World J Microbiol Biotechnol DOI 10.1007/s11274-014-1755-4
ORIGINAL PAPER
Synthesis and quantitative structure activity relationship (QSAR) of arylidene (benzimidazol-1-yl)acetohydrazones as potential antibacterial agents Yeldez El-Kilany • Nariman M. Nahas • Mariam A. Al-Ghamdi • Mohamed E. I. Badawy El Sayed H. El Ashry
•
Received: 16 June 2014 / Accepted: 3 October 2014 Ó Springer Science+Business Media Dordrecht 2014
Abstract Ethyl (benzimidazol-1-yl)acetate was subjected to hydrazinolysis with hydrazine hydrate to give (benzimidazol-1-yl)acetohydrazide. The latter was reacted with various aromatic aldehydes to give the respective arylidene (1H-benzimidazol-1-yl)acetohydrazones. Solutions of the prepared hydrazones were found to contain two geometric isomers. Similarly (2-methyl-benzimidazol-1-yl)acetohydrazide was reacted with various aldehydes to give the corresponding hydrazones. The antibacterial activity was evaluated in vitro by minimum inhibitory concentration (MIC) against Agrobacterium tumefaciens (A. tumefaciens), Erwinia carotovora (E. carotovora), Corynebacterium fascians (C. fascians) and Pseudomonas solanacearum (P. solanacearum). MIC result demonstrated that salicylaldehyde(1H-benzimidazol-1-yl)acetohydrazone (4) was the most active compound (MIC = 20, 35, 25 and 30 mg/L against A. tumefaciens, C. fascians, E. carotovora and P. solanacearum, respectively). Quantitative structure activity relationship (QSAR) investigation using Hansch analysis was applied to find out the correlation between antibacterial Electronic supplementary material The online version of this article (doi:10.1007/s11274-014-1755-4) contains supplementary material, which is available to authorized users. Y. El-Kilany N. M. Nahas M. A. Al-Ghamdi Chemistry Department, Faculty of Applied Science, Umm Al-Qura University, Mekkah, Kingdom of Saudi Arabia M. E. I. Badawy (&) Department of Pesticide Chemistry and Technology, Faculty of Agriculture, Alexandria University, El-Shatby, Alexandria 21545, Egypt e-mail: [email protected] E. S. H. El Ashry Chemistry Department, Faculty of Science, Alexandria University, Alexandria, Egypt
activity and physicochemical properties. Various physicochemical descriptors and experimentally determined MIC values for different microorganisms were used as independent and dependent variables, respectively. pMICs of the compounds exhibited good correlation (r = 0.983, 0.914, 0.960 and 0.958 for A. tumefaciens, C. fascians, E. carotovora and P. solanacearum, respectively) with the prediction made by the model. QSAR study revealed that the hydrophobic parameter (ClogP), the aqueous solubility (LogS), calculated molar refractivity, topological polar surface area and hydrogen bond acceptor were found to have overall significant correlation with antibacterial activity. The statistical results of training set, correlation coefficient (r and r2), the ratio between regression and residual variances (f, Fisher’s statistic), the standard error of estimates and significant (s) gave reliability to the prediction of molecules with activity using QSAR models. However, QSAR equations derived for the MIC values against the tested bacteria showed negative contribution of molecular mass. Keywords Hydrazones Geometric isomers Benzimidazole Benzimidazolyl-acetohydrazones Antibacterial activity QSAR
Introduction The resistance to antimicrobial agents is widespread, therefore, the development of new antimicrobial agents and understanding their mechanisms of action are becoming vital nowadays (Foroumadi et al. 2003). The discovery of the presence of 5,5-dimethyl-1-(a-D-ribofuranosyl)benzimidazole as an integral part of the structure of vitamin B12, led to the consideration of the benzimidazole a promising pharmacophore (Shah et al. 2013), that play an important role in
123
World J Microbiol Biotechnol
medicinal chemistry. The potential biological activity of benzimidazole derivatives as lead compounds has attracted much attention and led to the synthesis of variety of compounds having the benzimidazole ring. The incorporation of the benzimidazole nucleus is an important synthetic strategy in studies of antimicrobial drug discovery (Sharma et al. 2009; Khalafi-Nezhad et al. 2005). Benzimidazole and its derivatives have received much attention because of their biological activity and commercial application and they are considered as bacteriostatic and bactericidal (Chaudhary et al. 2011; Ingle and Magar 2011). Benzimidazole is present in many naturally occurring products and various drugs. Some of these compounds have antibacterial, antifungal, antiviral, antiinflammatory, antihypertensive, arteriosclerosis and anti-HIV activities (Chohan and Hanif 2013; Goker et al. 1999; Kalinowska-Lis et al. 2014; Podunavac-Kuzmanovic´ et al. 2002). During the last two decades, quantitative structure–activity relationship (QSAR) models have gained an extensive recognition in physical, organic, analytical, pharmaceutical and medicinal chemistry. QSAR models are highly effective in describing the structural basis of biological activity. The success of QSAR approach can be explained by the insight offered into the structural determination of chemical properties and the possibility to estimate the properties of new chemical compounds without the need to synthesize and test them (Hansch and Fujita 1964; Pal et al. 2011; Khatkar et al. 2013). Our thrust for the search of some newer and potent antimicrobial agents against bacteria, which is driven by the fact that the compound resistance is the major problem in treatment of diseases, prompted us to carry out QSAR studies of the molecules mentioned in above series (Gupta et al. 2002; Hansch and Fujita 1964; Kumar et al. 2007; Narasimhan et al. 2007). In the present study, we performed 2-D QSAR on a series of arylidene (benzimidazol-1-yl)acetohydrazones compounds to establish the relationship with the antibacterial activity against A. tumefaciens, C. fascians, E. carotovora and P. solanacearum. The resultant QSAR models were validated using multiple linear regression (MLR) using SPSS software. It was observed that molecular mass, polarizability, and lipophilicity parameters were major contributors to the antibacterial activity.
Materials and methods
N
N
R"
N
Aldehyde
H2C
H2C
O
H N
N
R"
N
H N
H
H
O Ar
N
H
OH
1 R" = H 2 R" = Me N
N
R"
N H2C Ar
H2C
O
N
N
H H
H
R R' R" = H R' H H OH OMe
N
Ar
E
R 3H 4 OH 5H 6H
R"
N
R
N
O H
Z
O
R' R" = Me R R' H 7H 8 OH H OH 9H
10 R" = H 11 R" = Me
Scheme 1 Synthesis of arylidene benzimidazolyl-acetohydrazones
monitored by TLC using Merck silica gel 60. Melting points were determined in open capillaries and are uncorrected. IR spectra were recorded on Fourier transform infrared 8,400 spectrophotometer with KBr pellets. 1H and 13 C (Proton decoupled) NMR spectra were recorded on Jeol JNM ECA 500 MHz using TMS as an internal standard (chemical shifts in d, ppm). The number of protons will be given as (c) when it is a fraction of a proton. The fraction of a proton is shown by X1. Satisfactory C, H, and N analyses were obtained for the compounds at the micro analytical laboratory of Cairo-University, Egypt. Four bacteria strains, Agrobacterium tumefaciens (Family: Rhizobiaceae), Erwinia carotovora (Family: Enterobacteriaceae), Corynebacterium fascians ((Family: Nocardiaceae) and Pseudomonas solanacearum (Family: Pseudomonadaceae) were provided by Microbiology Laboratory, Department of Plant Pathology, Faculty of Agriculture, Alexandria University, Egypt and were maintained on nutrient agar (NA) medium at 37 °C.
Materials and test microorganisms Antibacterial evaluation All reagents and solvents used in the study were of analytical grade and procured locally. Purification of compounds was made by column chromatography using Merck silica gel 60 (230–400 mesh) and a mixture of hexane and ethyl acetate was used for elution. All the reactions were
123
The in vitro antibacterial activities as minimum inhibitory concentration (MIC) of the compounds were determined by microdilution broth assay method using 2,3,5,-triphenyltetrazolium chloride (TTC, Sigma) as chromogenic marker
World J Microbiol Biotechnol Table 1 In vitro antibacterial activity of arylidene (benzimidazol-1yl)acetohydrazones derivatives against A. tumefaciens, C. fascians, E. carotovora and P. solanacearum
indicating MIC. A change in color from colorless to red formazan shows the growth of the organism.
Compound
Quantitative structure activity relationship analysis
1
MIC (mg/L) A. tumefaciens
C. fascians
E. carotovora
P. solanacearum
400
700
300
450
2
550
[1,000
[1,000
[1,000
3
75
400
200
350
4
15
45
30
35
5
20
35
25
30
6
75
500
275
400
7
200
475
250
425
8
50
300
100
125
9
125
375
175
200
10
45
350
75
150
11
150
425
225
250
MIC is the minimum inhibitory concentration obtained for each microorganism
(Thom et al. 1993; Yajko et al. 1995). Nutrient broth (NB) medium was used to grow the bacterial strains to a final inoculum size of 5 9 105 CFU/mL. The arylidene (benzimidazol-1-yl)acetohydrazones were dissolved in dimethyl sulfoxide (DMSO) and diluted with distilled water to obtain a final stock solution of 1,000 mg/L. For the broth microdilution test, 20 lL of each bacterial suspension in NB medium was added to the wells of a sterile 96-well microtitre plate already containing 40 lL of serially diluted arylidene (benzimidazol-1-yl)acetohydrazone compounds and 140 lL NB medium. The final volume in each well was 200 lL. Control wells were prepared with culture medium, bacterial suspension only, arylidene (benzimidazol-1-yl)acetohydrazone compounds only and DMSO in amounts corresponding to the highest quantity present. The contents of each well were mixed on a microplate shaker at 200 rpm for 1 min prior to incubation for 24 h in the cultivation conditions described above. The MIC was the lowest concentration where no viability was observed after 24 h on the basis of metabolic activity. To indicate respiratory activity the presence of color was determined after adding 10 lL/well of TTC dissolved in water (0.01 %, w/v) and incubated under appropriate cultivation conditions for 30 min in the dark. The absorbance was measured at 492 nm in an ultra microplate reader (Robonik, Pvt. Ltd). Positive controls were wells with a bacterial suspension in NB and a bacterial suspension in NB medium with DMSO in amounts corresponding to the highest quantity present in the broth microdilution assay. Negative controls were wells with growth medium and arylidene (benzimidazol-1-yl)acetohydrazone compounds. All measurements of MIC values were repeated in triplicate. The color change was assessed visually, with the highest dilution remaining colorless (inhibition of growth)
To develop QSAR, one needs to select a few descriptors from large physiochemical properties. The selected descriptors were used as independent variables to create the regression equations with activities as the dependent variable. Molecular weight (MW), CMR and topological polar surface area (tPSA) were computed by ChemDraw Ultra 11.0 software package. Calculated hydrophobic parameter (ClogP, logarithm of partition coefficient) was calculated by EPI SuiteTM v4.11 on the basis of the work of Hansch and Fujita (1964). The aqueous solubility of a compound (LogS) and hydrogen bond acceptors (HA) were calculated by using ChemSpider (http://www. chemspider.com/). The multiple linear regression model tested in this work was used to study the correlation between dependant variable (the biological activity expressed as MIC) and independent variables (MW, ClogP, logS, CMR, tPSA and HA). The regression analysis was performed using SPSS v.17 software (SPSS Inc., Chicago, IL, USA). Equations were justified by the correlation coefficient (r and r2), the standard error of estimates (s) and the value of the ratio between regression and residual variances (f, Fisher’s statistic). Only those parameters having good correlation coefficient and low standard error were considered to determine best equation. Regression was built using descriptor subsets containing only one of these highly correlated descriptors. Multiple Linear Regression (MLR) analysis was carried out to find out the factors responsible for variation in the biological activity. Successive regression equations were derived in which parameters are added, removed, or replaced until r2 and s values are optimized. To derive QSAR models, stepwise MLR analysis with cross validation technique was applied to both sets of ten compounds. Models with number of descriptors\7, f ratio higher than 2 and correlation coefficient, r more than 0.90 between predicted and the experimental antibacterial activities were validated. Significant descriptors were chosen based on the statistical data of analysis. Intercorrelation between the descriptors was checked for independence of the variables. The QSAR models with high statistical significance are reported herein.
Results and discussion Synthesis and spectroscopic characterizations of arylidene (benzimidazol-1-yl)acetohydrazones derivatives Reaction of (1H-benzimidazol-1-yl)acetohydrazine 3 with benzaldehyde, salicylaldehyde, p-hydroxybenzaldehyde,
123
World J Microbiol Biotechnol Table 2 Values of physiochemical molecular descriptors for arylidene (benzimidazol-1-yl)acetohydrazones derivatives used for QSAR analysis against A. tumefaciens Compound
PMIC
Residual
Values of descriptors
Actual
Predicted
MW
Clog P
LogS
CMR
TPSA
HA
1
2.68
2.82
-0.1403
2
2.57
2.41
0.1581
190.2 204.23
-0.12 0.152
1.91
5.3
70.72
5
2.14
5.76
70.72
3 4
3.57 4.29
3.61 4.28
-0.0365 0.0056
278.31 294.31
5
2.36 2.96
3.57 3.31
8.4 8.55
57.06 77.29
5 6
5
4.17
4.03
0.1354
294.31
2.33
3.34
8.55
77.29
6
6
3.61
3.54
7
3.16
3.14
0.0747
308.33
2.58
3.68
9.01
66.29
6
0.0187
292.34
2.63
3.74
8.86
57.06
8
3.79
5
3.80
-0.0065
308.33
3.23
3.45
9.01
77.29
9
6
3.39
3.55
-0.1567
308.33
2.6
3.48
9.01
77.29
6
10
3.78
3.76
0.0216
268.27
1.53
3.48
7.61
66.29
6
11
3.27
3.34
-0.0742
282.3
1.8
3.55
8.08
66.29
6
MIC minimum inhibitory concentration, pMIC -Log MIC, MW Molecular weight, ClogP Calculated hydrophobic parameter (logarithm of partition coefficient), LogS The aqueous solubility of a compound, CMR Calculated molar refractivity, tPSA Topological polar surface area, HA Hydrogen bond acceptors
Fig. 1 A Plot of predicted pMIC of arylidene (benzimidazol-1yl)acetohydrazones derivatives (n = 11) activity for A. tumefaciens against the experimental pMIC values for the linear regression developed model by Eq. 1. B Plot of residual pMIC against the experimental pMIC values
anisaldehyde or furfuraldehyde gave the respective hydrazones 3–6 (Scheme 1). The solution of 3 in DMSO-d6 showed the presence of two species as indicated from the presence of two signals for CH2 at d 5.07 and 5.54 and two signals for
123
H–C=N at d 8.20 and 8.21. Moreover 13C NMR showed two signals for C=N at dc 144.5 and 144.7. These data indicated that the hydrazone product was a mixture of the geometric isomers E and Z in a ratio 4:1, respectively. The selection of the higher ratio for E has been based their possible relative stability as a consequence of the crowded environment around the C=N in Z compared to E. Consequently, the expected intermediate gave preferably the isomer E over Z. The salicylaldehyde (1H-benzimidazol-1-yl)acetohydrazone 4 showed two signals for H–C=N at d 8.34 and 8.45 and two signals for C=N at dc 147.9 and 148.1, but the ratio of E to Z was 2:1, respectively. Similarly, the ratios of E and Z for 4-hydroxybenzaldehyde (1H-benzimidazol-1-yl)acetohydrazone 5 and anisaldehyde (1H-benzimidazol-1-yl)acetohydrazone 6 were found to be 4:1 and 3:1, respectively. Reaction of benzaldehyde, salicylaldehyde and 4-hydroxybenzaldehyde with (2-methyl-1H-benzimidazol1-yl)acetohydrazine gave the respective hydrazones 7–9 (Scheme 1) whose solutions in DMSO-d6 indicated their presence as mixture of the geometric isomers E and Z in a ratio 4:1, 2:1 and 3:1, respectively. The selection of the higher ratio for E has been based on their possible relative stability as a consequence of the crowded environment around the C=N in Z compared to E. The furfural (1Hbenzimidazol-1-yl)acetohydrazone 10 and (2-methyl-1Hbenzimidazol-1-yl)acetohydrazone 11 were having also a mixture of the two geometric isomers in the ratio 3:1. The in vitro antibacterial activity and QSAR study The antibacterial properties of arylidene (benzimidazol-1yl)acetohydrazone derivatives were performed in vitro by a
World J Microbiol Biotechnol
Fig. 2 A Plot of predicted pMIC of arylidene (benzimidazol-1yl)acetohydrazones derivatives (n = 10) activity for C. fascians against the experimental pMIC values for the linear regression developed model by Eq. 2. B Plot of residual pMIC against the experimental pMIC values
conventional microdilution broth assay method against A. tumefaciens, C. fascians, E. carotovora and P. solanacearum and the MIC values are given in Table 1. Salicylaldehyde(1H-benzimidazol-1-yl)acetohydrazone (4) having ortho –OH group on the phenyl ring was the most active compound among the tested compounds with MIC values of 15, 45, 30 and 35 mg/L against A. tumefaciens, C. fascians, E. carotovora and P. solanacearum, respectively. Followed in the descending order by 4-hydroxybenzaldehyde(1H-benzimidazol-1-yl)acetohydrazone (5) (MIC = 20, 35, 25 and 30 mg/L against A. tumefaciens, C. fascians, E. carotovora and P. solanacearum, respectively). This result confirms that the substitution with a hydroxyl group on the ortho or para positions on the phenyl ring enhanced the biological activity compared to other compounds. However, a substitution of the hydrogen by a methyl group on the benzimidazol led to decrease the activity (see compounds 1, 3, 4 and 5 vs. 2, 7, 8 and 9, respectively). 2-(2-Methyl-benzimidazol-1-yl)acetohydrazide (2) was the least active (MIC = 550 mg/L for A. tumefaciens, and higher than 1,000 mg/L for C. fascians, E. carotovora and P. solanacearum). When we consider the susceptibility
Fig. 3 A Plot of predicted pMIC of arylidene (benzimidazol-1yl)acetohydrazones derivatives (n = 10) activity for E. carotovora against the experimental pMIC values for the linear regression developed model by Eq. 3. B Plot of residual pMIC against the experimental pMIC values
of the microorganisms, another point deserves awareness; A. tumefaciens and E. carotovora were more sensitive to the tested compounds than C. fascians and P. solanacearum. For QSAR analysis, the respective MIC values in mg/L of these compounds were converted to mole/L and then to -log MIC (pMIC), where MIC value is the concentration of the compound required for 100 % inhibition of the growth of the microorganisms. Based on these values, statistically significant equations were generated. Different sets of equations were produced for A. tumefaciens, C. fascians, E. carotovora and P. solanacearum as follows. The various physiochemical molecular descriptors computed for investigations were molecular weight (MW), calculated hydrophobic parameter (logarithm of partition coefficient, ClogP), the aqueous solubility of a compound (LogS), calculated molar refractivity (CMR), topological polar surface area (tPSA) and hydrogen bond acceptor (HA). For A. tumefaciens, QSAR was run for 11 compounds and statistical parameters like r, r2, f, s and residual sum of squares were calculated (Table 2). They were found statistically significant and the regression analysis was run to yield the following equation:
123
World J Microbiol Biotechnol
s = 0.007) and high coefficient constant was obtained with CMR (14.421) compared to r = 0.737 in the case of MW. Further the plot of linear regression experimental pMIC values against the predicted pMIC values for A. tumefaciens also favors the model expressed by Eq. 1 as shown in Fig. 1A. To investigate the existence of a systemic error in developing the QSAR model, we have plotted pMIC observed against pMIC residual values (Fig. 1B). The propagation of the residuals on both sides of zero indicates that there is no systemic error in the development of linear regression model (Heravi and Kyani 2005). For C. fascians, variations in the biological activity of arylidene (benzimidazol-1-yl)acetohydrazone derivatives were analyzed using the physiochemical descriptors (Table S2) and the obtained equation with correlation coefficient was listed in Eq. 2.
Fig. 4 A Plot of predicted pMIC of arylidene (benzimidazol-1yl)acetohydrazones derivatives (n = 10) activity for P. solanacearum against the experimental pMIC values for the linear regression developed model by Eq. 4. B Plot of residual pMIC against the experimental pMIC values
pMIC ¼ 526ð0:113Þ MW þ 0:441ð 0:248Þ ClogP þ 0:932ð 0:702Þ LogS þ 14:421ð3:325Þ CMR þ 0:017ð0:028Þ tPSA þ 6:557ð1:430Þ HA 9:339ð2:898Þ ð1Þ where, number of compounds, n = 11, correlation coefficient, r = 0.983 and r2 = 0.966, Fisher test, f = 18.934, Residual sum of squares, RSS = 0.101; standard error of the estimate, SE = 0.158 and significant, s = 0.007. Equation 1 was found to be better model for antibacterial activity of this series of compounds against A. tumefaciens. This revealed that the molecular descriptors have significant effect on biological activity. All these parameters correlate positively with biological activity except MW indicating that the increase in molecular mass led to decrease the antibacterial activity. Table S1 shows the development of the QSAR Eq. 1 and all the calculated parameters were subjected to correlation studies with depending on their individual correlation with the biological activity. In this group, a combination of all descriptors significantly enhanced the biological activity where a high correlation was obtained (r = 0.983 and
123
pMIC ¼ 0:638ð0:280Þ MW þ 0:045ð0:520Þ ClogP þ 1:202ð1:482Þ LogS þ 17:893ð8:203Þ CMR þ 0:051ð0:062Þ tPSA þ 7:392ð3:617Þ HA 13:854ð6:948Þ ð2Þ where n = 10, r = 0.914, r2 = 0.835, f = 2.532, RSS = 0.331, SE = 0.332 and s = 0.238. The correlation coefficient of this equation is satisfied (r = 0.914) using the entire data set (n = 10). The positive coefficient of the Clog P, Log S, CMR, tPSA and HA indicate that an increase of these descriptors led to increase in the antibacterial activity. However, the negative coefficient of the MW indicates the high MW, the lower the activity against the bacterium. Table S3 shows the development of the QSAR Eq. 2 and all the calculated parameters were subjected to correlation studies with depending on their individual correlation with the biological activity. In this group, HA significantly enhanced the biological activity where a high correlation was obtained (r = 0.914) compared to r = 0.778 in the case of all descriptors except HA. The observed pMIC versus calculated pMIC values according to Eq. 2 were plotted in Fig. 2A and plot of pMIC observed against pMIC residual values is shown in Fig. 2B. The values of physiochemical molecular descriptors for arylidene (benzimidazol-1-yl)acetohydrazone compounds used for QSAR analysis against E. carotovora and actual and predicted pMIC with residuals are indicated in Table S4. Statistical parameters including r, r2, f, s and residual sum of squares were calculated and they were found statistically significant and the regression analysis was run to yield following QSAR equation:
World J Microbiol Biotechnol
pMIC ¼ 0:494ð0:171Þ MW þ 0:212ð0:316ÞClogP þ 1:925ð0:902Þ LogS þ 13:328ð4:993Þ CMR þ 0:070 ð0:038Þ tPSA þ 5:519ð2:201Þ HA 10:039ð4:229Þ ð3Þ 2
where n = 10, r = 0.960, r = 0.922, f = 5.919, RSS = 0.123, SE = 0.202 and s = 0.086. A combination of all molecular descriptors led to a significant increase in the biological correlation (r = 0.960) and high coefficient constants were obtained with CMR (13.328) and HA (5.519). The positive coefficient of Clog P, Log S, CMR, tPSA and HA indicate that an increase of these descriptors led to increase in the antibacterial activity. However, the negative coefficient (-0.494) of the MW indicates the high molecular mass, the lower the activity against the bacterium. Table S5 shows the development of the QSAR (Eq. 3) and all the calculated parameters were subjected to correlation studies with depending on their individual correlation with the biological activity. In this group, a combination of tPSA and HA with other descriptors significantly enhanced the biological activity where the correlation was increased from 0.695 to 0.960 with low standard deviation 0.086. The observed pMIC versus calculated pMIC values according to Eq. 2 were plotted in Fig. 3A and plot of pMIC observed against pMIC residual values is shown in Fig. 3B. QSAR analysis for P. solanacearum was run for ten compounds and the pMIC observed and calculated with the residuals and molecular descriptors for arylidene (benzimidazol-1-yl)acetohydrazone derivatives are shown in Table S6. Statistical parameters including r, r2, f and residual sum of squares were calculated and the regression analysis was run to yield following equation: pMIC ¼ 0:459ð0:190Þ MW þ 0:179ð0:353ÞClogP þ 1:753ð1:006Þ LogS þ 12:433ð5:566Þ CMR þ 0:074ð0:042Þ tPSA þ 5:159ð2:424Þ HA 10:180 ð4:715Þ ð4Þ where n = 10, r = 0.958, r2 = 0.917, f = 5.548, RSS = 0.152, SE = 0.225 and s = 0.094. Equation 4 was found to be a better model for antibacterial activity of this series of compounds against P. solanacearum (r = 0.958). This equation revealed that all descriptors contribute significantly to the biological activity. ClogP, LogS, CMR, tPSA and HA correlate positively with biological activity which emphasizes that increase in these parameters contribute to increase in biological activity. However, MW correlates negatively (-0.459) with biological activity which emphasizes that increase in
mass of molecule contributes to decrease in biological activity. Table S7 shows the development of the QSAR Eq. 4. A combination of CMR, tPSA and HA with other descriptors (MW, ClogP and LogS) significantly enhanced the biological activity where the correlation was increased from 0.605 to 0.958 with the lowest standard deviation 0.094. Correlation graph between actual and predicted MIC values was also generated (Fig. 4A) along with the plot of pMIC observed against pMIC residual values (Fig. 4B). In the present study, Clog P, LogS, CMR, tPSA and HA are very useful parameter for prediction of the biological activity and drug properties. CMR and HA were significantly the highest correlation found with all tested bacteria. For the arylidene (benzimidazol-1-yl)acetohydrazone derivatives, the QSAR regressions suggest that the same structural features describe the activities for bacteria. These results agree with those found for the QSAR of 5-nitroimidazole analogues and other related compounds (Gupta et al. 2002; Kumar et al. 2007; Narasimhan et al. 2007). A positive contribution of MR with activity against Escherichia coli was also reported (Gupta et al. 2002). In this study, it is seen that compounds with higher CMR showed higher antibacterial activity. A series of 2-(substituted phenyl)-1Hbenzimidazole and [2-(substituted phenyl)-benzimidazol-1yl]-pyridin-3-yl-methanone derivatives and tested in vitro for their antimicrobial activity was synthesized (Sharma et al. 2009). The results of QSAR investigation indicated the importance of molecular descriptors, dipole moment (m), log of octanol water partition coefficient (log P) and second order molecular connectivity index in describing the antimicrobial activity of the synthesized compounds. QSAR studies for some disubstituted-1,3,4-oxadiazole derivatives were tested and the in vitro antibacterial activity (MICs) of the compounds against Staphylococcus aureus and E. coli 2 exhibited good correlation (n = 36, r2 [ 0.7, rcv [ 0.5, s = 0.069) with the prediction made by the model of Pal et al. (2011). It was found that the polarizability, thermodynamic, and lipophilicity were major responsible factors for exhibiting the activity. Recent literature reveals that the QSAR has been applied to describe the relationship between narrow range of biological activity and physiochemical properties of the molecules. When biological activity data lie in a narrow range, the presence of minimum standard deviation of the biological activity justifies its use in QSAR studies (AbdelRahman et al. 2013; Kumar et al. 2007; Narasimhan et al. 2007; Nofal et al. 2013; Sharma et al. 2009). The minimum standard deviation (Eqs. 1, 2, 3 and 4) observed in the antimicrobial activity data justifies its use in QSAR studies. In conclusion, the models obtained through this QSAR study gave a better prediction capability of antibacterial activity against bacteria like A. tumefaciens, C. fascians,
123
World J Microbiol Biotechnol
E. carotovora and P. solanacearum. Clog P, Log S, CMR, tPSA and HA were the major factors responsible for positively affecting the antibacterial activity of these arylidene (benzimidazol-1-yl)acetohydrazones derivatives. The QSAR models are of immense importance for optimizing the biological activity of benzimidazol analogues, which could help in designing better molecules with enhanced anti-bacterial activity. The results suggest that all the newly synthesized compounds showed good to very good activity. Hence the fact that the compounds prepared in this study are chemically related to the current medication, suggests that further work with similar analogues is clearly warranted.
References Abdel-Rahman AAH, Nassar IF, El-Kattan IMH, Aly AA, Behalo MS, Abdelwahed NM (2013) Synthesis and antimicrobial evaluation of thioglycosides and acyclic C-nucleosides of 2-methylbenzimidazole. Der Pharma Chemica 5:210–217 Chaudhary M, Pareek D, Pareek PK, Kant R, Ojha KG, Pareek A (2011) Synthesis of some new biologically active benzothiazole derivatives containing benzimidazole and imidazoline moieties. Bull Korean Chem Soc 32:131–136 Chohan ZH, Hanif M (2013) Antibacterial and antifungal metal based triazole Schiff bases. J Enz Inhib Med Chem 28:944–953 Foroumadi A, Mansouri S, Kiani Z, Rahmani A (2003) Synthesis and in vitro antibacterial evaluation of N-[5-(5-nitro-2-thienyl)-1,3,4thiadiazole-2-yl] piperazinyl quinolones. Eur J Med Chem 38:851–854 Goker H, Ayhankilcilgil G, Tuncbilek M, Kus C, Ertan R, Kendi E, Oybey S, Fort M, Garcia C, Farre A (1999) Synthesis and antihistaminic Hl activity of 1,2,5(6)- trisubstituted benzimidazoles. J Heterocycl 51:2561–2573 Gupta MK, Mishra P, Prathipati P, Saxena AK (2002) 2D-QSAR in hydroxamic acid derivatives as peptide deformylase inhibitors and antibacterial agents. Bioorg Med Chem 10:3713–3716 Hansch C, Fujita T (1964) p–r–p analysis. A method for the correlation of biological activity and chemical structure. J Am Chem Soc 86:1616–1626 Heravi JM, Kyani A (2005) A use of computer-assisted methods for the modeling of the retention time of a variety of volatile organic compounds: a PCA-MLR-ANN approach. J Chem Inf Comput Sci 44:1328–1335
123
Ingle RG, Magar DD (2011) Heterocyclic chemistry of benzimidazoles and potential activities of derivatives. Int J Drug Res Tech 1:26–32 Kalinowska-Lis U, Felczak A, Che˛cin´ska L, Lisowska K, Ochocki J (2014) Synthesis, characterization and antimicrobial activity of silver(I) complexes of hydroxymethyl derivatives of pyridine and benzimidazole. J Organomet Chem 749:394–399 Khalafi-Nezhad A, Rad MNS, Mohbatkar H, Asrari Z, Hemmateenejad B (2005) Design, synthesis, antibacterial and QSAR studies of benzimidazole and imidazole chloroaryloxyalkyl derivatives. Bioorg Med Chem 13:1931–1938 Khatkar A, Nanda A, Kumar P and Narasimhan B (2013) Synthesis, antimicrobial evaluation and QSAR studies of gallic acid derivatives. Arab J Chem. doi:10.1016/j.arabjc.2013.11.014 Kumar A, Narasimhan B, Kumar D (2007) Synthesis, antimicrobial, and QSAR studies of substituted benzamides. Bioorg Med Chem 15:4113–4124 Narasimhan B, Judge V, Narang R, Ohlan R, Ohlan S (2007) Quantitative structure–activity relationship studies for prediction of antimicrobial activity of synthesized 2,4-hexadienoic acid derivatives. Bioorg Med Chem Lett 17:5836–5845 Nofal ZM, Srour AM, El-Eraky WI, Saleh DO, Girgis AS (2013) Rational design, synthesis and QSAR study of vasorelaxant active 3-pyridinecarbonitriles incorporating 1H-benzimidazol-2yl function. Eur J Med Chem 63:14–21 Pal T, Somani RR, Shirodkar PY, Bhanushali UV (2011) Quantitative structure activity relationship (QSAR) studies of a series of 2,5disubstituted-1,3,4-oxadiazole derivatives as biologically potential compounds. Med Chem Res 20:1550–1555 Podunavac-Kuzmanovic´ SO, Leovac VM, C´etkovic´ GS, Markov SL (2002) Synthesis, physico-chemical characterization and biological activity of copper(ii) and nickel(ii) complexes with l-benzoyl-2-methylbenzimidazole derivatives. Acta Periodica Technologica 33:151–157 Shah K, Chhabra S, Shrivastava SK, Mishra P (2013) Benzimidazole: a promising pharmacophore. Med Chem Res 22:5077–5104 Sharma D, Narasimhan B, Kumar P, Jalbout A (2009) Synthesis and QSAR evaluation of 2-(substituted phenyl)-1H- benzimidazoles and [2-(substituted phenyl)-benzimidazol-1-yl]-pyridin-3-ylmethanones. Eur J Med Chem 44:1119–1127 Thom SM, Horobin RW, Seidler E, Barer MR (1993) Factors affecting the selection and use of tetrazolium salts as cytochemical indicators of microbial viability and activity. J Appl Bacteriol 74:433–443 Yajko DM, Madej JJ, Lancaster MV, Sanders CA, Cawthon VL, Gee B, Babst A, Hadley WK (1995) Colorimetric method for determining MICs of antimicrobial agents for Mycobacterium tuberculosis. J Clin Microbiol 33:2324–2327
ORIGINAL PAPER
Synthesis and quantitative structure activity relationship (QSAR) of arylidene (benzimidazol-1-yl)acetohydrazones as potential antibacterial agents Yeldez El-Kilany • Nariman M. Nahas • Mariam A. Al-Ghamdi • Mohamed E. I. Badawy El Sayed H. El Ashry
•
Received: 16 June 2014 / Accepted: 3 October 2014 Ó Springer Science+Business Media Dordrecht 2014
Abstract Ethyl (benzimidazol-1-yl)acetate was subjected to hydrazinolysis with hydrazine hydrate to give (benzimidazol-1-yl)acetohydrazide. The latter was reacted with various aromatic aldehydes to give the respective arylidene (1H-benzimidazol-1-yl)acetohydrazones. Solutions of the prepared hydrazones were found to contain two geometric isomers. Similarly (2-methyl-benzimidazol-1-yl)acetohydrazide was reacted with various aldehydes to give the corresponding hydrazones. The antibacterial activity was evaluated in vitro by minimum inhibitory concentration (MIC) against Agrobacterium tumefaciens (A. tumefaciens), Erwinia carotovora (E. carotovora), Corynebacterium fascians (C. fascians) and Pseudomonas solanacearum (P. solanacearum). MIC result demonstrated that salicylaldehyde(1H-benzimidazol-1-yl)acetohydrazone (4) was the most active compound (MIC = 20, 35, 25 and 30 mg/L against A. tumefaciens, C. fascians, E. carotovora and P. solanacearum, respectively). Quantitative structure activity relationship (QSAR) investigation using Hansch analysis was applied to find out the correlation between antibacterial Electronic supplementary material The online version of this article (doi:10.1007/s11274-014-1755-4) contains supplementary material, which is available to authorized users. Y. El-Kilany N. M. Nahas M. A. Al-Ghamdi Chemistry Department, Faculty of Applied Science, Umm Al-Qura University, Mekkah, Kingdom of Saudi Arabia M. E. I. Badawy (&) Department of Pesticide Chemistry and Technology, Faculty of Agriculture, Alexandria University, El-Shatby, Alexandria 21545, Egypt e-mail: [email protected] E. S. H. El Ashry Chemistry Department, Faculty of Science, Alexandria University, Alexandria, Egypt
activity and physicochemical properties. Various physicochemical descriptors and experimentally determined MIC values for different microorganisms were used as independent and dependent variables, respectively. pMICs of the compounds exhibited good correlation (r = 0.983, 0.914, 0.960 and 0.958 for A. tumefaciens, C. fascians, E. carotovora and P. solanacearum, respectively) with the prediction made by the model. QSAR study revealed that the hydrophobic parameter (ClogP), the aqueous solubility (LogS), calculated molar refractivity, topological polar surface area and hydrogen bond acceptor were found to have overall significant correlation with antibacterial activity. The statistical results of training set, correlation coefficient (r and r2), the ratio between regression and residual variances (f, Fisher’s statistic), the standard error of estimates and significant (s) gave reliability to the prediction of molecules with activity using QSAR models. However, QSAR equations derived for the MIC values against the tested bacteria showed negative contribution of molecular mass. Keywords Hydrazones Geometric isomers Benzimidazole Benzimidazolyl-acetohydrazones Antibacterial activity QSAR
Introduction The resistance to antimicrobial agents is widespread, therefore, the development of new antimicrobial agents and understanding their mechanisms of action are becoming vital nowadays (Foroumadi et al. 2003). The discovery of the presence of 5,5-dimethyl-1-(a-D-ribofuranosyl)benzimidazole as an integral part of the structure of vitamin B12, led to the consideration of the benzimidazole a promising pharmacophore (Shah et al. 2013), that play an important role in
123
World J Microbiol Biotechnol
medicinal chemistry. The potential biological activity of benzimidazole derivatives as lead compounds has attracted much attention and led to the synthesis of variety of compounds having the benzimidazole ring. The incorporation of the benzimidazole nucleus is an important synthetic strategy in studies of antimicrobial drug discovery (Sharma et al. 2009; Khalafi-Nezhad et al. 2005). Benzimidazole and its derivatives have received much attention because of their biological activity and commercial application and they are considered as bacteriostatic and bactericidal (Chaudhary et al. 2011; Ingle and Magar 2011). Benzimidazole is present in many naturally occurring products and various drugs. Some of these compounds have antibacterial, antifungal, antiviral, antiinflammatory, antihypertensive, arteriosclerosis and anti-HIV activities (Chohan and Hanif 2013; Goker et al. 1999; Kalinowska-Lis et al. 2014; Podunavac-Kuzmanovic´ et al. 2002). During the last two decades, quantitative structure–activity relationship (QSAR) models have gained an extensive recognition in physical, organic, analytical, pharmaceutical and medicinal chemistry. QSAR models are highly effective in describing the structural basis of biological activity. The success of QSAR approach can be explained by the insight offered into the structural determination of chemical properties and the possibility to estimate the properties of new chemical compounds without the need to synthesize and test them (Hansch and Fujita 1964; Pal et al. 2011; Khatkar et al. 2013). Our thrust for the search of some newer and potent antimicrobial agents against bacteria, which is driven by the fact that the compound resistance is the major problem in treatment of diseases, prompted us to carry out QSAR studies of the molecules mentioned in above series (Gupta et al. 2002; Hansch and Fujita 1964; Kumar et al. 2007; Narasimhan et al. 2007). In the present study, we performed 2-D QSAR on a series of arylidene (benzimidazol-1-yl)acetohydrazones compounds to establish the relationship with the antibacterial activity against A. tumefaciens, C. fascians, E. carotovora and P. solanacearum. The resultant QSAR models were validated using multiple linear regression (MLR) using SPSS software. It was observed that molecular mass, polarizability, and lipophilicity parameters were major contributors to the antibacterial activity.
Materials and methods
N
N
R"
N
Aldehyde
H2C
H2C
O
H N
N
R"
N
H N
H
H
O Ar
N
H
OH
1 R" = H 2 R" = Me N
N
R"
N H2C Ar
H2C
O
N
N
H H
H
R R' R" = H R' H H OH OMe
N
Ar
E
R 3H 4 OH 5H 6H
R"
N
R
N
O H
Z
O
R' R" = Me R R' H 7H 8 OH H OH 9H
10 R" = H 11 R" = Me
Scheme 1 Synthesis of arylidene benzimidazolyl-acetohydrazones
monitored by TLC using Merck silica gel 60. Melting points were determined in open capillaries and are uncorrected. IR spectra were recorded on Fourier transform infrared 8,400 spectrophotometer with KBr pellets. 1H and 13 C (Proton decoupled) NMR spectra were recorded on Jeol JNM ECA 500 MHz using TMS as an internal standard (chemical shifts in d, ppm). The number of protons will be given as (c) when it is a fraction of a proton. The fraction of a proton is shown by X1. Satisfactory C, H, and N analyses were obtained for the compounds at the micro analytical laboratory of Cairo-University, Egypt. Four bacteria strains, Agrobacterium tumefaciens (Family: Rhizobiaceae), Erwinia carotovora (Family: Enterobacteriaceae), Corynebacterium fascians ((Family: Nocardiaceae) and Pseudomonas solanacearum (Family: Pseudomonadaceae) were provided by Microbiology Laboratory, Department of Plant Pathology, Faculty of Agriculture, Alexandria University, Egypt and were maintained on nutrient agar (NA) medium at 37 °C.
Materials and test microorganisms Antibacterial evaluation All reagents and solvents used in the study were of analytical grade and procured locally. Purification of compounds was made by column chromatography using Merck silica gel 60 (230–400 mesh) and a mixture of hexane and ethyl acetate was used for elution. All the reactions were
123
The in vitro antibacterial activities as minimum inhibitory concentration (MIC) of the compounds were determined by microdilution broth assay method using 2,3,5,-triphenyltetrazolium chloride (TTC, Sigma) as chromogenic marker
World J Microbiol Biotechnol Table 1 In vitro antibacterial activity of arylidene (benzimidazol-1yl)acetohydrazones derivatives against A. tumefaciens, C. fascians, E. carotovora and P. solanacearum
indicating MIC. A change in color from colorless to red formazan shows the growth of the organism.
Compound
Quantitative structure activity relationship analysis
1
MIC (mg/L) A. tumefaciens
C. fascians
E. carotovora
P. solanacearum
400
700
300
450
2
550
[1,000
[1,000
[1,000
3
75
400
200
350
4
15
45
30
35
5
20
35
25
30
6
75
500
275
400
7
200
475
250
425
8
50
300
100
125
9
125
375
175
200
10
45
350
75
150
11
150
425
225
250
MIC is the minimum inhibitory concentration obtained for each microorganism
(Thom et al. 1993; Yajko et al. 1995). Nutrient broth (NB) medium was used to grow the bacterial strains to a final inoculum size of 5 9 105 CFU/mL. The arylidene (benzimidazol-1-yl)acetohydrazones were dissolved in dimethyl sulfoxide (DMSO) and diluted with distilled water to obtain a final stock solution of 1,000 mg/L. For the broth microdilution test, 20 lL of each bacterial suspension in NB medium was added to the wells of a sterile 96-well microtitre plate already containing 40 lL of serially diluted arylidene (benzimidazol-1-yl)acetohydrazone compounds and 140 lL NB medium. The final volume in each well was 200 lL. Control wells were prepared with culture medium, bacterial suspension only, arylidene (benzimidazol-1-yl)acetohydrazone compounds only and DMSO in amounts corresponding to the highest quantity present. The contents of each well were mixed on a microplate shaker at 200 rpm for 1 min prior to incubation for 24 h in the cultivation conditions described above. The MIC was the lowest concentration where no viability was observed after 24 h on the basis of metabolic activity. To indicate respiratory activity the presence of color was determined after adding 10 lL/well of TTC dissolved in water (0.01 %, w/v) and incubated under appropriate cultivation conditions for 30 min in the dark. The absorbance was measured at 492 nm in an ultra microplate reader (Robonik, Pvt. Ltd). Positive controls were wells with a bacterial suspension in NB and a bacterial suspension in NB medium with DMSO in amounts corresponding to the highest quantity present in the broth microdilution assay. Negative controls were wells with growth medium and arylidene (benzimidazol-1-yl)acetohydrazone compounds. All measurements of MIC values were repeated in triplicate. The color change was assessed visually, with the highest dilution remaining colorless (inhibition of growth)
To develop QSAR, one needs to select a few descriptors from large physiochemical properties. The selected descriptors were used as independent variables to create the regression equations with activities as the dependent variable. Molecular weight (MW), CMR and topological polar surface area (tPSA) were computed by ChemDraw Ultra 11.0 software package. Calculated hydrophobic parameter (ClogP, logarithm of partition coefficient) was calculated by EPI SuiteTM v4.11 on the basis of the work of Hansch and Fujita (1964). The aqueous solubility of a compound (LogS) and hydrogen bond acceptors (HA) were calculated by using ChemSpider (http://www. chemspider.com/). The multiple linear regression model tested in this work was used to study the correlation between dependant variable (the biological activity expressed as MIC) and independent variables (MW, ClogP, logS, CMR, tPSA and HA). The regression analysis was performed using SPSS v.17 software (SPSS Inc., Chicago, IL, USA). Equations were justified by the correlation coefficient (r and r2), the standard error of estimates (s) and the value of the ratio between regression and residual variances (f, Fisher’s statistic). Only those parameters having good correlation coefficient and low standard error were considered to determine best equation. Regression was built using descriptor subsets containing only one of these highly correlated descriptors. Multiple Linear Regression (MLR) analysis was carried out to find out the factors responsible for variation in the biological activity. Successive regression equations were derived in which parameters are added, removed, or replaced until r2 and s values are optimized. To derive QSAR models, stepwise MLR analysis with cross validation technique was applied to both sets of ten compounds. Models with number of descriptors\7, f ratio higher than 2 and correlation coefficient, r more than 0.90 between predicted and the experimental antibacterial activities were validated. Significant descriptors were chosen based on the statistical data of analysis. Intercorrelation between the descriptors was checked for independence of the variables. The QSAR models with high statistical significance are reported herein.
Results and discussion Synthesis and spectroscopic characterizations of arylidene (benzimidazol-1-yl)acetohydrazones derivatives Reaction of (1H-benzimidazol-1-yl)acetohydrazine 3 with benzaldehyde, salicylaldehyde, p-hydroxybenzaldehyde,
123
World J Microbiol Biotechnol Table 2 Values of physiochemical molecular descriptors for arylidene (benzimidazol-1-yl)acetohydrazones derivatives used for QSAR analysis against A. tumefaciens Compound
PMIC
Residual
Values of descriptors
Actual
Predicted
MW
Clog P
LogS
CMR
TPSA
HA
1
2.68
2.82
-0.1403
2
2.57
2.41
0.1581
190.2 204.23
-0.12 0.152
1.91
5.3
70.72
5
2.14
5.76
70.72
3 4
3.57 4.29
3.61 4.28
-0.0365 0.0056
278.31 294.31
5
2.36 2.96
3.57 3.31
8.4 8.55
57.06 77.29
5 6
5
4.17
4.03
0.1354
294.31
2.33
3.34
8.55
77.29
6
6
3.61
3.54
7
3.16
3.14
0.0747
308.33
2.58
3.68
9.01
66.29
6
0.0187
292.34
2.63
3.74
8.86
57.06
8
3.79
5
3.80
-0.0065
308.33
3.23
3.45
9.01
77.29
9
6
3.39
3.55
-0.1567
308.33
2.6
3.48
9.01
77.29
6
10
3.78
3.76
0.0216
268.27
1.53
3.48
7.61
66.29
6
11
3.27
3.34
-0.0742
282.3
1.8
3.55
8.08
66.29
6
MIC minimum inhibitory concentration, pMIC -Log MIC, MW Molecular weight, ClogP Calculated hydrophobic parameter (logarithm of partition coefficient), LogS The aqueous solubility of a compound, CMR Calculated molar refractivity, tPSA Topological polar surface area, HA Hydrogen bond acceptors
Fig. 1 A Plot of predicted pMIC of arylidene (benzimidazol-1yl)acetohydrazones derivatives (n = 11) activity for A. tumefaciens against the experimental pMIC values for the linear regression developed model by Eq. 1. B Plot of residual pMIC against the experimental pMIC values
anisaldehyde or furfuraldehyde gave the respective hydrazones 3–6 (Scheme 1). The solution of 3 in DMSO-d6 showed the presence of two species as indicated from the presence of two signals for CH2 at d 5.07 and 5.54 and two signals for
123
H–C=N at d 8.20 and 8.21. Moreover 13C NMR showed two signals for C=N at dc 144.5 and 144.7. These data indicated that the hydrazone product was a mixture of the geometric isomers E and Z in a ratio 4:1, respectively. The selection of the higher ratio for E has been based their possible relative stability as a consequence of the crowded environment around the C=N in Z compared to E. Consequently, the expected intermediate gave preferably the isomer E over Z. The salicylaldehyde (1H-benzimidazol-1-yl)acetohydrazone 4 showed two signals for H–C=N at d 8.34 and 8.45 and two signals for C=N at dc 147.9 and 148.1, but the ratio of E to Z was 2:1, respectively. Similarly, the ratios of E and Z for 4-hydroxybenzaldehyde (1H-benzimidazol-1-yl)acetohydrazone 5 and anisaldehyde (1H-benzimidazol-1-yl)acetohydrazone 6 were found to be 4:1 and 3:1, respectively. Reaction of benzaldehyde, salicylaldehyde and 4-hydroxybenzaldehyde with (2-methyl-1H-benzimidazol1-yl)acetohydrazine gave the respective hydrazones 7–9 (Scheme 1) whose solutions in DMSO-d6 indicated their presence as mixture of the geometric isomers E and Z in a ratio 4:1, 2:1 and 3:1, respectively. The selection of the higher ratio for E has been based on their possible relative stability as a consequence of the crowded environment around the C=N in Z compared to E. The furfural (1Hbenzimidazol-1-yl)acetohydrazone 10 and (2-methyl-1Hbenzimidazol-1-yl)acetohydrazone 11 were having also a mixture of the two geometric isomers in the ratio 3:1. The in vitro antibacterial activity and QSAR study The antibacterial properties of arylidene (benzimidazol-1yl)acetohydrazone derivatives were performed in vitro by a
World J Microbiol Biotechnol
Fig. 2 A Plot of predicted pMIC of arylidene (benzimidazol-1yl)acetohydrazones derivatives (n = 10) activity for C. fascians against the experimental pMIC values for the linear regression developed model by Eq. 2. B Plot of residual pMIC against the experimental pMIC values
conventional microdilution broth assay method against A. tumefaciens, C. fascians, E. carotovora and P. solanacearum and the MIC values are given in Table 1. Salicylaldehyde(1H-benzimidazol-1-yl)acetohydrazone (4) having ortho –OH group on the phenyl ring was the most active compound among the tested compounds with MIC values of 15, 45, 30 and 35 mg/L against A. tumefaciens, C. fascians, E. carotovora and P. solanacearum, respectively. Followed in the descending order by 4-hydroxybenzaldehyde(1H-benzimidazol-1-yl)acetohydrazone (5) (MIC = 20, 35, 25 and 30 mg/L against A. tumefaciens, C. fascians, E. carotovora and P. solanacearum, respectively). This result confirms that the substitution with a hydroxyl group on the ortho or para positions on the phenyl ring enhanced the biological activity compared to other compounds. However, a substitution of the hydrogen by a methyl group on the benzimidazol led to decrease the activity (see compounds 1, 3, 4 and 5 vs. 2, 7, 8 and 9, respectively). 2-(2-Methyl-benzimidazol-1-yl)acetohydrazide (2) was the least active (MIC = 550 mg/L for A. tumefaciens, and higher than 1,000 mg/L for C. fascians, E. carotovora and P. solanacearum). When we consider the susceptibility
Fig. 3 A Plot of predicted pMIC of arylidene (benzimidazol-1yl)acetohydrazones derivatives (n = 10) activity for E. carotovora against the experimental pMIC values for the linear regression developed model by Eq. 3. B Plot of residual pMIC against the experimental pMIC values
of the microorganisms, another point deserves awareness; A. tumefaciens and E. carotovora were more sensitive to the tested compounds than C. fascians and P. solanacearum. For QSAR analysis, the respective MIC values in mg/L of these compounds were converted to mole/L and then to -log MIC (pMIC), where MIC value is the concentration of the compound required for 100 % inhibition of the growth of the microorganisms. Based on these values, statistically significant equations were generated. Different sets of equations were produced for A. tumefaciens, C. fascians, E. carotovora and P. solanacearum as follows. The various physiochemical molecular descriptors computed for investigations were molecular weight (MW), calculated hydrophobic parameter (logarithm of partition coefficient, ClogP), the aqueous solubility of a compound (LogS), calculated molar refractivity (CMR), topological polar surface area (tPSA) and hydrogen bond acceptor (HA). For A. tumefaciens, QSAR was run for 11 compounds and statistical parameters like r, r2, f, s and residual sum of squares were calculated (Table 2). They were found statistically significant and the regression analysis was run to yield the following equation:
123
World J Microbiol Biotechnol
s = 0.007) and high coefficient constant was obtained with CMR (14.421) compared to r = 0.737 in the case of MW. Further the plot of linear regression experimental pMIC values against the predicted pMIC values for A. tumefaciens also favors the model expressed by Eq. 1 as shown in Fig. 1A. To investigate the existence of a systemic error in developing the QSAR model, we have plotted pMIC observed against pMIC residual values (Fig. 1B). The propagation of the residuals on both sides of zero indicates that there is no systemic error in the development of linear regression model (Heravi and Kyani 2005). For C. fascians, variations in the biological activity of arylidene (benzimidazol-1-yl)acetohydrazone derivatives were analyzed using the physiochemical descriptors (Table S2) and the obtained equation with correlation coefficient was listed in Eq. 2.
Fig. 4 A Plot of predicted pMIC of arylidene (benzimidazol-1yl)acetohydrazones derivatives (n = 10) activity for P. solanacearum against the experimental pMIC values for the linear regression developed model by Eq. 4. B Plot of residual pMIC against the experimental pMIC values
pMIC ¼ 526ð0:113Þ MW þ 0:441ð 0:248Þ ClogP þ 0:932ð 0:702Þ LogS þ 14:421ð3:325Þ CMR þ 0:017ð0:028Þ tPSA þ 6:557ð1:430Þ HA 9:339ð2:898Þ ð1Þ where, number of compounds, n = 11, correlation coefficient, r = 0.983 and r2 = 0.966, Fisher test, f = 18.934, Residual sum of squares, RSS = 0.101; standard error of the estimate, SE = 0.158 and significant, s = 0.007. Equation 1 was found to be better model for antibacterial activity of this series of compounds against A. tumefaciens. This revealed that the molecular descriptors have significant effect on biological activity. All these parameters correlate positively with biological activity except MW indicating that the increase in molecular mass led to decrease the antibacterial activity. Table S1 shows the development of the QSAR Eq. 1 and all the calculated parameters were subjected to correlation studies with depending on their individual correlation with the biological activity. In this group, a combination of all descriptors significantly enhanced the biological activity where a high correlation was obtained (r = 0.983 and
123
pMIC ¼ 0:638ð0:280Þ MW þ 0:045ð0:520Þ ClogP þ 1:202ð1:482Þ LogS þ 17:893ð8:203Þ CMR þ 0:051ð0:062Þ tPSA þ 7:392ð3:617Þ HA 13:854ð6:948Þ ð2Þ where n = 10, r = 0.914, r2 = 0.835, f = 2.532, RSS = 0.331, SE = 0.332 and s = 0.238. The correlation coefficient of this equation is satisfied (r = 0.914) using the entire data set (n = 10). The positive coefficient of the Clog P, Log S, CMR, tPSA and HA indicate that an increase of these descriptors led to increase in the antibacterial activity. However, the negative coefficient of the MW indicates the high MW, the lower the activity against the bacterium. Table S3 shows the development of the QSAR Eq. 2 and all the calculated parameters were subjected to correlation studies with depending on their individual correlation with the biological activity. In this group, HA significantly enhanced the biological activity where a high correlation was obtained (r = 0.914) compared to r = 0.778 in the case of all descriptors except HA. The observed pMIC versus calculated pMIC values according to Eq. 2 were plotted in Fig. 2A and plot of pMIC observed against pMIC residual values is shown in Fig. 2B. The values of physiochemical molecular descriptors for arylidene (benzimidazol-1-yl)acetohydrazone compounds used for QSAR analysis against E. carotovora and actual and predicted pMIC with residuals are indicated in Table S4. Statistical parameters including r, r2, f, s and residual sum of squares were calculated and they were found statistically significant and the regression analysis was run to yield following QSAR equation:
World J Microbiol Biotechnol
pMIC ¼ 0:494ð0:171Þ MW þ 0:212ð0:316ÞClogP þ 1:925ð0:902Þ LogS þ 13:328ð4:993Þ CMR þ 0:070 ð0:038Þ tPSA þ 5:519ð2:201Þ HA 10:039ð4:229Þ ð3Þ 2
where n = 10, r = 0.960, r = 0.922, f = 5.919, RSS = 0.123, SE = 0.202 and s = 0.086. A combination of all molecular descriptors led to a significant increase in the biological correlation (r = 0.960) and high coefficient constants were obtained with CMR (13.328) and HA (5.519). The positive coefficient of Clog P, Log S, CMR, tPSA and HA indicate that an increase of these descriptors led to increase in the antibacterial activity. However, the negative coefficient (-0.494) of the MW indicates the high molecular mass, the lower the activity against the bacterium. Table S5 shows the development of the QSAR (Eq. 3) and all the calculated parameters were subjected to correlation studies with depending on their individual correlation with the biological activity. In this group, a combination of tPSA and HA with other descriptors significantly enhanced the biological activity where the correlation was increased from 0.695 to 0.960 with low standard deviation 0.086. The observed pMIC versus calculated pMIC values according to Eq. 2 were plotted in Fig. 3A and plot of pMIC observed against pMIC residual values is shown in Fig. 3B. QSAR analysis for P. solanacearum was run for ten compounds and the pMIC observed and calculated with the residuals and molecular descriptors for arylidene (benzimidazol-1-yl)acetohydrazone derivatives are shown in Table S6. Statistical parameters including r, r2, f and residual sum of squares were calculated and the regression analysis was run to yield following equation: pMIC ¼ 0:459ð0:190Þ MW þ 0:179ð0:353ÞClogP þ 1:753ð1:006Þ LogS þ 12:433ð5:566Þ CMR þ 0:074ð0:042Þ tPSA þ 5:159ð2:424Þ HA 10:180 ð4:715Þ ð4Þ where n = 10, r = 0.958, r2 = 0.917, f = 5.548, RSS = 0.152, SE = 0.225 and s = 0.094. Equation 4 was found to be a better model for antibacterial activity of this series of compounds against P. solanacearum (r = 0.958). This equation revealed that all descriptors contribute significantly to the biological activity. ClogP, LogS, CMR, tPSA and HA correlate positively with biological activity which emphasizes that increase in these parameters contribute to increase in biological activity. However, MW correlates negatively (-0.459) with biological activity which emphasizes that increase in
mass of molecule contributes to decrease in biological activity. Table S7 shows the development of the QSAR Eq. 4. A combination of CMR, tPSA and HA with other descriptors (MW, ClogP and LogS) significantly enhanced the biological activity where the correlation was increased from 0.605 to 0.958 with the lowest standard deviation 0.094. Correlation graph between actual and predicted MIC values was also generated (Fig. 4A) along with the plot of pMIC observed against pMIC residual values (Fig. 4B). In the present study, Clog P, LogS, CMR, tPSA and HA are very useful parameter for prediction of the biological activity and drug properties. CMR and HA were significantly the highest correlation found with all tested bacteria. For the arylidene (benzimidazol-1-yl)acetohydrazone derivatives, the QSAR regressions suggest that the same structural features describe the activities for bacteria. These results agree with those found for the QSAR of 5-nitroimidazole analogues and other related compounds (Gupta et al. 2002; Kumar et al. 2007; Narasimhan et al. 2007). A positive contribution of MR with activity against Escherichia coli was also reported (Gupta et al. 2002). In this study, it is seen that compounds with higher CMR showed higher antibacterial activity. A series of 2-(substituted phenyl)-1Hbenzimidazole and [2-(substituted phenyl)-benzimidazol-1yl]-pyridin-3-yl-methanone derivatives and tested in vitro for their antimicrobial activity was synthesized (Sharma et al. 2009). The results of QSAR investigation indicated the importance of molecular descriptors, dipole moment (m), log of octanol water partition coefficient (log P) and second order molecular connectivity index in describing the antimicrobial activity of the synthesized compounds. QSAR studies for some disubstituted-1,3,4-oxadiazole derivatives were tested and the in vitro antibacterial activity (MICs) of the compounds against Staphylococcus aureus and E. coli 2 exhibited good correlation (n = 36, r2 [ 0.7, rcv [ 0.5, s = 0.069) with the prediction made by the model of Pal et al. (2011). It was found that the polarizability, thermodynamic, and lipophilicity were major responsible factors for exhibiting the activity. Recent literature reveals that the QSAR has been applied to describe the relationship between narrow range of biological activity and physiochemical properties of the molecules. When biological activity data lie in a narrow range, the presence of minimum standard deviation of the biological activity justifies its use in QSAR studies (AbdelRahman et al. 2013; Kumar et al. 2007; Narasimhan et al. 2007; Nofal et al. 2013; Sharma et al. 2009). The minimum standard deviation (Eqs. 1, 2, 3 and 4) observed in the antimicrobial activity data justifies its use in QSAR studies. In conclusion, the models obtained through this QSAR study gave a better prediction capability of antibacterial activity against bacteria like A. tumefaciens, C. fascians,
123
World J Microbiol Biotechnol
E. carotovora and P. solanacearum. Clog P, Log S, CMR, tPSA and HA were the major factors responsible for positively affecting the antibacterial activity of these arylidene (benzimidazol-1-yl)acetohydrazones derivatives. The QSAR models are of immense importance for optimizing the biological activity of benzimidazol analogues, which could help in designing better molecules with enhanced anti-bacterial activity. The results suggest that all the newly synthesized compounds showed good to very good activity. Hence the fact that the compounds prepared in this study are chemically related to the current medication, suggests that further work with similar analogues is clearly warranted.
References Abdel-Rahman AAH, Nassar IF, El-Kattan IMH, Aly AA, Behalo MS, Abdelwahed NM (2013) Synthesis and antimicrobial evaluation of thioglycosides and acyclic C-nucleosides of 2-methylbenzimidazole. Der Pharma Chemica 5:210–217 Chaudhary M, Pareek D, Pareek PK, Kant R, Ojha KG, Pareek A (2011) Synthesis of some new biologically active benzothiazole derivatives containing benzimidazole and imidazoline moieties. Bull Korean Chem Soc 32:131–136 Chohan ZH, Hanif M (2013) Antibacterial and antifungal metal based triazole Schiff bases. J Enz Inhib Med Chem 28:944–953 Foroumadi A, Mansouri S, Kiani Z, Rahmani A (2003) Synthesis and in vitro antibacterial evaluation of N-[5-(5-nitro-2-thienyl)-1,3,4thiadiazole-2-yl] piperazinyl quinolones. Eur J Med Chem 38:851–854 Goker H, Ayhankilcilgil G, Tuncbilek M, Kus C, Ertan R, Kendi E, Oybey S, Fort M, Garcia C, Farre A (1999) Synthesis and antihistaminic Hl activity of 1,2,5(6)- trisubstituted benzimidazoles. J Heterocycl 51:2561–2573 Gupta MK, Mishra P, Prathipati P, Saxena AK (2002) 2D-QSAR in hydroxamic acid derivatives as peptide deformylase inhibitors and antibacterial agents. Bioorg Med Chem 10:3713–3716 Hansch C, Fujita T (1964) p–r–p analysis. A method for the correlation of biological activity and chemical structure. J Am Chem Soc 86:1616–1626 Heravi JM, Kyani A (2005) A use of computer-assisted methods for the modeling of the retention time of a variety of volatile organic compounds: a PCA-MLR-ANN approach. J Chem Inf Comput Sci 44:1328–1335
123
Ingle RG, Magar DD (2011) Heterocyclic chemistry of benzimidazoles and potential activities of derivatives. Int J Drug Res Tech 1:26–32 Kalinowska-Lis U, Felczak A, Che˛cin´ska L, Lisowska K, Ochocki J (2014) Synthesis, characterization and antimicrobial activity of silver(I) complexes of hydroxymethyl derivatives of pyridine and benzimidazole. J Organomet Chem 749:394–399 Khalafi-Nezhad A, Rad MNS, Mohbatkar H, Asrari Z, Hemmateenejad B (2005) Design, synthesis, antibacterial and QSAR studies of benzimidazole and imidazole chloroaryloxyalkyl derivatives. Bioorg Med Chem 13:1931–1938 Khatkar A, Nanda A, Kumar P and Narasimhan B (2013) Synthesis, antimicrobial evaluation and QSAR studies of gallic acid derivatives. Arab J Chem. doi:10.1016/j.arabjc.2013.11.014 Kumar A, Narasimhan B, Kumar D (2007) Synthesis, antimicrobial, and QSAR studies of substituted benzamides. Bioorg Med Chem 15:4113–4124 Narasimhan B, Judge V, Narang R, Ohlan R, Ohlan S (2007) Quantitative structure–activity relationship studies for prediction of antimicrobial activity of synthesized 2,4-hexadienoic acid derivatives. Bioorg Med Chem Lett 17:5836–5845 Nofal ZM, Srour AM, El-Eraky WI, Saleh DO, Girgis AS (2013) Rational design, synthesis and QSAR study of vasorelaxant active 3-pyridinecarbonitriles incorporating 1H-benzimidazol-2yl function. Eur J Med Chem 63:14–21 Pal T, Somani RR, Shirodkar PY, Bhanushali UV (2011) Quantitative structure activity relationship (QSAR) studies of a series of 2,5disubstituted-1,3,4-oxadiazole derivatives as biologically potential compounds. Med Chem Res 20:1550–1555 Podunavac-Kuzmanovic´ SO, Leovac VM, C´etkovic´ GS, Markov SL (2002) Synthesis, physico-chemical characterization and biological activity of copper(ii) and nickel(ii) complexes with l-benzoyl-2-methylbenzimidazole derivatives. Acta Periodica Technologica 33:151–157 Shah K, Chhabra S, Shrivastava SK, Mishra P (2013) Benzimidazole: a promising pharmacophore. Med Chem Res 22:5077–5104 Sharma D, Narasimhan B, Kumar P, Jalbout A (2009) Synthesis and QSAR evaluation of 2-(substituted phenyl)-1H- benzimidazoles and [2-(substituted phenyl)-benzimidazol-1-yl]-pyridin-3-ylmethanones. Eur J Med Chem 44:1119–1127 Thom SM, Horobin RW, Seidler E, Barer MR (1993) Factors affecting the selection and use of tetrazolium salts as cytochemical indicators of microbial viability and activity. J Appl Bacteriol 74:433–443 Yajko DM, Madej JJ, Lancaster MV, Sanders CA, Cawthon VL, Gee B, Babst A, Hadley WK (1995) Colorimetric method for determining MICs of antimicrobial agents for Mycobacterium tuberculosis. J Clin Microbiol 33:2324–2327
