Pharyngeal Early head trunk endoderm mesoderm
FIGURE 4.8 Transcriptional control of genes required for heart specification in Drosophila. (A-E) tinman (tin). This gene encodes a homeodomain regulator required for specification of dorsal mesoderm derivatives including cardiac mesoderm (CM), visceral mesoderm (VM), and dorsal muscles (Azpiazu and Frasch, 1993; Bod-mer, 1993; Yin et al., 1997). (A) Diagram of morphological gene expression domains in ectoderm and mesoderm of postgastrular embryos. At stage 9 the Dpp ligand is expressed in dorsal ectoderm, to which it is confined by its short gastrulation (Sog) antagonist. Transcription of the sog gene is controlled by the initial gradient of the Dorsal transcription factor (cf. Fig. 2.6). Twist (Twi) is an early mesodermal regulator that also responds to Dorsal, and is responsible for the initial activation of the tin gene. Expression of the tin gene then becomes Dpp-dependent. This maintains tin expression in the dorsal mesoderm. Htl (Heartless) is an FGF receptor required for the migratory behavior of the invaginated mesoderm, in the absence of which no dorsal mesoderm appears (Gisselbrecht et al., 1996). At stage 10 combinatorial patterning devices divide the tin mesodermal expression domain into metameric stripes within each parasegment (Park et al., 1996; Su et al., 1999). The wingless (wg) gene is activated downstream of sloppy paired (sip), and hedgehog (hh) is activated downstream of evenskipped (eve). The sip domains later give rise to cardiac cell types, and the eve domains, in which tin initially activates the gene encoding the Bagpipe (Bap) transcription factor, become visceral mesoderm; DSM, dorsal somatic mesoderm. Black arrows and bars indicate signaling interactions; blue arrows indicate transcriptional interactions. (B) Metameric distribution of cardiac and visceral mesoderm at stage 10, followed by inward migration of visceral mesoderm to produce apposed mesodermal layers. [(A, B) From Bodmer and Frasch (1999) In "Heart Development" (R. P. Harvey and N. Rosenthal, Eds.), pp. 65-90, Academic Press, San Diego.] (C) Location of transcription factor encoded by tin gene in normal embryos visualized by immunocytology: (CI) Gastrulation, Tin factor in all cells of the trunk mesoderm; (C2) stage 11, tin product in dorsal mesoderm (d.ms.), not in ventral mesoderm (v. ms.); expression also in head cap. (C3) Stage 14, Tin factor in cardioblasts (cb's) and in pericardial cells (pc's); (C4) stage 8, high magnification view of head region; tin expression anterior of arrowheads but not in region indicated by bracket (hemocyte domain). (D) Expression of lacz constructs driven by tin c/s-regulatory modules B (Dl), D (D2), C (D3), and A (D4). Embryos in DI-3 are about the same stages, and are in same respective orientations as in (CI-3); D4 is stage 12. [(C, D) From Yin et al. (1997) Development 124, 4971-4982 and The Company of Biologists Ltd.] (E) Organization of tin c/'s-regulatory sequences responsible for patterns of expression shown in (C) and (D). Question mark indicates that identity of factor is not known; dotted lines mean that the factor listed may act indirectly in that no direct interaction in tin c/s-regulatory DNA has been demonstrated. Proteins indicated in parentheses are present in ectoderm and act upstream of mesodermal signal response system. The factors constituting the inputs are present in different regions of the embryo: AE, anterior endoderm; IM, invaginated mesoderm; HH, head hemocyte domain; ES, eve stripes; SS, sip stripes; DM, dorsal mesoderm. Exons are black; the four tin cis-regulatory modules in the figure are located in the first intron (A, B) and 3' of the coding region (C, D). [(E) Adapted from Yin et al. (1997)
The c¿s-regulatory systems that determine the locations in which individual box genes are expressed in the rhombomeres (see Fig. 4.5) are again remarkably modular in organization. Though the context is about as different as could be imagined, this is the same prominent feature that we confronted in the patterning systems of the Drosophila wing and heart. The disposition of modular enhancers governing the domains of expression of hoxa3 is shown in Fig. 4.9A (Manzanares et al., 1999b), and the modular neural (N) and other enhancers (M, mesodermal) controlling expression of the paralogous hoxa4, b4, and d4 genes along the dorsal axis of the embryo are shown in Fig. 4.9B (Morrison etal., 1997). Typically for box genes, boxa3 is expressed in many regions of the postgastrular embryo. Its complex pattern of utilization is the sum of the subpatterns mediated by the different enhancer modules, some of which may indeed be divisible into even finer subelements. A discrete enhancer among these mediates expression of boxa3 exclusively in r5/r6 of the hindbrain (Fig. 4.9A; purple element). Similarly, discrete enhancers control expression of hoxa4, b4, and d4 in r6/r7 (Fig. 4.9B, N modules); others in the somites (M modules). Hoxb3 expression is similarly the sum of multiple modualr control functions (kwan et al., 2001). The experimental demonstrations in Fig. 4.9C-F illustrate the precision with which these and other cis-regulatory modules of the anterior box genes control rhombomeric expression pattern, and indicate some of the inputs which activate each (see legend for details and references). Note, for example, that the solid A/P pattern illustrated in Fig. 4.9E1 for boxb2 is actually a composite of several discrete genomic patterning devices: expression in r3-r5 depends on two such, one that operates in r3 and r5, and one that operates in r4. The explicit nature of these we see below.
Within less than a day (E8.0-E8.5) the rhombomeres achieve their distinct transcriptional identities. Figure 4.9G shows the rhombomeric expression domains just before and immediately after establishment of the respective transcriptional patterns for those genes considered in the following, viz, the box group 1, 2, and 3 genes; krox20 (k20); and kreisler (kr). Some relevant pieces of the control network are schematized in Fig. 4.9H, which is based on
Development 124, 4971-4982 and The Company of Biologists Ltd, including additional information from Reichmann et al. (1997); Bodmer and Frasch (1999).] (F) Combined spatial inputs to the cardial cell module of the mef2 differentiation regulator (Gajewski et al., 1997; Cripps et al., 1998; Nguyen and Xu, 1998). Cardial cells are a subset of mesodermal cells expressing tin. The Tin factor is essential for mefl expression (Gajewski et a!., 1997, 1998; Cripps et al, 1999) and so is Pannier (Pnr), a Gata class factor of heart cells, the gene for which is also activated bt Tin. Ectopic expression of the pnr gene causes cardiac gene expression in all domains in which tin is also expressed (Gajewski et al., 1998, 1999, 2001). The bHLH factor is unknown. Factors binding at the Pnr target site repress the mefl enhancer in pericardial cells (Gajewski et al., 1998; another module of the tin gene generates expression in additional cardial cells independently of the Tin factor).
FIGURE 4.9 c/s-Regulatory control of rhombomere specification during hindbrain development in mouse. (A) Modular regulatory elements of mouse hoxa3-hoxa4 intergenic control region: r, rhombomere; s, somite; hb, hindbrain; sc, spinal cord. [(A) From Manzanares et al. (1999b) Development 126, 759-769 and The Company of Biologists Ltd.] (B) Modular regulatory elements surrounding mouse hoxa4, hoxb4, and human hoxd4 (HOXD4) genes. CRI is a conserved intron sequence element the function of which is not known: M, mesoderm expression module; N, neural expression module; RA, retinoic acid response module. Each orthologue has a distinct c/s-regulatory organization, though they all share the same set of elements. The gray oval in (A) denotes the S7/8 enhancer 3' of the hoxa4 gene in (B). (C) The hoxd4 r6/7 module: (CI) expression of endogenous hoxd4 gene, by in situ hybridization, 9.5 dpc; (C2) expression of lacz transgene driven by neural hoxd4 r6/7 module, 10.5 dpc; no expression occurs in somites, in contrast to (CI). (D) Retinoic acid (RA) sensitivity of neural expression module of hoxb4 gene (cf. B; for review of RA effects on pattern formation in hindbrain, Gavalas and Krumlauf, 2000): (Dl) expression of lacz
FIGURE 4.9 (Continued)
Was this article helpful?