Introduction

Signal transduction is the means by which an electrical, physical, or chemical signal elicits a cellular response by activating a cellular receptor and initiating a chain of biochemical events. Most commonly, signal transduction occurs when an extracellular molecule binds to a transmembrane protein receptor, triggering intracellular protein kinase activation, and protein phosphorylation. Receptor tyrosine kinases (RTKs) constitute one class of transmembrane receptors and are characterized by the intrinsic tyrosine kinase activity of their cytoplasmic regions.

RTKs are expressed in tissues throughout the body both during intrauterine development and in adulthood and play a critical role in regulating cell differentiation, proliferation, survival, metabolism, and migration. Unsurprisingly then, research has found RTK dysfunction to play a role in a variety of human diseases, most notably cancers. An in-depth understanding of specific RTKs, their signaling cascades, and their effects on cell function have allowed for the development of a multitude of targeted drug therapies, with significant improvements in clinical outcomes.

Cellular Level

There are 58 unique known RTKs that subdivide into 20 subfamilies. All have an extracellular region with ligand-binding domains, a single transmembrane alpha-helix, and a cytoplasmic region consisting of a juxtamembrane domain, a tyrosine kinase domain (TKD), and a C-terminal tail. The ligand-binding domains of the extracellular region vary based on receptor subfamily. For the majority of RTKs, the ligand that is bound is a soluble growth factor peptide. Exceptions include the ephrin (Eph) family of receptors, whose ligands are membrane-bound ephrins on nearby cells, and discoidin receptors, activated by collagen fibers.[1]

The intracellular signaling pathways associated with RTKs are triggered when adaptor proteins (e.g., Grb2) or protein kinases bind to phosphotyrosines on the RTK or its associated docking protein and become phosphorylated. A discussion of how ligand binding results in these phosphotyrosine-binding sites is in the “Mechanism” section. Phosphotyrosine binding is permitted by the Src homology 2 (SH2) and phosphotyrosine-binding (PTB) domains located on the adaptor proteins and protein kinases. Adaptor proteins also contain Src homology 3 (SH3) or WW domains that target polyproline regions on protein kinases and guanine nucleotide exchange factors. These allow for the formation of signaling complexes and further protein kinase activation. RTKs utilize several signaling pathways, including MAPK/ERK, PI3K/Akt/mTOR, and PLCG1/PKC. These pathways regulate the transcription of numerous genes whose products promote cell differentiation, cell cycle progression, and cytoskeletal alteration and inhibit cell senescence and apoptosis. Interestingly, RTK signaling occurs in endosomes as well as at the cell surface, with different signaling pathways recruited depending on the location of the receptor.

Regulation of the activity of RTKs and their downstream pathways is critical, given the vital cellular processes they modulate. Negative feedback mechanisms are varied and occur at different levels in the signaling networks. The simplest example is the direct activation of protein tyrosine phosphatases, which dephosphorylate the tyrosine residues on active RTKs. In some RTKs, like epidermal growth factor receptor (EGFR), downstream effector kinases PKC and MAPK phosphorylate threonine residues in the juxtamembrane region, inhibiting high-affinity ligand binding. In the MAPK/ERK pathway specifically, MAPK phosphorylation of the upstream docking protein Gab1 and protein kinase Sos impairs their ability to interact with other proteins, while phosphorylation of upstream Raf kinase dampens its enzyme activity. All these mechanisms serve to inhibit or terminate the cellular response to RTK activation. Alternatively, RTK activity can undergo attenuation by removing receptors from the cell surface. Active RTKs recruit Cbl, a ubiquitin ligase that tags the receptors for endocytosis and degradation via ubiquitination.[2]

Development

Signaling pathways triggered by growth factor-RTK binding play a pivotal role in embryonic and fetal development. This knowledge comes from studies in which RTK genes are intentionally knocked out or altered and from the study of clinical disorders of organogenesis. Almost all RTKs and their growth factors function in some capacity in organ formation; notable examples appear below.

c-KIT (stem cell factor/SCF) – c-KIT (i.e., stem cell growth factor receptor) is a member of the PDGFR subfamily, but it has roles in development that other members of the family do not. c-KIT/SCF drives the proliferation, survival, and migration of hematopoietic stem cells, melanocytes, and primordial germ cells. Receptors are expressed not only on the cells themselves but on their migratory pathways and final destinations (bone marrow, fetal liver, skin, gonads). c-KIT is also present on mast cells and stimulates their proliferation. Mice with mutated c-KIT have macrocytic anemia, decreased pigmentation, and impaired fertility. In humans, loss-of-function mutations in c-KIT cause piebaldism, a condition characterized by hypopigmented patches of skin and hair.[3]

ErbB (epidermal growth factor/EGF and neuregulins) – EGFR (i.e., ErbB1) receptors are present diffusely on embryonic and fetal tissue; mice lacking EGFR die in utero or the early postnatal period, and have defects in numerous tissues including skin, CNS, intestines, liver, and lung. ErbB2 and ErbB4 knock-out mice fail to form cardiac ventricle trabeculae, leading to intrauterine death. ErbB4 knock-outs also have defects in cranial sensory ganglia and peripheral nervous system myelination due to defects in Schwann cell migration. ErbB3 knock-out mice have impaired atrioventricular valve formation with early embryonic demise.[4] 

Eph (ephrin) – Eph/ephrin signaling is unique in that both the receptor and ligand are membrane-bound. Cell to cell interactions result in alteration of the actin cytoskeleton; thus, Eph receptors are important in cell migration. In particular, Eph receptors play a crucial role in neural crest cell migration and axon guidance in the developing nervous system. Bidirectional repulsive and attractive forces mediated by these receptors help create the topographic maps of various sensory systems, guide spinal motor neurons to their appropriate destinations, and aid in the formation of the cerebral cortex and midline structures like the corpus callosum. Mutations in the gene encoding for ephrin-B1 result in craniofrontal nasal dysplasia, characterized by hypertelorism, cleft palate, a central nasal groove, skeletal abnormalities, and corpus callosum agenesis.[1]

FGF (fibroblast growth factor/FGF) – FGFR/FGF signaling modulates chondrocyte and osteoblast differentiation, proliferation, and apoptosis. Consequently, it is necessary for normal osteogenesis, skeletogenesis, and limb bud development. In achondroplasia, activating mutations of FGFR3 promote excessive chondrocyte apoptosis, leading to impaired endochondral ossification and bone growth. In craniosynostosis, activating mutations in FGFR2 leading to accelerated osteoblast maturation and premature cranial suture fusion.[5]  

RET (glial cell-line derived neurotrophic factor/GDNF) – RET/GDNF signaling is necessary for the development of the enteric nervous system, Peyer’s patches, kidney, and ureters. Inactivating RET mutations are a significant risk factor for Hirschsprung’s disease. Knock-out studies in mice have revealed that the absence of RET leads to renal and ureter agenesis and malformation; however, studies in fetuses with these conditions have demonstrated RET mutations in only a small percentage, suggesting that they are not the driving mutation.[6][7][8]

Tie (angiopoietin-1/ANG1) – Tie2/ANG1 signaling plays a unique role in venous morphogenesis. It induces endothelial cells (EC) to adopt a venous EC morphology and is important for venous remodeling and maintenance. It has a smaller role in angiogenesis as well. Tie2 deficient mice die as a result of defective vascular organization/venogenesis, and Tie2 mutations are frequently present in patients with venous malformations.[9] 

Trk (neurotrophins) – The Trk subfamily of RTKs consists of three receptors, each with a plethora of roles in neuronal differentiation, development, and survival. For example, TrkA/neural growth factor signaling plays a pivotal role in the formation and survival of sympathetic ganglia and nociceptive neurons, while TrkC/neurotrophin-3 signaling is crucial for cochlear ganglia and proprioceptive neuron development. TrkB/brain-derived neurotrophic factor signaling is required for synaptic plasticity, long-term potentiation, maintaining the survival of CNS neurons, and forming ocular dominance columns. All three receptors promote motor neuron survival and axon growth.[10] The syndrome of congenital insensitivity to pain with anhidrosis has links to mutations in TrkA.[11] 

VEGF (vascular endothelial growth factor/VEGF) – VEGFR-1/2 are located primarily on ECs, and binding of VEGF-A regulates vasculogenesis, angiogenesis, and vascular permeability. VEGFR-3, when bound to either VEGF-C or D, regulates lymphangiogenesis. Disruption of these signaling pathways in mice embryos results in death from aberrant vasculature formation and impaired cardiac function.[12]

Function

With a few exceptions, RTKs bind growth factors that regulate cell differentiation, proliferation, survival, metabolism, and migration.  The biologic outcomes of these cellular events are essential not only in intrauterine development (see “Development” section), but in adult tissue function and homeostasis as well. For example, VEGFR/VEGF mediated angiogenesis allows for wound healing, Trk/BDNF remains necessary for neuronal survival, and c-KIT/SCF continues to stimulate spermatogenesis. Most RTK-ligand combinations have a plethora of functions in various tissues, and researchers continually discover additional roles. A small number of significant RTK functions from subfamilies not previously mentioned appear below.  

Axl (Gas6 and Protein S/ProS) – two receptors of the Axl subfamily, Axl and Mer, act to dampen the innate immune system by inhibiting TLR-induced cytokine production in dendritic cells. They also promote phagocytic uptake of apoptotic cells.[13]

Ins (insulin and insulin growth factor-1/IGF-1) – InsR/insulin binding regulates glucose homeostasis, stimulating glucose uptake, glycogenesis, and lipogenesis in cells throughout the body.  IGF1R/IGF-1 binding promotes symmetric growth in all tissues.

MuSK (agrin) – Skeletal muscle and functions express musk in the formation of the neuromuscular junction by clustering acetylcholine receptors and promoting the transcription of genes encoding synaptic proteins.[14] 

PDGF (platelet-derived growth factor/PDGF) – PDGF receptors are present on fibroblasts, ECs, and smooth muscle cells. They facilitate wound healing by stimulating vascular remodeling, smooth muscle cell migration, and fibroblast-induced collagen synthesis.[15] 

Mechanism

For RTKs to initiate a downstream signaling cascade and cellular response, three distinct processes must occur. These are ligand-induced dimerization, activation of the cytoplasmic kinase domains, and linking of the activated RTK to cell signaling.

Ligand-Induced Dimerization

Early studies of RTKs espoused that receptors existed on the cell surface as monomers and that bivalent ligand interactions crosslinked two monomers into an active dimer. Subsequent studies have revealed that ligand-induced activation is more varied and complex. Prior to ligand binding, some RTKs exist on the cell surface as dimers (e.g., Ins) or oligomers (e.g., EGFR). Ligand binding in these cases induces structural changes in the receptor, which stimulate tyrosine kinase activity. Additionally, some RTKs may require oligomers rather than dimers for activation (e.g., Tie2, Eph).

The extent to which the ligand makes up the interface between the two TRK dimers varies depending on the receptor as well. The dimerization of many RTKs, including TrkA, Tie2, and FLT1, follow the traditional format of being “ligand-mediated.” The extracellular regions of the two RTK monomers do not contact one another; the ligand forms the entire dimer interface. Conversely, receptors in the EGFR family are said to be “receptor-mediated.” The ligand constitutes no part of the dimer interface; instead, ligand binding to each receptor monomer induces conformational changes in the receptor extracellular region, exposing sites for dimerization. Between these two extremes are cases in which the dimer interface forms by both the ligand and extracellular receptor domains (e.g., c-KIT) and cases where the dimer interface forms by various interactions between the involved ligand, receptors, and an accessory molecule. An example of the latter is the FGFR family, which requires heparin to form the dimer interface.

Activation of Cytoplasmic Kinase Domains

TKDs of RTKs are comprised of a C-lobe, N-lobe, and “activation loop.” For a TKD to become functional, both the “activation loop” and N-lobe alpha-C helix must adopt a specific, “active” configuration. In the majority of cases, the release of cis-autoinhibition is required for this to occur. Cis-autoinhibition occurs in several ways. In Ins receptors, a tyrosine on the activation loop occupies the kinase active site, forming a stable configuration that blocks ATP and protein substrate entry.

Similarly, in Tie2 RTKs, tyrosines on the C-terminal tail occupy the active site. In several RTKs, (c-KIT, FLT3, and MuSK), tyrosines in the juxtamembrane region interact with the TKD and its activation loop, stabilizing it in an inactive conformation. In all cases, ligand dimerization permits the partner TKD to trans-phosphorylate these tyrosines, releasing the cis-autoinhibition and allowing the RTK to assume the active state.

The EGFR and RET RTKs are unique in that TKD activation do not require trans-phosphorylation. In the case of EGFR, allosteric interactions in the “receiver” N-lobe of the TKD of one RTK monomer maintain it in an inactive conformation. Upon dimerization, these interactions are disrupted by the “activator” C-lobe of the TKD of the other RTK monomer, allowing the receiver TKD to become active. The mechanism of TKD activation in RET RTKs is not as well elucidated. The unphosphorylated and phosphorylated TKDs both exist in an active configuration and have a similarly low level of enzymatic activity. Speculations are that RET may exist as a trans-inhibited dimer and that the inhibitory interactions are relieved by ligand binding.[16]

Linking of Activated RTK to Cell Signaling

Except for EGFR and RET, trans-autophosphorylation provides the means by which RTKs link to cell signaling pathways. Autophosphorylation occurs in two phases. The first phase has been described previously; trans-autophosphorylation of specific tyrosine residues relieves cis-autoinhibition, allowing the TKD to adopt the active configuration and optimizing enzyme kinetics. In second phase autophosphorylation, additional tyrosines in the juxtamembrane region, C-terminal tail, and TKD are phosphorylated, creating phosphotyrosine-binding sites that can be bound by kinases and adaptor proteins possessing SH2 or PTB domains. The TKD can also phosphorylate tyrosines on associated docking proteins, which have no enzymatic activity themselves but provide more potential binding sites. Examples of docking proteins include FRS2, IRS1, and Gab-1. The cell signaling pathways triggered by these interactions are discussed in the “Cellular” section.[2]

Related Testing

The detection of RTK gene mutations, amplifications, and rearrangements is of the utmost clinical importance, as they have significant therapeutic and prognostic implications. The DNA sequencing, fluorescence in situ hybridization (FISH), and immunohistochemistry (IHC) are the main laboratory tests that detect these gene alterations. DNA sequencing using the Sanger method continues to predominate in clinical labs, although the use of next-generation sequencing techniques is increasing. These high throughput methods use parallel sequencing to produce large volumes of short sequence data more quickly and at a lower cost than traditional methods. They are most useful in detecting single-nucleotide mutations.[17]

FISH utilizes fluorescent probes that complement a gene or DNA sequence of interest and is particularly helpful in the case of gene amplification, deletions, or translocations. Lastly, IHC employs antigen-specific antibodies linked to either a fluorescent molecule or an enzyme that reacts with its substrate to form a colored product. This method does not directly identify genetic aberrations like the previous two methods, but indirectly by detecting the abnormal expression or overexpression of a gene product on a cell’s surface.

Pathophysiology

Increased RTK signaling leads to uncontrolled cell growth and proliferation. Clinically, this results in cancer and other hyperproliferative disorders. One mechanism by which this occurs is gain-of-function mutations in RTK genes. Point mutations, deletions, or insertions result in RTK products with increased affinity for their ligands, an increased propensity to dimerize, or hyperactive TKD regions due to the loss of cis-autoinhibitory interactions. For instance, an exon 21 L858R substitution or exon 19 L747_A750 deletion in the EGFR gene both result in a receptor with a hyperactive TKD region and can present in patients with adenocarcinoma of the lung. They tend to be more common in nonsmokers, women, and persons of Asian ethnicity.[18] In acute myeloid leukemia, internal tandem duplications in the gene encoding for FLT3 result in ligand-independent activation of the receptor.[19] Other examples of gain-of-function mutations include c-KIT in gastrointestinal stromal tumor (GIST) and mastocytosis, RET in polyendocrine neoplasia, and medullary thyroid cancer, MET in papillary renal cell carcinoma, and FGFR3 in bladder cancer.[2][20][21]

DNA amplification of RTK genes, leading to receptor overexpression, is another means by which RTK signaling aberrantly increases. Clinically relevant examples of DNA amplification include ErbB2 (i.e., HER2) in breast and gastric cancers and FGFR1 in squamous cell lung cancer.[21]

Lastly, chromosomal rearrangements can result in RTK fusion genes producing constitutively active receptors. The EML4-ALK fusion gene, which results from a t(2;2) inversion, is seen in 2 to 7% of lung adenocarcinomas. ROS translocations present in 1 to 2%.[22] Various translocations in RET often occur in papillary thyroid cancer.[20]

Not only can the RTKs themselves be hyperactive or overexpressed, but so can their downstream targets. Gain-of-function point mutations in KRAS are present in a significant percentage of patients with pancreatic, colorectal, lung (adenocarcinoma), serous ovarian, and thyroid cancer. BRAF V600E substitution mutations may present in melanoma, hairy cell leukemia, papillary thyroid cancer, and lung adenocarcinoma. In the PI3K/AKT pathway, there are reports of Akt1 E17K point mutations in lung (squamous), breast, colorectal and ovarian cancers, while researchers have observed PIK3CA mutations in endometrial, breast, colorectal, anal, and cervical cancer.[23]

Clinical Significance

The presence of mutated or overexpressed RTKs has important therapeutic and prognostic implications in a number of cancers. In the past two decades, a multitude of drugs targeting these RTKs has been developed, with some resulting in significant improvements in clinical outcomes. These drugs come in two forms: monoclonal antibodies that prevent receptor dimerization or growth factor binding or as small molecule tyrosine kinase inhibitors that prevent TKD phosphorylation. The result in both cases in the inhibition of proliferation and the promotion of apoptosis in malignant cells. 

Examples of monoclonal antibodies are trastuzumab and pertuzumab, which target different epitopes on HER2 to prevent receptor dimerization. Both have indications for localized and metastatic breast cancers that overexpress HER2; trastuzumab is also useful in HER2+ metastatic gastric cancer. These drugs have had profound impacts on the outcomes of breast cancer patients. For instance, before the development of these drugs, overall survival (OS) in patients with HER2+ metastatic breast cancer treated with chemotherapy alone was approximately 20 months. Today, the addition of trastuzumab and pertuzumab to standard chemotherapy has increased the overall survival to approximately 56 months in these same patients. Once considered to have the worst prognosis, both local and metastatic HER2+ breast cancer now have a median survival similar to luminal breast cancers.[24]  

Cetuximab and panitumumab are two monoclonal antibodies that antagonize EGFR/EGF binding. Their most common use is in the treatment of metastatic colon cancer. When combined with standard FOLFOX or FOLFIRI chemotherapy in the primary treatment of metastatic colon cancer, these drugs increase progression-free survival (PFS) by approximately 1.5 mo (8.5 to 10 months). A significant increase in OS is seen only with panitumumab (23.8 vs. 19.4 mo). These benefits are present only in patients who are RAS wild type, as inhibition of RTK-ligand binding is rendered useless in the context of a constitutively active downstream protein kinase.[25][26] 

Ramucirumab is the only monoclonal antibody targeting the VEGFR, functioning to inhibit tumor angiogenesis. Specifically, it prevents VEGF binding to VEGFR-2. It is indicated only in advanced or metastatic colorectal, gastric, and non-small cell lung cancer that has progressed or relapsed after first-line therapy. Given this use, it has a more limited clinical significance. For example, in patients with advanced or metastatic gastric and gastroesophageal cancer who had failed fluoropyrimidine or platinum-based therapy, the median overall survival was 5.2 months in the ramucirumab group vs. 3.8 months in the placebo group.[27]  

The number of small-molecule tyrosine kinase inhibitors (TKI) greatly exceeds the number of RTK targeted monoclonal antibodies. And unlike the antibodies, most TKIs are capable of inhibiting numerous RTKs and can be used in a wider variety of cancers. For these reasons, a complete discussion of the clinical significance of RTKs is beyond the scope of this text, but some examples appear below. 

As described in the “Pathophysiology” section above, a portion of patients with adenocarcinoma of the lung have genetic alterations in the gene coding for EGFR, most notably an exon 21 substitution or exon 19 deletion. Monotherapy with the TKIs osimertinib, erlotinib, gefitinib, and afatinib are all approved in metastatic lung adenocarcinoma harboring these mutations. Studies have demonstrated these drugs have improved PFS, but have a similar OS to chemotherapy, alone in this population.[28][29] Osimertinib is preferred over the others as it demonstrates superior PFS and remains efficacious in patients with an additional EGFR substitution mutation in exon 20 (T790M).[30][31] In patients with EML4-ALK+ metastatic lung adenocarcinoma, the TKI alectinib is the first-line therapy, while crizotinib and brigatinib are alternatives.[32]

Other clinical uses of targeted TKI therapies include lapatinib as second or third-line treatment in metastatic HER2+ breast cancer,[24] midostaurin in FLT3 mutated acute myeloid leukemia,[19] larotrectinib as salvage therapy in unresectable or metastatic solid tumors containing Trk gene translocations,[33] and imatinib in c-KIT mutated GIST. In c-KIT, mutations in exon 11 convey a better prognosis and response to imatinib than mutations in exon 9.[34]

Drugs targeting RTK ligands have also undergone development, as have drugs that target mutated protein kinases in their downstream signaling pathways. An example of the former is the bevacizumab, which binds VEGF and prevents it from binding VEGFRs. Examples of the latter include dabrafenib and vemurafenib, which target RAF protein kinases with the V600E mutation, and trametinib, inhibiting MEK. The most extensive use of these drugs is in melanomas with the BRAF V600E mutation.[35]