Mutations in the Drosophila melanogaster Gene Encoding S-adenosylmethionine Suppress Position-Effect Variegation

In Drosophila melanogaster, the study of trans-acting modifier mutations of positioneffect variegation and Polycomb group (PC-G) genes have been useful tools to investigate genes involved in chromatin structure. We have cloned a modifier gene, Suppesssm of zeste 5 (Su(z)S), which encodes Sadenosylmethionine synthetase, and we present here molecular results and data concerning its expression in mutants and genetic interactions. The mutant alleles Su(z)5, 1(2)R23 and 1(2)M6 show suppression of wm4 and also of two white mutants induced by roo element insertions in the regulatory region i.e., w" (in combination with z') and d'. Two of the Su(z)S alleles, as well as a deletion of the gene, also act as enhancers of PoZycomb by increasing the size of sex combs on midleg. The results suggest that Su(z)5 is connected with regulation of chromatin structure. The enzyme Sadenosylmethionine synthetase is involved in the synthesis of Sadenosylmethionine, a methyl group donor and also, after decarboxylation, a propylamino group donor in the bio-synthesis of polyamines. Our results from HPLC analysis show that in ovaries from heterozygous Su(z)5 mutants the content of spermine is significantly reduced. Results presented here suggest that polyamines are an important molecule class in the regulation of chromatin structure. 0 NE approach to understanding the control of differential gene expression has been to study trans acting regulatory factors in Drosophila melanogaster, This is most often done through dominant mutations that alter the expression of a given genetic model system. Two genetic model systems, the zeste-white interaction and positioneffect variegation, have easily scored phenotypic changes and have consequently been subject to extensive studies. Recently, the effects of the PC-G (PoZycomb group) genes on the regulation of homeotic gene expression have also become a widely used assay to study gene regulation. Results indicate that the role in gene regulation of these three systems may be caused by a common molecular mechanism, namely changes of chromatin structure (reviewed by PIRROTTA 1991; ORLANDO and PARO 1995). The basis of the zeste-white system is a particular mutant allele of zeste, namely z'. This mutant represses the eye-specific expression of the white gene, resulting in a yellow eye color instead of the wild-type red eye color (GANS 1953; BINGHAM and ZACHAR 1985). This repression requires two copies of the white gene in close proximity, which occurs either by homologous chromosome pairing or by tandem duplications (JACK and JUDD 1979; reviewed by Wu and GOLDBERG 1989; PIRROTTA 1990). The product of the zeste gene is a nuclear DNAbinding protein found associated with over 60 specific Corresponding authm: &sa Rasmuson-Lestander, Department of Genetics, Umei University, S901 87 Umei, Sweden. E-mail: [email protected] Genetics 143 887-896 (June, 1996) loci on the polytene chromosomes including the major homeotic complexes (PIRROTTA et al. 1988; RASTELLI et al. 1993). A multimerization of the Zeste protein is thought to be required for chromosome pairing. The zeste' mutation has been shown to cause formation of larger Zeste aggregates than the wild-type gene and the repression of white by zeste' is correlated with the hyperaggregation of the zeste gene product (CHEN and PIRROTTA 1993). In the early 1970s, the first mutagenesis screen for dominant suppressors of z' repression of a specific white allele, w", was presented (KALISCH and RA~MUSON 1974). In this screen seven dominant modifiers of z1 w'" were isolated. Of these, Suppressor ojzeste 2 and Enhancer of zeste 1 have been extensively studied. Su(z)2 suppresses the z'-mediated repression of white not only in z1 wis males but also in z1 w+ females, suppressing the yellow eye color to wild-type red. E(z)l gives the opposite effect; in combination with z1 w'" males or z1 w' females, E(z)l enhances the repression and gives a light yellow eye color. Both these genes have been cloned (BRUNK et al. 1991; JONES and GELBART 1993). E(%) is considered to be a member of the PC group (PHILLIPS and SHEARN 1990), while Su(z)2 is at least functionally related to some of the members in the PcG on the basis of its interaction with Postm'or sex combs (Psc) and Sex combs on midleg (Scm) (ADLER et al. 1989; WU et al. 1989). The characterization of the two genes have led to a mechanistic connection between the zesdwhite interaction and the regulation of genes by the PcG gene products. It has been shown by RASTELLI et al. (1993) that the Psc and Su(z)2 proteins bind to -80888 J. Larsson, J. Zhang and A. Rasmuson-Lestander 90 locations on salivary gland polytene chromosomes, and a comparison of these locations with the chromosomal binding sites for Zeste, Polycomb and Polyhomeotic shows that the proteins colocalize at a large number of sites, suggesting that they act cooperatively in regulation of target genes. The PC-G genes encode proteins that maintain a repressed state of the homeotic segment identity genes in the ANT-C and BX-C gene complexes (FRANKE et al. 1992). The correct expression of PC-G genes together with the tm’thorux group genes are of vital importance for the correct maintenance of the determined state of cells, and thus necessary for the proper development of an organized body plan. The mechanism for this maintenance of a determined state of gene expression has been proposed to be the formation of closed heterochromatin-like structures (PARO and HOGNESS 1991). The main arguments for this are molecular similarities and shared physiological properties between PC-G proteins and modifiers of position-effect variegation (reviewed by ORLANDO and PARO 1995). Position-effect variegation (PEV) was first characterized by MULLER (1930) as the variable, but heritable, repression of euchromatic genes when rearrangements juxtapose them to heterochromatin. For example, when an inversion places a white+ gene next to heterochromatin, the inactivation is seen as a variegating eye with pigmented and nonpigmented clones of ommatidia (for reviews, see HENIKOFF 1990; REUTER and SPIERER 1992). The most commonly used model system for PEV is the wm4 inversion described by MULLER (1930). Different mechanisms have been proposed to explain the PEV phenomenon, where the model of a multimeric assembly of chromatin, and its impact on transcription activity is fundamental. Lately however, strong evidence for more complex mechanisms has been published, including nuclear compartmentalization and physical alterations of DNA (reviewed by KARPEN 1994). Correlation between gene regulation caused by the PC-G genes and by PEV was indicated when a 52-amino acid-long chromo domain was shown to be present in both the Polycomb protein and the heterochromatin binding protein 1 (HP1) (PARO and HOG NESS 1991). The HP1 protein is encoded by the Su(uar)205 gene and mutants act as dominant modifiers of PEV. Further correlation between the two systems has been demonstrated by the fact that a regulatory region from the phgene, which responds to PC-G repression, can induce variegation of an adjacent white gene (FAWARQUE and DURA 1993). The results reported here seem to link these three model systems together. We have studied the Su(z)5 gene and found that Su(z)5 mutations, apart from being dominant suppressors of z’ w”, also act as dominant enhancers of Pcand as suppressors of PEV. In this paper we present evidence that the mutant Su(z)5 is caused by changes in the gene encoding the enzyme S-adenosylmethionine synthetase. We also argue that the dominant suppressor effect on z’ wi’ and on wm4, together with the enhancer effect on PC, is likely to be caused by a decrease in spermine concentration and that polyamines are important molecules in the assembly of higher order chromatin structure. MATERIALS AND METHODS Drosophila stocks and culture: For a description of mutants used see LINDSLEY and GRELL (1968) and LINDSLEY and ZIMM (1992). Stocks were kindly provided by C. CAGGESE (1(2)M, 1(2)R, Z(2)PM and Df(2L)PM stocks), M. M. GREEN (Df(ZL)(net, lgl)78:30 and z’ wTt”), B. MECHLER (Ilf(2L)1(2)gl net3) and the Umei Drosophila Stock Center. All eye color comparisons were made on parallel cultures of equal age. All crosses were repeated at least twice. Crosses were made in vials with potatomash-yeast-agar medium at temperatures indicated in the text. We have found that several balancer chromosomes contain modifiers of PEV that can interfere with the experiments. Therefore, before examining the suppression of PEV, the mutants were first rebalanced to Df(ZL)S2 (2L:21C6 D1;22A6-B1), to homogenize the second chromosome and to use the Df(2L)SZ chromosome as a control. The balanced strains M//Df(ZL)SZwhere Mstands for su(z)5, 1(2)M6, 1(2)R23 or Df(2L)PM44 were checked every generation to prevent propagation of crossovers. DNA isolation and Southern analysis: Genomic DNA was prepared according to the protocols described in SAMBROOK et al. (1989). DNA was cut with restriction enzymes indicated in Figure 1, separated on 0.8% agarose gels, transferred to GeneScreenPlus filter membranes (DuPont-NEN Research Products Inc.) and hybridized according to the instructions of the manufacturer. poly(A)+ RNA extraction: poly(A) + RNA was extracted using Dynal biomagnetic separation system. Ovaries or testes were frozen in ethanol/C02-ice bath. The frozen tissue was homogenized in 0.1 M Tris-HC1 (pH 8.0), 0.5 M LiC1, 10 mM EDTA, 1% SDS, 5 mM dithiothreitol (DTT). After this step the instructions from the manufacturer, Dynal, were followed. Reverse Northern analysis: poly(A) + RNA from pupae was labeled radioactively as described in SAMBROOK et al. (1989) and used as a probe in hybridizations to overlapping genomic DNA fragments from the 21A-B region (cloned in A) blotted onto GeneScreenPlus filters (DuPont-NEN). Northern analysis: Approximately 0.31.0 pg of poly(A)+ RNA from ovaries were separated on an 1.0% formaldehydeagarose gel as described by HANSSON and LAMBERTSSON (1983) and blotted onto GeneScreenPlus filters (DuPontNEN) using VacuGene Vacuum Blotting System (Pharmacia LKB Biotechnology AB). The filters were prehybridized, hybridized and washed according to the instructions for GeneScreenplus filters. The cDNA clone #10 (LARSSON and RASMUSON-LESTANDER 1994) and a-tubulin (kindly provided by A. LAMBERTSSON) labeled by the random priming technique were used as probes. HPLC analysis: To determine the content of polyamines, ovaries were sonicated in 0.2 M perchloric acid and centrifuged. The supernatant was analyzed for spermine and spermidine content using the reverse phase HPLC method described by SEILER and KNODGEN (1985). Polyamines were determined by separation of the ion pairs formed with 1octanesulfonic acid on a reversed-phase column (Kromasil KR 100-5C18; Eka Nobel; 15 cm X 4.6 mm inside diameter). For each strain double samples were analyzed. Each sample contained 100 ovaries from 5-day-old females (25”) and two aliquots were run separately. Suppression of PEV 889 Enzyme activity assay: Proteins were extracted by grinding 40 ovaries in 400 pl of 2 X extraction buffer [ 100 mM Tris (pH 7.5), 2 mM EDTA, 20% glycerol, 20 mM P-mercaptoethanol, 1 mM DTT). After centrifugation at 13,000 rpm for 5 min, the supernatant was collected and protein concentration was determined (BRADFORD 1976) using the Bic-Rad Protein Assay kit. The AdoMet synthetase activity assay was modified from MUDD et al. (1965). One hundred micrograms protein extract was incubated in 250 p1 reaction buffer [ 100 mM Tris (pH 8.0), 30 mM MgS04, 10 mM KCl, 7.5 pCi ["S]-methionine (Amersham) and 10 mM ATP] at 37" for 30 min. Cdntrol reactions were without ATP. The reactions were stopped by adding 2 ml of ice-cold water. Reactions were loaded on Dowex AG 50W-X2 cation exchange columns (NH4+ form) and columns were washed with 20 ml cold water. The adsorbed S-adenosylmethionine was eluted with 5.0 ml of NH,OH (29.1%). The collected samples were measured by scintillation spectrometry. Protein extractions and activity assays were repeated eight times for each strain. Sequencing: Fragments from the genomic X-clones y361 and y343 (kindly provided by H.-P. LERCH) were subcloned into pUC19 vector using standard techniques. DNA for sequencing was prepared by the Wizard Mini Prep DNA Purification System (Promega Corp.) and sequenced with the dideoxy chain termination technique (SANGER et al. 1977) using the Promega Taq Track sequencing kit and [35S]dATP (Amersham) following instructions from the supplier. Forward and reverse primers for pUC/Ml3 vector were used as well as internal primers (Symbicom). Samples were run at 2.5,6, and 9 hr on a 5% polyacrylamide gel at constant power (75 W).