J.Med. Chem. 1986,29, 555-561
555
Relationship between Lipophilic Character and Urinary Excretion of Nitroimidazoles and Nitrot hiazoles in Rats Giorgio Cantelli-Forti,+ Maria Clelia Guerra,? Anna Maria Barbaro,t Patrizia Hrelia,? Gian Luigi Biagi,*t and Pier Andrea Boreat Istituto di Farmacologia, Universitd d i Bologna, Bologna, Italy, and Istituto d i Farmacologia, Universitd di Ferrara, Ferrara, Italy. Received M a y 16, 1985 Many processes are involved in the renal excretion of drugs, but very little is known about their quantitative structure-activity relationship. The relationship between urinary excretion and lipophilic character of a series of nitroimidazoles and nitrothiazoles was studied. The unmetabolized forms of the drugs were detected in the urine by means of UV and HPLC procedures. The urinary excretion of unmetabolized forms is parabolically related with the log P, as an expression of lipophilic character of molecules.
In recent years there has been a considerable interest in the study of the mutagenic activity of nitroimidazo1es.l In previous reports we have investigated the mutagenic activity in vitro of a series of nitroimidazoles and nitrothiazoles and its relationship with physicochemical par a m e t e r ~ . ~ -The ~ mutagenic activity was shown to be influenced by both the R, value, as an expression of the lipophilic character of the molecules, and an interaction term taking into account the positive effect exerted by substituents characterized by higher molar refractivity and capable of hydrogen b ~ n d i n g . More ~ recently we have determined the urinary excretion of some of the above compounds in their unchanged and metabolic forms by means of a HPLC technique.6 In an attempt to clarify the role played by the lipophilic character of drugs in the very complex process of renal e ~ c r e t i o nwe , ~ have investigated the relationship between the lipophilic character and the urinary excretion of the unchanged forms of a series of 26 nitroheterocyclic compounds. The present paper describes the results of that work. Results Relationship between Log BR and Log P Values. The amounts of the unchanged form of each drug were detected in the urine samples collected a t 0-18, 18-36, 36-54, and 54-72 h after the treatment. The total 18, 36, 54, and 72 h urinary excretion of the unchanged form of each drug was then calculated and the data reported in Table I1 as log (percent X 10) of the administered dose. As a first step in our work we detected the unchanged form of 11 compounds in urine by means of both the UV and HPLC technique. The very good correlation coefficient of eq 1seems to justify the use in the main part of our work log BRHPLC= -0.352 (h0.149) + 1.136 (h0.067) log BRuv (1)
r = 0.985 s = 0.063 F = 290.68 P C 0.005 of the only UV procedure for the detection of the unchanged forms in urine. The log BRw values showed that the unchanged forms of 25 test compounds were present in the urine collected at 0-18 h. In particular, for compounds 26, 10, and 13 the unmetabolized forms excreted in the 18-h period accounted for 48.6%, 43.6%, and 41.7%, respectively, of the administered dose. On the other hand, for 14,22,4, and 25 the unmetabolized forms excreted in the same time accounted only for 2.5%, 4.1%, 4.3% and 4.6%, respectively, of the administered dose. The only exception was represented by 15 for which it was not n = 11
t Universiti di Bologna.
* Universiti di Ferrara.
0022-2623/86/1829-0555$01.50/0
possible to detect its unchanged form in the 0-18-h urine. The HPLC technique also did not show any peak with the retention time of compound 15, and as shown in a previous paper,2 the urine collected in the first 18 h did not demonstrate any mutagenic activity in the Ames test, which could be related to the absence of any active form of 15. The lack of detectable traces of 15 in urine in the first 18 h after application could be due to higher biotransformation in the liver and/or higher renal reabsorption. This could be related to the fact that 15 shows the highest log P value in the series reported in Table I. However, the lipophilic character seemed to influence the excretion of all the test compounds. The relationship between the log P values of the compounds and their log BR values is described by eq 2 and 3. 18 hours log BRuv = 2.240 (f0.069) - 0.217 (h0.119) log P (2)
n = 25
r = 0.356
s = 0.347
F = 3.35
P c 0.10 log BRuv = 2.570 (k0.042) + 0.050 (f0.054) log P -0.964 (h0.089) (log P)2 (3)
n = 25
r = 0.928 s = 0.141 F = 68.37 P C 0.005 The introduction of the (log P)2term into eq 2 significantly improved the correlation coefficient of eq 3, showing a quadratic regression between lipophilic character and urinary excretion. Obviously, because of the lack in the 18-h sample of urine of the unchanged form of 15, eq 2 and 3 were calculated for only 25 compounds (Figure 1). The ideal log Po,i.e. the lipophilic character determining the maximum of the urinary excretion, was calculated from eq 3 and found to be log Po = 0.026 (0.079 and -0.055 being the 95% confidence limits). On the basis of eq 3, compound 15 with a log P = 2.03 should have a log BR of -1.30, which means a very low excretion percentage (0.05%). This might explain the lack of any detectable amount of 15 in the 18-h urine sample. Since 11compounds had been (1) Voogd, C. E. Mutat. Res. 1981, 86, 243. (2) Cantelli-Forti, G.; Aicardi, G.; Guerra, M. C.; Barbaro, A. M.; Biagi, G. L. Teratog., Carcinog., Mutagen. 1983, 3, 51. (3) Guerra, M. C.; Barbaro, A. M.; Cantelli-Forti, G.; Foffani, M. T.; Biagi, G. L.; Borea, P. A.; Fini, A. J . Chromatogr. 1981,216, 93. (4) Guerra, M. C.; Barbaro, A. M.; Cantelli-Forti, G.; Biagi, G. L.; Borea, P. A. J. Chromatogr. 1983, 259, 329. ( 5 ) Biagi, G. L.; Barbaro, A. M.; Guerra, M. C.; Cantelli-Forti, G.; Aicardi, G.; Borea, P. A. Teratog., Carcinog., Mutagen. 1983, 3, 429. (6) Cantelli-Forti, G.; Guerra, M. C.; Hrelia, P.; Barbaro, A. M.; Biagi, G. L. Drugs E x p . Clin. Res. 1984, 5, 325. (7) Kubinyi, H. Arzneim.-Forsch. 1979, 29, 1067.
0 1986 American Chemical Society
556 Journal of Medicinal Chemistry, 1986, Vol. 29, No. 4
Cantelli-Forti et al
Table I. Structural Formulas and Physicochemical Parameters of the Nitroheterocyclic Compounds
Empiricrl tonnula
1 I 1
~
2
1
i 1 3
-H
5 Nitroimidazole
2 Methyl 5 nifroimidrzole
I
N
-H
I
-CHICHIOH
INl
1 l2.elhyl.urb”thioic acid Omethyl eslcrl.2 methyl 5 nitroimidazole (Csrntd8zoleI
1 . (2-lethylrulfonvl~.e~vlt2. .methvl.5.nitroimid8zole ITinidazole)
-CH,
I
I
C l H l 1 NIOI
’
-010
155 11
-CH,
24427
0450072
-0067
0465
1 [Methyl 2 lmethanol csrbsma tel 15 nitroimidizole iRonid8zole)
j
-016
106
-069
090
081
I
I
-CH,CH,SO,CH,CH,
-CH,
247 26
-0 125
035
21963
0199
033
060
26223
-0 115
095
-030
40236
0028
158
-070
31027
01-
111
-036
I -CH (CH,CIICH,OH
-CH,
~
C1HlQN303C
-CH,COONH,
I
1 I2 N morpholinylethyll.5 nitroimidizole (Nimoruolel
n
-CH,-CH,-
N
-H
0
W
I
-CH,
I
-CH,OH
CmHmNeo,
IDA38041
-CHI-
H3
IDA38311
1 1 15
16
1
I
1 1 17
18
1
19
~
/
1 1
OH
IDA38391
q,-,, m
-CH=CH -
C34H13N30,
I -CH,
I -CH,
IDA38401
IDA38531
-0
I
I -CH1-CH-CH,OH
I
1
I -CHl-CH,OH
1
iOA3838l
u
-CH3
IDA38321
t i
NAN
,CWH
~
j
om
I 13
14
/
-0.16
I
I
~
16918
-CHO
I . Mathyl.2-l2-amino4-ethinyl-pyrimidine)-5-nitroimidizole IAzanidizolel
I
-0087
I
n1troimid8zole
!
171 16
0 . W -0.19
9
1 Methyl Zhydroxymethvl 5
j
1-
C6HSN301
-CHICH,)CH,
-CH,CH,NHCOCH,
1 .[I~hlorome~yll.eth~nol~2-mefhyl.5-nitroimid8zole IOrnidazolel
10
-CH,
-CH,
1
j
C4H5N301
113.08
I
1 . Mathyl.2-formyl.5-nitro. imidazole
7
-CH,
I 1 Ethanol.2.mrthvl.5.nilro imidrzole IMetrondazoIe) 1 .Methyl.2 I l . m t h y l e l h y l l ~ 5 n~tro~midrzolc IIpronidazolel
6
-H
-CHI
CIH N3°4 ,l
031
Journal of Medicinal Chemistry, 1986, Vol. 29, No. 4 557
Nitroimidazoles and Nitrothiazoles Table I (Continued)
Empirical formulr
Structun
-pard
20
(DA38541
N
21
IMY 40/201
N
21
IDA38511
N
23
(DA38281
N
-CH, -CH,-0-CH,
-CH,-N/CH2cn,
\CH,CH,
-CH, -CH,-SO,-CH,
Wk‘
Rra
C,HzN,Ol
270.34
-%
C,H,,N,SO,
233.24
C,HeN302
187.18
0.455
0.88
-CH,
C,,H,N,U,
256.30
0.058
0.S
-H
C,H,N,O,
235.25
0.021 -0.30
0.383
-CH-CHZ
’
woi#lt
-cn,
/““2
-cn,
Molrubr
0.021
1.58
-0.007 -0.30
CH CH
On CH,-h-CH,’
‘CH,CH,
24
(DA59151
N
-CH,-CH,-
25
1 .(5-nitro.?.thuzdvlkZ-
C,H,N,O,S
215.20
1
2 . mi~b5(pnitrod*nyl
CsH,$O,S
185.30
N/q(
u
-CH,
‘r’ 4 m i d u o l i d i m (Niriblolel
sulfonyl)-thiazole
,
-
detected by means of the HPLC technique in the urine collected in the first 18h after application, eq 4 and 5 have been calculated with those log BRHpLC values. log BRHpLC = 2.226 (f0.074) - 0.417 (f0.122) log P (4)
n = 11
r = 0.752
s = 0.240 P =
