The Size and Internal Structure of a Heterochromatic Block Determine Its Ability to Induce Position Effect Variegation in Drosophila melanogaster

Corresponding author: Eugene V. Tolchkov, Institute of Molecular Genetics, Kurchatov Sq. 2, Moscow 123182, Russia., teugene{at}img.ras.ru (E-mail)

Communicating editor: S. HENIKOFF

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

In the In(1LR)pn2a rearrangement, the 1A-2E euchromatic segment is transposed to the vicinity of X heterochromatin (Xh), resulting in position effect variegation (PEV) of the genes in the 2BE region. Practically the whole X-linked heterochromatin is situated adjacent to variegated euchromatic genes. Secondary rearrangements showing weakening or reversion of PEV were obtained by irradiation of the In(1LR)pn2a. These rearrangements demonstrate a positive correlation between the strength of PEV of the wapl locus and the sizes of the adjacent heterochromatic blocks carrying the centromere. The smallest PEV-inducing fragment consists of a block corresponding to ~10% of Xh and containing the entire XR, the centromere, and a very proximal portion of XL heterochromatin. Heterochromatic blocks retaining the entire XR near the 2E region, but lacking the centromere, show no PEV. Reversion of PEV was also observed as a result of an internal rearrangement of the Xh blocks where the centromere is moved away from the eu-heterochromatin boundary but the amount of X heterochromatin remaining adjacent to 2E is unchanged. We propose a primary role of the X pericentromeric region in PEV induction and an enhancing effect of the other blocks, positively correlated with their size.

CHROMOSOMAL rearrangements that juxtapose euchromatin and heterochromatin induce a mosaic inactivation of euchromatic genes known as position effect variegation (PEV; see for reviews HENIKOFF 1990 ; LOHE and HILLIKER 1995 ; WEILER and WAKIMOTO 1995 ; ZHIMULEV 1998 ). Recently, a number of euchromatic genes acting as enhancers or suppressors of PEV have been molecularly characterized and have been shown to encode for chromosomal proteins (GRIGLIATTI 1991; REUTER and SPIERER 1992 ; LOHE and HILLIKER 1995 ; JENUWEIN et al. 1998 ). By contrast, the molecular nature of PEV-inducing heterochromatic blocks is still largely unknown. Analysis of secondary rearrangements resulting in reversion of PEV indicates that remnants of heterochromatin in the regions of eu-heterochromatic junctions may be insufficient to maintain inactivation (TARTOF et al. 1984 ; POKHOLKOVA et al. 1993 ; MAKUNIN et al. 1995 ; TOLCHKOV et al. 1997 ).

Early studies indicated that the strength of PEV correlates positively with the size of the heterochromatic block relocated near euchromatin (PANSHIN 1938 ). However, recent studies demonstrated that the strength of PEV of white does not correlate with the amount of heterochromatin adjacent to this gene (HOWE et al. 1995 ).

The different ability of different heterochromatic blocks to cause PEV has been widely discussed (SPOFFORD 1976 ). An important role of centromeric regions was proposed in the studies of PEV of heterochromatic genes removed from their heterochromatic location (HILLIKER and SHARP 1988 ) but this hypothesis was later abandoned (EBERL et al. 1993 ; WEILER and WAKIMOTO 1995 ). Recently, it has been shown that the strength of PEV may depend not only on the size of the heterochromatic blocks adjacent to the euchromatic genes, but also on the distance of the eu-heterochromatic junction from the main heterochromatic block (TALBERT et al. 1994 ; HENIKOFF et al. 1995 ; HENIKOFF 1997 ). It was proposed that failure of this junction to coalesce with the chromocenter may help the gene to escape inactivation (HENIKOFF et al. 1995 ). A similar hypothesis was put forward in the case of the inactivation of heterochromatic genes due to their transposition into a euchromatic environment (WAKIMOTO and HEARN 1990 ; EBERL et al. 1993 ).

Here we study a set of related rearrangements causing different extents of PEV of the genes located in the 2BE region. We have examined the effect on PEV not only of a stepwise decrease in the size of the cis-acting X-linked heterochromatic blocks, but also of their internal structure and their distance from the chromocenter.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Fly stocks and crosses:
The pericentric inversion In(1LR)pn2a (Fig 1A) and its derivatives are maintained over the FM7 balancer, marked with y^31dsc⁸w^av^Of B g⁴. To evaluate the effect of the Y chromosome on wapl variegation in pn2a derivatives R/FM7, y^31dsc⁸w^av^Of B g⁴ females (R, secondary rearrangements) were crossed to y l(1)74 pn w^a ct v · B^SY^L/w⁺ Y males where l(1)74 is a lethal allele of the wapl gene (GVOZDEV et al. 1975 ), hereafter referred to as l74. For details of mutants, balancers, and chromosome deficiencies, see LINDSLEY and ZIMM 1992 .

View larger version (26K): In this window In a new window Download PPT slide   Figure 1. Origin and structure of the In(1LR)pn2a rearrangement (A) and genetic systems for recovery of secondary rearrangements (B and C). (A) Diagrams representing the Batumi X chromosome (top) and the In(1LR)pn2a chromosome (bottom). Heterochromatin is depicted as a solid block; euchromatin is depicted as a thin line. Euchromatic boundaries are indicated according to the Bridges polytene map. The numbers above each diagram indicate vital loci in the region saturated by lethals (GVOZDEV et al. 1975 ). Pgd, wapl, and kz correspond to vital loci 1, 2, and 8, respectively; pn corresponds to the nonvital locus 4. Vertical arrows indicate the breakpoints resulting in the pn2a rearrangement. The region uncovered by Df(1)64c18 is also shown. C, centromere. XR, right arm of X chromosome. (B) Selective system for the recovery of secondary rearrangements. Rev, reversions. Irradiated In(1LR)pn2a/BSY males were crossed to y l74 pn wa ct v/y Df(1)64c18 pn wa ct v females. Viable Rev(pn2a)/l74 and Rev(pn2a)/Df(1)64c18 females were crossed to FM7, y31d sc8 wa vOF B g4/Y males to balance putative revertant chromosome. Recessive markers were used to distinguish Rev(pn2a) from recombinant chromosomes in the progeny of Rev(pn2a)/l74 or Rev(pn2a)/Df(1)64c18 females. (C) Modified scheme of crosses to obtain secondary rearrangements using the y l74 pn wa ct v · BSYL chromosome.

Selective system to obtain secondary rearrangements:
Secondary rearrangements (R) were induced by irradiation (4 kR) of pn2a/B^SY males (Fig 1B). Irradiated males were crossed to y l74 pn w^a ct v/y Df(1)64c18w^act v females. Selection was based on the strong inactivation of several vital genes due to PEV in the pn2a chromosome. Suppression of PEV in secondary rearrangements restores gene activity and results in survival of the R/l74 or R/Df(1)64c18 females carrying the rearrangement. Selected females were crossed to FM7, y^31dsc⁸w v^OfB g⁴Y males to balance the revertant chromosome. Recessive markers (w^a, ct, v) were introduced to distinguish R from a recombinant chromosome in the progeny of R/l74 or R/Df(1)64c18 females. Using this selective system, only rearrangements causing complete PEV loss were obtained.

To select rearrangements resulting in incomplete suppression of PEV, Y chromosome material was introduced into the system using the l74 · B^SY^L chromosome (see above; Fig 1C). As a result, eclosed pn2a/l74 · B^SY^L females survived, although most individuals showed an extreme wapl phenotype ("cut wings," angle between wings amounts to 180°; see below) and died before laying eggs. Selected females were balanced over FM7 by crossing to FM7/Y males.

Estimation of PEV:
wapl locus inactivation was tested in R/l(1)74 females carrying a secondary rearrangement over a normal X chromosome. The following phenotypic traits were monitored: "wings apart"; cut wings (or excised); irregular rows of facets; and decrease of viability.

The penetrance of the cut wings phenotype was measured as percentage of wings with cuts among all individuals. The wings apart phenotype was quantitatively evaluated both by counting the percentage of flies with an enlarged angle between wings and by measuring the angle between wings. Angle was measured using a scaled plate, where micro-dials were graduated in 10°. Increase of angle value over 30° (experiment 1) or over 45° (experiment 2) was taken as an indication of a mutant wapl phenotype (see Table 2), as compared to the angle of 20–30° in wild-type flies. The "irregular facets" phenotype was measured as the mean percentage of eye surface with irregular ommatidial packing. A disturbance of faceting was taken into account starting with a detectable spot comprising 2–3% of eye surface (~10 irregularly arranged facets out of 400). The eye surface was arbitrarily divided into eight sectors, each encompassing ~12–13% of the surface. The ratio of mutant to the total eye surface was estimated using this unit of evaluation. The low level of variegation was estimated directly by counting the number of facets in the area of mutant tissue. Fortuitous faceting disturbance amounts barely to 0.03% of eye surface in Batumi-L and other wild-type stocks.

The inactivation of the dor locus was tested in R/dor^l females. PEV intensity of the dor gene was evaluated as the percentage of eyes with yellowish spots as well as by estimation of the yellow area of the eye surface.

Analysis of segregation between the Y chromosome and recombinants containing parts of different rearrangements:
To analyze the segregation between the y²Y43T chromosome and the recombinant (Rec) sc^S1pm141^d (see legend to Fig 4A) or r16^pm141^d (see legend to Fig 4C) chromosomes, C(1),dor/Dp(1)y²Y43T females were crossed to Rec/Dp(1)y²Y43T males. Chromosome y²Y43T carries the 1A-2F duplication covering dor. Appearance of dor females in the progeny indicates the occurrence of X-Y nondisjunction in males.

View larger version (68K): In this window In a new window Download PPT slide   Figure 2. Diagramatic representation of rearrangements after Hoechst 33258 staining. Heterochromatic segments are indicated according to GATTI et al. 1994 ; euchromatin is depicted as a thin line. Solid segments indicate bright fluorescence; cross-hatched segments indicate moderate fluorescence; hatched segments indicate dull fluorescence; and open segments indicate no fluorescence. The horizontal bracket indicates the pericentromeric Quinacrine-fluorescent region (h33). The h34 segment is differentiated by N-banding. C, centromere.

View larger version (167K): In this window In a new window Download PPT slide   Figure 3. Cytological characterization of rearrangements by sequential Quinacrine staining, Hoechst staining, and N-banding. The numbers identify the heterochromatic regions (cf. with Fig 2 and GATTI et al. 1994 ). p and d indicate the proximal and distal portions of a broken region, respectively. In e note the reduced h45 N-band present on the second chromosome involved in the r24 rearrangement (45p), as compared to a normal-sequence second chromosome (45).

View larger version (33K): In this window In a new window Download PPT slide   Figure 4. Genetic analysis of rearranged chromosomes. Open circles indicate the centromere. Recombination events are indicated by dashed oblique lines. Cytological localization is according to the Bridges map. Other designations are as in Fig 2. (A) Recombination between m141 and In(1)scS1. The recombinant transmittable chromosome scS1pm141d carries no NO (see text for further details). The y2Y43T chromosome carries a 1A-2F duplication covering the 1AB deficiency in the recombinant chromosome. (B) Mapping of the centromere near the 2E region in the r24 rearrangement involving chromosomes 1, 2, and 3 (see Fig 2E). The wavy line represents euchromatin of chromosome 2. We infer that the centromere is not contained in the Q-bright material transposed to the 99A region because we failed to recover the putative recombinant chromosome Rec (bottom) lacking the 1A-2E region. The deletion of the 1A-2E region may be detected by a recovery of y2 males in a progeny of cross[r24/FM7 x Df(1)pn2a·BSYL/y2Y43T]. No y2 males were detected among 369 y+ males (genotype r24/y2Y43T) and 236 FM7/y2Y43T males, although ~20 y2 [Df(1)1A-2E/y2Y43Y] recombinant males may be recovered as a result of homologous recombination in the 86C-98F region (10 cM). Interchromosomal effect of FM7 balancer may compensate for the decrease of recombination because of rearrangement. The absence of y2 males indicates that putative recombinant chromosome Rec is acentric. (C) Centromere and NO mapping in the r16 rearrangement. Heterochromatin of chromosome 1 is shaded. Recombination between r16 and m141 (see Fig 2B and Fig G) leads to a recombinant chromosome lacking the 1A-2E region. The observed normal segregation of centric recombinant and y2Y43T chromosomes can be explained by the presence of NO fragment adjacent to the 2F euchromatic region, taking into account the role of rDNA in segregation of X and Y chromosomes (MCKEE 1996 ). Thus, a heterochromatic break of the r16 rearrangement must have occurred within the NO region, splitting it into two unequal parts (Fig 2G).

Cytological analysis of polytene chromosomes:
Fertile R/Y males (r9, r30, r20, pn2a, m100, and m141) and R/FM7 females (males carrying r4, r24, r16, and r35 were sterile or inviable) were crossed to y ac sc w females or males, respectively. Salivary glands were dissected from R/y ac sc w third instar female larvae with yellow malpighian tubules (w⁺ phenotype).

Mitotic chromosomes:
Preparation and sequential staining of mitotic chromosomes with Quinacrine, Hoechst, and N-banding were carried out as described (GATTI et al. 1994 ).

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Recovery of secondary rearrangements:
The In(1LR) pn2a original rearrangement (hereafter referred to as pn2a; ILYINA et al. 1980 ; TOLCHKOV et al. 1984 ) has a euchromatic breakpoint in 2E, which disrupts the Vinculin gene (ALATORTSEV et al. 1997 ), and a heterochromatic breakpoint in the right arm of the X chromosome (TOLCHKOV et al. 1997 ; Fig 1A). As a consequence, the 1A-2E segment, encompassing seven identified loci (including pn, wapl, and Pgd; GVOZDEV et al. 1975 ), is transposed adjacent to the centric heterochromatic block. These genes exhibit strong PEV (ALATORTSEV et al. 1982 ; TOLCHKOV et al. 1984 , TOLCHKOV et al. 1997 ). Inactivation spreads far from the breakpoint, affecting the dor locus situated ~700 kb apart from heterochromatin; kz and four additional loci adjacent to XR telomere in pn2a show no PEV (Fig 1A). Inactivation of euchromatic genes in the pn2a inversion resulted in inviability of pn2a/l74 as well as pn2a/Df(1)64c18 females (Fig 1B). l74 is a lethal allele of the wapl gene and Df(1)64c18 uncovers pn and several vital loci designated as complementation groups 5, 6, and 7 (Fig 1A).

Selection of pn2a secondary rearrangements with a decrease or complete loss of PEV was based on the restoration of activity of several vital genes affected by PEV in the pn2a chromosome. Two different cross schemes have been used to select secondary rearrangements with complete (Fig 1B) and partial (Fig 1C) reversion of PEV. Using the first scheme of selection, 4 secondary rearrangements were obtained (Fig 1B), while 18 secondary rearrangements were selected using a modified scheme of screening (Fig 1C). In a set of secondary rearrangements, the correlation between the decrease of PEV and the sizes and structures of cis-acting heterochromatic blocks was studied.

The structures of rearrangements were defined by mitotic chromosome staining as well as by genetic and polytene chromosome analysis. The centromere region (h33) and the eu-heterochromatic boundary (2E-h34) were identified by Quinacrine staining and N-banding, respectively. Centromere position was determined by genetic tests when the Quinacrine region was split. The nucleolus organizer region (NO) h29 is detected as a constriction, and the other Xh segments are identified by their peculiar fluorescence patterns after Hoechst staining. These results are summarized in Fig 2 (schemes of rearrangements) and Fig 3 (mitotic chromosomes).

On the basis of the structure of the heterochromatic block juxtaposed to 2E (see below), the rearrangements can be grouped into three classes.

Rearrangements with decreasing amounts of the cis-acting heterochromatic block (class 1):
m141 is an inversion with a euchromatic breakpoint in 2F and a heterochromatic breakpoint in h26 that separates the distal portion of h26 from the main block of Xh (Fig 2B and Fig 3B). Detection of free recombination with In(1)sc^S1 and recovery of males carrying the 1AB deletion (Fig 4A) demonstrate the presence of the 2F-20F inversion. The presence of the 1AB deletion was inferred from the recovery of the yellow² males carrying a recombinant chromosome and the y²Y43T duplication (Fig 4A). The observed high frequency of nondisjunction of sex chromosomes in these males was attributed to a deletion of NO in the recombinant chromosome, taking into account the role of NO region in sex chromosome disjunction (MCKEE 1996 ).

r4 is a complex rearrangement that splits the Xh block into a proximal part (h30-h34^p) that remains adjacent to 2E and a distal part (h26-h29) that is separated from the centric segment by the 13B-20F euchromatic fragment (Fig 2C and Fig 3C). Polytene chromosome analysis revealed also the presence of a translocation, T(1;2)2A;55F5-12, which in mitotic figures results in an increased length of the euchromatic fragment attached to the proximal Xh block (Fig 3C).

m100 is a translocation, T(1;2)h32/h33;42F1-3, that results in a small acrocentric chromosome, carrying the N-banded h34^p region and the whole Quinacrine-stained block (h33) attached to the euchromatic 1A-2E region (Fig 2D and Fig 3D). The rest of the Xh block is transposed to 2R and is split into two fragments by an additional inversion, In(1)5D;h26 (Fig 2D and Fig 3D).

r24 is a complex translocation involving chromosomes X, 2, and 3. In mitotic chromosome preparations the X-linked Quinacrine-bright material appears split into two equal parts, located at the opposite ends of a chromosome containing two Xh blocks separated by a euchromatic segment (Fig 3E). Polytene chromosome analysis allowed us to show that this euchromatic segment comprises both chromosome 2 and 3 bands (Fig 2E). Moreover, genetic analysis (see Fig 4B) confirmed that the X centromere is located near the 2E euchromatic region in this rearrangement.

Internal rearrangements that relocate distal Xh adjacent to acentric XR material (class 2):
r9 is a secondary rearrangement of the h26-h33 region. The h26 region is flanked by the NO constriction (h29) and the N-banded h34^p region, while the Quinacrine (Q)-bright band is located adjacent to the dull fluorescent segment h27-h28 (Fig 2F and Fig 3F). Analysis of NO-mediated recombination between r9 and the Y chromosome corroborates the presented structure (data not shown).

r16 is a complex rearrangement, Tp(1;1)h33;h29;2F+ In(1)16;h26/h27 + T(1;2), resulting in a marked decrease of the heterochromatic mass adjacent to the 2E region (Fig 2G). The transposed centric block comprises most (75–80%) of the X-linked Q-bright material, the rest of which is detected near h34^p (Fig 3G). Recovery of the r16^pm141^d recombinant chromosome carrying no small block of Q-bright material (data not shown) allowed us to map the centromere within the major Q-stained band (Fig 4C). NO appears to be located near the tiny Q-band and h34^p (Fig 3G), but genetic analysis indicated that a piece of NO must be left adjacent to the h30 region (Fig 4C).

Rearrangements that separate the acentric XR region from the main heterochromatic block (class 3):
Here the structures of three representative rearrangements are presented:

• r30 is a previously described (TOLCHKOV et al. 1997 ) inversion whose structure is diagramatically reported in Fig 2H.

• r35 is a complex rearrangement involving chromosome 3 (Fig 2I). The 1A-2E region is translocated to the rearranged third chromosome together with an adjacent Xh segment consisting of region h34^p and an acentric fragment of the Q-bright h33 region (Fig 2I and Fig 3I). The structure of the main Xh block associated with 3L is also modified by an additional inversion involving regions h28 and h29 (Fig 2I and Fig 3I).

• r20 is a secondary inversion with a heterochromatic breakpoint in the distal portion of h33 and a euchromatic breakpoint in region 1B. As a result, a small heterochromatic segment comprising h34^p and h33^d is separated from the main Xh block containing the centromere by the 2E-1B euchromatic region (Fig 2J and Fig 3J).

A total of 13 rearrangements leading to full suppression of PEV carry a putative heterochromatic breakpoint near the 2E-heterochromatin boundary and represent translocations of acentric 1A-2E region to distal euchromatin of autosomes (10 rearrangements) or X chromosomal inversions (3 rearrangements). These rearrangements were characterized by polytene chromosome analysis and represent the most numerous group of rearrangements.

PEV of the wapl locus in secondary rearrangements:
The strength of PEV exerted by the various rearrangements was tested by studying the degree of wapl locus inactivation in R/l74 female progeny from crosses R/FM7 x l74/w⁺Y (see MATERIALS AND METHODS). The following phenotypic traits were detected: wings apart, cut wings (or excised), and irregular ommatidial packing. Decrease of viability was also evaluated.

In experiment 1, stocks carrying a nonmarked Y chromosome were used (except the r24 stock, carrying a B^SY chromosome). However, in some rearrangements a high level of X chromosome nondisjunction was observed (Table 1), resulting in females carrying a Y chromosome, a well-known suppressor of PEV. To avoid artifacts in measuring PEV, all the stocks used for experiment 2 carried a B^SY chromosome. The percentage of eye surface with irregular faceting in R/l74^· B^SY^L females could not be estimated due to effects of the B^S marker (Bar eyes). However, correlation of the strength of irregular faceting with the wing phenotypes indicates that faceting is also affected by PEV. In Table 2 the rearrangements are reported in order of decreasing strength of PEV. As can be seen, the introduction of a Y chromosome decreases wapl inactivation in all cases.

A stepwise reduction of the heterochromatic block adjacent to the 2E region (class 1 rearrangements) causes a gradual decrease of PEV:
All four rearrangements (m141, r4, m100, and r24) belonging to class 1 exhibit a positive correlation between the size of the heterochromatic block remaining adjacent to 2E and the extent of wapl inactivation.

As can be seen in Table 2, viability is affected only by the secondary inversion m141, which, however, exhibits a significant increase of viability as compared to the original rearrangement pn2a. In both cases, addition of a Y chromosome fully restores viability (Table 2, experiment 1).

Judging from the penetrance of the cut wings phenotype (percentage of wings with cuts among total scored flies), the extent of wapl inactivation diminishes gradually in the four rearrangements belonging to class 1 (m141, r4, m100, and r24). The percentage of cut wings drops from 94% in m141/l74 females to 35% in r24/l74 females (experiment 1). These results were confirmed in experiment 2, although a comparison of r4/l74 with m100/l74 females revealed no significant difference in the number of cut wings. The number of cut wings drastically diminished after addition of a Y chromosome (experiment 1). Experiment 1 also demonstrates a significant decrease in the number of cut wings among the flies bearing the m141 rearrangement as compared to the individuals with pn2a inversion.

The decrease of PEV in m141 was confirmed when the penetrance of the wings apart phenotype was evaluated (Table 2). A further decrease in the number of wings apart flies was observed for the r4 and r24 rearrangements while the estimation of the mean angle between wings demonstrated the strong decrease of wapl inactivation occurring in r24 as compared to m100.

Evaluation of the expression of the irregular facets phenotype confirmed the progressive decrease of PEV in these secondary rearrangements, from m141 to r4.

Thus, a careful estimation of the pleiotropic effects of wapl inactivation allows us to conclude that the extent of variegation is correlated with the quantity of adjacent X-linked heterochromatin. We detected a gradual decrease of wapl variegation for four rearrangements (m141, r4, m100, and r24), where the heterochromatic block remaining adjacent to 2E was stepwise reduced. A pronounced wapl variegation is detected even in the r24 rearrangement carrying barely 10% of the whole Xh block, comprising approximately half of the Quinacrine-positive band h33, the centromere, and the N-banded region h34 (Table 2; Fig 2E). Elimination of the heterochromatic segments h26-h27 distal to the NO region (h29) resulted in a noticeable decrease of variegation strength. Thus, practically the whole Xh region appears to contribute to the strength of PEV.

Internal rearrangements in the Xh block (class 2) suppress PEV:
The extent of PEV was shown to be decreased drastically in r9 or even eliminated in r16, although the whole (r9) or a significant part (r16) of the Xh block remained adjacent to 2E in these two rearrangements (Fig 2F and Fig G). In the r9 rearrangement, an internal inversion of most of the Xh block, determining a separation of the centromere region from the eu-heterochromatic boundary, results in a drastic weakening of wapl variegation (Fig 2F; Table 2). In fact, inactivation of the wapl gene was detected only by disturbance of ommatidial packing affecting small areas of eye (from fractures of a single row of facets up to 5% of altered faceting, with a 2% mean value of mutant eye surface). No wapl wing phenotype was observed in r9/wapl females. To detect wapl inactivation distinctly, r9/wapl females were produced by reciprocal cross (l74/FM7 x r9/Y), taking into account the known paternal effect resulting in enhanced variegation (SPOFFORD 1976 ). Of these r9/l74 females, 4% exhibited cut wings.

No wapl variegation was observed in r16, where the Xh block juxtaposed to 2E is rearranged and lacks the centromere (Fig 2G). However, the amount of heterochromatic material that remains near 2E in r16 is larger than in r24, a secondary rearrangement that causes strong variegation. These results indicate that the strength of PEV may be determined by the specific arrangement of the X heterochromatic segments.

Small acentric pieces of XR juxtaposed to the 2E region (class 3 rearrangements) do not induce variegation:
Most of the secondary rearrangements belonging to class 3 have no centromere in the XR heterochromatic block that remains juxtaposed to 2E and cause no variegation. In these rearrangements centromere is separated from the XR portions by eight or more sections of the Bridges map. In particular, no variegation of the wapl gene was detected in r30 (TOLCHKOV et al. 1997 ) as well as in r35, where the 2E region is associated with the h34 and h34-h33 material, respectively (Fig 2H and Fig I). The r20 rearrangement (Fig 2J) is an exception in that it causes a variegation of the wapl gene that is stronger than in r24 (Table 2). A possible explanation of this unexpected result will be provided in the next section.

Peculiarities of the r20 rearrangement:
Several observations exclude the possibility that in r20 inactivation may reach wapl (2E region) starting from the main heterochromatic block containing the centromere (Fig 5). First, polytene chromosome analysis revealed no heterochromatization of the 1B-2C region juxtaposed to the centromere. Second, wapl variegation is stronger than dor variegation both in r20 and in the original pn2a rearrangement. The area of irregular facets achieves one-fifth of the eye surface in r20/wapl females (Table 2), in which r20 is maternally inherited, whereas barely 2% of eye surface with altered pigmentation is detected in r20/dor females (Table 3; Fig 5), despite the fact that in this case r20 is paternally inherited and should therefore exert an enhanced variegation (SPOFFORD 1976 ). Thus, inactivation in both r20 and pn2a appears to start from region h34, consistently affecting only the expression of the juxtaposed wapl gene (Fig 5).

View larger version (25K): In this window In a new window Download PPT slide   Figure 5. Patterns of the dor (percentage of yellowish eye surface) and wapl (percentage of eye surface with disturbed faceting) loci inactivation in In(1LR)pn2a (top) and r20 (bottom). The percentages of mutant eye tissue are depicted as blackened sectors. The number of essential loci uncovered by Df(1)sta and Df(1)JA52 is indicated above the bars designating the deficiencies (LINDSLEY and ZIMM 1992 ). Arrows indicate the direction of inactivation.

Finally, comparison of the viability of females carrying the r20 rearrangement over a series of deficiencies uncovering Bridges sections 1 and 2 strongly supports our hypothesis that in r20 variegation spreads from the small heterochromatic block h34 (Fig 5; Table 4). Deficiency Df(1)JA52 removes 4 vital loci (LINDSLEY and ZIMM 1992 ). Inviability of both r20/Df(1)JA52 and pn2a/Df(1)JA52 females indicates a strong inactivation of several vital loci localized distal to the wapl gene (Fig 5; Table 4). By contrast, no decrease in viability of r20/Df(1)sta females was observed, although this deletion is more extended and uncovers at least 17 vital loci (LINDSLEY and ZIMM 1992 ) that in r20 are situated closer to the main heterochromatic block than the dor locus (Fig 5). Thus, we have compelling evidence that in r20 wapl inactivation starts from the h34 block. The inability of comparably small (r30) or even larger acentric heterochromatic blocks (r35) to induce PEV might depend on their distance from the bulk of X heterochromatin and thus their distance from the chromocenter (see DISCUSSION).

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

In contrast with the advanced understanding of trans-acting factors, little is known of cis-acting requirements for PEV. Here we have addressed this problem by dissecting a heterochromatic block capable of inducing PEV. In the In(1LR)pn2a rearrangement virtually the whole X-linked heterochromatin is moved adjacent to the euchromatic region 2E, causing inactivation of genes located up to ~600 kb from the breakpoint region (TOLCHKOV et al. 1984 , TOLCHKOV et al. 1997 ). We used an efficient genetic system to recover partial or full PEV revertants represented by secondary rearrangements. This approach allowed us to correlate the size and the internal structure of different portions of Xh with the strength of PEV affecting the wapl gene, spaced ~50 kb from the eu-heterochromatin boundary.

The strength of PEV positively correlates with the size of the cis-acting heterochromatic blocks:
We characterized PEV strength by both the wing phenotypes and the relative size of mutant eye surface. PEV strength was shown to decrease with the size of the adjacent centromere containing heterochromatic block in rearrangements pn2a, m141, r4, m100, and r24 (Table 2; Fig 6). These results suggest that the whole X-linked heterochromatic block, including nine cytologically defined segments (GATTI and PIMPINELLI 1992 ; GATTI et al. 1994 ), is involved in PEV induction. Removal of the distal h26 section of heterochromatin in the m141 rearrangement is enough to produce a marked decrease of variegation (Table 2). Thus, a relatively small Xh region, placed at ~15–20 Mb from the eu-heterochromatic boundary, appears to play a significant role in the inactivation of euchromatic genes. Removal of Xh segments containing the NO (r4), the 1.688 complex satellite (m100), and virtually the entire heterochromatic material of the left arm (r24) resulted in further and gradual weakening of inactivation (Fig 6).

View larger version (27K): In this window In a new window Download PPT slide   Figure 6. Correlation between decrease of wapl gene inactivation and reduction of the cis-acting heterochromatic block. The arrangement of NO, complex satellite 1.688, and AATAT and AAGAG simple satellite sequences is shown. The numbers separated by slashes indicate the percentage of "wing with cuts" and "disturbed eye faceting areas," respectively. Segments h34 and h33 are disproportionally enlarged.

Evidence of a positive correlation between the size and the PEV-inducing potential of a given heterochromatic block had been reported previously for the derivatives of a rearrangement transposing the white gene to the heterochromatin of chromosome 4 (PANSHIN 1938 ). The diminishing of heterochromatic masses adjacent to the white locus was indirectly estimated judging by the extent of coupled variegation of chromosome 4 heterochromatic ci gene. More recently, however, the strength of white variegation was shown to be independent of the amount of adjacent 2h material (HOWE et al. 1995 ). The apparent contradiction between this and our results might be explained by taking into account that in the experiments by Howe and co-workers variegation was estimated in rearrangements with different euheterochromatic junctions, whereas in our rearrangements this junction is unchanged. DORER and HENIKOFF 1997 presented a similar conclusion in their studies of cis-silencing effects caused by transgene arrays. These arrays are considered as heterochromatic insertions and the degree of silencing was shown to be dependent on the size of arrays.

The observed conspicuous contribution to PEV of the small distal h26 fragment, comprising a negligible amount (~5%) of the genomic heterochromatin, suggests that X heterochromatin may act as an autonomous unit, relatively independent of the rest of heterochromatin. Interestingly, recent fluorescent in situ hybridization (FISH) analysis of the arrangement of heterochromatic components within interphase nuclei indicates that each chromosome has its own "heterochromatic compartment" in the Drosophila genome (DERNBURG et al. 1996 ).

Taking into account that dominant trans-acting modifiers of PEV are not known on the X chromosome, one may suppose that some of the observed effects might be attributed to the action of dominant trans-acting suppressors induced by irradiation in autosomes. However, the following arguments favor decreased cis-action of truncated heterochromatic blocks.

First, the observed conspicuous correlation between the size of the adjacent heterochromatic block and the degree of silencing of the wapl gene is difficult to explain by the action of trans-modifiers. Second, no rearrangements without X chromosome heterochromatin breakage were detected among the 22 selected reversions. The cases of full reversion of PEV (r16 and r35) can hardly be attributed to the effect of a putative extremely strong autosomal suppressor, since the tested strong autosomal suppressors [Su-var(3)9 and others] cause only negligible effects, resulting in no more than a 10% decrease of the number of cut wings in pn2a and r24, as compared to the drastic influence of the heterochromatic Y chromosome resulting in full reversion of PEV in r24.

Removal or distancing of the centromere region from the eu-heterochromatic boundary results in reversion or attenuation of PEV:
PEV was not detected in r30 (TOLCHKOV et al. 1997 ) and r35 rearrangements (Fig 6) where the 2E polytene region is situated near XR material lacking the centromere. This result could depend on the sensitivity of our genetic test, which might be inadequate to detect drastically weakened PEV induced by small heterochromatin masses. However, PEV was observed in r24, a rearrangement bearing a comparably small, but centromere-containing, heterochromatic block near 2E. The following indirect observations also suggest a crucial role of centromere in inactivation. The size of the heterochromatic block near 2E is not changed in r9, but the rearrangement results in spacing of the centromere from the eu-heterochromatic junction (Fig 7) and is associated with negligible wapl variegation. No PEV was detected in r16, where a conspicuous heterochromatic block lacking the centromere remains in the vicinity of 2E (Fig 7). Thus, the preservation of a significant mass of the Xh block may be insufficient to induce PEV on the adjacent euchromatic genes, if it lacks the centromere. The disappearance of PEV in r16 and its drastic weakening in r9 may be caused either by peculiar properties of the heterochromatic blocks fused to the h34 segment or by a disruption of the continuity of the heterochromatic region encompassing h34 and the centromere (see below).

View larger version (18K): In this window In a new window Download PPT slide   Figure 7. PEV and structures of rearrangements pn2a, r9, and r16. Arrow lengths arbitrarily indicate the strength of PEV exerted by rearrangements. The sizes of the putative basic (B) and modulating (M) regions are indicated.

If the centromere (h33 region) is the inactivation center, it is likely that its separation from the 2E euchromatin would result in complete reversion of inactivation. However, the absence of strong inactivation of the euchromatic genes localized near the centromeric heterochromatin in r20 (Fig 5; Table 4) contradicts this hypothesis. Alternatively, inactivation may be exerted via the concerted action of the h34 segment and the centromere region. Separation of the components of this putative integral block may result in the reversion of PEV observed in r30 and r35 (Fig 6). Actually, substitution of the centromere-associated region with other Xh segments (r16) fails to restore inactivation (Fig 7) and spacing of the centromere region from the eu-heterochromatic junction results in similar effects (r9).

We suppose that the centromere affects by distance-dependent interactions the state of at least some segments of the heterochromatic block harboring it. The published data show that the above hypothesis may be true for other models as well as for our model. Centromere separation resulted in variegation of the heterochromatic peach gene in Drosophila virilis, although the bulk of heterochromatin remains attached to the peach gene in the T(3;5)pe^m5 rearrangement (BAKER 1954 ). The participation of the centromere in the maintenance of the light (lt) gene activity can be deduced by the localization of the breakpoints inducing lt variegation. Actually, variegation of lt, localized in the distal 2L heterochromatin, is observed as a result of proximal breakages in 2L heterochromatin, just distal to the centromere (LOHE et al. 1993 ), whereas no lt variegation was detected as a result of 2R heterochromatin separation from the centromere.

The lack of full reversion of white variegation in a T(1;4)w^m11 derivative resulting in centromeric region detachment (PANSHIN 1938 ) indicates that the effect of the centromere on PEV intensity may vary in different rearrangements. The nature and size of the heterochromatic block, the localization of the centromere inside this block, and the distance between the breakpoint and the reporter gene may modulate centromere effects.

PEV in r20 and association of heterochromatic blocks:
The strong PEV observed in r20 is intriguing, since an acentric piece of XR in all the other rearrangements of class 3 is incapable of causing variegation of the adjacent 2E region. The sizes of blocks carrying the h34 segment in r20 and in r30 are comparable, while the size of the h33d Q-bright segment adjacent to 2E is even larger in r35 than in r20 (Fig 3). On the other hand, the distance of the eu-heterochromatic boundary from the pericentric heterochromatin amounts to 8 or more sections of the Bridges map in r30, r35, and other rearrangements of class 3, and only to 1.5 sections in r20. We suppose that in r20 a contact between the small acentric block and the chromocenter can occur, which in the other rearrangements of this class is impaired by the distance of the acentric XR block from the centromeric region. Actually, close pairing of the h34 region with the chromocenter was consistently detected in 100% of salivary gland nuclei bearing the r20 chromosome, where the 1B-2E polytene region looks like a loop associated with the chromocenter. Other rearrangements of this class (r30 and r35) in heterozygotes with structurally normal chromosomes show no notable association of the separated heterochromatic blocks with the chromocenter (~40 nuclei were analyzed for each rearrangement). This pairing would be the basis of the PEV-inducing capability observed in r20. These observations are in agreement with previous data indicating that PEV strength depends on the distance between variegating breakpoints and pericentromeric heterochromatin (WAKIMOTO and HEARN 1990 ; EBERL et al. 1993 ; KONEV 1994 , KONEV 1995 ; TALBERT et al. 1994 ; HENIKOFF et al. 1995 ; CSINK and HENIKOFF 1996 ).

The r20, r30, and r35 rearrangements carry near the 2E region the h34 block, containing the AAGAG satellite (LOHE et al. 1993 ; TOLCHKOV et al. 1997 ). The block of AAGAG satellite of comparable size (CSINK and HENIKOFF 1996 ; DERNBURG et al. 1996 ) inserted into the bw gene (bw^D) causes trans-inactivation of bw⁺ in bw^D/bw⁺ heterozygotes, but no cis-inactivation of neighboring genes (TALBERT et al. 1994 ). However, bw^D insertion starts to cause cis-inactivation if the distance between the insertion and the chromocenter is shortened (TALBERT et al. 1994 ). Thus, h34 block and bw^+D insertion are similar in their ability to cause cis-inactivation.

The strong effect of a small heterochromatic segment on PEV:
A comparison of the variegation strength in r4, m100, and r24 rearrangements reveals the crucial role of a restricted block of pericentric Xh in inducing PEV. The difference in PEV severity between m100 and r4 is modest, consisting only in a decreased amount of mutant eye surface in m100, with other traits showing no significant differences of variegation (Table 2; Fig 6). The size of the cis-acting Xh block is much smaller in m100 than in r4; m100 lacks segments h30-h32, encompassing ~11 Mb of the X-specific complex 1.688 satellite (LOHE et al. 1993 ; PIMPINELLI et al. 1995 ). On the other hand, a slight decrease of the cis-acting Xh block in r24 results in a significant decrease of PEV (Fig 6). Roughly estimating, half of the Q-bright material is removed in r24 as compared to m100 (Fig 3). It is assumed that Q-bright regions mainly contain AATAT satellite sequences (LOHE et al. 1993 ). Taking into account the calculated amount of this satellite within Xh (0.4–0.6 Mb; LOHE et al. 1993 ; SUN et al. 1997 ), cis-acting Xh regions in m100 and r24 should differ for ~0.2–0.3 Mb of AATAT satellite. Thus, the 0.3-Mb segment of AATAT satellite present in m100 seems to contribute to PEV much more than the 11-Mb block of complex satellite present in r4.

The amount of Q-bright material is comparable in both the centric r24 and acentric r35 heterochromatic fragments (Fig 3), but only r24 exerts a significant PEV on the adjacent euchromatic genes, thus suggesting that a functional centromere plays a central role in inducing variegation.

Interactions of the Xh segments contribute to PEV:
On the basis of the obtained results, we propose a model in which the pericentric region encompassing AATAT (h33) and AAGAG (h34) satellites can be referred to as basic region (B region), indispensable in causing strong PEV (Fig 7). The distal part of Xh can be considered as a modulating element (M region), capable of significantly enhancing PEV. Disruption of the B region can lead to full reversion (r30, r35, and r16), strong suppression (r9), or substantial decrease of PEV even in the presence of an insignificant decrease of the cis-acting heterochromatic mass (r24). We suppose that putative components of a disrupted B region may cause PEV if they are sufficiently close to each other to be able to associate. This could explain why r20, which carries moderately spaced pieces of the interrupted B region (h34 and h33), can induce strong variegation (Fig 5), while r9, in which the B-region components are more widely spaced and thus almost incapable of associating with each other, can induce only weak PEV. In other words, the spatial interaction of the B-region "modules," rather than its integrity, seems to be important for the induction of strong PEV. We propose that the centromere per se and simple satellite sequences contained in the B region can interact to form a spatial complex with a definite interior architecture that is indispensable in inducing strong PEV. Chromatin conformation in this "inactivation complex" would be dramatically changed, so that the inactivation potential of the complex largely exceeds the additive potentials of its components. The same sequences, once excluded from the inactivation complex, could exert only a weak influence on the neighboring euchromatic genes. If the B-region modules are able to form the putative inactivation complex, then the M-region material can affect it by enhancing its inactivation potential, probably by attracting proteins common to the M and B regions.

HOWE et al. 1995 showed that small reduction of the heterochromatic block size at the expense of euchromatin-adjacent sequences caused drastic changes in w variegation severity. The variegation strength in this case does not correlate with the size of the heterochromatic block. On the basis of the results the authors concluded "that the severity of variegation of the euchromatic w gene was not indicative of the quantity of adjacent heterochromatin ... rather w variegation was sensitive to the nature of the juxtaposed repetitive DNA" (WEILER and WAKIMOTO 1998 ). This conclusion is in accordance with our concept; a relatively small heterochromatic segment adjacent to euchromatin (B region) exerts a crucial effect on the intensity of PEV induced by the large heterochromatic block.

A B-like complex could also be formed in autosomal heterochromatin (Ah). This assumption might explain the differences of bw^D interactions with Xh and Ah (TALBERT et al. 1994 ; HENIKOFF et al. 1995 ). In the case of the bw^D translocations to the X chromosome proximal euchromatin the huge M region lacking simple satellites may impede the interaction of the bw^D insertion with B region. In this case bw^D variegation is suppressed (TALBERT et al. 1994 ). By contrast, Ah with centrally located centromere and dispersed simple satellites segments (LOHE et al. 1993 ) may provide more suitable conditions for such interactions.

Our results imply that strong PEV can be induced as a result of specific interactions between various heterochromatic repeats, whereas sometimes variegation is known to be induced solely due to the presence of a definite number of identical repeated sequences (DORER and HENIKOFF 1994 ). We propose the existence of two types of interaction between heterochromatin segments: specific interactions mediated by the centromere region and simple satellites as well as less specific interactions responsible for the effect of the size of cis-acting heterochromatin (M-B interactions in our terms). The centromere probably affects the state of the whole heterochromatic block or at least some of its segments.

	ACKNOWLEDGMENTS

We thank G. L. Kogan and S. A. Lavrov for providing advice and unpublished results. We thank V. E. Alatortsev, S. G. Balashova, B. O. Glotov, A. B. Devin, E. G. Pasyukova, J. M. Rozovsky, and Y. Y. Shevelev for their relevant comments and suggestions. We thank S. A. Lavrov for his help in preparing figures. We also thank two anonymous reviewers whose comments considerably improved this manuscript. This research was supported by grants from the Russian Foundation of Basic Researches (99-04-48561 and 96-15-98072) and the Russian Program "Frontiers in Genetics" (99-1-069) to V.A.G.

Manuscript received July 2, 1999; Accepted for publication November 24, 1999.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

ALATORTSEV, V. E., E. V. TOLCHKOV, S. Y. SLOBODYANYUK, and V. A. GVOZDEV, 1982  Cellular variegation and antigen disappearance as a result of the position effect of the Pgd locus in Drosophila melanogaster.. Genetika (Russ.) 18:13-23.

ALATORTSEV, V. E., I. A. KRAMEROVA, M. V. FROLOV, S. A. LAVROV, and E. D. WESTPHAL, 1997  Vinculin gene is non-essential in Drosophila melanogaster.. FEBS Lett. 413:197-201[Medline].

BAKER, W. K., 1954  The anatomy of the heterochromatin. Data on the physical distance between breakage point and the affected locus in V-type position effect. J. Hered. 45:65-68[Abstract/Free Full Text].

CSINK, A. K. and S. HENIKOFF, 1996  Genetic modification of heterochromatic association and nuclear organization in Drosophila. Nature 381:529-531[Medline].

DERNBURG, A. F., K. W. BROMAN, J. C. FENG, W. F. MARSHALL, and J. PHILIPS et al., 1996  Perturbation of nuclear architecture by long-distance chromosome interaction. Cell 85:745-759[Medline].

DORER, D. R. and S. HENIKOFF, 1994  Expansions of transgene repeats cause heterochromatin formation and gene silencing in Drosophila. Cell 77:993-1002[Medline].

DORER, D. R. and S. HENIKOFF, 1997  Transgene repeat arrays interact with distant heterochromatin and cause silencing in cis and trans.. Genetics 147:1181-1190[Abstract].

EBERL, D. F., B. J. DUYF, and A. J. HILLIKER, 1993  The role of heterochromatin in the expression of a heterochromatic gene, the rolled locus of Drosophila melanogaster.. Genetics 134:277-292[Abstract].

GATTI, M. and S. PIMPINELLI, 1992  Functional elements in D. melanogaster heterochromatin. Annu. Rev. Genet. 26:239-275[Medline].

GATTI, M., S. BONACCORSI, and S. PIMPINELLI, 1994  Looking for Drosophila mitotic chromosomes. Methods Cell Biol. 44:371-391[Medline].

GRIGLIATTE, T., 1991 Position effect variegation: an essay for nonhistone chromosomal proteins and chromatin assembly and modifying factors, pp. 587–627 in Functional Organization of the Nucleus: A Laboratory Guide, edited by B. A. HAMKALI and S. C. R. ELGIN. Academic Press, San Diego.

GVOZDEV, V. A., S. A. GOSTIMSKY, T. I. GERASIMOVA, E. S. DUBROVSKAYA, and O. Y. BRASLAVSKAYA, 1975  Fine genetic structure of the 2D3-2F5 region of the X chromosome of D. melanogaster.. Mol. Gen. Genet. 141:269-275[Medline].

HENIKOFF, S., 1990  Position effect variegation after 60 years. Trends Genet. 6:422-426[Medline].

HENIKOFF, S., 1997  Nuclear organizations and gene expression: homologous pairing and long range interactions. Curr. Opin. Cell Biol. 9:388-395[Medline].

HENIKOFF, S., J. M. JACKSON, and P. B. TALBERT, 1995  Distance and pairing effects on the brown^Dominant heterochromatic elements in Drosophila. Genetics 140:1007-1017[Abstract].

HILLIKER, A. J., and C. B. SHARP, 1988 New perspectives on the genetic and molecular biology of constitutive heterochromatin, pp. 91–115 in Chromosome Structure and Function, edited by J. P. GUSTAFSON, R. APPELS and R. KAUFMAN. Plenum, New York.

HOWE, M., P. DIMITRI, M. BERLOCO, and B. T. WAKIMOTO, 1995  Cis-effects of heterochromatin on heterochromatic and euchromatic gene activity in D. melanogaster.. Genetics 140:1033-1045[Abstract].

ILYINA, O. V., A. V. SOROKIN, E. S. BELYAEVA, and I. F. ZHIMULEV, 1980  New mutants. Dros. Inf. Serv. 55:205.

JENUWEIN, T., G. LAIBLE, R. DORN, and G. REUTER, 1998  SET domain proteins modulate chromatin domains in eu- and heterochromatin. Cell. Mol. Life Sci. 54:80-93[Medline].

KONEV, A. Y., 1994 Proximity dependent interaction between heterochromatic regions may be important for euchromatic loci position effect variegation. 35th Annual Drosophila Research Conference, Chicago. 325B.

KONEV, A. Y., 1995 Cytogenetical study of 44F-45D region of chromosome 2 in D. melanogaster. Ph.D. Thesis, Sanct-Petersburgh. (Russian).

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

LOHE, A. R. and A. J. HILLIKER, 1995  Return of the H-word (heterochromatin). Curr. Opin. Genet. Dev. 5:746-755[Medline].

LOHE, A. R., A. J. HILLIKER, and P. A. ROBERTS, 1993  Mapping simple repeated DNA sequences in heterochromatin of D. melanogaster.. Genetics 134:1149-1174[Abstract].

MAKUNIN, I. V., G. V. POKHOLKOVA, S. O. ZAKHARKIN, N. G. KHOLODILOV, and I. F. ZHIMULEV, 1995  Isolation and characterization of the repeated DNA sequences from the centric heterochromatin of the second chromosome of D. melanogaster.. Dokl. Akad. Nauk (Russ.) 344:266-269.

MCKEE, B. D., 1996  The licence to pair: identification of meiotic pairing sites in Drosophila. Chromosoma 105:135-141[Medline].

PANSHIN, I. B., 1938  The cytogenetic nature of the position effect of the genes white (mottled) and cubitus interruptus.. Biologicheskij Z. 7:837-868.

PIMPINELLI, S., M. BERLOCO, L. FANTI, P. DMITRI, and S. BONACCORSI et al., 1995  Transposable elements are stable structural components of Drosophila melanogaster heterochromatin. Proc. Natl. Acad. Sci. USA 92:3804-3808[Abstract/Free Full Text].

POKHOLKOVA, G. V., I. V. MAKUNIN, E. S. BELYAEVA, and I. F. ZHIMULEV, 1993  Observation on the induction of position effect variegation of euchromatic genes in D. melanogaster.. Genetics 134:231-242[Abstract].

REUTER, G. and P. SPIERER, 1992  Position effect variegation of euchromatic genes in D. melanogaster.. Bioessays 14:605-612[Medline].

SPOFFORD, J. B., 1976 Position effect variegation in Drosophila, pp. 955–1019 in Genetics and Biology of Drosophila, edited by M. ASHBURNER and E. VOVITSKI. Academic Press, London.

SUN, X., J. WAHLSTROM, and G. KARPEN, 1997  Molecular structure of a functional Drosophila centromere. Cell 91:1007-1019[Medline].

TALBERT, P. B., C. D. S. LECIEL, and S. HENIKOFF, 1994  Modification of the Drosophila heterochromatic mutation brown^Dominant by linkage alteration. Genetics 136:559-571[Abstract].

TARTOF, K. D., C. HOBBS, and M. JONES, 1984  A structural basis for variegation position effects. Cell 37:869-878[Medline].

TOLCHKOV, E. V., M. D. BALAKIREVA, and V. E. ALATORTSEV, 1984  Inactivation of the X chromosome region with a known fine genetic structure as a result of the variegated position effect in D. melanogaster.. Genetika (Russ.) 20:1846-1856.

TOLCHKOV, E. V., I. A. KRAMEROVA, S. A. LAVROV, V. I. RASHEVA, and S. BONACCORSI et al., 1997  Variegated position effect in D. melanogaster X chromosome inversions with a breakpoint in satellite block. Chromosoma 106:520-525[Medline].

WAKIMOTO, B. T. and M. G. HEARN, 1990  The effect of chromosome rearrangements on the expression of heterochromatic genes in chromosome 2L of D. melanogaster.. Genetics 125:141-154[Abstract].

WEILER, K. S. and B. T. WAKIMOTO, 1995  Heterochromatin and gene expression in Drosophila. Annu. Rev. Genet. 29:577-605[Medline].

WEILER, K. S. and B. T. WAKIMOTO, 1998  Chromosome rearrangements induce both variegated and reduced, uniform expression of heterochromatic genes in a development-specific manner. Genetics 149:1451-1464[Abstract/Free Full Text].

ZHIMULEV, I. F., 1998  Polytene chromosomes, heterochromatin and position effect variegation. Adv. Genet. 37:1-566[Medline].

This article has been cited by other articles:

				Y. A. Abramov, G. L. Kogan, E. V. Tolchkov, V. I. Rasheva, S. A. Lavrov, S. Bonaccorsi, I. A. Kramerova, and V. A. Gvozdev Eu-heterochromatic Rearrangements Induce Replication of Heterochromatic Sequences Normally Underreplicated in Polytene Chromosomes of Drosophila melanogaster Genetics, December 1, 2005; 171(4): 1673 - 1681. [Abstract] [Full Text] [PDF]

K. A. Maggert and G. H. Karpen The Activation of a Neocentromere in Drosophila Requires Proximity to an Endogenous Centromere Genetics, August 1, 2001;