Unlocking New Prenylation Modes: Azaindoles as a New Substrate Class for Indole Prenyltransferases

Abstract

Aza-substitution, the replacement of aromatic CH groups with nitrogen atoms, is an established medicinal chemistry strategy for increasing solubility, but current methods of accessing functionalized azaindoles are limited. In this work, indole-alkylating aromatic prenyltransferases (PTs) were explored as a strategy to directly functionalize azaindole-substituted analogs of natural products. For this, a series of aza-l-tryptophans (Aza-Trp) featuring N-substitution of every aromatic CH position of the indole ring and their corresponding cyclic Aza-l-Trp-l-proline dipeptides (Aza-CyWP), were synthesized as substrate mimetics for the indole-alkylating PTs FgaPT2, CdpNPT, and FtmPT1. We then demonstrated most of these substrate analogs were accepted by a PT, and the regioselectivity of each prenylation was heavily influenced by the position of the N-substitution. Remarkably, FgaPT2 was found to produce cationic N-prenylpyridinium products, representing not only a new substrate class for indole PTs but also a previously unobserved prenylation mode. The discovery that nitrogenous indole bioisosteres can be accepted by PTs thus provides access to previously unavailable chemical space in the search for bioactive indolediketopiperazine analogs.

Graphical Abstract

Five isomers of aza-tryptophan and the corresponding series of cyclic Aza-tryptophan-proline dipeptides were tested as unnatural substrates for three well-known indole-prenylating enzymes, FgaPT2, CdpNPT, and FtmPT1. Most substrates were found to produce prenylated products, which revealed isomer-dependent regioselectivity and a previously unreported class of cationic N-prenyl pyridinium products.

Introduction

Indole diketopiperazines (IDPs) are a class of natural products (NPs) composed of differently functionalized, tryptophan (Trp)-containing cyclic dipeptides that display diverse biological activities.^[1] Specific examples include tryprostatins A and B, which display moderate cytotoxicity through inhibition of microtubule assembly;^[1-2] avrainvillamide, a novel antibacterial agent with activities against multidrug-resistant strains of Enterococcus faecalis, Staphylococcus aureus, and Streptococcus pyogenes;^[1] and fumitremorgin C, which can potentiate cytotoxic agents via inhibition of breast cancer resistance protein (BCRP).^[3] Such utilities make IDPs a promising class of drug leads, but a combination of low natural abundance, high structural complexity, and low aqueous solubility make moving them through the drug development pipeline an arduous endeavor.^[4] The former two issues can be tackled through the reconstruction and optimization of natural biosynthetic pathways in bioreactors, which show promise in overcoming the challenges of traditional synthetic methods.^[5] However, the poor aqueous solubility of IDPs is a much more fundamental problem caused by the formation of extensive networks of solid-state π-π stacking and hydrogen bonding interactions.^[6] Therefore, optimizing the IDP scaffolds to enhance solubility represents a legitimate strategy for better accessing their biological activities.

To fine-tune physical properties while preserving or improving biological activity, medicinal chemists often utilize isosteric substitution, the systematic replacement of chemical moieties in a manner that generally retains the steric and electronic profile of the original functional group.^[7] A relevant example in medicinal chemistry is the substitution of a relatively hydrophobic CH moiety in an aromatic ring with a more hydrophilic N atom. The similar atomic sizes of C and N allow the substituted molecule to retain its overall shape, while the incorporation of a lone pair introduces additional, generalized polarity, as well as new opportunities for hydrogen bonding.^[7] The latter has additional relevance in the context of binding affinity with the desired biological target, but both are invaluable for their effects on aqueous solubility.^[7-8]

Replacement of an aromatic CH with N is particularly effective in the case of the indole ring (Figure 1A). These “azaindole” isosteres have been incorporated into a number of drug scaffolds with highly varied biological activities: kinase inhibiting;^[9] anti-tumor,^[10] anti-HIV,^[11] anti-tuberculosis,^[12] antibiotic,^[13] and anti-diabetic agents;^[14] and as cannabinoid receptor modulators^[15] (Figure 1B). Azaindoles themselves can also act as versatile organometallic ligands,^{[10c, 16]} and their luminescent properties provide additional utility in the areas of light emitting diodes,^{[16d, 17]} catalysis,^{[16b, 18]} and methods of probing biochemical/protein interactions.^[19]

Figure 1.

(A) Azaindoles as isosteres of the natural indole ring. The isosteric N atoms are shown in blue. (B) Representative azaindole-containing compounds displaying diverse bioactivities. The azaindole moieties are highlighted in red.

While the advantages of azaindole incorporation into NPs are robust, the synthetic methods used to achieve the isosteric substitution represent a significant trade-off. Poly-functionalized azaindole scaffolds are most commonly accessed by late-stage generation of the isosteric moiety from functionalized precursors, a strategy which inevitably imposes restrictions on functional group tolerance and synthetic practicality.^[20] While synthetic methods for functionalizing pre-existing azaindoles do exist, they require either directed metalation groups to provide regiospecificity^[21] or activated scaffolds capable of participating in organometallic coupling reactions.^{[16b, 22]} As such, there is a high demand for innovative methodologies that functionalize azaindoles with high regioselectivity.

Similar issues with traditional synthetic methods continue to linger at the forefront of medicinal chemistry, and an increasingly common response is the utilization of chemoenzymatic/biocatalytic approaches to synthesize and modify drug compounds.^[23] Compared to more traditional synthetic catalysts, enzymes are highly enantio- and regioselective in the reactions they perform, and these factors can even be tuned through rational design or directed evolution.^[23-24] Additional advantages include the lower cost and toxicity of both the catalysts themselves and their waste products. Thus, biocatalysts capable of performing late-stage modification on azaindoles have the potential to dramatically increase access to NP analogs bearing these isosteres. To date, only a handful of enzymes have been shown to utilize azaindoles as substrates: Trp synthase from Salmonella typhimurium, which can generate azaindole-containing Trp analogs from free azaindoles;^{[5b, 25]} the viral halogenase VirX1 from the cyanophage Syn10, which can iodinate 6-azaindole;^[26] and the strictosidine synthase from Catharanthus roseus, which can accept azaindole-containing analogs of tryptamine to produce similar analogs of isositsrikine.^[27] Taken together, the established chemoenzymatic methods for functionalizing azaindoles are limited in both the positions that can be reacted and the groups that can be added. However, the known activity of these enzymes with azaindole-containing substrate analogs implies others may be similarly promiscuous.

Within this context, aromatic prenyltransferases (PTs) are a class of enzymes responsible for the late-stage functionalization of various NP scaffolds, several reactions of which encompass direct alkylation of an indole moiety.^[28] Native PT reactions involve the transfer of an allylic prenyl group from a high-energy dimethylallyl diphosphate (DMAPP) donor to an aromatic acceptor, typically proceeding via electrophilic aromatic substitution.^[29] Furthermore, recent studies reveal the enzyme class to exhibit high levels of promiscuity toward both donor and acceptor substrates, and the regioselectivity of each enzyme has been shown to change when catalyzing reactions involving non-natural substrates.^{[28d, 30]} In terms of indole PTs, at least one enzyme has been found to alkylate each available position on the indole ring, though the acceptor substrate for each of these reactions was not necessarily the same.^{[28b, 31]} In fact, a variety of indole-containing substrates have been described: Trp-containing cyclic dipeptides,^[32] 5-hydroxy-brevianamide,^[32c] 2-methyl-brevianamide,^[32c] Trp,^[33] methyl-Trp,^[33] fluoro-Trp,^[33] and indolylbutanone.^[34] Though the structural diversity of these indole-containing analogs is indeed vast, alteration of the core aromatic scaffold (i.e., isosteric substitution) remains largely unexplored. Therefore, the goal of the current study was to evaluate the ability of indole PTs to catalyze the alkylation of azaindole-substituted substrate analogs.

Towards this goal, a series of azaindole isomers (2-, 4-, 5-, 6-, and 7-azaindole; Figure 1A) were incorporated into substrate analogs of PT-catalyzed reactions. Specifically, all five azaindole-containing analogs of Trp (Aza-Trp) and their corresponding Trp-Pro cyclic dipeptides (Aza-CyWP) were synthesized, and the purified compounds were tested for activity with a set of representative indole-alkylating PTs (FgaPT2, CdpNPT, FtmPT1) using DMAPP as the prenyl donor. The analytical-scale reactions were evaluated for turnover using reverse-phase high performance liquid chromatography (RP-HPLC) and revealed each enzyme could utilize at least one azaindole-substituted analog as substrate. Subsequent structural characterization of scaled-up reactions established the regioselectivity of all major products. The FgaPT2-catalyzed alkylations were found to exclusively generate cationic N-prenylpyridinium products, which to the best of our knowledge are the first and only reported examples of chemoenzymatic N-prenylation of a pyridine moiety. Meanwhile, the CdpNPT- and FtmPT1-catalyzed prenylation of Aza-CyWP analogs resulted in a variety of products whose main alkylation sites were N1, C2, or C3. Thus, the current study has established novel Aza-Trp and Aza-CyWP analogs as substrates for three representative indole PTs, thereby expanding the utility of PT biocatalysis to the late-stage modification of pharmaceutically relevant azaindole-containing scaffolds.

Results and Discussion

Synthesis of Aza-Trp and Aza-CyWP Analogs

The synthesis of 7-aza-tryptophan (7f) has been accomplished previously^[35] and was thus repeated for the present study (Scheme 1A). Briefly, the synthesis began with 7-azaindole (7a) undergoing a Mannich reaction to form 7-aza-gramine (7b). This was followed by methylation of 7b’s tertiary amine to form a trimethylammonium leaving group (7c), which was subsequently displaced by a protected amino acid functionality (N-(diphenylmethylene)glycine ethyl ester). The final 7f racemate was obtained after imine hydrolysis (Scheme 1A), purification, and saponification of the ethyl ester (Scheme 1C). A similar procedure was attempted for the synthesis of 4-, 5-, and 6-Aza-Trp analogs (4f, 5f, and 6f, respectively) based on the reasoning that the similar electronic properties of 4-, 5-, and 6-azaindole (4a, 5a, and 6a, respectively) to 7a would allow for their substitution in the reaction scheme. In the case of 5a and 6a, the synthesis was indeed successful and demonstrated for the first time that the established method of 7f synthesis is also capable of producing 5f and 6f isomers. However, in the case of 4a, 4-azagramine (4b) was found to be robustly resistant to this alkylation strategy, as the methylation of 4b produced an insoluble brick red tar which promptly decomposed under subsequent reaction conditions. Therefore, 4f was synthesized using an alternative route (Scheme 1A).^[36] In this method, 4b was refluxed in xylenes in the presence of acetamidomalonate and catalytic NaOH to generate intermediate 4c, which was then saponified and decarboxylated/deacylated via extended reflux in 20% aqueous HCl. The pure 4f racemate was recovered via recrystallization without the need for column purification.

Scheme 1.

Synthetic methods used to produce azaindole-containing substrate analogs. (A) Known synthetic method for ethyl ester 7e^[35] applied to 5a, 6a, and 7a, alongside an alternative route adapted for 4a.^[36] (B) Method for synthesizing ethyl ester 2e modified from literature.^[37] (C) Saponification of amino acid ethyl esters. (D) Synthesis of aza-CyWP analogs by coupling Aza-Trp esters to Fmoc-l-Proline using 4-(4,6-Dimethoxy-1,3,5-triazin-2-yl)-4-methylmorpholinium chloride (DMT-MM). Fmoc deprotection and cyclization were completed by stirring in morpholine for 7 days.

To complete the set of Aza-Trp analogs, the side chain’s indole ring also needed to be substituted with indazole (2-azaindole). However, the increased withdrawal of electron density from the reactive C3 position by the isosteric nitrogen is known to prevent typical indole reactions, as indazole lacks the electron rich pyrrole moiety. Therefore, racemic 2-azatryptophan (2f) was prepared from 3-methyl-indazole (2a) using slight modifications to existing methods (Scheme 1B).^[37] Following N-Boc protection of the primary amine (2b), the indazole was subjected to radical bromination of the methyl group by NBS, forming the brominated indazole intermediate 2c. This compound was then converted into the amino acid ethyl ester 2d using the same strategy employed for 5c, 6c, and 7c, with minor modifications to the purification protocol to account for the lower polarity and basicity of the indazole moiety compared to the other azaindoles. After hydrolysis and purification of 2d, the N1-Boc group was quantitatively removed using trifluoracetic acid (TFA) in DCM to yield pure 2e. As with 5f, 6f, and 7f, a racemic mixture of 2f was obtained by saponification of 2e in basified ethanol and subsequent purification using RP-HPLC (Scheme 1C), with nearly quantitative yields. The synthesis and purity of all five Aza-Trp analogs were confirmed by high resolution electrospray ionization mass spectrometry (HRMS) and 1D- and 2D-nuclear magnetic resonance (NMR) spectroscopy (see Supporting Information).

For the preparation of the Aza-CyWP analogs (2g, 4g, 5g, 6g, 7g), the Aza-Trp ester intermediates 2e, 4e (synthesized by esterifying 4f, Scheme 1A), 5e, 6e, and 7e were coupled to Fmoc-l-proline using 4-(4,6-Dimethoxy-1,3,5-triazin-2-yl)-4-methylmorpholinium chloride (DMT-MM, Scheme 1D). The crude dipeptides were then deprotected and cyclized via prolonged stirring in morpholine,^{[30e, 38]} generating mixtures of diastereomers due to the racemic Aza-Trp analogs. While the diastereomers of 2g were partially separated using traditional silica gel chromatography, the other Aza-CyWP isomers co-eluted and had to be separated using RP-HPLC. This technique was also used to further purify the diastereomers of 2g, and the process yielded enantiopure cyclo-(L)-Aza-Trp-(L)-Pro for each of 2g, 4g, and 6g. Diastereomers of 5g and 7g could not be separated even on RP-HPLC, so these compounds were carried forward as a mixture of diastereomers for enzymatic studies. The isolated yields of Aza-CyWP analogs from their corresponding Aza-Trp ethyl esters ranged 67–88%, and their synthesis and purity were confirmed by HRMS and 1D- and 2D NMR spectroscopy (see Supporting Information).

Screening Aza-Trp and Aza-CyWP Analogs as Substrates for PTs

Three representative indole-alkylating PTs (all from Aspergillus fumigatus)^{[32b, 39]} were selected to assess the ability of Aza-Trp isomers (2f, 4f, 5f, 6f, 7f) and their corresponding Aza-CyWP analogs (2g, 4g, 5g, 6g, 7g) to act as prenyl acceptors. These included FgaPT2, a Trp normal-C4-PT involved in fumigaclavine C biosynthesis;^{[39a, 39b]} CdpNPT, a CyWP reverse-C3-PT implicated in the aszonalenin biosynthetic pathway;[39c, 39d] and FtmPT1, a CyWP normal-C2-PT from the biosynthetic pathway of fumitremorgin B.^[32b] As an initial test, the ten aza-substituted analogs were evaluated with all three PTs in analytical-scale reactions using DMAPP as the prenyl donor. All reactions were conducted under standardized conditions, and appropriate controls were included to distinguish between enzymatic and non-enzymatic transformations. The generation of products was monitored on RP-HPLC using the retention time differences between the substrate and products.

Overall, this initial screen revealed eight positive reactions between the three enzymes: two with FgaPT2, one with CdpNPT, and five with FtmPT1. Consistent with its low prenylation efficiency with CyWP,^[39b] FgaPT2 did not accept any of the Aza-CyWP analogs as substrates and instead exclusively formed products with Aza-Trp analogs. Among these, only 4f and 5f displayed measurable turnover to prenylated products (23.6% and 71% conversion, respectively, at 254 nm; Figure 2), each of which resulted in a single product. Based on an analysis of FgaPT2’s reaction products (discussed later), the reason for its substrate selectivity appeared to be the electronic effects of the differential nitrogen substitutions, particularly in relation to FgaPT2’s known regioselectivity on the indole ring. In any case, the current data has established that Aza-Trp analogs, specifically 4f and 5f, can indeed serve as substrates in PT-catalyzed reactions.

Figure 2.

RP-HPLC traces for analytical-scale reactions of azaindole-containing substrate analogs catalyzed by indole-alkylating PTs. Traces for substrate standards are shown in gray. Traces for reactions catalyzed by FgaPT2, CdpNPT, and FtmPT1 are shown in blue, green, and magenta, respectively. Traces for reactions involving 4f, 5f, 2g, and 6g were recorded at 254 nm. Traces for reactions involving 5g and 7g were recorded at 280 nm. The trace for the reaction involving 4g was recorded at 340 nm. Product peaks are labeled using the names listed in Figure 3 and Table 1. Reactions catalyzed by FgaPT2 and CdpNPT were conducted in a final volume of 20–50 μL and contained 1 mM substrate analog, 1.5 mM DMAPP, 50 mM Tris (pH 7.8), 5 mM MgCl₂, and either 21 μM FgaPT2 or 200 μM CdpNPT. Reactions catalyzed by FtmPT1 were conducted in a final volume of 20 μL and contained 1 mM substrate analog, 1.5 mM DMAPP, 25 mM Tris (pH 7.8), 5 mM CaCl₂, and 45 μM enzyme. Control reactions lacked PTs, but their compositions otherwise mirrored the enzyme-containing reactions. All positive reactions were subsequently confirmed by HRMS.

Similar to FgaPT2’s inability to utilize Aza-CyWP analogs, CdpNPT did not prenylate any of the Aza-Trp analogs, though this outcome was anticipated given the enzyme’s known preferences for larger substrates (CyWP, daptomycin, etc.).^{[28d, 40]} More surprising was CdpNPT’s low promiscuity toward the Aza-CyWP analogs. As shown in Figure 2, only 7g displayed measurable turnover in the presence of CdpNPT, forming two distinct products with a cumulative conversion of 28.7% (at 280 nm). This unexpected selectivity for 7g likely arose from the variations in electron density on CdpNPT’s preferred site of prenylation (C3) caused by differential nitrogen replacement on the indole. Since the C3 electron density of 7g was closer to the native CyWP than the other potential substrates, this analog was most likely to proceed through analogous transformations. The profile itself demonstrated the formation of a major and a minor product (Figure 2), with the major product displaying a dramatically altered absorbance profile compared to the parent 7g (see Figure S106, Supporting Information). Such changes indicated dearomatization of the pyrrole ring likely caused by alkylation at C3, which is consistent with the native reaction of CdpNPT with CyWP.^{[30b, 32c]}

Interestingly, FtmPT1 was far more promiscuous than either FgaPT2 or CdpNPT in the analytical-scale reactions, accepting all five Aza-CyWP analogs as substrates in the presence of DMAPP (Figure 2) while being unable to prenylate any of the Aza-Trp analogs. These observations are consistent with FtmPT1’s known substrate preferences,^{[28a, 32a, 41]} but analysis of the resulting chromatograms revealed a surprising trend in product formation. The native FtmPT1 reaction with CyWP displays strict regioselectivity for C2 resulting in a single product, but exclusive generation of a single prenylated regioisomer was only observed with 2 of 5 reactions involving Aza-CyWP (2g and 6g, Figure 2). The three remaining substrates (4g, 5g, and 7g) produced 2–3 products each (Figure 2C). The percent conversion for each reaction was calculated to be 35% for 2g, 51% for 4g, 44% for 5g, 42% for 6g, and 44% for 7g. Furthermore, the major products of reactions containing 4g, 5g, and 6g had substantially altered absorbance maxima compared to the starting material (see Figure S106, Supporting Information), likely indicating the formation of C3-alkylated dearomatized products (discussed later).^{[32a, 32c]} Overall, the results of the current study have revealed Aza-CyWP analogs to represent an additional substrate class of the aromatic PTs, further establishing the enzyme class’s substrate promiscuity.

Structural Determination of Prenylated Azaindole-substituted Products

To assess the regioselectivity of the enzymatically-generated products, all eight analytical-scale reactions that demonstrated product formation (FgaPT2 + 4f/5f, CdpNPT + 7g, and FtmPT1 + 2g/4g/5g/6g/7g) were scaled up using 5–7 mg of Aza-Trp or Aza-CyWP analog under standard conditions. Following extraction from the aqueous reactions, semi-preparative RP-HPLC was used to purify all reaction products. Structural determination was carried out using 1D ¹H and 2D COSY, HSQC, HMBC, and ROESY NMR spectroscopy in D₂O (Aza-Trp products) or d₆-DMSO (Aza-CyWP products). Assigned structures are shown in Figure 3, and the reactions from which they originated are summarized in Table 1.

Figure 3.

Products of PT-catalyzed reactions involving azaindole-containing substrate analogs as determined by NMR spectroscopy. Isosteric N atoms are shown in blue, while prenyl groups transferred onto the azaindole scaffolds are shown in red. Gray arrows indicate key HMBC correlations confirming the site of prenylation. A superscripted C denotes cyclization of the azaindole’s C2 to the diketopiperazine, while a subscripted R denotes reverse prenylation.

Table 1.

Summary of PT-catalyzed reactions and generated products.

Enzyme	Substrate	Major Product	Minor Product
FgaPT2	4f	N4-4f	-
FgaPT2	5f	N5-5f	-
CdpNPT	7g	C3 C R-7g	N1-7g
FtmPT1	2g	N1-2g	-
FtmPT1	4g	C3 C -4g	N1-4g
FtmPT1	5g	C3 C -5g	-
FtmPT1	6g	C3 C -6g	-
FtmPT1	7g	N1-7g	C2-7g

^Cindicates cyclization of the product at C2 of the azaindole.

^Rindicates reverse prenylation.

As discussed previously, two of the five aza-Trp analogs (4f and 5f) acted as substrates for FgaPT2 in the presence of DMAPP, and each of these reactions generated a single product peak in the analytical-scale reactions. Following scale-up and purification, cumulative NMR analysis of these products revealed the isosteric pyridine N as the site of prenylation in both cases (Figure 3). Substitution of the aromatic CH positions was quickly ruled out by ¹H and HSQC experiments featuring the same number of aromatic CH signals as the non-prenylated Aza-Trp substrates. Meanwhile, the chemical shift values of the prenyl methylene group (C1’/H1’ = 54.5/5.26 ppm for N4-4f and 57.6/5.10 ppm for N5-5f) pointed to attachment at one of the azaindole N atoms. HMBC correlations between H1’ of the prenyl group and the aromatic carbons adjacent to the pyridine nitrogen (Figure 3) unequivocally established the prenylation sites as N4 and N5 on 4f and 5f, respectively. In terms of stereochemistry, the NMR experiments could not differentiate enantiomers, but the products were likely to be enriched in the l-isomer based on the enzymes’ known activities with l- and d- Trp.^[39a] Thus, FgaPT2 was found to catalyze N4- and N5-prenylation with its two Aza-Trp substrates, which is consistent with its known regioselectivity when l-Trp acts as the alkyl acceptor.^[30d] To the best of our knowledge, these reactions are the only known examples of aromatic PTs alkylating a basic aromatic nitrogen atom or forming charged prenylation products.

For the six reactions involving Aza-CyWP analogs (Figure 2), each major product was successfully isolated and characterized, as were the first minor products of CdpNPT + 7g and FtmPT1 + 4g/7g (Figure 3, Table 1). Cumulative NMR analysis revealed three sites of attachment (N1, C2, C3) across the six total reactions, as well as a preference for normal vs. reverse prenylation. For N1-2g/4g/7g, the site of attachment was assigned based on a combination of three factors: a lack of N1 proton signals in all spectra, the chemical shift values of the prenyl methylene group (C1′/H1′ = ~50/~5 ppm), and HMBC correlations between the prenyl H1’ and the azaindole C2/C7a positions. Meanwhile, the single C2-prenylated product (C2-7g) was easily identified by the loss of the associated proton signal in its 1D ¹H spectrum, which largely resembled that of the previously characterized tryprostatin B.^{[28a, 30e]} Perhaps the most interesting spectral changes were observed when C3 was the site of prenylation (C3^C-4g/5g/6g, C3^C_R-7g), as these products also showed signs of cyclization and dearomatization as predicted by their absorbance profiles (see Figure S106, Supporting Information). HMBC correlations between C1′/C3′ and the surrounding carbons of the Aza-CyWP core (C2, C7a, Cβ) clearly indicated C3 as the prenylation site, while cyclization and dearomatization were assigned based on multiple lines of evidence. These included the large upfield shift of the indole C2 ¹³C resonance (~125–145 ppm to ~73–78 ppm), the associated H2 signal shifting from >7 ppm to ~5.3 ppm, the altered C3 chemical shifts from ~108 ppm to ~55 ppm, and the disappearance of the Trp residue’s amide proton in all spectra. Normal vs. reverse prenylation was distinguished by the appearance of a terminal alkene signal in the HSQC spectra (C1’/H1’ = 114.4/5.04–5.11 ppm for C3^C_R-7g) in addition to HMBC correlations, all of which indicated connectivity between the indole C3 and the prenyl C3’ in C3^C_R-7g (Figure 3). Determination of the absolute stereochemistry of the products’ chiral centers was achieved using diagnostic NOE interactions. The (l,l) configuration of the diketopiperazine ring was confirmed by 2D ROESY experiments displaying clear interactions between the α protons of the Aza-Trp and l-Proline. In the case of cyclized products, the syn-cis configuration of the new chiral centers at C2 and C3 was confirmed by observed interactions between H2 and H2’ and the absence of interactions between H2 and the Aza-Trp α proton.

While all major products were successfully scaled up for characterization, not all minor products could be isolated in high enough quantity. Specifically, the two minor products of FtmPT1 + 5g and the second minor product of FtmPT1 + 7g (Figure 2) did not yield enough product following scale-up, even for ¹H NMR.

Mechanistic Discussions of Regioselectivity in PT-catalyzed Reactions

Considered together, the substrate specificity and regioselectivity of PTs with azaindole-containing substrate analogs provide insights into mechanistic changes that occur in response to isosteric N substitution. From a synthetic perspective, the replacement of aromatic CH groups in benzene with N is known to deactivate the resulting pyridine ring towards electrophilic aromatic substitution (i.e., the mechanism of PT-catalyzed reactions). This is because the more electronegative N atom pulls considerable electron density away from the remaining carbon positions. Thus, similar deactivation is expected when aza-substitution occurs at the 4, 5, 6, and 7 positions of indole, forming what are essentially pyrrole-fused pyridine rings. Though 2-azaindole does not contain this specific functional group, deactivation is still expected because the system’s aromaticity allows electron density to be pulled even from the distant carbon positions. Therefore, successful PT-catalyzed alkylation of azaindole-containing substrate analogs are likely the product of the specific enzyme’s native regioselectivity and its entropic (i.e., steric) control of the substrates in the binding pocket.

In the case of FgaPT2, native prenylation occurs at C4 on tryptophan’s indole and is proposed to occur either directly or indirectly following Cope rearrangement of a C3-reverse-prenylated iminium intermediate.^[42] Both mechanisms are likely to be disfavored in the case of Aza-Trp analogs based on the reduced electron density of these carbons (when present) compared to indole.^[20e] Thus, it was somewhat unsurprising that both no activity was observed with 2f, 6f, or 7f and no C-prenylated products were isolated with 4f and 5f. In contrast, the characterization of N4- and N5-prenylated single products from reactions of 4f and 5f (respectively) suggest a potential new mechanism. When C4 and C5 are replaced with N (4f and 5f, respectively), a nucleophilic lone pair is likely positioned near C1′ of the carbocation (~3.5 Å based on the published structure of FgaPT2 bound to l-Trp),^[42-43] such that direct nucleophilic attack is now plausible. Thus, the promiscuity and regioselectivity of FgaPT2 toward Aza-Trp analogs appeared to arise from a change in enzymatic mechanism from electrophilic aromatic substitution to nucleophilic attack by a nitrogenous lone pair.

In contrast, more prominent steric factors likely contributed to the limited substrate promiscuity of CdpNPT. The enzyme itself is known to bind less specifically to CyWP (brevianamide F) than its putative substrate S-benzodiazepinedione, resulting in comparatively lower activity with the parent compound of the current study.^[44] Furthermore, the demonstrably larger hydrophobic acceptor binding pocket of CdpNPT vs. both FgaPT2 and FtmPT1 likely decreased conformational control of the azaindole substrates.^[44] Compared to indole, the significantly enhanced polarity of azaindole substrates would be expected to have significantly less affinity for the large hydrophobic acceptor binding site of CdpNPT. Combined with the increased hydrogen bonding capabilities of azaindoles compared to indole, this movement may have stabilized the Aza-CyWP analogs in conformations not conducive to prenylation. The non-specific nature of CdpNPT’s binding site is a key factor for its remarkable promiscuity towards indole containing substrates, but this non-specificity appears to work against successful prenylation reactions upon introduction of more polar aromatic moieties. Still, CdpNPT’s successful prenylation of 7g indicates there may also be a considerable contribution by the substrate analogs’ electronic properties, as 7-azaindole is known to be the closest electronically to the parent indole ring with respect to the other isomers explored in this study. In terms of regioselectivity, CdpNPT is known to catalyze C3-reverse-prenylation of cyclic dipeptides accompanied by dearomatization. Thus, the major product of its reaction with 7g (C3^C_R-7g) was somewhat expected. Conversely, we were surprised to observe a minor N1-prenylated product (N1-7g), which may indicate a Cope rearrangement occurs in the active site following C3-reverse-prenylation.^[42] A similar but non-enzymatic rearrangement has also been observed after treatment of C3-reverse-prenylated cyclic l-Trp-l-Trp with acid at 37°C.^[32c] Still, CdpNPT’s inability to prenylate the other four Aza-CyWP analogs implied it was not an ideal choice for functionalizing this substrate class.

Meanwhile, FtmPT1 prenylated every Aza-CyWP substrate tested and exhibited differential regioselectivity depending on the location of the aza-substitution. As such, the resulting set of products may provide new mechanistic details for FtmPT1-catalyzed prenylations. In terms of literature precedent, the mechanism by which FtmPT1 selectively prenylates the C2 position of CyWP has been highly disputed. Crystallographic data of the enzyme bound to CyWP and a DMAPP mimetic, dimethylallyl S-thiolodiphosphate, indicated C3` of the prenyl group was positioned over C3 of the pyrrole moiety, suggesting a possible C3-reverse-prenylation and subsequent rearrangement to the C2 position.^{[32c, 43]} This was further supported by studies involving functionalized CyWP derivatives (5-hydroxybrevianamide F, 2-methylbrevianamide F), though the authors also proposed a potential second mechanism involving C3-normal-prenylation followed by a 1,2-alkyl shift to C2.^[32c] The latter mechanism was based on the isolation of C3-normal-prenylated products in reactions involving 2-methylbrevianamide F,^[32c] though similar regioselectivity has also been observed with FtmPT1-catalyzed prenylations of non-Pro-containing IDPs.^[32a] Alternatively, a third mechanism has been proposed in which the prenyl group rotates within the active site after pyrophosphate dissociation, allowing for direct nucleophilic attack by the C2 position.^[43] This mechanism has been supported by kinetic isotope experiments that reveal two partially rate-limiting steps in the reaction (phosphate ester hydrolysis, and C-C bond formation),^[42] both of which occur on a timescale orders of magnitude slower than carbocation rotation.^[42-43]

As they pertain to the current study, carbocation rotation appeared to be the most relevant based on the product profiles of the Aza-CyWP analogs (Figure 3, Table 1). The strongest evidence for this was the lack of C3-reverse-prenylation among the scaled-up products, which would be expected based on previous structural data.^[43] For 2g, carbocation rotation and subsequent attack by N1’s lone pair is the most plausible mechanism, based on significantly lower C3 electron density compared to indole and the other pyrrole-containing azaindole isomers. Additionally, N2-alkylation of indazole is highly disfavored because of the resulting loss of aromaticity,^[45] leaving the nucleophilic N1 position available to form N1-2g as the exclusive product. In the case of 7g, being the only Aza-CyWP isomer to undergo the native C2-prenylation pointed to a potential retention of mechanism, especially when considering the more similar electronic distribution between indole and 7-azaindole compared to the other isomers used in this study.^[32b]

In any case, conclusions we could offer on mechanistic details are limited by our utilization of non-natural substrates. In addition to electronic differences between CyWP and the Aza-CyWP isomers, the introduction of nitrogenous lone pairs provides new opportunities for hydrophilic interactions, which could significantly alter binding in the FtmPT1 active site. This is further bolstered by the results of previous studies using non-natural substrates, which also could not be unified into a single mechanism.^{[32c, 43]} Still, the prenylation of every azaindole examined here demonstrates wide substrate tolerance of PTs, though the enzyme’s activity, as denoted by the observation of differential regioselectivity, is influenced by isosteric substitution.

Cytotoxicity Studies of Prenylated Aza-CyWP Analogs

Given the established cytotoxic activity of tryprostatins A and B ^[1-2] towards human cancer cell lines, we tested the cytotoxicity of the prenylated Aza-CyWP analogs against human leukemia K562 cells. However, the results of the cell titer-blue viability assays^[30e] revealed that none of the compounds exhibited cytotoxic effects, even at a concentration of 100 μM. This is most likely due to the decreased lipophilicity of the analogs, which would decrease their ability to enter the cells. In any case, generating these analogs at all using PT-catalyzed reactions was a significant step in accessing azaindole-containing bioisosteres.

Conclusion

The current study has demonstrated that the indole PTs FgaPT2, CdpNPT, and FtmPT1 can all tolerate isosteric replacement of the acceptor indole moiety by azaindoles. Furthermore, the position of N-substitution within the indole ring was shown to heavily influence the reactivity of the azaindole-containing substrates and the regioselectivity of each prenylation. Notably, in the case of FgaPT2 with 4f and 5f, N-prenylation of a pyridine ring was observed for the first time in a PT-catalyzed reaction. Selective N-alkylation of this moiety in the presence of an unprotected primary amine without the use of protecting groups is not possible using modern synthetic methods, so the current study is an excellent illustration of the advantages of PT-based biocatalytic transformations. The diverse reactions observed in this study will also help guide future substrate and protein engineering efforts to improve the synthetic utility of PTs in drug discovery.

Experimental Section

General Materials:

Unless otherwise stated, all chemicals and reagents were purchased from Acros Organics (Geel, Antwerp, Belgium), Sigma-Aldrich (St. Louis, MO, USA), Ambeed (Arlington Heights, IL, USA), or AK Scientific (Union City, CA, USA) and were reagent grade or better. The Ni-NTA columns and PD-10 columns used for protein purification were purchased from GE Healthcare (Chicago, IL, USA).

General Methods:

All reported reactions were conducted in oven-dried glassware under an anhydrous nitrogen atmosphere with anhydrous solvents unless otherwise noted. Column chromatography purification was performed using silica gel (SiliCycle Inc, P60, particle size 40–63 μm). HPLC analysis and purification was conducted using an Agilent 1220 system equipped with a diode array detector (DAD). HRMS confirmation of products and substrates was conducted on an Agilent 6545-QTOF W/1290 HPLC mass spectrometer at the University of Oklahoma Department of Chemistry and Biochemistry. NMR spectra were collected on a Varian VNMRS 500 MHz, a Varian INOVA 600 MHz, a Varian Mercury VX-300, or a Varian VNMRS 400 MHz NMR spectrometer at the NMR facility of the Department of Chemistry and Biochemistry of the University of Oklahoma. All NMR spectra were obtained in 400–600 μL volumes of Methanol-d₄ (99.8%), DMSO-d₆ (99.9%), CDCl₃ (99.8%), or D₂O (99.9%) (Cambridge Isotope Laboratories, MA, USA), and processed using MestReNova (Version 12.0.3 and 14.2.1, Mestrelab Research, S.L., Compostela, Spain). Structural elucidation and NMR peak assignment of chemoenzymatically prenylated products was performed using 2D-NMR experiments (¹H-COSY, ¹H-¹³C HSQC, ¹H-¹³C HMBC or ROESY).

Aza-gramine Synthesis (4b, 5b, 6b, 7b):

To a mixture of dimethylamine (40% by weight in H₂O, 0.7 mL, 5.51 mmol), acetic acid (0.279 mL, 4.88 mmol), and isopropanol (1 mL) cooled to 0°C, 37% aqueous formaldehyde (0.363 mL, 4.88 mmol) were added and stirred at 0°C for 30 min. Azaindole (4a, 5a, 6a, or 7a) (500 mg, 4.24 mmol) in isopropanol (2 mL) was added to the reaction mixture and continued stirring at 0°C for another 30 min. The resulting solution was heated to 90°C overnight, allowed to cool down to room temperature, diluted with 10 mL water, and basified with 5 M NaOH. The aqueous solution was extracted with CHCl₃:Isopropanol (3:1; 3 x 10 mL), and the combined organic layers were dried over anhydrous Na₂SO₄, filtered, and concentrated in vacuo to yield a light-yellow oil which slowly crystalized upon standing at room temperature. If contaminates were detected via NMR, purification via column chromatography was carried out using CHCl₃:MeOH:NH₄OH (90:10:1 followed by 40:10:1). Isolated yields of aza-gramines were >85%.

N,N-Dimethyl-1-(1H-pyrrolo[3,2-b]pyridin-3-yl)methanamine (4b):

Isolated as a white crystalline solid using the procedure described above. ¹H NMR (300 MHz, Methanol-d₄) δ 8.29 (dd, J = 4.8, 1.5 Hz, 1H), 7.79 (dd, J = 8.2, 1.5 Hz, 1H), 7.54 (s, 1H), 7.14 (dd, J = 8.2, 4.8 Hz, 1H), 3.79 (s, 2H), 2.27 (s, 6H).

N,N-dimethyl-1-(1H-pyrrolo[3,2-c]pyridin-3-yl)methanamine (5b):

Isolated as a viscous clear oil using the procedure described above. ¹H NMR (400 MHz, Methanol-d₄) δ 8.91 (s, 0H), 8.16 (d, J = 5.8 Hz, 1H), 7.42 (d, J = 5.8 Hz, 1H), 7.37 (s, 1H), 3.76 (s, 2H), 2.32 (d, J = 1.1 Hz, 6H).

N,N-dimethyl-1-(1H-pyrrolo[2,3-c]pyridin-3-yl)methanamine (6b):

Isolated as a white crystalline solid using the procedure described above. ¹H NMR (300 MHz, Methanol-d₄) δ 10.07 – 9.96 (m, 1H), 9.40 (d, J = 5.5 Hz, 1H), 8.92 – 8.83 (m, 1H), 8.73 (s, 1H), 3.47 (s, 6H).

N,N-dimethyl-1-(1H-pyrrolo[2,3-b]pyridin-3-yl)methanamine (7b):

Isolated as a white crystalline solid using the procedure described above. ¹H NMR (500 MHz, Methanol-d₄) δ 8.19 (dd, J = 4.8, 1.5 Hz, 1H), 8.10 (dd, J = 7.9, 1.5 Hz, 1H), 7.37 (s, 1H), 7.12 (dd, J = 7.9, 4.8 Hz, 1H), 3.68 (s, 2H), 2.28 (s, 6H).

Aza-gramine Methylation (5c, 6c, 7c):

A 0.15 M solution of aza-gramine (5b, 6b, 7b; 420 mg, 2.4 mmol) was prepared in anhydrous THF under a nitrogen atmosphere and cooled to 0°C. The reaction mixture was treated with methyl iodide (1.5 eq, 510 mg) and stirred for 5 minutes at 0°C before warming to room temperature. After stirring at room temperature for 1 h, the mixture was diluted to triple its initial volume using pentane. The resulting suspension was filtered, washed with pentane, and taken forward without purification. The isolated yields of 5-aza-gramine methiodide (5c) and 6-aza-gramine methiodide (6c) were 72.2% and 57.2 %, respectively, and the yield for 7-aza-gramine methiodide (7c) was nearly quantitative.

3-Methyl-indazole Boc Protection (2b):

Based on a previously reported procedure,^[37] a solution of 3-methyl-indazole (600 mg, 4.54 mmol) in 20 mL anhydrous acetonitrile was prepared, cooled to 0°C, and treated with 4-dimethylaminopyridine (1.11 g, 9.08 mmol), triethylamine (1.27 mL, 9.08 mmol), and (Boc)₂O (1.98 g, 9.08 mmol). The reaction mixture was stirred for 1 h at 0°C before being allowed to warm to room temperature. After 2 h, TLC indicated consumption of starting material, and the solvent was removed in vacuo. The crude residue was dissolved in ethyl acetate (EtOAc) and washed with H₂O. The aqueous layer was extracted twice with EtOAc, and the combined organic layers were dried over anhydrous Na₂SO₄, filtered, and concentrated in vacuo. The crude product was purified by flash column chromatography on silica gel using 25% EtOAc in cyclohexane. Pure product was recovered as a pale orange oil (1.03 g, 97%), and its NMR spectra matched those reported previously.^[37]

Synthesis of 3-(Bromomethyl)-1-(tert-butoxycarbonyl)indazole (2c):

A solution was prepared containing Boc-protected 3-methyl-indazole (2b, 716 mg, 3.08 mmol) and N-bromosuccinamide (658 mg, 3.70 mmol) in 16 mL anhydrous CCl₄. The reaction mixture was then treated with benzoyl peroxide (74 mg, 0.31 mmol) and heated at reflux in an 80°C oil bath for 4 h. The mixture was cooled to room temperature, filtered through celite, and concentrated in vacuo. The resulting residue was purified via silica gel chromatography using a gradient from 100% hexane to 10% EtOAc in hexane. Unreacted starting material was successfully recovered during purification. The NMR spectra of the isolated product matched those reported previously.^[37]

Synthesis of tert-Butyl 3-(2-((diphenylmethylene)amino)-3-ethoxy-3-oxopropyl)-1H-indazole-1-carboxylate (2d):

A solution containing NaH (23.1 mg, 0.964 mmol, 60% suspension in mineral oil) in 6 mL anhydrous THF under dry nitrogen was treated with ethyl N-(diphenylmethylene)glycinate (257.7 mg, 0.964 mmol) and stirred at 0°C for 30 minutes. A solution of 2c in 2 mL THF was added to this mixture, and it continued stirring for 3 h at 0°C. The reaction was quenched with saturated aqueous NH₄Cl and extracted with EtOAc (3 x 10 mL), and the combined organic layers were dried over anhydrous Na₂SO₄, filtered, and concentrated in vacuo. The resulting residue was purified using flash column chromatography on silica gel using a gradient of 25-50% DCM in hexanes, resulting in 200 mg (62.5%) of pure 2d after removal of the solvent in vacuo. ¹H NMR (600 MHz, CDCl₃) δ 8.05 (d, J = 8.4 Hz, 1H), 7.61 (d, J = 8.0 Hz, 1H), 7.53 – 7.49 (m, 2H), 7.45 – 7.41 (m, 2H), 7.33 (dt, J = 13.8, 7.4 Hz, 2H), 7.28 – 7.23 (m, 2H), 7.14 (t, J = 7.5 Hz, 1H), 6.73 (d, J = 7.4 Hz, 2H), 4.63 (dd, J = 8.9, 4.6 Hz, 1H), 4.26 – 4.12 (m, 2H), 3.65 (dd, J = 14.3, 4.6 Hz, 1H), 3.61 (dd, J = 14.3, 8.9 Hz, 1H), 1.68 (s, 9H), 1.25 (t, J = 7.1 Hz, 3H).

Synthesis of Ethyl 2-((diphenylmethylene)amino)-3-(1H-pyrrolo[3,2-c]pyridin-3-yl)propanoate (5d):

Ethyl N-(diphenylmethylene)glycinate (168 mg, 0.63 mmol) was dissolved in 1.5 mL dry DMF and cooled to 0°C. NaH (15 mg, 0.63 mmol, 60% suspension in mineral oil) was added, the solution was stirred for 15 min, and half of the 5-aza-gramine methiodide (5c) was then added as a suspension in 1.5 mL DMF. After stirring at 0°C for 1 h, the remaining 5c was added as a powder in one portion. The reaction mixture was stirred at 0°C for 6 hours and subsequently left at 4°C overnight without stirring in a refrigerator. The reaction was quenched with saturated aqueous sodium bicarbonate and extracted 3 times with EtOAc, and the combined organic layers were dried over anhydrous Na₂SO₄, filtered, and concentrated in vacuo. Purification was achieved via silica gel chromatography with a gradient from 25–100% EtOAc in hexanes, followed by a mixture of CHCl₃:MeOH:NH₄OH (90:10:1) to elute the product. Azaindole-containing fractions were identified by their bright blue fluorescence under a UV lamp (365 nm), and the product 5d was collected as a brown viscous liquid in a final yield of 75 mg (29.9%). ¹H NMR (400 MHz, CDCl₃) δ 8.53 (s, 1H), 8.23 (d, J = 5.8 Hz, 1H), 7.61 – 7.56 (m, 3H), 7.33 (dd, J = 21.7, 7.3 Hz, 4H), 7.22 (q, J = 7.5, 6.4 Hz, 3H), 7.04 (s, 1H), 6.68 (d, J = 7.2 Hz, 2H), 4.40 (dd, J = 8.3, 5.1 Hz, 1H), 4.24 – 4.10 (m, 2H), 3.47 (dd, J = 14.3, 5.1 Hz, 1H), 3.28 (dd, J = 14.3, 8.3 Hz, 1H), 1.22 (t, J = 7.1 Hz, 3H).

Synthesis of Ethyl 2-((diphenylmethylene)amino)-3-(1H-pyrrolo[2,3-c]pyridin-3-yl)propanoate (6d):

The benzophenone-protected 6-aza-tryptophan intermediate 6d was prepared using the above synthetic method for 5d. The product 6d was isolated as a viscous brown-red oil with yields ranging from 16–25%. ¹H NMR (500 MHz, CDCl₃) δ 8.81 (s, 1H), 8.06 (d, J = 5.6 Hz, 1H), 7.60 (dd, J = 8.3, 1.3 Hz, 2H), 7.42 – 7.37 (m, 1H), 7.35 – 7.29 (m, 3H), 7.27 – 7.25 (m, 2H), 7.22 (t, J = 7.8 Hz, 2H), 6.67 (d, J = 7.4 Hz, 2H), 4.40 (dd, J = 8.4, 4.8 Hz, 1H), 4.20 (p, J = 7.1 Hz, 2H), 3.45 (dd, J = 14.4, 4.7 Hz, 1H), 3.31 (dd, J = 14.4, 8.5 Hz, 1H), 1.26 (t, J = 7.1 Hz, 3H).

Synthesis of Ethyl 2-((diphenylmethylene)amino)-3-(1H-pyrrolo[2,3-b]pyridin-3-yl)propanoate (7d):

A solution containing NaH (490 mg, 20.43 mmol, 60% suspension in mineral oil) in 100 mL anhydrous THF under dry nitrogen was treated with ethyl N-(diphenylmethylene)glycinate (5.46 g, 20.43 mmol) and stirred at 0°C for 30 min. A suspension of 7-aza-gramine methiodide (7c, 5.4 g, 17.03 mmol) in 50 mL anhydrous THF was added, and the mixture was stirred for 6 h at 0°C before being left overnight at 4°C. The reaction was quenched with a saturated solution of NaHCO₃ and extracted 3 times with EtOAc, and the combined organic layers were dried over anhydrous Na₂SO₄, filtered, and concentrated in vacuo. Purification was achieved via silica gel chromatography with a gradient from 25–100% EtOAc in hexanes, yielding 3.66 g (54%) of clear brown viscous liquid. The NMR spectra of the isolated product matched those reported previously.^[35]

Benzophenone imine hydrolysis (2e, 5e, 6e, 7e)

Ethyl 2-Amino-3-(1H-indazol-3-yl)propanoate (2e):

The protected 2-aza-tryptophan ester 2d (200 mg, 0.4 mmol) was dissolved in THF to make a 0.1 M solution, which was then treated with 5 M HCl (10 eq) and stirred at room temperature overnight. The resulting mixture was diluted with EtOAc, neutralized with saturated NaHCO₃, and extracted 3 times with EtOAc to yield 63 mg (47%) of the Boc-protected 2e after purification via silica column chromatography. Quantitative deprotection of the N-Boc functionality was achieved by dissolving 20 mg of the purified material in 1 mL anhydrous DCM and adding 2 mL TFA dropwise at 0°C. The reaction was allowed to warm to room temperature and was then stirred overnight. The solvent was removed via nitrogen flow, and the resulting residue was dissolved in aqueous K₂CO₃ and extracted 3 times with EtOAc. The combined organic layers were dried over Na₂SO₄, filtered, and concentrated in vacuo to yield pure 2-aza-tryptophan ethyl ester (2e). ¹H NMR (400 MHz, CDCl₃) δ 7.66 (d, J = 8.1 Hz, 1H), 7.41 (d, J = 8.4 Hz, 1H), 7.37 – 7.29 (m, 1H), 7.16 – 7.09 (m, 1H), 4.12 (qd, J = 7.1, 5.1 Hz, 2H), 4.05 (s, 1H), 3.47 (dd, J = 14.9, 4.6 Hz, 1H), 3.34 (dd, J = 14.9, 7.6 Hz, 1H), 1.17 (t, J = 7.1 Hz, 3H).

Ethyl 2-Amino-3-(1H-pyrrolo[2,3-c]pyridin-3-yl)propanoate (6e):

The protected 6-aza-tryptophan ester 6d (36 mg, 0.09 mmol) was dissolved in THF to make a 0.1 M solution, which was then treated with 2 M HCl (3.5 eq) and stirred at room temperature. The mixture was quenched with saturated NaHCO₃ and washed twice with Et₂O to remove the benzophenone. The aqueous layer was further basified with excess K₂CO₃, extracted 3 times with EtOAc, and extracted again with a 3:1 mixture of CHCl₃/isopropanol. The combined organic layers were dried over NaSO₄, filtered, and concentrated in vacuo, and the crude product was purified via silica gel column chromatography starting with 100% EtOAc, followed by a gradient of CHCl₃:MeOH:NH₄OH (90:10:1 → 40:10:1) to elute the product. Fluorescent fractions with high relative polarity were combined to provide 17 mg pure 6-aza-tryprophan ethyl ester 6e (80% yield) as a yellow oil. ¹H NMR (500 MHz, CDCl₃) δ 8.78 (s, 1H), 8.23 (d, J = 5.6 Hz, 1H), 7.55 (d, J = 5.5 Hz, 1H), 7.30 (s, 1H), 4.14 (qd, J = 7.1, 4.8 Hz, 2H), 3.80 (dd, J = 7.3, 5.2 Hz, 1H), 3.24 (dd, J = 14.5, 5.2 Hz, 1H), 3.09 (dd, J = 14.5, 7.3 Hz, 1H), 1.22 (t, J = 7.1 Hz, 3H).

Synthesis of Ethyl 2-Amino-3-(1H-pyrrolo[3,2-c]pyridin-3-yl)propanoate (5e):

The title compound was obtained using the conditions listed above (Synthesis of 6e) with comparable yields. Due to challenges in obtaining highly pure 5d, the crude reaction mixture containing 5d was taken forward using the above method, enabling simplified purification of the highly polar 5e. ¹H NMR (600 MHz, Methanol-d₄) δ 8.76 (s, 1H), 8.10 (d, J = 5.9 Hz, 1H), 7.36 (d, J = 5.8 Hz, 1H), 7.22 (s, 1H), 4.04 (t, J = 7.1 Hz, 2H), 3.76 (t, J = 6.3 Hz, 1H), 3.18 (d, J = 6.3 Hz, 2H), 1.10 (t, J = 7.2 Hz, 3H).

Synthesis of Ethyl 2-Amino-3-(1H-pyrrolo[2,3-b]pyridin-3-yl)propanoate (7e):

The benzophenone-protected 7-aza-tryptophan ester (7d) (3.66 g, 9.21 mmol) was hydrolyzed using a previously reported procedure.^[35] The desired product was obtained as a transparent brown oil (1.54 g, 71% yield). ¹H NMR (400 MHz, CDCl₃) δ 9.66 (s, 1H), 8.31 (d, J = 4.9 Hz, 1H), 7.95 (d, J = 7.8 Hz, 1H), 7.21 (s, 1H), 7.09 (dd, J = 7.9, 4.7 Hz, 1H), 4.14 (qd, J = 7.2, 2.2 Hz, 2H), 3.79 (dd, J = 7.3, 5.2 Hz, 1H), 3.23 (dd, J = 14.4, 5.2 Hz, 1H), 3.07 (dd, J = 14.4, 7.3 Hz, 1H), 1.22 (t, J = 7.2 Hz, 3H).

Synthesis of Ethyl 2-Amino-3-(1H-pyrrolo[3,2-b]pyridin-3-yl)propanoate (4e):

To a flask containing 7 mL xylenes, 4-aza-gramine (4b, 700 mg, 4 mmol), diethyl acetamidomalonate (868.9 mg, 4 mmol), and NaOH (50 mg, 1.25 mmol) were added and refluxed under nitrogen for 5 h. The hot mixture was then filtered and cooled, and the resulting precipitate was filtered, washed with hexane, and dried in vacuo to give a pure intermediate product. The residue was dissolved in 15.6 mL EtOH and treated with 1.56 mL water containing KOH (281mg, 5 mmol). After stirring at room temperature for 3 h, the reaction was acidified with HCl and concentrated in vacuo. Water (7 mL) was added to the reaction mixture, the pH was adjusted to 1 with conc. HCl, and the solvent was removed using a stream of nitrogen. The solid residue was heated at 130°C under nitrogen for 1 h to decarboxylate the intermediate. The residue was then dissolved in 20% aqueous HCl and heated at reflux for 22 hours. After cooling, the reaction mixture was basified with K₂CO₃ and washed with CHCl₃:isopropanol (3:1). The mother liquor was then saturated with K₂CO₃ and extracted with isopropanol (3 x 10 mL). The solvent was removed in vacuo to yield a tan/white solid 4f. Additional product was obtained by acidification of the mother liquor with conc. HCl, and subsequent removal of the solvent with low heat and nitrogen flow to produce a salt cake. The salt cake was extracted with boiling EtOH (3 x 20 mL), and the combined EtOH layers were evaporated to give a residue that was re crystallized by dissolution in 5 mL water, basifiied with 4 drops of 30% aqueous ammonia, and then diluted with 5 mL acetone. The mixture was left overnight at 4°C and the solid precipitate was recovered as pure 4f. The free 4f was converted into the ethyl ester 4e by refluxing thionyl chloride in EtOH overnight, resulting in quantitative yields of 4e.

Synthesis of Aza-Trp isomers (2f, 5f, 6f, and 7f) from Aza-Trp Ethyl Esters:

Saponification was achieved by first dissolving 20 mg of the Aza-Trp ethyl ester (2e, 5e, 6e, 7e) in 4 mL ethanol. The resulting solution was then basified through the addition of 2 mL 1 M NaOH, followed by overnight stirring at 40°C in a sealed vial. The solution was then concentrated to ~2 mL via nitrogen flow and combined with 2 mL saturated NaHCO₃. Due to the extreme polarity of the products, traditional extraction solvents were ineffective, so the aqueous solution was saturated with K₂CO₃ and extracted with isopropanol (3 x 10 mL). These layers were then combined, dried over anhydrous K₂CO₃, and concentrated, and the resulting residues were purified by semi-preparative RP-HPLC (Method 1, described below) to ensure minimal salt contamination. NMR characterization of the resulting products (1D and 2D) was performed with samples dissolved in D₂O to a final concentration of 5 mM.

2-Amino-3-(1H-indazol-3-yl)propanoic Acid (2f):

¹H NMR (500 MHz, D₂O) δ 7.87 (d, J = 8.2 Hz, 1H), 7.61 (d, J = 8.6 Hz, 1H), 7.50 (t, J = 7.4 Hz, 1H), 7.25 (t, J = 7.5 Hz, 1H), 3.90 (dd, J = 7.5, 5.2 Hz, 1H), 3.49 (dd, J = 14.9, 5.0 Hz, 1H), 3.36 (dd, J = 14.9, 8.0 Hz, 1H). HRMS-ESI: Calc for C₁₀H₁₂N₃O₂ [M+H]⁺: 206.09295; Found: 206.0920.

2-Amino-3-(1H-pyrrolo[3,2-b]pyridin-3-yl)propanoic Acid (4f):

¹H NMR (500 MHz, D₂O) δ 8.47 (dd, J = 8.3, 1.1 Hz, 1H), 8.41 (dd, J = 5.9, 1.1 Hz, 1H), 7.95 (s, 1H), 7.59 (dd, J = 8.3, 5.9 Hz, 1H), 4.20 (dd, J = 7.2, 5.8 Hz, 1H), 3.54 – 3.40 (m, 2H). HRMS-ESI: Calc for C₁₀H₁₂N₃O₂ [M+H]⁺: 206.09295; Found: 206.0927.

2-Amino-3-(1H-pyrrolo[3,2-c]pyridin-3-yl)propanoic Acid (5f):

¹H NMR (500 MHz, D₂O) δ 9.05 (s, 1H), 8.46 (s, 1H), 8.28 (d, J = 6.7 Hz, 1H), 7.88 (d, J = 6.7 Hz, 1H), 7.70 (s, 1H), 4.08 (t, J = 5.9 Hz, 1H), 3.57 – 3.46 (m, 2H). HRMS-ESI: Calc for C₁₀H₁₂N₃O₂ [M+H]⁺: 206.09295; Found: 206.0927.

2-Amino-3-(1H-pyrrolo[2,3-c]pyridin-3-yl)propanoic Acid (6f):

¹H NMR (500 MHz, D₂O) δ 8.95 (s, 1H), 8.46 (s, 1H), 8.18 (d, J = 6.5 Hz, 1H), 8.09 (d, J = 6.5 Hz, 1H), 8.03 (s, 1H), 4.08 (t, J = 6.1 Hz, 1H), 3.50 (dd, J = 16.8, 6.1 Hz, 2H). HRMS-ESI: Calc for C₁₀H₁₂N₃O₂ [M+H]⁺: 206.09295; Found: 206.0927.

2-Amino-3-(1H-pyrrolo[2,3-b]pyridin-3-yl)propanoic Acid (7f):

¹H NMR (500 MHz, D₂O) δ 8.73 (dd, J = 8.0, 1.2 Hz, 1H), 8.40 (dd, J = 6.0, 1.2 Hz, 1H), 7.66 (s, 1H), 7.60 (dd, J = 8.0, 6.0 Hz, 1H), 4.23 (t, J = 6.1 Hz, 1H), 3.52 (dd, J = 14.3, 6.2 Hz, 2H). HRMS-ESI: Calc for C₁₀H₁₂N₃O₂ [M+H]⁺: 206.09295; Found: 206.0927

Synthesis of Aza-CyWP Analogs (2g, 4g, 5g, 6g, 7g):

Fmoc-l-proline (25.5 mg, 0.135 mmol) was added to 3 mL EtOH containing Aza-Trp ethyl ester (2e, 4e, 5e, 6e, 7e; 30 mg, 0.128 mmol), and the mixture was cooled to 0°C in an open flask. DMT-MM (39.2 mg, 0.142 mmol) was then added in a single portion, and the reaction mixture was stirred for 2 h at 0°C before removal of the solvent in vacuo. Morpholine (4 mL) was added to the residue and stirred at room temperature for 7 days. Solvent was then removed by nitrogen flow, and the crude residue was redissolved in aqueous K₂CO₃. The mother liquor was extracted with a 3:1 mixture of CHCl₃:isopropanol (3 x 10 mL), and the combined organic layers were dried over anhydrous K₂CO₃, filtered, and concentrated in vacuo. The crude product was purified via silica gel column chromatography starting with 100% EtOAc, followed by a gradient of CHCl₃:MeOH:NH₄OH (90:10:1 → 40:10:1) to elute the product. RP-HPLC purification was then carried out using a Gemini-NX C18 (5 μm, 10 mm x 250 mm) column (Phenomenex, Torrance, CA, USA) and the methods specified below.

(8aS)-3-((1H-Indazol-3-yl)methyl)hexahydropyrrolo[1,2-a]pyrazine-1,4-dione (2g):

Semi-preparative Method 2 (described below) was used to purify 2g as a single diastereomer (l,l). ¹H NMR (500 MHz, CDCl₃) δ 7.75 (d, J = 8.1 Hz, 1H), 7.48 (d, J = 8.4 Hz, 1H), 7.40 (t, J = 7.6 Hz, 1H), 7.27 (s, 1H), 7.17 (t, J = 7.5 Hz, 1H), 4.46 (dt, J = 10.5, 2.5 Hz, 1H), 4.11 (t, J = 7.7 Hz, 1H), 3.93 (dd, J = 15.7, 3.3 Hz, 1H), 3.74 – 3.55 (m, 4H), 3.25 (dd, J = 15.7, 10.4 Hz, 1H), 2.42 – 2.28 (m, 1H), 2.13 – 1.99 (m, 4H), 1.98 – 1.84 (m, 1H), 1.26 (t, J = 5.0 Hz, 2H). HRMS-ESI: Calc for C₁₅H₁₆N₄O₂ [M+H]⁺: 285.135151; Found: 285.1350.

(8aS)-3-((1H-Pyrrolo[3,2-b]pyridin-3-yl)methyl)hexahydropyrrolo[1,2-a]pyrazine-1,4-dione (4g):

Semi-preparative Method 3 (described below) was used to purify 4g as a single diastereomer (l,l). ¹H NMR (500 MHz, d₆-DMSO) δ 11.23 (s, 1H), 8.88 (s, 1H), 8.32 (dd, J = 4.6, 1.4 Hz, 1H), 7.79 (dd, J = 8.2, 1.4 Hz, 1H), 7.58 (s, 1H), 7.14 (dd, J = 8.2, 4.6 Hz, 1H), 4.27 (dt, J = 7.4, 2.0 Hz, 1H), 4.16 (t, J = 7.7 Hz, 1H), 3.51 (dd, J = 15.1, 3.3 Hz, 1H), 3.45 – 3.36 (m, 2H), 2.95 (dd, J = 14.9, 8.8 Hz, 1H), 2.10 (dt, J = 11.1, 6.5 Hz, 1H), 1.92 – 1.72 (m, 3H). HRMS-ESI: Calc for C₁₅H₁₆N₄O₂ [M+H]⁺: 285.135151; Found: 285.1349.

(8aS)-3-((1H-Pyrrolo[3,2-c]pyridin-3-yl)methyl)hexahydropyrrolo[1,2-a]pyrazine-1,4-dione (5g):

Semi-preparative Method 4 (described below) was used to purify 5g as a 1:1 mixture of diastereomers (l,l:d,l). ¹H NMR (600 MHz, Methanol-d₄) δ 9.17 (s, 1H), 9.00 (s, 1H), 8.44 (s, 3H), 8.28 (d, J = 6.6 Hz, 2H), 7.82 (d, J = 6.7 Hz, 1H), 7.78 (d, J = 6.5 Hz, 1H), 7.64 (s, 1H), 7.57 (s, 1H), 4.57 (s, 1H), 4.24 (t, J = 5.5 Hz, 1H), 4.19 – 4.13 (m, 1H), 3.61 – 3.52 (m, 2H), 3.47 (dd, J = 14.8, 5.9 Hz, 1H), 3.42 – 3.32 (m, 3H), 3.29 (d, J = 4.8 Hz, 1H), 3.22 (t, J = 10.9 Hz, 1H), 3.12 (dd, J = 11.1, 6.5 Hz, 1H), 2.16 (dt, J = 13.0, 6.8 Hz, 1H), 2.11 (dt, J = 12.8, 6.6 Hz, 1H), 1.91 (q, J = 8.7 Hz, 1H), 1.81 (p, J = 9.3 Hz, 1H), 1.77 – 1.68 (m, 2H), 1.61 (p, J = 10.0 Hz, 1H), 1.43 – 1.33 (m, 1H). HRMS-ESI: Calc for C₁₅H₁₆N₄O₂ [M+H]⁺: 285.135151; Found: 285.1355.

(8aS)-3-((1H-Pyrrolo[2,3-c]pyridin-3-yl)methyl)hexahydropyrrolo[1,2-a]pyrazine-1,4-dione (6g):

Semi-preparative Method 5 (described below) was used to purify 6g as a single diastereomer (l,l). ¹H NMR (600 MHz, d₆-DMSO) δ 11.36 (s, 1H), 8.67 (s, 1H), 8.05 (d, J = 5.5 Hz, 1H), 7.96 (s, 1H), 7.57 (d, J = 5.5 Hz, 1H), 7.39 (d, J = 2.3 Hz, 1H), 4.33 (t, J = 5.1 Hz, 1H), 4.05 (t, J = 8.0 Hz, 1H), 3.77 (dd, J = 8.4, 5.7 Hz, 1H), 3.24 (ddd, J = 12.0, 9.1, 3.8 Hz, 1H), 3.21 – 3.11 (m, 2H), 2.96 (dt, J = 11.3, 6.0 Hz, 1H), 2.64 – 2.57 (m, 1H), 2.01 – 1.89 (m, 2H), 1.74 – 1.50 (m, 3H), 1.30 (dt, J = 20.0, 10.5 Hz, 1H). HRMS-ESI: Calc for C₁₅H₁₆N₄O₂ [M+H]⁺: 285.135151; Found: 285.1348.

(8aS)-3-((1H-Pyrrolo[2,3-b]pyridin-3-yl)methyl)hexahydropyrrolo[1,2-a]pyrazine-1,4-dione (7g):

Semi-preparative Method 6 (described below) was used to purify 7g as a mixture of diastereomers. The 7g obtained after chromatographic separation was determined to be a 2:1 mixture of diastereomers (l,l:d,l) as shown by NMR analysis. ¹H NMR (500 MHz, d₆-DMSO) (l, l-Diastereomer Peaks): δ 11.33 (s, 1H), 8.17 – 8.10 (m, 1H), 7.96 (dd, J = 7.9, 1.6 Hz, 1H), 7.88 (s, 1H), 7.23 (d, J = 2.4 Hz, 1H), 6.98 (td, J = 7.8, 4.6 Hz, 1H), 4.30 (t, J = 5.0 Hz, 1H), 4.01 (dd, J = 9.8, 7.1 Hz, 1H), 3.38 – 3.25 (m, 1H), 3.20 (ddd, J = 13.1, 9.3, 4.2 Hz, 1H), 3.16 – 3.05 (m, 2H), 1.93 (dtd, J = 12.2, 6.8, 2.7 Hz, 1H), 1.74 – 1.46 (m, 2H), 1.27 (ddt, J = 12.2, 10.4, 5.2 Hz, 1H). HRMS-ESI: Calc for C₁₅H₁₆N₄O₂ [M+H]⁺: 285.135151; Found: 285.1341.

Overexpression and Purification of PTs:

The recombinant PTs (FgaPT2, FtmPT1, and CdpNPT) were overproduced in Escherichia coli Rosetta2 cells transformed with pET28a vectors containing codon-optimized synthetic genes. All three N-HiS₆-fusion proteins were purified via Ni-NTA affinity chromatography as described previously,^{[28d, 30b, 30d, 30e]}

Analytical-scale PT-catalyzed Reactions:

FgaPT2- and CdpNPT-catalyzed reactions were conducted in a total volume of 20–50 μL and were composed of 21 μM FgaPT2 or 200 μM CdpNPT, 1 mM substrate analog (2f, 4f, 5f, 6f, 7f, 2g, 4g, 5g, 6g, or 7g), 1.5 mM DMAPP, 50 mM Tris (pH 7.8), and 5 mM MgCl₂. FtmPT1-catalyzed reactions were conducted in a total volume of 20 μL and were composed of 45 μM FtmPT1, 1 mM substrate analog (2f, 4f, 5f, 6f, 7f, 2g, 4g, 5g, 6g, or 7g), 1.5 mM DMAPP, 25 mM Tris (pH 7.8), and 5 mM CaCl₂. Negative controls without enzyme were conducted in parallel to verify that products did not emerge via non-biocatalytic mechanisms. The reaction mixtures were incubated at 35°C for 18 hours and then quenched with the addition of 40 μL cold methanol. The precipitated protein was removed via centrifugation (6000 x g for 30 min), and the supernatants were analyzed by analytical-scale RP-HPLC. Reactions were separated by a Gemini-NX C18 (5 μm, 4.6 mm x 250 mm) column (Phenomenex, Torrance, CA, USA) attached to an Agilent 1220 HPLC system equipped with a DAD. Analytical-scale Method 7 (described below) was used for reactions containing Aza-Trp analogs, and analytical-scale Method 8 (described below) was used for reactions containing Aza-CyWP analogs. For all positive reactions, percent turnover was calculated by either 1) dividing the area under the unreacted substrate peak by the area under the substrate peak from the negative control and subtracting from 1, or 2) using a calibration curve in cases where the absorbance profiles of products differed from their substrates. All positive reactions were subsequently confirmed using HRMS as described in the Supporting Information.

Enzymatic Scale-up Reactions for Aza-Trp Prenylation:

Scale-up reactions of Aza-Trp prenylation were conducted in a volume of 20 mL consisting of 1 mM Aza-Trp analog (4f or 5f), 1.25 mM DMAPP, 50 mM Tris (pH 7.8) and 5 mM MgCl₂. Reactions were initiated by the addition of 4 mg purified FgaPT2 (~500 μL) and incubated at 35°C for 16–48 h. Reaction progress was monitored on analytical-scale HPLC (Method 7) by taking 50 μL aliquots at regular time intervals. An additional 4 mg enzyme was added after 24 hours for reactions containing 4f due to poor conversion (<25%), and these reactions were incubated for an additional 24 hours at 35°C. Once completed, the reaction mixtures were concentrated via lyophilization to <5 mL, saturated with K₂CO₃, and extracted with isopropanol (3 x 10 mL). The combined organic layers were then dried over K₂CO₃, filtered via syringe filter, and evaporated via nitrogen flow. The residue was redissolved in 1 mL methanol and purified by semi-preparative RP-HPLC (Method 1) using a Gemini-NX, C-18 (5 μm, 10 × 250 mm) column (Phenomenex, Torrance, California, USA) attached an Agilent 1220 system equipped with a DAD detector. This procedure yielded purified N4-4f and N5-5f.

Enzymatic Scale-up Reactions for Aza-CyWP Prenylation:

Scale-up reactions of Aza-CyWP prenylation were conducted in a volume of 15–20 mL consisting of 1 mM Aza-CyWP analog (2f, 4f, 5f, 6f, or 7f), 1.5 mM DMAPP, 25 mM Tris (pH 7.8) and 5 mM CaCl₂. Reactions were initiated by the addition of 10 mg purified FtmPT1 or 20 mg of CdpNPT (~350 μL) and were then incubated at 35°C for 16–18 h. Reaction progress was monitored on analytical-scale HPLC (Method 8) by taking 50 μL aliquots at regular time intervals. An additional 10 mg enzyme was added if poor conversion was observed (<25%), followed by another 24 h incubation at 35°C. Once completed, the reaction mixtures were basified with K₂CO₃ and extracted with EtOAc (3 x 15 mL). Brief centrifugation was used to accelerate organic layer separation. The combined organic layers were then dried over Na₂SO₄ and concentrated under reduced pressure. Products were isolated by semi-preparative RP-HPLC using Gemini-NX, C-18 (5 μm, 10 × 250 mm) column (Phenomenex, Torrance, California, USA) attached to an Agilent 1220 system equipped with a DAD detector. The following semi-preparative methods (described below) were utilized for the corresponding products: Method 2 for N1-2g, Method 4 for N1-4g and C3^C-4g, Method 3 for C3^C-5g, Method 9 for C3^C-6g, and Method 6 for N1-7g and C2-7g.

HPLC Methods

Method 1:

Semi-preparative; used for the purification of 2f, 4f, 5f, 6f, 7f, N4-4f, and N5-5f [gradient of 1% B for 3 min, 1% B to 25% B for 15 min, 25% B to 100% B for 5 min, 100% B for 4 min, 100% B to 1% B in 0.5 min, 1% B for 6 min (A = ddH₂O with 0.1% TFA or formic Acid; B = acetonitrile); flow rate = 2.5 ml min⁻¹; A₂₃₀].

Method 2:

Semi-preparative; used for the purification of 2f and N1-2g [gradient of 10% B to 100% B over 25 min, 100% B for 4 min, 100% B to 10% B in 0.1 min, 10% B for 6 min (A = ddH₂O with 0.1% formic acid; B = acetonitrile); flow rate = 2 mL min⁻¹; A₂₈₀].

Method 3:

Semi-preparative; used for the purification of 4f and C3^C-5g [gradient of 10% B to 100% B over 30 min, 100% B for 7 min, 100% B to 10% B in 0.1 min, 10% B for 8 min (A = ddH₂O with 0.1% formic acid; B = acetonitrile); flow rate = 1.5 mL min⁻¹; A₂₈₀].

Method 4:

Semi-preparative; used for the purification of 5f, N1-4g, and C3^C-4g [gradient of 5% B to 70% B over 18 min, 70% B to 100% B over 2 min, 100% B for 5 min, 100% B to 25% B in 0.1 min, 5% B for 5 min (A = ddH₂O with 0.1% ammonia; B = acetonitrile); flow rate = 2 mL min⁻¹; A₂₈₀, A₃₄₀].

Method 5:

Semi-preparative; used for the purification of 6f [gradient of 20% B to 100% B over 30 min, 100% B for 5 min, 100% B to 20% B in 0.2 min, 20% B for 5 min (A = ddH₂O with 0.1% ammonia; B = acetonitrile); flow rate = 2 mL min⁻¹; A₂₈₀].

Method 6:

Semi-preparative; used for the purification of 7f, N1-7g, C2-7g, and C3^C_R-7g [gradient of 10% B to 60% B over 20 min, 60% B to 100% B in 5 min, 100% B for 5 min, 100% B to 10% B in 0.1 min, 10% B for 5 min (A = ddH₂O with 0.1% ammonia; B = acetonitrile); flow rate = 2 mL min⁻¹; A₂₈₀].

Method 7:

Analytical; used for separation and analysis of analytical-scale PT-catalyzed reactions containing Aza-Trp analogs [1% B for 3 min, gradient of 1% B to 5% B over 7 min, gradient of 5% B to 25% B over 10 min, gradient of 25% B to 100% B over 10 min, 100% B for 4 min, 100% B to 1% B over 1 min, 1% B for 5 min (A = ddH₂O with 0.1% TFA; B = acetonitrile); flow rate = 1 mL min⁻¹; A₂₃₀ + A₂₅₄ + A₂₈₀ + A₃₁₆ + A₃₄₀].

Method 8:

Analytical; used for separation and analysis of analytical-scale PT-catalyzed reactions containing Aza-CyWP analogs [gradient of 1% B to 100% B over 25 min, 100% B for 3 min, 100% B to 1% B over 0.1 min, 1% B for 7 min (A = ddH₂O with 0.1% TFA; B = acetonitrile); flow rate = 1 mL min⁻¹; A₂₃₀ + A₂₅₄ + A₂₈₀ + A₃₁₆ + A₃₄₀].

Method 9:

Semi-preparative; used for the purification of C3^C-6g [gradient of 10% B to 100% B over 30 min, 100% B for 7 min, 100% B to 10% B in 0.1 min, 10% B for 8 min (A = ddH₂O with 0.1% formic acid; B = acetonitrile); flow rate = 1.5 mL min⁻¹; A₂₈₀].

Cytotoxicity Assay for Prenylated Aza-CyWP Analogs:

Cell titer-blue viability assays for all prenylated Aza-CyWP analogs were performed using human leukemia K562 cells as described previously.^[30e]