BIOLOGICAL OVERVIEW

For many years immunologists have held the view that the primary goal of the immune system is discrimination between self and non-self. Because Drosophila does not have an immunoglobulin-based immune response, the fly affords a unique opportunity to examine the nature of an immune system that evolved prior to the evolution of immunoglobulin as a primary driving force in an organism's response to pathogens. An examination of Drosophila suggests that an inducible system exists and it is based on the need to detect and protect against danger. The immunologist Matzinger (1994) has argued that such a system also remains functional as a controlling influence in the vertebrate immune response, in addition to the immunoglobulin-based system. The protein JNK is one of the pivotal actors in the Drosophila immune response, and appears to be functionally conserved in the vertebrate immune response as well. What does JNK do in both Drosophila and vertebrates, and what does this teach about the nature of the immune response?

Before taking a closer look at basket/JNK, a word on the pathway in which it functions is in order. Cascades of mitogen activated protein kinases (MAPKs) transduce signals from multiple extracellular stimuli, mediating responses such as cell proliferation, differentiation, and the regulation of metabolic pathways. There are multiple MAPKs in eukaryotes. In Drosophila, two such pathways have been identified. The first is a MAPK pathway involving Ras and Rolled, two components of the Ras pathway. It is required at least three times during development: as the terminal system which mediates responses to the Torso receptor, as neurogenic and wing vein pathways mediate responses to the EGF-receptor, and for the differentiation of photoreceptors, which mediate responses to the Sevenless receptor.

The second MAPK pathway in Drosophila involves the MAPK called Jun-N-terminal kinase (DJNK), also known as Basket. This pathway involves an additional kinase, Hemipterous, which serves to phosphorylate DJNK. Both proteins are required for dorsal closure, the sealing of the dorsal region of the embryo late in embryonic development. Dorsal closure involves coordinated changes in the shape of ectodermal cells, driven by cytoskeletal changes. (For a discussion of dorsal closure see Hemipterous). The phosphorylation cascade involving HEP and DJUN terminates in the phosphorylation of DJUN. Both DJNK and DJUN are activated in the the immune response of the fly. DJNK is closely associated on chromosome 2 with dror, a Drosophila receptor tyrosine kinase homolog of vertebrate TRK kinases, which mediate responses to neurotropins. Close chromosomal linkage often, but not always, occurs as a consequence of shared developmental function.

Basket (Bsk) is regulated by Src during dorsal closure. Mutants for Src42A, a Drosophila c-src protooncogene homolog, are described. Src42A functions in epidermal closure during both embryogenesis and metamorphosis. The severity of the epidermal closure defect in the Src42A mutant depends on the amount of Bsk activity, and the amount of Bsk activity depends on the amount of Src42A. Thus, activation of the Bsk pathway is required downstream of Src42A in epidermal closure. This work confirms mammalian studies that demonstrate a physiological link between Src and JNK (Tateno, 2000).

Genes that regulate cell shape changes in Drosophila are required for dorsal closure of the embryonic epidermis and thorax closure of the pupal epidermis. Mutations in genes such as hemipterous (hep) and basket (bsk, also known as DJNK) result in abnormal embryos with a dorsal hole or abnormal adults with a dorsal midline cleft. Hep and Bsk are homologous to the mammalian MKK7 (MAPK kinase 7) and JNK, and they are components of a MAPK (mitogen-activated protein kinase) cascade. Although the role of the Hep-Bsk cascade during dorsal closure has been extensively studied, the upstream trigger of this cascade is poorly understood. To identify the trigger, a screen was carried out for mutants showing the dorsal midline cleft phenotype, like a mild hep mutant. The mutant for Src42A shows this phenotype and Src42A regulates Bsk during Drosophila development (Tateno, 2000).

One line, Jp45, from the mutant collection of the P-element-inserted semilethal lines has been identified that survives to adulthood but shows various degrees of the dorsal midline cleft phenotype. Excision of the P-element eliminates the semilethality and restores the cleft phenotype. The P-element is inserted in the 5' untranslated region (UTR) of the Src42A gene, which encodes a Src-family nonreceptor tyrosine kinase. Ethyl methanesulfonate (EMS) mutant screening was used to isolate two strong alleles of Src42A, Src42A^E1 and Src42A^{myristylation} (^myri). In Src42A^E1, a stop codon at codon 483 eliminates the COOH-terminal part of the kinase domain of Src42A. Src42A^myri has a point mutation in codon 2, which causes an amino acid substitution from Gly² to Asp. Gly² is conserved in all members of the Src family and must be myristylated for localization of Src to the cellular membrane in mammals. About 50% of the Src42A^myri homozygotes die before they hatch, and most of the remainder die during the first-instar larval stage. Therefore, Gly² is required for development (Tateno, 2000).

Because the adult Src42A^Jp45 phenotype resembles that of hep, it was suspected that Src42A is involved in Hep and Bsk function. A mutation in hep or bsk dominantly enhances the lethality and the phenotypic severity of Src42A^Jp45 homozygotes. Conversely, reducing the gene dosage of puckered (puc), a gene encoding a phosphatase that inactivates Bsk, suppresses the lethality and the severity of the cleft phenotype of Src42A^Jp45. Thus, Src42A may function in the Bsk pathway during metamorphosis (Tateno, 2000).

Dorsal closure is the process by which a pair of epidermal layers elongate dorsally and fuse at the dorsal midline of the embryo. This process is not completed in hep and bsk mutants, yielding a dorsal open phenotype. Strong Src42A mutants do not show the dorsal open phenotype but display malformed mouth parts. This defect is similar to the defect in the embryo of the Tec29 mutant. Tec29 is a Src-related nonreceptor tyrosine kinase and is regulated by SRC64, another Drosophila Src homolog, during oogenesis. Thus, Tec29 may be involved in the function of Src42A. A mutation in Tec29 dominantly enhances the lethality of Src42A^Jp45 (Tateno, 2000).

Furthermore, the Tec29 Src42A double mutant shows complete embryonic lethality, and a certain fraction of the dead embryos show the dorsal open phenotype. Activated DJun, a transcription factor downstream of Bsk, partially rescues the dorsal open phenotype in the Tec29 Src42A double mutant. Thus, Src42A appears to regulate Bsk in the fusion of epithelial sheets during embryogenesis and metamorphosis, and Tec29 is involved in this regulation. The double mutant for Src64 and Src42A manifests a mild but clear dorsal open phenotype, which suggests a functional redundancy between Src64 and Src42A (Tateno, 2000).

Expression of puc is known to be induced by the Bsk signal. In the wing disc of the wild-type third-instar larva, puc is expressed in the dorsal midline of the adult notum. In the wing disc of the Src42A^Jp45 mutant, puc expression is reduced. In contrast, larvae with a constitutively activated form of Src42A (Src42A^CA) shows ectopic expression of puc. Further, introduction of a hep null mutation reduces the amount of ectopic puc expression. It is known that Bsk induces expression of puc and decapentaplegic (dpp) during embryonic dorsal closure. The embryos of the Tec29 Src42A double mutant do not show any puc or dpp expression in the leading edge cells. These results indicate that Src42A, Tec29, Hep, and Bsk regulate dpp and puc expression during embryonic dorsal closure (Tateno, 2000).

To investigate the ability of Src42A^CA to activate Bsk, the amount of phosphorylated Bsk was directly assessed by immunoblot analysis. Forced expression of Src42A^CA does not affect the quantity of total Bsk protein but induces more phosphorylated Bsk than the controls. Thus, Src42A appears to regulate the phosphorylation level of Bsk (Tateno, 2000).

During embryonic dorsal closure, the Hep-Bsk signal is required for elongation of the leading edge cells. In the absence of the Bsk signal, these cells do not fully elongate. The accumulation of F-actin and phosphotyrosine (P-Tyr) in leading edge cells is associated with the elongation of these cells. Accumulation of these substances is disturbed in the DJun and the puc mutants. In the double mutant for Tec29 and Src42A, the leading edge cells contain reduced quantities of F-actin and P-Tyr, and these cells are only partially elongated. Thus, the defect in embryonic dorsal closure in the Tec29 Src42A double mutant is caused by this failure in cell shape change, as is the case in the DJun mutant (Tateno, 2000).

A model is proposed in which Src42A, upon receiving an unidentified signal, activates the Hep-Bsk pathway to regulate cell shape change and epidermal layer movement. This is consistent with the observation in mammals that c-Src regulates the cell morphogenetic and migratory processes and is known to activate JNK. As in Drosophila, c-Src definitely affects F-actin organization and P-Tyr localization during cell morphogenesis. Therefore, Src regulation of JNK activity toward a change in cell shape may be conserved (Tateno, 2000).

It can be also interpreted that Src42A acts upstream of DFos, a dimerization partner of DJun. Although the Src42A, Tec29, and Src64 single mutants do not show a dorsal open phenotype, the DFos mutant clearly exhibits it. This relationship is also analogous to that in mammals. Both c-src and c-fos knockout mice have a similar defect, osteopetrosis caused by reduced osteoclast function. But the phenotypic severity is milder in c-src than in c-fos knockouts; this can be explained by the functional overlap in multiple Src-family tyrosine kinases. Accordingly, in both Drosophila and mammals, multiple nonreceptor tyrosine kinases may cooperate to regulate the function of the Jun/Fos complex (Tateno, 2000).

Bacterial infection causes insect immune tissues to synthesize a spectrum of antibacterial peptides, including Cecropin and Diptericin, which lyse the invading microorganisms (Hultmark, 1993 and Peterson, 1995). Exposure of Drosophila to environmental stress (such as a pathogenic onslaught) activates two groups of transcription factors: AP-1 (coded for by Drosophila jun and fos) and two related proteins; Dorsal and Dorsal related factor (Dif).

Dorsal and Dif are homologs of NFkappaB (also known as Rel), an essential transcription factor involved in the vertebrate immune response (Ip, 1993). Both Dif and Dorsal act in the insect immune defense against bacterial infection (see Dorsal for more information).

Drosophila JNK is activated by endotoxic lipopolysaccharide (LPS). LPS is a component of bacterial cell walls, and is known to be a stimulant for the immune response in both insects and mammals. The response of Drosophila to bacterial infection can be studied by examining the effect of LPS on cultured cell lines. Addition of LPS to cultured cell lines, including mbn-2 hemocytes and Schneider S2 embryonic cells, causes marked induction of cecropin and diptercin genes. basket is activated within 5 minutes of LPS addition. However, basket activation is transient and returns to basal levels after 1 hour. The activation of DJun by DJNK in LPS-treated cells may lead to increased AP-1 (a heterodimer of DFos and DJun) transcriptional activity. Targets of Drosophila AP-1 may include the DJun promoter (Sluss, 1996).

Although the pathway in Drosophila for pathogen activation of the DJNK-DJun system and the immune system targets of DJNK and DJun are not yet known, the cloning of DJNK should help clarify the roles of these proteins in the fly's immune response.

Mammalian JNK has been implicated in a variety of immune-related signaling pathways, including the response to pro-inflammatory cytokines, the respone to LPS, and T-cell activation (Su, 1994, Westwick, 1994 and 1995 and Raingeaud, 1995). The role of JNK and of Rel proteins in the vertebrate immune response harkens back to the roles of these proteins in Drosophila: both proteins are involved in systems that have developed to detect and protect against danger. When viewed from the perspective of signal transduction cascades and transcription factors it is clear that at its core, the vertebrate immune response system is one which detects and protects against danger. This conclusion is made clearer by looking at the roles of JNK and Rel in vertebrate responses to danger and in the related immune responses (see the basket/JNK Evolutionary homologs section).

The "simple" immune response of Drosophila, involving two signal transduction pathways and two groups of transcription factors, serves as a model for the inducible response to pathogens that is found in both invertebrates and vertebrates. This system functions quite adequately without the added level of antigen specific immunoglobulins.

Where then does the vertebrate immunoglobulin system come from? Invertebrates have a dozen or more cell surface proteins possessing domains resembling those found in vertebrate immunoglobulins. Examples in Drosophila include, Dlar, Fasciclin II, Fibroblast growth factor receptor 1, Frazzled, Hikaru genki, Neuroglian, and Semaphorin 2. The Ig domain, as it is found in invertebrates, is involved in cell-cell interactions, and many of the proteins have roles in the complex process whereby axons find their way to target tissues. It is likely that the immunglobulin system, both T and B cell based, evolved from invertebrate Ig domain proteins. What particular aspect of the Ig domain makes it ideal for functions as diverse as homophilic cell adhesion and antigen recognition? The folded structure of the Ig domain allows for a rigid central core that acts to stabilize exposed polypeptide loops. These exposed loops can freely vary in sequence without seriously disrupting the stable structure of the protein. This special structure of the Ig domain has permitted the multitude of functions for which it serves.

GENE STRUCTURE

cDNA clone length - 1831 bases

Bases in 5' UTR - 535

Exons - 6

Bases in 3' UTR - 141

PROTEIN STRUCTURE

Amino Acids - 372

Structural Domains

Drosophila Basket/JNK contains kinase subdomains I-XI. The dual phosphorylation motif Thr-Xaa-Tyr required for MAP kinase activation is located in kinase subdomain VIII (Thr-Pro-Tyr). Sequence comparison with other MAP kinases demonstrates that DJNK is a member of the JNK group of protein kinases. The amino acid sequence of DJNK is 65% identical to human JNK1. The amino acid sequence between kinase subdomains IX and X varies between the JNKs, and determines the efficiency of binding to c-Jun. In this region, DJNK is most homologous to JNK2. DJNK is related more distantly to the Drosophila rolled gene product ERKA (Sluss, 1996 and Riesgo-Escovar, 1996).

date revised: 16 Dec 96  

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