(C) PLOS One [1]. This unaltered content originally appeared in journals.plosone.org.
Licensed under Creative Commons Attribution (CC BY) license.
url:https://journals.plos.org/plosone/s/licenses-and-copyright

------------



EGFRAP encodes a new negative regulator of the EGFR acting in both normal and oncogenic EGFR/Ras-driven tissue morphogenesis

['Jennifer Soler Beatty', 'Centro Andaluz De Biología Del Desarrollo', 'Universidad Pablo De Olavide Csic Ja', 'Sevilla', 'Cristina Molnar', 'Centro De Biología Molecular Severo Ochoa', 'Uam Csic', 'Univ. Autónoma De Madrid', 'Madrid', 'Carlos M. Luque']

Date: 2021-11

Activation of Ras signaling occurs in ~30% of human cancers. However, activated Ras alone is insufficient to produce malignancy. Thus, it is imperative to identify those genes cooperating with activated Ras in driving tumoral growth. In this work, we have identified a novel EGFR inhibitor, which we have named EGFRAP, for EGFR adaptor protein. Elimination of EGFRAP potentiates activated Ras-induced overgrowth in the Drosophila wing imaginal disc. We show that EGFRAP interacts physically with the phosphorylated form of EGFR via its SH2 domain. EGFRAP is expressed at high levels in regions of maximal EGFR/Ras pathway activity, such as at the presumptive wing margin. In addition, EGFRAP expression is up-regulated in conditions of oncogenic EGFR/Ras activation. Normal and oncogenic EGFR/Ras-mediated upregulation of EGRAP levels depend on the Notch pathway. We also find that elimination of EGFRAP does not affect overall organogenesis or viability. However, simultaneous downregulation of EGFRAP and its ortholog PVRAP results in defects associated with increased EGFR function. Based on these results, we propose that EGFRAP is a new negative regulator of the EGFR/Ras pathway, which, while being required redundantly for normal morphogenesis, behaves as an important modulator of EGFR/Ras-driven tissue hyperplasia. We suggest that the ability of EGFRAP to functionally inhibit the EGFR pathway in oncogenic cells results from the activation of a feedback loop leading to increase EGFRAP expression. This could act as a surveillance mechanism to prevent excessive EGFR activity and uncontrolled cell growth.

Activation of Ras signalling occurs in ~30% of human cancers. However, activated Ras alone is insufficient to produce malignancy. Thus, the discovery of genes cooperating with Ras in cancer is imperative to understand tumoral growth driven by Ras activating mutations. A key output of over-activated EGFR/Ras signalling is the induction of a complex and dynamic set of transcriptional networks leading to changes in gene expression. As a result of these changes, the normal function of some genes can become adjusted in a tumorigenic context. In this work, using the Drosophila wing imaginal disc as model system, we have identified a new EGFR inhibitor, EGFRAP, which function is redundant for proper morphogenesis, yet becomes an important limiter of the overgrowth driven by oncogenic EGFR/Ras activity. We show that the specificity of EGFRAP in cells with high levels of EGFR activity arises from activation of a negative feedback loop resulting in increased EGFRAP levels. This could act to prevent excessive EGFR activity and uncontrolled cell growth. We believe the identification of other factors behaving like EGFRAP, will help in our fight against cancer, as it might lead to the identification of new therapeutic drugs affecting cancer but not normal cells, a top priority in cancer research.

Funding: This work was funded by the by the Spanish Agencia Estatal de Investigación (MCUI/AEI, http://www.ciencia.gob.es/ ; Grant numbers BFU2016-80797-R, PID2019-109013GB-100 and MDM-2016-0687 to MDM-B; PGC2018-094476-B-100 to JFC and BFU2015-67266-R to CML) and from the Consejería de Educación e Investigación, Comunidad de Madrid ( https://www.comunidad.madrid/ ; Grant number S2010/BMD-2305 to CML) and by the European Regional Development Fund ( http://ec.europa.eu/regional_policy/en/funding/erdf/ ). Core funding to the CABD from the Junta de Andalucía is acknowledged. JSB was supported by a FPI Fellowship from MICINN. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

In this work, we have identified the gene CG33993, which we have named EGFRAP for EGFR adaptor protein, as a novel modulator of oncogenic Ras in the Drosophila wing imaginal disc. Elimination of EGFRAP, either by mutation or RNAi, enhances Ras V12 -mediated tissue hyperplasia. EGFRAP encodes a protein containing a conserved SH2 domain, which physically interacts with the active form of EGFR. EGFRAP localizes to the apical region of cells with high levels of EGFR activity, such as wing margin cells. The apical accumulation of EGFRAP in wing margin cells is regulated by the Notch pathway. In addition, we found that ectopic activation of the EGFR/Ras pathway in wing discs drives EGFRAP expression in the pouch territory. This EGFR/Ras-driven ectopic EGFRAP expression depends on Notch signaling, which is itself upregulated in response to EGFR hyperactivation. Based on these results, we propose that EGFRAP is part of a novel negative feedback loop acting as an important regulator of cell proliferation driven by excessive EGFR/Ras signaling. EGFRAP mutations are viable and don´t display EGFR-related phenotypes, suggesting that EGFRAP function is not required during normal development. However, we find that EGFRAP and its homologous gene PVRAP act redundantly in normal cells to restrain EGFR/Ras signaling. EGFRAP specificity for cells with high EGFR activity is consistent with activation of a Notch pathway-mediated negative feedback loop, where it acts to help prevent excessive EGFR activity and uncontrolled cell growth.

The wing disc also offers an effective system to study tumor progression and oncogenic cooperation (reviewed in [ 18 ]). The induction of cell mosaics over-expressing activated Ras (Ras V12 ) or EGFR (λTop) in the wing disc gives rise to hyperplastic growth [ 19 ]. This has been exploited in genetic screens to identify additional genes that can either suppress or enhance the growth of cells overexpressing Ras V12 or λTop mutations. EGFR/Ras pathway interactors include genes regulating apico-basal cell polarity, RhoGEF, genes encoding lysosomal proteins, proteins involved in mitochondrial respiratory function and several microRNAs (reviewed in [ 20 ]). Intriguingly, one of the genes found to cooperate with oncogenic Ras, both in Drosophila and in human cells, is EGFR itself [ 21 – 24 ]. However, in this context, it is noteworthy that none of the factors regulating EGFR during normal development have yet been identified as oncogenic EGFR/Ras cooperating genes.

The Drosophila wing is a very convenient system to analyze EGFR signaling, because of its simplicity and our ability to alter and visualize pathway activity in situ with relative ease. The wing develops from a relatively flat epithelial sac, known as the wing imaginal disc, which grows by cell proliferation during larval stages and differentiates during pupal development into the fly thorax and wing. The EGFR pathway is essential for development of the wing imaginal disc, acting to promote the early proximo-distal patterning of the disc and the formation of wing veins and sensory organs at later stages [ 13 – 16 ]. In this and other developmental contexts, EGFR signaling is tightly regulated by a carefully orchestrated spatial and temporal distribution of activating and inhibitory factors (reviewed in [ 17 ]).

The EGFR/Ras signaling pathway has been systematically studied during Drosophila embryonic and imaginal disc development, where it participates in the regulation of cell proliferation, growth, differentiation, migration and survival (reviewed in [ 10 ]. A particular contribution of the fly model has been the identification of positive and negative regulators through genetic screens. Some of these components are transcriptionally-regulated by the pathway, generating both positive and negative feedback loops (reviewed in [ 10 ]). For example, the expression of rhomboid, a gene encoding a serine protease that cleaves the ligand Spitz, is activated by EGFR signaling to further increase pathway activity. On the other hand, proteins encoded by EGFR-activated genes kekkon, argos, sprouty and MKP3 antagonize the pathway at different levels (reviewed in [ 11 , 12 ]).

True Ras proteins (H-, N- and K-RAS in humans and DRas1 in Drosophila) connect activated tyrosine kinase receptors, such as EGFR and FGFR, to intracellular transducers. Upon ligand binding, EGFR dimerizes and is activated via trans-phosphorylation, which leads to the recruitment of signaling molecules with Src homology 2 (SH2) domains, such as Grb2 and Shc (reviewed in [ 4 ]. These proteins mediate the recruitment of the guanine-nucleotide exchange factor (GEF) Son of Sevenless (Sos) to the receptor complex, which results in Ras activation [ 5 , 6 ]. Similar to Ras, EGFR is also mutationally-activated and/or overexpressed in one-fifth of all human cancers [ 7 ]. Likewise, even though EGFR expression is altered in many epithelial tumors, cooperation with other oncogenic lesions is required for malignant transformation and invasion [ 8 , 9 ].

Cancer is a devastating disease that threatens human health worldwide [ 1 ]. One of the most commonly affected genes in cancer is the proto-oncogene Ras. In fact, mutations that elevate its activity are present in ∼30% of human cancers and give rise to some of the most aggressive tumors [ 2 ]. However, hyperactivation of Ras signaling alone is insufficient to produce malignancy (reviewed in [ 3 ]. Additional mutations in other genes are required for Ras-driven malignant tumorigenesis. Thus, identifying genes that modulate the oncogenic capacity of Ras is imperative in our fight against cancer. Of particular interest are factors important for oncogenesis but that are dispensable for normal development and homeostasis. These molecules are ideal targets for cancer therapy, as they hold the potential to prevent the growth of cancer cells while having little or no effect on normal tissues.

(A-C’) Maximal projection of confocal images of third instar wing imaginal discs expressing the indicated UAS transgenes under the control of en-Gal4, stained with anti-RFP (red) and anti-pERK (green). (D) Box plot of the posterior/anterior pERK intensity ratio of wing discs of the designated genotypes. (E, F) Adult flies expressing the indicated transgenes under the control of ap-Gal4. Insets in E and F show confocal images of third instar wing imaginal discs expressing GFP and EGFRAP-GFP, respectively, under the control of ap-Gal4, stained with anti-GFP (green) and anti-EGFRAP (red). (G) Maximal projection of confocal images of an EGFR-sfGFP third instar wing imaginal disc expressing EGFRAP under the control of Hh-Gal4 stained with anti-GFP (green), anti-EGFRAP (red) and Hoechst (DNA, blue). (G’-G”) Confocal yz sections, along the white dotted lines on the anterior (A) and posterior regions (P) of the wing disc shown in G. Apical side of the wing disc is to the left. The statistical significance of differences was assessed with a t-test, **** and ** P values <0.0001 and <0.01, respectively. Scale bars, 50 μm (A-C’, G).

To further test whether EGFRAP acts as a negative regulator of the EGFR/Ras pathway, we altered EGFRAP expression levels and monitored the effect on pathway activity by visualizing pERK accumulation. We found that reducing (en>EGFRAP RNAi ) EGFRAP expression levels in the posterior compartment of wing discs, increased pERK levels ( Fig 7A–7B’ and 7D ). To increase EGFRAP levels, we used a transgenic line expressing a GFP-tagged version of the protein (UAS-EGFRAP-GFP; see Materials and Methods ). In this case, we found that by increasing EGFRAP expression levels (en>RFP; EGFRAP-GFP), pERK expression was reduced ( Fig 7C and 7D ). We also found that EGFRAP overexpression in the dorsal region of the wing disc resulted in the loss of notum and wing tissue, a phenotype similar to that observed upon reduction of EGFR/Ras activity ( Fig 7E and 7F ) [ 48 ]. Other SH2 domain-containing negative regulators of EGFR act by promoting the internalization and degradation of the receptor. However, we found that EGFRAP overexpression did not affect the levels or localization of EGFR-sfGFP ( Fig 7G ). These results suggest that the phenotypes associated with reduced EGFR signaling mediated by EGFRAP overexpression are not a consequence of defects in receptor trafficking.

EGFR is also expressed and required for the proliferation, survival and differentiation of most cell types in the Drosophila eye disc, including all photoreceptors (R1-R7), cone and pigment cells. Accordingly, we tested whether EGFRAP was also expressed in the eye disc. Interestingly, we found that EGFRAP was only expressed in a subset of the cells requiring EGFR activity ( S7A–S7B’ Fig ). More specifically, using the Hedgehog-LacZ marker (Hh-LZ,), which is expressed in photoreceptors R2-R5 ( S7C and S7C’ Fig , [ 46 ]), and an anti-Senseless (Sens) antibody, which labels R8 ( S7D and S7D’ Fig , [ 47 ]), we could determine that the photoreceptor expressing high EGFRAP levels was R7. In addition, ectopic expression of λtop in eye disc clones (hs>GFP; λtop) induced ectopic expression of EGFRAP ( S7E and S7E’ Fig ). Thus, EGFRAP expression and its relationship with EGFR during Drosophila eye development appears analogous to the situation in wing imaginal discs.

The EGFRAP expression pattern resembles that of Senseless, a transcription factor regulated by wingless signaling, downstream of the Notch pathway [ 44 , 45 ]. Thus, we tested if EGFRAP expression was regulated by Notch. We found that RNAi knockdown targeting Notch (hh>GFP; N RNAi , Fig 6I and 6J ) reduced EGFRAP levels. Conversely, ectopic expression of an active intracellular fragment of the Notch receptor (ap>GFP; N i , Fig 6K ) increased EGFRAP expression in the wing pouch ( Fig 6K ). Furthermore, we found that downregulation of Delta was sufficient to rescue the increase in EGFRAP levels due to EGFR pathway overactivation ( Fig 6L ).

The expression of many modulators of the EGFR/Ras pathway in the wing pouch is itself regulated by the EGFR/Ras pathway. To test whether this was also the case for EGFRAP, we analyzed its expression in wing discs expressing either activated (λtop, [ 41 ]) or dominant negative (EGFR DN , [ 42 ]) forms of EGFR in the dorsal compartment. We found that ectopic expression of λtop induced ectopic EGFRAP expression in the wing pouch ( Fig 6E–6F” ). However, we observed that downregulation of EGFR did not affect EGFRAP expression ( Fig 6H ). Note that in this case we used hedgehog-Gal4, which drives expression in the posterior compartment [ 43 ]. The specific expression of EGFRAP in cells flanking the D/V boundary and unresponsiveness to decreased EGFR activity suggest that other factors also regulate EGFRAP expression in the wing pouch.

(A) In situ hybridization of a third instar wing imaginal disc with a probe for EGFRAP mRNA. (B) Maximal projection of confocal images of a wild-type (wt) third instar wing imaginal disc stained with anti-EGFRAP. (C) Quantification of EGFRAP levels in the region of the white dotted line in (B). (D) Maximal projection of confocal images of an EGFR-sfGFP third instar wing imaginal disc stained with anti-EGFRAP (red) and anti-GFP (green). (D’) Confocal yz section, along the white dotted line in D. (E, F) Maximal projection of confocal views of wing discs expressing the indicated UAS transgenes under the control of ap-Gal4 stained with anti-GFP (green), anti-EGFRAP (red) and Hoechst (DNA, blue). (E’-E” and F’-F”) Confocal yz sections along the white dotted lines shown in E and F, respectively. Apical side of wing discs is to the left. White arrows indicate wing margin cells (wm). (G, H, I, K, L) Maximal projection of confocal images of wing discs expressing GFP and the specified UAS transgenes under the control of hedgehog-Gal4 (hh-Gal4) (G-I) and ap-Gal4 (K, L), stained with anti-EGFRAP (red), anti-GFP (green) and Hoechst (DNA, blue). (J) Box plot of the posterior/anterior EGFRAP intensity ratio of wing discs of the designated genotypes. (I’, L’) Confocal yz sections along the white dotted lines are shown in I and L, respectively. Apical side of wing discs is to the left. The statistical significance of differences was assessed with a t-test, *** P value<0.001. wm (wing margin). Scale bars, 60 μm (A, B) and 50 μm (D, E, F, G, H, I, K, L).

The absence of a loss-of-function phenotype for EGFRAP could be due to compensation by other adaptor proteins performing similar functions. We noticed that adjacent to EGFRAP lies the gene PVRAP, which also encodes an SH2 domain. Furthermore, the amino acid sequences of their SH2 domains share 74% identity, versus 26%-35% identity for pairwise comparisons of EGFRAP with SH2 domains found in other Drosophila proteins. We found that RNAi knockdown of PVRAP in wing imaginal discs enhanced the Ras V12 overgrowth and folding phenotype ( Fig 5A–5C’ ), in a similar way to EGFRAP. Furthermore, downregulation of PVRAP enhanced the effects of eliminating EGFRAP in Ras V12 expressing wing discs (ap>GFP; Ras V12 ; PVRAP RNAi ; EGFRAP L/S ; Fig 5D and 5D’ ). In addition, while downregulation of either EGFRAP (ap>EGFRAP RNAi ) or PVRAP (ap>PVRAP RNAi ) on their own did not cause any visible phenotype in the adult ( Fig 5E–5G and 5I–5K ), the simultaneous downregulation of both genes (ap>EGFRAP RNAi ; PVRAP RNAi ) resulted in ectopic sensory organs in the notum ( Fig 5H ), a phenotype that resembles increased EGFR activity [ 14 ], and blisters in the wing ( Fig 5L ).

All EGFRAP mutant alleles were homozygous viable and displayed no obvious morphological abnormalities, indicating that EGFRAP is dispensable for viability in Drosophila. To confirm the role of EGFRAP as a modulator of Ras V12 -mediated hyperplasia, we tested for synergetic interactions between EGFRAP mutations and Ras V12 in wing imaginal discs ( Fig 4 and S5 Fig ). We found that expression of Ras V12 in the posterior compartment of EGFRAP mutant discs ( Fig 4D–4D” and S5B–S5C” Fig ) resulted in an enhanced folding phenotype, similar to that observed in Ras V12 ; EGFRAP RNAi discs ( Fig 1D and 1D’ ). The strong Ras V12 enhancement by EGFRAP L and EGFRAP ΔSH2 alleles suggests that the function of EGFRAP as a modulator of Ras V12 -mediated hyperplasia is performed mainly by the long isoform and that the SH2 domain is crucial for this function.

(A) Schematic representation of the EGFRAP locus (3 rd chromosome), EGFRAP mutants generated, sgRNAs used for generation of mutants (green boxes 1–3) and sequence targeted by EGFRAP RNAi construct (purple box). (B-D) Maximal projection of confocal images of wing imaginal discs from third-instar larvae of the indicated genotypes stained with anti-GFP (green), RhPh (red) and Hoechst (DNA, blue). (B’-B”, C’-C” and D’-D”) Confocal sections of wing discs of the specified genotypes along the white dotted lines shown in B, C and D, respectively, parallel (B’, C’ and D’) or perpendicular (B”, C” and D”) to the A/P border. Apical sides of wing discs are to the left (B’, C’ and D’) or to the top (B”, C” and D”). Scale bars, 60 μm (B-D).

To further characterize the role of EGFRAP as a modulator of Ras V12 -mediated hyperplasia, we used CRISPR/Cas9 to generate specific EGFRAP alleles (see Materials and Methods ). The EGFRAP gene encodes two isoforms, a full-length long isoform (EGFRAP-RA) and a short isoform (EGFRAP-RB), whose transcription start site maps to the beginning of exon 3 ( Fig 4A ). We generated two EGFRAP mutant alleles, in which both isoforms (EGFRAP L/S ) or only the long isoform (EGFRAP L ) were truncated ( Fig 4A ). The EGFRAP L/S allele truncates 90.7% of the short isoform and 67% of the long isoform ( Fig 4A ). The EGFRAP L allele truncates 78.2% of the long isoform ( Fig 4A ). In addition, as exon 5 of EGFRAP encodes a conserved SH2 domain known to interact with phosphorylated tyrosines [ 36 ], we also generated a mutant allele whose SH2 domain was completely eliminated (EGFRAP ΔSH2 ; Fig 4A ).

Although ectopic Ras V12 expression in wing disc cells alone does not affect cell polarity ([ 34 ]; S4A–S4B” and S4D–S4E” Fig ), the removal of polarity genes enhances the hyperplastic phenotype of Ras V12 [ 35 ]. Thus, we tested whether cell polarity was affected in Ras V12 ; EGFRAP RNAi disc cells. We found that downregulation of EGFRAP did not alter the polarity of Ras V12 cells ( S4C–S4C” and S4F–S4F” Fig ).

(A-G) Maximal projection of confocal views of wing imaginal discs from third-instar larvae expressing GFP (green) and the indicated UAS transgenes under the control of ap-Gal4, stained with anti-GFP (green), anti-Dcp-1 (red in A-C, white in A’-C’), anti-pJNK (red in E-G) and Hoechst (DNA, blue). (D, H) Box plots of mean fluorescent Dcp-1 (D) and pJNK (H) intensities of wing discs of the designated genotypes. (I-K) Confocal yz sections parallel to the A/P axis of wing imaginal discs from third-instar larvae expressing GFP (green) and the indicated UAS transgenes under the control of ap-Gal4, stained with anti-GFP (green), anti-Dlg (red) and Hoechst (DNA, blue). Apical side of wing discs is to the top. Arrows in I-K point to the border between dorsal experimental and ventral control cells. The statistical significance of differences was assessed with a t-test, **** and *** P values <0.0001 and <0.001, respectively. Scale bars, 50 μm (A-G) and 30 μm (I-K).

Occasionally, some GFP positive cells were found outside of the dorsal domain in both ap>GFP; Ras V12 and ap>GFP; Ras V12 ; EGFRAP RNAi expressing discs, with the latter having the stronger phenotype ( S3A–S3C’ and S3E Fig ). Invading cells were always found basally and most of them were positive for the apoptotic marker, cleaved Dcp-1 (white arrow in S3D–S3D” Fig ). Effector caspases are active in tumors and are associated with metastasis [ 28 ]. Furthermore, caspase activity has been proposed to direct the migration of transformed cells in wing imaginal discs [ 29 ]. In this context, we found that co-expression of Diap1, an inhibitor of apoptosis, blocked the invasiveness of both Ras V12 and Ras V12 ; EGFRAP RNAi cells ( S3E Fig ). These results suggest that EGFRAP downregulation might cooperate with Ras V12 to promote tumor invasion associated with effector caspase activity.

(A-C) Maximal projection of confocal images of wing imaginal discs from third-instar larvae expressing myristoylated-Tomato (MyrT, red) and the indicated UAS transgenes under the control of ap-Gal4, stained with anti-Tomato (red) and the nuclear marker Hoechst (DNA, blue). (A’-C’) Confocal xz sections along the white dotted lines of wing discs shown in A-C. The apical side of wing discs is at the top. Brackets indicate cell height. (D-F) Apical and (G-I) basal surface views of the indicated genotypes. (J-L) Box plots of the cell height (J), apical cell area (K) and basal cell area (L) of the indicated genotypes. The statistical significance of differences was assessed with a t-test, ****P value<0.0001. Scale bars, 50 μm (A-C), 30 μm (A’-C’) and 10 μm (D-I).

Ectopic expression of activated Ras (Ras V12 ) in Drosophila wing imaginal discs produces hyperplasia due to increased cell growth, accelerated G1-S transition and cell shape changes [ 25 , 26 ]. To isolate new modulators of Ras V12 activity, we used RNAi to knockdown a battery of candidate genes and search for knockdowns that enhanced the Ras V12 phenotype. Ectopic expression of Ras V12 in the dorsal compartment of wing imaginal discs, by means of apterous-Gal4 (ap>GFP; Ras V12 ), or in a discrete stripe along the A/P boundary, using a tub-Gal80 ts ; patched-Gal4 combination (ptc 80ts > GFP; Ras V12 ), induces overgrowth of the tissue and the formation of ectopic folds [ 26 ] ( Fig 1A–1B’ , S1A–S1B’ and S1D Fig ). Using this model to screen for enhanced Ras hyperplasia, we identified the gene EGFRAP. Although EGFRAP RNAi had no detectable effect in wild-type cells (ap>GFP; EGFRAP RNAi , Fig 1C and 1C’ ), it enhanced the phenotype of Ras V12 cells (ap>GFP; Ras V12 ; EGFRAP RNAi , Fig 1D and 1D’ and ptc 80ts >GFP; Ras V12 ; EGFRAP RNAi , S1C , S1 C’ and S1D Fig ).

Discussion

EGFR signaling plays important roles during multicellular development and tissue homeostasis, affecting cell division and growth, cell fate choices, cell viability and organ morphogenesis. During morphogenesis or in normal differentiated tissues, cells are not persistently exposed to high concentrations of activating epidermal growth factors. However, over recent years it has been found that this scenario can change in conditions of pathway hyperactivation, where cells have to impose more rigorous control mechanisms to avoid the consequences of aberrant signaling (reviewed in [50]). Here, we have identified a novel EGFR inhibitor, EGFRAP, which exhibits a limited capacity to act as a negative regulator of EGFR during Drosophila wing morphogenesis, most likely due to redundancy with the SH2 domain containing protein PVRAP, yet behaves as an important regulator of the overgrowth driven by excessive activation of the EGFR/Ras pathway (Fig 8). We propose that the specificity of EGFRAP in cells with high levels of EGFR activity arises from activation of a negative feedback loop, which reduces excessive EGFR activity.

PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 8. Model of EGFRAP function as a modulator of EGFR/Ras-dependent tissue hyperplasia. Schematic drawing depicting the mechanisms by which EGFRAP could limit EGFR/Ras activity in normal (A) and RasV12-dependent oncogenic cells (B). (A) In normal cells EGFRAP and PVRAP act as negative regulator of the EGFR. Expression of EGRAP is confined to cells with high levels of EGFR activity via the Notch pathway. EGFRAP elimination results in a slight increase in EGFR activity, which does not seem to affect normal morphogenesis. However, the simultaneous elimination of EGFRAP and PVRAP leads to a further enhancement of EGFR signaling with consequences in normal morphogenesis. (B) Oncogenic EGFR/Ras activity promotes, via activation of the Notch pathway, an increase in EGFRAP expression that, in turn, restrains both EGFR activity and its capacity to induce hyperplasia. Downregulation of EGFRAP releases this restraint leading to a further increase in EGFR/Ras pathway activity and tumor growth, which is enhanced by elimination of PVRAP. https://doi.org/10.1371/journal.pgen.1009738.g008

The function of EGFRAP during morphogenesis and oncogenic transformation The EGFR signaling pathway is a network with highly redundant and overlapping input signals and feedback controls (reviewed in [50]. This redundancy enhances robustness of the pathway, as dysfunction of a given regulator of the pathway can be compensated by others. The identification of EGFRAP, an SH2-domain containing protein that acts as a novel EGFR negative regulator, adds to the regulatory tool-box of the pathway. Elimination of EGFRAP in normal cells marginally disrupts EGFR pathway activity, without affecting overall morphogenesis or viability. However, downregulation of EGFRAP in RasV12-expressing cells increases their oncogenic phenotype, suggesting that EGFRAP is needed to control high EGFR pathway activity. In this context, we report that next to EGFRAP lies PVRAP, a gene coding for another SH2 domain-containing protein, which we show that, similar to PVRAP, controls RasV12-mediated tissue hyperplasia without affecting normal morphogenesis. PVRAP has been shown to physically interact with PVR [51], but based on the degree of homology of its SH2 domain with that of EGFRAP, we surmise it could also interact with EGFR. We therefore propose that EGFRAP and PVRAP might constitute a novel SH2 domain containing family of proteins that could function redundantly to regulate EGFR activity. In fact, the simultaneous downregulation of EGFRAP and PVRAP in otherwise normal cells leads to defects consistent with EGFR hyperactivation, further emphasizing their regulatory role in the pathway. Additional biochemical, genetic and functional characterization of PVRAP will be required to confirm these findings and to gain a deeper understanding of how PVRAP interacts with EGFR in normal and pathological conditions. Although, EGFRAP and PVRAP seem to work redundantly in normal cells, each of them can individually restrain EGFR/Ras signaling in the context of pathway hyperactivation. It is known that the main output of over-activated EGFR/Ras signaling is the induction of a complex and dynamic set of transcriptional networks, leading to changes in the gene expression signature and metabolic state of cancer cells [52]. As a result of these changes, new genetic interactions can arise in tumor cells [53]. In this context, genes with non-essential or redundant functions in normal cells can become highly relevant in tumorigenic conditions. For example, expression of the copper/zinc dismutase SOD1 is dispensable for normal mammary gland development, yet is essential for the survival of breast cancer cells. SOD1 is specifically required in cancer cells to counter the increased superoxide production associated with oncogene activation [54]. Similarly, the MTH1 protein, which sanitizes oxidized dNTP pools, is required for cancer cell survival, but is dispensable for untransformed cells viability. The requirement of MTH1 in cancer cells arises from their need to adapt to an imbalanced redox state not seen in normal cells [55]. Drosophila Socs36E and its human ortholog, SOCS5, have a limited capacity to act as growth regulators under normal conditions but have proved to be important regulators of EGFR-dependent cellular transformation [56–58]. Likewise, FAK moderates RTK signaling only in situations of pathway hyper-activation [59]. Interestingly, in all these cases, oncogene activation promotes increased expression of the otherwise non-essential genes by mechanisms that remain poorly understood. Similarly, we show here that EGFRAP expression is upregulated by EGFR overactivation. Based on the above and on our results, we propose that the ability of EGFRAP to functionally inhibit EGFR/Ras activity in RasV12-expressing cells might arise from its increased levels of expression in these oncogenic cells. The fact that overexpression of EGFRAP in normal cells results in phenotypes associated with EGFR loss-of-function supports this view. Finally, we find that the Notch pathway, which is itself upregulated by hyperactivated EGFR/Ras signaling [60], is required for the increase in EGFRAP expression due to EGFR/Ras hyperactivation. Thus, we propose a model by which cells minimize the effect of excessive EGFR/Ras signaling by inducing the expression of negative regulators of the EGFR/Ras pathway, such as EGFRAP, via Notch activation. EGFR autocrine loop is required for transformation by activated Ras in different mammalian cell systems, including fibroblasts, keratinocytes, intestinal epithelial cells, melanomas and pancreatic cells [21–23]. The contribution of EGFR to oncogenic transformation seems to be evolutionary conserved, as it has also been found in Drosophila [30] and in Xiphophorus fish [61]. Furthermore, EGFR stimulates Ras-dependent tumor overgrowth through canonical EGFR signaling. Thus, even though oncogenic Ras isoforms are constitutively active, EGFR activation can induce GTP loading on wild type Ras isoforms, enhancing effector pathway signaling [62]. In this context, recent studies in mice indicate the existence of an optimal level of Ras activation, the so-called “sweet spot”, conducive to tumor formation and progression (reviewed in [63]. Thus, any negative regulator of the EGFR, such as EGFRAP and PVRAP, could act as tumor suppressors of the overgrowth of wild type cells ectopically expressing RasV12. Elimination of EGFRAP and/or PVRAP in these cells would lead to hyperactivation of the EGFR, which, in turn, would result in an enhancement of endogenous wild-type Ras signaling, thus increasing pathway activity and overgrowth. Alternatively, or in addition, the increase in EGFR activity due to elimination of EGFRAP and/or PVRAP could induce further overgrowth of RasV12 cells by stimulating non-canonical EGFR pathways. In fact, the EGFR stimulates the growth of fly and human cancer cells harboring oncogenic Ras mutations through the activation of pathways controlling cell growth and survival, such as the Hedgehog pathway [24].

EGFRAP enhances RasV12-dependent tissue hyperplasia Here, we find that the most prevalent change in wing disc morphology caused by RasV12 overexpression is the formation of extra folds, a phenotype that increases upon the removal of EGFRAP. We also show that the appearance of ectopic folds is associated with changes in cell shape, from columnar to cuboidal, and to a rise in cell volume. Increased EGFR signaling has been shown to affect cell shape through the regulation of Myosin II dynamics and FAK inactivation [64,65]. Thus, the increased cell shape changes observed in RasV12 cells upon EGFRAP down-regulation could be a direct consequence of the role of EGFRAP as a negative regulator of EGFR activity. The EGFRAP SH2 domain most closely resembles those of the Tensins (FlyBase). Tensins are a family of focal adhesion proteins, composed of four members (Tensin 1–4, TNS1-4), which link the cell membrane to the actin cytoskeleton and are lost in most cancer cell lines [66,67]. EGFRAP and PVRAP have been proposed to be orthologs of human TNS2 and TNS4 (FlyBase), proteins that in knockdown conditions increase tumorigenicity in several cancer lines [68]. In addition, TNS4 levels increase following EGF stimulation [69], in the same way as we observed for EGFRAP. Furthermore, TNS4 regulates cell shape downstream of EGFR, via interaction through their SH2 domains [69]. Finally, the role of Tensins as tumor suppressors has also been linked to their ability to bind and regulate integrins (reviewed in [66]. In this work, we observed that downregulation of both EGFRAP and PVRAP results in wing blisters, a phenotype associated with the loss of integrin function [70]. In this context, we propose that EGFRAP, and possibly PVRAP, could regulate cell shape changes downstream of EGFR hyperactivation in a similar way to the mammalian tensins.

[END]

[1] Url: https://journals.plos.org/plosgenetics/article?id=10.1371/journal.pgen.1009738

(C) Plos One. "Accelerating the publication of peer-reviewed science."
Licensed under Creative Commons Attribution (CC BY 4.0)
URL: https://creativecommons.org/licenses/by/4.0/


via Magical.Fish Gopher News Feeds:
gopher://magical.fish/1/feeds/news/plosone/

You are viewing proxied material from magical.fish. The copyright of proxied material belongs to its original authors. Any comments or complaints in relation to proxied material should be directed to the original authors of the content concerned. Please see the disclaimer for more details.