Vibrio harveyi: a serious pathogen of fish and invertebrates in mariculture

Vibrio harveyi, which belongs to family Vibrionaceae of class Gammaproteobacteria, includes the species V. carchariae and V. trachuri as its junior synonyms. The organism is a well-recognized and serious bacterial pathogen of marine fish and invertebrates, including penaeid shrimp, in aquaculture. Diseased fish may exhibit a range of lesions, including eye lesions/blindness, gastro-enteritis, muscle necrosis, skin ulcers, and tail rot disease. In shrimp, V. harveyi is regarded as the etiological agent of luminous vibriosis in which affected animals glow in the dark. There is a second condition of shrimp known as Bolitas negricans where the digestive tract is filled with spheres of sloughed-off tissue. It is recognized that the pathogenicity mechanisms of V. harveyi may be different in fish and penaeid shrimp. In shrimp, the pathogenicity mechanisms involved the endotoxin lipopolysaccharide, and extracellular proteases, and interaction with bacteriophages. In fish, the pathogenicity mechanisms involved extracellular hemolysin (encoded by duplicate hemolysin genes), which was identified as a phospholipase B and could inactivate fish cells by apoptosis, via the caspase activation pathway. V. harveyi may enter the so-called viable but nonculturable (VBNC) state, and resuscitation of the VBNC cells may be an important reason for vibriosis outbreaks in aquaculture. Disease control measures center on dietary supplements (including probiotics), nonspecific immunostimulants, and vaccines and to a lesser extent antibiotics and other antimicrobial compounds.


Introduction
Vibrio harveyi, which occurs naturally in marine habitats (Firmino et al. 2019;Zhang et al. 2018), has developed into a significant pathogen of wild and cultured marine fish and invertebrates (notably penaeid shrimp), especially in warm waters of Asia, southern Europe, and South America. However, V. harveyi is not always recovered as a pure culture from diseased animals. Instead, there is some evidence that the pathogen may be recovered in mixed microbial populations, with, for example, V. chagassi (Fabbro et al. 2011). Moreover, V. alginolyticus, V. harveyi, and V. parahaemolyticus were detected in diseased yellow croaker (Pseudosciaena crocea) in China; none of the three bacterial pathogens dominated (Liu et al. 2016). Similarly, V. harveyi was recovered with V. alginolyticus from diseased hybrid groupers (Epinephelus polyphekadion × E. fuscoguttatus) in Malaysia (Mohamad et al. 2019). Of course, it is uncertain whether the presence of two or more bacterial taxa from the same pathological material represents co-culture or the presence of secondary invaders or even chance contaminants. Unfortunately, this aspect of pathology is rarely investigated. This review strives to discuss the biology of V. harveyi particularly aspects which appertain to its role as a serious pathogen of mariculture.
Edited by Xin Yu.

Taxonomy
Vibrio harveyi is a member of the genus Vibrio, in the family Vibrionaceae, order Vibrionales, class Gammaproteobacteria, and phylum Proteobacteria. The classification of the organism has progressed from its initial name of Achromobacter harveyi (E. N. Harvey was a pioneer in the systematics of bioluminescence; Johnson and Shunk 1936) to Lucibacterium harveyi and Beneckea harveyi to the currently accepted name, V. harveyi (Farmer et al. 2005). Identification of isolates has progressed from phenotypic, serologic to genotypic or polyphasic, including the use of DNA:DNA hybridization when Ishimaru and Muroga (1997) confirmed that pathogenic isolates from milkfish (Chanos chanos) in Japan were definitely V. harveyi. A brief phenotypic description mentions that V. harveyi comprises Gram-negative, fermentative rod-shaped bacteria, which require for sodium chloride for growth and are motile by polar flagella (Fig. 1a, b). Growth occurs on TCBS (cholera agar = thiosulfate citrate bile salts sucrose agar) (Fig. 1c;Farmer et al. 2005). Some cultures are luminous (Fig. 1d). However, the phenotype may be changed by infection with the Vibrio harveyi myovirus like bacteriophage (VHML) (Vidgen et al. 2006). By 16S rRNA gene or whole-genome sequencing, V. harveyi is clearly a core species of Vibrio (Dorsch et al. 1992;Lin et al. 2018).
Vibrio harveyi was linked to V. campbellii by phenotyping and genotyping (Gomez-Gil et al. 2004). They noted that some isolates, which were labeled as V. harveyi, belonged in V. campbellii. However, multilocus sequence analysis confirmed the distinctness of bona fide isolates of V. harveyi and V. campbellii, and verified that both species are, indeed, separate entities (Thompson et al. 2007). Certainly, questions have been raised concerning the accuracy of identifying vibrios within the Harveyi clade. This problem may reflect on the associated studies discussed below (Hoffmann et al. 2012;Ke et al. 2017;Urbanczyk et al. 2013).
Vibrio carchariae was named after recovery of an isolate from a dead sandbar shark (Carcharhinus plumbeus)  and lemon sharks (Negraprion brevirostris) . However, detailed taxonomic examination, including the use of 16S rRNA gene sequencing (Gauger and Gomez-Chiarri 2002), led to the recognition that V. carchariae and V. harveyi are virtually identical, with the name of harveyi having precedence (Farmer and Hickman-Brenner 1992;Gauger and Gomez-Chiarri 2002;Pedersen et al. 1998). It is noteworthy that isolates identified as V. carchariae have been isolated from humans, who was bitten by sharks (Pavia et al. 1989).
The fish pathogen V. trachuri was named by Iwamoto et al. (1995), but was deduced to be synonymous with V. harveyi, with the latter having precedence in taxonomy (Thompson et al. 2002). These workers carried out a detailed polyphasic taxonomy study employing phenotyping, 16S rRNA gene sequencing, fluorescent amplified fragment length polymorphisms (FAFLP), and DNA:DNA hybridization. The outcome was the realization that V. trachuri was highly related to V. harveyi.
Generally, V. carchariae and V. trachuri have been recovered from diseased fish, whereas V. harveyi was predominantly recovered from invertebrates. Therefore, the combined taxon is especially significant as a pathogen of marine vertebrates and invertebrates.

The diseases
Vibrio harveyi has been linked with disease in a broad spectrum of warm water fish and invertebrates ( Table 1). The list of hosts is extensive, and includes silvery black porgy (Acanthopagrus cuvieri) and brown spotted grouper (Epinephelus tauvina) in Kuwait (Saeed 1995), common dentex (Dentex dentex) and farmed sole (Solea senegalensis) in Spain (Company et al. 1999;Zorrilla et al. 2003), and farmed sea perch (Lateolabrax japonicus) and cage-reared grouper (Epinephelus awoara) in China (Qin et al. 2006;Wang et al. 2002). V. harveyi has been often associated with eye disease, for example in common snook (Centropomus undecimalis) (Kraxberger-Beatty et al. 1990), milkfish (Chanos chanos)  (Ishimaru and Muroga 1997), and short sunfish (Mola mola) (Hispano et al. 1997). Without effective chemotherapy, the infected fish would inevitably become blind. In fish, the pathogen has also been attributed with gastro-enteritis, necrotizing enteritis, nodules on the opercula, scale drop and muscle necrosis, skin ulcers, tail rot, and vasculitis in different fish species (Table 1). With the latter infection, the fish were inappetant, lethargic, disoriented, and displayed necrotic subdermal cysts. Internally, meningitis, encephalitis, vasculitis, kidney necrosis, and liver and kidney damage were recorded (Grimes et al. 1985). Also, V. harveyi has been associated with a vibriosis-like disease in Japanese horse mackerel (Trachurus japonicus) when water temperatures were > 25 °C (Iwamoto et al. 1995). Here, fish displayed a diverse pathology, namely melanosis, erratic swimming behavior, bilateral exophthalmia, and hemorrhaging on/in the internal organs (Iwamoto et al. 1995). V. harveyi appears to exert great severity on/in immunosuppressed fish (Grimes et al. 1985).
An interesting situation occurred in Malaysia whereby farmed juvenile hybrid grouper (E. polyphekadion × E. fuscoguttatus) experienced losses of 29% in 10 days with diseased fish becoming lethargic, displaying excessive mucus production, fin rot, congestion of the brain, liver and kidneys, and splenic enlargement. V. alginolyticus and V. harveyi were cultured from diseased fish; both organisms exhibited pathogenicity in laboratory-based infectivity experiments. Indeed, use of both organisms administered concurrently led to more severe disease than when single cultures were used (Mohamad et al. 2019).
Vibrio harveyi has been particularly troublesome in shrimp culture. The pathogen has been recovered from Japan Nishimori et al. (1998) outbreaks of Bolitas negricans (Spanish = small balls; diseased shrimps contain balled epidermal tissue which can block the digestive tract) in Ecuador (Robertson et al. 1998) and luminous vibriosis (the diseased animals can glow in the dark; Prayitno and Latchford 1995), and linked with black shell disease in tiger shrimp (Penaeus monodon) in India (Selvin et al. 2005) and bacterial white tail disease in whiteleg shrimp (Litopenaeus vannamei) in China (Zhou et al. 2012). With the latter disease, mass mortalities have been reported in which the affected shrimp displayed white or opaque lesions in the tail, which were attributed to muscle necrosis (Zhou et al. 2012). The pathogen has been cultured from white patch disease in seahorses (Hippocampus kuda) (Raj et al. 2010). This disease, which was characterized by white patches on the surface and anorexia, resulted in high mortalities among the captive seahorses. Also, the pathogen has been associated with disease in lined seahorses (Hippocampus erectus) (Qin et al. 2017). Moreover, Chinese production of Japanese abalone (Haliotis discus hannai) experienced heavy mortalities attributed to V. harveyi (Nishimori et al. 1998). The disease signs centered on the presence of white spots on the foot and eventually mortalities (Nishimori et al. 1998). Foot pustule disease was described in Japanese abalone, which were lethargic and weak, and exhibited pustles and atrophying of the foot muscle, leading to mortalities .
Luminous vibriosis caused by V. harveyi was reported in the packhorse rock lobster (Jasus verreauxi) larvae (Diggles et al. 2000). At the beginning of the infection, the larvae developed quickly expanding white spots, with death occurring within three days when the water temperatures were 20-23 °C, and the epidermis within the spots was completely destroyed (Diggles et al. 2000). Skin ulceration caused by V. harveyi was also documented in juvenile sea cucumber (Holothuria scabra) in Madagascar (Becket et al. 2004).

Diagnosis
Confusion may have resulted previously from reliance on phenotypic diagnostic approaches when misidentification was unknown. Key diagnostic traits, which could allow confusion with V. alginolyticus, included the presence of fermentative motile Gram-negative rods that grew at 11-40 °C in 3-8% sodium chloride, produced catalase, oxidase, indole, and lysine and ornithine decarboxylase, but not arginine dihydrolase, reduced nitrates, was negative for the Voges Proskauer reaction, and decomposed blood, DNA, and gelatin but not aesculin, casein or starch, and utilized arabinose, cellobiose, glucose, and sucrose but not inositol or lactose (Austin and Austin 2016). However, modern molecular approaches have heightened the accuracy of diagnoses, and include 16S rRNA gene sequencing, which has become the favored method of identifying isolates accurately (Ransangan and Mustafa 2009) and multilocus sequence analysis, which successfully identified 36 isolates from Australia (Cano-Gomez et al. 2011). It is noteworthy that a polymerase chain reaction (PCR) based on the V. harveyi toxR gene was sensitive, i.e., capable of detecting 4.0 × 10 3 cells/ml, and specific to V. harveyi but no other vibrios (Pang et al. 2006). However, a PCR was developed, which was based on the conserved sequence of the gene encoding the OMP vhhP2, and was shown to be quick and fairly specific, although some V. campbellii cultures cross reacted and were misidentified as V. harveyi (Cano-Gomez et al. 2011). Success has been reported for a multiplex PCR, which was both specific and sensitive (Pinto et al. 2017). Similarly, an enterobacterial repetitive intergenic consensus (ERIC)-PCR was developed and regarded as effective for detection of the pathogen (Xu et al. 2017). A highly sensitive [60 colonyforming units (CFU) per reaction] and specific recombinase polymerase amplification (RPA) was designed specifically to recognize the V. harveyi toxR gene. The researchers considered that the technique was fast and accurate, permitting effective monitoring of disease outbreaks (Pang et al. 2019).
Serological approaches have garnered support. For example, a chemiluminescent-based dot blot incorporating monoclonal antibodies was described, and invoked a brief 2-h incubation leading to a limit of detection of 2 × 10 5 /ml; this was regarded as 50 times more sensitive than the conventional dot blot, and was superior (~ 1000 times more sensitive) to the indirect enzyme-linked immunosorbent assay ).

Pathogenic mechanisms
We will preface remarks by stating that much of the research on pathogenicity has involved use of laboratory cultures, some of which are of uncertain authenticity/identity. Therefore, there may be questions concerning whether or not the research has actually used genuine V. harveyi cultures, and the actual relevance of the data to the host. An initial conclusion may be reached that the pathogenicity mechanisms remain to be properly resolved. Nevertheless, infections and mortalities have been achieved using fish and cultures. For example, using V. trachuri and Japanese horse mackerel (Trachurus japonicus), which were maintained at a water temperature of 26 °C, 1.1 × 10 8 cells/fish were injected intraperitoneally, and led to 100% mortalities within 24 h. However, using only 1.1 × 10 7 cells/fish, a total of 50% mortalities occurred in four days. In comparison, immersion in 3.6 × 10 7 cells/ml for 2 min led to 100% mortalities in three days. The disease signs matched those of naturally infected fish, namely erratic swimming and melanosis (Iwamoto et al. 1995). Yii et al. (1997) reported an LD 50 dose for grouper (Epinephelus coioides) of 2.53 × 10 7 CFU/g of fish. Moreover, they noted that the disease signs were similar to those of the natural infection, namely that the intestine was swollen and full of yellowish fluid. In comparison, Pujalte et al. (2003) reported an LD 50 value of 1.5 × 10 5 to 1.6 × 10 6 CFU/fish for European sea bass (Dicentrarchus labrax). Similarly, Won and Park (2008) found the LD 50 dose to Japanese flounder (Paralichthys olivaceus) and black rockfish (Sebastes schlegeli) to be 2.48 × 10 5 -8.76 × 10 7 and 2.0 × 10 4 -2.52 × 10 6 CFU/g of fish, respectively. Therefore, how does the host respond to invasion by V. harveyi? Using suppression subtractive hybridization (SSH) technique, Wang et al. (2008) identified several immunerelated genes, such as a heat shock protein (HSP) 70 gene and a major histocompatibility complex (MHC) class 1a gene, from kidney and spleen of turbot (Scophthalmus maximus) infected by V. harveyi. The rapid transcriptional upregulation after V. harveyi infection may be important for the survival of fish . Moreover, pathogenic V. harveyi isolates from diseased Japanese flounder were more resistant to the bactericidal effects of fish serum, and caused elevated respiratory burst activity in head kidney macrophages of the host. Apoptosis was induced in cultured fish cells, leading to apoptotic bodies, DNA fragmentation, and enhanced caspase-3 activity (Li et al. 2011).
Many virulence factors, such as hemolysins, proteases, lipopolysaccharide (LPS), the capacity to bind iron, interaction with bacteriophages, biofilm formation, and quorum sensing, have been identified (Table 2 and see below for details). It is certainly possible that pathogenicity reflects the interaction between two or more virulence factors functioning together or sequentially (Won and Park 2008). Moreover, it is important to realize that the pathogenicity mechanisms of V. harveyi to fish and invertebrates may well be different (Owens et al. 1996;Zhang 2001).

Ability to bind iron
It has been argued that the ability of pathogens to bind iron may be important for fish but not for invertebrates (Owens et al. 1996).

Bacteriocin-like substances (BLIS)
BLIS was isolated from a culture, which was pathogenic to salmonids, i.e., V. harveyi VIB 571, and determined to be inhibitory to four other V. harveyi isolates and to other Vibrio spp., e.g., V. parahaemolyticus. BLIS was prepared from culture supernatants, and characterized as a unique protein with a molecular weight of ~ 32 kDa (Prasad et al. 2005).

Bacteriophage
Bacteriophages have been associated with virulence (Austin et al. 2003;Oakey and Owens, 2000). In particular, V. harveyi myovirus like (VHML) phage is a temperate bacteriophage with a narrow host range that could enhance and even restore virulence of V. harveyi isolates to Atlantic salmon (Salmo salar L.) and increase hemolytic activity (Austin et al. 2003;Oakey et al. 2002). Infecting an avirulent V. harveyi strain VIB 642 with VMHL led to virulence and upregulation of hemolysin secretion (Munro et al. 2003). Knowledge of VMHL increased as the complete nucleotide sequence and the open-reading frame (ORF) were determined (Oakey et al. 2002). Furthermore, it has been suggested that bacteriophage may mediate toxicity of isolates in tiger shrimp by transfer of one or more toxin genes or genes   Sun et al. (2008) controlling toxin production (Ruangpan et al. 1999). Subsequently, a novel siphovirus-like phage 1 (VHS1) was recovered, but a connection with virulence remains to be established (Khemayan et al. 2006;Pasharawipas et al. 2005).

Biofilms
Biofilms form when organisms are attracted towards or arrive by accident at a surface, become attached, and then multiply. Specific attachment to chitin by means of chitinbinding proteins is an important mechanism for adhesion and colonization of V. harveyi Kirchman 1993, 1994). Research has demonstrated that V. harveyi may survive in shrimp hatcheries due to the ability to form biofilms, which confer resistance to antibiotics and disinfectants (Karunasagar et al. 1994).

Extracellular products (ECP)
Vibrio harveyi has been reported as slightly cytotoxic to fish and invertebrates, and produces ECP (Table 2; Liu et al. 1996;Won and Park 2008;Zhang and Austin 2000), which comprise hemolysins (Li et al. 2011;Zhang and Austin 2000), caseinase, gelatinase, lipase, and phospholipase (Liu et al. 1996;Zhang and Austin 2000). When brine shrimp (Artemia franciscana) nauplii were injected with a range of V. harveyi strains, correlations between mortality of brine shrimp and production of proteases, phospholipases, or siderophores of V. harveyi cultures were observed, but there is no correlation between mortality and hemolytic activity, hydrophobicity, or lipase or gelatinase production (Soto-Rodriguez et al. 2003). A 22 kDa extracellular cysteine protease has been purified (Fukasawa et al. 1988a, b;Liu et al. 1997) and considered to be the major exotoxin to tiger prawns (Lee 1996;Lee et al. 1997;Liu and Lee 1996). Moreover, Harries and Owens (1999) isolated two proteins, which caused mortalities in tiger prawns, and were regarded as both exotoxins and the probable virulence factors. When the ECP of pathogenic strain E2 was heated at 100 °C for 10 min or digested with proteinase K, it produced the same pathology in Dublin Bay prawns (Nephrops norvegicus) as crude, untreated ECP. Western blotting revealed the presence of low-molecular-weight LPS, which was considered to constitute the lethal toxin (Montero and Austin 1999). However, the same strain was not pathogenic to rainbow trout (Oncorhynchus mykiss) (Zhang 2001), indicating that the pathogenicity mechanisms of V. harveyi to shrimp and fish may well be different. Zhang and Austin (2000) reported that the most highly pathogenic culture in their study was V. harveyi VIB 645, which produced ECPs with the highest titer of hemolytic activity to both Atlantic salmon and rainbow trout erythrocytes. They further correlated virulence of V. harveyi to salmonids with the possession of duplicate hemolysin genes, i.e., vhhA and vhhB. The majority (19/20) of the less virulent or avirulent cultures possessed single genes or none at all ). The open-reading frame (ORF) of the vhhA and vhhB genes was calculated to be 1257 nucleotides in length, which matches the tlh gene of V. parahaemolyticus (Nishibuchi and Kaper 1985). VHH hemolysin was determined to be homologous with lecithinase of V. cholerae and V. mimicus.

Hemolysins in the ECP
The vhhA gene was overexpressed in Escherichia coli, and the purified recombinant VHH hemolysin was found to be cytotoxic to Japanese flounder gill cells. The protein caused mortalities in Japanese flounder with the LD 50 dose calculated as 18.4 µg of protein/fish ). Sun et al. (2007) identified VHH hemolysin as a phospholipase B, and reported that a single residue change (Ser153) in the VHH hemolysin resulted in loss of hemolysin and phospholipase activities, and pathogenicity to turbot. In experiments involving immunolocalization of VHH in tissues of Japanese flounder challenged by the protein, the hemolysin was observed to be confined mainly to intestinal epithelial and gill epithelial cells (Fig. 2), indicating that the intestines and gills may be the primary targets for the invasion of V. harveyi in fish (unpublished data). In addition, recombinant VHH hemolysin induced various apoptotic features (such as Fig. 2 Immunohistochemical localization of recombinant VHH from Vibrio harveyi in tissues of Japanese flounder injected with VHH. Rabbit anti-VHH antibody was used as primary antibody and horseradish peroxidase (HRP)-labeled goat anti-rabbit antibody was used as the secondary antibody. Positive staining (arrows) in intestine (a) and gill (b). Bar = 20 µm apoptotic bodies, chromatin condensation, increased number of TUNEL-positive apoptotic cells, and caspase-3 activity) in Japanese flounder gill cell line. These data led to the conclusion that VHH inactivated these cells by apoptosis, via the caspase activation pathway (Bai et al. 2010). Moreover, VHH induced apoptosis in silver sea bream (Sparus sarba) and black sea bream (Mylio macrocephalus) fibroblast cell lines. This featured reduced mitochondrial membrane potential prior to increased caspase-3 activity. In short, it was highlighted that hemolysin led to cell death by induction of apoptosis (Deane et al. 2012).

Luminescence and quorum sensing
Some V. harveyi strains possess luminescence characteristics when grown either in vitro or in vivo ( Fig. 1b; Zhang 2001). Luminous vibriosis is so named, because diseased shrimp luminesce, i.e., glow in the dark (Prayitno and Latchford 1995). This raises the question concerning the role of luminescence in virulence. Luminescence may be verified by monitoring the expression of the LuxR gene in quorum sensing. Research has shown that the expression of LuxR and toxicity to shrimp was most important in relation to luminescence (Nakayama et al. 2005). V. harveyi had been confirmed to have all three types, i.e., AHL, AI-2, and CAI-1, of quorum-sensing activities (Yang et al. 2011). Additionally, quorum sensing regulates other virulent factors, such as type III secretion system (Henke and Bassler 2004).

Resuscitation from the viable but nonculturable (VBNC) state
The VBNC state was first described by Xu et al. (1982), and referred to a state where bacteria remain viable and metabolically active, but are not culturable on standard bacteriological growth media. It is reasoned that the VBNC state is a survival strategy, which has been used by bacteria in response to harsh environmental conditions, and the bacterial cells in VBNC state can return to activity and culturability under appropriate circumstances (Oliver 2010

Disease control
The spread of disease associated with V. harveyi has inevitably led to a multiplicity of studies aimed at control. Interest started with the use of antimicrobial compounds of which Prefuran (Argent) administered at 0.1 mg/L for an unspecified period (Kraxberger-Beatty et al. 1990) and oxytetracycline (Saeed 1995) were successful. However, the concern over tissue residues and the development and spread of antibiotic resistance contributed to research in other areas of disease control, notably preventative rather curative.

Bacteriophage therapy
Some research has focused on bacteriophage therapy to control luminous vibriosis in shrimp aquaculture (Choudhury et al. 2019). A lytic Myoviridae, VhKM4, was recovered (Lal et al. 2017). Moreover, a cocktail of three bacteriophages of the Myoviridae and Siphoviridae, which were isolated from shrimp farms in India, inhibited V. harveyi, led to greater survival of tiger shrimp larvae compared to controls (Stalin and Srinivasan 2017).
Probiotics have been evaluated for control of V. harveyi infections. For example, Bacillus coagulans achieved success when fed at 10 9 -cells/g of feed to the freshwater prawn, Macrobrachium rosenbergii. After 60 days, there was improved growth, significantly enhanced lysozyme and respiratory burst activities, and increased resistance to challenge (Gupta et al. 2016). Schaeck et al. (2016) were an advocate of V. lentus, which protected gnotobiotic (= germfree) sea bass (Dicentrarchus labrax) larvae against challenge with V. harveyi. The benefits of feeding marine red yeast, Rhodotorula mucilaginosa, at 1.0 g/kg for eight weeks have been extolled ). In addition, it was found that a marine antagonistic strain Pseudoalteromonas flavipulchra JG1 could inhibit the growth of V. harveyi (Jin et al. 2010). From this culture, several small molecule compounds (p-hydroxybenzoic acid, trans-cinnamic acid, and 6-bromoindolyl-3-acetic acid) with inhibitory effect to V. harveyi were purified (Yu et al. 2012). This bacterial strain could be used as a potential probiotic to control V. harveyi disease.
A common theme with the dietary supplements is evidence for improved growth performance, immunomodulation notably of cellular and innate immunity, and protection against challenge with V. harveyi. This raises the question concerning the reason(s) for the profound benefit of what is essentially the addition a small quantity of dietary supplement. Could the reason reflect inadequacies in the formulation of diets or are these truly supplements, i.e., supplementing already adequate feed?

Inhibition of quorum sensing
Quorum-sensing signal molecules may be involved in regulating virulence. In this connection, the 28 kDa AiiA protein of Bacillus thuringiensis interfered with quorum sensing in V. harveyi, and could be considered for involvement in disease control strategies (Bai et al. 2008).

Vaccines
Although vaccines are primarily used to control diseases of fish, some commercial products containing V. harveyi antigens have been marketed for the control of shrimp diseases in the Far East. Nevertheless, most of the research has been directed at the control of fish disease.
There has been a dramatic improvement in the success of vaccines for the control of V. harveyi diseases reflecting some excellent science (Table 3). Efforts started with the traditional approach of inactivated whole-cell preparations before progression to consideration of purified subcellular components, and subsequently to the modern era of DNA vaccines.
Whole-cell vaccines In their simplest form, whole-cell vaccines comprise bacterial suspensions, which have been inactivated with chemicals, typically formalin. Whole-cell vaccines have gained considerable success with the control of some fish diseases, and numerous examples have been commercialized, worldwide. The approach has been used to control diseases attributed to V. harveyi. For example, a   ; Wang et al. (2011Wang et al. ( , 2017) Immunostimulation Baletal and Gomez-Chiarri (2016) whole-cell vaccine administered by a variety of techniques, i.e., intraperitoneal injection, anal intubation, and, by immersion, to barramundi (Lates calcarifer) led to demonstrable antibody production. Clearly, the data revealed that barramundi could mount an immune response to vaccination (Crosbie and Nowak 2004). A bivalent vaccine comprising formalized cells of V. harveyi and Photobacterium damselae subsp. piscicida together with ECP was used in sole both involving immersion together with a booster dose or by i.p. injection. The outcome was excellent protection [relative percent survival (RPS = ~ 88%) for a 4-month period after vaccination]. Subsequently, protection declined (Arijo et al. 2005). A multivalent inactivated vaccine including cells of V. alginolyticus, V. harveyi, V. vulnificus, and infectious spleen and kidney necrosis virus was administered to groups of orange-spotted grouper (Epinephelus coioides) leading to excellent protection after challenge (RPS = 80%; Huang et al. 2012). The commercial attraction to inactivated whole vaccines is that they are comparatively inexpensive to produce. Consequently, research has been directed towards improving the performance of these vaccines, such as by entrapping the antigens in liposomes (Harikrishnan et al. 2012c). Use of CpG oligodeoxynucleotide 1668-enriched plasmid (p60CpG) as an adjuvant with formalized cells led to protection (RPS = 96.2%) in orange-spotted grouper (Nguyen et al. 2017). This led to immunomodulation and increased protection (Harikrishnan et al. 2012b). A similar positive effect resulted from the addition of adjuvant, Montanide, to formalin-inactivated cells, which were used to inject turbot. The vaccine was administered at 1.0 × 10 8 cells/fish leading to an RPS of 75% after four weeks (Xu et al. 2019). Subunit vaccines. Many articles have been published showing the success of subunit vaccines. ECPs, outer membrane proteins (OMP) (Arijo et al. 2008), and the hemolysin protein expressed in yeast (Saccharomyces cerevisiae) (Zhu et al. 2006) were immunogenic. Pang et al. (2010) described the use of OmpN mixed with Freunds complete adjuvant (FCA) to administer by i.p. injection (dose = 100 µl/fish) estuary cod (Epinephelus coioides) leading to an RPS or 60% or 70% depending on the challenge culture. A recombinant fusion protein r-OmpK-OmpU controlled disease in orange-spotted grouper (RPS = 81.8%) (Nguyen et al. 2018). Subsequently, the OMP To1C, which is a major adhesin, was used successfully to protect hybrid grouper (E. fuscoguttatus × E. lanceolatus) against experimental challenge (Zhu et al. 2019).
Molecular approaches to vaccine development have met with some success as illustrated by the study of Zhang et al. (2008), who cloned OMP genes, OmpK and GADPH; the recombinant proteins from which were expressed in the vector pET-30a(+). After purification of the protein, 100 µg quantities/fish were injected intraperitoneally into groups of yellow croaker (Pseudosciaena crocea). A booster dose was administered at three weeks. Use of r-OmpK and r-GADPH led to RPS values of 37.7% and 40%, respectively. Although some protection occurred, the levels were not as good as some of the studies using simpler whole-cell vaccines. In a subsequent development, expression of OmpK was achieved using the yeast Pichia pastoris. Experiments used groups of sea bass (Lateolabrax japonicus) and oral uptake of OmpK in alginate microspheres for five days. The outcome was an RPS of 61.5% (Mao et al. 2011). A recombinant protein vaccine, which was developed from the VirB11 gene, was trialed in groups of orange-spotted grouper. After challenge, an RPS of 88% was achieved (Atujona et al. 2019).
Live vaccines. We have two concerns about live vaccines that need to be voiced. First, what is the possibility that the presence of live vaccines cells in and around a fish holding facility could be mistaken for the presence of a population of viable pathogens cells? The presence of such viable cells could confuse diagnostics. The second concern revolves around whether or not it is possible for an attenuated vaccine cell to acquire virulence determinants in the aquatic environment. There is not any evidence for these concerns, but it is considered necessary for scientists to be aware of and not ignore the issues. The attraction for live vaccines is that their use is more likely to mimic the infection process of viable pathogens. Therefore, the host responses are likely to be more natural and protective! A live recombinant vaccine based on the OMP VhhP2 was injected intraperitoneally leading to excellent protection of Japanese flounder (RPS = 92.3%) (Sun et al. 2009). This was followed by a series of vaccination studies leading to commendable levels of protection. Thus, Cheng et al. (2010) used a recombinant protein, Vhp1, which when expressed in Escherichia coli and used as a live vaccine achieved an RPS of 90%. A live recombinant vaccine was developed by , who expressed DegQ. The vaccine was injected intraperitoneally into turbot leading to an RPS of 90.9%. By oral and immersion, ROS values of 60.5% and 47.1%, were recorded, respectively. The vaccine was used by oral and immersion in a field trial leading to RPS values of 77.8% after a month and 81.8% after two months   (Chin et al. 2020).
DNA vaccines. A DNA vaccine, pcDNA-GPx, was developed by incorporating the glutathione peroxidase GPx gene into the pcDNA3.1 ( +) plasmid, which was injected intramuscularly into orange-spotted groupers leading after challenge to an RPS of 77.5% . Wang et al. (2011) achieved as RPS of 100% in turbot using the purified 35 kDa OMP OmpU at a 50 µg dose applied by intramuscular injection. A DNA vaccine in which the ompU gene was inserted into pEGFP-N1 plasmid was less successful when injected in 10 µg quantities intramuscularly in turbot. With this, the RPS was only 51.4%. DegQ and Vhp1 were incorporated into a DNA vaccine for use in Japanese flounder in which the resulting RPS was 84.6% .
Although many publications have been encouraging in terms of success of vaccines after challenge, it remains to be determined which vaccines could be produced at the price that the users would be prepared to pay.

Immunostimulation
Summer flounder (Paralichthys dentatus) were immersed or injected intraperitoneally with a hot water extract of brown seaweed (Sargassum oligocystum) leading to significant stimulation of respiratory burst and hematocrit, and less mortalities after challenge (Sargassum oligocystum) (Baletal and Gomez-Chiarri 2016).

Conclusions
Vibrio harveyi is a serious pathogen for multiple species of marine fish and invertebrates particularly occurring in the warm waters of Asia, southern Europe, and South America. Some confusion has resulted with identification of cultures, particular centering on distinguishing V. harveyi and V. campbellii. However, taxonomy studies have confirmed the uniqueness of both of these species. Although virulence has been demonstrated with multiple isolates, there is a lack of clarity over the precise pathogenicity mechanisms, possibly reflecting differences between isolates as well as between fish and invertebrates, and overreliance on laboratory cultures, which are not necessarily reflective of the abilities of fresh isolates. Certainly, hemolysin has been the most extensively studied virulence factor of V. harveyi to date. Some excellent work has been accomplished, and it is realized that hemolysin is capable of killing fish cells by apoptosis via the caspase activation pathway. Maybe, key virulence genes, such as occur on plasmids and bacteriophages, become lost with storage of the so-called laboratory cultures. Fortunately, there has been much progress with the development of effective disease control strategies, including use of dietary supplements and vaccines. Clearly, further work is justified to better under this serious pathogen of marine animals.
Author contributions X-HZ conceived the idea to this work and wrote the manuscript; BA wrote the manuscript; XH performed the experiments and prepared the figures. All authors edited and approved the final manuscript.

Compliance with ethical standards
Conflict of interest The authors declare that they have no conflict of interest.
Animal and human rights statement This article does not contain any studies with human participants or animals performed by any of the authors.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creat iveco mmons .org/licen ses/by/4.0/.