Biocatalysis: An Industrial Perspective

EntreChem: Building a Sustainable Business Case in Biotechnology: From Biocatalysis to Synthetic Biology

EntreChem S.L., Vivero de Ciencias de la Salud, Calle Colegio Santo Domingo Guzmán s/n, 33011, Oviedo, Spain

4.2 Biocatalysis

4.2.1 Enantiopure Chiral Building Blocks

Chiral building blocks are essential to create sophisticated libraries in medicinal chemistry or as raw materials for process development of chiral APIs. EntreChem's technology, based on green biocatalytic technology offered by lipases, ketoreductases and transaminases, enables a growing offer of chiral products, especially chiral amino alcohols and diamines of high enantiomeric and chemical purity. The processes used for the laboratory synthesis are fully scalable, to ensure satisfactory re-supply of any of the isomers listed in Figure 4.1. As can be seen, we have developed chemoenzymatic routes for every enantiomer of both cis- and trans-1,2-cyclopentyl- and 1,2-cyclohexyl amino-alcohols and diamines, including monoprotected derivatives to facilitate the use of these building blocks in medicinal chemistry.

Figure 4.1 Portfolio of building blocks offered by EntreChem.

In this sense, enzymatic syntheses are an exciting and emerging opportunity in sustainable chemistry due to the intrinsic green features of enzymes which have water (a safe, non-toxic, biorenewable and cheap solvent) as their natural environment, enhancing the expectations of biocatalysis-based chemical manufacturing.

EntreChem also provides custom services to clients, in order to prepare other building blocks similar to those of Figure 4.1 or to develop novel chemoenzymatic routes to key pharmaceutical intermediates. A recently disclosed example is the synthesis of the heart-rate reducing drug ivabradine (Scheme 4.1).^1–3 Introducing chirality in the intermediate is challenging because the amino hot spot is primary and therefore remote to the chiral center. However, this was solved by a lipase-catalyzed alkoxycarbonylation with diethyl carbonate, and to kill two birds with one stone the carbonate product itself can be reduced to install the methylamino group present in the historical chemical processes of the sponsor.⁴ This is particularly convenient since it does not alter the end-game of the synthesis, the use of the same penultimate molecule as in the original process helps keep regulatory changes to a minimum. However, despite being operationally simple, this approach suffered from important drawbacks such as the recalcitrant 50% maximum yield inherent to KRs, the low enantioselectivity and the need for flash chromatography (enantiopure ivabradine in 30% overall yield for a three-step sequence). Alternatively, a route based on transaminases (TAs) or ketoreductases (KREDs) starting from an aldehyde precursor provided a chiral amine or alcohol, respectively. The amine would require two steps to the final product (including the same penultimate as above), but the alcohol could be converted in one step into the final ivabradine. Meanwhile the bioreduction displayed null enantioselectivity, the biocatalytic amination, performed at 25 mM and laboratory scale and involving a DKR, led to the target ivabradine in 50% yield in a four-step sequence without the need of chromatographic purification.

Scheme 4.1 Chemoenzymatic routes to ivabradine.

4.2.2 Cascade Processes Taking Advantage of Biocatalysis

Cascade and one-pot reactions represent an exciting recent development in white biotechnology.^5–9 Although the concept of performing enzymatic multistep syntheses in a concurrent fashion is not very recent (multistep reactions in whole cells were used as early as the 1980s for the production of amino acids),¹⁰ it has received increased attention in the past few years. Especially from a green chemistry point of view, cascades represent a very promising approach, particularly due to the avoidance of intermediate downstream and purification steps. These often contribute significantly to the overall environmental impact of a reaction. Furthermore, workup and isolation of intermediate products binds production capacities and resources thereby also significantly contributing to production costs on industrial scale. On a more fundamental level, multistep cascades can be considered as biomimetic approaches. Although metabolic engineering was developed as a successful technology for the targeted manipulation of naturally evolved metabolic pathways (see next section), the synthesis of complex, high-value pharmaceutical compounds often lies beyond the alteration of existing pathways.

Figure 4.2 summarizes the concept behind the design of successful cascade processes. The idea is to combine cascade units (individual reactions) in a one-pot, if not fully concurrent, procedure. Typically, each unit will rely on a catalytic system, which could be based on biocatalysis (isolated or whole cells), metal-catalysis or organocatalysis. Each cascade unit must be carefully studied in terms of key parameter robustness (e.g., temperature, solvent, concentrations of enzyme and substrate, cofactors or other reagents), not only to find the optimal ones, but to identify the upper and lower limits of each parameter, since the assembly of the cascade may require operating a particular cascade unit outside optimal parameters when practiced as a standalone reaction. Importantly, not only “within-unit” parameters are to be considered, but also parameters from other cascade units that could cause incompatibility of operation between those units. This is the case for cofactors or auxiliary reagents needed in one but perhaps very detrimental for another unit, not to mention the presence of the other catalysts themselves. The compatibility of such cross-parameters determines more often than not the final operation of the cascade, which might need to sacrifice the all-concurrent operational feasibility and settle for the one-pot operation. In any case, the advantages of such a setting as mentioned above with respect to step-wise, or even telescoped, processes are still significant.

Figure 4.2 Concept of cascade processes involving bio-, metal- and organo-catalysis.

As for the possibilities to combine several isolated enzyme classes to facilitate a one-pot cascade, the use of redox enzymes has been successfully demonstrated in recent years.^5,6 In this context, EntreChem has developed very recently an enzymatic cascade in water consisting of a stereoselective bioreduction of cyclic β-ketonitriles concurrently coupled to a whole cell-catalyzed nitrile hydrolysis in one-pot (Scheme 4.2).¹¹ The first step, mediated by NADPH-dependent ketoreductases (KREDs), involved a dynamic reductive kinetic resolution (DYRKR),¹² which led to β-hydroxynitriles in very high enantio- and diastereomeric ratios. Then, the simultaneous exposure to nitrile hydratase and amidase from whole cells of Rhodococcus rhodochrous provided the corresponding cyclic β-hydroxyacids with excellent overall yield and optical purity for the all-enzymatic cascade.¹³

Scheme 4.2 Enzymatic cascade syntheses of cyclic β-hydroxyacids.

The success of the strategy lay both in the synchronization between a fast racemization of the epimerizable stereocenter before ketoreduction in the first step and a microorganism that converts the cyano group of the β-hydroxynitrile rather than that contained in the starting β-ketonitrile. Interestingly, the KREDs showed marked cis-diastereoselectivity, complementing the existing methodologies, mostly aimed at trans-diastereoisomers, readily available from epoxides and aziridines. The strategy was particularly effective with the cyclohexyl and cycloheptyl substrates, and the adequate selection of stereocomplementary KREDs delivered both antipodes of the cis-2-hydroxycycloxane- and cis-2-hydroxycycloheptane-carboxylic acids in >95% yield and total selectivity of >99 : 1 dr and >99% ee at 20 mM substrate concentration and lab-scale (Scheme 4.3). Although operationally viable, the concurrent process for the five-membered ring homologue was challenging since the optimal selectivity in the ketoreduction was achieved at pH 5.0, which inhibits the activity of R. rhodochrous. This issue was circumvented by means of a sequential setup and, once the bioreduction was completed, the only experimental setting consisted of a slight pH increase to 7.0 and further addition of the bacterial suspension.

Scheme 4.3 Biocascades towards 2-hydroxycycloahexane- and cyclopentanecarboxylic acids.

Besides enzymatic cascades, the integration of enzymatic and chemocatalysis in one-pot sequences represents a more attractive and unexplored challenge since chemists have historically focused on combining reactions belonging to one of these technologies exclusively.^14–16 Actually, examples of metal catalysts (or organocatalysts) working hand in hand with enzymes are still very scarce, especially in aqueous media,¹⁷ and imposes several challenges, namely (i) catalyst compatibility and stability, (ii) cross-reactivity, (iii) shifting the reaction equilibrium to the product side, and (iv) scalability. Despite these facts, the pool of metal catalysts able to operate in water has expanded enormously in recent years, providing novel opportunities to combine with enzymes which have water as their natural environment.

In this context, EntreChem developed the first combination of ω-transaminases (ω-TAs), which catalyze the transfer of an amino group from an amino donor onto a carbonyl moiety, with a metal-catalyzed reaction, namely the ruthenium-promoted isomerization of allylic alcohols which delivers carbonyl compounds (Scheme 4.4).¹⁸ Actually, this widely-studied isomerization had been recently performed in non-conventional solvents such as water or deep eutectic solvents, enabling such a combined process in aqueous media. Once both steps had been independently optimized, studies of compatibility revealed that both ω-TA and cofactor (pyridoxal-5′-phosphate, PLP) impacted negatively the metal catalyst. Consequently, the process was accomplished in a sequential fashion at lab-scale. Indeed, the aqueous medium from the metal-catalyzed reaction (200 mM substrate concentration) was used directly to feed the bioamination with the only telescopic adjustment of a ten-fold dilution of the medium before adding the ω-TA and PLP. As a result, and taking advantage of the available ω-TAs, both antipodes of chiral amines were isolated in very high yield and enantiomeric excess without the need for further purification.

Scheme 4.4 Ruthenium-catalyzed allylic alcohol isomerization combined with an enantioselective enzymatic amination in a one-pot sequential process.

In a further study, the metal-catalyzed allylic alcohol isomerization was also combined with a bioreduction mediated by KREDs, enabling in this case the isolation of the analogue chiral alcohols (Scheme 4.5).¹⁹ On this occasion, both reaction conditions and reagents of each step were fully compatible and a concurrent process was efficiently implemented at 200 mM substrate concentration at lab-scale. Thus, the chemical and the enzymatic step proceeded simultaneously from the beginning in the same pot to furnish the products in high yield and enantioselectivity. Furthermore, a detailed study of the reaction kinetics revealed a rapid deactivation of the enzyme under the reaction conditions and, consequently, the faster the isomerization takes place, the higher the overall yield of the cascade. Thus, and despite a slight drop in the yield compared to a parallel sequential process, this first example of a genuine concurrent metal and biocatalyzed reaction in water is an important contribution to the field of cascade processes, as it does not require site-isolating techniques such as compartmentalization or encapsulation.

Scheme 4.5 Ruthenium-catalyzed allylic alcohol isomerization combined with an enantioselective bioreduction in a one-pot concurrent process.

4.3 Drug Development

4.3.1 Natural Products in Drug Discovery

Nature offers an endless pool of bizarre and sophisticated molecular entities with desirable drug-like properties, rendering them ideal starting points for development of pharmaceuticals.^20–22 The chemical structures of natural products (NPs) are the result of an on-going combinatorial chemistry performed by living organisms over millions of years, providing them multiple advantages related to their growth and survival. This explains the unique ability of these entities to specifically interact with biological target molecules. Thus, the history of NPs in drug discovery has been extraordinarily successful in the second half of the 20th century, highlighted by prominent examples such as the antitumor agents taxol, vinblastine or doxorubicin, the immunosuppressants for organ transplants cyclosporine and rapamycin, or the cholesterol-lowering agents statins, many of them top-selling drugs in recent decades.²² However, in the 1990s, at the height of NPs success, the discovery focus shifted to combinatorial synthetic chemistry. This highly novel technology changed the workflow by delivering thousands of molecules in weeks if not days, in contrast to the slow pace of NPs lead discovery, further aggravated by the lack of innovation, limited to screening strategies that by then provided many repeated leads after decades mining the “wild type” NPs space. Additionally, most discovery programs in NPs relied not on purified molecules but on extracts, whose phenotypic bioactivity measurement fell out of favor on the wake of the imatinib success story, the first targeted inhibitor, in the mid-1990s. Thus, owing to the continuous pressure within the market to find “blockbuster drugs” and dazzled by the novel breakneck assay speed of combinatorial chemistry, most firms cut back their programs of natural product drug discovery. The main argument against doing further screening of natural sources was related to the incompatibility with high-throughput screening (HTS) analysis, based on issues such as (i) high structural complexity of NPs, (ii) problematic supply and re-supply of the drug lead/candidate, and (iii) lack of efficient dereplication methodologies.

Over time and despite massive investment and millions of pure synthetic compounds screened, the initial expectations of synthetic medicinal chemistry were not met, the number of new drug approvals decreased year by year. Thus, synthetic libraries yielded a “hit rate” of <0.001%, significantly lower than the 0.3% hit rate obtained in the same period for the family of polyketide metabolites (20 commercial drugs from just over 7000 known structures).²³ Actually, according to a quite recent study by Bauer et al.,²⁴ principal component analysis of 20 structural and physicochemical characteristics of 40 top-selling drugs, 60 natural products and 20 compounds from commercial drug-like libraries illustrates that NPs interrogate a different area of chemical space than synthetic compounds. Indeed, most non-NP drugs and drug-like leads explore narrow and underrepresented regions of biologically relevant chemical space, effectively missing opportunities to discover new targets as evidenced by the space covered by approved NP drugs. As a result, libraries of drug-like molecules have proven ineffective against a variety of challenging targets, such as protein–protein interactions, nucleic acid complexes, and antibacterial modalities, highlighting the need for phenotypic screenings to discover new medicines beyond the target–mechanism of action approaches.²⁵ Additionally, since up to 99% of microorganisms have yet to be discovered and the currently high medical need in oncology or diseases in the Third World, the dispensing of NPs is a luxury we cannot afford.²⁶ Besides, the innovative approaches for screening living organisms as well as advances in emerging areas such as chemical biology or enzyme catalysis fuel the idea of an imminent second youth for NPs in drug discovery.²⁷

4.3.2 EntreChem’s Approach to Natural Products Drug Discovery

EntreChem’s strategy for bringing NPs back to the forefront of drug discovery avoids the early, now unproductive, stages of classical NPs-based drug discovery programs (isolation of environmental samples, extract analysis, dereplication, lead identification) and starts with a known lead compound (either an approved, clinical-stage or bioactive molecule) over which different technological approaches are applied to obtain new analogues difficult to obtain by other methods. These new analogues are the source of better candidates for preclinical development, since the small-size, focused libraries allow phenotypic screening including in vivo testing to prioritize low toxicity, high bioactivity analogues. Additionally, since the novel analogues come from molecules with somewhat known properties, technical and commercial risks associated with completely unknown product families are minimized (Figure 4.3).

Figure 4.3 EntreChem’s strategy to natural products drug discovery.

Advances in bacterial recombinant DNA technology have enabled the cloning of gene clusters involved in the biosynthesis of many bioactive NPs produced by microorganisms, as well as detailed knowledge of their metabolic pathways, significantly raising the potential of the combinatorial biosynthesis field.²⁸ Complementary to chemical synthesis and microbial fermentation, the manipulation of genes governing secondary metabolic pathways offers a promising alternative for preparing complex NPs and their analogues biosynthetically. Gene clusters encoding many NPs have been cloned and characterized, and it is now possible to introduce specific structural alterations into a NP in the presence of abundant functional groups by rational manipulation of the gene cluster governing its biosynthesis. The resulting molecules can be produced in recombinant bacteria by large-scale fermentation and purified and isolated by conventional downstream processes (Figure 4.4).

Figure 4.4 EntreChem’s scheme for generating novel compounds by combinatorial biosynthesis.

On the other hand, biocatalysis has become a common tool in both academic and industrial chemistry. The efforts of chemists over recent decades have led to the rationalization of the mechanism of action of biocatalysts, which have been routinely incorporated into many synthetic sequences. Nowadays, a further step consists in expanding the application of the biocatalytic toolbox to the modification of complex molecular scaffolds common to many pharmaceutical leads isolated from nature, leading to new drugs with remarkable improved activity, stability and pharmacokinetic properties. Traditionally, some recalcitrant issues associated with NPs were the chemical fragility (sensitivity to pH, heat, metal ions, light, acid or basic media, etc.), structural complexity and functional diversity which make the transformation of a specific functional group a truly challenging task. Enzymes can circumvent most of the aforementioned problems since they exhibit high selectivity and operate under mild conditions in both aqueous and organic media. As a result, a biocatalytic approach could hypothetically reduce the number of protection/deprotection steps and introduce structural diversity inaccessible via conventional chemical synthesis.^29–31 This is particularly relevant given that only 6% of the 1024 new molecule drugs approved for the treatment of human diseases in the past three decades (1981–2008) are directly isolated NPs. In contrast, over half these molecules (57%) are either NP derivatives obtained by semi-synthetic approaches or formally synthetic compounds that are structurally related to, or inspired by, NPs.³²

EntreChem’s selection of parental compounds with known anti-proliferative properties was based on the previous research of the co-founders during the past 20 years in two families of natural products: (i) aureolic acids, a group of polyglycosylated aromatic polyketides bearing a tricyclic core, which bind DNA and are considered antitumor antibiotics; (ii) indolocarbazole alkaloids, originally described as protein kinase C and topoisomerase I inhibitors. Thus, inspired by early accomplishments, the aim was to explore the chemical space around these molecular entities in order to improve efficacy and, especially important, reduce toxicity, since both promising chemical families come with toxicity liabilities that had hampered their clinical development.

4.3.3 Aureolic Acids: The Quest for Clinically Viable “Mithralogs”

The aureolic acid family includes mithramycin A, chromomycin A₃, olivomycin A, UCH9 and durhamycin A.³³ They are all antineoplastic antibiotics against Gram-positive bacteria and also stop the proliferation of tumor cells. Several studies have pointed out that the basis for the antitumor properties of mithramycin and its analogues (mithralogs) is the inhibition of replication and transcription processes during macromolecular biosynthesis by interacting, in the presence of Mg²⁺, with GC-rich nucleotide sequences,³⁴ especially at the site of union of Sp1 transcription factor.³⁵ Mithramycin A (MTM), the most representative member of the family, was approved as an anticancer drug by the FDA in 1970, and used originally for the treatment of several types of cancer,³⁶ Paget’s bone disease and hypercalcemia³⁷ until its discontinuation at the turn of the 21st century – indeed commercial clinical batches have not been reported since 2000. MTM clinical limitations are due to severe toxicity (hepatotoxicity),³⁸ a common problem for other aureolic acids, like the ten-times more potent chromomycin A₃ (CRM), which, unlike MTM, was not able to progress to the clinical stage.³⁹

Notwithstanding this, recent literature evidence linking MTM’s mechanism of action to both antitumoral and other therapeutic activities has fueled renewed interest in this privileged natural scaffold. Actually, the growing body of literature on MTM (more than 400 PubMed-indexed articles since 2000 only) describing promising biological activity, especially in oncology, reflects the potential of this molecular class for the treatment of different types of cancer. This trend is crowned by two recent MTM clinical trials sponsored by the NCI (the first in decades) in two unrelated indications: Ewing sarcoma (clinical trial identifier: NCT01610570) and non-small-cell lung cancer (clinical trial identifier: NCT01624090), which address two different targets in each histology (EWS-FLI1 fusion gene in sarcoma⁴⁰ and ABCG2 transporter in lung cancer⁴¹). Likewise, CRM was identified through in silico analysis of the publicly available drug profiles from the NIH (National Cancer Institute) as an agent suitable to selectively target the loss of the von Hippel-Lindau (VHL) tumor suppressor gene in clear cell renal carcinoma.⁴² Seeing these facts, EntreChem planned the search for novel mithralogs with improved therapeutic window.

The complex chemical structure of MTM can be divided into two well-defined fragments, namely a sugar domain (disaccharide chain of D-olivose-D-olivose and trisaccharide chain of D-olivose-D-oliose-D-mycarose) and an aglycone unit constituted by a tricyclic core and an alkylic side chain (Figure 4.5). During the past decade, a plethora of new analogues have been obtained by combinatorial biosynthesis approaches such as gene inactivation, gene expression, the use of sugar biosynthesis plasmids for sugar modification, and combinations of both.⁴³ As a result, this research has revealed valuable information about the role displayed by each structural unit on the activity: (i) truncated derivatives lacking some sugar show, in general, very low activities; (ii) tetracyclic analogues generated by gene expression are also less effective than the parental; (iii) the glycosylation pattern is crucial for the activity, with some promising analogues generated by sugar biosynthesis plasmids.

Figure 4.5 Chemical structures of selected aureolic acids.

Drawing on our technology position, EntreChem introduced mithralogs with modifications in the alkylic side chain of the aglycone by a gene inactivation approach, in particular mithralogs bearing a shorter ketone side chain. Among them, MTM-SK (EC-7072) is generally as active, but one order of magnitude less toxic, as MTM. Another analogue, MTM-SDK (EC-7073) showed one order higher antitumor activity but it is even more toxic than the parental MTM.⁴⁴ Next, the focus was on modifying the glycosylation pattern by different techniques. Actually, MTM sugars are 2,6-deoxyhexoses synthesized from glucose-1-phosphate through four enzymatic steps to give rise to NDP-4-keto-2,6-dideoxy-D-glucose, which is further methylated and/or reduced to render activated deoxy-sugars that are substrates for glycosyltransferases. The most prominent from many novel analogues obtained by introducing different sugar biosynthesis plasmids was the one exclusively differing from MTM in the absence of a C-methyl group in the distal sugar (sugar E) of the trisaccharide (demycarosyl-3D-β-D-digitoxosyl-MTM, EC-7092).⁴⁵ A further iteration consisted of testing the simultaneous effect of these two structural features, namely the shorter side ketone chain and the demethylated E-sugar. Thankfully, the resulting demycarosyl-3D-β-D-digitoxosyl-MTM SK (EC-8042) exhibited potent antitumor activity in a National Cancer Institute hollow fiber assay, indeed one of the most potent ever registered on the NCI records (higher than MTM), emphasizing its potential as a broad antitumor agent.⁴⁶ EC-8042 has demonstrated activity in mice models of human tumors like triple negative breast cancer,⁴⁷ and liposarcoma,⁴⁸ among others. Importantly, EC-8042 is even less toxic than MTM-SK in mice, and also rats and dogs show one order less magnitude in maximum tolerated dose. This and other in vivo data in murine models of human tumors provided the basis for EC-8042 to be selected by EntreChem as a drug candidate, which is currently undergoing regulatory preclinical assays in order to file an Initial New Drug (IND) application that would allow testing in humans.

On the other hand, the biosynthesis gene cluster of chromomycin (CRM) has been cloned and characterized, showing that the CmmA gene encodes the acetyltransferase responsible for transferring the acetyl groups to the sugars. Moreover, inactivation of this gene led to a compound approximately 100 times less active, which highlights the importance of acetyl decorating groups for the antitumor activity of CRM. Based on these facts, and aiming at producing new acetylated mithralogs with enhanced activity, EntreChem used the CmmA acetyl transferase as a biosynthetic enzyme, in a bioconversion with whole cells set up. As a result, seven new MTM derivatives were generated bearing one, two or three acetyl groups attached at specific positions of the trisaccharide unit (sugars D and E).⁴⁹ All these compounds showed activity at micromolar or lower concentrations against a panel of four tumor cell lines, some of them exerting improved activity against selected cell lines.

Finally, a biocatalytic approach based on a lipase-catalyzed acylation enabled decoration of MTM and CRM with acyl moieties in complementary positions to those functionalized with the CmmA acetyl transferase. Thus, lipase A and B from Candida antarctica (CAL-A and CAL-B) provided 24 new mithralogs derived from MTM and the “first generation” MTM-SK and MTM-SDK.⁵⁰ Specifically, CAL-B was highly regioselective towards the 4′-hydroxyl group of the aglycone side chain in MTM, and sterically hindered or poorly reactive esters provided 4′-monoacyl derivatives in excellent yields. On the other hand, more reactive acyl donors led to mixtures of mono- and diacylated derivatives in the aglycone side chain and the disaccharide (4′ and 4B-hydroxyl group). CAL-A, meanwhile, showed regioselectivity towards the dissacharide subdomain, acylating the 3B- or 4B-hydroxyl groups exclusively. As a result, and just by changing the acylating agent, a plethora of mono- and diacylated mithralogs were obtained. Similarly, MTM-SK and MTM-SDK were also submitted to enzymatic acylation leading to novel mithralogs by combining genetic engineering and biocatalysis. These compounds were acylated by CAL-B or CAL-A in the hydroxyl groups of the 3B and 4B positions of the disaccharide only. Along these libraries of acyl derivatives some exhibited activity comparable to, or slightly greater than, the parent drugs. A remarkable example is 3B-allyloxycarbonyl-MTM (EC-8105), which turned out to be approximately eight times more potent than MTM in the suppression of the EWS/FLI1 gene signature (Figure 4.6). Recently, MTM was identified in a high throughput screen at the NCI as an inhibitor of EWS/FLI1 and translated to the clinic in a phase I-II trial. Thus, the continued expression of this transcription factor proved to be critical for the cell survival of Ewing sarcoma, a bone and soft tissue sarcoma with a poor overall survival.⁴⁰ Importantly both EC-8105 and also EC-8042 suppress EWS/FLI1 activity, the former at lower concentrations than MTM and the latter at concentrations similar to MTM that do not appear to cause liver toxicity.⁵¹

Figure 4.6 Representative mithralogs generated by EntreChem.

4.3.4 Collismycin Analogs as Immunosuppressive and Neuroprotective Drugs

Within natural products synthesized by microorganisms, the bipyridyl family of compounds has been recently studied in some detail. Representatives of this family are SF2738A-F, pyrisulfoxins, collismycins and caerulomycins.^52–55 All these compounds contain a 2,2′-bipyridyl ring system that is further modified with some post-bipyridyl tailoring modifications. Early biosynthetic studies using labeled precursors suggested that picolinic acid was a biosynthesis precursor derived from lysine. Members of this family have been shown to exhibit antibacterial, antifungal and cytotoxic activities and they also show potential as anti-inflammatory agents through the binding to the dexamethasone-glucocorticoid receptor or as potential neuroprotectant agents by reducing oxidative stress in neurons.⁵⁶ Recently, EntreChem collaborated in the isolation and characterization of the biosynthetic gene cluster of collismycin A from Streptomyces sp. CS40.⁵⁷ Then, the chemical space of this natural scaffold was expanded by the combination of three different approaches (insertional inactivation, mutasynthesis and biocatalysis).^58,59 Thus, 15 derivatives were generated through the introduction of modifications in both pyridine rings of collismycin A (Scheme 4.6). Although none of these compounds displayed enhanced cytotoxicity, two of them showed better neuroprotective activity against oxidative stress in a zebrafish model. More interestingly, the most potent derivative showed very poor cytotoxic activity which is a required feature for a neuroprotectant compound. Recently, a natural member of this family, caerulomycin, has been described as immunosuppressant, a novel kind of activity for this class of molecules.

Scheme 4.6 Collismycin analogues generated by combinatorial biosynthesis and biocatalysis.

On the other hand, during the course of a project aimed at generating novel collismycin analogues by using biocatalysts, EntreChem discovered an efficient bio-deoximation method promoted by laccases.⁶⁰ Actually, this was the first enzymatic approach for the cleavage of oximes and a green alternative to the existing methods due to its mildness and simplicity. In particular, the system that used TEMPO or ABTS (10 mol%) as mediator and aerial O₂ as oxidant enabled the deprotection at room temperature of both aromatic and aliphatic ketoximes into the corresponding ketones (Scheme 4.7) at 40 mM and lab-scale. On the other hand, although the oxidative system suffered from low yields when applied to aldoximes, the presence of an intramolecular stabilizing motif such as the bipyridyl unit in collismycin made possible the efficient conversion into the corresponding carboxylic acid, leading to novel analogues (Scheme 4.7). This new and unexpected biocatalytic activity – namely the conversion of an aldoxime into a carboxylic acid – would presumably take place in two steps: (i) oxime disproportionation into the corresponding aldehyde and hydroxylamine; (ii) oxidation of the triggered aldehyde to carboxylic acid. The first step was supported by the detection of the aldehyde by HPLC/MS. Likewise, deconvolution experiments demonstrated that both enzyme and mediator must be present for the reaction to happen. Then, the concomitant oxidation of the resulting aldehyde to acid would be easily accomplished by the oxidizing system as in literature reports.⁶¹

Scheme 4.7 Laccase-catalyzed deoximation of ketoximes and aldoximes.

4.3.5 Glycosylated Indolocarbazoles as Potent and Selective Kinase Inhibitors

The indolocarbazole alkaloids are a family of natural products that has attracted much attention because of their ability to inhibit early identified targets like protein kinase C and topoisomerase I.^62,63 Rebeccamycin (REB) and staurosporine (STA) are the lead members of this family (Figure 4.7).^64,65 The main structural features of most natural indolocarbazoles are a heterocyclic aglycone chromophore bearing alternatively an imide (as in REB) or an amide function (as in STA) and a sugar moiety attached through a single N-glycosidic bond (as in REB) or by two bonds consisting of a N-glycosidic bond and a N,O-ketal (as in STA). These structural differences could be the basis of the different target selectivity, since REB inhibits DNA topoisomerase I while STA is a protein kinase inhibitor.^66,67

Figure 4.7 Chemical structures of lead glycosylated indolocarbazoles.

The discovery of novel kinase inhibitors represents a major challenge in R&D for pharmaceutical companies nowadays. The goal is to develop selective and potent inhibitors of kinases involved in signal transduction pathways related to a variety of diseases, especially those unlocking new drug targets. Actually, the lack of selectivity of STA led to intolerable toxicity and subsequent drag for clinical development. In recent years, many indolocarbazole analogues have been developed, with some successful examples of natural and synthetic derivatives reaching clinical stage research, indicating they are safe in humans. In particular, the gene clusters of both leads were characterized and their biosynthetic pathways deciphered, making it possible to change by genetic engineering both sugar moiety and aglycone. In this regard, EntreChem’s unique technology allowed combinatorial biosynthesis of genes from the metabolic pathways of REB and STA, providing a library of hybrid indolocarbazoles, whose mechanism of action consists of potent and selective inhibition of protein kinases.⁶⁸ Particularly, 14 novel analogues were described (seven from each lead) with a variety of glycosylation patterns. All of them showed antiproliferative activity against MDA-MB-231 (breast), A549 (lung) and HT29 (colon) human tumor cell lines with GI₅₀ in the submicromolar range. When assayed with a panel of 57 kinases, compounds with the sugar linked through two bonds showed stronger inhibition than those linked by a single bond. In particular, some of them were far less promiscuous than staurosporine (>70% inhibition at 10 nM for 23 kinases) with comparable inhibition only in a few kinases. From this library, one molecule bearing a STA-type aglycone and a L-olivose sugar (EC-70124) emerged as candidate since the in vitro profile, confirmed by cellular experiments, unveiled a mode of action impacting tumors dependent on NF-kB activation, a desirable target in cancer and potentially other diseases.^69,70 This compound displays a potent multi-kinase inhibitor spectrum affecting key intracellular kinases implicated in prosurvival and proliferative pathways. Thus, EC-70124 induces senescence of glioblastoma-initiating cells by inhibition of the NF-kB pathway⁷¹ and exerts antitumoral activity in triple negative breast cancer by the inhibition of both PI3K/mTOR and JAK/STAT pathways.⁷² Similar antitumoral activity has been also described in colorectal cancer mediated by PI3K/Akt pathway inhibition.⁷³ Interestingly, EC-70124 also acts as a dual STAT3/NF-kB inhibitor, reverting both tumorigenic and stem cell properties in prostate cancer.⁷⁴ Dose Range Finding studies in rats and dogs have been carried out, and pharmacokinetic data show circulating plasma levels well above those needed for therapeutic action in vitro, emphasizing the development potential of EC-70124 for oncology applications. Currently the oral formulation is under regulatory preclinical development to generate the IND data that will allow translation of this promising drug into humans.

4.4 Business Models in Biocatalysis and Natural Products Drug Discovery

As discussed in the previous sections, EntreChem is active in two areas of biotechnology research susceptible of making an impact in the market: biocatalysis process development and drug discovery and development. However, there are significant differences in the business potential of each of these biotechnology market segments.

The main motivation behind drug discovery and development programs is to find cures to threatening diseases. This represents a common goal for actors with diverse interests such as academics (both at Universities and Hospitals), contract research organizations (CROs), small biotech and Big pharma companies. Each of these actors brings to the table complementary capabilities better applicable to certain aspects of the discovery and development process, so partnership in innovation has become customary to successfully deliver quality drugs in a cost-effective system. A drug discovery and development program is an expensive (0.8–1.8B$)⁷⁵ and time consuming effort (>12 years),⁷⁶ which, economically speaking, translates to a high cost–high reward investment exercise, with the added difficulty that no actor (with the exception of Big pharma) has the economic resources needed to cover the entire process from concept discovery to marketing. As a result, a well-defined value chain has been established covering all stages, despite the complexity of the activities required to bring a drug from the benchtop to the bedside.

As can be seen in Figure 4.8, the itemization of every stage of the process allows one to address the “high-risk” of the overall program by setting milestones with associated costs and value potential, allowing investors to partake in specific portions of the value chain, according to their investing power, and cooperate with other investors, the stock market or a corporate client to provide an exit route to their investment. In the particular case of cancer, there is a bonus since once the first application (indication) for a given drug is approved (for a specific tumor) the sponsor can tackle other indications (increasing the market sales of the drug) by moving just one or two steps upstream in the value chain, so going back to preclinical stages or even “square number one” is not necessary to grow a strong business case.

Figure 4.8 Value chain in drug discovery and development.

On the other hand, the field of biocatalysis process development seems to lack the strong motivation that genuinely brings diverse actors together for a common goal. Of course, we could consider processes to make life-saving drugs or advanced pharmaceutical intermediates, but we should realize that the manufacturing process itself is not the investment driver of whoever puts the money on the table. This makes the typical biocatalysis project a problem that must be solved on a case by case basis, and, importantly, existing technology often enables the (imperfect) supply without the need to wait for the development of a new solution. The common ground for many actors, generally expressed as “more efficient processes”, “green chemistry”, have failed to elicit a response by the market defining a “value chain” where academics, industry and CROs can cooperate in an efficient manner and bring to fruition those common goals. As compared to the field of drug discovery and development, where there are numerous cases of start-up companies, exploiting academic ideas, successfully cooperating with CROs and large client companies, very few, if any, successful examples exist (business-wise) in the field of biocatalysis. The impression is that most companies in the field just come and go, fueling a negative feedback loop in which no investor is willing to bet on this kind of activity where few, if any, have been rewarded with handsome returns.

Nevertheless, we can try to define a value chain for biocatalysis process development. Similar to the drug development value chain, there will be higher risk at earlier stages and investment will grow as the program progresses through the milestones. The overall progress could be measured by, for example, the scale the process reaches, and this will be related to technical milestones (suitability of the biocatalyst, engineering of the same, robustness for operational parameters or recyclability, supply of the final biocatalyst formulation…). A simplification of this, according to Figure 4.9, divides the optimization stages into “reactions” early on, to move later into “processes” as the project matures. One can imagine a high attrition rate (as in the drug discovery industry) providing a funnel shape defined by those few reactions worth scaling, and once the value inflexion point is reached (which separates the academic exercise and the industrial application), a reasonable number of projects may come to a successful conclusion. At this point, one must realize that although biocatalysts offer high activity and selectivity, there are still drawbacks for its wide adoption by the industry: lack of designability; stability limits; and the insufficient number of well-characterized, ready-to-use biocatalysts are still considerable barriers, although steady progress is being made to solve some of these issues.⁷⁷

Figure 4.9 Potential value chain in biocatalysis process development.

As compared to the drug development in oncology, once the market is reached with a process, repeated success will require one to go many squares back in the game to find potentially scalable biocatalytic reactions (applicable to a new substrate) and start the optimization cycle again.

A problem with this scenario is how to establish a program valuation. There are similitudes with the drug development value chain, including the participation of several actors for a successful project: an academic or company experts in enzyme engineering, another in enzyme immobilization, high throughput screening, etc. However, the ideal situation is when most of these activities take place under the same roof, or in its absence, with a minimum number of partners, including the sponsor of the project.

This, practically, means that the value creation of the overall chain does not support many players, unlike typical drug discovery projects. The main reason for this is that value creation in process development is limited, as compared to that generated from addressing a medical need. Another important reason is the nature of the sponsor: in a drug development project the sponsor can be a small company that owns the substance patents; however, in a process development project the owner is normally an industrial sponsor, leaving little upside for innovators. Such an industrial sponsor almost invariably has a first-generation process in place, so the innovator role is limited to helping to develop (often with limited or no ownership rights) a robust cost-effective process. The optimization parameters (key to understanding value creation) vary on a case by case basis, and are difficult to benchmark by an outsider (innovators rarely get a full picture of what they are up against), while in drug development plenty of benchmarks exist based on the particular market segment that the drug addresses.

To make matters worse, the exit route of a successful project necessarily goes through beating handily the performance of the sponsor in-house process. If we examine the trajectory of drug development projects, a successful investment exit does not mean reaching the market after passing all technical milestones. From 1950 to 2009 the pharmaceutical industry had seen 4300 players, and only 261 (6.0%) had at least one approved drug.⁵⁰ The market assumes that a small innovative company may have the resources to reach, say, Phase II clinical trials and then the program is sold to a bigger partner that takes care of the rest of the now seemingly de-risked project. This means validation and significant income for the small innovator, even if the drug fails later trials and is never approved. It is the expectations of addressing a medical need that drives this kind of market.

As shown in Figure 4.10, the drug program can be perceived as a “space capsule”, effectively constituting a package of data and patent rights, that will be “abducted” by an “interstellar ship” when the time is right. Market conditions vary depending on external circumstances and this fact impacts the milestone at which the deal is more likely to happen; however, the trajectory, at least on paper, represents a clear pathway to success.

Figure 4.10 Exit route in drug development process may miss the final goal.

However, in the process development scenario shown in Figure 4.11, the “space capsule” of the small innovator, even if it contains all the elements required to make a positive impact, misses the most important driver of value creation: ownership of the molecule. This makes establishing a project trajectory difficult, and the expectations driver is unlikely to trigger an “abduction” before clearly superior data are on the table.

Figure 4.11 Project trajectory is difficult to define in biocatalytic process development.

Based on our limited ten-year experience in the biotech business, it seems unlikely that a business case in biocatalysis would emulate the success seen in more complex, but also more developed, business models. The similarities between the value chains of drug and bioprocess development are outweighed by the differences, and the relationship between innovation providers and industrial customers seems to require close cooperation from the beginning of the project, resembling the relationship model of CROs, but bearing the burden of innovation and left with a limited upside potential. To avoid this scenario, perhaps the biocatalysis sector would improve its sustainability if clear rules for a robust risk-shared partnership model are implemented to suit the needs of all actors involved.

References

1. J. P. Vilaine, Pharmacol. Res., 2006, 53, 424.

2. M. Pascual Izco, G. L. Alonso Salinas, M. Sanmartin Fernández, H. Del Castillo Carnevalli, M. Jiménez Mena, A. Camino López and J. L. Zamorano Gómez, Cardiology, 2016, 134, 372.

3. F. J. Hidalgo, M. Anguita, J. C. Castillo, S. Rodríguez, L. Pardo, E. Durán, J. J. Sánchez, C. Ferreiro, M. Pan, D. Mesa, M. Delgado and M. Ruiz, Int. J. Cardiol., 2016, 217, 7.

4. S. Pedragosa-Moreau, A. Le Flohic, V. Thienpondt, F. Lefoulon, A. M. Petit, N. Ríos-Lombardía, F. Morís and J. González-Sabín, Adv. Synth. Catal., 2017, 359, 485.

5. E. Ricca, B. Brucher and J. H. Schrittwieser, Adv. Synth. Catal., 2011, 353, 2239.

6. R. C. Simon, N. Richter, E. Busto and W. Kroutil, ACS Catal., 2014, 4, 129.

7. V. Kohler, Y. M. Wilson, M. Durrenberger, D. Ghislieri, E. Churakova, T. Quinto, L. Knorr, D. Häussinger, F. Hollmann, N. J. Turner and T. R. Ward, Nat. Chem., 2013, 5, 93.

8. C. A. Denard, J. F. Hartwig and H. Zhao, ACS Catal., 2013, 3, 2856.

9. J. H. Schrittwieser, F. Coccia, S. Kara, B. Grischek, W. Kroutil, N. d’Alessandro and F. Hollmann, Green Chem., 2013, 15, 3318.

10. H. Yamada and S. Shimizu, Angew. Chem., Int. Ed. Engl., 1988, 27, 622.

11. E. Liardo, N. Ríos-Lombardía, F. Morís, J. González-Sabín and F. Rebolledo, Org. Lett., 2016, 18, 3366.

12. G. A. Applegate and D. B. Berkowitz, Adv. Synth. Catal., 2015, 357, 1619.

13. R. rhodochrous IFO-15564, formerly known as Rhodococcus butanica ATCC 21197. See H. Kakeya, N. Sakai, T. Sugai and H. Ohta, Tetrahedron Lett., 1991, 32, 1343, , KREDs from the Codex KRED Screening Kit (Codexis) .

14. H. Gröger and W. Hummel, Curr. Opin. Chem. Biol., 2014, 19, 171.

15. E. Burda, W. Hummel and H. Gröger, Angew. Chem., Int. Ed., 2008, 47, 9551.

16. C. A. Denard, H. Huang, M. J. Bartlett, L. Lu, Y. Tan, H. Zhao and J. F. Hartwig, Angew. Chem., Int. Ed., 2014, 53, 465.

17. H. Sato, W. Hummel and H. Gröger, Angew. Chem., Int. Ed., 2015, 54, 4488.

18. N. Ríos-Lombardía, C. Vidal, M. Cocina, F. Morís, J. García-Álvarez and J. González-Sabín, Chem. Commun., 2015, 51, 10937.

19. N. Ríos-Lombardía, C. Vidal, E. Liardo, F. Morís, J. García-Álvarez and J. González-Sabín, Angew. Chem., Int. Ed., 2016, 55, 8691.

20. G. M. Cragg and D. J. Newman, Biochim. Biophys. Acta, 2013, 1830, 3670.

21. L. Katz and R. H. Baltz, J. Ind. Microbiol. Biotechnol., 2016, 43, 155.

22. D. J. Newman and G. M. Cragg, J. Nat. Prod., 2016, 79, 629.

23. K. J. Weissman and P. F. Leadlay, Nat. Rev. Microbiol., 2005, 3, 925.

24. R. A. Bauer, J. M. Wurst and D. S. Tan, Curr. Opin. Chem. Biol., 2010, 14, 308.

25. D. C. Swinney and J. Anthony, Nat. Rev. Drug Discovery, 2011, 10, 507.

26. K. E. Davis, S. J. Joseph and P. H. Janssen, Appl. Environ. Microbiol., 2005, 71, 826.

27. J. González-Sabín, Biochem. Pharmacol., 2012, 1, e1112.

28. To learn more about EntreChem co-founders contribution to the field, see: C. Olano, J. Antibiot. (Tokyo), 2011, 64, 51.

29. D. H. Altreuter and D. S. Clark, Curr. Opin. Biotechnol., 1999, 10, 130.

30. S. Riva, Curr. Opin. Chem. Biol., 2001, 5, 106.

31. J. González-Sabín, R. Morán-Ramallal and F. Rebolledo, Chem. Soc. Rev., 2011, 40, 5321.

32. G. M. Cragg, P. G. Grothaus and D. J. Newman, Chem. Rev., 2009, 109, 3012.

33. F. Lombó, N. Menéndez, J. A. Salas and C. Méndez, Appl. Microbiol. Biotechnol., 2006, 73, 1.

34. S. Mansilla, I. García-Ferrer, C. Méndez, J. A. Salas and J. Portugal, Biochem. Pharmacol., 2010, 79, 1418.

35. C. Vizcaíno, S. Mansilla and J. Portugal, Pharmacol. Ther., 2015, 152, 111.

36. J. H. Brown and B. J. Kennedy, N. Engl. J. Med., 1965, 272, 111.

37. W. G. Ryan, T. B. Schwartz and G. Northrop, J. Am. Med. Assoc., 1970, 213, 1153.

38. B. J. Kennedy, Am. J. Med., 1970, 49, 494.

39. H. L. Davis Jr, D. D. von Hoff, J. E. Henney and M. Rozencweig, Cancer Chemother. Pharmacol., 1978, 1, 83.

40. P. J. Grohar, G. M. Woldemichael, L. B. Griffin, A. Mendoza, Q. R. Chen, C. Yeung, D. G. Currier, S. Davis, C. Khanna, J. Khan, J. B. McMahon and L. J. Helman, J. Natl. Cancer Inst., 2011, 103, 962.

41. M. Zhang, A. Mathur, Y. Zhang, S. Xi, S. Atay, J. A. Hong, N. Datrice, T. Upham, C. D. Kemp, R. T. Ripley, G. Wiegand, I. Avital, P. Fetsch, H. Mani, D. Zlott, R. Robey, S. E. Bates, X. Li, M. Rao and D. S. Schrump, Cancer Res., 2012, 72, 4178.

42. P. D. Sutphin, D. A. Chan, J. M. Li, S. Turcotte, A. J. Krieg and A. J. Giaccia, Cancer Res., 2007, 67, 5896.

43. C. Méndez, J. González-Sabín, F. Morís and J. A. Salas, Planta Med., 2015, 81, 1326.

44. L. L. Remsing, A. M. González, M. Nur-e-Alam, M. J. Fernández-Lozano, A. F. Braña, U. Rix, M. A. Oliveira, C. Méndez, J. A. Salas and J. Rohr, J. Am. Chem. Soc., 2003, 125, 5745.

45. I. Baig, M. Pérez, A. F. Braña, R. Gomathinayagam, C. Damodaran, J. A. Salas, C. Méndez and J. Rohr, J. Nat. Prod., 2008, 71, 199.

46. L. E. Núñez, S. E. Nybo, J. González-Sabín, M. Pérez, N. Menéndez, A. F. Braña, K. A. Shaaban, M. He, F. Morís, J. A. Salas, J. Rohr and C. Méndez, J. Med. Chem., 2012, 55, 5813.

47. A. Pandiella, F. Morís, A. Ocaña, L. E. Núñez and J. C. Montero, Oncotarget, 2015, 6, 32856.

48. J. Tornín, L. Martínez-Cruzado, L. Santos, A. Rodríguez, L. E. Núñez, P. Oro, M. A. Hermosilla, E. Allonca, M. T. Fernández-García, A. Astudillo, C. Suárez, F. Morís and R. Rodríguez, Oncotarget, 2016, 7, 30935.

49. B. García, J. González-Sabín, N. Menéndez, A. F. Braña, L. E. Núñez, F. Morís, J. A. Salas and C. Méndez, Microb. Biotechnol., 2011, 4, 226.

50. J. González-Sabín, L. E. Núñez, A. F. Braña, C. Méndez, J. A. Salas, V. Gotor and F. Morís, Adv. Synth. Catal., 2012, 354, 1500.

51. C. L. Osgood, N. Maloney, C. G. Kidd, S. Kitchen-Goosen, L. Segars, M. Gebregiorgis, G. M. Woldemichael, M. He, S. Sankar, S. L. Lessnick, M. Kang, M. Smith, L. Turner, Z. B. Madaj, M. E. Winn, L. E. Nuñez, J. González-Sabín, L. J. Helman, F. Morís and P. J. Grohar, Clin. Cancer Res., 2016, 22, 4105.

52. S. Gomi, S. Amano, E. Sato, S. Miyadoh and Y. Kodama, J. Antibiot. (Tokyo), 1994, 47, 1385.

53. N. Tsuge, K. Furihata, K. Shin-Ya, Y. Hayakawa and H. Seto, J. Antibiot. (Tokyo), 1999, 52, 505.

54. A. McInnes, D. Smith, J. Walter, J. Wright, L. Vining and G. Arsenault, Can. J. Chem., 1978, 56, 1836.

55. K. Shindo, Y. Yamagishi, Y. Okada and H. Kawai, J. Antibiot. (Tokyo), 1994, 47, 1072.

56. A. Martínez, P. Usán, M. Medina, E. García, J. Pérez, R. I. Fernández, A. Medarde, L. M. Canedo, F. Romero, M. Alonso and J. Sánchez, WO2007017146 A3, 2007.

57. I. García, N. M. Vior, A. F. Braña, J. González-Sabín, J. Rohr, F. Morís, C. Méndez and J. A. Salas, Chem. Biol., 2012, 19, 399.

58. I. García, N. M. Vior, J. González-Sabín, A. F. Braña, J. Rohr, F. Morís, C. Méndez and J. A. Salas, Chem. Biol., 2013, 20, 1022–1032.

59. C. Sialer, I. García, J. González-Sabín, A. F. Braña, C. Méndez, F. Morís and J. A. Salas, Bioorg. Med. Chem. Lett., 2013, 23, 5707.

60. J. González-Sabín, N. Ríos-Lombardía, I. García, N. M. Vior, A. F. Braña, C. Méndez, J. A. Salas and F. Morís, Green Chem., 2016, 18, 989.

61. M. Mogharabi and M. A. Faramarzi, Adv. Synth. Catal., 2014, 356, 897.

62. G. W. Gribble and S. J. Berthel, Studies in Natural Product Chemistry, Elsevier, Amsterdam, 1993, vol. 12, p. 365.

63. C. Sánchez, C. Méndez and J. A. Salas, Nat. Prod. Rep., 2006, 23, 1007.

64. D. E. Nettleton, T. W. Doyle, B. Krishnan, G. K. Matsumoto and J. Clardy, Tetrahedron Lett., 1985, 26, 4011.

65. S. Omura, Y. Iwai, A. Hirano, A. Nakagawa, J. Awaya, H. Tsuchya, Y. Takahashi and R. Masuma, J. Antibiot. (Tokyo), 1977, 30, 275.

66. E. R. Pereira, L. Belin, M. Sancelme, M. Prudhomme, M. Ollier, M. Rapp, D. Severe, J. F. Riou, D. Fabbro and T. Meyer, J. Med. Chem., 1996, 39, 4471.

67. T. Tamaoki, H. Nomoto, I. Takahashi, Y. Kato, M. Morimoto and F. Tomita, Biochem. Biophys. Res. Commun., 1986, 135, 397.

68. C. Sánchez, A. P. Salas, A. F. Braña, M. Palomino, A. Pineda-Lucena, R. J. Carbajo, C. Méndez, F. Morís and J. A. Salas, Chem. Commun., 2009, 4118.

69. L. Nogueira, P. Ruiz-Ontañón, A. Vázquez-Barquero, F. Morís and J. L. Fernández-Luna, Oncotarget, 2011, 2, 646.

70. D. Verzella, M. Fischietti, D. Capece, D. Vecchiotti, F. Del Vecchio, G. Cicciarelli, V. Mastroiaco, A. Tessitore, E. Alesse and F. Zazzeroni, Curr. Drug Targets, 2016, 17, 311.

71. L. Nogueira, P. Ruiz-Ontañón, A. Vázquez-Barquero, M. Lafarga, M. T. Berciano, B. Aldaz, L. Grande, I. Casafont, V. Segura, E. F. Robles, D. Suárez, L. F. García, J. A. Martínez-Climent and J. L. Fernández-Luna, Oncogene, 2011, 30, 3537.

72. M. D. Cuenca-López, G. Serrano-Heras, J. C. Montero, V. Corrales-Sánchez, M. Gómez-Juárez, M. J. Gascón-Escribano, J. C. Morales, V. Voisin, L. E. Núñez, F. Morís, G. D. Bader, A. Pandiella and A. Ocaña, Oncotarget, 2015, 6, 27923.

73. G. Serrano-Heras, M. D. Cuenca-López, J. C. Montero, V. Corrales-Sánchez, J. C. Morales, L. E. Núñez, F. Morís, A. Pandiella and A. Ocaña, Oncotarget, 2015, 6, 31272.

74. G. Civenni, N. Longoni, P. Costales, C. Dallavalle, C. García Inclán, D. Albino, L. E. Núñez, F. Morís, G. M. Carbone and C. V. Catapano, Mol. Cancer Ther., 2016, 15, 806.

75. B. Munos, Nat. Rev. Drug Discovery, 2009, 8, 959.

76. I. Khanna, Drug Discovery Today, 2012, 17, 1088.

77. A. S. Bommarius, Annu. Rev. Chem. Biomol. Eng., 2015, 6, 319.

4.1 Introduction