Comparison of solid-phase cytometry and the plate count method for the evaluation of the survival of bacteria in pharmaceutical oils
Hans J. Nelis, Laboratory for Pharmaceutical Microbiology Ghent University Harelbekestraat 72, B-9000 Ghent, Belgium. E-mail: email@example.com
Aim: To compare the survival of four bacterial strains (Escherichia coli, Proteus mirabilis, Staphylococcus aureus, Pseudomonas aeruginosa) in pharmaceutical oils, including jojoba oil/tea tree oil, carbol oil, jojoba oil and sesame oil.
Methods and Results: Oils were spiked with the test bacteria in a concentration of 104 CFU ml−1. Bacteria were extracted from oils with phosphate-buffered saline containing 0·5% Tween 20. Aliquots of the pooled water layers were analysed by solid-phase cytometry and plate counting. Plate counts dropped to zero for all test strains exposed for 24 h to three of the four oils. In contrast, significant numbers of viable cells were still detected by SPC, except in the jojoba oil/tea tree oil mixture and partly in sesame oil.
Conclusions: Exposure of bacteria for 24 h to the two oils containing an antimicrobial led to a loss of their culturability but not necessarily of their viability. The antibacterial activity of the jojoba oil/tea tree oil mixture supersedes that of carbol oil.
Significance and Impact of the Study: These in vitro data suggest that the jojoba oil/tea tree oil mixture more than carbol oil inhibits bacterial proliferation when used for intermittent self-catherization.
The effect of antimicrobials on micro-organisms in oil has rarely been addressed in the literature. In keeping with the general principle that antimicrobial preservatives are only active in an aqueous medium, Ibrahim and Sonntag (1993) have demonstrated the inactivity of preservatives in anhydrous oils such as liquid paraffin and sunflower oil. However, certain pharmaceutical oils may contain antimicrobials, as for example carbol oil (CO), consisting of 2% phenol in sesame oil (SO, w/v). This formulation, of rather obscure origin, is used as a lubricant for clean intermittent self-catherization in children with neurogenic bladder dysfunction. Phenol is supposed to inactivate micro-organisms originating from the repeated dipping of the catheter in the oil when insertion attempts fail. Concerns about the detrimental effects of phenol on the urinary tract mucosa have led to the investigation of alternative lubricants, i.e. jojoba oil (JJO), a liquid wax, containing higher fatty acids and alcohols, supplemented with presumably antibacterial additives. One such experimental formulation contains 5% (v/v) tea tree oil (TTO), an essential oil displaying potent antimicrobial activity in aqueous medium (Gustafson et al. 1998; Mann et al. 2000).
In this study, the survival of bacteria in four oils has been investigated to assess the potential infection risk during the use of the latter as lubricants for self-catheterization. Both the total viable count and the plate count were determined to verify if exposure of bacteria to antimicrobials in oil results in the presence of viable but nonculturable bacteria.
Materials and methods
TTO was incorporated in JJO in a concentration of 5% (v/v). CO (2% phenol in SO) and SO were provided by the Department of Paediatrics of the Ghent University Hospital.
Bacterial strains and culture conditions
Escherichia coli ATCC 8739 (American Type Culture Collection, Manasses, VA, USA), Proteus mirabilis ATCC 14153, Staphylococcus aureus ATCC 6538 and Pseudomonas aeruginosa ATCC 15442 were used to spike the oils. Each strain was grown in 10 ml of Trypticase Soy Broth (TSB; BD, Sparks, MD, USA) after inoculation with a loopful of a 24-h old culture and incubated for 24 h at 37°C.
Spiking of oils
Oils were supplemented with test bacteria in a concentration of 104 CFU ml−1. One-ml aliquots of freshly prepared overnight cultures were serially diluted in PBS to a concentration of 106 CFU ml−1. One hundred microlitres of this suspension was added to 9·9 ml of the four oils, to yield a final concentration of 104 CFU ml−1. The bacterial suspensions were vigorously mixed using a vortex mixer and 5 s sonication (Bransonic® 3510; Branson Ultrasonics B.V., Soest, the Netherlands) followed by homogenization for 1 h using the Rotamix RK (Analis, Namur, Belgium), to obtain a homogeneous distribution of bacteria in the oils.
Extraction of bacteria from spiked oils
Bacteria were extracted from the oils by adding 5 ml of PBS 0·5% Tween 20 to 1-ml aliquots of the oil. The samples were mixed thoroughly by vortexing at maximum speed and centrifuged at 15 000 g for 10 min. The aqueous layer was removed and the oil layer was re-extracted with 5 ml of PBS. Both aqueous fractions were pooled and subjected to solid-phase cytometry (SPC) and plate counting.
Solid-phase cytometry: apparatus
The solid-phase cytometer consisted of a laser-scanning unit (ChemScan® RDI; Chemunex, lvry-sur-Seine, France) equipped with a 488-nm argon laser that scanned a 25-mm diameter membrane in 3 min. Two photomultiplier tubes with wavelength windows set for the green (500–530 nm) and amber (540–585 nm) regions of the emission spectrum of fluorescein detected the fluorescence light emitted by the labelled cells. The signals produced were processed by a PC using a series of software discriminants that enabled the instrument to differentiate between valid signals (labelled cells) and background (autofluorescent particles). Results were displayed as green spots on a membrane filter image in a primary and, following elimination of background by software discrimination, a secondary scan map. To confirm the identity of the fluorescent cells after the scan, their appearance and morphology were visually inspected with an Olympus BX40 (Olympus, Tokyo, Japan) epifluorescence microscope, equipped with a moving stage (Prior Scientific, Fulbourn, Cambridge, UK) driven by the ChemScan® User Interface. The metal holder containing the membrane filter was removed from the ChemScan® and placed on the locating plate of the moving stage so as to retain the holder and the membrane filter in exactly the same position as in the ChemScan®. By highlighting a green spot in the secondary scan map, the motorized stage was automatically directed to the corresponding position on the membrane filter.
Determination of the total viable count (SPC) and of the total plate count
One-ml aliquots of the pooled aqueous fractions were filtered over a 0·45-μm pore size CycloblackTM-coated polyester membrane filter (25 mm diameter). The filter was rinsed with CSE/2 solution to counterstain particles and dead cells and hence to reduce background fluorescence. The filter was incubated on a cellulose pad saturated with 600 μl of viability reagent (ChemChrome V6; 1:100 in ChemSol B16), incubated for 30 min at 37°C and subjected to SPC. All reagents came from Chemunex and were prepared fresh daily. The number of viable, culturable cells in the pooled aqueous fractions [plate count (PC)] was determined after serial dilution in PBS. From each dilution, 1 ml was plated on TSA. The number of colonies was counted following incubation of the plates at 37°C for 24 h.
Results and discussion
Membrane filtration was investigated for use in SPC analysis to evaluate the survival of bacteria in pharmaceutical oils. However, upon direct filtration of oil, a layer covering the surface of the membrane filter precluded the detection of fluorescent cells. Likewise, no colonies were formed after incubation of a membrane filter on an agar plate. In contrast, detection of fluorescent cells [total viable count (TVC)] or colonies (PC) did become possible after aqueous extraction of the bacteria from the oil and thorough centrifugation. One-ml aliquots of the pooled aqueous fractions could be analysed without interference of oil droplets. TVCs and PCs obtained after 1 and 24 h exposure of four bacteria to the various oils are presented in Table 1. The data represent means ± SD for three independent analyses. All oils tested were spiked with approx. 104 CFU of test bacteria. Initial PCs and TVCs in the overnight cultures used for spiking differed by less than 0·4 log units. After 1 h of contact time, the PC for the nonantimicrobial oils JJO and SO approximated the theoretical numbers (except for P. aeruginosa in SO). For the antimicrobial oils JJO/TTO and CO, a moderate (E. coli, P. mirabilis) to strong (P. aeruginosa) decrease in PC was noted. The TVC remained unaffected after 1 h of contact time of the test bacteria with three of the four oils, including the antimicrobial CO. The near-quantitative recovery of the test bacteria from the two nonantimicrobial oils, as obtained both by SPC and plating indicated that there was little, if any, partition of cells into the oils. Only in the JJO/TTO mixture, the TVC deviated from the theoretical numbers. Twenty-four hours of contact resulted in PCs below the detection limit for both antimicrobial oils and SO. However, several log units of viable bacteria were demonstrated by SPC in CO as well as in JJO and SO. No viable cells of the four bacteria were detected in the JJO/TTO mixture after 24 h.
Table 1. Total viable counts and plate counts obtained after exposure of test bacteria for 1 and 24 h to four oils
|JJO/TTO||1||3·34 ± 0·26||3·45 ± 0·07||3·52 ± 0·10||2·29 ± 0·35||2·47 ± 0·29||3·25 ± 0·17||3·95 ± 0·01||ND|
|CO||1||3·24 ± 0·16||3·72 ± 0·07||4·65 ± 0·16||3·76 ± 0·04||1·91 ± 0·27||2·58 ± 0·07||4·64 ± 0·02||ND|
|CO||24||3·50 ± 0·07||2·03 ± 0·44||2·11 ± 0·16||ND||ND||ND||ND||ND|
|JJO||1||4·00 ± 0·10||3·86 ± 0·06||4·25 ± 0·04||4·39 ± 0·07||4·04 ± 0·02||4·09 ± 0·05||5·00 ± 0·05||4·22 ± 0·07|
|JJO||24||4·57 ± 0·16||3·38 ± 0·06||4·27 ± 0·10||3·96 ± 0·08||4·63 ± 0·12||3·84 ± 0·07||4·86 ± 0·13||3·48 ± 0·07|
|SO||1||4·08 ± 0·05||3·82 ± 0·10||3·38 ± 0·08||3·93 ± 0·08||4·03 ± 0·06||3·95 ± 0·06||4·28 ± 0·08||2·47 ± 0·07|
|SO||24||2·56 ± 0·05||2·14 ± 0·15||ND||ND||ND||ND||ND||ND|
The two enumeration methods used gave complementary information about the possible antibacterial effects of the oils. While the PC method relies on the capacity of the bacteria to multiply, and hence to form colonies, SPC detects all viable cells, including the nonculturables, having esterase activity and an intact cytoplasmic membrane.
Particularly after 24 h exposure of the test bacteria to the oils, TVCs and PCs did not coincide. CO and the JJO/TTO mixture inactivated bacteria to a different extent. While 24 h of contact with CO resulted in a loss of their culturability (PC below detection limit), a major fraction of cells remained viable with TVC values of more than 2 log units (except for P. aeruginosa). However, in JJO/TTO, no viable bacteria were detected after 24 h. Surprisingly, plain SO proved to inactivate the test bacteria to the same extent as when supplemented with phenol, indicating that phenol did not exhibit an additional antibacterial effect. In contrast, the antibacterial activity of JJO/TTO was clearly due to the TTO, as unsupplemented JJO failed to reduce both the TVC and the PC. Hence, the present in vitro data show the superiority of JJO/TTO over CO and suggest it may be a valuable alternative lubricant for intermittent self-catheterization.
The authors are indebted to Katerina Gyoreva and David Tonnard for their assistance.