Photobacteriosis: Prevention and Diagnosis

Photobacteriosis or fish pasteurellosis is a bacterial disease affecting wild and farm fish. Its etiological agent, the gram negative bacterium Photobacterium damselae subsp. piscicida, is responsible for important economic losses in cultured fish worldwide, in particular in Mediterranean countries and Japan. Efforts have been focused on gaining a better understanding of the biology of the pathogenic microorganism and its natural hosts with the aim of developing effective vaccination strategies and diagnostic tools to control the disease. Conventional vaccinology has thus far yielded unsatisfactory results, and recombinant technology has been applied to identify new antigen candidates for the development of subunit vaccines. Furthermore, molecular methods represent an improvement over classical microbiological techniques for the identification of P. damselae subsp. piscicida and the diagnosis of the disease. The complete sequencing, annotation, and analysis of the pathogen genome will provide insights into the pathogen laying the groundwork for the development of vaccines and diagnostic methods.

Differences in susceptibility to the disease have been described on the basis of fish age. Larvae and juveniles are more susceptible to photobacteriosis, and acute infection induces 90-100% mortality of juvenile sea bream, whereas fish over 50 g are more resistant due to more efficient phagocytosis and killing of the bacteria by neutrophils and macrophages [6]. Bacteria that reside in different tissues and inside phagocytes cause chronic and acute forms of photobacteriosis. In its acute form, multifocal necrosis is present in the liver, spleen, and kidney and bacteria accumulate freely in phagocytes, capillaries, and interstitial spaces. Chronic lesions in the internal organs are characterized by the presence of white tubercles about 0.3-0.5 mm in diameter [7].
Adherence and invasive capacities are essential in the first stage of infection [3]. P. damselae subsp. piscicida has been reported to be weakly or moderately adherent and invasive in various fish cell lines but has shown a high binding capacity to fish intestines [8]. The adherence seems to be mediated by a protein or glycoprotein receptor of the bacterial cell surface, and the internalization of the bacteria occurs through an actin microfilament-dependent mechanism [8], with cell metabolism playing an active role [9]. However, the precise nature of the mechanism responsible for adherence and interaction with host cell receptors and virulence factors contributing to the invasion of fish nonphagocytic cells is still unknown [9].
Several virulence mechanisms of P. damselae subsp. piscicida have been described. The polysaccharide capsular material plays an important role in the pathogenesis of the bacterium, conferring resistance to serum killing and increasing fish mortality [10]. Furthermore, the intracellular 2 Journal of Immunology Research survival of the pathogen is likely to confer protection against specific and nonspecific host defenses and exogenous antimicrobial agents including antibiotics [8]. Extracellular products with phospholipase, cytotoxic, and hemolytic activities may account for the damage to the infected cells, the consequent release of the microorganisms, and the invasion of adjacent cells. In particular, a key pathogenicity factor of P. damselae subsp. piscicida is an exotoxin, the plasmid-encoded apoptosis-inducing protein of 56 kDa (AIP56), abundantly secreted by virulent strains and responsible for apoptogenic activity against sea bass macrophages and neutrophils in acute fish photobacteriosis [11]. The AIP56 toxin is a zincmetalloprotease that acts by cleaving NF-B p-65, with the catalytic activity located in the N-terminal domain and the C-terminal domain involved in binding and internalization into the cytosol of target cells [12]. The AIP56 induces activation of caspases 8, 9, and 3, loss of mitochondrial membrane potential, release of cytochrome c into the cytosol, and overproduction of ROS, suggesting activation of both extrinsic and intrinsic apoptotic pathways [13]. Through the activation of the cell death process involving macrophages and neutrophils, the pathogen is able to subvert the immune defenses of the host and to produce infectious disease.
Another important virulence mechanism of P. damselae subsp. piscicida is the acquisition of iron from its host by using high-affinity iron-binding siderophores, low molecular weight iron-chelating molecules that interact with bacterial membrane receptors to transport iron into the bacterium [14]. Furthermore, P. damselae subsp. piscicida is able to acquire iron from hemin and hemoglobin as unique iron sources in vitro [14], and iron limitation results in an increased binding of hemin in virulent strains [15]. The heme uptake of the bacterium includes a TonB system to transport heme into the cytoplasm and an ATP-binding cassette (ABC) system to drive it across the cytoplasmic membrane [16,17].
Little is known about the fish immune response to the bacterium and the factors responsible for its failure to protect against P. damselae subsp. piscicida. A transcriptomic approach has recently been applied to elucidate the early immune responses of juvenile gilthead sea bream to P. damselae subsp. piscicida infection. A rapid recognition of the pathogen is shown by the upregulation of lectins, peptides with antimicrobial activity, chemokines, and chemokine receptors, as well as protein of iron and the heme metabolism as a response against bacteria that are dependent on iron. However, this defensive reaction can be either beneficial or devastating to the host [18]. Moreover, the upregulation of genes with highly specialized suppressive functions has been observed indicating an active suppression of immunity that can be induced by the host to reduce tissue damages or by the pathogen to evade the host response [18].

Prevention of P. damselae subsp. piscicida Infection
Antibiotics have been the first line of defense in fish aquaculture to control photobacteriosis outbreaks, but after only a few years the pathogen acquired resistance to various antibiotics. In fact, different transferable genetic elements (R plasmids) carrying genes for resistance against kanamycin, sulphonamide, tetracycline [19][20][21][22], ampicillin [22,23], chloramphenicol [22,24], florfenicol [25], and erythromycin [26] have been documented in P. damselae subsp. piscicida. Differences in the geographic distribution of multidrug transferable elements have been observed among several strains collected in Japan and United States [22,27]. Furthermore, the intracellular parasitism of P. damselae subsp. piscicida within macrophages undermines the effectiveness of chemotherapy.
Taking into account all of these issues, research has been focused on the development of effective vaccines to prevent photobacteriosis and reduce the use of antibiotics in fish farming with benefits at biological and environmental point [28]. Conventional P. damselae subsp. piscicida vaccines are based on inactivated products containing cellular (heato formalin killed bacteria) and soluble antigens (LPS and ribosomal formulations) for immersion and injection administration (Table 1). However, they appear to be ineffective in protecting against pasteurellosis [2,3,[28][29][30][31][32][33][34][35][36][37][38]. Bacterins overexpressing a 97 kDa OMP and a 52 kDa ECP protein, involved in the internalization of the bacterium, are reported to be effective in both sea bass and yellowtail when delivered by immersion eliciting a strong antibody response in the gills and mucosae that may block pathogen entry and colonization [2]. However, the only commercially available vaccine is an ECP-enriched bacterin preparation that has been employed in several European countries with mixed results ranging from good in Spain, Turkey, and Greece to poor in Italy [28,39]. The recommended vaccination protocol consists of two bath immersions at monthly intervals starting at the larval stage when the fish is 50 mg and an oral booster immunization when fish reaches 2 g body weight [3].
Recombinant DNA technology and biotechnological approaches have thus far been used to a very limited extent for the development of bacterial vaccines for fish and effective preventive measures against fish pasteurellosis do not yet exist. Studies on the development of subunit vaccines have recently been reported in cobia from a Taiwan P. damselae subsp. piscicida isolate [4]. Immunoproteomics, using western blotting on protein analyzed with 2DE and LC-MS/MS to isolate immune-reactive proteins, has been applied to identify P. damselae subsp. piscicida antigens that were then cloned and produced as recombinant proteins. In particular, three antigens were shown to induce a protective effect in cobia and therefore were reported as potential vaccine candidates for the development of a subunit vaccine against the pathogen. However, the protection of these vaccine candidates has not been investigated in other fish species, where P. damselae subsp. piscicida causes serious disease and high mortality, and against other P. damselae subsp. piscicida isolates [4]. Moreover, antigen combinations were studied revealing that bivalent subunit vaccines may achieve a better efficiency than monovalent or trivalent antigens [41].
In our laboratory a biotechnological approach based on the reverse vaccinology has been applied to design a vaccine against fish pasteurellosis [40]. New genomic sequences of P. damselae subsp. piscicida were the starting point for bioinformatic analysis aiming to identify new proteins localized  [30], and Kusuda and Salati [31] on the bacterial surface. In fact, the primary condition in selecting a bacterial protein as a vaccine candidate is its cellular localization. Cytosolic proteins are unlikely to be immunological targets, whereas surface exposed and secreted proteins are more easily accessible to the host immune system [47]. In vitro screening of the in silico selected vaccine candidates by an inhibition adherence assay revealed that immunoglobulins from mice immunized with one of the recombinant vaccine candidates were able to affect the adherence of P. damselae subsp. piscicida to fish epithelial cells. The candidate antigen, found to be involved in the adherence and internalization of P. damselae subsp. piscicida in CHSE-214 cells, was predicted in silico as likely lipoprotein with outer membrane localization. The N-terminal signal peptide of 20 amino acids contains the lipobox motif, 2 positively charged residues within the first 7 amino acids and a transmembrane helix of 10 residues. A database search revealed homology with hypothetical proteins and no putative conserved domain; therefore, no putative biological role could be assigned to this lipoprotein. Vaccination and challenge experiments in a laboratory trial indicated that immunization of sea bass with the recombinant antigen induced the production of specific antibodies and conferred protection against P. damselae subsp. piscicida challenge [40]. In vivo long persistence of lipoprotein antibodies was obtained with a single antigen administration in agreement with Ho et al. [4] who reported that multiple administrations do not increase protection in fish. The recombinant lipoprotein is potentially able to protect sea bass against P. damselae subsp. piscicida and could be an interesting candidate for the design of a recombinant vaccine against photobacteriosis. However, protection efficacy over time, increasing doses of the antigen, and its use in combination with different adjuvants must be further investigated in field experiments. Due to the inconsistency of effective measures to prevent photobacteriosis, research has also focused on alternative methods to control the disease. Such methods include probiotics, to be applied in aquaculture to improve health, and a strain of Pediococcus pentosaceus, a lactic acid bacterium isolated from the intestine of adult cobia, has been investigated for its probiotic potential [48]. The acidic pH derived from metabolic acids in lactic acid bacteria culture supernatant has been shown to inhibit P. damselae subsp. piscicida growth in vitro. Dietary supplementation with P. pentosaceus in cobia enhances the growth rate and respiratory burst of peripheral blood leukocytes in fed fish. Furthermore, lactic acid bacteria feeding increased the survival rate of cobia after P. damselae subsp. piscicida immersion challenge. The mechanism affording this protection is still unclear. Although feeding with lactic acid bacteria did not increase specific antibody response after the immunization of cobia with inactivated P. damselae subsp. piscicida vaccine, it heightened the synergic protection against P. damselae subsp. piscicida challenge by 22% and could be administered by itself as a probiotic or with vaccination [48].
Furthermore, selective breeding for fish strains genetically resistant to photobacteriosis constitutes a potential strategy to reduce the probability of disease outbreak and avoid the dramatic consequences of high mortality in fish farms [49]. Quantitative trait loci mapping is applied to detect the regions of the host genome that are associated with resistance to the disease and marker-assisted selection is a useful approach used in several aquaculture species [50][51][52].
A study investigating quantitative trait loci for resistance to fish pasteurellosis in the gilthead sea bream identified two significant quantitative trait loci, one affecting late survival and another impacting overall survival, and a potential marker for disease resistance [49]. The identification of phase-specific quantitative trait loci in gilthead sea bream supports the hypothesis of a biphasic defense response with a primary infection by experimental exposure to the pathogen and a secondary infection with bacteria released from moribund and dead fish [49,53]. Results of quantitative trait loci, mapped by identifying regions of the genome that explain complex traits such as survival, could also be used to gain a better understanding of the mechanisms of disease resistance and defense response. Further insights might also be gained through comparative mapping with other species susceptible to photobacteriosis.

Identification of P. damselae subsp. piscicida and Diagnosis of Infection
Rapid diagnosis of fish photobacteriosis outbreaks is essential for proper management and effective control in aquaculture.
Disease diagnosis is usually made using standard microbiological methods, based on pathogen culturing and isolation steps. Biochemical and serological confirmation is also necessary to characterize the bacterium and to discriminate between the two closely related subspecies, piscicida and damselae of P. damselae. The miniaturized system AIP20E is usually used for a presumptive identification of the P. damselae subsp. piscicida. Although P. damselae generally displays a unique code of 2005004 for the piscicida [54] and 2015004 for the damselae subspecies [1], some strains exhibit aberrant reactions that can lead to misleading results [55]. Hence, differentiation of the subspecies P. damselae subsp. damselae can be achieved when three or more positive results are obtained in the lysine decarboxylase (LDC) production, motility, nitrate reduction to nitrite, gas production from glucose, thiosulfate citrate bile salts-sucrose (TCBS-1) growth, and urease tests, because all these tests yield negative results for all P. damselae subsp. piscicida strains [55,56]. Serological methods such as agglutination or the ELISA have also been developed and commercialized [3].
To overcome the problem of time-consuming and laborious procedures, in the last few years molecular methods have been developed in order to achieve an accurate and specific identification of P. damselae subsp. piscicida and a rapid diagnosis of photobacteriosis ( Table 2). The point at issue is the strong similarity at the DNA level between the two subspecies that makes it difficult to identify sequences useful for designing a subspecies-specific method [3,42,57]. rRNA sequences have been considered for this purpose [42], but strong similarities have been detected both in the 16S, 23S, and 5S (>99%) and the intergenic spacer regions (98-99.5%) between the two subspecies of P. damselae. Moreover, the mosaic-like structure of the latter makes them unsuitable for diagnostic purposes [42,58]. Only a PCR-based method at species level has been developed using the 16S sequences [42].
Integrated sets of methods combine the amplification of the capsular polysaccharide gene to identify the species P. damselae with an additional culture step on TCBS-1 agar to differentiate P. damselae subsp. piscicida from P. damselae subsp. damselae [44] or the amplification of two P. damselaespecific targets with restriction analysis of PCR products to obtain a unique digestion profile for P. damselae subsp. piscicida strains [46].
A multiplex PCR method based on the 16S rRNA and ureC genes has been proposed to discriminate between the two subspecies. The ureC gene is present in P. damselae subsp. damselae genome but is not found in P. damselae subsp. piscicida [1]. On the contrary, a P. damselae subsp. piscicidaspecific target sequence, conserved among strains of different geographical origin but not shared by P. damselae subsp. damselae, has not yet been reported [42,44].
An additional multiplex PCR protocol has been developed in our laboratory as a valid alternative to standard culture methods for the rapid and specific diagnosis of photobacteriosis in fish [43]. The gene coding for a penicillin binding protein 1A (GenBank accession number EU164926) was selected from a large-scale genome project as the PCR target for the identification of P. damselae subsp. piscicida because of several mismatches with the corresponding P. damselae subsp. damselae gene mainly clustered in the 3 end of the gene. However, specificity analysis also indicated amplification of the target gene in two P. damselae subsp. damselae strains. This is due to the fact that a stronger sequence similarity to P. damselae subsp. piscicida than to other P. damselae subsp. damselae strains was found in these two P. damselae subsp. damselae strains. Hence, an additional PCR target, the ureC gene, lacking in the P. damselae subsp. piscicida genome, was introduced in the assay with the aim of differentiating each strain at the subspecies level together with an internal amplification control to obtain a clear distinction between truly negative and false negative results. The optimized multiplex PCR is able to correctly identify and discriminate both subspecies of P. damselae with a detection limit of 500 fg DNA, corresponding to 100 genomic units, twofold higher than that of immunodiagnostic systems (i.e., Bionor Aquaeia-Pp kit) [45].

Conclusions
Partial genome sequencing of several P. damselae subsp. piscicida strains has been previously reported [40,59] and recently a draft of the complete genome sequence of P. damselae subsp. piscicida DI21 strain has been deposited in the public databases (GeneBank accession number PRJNA168653), but the complete gene annotation is not yet available. This information together with the comparative analysis of the genome sequence of different strains of P. damselae subsp. piscicida and P. damselae subsp. damselae will provide further insights laying the groundwork for the development of effective vaccines and diagnostic tools for the causative agent of fish pasteurellosis.