A function-structure model for NGF-activated TRK
1998; Springer Nature; Volume: 17; Issue: 24 Linguagem: Inglês
10.1093/emboj/17.24.7282
ISSN1460-2075
Autores Tópico(s)Protein Tyrosine Phosphatases
ResumoArticle15 December 1998free access A function–structure model for NGF-activated TRK Matthew E. Cunningham Matthew E. Cunningham Department of Pathology and Center for Neurobiology and Behavior, College of Physicians and Surgeons, Columbia University, New York, NY, 10032 USA Search for more papers by this author Lloyd A. Greene Corresponding Author Lloyd A. Greene Department of Pathology and Center for Neurobiology and Behavior, College of Physicians and Surgeons, Columbia University, New York, NY, 10032 USA Search for more papers by this author Matthew E. Cunningham Matthew E. Cunningham Department of Pathology and Center for Neurobiology and Behavior, College of Physicians and Surgeons, Columbia University, New York, NY, 10032 USA Search for more papers by this author Lloyd A. Greene Corresponding Author Lloyd A. Greene Department of Pathology and Center for Neurobiology and Behavior, College of Physicians and Surgeons, Columbia University, New York, NY, 10032 USA Search for more papers by this author Author Information Matthew E. Cunningham1 and Lloyd A. Greene 1 1Department of Pathology and Center for Neurobiology and Behavior, College of Physicians and Surgeons, Columbia University, New York, NY, 10032 USA *Corresponding author. E-mail: [email protected] The EMBO Journal (1998)17:7282-7293https://doi.org/10.1093/emboj/17.24.7282 PDFDownload PDF of article text and main figures. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InMendeleyWechatReddit Figures & Info Mechanisms regulating transit of receptor tyrosine kinases (RTKs) from inactive to active states are incompletely described, but require autophosphorylation of tyrosine(s) within a kinase domain 'activation loop'. Here, we employ functional biological assays with mutated TRK receptors to assess a 'switch' model for RTK activation. In this model: (i) ligand binding stimulates activation loop tyrosine phosphorylation; (ii) these phosphotyrosines form specific charge pairs with nearby basic residues; and (iii) the charge pairs stabilize a functionally active conformation in which the activation loop is restrained from blocking access to the kinase catalytic core. Our findings both support this model and identify residues that form specific charge pairs with each of the three TRK activation loop phosphotyrosines. Introduction Intensive research over the past several decades has revealed that receptor tyrosine kinases (RTKs) respond to ligand occupancy through sequential steps of oligo- or dimerization, rapid autophosphorylation of the 1–3 tyrosine residues that lie within the 'activation loop' of the kinase domain, full activation of the kinase, continued autophosphorylation at sites outside of the kinase domain and, finally, association with and phosphorylation of intracellular targets that propagate the signal (Ullrich and Schlessinger, 1990; Schlessinger and Ullrich, 1992; McDonald et al., 1995). A key question about this process is how occupancy converts receptors from an 'off' state to a fully functional active state. Structure–function studies with a variety of RTKs indicate that ligand-stimulated autophosphorylation of kinase domain 'activation loop' tyrosines is an early obligatory step in ligand-mediated kinase activation (Rosen et al., 1983; Herrera and Rosen, 1986; Tornqvist and Avruch, 1988; White et al., 1988; Li et al., 1994; Longati et al., 1994; Hernández-Sánchez et al., 1995; McCarty and Feinstein, 1998). However, the molecular mechanisms by which such autophosphorylation leads to receptor activation are only beginning to be understood. A crystallographic study of the non-phosphorylated insulin receptor kinase domain (IRK) indicated that in this state, the activation loop Tyr1162 acts as a pseudosubstrate that blocks access of substrate tyrosines and of ATP to the catalytic core (Hubbard et al., 1994). It was suggested, in accord with previous reports of conformational changes in insulin receptor activation (Baron et al., 1992; Kaliman et al., 1993; reviewed in Van Obberghen et al., 1993) that ligand-stimulated autophosphorylation of activation loop tyrosines results in loss of this inhibitory interaction which, in turn, promotes maximal kinase activation. We have utilized the TRK receptor of nerve growth factor (NGF) as a model for RTK activation. In a neuronal environment, TRK mediates NGF responses including long- and short-term gene regulation, neuronal differentiation and survival in serum-free medium (reviewed in Greene, 1984; Green et al., 1986; Loeb et al., 1991; Loeb and Greene, 1993; Greene et al., 1998). In vitro assessment has identified TRK tyrosine residues 490, 670, 674, 675 and 785 as autophosphorylation sites (Stephens et al., 1994). The first and last of these are required for binding and tyrosine phosphorylation of signaling molecules such as Shc and phospholipase Cγ-1 (PLCγ-1) (Loeb et al., 1994; Stephens et al., 1994); the other three tyrosines lie within the kinase activation loop and are involved in TRK activation (Segal et al., 1996; Cunningham et al., 1997). In a biologically based study of TRK activation loop tyrosines, we found that mutation of either Tyr670, Tyr674 or Tyr675 significantly impairs NGF-inducible responses including promotion of neurite outgrowth and tyrosine phosphorylation of Shc/ERKs and of PLCγ-1 (Cunningham et al., 1997). Although neuritogenic capacity and Shc/ERK signaling were substantially rescued by high overexpression of mutant receptors, loss of PLCγ-1 phosphorylation was not. The latter correlated with loss of TRK autophosphorylation at its C-terminal binding site for PLCγ-1, Tyr785. The observation that all three activation loop tyrosines participate in TRK activation led us to suggest that ligand binding and consequent phosphorylation of activation loop tyrosines leads not only to removal of an inhibitory interaction, but also to stabilization of the TRK intracellular domain in a conformation that promotes kinase activity and ability to access Tyr785. We further hypothesized that this functionally active conformation would be stabilized, at least in part, by creation of charge pairs between the activation loop phosphotyrosines and nearby positively charged residues. Our hypothesis is supported by a recent crystallographic study of activated IRK which indicated that two of the three phosphorylated activation loop tyrosines each interacts pairwise with basically charged residues, and that activation is accompanied by a large conformational change (Hubbard, 1997). The purposes of the present study were several-fold. The first was to devise a suitable biological test for the concept that charge pairs formed between phosphorylated activation loop tyrosines and nearby positively charged residues are required for ligand-mediated functional activation of RTKs. The second was, for the case of the TRK receptor, to identify the residues involved in such interactions. We also sought to exploit our biological screens of function to identify additional charged residues that might be important for stabilizing the active TRK conformation. Results Functional bioassays for TRK activation loop mutants To study the role of the TRK activation loop tyrosines in receptor function, we mutated these sites (Tyr670, Tyr674 and Tyr675), individually and in all combinations, to phenylalanines and inserted the wild-type and mutant TRKs into recombinant retroviruses. The latter were used to infect PC12nnr5 cells which lack endogenous TRK. Polyclonal cultures of infected cells were generated and assessed for NGF-promoted neuritogenesis. Results were similar to our previous report of NGF-induced neuritogenesis in multiple clonally derived PC12nnr5 lines with varying levels of wild-type and mutant receptor expression (Cunningham et al., 1997); singly mutated receptors Y670F and Y674F retained the highest residual function, and the Y675F mutant and all double combination mutants exhibited little or no neuritogenic activity (Figure 1A). This behavior of the polyclonal cultures is similar to that of clonal lines expressing the respective TRKs at levels at least 20-fold that found in PC12 cells (Cunningham et al., 1997). Western immunoblotting confirmed this level of expression in the polyclonal cultures (not shown). Figure 1.NGF-promoted mediation of NIH 3T3 cell transformation and polyclonal PC12nnr5 cell differentiation by wild-type and activation loop tyrosine mutant TRKs. (A) PC12nnr5 cells were stably transfected with the indicated TRKs, were grown as polyclonal cultures and were assessed for NGF-induced neuritogenesis. Reported values were assessed at day 7 of NGF treatment and are averages ± SEM. Comparable results were obtained in a duplicate experiment. (B) Recombinant LNC retroviruses containing wild-type or the indicated mutant TRKs were used to infect NIH 3T3 'N' cells which were expanded and subjected to NGF, G418 or no treatment as described in Materials and methods. After 7–10 days of treatment, transformed colonies of cells were apparent under low magnification phase-contrast microscopy. Transformation and viral titers were assessed under dark field microscopy, and efficiencies were calculated as described in Materials and methods. Reported values are averages ± SEM from triplicate, independent experiments. No transformation was observed in the absence of NGF here or in subsequent experiments. Download figure Download PowerPoint To extend and complement the neuritogenic assay, we also delivered the wild-type and mutant TRKs into NIH 3T3 cells. Fibroblasts transfected with TRK (Cordon-Cardo et al., 1991) respond to NGF by transformation, focus formation and anchorage-independent growth. The infected NIH 3T3 cells were subcultured and subjected to either G418 selection (100 μg/ml), no treatment or NGF stimulation (150 ng/ml). Cultures infected with wild-type TRK and maintained for 4–7 days past confluence developed NGF-dependent populations of transformed colonies that were much more rounded and phase-bright than the background cells and that were easily detected in dark field illumination by their distinctive refractiveness (not shown). The transformed cells also showed TRK immunoreactivity (not shown). No transformed colonies were detected without NGF. Transformation efficiency with virus expressing wild-type TRK was 68 ± 10% (n = 12) (Figure 1B) as defined by the ratio of the total number of NGF-induced transformed colonies observed to the total number of virions delivered (see Materials and methods). Functional activities of the mutant receptors in this assay (Figure 1B) were identical to those achieved with the polyclonal PC12nnr5 cells (Figure 1A); the Y670F and Y674F mutants showed strong transformation activity while the Y675F mutant and all double combination mutants were only weakly active, as was the K538N kinase-deficient TRK. Taken together, these observations indicate that mass cultures of transfected NIH 3T3 and PC12nnr5 cells are suitable to bioassay the relative functional activities of TRK mutants. A functional screen for contacts of phosphorylated activation loop tyrosines Our prior findings indicated that mutation of any of the activation loop tyrosines leads to diminished TRK function (Cunningham et al., 1997). Also indicated in that study, and confirmed here, was that this impairment can be rescued by overexpression in the cases of Y670F and Y674F, but not for Y675F or any double combination mutation of these sites. We suggested that one mechanism to explain the role of these tyrosines is that after phosphorylation they interact with nearby charged (or polar) residues and that such contacts in turn contribute to formation and stabilization of a fully activated TRK conformation. Disruption of any of the phosphorylations would impair such interactions, destabilize the activated conformation and consequently diminish receptor function. We further reasoned that if such interactions occur, then function should be equally affected by mutation of the corresponding charge partner of an activation loop phosphotyrosine. Employing these points, we developed a strategy to test our model and to identify charge partners for each of the activation loop phosphotyrosines. In this screen, candidate charge pair partners were mutated to alanine and evaluated for function (NGF-dependent transformation and/or neurite outgrowth) in backgrounds in which none (wild-type) or each of the activation loop tyrosines were mutated individually to phenylalanine. As an example (Figure 2A), a TRK bearing a mutation at either Tyr670, the charge partner for Tyr670, or both of these sites would show little loss of function when overexpressed. The same is true for Tyr674 and its corresponding charge partner. In contrast, when TRKs with paired mutations of Y670F and the charge partner for Tyr674, or of Y674F and the charge partner for Tyr670 are overexpressed, function would be expected to be lost because the YY670/674FF double mutation would be simulated (Figure 2A). These concepts allow a definition to be made of the phosphotyrosine charge pair partner that is grounded in preserved background-specific function. Namely, a charge pair partner candidate, when mutated alone, must have NGF-inducible function comparable with that of the singly mutated activation loop tyrosines (Y670F or Y674F), and must retain a similar level of inducible function in one of the backgrounds of mutant activation loop tyrosines while losing function in the other two. The specific phosphotyrosine involved in charge pairing with the candidate residue is then the background in which function is preserved. Initially, a potential limitation of our proposed screen was that the Y675F mutation alone shows little activity and therefore might not show further functional loss when paired with additional mutations. Figure 2.Rationale for functional screen-in-backgrounds and model for candidate residue locales. (A) NGF-induced TRK function requires two fully intact phosphotyrosine-based bridging interactions. Disruptions of a candidate charge, or of this and its phosphotyrosine partner retain function (left and upper situations), but disruptions involving two bridges result in loss of function (bottom). (B) Structural coordinates from the active IRK crystal structure (Brookhaven PDB, accession No. 1IR3) were adapted to depict graphically the candidate polar and basic TRK residues evaluated in this study using SETOR (Evans, 1993). TRK residues are identified on the structure as deduced by comparison of the primary sequences of the kinase domains. Mutant sites are color coded as follows: yellow indicates constructs that were not expressible, magenta indicates loss of function mutants, green indicates neutral mutations and red indicates background-specific function as predicted by our model. Download figure Download PowerPoint Identification of charge pair partners for phosphotyrosines 670 and 674 To implement the above screen, candidate residues for mutation were chosen on the basis of their positive charge or polar character, and of their potential three-dimensional positions relative to the activation loop tyrosines as predicted by imposition of the homologous TRK sequence on the crystal structures of the insulin and fibroblast growth factor (FGF) receptor kinase domains (Hubbard et al., 1994; Mohammadi et al., 1996; Hubbard, 1997). Comparison of the primary sequence of TRK (amino acids 486–790) with the construct used for crystallizing IRK (Hubbard et al., 1994) demonstrated 44% identity and 61% homology between the proteins. To aid the interpretation of the findings presented here, Figure 2B shows the TRK sites evaluated in this study superimposed onto the crystal structure of the activated IRK (Hubbard, 1997), which are color coded as defined in the legend. Because of the large number of candidate residues to be assessed, initial studies were carried out with the NIH 3T3 cell transformation assay. Several of the mutations, Q552A (Figure 3A), Q549A and T559A (data not shown), had no effect on function in all four backgrounds (wild type, Y670F, Y674F or Y675F) (compare Figures 1B and 3A). Likewise, preliminary findings showed that mutation of Gln724, Thr728, Gln736 or Arg738 did not affect TRK function in the Y674F background (data not shown). This indicates that mutations of nearby charged or polar residues do not lead indiscriminately to loss of activity. In contrast, TRKs bearing mutations at Lys541 or His642 showed loss of function in all backgrounds. The latter point mutant lost the capacity to become tyrosine autophosphorylated in response to NGF (data not shown), indicating that His642 plays an indispensable role in TRK kinase function. The corresponding IRK residue (His1130) was reported to form main chain interactions that stabilize the catalytic loop (Hubbard et al., 1994). Interestingly, despite the loss of NGF-inducible biological activity, the K541A point mutant retained a significant level of NGF-dependent autophosphorylation (data not shown). Substitution of the TRK residues Glu512 and Lys541 in the IRK activated structure reveals the potential for interaction of these two sites. Thus, the mechanism for impairment of the Lys541 mutant may involve disruption of a standing electrostatic interaction that in turn leads to disordering of the activated state. Three other candidates, Arg553, His563 and Lys659, showed significantly diminished function in the wild-type background (each with 21–25% transformation efficiency) and showed little to no function in the other backgrounds. Because none of these mutants fulfill the criteria for a phosphotyrosine charge pair partner, they were not investigated further in this study. Mutation of several other potential interaction sites resulted in loss of expression and thus their potential contributions as phosphotyrosine interactors could not be assessed. As inferred from the IRK crystal structure, these poorly expressible mutants either precede the kinase domain (His497 and His498), are contained in α-helix C (Arg548) or are located in the interdomain hinge region of the kinase (His565 and His568) (Figure 2B). Figure 3.Mutation of Gln552 does not impair NGF-inducible TRK function, and His639 and arginines 643, 667, 676 and 680 form functional charge pairs with phosphorylated activation loop tyrosines. NIH 3T3 cells were infected with recombinant retroviruses bearing the indicated TRKs, expanded and subjected to NGF, G418 or no treatment as described in Materials and methods. Values reported are the averages ± SEM of three independent experiments. (A and B) Receptor mutants bearing mutations of the indicated positively charged residues in the backgrounds of wild-type (filled white) or with each of the activation loop tyrosines mutated individually (Y670F filled gray, Y674F cross-hatched and Y675F filled black). (C) Receptor mutants bearing the R643A/R676A and R667A/R676A double mutations in the background of wild-type activation loop tyrosines. Download figure Download PowerPoint Several mutations of positively charged residues exhibited background-specific effects on function that conformed to the predictions of our model (Figure 3). The H639A mutation showed significant activity in the wild-type and Y670F backgrounds and little to no activity in the Y674F or Y675F backgrounds (Figure 3A). Although the absence of activity in the Y675F background is not informative, as described above, the loss of activity in the Y674F background identifies His639 as a charge partner for phosphotyrosine 670 (pY670). Similarly, the R676A mutation is active in the wild-type and Y674F backgrounds, but not when combined with the Y670F or Y675F mutations (Figure 3A) and, therefore, this arginine behaves as a functional charge pair partner for pY674. Likewise, the R680A mutation exhibited a partial, but significant (p = 0.1, by Student's t-test) diminution of activity in the Y670F background when compared with the wild-type background, and preserved activity in the Y674F background (Figure 3A); this suggests that this residue may be a weak charge pair partner for pY674. Consistent with this, the functional activity of the combined RR676/680AA mutations assessed in the four test backgrounds showed no further loss when compared with that of the R676A mutation alone (compare Figure 3A with B). Mutation of Arg643 restores the activity of Y675F and permits the identification of arginines 643 and 667 as charge pair partners for pY675 Unexpectedly, mutation of Arg643 selectively restored function in the Y675F background so that the R643A/Y675F double mutant was as functional in the transformation assay as the Y670F and Y674F single mutants (compare Figures 1B and 3B). In contrast, the R643A/Y670F and R643A/Y674F double mutants showed nearly total loss of activity, while the R643A mutation in the wild-type background caused no loss of activity. These observations suggest several points. First, rescue by the R643A mutation is not simply a result of general activation since it was background specific, and since, irrespective of the background, receptors with this mutation failed to show activity without NGF (data not shown). Secondly, R643A follows the pattern of function predicted by our model, with specific loss of function in the Y670F and Y674F backgrounds, and thus Arg643 appears to be a charge pair partner for pY675. Thirdly, these findings suggest a mechanism for the dramatic loss of function in the Y675F single mutant as compared with the Y670F or Y674F single mutants: in the absence of pY675, another charged residue (perhaps pY674) may interact inappropriately with Arg643. This interpretation would explain why the R643A mutation restores function in the Y675F background. When Arg643 is mutated, the incorrect charge pair cannot form and the phenotype would be equivalent to the loss of only a single charge pair interaction (i.e. like Y670F or Y674F). The observation that the R643A mutation rescues Y675F function permitted us to seek additional partners for phosphorylated Tyr675. Our initial screen revealed that the R667A mutant TRK possesses efficient NGF-induced transformation activity in the wild-type background, but little or no activity when combined with any of the three mutant activation loop tyrosines (Figure 3B). The loss of activity in the Y670F and Y674F backgrounds is consistent with the possibilities that Arg667 is a charge pair partner for pY675 and that, unlike R643A, mutation of Arg667 does not restore function to the Y675F background. Support for these possibilities is provided by the crystal structure of the activated LCK kinase domain which indicates charge pairing homologous to both the R643–pY675 and R667–pY675 TRK interactions. To test this hypothesis, the double TRK mutant RR643/667AA was prepared and assessed in all four backgrounds. The rationale was that if Arg667 were a charge pair partner for pY675, combination with the R643A mutation would rescue function selectively in the Y675F background, but not the Y670F or Y674F backgrounds. As shown in Figure 3B, this is the case, thus indicating that Arg667 and Arg643 are both charge partners for pY675. When this strategy was applied to additional candidates, Gln724, Thr728, Gln736 and Arg738 were excluded as partners for pY675 as were Arg676 and Arg680 (data not shown). Combination mutants of charge pair partners An important prediction of our model is that combined mutation of charge pair partners for different phosphotyrosines should lead to the same loss of function seen with combined mutation of the phosphotyrosines themselves. To test this, combined mutations of pY674 and pY675 charge pair partners were assessed in a background of wild-type phosphotyrosines (Figure 3C). In agreement with the model, the R676A/R667A combination showed nearly total loss of function. The R643A/R676A mutant also lost significant, but not total, function. The partial function remaining in this mutant suggests a partial stabilizing influence by the remaining pY675 interactor, Arg667. Similarly, in accordance with its weak putative interaction with pY674, paired mutation of Arg680 with pY675 interactors showed little loss of function (data not shown). Arg696 and Lys697 play roles in stabilization of activated TRK, but do not appear to be charge pair partners for activation loop phosphotyrosines Two additional mutations of charged residues (R696A and K697A) were identified that, like R667A, resulted in loss of NGF-dependent function in all three backgrounds of activation loop mutations, but not in the wild-type background (Figure 4A). These observations seem to rule out interactions with pY670 and pY674 but, as discussed for the R667A mutation, do not rule out interaction with pY675. To assess these possibilities further, combination mutations of charge pair partners (Arg643, Arg667, Arg676 or Arg680) with R696A or K697A were evaluated for function in the background of wild-type activation loop tyrosines (Figure 4B). The RR676/696AA and R676A/K697A mutant TRKs were not functional, confirming that Arg696 and Lys697 do not interact with pY674. Furthermore, all mutant TRK combinations containing R696A or K697A and charge pair partners for pY675 were highly to totally impaired, which eliminates the possibility that either Arg696 or Lys697 interacts with pY675. In addition, when R643A was superposed onto the R696A/Y675F mutant, function in this background was not rescued (transformation efficiency 2.4%). The partial activity retained by the R643A/K697A receptor is similar to that of the RR643/676AA receptor, and thus probably reflects the stabilizing influence of the remaining pY675 interactor, Arg667. Taken together, these observations indicate that Arg696 and Lys697 help to stabilize the TRK active state, but do so by a mechanism other than by interaction with activation loop phosphotyrosines. Another interesting point here is the functional loss shown in the R696A/K697A double mutant receptor (Figure 4B), which indicates that either the two residues act independently or that they interact with a common site that is tolerant to loss of one, but not both interactions. Figure 4.Arg696 and Lys697 do not interact with activation loop phosphotyrosines, but stabilize NGF-activated TRK. NIH 3T3 cells were infected with recombinant retroviruses bearing the indicated TRKs, expanded and subjected to NGF, G418 or no treatment, as described in Materials and methods. Values reported are the averages ± SEM of three independent experiments. (A) Receptor mutants bearing mutations of the indicated positively charged residues in the backgrounds of wild-type (filled white) or with each of the activation loop tyrosines mutated individually (Y670F filled gray, Y674F cross-hatched and Y675F filled black). (B) Receptor mutants bearing the indicated combination mutations in the background of wild-type activation loop tyrosines. Download figure Download PowerPoint Assessment of function in the PC12nnr5 cell neurite outgrowth assay confirms assignment of charge pair partners To determine if the above findings are applicable to a neuronal environment (which is more typical for TRK expression and function), selected mutant TRKs from the preceding experiments were assessed in the PC12nnr5 cell neurite outgrowth rescue assay (Loeb et al., 1991). Comparison of results from the neurite outgrowth assay (Figure 5) with those obtained in the transformation assay (Figure 3) revealed nearly identical findings, which confirmed that mutation of Gln552 has no evident phenotype in any background, and that His639, Arg676 and Arg643 are charge pair partners for pY670, pY674 and pY675, respectively. The parallel degrees of function observed in both assays suggest that the levels of biological 'readout' in each system (i.e. mitogenesis or differentiation) is dependent more upon the functional capacity of the mutant TRK receptor than upon its cellular context. Figure 5.Neuritogenic activity of TRKs bearing mutations of identified charge pair partners. Polyclonal cultures of PC12nnr5 cells stably transfected with the indicated TRKs were assessed for NGF-induced neuritogenesis at a 7 day time point. TRKs contain mutations of positively charged residues in backgrounds of wild-type (filled white) or with each of the activation loop tyrosines mutated individually (Y670F filled gray, Y674F cross-hatched and Y675F filled black). Reported values are the averages ± SEM for a representative, duplicated experiment. Download figure Download PowerPoint Mutation of charge pair partners for activation loop phosphotyrosines selectively impairs NGF-dependent phosphorylation of PLCγ-1 In our study of clonal PC12nnr5 cells, mutation of any one of the TRK activation loop tyrosines resulted in selective, severe impairment of NGF-dependent PLCγ-1 tyrosine phosphorylation, even when the receptors were highly overexpressed (Cunningham et al., 1997). This appeared to be due to loss of the capacity of the mutated receptors to autophosphorylate the PLCγ-1-binding site, Tyr785. In contrast, the extent of Shc phosphorylation was diminished, but, in parallel with biological function, was substantially rescued by receptor overexpression. We interpreted these findings to suggest that phosphorylation of activation loop tyrosines not only regulates TRK activation, but also contributes to stabilization of a conformational change that permits phosphorylation of Tyr785. If our charge pair model is correct, then mutation of phosphotyrosine charge pair partners should also selectively compromise NGF-promoted PLCγ-1 phosphorylation. When this was assessed in polyclonal PC12nnr5 cell cultu
Referência(s)