BIOLOGICAL OVERVIEW

Mammalian TAK1 was initially identified by a complementation assay in yeast for its ability to substitute for the MAPKKK Ste11p in the yeast MAPK pheromone pathway (Yamaguchi, 1995). TAK1 is a MAPKKK and contains a protein kinase domain in its N-terminal region that is ~30% identical to the catalytic domains of RAF1 and MEKK1. In mammals TAK1 can function in a signal transduction pathway that is triggered by the TGFß superfamily of ligands. In addition, the overexpression of Xenopus TAK1 and an upstream activator called TAB1 induce embryonic ventralization. Studies of dominant negative mutant forms of TAK1 place TAK1 activity downstream of the BMP2/4 receptors (Shibuya, 1998; Vidal, 2001 and references therein).

Besides TGFß signaling, TAK1 appears to be involved in various other signal transduction pathways. Cell culture experiments indicate that ceramide stimulates the TAK1 kinase activity and that ceramide-induced JNK/SAPK activation is blocked by a dominant negative TAK1 mutant (Shirakabe, 1997). Biochemical studies also show that MAPKK4 and MAPKK3/MAPKK6 can be phosphorylated by TAK1, suggesting that TAK1 can activate the JNK pathway and/or the P38 MAPK pathways (Moriguchi, 1996; Wang, 1997), and other experiments demonstrate interactions between TAK1 and the Wnt signaling pathway (Ishitani, 1999). Several studies have pointed toward a role for TAK1 in the IL1-R/NF-kappaB signaling pathway (Sakurai, 1998; Ninomiya-Tsuji, 1999; Sakurai, 1999). Dominant negative TAK1 mutants prevent NF-kappaB activation by IL1 or the TLRs (Ninomiya-Tsuji, 1999; Irie, 2000), and TAK1 appears to activate NF-kappaB by stimulating the IkappaB kinase complex (IKK) that induces IkappaB degradation by the proteasome (Ninomiya-Tsuji, 1999; Sakurai, 1999). Furthermore, TAK1 interacts with TRAF6, which is a downstream effector of the IL1 receptor, suggesting that TAK1 links TRAF6 to the NIK-IKK cascade in the IL1-R/TLR signaling pathways (Ninomiya-Tsuji, 1999; Vidal, 2001 and references therein).

Drosophila Tak1 is required for Relish cleavage and antibacterial immunity (Vidal, 2001). Ectopic activation of Tak1 signaling leads to a small eye phenotype, and genetic analysis reveals that this phenotype is a result of ectopically induced apoptosis. Genetic and biochemical analyses also indicate that the c-Jun amino-terminal kinase (JNK) signaling pathway is specifically activated by Tak1 signaling. Expression of a dominant negative form of Tak1 during embryonic development results in various embryonic cuticle defects including dorsal open phenotypes (Takatsu, 2000). Phenotypes generated by overexpressing a dominant negative form of Tak1 (Takatsu, 2000; Mihaly, 2001) are not observed in the Tak1-deficient mutants (Vidal, 2001), although the dominant negative Tak1 mutant does phenocopy the Tak1 mutant immune phenotype (Vidal, 2001).

Dorsal closure, taking place in mid-embryogenesis, describes the morphogenetic movements of the epidermis in order to replace the amnioserosa on the dorsal side of the embryo. This event is driven by the concerted spreading of epidermal cells towards the dorsal midline, where the two contralateral epidermal cell layers meet and remain connected. The JNK signaling module and nuclear targets of JNK, the AP-1 transcription factors dJun and dFos, control the process of dorsal closure. Uncompleted or failed dorsal closure is indicative of disrupted JNK signaling. Mutations in all known components of the JNK signaling pathway result in dorsal open embryos (Mihaly, 2001).

During dorsal closure, the expression of the dpp and puc genes in cells of the leading edge is controlled by the JNK kinase module and the AP-1 transcription factors Jun and Fos. Leading edge cells show loss of puc and dpp expression when deficient for JNK signaling. Conversely, constitutive activation of JNK signaling in the embryonic epidermis by expressing Rac^V12 or Dcdc42^V12 (via the UAS/GAL4 system) induces the upregulation of dpp and puc (Mihaly, 2001 and references therein).

To address the question of whether Tak1 can activate and act through the JNK MAPK module, Tak1 was expressed under the control of the en-GAL4 and pnr-GAL4 drivers and the induction of dpp and puc were monitored in the epidermis of stage 12-15 embryos (pnr is strongly expressed in cells of and neighboring the leading edge). In wild-type, dpp is expressed in two lateral stripes along the Drosophila embryo, and the dorsal most stripe corresponds to the leading edge of the epidermis. Overexpression of Tak1 with either GAL4 driver causes ectopic upregulation of dpp, as monitored by RNA in situ hybridization. Similarly, the analysis of embryos carrying one copy of the puc^E69 lacZ enhancer trap by beta-galactosidase activity staining shows a clear and robust ectopic puc expression when Tak1 is overexpressed. These patterns of dpp and puc activation by Tak1 are identical to those observed with activated Jun, suggesting that the effect is direct and mediated by the JNK signaling pathway (Mihaly, 2001).

In summary, these data indicate that overexpression of Tak1 in the embryonic ectoderm is sufficient to induce high-level expression of both known JNK target genes. As the same upregulation of puc and dpp is observed with JNKK/Hep^Act and Jun (a JNK activated transcription factor), the data strongly suggest that Tak1 acts through the JNK/Jun(AP-1) module in the context of dorsal closure (Mihaly, 2001).

In the absence of a Tak1 mutation, fly stocks carrying a dominant negative (DN) version of Tak1 (UAS-Tak1^K46R or UAS-Tak1^D159A) were generated to mimic the LOF situation. Embryonic expression of such kinase dead Tak1 proteins by using the 69B-Gal4, pnr-Gal4 or arm-Gal4 driver lines does not significantly affect the development of the embryos. Although various cuticle defects were detected, the penetrance of these phenotypes, including segmentation, head involution and dorsal closure defects, were below 3% and thus were considered insignificant (Mihaly, 2001).

The expression of DN proteins might induce unspecific pleiotropic effects by interfering with other kinases or signaling pathways, and/or only reduce the function of a given protein partially. Moreover, although kinase dead DN-Tak1 constructs worked in other tissues, the results with a DN-Tak1 are in contrast to the observation by Takatsu (2000) that reports 7% of embryos with dorsal closure phenotypes and an additional 25% displaying head involution defects. The experiments reported in this paper do not reveal any significant embryonic phenotypes. Thus, the double-stranded RNAi technique was employed to analyze a Tak1 LOF phenotype in the embryo. RNAi has been shown to effectively block the function of the gene of interest. In addition, it is strictly gene-specific for the RNA injected (Mihaly, 2001).

Injection of double-stranded Tak1 RNA into wild-type embryos results in a variety of embryonic defects. About 17.5% of the embryos exhibit an anterior open phenotype. However, dorsal closure defects are seen rarely in only about 3% of embryos, raising the question of whether Tak1 is required for dorsal closure as expected from the overexpression studies. Nevertheless, the anterior open phenotype appears to be significant since control injections of other double-stranded RNAs (dsRNAs), where the background of such phenotypes was only 1% (approximately) (Mihaly, 2001).

An explanation for the very low penetrance of the dorsal closure phenotype might be a general insensitivity to interfere with the process by means of RNAi. In order to address this issue, a control experiment was performed injecting D-jun (a well established factor required for dorsal closure) dsRNA into preblastoderm embryos. For D-jun injections, over 90% of the embryos display strong dorsal open cuticles, mimicking the D-jun LOF null phenotype. These results indicate that dorsal closure can be disrupted by RNAi-based methodology. Since Tak1 transcripts are abundant in preblastoderm embryos, determined by RNA in situ hybridization, a possible caveat to the RNAi Tak1 experiment could be maternally-derived protein that cannot be overcome by means of RNAi. However, this maternal product should still be inhibited by the overexpression of DN-Tak1, which was not seen (Mihaly, 2001).

Taken together, these DN-Tak1 overexpression and RNAi experiments suggest that Tak1 is not strictly required for dorsal closure and rather might serve an auxiliary role in the process. However, as this is in contrast to the results of Tak1 overexpression (when JNK target gene expression is upregulated very efficiently), it might be possible that Tak1 is (at least in part) redundant with another JNKKK during embryogenesis, and thus even a complete removal of it would be compensated by other kinases (Mihaly, 2001).

In contrast to the lack of a critical role for Tak1 in dorsal closure, Tak1 is required during morphogenetic changes and the fusion of the epithelial wing disc cell layers that takes place in thoracic closure, acting in the context of JNK signaling. JNK signaling is required in thoracic closure. The notum of the adult animal is formed by tissue of the two collateral wing imaginal discs, which undergo extensive morphogenetic rearrangements during metamorphosis. LOF in hep/JNKK and kayak/D-Fos results in aberrant wing disc morphogenesis and failure of wing disc fusion, giving rise to a thoracic cleft along the dorsal midline in the adult. To test whether Tak1 can also act in this context, DN (kinase dead) forms of Tak1 (UAS-Tak1^K46R or UAS-Tak1^D159A) were overexpressed with ap-GAL4 and pnr-GA66L4 in the thoracic parts of the wing discs. Examination of such flies shows incomplete closure of the thorax, giving rise to a mild thoracic cleft at the dorsal midline of the notum. Although this phenotype is relatively weak, it has a very high penetrance of 91% and is highly reminiscent of that observed in hypomorphic allelic combinations of either hep/JNKK or kay/d-fos. This is also in agreement with the phenotypic result of expressing Puckered, a negative regulator of JNK signaling at the time when wing disc fusion occurs. Puc overexpression driven by pnrGAL4 leads to the same phenotypic thorax cleft defects. These observations suggest that Tak1 is required during morphogenetic changes and the fusion of the epithelial wing disc cell layers, acting in the context of JNK signaling (Mihaly, 2001).

Overexpression of wild-type forms of Tak1 in eye imaginal discs results in defects in ommatidial planar polarity. The Drosophila JNK cascade has also been implicated in the establishment of correct ommatidial polarity in the compound eye where it acts downstream of Fz/Dsh. In the Drosophila eye, planar polarization is reflected in the mirror-symmetric arrangement of the ommatidial units relative to the dorso-ventral midline, the so-called equator. This pattern is generated early in the third instar imaginal disc, when immature ommatidia rotate 90° towards the equator. Subsequently, they adopt opposite chiral forms depending upon whether they lie dorsally or ventrally off the equator. Mutations in planar polarity genes like frizzled (fz) or dishevelled (dsh) result in the loss of mirror-image symmetry, with the ommatidia being misrotated and having acquired random chirality. Similarly, overactivation of any known component of the Fz/planar polarity pathway (including the JNK components) in both cells of the R3/R4 pairs leads to randomization of rotation and chirality causing a phenotype similar that of the LOF mutants (Mihaly, 2001).

An examination was carried out to see whether directed overexpression of Tak1 in the eye imaginal disc of third instar larvae (at the time of planar polarity Fz/JNK signaling) can interfere with the correct establishment of planar polarity. To this end UAS-Tak1 was expressed in photoreceptor precursors R3/R4 in the eye imaginal disc (under the sev-enhancer GAL4 driver: sev>Tak1). This type of overexpression creates specific eye planar polarity phenotypes with Fz, Dsh and other components of planar polarity signaling. Weak Tak1 expression (by rearing the flies at 18°C) causes a specific phenotype reminiscent of that caused by the components of planar polarity signaling, with polarity defects affecting both rotation and chirality, and also some loss of photoreceptors. This phenotype is already evident with the appropriate markers (e.g. svp-lacZ) at the time of planar polarity establishment in the third instar eye disc, indicating that it is a primary defect in polarity establishment, and not due to late differentiation defects (Mihaly, 2001).

The GOF sev>Tak1 phenotype provides a tool to test for genetic interactions with mutations in components of the Fz/planar polarity pathway and other signaling cascades. In such genetic interaction assays, it was found that reducing the dosage of the JNK signaling components (hep, bsk and D-jun) causes a strong suppression of the sev>Tak1 phenotype. These results are consistent with Tak1 acting upstream of the JNK module in polarity signaling, and support the notion that Tak1 can act generally upstream of JNK signaling (Mihaly, 2001).

Several other signaling pathways and kinases were tested for genetic interaction with sev>Tak1. Whereas no interaction with components of the Dpp signaling pathway was found, dominant suppression, comparable to that of the JNK components, was found with deficiencies removing the p38 kinases (p38a and p38b) and mutant alleles of nemo. These interactions suggest that the sev>Tak1 eye phenotype depends in part on the activities of these other MAPKs as well and is consistent with the previously reported tissue culture experiments. The observation that the sev>Tak1 phenotype is less sensitive to the dosage of msn/STE20 might indicate that msn is acting upstream of Tak1 (Mihaly, 2001).

It is concluded that during Drosophila development Tak1 can act as a component of the JNK signal transduction module. In addition, genetic interactions during eye development suggest that Tak1 can also activate other kinases, for example p38 type MAPKs and the Nemo kinase, in this context. Although TAK was initially identified as a TGF-beta activated kinase in mammalian cell culture systems (Yamaguchi, 1995), no evidence has been found by either phenotypic or genetic interaction analysis for Tak1 acting in a dpp/TGF-beta-dependent manner during Drosophila development (Mihaly, 2001).

Initially, the TAK subfamily of MAPKKK kinases has been identified as being activated by TGF-beta signaling (Yamaguchi, 1995). Subsequently, tissue culture experiments have shown that these kinases are able to activate both JNK and p38 MAPK pathways. These studies with Drosophila indicate that Tak1 can act in the context of the JNK cascade. This conclusion is based on several lines of evidence: (1) overexpression of Tak1 in embryos activates the expression of the JNK target genes dpp and puc; (2) expression of DN Tak1 with pnrGAL4 or apGAL4 leads to thorax closure defects, also a reported JNK LOF phenotype; (3) overexpression of Tak1 in third instar eye imaginal discs interferes with the correct establishment of ommatidial polarity, reminiscent of Hep, Bsk and Jun; (4) the GOF Tak1 planar polarity phenotype is suppressed by hep, bsk and jun LOF alleles. Taken together, these observations argue for Tak1 acting through the JNK cascade in Drosophila (Mihaly, 2001).

However, the analyses of the Tak1 LOF phenotypes generated by RNAi and DN isoforms during embryogenesis are in conflict with this conclusion and argue for a requirement that is in part redundant. The RNAi analysis has revealed predominantly anterior open embryos, and only very rarely the typical dorsal closure defects, normally seen in mutants of JNK signaling components. In addition, expression of a DN-Tak1 in the embryo in this study did not produce any dorsal closure phenotypes. Thus, although anterior open phenotypes can also be observed in mutants defective for JNK signal transduction (usually associated with hypomorphic alleles), the lack of clear dorsal closure defects suggests that Tak1 could be either partially redundant or plays a neglectable role in this JNK-mediated process. Taken together with the GOF produced by overexpression of Tak1, resulting in the robust activation of JNK target genes, these observations support the idea that Tak1 might act partially redundantly in this process during embryogenesis. Alternatively, the data can be viewed as Tak1 being able to generally act through the JNK signaling module wherever it is activated, leaving the question of the actual physiological role open. It is quite possible that Tak1 mediates JNK signaling only in a subset of the wide range of biological processes assigned so far to JNK signal transduction in Drosophila (Mihaly, 2001).

The studies reported by Takatsu (2000) found that Tak overexpression behind the MF can affect photoreceptor differentiation and causes cell death, but has otherwise no major effect on eye development (although mild defects in pigment and bristle cell differentiation can be revealed). In contrast, Tak1 overexpression with the sev-Gal4 driver severely interferes with the correct establishment of ommatidial polarity. The difference between these results can be explained by the different GAL4 drivers used. Whereas sevGAL4 is expressed at high levels early (and transiently) in the R3/R4 precursors, GMR-GAL4 (used by Takatsu) only affects later processes and can cause cell death. It is worth noting that the GMR-GAL4 driver does not cause planar polarity defects even with known planar polarity genes like fz and RhoA. The GMR-GAL4-driven expression starts in all cells behind the MF but early on levels are too low to cause defects. For example, GMR-driven expression of RhoA does not cause any appreciable defects in the third instar eye imaginal disc (when photoreceptors are recruited and polarity is established), although the corresponding adult eye is strongly reduced and malformed (Mihaly, 2001).

The sev>Tak1 eye phenotype is highly reminiscent of those caused by sev-Fz and sev-driven activated JNK pathway components. Moreover, since it can be suppressed by JNK LOF alleles, it supports a link between Tak1 and JNK and their role in regulating planar polarity in the eye downstream of Fz (Mihaly, 2001).

Genetic studies indicate that Tak1 can act through kinases other than JNK. Consistent with previous cell culture studies (Moriguchi, 1996), the p38 type MAP kinases also suppress the Tak1 GOF phenotype. This observation is also in line with the proposal that the Drosophila bsk/JNK and p38a and p38b kinases act as redundant elements in planar polarity signaling. The TAK MAPKKK family has also been implicated in activating a Nemo-like kinase (Nlk)/LIT-1 in cell culture and in C. elegans embryos (Ishitani, 1999; Meneghini, 1999 and Shin, 1999). Interestingly, nemo (the founding member of the Nlk subfamily) has been first described as a Drosophila planar polarity gene, affecting ommatidial rotation. Mutations in nemo strongly suppress the Tak1-induced eye polarity phenotype, raising the possibility that a TAK-Nemo/LIT-1 signaling cassette is evolutionarily conserved in worms, flies and vertebrates. Moreover, the existence of a Tak1-Nemo signal relay would suggest an interesting new link between the JNK signaling pathway and the Nemo kinases. The upstream activators and downstream targets of the latter and their role in planar polarity generation are unknown. In such a scenario Tak1 could be acting upstream at a branching point of the JNK and NLK signaling routes (Mihaly, 2001).

Strikingly, the TAK function mediated by Nemo-like kinases in vertebrate cell culture studies and C. elegans appears to antagonize canonical Wnt signaling at the level of Arm/TCF. As the canonical Wnt pathway and Fz/planar polarity signaling split at the level of Dsh, it is interesting to note that they might again intersect further downstream to reinforce the difference in signaling by antagonizing the other signaling branch. Further insight into the TAK-Nemo link will help to clarify the pathway specificity (Mihaly, 2001).

Several studies have demonstrated an involvement of the JNK pathway in stress-induced apoptosis in vertebrate cells. The Drosophila JNK pathway has also been implicated in regulating apoptosis upon deregulation of dpp and wg signaling in the developing wing disc. Tissue culture and Xenopus embryo injection experiments have shown that members of the TAK family of MAPKKKs induce apoptosis when overexpressed. Consistent with this, Tak1 overexpression in imaginal discs also induces apoptosis. Is the Tak1-induced apoptosis mediated through the JNK pathway? There are at least two observations supporting this possibility: (1) overexpression of a constitutively activated form of Hep (Hep^Act) or Jun (Jun^Asp) can lead to a qualitatively very similar phenotype as Tak1 overexpression with reduced eye size; (2) the apoptosis phenotype generated by Tak1 overexpression is strongly suppressed by hep, bsk and jun LOF alleles. Thus, a TAK-JNK signaling module might control apoptosis not only in vertebrates, but also in flies. It remains unclear whether the overexpression of Tak1 and high JNK activation directly activates the apoptotic pathway or whether cell death is a secondary consequence of it (Mihaly, 2001).

In summary, it has been demonstrated that the Drosophila TAK ortholog Tak1 can act as a JNKKK in the context of all described JNK-mediated morphogenetic processes during fly development, as well as during JNK-induced apoptotic cell death. However, the lack of a statistically significant dorsal closure phenotype in the embryo suggests that it is not the only JNKKK involved in this context (Mihaly, 2001).

GENE STRUCTURE

cDNA clone length - 3349

Bases in 5' UTR - 915

Exons - 8

Bases in 3' UTR - 397

PROTEIN STRUCTURE

Amino Acids - 678

Structural Domains

The overall Drosophila Tak1 protein is similar to the vertebrate TAK1s, containing an NH₂-terminal protein kinase domain as well as a long COOH-terminal domain. The kinase domain shows 56% identity and 73% similarity with the amino acid sequences of mTAK1. Phylogenetic analysis of the catalytic domain of Drosophila Tak1 with those of other Drosophila and vertebrate MAPKKK proteins indicate that its closest relatives are the vertebrate TAK1s. In the COOH-terminal region, Drosophila Tak1 is less well conserved with its vertebrate homologs. However, there is a stretch of 65 amino acids that is relatively well conserved between Drosophila and vertebrates (37% identity and 57% similarity to mTAK1). Interestingly, this region is almost completely missing in one of the alternative spliced forms of the human TAK1 (hTAK1c) (Takatsu, 2000 and references therein).

date revised: 20 August 2001

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.