An unusual genomic position effect on Drosophila white gene expression: pairing dependence, interactions with zeste, and molecular analysis of revertants

The Differences Between Cis- and Trans-Gene Inactivation Caused by Heterochromatin in Drosophila

Position-effect variegation (PEV) is the epigenetic disruption of gene expression near the de novo-formed euchromatin-heterochromatin border. Heterochromatic cis-inactivation may be accompanied by the trans-inactivation of genes on a normal homologous chromosome in trans-heterozygous combination with a PEV-inducing rearrangement. We characterize a new genetic system, inversion In(2)A4, demonstrating cis-acting PEV as well as trans-inactivation of the reporter transgenes on the homologous nonrearranged chromosome. The cis-effect of heterochromatin in the inversion results not only in repression but also in activation of genes, and it varies at different developmental stages. While cis-actions affect only a few juxtaposed genes, trans-inactivation is observed in a 500-kb region and demonstrates а nonuniform pattern of repression with intermingled regions where no transgene repression occurs. There is no repression around the histone gene cluster and in some other euchromatic sites. trans-Inactivation is accompanied by dragging of euchromatic regions into the heterochromatic compartment, but the histone gene cluster, located in the middle of the trans-inactivated region, was shown to be evicted from the heterochromatin. We demonstrate that trans-inactivation is followed by de novo HP1a accumulation in the affected transgene; trans-inactivation is specifically favored by the chromatin remodeler SAYP and prevented by Argonaute AGO2.

Identifying genes that interact with Drosophila presenilin and amyloid precursor protein

The gamma-secretase complex is involved in cleaving transmembrane proteins such as Notch and one of the genes targeted in Alzheimer's disease known as amyloid precursor protein (APP). Presenilins function within the catalytic core of gamma-secretase, and mutated forms of presenilins were identified as causative factors in familial Alzheimer's disease. Recent studies show that in addition to Notch and APP, numerous signal transduction pathways are modulated by presenilins, including intracellular calcium signaling. Thus, presenilins appear to have diverse roles. To further understand presenilin function, we searched for Presenilin-interacting genes in Drosophila by performing a genetic modifier screen for enhancers and suppressors of Presenilin-dependent Notch-related phenotypes. We identified 177 modifiers, including known members of the Notch pathway and genes involved in intracellular calcium homeostasis. We further demonstrate that 53 of these modifiers genetically interacted with APP. Characterization of these genes may provide valuable insights into Presenilin function in development and disease.

Chromatin structure and the regulation of gene expression: the lessons of PEV in Drosophila

Position-effect variegation (PEV) was discovered in 1930 in a study of X-ray-induced chromosomal rearrangements. Rearrangements that place euchromatic genes adjacent to a region of centromeric heterochromatin give a variegated phenotype that results from the inactivation of genes by heterochromatin spreading from the breakpoint. PEV can also result from P element insertions that place euchromatic genes into heterochromatic regions and rearrangements that position euchromatic chromosomal regions into heterochromatic nuclear compartments. More than 75 years of studies of PEV have revealed that PEV is a complex phenomenon that results from fundamental differences in the structure and function of heterochromatin and euchromatin with respect to gene expression. Molecular analysis of PEV began with the discovery that PEV phenotypes are altered by suppressor and enhancer mutations of a large number of modifier genes whose products are structural components of heterochromatin, enzymes that modify heterochromatic proteins, or are nuclear structural components. Analysis of these gene products has led to our current understanding that formation of heterochromatin involves specific modifications of histones leading to the binding of particular sets of heterochromatic proteins, and that this process may be the mechanism for repressing gene expression in PEV. Other modifier genes produce products whose function is part of an active mechanism of generation of euchromatin that resists heterochromatization. Current studies of PEV are focusing on defining the complex patterns of modifier gene activity and the sequence of events that leads to the dynamic interplay between heterochromatin and euchromatin.

Genetic factors controlling white gene expression of the transposon A(R) 4-24 at a telomere in Drosophila melanogaster

The position effect of the AR 4-24 P[white, rosy] transposon was studied at cytological position 60F. Three copies of the transposon (within approximately 50-kb region) resulted in a spatially restricted pattern of white variegation. This pattern was modified by temperature and by removal of the Y chromosome, suggesting that it was due to classical heterochromatin-induced position effect variegation (PEV). In contrast with classical PEV, extra dose of the heterochromatin protein 1 (HP1) suppressed white variegation and one dose enhanced it. The effect of Pc-G, trx-G, and other PEV suppressors was also tested. It was found that E(Pc)1, TrlR85, and mutations of Su(z)2C relieve A(R) 4-24-silencing and z1 enhances it. To explain the results obtained with these modifiers, it is proposed that PEV and telomeric position effect can counteract each other at this particular cytological site.

The effects of ectopic white and transformer expression on Drosophila courtship behavior

The sex determining genes of Drosophila males and females function to establish the potential for sex-specific behaviors. Previous studies suggest that ectopic GAL4-directed misexpression of the female-specific isoform of the sex-determining gene transformer (tra) in specific sub-domains of an otherwise male brain can lead to bisexual courtship behavior, thus identifying brain domains that may mediate sex-specific behavior. However, expression of mini-white, the marker gene used in both P[GAL4] and P[UAS(G)] constructs, also induces males to court other males, questioning whether GAL4-mediated tra expression alone can induce bisexual behavior. Here we demonstrate the consequences of inducing mutations in the mini-white genes within P[GAL4] and P[UAS(G)] constructs to generate flies in which a white mutant phenotype is revealed. In these mini-white mutant strains, P[GAL4]-mediated transformer expression alone is both sufficient and necessary to generate bisexual behavior. In addition, using RT-PCR, we reveal the presence of female transcripts of doublesex and fruitless in the brains of otherwise male (XY) flies exhibiting P[GAL4]-directed tra-expression, demonstrating that P[GAL4]-directed tra is functional at the molecular level. We conclude that P[GAL4]-directed misexpression of tra is responsible for the bisexual behavior previously described and that this is mediated via sex-specific splicing of dsx and fru. Our results support the validity of such strategies for identifying regions of the fly brain that underlie sex-specific behaviors.

An analysis of transvection at the yellow locus of Drosophila melanogaster

Studies of a wide variety of organisms have shown that homologous sequences can exert a significant impact on each other, resulting in changes in gene sequence, gene expression, chromatin structure, and global chromosome architecture. Our work has focused on transvection, a process that can cause genes to be sensitive to the proximity of a homologue. Transvection is seen at the yellow gene of Drosophila, where it mediates numerous cases of intragenic complementation. In this article, we describe two approaches that have characterized the process of transvection at yellow. The first entailed a screen for mutations that support intragenic complementation at yellow. The second involved the analysis of 53 yellow alleles, obtained from a variety of sources, with respect to complementation, molecular structure, and transcriptional competence. Our data suggest two ways in which transvection may be regulated at yellow: (1) a transcriptional mechanism, whereby the ability of an allele to support transvection is influenced by its transcriptional competency, and (2) a structural mechanism, whereby the pairing of structurally dissimilar homologues results in conformational changes that affect gene expression.

Two modes of transvection: enhancer action in trans and bypass of a chromatin insulator in cis

Ed Lewis introduced the term "transvection" in 1954 to describe mechanisms that can cause the expression of a gene to be sensitive to the proximity of its homologue. Transvection since has been reported at an increasing number of loci in Drosophila, where homologous chromosomes are paired in somatic tissues, as well as at loci in other organisms. At the Drosophila yellow gene, transvection can explain intragenic complementation involving the yellow2 allele (y2). Here, transvection was proposed to occur by enhancers of one allele acting in trans on the promoter of a paired homologue. In this report, we describe two yellow alleles that strengthen this model and reveal an unexpected, second mechanism for transvection. Data suggest that, in addition to enhancer action in trans, transvection can occur by enhancer bypass of a chromatin insulator in cis. We propose that bypass results from the topology of paired genes. Finally, transvection at yellow can occur in genotypes not involving y2, implying that it is a feature of yellow itself and not an attribute of one particular allele.

white+ transgene insertions presenting a dorsal/ventral pattern define a single cluster of homeobox genes that is silenced by the polycomb-group proteins in Drosophila melanogaster

We used the white gene as an enhancer trap and reporter of chromatin structure. We collected white+ transgene insertions presenting a peculiar pigmentation pattern in the eye: white expression is restricted to the dorsal half of the eye, with a clear-cut dorsal/ventral (D/V) border. This D/V pattern is stable and heritable, indicating that phenotypic expression of the white reporter reflects positional information in the developing eye. Localization of these transgenes led us to identify a unique genomic region encompassing 140 kb in 69D1-3 subject to this D/V effect. This region contains at least three closely related homeobox-containing genes that are constituents of the iroquois complex (IRO-C). IRO-C genes are coordinately regulated and implicated in similar developmental processes. Expression of these genes in the eye is regulated by the products of the Polycomb-group (Pc-G) and trithorax-group (trx-G) genes but is not modified by classical modifiers of position-effect variegation. Our results, together with the report of a Pc-G binding site in 69D, suggest that we have identified a novel cluster of target genes for the Pc-G and trx-G products. We thus propose that ventral silencing of the whole IRO-C in the eye occurs at the level of chromatin structure in a manner similar to that of the homeotic gene complexes, perhaps by local compaction of the region into a heterochromatin-like structure involving the Pc-G products.

A Polycomb and GAGA dependent silencer adjoins the Fab-7 boundary in the Drosophila bithorax complex

The homeotic genes of the Drosophila bithorax complex are controlled by a large cis-regulatory region that ensures their segmentally restricted pattern of expression. A deletion that removes the Frontabdominal-7 cis-regulatory region (Fab-7') dominantly transforms parasegment 11 into parasegment 12. Previous studies suggested that removal of a domain boundary element on the proximal side of Fab-7' is responsible for this gain-of-function phenotype. In this article we demonstrate that the Fab-7' deletion also removes a silencer element, the iab-7 PRE, which maps to a different DNA segment and plays a different role in regulating parasegment-specific expression patterns of the Abd-B gene. The iab-7 PRE mediates pairing-sensitive silencing of mini-white, and can maintain the segmentally restricted expression pattern of a BXD, Ubx/lacZ reporter transgene. Both silencing activities depend upon Polycomb Group proteins. Pairing-sensitive silencing is relieved by removing the transvection protein Zeste, but is enhanced in a novel pairing-independent manner by the zeste' allele. The iab-7 PRE silencer is contained within a 0.8-kb fragment that spans a nuclease hypersensitive site, and silencing appears to depend on the chromatin remodeling protein, the GAGA factor.

Genetic and molecular analysis of a set of unstable white mutants in Drosophila melanogaster

Two related unstable mutants at the white locus of Drosophila melanogaster show different interactions with the zeste1 mutant: one mutated white gene becomes repressed in males, whereas the other is unaffected by z1. By use of Southern blot techniques and by constructing genomic lambda-libraries, molecular analyses of the white regions of these two strains were performed. The results showed a single difference at a site 2.5 kb (kilobases) downstream of the white transcription unit. In both strains, FB (foldback) elements were integrated at this site, but the repressed strain also harboured a 4 kb NOF (Nofretete) element. No other restriction site polymorphisms between the two strains were observed within a 120 kb region surrounding the white gene. The extent of twelve white deletions and twelve white transpositions deriving from these unstable strains was analysed by in situ hybridisation and Southern blot techniques. The results revealed that the distal breakpoint of all aberrations coincided with the insertion site of the mobile elements, but that the centromere proximal breakpoints varied. The mechanisms for the instability and the interaction with the zeste1 mutant are discussed.

Meiotic transvection in fungi

The Neurospora crassa Asm-1+ (ascospore maturation 1) gene encodes an abundant nucleus-localized protein required for formation of female structures and for ascospore maturation. Deletion mutants of Asm-1+ are "ascus-dominant," i.e., when crossed to wild type, neither Asm-1+ nor Asm-1 delta spores mature. To explain this behavior, we considered three models: an effect of reduced dosage of the gene product, failure of internuclear communication, and failure of transvection (regulation dependent on pairing of alleles). We found that for proper regulation of subsequent sexual sporulation, Asm-1+ must be in proximity, probably paired, to its allelic counterpart in the zygote: i.e., transvection must occur. Disruption of pairing causes failure of ascospore progeny to mature. Transvection in Neurospora, unlike in Drosophila, occurs immediately before meiosis, and can be demonstrated between wild-type alleles.

A quantitative measure of the mitotic pairing of alleles in Drosophila melanogaster and the influence of structural heterozygosity

In Drosophila there exist several examples of gene expression that can be modified by an interaction between alleles; this effect is known as transvection. The inference that alleles interact comes from the observations that homologous chromosomes pair in mitotically dividing cells, and that chromosome rearrangements can alter the phenotype produced by a pair of alleles. It is thought that heterozygous rearrangements impede the ability of alleles to pair and interact. However, because the existing data are inconsistent, this issue is not fully settled. By measuring the frequency of site-specific recombination between homologous chromosomes, we show that structural heterozygosity inhibits the pairing of alleles that lie distal to a rearrangement breakpoint. We suggest that some of the apparent conflicts may owe to variations in cell-cycle lengths in the tissues where the relevant allelic interactions occur. Cells with a longer cell cycle have more time to establish the normal pairing relationships that have been disturbed by rearrangements. In support, we show that Minute mutations, which slow the rate of cell division, partially restore a transvection effect that is disrupted by inversion heterozygosity.

White as a reporter gene to detect transcriptional silencers specifying position-specific gene expression during Drosophila melanogaster eye development

The white+ gene was used as a reporter to detect transcriptional silencer activity in the Drosophila genome. Changes in the spatial expression pattern of white were scored in the adult eye as nonuniform patterns of pigmentation. Thirty-six independent P[lacW] transposant lines were collected. These represent 12 distinct pigmentation patterns and probably 21 loci. The spatial pigmentation pattern is due to cis-acting suppression of white+ expression, and the suppression probably depends on cell position rather than cell type. The mechanism of suppression differs from inactivation by heterochromatin. In addition, activation of lacZ in P[lacW] occurs also in specific patterns in imaginal discs and embryos in many of the lines. The expression patterns of white+ and lacZ may reflect the activity of regulatory elements belonging to an endogenous gene near each P[lacW] insertion site. We speculate that these putative POSE (position-specific expression) genes may have a role in pattern formation of the eye as well as other imaginal structures. Three of the loci identified are optomotor-blind, engrailed and invected. teashirt is also implicated as a candidate gene. We propose that this "silencer trap"' may be an efficient way of identifying genes involved in imaginal pattern formation.

Mutations of zeste that mediate transvection are recessive enhancers of position-effect variegation in Drosophila melanogaster

Evidence is presented demonstrating that mutations of zeste, particularly the null state, are strong recessive enhancers of position-effect variegation (PEV) for the white, roughest and Notch loci. The zeste locus encodes a DNA-binding protein that acts as a transcription factor and mediates transvection phenomena at several loci. Its involvement with these seemingly diverse phenomena suggests that the normal zeste product functions in the decondensation of chromatin. A model is presented proposing that zeste is important for opening and stabilizing domains of chromatin, a step in gene determination and the establishment of cell memory. It postulates that chromatin domains that have been structurally modified by chromosomal rearrangement or by insertion of transposable elements are particularly sensitive to the absence or modification of the zeste protein. Such a view unifies the role of zeste in transcription, transvection and PEV.

Regulatory regions of the homeotic gene proboscipedia are sensitive to chromosomal pairing

We have identified regulatory regions of the homeotic gene proboscipedia that are capable of repressing a linked white minigene in a manner that is sensitive to chromosomal pairing. Normally, the eye color of transformants containing white in a P-element vector is affected by the number of copies of the transgene; homozygous flies have darker eyes than heterozygotes. However, we found that flies homozygous for select pb DNA-containing transgenes had lighter eyes than heterozygotes. Several pb DNA fragments are capable of causing this pairing sensitive (PS) negative regulation of white. Two fragments in the upstream DNA of pb, 0.58 and 0.98 kb, are PS; additionally, two PS sites are located in the second intron, including a 0.5-kb region and 49-bp sequence. This phenotype is not observed when two PS sites are located at different chromosomal insertion sites (in trans-heterozygous transgenic animals), indicating that the pb-DNA-mediated repression of white is dependent on the pairing or proximity of the PS regions. The observed phenomenon is similar to transvection in which certain alleles of a gene can complement each other, but only when homologous chromosomes are paired. Interestingly, the intronic PS regions contain positive regulatory sequences for pb, whereas the upstream PS sites contain pb negative regulatory elements.

A genetic analysis of the Suppressor 2 of zeste complex of Drosophila melanogaster

The zeste1 (z1) mutation of Drosophila melanogaster produces a mutant yellow eye color instead of the wild-type red. Genetic and molecular data suggest that z1 achieves this change by altering expression of the wild-type white gene in a manner that exhibits transvection effects. There exist suppressor and enhancer mutations that modify the z1 eye color, and this paper summarizes our studies of those belonging to the Suppressor 2 of zeste complex [Su(z)2-C]. The Su(z)2-C consists of at least three subregions called Psc (Posterior sex combs), Su(z)2 and Su(z)2D (Distal). The products of these subregions are proposed to act at the level of chromatin. Complementation analyses predict that the products are functionally similar and interacting. The alleles of Psc define two overlapping phenotypic classes, the hopeful and hapless. The distinctions between these two classes and the intragenic complementation seen among some of the Psc alleles are consistent with a multidomain structure for the product of Psc. Psc is a member of the homeotic Polycomb group of genes. A general discussion of the Polycomb and trithorax group of genes, position-effect variegation, transvection, chromosome pairing and chromatin structure is presented.

Identification of Polycomb and trithorax group responsive elements in the regulatory region of the Drosophila homeotic gene Sex combs reduced

The Drosophila homeotic gene Sex combs reduced (Scr) is necessary for the establishment and maintenance of the morphological identity of the labial and prothoracic segments. In the early embryo, its expression pattern is established through the activity of several gap and segmentation gene products, as well as other transcription factors. Once established, the Polycomb group (Pc-G) and trithorax group (trx-G) gene products maintain the spatial pattern of Scr expression for the remainder of development. We report the identification of DNA fragments in the Scr regulatory region that may be important for its regulation by Polycomb and trithorax group gene products. When DNA fragments containing these regulatory sequences are subcloned into P-element vectors containing a white minigene, transformants containing these constructs exhibit mosaic patterns of pigmentation in the adult eye, indicating that white minigene expression is repressed in a clonally heritable manner. The size of pigmented and nonpigmented clones in the adult eye suggests that the event determining whether a cell in the eye anlagen will express white occurs at least as early as the first larval instar. The amount of white minigene repression is reduced in some Polycomb group mutants, whereas repression is enhanced in flies mutant for a subset of trithorax group loci. The repressor activity of one fragment, normally located in Scr Intron 2, is increased when it is able to homologously pair, a property consistent with genetic data suggesting that Scr exhibits transvection. Another Scr regulatory fragment, normally located 40 kb upstream of the Scr promoter, silences ectopic expression of an Scr-lacZ fusion gene in the embryo and does so in a Polycomb-dependent manner. We propose that the regulatory sequences located within these DNA fragments may normally mediate the regulation of Scr by proteins encoded by members of the Polycomb and trithorax group loci.

P-element-induced variation in metabolic regulation in Drosophila

Movement of transposable elements have demonstrated to be a cause of genetic variation that is relevant to quantitative characters in Drosophila. Here a particular class of P-element-induced variation known to be mediated through changes in expression of targeted enzyme-encoding genes is examined. Balancer chromosome and 139 third-chromosome lines of Drosophila melanogaster bearing unique stable P-element insertions in a common genetic background. Lines that were homozygous for second-chromosome P-element insertions were significantly more heterogeneous than control lines in 10 of 16 characters, whereas third-chromosome insertion lines were heterogeneous in 11 of the 16 traits. The average mutational variance per insertion relative to environmental variance (Vm/Ve) was 5.7 x 10(-2), and estimates varied widely across characters. The distributions of mutational effects tended to be skewed, with a longer tail toward high enzyme activities. Mutational effects deviated from a normal distribution in 15 of the 16 traits and significant outlier lines were found in both a positive and negative direction in several characters. Pleiotropic effects of single P-element insertions were quantified by correlation, and, after correcting for simultaneous tests, of the 91 correlations, 37 were significant at the 5% level. The pattern of pleiotropic effects derviated both from the equilibrium genetic correlations quantified in a previous study and from the correlations of mutational effects in a mutation-accumulation experiment, suggesting that multiple forces are at play that shape extant variation.

White gene expression, repressive chromatin domains and homeotic gene regulation in Drosophila

The use of Drosophila chromosomal rearrangements and transposon constructs involving the white gene reveals the existence of repressive chromatin domains that can spread over considerable genomic distances. One such type of domain is found in heterochromatin and is responsible for classical position-effect variegation. Another type of repressive domain is established, beginning at specific sequences, by complexes of Polycomb Group proteins. Such complexes, which normally regulate the expression of many genes, including the homeotic loci, are responsible for silencing, white gene variegation, pairing-dependent effects and insertional targeting.

Unusual properties of regulatory DNA from the Drosophila engrailed gene: three "pairing-sensitive" sites within a 1.6-kb region

We have previously shown that a 2-kb fragment of engrailed DNA can suppress expression of a linked marker gene, white, in the P element vector CaSpeR. This suppression is dependent on the presence of two copies of engrailed DNA-containing P elements (P[en]) in proximity in the Drosophila genome (either in cis or in trans). In this study, the 2-kb fragment was dissected and found to contain three fragments of DNA which could mediate white suppression [called "pairing-sensitive sites" (PS)]. A PS site was also identified in regulatory DNA from the Drosophila escargot gene. The eye colors of six different P[en] insertions in the escargot gene suggest an interaction between P[en]-encoded and genome-encoded PS sites. I hypothesize that white gene expression from P[en] is repressed by the formation of a protein complex which is initiated at the engrailed PS sites and also requires interactions with flanking genomic DNA. Genes were sought which influence the function of PS sites. Mutations in some Polycomb and trithorax group genes were found to affect the eye color from some P[en] insertion sites. However, different mutations affected expression from different P[en] insertion sites and no one mutation was found to affect expression from all P[en] insertion sites examined. These results suggest that white expression from P[en] is not directly regulated by members of the Polycomb and trithorax group genes, but in some cases can be influenced by them. I propose that engrailed PS sites normally act to promote interactions between distantly located engrailed regulatory sites and the engrailed promoter.

The determined state of white expression in the Drosophila eye is modified by zeste1 in the wzm family of mutants

Analysis of the whitezeste mottled (wzm) mutant family suggests that the zeste gene product functions in establishing and stabilizing a transcriptionally active chromatin domain for white locus expression. The z1 mutation reduces expression of paired or proximate copies of white, while single or unpaired copies maintain wild-type levels of expression. The wzm mutation, caused by the insertion of the retrotransposon BEL into the 5' intron of white, alters the zeste-white interaction to produce a mottled eye phenotype in hemizygous z1 wzm males. We have determined the molecular structure of four wzm derivatives. wzl results from the insertion of an additional transposable element into the 5' regulatory region of white. wzvl is a deletion of sequences upstream of the white locus. Two others, whalo and wcres, result from the transposition of wzm plus the entire verticals-roughest region into heterochromatin near the tip of chromosome 3L. They variegate for roughest but not for white; rather, the z1 effect on wzm now causes white expression to become non-autonomous and non-clonal. The analysis of these five mutations shows that the neomorphic zeste1 product, in combination with structural changes imposed by transposons and intercalary heterochromatin, modifies the determination and stability of white expression. We propose that the normal zeste product functions as part of a complex that stimulates transcription by changing chromatin conformation to establish and maintain transcriptionally active domains. The unpairing of homologs is proposed to be one of the initial results of conformational change, providing an explanation for the role of zeste in transvection.

Isolation of silencer-containing sequences causing a tissue-specific position effect on alcohol dehydrogenase expression in Drosophila melanogaster

A transient expression assay has been used to investigate the cause of a tissue-specific position effect on Adh expression from a transgene insertion in Drosophila. A 15.4-kb genomic clone containing the 3.2-kb Adh insert along with flanking regions of genomic DNA is expressed in this assay in a tissue-specific pattern resembling the abnormal expression pattern of the position effect. The 3.2-kb Adh insert is expressed normally without the flanking sequences. A silencer element is located upstream of the Adh gene within a 2-kb fragment that acts in both orientations and at a distance of at least 6.5 kb from the larval Adh promoter to suppress ADH expression in a nontissue specific fashion. The DNA sequence of the 2-kb fragment indicates that it is a noncoding region. A 17-bp sequence is repeated within this region and may be associated with the silencer activity, since subclones from the 2-kb fragment, each containing one of the repeated regions, both retain full silencer activity. This silencer fails to suppress expression from an alpha 1-tubulin promoter-LacZ fusion construct or an hsp70 promoter-Adh fusion construct. In addition to the silencer, another element is located downstream of the Adh gene that produces a higher level of anterior than posterior midgut expression. These results suggest that the 5' silencer and the 3' element act together to create the tissue specific position effect characteristic of the GC-1 line.

The su(Hw) protein insulates expression of the Drosophila melanogaster white gene from chromosomal position-effects

Mutations in the suppressor of Hairy-wing [su(Hw)] locus reverse the phenotype of a number of tissue-specific mutations caused by insertion of a gypsy retrotransposon. The su(Hw) gene encodes a zinc finger protein which binds to a 430 bp region of gypsy shown to be both necessary and sufficient for its mutagenic effects. su(Hw) protein causes mutations by inactivation of enhancer elements only when a su(Hw) binding region is located between these regulatory sequences and a promoter. To understand the molecular basis of enhancer inactivation, we tested the effects of su(Hw) protein on expression of the mini-white gene. We find that su(Hw) protein stabilizes mini-white gene expression from chromosomal position-effects in euchromatic locations by inactivating negative and positive regulatory elements present in flanking DNA. Furthermore, the su(Hw) protein partially protects transposon insertions from the negative effects of heterochromatin. To explain our current results, we propose that su(Hw) protein alters the organization of chromatin by creating a new boundary in a pre-existing domain of higher order chromatin structure. This separates enhancers and silencers distal to the su(Hw) binding region into an independent unit of gene activity, thereby causing their inactivation.