Chromosome architecture and distant interactions between regulatory elements are one of the key fields of research in modern biology. Today, it became obvious that, to establish correct expression of genes, their regulatory elements specifically interact with each other [1]. Recently, genome-wide 3D studies have shown that chromosomes of higher eukaryotes are organized into topologically associated domains (TADs) [2]. The architectural protein CTCF, which so far is the only comprehensively studied mammalian insulator protein, plays the key role in the organization of TAD boundaries in vertebrates [3].

One of the most convenient models for studying the organization of spatial specific interactions between regulatory elements in vivo is the Drosophila melanogaster Bithorax complex (BX-C). It consists of three homeotic genes Ultrabithorax (Ubx), abdominal-A (abd-A), and Abdominal-B (Abd-B), which are responsible for the formation of the third thoracic (T3) and all abdominal (A2–A7) segments [4–6]. These genes are regulated by tissue-specific regulatory domains (abx/bx, bxd/pbx, and iab-2–iab-8), alternating in the order of the location of the segments that they control (Fig. 1a). Each domain is responsible for the expression of one of the three genes and functions autonomously due to the surrounding insulator boundaries [7, 8]. The regulatory region of the Abd-B gene, represented by the iab-5–iab-8 domains, which are flanked by the Mcp, Fab-6, Fab-7, and Fab-8 boundaries, has been studied in most detail [4–6] (Fig. 1a). The regulatory elements PRE (Polycomb response element), which recruit the Polycomb complexes, are located near the boundaries [9, 10]. Using Fab-7 and Fab-8 as an example, it was shown that the boundaries have two functions: they maintain the autonomy of neighboring iab-domains and, at the same time, ensure specific interactions between iab- domains and the Abd-B promoter [11]. The Mcp, Fab-6, and Fab-8  boundaries contain the binding sites for the homologue of mammalian CTCF protein (dCTCF), and the Fab-7 and Mcp boundaries contain the binding sites for the architectural protein Pita [12, 13].

Fig. 1.
figure 1

(a) Scheme of the regulatory region of abd-A and Abd-B genes. Blue and green arrows show the abd-A gene transcripts, and blue color show the regulatory region of the Abd-B gene, respectively. Curly braces denote iab-domains containing segment-specific enhancers (the regulatory region of the abd-A gene is shown in blue, and the regulatory region of the Abd-B gene is shown in green). Vertical black lines mark the boundaries (Fub, Fab-3, Fab-4, Mcp, Fab-6, Fab-7, and Fab-8) of the iab-2–iab-8,9 regulatory domains, which are responsible for gene regulation and differentiation of segments T3–A8. The dCTCF-binding sites at the boundaries are shown with red circles. Below is a map of the Fab-7 boundary region that was removed and replaced with attP, lox, and frt sites in the Fab-7attP50 platform. Hypersensitivity sites, HS *, HS1, and HS2, are shown with gray rectangles, and the HS3 site, which is the PREiab-7 silencer, is shown in blue. (b) Photographs of the male abdomen cuticle in the light and dark field: wt (wild type), Fab-7attP50, F4 ry, F3 ry, F4, F3, F4 + PREiab-7, and F3 + PREiab-7. In adult wt males, the A7 segment is absent, the A6 sternite is banana-shaped and devoid of bristles, and the A5 sternite is quadrangular and covered with bristles. The A6 tergite has trichomes (dark field) only along the anterior and ventral margins, whereas the entire A5 tergite is almost completely covered with trichomes. In Fab7attP50 males, A6 is absent. The A6 segment of lines F3 ry and F4 ry is devoid of the sternite (partial transformation into A7) but has a deformed tergite. However, the distribution of trichomes indicates partial transformation into A5. In F3 and F4 lines, the complete disappearance of A6 is observed again. In the case of F4+PREiab-7 and F3+PREiab-7, the sternite is absent, and the tergite is markedly reduced in size but has extraneous trichomes, which indicates the transformation into A5.

Full size image

The aim of this study was to attempt to identify the boundaries in the regulatory region of the abd-A gene. Previously, regions corresponding to the potential boundaries Fab-3 and Fab-4 surrounding the iab-3 domain, containing two (Fab-3) and one (Fab-4) binding sites for the dCTCF protein, were found in the regulatory region of the abd-A gene [12]. It was shown previously that DNA fragments Fab-3 (626 bp, F3, (3R:16835097..16834472)) and Fab-4 (784 bp, F4 (3R:16857582..16856799)) (coordinates are indicated in accordance with Genome Release r6. 41, FlyBase), despite the presence of dCTCF sites, did not exhibit the properties of insulators at the imago stage in the studies of transgenic Drosophila lines using model regulatory systems of the yellow and white genes [14]. Four dCTCF sites also did not block the enhancers of the yellow and white genes; however, when located in the regulatory region of the Abd-B gene instead of the Fab-7 boundary, they effectively isolate the iab-6 domain, which ensures the formation of the A6 segment, from the iab-7 domain, which is responsible for the lysis of the A7 segment in adult male flies [15, 16]. Therefore, it can be expected that fragments F3 and F4 can also form a boundary in the BX-C context. To study this issue, we used the previously created Fab-7attP50 platform, in which the Fab-7 boundary was replaced with attP and frt sites (Fig. 1a) [17, 18]. The Fab-7 boundary consists of four hypersensitivity sites for DNAse I: HS * + HS1 + HS2, which form an insulator, and HS3, which is considered a PREiab-7 silencer [10].

All four sites in the Fab-7attP50 line were removed, which led to premature ectopic activation of the iab-7 domain by the iab-6 initiator in all PS11 cells, as a result of which the A6 segment in males was lysed (Fig. 1b), and in females it was transformed into A7. On the basis of the pBluKS plasmid, we created constructs carrying the attB site for integration of F3 or F4 fragments into the desired place site in the genome at the attP site, the frt site for subsequent excision of the reporter and plasmid sequences, and the rosy (ry) reporter gene for selection of positive events. The resulting constructs were injected into the embryos of the ry– Fab-7attP50 line, which contains a construct on the X chromosome that expresses the φc31 integrase at the preblastoderm stage [18]. The survived flies were crossed with the yw;TM2/MKRS line carrying mutations in the ry gene. Positive events of integration into the platform were selected by the restoration of the color of the eyes by the transgenic ry gene to the wild type.

As a result, the lines F3 ry and F4 ry were created. In both cases, slight restoration of boundary functions was observed, which was manifested as an incomplete formation of the A6 segment (the tergite was deformed, and the sternite either did not develop or consisted of small islets of cells) (Fig. 1b). The distribution pattern of trichomes on the tergite and bristles on the sternite remnants indicated that A6 cells in these lines have the specification characteristic of the A5 segment. This fact indicates that iab-6 in these tissues is isolated not only from iab-7 but also from the Abd-B gene promoter. However, after deletion of ry as a result of Flp-induced recombination between the frt sites, a complete lysis of the A6 segment in the F3 and F4 lines in males (Fig. 1b) and the transformation of A6 to A7 in females was observed, which suggests a complete loss of insulation. Since the ry gene itself does not form the boundary between iab-6 and iab-7, it can be concluded that, in the F3 ry and F4 ry lines, the F3 and F4 fragments cooperate with the ry gene in forming the boundary in individual cells of the A6 segment. It is most likely that the ry gene promoter is involved in this process. Thus, the studied fragments F3 and F4 have a very weak insulator activity. It is important to note that the F3 and F4 fragments inserted instead of Fab-7 caused male and female sterility. In this regard, it can be noted that the reproductive apparatus of homozygous males in both lines is rotated by 30°–180°. In addition, both in F3 and F4 lines, a significant decrease in the survival rate of homozygotes was observed. When heterozygous females and males on the MKRS or TM2 balancer were crossed, instead of expected ~33% in progeny no more than 5% of homozygotes survived, and they showed a delay in development (hatched much later than the wild type); when the temperature dropped to 18° C, both F3 and F4 homozygotes were not detected at all.

Earlier, it was shown for the Fab-6 and Fab-7 boundaries that PREs located in the iab-6 and iab-7 domains were involved in the formation of effective boundaries [19, 20]. To investigate the possible involvement of PRE in the activity of the boundaries of the abd-A regulatory region, we created constructs in which PREiab- 7 (227 bp) was added to the F3 and F4 fragments (Fig. 1a) and obtained transgenic lines in which Fab-7 was replaced with F3 + PREiab-7ry and F4 + PREiab-7ry. However, after the excision of ry at the frt sites, no complete restoration of the boundary between iab-6 and iab-7 in the obtained lines was observed. The phenotype of flies of lines F3 + PREiab-7 and F4 + PREiab-7 after the excision of ry resembled the phenotype of flies of lines F3 ry+ and F4 ry+ (Fig. 1b). Thus, PREiab-7 only partially enhances the weak insulator activity of the F3 and F4 fragments in individual cells of the A6 segment.

The obtained results show that the CTCF sites present in the 626-bp fragment of F3 and 784-bp fragment of F4 do not form strong insulators, in contrast to the 337-bp fragment of the Fab-8 boundary, containing two binding sites for the dCTCF protein, which completely blocked the interaction between iab-6 and iab-7 [15]. These results are consistent with the conclusions that unknown architectural proteins together with dCTCF are involved in the formation of strong BX-C boundaries [13]. It can be assumed that, in the iab-3–iab-4 region, strong insulators within the boundaries are not required, which is associated with the peculiarities of regulation in this region. At the same time, the replacement of the Fab-7 boundary with Fab-3 or Fab-4 boundary fragments adversely affects the viability and fertility of flies. This may be due to the ability of these boundaries to form incorrect contacts with their endogenous copies, which disrupts the correct functioning of abd-A and Abd-B genes. This assumption requires further studies.