The Drosophila homolog of the mammalian imprint regulator, CTCF, maintains the maternal genomic imprint in Drosophila melanogaster

Abstract

Background: CTCF is a versatile zinc finger DNA-binding protein that functions as a highly conserved epigenetic transcriptional regulator. CTCF is known to act as a chromosomal insulator, bind promoter regions, and facilitate long-range chromatin interactions. In mammals, CTCF is active in the regulatory regions of some genes that exhibit genomic imprinting, acting as insulator on only one parental allele to facilitate parent-specific expression. In Drosophila, CTCF acts as a chromatin insulator and is thought to be actively involved in the global organization of the genome.

Results: To determine whether CTCF regulates imprinting in Drosophila, we generated CTCF mutant alleles and assayed gene expression from the imprinted Dp(1;f)LJ9 mini-X chromosome in the presence of reduced CTCF expression. We observed disruption of the maternal imprint when CTCF levels were reduced, but no effect was observed on the paternal imprint. The effect was restricted to maintenance of the imprint and was specific for the Dp(1;f)LJ9 mini-X chromosome.

Conclusions: CTCF in Drosophila functions in maintaining parent-specific expression from an imprinted domain as it does in mammals. We propose that Drosophila CTCF maintains an insulator boundary on the maternal X chromosome, shielding genes from the imprint-induced silencing that occurs on the paternally inherited X chromosome. See commentary: http://www.biomedcentral.com/1741-7007/8/104.

Figure 1

Effect of dCTCF alleles on the imprinted Dp(1;f)LJ9 garnet (g) gene expression. (a) Maternally transmitted Dp(1;f)LJ9 mini-X chromosome; the control (y¹z^ag^53d/Dp(1;f)LJ9) displays full garnet expression with no variegation observed. Both dCTCF mutant alleles tested (y¹z^ag^53d/Dp(1;f)LJ9; CTCF^EY15833/+ and y¹z^ag^53d/Dp(1;f)LJ9; CTCF³⁰/+) disrupt maintenance of the maternal imprint, causing variegated garnet gene expression. Significant reduction in both red and brown pigment levels is observed in the presence of CTCF^EY15833 or CTCF³⁰ alleles. (b) Paternally transmitted Dp(1;f)LJ9 mini-X chromosome; the control (y¹z^ag^53d/Dp(1;f)LJ9) exhibits variegated garnet gene expression, whereas the introduction of CTCF^EY15833 and CTCF³⁰ alleles had no significant effect on garnet gene variegation. Pigment assay values are expressed as a percentage of wild-type pigment levels ± standard deviation. Values that are significantly different from the controls are marked with an asterisk signifying P < 0.001.

Figure 2

Eye phenotype of Dp(1;f)LJ9 Drosophila with dCTCF alleles. (a) Phenotypes of maternally inherited Dp(1;f)LJ9 mini-X chromosome ranging from 0% to 100% pigmentation; control n = 300, CTCF^EY15833 n = 300, CTCF³⁰ n = 300. (b) Phenotypes of maternally inherited Dp(1;f)LJ9 mini-X chromosome ranging from >80% to 100% pigmentation; control n = 150, CTCF^EY15833 n = 132 (Kolmogorov-Smirnov test; P < 0.001) and CTCF³⁰ n= 122 (Kolmogorov-Smirnov test; P < 0.001). (c) Phenotypes of paternally inherited Dp(1;f)LJ9 mini-X chromosome ranging from 0% to 100% pigmentation; control n = 300, CTCF^EY15833 n = 300, CTCF³⁰ n = 300. Each eye was scored depending on its phenotypic class, and the prevalence of each phenotypic class is expressed as a percentage versus the total number of eyes scored (n).

Figure 3

Absence of maternal or paternal effects from mutant dCTCF on Dp(1;f)LJ9 garnet expression. External control progeny (no modifier allele) have the same genotype as internal control progeny (y¹z^ag^53d/Dp(1;f)LJ9; TM3, Sb Ser/+) but are generated from a separate cross with parents that have never encountered a mutant CTCF^X allele. (a) Maternally inherited Dp(1;f)LJ9 CTCF^X internal control eye pigment levels; no significant difference in pigment levels was observed between external control progeny (no modifier allele) and internal control progeny from fathers carrying CTCF^X (CTCF^X internal), demonstrating that no paternal effect occurs. (b) Phenotypes of maternally inherited Dp(1;f)LJ9 ranging from 0% to 100% pigmentation; No modifier allele n = 300, CTCF^EY15833 internal n = 300, CTCF³⁰ internal n = 300. No garnet variegation was detected from the internal controls. (c) Paternally inherited Dp(1;f)LJ9 CTCF internal control eye pigment levels; no significant difference in pigment levels was observed between the external control progeny (no modifier allele) and internal control progeny from mothers carrying CTCF^X (CTCF^X internal), demonstrating that no maternal effect occurs. (d) Phenotypes of paternally inherited Dp(1;f)LJ9 ranging from 0% to 100% pigmentation; no modifier allele n = 300, CTCF^EY15833 internal n = 300, CTCF³⁰ internal n = 300. No significant change in garnet variegation was detected from the internal controls. Red eye pigment levels are measured by absorbance at 480 nm, and pigment quantification mean values for each group are based on n = 5 samples (40 heads total); error bars represent standard deviation.

Figure 4

dCTCF does not affect establishment of the Dp(1;f)LJ9 imprint. All genotypes tested are y¹z^ag^53d/Dp(1;f)LJ9; +/+ but differ in parental genotype. External control progeny (Ex. Control) are generated from parents carrying Dp(1;f)LJ9 that have never been exposed to a mutant dCTCF allele. Progeny generated from a parent carrying both the imprinted Dp(1;f)LJ9 mini-X chromosome and a mutant dCTCF allele test for effects on imprint establishment (Mat or Pat-Est. CTCF³⁰), whereas parental siblings carrying Dp(1;f)LJ9 and wild type for CTCF serve as an internal control (Mat or Pat-Est. CTCF⁺). (a) Maternal establishment of the Dp(1;f)LJ9 imprint is not affected by CTCF³⁰. No significant change in red pigment levels was detected between external control progeny (Ex. Control; n = 23), mothers carrying Dp(1;f)LJ9^MAT, and CTCF³⁰ (Mat-Est. CTCF³⁰; n = 10), and mothers carrying Dp(1;f)LJ9^MAT and CTCF⁺ (Mat-Est. CTCF⁺; n = 10). (b) No phenotypic difference is present between progeny generated from Dp(1;f)LJ9 carrying mothers, either mutant (Mat-Est. CTCF³⁰) or wild type (Mat-Est. CTCF⁺), for CTCF. (c) Paternal establishment of the Dp(1;f)LJ9 imprint is not affected by CTCF³⁰. No significant change in red pigment levels was detected between external control (Ex. Control; n = 44), fathers carrying Dp(1;f)LJ9^PAT and CTCF³⁰ (Pat-Est. CTCF³⁰; n = 18), and fathers carrying Dp(1;f)LJ9^PAT and CTCF⁺ (Pat-Est. CTCF⁺; n = 29). (d) No phenotypic difference is present between progeny generated from Dp(1;f)LJ9 carrying fathers, either mutant (Pat-Est. CTCF³⁰) or wild type (Pat-Est. CTCF⁺) for dCTCF.

Figure 5

The effect of CTCF³⁰ on white variegation in In(1)w^m4 and the fourth chromosome variegating strains 6-M193 and 39C-33. (a) Pigment levels were measured for In(1)w^m4/w¹¹¹⁸;CTCF³⁰/+ and In(1)w^m4/Y;CTCF³⁰/+ genotypes compared with the corresponding sibling In(1)w^m4;Tb/+ controls. In(1)w^m4 heterozygous for CTCF³⁰ results in an increase in white expression; however, this increase is only significant in female progeny (white bars). Red pigment quantification mean values for each group are based on n = 10 (50 heads total). Error bars represent standard deviation, and values significantly different from the controls are marked with an asterisk (P < 0.001). (b) Pigment levels were independently measured for both the maternally (6-M193^MAT) and paternally (6-M193^PAT) inherited fourth chromosome variegator 6-M193. 6-M193 heterozygous for CTCF³⁰ results in an increase in white expression when either maternally or paternally inherited; however, this increase is only significant in female progeny (white bars). Red pigment quantification mean values for each group are based on n = 9 (45 heads total) for 6-M193^MAT male and female progeny; n = 10 (50 heads total) for 6-M193^PAT male and female progeny; 6-M193;+/+ control male progeny; and n = 12 (60 heads total) for 6-M193;+/+ control female progeny. Error bars represent standard deviation, and values significantly different from the controls are marked with an asterisk (P < 0.001). (c) Pigment levels were independently measured for both the maternally (39C-33^MAT) and paternally (39C-33^PAT) inherited fourth chromosome variegator 39C-33. 39C-33 heterozygous for CTCF³⁰ results in no significant change in white expression when either maternally or paternally inherited. Red pigment quantification mean values for each group are based on n = 10 (50 heads total) for 39C-33^MAT and 39C-33^PAT male and female progeny, and n = 20 (100 heads total) for 39C-33;+/+ control male and female progeny. Error bars represent standard deviation.

Figure 6

Mating schematic for testing the effect of dCTCF on the maintenance of the Dp(1;f)LJ9 imprint. Two sets of progeny are generated from this cross: progeny that have independently inherited the Dp(1;f)LJ9 mini-X chromosome and a CTCF^X mutant allele (modifier progeny) and progeny that have inherited Dp(1;f)LJ9 and the TM3, Sb Ser balancer, but had a parent carrying a CTCF^X mutant allele (maternal and paternal effect test progeny).

Figure 7

Mating schematic for testing dCTCF for an effect on maternal establishment of the Dp(1;f)LJ9 imprint. Two primary sets of progeny and an external control were generated from this cross: progeny with a maternally-inherited Dp(1;f)LJ9 mini-X chromosome from mothers with the CTCF³⁰ mutation (Mat-Est. CTCF³⁰) and progeny that also have a maternally imprinted Dp(1;f)LJ9 chromosome but from mothers wild type for CTCF (Mat-Est. CTCF⁺). External control crosses were produced by crossing F1 generation X^X/Dp(1;f)LJ9; e/e females to y¹z^ag^53d/Y males (not depicted).

Figure 8

Mating schematic for testing dCTCF for an effect on the paternal establishment of the Dp(1;f)LJ9 imprint. Two primary sets of progeny and an external control were generated from this cross: progeny with a paternally-transmitted Dp(1;f)LJ9 mini-X chromosome from fathers with the CTCF³⁰ mutation (Pat-Est. CTCF³⁰) and progeny that also have a paternally imprinted Dp(1;f)LJ9 mini-X chromosome but from fathers wild type for CTCF (Pat-Est. CTCF⁺). External control crosses were produced by crossing F1 generation X^Y/Dp(1;f)LJ9; e/e males to y¹z^ag^53d/y¹z^ag^53d females (not depicted).