A Reinvestigation of the Preparation, Properties, and Applications of Aminomethyl and 4-Methylbenzhydrylamin

Article abstract of DOI:10.1021/jo9802269

Mild, efficient conditions have been developed for the preparation of 4-methylbenzhydrylamine polystyrene (MBHA) and aminomethyl polystyrene (AMPS) resins by a two-step procedure with synthons 1a and 1c. The products possess excellent swelling characteristics and acylate readily with linkers yielding useful derivatives, which retain good swelling and reactivity. Comparative studies with these resins, and their poly(ethylene glycol) (PEG) derivatives, yield insights into the role of spacer arm and environment effects in synthesis facilitation.

Full text of DOI:10.1021/jo9802269

3706
J . Org. Chem. 1998, 63, 3706-3716
A Rein vestiga tion of th e P r ep a r a tion , P r op er ties, a n d Ap p lica tion s
of Am in om eth yl a n d 4-Meth ylben zh yd r yla m in e P olystyr en e
Resin s1
J . Howard Adams, Ronald M. Cook, Derek Hudson,* Vasu J ammalamadaka,2
Matthew H. Lyttle, and Michael F. Songster
Solid Phase Sciences, 81 Digital Drive, Novato, California 94949
Received February 6, 1998
Mild, efficient conditions have been developed for the preparation of 4-methylbenzhydrylamine
polystyrene (MBHA) and aminomethyl polystyrene (AMPS) resins by a two-step procedure with
synthons 1a and 1c. The products possess excellent swelling characteristics and acylate readily
with linkers yielding useful derivatives, which retain good swelling and reactivity. Comparative
studies with these resins, and their poly(ethylene glycol) (PEG) derivatives, yield insights into the
role of spacer arm and environment effects in synthesis facilitation.
Sch em e 1
In tr od u ction
Amino-functionalized polystyrenes provide indispen-
sable platforms for attachment of handles and linkers,3,4
spacer arms,5 and other moieties [i.e., poly(ethylene
glycol) (PEG)]6 for facilitation of solid-phase-mediated
organic transformations (e.g., in peptide, DNA, and
combinatorial synthesis applications). In this regard,
4-methylbenzhydrylamine polystyrene (MBHA)7 and
aminomethyl polystyrene (AMPS)8 resins have been of
lasting durability and usefulness. Nevertheless, there
is a growing recognition that drastic processing condi-
tions during resin modification and incompleteness in the
multistep transformations involved can cause modifica-
tion of cross-linking and other undesired side reactions.
Standard procedures for the production of aminomethyl
polystyrenes,8 for example, use (hydroxymethyl)- or
(chloromethyl)phthalimide in TFA-CH2Cl2 mixtures at
elevated temperatures and trifluoromethanesulfonic acid
as the strong acid catalyst, conditions that result in
relatively inefficient incorporations and poorly swelling
materials. Modified procedures to avoid some of these
problems have been described recently.9,10 Similar prob-
lems occur as a result of the Friedel-Crafts acylation
with toluoyl chloride in the preparation of MBHA resin,
and additional complications arise from the Leuckart
reaction used to convert the resultant ketone to an amine
(i.e., molten ammonium formate at 165 °C). Incompletion
of the latter transformation, which leaves residual keto
functionality, can give rise to other side reactions, for
example, through Schiff’s base formation with incoming
amino groups.
These problems highlight the need to develop ad-
ditional mild, efficient means for the preparation of these
and other important resins for solid-phase synthesis.
Furthermore, much has been theorized and reported
about the role played by spacer arms and environment
effects in synthesis facilitation. For example, it is known
that the addition of poly(ethylene glycol) spacer arms to
solid-phase resins can often greatly improve synthetic
results; however, not much is yet known about the exact
causes of these observed improvements.
(1) Portions of this work were reported in preliminary form: Fifth
International Symposium on Solid-Phase Synthesis & Combinatorial
Chemical Libraries, London, England, Sep 2-6, 1997. Adams, J . H.;
Cook, R. M.; Hudson, D.; J ammalamadaka, V.; Lyttle, M. H.; Songster,
M. F. In Innovations and Perspectives in Solid-Phase Synthesis and
Combinatorial Chemical Libraries, 1998: Collected Papers, Fifth
International Symposium; Epton, R., Ed.; Mayflower Scientific: King-
swinford, England; in press.
(2) Present address: Siddco, Inc., 9000 S. Rita Rd., Bldg. 40, Tucson,
AZ 85747.
(3) For example, see: Albericio, F.; Kneib-Cordonier, N.; Biancalana,
S.; Gera, L.; Masada, R. I.; Hudson, D.; Barany, G. J . Org. Chem. 1990,
55, 3730-3743 and references therein.
(4) Review: Songster, M. F.; Barany, G. Methods Enzymol. 1997,
289, 126-174.
Resu lts
(5) Cook, R. M.; Adams, J . H.; Hudson, D. Tetrahedron Lett. 1994,
35, 6777-6780.
The route outlined in Scheme 1, which avoids use of
the Leuckart amination reaction, was investigated as a
means to access the desired resins from the related
synthons 1a and 1c. The preparation of (chloromethyl)-
phthalimide (1a ), which is readily commercially avail-
(6) Review: Barany, G.; Albericio, F.; Kates, S. A.; Kemp, M. Polym.
Prepr., Am. Chem. Soc. Div. Polym. Chem. 1997, in press.
(7) Matsueda, G. R.; Stewart, J . M. Peptides 1981, 2, 45-50.
(8) (a) Mitchell, A. R.; Kent, S. B. H.; Erickson, B. W.; Merrifield,
R. B. Tetrahedron Lett. 1976, 42, 3795-3798. (b) Mitchell, A. R.; Kent,
S. B. H.; Engelhard, M.; Merrifield, R. B. J . Org. Chem. 1978, 43,
2845-2852.
(9) Hider, R. C.; Goodwin, B. L. PCT Int. Patent 96/25440, 1996;
Chem. Abstr. 1996, 125, 248887.
(10) Zikos, C. C.; Ferderigos, N. G. Tetrahedron Lett. 1995, 36,
3741-3744.
S0022-3263(98)00226-6 CCC: $15.00 © 1998 American Chemical Society
Published on Web 05/08/1998

Polystyrene Resins
J . Org. Chem., Vol. 63, No. 11, 1998 3707
Sch em e 2
Ta ble 1. Op tim iza tion of Con d ition s for Dep h th a loyla tion of Resin s 2a a n d 2c
dephthaloylation efficiency (%)
line
reagents and reaction condns
2a f 3a
2c f 3c
1
2
3
4
5
6
7
10% hydrazine in ethanol, reflux, 6 h
25% hydrazine in 1,4-dioxane, 70 °C, 18 h
90
100
89
98
100
n/d
80
83
100
100
40% aqueous methylamine-CH2Cl2 (2:1), 25 °C, 72 h
40% aqueous methylamine-CH2Cl2-1,4-dioxane (2:1:1), 25 °C, 72 h
40% aqueous methylamine-1,4-dioxane (1:1), 25 °C, 72 h
40% aqueous methylamine-1,4-dioxane (1:1), 55 °C, 72 h
40% aqueous methylamine-1,4-dioxane (1:3), 55 °C, 72 h
99
100
100
100
able, has been extensively studied;11 however, no direct
details are to be found in the literature concerning the
preparation of synthon 1c. A method was described by
Worley12 for the production of the related synthon 1b,13
which involved radical-induced chlorination of the adduct
formed between thiophenol and benzyl chloride (Scheme
2, upper left). This procedure was satisfactory for the
preparation of the methyl substituted variant, 1c, start-
ing with 4-methybenzyl chloride; however, it was con-
sidered impractical for large-scale synthesis. The
alternative preparation (Scheme 2, lower left) from
tolualdehyde is novel, although a similar method14 has
been applied in an analogous, poorly characterized
synthesis.15 In contrast to the literature reports, HCl gas
was found to be an inefficient agent in the conversion of
the initially formed hemithioacetal into chloride 3. Ad-
dition of thionyl chloride drove the equilibrium involved
to completion. The product obtained, without isolation,
was then reacted with potassium phthalimide in aceto-
nitrile, a superior solvent to the DMF previously used.
Subsequent treatment of 5 with sulfuryl chloride gave
1c in good overall yield and high purity.
Initial resin modification studies focused on prepara-
tion of aminomethyl polystyrene according to the meth-
ods of Mitchell,8 but the transformations proceeded
inefficiently and gave poorly swelling products. The
variation of Zikos and Ferderigos,10 which used 1a with
ferric chloride as the Lewis acid, followed by dephtha-
loylation with hydrazine in refluxing EtOH, was a
considerable improvement, although dephthaloylation
was incomplete (Table 1, line 1). Substitution of 1,4-
dioxane for EtOH in combination with increased hydra-
zine excess provided quantitative removal (Table 1, line
2); however, removal of the highly insoluble byproduct
bis(hydrazide) was still nearly impossible. The relatively
mild methylamine-based reagent and conditions of Sko-
rna et al.16 (Table 1, line 3) formed two phases but was
comparable in efficiency to hydrazine in EtOH and
eliminated problems with byproduct removal. Further
refinement resulted from the addition of 1,4-dioxane to
give a single-phase reagent (Table 1, line 4), but 1,4-
dioxane-methylamine (with no CH2Cl2) was an even
superior reagent, giving excellent results at room tem-
perature (Table 1, line 5) and forming a readily separable
byproduct.
Ferric chloride proved a disappointing selection for the
analogous preparation of MBHA resin 3c, resulting in
inefficient incorporations of 1c and poorly swelling,
discolored products. Alternative Lewis acids were in-
vestigated, and titanium tetrachloride gave the best
substitution level, swelling properties, and product ap-
pearance; dichloroethane proved a better solvent for the
reaction than CH2Cl2. In this preparation of resin 3c,
as well as with resin 3a , complexes are formed between
the polystyrene products and the Lewis acid employed,
and careful washing protocols are necessary, as detailed
in the Experimental Section, to ensure their complete
elimination (as confirmed by negligible ash contents
found during elemental analyses). Dephthaloylation
conditions, again, required refinement. Hydrazine was
a more effective reagent in this series (Table 1, lines 1
and 2), whereas methylamine cocktails at ambient tem-
perature were less effective (Table 1, line 4). Although
hydrazine in 1,4-dioxane at elevated temperature worked
well, equivalent results were obtained with methylamine
at 55 °C for 3 days (Table 1, line 6). The application of
1,4-dioxane-40% aqueous methylamine (3:1) in both
series provided a more effective, safer, and more eco-
(11) (a) Sakellarios, E. J . J . Am. Chem. Soc. 1948, 70, 2822. (b)
Truchlik, S.; Macko, J .; Mojik, I.; Bytricky, L.; Dulak, K.; Paldan, M.;
Handlovsky, A. CS Patent 248 250, 1988; Chem. Abstr. 1988, 110,
75315.
(12) Worley, J . W. J . Org. Chem. 1979, 44, 1178-1180.
(13) (a) Bryan, W. H. J . Org. Chem. 1986, 51, 3371. (b) Bryan, W.
H. U.S. Patent 4 478 984, 1984.
(14) Bordwell, F. G.; Pitt, B. M. J . Am. Chem. Soc. 1955, 77, 572.
(15) Tuleen, D. L.; Markham, V. C. J . Org. Chem. 1967, 32, 204.
(16) Skorna, G.; Stemmer, R.; Ugi, I. Chem. Ber. 1978, 111, 806-
810.

3708 J . Org. Chem., Vol. 63, No. 11, 1998
Adams et al.
Sch em e 3
nomical reagent than hydrazine for dephthaloylation. A
very close correlation has been found between values for
nitrogen content determined by either a derivatization
assay (i.e., Fmoc-Nle-OH loading) or elemental analysis
for both 3a and 3c, indicating that the functionalization
produced is accessible and reactive.
had excellent swelling in CH2Cl2 led us to measure the
swelling of all our resins against a panel of six solvents
ranging in polarity from THF to MeOH (Table 2).18
The synthetic usefulness of all of these resins has been
assessed and directly compared in a variety of tests
(Tables 3 and 4). The simultaneous comparison tech-
nique19,20 is ideally suited as all reactions are performed
under identical conditions. Surprising results were found
in experiments designed to incorporate Fmoc-LinkerAm
[4-[(R,S)-R-[1-(9H-fluoren-9-yl)methoxycarbonylamino]-
2,4-dimethoxybenzyl]phenoxyacetic acid, also known as
Rink linker] to a level of 0.2 mmol/g onto resins having
widely ranging levels of amino functionalization (Table
3). These linker-substituted materials were desired to
allow synthesis comparisons at similar loading levels.
Very poor incorporation efficiencies were apparently
obtained with the high loaded supports studied (lines 7,
9, and 11), with 2 mmol/g AMPS giving no detectable
Fmoc incorporation (line 9). These findings are to be
contrasted with the excellent loading that can be achieved
when total substitution is desired (lines 2, 6, and 8).
Presumably, premature removal of the Fmoc group
occurred, after linker addition, as a result of the high
concentration of unreacted primary amino groups. This
conclusion is supported by the observation that the 2
mmol/g AMPS promoted rapid Fmoc removal from a
variety of neutral derivatives and multifunctional scaf-
folds in neat DMF solution (e.g., cleavage of N-Fmoc-1-
Three types of novel PEG graft copolymers have been
prepared using amino-functionalized resins 3a and 3c
(Scheme 3).17 The most simple, 8, was prepared by
p-nitrophenyl (Np) chloroformate-mediated exact under-
loading of monomethoxy-PEG onto the amino groups of
the 2 mmol/g AMPS resin (3a ), producing a highly stable
urethane bond. Adjustment of the molecular weight of
the PEG, and ratio to the amount of resin, allows the
production of resins of various compositions. The pro-
cedure described herein (see Experimental section) re-
producibly gives close to 60% PEG content in the
resulting copolymer with MeO-PEG2000 and an amino
substitution of 0.4 mmol/g. The chemistry involved has
been modified further to permit attachment of PEG-diols;
subsequent transformation of the pendant hydroxyl
group to an amino functionality, via Mitsunobu chemis-
try, provides 9 (Scheme 3). An alternative multibranched
variation, 10, based on a low-loaded resin, was also
prepared, using trimellitic anhydride chloride to establish
a branching point for the bis-addition of PEG-diamines
[partially protected at one terminus with a monomethoxy-
triphenylmethyl (MMT) group].5
With these materials at hand, we next sought to
characterize their properties and relative synthetic util-
ity. The qualitative observation of Zikos and Ferderigos10
that a 2 mmol/g AMPS resin, produced by their similar
approach using ethanolic hydrazine dephthaloylation,
(18) Solvation of peptide resin matrices has been studied in a wide
range of solvents, see: (a) Fields, G. B.; Fields, C. G. J . Am. Chem.
Soc. 1991, 113, 4202. (b) Cilli, E. M.; Oliveira, E.; Marchetto, R.;
Nakaie, C. R. J . Org. Chem. 1996, 61, 8992-9000.
(19) Hudson, D. J . Org. Chem. 1988, 53, 617-624.
(20) Meister, S. M.; Kent, S. B. H. In Peptides: Structure and
Function: Proceedings of the Eighth American Peptide Symposium;
Hruby, V. J ., Rich, D. H., Eds.; Pierce: Rockford, IL, 1983; pp 103-
106.
(17) The PEG-PS graft copolymers have been assigned the following
trade names: Champion I (8), Champion II (9), and Dendrogel (10).

Polystyrene Resins
J . Org. Chem., Vol. 63, No. 11, 1998 3709
Ta ble 2. Sw ellin g Ch a r a cter istics of Selected Resin s in Va r iou s Solven ts
swollen volume (mL per g dry resin)a
b
line
resin
notes
1% DVB; native
0.9 mmol/g
1 mmol/g
1.5 mmol/g
2 mmol/g
63% PEG; 0.39 mmol/g
56% PEG; 0.27 mmol/g
61% PEG; 0.20 mmol/g
PEG content unknown; 0.40 mmol/g
75% PEG; 0.20 mmol/g
THF
CHCl3
DMF
DMSO
CH3CN
methanol
1
2
3
4
5
6
7
8
9
polystyrene
3c (MBHA)c
3a (AMPS)c
3a (AMPS)c
3a (AMPS)c
8
7.9
9.0
8.4
8.6
8.2
5.9
6.3
5.8
6.4
4.0
7.8
8.8
8.5
8.6
8.5
7.9
8.1
5.9
8.8
4.5
5.0
6.2
6.5
8.1
7.6
6.5
6.1
5.5
6.8
4.0
2.7
2.6
3.9
6.0
6.9
6.4
5.5
4.8
6.5
3.7
2.6
2.5
3.4
4.0
3.5
6.0
4.9
4.4
6.5
3.7
2.2
2.3
3.0
3.0
3.2
5.0
4.3
3.7
6.0
3.5
9
10b
ArgoGel
TentaGel
10
a
b
Values given were averaged from multiple determinations (n g 3, SD e (0.2 mL). Methylene chloride gives values similar to those
from CHCl3. c Corresponding resins obtained from other sources were also evaluated and found to swell to a lesser extent in THF, CHCl3,
DMF, and DMSO.
Ta ble 3. Ad d ition of F m oc-Lin k er Am to Am in o-Su bstitu ted Resin s
amino loading
(mmol/g)
target loading
(mmol/g)
obsd loading
(mmol/g)
efficiency of
linker addition (%)
line
resin
1
2
3
4
5
6
7
8
9
ArgoGel
TentaGel
10b
8
9
3a (AMPS)
3a (AMPS)
3a (AMPS)
3a (AMPS)
12c
0.40a
0.21a
0.23
0.37
0.27
1.0
1.0
2.0
2.0
0.20
max
0.20
0.20
0.20
max
0.20
max
0.20
0.26
0.20
0.11
0.18b
0.12
0.13
0.19
0.60b
0.06
0.95b
0
55
n/a
60
65
95
n/a
30
n/a
0
10
11
2.0
0.24
0.05
92d
25
ArgoPore
0.74a
a
b
Loadings for these resins are as reported by the respective manufacturers. The observed loading was obtained after coupling with
1.1-1.5 equiv of Fmoc-LinkerAm (at which point the resin was completely ninhydrin negative). c Resin 12 was obtained from standard
AMPS resin 3a (2 mmol/g) by underloading as described in Scheme 4. This combined value reflects the accuracy of stoichiometric
d
underloading and linker addition.
Ta ble 4. Eva lu a tion of Resin P er for m a n ce in Va r iou s Syn th etic P r oced u r es
R-CT digestion
synthesis of ACP 65-74
use of Aloc/OAl protection
during peptide synthesisb
of Trp-Gly-
decapeptide test sequencea
stilbene synthesis derivatized resins
purity with
purity with
30 min
purity of
1-2 h
allyl-protected allyl removal
yieldg
(%)
purityh
(%)
amt cleaved
(%)
line
resinc
couplingsd (%) couplingsd (%)
peptidee (%)
efficiencye,f (%)
1
2
3
4
5
6
7
8
9
ArgoGel
90
85j
85j
85
85
85
90
90j
90j
90
85
85
75
90
75
70
90
95
>95i
>95
80i
40
1.5i
6.0
TentaGel
300j
>90j
10bk
8
9
95
85
>95
>95
75
50
>90
>90
1.5
6.0l
3a (underloaded to 0.06 mmol/g)m
3a (1 mmol/g)
12n
90
>95
75
55
50
>90
>90
>90
<0.5
4.5
5
ArgoPore
75
10 Macroreticular PS (300 Å)
a
b
The sequence of the ACP 65-74 decapeptide is VQAAIDYING. The pentapeptide sequence used for this test was Y(t-Bu)K-
(Aloc)E(OAl)K(Boc)G. c The Fmoc-LinkerAm-resins used for the synthesis of the ACP 65-74 decapeptide were the products described in
Table 3, except for entry 10, which was maximally loaded to 220 µmol/g; all other tests in this table were performed using maximally
d
loaded resin samples. Purities of the cleaved peptides were estimated by integration of the HPLC profiles of the crude products obtained
directly from ether precipitation of the TFA cleavage cocktails. e Purities of the cleaved peptides were estimated by integration of the
HPLC profiles of the crude products obtained directly from evaporation of the TFA-water cleavage cocktails. f Efficient allyl removal
g
was found to be very dependent on the quality of the tetrakis(triphenylphosphine)Pd(0) complex. Yields are based on the theoretical
h
weight of product calculated from the amino substitution of the starting resin. Purities were estimated by TLC of the crude products on
silica gel GF254 developed in toluene-1,4-dioxane-acetic acid (95:25:4); the trans-stilbene product obtained photodimerized quite readily
giving a lower Rf impurity, which was counted as “product” for these calculations. i The resin beads were observed to undergo agglomeration
during the evaluation procedure. j The crude product also contained other byproducts resulting from PEG cleavage by TFA. These impurities
k
are not included in HPLC or TLC estimates of purity but do contribute to the >100% mass yield in the stilbene synthesis test. The
l
formulation of resin 10b used in these comparisons was based on a 0.45 mmol/g MBHA resin. A value of 12% cleavage was observed for
m
n
a 200-400 mesh version of resin 9 under similar conditions. Entry 6 is the resin described in Table 3, line 7. Resin 12 was obtained
from standard AMPS resin 3a (2 mmol/g) by underloading as described in Scheme 4.
amino-6-hexanol by the resin had t1/2 ) 4 h). To avoid
these problems, an alternative procedure was developed
(Scheme 4) for underloading the 2 mmol/g AMPS resin,
giving resin 12. The method involved coupling a stoi-
chiometrically known mixture of Boc- and Fmoc-amino
acids and relied on the equal reactivity of the two
derivatives. A homogeneous distribution of amino sites
within beads is thus ensured, which may not necessarily
result from the alternative underloading method (i.e.,
coupling with less than 1 equiv of linker followed by

3710 J . Org. Chem., Vol. 63, No. 11, 1998
Adams et al.
Sch em e 4
capping unreacted amino groups). Resin 12 also main-
tains the excellent swelling characteristics of the parent
2 mmol/g AMPS resin (data not presented). For all linker
addition reactions to AMPS resins, the use of carbodi-
imide coupling in the presence of excess 1-hydroxyben-
zotriazole (HOBt) is recommended. Representative base-
labile, acid-labile, and photolabile linkers have been
attached to 1 and 2 mmol/g AMPS resins, with excellent
incorporations observed using very modest excesses of
reagents. For example, the substitution of 1 and 2
mmol/g AMPS with just a 10% excess of Fmoc-LinkerAm
using DIPCDI, HOBt-mediated coupling resulted in
quantitative addition and resins that gave negative
ninhydrin tests (Table 3, lines 6 and 8). These linker-
functionalized resins retain swelling profiles similar to
those of the parent resins (data not presented).
Two sets of simultaneous syntheses (Table 4) were
performed with the “notorious” ACP 65-74 decapeptide
test sequence, reportedly difficult to synthesize because
of the combined effects of interpeptide aggregation and
steric hindrance.21 Control supports, macroporous poly-
styrenes (lines 9 and 10) and a more highly loaded amino-
functionalized polystyrene (line 7), gave inferior results;
loading was clearly a critical issue. However, no signifi-
cant difference could be distinguished between any of the
PEG graft copolymers studied (lines 1 to 5), even with
brief 30 min couplings for all residues using the modestly
active (benzotriazol-1-yl-N-oxy)tris(dimethylamino)phos-
phonium hexafluorophosphate (BOP)/HOBt/N-methyl-
morpholine (NMM) procedure at suboptimal concentra-
tion. Even more remarkable, the underloaded AMPS
resin 12 (line 8), prepared by the method shown in
Scheme 4, synthesized the target in just as high yield
and efficiency (as judged by HPLC and MS analyses).
In further tests (Table 4), a pentapeptide bearing both
allyl22 [i.e., allyloxycarbonyl (Aloc) for Lys and allyl ester
(OAl) for Glu] and tert-butyl side-chain protecting groups
was synthesized with good efficiency on several of these
supports. Allyl removal was >95% complete on all resins
tested. Although incomplete under the suboptimal con-
ditions employed, the similarity in cleavage efficiencies
showed that allyl removal proceeded at closely similar
rates with all of the resins studied. Activation of the
product resin-bound, partially protected peptides with
BOP, to establish an amide bond between the deprotected
side-chain Lys and Glu functional groups, gave similar
product distributions upon cleavage. Intramolecular
cyclization between the side chains predominated, but
numerous oligomeric species were generated through
interpeptide interactions. Additionally, Trp-Gly-func-
tionalized versions were subjected to digestion with
R-chymotrypsin (R-CT), a method that has been used to
“shave” external regions of TentaGel beads.23 Detectable
cleavage occurred with the macroreticular PS resins
(Table 4, lines 9 and 10), but none could be detected with
12 (line 8), which does not swell in water. In the PEG
graft copolymer series, cleavage was less with ArgoGel
and 8 (lines 1 and 4) than with TentaGel or 9 (lines 2
and 5).
Standard phosphoramidite-mediated DNA syntheses
were performed using 8, derivatized with DMT-T-succin-
ate to a level of 50 µmol/g, comparing synthesis efficiency
to results obtained with a panel of alternative standard
support materials: controlled pore glass (CPG), macrore-
ticular polystyrene, and polymethacrylate.24 Average
coupling efficiencies and product purities in the syntheses
of test sequences were closely comparable for all these
materials.
The usefulness of the subject resins for combinatorial
chemistry was initially demonstrated with 2 mmol/g
AMPS using a multistep protocol, which included a
Horner-Emmons condensation, to form a stilbene ana-
logue (Scheme 5).25 The immobilization chemistry is a
variant of a method described by Ellman et al.,26 which
involved the off-resin production of a corresponding
substituted linker allyl ester. Noteworthy aspects of our
synthesis include the completeness of the linker addition
(13a f 14a ), subsequent bromination (giving 15a ), and
displacement reactions (to 16a ). A good yield of column-
purified stilbene 19 was obtained. Representative alter-
native supports (Table 4) were then taken through a
(21) The original publication describing synthetic difficulties with
acyl carrier protein 65-74 was: Hancock, W. S.; Prescott, D. J .;
Vagelos, P. R.; Marshall, G. R. J . Org. Chem. 1973, 38, 774-781.
Improved protocols were later described on polyacrylamide based
supports: Arshady, R.; Atherton, E.; Clive, D. L. J .; Sheppard, R. C.
J . Chem. Soc., Perkin Trans. 1 1981, 529-537. Other studies showed
that, for Boc chemistry, implementing coupling in DMF was the
primary modification necessary: Live, D. H.; Kent, S. B. H. In
Peptides: Structure and Function: Proceedings of the Eighth American
Peptide Symposium; Hruby, V. J ., Rich, D. H., Eds.; Pierce: Rockford,
IL, 1983; 65-68. This observation was later confirmed by detailed
study with Fmoc chemistry on polystyrene (see ref 19). Nevertheless,
this sequence still poses significant challenges; see, for example:
Alewood, P.; Alewood, D.; Miranda, L.; Love, S.; Meutermans, W.;
Wilson, D. Methods Enzymol. 1997, 289, 14-29 and other examples
in this volume.
(23) Va´gner, J .; Krchnˇa´k, V.; Sepetov, N. F.; Sˇtrop, P.; Lam, K. S.;
Barany, G.; Lebl, M. In Innovations and Perspectives in Solid-Phase
Synthesis: Peptides, Proteins and Nucleic Acids: Biological and
Biomedical Applications, 1994: Collected Papers, Third International
Symposium; Epton, R., Ed.; Mayflower Worldwide: Birmingham,
England, 1994; pp 347-352.
(24) Reddy, M. P.; Michael, M. A.; Farooqui, F.; Girgis, N. S.
Tetrahedron Lett. 1994, 35, 5771-5774.
(25) Williard, R.; J ammalamadaka, V.; Zava, D.; Benz, C. C.; Hunt,
C. A.; Kushner, P. J .; Scanlan, T. S. Curr. Biol. 1995, 2, 45-51.
(26) (a) Bunin, B. A.; Ellman, J . A. J . Am. Chem. Soc. 1992, 114,
10997-10998. (b) Bunin, B. A.; Plunkett, M. J .; Ellman, J . A. Proc.
Natl. Acad. Sci. U.S.A. 1994, 91, 4708-4712.
(22) Lyttle, M. H.; Hudson, D. In Peptides: Chemistry and Biology:
Proceedings of the Twelfth American Peptide Symposium; Smith, J .
A., Rivier, J . E., Eds.; ESCOM: Leiden, The Netherlands, 1992; pp
583-584.

Polystyrene Resins
J . Org. Chem., Vol. 63, No. 11, 1998 3711
Sch em e 5
similar five-step protocol, but using linker 13b with
abbreviated reaction times and no intermediate monitor-
ing, in an attempt to determine differences in their
synthetic performance. Indeed, significantly worse prod-
uct purity was observed for the ArgoGel-mediated reac-
tion (Table 4, line 1). In addition, a large proportion of
PEG was cleaved from the TentaGel resin (Table 4, line
2). Of the unmodified 1% cross-linked polystyrenes, the
underloaded 2 mmol/g AMPS was marginally superior
to the fully loaded version. Both of the macroreticular
PS resins (Table 4, lines 9 and 10) performed very well,
but with lower overall yields.
mediated “onium”-type coupling reactions are employed,
is not generally recommended; premature Fmoc removal,
as previously discussed,28 will lead to numerous byprod-
ucts. Because benzylamine itself is not known to be an
effective reagent, pendant amino groups appear to be able
to act in consort; clearly “pseudo-dilution” or effective site-
isolation effects are not operable. Carbodiimide-mediated
coupling in the presence of excess HOBt is strongly
recommended for linker addition and chain assembly on
these highly substituted resins.
The syntheses of peptides and DNA by solid-phase
methodology is well-known to be very dependent on the
nature of the support. Recent reports describe similar
variations for other chemical transformations.29 Many
practitioners of the art of performing organic transforma-
tions on resin-bound substrates have elected to use PEG-
PS graft copolymers and describe consequent improve-
ments in synthesis efficiency.6,30 In addition to bearing
what have been thought to be beneficial spacer arms,
they are pressure stable, can advantageously be used in
column reactors, and have been proven to be generally
useful. Such graft copolymers can be classified into two
types: those that incorporate PEG by in situ polymeri-
zation of ethylene oxide onto hydroxyl groups on the
polystyrene backbone, such as TentaGel,31 and the re-
cently introduced branched variant ArgoGel,32 and those
that incorporate preformed, appropriately protected PEGs
onto amino-modified resins, such as with PEG-PS.33,34
Remarkably, the simple PEG-derivatized resin, 8 (Table
2, line 6) has a similar swelling profile to that of the more
complex ArgoGel (Table 2, line 9). Deductions made from
the swelling properties of products from different base
resins are complicated by many considerations, for
example, variation in particle size and particle size
Discu ssion
Convenient, practical, and economical large-scale meth-
ods for the preparation of AMPS and MBHA resins are
described. By careful optimization of resin modification
chemistry, combining mild conditions with efficient con-
versions, products endowed with excellent swelling char-
acteristics have been obtained. Dramatic changes in the
swelling profiles were obvious with progressively higher
loaded AMPS; swelling reached a plateau level at a
substitution of 1.5 mmol/g for THF, CHCl3, DMF, and
DMSO (Table 2, line 4), with no significant improvement
found for the 2 mmol/g loaded resin (Table 2, line 5),
which represents a practical ceiling for the use of this
resin and is a substitution level beyond which incorpora-
tions became progressively less efficient. DMSO, a
solvent in which the parent PS swells negligibly (Table
2, line 1),27 is particularly useful for promoting nucleo-
philic substitution reactions, and the improved swelling
is expected to be beneficial. Similar swelling profiles are
retained in acetylated, protected amino acid and handle
substituted forms, suggesting that the phenomenon does
not simply result from the presence of polar free amino
groups and that good swelling should be maintained
during small molecule assembly. This conclusion was
substantiated by a stilbene synthesis involving the Hor-
ner-Emmons condensation reaction (Scheme 5). How-
ever, a note of caution regarding unexpected interresidue
side reactions is raised by our observations of premature
Fmoc removal mediated by pendant amino groups on the
1 and 2 mmol/g AMPS. Their use for high-load peptide
synthesis by Fmoc chemistry, particularly when base-
(28) (a) Bodanszky, M.; Deshmane, S. S.; Martinez, J . J . Org. Chem.
1979, 44, 1622. (b) Bodanszky, A.; Bodanszky, M.; Chandramouli, N.;
Kwei, J . Z.; Martinez, J .; Tolle, J . C. J . Org. Chem. 1980, 45, 72.
(29) Burgess, K.; Lim, D. J . Chem Soc., Chem. Commun. 1997, 785-
786; also see ref 38.
(30) Review: Meldal, M. Methods Enzymol. 1997, 289, 83-104.
(31) (a) Bayer, E.; Hemmasi, B.; Albert, K.; Rapp, W.; Dengler, M.
In Peptides: Structure and Function: Proceedings of the Eighth
American Peptide Symposium; Hruby, V. J ., Rich, D. H., Eds.; Pierce:
Rockford, IL, 1983; pp 87-90. (b) Rapp, W.; Zhang, L.; Ha¨bich, R.;
Bayer, E. In Peptides 1988: Proceedings of the Twentieth European
Peptide Symposium; J ung, G., Bayer, E., Eds.; de Gruyter: Berlin,
Germany, 1989; pp 199-201.
(32) Gooding, O.; Hoeprich, P. D. J .; Labadie, J . W.; Porco, J . A. J .;
van Eikeren, P.; Wright, P. In Molecular Diversity and Combinatorial
Chemistry Libraries and Drug Discovery; Chaiken, I. M., J anda, K.
D., Eds.; American Chemical Society: Washington, DC, 1996; p 199.
(33) Zalipsky, S.; Chang, J . L.; Albericio, F.; Barany, G. React. Polym.
1994, 22, 243 and references therein.
(27) Reports have described improvement in peptide synthesis using
DMSO on a highly loaded benzhydrylamine resin, even though the
starting resin did not swell in this solvent; see: Marchetto, R.; de
Oliveira, E.; Paiva, A. C. M.; Nakaie, C. R. In Peptides 1990:
Proceedings of the Twenty-First European Peptide Symposium; Giralt,
E., Andreu, D., Eds.; ESCOM: Leiden, The Netherlands, 1990; pp 122-
124.
(34) Ho, D. D.; Neumann, A. V.; Perelson, A. S.; Chen, W.; Leonard,
J . M.; Markowitz, M. Nature 1995, 373, 123.

3712 J . Org. Chem., Vol. 63, No. 11, 1998
Adams et al.
distribution may confer different packing characteristics.
Clearly, however, all three novel PEG graft copolymers
possessed excellent swelling characteristics.
larity in product profiles from the side-chain cross-linking
experiment on the partially protected peptide products
resulting from allyl removal provided further confirma-
tion of the comparable efficiency of the resins, and the
extent of interpeptide reactions provided no evidence for
site isolation.
Study of the relative synthetic efficiency of the two
closely related resins, 8 and 9, for peptide synthesis was
expected to provide an answer to the question of whether
the PEG in such graft copolymers serves principally to
modify the environment35 or whether the spacer arm
effect is additionally beneficial. Our results (Table 4)
found no significant difference between 8 and 9, which
bear backbone and PEG terminal amino groups, respec-
tively. These findings are in accord with qualitative
studies by Barany et al.,6 who obtained excellent syn-
thetic results with a branched PEG-PS resin prepared
as follows: (i) coupling Fmoc-Orn(Boc)-OH to 1 mmol/g
AMPS; (ii) removal of the Fmoc group; (iii) attachment
of a monomethoxy-PEG acid derivative; and (iv) peptide
synthesis on the Orn side-chain following Boc deprotec-
tion. With resin 8, synthesis is performed on amino-
methyl groups distant from the site of PEGylation, not
on amino groups proximal to a branch point. This
formulation is simple and reproducible, and the PEG-
urethane attachment linkage is highly stable to aggres-
sive reagents (even if some leaching were to occur no loss
of product would result).
The novel graft copolymer 8 gave excellent results in
the phosphoramidite-mediated synthesis of a DNA target
sequence. The application of this and other PEG-PS graft
copolymers is truly beneficial for oligonucleotide as-
sembly,37 as compared to unmodified 1% cross-linked
polystyrene, because rapid reactions are required in a
range of solvents (e.g., acetonitrile, THF-water, and
CH2Cl2). Any reaction requiring aqueous reagents will
also be facilitated on PEG supports, as evidenced by a
recent example involving a Suzuki cross-coupling reaction
of an immobilized arylboronic acid with a variety of
substituted aryl iodides.38 Last, the ability of R-chymo-
trypsin to partially cleave a resin-immobilized peptide
indicates PEG-PS copolymers are superior for bead-based
assays because these resins allow reasonably efficient
interactions between displayed ligands and their biologi-
cal targets; in this application, PEG spacer arms do have
a significant effect.
Furthermore, the studies provide a valid comparison
of both 8 and 9 with a similarly loaded amino-function-
alized polystyrene, 12, which is derived from the same
parent AMPS resin but lacks PEG derivatization. This
resin performed equivalently in all but the enzymatic
hydrolysis comparisons. We conclude that for the chemi-
cally mediated transformations studied PEG does not
play an active role in promoting synthesis efficiency,
neither by its influence on the microenvironment nor by
its spacer arm effect. Appropriate loading, the mainte-
nance of good swelling characteristics, and the absence
of extraneous cross-linking or other undesired functional-
ity on the resin are the principal factors responsible for
good synthesis efficiency. Another important influence
is the poorly understood phenomenon of bead aggrega-
tion, which results in poor reaction efficiency and inef-
ficient removal of reactants during washing. This un-
desirable effect was only significant for one of the
supports studied (Table 4, line 1). Instability of the PEG
attachment during synthesis and cleavage is also highly
undesirable and was observed in two cases (Table 4, lines
2 and 3). Regarding peptide chain assembly, clearly the
ACP 65-74 test decapeptide is not, intrinsically, difficult
to synthesize. Because little difference was observed in
product purity on the optimized supports with the
different coupling times studied, the half-lives of the
coupling reactions must, even for the sterically hindered
residues incorporated, be very short indeed. Diminished
coupling rates, resulting from steric effects or peptide
aggregation mediated by hydrogen bonding or hydropho-
bic effects, were obvious at high loading with 3a (Table
4, line 7) and the structurally dissimilar macroreticular
resins (Table 4, lines 9 and 10). In addition, our results
failed to confirm literature reports that allyl removal by
Pd(0)-mediated transfer is facilitated on PEG-substituted
polystyrenes over unmodified polystyrenes.22,36 The simi-
Con clu sion s
Careful initial derivatization and subsequent modifica-
tion provided improved resins possessing excellent swell-
ing properties and performance characteristics. All PEG-
derivatized polystyrenes tested performed similarly. Our
results provide experimental confirmation of the well-
established belief that resins that swell better provide
better synthetic reactivity. In all probability, reported
synthesis facilitation observed with PEG-polystyrenes
results, simply, from the maintenance of good swelling
during all stages of synthetic processes, rather than from
spacer arm or environment effects. Of the novel PEG
graft copolymers described, the simplest variant, 8,
proved both stable39 and the most generally useful,40
whereas the multibranched PEG-PS resins, ArgoGel and
the novel dendromeric 10, although for the most part
excellent, had no obvious advantages.
Exp er im en ta l Section
Gen er a l Meth od s. Reagents were supplied by Aldrich
(Milwaukee, WI); all resins and linkers were supplied by Solid
Phase Sciences (Novato, CA), except for ArgoGel and ArgoPore
from Argonaut Technologies (San Carlos, CA), and TentaGel
from Rapp Polymere (Tubingen, Germany); BOP and Fmoc-
amino acids were supplied by Chem Impex (Wood Dale, IL).
NMR spectra were provided by Acorn NMR (Fremont, CA),
and MS analyses were performed on a Voyager MALDI-TOF
spectrometer at the University of Michigan (Ann Arbor,
(36) Aldrich, J . V.; Leelasvatanakij, L.; Maeda, D. Y. In Peptides:
Chemistry, Structure and Biology: Proceedings of the Fourteenth
American Peptide Symposium; Kaumaya, J . T. P., Hodges, R. S., Eds.;
Mayflower Scientific: Kingswinford, U.K., 1996; pp 36-38.
(37) Resin 10b has proved an efficient support for the assembly of
a variety of complex peptide-DNA hybrids: Songster, M. F. Solid
Phase Sciences, unpublished results.
(38) Ruhland, B.; Bombrun, A.; M. Gallop, M. A. J . Org. Chem. 1997,
62, 7820-7826.
(39) Negligible PEG loss was observed with either 2 h TFA treat-
ment or concentrated NH3 at 55 °C for 6 h.
(35) Barany, G.; Sole, N. A.; van Abel, R. J .; Albericio, F.; Selsted,
M. E. In Innovations and Perspectives in Solid-Phase Synthesis:
Peptides, Polypeptides and Oligonucleotides, 1992: Collected Papers,
Second International Symposium; Epton, R., Ed.; Intercept: Andover,
England, 1992; pp 29-38.
(40) In addition to the detailed experiments presented, resin 8 has
been used to synthesize many complex biomolecules, including a 68-
residue peptide, and a variety of unmodified and fluorescently labeled
DNA sequences.

Polystyrene Resins
J . Org. Chem., Vol. 63, No. 11, 1998 3713
MI). Analytical HPLC analyses were performed as reported
previously,19 but using a Waters 996 photodiode array detector,
with data processed at 220 nm using the Waters Millennium
system (integration values are quoted in increments of 5% and
are significant to (2.5%). Peptide syntheses were performed
as described previously,19 except 2 or 10 mL “syringe” reac-
tors41 were used in place of macro DNA synthesis columns.
Elemental analyses were provided by Desert Analytics (Tus-
con, AZ). Amino substitution levels were also determined by
a derivatization procedure in which excess Fmoc-Nle-OH was
coupled to the resin in a 2 mL syringe reactor for 2 h. The
resin was thoroughly washed and dried, and then aliquots (ca.
5-10 mg) were accurately weighed into glass scintillation
vials, treated with 30% piperidine in DMF (0.5 mL) for 10 min
with vortex mixing, and diluted with MeOH (20 mL), and the
absorption was determined at 301 nm. Quantitation was
achieved by comparison with known standards, providing a
calibration curve to correct for the weight increase resulting
in derivatization and thereby giving original amino content.
All amino acids used were of the L configuration; all solvent
ratios are volume/volume.
N-(r-Ch lor o-4-m eth ylben zyl)p h th a lim id e (1c). Cau-
tionary note: the following procedure should be performed in
a fume hood because of the toxic and irritant nature of the
starting materials and byproducts. Phthalimido derivative 7
(660 g, 1.8 mol) was stirred in CH2Cl2 (3 L) under argon, and
sulfuryl chloride (150 mL, 250 g, 1.9 mol) was added dropwise
over 3 h, turning the mixture bright red. The reaction was
stirred overnight and then filtered and evaporated to about 1
L under reduced pressure. Petroleum ether (2 L) was added,
and the mixture was chilled at -15 °C overnight. A volumi-
nous solid was collected in a 3 L sintered glass funnel, washed
with cold petroleum ether (2 L), and dried to a constant weight
under high vacuum to give 395 g (85%) of a light yellow solid:
mp 110-113 °C; 1H NMR (CDCl3) δ 7.9 (dd, 2H), 7.75 (dd, 2H),
7.6 (d, 2H), 7.23 (s, 1H), 7.2 (d, 2H), 2.4 (s, 3H). Anal. Calcd
for C16H12NO2Cl: C, 67.26; H, 4.23; N, 4.90. Found: C, 67.40;
H, 4.24; N, 5.17.
(Am in om eth yl)p olystyr en e (AMP S, 3a ). Different load-
ings, up to 2 mmol/g, are obtained by scaling the amount of
(chloromethyl)phthalimide used; the following procedure, to
obtain 1 mmol/g, is typical. Underivatized 1% cross-linked
100-200 mesh polystyrene (1.6 Kg) was placed in a 22 L
reactor equipped with argon inlet, addition funnel, mechanical
stirrer, reflux condenser, and heating mantle. Anhydrous
CH2Cl2 (14 L), (chloromethyl)phthalimide (1a , 340 g, 1.8 mol),
and anhydrous ferric chloride (68 g) were added, and the
suspension was stirred under argon at reflux for 7 h, at which
time all HCl evolution had ceased. The reaction was cooled
overnight and the suspension filtered, and the beads were
washed successively with CH2Cl2 (4 L), 1,4-dioxane (6 L), 1,4-
dioxane-10% HCl (4:1, 6 L), 1,4-dioxane-water (4:1, 6 L), and
MeOH (6 L). The resin was air-dried (giving 2.0 kg) and then
returned to the reactor, resuspended in 1,4-dioxane (14 L), and
stirred until fully swollen. Aqueous methylamine (40%, 4.5
L) was added and the reaction stirred at 25 °C for 3 days. The
suspension was filtered, and the beads were washed with 1,4-
dioxane-water (4:1, 6 L), 1,4-dioxane (6 L), DMF (6 L), CH2Cl2
(12 L), and MeOH (20 L). The resin was again air-dried, then
passed through sieves to remove particles outside the 100-
200 mesh range, and finally dried to constant weight in vacuo
to provide 1600 g of AMPS resin: Elemental analysis found
1.5% N (corresponding to a loading of 1.07 mmol/g) and <0.05%
Cl; Fmoc loading assay gave 1.00 mmol/g.
were added, and the suspension was stirred under argon at
reflux for 7 h, at which time all HCl evolution had ceased. The
reaction was worked up analogously to the method described
above for 3a , except that dephthaloylation was performed with
1,4-dioxane-40% aqueous methylamine (3:1) at 55 °C for 3
days to give 44 g of product: Elemental analysis found 1.2%
N (corresponding to a loading of 0.84 mmol/g) and 0.10% Cl;
Fmoc loading assay gave 0.92 mmol/g. Higher loading levels
can be achieved with a greater excess of 1c but the reaction
becomes progressively more inefficient.
N-[r-(P h en ylt h io)-4-m et h ylb en zyl]p h t h a lim id e (7).
Cautionary note: the following procedure should be performed
in a fume hood because of the toxic and irritant nature of the
starting materials and byproducts. Thiophenol (420 mL, 460
g, 4.2 mol) was dissolved in Et2O (3 L) in a 5 L three-neck
flask equipped with an HCl gas inlet bubbler, 500 mL dropping
funnel, and argon inlet. The stirred solution was cooled in a
dry ice-water bath for 1 h under argon, and HCl was bubbled
in for 30 min before tolualdehyde (480 mL, 490 g, 4.1 mol)
was finally added dropwise over 3 h; slow HCl addition was
continued during aldehyde addition and for 2 h afterward. The
reaction was stirred overnight and allowed to warm to 25 °C.
Next, thionyl chloride (300 mL, 490 g, 4.1 mol) was added
dropwise over 3 h and the solution stirred for 24 h, during
which time the argon was turned off to allow the evolved gas
(SO2) to escape through the argon bubbler. The dark solution
was poured cautiously into a 5 L round-bottom flask and the
Et2O removed by vacuum transfer (in the hood). The resulting
green oil, which contained intermediate 5, was dried under
high vacuum (with a NaOH trap) for 1 h. Acetonitrile (3 L)
and potassium phthalimide (760 g, 4.1 mol) were then added,
accompanied by evolution of heat and a color change from dark
brown to pale yellow; after natural heat evolution subsided,
the solution was heated to reflux overnight. The brown
solution was cooled, most of the acetonitrile removed under
reduced pressure, and the resulting residue taken up in ethyl
acetate (4 L) and washed with water (6 L). The aqueous phase
was back-extracted with CH2Cl2 (2 L), and the two organic
phases were washed separately with brine (1 L), dried over
Na2SO4, and concentrated to ca. 1 L. Petroleum ether (1.4 L)
was then added to each flask, and they were chilled to -15 °C
for 3 days. The crystalline products were isolated by filtration,
washed with cold petroleum ether, and combined to give a tan
solid that was dried to a constant weight (660 g, 45% yield).
An analytical sample of the product was recrystallized from
ethyl acetate and petroleum ether: mp 119-121 °C; 1H NMR
(CDCl3) δ 7.76 (dd, 2H), 7.7 (d, 2H), 7.64 (dd, 2H), 7.5-7.4 (m,
2H), 7.2 (m, 3H), 7.16 (d, 2H), 6.72 (s, 1H), 2.3 (s, 3H). Anal.
Calcd for C22H17NO2S: C, 73.51; H, 4.77; N, 3.90. Found: C,
72.95; H, 4.87; N, 4.39.
P oly(eth ylen e glycol)-gr a ft-p olystyr en e (8). Poly(eth-
ylene glycol) methyl ether (400 g, Mn ca. 2000, 200 mmol) was
dried overnight in vacuo, dissolved in warm anhydrous pyri-
dine (2.3 L), and evaporated to a viscous state. Fresh pyridine
(2 L) was added, the solution evaporated to ca. 1 L, and
p-nitrophenyl chloroformate (40 g, 190 mmol) dissolved in
CH2Cl2 (150 mL) was added dropwise to the stirred solution
under argon. The reaction was stirred until all the initially
formed white precipitate had dissolved (4 h) and for an
additional 2 h afterward, whereupon the mixture was concen-
trated in vacuo to ca. 700 mL. Resin 3a (200 g, 2 mmol/g, 400
mmol) was mechanically stirred in DMF-CH2Cl2 (2:1, 2.1 L)
with HOBt (30 g, 220 mol) until fully swollen, and then the
activated poly(ethylene glycol) solution was poured in as a
constant stream with vigorous stirring. After being stirred
overnight, the resin had swollen further, adsorbing all solvent,
so CH2Cl2 was added to a total reaction volume of 4 L and
stirring continued for another 24 h. The suspension was
filtered, and the resin beads were washed successively with
DMF (3×), CH2Cl2 (3×), MeOH (3×), Et2O (2×), and petroleum
ether (2×) and then dried in vacuo to constant weight yielding
535 g (corresponding to 62% PEG content) of resin 8: Final
loading was 0.39 mmol/g as determined by Fmoc loading assay.
P oly(eth ylen e glycol)-gr a ft-p olystyr en e (9). Poly(eth-
ylene glycol) (50 g, Mn ca. 2000, 25 mmol) was dried in vacuo
4-Meth ylben zh yd r yla m in e P olystyr en e (MBHA, 3c).
Underivatized 1% cross-linked 100-200 mesh polystyrene (40
g) was placed in a 1 L reactor equipped with argon inlet,
addition funnel, mechanical stirrer, reflux condenser, and
heating mantle. Anhydrous dichloroethane (400 mL), phthal-
imido derivative 1c (16 g, 56 mmol), and titanium tetrachloride
(1.6 mL, care must be observed when handling this reagent)
(41) Krchnˇa´k, V.; Weichsel, A. S.; Cabel, D.; Flegelova´, Z.; Lebl, M.
Mol. Diversity 1996, 1, 149.

3714 J . Org. Chem., Vol. 63, No. 11, 1998
Adams et al.
and dissolved in CH2Cl2-pyridine (1:1, 200 mL), and the
solution was evaporated. Fresh pyridine (300 mL) was added
to the viscous mixture and the concentration repeated. p-
Nitrophenyl chloroformate (2.5 g, 12 mmol) dissolved in CH2Cl2
(13 mL) was then added dropwise over 15 min to the vigorously
stirred solution. A suspension of resin 3a (6.25 g, 2 mmol/g,
12.5 mmol) and anhydrous HOBt (2.0 g, 2.7 mmol) in DMF-
CH2Cl2 (1:1, 100 mL) was added to the activated PEG mixture,
and the reaction was stirred overnight. The resin beads were
isolated by filtration, washed with DMF (3×), capped with a
solution of acetic anhydride (0.3 M) and HOBt (0.3 M) in DMF
(200 mL) for 1 h, and washed with DMF (3×), at which point
the resin was ninhydrin negative. The resin was then
resuspended in DMF (200 mL) and treated with hydrazine
hydrate (20 mL), and the mixture was vortexed gently for 2
h. The beads were filtered, washed successively with DMF
(3×) and THF (3×), transferred to a round-bottom flask
equipped with a pressure-equalizing dropping funnel, and
stirred in a minimum volume of THF. Triphenylphosphine
(9.0 g, 34 mmol) and phthalimide (4.5 g, 31 mmol) were added,
followed by the dropwise addition of DEAD (5 mL, 32 mmol).
The suspension was stirred for 2 days, and then the beads were
filtered and washed successively with DMF (3×) and 1,4-
dioxane (3×). The intensely yellow beads were resuspended
in 1,4-dioxane-40% aqueous methylamine (3:1, 200 mL) and
stirred for 2 days. After filtration and washing with DMF
(3×), CH2Cl2 (3×), and MeOH (3×), the product was dried to
constant weight to yield 14 g (corresponding to 56% PEG
content) of resin 9 as pale yellow beads: Final loading was
0.27 mmol/g as determined by Fmoc loading assay.
P oly(eth ylen e glycol)-gr a ft-p olystyr en e (10b). Poly-
(ethylene glycol) bis(2-aminopropyl) ether (400 g, Mr ca. 2000,
200 mmol) was dissolved in pyridine (1.5 L) and evaporated
to about one-half volume. The solution was vigorously stirred
and MMT-Cl (64 g, 200 mmol) in CH2Cl2 (500 mL) added
dropwise over 1 h from a pressure-equalizing dropping funnel.
Triethylamine (29 mL, 210 mmol) was added to the now dark
red solution, and stirring was continued overnight. The
reaction was evaporated to dryness and the residue dissolved
in ethyl acetate (500 mL) and refrigerated overnight. Pre-
cipitated triethylamine hydrochloride was removed by filtra-
tion and washed with cold CH2Cl2 (500 mL). The combined
filtrates were evaporated and redissolved in CH2Cl2-DMF (1:
1, 500 mL) containing N-methylimidazole (20 mL). MBHA
resin (3c, 44.5 g, 0.45 mmol/g) was suspended in CH2Cl2-DMF
(1:1, 500 mL), diisopropylethylamine (DIEA, 10 mL, 57 mmol)
and pyridine (20 mL) were added, and the stirred mixture was
placed in an ice bath. Trimellitic anhydride chloride (73 g,
350 mmol) dissolved in CH2Cl2 (100 mL) was added and the
suspension stirred at 25 °C for 2 h. The beads were then
isolated by filtration and washed with CH2Cl2-DMF (1:1, 3×).
One-half of the mono-MMT-protected poly(ethylene glycol)
solution obtained above was added to the beads with thorough
mixing. The reaction was transferred to a 2 L glass screw-
cap bottle, which was tightly sealed and shaken overnight.
BOP (32 g, 71 mmol) and DIEA (15 mL, 86 mmol) were then
added, and the reaction was shaken for another 24 h. The
beads were isolated by filtration and thoroughly washed with
DMF (3×) and CH2Cl2 (3×). The beads were repeatedly
subjected to 5 min treatments with TFA-dimethyl phosphite-
CH2Cl2 (5:5:90) until a colorless filtrate was obtained (1 h) and
then washed with CH2Cl2 (3×), triethylamine-DMF (1:9, 2×),
and DMF (3×). The resin was then retreated with trimellitic
anhydride chloride, followed by the remaining one-half of the
mono-MMT-protected poly(ethylene glycol) solution, exactly as
described above, including MMT deprotection and washing
with triethylamine. Last, the resin was washed with CH2Cl2
(3×) and MeOH (3×) and then dried in vacuo to constant
weight, giving 113 g (corresponding to 61% PEG content) of
dendromeric resin 10b as off-white beads: Final loading was
0.20 mmol/g as determined by Fmoc loading assay.
CH2Cl2-DMF (1:1, 50 mL) and the solution mixed in a
stoppered flask for 10 min. The activated amino acid mixture
was then added to a suspension of resin 3a (5 g, 2 mmol/g) in
CH2Cl2-DMF (1:1) and shaken for 4 h. The resin beads were
isolated by filtration and washed with DMF (3×) and CH2Cl2
(3×). The resin was resuspended in CH2Cl2, TFA was added
(to 40% v/v), and the mixture was shaken for 30 min and then
washed with CH2Cl2 (3×), DMF (3×), triethylamine-DMF (1:
9, 2 × 5 min), and DMF (3×). The Boc-deprotected resin was
capped with a solution of acetic anhydride (0.5 M) and HOBt
(0.5 M) in DMF (100 mL) for 1 h, washed with DMF (3×),
CH2Cl2 (3×), and MeOH (3×), and dried to constant weight to
give 6.2 g of white beads.
Der iva tiza tion of 12 w ith F m oc-Lin k er Am . Approxi-
mately half of this material (3 g) was treated with 30%
piperidine in DMF, shaken for 15 min, and washed with DMF
(5×). Fmoc-LinkerAm (0.86 g, 1.6 mmol), BOP (0.74 g, 1.6
mmol), and HOBt (0.2 g, 1.6 mmol) in DMF (10 mL) were
treated with NMM (0.41 mL, 3.7 mmol) and added to the resin
suspension. The mixture was shaken for 4 h, and then the
beads were isolated by filtration, washed with DMF (3×),
CH2Cl2 (3×), and MeOH (3×), and dried to constant weight to
give 3.4 g of white beads: Final loading was 0.24 mmol/g as
determined by Fmoc analysis.
Der iva t iza t ion a n d E va lu a t ion of R esin 8 for DNA
Syn th esis. 5′-DMT-thymidine-3-succinate (0.37 g, 0.5 mmol),
BOP (0.27 g, 0.6 mmol), and HOBt (75 mg, 0.55 mmol) in DMF
(5 mL) were treated with NMM (0.14 mL, 1.3 mmol) and gently
mixed by hand for 5 min. The activated solution was added
to resin 8 (5 g, 0.39 mmol/g), preswollen in DMF, and the
reaction was shaken for 6 h. The beads were isolated by
filtration, washed with DMF (3×) and THF (3×), resuspended
in acetic anhydride-pyridine-N-methylimidazole-THF (5:5:
10:80), and mixed for 1 h. The beads were then washed with
DMF (3×), CH2Cl2 (3×), and MeOH (3×), and dried in vacuo
to give 5.1 g of a ninhydrin-negative resin: Final loading was
96 µmol/g as determined by DMT analysis. Equally efficient
incorporation was observed using twice the amount of 8 to
provide a resin (50 µmol/g final loading) that was used for the
comparative DNA synthesis evaluations. These experiments
were performed on a Biosearch 8750 synthesizer using col-
umns packed with 200 nmol of either resin 8 (derivatized as
described here), polymethacrylate, CPG, or 1000 Å macrore-
ticular polystyrene. Syntheses were performed in duplicate,
and product yield and purity were determined by anion
exchange HPLC using a Dionex DNAPac PA-1000 (4 mm ×
25 cm) column.
3-Hyd r oxy-4′-n itr o-tr a n s-stilben e (19). P r ep a r a tion of
(Hyd r oxym eth yl)p h en oxya cetic Acid (HMP A)-Der iva -
tized AMP S Resin (14a ). 4-(Hydroxymethyl)phenoxyacetic
acid (HMPA, 13a , 0.73 g, 4.0 mmol) and HOBt (0.81 g, 4.4
mmol) in CH2Cl2-DMF (1:1, 20 mL) were treated with DIPCDI
(0.7 mL, 4.4 mmol), stirred for 2 min, and added to AMPS resin
3a (1.95 mmol/g, 1 g) preswollen in a minimum amount of
CH2Cl2-DMF (1:1). This suspension was shaken overnight,
and then the resin was isolated by filtration, washed with DMF
(3×), CH2Cl2 (3×), and MeOH (3×), and dried to give 1.5 g of
ninhydrin negative HMPA-AMPS resin.
P r ep a r a tion of (Br om om eth yl)p h en oxya cetyl-AMP S
Resin (15a ). Triphenylphosphine (0.31 g, 1.2 mmol) and
carbon tetrabromide (0.39 g, 1.2 mmol) were added to a
suspension of HMPA-AMPS resin 14a (0.5 g, 0.65 mmol) in
CH2Cl2 (10 mL) cooled to 0 °C. The resulting suspension was
shaken overnight and filtered, and the beads were washed
with CH2Cl2 (4×) and acetonitrile (2×) and dried in vacuo to
give 0.55 g of resin 15a : Final loading was 1.15 mmol/g as
determined by Br analysis.
P r ep a r a tion of 3-F or m ylp h en ol Su bstitu ted Resin
In ter m ed ia te 16a . 3-Hydroxybenzaldehyde (0.37 g, 2.9
mmol) and potassium bis(trimethylsilyl)amide (0.58 g, 2.9
mmol) were shaken in DMF (5 mL) for 5 min, resin 15a (0.5
g, 0.57 mmol) was then added, and the suspension was shaken
overnight. The resin beads were isolated by filtration, washed
with DMF (3×), CH2Cl2 (3×), and MeOH (3×), and dried in
vacuo to give 0.5 g of resin-bound aldehyde 16a : residual Br
Un d er loa d in g of (Am in om eth yl)p olystyr en e Resin s.
P r ep a r a tion of Resin 12. DIPCDI (3.1 mL, 20 mmol) was
added to a solution of Fmoc-alanine (0.78 g, 2.5 mmol), Boc-
alanine (2.8 g, 15 mmol), and HOBt (3.1 g, 23 mmol) in

Polystyrene Resins
J . Org. Chem., Vol. 63, No. 11, 1998 3715
content <0.05 mmol/g as determined by Br analysis. Viability
of immobilization and cleavage was confirmed when a portion
of this resin (50 mg) was treated with TFA-water (19:1) for 2
h and the filtrate collected and concentrated to give 6 mg
(∼100%) of a product that was identical to the starting
3-hydroxybenzaldehyde.
ficiency was determined by analysis of Fmoc as described
above (see Table 3 for detailed results).
P r oced u r e for Ma xim a lly Loa d in g Resin s (Descr ibed
Her e for AMP S 3a ). Fmoc-LinkerAm (430 g, 0.80 mol) and
HOBt (150 g, 1.1 mol) were dissolved in CH2Cl2-DMF (1:1,
2.25 L), and DIPCDI (140 mL, 0.90 mol) was slowly added with
stirring over 10 min. The activated linker solution was rapidly
added to resin 3a [700 g, 0.95 mmol/g, 0.66 mol, preswollen in
CH2Cl2-DMF (1:1, 4 L) for 30 min] and washed in with a small
amount of additional solvent. The reaction was stirred
overnight, and then the beads were isolated by filtration and
washed with DMF (4 L), 1,4-dioxane (4 L), CH2Cl2 (8 L), and
MeOH (4 × 4 L). The ninhydrin-negative resin was dried in
vacuo to constant weight to yield 1.0 kg of Fmoc-
LinkerAm-AMPS: Final loading was 0.63 mmol/g as deter-
mined by Fmoc analysis. Samples of the 2 mmol/g AMPS 3a
and TentaGel were also loaded by analogous procedures (see
Table 3 for detailed results).
Syn th esis of ACP 65-74 Deca p ep tid e, Va l-Gln -Ala -Ala -
Ile-Asp -Tyr -Ile-Asn -Gly (Ta ble 4). Peptide syntheses were
performed in syringe reactors with 40 mg of the indicated
Fmoc-LinkerAm-resin, underloaded as described above for
Table 3 (except for the macroreticular polystyrene, which was
maximally loaded). Side-chain protection for Asp and Tyr was
provided by t-Bu; Asn and Gln side chains were protected with
Trt. The basic synthesis cycle involved an initial DMF wash
(2×), followed by Fmoc cleavage with piperidine-DMF (3:7, 1
min + 10 min), and then washing with DMF (5×), and coupling
with the appropriate activated amino acid solution for the
times indicated in Table 4. Fresh stock solutions of BOP/
HOBt/NMM activated Fmoc-amino acids were prepared as
follows: 2 mmol each of Fmoc-amino acid, BOP, and HOBt
were activated with NMM (0.3 M in DMF, 15 mL), and the
final volume was adjusted to 20 mL to give 0.1 M solutions.
After final Fmoc removal, the resins were washed with CH2Cl2
(3×), and MeOH (1×) and dried. The resins were cleaved in
the syringe reactors with TFA-triisopropylsilane-MeOH (38:
1:1) for 2 h; the cleavage solutions were collected in glass
scintillation vials and concentrated under nitrogen to a small
volume. Anhydrous Et2O was added, and the mixtures were
refrigerated for 2 h. The peptides were isolated by centrifuga-
tion, washed with Et2O (3×), and dried. The samples were
then analyzed by analytical HPLC and MS; where baseline
HPLC resolutions were not observed, forced droplines were
used from slope inflection points or valleys, a method that may
overestimate levels of impurities (see Table 4 for complete
results).
Eva lu a tion of Allyl P r otection d u r in g P ep tid e Syn -
th esis (Ta ble 4). P r ep a r a tion , Dep r otection , a n d On -
Resin Cyclization of P en tapeptide Test Sequ en ce: Tyr (t-
Bu )-Lys(Aloc)-Glu (OAl)-Lys(Boc)-Gly. Peptide syntheses
were performed as described above for the ACP decapeptide,
except 10 mL syringe reactors containing 250 mg of the
indicated resin maximally loaded with Fmoc-LinkerAm were
used. NR-Fmoc-protected Tyr(t-Bu), Lys(Aloc), Glu(OAl), Lys-
(Boc), and Gly were each coupled using the protocol described
above, except that the Fmoc group on the N-terminal Tyr was
left intact before the resins were dried. Following peptide
chain assembly, the dry resins were divided into two portions.
The first portion was placed in a 2 mL syringe reactor, the
N-terminal Fmoc group removed, and the resin washed with
DMF (3×), and then MeOH (3×), and the allyl-protected
peptide was cleaved with TFA-water (19:1) for 2 h. The
peptide was isolated by evaporation of the cleavage cocktail
and the residue lyophilized from acetonitrile-water (1:1). The
second portion of each resin was placed in Pyrex culture tubes
(Corning, 13 × 100 mm) equipped with Teflon lined screw-
caps. A stock solution of triphenylphosphine (1.7 g, 6.7 mmol)
and morpholine (0.5 mL, 5.7 mmol) in THF (25 mL) was
prepared and flushed with argon. Aliquots (4 mL) of this
solution were added to each vial, followed by 5-10 mg of
tetrakis(triphenylphosphine)Pd(0) complex that had been
freshly opened under argon. The vials were tightly sealed and
the suspensions shaken at 50 °C for 2 h. The contents of each
vial were transferred to a syringe reactor and the beads
P r ep a r a tion of Stilben e In ter m ed ia te 19 by On -Resin
H or n er -E m m on s Con d en sa t ion . 4-Nitrobenzyl dieth-
ylphosphonate [0.27 g, 1 mmol, previously prepared by reaction
of 4-nitrobenzyl bromide (5 g, 23 mmol) with triethyl phosphite
(3.8 g, 23 mmol) at 100 °C for 3 h] was dissolved in DMF (5
mL), treated with sodium methoxide (81 mg, 1.5 mmol), and
stirred for 15 min. The bright red solution was then added to
the remaining portion of resin 16a and the reaction shaken
overnight. The resin beads were isolated by filtration and
washed with water (5×), DMF-water (1:1, 5×), CH2Cl2 (5×),
and MeOH (2×). After being dried in vacuo, the product
stilbene was liberated by treatment with TFA-water (19:1)
for 2 h. The filtrate was collected and evaporated and the
crude product purified by flash chromatography (ethyl acetate-
hexane 3:7-1:0) to give 68 mg (54% from 13a ) of pure title
product, which was shown by 1H NMR, high-resolution MS
(calcd for C14H11NO3 241.0739, found 241.0740), and thin-layer
chromatography to be identical to an authentic sample of this
stilbene.25
Eva lu a tion of Dep h th a loyla tion Con d ition s (Ta ble 1).
Resin samples (2a or 2c, 50 mg) were placed in Pyrex culture
tubes (Corning, 13 × 100 mm) equipped with Teflon lined
screw-caps. Appropriate dephthaloylation cocktails (2 mL)
were added (exact descriptions are given in Table 1) and the
caps tightly sealed. The tubes were vortexed and then placed
in heating blocks at the temperatures and for the times
indicated in Table 1. Following this heating period, the
contents of each vial were transferred to a standard peptide
synthesis “syringe” reactor and the resins washed thoroughly
with DMF, CH2Cl2, and MeOH. Dephthaloylation efficiencies
were calculated from the amino content of the dried samples
determined by Fmoc-Nle-OH loading as described above and
are detailed fully in Table 1.
Det er m in a t ion of R esin -Sw ellin g Ch a r a ct er ist ics
(Ta ble 2). The bottom frits from 20 mL disposable polypro-
pylene columns (Bio-Rad, Hercules, CA) were replaced with
porous polyethylene frits punched from 1/16th in. medium-
porosity sheets (Porex Technology, Fairburn, GA). Each
column was then mounted vertically on a ring stand and
equipped with a two-way, luer-fitting, disposable stopcock
(Cole-Parmer, Vernon Hills, IL). The resin sample (2 g) was
placed in a 30 mL screw-capped polyethylene vial (VWR, So.
Plainfield, NJ ), THF (20 mL) was added, and the mixture was
sonicated and vortexed until the resin was fully swollen (30
min). The suspension was transferred to the column and
washed with THF until the bed volume was constant, at which
point the measurement was recorded. The next solvent to be
studied was then continuously eluted through the column until
the refractive index of eluent was the same as the fresh solvent
added and the bed volume was constant, at which point the
next measurement was recorded. The solvent series was run
in the order indicated in Table 2.
Ad d ition of F m oc-Lin k er Am to Am in o-Su bstitu ted
Resin s (Ta ble 3). P r oced u r e for Un d er loa d in g Resin s
(Ta r get Loa d in g ) 0.20 m m ol/g). Fmoc-LinkerAm (2.7 g,
5.0 mmol), BOP (2.3 g, 5.1 mmol), and HOBt (0.7 g, 5.2 mmol)
were dissolved in DMF and then treated with NMM (1.3 mL,
12 mmol). The solution was mixed for 3 min and then carefully
diluted to 25 mL. Aliquots of the activated LinkerAm solution
(1.0 mL, 0.4 mmol) were added to resin samples (2 g)
preswollen in the minimum volume of CH2Cl2-DMF (1:1, ca.
15 mL) in 20 mL glass scintillation vials using screw-caps with
conical polypropylene inserts (VWR), and the resultant sus-
pensions shaken overnight. The beads were isolated by
filtration, washed with DMF (3×), and capped with a solution
of acetic anhydride (0.3 M) and HOBt (0.3 M) in DMF for 2 h.
The beads were then washed with DMF (3×), CH2Cl2 (3×),
and MeOH (1×) and dried in vacuo to constant weight. All
samples were found to be ninhydrin negative. Loading ef-

3716 J . Org. Chem., Vol. 63, No. 11, 1998
Adams et al.
washed with DMF (1×), 0.4% concd HCl in DMF (2×), and
DMF (1×). One-half of each suspension was then treated to
remove the N-terminal Fmoc group, and, following TFA
cleavage, the allyl-deprotected peptides were isolated. Solu-
tions of BOP (45 mg, 0.10 mmol), HOBt (14 mg, 0.10 mmol),
and DIEA (35 µL, 0.20 mmol) in DMF (1 mL) were added to
each remaining peptide resin sample, and the suspensions
were shaken overnight. The N-terminal Fmoc groups were
then removed and the cyclized peptides isolated after TFA
cleavage. The peptide samples were then analyzed by analyti-
cal HPLC and MS as described above (see Table 4 for complete
results).
Com p a r a tive Syn th eses of Stilben e 19 (Ta ble 4). These
experiments were performed essentially as described above for
the basic synthesis of 19, but with minimized reaction times
and no monitoring of intermediates. Resin samples (200 mg)
were placed in 10 mL syringe reactors and, after several
washes, were left suspended in CH2Cl2-DMF (1:1). 4-(Hy-
droxymethyl)phenoxybutyric acid (HMPB, 13b, 0.42 g, 2.0
mmol) and HOBt (0.45 g, 3.3 mmol) dissolved in CH2Cl2-DMF
(1:1, 8 mL) were treated with DIPCDI (0.32 mL, 2.0 mmol)
for 2 min. Aliquots (1 mL) of the activated linker solution were
added to each syringe, which were then shaken for 2 h. The
samples were then washed with CH2Cl2-DMF (1:1, 3×) and
CH2Cl2 (3×). Triphenylphosphine (1.0 g, 3.8 mmol) and carbon
tetrabromide (1.3 g, 3.8 mmol) were dissolved in CH2Cl2 (20
mL). Two milliliters of this stock solution was added to each
syringe reactor, and the reactions were shaken for 2 h. The
resins were washed with CH2Cl2 (4×) and acetonitrile (2×)
and then treated for 2 h with 2 mL of a solution of 3-hydroxy-
benzaldehyde (0.5 g, 4.0 mmol) and potassium bis(trimethyl-
silyl)acetamide (0.88 g, 4.0 mmol) in DMF (20 mL). Next, the
resins were washed with CH2Cl2-DMF (4×) and DMF (4×)
and reacted with 2 mL of a solution of 4-nitrobenzyl dieth-
ylphosphonate (1.0 g, 3.7 mmol) and sodium methoxide (0.32
g, 5.9 mmol) in DMF (20 mL). The samples were shaken for
5 h, washed with DMF (5×), DMF-water (1:1, 5×), CH2Cl2
(3×), and MeOH (3×), and dried. Finally, the resins were
treated with TFA-water (19:1) for 2 h, the filtrate was
collected in tared scintillation vials, and the products were
isolated by evaporation and thorough drying. All products
were characterized as described above (see Table 4 for detailed
results).
r-Ch ym otr yp sin Digestion of Tr p -Gly Der iva tized
Resin s (Ta ble 4). A standard solution of activated Boc-Trp-
Gly-OH was prepared by dissolving the dipeptide (1.6 g, 4.5
mmol), BOP (2.0 g, 4.5 mmol), and HOBt (0.6 g, 4.5 mmol) in
DMF (30 mL) and adding NMM (1.0 mL, 9.0 mmol). Samples
of each support (50 mg) were placed in 2.5 mL syringe reactors,
and the dipeptide solution (2.0 mL) was added. The reactors
were shaken overnight, and then their contents were washed
with DMF (4×). A capping solution of acetic anhydride (0.3
M) and HOBt (0.3 M) in DMF (2 mL) was then added, and
the resins were shaken for 2 h. The resins were then washed
with DMF (1×), 1,4-dioxane (1×), and the enzyme buffer (15%
1,4-dioxane in 0.1 M aqueous ammonium bicarbonate, 3×). A
stock solution of R-chymotrypsin (50 mg) in buffer (50 mL) was
prepared and each reactor incubated with aliquots (2 mL) for
48 h. The resins were washed with water (3×) and DMF (1×),
and then 2 mL of an activated Fmoc-Nle solution [Fmoc-Nle-
OH (2.1 g, 6.0 mmol) and HOBt (1.0 g, 7.4 mmol) dissolved in
CH2Cl2-DMF (1:1, 30 mL) and treated with DIPCDI (0.95 mL,
6.0 mmol)] was added. The reactors were shaken for 2 h, and
then the contents were washed with DMF (3×), CH2Cl2 (3×),
and MeOH (3×). The resins were thoroughly dried and the
loadings determined by UV analysis of Fmoc. The percent
cleavage was then determined by comparison with Fmoc-
loading values obtained by addition of Fmoc-Nle to the resins
without prior dipeptide addition and enzyme incubation (see
Table 4 for detailed results).
Ack n ow led gm en t. We are grateful to the SBIR
program of the NIH, NIGMS for financial support
(Grant Nos. 2R44 GM50116-02A1, 2R44 GM50593-01,
and 2R44 GM55013-01). We thank Sara Biancalana
(Berlex Biosciences, Richmond, CA) for the synthesis of
the 68-residue peptide sequence.
J O9802269