In shrimp, the microbiota plays an important role in the metabolism, besides influencing the control and proliferation of some pathogens (Holt et al., 2021; de Paiva Maia et al., 2013). The principal changes in shrimp microbiota can be related to factors such as water temperature, pH, salinity, sulfide concentration, and diet components. In aquaculture, some bacteria are used as probiotics due to their ability to modify the microbiota composition of the shrimp intestine, decreasing the proliferation of some pathogenic bacteria such as the Vibrio genus (Van Hai and Fotedar, 2010). In this regard, infections caused by Vibrio are one of the most recurrent problems faced by commercial fish farms and marine invertebrates (Leyton and Riquelme, 2008).

In 2009, the aquaculture industry experienced the worst disease outbreak ever, caused by unknown pathology, which generated losses of over one billion dollars (FAO, 2013). This pathology was named “early mortality syndrome” (EMS); the ponds infected with this pathology present massive shrimp mortalities of up to 70 to 100 %. It appears within 30 to 35 days post-stocking of postlarvae into a newly prepared pond; the gross signs of this disease are characterized by the presence of a crystalline hepatopancreas, stomach, and empty intestine as well as erratic swimming. This disease affects the hepatopancreas, the primary organ responsible for the absorption and storage of food nutrients (Hong et al., 2015).

The hepatopancreas is composed of different cell types: E (embryonic), involved in a mitotic activity to produce the other cell types, F (fibrillar), which participates in the secretion of digestive enzymes during feeding, R (absorptive), whose principal function is to absorb diffusible metabolites, and B cells (blister-like) involved in intracellular digestion and assimilation (Wang et al., 2014; Vogt, 2019). The progressive degeneration of the B-, R-, E-, and F- cells can be observed by histological tests in shrimp undergoing AHPND (Tran et al., 2013; Hong et al., 2015). The disease rapidly spread from China to Vietnam in 2010, Malaysia in 2011, Thailand and Mexico in 2013, the Philippines in 2015 (FAO, 2013; Tran et al., 2013; Dabu et al., 2015), Central America in 2015 (Han et al., 2015a), the USA in 2017 (Dhar et al., 2019), Korea in 2018 (Hanet al., 2020), Japan in 2020 and is suspected in others countries in both Asia and Latin America, although not reported (Tang et al., 2020).

In 2013, the causal agent of this pathology was identified as a strain of Vibrio showing close similarity to Vibrio parahaemolyticus. Based on those results, the EMS was renamed Acute Hepatopancreatic Necrosis Disease (AHPND). These strains can produce a binary toxin called PirABvp, and the genes involved in their production were identified in a ~ 69 Kb pVPA3-1 plasmid detected in V. parahaemolyticus strain 13-028/A3 (Yang et al., 2014; Han et al., 2015). These binary toxins found in AHPND-causing V. parahaemolyticus show homology to the PirA and PirB toxins found in Gram-negative enterobacteria Photorhabdus luminescens.

The study of this pathology focuses mainly on isolating strains on semi-selective culture media, thus ruling out the intervention of other non-cultivable bacteria and their possible role within this pathology. However, current technologies can generate a large amount of sequence data from a single sample (Weinstock, 2013). Recently, Next Generation Sequencing (NGS) and universal primers have been used for metagenomics analysis of the 16S rRNA gene, which has allowed the identification of bacterial diversities in different samples from clinical, ecological, marine fields, etc. (Salipante et al., 2013; Staley et al., 2013; Větrovský and Baldrian, 2013). The present study was conducted to analyze the influence of AHPND on the shrimp hepatopancreatic microbiota.

From the total number of samples (186) analyzed by PCR, eight were positive to AHPND (four for each farm). Sequencing was performed for the libraries prepared from the positive AHPND samples and two non-AHPND. According to the sequence data, the microbiota in the non-AHNPD group hepatopancreas is composed mainly of phylum Proteobacteria, Firmicutes, Actinobacteria, Cyanobacteria, Bacteroides, Fusobacteria, and Thermi in both farms (Fig. 1). Although the non-AHNPD group hepatopancreas of both farms shares the same phyla, the relative abundance for each phylum is different between each farm, finding a higher percentage of Proteobacteria and Actinobacteria in samples from farm 1 and most abundance of Firmicutes in samples from farm 2. Meanwhile, the AHPND group in both farms showed higher bacterial diversity of proteobacteria than those from non-AHPND.

Figure 1

The microbiota in non-AHPND group hepatopancreas at the genus level was composed of Vibrio spp. (10.43 - 26.14 %), Staphylococcus spp. (1.60 - 17.13 %), Pseudomonas spp. (7.55 - 12.86 %), and Propionibacterium spp. (1.63 - 5.97 %), with variations on their relative abundance (%) between farms (Figs. 2 - 3). It is worth mentioning that genera Atopobium spp., Lactobacillus spp., Fusobacterium spp., Mycobacterium spp., Tannerella spp., and Parabacteroides spp. were only found in farm 1 (Fig. 2), and genera Corynebacterium spp., Thermus spp., Acinetobacter spp., Bacillus spp., Ochrobactrum spp., Leuconostoc spp., and Amycolatopsis spp. were detected only in farm 2 (Fig. 3), in non-AHPND group hepatopancreas. In addition to this, it was observed that the organisms infected with AHPND (AHPND group) had lower bacterial diversity than the non-AHPND group, and the relative abundance of the Vibrio genus triggered for both farms (Figures 2 - 3). For the case of farm 1, other genera were more abundant, specifically Corynebacterium spp. (8.50 %), Weissella spp. (5.86 %), and others (21.4 %) when compared with farm 2. Regarding alpha diversity, the Shannon index revealed a lower diversity in shrimp infected with AHPND (0.96) than in those negative to the pathogen (1.43).

Figure 2

Figure 3

In this study, Proteobacteria was the phylum with the highest abundance in the non-AHPND and AHPND groups, regardless of the sampled farm (Fig. 1). Remarkably, the higher abundance of Proteobacteria in the AHPND group is due to the abundance increased of the genus Vibrio spp. (Figs. 2 - 3). Similar studies on shrimp microbiota have shown the predominance of Proteobacteria in different parts of the shrimp digestive tract. Research carried out by Gainza et al. (2017) and Garibay-Valdez et al. (2020) showed a higher prevalence of the Proteobacteria phylum in the shrimp intestine at the nursery phase and harvest stages. Rungrassamee et al. (2014) demonstrated the predominance of Proteobacteria in the intestine of shrimp Penaeus monodon in wild-caught and domesticated in Thailand. The natural distribution of the bacteria belonging to the phylum Proteobacteria in seawater and its presence in the sediments of culture is the principal pathway of infiltration of these bacteria in different parts of the shrimp digestive tract (Sung et al., 2001; Velazquez-Roman et al., 2012).

Phylum Firmicutes represented the second most abundant phylum in the hepatopancreas of the non-infected shrimp. Some bacteria belonging to this phylum are considered probiotics in aquaculture, improving the shrimp’s immune system, and controlling pathogens; therefore, its high abundance could be attributed to the feeding practices during shrimp production (Qi et al., 2009; Ruiz-Zarzuela and Blas, 2013; Vargas-Albores et al., 2016).

On the other hand, the phylum Firmicutes were reduced in several samples of shrimp infected with AHPND, particularly compared with Proteobacteria. This reduction was mostly observed in samples HP7 and HP9 from farm 2 (Fig. 3), where the phylum Proteobacteria is the most abundant. Although Firmicutes bacteria were not eliminated in infected shrimp, it was evident that the disease affected this phylum. Similarly, Rungrassamee et al. (2016) studied the bacterial dynamics in the intestines of the black tiger shrimp under the pathogenic Vibrio harveyi challenge, showing the prevalence of four major phyla Firmicutes Actinobacteria, Bacteroidetes, and Proteobacteria, where Proteobacteria predominated, and the others showed less prevalence.

The third phylum with the higher abundance in all samples of non-AHPND hepatopancreas was Actinobacteria (Fig. 1). Although the participation of this phylum in animal nutrition and the production of some bioactive components has been addressed in aquaculture (Das et al., 2008), it showed a considerable reduction when shrimp were infected with AHPND. This reduction of some genera, such as Atopobium spp. (8.23 to 1.54 %) (Fig. 2), Amycolatopsis spp. (0.99 % to undetected), Corynebacterium spp. (24.24 to 0.62 %), and Propionibacterium spp. (5.97 to 0.07 %) (Fig. 3), might be related to poor nutrients absorption, which generates a slow growth, being one of the main symptoms of AHPND, which could have favorable conditions for the development of this disease.

To understand the hepatopancreas microbiota influence in AHPND disease, it is important to understand the composition at the genus level in the non-AHNPD group. Based on the PCA analysis, it is evident that the microbiota composition differs significantly between healthy organisms (HP1, HP6) and infected ones (HP2-HP5, HP7-HP10). Vibrio spp., which was found to be more abundant in all samples, has a significant influence and strongly determines the direction (vector) of the microbiota. Other microorganisms, such as Corynebactrrium spp., Propionibacterium spp., Pseudomonas spp., and Streptococcus spp., were also abundant and influential, but the presence of Vibrio spp. appears to have a more significant impact (Fig. 4).

Figure 4

A microbiota with a diverse relative abundance among genera was observed, registering Vibrio spp. as the most abundant genus (10.43 - 26.14 %), followed by Pseudomonas spp. (7.55 - 12.86 %), Propionibacterium spp. (2.25 - 5.97 %), among others in both farms (Figs. 2 - 3). In some marine species, such as crabs, oysters, and fish, the microbiota is composed of a wide bacterial diversity, which varies depending on the developmental stages (Ward et al., 2009; Trabal et al., 2014), and this diversity is related to the state of health (Martin-Antonio et al., 2007; Xiong et al., 2016).

The presence of PirABvp toxins was determined by PCR analysis; this detects the toxin-producing genes in plasmid pVPA3-1, so the Vibrio genus found in the non-AHPND do not present this plasmid. However, the Vibrio genera found in the AHPND group carry the genes producing PirABvp toxins. The genus Vibrio exceeds more than 50 % relative abundance in farm 1 samples, reaching 82.15 % (Fig. 2); on farm 2, the samples exceeded more than 95 % relative abundance (Fig. 3). The ability of the Vibrio species to produce PirABvp toxins generates an advantage over the various bacterial genera present in the hepatopancreas; this may be involved in such a high reduction in the abundances relative to the other genera. Previous works have considered V. parahaemolyticus as the causative agent of AHPND (FAO, 2013; Tran et al., 2013; Kondo et al., 2014). Other studies have mentioned that AHPND has occurred not only due to V. parahaemolyticus but also to V. owensii and V. harveyi, which have acquired the ability to reproduce this pathology (Kondo et al., 2015; Liu et al., 2018).

Based on the Figure 5, it is evident that the organisms infected with AHPND can be grouped based on their microbiota, which clearly distinguish them between the healthy and uninfected samples (pirABVP genes not detected). The microbiota found in healthy organisms or those without AHPND is different from the one found in organisms with AHPND (HP1, HP6). This is because they are grouped in a separate vector or component that explains 34.5 % of the variance (CP1+CP2). This correlation can be observed and confirmed through the Clustered Heat Map (Fig 6 - supplementary material).

Figure 5

In this study, some bacterial pathogens were detected at the species level, belonging to the genus Vibrio related to diseases of aquatic animals (unpubl. data). Our research found that the increased numbers of Vibrio in shrimp undergoing AHPND influenced the bacterial structure at the genus level.

Finally, the AHPND group showed a minor presence of some bacteria from the genus Geobacillus spp. (0.13 %), Brachybacterium spp. (0.12 %), Streptosporangium spp. (0.10 %), Actinomyces spp. (0.10 %), and Drosophila spp. (0.09 %) when the genus Vibrio is more abundant. In contrast, these genera of bacteria were not found in the microbiota of the non-AHPND group, suggesting the possibility of other opportunistic bacteria colonizing the hepatopancreas when the relative abundance of Vibrio species exceeds 90 %. In addition, the Shannon index in the AHPND organism confirmed the decrease in microbiota diversity, which in addition to the modification of the taxonomic profile, indicates a dysbiosis that could harm the host’s health since the microbiota is considered an annexed organ that provides vital functions (Xiong et al., 2016). It is important to mention that microbiota found among the farms identified at the genus level, showed a difference in the species and their relative abundance; this variation may be attributed to factors such as the location of the farms and the type of feeding provided by the farms and others.

In conclusion, the non-AHNPD and AHPND hepatopancreas microbiota is mainly composed of the phylum Proteobacteria, Firmicutes and Actinobacteria, considered core microbiota. The presence of the pirABvp toxins, causal of AHPND, is a clear advantage of the Vibrio genus bacteria over the other microorganisms present, although some of these have diverse functions in shrimp development. Only some bacteria of the phylum Firmicutes remained in the presence of the Vibrio genus; these could be used as an alternative to attenuate the dysbiosis caused by the pathogen. Understanding the microbiota composition brings new insights into controlling and reducing the risk of the appearance of AHPND in shrimp culture ponds.