Original Article

Oncogene (2007) 26, 6546–6559; doi:10.1038/sj.onc.1210480; published online 30 April 2007

SH2B1bold italic beta adaptor is a key enhancer of RET tyrosine kinase signaling

S Donatello1,5, A Fiorino1,2,5, D Degl'Innocenti1,5, L Alberti1, C Miranda1, L Gorla1, I Bongarzone1, M G Rizzetti1, M A Pierotti3,4 and M G Borrello1

  1. 1Department of Experimental Oncology, Research Unit no. 3, Milan, Italy
  2. 2Department of Experimental Oncology, Research Unit no. 4, Milan, Italy
  3. 3Scientific Direction, IRCCS Istituto Nazionale Tumori Foundation, Milan, Italy
  4. 4FIRC Institute of Molecular Oncology Foundation, Milan, Italy

Correspondence: Dr MG Borrello, Department of Experimental Oncology, UO3, IRCCS Istituto Nazionale Tumori Foundation, Via Venezian 1, 20133 Milan, Italy. E-mail: mariagrazia.borrello@istitutotumori.mi.it; Dr MA Pierotti, Scientific Direction, IRCCS Istituto Nazionale Tumori Foundation, Via Venezian 1, 20133, Milan, Italy. E-mail: marco.pierotti@istitutotumori.mi.it

5These authors contributed equally to this work.

Received 24 July 2006; Revised 21 March 2007; Accepted 22 March 2007; Published online 30 April 2007.



The RET gene encodes two main isoforms of a receptor tyrosine kinase (RTK) implicated in various human diseases. Activating germ-line point mutations are responsible for multiple endocrine neoplasia type 2-associated medullary thyroid carcinomas, inactivating germ-line mutations for Hirschsprung's disease, while somatic rearrangements (RET/PTCs) are specific to papillary thyroid carcinomas. SH2B1beta, a member of the SH2B adaptors family, and binding partner for several RTKs, has been recently described to interact with proto-RET. Here, we show that both RET isoforms and its oncogenic derivatives bind to SH2B1beta through the SRC homology 2 (SH2) domain and a kinase activity-dependent mechanism. As a result, RET phosphorylates SH2B1beta, which in turn enhances its autophosphorylation, kinase activity, and downstream signaling. RET tyrosine residues 905 and 981 are important determinants for functional binding of the adaptor, as removal of both autophosphorylation sites displaces its recruitment. Binding of SH2B1beta appears to protect RET from dephosphorylation by protein tyrosine phosphatases, and might represent a likely mechanism contributing to its upregulation. Thus, overexpression of SH2B1beta, by enhancing phosphorylation/activation of RET transducers, potentiates the cellular differentiation and the neoplastic transformation thereby induced, and counteracts the action of RET inhibitors. Overall, our results identify SH2B1beta as a key enhancer of RET physiologic and pathologic activities.


RET, SH2B1, signaling



RET is a proto-oncogene encoding a receptor tyrosine kinase (RTK) that forms a transmembrane receptor complex with the glial cell line-derived neurotrophic factor (GDNF) coreceptors alpha (GFRalpha1-4) for the GDNF ligands (Airaksinen and Saarma, 2002). This transducing complex bridges two molecules of RET together and, by triggering their transphosphorylation at specific tyrosyl residues, recruits either adaptor or effector proteins (Ichihara et al., 2004; Arighi et al., 2005) and activates several downstream cascades. Ret signaling is critical to the neural crest and excretory system development (Arighi et al., 2005). The two major alternatively spliced isoforms (RET-9 or short, 1072 aa; RET-51 or long, 1114 aa) differ at their cytoplasmic tail and, by recruiting a specific set of adaptors, exhibit a distinct role in mouse development (de Graaff et al., 2001).

Specific RET mutations are implicated in several pathogenic disorders. Activating germline mutations at specific amino-acid residues are responsible for hereditary multiple endocrine neoplasia (MEN) syndromes (MEN2A, MEN2B and FMTC), while loss-of-function mutations confer predisposition to Hirschsprung's disease (HSCR) (Arighi et al., 2005). Moreover, somatic chromosomal rearrangements at the RET locus, present in about one-third of papillary thyroid carcinomas (PTCs), lead to aberrant expression of chimeric proteins composed of RET cytoplasmic domain fused to the 5'-region of unrelated donor genes (Alberti et al., 2003). These oncogenes, referred as RET/PTCs, are constitutively activated through ligand-independent dimerization and subsequent transphosphorylation, and lead to neoplastic transformation (Arighi et al., 2005). They have been extensively studied to identify the cytoplasmic mediators recruited by RET-specific docking sites. Several interacting proteins, such as PLC-italic gamma, GRB2, GRB7, GRB10, SHC, Enigma, IRS1/2, FRS-2, SRC, SHANK3, DOKs, PDK1 and STATs (Takahashi, 2001; Arighi et al., 2005) have been so far identified, with certain adaptors exhibiting binding specificity for distinct RET isoforms. The recruitment of most transducers has been demonstrated to be crucial for activation of RET downstream pathways, including RAS/MAPK, p38alpha/JNK, PI3K/Akt and others typical of RTKs (Ichihara et al., 2004), and cellular responses. However, the complexity of mechanisms regulating RET intracellular signaling appears to suggest new functional links between RET and its mediated pathways.

To study novel mediators of RET signaling, SH2B1beta (formerly SH2-Bbeta) was identified as RET-binding protein through two-hybrid screening methodologies (Donatello et al., 2004; Zhang et al., 2006). SH2B1beta is a member of the structurally related SH2B family of adaptor proteins, characterized by a SRC homology 2 (SH2), a pleckstrin homology (PH) domain (Osborne et al., 1995; Riedel et al., 1997), and a recently identified 'phenylalanine zipper' dimerization domain (DD) (Dhe-Paganon et al., 2004). Genetic disruption of the SH2B1 gene results in obesity, type-II diabetes and impaired fertility (Duan et al., 2004; Ren et al., 2005). Four SH2B1 splice variants have been identified in mouse (alpha, beta, italic gamma and delta), and three in humans (alpha, beta and italic gamma) (Yousaf et al., 2001), coding for polypeptides with distinct C-terminal sequences and specific biological properties. The beta isoform has been reported as a binding partner for janus kinase (JAK), platelet-derived growth factor receptor (PDGFR), insulin-like growth factor 1 receptor (IGF1R), fibroblast growth factor receptor (FGFR), growth hormone receptor (GHR), insulin receptor (IR) and nerve growth factor receptor (NGFR) (Rui et al., 1997; Rui and Carter-Su, 1998; Wang and Riedel, 1998; Ahmed and Pillay, 2001; Qian and Ginty, 2001), but is unable to associate with epidermal growth factor (EGFR) (Yousaf et al., 2001). Its recruitment by the above tyrosine kinase (TKs) is mostly mediated by its SH2 domain. At variance with the majority of adaptor proteins acting downstream to RTKs, SH2B1beta was shown to enhance autophosphorylation of IR, NGF receptors and cytokine receptor-associated JAK2 (Qian and Ginty, 2001; Ahmed and Pillay, 2003), as well as to potentiate NGF-induced neuronal differentiation (Qian et al., 1998). Consistently, leptin (through JAK2) and insulin sensitivity were found positively regulated by SH2B1 in mouse (Duan et al., 2004; Ren et al., 2005).

In this study, we show that RET isoforms and their oncogenic derivatives associated to thyroid carcinomas, bind and phosphorylate SH2B1beta. We provide evidence that coexpression of this adaptor increases RET autophosphorylation, kinase activity and downstream signaling, thus potentiating its differentiating and transforming ability, and give some insights into the mechanisms underlying these processes. Furthermore, we report that SH2B1beta coexpression protect RET oncoproteins by specific RET inhibitors.



RET oncoproteins bind and phosphorylate SH2B1beta through a RTK-dependent mechanism

RTKs' adaptors/effectors are known to interact with their intracellular domain. To identify new potential RET-binding proteins, we used the RET/PTC2-iso9 protein fused to the GAL4-DBD as bait, and screened a GAL4 activation domain-tagged human placenta cDNA library by yeast two-hybrid system. RET/PTC2 oncogene arises from somatic rearrangements between the RET intracellular portion and the 5'-end of PRKAR1A (encoding the R1alpha subunit of PKA) in PTC (Bongarzone et al., 1993) (Figure 1a). The 21 clones identified as specifically interacting with RET/PTC2 (and not with the unrelated GAL4-DBD-lamin) correspond to known RET cofactors, such as SHC, PLCitalic gamma, GRB7, GRB10, Nck and p85alpha(PI3K), and to novel putative interacting proteins. Among them, two clones were a 639- and a 474-bp overlapping fragments, mapping at the 3'-end of SH2B1 cDNA. The isolated region includes the entire SH2 domain of SH2B1 and a small portion of flanking sequence (463–670 aa sequence), which allows identification of the isolated clones as coding for the SH2B1beta isoform (Figure 1a) (Donatello et al., 2004).

Figure 1.
Figure 1 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact help@nature.com or the author

SH2B1beta interacts with different RET forms and is phosphorylated at tyrosyl residues. (a) Schematic representation of GAL4-DBD-RET/PTC2 (bait) and SH2B1beta adaptor protein (prey). Proline rich (P), DD, PH and SH2 domains, the region isolated by two-hybrid screening (aa463–670) (black arrow), and the isoform specific C-terminus (open arrow) are indicated. (b) RET binds in vitro to recombinant GST-SH2B1beta protein. Lysates from RET/PTC2-transfected 293T cells were pulled down with the GST-fused proteins or with GST only and subjected to immunoblotting with the indicated antisera. CIAP, calf intestine alkaline phosphatase; RE, R555E mutant. (c and d) RET proteins interact with SH2B1beta in intact cells. RET/PTC2 (c) RETC634R (MEN2A-associated) and RETM918T (MEN2B-associated) (d) cotransfected with SH2B1beta constructs in 293T cells, were immunoprecipitated with alphaRET or alphaMYC, followed by immunoblotting with alphaRET, alphaGFP, alphaMYC and alphaphosphotyrosine (alphapTyr) antibodies. KR, K282R mutant. WCL, whole-cell lysate. Iso9 and Iso51, short and long RET isoforms. (e) SH2B1beta is directly phosphorylated by RET in vitro. RET/PTC2 transiently transfected in 293T cells was immunoprecipitated with alphaRET and subjected to kinase assay on GST, GST-SH2B1beta and GST-SH2BbetaRE proteins.

Full figure and legend (133K)

To rule out the contribution of the RIalpha portion in RET/PTC2 interaction with SH2B1beta, we tested the ability of the two SH2Bbeta1 clones to interact with the RET/PTC2(236–595) construct (Figure 1a). This construct, which encompasses both the functional RET TK domain and the C-terminal tail (aa sequence 714–1071 of proto-RET), retained the ability to bind SH2B1beta, thus indicating that RET intracellular region is directly involved in the interaction with the docking protein (data not shown).

To confirm the specificity of this interaction, we pulled-down lysates of RET/PTC2-transfected cells with GST-fused SH2B1beta proteins (Figure 1b), and evaluated the extent of binding. RET/PTC2 directly binds the wild-type adaptor, while its SH2Bbeta1R555E counterpart, mutated at the FLVR motif of the SH2 domain and unable to bind active RTKs (Rui and Carter-Su, 1999), showed no interaction. Furthermore, the inability of RET/PTC2 to be retained by the resin-bound SH2B1beta upon cell lysate treatment with phosphatase, strongly indicates that RET phosphorylation is required for this interaction.

As proto-RET and its oncogenic derivatives recruit several adaptors/effectors, and SH2B1beta has been described as an adaptor of several TKs, we sought to demonstrate whether (i) SH2B1beta interaction with RET is associated to specific RET isoforms; (ii) it is kinase-activity-dependent; and (iii) RET/PTC2 directly phosphorylates SH2B1beta. We therefore cotransfected 293T cells with either the short (-9) or the long (-51) RET/PTC2 isoform and the GFP-tagged SH2B1beta (or empty vector). Both RET/PTC2 isoforms, but not RET/PTC2K282R kinase-dead, coprecipitated with SH2B1beta and not with its SH2-inactive mutant (Figure 1c). Transmembrane forms of RET, including proto-RET (not shown), and MEN2A- and MEN2B-associated constitutive active mutants (Figure 1d), were all able to interact with the Myc-tagged SH2B1beta, which indeed excluded undesired tag-specific effects. A significant level of phosphorylated adaptor was detected in RET/SH2B1beta-wt coprecipitates.

A further demonstration of SH2B1beta phosphorylation by RET was provided by immunoprecipitating lysates from RET/PTC2- and mock-transfected cells with antiserum to RET and assaying the kinase activity on purified GST-SH2B1beta as substrate (Figure 1e). After blotting, we observed that the wild-type adaptor protein yielded a distinct [32P]-labeled band, while no phosphorylation was detected with the SH2B1betaR555E mutant. Together, the data obtained in vitro and in intact cells demonstrated that both kinase activity and a functional SH2 domain are indispensable for SH2B1beta recruitment to RET, which in turn phosphorylates the adaptor.

Y905 and Y981 on RET are major determinants of its interaction with SH2B1beta

The complex pattern of RET autophosphorylation has been already revealed. To map on RET the tyrosines putatively binding SH2B1beta, affinity-bound GST-SH2B1beta resin was used to pull-down RET/PTC2 constructs devoid of distinct autophosphorylation sites (shown in Figure 2a). Kinase-dead K282R and truncated N538 mutants, Y424F/Y429F, Y505F/539F and Y424D/Y429E (DE) double mutants were also included in the assay. RET/PTC2DE, which mimics constitutive autophosphorylation in the activation loop (Gryz and Meakin, 2000), maintains comparable kinase activity with the RET/PTC2 (not shown), but deprives the corresponding residues of potential docking activity. RET/PTC2N538, lacking the C terminus corresponding to residues 1015–1071 on RET, is devoid of the 1015 and 1062 docking sites, but still able to undergo autophosphorylation. As shown in Figure 2b, mutation at Y276 and Y452 (corresponding to Y752 and Y928, docking site for STATs), Y424F (Tyr900, activation loop, comparable autophosphorylation with RET-wt), Y539 (Tyr1015, docking site for PLCitalic gamma), Y586 (Tyr1062, multidocking site for SHC, FRS2, Enigma, DOKs and IRS1/2) (Arighi et al., 2005), and deletion of 1015–1071 sequence only slightly affected SH2B1beta binding to RET/PTC2, indicating that these docking sites and the C-terminal tail are dispensable for this interaction. In contrast, Y505F mutation (Tyr981, putative binding site for SRC) reduced RET binding to the resin-bound adaptor by more than 60%, and Y429F construct (Tyr905, in the activation loop), exhibiting the least autophosphorylation among the single mutants (Figure 2b, and (Kawamoto et al., 2004)), was barely retained by the resin. Further, Y424F/Y429F and Y505F/Y539F double mutants showed comparable binding, respectively, with Y429F and Y505F single mutants, and RET/PTC2DE construct, mimicking phosphorylation, partially restored this binding.

Figure 2.
Figure 2 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact help@nature.com or the author

Binding of RET to SH2B1beta is regulated by specific tyrosyl residues. (a) Schematic representation of the two isoforms of RET/PTC2 and proto-RET proteins. Signal peptide (SP), cadherin-like (CAD), cysteine-rich (CYS), transmembrane (TM) and TK domains in the two proteins are indicated, together with the corresponding tyrosine residues. (b) Specific Y to F mutations impair interaction with SH2B1beta. Lysates of 293T cells transfected with RET mutants were pulled down with the immobilized GST-SH2B1beta. The binding efficiency of RET/PTC2 mutants determined as bound protein (detected with alphaRET), relative to input expression, and normalized to the bound RET/PTC2-wt is indicated. Input, WCLs were immunoblotted with alphaRET and alphapTyr as control. (c) Interaction of SH2B1beta to RET/PTC2 and phosphorylation is regulated by both Y429 and Y505 residues on RET. RET/PTC2 and SH2B1beta cotransfected in 293T cells were precipitated with alphaRET and blotted as specified. Control of SH2B1beta expression was performed on WCLs.

Full figure and legend (190K)

The above pull-down experiments suggest that the interaction of RET with SH2B1beta is mediated by Y905 and Y981 sites. To confirm this finding, we constructed a RET/PTC2Y429/Y505F double mutant (FF), and immunoprecipitated with RET antiserum lysates from 293T cells cotransfected with SH2B1beta and RET/PTC2 mutants. As shown in Figure 2c, the adaptor protein was present and phosphorylated in immunoprecipitates from RET/PTC2-Y505F and, at lesser extent, from Y429F-transfected cells. In contrast, no SH2B1beta was detected in RET precipitate from FF mutant. Therefore, these results support a RET/SH2B1beta interaction mediated by both Y905 and Y981 of RET kinase domain.

SH2B1beta enhances RET phosphorylation and kinase activity

We have observed that cotransfection of RET/PTC2 with SH2B1beta increases RET protein level and enhances its autophosphorylation. To study this effect better, we cotransfected RET/PTC2-iso51 with either SH2B1beta or GRB2, as a control. As shown in Figure 3a, coexpression of GRB2, an adaptor known to bind RET-iso51 (Alberti et al., 2003), had no effect on RET autophosphorylation. In contrast, SH2B1beta, but not its R555E inactive mutant, dramatically increased tyrosine phosphorylation and raised protein levels.

Figure 3.
Figure 3 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact help@nature.com or the author

SH2B1beta coexpression upregulates RET autophosphorylation and kinase activity. (a) Wild-type SH2B1beta (but not GRB2 adaptor) enhances RET protein phosphorylation. RET/PTC2-iso51 was cotransfected with GFP-tagged GRB2 or SH2B1beta adaptors, immunoprecipitated and immunoblotted with alphaRET and alphapTyr. (b) Phosphorylated SH2B1beta-wt enhances in vitro RET kinase activity. RET/PTC2 isolated from 293T lysates with alphaRET was preincubated with 0.1 mug of GST, GST-SH2B1beta or GST-SH2B1betaRE proteins in presence or absence of ATP and its kinase activity determined on MBP substrate by autoradiography of 32P-labeled bands. The relative intensity of 32P-labelled MBP is shown. (c) RET kinase inhibition by PP1 inhibitor is hampered by SH2B1beta. 293T lysates coexpressing RET/PTC2 and SH2B1beta were subjected to immunokinase assay. The extent of phosphorylation was evaluated by densitometry and the relative intensity of 32P-labelled MBP is shown. (d) SH2B1beta counteracts RET dephosphorylation by SHP-1 tyrosine phosphatase. RET/PTC2 was co-transfected with wild-type SHP-1 or its inactive C455S mutant, together with increasing amounts of the adaptor vector (0–1.5 ratio over RET cDNA). RET phosphorylation was detected on WCLs with alphapTyr.

Full figure and legend (196K)

To investigate whether SH2B1beta directly affects RET kinase activity, an in vitro kinase immunocomplex assay was performed by preincubating RET/PTC2 and GST-SH2B1beta with or without cold adenosine triphosphate (ATP), and by adding the myelin basic protein (MBP) substrate and [italic gamma32-P]ATP. Assembling of RET/SH2B1beta-wt complex in presence of ATP, which allows phosphorylation of the adaptor (see Figure 1e), resulted in improved incorporation of [32P] by both RET/PTC2 and MBP (Figure 3b, right panel). At variance, similar levels of labeled MBP were observed when ATP was added together with the substrate (100-fold molar excess over the recombinant SH2B1beta), indicating that efficient phosphorylation of the adaptor is required to enhance RET activity. To evaluate the contribution of SH2B1beta to RET activity in intact cells, we determined the effect of the TK inhibitor PP1 (Carlomagno et al., 2002) on RET/PTC2 coexpressed or not with the adaptor. As expected, SH2B1beta-wt improved its kinase activity (up to 2.5-fold), even in the presence of PP1, which induces a dose-dependent inhibition of RET (Figure 3c). These data, while suggesting that SH2B1beta incorporates in a fully active complex with RET, demonstrate that the adaptor markedly protected the kinase from the inhibitory effect of PP1 compound at relative low inhibitor dose (0.5 muM). Comparable results were obtained by using the RPI-1 (Cuccuru et al., 2004) specific RET blocker (not shown), and by assessing RET-driven transcriptional activity in presence of TK inhibitors (Supplementary Figure 1S).

Further, as SHP-1 protein tyrosine phosphatase modulates the activity of RET-derived oncogenes by reducing their phosphorylation (Hennige et al., 2001), we asked whether SH2B1beta interaction with RET may modulate SHP-1-mediated dephosphorylation in intact cells. We coexpressed RET/PTC2 and SHP-1 (or its inactive C455S mutant (Roccato et al., 2005) together with increasing amounts of SH2B1beta, and examined the steady-state phosphorylation of RET. As shown in Figure 3d, in absence of SH2B1beta, RET/PTC2 phosphorylation was greatly reduced in cells transfected with SHP-1-wt. In presence of the adaptor protein, the SHP-1-expressing cells exhibited a SH2B1beta concentration-dependent RET phosphorylation. Consistently, no significant dephosphorylation was detected in cells transfected with the C455S mutant. This finding suggests that the complex formed by the interaction of RET with the adaptor protein might increase its autophosphorylation by protecting it from the action of phosphatases.

We also evaluated whether the elevated levels of RET observed in presence of SH2B1beta were due to a stabilizing effect of the binding complex. The steady-state RET/PTC2 levels were increased by cotransfection with SH2B1beta-wt, but no significant augmentation of its half-life was observed in protein turnover studies with cycloheximide protein translation blocker (Supplementary Figure 2S). Nevertheless, the wild-type adaptor had a positive effect on the oncoprotein level, as the increase in RET expression by SH2B1beta was maintained during the cycloheximide treatment.

RET activation of downstream effectors is enhanced by SH2B1beta

The observed effects of SH2B1beta on RET autophosphorylation and kinase activity suggest an upregulation of RET downstream signaling through enhancement of both the recruitment of docking proteins and the activation of its direct substrates. Therefore, we first studied the effect of RET/SH2B1beta complex on STAT3 activation, a direct substrate of several TKs, which mediates oncogenesis through phosphorylation of its Y705 residue (Bromberg et al., 1998). We assayed the extent of STAT3 activation by immunoblotting, and studied transactivation of the responsive pm674- and pCD1-luciferase reporter constructs, including respectively a high-affinity binding site for activated STAT3 (Besser et al., 1999) and the 1700-bp cyclin D1 promoter (Lee et al., 1999), which requires STAT3 and other factors for full activation. As shown in Figure 4a, at low RET/PTC2 expression levels in HeLa cells we could barely detect activation of endogenous STAT3, and measured a negligible effect on m67- and CD1-driven activity (Figure 4b). Coexpression of SH2B1beta resulted in a marked phosphorylation of endogenous STAT3, more elevated RET levels, and a significant increase in pm674luc and pCD1luc transactivation (respectively by five- and threefold) over its R555E mutant. This trend was confirmed by exogenous STAT3 expression.

Figure 4.
Figure 4 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact help@nature.com or the author

SH2B1beta improves SHC and STAT3 activation. (a) Phosphorylation of STAT3 by RET increases upon SH2B1beta coexpression. HeLa cells were transfected with combination of RET/PTC2, FLAG-tagged wild type or mutated STAT3, and GFP-SH2B1beta, and STAT3 phosphorylation was detected with alphapY705-STAT3. (b) RET-driven activation of STAT3-responsive promoters is enhanced by SH2B1beta. Cells were transiently transfected with the pm674luc or pCD1luc reporters, along the reported cDNAs, their activation evaluated 36 h post-transfection and normalized to the activity of cotransfected pRL-TKluc (RLU, relative luciferase units). Data represent the meansplusminuss.d. of three wells. (c) SHC recruitment to RET and phosphorylation are increased by SH2B1beta. Endogenous SHC was isolated by immunoprecipitation from 293T cells transfected with RET/PTC2 and SH2B1beta, and RET/SHC phosphorylation evaluated by alphapTyr.

Full figure and legend (168K)

Further, to assess whether increased RET phosphorylation by SH2B1beta may positively modulate recruitment of effectors, we assayed the phosphorylation of Shc, a key transducer of RTKs, that associates to RET through the Y1062 docking site (Arighi et al., 1997). Shc was immunoprecipitated from cells transfected with either RET/PTC2 or RET/PTC2Y586F, in combination with SH2B1beta, and the Tyr-phosphorylation of the immunocomplexes was evaluated. The SH2B1beta-Y439F/Y494F (FF) double mutant was included to test whether these tyrosines, required for the biological activity of JAK/and PDGFR/SH2B1beta complexes (O'Brien et al., 2003), and putative binding sites for the SHC SH2 domain (YXXL/I motif) (Ravichandran, 2001), may influence the binding of RET/SH2B1beta to Shc. As expected, RET/PTC2, but not its Y586F mutant (devoid of Shc docking site, Y1062 on RET), binds to and phosphorylates Shc isoforms (Figure 4c). Consistent with its positive effect on RET, Shc phosphorylation was significantly enhanced by SH2B1beta-wt coexpression, an effect also observed with the FF mutant. This result thus excludes the requirement of these residues for the SH2B1beta effect on Shc recruitment. Intriguingly, the inability of Y586F mutant to recruit Shc was partly reversed by SH2B1beta, thus suggesting that RET/SH2B1beta interaction may unveil additional docking sites yet to be explored.

A further evidence of positive effect of SH2B1beta on RET-induced signaling comes from comparative proteomic studies performed on lysates of RET/PTC2- and PTC2/SH2B1beta-transfected cells. Proteins immunopurified with immobilized anti-pTyr antibody were analysed by peptide mass fingerprinting (Gorla et al., 2006), and 29 phosphorylated proteins, including transducers, transcription factors, proteins involved in cytoskeleton organization and in RNA metabolism, were identified as specifically phosphorylated by RET/PTC2, mostly representing novel substrates and known transducers phosphorylated by RET. The same phosphoproteins showed a significantly increased staining intensity in cells coexpressing RET/PTC2 and SH2B1beta (Supplementary Figure 3S). These results suggest a quantitative effect of SH2B1beta on RET-induced phosphoproteome, without altering its specificity.

SH2B1beta enhances RET-mediated biological activity

RET/PTC oncogenes, isolated as transforming genes from thyroid cancers (Alberti et al., 2003), are exclusive to PTC. By evaluating SH2B1beta expression in a thyroid environment, we found that equivalent levels of SH2B1 mRNA are expressed in RET/PTC1-infected human thyrocytes and in their parental cells (microarray data deposited in the ArrayExpress database, E-MEXP-429). In addition, the specific expression of the beta-isoform was confirmed by reverse transcription–polymerase chain reaction (RT–PCR) in thyroid tumorigenic and matched normal tissues (not shown). To determine whether SH2B1beta impacts on RET-induced transformation, we stably cotransfected NIH3T3 cells with RET/PTC2 and SH2B1beta, and isolated colonies expressing distinct combinations of RET and the adaptor. Among them, we chose one SH2B1beta-positive clone (S1); one clone expressing high RET/PTC2 level (R1); and three clones expressing low RET/PTC2 amounts, but distinct level of the adaptor: R2, with no SH2B1beta; RS1, showing intermediate SH2B1beta level; RS2, with the most adaptor protein (Figure 5a). The corresponding cells were plated for growth in soft agar, and transformants were thus detected (Figure 5b). S1 was unable to form colonies of more than 8–16 cells, as parental NIH3T3, whereas RET/PTC2 and RET/SH2B1beta coexpressing cells formed colonies of more than 50 cells. As expected, the colony size in R clones was related to the oncogene levels. Interestingly, RS1 showing equivalent oncogene expression with the R2 clone and intermediate SH2B1beta, exhibited slightly larger colonies. Consistently, RS2 clone showing higher SH2B1beta expression than RS1, resulted in considerable large colonies spreading through the matrix. Overall, the size of transformed colonies on soft-agar parallels the extent of oncogene phosphorylation, which is augmented by SH2B1beta in the RS clones. Thus, coexpression of the adaptor protein potentiates RET/PTC2 transforming ability by enhancing its autophosphorylation.

Figure 5.
Figure 5 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact help@nature.com or the author

Clonogenic activity in soft agar of NIH3T3 clones expressing RET/PTC2 and SH2B1beta. (a) Biochemical analysis of NIH3T3 clones. RET/PTC2 and SH2B1beta expression levels and oncogene Tyr phosphorylation of different NIH3T3 clones were analysed by immunoblotting. Actin is shown as a control for protein loading. (b) Colony formation ability in soft agar of the NIH3T3 clones stably expressing RET/PTC2 and different amounts of SH2B1beta shown in (a). Transfectants were seeded in soft agar, and colonies were stained and photographed 2 weeks after plating. Representative plates are shown. S, SH2B1beta transfectants; R, RET/PTC2 transfectants; RS, RET/PTC2 and SH2B1beta coexpressing clones.

Full figure and legend (96K)

SH2B1beta has been reported to enhance the neuronal differentiation of rat PC12 pheochromocytoma induced by the activated NGF receptor TrkA (Rui and Carter-Su, 1999). Since activated RET kinases also cause differentiation of PC12 cells (Borrello et al., 1995), we assessed the effect of SH2B1beta on RET-induced neurite elongation. PC12 cells, expressing endogenous proto-RET and transfected with its GFR1alpha coreceptor exhibited a discrete neuronal differentiation by GDNF, as assessed by both neurite outgrowth and activation of the pvgf8luc reporter (a marker of neuronal differentiation upstream of the luciferase gene) (Canu et al., 1997), and a significant enhancement of the number of differentiated cells and neuritis lengths by SH2B1beta (Supplementary Figure 4S). This effect, which is in accordance with what previously reported (Zhang et al., 2006), was best evaluated by the higher vgf8 response. Moreover, as GDNF can signal independently of RET through GFRalpha, we tested the effect of constitutively active RETC634R (associated to MEN2A) and RET/PTC2 on GDNF-independent differentiation. The enhancing effect of SH2B1beta was confirmed, as assessed by luciferase activity (Figure 6a), and activation of RET downstream effectors was also augmented by the adaptor. In fact, pSTAT3 and pSTAT1 levels are markedly enhanced in cells coexpressing RET/PTC2 and SH2B1beta-wt, as well as pERK1/2 downstream to Ras/MAPK pathway (Figure 6b). Consistent with previous experiments, RET expression was found to be upregulated in PC12 cells cotransfected with the adaptor, which might contribute to the overall differentiation enhanced by the RET/adaptor complex.

Figure 6.
Figure 6 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact help@nature.com or the author

RET/PTC2- and RET/MEN2A-induced differentiation of PC12 pheochromocytoma cells is enhanced by SH2B1beta. (a) PC12 cells, seeded at 200 000 cell/well (12-wells plates), were transfected with pvgf8luc reporter vector in combination with the indicated constructs. Luciferase activity was measured 48 h post-transfection. The data represent the mean valuesplusminuss.d. of triplicate samples. (b) Analysis of RET-driven signaling in PC12 cells transfected with RET/PTC2 and SH2B1beta.

Full figure and legend (81K)

As SH2B1beta in vitro binding to RET is affected by specific Y to F mutations (see Figure 2), we investigated whether mutations at RET docking sites may downregulate the effect of SH2B1beta on PC12. RET/PTC2 short and long isoforms stimulated comparable differentiation, as measured by luciferase, and SH2B1beta coexpression increased this activity by approximately fourfold (Figure 7a). While RET/PTC2 mutants RET/PTC2-Y505F, -Y539F and -Y505F/Y539F induced differentiating effect comparable with the wild-type kinase, others, as RET/PTC2-Y276F, -Y586F, -Y424F, -Y429F, -Y424F/Y429F and -Y429F/Y505F, as well as the negative control mutant -K282R, displayed lower or no differentiating activity on PC12. Coexpression of SH2B1beta stimulated significantly (from four- to sixfold) the pvgf8luc response associated to the single mutants, and to RET/PTC2Y424F/Y429F and RET/PTC2Y505F/Y539F double mutants, while the RET/PTC2Y429F/Y505F-driven differentiation was negligibly affected. As expected, the activity of kinase-dead K282R control was unaffected by the adaptor. These data reveal that either Y429 or Y505 residues (Y905 and Y981 in proto-RET) are required for functional SH2B1beta recruitment, and appear to mediate its effect on RET-induced biological activity. Accordingly, as shown in Figure 7c, ERK1/2 phosphorylation elicited by RET/PTC2 is enhanced by SH2B1beta in RET/PTC2Y505F and, at lesser extent, in RET/PTC2Y429F single mutants cotrasfection experiments, as in -wt and -Y586F controls. The enhancing effect on ERKs is instead impaired by mutations at both Y429 and Y505 residues of the oncogene.

Figure 7.
Figure 7 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact help@nature.com or the author

Mutations at both Y429 and Y505 inhibit the ability of SH2B1beta to enhance RET/PTC2-triggered neuronal differentiation. (a) SH2B1beta augmentation of RET response on PC12 is impaired in RET/PTC2Y429F/Y505F mutant. Cells were transfected with pvgf8luc, SH2B1beta, and varying concentration of RET constructs. DNA was held constant (750 ng/well) by addition of empty pRc/CMV vector. Luciferase activity was measured 48 h post-transfection on triplicate samples. Iso51 and Iso9, long and short RET isoforms. (b) Enhancement by SH2B1beta of RET/PTC2-induced ERK phosphorylation is abolished by the Y429F-Y505F double mutation. Lysates from cells transiently transfected with RET/PTC2 and GFP-SH2B1beta were immunoblotted and the extent of ERK1/2 phosphorylation was visualized by alphapERK. (c) SH2B1beta contribution to RET-associated differentiation is not affected by its Y439F/Y494F mutations and Delta24–48 deletion. Transactivation of pvgf8luc reporter was assayed in PC12 cells (n=3) cotransfected with RET/PTC2 and mutants of the adaptor protein.

Full figure and legend (112K)

Finally, SH2B1beta mutations at Y439 and Y494 (required for regulation of JAKs- and PDGFR-triggered biological activities) (O'Brien et al., 2003) and Delta24–88 deletion (of the phenylalanine zipper DD) (Dhe-Paganon et al., 2004) were tested on PC12 differentiation. SH2-BbetaFF (see Figure 4c) and SH2-BbetaDelta24-88 proteins coimmunoprecipitate with and are phosphorylated by RET/PTC2, although to a lesser extent than the wild type (Supplementary Figure 5S). These mutants display an enhancing activity on RET/PTC2-driven differentiation comparable with the wild-type protein (Figure 7b), suggesting that both the above tyrosines and dimerization of the adaptor are not primarily involved in RET-mediated biological effects.



SH2B1beta, a splicing product of SH2B1 gene primarily identified as substrate of the TK JAK2 (Rui et al., 1997), has been recognized as an adaptor for several RTKs (Rui et al., 1997; Rui and Carter-Su, 1998; Wang and Riedel, 1998; Ahmed and Pillay, 2001; Qian and Ginty, 2001; Yousaf et al., 2001). We previously reported its identification, via yeast two-hybrid system, as a putative adaptor for both proto-RET and RET/PTC2 oncogene (Donatello et al., 2004). A clear demonstration that interaction of SH2B1beta with proto-RET is involved in signaling of GDNF-induced neuronal differentiation has been recently provided by He and co-workers (Zhang et al., 2006). The results reported in this paper, while supporting their findings, add novel information that highlights the importance of SH2B1beta in mediating RET-induced differentiation and neoplastic transformation, and gives some insights into the mechanisms underlying their interaction.

RET short and long isoforms, exhibiting distinct physiological role and common and isoform-specific transducers (de Graaff et al., 2001), as well as their thyroid carcinoma-associated oncogenic forms, incorporate in a fully active complex with SH21Bbeta through the adaptor's SH2 domain. The active autophosphorylated RET is required for this interaction to occur. As a consequence, SH2B1beta undergoes phosphorylation and RET displays increased phosphorylation and downstream signaling activation. Therefore, RET functions, including neuronal differentiation and cellular transformation, are positively modulated by SH2B1beta. These findings may contribute to our understanding of RET signaling in development and carcinogenesis. RET mediates, in fact, several biological processes, including the development of the enteric nervous system, kidney organogenesis and spermatogenesis (Arighi et al., 2005). Moreover, RET oncogenic activation is a crucial initiating event in thyroid cancers. Thus, the demonstration that SH2B1beta is a binding partner for proto-RET, RET/PTC, and MEN2A- and MEN2B-associated oncoproteins, reveals this adaptor as a critical player of RET-driven signaling. Furthermore, similar to the SH2B1beta's potentiation of specific pathways activated by NGF or insulin (Qian and Ginty, 2001; Ahmed and Pillay, 2003), the upregulation by SH2B1beta of RET-activated downstream pathways strongly supports the idea of a positive regulatory mechanism for this adaptor on RET activity. We have extended this concept by demonstrating that the RET/PTC-associated phosphoproteome is quantitatively enhanced by SH2B1beta coexpression, in agreement with the increased TK activity of the adaptor/RET complex. Accordingly, the finding that the GDNF-dependent and -independent RET-induced neuronal differentiation is markedly enhanced by SH2B1beta coexpression in PC12 cells (Zhang et al., 2006) and our results), appears a consequence of the general SH2B1beta-mediated upregulation of RET signaling. This is consistent with the previously demonstrated effect on NGF-induced TRK phosphorylation and neurite outgrowth (Rui and Carter-Su, 1999; Qian and Ginty, 2001). Interestingly, transgenic mice expressing DN-RET display defective early spermatogenesis (Arighi et al., 2005), a phenotype partly overlapping the impaired fertility of SH2B1-null mice (Ohtsuka et al, 2002; Duan et al., 2004). Whether RET/SH2B1beta complex is involved in spermatogenesis is an interesting point to be addressed.

The binding requirements for SH2B1beta are yet to be fully elucidated. It has been reported that tyrosines within pYXXL/M motif of TK domains are putative binding sites for SH2B1beta (Kong et al., 2002; Kurzer et al., 2004). By addressing Y to F mutations in RET cytoplasmic region (see Figure 2a for comparison of tyrosine residues between proto-RET and RET/PTC2), we identified Y905 (within YVKR motif) and Y981 (YRLM motif) as crucial sites for SH21Bbeta binding. Mutations at either Y905 or Y981 significantly reduce RET binding to the adaptor in pull-down experiments. However, they appear to complement in intact cells, and the interaction is prevented only when both residues are mutated. Consistently, SH2B1beta shows negligible effect on the poor neuronal differentiation induced by the RETY905F/Y981F double mutant, while significantly enhancing the differentiating activity associated to single mutants. Our finding of Y905 and Y981 as a binding sites for the adaptor is partly in contrast with the previous observation that led to propose RET-Y981 as unique SH2B1beta-docking site (Zhang et al., 2006). Thus, our results are consistent with previous finding showing SH2B1beta binding to pYXXL/M motif, and support the view that conserved tyrosines in the autocatalytic loop may be also required in order for RTKs to recruit members of SH2B family. In fact, SH2B1alpha and -italic gamma, and SH2B2 adaptors have been demonstrated to bind to phosphotyrosines within the YETD and YYRK motifs of the IR's activation loop (Kotani et al., 1998; Nelms et al., 1999; Hu et al., 2003), and evolutionarily conserved in TrkA and IGF1R. RET-Y905 lies into YVKR sequence exhibiting elevate homology with an IR's motif. Although the poor binding of SH2B1beta to RET-Y905F in vitro might also reflect its low tyrosine phosphorylation, a role as docking site for the activation-loop's tyrosine residues is demonstrated. Therefore, Y905 autocatalytic tyrosine may contribute both directly and indirectly to this binding.

Finally, the neuronal differentiation observed following SH2B1beta coexpression, over the one associated to RET devoid of Y1062 multidocking site, is consistent with the binding of the adaptor to this mutant, and with the recruitment of SHC adaptor (required for RTK-induced differentiation) to this complex. Whether SH2B1beta's phosphorylatable tyrosine residues are directly involved in the recruitment of SHC to its complex with RET1062F is an interesting aspect to be investigated. Moreover, as RET loss-of-function alterations associated with HSCR disrupt the binding of transducer to the multidocking site (Arighi et al., 2005), it would be interesting to test whether SH2B1beta could restore these partially impaired signals.

Clustering of SH2B proteins to RTKs increases their transphosphorylation and activity (Qian and Ginty, 2001). Mutants deleted of the phenylalanine-zipper DD in the SH2B N-terminus (Dhe-Paganon et al., 2004) are unable to enhance NGF-mediated PC12 differentiation (Qian and Ginty, 2001). We show that SH2B1beta devoid of this domain, SH2B1betaDelta24-88, exhibits reduced phosphorylation by RET/PTC2 and coprecipitation, but still enhances oncogene autophosphorylation and biological activity. This appears in agreement with previous results showing that the adaptor's SH2 domain binds to active phosphorylated TKs and enhances their autophosphorylation (Rui et al., 2000; Kurzer et al., 2004), which might suggest a regulatory mechanism for the N-terminal portion of SH2B1beta. Further investigations would help to elucidate the contribution of this domain to RET signaling.

A final important question is the impact of RET/SH2B1beta functional interaction in neoplastic processes of PTC-associated RET oncogenes. By testing the transforming ability of NIH3T3 cells cotransfected with RET/PTC2 and SH2B1beta, we could demonstrate that growth of RET/PTC2-expressing clones on soft agar is enhanced by SH2B1beta. In fact, cells expressing equivalent RET/PTC2 amounts show a direct relationship between RET autophosphorylation, the colony size and the level of the adaptor protein, thus suggesting a role for the adaptor in RET/PTC-driven transformation. This is further supported by the observed expression of SH2B1beta in primary thyrocytes, the human cell type from which PTC arises through RET/PTC oncogenes (data not shown).

Taken together, by demonstrating that SH2B1beta enhances RET-induced signaling and biological activities, we provide evidence and extend the concept that this adaptor is a key regulator exerting its action primarily on RTKs. Our studies give some insights into the understanding of the mechanisms by which SH21Bbeta upregulates RET activity: (i) by binding to RET-Y905 and -Y981; and (ii) by preventing RET dephosphorylation by SHP-1 tyrosine phosphatase. We may hypothesize that SH2B1beta, by protecting autocatalytic tyrosines from protein tyrosine phosphatases, enhances the steady-state autophosphorylation of RET and potentiates its catalytic activity. SHP-1, expressed in thyroid tissues, has been in fact shown to dephosphorylate RET, with consequent reduction of the transforming potential associated with RETMEN2A oncogene (Hennige et al., 2001). Moreover, we have shown that SH2B1beta binding to RET even hampers the action of TK inhibitors.

In conclusion, our investigation supports a role for SH2B1beta as a key enhancer of RET activity. Since both loss- and gain-of-function pathogenic RET mutations are involved in human diseases, the possibility to enhance RET residual signals or to attenuate RET-mediated oncogenic processes through SH2B1beta adaptor may represent an intriguing possibility to address novel therapeutic strategies.


Materials and methods

Plasmid construction

The MYC-tagged full-length SH2B1beta cDNA cloned in pRK5- vector was kindly provided by C Carter-Su (University of Michigan, Ann Arbor, MI, USA) (Rui and Carter-Su, 1999). The GFP-tagged cDNA was derived by EcoRI-XhoI digestion and subcloning into pEGFPC3 vector (Clontech, Mountain View, CA, USA). The corresponding R555E, Y439F, Y494F, Y439F/Y494F (SH2B1betaFF) and Delta24–88 mutants were obtained by site-directed mutagenesis using the Quick Change site-directed mutagenesis XL kit (Stratagene, La Jolla, CA, USA). The GST-SH2B1beta and -SH2B1betaR555E expression vectors were obtained by BamHI/EcoRI excision of the pRK5-SH2B1beta plasmids and subcloning into pGEX4T2. Cloning of proto-RET-9, proto-RET-51, RET/PTC2–9, RET/PTC2–51 and RETC634R cDNAs into pRC/CMV mammalian expression vector was previously reported (Borrello et al., 1995; Arighi et al., 1997; Lorenzo et al., 1997). RET/PTC2Y586F, Y505F, Y539F, Y505F/Y539F and K282R mutants were described earlier (Borrello et al., 1996; Arighi et al., 1997). RET/PTC2Y276F, Y424F, Y429F, Y452F, Y424F/Y429F, Y424D/Y429E (RET/PTC2DE) and Y429F/Y505F mutants were obtained by site-direct mutagenesis. The RET/PTC2N538 construct was obtained by PCR amplification of the 1–538 coding region and subcloned via HindIII/XbaI into pRC/CMV. The pcDNA3hygro-RET/PTC2–9 construct was generated by XbaI excision from its pRC/CMV homolog and subcloning into the recipient plasmid. All mutant constructs were confirmed by sequencing. The pRK5RS-SHP-1 expression vector was provided by A Ullrich (Max-Planck-Institut, Martinsried, Germany), and its SHP-1C455S derivative previously reported (Roccato et al., 2005). The pcDNA3-GFRalpha1 plasmid was from DP Smith (University of Cambridge, Cambridge, UK). The FLAG-tagged STAT3 constructs in pRc/CMV (Bromberg et al., 1998) were a gift of J Darnell (The Rockfeller University, New York, NY, USA).

Antibodies and other reagents

Antisera to RET (common I and II), raised against C-terminal epitopes were reported (Borrello et al., 1996). Antibodies specific to RET-9 (C-19) and to RET-51 (C-20) isoform, to proto-RET (H-300), and to GFP (B-8) were from Santa Cruz Biotechnology (Santa Cruz, CA, USA). Anti-pTyr 4G10 mouse monoclonal antibody and anti-SHC were from Upstate Biotechnology (Charlottesville, VA, USA). Antibodies to phospho-ERK1/2 and total ERK1/2, to pSTAT3 and STAT3, to pSTAT1 and STAT1 were from Cell Signaling (Danvers, MA, USA). Anti-MYC (9E10) and anti-FLAG monoclonal antibodies were respectively from Calbiochem (San Diego, CA, USA) and from Sigma (St Louis, MO, USA). Goat anti-GFP antibody was from Rockland (Gilbertsville, PA, USA). GDNF was from Alomone Labs (Jerusalem, Israel). Sepharose-protein A, Sepharose-protein A/G and Sepharose-glutathione beads, HRP-linked anti-IgG secondary antibodies, ECL reagents, and [italic gamma32-P]ATP (10 mCi/ml) were from Amersham Bioscience (Piscataway, NJ, USA).

Yeast two-hybrid screening

The full-length RET/PTC2 and RET(712–1071) were subcloned into the NdeI/SalI sites of pAS2–1 vector, downstream to the GAL4-DBD (Clontech). Human Placenta Matchmaker cDNA library, cloned into the EcoRI/XhoI sites of pACT2 yeast expression vector, was screened with the RET/PTC2 construct as a bait, according to the Matchmaker GAL4 Two-Hybrid System protocol (Clontech). Positive colonies were rescreened with the pAS2–1-RET(712–1071) construct, and prey fragments of positive clones were sequenced at both 5'- and 3'-end for identification.

Cell culture

293T embryo kidney and HeLa cervical carcinoma cells were grown on Dulbecco's modified Eagle's medium-based media plus 10% fetal bovine serum. PC12 rat pheochromocytoma cells were cultured in RPMI medium supplemented with 10% horse serum and 5% FBS. PC12 and 293T were maintained at 37°C in 5% humidified CO2 atmosphere, HeLa and NIH3T3 in 10% CO2. Before experiments, PC12 cells were seeded at 50 000 cells/cm2 density on multi-well plates previously coated with Collagen type-IV (BD Bioscience, Franklin Lakes, NJ, USA).

Transient and stable transfections and soft-agar assay

293T cells seeded into 10-cm plates were cotransfected by the CaPO4 precipitation method. Multi-well plates were transfected by Fugene6 lipid transfectant (Roche, Basel, Switzerland). Proteins were extracted 36–48 h post-transfection. Stable NIH3T3 cotransfectants expressing RET/PTC2 and GFP-SH2B1beta gene, or combinations with their empty vectors, were obtained via Lipofectamine 2000 (Invitrogen, Carlsbad, CA, USA) transfection of subconfluent cells with 2 mug each of plasmids carrying neomycin (pEGFPC3-SH2B1beta) and hygromycin (pcDNA3hygro-RET/PTC2–9) resistance genes. Clones were isolated by G418 (0.5 mg/ml) and hygromicin (50 mug/ml) selection, and tested for their ability to growth in soft agar. Briefly, cells were suspended in complete selective medium containing 0.33% (w/v) Type-VII agarose (Sigma) and overlaid onto previously solidified 0.5% agarose in complete medium. Following two weeks culture, colonies were stained with 1 mg/ml p-iodonitotetrazolium-violet (Sigma), counted and photographed.

Immunoprecipitation, immunokinase and pull-down assays

Cultured cells were lysed in ice-cold PLCLB buffer (50 mM (4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid) (HEPES), pH 7.5, 150 mM NaCl, 1.5 mM MgCl2, 1 mM ethylene glycol-bis(beta-aminoethyl ether)-N,N,N',N'-tetraacetic acid, 10 mM Na4P2O7, 50 mM NaF, 10% glycerol, 1% Triton X-100) supplemented with protease/phosphatase inhibitors. Cleared cell lysates (approximately1 mg protein) were incubated with the indicated antibodies and Sepharose-protein A/G beads, rotated for 3–4 h at 4°C, washed three times with HNTG buffer (20 mM HEPES, pH 7.5, 140 mM NaCl, 10% glycerol, 0.2% Triton X-100) plus protease/phosphatase inhibitors and boiled 5 min in sodium dodecyl sulfate (SDS) sample buffer. The solubilized proteins were separated by SDS–polyacrylamide gel electrophoresis (PAGE) on polyacrylamide gels, western blotted with selected antibodies and detected by ECL. Lysate aliquots (40 mug) were also immunoblotted to evaluate the level of specific protein expression. Immunokinase assays were performed on cell lysates incubated with antibody to RET-9 (C-19) and collected on Sepharose-protein A beads after rotation (3 h, 4°C), two washes with HNTG buffer, and two with kinase buffer (50 mM HEPES pH 7.2, 1 mM MgCl2, 5 mM MnCl2, 1 mM dithiothreitol, 0.2 mM phenylmethanesulfonylfluoride, 0.2 mM Na2VO7). The reaction was carried out in 30 mul of kinase assay buffer containing 2 muM ATP, 50 muM MBP and 2.5 muCi of [italic gamma32-P]ATP, and proceeded for 20 min at 4°C before stopping with 10 mul of 4 times SDS sample buffer and boiling. Samples were separated on 12% polyacrylamide gel, transferred onto nitrocellulose membrane, and radioactive blots exposed to phosphorscreen (32P-labelled bands were quantified by densitometric scanning on Thyphoon 8600 (Amersham) and normalized to the corresponding RET protein). To evaluate the in vitro phosphorylation of SH2B1beta protein by RET/PTC2, the kinase reactions were carried out in the kinase assay buffer added with 2 muM ATP, 5 mM reduced glutathione, 2 mug of resin-bound GST-SH2B1beta and 2.5 muCi of [italic gamma32-P]ATP. Following separation on 7.5% gel and blotting onto membranes, the radioactive blots exposed to phosphorscreen. Pull-down experiments were performed by using GST-SH2B1beta protein and its R555E mutant counterpart. Briefly, RET-transfected 293T cells were lysed in TNE-T 1% lysis buffer (50 mM Tris, pH 7.4, 140 mM NaCl, 2 mM ethylenediaminetetraacetic acid, 40 mM NaF, 1% Triton X-100 plus protease/phosphatase inhibitors), rotated 4 h on immobilized GST-SH2B1beta beads, washed twice with TNE-T 0.5%, and immunoblotted for detection of bound RET proteins, in parallel with the respective inputs.

Proteomic analysis

Lysates from transfected 293T cells were pooled (approx10 mg of protein) and precleared with Sepharose-protein A. Phosphorylated proteins were affinity-purified with agarose-coniugated antibody to pTyr (Sigma). Following 3 h incubation, the beads were washed trice with HNTG buffer, boiled in SDS sample buffer, and the eluted phosphoproteins resolved by SDS–PAGE (4–12% precast NuPAGE gel, Invitrogen) before standard silver staining.

Protein bands were picked and treated overnight with sequencing-grade modified trypsin (Promega, Madison, WI, USA) and identified by peptide mass fingerprint method. MALDI-TOF mass spectrometry analysis was performed on a Voyager-DE STR (Applied Biosystems), and spectra analysed with ProFound software (http://www.expasy.org/tools/ProFound) searching in nonredundant protein database (NCBI). Peptide mass tolerance was plusminus20 p.p.m.

Reporter assay

PC12 and HeLa cells, seeded on 12-wells plates in complete medium with no antibiotics, were transfected respectively via Cellfectin (Invitrogen) and Fugene6, with combination of reporter vectors, pRL-TK co-reporter, RET and SH2B1beta constructs. pvgf8luc (Canu et al., 1997), pCD1luc (Lee et al., 1999) and pm674TATAluc (Besser et al., 1999) reporters were respectively provided by R Possenti (Tor Vergata University, Rome, Italy), R Pestell (Georgetown University, Washington, DC, USA) and J Darnell. At 12–16 h post-transfection, PC12 cells were treated with GDNF and left 48 h before neurite outgrowth and reporter activity were evaluated. Luciferase activity was measured with the dual-luciferase assay kit (Promega). The activity associated to the reporter plasmids was normalized to the activity of the Renilla luciferase.



  1. Ahmed Z, Pillay TS. (2001). Functional effects of APS and SH2-B on insulin receptor signalling. Biochem Soc Trans 29: 529–534. | Article | PubMed | ISI | ChemPort |
  2. Ahmed Z, Pillay TS. (2003). Adapter protein with a pleckstrin homology (PH) and an Src homology 2 (SH2) domain (APS) and SH2-B enhance insulin-receptor autophosphorylation, extracellular-signal-regulated kinase and phosphoinositide 3-kinase-dependent signalling. Biochem J 371: 405–412. | Article | PubMed | ISI | ChemPort |
  3. Airaksinen MS, Saarma M. (2002). The GDNF family: signalling, biological functions and therapeutic value. Nat Rev Neurosci 3: 383–394. | Article | PubMed | ISI | ChemPort |
  4. Alberti L, Carniti C, Miranda C, Roccato E, Pierotti MA. (2003). RET and NTRK1 Proto-Oncogenes in Human Diseases. J Cell Physiol 195: 168–186. | Article | PubMed | ChemPort |
  5. Arighi E, Alberti L, Torriti F, Ghizzoni S, Rizzetti MG, Pelicci G et al. (1997). Identification of SHC docking site on Ret tyrosine kinase. Oncogene 14: 773–782. | Article | PubMed | ISI | ChemPort |
  6. Arighi E, Borrello MG, Sariola H. (2005). RET tyrosine kinase signaling in development and cancer. Cytokine Growth F R 16: 441–467. | Article | ISI | ChemPort |
  7. Besser D, Bromberg JF, Darnell Jr JE, Hanafusa H. (1999). A single amino acid substitution in the v-Eyk intracellular domain results in activation of Stat3 and enhances cellular transformation. Mol Cell Biol 19: 1401–1409. | PubMed | ISI | ChemPort |
  8. Bongarzone I, Monzini N, Borrello MG, Carcano C, Ferraresi G, Arighi E et al. (1993). Molecular characterization of a thyroid tumor-specific transforming sequence formed by the fusion of ret tyrosine kinase and the regulatory subunit RIa of cyclic AMP-dependent protein kinase A. Mol Cell Biol 13: 358–366. | PubMed | ISI | ChemPort |
  9. Borrello MG, Alberti L, Arighi E, Bongarzone I, Battistini C, Bardelli A et al. (1996). The full oncogenic activity of Ret/ptc2 depends on tyrosine 539, a docking site for phospholipase Cgamma. Mol Cell Biol 16: 2151–2163. | PubMed | ISI | ChemPort |
  10. Borrello MG, Smith DP, Pasini B, Bongarzone I, Greco A, Lorenzo MJ et al. (1995). RET activation by germline MEN2A and MEN2B mutations. Oncogene 11: 2419–2427. | PubMed | ISI | ChemPort |
  11. Bromberg JF, Horvath CM, Besser D, Lathem WW, Darnell Jr JE. (1998). Stat3 activation is required for cellular transformation by v-src. Mol Cell Biol 18: 2553–2558. | PubMed | ISI | ChemPort |
  12. Canu N, Possenti R, Rinaldi AM, Trani E, Levi A. (1997). Molecular cloning and characterization of the human VGF promoter region. J Neurochem 68: 1390–1399. | PubMed | ISI | ChemPort |
  13. Carlomagno F, Vitagliano D, Guida T, Napolitano M, Vecchio G, Fusco A et al. (2002). The kinase inhibitor PP1 blocks tumorigenesis induced by RET oncogenes. Cancer Res 62: 1077–1082. | PubMed | ISI | ChemPort |
  14. Cuccuru G, Lanzi C, Cassinelli G, Pratesi G, Tortoreto M, Petrangolini G et al. (2004). Cellular effects and antitumor activity of RET inhibitor RPI-1 on MEN2A-associated medullary thyroid carcinoma. J Natl Cancer Inst 96: 1006–1014. | PubMed | ChemPort |
  15. de Graaff E, Srinivas S, KilKenny C, D'Agati V, Mankoo BS, Costantini F et al. (2001). Differential activities of the RET tyrosine kinase receptor isoforms during mammalian embryogenesis. Genes Dev 15: 2433–2444. | Article | PubMed | ISI | ChemPort |
  16. Dhe-Paganon S, Werner ED, Nishi M, Hansen L, Chi YI, Shoelson SE. (2004). A phenylalanine zipper mediates APS dimerization. Nat Struct Mol Biol 11: 968–974. | Article | PubMed | ChemPort |
  17. Donatello S, Alberti L, Fiorino A, Degl'Innocenti D, Rizzetti MG, Gorla L et al. (2004). Identification of SH2-Bbeta as a RET adaptor protein. Tumori 4: 122. | Article |
  18. Duan C, Yang H, White MF, Rui L. (2004). Disruption of the SH2-B gene causes age-dependent insulin resistance and glucose intolerance. Mol Cell Biol 24: 7435–7443. | Article | PubMed | ISI | ChemPort |
  19. Gorla L, Cantu M, Micciche' F, Patelli C, Mondellini P, Pierotti MA et al. (2006). Ret oncoproteins induce tyrosine phosphorylation changes of proteins involved in RNA metabolism. Cell Signal 18: 2272–2282. | Article | PubMed | ISI | ChemPort |
  20. Gryz EA, Meakin SO. (2000). Acidic substitution of the activation loop tyrosines in TrkA supports nerve growth factor-independent cell survival and neuronal differentiation. Oncogene 19: 417–430. | Article | PubMed | ISI | ChemPort |
  21. Hennige AM, Lammers R, Hoppner W, Arlt D, Strack V, Teichmann R et al. (2001). Inhibition of Ret oncogene activity by the protein tyrosine phosphatase SHP1. Endocrinology 142: 4441–4447. | Article | PubMed | ISI | ChemPort |
  22. Hu J, Liu J, Ghirlando R, Saltiel AR, Hubbard SR. (2003). Structural basis for recruitment of the adaptor protein APS to the activated insulin receptor. Mol Cell 12: 1379–1389. | Article | PubMed | ISI | ChemPort |
  23. Ichihara M, Murakumo Y, Takahashi M. (2004). RET and neuroendocrine tumors. Cancer Lett 204: 197–211. | Article | PubMed | ISI | ChemPort |
  24. Kawamoto Y, Takeda K, Okuno Y, Yamakawa Y, Ito Y, Taguchi R et al. (2004). Identification of RET autophosphorylation sites by mass spectrometry. J Biol Chem 279: 14213–14224. | Article | PubMed | ISI | ChemPort |
  25. Kong M, Wang CS, Donoghue DJ. (2002). Interaction of fibroblast growth factor receptor 3 and the adapter protein SH2-B. A role in STAT5 activation. J Biol Chem 277: 15962–15970. | Article | PubMed | ISI | ChemPort |
  26. Kotani K, Wilden P, Pillay TS. (1998). SH2-Balpha is an insulin-receptor adapter protein and substrate that interacts with the activation loop of the insulin-receptor kinase. Biochem J 335: 103–109. | PubMed | ISI | ChemPort |
  27. Kurzer JH, Argetsinger LS, Zhou YJ, Kouadio JL, O'Shea JJ, Carter-Su C. (2004). Tyrosine 813 is a site of JAK2 autophosphorylation critical for activation of JAK2 by SH2-B beta. Mol Cell Biol 24: 4557–4570. | Article | PubMed | ISI | ChemPort |
  28. Lee RJ, Albanese C, Stenger RJ, Watanabe G, Inghirami G, Haines GK et al. (1999). pp60(v-src) induction of cyclin D1 requires collaborative interactions between the extracellular signal-regulated kinase, p38, and Jun kinase pathways. A role for cAMP response element-binding protein and activating transcription factor-2 in pp60(v-src) signaling in breast cancer cells. J Biol Chem 274: 7341–7350. | Article | PubMed | ISI | ChemPort |
  29. Lorenzo MJ, Gish GD, Houghton C, Stonehouse TJ, Pawson T, Ponder BAJ et al. (1997). RET alternate splicing influences the interaction of activated RET with the SH2 and PTB domains of Shc, and the SH2 domain of Grb2. Oncogene 14: 763–771. | Article | PubMed | ISI | ChemPort |
  30. Nelms K, O'Neill TJ, Li S, Hubbard SR, Gustafson TA, Paul WE. (1999). Alternative splicing, gene localization, and binding of SH2-B to the insulin receptor kinase domain. Mamm Genome 10: 1160–1167. | Article | PubMed | ISI | ChemPort |
  31. O'Brien KB, Argetsinger LS, Diakonova M, Carter-Su C. (2003). YXXL motifs in SH2-Bbeta are phosphorylated by JAK2, JAK1, and platelet-derived growth factor receptor and are required for membrane ruffling. J Biol Chem 278: 11970–11978. | Article | PubMed | ISI | ChemPort |
  32. Ohtsuka S, Takaki S, Iseki M, Miyoshi K, Nakagata N, Kataoka Y et al. (2002). SH2-B is required for both male and female reproduction. Mol Cell Biol 22: 3066–3077. | Article | PubMed | ISI | ChemPort |
  33. Osborne MA, Dalton S, Kochan JP. (1995). The yeast tribrid system-genetic detection of trans-phosphorylated ITAM-SH2-interactions. Biotechnology 13: 1474–1478. | Article | PubMed | ChemPort |
  34. Qian X, Ginty DD. (2001). SH2-B and APS are multimeric adapters that augment trkA signaling. Mol Cell Biol 21: 1613–1620. | Article | PubMed | ISI | ChemPort |
  35. Qian X, Riccio A, Zhang Y, Ginty DD. (1998). Identification and characterization of novel substrates of Trk receptors in developing neurons. Neuron 21: 1017–1029. | Article | PubMed | ISI | ChemPort |
  36. Ravichandran KS. (2001). Signaling via Shc family adapter proteins. Oncogene 20: 6322–6330. | Article | PubMed | ISI | ChemPort |
  37. Ren D, Li M, Duan C, Rui L. (2005). Identification of SH2-B as a key regulator of leptin sensitivity, energy balance, and body weight in mice. Cell Metab 2: 95–104. | Article | PubMed | ISI | ChemPort |
  38. Riedel H, Wang J, Hansen H, Yousaf N. (1997). PSM, an insulin-dependent, pro-rich, PH, SH2 domain containing partner of the insulin receptor. J Biochem (Tokyo) 122: 1105–1113. | PubMed | ChemPort |
  39. Roccato E, Miranda C, Raho G, Pagliardini S, Pierotti MA, Greco A. (2005). Analysis of SHP-1 mediated down-regulation of the TRK-T3 oncoprotein identifies TFG as a novel SHP-1 interacting protein. J Biol Chem 280: 3382–3389. | Article | PubMed | ISI | ChemPort |
  40. Rui L, Carter-Su C. (1998). Platelet-derived growth factor (PDGF) stimulates the association of SH2-Bbeta with PDGF receptor and phosphorylation of SH2-Bbeta. J Biol Chem 273: 21239–21245. | Article | PubMed | ISI | ChemPort |
  41. Rui L, Carter-Su C. (1999). Identification of SH2-bbeta as a potent cytoplasmic activator of the tyrosine kinase Janus kinase 2. Proc Natl Acad Sci USA 96: 7172–7177. | Article | PubMed | ChemPort |
  42. Rui L, Gunter DR, Herrington J, Carter-Su C. (2000). Differential binding to and regulation of JAK2 by the SH2 domain and N-terminal region of SH2-Bbeta. Mol Cell Biol 20: 3168–3177. | Article | PubMed | ISI | ChemPort |
  43. Rui L, Mathews LS, Hotta K, Gustafson TA, Carter-Su C. (1997). Identification of SH2-Bbeta as a substrate of the tyrosine kinase JAK2 involved in growth hormone signaling. Mol Cell Biol 17: 6633–6644. | PubMed | ISI | ChemPort |
  44. Takahashi M. (2001). The GDNF/RET signaling pathway and human diseases. Cytokine Growth F 12: 361–373. | Article | ISI | ChemPort |
  45. Wang J, Riedel H. (1998). Insulin-like growth factor-I receptor and insulin receptor association with a Src homology-2 domain-containing putative adapter. J Biol Chem 273: 3136–3139. | Article | PubMed | ISI | ChemPort |
  46. Yousaf N, Deng Y, Kang Y, Riedel H. (2001). Four PSM/SH2-B alternative splice variants and their differential roles in mitogenesis. J Biol Chem 276: 40940–40948. | Article | PubMed | ISI | ChemPort |
  47. Zhang Y, Zhu W, Wang YG, Liu XJ, Jiao L, Liu X et al. (2006). Interaction of SH2-B{beta} with RET is involved in signaling of GDNF-induced neurite outgrowth. J Cell Sci 119: 1666–1676. | Article | PubMed | ISI | ChemPort |


We are grateful to C Carter-Su (University of Michigan, Ann Arbor, MI, USA) for generously providing SH2B1beta cDNA, to J Darnell (The Rockfeller University, New York, NY, USA) for STAT3 and pm674TATAluc constructs, to A Ullrich (Max-Planck-Institut, Martinsried, Germany) for SHP-1 plasmid, to R Possenti (Tor Vergata University, Rome, Italy) for pvgf8luc, and R Pestell (Georgetown University, Washington, DC, USA) for pCD1luc reporter. We thank Piera Mondellini for SH2B1beta expression studies on thyroid tissues, Marco Cantù for MS analysis, Maria Teresa Radice for technical assistance, Vijay Kumar and Elena Arighi for critical reading of the article. This research was supported by the Associazione Italiana per la Ricerca sul Cancro (AIRC) and by the European Community's Sixth Framework Programme under the SIMAP Project.

Supplementary Information accompanies the paper on the Oncogene website (http://www.nature.com/onc).