Lack of Evidence that CYTH2/ARNO Functions as a Direct Intracellular EGFR Activator

The EGFR instructs complex cellular programs by generating robust signals of defined strength and duration. To preserve signal identity and guard cells against the oncogenic risk generated by unabated signaling, it is imperative that EGFR be tightly regulated. In this framework, the ligand-dependent transition of EGFR from an inactive state to a catalytically competent one is regarded as a key control step. In absence of ligand, the EGFR adopts a default auto-inhibited conformation that is relieved by EGF binding and attendant stabilization of EGFR dimers. These structural transitions result in allosteric activation of the EGFR tyrosine kinase domain (TKD) through formation of asymmetric dimers between juxtaposed TKDs. Downstream signaling is then initiated via in trans phosphorylation of Tyr residues located in the COOH tail of dimerized receptors (Lemmon et al., 2014). Hence, EGFR activation is essentially driven by ligand: receptor binding kinetics and conformational changes caused by receptor dimerization. A study published in Cell (Bill et al., 2010) expanded this model, suggesting a novel layer of intracellular EGFR regulation implemented by cytohesins. The cytohesin (CYTH) family includes four members, which are well known guanyl nucleotide exchange factors (GEF) for ARF GTPases (Casanova, 2007). Through work mainly focused on ARNO, i.e., CYTH2, Bill et al. proposed that cytohesins facilitate allosteric activation of the EGFR tyrosine kinase domain by stabilizing an optimal conformation of asymmetric TKD dimers. At the cell biological level, the Bill et al. model posited that the intracellular concentration of cytohesins was a relevant determinant of cellular responsiveness to EGFR ligands. The above model was extended to envisioning cytohesins as oncogenic modifiers of EGFR and potential therapeutic targets in tumors addicted to oncogenic EGFR signaling (Bill et al., 2010). Allosteric activation of the EGFR kinase is inhibited by activity-dependent binding of MIG6 (via its EBR, i.e., ErbB binding region) to an extended surface of the TKD that includes the peptide substrate-binding pocket and C-lobe distal region (Park et al., 2015; Zhang et al., 2007). Because of their functional antagonism, we hypothesized that MIG6 and cytohesins could bind the EGFR TKD competitively. While pursuing this hypothesis, we did not observe significant regulation of EGFR activity by ARNO. Bill et al. showed that ectopic expression of ARNO in H460 lung cancer cells increased EGFR autophosphorylation in a dose-dependent manner. While the Sec7 domain (carrying ARNO’s GEF activity) was necessary and sufficient to upregulate EGFR catalytic activation, GEF activity per se was not required, because the GEF-dead E156K mutant was as efficient as wtARNO at increasing EGFR autophosphorylation (Bill et al., 2010). In contrast, we found that graded overexpression of MYC-tagged ARNO in HeLa cells did not increase EGFR autophosphorylation in experiments that tested a range of EGF doses over different times of cell stimulation. MYC-ARNO E156K was likewise ineffective at enhancing EGFR autophosphorylation (Figure 1A, 1B, S1A, and S1B). However, expression of MYC-MIG6 at levels comparable to those of MYC-ARNO induced robust suppression of EGFR autophosphorylation (Figure 1B), as reported (Anastasi et al., 2007). Similar results were obtained in NR6-EGFR cells (Figures S1C– S1F). Finally, also in cells stimulated with EGF at 4 C, i.e., in absence of endocytosis, MYC-ARNO did not increase EGFR autophosphorylation, whereas MYC-MIG6 readily suppressed EGFR activation under these conditions (data not shown). ARNO was proposed to enhance EGFR activation through its ability to bind to activated receptor dimers (Bill et al., 2010), based on co-localization of