Solid phase synthesis of α-amino squaric acid-containing peptides

Abstract

A new method has been developed for the synthesis of 3-(1-aminoalkyl)-4-hydroxycyclobut-3-ene-1,2-dione [α-amino squaric acid (α-Asq)]-containing peptides using solid phase peptide synthesis according to an Fmoc protecting group strategy. FmocHN-Gly-[α-Asq]-Oi-Pr was successfully used as a coupling unit in this method, which allowed for the construction of α-Asq-containing hexapeptide libraries including Sq-Gly and Sq-Phe units using the Wang-resin. Peptides containing the α-Asq moiety exhibited inhibitory activity towards a digestive enzyme.

---

The solid phase synthesis of peptides using the Fmoc protecting group (i.e., the Fmoc strategy) represents a robust technology for the rapid construction of various peptides, which is also amenable to the automated synthesis of peptide libraries.¹ The application of this technique to the synthesis of peptides containing amino acid isosteres has recently attracted considerable levels of attention because the resulting peptide analogues exhibit improved levels of potency and selectivity towards biological targets, as well as enhanced metabolic and chemical stabilities compared with the corresponding unmodified peptides. Several amino acid isosteres have been developed for use in the solid phase synthesis of peptide analogues.²

We recently reported 3-(1-aminoalkyl)-4-hydroxycyclobut-3-ene-1,2-dione [α-amino squaric acid (α-Asq)] (1) as a novel amino acid analogue, where the carboxylic acid moiety of the α-amino acid had been replaced with a 4-hydroxy-2,3-dioxocyclobut-1-enyl (Sq) group (Fig. 1).³ The Sq group and squaric acid 2 have been studied extensively as carboxylic acid surrogates in medicinal chemistry, bioorganic chemistry, chemical biology, and materials chemistry.³ It was envisaged that the incorporation of α-Asq into peptides would allow for the preparation of novel α-Asq-containing peptides, because the inherent reactivity of the Sq group would allow this moiety to react with various amines to form amide like N-Sq bonds.^3,4 In our previous study, we reported for the first time the synthesis of the Sq-Gly-containing Leu-enkephalin analogue 3 (Fig. 1),^4b which exhibited promising binding activities towards opioid receptors. This compound was also found to be stable under acidic conditions (i.e., 30% HBr/AcOH), as well as being stable to the conditions required for its purification by HPLC (i.e., H₂O/MeCN/TFA), and its subsequent analysis by NMR spectroscopy in DMSO-d₆ and D₂O.^4b These observations led us to explore the possibility of developing a method for the rapid construction of α-Asq-containing peptides using the Fmoc strategy. Herein, we report the first solid phase synthesis of α-Asq-containing peptides and their application to the construction of peptide analogue libraries containing the α-Asq unit.

Fig. 1 Structures of α-amino acid, α-Asq (1), squaric acid (2), and α-Asq-containing Leu-enkephalin (3).

FmocHN-[α-Asq]-Oi-Pr 4 was selected as a coupling unit for the Fmoc strategy used in this study (Scheme 1). The synthesis of FmocHN-[Sq-Gly]-Oi-Pr 9 was commenced with N-Cbz-hydroxycyclobutenone 5.^4b Removal of the Cbz group from 5 followed by protection of the resulting free amino group with the Fmoc group gave 7 in 81% yield, which was converted to N-Fmoc α-Sq glycine 8 in 84% yield using the conventional condition in the presence of a small amount of 12 N HCl.^3,4 Subsequent removal of the t-butyl group in compound 8 provided the corresponding carboxylate, which underwent a spontaneous Sq group-promoted decarboxylation reaction⁴ in situ to give FmocHN-[Sq-Gly]-Oi-Pr 9 in 76% yield.

The Fmoc strategy used in the current study consists of a series of sequential transformations, including (i) linkage of the Fmoc-protected amino acids (FmocHN-AA-OH) to the acid-labile linker on solid phase; (ii) removal of the Fmoc protecting group from the AA; (iii) condensation of FmocHN-AA-OH; (iv) repetition of steps (ii) and (iii) to elongate the peptide chain; and (v) treatment of the resulting peptide with acid to cleave the peptide from the resin. We initially investigated the compatibility of FmocHN-[Sq-Gly]-Oi-Pr 9 under the proposed reaction conditions prior to the solid phase synthesis using a model system involving the dipeptide analogue FmocHN-[Sq-Gly]-Phe-OBn 10. Treatment of 9 with H₂N–Phe-OBn gave 10 in 71% yield (Scheme 1). Our initial attempt to remove the Fmoc group in 10 using 20% piperidine in THF resulted in a complex mixture. We also investigated the use of a variety of different reagents and conditions [e.g., Et₂NH, DBU, 2,2,6,6-tetramethyl piperidine, lower temperature, different concentrations, in situ trapping with Boc₂O and Ac₂O], but none of these reactions resulted in the formation of the desired amine 12. 9-Methylidenefluorene (11) was detected as major by-product in the complex mixture resulting from this reaction, which indicated that the Fmoc group was being liberated during the course of the reaction to give amine 12, which must have underwent rapid decomposition.

Scheme 1 Reaction conditions: (a) H₂/Pd–C (20 wt%), MeOH, rt, 3 h; (b) FmocOSu (1 eq.), MeCN, rt, 4 h; (c) 12 N HCl (1 eq.), CH₂Cl₂, rt, 3 h; (d) TFA (30 eq.), CH₂Cl₂, 0 °C to rt, 12 h.

It was envisaged that the protection of the free amine group as an amide such as dipeptide unit 13 would overcome the decomposition issues observed above and allow for the use for the solid phase synthesis (Scheme 2). FmocHN-Gly-[Sq-Gly]-Oi-Pr 16 was synthesized from compound 5 according to the route shown below.^4a Compound 6 was coupled with FmocHN-Gly-OH using EDCI to give 14 in 98% yield. Compound 14 was then converted to 16 in a similar manner to that used for the synthesis of 9 via a Sq group-promoted decarboxylation reaction. FmocHN-Gly-[Sq-Gly]-Oi-Pr 16 was then condensed with H₂N–Phe-OBn to give the model tripeptide analogue 17 in 64% yield. Compound 17 was treated with Et₂NH to allow for the removal of the Fmoc protecting group, and the resulting amine was acetylated with Ac₂O to give the corresponding amide 18 in 72% yield, without decomposition of the free amine (Scheme 2).

Scheme 2 Reaction conditions: (a) H₂/Pd–C (20 wt%), MeOH, rt, 3 h; (b) FmocHN-Gly-OH (1 eq.), EDCI (1 eq.), THF, rt, 3 h; (c) 12 N HCl (1 eq.), CH₂Cl₂, rt, 3 h; (d) TFA (30 eq.), CH₂Cl₂, 0 °C to rt, 12 h; (e) H₂N–Phe-OH (3 eq.), THF, rt, 3 h; (f) 20% Et₂NH/THF, rt, 0.5 h then Ac₂O, THF, rt, 0.5 h.

With a suitably stable linkage unit in hand, we proceeded to evaluate the use of solid-supported 16 in a series of sequential transformations, including (i) removal of the Fmoc group of the FmocHN-Phe-functionalised Merrifield resin; (ii) treatment of the resulting amine with 2 equivalents of 16; and (iii) cleavage of the resulting peptide from the resin by treatment with TFA (Scheme 3).

Scheme 3 Solid phase support of dipeptide unit 16.

The yield of FmocHN-Gly-[Sq-Gly]-Phe-OH 19 following the cleavage of the product from the resin was used to evaluate the linkage efficiency of the peptide to the resin. Unfortunately, however, this protocol only resulted in the formation of a trace amount of 19. Increasing the amount of 16, as well as extending the reaction period or switching to trityl resin and Clear resin® resulted in similarly low yields of the product. Following an extensive period of investigation, we found that the Wang resin® gave the best results of all of the resins tested, with 19 being isolated in 43% yield following a single loading of 16. The coupling efficiency could be further improved by twice repeating the loading of 16. Under the optimized conditions, 19 was isolated in 80% yield following its cleavage from the resin.

The α-substituted α-Asq, Sq-Phe, could also be incorporated into the peptide using the Fmoc strategy. To shorten the number of synthetic steps required of this protocol, we investigated the use of the dipeptide unit 20 bearing a –COOt-Bu group at the α-position of Sq-Phe as the coupling unit.⁵ It was envisaged that the –COOt-Bu group could be readily removed under the conditions used for the cleavage of the peptide from the resin using TFA, which would also allow for the removal of the t-Bu group and decarboxylation.

Compound 20 was prepared from 15 in 90% yield following a benzylation reaction^4c in the presence of Et₃N. The suitability of 20 for use in the solid phase synthesis protocol was evaluated by the synthesis of Leu-enkephalin analogue 21. Treatment of the Sq-Phe-containing dipeptide unit 20 (4 equiv.) with the Wang resin in the presence of DIEA (2 equiv.) allowed for the dipeptide to be efficiently immobilized on the resin. In this case, excess amount of 20 could be recovered cleanly from the filtrate after coupling reaction. The remaining two amino acid residues Gly and Tyr were successfully condensed with the resin-supported peptide using HOBt (4 equiv.) and HBTU (3.6 equiv.). Treatment of the peptide analogue-linked resin with TFA provided the Sp-Phe-containing enkephalin analogue 21 in an overall yield of 30% (Scheme 4).

Scheme 4 Synthesis of Sq-Phe-containing peptide. Reaction conditions: (a) 20% piperidine/DMF, rt, 15 min; (b) 20 (4 eq.), DIEA (2 eq.), AcOEt, rt, 60 h; (c) 20% Et₂NH/DMF, rt, 15 min; (d) FmocHN-Gly-OH (4 eq.), HOBt (4 eq.), HBTU (3.6 eq.), DIEA (8 eq.), DMF, rt, 1 h; (e) 20% Et₂NH/DMF, rt, 15 min; (f) FmocHN-Tyr(t-Bu)-OH (4 eq.), HOBt (4 eq.), HBTU (3.6 eq.), DIEA (8 eq.), DMF, rt, 1 h; (g) 20% Et₂NH/DMF, rt, 15 min; (h)TFA, rt, 3 h.

The synthetic strategy described above was extended to the construction of an α-Asq-containing peptide library using the dipeptide units 16 and 20 (Scheme 5). Starting from FmocHN-Phe-Wang resin, a mixture of FmocHN-Phe-OH, FmocHN-Tyr(t-Bu)-OH, FmocHN-His(Trt)-OH and dipeptide unit 16 were linked on the solid phase. The three peptide libraries 22, 23, and 24, which differed only in the positioning of the Gly-[Sq-Gly] unit, were generated in a parallel manner. AA in the each library represents either Phe, Tyr, or His. The composition of each peptide library (including 27 hexapeptide analogues) was confirmed by MALDI-TOF MS analysis (Scheme SI-1†). The Sq-Phe-containing peptide libraries 25, 26, and 27 were successfully synthesized in a similar manner (Scheme SI-2†).

Scheme 5 Syntheis of α-Asq-containing eptide libraries 22–27. Reagents and coupling sequences are depicted in ESI.†

The degradation of biologically active peptides by digestive enzymes represents a significant limitation to the development of peptide drugs. The incorporation of an amino acid isostere into biologically active peptides has emerged as an effective strategy for increasing the metabolic stability of peptides, and it was envisaged that the use of the α-Asq moiety in this regard would enhance the stability of the resulting peptides towards digestive enzymes. To evaluate the feasibility of this strategy, we developed a model study using peptide 28 and carboxypeptidase Y (Fig. 2). The α-Asq-containing peptide 28 was prepared using solid phase synthesis (Scheme SI-3†) and subsequently subjected to an enzymatic hydrolysis reaction using carboxypeptidase Y. The time course of the enzymatic hydrolysis reaction was monitored by MALDI-TOF MS (Fig. 2). After 2 hours, all of the starting peptide was converted to peptide 29 [m/z 880], where the three amino acid residues –Ser–Tyr–Phe-OH had been cleaved from the C-terminal of 28 (Fig. 2(c)). These results indicated that the α-Asq moiety was stable to carboxypeptidase Y. The α-Asq-containing peptide libraries was employed to random screening using two assay systems: (i) rat hepatoma cells (dRLh-8) proliferation effects and (ii) inihibitory effects toward acetylcholinesterase (see SI†). Among them, peptide library 23 exhibited moderate inhibitory activity towards dRLh8 cancer cells at 100 μM whereas the other libraries appeared to be inactive at this concentration. None of the libraries exhibited inhibitory activity towards acetylcholinesterase at a dose of 100 μg mL⁻¹.

Fig. 2 Enzymatic hydrolysis reaction of peptide 28 with carboxypeptidase Y. MALDI-TOF MS spectra of the crude mixture at (a) 0 h, (b) 0.5 h, (c) 2 h. Matrix: α-CHCA, results recorded in the positive ion mode.

In summary, we have developed a method for the solid phase synthesis of α-Asq-containing peptides and peptide libraries. We examined the inhibitory effect of α-Asq towards protease using carboxypeptidase Y for the first time. The α-Asq moiety was found to be resistant to carboxypeptidase Y. This results would aid to the design and synthesis of stable α-Asq-incorporating peptides. The biological screening of the libraries generated in this study revealed that 23 exhibited moderate inhibitory activity towards dRLh-8 cells at 100 μM. Further library screening studies, and studies directed towards the application of novel peptide analogues in chemical biology,⁶ are currently underway in our laboratory.

Acknowledgements

This work was financially supported by grants from the Ministry of Education, Culture, Sports, Science and Technology (MEXT) (no. 23102009), and the Japan Society for the Promotion of Science (KAKENHI, Grant no. 23228001 & 25282233).

Notes and references

1. (a) M. Gongora-Benitez, J. Tulla-Puche and F. Albericio, ACS Comb. Sci., 2013, 15, 217; (b) Amino Acids, Peptides and Proteins in Organic Chemistry, Protection Reactions, Medicinal Chemistry, Combinatorial Synthesis, ed. A. B. Hughes, WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim, 2011, vol. 4; (c) Peptides: Chemistry and Biology, ed. N. Sewald and H.-D. Jakubke, WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim, 2nd edn, 2009.
2. (a) O. Melnyk, N. Ollivier, S. Besret and P. Melnyk, Bioconjugate Chem., 2014, 25, 629; (b) L. Kiss and F. Fülöp, Chem. Rev., 2014, 114, 1116; (c) Y. Ishibashi and M. Kitamura, Chem. Commun., 2009, 6985; (d) P. Cudic and M. Stawikowski, Peptidomimetics: Fmoc Solid-Phase Pseudopeptide Synthesis, in Peptide-Based Drug Design, ed. L. Otvos, Humana Press, 2008, p. 223; (e) N. G. Angelo and P. S. Arora, J. Am. Chem. Soc., 2005, 127, 17134.
3. For reviews: (a) F. R. Wurm and H.-A. Klok, Chem. Soc. Rev., 2013, 42, 8220; (b) C. Ballatore, D. M. Huryn and A. B. Smith III, ChemMedChem, 2013, 8, 385; (c) C. M. Marson, Chem. Soc. Rev., 2011, 40, 5514; (d) R. I. Storer, C. Aciro and L. H. Jones, Chem. Soc. Rev., 2011, 40, 2330; (e) T. Shinada, T. Ishida and Y. Ohfune, J. Synth. Org. Chem., Jpn., 2007, 65, 32; (f) M. Ohno and S. Eguchi, in Bioactive Heterocycles I: Topics in Heterocyclic Chemistry, ed. S. Eguchi, Springer, Berlin, 2006, vol. 6, pp. 1–37; (g) G. Seitz and P. Imming, Chem. Rev., 1992, 92, 1227; (h) F. Serratosa, Acc. Chem. Res., 1983, 16, 170; (i) A. H. Schmidt, Synthesis, 1980, 961; (j) R. West, in Oxocarbons, Academic, New York, 1980; (k) R. West and J. Niu, in Nonbenzenoid Aromatics, ed. J. P. Snyder, Academic, New York, 1969, p. 311.
4. (a) T. Shinada, A. Yamasaki, Y. Kiniwa, K. Shimamoto and Y. Ohfune, Tetrahedron Lett., 2009, 50, 4354; (b) T. Shinada, T. Ishida, K. Hayashi, Y. Yoshida, Y. Shigeri and Y. Ohfune, Tetrahedron Lett., 2007, 48, 7614; (c) T. Ishida, T. Shinada and Y. Ohfune, Synthesis, 2005, 2723; (d) T. Ishida, T. Shinada and Y. Ohfune, Tetrahedron Lett., 2005, 46, 311; (e) T. Shinada, K. Hayashi, T. Hayashi, Y. Yoshida, M. Horikawa, K. Shimamoto, Y. Shigeri, N. Yumoto and Y. Ohfune, Org. Lett., 1999, 1, 1663.
5. Compound 15 was labile under the coupling reaction conditions (basic conditions) to give a complex mixture due provably to the presence of a reactive active methine in the molecule. On the other hand, compound 20 is more stable due to the protecting effect of the active methine with a benzyl group.
6. (a) S. Ziegler, V. Pries, C. Hedberg and H. Waldmann, Angew. Chem., Int. Ed., 2013, 52, 2744; (b) M. Ueda, Chem. Lett., 2012, 41, 658.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c4ra10442k

---