Assessing Efficacy of Combined Riboflavin and UV-A Light (365 nm) Treatment of Acanthamoeba Trophozoites

Abstract

Purpose.: To assess the Acanthamoeba trophozoite viability in vitro and treatment of Acanthamoeba keratitis in a hamster model using ultraviolet light A (UV-A) and riboflavin (B2).

Methods.: A sample of Acanthamoeba sp. cultured was transferred to a 96-well plate and exposed to B2 and the UV-A light (365 nm wavelength) at a power density of 3 mW/cm², 8 mm spot diameter, for 30 minutes. The exposure was done in triplicate. Control groups were prepared in triplicate as well: blank control, UV-A only, riboflavin only, and dead control. Cell viability assessment was done using the trypan blue dye exclusion method. Acanthamoeba keratitis was induced in Chinese hamsters; who were randomly assigned to one of the animal groups: UV-A + B2, propamidine isethionate (Brolene; Sanofi-Aventis, Ellerslie, Auckland, Australia), UV-A + B2 + propamidine isethionate (Brolene), only UV-A, only B2, and blank. Throughout the 14 days after treatment the animals were examined clinically. Histology and clinical scores of all groups were compared.

Results.: The in vitro study showed no difference between the treatment group UV-A + B2 and the control groups. In the hamster keratitis model a significant improvement of clinical score was observed for the groups propamidine isethionate (Brolene) and UV-A + B2 + propamidine isethionate (Brolene) (P = 0.0067). Also a significant worsening of clinical score was observed in the other groups: UV-A + B2 group (P = 0.0084), only UV-A (P = 0.0078), B2 only (P = 0.0084), and blank (P = 0.0082). No difference was observed between propamidine isethionate (Brolene) and UV-A + B2 + propamidine isethionate (Brolene).

Conclusions.: The adjunctive use of UV-A and B2 therapy did not demonstrate antitrophozoite activity; in vivo UV-A and B2 did not demonstrate efficacy in this model.

Free-living amoebae of Acanthamoeba genus are eukaryotic organisms, aerobic, and widely dispersed in nature, which can inhabit aquatic and terrestrial environments. ¹ Over the past few years, this protozoan has played different roles in human pathology, including disabling painful infection of the cornea and ocular structures. ¹ Despite efforts at early diagnosis and specific treatment, Acanthamoeba keratitis (AK) continues as a threatening eye infection and clinical reports reveal a trend of yearly increase in developing countries. ² Current therapeutic procedures of AK are based on antimicrobial compounds and use of topical corticosteroids. ^3,4 Adverse effects with treatment of AK have been previously described, including parasite resistance to one or more chemical agents, the potentiation by steroids of the pathogenesis of Acanthamoeba sp. at the site of infection and changes in the structure of the iris, lens, and anterior chamber due to prolonged therapy using antimicrobial compounds at higher concentrations. ^{5 –8} When Acanthamoeba infections do not respond to medical therapy, penetrating keratoplasty (PK) with or without adjunctive cryotherapy has been proposed as alternative therapeutic option for visual rehabilitation. ^9,10 Graft survival after PK in this population is poor and can lead to postoperative complications, including graft rejection, stromal thinning, leaking incisions, glaucoma, and cataract. ^9,11 Recurrence of AK can follow corneal transplantion if viable cysts or trophozoites of protozoa persist in the tissue adjacent to the site of infection or are not completely eradicated from the host corneal tissue before surgery. ^12,13  

Irradiation of corneal infection using long-wavelength ultraviolet light (UV-A) associated with a photosensitizer riboflavin has been proposed as an alternative therapeutic approach in the treatment of AK. ^14,15 Riboflavin UV-A is thought to increase collagen fiber diameter and rigidity by collagen cross-linking and the method has been used to stop the progression of keratectasia in patients with keratoconus. ¹⁶ Photochemical inactivation of corneal pathogens (i.e., bacteria and fungus, and blood-borne parasites) has been evaluated. ^17,18 However, the antimicrobial effect of UV light at a long-wavelength and riboflavin combination in the viability of Acanthamoeba cysts and trophozoites has not been determined. 

In the present study, we describe the effect of UV-A and riboflavin (B2 vitamin), alone and in combination, on the cell viability of Acanthamoeba cysts and trophozoites, and we correlate in vitro and in vivo findings. Furthermore, we report the efficacy of the photochemical process compared with an antimicrobial chemical agent (propamidine isethionate; Brolene [Sanofi-Aventis, Ellerslie, Auckland, Australia]) commonly used for the inactivation of this protozoan in corneal infections. 

Materials and Methods

Acanthamoeba Strain

The research was approved by the local ethics committee and was conducted in accordance with the tenets of the Declaration of Helsinki. Informed consent of patient was obtained before sample collection. Acanthamoeba strain was isolated from a corneal scraping sample obtained from the left eye of a female 56-year-old soft contact lens wearer. The patient presented with severe pain, tearing, foreign body sensation, conjunctival hyperemia, burning, discharge, photophobia, and visual impairment. Briefly, the amoeba was isolated from corneal tissue by plate-culture procedure onto a Foronda medium covered with avirulent, plasmidless, and heat-inactivated Escherichia coli strain at 25°C. ¹⁹ Occurrence of Acanthamoeba sp. was observed after 8 days postinoculation, when cysts and trophozoites were picked up aseptically from nonselective agar medium and transferred to a tissue culture flask containing 5 mL of Neff's broth medium ²⁰ with gentamicin (10 mg/mL) to proceed with axenization process of primary isolate. To provide phylogenetic information, the Acanthamoeba strain was identified at the genotype level by nucleotide comparison with the available DNA sequences in the GenBank data library. Briefly, chromosomal DNA of trophozoites was extracted and purified using a DNA mini kit (QIAamp; Qiagen, Valencia, CA) according to the manufacturers' instructions. The amplimer ASA.S1 was amplified and sequenced in both strands twice using JDP primer pairs as previously described. ²¹ Molecular identification of T4 genotype related with Acanthamoeba isolate was carried out with software (BLAST; Pittsboro, NC). ²² The partial 18S rDNA nucleotide sequence from the Acanthamoeba isolate was deposited in the GenBank database under accession number JN222978. 

In Vitro Assays

Acanthamoeba sp. was cultivated axenically in Neff's broth medium. Trophozoites were harvested at the end of the logarithmic phase of growth by draining the supernatant medium out and adding to the culture flask a cell dissociation reagent (TrypLE Express; Invitrogen, Carlsbad, CA) following the manufacturer's protocol. Amoebas were pelleted by centrifugation at 1400g for 10 minutes at 25°C. The supernatant was removed and trophozoites were resuspended in diluted saline buffer (4 mL distilled water, 1 mL 0.9% saline). The amount of viable cells was measured using Fuchs-Rosenthal counting chambers and the trypan blue dye exclusion method. Experimental cellular concentration was adjusted to 3.6 × 10⁴ cells/ mL; for each trial, the count of the chambers was performed in duplicate. Four control groups were performed, as follows: blank control (control 1), the cells recovered from each well at the end of the experiment without exposure to UV light or riboflavin solution; UV light (control 2), the cells recovered from the well previously submitted to 30 minutes of UV-A exposure without riboflavin instillation; B2 (control 3), the cells recovered from the well previously submitted to riboflavin instillation for 30 minutes without UV-A exposure; dead control (control 4), the cells recovered from the well previously submitted to propamidine isethionate (Brolene) for 30 minutes. Each control group was assayed in triplicate. 

An aliquot of 20 μL of cell-containing solution was placed into each well of a sterile 96-well microplate (Corning Life Science, Lowell, MA). Each well of the microplate has an internal diameter of 6.85 mm, thus ensuring that the entire area can be exposed to the UV light spot diameter of 8 mm. To promote the adhesion of trophozoites in the bottom of the microplate, 200 μL of Neff's medium were added to each well. The cell's attachment was observed microscopically, the culture medium was drained out carefully, and 30 μL of diluted saline buffer was added to each well to wash the adhered cellular content. A volume of 40 μL of riboflavin-5-phosphate 0.1% (Sigma-Aldrich, St. Louis, MO) was dispensed into each well followed by irradiance of UV-A light source (Opto Xlink, Opto, Sao Carlos, Brazil) with 365 nm wavelength, at a power density of 3 mW/cm², for 30 minutes. The light exposure was done once per well in triplicate. After the UV-A light exposure the riboflavin was drained, dissociation reagent solution (TrypLE Express; Invitrogen) was added to each well and the cells were recovered. The cell viability of each assay was measured by trypan blue dye exclusion method in a Fuchs-Rosenthal counting chamber under microscope. 

In Vivo Assays

All animals were treated in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research, using protocols approved and monitored by the Johns Hopkins University School of Medicine Animal Care and Use Committee. To study the functional role of riboflavin associated with UV-A in the increased resistance of collagen fibers to the enzymatic activity of the protozoan, a complementary experimental group was added to in vivo assays, consisting of animals submitted to a previous riboflavin-UV-A procedure followed by chemical therapy with propamidine isethionate (Brolene) daily for 14 days. Five female Chinese hamsters 5 to 6 weeks of age were randomly assigned to each experimental group (n = 6). All corneas were evaluated at baseline and no pre-existing lesions were present. Animals were anesthetized with a mixture of ketamine/xylazine (100 mg/kg/10 mg/kg) intraperitonially and topical anesthetic proparacaine (Alcon Laboratories, Inc., Ft. Worth, TX) was applied to each eye. Induction of AK was carried out by scraping a 4-mm diameter of corneal epithelium with a sterile cotton applicator under an operating microscope. Silicone hydrogel contact lenses (Johnson & Johnson, New Brunswick, NJ) containing Acanthamoeba sp. at concentration of 3.5 × 10⁴ cells/mL (90% trophozoites) were previously prepared using a 3-mm dermal punch (Miltex Instruments, Plainsboro, NJ). The contact lenses were placed aseptically on the corneal surface of each animal and the eyelids were closed by tarsorrhaphy with 10 to 0 sutures (Mersilene, Ethicon, Inc., Somerville, NJ). Sutures were removed 7 days after exposure to parasites and the contact lenses were removed. Eyes were observed under an operating microscope and slit lamp for 14 days. The severity of corneal infection was scored according to the criteria proposed previously (Table 1). ²³ Briefly, the clinical score system is based on the evaluation of epithelial defects, stroma edema, vascularity, and stromal opacity. Each finding was scored from zero to four points. Thus, the higher clinical score was related with the worse clinical status. Treatment profiles for each animal group are shown in Table 2. After 14 days of treatment, the animals were euthanized and the whole globes were harvested. Under a dissecting microscope, the cornea was excised and fixed in 10% buffered formaldehyde for 72 hours, followed by immersion and orientation in paraffin solution. The samples were cut, mounted, and stained with hematoxylin and eosin according to standard methods. The sections were analyzed using an inverted light microscope (Olympus LH50A; Olympus, Center Valley, PA) with a 40 × objective. Digital images of the cornea were captured using a 10 × objective microscope camera (JENOPTIK Laser Technology Corp., Brighton, MI) equipped with specialized software (CapturePro 2.5 Image Aquisition Software for ProgRes Microscope Cameras [JENOPTIK Laser Technology Corp.]). 

Table 1.

 

Clinical Scoring Criteria for Acanthamoeba Keratitis in Chinese Hamsters ²³

Score	Epithelial Defects	Stromal Edema	Vascularity	Stromal Opacity
0	No defects	No edema	No vessels	Stroma clear
1	Defect < 25% of surface	Edema < 25%	Vessels on < 25% of circumference of the cornea	Opacity < 25% of cornea area, pupil easy visible
2	Defect 25% to < 50%	Edema 25% to < 50%	Vessels 25% to < 50%	Opacity 25% to < 50%, pupil hardly visible
3	Defect 50% to < 75%	Edema 50% to < 75%	Vessels 50% to < 75%	Opacity 50% to < 75%, pupil not visible
4	Defect > 75% of surface area	Edema > 75%	Vessels > 75%	Opacity > 75%, pupil and iris not visible

Table 2.

 

Therapeutic Patterns Established for Animals Not Infected (Blank Group) and Animals Infected with AK

Animal Treatment Group	Schematic Therapeutic Pattern
Blank group	No treatment
UV-A light exposure	UV-A light exposure for 30 min
Riboflavin instillation	Riboflavin drops (5 μL) were instilled onto the cornea every 5 min for 60 min
Propamidine isethionate (Brolene; Sanofi-Aventis, Ellerslie, Auckland, Australia) instillation	Propamidine isethionate (Brolene) drops (5 μL) were instilled onto the cornea every hour for 6 hours during 14 days
Riboflavin + UV-A	Riboflavin drops (5 μL) were instilled onto the cornea every 5 min for 30 min followed by irradiance of UV-A light source for 30 min, resulting in a total dose of 3.4 J or a total radiant exposure of 5.4 J/cm2 to the cornea. Riboflavin drops and topical anesthetic were instilled every 5 min during the light exposure
Riboflavin + UV-A + propamidine isethionate (Brolene)	After the riboflavin and UV-A light exposure procedure, propamidine isethionate (Brolene) drops (5 μL) were instilled onto the cornea every hour for 6 hours for 14 days

Statistical Analyses

The mean value of all three measurements was calculated and compared among groups with the Kruskall-Wallis test. Bonferroni method was used to correct for multiple comparisons. A P value <0.05 was considered statistically significant. Wilcoxon test was used to evaluate pre- and posttreatment clinical scores. All analyses were performed with Stata version 10 program (Stata Corp., College Park, TX). 

Results

The in vitro experiment showed no difference among the groups: UV-A + B2 versus control 1 (blank) (P = 1), UV-A + B2 versus control 2 (UV-A light only) (P = 1), UV-A + B2 versus control 3 (B2 only) (P = 1), control 1 versus control 2 (P = 1), control 1 versus control 3 (P = 1), and control 2 versus control 3 (P = 1). Propamidine isethionate (Brolene) presented most of the cells dead, while the percentage of dead cells on the other groups was approximately 10%. Thus UV-A light and vitamin B2, alone and in combination did not demonstrate any antitrophozoite activity in this system (Figs. 1, 2). 

Figure 1.

 

Bright field photomicrographs (magnification, × 400) illustrating the effect of UV-A + B2 against monolayer cell culture of Acanthamoeba trophozoites. The antimicrobial spectrum between control (A, B) and experimental (C, D) groups were compared. Black and white arrows indicate live and dead protozoa, respectively. Trophozoites before (A) and after (B) antimicrobial assay with propamidine isethionate (Brolene). Trophozoites before (C) and after (D) antimicrobial assay with UV-A + B2. Different colors among pictures were generated by background staining due to white illumination incidence against specific compound in each aqueous solution assayed (i.e., Neff medium; slightly yellow, as seen in A, C); propamidine isethionate (Brolene) (slightly blue, as seen in B), and B2 (strongly yellow, as seen in D).

Figure 2.

 

Quantitative data of the in vitro experiments from control and experimental groups in the cell viability of Acanthamoeba trophozoites.

Acanthamoeba keratitis was induced in all right eyes. The results shown in Figure 3 indicate that clinical scores before treatment did not differ among the groups (P = 1). After treatment, two animal groups composed of propamidine isethionate (Brolene) instillation only and UV-A + B2 + propamidine isethionate (Brolene) presented clinical improvement. The other animal groups which were composed of UV-A + B2, UV-A only, B2 only, and blank presented higher clinical scores, and clinical worsening. After 14 days follow-up there was a significant improvement of clinical score for animals in the groups treated with propamidine isethionate (Brolene) alone and the combination of UV-A + B2 + propamidine isethionate (Brolene) (P = 0.0067). However there was no difference between these groups; the response of the group treated with the combination of UV-A + B2 + propamidine isethionate (Brolene) was not superior to that seen in the group treated with propamidine isethionate (Brolene) alone (Fig. 4). A significant worsening of clinical score was observed in all other groups: UV-A + B2 group (P = 0.0084), only UV-A (P = 0.0078), B2 only (P = 0.0084), and blank (P = 0.0082), with no difference among those (Fig. 3). Histologic examination after 14 days of treatment revealed mild disruption of stromal lamellae and fibrosis in the 2 groups with propamidine isethionate (Brolene). In the groups without propamidine isethionate (Brolene), disruption of stromal lamellae, neovascularization, and a dense inflammatory infiltrate was observed. Similar to the clinical scores, no difference among the nonpropamidine isethionate (Brolene) groups was noticed in the histology samples (Fig. 5). 

Figure 3.

 

Comparison of clinical scores between control and experimental before and after of the establishment of therapeutic patterns. Quantitative data were provided by clinical observation daily of corneal hamsters infected with Acanthamoeba trophozoites during 14 days. Lower clinical score equals better clinical status.

Figure 4.

 

Progression/regression analyses of keratitis induced by Acanthamoeba trophozoites in Chinese hamsters by in vivo assays. Scarified corneas of Chinese hamsters were infected with 10⁴ trophozoites/mL of Acanthamoeba strain and were photographed on Day 7 (before treatment, as shown in A1, B1, C1) and day 14 (after treatment, as shown in A2, B2, C2) post amoebal inoculation. (A) Propamidine isethionate (Brolene) instillation group, (B) riboflavin + UV-A + propamidine isethionate (Brolene) group, (C) riboflavin + UV-A group. Corneas from (A1), (B1), and (C1) showed clinical score four, while corneal outcomes posttreatment as shown in (A2), (B2), and (C2) indicated clinical scores 2, 2, and 13, respectively.

Figure 5.

 

Comparison of stromal changes caused by the UV-A + B2 in rabbit and Chinese hamsters by histopathologic analysis. Chinese hamster AK was induced by inoculation of 10⁴ trophozoites/mL. Corneal keratocyte density of both animals was cryosectioned 1 week after UV-A + B2 exposure and tissue sections were stained with hematoxylin-eosin. (A) Keratocyte apoptosis in the anterior stroma of rabbit cornea; (B) corneal aspect of Chinese hamster after UV-A + B2 exposure; (C) blank control of Chinese hamster cornea (no treatment); (D) AK in Chinese hamsters after treatment with UV-A + B2. Original magnification, ×100.

Discussion

Severe infectious keratitis cases have been reported to respond to combination UV-A light and riboflavin. Among the infectious agents reported in these case reports and small series are Mycobacterium chelonae, nontuberculous Mycobacterium species, filamentous fungus such as Acremonium sp. and Fusarium sp., ²⁴ E. coli, ²⁵ Moraxella lacunata, Haemophilus influenzae, S. aureus, S. epidermidis, ²⁶ Acanthamoeba spp, ¹⁵ and one case in which no specific pathogen was determined. ¹⁴ To the best of our knowledge this is the first report to assess the in vitro antitrophozoite activity of UV-A + B2, as well as the clinical treatment of AK in a hamster model. Recently, del Buey presented a poster showing that the UV-A + B2 therapy has no cysticidal effect (del Buey MA, et al. IOVS 2011;52:ARVO E-Abstract 1114). In this study the antitrophozoite effect of UV-A + B2 was preferred considering the high resistance of the Acanthamoeba cysts and that trophozoite is the amoeba infective form. ³  

Acanthamoeba keratitis has been characterized as a painful and vision-threatening disease. ²⁷ The infection cascade starts with adhesion of protozoa to the corneal surface and subsequent virulence processes involve the invasion and destruction of the corneal stroma. ^1,27 Therapy for AK includes application of chemical compounds, such as biguanides and diamidines, hourly for months. ^3,4 Although anterior segment complications related to use of antimicrobial agents have been reported in observational case series, amoebicidal activity of these compounds has improved infection during the management of AK. ^8,9,19,28 Due to the aggressive, prolonged, and not uniformly successful treatment currently employed, alternative therapeutic procedures have been proposed for AK. For this reason, the effect of UV-A light associated with a photosensitizer (riboflavin) in the viability of Acanthamoeba trophozoites was evaluated in this study. Although the successful combination of UV-A and riboflavin in therapy for AK has been described, ¹⁵ our data demonstrate resistance of this protozoa strain to the photochemical process in our in vitro and in vivo models. In vivo experiments showed a progressive clinical worsening after corneal cross-linking, similarly to the control groups. This agrees with a previous report by Rama et al., ²⁹ which showed the progressive ulceration and corneal melting by Acanthamoeba sp. in a patient previously submitted to a UV-A + riboflavin cross-linking procedure. 

The rationale of using UV-A and B2 for infectious keratitis is based on the riboflavin-based pathogen reduction technology (PRT) ^17,30 used to treat blood products. The infectivity of pathogens is reduced by three combined mechanisms. First, the direct damage of nucleic acids of pathogens with the UV light source; secondly, damage of pathogen nucleic acids, proteins, and membranes by reactive oxygen species generated when riboflavin absorbs light and interacts with dissolved oxygen in solution. ³¹ The third is the damage of pathogen nucleic acid by the interaction of riboflavin with nucleic acids. ³²  

Type I collagen is the main constituent of the corneal stromal matrix and the increase of collagen resistance against enzymatic degradation would theoretically be another possible advantage of using UV-A + B2 for infectious keratitis. ³³ In addition, collagenolytic enzymes with type I collagen degradation activity among Acanthamoeba strains have been suggested as a differential virulence factor and seemed to be correlated with the severity of clinical manifestation in patients. ³⁴ This study tested the hypothesis that riboflavin associated with UV-A followed by chemical treatment with propamidine isethionate (Brolene) could provide increasing resistance of collagen fibers against enzymatic activity of the protozoan, and a Chinese hamster was chosen as an animal model susceptible to amoeba infection after a previous protocol described by van Klink and collaborators. ²³  

The groups treated with propamidine isethionate (Brolene) had significant clinical improvement. The group treated with UV-A + B2 and propamidine isethionate (Brolene) did not present better clinical results compared with the group treated with propamidine isethionate (Brolene) alone. In contrast with a previous report, ³⁵ the histologic changes caused by UV-A + B2 described for rabbit corneas like apoptotic damage of keratocytes throughout the stroma were not observed. Also the inflammatory cell infiltration and neovascularization do not differ from the group propamidine isethionate (Brolene) only. 

To confirm the stromal changes caused by the UV-A + B2, we treated one extra rabbit and three Chinese hamsters after the protocols recommended for keratoconus. ¹⁶ Corneas were harvested 10 days after the exposure to UV-A + B2. As shown in Figure 5, surprisingly the stromal histology changes described for rabbits like keratocyte apoptosis down to 200 μm was observed in the rabbit cornea, but not for the Chinese hamster keratocytes. One possibility is therefore that the UV-A + B2 + propamidine isethionate (Brolene) group did not experience better results because the Chinese hamster corneal stroma does not respond with a strengthening of the collagen matrix as do rabbit and human corneas. 

The lack of in vitro activity against trophozoites of UV-A + B2 may relate to the different parameters used for riboflavin-based pathogen reduction technology (PRT) compared with those used for keratoconus. In PRT, the illuminator delivers 6.2 J per mL and the wavelength output ranging from 265 to 370 nm. ³⁶ Safety studies have recommended for keratoconus, a wavelength of 370 ± 5 nm and total irradiance dose of 5.4 J. ³⁷ Thus not only the total energy is higher in PRT but also the wavelength is shorter, likely causing more damage to the cells and DNA. ³⁷  

In conclusion, the UV-A + B2 technology does not present antitrophozoite activity in vitro or in vivo when applied using the parameters used in the present study. 

Footnotes

 Disclosure: R.T. Kashiwabuchi, None; F.R.S. Carvalho, None; Y.A. Khan, None; D. de Freitas, None; A.S. Foronda, None; F.E. Hirai, None; M.S. Campos, None; P.J. McDonnell, None