A new generation of reversible backbone-amide protection for the solid phase synthesis of difficult sequences

Article abstract of DOI:10.1016/S0040-4039(00)00537-2

A versatile safety catch backbone-amide protection system (1, 2) has been developed to inhibit interchain aggregation during solid phase peptide synthesis. The strategy utilises the N(α)-Fmoc derivative of an N-(3- methylsulfinyl-4-methoxy-6-hydroxybenzyl) (SiMB, 2) substituted amino acid (5b), a group which exhibits excellent all round coupling kinetics. The value of these improved derivatives is demonstrated through the syntheses of leucine enkephalinamide (9) and the well known 'difficult sequence' from the acyl carrier protein, residues (65-74) (10) via standard uronium activation and pentafluorophenyl ester chemistry. (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0040-4039(00)00537-2

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 41 (2000) 3997–4001
A new generation of reversible backbone-amide protection for the
solid phase synthesis of difficult sequences
Jo Howe, Martin Quibell and Tony Johnson ∗
Peptide Therapeutics, Peterhouse Technology Park, 100 Fulbourn Road, Cambridge CB1 9PT, UK
Received 23 February 2000; accepted 28 March 2000
Abstract
A versatile safety catch backbone-amide protection system (1, 2) has been developed to inhibit interchain
aggregation during solid phase peptide synthesis. The strategy utilises the Nα-Fmoc derivative of an N-(3-
methylsulfinyl-4-methoxy-6-hydroxybenzyl) (SiMB, 2) substituted amino acid (5b), a group which exhibits
excellent all round coupling kinetics. The value of these improved derivatives is demonstrated through the syntheses
of leucine enkephalinamide (9) and the well known ‘difficult sequence’ from the acyl carrier protein, residues
(65–74) (10) via standard uronium activation and pentafluorophenyl ester chemistry. © 2000 Elsevier Science Ltd.
All rights reserved.
Keywords: amino acids and derivatives; solid phase synthesis; peptides and polypeptides.
In solid phase peptide synthesis ‘difficult sequences’1 are characterised by the sudden appearance of
sluggish deprotection and acylation reactions. It has been proposed that this is a consequence of resin-
bound peptide chain self-association leading to the formation of aggregated secondary structures such as
β-sheets.1 The dramatic reduction in acylation and deprotection rates, which may persist throughout the
remainder of a synthesis, frequently yields poor quality crude products. A major contributing factor to
‘aggregation’ is interchain backbone-amide hydrogen bonding between growing peptide chains.2 Based
upon these observations, we have previously described the N-(2-hydroxy-4-methoxybenzyl) (Hmb, 3)
backbone-amide protecting group,3 for the temporary protection of the secondary amide bond. This was
designed to be compatible with the Fmoc/t-butyl† approach to solid phase peptide synthesis.4 The judicial
use of Hmb backbone protection inhibits the formation of aggregated secondary structures and has
provided excellent improvements in the quality of crude product from many previously poor syntheses.5
∗
†
Corresponding author. Fax: 01223 416300; e-mail: tony.johnson@peptide.co.uk (T. Johnson)
BOP, benzotriazol-1-yloxytris(dimethylamino)phosphonium hexafluorophosphate; DIEA, diisopropylethylamine; Fmoc,
fluoren-9-ylmethoxycarbonyl; Fmoc–Cl, fluoren-9-ylmethoxycarbonylchloroformate; HOBt, 1-hydroxybenzotriazole; MeTf,
methyl triflate; NMM, N-methylmorpholine; TFA, trifluoroacetic acid; TMS–Cl, trimethylsilyl chloride.
0040-4039/00/$ - see front matter © 2000 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(00)00537-2
tetl 16816

3998
Hmb protection is introduced into a peptide chain via N,O-bisFmoc-Nα-(2-hydroxy-4-methoxybenzyl)
amino acid derivatives.3 Direct Nα-acylation of the incorporated Nα-(Hmb) residue is hindered, but is
mechanistically facilitated by initial acylation of the 2-hydroxyl function followed by intramolecular
O→N acyl transfer.3 The optimal reagent for O-acylation of Nα-Hmb protected residues is the sym-
metric anhydride, which, despite the acyl transfer mechanism, may require extended reaction times
for quantitative coupling.3 Thus, for routine use, the major shortcomings of the Hmb group are the
time-consuming preparation of amino acid derivatives and their slow (albeit quantitative) N-acylation.
Building upon the undoubted success of the Hmb group, we have incorporated new design elements into
a second generation of amide-bond protection, SiMB (2), which provides the synthetic benefits afforded
by backbone protection together with improved acylation kinetics.
Our own early investigations and those of others, noted that mono-Nα-Fmoc-Nα-(Hmb)amino acids
(5a), although relatively simple to prepare, formed, upon carboxyl activation, a 4,5-dihydro-8-methoxy-
1,4-benzoxapin-2(3H)-one (6a).6 Species 6a was poorly reactive, which necessitated the use of N,O-
bisFmoc-Nα-(Hmb)amino acid residues for quantitative Hmb amino acid coupling.3 In order to exploit
the relative ease of synthesising Nα-Fmoc-Nα-substituted amino acids, we reasoned that addition of an
electron withdrawing substituent para to the phenolic hydroxyl functionality should improve acylation
kinetics of intermediate benzoxapin-2(3H)-ones (6a→6b) and yield synthetically useful activated amino
acid derivatives. In addition, the rate of intramolecular O→N acyl transfer would also be expected
to increase significantly. An attractive candidate for this purpose, compatible with other procedures
encountered during solid phase peptide synthesis, was an oxidation–reduction system based upon
the readily reversible sulphide (TFA labile, poorly active benzoxapin-2(3H)-one) to sulphone (TFA
stable, highly active benzoxapin-2(3H)-one)7 transition. We initially introduced this safety catch concept
through the 6-hydroxy-5-methyl-1,3-benzoxathiolyl group (4), which gave excellent coupling kinetics
and low epimerisation as the C-terminal residue of a protected peptide segment.8
Recently, we have found that the open chain version of (4), i.e. Fmoc-(SiMB)amino acids (8a and 8b)
are readily prepared (Scheme 1), from the commercially available 6-hydroxybenzoxathialone (Aldrich),
yielding stable, crystalline solids.9
The performance of derivatives 8a and 8b has been investigated in the preparation of two test
sequences (9 and 10). The two diastereoisomers of (9) were initially prepared by standard solid phase

3999
Scheme 1. Reagents and conditions: (i) isobutylene/MeTf; (ii) CH3l/Cs2CO3/2-butanone; (iii) POCl3/DMF; (iv) (a) H-amino
acid-OH/NaOH; (b) NaBH4; (v) (a) TMS-Cl/DIEA; (b) Fmoc-Cl; (vi) NaIO4
peptide synthesis techniques,10 to enable the chiral integrity and acylation kinetics of Nα-Fmoc-Nα-
(SiMB)-L-phenylalanine (8a) to be examined. As expected, amino acid derivative 8a was incorporated
using a standard coupling method (BOP/HOBt/NMM, 5 equiv.) within 45 min. Following Nα-Fmoc
deprotection, the anticipated increase in the rate of acylation was observed with quantitative coupling of
Phe onto Nα-(SiMB)Phe using pentafluorophenyl ester chemistry in 45 mins. It should be noted that the
coupling of Phe to Nα-(Hmb)Phe requires the use of symmetric anhydrides for expended periods (24 h).3
Completion of the assembly followed by cleavage using reductive acidolysis11 gave chirally intact crude
product (Fig. 1a,b).
H-Tyr-Phe-(D/L)Phe-Gly-Leu-NH2 (9)
Fig. 1. Analytical HPLC (Phenomenex Jupiter C4 column: 10% B in A to 90% B in A over 25 min: A=0.1% aq. TFA, B=90%
CH3CN, 10% A: 1.5 ml min−1 flow rate monitoring at 215 nm). Compound 1a crude [Leu5]-enkephalinamide (9: L-Phe): 1b
crude [Leu5]-enkephalinamide (9: L+D-Phe; co-injection of enantiomers)
The second test peptide was chosen to highlight the use of Nα-Fmoc-Nα-(SiMB)-L-alanine (8b) in
the synthesis of the well known difficult sequence from the acyl carrier protein, residues (65–74) (10).12

4000
ACP (65–74) is known to aggregate significantly on deprotection of the penultimate glutamine residue.
Incorporation of the final valine residue is often incomplete, typically reaching 85–90% via traditional
solid phase procedures.12
H- Val65-Gln-Ala-Ala68-Ile69-Asp-Tyr-Ile-Asn-Gly74-NH2 (10)
Using standard pentafluorophenol ester coupling conditions,10 the peptide was assembled upto Ile69.
Ala68 was incorporated by coupling Nα-Fmoc-Nα-(SiMB)-L-alanine (8b) via BOP/HOBt/NMM (5
equiv. 45 mins) activation. Removal of the Fmoc group was followed by acylation of the secondary
amine using Fmoc-Ala-OPfp/HOBt in a standard 45 min coupling. Continuation of the sequence gave
the desired peptide, with no detectable level of the desVal nonapeptide13 (Fig. 2). Amino acid analysis
confirmed the quantitative incorporation of the final valine residue.
Fig. 2. Analytical HPLC (Phenomenex Jupiter C4 column: 10% B in A to 40% B in A over 25 min: A=0.1% aq. TFA, B=90%
CH3CN, 10% A: 1.5 ml min−1 flow rate, monitoring at 215 nm). Crude ACP (65–74) (10)
Based on a sound understanding of backbone protection, as applied to solid phase peptide synthesis,
we have designed and prepared a novel amide protecting group, SiMB (2). Nα-Fmoc-(Nα-SiMB) amino
acid derivatives (5b, 8a and 8b) are prepared readily and couple quantitatively through standard uronium
carboxyl activation (via the intermediate benzoxapin-2(3H)-one). Efficient (quantitative) acylation of
Nα-SiMB amino acids is achieved with commercially available active esters opening the way for fully
automated synthesis of ‘difficult sequences’.
References
1. (a) Kent, S. B. H. In Peptides, Structure and Function; Proceedings of the 9th American Peptide Symposium; Deber, C. M.;
Hruby, V. J.; Kopple, K. D., Eds.; Pierce Chemical Co.: Rockford, IL, 1985; 407. (b) Atherton, E.; Sheppard, R. C. ibid,
415.
2. Kent, S. B. H.; Alewood, D.; Alewood, P.; Baca, M.; Jones, A.; Schnollzer, M. In Innovations and Perspectives in Solid
Phase Synthesis, 2nd International Symposium; Epton, R., Ed.; Intercept: Andover, 1992; 1.
3. Johnson, T.; Quibell, M.; Sheppard, R. C. J. Pept. Sci. 1995, 1, 11. A range of derivatives are commercially available from
Calbiochem-Novabiochem.
4. Atherton, E.; Sheppard, R. C. In Solid Phase Peptide Synthesis: A Practical Approach; Oxford University Press, Oxford,
1989; p. 1.
5. Hyde, C.; Johnson, T. J.; Owen, D.; Quibell, M.; Sheppard, R. C. Int. J. Pept. Protein Res. 1994, 43, 431.
6. (a) Nicolas, E.; Pujades, M.; Bacardit, J.; Giralt, E.; Albericio, F. Tetrahedron Lett. 1997, 38, 2317. (b) Quibell, M.,
unpublished observations, 1993.

4001
7. (a) Patek, M.; Lebl, M. Tetrahedron Lett, 1991, 32, 3891. (b) Kiso, Y.; Fukui, T.; Tanaka, S.; Kimura, T.; Akaji, K.
Tetrahedron Lett. 1994, 35, 3571.
8. Offer, J.; Johnson, T.; Quibell, M. Tetrahedron Lett. 1997, 38, 9047.
9. Compound (8a) was prepared in 37% yield from resorcinol. Analytical HPLC showed the presence of a single peak with
ESMS 586.1 (MH+); 1171.3 (2MH+), calculated mass: C33H31NO7S, 585.6.
10. All work was carried out using Novasyn® TGR resin (Novabiochem). Couplings were carried out in DMF using 5 equiv.
activated amino acid derivatives. Coupling yields were quantified by amino acid analysis (Biochrom 20 analyser).
11. Reductive acidolysis conditions; peptide-resin was treated with SiCl4:TFA:anisole:EDT (5:90:2.5:2.5 v/v) for 2 h at room
temperature. Following filtration, TFA was removed by nitrogen sparging, peptides were precipitated and washed with tert-
butylmethyl ether precipitation. Peptide 9 via derivative 8a, gave a single species (see Fig. 1a) with ESMS 645.5 (MH+)
and 1288.9 (2MH+) (expected MH+ 645.8). Amino acid analysis: Gly 1.05 (1), Leu 1.01 (1), Tyr 1.00 (1), Phe 1.93 (2).
12. (a) Hancock, W. S.; Prescott, D. J.; Vagelos, P. R.; Marshall, G. R. J. Org. Chem. 1973, 38, 774. (b) Johnson, T. J.; Quibell,
M.; Owen, D.; Sheppard, R. C. J. Chem. Soc., Chem. Commun. 1993, 4, 369.
13. Peptide 10 via derivative 8b gave a single species (see Fig. 2) with ESMS 1062.7 (MH+) (expected MH+ 1063.2). Amino
acid analysis: found (expected): Asp 1.96 (2), Glu 0.92 (1), Gly 1.05 (1), Ala 2.05 (2), Val 0.98 (1), Ile 2.08 (2), Tyr 1.04
(1).