**3. Kinase inhibitors**

and enable considerable conformational flexibility. In the inactive conformation the P-loop may be largely disordered. In the active kinase conformation the P-loop may resemble more of a classic beta-turn and is able to interact with the triphosphate group of bound ATP. The Aloop includes one or more tyrosine residues that when phosphorylated induce the conversion from the inactive to the active kinase conformation. The A-loop is long, containing around 27 amino acid residues that may fold into a number of isolated single turn helices. The A-loop has a conserved DFG sequence, where in the non-phosphorylated inactive kinase conforma‐ tion the phenylalanine sidechain is oriented toward the triphosphate binding area of the ATP binding site, the so called DFG-out conformation. In the phosphorylated active kinase conformation, the DFG phenylalanine sidechain is oriented toward the interior of the protein, the so-called DFG-in conformation [5-7]. While the DFG orientation related to active and inactive kinase conformations is clear in some cases it is not clear in others, and is not as general in differentiating active and inactive conformations as first thought. It has subsequently been found through surface comparison of x-ray derived crystal structures of both active and inactive forms of specific kinases, that the assembly of a hydrophobic spine through sidechain interaction of discontiguous amino acids within the N-lobe, C-lobe and activation loop regions offers a better structural definition of kinase activation for all kinase families [8]. Figure 2 demonstrates the hydrophobic spine assembly of a representative kinase. Both active and inactive kinase conformations are able to bind ATP. However, only the active kinase confor‐

Protein kinases are important to many biological processes, as is evidenced by the more than 500 kinases encoded in the human genome. Kinases are involved in processes ranging from modulation of intracellular trafficking of proteins, to assembly of complicated multiprotein complexes, to gene expression. Kinase modulation of gene expression through signal trans‐

The underlying cause of cancer is genomic instability. The emergence of a tumor cell typically requires multiple mutations in the genome that both activate oncogenes and inactivate tumor suppressor genes. These mutations synergize to transform a cell to a state of uncontrolled growth, proliferation, and enhanced survival. Mutations and/or epigenetic alterations that increase expression of, or produce constitutive active forms of some kinase proteins are known to drive the proliferation and/or survival of some tumors [9]. The types of kinases that are often mutated in tumor cells, although not exclusively, are found in signal transduction pathways. These pathways are kinase cascades that transmit extracellular signals to the nucleus and ultimately affect gene expression patterns. When a kinase in such a pathway is mutated where it is constitutively active, the extracellular signal is no longer needed and the pathway is always active and not regulated. Alternatively, epigenetic alterations or an increase in copy number of a protein kinase gene may cause overexpression of the kinase to increase activation of a signaling pathway. For example, a significant portion of non-small cell lung cancers (NSCLC) over express or have EGFR mutations. This causes over active signaling through one or more of the Ras/Raf/ERK, PI3K/AKT/mTOR, and STAT pathways that lead to increased growth,

mation can perform the kinase function on substrates.

16 Updates on Cancer Treatment

duction is of particular interest for cancer therapy.

proliferation, motility and survival of the tumor cells.

The natural product staurosporin was one the first small molecules discovered that inhibits the function of protein kinases [14]. Staurosporin is a non-selective inhibitor and is not amenable to development as a drug. However, one staurosporin-like molecule, midostaurine is in clinical development for oncology indications such as indolent systemic mastocytosis [15].

Imatinib, the first marketed kinase inhibitor for treatment of cancer, was discovered by scientist at Ciba-Geigy in the 1990s. It was determined that imatinib inhibited the function of the BCR-Abl fusion kinase that drives chronic myeloid leukemia [16]. Since this time (and perhaps before) the primary rational approach to design of kinase inhibitors has been to target the ATP binding site. The first experimentally determined structure of a small molecule bound to the ATP binding site of a kinase appeared in the year 2000 [17]. This structure was groundbreaking, in that it showed imatinib to be buried deeply in the ATP binding site of the kinase thus locking it in an inactive conformation. This structure provided data that could be used for a structure based design approach of future kinase inhibitors. The structure revealed a pose of imatinib in the ATP binding site that is today designated the type II binding mode.

Clinically used ATP site binding direct kinase inhibitors fall within two categories that describe their binding modes. Type I inhibitors may bind both active and inactive conformations of a particular kinase target. Type II kinase inhibitors bind only inactive conformations of a kinase. Both type I and II inhibitors bind in the ATP binding site and make use of hydrogen bonds to the amide backbone of hinge region amino acid residues. The difference between type I and II inhibitors is that type II inhibitors typically penetrate deeper into the pocket accessing what is termed an allosteric binding site, the effect of which is to displace the C-helix and locking the kinase in an inactive conformation. Knowledge of binding geometries has been applied to kinase inhibitor design projects.

The majority of kinase inhibitor design projects have focused on ATP competitive inhibi‐ tors targeting the ATP binding site. The popularity of such programs have spurred the growth of contract research screening services that screen potential inhibitors in a high throughput format for binding to the ATP binding site and/or inhibition of target kinase function [18]. Design programs may be conducted in a variety of manners from high throughput screening of compound libraries to fragment based and/or structure guided design. One approach is to start from known privileged fragments or templates - chemi‐ cal moieties that appear in numerous kinase inhibitors - such as the quinazoline ring system [19]. The quinazoline ring system appears in many experimental kinase inhibitors, clinical candidates (tandutinib), and kinase inhibitor drugs such as afatinib, vandetinib, lapatinib, erlotinib and gefitinib (see Table 1).

**Figure 3.** Kinase hinge peptide backbone (A) hydrogen bonding pattern to ATP heterocycle (B) and other heterocycles found in drug molecules. (C) quinazoline, e.g. erlotinib. (D) pyridine, e.g. imatinib. (E) pyrrolopyridine, e.g. vemurafe‐ nib. Dashed lines represent hydrogen bonds.

It appears that a unifying characteristic of kinase inhibitors that bind to the ATP binding site is a heterocyclic moiety that serves as a mimic of the adenine ring system and forms Van der Waals contacts with hydrophobic groups of the floor and ceiling of the ATP binding site as well as one or more hydrogen bonds with the amide backbone of hinge region amino acid residues. Figure 3 describes how ATP and some heterocycles found in kinase inhibitors form their hydrogen bonding patterns to the hinge region. Because the ATP binding site is conserved among kinases, kinase inhibitors tend to be promiscuous and inhibit numerous kinases other than their primary target. How the final kinase inhibitor is adorned about its hinge binding moiety contributes to its overall shape and possible intermolecular interactions, ultimately determines its degree of selectivity for its intended target.

Another unifying theme of kinase inhibitors is that their binding to their kinase targets displaces the alignment of the hydrophobic spine that is formed by kinase activation [8]. Besides the occupation of the ATP binding site by the kinase inhibitor, disallowing binding and use of ATP, the kinase inhibitor disrupts the formation of the hydrophobic spine necessary for attainment of the active conformation (Figure 4). This observation leads to the inevitable question of whether inhibitors, that disrupt the hydrophobic spine, but do not bind to the ATP binding site, may be developed.

the amide backbone of hinge region amino acid residues. The difference between type I and II inhibitors is that type II inhibitors typically penetrate deeper into the pocket accessing what is termed an allosteric binding site, the effect of which is to displace the C-helix and locking the kinase in an inactive conformation. Knowledge of binding geometries has been applied to

The majority of kinase inhibitor design projects have focused on ATP competitive inhibi‐ tors targeting the ATP binding site. The popularity of such programs have spurred the growth of contract research screening services that screen potential inhibitors in a high throughput format for binding to the ATP binding site and/or inhibition of target kinase function [18]. Design programs may be conducted in a variety of manners from high throughput screening of compound libraries to fragment based and/or structure guided design. One approach is to start from known privileged fragments or templates - chemi‐ cal moieties that appear in numerous kinase inhibitors - such as the quinazoline ring system [19]. The quinazoline ring system appears in many experimental kinase inhibitors, clinical candidates (tandutinib), and kinase inhibitor drugs such as afatinib, vandetinib, lapatinib,

**Figure 3.** Kinase hinge peptide backbone (A) hydrogen bonding pattern to ATP heterocycle (B) and other heterocycles found in drug molecules. (C) quinazoline, e.g. erlotinib. (D) pyridine, e.g. imatinib. (E) pyrrolopyridine, e.g. vemurafe‐

It appears that a unifying characteristic of kinase inhibitors that bind to the ATP binding site is a heterocyclic moiety that serves as a mimic of the adenine ring system and forms Van der Waals contacts with hydrophobic groups of the floor and ceiling of the ATP binding site as well as one or more hydrogen bonds with the amide backbone of hinge region amino acid residues. Figure 3 describes how ATP and some heterocycles found in kinase inhibitors form their hydrogen bonding patterns to the hinge region. Because the ATP binding site is conserved among kinases, kinase inhibitors tend to be promiscuous and inhibit numerous kinases other than their primary target. How the final kinase inhibitor is adorned about its hinge binding moiety contributes to its overall shape and possible intermolecular interactions, ultimately

Another unifying theme of kinase inhibitors is that their binding to their kinase targets displaces the alignment of the hydrophobic spine that is formed by kinase activation [8].

kinase inhibitor design projects.

18 Updates on Cancer Treatment

erlotinib and gefitinib (see Table 1).

nib. Dashed lines represent hydrogen bonds.

determines its degree of selectivity for its intended target.

Considering the mountain of published research and plethora of X-ray diffraction and NMR derived structures of kinase inhibitors bound to their target kinases, the design of kinase inhibitors is fairly well understood. The remaining challenges for drug designers, in the area of kinase inhibitors, are concerned with selectivity. Selectivity is a potential problem for kinase inhibitors, thus far, because most target the highly conserved ATP binding site. Targeting a single family of kinases, let alone a single member of a family, is difficult because of the inherent promiscuity of this drug class. An even more difficult challenge is targeting a cancer relevant mutant kinase selectively with respect to the wild type kinase. The selective targeting of a mutant kinase that drives the proliferation and survival of a patients' tumor has great promise not only in potential survival benefit but also reduction in drug side effects.

**Figure 4.** EGFR kinase domain (ribbon) bound to kinase inhibitor lapatinib (stick). Note that the amino acid sidechains that form the hydrophobic spine (surface) are bowed out of alignment.

Common side effects caused by kinase inhibitors include rash, fatigue, and gastrointestinal disturbances such as nausea, vomiting, diarrhea, or constipation. For example, patients receiving therapy with EGFR inhibitors may experience severe acneiform eruption and diarrhea which may be dose limiting [20]. Additionally, the EGFR inhibitor gefitinib has been associated with interstitial lung disease, especially in patients with underlying pulmonary diseases [21]. The more promiscuous kinase inhibitors such as sunitinib and sorafenib exhibit some cardiotoxicites such as hypertension, LVEF and QT prolongation [22]. Additional less severe side-effects may be experienced by patients receiving kinase inhibitors such as hair depigmentation which accompanies KIT inhibitors [23]. Even though kinase inhibitors are considered targeted therapy the patient will experience side effects that can be associated with the mechanism of action, off target effects and intrinsic chemical effects. Almost as common as side effects are associated with anticancer drugs so is tumor resistance. And just as other anticancer drugs may encounter tumor resistance, a kinase inhibitor may also encounter resistance either acquired or intrinsic.

Tumor resistance to anticancer drugs can be a challenging battle even though a lot has been learned about the mechanisms that allow tumor cells to escape the chemical assault. One major mechanism of tumor resistance to kinase inhibitors is mutant forms of the target kinase, either preexisting or acquired during or after treatment, that reduce the kinase inhibitors binding affinity. Early clinical experience with imatinib identified patients that developed resistance after an initial response to the drug due to mutation of the ACR-Abl kinase that prevented imatinib binding [24, 25]. Experience with KIT inhibitors has revealed drug resistant mutations in this kinase [10]. Pharmacokinetic mediated tumor resistance to kinase inhibitors include, polymorphisms in drug metabolizing enzymes and transporters. For example, CML patients with low human organic cation transporter-1 (hOCT1) activity have suboptimal response to imatinib but not to nilotinib [26, 27]. Other mechanisms of tumor resistance to kinase inhibitors may include unique features of the tumors microenvironment and activation of alternative signaling pathways. For a more comprehensive discussion of tumor resistance to kinase inhibitors the reader is directed to the references [28, 29].

Since the year 2001 there have been no less than 24 small molecule, ATP site binding kinase inhibitors approved by the USFDA for oncology indications (Table 1). As the concept of oncogene addiction is explored more and a better understanding of the role of kinases in oncology is realized, the class of kinase drugs may be ideally situated to have a profound effect on personalized cancer medicine.



anticancer drugs may encounter tumor resistance, a kinase inhibitor may also encounter

Tumor resistance to anticancer drugs can be a challenging battle even though a lot has been learned about the mechanisms that allow tumor cells to escape the chemical assault. One major mechanism of tumor resistance to kinase inhibitors is mutant forms of the target kinase, either preexisting or acquired during or after treatment, that reduce the kinase inhibitors binding affinity. Early clinical experience with imatinib identified patients that developed resistance after an initial response to the drug due to mutation of the ACR-Abl kinase that prevented imatinib binding [24, 25]. Experience with KIT inhibitors has revealed drug resistant mutations in this kinase [10]. Pharmacokinetic mediated tumor resistance to kinase inhibitors include, polymorphisms in drug metabolizing enzymes and transporters. For example, CML patients with low human organic cation transporter-1 (hOCT1) activity have suboptimal response to imatinib but not to nilotinib [26, 27]. Other mechanisms of tumor resistance to kinase inhibitors may include unique features of the tumors microenvironment and activation of alternative signaling pathways. For a more comprehensive discussion of tumor resistance to kinase

Since the year 2001 there have been no less than 24 small molecule, ATP site binding kinase inhibitors approved by the USFDA for oncology indications (Table 1). As the concept of oncogene addiction is explored more and a better understanding of the role of kinases in oncology is realized, the class of kinase drugs may be ideally situated to have a profound effect

**Structure Generic Name Target(s) Indication(s)**

afatinib EGFR NSCLC

axitinib VEGFR 1/2/3 RCC

Src

CML

bosutinib BCR-Abl, Hck, Lyn,

resistance either acquired or intrinsic.

20 Updates on Cancer Treatment

on personalized cancer medicine.

inhibitors the reader is directed to the references [28, 29].



**Table 1.** USDA approved kinase inhibitors for oncology indication.

**Structure Generic Name Target(s) Indication(s)**

22 Updates on Cancer Treatment

ibrutinib BTK

imatinib BCR-Abl, PDGFR,

nintedanib VEGFR, FGFR,

pazopanib

ponatinib

regorafenib

KIT

lapatinib EGFR, ERBB2 Breast cancer

nilotinib BCR-Abl, PDGFR CML

PDGFR

VEGFR, PDGFR, FGFR, KIT, Lck, FMS, ITK

BCR-Abl, VEGFR, PDGFR, Src, KIT, Flt3, RET, Tie2

VEGFR, BCR-Abl, B-Raf, PDGFR, FGFR, Tie2

ruxolitinib JAK myelofibrosis

sorafenib Multi-kinase inhibitor RCC, DTC

CLL, mantle cell lymphoma

CML, ALL, GIST, ASM

IPF

RCC, soft tissue sarcomas

CML, Ph+ All

CRC
