Deprotection of N-Alloc amines by Pd(0)/DABCO - An efficient method for in situ peptide coupling of labile amino acids

Article abstract of DOI:10.1016/S0040-4039(01)01453-8

A highly efficient one-pot deprotection/peptide coupling protocol of N-Alloc amino acids with activated N-Boc or N-Fmoc amino acids was developed in solution and on solid phase. DABCO was found to be especially effective for the deprotection of the N-Allo

Full text of DOI:10.1016/S0040-4039(01)01453-8

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 42 (2001) 7049–7053
Deprotection of N-Alloc amines by Pd(0)/DABCO—an efficient
method for in situ peptide coupling of labile amino acids
Chiara Zorn,a Frieder Gnad,a Sunnhild Salmen,a Timothy Herpinb,† and Oliver Reisera,*
aInstitut fu¨r Organische Chemie, Universita¨t Regensburg, Universita¨tsstraße 31, D-93040 Regensburg, Germany
bRhone-Poulenc Rorer Research and Development, 500 Arcola Road, Collegeville, PA 19426, USA
Received 24 July 2001; revised 7 August 2001; accepted 8 August 2001
Abstract—A highly efficient one-pot deprotection/peptide coupling protocol of N-Alloc amino acids with activated N-Boc or
N-Fmoc amino acids was developed in solution and on solid phase. DABCO was found to be especially effective for the
deprotection of the N-Alloc group, resulting in short reaction times (10–20 min) and allowing the coupling of amino acids that
are unstable in unprotected forms. © 2001 Elsevier Science Ltd. All rights reserved.
The synthesis of amides by coupling of a carboxylic
acid with an amine is well established, most notably in
peptide synthesis. Commonly, a carboxylic acid is acti-
vated by appropriate reagents in situ and is then
reacted with an unprotected amine.1 However, if the
amine is difficult to prepare or even unstable in unpro-
tected form, a technique which allows its in situ genera-
tion and coupling with a carboxylic acid is required.
For example, b-aminocyclopropane carboxylic acids2
(b-ACCs), e.g. 1,3 have been recognized recently as
conformationally rigid amino acids for the synthesis of
peptides.4,5 Since donor–acceptor 1,2-disubstituted
cyclopropanes rapidly undergo ring opening,6 deriva-
tives of 1 are only stable if an electron-withdrawing
group protects the amino functionality.7 We were able
to couple 1a with amino acids via its ammonium chlo-
ride salt (Scheme 1);5b however, this strategy is not
compatible with acid-sensitive groups that are com-
monly encountered as protecting groups of amino acid
side chains. Moreover, the protocol requires saturated
HCl gas in ethyl acetate, which is also not amenable to
automated solid-phase synthesis.
synthesis. However, deprotection of Fmoc by base in
the presence of activated carboxylic acids is inherently
problematic and, indeed, we were unsuccessful when
trying to convert derivative 1b to 2 under various
conditions for Fmoc cleavage (TBAF, piperidine, mor-
pholine); only ring opening products of the cyclo-
propane moiety could be obtained (Scheme 1).
We then turned our attention to the N-Alloc protecting
group which is known to be readily cleaved by catalytic
amounts of palladium(0) in the presence of a nucle-
ophile.8 Among the allyl scavengers that have been
already used in the context of peptide chemistry for
N-Alloc cleavage,9 metal hydrides10 and especially
phenylsilane introduced by Guibe´11 seemed to be most
promising for our purpose, since it has already been
successfully employed for domino deprotection/cou-
pling reactions with active esters of amino acids, giving
excellent results. In situ activation of the amino acid to
be coupled is also possible, as was shown by Hiemstra
and Speckamp10a,b or Zhu,10c respectively. However, all
examples reported so far use N-Alloc amino acids that
O
Consequently, we were looking for a milder method
that would allow in situ deprotection and coupling of
amines with carboxylic acids. An attractive choice
would have been a strategy based on the Fmoc group,
taking into account its great importance for peptide
a (1a)
NHR
NH
R
b (1b)
BnO2C
CO2Me
BnO2C
CO2Me
1a: R = Boc
2
1b: R = Fmoc
* Corresponding author. Tel.: +49-941-9434631; fax: +49-941-
9434121; e-mail: oliver.reiser@chemie.uni-regensburg.de
Scheme 1. (a) (i) 1a, HCl/EtOAc, (ii) N-Boc-XXa, EDC,
HOBt, 76–86%; (b) 1b, base (TBAF, piperidine, or morpho-
line), RCOX (X=Cl, OBt)“decomposition.
† Present address: Bristol-Myers Squibb Pharmaceutical Research
Institute, PO Box 5400, Princeton, NJ 08543-5400, USA.
0040-4039/01/$ - see front matter © 2001 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(01)01453-8

7050
C. Zorn et al. / Tetrahedron Letters 42 (2001) 7049–7053
Using DABCO instead of pyridine, however, dramati-
cally accelerated the deprotection/coupling sequence,
giving rise to 5 after a reaction time of only 10–20 min
when 10 mol% of palladium(0) were employed (87–99%
yields, entries 5–8).14 Most notably, under these reac-
tion conditions N-Fmoc-protected amino acids are
again tolerated as coupling partners (entry 8), so that
the synthesis of peptide fragments suitable for Fmoc
strategy becomes possible. Moreover, the protection of
carboxylic acids as benzyl esters is possible (entries 1–3,
5, and 6), since the latter functionality is not cleaved by
Pd(0) using the protocol described here. With both
protocols, little (52%) or no racemization was
observed in the coupling step.
are stable in N-unprotected forms and, therefore, do
not require an in situ-coupling protocol.
On the other hand, morpholine, pyrrolidine or pipe-
ridine have been shown to be also highly effective for
the deprotection of the Alloc group in peptide synthe-
sis,12 having the distinct advantage of hydrolytic stabil-
ity—especially in the presence of transition
metals—and lower cost in comparison to phenylsilane.
Although the presence of a secondary amine should not
be compatible with amino acids that are Fmoc pro-
tected at the N-terminus or activated with a carbodi-
imide/HOBt13 at the C-terminus, we envisioned that a
tertiary amine such as DABCO13 or pyridine would
either not interfere with the amino acid or could even
aid its activation. Moreover, these reagents should offer
similar operational advantages like secondary amines,
and indeed, pyridine, triethylamine, or N-methylmor-
pholine have been used as allyl scavengers before,8
albeit not in a one-pot peptide-coupling protocol.
While the examples given so far have only been advan-
tageous from an operational point of view due to the
convenience of a one-pot deprotection/coupling pro-
cess, a more challenging task from a synthetic point of
view is the coupling of b-ACC derivatives, such as 6,
being unstable in the N-unprotected form.
Using N-Alloc-alanine benzylester (3a) and N-Alloc-
valine methylester (3b) as model compounds, we first
tested if pyridine or DABCO could be used for N-Alloc
deprotection in the presence of an activated carboxylic
acid to ultimately achieve peptide coupling (Scheme 2
and Table 1). Indeed, the dipeptides 5 were obtained in
high yields (94–99%) using pyridine/Pd[PPh3]4 as the
deprotecting agent in the presence of N-protected
amino acids activated by EDC/HOBt.13 N-Boc- and
especially N-Fmoc-protected amino acids are suitable
coupling partners. Nevertheless, the reactions pro-
ceeded slowly (entry 1) making 20 mol% of palladium
catalyst necessary to stay within acceptable reaction
times (entries 2–4).
Gratifyingly, both protocols were also applicable for
the coupling of 6 (Table 2); the dipeptides 7 could be
obtained in high yields (92–97%). Again the use of
DABCO resulted in considerable shortened reaction
times, but another advantage of this procedure in com-
parison to the use of pyridine became apparent:
although ring opening of the in situ-generated cyclo-
propylamino carboxylic acid could be suppressed in
both cases, epimerization of the stereocenter bearing
the amino group was observed to a substantial extent
(up to 20%) when pyridine was employed. In contrast,
epimerization was minimal (<3%) when DABCO was
used, indicating that this reagent is not responsible
solely for the cleavage of the allyl group but might also
activate the HOBt ester of the amino acid used as the
coupling partner.
O
Pd[PPh3]4
NHPg
CO2H
NHAlloc
CO2R1
NHPg
scavenger
+
HN
Using phenyl silane as allyl scavenger, the coupling of 6
with N-Fmoc-alanine not only afforded a good yield
(90%) of 7 (entry 7), but also 5% of the ring-opening
product, which made the purification of the dipeptide
difficult.
R2
R2
CO2R1
R
EDC/HOBt
CH2Cl2
R
3a: R=Me, R1=Bn
3b: R=iPr, R1=Me
5
4a: R2=Me, Pg=Boc
4b: R2=Bn, Pg=Boc
4c: R2=Me, Pg=Fmoc
We were also able to apply the in situ-coupling proto-
col to solid-phase synthesis, thereby demonstrating that
the building block 8 can be used as an unnatural amino
Scheme 2. One-pot peptide coupling of amino acids 3 and 4
(cf. Table 1).
Table 1. One-pot peptide coupling of amino acids 3 and 4 (cf. Scheme 2)
Entry
Amino acidsa
Scavengerb
Pd(PPh3)4 (mol%)
Time (min)
Yield (%)
1
2
3
4
5
6
7
8
3a
3a
3a
3b
3a
3a
3b
3b
4a
4b
4c
4c
4a
4b
4a
4c
Py
Py
Py
Py
DABCO
DABCO
DABCO
DABCO
10
20
20
20
10
10
10
10
900
90
90
90
10
10
20
20
94
99
94
98
99
99
97
87
a 3 equiv. 4a,4b or 1.5 equiv. 4c, respectively, were employed.
b 20 equiv. pyridine or 5 equiv. DABCO were employed.

C. Zorn et al. / Tetrahedron Letters 42 (2001) 7049–7053
Table 2. One-pot peptide coupling of amino acids 4 and 6 (cf. Scheme 3)
7051
Entry
Amino acida
Scavengerb
Pd(PPh3)4 (mol%)
Time (min)
Yield (%)
1
2
3
4
5
6
7
4a
4d
4a
4b
4c
4d
4c
Py
Py
20
10
10
10
10
10
10
900
120
15
15
15
93
95
96
93
92
97
90c
DABCO
DABCO
DABCO
DABCO
PhSiH3
15
15
a 3 equiv. 4a,4b or 1.5 equiv. 4c, respectively, were employed.
b 10 equiv. pyridine or 5 equiv. DABCO or PhSiH3 were employed.
c Contaminated with 5% of the ring-opening product of 6.
O
PgHN
Pd[PPh3]4
NHAlloc
NH
NHPg
CO2H
scavenger
R2
+
R2
EDC/HOBt
CH2Cl2
BnO2C
E
BnO2C
E
E=CO2Me
6
4a-c
4d: R2=H, Pg=Boc
7
Scheme 3. One-pot peptide coupling of amino acids 4 and 6
(cf. Table 2).
Scheme 5. (a) (i) Fmoc-Pro-OH, DIC, HOBt, DMF, (ii)
piperidine, DMF, (iii) Alloc-Pro-OH, DIC, HOBt, DMF; (b)
Pd[PPh3]4, DABCO or PhSiH3 (12 equiv.), Fmoc-Ala-OH,
EDC, HOBt, CH2Cl2, 2 h, rt; (c) TFA, CH2Cl2 2:1; isolated
yield based on loading of Phe to the resin: 99% (DABCO),
99% (PhSiH3).
acid in automated peptide synthesis (Scheme 4). Using
standard DIC/HOBt13 peptide-coupling techniques,
dipeptide 9 was synthesized on a Wang™ resin with an
Advanced Chem Tech ACT-90 synthesizer. Palla-
dium(0)-catalyzed deprotection with DABCO or
phenylsilane in the presence of Fmoc-Ala-OH resulted
in the resin-bound peptide 10. Upon cleavage from the
resin with trifluoroacetic acid, tripeptide 11 was formed
without any detectable epimerization in 83–85% iso-
lated yield. While we chose to add PhSiH3 manually15
just prior to the Alloc-deprotection/coupling step to
insure its integrity, we used a solution of DABCO in
CH2Cl2 without special precaution other than carrying
out the reaction under a nitrogen atmosphere within
the setup of an automated peptide synthesizer.
Scheme 4. (a) (i) Fmoc-PheOH, DIC, HOBt, DMF, (ii) pipe-
ridine, DMF, (iii) 8, DIC, HOBt, DMF; (b) Pd[PPh3]4,
DABCO or PhSiH3 (12 equiv.), Fmoc-Ala-OH, EDC, HOBt,
CH2Cl2, 2 h, rt; (c) TFA, CH2Cl2 2:1; isolated yield based on
loading of Fmoc-Phe-OH to the resin: 85% (DABCO), 83%
(PhSiH3).
Finally, the peptide protocols described here also
allowed the linear coupling of amino acids on solid
phase between the second and third position. This
transformation is particularly difficult to accomplish

7052
C. Zorn et al. / Tetrahedron Letters 42 (2001) 7049–7053
since intramolecular diketopiperazine (DKP) formation
with concurrent cleavage from the resin, being initiated
by the free N-terminus of the peptide, is often encoun-
tered.16 The Alloc tandem deprotection/coupling strat-
egy was already successfully applied to suppress DKP
formation.11b We demonstrate here the application of
this strategy in combination with the in situ activation
of the amino acid towards dipeptides containing proline
residues, which are especially prone to DKP formation.
Gratifyingly, even the N-Alloc-protected Pro-Pro
dipeptide 12 smoothly underwent in situ coupling with
Fmoc-Ala either with DABCO or PhSiH3 acting as
allyl scavenger.17 After cleavage from the resin, 14 was
obtained in quantitative yield (Scheme 5).
5. (a) Voigt, J.; Noltemeyer, M.; Reiser, O. Synlett 1997,
202–204; (b) Bubert, C.; Cabrele, C.; Reiser, O. Synlett
1997, 827–829.
6. This property has proved to be of great value in Organic
Synthesis. For a review, see: (a) Reissig, H.-U. Top. Curr.
Chem. 1988, 144, 73. For recent examples, see: (b) Patra,
P. K.; Reissig, H.-U. Synlett 2001, 33–36; (c) Bo¨hm, C.;
Reiser, O. Org. Lett. 2001, 3, 1315–1318.
7. For an exception, see: Kraus, G. A.; Kim, H.; Thomas,
P. J.; Metzler, D. E.; Metzler, C. M.; Taylor, J. E. Synth.
Commun. 1990, 20, 2667.
8. (a) Guibe´, R. Tetrahedron 1998, 54, 2967–3042; (b)
Albericio, F. Biopolymers 2000, 55, 123–139.
9. (a) Amine-boranes: Gomez-Martinez, P.; Dessolin, M.;
Guibe, F.; Albericio, F. J. Chem. Soc., Perkin Trans. 1
1999, 2871–2874; (b) NDMBA: Kunz, H.; Ma¨rz, J.
Angew. Chem., Int. Ed. Engl. 1988, 27, 1375; (c) TSA:
Genet, J. P.; Blart, E.; Savignac, S.; Lemeune, S.;
Lemaire-Audoire, S.; Bernard, J.-M. Synlett 1993, 680.
10. (a) Roos, E. C.; Bernabe´, P.; Hiemstra, H.; Speckamp, N.
Tetrahedron Lett. 1991, 32, 6633–6636; (b) Roos, E. C.;
Bernabe´, P.; Hiemstra, H.; Speckamp, W. N. J. Org.
Chem. 1995, 60, 1733–1740; (c) Beugelmans, R.; Bourdet,
S.; Bigot, A.; Zhu, J. Tetrahedron Lett. 1994, 35, 4349; (d)
Beugelmans, R.; Neuville, L.; Bois-Choussy, M.; Chas-
tanet, J.; Zhu, J. Tetrahedron Lett. 1995, 36, 3129–3132.
11. (a) Thieriet, N.; Guibe´, F.; Loffet, A. Peptides 1996,
823–824; (b) Thieriet, N.; Alsina, J.; Giralt, E.; Guibe´, F.;
Albericio, F. Tetrahedron Lett. 1997, 38, 7275–7278; (c)
Thieriet, N.; Gomez-Martinez, P.; Guibe´, F. Tetrahedron
Lett. 1999, 40, 2505–2508.
12. (a) Kunz, H.; Waldmann, H. Angew. Chem., Int. Ed.
Engl. 1984, 23, 71–72; (b) Jouin, P.; Poncet, J.; Dufour,
M.-N.; Pantaloni, A.; Castro, B. J. Org. Chem. 1989, 54,
617–627; (c) Schmidt, U.; Mundinger, K.; Mangold, R.;
Lieberknecht, A. J. Chem. Soc., Chem. Commun. 1990,
1216–1219; (d) Baldwin, J. E.; Moloney, M. G.; North,
M. Tetrahedron 1989, 45, 6319–6330; (e) Stanley, M. S. J.
Org. Chem. 1992, 57, 6421–6430.
In conclusion, a new variant of the tandem deprotec-
tion/coupling strategy for the synthesis of peptides
using N-Alloc-protected amino acids has been devel-
oped, which seems to be especially useful for amino
acids and peptide fragments that are labile in the
N-unprotected form.
Acknowledgements
This work was initiated through Scientia EuropÆa No.
3 (organized by Professor Dr. Guy Ourisson and Dr.
Steve Brooks) and was supported by the Deutsche
Forschungsgemeinschaft (RE 948-4/1), the Fonds der
Chemischen Industrie (Strukturhilfe Bioinformation)
and by the Rhone-Poulenc Foundation, France.
References
1. Bailey, P. D. An Introduction to Peptide Chemistry;
Wiley: New York, 1990.
13. Abbreviations: DABCO, 1,8-diazabicyclo[2.2.2]cyclo-
octane; EDC, 1-(3(dimethylamino)propyl)-3-ethylcar-
bodiimide; DIC, diisopropylcarbodiimide.
2. (a) Cannon, J. G.; Garst, J. E. J. Org. Chem. 1975, 40,
182–184; (b) Shroff, C. C.; Stewart, W. S.; Uhm, S. J.;
Wheeler, J. W. J. Org. Chem. 1971, 22, 3356–3361; (c)
Paulini, K.; Reissig, H.-U. Liebigs Ann. Chem. 1991,
455–461; (d) Vilsmaier, E. In The Chemistry of the Cyclo-
propyl Group; Rappoport, Z., Ed.; VCH-Wiley: New
York, 1987; p. 1341; (e) Martin-Vila, M.; Muray, E.;
Aguada, G. P.; Alvarez-Larena, A.; Branchadell, V.;
Minguillon, C. G. E.; Ortuno, R. M. Tetrahedron: Asym-
metry 2000, 11, 3569.
3. (a) Beumer, R.; Reiser, O. Tetrahedron 2001, 57, 6497–
6503; (b) Beumer, R.; Bubert, C.; Cabrele, C.; Vielhauer,
O.; Pietzsch, M.; Reiser, O. J. Org. Chem. 2000, 65,
8960–8969; (c) Bubert, C.; Voigt, J.; Biasetton, S.; Reiser,
O. Synlett 1994, 675–677.
4. (a) Paulini, K.; Reissig, H.-U. Liebigs Ann. Chem. 1994,
549–554; (b) Beck-Sickinger, A. G.; Hoffmann, E.;
Paulini, K.; Reissig, H.-U.; Willim, K.-D.; Wieland, H.
A.; Jung, G. Biochem. Soc. Trans. 1994, 22, 145–149; (c)
Hibbs, D. E.; Hursthouse, M. B.; Jones, I. G.; Jones, W.;
Malik, K. M. A.; North, M. Tetrahedron 1997, 53,
17417–17424; (d) North, M. J. Peptide Sci. 2000, 6,
301–313.
14. Representative procedure: All solvents were dried by
standard laboratory methods and degassed with nitrogen
prior to use. The N-Boc or N-Fmoc amino acid 4 (3
equiv.) was activated by stirring in CH2Cl2 with EDC (3
equiv.) and HOBt (3 equiv.) for 1 h at 0°C and 1 h at rt
under a nitrogen atmosphere. The palladium(0) catalyst
(10–20 mol%) and, subsequently, a solution of the N-
Alloc amino acid 3 or 6 (0.4 mmol, 1 equiv.) in CH2Cl2
were added. Finally, DABCO (5 equiv.) was added in one
portion and the reaction mixture was stirred for the
indicated time. Extractive work-up of the organic layer
(saturated NaHCO3, 1N KHSO4, and saturated
NaHCO3), drying (MgSO4), concentration and purifica-
tion of the residue on silica gel afforded 5 or 7 (Tables 1
and 2).
15. For a recent report on an automated solid-phase synthe-
sis using Alloc deprotection with Pd(0)/phenylsilane, see:
Grieco, P.; Gitu, P. M.; Hruby, V. J. J. Peptide Res.
2001, 57, 250–256.
16. For other solutions to the DKP problem, see: (a) Suzuki,
K.; Nitta, K.; Endo, N. Chem. Pharm. Bull. 1975, 23,

C. Zorn et al. / Tetrahedron Letters 42 (2001) 7049–7053
7053
222–224; (b) Shute, R. E.; Rich, D. H. J. Chem. Soc.,
Chem. Commun. 1987, 1155–1156; (c) Tung, R. D.;
Dahon, M. K.; Rich, D. H. J. Org. Chem. 1986, 51, 3350;
(d) Gair´ı, M.; Lloyd-Williams, P.; Albericio, F.; Giralt,
E. Tetrahedron Lett. 1990, 31, 7363–7366; Dalcol, I.;
Rabanal, F.; Ludevid, M.-D.; Albericio, F.; Giralt, E. J.
Org. Chem. 1995, 60, 7575–7581; (e) Goolcharran, C.;
Borchardt, R. T. J. Pharm. Sci. 1998, 87, 283–288; (f)
Humphrey, J. M.; Chamberlin, A. R. Chem. Rev. 1997,
97, 2243–2266; (g) Lloyd-Williams, P.; Albericio, F.;
Giralt, E. Solid-phase Peptide Synthesis; CRC Press:
Boca Raton, FL, 1997; pp. 59–61.
DABCO can be rigorously defined as an allyl scavenger,
since the fate of the allyl group remained unclear. It was
suggested that the allyl group might end up as allyl ether
of HOBt or even as an N-allyl derivative of HOBt. This
intriguing speculation implied that DABCO could act
simply as a catalyst. We tested this hypothesis by cou-
pling 6 with benzoic acid using our general procedure (see
Ref. 14, activation with DIC) but only in the presence of
10 mol% DABCO. Unfortunately, we were able to obtain
very little of the expected coupling product. Therefore,
we must conclude at this point that the allylammonium
salt of DABCO is formed and that the latter indeed
serves as the primary allyl scavenger.
17. One referee raised the important question whether