The pivaloylglycol anchor group: A new platform for a photolabile linker in solid-phase synthesis

Article abstract of DOI:10.1021/jo9816055

We have designed a new photolabile linker (2) based on 2- pivaloylglycerol for the solid-phase synthesis of acids. The linker was prepared in six steps and anchored to the support via an amide bond. Photocleavage is a two-step process, in which the immobilized acids are released by photolyric generation of a radical center and subsequent spontaneous β-C,O bond scission. The pivaloyl linker (2) was found to cleave with high yields and purities the acids in various solvents (THF, CH₂Cl₂, dioxane, DMSO) by irradiation with light above 320 nm. Using this linker, we have demonstrated the solid-phase synthesis of test compounds by peptide synthesis, palladium-catalyzed cross coupling, and epoxidation. The linker proved to be stable toward the treatment with acids and bases. The photolysis rates of our pivaloyl linker (2) were compared with the rates of a o- nitrobenzyl photolinker (1) and proved to be superior.

Full text of DOI:10.1021/jo9816055

J . Org. Chem. 1998, 63, 9045-9051
9045
Th e P iva loylglycol An ch or Gr ou p : A New P la tfor m for a
P h otola bile Lin k er in Solid -P h a se Syn th esis
Stefan Peukert and Bernd Giese*
Department of Chemistry, University of Basel, St. J ohanns-Ring 19, CH-4056 Basel, Switzerland
Received August 6, 1998
We have designed a new photolabile linker (2) based on 2-pivaloylglycerol for the solid-phase
synthesis of acids. The linker was prepared in six steps and anchored to the support via an amide
bond. Photocleavage is a two-step process, in which the immobilized acids are released by photolytic
generation of a radical center and subsequent spontaneous â-C,O bond scission. The pivaloyl linker
2 2
(2) was found to cleave with high yields and purities the acids in various solvents (THF, CH Cl ,
dioxane, DMSO) by irradiation with light above 320 nm. Using this linker, we have demonstrated
the solid-phase synthesis of test compounds by peptide synthesis, palladium-catalyzed cross coupling,
and epoxidation. The linker proved to be stable toward the treatment with acids and bases. The
photolysis rates of our pivaloyl linker (2) were compared with the rates of a o-nitrobenzyl photolinker
(1) and proved to be superior.
In tr od u ction
The rapidly growing field of combinatorial chemistry
has renewed interest in the use of solid-phase organic
synthesis techniques as a convenient means of as-
sembling molecules.1 In addition to synthetic methods
amenable to a solid-phase approach, the need for im-
proved and novel linkers anchoring the molecules to the
support has greatly expanded. Since their first descrip-
tion, photolabile linkers have received considerable at-
tention during the past 25 years.3
,2
It is widely recognized that photolysis offers a mild
method of cleavage that takes place under neutral
conditions. This detachment is orthogonal to acidic and
basic reaction conditions and therefore affords additional
flexibility in the synthesis on solid support. The release
of the molecules can occur after removal of any protecting
groups and washing of the resin, thereby affording the
compounds suitable for biological screening without
contamination by cleavage reagents.
F igu r e 1.
tagging moieties.8 Less extensively, phenacyl-9 and
benzoin-based10 linkers have been utilized in solid-phase
chemistry in recent years. A new photolinker based on
a benzyl thioether that releases molecules without an
additional functional group has also been reported.
However, its use is restricted to molecules with a biaryl
unit.11
To provide more options in this area, we wish to report
the development of a new photolabile linker (2) with a
pivaloylglycol moiety (bold type in compound 2, Figure
Photocleavage facilitates a controlled release of mol-
ecules with the possibility to control the amount of
liberated compound by the exposure time to UV light.
Photolabile linkers proved to be valuable for the release
of both ligands and tagging molecules.
Several photolabile linkers are currently available.
1
), which is based on radical-induced â-C,O bond cleav-
o-Nitrobenzyl compounds (1) have been widely employed
4
age.
by many researchers for the generation of peptides,
5
6
7
oligosaccharides, oligonucleotides, small molecules, and
Ba ck gr ou n d
(
1) For reviews on combinatorial chemistry, see: (a) Frobel, K.;
Our mechanistic studies on biological processes showed
that radical-induced â-C,O bond cleavage plays a pivotal
Kr a¨ mer, T. Chem. Unserer Zeit 1996, 30, 270. (b) Balkenhohl, F.; von
dem Bussche-H u¨ nnefeld, C.; Lansky, A.; Zechel, C. Angew. Chem., Int.
Ed. Engl. 1996, 35, 2288.
(
2) For reviews on solid-phase organic synthesis, see: (a) Thompson,
(6) Matray, T. J ., Greenberg, M. M. J . Am. Chem. Soc. 1994, 116,
6931.
(7) (a) Holmes, C. P. J . Org. Chem. 1997, 62, 2370. (b) Ruhland, B.;
Bhandari, A.; Gordon, E. M.; Gallop, M. A. J . Am. Chem. Soc. 1996,
118, 253. (c) Holmes, C. P.; J ones, D. G. J . Org. Chem. 1995, 60, 2318.
(8) Brown, B. B.; Wagner, D. S.; Geysen, H. M. Mol. Diversity 1995,
1, 4.
L. A.; Ellman, J . A. Chem. Rev. 1996, 96, 555. (b) Fr u¨ chtel, J . S.; J ung,
G. Angew. Chem., Int. Ed. Engl. 1996, 35, 17.
(3) For a general review of the use of photolabile supports, see:
Lloyd-Williams, P.; Albericio, F.; Giralt, E. Tetrahedron 1993, 49,
1
1065.
(4) (a) Rich, D. H.; Gurwara S. K. J . Chem. Soc., Chem. Commun.
1
973, 610. (b) Lloyd-Williams, P.; Albericio, F.; Giralt, E. Int. J . Pept.
(9) (a) Wang, S.-S. J . Org. Chem. 1976, 41, 3258. (b) Bellof, D.;
Mutter M. Chimia 1985, 39, 317.
(10) Rock, R. S.; Chan, S. I. J . Org. Chem. 1996, 61, 1526.
(11) Sucholeiki, I. Tetrahedron Lett. 1994, 35, 7307.
Protein Res. 1991, 37, 58.
5) Nicolaou, K. C.; Watanabe, N.; Li, J .; Pastor, J .; Winssinger, N.
Angew. Chem., Int. Ed. Engl. 1998, 110, 1559.
(
1
0.1021/jo9816055 CCC: $15.00 © 1998 American Chemical Society
Published on Web 11/10/1998

9
046 J . Org. Chem., Vol. 63, No. 24, 1998
Peukert and Giese
Sch em e 1a
Sch em e 2a
a
Reagents: (a) TentaGel S NH2, DIC, CH2Cl2; (b) (HF)3‚Et3N,
THF; (c) Fmoc-Gly-anhydride, DMAP; (d) biphenylcarboxylic acid,
DIC, DMAP, THF.
a
Reagents: (a) MeC(OMe)3, CSA, dioxane, 81%; (b) TBDMSCN,
t
KCN, 18-crown-6, DMF, 84%: (c) (1) BuLi, CuI, Et2O, (2) AcOH,
4%; (d) K2CO3, MeOH, quant; (e) BrCH2CO2 Bu, Ag2O, DMF,
2%; (f) TMSOTf, 2,6-lutidine, 96%.
t
butyl bromoacetate in the presence of silver(I) oxide to
yield 8. The use of the tert-butyl ester was requisite to
avoid lactonization by participation of the tertiary hy-
droxyl group. This was followed by release of the acid
and protection of the tertiary alcohol with trimethylsilyl
triflate to afford linker 9. This procedure requires
chromatography only after the alkylation step. Inter-
mediates 4 and 5 can be distilled, and intermediates 6
and 7 and the final product are isolated by filtration over
silica gel to afford gram quantities of 9 in 40% overall
yield.
8
7
role in several systems. C-4′ DNA radicals can undergo
a spontaneous heterolytic â-C,O bond scission, resulting
in the cleavage of the DNA strand.12 Radical-induced
cleavage also takes place in the enzymatic deoxygenation
of ribonucleotides by ribonucleotide reductase via a C-3
nucleoside radical.13 A similar mechanism can operate
in the â-elimination of phosphocholine from C-2 lysoleci-
thin radicals.14 Radicals were generated for our mecha-
nistic studies on biological systems by photolysis of
suitable radical precursors. The combination of the two
reactions, photolytic generation of a radical center and
subsequent spontaneous â-C,O bond scission, is utilized
in our concept for a new photolabile linker.
Cou p lin g to th e Solid P h a se a n d Atta ch m en t of
Acid s. With linker 9 in hand, we set out to explore its
photolytic properties and its use in solid-phase chemistry.
Coupling of 9 to commercially available amino support
(
TentaGel S NH
silyl-protected linker with diisopropylcarbodiimide (Scheme
). The anchoring proceeded with complete conversion
2
) was conducted by activation of the
2
Resu lts a n d Discu ssion
1
6
as judged by the Kaiser test. Desilylation of the two
hydroxyl groups was conducted with triethylamine tris-
hydrofluoride on the solid phase to afford the correspond-
ing photolabile support 11. Experiments in the homo-
geneous phase showed that this deprotection reaction
proceeds with >75% yield.
Syn th esis. The linker described below is based on a
C-2 pivaloyl-substituted glycerol. tert-Butyl ketones are
suitable as photolabile radical precursors when exposed
to light below 340 nm.15 The most common method of
preparing photolabile supports is to anchor linkers bear-
ing a carboxylic acid tether onto the support as an amide.
Combining these facts, a suitable linker may be derived
from pivaloyl-substituted glycerol by incorporation of a
carboxylic acid moiety on one of the two primary alcohols.
The remaining primary alcohol can then be used to
immobilize the substrate.
To study the photolytic properties of the linker, the
photolabile support was loaded with Fmoc-glycine (12a )
and biphenylcarboxylic acid (12b). Attachment of amino
acids in general was achieved using the symmetrical
anhydride approach. The level of first amino acid at-
tachment was estimated by piperidine-mediated cleavage
of the Fmoc-protecting group and UV-spectroscopic analy-
sis. Aromatic carboxylic acids were attached using the
standard carbodiimide coupling protocol. These reactions
The synthesis of carboxylic acid 9 is outlined in Scheme
1
. The dimer of 1,3-dihydroxyacetone 3 was transformed
1
4
via ortho ester 4 into the silyl-protected cyanohydrin
. Copper(I)-catalyzed nucleophilic attack of tert-butyl-
5
13
17
were checked qualitatively by gel-phase C NMR. In
no case was esterification of the tertiary hydroxyl group
observed.
lithium and subsequent acidic hydrolysis yielded the tert-
butyl ketone 6. Deacetylation occurred with concomitant
migration of the silyl group, affording compound 7. The
carboxyl group was introduced by alkylation with tert-
Mech a n ism of P h otoclea va ge. Photocleavage starts
with generation of an R-hydroxyalkyl radical (12b f 13)
via Norrish I reaction. To gain insight into the photo-
cleavage mechanism not only resin 12b but also model
compound 17 was studied (Scheme 3).
(12) (a) Giese, B.; Beyrich-Graf, X.; Erdmann, P.; Petretta, M.;
Schwitter, U. Chem. Biol. 1995, 2, 367. (b) Von Sonntag, C.; Hagen.
U.; Sch o¨ n-Bopp, A.; Schulte-Frohlinde, D. Adv. Radical Biol. 1981, 9,
1
09.
13) (a) Licht, S.; Gerfer, G. J .; Stubbe, J . Science 1996, 271, 477.
b) Lenz, R.; Giese, B. J . Am. Chem. Soc. 1997, 119, 2784.
14) M u¨ ller, S. N.; Batra, R.; Senn M.; Giese, B.; Kisel, M.; Shadyro,
O. J . Am. Chem. Soc. 1997, 119, 2795.
15) Bamford, C. H.; Norrish, R. G. W. J . Chem. Soc. 1935, 1504.
(
(
(16) Kaiser, E.; Colescott, R. L.; Bossinger, C. D.; Cook, P. I. Anal.
Biochem. 1970, 34, 595.
(17) Look, G. C.; Holmes, C. P.; Chinn, J . P.; Gallop, M. A. J . Org.
Chem. 1994, 59, 7588.
(
(

Pivaloylglycol Anchor Group
Sch em e 3
J . Org. Chem., Vol. 63, No. 24, 1998 9047
Ta ble 1. P u r ity a n d Yield of P h otoclea va ge in Differ en t
Solven ts
solvent
2 2
THF CH dioxane DMSO 2-propanol acetonitrile
Cl a
purity (%) 90
92
90
91
90
91
87
85
78
28
yield (%)
92
a
Contains 0.02% 2-methyl-2-butene.
F igu r e 2. Photolysis rates for the pivaloylglycol linker 12a
at different wavelengths: triangles, λ > 305 nm; circles, λ >
3
20 nm; stars, λ > 335 nm.
determined here. Apart from this, the cleavage is strik-
ingly solvent independent, affording yields from 78% to
The UV spectrum of 17 exhibits n-π* carbonyl absor-
bances between 260 and 340 nm with a maximum at 297
nm (ꢀ ) 32). The quantum yield for the photolytic radical
generation (17 f 18) at 308 nm in acetonitrile was
determined to be Φ ) 0.56. Our proposed mechanism of
9
2% and purities of 85% to 92%. THF was used as
solvent of choice for all other experiments except for
photolysis in open well plates where dioxane was pre-
ferred because of its lower volatility.
The next step was to investigate the wavelength
dependence of photolysis rates (Figure 2). Above 345 nm,
no measurable photocleavage took place. Photocleavage
started when using the 335 nm cutoff filter (half-life of
the subsequent â-C,O bond scission (13 f 14) is based
on earlier investigations of C-2 glycerol radicals.14 These
experiments suggest a concerted elimination, where the
glycerol radical 19 is converted into enolate radical 20
and acetic acid via a seven-membered hydrogen-bridged
transition state. The fate of enolate radical 14 was
2
5 min). Use of the 320 nm filter greatly increased the
photolysis rate (half-life of 2 min), which could be further
increased with the 305 nm filter (half-life of 0.5 min).
In all cases, photocleavage stopped at roughly 90%
conversion even after extended periods of exposure to UV
light.
1
3
investigated on the solid phase by gel-phase C NMR of
the cleaved support. Irradiation of resin 12b in THF
resulted in a clean conversion to resin 16. This can be
explained by H atom abstraction of the immobilized
enolate radical 14 from the solvent to give resin-bound
acetone (16).
All further irradiations were conducted using the 320
nm cutoff filter. This allowed fast photolytic cleavage (12
min irradiation time for 90% conversion) and minimiza-
tion of the possible damage by short-wavelength UV light.
Ap p lica tion s of th e Lin k er in Solid -P h a se Syn -
th esis. The use of the photolabile linker was explored
by various reactions on the solid phase and subsequent
photolytic release. As our first example, we chose a
P h otolytic P r op er ties. Cleavage yield, purity, and
photolysis rates of the released Fmoc-glycine in various
solvents and at different wavelengths were determined
by photolyzing the suspended resin beads 12a in quartz
glass cells. The cleavage yield and photolysis rates were
measured by UV spectroscopy of the supernatant solution
at 290 nm, whereas the purity of the released acid was
checked via RP-HPLC analysis. The lamp used was a
1
8
peptide synthesis employing Fmoc chemistry.
Leu-
enkephalin, chosen as a model peptide, was prepared by
stepwise coupling of Fmoc-protected amino acids to resin
12c (Scheme 4). After four reaction cycles, resin 21 was
obtained with 0.147 mmol/g loading of protected Leu-
enkephalin. By treatment with piperidine and TFA, Leu-
enkephalin was deprotected to yield resin 22 prior to
photocleavage. Both peptides (21 and 22) could be
cleaved photolytically in quartz glass cells after 12 min
of irradiation and afforded the desired peptides in high
purity and yield. The principal HPLC peak of the
photocleavage from resin 22 comigrated with authentic
5
00 W Hg high-pressure lamp equipped with cutoff
filters, thus affording light above the cutoff wavelengths.
The power level of the light beam was adjusted to 1000
2
mW/cm (measured between 300 and 400 nm).
Six different solvents were chosen for the photolysis
study: THF, dioxane, dichloromethane, and acetonitrile
represent aprotic organic solvents, 2-propanol represents
a protic organic solvent, and DMSO since it is often used
as a solvent for drugs in biological assays. Table 1
summarizes the results for the solvents examined.
Acetonitrile turned out to be unsuitable under the
conditions of photocleavage; therefore, the yield was not
(18) Field, G. B. et al. Int. J . Peptide Protein Res. 1990, 35, 161.

9
048 J . Org. Chem., Vol. 63, No. 24, 1998
Peukert and Giese
Sch em e 4a
a
Reagents: (a) (1) piperidine, DMF; (2) Fmoc-amino acid, PyBOP, DIPEA, DMF; (1) + (2): 4×; (b) (1) piperidine, DMF; (2) TFA.
Sch em e 5
Sch em e 6
Leu-enkephalin, and its identity was confirmed by mass
spectrometry.
Additionally, to demonstrate the wide applicability of
the photolabile anchor group, palladium-catalyzed cross-
couplings were performed with resin-bound iodobenzoic
acid 12d (Scheme 5). Stille19 and Suzuki coupling with
tributylphenyltin and phenylboronic acid, respectively,
gave immobilized biphenylcarboxylic acid 12b.
20
Ta ble 2. Sta bility of th e Lin k er tow a r d Tr ea tm en t w ith
Acid s a n d Ba ses
yielda
(%)
yielda
(%)
In the course of the reactions, the beads turned dark.
However, the acid 15b was released in both cases with
acid
base
9
0
3% purity. On the basis of the initial resin loading of
.29 mmol/g, the amount of cleaved biphenylcarboxylic
TFA/CH2Cl2, rt
93 DIPEA/THF, 60 °C
90 2,6-lutidine/THF, 60 °C
89 DBU/toluene, 80 °C
100 K(NSiMe2)/THF, -78 °C
99
99
92
81
70
BF3‚Et2O/CH2Cl2, rt
AcOH/H2O/THF, 60 °C
HCl1M/THF, rt
acid corresponded to an overall yield of 50% (Stille
coupling) and 72% (Suzuki coupling), respectively, for five
steps.
p-TsOH/toluene, 80 °C
a
Yield of photolysis compared to yield of photolysis of untreated
The detachment of acid-labile substrates from the solid
phase is often foiled, as most linkers are cleaved by strong
acids. Because of the inherent neutral cleavage condi-
tions, a photolabile support can serve as an alternative
for theses substrates, as was demonstrated with the
immobilized acetal 12e and epoxide 12g. The synthesis
of the epoxide was carried out by epoxidation of resin-
bound vinylbenzoic acid (12f). Photolysis in THF liber-
ated the substrates 15e and 15g in good yield and high
purity (Scheme 6).
resin.
Sta bility of th e Lin k er . The stability of the linker
toward acids and bases was further examined by incu-
bating resin 12b with various commonly used acids and
bases for 2 h. Then, the resins were washed, dried, and
subjected to photolysis. Yields of the photocleavage were
compared with results of the photolysis of untreated resin
1
2b and are given in Table 2. The linker shows good to
excellent stability toward Brønsted and Lewis acids and
toward nonnucleophilic bases. Strong nucleophilic re-
agents such as hydrazine or sodium methanolate, how-
ever, result in complete cleavage of the substrates by
attacking the ester linkage.
Com p a r ison of P iva loyl a n d o-Nitr oben zyl Lin k -
er s. Among the known photolabile linkers, the veratryl-
based o-nitrobenzyl linkers developed by Affymax are
The results demonstrate that the pivaloylglycol linker
is compatible with a wide range of reaction conditions
(peptide syntheses, Pd-catalyzed cross-couplings, oxida-
tions) and allows for the detachment of acids from the
polymeric support under mild and neutral conditions.
(
19) Stille coupling on solid phase: Deshpande, M. S. Tetrahedron
Lett. 1994, 35, 5613.
20) Suzuki coupling on solid phase: Ruhland, B.; Bombrun, A.;
Gallop, M. A. J . Org. Chem. 1997, 62, 7820.
7
c,21
commercially available and most frequently used.
Photolabile support 1 (compound ) biphenylcarboxylic
(

Pivaloylglycol Anchor Group
J . Org. Chem., Vol. 63, No. 24, 1998 9049
F igu r e 3. Photolysis rates in dependency of the power level for pivaloylglycol linker 12b (bold lines) and o-nitrobenzyl linker 1
2
2
2
(
dotted lines): circles, 500 mW/cm ; triangles, 200 mW/cm ; squares, 50 mW/cm .
the resin was covered with dioxane, which is reported to
7
a
be the best solvent for o-nitrobenzyl linkers. With a
5
00 W Hg lamp equipped with a 280-400 nm dichroic
mirror and 320 nm cutoff filter (power level 100 mW/
cm ), the samples were irradiated from above. The
2
results confirmed the findings of the photolyses in quartz
glass cells. The 10 min irradiation time for 90% conver-
sion for the pivaloyl linker 12b compared to 40 min for 1
underscored the advantage of our pivaloyl linker (Figure
4
).
Con clu sion
The new photolabile linker based on 2-pivaloylglycerol
F igu r e 4. Photolysis rates in 96-well microtiter plates: bold
line, pivaloylglycol linker 12b; dotted line, o-nitrobenzyl linker
represents an excellent alternative for the cleavage of
carboxylic acids. Photolytic cleavage is rapid and pro-
duces high yields of released acids with very high purity.
The linker may be photolyzed by 320-340 nm light,
which allows easy handling in the laboratory without
precautions. The photobyproducts are either volatile (CO
and isobutene/isobutane) or inert (resin-bound acetone)
and are transparent above 320 nm to prevent inner filter
effects from interfering with the course of the photolysis.
Finally, the linker is compatible with many reagents and
reactions, which allows broad applicability in combina-
torial chemistry. Further investigations in the utility of
photolabile supports based on this concept for the detach-
ment of compounds other than acids are underway and
will be reported in due course.
1
.
acid) was therefore chosen as a reference for the evalu-
ation of the rates of photochemical cleavage in compari-
son to our photolabile support 12b. To establish the
cleavable loading of acid on the resins, photolyses were
performed in quartz glass cells until no further conver-
sion was observed. The purity of the liberated biphenyl-
carboxylic acid was >95% with both linkers. In all
subsequent time- and power-dependent irradiations, the
yields are expressed relative to the cleavable loading
determined as described above. The results of photo-
cleavage in quartz glass cells are shown in Figure 3.
Photolysis of pivaloyl resin 12b is roughly three to four
times faster than that of o-nitrobenzyl resin 1 at the same
power level. Strikingly, the photocleavage from resin 1
became significantly slower with increasing conversion.
This effect is presumably due to the formation of support-
bound nitroso ketone, which may act as an internal light
filter. With the pivaloyl-based linker that generates only
CO, isobutene, and resin-bound acetone upon photocleav-
age this effect is not observed as the photoproducts are
transparent above 300 nm. With both linkers, the
support-bound photolysis becomes faster with increasing
power level of the UV light, albeit no linearity was
observed. This is probably due to light scattering,
shielding, or shadowing effects of the resin beads, which
causes a large proportion of the light to remain unused.
Exp er im en ta l Section
Gen er a l Meth od s. All quoted temperatures are uncor-
rected. Materials and reagents were of the highest grade
available commercially and used without further purification.
2
Commercial resin (TentaGel S NH ) was obtained from Rapp
Polymere, T u¨ bingen, Germany. THF was distilled from potas-
sium/benzophenone before use. The level of attachment of
amino acids was estimated according to the protocol in the
Novabiochem handbook.21 H and C NMR data were mea-
sured in the indicated solvent with tetramethylsilane as
internal standard. The gel phase marks data that are taken
1
13
3
from resin swollen in CDCl . Fast atom bombardment (FAB)
mass spectra were obtained with m-nitrobenzyl alcohol as
matrix. Combustion analyses were performed by the Mik-
roanalytisches Labor, University of Basel. Quantum yield was
2
Therefore, a power level of 50-100 mW/cm seems to be
sufficient for irradiation of photolabile supports.
The kinetics of the photolytic cleavage of the two
linkers were also examined in 96-well microtiter plates.
In each well was placed a few milligrams of resin, and
determined at 308 nm and di-tert-butyl ketone as reference
compound.22 Analytical HPLC was performed on (a) Kontron
Instruments and (b) Waters Alliance 2690 using columns as
follows: (a) Merck Lichrospher 100, RP 18 (5 µm), 250 mm ×
(
21) Novabiochem AG, L a¨ ufelfingen, Switzerland, Catalog & Peptide
(22) Adam, W.; Fragale, G.; Klapstein, D.; Nau, W. M.; Wirz, J . J .
Am. Chem. Soc. 1995, 117, 12578.
Synthesis Handbook, 1997/1998.

9
050 J . Org. Chem., Vol. 63, No. 24, 1998
Peukert and Giese
4
4
mm and (b) Rainin Microsorb, Type C18 (3 µm), 50 mm ×
.6 mm. Gradients used for elution were as follows: (a) 0-30
1H), 4.03 (d, 1H, J ) 16.7 Hz), 3.93 (d, 1H, J ) 16.7 Hz), 3.87
(d, 1H, J ) 9.6 Hz), 3.85 (d, 1H, J ) 10.0 Hz), 3.63 (d, 1H, J
) 10.0 Hz), 3.53 (d, 1H, J ) 9.6 Hz), 1.48 (s, 9H), 1.27 (s, 9H),
min, 30-100% B in A with A ) 0.1% TFA in water and B )
acetonitrile; (b) 0-1 min, 100% A; 1-13 min, 0-100% B in A;
1
0
0.88 (s, 9H), 0.06 (s, 3H), 0.05 (s, 3H); 13C NMR (CDCl
) δ
216.9, 170.0, 85.2, 82.1, 75.4, 69.2, 66.9, 44.9, 28.1, 26.6, 25.9,
3
3-15 min, 100% B with A ) 0.1% TFA in water and B )
.1% TFA in water/acetonitrile (1:9).
-
1
18.3, -5.4, -5.5; IR (film) 3440, 1732, 1694, 1253 cm ; MS
+
5
-[(ter t-Bu tyld im eth ylsilyl)oxy]-5-cya n o-2-m eth oxy-2-
(FAB) 405 (MH) . Anal. Calcd for C20
9.96. Found: C, 59.21; H, 9.98.
40 6
H O Si: C, 59.37; H,
1
4
m eth yl-1,3-d ioxa n e (5). To a stirred solution of ketone 4
1.75 g, 12.0 mmol) in DMF (13 mL) were added a few crystals
(
6-[(ter t-Bu tyldim eth ylsilyl)oxy]-3-oxa-5-pivaloyl-5-[(tr i-
m eth ylsilyl)oxy]h exa n oic Acid (9). To a solution of ester
8 (2.30 g, 5.68 mmol) in dry THF (20 mL) were added at 0 °C
2,6-lutidine (1.98 mL, 17 mmol) and trifluoromethanesulfonic
acid trimethylsilyl ester (TMSOTf, 2.57 mL, 11.4 mmol). The
solution was allowed to warm to room temperature and heated
to 40 °C after 1 h. After an additional 2 h, 2,6-lutidine (1.32
mL, 11.4 mmol) and TMSOTf (1.54 mL, 8.5 mmol) were added.
The reaction was quenched by addition of acetic acid (20 mL,
30% in water) and extracted with ether (200 mL). The organic
of potassium cyanide and 18-crown-6, and the mixture was
cooled to 0 °C under argon. tert-Butyldimethylsilyl cyanide
(1.95 g, 13.8 mmol) in DMF (3 mL) was added slowly. After
3
0 min, the mixture was diluted with ether (200 mL) and
washed with water (2 × 200 mL) and brine (100 mL), the
organic solvent was removed, and the residue was distilled in
a Kugelrohr oven (at 150 °C at 0.08 mbar). A diastereomeric
mixture of 5a and 5b (2.90 g, 84% yield) in a ratio of 8:1 was
1
obtained as a clear oil. Data for 5a : H NMR (CDCl
3
) δ 3.84
(
s, 4H), 3.30 (s, 3H), 1.50 (s, 3H), 0.87 (s, 9H), 0.24 (s, 6H); 13
C
4
phase was washed with brine (50 mL), dried (MgSO ), and
NMR (CDCl
3
) δ 120.2, 111.5, 66.2, 63.6, 51.0, 25.4, 21.2, 17.9,
concentrated in vacuo. Filtration over silica gel gave acid 9
1
1
-
3.5. Data for 5b: H NMR (CDCl
H), 1.46 (s, 3H), 0.90 (s, 9H), 0.24 (s, 6H); C NMR (CDCl
δ 118.8, 111.5, 66.2, 63.6, 50.8, 25.7, 20.9, 17.9, -3.8; IR (film)
3
) δ 3.84 (s, 4H), 3.31 (s,
(2.29 g, 96% yield) as off-white crystals: mp 84-86 °C; H
1
3
3
3
)
3
NMR (CDCl ) δ 4.06 (s, 2H), 3.89 (d, 1H, J ) 9.3 Hz), 3.79 (d,
1H, J ) 10.4 Hz), 3.73 (d, 1H, J ) 10.4 Hz), 3.61 (d, 1H, J )
237, 1263, 1152 cm-1; MS (CI) m/z 288 (MH) . Anal. Calcd
for C13 Si: C, 54.32; H, 8.77; N, 4.87. Found: C, 54.68;
H, 8.57; N, 4.83.
-Acetoxy-2-[(ter t-bu tyld im eth ylsilyl)oxy]-3-h yd r oxy-
-p iva loylp r op a n e (6). Cyanohydrin 5a ,b (2.60 g, 9.0 mmol),
+
9.3 Hz), 1.25 (s, 9H), 0.89 (s, 9H), 0.24 (s, 9H), 0.07 (s, 3H),
2
H
25NO
4
0.05 (s, 3H); 13C NMR (CDCl
3
) δ 216.7, 172.2, 88.6, 75.6, 68.4,
67.4, 45.1, 26.7, 25.9, 18.5, 2.1, -5.5; IR (film) 3100, 1732, 1698,
-
1
+
1
1250 cm ; MS (FAB) m/z 421 (MH) . Anal. Calcd for
Si : C, 54.25; H, 9.58. Found: 54.42; H, 9.57.
1,3-Bis[(ter t-bu tyld im eth ylsilyl)oxy]-2-h yd r oxy-2-p iv-
2
C
19
H
40
O
6
2
CuI (17.5 mg, 0.09 mmol), and dry diethyl ether (150 mL) were
stirred and cooled to -78 °C under argon. tert-Butyllithium
a loylp r op a n e (17). To a solution of monosilyl ether 7 (200
mg, 0.69 mmol) in DMF (0.7 mL) were added tert-butyldi-
methylsilyl chloride (167 mg, 1.1 mmol) and imidazole (150
mg, 3.2 mmol) and the mixture stirred for 18 h. The mixture
was diluted with ether (10 mL) and washed with water (3 ×
(22.6 mL, 36 mmol, 1.6 M in pentane) was added in 10 min,
upon which the reaction mixture turned dark green. After 3.5
h at -78 °C, acetic acid (30 mL, 50% in water) was rapidly
added. The reaction mixture was allowed to warm to room
temperature, additional acetic acid (100 mL, 30% in water)
was added, and the mixture was vigorously stirred overnight.
The phases were separated, the aqueous phase was extracted
with diethyl ether (2 × 50 mL), and the combined organic
phases were dried over magnesium sulfate and concentrated
in vacuo. Acetic acid was removed by coevaporation with
4
10 mL), and the organic phase was dried (MgSO ) and
concentrated in vacuo. Flash chromatography on silica gel
(hexanes/EtOAc 30:1) afforded the bissilyl ether 17 (251 mg,
90% yield) as a colorless oil: 1H NMR (CDCl
) δ 3.82 (d, 2H,
J ) 9.9 Hz), 3.58 (d, 2H, J ) 9.9 Hz) 3.38 (s, 1H), 1.25 (s, 9H),
0.88 (s, 18H), 0.05 (s, 6H), 0.04 (s, 6H); 13C NMR (CDCl
) δ
216.8, 85.2, 66.7, 44.7, 26.6, 25.8, 18.3, -5.4, -5.5; IR (film)
3
3
toluene and the residue filtered over silica gel to give ketone
1
-1
+
6
(2.45 g, 84% yield) as a white solid: mp 61-62 °C; H NMR
3541, 1698, 1258 cm ; MS (FAB) 405 (MH) ; UV (MeCN) 297
(CDCl
3
) δ 4.37 (d, 1H, J ) 11.7 Hz), 4.22 (d, 1H, J ) 11.7 Hz),
44 4 2
(ꢀ ) 32). Anal. Calcd for C20H O Si : C, 59.35; H, 10.96.
Found: C, 59.24; H, 10.80.
3
.83 (dd, 1H, J ) 9.2, 11.2 Hz), 3.65 (dd, 1H, J ) 3.6, 11.2
Hz), 2.21 (dd, 1H, J ) 3.6, 9.2 Hz), 2.05 (s, 3H), 1.30 (s, 9H),
P h otolin k er -Ten ta Gel (11). TentaGel S NH
0.29 mmol/g loading) was suspended in dry CH
2
2
(1.0 g, 0.28/
Cl (5 mL),
0
2
-
3
.98 (s, 9H), 0.23 (s, 3H), 0.21 (s, 3H); 13C NMR (CDCl
17.5, 170.3, 86.9, 68.2, 66.9, 45.1, 26.3, 26.0, 20.8, 18.7, -2.2,
3
) δ
2
and photolinker 9 (183 mg, 0.435 mmol), DMAP (9 mg, 0.07
mmol), and diisopropylcarbodiimide (DIC, 112 µL, 0.73 mmol)
were added. The resin was shaken for 18 h at room temper-
-
1
2.4; IR (film) 3518, 1748, 1698, 1254 cm ; MS (FAB) m/z
33 (MH) . Anal. Calcd for C16H O Si: C, 57.79; H, 9.70.
32 5
+
Found: C, 57.77; H, 9.77.
ature and washed with CH
revealed complete reaction. The resin was then suspended in
THF (4 mL) and (HF) ‚Et N (1 mL) was added. After 24 h
the resin was washed with THF, CH Cl and dried to yield
2 2 2 2
Cl , DMF, and CH Cl . Kaiser test
1
-[(ter t-Bu t yld im et h ylsilyl)oxy]-2,3-d ih yd r oxy-2-p iv-
a loylp r op a n e (7). To a solution of acetate 6 (3.90 g, 11.7
mmol) in methanol (25 mL, 95% in water) was added potas-
sium carbonate (0.25 g), and the mixture was stirred for 5 h
at room temperature. The reaction mixture was filtered over
silica gel and the solvent removed to yield silyl ether 7 (3.40
g, quantitative) as a clear oil: H NMR (CDCl
3
3
2
2
1
3
the photolabile support 11 (1.01 g); gel phase C NMR (CDCl
δ 216.5, 170.1, 84.8, 76.1, 69.8, 65.7, 45.0, 38.8, 25.6.
3
)
Gen er a l P r oced u r e for Atta ch m en t of th e F ir st Am in o
Acid . Photolinker-TentaGel (11, 100 mg, 29 µmol) was
washed with DMF, a 0.5 M solution of symmetrical anhydride
(prepared from 10 equiv of Fmoc amino acid and 5 equiv of
1
3
) δ 3.86 (d, 1H,
J ) 9.8 Hz), 3.74 (d, 1H, J ) 9.8 Hz), 3.69 (dd, 1H, J ) 8.0,
1.3 Hz), 3.65 (s, 1H), 3.59 (dd, 1H, J ) 5.3, 11.3 Hz), 2.39
dd, 1H, J ) 5.3, 8.0 Hz), 1.26 (s, 9H), 0.88 (s, 9H), 0.07 (s,
1
(
3
4
DIC in 5 mL of CH
amino acid coupled to the resin for 1 h by the addition of
DMAP (1 equiv). The resin was washed successively with CH
Cl , DMF, and CH Cl and dried. Fmoc-Gly photolinker-
2 2
Cl at 0 °C) in DMF was added and the
H), 0.06 (s, 3H); 1 C NMR (CDCl
3
) δ 217.4, 84.2, 66.5, 65.8,
3
4.8, 26.4, 25.8, 18.2, -5.5, -5.6; IR (film) 3482, 1693, 1256
2
-
-
1
+
cm ; MS (FAB) m/z 291 (MH) . Anal. Calcd for C14
H
30
O
4
Si:
2
2
2
C, 57.89; H, 10.41. Found: C, 57.70; H, 10.23.
TentaGel (12a ): 0.183 mmol/g. Fmoc-Leu-photolinker-
TentaGel (12c): 0.191 mmol/g.
6
-[(ter t-Bu tyld im eth ylsilyl)oxy]-5-h yd r oxy-3-oxa -5-p iv-
a loylh exa n oic Acid ter t-Bu tyl Ester (8). To a solution of
alcohol 7 (2.80 g, 9.66 mmol) in DMF (20 mL) were added at
Gen er a l P r oced u r e for Atta ch m en t of Ar om a tic Ca r -
boxylic Acid s. To a suspension of photolinker resin 11 (200
mg, 58 µmol) in THF (2 mL) were added the aromatic
carboxylic acid (2 equiv), DMAP (0.2 equiv), and DIC (1.2
equiv), and the mixture was shaken for 18 h at room temper-
0
°C silver(I) oxide (3.35 g, 14.5 mmol) and tert-butyl bromo-
acetate (4.25 mL, 29 mmol). The suspension was allowed to
warm to room temperature and stirred in the dark for 18 h.
The reaction mixture was diluted with ether (200 mL) and
washed with water (3 × 50 mL), and the organic phase was
ature. Washing (THF and CH
acid-photolinker resins 12b,d,e,f. Gel phase C NMR (CDCl
2 2
Cl ) and drying afforded the
1
3
3
)
dried (MgSO
4
) and concentrated in vacuo. Flash chromatog-
data: 12b δ 215.1, 169.7, 166.4, 146.0, 139.8, 130.3, 129.1,
128.4, 127.3, 127.2, 83.6, 75.6, 69.7, 68.0, 45.3, 38.9, 26.8; 12d
δ 215.1, 169.7, 165.9, 137.8, 131.2, 129.2, 101.2, 83.4, 75.5, 69.7,
raphy on silica gel (hexanes/EtOAc 9:1) yielded ester 8 (2.80
1
g, 72% yield) as a colorless oil: H NMR (CDCl ) δ 4.08 (s,
3

Pivaloylglycol Anchor Group
J . Org. Chem., Vol. 63, No. 24, 1998 9051
6
1
1
1
8.1, 45.3, 38.9, 26.7; 12e δ 215.0, 169.7, 166.2, 143.4, 130.2,
29.8, 126.6, 102.9, 83.6, 75.6, 69.7, 68.1, 65.4, 45.3, 38.9, 26.8;
2f δ 215.1, 169.7, 166.2, 142.4, 136.0, 130.1, 126.3, 126.2,
17.0, 83.9, 75.6, 69.8, 68.0, 45.3, 38.9, 26.8.
R esin -Bou n d Acet on e (16). Biphenyl carboxylic acid-
photolinker-TentaGel (12b, 100 mg, 29 µmol) was suspended
in four portions of THF (3 mL, quartz glass cells) and
irradiated according to the general photolysis conditions (see
t
Leu -E n k ep h a lin (F m oc, Bu )-P h ot olin k er -Ten t a Gel
2 2
below), followed by washing with THF and CH Cl to yield
(
21) a n d Leu -En k ep h a lin -P h otolin k er -Ten ta Gel (22).
resin 17 (93 mg) and biphenyl carboxylic acid (4.0 mg, 68%
The peptide resins were prepared on Leu-photolinker-resin
yield): gel phase 13C NMR (CDCl
38.7, 26.2.
3
) δ 204.9, 169.0, 76.4, 69.6,
1
2c (200 mg, 0.191 mmol/g) according to the synthesis cycle
described below. All amino acids were N-Fmoc protected. The
side chain functionality of Tyr was protected as the tert-butyl
ether.
Cou p lin g. To the resin were added a 0.1 M solution of
amino acid (2.5 equiv) in DMF, a 0.5 M solution of PyBOP
Gen er a l P h otolysis Con d ition s. Photolyses in quartz
glass cells (1 cm path length, equipped with stir bar) were
conducted with 2-8 mg of resin suspended in 3 mL of solvent
in the beam of a 500 W Hg high-pressure lamp fitted with a
water filter and various cutoff filters. The power level
(measured between 300 and 400 nm) at the cell was adjusted
(
2.5 equiv) in DMF, and DIPEA (5 equiv). The suspension was
mixed at room temperature for 30 min, the supernatant was
sucked off, and the resin washed successively with DMF, CH
Cl , and DMF. The Kaiser test was negative in all cases.
Am in e Dep r otection . Piperidine (20%) in DMF (2 mL)
2
by a collimating lens between 50 and 1000 mW/cm . The cells
2
-
were maintained at 20 °C and irradiated horizontally with
gentle mixing of the beads by means of a magnetic stirrer.
After photolysis, the supernatant was analyzed by UV spec-
troscopy and reversed-phase HPLC.
2
was added, and the suspension was mixed for 7 min at room
temperature. The supernatant was sucked off, and the resin
was washed with DMF.
Following Fmoc removal from the final tyrosine residue,
side-chain protection was removed by treating the resin with
TFA/water/triethylsilane (95:2.5:2.5, 2 mL) and allowing the
Photolyses in 96-well microtiter plates were conducted with
2
-7 mg of resin and 0.2 mL of dioxane. Photolyses were
performed by irradiating the samples with a 500 W Hg high-
pressure lamp fitted with a 280-400 nm dichroitic mirror and
a 320 nm cutoff filter from above. The power level was
suspension to stand 1 h at room temperature. Leu-Enkephalin
2
adjusted to 100 mW/cm . After photolysis, a defined amount
t
(
Fmoc, Bu)-photolinker-TentaGel (21): 0.147 mmol/g.
of benzoic acid was added as internal standard. After mixing,
an aliquot was removed and analyzed by reversed-phase HPLC
with detection at 214 nm.
Ep oxid a tion : 4-Oxir a n ylben zoic Acid -P h otolin k er -
Ten ta Gel (12g). Resin-bound vinyl benzoic acid (12f, 100 mg,
2
9 µmol) was suspended in CH
2 2
Cl (1 mL), m-CPBA (55%, 31
Sta bility tow a r d Acid s a n d Ba ses. Biphenylcarboxylic
acid-photolinker-TentaGel (12b, 0.204 mmol/g, 3-6 mg each)
mg, 0.1 mmol) was added, and the suspension was shaken for
2
4 h. The resin was washed successively with CH
2
Cl
2
, THF,
was incubated for 2 h with 2 mL of (a) TFA (50% in CH
rt; (b) BF ‚Et O (5% in CH Cl
rt; (d) AcOH/H O/THF (3:1:1), 60 °C; (e) p-TsOH (1% in
toluene), 80 °C; (f) DIPEA (10% in THF), 60 °C; (g) 2,6-lutidine
10% in THF), 60 °C; (h) DBU (5% in toluene), 80 °C; and (i)
2
Cl
2
),
1
3
and CH
2
Cl
2
and dried: gel phase C NMR (CDCl
3
) δ 215.0,
3
2
2
2
), rt; (c) 1 M HCl (20% in THF),
1
69.7, 166.1, 143.6, 130.0, 129.4, 125.6, 83.6, 75.6, 69.8, 68.0,
1.8, 51.5, 45.2, 38.9, 26.8.
Stille Rea ction w ith Resin 12d . To a degassed suspen-
sion of polymer-bound iodobenzoic acid (12d , 100 mg, 28 µmol)
in anhydrous NMP (1 mL) were added Pd dba ‚CHCl (3.1 mg,
µmol) and AsPh (3.7 mg, 12 µmol). After 5 min, Bu SnPh
20 µL, 58 µmol) was added and the reaction allowed to stand
at 50 °C. After 6 h, again Pd dba ‚CHCl (1.5 mg, 1.5 µmol),
AsPh (1.8 mg, 6 µmol), and Bu SnPh (10 µL, 29 µmol) were
added. The reaction mixture was maintained at 50 °C for 36
2
5
(
K(NSiMe
2
) (5 equiv in THF), -78 °C; washed (THF and CH
2
-
)
2
3
3
2
Cl , in case of basic reagents first with 5% HOAc in CH
2
Cl
2
3
(
3
3
and dried, followed by photolysis. Yields were compared with
the photolysis yield of untreated resin 12b.
2
3
3
3
3
Ack n ow led gm en t. Support by the Swiss National
Science Foundation and by Novartis is appreciated. S.P.
is grateful for partial support in the form of a Kekul e´ -
Stipendium by the Fonds der Chemischen Industrie.
The authors thank Dr. M. Diggelmann (Novartis) for
valuable discussions and Dr. W. Nau (University of
Basel) for assistance in determining quantum yields.
2 2
h, washed with NMP, THF, and CH Cl , and dried to afford
dark-brown resin 12b.
Su zu k i Rea ction w ith Resin 12d . To a degassed suspen-
sion of polymer-bound iodobenzoic acid (12d , 100 mg, 28 µmol)
in DMF (1 mL) were added PhB(OH)
PdCl (dppf) (4 mg, 6 µmol), and Et N (39 µL, 0.28 mmol). After
8 h at 65 °C, the reaction mixture was washed with DMF
and CH Cl and dried to afford a brown resin 12b.
2
(14 mg, 112 µmol),
2
3
1
2
2
J O9816055