Rational design of novel immunosuppressive drugs: Analogues of azathioprine lacking the 6-mercaptopurine substituent retain or have enhanced immunosuppressive effects

Article abstract of DOI:10.1021/jm960132w

Clinical use of the immunosuppressive drug azathioprine is limited by potentially serious toxic effects related to depression of bone marrow function. The immunosuppressive and toxic properties of azathioprine are regarded as being properties of the cytotoxicity of its metabolite, 6- mercaptopurine (6-MP). However, azathioprine has an immunosuppressive effect additional to that attributable to 6-MP alone, and we propose that this is associated with an action of the methylnitroimidazolyl substituent. This suggests a route to the rational design of nontoxic immunosuppressants by replacing the 6-MP component of azathioprine with nontoxic thiols. We have synthesized and tested in vitro 24 such analogues, with two being further tested in vivo. In the human mixed lymphocyte reaction, virtually all compounds showed some degree of activity, 10 compounds being more active than azathioprine. In vivo, two compounds were more effective than azathioprine at prolonging graft survival in mice. In an oral toxicity study in male CD1 mice at doses equivalent to those at which azathioprine caused severe bone marrow depression both analogues had no toxic effects. Our results show that the immunosuppressive effects and bone marrow toxicity of azathioprine are not a consequence of release of 6-MP alone, and with appropriate modification can be separated, an approach which may lead to less toxic immunosuppressive drugs.

Full text of DOI:10.1021/jm960132w

2690
J . Med. Chem. 1996, 39, 2690-2695
Ra tion a l Design of Novel Im m u n osu p p r essive Dr u gs: An a logu es of
Aza th iop r in e La ck in g th e 6-Mer ca p top u r in e Su bstitu en t Reta in or Ha ve
En h a n ced Im m u n osu p p r essive Effects
Duncan J . K. Crawford,* J ohn L. Maddocks, D. Neville J ones, and Paul Szawlowski
University Department of Medicine and Pharmacology, Royal Hallamshire Hospital, and Department of Chemistry,
University of Sheffield, Sheffield S10 2J F, U.K.
Received February 12, 1996X
Clinical use of the immunosuppressive drug azathioprine is limited by potentially serious toxic
effects related to depression of bone marrow function. The immunosuppressive and toxic
properties of azathioprine are regarded as being properties of the cytotoxicity of its metabolite,
6-mercaptopurine (6-MP). However, azathioprine has an immunosuppressive effect additional
to that attributable to 6-MP alone, and we propose that this is associated with an action of the
methylnitroimidazolyl substituent. This suggests a route to the rational design of nontoxic
immunosuppressants by replacing the 6-MP component of azathioprine with nontoxic thiols.
We have synthesized and tested in vitro 24 such analogues, with two being further tested in
vivo. In the human mixed lymphocyte reaction, virtually all compounds showed some degree
of activity, 10 compounds being more active than azathioprine. In vivo, two compounds were
more effective than azathioprine at prolonging graft survival in mice. In an oral toxicity study
in male CD1 mice at doses equivalent to those at which azathioprine caused severe bone marrow
depression both analogues had no toxic effects. Our results show that the immunosuppressive
effects and bone marrow toxicity of azathioprine are not a consequence of release of 6-MP alone,
and with appropriate modification can be separated, an approach which may lead to less toxic
immunosuppressive drugs.
In tr od u ction
ment of which is currently limited or prevented by the
undesirable side effects of currently available therapeu-
tic agents.
Immunosuppressants are of use clinically for three
main purposes: suppression of the hosts’ response to
organ allografts, suppression of the response of lym-
phocytes in the graft to host antigens in bone marrow
transplants, and treatment of a variety of conditions
when the immune response has been inappropriately
stimulated. Drugs used for immunosuppression include
the fungal peptide cyclosporin 1, the structurally dis-
similar but mechanistically related macrolide FK-506
2, and glucosteroids and cytotoxic agents such as
cyclophosphamide 3 and azathioprine 4 (Chart 1).
Although FK 506 and cyclosporin are more selective and
potent agents than the earlier drugs, they possess severe
adverse side effects such as nephrotoxicity, neurotox-
icity, and gastric toxicity which limit their clinical
utility.1-3 Consequently, although azathioprine has
been available for approximately 30 years, it remains
widely used as an immunosuppressive drug to prevent
the rejection of organ transplants and in diseases
involving the immune system. However, the use of
azathioprine itself is limited by potentially serious toxic
effects related to depression of bone marrow function.4,5
There is therefore an ongoing requirement for less toxic
and thus safer immunosuppressant drugs for clinical
use. These would not only be of use as replacement
drugs but also may be of potential use in a number of
long term inflammatory conditions, such as psoriasis,
rheumatoid arthritis, haemolytic anaemia, ulcerative
colitis, Guillain-Barre syndrome, Polyarteritis nodosa,
glomerulonephritis, and multiple sclerosis, the treat-
Chemically, azathioprine is the 1-methyl-4-nitro-
imidazyl derivative of 6-mercaptopurine, and it has
always been assumed that the immunosuppressive and
toxic properties of azathioprine are both a consequence
of release of 6-mercaptopurine as the main metabolite
in vivo.6 As azathioprine was originally developed as a
prodrug of 6-mercaptopurine, this view point is perhaps
not surprising. Physiologically, formation of 6-mercap-
topurine is a consequence of the nonenzymatic reaction
with the thiol groups of both glutathione and cysteine
(Scheme 2) which also gives rise to the alkylation
reaction products such as 5-glutathionyl-1-methyl-4-
nitroimidazole, 5-cysteinyl-1-methyl-4-nitroimidazole,
and 5-mercapto-1-methyl-4-nitroimidazole, which them-
selves have little or no immunosuppressive activity.7,8
Although it is tempting to explain both azathioprine’s
immunosuppression and bone marrow toxicity on the
basis of cytotoxicity caused by its nucleotide metabolites,
there are several pieces of evidence which suggest that
the action of azathioprine cannot be solely explained by
release of 6-mercaptopurine.
Clinically, observations in patients with kidney trans-
plants have shown a number of unusual differences
between the expected and observed mode of action; for
example, immunosuppression produced by azathioprine
is sufficient to prevent rejection in the absence of
neutropenia and is highly dependent on repeatedly
taking the drug. Cessation of azathioprine therapy,
even for a few days, is frequently associated with
transplant rejection.4,5,9 Moreover, although clonal
expansion through rapid cell division occurs in both
cellular and humoral immune responses, azathioprine
* Author for correspondence. Present Address: Tocris Cookson Ltd.,
Test Valley Business Centre, Test Lane, Nursling, Southampton SO16
9J W, U.K.
X Abstract published in Advance ACS Abstracts, J une 1, 1996.
S0022-2623(96)00132-X CCC: $12.00 © 1996 American Chemical Society

Rational Design of Novel Immunosuppressive Drugs
J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 14 2691
Ch a r t 1. Chemical Structures of Cyclosporin (1), FK-506 (2), Cyclophosphamide (3), and Azathioprine (4)
Sch em e 1. Synthetic Route to Azathioprine Analogues
has a relatively selective effect on the cellular immune
system.10 These phenomena, while somewhat circum-
stantial, are not consistent with an immunosuppressive
mechanism involving inhibition of DNA synthesis,
which could be expected to be much slower to respond.
In vitro studies using the human mixed lymphocyte
reaction (MLR) as a model of the cellular immune
reaction have shown that azathioprine is more active
in inhibiting the reaction when added within the first
24 h and is much less effective when added later, when
cellular proliferation is maximal.10,11 Further, 6-mer-
captopurine inhibition of the MLR is completely pre-
vented by hypoxanthine and inosine whereas that
resulting from azathioprine is unaffected.12 Addition-
ally, azathioprine inhibits the MLR using lymphocytes
from patients with Lesch-Nyhan syndrome that are
unable to form active nucleotide metabolites, responsible
for cytotoxicity.13 Taken together, these inconsistencies
indicate that azathioprine has an effect subtly different
from that which might be expected from the action of
6-MP alone.
Sch em e 2. Mechanism of Physiological Release of 6-MP
Some of these differences have been discussed, and
the suggestion has been made that azathioprine has an
effect related to its chemical reactivity with thiols or
amino groups by Elion,14 but the generally accepted
view is that the purine component is essential for
immunosuppression and the effect of azathioprine is a
consequence of its greater metabolic stability, giving a
more controlled release of 6-mercaptopurine. Whereas
we acknowledge that release of 6-MP plays a role in the
immunosuppressive action of azathioprine, we now
propose a mechanism which can explain the different,
enhanced effectiveness of azathioprine over 6-mercap-
topurine based on an investigation of its chemical mode
of action. As shown in Scheme 2, release of mercapto-
purine is a consequence of conjugate addition-elimina-
tion of a thiol at the 5 position of the imidazole ring.
This process can also be regarded as heteroarylation of
the attacking thiol. We suggest that this is a secondary
effect of azathioprine which also contributes to its
immunosuppressive action. As the lymphocyte cell
membrane is richly endowed with thiol groups,15 we
postulate that heteroarylation at thiols (or amino groups)
on one or more key sites is responsible for the additional
immunosuppressive effects seen (Scheme 3). With this
working hypothesis as a basis, we set out to explore the
possibility that by replacing 6-MP the immunosuppres-
sive and toxic effects of azathioprine may be separable
properties. We have therefore synthesized a number
of analogues of azathioprine in which the “upper” ,
proposed arylating portion of the molecule is retained
Sch em e 3. Proposed Model of Inactivation of
Lymphocytes by Heteroarylation
whereas the “lower” part of the molecule (6-mercapto-
purine, the toxic component) is replaced with different,
nontoxic mercaptans. Development of less toxic immu-
nosuppressants based on a structure-activity-related

2692 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 14
Ta ble 1. Experimental Conditions and Physical Data
Crawford et al.
separation of the immunosuppressive and toxic proper-
ties of azathioprine has not to our knowledge been
undertaken.
Resu lts a n d Discu ssion
As a primary screen, analogues were initially tested
in the human mixed lymphocyte reaction. Briefly,
antigen responding and stimulating human lympho-
cytes were incubated together for 5 days at 37 °C in a
desiccator flushed with a mixture of 5% CO2 and 95%
air. The response was measured by the incorporation
of tritiated thymidine into DNA, with the results being
expressed as either the percentage inhibition of the
uptake of tritiated thymidine at 10 or 25 µM or the ED50
value, unless otherwise indicated, and are presented in
Table 2. For comparison at 25 µM azathioprine alone
causes 79% inhibition and has an ED50 of 7.9 µM.
The results obtained for the first category of com-
pounds, the benzimidazoles 9-14, the benzoxazole 15,
and the benzthiazole 16, which are somewhat related
by the fact that their corresponding thiols are roughly
isosteric with 6-MP, show immediately that substitution
of the mercaptopurine substituent results in compounds
with retained or enhanced ability to inhibit cell activity,
with compounds 9, 10, 13, and 15 outscoring azathio-
prine in this test. Although we were very encouraged
by this finding, these compounds were not further
Ch em istr y
Analogues 9-32 were synthesized by reacting 5-chloro-
1-methyl-4-nitroimidazole 8 with the appropriate thiol
using either sodium hydroxide and water (method A)
or potassium carbonate in acetone (method B) as
reagents (Scheme 1). Key intermediate 8 was synthe-
sized following the methods given by Wallach16,17 and
Sarasin and Wegmann.18 Briefly, diethyl oxalate 5 is
converted to dimethyloxamide 6 on treatment with
aqueous methylamine, which on treatment with phos-
phorus pentachloride gives 5-chloro-1-methylimidazole
(7). This material can be nitrated using conventional
conditions to give 8, a white, highly crystalline com-
pound in good overall yield. The methylethyl analogues
31 and 32 are made by simply substituting ethylamine
for methylamine in the first step. All compounds made
were characterized by proton NMR, mass spectroscopy,
and microanalysis. (See Table 1 for a full list of
compounds made and characterisation details.)

Rational Design of Novel Immunosuppressive Drugs
J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 14 2693
Ta ble 2. Inhibitory Potency of Azathioprine Analogues in the
Human Mixed Lymphocyte Reaction
% inhibition of
[3H]thymidine incorporationa
compd no.
10 µM
NDc
25 µM
ED50 (µM)b
azathioprine
79
90
86
7.9
9
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
89
ND
ND
ND
ND
ND
ND
ND
ND
24
ND
ND
ND
ND
ND
ND
ND
ND
ND
14
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
2
59d
94
41
87
27
50
47
34
21
76
47
47
29
65
100
58
F igu r e 1. Effect of treatment with immunosuppressive drugs
by intraperitoneal injection on BALB/C skin allograft survival
in CBA mice. Treatment groups: saline control, 55 mice;
azathioprine positive control, 29 mice; compound 29, 20 mice;
compound 30, 46 mice. The results are shown as a box,
representing the median and 25th and 75th centiles, and
whiskers representing the lowest and highest values. NS )
not significant.
100
100
100
100
100
2.8
3.15
2.7
2.5
2.9
82
98
98
98
a
Values are the average of at least three experiments performed
for further testing for their ability to prevent rejection
of BALB/C skin allografts on CBA mice. Four groups
of mice were tested and treated with saline, azathio-
prine, and the analogues 29 and 30, respectively (Figure
1). In this model azathioprine did not prolong skin graft
survival when compared to the saline control group with
the following results being obtained at equimolar
doses: azathioprine median 12, range 9-15 days, n )
29 vs saline median 12, range 9-17 days, n ) 55; P )
0.19, 95% CIdiff 2, 0 days. However, the two analogues
tested were able to prolong skin graft survival, and the
results were statistically significant: the 1-methylimid-
azole analogue 29 median 13, range 11-16 days, n )
20 vs saline median 12, range 9-17 days, n ) 55; P )
0.003, 95% CIdiff 2, 1 days, and the 4-methyltriazole
derivative 30 median 14, range 11-20, n ) 46 vs saline
median 12, range 9-17 days, n ) 55; P ) 0.0001, 95%
CIdiff 2, 3 days. Compound 30 was superior to azathio-
prine in prolonging graft survival (P ) 0.003, 95% CIdiff
1, 3 days). Although these results appear relatively
modest, this model is a very severe test for immuno-
suppressive drugs, and the key observation is that the
analogues are more effective than the control drug.
In a 14 day oral dosing study, the toxic effects of the
1-methylimidazole (29) and the 4-methyltriazole (30)
were compared. Daily treatment of male CD1 mice with
100, 200, and 400 mg/kg doses of azathioprine consis-
tently produced marked leucopenia, severe bone marrow
depletion, and necrosis, whereas at equimolar doses
neither of the analogues produced these effects. That
the drugs were sufficiently orally absorbed to cause
immunosuppression was shown by T-cell priming of
regional lymph nodes.
b
in triplicate. ED50 values determined from a graph with at least
four points, each derived from the mean of 3 -10 experiments.
d
c ND ) not determined. Determined at 50 µM.
investigated at this stage as their solubility was rather
low, and 25 µM was about the maximum concentration
obtainable in the cell culture medium.
With the exception of compound 17, compounds 17-
25 represent derivatives made from commercially avail-
able mercaptans with no particular attempt to ration-
alize their design, but to test the generality of the
approach. Once again, all compounds tested showed
some degree of inhibition of cell stimulation, although
none of these is more potent than azathioprine. Com-
pound 17, 8-hydroxyazathioprine, is worthy of note for
its ability to inhibit the mixed lymphocyte reaction,
albeit with approximately one-third the activity of
azathioprine, as 8-hydroxymercaptopurine is known to
lack cytotoxicity in vitro.7
The final seven compounds (26-32) all have in
common 5-membered heterocyclic rings attached to the
primary imidazole, and these compounds proved to be
the most potent. The solubility of these compounds was
also optimized so these have been most widely investi-
gated. At 25 µM all except 27 caused complete inhibi-
tion of the MLR and for these compounds results
obtained at 10 µM and ED50 values more accurately
express their potency. The imidazoles 28 and 29
showed much better inhibition than azathioprine and
were determined to have ED50 values of 2.8 and 3.15
µM, respectively, compared to 7.9 µM for azathioprine.
The isosteric triazole 30 proved yet more potent, causing
98% inhibition of cell activity at 10 µM and having an
ED50 value of 2.7 µM. Finally, the methylethyl ana-
logues of 29 and 30, numerically 32 and 31, were also
tested with 31 giving the lowest ED50 value of the whole
series.
These results show clearly that the purine moiety of
azathioprine is not a prerequisite for its immunosup-
pressive effects, and furthermore, the imidazole is in
fact very tolerant of substitution at this position.
Replacement by almost any thiol-containing compound
gives rise to molecules that retain the immunosuppres-
Two of the most active compounds, the 1-methyl-
imidazole 29 and the 4-methyltriazole 30 were selected

2694 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 14
Crawford et al.
g, 61%): mp 210-212 °C; NMR (DMSO) δ 8.58 (1H, s, triazole
proton), 8.05 (1H, s, imidazole proton), 3.76 (3H, s, methyl
protons), 3.69 (3H, s, methyl protons); MS m/z 240 (M+). Anal.
Found: C, 35.19; H, 3.47; N, 34.84; S, 13.12. C7H8N6SO2
requires: C, 34.99; H, 3.35; N, 34.98; S, 13.34%.
sant properties of azathioprine with the overall level of
activity being about the same order of magnitude.
Our heteroarylation hypothesis to help explain the
mechanism of immunosuppression produced by azathio-
prine receives considerable support from the results of
this work. The wide range of chemical structures of
analogues possessing immunosuppressive activity over
a fairly narrow concentration range favors the idea of a
chemical reaction rather than a specific interaction with
a receptor. Their precise locus of action is, however,
unknown, and we have made no attempt to investigate
it in this study.
Meth od B: Typ ica l Exa m p le. 2-[(1-Meth yl-4-n itr o-5-
im id a zoyl)th io]th ia zolin e (26). A solution of 5-chloro-1-
methyl-4-nitroimidazole (3 g, 18.63 mmol) and 2-mercapto-
thiazoline (2.21 g, 18.63 mmol) in acetone (60 mL) was treated
with potassium carbonate (7.71 g, 55.89 mmol) and the
resulting solution allowed to stir at room temperature over-
night, after which time TLC showed no starting materials. The
solid material was removed by filtration, and the solvent was
removed under reduced pressure to give a crude product which
was recrystallized from ethanol to give the title compound as
light yellow crystals (3.266 g, 71%): mp 100-101 °C; ΝMR
(DMSO) δ 8.19 (1H, s, imidazole proton), 4.16 (2H, t, J ) 8
Hz, thiazoline protons), 3.44 (2H, t, J ) 8 Hz, thiazoline
protons), 3.70 (3H, s, methyl protons); MS m/z 198 (M - 46).
Anal. Found: C, 34.33; H, 3.60; N, 22.90; S, 25.96. C7H8N4S2O2
requires: C, 34.4; H, 3.6; N, 22.95 and S, 26.2%.
Con clu sion s
We have synthesized 24 analogues of azathioprine
lacking a 6-mercaptopurine substituent and found that
immunosuppressive effects are retained or even en-
hanced in these molecules, with 10 more potent than
azathioprine in vitro and two (29 and 30) being shown
to be more potent and less toxic in vivo.
Biologica l Stu d ies. Heparinized venous blood samples
were obtained from normal human donors. Lymphocytes were
separated from 20 mL blood samples by treatment with
dextran (MW 150 000) in normal saline and carbonyl iron
powder (type SF, from G. A. F. Ltd., Manchester, U.K.) before
isolation on Ficoll-Metrizoate (Lymphoprep, Nygaard AS, Oslo,
Norway). Stimulating cells were prepared with mitomycin-C
(Sigma Chemical Co. Ltd., Dorset, U.K.) as described by Bach
and Voynow.19 For culture, lymphocytes were suspended in
RPMI 1640 (Flow Laboratories, Irvine, U.K.) prepared with
glutamine (2 mM), HEPES buffer (20 mM), gentamicin 50 (µg/
mL), and amphotericin sodium desoxycholate (Fungizone, 2.5
µg/mL). They were cultured in triplicate in round-bottomed
microtitre plates (Flow Laboratories). Each well contained 200
µL, consisting of 50 µL each of antigen responding lymphocytes
(105), stimulating lymphocytes (105), inactivated, pooled hu-
man AB serum diluted with culture medium to give a final
concentration of 20%, and the test drug dissolved in culture
medium. Lymphocytes were incubated for 5 days at 37 °C in
a humidified desiccator flushed with a gas mixture of 5% CO2
and 95% air.
These results indicate that release of 6-MP, and its
effects on DNA synthesis, is not the sole explanation
for the immunosuppressive effects of azathioprine.
Instead it appears that a secondary effect, associated
with the nitroimidazole portion of the molecule is very
important and with appropriate modification can pro-
duce compounds with more potent immunosuppressive
properties. It seems possible that rather than release
of 6-mercaptopurine being responsible for the immuno-
suppressive effect of azathioprine, due to its toxicity
6-mercaptopurine may have been an unfortunate choice
from among the wide range of thiols which can be
substituted on the imidazole ring to produce immuno-
suppression. These compounds, or others designed
using the same principles, offer a lead to immunosup-
pressive drugs with greatly reduced side effects, which
will offer alternative and safer treatment of conditions
in which immunosuppressants are already used, both
directly and as steroid sparing agents.
The response was measured by the incorporation of tritiated
thymidine ([3H]Tdr) into DNA. Full details have been de-
scribed previously.11,12 Results were expressed as the concen-
tration of drug required to inhibit [3H]thymidine incorporation
into DNA in the MLR by 50%. The EC50 was determined from
a graph with at least four points, each derived from the mean
of 3-10 experiments.
Exp er im en ta l Section
Melting points are uncorrected. 1H NMR spectra were
recorded on a Bruker AM-250 (250 MHz) spectrometer sup-
ported by an Aspect 3000 data system. Chemical shifts (δ)
are expressed in parts per million relative to internal tetra-
methylsilane; coupling constants (J ) are in hertz. Elemental
analyses were within 0.4% of the theoretical values for all
compounds. Mass spectra were obtained using either a Kratos
MS25 or MS50 spectrometer supported by a DS 55 data
system. Thin layer chromatography was run on Merck 7736
60GF silica gel.
5-Ch lor o-1-m eth yl-4-n itr oim id a zole (8) was synthesized
using methods previously described in the literature.16-18
Analogues were prepared from this intermediate by using the
appropriate thiol under the conditions described below.
Meth od A: Typ ica l Exa m p le. 3-[(1-Meth yl-4-n itr o-5-
im id a zoyl)th io]-4-m eth yl-4H-1,2,4-tr ia zole (30). A solu-
tion of sodium hydroxide (1.366 g, 34.1 mmol) in water (100
mL) at room temperature was treated with 3-mercapto-4-
methyl-4H-1,2,4-triazole (3.571 g, 31.05 mmol) and the result-
ing suspension stirred until all the solid material was in
solution. 5-Chloro-1-methyl-4-nitroimidazole (5 g, 31.05 mmol)
was then added and the resulting solution allowed to stir for
1 h at room temperature, after which time TLC showed no
starting materials. Neutralization with acetic acid caused the
product to precipitate, and this was collected and recrystallized
from acetone to give the title compound as fawn crystals (4.53
For skin grafting, the technique of Billingham and Medawar
was used.20 A full thickness piece of Balb/c (white H-2d) mouse
tail skin was grafted onto the flank of a CBA (brown skin, H-2)
mouse (OLAC, Bicester, U.K.). The graft was protected by a
gypsum plaster cast which was removed after 8 days when
the condition of the graft was assessed visually by an inde-
pendent observer, previously blinded to the treatment. Ana-
logues were dissolved in 0.9% sterile saline at 37 °C to give a
final concentration of 5 mg/mL. Each was given in a dose of
45 mg/kg/day. Azathioprine (Imuran, Wellcome, U.K.) was
prepared in the same way and given in equimolar dosage of
52 mg/kg/day. Drugs were given to groups of 10 mice by
intraperitoneal injection 3 h before surgery and then once daily
until the graft had been fully rejected. Controls were treated
similarly but were given 0.9% saline. Skin graft survival data
was analyzed by stem and leaf displays and normal probability
plots and found to be not normally distributed. Statistical
analysis was performed using the Mann-Whitney U-test. The
data is presented as medians, 25th and 75th centiles, and the
range. Where comparisons are made, the 95% CIs for the
differences are presented (95% CIdiff). A P value of <0.05 was
considered significant.
Ack n ow led gm en t. This work was supported by
project grants from the Medical Research Council and

Rational Design of Novel Immunosuppressive Drugs
J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 14 2695
(10) Bach, J . F. The mode of action of immunosuppressive agents,
2nd ed.; North-Holland Publishing Co.: Amsterdam, 1975; pp
94-172.
(11) Al-Safi, S. A.; Maddocks, J . L. Effects of azathioprine on the
mixed lymphocyte reaction. Br. J . Clin. Pharmacol. 1983, 15,
203-209.
the National Kidney Research Foundation. The authors
thank Stuart Prismall for his assistance in preparing
the manuscript and drawing the diagrams.
Refer en ces
(12) Al-Safi, S. A.; Maddocks, J . L. Azathioprine and 6-mercapto-
purine (6-MP) suppress the human mixed lymphocyte reaction
(MLR) by different mechanisms. Br. J . Clin. Pharmacol. 1984,
17, 417-422.
(13) Szawlowski, P. W. S.; Al-Safi, S. A.; Dooley, T.; Maddocks, J . L.
Azathioprine suppresses the mixed lymphocyte reaction of
patients with the Lesch-Nyhan syndrome. Br. J . Clin. Pharma-
col. 1985, 20, 489-492.
(14) Elion, G. B. The Pharmacology of Azathioprine. Ann. N. Y. Acad.
Sci. 1993, 685, 401-407.
(15) Merrishi, J . N.; Grassetti, D. R. Sulphydryl groups on the surface
of intact Ehrlich ascites tumour cells, human blood platelets and
lymphocytes. Nature 1969, 224, 563-564.
(16) Wallach, O.; Boehringer, A. Verhalten des Dimethyloxamids.
(The preparation of dimethyloxamides.) Annalen 1877, 184, 50-
127.
(17) Wallach, O. Ueber die Einwirking von Phosphorpentachlorid auf
Saureamide. (On the action of phosphorous pentachloride with
acid amides.) Annalen 1882, 214, 257-327.
(18) Sarasin, J .; Wegmann, E. Synthese de l’heteroxanthine a partir
d’un derive de l’imidazol. (Synthesis of heteroxanthines starting
from a derivative of imidazole.) Helv. Chim. Acta. 1924, 7, 713-
719.
(19) Bach, F. H.; Voynow, N. K. One-way stimulation in mixed
lymphocyte cultures. Science 1966, 153, 545-547.
(20) Billingham, R. E. and Medawar, P. B. The technique of free skin
grafting in mammals. J . Exp. Biol. 1951, 28, 385-405.
(1) Myers, B. D.; Ross, J .; Newton, L.; Luetscher, J .; Perlroth, M.
Cyclosporin-Associated Chronic Nephropathy. N. Engl. J . Med.
1984, 311, 699-705.
(2) The US Multicenter FK506 Liver Study Group. A Comparison
of Tacrolimus (FK 506) and Cyclosporine for Immunosuppression
in Liver Transplantation. N. Engl. J . Med. 1994, 331, 1110-
1115.
(3) Atkinson, P.; J oubert, G.; Barron, A.; Grant, D.; Paradis, K.;
Seidman, E.; Wall, W.; Rosenberg, H.; Howard, J .; Williams, S.;
Stiller, C.; Hypertrophic Cardiomyopathy associated with Tac-
rolimus in Paediatric Transplant Patients. Lancet 1995, 345,
894-896.
(4) Oesterwitz, H.; Horpacsy, G.; May.; Mebel, M. Frequency of
Leucopenia incidents following azathioprine therapy after kidney
transplantation. Eur. Urol. 1978, 4, 167-170.
(5) Pollak, R.; Nishikawa, R. A.; Mozes, M. F.; J onasson, O.
Azathioprine-induced leucopenia. Clinical significance in Renal
transplantation. J . Surg. Res. 1980, 29, 258-264.
(6) Elion, G. B. The purine path to chemotherapy. Science 1989,
244, 41-47.
(7) Chalmers, A. H. Studies on the mechanism of formation of
5-mercapto-1-methyl-4-nitroimidazole. A metabolite of the im-
munosuppressive drug azathioprine. Biochem. Pharmacol. 1974,
23, 1891-1901.
(8) Chalmers, A. H.; Knight, P. R.; Atkinson, M. R. Conversion of
azathioprine into mercaptopurine and mercaptoimidazole de-
rivatives in vitro and during immunosuppressive therapy. Aust.
J . Exp. Biol. Med. Sci. 1967, 45, 681-691.
(9) Haesslein, H. C.; Pierce, J . C.; Lee, H. M.; Hume, D. M.
Leucopenia and azathioprine management in renal transplanta-
tion. Surgery 1972, 71, 598-604.
J M960132W