Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2020 Oct 30.
Published in final edited form as: J Am Chem Soc. 2019 Oct 17;141(43):17361–17369. doi: 10.1021/jacs.9b09067

Total Synthesis and Stereochemical Assignment of Streptide

Nicholas A Isley †,#, Yusuke Endo †,#, Zhi-Chen Wu †, Brett C Covington ‡, Leah B Bushin ‡, Mohammad R Seyedsayamdost ‡, Dale L Boger †,*
PMCID: PMC6821584  NIHMSID: NIHMS1053319  PMID: 31577142

Abstract

Streptide (1) is a peptide-derived macrocyclic natural product that has attracted considerable attention since its discovery in 2015. It contains an unprecedented post-translational modification that intramolecularly links the β-carbon (C3) of a residue 2 lysine with the C7 of a residue 6 tryptophan, thereby forming a 20-membered cyclic peptide. Herein, we report the first total synthesis of streptide that confirms the regiochemistry of the lysine-tryptophan crosslink and provides an unambiguous assignment of the stereochemistry (3R vs 3S) of the lysine-2 C3 center. Both the 3R and the originally assigned 3S lysine diastereomers were independently prepared by total synthesis and it is the former, not the latter, that was found to correlate with the natural product. The approach enlists a powerful Pd(0)-mediated indole annulation for the key macrocyclization of the complex core peptide, utilizes an underdeveloped class of hypervalent iodine(III) aryl substrates in a palladium-catalyzed C–H activation/β-arylation reaction conducted on a lysine derivative, and provides access to material with which the role of streptide and related natural products may be examined.

Graphical Abstract

graphic file with name nihms-1053319-f0001.jpg

INTRODUCTION

Streptide (1) is a peptide-derived macrocyclic natural product that was first isolated from Streptococcus thermophilus (Figure 1). Its production is controlled by a quorum sensing system, consisting of a short hydrophobic peptide signal and a cognate transcriptional regulator, that is common to many streptococci, including pathogenic organisms.1,2 Streptide biosynthesis commences with the ribosomal production of a 30 amino acid peptide that is post-translationally modified by macrocyclization and further processed by proteolysis to deliver the mature nine amino acid cyclic peptide.1–3 The structure of streptide and the biosynthetic origin of the macrocycle were disclosed in 2015 through a combination of spectroscopic and enzymatic studies.3 NMR spectroscopic analyses located the crosslink to the β-carbon (C3) of the residue 2 lysine and C7 of the residue 6 tryptophan, giving rise to a 20-membered cyclic peptide. The α-carbon configurations were experimentally determined to be L while computational studies in conjunction with NMR NOESY constraints suggested a lysine-2 (S)-C3 configuration at the new chiral center.3 Subsequent biosynthetic investigations showed that the crosslink is installed in a single step by StrB, a radical SAM metalloenzyme, via radical-mediated chemistry. More recently, homologs of StrB have been identified in pathogenic streptococci.4,5 They have been shown to catalyze the same crosslinking reaction on a common AK2GDGW6 peptide core that contains the identical leader sequence but altered C-termini, depending on the producing strain, indicating that streptide is the first characterized member of a growing family of such post-translationally modified macrocyclic peptides.6 At present, the role of streptide and its congeners upon secretion remains undefined.

Figure 1.

Figure 1.

Open in a new tab

Assigned lysine (3S) and revised (3R) structure of streptide (1).

In efforts that culminated in the total synthesis of the chloropeptins and complestatins,7 we introduced an intramolecular palladium(0)-mediated indole annulation for macrocyclization of their strained biaryl ring systems that contain crosslinks to either C6 or C7 of the indole of a (R)-tryptophan embedded within their peptide cores.7,8 Not only did this reaction permit synthesis of their strained 16- or 17-membered macrocyclic rings, but it did so in the context of fully functionalized peptides, providing a powerful new macrocyclization reaction that was more effective than conventional alternatives.9 Given the unusual structure of streptide, its unprecedented crosslink, its unknown function, and its low natural abundance (1.7 mg/20 L production culture), we embarked on a total synthesis of streptide that would enlist this palladium(0)-mediated macrocyclization reaction as a key step in the assembly of the 20-membered cyclic peptide core. This approach would afford a total synthesis of 1 amenable to the preparation of analogues or congeners, and at the same time provide a formidable test of the generality of the indole annulation macrocyclization. Thus, key to our approach was the use of this intramolecular Larock indole synthesis to close the 20-membered cyclic peptide, enlisting conditions that permit the use of a 2-bromoaniline and its reaction with an alkyne that contains a removable large terminal substituent (–SiEt3) to control the indole cyclization regioselectivity (Figure 2).10 Not only would this cyclization provide the fully functionalized ring system of 1, but its use would represent only the second reported application of what has proven to be a powerful macrocyclization reaction for the synthesis of complex natural products. Additionally, a palladium-catalyzed C–H activation reaction11 on a lysine derivative was planned for diastereoselective installation of a suitably functionalized β-aryl substituent needed for implementation of the indole annulation.

Figure 2.

Figure 2.

Open in a new tab

Key retrosynthetic steps for the total synthesis of streptide.

The efforts detailed herein culminated in the first total synthesis of streptide and provided an unambiguous assignment of the regiochemistry and stereochemistry (3R vs 3S) of the lysine C3 center. Both the 3R and the originally proposed 3S lysine diastereomers were independently prepared by total synthesis and subsequent NMR and HPLC analysis demonstrated that the 3R diastereomer correlates with the natural product. In addition to demonstrating the generality of the indole annulation macrocyclization reaction, the approach led to the use of an underutilized class of hypervalent iodine(III) aryl substrates in the palladium-catalyzed C–H activation/β-arylation functionalization of a lysine derivative. With an approach at hand to synthesize streptide and related analogues, the function of this natural product and its congeners can now be explored.

RESULTS AND DISCUSSION

β-Arylation of lysine derivatives by palladium-catalyzed C–H activation.

Since the initial report by Corey12 utilizing Daugulis’ bidentate 8-aminoquinoline directing group,13 additional advances in Pd-catalyzed β-functionalization of amino acids have been reported,14 including recent functionalizations reported by Yu without added directing groups.15 Despite these reports, Pd-catalyzed β-aryl functionalization of amino acids at non-benzylic secondary alkyl centers possessing β-hydrogens is still underdeveloped.16 This challenge is compounded when intercepting an intermediate Pd(II) alkyl species with a sterically demanding aryl halide coupling partner as required for our work (i.e. ortho-substitution).17 To effectively install the appropriately functionalized aryl unit for the key indole annulation macrocyclization reaction, we found that the use of a diaryliodonium salt allowed introduction of a sterically hindered 3-bromo-2-acetamidophenyl group without competitive reactions of the requisite aryl bromide in a reaction in which the empirically-derived conditions also proved important, perhaps accelerating the reaction or suppressing competitive β-hydride elimination and C–Pd homolysis.18,19 This β-arylation reaction was efficient, scalable, diastereoselective and conducted with an unsymmetrical diaryliodonium salt, bearing a non-transferable sterically large mesityl group.18 Thus, treatment of 4, prepared from L-lysine (Supporting Information), and the iodonium salt 3,20 prepared in two steps (86%) from commercially available 6-bromo-2-iodoaniline, with Pd(OAc)2 (0.2 equiv) and AgOAc (2.0 equiv) in t-BuOH (0.1 M) provided 5 in good yield and diastereoselectivity (60%, 4–6:1 dr) under mild reaction conditions (80 °C, 18 h). This was observed without competitive α-epimerization of the product (>99% ee), intramolecular trap of the intermediate Pd(II)-palladacycle by the distal carbamate, or β-hydride elimination (Scheme 1). The outcome of the reaction allowed us to carry both the 3R (major) and 3S (minor) diastereomers forward to complete the synthesis of both streptide stereoisomers. Fortuitously, though unanticipated, it was the major 3R diastereomer 5, not the minor 3S diastereomer 6, of this reaction that afforded streptide.

Scheme 1.

Scheme 1

Open in a new tab

Initial efforts to install a functionalized aryl group at the β-position of lysine were conducted with N-acetyl-6-bromo-2-iodoaniline (2b) and N-acetyl-2,6-dibromoaniline and were only modestly successful. The native directing group ability of the carboxylic acid was explored15 as well as removable carboxylic acid-derived directing groups (e.g., -CONHOMe and -CONHC6F5 carboxamides), albeit unsuccessfully.11 Notably, when 4 was employed in early studies, the intermediate Pd(II) C–H insertion product was often intercepted in an intramolecular amination reaction by the distal carbamate (NHBoc) to give a pyrrolidine. Thus, a screen was conducted with the bis-phthalimide protected lysine derivative 10, lacking a distal carbamate NH and incorporating the strongly coordinating 8-aminoquinoline directing group.13 The screen, which enlisted 2b as the coupling partner, was conducted with varied solvents, palladium sources, bases, and additives, leading to confirmed product formation. However, even with extensive optimization, product 11 was only observed at near stoichiometric loadings of Pd(OAc)2 (0.4 equiv) with excess aryl iodide (4 equiv), and then only afforded 11 in low yield (27%, 0.4 equiv of Pd(OAc)2, 1.5 equiv of AgOAc, 0.07 M t-BuOH, 110 °C, 48 h) (Figure 3). As a result, efforts were refocused on alternatives to the coupling partner 2b, expecting that a stronger oxidizing reagent may better access the needed Pd(IV) oxidative insertion intermediate than an aryl iodide with our hindered substrate. Unsymmetrical diaryliodonium salts such as 3, bearing a nontransferable aryl group, were explored as coupling partners. Although such reagents have been used effectively,18,19 Sanford’s early mechanistic studies indicated that their participation in oxidative addition at Pd(II) is still the rate-determining step in a PdII/IV cycle.18 This likely has discouraged their subsequent exploration in couplings that might suffer competing reactions of the intermediate Pd(II) palladacycle. Treatment of 10 even with near stoichiometric (vs excess) amounts of the iodonium salt 3 (1.2 equiv, 0.2 equiv of Pd(OAc)2, 1.2 equiv of K2CO3) provided 11 in good yield (60%) with excellent diastereoselectivity (>20:1 dr) under mild reaction conditions (110 °C, 16 h) without the need for other heavy metal additives (e.g., Ag(I)) (Figure 3). In the optimization of the reaction of 10 with 3, no reaction was observed at 25–60 °C, an increased loading of base did not further improve conversions and use of toluene versus ClCH2CH2Cl led to competitive β-hydride elimination. The extension of these observations to 4, as shown in Scheme 1, now proved possible and provided 5 under mild conditions (t-BuOH, 80 °C, 18 h) but required inclusion of AgOAc in the reaction mixture to avoid distal carbamate interception of the intermediate Pd(II) C–H insertion product.

Figure 3.

Figure 3.

Open in a new tab

Initial key observations on C–H activation and subsequent arylation of a lysine derivative.

The stereochemistry of 5 and 6 was assigned initially based on the original work of Corey12 for Pd-catalyzed β-arylation of phthaloyl-protected L-amino acids that contain the 8-aminoquinolyl directing group (>20–4:1 dr favoring 2S,3R), in which the initial C–H insertion preferentially forms the sterically favored intermediate 5-membered trans-palladacycle. In Corey’s work, the stereochemical assignments were unambiguously established by X-ray structures of several of the major products,12 and such work subsequently has been extended to crystallographic characterization of the intermediate trans-palladacycles themselves by Yu and Chen.14f,g Our initial assignments based on this work were confirmed with the single crystal X-ray structure determinations of 12 and 13, derived from the major and minor diastereomers 5 and 6, respectively (Figure 4).20

Figure 4.

Figure 4.

Open in a new tab

Relative and absolute stereochemistry of 5 and 6 established by X-ray diffraction analysis.20

The major diastereomer 5 and its accompanying minor isomer 6 were not easily separated by chromatography at this stage but were found to be readily separated after installation of the protected N-terminal alanine residue. Thus, phthalimide deprotection of the diastereomeric mixture of 5 and 6 (4–6:1) with 1,2-diaminoethane (EtOH, 0–25 °C, 18 h) provided the corresponding free amines (99%), which were immediately coupled with BocHN-Ala-OH (1.2 equiv, 1.2 equiv of 1-cyano-2-ethoxy-2-oxoethylidenaminooxy)-dimethylamino-morpholino-carbenium hexafluorophosphate (COMU), 1.0 equiv of N-methylmorpholine (NMM), DMF, 0–25 °C, 18 h, 75–87% combined yield) to provide the alanine-linked lysine-2 2S,3R (7) and 2S,3S (8) diastereomers (Scheme 1). At this point, the lysine C3 diastereomers were easily separated by column chromatography. Removal of the 8-aminoquinoline directing group from the major isomer 7 was accomplished by mild oxidative deprotection (via ozonolysis),21 which cleaves the quinolyl group to a reactive imide intermediate that we found can be hydrolyzed selectively to provide not only a carboxamide (NH4OH, THF, 25 °C, 18 h, 67%), but also the carboxylic acid. Thus, ozonolysis of 7 in the presence of pyridine22 (3 equiv) at -78 °C (CH2Cl2, 5–10 min) followed by treatment of the isolated reaction product with H2O2–LiOH (THF/H2O, 0 °C, 1 h) directly provided the carboxylic acid 9 in good yield (63%, Scheme 1). Whereas ozonolysis in absence of pyridine led to substantial amounts of carboxamide product upon hydrolysis, its use as an ozonolysis additive22 tempered the oxidation reaction and thereby helped suppress the competitive carboxamide formation in the subsequent intermediate imide hydrolysis.

Assembly of the N-terminal core hexapeptide and Larock macrocyclization.

The linear peptide 19, incorporating 9 and the partner triethylsilyl alkyne that serve as the precursors for the crosslinked tryptophan, was prepared by coupling H2N-Gly-OBn with FmocHN-(OtBu)Asp-OH promoted by N-(3-dimethylaminopropyl)-N′-ethylcarbodiimide hydrochloride and 1-hydroxy-7-azabenzotriazole (EDCI–HOAt, DMF) to afford the dipeptide 14 in 89% yield (Scheme 2). Deprotection of 14 (1 equiv of 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU), CH2Cl2, 0 °C, 1.5 h) under conditions found to suppress cyclization followed by coupling of the released free amine in the same vessel with FmocHN-Gly-OH (EDCI–HOAt, 16 h) provided 15 in 85% yield. Benzyl ester removal by hydrogenolysis (H2, Pd/C, THF, 23 °C, 16 h) afforded the tripeptide 16 (90%), which was coupled with 177a (preparation supplied in Supporting Information, 1.07 equiv of HBTU, 1.07 equiv of NMM, DMF, 0–23 °C, 16 h) to provide 18. (2-(1H-Benzotriazol-1-yl)-1,1,3,3-tetramethyluronium hexafluorophosphate (HBTU) was found to be the optimal coupling reagent of those examined, providing 18 in good yield (84%) and under conditions that do not result in α-epimerization adjacent to the methyl ester. After in situ Fmoc deprotection of 18 (DBU, CH2Cl2, 0 °C, 1 h), the linear hexapeptide 19 of the streptide macrocyclic core was accessed by coupling the liberated amine with 9 (1.2 equiv of COMU, 1.0 equiv of NMM, CH2Cl2, 0–25 °C, 18 h, 83%).

Scheme 2.

Scheme 2

Open in a new tab

Macrocyclization of 19 was first examined under conditions we previously introduced (1.1 equiv of Pd(OAc)2, 1.3 equiv of 1,1’-bis(di-tert-butylphosphino)ferrocene (dtbpf), 1.3 equiv of Et3N, toluene/MeCN (1:1), 100 °C, 1 h), permitting use of an aryl bromide.7 The reaction provided 20 in good albeit variable yield (50%, Figure 5). The addition of Bu4NBr (1 equiv), which is thought to increase catalyst stability by preventing Pd black formation,23 maintained but did not improve the yield of 20 (45%). However, the reaction remained variable, especially on small scale, complicating its optimization and scale up. The use of other ammonium salts or additives (e.g., LiCl) did not improve this further. Whereas other Pd catalysts such as PdCl2(MeCN)2, Pd2(dba)3, PdCl2(dtbpf), and (o-tol3P)3Pd failed to provide the macrocyclization product or did so poorly as noted in our previous work,7,8 (t-Bu3P)2Pd also proved to be effective and provided consistent conversions as reported by Reisman.24 Further optimization focused on the use of this catalyst (Figure 5). Macrocycle 20 was obtained in 60% yield under even milder reaction conditions (80 °C, 0.001 M MeCN, 1 h), using Et3N or dicyclohexyl methylamine (Cy2NMe, 1.3 equiv) and (t-Bu3P)2Pd (1.1 equiv) (entries 4 and 8). Although Et3N could be used as the added base (entry 4), its lower boiling point led to more variable yields on small scale (entry 4 vs 3), and this is likely the source of the variable conversions we initially observed with its use with Pd(OAc)2. Prolonged reaction times beyond 1 h in addition to increased base loading also provided the desired product in comparable, but not improved, yields (entries 7 and 9). Lastly, a catalyst loading of 0.6 equiv of (t-Bu3P)2Pd matched the conversion of its stoichiometric use (58%, 2 h, entry 6 vs 8 and 9), whereas use of 0.2 equiv of (t-Bu3P)2Pd also provided 20, albeit in lower yield (34%, 4 h, entry 5). Because such reactions work effectively with catalytic Pd(OAc)2 or (t-Bu3P)2Pd on simpler non-peptide substrates,8,24 it is likely that the precursor 19 or the macrocycle 20 act to sequester the catalyst, requiring the higher catalyst loading.7,8 However, further efforts to optimize the reaction for sub-stoichiometric use of the Pd catalyst were not pursued as it was not central for the work detailed herein. Nonetheless, it is remarkable that the Larock cyclization, adapted for use with an aryl bromide, performs effectively for formation of a 20-membered macrocycle on a complex substrate that contains 6 amides, 2 carbamates and 2 esters, one of which is prone to epimerization.

Figure 5.

Figure 5.

Open in a new tab

Representative results of the macrocyclization optimization.

With 20 in hand, its selective or exhaustive deprotections were examined. This not only set the stage for completion of the total synthesis of streptide, but it also permitted the examination of the conformational properties of 1 and the core macrocycle. TES removal under mild conditions (3 equiv of Bu4NF, THF, 0 °C, 78%) provided 21 (Figure 6). Whereas hydrolysis of the C-terminal methyl ester of 20 with LiOH or related reagents was challenging due to competing epimerization and partial indole N-acetyl deprotection, the hydrolysis to provide 22 was accomplished cleanly with Me3SnOH (88–93%).25 Finally, studies on the single-step global deprotection of 22 provided complex mixtures unless a trialkylsilane was included in the reaction mixture, acting to trap cationic intermediates arising from the deprotections. Thus, both 23 (71%) and 24 (94%) were obtained in excellent yields from 22 following HCl or trifluoroacetic acid (TFA) treatment and are indicative of clean TES removal along with cleavage of the t-butyl carbamates and t-butyl ester and a slower hydrolysis of the N-acetylindole.

Figure 6.

Figure 6.

Open in a new tab

Deprotection studies on the macrocyclic core.

Macrocycle 20 and each of the partial or complete deprotection products 21–24 displayed relatively sharp 1H NMR spectra consistent with adoption of rigid or rapidly interconverting conformations. At 1 °C but not at 25 °C, only 20 displayed a limited number of distinct (e.g., indole C4 CH) and sharp 1H NMR signals (CD3OD) attributable to two (major and minor) conformations (Supporting Information Figures S1–S6). Although not examined in detail, these comparisons of 20 and 21 suggest that 20, as well as 21, do not display atropisomeric behavior at room temperature.

Completion of the total synthesis of streptide.

As detailed above, selective hydrolysis of the methyl ester 20 to provide 22 was achieved with use of Me3SnOH (4–10 equiv, ClCH2CH2Cl, 80 °C, 4.5–2 h, 88–93%), leaving the indole N-acetyl group untouched and avoiding detectable epimerization (Figure 6). Typical hydrolysis reactions mediated by LiOH, with or without added H2O2, either led to apparent competitive epimerization or partial deprotection of the indole N-acetyl group. Fmoc deprotection of the tripeptide 25 (2 equiv, synthesis in Supporting Information) with DBU (1.5 equiv, 0 °C, 1 h) followed by coupling of the free amine with 22 (2 equiv of HBTU, 1 equiv of NMM, DMF/CH2Cl2, 0–25 °C, 18 h) provided 26 (75%, >99:1 dr), constituting a fully protected streptide (Scheme 3). Although not investigated in detail, alternative amide coupling reagents such as COMU provided only trace product formation, whereas use of EDCI–HOAt (CH2Cl2) and HATU (DMF) led to apparent competitive α-epimerization but provided the products in good yield (56% and 67%, respectively). A single-step global deprotection of 26 to provide streptide (1) with removal of 7 protecting groups was achieved by treatment with aqueous 5 N HCl/Et3SiH/1,2-ethanedithiol (8:1:1) at room temperature. This smoothly and cleanly provided streptide (1) in high yield (82–94%) after purification by reverse-phase HPLC, completing the first reported total synthesis of streptide.

Scheme 3.

Scheme 3

Open in a new tab

Comparison of the spectroscopic properties of synthetic 1 with the 1H and 13C NMR spectra reported for authentic streptide provided a near identical match. In the course of the initial characterization of synthetic 1, we found that trace amounts of acid (TFA) residual from our HPLC purification can impact the chemical shifts observed in the NMR spectra and it is possible concentration dependent distinctions were also observed. This initial and unexpected correlation with natural streptide was particularly strong for the 1H NMR spectrum, whereas there were a few larger chemical shift differences in the 13C NMR. It was not clear whether the unexpected spectroscopic correlation of synthetic 1, containing the lysine 3R (not the proposed 3S) stereochemistry, with natural streptide represented a true correlation or one that was fortuitously close. Consequently, and as originally planned, the lysine 3S-diastereomer of streptide was also prepared.

Total synthesis of the lysine-2 3S-diastereomer of streptide.

The lysine-2 C3 configuration in streptide could not be unambiguously established by spectroscopic methods in the original structure determination studies.3 Instead, a computational approach was employed to assess correspondence between the computed streptide structures and constraints derived from NOESY correlations using the CYANA algorithm. The diastereomer that satisfies more constraints would give a lower target function (f). Both S and R C3 configurations could be acceptably fit to the 1H NMR data (e.g.; S vs R: f = 0.042 vs 0.054, rmsd = 1.62 vs 1.72 Å, and accounting for 29 vs 27 observed NOEs between calculated and observed values)3 with the S-diastereomer providing a slightly better fit. In addition, the original characterization of the natural product and its spectroscopic data were collected on limited amounts of sample. To unambiguously assign the configuration at the lysine C3, the lysine 3S-diastereomer 8, arising from the minor diastereomer 6 generated in the C–H functionalization reaction, was carried through the identical synthetic sequence (Scheme 4). Notably, the Pd(0)-promoted macrocyclization of 28 provided 29 in conversions (60%) identical to that observed with the diastereomer 19 without alteration of the reaction conditions (80 °C, 0.001 M CH3CN, 1 h). This synthesis provided 32, the lysine-2 C3 S-diastereomer.

Scheme 4.

Scheme 4

Open in a new tab

Reisolation of and comparison with natural streptide.

To facilitate the comparison of the synthetic streptide diastereomers with the authentic natural product, streptide was isolated from 20 L of S. thermophilus production cultures using previously established procedures,3 yielding approximately 1.7 mg of pure material. After completion of the synthetic work and with the benefit of having freshly isolated streptide available, detailed NMR and HPLC-MS comparisons were conducted. Upon side-by-side examination, natural streptide and the synthetic lysine-2 3R diastereomer clearly correlated. The 1H and 13C NMR data were essentially identical, whereas the 3S diastereomer was easily distinguishable with divergent chemical shifts at the lysine-2 α−1H (4.81 ppm for the 3R diastereomer and authentic streptide, 5.02 for the 3S diastereomer) and lysine-2 β−1H (3.52 ppm for the 3R diastereomer and authentic streptide, 4.03 ppm for the 3S diastereomer), among others (Supporting Information Figure S20–S23). Moreover, we were able to demonstrate that the 1H NMR and 13C NMR spectra of 1 display subtle condition-dependent shifts and changes largely centered within or adjacent to the acidic residues (Met9, Asp4, Gly3 and Gly5), accounting for the uncertainty in our initial correlation with the original published spectra (Supporting Information Figures S17–S19 and accompanying discussion). However, changes in the chemical shifts for Ala1, Lys2 and Trp6 were not observed (< 0.01 ppm differences), indicating a lack of any impact of either the protonation state or any major conformational change of the macrocycle under the conditions examined. In agreement with the NMR comparisons, high-resolution HPLC-MS analysis unambiguously demonstrated co-elution of authentic streptide with the 3R diastereomer 1, but not with the 3S diastereomer 32, which could be separated from both when injected as mixtures (Figure 7). Together, these results confirm the first chemical synthesis and establish the full structural assignment of streptide.

Figure 7.

Figure 7.

Open in a new tab

HR-HPLC-MS analysis of authentic and synthetic streptides. The sample loaded is labeled on top of the resulting elution profile. Streptide elution was monitored by absorbance at 280 nm (not shown) and by MS total ion count. Shown is the HR-MS-extracted ion chromatogram for streptide (m/z 989.489) in each trace. Authentic streptide co-elutes with lysine-2 C3 diastereomer 1 (3R) but not with 32 (3S). See Supporting Information for details.

With knowledge of the absolute configuration of the lysine-2 C3, we generated a computational model of streptide in CYANA using constraints derived from NOESY data that were collected on authentic 1.3,26 The crosslinked Lys/Trp residues were added manually to the CYANA residue library and the configuration of the lysine-2 C3 was explicitly defined as R. The calculation provided possible solution conformers, with the conformer that had the lowest target function, and therefore exhibited the best agreement with the experimental constraints, depicted in Figure 8. It features an indole-based cyclophane, a (3,7)indolophane, in which the encircling peptide adopts a rigid conformation enforced by the amide bonds within the macrocycle and the bridging indole. The three-dimensional model of 1 facilitates future docking studies to possible target sites.

Figure 8.

Figure 8.

Open in a new tab

Computational model of the three-dimensional structure of streptide.

CONCLUSION

Herein, we report the first total synthesis of streptide that provided confirmation of the site (regiochemistry) and established the configuration (stereochemistry, 3R vs 3S) of the residue 2 lysine C3 central to the unusual crosslink in this natural product. Both the 3R and the initial computationally assigned 3S lysine diastereomers were independently prepared by total synthesis, involving a longest linear synthetic sequence of 11 steps, and it was the 3R diastereomer that was found to correlate with the natural product. The approach utilized a powerful Pd(0)-mediated indole annulation for macrocyclization of the core 20-membered cyclic peptide, enlisted an underutilized hypervalent iodine(III) aryl substrate in a key palladium-catalyzed C–H activation and β-arylation reaction conducted on a lysine derivative, and provides access to material with which the role of streptide and related natural products may be examined.

Supplementary Material

Supporting Information

ACKNOWLEDGMENTS

We gratefully acknowledge financial support from the NIH (CA041101 to D.L.B. and postdoctoral fellowship F32 GM114948 to N.A.I.), Kaken Pharmaceutical Co., Ltd. (to Y.E.), the Burroughs Wellcome Fund (PATH Investigator Award to M.R.S.), and the NSF (NSF GRF Award to L.B.B.). We thank Milan Gembicky, C. Moore and A. Rheingold of the Crystallography Facility at the University of California, San Diego for X-ray structure determinations of 3, 12 and 13.

Footnotes

The authors declare no competing financial interest.

REFERENCES

  • 1.Ibrahim M; Guillot A; Wessner F; Algaron F; Besset C; Courtin P; Gardan R; Monnet V Control of the transcription of a short gene encoding a cyclic peptide in Streptococcus thermophilus: a new quorum-sensing system? J. Bacteriol 2007, 189, 8844–8854. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Fleuchot B; Gitton C; Guillot A; Vidic J; Nicolas P; Besset C; Fontaine L; Hols P; Leblond-Bourget N; Monnet V; Gardan R Rgg proteins associated with internalized small hydrophobic peptides: a new quorum-sensing mechanism in streptococci. Mol. Microbiol 2011, 80, 1102–1119. [DOI] [PubMed] [Google Scholar]
  • 3.Schramma KR; Bushin LB; Seyedsayamdost MR Structure and biosynthesis of a macrocyclic peptide containing an unprecedented lysine-to-tryptophan crosslink. Nat. Chem 2015, 7, 431–437. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Davis KM; Schramma KR; Hansen WA; Bacik JP; Khare SD; Seyedsayamdost MR; Ando N Structures of the peptide-modifying radical SAM enzyme SuiB elucidate the basis of substrate recognition. Proc. Natl. Acad. Sci. USA 2017, 114, 10420–10425. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Schramma KR; Forneris CC; Caruso A; Seyedsayamdost MR Mechanistic investigations of lysine-tryptophan cross-link formation catalyzed by streptococcal radical S-adenosylmethionine enzymes. Biochemistry 2018, 57, 461–468. [DOI] [PubMed] [Google Scholar]
  • 6.Schramma KR; Seyedsayamdost MR Lysine-tryptophan crosslinked peptides produced by radical SAM enzymes in pathogenic streptococci. ACS Chem. Biol 2017, 12, 922–927. [DOI] [PubMed] [Google Scholar]
  • 7.(a) Garfunkle J; Kimball FS; Trzupek JD; Takizawa S; Shimamura H; Tomishima M; Boger DL Total synthesis of chloropeptin II (complestatin) and chloropeptin I. J. Am. Chem. Soc 2009, 131, 16036–16038. [DOI] [PMC free article] [PubMed] [Google Scholar]; (b) Shimamura H; Breazzano SP; Garfunkle J; Kimball FS; Trzupek JD; Boger DL Total synthesis of complestatin: development of a Pd(0)-mediated indole annulation for macrocyclization. J. Am. Chem. Soc 2010, 132, 7776–7783. [DOI] [PMC free article] [PubMed] [Google Scholar]; (c) Breazzano SP; Boger DL Synthesis and stereochemical determination of complestatin A and B (neuroprotectin A and B). J. Am. Chem. Soc 2011, 133, 18495–18502. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.(a) Breazzano SP; Poudel YB; Boger DL A Pd(0)-mediated indole (macro)cyclization reaction. J. Am. Chem. Soc 2013, 135, 1600–1606. [DOI] [PMC free article] [PubMed] [Google Scholar]; (b) Lee K; Poudel Y; Glinkerman CM; Boger DL Total synthesis of dihydrolysergic acid and dihydrolysergol: development of a divergent strategy applicable to rapid assembly of D-ring analogs. Tetrahedron 2015, 71, 5897–5905. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.(a) Okano A; Isley NA; Boger DL Total synthesis of vancomycin related glycopeptide antibiotics and key analogues. Chem. Rev 2017, 117, 11952–11993. [DOI] [PMC free article] [PubMed] [Google Scholar]; (b) Parenty A; Moreau X; Campagne JM Macrolactonizations in the total synthesis of natural products. Chem. Rev 2006, 106, 911–939. [DOI] [PubMed] [Google Scholar]; (c) Marsault E; Peterson ML Macrocycles are great cycles: Applications, opportunities, and challenges of synthetic macrocycles in drug discovery. J. Med. Chem 2011, 54, 1961–2004. [DOI] [PubMed] [Google Scholar]; (d) White CJ; Yudin AK Contemporary strategies for peptide macrocyclization. Nat. Chem 2011, 3, 509–524. [DOI] [PubMed] [Google Scholar]; (e) Ronson TO; Taylor RJK; Fairlamb IJS Palladium‐catalysed macrocyclisations in the total synthesis of natural products. Tetrahedron 2015, 71, 989–1009. [Google Scholar]; (f) Ronson TO; Unsworth WP; Fairlamb IJS In Practical Medicinal Chemistry with Macrocycles: Design, Synthesis, and Case Studies, Marsault E, Peterson ML Eds; John Wiley & Sons, Inc, 2017, Chapter 12, pp 281–305. [Google Scholar]; (g) Smolyar IV; Yudin AK; Nenajdenko VG Heteroaryl rings in peptide macrocycles. Chem. Rev 2019, 119, 10032–10240. [DOI] [PubMed] [Google Scholar]
  • 10.(a) Larock RC; Yum EK Synthesis of indoles via palladium-catalyzed heteroannulation of internal alkynes. J. Am. Chem. Soc 1991, 113, 6689–6690. [Google Scholar]; (b) Larock RC; Yum EK; Refvik MD Synthesis of 2,3-disubstituted indoles via palladium-catalyzed annulation of internal alkynes. J. Org. Chem 1998, 63, 7652–7662. [Google Scholar]; (c) Ma C; Liu X; Li X; Flippen-Anderson J; Yu S; Cook JM Efficient asymmetric synthesis of biologically important tryptophan analogues via a palladium-mediated heteroannulation reaction. J. Org. Chem 2001, 66, 4525–4542. [DOI] [PubMed] [Google Scholar]; (d) Shen M; Li G; Lu BZ; Hossain A; Roschangar F; Farina V; Senanayaka CH The first regioselective palladium-catalyzed indolization of 2-bromo- or 2-chloroanilines with internal alkynes: A new approach to 2,3-disubstituted indoles. Org. Lett 2004, 6, 4129–4132. [DOI] [PubMed] [Google Scholar]
  • 11.(a) He J; Wasa M; Chan KSL; Shao O; Yu J-Q Palladium-catalyzed transformations of alkyl C–H bonds. Chem. Rev 2017, 117, 8754–8786. [DOI] [PMC free article] [PubMed] [Google Scholar]; (b) Chen X; Engle KM; Wang D-H; Yu J-Q Palladium(II)‐catalyzed C–H activation/C–C cross‐coupling reactions: Versatility and practicality. Angew. Chem. Int. Ed 2009, 48, 5094–5115. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Reddy BVS; Reddy LR; Corey EJ Novel acetoxylation and C–C coupling reactions at unactivated positions in α-amino acid derivatives. Org. Lett 2006, 8, 3391–3394. [DOI] [PubMed] [Google Scholar]
  • 13.(a) Zaitsev VG; Shabashov D; Daugulis O Highly regioselective arylation of sp3 C–H bonds catalyzed by palladium acetate. J. Am. Chem. Soc 2005, 127, 13154–13155. [DOI] [PubMed] [Google Scholar]; (b) Corbet M; De Campo F 8‐Aminoquinoline: A powerful directing group in metal‐catalyzed direct functionalization of C–H bonds. Angew. Chem. Int. Ed 2013, 52, 9896–9898. [DOI] [PubMed] [Google Scholar]
  • 14.(a) Giri R; Chen X; Yu J-Q Palladium-catalyzed asymmetric iodination of unactivated C–H bonds under mild conditions. Angew. Chem. Int. Ed 2005, 44, 2112–2115. [DOI] [PubMed] [Google Scholar]; (b) Zhang S-Y; Li Q; He G; Nack WA; Chen G Stereoselective synthesis of β‑alkylated α‑amino acids via palladium-catalyzed alkylation of unactivated methylene C(sp3) -H bonds with primary alkyl halides. J. Am. Chem. Soc 2013, 135, 12135–12141. [DOI] [PubMed] [Google Scholar]; (c) Wang D-H; Wasa M; Giri R; Yu J-Q Pd(II)-catalyzed cross-coupling of sp3 C–H Bonds with sp2 and sp3 boronic acids using air as the oxidant. J. Am. Chem. Soc 2008, 130, 7190–7191. [DOI] [PubMed] [Google Scholar]; (d) Wasa M; Engle KM; Yu J-Q Pd(0)/PR3-catalyzed intermolecular arylation of sp3 C–H Bonds. J. Am. Chem. Soc 2009, 131, 9886–9887. [DOI] [PMC free article] [PubMed] [Google Scholar]; (e) Shabashov D; Daugulis O Auxiliary-assisted palladium-catalyzed arylation and alkylation of sp2 and sp3 carbon–hydrogen bonds. J. Am. Chem. Soc 2010, 132, 3965–3972. [DOI] [PMC free article] [PubMed] [Google Scholar]; (f) He J; Li S; Deng Y; Fu H; Laforteza BN; Spangler JE; Homs A; Yu J-Q Ligand-controlled C(sp3)–H arylation and olefination in synthesis of unnatural chiral α-amino acids. Science 2014, 343, 1216–1220. [DOI] [PMC free article] [PubMed] [Google Scholar]; (g) Chen K; Shi B-F Sulfonamide-promoted palladium(II)-catalyzed alkylation of unactivated methylene C(sp3)–H bonds with alkyl halides. Angew. Chem. Int. Ed 2014, 53, 11950–11954. [DOI] [PubMed] [Google Scholar]; (h) Zhu R-Y; Tanaka K; Li G-C; He J; Fu H-Y; Li S-H; Yu J-Q Ligand-enabled stereoselective β-C(sp3)–H fluorination: Synthesis of unnatural enantiopure anti-β-fluoro-α-amino acids. J. Am. Chem. Soc 2015, 137, 7067–7070. [DOI] [PMC free article] [PubMed] [Google Scholar]; (i) Chen G; Shigenari T; Jain P; Zhang Z; Jin Z; He J; Li S; Mapelli C; Miller MM; Poss MA; Scola PM; Yeung K-S; Yu J-Q Ligand-enabled β-C–H arylation of α-amino acids using a simple and practical auxiliary. J. Am. Chem. Soc 2015, 137, 3338–3351. [DOI] [PMC free article] [PubMed] [Google Scholar]; (j) Zhang X; Lu G; Sun M; Mahankali M; Ma Y; Zhang M; Hua W; Hu Y; Wang Q; Chen J; He G; Qi X; Shen W; Liu P; Chen G A general strategy for synthesis of cyclophane-braced peptide macrocycles via palladium-catalysed intramolecular sp3 C−H arylation. Nat. Chem 2018, 10, 540–548. [DOI] [PubMed] [Google Scholar]; (k) Gong W; Zhang G; Liu T; Giri R; Yu J-Q Site-selective C(sp3)–H functionalization of di-, tri-, and tetrapeptides at the N-terminus. J. Am. Chem. Soc 2014, 136, 16940–16946. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Chen G; Zhuang Z; Li G-C; Saint-Denis TG; Xiao Y; Joe CL; Yu J-Q Ligand-enabled β-C–H arylation of α-amino acids without installing exogenous directing groups. Angew. Chem. Int. Ed 2017, 56, 1506–1509. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Noisier AFM; Brimble MA C–H functionalization in the synthesis of amino acids and peptides. Chem. Rev 2014, 114, 8775–8806. [DOI] [PubMed] [Google Scholar]
  • 17.He G; Zhang S-Y; Nack WA; Pearson R; Rabb-Lynch J; Chen G Total synthesis of hibispeptin A via Pd-catalyzed C(sp3)–H arylation with sterically hindered aryl iodides. Org. Lett 2014, 16, 6488–6491. [DOI] [PubMed] [Google Scholar]
  • 18.(a) Deprez NR; Sanford MS Synthetic and mechanistic studies of Pd-catalyzed C−H arylation with diaryliodonium salts: Evidence for a bimetallic high oxidation state Pd intermediate. J. Am. Chem. Soc 2009, 131, 11234–11241. [DOI] [PMC free article] [PubMed] [Google Scholar]; (b) Kalyani D; Deprez NR; Desai LV; Sanford MS Oxidative C–H activation/C–C bond forming reactions: Synthetic scope and mechanistic insights. J. Am. Chem. Soc 2005, 127, 7330–7331. [DOI] [PubMed] [Google Scholar]
  • 19.(a) Daugulis O; Zaitsev VG Anilide ortho-arylation by using C–H activation methodology. Angew. Chem. Int. Ed 2005, 44, 4046–4048. [DOI] [PubMed] [Google Scholar]; (b) Pan F; Shen P-X; Zhang L-S; Wang X; Shi Z-J Direct arylation of primary and secondary sp3 C–H bonds with diarylhyperiodonium salts via Pd catalysis. Org. Lett 2013, 15, 4758–4761. [DOI] [PubMed] [Google Scholar]
  • 20.The structure of 3 (CCDC 1909063) was confirmed with an X-ray structure determination conducted on colorless needles grown by vapor diffusion of Et2O into a solution of 3 in MeOH. The structure and absolute configuration of 12 (CCDC 1908374) and 13 (CCDC 1908375) were confirmed in single-crystal X-ray structure determinations conducted on crystals grown from Cl2C6H4/heptane (12, colorless needles) or EtOAc/pentane (13, colorless needles) and have been deposited with the Cambridge Crystallographic Data Center.
  • 21.Berger M; Chauhan R; Rodrigues CAB; Maulide N Bridging C–H activation: Mild and versatile cleavage of the 8‐aminoquinoline directing group. Chem. Eur. J 2016, 22, 16805–16808. [DOI] [PubMed] [Google Scholar]
  • 22.Willand-Charnley R; Fisher TJ; Johnson BM; Dussault PH Pyridine is an organocatalyst for the reductive ozonolysis of alkenes. Org. Lett 2012, 14, 2242–2245. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Jackson JJ; Kobayashi H; Steffens SD; Zakarian A 10-Step asymmetric total synthesis and stereochemical elucidation of (+)-dragmacidin D. Angew. Chem. Int. Ed 2015, 54, 9971–9975. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Chuang KV; Kieffer ME; Reisman SE A mild and general Larock indolization protocol for the preparation of unnatural tryptophans. Org. Lett 2016, 18, 4750–4753. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Nicolaou KC; Estrada AA; Zak M; Lee SH; Safina BS A mild and selective method for the hydrolysis of esters with trimethyltin hydroxide. Angew. Chem. Int. Ed 2005, 44, 1378–1382. [DOI] [PubMed] [Google Scholar]
  • 26.Güntert P; Mumenthaler C; Wüthrich K Torsion angle dynamics for NMR structure calculation with the new program DYANA. J. Mol. Biol 1997, 273, 283–298. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supporting Information
NIHMS1053319-supplement-Supporting_Information.pdf (16MB, pdf)

RESOURCES