Two Drosophila suppressors of cytokine signaling (SOCS) differentially regulate JAK and EGFR pathway activities

Abstract

Background: The Janus kinase (JAK) cascade is an essential and well-conserved pathway required to transduce signals for a variety of ligands in both vertebrates and invertebrates. While activation of the pathway is essential to many processes, mutations from mammals and Drosophila demonstrate that regulation is also critical. The SOCS (Suppressor Of Cytokine Signaling) proteins in mammals are regulators of the JAK pathway that participate in a negative feedback loop, as they are transcriptionally activated by JAK signaling. Examination of one Drosophila SOCS homologue, Socs36E, demonstrated that its expression is responsive to JAK pathway activity and it is capable of downregulating JAK signaling, similar to the well characterized mammalian SOCS.

Results: Based on sequence analysis of the Drosophila genome, there are three identifiable SOCS homologues in flies. All three are most similar to mammalian SOCS that have not been extensively characterized: Socs36E is most similar to mammalian SOCS5, while Socs44A and Socs16D are most similar to mammalian SOCS6 and 7. Although Socs44A is capable of repressing JAK activity in some tissues, its expression is not regulated by the pathway. Furthermore, Socs44A can enhance the activity of the EGFR/MAPK signaling cascade, in contrast to Socs36E.

Conclusions: Two Drosophila SOCS proteins have some overlapping and some distinct capabilities. While Socs36E behaves similarly to the canonical vertebrate SOCS, Socs44A is not part of a JAK pathway negative feedback loop. Nonetheless, both SOCS regulate JAK and EGFR signaling pathways, albeit differently. The non-canonical properties of Socs44A may be representative of the class of less characterized vertebrate SOCS with which it shares greatest similarity.

Figure 1

Protein sequence comparison of Drosophila and mouse SOCS. (A) The predicted carboxyl terminal protein sequences of Drosophila (d), mouse (m), and C. elegans (ce) SOCS genes, including the SH2 and SOCS box domains, are aligned and shaded to indicate similarities and identities. (B) Based on the protein alignments, the neighbor-joining method was used to construct a phylogenetic tree of these SOCS.

Figure 3

Loss of JAK activity does not affect Socs44A expression. As compared with wild-type at various embryonic stages (A and B), germline clone derived embryos from hop^c111 mothers (C-H) display dramatically reduced or eliminated expression of Socs36E (C and D). Only a stripe of mesodermal staining in germ band extended embryos (D) remains at nearly normal intensity in the mutant embryos. In contrast, expression of Socs44A in trachea persists in hop^c111 germline clone-derived embryos that are unrescued (E) or paternally rescued (F). However, the trachea are morphologically altered and drastically reduced in unrescued (G) and paternally rescued (H) animals, as compared with wild-type (I), as evidenced by a trachealess enhancer trap (G-I).

Figure 2

Socs36E and Socs44A are expressed in different spatio-temporal patterns. The embryonic expression patterns of upd and Socs36E are dynamic from early blastoderm throughout embryogenesis [see 28, 29 and Fig. 3]. Socs44A expression is not detected until very late stages in the trachea (A). Although such staining can be artifactual, sense strand probe never showed any staining (B). In the ovary, upd is expressed specifically in the polar follicle cells at each end of the chamber (C). Socs36E expression encompasses the anterior and posterior follicular epithelium, with highest expression at the poles (D). This is consistent with activation of Socs36E transcription due to reception of the Upd ligand which is secreted from the polar follicle cells and diffuses toward surrounding cells. Socs44A expression is restricted to the germline and only during later stages of oogenesis (E)

Figure 4

Socs44A misexpression reduces JAK signaling in the wing. Wild-type venation (A) is compared with a viable hop mutant, hop^msv/hop^M38 (B). hop reduction causes ectopic vein (arrow) near the posterior crossvein. (C) Expression of UAS-Socs36E using the engrailed-GAL4 driver (e16E-GAL) produces a similar ectopic vein phenotype, plus the loss of the anterior crossvein (arrowhead). (D) Similar misexpression of Socs44A causes ectopic wing vein production near the posterior crossvein (arrow) and arching of vein L3 (arrowhead). (E) Reduction of the dosage of hop enhances the Socs44A misexpression phenotype. (F) Misexpression of hop in the posterior compartment causes dramatic vein loss, but that loss is restored by the simultaneous expression of Socs44A (G).

Figure 5

Socs44A increases activity of EGFR signaling. The ectopic wing vein phenotype of Socs44A misexpression (A) is rescued by reduction of Egfr (B), Sos (C) or Ras85D (D), positive effectors of EGFR signaling. In contrast, reduction of argos, a negative regulator of EGFR signaling, enhances the Socs44A misexpression phenotype (E). The argos allele combined with en-GAL have no effect on venation without the UAS-Socs44A transgene (F). Certain heteroallelic Egfr mutants possess a distinct wing vein phenotype, whereby the anterior crossvein and the central portion of L4 is missing (G, arrows). Engrailed-driven misexpression of argos has a similar phenotype (H and J). Concurrent misexpression of Socs44A antagonizes argos misexpression to restore near normal wing venation (I and K). The designation "2xUAS-argos" refers to presence of 2 total copies of the transgene in the genome.

Figure 6

Socs44A deficiencies enhance argos misexpression phenotypes. (A) The engrailed-GAL4 driven misexpression of argos produces a range of phenotypes which were classified based on severity. The combination of en-GAL and Df(2)CA53 had no effect on venation. (B) In flies that were also heterozygous for Df(2)CA53, which removes the Socs44A locus, the distribution of phenotypes was significantly shifted to more severe classes as compared to animals heterozygous for Df(2)Drl^rv18, an overlapping deficiency that does not remove Socs44A or for Sco, a chromosome wild-type for the 44A region.

Figure 7

Socs36E and Socs44A have different activities during oogenesis. In wild-type ovaries (A, B), pnt-lacZ (red) is expressed in a gradient in the posterior terminal cells. Cells that lack hop activity (marked by a lack of green, see outline), also fail to express pnt-lacZ (C-E). Similarly, UAS-Socs36E misexpressed in clones (marked by presence of green, see outline), lack pnt-LacZ expression (F-H, see insets). In contrast, UAS-Socs44A misexpressed in clones (marked by presence of green, see outline), had no effect on pnt-LacZ expression (I-K). DAPI nuclear staining is shown in blue.