersion of the heterochromatic chromocenter(supporting information, Figure S1). H1 is thus a majordeterminant of chromosome structure in vivo.

Similar chromosome defects result from the loss ofH1 and ISWI function: To clarify the functional re-lationship between ISWI and histone H1, we comparedphenotypes resulting from their loss of function. Wepreviously demonstrated that ISWI plays a global role inchromatin compaction that is antagonized by theacetylation of lysine 16 of the histone H4 tail (H4K16)(Deuring et al. 2000; Corona et al. 2002). The male Xchromosome— which is acetylated on H4K16 by thedosage compensation complex—is therefore particu-

larly sensitive to the loss of ISWI function. The partialloss of ISWI function leads to the decondensation of themale X chromosome (Figure 1, B, D, and G) accompa-nied by the loss of H1 (Corona et al. 2007). A furtherreduction in ISWI function (due to the expression ofthe dominant-negative ISWIK159R protein) leads to thedecondensation of all chromosomes (Figure 1, B andH) accompanied by the loss of H1 (Corona et al. 2007).

The spectrum of chromosome defects resulting fromthe loss of ISWI function and H1 are similar, but notidentical. The X chromosome of ISWI mutant malesappears much broader than normal, but usually retainsits banding pattern (Figure 1, D and G). H1 knockdowncaused similar defects (Figure 1, C, D, and G) but didnot have as pronounced an effect on chromosome

Figure 1.—Loss of histone H1 alters chromosome structure. (A) Reduced histone H1 expression is observed in the salivaryglands of P[w1, UAS-His1-dsRNA-8-4]/1; da-GAL4/1 larvae, compared to control da-GAL4/1 larvae, as assayed by protein blotting.Protein sizes can be determined by referring to molecular weight markers alongside the gel. (B–D and F–H) Polytene chromo-somes stained with DAPI. Control da-GAL4/1 chromosomes (B) exhibit normal morphology while His1 RNAi leads to chromo-some decondensation (C–D and F). (D) A magnification of the boxed regions of B, C, and G. P[w1, UAS-His1-dsRNA-8-4]/1;da-GAL4/1 chromosomes and the male Iswi1/Iswi2 X chromosome are decondensed relative to the control chromosome, but thebanding pattern is maintained. (E) Quantification of DNA in P[w1, UAS-His1-dsRNA-8-4]/1; da-GAL4/1 and control da-GAL4/1 chro-mosomes . (F) Individual chromosome arms are no longer distinguishable in some nuclei. (G) The male X chromosome (arrowhead)is decondensed in Iswi1/Iswi2 larvae. (H) Expression of ISWIK159R leads to disorganized chromatin (arrowhead) and decondensation(arrow) of all chromosomes. (I) Quantification of DNA in P[w1, eyGAL4], P[w1, UAS-ISWIK159R-HA-6His]11-4/H2AvD-GFP and controlH2AvD-GFP/TM3 chromosomes. Bars, 20 mm.

664 G. Siriaco et al.

http://www.genetics.org/cgi/data/genetics.109.102053/DC1/1

length (Figure 1, B and C). In general, the expression ofHis1 dsRNA led to a much greater increase in the size ofpolytene chromosomes than the expression of ISWIK159R

(compare Figure 1, F to H). This may be due to areduction in DNA replication in larvae expressingISWIK159R (Figure 1I). Overall, the similarities betweenthe phenotypes resulting from the loss of His1 and ISWIfunction support our proposal that ISWI regulateschromatin structure by promoting the incorporationof H1 into chromatin.

To verify that H1 acts downstream of ISWI to regulatechromatin structure, we examined whether the loss ofH1 altered either the expression of ISWI or its associ-ation with chromatin. The expression of His1 dsRNAdramatically reduced the expression of H1 in salivarygland nuclei without decreasing the overall level of ISWIprotein (Figure 2A). We also failed to observe obviousdifferences in the level of ISWI associated with thepolytene chromosomes of larvae expressing His1 dsRNAand control larvae (Figure 2, B and C). Thus, H1 doesnot appear to modulate chromosome structure byaltering the expression of ISWI or its association withchromatin.

Histone H1 undergoes rapid, replication-indepen-dent exchange in vivo: How does ISWI promote theassociation of H1 with chromatin? ISWI can promotethe assembly of nucleosome arrays containing H1in vitro (Lusser et al. 2005; Maier et al. 2008), suggestingthat it may be required for replication-coupled chroma-tin assembly in vivo. ISWI could also promote H1incorporation via a replication-independent mecha-nism, since H1 undergoes rapid, ATP-dependent ex-change throughout the cell cycle in other organisms(Catez et al. 2006). As a first step toward distinguishingbetween these possibilities, we used fluorescence re-covery after photobleaching (FRAP) to analyze inter-actions between H1 and chromosomes in a strainexpressing CFP-tagged histone H1. We found that themajority of H1 associated with chromosomes undergoesrapid exchange in vivo (Figure 3A). As observed inmammalian cells (Misteli et al. 2000), approximatelyhalf the H1 underwent exchange within 50 sec. Thisexchange must be replication independent due to theshort duration of our experiment and the fact that onlytwo to three rounds of DNA replication occur over 48 hrin the salivary glands of third instar larvae (Rodman1967). Furthermore, treatment of salivary glands withaphidicolin, an inhibitor of DNA replication, did notaffect the rate of H1 exchange (Figure 3A). H1exchange therefore occurs independently of replica-tion-coupled chromatin assembly in this tissue.

H1 knockdown decreases nucleosome repeat length:The above findings indicated that ISWI might promotethe association of H1 with chromatin during eitherreplication-coupled chromatin assembly or replication-independent H1 exchange. To help distinguish be-tween these possibilities, we compared changes in

nucleosome repeat length (NRL) resulting from theloss of H1 and ISWI function. Incorporation of H1during de novo chromatin assembly increases the averagedistance between nucleosomes and there is a strongcorrelation between NRL and the amount of H1 in-corporated during chromatin assembly (Blank andBecker 1995; Woodcock et al. 2006; Routh et al.2008). Thus, the reduced expression of H1—or factorsthat promote replication-coupled H1 assembly—shouldcause a significant decrease in NRL. By contrast, the lossof factors required for replication-independent H1exchange should have little or no effect on NRL, sincethis process occurs after genomewide nucleosomedensity has been established. We previously demon-strated that the loss of ISWI function has no apparenteffect on NRL in the larval salivary gland, even though itleads to the loss of H1 from chromosomes (Coronaet al. 2007). By contrast, reducing the level of H1 in thesalivary gland via expression of His1 dsRNA leads to areproducible 14-bp decrease in NRL from 172 to 158 bp(Figure 3B). These data suggest that ISWI is notrequired for replication-dependent H1 assembly insalivary gland nuclei.

ATP is required for the association of histone H1with interphase chromosomes: If ISWI is required forreplication-independent H1 assembly, a reduction in

Figure 2.—Histone H1 is not required for the expressionof ISWI or its binding to chromatin. (A) Levels of ISWIprotein are not affected in the salivary glands of P[w1, UAS-His1-dsRNA-8-4]/1; da-GAL4/1 larvae, compared to controlda-GAL4/1 larvae, as assayed by protein blotting. A compara-ble blot was probed with antibodies against histone H3 as acontrol. Protein sizes can be determined by referring to mo-lecular weight markers alongside the gel. (B and C) Polytenechromosomes of da-GAL4/1 (B) and P[w1, UAS-His1-dsRNA-8-4]/1; da-GAL4/1 (C) larvae were stained with an antibodyagainst ISWI. Polytene chromosomes were prepared and pro-cessed in parallel, and images were captured using identicalexposure times.

Drosophila ISWI Regulates H1 Assembly 665

cellular ATP levels should lead to the loss of H1 fromchromosomes. To test this prediction, we monitored theassociation of histone H1-CFP with polytene chromo-somes by live analysis following exposure to inhibitors ofoxidative phosphorylation. Within 1 hr of azide treat-ment, H1-CFP was detected in the nucleoplasm in.90% of nuclei (Figure 3C, second panel); this wasnever observed in untreated salivary glands. In�15% ofnuclei, all the H1-CFP had dissociated from chromo-somes and was found in the nucleoplasm (Figure 3C,third panel). Similar results were obtained with otherinhibitors of oxidative phosphorylation, including anti-mycin A and rotenone (data not shown). By contrast,

azide treatment had no effect on the association of atagged core histone, H2AvD-GFP (Clarkson and Saint1999), with chromosomes (Figure 3C, fourth panel).Since ATP-depletion affects many cellular processes, it ispossible that azide treatment triggers the dissociation ofH1 from polytene chromosomes via an ISWI-indepen-dent mechanism. However, our data suggest that repli-cation-independent H1 assembly is an energy-dependentprocess that is subject to regulation by ISWI or otherfactors.

Characterization of chromosome defects resultingfrom the loss of H1 and ISWI in living cells: Liveanalysis revealed that the dissociation of H1 from

Figure 3.—Histone H1 is rapidly exchangedin salivary gland nuclei and increases NRL. (A)Quantitative analysis of FRAP experiments. Therecovery curve for aphidicolin-treated nucleiare shown in dark shading, the recovery curvefor control DMSO-treated nuclei in light shad-ing. Salivary glands were incubated in aphidico-lin or DMSO for 4 hr prior to FRAP analysis.(B) Partial micrococcal nuclease digestionof chromatin isolated from salivary glands ofda-GAL4/1 and P[w1, UAS-His1-dsRNA-8-4]/1;da-GAL4/1 larvae. Loss of histone H1 leads toa reduction in NRL compared to control. DNAfragment sizes can be determined by referringto the 100-bp ladder alongside the gel. (C)Examples of phenotypes resulting from azidetreatment of nuclei expressing H1-CFP orH2AvD-GFP; dashed line identifies the nuclearboundary. Histone H1 dissociates from chromo-somes and appears in the nucleoplasm. Azidetreatment had no effect on H2AvD associationwith chromosomes. Bars, 20 mm.

666 G. Siriaco et al.

chromosomes following treatment with inhibitors ofoxidative phosphorylation was not accompanied byobvious changes in nuclear diameter or chromosomevolume (Figure 3C). This was surprising, since tradi-tional methods for fixing and squashing polytenechromosomes showed that the loss of H1 significantlyincreased the size of polytene chromosomes (seeabove). To gain a more accurate impression of therelative roles of H1 and ISWI in chromosome organiza-tion, we visualized chromosomes in living cells express-ing H2AvD-GFP. The expression of His1 dsRNA caused atwo- to fivefold increase in their volume (Figure 4, A, B,E, and G–I). This increase was not due to extra rounds ofreplication, since the DNA content of chromosomes ofcontrol larvae and larvae expressing His1 dsRNA weresimilar (Figure 1E). Interestingly, we never observedobvious changes in the banding pattern of chromo-somes following H1 knockdown in living cells, evenwhen the chromosome volume increased dramatically(Figure 4, H and I). Thus, His1 RNAi caused muchgreater changes in salivary gland polytene chromosomestructure than ATP depletion, even though both con-ditions led to a significant reduction in the level of H1associated with chromatin.

Live analysis of salivary gland nuclei expressing H2AvD-GFP also revealed differences between the chromosomedefects resulting from the expression of His1 dsRNA andthe dominant-negative ISWIK159R protein. The expressionof ISWIK159R did not cause obvious changes in chromo-some size (Figure 4, C, D, and F), even though thesechromosomes contain reduced levels of H1 (Corona et al.2007). Indeed, when normalized for DNA content, thevolume of chromosomes of control larvae and larvaeexpressing ISWIK159R were indistinguishable (Figure 4G).The banding pattern of polytene chromosomes was oftendisrupted, however, and we frequently observed ‘‘holes,’’which may represent regions of decondensed chromatin(Figure 4, C, D, H, and J). These defects are similar tothose observed following the treatment of salivary glandswith inhibitors of oxidative phosphorylation (compareFigures 3C and 4D).

DISCUSSION

Our findings provide direct evidence that H1 is amajor determinant of interphase chromosome struc-ture and support our proposal that ISWI regulateshigher-order chromatin structure by promoting theassociation of H1 with chromatin. The incorporation

Figure 4.—Loss of histone H1 increases chromosome vol-ume. (A–D and H–J) Live analysis of nuclei expressingH2AvD-GFP reveals decondensation of P[w1, UAS-His1-dsRNA-8-4]/1; da-GAL4, H2AvD-GFP/1 chromosomes (B)compared to da-GAL4, H2AvD-GFP/1 chromosomes (A).(D) P[w1, eyGAL4], P[w1, UAS-ISWIK159R-HA-6His]11-4/H2AvD-GFP chromosomes show disorganized chromatinstructure but no increase in chromosome volume comparedto control H2AvD-GFP/TM3 chromosomes (C). Bars, 20 mm.(E) Volume quantification of P[w1, UAS-His1-dsRNA-8-4]/1;da-GAL4, H2AvD-GFP/1 and control da-GAL4, H2AvD-GFP/1chromosomes. (F) Volume quantification of P[w1, eyGAL4],P[w1, UAS-ISWIK159R-HA-6His]11-4/H2AvD-GFP and controlH2AvD-GFP/TM3 chromosomes. (G) Quantification of thechange in chromatin compaction relative to control, estab-

lished by calculating the ratio of volume to DNA for each nu-cleus. The ratio of control samples was normalized to 1. (H–J)A magnification of arms from H2AvD-GFP/TM3 chromosomes(H), P[w1, UAS-His1-dsRNA-8-4]/1; da-GAL4, H2AvD-GFP/1chromosomes (I) and P[w1, eyGAL4], P[w1, UAS-ISWIK159R-HA-6His]11-4/H2AvD-GFP chromosomes (J). Bars, 5 mm.

Drosophila ISWI Regulates H1 Assembly 667

of H1 during replication-coupled chromatin assemblyhas a particularly dramatic effect on chromatin com-paction. After chromatin has been assembled, thecontinued association of H1 with chromosomes, whileimportant, appears to have more subtle effects onchromosome structure.

An independent analysis of phenotypes resultingfrom the knockdown of Drosophila His1 by RNAi wasrecently reported (Lu et al. 2009). Consistent with ourdata, the authors of this study found that histone H1 isessential for Drosophila development. However, theyobserved relatively mild defects in salivary gland poly-tene chromosome structure following H1 knockdown.These defects appear similar to the weakest phenotypeswe observed following H1 knockdown (Figure 1C),which may reflect differences in the extent of H1knockdown achieved in our studies. On the basis ofthe analysis of fixed polytene chromosome squashesfollowing H1 depletion, Lu et al. (2009) concluded thatH1 is required for the alignment of sister chromatids inpolytene chromosomes. Although we observed an evenstronger disruption of the banding pattern of polytenechromosome squashes following H1 knockdown, werarely observed such defects via live analysis. Our datatherefore argue against a major role for H1 in sisterchromatid alignment and illustrate the importance ofusing live analysis to study factors involved in the reg-ulation of higher-order chromatin structure.

The incorporation of H1 during replication-coupledchromatin assembly increases the average distancebetween nucleosomes, thus leading to a decrease ingenomewide nucleosome density (Woodcock et al.2006). Accordingly, we observed a significant decreasein NRL following H1 knockdown (Figure 3B). Bycontrast, the loss of ISWI function leads to a dramaticreduction in the level of H1 associated with chromo-somes without causing obvious changes in NRL (Coronaet al. 2007). These data strongly suggest that ISWIpromotes the association of H1 with salivary glandpolytene chromosomes via a replication-independentmechanism. It remains possible that an additional rolefor ISWI in replication-coupled H1 assembly escapeddetection in our genetic studies due to the failure tocompletely eliminate ISWI function during the stages ofsalivary gland development when the bulk of DNAreplication occurs. Further experiments, including theanalysis of fast-acting conditional ISWI alleles, will berequired to address this issue.

How does ISWI promote the association of H1 withchromatin? By altering the structure, accessibility orfluidity of chromatin, ISWI may facilitate the binding ofH1 to chromatin during dynamic exchange. Consistentwith this possibility, we found that inhibitors of oxidativephosphorylation lead to the dissociation of H1 frompolytene chromosomes accompanied by its accumula-tion in the nucleoplasm. Alternatively, ISWI may stabilizethe binding of H1 to chromatin by influencing its

phosphorylation. H1 is phosphorylated in most organ-isms, including Drosophila (Villar-Garea and Imhof2008). In both Tetrahymena and mammals, the phos-phorylation of H1 weakens its association with chroma-tin, leading to an increased frequency of H1 exchange(Dou et al. 2002; Contreras et al. 2003). Thus, ISWI mayindirectly promote the association of H1 with chromatinby altering the level or activity of an H1 kinase orphosphatase.

The chromatin of stem cells is hyperdynamic, withboth histone H1 and other chromatin-associated pro-teins undergoing highly elevated rates of exchange(Meshorer and Misteli 2006). This property ofpluripotent cell types appears to be functionally impor-tant, since a mutant form of H1 that tightly bindschromatin blocks stem cell differentiation (Meshoreret al. 2006). These findings suggest that ISWI and otherfactors that regulate the association of H1 with chroma-tin may play important roles in the regulation of cellularpluripotency and differentiation. This possibility isintriguing in light of recent studies implicating ISWIin both nuclear reprogramming and stem cell self-renewal (Kikyo et al. 2000; Xi and Xie 2005).

Previous studies have shown that the dosage com-pensation machinery antagonizes ISWI function viathe acetylation of its nucleosome substrate on H4K16(Corona et al. 2002; Shogren-Knaak et al. 2006). Further-more, increased linker histone exchange has beenobserved in active chromatin enriched in core histoneacetylation (Misteli et al. 2000). It is therefore temptingto speculate that the dynamic association of H1 withchromatin is modulated by the interplay of chromatin-remodeling and -modifying enzymes, thus providing astraightforward mechanism for creating rapid, readilyreversible changes in higher-order chromatin structureand gene expression. Further work will be required totest this hypothesis and clarify the molecular mecha-nisms that regulate the association of H1 with chromatinin vivo.

We thank the Bloomington Stock Center for the strains and GrantHartzog, Susan Strome, Rohinton Kamakaka, and the members of ourlaboratories for numerous helpful discussions. This work was sup-ported by National Institutes of Health grant GM49883 (to J.W.T.).

LITERATURE CITED

Blank, T. A., and P. B. Becker, 1995 Electrostatic mechanism of nu-

cleosome spacing. J. Mol. Biol. 252: 305–313.Bouazoune, K., and A. Brehm, 2006 ATP-dependent chromatin re-

modeling complexes in Drosophila. Chromosome Res. 14: 433–449.

Brand, A. H., A. S. Manoukian and N. Perrimon, 1994 Ectopic ex-

pression in Drosophila. Methods Cell Biol. 44: 635–654.Brown, D. T., 2003 Histone H1 and the dynamic regulation of chro-

matin function. Biochem. Cell Biol. 81: 221–227.Carruthers, L. M., J. Bednar, C. L. Woodcock and J. C. Hansen,

1998 Linker histones stabilize the intrinsic salt-dependent fold-

ing of nucleosomal arrays: mechanistic ramifications for higher-order chromatin folding. Biochemistry 37: 14776–14787.

668 G. Siriaco et al.

Catez, F., T. Ueda and M. Bustin, 2006 Determinants of histoneH1 mobility and chromatin binding in living cells. Nat. Struct.Mol. Biol. 13: 305–310.

Clarkson, M., and R. Saint, 1999 A His2AvDGFP fusion gene com-plements a lethal His2AvD mutant allele and provides an in vivomarker for Drosophila chromosome behavior. DNA Cell Biol. 18:457–462.

Contreras, A., T. K. Hale, D. L. Stenoien, J. M. Rosen, M. A. Manciniet al., 2003 The dynamic mobility of Histone H1 is regulated bycyclin/CDK phosphorylation. Mol. Cell. Biol. 23: 8626–8636.

Corona, D. F., C. R. Clapier, P. B. Becker and J. W. Tamkun,2002 Modulation of ISWI function by site-specific histone acet-ylation. EMBO Rep. 3: 242–247.

Corona, D. F., G. Siriaco, J. A. Armstrong, N. Snarskaya, S. A.McClymont et al., 2007 ISWI regulates higher-order chroma-tin structure and histone H1 assembly in vivo. PLoS Biol. 5: e232.

Deuring, R., L. Fanti, J. A. Armstrong, M. Sarte, O. Papoulas et al.,2000 The ISWI chromatin-remodeling protein is required forgene expression and the maintenance of higher order chromatinstructure in vivo. Mol. Cell 5: 355–365.

Dou, Y., J. Bowen, Y. Liu and M. A. Gorovsky, 2002 Phosphorylationand an ATP-dependent process increase the dynamic exchange ofH1 with chromatin. J. Cell Biol. 158: 1161–1170.

Fan, Y., T. Nikitina, J. Zhao, T. J. Fleury, R. Bhattacharyya et al.,2005 Histone H1 depletion in mammals alters global chroma-tin structure but causes specific changes in gene regulation. Cell123: 1199–1212.

Gerber, M., J. C. Eissenberg, S. Kong, K. Tenney, J. W. Conawayet al., 2004 In vivo requirement of the RNA polymerase II elon-gation factor elongin A for proper gene expression and develop-ment. Mol. Cell. Biol. 24: 9911–9919.

Godde, J. S., and K. Ura, 2008 Cracking the enigmatic linker his-tone code. J. Biochem. 143: 287–293.

Khochbin, S., 2001 Histone H1 diversity: bridging regulatory sig-nals to linker histone function. Gene 271: 1–12.

Kikyo, N., P. A. Wade, D. Guschin, H. Ge and A. P. Wolffe,2000 Active remodeling of somatic nuclei in egg cytoplasmby the nucleosomal ATPase ISWI. Science 289: 2360–2362.

Leenders, H. J., A. Kemp, J. F. Koninkx and J. Rosing,1974 Changes in cellular ATP, ADP and AMP levels followingtreatments affecting cellular respiration and the activity of cer-tain nuclear genes in Drosophila salivary glands. Exp. Cell Res.86: 25–30.

Lever, M. A., J. P. H. Th’ng, X. Sun and M. J. Hendzel, 2000 Rapidexchange of histone H1.1 on chromatin in living human cells.Nature 408: 873–876.

Lu, X., S. N. Wontakal, A. V. Emelyanov, P. Morcillo, A. Y. Konevet al., 2009 Linker histone H1 is essential for Drosophila devel-opment, the establishment of pericentric heterochromatin, and anormal polytene chromosome structure. Genes Dev. 23: 452–465.

Lusser, A., D. L. Urwin and J. T. Kadonaga, 2005 Distinct activitiesof CHD1 and ACF in ATP-dependent chromatin assembly. Nat.Struct. Mol. Biol. 12: 160–166.

Maier, V. K., M. Chioda and P. B. Becker, 2008 ATP-dependentchromatosome remodeling. Biol. Chem. 389: 345–352.

Maresca, T. J., B. S. Freedman and R. Heald, 2005 Histone H1 isessential for mitotic chromosome architecture and segregationin Xenopus laevis egg extracts. J. Cell Biol. 169: 859–869.

Meshorer, E., and T. Misteli, 2006 Chromatin in pluripotent em-bryonic stem cells and differentiation. Nat. Rev. Mol. Cell. Biol. 7:540–546.

Meshorer, E., D. Yellajoshula, E. George, P. J. Scambler, D. T.Brown et al., 2006 Hyperdynamic plasticity of chromatin pro-teins in pluripotent embryonic stem cells. Dev. Cell 10: 105–116.

Misteli, T., A. Gunjan, R. Hock, M. Bustin and D. T. Brown,2000 Dynamic binding of histone H1 to chromatin in livingcells. Nature 408: 877–881.

Ner, S. S., and A. A. Travers, 1994 HMG-D, the Drosophila mela-nogaster homologue of HMG 1 protein, is associated with earlyembryonic chromatin in the absence of histone H1. EMBO J.13: 1817–1822.

Robinson, P. J., and D. Rhodes, 2006 Structure of the ‘30 nm’ chro-matin fibre: a key role for the linker histone. Curr. Opin. Struct.Biol. 16: 336–343.

Rodman, T. C., 1967 DNA replication in salivary gland nuclei of Dro-sophila melanogaster at successive larval and prepupal stages. Ge-netics 55: 375–386.

Routh, A., S. Sandin and D. Rhodes, 2008 Nucleosome repeatlength and linker histone stoichiometry determine chromatin fi-ber structure. Proc. Natl. Acad. Sci. USA 105: 8872–8877.

Shogren-Knaak, M., H. Ishii, J. M. Sun, M. J. Pazin, J. R. Davie et al.,2006 Histone H4–K16 acetylation controls chromatin structureand protein interactions. Science 311: 844–847.

Tsukiyama, T., C. Daniel, J. Tamkunand C. Wu, 1995 ISWI, a mem-ber of the SWI2/SNF2 ATPase family, encodes the 140 kDa sub-unit of the nucleosome remodeling factor. Cell 83: 1021–1026.

Villar-Garea, A., and A. Imhof, 2008 Fine mapping of posttransla-tional modifications of the linker histone H1 from Drosophilamelanogaster. PLoS ONE 3: 1–12.

Woodcock, C. L., A. I. Skoultchi and Y. Fan, 2006 Role of linkerhistone in chromatin structure and function: H1 stoichiometryand nucleosome repeat length. Chromosome Res. 14: 17–25.

Xi, R., and T. Xie, 2005 Stem cell self-renewal controlled by chroma-tin remodeling factors. Science 310: 1487–1489.

Communicating editor: F. Winston

Drosophila ISWI Regulates H1 Assembly 669

Supporting Information

http://www.genetics.org/cgi/content/full/genetics.109.102053/DC1

Drosophila ISWI Regulates the Association of Histone H1 with

Interphase Chromosomes in Vivo  

Giorgia Siriaco, Renate Deuring, Mariacristina Chioda, Peter B. Becker

and John W. Tamkun

Copyright © 2009 by the Genetics Society of America

DOI: 10.1534/genetics.109.102053

G. Siriaco et al. 2 SI

Figure S1.—Histone H1 does not affect H4 acetylation or H3K9 methylation. (A-D) Polytene

chromosomes were dissected from larval salivary glands and fixed with 6mM MgCl2/1% citric acid/ 1% TritonX-100. (A, C) Chromosomes were stained with an antibody recognizing tetra-acetylated histone H4. Levels of H4 acetylation are comparable in da-GAL4/+ (A) and P[w+, UAS-his1-dsRNA-8-4]/+; da-GAL4/+ (C) chromosomes. (B, D) Chromosomes were stained with an antibody recognizing H3K9me3. Levels of H3K9me3 are comparable in da-GAL4/+ (B) and P[w+, UAS-His1-dsRNA-8-4]/+; da-GAL4/+ (D) chromosomes. Arrowheads indicate the chromocenter. The images were captured with comparable exposure times.