BSJ-4-116

Molecular hybrids: A five-year survey on structures of multiple targeted hybrids of protein kinase inhibitors for cancer therapy

Osama M. Soltan a, Mai E. Shoman b, *, Salah A. Abdel-Aziz a, c, Atsushi Narumi d,
Hiroyuki Konno e, Mohamed Abdel-Aziz b, **
a Department of Pharmaceutical Chemistry, Faculty of Pharmacy, Al-Azhar University, Assiut Branch, Assiut, 71524, Egypt
b Department of Medicinal Chemistry, Faculty of Pharmacy, Minia University, 61519, Minia, Egypt
c Department of Pharmaceutical Chemistry, Faculty of Pharmacy, Deraya University, 61111, Minia, Egypt
d Department of Organic Materials Science, Graduate School of Organic Materials Science, Yamagata University, Jonan 4-3-16, Yonezawa, 992-8510, Japan
e Department of Biological Engineering, Graduate School of Science and Engineering, Yamagata University, Jonan 4-3-16, Yonezawa, 992-8510, Japan

A R T I C L E I N F O

Article history:
Received 22 June 2021 Received in revised form 23 July 2021
Accepted 8 August 2021
Available online 14 August 2021

Keywords: Protein kinase inhibitors Molecular hybridization Multiple target therapy Anticancer

Abstract

Protein kinases have grown over the past few years as a crucial target for different cancer types. With the multifactorial nature of cancer, and the fast development of drug resistance for conventional chemo- therapeutics, a strategy for designing multi-target agents was suggested to potentially increase drug efficacy, minimize side effects and retain the proper pharmacokinetic properties. Kinase inhibitors were used extensively in such strategy. Different kinase inhibitor agents which target EGFR, VEGFR, c-Met, CDK, PDK and other targets were merged into hybrids with conventional chemotherapeutics such as tubulin polymerization and topoisomerase inhibitors. Other hybrids were designed gathering kinase inhibitors with targeted cancer therapy such as HDAC, PARP, HSP 90 inhibitors. Nitric oxide donor molecules were also merged with kinase inhibitors for cancer therapy. The current review presents the hybrids designed in the past five years discussing their design principles, results and highlights their future perspectives.

1. Introduction
1.1. Protein kinases

Kinases (PKs) are enzymes responsible for transferring a phos- phate group from one molecule of ATP to the amino acid residue of the target molecule. They play fundamental roles in metabolism regulation, gene expression, cell division, cell growth and cell dif- ferentiation. With regards to the source of the phosphorylated -OH group, these enzymes are classified as serine/threonine kinases and tyrosine kinases [1,2].

The protein tyrosine kinase (PTK) family constitutes the largest subgroup of kinases encoded in the human genome, with 90 members out of 518 total kinases [3]. Most protein kinases belong to a single superfamily containing a eukaryotic protein kinase (478 ePK) catalytic domain. There are also atypical protein kinase (40 aPK) families were identified, these contain proteins reported to have biochemical kinase activity. These kinases are classified into a hierarchy of groups, families, and subfamilies. Human kinase clas- sification of five broad groups 44 families and 51 subfamilies by adding four new groups, 90 families, and 145 subfamilies. Between the four new groups, STE consists of MAPK cascade families (Ste7/ MAP2K, Ste11/MAP3K, and Ste20/MAP4K). The CK1 group contains CK1, TTBK (tau tubulin kinase), and VRK (vaccinia-related kinase) families. TKL (tyrosine kinaseelike) is a diverse group of families that resemble both tyrosine and serine threonine kinases. It con- sists of the MLK (mixed-lineage kinase), LISK (LIMK/TESK), IRAK [interleukin-1 (IL-1) receptoreassociated kinase], Raf, RIPK [receptor-interacting protein kinase (RIP)], and STRK (activin and TGF- receptors) families. Members of the RGC (receptor guanylate cyclase) group are also similar in domain sequence to tyrosine ki- nases. The distribution of these 518 kinases in the human genome can be illustrated in Table 1 [3].

Protein kinases are the most often mutated family of genes that are involved in cancers, with an approximately 4-fold over- representation when compared to a random selection of the same number of other genes. Some overrepresentation of kinases might be attributable to ascertainment bias, given the huge capacity of this protein family. However, mutations in most protein kinases were identified through positional cloning approaches. As a result, ascertainment bias is unlikely to fully explain their overabundance among cancer genes. Moreover, Protein kinases may operate as tumor suppressors or proto-oncogenes in normal, healthy cells. So, mutations in protein kinases may cause cancers via several mech- anisms, such as activation of proliferative pathways, genomic instability, reduced DNA damage response, deactivation of apoptotic pathways and/or the promotion of angiogenesis and cell motility [8].

Furthermore, mutations in PTKs may result in many diseases, especially cancer, and thus PTKs are widely investigated as crucial drug targets, particularly for cancer therapy [9]. In addition, ki- nases’ ability to trigger various signaling pathways involved in cancer cell survival, proliferation, and metastasis prompted their use as a drug target for cancer. Their over-expression in tumor cells usually correlates to poor cancer prognosis [10e12].

The PTK family includes 58 receptor types (RTKs) and 32 non- receptor type kinases (nRTKs) in the human genome [13]. The re- ceptor tyrosine kinases (RTKs) include platelet derived growth factor receptors (PDGFR), epidermal growth factor receptor (EGFR),vascular endothelial growth factor receptor (VEGFR) and fibroblast growth factor receptor (FGFR), while the non-receptor tyrosine kinases (nRTKs) include Src, JAK, Abl, among others [2].

Other kinases are involved in cancer pathophysiology like, cyclin dependent kinases (CDKs). They constitute a crucial serine/threo- nine protein kinase family that is involved in the regulation of fundamental cellular processes such as cell division and gene transcription [14]. The regulation of cell cycle is a well-organized pairing of different family members of CDKs. When activated with their cyclins, these kinases modulate the activity of many substrates such as organizational proteins, transcription factors and proteins involved in the replication and machinery of cells. Among these CDKs, CDK2, CDK4 and CDK6 is especially important, which contributes to controls in the majority of cancer types [15]. Over- activation of CDKs leads to failure of checkpoints and the loss of apoptotic response [16], consequently the inhibition of CDK for treatment of cancers has been widely investigated. Several CDKs inhibitors are subject to clinical trials [17].

The interaction between CDKs and cyclins are tightly linked to the correct timing and order of each cell phase. CDK1, CDK2, CDK3, CDK4 and CDK6 directly engaged in the cell cycle regulation, while CDK7, CDK8, CDK9 and CDK12 are involving in transcription regu- lation [18]. CDK4 formed a complex with cyclinD1, which is crucial component of cell cycle activation that regulate G1 phase, during cells grow and produce proteins in preparation for DNA synthesis. Deregulation of the CDK4-cyclinD-Rb pathway and CDK4 overex- pressed in cancer. Inhibition of CDK4 linked to cell cycle regulation by inhibitors can result in a G1 arrest and cell cycle progression are halted allowing an effective approach to the control of tumor growth [19]. The FDA approved CDK inhibitors such as Palbociclib, Ribociclib (Table 2) and Abemaciclib have been launched on the market for the treatment of breast cancer and HER2-negative advanced breast cancer [18].

Mitogen-activated protein kinase (MAPK) are evolutionarily conserved kinase modules that link extracellular signals to the machinery parts that control fundamental cellular processes such as proliferation, growth, migration, differentiation and apoptosis. To date six distinct groups of MAPKs have been characterized in mammals; extracellular signal-regulated kinase (ERK) 1/2, ERK3/4, ERK5, ERK7/8, Jun N-terminal kinase (JNK)1/2/3 and others [20].

There are many kinase inhibitors approved by the FDA having significant therapeutic effects against different cancers (Fig. 1), in addition to several others under clinical trials [21]. For instance, Imatinib was approved in 2002 as an inhibitor for BCR-ABL kinase and has shown outstanding efficacy in the treatment of Chronic Myelogenous Leukemia [22]. Erlotinib, Gefitinib were approved as EGFR inhibitors. Erlotinib has been the first EGFR inhibitor assessed as first-line treatment for NSCLC patients carrying EGFR mutations [23,24], while Sorafenib has been the first kinase inhibitor approved for renal cell.

carcinoma [25] (RCC) treatment. Sunitinib was also approved for the treatment of metastatic RCC and imatinib-resistant gastroin- testinal stromal tumor (GIST) [26] and Vemurafenib was approved for the treatment of metastatic melanoma associated with BRAF V600E mutation [27].
Moreover, a few of kinase inhibitors were approved for non- malignant indications, such as Fedratinib that is utilized for the treatment of myelofibrosis [28], whereas Ruxolitinib is employed for the treatment of myelofibrosis and polycythemia vera [29] and Tofacitinib is used for the treatment of rheumatoid arthritis [30]. In addition, Sirolimus [31,32] and Ibrutinib [33,34] are prescribed for the treatment of both cancerous and non-cancerous diseases.

A list of some FDA approved kinase inhibitors that will be dis- cussed later in this review is collected in Table 2 along with their therapeutic indication and potential targets. Despite the fact that kinase inhibitors introduced a new concept of targeted chemotherapeutics in cancer therapy, there are cases where kinase inhibitors did not show full efficiency due to the complex and dynamic nature of cancer cells that lead to increase in non-responsiveness rates and the emergence of acquired drug resistance [35]. Resistance to Gefitinib, Erlotinib and Afatinib was detected in NSCLC patients [36,37]. Moreover, Sunitinib efficacy was influenced by KIT gene mutation and its resistance was asso- ciated with up-regulation of FGF1 in RCC cell line [38,39]. A short- term increase in patient survival and resistance was noticed with Vemurafenib though it is originally described for metastatic mel- anoma associated with BRAFV600E mutation [27]. Additionally, the success of single chemotherapeutic agents is usually hindered by the multifactorial nature of most cancer types [40].

To overcome such limitations different strategies have been tested, including drug combination or cocktails. Fortunately, the co- use of kinase inhibitors with other chemotherapeutics resulted in reduced resistance rates. For example, improved response and survival rates were observed using a combination of Sorafenib with Capecitabine for advanced solid tumors. Likewise, a synergistic response with reduced tumor cells proliferation was reported using the combination of Sorafenib with Doxorubicin whereas its com- bination with low-dose of 5-Fluorouracil (5-FU) caused reduced side effects in hepatocellular carcinoma patients [41e44]. However, such combination frequently hampered by possible adverse drug- drug interactions, increased toxicities, unpredictable pharmacoki- netic (PK) properties, and/or poor patient compliance [45,46].

An alternate strategy is the design of smart hybrid molecules carrying multiple pharmacophore entities in a monocular drug thus binds simultaneously to various cancer targets [47,48]. Hybrids between kinases and other important therapeutic targets (such as tubulin, topoisomerases, HDACs and others) that aims to syner- gistic effect, suppress proliferation and induce apoptosis in cancer cell that make the cancer cell more sensitive for treatment with kinase inhibitors and to defeat the drug resistance [49e51].

Dual-target agents have benefits rather the combination ther- apies compared to single-target agent, dual-target drugs may display either their additive or synergistic effects. In addition, drug repurposing saves time and costs, which allows an opportunity for design of dual targeted drugs. Multiple features of computational techniques can be developed to solve the issue of dual-target drug design, such as higher-quality datasets, new hypotheses and ma- chine learning models, and more innovative algorithms and soft- ware packages.
On the other hand, despite their apparent success, dual-targeted inhibitors face a number of obstacles. Controlling the side effects of dual-targeted activities is a major challenge. The dual inhibitors continue to be selective against two targets. Primary targets are regulated at low concentrations, but secondary targets require high dosages, resulting in hazardous side effects. Not just biochemical or cell-based bioactivity, but also quantitative information about the potency of dual targeted inhibitors is required. Another significant drawback is that cancer pathways/mechanism are partially un- derstood at the molecular levels. It limits the development of pol- ypharmacological network without the full data. The toxicities or targets of many reported inhibitors are insufficiently understood. The current dual-target design strategies still have several limita- tions such as inability to handle target flexibility. Addressing these issues in their respective regions will be overcome gradually.

Furthermore, the multi-target molecules are often large in molecular weight, with higher logP and having two mechanisms of action may create different adverse effects and unpredictable pharmacokinetic profile. Acute and delayed toxicities have been found to be more common in multitarget drug, especially when non-selective drug is used. And therefore, Comprehensive target combined with in vivo as well as phenotypic screening combined with target-based techniques.

1.2. Molecular hybridization

Molecular hybridization is one of the most valuable structure modification tools currently used in drug construction. It is based on combining two or more distinct biologically active pharmaco- phoric moieties; usually via a linker or a spacer; to form a new molecular hybrid (Fig. 2) with potentially higher activity and effi- cacy compared to the parent moieties [80e83] and/or alleviated side effect of single drug usage [84].

The typical and successful example for the synthesis of hybrid molecule was the synthesis of the antimalarial drug Trioxaquine (hybridization between trioxane and quinoline, Fig. 1). The devel- opment of Trioxaquine hybrid molecules with a dual mode of action (a “double-edged sword”) capable of killing multi-resistant strains by oral administration is being pursued in order to obtain new antimalarial drugs that are both affordable and capable of pre- venting the emergence of resistant strains. This hybrid molecule has two pharmacophores capable of interacting with the heme target and is made with a trioxane motif covalently linked to an aminoquinoline entity. The antimalarial activity of the two inde- pendent precursors of trioxaquines is limited when compared to the hybrid entity, revealed the synergistic effect of the covalent binding of both pharmacophores. This hybrid was active on the immature erythrocytic stages of P. falciparum as artemisinin de- rivatives, whereas chloroquine is active on the late stages. Like a result, trioxaquine hybrid have all features of trioxane-containing molecules, but this hybrid molecule is also able to prevent the polymerization of b-hematin as chloroquine. These findings sug- gest that both moieties of trioxaquine can interact with heme, as a common target [85,86].

This rational approach was developed to create a single mole- cule hitting multiple targets and has a dual function to facilitate the election of new drugs [87e89]. Potentially, these new hybrids will reduce the risk of drug-drug interaction, improve pharmacokinetic properties [90], minimize toxicity of the parent drug and enhance patient compliance [91,92].

The concept of hybridization is usually achieved via chemical connection of two or more biologically active subunits to be joined directly or indirectly forming a new chemical entity possessing pharmacological profiles of joint subunits (Fig. 2) [93,94]. In gen- eral, a symmetrical hybrid connects two identical fragments and is expected to produce more potent and/or selective pharmacological effects compared to single molecules, whereas the non- symmetrical hybrid drug is perceived to show both pharmacolog- ical activities resulting from the individual pharmacophores (dual action) with additional synergic effect [95].

In the hybrid models shown in Fig. 3, a possible combination between two active moieties can be produced via direct method (non-linker mode called fused, model A or merged mode, model B) or through indirect linked hybridization (linked by cleavable linker, model C or linked by stable linker, model D) [94,96].

1.2.1. Design strategies for dual-target agents in medicinal chemistry
1.2.1.1. Drug repurposing. It is important to mention the drug repurposing as a method of drug discovery, it is different from the traditional drug development method. Classical method is a time- consuming, laborious, expensive and risky process. it usually takes 10e15 years to develop a novel drug [97]. However, the success rate of developing a new drug does not exceed 2.01% [98]. Typically, traditional drug development strategies include five phases: detection and preclinical, safety review, clinical research, FDA review, and post-market safety monitoring by the FDA [99].

Drug repurposing (also known as drug repositioning) aims at identifying new uses for already existing drugs [100] and there are only four steps in drug repositioning: compound identification, compound acquisition, development, and FDA post-market safety monitoring [99]. The approved drug may affect more pathologies and regulate the activity of unpredicted targets. Drug repositioning is applied to investigate a new therapeutic area. The cost and time of research would be significantly reduced, because of the past understanding of the metabolism and possible adverse effects [101].

1.2.1.2. Skeleton modification of dual-target drugs. Scaffold modifi- cation of lead agents is an effective and popular technique to afford potential small-molecule drugs. Modifying a compound’s skeleton is typically done to improve medication absorption/solubility, lessen toxic side effects, and improve drug specificity. Skeleton modification of lead drugs can be a good way to get compounds, decrease toxicity and enhance metabolic stability and water solu- bility, but it only works for certain scaffold types of small-molecule compounds [101].

Fig. 2. Design concept of molecular hybridization.

Fig. 3. Direct and indirect pharmacophoric hybridization approaches with examples.

1.2.1.3. Pharmacophore based approaches as strategy for design of dual target inhibitor. Pharmacophore-based combination of drugs is carried out from the perspective of structural chemistry by considering the common pharmacophores of two inhibitors acting on two different targets. Combining distinct pharmacophores from these inhibitors is lead to the rational design of a dual- or multi- target agent that having appropriate physicochemical characters. The following are the three types of pharmacophore-based design strategies based on the pharmacophoric similarity of the inhibitors on two targets: linked pharmacophore approach, fused pharma- cophore approach, and merged pharmacophore approach.

● Linked pharmacophore method

A long linker is introduced between two separate drugs (Fig. 3C and D) when the linker pharmacophore method is used. A novel drug usually contains the major structural features of each drug and may be effective against two targets. When conjugated linkers are utilized, two different pharmacophores are linked by a linker that does not exist in each pharmacophore. The expanding amount of structural information of each inhibitor is considered in linker-based drug design. This method could aid in the rational development of more powerful dual-target drugs. Although it still has problems with metabolism and in vivo kinetics, this method could be useful in the development of more effective dual-target medicines [93].

● Fused pharmacophore method

The second method for linking frameworks involves combining molecules directly (fusion) without the use of a linker. Both phar- macophore components do not overlap. The size of the linker in directly fused compounds is reduced to the point that the phar- macophore frameworks are essentially touching, as observed in the case of aspirin duplications (Fig. 3, A). This approach has an advantage that the designed molecule having the ability to release two drugs with different activity in vivo [96,102].

● Merged pharmacophore method

The merged pharmacophore approach needs lead compounds that have the same pharmacophore and share mutual fragments. As a result, the two small molecules are joined and synthesized into one molecule, but their pharmacophores stay contact, resulting in simpler and smaller merged small molecule. The merged phar- macophore strategy is similar to the hybridizing method in that it uses smaller molecular weight molecules. The merged and fused pharmacophore techniques were created using previously pub- lished small compounds, limiting the diversity of dual target mol- ecules and necessitating the development of new pharmacophores (Fig. 3, B) [96,101].

1.2.2. Computational approaches for designing dual-target drugs

Computational design is another effective approach to come up with new dual-target agents, particularly for new targets with not reported drugs. Due to the structural diversity of these single target inhibitors and the limitations of their target combinations, several inhibitors have recently gained increased attention. Some in- hibitors are designed based upon target structures, and multi- target compounds are designed by the involvement of pharma- cochemistry, proteomics, and computational chemistry.

The selection of target combinations and virtual screening of multi-target small molecules using techniques such as data mining, structural analysis, and pharmacophore construction are all part of the dual-target drug design process. The small-molecule com- pounds produced from some virtual compound libraries are structurally diverse [103]. As a result, the following are some computational design methodologies for dual target inhibitors: multiple docking, fragment-based drug design, and the usage of common pharmacophores.

● Docking-based drug design

The complementarity of a library of small-molecule compounds with the binding sites in the two targets is assessed using a multiple docking technique, and candidate compounds with the best dock- ing scores are chosen for further structural modification. The mo- lecular docking method cannot provide highly accurate predictions, because identical binding sites exist in targets. At the current stage, inverse docking technique suffer from a lack of scoring algorithm performance and the difficulties of addressing target flexibility. Otherwise, enhancing target selectivity prediction accuracy and false-positive rate remains a difficulty [104].

● Fragment-based drug design (FBDD)

Another successful pathway to construct multi-target inhibitor hybrids is fragment-based lead generation. Fragments are scaffolds that are usually part of drugs responsible for biological activities. These fragments are then joined to produce lead compounds [105]. However, this approach differs from the hybridization strategy. The fragment-based method usually applied to improve the biological activity of a molecule fragment by the addition of chemical func- tions [85] or from fragment libraries MOE, several fragment moi- eties were defined as building blocks and are used to make a drug molecule such as ring, linker, and side-chain [106]. Walters et al. [107] analyzed compounds reported in Journal of Medicinal Chem- istry between 1959 and 2009. A total of 415 284 molecules are included in this assessment, he founded that 70% of the compounds are composed of a relatively small number of building blocks (e.g, only 37 rings, 53 linkers, and 16 side-chains) [107].

● Pharmacophore-based drug design

Development of a pharmacophore model to identify the struc- tural features required to access the binding conformation, after the model establishes the essential component. A structure-based approach allows for the identification of drugs active against un- related targets by taking advantage of shared binding site proper- ties that are not apparent from protein sequence or ligand similarities. Pharmacophore-based drug design is a time- consuming process, and parametrization has had a significant impact on virtual screening [108].The linker or spacer is used to connect the selected components properly designed to form a hybrid system, the choice of the suit- able linker and the proper method of attachment is the crucial part for designing the hybrid [93].

Linkers can be classified to different categories according to the mechanism of drug release and drug stability in blood circulation system, including: cleavable and non-cleavable linkers [109], hy- drophilic or hydrophobic linkers [110], acid labile linker [111], enzyme cleavable [112], photocleavable linker [113], prodrug linker [114]. One potential drawback of using a labile linker is that the drug may be released prematurely before reaching the target cells [115]. Chemically, linkers can be ester, amide, thioester, hydrazone, diazo, triazole, disulfide or Mannich base linkers [116].

In discussing molecular hybrids, few terms should be moni- tored. Hybrids, alternatively named multi-target directed com- pounds, dual-acting compounds, me-better drugs, multiple ligands combi molecules [96], usually involves connecting distinguished chemical entities with two or more structural moieties with varied biological functions [93,94]. Molecular hybrids, Conjugates [117], though this term is mostly used interchangeably with the word hybrid [117,118], some reports distinguish conjugates as hybrids using the whole drug non fragmented or using a linker that is not originally seen in both parent drugs [94]. It usually uses the chemical reactivity of the designated biologically active subunits to be joined directly or indirectly through drug- or non-drug linkers or carriers [93,94].

In the upcoming sections, the design of multi-targeted hybrid molecules employing PK inhibitors will be discussed. We will focus on tools used in the last five years as essential tools for the rational design and discovery of new anti-cancer agents. The hybrids dis- cussed will be merging PK with other chemotherapeutic targets such as tubulin, topoisomerase and others, while multi-kinases are beyond the scope of this review.

During the study of dual protein kinase hybrids, all the three previously mentioned strategies were used to design multi- targeted entities with the use of fragment-based hybrids account- ing for more than 50% of the hybrids designed. Most hybrids designed also used non-cleavable stable linkers, with very few ones acting as prodrugs releasing both parent drugs into the target site (compounds 35 and 36; section 2.6).

Though the hybridization strategy looks very appealing to me- dicinal chemists and creates an enormous pile of new leads for opti- mization, limited success stories reached a development level (compounds 33 and 34; section 2.5 are currently in clinical trials). A few limitations might be causing such a lack of developmental progress. A proper selection of the linker offers a significant challenge to the design of those hybrids. One limitation is that utilizing a non- cleavable linker would create a new molecule that could hold its own pharmacological profile and toxicities. On the other hand, attention should bepaid to thestability of the formed molecules inthe systemic circulation, especially with the design of liable linkers; premature release of the used drugs will be disadvantageous. Flexible linkers might introduce altered affinities and selectivity to the designated targets. An increase in molecular weight prompted by hybridization causes another challenge as it might affect oral bioavailability and the ability of the hybrid to cross cellular barriers. Significant care also should be addressed towards choosing the point of attachment of the two drugs. Incompatible connections might affect or destroy the pharmacophore leading to loss of activity; thus, non-pharmacophore points of association should be selected, Addi- tionally, the lack of selectivity of most developed kinase inhibitors adds to the challenges faced during the design of such hybrids.

2. Dual kinase/tubulin polymerization inhibitor hybrids

Microtubules are formed by noncovalent polymerization of a- and b-tubulin and function as an important portion of the cellular cytoskeleton. They are vital for the development and function of the mitotic spindle during mitosis, and also for establishing cell shape, intracellular transport, secretion and cell motility [119,120]. Their importance for cell function and structure, along with their long history of utility in cancer therapy, have made microtubules crucial for anticancer drug discovery, and agents that target them are among the most reliable chemotherapeutics [121,122].

The naturally occurring tubulin polymerization inhibitor alka- loid; Colchicine 2 used in gout treatment, is currently established as a potent cytotoxic agent. It binds to tubulin and inhibits formation of microtubules resulting in mitotic arrest followed by apoptosis [123,124]. In spite of its anticancer efficacy, this drug has limited therapeutic use because of its toxicity. Therefore, several efforts are being made to enhance its anticancer potency, one of these stra- tegies made via innovation of a hybrid anticancer agent that con- tains the colchicine tubulin inhibitor nucleus [125].Surprisingly, some kinase inhibitors were also able to modulate microtubular function thus it might potentiate the action of tubulin-targeting drugs. Tivantinib, a c-MET inhibitor, binds directly to tubulin via the colchicine binding site. Further kinase inhibitors such as PI3-Akt, FLT3 and JAK-2 inhibitors have affected microtubule polymerization [126]. Activity of such agents is usually accompanied by enhanced apoptosis and reduced resistance compared to single targeted inhibitors. Another encouraging aspect is the ability to employ lower doses, resulting in lower toxicity levels [127]. These findings suggest the approach of designing hy- brids involving both tubulin and kinase modulators is largely beneficial in cancer therapy.

2.1. Dual VEGFR/tubulin polymerization inhibitor hybrids

The main assumption for designing dual VEGFR/tubulin poly- merization inhibitors is the constitutional similarity found between certain inhibitors for both targets. VEGFR as a receptor tyrosine can be blocked either competitively via blocking the ATP binding pocket or non-competitively via extending the structure to fill the allosteric site adjacent to the ATP binding site. The ATP binding pocket is usually filled with an orthogonal biaryl ring system con- nected to a linker usually occupied by hydrogen bond (Hebond) donoreacceptor groups to interact with the kinase “hinge” region, in addition to two hydrophobic groups (moiety1 and 2, Fig. 4).

Fig. 4. Structures of Sorafenib 1, Colchicine 2, Motesanib 3, ABT-751 4, Model of dual VEGFR/tubulin polymerization inhibitor and structures of dual VEGFR/antitubulin hybrids 5 and 6 with their inhibitory activities against VEGFR-2 and b-tubulin with the black color is the main core, the blue color highlighting the hydrophobic groups and the red color is highlighting the linker. While, noncompetitive inhibitors as Sorafenib 1 and Motesanib 3, which demonstrate that the second generation of RTK inhibitors retain a head part responsible for binding to the ATP binding site [128], with similar structural features of ATP binding competitors and adds up a tail that binds to the allosteric site via Hebond forming linkers usually urea (i.e. N,N-diaryl urea as in Sorafenib 1) or amide (i.e. N-arylamino-N,N-diaryl carboxamide as in Mote- sanib 3). These groups can form a Hebond acceptor donor pair and connect the head part to another aryl group forming hydrophobic interactions (moiety 3, Fig. 4).

Similarly, the essential pharmacophoric features optimal for colchicine binding site inhibitor are: a biaryl ring system, methoxyphenyl fragment, three Hebond acceptors, one Hebond donor, two hydrophobic centers (moiety1, 2, Fig. 4) and one planar group as seen in ABT-751 [129] 4 (Fig. 4) (ortho-N-aryla- mino-N´-aryl-benzenesulphonamide). When analyzing the fea- tures of both VEGFR and tubulin inhibitors, a pivotal point was found for designing such formula having dual inhibitor effect, is the linker, to increases the probability to anchor colchicine and kinase binding sites by hydrogen bond formation. Furthermore, similar constitutional characters were found especially when investigating N,N-diaryl carboxamide in VEGFR inhibitors and ortho-amino-N,N`- diaryl sulfonamide in tubulin inhibitors as isosteres to each other. Mohammed et al. [130], used the aforementioned principle in the design of two series of ortho-amino thiophene carboxamide derivatives as dual VEGFR and mitotic inhibitors via connecting the thiophene moiety with different hydrophobic moieties through either aminoacetamide linker or methyledine hydrazide linker as in Soferanib 1 (Fig. 4) or amides in Motesanib 3 and the tubulin polymerization inhibitor ABT-751 4 (Fig. 4) to afford single struc- tures having a dual VEGFR/tubulin polymerization inhibitors 5 and
6. It is noteworthy that the most active compound was 5 (Fig. 4), as
it shares a common pharmacophoric feature (a halogen in the para position) in the tail region with Sorafenib, which explain the high cytotoxicity observed.

2.2. Dual EGFR/tubulin polymerization inhibitor hybrids

The clinical evidence for the success of tyrosine kinase inhibitors in combination with microtubule-targeting agents in cancer ther- apy prompted the development of a single molecule that possess both epidermal growth factor receptor (EGFR) kinase and tubulin polymerization inhibitory properties [131,132]. Evidence also sug- gests a synergistic role for both agents as microtubule inhibition induced inactivation of EGFR receptor in certain cancer types [133] and causes desensitization of resistant lung cancer cell lines to EGFR inhibitors [134].
The hybrids of this series used fragment-based conjugation strategies that involve the use of common pharmacophores found in both biological groups. Famous fragments used were chalcones and quinazoline ring among others. The choice of chalcones were supported by their ability to exhibit anticancer activity by several mechanisms including tubulin polymerization and kinase inhibi- tion such as Phloretin 8, Lichochalcone 9 or benzo[c]furan-chalcone found in 11 [135e139]. Hybrids 10 [51] and 12 [140] (Fig. 5) incorporating chalcones were equipotent to standard EGFR and tubulin polymerization inhibitors used. The success was associated usually with the presence of a trisubstituted phenyl group similar to that present in Colchicine 2, and podophyllotoxin 7.

A second major fragment used here was quinazoline, being a primary classic fragment seen in EGFR inhibitors such as Erlotinib 13, Gefitinib 14, and Lapatinib 15 (Fig. 6). It is also used in the design of tubulin polymerization inhibitors [141]. A promising lead for using quinazoline is compound 16. It displayed higher anti- proliferative activity than the reference drug Erlotinib against breast cancer cell lines MCF-7 and MDA-MBA-231.

Fig. 5. Structures of Cholchicine 2, Podophyllotoxin 7, Phloretin 8, Lichochalcone 9, structures of dual EGFR/tubulin polymerization inhibitor hybrid 10, reported EGFR inhibitor 11 and dual EGFR/tubulin polymerization 12.

Fig. 6. Structure of Erlotinib 13, Gefitinib 14, lapatinib 15, kinase inhibitor CI-1040 20 and dual EGFR/tubulin polymerization inhibitor hybrids 16-19 and 21.

Various heterocyclic systems such as thienopyrimidines, oxa- zoles and triazoles were used as an isosteric replacement of qui- nazoline. Thienopyrimidine was used in the design of both EGFR and the tubulin polymerization inhibitors. The success of the thienopyrimidine-containing hybrids, in compound 17 has been typically associated with incorporating the trimethoxy group phenyl group common in tubulin inhibitors. It showed potent cytotoxicity versus five cancer cell lines (A549, Hela, HT-29, Jurkat and RS4; 11)with IC50 values of 0.019, 0.001, 0.02, 0.001 and 0.002 mM, respectively [142].

In 2020, Mohamed et al. [143], used 1,2,4-triazole to potentially mimic quinazoline moiety in FDA approved EGFR inhibitors Erlo- tinib 13, Gefitinib 14 and Lapatinib 15. They replaced the ether linker with a thioether to yield the designed hybrids 18 (Fig. 6) aiming at introducing a multi-target anticancer agent with improved pharmacokinetic properties compared to the reference quinazoline drugs.
Using the isoxazole scaffold, Warda and co-workers [144] designed hybrid 19 (Fig. 6) targeting EGFR/tubulin polymerization based. Additionally, it showed multiple kinase activity against VEGFR-2 andCK2a, as well as acceptable topoisomerase IIb and tubulin polymerization inhibitory activities.

Similarly, Ihmaid et al. [145], developed anthranilic acid-based diamides hybrids 21 with dual target EGFR/tubulin polymeriza- tion inhibition based on the anthranilic acid pharmacophore from the kinase inhibitor drugs CI-1040 20 (Fig. 6). Conjugation was only used once in developing this type hybrids by Sun et al. [146] They designed a new class of dual EGFR/tubulin polymerization in- hibitors through conjugating a reported anti-tubulin and potent anti-cancer agent shikonin 22 with substituted benzoylacrylic acid pharmacophore which is structurally similar to chalcones pos- sessing EGFR-targeting activity in one compact structure 23 (Fig. 7).

2.3. Dual CDK/tubulin polymerization inhibitor hybrids

The ability of CDK1 to phosphorylate both tubulin and tubulin binding proteins [147] encouraged several trials to design new agents that jointly target both cyclin-dependent kinases (CDKs) and tubulin that are involved in regulating the cell cycle, promoting tumor cell proliferation and apoptosis [148,149]. Aminothiazole was selected, being a common fragment in both the CDK inhibitor 24 and the potent tubulin polymerization inhibitor 25 to furnish hybrid 26 (Fig. 8) as dual CDK/Tubulin inhibitors [150]. Unfortu- nately, cytotoxicity screening revealed weak to moderate activities, and negligible activity on both CDK and tubulin polymerization.

A more potent hybrid was reached in 2015 by Mahale et al. [151] They used hybridization in joining tryptoline fraction of the natural CDK inhibitor Fascaplysin 27 (Fig. 8) with different aromatic car- boxylic acid based on the structure of the compound CA224 28 a non-planar analogue of Fascaplysin showing selective Cdk4 inhibition.Though there were no fragments used from a tubulin inhibitor, the resulting structure 29 was a multi targeted compound with better therapeutic window than Fascaplysin 27. It inhibited the growth of several human cancer cells with an IC50 < 1 mM. 2.4. Dual Bcr-abl/tubulin polymerization inhibitor hybrids Bcr-Abl tyrosine kinase is a chimeric oncogene that results from abnormality in chromosomes upon fusion between (chromosome 9) Abelson tyrosine kinase (Abl) gene and (chromosome 22) break point cluster (Bcr) gene, and its overexpression is closely associated with chronic myelogenous leukemia (CML) [152e154]. Imatinib 30 is the first line in treatment of CML and is a Bcr-Abl kinase inhibitor. Most newly diagnosed CML patients treated with Imatinib achieve durable responses; however, a small percentage of them and most advanced- phase patients relapse on Imatinib therapy [155e158]. Nevertheless, Imatinib resistance emerged affecting its binding to Bcr-Abl kinase. Additionally, Imatinib and other kinase inhibitors used failed to induce high rates of apoptosis in CML cells [159]. Use of drug cocktails or combination therapy with additional anticancer agents could lead to gradual reduction in patients resistant to Imatinib. Fig. 7. Structure of Shikonin 22, and dual EGFR/tubulin polymerization inhibitor hybrid 23. Fig. 8. Structures of reported CDK inhibitor 24, reported tubulin polymerization in- hibitor 25, Fascaplysin 27, CA224 28 and dual CDK/tubulin polymerization inhibitor hybrids 26 and 29. A close connection was established between tubulin polymeri- zation and Abl-Bcr tyrosine kinase since the discovery of the mul- titargeted drug MPT0206 31 (Fig. 9). MPT0B206 31 was developed as a highly potent tubulin polymerization inhibitor. It came up as a down-regulator for Bcr-Abl expression in both Imatinib-sensitive and Imatinib-resistant CML cells [160]. MPT0B206 could be used as a potential lead for the development of a dual kinase/tubulin inhibitor, but unfortunately this area is not fully explored. 2.5. Dual Src/tubulin polymerization inhibitor hybrids The Src family comprises a group of non-receptor tyrosine ki- nases (SFKs), a proto-oncogene that regulates various aspects of tumor progression via multiple signaling pathways, including growth (Ras/MEK/ERK), cell survival (PI3K/Akt), angiogenesis (STAT3/VEGF) and metastasis (FAK/paxillin) [161,162]. Over- expression of Src has been demonstrated in colorectal, lung, pancreatic, breast, ovarian, and prostate carcinomas, highlighting Src as a potential therapeutic target [163e165]. The FDA approved several agents targeting Scr family such as Dasatinib 32 (Fig. 10). Structure-based drug design introduced a true success story to the field of designing dual kinase/tubulin hybrids providing 2 hy- brids into the drug development stage. Based on the ability of tubulin inhibitors to downregulate Src, molecular modeling was utilized in evaluating a small library of targeted inhibitors to identify their tubulin polymerization inhibition. Two compounds KX2-391 33 and KX2-361 34 came up as potent Scr/tubulin in- hibitors. Compound KX2-391 33 is currently in phase 3 clinical trial that is likely to provide a first in class topical treatment for actinic keratosis with excellent efficacy and less toxicity compared to existing standard therapy. Compound KX2-361 34 is a phase 1 clinical trial candidate that is likely to provide a first in class treatment for malignant glioblastoma [131]. Fig. 9. Structure of imatinib 30 and MPTO206 31. Fig. 11. Structures and design of dual PDHK1/tubulin polymerization inhibitor hybrids 35 and 36. 3. Dual kinase/Topoisomerase inhibitors Fig. 10. Structures of Dasatinib 32, and dual Src/tubulin polymerization inhibitors KX2- 391 33 and KX2-361 34. 2.6. Dual PDK1/tubulin polymerization inhibitor hybrids PDK; pyruvate dehydrogenase kinase-1, a member of the Gyr- ase, Hsp90 and Histidine kinase ATPase/kinase superfamily, con- tains four mammalian isoforms (PDK1-4) in mitochondria characterized in terms of their differences in activity, regulations and tissue distribution. Among these isoforms, PDK1 is predominantly associated with malignancy by reducing pyruvate dehydrogenase complex activity through phosphorylation of specific serine residues in E1a subunit of pyruvate dehydrogenase (PDH) [166e168]. Remarkably over- expression of PDK1 has been reported in multiple clinical cancer specimens and has become popular pharmacological target for cancer chemotherapy [169e171]. To date, the use of PDK1 in- hibitors has been disappointing in clinical trials, raising concerns about the rationale of designing dual tubulin/PDK1 inhibitors. Nevertheless, the results showed that it can be critical, leading to an innovative strategy for the design of new anticancer compounds. Conjugation was the main tool used in designing dual PDK1/ tubulin inhibitors. These conjugates were able to modulate the poor pharmacokinetic parameters of PDK inhibitors used. More- over, they serve as mutual prodrugs, as they hydrolyze in the tumor cell promoted by the action of hepatic microsomal esterases or amidases, yielding the two potential codrugs. Lin et al. [172], designed a conjugate between a-lipoic acid the naturally occurring co-factor of pyruvate dehydrogenase and shikonin 22 a reported anti-tubulin agent, with substitutions optimization to design the conjugated drug 35 via an ester connection (Fig. 11). Another con- jugate 36 (Fig. 11) was designed joining benzophenone derived drug Phenstatin, a known potent tubulin polymerization inhibitor with dichloroacetic acid (DCA), a potent non-specific inhibitor of the mitochondrial pyruvate dehydrogenase kinases (PDK1-4 or PDHK1-4) [173]. Topoisomerases (topo); I and II, they are enzymes responsible for resolving sophisticated DNA topological intermediates gener- ated during DNA replication, transcription, recombination and repair processes, such as supercoiled, relaxed, knotted, and cate- nated DNA mainly via catalyzing DNA cleavage and relegation [174,175]. While topo I cleaves one single-stranded DNA during each catalytic cycle, topo II breaks one double-stranded DNA strand, allowing another segment of duplex DNA to pass through the transient breakage before resealing the broken strand to resolve DNA knots and tangles [176]. The FDA has approved drugs targeting topoisomerase enzymes such as DACA 37 and Daunorubicin 38 (Fig. 12). These drugs disturb the catalytic function of mainly topo II either by trapping the enzyme in its complex with cleaved DNA [177], by preventing the process of relegation, by enhancing the formation of cleavable complexes, or by intercalating DNA, preventing the enzyme from performing its catalytic function [178]. Some agents act through more than one of these mechanisms. The expression of topo II is tightly associated to cell division especially in the case of abnormal rapidly dividing cells. Consequently, compounds targeting this enzyme can act as effective anticancer agents [179]. Despite the clinical success of drugs that target topo II, the evolution of resistant cancer cells can restrict their clinical effec- tiveness [180]. Moreover, similarly like other anticancer drugs, most topo II inhibitors induce severe side effects [181]. Combina- tions of different therapeutics such as platinum compounds, taxols, and even with targeted therapies as EGFR and Her2/neu inhibitors were used with topo inhibitors to overcome these limitations. Purposes listed for such combinations could be merely clinical consideration or to reduce the toxicity of active agents used [182,183]. A mutual connection between kinases as EGFR and topo II was reported; they modify each other's expression in vivo in many cancer types; high levels of EGFR have been shown to correlate with the resistance of cancer cells to topo II inhibitors. EGFR is overexpressed in a variety of cancers and has a regulatory role in their growth, thus, synchronized inhibition of topo II and EGFR could be an effective strategy for anticancer treatments [180,184]. Dovitinib 39 (Fig. 12) is a multi-kinase inhibitor and a dual topo I and II inhibitor, nonetheless, kinase inhibition is still considered to be the main mechanism of action [185]. Fig. 12. Structures of topoisomerase inhibitors DACA 37, Daunorubicin 38 and Kinase/ topoisomerase inhibitor Dovitinib 39. 3.1. Dual EGFR -VEGFR/topoisomerase inhibitors The design of dual EGFR/topo inhibitors used the same princi- ples implied earlier for tubulin/EGFR hybrids. The most used design strategy was incorporating a heterocyclic fragment as the primary core for inducing the kinase inhibition activity. Similar to tubulin/ kinase hybrids, quinazoline is the main fragment used to design topo/kinase hybrids. Different thioquinazolinone based hybrids were designed such as 40; Fig. 13) with dual kinase/topoisomerase inhibitory activity using biologically active linkers such as 1,2,3- triazole, thiourea or acylhydrazone moiety [186]. These series were not very successful in supporting hybrid rational, whereas most of the derivatives showed moderate anti-proliferative activity against Hela, HCT-116 and MCF-7 cancer cell lines, only one hybrid 40 carrying a triazole linker, possessed potent multi-target inhibi- tory activities against EGFR, VEGFR-2 and topo II. A similar conclusion was reached using a different heterocyclic core; quinoline. Quinoline is an isostere to quinazoline and is also detected in topo inhibitors such as Camptothecin 41. Using a fused quinoline core in a non-planner compound (42, Fig. 13) retained the Topo regulatory activity of Camptothecin with the addition of an EGFR inhibitory action [187]. Fig. 13. Structure and principles of design of dual EGFR-TK/topoisomerase inhibitors 40, 42-45 and 48. The use of 4-amino-3-nitroquinoline scaffold in 43 (Fig. 13) substantially forms C2-N bond that might occupy extra space of ATP-binding pocket of EGFR and mutant EGFR decreasing incidence of acquired resistance [188].Quinoline were also found in antibacterial topo II inhibitor Ciprofloxacin. A symmetrical hybrid from ciprofloxacin. 44 con- necting 2 Ciprofloxacin molecules in a chalcone like spacer (Fig. 13) successfully inhibited tumor cell proliferation of six different cancer cell lines (T24, Panc-1, BGC-823, HCG-27, PU145 and Capan-1 cells) with IC50s ranged from 0.312 to 10 mM. Compound 44 significantly inhibited tyrosine kinase activity in vitro with an IC50 of 0.64 mM in Capan-1 cells and 3.1 mM in Panc-1 cells and inhibited the ability of topo II to relax supercoiled pBR322 DNA [181]. Tricyclic tetrahydrobenzothieno[2,3-d]pyrimidine could be used as a planner fragment required for topo inhibition. Multi-sited enzyme inhibitors such as 45 (Fig. 13) were obtained joining thei- nopyrimidines with diaryl urea kinase fragment found in Sorafenib 1 and Doxorubicin [189]. The hybrids are promising multi-targeted lead specially against MCF-7 cell line. Hybrids 45 showed 1.5 to 1.03 folds more potent activity than Doxorubicin. Keeping the CF3, Cl substitution of Sor- afenib is essential for retaining the kinase activity of these hybrids. Another effective strategy was to incorporate chalcones into the designed hybrids as mentioned earlier with tubulin/kinase hybrids. Chalcones was either joined quinoline as in 44 or merged with a small triazole group as in 48 (Fig. 13). The design used the naturally occurring, Flavokavain B 46 with the bioactive 1,2,3-triazole core in 47 (Fig. 13) to yield hybrids 48 with dual inhibition effect on both VEGFR and topoisomerase enzymes [190]. The hybrid was acting against several molecular targets including EGFR, VEGFR2 and topoisomerase with lower inhibitory activities. 4. Dual kinase inhibitors/estrogen modulators Estrogen receptors (ERs) are ligand-modulated transcriptional factors which play a key role in induction, proliferation, and development of certain types of malignancies such as breast cancer. ERs overexpression indicates a negative breast cancer prognosis with more than 70% breast tumors being Era (a member of ER family) positive and estrogen dependent [191]. Selective estrogen receptor modulators (SERMs) are a special group of non-steroidal ER ligands, which act to modulate estrogen receptors. They act as antagonists in breast tissue but agonists in other tissues such as CVS and bones. The special action mode of SERMs makes them important anti-breast cancer agents with benefits in the cardiovascular system while maintaining bone density. Tamoxifen 49 (Fig. 14) is the first SERM developed and is currently widely used in the prevention and treatment of breast cancer [192e194]. In fact, many women with breast cancer initially respond well to Tamoxifen; however, the side effects often emerge with long-term usage such as increasing the occurrence of endo- metrial cancer, and acquired drug resistance [195,196]. A mutual relationship between ER modulators with kinases is proven; for example Ras/Mitogen-activated protein kinase pathway that is activated in VEGF/VEGFR signal transduction is closely associated with Tamoxifen resistance [197]. Further investigations revealed that the MAPK pathway can phosphorylate and activate ERa in a ligand-independent manner, resulting in transcription of estrogen-regulated genes and cell proliferation [198,199]. More- over, the fact that VEGFR-2 inhibitors are not sufficient as mono- therapy in treatment of breast cancer [200], encouraged the use of a combination of a VEGFR inhibitor (Brivanib alaninate) with Tamoxifen. This combination was reported to maximize thera- peutic efficacy as well as to retard SERM resistant tumor growth [201]. Fig. 14. Structures of Tamoxifin 49, Cabozatinib 50, reported dual VEGFR/ER ligand compound 51 and structures of dual VEGFR/ER targeting hybrids 52-54. The use of heterocyclic fragment was again the fundamental aspect in designing kinase/ER inhibitors. Quinoline found in Cabozantinib 50 was used to carry a vicinal diaryl system of essential for Tamoxifen 49 or the reported dual ligand of ERa and VEGFR compound 51, to form a series of 3-aryl-quinoline de- rivatives 52 (Fig. 14) with various basic side chain at the end to mimic the structural of the SERMs [202]. Hybrid 52 was an effective inhibitor of ERa (binding affinity similar to Tamoxifen) and VEGFR-2, simultaneously. SAR showed that methyl piperazine, a fragment that is common in kinase in- hibitors, at the end of side chains contribute most to the anti- proliferative activity against MCF-7 breast cancer cells and poten- tial in vitro anti-angiogenesis effects. Replacement of quinoline with coumarin in 53 [203] or fused coumarin in 54 [204] (Fig. 14) also was a key element in designing hybrids of this type. The presence of oxygen atom at the C-4 po- sition of coumarin scaffold proved to be beneficial in enhancing the Era binding affinity while retaining an excellent inhibition on VGFR-2 and the subsequent signaling transduction of Raf-1/MAPK/ ERK pathway in MCF-7 cells. 5. Dual kinases/HDAC inhibitors Histone deacetylase enzymes (HDACs) are a family of enzymes that catalyze the removal of an acetyl group from the lysine resi- dues located in the N terminal tails of core histones and promote a more closed chromatin structure where transcription is repressed [205,206]. There are eighteen HDACs that have been identified in humans and have been classified into four classes depending on their sequence homology to yeast HDACs, their subcellular locali- zation and their enzymatic activities. Amongst them, Classes I (HDAC1-3, and HDAC8), II (HDAC4-7, and HDAC9-10) and IV (HDAC11) are zinc dependent metalloproteins [207e209]. It has been widely recognized that HDACs play a vital role in epigenetic regulation of gene transcription and regulate other cellular functions such as proliferation, cell death, and motility. In addition, HDAC inhibitors (especially HDACI) can affect cancer cell growth and survival by blocking the deacetylation of histone or non-histone proteins (such as Hsp90, a-tubulin, and transcription factors p53 and NF-jB), inducing angiogenesis suppression, cell cycle arrest and apoptosis [207e209]. The dysregulation of histone deacetylases has been frequently recognized as a crucial factor in cancer initiation [210]. Typical pharmacophoric features of most HDAC inhibitors include a cap group, a linker and a zinc binding group (ZBG). A cap structure is usually composed of a hydrophobic ring that interacts with amino acids residues around the entrance of the active site; the linker group occupies the tubular channel to connect the cap group and the zinc site; the zinc-binding group, such as hydroxamic acid and 2-aminobenzamide, chelates with zinc ion by forming coordinate bond within the active site (Fig. 15) [211,212].To date, the FDA approved many HDACs inhibitors (Fig. 15) for treatment of various cancer types including Vorinostate (SAHA) 55 and Belinostate 56. In addition, Pracinostate 57 is in phase 3 clinical trials. Chidamide 58, is a histone acetylase (HDAC) inhibitor approved in China for the treatment of advanced hematological malignancies. Nonetheless, many of HDACs, when used as single agent, showed a limited therapeutic application and resistance was often observed. The combination therapy was used to maximize their efficacy, whilst reducing toxicity and resistance [213,214].Hybridization with other drugs or combination therapy, have been tested in order to overcome these limitations especially with Kinase inhibitors due to their flexible SAR and the potent syner- gistic effects of HDAC with kinase inhibitors [215]. concurrent in- hibition of both HDAC and certain kinases would enhance HDAC efficacy and minimize the incidence of resistance [216]. Hybrids of dual kinase/HDAC targeting ability is the major class of hybrids designed. The clarity of the pharmacophores required for this groups and the minimal structural modifications done to incorporate the HDAC targeting into different classes of kinase inhibitors probably contributed to such versatility. The conjuga- tion concept prevails in the design of dual kinase/HDAC inhibition. It usually involves retaining a cap group carrying the kinase in- hibitor attached to a zinc-binding group from HDAC inhibitors via a suitable linker. The ZBGs mostly used were either hydroxamic acid or aminoanilides with promising results with both groups. Linkers used are usually a 4e6 carbon chain, a simple benzyl group or in fewer cases a small heterocyclic group as triazole were included as a part of the linker. The use of triazole offered an additional site for H-bond formation and is usually associated with better fit into the HDAC active site as seen in conjugates 64 and 67 discussed later. Fig. 15. Structural features and structures of FDA approved HDAC inhibitors SAHA 55, Belinostat 56, Praconstat 57, and Chidamide 58. 5.1. Dual VEGFR/HDAC inhibitors Using VEGFR inhibitor Pazopanib 59 as a Cap group with or without a linker to bind to ZBG (aminoanilide or hydroxamic acid) formed a dual VEGFR/HDAC targeting conjugates. Conjugate 60 with aminoanilide ZBG (Fig. 16) was more potent than its parent compounds in inhibiting both class I HDAC and VEGFR-2 [217]. Similarly, Vandetanib 61 was employed as a cap group with a six carbon linker to a hydroxamic acid ZBG in conjugates 62 and 63 (Fig. 16) [218,219]. Further investigations of these conjugates demonstrated the crucial role for phenyl ring substitutions on the VEGFR inhibition while HDAC activity is optimum with 5e6 carbon length linker. Incorporation of a 1,2,3-triazole linker to these conjugates (com- pound 64, Fig. 16); induced a better fit for the hydroxamic acid group the hydrophilic zone of kinase as it acts as a bioisostere of amide group in SAHA 55 and aids the cap group to recognize the surface outside the HDAC active pocket [220]. 5.2. Dual EGFR/HDAC inhibitors An additional successful conjugate was designed using the EGFR inhibitor Osimertinib 65 as a cap group connected to ZBC (hydroxamic acid or N-(2-aminophenyl)acetamide moiety) through different alkyl linkers to construct the new conjugate 66 (Fig. 17) [221]. These conjugates were more potent anti-cancer agents against MDA-MB-231, MDA-MB-468, HeLa, KG-1and HT-29 cancer cell lines than Osimertinib and SAHA. Fragment based design was also used in this context using only 4-aminoquinazoline fragment of Lapatinib 15, as a cap moiety, 1,2,3-triazole moiety as a linker and the hydroxamic acid moiety as ZBG forming a series of EGFR/HDAC inhibitors [222] 67 (Fig. 17). 5.3. Dual FGFR/HDAC inhibitors The fibroblast growth factor (FGF) family and their four receptor tyrosine kinases (FGFR1, 2, 3, and 4) play a basilar function in many physiologic processes, including embryogenesis, tissue repair, tis- sue homeostasis, wound healing, and inflammation. FGFR1 over- expression is also common in breast cancer and has been reported in nearly 15% of hormone receptor-positive breast cancers as well as in approximately 5% of the more aggressive triple-negative breast cancers. Also, FGFR2 is diffuse in about 10% of gastric tumor [223] and 4% of triple-negative breast cancers. FGFR3 is commonly amplified in bladder cancer, oral, cervical [224] and hepatological cancers and FGFR4 is magnified in hepatocellular carcinoma [225] and ovarian cancer [226]. Thus, FGFR has emerged as an attractive target in cancer therapy [227]. Due to the lack selectivity of kinases, the first generation small-molecule FGFR inhibitors were multi- targeted resulting in a variety of side and toxic effects in clinical and preclinical studies. Moreover, the second-generation FGFR- selective inhibitors, including AZD4547 68 (Fig. 18) have obtained great success, owing to excellent in vivo efficacy and favorable pharmacokinetic properties [228e230]. Fig. 16. Structures of Pazopanib 59, dual VEGFR/HDAC inhibitors 60, Vandetatinib 61, and dual VEGFR/HDAC inhibitors hybrids 62-64. Fig. 17. Structure of Osimertinib 65, Lapatinib 15, dual EGFR/HDAC inhibitor hybrids 66 and 67. Combining FGFR with HDAC has also proven to be a promising strategy for dual acting anti-cancer agents. A conjugate 70 (Fig. 18) was designed [231] via merging compound 69, a reported potent FGFR inhibitor as a cap moiety with the ZBG (hydroxamic acid). The absence of a small piperazine group from the parent FGFR inhibitor in the designed hybrid affected the FGFR inhibition pointing out at one of the challenges facing the design of such hybrids being the loss of activity when small fragments of the used pharmacophores are omitted. 5.4. Dual CDK/HDAC inhibitors Reports acknowledged that SAHA combination with Flavopir- idol, which is a pan-CDK inhibitor, was utilized for the treatment of breast, lung cancer and melanoma [232,233]. A fragment based design integrated purine-based pharmacophore from the CDK in- hibitor Roscovitine 71 as a cap group with the ZBG as 2- aminobenzamide group using different aromatic spacers (example: hybrid 72, Fig. 19) [234] or hydroxamic acid group and the spacer found in the HDAC inhibitor drug Panobinostat 73 forming hybrid 74 (Fig. 19) [235]. Incorporation of the NHOH group was more promising as hybrid 76 (Fig. 19) that demonstrated inhibitory activities against HDAC1 and CDK2 at nanomolar doses. Concept of conjugation was also applied using two ZBGs with substituted-pyrazole-carboxamide scaffold from a reported CDK inhibitor 75 (Fig. 19) as a cap group to obtain a single molecule that could inhibit both CDK as well as HDAC. A more prominent activity was observed with the diaminobenzene as a ZBG as HDAC2 activity was highly potent with subnanomolar concentrations (compound 76, Fig. 19) [236]. Fig. 18. Structures of AZD4547 68, reported FGFR inhibitor 69 and structure of dual FGFR/HDAC inhibitor hybrids 70. Conjugation were also used combing hydroxamic acid from SAHA with the selective CDK4/9, Ribociclib 77 using different alkyl linkers (78, Fig. 19) [19]. Compound 78 was a multi-kinase inhibitor, which potently inhibited CDK4, CDK9, and HDAC1. It also inhibited several kinases, such as the Aurora-A/B/C, Flt4, LIMK1, and TrkA, with IC50 < 0.05 mM. The in vitro cellular assays showed potent activities against T47D and MDA-MB-231, A549, H460, HepG2 and Hep3B cancer cell lines. 5.5. Dual c-Met/HDAC inhibitors Mesenchymal-epithelial transition factor (c-Met) is a prototype member of a subfamily receptor tyrosine kinases, also known as hepatocyte growth factor receptor (HGFR). Binding of HGF to its receptor initiate cellular signaling associated with cell proliferation, migration, motility and invasion. Therefore, c-Met considers as a promising target for cancer drug development and FDA approved Cabozantinib 50 and BMS-777607 79 (Fig. 20) as selective and potent c-Met Kinase inhibitor suppresses (HGF)eStimulated Pros- tate Cancer Metastatic Phenotype In vitro [237]. Cabozantinib acts as a dual inhibitor of both c-MET and VEGFR-1 inhibitor in treat- ment of patient with medullary thyroid cancer [238]. However, it is not enough to block tumor progression by exclusive use of c-Met inhibitors due to low efficacy or acquired resistance [239,240]. Actually, inhibition of HDAC affects c-Met and its downstream signaling pathways directly and indirectly. In addition, HDAC in- hibitors were found to suppress the expression of c-Met itself [241e243]. Only fragment-based hybridization was used in developing HDAC/c-Met dual inhibitors. A fragment 4-phenoxyquinoline moiety from Cabozantinib 50 as a cap group connected to hydroxamic acid group as ZBG using alkyl linkers (Fig. 20). Urea, semicarbazone or amido urea was introduced instead of cyclopropane-dicarboxamide moiety of Cabozantinib 50 for providing structurally diverse dual c-Met/HDAC inhibitors (80, Fig. 20) [243].Similarly, a fragment of the potent selective c-Met inhibitor 81 was joined with benzendiamine as ZBG (82, Fig. 20) using different linkers [244]. The potent hybrid 82 resulted in inducing c-Met ki- nase and HDAC1 inhibition with considerable antiproliferative activities. 5.6. Dual Mnk/HDAC inhibitors Mitogen-activated protein kinases belong to the serine/threo- nine kinase family, that is presented in two isoforms: Mnk1 and Mnk2 that share 80% sequence identity. Mnks can mediate the resistance of many clinical drugs, such as Imatinib and Cytarabine. Fig. 19. Structures and design of dual CDK/HDAC inhibitors 72, 74, 76 and 78. Fig. 20. Structures of c-Met inhibitors, Cabozantinib 50, BMS-777607 79, and design of dual c-Met/HDAC inhibitor hybrids 80 and 82. Therefore, the combination of Mnk inhibitors or the design of dual- target inhibitors that simultaneously targeting Mnk and other tar- gets may have better anti-tumor effects [245e247].Since both Mnks and HDAC play axial role in translating mul- tiple oncogenic signaling pathways during oncogenesis and both are highly expressed in a variety of cancers; concurrent inhibition of both HDAC and Mnk could lead to inhibition of tumor cell proliferation.To date, no Mnk inhibitors have been approved by FDA, thus the hybrids designed depended mainly on reported potent Mnk in- hibitor, D06 or 83 (Fig. 21) [248e250]. The usual design concept was used again connecting the pharmacophore from 83 with ZBG (hydroxamic acid or benzendiamine) using different alkyl linkers with potential lead carrying the hydroxamic acid group (hybrid 84, Fig. 21) [250]. 5.7. Dual janus kinase/HDAC inhibitors The Janus kinases (JAK) are a class intracellular, non-receptor tyrosine kinase that includes JAK1, JAK2, JAK3 and TYK2, which play important roles in the signaling of a variety of cytokines. Activation of JAKs by different cytokines results in phosphorylation and dimerization of the STAT (signal transducers and activators of transcription) proteins, which further translocate to the nucleus and activate gene transcription. The JAK/STAT signaling system is associated with various biological functions including cancer [251e254]. JAK2, among the other JAKs subtypes, was proven to be crucial for tumor growth and progression. JAK2 inhibitors have been evaluated in clinical trials for the treatment of a wide spectrum of hematological malignancies as well as solid tumors. However, resistance of JAK2 inhibitors has been observed [255,256]. There- fore, to enhance the therapeutic effects and minimize the risk of resistance, bifunctional drug therapy or the development of JAK2- based multi-targeting anticancer drugs, as they provide new op- portunities [257,258]. Conjugation of hydroxamic acid with FDA approved JAK in- hibitors was one strategy used to generate hybrids. FDA approved agents which target JAKs such as the JAK1/2/3 and TYK2 inhibitor Tofacitinib 85 and Ruxolitinib 86, (selective JAK1/2 inhibitor) was used in such context (Fig. 22). Ruxolitinib-Vorinostate (SAHA) hybrid 87 was a potent dual JAK/HDAC inhibitor (Fig. 22) [259]. It inhibits JAK1 and HDACs 1, 2, 3, 6, and 10 with IC50 values of less than 0.02 mM, and is selective for the JAK family against a panel of 97 kinases. Broad cellular anti-proliferative potency of 87 is demonstrated by JAK-STAT and HDAC pathway blockade in hema- tological cell lines. The presence of methyl group at C1 of the used linker augmented selectivity profile towards JAK1, HDAC1 and HDAC6. Similarly, the macroloid structure of Pacritinib 88 (a phase III clinical trials JAK/FLT3 inhibitor for myelofibrosis) was used as cap group for a hydroxamic acid ZBG to yield 89 instead of pyrro- lidine ring of its structure (Fig. 22) [260]. Substituted pyrimidine fragment common in reported JAK in- hibitors (such as 90 and 91, Fig. 23) were joined with hydroxamic acid to produce dual JAK/HDAC inhibitor hybrids 92-95 (Fig. 23) [216,261,262]. Most of the hybrid series simultaneously inhibit JAK2 and total HDACs at nanomolar levels, also showed potent anti- proliferative activities against tumor cell lines (Hela, K562, MOLT-4 and Jurkat) than SAHA and Ruxolitinib. The SAR study revealed that; the hydroxamate group of the hybrids achieves a potent HDAC inhibitory activity, the amino- pyrimidine moiety achieves a potent JAK inhibitory activity, proper linker length is very important for HDAC inhibitory activity (the linker length five or six methylene groups for highest HDACs in- hibition; however, no obvious effect on JAK2 inhibition) and for hybrids with aromatic linkers the N-hydroxybenzamide exhibited more potent HDAC and JAK2 inhibitory activities than the N- hydroxycinnamide.The use of benzenediamine as a ZBG in this context did not produce as potent hybrids as those obtained with hydroxamic acid. 5.8. Dual Bcr/Abl/HDAC inhibitors Synergistic and additive effects were observed with the concomitant use of HDAC inhibitors with Bcr-Abl inhibitors [35,263]. The potential outcome of this combination was to over- come the emerged resistance of Bcr/Abl inhibitor drugs and severe side effects such as blood clotting as in case of Imatinib 30 and Dasatinib 32.That synergism was reached using a fragment of Dasatinib 32 as a cap group with 2-aminobenzendiamine moiety from Chidamide 58 as ZBG, forming a series inhibiting both Bcr/Abl and HDACS (96, Fig. 24) that exhibited higher potency than Dasatinib against K562 cell line [264]. 5.9. Dual PI3K/HDAC inhibitors The use of both HDAC and PI3K inhibitor drugs are limited by insufficient effectiveness and drug resistance, there is strong evi- dence that co-inhibition of both PI3K and HDAC can synergistically inhibit tumor growth and deal with these limitations by improving efficacy, decreasing resistance, and providing a better therapeutic profile than single agent [265e267]. Fig. 21. Structures of DO6 83, structure of dual Mnk/HDAC inhibitor hybrids 84. Fig. 22. Structures of Tofacitinib 85, Ruxolitinib 86, Pacritinib 88 and dual JAK/HDAC inhibitors 87 and 89. A dual activity with HDAC was introduced via conjugating the benzotriazole inhibitors with hydroxamic acid through a long chain alkyl chain (103 and 104, Fig. 26) [276,277]. The benzotriazole group offers different positions to incorporate the alkyl linker simply via alkylating any of the three nitrogen atoms present. Change the position of the linker on either nitrogen did not much affect the hybrid activity. Incorporation of a triazole group (as in 104, Fig. 26) into the alkyl linker is repetitively used to induce a hydrogen bond forming fragment into the linker and it usually induces enhancement in HDAC activity compared to non-triazole containing linkers. Triazole effect on physicochemical parameters also must be accounted for this activity, but this still requires further confirmation. 5.11. Dual Raf/HDAC inhibitors Fig. 23. Structures of reported JAK inhibiter 90 and 91 and the designed structure of dual JAK/HDAC inhibitor hybrids 92-95. In this context, quinazolin-4-one based hydroxamic acids was rationally designed as dual PI3K/HDAC inhibitors. One series (98 [268], Fig. 25) was carrying structural similarity to the pharmaco- phore of the PI3K inhibitor drug Idelalisib 97 while the other (100 [269], Fig. 25) hold fused quinazoline to mimic Pictilisib 99. Hybrids were potent and selective PI3Kg/d/HDAC6 inhibitors with considerable cytotoxicity against a panel of 60 different cancer cell lines (NCI-60). Particularly, compound 98 was identified as a potent dual inhibitor with high selectivity and potent anti- proliferative activity against various cancer cell lines. 98 also exhibited good potency in inducing cell death via necrosis in multiple AML cell lines. 5.10. Dual CK/HDAC inhibitors Casein Protein kinase CK2 is a serine/threonine protein kinase, which is composed of tetrameric complexes consisting of two subunits (CK2a, CK2a’). CK is important for cell viability. It is localized in different cell compartments and is involved in several basic functions such as signal transduction, cell growth and cell differentiation, gene expression and apoptosis [270,271]. A wide Chen et. al. [278] used an imidazole group as ZBG, it was not included in multitargeted compounds involving HDAC before that. Imidazole with joined with the pharmacophore from Sorafenib 1 and retained the key urea group in Sorafenib (Fig. 27). The resultant hybrids 105 (Fig. 27) were able to concurrently inhibit both HDACs and Raf kinases. Keeping the same substitutions as in Sorafenib gave 105a the most potent hybrids against BRafV600E and HDAC1, with IC50 values as 0.086 mM, 1.85 mM, respectively. It also showed better cytotoxic activities against A549 and SK-Mel-2 in cellular assay than Sor- afenib and SAHA, with IC50 values of 9.11 mM and 5.40 mM, respectively. 6. Dual kinase/BRD4 inhibitors The bromodomain and extraterminal (BET) proteins family constitutes a subclass of bromodomain containing proteins (BRDs), epigenetic modulator proteins. Bromodomains (BRDs) are about 110 amino acid domains that regulate transcription, gene splicing,chromatin remodeling, protein scaffolding and signal transduction. Therefore, play crucial roles in cell division and proliferation. Members of the bromodomain and extra terminal (BET) protein family (BRD2, BRD3, BRD4, and BRDT) have been involved in many disease pathways and are therefore considered promising drug targets [279,280]. Fig. 24. Structure of Dasatinib 32, structure of dual Bcr-Abl/HDAC inhibitor hybrids 96. Fig. 25. Structures of Idelalisib 97, Pictilisib 99 and dual PI3K/HDAC hybrids 98 and 100. Fig. 26. Structures of TBB 101, DMAT 102 and dual CK/HDAC inhibitor hybrids 103 and 104. BRD4, which is one of the bromodomain family, is associated with interphase chromatin and the chromosomes of mitotic and meiotic cells. BRD4 inhibition exerts a broad spectrum of desirable biologic effects including anticancer properties. Importantly, BRD4 inhibition downregulates oncogenic MYC transcription factors in several cancer cell lines. Many efforts are underway to develop chemically diverse and highly potent BRD4 inhibitors as new can- cer therapeutics [281,282].Furthermore, it is believed that dual kinase/epigenetic inhibitors would be a promising kind of anticancer drugs because they simultaneously influence diverse signaling pathways, as their ki- nase inhibitory activity could block the cell growth and prolifera- tive signaling while their epigenetic inhibitory activity would affect the expression of oncogenes and tumor-suppressors. Several drug combinations based on kinase inhibitors and epigenetic regulatory molecules have been proved more effective than a single agent in clinical trials or preclinical cancer models, which also support the development of dual kinase/epigenetic inhibitors as anticancer drugs [283e286]. Additionally, the concept of kinase-bromodomain dual in- hibitors has been explored in recent years. Several kinase inhibitors have been identified as bromodomain inhibitors by binding to the KAc binding pocket [287]. These finding has spurred many re- searchers to develop compounds that works to discourage both kinase and BDR4 proteins.Most hybrids designed with BDR4 was merged with PLK1 in- hibitors. PLK1 or Polo-like kinase 1 is a serine/threonine kinase which plays diverse regulatory functions in mitosis. It is always upregulated in various solid tumors and acute myeloid leukemia (AML), and its overexpression frequently associated with poor prognosis and survival. Thus, inhibition of PLK1 has proven to be a potential treatment for different cancers. In spite some patients with solid tumors responded well to single-agent PLK1 inhibitors, the overall antitumor activity was modest in clinical trials, and drug combination strategy appears to be essential for the application of PLK1 inhibitors in clinical [288e290]. Several recent reports high- lighted that the inhibition of polo-like kinase 1 (PLK1) synergizes with BRD4 inhibition in multiple cancer types, including prostate cancer and acute myeloid leukemia (AML) [286,291]. BI-2536 (106, Fig. 28) was originally developed as a selective PLK inhibitor and was subsequently proven to have dual PLK1/BRD4. The structural modifications induced to 106 was the only technique used to develop more potent mutual inhibitors of PLK/BRD4. Hy- brids 107-112 (Fig. 28) were designed as dual PLK1/BRD4 inhibitors mainly by extending structure of 106 either by replacement of N- methyl lactam of BI-2536 either by substituted-1,2,4-triazole group as in 107 [292], or simply varying substitutions as in the potent hybrids 108 [293] and 109 [294]. PROTAC strategy (proteolysis targeting chimeras) was used in identification of hybrid 110 as a dual inhibitor via extending the structure of 106 [295]. All the designed hybrids were active in a nano molar range but the use of only varied substitutions is usually a challenge in the design of such hybrids as varying substitutions would usually promote one ac- tivity over the other, thus optimization is critically required to reach a fixed pharmacophore for the development of this group of hybrids. Simple structural modifications even can shift activity toward other kinases. For example; changing cyclopentane into a thio- phene group in 111 (Fig. 28) resulted in designing the first dual ALK/ BRD4 inhibitor. The resulting compounds 111 showed improved ALK inhibitory activity with maintaining BRD4 activity, while the PLK1 activity was diminished [296]. More extensive modifications seen in 112 shifted the activity toward dual CDK9/BRD4 inhibition [297]. Fig. 27. Structures of Sorafenib 1, dual BRaf/HDAC inhibitor hybrids 105. Fig. 28. Structure of BI-2536 106, dual PLKI/BRD4 inhibitor hybrids 107-112 and their activities against kinases and BRD4. Another example of simple structure modification was seen on TG101209 structure (113, Fig. 29). TG101209 is primary targeting JAK2, IC50 for BRD4 is 120 nM accompanied by significant cyto- toxicity [298]. Transforming sulfonamide group to NHSO2 group along with the introduction of halogen substituents in 114 signifi- cantly increased binding potential to BRD4 (Fig. 29). 7. Dual kinase/PARP inhibitors Poly (ADP-ribose) polymerase or PARP is a class of enzymes that catalyze the transfer of ADP-ribose from nicotinamide adenine dinucleotide onto acceptor proteins. PARP-1 is the most abundant member of this family. It plays a significant role in single strand DNA repair via Base Excision Repair (BER) mechanisms [299e301]. In cancer cell PARP inhibition leads to BER inhibition that results in inhibition of DNA repairing and cancer cell apoptosis, and it is thus increasingly considered a vital target for cancer therapy [302]. Four PARP inhibitors, Olaparib, Niraparib, Rucaparib, and Talazoparib, are approved by the FDA or EMA for the treatment of BRCA mutant ovarian and breast cancer [303e306]. Mutual effects were recorded between PARP and certain kinases mainly EGFR and c-MET proteins. Combined inhibition of kinases and PARP pathways boosts the cytotoxic effects associated with the single use of either of them [307]. Interestingly, PARP inhibitors resistance in various tumors is reversed with the use of c-Met and EGFR inhibitors with great efficacy achieved in aggressive triple negative breast cancer cells [308]. The use of such combination is currently extended to advanced lung cancer as well as head and neck squamous carcinoma [309]. Thus, the design of PARP/TK in- hibitor hybrids is a promising tool to replace the use of combina- tions of both drugs in clinical settings. These hybrids were designed by simply adding up single entities together in one compact structure. This strategy keeps all drugs almost intact to potentially guarantee the fitting of each drug into its expected target. The total molecular size should be taken into consideration, so it does not exceed the individual enzyme pockets targeted.Julie Schmitt et al. [310] designed a single molecule JS230 117 (Fig. 30) as dual EGFR/PARP inhibitor by joining fragments from Gefitinib 14, Busulfan 115 and Olaparib 116 structures utilizing triazole linker. The hybrid was superior in both cytotoxic effects and in individual mechanisms to its 3 parent compounds. JS230 117 demonstrated triple effect by EGFR inhibition with 7-fold greater than Gefitinib, DNA damaging potential 3 to 6-fold more potent than Busulfan and sustaining the damage by inhibiting PARP in a manner like Olaparib. The major fragment of Olaparib 116 was joined with major pharmacophoric fragments the reported PI3K inhibitor BKM120 118 to design dual PI3K/PARP inhibitors 119, (Fig. 30). Though a large molecular structure is formed but the size is smaller than if all structures combined [311]. Compound 119 was equipotent as its parent drugs in cellular assays against BxPC3, A2780, Jurkat, DU145, A549, CaKi-1, Ramos and SW620 cancer cell lines. 8. Dual kinase/HSP90 inhibitors Heat shock protein 90 (HSP90) is a molecular chaperone that help other proteins to maintain their assembly, stabilizes proteins against heat stress, and aid in degradation of protein, many of them are tumor generating proteins which implicated in cell prolifera- tion, angiogenesis, metastasis and invasion. Overexpression of HSP90 is common in several types of tumors. Thus, HSP90 inhibi- tion would lead to degradation of client proteins and inhibit the proliferation and growth of tumor cells [312e314]. Reports suggested that HSP90 supports different kinases func- tion and simultaneous attenuation of HSP90 chaperone complexes with certain tumor-driving kinases prolongs their efficacy and re- duces the dose required for such drugs, hence reduces their asso- ciated adverse effects [315]. Based on these findings, a fragment- based design was used in collecting the essential fragments of some kinase inhibitors mainly; PI3K, JAK, ALK inhibitors with those for HSP90. For example, hybrids 122 and 123 used a fused tricyclic pipridine-thiolopyrimidine system similar to the thiolopyrimidine found in the PI3K inhibitor, Apitolisib 120 and 2,4-dihydroxy-5- isopropylbenzate pharmacophore from the reported HSP90 inhib- itor AT13387 121 (Fig. 31) [316,317]. Similarly, a series of dual ALK/HSP90 inhibitor hybrids were designed through merging the fragments of HSP90 inhibitor com- pound AUY922 124 and the reported ALK inhibitor 125 (Fig. 31) [318]. Among these series the two hybrids 126 and 127 (Fig. 31) displayed higher activities against ALK and HSP90a.The intrinsic pathway of apoptosis is governed by the anti- apoptotic BCL-2 (B cell lymphoma gene 2) family of proteins that prevents the release of cytochrome C from mitochondria. BCL-2 protein upregulation prevents the cell apoptosis, while inhibition of BCL-2-BH123 interaction can induce apoptosis. Moreover, BCL-2, is one such protein that is highly upregulated in many tumors as compared to normal cells, and thus, it has become promising target for the development of anticancer drugs [320e322]. Fig. 29. Structures of TG101209 113, dual TK/BRD4 inhibitors 114. Fig. 30. Structures of Gefitinib 14, Busulfan 115, Olaparib 116, JS230 117, BKM120 118 and dual PI3K/PARP inhibitors 119. Fig. 33. Structure of dual CDK/BCL-2 inhibitor hybrids 129. 9. Dual kinase/BCL-2 inhibitors Isatin based hybrid; 129 (Fig. 33) was developed during design of indole based anti-cancer agents and found to trigger both targets inhibiting both CDK and BCL-2 with showed considerable anti- proliferative activities [323]. Fig. 31. Structures of Apitolisib 120, AT13387 121, dual TK/HSP90 inhibitor hybrids 122, DHP1808 123, 126 and 127. It is noteworthy that hybrids 122,123, 126 and 127 was merged with no spacer used. Lianbin Yao et al. [319] designed a dual JAK kinase/HSP90 protein using fragments of the JAK inhibitor Rux- olitinib 86 and the HSP90 inhibitor BEP800 using a long chain alkyl linker 128 (Fig. 32). The activity of these hybrids was non compa- rable to the previous merged hybrids pointing out one challenge mentioned earlier that size of the designed hybrids is a significant factor during designing multi-targeted hybrids. Fig. 32. Design of dual JAK/HSP90 inhibitor 128. 10. Dual kinase inhibitors/nitric oxide releasing hybrids Nitric oxide (NO) is a small reactive gaseous molecule involved in several physiological and pathophysiological processes, including cancer biology. The small size and relative lipophilicity allow it to cross cell membranes without the presence of receptors. NO produced in vivo by three isoforms of nitric oxide synthase: endothelial (eNOS), neuronal (nNOS), and inducible (iNOS). Un- luckily, NO is rapidly metabolized before it reaches its site of action; therefore, NO donors became essential, as they can produce a NO- sustained release prolonging its half-life with the estimated dose [324]. In cancer, NO has a dichotomous effect (concentration- dependent), since it can act as a tumor suppressor or progressor. In pico-to nanomolar concentration range, NO contributes to tumor promotion, while at higher concentrations, micro-to millimolar; NO has antitumor effects [325,326]. A high concentration of NO is usually attained via iNOS, which is highly expressed in malignant tumors and has an established relationship with NO's anti-neoplastic functions. These functions are associated with inhibition of cancer cell proliferation, cell cycle arrest and cell necrosis [327], and they are also associated with stimulation of apoptosis through upregulation of the tumor sup- pressor p53, degradation of proteosome of anti-apoptotic agents, production of mitochondrial Smac (a protein which encourages cytochrome c-dependent caspase activation) release, the release of cytochrome c upon an increase in mitochondrial permeability, generation of RNS and ROS, S-nitrosylation of Fas receptor, Bcl-2, glyceraldehyde- 3-phosphate dehydrogenase, and reasonable levels of DNA fragmentation [328,329]. Generation of NO in blood vessels adjacent to the tumor is usually correlated to protection against cancer associated with angiogenesis impairment [327]. Additionally, NO possesses antioxidant effects that trigger cryoprotection [327] as it breaks cell damaging radical propagation [330]. Many NO donors, such as sodium nitroprusside and glyceryl trinitrate, have been shown to possess potent cytotoxicity and in- duction of apoptosis, angiogenesis inhibition [331]. Several hybrids were developed to produce a known anticancer agent that is able to release NO. These hybrids may be beneficial in providing lower doses of the original chemotherapeutic drug to lower the toxicity, while releasing the gaseous transmitter potentially offering an additional anticancer effect [332,333]. For example, NO enhances the taxol-induced cytotoxicity in carcinoma cells by increasing influx of taxol intracellular compartment [334]. It also introduces a synergistic effect noticed in the case of 5-Flourouracil/Dia- zeniumdiolate conjugates [335], and promotes the cytotoxicity of Cisplatin [336]. Increased efficiency of cytostatic therapy and retardation of drug resistance to anticancer agents were also re- ported upon co-administration with NO donors [337]. Similarly, NO can potentiate the ability of kinase inhibitors to produce cell death or inhibit the proliferation of cancer cells as in the case of two EGFR signaling pathway inhibitors Erlotinib and ZD1839 [338,339]. The combination of H89 (a small molecule that has been known as PKA inhibitor) and nitroglycerin (NTG) en- hances cell death in colon cancer cells. Indeed, H89 induces reactive oxygen species (ROS) induction which combines with NO, triggers apoptotic cell death by caspase activation [340]. A complex regulatory relationship exists between protein ki- nases, kinase inhibitors and NO in the cancer network. PKIs may exert an influence on endogenous production of NO. In contrast, NO (from other sources, mainly NO donors) may have an impact on protein kinases [327]. NO can act as a protein kinase activator [341], since the intracellular production of NO by NOS is required for the phosphorylation of subsequent activation of EGFR signaling path- ways. Besides, NO produced from NO-releasing agents can also positively regulate the EGFR pathways [342]. On contrary, NO may act as a protein kinase inhibitor [327], in agreement with the well- documented ambivalent role of NO in cancer. Various studies re- ported a negative regulatory effect of nitric oxide on protein ki- nases. Actually, NO can inhibit pathways involved in the proliferation and survival of cancer cells via its potential to down- regulate the activity of several protein kinases [343]. 10.1. EGFR inhibitor/NO release hybrids The mutual relationship between EGFR and NO put a lot of sense into the concept of designing NO donating EGFR inhibitors. How- ever, to date, the design of such agents did not involve hybrids based on FDA approved inhibitors to validate the usefulness of such hybrids. The design of these hybrids mainly depends on small molecules carrying structure similarity to approved inhibitors. Although these hybrids reported a retained ability to induce cyto- toxicity, many research steps are required to allow clinical use of such hybrids due to the complex nature of NO. The strategy used in designing such hybrids was based on the use of a structure similar to a known inhibitor, followed by incor- porating a small alkyl group (4e5 carbon length chain) carrying a nitrate group. The main feature of the nitrate ester is that they are rapid producers of a high amount of NO that would probably need extensive care in designing an appropriate delivery method of these agents to tumor site. Examples used for this strategy were based upon using the steroidal nucleus of oleanolic acid 130; hederagenin 131 to develop hybrid 132 (Fig. 34) [344]. Similarly, bioisosteric similarity of pyrazolopyrimidine with the kinase inhibitor Erlotinib 13, was used in the design of a class of EGFR inhibitors/NO donors (133 and 134, Fig. 34) using the alkyl bridge strategy [345]. Fig. 34. Structure of oleanolic acid 130, hedragenin 131, and EGFR inhibitors/NO releasing hybrids 132-134. The hybrids designed by this strategy showed minimal devia- tion from the pharmacophore required for EGFR inhibition thus successfully retained the ability of their parent drugs to inhibit EGFR (IC50 ¼ 0.01 and 3e51 mM for hybrids 132 and 133, 134, respectively in both hybrids. The increased ability to produce NO seemed to be beneficial as the hybrids were superior to their anticancer effects and hybrid 132 the NO released might contribute to overcome resistance in gefitinib-resistant H1975 with IC50 of 8.1 mM and osimertinib-resistant H1975-LTC with IC50 of 7.6 mM. Another strategy used to design dual EGFR/NO hybrids used oxime as a nitric oxide group. This approach was probably adhered to induce control over NO release compared to uncontrolled release from a nitrate ester. NO release from an oxime is initiated by a thiol and requires particular oxidative enzymatic interference to pro- ceed. Using oxime is also advantageous in achieving proper phys- icochemical parameters as it does not introduce tremendous changes to lipophilicity of the parent compounds compared to the nitrate ester keeping the cellular entry rate almost constant. This strategy also did not use approved EGFR inhibitors but only small molecules with reported inhibitory activity. Abdel-Aziz group used this strategy to transform acylated 1,2,4-triazole 135 reported as EGFR inhibitor (Fig. 35), into NO releasing hybrid 136 [346] and also in order to design xanthine based EGFR inhibitor/NO donor hybrids [347] 139 and 140 based on the kinase inhibitors 137 and 138 (Fig. 35). Both hybrids were superior anti- cancer agents than their parent compounds although their ability to induce EGFR inhibition was slightly altered. It is worth mentioning that hydroxyiminophenethyl hybrids 140 were more active than the hydroxyimino-ethyl phenyl acetamide derivatives 139 suggesting the importance of a specific location for the oxime moiety and raising a question about the role of NO in the increased activity of such hybrids. Fig. 35. Structure of reported EGFR inhibitor 135, 137 and 138 and the designed structure of 1,2,4-triazole based EGFR inhibitor/NO releasing hybrids 136, 139 and 140. 10.2. VEGFR inhibitor/NO release hybrids Encouraged by the success observed by EGFR/NO hybrids, the same research group extended their work to develop dihydropyr- imidine derivatives containing oxime as NO releasing moiety (142, Fig. 36) joining dihydropyrimidine scaffold as in Monasterol 141, and the pharmacophoric features of the kinase inhibitor drug Sor- afenib 1 [348]. A very promising results were expected from such hybrids as they fully go with the general structural features of VEGR in- hibitors; consisting of four distinct moieties, these moieties include hinge-binding moiety, linker (3e5 atoms), hydrogen bonding moiety (amide/urea) and a hydrophobic tail that occupy the allo- steric hydrophobic pocket of the enzyme. Nonetheless, the cyto- toxicity screening for NO-hybrids and the non-oxime intermediate compounds revealed that the NO releasing hybrids 142 showed weak anticancer activity against most of the tested cancer cell lines than the non-oxime intermediates. One hybrid, 142b; the para- methoxy phenyl derivative; showed a moderate cell growth inhi- bition against most of the tested cell lines with growth inhibition percentages of 20.51%e41.90%, VEGFR-2-kinase was tested against 142c whereas Sorafenib was used as a reference (142c IC50 > 1 mM, Sorafenib IC50 ¼ 0.00017 mM).

10.3. PI3K/AKT inhibitors/NO release hybrids

Phosphoinositol-3-Kinases (PI3Ks) are a class of related intra- cellular signal transducer lipid kinase enzymes, which can phos- phorylate the 3-hydroxyl group of the phosphatidylinositol’s ring. Disorganization of PI3Ks and their downstream molecules has a significant impact on initiation, growth, proliferation, and survival of cancer cell [349]. It is the first signal transducer in the PI3K/ protein kinase B (Akt)/mammalian target of rapamycin (mTOR) signaling pathway that is one of the most common contributors to many human tumors [350,351]. Thus PI3K is considered as a po- tential target for cancer treatment.

Inhibition of PI3Ks showed potent cytotoxic activity. Four recently approved PI3K inhibitor drugs [72,352e354] including Idelalisib 97 and several others in ongoing clinical trials such as Pictilisib 99 (GDC- 0941) and Apitolisib 120 as monotherapy or in combination with other therapeutic agents for cancer therapy [355,356]. The latest reported PI3K inhibitors displayed high rates of resistance whilst other survival and growth-related pathways are simultaneously activated, rendering their use as a single agent unsuitable in cancer treatment. Therefore, PI3Ks inhibitors com- bined with agents of other pathways to target multiple pathways [356].

A successful hybrid of a dual PI3K inhibitor/NO releasing donor was designed through joining furoxan-based NO-donating hybrids
[357] with b-elemene like structure (143, Fig. 37). b-elemene is a sesquiterpene compound isolated from the traditional Chinese medicinal herb Curcuma aromatic Salisb that has been used as an anticancer drug in China for many years. The two pharmocophoric entities were linked via either alkyl or amino acids linker, to form hybrids 144 and 145 (Fig. 37) that possess potent kinase inhibition activity with the ability to release NO.

NO incorporation in both hybrids significantly enhanced anti- proliferative activity against three cancer cell lines (U87, SGC- 7901 and HeLa) compared to parent compound b-elemene. Inter- estingly, these hybrids showed excellent sensitivity to U87 cells with IC50 values ranging from 173 to 2 nM. The role of NO was clearly demonstrated by the ability of most hybrids to produce high concentration of NO in vitro, and by the fact that the anti-tumor activity of compound 144 in U87 cells was significantly attenu- ated by NO scavenger (hemoglobin or carboxy-PTIO).

10.4. MEK inhibitors/NO release hybrids

The RAF/RAS/ERK/MEK system is a crucial signaling pathway involved in regulating cell differentiation, proliferation and apoptosis. Irregular activation of this pathway leads to various cancers and other diseases. The mitogen-activated protein kinase enzymes (MEK1 and MEK2) are one of the critical kinases in this pathway. The extracellular signal-regulated kinase ERK appears to be the primary substrate for MEK. Allosteric MEKs inhibitors have demonstrated authentic anti-tumor efficacy and low toxicity in clinical trials. Drugs targeting MEK pathway have distinctive char- acteristics: (a) they bind to an allosteric binding site without competing with ERK or ATP, (b) they cause conformational changes to MEK and lock the unphosphorylated form in a catalytically inactive state, and (c) there is no symmetric sequence between the inhibitor binding site of MEK and other protein kinases. These properties made MEK a highly selective and very promising target for anticancer drug discovery [358e360].

Combination of a MEK inhibitor with an NO-donor can synergistically inhibit proliferation and invasion of cancer cells, and previously reported hybrids of MEK inhibitor and NO releasing agents exhibited potent anticancer effects in many cancer cell lines [361,362].
The design of MEK/NO-donating hybrids used merging the furoxan NO-donor group found in a week previously studied MEK inhibitor 146 with the potent reported inhibitors G8935 147 to yield RO5126766 148.148 is currently in phase I clinical trials [363]. Hybrid 182 were used in further develop more potent hybrids 149 and 150 [364] (Fig. 38). Interestingly, these compounds were more potent anti-proliferative agents than the parent drug 146, although they produced a lower concentration of NO compared to parent compound 146, but NO was probably responsible for activity against highly resistant cell lines and reduced toxicity compared to the parent drug. Though this activity might be attributed to incorporation of a chalcone that might contribute to activity.

Fig. 36. Structures of Sorafenib 1, Monasterol 141 and the design of dihydropyrimidine based VEGFR inhibitor/NO releasing hybrids 142.

Fig. 37. Structures of b-elelmene 143 and b-elemene based PI3K/AKT inhibitor/NO release hybrids models 144 and 145.

Fig. 38. Structures of reported MEK inhibitors 146, 147 and design and structure of phenyl-sulfonylfuroxan based MEK inhibitors/NO release hybrids 148-150.

The failure of these hybrids to produce consistent responses might be attributed to the structure extensions added to bind the NO donating group. In addition, result obtained from NO donating hybrids is complex and cannot be attributed mainly to NO pro- duction, structural consideration must also be accommodated. To date, there is insufficient evidence introduced correlating any observed synergism merely to NO release, and the activity of these hybrids will depend on multiple factors including NO release and extended to the type of linker used and structural and electronic factors that may initially affect the pharmacophore for the esti- mated target and the ability of the hybrid to release NO under different cellular conditions. The complexity of NO responses in biological systems dependent on thiol and ROS status at the target point might also hinder an exact prediction of these hybrids behavior at the site of action and prevents their use in clinical settings even with proven in vitro evidence of their activity.

11. Conclusion

The use of conventional chemotherapeutics for cancer treat- ment has always been a challenge. Although some of these agents made a leap in treating a disease that was rarely combated in the past, they have major limitations to their use in cancer treatment. These limitations involve many aspects; for instance; their high toxicity that includes hair loss, decreased immunity as well as emergence of resistance and failure to produce response in certain patients or certain tumor types. These limitations made the dis- covery of protein kinase inhibitors as a possible cancer treatment more appealing as they are offering a targeted therapy with fewer side effects. Since the introduction of Imatinib in fighting CML, medicinal chemistry research started to put huge efforts in the design and development of kinase inhibitors as potential anti- cancer agents. Several success stories were achieved with an average of 4 kinase inhibitors approved each year by FDA for cancer treatment.

Given the multifactorial nature of complex diseases such as cancer, a combination therapy was always a solution to accomplish more success in hitting the disease. Combination regimens are usually implemented via using two more therapies targeting different cancer pathways to either achieve synergy and/or decrease side effects. Despite its profound success, this strategy is a challenge to be utilized in cancer therapy due to emergence of mutual resistant malignancies and disrupted pharmacokinetic profiles that requires further investigation regarding dosages and metabolic studies.

Hybridization emerged as a potential strategy to circumvent these limitations by placing a single drug candidate attacking different targets and possibly offering advantageous characters of combination therapy without their observed limitations. One group of cancer therapeutics that is extensively used in hybridization strategy is the protein kinase inhibitors. Thus, this review intro- duced a survey of hybrids including kinase inhibitors published in the last five years. The hybrids discussed was designed was joining PKIs with non-kinase targets, since targeting more than one kinase was previously discussed in other articles. A huge pile of hybrids was designed with joining PKIs with tubulin polymerization in- hibitors as a conventional therapy. High percentage were also seen with targeted epigenetic inhibitors such as HDAC inhibitors. Hy- brids with endogenous bio mediators such as NO were also listed. Great number of the used PKIs were usually addressed at inhibiting EGFR and VEGFR tackling various tumor cell lines and affecting both tumor growth, angiogenesis and metastasis. The hybrids listed either joined two different motifs targeting two different pathways or using one structure motif that is common for targeting the two addressed pathways. The chemistry beyond the design is usually based on using biologically active heterocycles including quino- lines, quinazolines, pyrimidines, triazoles among othersingle and fused rings. Linker chemistry is also a major aspect studied in designing hybrids, with those introducing hydrogen bond donor or acceptor groups being the most beneficial.

Special attention should be paid to the attachment points for joining the hybrid fragments away from the pharmacophores and not destroying them. Considerations about total size of the hybrid and physicochemical parameters should be also counted for.In conclusion, the design of a dual target drug candidate remains in urgent need for achieving progress in cancer prognosis and im- proves life expectancies, though it is still in early stages and needs extensive efforts to be able to introduce more successful leads to clinical trials and approval stages.

Declaration of competing interest

The authors declare that they have no known BSJ-4-116 competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.