Biomolecular Negishi cross-couplings in the synthesis of amino acids

The Negishi cross-coupling is a powerful C – C bond-forming reaction widely utilised in many areas of organic synthesis. This review details the use of Negishi cross-couplings in the synthesis of unnatural amino acids. The application of this reaction in the preparation of aromatic, heteroaromatic, and, complex amino acid derivatives are reviewed and presented herein.


Introduction
Negishi cross-coupling, since its original development in 1977 by Negishi and co-workers, has become a widely-utilised reaction in the field of organic chemistry. 1 Employing either palladium or nickel as a transition metal catalyst, the Negishi crosscoupling will form a new carbon-carbon bond between an organic halide/triflate and an organozinc reagent. 2,3 As a versatile route to form carbon-carbon bonds the Negishi crosscoupling has seen widespread use in the synthesis of many novel structural motifs. [4][5][6] In recent years, developments in Negishi cross-couplings by researchers, such as Fu, has allowed the transformation to be applied to more substrates than ever before. [7][8][9][10] One application of the Negishi cross-coupling reaction is in the synthesis of unnatural amino acids. The formation of such compounds is of interest to a multitude of chemical and biological research arenas. Introduction of unnatural amino acids into peptide sequences has been used to probe protein structure and function in both in vitro 11,12 and in vivo systems. 13,14 Antibiotic peptides have utilized unnatural amino acids to increase organism selectivity, potency and metabolic stability. 15 In recent years asymmetric catalysis has employed unnatural amino acids as easily accessible chiral building blocks for the synthesis of new ligands. [16][17][18][19][20] For example, chiral ligands such as PyBOX and BOX rely on the chirality afforded by natural or unnatural amino acids in their preparation and design. [21][22][23][24] William D. G. Brittain With wide-ranging applications, methods to form unnatural amino acids are without doubt important. This review gives an overview and update 25 of the synthesis and derivatisation of amino acids through Negishi cross-couplings. Three major classes of derivatisation are discussed: aromatic-, heteroaromatic and more complex unnatural amino acids. Several mechanistic studies have been carried out on the Negishi cross-coupling of alkylamido zinc reagents, and these are also discussed.

Synthesis of aromatic amino acids
Aromatic amino acids have been shown to have a broad range of utility. Aromatic amino acids have been successfully employed in natural product synthesis, 26 peptide chemistry 27,28 and drug development. 29 Therefore, facile synthesis of aromatic amino acids is important.
The area of Negishi cross-coupling to form stereo-defined aromatic amino acids was pioneered in the seminal work by . 30 A range of novel single-enantiomer substituted phenylalanines 3a-d were reported from the crosscoupling between an iodozinc L-serine derivative 2 and a range of para-substituted iodobenzenes (Scheme 1). The transformation was found to be tolerant to Boc protecting groups and a range of electron-donating (Scheme 1, 3c) and withdrawing groups (Scheme 1, 3d) on the aryl iodide component. The novel phenylalanine derivatives were isolated in yields of between 35% and 65%.
Jackson and Dexter would adapt their initial Negishi crosscoupling methods to also produce β and γ amino acids from aspartic and glutamic acid derivatives. 31 Utilising activated zinc and Pd 2 (dba) 3 , a range of β and γ amino acids were furnished in yields between 33% and 89% (Scheme 2). The scope and reproducibility using this methodology was found to be superior to the conditions initially used to access α amino acids. 30 Modification of serine and phenylalanine via Negishi crosscoupling was reported by König and co-workers (Scheme 3). 32 Using Rieke zinc they were able to successfully form organozinc species from amino acids, which could then be coupled with electron-rich iodoanilines and iodobenzyl amines. Using this approach, a series of metal-chelating side chain containing unnatural amino acids 9a-e were successfully synthesised in yields ranging between 65% and 86% (Scheme 3).
The synthesis of β 2 -homophenylalanine derivatives using Negishi cross-couplings were first reported in 2008 (Scheme 4). 33 The synthesis of the corresponding β 3 -amino acids had already been reported in 1999, 34,35 but the methodology had not been used to access β 2 -amino acids. Conjugate addition of benzylamine to compound 10, followed by Boc protection and alkyl iodide formation, led to the protected amino acid coupling partner 13 (Scheme 4). Cross-coupling was then achieved using a range of aryl iodides and bromides. In the case of nitro-substituted aryls lower yields were observed. This was attributed to a competing zinc catalysed reduction of the nitro groups. It was found that removal of excess zinc before Negishi cross-coupling gave greatly improved yields, with an increase from 24% to 45% in the case of 1-iodo-4-nitrobenzene (Scheme 4, 14d).
In order to make the transformation as synthetically relevant as possible widening the scope of the haloaryl component of the Negishi cross-coupling was elaborated upon in 2008. 36 Taking inspiration from the reported reaction of a serine derivative 1 and a bromo-substituted naphthalene by Szczepankiewicz et al. (Scheme 5) 37 it was subsequently demonstrated that a wide range of bromides were applicable to the Negishi cross-coupling (Scheme 6).
Yields for the cross-coupling reaction when an aryl bromide was employed were lower across all examples compared to their iodo equivalents. 36 However, it should be noted that aryl bromides are often cheaper and more widely available than their corresponding iodides.
By exploiting the differences in cross-coupling reactivity between aromatic iodides and aromatic bromides it was demonstrated that asymmetrical bis-amino acids could be successfully synthesised. For example, the bis-amino acid 17 (Scheme 7) was synthesised through a double cross-coupling process in a 21% overall yield. This was an improvement over the previously reported palladium-catalysed Suzuki-Miyaura coupling approach to form asymmetrical biphenyl bis-amino acids which was found to return racemic products. 38 Biarylphosphine ligands such as SPhos are known to increase the rate and yield of palladium-catalysed reactions. 39 These ligands are believed to stabilise monoligated Pd(0), and in addition, have the ability to stabilise Pd(II) centres. 39,40 It was found that an active catalyst formed from 1 : 1 SPhos to Pd 2 (dba) 3 led to improved yields (compared to earlier reports) of Negishi cross-coupled phenylalanine derivatives 3a-d (Scheme 8). 41 Interestingly it was found that when 2-iodoaniline was used in the presence of SPhos the expected amino acid derivative was not delivered and instead a lactam 18 was formed in good yield (85%) (Scheme 9).
Jackson's group built upon the observation that SPhos improved the yields of Negishi cross-couplings. By taking inspiration from their work on the kinetics of β-elimination, the Boc protecting group (see mechanistic studies) 42 was Scheme 4 β 2 -Amino acids synthesised through Negishi cross couplings, Jackson and co-workers. 33 Scheme 5 Synthesis of a naphthyl-appended amino acid precursor by Szczepankiewicz et al. 37

Scheme 6
Negishi cross-couplings of aryl bromides with an iodo serine derivative.
replaced with a trifluoro-acyl (TFA) group (Scheme 10). 43 The TFA protecting group had been shown previously to perturb the alkyl zinc halide from undergoing β-elimination, (see later section on mechanistic studies for further information) so it was hoped that the combination of SPhos and a TFA protecting group on the zinc halide would lead to improved reaction conditions for the synthesis of phenylethylamine derivatives. TFA-protected zinc iodides derived from phenylalanine 19 and valine 21 were shown to react smoothly with a range of aryl iodides, delivering good to excellent yields of aryl appended protected amines 20a-c/22a-c.
The synthesis of trifluoromethyl appended aromatic amino acids were reported by Featherston and Miller (Scheme 11). 44 Cross-coupling was successfully carried out between Cbzprotected iodide 23 and a fluoro-aromatic to give 24 in a 54% yield. The aromatic amino acid 24 was then successfully transformed into the corresponding ketone through treatment with ozone. The peptides formed in the study were found to be cata-lytically active in the oxidation of substituted styrenes to their corresponding epoxides. In addition to their applications in catalysis, fluorinated amino acids have also been shown to be useful as biophysical probes and as structural components in peptide sequences. 45

Synthesis of heteroaromatic amino acids
Drug development and discovery often utilise heteroaromatic cores to deliver potent biological activity, 46,47 and elaboration of heterocyclic cores with other moieties is a cornerstone of fragment-based drug discovery (FBDD). [48][49][50] Within this context heteroaromatic amino acids have been shown to be valuable building blocks for the synthesis of many biologically active compounds. [51][52][53] The Negishi cross-coupling is one way in which heteroaromatic-appended amino acids can be synthesised in relatively few steps with great modularity.
Synthesis of heteroaromatic amino acid derivatives formed through Negishi cross-coupling was reported by Göbel and coworkers. 54 Negishi cross-coupling was successfully carried out using a range of aryl bromides and iodides with oxazolidine 27 (Scheme 12). Heteroaromatics such as quinoline 30e and pyrimidine 30f were found to be amenable to the methodology, and the inclusion of commonly used protecting groups (Boc and Fmoc) allowed the formed novel amino acids to be incorporated into longer peptide sequences by solid phase peptide synthesis (SPPS).
A detailed study into the synthesis of ortho-substituted nitroaromatic containing amino acid derivatives was carried out by Tuttle et al. (Scheme 13). 55 Reaction optimisation between 1-iodo-2-nitrobenzene and iodozinc compound 31 showed that Pd(OAc) 2 in combination with XPhos in a 1 : 2 ratio gave a 92% yield of the desired cross-coupling product. In addition, it was found that 1-chloro-2-nitrobenzene was also amenable to cross-coupling in 87% yield. A range of ortho and multiply-substituted nitro-bromobenzenes and pyridines were used as cross-coupling partners with iodozinc compound 31. Yields for nitropyridines ranged between 22% and 29% whilst nitrobenzenes ranged between 41% and 77% yield.
Using these novel protected amino acid derivatives, it was possible to deliver the intended hydroxamic acid targets 35 Scheme 9 Unexpected lactam formation observed in the cross-coupling of an iodoaniline with an iodo serine derivative.
Scheme 10 Amines prepared from amino acids by Negishi cross coupling, 19 and 21 prepared from phenylalanine and valine respectively in two steps. (Scheme 14) (similar to the species observed in Scheme 9). These heterocyclic motifs were of interest for their potential biological activity and the developed methodology allowed potential access to a library of these hydroxamic acids.
In 2014 Usuki et al. reported the improved Negishi crosscoupling of halopyridines with an iodozinc species derived from aspartic acid (Scheme 15). 56 Previous work such as that reported by Tuttle et al. (Scheme 13) had shown that yields for pyridine cross-coupling products significantly lagged behind their non-heteroaromatic equivalents. 57 After extensive reaction optimisation, a catalytic system of Pd 2 (dba) 3 in combination with a P(2-furyl) 3 ligand was shown to be the best for the cross-coupling of halopyridines, with a 90% yield recorded in the reaction between 4-iodopyridine and 36. In moving to bromo-substituted pyridines lower yields were recorded for all substrates, with 2-3-and 4-bromopyridines giving 44%, 33% and 23% yields respectively. Chloropyridines, regardless of halide position on the ring, were found to be ineffectual in all The ability to append furyl and thienyl functionalities onto amino acids was reported by Cobb and co-workers in 2015 (Scheme 16). 58 The synthesis of 2-thienyl-alanine was achieved in a moderate 42% yield but this was an improvement over the previously reported, 10% yield. 59 Both Boc and Fmoc protecting groups were shown to be tolerated by the cross-coupling conditions. A Boc-protected 2-furyl derivative was subsequently shown to be stable in a (Benzotriazol-1-yloxy)tripyrrolidinophosphonium hexafluorophosphate (PyBOP) mediated peptide coupling.

Synthesis of complex amino acids
Beyond aromatic and heteroaromatic examples, the Negishi cross-coupling reaction has been used to prepare a wide range of complex amino acids.
The formation of biphenyl-linked amino acids and biphenyl amino acid derivatives can be easily achieved using a crosscoupling approach (Scheme 18). 60 The synthesis of the macrocyclic tripeptide K13 61 and a non-natural macrocyclic protease inhibitor used this approach. 62 Building upon these studies the synthesis of biphenyl linked amino acids was reported where a strategy of coupling iodosubstituted bi-aryls with alkyl zinc reagents was adopted. Negishi cross-coupling was successfully carried out between iodozinc L-serine derivative 2 and a range of mono-and diiodo-substituted bi-phenyls.
The reaction was found to be applicable to both acyl and benzyl protecting groups on the bi-phenyl moiety. However, it was found that a large excess (4 equiv.) of the iodozinc species was required in order to achieve a satisfactory yield of 65% of the biaryl derivative 45 (Scheme 18).
Glycoproteins have been shown to be critical to cell adhesion, 63,64 differentiation 65 and growth, and they have also been used as new targets for therapeutics and diagnostics. 66 Ousmer et al. utilised a Negishi cross-coupling to form a p-(Cglucopyranosyl)phenylalanine derivative (Scheme 17). 67 An iodo-substituted phenylalanine derivative was successfully coupled to a glucal 41 in an excellent 90% yield. This was then further elaborated by stereo-selective hydroboration to yield the desired glucose modified phenylalanine 43 in a 37% overall yield.
Negishi and Suzuki-Miyaura cross-couplings were shown to be amenable to the preparation of imidazoles derived from amino acids (Scheme 19). 68 Imidazoles derived from (S)-Ala, (S)-Leu and (S)-Phe 46 protected with ethoxymethyl groups gave substrates 47 which were found to readily undergo Negishi cross-couplings to form corresponding bi-aryl systems.
Scheme 16 Furyl and thienyl-appended amino acids synthesised using Negishi cross-coupling, Cobb and co-workers. 58 Scheme 17 Synthesis of a p-(C-glucopyranosyl)phenylalanine derivative. The bi-aryl protected amines 48a-c were successfully obtained in yields ranging from 34% to 88% (Scheme 19). Jackson and co-workers would revisit the synthesis of the macrocyclic peptide K13 61 and expand their scope to the cyclic peptide OF4949-III 52 in 2009. 69 Building on their earlier synthesis of K13 and their reported Negishi-based methodology for bi-aryl linked amino acids, 60 Jackson et al. reported an intramolecular Negishi strategy to afford K13 and a Negishi cross-coupling key step synthesis of OF4949-III (Scheme 20). Utilising this approach OF4949-III 52 was synthesised in 12 steps with an overall yield of 20%.
Other natural products have also been successfully synthesised using Negishi cross-couplings of amino acid building blocks as a key synthetic step. The Kapakahine family of natural products has shown promising preliminary biological activity against P388 murine leukemia cells and thus accessing these compounds was of significant interest. 70 The synthesis of Kapakahine E and F, by Espejo and Rainier utilised a Negishi cross-coupling to form a key intermediate 54 which was subsequently taken through two end game synthetic routes to yield the two natural products (Scheme 21). 70 Negishi cross-couplings that can utilise triflates in the place of the halide component have been widely utilised in mainstream organic synthesis. 71,72 Within the field of peptide chemistry it was reported that a PdCl 2 (PPh 3 ) 2 mediated crosscoupling between iodozinc reagents and cycloalkenyl triflates could be carried out, but the yields obtained were only moderate, ranging between 41% and 46%. 73 The Negishi cross-coupling of iodozinc compounds with cycloalkenyl triflate derivatives was also reported by Jackson and co-workers in 2012 (Scheme 22). 74 In this study, the addition of a phosphine ligand (SPhos), in combination with a Pd 2 (dba) 3 catalyst, gives rise to an increase in the yield of the cross-coupling reaction product to 70% for the cyclopentenyl and cyclohexenyl derivatives 55a/b.
Yields for cycloheptenyl and cyclooctenyl products 55c/d, however, were moderate at 54% and 58%, respectively. This led the authors to propose that the addition of lithium chloride could be advantageous. Lithium chloride had been shown previously to increase the yields in Stille reactions, and can increase the reactivity of organozinc halides. 75 Indeed, the addition of 1.8 equivalents of LiCl to the Negishi cross-coupling reaction mixture led to increased yields for the cycloheptenyl and cyclooctenyl derivatives 55c/d, boosting yields to 62% and 63%, respectively. In addition to the Negishi methodology developed it was shown that the cross-coupled product 55a could react with HF pyridine to furnish the fluoro compound 56 in 30% isolated yield (Scheme 23). This compound had been reported earlier as a building block for an enzyme inhibitor. 76 The previous synthesis of compound 56 had been relatively low yielding and had required a chiral resolution step. 77 The cross-coupling methodology therefore offered a simpler and more efficient synthesis of these fluorinated motifs.
The Negishi cross-coupling of pyridines was extended to bis-halo pyridines by Usuki and co-workers (Scheme 24). 78 Utilising a di-bromo-substituted pyridine 57 a double Negishi and Sonogashira cross-coupling was employed to garner key intermediate 58, which was further elaborated to yield desmosine 59. In the same report, a deuterated version of desmosine was also successfully synthesised following the same crosscoupling strategy. Desmosine is a biomarker for degradation of elastin-containing tissues, which is a symptom of diseases such as atherosclerosis, aortic aneurysm, cystic fibrosis and chronic obstructive pulmonary disease (COPD). [79][80][81] The synthesis of desmosine therefore is of importance for drug discovery and the rapid diagnosis of diseases such as COPD.
In 2017 Usuki and co-workers elaborted their synthesis of Desmosine to build it into a cyclic peptide (Scheme 25). 82 This work was driven by the fact that the proposed stuctures of elastin suggest cross-linkage through Desmosine. 83,84 Through the use of a Desmosine analogue the cross-linking motif within elastin was hoped to be probed. Usuki and co-workers utilised the same general strategy to build the Deomsine core as they had previously published. 78 A Sonogashira cross-coupling afforded the intermediate In 2017 the total synthesis of Nannocystin A 64 was reported by Liu et al. 85 Within this report a Negishi cross-coupling of an iodo serine derviative was employed to build the key fragment 63 (Scheme 26). The final natural product was delivered with a longest linear sequence of 11 steps in a 5.3% overall yield. Nannocystin A 64 has been found to possess potent biological activity. It has been shown to be an inhibitor of elongation factor 1 and inhibits cancer cell growth. 86 The Negishi cross-coupling reaction to form a novel amino acid derivative was vital to access this potent natural product synthetically.

Mechanistic studies
Several mechanistic studies have been undertaken to try and gain fundamental knowledge of the reactivity of alkyl zinc amino acid derivatives. Using a range of analytical techniques, the species responsible for catalysis and mechanistic pathways have been probed. From these studies a greater understanding of the workings of Negishi cross-couplings of iodo amino acids has been established.
One of the major problems in the synthesis of β-amino acids using Negishi cross-coupling is that the iodozinc crosscoupling partner (derived from species such as aspartic or glutamic acid) is intrinsically susceptible to β-elimination to form an alkene. In order to understand these problems, NMR spectroscopy was utilised to probe the decomposition of β-amidozinc reagents. NMR kinetic studies showed that the rates of β-elimination in both THF-d 8 and DMF-d 7 were first order and pointed towards a unimolecular rate-limiting step. In addition, it was observed that the rate constant in the case of DMF was four times smaller than that for THF. Therefore the elimination process is significantly perturbed in DMF. Calculation of the activation parameters led to the conclusion that the elimination process was syn in nature, invoking an intramolecular coordination between the carbamate and the zinc centre (Scheme 27). DMF therefore is believed to compete with the carbamate for zinc, thus slowing the rate of elimination.
In 2003 it was reported that a β-amidozinc reagent was more resistant to elimination when a better leaving group was employed. 87 Jackson and Rilatt further studied this kinetic stability and reactivity of β-amino alkyl zinc iodides in 2008. 88 They found that the N-protecting group on the zinc iodide was critical in determining the rate of β-elimination. Boc protecting groups were found to increase the rate of elimination, and from this it was theorised that the carbonyl group (within the carbamate) was competing with the ester for coordination with zinc, leading to increased elimination. Replacing the Boc group with a TFA group perturbed the rate of elimination. The lower Lewis basicity of TFA on its carbonyl group helps prevent coordination to the zinc, and thus elimination. On top of this, it was observed that upon changing the Boc group for TFA, the rate constant changed from being first order to second order, indicating that the TFA elimination pathway follows a different mechanism from the Boc elimination. In the same report the reactivity of the β-amino alkyl zinc iodides towards Negishi cross-coupling was also studied. It was found that the N-protecting group made little difference to the rate of crosscoupling with the main factor being the proximity of the ester from the Zn site. It was concluded that a TFA N-protecting group offers much greater stability toward β-elimination whilst not affecting the rate of Negishi cross-coupling.
Negishi cross-coupling was found to be applicable to aqueous environments. It was reported that the presence of tetramethylethylenediamine (TMEDA) allows the Negishi cross-coupling of phenylalanine derivatives to be successfully carried out in the presence of water (Scheme 28). 89 Being able to carry out Negishi reactions in water is of significant interest to peptide chemists, as it allows the post-translational modification of peptides. It was found that the TMEDA helps prevent protonation of the alkylzinc iodides in water. It was proposed that the stabilisation occurs due to the TMEDA inducing ionisation in the zinc-iodine bond. This hypothesis was supported by evidence from tandum mass spectometery, infrared multiphoton dissociation and DFT calculations. Conditions for the successful synthesis of alkylzinc iodides and subsequent Negishi coupling in water were probed, and it was found that the use of nano zinc in combination with TMEDA and a PdCl 2 (AmPhos) 2 gave good yields ranging between 28% and 73% of phenylalanine derivatives.

Scheme 26 Total synthesis of nannocystin A, Liu et al. 85
Scheme 27 Syn elimination of Boc protected amidozinc reagents.

Scheme 28
Negishi cross coupling in water of an iodo serine derivative, Jackson and co-workers. 89

Conclusions
The Negishi cross-coupling reaction has been shown to be a highly versatile reaction allowing an expedient route to the synthesis of many unnatural amino acids. Since the pioneering work of Jackson and co-workers advances in available reagents (e.g. new ligands) has led to an increase in substrate tolerance and product yields. In addition mechanistic studies have helped researchers to understand the requirements for successful cross-couplings across a range of substrates and this has in turn allowed access to more novel unnatural amino acids than ever before. The ability to prepare novel aromatic, heteroaromatic and complex amino acids opens up a myriad of new potential applications for these motifs. For example, as highlighted in this review, unnatural amino acids accessed via Negishi cross-couplings have been exploited in the fields of natural product synthesis, drug discovery, materials/polymer chemistry and catalysis.

Conflicts of interest
There are no conflicts to declare.