Genetics, Vol. 161, 733-746, June 2002, Copyright © 2002

Transvection and Silencing of the Scr Homeotic Gene of Drosophila melanogaster

Jeffrey W. Southwortha and James A. Kennisona
a Section on Drosophila Gene Regulation, Laboratory of Molecular Genetics, National Institute of Child Health and Human Development, National Institutes of Health, Bethesda, Maryland 20892-2785

Corresponding author: James A. Kennison, Rm. 3B-331, NIH, Bethesda, MD 20892-2785., jim_kennison{at}nih.gov (E-mail)

Communicating editor: T. C. KAUFMAN


*  ABSTRACT
*TOP
 *ABSTRACT
*MATERIALS AND METHODS
 *RESULTS AND DISCUSSION
 *LITERATURE CITED

The Sex combs reduced (Scr) gene specifies the identities of the labial and first thoracic segments in Drosophila melanogaster. In imaginal cells, some Scr mutations allow cis-regulatory elements on one chromosome to stimulate expression of the promoter on the homolog, a phenomenon that was named transvection by Ed Lewis in 1954. Transvection at the Scr gene is blocked by rearrangements that disrupt pairing, but is zeste independent. Silencing of the Scr gene in the second and third thoracic segments, which requires the Polycomb group proteins, is disrupted by most chromosomal aberrations within the Scr gene. Some chromosomal aberrations completely derepress Scr even in the presence of normal levels of all Polycomb group proteins. On the basis of the pattern of chromosomal aberrations that disrupt Scr gene silencing, we propose a model in which two cis-regulatory elements interact to stabilize silencing of any promoter or cis-regulatory element physically between them. This model also explains the anomalous behavior of the Scx allele of the flanking homeotic gene, Antennapedia. This allele, which is associated with an insertion near the Antennapedia P1 promoter, inactivates the Antennapedia P1 and P2 promoters in cis and derepresses the Scr promoters both in cis and on the homologous chromosome.

THE molecular basis of differential gene activity is one of the fundamental problems of developmental biology. In different cell types, some genes are active while other genes are silent. The determination of segment identity in Drosophila provides a powerful system for the study of factors required for both gene activation and silencing. The homeotic genes in two clusters, the Antennapedia complex (ANTC) and the bithorax complex (BXC), specify the identities of most of the segments of the head and trunk (KAUFMAN et al. 1990 Down; LEWIS 1993 Down, LEWIS 1998 Down). Each of the homeotic (HOM) genes in these complexes encodes homeodomain-containing proteins that control the transcription of downstream target genes (GELLON and MCGINNIS 1998 Down). The protein products of the HOM genes are expressed in precise temporal and spatial patterns that are crucial for proper segment identity. As both loss of function and ectopic expression of HOM genes cause changes in segment identity, it is clear that both activation and silencing are important.

The Sex combs reduced (Scr) gene of the ANTC is required in the head and in the first segment of the thorax, both in the embryo and in the imaginal tissues that differentiate the adult structures (WAKIMOTO and KAUFMAN 1981 Down; STRUHL 1982 Down; PATTATUCCI et al. 1991 Down). The requirements for Scr function in the thorax are particularly useful for studying gene activation and silencing. Each thoracic segment in the larva includes two groups of cells, the imaginal leg discs, that will differentiate the adult legs at metamorphosis. SCR proteins are normally expressed at high levels only in the cells of the first imaginal leg disc; the cells of the second and third imaginal leg discs have little or no expression of SCR proteins (PATTATUCCI and KAUFMAN 1991 Down). A row of characteristic bristles (the sex comb) is present on the first tarsal segment of the first thoracic leg of the adult male. These characteristic bristles, called sex comb teeth, are not differentiated on either the second or the third leg. Loss of Scr function in the first thoracic segment causes a failure to differentiate sex comb teeth (KAUFMAN et al. 1990 Down; PATTATUCCI et al. 1991 Down). In contrast, ectopic expression of SCR proteins in the imaginal cells of the second and third thoracic segments causes the differentiation of sex comb teeth on the second and third legs of adult males (KAUFMAN et al. 1990 Down; PATTATUCCI and KAUFMAN 1991 Down). The appearance of ectopic sex comb teeth, the "extra sex combs" phenotype, has been used to identify proteins that are required for silencing of the Scr gene. These silencing proteins are known collectively as Polycomb (PC) group proteins (JURGENS 1985 Down; KENNISON 1995 Down; SIMON 1995 Down; PIRROTTA 1997 Down, PIRROTTA 1998 Down; VAN LOHUIZEN 1999 Down). Although the PC group proteins are required to maintain silencing of the HOM genes, they are not required to determine the initial patterns of gene repression in the early embryo. In addition to the trans-acting PC group proteins, silencing also requires cis-regulatory elements within or near the target genes. Cis-regulatory elements that silence reporter genes when included in transgenes and respond to PC group mutations have been identified and are called PC group response elements (PREs; reviewed in HAGSTROM and SCHEDL 1997 Down; PIRROTTA 1997 Down, PIRROTTA 1998 Down; MIHALY et al. 1998 Down). There may be multiple types of PREs that interact with different subsets of PC group proteins (RASTELLI et al. 1993 Down; STRUTT and PARO 1997 Down).

Chromosomal integrity also seems to play some role in homeotic gene silencing in Drosophila. Chromosomal rearrangements that disrupt silencing of Scr have been known for many years (TOKUNAGA 1966 Down; PATTATUCCI and KAUFMAN 1991 Down; PATTATUCCI et al. 1991 Down), but the molecular mechanisms responsible for this disruption of silencing are unclear. We have studied the effects of mutations and chromosomal rearrangements on Scr function in both the first thoracic segment (where the gene is normally active) and the second and third thoracic segments (where the gene is normally silenced). On the basis of these results, we propose a model in which proteins bound at a pair of regulatory elements flanking the Scr promoter interact to stabilize silencing.


*  MATERIALS AND METHODS
*TOP
 *ABSTRACT
 *MATERIALS AND METHODS
*RESULTS AND DISCUSSION
 *LITERATURE CITED

Stocks:
Flies were raised on a cornmeal-molasses-yeast-agar-Tegosept medium at 25°. Care was taken with all crosses to avoid overcrowding cultures. While culture conditions, temperature, and second-site modifier mutations can greatly suppress or enhance the extra sex combs phenotype caused by Scr derepression in the second and third leg imaginal discs (KENNISON and RUSSELL 1987 Down), there is no evidence that Scr function in the first leg imaginal discs is affected by either temperature or culture conditions. KENNISON and RUSSELL 1987 Down used 46 duplications that included >99% of the autosomes to search for dosage-dependent modifiers of either the extra sex combs phenotype (caused by Polycomb group mutations, Scr rearrangements, or the Scx allele of Antp) or the numbers of sex comb teeth on first legs (examined both in males with two wild-type Scr alleles and in males with only a single wild-type Scr allele). While several of the duplications appeared to include dosage-dependent modifiers that enhanced or suppressed the extra sex combs phenotypes by 200–300%, only two duplications appeared to have significant effects on the numbers of sex comb teeth on first legs (one duplication reduced the average number of sex comb teeth on the first leg by 23%, while the second duplication increased the average number by 21%). Unless otherwise noted, mutations and chromosome rearrangements in the ANTC are described in LINDSLEY and ZIMM 1992 Down or PATTATUCCI et al. 1991 Down. Chromosome rearrangements associated with the Scr and Antp alleles used in this study are listed in Table 1. We usually refer to the inversions, transpositions, and translocations by the Scr or Antp mutant that they carry, rather than by the complete aberration name. Because of the haplolethality of Tpl (LINDSLEY et al. 1972 Down), the Df(3R)Tpl9 chromosome used in all of the experiments described here carries a duplication that includes the Tpl locus [Dp(3;3)Dfdrv1; HAZELRIGG and KAUFMAN 1983]. To determine the approximate location of the breakpoint of Df(3R)LIN2, we used PCR on single embryos from the deficiency stock (one-quarter of the embryos should be homozygous for the deficiency chromosome). The sequences (from 5' to 3') of the pairs of primers used were (1) AGGTAGGACGCAAAAGTCTAGCCAGTTTGC and AGCCGACGGCGTTCACCTCCGTTCAAGTGA for the Scr promoter, (2) CAGGATCTGCCGCAGGACCAGCTCATTCGC and AGCAGCATCATCTTCGGCCTTGCGCTTGGT for the ftz promoter, and (3) CGTATGGGTTTGCATAATAGAGGTAAATGT and GTAACGAAAAGTGTGTGGCGTAAATTAGGT for a region ~9700 bp upstream of the Scr promoter.


 
View this table:
In this window
In a new window

  		Table 1. Chromosome aberrations associated with Scr and Antp mutations

Complementation tests:
Each chromosomal deficiency or Scr mutant allele was crossed to zen7, Dfd15, Scr4, ftz11, and Antp25 to verify that it fails to complement the expected mutations. Many of the mutations and chromosomal aberrations that map to the Scr region were also crossed to each other to test for complementation. Viability for a given genotype was determined by dividing the observed number of flies by the expected number and multiplying by 100%. The expected numbers were calculated from the surviving siblings of other genotypes. For example, in a cross of Scr4/Balancer females to Df(3R)Scr/Balancer males, the expected number of Scr4/Df(3R)Scr progeny was estimated as one-half of the total of Scr4/Balancer plus Df(3R)Scr/Balancer progeny. For a number of genotypes, the viability in our experiments was significantly greater than that previously observed by PATTATUCCI et al. 1991 Down. We believe that this is due to a difference in either the culture medium or the temperature. Because of these differences, we have compared different genotypes using only the data presented here.

Scr expression:
To estimate the degree of Scr expression in the imaginal leg cells of all three thoracic segments, the numbers of sex comb teeth were determined as described in KENNISON and RUSSELL 1987 Down. For preparations of adult cuticle, flies were boiled in 10% KOH, washed in distilled water, and mounted in Aqua Poly/Mount (Polysciences, Warrington, PA).


*  RESULTS AND DISCUSSION
*TOP
 *ABSTRACT
 *MATERIALS AND METHODS
 *RESULTS AND DISCUSSION
*LITERATURE CITED

The homeotic gene Scr is required for specifying the identity of the first thoracic segment in both embryonic and imaginal cells. Loss of Scr function in the cells of the first thoracic segment causes those cells to differentiate structures characteristic of the second thoracic segment (STRUHL 1982 Down; HAZELRIGG and KAUFMAN 1983 Down), while ectopic expression of Scr in the cells of the second or third thoracic segment causes those cells to differentiate structures characteristic of the first thoracic segment (PATTATUCCI and KAUFMAN 1991 Down). Quantitation of levels of expression for the Scr gene is difficult due to the rapid turnover of SCR proteins. However, earlier studies have shown a good correlation between the levels of expression of SCR proteins in the third instar imaginal leg cells and the degree of differentiation of first thoracic leg structures in the adult cuticle (PATTATUCCI and KAUFMAN 1991 Down). It has also been shown that the number of sex comb teeth (bristles specific to the first thoracic leg in adult males, shown in Fig 1) correlates well with the differentiation of other first leg-specific bristle patterns; adult legs with higher numbers of sex comb teeth more often have other bristle patterns characteristic of first legs than legs with lower numbers of sex comb teeth (HANNAH and STROMNAES 1955 Down; HANNAH-ALAVA 1958 Down). Because of the relative ease in counting sex comb teeth on the legs of living flies, we have used the numbers of sex comb teeth to estimate the levels of Scr expression in males of many different genotypes. These different genotypes include both mutations and chromosomal aberrations that affect function of one or more genes in the ANTC complex. Fig 2 shows ~200 kb of the ANTC that includes the Scr, Antp, and ftz transcription units, as well as the 3' end of the Dfd transcription unit. The molecular lesions associated with the Scr and Antp mutant chromosomes used in this study are depicted below the molecular map. The approximate positions of three insertions of repetitive DNA (ftz11, Antp23, and AntpScx), 17 Scr and Antp alleles associated with chromosomal aberrations (inversions, translocations, and transpositions), and 15 deficiencies with at least one breakpoint in this region have been included (GARBER et al. 1983 Down; SCOTT et al. 1983 Down; RILEY et al. 1987 Down; PATTATUCCI et al. 1991 Down; LINDSLEY and ZIMM 1992 Down). We localized the breakpoint of Df(3R)LIN2 between the Scr and ftz promoters using PCR with genomic DNA from homozygous deficiency embryos. We were unable to amplify a DNA fragment from the Scr promoter from homozygous embryos, but we were able to amplify DNA fragments from the ftz promoter (~16 kb upstream of the Scr promoter) and from sequences ~9.7 kb upstream of the Scr promoter. Both Df(3R)LIN2 and Df(3R)Tpl9 (which has a similar breakpoint in the region upstream of the Scr promoter) complement ftz11 but fail to complement Scr4, consistent with breakpoints between the Scr and ftz promoters.



View larger version (125K):
In this window
In a new window
Download PPT slide
  		Figure 1. The tarsal segments of the distal first leg of adult males. The genotypes are listed in the middle of each panel and are wild type (A), Scr2/+ (B), Df(3R)Scr/+ (C), and Df(3R)Tpl9/+ (D). Each bar spans the sex comb with the number of individual bristles (sex comb teeth) in that sex comb next to the bar. At the bottom are diagrams indicating which promoters (P) probably respond to which enhancers (boxes labeled E). The X in B indicates that transcription probably proceeds, but that the product is not functional. In C and D, the deleted DNA is indicated by an open bracket.



View larger version (27K):
In this window
In a new window
Download PPT slide
  		Figure 2. Molecular genetic map of the distal region of the ANTC. The top shows the location of the transcription units in 210 kb of genomic DNA from polytene chromosome bands 84B1–2. Both the P1 and P2 transcription units are shown for the Antp gene; however, only the 3' end of the Dfd gene is within the genomic region shown. Only the larger introns are shown (several small introns in Dfd, ftz, and Antp were omitted). The line below the transcription units refers to the coordinates on the genomic DNA as defined in SCOTT et al. 1983 Down and LINDSLEY and ZIMM 1992 Down. Just below the genomic DNA, the putative imaginal leg enhancer is indicated by the green bar. The approximate locations of the molecular lesions associated with various mutant chromosomes are shown below the map of the genomic DNA. All chromosomal aberrations in red cause ectopic sex comb teeth to differentiate on the second and third legs of heterozygous adult males (Table 3). The triangles are insertions of repetitive DNA and the vertical arrows are inversions, translocations, and transpositions. The horizontal arrows below the Scr and Antp alleles show the extent of the chromosome regions defined by molecular lesions associated with either partial or complete loss-of-function phenotypes. The black bars at the bottom show the extents of chromosomal deletions with one breakpoint within this genomic region. For the deletions listed on the left, only the right breakpoint is within this region. For the deletions listed on the right, only the left breakpoint is within this region. Both breakpoints of Df(3R)4SCB are within the genomic region depicted. The molecular analyses of the chromosome aberrations are from GARBER et al. 1983 Down, SCOTT et al. 1983 Down, RILEY et al. 1987 Down, PATTATUCCI et al. 1991 Down, or LINDSLEY and ZIMM 1992 Down.

Scr imaginal leg enhancer:
Genetic studies by Kaufman and co-workers have shown that a region upstream of the Scr promoter is required for Scr function in the imaginal first leg (HAZELRIGG and KAUFMAN 1983 Down; PATTATUCCI et al. 1991 Down). We have used chromosomal aberrations to further map the imaginal first leg requirement to a region ~35–40 kb upstream of the Scr promoter [between the breakpoints of Df(3R)Antp7 and Df(3R)A41]. The data from these experiments to map the imaginal leg enhancer are shown in Fig 3. Scr4 is a null mutation that makes no detectable protein (RILEY et al. 1987 Down). Although Df(3R)Antp7 was reported to show no viability when heterozygous to Scr4 (PATTATUCCI et al. 1991 Down), we recovered 11% of the expected flies of this genotype. Scr4/Df(3R)Antp7 males had 4–7 sex comb teeth per leg (an average of 5.6), only slightly less than the 4–8 sex comb teeth per leg (an average of 6.2) observed in Scr4/+ heterozygotes. In contrast, Scr4/Df(3R)A41 males, which were recovered at 1% of the expected frequency, rarely had sex comb teeth on the first legs (an average of 0.08). Similar results were observed with Df(3R)Hu, which deletes ~10 kb more from this upstream region. Although only 2% of the expected Scr4/Df(3R)Hu males survived, none of them had any sex comb teeth on the first legs. We observed similar results with the hypomorphic mutation Scr1, but because the viability of flies heterozygous for the deletions was much greater with Scr1 than with Scr4, we were able to examine more males. Of 40 first legs examined from Scr1/Df(3R)Antp41 males, only 11 total sex comb teeth (an average of 0.3 per leg) were observed. On the same number of first legs of Scr1/Df(3R)Antp7 males, 211 total sex comb teeth (an average of 5.3 per leg) were observed. Deletions that remove more DNA from this region did not give any surviving adult males when heterozygous to either Scr1 or Scr4 nor did deletions that extend into the Scr gene from the proximal end of the ANTC complex. These results show that sequences between the breakpoints of Df(3R)Antp7 and Df(3R)A41 (~35–40 kb upstream of the Scr promoter) are required for expression of Scr in the imaginal first leg.



View larger version (28K):
In this window
In a new window
Download PPT slide
  		Figure 3. The ranges and average numbers of sex comb teeth per first leg for males heterozygous for Scr1 or Scr4. The genotypes are shown at the bottom. N, the number of legs scored. The open bars indicate the ranges and the horizontal solid bars are the average numbers of sex comb teeth per leg for each sample. The horizontal dotted line indicates 6.1 sex comb teeth, the average for males with one wild-type Scr gene.

Transvection at the Scr gene in the first thoracic segment:
To further confirm the presence of the imaginal leg enhancer upstream of the Scr promoter, we also counted the number of sex comb teeth on the first legs of many mutations and chromosomal aberrations when heterozygous to a wild-type allele (Fig 4). Males with two wild-type copies of the Scr gene had 9–12 sex comb teeth per first leg (an average of 10.8), while males with only one wild-type copy had 4–8 (averages of 6.1 per leg for Scr1 heterozygotes, 6.3 per leg for Scr2 heterozygotes, and 6.2 per leg for Scr4 heterozygotes). Similar results were observed with two deletions that removed the entire Scr gene, Df(3R)Scr and Df(3R)Antp17. Males heterozygous for Df(3R)Antp7 had only a slight reduction in the average number of sex comb teeth per first leg (9.4), while males heterozygous for deletions that removed more DNA from the distal end of the Scr gene [Df(3R)A41, Df(3R)Scx4, and Df(3R)Hu] had about the same number of sex comb teeth per leg as males heterozygous for complete loss of Scr. Thus, loss of the imaginal leg enhancer appears to be equivalent to complete loss of the gene.



View larger version (27K):
In this window
In a new window
Download PPT slide
  		Figure 4. The ranges and average numbers of sex comb teeth per first leg for Scr mutant males. The genotypes are at the bottom. N, the number of legs scored. The open bars indicate the ranges and the horizontal solid bars are the average numbers of sex comb teeth per leg for each sample. The top horizontal dotted line indicates 10.8 sex comb teeth, the average for males with two wild-type Scr genes. The bottom horizontal dotted line indicates 6.1 sex comb teeth, the average for males with only one wild-type Scr gene.

The results from deleting DNA from the proximal end of the Scr gene were not as expected, however. We found that deletions that remove only proximal parts of the Scr gene, as well as most chromosomal aberrations with one breakpoint within the Scr gene, have far more Scr function than expected (Fig 4). We can explain these results by transvection at the Scr gene. The ability of cis-regulatory elements to activate transcription in trans on the homologous chromosome has been observed for a number of genes in Drosophila and is an example of transvection (LEWIS 1954 Down; GELBART 1982 Down; KORNHER and BRUTLAG 1986 Down; GEYER et al. 1990 Down; LEISERSON et al. 1994 Down; HOPMANN et al. 1995 Down). Although cis-regulatory elements in the Scr gene have previously been shown to activate transcription of the wild-type homolog when derepressed in the second and third thoracic segments (PATTATUCCI and KAUFMAN 1991 Down), transvection at the Scr gene in the first thoracic segment has not been described. Deletions that remove part of the Scr transcription unit, but not the promoter, do not affect the expression of the homologous gene. For example, males heterozygous for either Df(3R)WIN11 or Df(3R)CP1 (Fig 4) have only 4–6 sex comb teeth per first leg, the number expected from normal function of the single Scr gene on the wild-type homolog. Deletions that remove the entire Scr transcription unit, including the promoter and considerable upstream regulatory elements (proximal deletions that are also lacking ftz function), also fail to affect function of the wild-type homolog in the first leg. These include Df(3R)9A99 (an average of 6.1 sex comb teeth per first leg) and Df(3R)Dfd13 (an average of 5.0 sex comb teeth per first leg). Surprisingly, proximal deletions that remove the Scr promoter but not upstream regulatory sequences have increased function of Scr from the homologous promoter. Df(3R)Tpl9 and Df(3R)LIN2 remove the entire Scr transcription unit (Fig 2), but heterozygous males have 8–10 sex comb teeth per first leg (averages of 8.8 and 8.2). This suggests that the imaginal leg enhancer is able to activate transcription of the homologous promoter when the promoter in cis is deleted. Fig 1 shows first legs from a wild-type male, a male heterozygous for a putative point mutation (Scr2), a male heterozygous for a deletion that includes the entire Scr gene [Df(3R)Scr], and a male heterozygous for a deletion that removes the Scr promoter and transcription unit, but not the upstream regulatory DNA [Df(3R)Tpl9], with diagrams below each leg to illustrate transvection in the Df(3R)Tpl9 heterozygote.

Males heterozygous for inversions, translocation, and transpositions in the Scr genomic region also show transvection in the first thoracic segment. As shown in Fig 4, males heterozygous for Scr alleles associated with chromosomal aberrations that have breakpoints between the Scr promoter and the putative imaginal leg enhancer have more sex comb teeth per first leg than expected for a single functional Scr allele. For example, males heterozygous for T(2;3)ftzRpl (which has a breakpoint about midway between the Scr promoter and the imaginal leg enhancer) have 8–12 sex comb teeth per first leg (an average of 10.4), which shows almost no difference from males with two wild-type Scr alleles. Males heterozygous for similar chromosomal aberrations in which the breakpoint is within the Scr transcription unit (and therefore not between the promoter and imaginal leg enhancer) have the expected reduction in the number of sex comb teeth per first leg: averages of 6.4–7.4 for In(3LR)Scr10, In(3LR)Scr9, Tp(3;3)ScrT1, and In(3R)ScrMsc .

The two proximal Scr deletions, Df(3R)Tpl9 and Df(3R) LIN2, that appear to have more Scr activity than null mutations in heterozygous males also have more activity than null mutations by other genetic tests. Males heterozygous for Df(3R)Tpl9 or Df(3R)LIN2 always have a higher number of sex comb teeth per first leg (and usually higher viability) compared to males heterozygous for Scr4. We have examined this in males in which the homologous chromosome carries a variety of mutations or chromosomal aberrations that reduce Scr expression or function. These results are shown in Fig 5 and Fig 6. For example, Scr4/Df(3R)A41 males were only 1% as viable as expected and rarely had sex comb teeth on the first legs (an average of only 0.08 per leg); Df(3R)Tpl9/Df(3R)A41 males had greater viability (13% of the expected) and greater numbers of sex comb teeth per first leg (an average of 1.8). Df(3R)Tpl9 and Df(3R)LIN2 do not behave as null mutations by genetic tests, even though both delete the entire Scr transcription unit.



View larger version (32K):
In this window
In a new window
Download PPT slide
  		Figure 5. The ranges and average numbers of sex comb teeth per first leg for Scr mutant males. The genotypes are shown at the bottom. N, the number of legs scored. The open bars indicate the ranges and the horizontal solid bars are the average numbers of sex comb teeth per leg for each sample.



View larger version (35K):
In this window
In a new window
Download PPT slide
  		Figure 6. Distributions of first thoracic legs with various numbers of sex comb teeth. The horizontal axes show the numbers of sex comb teeth per leg. For each sample, the distributions are shown by solid boxes, with the height of each box proportional to the percentage of legs in the sample with that number of sex comb teeth. All distributions are drawn to the same scale for comparison. N, number of legs scored for each sample. The genotype is shown to the right of each distribution.

The test for transvection is to show pairing dependence. We examined males heterozygous for Df(3R)Tpl9 and one of four translocations with breakpoints in proximal 3R (Table 2 and Fig 6). Breakpoints in 3R proximal to Scr are expected to reduce transvection, while breakpoints distal should have little or no effect. As expected, all three translocations that have breakpoints proximal to Scr [T(2;3)MAP3, T(2;3)P607, and T(2;3)bwvDe3] reduce the average number of sex comb teeth per first leg in heterozygous Df(3R)Tpl9 males (from 8.8 to 7.2 or less). Heterozygosity for T(2;3)Antp17, which has a breakpoint distal to Scr and should disrupt pairing of Scr very little (if at all), does not appear to have any effect on the number of sex comb teeth per first leg in Df(3R)Tpl9 heterozygous males. We have also used a duplication [Dp(3;Y)77ab] on the Y chromosome for the ANTC to examine transvection of Scr. The duplication appears to provide wild-type Scr function, but does not appear to pair with the endogenous Scr region. Df(3R) Tpl9/Scr4; Dp(3;Y)77ab males have an average of only 6.4 sex comb teeth per first leg (the number expected for the one wild-type Scr allele on the duplication).


 
View this table:
In this window
In a new window

  		Table 2. Transvection of Scr in the first leg is sensitive to pairing but not zeste mutations

Although transvection at some genes in Drosophila is dependent on the protein product of the zeste gene (KAUFMAN et al. 1973 Down; GELBART and WU 1982 Down), other examples of transvection appear to be zeste independent. For the Ubx gene in which transvection was first identified, both zeste-dependent and zeste-independent examples of transvection have been characterized (KAUFMAN et al. 1973 Down; MARTINEZ-LABORDA et al. 1992 Down). To determine whether the transvection at Scr that we observed with Df(3R)Tpl9 is dependent on zeste, we used three different alleles of zeste, z1, za, and z11G3 (KAUFMAN et al. 1973 Down; GELBART and WU 1982 Down). None of the zeste alleles appeared to affect the viability or average numbers of sex comb teeth for Df(3R)Tpl9 males that were also heterozygous for wild type, Df(3R)A41, or Df(3R)Scx4 (Table 2 and Fig 6). Transvection of Scr in the first leg appears to be zeste independent, at least for the example presented here. The pairing-dependent derepression of Scr in the second and third legs described in the next section also appears to be zeste independent.

Silencing of Scr in the second and third thoracic segments:
The imaginal leg enhancer appears to function in all three leg discs, but is normally silenced in the second and third leg discs. Silencing of Scr in the second and third thoracic segments has long been known to require trans-acting factors encoded by the Polycomb group genes (reviewed in KENNISON 1995 Down; SIMON 1995 Down; PIRROTTA 1997 Down; VAN LOHUIZEN 1999 Down). Mutations in Polycomb group genes disrupt silencing and allow ectopic expression of Scr in the imaginal cells that form the second and third legs. This leads to the differentiation of sex comb teeth (a characteristic first leg structure) on the second and third legs of adult males. In contrast, mutations in Polycomb group genes do not appear to affect the expression of Scr in the first thoracic segment (PATTATUCCI et al. 1991 Down; Fig 7).



View larger version (25K):
In this window
In a new window
Download PPT slide
  		Figure 7. The ranges and average numbers of sex comb teeth per first leg for Scr mutant males. The genotypes are shown at the bottom. N, the number of legs scored. The open bars indicate the ranges and the horizontal solid bars are the average numbers of sex comb teeth per leg for each sample.

Mutations and chromosomal aberrations with breakpoints in the ANTC are also known to disrupt silencing of Scr and cause the appearance of ectopic sex comb teeth on the second and third legs of mutant males (TOKUNAGA 1966 Down; SCOTT et al. 1983 Down; PATTATUCCI et al. 1991 Down). HAZELRIGG and KAUFMAN 1983 Down first made the very important observation that the ScrW mutation causes ectopic expression from the Scr gene on the wild-type homolog, not from the mutant chromosome. PATTATUCCI and KAUFMAN 1991 Down extended this to show that many of the chromosomal aberrations in the ANTC that cause ectopic expression of Scr in the second and third legs are disrupting silencing on the homolog, rather than on the chromosome that is actually involved in the chromosomal aberration. Although some chromosomal aberrations with breakpoints within or near the Scr transcription unit disrupted silencing, other aberrations with similar breakpoints did not (PATTATUCCI and KAUFMAN 1991 Down; PATTATUCCI et al. 1991 Down). We have repeated and extended these observations, with slightly different results. We have examined adult males heterozygous for each of the chromosome aberrations shown in Fig 2, as well as several putative point mutations in Scr, Antp, ftz, and Dfd. Most of the mutations and chromosomal aberrations that we tested had no effects on Scr silencing. The exceptions are shown in red in Fig 2. The total numbers of ectopic sex comb teeth (and the averages per leg) for each genotype that showed any Scr derepression are included in Table 3. Fig 8 shows two examples of ectopic sex comb teeth on second legs from mutant males. Since most of the genotypes that we examined did not include any males with ectopic sex comb teeth, we have given the data for only five of these genotypes (homozygous wild type and males heterozygous for Scr1, Scr2, Scr4, or Scr10) in Table 3 for comparison. Although ScrT1, ScrT2, and ScrT3 were originally isolated as weak dominant mutants with ectopic sex comb teeth, no ectopic sex comb teeth were observed in the males examined by PATTATUCCI et al. 1991 Down. We initially examined males heterozygous for these three mutations >15 years ago and observed ectopic sex comb teeth in a few males of each genotype; we recently repeated these experiments and observed ectopic sex comb teeth in almost identical frequencies to those observed in our earlier experiments. We also observed a small number of ectopic sex comb teeth in males heterozygous for Scr9 (Table 3); PATTATUCCI et al. 1991 Down observed no ectopic sex comb teeth in their sample of heterozygous Scr9 males. We believe that these differences are probably due to different culture conditions. Most mutations that have ectopic sex comb teeth in heterozygous males (for example, ScrMsc and AntpScx) can be suppressed by low temperature, overcrowding, or Minute mutations (PURO and NYGREN 1975 Down; DENELL 1978 Down; SINCLAIR et al. 1984 Down; KENNISON and RUSSELL 1987 Down).



View larger version (108K):
In this window
In a new window
Download PPT slide
  		Figure 8. Ectopic sex comb teeth on second legs of adult males. The leg on the left is wild type and has no ectopic sex comb teeth. The legs in the middle and on the right are examples from two AntpScx heterozygous males with different numbers of ectopic sex comb teeth. The ectopic sex comb teeth are indicated by the bars. The number of teeth is to the left of each bar.


 
View this table:
In this window
In a new window

  		Table 3. Ectopic sex comb teeth on second and third legs

We observed ectopic sex comb teeth on the second and third legs of males heterozygous for any translocation, transposition, or inversion that has one breakpoint within a 60- to 70-kb region of the Scr gene (Table 3 and Fig 2). We did not observe ectopic sex comb teeth on the legs of males heterozygous for any translocation, transposition, or inversion that had breakpoints proximal or distal to the 60–70 kb region, nor did we observe ectopic sex comb teeth on the legs of males heterozygous for any deletion. With the exception of AntpScx, which is associated with the insertion of repetitive DNA near the Antp P1 promoter (SCOTT et al. 1983 Down) and is discussed further in considerable detail below, we never observed ectopic sex comb teeth on the legs of males that are heterozygous for transposable element insertional mutations or putative point mutations. Only chromosomal aberrations in which the Scr gene is broken, but both pieces are still present in the genome, appear able to disrupt Scr silencing. Interestingly, complete disruption of silencing in chromosome rearrangements can occur independently of Polycomb group mutations. This is illustrated by AntpScx/Df(3R)Tpl9 flies (Table 3). We observed almost the same number of sex comb teeth in the first and second legs (averages of 8.9 and 8.7), suggesting that Scr is completely derepressed in the second leg. Further, there were no increases in the numbers of sex comb teeth when these flies were also heterozygous for Pc21 (averages of 8.9 and 8.5 for the first and second legs, respectively).

Consistent with earlier work (DENELL 1973 Down; PATTATUCCI and KAUFMAN 1991 Down), the Scr derepression that we observed in males heterozygous for the chromosomal aberrations is pairing dependent, but does not require wild-type ZESTE proteins (Table 3). For example, we observed no difference in the frequency of ectopic sex comb teeth in ScrMsc heterozygous males that are z+ (an average of 3.2 sex comb teeth per second leg), z1 (an average of 3.3 sex comb teeth per second leg), or za (an average of 3.0 sex comb teeth per second leg). We have examined the effects of pairing on Scr derepression in ScrMsc males, using a duplication of the ANTC complex on the Y chromosome, Dp(3;Y)77ab. This duplication rescues the Scr loss-of-function phenotype, but does not appear to pair very well with the endogenous Scr gene in the third chromosome (Table 2 and Fig 6). In males that carry the ScrMsc mutant chromosome heterozygous to a wild-type chromosome, we observed an average of 3.2 sex comb teeth per second leg (Table 3). When the wild-type Scr gene was on the Y chromosome duplication, we observed an average of <0.3 sex comb teeth per second leg [the ScrMsc/Scr4;Dp(3;Y)77ab and ScrMsc/Df(3R)Scr;Dp(3;Y)77ab males in Table 3]. This shows that the extra sex combs phenotype of ScrMsc requires the presence of a wild-type transcription unit on the homolog. The lack of ectopic sex comb teeth in ScrMsc/Df(3R)Scx4 males (Table 3) shows that the imaginal leg enhancer on the homolog is also required for the extra sex combs phenotype of ScrMsc. If the ScrMsc chromosome provides neither the transcription unit nor the enhancer, how does it cause derepression of the wild-type homolog? Similar results have been observed with other Scr mutations that derepress Scr in the second and third legs (Table 7 in PATTATUCCI et al. 1991 Down; Table 3 of this article).

To account for these data, we propose that two regulatory elements normally interact to maintain silencing of the wild-type gene. These putative regulatory elements are located distal and proximal to the 60–70 kb region that includes the chromosome rearrangements that cause the appearance of ectopic sex comb teeth. Although the distal and proximal elements may be different, we refer to both putative regulatory elements as maintenance elements for silencing (MES). In our model, when the Scr gene is active, flanking MESs fail to interact (Fig 9A). When the Scr gene is silenced, the flanking MESs preferentially interact in cis to stabilize silencing of genes in between (Fig 9B). The interaction of MESs may occur through the binding of different proteins to these elements when silencing is specified, or it may occur by the modification of proteins already bound even when the gene is active. Maintenance of silencing, however, affects only genes that lie between two elements; i.e., silencing requires the ability to form a physical loop of DNA between the two elements. Fig 9C shows what we believe occurs in flies heterozygous for chromosomal aberrations with one breakpoint between the two silencing elements. Interaction of the elements on the wild-type homolog would preferentially occur in cis, maintaining silencing in most cells. However, because the silencing elements on the broken chromosome are no longer in cis, they could compete for interactions with the silencing elements on the wild-type homolog. If both elements on the aberration chromosome interact with the elements on the homolog, the configuration shown in Fig 9C (right) might be stable enough to prevent interaction of the two elements in cis on the wild-type chromosome. This would disrupt silencing of the Scr promoter between these two elements, allowing derepression of the wild-type Scr gene. We believe that deletion chromosomes that contain only one MES are not able to effectively compete with the cis interactions on the wild-type homolog. Our model can account for all of the data described so far, and it can also explain the behavior of an old mutation with very anomalous properties. This is the AntpScx mutation isolated in 1953 (HANNAH and STROMNAES 1955 Down).



View larger version (24K):
In this window
In a new window
Download PPT slide
  		Figure 9. Model for derepression of a silenced gene by chromosome aberrations. Solid lines are chromosomes; arrows indicate the transcription units; open circles indicate maintenance elements for silencing (MESs) that have not bound proteins; and the shaded and crosshatched circles are MESs with proteins bound. (A) A pair of chromosomes in which both Scr alleles are active. (B) A cell in which the MESs have bound proteins to interact and silence both Scr alleles. (C) A cell in which a chromosome aberration in one chromosome destabilizes the interactions of MESs in cis and allows them to interact with MESs in the homolog, causing the silenced allele in the homolog to become activated. (D) The model for activation of the silenced gene (in cis at the bottom left and in trans at the bottom right) by the insertion of an additional MES upstream of the gene in the Scx (AntpScx) mutant chromosome.

The AntpScx mutant was isolated originally on the basis of a dominant extra sex combs phenotype (HANNAH and STROMNAES 1955 Down). It is lethal when heterozygous to Antp mutant alleles, but is viable when heterozygous to Scr mutant alleles (HANNAH-ALAVA 1958 Down; DENELL et al. 1981 Down). The AntpScx mutant chromosome is cytologically normal and the only molecular lesion identified in the ANTC was the insertion of repetitive DNA very close to the Antp P1 promoter (SCOTT et al. 1983 Down). Given the physical location of the insertion, it is not surprising that the AntpScx mutant chromosome fails to complement Antp alleles that specifically lack P1 function, such as AntpB, Antp73b, AntpCB, and Antp17 (HANNAH and STROMNAES 1955 Down; DENELL et al. 1981 Down). There is no difference in the average number of sex comb teeth per first leg in AntpScx heterozygous males compared to homozygous wild type (Fig 4) or in AntpScx/Scr4 males compared to +/Scr4 males (Fig 3). Males heterozygous for AntpScx, however, do have a considerable number of ectopic sex comb teeth (an average of 2.7 per second leg, Table 3). The ectopic sex comb teeth result from misexpression of Scr in cis and in trans (Table 3). Males with Scr mutations in cis to AntpScx (ScrE2 AntpScx/+ and ScrE3 AntpScx/+) have fewer sex comb teeth per second leg (an average of 0.8); males with Scr mutations on the homolog (AntpScx/Scr2 and AntpScx/Scr4) also have fewer sex comb teeth per second leg (an average of 1.3–1.6). Scr mutations both in cis and in trans to AntpScx (ScrE2 AntpScx/ Scr4;Dp(3;Y)77ab) almost completely eliminate the ectopic expression of Scr (an average of only 0.02 sex comb teeth per second leg). Comparison of the effects of Scr mutations in cis and in trans also suggest that AntpScx derepresses the Scr promoter in cis about twice as much as the Scr promoter on the homolog. A molecular mechanism through which the insertion of repetitive DNA ~150 kb upstream of the Scr promoter might be responsible for transcriptional derepression of both the cis promoter and the Scr promoter on the homolog has not been previously suggested. To our knowledge, the model shown in Fig 9D is the first attempt to explain the unusual properties of the AntpScx mutant chromosome. We believe that the repetitive DNA inserted near the Antp P1 promoter on the AntpScx mutant chromosome mimics the endogenous regulatory elements involved in the maintenance of silencing (the MES elements in Fig 9). By competing for interactions with the endogenous elements either on the same chromosome (Fig 9D, bottom left) or on the homolog (Fig 9D, bottom right), the AntpScx insertion disrupts silencing of the Scr promoter in cis or in trans, respectively. In this respect, the AntpScx insertion appears to be more effective than a wild-type MES, since deletion chromosomes with a single MES do not interfere with silencing on the homolog (see preceding paragraph). Not only does this model explain the existing data, but it also makes a prediction. The Antp P2 promoter is between the repetitive insertion on the AntpScx mutant chromosome and the endogenous regulatory elements in the Scr gene. Interactions between the AntpScx insertion and the endogenous elements in the Scr gene in cis should not only derepress the Scr promoter, but should also silence the Antp P2 promoter (Fig 10). Interactions between the AntpScx insertion and the endogenous elements in the Scr gene in trans should not silence the Antp P2 promoter. Since the AntpScx mutation appears to derepress Scr in cis about twice as much as in trans (Table 3), we expect about two-thirds of the cells to lack Antp P2 function from the AntpScx chromosome. Two mutations (Antp1 and Antp23) have been characterized that inactivate the Antp P2 promoter but appear to have normal function for the Antp P1 promoter (ABBOTT and KAUFMAN 1986 Down). These two mutations can be used to examine Antp P2 function on the homologous chromosome in heterozygotes. As expected from our model, AntpScx interferes significantly with function of the P2 promoter; AntpScx fails to complement both Antp1 and Antp23 for viability (we have found no surviving adults among several hundred expected). In contrast, deletions that remove the Antp P1 promoter [Df(3R) Antp1P and Df(3R)Antp2 shown in Fig 10] and chromosome aberrations that physically separate the P1 and P2 promoters [In(3R)AntpB , In(3LR)AntpPW, In(3R)AntpR, In(3R)AntpLC, T(2;3)AntpCtx, In(3R)Antp73b, T(2;3)Antp17, and In(3R)AntpCB in Fig 10] are all viable when heterozygous to either Antp1 or Antp23 (ABBOTT and KAUFMAN 1986 Down; data not shown). With these results, four genetic properties are now associated with the AntpScx mutant chromosome: (1) loss of Antp P1 function, (2) loss of Antp P2 function, (3) derepression of the Scr promoter on the mutant chromosome, and (4) derepression of the Scr promoter on the homolog.



View larger version (18K):
In this window
In a new window
Download PPT slide
  		Figure 10. Derepression of the Scr promoter in cis should silence the Antp P2 promoter. The genomic region that includes the Scr and Antp transcription units is shown at the top (see the Fig 2 legend). The locations of the endogenous silencing elements in the Scr gene are indicated by circles just below the genomic map and the insertion of repetitive DNA in the AntpScx chromosome is indicated by the triangle labeled Scx. Curved lines joining the circles and triangle indicate the interactions between proteins bound to these sites. The large X indicates that the interaction on the left may fail due to competition from the interaction on the right. The Antp P1 and P2 promoter mutants are shown at the bottom. As in Fig 2, the solid triangle indicates the insertion of a transposable element; the vertical arrows indicate breakpoints of inversions, translocations, and transposition; and the black bars indicate DNA missing in Df(3R)Antp1P and Df(3R)Antp2.

It is possible that there are multiple molecular lesions on the AntpScx mutant chromosome that were not detected in the molecular analyses. However, it should be emphasized that the AntpScx mutant chromosome is cytologically normal, is wild type for Scr function, and has the ability to derepress the Scr gene in trans. Only Scr mutations that have chromosome aberration breakpoints within the Scr locus have the ability to derepress Scr in trans. Our model explains how the identified molecular lesion could lead to all of the mutant phenotypes observed.

In our model, trans interactions between MESs occur when the cis interactions are disrupted. Although PREs are believed to normally act in cis to maintain silencing, they are also able to act in trans when included within transgenes. These trans interactions of PREs are enhanced when cis interactions are blocked (SIGRIST and PIRROTTA 1997 Down; MULLER et al. 1999 Down). In addition, while single PREs appear to partially silence transgenes, silencing is often greater when multiple PREs can interact (SIGRIST and PIRROTTA 1997 Down; MULLER et al. 1999 Down). A pair of major PREs has also been characterized in about the same position in the Ultrabithorax (Ubx) homeotic gene as the MESs in the Scr gene; i.e., one PRE is ~25 kb upstream of the Ubx promoter and a second PRE is within an intron in the middle of the transcription unit (PIRROTTA 1997 Down). Therefore, an important question is whether MESs are the same as PREs. We believe that they are distinct elements, but are often in close proximity. Many DNA fragments that contain PREs may also contain MES elements, but these activities may be separable. For example, a 2.9-kb DNA fragment from the Mcp region of the bithorax complex appears to contain at least two different types of regulatory elements (MULLER et al. 1999 Down). An 800-bp DNA fragment from the central region of the larger fragment is not sufficient for silencing, but it is sufficient for mediating pairing-sensitive interactions between transgenes on different chromosomes. It is also sufficient for mediating long-range interactions between enhancers and promoters in transgenes. Two XbaI restriction fragments from Scr (an 8.2-kb fragment from the second intron and a 10.0-kb fragment 35–45 kb upstream of the promoter) that have been tested in transgenes for PRE activity overlap with the putative MESs (GINDHART and KAUFMAN 1995 Down). Both fragments appear to partially silence the reporter gene in a transgene assay. This silencing is sensitive to some Polycomb group mutations; however, the two tested fragments differ as to which Polycomb group mutations had effects. Interestingly, only the 8.2-kb fragment exhibited pairing-sensitive silencing, while only the 10.0-kb fragment functioned as a PRE in embryos. The apparent independence of MES function and Polycomb group repression also suggests that MESs may be separate elements that are in close proximity to PREs. It is possible that MESs act to maintain interactions between nearby PREs, thus facilitating the maintenance of silencing. In this respect, MESs may be similar to the pairing-sensitive regulatory elements identified upstream of the engrailed promoter (KASSIS 1994 Down).


*  ACKNOWLEDGMENTS

We thank Judy Kassis, Jürg Müller, Susan Haynes, Sam Stoler, Rob Scott, Angela Pattatucci, Tom Kaufman, and John Tamkun for many helpful discussions and ideas and Judy Kassis and Charlene Southworth for comments on the manuscript. We also acknowledge Stanley Tiong, Rob Denell, Tom Kaufman, and the Drosophila stock centers at Bloomington (Indiana) and Umea (Sweden) for providing mutant strains.

Manuscript received March 9, 2001; Accepted for publication March 15, 2002.


*  LITERATURE CITED
*TOP
 *ABSTRACT
 *MATERIALS AND METHODS
 *RESULTS AND DISCUSSION
 *LITERATURE CITED

ABBOTT, M. K. and T. C. KAUFMAN, 1986  The relationship between the functional complexity and the molecular organization of the Antennapedia locus of Drosophila melanogaster.. Genetics 114:919-942[Abstract/Free Full Text].

DENELL, R. E., 1973  Homeosis in Drosophila. I. Complementation studies with revertants of Nasobemia. Genetics 75:279-297[Abstract/Free Full Text].

DENELL, R. E., 1978  Homeosis in Drosophila. II. A genetic analysis of Polycomb. Genetics 90:277-289[Abstract/Free Full Text].

DENELL, R. E., K. R. HUMMELS, B. T. WAKIMOTO, and T. C. KAUFMAN, 1981  Developmental studies of lethality associated with the Antennapedia gene complex in Drosophila melanogaster.. Dev. Biol. 81:43-50[Medline].

GARBER, R. L., A. KUROIWA, and W. J. GEHRING, 1983  Genomic and cDNA clones of the homeotic locus Antennapedia in Drosophila.. EMBO J. 2:2027-2036[Medline].

GELBART, W. M., 1982  Synapsis-dependent allelic complementation at the decapentaplegic gene complex in Drosophila melanogaster.. Proc. Natl. Acad. Sci. USA 79:2636-2640[Abstract/Free Full Text].

GELBART, W. M. and C.-T. WU, 1982  Interactions of zeste mutations with loci exhibiting transvection effects in Drosophila melanogaster.. Genetics 102:179-189[Abstract/Free Full Text].

GELLON, G. and W. MCGINNIS, 1998  Shaping animal body plans in development and evolution by modulation of Hox expression patterns. Bioessays 20:116-125[Medline].

GEYER, P. K., M. M. GREEN, and V. G. CORCES, 1990  Tissue-specific transcriptional enhancers may act in trans on the gene located in the homologous chromosome: the molecular basis of transvection in Drosophila.. EMBO J. 9:2247-2256[Medline].

GINDHART, J. G., JR. and T. C. KAUFMAN, 1995  Identification of Polycomb and trithorax group responsive elements in the regulatory region of the Drosophila homeotic gene Sex combs reduced. Genetics 139:797-814[Abstract].

HAGSTROM, K. and P. SCHEDL, 1997  Remembrance of things past: maintaining gene expression patterns with altered chromatin. Curr. Opin. Genet. Dev. 7:814-821[Medline].

HANNAH, A. and O. STRÖMNAES, 1955  Extra sex combs mutants in Drosophila melanogaster.. Dros. Inf. Serv. 29:121-123.

HANNAH-ALAVA, A., 1958  Developmental genetics of the posterior legs of Drosophila melanogaster.. Genetics 43:878-905[Free Full Text].

HAZELRIGG, T. and T. C. KAUFMAN, 1983  Revertants of dominant mutations associated with the Antennapedia gene complex of Drosophila melanogaster: cytology and genetics. Genetics 105:581-600[Abstract/Free Full Text].

HOPMANN, R., D. DUNCAN, and I. DUNCAN, 1995  Transvection in the iab-5,6,7 region of the bithorax complex of Drosophila: homology independent interactions in trans.. Genetics 139:815-833[Abstract].

JÜRGENS, G., 1985  A group of genes controlling the spatial expression of the bithorax complex in Drosophila.. Nature 316:153-155.

KASSIS, J. A., 1994  Unusual properties of regulatory DNA from the Drosophila engrailed gene: three "pairing-sensitive" sites within a 1.6-kb region. Genetics 136:1025-1038[Abstract].

KAUFMAN, T. C., S. E. TASAKA, and D. T. SUZUKI, 1973  The interaction of two complex loci, zeste and bithorax in Drosophila melanogaster.. Genetics 75:299-321[Abstract/Free Full Text].

KAUFMAN, T. C., M. A. SEEGER, and G. OLSEN, 1990  Molecular and genetic organization of the Antennapedia gene complex of Drosophila melanogaster.. Adv. Genet. 27:309-362[Medline].

KENNISON, J. A., 1995  The Polycomb and trithorax group proteins of Drosophila: trans-regulators of homeotic gene function. Annu. Rev. Genet. 29:289-303[Medline].

KENNISON, J. A. and M. A. RUSSELL, 1987  Dosage-dependent modifiers of homoeotic mutations in Drosophila melanogaster.. Genetics 116:75-86[Abstract/Free Full Text].

KORNHER, J. S. and D. BRUTLAG, 1986  Proximity-dependent enhancement of Sgs-4 gene expression in D. melanogaster. Cell 44:879-883[Medline].

LEISERSON, W. M., N. M. BONINI, and S. BENZER, 1994  Transvection at the eyes absent gene of Drosophila. Genetics 138:1171-1179[Abstract].

LEWIS, E. B., 1954  The theory and application of a new method of detecting chromosomal rearrangements in Drosophila melanogaster.. Am. Nat. 88:225-239.

LEWIS, E. B., 1993  Clusters of master control genes regulate the development of higher organisms. Curr. Sci. 64:640-649.

LEWIS, E. B., 1998  The bithorax complex: the first fifty years. Int. J. Dev. Biol. 42:403-415[Medline].

LINDSLEY, D. L., and G. G. ZIMM, 1992 The Genome of Drosophila melanogaster. Academic Press, San Diego.

LINDSLEY, D. L., L. SANDLER, B. S. BAKER, A. T. C. CARPENTER, and R. E. DENELL et al., 1972  Segmental aneuploidy and the genetic gross structure of the Drosophila genome. Genetics 71:157-184[Abstract/Free Full Text].

MARTÍNEZ-LABORDA, A., A. GONZÁLEZ-REYES, and G. MORATA, 1992  Trans regulation in the Ultrabithorax gene of Drosophila: alterations in the promoter enhance transvection. EMBO J. 11:3645-3652[Medline].

MIHALY, J., I. HOGGA, S. BARGES, M. GALLONI, and R. K. MISHRA et al., 1998  Chromatin domain boundaries in the bithorax complex. Cell. Mol. Life Sci. 54:60-70[Medline].

MULLER, M., K. HAGSTROM, H. GYURKOVICS, V. PIRROTTA, and P. SCHEDL, 1999  The Mcp element from the Drosophila melanogaster bithorax complex mediates long-distance regulatory interactions. Genetics 153:1333-1356[Abstract/Free Full Text].

PATTATUCCI, A. M. and T. C. KAUFMAN, 1991  The homeotic gene Sex combs reduced of Drosophila melanogaster is differentially regulated in the embryonic and imaginal stages of development. Genetics 129:443-461[Abstract].

PATTATUCCI, A. M., D. C. OTTESON, and T. C. KAUFMAN, 1991  A functional and structural analysis of the Sex combs reduced locus of Drosophila melanogaster.. Genetics 129:423-441[Abstract].

PIRROTTA, V., 1997  Chromatin-silencing mechanisms in Drosophila maintain patterns of gene expression. Trends Genet. 13:314-318[Medline].

PIRROTTA, V., 1998  Polycombing the genome: PcG, trxG, and chromatin silencing. Cell 93:333-336[Medline].

PURO, J. and J. NYGREN, 1975  Mode of action of a homeotic gene in Drosophila melanogaster.. Hereditas 81:237-248[Medline].

RASTELLI, L., C. S. CHANG, and V. PIRROTTA, 1993  Related chromosome binding sites for zeste, suppressors of zeste and Polycomb group proteins in Drosophila and their dependence on Enhancer of zeste function. EMBO J. 12:1513-1522[Medline].

RILEY, P. D., S. B. CARROLL, and M. P. SCOTT, 1987  The expression and regulation of sex combs reduced protein in Drosophila embryos. Genes Dev. 1:716-730[Abstract/Free Full Text].

SCOTT, M. P., A. J. WEINER, T. I. HAZELRIGG, B. A. POLISKY, and V. PIRROTTA et al., 1983  The molecular organization of the Antennapedia locus of Drosophila. Cell 35:763-776[Medline].

SIGRIST, C. J. A. and V. PIRROTTA, 1997  Chromatin insulator elements block the silencing of a target gene by the Drosophila Polycomb response element (PRE) but allow trans interactions between PREs on different chromosomes. Genetics 147:209-221[Abstract].

SIMON, J., 1995  Locking in stable states of gene expression: transcriptional control during Drosophila development. Curr. Opin. Cell Biol. 7:376-385[Medline].

SINCLAIR, D. A. R., T. A. GRIGLIATTI, and T. C. KAUFMAN, 1984  Effects of a temperature-sensitive Minute mutation on gene expression in Drosophila melanogaster. Genet. Res. 43:257-275.

STRUHL, G., 1982  Genes controlling segmental specification in the Drosophila thorax. Proc. Natl. Acad. Sci. USA 79:7380-7384[Abstract/Free Full Text]<