The Epidermal Growth Factor Receptor (PDB ID: 2gs6)
Created by Emily Steen
THE BASIC STRUCTURE OF EPIDERMAL GROWTH FACTOR RECEPTOR
The EGF receptor as a whole is part of the ErbB superfamily of proteins with intrinsic kinase (RTK) activity, which control aspects of cell growth and development, such as apoptosis, protein secretion, and organ morphogenesis and repair. It is therefore expressed and active in many cell types and is a highly conserved gene sequence. Null mutations in the ErbB gene loci on the short arm of chromosome 7 are embryonic lethal at day 10.5-13.5 (8), demonstrating their essentiality.
In vertebrates, four EFGR family members are known - ErbB1 (otherwise known as EGFR, or HER1), ErbB2 (otherwise known as Neu or HER2), ErbB3 (or HER3), amd ErbB4 (or HER4) - with overall amino acid identity between them of approximately 50%.(8) Notably, ErbB2 has no known ligand and ErbB3 has no kinase activity. In the dimerization action necessary to activate these receptors and send a cascade of downstream signals into the cell, these four related proteins can mix and match, and in this way initiate a huge diversity of downstream signals to fit the needs of the many kinds of cells in which they are present.
ErbB1, or Epidermal Growth Factor Receptor, contains five labile and independent domains: the extracellular binding domain, a single transmembrane alpha helix, a juxtamembrane domain, the kinase domain, and the C-terminal tail. It is comprised of 1186 residues in total (8).
The large and highly glycosylated extracellular binding domain of ErbB1 ( PDB ID = 1ivo), shown here in its biological dimer form, is comprised of residues 1-621, is made up of four separate chains, and can be activated by any of six small polypeptide ligands: Epidermal Growth Factor (EGF; PDB ID = 1egf), Transforming Growth Factor-alpha (TGF-a), Amphiregulin, heparin-binding EGF-like Growth Factor, Betacellulin, and Epiregulin (5,8).
A single-pass transmembrane alpha helix ( PDB ID = 2jwa)occurs at residues 622-644.
The short cytosolic juxtamembrane region, comprised of comprised of residues 645-687, has a number of regulatory functions. is "required for feedback attenuation by Protein Kinase C" - phosphorylation of the residue Threonine-654 by PKC disrupts the ras-raf-MEK-MAPk pathway (8,9).
The kinase domain ( PDB ID = 2gs6), comprised of residues 688-955 and the focus of this molecular document as a whole, is the most highly conserved domain of the five. It regulates the phosphorylation, and thus the activation or inactivation, of the protein overall. The EGFR kinase domain auto-phosphorylates six Tyrosine residues on the C-terminal tail and also catalyzes the transfer of a gamma-phosphate of bound ATP to outside substrates (6).
The long, flexible C-terminal tail, containing residues 956-1186, contains the six main phosphorylatable Tyrosine residues that instigate and modulate the rest of the signalling cascade: Y-992, Y-1045, Y-1068, Y-1086, Y-1148, and Y-1173 (8).
Specifically, the GRB-2 adaptor binds pY1068 and pY1086, the Shc adaptor binds pY1148 and pY1173, the Dok-R adaptor binds pY-1086 and pY1148, phospholipase (PLC) is recruited by pY1173 and pY992, the c-Cbl ubiquitization factor binds pY1045, the SHP-1 phosphatase binds pY1173, and the Abl tyrosine kinase binds pY1086 (6).
THE BASIC MECHANISM OF EGFR ACTION
When EGF binds to EGFR, it induces a conformational change in EGFR that exposes a receptor-receptor interaction site in the extracellular domain of EGFR. It is assumed that the dimerization loop in the unoccupied receptor adopts a conformation that does not facilitate receptor-receptor interactions - only when bound to EGF can two EGFR monomers dimerize. Therefore, when the above-listed ligands are absent, the extracellular binding domain folds back on itself and autophosphorylation does not occur.
The extracellular domain of EGFR dimerizes when a specific loop projects from one of two cysteine-rich clusters on the extracellular region of one receptor and interacts with a specific region (notably, not the same "dimerization loop" structure) in an the adjoining receptor. Point mutations in key residues or deletion of the dimerization loop results in an uncoupling or preventing of EGF to its downstream signal - ligand-induced tyrosine phosphorylation within the cell is strongly reduced.
The extracellular domain then unfolds and symetrically (that is, head-to-head or back-to-back) dimerizes with a similarly-activated neighbor. Thus, in summation, the dimerization of EGFR as a whole requires the binding of two monomers of EGF to two EGFR molecules in a 2:2 EGF:EGFR complex formed from stable intermediates of 1:1 EGF:EGFR complexes (7).
This homodimer or heterodimer brings together the two intracellular kinase domains which bind asymmetrically (that is, head-to-tail), in contrast to the symmetrical extracellular dimer. This ability of the kinase domains to invert is attibuted to the flexible juxtamembrane portion of the protein. The interaction between the kinase dimer components closely resembles that of CDK-2 and cyclin A; further explanation of dimer activity is under a later heading (2).
These united kinase domains then phosphorylate Tyr residues on the long C-terminal tails of the receptors. The unique autophosphorylated dimeric receptor complex initiates a cascade of distinct downstream signaling pathways by recruiting different Src-homology 2 (SH2)- or PTB-containing effector proteins to itself (2).
Now activated, the EGFR dimer forms a complex with the adaptor protein, Grb, which itself is bound to SOS, a guanine nucleotide releasing factor. The Grb-SOS complex either binds directly to phosphotyrosine sites on the receptor's C-terminal tail or indirectly through Shc. These protein interactions bring SOS in close proximity to Ras, activating it. Ras subsequently activates signaling pathways that in turn activate transcription factors (c-fos, AP-1, Elk-1) that promote gene expression (2).
However, receptor activation "not only engages the multiple positively-acting pathways described above, but also sets in motion mechanisms that will ultimately terminate signalling" (6). The signal can be arrested in either of 2 ways. Firstly, by dephosphorylation of the C-terminal Y residues by the enzyme Protein Tyrosine Phosphatase 1B ( PDB ID = 1ptu), acting specifically on residues pY992 and pY1148 (6). Secondly, upon ligand activation of EGFR, c-Cbl proteins are recruited rapidly to the EGFR and mediate its ubiquitination. there is a rapid decrease in the cell surface number of the receptor and an eventual decrease in the cellular content of activated receptors, a process known as down-regulation (9). After this, the EGFR receptor and ligand undergo endocytosis and internalization into clathrin-coated pits, to be degraded in the acidic lysosome or the late endosome (9). This removal of receptors results in signal attenuation, with the different ligand-receptor complexes determining the lifespan and therefore the strength of these signals.
THE BASIC STRUCTURE OF EGFR KINASE DOMAIN
Subunits
The EGFR kinase domain is comprised of 2 subunits, each of which have distinct functions.
The larger A chain (colored aqua), which is 330 residues in length, has a weight of 37563.80 D. 74 waters of hydration are associated with the A chain. The chloride anion ligand binds to chain A at Arg812. Because the A chain is the site of dimer interface in the active conformation of the protein, it is the subunit that truly produces tyrosine kinase activity through trans-phosphorylation of the C-terminal tail and of cytoplasmic signaling molecules.
Its sequence is as follows: GAMGEAPNQALLRIL KETEFKKIKVLGSGA FGTVYKGLWIPEGEK VKIPVAIKELREATS PKANKEILDEAYVMA SVDNPHVCRLLGICL TSTVQLITQLMPFGC LLDYVREHKDNIGSQ YLLNWCVQIAKGMNY LEDRRLVHRDLAARN VLVKTPQHVKITDFG LAKLLGAEEKEYHAE GGKVPIKWMALESIL HRIYTHQSDVWSYGV TVWELMTFGSKPYDG IPASEISSILEKGER LPQPPICTIDVYMIM VKCWMIDADSRPKFR ELIIEFSKMARDPQR YLVIQGDERMHLPSP TDSNFYRALMDEEDM DDVVDADEYLIPQQG
The A chain contains 16 separate alpha helices, the length of which range from 3 residues to 22 residues; This subunit also contains 11 beta-sheets, aligned as a 5 intercalating sheets in the N lobe, then as 3 sets of sheet pairs elsewhere. Details of the secondary structures by location in the A-chain are as follows:
N lobe (lime green superimposed on aqua A-chain):
Short tripeptide helical section: ETE (residues 685-687).
Beta sheet 1: FKKIKVLG (684-695).
Beta sheet 2: GTVYKGLW (700-707).
Beta sheet 3: IPVAIKEL (715-723).
Alpha helix C: PKANKEILDEAYVMA (729-743).
Beta sheet 4: LLGICL (753-758).
Beta sheet 5: VQLITQ (762-767).
C-lobe (superimposed orange on aqua A-chain):
Alpha helix F: THQSDVWSYGVTVWELM (868-884).
Alpha helix G: ASEISSILEK (896-905).
Alpha helix H: IDVYMIMVKC (917-926).
Short tripeptide helical section: ADS (931-933).
Alpha helix I: FRELIIEFSKMA (937-948).
Short tripeptide helical section: PQR (951-953).
Also present is a much smaller polypeptide molecule, the B chain (colored aqua) which is only 13 residues in length and is only associated with 1 water molecule. Its weight is 1544.7 D. The ATP analog-peptide conjugate (colored lime green), Thiophosphoric Acid O-((Adenosyl-Phospho)Phospho)-S-Acetamidyl-Diester (C12H19 N6 O13 P3 S) - hereafter referred to as "ATP Analog 112" - is direct contact with the smaller chain B at the site of Tyr6 (colored purple), and the fact of these minor interactions is its primary structural and functional contribution to the protein as a whole.
Its sequence is as follows: AEEEIYGEFEAKK
A Chain, In Depth
The EGFR kinase domain takes on a bilobate-fold conformation that is inherent in most kinase domains of proteins. The NH2-terminal lobe (N-lobe) is formed from mostly beta-strands and one alpha-helix (hereafter referred to as the "C-helix"), while the large COOH-terminal lobe (C-lobe) is mostly alpha-helical. The two end lobes are separated by a cleft region in which ATP and ATP analogs, like the topical analog Residue 112, can bind. The relationship between N-lobe and C-lobe - that is, the "opening angle" to the cleft region - differs depending on the presence or absence of ATP or an analog (2). It appears that the interlobe angle reduces when ATP binds (1).
Other N-Lobe Points of Interest
The NH2-terminal nine amino acids form several intermolecular contacts including H-bonds involving main chain atoms of residues Asn676 and Leu680, although intramolecular H-bonds between Asn-676 and both Tyr-740 and Ser-744 also contribute.
A salt bridge between the side chains of Lys721and Glu738 interact with the alpha- and beta-phosphates of ATP - this bridge is a highly conserved ionic structure among most N-lobes of activated kinases (1).
"Important [N-lobe] elements of the catalytic machinery bordering the [central cleft region]... include the glycine-rich nucleotide phosphate-binding loop (Gly695-Gly700)" (1).
Other C-Lobe Points of Interest
The C-lobe contributes four notable structures to the area around the cleft: "the presumptive catalytic base Asp813, the catalytic loop (Arg812-Asn818), the activation loop (Asp831-Val852)", and the DFG motif (Asp831-Gly833) within the A-loop(1). The DFG motif of the kinase is involved in geometrically coordinating ATP within the cleft.
The LVI Motif
The C-lobe of EGFRK contains this crucial and distinctive tripeptide sequence Leu955-Val956-Ile957. It regulates the transphosphorylation of substrate - not receptor - tyrosines. Of the three amino acids in this sequence, substitutions for Leu955 are the most deleterious for phospho-transfer activity. The Leu955 side chain is not in contact with outside proteins (including other EGFRs), and is in fact completely buried in a hydrophobic pit; its replacement by an Alanine would seem to disconnect it and nearby residues from the C-lobe and deleteriously impact the transphosphorylation ability of the kinase domain (1).
Catalytic Site
The catalytic-loop sequence RDLAARN is highly conserved and adjacent to the DFG motif (2). In the inactive conformation, Leu834 and L837 pack against alpha helix C, preventing the formation of a Lys721-Glu738 ion pair. In the active conformation of the kinase domain's catalytic site, however, Leu834 and L837 are surface exposed and the Lys721-Glu738 ion pair is intact (2).
The Activation Loop
Protein kinases contain a large flexible loop, called the activation loop, that regulates kinase activity. "In most protein kinases, the activation loop assumes its catalytically competent conformation" only if it is first phosphorylated on a Tyr (1). For these kinases, the "unphosphorylated activation loop is positioned many Angstroms from the active conformation" (1). The A-loop in apo-EGFRK (and EGFRK-erlotinib), however differs significantly: Tyr845 of the EGFRK A-loop, at a position that must be phosphorylated in other RTKs, can be replaced by Phe without loss of function.
Many energetically beneficial interactions stabilize the unique unphosphorylated active EFGR A-loop conformation, most of which are also found in other active kinase A-loops. For example, Tyr845 makes van der Waals contact with the aliphatic part of neighboring Lys836 and the side chain of Arg808 H-bonds to the main chain oxygen of Gly839 (1). Overall, the conformation adopted by the EGFRK A-loop appears to result more from an energetic advantage for the active conformation rather than an energetic disadvantage for an alternate inactive one. It seems, therefore, that the kinase domain is "primed and ready for phospho-transfer" (1).
THE BASIC FUNCTION OF EGFR KINASE DOMAIN
Epidermal growth factor receptor (EGFR) signalling is initiated by ligand binding, followed by receptor homo- and hetero-dimerization, and subsequent autophosphorylation by its kinase domain as well as phosphorylation of other cytoplasmic substrates. This process of receptor activation subsequently results in a signalling cascade that drives a wide range of cellular responses, such as cytoskeletal rearrangements, changes in gene expression, and increased cell proliferation (5). "The regulation of the vital cellular processes influenced by EGFR signaling must be exerted by control of the delivery of COOH-terminal substrate tyrosines to the active site" (1).
The EGFR superfamily differs from the overarching Protein Tyrosine Kinase group in that they do not require "initial [trans-]phosphorylation of kinase domain residues [specifically, the Activation-loop] for full catalytic competency" (7). Instead, the kinase activity of EGFR kinase domain is activated by "ligand engagement in a manner that depends on intermolecular interactions" - that is to say, dimerization (2). The EGFR family contains an dimerization motif that resides between the kinase domain and the carboxyl-terminal phosphorylation sites, with the "greatest effects on receptor function... concentrated in the Leu955-Val956-Ile957 [LVI] segment" (2). This necessary motif leads to ligand-independent dimerization of EGFR intracellular domains - an "inherent catalytic activity of the EGFR family kinases [which is] unique among RTKs" (1). Dimerization results in the phosphorylation of tyrosine residues in the C-terminal tail segments and of the activation loop of the kinase domain, and this "phosphorylation in the A-loop causes [the EGFR] to undergo a large structural reorganization that relieves steric and/or chemical restraints on the catalytic active site" (7). The phosphorylated residues also serve as docking sites for signaling molecules that contain SH2 or PTB domains that are responsible for downstream signal transmission.
The isolated kinase domain has low basal activity, but the activating heterozygous mutation L834R in the activation loop of the kinase domain increases its activity substantially, and is frequently implicated in lung cancers (2). This low basal activity of the monomeric receptor proves that the kinase domain is autoinhibited when alone. In the active conformation of the kinase domain of the receptor, Leu834 - sometimes in conjunction with an adjacent Leu837 - do not play an active structural role. However, in the inactive conformation, these residues pack against hydrophobic side chains in the interior of the kinase domain and also support the orientation of a critical alpha-helix in the N-lobe of the kinase domain. Replacing either of the previously mentioned leucines with an amino acid with a polar side chain, like arginine or glutamine, destabilizes the inactive conformation, prevents the formation of a Lys721-Glu738 ion pair in the helix, and activates the kinase domain improperly (2). This low basal activity of the unmutated monomeric receptor proves that the kinase domain is autoinhibited when alone.
DIMERIZATION ANALOGY
As the specific action of the EGFR Kinase Domain dimer is often compared to the mechanism displayed by Cyclin-A and Cyclin-Dependent Kinase-2 ( cyclin-CDK complex PDB ID = 1hcl) , which are proteins that coordinate and regulate the eukaryotic cell cycle. A brief overview of the Cyclin-CDK pathway therefore informs later discussion of the topical protein.
When the appropriate extracellular signal is received by the cell (often a growth factor, as in EGFR), a cyclin protein and a CDK protein combine to form an activated heterodimer - cyclins have no intrinsic catalytic activity and CDKs are inactive in the absence of a partner cyclin (10). When cyclin is activated by binding a CDK, however, this dimer is able phosphorylates various target proteins, initiating the downstream steps in the cell cycle. Within this dimer, the cyclin performs a regulatory function, while the CDK is the catalytic subunit. One major problem with this analogy is the expression of the receptors: EGFR is found in most cells at all times; CDKs are also constitutively expressed in cells. Cyclins, on the other hand, are only made in cells at certain junctions in the cell cycle, in response to biological signals and timing.
This model, in which one EGFR kinase domain in a dimer acts as a cyclin-like activator for the other kinase, provides a precise molecular basis for the autoinhibition and EGF-induced acquisition of catalytic activity by EGFR (2).
Notably, this discussion deals only with a monomer of the kinase domain, though in its putative active conformation and in complex with an ATP analog.
THE MECHANISM OF EGFR KINASE DOMAIN ACTION
It is postulated that EGFR activation stems from substrate binding and the formation of an asymmetric homo- or heterodimer, which would contain different molecules within the same EGFR family (including ErbB2, ErbB3, and ErbB4). The accepted model for EGFR activation "contains as its central feature the direct communication of the arrival of an external signal to the interior of the cell by the conversion of a monomeric receptor to a ligand-induced dimer" (2). Because the "activation of EGFR has such serious consequences for the cell... [that] nature has clearly evolved secondary levels of control that minimize inappropriate activation of the receptor through [random] intermolecular collisions" (2). The dimerization interface of the extracellular domains is kept hidden until unmasked by EGF, and only then can the "fundamental on/off switch" that is the dimerization of the kinase domains occur (2).
The ligand-induced activation of most Receptor Tyrosine Kinases (RTKs) are mediated and controlled by autophosphorylation of key tyrosine(s) residues in the activation loop of the catalytic kinase domain - in a typical inactive kinase domain dimer, complexed with AMP-PNP (PDB ID = 2gs7), the A-loop is bent so as to block ATP and substrate from accessing its catalytic site. "Upon tyrosine phosphorylation, however, the typical RTK adopts an 'open configuration' enabling access to ATP and substrate" which leads to increased kinase activity (7). This is seen in the fact that point mutations of certain conserved Y residues, such as those listed above, either stem or wholly prevent activation of the kinase domain.
In contrast to most RTKs, in which the activation loop assumes a catalytically significant conformation only when it is first phosphorylated on a Tyr in the A-loop, mutation of the conserved Tyrosine residues in the A-loop of EGFR (ErbB1) is not seriously consequential to further activation and signalling. Receptor activation, therefore, is instead primarily dependent upon the initial ligand-binding and subsequent asymmetric dimer formation, not upon phosphorylation (1,2,7). It may also "indicate that activation loop of EGFR does not play a prominent role in autoinhibition of the PTK domain of EGFR" and that the EGFR is not as strictly autoinhibited in its inactive form (7).
The placement of the A-loop over the catalytic site does contribute to the inactive EGFR structure, however, but in a secondary way. Binding of N-terminal SH2 and SH3 domains to the C-terminal Tyr residues keeps the enzyme in the inactive state by packing in the C-terminal tail. With Y-phosphorylation, the activation loop assumes the described "open configuration" and the C-tail extends, able to interact with downstream receptor protein interactions (7). Still, it is asymmetrical dimerization that is the is the key factor that regulates and facilitates the initial pro-phosphorylation conformation in the activation loop and C-helix, by releasing the C-terminal tail from its association to the SH2 regulatory domain.
As stated before, the interaction between the dimer components closely resembles that of CDK2 and cyclinA; one monomer partner - save ErbB2 - can act as the activated kinase (hereafter referred to as "monomer A" for simplicity), and the other EGFR family homolog "serve[s] as [its] own 'cyclin'" and is not actually activated itself ("monomer B") (2). The C-lobe of one of the 2 interacting kinase domains is arranged to interact with the N-lobe (especially the C alpha-helix) of the other kinase domain in a "3-fold screw rotation" that is facilitated by the dynamic flexibility of the juxtamembrane segment of the receptor (2). The CDK/cyclin-like dimer interface buries 2019 sq. Angstroms of surface area, and is formed by monomer A's contribution of N-terminal residues 672-685 (purple), its C a-helix (red), and the loop between beta-sheets 4 and 5 (green); monomer B's contribution to the interface is the loop between a-helices G and H (purple) and the end of a-helix I (red) (2). The core of this dimer is dominated by hydrophobic interactions, chiefly involving L680, I682, L736, and L758 in monomer A (residues colored yellow)and I917, Y920, M921, V924, M928 in monomer B (residues colored yellow) (2).
The region of dimer interface is shown to be the crucial functionary in kinase activity by various induced point mutation tests. These mutations include " P675G, L680N, I682Q contributed to the interface by monomer A (the activated kinase), and I917R, M921R, V924R, and M928R, involving residues that are contributed to the interface by monomer B (the cyclin-like partner)" (2). All of these mutations almost completely "abrogate the ability of EGFR to phosphorylate all three of the tested autophosphorylation sites, either before or after EGF stimulation" (2).
Because the dimer interface is asymmetric, the model predicts that "an EGFR molecule with a mutation in the C-lobe face of the dimer interface (i.e., with the cyclin-like face disrupted) can be activated by another EGFR molecule that has an intact C-lobe interface... [and that] an EGFR molecule with a mutation in the N-lobe face of the dimer interface (i.e., one that is predicted to be resistant to activation) can act as a cyclin-like activator for another EGFR molecule in which the N-lobe face is intact" (2). Zhang et al. tested these predictions by constructing a "catalytically-dead" form of EGFR, where "Asp813 (the catalytic base in the N-lobe of the kinase domain)[was] replaced by asparagine", denoted as "EGFR kinase-dead" (2). Cotransfecting the kinase-dead N-lobe mutant with the C-lobe mutant V924R still resulted in robust autophosphorylation, as they were able to rescue each other's lost lobal function. The symmetrical dimer does not play a significant role in EGFR activation and is a biological anomaly rather than an alternate conformation to the asymmetric dimer.
Ligand-induced activation of the kinase domain is, to a lesser extent than dimerization, mediated by intermolecular autophosphorylation of certain Tyr residues, in the activation loop of the kinase domain.
THE IMPLICATION OF EGFR KINASE DOMAIN IN CANCER
Over-activated EGFR can convert a normal cell into a malignant one by providing sustained signals for cell proliferation, anti-apoptosis, angiogenesis and metastasis, which are the basic properties of cancer (6).
Illustrating this is an EGFR vIII mutation that is highly prevalent in human glioblastoma, which results in the deletion of residues 6-273. Though this deletion does not directly effect the kinase domain of the EGFR, the in-frame removal of 801 base pairs, with a new glycine codon at the fusion junction, in the extracellular portion of the now-shortened receptor leads to constitutive, ligand-independent kinase domain activity (4,6). Also supporting this observation is that the cellular homolog of EFGR is the avian erythroblastosis virus v-erbB oncogene, which also encodes a constitutively-expressed but extracellularly-truncated EGF receptor (6).
EGFR gene overexpression and EGFR activation, through gene amplification and/or increased transcription of EGFR, are two of the main mechanisms implicated in cancer development (6). Because of this, EGFR and its downstream signaling molecules are relevant targets for therapeutic intervention. Difficulties in developing treatment strategies lie in the fact that there are four ErbB receptors, with different dimerization and ligand-binding patterns; prediction of how an ErbB receptor combination needs to be targeted is therefore an inherent challenge. Also presenting a roadblock are the cases in which tumorigenesis is not dependent on EGFR activity, as these treatements would obviously "not be expected to produce clinical benefit" (6).
Still, there are two main ways to inhibit improper EGFR activation and in this way selectively eradicate cancer cells. The first is to direct monoclonal antibodies (mAbs), like Herceptin or Trastuzumab, against the ligand-binding extracellular domain of EGFR. The second, and more topical, agent is that of small molecule tyrosine kinase inhibitors. In addition, approaches using antisense oligonucleotides and ribozymes that block receptor translation have been developed (6).
Small-molecule Tyrosine Kinase Inhibitors (TKIs) block the activation of the EGFR tyrosine kinase (TK) domain, as the name would suggest. "The basic mechanism of action of these agents is their competitive inhibition of the binding of ATP through interaction with the TK domain of the receptor... resulting in selective inhibition of EGFR autophosphorylation" - that is, they inhibit ligand- and dimerization-induced receptor phosphorylation by competitively binding to the intracellular ATP-binding site of the kinase domain (6). Without kinase phosphorylation, EGFR is unable to activate itself and therefore cannot attract or bind intermediate adaptor proteins. By blocking the signaling cascade in cells that rely on the proliferative pathway, tumor growth should be lessened. These agents are synthetic and generally low molecular weight quinazoline derivatives.
PROTEIN OF COMPARISON: LAPATINIB-EGFR KINASE DOMAIN COMPLEX (GW572016-EGFR)
LAPATINIB / GW572016 / TYKERB
The drug lapatinib (IUPAC name N-[3-chloro-4-[(3-fluorophenyl)methoxy]phenyl]-6- [5-[(2-methylsulfonylethylamino)methyl]-2-furyl] quinazolin-4-amine) is a quinazoline inhibitor that binds to the active site of the kinase domain of the receptor, blocking inappropriate signaling inside the cell; the inactive crystal structure of EGFR kinase domain's A chain when complexed with this drug is given by PDB ID = 1xkk. The Protein Blast and Dali Structural Alignment tests present an E Value of 0, a Z-score of 36.7 and RMSD = 2.4 between the homo sapiens sapiens EGFR kinase domain complexed with a quinazoline inhibitor (lapatinib/GW572016) and the crystal structure of the active homo sapiens sapiens EGFR kinase domain in complex with an ATP analog-peptide conjugate, which tends to indicate similar secondary and three-dimensional structures (3). Though the structure of this comparison protein is very similar to both the inactive apo-protein structure and the structure of the topical protein, which is in the active form and bound to an ATP analog, there are major structural differences in the N- and C-terminal lobes, the C-terminal tail, and the C-helix, and therefore in the protein's ability to bind ATP and remove its phosphates.
NOTE: due to poor electronic density, there are a few residues not included in the final model: six residues at the NH2 terminus, 8 residues at the COOH-terminus, and five short loop regions (residues 710 to 713, 726 to 730, 844 to 851, 964 to 970, and 980 to 983) are not displayed. The illustrated residues and the actual residue numbers, therefore, may not correspond exactly.
The shape of the ATP-binding cleft of EGFR kinase domain depends on the orientation of the NH2- and COOH- terminal lobes of the kinase monomers A and B, which are connected by a flexible "hinge region" (7). Lapatinib binds in this ATP-binding cleft of the EFGR kinase domain, expanding this "back pocket" into a shape larger than that of apo-EGFR or of the Tarceva-EGFR complex (more fully described in a later section). This larger pocket is a result of a 9 Angstrom shift in one end of the C helix to accommodate a bulky 3-fluorobenzyl-oxy group on lapatinib. The 3-fluorobenzyl-oxy group occupies roughly the same space as Met742 in the C-helix, forcing this Met residue to shift. This shift breaks a highly-conserved salt bridge between Glu738 and Lys721 , one that cleaves phosphate groups from ATP (3). Lys721 is then induced to hydrogen bond to the side chain of Asp831, located in the COOH-terminal lobe, while Glu738 points out toward solvent (3). The DFG motif of kinases - part of the activation-loop and involved in coordinating ATP - is also altered by lapatinib's presence. The side chains of Asp831 and Phe832 of that DFG motif form part of the expanded "back pocket" (3).
Within this binding cleft, lapatinib participates in many covalent interactions that stabilize its position. The quinazoline ring is hydrogen-bonded to the "hinge region" of the kinase: N1 of the quinazoline is hydrogen bonded to the main chain NH of Met769, and N3 makes a water-mediated hydrogen bond to the side chain of Thr830. The ring is "sandwiched from the top and bottom by the side chains of Ala719 and Leu820, respectively" (3). The 3-chloro-aniline group is positioned in a pocket formed by the side chains of Val702, Lys721, Leu764, Thr766, Thr830, and Asp831 (3). The 3-fluorobenzyloxy group occupies a pocket formed by the side chains of Met742, Leu753, Thr766, Thr830, Phe832, and Leu834 (3). The aniline nitrogen and the ether oxygen of lapatinib are not involved in any direct hydrogen bonding interactions with the protein.
In addition, residues 971-980 in the COOH-terminal tail form a short alpha helix that partially blocks the ATP binding cleft. The ATP binding site of the protein complexed with lapatinib is therefore effectively in a closed conformation.
GW572016-EGFR is found in an inactive-like conformation. It is unknown as to whether the drug binding binding induces this change from active (or partially-active) to inactive or whether lapatinib binds to the pre-existing pool of EGFR and maintains its inactive status (3). The fact that lapatinib binds to or induces the inactive conformation is advantageous in that it dramatically reduces the rate of inhibitor dissociation (or, alternatively, increases the dissociation constant), which means the drug has long-lasting effects even after its concentration levels drop. It also appears that the "dissociation of [lapatinib] may require a conformational change in EGFR... a slow protein conformational change may explain the slow dissociation rate" (3).
TWO NOTABLE DRUG-PROTEIN COMPLEXES
Tykerb (lapatinib), Tarceva (erlotinib), and Iressa (gefetinib), as small molecule tyrosine kinase inhibitors (TKIs), down-regulate tyrosine-phosphorylation on the C-terminal tails of the EGF family of receptors. These phosphorylated sites often act as docking sites for downstream signal transduction molecules, and thus can be inhibited in order to block inappropriate biological activity by an activated EGFR. These drugs all share a common 4-anilinoquinazoline core, but all have different inhibition strategies and mechanisms of action. Thus, specific drugs may select or target different conformations (inactive, phosphorylated active, dephosphorylated active) of what is technically the enzyme, but with a different tertiary structure (3).
A key challenge with EGFR kinase domain treatments is that after 8 to 12 months, the cancer cells become resistant. The primary source of resistance is a mutation that appears to recruit a mutated Insulin-like growth Factor-1 Receptor ( IGF-1; PDB ID = 1gzr) to act as one of the EGFR partners in the catalytic heterodimer. The signal to grow is thus communicated downstream even in the presence of an EGFR inhibitor drug - a consequence of the asymmetric dimer formation, in which one partner can be "catalytically dead" and still have its functionality rescued.
Erlotinib and gefitinib are selective inhibitors of EGFR (ErbB1), while lapatinib is a potent inhibitor of both EGFR and ErbB2. Also, lapatinib captures the inactive conformation of the EFGR kinase domain, while erlotinib and gefetinib inhibit the active conformation. The off-rates of these drugs are also highly disparate, and may link into which conformation of the receptor that the drug captures - that is to say, because erlotinib and gefitinib inhibit a conformation of the receptor that is almost identical to the apo-EGFR protein, the fast "off-rate may reflect the fact that the inhibitor binds and dissociates from an active form of the enzyme without requiring any major changes in protein conformation" (3).
ERLOTINIB / OSI-774 / TARCEVA
N-(3-ethynylphenyl)-6,7-bis(2-methoxyethoxy)quinazolin-4-amine, the IUPAC name of erlotinib (colored mint), in complex with EGFR ( PDB ID = 1m17) has a rather different structure than that of the lapatinib-EGFR complex, though they both bind in a reversible fashion to the ATP binding site of the receptor. These differences include the shape of the ATP binding site, the position of the C helix, the conformations of the COOH-terminal tail and activation loop, and the hydrogen bonding pattern with the quinazoline ring of the inhibitors. As stated previously, the NH2- and COOH-terminal lobes of kinases are connected by a flexible hinge region. The relative orientation of the two lobes with respect to one another influences the shape of the ATP binding cleft and is dependent on the activation state of the kinase and the presence of ligands. Both lapatinib and erlotinib cause EGFR inhibition by competitively binding to the site occupied by ATP during phospho-transfer; however, while the opening to the ATP binding cleft of lapatinib-EGFR is in a relatively closed conformation, the erlotinib-EGFR complex ATP binding cleft (residues in cleft colored red) is in a more open conformation.
Finally, the quinazoline rings of erlotinib and lapatinib hydrogen bond with EGFR differently. The quinazoline N1 of both compounds accepts a hydrogen bond from the main chain amide nitrogen of Met769. However, while the N3 of lapatinib makes a water-mediated hydrogen bond to the side chain of Thr830, the N3 of erlotinib makes a water-mediated hydrogen bond with the side chain of Thr766 (in red).
GEFITINIB / ZD-1839 / IRESSA
Gefitinib ( PDB ID = 2ity), or 4-(3-chloro-4-fluoroanilino)-7-methoxy-6-(3-morpholinopropoxy) quinazoline, was the first EGFR tyrosine kinase inhibitor, approved for lung cancer treatment by the FDA in May 2003 by the FDA (5). When administered with other TKIs or with monoclonal antibodies, gefitinib can display synergistic inhibitory effects on angiogenesis and rapid regression of tumors. When drug administration was discontinued, however, the tumors regrew, "suggesting that long-term drug administration will be required to maintain a tumor response in patients" (5).
Gefitinib does not cause the dephosphorylation of the pre-existing activated receptor or block the kinase domain, but instead inhibits phosphorylation on these residues by competing with ATP for its binding site.
DISSOCIATION RATES COMPARED
The dissociation rate of the inhibitor-receptor complex relates to and influences the potency or duration of receptor inhibition. This inhibition is best gauged by the duration of down-regulation of autophosphorylation in oncogenic cells, quantified by the "dissociation constant for an inhibitor, Ki, [which] is the best standard of comparison for potency and selectivity" (3).
For gefetinib, Ki was .4 nM (peptide concentration fixed and drug concentration varied); for erlotinib, Ki was .7 nM. Notably, lapatinib's Ki was incredibly high at 3 nM, while still maintaining the inhibition reversibility (3).
In a "washout" test, where the recovery of receptor autophosphorylation action was compared with the time from which the inhibitory drug dissociated, lapatinib-treated cells were notably slow to recover: the receptors had only recovered 15% of their control-level phosphorylation after 96 hours (3). In contrast, erlotinib-treated cells recovered control levels of phosphorylation 24 hours after washout, and gefetinib-treated cells recovered 60% control activity after 72 hours (3).