Detection of Tilapia Lake Virus in Egyptian fish farms experiencing high mortalities in 2015

Institute of Veterinary Bacteriology, University of Bern, Bern, Switzerland WorldFish, Abbassa Research Center, Sharkia, Egypt National Institute of Oceanography and Fisheries, Cairo, Egypt International Livestock Research Institute, Nairobi, Kenya WorldFish, Bayan Lepas, Penang, Malaysia Center for Infection and Immunity, Mailman School of Public Health, Columbia University, New York, NY, USA Institute of Geography, University of Bern, Bern, Switzerland International Livestock Research Institute, Addis Ababa, Ethiopia

Since 2013, an increasing number of tilapia farms in Egypt have been experiencing high summer morbidity and mortality rates (WorldFish 2015). Such disease outbreaks are a major constraint to the aquaculture trade and have devastating economic and social consequences (Aly, 2013). Tilapia have become such an important global food source, not only because they are an inexpensive source of protein but also due to their favourable culturing characteristics such as, their mode of reproduction, omnivorous diet, fast growth, tolerance for high-density aquaculture and because they are relatively resistant to poor water quality and disease (FAO 2005). Despite this, various bacteria, fungi, protozoa and viruses have been associated with disease in tilapia aquaculture (FAO 2005). There have been many reports implicating bacteria in tilapia diseases and in particular, bacterial infections caused by Aeromonas or Streptococcus species.
Our aim was to detect potential pathogens in diseased tilapia that might account for the high summer mortalities observed in Egyptian fish farms. To this end, diseased fish were examined from eight farms in different areas of the Nile delta. Here, we report the detection of Aeromonas species and, for the first time, the presence of TiLV in Egyptian tilapia aquaculture.
In September 2015, we visited eight commercial farms experiencing the so-called summer mortality spanning a large area of the Nile Delta ( Figure 1). We randomly sampled 13-40 fish per farm and examined them macroscopically to determine disease prevalence (Table 1 and Table S1). External clinical examinations considered haemorrhagic patches, detached scales, open wounds, dark discoloration and fin rot as signs of disease. Two diseased fish per farm were dissected, and tissues from the head kidney were collected in Universal Transport Medium tubes (UTM TM , Copan, Italy) and kept at 4°C during transportation before being stored at À80°C. Additionally, the head kidney, spleen and liver tissues were swabbed using Amies Agar Gel (with charcoal) swabs (Copan, Italy). As controls, 20 healthy tilapia were examined from WorldFish in Abbassa and two were dissected and processed as outlined above.
The presence of TiLV was tested using a PCR protocol published elsewhere (Eyngor et al., 2014). Briefly, 20-60 mg of head kidney tissue was homogenized in 1 ml TRIzol reagent (Thermo Fisher Scientific, USA) using Lysing Matrix B and D (MP Biomedicals, USA,) in a MP FastPrep-24 sample preparation system (MP Biomedicals) for 2 9 40 s, with cooling in between, at a speed of 6.5 m/s. Thereafter, the TRIzol reagent protocol was followed, with the exception that 1-   et al., 2014), while the Egyptian TiLV isolates showed only 93% identity to the Israeli strain when comparing sequences from segments 3, 4 and 9. We built phylogenetic trees for each segment using phyML (Guindon et al., 2010). All trees were congruent pointing towards the absence of reassortment (Figure 2 (Bizzini, Durussel, Bille, Greub, & Prod'hom, 2010). We did not detect any Streptococcus species, but we found Aeromonas species in every farm (Table 1).
Our aim was to identify candidate pathogens potentially responsible for the recent summer mortalities reported from Egyptian tilapia aquaculture (Worldfish, 2015). We visited eight commercial farms, covering different regions of the Nile delta and observed morbidities ranging from 43% to 100%. In every farm, we identified Aeromonas species which have been detected in Egyptian tilapia farming for many decades and have been linked to previous severe disease outbreaks (Aly, 2013). The new finding in trying to account for the recent Egyptian tilapia mortalities is the discovery of TiLV in 50% of the farms investigated. Notably, the application of a recently published Nested Reverse Transcription-PCR (Kembou Tsofack et al.,

2017) probably would have resulted in a higher detection rate of
TiLV. So far, TiLV has been implicated in mass tilapia deaths occurring also during the hot seasons in Israel and Ecuador. It is a segmented negative sense RNA virus that appears to be a real threat to global tilapia aquaculture. The absence of re-assortment and the closer similarity of Israeli to Ecuadorian isolates than to geographically closer Egyptian isolates could be due to anthropogenic influence via movement of feeds, live fish or water. Given our data, we cannot conclusively state that the emergence of this virus in Egyptian tilapia aquaculture is solely responsible for their summer mortalities, but it

(a)
F I G U R E 2 Unrooted phylogenetic trees displaying the relationship between the Egyptian, Israeli and Ecuadorian TiLV strains. The trees are based on a 249-bp gene sequence of segment 3 (a), a 263-bp gene sequence of segment 4 (b) and a 252-bp gene sequence of segment 9 (c). It was generated using phyML 3.0 under the HKY model of nucleotide substitution (Guindon et al., 2010). Bootstrap values are for 1,000 bootstraps is a significant finding. Therefore, a prompt investigation into the prevalence of this virus in further regions of the Nile delta and how the Egyptian TiLV, which is divergent from the Ecuadorian and Israeli strain, manifests itself in Nile tilapia is required. Additional Future studies should investigate the effect of co-infections of pathogenic Aeromonas species with TiLV. Moreover, abiotic factors, such as water quality and temperature, should be investigated with respect to their effect on tilapia susceptibility to TiLV. Crucially, future work must focus on implementing existing diagnostic methods (Kembou Tsofack et al., 2017) into disease control policies. The development and application of an efficient vaccine would be the most effective disease control measure, but this endeavour requires more time.