Water samples were collected in 2009 and in 2010 from the inlets and fish tanks of 22 independent Swiss fish farms. Inlet water flew directly from the river into separate tanks; the water volume ranged from 2 to 105 m3. The water flow was continuous. The detailed sampling structure is described in Table 2.
During 2009, water and different fish species were sampled every second week in 4 fish farms located in the Ticino Canton (Switzerland) (60 sampling actions).
In 2010, sampling was carried out in 22 fish farms all over Switzerland at 3 different periods (85 sampling actions). The first was in winter shortly before fishes started hatching (only water), the second was carried out 6 and the third 12 weeks after hatching and when fishes started feeding. At each sampling date six fishes [Rainbow trout (Oncorhinchus mykiss) and brown trout (Salmo trutta fario and Salmo trutta lacustris)] were collected concomitantly with two water samples, one from the inlet and one from the tank. 50 ml of water were collected in 50 ml Falcon tubes (Becton Dickinson BD, Switzerland), while fishes were collected in a container with water and brought back to the laboratory within 24 h after collection in refrigerating bags. Plating and fixation of water samples were carried out immediately upon arrival in the laboratory.
Population density of fishes in the tanks, physical (temperature, water conductibility, oxygen saturation, water volume) and chemical (disinfectant and antibiotic use) water parameters were recorded directly at the fish farm.
In the laboratory, 100 μl of water collected were plated on Cytophaga enriched Agar Medium (CAM, medium 1133 DSMZ: 0.2% tryptone, 0.05% beef extract, 0.05% yeast extract, 0.02% sodium acetate, 1.5% agar). All plates were incubated at 15°C during 5 to 10 days. Yellow colonies (i.e. putative flavobacteria) were transferred onto fresh plates and screened with a Flavobacterium spp. and F. psychrophilum specific FISH . Pure cultures of Flavobacterium spp. and F. psychrophilum were conserved at −80°C in 1 ml skimmed milk (Becton Dickinson, Switzerland) supplemented with 10% bovine serum and 20% glycerol.
Fixation of water samples was carried out according to Tonolla et al.  with the following modifications: 15 ml of each water sample were filtered with a Millipore filtration system (Merck Millipore) with 3.0 μm mesh size filters overlaid with 0.2 μm mesh size filters. Each sample was covered with 4% Paraformaldehyde Fixation Buffer (PBS: 0.13 M NaCl, 7 mM Na2-HPO4, 3 mM NaH2PO4, pH 7.2) for 30 min and then washed twice with 1× Phosphate Buffered Saline (PBS). The overlay filters were transferred into plastic bags; 600 μl of a 50% PBS-ethanol solution were added, the bags sealed and bacteria re-suspended by slightly rubbing the filter between thumb and forefinger. The suspension was then transferred into a 1.5 ml Eppendorf tube and stored at −20°C until DNA extraction. The DNeasy Blood & Tissue Kit (QIAGEN - Switzerland) was used for DNA extraction of all fixed water samples.
For pathogen detection in animals, fish collected were killed by immersion in 0.01% benzocaine followed by section of the vertebral column. Spleen of rainbow trout, brown trout fario and brown trout lacustris were homogenized separately in 200 μl of sterile water. 190 μl of the homogenates were plated on CAM medium and incubated at 15°C for 5 to 10 days while the remaining 10 μl were used for FISH .
Approval for animal experiments and water collection was obtained from the Federal Veterinary Office (FVO, Switzerland) and the Ticino Cantonal Veterinary Office (Authorization 03/2010 and 04/2010).
Identification of colonies and diagnosis of outbreaks by FISH
Identification of flavobacteria in general and F. psychrophilum in particular was carried out using a published FlSH protocol . F. psychrophilum (DSM 3660), environmental Flavobacterium spp. and Chryseobacterium spp. isolates were used as positive and negative controls.
rpoC qPCR design and test of primers
DNA was extracted using InstaGene kit [Bio-Rad, Hercules (CA), USA]. Partial DNA dependent β’ subunit RNA polymerase (rpoC) gene sequences were amplified based on the RNA polymerase β’ subunit primers sequences described by Griffiths et al.  with the addition of sequence tags UP1s and UP2sr (rpoC_F 5’- GAAGTCATCATGACCGTTCTGCAATHGGNGARCCNGGNACNCA-3’ and rpoC_R 5’- AGCAGGGTACGGATGTGCGAGCCGGNARNCCNCCNGTDATRTC-3’; synthesized by Microsynth, Switzerland) to increase sequencing performance . The PCR reaction was carried out in a total volume of 50 μl using 2.5 U HotStarTaq DNA Polymerase (QIAGEN-Switzerland), 7 mM MgCl2, PCR Buffer 1X (QIAGEN-Switzerland), 0.2 mM dNTP (Roche, Switzerland), 0.2 μM of each forward and reverse primer, and 5 μl of InstaGene DNA extract. The thermal cycle started with 15 min HotStarTaq activation at 95°C followed by 36 cycles of 1 min at 94°C, 90 s at 55°C, 1 min at 72°C and eventually an elongation cycle of 7 min at 72°C.
Sequences (GenBank access numbers JX657163- JX657284) obtained from the rpoC gene general PCR were aligned using MEGA4  and screened for a conserved species-specific fragment that would be used to design a set of primers and a TaqMan probe targeting specifically F. psychrophilum. Primers F.psychro_P1F 5’-GAAGATGGAGAAGGTAATTTAGTTGATATT-3’, F. psychro_P1R 5’- CAAATAACATCTCCTTTTTCTACAACTTGA-3’ and a minor groove binder (MGB), and probe F. psychrophilum_probe 5’- AAACGGGTATTC TTCTTGCTACA -3’ (Applied Biosystems) labeled with FAM were tested in silico and with BLAST (Basic local alignment search tool ). The primers amplified a fragment of 164 bp. PCR was carried out in a final volume of 25 μl containing 1X Taq PCR Master Mix Kit (QIAGEN, Switzerland), 0.3 μM primers F. psychro_P1F and F. psychro_P1R, and 2.5 μl of genomic DNA. Conditions for amplification were 94°C for 1 min followed by 35 cycles of 94°C for 30 s, 56°C for 35 s and 72°C for 30 s, with a final elongation cycle of 7 min at 72°C.
DNA of F. psychrophilum, Flavobacterium spp. and other bacterial species isolated from soil, water and fish were used to test sensitivity and specificity of the primers. All tested bacteria and their origin are listed in Table 1.
qPCR cycling parameters
The qPCR was carried out in a final volume of 20 μl containing 1× TaqMan Environmental Master Mix v.2.0 (Applied Biosystems), 0.9 μM of each primer, 0.2 μM of F. psychrophilum probe, 1X of internal control Exo IPC Mix, 1× of IC DNA (TaqMan Univ. MMix w Exog IntPostC, Applied Biosystems), and 2 μl of template DNA. An internal control was added to each reaction to check for PCR inhibitors. The run consisted of two cycles at 50°C for 2 min and 95°C for 10 min, followed by 40 cycles at 95°C for 15 s and 60°C for 1 min. All assays were carried out in triplicates. Water was used as negative control and series of quantified DNA dilutions as standards.
Preparation of standards
F. psychrophilum DNA was amplified by PCR with primers F. psychroP1F and F.psychroP1R. The products were purified with PCR clean-up NucleoSpin® ExtractII (Macherey-Nagel, Germany) and quantified with a Nanodrop spectrophotometer (ND1000, Witek, Switzerland). The total amount of DNA measured was divided by 1.797 × 10−7 pg [the weight of one rpoC fragment (164 bp) [54–56]]. The result was an estimate of the number of gene copies in 1 μl of purified product. Serial dilutions from 1 × 107 to 1 × 100 copies/μl of amplified DNA were used to calculate the Limit of Detection (LOD) of the qPCR and as quantitative standards for further analyses.
Serial 10-fold dilutions were made starting from F. psychrophilum suspensions [Optical Density (OD595) 0.3 ± 0.02] corresponding to (3 × 109) ± (7 × 108) cells/ml . Each suspension was extracted with DNeasy Blood & Tissue Kit (QIAGEN - Switzerland) and used to determine the quantification limit (QL).
Limit of detection and quantification limit
Calibration curves were obtained by plotting cycle time (Ct) values against log10 (gene copies number). The coefficients of regressions as well as the R2 values were calculated. The LOD was calculated using a serial dilution from 2 × 107 to 2 × 100 amplified fragments per reaction of 20 F. psychrophilum amplified DNA standards.
Suspensions of 24 F. psychrophilum isolates (serial dilutions from 2 × 104 to 2 × 10−1 cells per reaction) were analyzed to determine the QL. Genomic DNA standards from bacteria suspensions were used to check the reliability of the quantification.
qPCR specificity and potential cross-amplifications with other Flavobacterium spp. were checked using dilutions of DNA extracted from F. branchiophilum (concentrations: 106 and 105 cells per reaction), F. columnare, F. johnsoniae, F. psychrolimnae, F. fryxellicola, Flavobacterium sp. and Chryseobacterium sp. (concentrations: 107 and 106 cells per reaction). In addition, diluted DNA samples of F. columnare (107, 104, 103, 102 cells per reaction) and F. branchiophilum (106, 104, 103, 102 cells per reaction) were mixed with decreasing concentrations of F. psychrophilum DNA (from 106 to 103 cells per reaction).
For the results of the qPCR to be reliable, the coefficient of the standards regression had to be in the range −3.6 – -3.0 (Applied Biosystems, manufacturer’s instructions for qPCR), the coefficient of variation of quantification within each standard and sample in triplicates <25% and the non target control (water) had to show no amplification within the run [54, 57].
qPCR of spleen samples
Spleens of diseased and symptomless rainbow trout and brown trout were gathered during 2011 and 2012 in the Ticino fish farms and treated as described before. Fish were considered healthy when they showed no disease symptoms and, additionally, no signs of infection or extraordinary mortality were reported in the fish farm.
In total 15 rainbow and brown trout spleens were collected and analyzed during 4 outbreaks while 43 spleens from symptomless fish (rainbow and brown trout) were collected in 2 different fish farms showing no sign of infection.
Spleens from symptomless fish were removed, weight calibrates and stored at −20°C until further processing. Mean spleen weight was 0.013 ± 0.007 g for rainbow trout and 0.007 ± 0.002 g for brown trout.
At the time of the experiments, spleens from healthy fishes were thawed and homogenized in 200 μl of sterile water. 100 μl of the suspension were spiked with known amounts of F. psychrophilum (106 to 101 cells per reaction) to a final volume of 100 μl and extracted using DNeasy Blood & Tissue Kit (QIAGEN). The remaining 100 μl were used as controls in FISH and DNA extraction for F. psychrophilum qPCR screening and quantification purpose.
Spleens from diseased fish were used to quantify levels of infection under real-life conditions. They were removed and homogenized in 200 μl of sterile water. It was, however, not possible to weight them. 90 μl of the spleen homogenates were plated on CAM and incubated at 15°C for 5 to 10 days while 10 μl were analysed using FISH with the PanFlavo and F. psychrophilum probes . DNA was extracted from the remaining 100 μl.
Primer specificity (SP) and sensitivity (SE) as well as positive and negative predicted values were assessed by standard PCR. The efficiency of qPCR was calculated as E = 10-1/slope-1. A linear regression was used to calculate the LOD and the QL at the fifth percentile of all analyzed samples correctly detected (LOD) or quantified (QL) by the technique using SPSS Statistics for Windows, Version 20.0 (IBM Corp., Armonk, NY).