Development of Standardized Methods for Assessing Biocidal Efficacy of Contact Lens Care Solutions Against Acanthamoeba Trophozoites and Cysts

Abstract

Purpose.: To investigate experimental variables in the development of standardized methods to assess the efficacy of contact lens disinfection systems against the trophozoite and cysts of Acanthamoeba spp.

Methods.: A. castellanii (ATCC 50370), A. polyphaga (ATCC 30461), and A. hatchetti (CDC: V573) were adapted to axenic culture and used to produce cysts either with Neff's encystment medium (NEM) or starvation on nonnutrient agar (NNA). Challenge test assays and a most probable number approach were used to compare the trophozoite and cysticidal efficacy of four multipurpose disinfectant solutions (MPDSs) and a one-step hydrogen peroxide system (with and without the neutralizing step).

Results.: With trophozoites, four of four MPDSs and the one-step peroxide system gave ≥3 log₁₀ kill for all strains 6 hours, regardless of culture medium used. Greater resistance was found against cysts, with results for MPDSs varying by species and method of cyst production. Here, 1–3 log₁₀ kill was found with NEM cysts for three of four MPDSs compared with one of four for the NNA cysts at 6 hours ( A. castellanii and A. polyphaga , only). The one-step peroxide system gave 1–1.9 log₁₀ kill with NEM cysts and 0.8–1.1 for NNA cysts. Only 3% hydrogen peroxide gave total kill (>3 log₁₀) of NNA cysts at 6 hours.

Conclusions.: A reproducible method for determining the susceptibility of Acanthamoeba trophozoites and cysts to contact lens care systems has been developed. This will facilitate assay standardization for assessing the efficacy of such products against the organism and aid development of improved disinfectant and therapeutic agents.

Introduction

Keratitis due to the small free-living amoeba Acanthamoeba is a rare but potentially blinding infection, with contact lens wearers accounting for approximately 90% of cases. ^1–4 Poor compliance with recommended contact lens and storage case hygiene procedures, notably the rinsing and storing of contact lenses in tap water, are recognized risk factors for acquiring the infection. ^5,6 The incidence of Acanthamoeba keratitis varies by geographic location and estimates vary from one to two cases per million contact lens wearers in the United States, up to 17–21 cases per million in the United Kingdom. ^3,7 This compares with approximately 40 per 100,000 microbial keratitis cases of mainly bacterial origin. ^8–10 A significant rise in the number of Acanthamoeba keratitis cases in the United States and other countries was reported in 2007. According to the US Centers for Disease Control and Prevention, those infected were 16.9 times (odds ratio) more likely to have used a specific multipurpose contact lens disinfectant solution, which was then globally recalled. ^6,11 However, since the recall the number of Acanthamoeba keratitis cases in the United States has remained above the prerecall estimates, indicating that the incidence of the infection may be significantly greater than previously indicated (Gupta S, et al. IOVS 2009;50:ARVO E-Abstract 3114; and Ref. 7). 

Acanthamoeba is characterized by a life cycle of a feeding and replicating trophozoite stage that, in response to adverse conditions, can transform into a resistant double-walled cyst stage. Various species have been reported to cause Acanthamoeba keratitis, with A. castellanii , A. polyphaga , and A. hatchetti being most common. ¹² More recent studies comparing 18S rRNA gene sequences of Acanthamoeba species and strains have greatly aided the taxonomic study of the organism and, currently, 17 lineages have been identified (T1–T17), with the majority causing keratitis occurring in T4. ^13,14 In the transformation process, the trophozoite rounds and produces a thin inner cyst (endocyst) wall composed mainly of cellulose. Later, the immature cyst forms a thicker outer (ectocyst) wall of protein to give the mature and resistant cyst stage. ^15,16 Several plugged openings, termed ostioles or opercula, span both the walls through which the trophozoite exits during excystment. ^15,17 The resistance of mature Acanthamoeba cysts to extremes of temperature, desiccation, and most disinfectant solutions at a working concentration accounts for the organism presence in virtually all soil and aquatic habitats, presenting a constant challenge to the contact lens wearer. ^2,5,18  

Accordingly, good hygiene practices, including the correct use of disinfection systems and care of the lens storage case, are fundamental to safe contact lens wear. Multipurpose disinfecting solutions (MPDSs) are most commonly used for soft contact lens care and represent a single solution composed of biocidal preservatives, buffer system, and other agents to aid lens comfort and cleaning for the rinsing, disinfection, and storage of lenses. ^19,20 Alternative methods use 3% hydrogen peroxide with a neutralization step, typically a platinum-coated disc inside the storage case (one-step system) or following the addition of a separate catalase enzyme tablet (two-step system). ^21,22 As a requirement for a product to be marketed as a contact lens disinfectant, it must comply with the international standard ISO 14729, which requires the demonstration of antimicrobial activity against specified bacteria, yeast, and mold within the manufacturer's recommended disinfection time. ²³ However, the standard does not require for testing against Acanthamoeba. 

Regardless, several studies have compared the efficacy of contact lens care solutions against Acanthamoeba. These have used both qualitative and quantitative assays, a variety of test strains, trophozoite culture, and cyst production methods. ^{12,21,24–32} This has resulted in conflicting findings on the amoebacidal efficacy of contact lens care solutions, particularly the cyst stage, with little information regarding factors that may affect the assay method and findings. ²⁶ In this present study, we describe the development of standardized methods for determining the efficacy of contact lens disinfectant systems against Acanthamoeba trophozoites and cysts and show how selection of test strain and method of cyst production can significantly affect the findings. 

Materials and Methods

Acanthamoeba Strains

A. castellanii (ATCC 50370) and A. polyphaga (ATCC 30461) were obtained from the American Type Culture Collection (ATCC, Manassas, VA). A. hatchetti (CDC: V573) was obtained from Govinda Visvesvara, Centers for Disease Control and Prevention, Atlanta, GA. ¹² All strains were isolated from Acanthamoeba keratitis patients and were of the T4 genotype, but showed some degree of genetic variation within the sequence. ^12,13,33  

Trophozoite Culture

Trophozoites were adapted and maintained at 28–30°C, in air, in tissue culture flasks (BD Falcon; BD Biosciences, Franklin Lakes, NJ) using a semidefined axenic medium (Ac#6) comprising: Biosate Peptone (BBL; Becton, Dickinson and Company, Oxford, UK) 20.0 g; glucose 5 g; KH₂PO₄ 0.3 g; vitamin B₁₂ 10 μg; l-methionine 15 mg per liter of deionized water. The pH was adjusted to 6.5 to 6.6 with 1 M NaOH before autoclaving at 121°C for 15 minutes and storage at room temperature for use within 2 months. Penicillin and streptomycin (Sigma-Aldrich, Dorset, UK) were added to a final concentration of 40 U/mL and 40 μg/mL, respectively, before use and the medium stored at 2 to 8°C for use within 14 days. 

A. castellanii (50370) trophozoites were also grown in peptone–yeast extract–glucose (PYG) medium (ATCC 712 medium). Aliquots of trophozoites were cryopreserved so that no strain was passaged >12 times before use in testing. ³⁴  

Trophozoite and Cyst Preparation

Late log-phase growing trophozoites were seeded into large (150 cm²) tissue culture flasks (catalog no. 355001; BD Falcon; BD Biosciences) at a density of 1 × 10⁵/mL in 50 mL of Ac#6 medium and incubated for 24 hours at 28 ± 2°C to yield approximately 1 × 10⁷ trophozoites for immediate testing. The flasks were vigorously shaken to dislodge adhering trophozoites, which were then pelleted and washed three times, with quarter-strength Ringer's solution, pH 7.0 (Oxoid, Basingstoke, UK) by centrifugation at 500g for 5 minutes. 

Two methods for cyst production were compared using Neff's encystment medium (NEM) and starvation on nonnutrient agar (NNA). Trophozoites were seeded into large tissue culture flasks at a density of 1 × 10⁵/mL in 50 mL of Ac#6 medium and incubated for 48 hours at 28 ± 2°C, to yield approximately 1 to 2 × 10⁷ trophozoites. The trophozoites were harvested by centrifugation at 500g for 5 minutes and washed three times with quarter-strength Ringer's solution and the final pellet inoculated into 100 mL of NEM at a density of 3 to 5 × 10⁵/mL in 500 mL polystyrene storage bottles (catalog no. 430282; Corning Life Sciences, Durham, NC). ¹⁵ The cultures were incubated at 28 ± 2°C for 14 days with gentle shaking (100 rpm). For some experiments, the cysts were harvested for testing after only 7 days incubation in NEM. 

The maturity of the cysts was then confirmed by phase-contrast microscopy before harvesting (with swabbing of the bottle sides to remove adherent cysts) and washing three times with quarter-strength Ringer's solution by centrifugation at 1000g for 10 minutes. 

NEM was prepared from sterile stocks of reagents that were combined to the following final concentrations in nanopure water: KCl (0.5 M) 0.1 M, MgSO₄ · 6H₂O (1 M) 0.008 M, NaHCO₃ (1 M) 0.001 M, Tris (1 M) 0.02 M, CaCl₂ · 2H₂O (0.5 M) 0.0004 M and phenol red (sodium salt: 1.5% w/v in H₂O) 25 μL/L. The solution was adjusted to pH to 8.9 ± 0.2, sterilized by filtration through a disposable 0.22 μm cellulose acetate membrane filter unit (catalog no. 430769; Corning Life Sciences), and stored at room temperature for use within 1 month. The osmolarity of the encystment medium was 220 mOsm/kg (Advanced Instruments, Inc., Norwood, MA). 

In the NNA encystment method, trophozoites were cultured and harvested as described above for NEM cysts and the final pellet was resuspended in 1 mL quarter-strength Ringer's solution. The trophozoites were spread gently over the surface of five NNA plates composed of 2.5% nonnutrient agar (catalog no. 214530; Becton, Dickinson and Company) in quarter-strength Ringer's solution in 100 mm × 15 mm petri dishes at 25 mL per plate (catalog no. 25384‐302; VWR International, Brisbane, CA). The uninoculated plates were dried overnight at 35 to 37°C and stored at 2 to 8°C for up to 7 days before use. Inoculated plates were incubated in air at 28 to 30°C for 48 hours before being sealed in polythene bags for a further 5-day incubation. Encystment occurs rapidly through starvation on the plates. The mature cysts were harvested by flooding the plates with quarter-strength Ringer's solution and gently rubbing the agar surface with a sterile cotton-tipped swab. The cysts were then pooled and washed with quarter-strength Ringer's solution as described above. 

The cysts from both preparation methods were stored at 4 to 8°C for testing within 7 days or, in some studies, 14 days. Immediately before testing, the cysts were passed through a 40-μm pore-sized cell strainer to eliminate clumps (catalog no. 352340; BD Falcon; BD Biosciences). The effect of culture in Ac#6 or PYG media on the susceptibility of A. castellanii (ATCC 50370) trophozoites to the test solutions was also investigated. 

Trophozoite and cyst counts were made using disposable Fuchs-Rosenthal hemocytometer chambers (INCYTO C-Chip Disposable Hemacytometers; VWR International). The appearance and maturity of the cyst preparations were observed microscopically (Olympus CKX41; Olympus Medical Equipment, Mokena, IL) with ×10, ×20, and ×40 phase-contrast objectives. Images were recorded using a digital camera and capture software (SPOT Insight; SPOT Imaging Solutions, Diagnostic Instruments, Inc., Sterling Heights, MI). Both NEM and NNA preparations contained >95% mature cysts and was a criterion for their use in the biocidal assays. 

Electron Microscopy

Mature cysts of A. castellanii (50370) prepared by NEM or NNA methods were examined by transmission electron microscopy (TEM). The cysts were harvested by centrifugation at 1000g for 10 minutes and the pellet was fixed with 0.05 M HEPES buffered 2.5% glutaraldehyde followed by postfixation with buffered 1% osmium tetroxide. Washed samples were subsequently embedded in 2% agar, en bloc stained with 2% uranyl acetate/30% ethanol, and then dehydrated through an ethanol series. The samples were gradually infiltrated with epoxy resin before final embedding and polymerization. Thin-cut sections, approximately 80 nm thick, were collected on copper mesh grids, counterstained with Reynolds' lead citrate, and viewed using a commercial microscope (Tecnai 12 BioTWIN; FEI Company, Hillsboro, OR) at 100 kV acceleration voltage and images were captured using a charge-coupled device camera (Orius SC1000; Gatan, Inc., Pleasanton, CA). 

Test Solutions

Four multipurpose disinfectant solutions (MPDSs) and a one-step hydrogen peroxide system were studied. The products and main biocidal components were: MPDS-1 (RevitaLens OcuTec, polyquaternium-1 [PQ-1] 0.0003% + alexidine dihydrochloride 0.00016%; Abbott Medical Optics Inc., Santa Ana, CA), MPDS-2 (OPTI-FREE RepleniSH, PQ-1 0.001% + myristamidopropyl dimethylamine [MAPD] 0.0005%; Alcon, Inc., Hünenberg, Switzerland), MPDS-3 (Biotrue, PQ-1 0.0001% and polyhexamethylene biguanide [PHMB], 0.00013%; Bausch & Lomb, Rochester, NY), and MPDS-4 (renu fresh, PHMB 0.0001%; Bausch & Lomb). PER-1 (Clear Care; Alcon, Inc.) comprised a 3% hydrogen peroxide solution with a storage case containing a platinum disk to obtain neutralization of the peroxide during the manufacturer's recommended 6-hour disinfection time. The PER-1 system was also tested without the presence of the neutralizing disc (PER-2). All solutions were tested from previously unopened bottles that were within at least 6 months of the manufacturers' expiry date. 

Assay Methods

A most probable number (MPN) approach, using trimmed Spearman-Karber computations, was used to determine the level of Acanthamoeba survival or kill following exposure to the test solutions. ³⁵ Quarter-strength Ringer's was used as a negative control in all experiments. 

For the MPDS, the bottles were thoroughly mixed by inversion six times and 5.0 g of solution or quarter-strength Ringer's solution was added to 14 mL round-bottom polypropylene tubes (catalog no. 352059; BD Falcon; BD Biosciences). The solutions were weighed into the test tubes from the manufacturers' bottles, rather than pipetted, to avoid possible loss of biocidal components through binding to the plastics of the pipettes. Fifty μL of 5 × 10⁶/mL trophozoites or cysts were added to give a final concentration of 5 × 10⁴/mL in the test or control solutions. 

The tubes were vortexed briefly and incubated at 22°C in the dark. At timed intervals of 0, 6, and 24 hours, the tubes were vortexed and, in quadruplicate, 20 μL aliquots were removed and added to 180 μL 0.1% Tween 80 and 0.35% lecithin in quarter-strength Ringer's solution (sterilized by autoclaving at 121°C for 15 minutes and stored at room temperature for up to 3 months) in a 96-well flat-bottom microtiter plate with lid (catalog no. 353075; BD Falcon; BD Biosciences) and left for 5 minutes. Serial, 10-fold, dilutions of 20 μL in 180 μL quarter-strength Ringer's solution were then made across the rows of a microtiter plate (catalog no. 353075; BD Falcon; BD Biosciences) in quadruplicate. Fifty μL Escherichia coli (ATCC 8739) at an OD₆₀₀ of 0.7 to 0.9 in quarter-strength Ringer's solution was then added to each well and the plate sealed and incubated at 32°C. The filling of the plates and subsequent performing of serial dilutions and feeding with E. coli was greatly aided by the use of multichannel and automated dispensing pipettes (e.g., Eppendorf, Mount Laurel, NJ). 

The E. coli stock was prepared from growing the bacterium in 100 mL trypticase soy broth (BBL, catalog no. 211768; Becton, Dickinson and Company) in 500 mL polystyrene storage bottles (catalog no. 430282; Corning Life Sciences) overnight at 30°C with shaking (100 rpm). The bacteria were harvested and washed once in quarter-strength Ringer's solution by centrifugation at 3000g for 30 minutes. The final pellet was resuspended in 5 mL quarter-strength Ringer's solution and stored aliquotted at 2 to 8°C for use within 14 days. 

The plate wells were observed microscopically (×10 and ×20 phase-contrast objectives) after 3, 5, 7, 10, and 14 days incubation and scored for the presence of amoebal growth. Here, where survival has occurred, the trophozoites feed and replicate on the E. coli bacteria and eventually form a near monolayer in the wells. For cysts, the presence of live bacteria stimulates excystment and the emergent trophozoites feed and replicate on the bacteria to also give rise to confluence in the wells. Careful microscopic examination can detect the presence of live motile trophozoites producing clearings in the E. coli lawn, enabling positive wells to be recorded before confluent growth in the wells is achieved. 

For the one-step peroxide system (PER-1), the storage case, with platinum-neutralizing disc, was filled with 10 mL of the manufacturer's hydrogen peroxide solution and immediately challenged with 100 μL 1 × 10⁷/mL trophozoites or cysts. At timed intervals of 0 and 6 hours, aliquots were removed and processed as described above, with a neutralizer composed of sodium thiosulfate, pentahydrate (6 g/L), Tween 80 (0.5 g/L), and 400 to 500 U/mL catalase from bovine liver (catalog no. C9322; Sigma-Aldrich) in quarter-strength Ringer's solution. The neutralizer was filter sterilized and stored at 2 to 8°C for use within 14 days. Experiments were also conducted using the hydrogen peroxide solution (Per-2) in polypropylene tubes without the presence of the platinum-neutralizing disc at 0-, 6-, and 24-hour exposure. 

The effectiveness of the biocidal neutralizers used for the MPDS and Per-1 solutions was determined by adding 0.5 mL test to 4.5 mL appropriate neutralizer, vortexing and leaving for 5 minutes before challenging with Acanthamoeba and proceeding to determine viability after 0, 6, and 24 hours by the MPN approach, as described above for assaying the test solutions. 

By scoring the presence or absence of Acanthamoeba growth in the wells for a given dilution, the MPN of surviving trophozoites or cysts can be calculated according to the formula X ₀ − (d/2) + d(∑ r_i /n_i ), where: X ₀ = log₁₀ of the reciprocal of the lowest dilution at which all test inocula are positive; d = log₁₀ of the dilution factor (i.e., the difference between the log dilution intervals); n_i is the number of test inocula used at each individual dilution; r_i is the number of positive test inocula (out of n_i ); ∑ (r_i /n_i ) = ∑ (P) = sum of the proportion of positive tests beginning at the lowest dilution showing a 100% positive result. ³⁵ At a given time point the test or control solution was sampled in quadruplicate and each diluted from 10⁻¹ to 10⁻⁴, using a total of 16 wells. If all wells subsequently scored positive (16/16), then the MPN of surviving organisms would be 31,623 or 4.5 log₁₀. If killing has occurred, then fewer wells will be positive and, for example, if there were no growth in all the 10⁻⁴ wells and the same in three of the 10⁻³ (i.e., 11/16 wells total positive) the value would be 1778 or 3.25 log₁₀. 

In all assays, the T ₀ count was derived from the quarter-strength Ringer's control values. The reduction in viable cysts was plotted as delta logs with SEM for each time point according to the formula log₁₀ T ₀ − log₁₀ T_n , where T ₀ is the initial viable count at the start of the experiment, as determined from the quarter-strength Ringer's control included in all experiments, and T_n is the viable count at an experimental time point. All experiments were performed in triplicate and repeated on at least three separate occasions. Statistical analysis was performed using one-way ANOVA from experimental mean values and SEM, with Tukey's posttest using a commercial analysis software program (GraphPad InStat version 3.06 for Windows; GraphPad Software, Inc., La Jolla, CA). 

Results

The efficacy of the test solutions against the trophozoites of the three Acanthamoeba spp. is shown in Table 1. All the solutions gave ≥3 log₁₀ kill for all the strains by the first time point of 6 hours (Table 1). No significant difference was found in the biocidal results at 6 or 24 hours for trophozoites of A. castellanii cultured in PYG or Ac#6 (P > 0.05). 

Table 1

 

Efficacy of Test Solutions Against the Trophozoites of Acanthamoeba spp.

Solution	Average log10 Trophozoite Kill and SEM
	A. castellanii , ATCC 50370 (SEM)		A. polyphaga , ATCC 30461 (SEM)		A. hatchetti , CDC: V573 (SEM)
	6 hour	24 hour	6 hour	24 hour	6 hour	24 hour
MPDS-1	3.6 (0.0)	3.6 (0.0)	3.7 (0.0)	3.7 (0.0)	3.6 (0.0)	3.6 (0.0)
MPDS-2	3.2 (0.5)	3.0 (0.0)	3.9 (0.0)	3.9 (0.0)	3.4 (0.0)	3.4 (0.0)
MPDS-3	3.0 (0.25)	3.7 (0.0)	3.9 (0.0)	3.9 (0.0)	3.6 (0.0)	3.6 (0.0)
MPDS-4	3.6 (0.0)	3.6 (0.0)	3.9 (0.0)	3.9 (0.0)	3.6 (0.0)	3.6 (0.0)
Per-1	3.5 (0.1)	NT	3.9 (0.0)	NT	3.7 (0.0)	NT
Per-2	3.7 (0.0)	3.7 (0.0)	3.9 (0.0)	3.9 (0.0)	3.7 (0.0)	3.7 (0.0)

 NT, not tested.

Greater resistance was found against cysts, with results for the MPDS varying by species and method of cyst production, as summarized in Table 2 and Figures 1 through 3. Typically, cysts prepared by the NNA method were significantly more resistant than those of the NEM type (Figs. 1–3). For example, at the 6-hour sample time MPDS-3 gave kill values for NEM-prepared cysts of A. castellanii , A. polyphaga , and A. hatchetti of 1.6, 2.9, and 2.3 log₁₀ compared with 0.7, 0.8, and 0 log₁₀ for NNA cysts, respectively (Table 2 and Figs. 1–3), which was all highly significant (P < 0.01–0.001). MPDS-4 gave NEM cyst kill values of 2.8, 3.5, and 3.4 log₁₀ for the species compared with 0.7, 0.4, and 0.1 log₁₀ for NNA-derived cysts, respectively (P < 0.001). MPDS-2 did not show any significant cysticidal activity, even after 24-hour exposure (≤0.8 log₁₀) regardless of species or cyst type (Table 2) and the findings were not significant (P > 0.05). MPDS-1 showed good activity against A. castellanii and A. hatchetti, regardless of cyst type at 6 hours (2.7–3.7 log₁₀ kill, P > 0.05). However, A. polyphaga was more resistant, giving NEM and NNA cyst kill values of 1.3 and 0.6 log₁₀, respectively (Table 2 and Figs. 1–3), which was statistically significant (P < 0.05). 

Figure 1

 

Efficacy of test solutions against A. castellanii (ATCC 50370) 7-day and 14-day NEM- and NNA-produced cysts. *7-day NEM-prepared cysts were statistically (P < 0.001) more susceptible to disinfection than those formed from a 14-day incubation.

Figure 2

 

Efficacy of test solutions against A. polyphaga (ATCC 30461) 7-day and 14-day NEM- and NNA-produced cysts.

Figure 3

 

Efficacy of test solutions against A. hatchetti (CDC:V573) 7 day and 14 day NEM and NNA produced cysts.

Table 2

 

Efficacy of Test Solutions Against the Cyst of Acanthamoeba spp. Prepared by Two Different Methods

Solution	Average log10 Cyst Kill and SEM
	A. castellanii , ATCC 50370 (SEM)			A. polyphaga , ATCC 30461 (SEM)			A. hatchetti, CDC: V573 (SEM)
	Cyst Type	6 hours	24 hours	Cyst Type	6 hours	24 hours	Cyst Type	6 hours	24 hours
MPDS-1	NEM	3.4 (0.0)	3.4 (0.0)	NEM	1.3 (0.1)	2.6 (0.2)	NEM	3.7 (0.0)	3.7 (0.0)
	NNA	2.7 (0.2)	3.7 (0.0)	NNA	0.6 (0.2)	2.7 (0.3)	NNA	3.6 (0.1)	3.7 (0.0)
MPDS-2	NEM	0.2 (0.2)	0.3 (0.1)	NEM	0.1 (0.3)	0.6 (0.2)	NEM	0.3 (0.3)	0.4 (0.1)
	NNA	0.8 (0.0)	0.6 (0.1)	NNA	0.1 (0.2)	0.6 (0.1)	NNA	0 (0.1)	0.2 (0.3)
MPDS-3	NEM	1.6 (0.2)	3.1 (0.1)	NEM	2.9 (0.2)	3.5 (0.0)	NEM	2.3 (0.2)	3.2 (0.3)
	NNA	0.7 (0.1)	1.7 (0.2)	NNA	0.8 (0.0)	1.2 (0.1)	NNA	0 (0.5)	1.5 (0.3)
MPDS-4	NEM	2.8 (0.2)	3.2 (0.0)	NEM	3.5 (0.0)	3.5 (0.0)	NEM	3.4 (0.0)	3.4 (0.0)
	NNA	0.7 (0.1)	2.7 (0.3)	NNA	0.4 (0.2)	1.3 (0.3)	NNA	0.1 (0.3)	1.6 (0.2)
Per-1	NEM	1.9 (0.2)	NT	NEM	1.0 (0.2)	NT	NEM	1.4 (0.3)	NT
	NNA	0.8 (0.2)	NT	NNA	1.1 (0.2)	NT	NNA	1.0 (0.2)	NT
Per-2	NEM	3.6 (0.0)	3.6 (0.0)	NEM	3.5 (0.0)	3.5 (0.0)	NEM	3.7 (0.0)	3.7 (0.0)
	NNA	3.8 (0.0)	3.8 (0.0)	NNA	3.5 (0.0)	3.5 (0.0)	NNA	3.5 (0.0)	3.5 (0.0)

The one-step hydrogen peroxide system (PER-1) showed less significant variation in disinfection based on cyst type. Here, 6-hour kill values for A. castellanii , A. polyphaga , and A. hatchetti were 1.9, 1.0, and 1.4 log₁₀, with NEM cysts compared with 0.8, 1.1, and 1.0 log₁₀ for NNA cysts (Table 2 and Figs. 1–3), which was significant for only the A. castellanii findings (P < 0.05). When the hydrogen peroxide solution was tested without the neutralizing disc (PER-2), ≥3.5 log₁₀ kill after 6 hours was achieved for all species regardless of cyst type (Table 2 and Figs. 1–3). 

Exposure of the cysts to the test solutions for up to 24 hours resulted in increased cyst kill with those solutions not achieving total kill at 6 hours (Table 2). For example, with NNA cysts, MPDS-3 and MPDS-4 gave 1.2 to 2.7 log₁₀ kill at 24 hours compared with 0 to 0.8 log₁₀ at 6 hours (Table 2). With MPDS-1 and A. polyphaga , the kill for NNA cysts increased from 0.6 log₁₀ to 2.7 log₁₀. With NEM cysts, increased kill was observed with MPDS-1 and A. polyphaga (1.3–2.6 log₁₀ kill) and with MPDS-3 for all species (1.6, 2.9, and 2.3 log₁₀ at 6 hours and 3.1, 3.5, and 3.2 log₁₀ by 24 hours). MPDS-2 showed no cysticidal activity after 24-hour exposure, regardless of cyst type (Table 2). 

No loss in trophozoite or cyst viability over 24-hour exposure was detected in the neutralizer control experiments, indicating that these were appropriate for the purpose and nontoxic to the amoebae (results not shown). 

For the cysts but not the trophozoites, positive wells continued to be detected between 7 and 14 days incubation and affected the results in many cases (results not shown). This confirms the need to incubate and read plates up to 14 days before reporting results. Although occasional wells continued to turn positive up to 21 days incubation, this was not sufficient to affect significantly the results obtained at 14 days (results not shown). 

NEM cysts harvested for testing after 7 days incubation in the encystment medium were more susceptible to 6-hour disinfection than those left for 14 days (Table 3 and Figs. 1–3). This was statistically significant with A. castellanii and MPDS-3, which gave 3.2 log₁₀ kill for 7-day cysts compared with 1.6 log₁₀ for 14-day cysts (P < 0.01) and also MPDS-4 with 3.4 log₁₀ vs. 2.8 log₁₀ (P < 0.01). For the other species, although greater kill was found for the 7-day cysts, the results were not significant (P > 0.05). 

Table 3

 

Efficacy of Test Solutions Against 7-Day vs. 14-Day NEM-Derived Cysts of Acanthamoeba spp.

Solution	Average log Kill for NEM Cysts 7 vs. 14 Day (6-Hour Exposure)
	A. castellanii, ATCC 50370 (SEM)		A. polyphaga, ATCC 30461 (SEM)		A. hatchetti, CDC: V573 (SEM)
	7 day	14 day	7 day	14 day	7 day	14 day
MPDS-1	3.4 (0.0)	3.4 (0.0)	1.6 (0.4)	1.3 (0.1)	3.7 (0.0)	3.7 (0.0)
P value	>0.05		>0.05		>0.05
MPDS-2	0.3 (0.2)	0.2 (0.2)	0.5 (0.0)	0.1 (0.3)	0.3 (0.0)	0.3 (0.3)
P value	>0.05		>0.05		>0.05
MPDS-3	3.2 (0.3)	1.6 (0.2)	3.5 (0.0)	2.9 (0.2)	2.7 (0.2)	2.3 (0.2)
P value	<0.01		>0.05		>0.05
MPDS-4	3.4 (0.0)	2.8 (0.2)	3.5 (0.0)	3.5 (0.0)	3.7 (0.0)	3.4 (0.0)
P value	<0.01		>0.05		>0.05
Per-1	2.3 (0.1)	1.9 (0.2)	1.4 (0.1)	1.0 (0.2)	1.4 (0.1)	1.4 (0.3)
P value	>0.05		>0.05		>0.05
Per-2	NT	3.6 (0.0)	NT	3.5 (0.0)	NT	3.7 (0.0)

The effect of storage at 2 to 8°C on the susceptibility of A. castellanii and A. polyphaga NNA cysts to MPDS-1 gave 6-hour kill values of 2.5 log₁₀ (SEM 0.2) and 0.9 log₁₀ (SEM 0.3), respectively, after 7 days storage and 3.7 log₁₀ (SEM 0.3) and 0.5 log₁₀ (SEM 0.3) after 14 days. Although this suggests that the A. castellanii cysts may become more susceptible to disinfection on storage, the finding was not statistically significant (P > 0.05). 

The microscopic appearance of the NEM and NNA cysts of A. castellanii by phase contrast and TEM is shown in Figure 4. Under light microscopy, the NEM cysts appeared mature with the presence of an outer (ecto) and inner (endo) cyst wall (Fig. 4a). However, with the NNA cysts, the ectocyst wall was thicker and more associated with the endocyst wall, giving the characteristic angular appearance of the cyst (Fig. 4b). Comparison using TEM also revealed the greater thickness of the ectocyst in NEM compared with NNA cysts (Figs. 4c, 4d). 

Figure 4

 

Appearance of A. castellanii (ATCC 50370) cysts under phase contract (a, b) and transmission electron microscopy (c, d). NEM cysts (a, c). NNA cysts (b, d).

Discussion

Acanthamoeba keratitis is being reported with increasing frequency with contact lens wear being a factor in 90% of cases. ^3,8,11 Although poor lens and storage case hygiene practices are known to increase significantly the risk of infection, lack of amoebacidal activity of care solutions has also been suggested as a possible compounding factor. ^6,11 Unlike the requirements for bacteria and fungi, no standard method exists for conducting efficacy testing of contact lens disinfectant solutions or systems against Acanthamoeba. ²³ Where such studies have been reported, a range of test strains, culture and cyst preparation methods, and assay techniques have been used, often with conflicting findings. ^{12,21,24–32,36–38} In this study, we have developed a reliable and reproducible method for quantifiably determining the efficacy of contact lens disinfectant solutions against Acanthamoeba trophozoite and cyst stage and investigated the effect of various test parameters on the findings. 

Acanthamoeba trophozoites can be readily cultured on plain nonnutrient agar (NNA) seeded with a lawn of Gram-negative bacteria such as E. coli or E. aerogenes (monoxenic culture) or in semidefined, bacteria-free, broth media (axenic culture) as used here. ^2,18,39 Axenic cultivation enables the controlled growth of trophozoites to large numbers required to perform quantifiable, time kill assays. ^21,25 Although no comparison was made regarding the age of the cultures and their susceptibility to disinfection, care was taken to ensure that the trophozoites of the strains tested were fresh and still in a state of replication in the culture flasks. No difference in disinfectant susceptibility was observed between trophozoites of strains grown in PYG or Ac#6 medium, although we find the latter to be better for adapting clinical strains to axenic culture, giving higher cell yields, and somewhat easier to prepare (personal observations). In contrast, culture using NNA-bacteria plates results in low yields of trophozoites, which are prone to spontaneous encystment. The presence of resistant cysts in the trophozoite inoculum may give rise to erroneous results. Furthermore, even with repeated washing prior to testing, the trophozoites remain cocontaminated with the host bacterium. This can result in >10⁶ bacteria in the challenge inoculum (personal observation), which may interfere with the disinfecting ability of the test solution. 

As with trophozoite testing, large numbers of cysts are required to conduct quantifiable and time-dependent disinfection assays. Typically, these are obtained from axenically grown trophozoites, which are induced to encyst using a variety of methods. Most commonly, this is achieved using the chemically defined encystment medium developed by Neff and colleagues, ¹⁵ starvation through prolonged culture in axenic broth media, ^29,30 addition of Mg²⁺ cations to axenic broth media, ³⁷ or on nonnutrient agar (NNA), ²⁵ sometimes supplemented with taurine and magnesium. ¹⁶ Cysts can also be produced by prolonged incubation of trophozoites on NNA plates seeded with E. coli or E. aerogenes , ^12,26 although, as with monoxenically grown trophozoites, such cyst preparations cannot be freed of the host bacterium. Of these, the NEM method has been most widely used to study Acanthamoeba transformation and analysis of the cysts formed. ^15,16 A requisite for any encystment method is that the cysts produced should be in the mature and most resistant form, because immature or developing stages are more susceptible to disinfection. ^32,38  

The direct comparison of findings from this study with those reported previously for Acanthamoeba biocidal testing may not be relevant, given the differences in assay procedures, strains, test solutions, and method of trophozoite culture and cyst production. However, the majority of studies have used axenically grown trophozoites and, where the same solution has been tested, found similar results against this stage of the organism. For example, MPDS-4 and a solution similar to MPDS-2 (based on PQ-1 and MAPD) were reported to give >2 to ≥3 log₁₀ kill with trophozoites of A. castellanii and A. polyphaga strains after a 6-hour exposure, which is comparable to the findings of this study. ^20,25,29,31 Similarly, the one-step hydrogen peroxide system (PER-1) gave >3 log₁₀ kill of trophozoites with strains of A. castellanii (including 50370) and A. polyphaga . ^21,29,31 In another study no difference in the degree of disinfection by PHMB-based MPDS was observed with two A. castellanii strains (including 50370) cultured monoxenically or axenically. ²⁶ However, the monoxenic trophozoite cultures were reported to be more resistant to MPDS based on PQ-1 + MAPD. ²⁶ Because no such differences were noted for the PHMB MPDS, the greater resistance of the monoxenically grown trophozoites for the solution might have been caused by the neutralizing presence of bacteria in the challenge inoculum. 

In this study, cysts of three Acanthamoeba species produced with NEM and NNA (without bacteria) were compared. Overall, cysts prepared using the NNA method were significantly more resistant to disinfection. Although mature cysts were produced by the NEM method, as judged by light and electron microscopy, testing showed them to be significantly more susceptible to disinfection by PHMB-based MPDS. No cysticidal activity was found with MPDS-2 and only slight differences were found with PER-1, regardless of the encyst method, and indicates that the NEM cysts were not of an inherently weakened or immature state, in that this may have resulted in some degree of kill occurring. ^32,38 Differences in susceptibility were observed between the test strains, with A. castellanii and A. hatchetti being more susceptible to MPDS-1 than A. polyphaga , but the latter showing greater kill by MPDS-3 for both encystment methods. Of the MPDS studied here, only MPDS-1 showed significant activity, giving a 2.7 to 3.6 log₁₀ kill for A. castellanii and A. hatchetti but <1 log₁₀ with A. polyphaga after a 6-hour exposure with NNA cysts. The other MPDS showed <1 log₁₀ kill for all species at this time point and indicates that the alexidine-based MPDS may have greater cysticidal activity than that of those based on PHMB or MAPD, at least for A. castellanii and A. hatchetti strains tested. With PER-1, the results were similar to those found with the NEM cysts, giving an approximately 1 log₁₀ kill after 6 hours. NEM was originally developed to study differentiation in an environmental isolate of A. castellanii (Neff strain, ATCC 30010) and it is possible that other species and strains of the organism behave differently in their encystment response to the medium and produce cysts that are inherently more susceptible to disinfection, particularly with PHMB-based MPDS. ^15,40  

Overall, the dual biocide MPDS showed good efficacy against the trophozoites of the Acanthamoeba spp. tested. However, greater variation was found against the NNA-derived cysts, with only MPDS-1 showing significant cysticidal efficacy at the manufacturer's recommended disinfection time of 6 hours but only against A. castellanii and A. hatchetti and not A. polyphaga . Good cysticidal activity against all species was found only with 3% hydrogen peroxide (PER-2) with an exposure time of 6 hours prior to neutralization. Where neutralization is initiated at the start of disinfection, as with the one-step system, a marked reduction in efficacy is obtained due to the rapid decomposition of the hydrogen peroxide. ²¹ Based on the observations of this study, 3% hydrogen peroxide for at least 6 hours prior to neutralization is the most efficacious against the cyst stage of Acanthamoeba spp. 

With respect to trophozoites, the direct comparison of published methods regarding the disinfectant efficacy against Acanthamoeba cysts is complicated by the variety of strains and methods used. Kobayashi and colleagues (2011) ³¹ compared 7-day and 14-day incubated NEM-derived cysts and found the latter to be significantly more resistant to disinfection against A. castellanii . With MPDS-4 they observed >3 log₁₀ kill of 7-day cysts but <1 log₁₀ with 14-day preparations with strain ATCC 50515. Although we also observed significant differences in resistance between such cysts, this was statistically significant only with A. castellanii and MPDS-3 and MPDS-4. The authors also found ≤1 log₁₀ kill with the one-step peroxide system using 14-day cysts, but >3 log₁₀ kill with 7-day cysts, compared with the 1–2 log₁₀ kill obtained here with 7- or 14-day preparations. The reasons for the differences between the studies are unclear, but our results were consistent on repeated testing with different batches of such cysts. 

In a study using NNA cysts and an MPN approach to assessing viability with A. castellanii CCAP 1501/1A (ATCC 30010), Beattie and colleagues (2003) ²⁵ found a 1 log₁₀ kill after 6 hours and 3 log₁₀ reduction at 24 hours within MPDS-4, which compares well with the 0.7 log₁₀ and 2.7 log₁₀ found here for ATCC 50370 of the species and these exposure times. For a solution similar to MPDS-2 (based on PQ-1 and MAPD) they obtained ≤1 log₁₀ kill after 6- or 24-hour exposure, which is also comparable to our findings. ²⁵ In our study, the concentration of agar was increased to 2.5% in quarter-strength Ringer's solution and mature cysts were formed within 7 days. The higher agar concentration aids cyst harvesting and prevents trophozoites burrowing into the agar, which can occur with some Acanthamoeba strains (unpublished observations). 

Other studies have used cysts obtained from prolonged incubation of trophozoites on NNA- E. coli or NNA- E. aerogenes plates. In a qualitative method that determined the ability of a solution to give total kill of a fixed inoculum of 100 cysts in 1 mL solution, Johnston and colleagues (2009) ¹² found survival with three strains of Acanthamoeba, including A. hatchetti (CDC: V573) studied here, after 6 or 24 hours in MPDS-4 and MPDS-2. Trophozoites were not tested, but the cysticidal findings with these solutions are similar to those found in this study. Although such an approach to Acanthamoeba testing is less complex than that required for MPN-based assays, it does not enable the kinetics of killing to be determined. Furthermore, a single surviving organism in the fixed inoculum would give rise to a positive result and failure of the test. 

Shoff and Eydelman (2012) ²⁶ compared the effect of encystment method on disinfection using both monoxenic (NNA- E. aerogenes ) and axenic (PYG) trophozoite cultures of two strains of A. castellanii (including ATCC 50370) using an MPN approach. Significant differences in the level of kill were found depending on the encystment method, with NEM cysts being more susceptible to disinfection than those from NNA or NNA- E. aerogenes cultures using axenically grown trophozoites but not when monoxenic cultures were used. For the MPDS studied, the NNA and NNA- E. aerogenes derived cysts showed similar resistance with <1 log₁₀ kill after 6-hour exposure, which is in agreement with the findings of our study using NNA cysts from axenic trophozoite cultures. ²⁶ As noted for monoxenically cultured trophozoites, the presence of bacteria from NNA-bacteria–derived cysts may interfere with the biocidal capability of the test solution. Therefore, axenic NNA cyst preparations should be preferred for such testing and these can be readily obtained using the methods described herein. Light and electron microscopy demonstrated that such cysts appeared more typical of the group II morphology for the species and showed a thicker outer cyst wall compared with that produced using NEM. ⁴¹  

It has been reported that Acanthamoeba can become attenuated in their virulence and encystment ability on continuous axenic culture. ^42–44 In one study it was found that Acanthamoeba strains grown on Hep-2 cells, rather than PYG medium, prior to encystment in NEM, were more resistant to physical and chemical disinfection, which was attributed to the presence of a thicker cyst wall with higher cellulose content. ⁴² In another study, cysts from an Acanthamoeba keratitis strain that had been in continuous culture for several years were significantly more susceptible to disinfection than the parent culture, which had been axenized and cryopreserved shortly after isolation. ²⁸ In our study, the Acanthamoeba test strains were cryopreserved in batches and passaged no more than 12 times in the preparation for testing. However, the culture history of the strains prior to receipt in the laboratory was unknown. This potential variable is addressed in the standard protocol for assessing the efficacies of contact lens disinfectant solutions against bacteria and fungi, with instruction that strains be passaged no more than five times from the original culture prior to testing. ²³ Accordingly, the effect of passage number on the biocidal susceptibility of Acanthamoeba should be considered when conducting such testing. The development of a reliable and reproducible reference method for conducting Acanthamoeba biocidal testing will enable further evaluation of such potential experimental variables and it would be of interest to investigate whether strains should be passaged through Hep-2 cells prior to axenization and use in disinfection assays. However, care should be taken when culturing and passaging Acanthamoeba strains because they can become infected with other microbes that may hinder their growth and encystment capabilities. ^45,46 It has also been reported that cysts stored at 4°C for several weeks were more susceptible to disinfection. ²⁸ Although not rigorously evaluated here, cysts of A. castellanii were found to be more susceptible after storage at 2 to 8°C for 14 days compared with those tested within 7 days and this may be a further important consideration when standardizing Acanthamoeba biocidal testing. It should also be emphasized that, for the cysticidal assays, the E. coli –containing microtiter plates require incubation for 14 days in that excystment continued after 7 days and any such early reporting of results would overestimate the biocidal capacity of the test solution. 

It has been reported that certain contact lens care solutions can induce Acanthamoeba trophozoites to transform into a cyst stage, notably through the presence of propylene glycol in the formulations. ^47–51 However, the cysts formed appear as an immature state and fail to develop, which has led to their being termed “pseudocysts.” ⁴⁷ Such cysts are significantly more susceptible to disinfection than the mature form and appear similar to the trophozoite stage in their sensitivity and unlikely to influence the assay findings. None of the solutions used in this study induced Acanthamoeba encystment (unpublished observations), although the phenomenon of solution-induced encystment should be considered, and examined for, when conducting trophozoite biocidal assays. 

In conclusion, given the wide variation in test methods and findings, standardization of Acanthamoeba contact lens disinfection testing is required as is currently defined for bacteria and fungi under ISO 14729. ²³ Here, various experimental variables were investigated in the development of a reproducible method for determining the efficacy of such systems against both the trophozoites and cysts of Acanthamoeba spp. Based on the observations of this study, it is recommended that axenic trophozoites and NNA cysts be tested using the MPN approach, which allows the quantified determination of Acanthamoeba viability to be assessed over time. Given the variation in susceptibility observed between the test strains, both A. castellanii (ATCC 50370) and A. polyphaga (ATCC 30461) should be evaluated; these are derived from keratitis cases and available from a recognized culture depository. It is hoped that the adoption of standardized methods for determining the efficacy of antiacanthamoebal compounds will aid in the development of safer contact lens care products and improved therapeutic agents. 

Acknowledgments

Parts of this study were undertaken when Simon Kilvington was employed at Abbott Medical Optics Inc. 

Disclosure: S. Kilvington, Abbott Medical Optics Inc. (E); A. Lam, Abbott Medical Optics Inc. (E)