Quantitative Structure-Activity Studies of Hydrazones

S T R U C T U R E - A C T I V I T Y STUDIES OF

The authors want to thank Mr. S . STRAUB and his coworkers for their excellent technical work and Dr. 1

K.

HAN,

D.

TETAERT,

M.

DAUTREVAUX,

Y.

MOSCHETTO,

[1971].

6

F. W . J. TEALE, Biochem. biophysica Acta 35, 543 [1959].

7

H . HAGENMAIER, W . EBBIGHAUSEN, G . NICHOLSON, a n d

P. EDMAN and G. BEGG, Europ. J. Biochem. 1, 80 [1967].

3

A . B. EDMUNDSON, N a t u r e

4

D . D A U T R E V A U X , Y . BOULANGER, K . H A N , a n d G . BISERTE,

[London]

205, 883

[1965].

5

K.

159

G. NICHOLSON for performing the mass spectrometric analysis.

and G. BISERTE, Hoppe-Seyler's Z. physiol. Chem. 352, 15 2

HYDRAZONES

[Amsterdam] W.

VÖTSCH, Z. Naturforsch. 25 b, 681 [1970]. 8

K.

HAN,

M.

DAUTREVAUX,

C.

CHAILA,

and

G.

BISERTE,

Europ. J. Biochem. 16, 465 [1970].

Europ. J. Biochem. 11, 267 [1969]. BÜNNIG

and

R.

HAMM,

J.

Chromatography

43,

450

[1969],

Quantitative Structure-Activity Studies of Hydrazones, Uncouplers of Oxidative Phosphorylation * W . DRABER, K . H . BÜCHEL Farbenfabriken Bayer AG, Forschungszentrum, D-5600 Wuppertal 1, Postfach 13 01 05 a n d G . SCHÄFER Medizinische Hochschule Hannover, Department Biochemie ( Z . Naturforsch. 27 b, 159—171 [1972] ; received November 3, 1971)

a-Acyl-a-cyanocarbonyl-phenylhydrazones are effective uncouplers of oxidative phosphorylation. The p/ 50 -values of a series of 60 hydrazones with substituent variations in six positions of the molecule were determined in rat liver mitochondria. They ranged from 4.96 to 7.06. The biological data together with physico-chemical compound and substituent parameters were analysed by multiple regression techniques to establish the structure activity relationship. The integral parameters p K a and log P (partition coefficient) gave correlations of only moderate significance. Good agreement of found and calculated p/ 50 -values was obtained by an equation with electronic (o) and hydrophobic (it) substituent parameters in linear and quadratic terms. It is concluded that the contribution of a substituent to uncoupling activity depends on its position in the molecule. The activity is enhanced by hydrophobic shielding of the acidic NH-group. The relevancy of these results in relation to current theories on the mechanism of oxidative phosphorylation is discussed.

Reagents dissipating the link between formation of ATP and electron transport in mitochondria, the so-called uncouplers of oxidative phosphorylation, belong to different chemical classes. Recent compilations of uncouplers *>la comprise such diverse compounds as dicoumarol, lauric acid, aromatic hydrazones, benzimidazoles, salicylanilides, methylamine and atebrin, apart from the classical nitrophenols. Furthermore, l,l,3-tricyano-2-amino-l-propene 2 , 3-nitropropionate 3 , 2-anilinothiophenes 4 , and 3'-trifluoromethyl-yV-phenylanthranilic acid 5 are reported to be uncouplers. However, a critical comparison of the activity of these chemicals, expressed as the concentration necessary either for half-maximal stimulation of electron-transport in absence of ADP

and Pi, 50%-inhibition of ATP-synthesis or halfmaximal stimulation of ATP-ase, reveals that all those compounds with a high degree of activity have an acidic OH- or NH-group in common. This is true for the classical uncouplers, 2,4-dinitro-phenol as for other phenols, for salicylanilides, dicyanocarlbonyl-hydrazones, benzotriazoles, benzimidazoles, trihaloimidazoles and phenylaminodinitrothiophenes. In contrast, carboxylic acids like lauric acid and 3-nitro-propionic acid or amines like methylamine and chloropromazine are comparatively weak uncouplers. Discussions on the mechanism of uncoupling action are usually based on the assumption that all acidic uncouplers act by the same general mechanism. For the OH-

* Dedicated to Prof. R. WEGLER on the occasion of his 65th birthday.

Request for reprints should be sent to Dr. K. H. BÜCHEL, Farbenfabriken Bayer AG, Forschungszentrum, D-5600 Wuppertal 1, Postfach 13 01 05.

Unauthenticated Download Date | 11/24/15 1:34 PM

W. D R A B E R , K. H. BÜCHEL, AND G. SCHÄFER

160

and NH-acidic uncouplers, this assumption is the most probable one, while it appears questionable when all the different classes of compounds including those with low activity are taken into consideration. Apart from the necessity of an acidic — OH or — NH group, there must be other structural requirements for high uncoupling activity, which are less evident. Information about them can be obtained by quantitative structure-activity studies and should provide a valuable background in discussions on the mechanism of energy coupling. As early

as

1961,

HEMKER

and

HÜLSMANN

have

6

stated that the uncoupling activity of alkyl-dinitrophenols is determined not only by their acidity but also by their lipophilic character which can be measured by partitioning between a lipophilic and a hydrophilic phase. In 1965, HANSCH and coworkers described the uncoupling activity of phenols by a linear regression function of pKa and the lipophilic substituent parameter n 7 . Similar MURAOKA

results

were

a n d TERADA

published tried to

by

FUJITA 8 .

correlate the

un-

coupling activity of /V-phenyl-anthranilic acids either with the partition coefficient (eq. 1) or with pKa (eq. 2 ) , but with unsatisfactory results9. log

log

c

c

= 2.043+ 0.777 log P0 n 11

r 0.723

(1) *

s 0.369

(2)

=0.494 +0.586 P K 13

0.808

0.293

However, our recalculation of their data gave a good correlation (eq. 3 ) , using TI and o as parameters 9a . log— = 3 415 + 0.598 o + 0.597 n c

(3)

11 0.935 0.168 All published equations correlating the activity of uncouplers with physico-chemical parameters are of the simple linear form log * —a-\-b- o + c-71 c A naive interpretation would lead to the conclusion that it is just acidity and lipophilicity which * n is the number of examples used in the regression analysis, r is the correlation coefficient and s the standard deviation.

make an uncoupler. This is certainly an over-simplioation. In fact, lipophilicity as a measure of protein binding has been interpreted in this way. No conclusive explanation, however, answers the question why no potent uncouplers are found e. g. among the fatty acids. Moreover, good to excellent correlations which have been obtained with the above linear regression equation, are mainly the result of limited variance in the substituents. For instance, in FUJITA'S analysis 8 of WEINBACH'S and GARBUS' data 10 , which resulted in the equation log

c

= - 1 . 3 4 0 + 0.928 zfpKa

0.172 ti

(with r = 0.995 and 5 = 0.226) only chlorine, bromine and nitro substituents of the phenolic nucleus were included. The number of compounds was limited to 10. In a recently published investigation on 23 phenols 10:1 of low to moderate uncoupling activity a linear H a n s c h equation was obtained. The variation of substituents was confined to halogens, nitro and methyl, and the concept of additivity of the free energy parameters o and TI had to be left with poly-substituted componds. In spite of the high statistical significance of the correlations obtained, one has to be very cautious in discussing this result in connection with the mechanism of action, when the variance of substituents or data is low. Carelessness in the interpretation of quantitative structure-activity correlations will inevitably lead to wrong conclusions. This paper presents an analysis, based on the TI — o-approach n , of the biological and physicochemical data of 60 a-acyl-a-cyanocarbonyl-phenylhydrazones. These compounds are uncouplers of oxidative phosphorylation with p/50-values ranging from 7.09 to 4.96, a 110-fold spread of potency. 14 different substituents are varied in six different positions of the molecule. A number of them have insecticidal and acaricidal activity connected 12 with more favourable mammalian toxicity and low er phytotoxicity than the related dicyano-carbonylphenylhydrazones 13. Methods The a-acyl-a-cyano-carbonylphenylhydrazones were prepared in analogy to known methods by coupling of diazotized anilines with cyanomethyl-compounds14. Their melting points are given in Table 1. Satisfactory


STRUCTURE-ACTIVITY STUDIES OF HYDRAZONES

4 22. This offers the advantage of greater experimental simplicity which makes it possible to overcome the usually lower accuracy by a larger number of replicas. In addition, one can obtain data also from impure or instable samples. Commercial silica gel plates (SIL G/UV254, Macherey and Nagel, Düren, Germany) were used which were impregnated with 10% paraffin oil in chloroform, dried at 100 °C and aged for a few days, as the stationary phase. Dioxane/acetone/0.01 N H 2 S0 4 was the mobile phase. The values of Table 1 were each obtained from 6 — 12 replicas. The 7?M-values did not correlate with the log p-values as eq. (5) shows, nor do they correlate with n. 7?M = - 0 . 6 6 3 + 0.025 log P0 n 50

r 0.078

(5)

s 0.269

This result was not improved by using additional solvent or thin-layer plate systems. Since with several other classes of compounds, good inter-correlations of the different lipophilic parameters were found 23 ' 24 , some property of the hydrazones seems to be the reason of the above-mentioned disagreement. We believe that it is the acidic NH-group which causes interaction with the material of the thin-layer plate, thus disturbing the partitioning process. Biological data For detection of uncoupling activities of the compounds listed in Table 1 their effects on respiratory


STRUCTURE-ACTIVITY

STUDIES OF

rate, ATP-ase activity, redox state of pyridinenucleotides, and of cytochrome b were determined according to methods previously described25'26. Mitochondria from rat liver were supplied with succinate as respiratory substrate. Quantitative results were obtained by determination of the concentration dependency of uncouplers for state-4 release 27. Oxygen uptake was recorded polarographically in absence of ADP. The uncoupling activity was expressed in terms of the p/50-value. The latter has been defined earlier as the negative log of the concentration of an uncoupler yielding half-maximal stimulation of oxygen uptake in the above system 28. The dependency of succinate respiration from uncoupler concentration is given by sigmoidal curves in a double log plot 4. As already reported on uncoupling thiophen-derivatives at concentrations roughly one order of magnitude above those causing maximum uncoupling an inhibiting action on mitochondrial respiration could be observed with active compounds28. In accordance with results of SLATER et al. 29 and V A N D A M et al. 30 this inhibition is probably due to competition of uncoupleranions with substrate anions for penetration into the mitochondria. Fig. 1 elucidates the dependency of succinate accumulation by rat liver mitochondria on uncoupler concentration, including a L i n e w e a v e r B u r k plot demonstrating the compotitive type of inhibition. However, thermodynamic calculations of FINKELSTEIN make it conceivable that a dimer formed between the undissociated and the dissociated form of an uncoupling weak acid may be the penetrating species 31. In contrast to 2-anilinothiophenes, cyclic uncoupling phenomena were never detected 4 ' 33. Whereas TV-alkylated compounds did not exhibit uncoupling activities. iV-acylated compounds generally did. Obviously, an esterase activity is attached to liver mitochondria through which the TV-acyl compounds are hydrolyzed thus liberating the active form of uncouplers. The enzymatic nature of the reaction was concluded from kinetic data 34 demonstrating saturation characteristics

163

of the reaction rate with respect to substrate concentration. This esterase activity to a large extent was also found in the 10,000 g supernatant of rat liver homogenate and in purified microsomal fractions from rat liver. It should be mentioned that most of the uncouplers used in this study form coloured anions, and represent pH-indicators with a broad absorption band in the visible. The relative concentration of the anions in mitochondrial suspensions can be detected by means of dual wavelength spectrophotometry. From such experiments it was seen that the amount of anions decreased during the state 3 u (uncoupled) — state 5 (anaerobic) transition. This points towards an unequal distribution of not only protons, but also the uncoupler anion during stimulated respiration, as being expected from the existence of a respiratory induced pH-gradient across the mitochondrial membrane 35- 36. Substituent parameters Our 60-compound set contains a number of substituents for which the corresponding parameters were not available. First of all, o-values for the groups R were calculated /7\-NH-N=C^CN

II 0

x

by regression equations of the form pKA = a + 6--2' o with the pK*-values listed in Table 1. By using only published m- and p-a-values37 for the four different R-substituents the following equations were obtained. R

pK*

n

r

s

Eq.

CH 3

6 . 0 0 - 1.59 a m , p

5

0.85

0.25

(6)

C(CH3)3

7 . 0 6 - 2.22 0m,p

4

0.94

0.17

(7)

OCH3

7 . 1 8 - 1.83 a m > p

6

0.99

0.07

(8)

OC3H5

2 . 1 2 - •1.49 o m , p

5

0.97

0.11

(9)

fnmoles Succ. 7 20 ~L mgPr.-m in J

[11A [~mg Pr.-min 7 [yj~[nmoles SuccJ

T=15°C 4

HYDRAZONES

T=15°C;

15

-

cSucc.=3.9mM

10

(I) (E)

1-10'7

mol/l 2-W5mol/l Uncoupler 1.0

05

[cSucc]=[mmol]

2.0

C Uncoupler

3.0

>

4.0

'•10-j-

5.0

F i g . 1. Influence of u n c o u p l e r ( c o m p o u n d Nr. 4) on uptake of 14-C-succinate by rat liver mitochondria. Uptake was measured b y the filtration centrifugation t e c h n i q u e 3 2 ; corrections were m a d e f o r the sucrose-perm e a b l e space by determinations with 14-C-sucrose in parallel experiments. For details c. f . 8 2 a .


W . D R A B E R , K. H. BÜCHEL, A N D G. S C H Ä F E R

164

Contrary to expectation the regression lines show different slopes, probably due to the limited variance of the data. A graphical evaluation of the four regression lines at an average pK*-value with Or = 0 for R = CH3 as an arbitrary value gave the following figures: R CH3 C(CH3)3 OCH3 OC2H5

OR

-

crm -0.07 -0.12 0.12 0.15

0.00 0.43 0.60 0.65

-

7im , to obtain eq. (13) and its inverse function (14). log Pq —

2.873 + 0,996 n2_0 + O.398 TTR

(13)

712—6 = - 2.785 - 0.384 jir + 0.977 log P0 n r 23 0.960 23 0.951

(14) s 0.165 0.174

The lacking 71-values calculated by eq. (14) are listed in Table 4.

0.17 0.20 0.27 0.25

Substituent

71

Derived f r o m c o m p o u n d s N o .

Table 2. a-values.

2-CF3 4-CF3

A comparison with the H a m m e 11 constants for aromatic m- and p-substituents shows that the effect of the group R is mainly through resonance interaction with the electron attracting centre. By using the or-values from Table 2, eq. (10) and the inverse eq. (11) were obtained.

2-NO2

1.08 0.87 0.11 0.21 0.07 0.54 1.64 1.29 1.67

28, 30, 34, 58 39 31 12, 38

pK* = 7.284-1.838 am>PlR am,p>R = 3.863-0.535 pK*

n 20 20

r 0.978 0.964

s 0.146 (10) 0.097 (11)

Eq. (11) was used to calculate the lacking a-values which are listed in Table 3. Substituent

0

Derived f r o m c o m p o u n d s N o .

2-CH3 2-CF3 2-C1

0.00 0.51 0.30

2-NO2

0.39 0.51 1.01 0.95 0.28 0.33 0.96 0.74

19, 44 22, 28, 32, 56 13, 15, 16, 26, 29, 30, 43, 45, 47, 49, 57 20, 52, 54 58 39 31 46 12, 17, 41 38 1, 7, 14

2-CN 2-S02CH3 2-S02C2H5 3-CHF2 4-SCFg 4-S02CF3 3,4-CF20CF20

Table 3. a-values.

A regression analysis of the pK*-values of the 60compound set resulted in eq. (12) n pK* = 7.326-1.957 a2-C>R 60

r 0.922

s 0.322

(12)

which differs only slightly from eq. (10). An attempt to calculate ^-values for CH 3 , C(CH3)3, OCH3 and O C 2 H 5 by using m, and p-nvalues according to HANSCH and FUJITA 11 and the measured log P0-values of the four different sets in a manner analogously to that applied to a-values failed, because the slopes of the four lines differed too much and one of them was statistically not significant. Therefore, we used ^-values reported for the phenoxyacetic acid system 38 also for the R-substituent setting 7TR =

2-CN 2-S02CH3 2-S02C2H5 4-SCFg 4-S02CF3 3,4-OCF2OCF2

32, 56 42, 52, 54

41

1» 7

Table 4. ^-values.

The reliable log P0-values were correlated with ti and gave the eq. (15). log P0 = 2.876 + 0.933 ti2~6 + 0.497 mi n 47

r 0.905

(15) s 0.311

This correlation is not particularly good because of the only moderate precision of log P0. It reflects the problems involved in the determination of partition coefficients of compounds of limited hydrolytic stability. As already mentioned above, the i?m-values from reversed-phase thin-layer chromatography gave no correlation at all either with log P 0 or with n. Results Current theories on the mechanism of oxidative phosphorylation and the hitherto available structureactivity correlations with uncouplers are consistent with the assumption, that electronic as well as lipophilic properties of an uncoupler determine its activity. Steric effects may also play a significant role and have been taken into account in this analysis, although steric parameters are more difficult to obtain and more problematic to use than the wellestablished a- and ^-constants. The biological data expressed as p/ 50 -values were taken as dependent variables. Together with the molecular and substituent parameters as independent variables, the data were treated by computerized multiple regression techniques. Multiple correlation coefficient ( r ) , standard deviation (5), and 95%-


S T R U C T U R E - A C T I V I T Y STUDIES OF H Y D R A Z O N E S

confidence intervals of the coefficients were taken as statistical criteria. The analysis was carried out step-wise starting from simple equations with one or two variables. Improvement was achieved by adding further parameters and discarding those which proved statistically insignificant. For practical reasons and in agreement with the current application of regression analysis on structure-activity relationships, only linear and quadratic equations were used. The 60 p/50-values were used in all calculations. Correlations with molecule parameters A number of regression equations with the molecule parameters pK*, log.P0 and R\,[ containing linear, quadratic and interaction terms are shown in Table 5. Neither of these equations gave a very satisfactory correlation. Some conclusions, however, could be drawn. The negative sign of the pK2-term in all equations indicates a pK-optimum which is in agreement with expectance, although the earlier reported uncoupler correlations do not contain such a term 7 ' 8 , cf. also eq. (3). The improvement gained by introduction of the lipophilic parameters log P and i?M is small, and the coefficients show little if any significance, log P is a slightly better hydrophobic parameter than R ^ , which is in agreement with the observation, that Rm does not correlate with lipophilic constants (eq. 5). By the interaction terms in eq. (20) and (21), the correlations are slightly improved, but still the significance of the hydrophobic terms is low as judged by the 95%Eq. No.

16 17 18 19 20 21

«0

pK* exp.I1

exp.22

-

+ 3.29 1.26 + 2.30 1.28 + 2.20 0.12 + 2.86 1.58 + 0.95 1.55 + 2.67 1.46

-0.32 0.11 -0.23 0.12 -0.22 0.12 - 0.28 0.14 - 0.16 0.12 - 0.28 0.13

1.89 3 3.454 -0.05 3.30 + 0.71 3.40 + 0.88 4.30 + 5.36 4.83 + 0.11 4.03

logP exp. 1

+ 0.18 0.10 - 0.17 0.45

exp. 2

confidence-intervals. This led us to believe that the substituents in the different positions of the molecule contribute in different ways to the activity. Thus it was necessary to separate the substituent effects in the regression equations by replacing the molecular parameters by substituent constants. The use of substituent parameters has another practical advantage. The attempt to use multiple regression analysis in a predictive sense as a tool to make more active compounds, makes it necessary to base the correlations on parameters which are ideally structure-independent and can be collected either from compilations in the literature or are easily available hy own measurements. Actually, the extrathermodynamic constants o and TZ have proved their usefulness in many structure-activity correlations 38 , in spite of considerable multiplicity in o 39 ' 40 and, to a lesser extent, in JI. More recently, examples of excellent correlations with quantum chemical indexes have appeared in 1. c. 41 ~ 43 , thus apparently terminating a period of numerous less successful approaches. Quantum chemical substitutes for TI and o have been developed, but the two extrathermodynamic parameters are probably still the more preferable ones for correlation studies. As regards multiplicity, the choice of quantum chemical indexes is considerable and it has yet to be established, which ones are of general importance for biological activity. The amount of work needed in each single instance to calculate them, is often greater than that needed for e. g. pK- or log P-determinations. RM

exp. 1

0.86 0.69

pK * •logP

p K * • i?M r

exp. 2

+ 0.05 0.06 +

-

165

0.04 0.40

+ 0.25 0.37

+ 0.03 0.06 -

2.56 1.67

-

+

0.16 0.12

0.13 0.42

+ 0.58 0.36

s

0.664

0.335

0.738

0.306

0.750

0.302

0.691

0.330

0.782

0.287

0.749

0.309

Table 5. Regression equations with molecule parameters. Notes on the representation of regression equations in tabulated f o r m : The left side of the equations, p/ 5 0 , is omitted. 1 and 2 denote the first and the second power of the variable on top of the column. 3 is the coefficient of the variable, or in case of a 0 , the intercept. 4 is the 95%-confidence interval of the coefficient. For example, in explicit form, eq. (20) reads: p/ 5 0 = 5.36 (4.83) + 0 . 9 5 ( 1 . 5 5 ) - p K - 0 . 1 6 ( 0 . 1 2 ) - p K 2 - 0 . 8 6 ( 0 . 6 9 ) -log P + 0 . 0 3 ( 0 . 0 6 ) • (log P) 2 + 0.16 (0.12) - p K - l o g P.


166

W . D R A B E R , K . H. BÜCHEL, A N D G. S C H Ä F E R

i. no.

00

+ 5.6G 0.19 + 5.52 0.29

22 23

24

+ 5.49 0.31 + 5.41 0.37

25

Electronic Parameters 02-6, R exp. 1 exp. 2

H y d r o p h o b i e Parameters ^2-6, R exp. 1 exp. 2

+ 0.48 0.29 + 0.45 0.29 02-6 exp. 1 + 0.69 0.83 + 0.68 0.83

+ 0.31 0.09 + 0.50 0.29 712-6 exp. 1 + 0.36 0.14 + 0.51 0.46

-

exp. 2 -0.28 0.45 -0.26 0.46

0.28 0.28 -0.26 0.28 0R exp. 1 + 0.22 0.45 + 0.24 0.45

exp. 2 - 0.04 0.71 - 0.01 0.72

-

XR exp. 1 + 0.26 0.14 + 0.26 0.14

exp. 2

-

0.05 0.07

0.07 0.18

r

s

0.783

0.27

0.791

0.27

0.783

0.27

0.785

0.28

Table 6. Regression equations with substituent paratemers.

(o 2 ) 6 ) from the others, as shown in eq. ( 2 6 ) . The hydrophobic parameters at the same time became more significant. When o 2 6 was replaced by T a f t ' s £ s -values, modified according to KuTTER and HANSCH 44 , the correlation was not improved. This is not surprising since o2>6 and Es are highly intercorrelated within the given set of compounds (r = 0 . 9 1 ) . In eqs. (26) and ( 2 7 ) , the electronic and steric parameters in o-position are of little significance as shown by the 95%-confidence intervals. When those were omitted, a better correlation was obtained, which was further improved by separation of the rc-terms (eq. 28) and a ;r2>62-term ( e 1- 2 9 ) . The latter equation was very slightly improved by introduction of a 7r3_52-term. This four-parameter equation gave the best agreement of experimental and calculated data:

Correlations with substituent parameters Table 6 shows a number of regression equations with substituent parameters. Replacement of pK and l o g P in eqs. (17) and (18) by the equivalent substituent parameters without separation results in some improvement of the correlation. Nothing is gained by separation of the substituents at the aromatic ring and the side chain. A consistent feature of eqs. (22) - (25) is the negative sign of the o2-terms. The significance of several coefficients is low, particularly in the four-parameter eqs. (24) and ( 2 5 ) . From these results it was inferred that further separation of the substituents was necessary. Some steps are shown in Table 7. Substantial improvement was achieved by separation of the electronic parameter of the o-substituents p/50=

5.813 + 0.515 a 3 - 5 R - 0 . 8 0 4 0 3 - 5 . R 2 (0.149) (0.163) (0.299)

+0.306ji3-5 (0.323)

+ 0 . 0 5 0 t i 3 - 5 2 - 0 . 1 3 4 j i 2 , 6 + 0 . 3 6 0 7r 2 ,c 2 + 0 . 1 8 9 t i r (0.104) (0.294) (0.303) (0.083)

r = 0.921 s = 0.172

(30)

Eq. no.

ao

Electronic Parameters 03-5, R 02,6 exp. 1 exp. 2 exp. 1

26

+ 5.79 0.18 + 5.68 0.17

+ 0.54 0.19 + 0.53 0.19

27

-0.39 0.28 -0.40 0.23

-

0.58 0.51

28

+ 5.83

29

+ 5.78

+ 0.60 0.59

+ 0.36 0.10 + 0.36 0.09

713-5

03-5, R exp. 1 + 0.56 0.16 + 0.51 0.16

exp. 2

H y d r o p h o b i c / S t e r i c Parameters 712-6 tir Es 2,6 exp. 1 exp. 1 exp. 1 exp. 2

exp. 2 - 0.82 0.30 - 0.77 0.29

exp. 1 + 0.38 0.08 + 0.41 0.08

+ 0.22 0.11 + 0.22 0.10

— 0.13 0.13

exp. 1 + 0.18 0.12 - 0.15 0.29

Table 7. Separation of substituent paratemers.


0.22

0.877

0.21

0.910

0.18

0.920

0.17

^2,6

TlR exp. 1 +0.17 0.09 + 0.19 0.08

+0.07 0.05

0.871

exp. 2

+0.37 0.30

STRUCTURE-ACTIVITY STUDIES OF HYDRAZONES

No.

X2

X3

X4

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

H H H H H H H H H H H H CL H CL CL H H CH 3 N02 H CF 3 H H CL CL CL CF 3 CL CL S0 2 Et CF 3 H N02 H H CL H S0 2 Me H H H CL CH 3 CL H CL CL CL S0 2 Et H N02 H N02 H CF 3 CL CN CL H

H CF 3 CL CL CF3 CL H CF3 CF3 CL CL H H H H H H CL H H CL H CL CL H CN H H H H H H CL H CF3 CL H H H H H H H H H CHF 2 H H H H CF 3 H CL H CL H H H H H

OCF 2 - O - C F 2 CF3 H CL CL H CL CF3 H H CL OCF 2 - O - C F 2 H CF3 H CL CL CL H CL H SCF 3 CF3 H OCF2- O-CF2 CF3 H CL CL SCF3 H H H H CL H NOo H CL H CL H CL H CL CL CL H CL CF 3 H H CL CF 3 H H CF 3 H CF3 H CL H CL CF 3 H CF3 H H H CL CL SO 2 CF 3 H H NO 2 H CL SCF3 H H CF 3 CL CL CL H H NO 2 H CL H CF3 CL CL H NO 2 H CF3 H H NOo H H H H NO 2 H H H H H H CN H H H H H

X5

X6

H H H H H H H H H H H H H H H H H H H H H H H H H CL H H CL CL H H H H H H H H H H H H H H CL H H H CL H H H H H H H H H H H

R