**5. Impact of Cross-Coupling Reactions on New Chemical Modalities**

In the past few years, there has been an increasing interest in the medicinal chemistry community to explore additional chemical modalities beyond the traditional small molecules. These new chemical modalities [51] offer new opportunities to modulate targets and in particular, those challenging targets previously considered undruggable. As forward thinking, we want to offer an overview of the applications of cross-coupling reactions to these new chemical modalities.

#### *5.1. Application of Cross-Coupling Reactions to DNA-Encoded Libraries (DELs)*

Novel screening methodology using DNA-encoded libraries (DELs) [52] has been a focus of attention for multiple pharmaceutical companies as a way to enhance their high-throughput capabilities in early Lead Generation [53]. DNA-encoded synthesis allows a much broader evaluation of the traditional chemical space (~5-fold orders of magnitude) [54] coupled to miniaturization of high-affinity assays to test the different molecules. To design and synthesize DELs, reactions are required to be mainly feasible in aqueous conditions, and mild enough to preserve the DNA. Other considerations are assembly of building blocks, oligonucleotide conjugation, polymerase chain reaction (PCR) sequencing, and analysis of large data sets.

During the design process of a DEL library, the researcher needs to keep in consideration the types of reactions, specifically, they must be high yielding, broad scope, and primarily maintain the integrity of the DNA code. Many types of reactions have been able to adapt to mild and aqueous conditions required for DEL synthesis. For cross-coupling reactions, the Suzuki–Miyaura C–C coupling has been demonstrated in an example from the GlaxoSmithKline (GSK) group [55] towards hit identification of phosphoinositide 3-kinase α (PI3Kα) ligands, preparing a three-cycle library of over 3.5 million of diverse compounds. More recently, Torrado and colleagues [56] developed a methodology to tackle, for the first time, the C–N cross-coupling on DNA. Through a series of parallel screening conditions, the group identified mild and efficient reaction conditions for the *n*-arylation of anilines on-DNA aryl bromides. The strategy focused on identifying the palladium catalyst in multi-array aqueous conditions, using *t*BuOK as the base and temperatures between 50 and 80 ◦C, using first as a pilot aryl-iodides on DNA. After investigation of more than 12 different sources of palladium, the reaction only provided the desired product when *t*-butyl-XPhos Pd precatalyst G3 was used. Overall, a combination of *t*-butyl-XPhos Pd precatalyst G3 (15 eq.) using NaOH (300 eq.) as the base proved to be the most effective conditions for the generation of the C−N bond between conjugated DNA aryl iodides and aromatic amines at 30 ◦C. Based on these results, the group explored reaction conditions for the use of aryl bromides on DNA (**73**, Figure 12). Fortunately, minor adjustments in the reaction protocol, like increasing the number of equivalents of the aniline (from 80 eq. to 150 eq.) and the temperature to 60 ◦C, led to comparable catalytic abilities for DNA-conjugated aryl bromides, less reactive electrophiles, but more versatile building blocks. In Figure 12, it is shown further application of this methodology to heteroaryl cross-coupling reactions expanding the scope of the reaction for medicinal chemistry applications during hit identification. After applying this protocol to >850 structurally diverse (hetero)aromatic anilines and >450 aryl bromides conjugated to DNA, the authors reported the successful application of this new method to the production of a DEL during cycle 3.

**Figure 12.** Examples of C–N cross-coupling reaction in DNA using various heteroaromatic amines. (Conversion was calculated by LCMS signal integration, not isolated yield.).

#### *5.2. Application of Cross-Coupling Reactions to Macrocycle and Cyclopeptides*

Macrocycle and cyclopeptides are another chemical modality that it is currently attracting vast attention in medicinal chemistry [57] to modulate targets with larger binding pockets and provide novel mechanism of action. Strategies targeting traditional small molecules applying "rule-of-five" (Ro5) guidelines have been recognized to be unsuccessful against difficult targets, like protein–protein interactions, nucleic acid complexes, or antibacterial modalities. However, natural products have demonstrated to be effective at modulating such targets, directing a renewed interest on investigating underrepresented chemical scaffolds associated with natural products. Naturally derived cyclopeptides [58] offer advantages in affinity, selectivity, stability and druggability in comparison to linear peptides.

Using natural products as starting points, medicinal chemists have modified the core structure of the cyclopeptide to increase pharmacological activity and drug-like properties. There are examples of introducing biaryl systems within the cyclopeptide core like largazole [59] derivatives resulting in higher selectivity and potency. Recently Sewald [60] and colleagues developed a new methodology applying Suzuki−Miyaura cross-coupling reactions (in solution or on-resin) to introduce a biaryl moiety into an Arg-Gly-Asp (also known as RGD) cyclopeptide providing new SAR direction. Cilengitide (**75**, Figure 13) is a cyclic Arg-Gly-Asp peptide (displayed in blue) and potent αvβ3 and αvβ5 integrin inhibitor that received significant attention for its advanced clinical studies for glioblastoma. The critical role of integrins as intermediaries in a broad range of cancer cell activities resulted in multiple drug discovery efforts of this target for oncology.

**Figure 13.** Structure of cyclopeptide cilengitide (**75**), RGD displayed in blue.

Sewald and his group used the Suzuki-Miyaura reaction as an approach towards side-to-tail cyclization incorporating the biaryl moiety. They incorporated bromo-tryptophan isomers within the peptide chain to enable an intramolecular palladium-mediated Suzuki-Miyaura cyclization reaction with a boronic acid in the other extreme of the peptide. The versatility of the Suzuki-Miyaura reaction allowed for the synthesis of multiple cyclopeptides and SAR evaluation. The initial conditions in solution were Na2PdCl4 with water-soluble sSPhos (sodium 2 -dicyclohexylphosphino-2,6-dimethoxy-1,1 -biphenyl-3-sulfonate hydrate). However, high dilution conditions (0.2 mM) to avoid intermolecular cross-coupling did not afford the desired cyclic peptide. Increasing the concentration of the cross-coupling reaction to 2.0 mM provided higher conversion without the dimerization by-product. Further optimization included increasing the reaction temperature to 100 ◦C. The cross-coupling reaction on-resin was adapted using Pd2(dba)3, potassium fluoride, and a solvent mixture of 1,2-dimethoxyethane (DME), ethanol (EtOH) and water (H2O) (Scheme 8). Using this approach more than 20 cyclopeptides were synthesized and evaluated in vitro. Interestingly, the connectivity between the aromatic group and the indole influenced the conformation of the cyclopeptide, the affinity and selectivity. As a result, cyclopeptide **77** was identified as a low nanomolar (5.4 nM) αvβ3 integrin with a connection 3 to 7– aromatic-indole respectively and high human plasma stability (t1/<sup>2</sup> > 24 h).

**Scheme 8.** Suzuki-Miyaura Cross-Coupling for the synthesis of biaryl cyclopeptide **77**.

#### *5.3. Application of Cross-Coupling Reactions to Allosteric Modulators*

Another interesting chemical modality that is currently receiving much attention is the design and synthesis of allosteric modulators [61–63]. This modality offers significant advantages over the development of orthosteric agonists, enabling much more fine-tuned modulation of the receptor only in the presence of the endogenous ligand. A successful allosteric modulator should be able to bind to a "remote" or secondary active site of the receptor and might produce changes in the conformation of the primary active site (or orthosteric site). Overall, this chemical modality provides ligands that offer suited compounds less prone to produce desensitization (like orthosteric agonists) and favorable toxicological profiles as allosteric modulators have a "ceiling" on the magnitude of their allosteric effect.

One recent example is the discovery of AG−120 [64] (ivosidenib) (**80**) (Figure 14), a highly specific, allosteric, reversible inhibitor of the isocitrate dehydrogenase−1 (IDH1) mutant enzyme (a mutation present in several types of cancer). The identification of IDH mutations among numerous cancer types has transformed the knowledge of oncogenesis and the opportunity for targeted therapeutics focus on small molecule inhibitors. Ivosidenib (**80**) was approved in the United States in 2018 for the treatment of relapsed or refractory acute myeloid leukemia (AML) for patients with the IDH1 mutation, followed by a 2019 FDA-approval for patients susceptible to the IDH1 mutation and upon first diagnosis.

**Figure 14.** Evolution of the key breakthrough compounds for IDH1 inhibitors.

An initial prototype mutant IDH1 inhibitor AGI−5198 (**78**, Figure 14) was identified from a high-throughput screen. Although AGI−5198 (**78**) showed a strong tumor inhibition in preclinical models, the poor pharmacokinetic profile (high clearance) prevented its further advancement into clinical studies. Medicinal chemistry approaches toward decreasing clearance rates by blocking metabolism were undertaken by incorporating fluorinated cycloalkyl groups (green box, Figure 14) and replacing the imidazole in the right portion of the molecule (orange box), which led to the discovery of AGI−14100 (**79**, Figure 14). AGI−14100 (**79**) exhibited high in vitro potency for IDH1 and suitable metabolic stability; however, further evaluation for the human pregnane X receptor (hPXR) indicated that the compound was a strong inducer of the cytochrome P450 (CYP) 3A4. Introduction of additional polarity in the central core of the molecule to minimize CYP induction, led to the replacement of the 3,5-di-fluoro-phenyl by a 5-fluoro-pyridine (purple box). The resulting compound, AG−120 (**80**), had the appropriate combination of in vitro potency and pharmacokinetic profile with reduced hPXR activation.

The synthesis of AG−120 (**80**) is shown in Scheme 9. After preparing the intermediate **83** using standard literature procedures, the Agios group used a Ugi four-component reaction to prepare the key precursor **84** as a racemate in a 46% yield. Palladium-mediated cross-coupling reaction using Pd2(dba)3, Xantphos as the catalyst, and Cs2CO3 as the base with 2-bromo-isonicotinonitrile led to a diastereomeric mixture **85a**/**b**. Furthermore, after crystallization, the resulting mixture was submitted for chiral resolution giving the enantiomerically pure compound AG−120 (**80**). As the authors noted, the synthetic route was robust enough to be scalable to >100 g to enable the preclinical pharmacology and toxicology studies.

**Scheme 9.** Synthesis of AG-120 (**80**).

#### *5.4. Application of Cross-Coupling Reactions in Proteolysis Targeting Chimeras (PROTACs)*

Targeted protein degradation, in particular PROTACs, has attracted the attention of many academic institutions and pharmaceutical companies as a novel therapeutic approach in drug discovery [65,66]. This emerging technology has the ability to target the "undruggable" proteome, a limitation of traditional drugs [67,68]. These molecules are designed to enable ubiquitination of the target protein of interest, ultimately flagging it for degradation by the proteasome. PROTAC molecules possess a heterobifunctional structure that contains three chemical elements: a ligand that binds the desired target protein, a moiety that binds a specific E3 ligase, and a linker between the two. There are advantages of this strategy for drug development as maintaining a high level of drug dosing is not required as these degraders are catalytic in their mode of action. PROTACs have become a new chemical modality [51] with some companies like Arvinas being pioneers [69] and many other pharmaceutical companies focusing on this approach. As such, we have witnessed a revolution in this field with multiple publications across therapeutic areas. For the synthesis of PROTACs, we are observing different methodologies to synthesize the ligand that binds to the receptor and publications are emerging incorporating the application of cross-coupling reactions [70–72]. Among them, we have selected a specific example below where the use of Pd-mediated cross-coupling reaction connecting the ligand to the PROTAC linker is essential.

GSK focused their attention on targeting Interleukin−1 receptor-associated kinase 4 (IRAK4) degradation utilizing the small molecule PROTAC approach [73]. The effort aimed to target both IRAK4 s kinase and scaffolding protein functionality, currently unachievable by small molecule inhibitors. This strategy has the potential to have a greater therapeutic benefit and could lead to new therapeutic opportunities in the treatment of autoimmune, inflammatory, and cancer diseases.

Efforts began by preparing IRAK4 degrader compounds utilizing a modified analog (not displayed) of the known IRAK4 kinase inhibitor PF−06650833 (**89**) (Figure 15) [74], in combination with either the von Hippel–Lindau (VHL), Cereblon (CRBN), or Inhibitor of Apoptosis (IAP) E3 ligase ligand

attached via a linker moiety designed *in silico*. Rapid production and subsequent evaluation of these analogs' degrader capabilities were facilitated by the development of a synthetic strategy, utilizing cross-coupling capabilities that productively enabled variations of the E3 ligase ligand and linker. This approach employed a Sonogashira coupling, attaching the IRAK4 ligand analog to the desired E3 ligase ligand and linker moiety (general reaction conditions displayed in Scheme 10). The conditions for the coupling reaction utilized Pd2(dba)3 as the palladium source in the presence of XPhos, potassium carbonate as the base in *N*,*N*-dimethylacetamide (DMA). Although variations in catalyst, reagent loadings, and reaction times were required, this methodology proved to be a robust route for the rapid production of analogs.

**Figure 15.** Evolution of small molecule IRAK4 ligand, PF−06650833 (**89**), to PROTAC degrader molecules **88** and **90**.

**Scheme 10.** Synthesis of IRAK4 PROTACs linking E3 ligase ligand and IRAK4 ligand.

After several iterations of IRAK4 PROTAC design, the fully functionalized, 4-bromo derivative of PF−06650833 (**86**) was installed onto the identified IRAK4 PROTAC VHL linked warhead (**87**), utilizing the established coupling conditions previously detailed, resulting in the discovery of compound **88** (Scheme 10/Figure 15). Compound **88** was found to display potent degradation with a half-maximal degradation concentration (DC50) in peripheral blood mononuclear cells (PBMC) cells of 259 nM. Further optimization efforts focused around modifying the polarity and flexibility of the linker moiety. This exploration was expedited by the previous screening cross-coupling reaction conditions and led to a more rigid, polar spirocyclic pyrimidine linker, compound **90** (Figure 15), which displayed a further improvement in potency (DC50 of 151 nM in PBMC cells). Additional studies revealed that the degradation took place in a proteasome dependent manner.

Further studies of the compound elucidated that not all IRAK4 functions were inhibited despite IRAK4 degradation, highlighting the need for a more thorough understanding of the target itself. Regardless, these compounds achieved potent intracellular degradation of IRAK4 as proof of concept. This strategy showed promise in the identification of novel therapeutic indications.

#### **6. Conclusions**

In this overview of the applications of cross-coupling reactions to medicinal chemistry and drug discovery, we have presented a selection of recent examples from a diverse set of therapeutic indications including migraine, oncology, pain, antiviral, autoimmune diseases, and cystic fibrosis. Cross-coupling reactions are used nowadays as a common and reliable approach to build SAR assessments. Furthermore, the application of cross-coupling methodology has become more frequent for the large scale-up synthesis in the development phase, as illustrated in this manuscript. Current advances are directed to develop "greener" conditions for process cross-coupling transformations. New chemical modalities are an emerging area of interest in drug discovery, where researchers are looking for novel chemical space to modulate challenging targets. We thought it would be important to showcase how cross-coupling methodology can be instrumental in identifying novel starting points for Lead Generation campaigns through DELs, macrocycles and cyclopeptides, allosteric approaches, as well as PROTACs efforts. At the core of this attention, lies the flexibility and reproducibility of the reaction conditions, to enable a vast set of transformations. We are envisioning a bright future for further development in this field with newer and milder cross-coupling conditions to enable reactions even in physiological relevant environments.

**Author Contributions:** The manuscript was written through contributions of all authors. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Conflicts of Interest:** The authors declare no conflicts of interest.

#### **Abbreviations**

