RET Proto-Oncogene

Abstract

The rearranged during transfection (RET) proto-oncogene encodes a transmembrane receptor tyrosine kinase and its alterations cause various cancers and developmental disorders. Gain-of-function mutations caused by gene rearrangements have been found in papillary thyroid carcinoma, non-small-cell lung carcinoma, and other cancers, while point mutations are responsible for hereditary cancer syndrome, multiple endocrine neoplasia type 2, and sporadic medullary thyroid carcinoma. Loss-of-function point mutations or deletions lead to Hirschsprung disease, a developmental disorder associated with aganglionosis of the intestinal tract. RET is also involved in various physiological and developmental functions through activation by glial cell line-derived neurotrophic factor (GDNF) family ligands (GFLs). Gene knockout studies have revealed that GDNF-RET signaling plays an essential role in the development of the enteric nervous system, kidney, and urinary tract, as well as in the self-renewal of spermatogonial stem cells. Moreover, recent progress in developing RET-selective inhibitors has significantly contributed to treating patients with RET-altered cancers. This chapter describes and discusses the functions associated with disease and physiology.

1. Introduction

The rearranged during transfection (RET) oncogene was identified as a new transforming gene by the transfection of NIH3T3 cells with human lymphoma DNA in 1985 [1]. The transforming gene was generated by recombining two unlinked human DNA sequences that most likely occurred during transfection. Hence, the term RET stems for “REarranged during Transfection.” The active RET transforming gene encodes a fusion protein comprising a carboxy-terminal tyrosine kinase domain and an amino-terminal dimerizing domain, fused by rearrangement [2]. The dimerizing domain is necessary for tyrosine kinase activation. Subsequently, the name RET was retained to designate the carboxy-terminal of the gene as a tyrosine kinase domain (RET proto-oncogene).

Alterations in the RET proto-oncogene have been found in various human cancers and developmental disorders, including thyroid cancer, non-small cell lung cancer, multiple endocrine neoplasia type 2 (MEN2), and Hirschsprung disease (HSCR) [3]. In addition, next-generation DNA and RNA sequencing approaches have identified less frequent RET rearrangements and mutations in various cancers, including colon and breast cancers. The mechanisms by which these mutations lead to RET activation or inactivation have been extensively studied [4, 5]. Ongoing clinical trials on RET-selective inhibitors have demonstrated remarkable efficacy and continue to improve the outcomes of patients with RET alterations without increased toxicity [6].

2. Structure and expression of RET proto-oncogene

The RET proto-oncogene encodes a receptor tyrosine kinase with a unique extracellular domain that consists of four cadherin-like domains and a cysteine-rich region with 16 cysteine residues in a stretch of 120 amino acids (Figure 1) [3, 7, 8, 9]. Alternative 3′ splicing produces three different isoforms (1072, 1106, and 1114 amino acids) with short (9 amino acids, referred to as RET9), intermediate (43 amino acids, RET43), and long (51 amino acids, RET51) carboxy-terminal tails. RET9 and RET51 are two major isoforms highly conserved among various species [4, 5]. The human RET proto-oncogene is located on chromosome 10q11.2 and comprises 21 exons.

Figure 1.

RET activation by GDNF family ligands (GFLs). GDNF, NRTN, ARTN, and PSPN preferentially bind to GFRα1, GFRα2, GFRα3, and GFRα4, respectively, and activate RET. GDNF/GFRα1/RET signaling complex is essential for developing the enteric nervous system, kidney, and urinary tract and for self-renewal/survival of spermatogonial stem cells. GDF15 is a stress response cytokine. GDF15 binds to GFRAL and activates RET, as observed for GFLs. GDNF, glial cell line-derived neurotrophic factor; NRTN, neurturin; ARTN, artemin; PSPN, persephin; GDF15, growth differentiation factor-15; GFRAL, GDNF family receptor α-like; CLD, cadherin-like domain; CRD, cysteine-rich domain.

RET expression is observed in the developing excretory and nervous systems during the embryogenesis of mice and rats [10, 11]. It is highly expressed in the nephric duct, ureteric bud, and collecting ducts of developing kidneys. RET is also expressed in enteric neural crest-derived cells, the autonomic and dorsal root ganglia of the peripheral nervous system, the neuroepithelial cells of the ventral neural tube, and several cranial ganglia in the central nervous system. In agreement with this expression pattern, Ret-deficient mice exhibit a lack of enteric neurons in the entire gastrointestinal tract and kidney agenesis or severe dysgenesis, elucidating the pivotal roles of RET in development [3]. After birth, various peripheral and central nervous system neurons continue to express RET, whereas their expression disappears in adult kidneys.

3. Activation of RET by glial cell line-derived neurotrophic factor (GDNF)-family ligands

In 1993, GDNF was purified and cloned as a neurotrophic factor that enhances the survival of midbrain dopaminergic neurons [12]. GDNF has also been shown to be a potent trophic factor for spinal motor and central noradrenergic neurons. Additionally, GDNF is essential for the survival and differentiation of peripheral sympathetic, parasympathetic, sensory, and enteric neurons. GDNF is structurally related to the transforming growth factor (TGF)-β and contains seven cysteine residues called cysteine knot motifs. Furthermore, three other proteins of the GDNF family ligands (GFLs), including neurturin (NRTN), artemin (ARTN), and persephin (PSPN), were identified, sharing approximately 40% amino acid identity with each other and possessing neurotrophic effects on various neurons [3].

Physiologically, RET is activated by GFLs through a unique multicomponent receptor complex consisting of a glycosylphosphatidylinositol-anchored co-receptor (GDNF family receptor α 1–4, GFRα1–4) as a ligand-binding component and RET tyrosine kinase as a signaling component (Figure 1). The formation of the GFL-GFRα-RET 2:2:2 ternary complex results in the activation of various intracellular signaling pathways necessary for physiological function [13, 14, 15]. GDNF, NRTN, ARTN, and PSPN use GFRα1, GFRα2, GFRα3, and GFRα4, respectively, as their preferred ligand-binding receptors (Figure 1), although crosstalk occurs between the ligand and co-receptor pairs to a certain extent [16, 17]. Despite the crosstalk demonstrated in vitro, knockout mouse studies have exhibited specific roles for each GFL-GFRα-RET complex. For example, Gdnf-, Gfrα1-, and Ret-deficient mice share phenotypes characterized by a lack of enteric neurons in the entire gastrointestinal tract, kidney agenesis, and severe dysgenesis [18, 19, 20]. Both Nrtn- and Gfrα2- deficient mice showed a reduced number of myenteric neurons in the intestine and a drastic reduction in cholinergic innervation in the lacrimal and salivary glands, indicating that the preferred GFL-GFRα-RET complex plays specific roles in vivo [16].

More recently, another ligand of the TGF-β superfamily, growth differentiation factor 15 (GDF15), has been shown to bind to GFRα-like (GFRAL) and activate RET (Figure 1), regulating food intake and body weight [21, 22, 23, 24]. GDF15 is a stress-induced hormone, and its plasma levels are markedly increased in various human diseases, including cardiovascular and chronic kidney diseases, diabetes, advanced cancer, and serious infections. Co-expression of GFRAL and RET was detected in neurons of the hindbrain area postrema and the nucleus of the solitary tract, where RET activation by GDF15 plays a pivotal role in body weight control.

4. Activation of RET intracellular signaling pathways by GFLs and their roles in the development

Following RET activation by GFLs, many tyrosine residues in the intracellular region are phosphorylated, activating various signaling pathways. The intracellular region of RET contains 18 tyrosine residues, two of which are in the juxtamembrane domain, 11 in the kinase domain, and five in the C-terminal region. Of the five tyrosine residues in the C-terminus, three (Y1015, Y1029, and Y1062) were common between RET9 and RET51, and two (Y1096 and Y1102) were present only in RET51. Each phosphorylated tyrosine interacts with specific adaptor proteins. For example, phosphorylated Y981, Y1015, Y1062, and Y1096 represent binding sites for SRC, phospholipase Cγ (PLCγ), SHC/FRS2/DOK1–6/IRS1–2, and GRB2, respectively (Figure 2). The signal transducer and activator of transcription 3 (STAT3) bind to phosphorylate Y752 and Y928 [3, 17]. As a result, the RAS/mitogen-activated protein kinase and/or phosphatidylinositol-3 kinase pathways were activated through the phosphorylation of Y1062 or Y1096 (Figure 2). Activation of the PLCγ pathway through phosphorylated Y1015 regulates protein kinase C activity and Ca²⁺ release from the endoplasmic reticulum, increasing intracellular Ca²⁺ levels and inducing Ca signaling.

Figure 2.

RET activates intracellular signaling pathways via phosphotyrosines and CCDC6-RET fusion. Phosphorylated tyrosines in RET intracellular domain interact with a wide range of adaptor proteins, activating downstream signaling pathways, including the RAS/MAPK, PI3K-AKT, and PLCγ-PKC pathways. For example, phosphotyrosine 1062 represents a multifunctional docking site for SHC, FRS2, and DOK family proteins. CCDC6-RET fusion identified in human cancers dimerizes via the coiled-coil domain present in the amino terminus of CCDC6.

The importance of each intracellular signaling pathway has been demonstrated in mouse gene-targeting studies. Ret mutant mice in which tyrosine 1062 was replaced with phenylalanine (Y1062F) exhibited severe defects in enteric neurons in the intestinal tract, small kidneys, and a lack of spermatogonial stem cells, indicating a crucial role for Y1062 signaling in development [25, 26, 27]. Ret Y1015F mutant mice show abnormalities in the kidney and ureter, such as multicystic kidneys and megaureters, but produce only minor abnormalities in the enteric nervous system [28]. The PLCγ pathway is critical for upper and lower urinary tract development, and mutations that abrogate this pathway generate features reminiscent of congenital anomalies of the kidneys or urinary tract (CAKUT).

HSCR is a relatively common congenital malformation associated with aganglionosis of the gastrointestinal tract (prevalence: one in 5000 live births). The disease is characterized by the absence of the intramural nervous plexuses, myenteric plexus (Auerbach plexus), and submucosal plexus (Meissner plexus). RET is a major causative gene of HSCR in which various mutations, including missense, nonsense, and frameshift mutations or partial/complete deletions, have been detected [29, 30]. RET mutations are found in approximately 50% of patients with familial HSCR and 10–20% of sporadic cases. These mutations inactivate and abrogate RET signaling, which is responsible for the migration and proliferation of enteric neural crest-derived cells during embryogenesis. Since HSCR mutations have been identified along the entire coding sequence, various mechanisms that perturb RET intracellular signaling have been demonstrated [3, 31]. For example, most mutations identified in the extracellular domains impair RET cell surface expression, most likely because of protein misfolding. Mutations in the kinase domain result in complete or partial impairment of kinase activity. Some mutations in the C-terminal tail impair intracellular signaling due to decreased binding of adaptor proteins such as SHC.

5. RET rearrangement in human cancer

In human cancers, RET rearrangement or fusion was first identified in papillary thyroid carcinoma (PTC) at a relatively high frequency (~25%) [32]. Subsequently, an extensive series of studies by the same group, including 177 papillary thyroid carcinomas, 37 follicular carcinomas, 15 anaplastic carcinomas, 18 medullary thyroid carcinomas, and 34 benign adenomas by Southern blot analysis, revealed rearrangements in 19% of papillary thyroid carcinomas, but not in other malignant and benign tumors [33]. However, at a lower frequency, RET rearrangements have been reported in other types of cancers such as follicular, anaplastic, and medullary thyroid carcinomas, as well as in benign thyroid lesions, such as follicular adenoma, at variable frequencies [34]. Somatic rearrangements involve the 3′ sequence of RET with a tyrosine kinase domain and the 5′ sequence of partner genes with a dimerizing domain such as the coiled-coil domain. RET breakpoints often occur within intron 11 and less frequently within introns 7, 10, and others [3].

Further studies reported that the prevalence of RET rearrangements in PTC varies significantly in different geographic regions, ranging from 3% in Saudi Arabia to 85% in Australia [34]; this may be because of racial differences in genetic backgrounds, environmental factors, or variations in screening techniques such as Southern blot analysis, reverse transcription polymerase chain reaction, and in situ hybridization. Integrated multi-platform analyses performed using The Cancer Genome Atlas network yielded RET rearrangements in 6.8% of PTC samples (33/484 cases) [35]. In addition, it is notable that BRAF mutations (prevalence: approximately 60%) and RET fusions are largely mutually exclusive in PTCs.

The most prevalent fusion genes were coiled-coil domain containing 6 (CCDC6)-RET (named RET/PTC1) (Figure 2) and nuclear receptor coactivator 4 (NCOA4)-RET (named RET/PTC3), which accounted for >90% of all RET rearrangement-positive PTCs (Table 1) [36]. Since CCDC6 and NCOA4 are located on chromosome 10q, as observed for RET, both RET/PTC1 and RET/PTC3 are created by paracentric inversion of this chromosome. Other partner genes, including PRKAR1A (RET/PTC2), GOLGA5, TRIM24, TRIM27, TRIM33, KTN1, RFG9, ELKS, PCM1, and HOOK3, and RET fusions with these genes are created by intrachromosomal translocation. There are at least 19 different 5′ partner genes for RET fusion in PTC [34, 37].

Cancer	Prevalence	RET fusions
Papillary thyroid carcinoma in general population	5–20%	CCDC6-RET; NCOA4-RET
Papillary thyroid carcinoma after Chernobyl reactor accident	50–80%	NCOA4-RET; CCDC6-RET
Non-small cell lung carcinoma	1–2%	KIF5B-RET; CCDC6-RET
Colon carcinoma	~0.2%	CCDC6-RET; NCOA4-RET
Salivary intraductal carcinoma	~40%	NCOA4-RET; TRIM27-RET

Table 1.

Prevalence of RET fusions in human cancer.

The prevalence of RET rearrangement is much higher (49–87%) in radiation-induced PTCs following the Chernobyl nuclear accident or the atomic bomb in Japan (Table 1). The highest frequency was observed in post-Chernobyl children [34]. Interestingly, in post-Chernobyl PTCs that developed less than 10 years after the accident, RET/PTC3 was most frequently detected, whereas tumors found after a longer latency preferentially carried RET/PTC1. In the case of PTC patients exposed to radiation from the atomic bomb, RET fusions occurred at a higher frequency of 50% in patients with high doses (>0.5 Gy) [38]. In patients with thyroid tumors (benign or malignant) who received external therapeutic radiation, the prevalence of RET/PTC was reported to be high (52–84%). In contrast, the prevalence of BRAF mutations was low in patients with radiation-induced PTCs.

RET rearrangement is also found in a portion (1–2%) of non-small-cell lung carcinomas (NSCLC) (Table 1) [39, 40, 41]. The most common RET fusions in lung cancer are kinesin family member 5B (KIF5B)-RET (~80%) and CCDC6-RET (~15%), followed by fusions such as NCOA4-RET, TRIM33-RET, and CUX1-RET [36]. KIF5B is located on the short arm of chromosome 10, and the pericentric inversion of this chromosome creates the KIF5B-RET fusion. Patients with RET fusion-positive NSCLCs show unique clinicopathological characteristics: relatively young (<60 years old), female, and nonsmokers.

Next-generation DNA and RNA sequencing technologies have identified less frequent RET fusions in various cancers. These include colorectal, breast, ovarian, salivary gland intraductal carcinomas, chronic myelomonocytic leukemia, and spitzoid tumors. Large-scale analyses have revealed that RET fusions can be detected in 0.2% of colorectal cancers (6/3117 cases) [42] and 0.1% of breast cancers (8/9693 cases) [43]. The CCDC6-RET and NCOA4-RET fusion genes were identified in both tumors. Interestingly, a high frequency (>40%) of RET fusions, including NCOA4-RET and TRIM27-RET, was detected in salivary intraductal carcinoma but not in salivary duct carcinoma (Table 1) [44]. In addition, new RET fusions ETV6-RET and VIM-RET were identified in 16% (8/49 cases) and 2% (1/49 cases) of secretory carcinomas of the salivary gland [45]. These findings suggest that detecting specific RET fusions helps in diagnosing specific salivary gland carcinomas.

6. RET mutations in MEN2 and sporadic cancer

Medullary thyroid carcinoma (MTC) is a malignant tumor of the neural crest-derived parafollicular C cells that produce calcitonin. MTC develops either sporadically (~75% of cases) or as a component of the hereditary cancer syndrome MEN2 (~25% of cases). MEN2 is an autosomal-dominant cancer syndrome characterized by the development of MTC and pheochromocytoma derived from adrenal chromaffin cells. Based on clinical phenotypes, MEN2 is classified into three subtypes: MEN2A, MEN2B, and familial medullary thyroid carcinoma (FMTC). The affected family members by MEN2A develop MTC (~100% of cases), pheochromocytoma (~50% of cases), and parathyroid hyperplasia/adenoma (hyperparathyroidism, ~20% of cases). Lichen amyloidosis is occasionally observed in MEN2A patients. MEN2B is a more aggressive subtype with early onset of MTC (~100%) and pheochromocytoma (~50%). In addition, MEN2B patients display a more complex phenotype, including mucosal neuroma, hyperganglionosis of the intestine, medullated corneal nerve, and marfanoid habitus, but not hyperparathyroidism. FMTC is characterized by MTC, which usually develops later in life and is now considered an indolent subtype of MEN2A.

Germline RET mutations are responsible for MEN2 syndrome [46, 47, 48, 49]. The majority of MEN2A mutations (>95%) have been identified in one of six cysteine residues (codons 609, 611, 618, and 620 in exon 10 and codons 630 and 634 in exon 11) in the cysteine-rich region of the RET extracellular domain (Figure 3A). Among them, approximately 85% of MEN2A mutations affect codon 634. The same cysteine mutations were also found in FMTC, with a high frequency of ~60% for codon 609, 611, 618, 620, or 630 substitutions and a low frequency of ~30% for codon 634. In addition, other point mutations, including Gly533Cys (G533C) (exon 8 in the extracellular domain), Glu768Asp (E768D), Leu790Phe (L790F), Tyr791Phe (Y791F), Val804Met/Leu (V804M/L), and Ser891Ala (S891A) substitutions (exons 13–15 in the kinase domain), have been reported in some FMTC and/or MEN2A families (Figure 3A) [3, 31]. Moreover, a 9- or 12-base pair duplication in exon 11 and a 9-base pair duplication in exon 8, which creates an additional cysteine residue, have been reported in two MEN2A families and one FMTC family, respectively [5].

Figure 3.

Germline RET mutations in MEN2. A. the majority of MEN2A mutations (>95%) are identified in one of six cysteine residues (codons 609, 611, 618, and 620 in exon 10 and codons 630 and 634 in exon 11) in the cysteine-rich domain (CRD) of the RET extracellular region. In addition to cysteine substitutions, MEN2A/FMTC mutations are less frequently found at non-cysteine residues in both the extracellular and intracellular regions. The M918 mutation is detected in >95% of MEN2B patients. B. Mechanism of RET activation by cysteine mutations. When a cysteine residue is replaced with another amino acid (designated X) in MEN2A/FMTC, mutant RET proteins form aberrant ligand-independent dimerization, resulting in constitutive activation.

We and others have demonstrated that cysteine mutations in MEN2A or FMTC result in ligand-independent constitutive activation (dimerization) of mutant RET by forming aberrant intermolecular disulfide bonds (Figure 3B). Cysteine residues are thought to form the intramolecular disulfide bonds necessary for the appropriate tertiary structure of the RET protein. The hypothesis is that when a cysteine residue is replaced with a non-cysteine residue by MEN2A mutations, the partner cysteine involved in the disulfide bond becomes free and forms an aberrant intermolecular disulfide bond with mutant RET, leading to constitutive activation (dimerization) (Figure 3B) [50, 51].

Two specific missense mutations, Met918Thr (M918T) and Ala883Phe (A883F), were associated with the development of MEN2B (Figure 3A). The M918T mutation accounts for more than 95% of MEN2B patients, while fewer than 4% are accounted for by the A883F mutation [3, 31]. In addition, double germline mutations at codons 804 and 806 (V804M and Y806C) were found in a Japanese patient with clinical features characteristic of MEN2B [52]. MEN2B mutations identified in the tyrosine kinase domain appear to activate RET in a monomeric form, probably because of a conformational change in the catalytic core of the kinase domain. Some reports have suggested that MEN2B mutations alter the substrate specificity of RET [3], although the signaling pathway crucial for developing MEN2B clinical phenotypes remains elusive.

According to data published in a public database in 2015 (COSMIC; Catalog of Somatic Mutations in Cancer), somatic RET mutations have been identified at a high frequency (>40%; 667/1662 cases) in sporadic MTC patients. M918T mutations were the most frequent in these patients, and cysteine mutations (C634, C630, C620, C618, and C611) and E768D, A883F, and S891A mutations in the kinase domain have also been observed in sporadic MTC [36, 37]. Oncogenic RET mutations have also been detected in 0.5% (8/1489 cases) of colorectal cancer [42] and 0.2% (16/9693 cases) of breast cancers [43]. Moreover, studies using next-generation sequencing have revealed the presence of RET mutations in a variety of cancers at a low frequency, including endometrial and ovarian cancer, hepatoma, skin melanoma, glioblastoma multiforme, meningioma, gastrointestinal stromal tumor, Merkel cell carcinoma, paraganglioma, and atypical lung carcinoid [53].

7. RET mutations in molecular diagnostic

RET genetic screening using next-generation sequencing is essential for the diagnosis of patients with MTC. The majority of MEN2 mutations are found in exons 10-11 and exons 13-16 of the RET proto-oncogene (Figure 3). After diagnosis, annual screening for MTC and pheochromocytoma should be carried out and prophylactic thyroidectomy for preventing metastasized MTC is recommended. The timing of prophylactic thyroidectomy is based on risk stratification of the RET mutation proposed by the American Thyroid Association (ATA). The revised ATA guidelines use ‘highest risk’ (HST), ‘high risk’ (H), and ‘moderate risk’ (MOD) that are associated with aggressiveness [54]. The ATA-HST category incudes patients with MEN2B and RET M918T mutation, and the ATA-H category includes patients with RET C634 mutations and the RETA883F mutation. The ATA-MOD category includes patients with RET mutations other than M918T, C634 and A883F. RET genetic screening is beneficial for early diagnosis and optimal treatment of patients with both hereditary and sporadic MTC.

8. RET overexpression in breast cancer

RET has been reported to be overexpressed in 40–60% of breast tumors across multiple tumor subtypes [55]. In particular, its expression is common in the estrogen receptor (ER)-positive subtype and is associated with larger tumor size, higher stage, and reduced overall survival. Elevated RET expression was also observed in ER-positive breast cancer cell lines, mirroring observations from patient samples. Treatment with estradiol induced the transcription of RET, GFRα, and ARTN, suggesting a regulatory mechanism for RET function. Multiple estrogen response elements have been identified within the RET enhancer region, which is approximately ~50 k bp relative to the RET transcriptional start site. In vitro study revealed that GDNF stimulation in breast cancer cells induces ER phosphorylation and estrogen-independent activation of the ER pathway, resulting in increased tumor cell proliferation and survival. RET-mediated ER phosphorylation involves the mammalian target of the rapamycin signaling pathway and is correlated with resistance to endocrine therapies such as tamoxifen [56, 57]. In addition, RET overexpression was observed in some human epidermal growth factor receptor 2 (HER2)-positive breast cancers. Using patient-derived xenograft models and cell lines, GDNF has been demonstrated to induce crosstalk between RET and HER2 via Src kinase, thereby conferring resistance to HER2-targeting therapy [58]. Thus, RET may be a useful target against endocrine- and trastuzumab-resistant breast cancers.

9. Development of selective RET kinase inhibitors for targeted therapy

Various tyrosine kinase inhibitors (MTKIs) have recently been used to treat RET-altered cancers. MTKIs include vandetanib, cabozantinib, sorafenib, sunitinib, ponatinib, lenvatinib, alectinib, and RXDX-105. However, MTKIs have demonstrated limited clinical efficacy, as indicated by their lower objective response ratios (ORRs), shorter progression-free survival rates, and significant off-target adverse effects [36, 59].

The RET-selective TKIs selpercatinib and pralsetinib have been developed to treat RET-altered cancers to overcome these problems. In phase1/2 trials, selpercatinib and pralsetinib demonstrated remarkable efficacy in RET fusion-positive NSCLC, with ORRs among treatment-naïve and platinum-based chemotherapy-pretreated patients of 85% and 64%, respectively, for selpercatinib, and 70% and 61%, respectively, for pralsetinib. Notably, both inhibitors also exhibited antitumor activity in patients with brain metastasis of NSCLC [36, 60, 61, 62].

In RET-mutant MTC patients who previously received MTKI treatment (cabozantinib, vandetanib, or both) and patients with treatment-naïve RET-mutant MTC, the ORRs were 69 and 73%, respectively, for selpercatinib [63]. In patients with previously treated RET-fusion-positive thyroid cancer, the ORR was 79% for selpercatinib. Patients who received pralsetinib also showed similar safety and remarkable efficacy in RET-mutant or fusion-positive thyroid cancer [64].

As observed for inhibitors of other tyrosine kinases, resistance to selective RET inhibitors occurs via both on-target and off-target mechanisms. After a dramatic initial response, the G810C/R/S and Y806C/N mutations demonstrated acquired resistance to inhibitors [65, 66]. They were detected in the RET tyrosine kinase domain’s solvent front and hinge regions. In addition, L730V/I resistance mutations at the roof of the solvent-front site were identified as strongly resistant to pralsetinib, but not to selpercatinib [67]. Novel RET kinase inhibitors have been developed to overcome these mutations. In addition, several reports have shown bypass mechanisms (off-target mechanisms) of resistance, including MET or FGFR1 amplification and KRAS NRAS or BRAF mutations [66]. Thus, developing multiple therapeutic strategies is necessary to enhance the clinical benefits of selective RET kinase inhibitors.

10. Conclusion

Since the discovery of the RET gene in 1985 [1], research on RET function has made significant progress in cancer research, developmental biology, and neuroscience [3]. Many point mutations or rearrangements of the RET proto-oncogene have been identified in various hereditary and nonhereditary cancers, and the mechanisms of RET activation have been elucidated. Long-sought RET-selective inhibitors (selpercatinib and pralsetinib) have been developed, significantly increasing the therapeutic efficacy of RET-mutant cancers. Further development of new inhibitors is expected to overcome resistance to selective inhibitors.

In developmental biology, RET functions, including intracellular signaling, are being extensively studied to understand the mechanisms underlying the development of the enteric nervous system, kidneys, and spermatogenesis. RET is a major causative gene of HSCR, an enteric nervous system developmental disorder. In contrast to the mutations identified in cancers, HSCR mutations are loss-of-function mutations. In addition, the activation of RET by GFLs enhances the survival of dopaminergic, motor, and sensory neurons. Therapeutic strategies targeting RET are expected for treating neurodegenerative diseases such as Parkinson’s disease and amyotrophic lateral sclerosis, as well as for nerve pain.

Moreover, recent findings demonstrating the association between RET activation by GDF15 in neurons of the brainstem and a decrease in food intake are opening up a new field of research [21, 22, 23, 24]. These findings may advance our understanding of the mechanisms underlying cachexia in patients with cancer. It is expected that future research on the RET function will continue to have a profound impact on life and medical science.

Acknowledgments

The author thanks all laboratory members and research collaborators for insightful discussion. Work at the author’s laboratory is partly supported by Grant-in-aid for Scientific Research from the Ministry of Education, Culture, Sports, Science and Technology (MEXT) of Japan. I would like to thank Editage (www.editage.com) for English language editing.

Conflict of interest

The author declares no conflict of interest.