Exploiting an inherent neighboring group effect of α-amino acids to synthesize extremely hindered dipeptides

Article abstract of DOI:10.1021/ja806063k

The creation of highly hindered peptides that contain combinations of non-natural N-alkyl amino acids and N-alkyl-α,α-disubstituted amino acids presents a formidable challenge. Hindered, non-natural amino acids are of interest because they import resistance to proteolysis and unusual conformational properties to peptides that contain them. Toward a solution to this problem, we describe a new approach to creating extremely hindered dipeptides that is operationally simple and uses mild conditions and commercially available amino acids. The approach reduces the need for protecting groups and yields urethane-protected dipeptide acids that can be used as building blocks in the synthesis of larger peptides. We propose that the reaction proceeds through a previously unexploited intramolecular O,N-acyl transfer pathway. Copyright

Full text of DOI:10.1021/ja806063k

Published on Web 10/08/2008
Exploiting an Inherent Neighboring Group Effect of r-Amino Acids To
Synthesize Extremely Hindered Dipeptides
Zachary Z. Brown and Christian E. Schafmeister*
Chemistry Department, Temple UniVersity, Philadelphia, PennsylVania 19122
Received August 1, 2008; E-mail: meister@temple.edu
Scheme 1. The Challenge of Hindered Amide Bond Synthesis
The synthetic methodology for creating peptides is among the
most highly developed synthetic methodologies in chemistry.1,2
While the creation of proteogenic peptides is largely a solved
problem, the creation of highly hindered peptides that contain
combinations of non-natural N-alkyl amino acids and N-alkyl-R,R-
disubstituted amino acids remains a formidable challenge.3 Hin-
dered, non-natural amino acids are of interest because they import
resistance to proteolysis and unusual conformational properties to
peptides.4 Toward a solution to this problem, we describe a new
approach to creating extremely hindered dipeptides that is opera-
tionally simple and uses mild conditions and commercially available
amino acids. The approach reduces the need for protecting groups
and yields urethane-protected dipeptide acids that can be used as
building blocks in the synthesis of larger peptides. This approach
appears to take advantage of a previously unexploited intramolecular
acyl transfer.
Scheme 2. Synthesis of Fmoc-Aib-NMeVal-OH 3
Table 1. Yields of a Variety of Hindered Dipeptidesa
entry
dipeptide
yield (%)
time (min)
Urethane-based amine protecting groups including the carboxy-
benzoyl group (Cbz),5 the tert-butylcarboxyl group (t-Boc)2 and
the 9-fluorenylmethylenecarboxyl group (Fmoc)6 are by far the most
common amine protecting groups used in peptide synthesis.
Urethane-protected amino acids are commercially available for all
proteogenic amino acids and many unnatural amino acids. These
groups protect the amine of amino acids from unwanted reactions,
they protect the amino acid from racemization during activation of
the carboxyl group, and they are easily removed using mild
conditions. Many approaches have been demonstrated for coupling
proteogenic and hindered amino acids in which either the amino
group (i.e., N-alkyl amino acids) or the carboxylic group (i.e., R,R-
disubstituted amino acids) are relatively hindered.7,8 On the other
hand, there is as yet no efficient way to form amide bonds between
hindered urethane-amino acids (such as Fmoc-Aib-OH) with
carboxyl-masked hindered amino acids such as N-alkyl amino
acids.3,9 The challenge is that the amine of an N-alkyl amino acid
is a very poor nucleophile, and, to compensate, strong activation
of the urethane-amino acid, such as through use of a urethane-
amino acid chloride, would be required to achieve coupling.
However, acid chlorides are difficult to handle and strong activation
of urethane-amino carboxyls leads to the rapid formation of
oxazalones, which are poor electrophiles (Scheme 1) and prone to
racemization. Sulfonamides have been developed as alternatives
to urethane protecting groups10 and used to couple N-methyl amino
acids, but they cannot yet match the low cost, commercial
availability, excellent stability, and convenience of urethane-
protected amino acids.
4
5
6
7
8
9
10
11
12
13
Fmoc-Aib-(S)-Tic-OH
Fmoc-Aib-NMeAib-OH
86
60
78
80
68
78
74
79
60
78
60
60
5
5
5
Fmoc-(S)-NMeVal-(S)-NMeVal-OH
Fmoc-(S)-NMeVal-(S)-Tic-OH
Fmoc-(S)-NMeVal-NMeAib-OH
Fmoc-(S)-NMeVal-Sar-OH
Fmoc-Ac5c-NMeVal-OH
Fmoc-Ac5c-(S)-Tic-OH
5
45
45
45
45
Fmoc-Ac5c-NMeAib-OH
Fmoc-Ac5c-Sar-OH
a All reactions were carried out as described in Scheme 2 varying
only the duration of the reaction.
prewarmed to 55 °C (Scheme 2). The solution was allowed to react
at 55 °C for one hour and then subjected to reverse-phase high
performance liquid chromatography with mass spectrometry
(HPLC-MS). The same approach was used to synthesize a variety
of other dipeptides with good yields (Table 1). Fmoc-amino acids
activated as 1-oxy-7-azabenzotriazole (OAt)) esters using diiso-
propylcarbodiimide and HOAt produced dipeptide, but the reactions
were ∼50% slower than those using acid fluorides.7
Reverse phase chromatography of the product of Scheme 2
provides the dipeptide Fmoc-Aib-NMeVal-OH (80%), the recovered
starting material Fmoc-Aib-OH (11%), and the ester Fmoc-Aib-
OCH(CF3)2 (9%) (Figure 1). Hexafluoroisopropanol is used as the
solvent for the dipeptide coupling because it is the only solvent
that we have as yet identified that consistently maintains solubility
of both the amino acids and the Fmoc-amino acid fluoride. The
drawback of using this solvent is that it reacts with the acid fluoride
to produce small amounts of hexafluoroisopropyl ester. It is
important to completely avoid base in these reactions because base
leads to quantitative esterification with solvent. The byproducts and
unreacted amino acid starting material can be removed by extraction
and chromatography. For precious Fmoc-amino acids the ester can
To form a dipeptide we simply combine an Fmoc-amino acid
fluoride with an excess of unprotected amino acid (Scheme 2)
dissolved in hexafluoroisopropanol. We followed the literature
procedure11 with slight modifications to form the acid fluoride 1.
The Fmoc-amino acid fluoride was used without purification and
added directly to a solution of 105 mg (0.8 mmol) of N-methylvaline
2 dissolved in 4 mL of hexafluoroisopropanol that had been
9
14382 J. AM. CHEM. SOC. 2008, 130, 14382–14383
10.1021/ja806063k CCC: $40.75
2008 American Chemical Society

C O M M U N I C A T I O N S
Scheme 3. Proposed Mechanism for Dipeptide Formation
combined 1 with 4 equiv N-(2-methylnaphthylene)-sarcosine in
HFIP and observed that almost 8% of it was converted to the HFIP
ester, consistent with the formation of an anhydride that undergoes
solvolysis because it can not undergo O,N-acyl transfer. The
proposed amino-anhydride intermediate and five-member ring
transition state is similar to that of the Ugi reaction,13 which has
been described as a “remote Mumm rearrangement”.14 It is
reminiscent of the S,N-acyl shift invoked in native chemical
ligation15 and the O,N-acyl shift16 invoked in the “depsipeptide
technique”17 and “switch peptides”.18 It is also reminiscent of the
pioneering work of Kemp and co-workers in their development of
O,N-acyl transfer groups.19
In conclusion, we have demonstrated an operationally simple
approach to synthesizing extremely hindered dipeptides using
commercially available amino acids and mild reaction conditions
that provides high yields of dipeptide and maintains stereochemical
integrity.
Figure 1. HPLC chromatogram of the reaction mixture of Fmoc-Aib-
NMeVal-OH 3. The unlabeled peaks did not have identifiable masses.
be recovered and recycled using acid or base mediated hydrolysis
depending on whether N-Fmoc or N-Boc urethane protection is
used.
Low resolution HPLC-MS gave m/z values that were consistent
with the structure of each dipeptide 3 to 13. High resolution mass
spectrometry of 3, 4, and 10 were consistent with the predicted
masses of these dipeptides. The structures of 3, 4, and 10 were
confirmed using HMQC and HMBC two-dimensional NMR experi-
ments, which in the case of 3 demonstrated correlations between
the hydrogens of the N-Me group and the carbonyl carbon of the
Aib residue as well as to the R carbon of the NMeVal residue. In
addition, the methyl groups of the Aib residue of 3 have different
chemical shifts, indicating that they are diastereotopic. We collected
proton spectra of 3 at a variety of temperatures and the peaks are
broad at room temperature and become sharp at 360 K. This is
consistent with two amide rotamers interconverting slowly at room
temperature and interconverting rapidly at high temperature. Other
examples of dipeptides that we synthesized demonstrated proton
NMR spectra that were consistent with mixtures of slowly inter-
converting amide rotamers at room temperature (e.g., dipeptide 6).
We carried out the dipeptide formation reaction using (S)-Fmoc-
NMeVal-F and racemic NMeVal-OH, and by reverse-phase HPLC
we saw two peaks consistent with two diastereomeric dipeptides.
The reverse-phase HPLC of 6 displayed only one dipeptide peak,
which demonstrates that the dipeptide formation reaction maintains
the stereochemical integrity of both amino acid stereocenters.
At 55 °C the reactions with Fmoc-NMeVal-F (6 to 9) are over
in 5 min. When the Fmoc-amino acid fluoride is more hindered
(such as Fmoc-Aib-F), the dipeptide formation is slower but the
reaction with solvent also slows down so the yields remain high.
The time-course of the formation of 3 is shown in the Supporting
Information. We observed no premature deprotection of the Fmoc
group in any of these reactions. Our attempts to couple Fmoc-Aib-
OH with NMeVal-OMe in dimethylformamide using a variety of
activated species, including acid fluorides, yielded no significant
dipeptide and extensive Fmoc deprotection as seen by others.12 We
also carried out a competition experiment in which we added Fmoc-
Aib-F to 4 equiv of NMeVal-OH and 4 equiv of NMeVal-OMe.HCl
in HFIP in one pot. By RP-HPLC we observed a ratio of Fmoc-
Aib-NMeVal-OH/Fmoc-Aib-NMeVal-OMe of 97:3. The free car-
boxylic acid of the unprotected amino acid is necessary to achieve
efficient acylation.
Acknowledgment. The authors would like to thank Dr. Jason
Moss for valuable conversations. This work was supported by the
NIH/NIGMS (GM067866).
Supporting Information Available: Synthesis and characterization
of the dipeptides. This material is available free of charge via the
Internet at http://pubs.acs.org.
References
(1) Merrifield, B. Methods Enzymol. 1997, 289, 3–13. Albericio, F. Curr. Opin.
Chem. Biol 2004, 8, 211–221.
(2) Bodanszky, M. Principles of Peptide Chemistry; Springer-Verlag: New
York, 1984.
(3) Carpino, L. A.; Ionescu, D.; El-Faham, A.; Henklein, P.; Wenschuh, H.;
Bienert, M.; Beyermann, M. Tetrahedron Lett. 1998, 39, 241–244.
(4) Humphrey, J. M.; Chamberlin, A. R. Chem. ReV. 1997, 97, 2243–2266.
(5) Bergmann, M.; Zervas, L. Chem. Ber. 1932, 65, 1192–1201.
(6) Carpino, L. A. Acc. Chem. Res. 1987, 20, 401–407.
(7) Carpino, L. A. J. Am. Chem. Soc. 1993, 115, 4397–4398.
(8) Wenschuh, H.; Beyermann, M.; Krause, E.; Brudel, M.; Winter, R.;
Schumann, M.; Carpino, L. A.; Bienert, M. J. Org. Chem. 1994, 59, 3275–
3280. Thern, B.; Rudolph, J.; Jung, G. Angew. Chem., Int. Ed. 2002, 41,
2307–2309. Katritzky, A. R.; Todadze, E.; Angrish, P.; Draghici, B. J.
Org. Chem. 2007, 72, 5794–5801. Cianci, J.; Baell, J. B.; Harvey, A. J.
Tetrahedron Lett. 2007, 48, 5973–5975.
(9) Moretto, V.; Valle, G.; Crisma, M.; Bonora, G. M.; Toniolo, C. Int. J. Biol.
Macromol. 1992, 14, 178–184.
(10) Miller, S. C.; Scanlan, T. S. J. Am. Chem. Soc. 1998, 120, 2690–2691.
(11) Kaduk, C.; Wenschuh, H.; Beyermann, M.; Forner, K.; Carpino, L.; Bienert,
M. Lett. Pept. Sci. 1996, 2, 285–288.
(12) Wenschuh, H.; Beyermann, M.; Winter, R.; Bienert, M.; Ionescu, D.;
Carpino, L. A. Tetrahedron Lett. 1996, 37, 5483–5486.
(13) Ugi, I. Angew. Chem., Int. Ed. 1962, 1, 8–21. Do¨mling, A.; Ugi, I. Angew.
Chem., Int. Ed. 2000, 39, 3168–3210.
(14) Mumm, O. Chem. Ber. 1910, 43, 886–893.
(15) Dawson, P. E.; Muir, T. W.; Clark-Lewis, I.; Kent, S. B. H. Science 1994,
266, 776–779.
(16) Bergmann, M.; Brand, E.; Weinmann, F. Hoppe-Seyler’s Z. Physiol. Chem.
1923, 131, 1–17.
The observation that the degree of steric crowding of the amine
of the amino acid has little effect on the dipeptide yield led us to
propose the mechanism shown in Scheme 3. We propose that the
carboxylic acid of the amino acid attacks the acid-fluoride to form
a transient anhydride that spontaneously rearranges through a five-
membered ring transition state to form the hindered amide bond in
an intramolecular O,N-acyl transfer. To test this hypothesis we
(17) Coin, I.; Dolling, R.; Krause, E.; Bienert, M.; Beyermann, M.; Sferdean,
C.; Carpino, L. J. Org. Chem. 2006, 71, 6171–6177.
(18) Dos Santos, S.; Chandravarkar, A.; Mandal, B.; Mimna, R.; Murat, K.;
Saucede, L.; Tella, P.; Tuchscherer, G.; Mutter, M. J. Am. Chem. Soc. 2005,
127, 11888–11889.
(19) Kemp, D. S.; Carey, R. I. J. Org. Chem. 1993, 58, 2216–2222.
JA806063K
9
J. AM. CHEM. SOC. VOL. 130, NO. 44, 2008 14383