BIOLOGICAL OVERVIEW

The gap genes hunchback, giant, knirps and Krüppel function in concert to subdivide the anterior-posterior axis of the early embryo. They form an interactive system whose functional domains are set, at least in part, by their effects on one another. Krüppel expression forms a single broad band, like a belt, around the middle of the embryo. Two broad stripes of knirps expression lie adjacent to the anterior and posterior edges of the Krüppel band. Two stripes of giant expression abut the two knirps domains, and consequently lie even closer to the embryo's terminal ends.

Even when functioning as transcription factors, these genes do not necessarily operate independently. All transcription factors form multiprotein complexes when bound to promoters, and many interact when in the free, unbound state. Krüppel can physically associate with Knirps and Hunchback to form heterodimers (Sauer, 1995), and often their binding sites on promoters interdigitate (Langland, 1995 and Gutjahr, 1994). Further refuting hopes for functional simplicity, these proteins may act either as inhibitors or activators. Knirps, for example, activates hairy stripe 6 (Langeland, 1994) and represses stripe 7 (Pankratz, 1990). Despite the interactive associations among these three transcription factors, they belong to three distinctly different molecular families: Knirps is a steroid receptor; Krüppel, a zinc finger, and Giant, a basic leucine zipper. There are no simple answers here.

Knirps is considered a short range repressor. Rather than functioning competitively to block the binding of inducers, such repressors are thought to engender local changes in chromatin structure resulting in enhancer silencing. Human CtBP attenuates transcriptional activation and tumorigenesis mediated by the adenovirus E1A protein. CtBp is known to interact with a conserved sequence in the adenovirus E1A protein, Pro-X-Asp-Leu-Ser-X-Lys (P-DLS-K). Mutations in this sequence eliminate E1a-CtBP interactions so that CtBP no longer inhibits E1A mediated transcriptional activation and tumorigenesis in mammalian cell cultures. The P-DLS-K sequence is present in the repression domains of two unrelated short-range repressors, Knirps and Snail; the latter protein also contains the related sequence P-DLS-R. The P-DLS-K sequence is essential for the interaction of these proteins with Drosophila CtBP (dCtBP). A P-element-induced mutation in dCtBP exhibits gene-dosage interactions with a null mutation in knirps, which is consistent with the occurrence of Knirps-dCtBP interactions in vivo. Mutation of the D-DLS-K motif in Knirps, results in a Knirps protein that cannot mediate repression. These observations suggest that CtBP and dCtBP are engaged in an evolutionarily conserved mechanism of transcriptional repression, which is used in both Drosophila and mammals. Mammalian and Drosophila CtBP may mediate repression through the enzymic modification of chromatin because both proteins are related to D-isomer 2-hydroxy acid dehydrogenases. Despite this rather unexpected homology, immunolocalization assays indicate that the Drosophila CtBP protein accumulates in nuclei. Perhaps CtBP and dCtBP cause local changes in chromatin structure by introducing subtle changes in core histones. Alternatively, it is possible that CtBPs are components of an enzymatic cascade that modulates the activities of histone deacetylases or other co-repressor proteins (Nibu, 1998).

The knirps and knirps-related genes organize the development of the second wing vein in Drosophila. The position of the L2 wing vein is determined by a chain of known developmental events, beginning with the primary subdivision of the wing imaginal disc into anterior versus posterior lineage compartments. The subdivision of body segments such as the wing primordium into anterior and posterior compartments, in turn, can be traced back to early A/P patterning in the blastoderm stage embryo. To summarize these events briefly:

1. The posterior compartment fate is defined by expression of engrailed (en), which activates expression of the short-range Hedgehog (Hh) signal in posterior compartment cells and prevents posterior compartment cells from responding to Hh.
2. Secreted Hh travels a short distance (6-8 cells) into the anterior compartment where it initiates a sequence of signaling events, culminating in the activation of several Hh target genes including decapentaplegic.
3. Dpp synthesized in this narrow strip of cells travels significant distances in both the anterior and posterior directions to activate expression of Dpp target genes, such as the neighboring spalt-major (salm) and spalt-related (salr) genes, in a threshold-dependent fashion.
4. Juxtaposition of salm expressing and salm non-expressing cells induces expression of the rhomboid (rho) gene in a stripe 1-2 cells wide, corresponding to the L2 vein primordium.
5. rho then promotes differentiation of all longitudinal veins during late larval and early pupal development by potentiating signaling through the Egf-R/RAS pathway (Lunde, 1998 and references).

Evidence is presented that kni and knrl link A/P positional information in larval wing imaginal discs to morphogenesis of the second longitudinal wing vein (L2). kni and knrl are expressed in similar narrow stripes corresponding to the position of the L2 primordium. The kni and knrl L2 stripes abut the anterior border of the broad central expression domain of the Dpp target gene spalt major. Evidence is provided that radius incompletus (ri), a well-known viable mutant lacking the L2 vein, is a regulatory mutant of the kni/knrl locus. In ri mutant wing discs, kni and knrl fail to be expressed in the L2 primordium. In addition, the positions of molecular breakpoints in the kni/knrl locus indicate that the ri function is provided by cis-acting sequences upstream of the kni transcription unit. Consistent with kni and knrl playing a role in L2 vein formation, kni and knrl are expressed in similar narrow stripes corresponding to the position of the L2 primordium. kni-expressing cells abut the anterior border of strong sal-lacZ expression and express little or no detectable lacZ. For convenience, these kni expressing cells are referred to as salm non-expressing cells. Consistent with the genetic evidence that ri is a regulatory mutant of the kni/knrl locus, the L2 stripes of kni and knrl expression are absent in ri mutant discs. However, outside the wing pouch of ri discs, kni and knrl are expressed normally. In support of the genetic evidence suggesting that ri is a cis-acting regulatory allele of the kni/knrl locus, ri function has been mapped to a region lying immediately upstream of the kni transcription unit (Lunde, 1998).

Knirps targets rhomboid, which is required for the formation of the L2 vein. ri function is required to initiate expression of the vein-promoting gene rho in the L2 primordium, but is not essential for rho expression in other vein stripes. As would be expected if the kni/knrl locus acted upstream of rho, initiation of kni expression in the L2 primordium precedes that of rho. Another early marker for the L2 vein primordium is down-regulation of the key intervein gene blistered (bs). In ri mutants, down-regulation of Bs in L2 is not observed. Consistent with the kni/knrl locus functioning upstream of both rho and Egf-receptor signaling, kni and knrl are expressed normally in rho ^ve; vn ¹ double mutant wing discs. rho ^ve; vn ¹ mutants, which lack rho expression in vein primordia and have reduced levels of the Egf-R ligand encoded by the vn gene, are devoid of veins. Rescue of ri mutants by a ubiquitously expressed kni transgene also suggests that kni controls rho expression (Lunde, 1998).

The salm transcription factor has been shown to function upstream of rho in the L2 primordium; rho expression in L2 has been shown to be induced at the boundary between salm expressing cells and salm non-expressing cells (Sturtevant, 1997). The L2 vein primordium abuts salm-expressing cells but is comprised largely of salm non-expressing cells (Sturtevant, 1997). Like rho, expression of kni in the L2 primordium abuts the anterior edge of the broad salm expression domain in wild-type third instar wing discs, and is displaced along with the anterior border of salm expression in hedgehog ^Moonrat (hh ^Mrt) wing discs. In hh ^Mrt wing discs, the anterior limit of the salm expression domain on the ventral surface is frequently shifted forward, relative to the border on the dorsal surface. Associated with the asymmetry in sal-lacZ expression, the dorsal and ventral components of the kni L2 stripe are driven out of register. The coordinate shift of salm and kni expression is consistent with salm functioning upstream of kni. Strong ectopic expression of either salm or spalt related using the GAL4/UAS system eliminates kni and knrl expression, and leads to the production of small wings lacking the L2 and L5 veins. The loss of kni and knrl expression in discs mis-expressing salm or salr and the subsequent elimination of L2 presumably results from obscuring the sharp boundary of endogenous salm and salr expression. Clonal analysis also indicates that salm acts upstream of kni/knrl. salm - clones generated in the anterior compartment between L2 and L3 induce ectopic forks of the L2 vein, which lie along the inside edge of the salm - clones. In contrast, salm - clones produced in corresponding positions of ri mutant wings never induce L2 forks. However, other phenotypes associated with salm - clones, such as ectopic islands of triple row bristles at the margin, are observed with regularity in an ri background. It is proposed that a short-range diffusible signal X, functioning downstream of salm and salr induces expression of kni/knrl along the anterior border of the salm expression domain (Lunde, 1998).

To determine the importance of restricting kni expression to the L2 primordium, the GAL4/UAS system was used to mis-express kni or knrl at high levels in various patterns. The GAL4-MS1096 line drives expression of UAS-target genes ubiquitously throughout the dorsal surface of third instar wing discs, and weakly on the ventral surface in the anterior region of the disc. GAL4-MS1096-driven expression of either the UAS-kni or UAS-knrl transgenes eliminates expression of vein markers such as rho, the provein/proneural gene caupolican (caup), the lateral inhibitory gene Delta (Dl), and the proneural gene achaete on the dorsal surface of the wing disc. In contrast, these vein markers are expressed in normal patterns on the ventral surface, albeit at reduced levels, presumably reflecting the weak expression of GAL4 in ventral cells of GAL4-MS1096 discs. In addition, modulated expression of blistered (bs), which is lower in vein than intervein cells of wild-type discs, also disappears on the dorsal surface of GAL4-MS1096 wing discs. Thus, strong expression of kni or knrl on the dorsal surface of wing discs eliminates expression of both vein and intervein markers. Similarly, when GAL4-71B is used to drive UAS-kni or UAS-knrl expression in a central domain slightly broader than that of salm, distinctions between vein and intervein cells are eliminated within the region of GAL4 expression. In these discs, vein and intervein markers are expressed normally in the L5 primordium, which lies outside of the GAL4-71B expression domain. These data reveal that ectopic kni or knrl expression does not simply favor vein over intervein cell fates. Since strong uniform kni or knrl mis-expression is required to eliminate veins, higher levels of kni/knrl activity are necessary to inhibit vein formation than are required to induce expression of rho in or near the L2 primordium (Lunde, 1998).

In addition to activating rho expression, kni and knrl also are likely to positively autoregulate. Patterned mis-expression of kni using the GAL4/UAS system induces corresponding expression of the knrl gene and vice versa. Since kni and knrl appear to share cis-regulatory elements in third instar larval wing discs and during other stages of development, the reciprocal cross-regulation observed between kni and knrl is likely to reflect an autoregulatory function for these genes. kni function does not appear to be necessary for activating knrl expression in the L2 primordium, however, since elimination of kni function in large kni - clones covering both the dorsal and ventral components of L2 does not lead to any loss of the L2 vein. Another consequence of high level ectopic kni expression is strong down-regulation of salm expression. Since kni and knrl are normally expressed immediately adjacent to the anterior salm border, suppression of salm expression by kni appears to sharpen the anterior salm border and refine the position of the L2 primordium. Thus, the kni/knrl locus functions downstream of spalt major (salm) and upstream of genes required to initiate vein-versus-intervein differentiation. Mis-expression experiments suggest that kni and knrl expressing cells inhibit neighboring cells from becoming vein cells. kni and knrl are likely to refine the L2 position by positively auto-regulating their own expression and by providing negative feedback to repress salm expression (Lunde, 1998).

It is possible to trace formation of the L2 vein that occurs during the larval stage back to early A/P patterning that occurs during segmentation in the embryo. This chain of events leads to activation of the kni and knrl genes in narrow stripes at the anterior edge of the salm expression domain, thus linking positional information to morphogenesis. It is proposed that salm activates expression of a short-range signal X, which induces expression of kni and knrl in adjacent salm non-expressing cells. Since Kni and Knrl are members of the steroid hormone receptor superfamily, it is possible that the signal X could be a lipid-soluble factor, which binds and activates Kni and Knrl. However, given the minimal sequence conservation between Kni and Knrl in the putative ligand binding regions of these proteins, this direct form of signaling seems unlikely. Once activated, kni and knrl organize formation of the L2 primordium. kni and knrl are proposed to organize development of the L2 vein primordium through a variety of concerted actions. A key target gene activated by kni and knrl in the L2 primordium is the vein-promoting gene rho, which potentiates signaling through the Egf-receptor/RAS pathway. Because low levels of ubiquitous kni and knrl expression preferentially promote vein development near the location of L2, another activity provided at the anterior boundary of the salm expression domain is likely to act in parallel with the kni and knrl genes to define the position of the L2 primordium. This parallel genetic function may be supplied by the signal X, hypothesized to induce kni and knrl expression in salm non-expressing cells (Lunde, 1998).

kni and knrl are also likely to suppress vein development in neighboring intervein cells since strong uniform mis-expression of kni and knrl eliminates veins. This result could be explained if kni and knrl normally activate expression of a signal that suppresses vein development in neighboring intervein cells. Such a lateral inhibitory function presumably restricts formation of the L2 primordium to a narrow linear array of cells. To account for the fact that kni and knrl do not turn themselves off in L2 as a consequence of the proposed lateral inhibitory signaling, it is imagined that these cells are refractory to the lateral inhibitory mechanism. Alternatively, the hypothetical signal X, which promotes kni and knrl expression in cells adjacent to the salm expression domain, might continue to exert an inductive influence that overrides lateral inhibitory signaling in the L2 primordium. This possibility is consistent with low levels of ubiquitous kni expression rescuing rho expression in the vicinity of the normal L2 primordium in ri mutants. Although the nature of the proposed lateral inhibitory mechanism is unknown, the Notch signaling pathway is an obvious candidate, since loss of Notch function during late larval stages results in the formation of much broadened rho expressing stripes. Since Delta is unlikely to be the ligand mediating lateral inhibition, due to its absence in the L2 primordium, another Notch ligand might be activated in response to kni and knrl to suppress the vein fate in neighboring cells. It is also possible that a different type of signaling pathway is involved in this process. Finally, kni and knrl are likely to maintain and sharpen the anterior salm border through a combination of autoactivation and negative feedback on salm expression. Kni and Knrl may repress salm expression directly or could function indirectly through an intermediate tier of regulation. The ability of ectopic kni or knrl expression to suppress expression of salm as well as vein markers, but not to suppress expression of genes involved in defining the A/P organizing center (i.e. hh, dpp and ptc), is consistent with kni and knrl functioning at the last step in defining positional information required for placement of the L2 primordium. It will be interesting to determine whether there are genes functioning analogously to kni and knrl, that specify the positions of other longitudinal veins along the A/P axis of wing imaginal discs (Lunde, 1998).

The model proposed for activating expression of kni and knrl in a narrow stripe of cells is analogous to the earlier induction of dpp in a narrow stripe of anterior compartment cells by the short-range Hh signal emanating from the posterior compartment. In both cases a domain-defining gene (i.e. en or salm) activates expression of a short-range signal (i.e. Hh or X), while preventing these same cells from responding to the signal. According to such a genetic wiring diagram, only cells that are immediately adjacent to cells producing the short-range signal are competent to respond to it. This set of constraints restricts the expression of target genes to narrow stripes or sharp lines. An exquisite example of linear gene activation is the initiation of single minded expression in a single row of mesectodermal cells abutting the snail expression domain in the mesoderm of blastoderm embryos. Direct mechanisms contribute to activating sim in this precise pattern, because snail represses sim expression in ventral cells and Dorsal and Twist collaborate to define a relatively sharp threshold for activating sim that extends a short distance beyond the snail border. However, these direct transcriptional mechanisms alone do not seem sufficient to explain the absolutely faithful linear path of sim expression in a single row of cells along the irregular contour of snail expressing mesodermal cells. Perhaps communication between snail expressing cells and their immediate dorsal neighbors plays a role in achieving the invariant registration of the sim and snail expression patterns. In support of a role for cell-cell communication in this process, initiation of sim expression in the blastoderm embryo requires signaling through the Notch/Delta/E(spl) pathway. Furthermore, in the mesoderm, ubiquitously supplied maternal Delta protein is rapidly retrieved from the surface in the form of multi-vesicular bodies, which is typical of ligands involved in active signaling. Thus, Snail may regulate expression of some co-factor required for membrane bound Delta to productively activate the Notch signaling pathway in adjacent cells, which are free to respond by activating sim expression. It is noteworthy that in each of three cases considered above, products of entirely distinct domain-defining genes (e.g. En, Salm and Sna) induce the linear expression of genes in adjacent cells by activating production of short-range signals (e.g. Hh, X, Dl) while simultaneously suppressing response to those signals. The width of the target gene stripes presumably depends on the range of the signal and on the level of signal required to activate expression of specific genes. Thus, Hh activates expression of the target gene dpp in a domain 6-8 cells wide; the hypothetical factor X acts more locally to induce expression of kni and knrl in a stripe 2-3 cells wide, and the putative (activated) form of membrane tethered Delta induces sim expression in a single row of abutting mesectodermal cells. Perhaps this ëfor export onlyí signaling mechanism is a general scheme for drawing lines in developing fields of cells (Lunde, 1998 and references).

GENE STRUCTURE

Exons - three

Bases in 3' UTR - 508

PROTEIN STRUCTURE

Amino Acids - 429

Structural Domains

Knirps is a zinc finger steroid/thyroid orphan receptor-type transcription factor (Nauber, 1988).

knirps and knirps-related are closely linked in the 77E1,2 region on the left arm of the third chromosome. They encode nuclear hormone-like transcription factors containing almost identical Cys2/Cys2 DNA-binding zinc finger motifs which bind to the same target sequence (Rothe, 1994).

A comparative tree of DNA-binding domain amino acid sequences reveals the evolutionary affinities of Drosophila nuclear receptor proteins. Knirps shows no close affinities to other nuclear receptor proteins. Drosophila Ecdysone receptor sequence is most similar to murine RIP14. Tailless has a close affinity to murine Tlx. Drosophila E78 and E75 fall in the same subclass as Rat Reverb alpha and beta, and C. elegans "CNR-14." Drosophila HR3 is in the same subclass as C. elegans "CNR-3." Drosophila HNF-4 is most closely related in sequence to Rat HNF-4. Drosophila Ftz-F1 and Mus ELP show sequence similarity to each other. Drosophila Seven up is closely related to Human COUP-TF. Drosophila Ultraspiracle is in the same subfamily as Human RXRalpha, Human RXRbeta, and Murine RXRgamma. The latter two groups, containing Ultraspiracle and Seven up, show a distant affinity to each other. Four other subfamilies show no close Drosophila affinities. These are: 1) C. elegans rhr-2, 2) Human RARalpha, beta and gamma, 3) Human thyroid hormone receptor alpha and beta, and 4) Human growth hormone receptor, glucocorticoid receptor, and progesterone receptor (Sluder, 1997).

date revised: 15 November 98

Home page: The Interactive Fly © 1995, 1996 Thomas B. Brody, Ph.D.

The Interactive Fly resides on the
Society for Developmental Biology's Web server.