## Introduction

Amoebae from the genus Acanthamoeba are ubiquitous protozoa which can be opportunistic pathogens in humans. They are the etiologic agent of Acanthamoeba keratitis (AK), a sight-threatening disease related to the use of contact lenses, and granulomatous amoebic encephalitis, a fatal central nervous system disease more common in immunocompromised individuals1,2,3.

The ameba life cycle includes the trophozoite form, which is able to feed, reproduce, and infect, and the cyst stage, which is the resistant form1,2,3. The change from trophozoite to cyst is usually induced by adverse environmental conditions. Cysts can resist extreme environmental pressures such as high temperatures, ultraviolet radiation, gamma radiation and changes in pH, among others4,5. Moreover, some drugs used to treat AK induce cyst formation6. Therefore, eradication is challenging and requires the development of new treatments.

For AK infections, standard treatment drugs are biguanides (such as chlorhexidine (CLX)), diamidines (such as propamidine isethionate) and the aminoglycoside antibiotic, neomycin3,7. Biguanides and diamidines are membrane-acting agents while neomycin is a protein synthesis inhibitor8. They are generally used in combination to overcome drug resistance, although their high cytotoxicity along with their inability to eliminate cysts (recurrences occur in about 10% of cases) has led to the search for new compounds with therapeutic applications7,8. Although the posology may vary, hourly administration for three days is routinely recommended. Total therapy duration is 3–4 weeks, although patient withdrawal is an additional cause of treatment failure4. Furthermore, resistance to standard treatments has been reported, mainly when CLX is used as monotherapy9.

In recent years, newly synthesized dendritic compounds have been described as an alternative for treating bacterial, viral, and protozoal infections10,11,12,13. These compounds have shown anti-Acanthamoeba activity against trophozoites and cysts, as well as low cytotoxicity in vitro11. Dendritic compounds can be combined with standard drugs used in therapy. Consequently, there would be a reduction in the required effective drug concentrations and in the cytotoxic effects on patients. In some cases, synergy has also been described14. Moreover, the strategy of combined therapy with dendritic compounds could contribute to prevent drug resistance development15.

In an attempt to identify new therapeutic approaches, our study focused on the evaluation of different generations (1–3) of carbosilane dendrons derived with 4-phenylbutyric acid (PBA) against trophozoites and cysts of Acanthamoeba spp. clinical isolates. Additionally, possible synergistic effects of treatments combined with CLX were tested. Cytotoxicity assays were then performed with effective concentrations. Finally, scanning electron microscopy was used to elucidate the effects of the most effective dendritic compound alone and in combination with CLX on trophozoite structure.

## Materials and methods

### Acanthamoeba spp. clinical strains

A. polyphaga 2961 (kindly supplied by Dr. E. Hadas, Poznan University of Medical Sciences, Poland) and A. griffini MYP2004 (isolated by our research group, University of Alcalá, Spain) were used16. A. polyphaga 2961 was grown at 32 °C in Peptone Yeast Glucose broth supplemented with Bacto Casitone (PYG + B) and A. griffini MYP2004 was grown at 37 °C in CERVA16,17. The strains were maintained by weekly media changes.

### Dendritic compounds and standard treatment drugs used

Different generations of cationic carbosilane dendrons derived using 4-phenylbutyric acid (PBA) at the ArCO2Gn(SNMe3I)m focal point (n = 1, m = 2; n = 2, m = 4 and n = 3, m = 8) were studied and are referred to as ArCO2G1(SNMe3I)2 (1), ArCO2G2(SNMe3I)4 (2) and ArCO2G3(SNMe3I)8 (3), respectively (Fig. 1). The focal point is the location from which the reactive peripheral groups emerge to form the hyperbranched wedge. These systems were prepared as described by Lozano-Cruz et al., 2020. To perform combination assays, chlorhexidine digluconate (CLX) (Sigma-Aldrich Ltd., St. Louis, MO, USA) was considered the reference drug3.

### Amoebicidal activity assays against trophozoites

The three dendrons, ArCO2G1(SNMe3I)2 (1), ArCO2G2(SNMe3I)4 (2) and ArCO2G3(SNMe3I)8 (3) were evaluated against A. polyphaga 2961 and A. griffini MYP2004 trophozoites. Assays were prepared in 96-well microtiter plates treated with Poly-L-Lysine (CellStar, Greiner Bio-one).

Firstly, the inoculum was adjusted to 10,000 trophozoites (A. polyphaga 2961) and 15,000 trophozoites (A. griffini MYP2004) per well, based on a previous strain growth analysis performed in our laboratory. For this purpose, a Fuchs-Rosenthal© counting chamber (Optic Labor) and 0.2% Congo Red in distilled water (Congo Red, Sigma Aldrich) were used under an optic microscope (Carl Zeiss)17.

Secondly, the dendrons were twofold serially diluted from stock solutions to reach final concentrations ranging from 2 to 512 mg/L and 100 μL of each concentration were added to each well. Then, 100 μL of the adjusted trophozoite solution (2 × PYG + B and 2 × CERVA for A. polyphaga 2961 and A. griffini MYP2004, respectively) were also added. Plates were incubated at 32 °C for A. polyphaga 2961 and 37 °C for A. griffini MYP2004. Manual counting using the 0.2% Congo Red exclusion assay was performed at 24 and 48 h of treatment and percent viability was defined as: % viability = (mean treated/mean control) × 100. The minimum trophozoite amoebicidal concentration (MTAC) was defined as the lowest concentration of test solution that produced a complete reduction in trophozoite viability18.

Each drug concentration was tested in triplicate and in at least two independent experiments. A control well containing the amebae with no treatment and control wells of medium were included. All microtiter plates were sealed with Parafilm®. These procedures apply to all tests carried out in this study.

### Cysticidal activity assays

Cysts were obtained from ameba cultures in logarithmic phase under optimal growing conditions. The medium was replaced with Neff’s encystment medium and flasks were agitated for 9 days at room temperature17. A. polyphaga 2961 and A. griffini MYP2004 cyst assays were run in 96-well microtiter plates (Deltalab).

The inoculum was adjusted in Neff’s encystment medium, and 10,000 cysts were inoculated per well for both strains. The dendrons were twofold serially diluted to final concentrations ranging from 2 to 512 mg/L; 100 μL of each concentration were mixed with 100 μL of adjusted cyst solution. The dendron solutions were removed after 24 and 48 h of treatment, then each well was washed with 1 × PBS (10x, Sigma Aldrich, St. Louis, MO, USA) and culture medium was added (1 × PYG + B for A. polyphaga 2961 and 1 × CERVA A. griffini MYP2004). Plates were incubated at optimal growing conditions (32 °C for A. polyphaga 2961 and 37 °C for A. griffini MYP2004). Wells were observed three times per week with an inverted microscope (Motic AE21) for 21 days to visualize excystment and determine the minimum cysticidal concentration (MCC), defined as the lowest concentration that completely inhibits excystment and trophozoite growth18.

### Combined treatment against trophozoites

To perform these assays, the procedure was the same as for the amoebicidal activity assays described above. For this purpose, the checkerboard method was used19. The most effective dendron was tested in combination with CLX in a final volume of 200 μL (1:1 ratio). Required concentrations were achieved by serial dilution. Dendron concentrations ranged from 0.5 to 16 mg/L (< MTAC), while CLX concentrations ranged from 0.5 to 4 mg/L (< MTAC). Manual counting was performed after 24 and 48 h of incubation under optimal growing conditions to determine trophozoite viability compared to the untreated control.

To determine synergy, the fractional inhibitory concentration index (FICI) was calculated:

$${\text{FICI }}\left( {{\text{combination}}} \right) \, = \, [{\text{MTAC}}_{{({\text{Dendron}} + {\text{ CLX}})}} /{\text{ MTAC}}_{{\left( {{\text{Dendron}}} \right)}} \left] { \, + \, } \right[{\text{MTAC}}_{{\left( {{\text{CLX }} + {\text{ Dendron}}} \right)}} /{\text{ MTAC}}_{{({\text{CLX}})}} ]$$

The result was interpreted as synergistic when FICI was $$\le$$ 0.5, additive when 0.5 $$<$$ FICI $$\le$$ 1, indifferent when 1 $$<$$ FICI $$<$$ 4 and antagonistic when FICI $$\ge$$ 415.

### Combined treatment against cysts

Experiments were prepared as explained above for the cysticidal activity assays. Briefly, 100 μL of adjusted cyst inoculum was added to each microtiter plate well. The combination of the most effective dendron in concentrations ranging from 2 to 64 mg/L (MCC) and CLX, in concentrations from 1 to 8 mg/L, were added to each well (50 μL of each compound). After 24 and 48 h of treatment, the compounds were discarded. Wells were then washed twice with 1 × PBS (10x, Sigma Aldrich, St. Louis, MO, USA) and fresh medium was added. Finally, plates were incubated under optimal growing conditions. They were observed three times per week using an inverted microscope (Motic AE21) for 21 days to determine excystment. To determine synergy, the FICI was calculated as previously described.

### Cytotoxicity assay in HeLa cells

HeLa cells were cultured in Dulbecco's Modified Eagle Medium (DMEM) (Gibco) supplemented with 10% fetal bovine serum and 1% of an antibiotic mixture. The inoculum was adjusted to 1 × 104 cells per well in a 24-well microtiter plate. Cells were grown for 3–4 days at 37 °C with 5% CO2 to reach confluence17. Then, the medium was discarded and 400 μL of the most effective dendron and CLX concentrations diluted in DMEM medium were added.

After 24 and 48 h of treatment, the medium was discarded once again, and the wells were washed twice with 1 × PBS (10x, Sigma Aldrich, St. Louis, MO, USA). After that, 500 μL of fresh medium and 50 μL of 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) was added (Sigma-Aldrich Ldt). The plates were incubated at 37 °C for 4 h. Then, the medium was discarded and 500 μL of DMSO were added20. Absorbance values were determined at the 570 nm wavelength using a spectrophotometer (BioTek Instruments Inc. Model: ELX 800).

Viability was calculated as (absorbance mean of treated/absorbance mean of non-treated) × 100. Values higher than 90% viability were considered non-cytotoxic, while values between 75 and 90% were considered low cytotoxicity. If viability descended to the range of 60–75%, it was deemed a moderate cytotoxic level. High cytotoxicity was established when viability values were lower than 60%21.

### Scanning electron microscopy (SEM)

SEM study was performed to evaluate the impact of the most effective dendron, CLX and combinations of these as previously described by our group16. To perform these studies, 200 μL of adjusted trophozoite suspension were placed on a glass coverslip. After 1 h of incubation under optimal growing conditions (32 °C for A. polyphaga 2961 and 37 °C for A. griffini MYP2004), the medium was discarded and treated for 24 or 48 h. Then, the medium was discarded, the wells were washed twice with PBS, and the fixative solution (2% glutaraldehyde and 1% CaCl2 in Milloning’s solution (NaH2PO4·H2O and NaOH)) was added for 1 h. Afterwards, the coverslips were washed with washing solution. The samples were dehydrated in increasing concentrations of ethanol (30, 50, 70, 95 and 100%) and finally in anhydrous acetone. Desiccation was performed in a critical point drying apparatus (Polaron E-3000), and the coverslips were mounted on aluminium stubs and coated with gold (Polaron E-5000/5100). The samples were examined using a Scanning Electron Microscope (JSM-IT500, JEOL) at the Medicine and Biology Research Support Center (University of Alcalá).

### Statistical analysis

For statistical analysis and graph generation, Microsoft Excel (Microsoft Office 365, Microsoft, Redmond, Washington, USA) and GraphPad Prism 8® (GraphPad Software, San Diego, California, USA) were used. Two-way ANOVA and One-way ANOVA (followed by Dunnett’s multiple comparisons test) were performed, and significance was established at p < 0.05. IC50 was obtained on GraphPad Prism 8® a using non-linear regression analysis22.

## Results

### Amoebicidal activity assays against trophozoites

All dendrons evaluated, ArCO2G1(SNMe3I)2 (1), ArCO2G2(SNMe3I)4 (2) and ArCO2G3(SNMe3I)8 (3) showed amoebicidal activity against trophozoites, although in variable concentrations (Fig. 2, Table 1). ArCO2G2(SNMe3I)4 (2) was the most effective, exhibiting a MTAC of 64 mg/L for A. griffini MYP2004 after 24 and 48 h of treatment (Fig. 2, Table 1). In the case of A. polyphaga 2961, this concentration was equal after 24 h of treatment and varied in the range of 32–64 mg/L after 48 h of treatment (Fig. 2, Table 1). The IC50 of each dendron is shown in Table 2.

### Cysticidal activity assays

The MMC of the dendritic compounds on the excystment of cysts was determined by observation under the microscope18. Only ArCO2G2(SNMe3I)4 (2) demonstrated a cysticidal activity in the concentration range tested (2–512 mg/L) (Table 3). For A. griffini MYP2004, the MCC was 64 mg/L, while for A. polyphaga 2961 it was higher, 128 mg/L (Table 3). There was no change between both times of incubation.

### Combined treatment against trophozoites

The effective concentrations used in the combination treatments were lower than the effective concentrations for each compound tested individually. Moreover, when the FICI was calculated, a synergistic effect was found when 16 mg/L ArCO2G2(SNMe3I)4 (2)/1 mg/L CLX were combined against A. griffini trophozoites, and 16 mg/L ArCO2G2(SNMe3I)4 (2)/0.5 mg/L CLX against A. polyphaga trophozoites. Therefore, the ArCO2G2(SNMe3I)4 (2) effective concentration was reduced from 64 to 16 mg/L and the CLX effective concentration was reduced from 2–4 mg/L to 0.5–1 mg/L (non-cytotoxic concentrations). Under these conditions, the combined therapy improved compound efficacy and reduced the effective concentrations and thus, the cytotoxicity of both compounds. Nevertheless, some of the other effective combinations had additive or indifferent effects (Tables 4 and 5, Fig. 3).

### Combined treatment against cysts

As described in the combination assay against trophozoites, effective treatment concentrations against cysts were lower in combination than for each of the drugs individually. FICI values showed a synergistic effect for the highest ArCO2G2(SNMe3I)4 (2) concentrations tested in combination. For example, synergy was found when 16 mg/L ArCO2G2(SNMe3I)4 (2)/2 mg/L CLX were combined against A. griffini cysts, and 32 mg/L ArCO2G2(SNMe3I)4 (2)/2 mg/L CLX against A. polyphaga cysts. Therefore, the ArCO2G2(SNMe3I)4 (2) effective concentration was reduced from 64–128 to 16–32 mg/L and the CLX effective concentration was reduced from 8 to 1–2 mg/L. Under these conditions, the combined therapy improved compound efficacy and reduced the effective concentrations against cysts and, in consequence, the cytotoxicity of both compounds. However, the primary outcome was an additive effect, once again (Tables 6 and 7).

### Cytotoxicity assay in HeLa cells

The cytotoxicity of treatment combinations with amoebicidal and amoebostatic activity was evaluated in HeLa cells. Individual evaluation of ArCO2G2(SNMe3I)4 (2) and CLX cytotoxicity in HeLa cells performed previously showed that ArCO2G2(SNMe3I)4 (2) demonstrated low cytotoxicity in the range of 4–8 mg/L, according to the aforementioned criteria, even after 48 h of exposure23. However, CLX exhibited high cytotoxicity at similar concentrations after only 24 h.

Effective amoebicidal combinations against trophozoites of both strains showed some cytotoxicity. Nevertheless, the combination that showed a moderate cytotoxicity (65% cell viability) could inhibit proliferation and eliminate 100% A. polyphaga 2961 while reducing A. griffini MYP2004 viability up to 97% after a 24 h treatment (8 mg/L ArCO2G2(SNMe3I)4 (2)/2 mg/L CLX) (Fig. 4, Table 8). In addition, the combination of 2 mg/L ArCO2G2(SNMe3I)4 (2) and 1 mg/L CLX eliminated 100% of A. polyphaga trophozoites after a 48 h treatment and showed moderate toxicity (75% cell viability) (Fig. 4, Table 8).

The effective treatment concentrations achieved in different combinations against cysts were higher than the ones used against trophozoites that showed high cytotoxicity. Therefore, further cytotoxicity was not evaluated, as it was assumed to be greater than in those treatment combinations previously studied, i.e., more cytotoxic than the combinations of concentrations that were effective against trophozoites.

### Scanning electron microscopy (SEM)

Images of untreated trophozoites obtained by SEM showed the characteristics of healthy amoebae with numerous acanthopodia and membrane integrity. However, membrane alterations were observed when the amoebae were treated with both the ArCO2G2(SNMe3I)4 (2) dendron and CLX, as monotherapy and in combination (Figs. 5 and 6).

As observed in the images, CLX completely disintegrates the Acanthamoeba membrane, causing the release of cytoplasmic constituents (Figs. 5D,E, 6D,E), reducing the number of acanthopodia and inducing a rounded shape in trophozoites (Figs. 5D,E, 6D,E). These observations may suggest that dendron ArCO2G2(SNMe3I)4 (2) caused the appearance of holes, possibly due the interaction with the negative charges on the amoeba surface and, in consequence, it may produces the disruption of the plasma membrane (Fig. 6B (arrow)). In addition, after treatments with drug combinations, no other changes were noticed apart from the ones aforementioned for each of the treatments as monotherapies (Figs. 5F,G, 6F,G). Alterations were more evident after 48 h of treatment.

## Discussion

AK is a rare, sight-threatening disease, although the severity of its outcomes together with the lack of effective treatments has made the development of new therapeutic agents an urgent line of research3,24.

On the one hand, CLX is a biguanide considered a standard treatment drug in AK cases. It can be used in combination with a diamidine or as a monotherapy24,25. However, it is highly cytotoxic at effective concentrations26,27. To solve this problem, combination strategies have been proposed15. On the other hand, dendritic compounds have shown effectiveness in previous studies not only against bacteria and viruses, but also against amoebae17,28,29,30. Moreover, these compounds are water-soluble, and their structure facilitates a directed design, characteristics that make them suitable therapeutic agents10. Their cationic structure, like CLX, may explain their biocidal activity, even though their mode of action has not yet been completely elucidated31. The three dendrons evaluated, ArCO2G1(SNMe3I)2 (1), ArCO2G2(SNMe3I)4 (2) and ArCO2G3(SNMe3I)8 (3), were effective against trophozoites at different concentrations. In addition, the second generation dendron, ArCO2G2(SNMe3I)4 (2), with four positive charges on the periferial groups and PBA at the focal point showed cysticidal activity. Cysts have the ability to resist treatment by different compounds because they form a protective wall, composed mainly of cellulose (endocyst), proteins and polysaccharides (exocyst)32. Nevertheless, A. griffini MYP2004 cysts seemed to be more sensitive than A. polyphaga 2961 cysts to treatment with ArCO2G2(SNMe3I)4 (2). Although this result may seem to be contradicted by the data shown in literature, the dendron could act differentially on each cyst wall depending on its exact composition3. Previous studies described differences in cystic proteins within isolates, so these data may also indicate protein variability between the Acanthamoeba strains used in our study and other amoebic strains33.

CLX, which caused the alterations observed by SEM after treatments, is a cationic molecule that interacts with the negatively charged plasma membrane, thus provoking membrane destabilization, pore formation, cytoplasmic content leakage and cell death14,34. Dendrons are also cationic molecules, thus their mode of action might be similar and alterations may be due to positive charges that interact with and disrupt cell structures31. In addition, dendrons have an aromatic ring at their focal point (growing dendron point where peripherial groups emerge) that could improve their antibacterial activity35.

In previous studies, amoebicidal activity against A. polyphaga 2961 trophozoites and cysts shown by dendritic wedge topology systems with isobutyric acid at the focal point (IC50 at 24 h treatment: ArCO2G2(SNMe3I)4 for trophozoites 8.24 mg/L and for cysts 128 mg/L) was greater than that shown by spherical topology systems containing a polyphenoxy nucleus (IC50 at 24 h treatment: G1O3(SNMe3I)6 for trophozoites 16.9 mg/L and for cysts > 512 mg/L) or a silicon atom nucleus (IC50 24 h G1O3(SNMe3I)6, 430.1 mg/L for trophozoites and > 512 mg/L for cysts)36. In view of these results and considering that they all have a similar number of positive charges in their structure, we suggest that the presence of the aromatic ring modulates the activity of the dendritic system, and the topology of the system also seems to have a clear influence on the activity. The dendritic wedge presents a more open structure that could provide the aromatic ring with the possibility for interacting with the amoeba membrane.

The FICI index was calculated in the combination assays15,37. The main goal was to find a combination that could reduce effective doses to less cytotoxic ones. To this end, the most effective dendron ArCO2G2(SNMe3I)4 (2) was tested in combination with CLX. The results showed a synergistic effect at highly cytotoxic concentrations, although other combinations with additive effects achieved a reduction in the effective concentration and had moderate cytotoxicity. For example, certain combinations were able to reduce A. griffini MYP2004 viability to only 3% and completely eliminate A. polyphaga 2961 trophozoites (0% viability). Consequently, we highlight those combinations which achieved a reduction in effective concentrations of CLX with an increase in amoebicidal activity and a significant reduction in cell cytotoxicity. We also propose that the use of these combinations in prevention and treatment strategies could be studied in in vivo models. Additionally, combined therapy has been shown to prevent the development of new resistant strains, thus combination strategies may have extra value in the fight against AK resistance15. In summary, our data support the suitability of these promising dendrons alone and in combination for treating AK.

## Conclusions

The cationic dendrons with isobutyric acid at the focal point evaluated in this study showed amoebicidal activity against trophozoites. Furthermore, the second generation dendron (ArCO2G2(SNMe3I)4) (2) with four positive charges on the surface had cysticidal activity and a synergistic effect when combined with CLX, a standard treatment drug. Besides, the combinations with a drastic reduction in trophozoite viability, even to 0%, showed only moderate cytotoxicity. Additionally, the ArCO2G2(SNMe3I)4 (2) dendron and CLX modes of action showed similarities as both were capable of disrupting trophozoite membranes.