Position Effect Variegation and Viability Are Both Sensitive to Dosage of Constitutive Heterochromatin in Drosophila

The dosage effect of Y-chromosome heterochromatin on suppression of position effect variegation (PEV) has long been well-known in Drosophila. The phenotypic effects of increasing the overall dosage of Y heterochromatin have also been demonstrated; hyperploidy of the Y chromosome produces male sterility and many somatic defects including variegation and abnormal legs and wings. This work addresses whether the suppression of position effect variegation (PEV) is a general feature of the heterochromatin (independent of the chromosome of origin) and whether a hyperdosage of heterochromatin can affect viability. The results show that the suppression of PEV is a general feature of any type of constitutive heterochromatin and that the intensity of suppression depends on its amount instead of some mappable factor on it. We also describe a clear dosage effect of Y heterochromatin on the viability of otherwise wild-type embryos and the modification of that effect by a specific gene mutation. Together, our results indicate that the correct balance between heterochromatin and euchromatin is essential for the normal genome expression and that this balance is genetically controlled.

heterochromatinization process diluting these proteins at the variegating sites (Zuckerkandl 1974). Many genetic dominant suppressors of PEV have been isolated (Grigliatti 1991;Reuter and Spierer 1992). These modifiers show dosage effects on PEV in that one dose suppresses and three doses enhance PEV, suggesting a limited production of their proteins (Locke et al. 1988). Intriguingly, it has been shown that some suppressors of PEV are recessive lethals, and their lethality depends on their interaction with the Y chromosome. For example, a dominant mutation, Su(var)2-1 01 , that suppresses position effect variegation (Reuter et al. 1982b) displays a lethal interaction with the Y chromosome (Reuter et al. 1982a): X/Y males homozygous for Su(var)2-1 01 do not survive, while X/0 males homozygous for the mutation are viable. Because Su(var)2-1 induces a significant hyperacetylation of histone H4 (Dorn et al. 1986), this lethal interaction has been interpreted as a hyperactivation of the chromatin, producing a strong genetic imbalance due to an accumulation of hyperacetylated histones induced by the suppressor and the titration of heterochromatic proteins by the Y chromosome. All these data strongly suggest that heterochromatic proteins are produced in limited amounts, and they raise an important question: if the amount of heterochromatic proteins is critical for the correct functionality of the genome, should hyper-amounts of heterochromatin per se affect viability? The phenotypic effects of Y heterochromatin dosage, even in wild-type flies, have long been well-known. Cooper (1955) showed that hyperploidy of the Y chromosome produces male sterility and many somatic defects including variegation and abnormal legs and wings. More recent data have suggested that quantitative Y chromosome polymorphism could be associated with phenotypic variation in both autosomic and X-linked gene expression, a phenomenon known as Y-linked regulatory variation (YRV) (Francisco and Lemos 2014). We stress that these data are intriguing because in Drosophila, the Y chromosome is essential only for fertility and it is completely dispensable for viability.
We tested the dosage effects of X chromosome and autosomal heterochromatin on PEV and the dosage effects of Y heterochromatin on viability. The results show that PEV suppression is a general feature of any type of constitutive heterochromatin, and that the intensity of suppression is related to its amount instead of some mappable heterochromatic factor. Likewise, the lethal interaction of the Y with Su(var)2-1 01 depends on the overall amount of Y heterochromatin and not on a specific site. Importantly, we also discovered a clear dosage effect of Y heterochromatin on the viability of otherwise wild-type embryos. All these results indicate that the dosage balance between heterochromatin and euchromatin is essential for viability and that it is genetically controlled.

MATERIALS AND METHODS
For a description of chromosome rearrangements and genetic markers, see FlyBase (http://flybase.bio.indiana.edu) and Lindsley and Grell (1968). The majority of free duplications that were generated by same chromosomal inversions share a small euchromatic  (2)e97, respectively. The numbers indicate the different autosome pairs, and letters X and Y, respectively, indicate the X and Y sex chromosomes.

Figure 2
Diagrammatic representation of the DAPI staining pattern of different heterochromatic free duplications of the X chromosome. The first diagram above represents the Dapi banding pattern of X heterochromatin from the wild-type Ore-R strain. The diagrams below show the portions of heterochromatin present in the different free duplications. C indicates the position of the centromere. Region 29 represents the nucleolar organizer. Filled segments indicate bright fluorescence, cross-hatched segments indicate moderate fluorescence, hatched segments indicate dull fluorescence, and open segments indicate no fluorescence. Euchromatin is depicted as a broken line. Note that the free duplications 1173, 1205, and 1187 have the centromere positioned on the opposite side with respect to the fluorescence pattern of the heterochromatin in wild-type X chromosome. This is because the free duplications were obtained from inversion In(1)sc 8 , in which the euchromatic breakpoint is proximal to the yellow locus and the heterochromatic breakpoint is close to the centromere. The free duplications derived from In(1)sc 8 share the same small distal euchromatic segment. The other free duplication was obtained from a wild-type X chromosome and shares a cytologically small euchromatic segment. Numbers inside the brackets indicate the size of the free duplications expressed as a percentage of the wild-type pericentromeric heterochromatin. Note that the duplication A140 has the whole wild-type heterochromatin. distal segment. Also, the other free duplications show a cytologically small distal euchromatic segment.

Culture conditions
Flies were maintained on a standard Drosophila medium containing cornmeal, yeast, sucrose, and agar with Nipagin added as a mold inhibitor instead of propionic acid (because the latter can suppress position effect variegation). All cultures were grown at 24°.

Eye pigment measurement
Heads were collected 3 d after eclosion of the flies by freezing the adults in Eppendorf tubes and vortexing for a few seconds. The red pigment was extracted according to Ephrussi and Herold (1944). Levels were measured using a spectrophotometric assay at 480 nm.

Mitotic chromosome preparation
Brains were dissected from third instar larvae and mitotic chromosomes were prepared according to Pimpinelli et al. (2010).

RESULTS AND DISCUSSION
To assay the effects of autosomal and X chromosome heterochromatin on PEV, several different sizes of heterochromatic free duplications derived from an X chromosome or a second chromosome (see Figure  1 for examples) were tested on three chromosome rearrangements causing gene variegation. Two of them are inversions of the X chromosome: In(1)l v231, which carries the variegating lethal l(1)v231, and In(1)w m4 , which shows PEV of the wild-type white gene. The third rearrangement is In(2)bw Vde2 , an inversion of the second chromosome that carries a variegating allele of the brown + gene. For all the inversions, the proximal breakpoints are located within the heterochromatin.
The suppression effect of X heterochromatin on position effect variegation To test possible effects of X heterochromatin on PEV, we used a series of X heterochromatic free duplications whose diagrammatic representation is shown in Figure 2. These free duplications were created by Krivshenko and Cooper from the In(1)sc 8 and from a wild-type X chromosome and described in Lindsley and Grell (1968). The size of these heterochromatic duplications are also reported in Parry and Sandler (1974) and ranges from the shortest Dp(1)1187 to the longest Dp(1)A140, which carries all of the X heterochromatin similar to the wild-type OR-R.
To test the effects on the variegating lethality caused by l(1) v231, we crossed heterozygous females carrying the l(1)v231 chromosome and a normal X chromosome marked with the recessive mutations yellow (y), white (w), and forked (f), with males carrying the attached-XY compound chromosome Y s X.Y L , In(1) EN yB, and one of the different free duplications of the X heterochromatin. This cross produces two types of male progeny, one with l(1)v231y and the other with ywf ; both carry the same X heterochromatic free duplication. The relative viability of the l(1) v231 chromosome was measured as the ratio of the l(1)v231y/Dp (1)y + males to their ywf/Dp(1)y + brothers. Table 1 shows the relative viability of l(1)v231 males carrying X heterochromatic free duplications of different sizes. It is apparent that the viability of the l(1)v231 males depends on the amount of X heterochromatin. Males carrying the smallest free duplication are only 6% as viable as their ywf brothers. The viability increases with the amount of heterochromatin to a maximum of 66% viability for the biggest duplication. It is obvious that X heterochromatin can suppress the lethality of l(1)v231 and that the suppression is n The results are of crosses of heterozygous l(1)v231/ywf females for males carrying the attached-XY chromosome, Y s X.Y L , In(1)EN yB, and the indicated heterochromatic free duplication of the X chromosome. a The suppression effect is expressed as relative male viability (%) = lð1Þv231=Dpð1Þy þ males ywf =Dpð1Þy þ males · 100. b Dp(1) Ã y + = X chromosome heterochromatic free duplications.
n The results are of crosses of ywf/ywf females carrying the different X heterochromatic free duplications, or a Y chromosome, for males In (1) dosage-dependent. However, the difference in suppression effect between Dp(1)1173 and Dp(1)A140 seems to be an exception. In this case, the small size difference (only 7%) corresponds to a remarkable difference in suppression effect (from 46.6% to 66.2%, respectively). Because the euchromatic breakpoint in the Dp (1) A140 fragment is not precisely mapped, it could be that some genes at the euchromatin-heterochromatin boundary contribute to a PEV-suppression effect.
The same X heterochromatic free duplications were tested for their effects on In(1)w m4 and In (2)bw Vde2 . In the first case, ywf/ywf females carrying either one of the X heterochromatic free duplications or a Y chromosome were crossed to In(1)yw m / Bs Y males. In the second case, the same females were crossed to X,y/Y; bw v /Cy males. In both the experiments, optical density levels in an eye pigment assay were used to determine the effects of heterochromatin dosage. The percentage of suppression was calculated as difference between the optical density levels of progeny with and without the heterochromatic free duplications. Table 2 and Table 3 show that the X heterochromatin is also able to suppress the variegation of the white and brown genes, and that the intensity of suppression is directly related to heterochromatin dosage.
The suppression effect of autosomal heterochromatin on position effect variegation To assess the capacity of the autosomal heterochromatin to suppress variegation induced by the same chromosome rearrangements, we used a series of different sizes of heterochromatic free duplications of the second chromosome (Brittnacher and Ganetzky 1989) (Figure 3). Again, we crossed heterozygous females carrying the l(1)v231 chromosome and a ywf X chromosome to males carrying the attached-XY compound chromosome Y s X.Y L , In(1)EN yB, and one of the free duplications of the second chromosome heterochromatin. Each cross produced two types of male progeny, one carrying the l(1)v231y and the other carrying ywf, both with the same autosomal heterochromatic free duplication. The relative viability of the l(1)v231 chromosome was measured as the ratio of the l(1)v231y/Dp(2)y + males to their ywf/Dp(2)y + brothers. In Table 4, where the relative viability of l(1)v231 males carrying autosomal heterochromatic free duplications of different sizes is reported, it appears that the viability of the l(1)v231 males depends on the amount of autosomal heterochromatin. The viability of males carrying the free duplications compared with their ywf brothers ranges from 25.8% for the smallest duplication to 57.9% for the biggest one. Autosomal heterochromatin is also able to suppress the lethality of l(1)v231 in a dosage-dependent manner.
The same autosomal free duplications were tested for their effects on In(1)w m4 and In (2)bw Vde2 chromosome rearrangements. The ywf/ ywf females carrying the different autosomal heterochromatic free duplications, or a Y chromosome, were crossed to In(1)yw m / Bs Y males or to X,y/Y; bw v /Cy males. Once again, we measured the optical density levels in eye pigment assays of progeny carrying or not carrying heterochromatic free duplications. The dosage effect was expressed as the percentage of suppression induced by the different free duplications compared with the Y chromosome. Table 5 and  Table 6 show a clear dosage effect of autosomal heterochromatin in suppressing the PEV of both white and brown genes.
All our data clearly show that heterochromatic free duplications are able to suppress the variegation of all the tested rearrangements and that the intensity of suppression is directly related to the size of n The results are of crosses of ywf/ywf females carrying the different X heterochromatic free duplications, or a y+ Y chromosome, for males y/Y; bw v /Cy.  the duplications, regardless of the chromosomic origin of the heterochromatin. Because a previous study showed a similar behavior of the entirely heterochromatic Y chromosome on PEV (Dimitri and Pisano 1989), this indicates that a dosage effect on PEV is a feature of all constitutive heterochromatin.

Interaction of different Y chromosome fragments with
Su-var(2)1 01 mutation As we mentioned, the dominant PEV suppressor Su-var(2)1 01 (Reuter et al. 1982b) displays a lethal interaction with the Y chromosome. X/Y males homozygous for Su-var(2)1 01 are completely lethal, whereas homozygous X/0 males are almost completely viable (Reuter et al. 1982a). To test whether the lethal interaction of the Y chromosome with Su(var)2-1 01 depends on the amount of Y heterochromatin, we analyzed the viability of Su (var)2-1 01 homozygous males carrying Y chromosome fragments of different sizes, as illustrated in Figure 4 (Pimpinelli et al. 1985). Table 7 shows the results from crosses of w m4 /w m4 ; Su(var)2-1 01 / Cy females with X-Y; Su(var)2-1 01 , Sco/+ males, which also carry Y chromosome fragments of different sizes, particularly Df(Y)S6, Df(Y)S12, and Df(Y)S10 fragments. Fragments Df(Y)S12 and Df (Y)S10 appear similar in size. However, the length of the nucleolar organizer (region 20 in the diagram of Figure 4) is not representative of its real length because the maps were elaborated from prometaphase chromosomes, where this region is less compact than the rest of heterochromatin. In metaphases, Df(Y)S12 is significantly longer than Df(Y)S10. For each cross, the proportion of the progeny of Su(var)2-1 01 , Sco/Su(var)2-1 01 homozygous males compared to their Su(var) 2-1 01 , Sco /Cy heterozygous brothers clearly shows that the lethal interaction is correlated with the size of the Y chromosome fragments, thus indicating a quantitative effect of heterochromatin on the lethality induced by the Su(var)2-1 01 mutation.
We then analyzed the phenocritical period of the larval lethality. Intriguingly, we found that the majority of the larvae reach the adult stage; the lethality is mainly concentrated at the embryo stage. This suggests a threshold effect of heterochromatin dosage at a restricted and sensitive period during embryo development. The embryos that surmount this stage of sensitivity are able to reach the adult stage.

The lethal dosage effect of Y chromosome hyperploidy
The phenotypic effects of Y heterochromatin dosage, even in wildtype flies, have long been well-known. Cooper (1955) showed that hyperploidy of the Y chromosome produces male sterility and many somatic defects, including variegation and abnormal legs and wings. The somatic dosage effect of the Y chromosome is intriguing because this chromosome is essential only for fertility, whereas it is completely dispensable for viability. These data above show a lethal interaction of Y heterochromatin with Su(var)2-1 01 depending on dosage and suggest that a hyperdosage of heterochromatin could also affect viability in wild-type flies.
We used free duplications of the Y chromosome ( Figure 4) to determine if altering the dosage of specific Y chromosome regions produces phenotypic abnormalities and affects viability. We crossed ywf/w + Y males, which also carry Y chromosome fragments of different sizes to ywf/ywf/B s Y females, and we analyzed the percentage of male progeny with three Y chromosomes, or with two Y chromosomes plus another Y chromosome fragment. The data reported in Table 8 strongly indicate a quantitative lethal effect of the Y heterochromatin. We found lethality among male progeny carrying two Y chromosomes plus an additional Y fragment, and its strength was related to the fragment size. The only exception seems to be the small reversed effect of S5 and S6 with respect to their length. At present, we do not have any plausible explanation. However, we think that this result is not so relevant to affect the general conclusions that lethality n The results are of crosses of heterozygous l(1)v231/ywf females for males carrying the attached-XY chromosome, Y s X.Y L , In(1)EN yB and the indicated heterochromatic free duplication of the second chromosome. a The suppression effect is expressed as relative male viability (%) = lð1Þv231=0; Dpð2Þy þ males ywf =0; Dpð2Þy þ males · 100 b Dp(2) Ã y + = second chromosome heterochromatic free duplications.
n The results are of crosses of ywf/ywf females carrying the different second chromosome heterochromatic free duplications, or y+ Y chromosome, for males In (1) is related to the increase of Y chromosome dosage. We also observed several phenotypic abnormalities in the surviving progeny with a high dosage of Y heterochromatin, such as those described by Cooper (1955).

CONCLUSIONS
Our results clearly show that the pericentromeric constitutive heterochromatin of different chromosomes suppresses PEV and that the intensity of suppression is directly related to dosage rather than to any mappable heterochromatic element. We stress that all the types of PEV that we analyzed were due to the relocation of the variegating genes close to pericentromeric heterochromatin. We cannot exclude different sensitivity of telomeric PEV to the dosage of heterochromatin. A different response of telomeric PEV to Su(var) mutations has already been shown (Cryderman et al. 1999;Wang et al. 2014). This suggests that telomeric PEV could be a peculiar silencing mechanism. More significantly, the present data show that viability in D. melanogaster is also sensitive to the amount of heterochromatin and that this sensitivity can be modified by specific mutations such as Su(var)2-1 01 . These data indicate that the correct genome expression depends on the amount of heterochromatin, thus suggesting a func-tional relationship between heterochromatin and euchromatin. We think that the heterochromatic and the euchromatic domains probably share many structural features involving several chromosomal proteins. Some evidence for a dynamic functional balance between heterochromatin and euchromatin has been already provided by Ebert et al. (2004). In this view, the imbalance of genome function produced by a variation in heterochromatin dosage could depend on an alteration in the distribution of chromatin factors between the two domains (Zuckerkandl 1974). This mechanism establishes a functional connection between heterochromatin and euchromatin with heterochromatin regulating euchromatic gene expression by controlling the chromatin structure (Pimpinelli 2000). It is not unreasonable to imagine how a quantitative imbalance of shared proteins between heterochromatin and euchromatin could produce phenotypic effects. A hyper-dosage of heterochromatic DNA may, in fact, accumulate several key regulatory proteins in heterochromatin, thus decreasing their availability for the regulation of normal euchromatic gene expression at various loci ( Figure 5). Although other scenarios cannot be ruled out, the demonstration by Fanti et al. (2008) that heterochromatin and euchromatin share many chromatin proteins involved in maintaining the expression state of several genes during development seems to support this view.
n The results are of crosses of ywf/ywf females carrying the different second chromosome heterochromatic free duplications, or y+ Y chromosome, for males y/Y; bw v /Cy. Optical density (O.D. 480 nm) levels were measured in y/ywf; bw v /+/Dp (2)

ACKNOWLEDGMENTS
We thank the Bloomington Stock Center for the mutant Drosophila strains. We also thank B. Ganetzky for the stocks carrying autosomal heterochromatic free duplications. We thank F. Spirito for several helpful discussions. This work was partly supported by the Epigenomics Flagship Project EpiGen, the Italian Ministry of Education and Research, National Research Council.