- Email: [email protected]
Structure–antimicrobial activity relationship for silanols, a new class of disinfectants, compared with alcohols and phenols
International Journal of Antimicrobial Agents 29 (2007) 217–222
Short communication
Structure–antimicrobial activity relationship for silanols, a new class of disinfectants, compared with alcohols and phenols Yun-mi Kim a , Samuel Farrah b , Ronald H. Baney a,∗ a
Department of Materials Science and Engineering, University of Florida, Gainesville, FL 32611, USA b Microbiology and Cell Science, University of Florida, Gainesville, FL 32611, USA Received 4 April 2006; accepted 25 August 2006
Abstract Triorganosilanols (R(CH3 )2 SiOH) were recently reported to exhibit unexpectedly strong disinfectant properties. The antimicrobial activities of silanols were significantly higher than their analogous alcohols. A study of the structural dependence of their antimicrobial activity was conducted against four bacteria, Escherichia coli, Staphylococcus aureus, Pseudomonas aeruginosa and Enterococcus faecalis. Silanols, alcohols with structures analogous to the silanols (R(CH3 )2 COH) and substituted phenols were evaluated as a single class of materials. The minimum lethal concentrations (MLCs), defined as the concentration required for a 7-log reduction in viable bacteria after 1 h exposure, were used to measure antimicrobial activity. Octanol–water partition coefficients (log P) and hydrogen bond acidities (␦), measured as the shift in frequency of the OH stretching band between free OH and hydrogen-bonded OH to diethyl ether oxygen by infrared spectroscopy, were utilised as dispersive and polar structural parameters, respectively. The correlation established by multiple regression analysis between antimicrobial activities and structural properties of silanols, alcohols and phenols against the four bacteria treated as a single family produced the following equation, log(1/MLC) = 0.679 log P + 0.0036␦ − 1.909 (n = 282, r = 0.96, s = 0.22). This equation and the significantly high correlation coefficient supported the hypothesis that the lipophilic properties and the H-bond acidities are primary factors for the antimicrobial action of silanols, alcohols and phenols. The high antimicrobial activity of silanols is explained by their greater H-bond acidity and their enhanced lipophilicity. © 2006 Elsevier B.V. and the International Society of Chemotherapy. All rights reserved. Keywords: Silanols; Structure–antimicrobial activity relationship; Regression analysis; Hydrophobicity; H-bond acidity
1. Introduction Silanols, i.e. silicon alcohols (R3 SiOH), are a new class of antimicrobials [1]. Their antimicrobial activity appears to be stronger than their analogous alcohols. Silanols are environmentally friendly materials because they are rather quickly degraded into environmentally benign silica, carbon dioxide and water in the environment [2]. Silanols have a hydrophilic portion, the hydroxyl group, and a hydrophobic region, the organic substituents, similar to alcohols and phenols widely used as antimicrobials. It is reason∗ Corresponding author. Present address: 168 Rhines Hall, University of Florida, Gainesville, FL 32611, USA. Tel.: +1 352 846 3785; fax: +1 352 846 3355. E-mail address: [email protected] (R.H. Baney).
able to predict that the mode of antimicrobial action of silanols would resemble those of the alcohols and phenols [3] because of the similarity of the chemical structures of these materials. The quantitative relationship between chemical structure and biological activity has received considerable attention in recent years because it allows one to predict bioactivity without an inordinate amount of time and effort. Moreover, insights gained from such studies can give insights into the mechanism of bioactivity. A method for quantitative correlation of biological activity and chemical structure was introduced by Hansch and Fujita in 1964 [4]. Structural dependence studies with antimicrobial activity have focused mainly on the effects of the hydrophobicity by testing a homologous series of samples, aliphatic alcohols, alkylated phenols and quaternary ammonium compounds [5–7].
0924-8579/$ – see front matter © 2006 Elsevier B.V. and the International Society of Chemotherapy. All rights reserved. doi:10.1016/j.ijantimicag.2006.08.036
218
Y.-m. Kim et al. / International Journal of Antimicrobial Agents 29 (2007) 217–222
Lien et al. [8] summarised the structure–activity relationship of antimicrobials by means of equations based on the method proposed by Hansch and Fujita [4]. The multiple regression analysis method was used to establish the correlation equations [9]. Lien et al. [8] observed that the correlations varied with bacteria and type of antimicrobials. These authors suggested that the lipophilicity of the molecule was the most important factor for the antimicrobial activities of the compounds, with a relatively minor contribution from electronic properties [8]. Our studies examine the relationship between antimicrobial activity and physicochemical properties for alcohols, phenols and previously unreported silanols as a single class of agents. The parameters utilised in our study were the octanol–water partition coefficient and the hydrogen bond acidity, which represent, respectively, the dispersive lipophilic property and the polar property of the hydroxylcontaining antimicrobials.
2. Materials and methods 2.1. Antimicrobials and their purities Eight silanols (R(CH3 )2 SiOH) (Fig. 1a) were prepared by hydrolysis of chlorosilane obtained from Gelest Inc. (Morrisville, PA). The chlorosilanes and water were mixed for 15–30 min in a diethyl ether solution with ammonium hydroxide and the silanols were collected by evaporation of the ether. The analogous alcohols, R(CH3 )2 COH (Fig. 1b) and nine phenols (Fig. 1c) were obtained from Acros Organics (Fairlawn, NJ). The purities of the silanols measured by 29 Si and 1 H nuclear magnetic resonance spectroscopy were 95 ± 3%, with the impurity consisting of disiloxanes. The antimicrobial effects of disiloxanes were also tested. The alcohols and phenols were used as received and their purities were >97% as determined by gas chromatography. 2.2. Preparation of bacteria and procedures for antimicrobial tests The bacterial strains employed were Escherichia coli C3000 (ATCC 15597), a laboratory strain of Staphylococcus
aureus (Department of Microbiology, University of Florida), Pseudomonas aeruginosa type strain (ATCC 10145) and Enterococcus faecalis type strain (ATCC 19433). Bacterial suspensions were prepared as follows. Bacteria were inoculated in a nutrient Columbia broth overnight at 37 ◦ C with constant agitation under aerobic conditions. The bacterial cells were collected by centrifugation at 500 relative centrifugal force for 10 min at 4 ◦ C and washed three times with sterilised distilled water. The bacterial pellet was resuspended in sterilised water after the final washing. The concentrations of the prepared bacterial suspensions were 2–6 × 108 colony-forming units (CFU)/mL. The antimicrobial activity tests of the materials were carried out by adding a given concentration of antimicrobial agent to 9 g of deionised water and 1 g of bacterial suspension containing a concentration of 2–6 × 108 CFU/mL. The solution was stirred for 1 h. The minimum lethal concentration (MLC) is defined as the lowest concentration of an agent required for a 7-log reduction in viable bacteria for the test organisms after a 1 h exposure. The MLCs were utilised as a measure of the biocidal activity of the antimicrobials. As a control, a sample was diluted in phosphate-buffered saline based on a serial 1:10 dilution. Then, 0.1 mL from each of the last two dilution tubes was plated on plate count agar (Difco, Sparks, MD). Serial dilution was not required to obtain the MLC of the agents because of complete killing, in this case a 7-log reduction of test bacteria. After incubating the plates for 24 h at 37 ◦ C, the colonies growing on the medium were counted to estimate the number of viable bacteria. The standard deviation values in Table 1 were determined by taking a mean of three tests. 2.3. Determination of hydrogen bond acidity and the octanol–water partition coefficient Relative H-bond acidities of the silanols, alcohols and phenols were determined by measuring ␦, defined as the shift in frequency of the OH stretching band between free OH and hydrogen-bonded OH to diethyl ether oxygen [10]. The shift is a measure of the relative proton-donating nature of the OH-containing compounds, because the protonaccepting ability remains constant since the same base (diethyl ether) was employed. The shift, ␦, is proportional
Fig. 1. Chemical structures of (a) silanols, (b) alcohols and (c) phenols.
Y.-m. Kim et al. / International Journal of Antimicrobial Agents 29 (2007) 217–222
219
Table 1 Minimum lethal concentration (MLC), H-bond acidity in ether (␦) and octanol–water partition coefficient (log P) of silanols, alcohols and phenols against Escherichia coli, Staphylococcus aureus, Pseudomonas aeruginosa and Enterococcus faecalis log P
␦
3.15 ± 0.05 1.32 ± 0.03 1.14 ± 0.05 0.50 ± 0.05 0.42 ± 0.03 0.23 ± 0.03 – 0.12 ± 0.02
1.14 1.5 1.63 2.12 2.36 2.85 2.62 3.34
239 260 237 237 267 257 236 252
13.33 ± 0.42 5.63 ± 0.18 5.67 ± 0.15 1.76 ± 0.04 0.93 ± 0.03 0.76 ± 0.06 0.75 ± 0.05 0.32 ± 0.02
0.73 1.08 1.22 1.71 1.95 2.2 2.44 2.93
126 130 125 125 137 126 129 129
1.51 2.06 2.16 2.55 3.04 3.28 3.53 4.02 4.52
278 270 318 269 270 252 270 269 271
MLC (%, g/g)a
Materials
R
Silanols (R(CH3 )2 SiOH)
Methyl Vinyl Ethyl n-Propyl Phenyl Benzyl n-Butyl Phenethyl
2.36 1.23 1.04 0.43 0.27 0.17 0.14 0.10
± ± ± ± ± ± ± ±
0.12 0.06 0.04 0.03 0.03 0.01 0.01 0.02
2.48 1.04 0.80 0.36 0.26 0.16 0.14 0.08
± ± ± ± ± ± ± ±
0.17 0.07 0.02 0.04 0.03 0.01 0.01 0.02
2.36 ± 0.01 1.0 ± 0.03 0.87 ± 0.04 0.40 ± 0.03 0.35 ± 0.05 – – –
Alcohols (R(CH3 )2 COH)
Methyl Vinyl Ethyl n-Propyl Phenyl n-Butyl Benzyl Phenethyl
13.54 5.23 5.09 1.68 0.96 0.67 0.7 0.26
± ± ± ± ± ± ± ±
1.27 0.06 0.04 0.04 0.06 0.03 0.05 0.01
10.61 4.37 4.17 1.69 0.78 0.65 0.59 0.26
± ± ± ± ± ± ± ±
0.19 0.05 0.12 0.04 0.03 0.02 0.02 0.01
9.79 ± 0.19 3.67 ± 0.08 3.96 ± 0.04 1.03 ± 0.03 0.71 ± 0.03 0.53 ± 0.02 0.63 ± 0.03 1.33 ± 0.28
Phenols (RC6 H5 OH)
Hydridob 4-Methyl 3-Chloro 4-Ethyl 4-Propyl 2-Phenyl 4-Butyl 4-Pentyl 4-Hexyl
0.70 0.410 0.13 0.13 0.053 0.12 0.013 0.010 0.055
± ± ± ± ± ± ± ± ±
0.01 0.01 0.02 0.02 0.006 0.03 0.002 0.002 0.005
0.61 0.35 0.11 0.14 0.045 0.085 0.015 0.008 0.004
± ± ± ± ± ± ± ± ±
0.02 0.02 0.01 0.02 0.005 0.015 0.005 0.002 0.001
0.62 ± 0.08 0.35 ± 0.05 0.11 ± 0.01 0.11 ± 0.01 0.052 ± 0.008 0.13 ± 0.03 0.060 ± 0.01 – –
E. coli
S. aureus
P. aeruginosa
E. faecalis
0.98 ± 0.03 0.45 ± 0.05 0.14 ± 0.01 0.17 ± 0.03 0.055 ± 0.005 0.13 ± 0.03 0.017 ± 0.003 0.012 ± 0.002 0.006 ± 0.001
–, indicates no MLC obtained. a Each MLC is the average value of three data points. b Hydrido, unsubstituented phenol.
to the strength of the H-bond. A transmission sampling technique was utilised for the infrared spectroscopy measurements. The octanol–water partition coefficient, log P, used as a measure of the lipophilicity of the agents, was calculated using the demo program ‘LogKow’ provided by Syracuse Research Corporation. The program was used to estimate log P using the atom/fragment contribution (AFC) method developed through multiple linear regressions of experimental log P values [11].
3. Results 3.1. Characterisation of the partition coefficient and the H-bond acidity for the antimicrobials The primary impurities in the silanols were identified as the silanol condensation products, disiloxanes. Antimicrobial activities of the disiloxanes were evaluated to determine their contribution to the antimicrobial activities of the silanols. The experiments performed using 10% of the disiloxanes tested against the four bacteria showed <1 log reduction, demonstrating that the silanols were responsible for the observed antimicrobial activities.
The H-bond acidity and the partition coefficient of the materials are shown in Table 1. The H-bond acidities of the silanols are almost two times higher than their analogous alcohols owing to electron back donation through the bond from the p orbital of oxygen to the vacant d orbital of silicon [10]. We propose that the higher acidities of the silanols play an important role in their enhanced biocidal properties. The partition coefficients estimated by the AFC method [11] showed a gradual increase as the alkyl chain length increased (Table 1). The partition coefficients of the silanols were higher than the analogous alcohols, as shown in Table 1. It is well known that silicon compounds, including the silanols, exhibit higher hydrophobicity than analogous organic compounds owing to flexible molecular chains and a lower group rotation energy barrier around a Si C bond than the carbon–carbon bond [12]. 3.2. Antimicrobial activities of silanols, alcohols and phenols against four bacteria The MLCs of the silanols, alcohols and phenols were determined against E. coli, P. aeruginosa, S. aureus and E. faecalis and are given in Table 1. The MLC decreases as the alkyl chain length of the silanols increases. The experimental results with the alcohols and phenols were similar to those
220
Y.-m. Kim et al. / International Journal of Antimicrobial Agents 29 (2007) 217–222
of the silanols, although their MLC values differed from the silanols. A lower MLC implies a higher antimicrobial activity. A change of substituents from a short alkyl chain to a longer alkyl chain resulted in increased biocidal activity. The susceptibility of bacteria to the antimicrobials can be compared based on the relative MLCs of the antimicrobials. The resistances of the four bacteria against the agents varied slightly. Overall, E. faecalis was the least susceptible bacterium, followed by E. coli, P. aeruginosa and S. aureus. It is significant to note that the silanols showed at least two times the antimicrobial activities of the analogous alcohols against the four bacteria. For E. coli, trimethylsilanol and phenethyldimethylsilanol showed 2.36% and 0.1% of the MLC, respectively, whereas the corresponding alcohols, t-butanol and phenethyldimethylcarbinol, exhibited 13.54% and 0.26% of the MLC, respectively. 4-Hexyl phenol, which contains the most carbon atoms of the materials tested, displayed the lowest MLC for S. aureus, as shown in Table 1. 3.3. Correlation between antimicrobial activity and structural parameters We propose that the physicochemical parameters of the silanols, which show a higher hydrophobicity and H-bond acidity compared with the alcohols, contribute to their enhanced antimicrobial activity. We demonstrated a quantitative structure–activity relationship using silanols, alcohols and phenols treated as a single class of antimicrobials. The correlation equations between their antimicrobial activities and their structural properties were created using a multiple regression analysis and are summarised in Table 2. The MLC values converted to the logarithms of (1/MLC) were plotted as a function of the partition coefficient and produced a linear free energy relationship [4] (Table 2) with statistically significant values. The significance and validity of the regression models could be evaluated by assessing the correlation coefficient r, the standard deviation s and the Fvalues [9]. If the correlation coefficient r is >0.9, the models are significant. If the standard deviation s is not much larger than 0.3, the models are acceptable. The F-value is a measure
of statistical significance of the regression model. If F-values are larger than the 95% significance limits, the model can be considered statistically significant. Statistical analysis data shown in Table 2 confirmed that the equations are statistically significant. Hydrophobicity was a major contributor to the antimicrobial effect against the bacteria. In this study, H-bond acidity was also considered as a structural parameter involved with antimicrobial activity. The H-bond acidity is dependent on the polar properties of materials and is a measure of the H-bond strength [10]. The significance of establishing the correlation with both parameters was clearly demonstrated by an increase in the correction coefficient as well as an improvement in the other statistical values (Table 2). The addition of H-bond acidity as a structural parameter can be justified if the partial F-value is larger than the 95% significance levels [9]. The partial F-values estimated were larger than the 95% confidence levels, suggesting that the H-bond acidity and the partition coefficient are primary contributors to antimicrobial activity. The overall equations for the four bacteria against the antimicrobials were also significant. 3.4. Variation of the cut-off points A fall off in antimicrobial activity was observed as the number of carbon atoms of the substituents of the antimicrobials increased beyond a cut-off point. This is the so-called ‘cut-off point’ where biological activity falls rapidly or disappears as chain length increases. The antimicrobial activity of 4-hexyl phenol (0.055% of the MLC) was lower than that of 4-pentyl phenol, with an activity of 0.010% of the MLC against E. coli (Table 1). The MLCs for 4-hexyl and 4-pentyl phenol and for n-butyl, benzyl and phenethyldimethylsilanol were not determined against P. aeruginosa because of a significant reduction in antimicrobial activity of the materials as the alkyl chain increased beyond this cut-off point. The cut-off points varied with the type of bacteria. In the case of the Gram-negative bacterium E. coli, a clear cutoff point for antimicrobial activity was observed (Table 1). In contrast, no cut-off point was detected for S. aureus.
Table 2 Correlation equations for antimicrobial activities and physicochemical properties of silanols, alcohols and phenols against Escherichia coli, Staphylococcus aureus, Pseudomonas aeruginosa and Enterococcus faecalis Bacteria Gram-negative bacteria E. coli P. aeruginosa Gram-positive bacteria S. aureus E. faecalis Overall equation
Equations log(1/MLC) = 0.7391 log P − 1.294, n = 75, r = 0.91, s = 0.32, F = 352 (Eq. 1) log(1/MLC) = 0.63 log P + 0.0037␦ − 1.84, n = 75, r = 0.95, s = 0.24, F = 333, partial F-test = 55 (Eq. 2) log(1/MLC) = 0.672 log P – 1.138, n = 60, r = 0.83, s = 0.34, F = 128 (Eq. 3) log(1/MLC) = 0.55 log P + 0.004␦ − 1.718, n = 60, r = 0.94, s = 0.23, F = 216, partial F-test = 95 (Eq. 4) log(1/MLC) = 0.84 log P − 1.45, n = 75, r = 0.95, s = 0.26, F = 676 (Eq. 5) log(1/MLC) = 0.743 log P + 0.0035␦ − 1.971, n = 75, r = 0.98, s = 0.17, F = 873, partial F-test = 105 (Eq. 6) log(1/MLC) = 0.832 log P − 1.54, n = 72, r = 0.95, s = 0.26, F = 648 (Eq. 7) log(1/MLC) = 0.738 log P + 0.0032␦ − 2.011, n = 72, r = 0.98, s = 0.17, F = 837, partial F-test = 101 (Eq. 8) log(1/MLC) = 0.786 log P − 1.379, n = 282, r = 0.92, s = 0.30, F = 1543 (Eq. 9) log(1/MLC) = 0.679 log P + 0.0036␦ − 1.909, n = 282, r = 0.96, s = 0.22, F = 1639, partial F-test = 268 (Eq. 10)
Y.-m. Kim et al. / International Journal of Antimicrobial Agents 29 (2007) 217–222
n-Butyldimethylsilanol did not show a 7-log reduction against E. faecalis. It was observed that n-butyldimethylsilanol required 2 h to show a 7-log reduction.
221
against S. aureus. The correlation equations for the antimicrobials against the four bacteria (Table 2) were significant enough to reveal that the antimicrobial activities of silanols, alcohols and phenols were dependent on the same structural features, namely their lipophilic nature and H-bond acidity.
4. Discussion 4.2. The cut-off point 4.1. Structure–antimicrobial activity relationship In this study, the H-bond acidity and the partition coefficient are considered as key structural parameters. The results reported in Table 2 revealed that the two parameters are primary contributors to antimicrobial activity. The hydrophobicity of materials is an important parameter because it is directly related to membrane permeation. Hunt [13] proposed that the potency of aliphatic alcohols is directly related to their lipid solubility through the hydrophobic interaction between the alkyl chain from alcohols and the lipid region in the membrane. We suggest that a similar hydrophobic interaction might occur between the organic groups of the silanols accumulated in the lipid of bacterial membranes. As a consequence of their hydrophobic interaction, bacteria lose their membrane permeability, ultimately causing death of the bacteria [13]. Kubo et al. [14] have proposed that the antimicrobial effects may be due to a balance between the hydrophilic and hydrophobic portions of the molecule. This concept is reasonable because antimicrobial agents having only a hydrophobic portion but no hydrophilic region or relatively long alkyl chains in their chemical structure showed either reduced or no bioactivity [7]. The hydroxyl function of the alcohols is to orient and preferentially localise the materials near the membrane by virtue of hydrogen bonding with ester linkages of fatty acyl residues and with water molecules [14,15]. The hydrogen bonding strength is directly related to the H-bond acidity [10]. It is reasonable to suggest that the higher H-bond acidity of the silanols compared with the analogous alcohols contributes to a better balance through strong hydrogen bonding. A contribution of the Hbond acidity, although minor, was clear when agents with lower log P and higher H-bond acidity were compared. The higher activity of trimethylsilanol or 4-ethylphenol compared with ethyldimethylcarbinol or n-butyldimethylsilanol, respectively, appears to be attributed to their higher H-bond acidity. It is reasonable to suggest that the silanols may disrupt the cell membrane more efficiently than the analogous alcohols owing not only to their higher hydrophobicity but also their higher H-bond acidity. A linear free energy relationship between antimicrobial activity and the partition coefficient and H-bond acidity was demonstrated. The correlations were achieved over a wide range of structural variations and microbial organisms. Twenty-five chemical agents were tested against four bacteria. A wide range of partition coefficients (0.73–4.52) was covered. Their biological responses (i.e. MLCs) varied from 13.54% to 0.01% against E. coli and from 10.61% to 0.004%
There are several concepts to explain the occurrence of the cut-off point. Hansch and co-workers suggested that the fall off in activity with increase in hydrophobicity was due to a slow diffusion rate of molecules strongly bound in the membrane [4,8]. Hansch and Fujita also reported that the limited solubility of materials in the hydrophilic phase led to the fall off in activity [4], as suggested by Ferguson in 1939 [16]. Klarmann et al. [6] reported that for Gram-negative bacteria the maximum activity was reached at the amyl derivative of chlorophenol against Salmonella enterica serovar Typhi and at the hexyl derivative against Shigella flexneri. However, in the case of the Gram-positive bacteria S. aureus, maximum activity was observed with the n-octyl derivative. Lien et al. [8] also reported that Gram-positive bacteria showed a higher cut-off point (optimum partition coefficient value of 6) compared with Gram-negative bacteria (optimum partition coefficient value of 4). Our study showed a similar result, i.e. the Gram-negative bacteria E. coli and P. aeruginosa showed lower cut-off points than the Gram-positive bacteria S. aureus and E. faecalis.
Acknowledgments The authors acknowledge the financial support of Air Force Research Laboratory — Tyndall Air Force Base and Particle Engineering Research Center (PERC) at the University of Florida and the Microbiology Laboratory at the University of Florida for characterisation and biological activity tests.
References [1] Kim Y, Farrah S, Baney RH. Silanol—a novel class of antimicrobial agent. Electron J Biotechnol 2006;9:176–80. [2] Graiver D, Farminer KW, Narayan R. A review of the fate and effects of silicones in the environment. J Polym Environ 2003;11:129–36. [3] Denyer SP. Mechanisms of action of biocides. Int Biodeterior 1990;26:89–100. [4] Hansch C, Fujita T. Rho–sigma–pi analysis. Method for correlation of biological activity + chemical structure. J Am Chem Soc 1964;86:1616–26. [5] Daoud NN, Dickinson NA, Gilbert P. Antimicrobial activity and physico-chemical properties of some alkyldimethylbenzylammonium chlorides. Microbios 1983;37:73–85. [6] Klarmann EG, Shternov VA, Gates LW. The alkyl derivatives of halogen phenols and their bactericidal action. 1. Chlorophenols. J Am Chem Soc 1933;55:2576–89. [7] Tanner FW, Wilson FL. Germicidal action of aliphatic alcohols. Proc Soc Exp Biol Med 1943;52:138–40.
222
Y.-m. Kim et al. / International Journal of Antimicrobial Agents 29 (2007) 217–222
[8] Lien EJ, Hansch C, Anderson SM. Structure–activity correlations for antibacterial agents on Gram-positive and Gram-negative cells. J Med Chem 1968;11:430–41. [9] Kubinyi H. QSAR: Hansch analysis and related approaches. New York, NY: VCH Publisher; 1993. [10] West R, Baney RH. Hydrogen bonding studies. 2. The acidity and basicity of silanols compared to alcohols. J Am Chem Soc 1959;81:6145– 8. [11] Meylan WM, Howard PH. Atom/fragment contribution method for estimating octanol–water partition coefficients. J Pharm Sci 1995;84:83–92.
[12] Owen MJ. Siloxane surface-activity. Adv Chem Ser 1990;224:705–39. [13] Hunt WA. The effects of aliphatic alcohols on the biophysical and biochemical correlates of membrane function. Adv Exp Med Biol 1975;56:195–210. [14] Kubo I, Muroi H, Himejima M, Kubo A. Antibacterial activity of longchain alcohols—the role of hydrophobic alkyl-groups. Bioorg Med Chem Lett 1993;3:1305–8. [15] Dombek KM, Ingram LO. Effects of ethanol on the Escherichia coli plasma membrane. J Bacteriol 1984;157:233–9. [16] Ferguson J. The use of chemical potentials as indices of toxicity. Proc R S Lond B Ser 1939;B127:387.
Short communication
Structure–antimicrobial activity relationship for silanols, a new class of disinfectants, compared with alcohols and phenols Yun-mi Kim a , Samuel Farrah b , Ronald H. Baney a,∗ a
Department of Materials Science and Engineering, University of Florida, Gainesville, FL 32611, USA b Microbiology and Cell Science, University of Florida, Gainesville, FL 32611, USA Received 4 April 2006; accepted 25 August 2006
Abstract Triorganosilanols (R(CH3 )2 SiOH) were recently reported to exhibit unexpectedly strong disinfectant properties. The antimicrobial activities of silanols were significantly higher than their analogous alcohols. A study of the structural dependence of their antimicrobial activity was conducted against four bacteria, Escherichia coli, Staphylococcus aureus, Pseudomonas aeruginosa and Enterococcus faecalis. Silanols, alcohols with structures analogous to the silanols (R(CH3 )2 COH) and substituted phenols were evaluated as a single class of materials. The minimum lethal concentrations (MLCs), defined as the concentration required for a 7-log reduction in viable bacteria after 1 h exposure, were used to measure antimicrobial activity. Octanol–water partition coefficients (log P) and hydrogen bond acidities (␦), measured as the shift in frequency of the OH stretching band between free OH and hydrogen-bonded OH to diethyl ether oxygen by infrared spectroscopy, were utilised as dispersive and polar structural parameters, respectively. The correlation established by multiple regression analysis between antimicrobial activities and structural properties of silanols, alcohols and phenols against the four bacteria treated as a single family produced the following equation, log(1/MLC) = 0.679 log P + 0.0036␦ − 1.909 (n = 282, r = 0.96, s = 0.22). This equation and the significantly high correlation coefficient supported the hypothesis that the lipophilic properties and the H-bond acidities are primary factors for the antimicrobial action of silanols, alcohols and phenols. The high antimicrobial activity of silanols is explained by their greater H-bond acidity and their enhanced lipophilicity. © 2006 Elsevier B.V. and the International Society of Chemotherapy. All rights reserved. Keywords: Silanols; Structure–antimicrobial activity relationship; Regression analysis; Hydrophobicity; H-bond acidity
1. Introduction Silanols, i.e. silicon alcohols (R3 SiOH), are a new class of antimicrobials [1]. Their antimicrobial activity appears to be stronger than their analogous alcohols. Silanols are environmentally friendly materials because they are rather quickly degraded into environmentally benign silica, carbon dioxide and water in the environment [2]. Silanols have a hydrophilic portion, the hydroxyl group, and a hydrophobic region, the organic substituents, similar to alcohols and phenols widely used as antimicrobials. It is reason∗ Corresponding author. Present address: 168 Rhines Hall, University of Florida, Gainesville, FL 32611, USA. Tel.: +1 352 846 3785; fax: +1 352 846 3355. E-mail address: [email protected] (R.H. Baney).
able to predict that the mode of antimicrobial action of silanols would resemble those of the alcohols and phenols [3] because of the similarity of the chemical structures of these materials. The quantitative relationship between chemical structure and biological activity has received considerable attention in recent years because it allows one to predict bioactivity without an inordinate amount of time and effort. Moreover, insights gained from such studies can give insights into the mechanism of bioactivity. A method for quantitative correlation of biological activity and chemical structure was introduced by Hansch and Fujita in 1964 [4]. Structural dependence studies with antimicrobial activity have focused mainly on the effects of the hydrophobicity by testing a homologous series of samples, aliphatic alcohols, alkylated phenols and quaternary ammonium compounds [5–7].
0924-8579/$ – see front matter © 2006 Elsevier B.V. and the International Society of Chemotherapy. All rights reserved. doi:10.1016/j.ijantimicag.2006.08.036
218
Y.-m. Kim et al. / International Journal of Antimicrobial Agents 29 (2007) 217–222
Lien et al. [8] summarised the structure–activity relationship of antimicrobials by means of equations based on the method proposed by Hansch and Fujita [4]. The multiple regression analysis method was used to establish the correlation equations [9]. Lien et al. [8] observed that the correlations varied with bacteria and type of antimicrobials. These authors suggested that the lipophilicity of the molecule was the most important factor for the antimicrobial activities of the compounds, with a relatively minor contribution from electronic properties [8]. Our studies examine the relationship between antimicrobial activity and physicochemical properties for alcohols, phenols and previously unreported silanols as a single class of agents. The parameters utilised in our study were the octanol–water partition coefficient and the hydrogen bond acidity, which represent, respectively, the dispersive lipophilic property and the polar property of the hydroxylcontaining antimicrobials.
2. Materials and methods 2.1. Antimicrobials and their purities Eight silanols (R(CH3 )2 SiOH) (Fig. 1a) were prepared by hydrolysis of chlorosilane obtained from Gelest Inc. (Morrisville, PA). The chlorosilanes and water were mixed for 15–30 min in a diethyl ether solution with ammonium hydroxide and the silanols were collected by evaporation of the ether. The analogous alcohols, R(CH3 )2 COH (Fig. 1b) and nine phenols (Fig. 1c) were obtained from Acros Organics (Fairlawn, NJ). The purities of the silanols measured by 29 Si and 1 H nuclear magnetic resonance spectroscopy were 95 ± 3%, with the impurity consisting of disiloxanes. The antimicrobial effects of disiloxanes were also tested. The alcohols and phenols were used as received and their purities were >97% as determined by gas chromatography. 2.2. Preparation of bacteria and procedures for antimicrobial tests The bacterial strains employed were Escherichia coli C3000 (ATCC 15597), a laboratory strain of Staphylococcus
aureus (Department of Microbiology, University of Florida), Pseudomonas aeruginosa type strain (ATCC 10145) and Enterococcus faecalis type strain (ATCC 19433). Bacterial suspensions were prepared as follows. Bacteria were inoculated in a nutrient Columbia broth overnight at 37 ◦ C with constant agitation under aerobic conditions. The bacterial cells were collected by centrifugation at 500 relative centrifugal force for 10 min at 4 ◦ C and washed three times with sterilised distilled water. The bacterial pellet was resuspended in sterilised water after the final washing. The concentrations of the prepared bacterial suspensions were 2–6 × 108 colony-forming units (CFU)/mL. The antimicrobial activity tests of the materials were carried out by adding a given concentration of antimicrobial agent to 9 g of deionised water and 1 g of bacterial suspension containing a concentration of 2–6 × 108 CFU/mL. The solution was stirred for 1 h. The minimum lethal concentration (MLC) is defined as the lowest concentration of an agent required for a 7-log reduction in viable bacteria for the test organisms after a 1 h exposure. The MLCs were utilised as a measure of the biocidal activity of the antimicrobials. As a control, a sample was diluted in phosphate-buffered saline based on a serial 1:10 dilution. Then, 0.1 mL from each of the last two dilution tubes was plated on plate count agar (Difco, Sparks, MD). Serial dilution was not required to obtain the MLC of the agents because of complete killing, in this case a 7-log reduction of test bacteria. After incubating the plates for 24 h at 37 ◦ C, the colonies growing on the medium were counted to estimate the number of viable bacteria. The standard deviation values in Table 1 were determined by taking a mean of three tests. 2.3. Determination of hydrogen bond acidity and the octanol–water partition coefficient Relative H-bond acidities of the silanols, alcohols and phenols were determined by measuring ␦, defined as the shift in frequency of the OH stretching band between free OH and hydrogen-bonded OH to diethyl ether oxygen [10]. The shift is a measure of the relative proton-donating nature of the OH-containing compounds, because the protonaccepting ability remains constant since the same base (diethyl ether) was employed. The shift, ␦, is proportional
Fig. 1. Chemical structures of (a) silanols, (b) alcohols and (c) phenols.
Y.-m. Kim et al. / International Journal of Antimicrobial Agents 29 (2007) 217–222
219
Table 1 Minimum lethal concentration (MLC), H-bond acidity in ether (␦) and octanol–water partition coefficient (log P) of silanols, alcohols and phenols against Escherichia coli, Staphylococcus aureus, Pseudomonas aeruginosa and Enterococcus faecalis log P
␦
3.15 ± 0.05 1.32 ± 0.03 1.14 ± 0.05 0.50 ± 0.05 0.42 ± 0.03 0.23 ± 0.03 – 0.12 ± 0.02
1.14 1.5 1.63 2.12 2.36 2.85 2.62 3.34
239 260 237 237 267 257 236 252
13.33 ± 0.42 5.63 ± 0.18 5.67 ± 0.15 1.76 ± 0.04 0.93 ± 0.03 0.76 ± 0.06 0.75 ± 0.05 0.32 ± 0.02
0.73 1.08 1.22 1.71 1.95 2.2 2.44 2.93
126 130 125 125 137 126 129 129
1.51 2.06 2.16 2.55 3.04 3.28 3.53 4.02 4.52
278 270 318 269 270 252 270 269 271
MLC (%, g/g)a
Materials
R
Silanols (R(CH3 )2 SiOH)
Methyl Vinyl Ethyl n-Propyl Phenyl Benzyl n-Butyl Phenethyl
2.36 1.23 1.04 0.43 0.27 0.17 0.14 0.10
± ± ± ± ± ± ± ±
0.12 0.06 0.04 0.03 0.03 0.01 0.01 0.02
2.48 1.04 0.80 0.36 0.26 0.16 0.14 0.08
± ± ± ± ± ± ± ±
0.17 0.07 0.02 0.04 0.03 0.01 0.01 0.02
2.36 ± 0.01 1.0 ± 0.03 0.87 ± 0.04 0.40 ± 0.03 0.35 ± 0.05 – – –
Alcohols (R(CH3 )2 COH)
Methyl Vinyl Ethyl n-Propyl Phenyl n-Butyl Benzyl Phenethyl
13.54 5.23 5.09 1.68 0.96 0.67 0.7 0.26
± ± ± ± ± ± ± ±
1.27 0.06 0.04 0.04 0.06 0.03 0.05 0.01
10.61 4.37 4.17 1.69 0.78 0.65 0.59 0.26
± ± ± ± ± ± ± ±
0.19 0.05 0.12 0.04 0.03 0.02 0.02 0.01
9.79 ± 0.19 3.67 ± 0.08 3.96 ± 0.04 1.03 ± 0.03 0.71 ± 0.03 0.53 ± 0.02 0.63 ± 0.03 1.33 ± 0.28
Phenols (RC6 H5 OH)
Hydridob 4-Methyl 3-Chloro 4-Ethyl 4-Propyl 2-Phenyl 4-Butyl 4-Pentyl 4-Hexyl
0.70 0.410 0.13 0.13 0.053 0.12 0.013 0.010 0.055
± ± ± ± ± ± ± ± ±
0.01 0.01 0.02 0.02 0.006 0.03 0.002 0.002 0.005
0.61 0.35 0.11 0.14 0.045 0.085 0.015 0.008 0.004
± ± ± ± ± ± ± ± ±
0.02 0.02 0.01 0.02 0.005 0.015 0.005 0.002 0.001
0.62 ± 0.08 0.35 ± 0.05 0.11 ± 0.01 0.11 ± 0.01 0.052 ± 0.008 0.13 ± 0.03 0.060 ± 0.01 – –
E. coli
S. aureus
P. aeruginosa
E. faecalis
0.98 ± 0.03 0.45 ± 0.05 0.14 ± 0.01 0.17 ± 0.03 0.055 ± 0.005 0.13 ± 0.03 0.017 ± 0.003 0.012 ± 0.002 0.006 ± 0.001
–, indicates no MLC obtained. a Each MLC is the average value of three data points. b Hydrido, unsubstituented phenol.
to the strength of the H-bond. A transmission sampling technique was utilised for the infrared spectroscopy measurements. The octanol–water partition coefficient, log P, used as a measure of the lipophilicity of the agents, was calculated using the demo program ‘LogKow’ provided by Syracuse Research Corporation. The program was used to estimate log P using the atom/fragment contribution (AFC) method developed through multiple linear regressions of experimental log P values [11].
3. Results 3.1. Characterisation of the partition coefficient and the H-bond acidity for the antimicrobials The primary impurities in the silanols were identified as the silanol condensation products, disiloxanes. Antimicrobial activities of the disiloxanes were evaluated to determine their contribution to the antimicrobial activities of the silanols. The experiments performed using 10% of the disiloxanes tested against the four bacteria showed <1 log reduction, demonstrating that the silanols were responsible for the observed antimicrobial activities.
The H-bond acidity and the partition coefficient of the materials are shown in Table 1. The H-bond acidities of the silanols are almost two times higher than their analogous alcohols owing to electron back donation through the bond from the p orbital of oxygen to the vacant d orbital of silicon [10]. We propose that the higher acidities of the silanols play an important role in their enhanced biocidal properties. The partition coefficients estimated by the AFC method [11] showed a gradual increase as the alkyl chain length increased (Table 1). The partition coefficients of the silanols were higher than the analogous alcohols, as shown in Table 1. It is well known that silicon compounds, including the silanols, exhibit higher hydrophobicity than analogous organic compounds owing to flexible molecular chains and a lower group rotation energy barrier around a Si C bond than the carbon–carbon bond [12]. 3.2. Antimicrobial activities of silanols, alcohols and phenols against four bacteria The MLCs of the silanols, alcohols and phenols were determined against E. coli, P. aeruginosa, S. aureus and E. faecalis and are given in Table 1. The MLC decreases as the alkyl chain length of the silanols increases. The experimental results with the alcohols and phenols were similar to those
220
Y.-m. Kim et al. / International Journal of Antimicrobial Agents 29 (2007) 217–222
of the silanols, although their MLC values differed from the silanols. A lower MLC implies a higher antimicrobial activity. A change of substituents from a short alkyl chain to a longer alkyl chain resulted in increased biocidal activity. The susceptibility of bacteria to the antimicrobials can be compared based on the relative MLCs of the antimicrobials. The resistances of the four bacteria against the agents varied slightly. Overall, E. faecalis was the least susceptible bacterium, followed by E. coli, P. aeruginosa and S. aureus. It is significant to note that the silanols showed at least two times the antimicrobial activities of the analogous alcohols against the four bacteria. For E. coli, trimethylsilanol and phenethyldimethylsilanol showed 2.36% and 0.1% of the MLC, respectively, whereas the corresponding alcohols, t-butanol and phenethyldimethylcarbinol, exhibited 13.54% and 0.26% of the MLC, respectively. 4-Hexyl phenol, which contains the most carbon atoms of the materials tested, displayed the lowest MLC for S. aureus, as shown in Table 1. 3.3. Correlation between antimicrobial activity and structural parameters We propose that the physicochemical parameters of the silanols, which show a higher hydrophobicity and H-bond acidity compared with the alcohols, contribute to their enhanced antimicrobial activity. We demonstrated a quantitative structure–activity relationship using silanols, alcohols and phenols treated as a single class of antimicrobials. The correlation equations between their antimicrobial activities and their structural properties were created using a multiple regression analysis and are summarised in Table 2. The MLC values converted to the logarithms of (1/MLC) were plotted as a function of the partition coefficient and produced a linear free energy relationship [4] (Table 2) with statistically significant values. The significance and validity of the regression models could be evaluated by assessing the correlation coefficient r, the standard deviation s and the Fvalues [9]. If the correlation coefficient r is >0.9, the models are significant. If the standard deviation s is not much larger than 0.3, the models are acceptable. The F-value is a measure
of statistical significance of the regression model. If F-values are larger than the 95% significance limits, the model can be considered statistically significant. Statistical analysis data shown in Table 2 confirmed that the equations are statistically significant. Hydrophobicity was a major contributor to the antimicrobial effect against the bacteria. In this study, H-bond acidity was also considered as a structural parameter involved with antimicrobial activity. The H-bond acidity is dependent on the polar properties of materials and is a measure of the H-bond strength [10]. The significance of establishing the correlation with both parameters was clearly demonstrated by an increase in the correction coefficient as well as an improvement in the other statistical values (Table 2). The addition of H-bond acidity as a structural parameter can be justified if the partial F-value is larger than the 95% significance levels [9]. The partial F-values estimated were larger than the 95% confidence levels, suggesting that the H-bond acidity and the partition coefficient are primary contributors to antimicrobial activity. The overall equations for the four bacteria against the antimicrobials were also significant. 3.4. Variation of the cut-off points A fall off in antimicrobial activity was observed as the number of carbon atoms of the substituents of the antimicrobials increased beyond a cut-off point. This is the so-called ‘cut-off point’ where biological activity falls rapidly or disappears as chain length increases. The antimicrobial activity of 4-hexyl phenol (0.055% of the MLC) was lower than that of 4-pentyl phenol, with an activity of 0.010% of the MLC against E. coli (Table 1). The MLCs for 4-hexyl and 4-pentyl phenol and for n-butyl, benzyl and phenethyldimethylsilanol were not determined against P. aeruginosa because of a significant reduction in antimicrobial activity of the materials as the alkyl chain increased beyond this cut-off point. The cut-off points varied with the type of bacteria. In the case of the Gram-negative bacterium E. coli, a clear cutoff point for antimicrobial activity was observed (Table 1). In contrast, no cut-off point was detected for S. aureus.
Table 2 Correlation equations for antimicrobial activities and physicochemical properties of silanols, alcohols and phenols against Escherichia coli, Staphylococcus aureus, Pseudomonas aeruginosa and Enterococcus faecalis Bacteria Gram-negative bacteria E. coli P. aeruginosa Gram-positive bacteria S. aureus E. faecalis Overall equation
Equations log(1/MLC) = 0.7391 log P − 1.294, n = 75, r = 0.91, s = 0.32, F = 352 (Eq. 1) log(1/MLC) = 0.63 log P + 0.0037␦ − 1.84, n = 75, r = 0.95, s = 0.24, F = 333, partial F-test = 55 (Eq. 2) log(1/MLC) = 0.672 log P – 1.138, n = 60, r = 0.83, s = 0.34, F = 128 (Eq. 3) log(1/MLC) = 0.55 log P + 0.004␦ − 1.718, n = 60, r = 0.94, s = 0.23, F = 216, partial F-test = 95 (Eq. 4) log(1/MLC) = 0.84 log P − 1.45, n = 75, r = 0.95, s = 0.26, F = 676 (Eq. 5) log(1/MLC) = 0.743 log P + 0.0035␦ − 1.971, n = 75, r = 0.98, s = 0.17, F = 873, partial F-test = 105 (Eq. 6) log(1/MLC) = 0.832 log P − 1.54, n = 72, r = 0.95, s = 0.26, F = 648 (Eq. 7) log(1/MLC) = 0.738 log P + 0.0032␦ − 2.011, n = 72, r = 0.98, s = 0.17, F = 837, partial F-test = 101 (Eq. 8) log(1/MLC) = 0.786 log P − 1.379, n = 282, r = 0.92, s = 0.30, F = 1543 (Eq. 9) log(1/MLC) = 0.679 log P + 0.0036␦ − 1.909, n = 282, r = 0.96, s = 0.22, F = 1639, partial F-test = 268 (Eq. 10)
Y.-m. Kim et al. / International Journal of Antimicrobial Agents 29 (2007) 217–222
n-Butyldimethylsilanol did not show a 7-log reduction against E. faecalis. It was observed that n-butyldimethylsilanol required 2 h to show a 7-log reduction.
221
against S. aureus. The correlation equations for the antimicrobials against the four bacteria (Table 2) were significant enough to reveal that the antimicrobial activities of silanols, alcohols and phenols were dependent on the same structural features, namely their lipophilic nature and H-bond acidity.
4. Discussion 4.2. The cut-off point 4.1. Structure–antimicrobial activity relationship In this study, the H-bond acidity and the partition coefficient are considered as key structural parameters. The results reported in Table 2 revealed that the two parameters are primary contributors to antimicrobial activity. The hydrophobicity of materials is an important parameter because it is directly related to membrane permeation. Hunt [13] proposed that the potency of aliphatic alcohols is directly related to their lipid solubility through the hydrophobic interaction between the alkyl chain from alcohols and the lipid region in the membrane. We suggest that a similar hydrophobic interaction might occur between the organic groups of the silanols accumulated in the lipid of bacterial membranes. As a consequence of their hydrophobic interaction, bacteria lose their membrane permeability, ultimately causing death of the bacteria [13]. Kubo et al. [14] have proposed that the antimicrobial effects may be due to a balance between the hydrophilic and hydrophobic portions of the molecule. This concept is reasonable because antimicrobial agents having only a hydrophobic portion but no hydrophilic region or relatively long alkyl chains in their chemical structure showed either reduced or no bioactivity [7]. The hydroxyl function of the alcohols is to orient and preferentially localise the materials near the membrane by virtue of hydrogen bonding with ester linkages of fatty acyl residues and with water molecules [14,15]. The hydrogen bonding strength is directly related to the H-bond acidity [10]. It is reasonable to suggest that the higher H-bond acidity of the silanols compared with the analogous alcohols contributes to a better balance through strong hydrogen bonding. A contribution of the Hbond acidity, although minor, was clear when agents with lower log P and higher H-bond acidity were compared. The higher activity of trimethylsilanol or 4-ethylphenol compared with ethyldimethylcarbinol or n-butyldimethylsilanol, respectively, appears to be attributed to their higher H-bond acidity. It is reasonable to suggest that the silanols may disrupt the cell membrane more efficiently than the analogous alcohols owing not only to their higher hydrophobicity but also their higher H-bond acidity. A linear free energy relationship between antimicrobial activity and the partition coefficient and H-bond acidity was demonstrated. The correlations were achieved over a wide range of structural variations and microbial organisms. Twenty-five chemical agents were tested against four bacteria. A wide range of partition coefficients (0.73–4.52) was covered. Their biological responses (i.e. MLCs) varied from 13.54% to 0.01% against E. coli and from 10.61% to 0.004%
There are several concepts to explain the occurrence of the cut-off point. Hansch and co-workers suggested that the fall off in activity with increase in hydrophobicity was due to a slow diffusion rate of molecules strongly bound in the membrane [4,8]. Hansch and Fujita also reported that the limited solubility of materials in the hydrophilic phase led to the fall off in activity [4], as suggested by Ferguson in 1939 [16]. Klarmann et al. [6] reported that for Gram-negative bacteria the maximum activity was reached at the amyl derivative of chlorophenol against Salmonella enterica serovar Typhi and at the hexyl derivative against Shigella flexneri. However, in the case of the Gram-positive bacteria S. aureus, maximum activity was observed with the n-octyl derivative. Lien et al. [8] also reported that Gram-positive bacteria showed a higher cut-off point (optimum partition coefficient value of 6) compared with Gram-negative bacteria (optimum partition coefficient value of 4). Our study showed a similar result, i.e. the Gram-negative bacteria E. coli and P. aeruginosa showed lower cut-off points than the Gram-positive bacteria S. aureus and E. faecalis.
Acknowledgments The authors acknowledge the financial support of Air Force Research Laboratory — Tyndall Air Force Base and Particle Engineering Research Center (PERC) at the University of Florida and the Microbiology Laboratory at the University of Florida for characterisation and biological activity tests.
References [1] Kim Y, Farrah S, Baney RH. Silanol—a novel class of antimicrobial agent. Electron J Biotechnol 2006;9:176–80. [2] Graiver D, Farminer KW, Narayan R. A review of the fate and effects of silicones in the environment. J Polym Environ 2003;11:129–36. [3] Denyer SP. Mechanisms of action of biocides. Int Biodeterior 1990;26:89–100. [4] Hansch C, Fujita T. Rho–sigma–pi analysis. Method for correlation of biological activity + chemical structure. J Am Chem Soc 1964;86:1616–26. [5] Daoud NN, Dickinson NA, Gilbert P. Antimicrobial activity and physico-chemical properties of some alkyldimethylbenzylammonium chlorides. Microbios 1983;37:73–85. [6] Klarmann EG, Shternov VA, Gates LW. The alkyl derivatives of halogen phenols and their bactericidal action. 1. Chlorophenols. J Am Chem Soc 1933;55:2576–89. [7] Tanner FW, Wilson FL. Germicidal action of aliphatic alcohols. Proc Soc Exp Biol Med 1943;52:138–40.
222
Y.-m. Kim et al. / International Journal of Antimicrobial Agents 29 (2007) 217–222
[8] Lien EJ, Hansch C, Anderson SM. Structure–activity correlations for antibacterial agents on Gram-positive and Gram-negative cells. J Med Chem 1968;11:430–41. [9] Kubinyi H. QSAR: Hansch analysis and related approaches. New York, NY: VCH Publisher; 1993. [10] West R, Baney RH. Hydrogen bonding studies. 2. The acidity and basicity of silanols compared to alcohols. J Am Chem Soc 1959;81:6145– 8. [11] Meylan WM, Howard PH. Atom/fragment contribution method for estimating octanol–water partition coefficients. J Pharm Sci 1995;84:83–92.
[12] Owen MJ. Siloxane surface-activity. Adv Chem Ser 1990;224:705–39. [13] Hunt WA. The effects of aliphatic alcohols on the biophysical and biochemical correlates of membrane function. Adv Exp Med Biol 1975;56:195–210. [14] Kubo I, Muroi H, Himejima M, Kubo A. Antibacterial activity of longchain alcohols—the role of hydrophobic alkyl-groups. Bioorg Med Chem Lett 1993;3:1305–8. [15] Dombek KM, Ingram LO. Effects of ethanol on the Escherichia coli plasma membrane. J Bacteriol 1984;157:233–9. [16] Ferguson J. The use of chemical potentials as indices of toxicity. Proc R S Lond B Ser 1939;B127:387.