Design and synthesis of novel hydrophilic spacers for the reduction of nonspecific binding proteins on affinity resins

Article abstract of DOI:10.1016/j.bmc.2004.03.052

Tubulin and actin often bind nonspecifically to affinity chromatography resins, complicating research toward identifying the cellular targets. Reduction of nonspecific binding proteins is important for success in finding such targets. We herein disclose the design, synthesis, and effectiveness in reduction of nonspecific binding proteins, of novel hydrophilic spacers (2-5), which were introduced between matrices and a ligand. Among them, tartaric acid derivative (5) exhibited the most effective reduction of nonspecific binding proteins, whilst maintaining binding of the target protein. Introduction of 5 on TOYOPEARL reduced tubulin and actin by almost 65% and 90% compared to that without the hydrophilic spacer, respectively, with effective binding to the target protein, FKBP12.

Full text of DOI:10.1016/j.bmc.2004.03.052

Bioorganic & Medicinal Chemistry 12 (2004) 2831–2841
Design and synthesis of novel hydrophilic spacers for the reduction
of nonspecific binding proteins on affinity resins
Takaaki Shiyama, Minoru Furuya, Akira Yamazaki, Tomohiro Terada
and Akito Tanaka*
Chemistry Department, Reverse Proteomics Research Institute Co., Ltd, 2-6-7 Kazusa-Kamatari, Chiba 292-0818, Japan
Received 23 February 2004; revised 23 March 2004; accepted 23 March 2004
Available online
Abstract—Tubulin and actin often bind nonspecifically to affinity chromatography resins, complicating research toward identifying
the cellular targets. Reduction of nonspecific binding proteins is important for success in finding such targets. We herein disclose the
design, synthesis, and effectiveness in reduction of nonspecific binding proteins, of novel hydrophilic spacers (2–5), which were
introduced between matricesand a ligand. Among them, tartaric acid derivative ( 5) exhibited the most effective reduction of
nonspecific binding proteins, whilst maintaining binding of the target protein. Introduction of 5 on TOYOPEARL reduced tubulin
and actin by almost 65% and 90% compared to that without the hydrophilic spacer, respectively, with effective binding to the target
protein, FKBP12.
ꢀ 2004 Elsevier Ltd. All rights reserved.
1. Introduction
molecular weights(50 and 42 kDa, respectively) to many
putative target proteins.
Affinity chromatography matricesbearing bioactive
compoundsplay an important role in the dicsovery of
novel drug targetsand the elucidation of drug mecha-
nisms. Their effectiveness has been demonstrated by the
discovery of FKBP12,1 HDAC,2 and cytosolic malate
dehydrogenase3 asthe psecific binding proteinsof
FK506, Trapoxin, and E-7070, respectively. The suc-
cessful isolation of target proteins by affinity chroma-
tography depends on the synthesis of polymeric resins
that can bind to the cellular target with maximum
selectivity and efficiency. Nonspecific binding of cellular
proteinsto affinity matricesistherefore a isgnificant
limitation to thisbiochemical approach, and reduction
of such binding by chemical modification is an impor-
tant goal and the responsibility of chemists and chemical
biologists. In particular, tubulin and actin are known as
representative nonspecific binding proteins because they
often interfere with affinity chromatography studies, due
to their high abundance and the similarity of their
There are now several commercially available matrices for
preparation of affinity resins. AffiGel, a polymer of sugar
derivatives, is one of the most popular resins for prepa-
ration of affinity matrices.4 However AffiGel isnot suit-
able for organic synthesis, because it is easily denatured
under organic synthesis conditions, and becomes dena-
tured even in N,N-dimethylformamide (DMF).5 Thus,
chemical approaches using this resin are limited. For
example, immobilization of ligandson AffiGel isusually
carried out under aqueous conditions, which usually re-
quiresues of excesamountsof ligand. Thislimitation
restrictsopportunitiesto find target proteinsusing affinity
resin technology because bioactive derivatives from the
bait compound bearing a linker moiety for immobiliza-
tion are usually synthesized through multi-steps and
sample amount is not large. In contrast, TOYOPEARL, a
poly(methacrylate) derivative, is stable under most syn-
thetic conditions, which allow the synthesis of more
effective affinity resins. Thisisattractive for chemistssince
the recent development of combinatorial chemistry allows
synthesis of a variety of compounds on functional poly-
mers.6–8 However, methacrylate polymersbearing bioac-
tive compounds often show high levels of nonspecific
protein binding with the target protein, FKBP12, in
comparison to AffiGel with the same ligands (Fig. 1B).9
Keywords: Nonspecific binding proteins; Affinity chromatography;
Hydrophilic spacer; FK506.
* Corresponding author. Tel.: +81-0438523990; fax: +81-0438523986;
e-mail: tanaka-a@reprori.jp
0968-0896/$ - see front matter ꢀ 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2004.03.052

2832
T. Shiyama et al. / Bioorg. Med. Chem. 12 (2004) 2831–2841
Figure 1. (A) Design of hydrophilic spacers studied in this study. (B) Binding proteins on TOYOPEARL bearing FK506 with and without previous
reported hydrophilic spacer 1. FKBP12 isknown asthe target protein of FK506, and other proteinssuch astubulin and actin are thought to be
nonspecific binding proteins.
Therefore, reduction of nonspecific binding proteins to
methacrylate derivativesisan important goal.
resins by the Fmoc strategy. Conversion of the hydroxy
group of mono benzyl protected penta(ethyleneglycol)
(7) to a bromide by carbon tetrabromide and triphe-
nylphosphine gave 8. Reaction of mono O-TBS pro-
tected 1,4-dihydroxybenzene (10) with mono trityl
protected penta(ethyleneglycol), prepared in a similar
manner to 7, by Mitsunobu reaction afforded 11. After
removal of the TBS group of 11 by tetrabutylammo-
nium fluoride (TBAF), a coupling reaction with 8 gave
13. Selective deprotection of the trityl group of 13 under
acidic conditions, followed by coupling with tert-butyl
a-bromoacetate, afforded 15. After removal of benzyl
group of 15, conversion of the hydroxy group to an
amino group wasperformed via azide 17, and wasfol-
lowed by introduction of an Fmoc group to the amine
and subsequent deprotection of the tert-butyl ester
under acidic conditions, resulting in the synthesis of the
desired N-Fmoc protected PEG type spacer (20).
The phrase ‘nonspecific protein binding’ is usually used
to represent proteins that bind to affinity resins based on
physical adsorption.10;11 In a previouspaper, 9 we carried
out a quantitative study and reported a linear relation-
ship between the CLOGP of ligands and the amount of
nonspecific binding proteins, and effectiveness of intro-
duction of a chemically stable and hydrophilic spacer (1,
Fig. 1). Compound 1 was designed based on the structure
of poly(ethyleneglycol) (PEG) that isoften uesd asa
hydrophilic spacer in biochemistry. However, the non-
specific binding proteins still remained even after intro-
duction of a hexamer of the PEG-type spacer (Fig. 1B).
This prompted the present study to develop novel
hydrophilic spacers to reduce them more effectively. We
herein report the design of novel hydrophilic spacers (2–
5) and their effectiveness on reduction of nonspecific
binding proteins.
We next designed other type of spacers based on tartaric
acid derivatives, which have an ethylene glycol moiety
(3–5, Scheme 2) because of their hydrophilic properties.
These spacers have amino acid-like structures and were
synthesized as N-Fmoc protected derivatives( 27, 29,
and 32, Scheme 2) aswell. Mono protected interme-
diates( 22, 24) were synthesized from symmetrical
starting materials, (4S,5S)-4,5-di(aminomethyl)-2,2-
2. Design and synthesis
We designed novel PEG like spacer (2, Fig. 1) bearing a
phenyl ring to join two PEG moieties. Compound 2 has
longer PEG unitscompared to the previousreported
spacer (1), aiming at greater hydrophilicity. Thisnovel
spacer has an amino group and carboxylic acid moiety
at each end, an amino acid like structure, which facili-
tates the synthesis of polymer on solid phase. Com-
dimethyldioxolane 21 and L-(+)-tartaric acid derivative
23, respectively. After coupling 22 and 24, deprotection
of the acetonide and benzyl groupsafforded the desired
compound 27 (Scheme 2). Compound 29 wasprepared
pound 2 wasysntheizsed asan N-Fmoc protected
derivative (20, (Scheme 1), for elongation reactionson
by condensation of N-Fmoc hydrazine 28 and
tartaric acid (Scheme 2). Compound 32 wassynthesized
L-(+)-

T. Shiyama et al. / Bioorg. Med. Chem. 12 (2004) 2831–2841
2833
a
b
O
O
O
HO
HO
OH
BnO
OH
BnO
Br
4
4
4
7
8
6
TBSO
TBSO
e
c
d
O
O
OTr
OH
10
OH
4
9
11
O
HO
BnO
O
g
f
4
O
O
O
OTr
O
OTr
4
4
13
12
O
X
O
O
h
OH
BnO
O
4
O
Y
4
O
O
O
O
4
4
14
15: X=BnO, Y=COOtBu
16: X=OH, Y=COOtBu
17: X=N3, Y=COOtBu
18: X=NH2, Y=COOtBu
i
j
k
l
19: X=NHF moc, Y=COOtBu
20: X=NHF moc, Y=COOH
m
Scheme 1. Synthesis of N-Fmoc protected 2 (20): Reagentsand conditions: (a) BnCl, KOH; (b) PPh 3, CBr4/CH2Cl2; (c) TBSCl, imidazole/DMF; (d)
TrO(CH2CH2O)5H, PBu3, DAMD/toluene; (e) TBAF/THF; (f) 8, NaH/THF; (g) TFA/CH2Cl2, H2O; (h) BrCH2COOtBu, NaH/THF, DMF; (i) H2,
Pd(OH)2/MeOH; (j) (i) TsCl, DMAP/Py, (ii) NaN3/DMF, (k) H2, Pd(OH)2/MeOH; (l) Fmoc-OSu, NEt3/THF; (m) 95% TFA/H2O.
a
O
O
O
O
H2N
NHF
O
moc
H2N
HO
NH2
c
22
21
O
O
O
O
H
N
NHFmoc
BnO
O
b
O
O
O
O
BnO
OH
OH
25
O
O
O
O
24
OH
23
HO OH
HO OH
HO
d
HO
O
OH
e
H
H
N
NHFmoc
HO
N
NHFmoc
BnO
O
O
O
27
26
O
OH
CO2H
f
FmocNHNH2
FmocHNHN
29
OH
28
O
OH
h
g
NH2
NH2
FmocHN
HCl H2N
FmocHN
N
H
CO2H
O
O
OH
30
32
31
Scheme 2. Synthesis of N-Fmoc protected 3 (27), 4 (29), and 5 (32): Reagentsand conditions: (a) Fmoc-OSu, NEt 3/THF; (b) BnOH, CDI, DBU/
CH3CN; (c) PyBOP, DIPEA/DMF; (d) 80% TFA/CH2Cl2; (e) H2, Pd(OH)2/MeOH, AcOEt; (f) -(+)-tartaric acid, EDCÆHCl/DMF; (g) Fmoc-OSu,
L
Na2CO3/acetone, water; (h) (i) PhI(OCOCF3)2/AcOEt–CH3CN–H2O, (ii)
L
-(+)-tartaric acid, EDCÆHCl/DMF.
from a mono N-Fmoc protected diaminomethane,
which wasysntheiszed from Fmoc glycinamide 31 by
Hoffman rearrangement using PhI(OCOCF3)2. This
intermediate wascondenesd with L-(+)-tartaric acid
without isolation to afford the desired intermediate 32
because of its instability (Scheme 2).

2834
T. Shiyama et al. / Bioorg. Med. Chem. 12 (2004) 2831–2841
In order to assess the effectiveness of novel spacers (2–5)
for reduction of nonspecific binding proteins, we syn-
thesized affinity matrices bearing a common ligand,
FK506 (33) on TOYOPEARL (36–40, Scheme 3).
FK506 is an immunosuppressive drug that targets to
FKBP12, the FK506 binding protein, with a Kd of
0.4 nM.12 The structure and function of the complex of
FK506 and FKBP12 have been well characterized at the
molecular level by S. L. Schreiber and co-workers.12–15
Since the molecular weight of FKBP12 (12 kDa) differs
significantly from that of actin (42 kDa) and tubulin
(50 kDa), it is easy to observe and compare the amounts
of specific and nonspecific protein binding. Introduction
of a linker moiety onto FK506 to connect it to the resin
wascarried out on the hydroxyl group at the 32 position
(34) because crystal structure studies indicate that this
position is not involved in the binding of FK506 with
FKBP12.15 Introduction of spacers was carried out by
the Fmoc strategy. Immobilization of Fmoc protected
spacers 20, 27, 29, 32, and the bait compound 34 on
resins was performed by using EDCÆHCl and HOBt.
hydrophilic spacers, we mixed affinity resins (36–40)
with a common lysate that was prepared from rat brains
using a simple buffer (buffer A; 0.25 M Sucrose, 25 mM
Tris–HCl, pH 7.4). Rat brain lysate was thought to be
suitable for our study since it included both the specific
binding protein, FKBP12, and nonspecific binding
9
proteinssuch astubulin and actin.
Binding proteinson
each affinity resin were completely eluted by a SDS
sample buffer solution (Nacalai, Sample Buffer Solution
with 2-ME(2·) for SDS-PAGE, cat. no 30566-22,
including 4%(w/v)-SDS, 20%(v/v)-glycerol, 0.01%(w/v)-
bromophenol
blue,
10%(v/v)-2-mercaptoethanol,
0.125 M Tris–HCl, pH 6.8) after extensive washes using
buffer A. Eluted proteinswere reoslved by SDS-Poly-
acrylamide gel electrophoresis, and identification of
tubulin, actin, and FKBP12 wasperformed by MALDI-
TOF mass spectrometry peptide mass fingerprinting
after in-gel digestion (Fig. 2). As determined by binding
competition with soluble FK506, we confirmed that only
FKBP12 specifically interacts with the FK506-bearing
resins and that other proteins such as tubulin and actin
are nonspecific binding proteins.
3. Results and discussion
Both AffiGel and TOYOPEARL matricesbearing
FK506 without spacers (36b, 36a) captured the target
protein, FKBP12 (lane 3, 4, Fig. 2A), from the com-
plicated protein mixture (lane 2). However, the amount
In order to compare the effect on reduction of nonspe-
cific binding proteinsby introduction of the novel
RO
MeO
Me
Me
O
O
OH
N
O
Me
Me
O
O
R: H (FK506, 33)
: HOOC-(CH2)6-CO- (34)
OH
O
Me
OMe
OMe
TM
a
TOYOPEARL
spacer
NH2
NHCO
35
NH2
:
or
AffiGelTM
n
O
O
b
spacer
NHCO
NOCH
n
MeO
Me
OH
Me
N
O
36a: no spacer(TOYOPEARL)
36b: no spacer(AffiGel)
O
O
Me
Me
O
O
Me
OH
O
spacer
2
3
4
5
n
OMe
OMe
1
2
3
4
5
37a
37b
37c
37d
37e
38a
38b
38c
38d
39a
39b
40a
40b
40c
40d
Scheme 3. Synthesis of affinity matrices bearing a FK506 derivative (34) with/without hydrophilic spacers (36–40): Reagentsand conditions: (a) (i)
N-Fmocprotectednovelhydrophilicspacer(20, 27, 29, or32),PyBOP,DIPEA/NMP;(ii)piperidine/DMF/CH2Cl2 ¼ 1:2:2;(b)34, EDCÆHCl,HOBt/DMF.

T. Shiyama et al. / Bioorg. Med. Chem. 12 (2004) 2831–2841
2835
Figure 2. Binding proteinson affinity matricesbearing FK506 with or without hydrophilic spacers( 36–40).
of nonspecific binding proteins such as tubulin and actin
were very different, since affinity matrices from AffiGel
only captured small amounts of nonspecific binding
proteinswhile the TOYOPEARL onesbound much
more. We reported that thisdifference mainly aroes
from differencesin hydrophilic propertiesin the previ-
We next analyzed proteinsbound to affinity reisns
bearing the tartaric acid derivativesaspsacers(
38–40,
Fig. 2B). All of these spacers also showed reductive ef-
fects on nonspecific binding proteins, whilst maintaining
capture of the target protein. In particular, introduction
of 4 (39a–b) and 5 (40a–d) were highly effective com-
pared to 2 (37a–e), while that of 3 (38a–d) wasalmost
equivalent to 2. This difference is interested because both
these derivativesare tartaric derivativesand 3 hasmore
hydroxyl groupscompared to 4 or 5 and isthought be
more hydrophilic. The reason for this difference is,
however, not clear. The nonspecific binding proteins
were reduced by introduction of 4 or 5 enough to carry
out researches on the target finding without being
interfered by them, while there were some proteins,
which were hardly reduced by introduction of the
hydrophilic spacers. We have not identified these pro-
teinsyet, and have no information about the reason for
not being susceptible to introduction of the hydrophilic
spacers.
9
ouspaper.
The PEG type spacer (2) could reduce nonspecific
binding proteinswhilst binding of the target protein was
almost constant in this series, as expected (n ¼ 1 to 5,
lane 5–9, Fig. 2A). The affinity resin bearing the pen-
tamer of 2 (37e, lane 9, Fig. 2A) yielded almost equiv-
alent amountsof FKBP12 compared to that without the
spacer (36a), while affording lower amountsof nonspe-
cific binding proteins. The amount of these nonspecific
binding proteinswaslower than that by introduction of
the previousPEG type ps acer 1 (Fig. 1B, lane 3). The
amount decreased in proportion to the number of 2,
indicating a repetition effect of 2. However reduction of
them by introduction of 2 were not sufficient since
affinity resins bearing pentamer of 2 still bound some
amount of the nonspecific binding proteins.
In order to assess the reductive effects quantitatively,
measurements of tubulin and actin were carried out by a

2836
T. Shiyama et al. / Bioorg. Med. Chem. 12 (2004) 2831–2841
GS-710 Calibrated Imaging Densitometer (BIO-RAD,
software; uantity ne-4.1.0), and were plotted in Fig-
pounds, for example, using deprotection reactions on
the resins after immobilization of bioactive derivatives.
Q
O
ure 3A and B, respectively. Figure 3A showed that the
most effective reduction of tubulin was found by intro-
duction of 5 among hydrophilic spacers employed in this
study. Introduction of tetramer of 5 reduced amount of
tubulin by 65% almost on the affinity resins. Introduc-
tion of 4 also exhibited a good reductive effect on
tubulin. Introduction of dimmer of 4 exhibited almost
the same reductive effect as that by introduction of the
same number of 5. A plot of the amount of actin was
shown in Figure 3B, which showed that the amount of
4. Conclusions
Tubulin and actin often bind nonspecifically to affinity
chromatography resins, complicating research toward
identifying the cellular targets of small molecules.
Reduction of nonspecific binding proteins is important
for success in discovery of such targets. In order to de-
velop strategies to circumvent this problem, we designed
and synthesized novel hydrophilic spacers 2–5 (Fig. 1).
Effectiveness was compared in experiments using a
common lysate prepared from rat brains after a common
bait compound, FK506 (33), wasimmobilized on them
(Fig. 2). Among spacers examined in this study, the
reductive effect of tartaric acid derivative 5 wassuperior.
Introduction of tetramer of 5 on TOYOPEARL reduced
tubulin and actin by almost 65% and 90% compared to
that without the hydrophilic spacer, respectively, with
effective binding to the target protein, FKBP12 (Figs. 2
and 3). While there were still small amounts of nonspe-
cific binding proteinsafter introduction of tetramer of 5
(lane 13, Fig. 2B), the reductive effect wassufficient for
the target finding by affinity chromatography. The use of
5 or other hydrophilic spacers studied in this study with
chemically stable matrices such as TOYOPEARL could
allow us to synthesize a variety of affinity matrices
bearing complexed compounds, while the use of the
popular matrix AffiGel is restricted because it is often
denatured under various synthetic conditions. We be-
lieve that these developments of the novel hydrophilic
spacers will accelerate the target-finding research and
applicationsin drug discovery.
actin wasmost effectively reduced by introduction of
4
(60% by the monomer, 80% by the dimer) and 5 (70% by
the dimmer, 90% by tetramer). While actin wastsill
observed even after introduction of tetramer of 5, it was
thought that the effect reached a maximum. Obvious
reductive effect on actin by introduction of 2 wasnot
observed in Figure 3B. The reason for the ineffectiveness
of 2 on reduction of actin wasnot clear.
Among the hydrophilic spacers examined in this study,
the introduction of 5 wasthe mots effective. Affinity
matricesof FK506 bearing the tetramer of 5 asa spacer
captured an equivalent amount of the target protein,
FKBP12, with only 35% and 10% of the nonspecific
binding proteins, tubulin and actin, respectively, com-
pared to the case without the hydrophilic spacer. These
results demonstrated that introduction of 5 can allow us
to carry out successful isolation of target proteins by
affinity chromatography without interference from
nonspecific binding proteins. Moreover, 5 ischemically
stable, therefore its use with chemically stable matrices
such as TOYOPEARL can allow us to synthesize a wide
variety of affinity matricesbearing complexed com-
3.0
(A)
2.5
2.0
1.5
1.0
0.5
5. Experimental
Thin-layer chromatography (TLC) wasperformed on
Merck silica gel 60 F-254 plates. For normal chroma-
tography, Merck silica gel type 60 (size 70–230) was
used. All evaporation was performed with a rotary
evaporator under reduced atmosphere. The structures of
all compoundswere confirmed by a LC-MS (Agilent
1100 SeriesLC/MSD) and 400 MHz proton nuclear
magnetic resonance spectrum (Brucker, Advance-series
400). The chemical shift values are reported in parts per
million on the d scale from internal standard tetra-
methylsilane. No attempt was made to maximize the
yields. Compound 34 was synthesized according to
Ref. 9.
0
1
2
3
4
5
2.5
2.0
1.5
1.0
0.5
0.0
(B)
0
1
2
3
4
5
5.1. 2-(2-{2-[2-(2-Benzyloxyethoxy)ethoxy]ethoxy}eth-
oxy)ethanol (7)
n mer of hydrophilic spacer (2-5)
Figure 3. Plotsof estimated amount of tubulin and actin with numbers
of hydrophilic spacer 2 (ꢀ), 3 (n), 4 (d), and 5 (m). The amountsof
tubulin and actin were represented as a ratio compared to amount of
FKBP12.
A mixture of potassium hydroxide (1.2 g, 21.4 mmol)
and penta(ethylene glycol) (6, 5.2 g, 21.8 mmol) was
stirred at 130 ꢁC for 15 min. After cooling to room

T. Shiyama et al. / Bioorg. Med. Chem. 12 (2004) 2831–2841
2837
temperature (rt), benzyl chloride (2.7 g, 21.3 mmol) was
added slowly, and the mixture was stirred at 130 ꢁC for
2 h. After cooling to rt, water wasadded, and the mix-
chromatography on silica gel (elution; 33% AcOEt in n-
1
hexane) to give 11 (colorless oil, 2.03g, 60%). H NMR
(CDCl3) d 0.16 (s, 6H), 0.97 (s, 9H), 3.23 (t, 2H,
J ¼ 5 Hz), 3.64–3.69 (m, 14H), 3.80 (t, 2H, J ¼ 5 Hz),
4.04 (t, 2H, J ¼ 5 Hz), 6.69–6.75 (m, 5H), 7.23–7.30 (m,
8H), 7.45–7.47 (m, 6H).
ture wasextracted with dichloromethane (CH Cl2). The
2
separated organic layer was dried over sodium sulfate
(Na2SO4), and the filtrate wasevaporated in vacuo and
purified by column chromatography on silica gel (elu-
tion; 5% methanol (MeOH) in ethyl acetate (AcOEt)).
Fractionscontaining product were collected and evap-
orated in vacuo to give 7 (light yellow oil, 2.3 g, 32%).
1H NMR (CDCl3) d 3.76–3.84 (m, 20H), 4.13 (s, 2H),
7.26–7.35 (m, 5H); MS m=z 329 (MHþ).
5.5. 4-[2-(2-{2-[2-(2-Trityloxyethoxy)ethoxy]ethoxy}eth-
oxy)ethoxy]phenol (12)
TBAF (1 M) in tetrahydrofuran (THF, 4 mL) wasadded
to a mixture of 11 (1.88 g, 2.74 mmol) and THF (20 mL)
at 0 ꢁC, and then stirred at rt for 10 min. The reaction
mixture waspoured into a mixture of water and AcOEt.
The separated organic layer was washed with brine and
dried over Na2SO4. After filtration, the filtrate was
evaporated in vacuo, and purified by column chroma-
tography on silica gel (elution; 10% MeOH in AcOEt).
Production fractionswere collected and evaporated in
vacuo to give compound 12 (light yellow oil, 875 mg,
85%). 1H NMR (CDCl3) d 3.23 (t, 2H, J ¼ 5 Hz), 3.64–
3.69 (m, 14H), 3.79 (t, 2H, J ¼ 5 Hz), 4.04 (t, 2H,
J ¼ 5 Hz), 4.68 (s, 1H), 6.71–6.79 (m, 4H), 7.20–7.30 (m,
9H), 7.45–7.47 (m, 6H).
5.2. [2-(2-{2-[2-(2-Bromoethoxy)ethoxy]ethoxy}eth-
oxy)ethoxymethyl]benzene (8)
Carbon tetrabromide (4.12 g, 12.4 mmol) wasadded to a
mixture of 7 (2.04 g, 6.21 mmol) and CH2Cl2 (20 mL).
After cooling to 0 ꢁC, triphenylphosphine (2.51 g,
12.4 mmol) wasadded. Thismixture wasstirred at room
temperature (rt) for 1 h, and evaporated in vacuo. The
resulting residue was purified by column chromatogra-
phy on silica gel (elution; 50% AcOEt in n-hexane).
Product fractionswere collected and evaporated
in vacuo to give 8 (light yellow oil, 2.1 g, 88%). 1H NMR
(CDCl3) d 3.46 (t, 2H, J ¼ 6:4 Hz), 3.62–3.69 (m, 16H),
3.80 (t, 2H, J ¼ 6:4 Hz), 4.57 (s, 2H), 7.26–7.35 (m, 5H);
MS m=z 391 (MHþ).
5.6. 1-[2-(2-{2-[2-(2-Benzyloxyethoxy)ethoxy]ethoxy}eth-
oxy)ethoxy]-4-[2-(2-{2-[2-(2-trityloxyethoxy)ethoxy]eth-
oxy}ethoxy)ethoxy]benzene (13)
5.3. 4-tert-Butyldimethylsilyloxyphenol (10)
A mixture of 12 (1.2 g, 2.1 mmol) and THF (20 mL) was
slowly added to a mixture of sodium hydride (NaH,
800 mg, 20 mmol; 60% in mineral oil) and THF (10 mL)
at 0 ꢁC, and stirred for 20 min under the same condi-
tions. A mixture of 8 (1.7 g, 4.3 mmol) and THF (20 mL)
wasadded, and the mixture tsirred for 20 min at 0 ꢁC
and at rt for 2 h. After addition of water (1 mL), the
resulting mixture was concentrated at reduced pressure.
After extraction with CH2Cl2, the combined organic
layerswere washed with brine and dried over Na 2SO4.
After filtration, the filtrate wasevaporated in vacuo and
purified by column chromatography on silica gel (elu-
tion; 50% AcOEt in n-hexane). Production fractions
were collected, and evaporated in vacuo to give 13 (light
yellow oil, 1.56 g, 84%). 1H NMR (CDCl3) d 3.23 (t, 2H,
J ¼ 5 Hz), 3.61–3.72 (m, 30H), 3.78–3.82 (m, 4H), 4.03–
4.07 (m, 4H), 4.56 (s, 2H), 6.82 (s, 4H), 7.09–7.34 (m,
14H), 7.44–7.47 (m, 6H).
A mixture of imidazole (3.1 g, 45 mmol), tert-butyldi-
methylsilyl chloride (4.5 g, 30 mmol), and DMF (50 mL)
wasadded to a mixture of hydroquinone ( 9, 17.3 g,
30 mmol) and DMF (30 mL), and the mixture wasstir-
red at rt overnight. Thismixture waspoured into a
mixture of water and diethyl ether. The separated
organic layer waswahsed with brine, and dried over
Na2SO4. After filtration, the filtrate wasevaporated in
vacuo and purified by column chromatography on silica
gel (elution; 10% AcOEt in n-hexane). Product fractions
were collected and evaporated in vacuo to give 10 (white
1
crystals, 3.6 g, 53%). H NMR (CDCl3) d 0.16 (s, 6H),
0.97 (s, 9H), 4.43 (s, 1H), 6.70 (d, 4H, J ¼ 2 Hz); MS
m=z 225 (MHþ).
5.4. tert-Butyldimethyl-{4-[2-(2-{2-[2-(2-trityloxyeth-
oxy)ethoxy]ethoxy}ethoxy)ethoxy]phenoxy}silane (11)
5.7. 2-(2-{2-[2-(2-{4-[2-(2-{2-[2-(2-Benzyloxyethoxy)eth-
oxy]ethoxy}ethoxy)ethoxy]phenoxy}ethoxy)ethoxy]eth-
oxy}ethoxy)ethanol (14)
A mixture of tributylphosphine (PBu3, 1.17 g, 5.8 mmol)
and toluene (2 mL) wasadded to a mixture of 2-(2-{2-[2-
(2-trityloxyethoxy)ethoxy]ethoxy}ethoxy)ethanol (2.8 g,
5.8 mmol) and toluene (10 mL) and stirred at rt for 1 h.
This solution was slowly added to a mixture of 10
(1.1 g, 4.9 mmol), 1,10-azo-bis-N,N-dimethylformamide
(DAMD, 1.0 g, 5.8 mmol), and toluene (10 mL), and
stirred at rt overnight. This mixture was poured into
AcOEt (20 mL) and the resulting precipitates removed
by filtration, and washed with AcOEt. The collected
filtrate wasevaporated in vacuo and purified by column
A solution of 10% trifluoroacetic acid (TFA) in CH2Cl2
(10 mL) wasadded to a mixture of
13 (1.43 g,
1.62 mmol), water (1 mL), and CH2Cl2 (20 mL) at 0 ꢁC,
and stirred at rt for 2.5 h. The reaction was poured into
saturated NaHCO3 and extracted with CH2Cl2 and
chloroform (CHCl3). The combined organic layerswere
washed with brine and dried over Na2SO4. After filtra-
tion, the filtrate wasevaporated in vacuo and purified by

2838
T. Shiyama et al. / Bioorg. Med. Chem. 12 (2004) 2831–2841
column chromatography on silica gel (elution; 10%
MeOH in AcOEt). Production fractionswere collected
and evaporated in vacuo to give compound 14 (light
yellow oil, 875 mg, 85%). 1H NMR (CDCl3) d 3.59–3.74
(m, 32H), 3.80–3.84 (m, 4H), 4.05–4.08 (m, 4H), 4.56 (s,
2H), 6.83 (s, 4H), 7.26–7.34 (m, 5H); MS m=z 641
(MHþ).
Sodium azide (NaN3, 26.4 mg, 4.1 mmol) wasadded to a
mixture of the above intermediate and DMF (1 mL),
and wasstirred at 60 ꢁC for 90 min. After cooling to rt,
thismixture waspoured into a mixture of AcOEt and
saturated NaHCO3. The organic layer wasthen washed
with brine and dried over Na2SO4. After filtration, the
filtrate wasevaporated in vacuo and purified by column
chromatography on silica gel (elution; 10% MeOH
in AcOEt). Product fractionswere collected, and
evaporated in vacuo to give tert-butyl [2-(2-{2-[2-(2-{4-
[2-(2-{2-[2-(2-azideethoxy)ethoxy]ethoxy}ethoxy)ethoxy]-
phenoxy}-ethoxy)ethoxy]ethoxy}ethoxy)ethoxy]acetate
5.8. tert-Butyl [2-(2-{2-[2-(2-{4-[2-(2-{2-[2-(2-benzyloxy-
ethoxy)ethoxy]ethoxy}ethoxy)ethoxy]phenoxy}ethoxy)-
ethoxy]ethoxy}ethoxy)ethoxy]acetate (15)
1
(17, colorless oil, 195 mg, 73%). H NMR (CDCl3) d
1.47 (s, 9H), 3.38 (t, 2H, J ¼ 5 Hz), 3.66–3.72 (m, 30H),
3.81–3.84 (m, 4H), 4.02 (s, 2H), 4.06–4.08 (m, 4H), 6.84
(s, 4H).
A mixture of 14 (817 mg, 1.28 mmol) and THF (6 mL)
waslsowly added to an ice-cooled mixture of NaH
(180 mg, 4.5 mmol; 60% in mineral oil), THF (5 mL) and
DMF (1 mL), and was stirred under the same conditions
for 30 min. A mixture of tert-butyl bromoacetate (1.0 g,
5.1 mmol) and THF (7 mL) wasadded and the mixture
stirred at rt for 2 h. This mixture was poured into water
and extracted with AcOEt, and washed with brine, then
dried over Na2SO4. After filtration, the filtrate was
evaporated in vacuo, and purified by column chroma-
tography on silica gel (elution; 5% MeOH in AcOEt).
Production fractionswere collected, and evaporated in
vacuo to give 15 (light yellow oil, 655 mg, 68%). 1H
NMR (CDCl3) d 1.47 (s, 9H), 3.61–3.73 (m, 32H), 3.81–
3.84 (m, 4H), 4.02 (s, 2H), 4.05–4.08 (m, 4H), 4.56 (s,
2H), 6.83 (s, 4H), 7.26–7.24 (m, 5H).
5.11. tert-Butyl [2-(2-{2-[2-(2-{4-[2-(2-{2-[2-(2-aminoeth-
oxy)ethoxy]ethoxy}ethoxy)ethoxy]phenoxy}ethoxy)eth-
oxy]ethoxy}ethoxy)ethoxy]acetate (18)
A mixture of 17 (190 mg, 0.323 mmol), Pd(OH)2–C
(18 mg), and MeOH (1 mL) wastisrred under an
hydrogen atmosphere at rt for 1 h. After filtration, the
filtrate wasevaporated in vacuo to give crude 18
(175 mg). MS m=z 664 (MHþ). Thiscompound wasused
in the following step without further purification.
5.12. tert-Butyl {2-[2-(2-{2-[2-(4-{2-[2-(2-{2-[2-(9H-fluo-
ren-9-yl-methoxycarbonylamino)ethoxy]ethoxy}ethoxy)-
ethoxy]ethoxy}phenoxy)ethoxy]ethoxy}ethoxy)ethoxy]-
ethoxy}acetate (19)
5.9. tert-Butyl [2-(2-{2-[2-(2-{4-[2-(2-{2-[2-(2-hydroxy-
ethoxy)ethoxy]ethoxy}ethoxy)ethoxy]phenoxy}ethoxy)-
ethoxy]ethoxy}ethoxy)ethoxy]acetate (16)
A mixture of 9-fluorenylmethylsucinimidyl carbonate
(Fmoc-OSu, 115 mg, 0.34 mmol), triethylamine (NEt3,
62 mg, 0.62 mmol), and THF (0.5 mL) wasadded to a
mixture of 18 (175 mg) and THF (1.2 mL) at 0 ꢁC, and
was stirred under the same conditions for 45 min. This
mixture waspoured into a mixture of water and AcOEt.
The separated organic layer was washed with brine and
dried over Na2SO4. After filtration, the filtrate was
evaporated in vacuo and purified by column chroma-
tography on silica gel (elution; 10% MeOH in AcOEt).
Product fractionswere collected and evaporated in va-
cuo to give compound 19 (colorless oil, 196 mg, 84%).
1H NMR (CDCl3) d 1.47 (s, 9H), 3.37–3.41 (m, 2H),
3.55–3.84 (m, 34H), 4.01 (s, 2H), 4.01–4.07 (m, 4H),
4.20–4.23 (m, 1H), 4.40 (d, 2H, J ¼ 7 Hz), 6.81 (s, 4H),
7.29–7.33 (m, 2H), 7.39–7.41 (m, 2H), 7.60 (d, 2H,
J ¼ 7 Hz), 7.76 (d, 2H, J ¼ 7 Hz).
A mixture of 15 (655 mg, 0.868 mmol), palladium
hydroxide on activated carbon (Pd(OH)2–C, 145 mg),
and MeOH (6 mL) wastsirred under an hydrogen
atmosphere at rt for 4 h. After filtration, the filtrate was
evaporated in vacuo to give crude 16 (575 mg). 1H NMR
(CDCl3) d 1.47 (s, 9H), 3.59–3.74 (m, 32H), 3.81–3.84
(m, 4H), 4.02 (s, 2H), 4.07–4.09 (m, 4H), 6.84 (s, 4H).
5.10. tert-Butyl [2-(2-{2-[2-(2-{4-[2-(2-{2-[2-(2-azidoeth-
oxy)ethoxy]ethoxy}ethoxy)ethoxy]phenoxy}ethoxy)eth-
oxy]ethoxy}ethoxy)ethoxy]acetate (17)
4-Dimethylaminopyridine (DMAP, 11 mg, 0.09 mmol)
wasadded to a mixture of 16 (254 mg, 0.451 mmol) and
pyridine (1.5 mL). After thismixture wascooled at 0 ꢁC,
p-toluenesulfonyl chloride (TsCl, 130 mg, 0.677 mmol)
wasadded and the mixture tsirred at rt for 2 h, then
poured into a mixture of ice, water, and AcOEt. The
organic layer was washed with saturated KHSO4, satu-
rated NaHCO3, and brine, and dried over Na2SO4.
After filtration, the filtrate wasevaporated in vacuo to
give crude tert-butyl [2-(2-{2-[2-(2-{4-[2-(2-{2-[2-(2-
tosyloxyethoxy)ethoxy]ethoxy}ethoxy)ethoxy]phenoxy}-
ethoxy)ethoxy]ethoxy}ethoxy)ethoxy]acetate, which was
used in the following step without further purification.
5.13. {2-[2-(2-{2-[2-(4-{2-[2-(2-{2-[2-(9H-Fluoren-9-yl-
methoxycarbonylamino)ethoxy]ethoxy}ethoxy)ethoxy]-
ethoxy}phenoxy)ethoxy]ethoxy}ethoxy)ethoxy]ethoxy}-
acetic acid (20)
A mixture of 19 (196 mg, 0.25 mmol), water (0.075 mL),
and TFA (1.5 mL) wastsirred at rt for 5 min. After
evaporation, the residue was purified by column chro-
matography on silica gel (elution; 10% MeOH in

T. Shiyama et al. / Bioorg. Med. Chem. 12 (2004) 2831–2841
2839
CHCl3). Fractionsincluding the target were collected,
and evaporated in vacuo to give compound 20 (colorless
oil, 175 mg, 96%). H NMR (CDCl3) d 3.37–3.39 (m,
2H), 3.56–3.84 (m, 34H), 4.02–4.07 (m, 4H), 4.12 (s,
2H), 4.22–4.23 (m, 1H), 4.40 (d, 2H, J ¼ 7 Hz), 5.47 (br
s, 1H), 6.81 (s, 4H), 7.29–7.33 (m, 2H), 7.38–7.41 (m,
2H), 7.60 (d, 2H, J ¼ 7 Hz), 7.76 (d, 2H, J ¼ 7 Hz); MS
m=z 830 (MHþ). Anal. Calcd for C43H59NO15/1.5H2O:
C, 60.27; H, 7.29; N, 1.63. Found: C, 60.09; H, 7.11;
N, 1.85.
tion; 50% AcOEt in n-hexane) to give 25 (0.49 g, 24%).
1H NMR (CDCl3) d 1.33–1.39 (m, 6H), 1.43 (s, 3H),
1.50 (m, 3H), 3.42 (br s, 2H), 3.53 (br s, 2H), 3.75 (m,
2H), 4.21 (m, 1H), 4.37–4.48 (m, 2H), 4.76–4.82 (m, 2H),
5.25 (s, 2H), 5.29 (m, 1H), 6.93 (m, 1H), 7.28–7.41 (m,
9H), 7.59 (d, 2H, J ¼ 7:5 Hz), 7.75 (d, 2H, J ¼ 7:5 Hz);
MS m=z 645 (MHþ).
1
5.17. N-[4-(9H-Fluoren-9-ylmethoxycarbonylamino)-
2S,3S-dihydroxybutyl]-2R,3R-dihydroxysuccinamic acid
benzyl ester (26)
5.14. (5S-Aminomethyl-2,2-dimethyl-[1,3]dioxolan-4S-
ylmethyl)carbamic acid 9H-fluoren-9-ylmethyl ester (22)
A mixture of 25 (0.46 g, 0.71 mmol), TFA (8 mL), and
CH2Cl2 (2 mL) wastsirred at rt for 5 h. The reaction
mixture wasconcentrated in vacuo, and the reuslting
residue recrystallized from n-hexane/AcOEt to give 26
Fmoc-OSu (2.1 g, 6.25 mmol) in THF (10 mL) and NEt3
(1.26 g, 12.5 mmol) were added to an ice-cooled solution
of 21 (1.0 g, 6.25 mmol) and THF (20 mL), and the
mixture stirred for 30 min under the same conditions.
The reaction mixture wasthen poured into a mixture of
water and AcOEt, and the separated organic layer was
washed with brine and was dried over anhydrous mag-
nesium sulfate (MgSO4). After filtration, the filtrate was
evaporated in vacuo and purified by column chroma-
tography on silica gel (elution; 5% ethanol (EtOH) in
AcOEt) to give 22 (1.22 g, 53%). 1H NMR (DMSO-d6) d
1.30 (s, 6H), 2.68 (d, 2H), 3.07 (m, 2H), 3.63–3.69 (m,
1H), 3.74–3.80 (m, 1H), 4.19–4.23 (m, 1H), 4.29 (m, 2H),
7.33 (m, 2H), 7.42 (t, 2H), 7.58 (m, 1H), 7.70 (d, 2H),
7.89 (d, 2H); MS m=z 383 (MHþ).
1
(220 mg, 55%). H NMR (DMSO-d6) d 2.99–3.06 (m,
2H), 3.09–3.19 (m, 2H), 3.44 (m, 2H), 4.18–4.23 (m, 1H),
4.26–4.27 (m, 3H), 4.47 (d, 1H),4.63 (m, 2H), 5.15 (m,
2H), 5.40 (d, 1H), 5.88 (t, 1H), 7.14 (m, 1H), 7.28–7.42
(m, 9H), 7.64 (m, 1H), 7.70 (d, 2H), 7.88 (d, 2H); MS
m=z 565 (MHþ).
5.18. N-[4-(9H-Fluoren-9-ylmethoxycarbonylamino)-
2S,3S-dihydroxybutyl]-2R,3R-dihydroxysuccinamic acid
(27)
A mixture of 26 (200 mg, 0.35 mmol), 10% palladium on
carbon (Pd-C, 50 mg), AcOEt (50 mL), and MeOH
(50 mL) wasstirred under a hydrogen atmosphere at rt
for 2 h. After filtration, the filtrate wasevaporated in
vacuo to give 27 (165 mg, 100%). 1H NMR (acetone-d6)
d 3.2–3.4 (m, 3H), 3.49–3.70 (m, 3H), 4.23 (m, 1H),
4.33–4.35 (m, 2H), 4.46 (m, 1H), 4.60 (m, 1H), 6.50 (m,
1H), 7.33 (t, 2H), 7.41 (t, 2H), 7.68 (m, 1H), 7.71 (d,
2H), 7.86 (d, 2H); MS m=z 475 (MHþ). Anal. Calcd for
C23H26N2O9/1.5H2O: C, 55.08; H, 5.82; N, 5.59. Found:
C, 54.88; H, 5.53; N, 5.91.
5.15. 5R-Hydroxycarboyl-2,2-dimethyl-[1,3]dioxolane-
4R-carboxylic acid benzyl ester (24)
A solution of 23 (740 mg, 3.89 mmol), 1,10-carbonyldi-
imidazole (CDI, 1.26 g, 7.78 mmol) and acetonitrile
(CH3CN, 50 mL) wastsirred at rt for 30 min. Benzyl
alcohol (420 mg, 3.89 mmol) and 1,8-diazabycy-
clo[5.4.0]undec-7-ene (DBU, 887 mg, 5.84 mmol) were
added and the mixture stirred at rt overnight. The
reaction mixture waspoured into a mixture of diluted
HCl and AcOEt and the separated organic layer was
washed with brine and dried over MgSO4. After filtra-
tion, the filtrate wasconcentrated in vacuo to give crude
24, which wasused in the next reaction without further
purification.
5.19. 4-[N0-(9H-Fluoren-9-ylmethoxycarbonyl)-hydraz-
ino]-2R,3R-dihydroxy-4-oxo-butyric acid (29)
A
mixture of Fmoc-hydrazine hydrochloride (28,
500 mg, 1.72 mmol), (+)-tartaric acid (634 mg,
4.22 mmol), DIPEA (0.300 mL, 1.72 mmol), EDCÆHCl
(485 mg, 2.53 mmol), and DMF (10 mL) wastsirred
overnight at rt. After concentration in vacuo, the
resulting residue was dissolved in saturated NaHCO3
and ethyl ether. The water layer wasacidified with 2 N
KHSO4 and extracted with AcOEt. The organic solution
waswahsed with brine and dried over Na 2SO4. After
filtration, the filtrate wasevaporated in vacuo, and re-
crystallized from MeOH/ethyl ether to give 29 (413 mg,
5.16. 5R-({5S-[(9H-Fluoren-9-ylmethoxycarbonyl-
amino)methyl]-2,2-dimethyl-[1,3]dioxolane-4S-ylmethyl}-
carbamoyl)-2,2-dimethyl-[1,3]dioxolane-4R-carboxylic
acid benzyl ester (25)
A mixture of crude 24, 22 (1.2 g, 3.14 mmol), bromo-
tris(pyrrolidino)phosphonium
hexafluorophosphate
(PyBOP, 3.37 g, 6.48 mmol), N,N0-diisopropylethyl-
amine (DIPEA, 1.25 g, 9.74 mmol), and DMF (10 mL)
wasstirred at rt for 17 h. The mixture waspoured into a
mixture of water and AcOEt and the separated organic
1
63%). H NMR (acetone-d6) d 4.29 (t, 2H, J ¼ 7:1 Hz),
4.35 (m, 1H), 4.60 (s, 1H), 4.63 (s, 1H), 7.33 (t, 2H,
J ¼ 7:4 Hz), 7.42 (t, 2H, J ¼ 7:4 Hz), 7.76 (d, 2H,
J ¼ 7:4 Hz), 7.85 (d, 2H, J ¼ 7:4 Hz); MS m=z 387
(MHþ). Anal. Calcd for C19H18N2O7/0.2H2O: C, 58.52;
H, 4.76; N, 7.18. Found: C, 58.42; H, 4.69; N, 7.37.
layer waswahsed with brine, and dried over MgSO
After filtration, the filtrate wasevaporated in vacuo and
purified by column chromatography on silica gel (elu-
.
4

2840
T. Shiyama et al. / Bioorg. Med. Chem. 12 (2004) 2831–2841
5.20. Fmoc-glycinamide (31)
(80%). After washing the resin with DMF carefully,
0.5 mL of mixture (acetic anhydride/CH2Cl2/
a
A solution of Fmoc-OSu (30.5 g, 90.5 mmol) in acetone
(250 mL) wasadded dropwies to a mixture of glycin-
amide hydrochloride (30, 10.0 g, 90.5 mmol), 10%
sodium carbonate solution (600 mL), and acetone
(250 mL) at 0 ꢁC. After stirring at rt for 1 h, the reaction
mixture wasconcentrated in vacuo and the residue was
dissolved in AcOEt. The solution was washed with 2 N
citric acid, saturated NaHCO3, and brine, and dried
over MgSO4. After filtration, the filtrate wasevaporated
in vacuo, and recrystallized from MeOH to give 31
(27.7 g, quant.). 1H NMR (CDCl3) d 3.87 (d, 2H,
J ¼ 5:6 Hz), 4.23 (t, 1H, J ¼ 6:6 Hz), 4.46 (d, 2H,
J ¼ 6:6 Hz), 7.30 (t, 2H, J ¼ 7:5 Hz), 7.39 (t, 2H,
J ¼ 7:5 Hz), 7.57 (d, 2H, J ¼ 7:5 Hz), 7.75 (d, 2H,
J ¼ 7:5 Hz); MS m=z 297 (MHþ).
NMP ¼ 1:8:2) wasadded for acetyl capping of remain-
ing amine groups. The reaction mixture was shaken at rt
for 3 h, and then washed with DMF cleaning five more
times.
ThisFmoc reisn wasmixed with 0.5 mL of a mixture
(piperidine/DMF/CH2Cl2 ¼ 1:4:4) at rt for 3 h. After
filtration, the resin was washed five times with DMF to
afford the objective resins bearing the hydrophilic spacer
(35).
5.23. Affinity resin bearing FK506 on TOYOPEARL
(36a)
A mixture of 34 (38.4 mg, 0.04 mmol), TOYOPEARLTM
(TOSOH, AF-Amino-650M, cat. 08002, 100 lL,
0.01 mmol), EDCÆHCl (9.2 mg, 0.048 mmol), HOBt
(6.5 mg, 0.048 mmol), and DMF (1 mL) wasshaken at rt
for 6 h. After removal of solvent by filtration, the resin
waswahsed with DMF. The reaction ratio wasdeter-
mined by the Ninhydrin test (82%). The resin was mixed
with a 20% DMF solution of acetic anhydride at rt for
1 h, washed with DMF, and 20% aqueous EtOH.
5.21. N-[(9H-Fluoren-9-ylmethoxycarbonylamino)-
methyl]-2R,3R-dihydroxysuccinamic acid (32)
A solution of 31 (1.50 g, 5.06 mmol), [bis(trifluoroacet-
oxy)iodo]benzene (PhI(OCOCF3)2, 2.83 g, 6.58 mmol),
AcOEt (15 mL), CH3CN (15 mL), and water (15 mL)
wasstirred at rt for 1.5 h. After acidification with 0.2 N
HCl, the water layer waswahsed with a mixture of
AcOEt and n-hexane (1:1), and basified with saturated
NaHCO3 at 0 ꢁC. After extraction with CHCl3, the
separated organic layer was washed with brine and dried
over MgSO4. After filtration, the filtrate wasevaporated
in vacuo to give N-Fmoc-diaminomethane, which was
immediately used in the next reaction.
The synthesis of affinity resins bearing FK506 on Affi-
Gele (36b, BIO-RAD, AffiGel 102 Gel, cat. 153-2401)
wascarried out in a ismilar manner. AffiGel wasuesd
after washing with DMF five times. Other resins (37–40)
were prepared in a similar manner.
A mixture of the above N-Fmoc-diaminomethane,
L
5.24. Preparation of rat brain lysate
-(+)-tartaric acid (1.42 g, 9.45 mmol), EDCÆHCl
(795 mg, 4.15 mmol), and DMF (100 mL) wasstirred at
rt overnight. After concentration in vacuo, the resulting
residue was dissolved in AcOEt. The organic solution
waswashed with 2 N citric acid and wasextracted with
saturated NaHCO3. The water solution was washed
with AcOEt, acidified with 2 N citric acid at 0 ꢁC, and
extracted with AcOEt. The separated organic layer was
washed with brine and dried over Na2SO4. After filtra-
tion, the filtrate wasevaporated in vacuo and the residue
Fresh rat brain was homogenized (1:10, w/v) in buffer A
(25 mM TrispH ¼ 7.4, 0.25 M sucrose). The homo-
genate wascentrifuged at 9500 rpm for 10 min. After the
supernatant was separated, it was centrifuged again at
50,000 rpm for 30 min. The obtained supernatant was
used for lysate and kept at )80 ꢁC before use.
5.25. Binding assay on affinity resins
1
recrystallized from MeOH to give 32 (588 mg, 29%). H
NMR (acetone-d6) d 4.24 (t, 1H, J ¼ 7:0 Hz), 4.33 (m,
2H), 4.44 (d, 1H, J ¼ 1:9 Hz), 4.58–4.71 (m, 3H), 7.33
(t, 2H, J ¼ 7:5 Hz), 7.41 (t, 2H, J ¼ 7:5 Hz), 7.71 (d, 2H,
J ¼ 7:5 Hz), 7.86 (d, 2H, J ¼ 7:5 Hz); MS m=z 401
(MHþ). Anal. Calcd for C20H20N2O7/0.67H2O: C, 58.25;
H, 5.21; N, 6.79. Found: C, 58.11; H, 5.05; N, 6.74.
The lysate as a crude tissue extract was diluted by buffer
A and total protein concentration wasabout 7 mg/mL.
This lysate was stirred with affinity resin at 4 ꢁC for
about 15 h to adsorb the non specific binding proteins or
specific binding proteins. A typical mixture has a total
volume of 1.0 mL, consisting of buffer A, 10 lL of beads
previously equilibrated by buffer A, and 0.5 mL of tissue
extract. After incubation, the resins were precipitated by
centrifugation in a microcentrifuge at 12,000 rpm for
1 min. The resins were washed five times with 1.0 mL of
buffer A. The washed beads were then resuspended in
20 lL of SDS sample buffer solution (Nacalai, Sample
Buffer Solution with 2-mercaptoethanol (2-ME, 2·) for
SDS-PAGE, cat. 30566-22, including 4%(w/v)-SDS,
20%(v/v)-glycerol, 0.01%(w/v)-bromophenol blue, 10%
(v/v)-2-mercaptoethanol, 0.125 M Tris–HCl, pH 6.8),
shaken at 25 ꢁC for 10 min, and centrifuged for 1 min.
5.22. Representative procedure for introduction of
hydrophilic spacer on resins
A mixture of 20 (66.4 mg, 0.08 mmol), TOYOPEARL
(200 lL, 0.02 mmol), PyBOP (52 mg, 0.1 mmol), DIPEA
(34 lL, 0.10 mmol), CH2Cl2 (0.8 mL), and N-methyl-2-
pyrrolidone (NMP, 0.2 mL) washsaken at rt for 4 h.
After filtration, the resin was washed with DMF 5 times.
The reaction ratio wasdetermined by the Ninhydrin test

T. Shiyama et al. / Bioorg. Med. Chem. 12 (2004) 2831–2841
2841
bioactive ligand and coupling reagentsusch asEDC.
The supernatant was subjected to SDS-PAGE followed
by CBB staining.
However, it is impossible for us to use an excess amount of
the ligand for immobilization since we usually synthesize
the compound with a linker moiety via a number of
synthetic steps, and only have small amounts. Thus,
immobilization reaction such as amide formation reaction
need to be carried out in organic solvent to avoid the
presence of water. AffiGel is sometimes irreversibly
denatured in organic solvent and sometimes give false-
positive targets. For example, we found some novel
nonspecific binding proteins when we synthesized affinity
material bearing FK506 in acetonitrile (data not shown)
whilst they were not detected when prepared in aqueous
solution.
Acknowledgements
We thank Drs. David Barrett (Fujisawa Pharmaceutical
Co., Ltd) and Masayuki Haramura (Reverse Proteomics
Research Institute Co., Ltd) for their critical reading of
this manuscript. The financial support from NEDO
(The New Energy and Industrial Technology Develop-
ment Organization) isgratefully acknowledged.
6. Borman, S. Chem. Eng. News 2001, 79, 49.
7. Bunin, B. A.; Ellman, J. A. J. Am. Chem. Soc. 1992, 114,
10997.
8. Ohlmeyer, M. H. J.; Swanson, R. N.; Dillard, L. W.;
Reader, J. C.; Asouline, G.; Kobayashi, R.; Wiegler, M.;
Still, W. C. Proc. Natl. Acad. Sci. U.S.A. 1993, 90,
10922.
9. Tamura, T.; Terada, T.; Tanaka, A. Bioconjugate Chem.
2003, 14, 1222.
10. Hofstee, B. H. J. Adv. Exp. Med. Biol. 1974, 42, 43.
11. Hofstee, B. H. J. Biochem. Biophys. Res. Commun. 1973,
53, 1137.
12. Siekierka, J. J.; Hung, S. H.; Poe, M.; Lin, C. S.; Sigal,
N. H. Nature 1989, 341, 755.
References and notes
1. Harding, M. W.; Galat, A.; Uehling, D. E.; Schreiber,
S. L. Nature 1989, 341, 758.
2. Taunton, J.; Hassig, C. A.; Schreiber, S. L. Science 1996,
272, 408.
3. Oda, Y.; Owa, T.; Sato, T.; Boucher, B.; Daniels, S.;
Yamanaka, H.; Shinohara, Y.; Yokoi, A.; Kuromitsu, J.;
Nagasu, T. Anal. Chem. 2003, 75, 2159.
4. Information on an agarose derivative, ‘AffiGel’ can be
found at the web site of Bio-Rad Laboratories, Inc. (http://
www.bio-rad.com).
13. Harding, M. W.; Galat, A.; Uehling, D. E.; Schreiber,
S. L. Nature 1989, 341, 758.
5. As described in the text, AffiGel is stable in aqueous
conditions, and is usually used in aqueous solutions.
Immobilization of a ligand on AffiGel isuusally carried
out in aqueous solution by addition of excess amounts of a
14. Liu, J.; Farmer, J. D.; Lane, W. S.; Friedman, J.;
Weissman, I.; Schreiber, S. L. Cell 1991, 66, 807.
15. Duyne, G. D. V.; Standaert, R. F. P.; Karplus, A.;
Schreiber, S. L.; Clardy, J. Science 1991, 252, 839.