Core-shell-type resins for solid-phase peptide synthesis: Comparison with gel-type resins in solid-phase photolytic cleavage reaction

Article abstract of DOI:10.1021/ol048815q

(Graph Presented) Novel core-shell-type resins with a rigid core and amino-functionalized flexible shell were prepared with 2,4,6-trichloro-1,3,5- triazine (CNC) and Jeffamine ED-600 starting from 1% cross-linked aminomethyl (AM) polystyrene resins. All o

Full text of DOI:10.1021/ol048815q

ORGANIC
LETTERS
2004
Vol. 6, No. 19
3273-3276
Core−Shell-Type Resins for Solid-Phase
Peptide Synthesis: Comparison with
Gel-Type Resins in Solid-Phase
Photolytic Cleavage Reaction
Hanyoung Kim, Jin Ku Cho, Woo-Jae Chung, and Yoon-Sik Lee*
School of Chemical Engineering, Seoul National UniVersity, Seoul 151-744, Korea
yslee@snu.ac.kr
Received June 22, 2004
ABSTRACT
Novel core−shell-type resins with a rigid core and amino-functionalized flexible shell were prepared with 2,4,6-trichloro-1,3,5-triazine (CNC)
and Jeffamine ED-600 starting from 1% cross-linked aminomethyl (AM) polystyrene resins. All of the amino groups were located outside the
resin beads, and the loading capacity was 0.2−0.4 mmol/g. The amount of CNC treated was a determining factor in the properties of the final
resins. The core−shell-type resins showed superior performances in terms of the initial loading of amino acid and the photocleavage reaction
compared to the gel-type resins.
Since solid-phase synthesis was introduced by Merrifield,
this protocol has been the standard for the preparation of
peptides/oligonucleotides, and a huge number of solid-phase
organic reactions have been developed for the purpose of
combinatorial approach to find hits.1 In general, lightly
(1-2%) cross-linked gel-type resins are employed as the
supports for solid-phase reactions. However, these gel-type
resins require a suitable swelling ability for the diffusion of
the reagents, which limits the choice of solvents and their
usage for continuous flow systems.2 Even if the resins are
fullyswollen, they present an obstacle to the access of large
molecules such as enzymes3 or the penetration of light into
the resin beads.4
In an effort to overcome these fundamental drawbacks,
we previously proposed the concept of the core-shell type
of resin, in which the rigid core can provide good mechanical
properties and the functionalized flexible shell can allow
facile contact of the reagents. These resins were prepared
by suspension polymerization and exhibited successful
behaviors in photolytic and enzymatic cleavage reactions.5
The resins, however, suffered from relatively low loadings
(0.05-0.2 mmol/g), and the preparation of the macromono-
mers was overly complicated. Here, we wish to report the
(1) (a) Merrifield, R. B. J. Am. Chem. Soc. 1963, 85, 2149. (b) Letsinger,
R. L.; Mahadevan, V. J. Am. Chem. Soc. 1965, 87, 3526. (c) Combinatorial
Chemistry: Synthesis, Analysis, Screening; Jung, G., Ed.; Wiley-VCH:
Weinheim, 1999. (d) The Combinatorial Index; Bunin, B. A., Ed.; Academic
Press: San Diego, 1998.
(2) (a) Kress, J.; Zanaletti, R.; Rose, A.; Frey, J. G.; Brocklesby, W. S.;
Ladlow, M.; Bradley, M. J. Comb. Chem. 2003, 5, 28. (b) Yan, B.; Fell, J.
B.; Kumaravel, G. J. Org. Chem. 1996, 61, 7467. (c) Hodge, P. Synthesis
and Separations using Functional Polymers; Sherrington, D. C., Hodge,
P., Eds.; John Wiley and Sons: New York, 1988. (d) Regen, S. L.
Macromolecules 1975, 8, 689.
(3) Kress, J.; Zanaletti, R.; Amour, A.; Ladlow, M. Frey, J. G.; Bradley,
M. Chem. Eur. J. 2002, 8, 3769.
(4) Patchnornik, A.; Amit, B.; Woodward, R. B. J. Am. Chem. Soc. 1970,
92, 6333.
(5) (a) Cho, J. K.; Park, B. D.; Lee, Y. S. Tetrahedron Lett. 2000, 41,
7481. (b) Cho, J. K.; Park, B. D.; Park, K. B.; Lee, Y. S. Macromol. Chem.
Phys. 2002, 203, 2211.
10.1021/ol048815q CCC: $27.50 © 2004 American Chemical Society
Published on Web 08/25/2004

Scheme 1. Preparation of Core-Shell Resins
synthesis of novel core-shell (CS) resins readily prepared
prepared from the AM-PS resin according to the amount of
CNC (0.5-2, 3-5, 7-10 equiv of initial amino groups,
respectively). After coupling with Jeffamine and coupling
of the resulting amino groups of each CS resin with FTIC
(5-fluorescein isothiocyanate), their cross-sectioned views
were observed using a confocal fluorescent microscope.8
TentaGel-NH2 (0.3 mmol/g) and AM-PS resins (1.1 mmol/
g) were both examined after FTIC coupling. From the
confocal microscope images, we found that the TentaGel and
AM-PS resin-bound FTIC was located across the whole
region of the resin beads (Figure 1a and 1b, respectively).
On the other hand, the CS resin-bound FTIC existed only
on the surface area. In particular, the interior structures of
the CS resins appeared to vary depending on the amount of
CNC treated. When a small amount of CNC was used, the
shape of the shell was irregular and some cracks were
observed (CS resin I, Figure 1c). When higher amounts of
CNC were used, thicker shell was formed (CS resin II and
III, Figure 1d and 1e). The loading level of each CS resins
was determined by means of the Fmoc photometric test after
Fmoc-Leu coupling.9 We found that the loadings of CS resin
I were rather low, namely, 0.03-0.1 mmol/g. Such a low
from commercially available aminomethyl polystyrene
(AM-PS, BeadTech6) and their characteristics and perfor-
mances in solid-phase reactions.
Initially, AM-PS resin (1% cross-linked) was further cross-
linked with 2,4,6-trichloro-1,3,5-triazine (cyanuric chloride,
CNC).7 Then, the resultant resins were treated with an excess
of O,O′-bis(2-aminopropyl) poly(ethylene glycol) 500
(Jeffamine ED-600, MW ∼600). At this stage, the long-
chained Jeffamines were expected to selectively couple to
the active chloride groups of the CNC outside the resin beads.
Finally the remaining chloride on the resins was capped with
dimethylamines (Scheme 1).
Although the resulting CS resins were less swellable than
the gel-type resins as a result of their higher degree of cross-
linking, their amphiphilic poly(ethylene glycol) shell permit-
ted them to be compatible with a wide range of hydrophilic/
hydrophobic solvents. In the procedure described above, we
considered the amount of CNC added to be a determining
factor in the final resin properties, because CNC plays the
role of both cross-linkers and precursor of the amino-
functional groups. Therefore, three types of resins were
(8) (a) Kang, J.; Schwendeman, S. P. Macromolecules 2003, 36, 1324.
(b) Heinemann, M.; Limper, U.; Buchs, J. J. Chromatogr., A 2004, 1024,
45.
(6) http://www.beadtech.co.kr.
(7) No remaining amino group was left, which was confirmed using a
qualitative ninhydrin test.
(9) Average of three experiments (Fmoc-titration).
3274
Org. Lett., Vol. 6, No. 19, 2004

Scheme 2. SPPS and Photocleavage Reactionsa
a Reagents and conditions: (i) Fmoc-protected 4-[4-(1-aminoethyl)-2-methoxy-5-nitrophenoxy]butyric acid, BOP, DIPEA, NMP; (ii)
25% piperidine in NMP; (iii) Fmoc-amino acid, BOP, HOBt, DIPEA, NMP; repeat (ii) and (iii); (iv) 25% piperidine in NMP; (v) 95%
TFA, 2.5% triethylslane, 2.5% H2O; (vi) hν, 360 nm, MeOH.
loading level proves that most of the CNC was used up in
the cross-linking reactions and, consequently, there were very
few active chlorides left to couple with the Jeffamine. The
loadings of CS resin II and CS resin III were 0.2-0.4 and
0.5-0.7 mmol/g, respectively. However, when we tried to
load Fmoc-Leu on the CS resin III, complete attachment was
possible only after four repeated couplings, which indicates
that the local density of amino groups on the shell was too
high for the reagents to obtain free access. On the basis of
the above results, we chose CS resin II as the optimal core-
shell resin.
phenoxy]butyric acid was attached to the resin and
Leu-enkepalinamide was synthesized by Fmoc-chemistry.10
The final peptide was released by UV light at 360 nm
(Scheme 2).11 The results with TentaGel-NH2 resin (0.30
mmol/g) and AM-PS (1% cross-linked) resin were compared
with those of CS resin (0.37 mmol/g). For the sake of
obtaining a fair comparison, the original loadings of PS resin
(1.1 mmol/g) were reduced by controlling the equivalence
of the photocleavable linker during the coupling. Finally,
0.27 and 0.58 mmol/g of the PS resin-bound photolinker were
obtained, and the remaining free amino groups were capped
by acetylation.
Interestingly, in the course of assembling the peptide on
the resins, we found that the loading of the first amino acid
on the CS resin proceeded much more efficiently than that
on the other resins. In the solid-phase organic synthesis, the
initial loading is known to be troublesome step because of
the matrix effect of the polymer backbone.12 However, this
problem could be minimized on the CS resins.
To examine the detailed kinetics, the coupling yields of
Fmoc-Leu were investigated according to the coupling time.
Figure 1. Cross-sectioned confocal microscope images of FITC-
coupled resins: (a) TentaGel resin (0.3 mmol NH2/g); (b) AM-PS
resin (0.3 mmol NH2/g); (c) CS resin I (0.5-2 equiv of CNC used,
0.03-0.1 mmol NH2/g); (d) CS resin II (3-5 equiv of CNC used,
0.2-0.4 mmol NH2/g); (e) CS resin III (7-10 equiv of CNC used,
0.5-0.7 mmol NH2/g).
To confirm the performance in solid-phase synthesis, CS
resins were applied to the photocleavage reaction, which is
often preferable to using acidic or basic conditions owing
to its orthogonality. The o-nitrobenzyl photocleavable linker,
Fmoc-protected 4-[4-(1-aminoethyl)-2-methoxy-5-nitro-
Figure 2. Initial loading time of Fmoc-Leu on the resins: (b) CS
resin, (3) TentaGel resin, (O) LL PS resin, (1) HL PS resin.
Org. Lett., Vol. 6, No. 19, 2004
3275

The coupling rate on the CS resin was approximately 2-4
times faster than that on TentaGel or the low-loading (LL)
PS resin and high-loading (HL) PS resin (Figure 2).13
After the solid-phase peptide synthesis, photocleavage
reactions were carried out on each resin. As expected,
compared to the other resins, the CS resin exhibited an
excellent performance in the photocleavage reaction. The CS
resin released the product rapidly and gave more than 95%
yield within 1 h. In contrast, TentaGel resin released just
72% of the product, and LL PS resin and HL PS resin
released only 41% and 34% of the product, respectively, even
after 3 h (Figure 3). The difference in yield between TentaGel
and PS resins may be partly due to the swelling difference
in methanol.14 Unlike the gel resins, the performance of the
CS resin was not related to the swelling ability.
In conclusion, we developed novel core-shell-type resins,
in which the functional groups were located on the shell
layer. They were readily prepared from 1% cross-linked PS
resins via two reaction steps. Regardless of the swelling
ability, the core-shell resin showed excellent results in the
photocleavage reaction. In addition, the initial loading onto
the CS resins progressed more efficiently than that on the
other gel-type resins. Further studies on the enzyme compat-
ibility of the CS resin and the optimization of the length of
the PEG chains are underway.
Acknowledgment. The authors wish to convey their
appreciation to the Brain Korea 21 Program supported by
the Ministry of Education for their financial supports. Also,
this research, under contract project code MS-04-211, has
been supported by the Intelligent Microsystem Center (IMC;
Http://www.microsystem.re.kr), which carries out one of the
21st century’s Frontier R&D Projects sponsored by the Korea
Ministry of Science & Technology.
Supporting Information Available: Experimental pro-
cedures for the preparation of CS resins, solid-phase peptide
synthesis, swelling properties of resins, HPLC chromatogram,
and MS spectra of peptides. This material is available free
of charge via the Internet at http://pubs.acs.org.
OL048815Q
(10) Holmes, C. P. J. Org. Chem. 1997, 62, 2370.
(11) The product was isolated with cold ether and analyzed with HPLC
and MALDI-TOF MS.
(12) (a) Li, W.; Yan, B. J. Org. Chem. 1998, 63, 4092. (b) Basso, A.;
Evans, B.; Pegg, N.; Bradley, M. Tetrahedron Lett. 2000, 41, 3763.
(13) We used rather stringent condition (equimolar addition of Fmoc-
Leu, HOBt, BOP, and DIEA, 2 equiv each) on purpose for comparison of
the CS resin and TentaGel. When we used an excess amount of DIEA (1.5
equiv of other reagents), the CS resin and TentaGel showed similar
properties in the first amino acid coupling.
(14) Usually, methanol or PBS buffer has been used as a solvent in this
type of reaction, because the cleaved product is water- or methanol-soluble
free peptide.10
Figure 3. Kinetic profiles of photolytic cleavage on each resin:
(b) CS resin, (3) TentaGel resinm (O) LL PS resinm (1) HL PS
resin.
3276
Org. Lett., Vol. 6, No. 19, 2004