Francisellosis in fish: an emerging challenge

Francisellosis is a bacterial disease with increasing economic impacts in the culture of tilapia and Atlantic cod since emerging in 1992. Two main strains – Francisella noatunensis subsp. orientalis (Fno) and F. noatunensis subsp. noatunensis (Fnn), have been identified, causing both acute and chronic granulomatous systemic disease. The piscine host range is increasing and Francisella culture should be included in routine diagnosis. Differentiation from the major zoonotic F. tularensis and opportunistic zoonotic F. philomiragia when dealing with environmental soil and water samples from fish farms is important. Diagnosis can be challenging but presentation of granulomatous pathology in fish should require use of cysteine supplemented selective media, culture at 15–28°C or culture in fish cell lines and specific PCR to exclude piscine Fno or Fnn. Control of infections in fish rely on appropriate antibiotic selection although in the long term an effective commercial vaccine that includes the pathogenic species of Francisella is required.

Tilapia (Oreochromis niloticus) production has increased from 2.6 million tons in 2005 to an estimated 4.5 million tons in 2014¹, being only second to carp in global aquaculture production². Atlantic cod (Gadus morhua) production peaked at 22.7 tons in 2009 and dropped to 4.3 tons in 2013³. Francisellosis has been reported in farmed tilapia from Taiwan (1992), United States (2003), Costa Rica (2009), Indonesia (2009), United Kingdom (2010), and Brazil (2012)⁴ with mortalities of 30–75%^5,6 and up to 95%⁷. During 2004–2005, outbreaks in farmed Atlantic cod in Norway resulted in approximately 40% losses, presenting a major impediment to the expansion of cod aquaculture⁷. Initially thought to be a Rickettsia-like organism^5,6,8 or Piscirickettsia-like organism^8,9, the pathogen was later confirmed as a γ-Proteobacteria in the family Francisellaceae, order Thiotrichales⁷. Francisella noatunensis subsp. orientalis (Fno) causes francisellosis in tilapia (a fresh and warm water fish species) and F. noatunensis subsp. noatunensis (Fnn) in Atlantic cod^7,8 (a marine and cold water fish species). F. philomiragia subsp. noatunensis subsp. nov. and F. piscicida were two different names proposed for the organisms isolated from Atlantic cod in Norwegian disease outbreaks, however it has been resolved to be synonymous with Fnn^7,10. Infections associated with F. philomiragia/Fnn in Atlantic salmon (Salmo salar), Francisella spp. in three-line grunt (Parapristipoma trilineatum) and ornamental cichlid species are reported⁷. Fno infected hybrid striped bass (Morone chrysops × M. saxatilis)⁸ and Francisella halioticida¹¹ infected the giant abalone (Haliotis gigantea)⁷. Recently, disease in marine ornamental fish species (wrasses and damselfish) was associated with Fno¹².

The gross pathology is typified by visceral granulomatosis causing splenomegaly and renomegaly due to multiple whitish-tan nodules with similar lesions in liver, gills or muscle. The degree and range of organ involvement differ between species. In Atlantic cod, emaciation, haemorrhagic skin and heart nodules also occur while in tilapia, gills can have the nodules in addition to exopthalmia and skin haemorrhages and scale loss^7,8. Histopathology in affected organs feature granulomas consisting of vacuolated macrophages with the Francisella organisms, associated central necrosis and fibrous encapsulation^{5,7–9,12,13}, in the sub-acute (7 days post challenge⁵) to chronic disease. Acute disease has been experimentally replicated causing 100% mortality by 72 h post intraperitoneal inoculation of approximately 10⁷ colony-forming unit (cfu) per fish where bloody ascites, increased melanomacrophage centres but no granulomas were observed⁶. For tilapia, epizootics typically occur in cooler, winter water temperatures with higher mortalities at 15°C than 30°C⁷ or no mortalities at 26.5–29.2°C⁸. Francisellosis causes more mortalities as water temperatures increase towards 20°C in summer for Atlantic cod⁷. Epidemiologically, piscine Francisellae cause disease in both fresh and marine waters and morbidity can be extremely elevated for Atlantic cod and tilapia⁸. Fish pathogenic Francisella can enter a viable but non-culturable state in cold water after 30 days at 8°C and 16 days at 12°C⁸, meaning that they are non-virulent. A reservoir of the F. philomiragia in the aquatic protozoan Acanthamoeba castellanii and the aquatic biofilm has been reported¹⁴ with implications of transmission to fish.

Francisella are 0.1–1.5 µm, strictly aerobic, facultatively intracellular, non-motile, Gram-negative coccobacilli⁷ to pleomorphic spherical^6,9,15,16, halophilic¹⁵ or freshwater⁶ organisms. Culture of Fno and Fnn from kidney, spleen, blood or granulomatous lesions is made on enriched blood agar plates supplemented with 0.1% cysteine and 1% glucose, cysteine heart agar with 1% haemoglobin (CHAH) or cysteine heart agar with 5% sheep blood (CHAB) or Thayer-Martin media^6–9. The organism fails to grow on trypticase soy agar (TSA) supplemented with 5% sheep blood⁶ and can be easily overgrown with or inhibited by contaminant or secondary bacteria^6,8,9. Polymixin B (100 U/mL) and/or ampicillin (50 µg/mL) maybe added to reduce these bacteria⁶. Incubation temperature is 15–20°C for Fnn and 25°C for Fno⁹ with Fnn growing poorly at 30°C and Fno preferring 28°C^4,8. Colonies develop slowly, taking up to 30 days but may appear as smooth, white to greyish within 3–6 days⁹. Differentiation can be made from the zoonotic F. tularensis and F. philomiragia in that these organisms can grow at 35–37°C while Fnn and Fno do not⁸. Further, F. philomiragia does not have an essential requirement for cysteine to grow^7,17. Biochemical reactions for Fno and Fnn are the same, with negative reactions for cytochrome oxidase activity, acid production from sucrose, β-galactosidase and no enzymatic activity for O-nitrophenyl N-acetyl-β-D-glucosamide (ONAG), P-nitrophenyl-β-D-galactopyranoside (PNPG), leucine arylamidase, and N-acetyl-β-glucosaminidase⁷. However, Fnn metabolises D-glucose but does not have indoxyl phosphate (IDP)^7,18 activity, while Fno is the reverse for these tests⁷. Molecular testing based on the G1,L1 primers targeting the internal transcribed sequence (ITS) with Eub A and Eub B primers targeting 16S rRNA, followed by sequence homology analysis is able to differentiate Fno (in tilapia and three-line grunt) from Fnn (Altlantic cod and Atlantic salmon)^7,18. Of note, Fnn shows 99.3% and Fno shows 98.6% 16S rRNA similarity to F. philomiragia, but they are more genetically dissimilar to F. tularensis^7,8,16. Cell culture isolation has been demonstrated for Fno using chinook salmon embryo (CHSE-214⁵) and tilapia ovary cells (TO)⁹. Similarly Fnn can be grown using salmon head kidney (SKK-1) and Atlantic salmon kidney (ASK)⁸. Serological testing using antiserum raised against Fnn detects Fno as well, with F. philomiragia agglutinating slightly to the Fnn antiserum⁷ but there was no cross reaction with monoclonal antibody against F. tularensis.

In terms of zoonotic risk, F. tularensis is a major environmental and tick or insect vector-borne human pathogen causing pneumonic tularemia^6,19, with F. tularensis subsp. tularensis being the most virulent strain and of biological weapon concern^6,7. Recently, tularemia in Turkey has been associated with beaver, muskrat and voles which infect surface waters suggesting that the aquatic environment is an important risk factor in its epidemiology^20,21. F. philomiragia is a rare disease and is associated with immune-compromised patients as in chronic granulomatous disease (CGD) and in near drowning events causing pneumonia or fever-bacteraemia^13,15,17. F. philomiragia has been isolated from brackish water in an area where repeated tularemia cases occurred¹⁹. Therefore, it may be prudent to consider that zoonotic species of Francisella could be transmitted through the aquatic environment when dealing with aquatic environmental samples, including those from fish farms. To date, Fno and Fnn are considered to have negligible zoonotic risk as they cannot grow at 37°C and for Fno in tilapia, there has been no documented case of human infection despite it being a major aquaculture product processed for human consumption^7,8.

Control of clinical infections of francisellosis in tilapia has been reported with 30–50 mg/kg oxytetracycline over a 10–14 day treatment, but the high infectivity, a low infective concentration, high morbidity and inappetance in severely infected fish may render sustainable management ineffective⁸. Isolates may be resistant to trimethoprim-sulfamethoxazole, penicillin, ampicillin, cefuroxime and erythromycin, gentamicin and ciprofloxacin⁷. Florfenicol at 15 mg/kg has been demonstrated experimentally to improve survival to challenge with Fno, and it is suggested that this antibiotic could penetrate intracellularly to clear the organism⁷. To date there is no commercial vaccine for piscine francisellosis although development work based on attenuation of Fno by mutation of the iglC* gene provided effective protection in tilapia^7,8. Formalin-killed Fno bacterin with a mineral oil adjuvant provided a relative percentage survival (RPS) value in tilapia of 100% at day 27 post intraperitoneal challenge, with a specific antibody response at 15, 30 and 45 days post vaccination¹.

There are a number of key issues with piscine francisellosis:

• improving the efficiency of definitive diagnosis to mitigate the inadvertent dissemination of infected carrier fish hosts. This will require veterinary pathologists and microbiologists to be up-to-date regarding the case presentation of the disease. As a standard approach, fish with granulomatous disease should be subject to Francisella sp. exclusion, as part of the differential diagnoses.

• research into the epidemiology (in particular the diversity of reservoir host species) and virulent factors or genes of Fno and Fnn as part of the process for development of commercial vaccine products. This is important as warm water and cold water francisellosis are likely to present different scenarios in terms of disease management.

• finally, regarding the zoonotic risk of F. tularensis and F. philomiragia with these being isolated also from aquatic environments^15,19, bacteriological culture conditions to exclude these zoonotic Francisellae from fish samples is an important exercise. This will avoid inadvertent human infection from aquatic or aquaculture environments.