The RET proto‐oncogene: A molecular therapeutic target in thyroid cancer

Abstract

The RET proto‐oncogene is responsible for the development of several human inherited and non‐inherited diseases. Germline point mutations were identified in multiple endocrine neoplasia types 2A and 2B, and familial medullary thyroid carcinoma. More than 10 rearranged forms of RET, referred to as RET/PTC 1–9, ELKS/RET and RFP/RET, have been cloned from sporadic and radiation‐associated papillary thyroid carcinomas. These mutations induced oncogenic activation of RET tyrosine kinase by different mechanisms. To date, various kinds of therapeutic approaches have been developed for the treatment of RET‐associated cancers, including tyrosine kinase inhibitors, gene therapy with dominant negative RET mutants, and RNA interference to abrogate oncogenic mutant RET expression. RET and some signaling molecules that function downstream of RET could be potential targets for the development of selective cancer therapeutics. (Cancer Sci 2005; 96: 143–148)

In 1985, the RET (REarranged during transfection) gene was identified as a novel oncogene activated by DNA rearrangement. ⁽¹⁾ The resulting chimeric oncogene encoded a fusion protein, consisting of an amino‐terminal half with a putative zinc finger motif, and a carboxyl‐terminal half with a tyrosine kinase domain. ⁽²⁾ This fusion resulted from recombination between two unlinked human DNA segments, which occurred during the transfection process.

The name RET (RET proto‐oncogene) has been retained to designate the gene coding for the tyrosine kinase. ⁽³⁾ The RET proto‐oncogene encodes a receptor tyrosine kinase with four cadherin‐related motifs and a cysteine‐rich region in the extracellular domain, and its ligands are glial cell line‐derived neurotrophic factor (GDNF) and related molecules, including neurturin (NRTN), artemin (ARTN) and persephin (PSPN). RET activation by these neurotrophic factors is mediated through a unique multicomponent receptor system, consisting of glycosyl‐phosphatidylinositol‐anchored coreceptor (GFRα1–4) as a ligand‐binding component and RET tyrosine kinase as a signaling component. ⁴ , ⁵ , ⁶ Gene knock‐out studies have revealed that GDNF/RET signal transduction is essential for the development of the kidney and the enteric nervous system. ⁷ , ⁸ , ⁹ , ¹⁰

RET mutations are responsible for the development of several human diseases, including multiple endocrine neoplasia 2A and 2B (MEN 2A and MEN 2B), familial medullary thyroid carcinoma (FMTC), papillary thyroid carcinoma (PTC) and Hirschsprung's disease (HSCR). ⁽¹¹⁾ MEN 2A and MEN 2B share the clinical features of medullary thyroid carcinoma (MTC) and pheochromocytoma, and FMTC is characterized by MTC alone. MEN 2A, MEN 2B and FMTC have been identified to be caused by gain‐of‐function germline mutations of RET, but somatic rearrangements of RET also lead to intrinsic tyrosine kinase activation in sporadic and radiation‐associated PTC. In contrast, loss‐of‐function mutations of RET are responsible for the development of HSCR, which is a congenital malformation associated with the absence of enteric neurons. ⁽¹¹⁾

Recently, useful therapeutic options for the treatment of neoplastic diseases have been developed, including the use of tyrosine kinase inhibitors, monoclonal antibodies or gene therapy. RET has also emerged as a potent target in some preclinical approaches to the treatment of RET‐associated cancer. This artile discusses the mechanisms involved in oncogenic RET signaling and the potential of RET as a therapeutic target.

Activation mechanisms of RET

Three splicing variants of RET (RET9, RET43, RET51) are generated by alternative spicing of the 3’ region (Fig. 1). Among them, RET9 and RET51 are major isoforms consisting of 1072 and 1114 amino acids, respectively. Glial cell line‐derived neurotrophic factor family ligands (GDNF, NRTN, ARTN, PSPN) activate RET via binding to GDNF family receptor α1–4 (GFRα1–4). Interaction of these ligands with GFRα1–4 induces RET dimerization, resulting in intrinsic tyrosine kinase activation (Fig. 2a). ⁶ , ¹¹

Figure 1.

Schematic structure of RET, and mutations identified in multiple endocrine neoplasia (MEN) types 2A, MEN 2B and familial medullary thyroid carcinoma (FMTC). CAD, cadherin‐related motif; CYS, cysteine‐rich region; S, signal sequence; TM, transmembrane domain; TK, tyrosine kinase domain.

Figure 2.

Activation mechanisms of RET tyrosine kinase. (a) Normal RET activation by glial cell line‐derived neurotrophic factor (GDNF). GDNF binds to glycosyl‐phosphatidylinositol‐anchored coreceptor (GFRα1) and leads to RET dimerization. (b) RET activation by multiple endocrine neoplasia (MEN) type 2A mutations. Cysteine mutations activate RET by inducing ligand‐independent and disulfide‐linked homodimerization. (c) Activation of RET by MEN 2B. Mutations in the tyrosine kinase domain activate monomeric RET, probably due to a conformational change in the kinase domain. (d) The chimeric RET/papillary thyroid carcinomas (PTC) oncoproteins localize to the cytoplasm, and the sequences fused to the RET tyrosine kinase domain are capable of inducing intracellular dimerization. Cys, cysteine residue; ‐, disulfide link; X, mutations in cysteine; PY‐, phosphorylated tyrosine.

MEN 2A mutations have been identified mainly in one of six cysteine residues (codons 609, 611, 618, 620 in exon 10, and codons 630, 634 in exon 11) in the RET extracellular domain (Fig. 1). Cysteine mutations were also found in some FMTC families. We and other investigators demonstrated that these cysteine mutations induced ligand‐independent, disulfide‐linked RET homodimerization, leading to its constitutive activation (Fig. 2b). ¹² , ¹³ It is hypothesized that when a cysteine is replaced with another amino acid as a result of a MEN 2A mutation, a partner cysteine that is involved in forming an intramolecular disulfide bond may become free and induce an aberrant intermolecular disulfide bond between two mutant RET proteins (Fig. 2b).

MEN 2B is caused by the Met918Thr (M918T) or the Ala883Phe (A883F) mutations in the RET kinase domain (Fig. 1). More than 95% of MEN 2B patients are accounted for by the M918T mutation and fewer than 4% are accounted for by the A883F mutation. These MEN 2B mutations appear to induce a conformational change of the catalytic core of the kinase domain and activate RET without dimerization (Fig. 2c). ¹³ , ¹⁴

Somatic chromosomal rearrangements of RET (RET/PTC) have been identified in 5–30% of sporadic and 60–70% of radiation‐associated PTC. To date, more than 10 rearranged forms of RET/PTC have been reported. In these rearrangements, the intracellular RET kinase domain is fused to amino‐terminal sequences of a variety of activating genes. These fusion proteins are capable of ligand‐independent dimerization (Fig. 2d). ¹⁵ , ¹⁶ , ¹⁷ , ¹⁸

Intracellular signaling pathways via RET

The intracellular domain of RET contains autophosphorylation sites, and phosphorylated tyrosines serve as docking sites for signaling molecules. ¹⁷ , ¹⁸ , ¹⁹ Among them, phosphorylated tyrosine 1062 (Y1062) represents a binding site for a variety of adaptor proteins including Shc, FRS2, Dok1/4/5, IRS1/2, and Enigma (Fig. 3), and is important for the transforming ability of mutant RET. ⁽²⁰⁾ In addition, it was found that tyrosine 905 binds to Grb7/10, tyrosine 981 to Src, tyrosine 1015 to phospholipase Cγ (PLCγ), and tyrosine 1096 to Grb2 (Fig. 3). ²¹ , ²² , ²³ Interestingly, rat sarcoma oncogene/extracellular signal‐regulated kinase (RAS/ERK), phosphatidylinositol‐3‐kinase (PI3K)/AKT, p38 mitogen‐activated protein kinase (p38MAPK), and c‐Jun N‐terminal kinase (JNK) pathways are activated mainly through tyrosine 1062. ⁽²⁴⁾ When the adaptor protein Shc binds to phosphorylated tyrosine 1062, it recruits the Grb2‐Gab1 and Grb2‐Sos complexes that then activate the PI3K/AKT and RAS‐ERK pathways, respectively (Fig. 4). ²⁵ , ²⁶ , ²⁷

Figure 3.

Signaling pathways mediated by RET. Dok1/4/5/6, downstream of tyrosine kinase 1/4/5/6; ERK, extracellular signal‐regulated kinase; FRS2, fibroblast growth factor receptor substrate 2; Gab1, Grb2 associated binding protein 1; Grb2, growth factor receptor bound protein 2; JNK, c‐Jun N‐terminal kinase; Nck, non‐catalytic region of tyrosine kinase; PI3K, phosphatidylinosytol 3‐kinase; PKC, protein kinase C; PLCγ, phospholipase Cγ; STAT3, signal transducer and activator of transcription 3; Ras, rat sarcoma oncogene; RasGAP, Ras‐GTPase activating protein.

Figure 4.

Biological and biochemical responses mediated by the RET signaling pathway.

Glial cell line‐derived neurotrophic factor‐dependent signaling, via tyrosine 1062 in RET, also plays a crucial role in organogenesis. We recently demonstrated that knock‐in mice, in which tyrosine 1062 is replaced with phenylalanine, show a severe defect of enteric neurons in addition to kidney hypodysplasia, indicating that signaling via tyrosine 1062 is important for the development of the enteric nervous system and the kidney. ⁽²⁸⁾ It is interesting to note that AKT activation is markedly impaired in these homozygous knock‐in mice.

However, oncogenic RET induces transforming activity and promotes cell invasiveness. ¹¹ , ¹⁷ AKT mediates multiple cellular responses, such as survival signaling by NFκB activation and BAD inactivation, and cell cycle progression through cyclin D1. Therefore, the activation of the PI3K/AKT pathway is required for RET‐mediated transformation. ²⁷ , ²⁹ , ³⁰ We previously reported that AKT and JNK phosphorylation significantly increases in MEN 2B transfectants compared to MEN 2A transfectants. ²⁵ , ³¹ In addition, Marshall et al. reported that the JNK pathway is involved in the ability of RET‐MEN 2B to metastasize. ⁽³²⁾ These findings suggest that higher levels of AKT and JNK activation may be responsible for the development of aggressive properties in MEN 2B.

Moreover, it was shown that signal transducer and activator of transcription 3 (STAT3) is an important downstream target of RET‐MEN 2A and RET‐MEN 2B. ³³ , ³⁴ Binding of STAT3 to Tyr752 and Tyr928 of RET enhances RET‐MEN 2A‐mediated cell proliferation and transformation. ⁽³³⁾ RET‐MEN 2B also constitutively activates STAT3. Nuclear enrichment of STAT3 and elevated expression of the chemokine receptor CXCR4 were detected in metastatic MTC from MEN 2B patients. In addition, STAT3 promoted the expression of metastasis‐related mucin genes (MUC1, MUC4, and MUC5B) via binding to their promoter regions. ⁽³⁴⁾

Effects of tyrosine kinase inhibitors on mutant RET‐expressing cells

RET is a potential target in the treatment of MEN 2A, MEN 2B, and PTC. There are several strategies to block the tyrosine kinase function of RET, including RNA interference to knock down RET expression, gene therapy with dominant negative mutants or small molecules that serve as tyrosine kinase inhibitors.

Imatinib (STI571; Gleevec), a 2‐phenylaminopyrimidine, specifically inhibits the tyrosine kinase activity of Abl, Kit and platelet‐derived growth factor receptor. It has been demonstrated that the administration of Imatinib (Gleevec) is effective in the treatment of chronic myelogenous leukemia, which are characterized by constitutively active Bcr‐Abl tyrosine kinase, and gastrointestinal stromal tumors, characterized by activating mutations of Kit tyrosine kinase. ⁽³⁵⁾ Therefore, it is rational to select tyrosine kinase inhibitors for the treatment of MTC and PTC, although a selective drug that targets oncogenic RET has not yet been found.

To date, a few tyrosine kinase inhibitors have been shown to inhibit oncogenic RET activity. The anilinoquinazoline ZD6474 is a selective inhibitor of the vascular endothelial growth factor (VEGF) receprot‐2 (flk‐1/KDR) tyrosine kinase. ZD6474 has a one‐half maximal inhibitory concentration (IC₅₀) of 40 nM against the isolated KDR enzyme and blocks VEGF‐stimulated endothelial cell migration and proliferation. ZD6474 is currently being used in clinical trials against non‐small cell lung cancer and breast cancer. ⁽¹⁷⁾ Recently, Carlomagno et al. reported that ZD6474 efficiently blocks in vivo phosphorylation and signaling of RET/PTC3 and RET‐MEN 2B oncoproteins with an IC₅₀ of 100 nM, and inhibits tumor growth when used to treat nude mice that have been injected with RET/PTC‐transformed NIH3T3 cells. ⁽³⁶⁾ ZD6474 has several advantages in cancer therapy, including anti‐angiogenetic effects, low toxicity, and the possibility of oral administration.

The pyrazolo‐pyrimidine PP1 inhibits the enzymatic activity and transforming ability of RET oncoprotein in NIH3T3 fibroblasts and thyroid carcinoma cell lines. ⁽³⁷⁾ PP2, a compound related to PP1, also blocks in vivo phosphorylation and signaling of RET/PTC1. ³⁸ , ³⁹ Because PP1 and PP2 blocks the Src family kinases, the inhibition of oncogenic activity by PP1 or PP2 may not depend solely on inhibition of RET kinase. Therefore, PP1 and PP2 can suppress cell growth by inhibiting not only RET but also Src. Interestingly, it was demonstrated that PP1 induces RET‐MEN 2A and RET‐MEN 2B oncoprotein degradation through proteosomal targeting. ⁽⁴⁰⁾ Therefore, the potent inhibitory effects of PP1, PP2 and ZD6474 on RET seem promising in the treatment of RET‐associated cancer, although mutation of valine 804 of RET causes resistance to these compounds. ⁽⁴¹⁾

Strock et al. ⁽⁴²⁾ showed that the indolocarbazole derivatives CEP‐701 and CEP‐751 inhibited RET autophosphorylation and proliferation of TT cells, a human MTC cell line, at concentrations <100 nM, and CEP‐751 and its prodrug, CEP‐2563, also inhibited tumor growth in TT cell xenografts. The 2‐indolinone derivative RPI‐1 is another RET tyrosine kinase inhibitor. ⁽⁴³⁾ In NIH3T3 cells expressing the RET‐MEN 2A mutant and TT cells, RPI‐1 treatment markedly reduced RET tyrosine phosphorylation as well as activation of downstream signaling molecules, and inhibited cell proliferation. The study also demonstrated that RPI‐1 has strong and reproducible antitumor activity against TT xenografts in nude mice. Importantly, RPI‐1 was delivered orally, the most suitable method for use in prolonged treatment.

The inhibition of RET may be an effective strategy for the treatment of MTC or PTC, although most of the small compounds examined lacked critical selectivity. Increasing numbers of compounds are now being discovered, and some compounds are in clinical trials. Specific tyrosine kinase inhibitors might soon become available for use in standard therapy of patients with RET‐associated cancers.

Alternative approaches for the treatment of RET‐associated cancer

Drosten et al. reported that adenovirus‐mediated transduction of dominant negative RET into TT cells reduced expression of oncogenic RET receptor on the cell surface, and that inoculation of dominant negative RET‐expressing MTC cells into nude mice led to an almost complete suppression of tumor growth. ⁴⁴ , ⁴⁵ These results suggest that inhibition of oncogenic RET expression by a dominant‐negative RET mutant is a powerful approach for MTC treatment, although many issues remain to be resolved before in vivo use of viral vectors.

Another powerful tool for cancer therapy is humanized monoclonal antibodies against oncogene products. Mouse chimeric humanized monoclonal antibodies have been developed against members of the ErbB tyrosine kinase family. Trastuzumab (Herceptin) and IMC‐C225 (cetuximab or Erbitux) are humanized monoclonal antibodies targeted against the ectodomains of ErbB2 and ErbB1, respectively. Trastuzumab was approved for the treatment of Erb2‐overexpressing breast cancer in 1998 and showed clinical efficacy. ⁽⁴⁶⁾ IMC‐C225, in combination with irinotecan (a DNA topoisomerase I inhibitor), was approved by the United States Food and Drug Administration for the treatment of metastasized colorectal cancer. ⁽⁴⁶⁾ Therefore, the development of monoclonal antibodies against the ectodomain of RET is expected to aid treatment of RET‐associated cancer.

As already mentioned above, tyrosine 1062 is a critical residue for the transduction of the intracellular signaling pathway downstream of RET, as well as the transforming activity of mutant RET proteins. The AKT and JNK pathways, which are implicated in the aggressive properties of MTC in MEN 2B, are also activated via phosphorylated tyrosine 1062. Tyrosine 1062 represents a binding site for the phosphotyrosine binding domain of several adaptor proteins, such as Shc and Dok, which are responsible for activation of AKT and JNK. Structural analysis of the interactions between RET and these adaptor proteins may provide useful information in the design of small compounds which selectively interfere with these interactions, thereby inhibiting oncogenic signaling.

Perspective and concluding remarks

Although there are several approaches in the treatment of RET‐associated cancers (Fig. 5), searching for specific inhibitors of RET kinase is a promising strategy. To identify small compounds with efficacy in vivo, it is important to develop MTC animal models. To this end, we previously generated RET‐MEN 2A protein‐expressing transgenic mice, which developed MTC with complete penetrance. ⁽⁴⁷⁾ In addition, the development of mammary or parotid gland adenocarcinomas was observed in approximately one‐half of the transgenic mice. ⁽⁴⁷⁾ We are now trying to treat transgenic mice using various signaling inhibitors as well as through gene therapy.

Figure 5.

Strategies to inhibit RET in cancer therapy.

RET dimerization is induced in MEN 2A‐associated MTC and sporadic or radiation‐associated PTC by different mechanisms, leading to the abnormal constitutive activation of RET enzymatic activity. Therefore, reagents such as antioxidants, which abrogate RET dimerization, may also be useful in the treatment of MTC and PTC. Moreover, recent advances in RNA interference technology are providing a novel tool for cancer therapy. The rational approach of molecularly targeted therapy will undoubtedly improve the future treatment of MTC and PTC.