Negative Regulation of EGFR/MAPK Pathway by Pumilio in Drosophila melanogaster

In Drosophila melanogaster, specification of wing vein cells and sensory organ precursor (SOP) cells, which later give rise to a bristle, requires EGFR signaling. Here, we show that Pumilio (Pum), an RNA-binding translational repressor, negatively regulates EGFR signaling in wing vein and bristle development. We observed that loss of Pum function yielded extra wing veins and additional bristles. Conversely, overexpression of Pum eliminated wing veins and bristles. Heterozygotes for Pum produced no phenotype on their own, but greatly enhanced phenotypes caused by the enhancement of EGFR signaling. Conversely, over-expression of Pum suppressed the effects of ectopic EGFR signaling. Components of the EGFR signaling pathway are encoded by mRNAs that have Nanos Response Element (NRE)–like sequences in their 3’UTRs; NREs are known to bind Pum to confer regulation in other mRNAs. We show that these NRE-like sequences bind Pum and confer repression on a luciferase reporter in heterologous cells. Taken together, our evidence suggests that Pum functions as a negative regulator of EGFR signaling by directly targeting components of the pathway in Drosophila.


Introduction
A variety of cellular processes such as cell fate specification, proliferation, and apoptosis utilize epidermal growth factor receptor (EGFR) signaling. Upon activation, the signaling proceeds through Drk, Sos, and Ras activation, to a phosphorylation cascade involving Raf (MAPKKK) and Dsor1 (MEK). The pathway culminates in activation of rolled (rl) MAP kinase (MAPK), which phosphorylates a suite of substrates to determine a specific cellular response [1]. Since aberrant signaling results in abnormal organ formation or tumorigenesis [2], intricate spatio-temporal regulation of the signaling is essential. Thus diverse negative regulators are employed to precisely regulate EGFR signaling.
In Drosophila melanogaster, adult wing blade has five wing veins, which are differentiated in the wing imaginal disc during larval and pupal stages. EGFR signaling during the larval period promotes wing vein cell differentiation. Enhanced EGFR signaling results in the development of extra-wing veins, whereas reduced signaling results in wing vein loss [3][4][5][6][7][8][9][10]. Apparently, the levels of EGFR signaling are carefully regulated to ensure normal vein development.
Large bristles (macrochaetes) on the notum of adult flies arise from a single sensory organ precursor (SOP) cell in the wing imaginal disc during larval development. Each SOP cell is selected from a group of equipotent cells in a proneural cluster that is specified by high level expression of proneural genes such as achaete (ac) and scute (sc) [11][12][13][14][15]. Persistent expression of proneural genes in SOP cells requires EGFR signaling [16,17]. Reduced EGFR signaling results in loss of SOP cells in the disc and macrochaetae in the adult [4,16]. Conversely, excess EGFR signaling evokes supernumerary SOP cells by stimulating proneural gene expression [16], in turn causing the formation of extra bristles on the thorax and notum of the adult. Thus, as in the case of wing vein specification, the selection of SOP cells from proneural clusters requires precise regulation of EGFR signaling.
In our study, we observed extra wing veins and thoracic macrochaetes in pum mutants, which is reminiscent of phenotypes associated with up-regulation of EGFR signaling. Our genetic interaction analysis suggests that Pum functions as a negative regulator of EGFR signaling. Pum is a translational repressor that binds to the Nanos Responsive Element (NRE) sequence at the 3'UTR of its target mRNAs [18][19][20]. We demonstrated that Pum binds to potential NRE sequence found in the EGFR, Drk, Sos, and MAPK (rl) 3'UTRs and represses reporters containing these NRE sequences. This study revealed a role for Pum in formation of wing veins and bristles by negatively regulating EGFR signaling.
We explored genetic interactions between Pum and EGFR signaling. Toward this end, we examined the effect of eliminating one copy of pum + on the wing vein phenotypes associated with rl [Sem] . Eliminating one copy of pum (pum/+) by itself does not produce ectopic wing vein (data not shown). However, eliminating one copy of pum greatly enhanced the extra wing vein phenotype in rl [Sem] flies (rl [Sem] /+; pum 7 /+) (arrowhead in Figure 1F) (penetrance . 90%, n .10). Likewise, greatly increased ectopic wing veins were generated when ras activation was combined with pum knock-down (en-gal4/UAS-Ras; UAS-Pum-IR/+) (arrow in Figure 1I). Taken together these results indicate that reduction of pum synergistically enhanced EGFR signaling, consistent with the idea that pum negatively regulates EGFR signaling in wing vein formation.
If pum negatively regulates EGFR signaling, over-expression of pum should override EGFR signaling. We tested this hypothesis and found that ectopic expression of EGFR under ser-Gal4 control (active in the dorsal compartment during the 2 nd instar and the dorso-ventral boundary in the third instar) resulted in extra wing veins around the wing boundary (arrowhead in Figure 2B). Ectopic expression of Pum via ser-GAL4 resulted in development of a distal wing notch (Figure 2A). Co-overexpresson of pum with EGFR (UAS-pum/+; ser-GAL4/UAS-EGFR) ( Figure 2C) suppressed the development of ectopic veins, generating wing indistinguishable from those in which Pum alone is mis-expressed. Likewise, Co-ectopic expression of pum with rl [Sem] (UAS-Pum/UAS-rl [Sem] ; dpp disk -Gal4/+) ( Figure 2F) resulted in suppression of ectopic veins caused by rl [Sem] overexpression (UAS-rl [Sem] /+; dpp disk -Gal4/+) ( Figure 2E). Thus, both loss-and gain-of Pum function modify EGFR pathway activity in a manner that suggests negative regulation by Pum.
We examined whether pum genetically interacts with EGFR during bristle formation. Eliminating one copy of pum (pum/+) did not affect the number of bristles by itself (data not shown), but greatly enhanced bristle number induced by ectopic expression of EGFR (sca-Gal4/UAS-EGFR; pum/+) ( Figure 3G, H, N; Table 1). This result suggests that Pum negatively regulates EGFR signaling in bristle formation. To examine whether ectopic Pum can suppress EGFR signaling, we co-expressed Pum and EGFR (sca-Gal4/UAS-EGFR; UAS-Pum/+) and found that co-expression of Pum and EGFR resulted in the elimination of EGFR-induced extra bristles, phenotypes similar to Pum over-expression (Figure 3J, 3N; Table 1). These data indicate that Pum effectively down-regulate EGFR signaling in bristle formation.

Pum Activity in the Wing Disc
Although EGFR pathway component expression and activity have been well characterized in imaginal discs, Pum activity in the discs has not been well documented. By histochemical methods, we found that Pum is uniformly expressed in wing imaginal discs (not shown). To distinguish uniform expression from uniform background, we performed two additional experiments. Using the dpp-GAL4 driver that is active near the anterior-posterior (A/P) compartment boundary ( Figure 4A), we over-expressed either wild type Pum or Pum RNAi. As shown in Figure 4, ectopic Pum antigen is detected where dpp-GAL4 is active in the former experiment; conversely the antigen is specifically depleted in the latter experiment. We conclude that Pum is expressed throughout the wing disc and thus available to regulate EGFR pathway components.
We also assayed Pum activity in wing discs using a GFP reporter mRNA bearing NRE sequences in its 3'-UTR [23]. Modulating the level of Pum near the A/P compartment boundary via dpp-GAL4 regulates accumulation of GFP encoded by this reporter ( Figure 4B, C) demonstrating that Pum is active in the wing disc. Furthermore, over-expression of Nos, which is a cofactor of Pum in other tissues, negatively regulates GFP ( Figure 4D). Thus, Pum is both expressed and active in the wing disc.
Pum can Repress Translation of EGFR and rl through Binding to NRE Sequences in their 3'-UTRs The genetic analysis described above indicates that Pum downregulates EGFR signaling. Thus we searched the 3'-UTRs encoding EGFR components for potential NRE sequences (UGUAN(N)AUA, where N is any nucleotide) as a first step in determining whether regulation by Pum might be direct [24,25] ( Table 2). A genome-wide screen had previously shown that some mRNAs of EGFR pathway can be co-precipitated with Pum [24,25]. We identified two putative NRE-like sequences in each of the EGFR, Raf, and Drk 3'-UTRs, termed NRE 1 and NRE 2; Ras, Rolled (Rl), and Sos each possess a single putative NRE-like sequence. We next examined whether Pum binds directly to these putative NREs. Using a well-characterized yeast three-hybrid assay for Pum binding [26] we found that Pum bound to the EGFR NRE1, Rl NRE, Sos NRE, and Drk NRE1 sequences ( Figure 5A; Table 2). Binding was abolished by mutation of the sequence in the NRE from UGU to ACA ( Figure 5A; Table 2). To investigate whether Pum binding to the NREs of EGFR, Rl, Sos, and Drk mediated translational repression, we introduced each NRE into the 3'-UTR of the luciferase (luc) gene, using wild type and mutant hunchback (hb) NRE sequences as controls [27]. Pum repressed translation of luc containing EGFR NRE1, Rl NRE, Sos NRE, and Drk NRE1; the putative Raf NRE1, which does not bind appreciably to Pum, does not mediate repression ( Figure 5B; Table 2; data not shown). Mutant forms of each NRE that do not bind Pum in the yeast assay do not confer regulation by Pum in the Luciferase assay ( Figure 5B; Table 2). These data indicate that Pum directly binds to various NREs in the mRNAs encoding EGFR signaling components and can repress their translation.

Discussion
We have shown that, in the absence of Pum, extra bristles and wing veins develop, while over-expression of Pum eliminates bristles and wing veins. Several lines of evidence show that the role of Pum is to negatively regulate development of wing veins and bristles. First, lossand gain-of Pum function produced aberrant wing vein and bristle phenotypes that are inverse to those produced by altered EGFR signaling. Second, reduction of Pum activity greatly enhanced phenotypes associated with reduced EGFR signaling. Third, concomitant expression of Pum suppressed phenotypes associated with ectopic EGFR signaling. In support of the genetic conclusion, we show that Pum binds the NRE-like sequence of EGFR, Rl, Sos, and Drk mRNAs and represses translation of a reporter containing these sequences in heterologous cells, suggesting that Pum is a negative regulator of EGFR signaling.
To define Pum's role in the development of wing veins and bristles precisely, we attempted to locate Pum protein and measure Pum activity through a GFP-NRE construct in the 3 rd -instar larval and pupal wing imaginal discs where wing vein and SOP cells are specified. We obtained a low-level ubiquitous expression of Pum and broad Pum activity, suggesting that Pum might function as general attenuator of EGFR signaling.
Our discovery of negative regulation of EGFR signaling by Pum is not confined to Drosophila somatic cells, since it has also been reported in germline cells of C. elegans, cultured human stem cells, and yeast cells [28,29]. Thus, it is likely that Pum regulation of EGFR signaling is universal and involves diverse developmental contexts, ranging from C. elegans to Drosophila and humans.

Luciferase Reporter Assays
A full-length Pum from LD44635 was cloned into the EcoRI/ XhoI sites of pcDNA3 (Invitrogen) by PCR. The same NRE fragments used for the yeast three-hybrid assay were inserted into the BamHI/XhoI sites of pcDNA3-LUC [27]. HEK293 cells were maintained with Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal bovine serum for 24 h and transiently transfected with the appropriate set of reporter and expression plasmids using SuperFect reagent (Qiagen). For reporter assays, 24 h after transfection, cells were harvested and assayed for luciferase activity as described previously [27,42]. The results from triplicate samples were averaged and normalized to LacZ expression from pSV-b-gal (Promega). The plasmid DNAs used for transfection included the reporter plasmid having NRE wild-type or its mutant, the pSV-b-gal control plasmid, and pcDNA3-pum.

Immunostaining of the 3rd Instar Larval Wing Disc
The 3rd instar larval wing discs were dissected in PBS. The tissues were fixed for 15 minutes with gentle rocking in 4% formaldehyde in PBS. After fixation, the tissues were washed three times in PBT (PBS, 0.1% Tween-20) at RT for 15 minutes. The tissues were then blocked for 1 hour by 5% normal Goat serum in PBT. Primary antibodies were incubated overnight at 4uC. The wing discs were then washed four times with PBT for 10 minutes, and incubated for two hours with secondary antibodies, then   Figure 5A. +++, strong binding; ++, moderate binding. ***Pum repression of a reporter containing NRE-like sequence as determined in Figure 5B. Y, repression; N, no repression; ND, not determined. doi:10.1371/journal.pone.0034016.t002