Identification, characterization and infection dynamics of Vibrio strains in phyllosoma of Thenus unimaculatus

ABSTRACT The present study investigates Vibrio outbreak in the hatchery system designed for larval rearing of shovel-nosed lobster Thenus unimaculatus in India. High larval mortalities during the episode led to an almost complete loss of the larval stock during the first 10 days. Microscopic examination revealed heavy fouling of the phyllosoma appendages. Histopathological analysis showed the presence of bacteria in larval tissues. Investigation by scanning electron microscopy revealed the formation of bacterial plaques on appendages and adhesion of bacteria on fine setae in heavily infected larvae. Bacteria isolated from larvae and identified by 16S rRNA sequencing and phylogenetic analysis indicated that the strains were closely related to luminescent Vibrio campbellii and Vibrio harveyi. The virulence potential of Vibrio strains isolated from larvae was evaluated through an experimental infection. The study helps to understand microbial infection and its impact on phyllosoma survival and may allow a more effective health management regime to improve larval survival in the hatchery phase.


Introduction
The hatchery phase represents a critical point in closing the life cycle of any cultured species. Generally, in commercially viable crustacean aquaculture ventures, the success depends on species with shorter larval phases (Wickins and Lee 2008). The shovel-nosed lobster, Thenus unimaculatus (Burton and Davie 2007), is emerging as a preferred candidate species of aquaculture interest because of its shorter larval phase (30 days), hardiness and reasonable market value (Kizhakudan et al. 2004;Kizhakudan 2005;Vijayakumaran and Radhakrishnan 2011;Kizhakudan and Krishnamurthy 2014). Though the Indian species was earlier known as Thenus orientalis, studies by Burton and Davie (2007) confirmed the presence of a different species, Thenus unimaculatus. It is one of India's important commercially exploited lobster (Radhakrishnan et al. 2005). The larval phase of T. unimaculatus consists of four stages of phyllosoma I-IV, followed by a post-larval, non-feeding, nisto stage that metamorphose to juveniles. Phyllosoma is dorsoventrally flat, leaflike, transparent, and is known as the first seed. The complete larval cycle of T. unimacultus has been studied in detail (Kizhakudan and Krishnamurthy 2014), but its seed production is yet to be developed commercially. Periodic mass mortalities in the early stages of hatchery-reared larvae hamper the commercialscale rearing of T. unimaculatus. The major technological challenge for T. unimaculatus culture involves maintaining good health throughout the larval phase.
Though an array of pathogenic bacteria are known to induce mortality in hatcheries, Vibrio species represent the major pathogenic group affecting crustacean larvae, juveniles and other aquatic organisms (Goulden et al. 2012;Prado et al. 2015;Rojas et al. 2015;Dubert et al. 2016). Vibrio are widely distributed in aquatic environments (Vandenberghe et al. 2003), in commensal or pathogenic relations, thereby playing a significant role in aquaculture (Beaz-Hidalgo et al. 2010;Romalde et al. 2014). They can seriously hamper the survival of larvae, especially when the larvae are stressed due to poor hatchery conditions (Lacoste et al. 2001;Tseng and Chen 2004). Despite the development of efficient rearing techniques, T. unimaculatus often suffers recurrent episodes of high mortality during the larval phases. These episodes are mainly associated with high Vibrio load. This study provides the details of our investigation on the Vibrio outbreak in the hatchery system designed for larval rearing of T. unimaculatus. We focused on identifying, characterization, and infection dynamics of Vibrio strains in phyllosoma of T. unimaculatus. Understanding this microbial aspect in the larval-rearing system and its impact on phyllosoma may allow a more effective health management regime and thus improve larval survival.

Materials and method
Larval rearing Wild T. unimaculatus brooders were collected from local fishers of Chennai-Mahabalipuram stretch in North Tamil Nadu, India and transported to the hatchery in Kovalam Field Laboratory, Central Marine Fisheries Research Institute, Chennai, India. Animals were kept in black coloured rectangular tanks with in situ sand bed and filter recirculatory system. Water quality parameters in the tanks like temperature, salinity and pH were maintained at 28-30°C, 33-35 ppt and 8-8.2, respectively. During the breeding season in winter (December to March), females were examined weekly for the estimated date of hatching. The brooders were held in separate tanks, with daily water exchange done manually at the hatching date. Healthy larvae hatched from the brooders and showing positive phototaxis were used for larval rearing in horizontal raceways with filtered seawater with 37-39 ppt salinity. Water exchange with 50% tank volume was carried out daily and aeration was restricted to mild disturbance of water using small airstones. The larvae were fed on freshly chopped hepatopancreas of the backwater clam, Meretrix casta.

Phyllosoma survival
Phyllosoma survival percentage of the whole population was measured over three standard larval-rearing trials. Larval survival estimates were calculated based on counts of dead animals siphoned out daily from the bottom of each larval-rearing tank.

Microscopic examination
Larvae were examined daily under a light microscope (Carl Zeiss, Germany) for any fouling or change in the typical morphology of different tissues.

Histology
Samples of live phyllosoma were randomly selected from larval-rearing tanks, though during die-off periods, samples were a mixture of living and moribund animals. During the disease outbreak, approximately five larvae were examined on days 1, 3, 5, 7 and 10. Phyllosoma larvae were fixed in Davidson's fixative (Hasson et al. 1997) for 48 h before processing for routine histological procedures (Bell and Lightner 1988). Briefly, the fixed tissues were dehydrated in graded ethanol series (70%, 90%, and 100%) for 60 min each. After dehydration, tissues were treated with xylene for 60 min (twice) and paraffin blocks were made using a tissue embedding system. Tissue sections (4-5 μm) were cut, further processed, stained with haematoxylin, and counter-stained with eosin following standard procedures (Bell and Lightner 1988). The stained tissue sections were mounted in dibutyl phthalate polystyrene xylene and visualized with light microscopy.

Scanning electron microscopy (SEM)
The larvae were washed three times with sterile seawater, fixed in 3% glutaraldehyde (prepared in filtered seawater) for 4 h. Following this, the larvae were washed thoroughly and postfixed in 1% osmium tetroxide in 0.1 M cacodylate buffer for 2 h (Bourne et al. 2004). After fixation, the larvae were washed and dehydrated in ethanol series and dried in a polaron critical point drying apparatus, mounted on carbon tabs and sputter-coated with platinum at 25 mA for 150 s in a Balzars MFD 020 sputter coating unit. The coated specimens were examined using a TESCAN VEGA3 scanning electron microscope at 30 kV.

Isolation of the pathogen
The bacteria were isolated from larvae on days 1, 3, 5, 7 and 10. Ten moribund phyllosoma larvae were washed in sterile saline to remove loosely attached epibionts and excess detritus. Samples were homogenized in 10-fold dilution of sterile saline, and serial dilutions of 1 × 10 −3 were prepared. From the different dilutions thus obtained, 100 µl from each was plated in triplicates on Zobell Marine Agar (ZMA) (Himedia) and Thiosulfate Citrate Bile Sucrose Agar (TCBS) (Himedia). The plates were incubated at 37°C for 48 h. Fifteen dominant and unique morphotypes were sub-cultured to purity and subsequently characterized by biochemical and phylogenetic analyses (Garrity et al. 2004).

Extraction of bacterial genomic DNA
The total bacterial DNA was extracted from single bacterial colonies using the HiPurA ™ Multi-Sample DNA Purification Kit (Himedia) as per the manufacturer's instructions. The 260/280 and 260/230 OD ratios were found to be 1.8 and 2-2.2, respectively, confirming the purity of the DNA for downstream sequencing.

Sequencing and phylogenetic analysis
Sanger sequencing of the PCR products was performed by Eurofins Genomics India Pvt. Ltd Bengaluru, India. The NCBI BLAST database was used to analyse sequence homology (98-100%) for extracted 16S rRNA gene sequences. The overlapping 820 bp fragment of 16S rRNA gene from sequencing reactions was then aligned using ClustalW. A phylogenetic tree was constructed by the neighbour-joining method (MEGA version 7 software) using Kimura 2-parameter model (Kimura 1980) and tested by bootstrap with 1000 repetitions (Kumar et al. 2008).

Inoculum preparation for experimental infection
Bacterial isolates obtained during the luminous outbreak in the hatchery were used for the challenge study. Isolates were subcultured on TCBS at 28°C and suspended in 10 ml of seawater sterilized by filtration (0.22 µm pore size). The final suspension was adjusted to an optical density at 600 nm (OD) of 0.1. Corresponding total viable counts were determined by plating on ZMA (Himedia). The mean of 3 replicate counts was used for the immersion challenge. Phyllosoma larvae (stage 1) were distributed into 12 plastic trays each with 30 larvae at a density of 6 per litre. The animals were acclimated in darkness at 28°C for 8 h. The final concentrations used in the trays were prepared by adding appropriate volumes of bacteria to final concentrations of 10 8 , 10 5 and 10 3 CFU ml −1 , representing a high, medium and low level of bacterial dose, respectively. Each treatment was run in triplicate. Experimental control was not exposed to any bacteria. The number of dead or luminous phyllosoma was assessed every 24 h for five days. Phyllosoma not displaying any active movement after prolonged inspection being recorded as dead. The water temperature and salinity were maintained at 28°C and 35 ppt respectively throughout the experiment, and aeration was not provided to any of the trays. Vibrio strains were reisolated from moribund phyllosoma and confirmed by PCR with the same conditions as mentioned above in the PCR section.

Ethics statement
We adhered to the ethical guidelines and the biosafety rules and regulations followed by the institution ICAR-Central Marine Fisheries Research Institute, India.

Phyllosoma survival
Larval rearing was characterized by mass mortality events due to the luminescent Vibrio outbreak. Phyllosoma appeared whitishopaque, lethargic and poorly phototactic. Survival graphs of three tanks showed a similar pattern in 10 days. Initially, there was a mortality of 10-20% in the stock of larvae in the tanks, which increased over a period of 10 days. Survival on the 10th day was 0-10% in the three tanks ( Figure 1).

Microscopic observation
The microscopic examination of infected phyllosoma showed changes as compared to healthy individuals. Heavy fouling was observed on larval setae in infected phyllosoma. The fouling was observed both on appendages and setae (Figure 2).

Histopathological examination
The hepatopancreatic tubules of moribund larvae was atrophied with necrosis of hepatopancreatocytes and (2 or 3/5) larvae contained bacteria in the lumen. Examination of eyes also indicated bacterial infiltration. Severe filamentous bacterial infections on the setae of the appendages were also identified ( Figure 3).

SEM analysis
Examination of P1 phyllosoma larvae by SEM revealed the presence of numerous bacteria over the appendage surface. These bacteria were present in large numbers surrounding small spines protruding from appendages (Figure 4).

Biochemical characters
All strains isolated were Gram-negative, oxidase-positive and negative for vibriostatic agent O/129 (10 µg); showed no growth in 0% and 10% sodium chloride, and positive growth in medium supplemented with 3% sodium chloride. Further analysis revealed facultative anaerobism in all strains. Vigorous growth was also observed on ZMA and Tryptic soy agar supplemented with 2.0% NaCl (w/v) at 28°C for 12 h. Luminescence was observed on plates in a dark room. Growth on TCBS agar was also checked after 72 h by yellow, convex, round colonies at incubation temperature of 28°C. This indicated that all the isolates were halophilic Vibrio species. The strains were negative for arginine dihydrolase reaction, and positive for indole production and gelatin production, establishing close identity with V. harveyi-related organisms. Majority of strains showed variable results in the ornithine decarboxylase test, which is considered as the key test for differentiating V. campbellii from V. harveyi (Alsina and Blanch 1994) (Supplementary  Table 1).

Identification by 16S rRNA
All isolates exhibited 99% sequence identity to the 16S rRNAgene of Vibrio species in the NCBI database. The strains V. campbellii and V. harveyi were repeatedly isolated from both ZMA and TCBS media and were dominant at different time points in a larval-rearing run. The nucleotide sequence data from strains have been submitted to GenBank database under accession numbers, MH231447.1, MH628047.1, MH707096.1 and MH707097. The lowest E value was considered for all comparisons. Construction of a phylogenetic tree with sequences of 16S rRNA with MEGA7 software using NJ analysis revealed a similar identification pattern. Constructed phylogenetic tree showed that the isolated bacteria were closely related to V. harveyi and V. campbellii and were farthest from V. cholera which acted as outgroup ( Figure 5).

Experimental infection of phyllosoma
Experimental infection at higher concentration (10 8 CFU ml −1 ) of V. harveyi and V. campbellii resulted in about 40% mortality of phyllosoma on the first day, and by the fifth day, there was almost 100% mortality. At medium (10 5 CFU ml −1 ) and low (10 3 CFU ml −1 ) bacterial concentrations, survival of about 25% and 40%, respectively, was observed on the 5th day. In the control group, up to 75% survived on the 5th day ( Figure 6).

Discussion
Vibriosis is a global disease in marine aquaculture (Austin et al. 2005). Outbreaks of vibriosis have been observed in several lobster species causing mass mortality of the larvae (Diggles et al. 2000;Handlinger et al. 2000;Webster et al. 2006). Although, microbial community dynamics of lobster larvae have been studied in detail (Bourne et al. 2004;Payne et al., 2006;Goulden et al., 2012), the studies are largely centred on Panulirus ornatus. The present study represents the first attempt to identify, characterize and understand the infection dynamics of Vibrio campbellii and Vibrio harveyi during phyllosoma stage of Thenus unimaculatus. We employed a polyphasic approach incorporating direct microscopic analysis, scanning electron microscopy, culture-based microbiological, biochemical, and molecular methods to understand the impact of Vibriosis on phyllosoma larvae.
Estimation of larval survival in large volumes with unevenly mixed larvae is difficult. Our previous attempts at harvesting live phyllosomas to obtain accurate counts led to appendage injury and increased mortality which could not be distinguished from mortality due to bacterial outbreak. To this end, an indirect method of survival estimation based on the count of dead larvae was employed. This method has been also followed by Bourne et al. (2004). Increased mortalities are often observed in early developmental stages around the time of moulting, which occurs approximately every 7-9 days (Nguyen et al., 2018). We found a reduction in the survival percentage by about 50% during days 7-9 of the outbreak, consistent with the notion that phyllosoma is more susceptible to infection during the moulting stages. Survival patterns were consistent across the three larval-rearing runs, with about 90% mortality within 10 days. Diggles et al. (2000) observed about 75% mortality over 4 weeks in Jasus verreauxi reared at 20-23°C, and infected with Vibrio harveyi. Among other factors, the temperature is a critical factor that determines the rate of a bacterial outbreak (Travers et al., 2009). In our study, the larvae were reared at 28-30°C, which could have significantly accelerated the bacterial growth in the rearing system.
Routine microscopy revealed heavy fouling of the phyllosoma appendages. Phyllosoma death has been reported to increase due to heavy epibiont growth, increasing the difficulty of moult shedding, and hampering respiration which makes it difficult to meet the oxygen requirements (Diggles, 1999, Handlinger et al., 1999, Bourne et al., 2007. This shortage of oxygen is all the more exacerbated during the moulting stage when demand for oxygen increases (Carvalho & Phan, 1998). Further, the epibiont growth is reported to interfere with the ability of phyllosoma to process and masticate feed, which leads to a progressive decline of the nutritional status and increases susceptibility to opportunistic pathogenic bacteria (Handlinger et al., 2000;Fernandez-Leborans et al., 2006). Similar observations have been made on phyllosomas of Panulirus ornatus (Bourne et al., 2007) and Jasus edwardsii (Handlinger et al., 1999) infected with the filamentous bacteria. Although we did not observe fouling around the mouth area of the infected phyllosoma, it is possible that fouling may still  have hindered the nutrition of phyllosoma by compromising its locomotion, thereby restricting capture of feed. SEM analysis confirmed the nature of fouling as bacterial plaques on the appendages and adhesion of bacteria on the fine setae in heavily infected larvae. Bourne et al. (2007) also observed the adhesion of bacteria on the surface of mouth and anus. These entangled appendages are believed to render the animal unable to feed adequately (Bourne et al., 2004;Payne et al., 2006). This may also have significantly contributed to the observed mortality that we report here. The epibiont associated with the fouling of phyllosoma appendages is predominantly Leucothrix mucor and the heavy infestation of this epibiont can cause larval mortality (Sadusky & Bullis, 1994;Kitancharoen, Hatai, & Hara, 1997). Although in the present study, we did not mainly focus on the epibionts which caused fouling, characterization of the fouling bacteria is an important aspect that should be considered as the nature of epibionts is known to, at least partly, determine the welfare of phyllosoma (Shields, 2011).
Histopathological analysis demonstrated that the bacterial infestation was not only restricted to the body surface of phyllosoma but had also affected the internal tissues of the larvae. Proliferation of bacteria was observed in the tubules of the hepatopancreas. Similar pathologies associated with Vibrio infections have also been reported in cultured phyllosomas of packhorse rock lobster, Jasus verreauxi (Diggles et al., 2000), southern rock lobster, Jasus edwardsii (Handlinger et al., 1999) and P. ornatus (Bourne 2004;, and in different life stages of penaeid shrimp (Lavilla-Pitogo et al., 1998;Soonthornchai et al. 2010). Larval feeds and cannibalism of dead larvae can contribute to Vibrio transmission during phyllosoma stages (Goulden, Hall, Bourne, Pereg, & Høj, 2012). Both these factors may be responsible for the occurrence of bacteria in the hepatopancreas (de Souza Valente & Wan, 2021). Infestation of the hepatopancreas also underlines the importance of research towards development of formulated larval feeds which can at least partly reduce the dependence on animal tissues (Gora et al., 2018). Heavily infested moribund or dead phyllosoma exhibit fluorescence that attracts healthy phyllosoma for cannibalism leading to rapid spread of vibriosis (Goulden et al., 2012;de Souza Valente & Wan, 2021). Among other factors, light is an important regulator of cannibalism in crustaceans (Romano & Zeng, 2017). Therefore, proper light conditions in the rearing facility which ensure cessation of cannibalism could be a strategy to reduce larval mortality (Gardner & Maguire, 1998). Examination of eyes indicated bacterial infiltration which suggests the progression to systemic infection in the late stages of moribidity or postmortem. The hepatopancreas, appendages and the eye region are the major organs of the phyllosoma and we were able to find Vibrio infection in these tissues. The systemic nature of Vibrio infection in phyllosoma indicates that although diet may be the major route for entry of Vibrio in the phyllosoma, it rapidy spreads to different organs making it difficult to ascertain the major predilection site for Vibrio in the phyllosoma. This development of rapid systemic infection in phyllosoma advocates for antibiotics to control the Vibrio infection, however, several studies have reported the development of resistance in Vibrio against antibiotics (Kitiyodom et al., 2010). Therefore, prophylactic approach like using natural bioactive compounds is a more sustainable and environment-friendly strategy to control Vibrio outbreaks in the hatcheries (Hall et al., 2013;Rossi et al., 2021).
In recent years, sequencing and comparison of the 16S rRNA gene have become an important tool for identifying bacterial species (Srinivasan et al., 2015). Dominant strains isolated from the culture included luminescent V. campbellii and V. harveyi and this showed a mixed dominance for disease outbreaks. Isolation of more than one species of Vibrio during disease occurrences have been previously reported in the other crustaceans (Davis & Sizemore, 1982;Harrison et al., 2022). Both V. harveyi and V. campbellii are widespread in marine environments. They are responsible for diseases in many wild and reared aquatic organisms, most notably penaeid shrimp, several fish species, and mollusks (de Souza Valente & Wan, 2021). V. campbellii has been isolated from diseased farm-shrimps from south India and has become a major emerging pathogen (Haldar et al., 2011). Furthermore, phylogenetic analysis of 16S rRNA gene sequences of isolated sequences clustered with known V. harveyi-related clade, including V. campbellii, V. harveyi and V. rotiferianus isolates. V. campbellii is closely related to several other Vibrio species, i.e. V. harveyi, V. rotiferianus, V. alginolyticus and V. parahaemolyticus (Thompson et al. 2004(Thompson et al. , 2007. The phyllosoma larvae are particularly vulnerable to opportunistic bacteria as their immune system is still developing (Gollas-Galvan et al. 2017) and culture conditions can further predispose larvae to a variety of stressors that may compromise the immune system (Rehman et al. 2017). Although through experimental infection we were able to ascertain the causative agents as Vibrio spp., it has been reported that Vibrio spp. in crustaceans are opportunistic bacteria which become pathogenic under stressful conditions (de Souza Valente and Wan 2021). Therefore, the contribution from the growth of other opportunistic pathogens to further deteriorate the health of phyllosoma and increase mortality cannot be ruled out. This aspect of bacterial disease outbreaks has not been thoroughly investigated in lobsters. Furthermore in our study, we focused on the culture-dependent technique to determine the causative agent. Future studies should also focus on the culture-independent high-throughput techniques like metagenomics to understand the contribution of other opportunistic pathogens towards disease outbreaks. Unfortunately, there is also no information regarding the microbiota profile of the phyllosoma larvae of T.unimaculatus in the wild and hatcheries. Understanding the shift in the microbiota profile of the phyllosoma during a disease outbreak is fundamental in maintaining the health of the phyllosoma (Holt et al. 2020).
The virulence potential of Vibrio strains isolated from luminous phyllosoma was evaluated and Vibrio strains were reisolated from moribund experimentally infected phyllosomas. Since we were able to find two species of Vibrio during the outbreak, we used three different doses of both species for the experimental challenge to ascertain which of the two Vibrio sp. predominantly accounts for the observed mortality. The selection of three doses of Vibrio sp. for the experimental challenge of phyllosoma in the present study was based on previous reports (Diggles et al. 2000;Goulden et al. 2012). We observed mortalities of 100% with dose 10 8 CFU ml −1 on 5th day at 28°C by V.harveyi and V. campbellii while as Diggles et al. (2000) reported mortality of 100% with dose 10 7 CFU ml −1 at 24°C on 7th day in the phyllosoma of Jasus verreauxi by V.harveyi. On the other hand, there was 60% mortality with dose 10 7 at 28°C in the phyllosoma of P.ornatus by V.owensii (Goulden et al. 2012). Employing different doses for the experimental challenge helped us to understand a dosedependent relationship of Vibrio spp. with the mortality. It appeared that both the species contributed equally to phyllosoma mortality.
The route of the experimental infection is an essential factor that determines the outcome of any bacterial challenge experiment. Goulden et al. (2012) reported a high interindividual variation during the immersion challenge in the phyllosoma of P.ornatus by V.owensii while a consistent mortality pattern was reported in the phyllosoma when they were challenged with the same bacteria using artemia as vector. We used immersion challenge as the route of infection to confirm the susceptibity with different doses, as it allows us more control in regulating the bacterial load exposed to larvae. However, future studies should also confirm the susceptibility using different transmission routes to ensure the phyllosoma susceptibility.

Conclusion
We report the investigation on a disease outbreak in a lobster hatchery. The pathogens were identified to a species level based on biochemical and molecular methods. The study revealed that Vibrio spp. are the primary pathogens which can lead to a wipe out of almost the whole stock. The findings of this study will help in the development of prophylactic measures which will improve the larval survival in the hatchery phase.