Synthesis and optimization of hyaluronic acid-methotrexate conjugates to maximize benefit in the treatment of osteoarthritis

Article abstract of DOI:10.1016/j.bmc.2009.12.053

We previously reported that a conjugate of hyaluronic acid (HA) and methotrexate (MTX) could be a prototype for future osteoarthritis drugs having the efficacy of the two clinically validated agents but with a reduced risk of the systemic side effects of MTX by using HA as the drug delivery carrier. To identify a clinical candidate, we attempted optimization of a lead, conjugate 1. Initially, in fragmentation experiments with cathepsins, we optimized the peptide part of HA-MTX conjugates to be simpler and more susceptible to enzymatic cleavage. Then we optimized the peptide, the linker, the molecular weight, and the binding ratio of the MTX of the conjugates to inhibit proliferation of human fibroblast-like synoviocytes in vitro and knee swelling in rat antigen-induced monoarthritis in vivo. Consequently, we found conjugate 30 (DK226) to be a candidate drug for the treatment of osteoarthritis.

Full text of DOI:10.1016/j.bmc.2009.12.053

Bioorganic & Medicinal Chemistry 18 (2010) 1062–1075
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Synthesis and optimization of hyaluronic acid–methotrexate conjugates
to maximize benefit in the treatment of osteoarthritis
a,
a
a
a
a
Akie Homma a, , Haruhiko Sato , Tatsuya Tamura , Akira Okamachi , Takashi Emura , Takenori Ishizawa ,
Tatsuya Kato a, Tetsu Matsuura a, Shigeo Sato a, Yoshinobu Higuchi a, Tomoyuki Watanabe a,
Hidetomo Kitamura a, Kentaro Asanuma a, Tadao Yamazaki a, Masahisa Ikemi b, Hironoshin Kitagawa b,
Tadashi Morikawa b, Hitoshi Ikeya b, Kazuaki Maeda b, Koichi Takahashi b, Kenji Nohmi b, Noriyuki Izutani b,
Makoto Kanda b, Ryohchi Suzuki b
*
*
a Research Division, Chugai Pharmaceutical Co., Ltd, 1-135, Komakado, Gotemba, Shizuoka 412-8513, Japan
b Central Research Institute, Denki Kagaku Kogyo K. K, 3-5-1 Asahimachi, Machida, Tokyo 194-8560, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
We previously reported that a conjugate of hyaluronic acid (HA) and methotrexate (MTX) could be a pro-
totype for future osteoarthritis drugs having the efficacy of the two clinically validated agents but with a
reduced risk of the systemic side effects of MTX by using HA as the drug delivery carrier. To identify a
clinical candidate, we attempted optimization of a lead, conjugate 1. Initially, in fragmentation experi-
ments with cathepsins, we optimized the peptide part of HA–MTX conjugates to be simpler and more
susceptible to enzymatic cleavage. Then we optimized the peptide, the linker, the molecular weight,
and the binding ratio of the MTX of the conjugates to inhibit proliferation of human fibroblast-like syno-
viocytes in vitro and knee swelling in rat antigen-induced monoarthritis in vivo. Consequently, we found
conjugate 30 (DK226) to be a candidate drug for the treatment of osteoarthritis.
Received 9 November 2009
Revised 18 December 2009
Accepted 19 December 2009
Available online 28 December 2009
Keywords:
Hyaluronic acid
Methotrexate
Drug delivery
Osteoarthritis
Ó 2009 Elsevier Ltd. All rights reserved.
1. Introduction
4. HA modified through its carboxylic acid because of the diversity
of the chemical reaction.
To provide a novel drug for the treatment of osteoarthritis (OA)
which has both the pain-eliminating effect of hyaluronic acid (HA)
and the anti-inflammatory effect of methotrexate1 (MTX), we pro-
posed a drug delivery system (DDS) that localizes MTX at the syno-
vial membrane and uses HA as the carrier. Additionally, we
thought the strategy would reduce the risks of the systemic side ef-
fects of MTX.2,3
The efficacy of conjugate 1 indicated the importance of the con-
jugation of HA and MTX because the mixture of HA and MTX did
not show efficacy.4 Although conjugate 1 was promising, it was
apparent that further design and modifications were required.
The following modifications were considered necessary:
In our previous work,4 we found conjugate 1 (Fig. 1) which
showed anti-proliferative effects on human synovial cells stimu-
1. For simple and easy scale-up for industrialization, the number
of amino acids had to be decreased.
lated by TNF-
a
. Moreover, it inhibited knee swelling in an anti-
2. To assure better pain-relieving effect, the average molecular
weight (Mw) of the HA–MTX conjugates had to be maintained
at the same level as native HA.5–12
gen-induced monoarthritis rat model. Conjugate 1 consisted of
the following four parts:
3. To exert comparable efficacy of oral MTX, the binding ratio of
MTX had to be sufficiently high.
1. MTX bonded through
enzymes.
a- or c-carboxylic acid, cleavable by
2. A peptide chain recognized and cleaved by intracellular
enzymes.
3. Linkers with polyethylene glycol (PEG) to play the role of avoid-
ing the steric hindrance of HA against the approach of enzymes.
Considering these requirements, we optimized conjugate 1 to a
clinical candidate HA–MTX conjugates in the following steps. First,
we optimized the peptide part of conjugate 1 to be simpler and
more susceptible to enzymatic cleavage through enzyme recogni-
tion experiments in vitro and in arthritic models in vivo. Second,
we determined the range of the Mw of HA and the MTX binding ra-
tio from the viewpoint of efficacy and quality control of the conju-
gates. We also selected a linker for the conjugates that would exert
* Corresponding authors.
E-mail address: honmaake@chugai-pharm.co.jp (A. Homma).
0968-0896/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2009.12.053

A. Homma et al. / Bioorg. Med. Chem. 18 (2010) 1062–1075
1063
Figure 1. Structure of conjugate 1. Peptide: AsnPhePhe. Linker: PEG13. MTX binding ratio which means n/(m + n): 1.2%.
a-MTX connected (with
c
-). HA molecular weight:
1780 kDa.
the best efficacy. Through the optimization process, we identified
conjugate 30 as a promising candidate conjugate for a future OA
drug.
amino acids. On the other hand, the c-connected compounds (3,
5 and 7) were tolerant to the enzymes. Some compounds became
‘unknowns’, with mass spectra showing protonated MTX. We also
observed that cathepsin D tended to break between Phe and Phe
and cathepsins B and L broke at end of the PhePhe fragment, sug-
gesting that the HA–MTX conjugate must have a PhePhe sequence
2. Results and discussion
and be a-connected to MTX.
2.1. Cathepsin fragmentations
2.2. Peptide selection
First, we examined cathepsin fragmentations of the peptide to
determine which pattern connected and which peptide was well
recognized by cathepsins. We considered the simplest amino acid
sequence recognized by enzymes, especially cathepsins B, D or L
because they are typical peptidases found in the synovial tis-
sues.13–15 We used compounds 2–9, which were models of the
HA–MTX conjugates, for a simple analysis of the structures
(Fig. 2). Their amino acid sequences were what are known as rec-
ognition sites.16–18 All analyses were performed by LC–MS
spectroscope.
To select the peptides, we introduced amino acids to the conju-
gates. The synthetic route of conjugate 17, one of the conjugates
used in the evaluation of efficacy, is summarized in Schemes 2
and 3. Compound 16 was the key precursor and the final com-
pound confirmed to be a small molecule. The other precursors
were synthesized using the same method as for compound 16
and were used in the conjugation. The MTX binding ratio was cal-
culated by dividing the concentration of the MTX by the concentra-
tion of the disaccharide unit (GlcA-GlcNAc). These concentrations
were obtained by gel permeation chromatography (GPC) analysis
of the conjugates. GPC analysis determined both the binding ratio
and the Mw of the conjugates at the same time. As shown in Ta-
ble 2, conjugates 17–24 had 1.3–1.9% MTX binding, PEG13 linkers
and Mws of 1380–2190 kDa. All conjugates were single isomers,
with the exception of conjugate 1 (AsnPhePhe) and conjugate 20
(Glu). The effects of the HA–MTX conjugates on cell proliferation
Compounds 2ꢀ9 were synthesized using the solid-phase syn-
thesis method19,20 (Scheme 1). The
a- and c-connected MTX were
provided by applying the method to
c-protected or a-protected
glutamic acid. The compounds were treated for 24 h at 37 °C with
cathepsins B, D or L. After incubation, media were analyzed to de-
tect the MTX released from the compound. MTX derivatives which
bind to one amino acid are also known to have anti-inflammatory
effects but less than the MTX itself.2,3,21–23 Fragment production
was determined by the calculated percentage of the UV area
(Table 1).
induced by TNF-a stimulation were examined using human fibro-
blast-like synoviocytes (HFLS), as shown in Figure 3. Conjugates 1,
17, 19 and 23, which inhibited cell proliferation more effectively
than the others, were evaluated in the rat antigen-induced arthritis
model.24–26 The percentages of inhibition for the conjugates com-
The three peptides with
a-connected compounds, GlyPheLeu-
Gly, AsnPhePhe and PhePhe (compounds 2, 4 and 6), were well rec-
ognized by cathepsins and released MTX or MTX and one or two
Figure 2. Structure of model compounds 2–9. Peptide chain = -GlyPheLeuGly- (2, 3), -AsnPhePhe- (4, 5), -PhePhe- (6, 7), -Phe- (8, 9).

1064
A. Homma et al. / Bioorg. Med. Chem. 18 (2010) 1062–1075
Scheme 1. Reagents and conditions: (a) Succinic anhydride (1.3 equiv), Et3N (1.3 equiv), THF, 0 °C to rt, 17 h; (b) H2, 10% Pd–C (0.08 w/w), MeOH, rt, 3 days; (c) Fmoc-OSu
(1.3 equiv), NaHCO3 (2.4 equiv), 1,4-dioxane–H2O, 0 °C to rt, 9 h, followed by solid-phase synthesis. Fmoc deprotection: Pip/DMF, two times, coupling: DAMPA, Fmoc-AA-OH
or Fmoc-AA-OPfp was treated with DIC–HOBt or Pfp ester-NMM or HATU–HOBt protocol, cleavage: 20% TFA/CH2Cl2.
pared with the HA-treated groups are shown in Figure 4. The effi-
cacy of conjugates 1 and 17 was almost equal to that of oral MTX
treatment, indicated by the bold line. In vitro data for conjugate
17 (PhePhe), conjugate 19 (Phe) and conjugate 23 (Ile) showed
strong efficacy. In contrast to conjugates 19 and 23, conjugate 17
also showed efficacy in vivo. We think the criteria for conjugates
for OA drug development includes in vivo efficacy comparable to
oral administrated MTX. Consistency with the cathepsin fragmen-
tation experiments supported our decision to select ‘PhePhe’ for
the next optimization.
2.4. Molecular weight of conjugates
To select conjugates with the appropriate Mw, we synthesized
HA–MTX conjugates 17 (PhePhe, PEG13) and conjugate 30
(PhePhe, Alkyl2) which had Mws ranging between 330 and
2180 kDa (Fig. 7A and B). Control of the Mw was accomplished
by the use of HA of various Mws as the starting material. The three
classes (330, 800 and 1980 kDa) of Mw of conjugate 17 with
1.1ꢀ1.4% MTX inhibited HFLS proliferation (Table 3). However,
the conjugates of Mw 330 kDa did not show efficacy in vivo
(Fig. 7A). Similar to conjugate 17, conjugate 30 (Mw 340 kDa) with
1.9% MTX showed insufficient inhibition of knee swelling (Fig. 7B).
It is known that high Mw HA tends to stay longer than low Mw HA
in the joint cavities and synovial tissues. We thought the low Mw
was either insufficient for a conjugate to accumulate at the syno-
vial cells or it metabolized rapidly. In addition, HA (Mw of ca.
2000 kDa) behaves the same as the conventional HA in the human
body and so we determined 2000 kDa to be the target Mw.
2.3. Linker selection
To select the linkers, we synthesized conjugates 25–31 using
the same method as for conjugate 17 (Schemes 2 and 3) with the
peptide fixed to PhePhe (Table 2). During the initial phase, we
thought a lengthy PEG13 was necessary to avoid the hindrance
of enzyme recognition from the HA backbone; however, shorter
linkers played this role well. PEG8, PEG5 and Alkyl2 (conjugates
27, 28 and 30) showed anti-proliferative effects on the synovial
cells similar to PEG13 (conjugate 17, Fig. 5). The in vitro results
suggested that the efficacy seemed mainly due to the binding ratio
of MTX, with the exception of conjugate 31, which had a Lys
branch next to the peptide. On the other hand, the in vivo results
suggested that the efficacy of the conjugates was not influenced
by the length of the linkers but by the direction or hydrophobicity
provided by the linkers. We calculated the log P with ISIS Draw
software (version 2.5 J) and the order was as follows: PEG8
(ꢀ2.167), PEG5 (ꢀ1.893), Alkyl2 (ꢀ1.618), PEG13 (ꢀ1.459),
PEG10 (ꢀ1.185), Lys(OMe) (ꢀ1.169), PEG12 (ꢀ0.203), and Alkyl5
(ꢀ0.145). The order of the results for the three best conjugates
in vitro was PEG8, PEG5, and Alkyl2, and thus high hydrophobicity
provided good efficacy in vitro. The in vivo efficacy also supported
by the in vitro results (Fig. 6). Consequently, we selected an Alkyl2
linker because it showed strong efficacy in both experiments.
2.5. Range of binding ratios
To evaluate the range of the binding ratio which would provide
sufficient effect on the animal models, we compared in vitro effi-
cacy of conjugates with different MTX binding ratios (Table 4). In
the case of conjugate 30 with a 3.8% binding ratio, the in vitro effi-
cacy was as strong as MTX (IC50 0.14
jugate 30 with lower binding ratios (1.3% and 0.5%) showed weak
or no effect (Table 4). In our in vitro assay system (TNF- -induced
lM). On the other hand, con-
a
proliferation of HFLS), free MTX showed anti-proliferative effect in
a dose dependent manner but hyaluronic acid did not. Therefore,
the efficacy of the HA–MTX conjugates in this assay system derives
from that of MTX. Our preliminary experiments suggested that
HA–MTX is incorporated into synovial cells via hyaluronic recep-
tors on the cell surface and degraded by lysosomal enzymes to re-
lease MTX inside cells. This indicates that the amount of MTX

A. Homma et al. / Bioorg. Med. Chem. 18 (2010) 1062–1075
1065
Table 1
Cathepsin fragmentation
Compound
Peptide
Cathepsina
Compound remainingb (%)
Products (MTX-derivatives)b (%)
Released MTX
MTX-(amino acids)
2
3
a
GlyPheLeuGly
B
D
L
0
82
0
0
0
0
100 (Gly)
18 (GlyPhe)
100 (Gly)
c
GlyPheLeuGly
B
D
L
>90
>90
20
0
0
0
<10 (Gly)
<10 (GlyPhe)
40 (GlyPheLeu),
20 (GlyPheLeuGly), 20 (Gly)
4
5
6
7
8
9
a
c
a
c
a
c
AsnPhePhe
AsnPhePhe
PhePhe
PhePhe
Phe
B
D
L
0
0
0
0
0
63
100 (Asn)
100 (AsnPhe)
37 (Asn)
B
D
L
90
25
80
0
0
0
10 (Asn)
70 (AsnPhe), 5 (Asn)
20 (AsnPhePhe)
B
D
L
0
0
0
60
13
26
40 (Phe)
87 (Phe)
74 (Phe)
B
D
L
60
10
80
0
0
0
40 (MTX+1H)
90 (Phe)
20 (PhePhe)
B
D
L
70
100
45
15
0
55
15 (MTX+1H)
Phe
B
D
L
90
100
100
0
0
0
10 (MTX+1H)
a
Enzyme reaction: 37 °C, 24 h.
(Compound remaining, 2–9) + (products) = 100% (calculated from the LC peak area).
b
released is likely to be proportional to that of the MTX released
from a HA–MTX conjugate. We therefore speculate that the
amount of MTX conjugated to compound 30 with a low binding ra-
tio (ex. 0.5%) was too small to exert its efficacy in this assay system.
Similar results were found with in vivo experiments (Fig. 8A and
B). Both in vitro and in vivo effects were stronger in proportion
to the increase in the binding ratio, suggesting that the MTX would
almost certainly be released from the HA and would act as an anti-
inflammation drug. However, the conjugates with higher MTX
binding ratios (ex. 4.4%) were no more effective than the 3.8% con-
jugate, and the Mw of conjugates tended to decrease with increas-
ing MTX binding ratio. So, we concluded that conjugates require at
least 1.3%, hopefully around 3.8% of MTX binding ratio, for suffi-
cient efficacy and maintaining desirable Mw. With a weekly dose
of MTX, the 3.8% MTX conjugate was only 23.6 mg/week, a mere
1/79 the amount of the MTX of the conjugate that showed obvi-
ously stronger effects than oral MTX. The decrease in total dose
of MTX in this case suggests a decrease in the risks of systemic side
effects.
that the efficacy of MTX resulted from modification of the a-car-
boxylic acid of MTX and when 2 or 3 amino acid peptides were well
recognized. Conjugate 30 with a modified PhePhe peptide chain
and ethylenediamine linker was the best drug candidate for the
treatment of OA. The Mw of conjugate 30 was targeted to be
2000 kDa. The binding ratio of MTX was over 1.3% and, at around
3.8%, has sufficient efficacy compared with orally administered
MTX. For industrial synthesis, we improved the synthetic route
of conjugate 30: the intermediates mostly crystallized and so there
was no need for column purifications. The localization of MTX and
reduction of the total amount of MTX administered should reduce
the risks of the side effects of MTX. Conjugate 30 is expected to
possess the pain-eliminating effect of HA preparations and the
anti-inflammatory effect of a MTX derivative. Additionally, conju-
gate 30 will be a useful and safe drug administered by intra-artic-
ular injection and thus avoid the severe systemic risks of MTX.
Conjugate 30 (DK226) is now under investigation as a drug candi-
date for the treatment of OA.
4. Experimental
4.1. Chemistry
2.6. Synthesis for industrial development
After the peptide and linker were selected, we reconsidered the
process for providing conjugate 30 industrially (Scheme 4). All
intermediates were solid, purified without columns and provided
yields over 85%. Conjugate 30 obtained from this route was used
to evaluate the range of the MTX binding ratios shown above.
All amino acid derivatives, 1-hydroxybenzotriazole monohy-
drate (HOBt) and 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide
hydrochloride (EDC) were purchased from Watanabe Chemical
Industries, Ltd. HA was supplied by Denki Kagaku Kogyo, K. K. All
solvents were purchased from Kanto Chemical Co., Inc. 4-meth-
eylaminobenzoic acid was purchased from Tokyo Chemical Indus-
try Co., Ltd. All other compounds were of the best available
commercial grade. 1H NMR spectra were recorded on a JEOL ECP
270 or 500 MHz or a Mercury 300 MHz instrument in CDCl3,
DMSO-d6 or deuterium oxide solutions. Low-resolution mass spec-
tra were determined on LC/PDA/MS (HP 1100/TSP UV 6000/LCQ
3. Conclusion
We obtained HA–MTX conjugates that showed excellent effi-
cacy including inhibition of the HFLS proliferation induced by
TNF-
a stimulation and inhibition of knee swelling in rat antigen-
induced arthritis. Cathepsin fragmentation experiments suggested

1066
A. Homma et al. / Bioorg. Med. Chem. 18 (2010) 1062–1075
Scheme 2. Reagents and conditions: (a) Z-Phe-OH (1.3 equiv), HOBt (1.0 equiv), EDC (1.2 equiv), DMF, 0 °C to rt, 2 days; (b) H2, 10% Pd–C, rt; (c) Z-Phe-OH (1.6 equiv), HOBt
(1.5 equiv), EDC (1.6 equiv), DMF, 0 °C to rt, 2 days; (d) Z-Glu(OMe)-OH (1.1 equiv), HOBt (1.0 equiv), EDC (2.2 equiv), DMF, 0 °C to rt, 16 h; (e) DAMPA (1.5 equiv), HOBt
(1.5 equiv), EDC (1.5 equiv), DMF, 0 °C to rt, 2 days; (f) TFA, 0 °C, 40 min.
Scheme 3. Conditions of conjugation.
Classic, LC–MS) system using Cadenza CD-C18 (3.0 mm
I.D. ꢁ 20 mm) column at 35 °C. k was 190–400 nm total area with
retention times evaluated in minutes. The solvents ran as follows:
(A) 0.05% TFA, H2O, (B) 0.05% TFA, MeCN, Gradient (A/B): 95/5–0/
100 (9.5 min), 0/100 (2.5 min), Flow rate: 1.0 mL/min. The binding
ratio of MTX and the Mw of HA–MTX conjugates were determined
by GPC on an LC/PDA/RI (Waters Alliance 2695/2996/2414) system
using Shodex OHpak SB-806 HQ (8.0 mm I.D. ꢁ 300 mm) column

A. Homma et al. / Bioorg. Med. Chem. 18 (2010) 1062–1075
1067
Table 2
Structure of the conjugates
% of knee swelling against HA treated group
0
20
40
60
80
100
Conjugate
Peptide
Linker
MTX (%)
Mw (kDa)
1
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
AsnPhePhe
PhePhe
PheGly
Phe
Glu
Trp
Tyr
Ile
bAla
PhePhe
PhePhe
PhePhe
PhePhe
PhePhe
PhePhe
PhePhe
PEG13
PEG13
PEG13
PEG13
PEG13
PEG13
PEG13
PEG13
PEG13
PEG12
PEG10
PEG8
1.2
1.3–1.4
1.4
1.7
1.5
1.9
1.7
1.7
1.7
1.7
1.6
2.0
1.9
1.2
1.9–2.1
1.4
1780
1380–2190
1860
1790
1830
1390
1760
1620
1430
1550
1620
1490
1960
1860
1860–2180
1720
HA
1
17
19
23
PEG5
Alkyl5
Alkyl2
Lys
MTX p.o.
Figure 4. Peptide screening to evaluate the inhibition of knee swelling in antigen-
induced arthritic rats. Conjugates 1, 17, 19 and 23: 0.5 mg as HA, once a week, ia
MTX: 0.1 mg/kg, five times a week, p.o. for two weeks. Knee swelling, the difference
between right and left knee width, was measured on days 0–14. The bold line
indicates the average for MTX treatment groups.
PEG13 = –HN–(CH2)3–O–(CH2)2–O–(CH2)2–O–(CH2)3–NH–.
PEG12 = –HN–(CH2)3–O–(CH2)4–O–(CH2)3–NH–.
PEG10 = –HN–(CH2)3–O–(CH2)2–O–(CH2)3–NH–.
PEG8 = –NH–(CH2)2–O–(CH2)2–O–(CH2)2–NH–.
PEG5 = –NH–(CH2)2–O–(CH2)2–NH–.
Alkyl5 = –NH–(CH2)5–NH–.
Alkyl2 = –NH–(CH2)2–NH–.
IC50 (µM)
0.1
1
10
100
IC50 (µM)
0.1
1
10
100
1
17
1
25
26
17
18
19
20
21
22
23
24
27
28
29
30
31
MTX
MTX
Figure 5. Linker screening, evaluated by anti-proliferative effect on human
fibroblast-like synoviocytes (HFLS).
Figure 3. Peptide screening to evaluate the anti-proliferative effect on human
fibroblast-like synoviocytes (HFLS).
was warmed to rt. After 17 h, ethyl acetate was added. The organic
layer was washed with 1 N HCl solution and brine, dried over sodium
sulfate and evaporated. The product was purified by silica gel column
chromatography (eluent: dichloromethane/methanol, 20/1) to af-
ford 6.10 g (quant.) of the desired compound 10. 1H NMR
(270 MHz, CDCl3) d: 0.971 (s, 9H), 1.75–1.80 (m, 4H), 2.54–2.64 (m,
4H), 3.21–3.68 (m, 16H), 4.24 (d, 1H, J = 9.2 Hz), 5.08 (s, 2H), 5.56
(br, 1H), 7.00–7.35 (m, 7H).
at 40 °C. RI was analyzed at 35 °C. k was detected at 259 nm. The
solvent ran as follows: 50 mM sodium phosphate (pH 6.0), Flow
rate: 0.3 mL/min, injection volume: 100 lL.
4.1.1. MTX-a-AsnPhePhe-NHC10H20O3NH-HA (1)
Synthesis was previously reported.4
4.1.2. Synthesis of compounds 2–9
4.1.2.2. N-{1-[3-(2-{2-[3-(9H-Fluoren-9-ylmethoxycarbonylami-
no)-propoxy]-ethoxy}-ethoxy)-propylcarbamoyl]-2,2-dimethyl-
propyl}-succinamic acid (11). Compound 10 was dissolved in
20 mL of methanol. To the mixture, 0.5 g (8% weight of 10) of 10%
PdꢀC was added. After stirring under hydrogen atmosphere for
3 days, the reaction mixture was filtered and evaporated. The residue
was dissolved in 1,4-dioxane–H2O (50–30 mL). N-(9-Fluorenylmeth-
oxycarbonloxy)succinimide (Fmoc-OSu, 4.33 g, 12.8 mmol) and
4.1.2.1. N-[1-(3-{2-[2-(3-Benzyloxycarbonylamino-propoxy)eth-
oxy]ethoxy}propylcarbamoyl)-2,2-dimethylpropyl]succinamic
acid (10). [3-(2-{2-[3-(2-Amino-3,3-dimethylbutyrylamino)-pro-
poxy]ethoxy}ethoxy)propyl]carbamic acid benzyl ester (KNC Labo-
ratories Co., Ltd, 5.0 g, 10.7 mmol) and succinic anhydride (1.28 g,
12.8 mmol) were dissolved in THF (50 mL) at 0 °C. After triethyl-
amine (Et3N, 1.78 mL, 12.8 mmol) was added, the reaction mixture

1068
A. Homma et al. / Bioorg. Med. Chem. 18 (2010) 1062–1075
Table 3
% of knee swelling against HA treated group
Effects of high, middle and low Mw conjugate 17 on human fibroblast-like
synoviocytes (HFLS).
0
20
40
60
80
100
Conjugate
MTX (%)
1.2
Mw (kDa)
1780
IC50
1.08
12.9
13.2
14.7
(lM)
1
1
17
1.1
1.4
1.4
330
800
1980
17
27
28
30
Table 4
Effects of high, middle and low binding ratios of MTX for conjugate 30 on human
fibroblast-like synoviocytes (HFLS).
Conjugate
MTX (%)
1.2
Mw (kDa)
1780
IC50 (lM)
1
1.08
MTX p.o.
30
0.5
1.3
3.8
2170
2090
1910
>12.0
1.13
0.09
Figure 6. Linker screening to evaluate the inhibition of knee swelling in rat
antigen-induced arthritis. Conjugates 27, 28 and 30: 0.5 mg as HA, once a week, i.a.
MTX: 0.1 mg/kg, five times a week, p.o. for two weeks. Knee swelling, the difference
between the right and left knee width, was measured on days 0–14. The bold line
indicates the average for MTX treatment groups.
4.1.2.3.1. Fmoc deprotection. Washing was carried out twice
with 1 mL of Pip/DMF (1 min/time), with 1.5 mL of Pip/DMF
(30 min) and nine times with 1 mL of DMF (1 min/time).
NaHCO3 (1.98 g, 23.5 mmol)were addedtothe solutionand stirredat
0 °C . After 1 h, the reaction mixture was warmed tort. After 8 h, ethyl
acetate was added. The organic layer was washed with 1 N HCl solu-
tion and brine, dried over sodium sulfate and evaporated. The prod-
uct was purified by silica gel column chromatography (eluent:
dichloromethane/methanol, 20/1) to afford 1.71 g (2.61 mmol, 24%
fortwo steps)of thedesired compound 11. 1H NMR (270 MHz, CDCl3)
d: 0.97 (s, 9H), 1.75 (br, 4H), 2.55–2.66 (m, 4H), 3.21–3.62 (m, 16H),
4.21–4.40 (m, 4H), 5.51 (br, 1H), 6.95–7.76 (m, 11H). LC–MS m/z:
656.1 (M+H)+.
4.1.2.3.2. Coupling conditions. DIC–HOBt protocol: N-(9-fluore-
nylmethoxycarboxyl)-
methoxycarboxyl)- -glutamic acid
Glu(OtBu)-OH), N-(9-fluorenylmethoxycarboxyl)-
L
-leucune (Fmoc-Leu-OH), N-(9-fluorenyl-
-tert-butyl ester (Fmoc-
-glutamic acid
-tert-butyl ester (Fmoc-Glu(OH)-OtBu) or 4-[N-(2,4-diamino-6-
L
c
L
a
pteridinylmethyl)-N-methylamino]benzoic
acid
(DAMPA,
2.0 equiv), DIC (2.0 equiv) and HOBt (2.0 equiv) were agitated un-
der nitrogen in 1.5 mL of DMF for 2 h and for 24 h for DAMPA.
Pfp ester protocol: N-(9-Fluorenylmethoxycarboxyl)-
pentafluorophenyl ester (Fmoc-Gly-OPfp) or N-(9-fluorenylmeth-
oxycarboxyl)- -phenylalanine pentafluorophenyl ester (Fmoc-
L-glycine
4.1.2.3. Solid-phase synthesis of compounds 2–9. Protected
amino acids, 20% piperidine in N,N-dimethylformamide (Pip/
DMF) and coupling reagents were purchased from Watanabe
Chemical Industries, Ltd.
L
Phe-OPfp) (2.0 equiv) and N-methylmorpholine (NMM, 2.0 equiv)
were agitated in 1.5 mL of DMF for 2 h. and with Pip/DMF for
30 min.
Fmoc-resin (BACHEM, 0.55 mmol/g, 72.7 mg, 0.04 mmol) was
agitated in 1.5 mL of Pip/DMF for 30 min. After washing the resin
nine times with 1 mL of DMF, 2.5 equiv of 11, N,N0-diisopropylcar-
bodiimide (DIC) and HOBt and 1.5 mL of DMF were added. After 6 h
of agitating for coupling, the resin was washed six times with 1 mL
of DMF.
HATU–DIPEA protocol: In the case of N-(9-fluorenylmethoxy-
carboxyl)-L-asparagine (Fmoc-Asn-OH) (2.0 equiv), O-(7-axabenzo-
triazol-1-yl)-N,N,N0,N0-tetramethyluronium hexafluorophosphate
(HATU, 2.0 equiv) and diisopropylethylamine (DIPEA, 2.0 equiv)
were agitated in 1.5 mL of DMF for 24 h. After each coupling, wash-
ing was carried out with 1 mL of DMF (six times, 1 min/time).
P < 0.005
P < 0.005
A
B
P < 0.05
P < 0.0005
40
40
30
30
MTX, p.o.
MTX, p.o.
20
10
0
20
10
0
Vehicle HA
330
800
17
1980
Vehicle HA
340 2180
kDa
30
Figure 7. Effects of HA–MTX with different molecular weights of HA on knee swelling in rat antigen-induced arthritis. Knee swelling, the difference between the right and left
knee width, was measured on days 0–14, and the AUC for knee swelling was calculated (A and B). Various Mws of conjugate 17, conjugate 30, HA (Mw 1900–2500 kDa) and
vehicle were i.a. injected once a week. Results are expressed as mean + SEM. P <0.05, compared to HA-treated group. The bold lines indicate the average for the MTX
treatment groups.

A. Homma et al. / Bioorg. Med. Chem. 18 (2010) 1062–1075
1069
P<0.01
B
A
4.0
3.0
2.0
1.0
0
P<0.0001
P<0.05
50
40
30
20
10
0
*
*
*
*
*
*
*
*
*
0
2
4
6
8
10 12 14
1.3
3.8
Days after challenge
Saline
HA
MTX, p.o.
30 (MTX, %)
Saline
HA: 0.5 mg, once/week, i.a.
Mean+SEM
30 (MTX 1.3%) : 0.5 mg as HA, once/week, i.a.
30 (MTX 3.8%) : 0.5 mg as HA, once/week, i.a.
MTX: 1.25 mg/kg, 5 times/week, p.o.
Weekly dose of MTX
µg/week
µg/week
30 (MTX 1.3%): 7.7
23.6
30 (MTX 3.8%):
*P < 0.05 vs HA
Mean SEM
µg/week
MTX (1.25 mg/kg): 1875
Figure 8. Effects of conjugate 30 with different binding ratios of MTX on knee swelling in rat antigen-induced arthritis. Knee swelling, the difference between the right and
left knee width, was measured on days 0–14 (A) and the AUC for knee swelling calculated (B). Results are expressed as mean + SEM. P <0.05, compared to HA-treated group.
Scheme 4. Reagents and conditions: (a) H-Phe-OtBu, EDC–HOBt, NMM, THF, 5–25 °C; (b) H2SO4, MeCN, 0–25 °C; (c) N-(tert-butoxycarbonyl)-1,2-diaminoethane, EDC–HOBt,
NMM, THF, 5–25 °C; (d) H2, Pd/C, DMF–HCl aq, rt; (e) DAMPA, EDC–HOBt, NMP, 55 °C; (f) 7 N HCl–ethanol, DMF, rt; (g) sodium hyaluronate, tris[2-(2-
methoxyethoxy)ethyl]amine, EDC–HOOBt, THF–H2O, 5 °C; (h) NaOH aq, 5 °C.
4.1.2.3.3. Cleavage. Peptide resin was washed with dichloro-
methane and dried. The dried resin was treated for 1 h with 20%
TFA/dichloromethane (1.5 mL). Then the resin was filtered off
and washed with dichloromethane. The products were purified
using the separable HPLC method. (Column: Inertsil ODS-3,
20.0 mm I.D. ꢁ 150 mm, GL Science Inc., Mobile Phase: (A)
0.05% TFA, H2O, (B) 0.05% TFA, MeCN, Gradient (A/B): 90/10–
50/50 (25 min), Flow Rate: 10.0 mL/min, Wavelength: 215,
308 nm).
4.1.2.6. Compound 4. Fmoc-Phe -OPfp, Fmoc-Asn-OH, Fmoc-
Glu(OtBu)-OH and DAMPA were used for coupling. The desired
compound was obtained as a 2ꢂTFA salt (8.5 mg, 14%). LC–MS tR:
3.8 min, m/z: 1277.5 (M+H)+.
4.1.2.7. Compound 5. Fmoc-Phe-OPfp, Fmoc-Asn-OH, Fmoc-
Glu(OH)-OtBu and DAMPA were used for coupling. The desired
compound was obtained as a 2ꢂTFA salt (7.2 mg, 12%). LC–MS tR:
3.7 min, m/z: 1277.7 (M+H)+.
4.1.2.4. Compound 2. Fmoc-Gly-OPfp, Fmoc-Leu-OH, Fmoc-Phe-
OPfp, Fmoc-Glu(OtBu)-OH and DAMPA were used for coupling.
The desired compound was obtained as a 2ꢂTFA salt (8.9 mg,
15%). LC–MS tR: 3.8 min, m/z: 1243.6 (M+H)+.
4.1.2.8. Compound 6. Fmoc-Phe-OPfp, Fmoc-Glu(OtBu)-OH and
DAMPA were used for coupling. The desired compound was ob-
tained as a 2ꢂTFA salt (7.3 mg, 13%). LC–MS tR: 4.0 min, m/z:
1163.7 (M+H)+.
4.1.2.5. Compound 3. Fmoc-Gly-OPfp, Fmoc-Leu-OH, Fmoc-Phe-
OPfp, Fmoc-Glu(OH)-OtBu and DAMPA were used for coupling.
The desired compound was obtained as a 2ꢂTFA salt (7.4 mg,
13%). LC–MS tR: 3.7 min, m/z: 1243.8 (M+H)+.
4.1.2.9. Compound 7. Fmoc-Phe-OPfp, Fmoc-Glu(OH)-OtBu and
DAMPA were used for coupling. The desired compound was ob-
tained as a 2ꢂTFA salt (8.9 mg, 16%). LC–MS tR: 3.9 min, m/z:
1163.7 (M+H)+.

1070
A. Homma et al. / Bioorg. Med. Chem. 18 (2010) 1062–1075
4.1.2.10. Compound 8. Fmoc-Phe-OPfp, Fmoc-Glu(OtBu)-OH and
DAMPA were used for coupling. The desired compound was ob-
tained as a 2ꢂTFA salt (8.8 mg, 18%). LC–MS tR: 3.6 min, m/z:
1016.7 (M+H)+.
4.1.3.1.4. MTX-(OMe)-a-PhePhe-NH–C10H20O3–NH-Boc (15) .
Compound 14 (514 mg, 0.576 mmol) was suspended in 30 mL
of methanol. 100 mg (19% weight of 14) of 10% PdꢀC was added
to the mixture. After stirring at rt for 1.5 h under hydrogen atmo-
sphere, the reaction mixture was filtered, evaporated, and the res-
idue was then dissolved in 5 mL of DMF. HOBt (132 mg,
0.864 mmol, 1.5 equiv), EDC (166 mg, 0.864 mmol, 1.5 equiv) and
DAMPA (281 mg, 0.864 mmol, 1.5 equiv) were added to this DMF
solution at 0 °C. After being stirred at rt for 2 days, 5% Na2CO3
was added to the reaction solution, and the generated precipitate
was purified by amino silica gel (NH-DM1020, 100–200 mesh,
from Fuji Silysia Chemical Ltd) column chromatography (eluent:
dichloromethane/methanol, 100/7 and chloroform/methanol,
100/4, two times) to provide 415 mg (0.390 mmol, 68%) of com-
pound 15 as a yellow powder.
4.1.2.11. Compound 9. Fmoc-Phe-OPfp, Fmoc-Glu(OH)-OtBu and
DAMPA were used for coupling. The desired compound was ob-
tained as a 2ꢂTFA salt (8.8 mg, 18%). LC–MS tR: 3.5 min, m/z:
1016.7 (M+H)+.
4.1.3. Synthesis of amine compounds as precursors of the
conjugates
4.1.3.1. 4,7,10-Trioxa-13-[N-[N-[N-[4-[[(2,4-diamino-6-pteridi-
nyl)methyl]methylamino]benzoyl]-a-(5-methylglutam-
yl)]phenylalanyl]phenylalanylamino]tridecanylamine: MTX-
a-
PhePhe-NH–C10H20O3–NH2 (16).
1H NMR (270 MHz, DMSO-d6) d: 1.36 (s, 9H), 1.48–1.61 (m, 4H),
1.81–1.92 (m, 2H), 2.24 (t, 2H, J = 7.9 Hz), 2.70–3.10 (m, 8H), 3.22
(s, 3H), 3.25–3.47 (m, 12H), 3.54 (s, 3H), 4.25–4.50 (m, 3H), 4.79
(s, 2H), 6.61 (br s, 2H), 6.76–6.83 (m, 3H), 7.06–7.24 (m, 10H),
7.45 (br s, 1H), 7.67–7.80 (m, 4H), 7.86 (d, 1H, J = 8.1 Hz), 8.09 (d,
1H, J = 7.4 Hz), 8.15 (d, 1H, J = 8.1 Hz), 8.56 (s, 1H). LC–MS m/z:
1087.5 (M+Na)+.
4.1.3.1.1. Z-Phe-NH–C10H20O3–NH-Boc (12). N-Carbobenzoxy-
-phenylalanine (Z-Phe-OH: 852 mg, 2.85 mmol), N-t-butoxycar-
L
bonyl-4,7,10-trioxa-1,13-tridecanediamine (NH2–C10H20O3–NH–
Boc, 760 mg, 2.37 mmol), and HOBt (363 mg, 2.37 mmol) were dis-
solved in 6 mL of DMF. To the solution EDC (546 mg, 2.85 mmol)
was added at 0 °C. The reaction mixture was warmed to rt and stir-
red for 2 days and then ethyl acetate was added. The organic layer
was washed with 10% citric acid, 5% Na2CO3 and brine, dried with
sodium sulfate and evaporated. The product was purified by silica
gel column chromatography (eluent: dichloromethane/methanol,
100/3) to afford 1.35 g (2.24 mmol, 79%) of the desired compound
12 as a colorless oil. 1H NMR (270 MHz, CDCl3) d: 1.43 (s, 9H), 1.56–
1.74 (m, 4H), 3.06 (d, 2H, J = 6.8 Hz), 3.17ꢀ3.58 (m, 16H), 4.30–4.39
(m, 1H), 4.98 (br, 1H), 5.08 (s, 2 H), 5.50 (br, 1H), 6.40 (br, 1H),
7.16–7.32 (m, 10H). LC–MS m/z: 624.3 (M+Na)+.
4.1.3.1.5. MTX-(OMe)-a-PhePhe-NH–C10H20O3–NH2 (16). 3 mL
of TFA was added to compound 15 (413 mg, 0.388 mmol) at 0 °C.
After stirring for 40 min, the reaction mixture was evaporated
and the residue was purifying by aminosilica gel column chroma-
tography (eluent: dichloromethane/methanol, 100/7, two times) to
provide 344 mg (0.346 mmol, 89%, 40% for five steps) of compound
16 as a yellow powder. 1H NMR (270 MHz, DMSO-d6) d: 1.49–1.95
(m, 4H), 1.81–1.92 (m, 2H), 2.24 (t, 2H, J = 7.9 Hz), 2.70–3.10 (m,
8H), 3.22 (s, 3H), 3.25–3.47 (m, 12H), 3.54 (s, 3H), 4.25–4.50 (m,
3H), 4.79 (s, 2H), 6.61 (br s, 2H), 6.76–6.83 (m, 3H), 7.06–7.24
(m, 10H), 7.45 (br s, 1H), 7.83 (br t, 1H, J = 5.8 Hz), 8.01 (d, 1H,
J = 7.9 Hz), 8.09 (d, 1H, J = 7.1 Hz), 8.15 (d, 1H, J = 7.8 Hz), 8.56 (s,
1H). LC–MS m/z: 965.5 (M+H)+.
4.1.3.1.2. Z-PhePhe-NH–C10H20O3–NH-Boc
(13). 1.35 g
(2.24 mmol) of compound 12 was dissolved in 12 mL of methanol.
To the mixture 200 mg (15% weight of 12) of 10% PdꢀC was added.
After stirring under hydrogen atmosphere for 4 h, the reaction
mixture was filtered and evaporated. The residue was dissolved
in 10 mL of dry DMF. To the solution, Z-Phe-OH (1.07 g, 3.57 mmol,
1.6 equiv), HOBt (514 mg, 3.36 mmol, 1.5 equiv) and EDC (688 mg,
3.59 mmol, 1.6 equiv) were added at 0 °C. The reaction mixture
was warmed to rt, stirred for 2 days and then ethyl acetate was
added to the mixture. Following the same procedure for 12, 13
was obtained (1.56 g, 2.08 mmol, 93%) as a white solid generated
by the addition of n-hexane to the concentrated residue. 1H NMR
(270 MHz, CDCl3) d: 1.43 (s, 9H), 1.60ꢀ1.78 (m, 4H), 2.96–3.60
(m, 20H), 4.42–4.59 (m, 2H), 4.96–5.07 (m, 3H), 5.41 (br d, 1H),
6.39 (br, 1H), 6.73 (br d, 1H), 7.08–7.31 (m, 15H). LC–MS m/z:
771.3 (M+Na)+.
The other precursors were obtained following the same proce-
dures used for 16.
4.1.3.2. MTX-(OMe)-a-PheGly-NH–C10H20O3–NH2 (precursor of
18). Following the same procedure for compound 16 but with N-
carbobenzoxyglycine (Z-Gly-OH) in place of Z-Phe-OH in the pro-
duction of 13, the target compound (528 mg, 0.604 mmol, 54%
for five steps) was obtained as
a
yellow powder. 1H NMR
(270 MHz, DMSO-d6) d: 1.51–1.64 (m, 4H), 1.84–1.94 (m, 2H),
2.21–2.30 (m, 2H), 2.55 (t, 2H, J = 6.3 Hz), 2.78–2.92 (m, 1H),
3.03–3.76 (m, 17H), 3.22 (s, 3H), 3.55 (s, 3H), 4.26–4.52 (m, 2H),
4.79 (s, 2H), 6.63 (br s, 2H), 6.82 (d, 2H, J = 8.7 Hz), 7.11–7.24 (m,
5H), 7.47 (br s, 1H), 7.62–7.72 (m, 4H), 8.04–8.16 (m, 2H), 8.28
(br t, 1H), 8.56 (s, 1H). LC–MS m/z: 875.5 (M+H)+.
4.1.3.1.3. Z-Glu(OMe)PhePhe-NH–C10H20O3–NH-Boc (14). Com-
pound 13 (500 mg, 0.668 mmol) was dissolved in 10 mL of metha-
nol. 150 mg (30% weight of 13) of 10% PdꢀC was added to the
mixture. After stirring at rt for 24 h under hydrogen atmosphere,
the reaction mixture was filtered and evaporated. The residue was
4.1.3.3. MTX-(OMe)-a-Phe-NH–C10H20O3–NH2 (precursor of
19). Following the same procedure for compound 16 but exclud-
ing the procedure for 4.1.3.2 (above), the target compound
(496 mg, 0.607 mmol, 47% for four steps) was obtained as a yellow
powder. 1H NMR (300 MHz, DMSO-d6) d: 1.49–1.59 (m, 4H), 1.82–
1.89 (m, 2H), 2.19–2.27 (m, 2H), 2.55 (t, 2H, J = 7.2 Hz), 2.73–3.10
(m, 4H), 3.23 (s, 3H), 3.17–3.48 (m, 12H), 3.55 (s, 3H), 4.21–4.28
(m, 1H), 4.38–4.45 (m, 1H), 4.80 (s, 2H), 6.61 (br s, 2H), 6.83 (d,
2H, J = 9.3 Hz), 7.11–7.20 (m, 5H), 7.46 (br s, 1H), 7.66 (br s, 1H),
7.73 (d, 2H, J = 9.0 Hz), 7.83 (t, 1H), 7.92 (d, 1H, J = 8.4 Hz), 8.12
(d, 1H, J = 7.5 Hz), 8.56 (s, 1H). LC–MS m/z: 818.4 (M+H)+.
dissolved in 5 mL of DMF. N-Carbobenzoxy-L-glutamic acid-c-
methyl ester (Z-Glu(OMe)-OH: 217 mg, 0.734 mmol, 1.1 equiv),
then HOBt (102 mg, 0.668 mmol, 1.0 equiv) and EDC (141 mg,
0.734 mmol, 2.2 equiv) were added at 0 °C and the reaction mixture
stirred at rt for 16 h. Following the same procedure for 12, the de-
sired residue was obtained followed by silica gel column chromatog-
raphy (eluent: dichloromethane/methanol, 100/5). 14 was obtained
(529 mg, 0.594 mmol, 89%) as a white solid which was generated by
the addition of n-hexane to the concentrated residue. 1H NMR
(270 MHz, DMSO-d6) d: 1.36 (s, 9H), 1.50–1.85 (m, 6H), 2.20 (t, 2H,
J = 7.9 Hz), 2.70–3.10 (m, 8H), 3.25–3.48 (m, 12H), 3.56 (s, 3H),
3.93–4.02 (m, 1H), 4.20–4.60 (m, 2H), 5.00 (s, 2H), 6.77 (br t, 1H),
7.10–7.45 (m, 16H), 7.82 (br t, 1H, J = 6.1 Hz), 7.91 (d, 1H,
J = 7.9 Hz), 8.22 (d, 1H, J = 7.9 Hz). LC–MS m/z: 914.3 (M+Na)+.
4.1.3.4. MTX-(OMe)-a-Glu(OMe)–NH–C10H20O3–NH2 (precursor
of 20). Following the same procedure for 4.1.3.3 (above) with Z-
Glu(OMe)-OH in place of Z-Phe-OH, the target compound
(600 mg, 0.738 mmol, 45% in four steps) was obtained as a yellow

A. Homma et al. / Bioorg. Med. Chem. 18 (2010) 1062–1075
1071
powder. 1H NMR (270 MHz, DMSO-d6) d: 1.50–2.03 (m, 8H), 2.24–
4.1.3.10. MTX-(OMe)-a-PhePhe-NH–C8H16O2–NH2 (precursor of
2.31 (t, 2H), 2.34–2.40 (t, 2H), 2.49–2.57 (t, 2H), 2.97–3.52 (m,
14H), 3.21 (s, 3H), 3.53 (s, 3H), 3.55 (s, 3H), 4.15–4.36 (m, 2H),
4.78 (s, 2H), 6.61 (br s, 2H), 6.81 (d, 2H, J = 8.7 Hz), 7.46 (br s,
1H), 7.67 (br s, 1H), 7.72 (d, 2H, J = 8.6 Hz), 7.84 (br t, 1H), 7.95
(d, 1H), 8.14 (d, 1H), 8.55 (s, 1H). LC–MS m/z: 814.4 (M+H)+.
26). Following the procedure for compound 16 with N-t-butoxy-
carbonyl-4,7-dioxa-1,10-decanediamine (NH2–C8H16O2–NH-Boc)
in place of NH2–C10H20O3–NH-Boc, the target compound
(407 mg, 0.442 mmol, 24% for five steps) was obtained as a yellow
powder. 1H NMR (400 MHz, DMSO-d6) d: 1.50–1.57 (m, 4H), 1.85–
1.91 (m, 2H), 2.21–2.28 (m, 2H), 2.60 (t, 2H), 2.70–3.13 (m, 6H),
3.22 (s, 3H), 3.25–3.45 (m, 8H), 3.55 (s, 3H), 4.27–4.49 (m, 3H),
4.79 (s, 2H), 6.60 (br s, 2H), 6.82 (d, 2H, J = 8.8 Hz), 7.07–7.21 (m,
10H), 7.43 (br s, 1H), 7.69 (br s, 1H), 7.71 (d, 2H, J = 8.8 Hz), 7.75
(br t, 1H), 7.85 (d, 1H), 8.08 (d, 1H), 8.13 (d, 1H), 8.56 (s, 1H).
LC–MS m/z: 921.4 (M+H)+.
4.1.3.5. MTX-(OMe)-
21). Following the same procedure for 4.2.3.3 with N-carboben-
zoxy- -tryptophan (Z-Trp-OH) in place of Z-Phe-OH, the target
a-Trp-NH–C10H20O3–NH2 (precursor of
L
compound (171 mg, 0.200 mmol, 20% in four steps) was obtained
as a yellow powder. 1H NMR (270 MHz, DMSO-d6) d: 1.50–1.61
(m, 4H), 1.84–1.97 (m, 2H), 2.23–2.32 (m, 2H), 2.50–2.56 (t, 2H),
2.92–3.15 (m, 4H), 3.22 (s, 3H), 3.29–3.45 (m, 12H), 3.55 (s, 3H),
4.29–4.49 (m, 2H), 4.78 (s, 2H), 6.64 (br s, 2H), 6.80 (d, 2H), 6.92
(t, 1H), 7.04 (t, 1H), 7.10 (s, 1H), 7.26 (d, 1H), 7.44 (br s, 1H), 7.51
(d, 1H), 7.65 (br s, 1H), 7.69 (d, 2H), 7.82 (br t, 1H), 7.93 (d, 1H),
8.10 (d, 1H), 8.55 (s, 1H), 10.80 (s, 1H). LC–MS m/z: 857.5 (M+H)+.
4.1.3.11. MTX-(OMe)-a-PhePhe-NH–C6H12O2–NH2 (precursor of
27). Following the procedure for compound 16 with N-t-butoxy-
carbonyl-3,6-dioxa-1,8-octanediamine (NH2–C6H12O2–NH-Boc) in
place of NH2–C10H20O3–NH-Boc, the target compound (148 mg,
0.166 mmol, 23% in five steps) was obtained as a yellow powder.
1H NMR (270 MHz, DMSO-d6) d: 1.81–1.91 (m, 2H), 2.20–2.25 (m,
2H), 2.61–2.64 (t, 2H), 2.70–2.97 (m, 6H),3.22 (s, 3H), 3.27–3.47
(m, 8H), 3.55 (s, 3H), 4.27–4.47 (m, 3H), 4.79 (s, 2H), 6.62 (br s,
2H), 6.82 (d, 2H, J = 8.7 Hz), 7.06–7.25 (m, 10H), 7.46 (br s, 1H),
7.67 (br s, 1H), 7.71 (d, 2H, J = 8.6 Hz), 7.85 (d, 1H), 7.92 (br t, 1H),
8.07 (d, 1H), 8.15 (d, 1H), 8.56 (s, 1H). LC–MS m/z: 893.6 (M+H)+.
4.1.3.6. MTX-(OMe)-
22). Following the same procedure for 4.2.3.3 with N-carboben-
zoxy- -tyrosine (Z-Tyr-OH) in place of Z-Phe-OH, the target com-
a-Tyr-NH–C10H20O3–NH2 (precursor of
L
pound (133 mg, 0.160 mmol, 62% in four steps) was obtained as
a yellow powder. 1H NMR (270 MHz, DMSO-d6) d: 1.51–1.62 (m,
4H), 1.85–1.95 (m, 2H), 2.23–2.31 (m, 2H), 2.51–2.58 (t, 2H),
2.63–2.91 (m, 2H), 2.95–3.16 (m, 2H), 3.22 (s, 3H), 3.27–3.54 (m,
12H), 3.56 (s, 3H), 4.22–4.35 (m, 2H), 4.79 (s, 2H), 6.57 (d, 2H,
J = 8.1 Hz), 6.61 (br s, 2H), 6.82 (d, 2H, J = 8.7 Hz), 6.92 (d, 2H,
J = 8.1 Hz), 7.47 (br s, 1H), 7.67–7.88 (m, 5H), 8.13 (d, 1H), 8.55
(s, 1H). LC–MS m/z: 834.4 (M+H)+.
4.1.3.12. MTX-(OMe)-a-PhePhe-NH–C4H8O–NH2 (precursor of
28). Following the procedure for compound 16 with N-t-butoxy-
carbonyl-3-oxa-1,5-pentanediamine (NH2–C4H8O–NH-Boc) in
place of NH2–C10H20O3–NH-Boc, the target compound (52 mg,
0.061 mmol, 7% in five steps) was obtained as a yellow powder.
1H NMR (270 MHz, DMSO-d6) d: 1.84–1.92 (m, 2H), 2.20–2.27
(m, 2H), 2.60–2.64 (t, 2H), 2.71–2.99 (m, 6H), 3.22 (s, 3H), 3.25–
3.45 (m, 4H), 3.54 (s, 3H), 4.27–4.50 (m, 3H), 4.79 (s, 2H), 6.61
(br s, 2H), 6.81 (d, 2H, J = 8.4 Hz), 7.05–7.21 (m, 10H), 7.45 (br s,
1H), 7.65 (br s, 1H), 7.70 (d, 2H, J = 8.6 Hz), 7.84 (d, 1H), 7.91 (br
t, 1H), 8.07 (d, 1H), 8.15 (d, 1H), 8.55 (s, 1H). LC–MS m/z: 849.4
(M+H)+.
4.1.3.7. MTX-(OMe)-
23). Following the same procedure for 4.2.3.3 with N-carboben-
zoxy- -isoleucine (Z-Ile-OH) in place of Z-Phe-OH, the target com-
a-Ile-NH–C10H20O3–NH2
(precursor
of
L
pound (562 mg, 0.717 mmol, 41% in four steps) was obtained as a
yellow powder. 1H NMR (270 MHz, DMSO-d6) d: 0.76–0.80 (m, 6H),
0.99–1.10 (m, 1H), 1.36–1.45 (m, 1H), 1.49–1.73 (m, 5H), 1.88–2.07
(m, 2H), 2.33–2.38 (m, 2H), 2.55 (t, 2H, J = 6.6 Hz), 2.98–3.48 (m,
14H), 3.21 (s, 3H), 3.56 (s, 3H), 4.05–4.13 (m, 1H), 4.40–4.48 (m,
1H), 4.78 (s, 2H), 6.60 (br s, 2H), 6.82 (d, 2H, J = 8.4 Hz), 7.46 (br
s, 1H), 7.66–7.72 (m, 3H), 7.98 (br t, 1H), 8.12 (d, 1H, J = 7.6 Hz),
8.56 (s, 1H). LC–MS m/z: 784.4 (M+H)+.
4.1.3.13. MTX-(OMe)-a-PhePhe-NH–C5H10–NH2 (precursor of
29). Following the procedure for compound 16 with N-t-butoxy-
carbonyl-1,5-pentanediamine (NH2–C5H10–NH-Boc) in place of
NH2–C10H20O3–NH-Boc,
the
target
compound
(148 mg,
0.175 mmol, 17% in five steps) was obtained as a yellow powder.
1H NMR (270 MHz, DMSO-d6) d: 1.16–1.56 (m, 6H), 1.81–1.97
(m, 2H), 2.21–2.29 (m, 2H), 2.69–3.06 (m, 6H), 3.23 (s, 3H), 3.55
(s, 3H), 4.25–4.50 (m, 3H), 4.80 (s, 2H), 6.65 (br s, 2H), 6.82 (d,
2H, J = 8.6 Hz), 7.08–7.24 (m, 10H), 7.50 (br s, 1H), 7.60–7.89 (m,
5H), 8.10–8.16 (m, 2H), 8.55 (s, 1H). LC–MS m/z: 847.4 (M+H)+.
4.1.3.8. MTX-(OMe)-a-bAla-NH–C10H20O3–NH2 (precursor of
24). Following the same procedure for 4.2.3.3 with N-carboben-
zoxy-b-alanine (Z-bAla-OH) in place of Z-Phe-OH, the target com-
pound (230 mg, 0.310 mmol, 37% for four steps) was obtained as a
yellow powder. 1H NMR (270 MHz, DMSO-d6) d: 1.49–1.62 (m, 4H),
1.79–2.02 (m, 2H), 2.21 (t, 2H, J = 6.9 Hz), 2.32 (t, 2H, J = 7.3 Hz),
2.56 (t, 2H, J = 6.6 Hz), 3.00–3.61 (m, 19H), 3.55 (s, 3H), 4.29–4.38
(m, 1H), 4.78 (s, 2H), 6.61 (br s, 2H), 6.81 (d, 2H, J = 8.6 Hz), 7.43 (br
s, 1H), 7.61–7.91 (m, 3H), 7.72 (d, 2H, J = 8.6 Hz), 8.02 (d, 1H,
J = 7.8 Hz), 8.55 (s, 1H). LC–MS m/z: 742.4 (M+H)+.
4.1.3.14. 2-[N-[N-[N-[4-[[(2,4-Diamino-6-pteridi-
nyl)methyl]methylamino]benzoyl]-
yl)]phenylalanyl]phenylalanylamino]-ethylamine: MTX-(OMe)-
-PhePhe-NH–C2H4–NH2 (precursor of 30). Following the proce-
a-(5-methylglutam-
a
dure for compound 16 with N-t-butoxycarbonylethylenediamine
(NH2–C2H4–NH-Boc) in place of NH2–C10H20O3–NH-Boc, the target
compound (275 mg, 0.342 mmol, 50% in five steps) was obtained
as a yellow powder. 1H NMR (270 MHz, DMSO-d6) d: 1.80–1.96
(m, 2H), 2.20–2.28 (m, 2H), 2.45 (t, 2H, J = 6.6 Hz), 2.70–3.10 (m,
6H), 3.22 (s, 3H), 3.55 (s, 3H), 4.26–4.52 (m, 3H), 4.79 (s, 2H),
6.61 (br s, 2H), 6.82 (d, 2H, J = 8.7 Hz), 7.06–7.21 (m, 10H), 7.46
(br s, 1H), 7.65–7.73 (m, 3H), 7.85 (d, 1H, J = 8.1 Hz), 8.08–8.16
(m, 2H), 8.56 (s, 1H). LC–MS m/z: 805.3 (M+H)+.
4.1.3.9. MTX-(OMe)-a-PhePhe–NH–C10H20O2–NH2 (precursor of
25). Following the procedure for compound 16 with N-t-butoxy-
carbonyl-4,9-dioxa-1,12-dodecanediamine (NH2–C10H20O2–NH-
Boc) in place of NH2–C10H20O3–NH-Boc, the target compound
(221 mg, 0.233 mmol, 14% in five steps) was obtained as a yellow
powder. 1H NMR (400 MHz, DMSO-d6) d: 1.47–1.60 (m, 8H),
1.80–1.95 (m, 2H), 2.20–2.29 (m, 2H), 2.60 (t, 2H), 2.70–3.10 (m,
6H), 3.22 (s, 3H), 3.25–3.50 (m, 8H), 3.54 (s, 3H), 4.25–4.49 (m,
3H), 4.79 (s, 2H), 6.60 (br s, 2H), 6.81 (d, 2H, J = 8.4 Hz), 7.06–
7.20 (m, 10H), 7.45 (br s, 1H), 7.65 (br s, 1H), 7.70 (d, 2H), 7.73
(br t, 1H), 7.83 (d, 1H), 8.10 (d, 1H), 8.11 (d, 1H), 8.55 (s, 1H).
LC–MS m/z: 949.5 (M+H)+.
4.1.3.15. MTX-(OMe)-
31). Following the same procedure for 16 with N-
bonyl-
a-PhePhe-Lys-OMe
(precursor
of
e
-t-butoxycar-
L
-lysine methyl ester (H-Lys(Boc)-OMe) in place of NH2–

1072
A. Homma et al. / Bioorg. Med. Chem. 18 (2010) 1062–1075
C10H20O3–NH-Boc, the target compound (178 mg, 0.l97 mmol, 12%
in five steps) was obtained as a yellow powder. 1H NMR (270 MHz,
DMSO-d6) d: 1.25–1.34 (m, 4H), 1.56–1.69 (m, 2H), 1.75–1.90 (m,
2H), 2.18–2.25 (br t, 2H), 2.50–2.60 (m, 2H), 2.65–3.07 (m, 4H),
3.22 (s, 3H), 3.54 (s, 3H), 3.60 (s, 3H), 4.15–4.60 (m, 4H), 4.79 (s,
2H), 6.63 (br s, 2H), 6.81 (d, 2H, J = 8.7 Hz), 7.00–7.25 (m, 10H),
7.45 (br s, 1H), 7.62 (br s, 1H), 7.69 (d, 2H, J = 8.6 Hz), 7.80 (d,
1H), 8.05 (d, 1H), 8.16 (d, 1H), 8.30 (d, 1H), 8.56 (s, 1H). LC–MS
m/z: 905.4 (M+H)+.
1H NMR (500 MHz, D2O) d: 1.67 (m), 1.79 (m), 1.84–1.94 (m),
2.02 (br s), 2.12–2.20 (m), 2.59 (m), 2.77 (m), 2.91 (m), 2.99 (m),
3.12–3.25 (m), 3.35 (br s), 3.49 (br s), 3.51 (br s), 3.57 (br s), 3.71
(br s), 3.83 (br s) 4.18 (t), 4.45 (br d), 4.55 (br d), 4.88 (d), 4.96
(d), 6.76 (d), 6.95–7.10 (m), 7.72 (d), 8.68 (s).
4.1.6. MTX-a-PheGly-NHC10H20O3NH-HA (18)
Following the general procedure, sodium hyaluronate (500 mg,
ca. 2300 kDa) and the compound obtained in 4.1.3.2. (27.1 mg,
0.031 mmol) provided an aqueous solution of conjugate 18. After
sterilization process, conjugate 18 with a Mw of ca. 1860 kDa
and binding ratio of 1.4% was obtained.
4.1.4. General procedure for the synthesis of conjugates 17–31
In the following, ‘water’ refers to extra pure water and HA was
treated under sterile conditions.
A solution of 3-hydroxy-3,4-dihydro-4-oxo-1,2,3-benzotriazine
(HOOBt) (0.125 mmol) and amines including 16 (0.031 mmol) in
20 mL of water–THF (1:1) solvent was added to a suspension of so-
dium hyaluronate (500 mg, Mw of ca. 2300 kDa) in 10 mL of THF
solution. To the mixture, tris[2-(2-methoxyethoxy)ethyl]amine
(TDA-1, 0.094 mmol) dissolved in 10 ml of water–THF (1:1) was
added, followed by stirring at 5 °C. After 30 min, EDC (0.125 mmol)
dissolved in water (10 mL) was added to the mixture, followed by
stirring at 5 °C for 20 h. A aqueous solution of 0.09 N NaOH solu-
tion (220 mL) was added to the reaction mixture, followed by stir-
ring at 5 °C for 3.5 h. 1 N HCl (20 mL) was added to the reaction
mixture for neutralization, then NaCl solution (9 g in 45 mL of
water) was added, followed by dropwise addition of ethanol
(600 mL). The precipitate which formed was separated by centrifu-
gation. The precipitate was dissolved in water (40 mL) to provide
an aqueous solution of the HA–MTX conjugate.
4.1.7. MTX-a-Phe-NH–C10H20O3–NH-HA (19)
Following the general procedure, sodium hyaluronate (500 mg,
ca. 2300 kDa) and the compound obtained in 4.1.3.3. (25.4 mg,
0.031 mmol) provided an aqueous solution of conjugate 19. After
sterilization, conjugate 19 with a Mw of ca. 1790 kDa and binding
ratio of 1.7% was obtained.
4.1.8. MTX-a-Glu-NH–C10H20O3–NH-HA (20)
Following the general procedure, sodium hyaluronate (500 mg,
ca. 2300 kDa) and the compound obtained in 4.1.3.4 (25.2 mg,
0.031 mmol) provided an aqueous solution of conjugate 20. After
sterilization, conjugate 20 with a Mw of ca. 1830 kDa and a binding
ratio of 1.5% was obtained. 1H NMR (500 MHz, D2O) d: 1.57 (m),
1.77 (m), 2.02 (br s), 2.25 (m), 2.37 (t), 3.24 (s), 3.25 (s), 3.35 (br
s), 3.51 (br s), 3.56 (br s), 3.71 (br s), 3.83 (br s), 4.13 (m), 4.22
(m), 4.36 (m), 4.46 (br s), 4.55 (br s), 4.91 (s), 6.94 (d), 7.76 (d),
8.66 (s), 8.68 (s).
To the conjugate solution, NaCl (6 g) dissolved in 160 mL of
water and 400 mL of ethanol were added. The precipitate which
formed was separated by centrifugation and dissolved in 500 mL
of water, to which NaCl (15 g) was added before filtration using a
The terms in italics indicate minor signals. From the signals,
conjugate 20 was deduced to be a mixture of
(minor) isomers.
a-(major) and c-
0.45
lm filter (Stervex HV: Millipore). Then ethanol (1000 mL)
was aseptically added dropwise. The precipitate was filtered and
dried in vacuo. The conjugate was dissolved in 40 mL of phosphate
buffer solution (2 mM sodium phosphate, 154 mM NaCl, pH 7.2) to
provide a sterile aqueous solution of the HA–MTX conjugate. The
Mw and the binding ratio of MTX were determined by GPC. The
Mw was calculated using a standard curve generated from HA
solutions having various Mws (340, 850, 1440, 1760 and
2200 kDa). The binding ratio of MTX was calculated by measuring
the ultraviolet absorption (259 nm). The procedure for calculating
the binding ratio of MTX was as follows: The concentration of HA
in the conjugate was determined by the RI peak area relative to
the standard HA solution. Next, the HA concentration was con-
verted to an equivalent concentration of a disaccharide unit with
one carboxyl group. The concentration of MTX in the conjugate
was determined by the UV peak area relative to the standard
MTX solution. The binding ratio of MTX was obtained by dividing
the concentration of MTX by the concentration of the disaccharide
unit.
4.1.9. MTX-a-Trp-NH–C10H20O3–NH-HA (21)
Following the general procedure, sodium hyaluronate (500 mg,
ca. 2300 kDa) and the compound obtained in 4.1.3.5 (26.6 mg,
0.031 mmol) provided an aqueous solution of conjugate 21. After
sterilization, conjugate 21 with a Mw of ca. 1390 kDa and binding
ratio of 1.9% was obtained.
4.1.10. MTX-a-Tyr-NH–C10H20O3–NH-HA (22)
Following the general procedure, sodium hyaluronate (500 mg,
ca. 2300 kDa) and the compound obtained in 4.1.3.6 (25.9 mg,
0.031 mmol) provided an aqueous solution of conjugate 22. After
sterilization, conjugate 22 with a Mw of ca. 1760 kDa and binding
ratio of 1.7% was obtained.
4.1.11. MTX-a-Ile-NH–C10H20O3–NH-HA (23)
Following the general procedure, sodium hyaluronate (500 mg,
ca. 2300 kDa) and the compound obtained in 4.1.3.7 (24.3 mg,
0.031 mmol) provided an aqueous solution of conjugate 23. After
sterilization, conjugate 23 with a Mw of ca. 1620 kDa and binding
ratio of 1.7% was obtained.
4.1.5. MTX-a-PhePhe-NH–C10H20O3–NH-HA (17)
Following the general procedure, an aqueous solution of conju-
gate 17 was prepared from sodium hyaluronate (500 mg, ca.
2300 kDa) and compound 16 (29.9 mg, 0.031 mmol). The first lot
after sterilization had a Mw of ca. 1980 kDa and binding ratio of
MTX of 1.4%. The second lot after sterilization had a Mw of ca.
1910 kDa and binding ratio of 1.3%. The third lot after sterilization
had a Mw of ca. 1380 kDa and binding ratio of 1.3%. The fourth lot
after sterilization had a Mw of ca. 2190 kDa and binding ratio of
1.3%.
4.1.12. MTX-a-bAla-NH–C10H20O3–NH-HA (24)
Following the general procedure, sodium hyaluronate (500 mg,
ca. 2300 kDa) and the compound obtained in 4.1.3.8 (23.0 mg,
0.031 mmol) provided an aqueous solution of conjugate 24. After
sterilization, conjugate 24 with a Mw of ca. 1430 kDa and binding
ratio of 1.7% was obtained.
4.1.13. MTX-a-PhePhe-NH–C10H20O2–NH-HA (25)
When 800 or 320 kDa of sodium hyaluronate was used, the con-
jugate which Mw of ca. 800 or 330 kDa and binding ratio was 1.4%
or 1.1% was obtained, respectively.
Following the general procedure, sodium hyaluronate (500 mg,
ca. 2300 kDa) and the compound obtained in 4.1.3.9 (29.4 mg,
0.031 mmol) provided an aqueous solution of conjugate 25. After

A. Homma et al. / Bioorg. Med. Chem. 18 (2010) 1062–1075
1073
sterilization, conjugate 25 with a Mw of ca. 1550 kDa and binding
ratio of 1.7% was obtained.
ethyl acetate. The product was dried in vacuo and the desired
dipeptide ester was obtained (149 g, quant.). To the residue,
745 mL (5 v/w) of acetonitrile was added. At 0 °C, concd. H2SO4
(37.3 mL, 0.25 v/w) was added dropwise to the solution. After stir-
ring for 5 h at 0 °C and for 15 min at 25 °C, H2O 1.49 L (10 v/w) was
added dropwise. The precipitate which formed was filtered,
washed with 749 mL (5 v/w) of water and dried at 30 °C. The de-
sired dipeptide 32 (132 g, quant.) was obtained as a white solid.
1H NMR (300 MHz, DMSO-d6) d: 1.16 (t, 2H, J = 7.2 Hz), 1.66–1.92
(m, 2H), 2.27 (t, 2H, J = 8.0 Hz), 2.52–3.08 (m, 2H), 4.00–4.07 (m,
3H), 4.37–4.44 (m, 1H), 5.00 (s, 2H), 7.18–7.41 (m, 10H), 7.40 (d,
1H, J = 8.1 Hz), 8.12 (d, 1H, J = 7.8 Hz). LC–MS m/z: 457.0 (M+H)+.
4.1.14. MTX-a-PhePhe-NH–C8H16O2–NH-HA (26)
Following the general procedure, sodium hyaluronate (500 mg,
ca. 2300 kDa) and the compound obtained in 4.1.3.10 (28.5 mg,
0.031 mmol) provided an aqueous solution of conjugate 26. After
sterilization, conjugate 26 with a Mw of ca. 1620 kDa and binding
ratio of 1.6% was obtained.
4.1.15. MTX-a-PhePhe-NH–C6H12O2–NH-HA (27)
Following the general procedure, sodium hyaluronate (500 mg,
ca. 2300 kDa) and the compound obtained in 4.1.3.11 (27.7 mg,
0.031 mmol) provided an aqueous solution of conjugate 27. After
sterilization, conjugate 27 with a Mw of ca. 1620 kDa and binding
ratio of 2.0% was obtained.
4.1.20.2. Z-Glu(OEt)PhePhe-OH (33). To
a solution of 100 g
(219 mmol) of 32, H-Phe-OtBu (56.3 g, 219 mmol, 1.0 equiv) and
HOBt (33.5 g, 219 mmol, 1.0 equiv) in THF–H2O (1 L and 100 mL,
10 v/w and 1 v/w) was added 24.4 g (241 mmol, 1.1 equiv) of
NMM. At 5 °C, EDC (43.5 g, 227 mmol, 1.04 equiv) was added and
the reaction mixture was stirred for 1 h then warmed to 25 °C.
After 3 h, 1 L (10 v/w) of H2O and 1 L (10 v/w) of ethyl acetate were
added to the reaction mixture. The organic layer was washed with
400 mL of 5% NaHCO3 aqueous solution (1 L, 10 v/w), 10% citric
acid/20% NaCl aqueous solution (1 L and 100 mL, 10 and 1 v/w)
and 900 mL (9 v/w) of 20% NaCl aqueous solution then dried over
MgSO4 (50 g, 0.5 w/w) and evaporated. The product was dissolved
in 1 L (10 v/w) of acetonitrile and evaporated in order to remove
ethyl acetate. The desired dipeptide ester was obtained (144 g,
quant.). To the residue was added 720 mL (5 v/w) of acetonitrile.
At 0 °C concd H2SO4 (36.0 mL, 0.25 v/w) was added dropwise to
the solution. The reaction mixture was kept at 0 °C for 0.5 h and
25 °C for 1.5 h. To the suspension, isopropyl ether (2.16 L, 15 v/
w) was added dropwise. After stirring for 1 h, the precipitate was
filtered and washed with 720 mL (5 v/w) of isopropyl ether. The
precipitate was dried at 30 °C in vacuo and the desired tripeptide
33 (132 g, quant.) was obtained as a white solid. 1H NMR
(300 MHz, DMSO-d6) d: 1.16 (t, 3H, J = 7.2 Hz), 1.60–1.83 (m, 2H),
2.17 (t, 2H, J = 7.7 Hz), 2.69–3.17 (m, 4H), 3.94–4.06 (m, 3H),
4.38–4.48 (m, 1H), 4.54 (br, 1H), 4.96–5.06 (m, 2H), 7.20–7.39
(m, 16H), 7.88 (d, 2H, J = 8.1 Hz), 8.32 (d, 1H, J = 8.1 Hz). LC–MS
m/z: 604.0 (M+H)+.
4.1.16. MTX-a-PhePhe-NH–C4H8O–NH-HA (28)
Following the general procedure, sodium hyaluronate (500 mg,
ca. 2300 kDa) and the compound obtained in 4.1.3.12 (26.3 mg,
0.031 mmol) provided an aqueous solution of conjugate 28. After
sterilization, conjugate 28 with a Mw of ca. 1490 kDa and binding
ratio of 1.9% was obtained.
4.1.17. MTX-a-PhePhe-NH–C5H10–NH-HA (29)
Following the general procedure, sodium hyaluronate (500 mg,
ca. 2300 kDa) and the compound obtained in 4.1.3.13 (26.3 mg,
0.031 mmol) provided an aqueous solution of conjugate 29. After
sterilization, conjugate 29 with a Mw of ca. 1960 kDa and binding
ratio of 1.2% was obtained.
4.1.18. MTX-a-PhePhe-NH–C2H4–NH-HA (30)
Following the general procedure, sodium hyaluronate (500 mg,
ca. 2300 kDa) and the compound obtained in 4.1.3.14 (25.0 mg,
0.031 mmol) provided an aqueous solution of conjugate 30. The
first lot, after sterilization, had a Mw of ca. 1860 kDa and binding
ratio of 2.1%. The second lot, after sterilization, had a Mw of ca.
2180 kDa and binding ratio of 1.9%. The third lot, after sterilization,
had a Mw of ca. 2090 kDa and binding ratio of 1.3%.
When 320 kDa of sodium hyaluronate was used, the conjugate
which Mw of ca. 340 kDa and binding ratio was 1.9% was obtained.
4.1.20.3. Z-Glu(OEt)PhePhe-NH–C2H4–NH-Boc (34). To a solu-
tion of 60.4 g (100 mmol) of 33, NH2–C2H4–NH-Boc (16.8 g,
105 mmol, 1.05 equiv) and HOBt (15.3 g, 100 mmol, 1.0 equiv) in
DMF (600 mL, 10 v/w) was added 11.1 g (110 mmol, 1.1 equiv) of
NMM at 25 °C. The solution was cooled to 5 °C and EDC (20.1 g,
105 mmol, 1.05 equiv) added. After stirring for 4 h at 5 °C,
660 mL (11 v/w) of H2O was added dropwise and stirred for
30 min at 25 °C. The precipitate which formed was filtered and
washed with 600 mL (10 v/w) of water and dissolved in 1.2 L (20
v/w) of ethanol at the temperature of reflux. After stirring for
1.5 h, the solution was cooled to 25 °C and stirred for 2 h. The pre-
cipitate which formed was filtered, washed with 300 mL (5 v/w) of
ethanol and dried at 30 °C in vacuo. The product 34 (64.6 g,
86.6 mmol, 86.6%, 97.8% purity) was obtained as a white solid.
1H NMR (300 MHz, DMSO-d6) d: 1.16 (t, 3H, J = 7.2 Hz), 1.36 (s,
9H), 1.62–1.84 (m, 2H), 2.18 (t, 2H, J = 7.8 Hz), 2.72–3.10 (m, 8H),
3.91–4.02 (m, 1H), 4.02 (q, 2H, J = 7.2 Hz), 4.35–4.54 (m, 2H),
5.00–5.06 (m, 2H), 6.68 (br, 1H), 7.20–7.34 (m, 16H), 7.84–7.91
(m, 2H), 8.20 (d, 1H, J = 8.1 Hz). LC–MS m/z: 746.0 (M+H)+, 768.2
(M+Na)+.
4.1.19. MTX-a-PhePhe-Lys-HA (31)
Following the general procedure, sodium hyaluronate (500 mg,
ca. 2300 kDa) and the compound obtained in 4.1.3.15 (28.1 mg,
0.031 mmol) provided an aqueous solution of conjugate 31. After
sterilization, conjugate 31 with a Mw of ca. 1720 kDa and binding
ratio of 1.4% was obtained.
4.1.20. Synthesis of 32 for industrial improvement
4.1.20.1. Z-Glu(OEt)Phe-OH (32). To
(290 mmol) of N-carbobenzoxy- -glutamic acid-
Glu(OEt)-OH), 74.8 g (290 mmol, 1.0 equiv) of
tert-butyl ester hydrochloride
(H-Phe-OtBuꢂHCl),
a
solution of 90.0 g
-ethyl ester (Z-
-phenylalanine-
44.4 g
L
c
L
(290 mmol, 1.0 equiv) of HOBt in 900 mL (10 v/w) of THF and
90 mL (1 v/w) of H2O was added 32.3 g (319.0 mmol, 1.1 equiv)
of NMM. At 5 °C, EDC (57.9 g, 302 mmol, 1.04 equiv) was added
and the reaction mixture was stirred for 1 h at 5 °C then warmed
to 25 °C. After 3 h, 900 mL (10 v/w) of H2O and 900 mL (10 v/w)
of ethyl acetate were added to the reaction mixture and stirring
for 5 min. The organic layer was washed with 900 mL of 5% NaH-
CO3 aqueous solution (10 v/w), 10% citric acid/20% NaCl aqueous
solution (900 mL and 90 mL, 10 and 1 v/w) and 810 mL (9 v/w)
of 20% NaCl aqueous solution then dried over magnesium sulfate
(45 g, 0.5 w/w) and evaporated. The product was dissolved in
900 mL (10 v/w) of acetonitrile and evaporated in order to remove
4.1.20.4. H-Glu(OEt)PhePhe-NH–C2H4–NH-Boc (35). To a solu-
tion of 55.0 g (73.7 mmol) of 34 in DMA (550 mL, 10 v/w) under
nitrogen atmosphere were added 14.9 mL (0.27 v/w) of concd
HCl and 11 g (0.2 w/w) of 10% PdꢀC (containing 50% H2O). After

1074
A. Homma et al. / Bioorg. Med. Chem. 18 (2010) 1062–1075
stirring under hydrogen atmosphere for 2 h, the reaction mixture
was filtrated through Celite to remove palladium reagent and
washed with 550 mL (10 v/w) of ethyl acetate. To the filtrate was
added 1.65 L (30 v/w) of ethyl acetate dropwise. After stirring for
1 h at 5 °C, the precipitate which formed was filtered, washed with
275 mL (5 v/w) of ethyl acetate and dried in vacuo. Compound 35
was obtained as a HCl salt (43.9 g, 67.7 mmol, 91.8%, 98.0% purity).
1H NMR (300 MHz, DMSO-d6) d: 1.17 (t, 3H, J = 7.1 Hz), 1.37 (s, 9H),
1.89–2.00 (m, 2H), 2.32–2.39 (m, 2H), 2.78–3.10 (m, 8H), 3.75 (t,
1H, J = 6.03 Hz), 4.05 (q, 2H, J = 7.1 Hz), 4.44 (q, 1H, J = 6.03 Hz),
4.56–4.62 (m, 1H), 6.71 (br, 1H), 7.15–7.27 (m, 10H), 7.94 (br,
with 1
and 493
l
L cathepsin L (0.3 mg/mL, 1584 mU/mg, from Calbiochem)
L of acetate buffer containing 5 mM reduced glutathione
l
and 1 mM EDTA. Each enzyme mixture was incubated at 37 °C for
24 h and analyzed by LC–MS (see the general analysis method).
4.2.2. In vitro experiments
HFLS (Cell Applications) was seeded at 5000 cells/well on a 96-
well plate (Falcon) and cultured for 3 h in Iscove’s modified Dul-
becco’s medium (IMDM) containing 5% FBS and 1X Antibiotic-Anti-
mycotic (GIBCO BRL). After cellular attachment, TNF-
(recombinant human TNF- , R&D Systems) (final concentration:
a
a
1H), 8.12 (br, 2H), 8.42 (d, 1H,
J
8
8.00 Hz), 8.57 (d, 1H,
10 ng/mL) and each HA–MTX conjugate at each concentration
was added, followed by cultivation for 5 days. Two days before
the end of culture, 37 kBq/well of [3H]-deoxyuridine was added
to the cells (MORAVEK), followed by determining the uptake quan-
tity (radioactivity) of [3H]-deoxyuridine using a scintillation coun-
ter. Cells were recovered by unsticking with 0.05% trypsin-0.2%
EDTA.
Radioactivity was calculated as a relative value (% of control),
using, as the control, radioactivity in the group of cells cultured
without any added test substance. Since the concentration of a free
carboxyl group is 2.49 ꢁ 10ꢀ3 mol/L (1 g/401/L; 401 is the Mw of
N-acetylglucosamine + glucuronic acid) for each 1 mg/mL of HA,
the MTX concentration in each HA–MTX was calculated by multi-
plying the value by the conjugation rate of MTX. (For 1 mg/mL of
HA–MTX conjugate with a conjugation rate of MTX of 1%, the con-
centration of MTX was 2.49 ꢁ 10ꢀ5 mol/L.) The value obtained was
used to calculate the activity of cell proliferation inhibition (the
IC50 value) by a 4-parameter logistic method using analysis soft-
ware (GraphPad Prism 3.02).
J = 8.0 Hz). LC–MS m/z: 612.1 (M+H)+, 634.2 (M+Na)+.
4.1.20.5. MTX-(OEt)-a-Phe-Phe-NH–C2H4–NH2 (36). To a solu-
tion of 13.3 g (41.0 mmol, 0.95 equiv) of DAMPA in N-methylpi-
peridone (NMP, 280 mL, 10 v/w) was added 28.0 g (43.2 mmol)
of 35, 1.99 g (13.0 mmol, 0.3 equiv) of HOBt, 6.55 g (64.8 mmol,
1.5 equiv) of NMM and 10.8 g (56.2 mmol, 1.3 equiv) of EDC. After
stirring for 4 h at 55 °C, the reaction mixture was cooled to rt and
420 mL (15 v/w) of water added. After stirring for 0.5 h, the precip-
itate which formed was filtered and washed with 420 mL of water.
The precipitate was dried in vacuo and the desired compound was
obtained as a yellow solid (36.7 g, 39.9 mmol, 92.4%). LCMS m/z:
919.2 (M+H)+.
This solid (18.0 g, 19.6 mmol) was suspended in 180 mL (10 v/
w) of 7 N HCl–ethanol and stirred for 1 h at rt. To the suspension
was added 360 mL of ethanol (20 v/w) and 180 mL of DMF (10 v/
w) and the reaction mixture became clear. After 1.5 days, the pre-
cipitate which was formed was filtered and washed with 90 mL (5
v/w) of ethanol. The titled compound 36 was dried under reduced
pressure at 30 °C and obtained as a yellow solid (3ꢂHCl salt, 12.4 g,
13.4 mmol, 68.2%, 97.8% purity). 1H NMR (300 MHz, DMSO-d6) d:
1.14 (t, 3H, J = 7.1 Hz), 1.78–1.93 (m, 2H), 2.19–2.24 (m, 2H),
2.49–2.55 (m, 2H), 2.71–3.06 (m, 6H), 3.22 (s, 3H), 4.01 (q, 2H,
J = 7.1 Hz), 4.26–4.52 (m, 3H), 4.79 (s, 2H), 6.58 (br, 2H), 6.82 (d,
2H, J = 9.1 Hz), 7.06–7.23 (m, 10H), 7.46 (br, 1H), 7.65 (br, 1H),
7.71 (d, 2H, J = 9.0 Hz), 7.75–7.78 (m, 1H), 7.83 (d, 1H, J = 8.0 Hz),
8.08–8.14 (m, 2H), 8.50 (s, 1H). LC–MS m/z: 820.2 (M+H)+.
4.2.3. In vivo experiments
Six-week-old male LEW/Crj rats were purchased from Charles
River Laboratories Japan, Inc. The rats were sensitized with
0.5 mL of an emulsion prepared from a 2-mg/mL mBSA (Calbio-
chem) aqueous solution and an equal amount of Freund’s complete
adjuvant (Difco) injected into the flank at 21 and 14 days before
inducing arthritis. The arthritis was induced by injecting 50 lL of
a 2-mg/mL mBSA aqueous solution into the right knee joint. The
left knee joint was untreated and served as the control. Knee joint
swelling was assessed by measuring the width of each knee joint
with calipers to define the left-right difference. The width of all
knee joints was measured twice a week from immediately before
inducing arthritis to two weeks after to calculate the area under
the curve (AUC) over time for joint swelling. In addition, the AUC
was calculated using the value relative to the HA-treated control
group (% of control) which HA Mw were in the range of
1900ꢀ2500 kDa. HA and HA–MTX conjugates were administered
4.1.20.6. MTX-a-PhePheNH–C2H4–NH-HA (30, high, middle and
low binding ratio of MTX). Following the general procedure, so-
dium hyaluronate (500 mg, ca. 2300 kDa), compound 36 and TDA-
1 provided an aqueous solution of conjugate 30. The amount of
compound 36 and TDA-1, and the specs of obtained conjugates
were as follows:
Compd 36
TDA-1
Binding ratio (%)
Mw (kDa)
into the right knee joint in an amounts of 50
1 day) and after (7 days) inducing arthritis.
lL before (7 and
(mmol)
(mmol)
0.008
0.020
0.063
0.118
0.105
0.063
0.5
1.3
3.8
2170
2090
1910
At each measurement, the mean and standard deviation of AUC
were calculated to perform an unpaired t-test between each test
substance-treated group and the HA-treated group, and was
judged to be significantly different if the probability level was less
than 5%. Statistical analysis was performed using SAS version 6.12
(SAS Institute Japan).
4.2. Biology
Acknowledgments
4.2.1. Cathepsin fragmentation
L of each compound (40 mM) was mixed with 1
g/ L, 3930 U/mg, from Sigma–Aldrich) and 494
tate buffer (40 mM sodium acetate buffer, pH 5.0) containing
5 mM reduced glutathione and 1 mM EDTA. 5 L of each com-
pound (40 mM) was mixed with 2 L cathepsin D (1.3 g/ L,
381 U/mg, from Sigma–Aldrich) and 493 L of acetate buffer con-
taining 1 mM EDTA. 5 L of each compound (40 mM) was mixed
5
l
l
L cathepsin
The authors thank Prof. Kunio Ogasawara for his helpful sugges-
tions concerning this study. We also thank Ms. Frances Ford for
proofreading the manuscript.
B (0.6
l
l
lL of ace-
l
l
l l
References and notes
l
l
1. Swierkot, J.; Szechinski, J. Pharmacol. Rep. 2006, 58, 473.

A. Homma et al. / Bioorg. Med. Chem. 18 (2010) 1062–1075
1075
2. Hamstra, D. A.; Page, M.; Maybaum, J.; Rehemtulla, A. Cancer Res. 2000, 60, 657.
3. Smal, M. A.; Dong, Z.; Cheung, H. T.; Asano, Y.; Escoffier, L.; Costello, M.;
Tattersall, M. H. Biochem. Pharmacol. 1995, 49, 567.
4. Homma, A.; Sato, H.; Okamachi, A.; Emura, T.; Ishizawa, T.; Kato, T.; Matsuura,
T.; Sato, S.; Tamura, T.; Higuchi, Y.; Watanabe, T.; Kitamura, H.; Asanuma, K.;
Yamazaki, T.; Ikemi, M.; Kitagawa, H.; Morikawa, T.; Ikeya, H.; Maeda, K.;
Takahashi, K.; Nohmi, K.; Izutani, N.; Kanda, M.; Suzuki, R. Bioorg. Med. Chem.
2009, 17, 4647.
16. Putnam, D.; Kopecek, J. Polymer Conjugates with Anticancer Activity; Springer,
1995.
17. Morimoto, H.; Ito, T.; Inoue, K.; Okuno, S. PTC Int. Appl. Patent JP05039306,
1993.
18. Inoue, K.; Susaki, H.; Ikeda, M.; Kuga, H.; Kumazawa, E.; Togo, A. PTC Int. Appl.
Patent WO9746260, 1997.
19. Franzyk, H.; Christensen, M. K.; Jorgensen, R. M.; Meldal, M.; Cordes, H.;
Mouritsen, S.; Bock, K. Bioorg. Med. Chem. 1997, 5, 21.
20. Meyer, J. P.; Davis, P.; Lee, K. B.; Porreca, F.; Yamamura, H. I.; Hruby, V. J. J. Med.
Chem. 1995, 38, 3462.
5. Wen, D. Y. Am. Family Physician 2000, 62, 565.
6. McCarty, M. F.; Russell, A. L.; Seed, M. P. Med. Hypotheses 2000, 54, 798.
7. Adams, M. E.; Lussier, A. J.; Peyron, J. G. Drug Saf. 2000, 23, 115.
8. Mihara, M.; Higo, S.; Uchiyama, Y.; Tanabe, K.; Saito, K. Osteoarthritis Cartilage
2007, 15, 543.
21. Smith, G. K.; Banks, S.; Blumenkopf, T. A.; Cory, M.; Humphreys, J.; Laethem, R.
M.; Miller, J.; Moxham, C. P.; Mullin, R.; Ray, P. H.; Walton, L. M.; Wolfe, L. A.,
3rd J. Biol. Chem. 1997, 272, 15804.
9. Pagnano, M.; Westrich, G. Osteoarthritis Cartilage 2005, 13, 751.
10. Hsieh, Y. S.; Yang, S. F.; Lue, K. H.; Chu, S. C.; Lu, K. H. J. Orthop. Res. 2008, 26, 475.
11. Kotevoglu, N.; Iyibozkurt, P. C.; Hiz, O.; Toktas, H.; Kuran, B. Rheumatol. Int.
2006, 26, 325.
22. Kuefner, U.; Lohrmann, U.; Montejano, Y. D.; Vitols, K. S.; Huennekens, F. M.
Biochemistry 1989, 28, 2288.
23. Castex, C.; Lalanne, C.; Mouchet, P.; Lemaire, M.; Lahana, R. Tetrahedron 2005,
61, 803.
12. Ghosh, P.; Guidolin, D. Semin. Arthritis Rheum. 2002, 32, 10.
13. Bayliss, M. T.; Ali, S. Y. Biochem. J. 1978, 171, 149.
24. Mahmoodi, M.; Sahebjam, S.; Smookler, D.; Khokha, R.; Mort, J. S. Am. J. Pathol.
2005, 166, 1733.
14. Maciewicz, R. A.; Wotton, S. F.; Etherington, D. J.; Duance, V. C. FEBS Lett. 1990,
269, 189.
15. Keyszer, G.; Redlich, A.; Haupl, T.; Zacher, J.; Sparmann, M.; Engethum, U.; Gay,
S.; Burmester, G. R. Arthritis Rheum. 1998, 41, 1378.
25. Roth, A.; Mollenhauer, J.; Wagner, A.; Fuhrmann, R.; Straub, A.; Venbrocks, R.
A.; Petrow, P.; Brauer, R.; Schubert, H.; Ozegowski, J.; Peschel, G.; Muller, P. J.;
Kinne, R. W. Arthritis Res. Ther. 2005, 7, R677.
26. Griffiths, R. J. Agents Actions 1992, 35, 88.