Synthesis of a ketomethylene isostere of the fibrillating peptide SNNFGAILSS

Article abstract of DOI:10.1021/jo901466b

(Chemical Equation Presented) The direct synthesis of a ketomethylene isostere of the fibril-forming decapeptide SNNFGAILSS is presented with the goal of understanding how small structural changes alter the ability of such peptides to recognize each other

Full text of DOI:10.1021/jo901466b

pubs.acs.org/joc
hIAPP20-29 SNNFGAILSS-NH2 decapeptide.7 The 3D
Synthesis of a Ketomethylene Isostere of the
Fibrillating Peptide SNNFGAILSS
structure of these β-sheet forming fibrils can be used to
provide atomic-resolution insight into the key interactions
that hold these fibrils together.
Tina Mittag, Daniel E. Otzen, Niels Chr. Nielsen, and
Troels Skrydstrup*
Parallel to this work, we have set out to prepare analogues
of SNNFGAILSS 1 incorporating small structural changes
in order to examine their influence on fibril formation and
growth (Figure 1). To this end, the ketomethylene isostere
(SNNF-Ψ(CO-CH2)-GAILSS) 2 represents an interesting
target. Such a structural modification will remove a single
amide bond, thereby eliminating one hydrogen bond inter-
action. How this variation will influence the ability of such
peptides to form β-sheet fibrils will be of particular interest
for understanding stability of the fibril structures formed
from the parent peptide.
The Center for Insoluble Protein Structures (inSPIN),
Department of Chemistry and the Interdisciplinary
Nanoscience Center, Aarhus University, Langelandsgade 140,
8000 Aarhus C, Denmark
ts@chem.au.dk
Received July 11, 2009
Synthesis of ketomethylene isosteres are generally not
simple operations and usually require multiple steps for
preparing these classes of dipeptide analogues.8 Recently,
we reported a novel and direct route to dipeptidyl keto-
methylene isosteres through the coupling of N-acyl oxazoli-
dinone derivatives of amino acids or peptides with acryl-
amides.9-14 In this way, these peptide isosteres can be
accessed directly from a single coupling step. Most impor-
tantly, the reaction conditions are mild and hence epimeriza-
tion is avoided at the adjacent stereogenic center to the newly
formed ketone. With this reaction in mind, we set out to
prepare a ketomethylene dipeptide isostere of the fibrillat-
ing peptide SNNFGAILSS at the structurally simplest
peptide bond position (Phe-Gly). The results of this work
again illustrate the usefulness of this lanthanide reagent for
performing carbon-carbon bond-forming reactions with
important biomolecules.
The direct synthesis of a ketomethylene isostere of the
fibril-forming decapeptide SNNFGAILSS is presented
with the goal of understanding how small structural
changes alter the ability of such peptides to recognize
each other for β-sheet formation. The key synthetic step
relies on a SmI2-mediated coupling of a N-tetrapeptidyl
oxazolidinone with a simple acrylate followed by depro-
tection of the carboxylic acid and a peptide coupling step
with the pentapeptide H-AILSS-NH2.
To commence this study, we initially focused our attention
on the synthesis of a ketomethylene isostere of the smaller
peptide NFGAIL as a model to examine the suitability of the
C-C bond-forming approach to this class of peptides. In a
recent publication, we reported that the SmI2-mediated
Deposition of human islet amyloid polypeptide (hIAPP)
as fibrillar amyloid in the pancreatic islets of Langerhans is a
characteristic histopathological marker for type II diabetes
mellitus (T2DM) and is found in more than 90% of the
affected patients.1,2 Extensive studies with fragments of
hIAPP have revealed that hIAPP20-29 (SNNFGAILSS),3,4
hIAPP22-27 (NFGAIL),5 and hIAPP22-29 (NFGAILSS)6 all
form amyloid fibrils with morphologies very similar to those
of the full-length polypeptide. This region of the peptide is
therefore believed to form the fibrillation core domain of
fibrils in the pancreas of type II diabetes patients. We have
recently determined the complete 3D β-sheet fibril structure
(7) Nielsen, J. T.; Bjerring, M.; Jeppesen, M. D.; Pedersen, R. O.;
Pedersen, J. M.; Hein, K. L.; Vosegaard, T.; Skrydstrup, T.; Otzen, D. E.;
Nielsen, N. C. Angew. Chem., Int. Ed. 2009, 48, 2118–2121.
(8) For a few pertinent examples, see: (a) Deziel, R.; Plante, R.; Caron, V.;
Grenier, L.; LlinasBrunet, M.; Duceppe, J. S.; Malenfant, E.; Moss, N.
J. Org. Chem. 1996, 61, 2901–2903. (b) Kaltenbronn, J. S.; Hudsepth, E. A.;
Lunney, B. M.; Michniewicz, B.; Woo, P. K. W.; Essenburg, A. D. J. Med.
Chem. 1998, 33, 838. (c) DeGraw, J. I.; Almquist, R. G.; Hiebert, C. K.;
Colwell, W. T.; Crase, J.; Hayano, T.; Judd, A. K.; Dousman, L.; Smith,
R. L.; Waud, W. R.; Uchida, I. J. Med. Chem. 1997, 40, 2386–2397.
(d) Kempf, D. J. J. Org. Chem. 1986, 51, 3921–3926. (e) Lin, W.; Theberge,
C. R.; Henderson, T. J.; Zercher, C. K.; Jasinski, J.; Butcher, R. J. J. Org.
Chem. 2009, 74, 645–651.
(9) Jensen, C. M.; Lindsay, K. B.; Taaning, R. H.; Karaffa, J.; Hansen, A.
M.; Skrydstrup, T. J. Am. Chem. Soc. 2005, 127, 6544–6545.
(10) Hansen, A. M.; Lindsay, K. B.; Sudhadevi Antharjanam, P. K.;
Karaffa, J.; Daasbjerg, K.; Flowers, R. A. II; Skrydstrup, T. J. Am. Chem.
Soc. 2006, 128, 9616–9617.
(11) Karaffa, J.; Lindsay, K. B.; Skrydstrup, T. J. Org. Chem. 2006, 71,
8219–8226.
(12) Lindsay, B. K.; Ferrando, F.; Christensen, K. L.; Overgaard, J.;
Roca, T.; Bennasar, M.-L.; Skrydstrup, T. J. Org. Chem. 2007, 72, 4181–
4188.
˚
of one of the hIAPP fragments with 0.5 A atomic resolution
using solid-state nuclear magnetic resonance spectroscopy
with a fully 15N,13C-labeled NFGAIL fragment of the
(1) Opie, E. L. J. Exp. Med. 1900, 5, 397–428.
(2) Westermark, P. Amyloid 1994, 1, 47–60.
(3) Glenner, G. G.; Eanes, E. D.; Wiley, C. A. Biochem. Biophys. Res.
Commun. 1988, 155, 608–614.
(4) Westermark, P.; Engstrom, U.; Johnson, K. H.; Westermark, G. T.;
Betsholtz, C. Proc. Natl. Acad. Sci. U.S.A. 1990, 87, 5036–5040.
(5) Tenidis, K.; Waldner, M.; Bernhagen, J.; Fischle, W.; Bergmann, M.;
Weber, M.; Merkle, M.-L.; Voelter, W.; Brunner, H.; Kapurniotu, A. J. Mol.
Biol. 2000, 295, 1055–1071.
(13) Mittag, T.; Christensen, K. L.; Lindsay, K. B.; Nielsen, N. C.;
Skrydstrup, T. J. Org. Chem. 2008, 73, 1088–1092.
~
(14) Taaning, R. H.; Thim, L.; Karaffa, J.; Campana, A. G.; Hansen,
A,-M.; Skrydstrup, T. Tetrahedron 2008, 64, 11884–11995.
(6) Azriel, R.; Gazit, E. J. Biol. Chem. 2001, 276, 34156–34161.
DOI: 10.1021/jo901466b
r
Published on Web 09/09/2009
J. Org. Chem. 2009, 74, 7955–7957 7955
2009 American Chemical Society

Mittag et al.
JOCNote
SCHEME 2. Synthesis of the Tripeptide NF-Ψ(CO-CH2)-G
Isostere
SCHEME 3. Synthesis of the Fully Protected NF-Ψ(CO-CH2)-
GAIL
FIGURE 1. SNNFGAILSS decapeptide and the corresponding
ketomethylene isostere.
SCHEME 1. Previously Published SmI2-Promoted Reactions
with Peptide-Derived Acrylamides
smaller and more soluble electron-deficient alkene as the
coupling partner.
Preliminary efforts were focused on the coupling of the
N-dipeptidyl oxazolidinones 3 and 8 with methyl acrylate, as
shown in Scheme 2. The coupling products 9 and 10 were
obtained by dropwise addition of SmI2 (0.1 M in THF) over
30 min to a cold solution of either peptidyl oxazolidinone 3 or 8
and methyl acrylate in the presence of water followed by stirring
for 45 h. Both 9 and 10 could be isolated in good yields, and as
expected from earlier work, the Fmoc and Boc protecting groups
on the N-terminal amino acid were well tolerated.13
To obtain the free carboxylic acid for further peptide coupling,
hydrolysis of the methyl ester was examined using several
protocols with the goal of avoiding epimerization of the chiral
carbon center adjacent to the ketone functionality. Whereas
LiOH in THF/water (4:1)16 and potassium trimethylsilanolate in
THF17 proved effective for the hydrolysis of the methyl ester 10,
considerable epimerization was observed in the product in both
coupling of the N-dipeptidyl oxazolidinone 3 and N-acryl-
oyl-L-alanine methyl ester 4 provided the fully protected keto-
methylene isostere 5 of the tetrapeptide NFGA in a good 82%
yield (Scheme 1).13 However, upon extrapolation to the hexa-
peptide analogue, the reaction between 3 and the larger acryl-
amide, N-acryloyl-L-alanyl-L-isoleucyl-L-leucine methyl ester 6
was impeded by the low solubility of the electron-poor alkene
under the reaction conditions leading to the formation of the
coupling product 7 in only a 20% yield.
To avoid this potential reactant solubility issue for this key
coupling step, which undoubtedly would be more prominent
if this synthetic approach were to be extended to the larger
decapeptide, we decided to pursue a slightly modified syn-
thetic route to the ketomethylene isostere. In this alternative
approach, the N-acyl oxazolidinone would initially be
coupled to a simple acrylate followed by liberation of the
carboxylic acid and a standard peptide coupling step with a
C-terminally protected peptide. The solubility issue in the
C-C bond-forming step is in this way resolved by utilizing a
(15) Mikkelsen, L. M.; Jensen, C. M.; Høj, B.; Blakskjær, P.; Skrydstrup,
T. Tetrahedron 2003, 59, 10541–10549.
(16) Schmuck, C.; Geiger, L. Chem. Commun. 2005, 772–774.
(17) (a) Lovric, M.; Cepanec, I.; Litvic, M.; Bartolincic, A.; Vinkovic, V.
Croat. Chem. Acta 2007, 80, 109–115. (b) Polyak, F.; Lubell, W. D J. Org.
Chem. 2001, 66, 1171–1180. (c) Motorina, I. A.; Huel, C.; Quiniou, E.;
Mispelter, J.; Adjadj, E.; Grierson, D. S. J. Am. Chem. Soc. 2001, 123, 8–17.
7956 J. Org. Chem. Vol. 74, No. 20, 2009

Mittag et al.
JOCNote
SCHEME 4. Synthesis of SNNF-Ψ(CHOH-CH2)-GAILSS-NH2
of L-leucine benzyl ester and N-Boc-L-isoleucine using EDC
as the coupling reagent. Removal of the Boc-protecting
group of the resulting N-Boc-L-isoleucyl-L-leucine benzyl
ester with trifluoroacetic acid in dichloromethane followed
by further extension of the peptide chain with alanine and
deprotection yielded the tripeptide 13 ready for peptide
coupling with 11. Applying standard EDC peptide coupling
conditions afforded the desired, fully protected NFGAIL
isostere 14 in an 80% overall yield for the last two steps.
With the successful synthesis of the hexapeptide analogue
as a model, we immediately proceeded to apply this synthetic
approach for accessing the ketomethylene isostere of the
decapeptide as illustrated in Scheme 4. Synthesis of the
required tetrapeptidyl oxazolidinone 16 was performed by
two consecutive cycles of N-terminal protecting group
removal with 3 M HCl in i-PrOH at 50 °C followed by an
EDC-mediated peptide coupling via the tripeptide 15. The
key C-C bond-forming step was initially performed under
standard conditions for peptidyl oxazolidinones and acrylate
couplings at -78 °C and provided the ketomethylene isostere
17 of the pentapeptide SNNFG in a 54% yield, which is
consistent with the yields previously published for this type
of reaction.13 However, increasing the reaction temperature
to -55 °C from -78 °C and doubling the amount of benzyl
acrylate had a beneficial effect on the yield and 17 was
isolated in a good 72% yield. Hydrogenolysis of the coupling
product afforded the partially deprotected isostere 18.
Finally, the peptide chain was elongated to the corres-
ponding decapeptide isostere 20 by liquid-phase peptide
coupling in DMF between the protected peptide amide
fragment 19, which was prepared by automated solid-phase
peptide synthesis, and the carboxylic acid 18. Treatment of
the protected decapeptide ketomethylene isostere 20 with
95% trifluoroacetic acid afforded the desired peptide isostere
21 in a 50% yield after HPLC purification. The decapeptide
analogue was prepared as its C-terminal amide, as we have
earlier shown that the same derivative of the parent peptide
can easily be fibrillated,7 thus facilitating the subsequent
fibrillation studies of this new ketomethylene isostere.
In summary, we have prepared a ketomethylene isostere of
the fibrillating SNNFGAILSS fragment of the human islet
amyloid polypeptide with the purpose of exploiting such
analogues for understanding how small structural altera-
tions influence the ability of such peptides to aggregate. The
isostere was prepared exploiting an efficient SmI2-mediated
C-C bond forming reaction as the key step for attaining a
peptidyl ketone intermediate. Work is now underway to
investigate whether this decapeptide analogue can fibrillate,
as well as examining its influence on the fibrillation of the
parent peptide. These studies will be reported in due course.
1
cases by H NMR spectroscopy. Trimethyltin hydroxide has
recently been reported as a mild and selective method for the
hydrolysis of epimerization-prone methyl esters.18 Treatment
of 9 with Me3SnOH (3.0 equiv) in dichloroethane at 70 °C for
4.5 h provided the desired carboxylic acid 11 with no epimeriza-
tion, but only a low yield of the product was obtained. Extended
reaction times afforded higher yields of 11, though with signs of
epimerization according to 1H NMR and HPLC analysis.
To circumvent these problems, the C-C bond-forming
reaction was carried out with benzyl acrylate and the
N-peptidyl oxazolidinone 8 leading to the formation of the
ketone 12 in a good yield (Scheme 3). The following hydro-
genolysis was accomplished using palladium on charcoal
under a hydrogen atmosphere and afforded the free carboxy-
Acknowledgment. We are deeply appreciative of generous
financial support from the Danish National Research Foun-
dation, the Lundbeck Foundation, the Carlsberg Founda-
tion, the iNANO and OChem Graduate Schools and Aarhus
University.
1
lic acid 11 in a 92% yield. H NMR analysis of tripeptide
NFG analogue revealed that no detectable epimeriza-
tion had occurred during either the coupling or the hydro-
genolysis step.
The tripeptide 13 required for the subsequent peptide
coupling was synthesized from the p-toluenesulfonate salt
Supporting Information Available: Experimental details
for all compounds including copies of 1H NMR, 13C NMR
spectra. This material is available free of charge via the
Internet at http://pubs.acs.org.
(18) Nicolaou, K. C.; Estrada, A. A.; Zak, M.; Lee, S. H.; Safina, B. S
Angew. Chem., Int. Ed. 2005, 44, 1378–1382.
J. Org. Chem. Vol. 74, No. 20, 2009 7957