Since the success of cyclosporine A (CsA) in treating organ
rejection in liver transplants of the first patient1 and subsequent
approval for use in 1983, it has been used against transplant
rejection, psoriasis, and autoimmune diseases such as chronic
urticaria and rheumatoid arthritis. Other blockbuster
immunosuppressants, tacrolimus (also known as FK506) and
sirolimus (also known as rapamycin) were first approved for liver
and kidney transplants in 1994 and 1999, respectively. These
three drugs are still the most prescribed immunosuppressive
ammonium salt 10 with methyl iodide. It afforded compound 11
as a mixture with its tropolonic regioisomer on Hoffman
degradation procedure with Pd on C and sodium hydroxide.13
Then, methylation of free alcohol followed by substitution with
sodium thiomethoxide afforded the desired compound 12 void of
7-amide group. Scheme 2 describes the synthesis of ring-
truncated analogs. The key precursor 16 of ring-truncated analogs
was prepared from 5-iodotropolone, which was developed by
Ebisawa et al,14 by O-methylation with MeI and K2CO3 in
acetone. 19, 20, 23 and 24, which are ring-truncated analogs with
2 or 3 carbon-linkers, were synthesized by continuous double
cross coupling reactions. Stille cross coupling reaction of 17 (or
21) with vinyl (or allyl) tributyltin produced intermediates with
terminal alkenes. They afforded desired ring-truncated analogs
on Heck reaction with 16. Cross coupling of 25 with 16 by
Pd(OAc)2 and PPh3 in presence of TBAF15 gave compound 26
which was readily converted to compound 27 by substitution
with sodium thiomethoxide.
3
drugs in the market.2, However, current immunosuppressants
including these three show serious side effects such as renal
toxicity, neurotoxicity, hypertension and pulmonary toxicity.4-6
Despite intensive efforts to overcome rejection in organ
transplants and autoimmune diseases, there still remains the need
for more effective and safer immunosuppressants.
Our group has discovered a new class of immunosuppressant
derived from the natural product colchicine, which inhibits
microtubule polymerization by binding to tubulin.7, 8 Colchicine
(3) was approved as a therapeutic agent for the treatment of acute
gout flares, prophylaxis, and familial Mediterranean fever.
Compound 1, a first-generation colchicine analog (Figure 1)
exhibited potent activity on in vitro MLR and in vivo skin-
allograft model but revealed poor pharmacokinetic properties
(F~0%) in vivo.9 Compound 2, a second-generation colchicine
analog (Figure 1), which was developed by the study of
First, we evaluated the inhibitory activity of analogs on LPS-
induced B cell- and Con A-induced T cell proliferation. As
shown in figure 3, compounds (7 and 12) eliminating 7-
carboxamide group exhibited high inhibitory activity on both B-
and T cell proliferation in dose-dependent manners. Compound
26 and 27, which are ring-truncated analogs with a simple
biphenyl structure, showed a moderate inhibitory activity but
those are less active than tricyclic 7 or 12. However, ring-
truncated compounds 19, 20, 23 and 24 with a carbon linker
exhibited the loss of activity as described in figure 2 of the
supporting information. We did not try to synthesize the analogs
with 10-thiomethoxy group of these inactive compounds on the
basis of our previous SAR report.9 This series of analogs have a
similar tendency in the activity as compared to the previous lead
compound 2, which is that B cell proliferation is more strongly
suppressed than T cell.10 Nitric oxide (NO) is known to
contribute to the acute rejection of the organ-tranplantation16 and
deemed a tissue damaging molecule by activated macrophages in
autoimmune diseases.17 We compared their inhibitory effects of
NO production on bone marrow-derived macrophages which are
primary macrophages, and transformed RAW 264.7 cells. (Figure
4) 7 and 12 suppressed NO production on RAW 264.7 cells but
not bone marrow-derived macrophages, suggesting the
compounds (especially 12) just suppress NO production of
proliferating cells. NO is produced from diverse NO synthases
(NOSs) which have been found in macrophages, dendritic cells,
NK cells, B- and T cells. NOSs are regulated by cytokines such
as IL-12 and IFN-γ.18 Both 7 and 12 showed dose-dependent
suppression of IL-12 on bone marrow-derived dendritic cells
without affecting cell viability. (Figure 4)
structure-activity
relationships
and
solubility-activity
relationships between the first generation analogs, exhibited
improved pharmacokinetic properties (F = 67.3%) and effective
in vivo activity on the Zymosan A-induced arthritis model,
Carrageenan-induced edema model and the local lymph node
assay (LLNA).10 However, compound 2 turned out to have severe
toxicity including intestinal toxicity, loss of spleen weight
suggesting immunotoxicity, and loss of weight during in vivo
tests. (Supporting information, Figure 1) Thus, we needed to
discover less toxic compounds that retained activity compared to
the previous lead compound 2. To achieve this goal, we
attempted the lead optimization of compound 2 to produce less
toxic analogs. Unfortunately, as shown in our previous structure-
activity relationships of colchicine analogs, alteration of the 7-
membered tropolone ring with 10-thiomethoxy group induced
significantly decreased activity or loss of activity. (data not
shown) Furthermore, it was undesirable to alter and optimize the
2-methylcyclopropane-1-carboxamide group at C7-position
because it had been selected to improve aqueous solubility and
PK profiles.10 Based on previous SAR studies and preliminary
optimization of compound 2, we decided to design structurally
different compounds from colchicine analogs which have 7-
carboxamide groups as a side chain. In our efforts to develop a
novel series of immunosuppressants from colchicine, all of our
analogs had retained the 7-carboxamide group which included
aliphatic, aromatic and pseudo-aromatic groups. For this study,
we investigated whether the deletion of the 7-carboxamide group
or further simplified analogs which have no middle 7-membered
ring would exhibit high activity with reduced toxicity compared
to compound 2. (Figure 2) Herein, we describe the synthesis and
biological evaluation of structurally simplified compounds from
colchicine as an immunosuppressant.
Next, we further investigated the in vivo immunosuppressive
efficacy of 7 and 12 which are the most potent compounds in this
study. In the adapted immune system, naïve B lymphocytes,
which have not yet recognized an antigen via their B cell
receptors, are activated by many protein antigens called T-
dependent antigens. After that, they proliferate, differentiate, and
mount an antibody response against T-dependent antigens.19 The
spleen cells from mice were immunized with SRBCs and treated
with compound 7 or 12, and then, the total specific (anti SRBC)
antibody formation was accessed.20 Both compounds exhibited
dose-dependent suppression of in vivo T-dependent antibody
response, which implies they strongly suppress B cell-mediated
immune response and it’s exactly correlated with the in vitro data.
(Figure 5a) We also evaluated their immunosuppressive effects
on the arthritis and edema model as we tested compound 2 in a
previous report.10 The C57BL/6 mice were treated with 7 and 12
after subcutaneous injection of Zymosan A into the hind footpad
As depicted in scheme 1, Ketone 5 was readily available from
deacetylthiocolchicine 4 using the formation of Schiff’ base
followed by anionic equilibration and hydrolysis procedure
12
developed by Rapoport et al.11,
Reduction with NaBH4
produced racemic alcohol 6 which afforded a deamidated alkene
7 on subsequent methanesulfonylation followed by elimination
with DBU. Hydrolysis of colchicine in 10% sulfuric acid gave a
10-hydroxy free amine 9 which was converted to quaternary
Chang, Dong-Jo
