Rapamycin analogs with reduced systemic exposure

Article abstract of DOI:10.1016/j.bmcl.2005.06.106

The synthesis and biological activities of rapamycin (I) analogs modified at the C-40 position are reported. Emphasis placed on compounds that potentially have an improved safety profile on account of their shorter in vivo half-life when compared with rap

Full text of DOI:10.1016/j.bmcl.2005.06.106

Bioorganic & Medicinal Chemistry Letters 15 (2005) 5340–5343
Rapamycin analogs with reduced systemic exposure
*
Rolf Wagner, Karl W. Mollison, Luping Liu, Cynthia L. Henry,
Teresa A. Rosenberg, Nwe Bamaung, Noah Tu, Paul E. Wiedeman,
Yatsun Or, Jay R. Luly, Benjamin C. Lane, James Trevillyan,
Yung-Wu Chen, Thomas Fey, Gin Hsieh, Kennan Marsh, Merrill Nuss,
Peer B. Jacobson, Denise Wilcox, Richard P. Carlson,
George W. Carter and Stevan W. Djuric
Abbott Laboratories, Global Pharmaceutical Research and Development, Abbott Park, IL 60064, USA
Received 11 May 2005; revised 20 June 2005; accepted 21 June 2005
Available online 26 September 2005
Abstract—The synthesis and biological activities of rapamycin (I) analogs modified at the C-40 position are reported. Emphasis
placed on compounds that potentially have an improved safety profile on account of their shorter in vivo half-life when compared
with rapamycin.
Ó 2005 Elsevier Ltd. All rights reserved.
Rheumatoid arthritis (RA) is an autoimmune disease
afflicting over 2 million individuals in the United States
and 5.7 million worldwide. RA is a painful disease that
results in progressive joint destruction, deformity, and
immobility, and can eventually lead to death. Pain relief
using non-steroidal anti-inflammatory drugs (NSAIDS)
is the most popular treatment for RA. Unfortunately,
this treatment cannot prevent disease progression. Other
disease modifying antirheumatic drugs (DMARDS) are
also currently available, but their use as first-line thera-
py is normally avoided due to slow onset of action, mod-
require demonstration of efficacy and an improved
safety profile with respect to currently available drugs.
Furthermore, this agent should act immediately and
produce complete relief from the disease by arresting
the disease process itself.
HO
40
CH3O
OCH3
O
O
O
HO
N
O
1
O
est efficacy, and frequent toxic side effects. The potent
O
CH3O
immunosuppressant, cyclosporine, has seen increased
use in the treatment of severe RA, despite its publicized
liabilities, demonstrating the marketability of more effi-
HO
O
1
1
cacious agents. Indeed, new classes of macromolecules,
targeting TNF-a and IL-1, have been introduced to the
market with astonishing success. While these com-
pounds demonstrate an improved benefit/risk profile
when compared to DMARDS, they do present patients
with a unique side effect profile, and therefore provide
additional opportunity for the development of alterna-
Rapamycin (1), a secondary fungal metabolite, has
shown comparable efficacy when substituted for cyclo-
sporine in clinical trials for renal transplantation.
3
Moreover, the side-effect profile is reported to be rela-
tively benign compared to that of cyclosporine, with ele-
vated triglyceride and cholesterol levels as the principal
dose limiting side-effects seen in some patients.
However, rapamycin is also beset with an overly long
2
tive treatments. The criteria guiding our search for a
potent immunosuppressant for the treatment of RA
4
half-life of 63 h. Prolonged exposure of patients to drug
*
Corresponding author. Tel.: +1 847 937 1138; e-mail:
rolf.wagner@abbott.com
Enanta Pharmaceuticals Inc., Watertown, MA 02472, USA.
levels becomes problematic once toxicity is encountered.
Patients suffering from autoimmune diseases, such as
RA, historically have not tolerated as much risk as the
0
960-894X/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2005.06.106

R. Wagner et al. / Bioorg. Med. Chem. Lett. 15 (2005) 5340–5343
5341
a
transplant community has been willing to accept. We
Table 1. IC50, nM (n)
therefore viewed rapamycin as an excellent starting
point for synthesizing an improved agent, emphasizing
similar in vivo efficacy, prompt measurable effect, and
a shorter half-life. This communication briefly describes
the synthesis and biological characterization of our most
promising compounds.
Compound FKBP12
binding
MLR Hu
inhibition
MLR Rt
inhibition
b
c
1
3
4
6
1.6 ± 0.9 (14) 1.6 ± 2.2 (72) 2600 ± 1900 (14)
3.3 ± 2.2 (3)
4.8 ± 1.0 (3)
5.8 ± 4.4 (8)
2.0 ± 4.4 (17) 1400 ± 600 (9)
3.3 ± 2.2 (3) 1500 ± 600 (6)
5.9 ± 6.1 (36) 2500 ± 1800 (11)
a
b
c
Data reported as the mean ± SEM for (n) determinations as noted.
Human mixed lymphocyte reaction, serum containing.
Rat mixed lymphocyte reaction, serum free, Lewis rat lymph node
cells stimulated by Brown–Norway splenocytes.
Treatment of rapamycin (1) with trifluoromethanesulf-
onic anhydride and 2,6-lutidine in CH Cl , followed
by filtration through a pad of silica gel (Et O), provided
2
2
2
2
in 95% yield (Scheme 1). The triflate 2 is used immedi-
ately in the subsequent reaction. 40-Epi-tetrazolyl-rap-
amycins 3 (ABT-578) and 4 were obtained in a 2:1
ratio in 34% yield after the reaction of 2 with tetrazole
in isopropyl acetate.
Likewise, analogs 3, 4, and 6, exhibit IC s in the range
50
from 3.3 to 5.8 nM. In vitro T-cell antiproliferative abil-
ity was assessed by comparison of the analogs to 1, in
the human (MLR Hu) and rat (MLR Rt) mixed lym-
phocyte reaction. The ability of all semi-synthetic rapa-
mycin analogs to inhibit alloantigen-induced T-cell
proliferation in the MLR Hu was similar to that of rap-
amycin (1), which measured 1.6 nM. The weakest inhib-
itor was carbamate 6, with a potency of 5.9 nM. All
compounds were much weaker inhibitors in the rat, with
no differentiation shown between them. The low appar-
ent potency with rat cells in vitro has been observed by
Another promising analog, carbamate 6, was synthe-
sized, according to Scheme 2. Rapamycin was treated
with finely ground p-nitrophenylcarbonate and pyridine
at 0 °C, and then warmed to room temperature. After
purification by flash chromatography, pure carbonate
5 was isolated in 73% yield. This key intermediate was
treated with N,O-dimethylhydroxylamine hydrochloride
in pyridine at room temperature, followed by heating to
8
others, but is little understood.
50 °C. Chromatographic purification provided pure 6 in
90% yield.
Having established reasonable efficacy in vitro, the next
stage required the evaluation of the compounds in auto-
immune disease rat models. The delayed-type hypersen-
sitivity (DTH Rt) response to antigen challenge is a
T-cell-mediated inflammatory response that may mimic
the pathologic response to altered self-proteins or micro-
bial antigens involved in RA, and may thus serve as a
Rapamycin (1) exerts its immunosuppressive activity by
the formation of a trimolecular complex, imitating the
first step in the biochemical mechanism of action of
FK506. The drug is required to bind to FKBP12, a pep-
tidyl-prolyl-isomerase, before this molecular partnership
5
6
inhibits the serine–threonine kinase TOR (or RAFT1).
The binding assay and MLR (mixed lymphocyte reac-
9
reasonable RA model (Table 2). Tetrazoles 3 and 4,
7
as well as carbamate 6, demonstrated nearly equal effica-
cy with rapamycin 1, with the weakest compound, 4,
only being fourfold less potent than 1.
tion) assays used below have been described previously.
Rapamycin (1) inhibits FKBP12 with an IC50 of 1.6 nM
(
Table 1).
N
N
X
HO
CH O
CF SO O
N
3
2
b
a
Y
CH O
CH O
3
3
3
1
2
3: X=N, Y=CH
: X=CH, Y=N
4
Scheme 1. Reagents and conditions: (a) i. 2,6-Lutidine (4.2 equiv), CH
2
Cl
2
ꢀ78 °C, ii. (CF
3
SO
0 min, 95%; (b) isopropyl acetate, tetrazole (2 equiv), diisopropylethylamine (2 equiv), RT 18 h, 34%, isomers 3 and 4 separated by silica gel
2
)
2
O added dropwise (2 equiv) ꢀ78 °C 15 min RT
2
chromatography eluting sequentially with hexane, hexane/ether (4:1, 3:1, 2:1, 1:1), ether, hexane/acetone (1:1).
OCH3
pNO PhO
H C
3
N
2
HO
CH O
O
O
a
b
O
O
3
CH O
3
CH O
3
1
5
6
Scheme 2. Reagents and conditions: (a) Bis(p-nitrophenyl)carbonate (2 equiv) 0 °C, pyridine addition, warmed to RT (1 h), 73%; (b) N,O-
dimethylhydroxylamine hydrochloride (4.4 equiv), pyridine, RT (18 h), 50 °C (1 h), 90%.

5342
R. Wagner et al. / Bioorg. Med. Chem. Lett. 15 (2005) 5340–5343
a
Table 2. ED50, mpk/day PO (n) (95% range)
b
c
d
Compound
DTH Rt
EAE Rt
AA Rt
1
3
4
6
0.5 (0.4–0.7) (35)
1.7 (0.9–7.7) (21)
2.0 (1.3–4.5) (7)
0.4 (0.1–1.0) (21)
0.3 (0.2–0.4) (10)
1.2 (0.3–1.7) (10)
NA
0.2 (0.2–0.3) (20)
0.7 (0.5–0.9) (10)
NA
0.5 (0.3–0.7) (10)
0.3 (0.2–0.5) (10)
a
b
c
9
5% confidence limits for n determinations as noted.
Sprague–Dawley rat delayed-type hypersensitivity.
Lewis rat experimental autoimmune encephalomyelitis.
Lewis rat developing adjuvant-induced arthritis.
d
The rat experimental autoimmune encephalomyelitis
EAE Rt) model is used as a model of multiple sclerosis.
EAE is an acute experimental neurological disease med-
The most favorable profiles for 3 and 6 would also in-
clude decreased risk of toxicity for the patient popula-
tion. One aspect of this is a more controllable
exposure of the drug toward the patient, which could
be influenced by an attenuation of some key pharmaco-
kinetic parameter. One can observe that in rats (Table
3), compounds 3 and 6 both have dramatically shorter
half-lives than rapamycin (1), as well as lower Cmax
and AUC values upon a single oral drug dosing of
2.5 mpk. The oral bioavailability (F) of both new com-
pounds is significantly lower than 1, 12.1% (3) and
(
+
iated by CD4 T cells, whose target is the central ner-
1
0
vous system. The disease is artificially induced by
subcutaneously injecting the rat with homogenized isol-
ogous spinal cord in modified complete FreundÕs adju-
vant. This disease is effectively abrogated by the three
agents tested, with rapamycin (1) being most potent with
an ED50 of 0.3 mpk/day and 3 being the least potent at
1
.2 mpk/d.
6
.2% (6) versus 43.2% for 1. Nevertheless, sufficient lev-
The pathogenesis of human rheumatoid arthritis and
adjuvant arthritis in the rat (AA Rt) shares several
els of each drug are achievable in the DTH, EAE, and
AA rat models to demonstrate a robust biological effect,
and it should be noted that the oral bioavailability of an
1
1
important features. Among these are included T-cell
mediation of the disease, deposition of fibrin in the
joints, leukocyte influx into the synovium, and destruc-
tion of cartilage and bone. While this pathogenic process
requires as long as 30–50 years in human RA, the AA Rt
model accelerates the disease progression to end-stage
disease after a mere 30–40 days post-antigen administra-
tion. Using a prophylactic approach, rapamycin (1) effec-
tively blocks the development of AA with an ED50 of
1
3
oral solution of 1 is advertised to be 14% in humans. It
is gratifying that the rat PK model has been a reliable
comparative predictor of the PK behavior of 1 versus
3 in human subjects. Indeed, in a multiple dose escala-
tion study, the half-life of 3 in humans was determined
to be 25–31 h, significantly shorter than the 63 h report-
1
4
ed for 1.
0
0
.2 mpk/day, whereas 3 and 6 show ED Õs of 0.7 and
Unsurprisingly, it was observed that rapamycin (1) was
a metabolite of 6. In a separate experiment, oral dosing
5
0
.3 mpk/day, respectively. Abrogation of the soft tissue
1
2
destruction in AA Rt is shown in Figure 1. All three
models, DTH Rt, EAE Rt, and AA Rt, therefore estab-
lish the efficacy of rapamycin (1), as well as 3 and 6, in
autoimmune disease models, despite the apparent weak
rat whole-cell activity shown in Table 1. Since we have
demonstrated efficacy in a species with weak whole cell
activity, we are optimistic that efficacy can be demon-
strated in human subjects since the whole cell humam
MLR response is more sensitive to all three compounds.
Table 3. Male Sprague–Dawley rat pharmacokinetic parameters
(2.5 mpk PO, solution)
Compound
n
t1/2 (h),
C
max
T
max AUC0–1
F (%)
(
ng/mL) (h)
(ng h/mL)
1
3
6
7
3
3
33.4
7.9
50.8
46.7
29.7
0.5
0.4
0.4
559
163
132
43.2
12.1
6.2
8.7
Normal
Adjuvant Control
#1
#3
#6
Figure 1. Magnetic resonance images showing soft tissue changes in sagittal sections from left hindpaws from normal, vehicle dosed 15 day adjuvant
controls, and rats dosed from day 0 to 15 with compounds 1 (1 mpk), 3 (3 mpk) and 6 (1 mpk).

R. Wagner et al. / Bioorg. Med. Chem. Lett. 15 (2005) 5340–5343
5343
Table 4. Male Sprague–Dawley rat pharmacokinetic parameters
2.5 mpk PO, solution)
References and notes
(
1
2
3
4
. Chikanza, I. C.; Jawed, S.; Naughton, D.; Blake, D. R.
J. Pharm. Pharmacol. 1998, 50, 357.
. Fleischman, R. M.; Iqbal, I.; Stern, R. L. Expert Opin.
Drug Saf. 2004, 3, 391.
. Groth, C. G. Presented at the ASTP 16th Annual Meeting,
Chicago, IL, 1997.
. Zimmerman, J. J.; Kahan, B. D. J. Clin. Pharmacol. 1997,
37, 405; Ferron, G. M.; Mishina, E. V.; Zimmerman, J. J.;
Jusko, W. J. Clin. Pharmacol. Ther. 1997, 61, 416;
Mahalati, K.; Kahan, B. D. Clin. Pharmacokinet. 2001,
40, 573.
Compound
C
max (ng/ml)
T
max (h)
AUC0–1 (ng h/mL)
6
1
23.2
9.9
0.3
1.0
127
83
of rats with 2.5 mpk of 6 revealed that the maximal con-
centration of 1 was significantly lower than the Cmax for
(Table 4). Yet since the oral half-life of 1 is significant-
ly longer than 6, the AUCs are very similar, yet the sum
of both is still much lower than the AUC generated by
an oral administration of 2.5 mpk of 1 alone. Whether
or not this attenuated exposure of the animal to 1 via
the active Ôpro-drugÕ 6 translates to a less insulting side
effect profile would need to be established by clinical tri-
als in human patients.
6
5
6
. Wagner, R.; Rhoades, T. A.; Or, Y. S.; Lane, B. C.; Hsieh,
G.; Mollison, K. W.; Luly, J. R. J. Med. Chem. 1998, 41,
1764.
. Brown, E. J.; Albers, M. W.; Shin, T. B.; Ichikawa, K.;
Keith, C. T.; Lane, W. S.; Schreiber, S. L. Nature 1994,
369, 756; Sabatini, D. M.; Erdjument, B. H.; Lui, M.;
Tempst, P.; Snyder, S. H. Cell 1994, 78, 35.
In summary, we have identified two promising analogs
of rapamycin that show impressive efficacy in several
models of autoimmune disease, comparing favorably
in efficacy to rapamycin. Furthermore, it is possible that
these agents may offer a safety advantage with respect to
exposure toward parent drug. Indeed, it has been found
that significant circulating levels of rapamycin were
found in rats that were dosed with 6. While 6 is several
fold less potent than 1, it has yet to be shown that sup-
pression of the maximal concentrations of either com-
pound (1 or 6) via this strategy can result in the
reduction of adverse side effects. Further characteriza-
tion of drug candidate 3 has already occurred in human
subjects and will be the subject of future reports, espe-
cially in regard to its use on drug-coated stents. The
use of 3 in autoimmune diseases, however, and contin-
uing toxicological studies, is clearly a tantalizing pros-
pect. Establishment of efficacy in rat autoimmune
disease models has provided a cause for optimism to-
ward its eventual use in the treatment of diseases of
the human immune system.
7
. Kawai, M.; Lane, B. C.; Hsieh, G. C.; Mollison, K. W.;
Carter, G. W.; Luly, J. R. FEBS Lett. 1993, 316, 107.
8. Olivera, D. L.; Kaplan, J. M.; Newman-Tarr, T.; Ruggieri,
E. V.; Badger, A. M. Clin. Immunol. Immunopathol. 1993,
6
8, 357.
. Henningsen, G. M.; Koller, L. D.; Exon, J. H.; Talcott, P.
9
A.; Osborne, C. A. J. Immunol. Methods 1984, 70, 153.
0. Swanborg, R. H. Clin. Immunol. Immunopathol. 1995, 77,
1
1
1
4
.
1. Newbould, B. B. Br. J. Pharmac. 1963, 21, 127; Winter, C.
A.; Nuss, G. W. Arthritis Rheum. 1966, 9, 394.
2. Jacobson, P. B.; Morgan, S. J.; Wilcox, D. M.; Nguyen,
P.; Ratajczak, C. A.; Carlson, R. P.; Harris, R. R.
Arthritis Rheum. 1999, 42, 2060.
13. Zheng, J.; Sarnbol, N. C.; Zimmerman, J.; Zaidi, A. Clin.
Pharmacol. Ther. 1996, 59, 150; PhysiciansÕ Desk Refer-
ence. 58th ed. Montvale, NJ: Medical Economics Co.,
2
004.
1
4. Karyekar, C. S.; Pradhan, R.; Freeney, T.; Edeki, T.; Ji,
Q.; Chiu, W.; Schwartz, L. B.; OÕDea, R. Multiple-dose
I.V. pharmacokinetics of a new immunosuppressive com-
pound, ABT-578, in healthy subjects. AAPS J. 2004, 6,
Abstract M1089.