Key words
arylmethylprolines - prolines - hydroboration - Suzuki cross-coupling
Proline is the only proteogenic amino acid with a cyclic backbone, which confers to this residue a uniquely restricted conformation. Nature and scientists have used proline and its derivatives to regulate numerous processes, ranging from ion-gating and the structural integrity of skin to asymmetric catalysis.[1]
[2]
[3]
[4] The development of proline analogues and their incorporation into peptides and other compounds is therefore of great interest. Proline derivatives with different substituents at Cγ are the most common, due to their natural occurrence and the ease of functionalization of (2S,4R)-4-hydroxyproline.[1]
[5] Examples include derivatives with heteroatoms at Cγ, e.g., F, Cl, N3, NH2, or alkyl groups, e.g., Me and
t
Bu.[1]
[6] In contrast, derivatives with arylmethyl substituents at Cγ are less commonly utilized, possibly due to a lack of a straightforward synthetic route.
We became interested in proline derivatives bearing naphthyl moieties, for their value in the molecular recognition of RNA.[7] Synthetic routes have been reported for the functionalization of proline at Cγ with benzylic or indolylmethyl substituents.[6a]
[8]
[9] However, we had limited success in transferring these reaction conditions, which rely on Wittig reactions of 4-oxoproline followed by hydrogenation, to larger aryl moieties (Scheme [1], top).
Scheme [
1
] Synthetic routes to 4-(arylmethyl)proline derivatives
We therefore sought an alternative route and we envisioned Suzuki reactions between an organoborane–proline derivative and aryl halides as a strategy that might provide access to proline derivatives with various aryl groups (Scheme [1], bottom). Here, we report a general synthetic route to arylmethyl proline derivatives that permits the introduction of a broad range of aryl moieties at Cγ.
Our synthetic route relies on the hydroboration of the Boc/
t
Bu-protected 4-methyleneproline 5, which was obtained from (2S,4R)-4-hydroxyproline (1) by slight modification of a previously published procedure (Scheme [2]).[10] This four-step synthesis started with Boc-protection of 1, followed by oxidation to ketone 3, protection of the carboxylic acid as the
t
Bu ester in 4, and introduction of an exocyclic methylene group by a Wittig reaction.[11]
Scheme 2 Synthesis of the common precursor tert-butyl N-(tert-butoxycarbonyl)-4-methyleneprolinate (5)
Hydroboration of the 4-methyleneproline 5 with 9-BBN provided the organoborane 6, which was used for the Suzuki reaction without further purification (Scheme [3], top). For the Suzuki reaction, various catalysts and conditions were explored by using 2-bromonaphthalene as a model aryl bromide. We focused in particular on catalysts that had proven valuable for cross-couplings with other amino acid derivatives (Scheme [3], bottom).[12] Among the tested palladium-based catalysts, reactions with PEPPSI[13] showed the highest conversion of 5 and 2-bromonaphthalene into the Suzuki reaction product 7a. Under optimized conditions [5 M aq KOH, ArBr (1.3 equiv), PEPPSI(3% mol)], the 4-(2-naphthylmethyl)proline derivative 7a was obtained in a yield of 83%. Note that 3 mol% of PEPPSI was enough to obtain these results. Because PEPPSI is more air-stable than other palladium catalysts,[14] this catalyst was used for all further experiments.
Scheme 3 Top: Suzuki cross-coupling reaction to yield various 4-(arylmethyl)proline derivatives 7a–f. Bottom: Catalysts tested in the Suzuki cross-coupling reaction.
Reassuringly, this route also permitted the synthesis of proline derivatives bearing substituted naphthyl moieties (7b and 7c) as well as phenyl (7d), 9-anthryl (7e), or pyren-1-yl (7f) substituents in good overall yields (60–83%; Scheme [3]).[15] All derivatives were obtained with a diastereoselectivity of ~3:2 in favor of the syn-product, as determined by analysis of 1H NMR NOE spectroscopy.[11]
Because peptide syntheses typically require Fmoc-protected amino acids, we converted 7a–c into the respective Fmoc-amino acids 8a–c. Simultaneous removal of the
t
Bu protecting groups in 6 M HCl in 1,4-dioxane, and subsequent Fmoc-protection afforded 8a–c in yields of 74–89% (Scheme [4]). The diastereoisomers were separated by preparative reverse-phase HPLC to obtain enantiomerically pure amino acids at a scale of up to 2.5 g.[15]
[16]
Scheme 4 Synthesis of Fmoc-protected amino acids 8a–c
In conclusion, we have introduced a synthetic route to access proline derivatives bearing a variety of arylmethyl substituents at the γ-position. The products were obtained in good yields for every tested aromatic moiety. The diastereoselectivity of the hydroboration step was modest, but the diastereoisomeric products could be separated on a gram scale. Installation of a Fmoc-protecting group was straightforward. Thus, the route provides access to proline derivatives with a variety of arylmethyl moieties at Cγ that are suitably protected for solid-phase peptide synthesis. We envision these derivatives as being valuable additions to the toolkit of proline analogues for applications in chemistry and chemical biology.
