Alkyl-carbon chain length of two distinct compounds and derivatives are key determinants of their anti-Acanthamoeba activities

The opportunistic pathogen, Acanthamoeba castellanii is the causative agent for the sight threatening infection Acanthamoeba keratitis (AK). It is commonly associated with contact lens wearers, and prevalence is increasing at an alarming rate due to an inadequate preventive strategy to protect the lens from this protist. This problem is compounded by the lack of an effective acanthamoebocide, particularly with cysticidal activity in the contact lens solutions. We have used cytotoxicity assays and a variety of biophysical approaches to show that two molecules with tails made of alkyl carbon, alkylphosphocholines (APCs) and quaternary ammonium compounds (QACs) had significant chain-length dependent efficacy against A. castellanii trophozoites, the latter producing death via permeabilization, and DNA complexing. QACs were more effective than APCs and had activity against cysts. Conversely, the QAC with 12 alkyl carbon chain, was non toxic, its presence increased A. castellanii trophozoites biomass and delayed encystation by 96 h. Interestingly, it was unable to induce excystation and increased trophozoite sensitivity to APC16. These results present a mono- and multi-inhibitor management strategy effective against trophozoites and cysts that may be useful for formulating into contact lense cleaning solutions and reducing AK incidence.

Examination of the genomic DNA (gDNA) of shrunk QAC18-treated A. castellanii trophozoites (18.75 µg/ ml) after 96 h, using DNA-agarose gel electrophoresis, the fragmented DNA profile consistent with death via apoptosis was not observed (data not shown). However, gDNA of A. castellanii incubated with QAC18 produced a dose-dependent compaction judged by a progressively increased absorbance at 260 nm (Fig. 4A). In contrast, the interaction of QAC12 with A. castellanii gDNA produced a dose-dependent decreased absorbance at 260 nm (Fig. 4B). In control experiments, neither QAC12 nor QAC18 had absorption at the concentrations used in the interaction assays. www.nature.com/scientificreports www.nature.com/scientificreports/  Table 3. Back transformation of QAC-treated cysts to trophozoites after drug wash-out. "+"Indicates cells reverting back to trophozoites after washout; and capable of oxidizing reduced alamar blue from blue to pink after 8 days "−" indicates cell that did not revert to trophozoites after wash out and unable to oxidise reduced alamar blue solution www.nature.com/scientificreports www.nature.com/scientificreports/ Long Apcs are toxic to A. castellanii trophozoites. The efficacy of the zwitterionic APCs with different alkyl-carbon chain lengths (12-16 carbons) tested against A. castellanii was also dependent on tail length, with APC12 and APC16 being the least and most potent with IC 50 s of 32.30 μg/ml and 7.60 μg/ml, respectively ( Table 4). The removal of APC16 from treated cells, and incubation in PG medium alone, showed that they were Figure 3. QAC12 increased A. castellanii cell density in vitro. Acanthamoeba trophozoites (10 5 cells/ml) treated with 150 µg/ml QAC12 doubly serially diluted to 0.015 µg/ml for 96 h at 25 o C. Cell viability estimated by the alamar blue assay and expressed as a percentage showed dose-dependent increase in mean absorbance, a measure for cell density, which was significant at 9.38 µg/ml and above, a 23% increase in cell density relative to control cells without QAC12 was observed at this concentration, with density peaking at 37.5 μg/ml and above (>32% increase). Average cell density of untreated cells is 94% (dashed line). Students t-test p < 0.01 as denoted by "*", data is mean viability ± SD; n = 6 independent experiments. respectively. Data is mean absorbance at 260 nm ± SD; n = 4 independent experiments performed in triplicates. Student's t-test showed significant difference between mean absorbance at 1.1, 1:10 and 1:20 relative to 1:0 at p < 0.01 for QAC18 (A) and QAC12 (B) respectively. www.nature.com/scientificreports www.nature.com/scientificreports/ able to oxidise resurzarin after 96 h when treated with up to 37.5 μg/ml, suggesting the compound to be amoebostatic or excystation was occuring (data not shown).

No of carbons
QAC12 delayed encystation. The fact that QAC12 promoted trophozoites growth suggested that it provided a favourable condition for trophozoites and delayed or stop encystation. The incubation of trophozoites in encystment medium, supplemented with 37.5 µg/ml of QAC12 delayed encystation by 96 h (Fig. 5). However, the same concentration of QAC12 was unable to induce excystation of matured cysts (data not shown).
QAC12 enhanced sensitivity of trophozoites to APC16 but not QAC18. As QAC12 delayed encystation of A. castellanii trophozoites, we postulated that they would be susceptible to the active QACs and APCs. The addition of 37.5 µg/ml or 18.75 µg/ml of QAC12 to QAC18 at concentrations starting from 150 μg/ml doubly diluted 12 times to 1.17 μg/ml produced estimated IC 50 s of 14.6 μg/ml and 11.7 μg/ml respectively, statistically significantly different from QAC18 alone in control experiments (IC 50 -6.9 μg/ml; p < 0.01, Fig. 6A).
In contrast, QAC16 combined with APC12 under the same conditions produced shifts from the dose response curves of APC16-treated cells (Fig. 6B). The estimated IC 50 s for APC16 with QAC12 at 37.5 μg/ml and 18.75 μg/ ml were 1.74 μg/ml and 2.03 μg/ml, statistically significantly different from the APC16 only in control experiments (IC 50 -7.6 μg/ml; p < 0.01, Fig. 6B). Interestingly, the removal of the inhibition of APC16 at concentrations of 3.12 μg/ml, 4.68 μg/ml, 6.25 μg/ml and 9.36 μg/ml with QAC12 (37.5 μg/ml and 18.75 μg/ml) after 96 h  . Cell viability estimated using the alamar blue assay showed the addition of QAC12 at both concentrations decreased and increased the potency for QAC18 and APC16 respectively relative to the monotherapies. Data are mean estimated viability ± SD; n = 4 independent experiments performed in triplicates. IC 50 s were statistically significant between both QAC18 and APC monotherapies and the combined treatment using QAC12 at both concentrations (p < 0.01, A,B).
QAcs have cysticidal activities against A. castellanii. Lastly, we observed that QAC16 and QAC18 and not QAC12 had activity against matured A. castellanii cysts (calcofluor white positive and SDS-resistant (0.5%, w/v; Fig. 7A-G) and their estimated IC 50 s were 19.00 ± 0.03 µg/ml and 15.00 ± 0.06 µg/ml respectively after 96 h ( Table 2). The cysticidal activity of QAC16 to QAC18 strongly correlated with (a) the alkyl-carbon chain lengths (Table 2) and the duration of cell exposure to the inhibitor (Fig. 7H). Like trophozoites, drug treatment resulted in leakage of proteins (Fig. 8A). Proteins concentrations in spent medium of QAC-treated cysts increased 6.6-fold and the extruded proteins are shown in Fig. 8B. Further, cyst viability estimated using a linked reversion-viability assay (that involved inhibitor removal, incubation of cells in PG medium alone for 7 days to  www.nature.com/scientificreports www.nature.com/scientificreports/ induce excystation and viability estimation via Alamar blue), with QAC18 and QAC16 at 37.5 μg/ml after 24 h and 48 h respectively and for QAC18 at 18.75 μg/ml after 48 h showed no resurgence or resazurin oxidation (Table 3).

Discussion
APCs and QACs are effective antimicrobials against fungi, bacteria, some protists and also have anti-tumour properties. In this study, we have confirmed their toxicity against Acanthamoeba trophozoites 34,35 , and showed that QACs were more effective than their APCs counterpart, perhaps due to the formation of larger micelles (QACs, 8.8 ± 0.8 nm; APCs, 5.8 ± 1.0 nm) 36 . Interestingly, the activities from both compound-types were strongly correlated with their alkyl-carbon chain lengths against trophozoites, a key indicator for micellar size and hydrophobicity. In addition, we noted that the duration of contact between the compound and the protist was important; longer alkyl-chain molecules required shorter contact times (<24 h) than their shorter counterpart. This is probably related to the nature of their interaction, with complementary alkyl carbon chains in the protist's plasma membrane. Metabolomics analysis have shown that fatty acids 20-30 carbon atoms long 37 in A. castellanii are abundant in its plasma membranes, presumably enough for long chain QACs to produce rapid solubilisation and death. This supports with similar study in the parasitic protist, Leishmania, where lipidomics analysis showed fatty acids with 18-25 alkyl-carbons long to be abundant 38 and coincidentally, the shorter APCs and QACs were more efficacious. Previously as 1.9 × 10 10 of QAC18 molecules have been shown to adhere to per unit surface area of the plasma membrane of A. castellanii trophozoites to cause leakage 39,40 . So, there is a possibility that in our study, micelles instead assembled on the surface of the cell, increased permeability and caused leakage of proteins and DNA in QAC-treated trophozoites.
Our results have also shown that it is possible that, the cationic but less so their zwitterionic APC counterpart, can induce rapid reversal of the net negative charge of the protist plasma membrane to positive, with detrimental effect as described for fungi 41,42 and Leishmania spp. [43][44][45] . Finally, the observed DNA compacting caused by QAC18 and not QAC12 suggested that QAC had an intracellular target and offered a third death mechanism for QACs in A. castellanii 41,42,[46][47][48] . Unexpectedly, death was not due to apoptosis. The data suggests that QAC18 may be complexing with compacted DNA, disrupting the normal cell cycle.
Compounds with activity against Acanthamoba cysts are limited. Our study has shown that the QACs used in this study have cysticydal activities and cause death via leaking, possibly emanating from conformational changes and weakening of the inter-fibre bonds of the cellulose cell wall. Evidence for this is indirect, and based on the loss of calcofluor white staining in QAC18-treated cysts. For example, the loss of the natural auto-fluorescent of carboxymethyl cellulose after QAC activity has been linked with conformational changes in cellulose 46 and as such, QACs are being used to alter the structural conformation of wood cellulose in paper manufacturing 49 .
Further, we observed that QAC12 promoted A. castellanii trophozoites growth, suggesting a role as an energy substrate, as its presence increased trophozoites biomass and delayed encystation by 96 h. Interstingly, the delayed encystation of trophozoites with QAC12 made them sensitive to APC16 but less so to QAC18; possibly allowing APC16 to exert its cytotoxicity without the induction of encystation. This is evidenced by the cytostatic to cytotoxic switch of APC16 without and with QAC12, respectively. At this stage, the interaction of these mixed surfactants with different charges to lower the surface interfacial tension of the molecule, their ease of forming micelles (CMCs), molar solubilisation ratio cannot be excluded 50,51 . More investigations are required to substantiate their involvement.
Finally, we have presented sufficient evidence to show QAC18 alone and QAC12 in combination with APC16 are efficacious against Acanthamoeba castellanii, trophozotes and cysts, enough for consideration as a preventative management strategy for cleansing contact lens 5,32,34 . At present, cytotoxicity information of these mixed surfactants provided in this study against human epithelial and corneal cell lines are absent and as such are a significant barrier for use as a medical device. However, Polyquad, a QAC, is a preferred preservative in contact lens cleaning solutions utilised by a number of manufacturers 52 . In addition, QAC "neutralizing" agents such as β-cyclodextrin and NeutraQuat ™ that bind free QACs, have very high neutralisation efficacy, such that their use as a post-sterilisation step should be suitable to minimise eye damage negating the use of tap water 53,54 . In short, we propose that combinations of different alkylphosphocholines could be introduced in disinfectant protocols as an effective preventative management protocol for contact lens care, to prevent Acanthamoeba contamination and protect compliant contact lens users.

Methods
cell lines used in this study. Acanthamoeba castellanii ATCC 50370 trophozoites and cysts were cultured in Peptone Glucose (PG) medium and encystment medium 55 respectively at 25 °C. Cysts were prepared by incubating trophozoites for 8 days in encystment medium and integrity validated with calcofluor (0.25 µg/ml) assay 56 or sodium dodecyl sulfate (SDS, 0.5% w/v) disintegration assay 57 and microscopic observation. cytotoxicity assay. The compounds used in this study all contained alkyl-carbon chains: three APCs For the combined assay, similar concentrations of QAC18 and APC16 were prepared and 18.75 µg/ml and 37.5 µg/ml of QAC12 added (to a final volume of 100 µl) were added to 10 5 cells/ml in a 96-well plate and later incubated at 25 °C for 96 hours.
www.nature.com/scientificreports www.nature.com/scientificreports/ Resazurin (5 mg/ml) 58 and trypan blue (0.4 µg/ml) 59 were then added to trophozoites and cysts respectively after exposed to the compounds and the cell viability estimated from changes in absorbance (at 570 nm and 595 nm) and microscopically respectively.
For the drug kinetic assay, viability was assessed by the linked Alamar blue-reversion assay. Drug-treated cysts were washed with PBS and fresh PG medium was added to allow reversion to trophozoites for 5 days, after which, alamar blue was added and viability estimated as described above.
Efficacy was expressed as a percentage of untreated controls for each drug concentration used and the data used to calculate the IC 50 . Cell density was estimated using the modified Neubaur hemocytometer and expressed as cells/ml. isolation of total genomic DnA. Genomic deoxynucleic acids (gDNA) from 10 6 A.castellanii trophozoites (treated with and without QAC) were extracted from cells harvested by centrifugation (850 × g, 10 min), lysed with UNSET buffer (8 M urea, 150 µM NaCl, 2% SDS, 1 µM ethylenediaminetetraacetic acid (EDTA), 100 µM Tris-HCl, pH 7.5) and the DNA extracted with phenol-chloroform (1:1 v/v). The biphasic suspension with DNA enriched on the lower trizol layer was centrifuged (12,000 × g for 15 min, 4 °C) for clear separation and transferred to a new microcentrifuge tube. The DNA was precipitated with cold ethanol (100%, v/v) and 0.3 M sodium acetate, washed twice with 70% (v/v) ethanol (12,000 × g for 10 min, 4 °C), air dried and resuspended in distilled water.
Morphometric analysis. A. castellanii trophozoites incubated with and without QAC at a duration determined by experimental protocol were examined microscopically and their sizes measured using the Open Laboratory (Improvision) calibration graticule and the the Holomonitor M4 live cell analyser and Holomonitor App Suite. The mean body sizes, area, volume and perimeter and the proportion of cells with a fixed interval were determined. potassium (K + ) determination assay. Potassium [K + ] concentration in spent medium of A. castellanii trophozoites and cysts were harvested as described above, filtered with the 0.22 µm syringe filter and used for [K + ] determination using the Atomic Absorption Spectrometry (ASS, absorbance 766.5 nm). A K + standard curve was used to convert absorbance values to concentration (expressed as mg/ml). protein and DnA determination assay. Total protein concentration in cell extract and media of QAC-treated and untreated A. castellanii trophozoites and cysts were estimated using the Quick Start Bradford Protein Assay (Bio-Rad) as described in the manufacturer's instructions and the absorbances obtained at 595 nm using the Tecan Infinite M1000 Pro plate reader and expressed as ng/µl protein using a BSA standard curve. QAC18 and PG medium were used in control experiments. Proteins were visualised by SDS-PAGE gel as described previously 38 . DNA concentration in cell extract and media of QAC-treated and untreated A. castellanii trophozoites were estimated at 260 nm using the Nanodrop ND-1000 and expressed as ng/µl DNA. DNA was visualised with DNA agaraose gel (0.75%) as described previously 38 . QAc-DnA interaction assay. The interaction between the gDNA from A. castellanii with QAC12 or QAC18 was investigated at DNA:QAC ratios of 1:0, 1:1, 1:10 and 1:20 for 15 min and the concentration of unreacted DNA in the complex estimated at 260 nm using the Nanodrop ND-1000 at 260 nm. Statistical analysis. Descriptive statistics of mean and standard deviation values were used to represent data for at least four independent experiments each done in triplicate. To explore differences, between baseline and assay characteristics, T-test statistics were calculated with a statistical threshold of significance set at p < 0.01 or p < 0.05.