Development 111, 437-449 (1991) Printed in Great Britain © The Company of Biologists Limited 1991 437 Control of the expression of the bithorax complex genes abdominal-A and Abdominal-B by c/s-regulatory regions in Drosophila embryos ERNESTO SANCHEZ-HERRERO Centro de Biologia Molecular, Facultad de Ciencias, Universidad Autdnoma de Madrid, Cantoblanco, 28049 Madrid, Spain Summary The abdominal-A (abd-A) and Abdominal-B (Abd-B) genes of the bithorax complex (BX-C) specify the identity of most of the Drosophila abdomen. Six different classes of infraabdominal {job) mutations within the BX- C transform a subset of the parasegments affected by the lack of these two genes. It is thought that these mutations define parasegmental m-regulatory regions that control the expression of abd-A and Abd-B. By staining embryos mutant for different iab mutations with anti-abd-A and anti-Abd-B antibodies I show here that the expression of Abd-B (and probably also abd-A) exhibit a parasegmen- tal regulation. I have also studied the significance of the chromosomal order of parasegmental iab regulatory sequences, and the possible presence of chromosomal 'boundaries' between them, by looking at the expression of abd-A and Abd-B in embryos carrying the Uab 1 and Mcp mutations. These data are discussed in the light of models of parasegmental-specific regulatory regions within the BX-C. Key words: bithorax complex, parasegments, cw-regulation, homeotic. Introduction The identity of segments in Drosophila is determined by a group of genes called homeotic genes, mutations in which transform a segment or a group of segments into another (Gehring and Hiromi, 1986; Mahaffey and Kaufman, 1988). These genes share a number of developmental and molecular characteristics, including the presence of a 180bp coding sequence, the homeo- box, which is present in many other species (McGinnis et al. 1984; Scott and Weiner, 1984; reviewed in Scott et al. 1989). It was this conservation of the homeobox that led to the identification in other organisms of genes homologous to those described as homeotic in the fruit fly (reviewed in Scott et al. 1989). In Drosophila most of these genes are clustered in two complexes: Antennapedia (ANT-C) (Kaufman etal. 1980; Kaufman, 1983) and Bithorax (BX-C) (Lewis, 1978; Sanchez-Herrero et al. 1985). A main feature of both complexes is that the order of the genes (and in the BX-C the regulatory sequences) on the chromosome correlates with their spatial expression and requirement in the anteroposterior axis of the animal: more proximal genes are expressed (and required) more anteriorly than more distal ones (Lewis, 1978; Kaufman, 1983). This relationship also occurs in vertebrates, revealing a surprising evolutionary conser- vation (Gaunt etal. 1988; Graham etal. 1989; Duboule and Dolle, 1989; Akam, 1989). This correlation was first discovered by the pheno- typic study of mutations in the BX-C (Lewis, 1978). This complex is composed of three genes: Ultrabithorax (Ubx), abdominal-A (abd-A) and Abdominal-B (Abd- B) (Sanchez-Herrero etal. 1985; Tiong etal. 1985), each with a homeobox (Regulski et al. 1985). Their concerted action is responsible for the specification of part of the thoracic and all of the abdominal (A) segments, although each gene has its proper realm of action. This is defined by the analysis of patterns of gene expression and by mutational studies, which also show that these genes act on units called parasegments (PS; Martinez-Arias and Lawrence, 1985), rather than segments: mutations in the Ubx gene transform PS5 and PS6 (and to a lesser extent PS7-12), in abd-A PS7-13 and in Abd-B PS10-14 (or PS10-15) (Sdnchez-Herrero etal. 1985; Tiong et al. 1985; Casanova et al. 1987; reviewed in Duncan, 1987). Many of the mutations in the BX-C, however, do not completely inactivate any of these three genes. Their phenotype is part of that produced by Ubx. abd-A or Abd-B mutations, which they fail to complement. Molecular analyses have shown that they map within introns or outside protein-coding transcription units (Bender et al. 1983; Hogness et al. 1985; Karch et al. 1985; Celniker etal. 1989; Zavortink and Sakonju, 1989; Karch et al. 1990), suggesting that their effect could be mainly, or exclusively, regulatory. The existence of mutations in regulatory regions that transform certain parasegments and their particular order on the chromosome prompted a model, based on the original of Lewis (1978), of how the BX-C could be activated (Peifer et al. 1987). This process would be