Repertoire of free-living protozoa in contact lens solutions

CL solution specimens were collected between 2009 and 2014 by CL wearers presenting to the Ophthalmology Department of the Timone Hospital in Marseille, France, for the clinical diagnosis of keratitis and corneal ulcers. Clinical criteria for diagnosis included evidence of a corneal infiltrate or corneal ulcer with underlying inflammation, which could lead to the necrosis of corneal tissue. CL solution provided by the patient was poured into a sterile can kept at room temperature for 4–24 h before it was analysed in the laboratory. The following standard protocol was used to search for protozoa. The CL solution was spread onto a non-nutrient agar plate, plated with a lawn of living Enterobacter aerogenes. The non-nutrient agar plate was incubated at 28 °C in a humidified atmosphere (contact with moistened gauze) and examined by microscope at × 4 and × 10 magnifications. Free-living protozoa were subcultured on a new non-nutrient agar plate with living E. aerogenes in order to obtain sufficient clonal populations. When the growth was sufficient, areas where protozoa were easily detected by microscope were cut and centrifuged at 2000 g for 10 min. The pellet was re-suspended in 1 mL of Amoeba Page’s saline PAS (Dunstaffnage Marine Laboratory, Oban, UK) for further DNA extraction.

CL solution specimens were seeded onto 5 % sheep-blood agar (COS, bioMérieux, La-Balme-les-Grottes, France) and BCYE (Buffered Charcoal Yeast Extract, bioMérieux) and incubated at 32 °C for 10 days in a 5 % CO₂ atmosphere. For the culture of yeasts and fungi, CL solution specimens were seeded onto Sabouraud agar containing chloramphenicol and gentamicin (bioMérieux), incubated at 32 °C for 10 days. All the bacterial isolates were identified using matrix-assisted laser desorption ionization-time of flight mass spectrometry (MALDI-TOF-MS; Microflex, Bruker Biospin S.A., Wissembourg, France) as previously described [6]. Briefly, colonies detached from the agar were directly applied to a MALDI-TOF MTP 384 target plate (Bruker) in order to analyze four spots per isolate. Each spot was overlaid with 2 μL of matrix solution, a saturated solution of α-cyano-4-hydroxycinnamic acid in 50 % acetonitrile mixed with 2.5 % tri-fluoracetic-acid. The matrix-sample was crystallized by air-drying at room temperature for 5 min. Measurements were performed using an Autoflex II mass spectrometer (Bruker Daltonik) equipped with a 337-nm nitrogen laser. Spectra were recorded in the 2–20 kDa mass range. Data were automatically acquired using AutoXecute acquisition control software. The two first raw spectra obtained for each isolate were imported into the BioTyper software, version 2.0 (Bruker Daltonik GmbH), and were analyzed by standard pattern matching (with default parameter settings) against 5625 references in the BioTyper database. When both spots yielded a score ⩾1.9, identification was complete. In this study, it was not necessary to complete accurate MALDI-TOF-MS identification of bacteria by DNA sequencing.

Total DNA was extracted using the QIAmp tissue kit according to the manufacturer’s protocol (QIAGEN SA, Courtaboeuf, France). A 328-bp fragment of the 18S rRNA gene was PCR-amplified using the primers NS5/F 5′AACTTAAAGGAATTGACGGAAG3′ and NS6/R 3′GCATCACAGACCTGTTGCCTC5′ and an annealing temperature of 60 °C [7]. All amplification reactions were performed using the 2720 thermal cycler (Applied Biosystems, Saint-Aubin, France) in a 50 μL-mixture containing 5 μL of dNTPs (2 mM of each nucleotide), 5 μL of DNA polymerase buffer (Qiagen), 2 μL of MgCL2 (25 mM), 0.25 μL HotStarTaq DNA polymerase (1.25 U) (Qiagen), 1 μL of each primer and 35.75 μL of DNAse-free water. The positive control consisted of Candida albicans DNA. Sterile distilled water was used as a negative control. PCR consisted of a 15-min initial denaturation Taq polymerase Hot-Star at 95 °C followed by 30-s denaturation at 95 °C, 30-s hybridation at 60 °C and 1-min elongation at 72 °C. After 35 cycles, extension was performed for 5 min at 72 °C. Amplified products were visualized under UV illumination with Syber Safe ® staining after electrophoresis using a 1.5 % agarose gel. PCR products were cloned by the pGEM® -T Easy Vector System Kit according to the manufacturer’s instructions (Promega, Lyon, France). They were sequenced in both directions using the Big Dye® Terminator V1.1 Cycle Sequencing Kit (Applied Biosystems). Original sequences have been submitted to GenBank.

Sequencing products were resolved using an ABI PRISM 3130 automated sequencer (Applied Biosystems). Sequences were compared with the GenBank database using the online BLAST program (www.​ncbi.​nlm.​nih.​gov). The highest percentage of sequence similarity was used to identify isolates. Sequence similarity higher than 97 % with a described species was considered to be indicative of identification at the species level. Phylogenetic analysis was established by the neighbor-joining method using MEGA5 software (www.​megasoftware.​net). Phylogenetic construct was based on the 18S rRNA gene sequences aligned with 52 references.

A total of 20/233 (8.6 %) CL solution specimens collected between 2009 and 2014 from 16 patients, cultured at least one free-living protozoa (Table 1). Protozoa identifications were made by partial sequencing of the 18S rRNA gene and by establishing the percentage of similarity of these sequences with reference sequences. authenticated by the validity of positive and negative controls. With one exception, confident identification was obtained at the genus level only. These identifications include Acanthamoeba in 16 (80 %) solution specimens collected from 12 different patients, Colpoda steini in specimen n°14, Cercozoa sp. in specimen n°12, Protostelium sp. in specimen n°15, and an identical 99 % sequence similarity with both Hartmanella and Vermamoeba genus in specimen 13.

Further phylogenetic analysis (Fig. 1) confirmed these identifications and indicated that the protozoa isolated in specimen n°13 was more closely related to Vermamoeba. Furthermore, phylogenetic analysis indicated that the same Acanthamoeba was isolated in left and right contact lens solutions in patients 6, 11 and 16.

Twelve of the 20 protozoa-positive (60 %) CL specimens cultured bacteria, while eight protozoa-positive CL specimens did not. Stenotrophomonas sp. and Pseudomonas sp. were most frequently identified and found in 8/20 (40 %) specimens, followed by Klebsiella sp. in 4/20 (20 %) specimens, Aeromonas sp. in 3/20 (15 %) specimens, Chryseobacterium sp. and Sphingobacterium sp. in 2/20 (10 %) specimens and Achromobacter sp., Agrobacterium sp., Alcaligenes sp., Citrobacter sp., Delftia sp., Enterobacter sp., Microbacterium sp., Mycobacterium sp., Raoultella sp., Serratia sp., Shewanella sp. and Wautersiella sp. in 1/20 specimens. Fungi were cultured in five protozoa-positive CL specimens. Fungi included Candida guilliermondii, Candida parapsilosis, Candida lipolytica and Candida colliculosa, Fusarium oxyoprum, Sacrocadium kiliense and Penicillium chrysogenum. In three cases, several fungi were co-cultured, including P. chrysogenum, C. parapsilosis and F. oxysporum in case 16–1, C. guilliermondii and F. oxyoprum in case 7 and C. parapsilosis and C. lipolytica in case 10.