Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

5.1 Introduction

Protein kinases are a group of enzymes which play a significant role in every aspect of cellular metabolism. [1] Regulation of kinase activity is crucial in various biological processes like proliferation, apoptosis, cell cycle, differentiation, development, and transcription. [2] Dysregulation of kinase activity has been associated with a variety of diseases including cancer [3,4,5], diabetes [6], autoimmune, cardiovascular [7], inflammatory [8, 9] and nervous system, etc [10].

The human genome contains 518 (478 typical and 40 atypical) protein kinase genes; they constitute about 2% of all eukaryotic genes. The typical protein kinases are grouped into three major groups (i) protein-serine/threonine kinases (385), (ii) protein-tyrosine kinases (90), and (iii) tyrosine-kinase like proteins (43) based on the phosphorylation of the –OH group of amino acid on target proteins.

The major function of protein kinases is the phosphorylation of proteins that can either stimulate or inhibit the protein functionality. The kinases as mediators of protein phosphorylation are very important in disease pathophysiology (e.g. cancer) by means of mutational activation or by helping the neoplastic growth. Though their activity is tightly regulated in normal cells, they may acquire transforming capabilities due to mutations, overexpression and autocrine-paracrine stimulation. Mutationally activated kinases are in constant state of activation, and with their transforming capability, they become ideal for survival and growth of the cancer cell leading to the dependence of cancer cells (oncogene addiction). It also renders these cancer cells highly susceptible to specific kinase inhibitors [11, 12]. Kinases are considered one of the most important classes of drug targets and design and development of specific kinase inhibitors, has therefore, became a major strategy in drug discovery programs. The interest of biomedical research in kinases can be gauzed from the fact that about 5000 crystal structures [13] and 400,000 publications have been reported in PDB and PubMed, respectively, in the last three decades.

5.2 The Structure of Protein Kinase Domain

The elucidation of the crystal structure of PKA in 1991 by Dr. Susan Taylor’s group was a significant achievement for understanding the molecular basis for kinase function. Since this discovery the crystal structures of nearly 200 different protein kinases have been determined, amounting to more than 5000 X-ray structures being publicly available.

The protein kinases have a small N-terminal lobe and a large C-terminal lobe in their highly conserved kinase domain. The two lobes are joined by a short hinge region which helps in opening and closing of the kinase structure [14] (Fig. 5.1a). The ATP binds near hinge region between these two lobes. The small lobe comprises five stranded anti-parallel β-strands (β1–β5) and a αC-helix. The N-terminal has a conserved glycine-rich (GxGxxG) loop in between the β1 and β2 strands. This loop is known as p-loop that helps in positioning β and γ-phosphates of ATP for catalysis. The glycine-rich loop is followed by conserved valine residue which makes hydrophobic contact with adenine of ATP. The third β strand has a sequence of Ala-xxx-Lys, from which the lysine residue couples with α and β phosphates of ATP to αC-helix. The conserved glutamate residue in the center of αC-helix plays an important role in activation and inactivation of the kinases. The formation of a salt bridge between the glutamate of αC-helix and lysine of β3 strand corresponds to αC-in (active) conformation [15], whereas the absence of it leads to αC-out (inactive) conformation of kinases (Fig. 5.1b). The large lobe comprises six α-helical segments (αD–αI) and four short conserved β-strands (β6–β9) [16]. The αE-helix is followed by the β6 strand, the catalytic loop, the β7 strand, β8 strand, and the activation segment containing the β9 strand. The activation segment forms an open structure by extending itself from the catalytic loop and forms the active state of the kinase, thus allowing the binding of substrates. The catalytic loop of kinases has a signature motif of K/E/D/D (Lys/Glu/Asp/Asp) from which the first aspartate residue facilitates the nucleophilic attack of the hydroxyl group onto the γ-phosphorous atom of ATP after withdrawing a proton from –OH group. The second aspartate introduces as the first residue of activation segment and plays an important role in activation and deactivation of protein kinases. In most kinase, the activation segment comprises 20–30 amino acids in length, starts with D/F/G (Asp/Phe/Gly), and ends with A/L/E (Ala/Leu/Glu) signature motifs [17]. The kinases, in general, can interconvert between two conformational states viz. active and inactive. In the active kinase conformation, the aspartate of DFG motif points into the ATP-binding site and coordinates with two Mg+2 ions; thus, the activation segment acquires an open and extended conformation (Fig. 5.1c). In contrast, an inactive kinase conformation is characterized by a flipped conformation (180°) of the DFG motif (i.e.) the aspartate of the DFG motif moves away from the ATP binding site (Fig. 5.1d). The structural features signifying the active and inactive kinase conformations are summarized in Table 5.1.

Fig. 5.1
figure 1

Structural features of protein kinases. (a) The 3D structure of a kinase showing N-terminal lobe, hinge region, and C-terminal lobe (b) Structures of kinases showing both αC-in and αC-out conformations, activation segment open and close conformations (Alignment of structures PDB ID: 1M17 and PDB ID: 4HJO) (c) Kinase structure showing an active DFG-in conformation (PDB ID: 2GQG) (d) Kinase structure showing an inactive DFG-out conformation (PDB ID: 3K5V)

Full size image
Table 5.1 Summary of regulatory features within the kinase domain
Full size table

5.3 Protein Kinase Inhibitors

The kinases gained great attention from medicinal chemists for design of their modulators. Since the discovery of Imatinib by Novartis in 2001 many pharmaceutical companies have reported a variety of inhibitors and about 33 kinase inhibitors have been approved by USFDA by March 2017. These inhibitors are classified into different types depending on their binding in the kinase domain. The Table 5.2 enlists all the small molecule kinase inhibitors approved by USFDA.

Table 5.2 The FDA-approved kinase inhibitors
Full size table

5.4 Classical or ATP-Competitive Inhibitors

The type I inhibitors bind to the active kinase conformation forming H-bonds with the kinase hinge region residues and occupy the adenosine binding pocket. Their binding is independent of the conformation of key structural elements, e.g., helix αC and the DFG. Since the residues at and near the ATP binding site are highly conserved and the unique shape of the adenine site allows for only a little variation in the heterocyclic system, almost all of the inhibitors share only a small number of heterocyclic rings and are typically entropically constrained. Consequently, it is difficult to design type I inhibitors with high selectivity [18].

The type II kinase inhibitors bind to and stabilize the inactive conformation of the kinase thereby preventing the binding of ATP and subsequent activation of the kinase. Similar to type I inhibitors they also occupy adenosine binding pocket. However, they induce a configuration of DFG residues termed DFG-OUT [19]. These inhibitors bind to the ATP binding site and thus are considered as ATP-competitive inhibitors. However, the inhibitors that bind to the inactive conformation face weaker competition from cellular ATP. They may act primarily by shifting the equilibrium between conformational states in a way that prevents kinase activation, rather than by inhibiting kinase activity directly [20]. An advantage of type II inhibitors over type I inhibitors is due to the fact that the amino acids surrounding the newly exposed pocket (due to DFG shift) are less conserved as compared to those in the ATP binding pocket. Thus, it is possible to design inhibitors exploiting this difference to achieve better kinase selectivity [21, 22].

The inhibitors of type-I and type-II are further divided into subtypes A and B depending on their interactions with the residues present in the kinase domain. The type A inhibitors mainly bind to the residues in the front and back cleft, gatekeeper area, and the region separating the small lobe from large lobe. The type B inhibitors bind to the residues present in the front cleft and gatekeeper area only.

5.5 Allosteric or Noncompetitive Inhibitors

Mutations resistant to classical ATP-competitive (Type I/II) inhibitors are quickly emerging and as such often limit the success of targeted cancer therapies. These mutations often result in a steric hindrance obstructing inhibitor binding to the hinge region of the ATP pocket. Some of these mutations may result in increased affinity for the ATP that shifts balance against a competitive inhibitor. The allosteric inhibitors constitute a group of structurally diverse compounds that bind in sites other than the ATP binding site. These inhibitors (e.g., type III inhibitors) have shown promise toward addressing mutation-dependent drug resistance. They have also been utilized to make more selective inhibitors as they bind in a remote pocket that may not be conserved. Therefore, the identification and development of such inhibitors is the focus of many drug discovery projects.

Type III inhibitors: Certain kinases have an allosteric pocket, adjacent to ATP binding site, where an inhibitor can bind along with ATP. These inhibitors do not interact with hinge residues; they block kinase activity without displacing ATP [23]. We will discuss some of these inhibitors below.

The ERK/MAP kinase cascade dysregulation is implicated in cancer. Mutations in upstream RAS and Raf occur often and contribute to the oncogenic phenotype through activation of MEK and then ERK. The Ras mutations lead to activation of the Raf-MEK-Erk kinases. A lot of research has been done on identification of inhibitors for these kinases. The MEK kinase is one of the targets for which type III inhibitors have been identified. The Parke-Davis identified an inhibitor (PD 098059) that prevented MEK activation of MAPK. This molecule showed a high degree of selectivity over closely related kinases. The co-crystal structure of MEK1 with another compound (PD 318088) showed co-binding of ATP and the inhibitor (PDB ID: 1S9J). This was the first type III inhibitor that binds in the active site without interfering with ATP binding [24, 25]. Another molecule Trametinib is a first US-FDA approved type III inhibitor targeting MEK1 (mitogen-activated kinase-1) for the treatment of B-raf mutated (V600K/E) metastatic melanoma [26, 27, 28]. It binds in the allosteric back pocket region with DFG-in configuration (Fig. 5.2a). The allosteric back pocket refers to a distinctive pocket adjacent to the ATP binding pocket.

Fig. 5.2
figure 2

Type III allosteric inhibitors (a) Inhibitor with MEK1 (PDB ID: 1S9J) (b) Inhibitor with FAK (PDB ID: 4EBW) (c) CDK2 kinase (PDB ID: 3PXF) (d) LIMK2 kinase (PDB ID: 4TPT)

Full size image

Iwatani et al. reported a highly selective series of 1,5-dihydropyrazolo[4,3-c][2, 1]benzothiazines for allosteric inhibition of focal adhesion kinase (FAK) involved in regulation of cellular survival and proliferation (Fig. 5.2b). These compounds were non-ATP compititive inhibitos of FAK [29]. The co-crystal structural analysis revealed that these inhibitors specifically bind to a novel allosteric site within the C-lobe and induce disruption of ATP pocket formation. Interestingly, the phosphorylation of FAK leads to a reduction in allosteric inhibition potency. The structure activity relationship analysis of these compounds indicated that N-substitution of the pyrazole ring is important for achieving allosteric binding and high selectivity among kinases. Tomita et al. employed synthetic medical chemistry approach for the development of potent and selective FAK inhibitors [30]. They used pyrazolo[4,3-c][2, 1]benzothiazines to target the FAK allosteric site. The lead molecule had significant FAK kinase inhibitory activities for cell-free (IC50 = 0.64 1M) and cellular assays (IC50 = 7.1 1M).

Diarylamine compounds form a major group of allosteric MEK1/2 inhibitors. The molecule cobimetinib has been studied in combination with vemurafenib for the treatment of B-Raf V600E/K mutation-positive advanced melanoma. In 2015, the U.S. Food and Drug Administration (FDA) approved the use of cobimetinib in combination with vemurafenib to treat patients with advanced stages of melanoma [31, 32].

The compound MK-2206 reported as Akt inhibitor binds to inactive Akt conformation by targeting an allosteric site at the interface of the kinase domain and the pleckstrin homology (PH) domain. It is currently under investigation in clinical studies on breast cancer, NSCLC, nasopharyngeal carcinoma, and other cancers [33].

The displacement of the structural αC helix with type III allosteric inhibitors has recently been exploited in drug discovery. A crystal structure of CDK2 with an open allosteric pocket adjacent to the αC helix has been reported (Fig. 5.2c). Rastelli et al. identified CDK2 allosteric inhibitors with micromolar potency through docking-based virtual screening. These compounds bind into an allosteric pocket of CDK2 formed following displacement of the αC-helix [34]. Godwin et al. reported a novel series of sulfonamides as potent and selective inhibitors of LIM-kinase 2 (LIMK2). The kinetic experiments further revealed that these molecules were non-ATP competitive inhibitors of LIMK2. Structural analysis by X-ray crystallography revealed that these molecules bind in a hydrophobic pocket near the ATP binding pocket with DFG-out orientation (Fig. 5.2d) [35].

Type IV inhibitors: These inhibitors bind at a site remote from the ATP-binding e.g., surface pockets and interfere with binding of key regulators. These sites can be present anywhere in the kinase domain other than the site adjacent to the ATP binding site. The type IV inhibitors are considered as allosteric or noncompetitive inhibitors of ATP as they don’t hamper the ATP binding site. Another significant characteristic of type IV inhibitors is that they induce conformational changes which make the kinase inactive [35,36,38].

A library of highly functionalized pyrazolo[3,4-d]pyrimidines, with a high level of molecular diversity, has been reported by Vignaroli et al. The enzymatic screening of this “privileged scaffold”-based compound collection, showed high activity against Src, Abl wt, and T315I ABL. The study has led to the development of a new allosteric inhibitor of the T315I ABL. The most potent compound showed an IC50 value of 3.16 μM against Abl T315I, independent of the concentration of ATP and the peptide substrate [39].

A molecule GNF-2 bind to the myristoyl pocket of the C-lobe of the kinase domain is the first reported type IV inhibitor of Abl. A series of 1,3,4-thiadiazole compounds were reported as promising Abl inhibitors. The lead compound BO1 inhibited T315I ABL in an ATP-independent manner signifying allosteric mechanism of inhibition [40, 41].

Yamada et al. reported ATP noncompetitive WNK1–4 kinase inhibitors as next-generation anti-hypertensive agents. The co-crystallization of the inhibitors with WNK1 revealed an allosteric binding mode consistent with the observed specificity for WNK1–4 kinases. The optimized compound inhibited rubidium uptake by sodium chloride co-transporter 1 (NKCC1) in HT29 cells [42] (Fig. 5.3a, b).

Fig. 5.3
figure 3

The crystal structures of WNK1. (a) Structure representing the binding of inhibitor (magenta) in human WNK1 with AMP-PNP. The allosteric pocket is located near the ATP binding site. (PDB ID: 5TF9) (b) Structure representing ATP competitive inhibitor WNK463 in human WNK1 (PDB ID: 5DRB)

Full size image

The type V inhibitors: They bind to two different sites on the kinase domain. They are further divided into two categories, i.e., bisubstrate analog inhibitors that target both the ATP and protein substrate binding sites, and bivalent inhibitors that target the ATP binding cleft and other surface (apart from substrate binding site) on the protein kinase. The design of bisubstrate analog inhibitors is based on the fact that different protein kinase group show significant variation in their substrate recognition [43]. Parang et al. designed a bisubstrate analog inhibitor by covalent linking of ATPγS to an analog of the peptide substrate, IRS 727, by a short two carbon spacer with a distance of approximately 5.7 Å between the tyrosine nucleophilic atom of the IRS 727 analog and γ phosphorous of the ATP moiety [44]. The bivalent inhibitors are designed to target both the catalytic and regulatory domains of the kinases. Profit et al. designed such an inhibitor for tyrosine kinase by tethering of an active site-directed peptide sequence with an SH2 domain recognition sequence through a flexible linker comprising γ-amino butyric acid [45].

The type VI inhibitors: They bind covalently to the kinases by formation of a covalent bond between the alkene portion of the inhibitor and the cysteine residue present within the ATP-binding site in some of the kinases. A beautiful example of selective inhibitor design by exploiting the presence of certain residues at specific positions has been reported by Cohen et al. The authors have noted that threonine, which is a small gatekeeper, provides only a partial discrimination between kinase active sites; therefore, if a second selectivity filter can be applied it may result in a more selective inhibitor. They discovered that there are only three kinases (RSK1, RSK2, and RSK4) which have a highly reactive cysteine on the P-loop (the high reactivity of cysteine owing to its proximity to solvent accessible surface and thus lower pKa compared to buried cysteine) and a threonine as gatekeeper. They designed and tested some analogs targeting these two residues and found two selective and irreversible kinase inhibitors [46].

The inhibitors from type-I to type-V are reversible in nature whereas type VI inhibitors are irreversible.

Another example of targeting a surafce exposed Cystein is the design of molecules afatinib and ibrutinib. The alkene part of afatinib makes a covalent interaction with the hinge cystein in the ATP binding pocket. [47, 48].

5.6 Concluding Remarks

The structural and physiological knowledge of protein kinases have improved the understanding of kinase function at molecular level. The drug discovery targeting protein kinases has achieved substantial progress since the discovery of first kinase inhibitor. The success of many kinase inhibitors has propelled it further, and it seems to be the most desirable field in biomedical research after oncological research [49].

In the current chapter, the story of kinase inhibitor design with special reference to allosteric site has been discussed. The review is focused mostly on the allosteric inhibitors, their binding in the kinase domain, and analysis of available crystal structures. Our suggestions for self-motivated researchers in form of future direction of research may include in-depth structural studies of kinases with special reference to DFG-in and DFG-out conformations. Analysis of residue–residue interactions through interaction networks to explore the transformation of signals during binding of different inhibitors and long-term molecular dynamics simulation of reported crystal structures with allosteric inhibitors to decipher the dynamics of the allosteric site(s).

The human kinome is classified into seven major kinase families. However, the discovery of small molecule inhibitors so far is limited to selected group of kinases. The majority of 33 inhibitors approved to date are confined to the tyrosine kinases, serine/threonine kinases, and tyrosine-like kinases family. It indicates that there is a tremendous scope in this field for the future discovery research.