ELSEVIER International Journal of Pharmaceutics 147 (1997) 51-60 intemational journal of pharmaceutics Effect of ethanol, propylene glycol and glycerol on the interaction of methyl and propyl p-hydroxybenzoate with Staphylococcus aureus and Pseudomonas aeruginosa Rula M. Darwish a, Sally F. Bloomfield a'b,* ~'Department of PharmacT, Kings' College London, University t~['London, London SW3 6LX, UK hFaculO' of Pharmacy, University o{ Jordan, Amman, Jordan Received 30 June 1996; revised 30 June 1996; accepted 30 October 1996 Abstract Co-solvents ethanol, propylene glycol and glycerol were used to increase the aqueous concentrations of methyl and propyl p-hydroxybenzoate above their saturation solubility. The increased aqueous concentration of parabens in co-solvent solutions was associated with increased antibacterial activity against Staphylococcus aureus and Pseu- domonas aeruginosa, but the extent of this effect increased with increasing hydrophobicity of the co-solvent. Addition of co-solvents to aqueous solutions of parabens caused a reduction in their octanol/water partitioning and a reduction in the uptake of the parabens by the cells, the extent of the reduction increasing as the hydrophobicity of the co-solvent increased. The addition of co-solvents also potentiated cell membrane damage as evidenced by leakage of radio-labelled phosphate from parabens-treated cells, but in this situation the extent of the damage was correlated with increasing hydrophobicity of the co-solvent. The results indicate that the increase in antibacterial activity of parabens preservatives in co-solvent solutions can largely be accounted for by their combined effects on the integrity of the cell membrane. © 1997 Elsevier Science B.V. Keywords: Parabens; Co-solvent; Preservation; Membrane damage 1. Introduction The alkyl esters ofp-hydroxybenzoic acid, com- monly known as the parabens, constitute one of * Corresponding author. Tel.: + 44 171 3334806; fax: + 44 171 3515307; e-mail: UDKJ077@kcl.ac.uk the most important groups of chemical agents used as preservatives in pharmaceutical, cosmetic and other products. Although various studies of the homologous series of parabens esters has shown that their activity increases with increasing ester chain length (Hansch et al., 1972; Eklund, 1985; Russell et al., 1985; Russell and Gould, 0378-5173/97/$17.00 ~5 1997 Elsevier Science B.V. All rights reserved, PII S0378-5173(96)04542-5
`
`Apotex, Inc. (IPR2019-00400), Ex. 1024, p. 001
`
`

`

`52 R.M. Darwish, S.F. Bloomfield / International Journal q/ Pharmaceutics 147 (1997) 51 60 1988), in practice the usefulness of the higher homologues is limited by their decreasing water solubility which in turn limits their bioavailablity within the system. In previous studies (Darwish, 1992; Darwish and Bloomfield, 1995) we have shown that pharmaceutical co-solvents such as ethanol, propylene glycol and glycerol can be used to increase the aqueous solubility of methyl and propyl parabens (MHB and PHB) preservatives, thereby increasing their antibacterial action against Staphylococcus aureus and Pseudomonas aeruginosa, but the extent of this effect was differ- ent for different co-solvents. The object of this investigation was to determine the mechanisms by which the different co-solvents exert their effect on the antibacterial activity of the parabens esters. trations higher than their saturation solubility (as determined by standard methods) could be pre- pared, which were stable over at least 48 h. Using solutions prepared in this way it was possible to compare the effects of the three co-solvents on the activity of MHB and PHB at concentrations up to 0.36% and 0.1%, respectively. 2.3. Preparation of solutions of [32p] [32P]Ortho-phosphate in acid solution (Amer- sham, UK; specific activity 6.06 MCi/ml)was diluted in 0.05 mM phosphate carrier and the pH adjusted to 7.0. 2.4. Evaluation of bactericidal activiO, 2. Material and methods 2.1. Bacterial test suspensions Staphylococcus aureus NCTC 10788 and Pseu- domonas aeruginosa NCTC 06750 were grown overnight at 37°C on nutrient agar containing (g 1-~) Lab-lemco, 1; yeast extract, 2.0; peptone 5.0; NaC1, 5.0; agar No. 3, 15.0; pH 7.4. Organisms were harvested, washed twice with peptone water, containing (g 1 ~) peptone, 1; NaC1, 5; pH 7.2, and diluted to contain the required number of colony forming units per ml (cfu ml ~). Media bases were obtained from Oxoid, Basingstoke, UK. 2.2. Preparation of test solutions Solutions of co-solvents were prepared by adding weighed quantities of ethanol (EtOH) (Hayman, Essex, UK), propylene glycol (PG) (Fisons, Loughborough, UK) and glycerol BP (Gly) (MacCarthy, Romford, UK) to sterile dis- tilled water. To prepare solutions of parabens, weighed quantities of methyl and propyl p-hy- droxybenzoate (BDH Chemicals, Dorset, UK) were added to sterile distilled water or co-solvent solution, and dissolved using sonication and/or gentle heat when necessary. By using heat or sonication, solutions containing parabens concen- Test solutions and bacterial test suspensions were placed in a water bath at 25°C to equilibrate. At zero time a sample of the bacterial test suspen- sion was added to each test solution to give an inoculum size of 106 l0 s cfu ml t in the test mixture. Samples were withdrawn at predeter- mined time intervals (t), neutralised by dilution in quarter strength Ringer's solution (1 in 100), and viable counts estimated by plating (in duplicate) 50-/~1 aliquots onto nutrient agar plates using a Model D spiral plater (Don Whitley, Shipley, UK). Plates were incubated at 37°C for 24 h. Colonies were counted using the laser counter (Model 500 Colony Counter, Don Whitley, Ship- ley, UK) and the log number of cfu ml- 1 deter- mined. The microbicidal effect (ME value) was calculated as log number of cfu ml 1 at time t minus the log number of cfu ml ~ at time zero. Where samples gave no detectable survivors (NDS) the actual number of survivors in the final dilutions is < 2 × l0 3 cfu ml ~ (i.e. < 1 cfu in 50 ml). Preliminary experiments showed that dilution of 1 ml of the reaction mixtures with 99 ml of quarter strength Ringer's solution was sufficient to neutralise the effect of the parabens. 2.5. The effect of co-solvents on the partitioning of parabens between water and octanol Accurately weighed quantities of parabens were dissolved in 1-octanol by sonication. The same
`
`Apotex, Inc. (IPR2019-00400), Ex. 1024, p. 002
`
`

`

`R.M. Darwish, S.F. Bloomfield / bTternational Journal ~?[ Pharmaceutics 147 (1997) 51 60 53 volumes of distilled water or co-solvent solutions (3 M) saturated with 1-octanol were added. Tubes were stoppered and inverted 20 times per min for 48 h at room temperature. The mixtures were centrifuged (1000 x g, 15 rain) and the concentra- tion of parabens in the aqueous layer determined by spectrophotometric analysis at 255 nm. A par- tition coefficient was calculated as the ratio of the parabens concentration in the oil phase:parabens concentration in the aqueous phase. 2.6. The effect oJ co-solvents on the adsorption of parabens by the bacterial cells Different concentrations of parabens were dis- solved in distilled water and co-solvent solutions using sonication or gentle heat when necessary. Bacterial test suspension was added to give a cell concentration of 6 x 109 cfu ml ~ and the test mixture maintained at 25°C for 1 h. Cells were removed by centrifugation (1000 x g, 15 rain) and the parabens concentration of the supernatant determined by spectrophotometric analysis at 255 rim. 2. 7. The effect of parabens and co-solvents on the bacterial membrane Bacterial suspensions of cell density 1 × 101° cfu ml ~ were incubated in 0.1% pepton e water with 1 ml of 0.05 mM [32p]phosphate for 2 h at 37°C. Cells were harvested by centrifugation (3000 x g, 10 rain) washed and resuspended in quarter strength Ringer's solution. Quantities of preloaded cells were added to the test solutions (parabens, co-solvents, parabens/co-solvents solu- tions) to give a cell density of 109 cfu ml - ~ in the test mixture. Controls were set up as described above using sterile distilled water instead of test solutions. To determine the initial [32p] content of the cells, one of the control tubes was heated at 100°C for 10-15 rain. Samples of test mixture were withdrawn at predetermined time intervals, filtered through membrane filters (millipore, pore size 0.2/~m). The membranes were added to 5 ml optiphase 'Hisafe' liquid scintillant (Amersham, UK) and the [32p] content of the cells determined using an LKB Wallac 1209 Liquid Scintillation Counter (Wallac, Milton Keynes, UK). Results were expressed as a percentage of the initial [32p] content of the cells. 2.8. Replication qf results All results were confirmed by not less than two replicate experiments. 3. Results and discussion Previous investigations (Darwish, 1992; Dar- wish and Bloomfield, 1995) showed that pharma- ceutical co-solvents can be used to increase the aqueous solubility of methyl and propyl parabens preservatives and thereby increase their antibacte- rial activity against S, aureus and Ps. aeruginosa. The least polar agent EtOH at 3 M, was the most effective co-solvent giving a saturation solubility of 0.44% and 0.09% for MHB and PHB, respec- tively, and producing the most significant increase in activity. By contrast, the most polar co-solvent Gly (3 M) was the least effective giving solubilities of 0.13% and 0.035% for MHB and PHB, respec- tively, with relatively less increase in activity at equivalent parabens concentrations. Saturation solubility determinations agree with the findings of Yalkowsky (1981) who showed that the solubil- ity of a drug depends on its polarity with respect to the co-solvent, the best solvent for a particular solute being the one which matches its polarity. Mechanism of action studies (Freese et al., 1973: Freese and Levin, 1978; Eklund, 1980, 1985) suggest that the parabens act mainly by causing disorganisation of the microbial cell membrane: at low (bacteriostatic) concentrations parabens ap- pear to cause energy uncoupling which inhibits the uptake of metabolites, whilst at higher (bacte- ricidal) loss of the membrane semipermeability occurs. From studies of the effects of hydrophilic/ hydrophobic balance on antibacterial activity Hansch et al. (1972) concluded that the activity of the membrane active antibacterials such as the parabens depends on their ability to move freely in the aqueous phase, and yet be lipophilic enough to partition through the microbial outer cell envelope (where present) and the cytoplasmic
`
`Apotex, Inc. (IPR2019-00400), Ex. 1024, p. 003
`
`

`

`54 R.M. Darwish, S.F. Bloomfield / International Journal of Pharmaceutics 147 (1997) 51 60 membrane. Studies of various alcohol and glycol compounds (Hugo and Russell, 1982; Russell and Chopra, 1996) show that these agents also have antimicrobial activity which is associated with damage to the cytoplasmic membrane. From the various studies described above a number of possible explanations for the observed effects of the co-solvents on the activity of the parabens can be proposed. On one hand the co-solvents may interact with the parabens in the aqueous phase thereby altering the hydrophilic/ lipophilic balance and affecting their uptake and/ or interaction with the cell. On the other hand, since the co-solvents themselves cause damage to the cell membrane, potentiation of antimicrobial action may result from a direct effect on the microbial outer cell envelope (where present) or cytoplasmic membrane. To evaluate these possi- bilities adsorption isotherms, octanol o/w parti- tion coefficients and cytoplasmic membrane damage were determined for MHB and PHB in aqueous solution and co-solvent systems against S. aureus and P. aeruginosa, and the results com- pared with the effects of the co-solvents on the antibacterial activity of the parabens assessed un- der comparable conditions. Results in Fig. 1 show the bactericidal activity of combinations of concentrations of PHB from 0.04% to 0.1% in 3 M co-solvent solutions, com- pared with that of the 3 M co-solvent solutions alone and 0.04% PHB alone. Aqueous solutions of PHB at saturation solubility (0.04%) had little effect on S. aureus but produced a 2-3 log reduc- tion against Ps. aeruginosa in 2-4 h. PG and Gly at 3 M had little or no activity against both S. aureus and Ps. aeruginosa, over 24 h but 3 M EtOH produced total kill (i.e. no detectable sur- vivors) within 2 h. For all the PHB/co-solvent solutions an increase in PHB concentration above the aqueous saturation solubility was associated with increased antibacterial activity but, for a given increase in parabens concentration, the in- crease in activity over and above that of the co-solvent alone was generally greater for PHB in EtOH compared with PHB in PG which in turn was greater than that observed with PHB in Gly. With PHB in Gly ME values of solutions up to 0.1% against Ps. aeruginosa were actually less than that of 0.04% saturated solution of PHB alone. Results with MHB which similarly also show that, for a given increase in concentration, the increase in activity was generally greater for MHB solutions in EtOH compared with PG and Gly solutions have been previously described (Darwish, 1992; Darwish and Bloomfield, 1995). Evidence from the literature (Higuchi and Kuramoto, 1954; Guttman and Higuchi, 1957; Patel and Kostenbauder, 1958; Van Doorne et al., 1988) indicates the tendency of some formulation components used in pharmaceuticals and cosmet- ics to form molecular complexes with parabens preservatives. Van Doorne et al. (1988) demon- strated that, although fl-cyclodextrins could be used to increase the aqueous solubility of parabens esters, the nature of the complexation was such that there was a decrease in antibacterial activity. Patel and Kostenbauder (1958) demon- strated a similar effect for solutions of parabens in Tween-80. To evaluate the possibility of interac- tion between EtOH, PG and Gly co-solvents and parabens, UV spectra of MHB and PHB solutions in water and 3 M co-solvent solutions were deter- mined and compared (results not shown). From the chemical structure of the parabens and the co-solvents it would be expected that the interac- tion between these molecules would be largely hydrophobic. This correlates with the fact that there was no apparent change in the UV spectrum of parabens following addition of co-solvents in- dicating no complex formation involving charge transfer. By contrast, results in Table 1 indicate that addition of co-solvents to solutions of MHB and PHB was associated with a decrease in the parti- tioning of parabens into octanol. Results indicate that in general the reduction in partitioning was greatest for the least polar co-solvent EtOH and least for the most polar co-solvent Gly, although there was some discrepancy in the results for MHB in EtOH and PG. It must be borne in mind that the octanol/water partitioning of EtOH, PG and Gly was not determined and may have had some effect on the parabens partitioning. The relative effects of co-solvents on the oc- tanol/water partitioning of PHB is in agreement with their effects on the uptake of PHB by cells of
`
`Apotex, Inc. (IPR2019-00400), Ex. 1024, p. 004
`
`

`

`R.M. Darwish, S.F. Bloomfield/International Journal of Pharmaceutics 147 (1997) 51-60 55 2 F c_ 8 1 = NDS ~3 z o ~ xos 3 2 1 0 Gry ~M a) PHB 0 04'~, lID PH80.04%,Gly 3M f PHB 00B%+Gly 3M 2. o 100 2O0 300 MIN$ PG lU | PHEI 0.04% -O-- PHB O.04%~PG 3 PH8 0.O6%÷PG $M ~,- PHS o ~*PG 3 /, tOO 200 300 MINS c) 100 200 300 MINS =~ ~ Q,y 3u -,~- PHB 004% ~PHa 0 04~,*Gq 3M ~ PHB 0.O6%+GPy 3M (~ 2 ~ PHB Ol%+Gly 3M 0 T ' l ' [ ' ~ os 3 I 0 ND5 3 z o o 2 _z 0 100 2OO 30O MINS -IF- PG ~M | pH6 o.o 4./. / eJ -0- PHB 0.04%+FG 3 PHB D.08%+pG 3 -~ PHB O.I%*PG 3M 100 200 300 MINS t) • i 100 200 30O MINS Fig. I. Log reduction in viable counts of (a), (b), (c) Staphyloeoccus aureus and (d), (e), (f) Pseudomonas aeruginosa in 3 M co-solvents glycerol (Gly), propylene glycol (PG) and ethanol (EtOH), propyl p-hydroxybenzoate (PHB) 0.04%, and combinations of PHB/co-solvents, (NDS, no detectable survivors). S. aureus and Ps. aeruginosa. Results in Fig. 2 show that, although the amount of PHB adsorbed by the bacterial cell increased with increasing concentration of parabens, the uptake of PHB from solutions of PHB in Gly was slightly greater than for PHB in PG, which in turn was greater than for PHB in EtOH. The results imply that, since the parabens are less soluble in Gly, which is the most polar co-solvent, than in EtOH which is the least polar co-solvent used, the tendency of parabens to leave the Gly containing system and adsorb onto the cells is greater than the tendency to leave the EtOH containing system. However, the order of the increase in the uptake and parti- tioning is in contradiction with the order of activ- ity, PHB in EtOH being more active than PHB in
`
`Apotex, Inc. (IPR2019-00400), Ex. 1024, p. 005
`
`

`

`56 R.M. Darwish, S.F. Bloomfield / International Journal oJ Pharmaceutics 147 (1997) 51-60 Table 1 Octanol/water partition coefficients for methyl and propyl p-hydroxybenzoate solutions in water and co-solvent solutions Aqueous phase Methyl p-hydrox- Propyl p-hydroxy- ybenzoate benzoate Water 77.20 670.12 Glycerol (3 M) 52.48 480.20 Propylene glycol 39.81 236.40 (3 M) Ethanol (3 M) 49.00 233.40 PG which in turn is more active than PHB in Gly. It was also noted that, although total uptake from 0.1% PHB solutions in all co-solvents was signifi- cantly greater than for 0.04% PHB alone, the additional uptake of parabens from PHB solution in Gly was associated with relatively little increase in activity. By contrast, the total uptake for solu- tions of PHB in EtOH and PG was less than for solutions of PHB in Gly, but a more marked increase in activity was observed. From this it is concluded that, although the co-solvents have an effect on uptake of parabens, it is not responsible (and may possibly have a counteracting effect) for the observed pattern of increased activity of parabens in the different co-solvent solutions. As found by Furr and Russell (1972), adsorption isotherms for MHB could not be determined since the amount of ester adsorbed was undetectable using the method used here which was based on the difference in parabens concentration in the supernate before and after addition of bacterial cells. As stated previously, the possibility exists that a direct interaction of the co-solvents with the bac- terial inner and outer cell membranes could be responsible for the observed effects of the co-sol- vents on the efficacy of the parabens. The gener- ally higher resistance of Gram-negative bacteria to parabens as reported by a number of workers (Hugo and Russell, 1982; Russell and Chopra, 0.03 0,03 0.02 > ~e 0- 0.01 "6 0.00 Gly 3M - PQ 3M 0 EtOH 3M ,3 weter i~/// / / b) 1 / / l- ely 3M / PG 3M / ~ EtOH 3M / ~ water / 0,02 "~ I// -- 0.01 o o_ 0.00 0,00 0,02 0.04 0.00 0.08 0.10 0.12 0.00 0.02 0.04 0.06 0.08 0.10 0.12 Initial concentration of PHB %w Initial concentration of PHB %w Fig. 2. Adsorption isotherms for suspensions of (a) Staphylococcus aureus and (b) Pseudomonas aeruginosa containing 6 × 10 9 CfU ml-~ in solutions of propyl p-hydroxybenzoate (PHB) in water and 3 M solutions of glycerol (Gly), propylene glycol (PG) and ethanol (EtOH) (contact time 1 h).
`
`Apotex, Inc. (IPR2019-00400), Ex. 1024, p. 006
`
`

`

`R.M. Darwish, S.F. Bloomfield ,' International Journal ~1" Pharmaceuties 147 (1997) 51 60 57 100 100 80 60 ?, ~. 4o 2O -I~ GIy 3M PG 3M EtOH 3M -: water J .? i o 40 .-~ 20 O- j~ . J" 4l~ Gly 3M PG 3M EtOH 3M waler y J 0 10 20 30 0 10 20 30 hours hours Fig. 3. The effect of 3 M solutions of glycerol (Sly), propylene glycol (Pa) and ethanol (EtOH) on the release of radio-labelled phosphate from (a) Staphylococcus aureus and (b) Pseudomonas aeruginosa suspensions containing 10 9 cfu ml i 1996) is usually attributed to the presence of the outer cell membrane which hinders access of the parabens to the cytoplasmic membrane. Uptake of parabens through the Gram-negative outer cell membrane is thought to occur by the hydrophobic route. This is confirmed by evidence showing that lipopolysaccha.ride-deficient mutants are more sensitive to parabens than the wild strains, and that the extent of the increase in sensitivity of the mutants is greater for butyl and propyl ester than for methyl and ethyl esters (E1-Falaha et al., 1983; Russell et al., 1985, 1987). Since the co-solvents are known to disrupt the cytoplasmic membrane (Eaton et al., 1982; Russell and Chopra, 1996) it might be expected that they could also interact with the Gram-negative outer cell membrane, thus facilitating access of parabens to the cytoplasmic membrane. The results in this paper indicate that this is unlikely to be the case since the effects of the co-solvents on the adsorption and bactericidal action of parabens was no greater for the Gram- negative bacteria Ps. aeruginosa than the Gram- positive organism S. aureus where the outer membrane is absent. Since the co-solvents EtOH, PG and Gly represent relatively small hydrophilic molecules it is likely that these molecules pass freely across the outer membrane through the porins (i.e. the hydrophilic routes) to reach the cytoplasmic membrane where they exert their ac- tion (Russell and Chopra, 1996). Cytoplasmic membrane damage associated with the effects of the parabens, co-solvents and parabens/co-solvent combinations was determined by measuring leakage of radio-labelled phosphate from cells of S. aureus and Ps. aeruginosa. Results in Fig. 3 indicate that the co-solvents alone at 3 M concentration produced an increase in the rate of leakage of radio-labelled phosphate as com- pared with untreated cells indicating significant membrane damage. The effect of the co-solvents on the release of phosphate correlates with their relative efficacy as antibacterial agents, the most
`
`Apotex, Inc. (IPR2019-00400), Ex. 1024, p. 007
`
`

`

`58 R.M. Darwish, S.F. Bloomfield / International Journal of Pharmaceutics 147 (1997) 51-60 60 4O ! "~ 20 ~e eo "l 5O 40 o 30 i "6 20 10 7O so i b) -l I ~ I ~ ! ' I I ' o 0.00 0,02 0,04 0,06 0,00 0,10 0.12 0.00 PHi] concenlration % wlv I ' I ~ I i ~ J I ' 0.02 004 o oe 008 olo o12 PHB conoentralion % wlv Fig. 4. The effect of propyl p-hydroxybenzoate (PHB) and combinations of PHB with 3 M solutions of glycerol (Gly), propylene glycol (PG) and ethanol (EtOH) on the release of radio-labelled phosphate from (a) Staphylococcus aureus and (b) Pseudomonas aeruginosa suspensions containing l09 cfn ml-1 (contact time 2 h). hydrophobic co-solvent EtOH causing the most extensive membrane damage and loss of viabi- lity and Gly the least. Results in Figs. 4 and 5 show that addition of MHB and PHB to the co-solvents solutions caused an increase in the release of phosphate which increased with in- creasing parabens concentration. The results also show a correlation between the extent of the membrane damage caused by the co-solvent/ parabens combinations and their antimicrobial action, the most effective combination PHB/ EtOH causing the greatest leakage of phosphate and the least effective combination, PHB/Gly, causing the least extensive leakage. It was noted that the osmotic effect of the co-solvents at 3 M concentration might be expected to have a protec- tive effect against the loss of cytoplasmic con- stituents. This in turn would cause a reduction in bactericidal activity of the co-solvents combina- tions whereas the opposite effect was found to occur. In general these results indicate that for any parabens/co-solvent combination the antimicro- bial activity is determined by their combined ef- fects on the integrity of the cell membrane permeability. Within the limits of the experimen- tal error, the dose/response curves (Figs. 4 and 5) for the three co-solvents in combination with PHB were more or less parallel suggesting that the effects are additive, a given increase in parabens concentration being associated with the same pro- portional increase in leakage. However, this does not correlate with the evaluations of bactericidal activity. Time/survivor curves as shownin Fig. 1 suggest that the proportional increase in activity for a given increase in parabens concentration was greater for EtOH than for PG which in turn was greater than for Gly. This latter result suggests a possible synergistic effect, but may equally be determined by the sigmoid shape of the time/sur- vivor curves which dictates that a relatively small increase in membrane damage caused by the ef-
`
`Apotex, Inc. (IPR2019-00400), Ex. 1024, p. 008
`
`

`

`R.M. Darwish, S.F. Bloomfield / International Journal of Pharmaceuties 147 (1997) 51-60 59 i i L 60 a) / , ] /' / 50 /,, = -Q. wster ] 0 I : I I ; 0,0 O.f 0.2 0.3 MHB concentration % wlv 40 © ~ 30 E ~ 2o • Giy 3M PG 3M -0 EtOH 3M ~-wale~ ! i [ i I : I ,' 0.4 0.0 01 0.2 0.3 0.4 MHB concentration % w/v Fig. 5. The effect of methyl p-hydroxybenzoate (MHB) and combinations of MHB with 3 M solutions of glycerol (Gly), propylene glycol (PG) and ethanol (EtOH) on the release of radio-labelled phosphate from (a) Staphylococcus aureus and (b) Pseudomonas aeruginosa suspensions containing 109 cfu ml 1 (contact time 2 h). fects of combining the two agents, could produce a marked increase in activity. The observed ef- fects of the co-solvent/parabens combination also do not account for the fact that, although the addition of PHB (0.04-0.1%) to glycerol 3 M caused some increase in leakage, this was associ- ated with a decrease in activity as compared to the activity of PHB (0.04%) alone. Overall the results of these studies indicate that the inclusion of a co-solvent has the potential to increase the usefulness of the parabens as preservative agents for pharmaceuticals, cosmetics etc. by facilitating an increase in aqueous concentration above their saturation solubility. Indications are that the ef- fectiveness of the system depends on the choice of co-solvent and can be optimised by selecting the most appropriate co-solvent which itself causes maximum damage to the cell membrane but does not significantly affect the bioavailability and con- sequently the adsorption of parabens by the bac- terial cell. References Darwish, R., The effect of co-solvents on the activity of preservatives in oral aqueous dosage forms. PhD Thesis, University of London, 1992. Darwish, R. and Bloomfield, S.F., The effects of co-solvents on the antibacterial activity of paraben preservatives, lnt. J. Pharm., 119 (1995) 183 192. Eaton, L.C., Tedder, T.D. and Ingrain, L.O., Substance and alcohol actions. Misuse, 3 (1982) 77. EI-Falaha, B.M.A., Russell, A.D. and Furr, J.R., Sensitivities of wild type and envelope-defective strains of Escheriehia eoli and Pseudomonas aeruginosa to antibacterial agents. Mierobios, 38 (1983) 99--105. Eklund, T., Inhibition of growth and uptake processes in bacteria by some chemical food preservatives. J. Appl. Bacteriol,, 48 (1980) 423-432. Eklund, T., The effect of sorbic acid and estersof p-hydroxy- benzoic acid on the proton motive force in Escherichia coil membrane vesicles. J. Gen. Microbiol., 313 (1985) 73 76. Freese, E. and Levin, B.C., Action mechanisms of preserva- tives and antiseptics. Dev. Incl. Mierobiol., 19 (1978) 207- 227, Freese, E., Shen, C.W. and Galliers, E., Function of lipophilic acids as antimicrobial food additives. ,Vature, 241 (1973) 321-324.
`
`Apotex, Inc. (IPR2019-00400), Ex. 1024, p. 009
`
`

`

`60 R.M. Darwish, S.F. Bloom[ieM / International Journal o]' Pharmaceutics 147 (1997) 51 60 Furr, J.R. and Russell, A.D., Some factors influencing the activity of esters of p-hydroxybenzoic acid against Serratia marceseens. Microbios, 5 (1972) 189 198. Guttman, D. and Higuchi, T., Possible complex formation between macromolecules and certain pharmaceuticals. X. The interaction of some phenolic compounds with polyethylene glycols, polypropylene glycols, and polyvinyl pyrrolidone. J. Am. Pharm. Assoc., 45 (1957) 659-664. Hansch, C., Coubeils, J.L. and Leo, A., The antimicrobial structure activity relationship in esters of 4-hydroxy ben- zoic acid. Chim. Ther., 6 (1972) 427 433. Higuchi, T. and Kuramoto, R., Study of possible complex formation between macromolecules and certain pharma- ceuticals. J. Am. Pharm. Assoc'., 43 (1954) 398 401. Hugo, W.B. and Russell, A.D., Types of antimicrobial agents. In Russell, A.D., Hugo, W.B. and Ayliffe, G.A.J. (Ed.), Principles and Practice of Disir!/bction. Preservation and Sterilisation, Blackwell, London, 1982, pp, 7 -88. Patel, N.K. and Kostenbauder, H.B., Interaction of preserva- tives with macromolecules. I. Binding of para-hydroxyben- zoic and esters and polyoxyethylene 20 Sorbitan Mono-oleate (Tween 80). J. Am. Pharm. Assoc., 47 (1958) 289 295. Russell, A.D. and Gould, G.W., Resistance of Enterobacteria- cae to preservatives and disinfectants. J. Appl. Baeteriol. Syrup. Suppl., (1988) 167S 195S. Russell, A.D. and Chopra, 1., Antiseptics, disinfectants and preservatives, their properties, mechanisms of action and uptake into bacteria. In Russell, A.D. andChopra, I. (Ed.), Understanding Antibacterial Action and Resistance, 2rid edn, Ellis Horwood, Chichester, 1996, pp. 97 131. Russell, A.D., Furr, J.R. and Pugh, W.J., Susceptibility of porin and lipopolysaccharide-deficient mutants of Es- cheriehia eo[i to a homologous series of esters ofp-hydroxy benzoic acid. hTt. J. Pharm., 27 (1985) 163 173. Russell, A.D., Furr, J.R. and Pugh, W.J., Sequential loss of outer membrane lipopolysaccharide and sensitivity of Es- cherichia coli to antibacterial agents. Int. J. Pharm., 35 (1987) 227 233. Van Doorne, H., Bosch, E.H. and Leek, C.F., Formation and antimicrobial activity of complexes of/~'-cyclodextrin and some antimycotic imidazole derivatives. Pharm. Weekl. Sci. Edn., 10 (1988) 80 85. Yalkowsky, H.S., Techniques g/ Solubilization of Drugs, Marcel Dekker, New York, 1981.
`
`Apotex, Inc. (IPR2019-00400), Ex. 1024, p. 010
`
`