Diversifying Amino Acids and Peptides via Deaminative Reductive Cross-Couplings Leveraging High-Throughput Experimentation

Abstract

A deaminative reductive coupling of amino acid pyridinium salts with aryl bromides has been developed to enable efficient synthesis of noncanonical amino acids and diversification of peptides. This method transforms natural, commercially available lysine, ornithine, diaminobutanoic acid (DAB), and diaminopropanoic acid (DAP) to aryl alanines and homologated derivatives with varying chain lengths. Attractive features include ability to transverse scales, tolerance of pharma-relevant (hetero)aryls and biorthogonal functional groups, and the applicability beyond monomeric amino acids to short and macrocyclic peptide substrates. The success of this work relied on high-throughput experimentation (HTE) to identify complementary reaction conditions that proved critical for achieving the coupling of a broad scope of aryl bromides with a range of amino acid and peptide substrates including macrocyclic peptides.

Graphical Abstract

INTRODUCTION

Over the past two decades, innovations in synthetic organic chemistry, biotechnology and computational design have yielded exquisitely selective peptide and protein therapies and probes that have significantly advanced drug discovery and chemical biology.¹ Within the pharmaceutical arena, the peptide renaissance can be attributed to advances toward in vitro transcription-translation platforms (e.g., mRNA display), which have evolved to accommodate the incorporation of noncanonical amino acids into diverse peptide libraries, granting access to previously undruggable targets through the fine-tuning of permeability, potency, and stability.² In addition, noncanonical amino acids have played a critical role in the understanding of biological processes within cells through the labeling of peptides and proteins with biorthogonal handles such as benzophenones, alkynes, azides, tetrazines, cyclopropenes, and diazirines.³

As evidence of their importance, 8 of the top 20 peptide drugs by sales contained a noncanonical amino acid and most recently, nirmaltrelvir (Paxlovid^™), Pfizer’s breakthrough oral antiviral for COVID-19, is a tripeptide constructed from only noncanonical amino acids (Scheme 1A).^1d In addition to approved drugs, there are emerging peptide pre-clinical candidates with high noncanonical amino acid content, including an investigational PCSK9 inhibitor which was identified using the mRNA display technology.⁴ While several classes of noncanonical amino acids exist, aryl alanines are highly prevalent in pharmaceuticals due to their versatility and ease of incorporation through solid-phase peptide synthesis or genetic code expansion techniques.⁵ Further, although carfilzomib features a phenethyl side chain, routes to homologated aryl alanines are underdeveloped, making them underexplored in pharmaceutical applications. As a result, general methods for efficient generation of aryl alanines and their homologated analogues would be highly desirable for advancing peptide drug discovery.

Scheme 1.

State of the Art in Aryl Alanines and Homologated Derivatives

Despite this importance of aryl alanines and their homologated derivatives, methods to access them remain limited, especially in high enantiopurity. Asymmetric synthesis via Strecker, Petasis, or hydrogenation reactions enable incorporation of various side chains, but can require empirical catalyst optimization for individual substrates.⁶ Alternatively, Michael additions of radical intermediates to dehydroalanines can provide noncanonical amino acids, but most of these methods are currently relegated to the preparation of racemic amino acids.⁷ Another attractive approach is to utilize the innate enantiopurity of a canonical amino acid (Scheme 1B). To this end, the Pd-catalyzed C-H arylation of alanine derivatives to noncanonical phenylalanines was pioneered by the Yu group and later expanded to peptides and homologated aryl alanines by the groups of Weng and Shi, respectively.⁸ However, these reports are limited to the use of aryl iodides as electrophiles. In addition, serine can be readily halogenated and transformed into noncanonical amino acids via reductive cross-electrophile coupling; however, halogenated amino acids are often unstable under reaction conditions, such as those used in solid phase peptide synthesis, hampering their broad applicability and use in diversity-oriented synthesis.⁹ Lastly, the Walczak group has recently demonstrated an umpolung approach using boronoalanines derived from aspartic acid.¹⁰

Recognizing the advantages of using the innate enantiopurity of a natural (or non-proteinogenic) amino acid to prepare enantiopure noncanonical variants, we have developed deaminative conditions that allow lysine, ornithine, 2,4-diaminobutyric acid (DAB), and 2,3-diaminopropanoic acid (DAP) to be converted to aryl alanines and homologated derivatives with varying chain lengths.^11,12 Because reductive cross-electrophile couplings occur under neutral (non-basic) conditions,^13,14,15 this approach would allow conservation of enantiopurity in the coupling of Katritzky pyridinium salts¹⁶ and aryl bromides to access diverse arylated amino acids (Scheme 1C). We prioritized the use of aryl bromides over aryl iodides because they are a much more available and diverse substrate class. For example, there are greater than three times as many aryl bromides as aryl iodides in Merck & Co., Inc. inventory. To maximize utility, it was critical that this method was scalable, tolerant of pharma-relevant (hetero)aryls and bioorthogonal functional groups, and amenable to both monomeric amino acid starting materials and peptide substrates. The amenability of these cross-couplings to be conducted on nano-, micro-, millimole scales also provides confidence for translatability across various stages of drug discovery. Finally, the stability of lysine-derived Katritzky salts makes them ideal building blocks for solid-phase peptide synthesis (SPPS) applications, allowing seamless peptide synthesis and late-stage functionalization (LSF) in solid or solution phase. Importantly, these goals were met by a successful academic/industry collaboration that facilitated multi-parameter optimizations across substrates varying in complexity to identify reaction conditions that exhibit the broadest scope of coupling partners.

RESULTS AND DISCUSSION

We selected the arylation of the pyridinium salt of N-Boc-lysine methyl ester (3a) for optimization. This pyridinium can be easily prepared in 3 steps from Boc-Lys(Z)-OH. Starting with our previously reported conditions for the reductive coupling of alkylpyridinium salts and aryl bromides,^15a we observed a promising 37% yield of product 4 (Table 1, entry 1). Notably, under these reductive conditions, the enantiomeric excess of the amino acid precursors was conserved. The use of DMF instead of NMP, as well as switching to 3 equiv of LiCl as the additive, led to 77% yield of 4 (entry 4). However, in a preliminary investigation of the aryl bromide scope using Mn as reductant and bipy as ligand, high yields were not broadly achieved. Although the use of Zn or TDAE led to low product yields under these conditions (entries 5 and 6), optimization of solvent, additive, and catalyst with TDAE afforded 93% yield of 4 (entries 7–11, see Supporting Information. Tables S1–S10 for detailed optimization experiments). Notably, the reaction proceeds in the absence of additive, albeit in somewhat lower yield (entries 9 vs. 10). Under these optimized conditions, the use of other commonly used ligands for reductive couplings was not advantageous (Table S10).

Table 1.

Optimization on 0.1 mmol scale^a

entry	[Red]	additive (equiv)	solvent ([1a], M)	yield (%)b
1	Mn	MgCl2 (1.0)	NMP (0.17)	37 (>99% ee)c
2	Mn	MgCl2 (1.0)	DMF (0.13)	53
3	Mn	MgCl2 (3.0)	DMF (0.10)	63
4	Mn	LiCl (3.0)	DMF (0.10)	77 (>99% ee)c
5	Zn	LiCl (3.0)	DMF (0.10)	35
6	TDAE	LiCl (1.0)	DMF (0.10)	37
7	TDAE	MgBr2 (1.0)	DMF (0.10)	41
8	TDAE	MgBr2 (1.0)	DMPU (0.10)	71
9	TDAE	NaCl (1.0)	DMPU (0.10)	78
10	TDAE	none	DMPU (0.10)	50
11d	TDAE	NaCl (1.0)	DMPU (0.10)	93 (>99% ee)c

^aConditions: 3a (0.10 mmol), Ar–Br (1.2 equiv), NiCl₂·DME (10 mol %), bipy (12 mol %), reductant (2.0 equiv), additive, solvent, 80 °C, 24 h, unless otherwise noted.

^bYields of single experiments determined by ¹H NMR analysis using 1,3,5-trimethoxybenzene as internal standard, which was added after work-up.

^cDetermined by SFC using a chiral stationary phase.

^dNiBr₂·diglyme in place of NiCl₂·DME. Bipy = 2,2’-bipyridine.

Under these optimized conditions (Table 1, entry 10), the scope of aryl bromide was promising. On 0.5 mmol scale, 92% yield of product 4 was obtained in 99% ee (Scheme 2). The reaction was scalable, resulting in 82% yield of 4 on 5.0-mmol scale. Both electron-rich (4–7) and electron-poor (8–13) aryl bromides led to good product yields. A range of functional groups are tolerated, including ethers (4, 6), dioxolane (5), secondary amide (7), nitrile (9), trifluoromethyl (10), ketones (8, 11), aldehyde (12), and α,β-unsaturated ester (13). Compatibility with benzophenone (8) and alkyne (14) are particularly notable due to their utility in bioorthogonal ligation.⁵ Interestingly, however, although some protic groups (e.g., amide 7) were tolerated, the use of aryl bromides containing phenols or carboxylic acids as well as free NH-bearing heteroaryl bromides gave little to no desired product (15–19). In these cases, the main products result from homocoupling of the arylbromide and protodehalogenation. Recovered pyridinium and aryl bromide were also observed.

Scheme 2. Initial Aryl Bromide Scope^a.

^a Conditions: 3a (0.50 mmol), Ar–Br (1.2 equiv), NiBr₂·diglyme (10 mol %), 2,2’-bypyridine (bipy) (12 mol %), TDAE (2.0 equiv), NaCl (1.0 equiv), DMPU (0.1 M), 80 °C, 24 h. Isolated yields, unless otherwise noted. ^b Yield determine by ¹H NMR using an internal standard.

These preliminary results set the stage to leverage HTE to address the limitations highlighted in Scheme 2. In particular, we wanted to re-evaluate the relationship of ligand and reductant to broaden the aryl bromides that could be used in this method. An HTE approach allows for multi-parameter optimization that enables observation of trends and interdependencies that may be missed in traditional linear reaction optimization.¹⁷ Due to the success of previous academic/industrial partnerships that leverage HTE,¹⁸ and to ensure that this method would be pharmaceutically relevant, we undertook this effort in collaboration with researchers at the state-of-the-art HTE facilities at Merck & Co., Inc., Rahway, NJ, USA. As described below, this collaboration and multi-parameter reaction optimization: 1) resolved the specific limitations highlighted in Scheme 2; 2) elucidated the scope and limitations with a broad range of pharmaceutically relevant diverse aryl halides and pyridinium substrates; and 3) demonstrated the compatibility of these couplings with material sparing microscale and nanoscale platforms.

Given the limitations we observed with acidic aryl bromides (Scheme 2), we re-evaluated the conditions for our original model reaction to see if other sets of conditions would also show promise. As illustrated in Scheme 3, the HTE investigations commenced with the optimization of ligand, additive, and reductant for the formation of product 4. These experiments were conducted at 10 μmol scale using 96 well aluminum blocks. Nanomolar scale, 1536 well blocks were not used due to the challenges with accurately dosing heterogeneous reductants (Mn and Zn). The solvent choice (DMA) for HTE studies was largely dictated by the enhanced solubility of certain pyridinium salts in DMA vs DMPU. Although several additives promote this transformation (Tables 1, S4, S6, and S8), TBAI was selected for HTE studies due to its solubility in DMA. Because extensive ligand optimization had been performed on 0.1-mmol scale with both Mn and TDAE (see representative results in Tables S2 and S10), a limited set of ligands were tested. Phenanthroline (L2) and bipyridine (L3) were two of the most promising ligands on 0.1-mmol scale. Although tridentate ligands had provided low yields on 0.1-mmol scale, pyridine-bis(carboxamidine) L1 and tri-tert-butylterpyridine (L4) were also included. Notably, L1, originally reported by Weix,¹⁹ has proven highly effective in reductive couplings and had not been investigated in the optimization studies on 0.1-mmol scale. These ligands were pre-stirred with NiBr₂·DME to ensure ligation(see Table S11 for additional details regarding order of addition and reproducibility). Reaction efficiency was judged based on LC area % (LCAP) of the product peak in the UPLC chromatogram. Consistent with the results in Table 1, TDAE is the most effective reductant using bipy (L3) as the ligand. Interestingly, however, Mn and Zn can be effective reductants using L1 as the ligand. Furthermore, TBAI promotes the TDAE reactions. However, couplings using Zn and Mn proceed well in the absence of TBAI. Ultimately, these results enabled the identification of several promising reaction conditions for evaluation of the scope using inorganic (Mn or Zn) or organic reductants (TDAE).

Scheme 3. Ligand/Reductant Interdependence^a.

^a Conditions: 3a (10 μmol), Ar–Br (1.5 equiv), NiBr₂·DME (10 mol %), ligand (12 mol %), reductant (2.0 equiv), TBAI, DMA (0.1 M), 80 °C, 24 h. Numbers in each box represent the experimental LCAP. Reaction success was determined by LCAP (LC area %) of the desired product from the UPLC-MS analysis of the crude reaction mixture. Conditional formatting was applied to highlight the range of product LCAPs (see the key above).

To elucidate whether these newly identified conditions afford a broader scope than depicted in Scheme 2, lysine pyridinium 3a was coupled with diverse (hetero)aryl bromides including those containing acidic functionalities (Scheme 4). This investigation was conducted using the 3 sets of highest yielding conditions identified above: A) TDAE with L3 and TBAI, B) Mn with L1 and TBAI, or C) Zn with L1. As shown Schemes 4 and 5, under TDAE/L3 conditions, this microscale HTE investigation revealed remarkable heterocycle and functional group tolerance. Aryl bromides bearing heterocyclic rings such as pyridine, indazoles, thiazole, furan, oxadiazoles, azaindoles, triazoles, indoline, and imidazole led to the observation of the desired products, as measured by product LCAP.²⁰ However, 5-membered heteroaryl bromides generally provided low product LCAPs. In general, TDAE led to the highest product yield across a range of halides. However, as highlighted by the arrows in Scheme 5 and shown for selected examples in Scheme 4, Zn or Mn was more effective for halides bearing acidic functional groups that had proved problematic under our original conditions (see Schemes 2 and 4, 15, 17, and 18 from Br-8, Br-5, and Br-12, respectively). We speculate that the basicity of TDAE may be partly responsible for the low product LCAPs with these acidic substrates. However, further studies are needed to fully elucidate the complex relationships between substrate, ligand, reductant, and additive. Importantly, the scalability of these observed trends was confirmed by performing 17 reactions in Scheme 4 on 0.5 mmol scale.

Scheme 4. HTE Investigation of Aryl Bromide Scope using TDAE as Reductant^a.

^a For microscale experiments, numbers represent the experimental product LCAP for reactions performed on 10 μmol scale. For selected examples, numbers at the bottom of each box represents isolated yields of the products on 0.5 mmol scale. Conditional formatting was applied to product LCAP (see the key above). Microscale conditions: 3a (10 μmol), Ar–Br (1.5 equiv), NiBr₂·DME (10 mol %), L1 or L3 (12 mol %), reductant (2.0 equiv), TBAI (0 or 1.0 equiv), DMA (0.1 M), 80 °C, 24 h. (A) L3, TDAE, TBAI. (B) L1, Mn⁰, TBAI, (C) L1, Zn⁰. For 0.5-mmol experiments, isolated yields (IY) are reported. 0.5-mmol conditions: 3a (0.50 mmol), Ar–Br (1.2 equiv), NiBr₂·diglyme (10 mol %), L1 or L3 (12 mol %), reductant (2.0 equiv), NaCl (0 or 1.0 equiv), DMPU (0.1 M), 80 °C, 24 h. (D) L3, TDAE, NaCl. (E) L1, Mn⁰.

Scheme 5. Interdependence of Aryl Bromide and Reaction Conditions^a.

^a Aryl bromide numbers on the x-axis are the same as those in Scheme 4. Experimental product LCAP are reported for reactions performed on 10 μmol scale. Conditions: 3a (10 μmol), Ar–Br (1.5 equiv), NiBr₂·DME (10 mol %), L1 or L3 (12 mol %), reductant (2.0 equiv), TBAI (0 or 1.0 equiv), DMA (0.1 M), 80 °C, 24 h. (A) L3, TDAE, TBAI. (B) L1, Mn⁰, TBAI, (C) L1, Zn⁰.

Encouraged by the generality of these reactions at microscale (10 μmol), the compatibility of analogous couplings at nanoscale (100 nmol) was evaluated. Over the past 6 years, nanoscale HTE has become a powerful technology in the context of medicinal chemistry programs since it enables the rapid and parallel elucidation of the reactivity landscape using minimal quantities of precious substrates.²¹ The results of such screens enable chemists to selectively scale the productive reactions while prioritizing the unsuccessful couplings for further optimizations using alternate conditions. Two key practical requirements to effectively translate the microscale reactions to nanoscale platforms are high boiling solvents and the homogeneity of the reagent stock solutions. Notably, the TDAE reaction conditions satisfy these requirements and hence were poised for these nanoscale studies.²² As such, we attempted the coupling of lysine pyridinium 3a with structurally diverse halides using TDAE conditions. Remarkably, significant product formation was observed from these nanoscale reactions, with reactivity trends across various halides mirroring those observed for analogous couplings performed at microscale (Scheme 6). Consistent with the microscale results, halides bearing acidic functionalities do not afford the desired product using TDAE.

Scheme 6. Comparison of Reaction Performance at Nanoscale vs. Microscale^a.

^a Experimental LCAP for reactions performed on 100 nmol or 10 μmol scale. Nanoscale conditions: 3a (0.1 μmol), Ar–Br (1.5 equiv), NiBr₂·DME (10 mol %), 4,4’-tBubipy (12 mol %), TDAE (2.0 equiv), TBAI (1.0 equiv), DMA (0.1M), 80 °C, 24 h. For microscale conditions, see Schemes 4 and 5, Conditions A.

Having optimal conditions compatible with a broad range of pharmaceutically relevant aryl halides, we next explored the applicability of these couplings using simple to complex pyridinium substrates. Katritzky salts derived from lysine, ornithine, DAB and DAP amino acids together with two tripeptides were evaluated. As depicted in Scheme 7, conditions using L3 and TDAE led to the highest product LCAPs for reactions using pyridiniums 3a–3e. In contrast, the L1/Zn conditions were more effective with challenging substrates such as DAP derivative 3f, where elimination competes with productive coupling, and the tripeptides 3g and 3h. For these more challenging substrates, TBAI seems to have a positive impact under these L1/Zn conditions, in contrast to the initial model reaction (see Scheme 3 vs Scheme 7); however, it is challenging to dose Zn consistently between plates on these scales, which may also contribute to this difference. We hypothesize that the basicity of TDAE may be problematic with DAP-derived pyridinium and the peptide substrates due to their susceptibility toward elimination reactions. Indeed, a Glorius screen of amino acid additives showed that the L3/TDAE conditions were sensitive to a number of canonical amino acid side chain functional groups (see Table S12). In contrast, the L1/Zn (and L1/Mn) conditions were much more robust. Importantly, these trends with respect to the pyridinium substrate were also manifested for reactions with various halides (Scheme 8). Together the screens described in Schemes 4–8 emphasize the importance of multi-parameter optimizations across substrates varying in complexity to identify reaction conditions that exhibit the broadest scope of coupling partners.

Scheme 7. Interdependence of Pyridinium Salt and Reaction Conditions^a.

^a Conditions: 3a-h (10 μmol), Ar–Br (1.5 equiv), NiBr₂·DME (10 mol %), ligand (12 mol %), reductant (2.0 equiv), TBAI (0 or 1.0 equiv), DMA (0.1M), 80 °C, 24 h. For TDAE, bipy (L3) used as ligand. For Mn and Zn, pyridine-2,6-bis(carboximidamide) dihydrochloride (L1) used as ligand. Conditional formatting was applied to highlight the range of product LCAP detected (see the key above).

Scheme 8. Optimal Conditions for Pyridinium/Bromide Pairs^a.

^a Conditions: 3a (10 μmol), Ar–Br (1.5 equiv), NiBr₂·DME (10 mol %), L1 or L3 (12 mol %), reductant (2.0 equiv), TBAI (0 or 1.0 equiv), DMA (0.1 M), 80 °C, 24 h. (A) L3, TDAE, TBAI. (B) L1, Mn⁰, TBAI, (C) L1, Zn⁰. Conditional formatting was applied to highlight the range of product LCAP detected (see the key above).

Having identified the most promising conditions for the various classes of amino acid and peptide pyridinium salts, we then demonstrated their scalability for the reactions using monomeric pyridinium amino acids (Scheme 9). Various side chain lengths were well tolerated, enabling synthesis of aryl alanine analogue 36, as well as homologated derivatives 34 and 35. Notably, the utility of HTE to identify appropriate conditions for various substrates is highlighted by the achievement of 70% yield of 36 using the L1/Mn⁰ conditions; in contrast product 36 was not observed using the L3/TDAE conditions. In addition, other common protecting groups could be used on both the amino and carboxylate groups.²³

Scheme 9. Scope of Pyridinium Salts^a.

^a Condition A: Pyridinium 3 (0.50 mmol), Ar–Br (1.2 equiv), NiBr₂·diglyme (10 mol %), bipy (12 mol %), TDAE (2.0 equiv), NaCl (1.0 equiv), DMPU (0.1 M), 80 °C, 24 h. Condition B: Pyridinium 3 (0.50 mmol), Ar–Br (1.2 equiv), NiBr₂·diglyme (10 mol%), L1 (12 mol%), Mn⁰ (2.0 equiv), DMPU (0.1M) 80 °C, 24 h. Isolated yields using Condition A except that Condition B was used to prepare 36. For 34–36, average yield of duplicate experiments (±5%). Ee’s determined by SFC using a chiral stationary phase.

As mentioned in the introduction, one advantage of the pyridinium substrates is their tolerance to solid-phase synthesis conditions, which provides opportunities for late-stage functionalization (LSF) of peptides, contrary to traditional halide substrates. Peptides 3g and 3k were made through solid-phase peptide synthesis (SPPS) using an Fmoc-OH lysine-derived pyridinium salt which after resin cleavage and subjection to our optimized conditions, afforded diversified peptides 40 and 41 in 32% and 40% yield respectively (Scheme 10A).

Scheme 10.

Solid-Phase Peptide Synthesis (SPPS) and Cross-Coupling

These encouraging results prompted us to assess whether the pyridinium peptides could be cross-coupled directly on resin. Using the Rink Amide resin, which is more stable to heat than the 2-chlorotrityl chloride resin (CTC-resin), pentapeptide 3l-resin was synthesized in 10 steps, with an estimated 82% crude yield (Scheme 10B, also see Supporting Information, page S45). The coupling conditions were then tested on this crude resin-bound peptide, yielding peptide 43 in 28% yield after cleavage and purification from corresponding peptide 42 (based on crude starting material, the true yield would be higher). Overall, pentapeptide 43 was prepared in 23% yield over 12 steps with a single chromatographic purification. This is a rare example of a Ni-catalyzed coupling on resin and will add to the diversification toolbox available to chemists for peptide derivatization.²⁴

Lastly, with an understanding of how to utilize SPPS to access short pyridinium peptides, we wanted to further exemplify the power of this methodology for LSF of a pharmaceutically relevant macrocyclic peptide. As such, we turned our attention to the structurally complex PCSK9 inhibitor in Scheme 1, as it contained a lysine residue that was the focus of extensive structure-activity relationship studies to modulate potency and rat mast cell degranulation.⁴ Following a similar SPPS procedure as shown in Scheme 10, we were able to readily access a pyridinium analog of the PCSK9 inhibitor in Scheme 1; however, our attempts to cross-couple with various aryl halides and conditions did not result in any product formation (see Supporting Information, page S59–65). We hypothesized that the thioether linkage might be interfering with the cross-electrophile coupling²⁵ and turned our attention towards synthesizing 44, a structurally related analogue with a fully peptidic backbone (Scheme 11). In addition to the thioether being challenging for LSF, there is a desire to remove thioether linkages in the context of peptide drug discovery, as they present a metabolic liability. To that end, the resin-bound linear precursor to 44 was obtained in 19 steps on solid-phase, and after resin cleavage and macrocyclization furnished the desired LSF substrate. Much to our delight, subjection of 44 to a subtle variation of our Zn-mediated reductive coupling conditions furnished derivatized product 45 in 10% isolated yield.

Scheme 11.

Application to synthesis of PCSK9 inhibitor analog

CONCLUSION

In conclusion, we have developed a reductive coupling of amino acid pyridinium salts with aryl bromides to enable efficient synthesis of noncanonical amino acids and diversification of peptides. The use of HTE strategies elucidated an interdependence between ligand and reductant, allowing identification of complementary sets of conditions, which proved crucial for maximizing the scope of aryl bromide and pyridinium substrates. L3/TDAE conditions were optimal for amino acids with longer side chains, and L1/Mn (or Zn) conditions were more effective for DAP and peptide substrates. Importantly, this reductive coupling can be applied to nanoscale synthesis, allowing these transformations to be performed across miniaturized scales, and enables preparation of aryl alanine homologues with various side chain tether lengths from natural and commercially available amino acid precursors. Further, these amino acid pyridinium salts and the Ni-catalyzed cross-coupling are amenable to solid-phase peptide synthesis, enabling efficient late-stage diversification of linear and macrocyclic peptide substrates thereby rendering them suitable for the optimization of peptide biopharmaceutical properties.