Introduction

Pseudomonas species, a group of gram-negative bacteria, are widely distributed in aquatic environments and are known to be associated with stress-related diseases in farmed fish, particularly in intensive culture systems. These bacteria have been identified as the primary cause of fish diseases, leading to significant economic losses, with reported mortality rates reaching 100% in valuable fish species such as rainbow trout (Oncorhynchus mykiss), sea bream (Sparus aurata), sea bass (Dicentrarchus labrax), and ayu (Plecoglossus altivelis) (López et al. 2012; Derome et al. 2016). The most commonly isolated species include Pseudomonas luteola, P. anguilliseptica, P. baetica, P. fluorescens, P. aeruginosa, P. koreensis, P. plecoglossicida, P. pseudoalcaligenes, P. putida, and P. chlororaphis (Austin and Austin 2016; Sherif et al. 2021).

P. anguilliseptica, a bacterium commonly found in aquatic environments, is characterized as a Gram-negative microorganism. It exhibits fastidious characteristics, displaying slow growth and weak reactions when cultured on bacterial media commonly employed in microbiology laboratories (Daly 1999; López-Romalde et al. 2003). P. anguilliseptica is an opportunistic microorganism responsible for causing a disease commonly referred to as red spot disease or “sekiten-byo” in Japan. This pathogen was initially documented in pond-farmed Japanese eel (Anguilla japonica) (Wakabayashi and Egusa 1972). Additionally, it is considered a prevalent infectious pathogen among farmed marine and brackish water fishes (Daly 1999). Outbreaks of P. anguilliseptica in farmed gilt-head seabream have been reported in several Mediterranean countries, including France, Portugal, and Spain. These outbreaks have resulted in a disease known as hemorrhagic septicemia, associated with high mortality rates and significant financial losses for the aquaculture industry (Doménech et al. 1999; Romalde et al. 2001; Ali et al. 2022). Similar occurrences have also been reported in other fish species such as turbot (Scophthalmus maximus), black sea bream (Acanthopagrus schlegeli), European eel (Anguilla anguilla), ayu (P. altivelis), sea trout (Salmo trutta), rainbow trout (O. mykiss), Atlantic salmon (Salmo salar), striped jack (Pseudocaranx dentex), whitefish (Coregonus sp.), Baltic herring (Clupea harengus membras), and orange-spotted grouper (Epinephelus coioides) (Al-Marzouk 1999; Austin and Austin 1999; Daly 1999; López-Romalde et al. 2002).

Identifying these bacterial species using microbiological analyses is challenging due to slow growth, low metabolic activity, and variability observed in their morphology and biochemical characteristics among different isolates. Consequently, molecular techniques accurately identify and distinguish P. anguilliseptica from other closely related genera (Mulet et al. 2012; Palleroni 2015). These techniques include 16 S rRNA sequencing for genus identification and multilocus sequence analysis (MLSA) for species classification (Mulet et al. 2012; Gomila et al. 2015; Abdelsalam et al. 2022). The multilocus sequence analysis (MLSA) technique was developed to assess molecular variations in multiple protein-coding housekeeping genes to compare closely related bacterial species (Mjølnerød et al. 2021). Multiple species of bacteria were subjected to standardization using MLSA, and the genetic variations derived from conserved housekeeping genes were revealed in the phylogenetic relationship. The MLSA technique is the most reliable method for studying the epidemiology, geographical distribution, and characterization of Pseudomonas spp. It can distinguish between different pathogenic strains of this pathogen (Sánchez et al. 2014; Gomila et al. 2015). The MLSA technique, which involves partial sequencing of four housekeeping genes (rpoB, rpoD, gyrB, and aroE), effectively identifies Pseudomonas spp. This technique has provided valuable insights into the phylogenetic relationships among different Pseudomonas strains and enabled discrimination between them (Mulet et al. 2012; Gomila et al. 2015).

Recently, antibiotic-resistant bacteria were frequently isolated from aquatic animals (Naik et al. 2018), aquatic water, and sediment (Pham et al. 2018; Shah et al. 2012). In Egypt, aquatic farmers and paramedics routinely use different antibiotics without a veterinary consultant during disease outbreaks, resulting in emerging antibiotic-resistant bacteria (WHO 2006; Sherif et al. 2021, 2022).

Over the past year, multiple reports of fish mortalities have occurred in farms that specifically reared seabass and seabream.

The objective of the current study was to conduct phenotypic and genetic characterizations of P. anguilliseptica strains isolated from moribund seabream and sea bass. Furthermore, the study aimed to investigate the weak response to antibiotic treatment. In addition, the quality of fish farm water associated with this particular bacterium was examined.

Materials and methods

Fish farms and samples collection

Specimens of moribund sea bream (Sparus aurata) and sea bass (Dicentrarchus labrax) were obtained from a collective of four fish farms located in the governorates of Dameitta (comprising three farms) and Kafrelsheikh (comprising one farm). These particular fish farms encountered high mortality rates during the winter season in November 2022. One hundred moribund fish were concurrently collected from affected fish ponds with water samples in triplicates, as described by Sherif et al. (2023) and Eldessouki et al. (2023). Water samples were collected at a depth of 0.5 m using aseptic techniques to guarantee the absence of contamination in the collected samples. The investigation was conducted to determine the presence of bacteria in fish samples. Bacterial cultures were obtained from the internal organs, specifically the kidney, spleen, and hepatopancreas tissues, using the protocols described by Woo and Bruno (2014) and Sherif and Kassab (2023).

Bacterial analyses

Bacterial isolation

Pooled specimens of livers, kidneys, and spleens, each fish species per farm considered one sample, were streaked on various agar mediums. Subsequently, these samples were incubated at 25 °C for 24 to 72 h (Wakabayashi and Egusa 1972). The agarmedia used consisted of tryptic soy agar (TSA) supplemented with 3% sodium chloride and 5% sheep blood agar. The prevailed colonies were selected and purified by subculturing onto TSA under the same conditions. Isolates were confirmed by selective isolation on Salmonella-Shigella agar SS agar. Gram staining, oxidase test, catalase activity, lactose fermentation on MacConkey agar, and motility testing using motility mediums were performed for the preliminary identification of bacterial isolates. The stock cultures were maintained at −70 °C in tryptic soy broth (Difco Laboratories), supplemented with 1% sodium chloride and 15% glycerol. The biochemical analyses were conducted in triplicate using API20 E following the guidelines provided by the manufacturer (BioMerieux, Marcy l’ Etoile, France) and the recommendations presented by Madigan and Martinko (2005). In this work, the biosafety protocols adhered to the Pathogen Safety Data Sheet for Infectious Substances-P. Anguilliseptica, as the Pathogen Regulation Directorate of the Public Health Agency of Canada (2022) outlines.

Molecular identification

DNA extraction

Genomic DNA from the pure 14 isolates of Pseudomonas spp. was extracted using the QIAamp DNA mini kit (Qiagen, Hilden, Germany, Cat. no. 513) following QIAGEN’s recommended protocol. The extracted DNA was eluted using the kit’s 100 µl elution buffer. The concentration and integrity of DNA were determined using the Nanodrop spectrophotometer following the manufacturer guidelines at 260/280 nm (ND-1000 UV-VIS, Thermo Scientific, USA). The extracted DNA was stored at − 20 °C for subsequent PCR amplifications and sequencing.

PCR amplifications

The identity of Pseudomonas spp. isolates was initially confirmed by amplifying and sequencing the 16 S rRNA gene using the universal primer pair 9 F and 1541R (Table 1), following the established protocol described by Osborne et al. (2005). For more in-depth phylogenetic analysis, four housekeeping genes—gyrB (703 bp), rpoB (529 bp), rpoD (766 bp), and aroE (653 bp)—were individually PCR amplified from each Pseudomonas isolate using multilocus sequence typing (MLST). The MLST primers were adopted from Mjølnerød et al. (2021). Additionally, bacterial strains were screened for resistance to tetracycline, erythromycin, sulfonamide, and quinolone by detecting the presence of resistance genes tetA, qnrs, sul1, and ermB, respectively, using PCR as described by Randall et al. (2004). All primers used in this study are documented in Table 1.

Table 1 The primers used in this study
Full size table

In brief, PCR amplifications were conducted in 25 µl total reaction volumes containing 4 µl DNA template (50 ng), 12.5 µl MyTaq Red Mix (Bioline), 1 µl of each primer (20 pmol), and 6.5 µl nuclease-free water. Each gene (16 S rRNA, gyrB, rpoB, rpoD, aroE, sul1, tetA, qnrs, and ermB) was individually amplified from every Pseudomonas isolate. Thermocycling conditions began with initial denaturation at 95 °C for 4 min followed by 35 cycles of denaturation at 95 °C for 1 min, annealing for 1 min, and extension at 72 °C for 1 min. Annealing temperatures were adjusted based on the target gene, specifically 52 °C for 16 S rRNA, 56 °C for the four MLST genes gyrB, rpoB, rpoD, and aroE, 55 °C for tetA, 68 °C for sul1, 53 °C for qnrs, and 62 °C for ermB. A final extension step at 72 °C for 7 min completed the PCR. Amplicons were purified using the QIAquick PCR Purification Kit (Qiagen) per the manufacturer’s instructions. Purified products were Sanger sequenced in both forward and reverse directions using the ABI 3730xl DNA Analyzer (Applied Biosystems). Raw sequence reads were trimmed, edited, and assembled into contigs using BioEdit version 7.2.5 (Hall 1999).

For species identification, the assembled 16 S rRNA sequences were aligned against other 16 S rRNA sequences of Pseudomonas spp. using the online BLAST tool of NCBI. The obtained 14 sequences were deposited in the database of GenBank under the following succession accession numbers starting from ON926995 to ON927008 (Tables 2 and 3). The identity of Pseudomonas isolates was confirmed at the species level when the BLAST similarity reached ≥ 99 % to the relevant 16 S rRNA sequences available in the GenBank database, as illustrated by Drancourt et al. (2000). The multiple sequence alignments were carried out using the MEGA program to match the interspecies and intrastrain similarities index (Kumar et al. 2018). The neighbor-joining phylogenetic analysis of the current 16 S rRNA sequences was performed against 14 typing strains including, P. anguilliseptica (3 isolates), P. resinovorans (1 isolate), P. aeruginosa (1 isolate), P. asplenii (1 isolate), P. monteilii (1 isolate), P. tolaasii (1 isolate), P. fluorescens (2 isolates), P. chlororaphis (2 isolates), and P. putida (2 isolates).

Table 2 Bacterial growth in different incubation temperatures and salinities
Full size table
Table 3 Pseudomonas anguilliseptica isolated in this study
Full size table

Multilocus sequence analysis

The raw sequences of gyrB, rpoB, rpoD, and aroE genes were trimmed and edited using BioEdit software ver. 7.2.5 (Hall 1999). The assembled sequences were submitted to the GenBank database under the following accession numbers: gyrB (OQ694532- OQ694545), rpoB (OQ726020- OQ726033), rpoD (ON981406- ON981419), and aroE (OQ744508- OQ744521). The final sequences were imported and concatenated in Microsoft Excel. The phylogenetic analysis of the concatenated sequences of housekeeping genes was constructed by neighbor-joining methods of the MEGA 11 with the pairwise deletion option (Kumar et al. 2018). The node supports were estimated using nonparametric bootstrapping with 1000 replicates and Kimura’s two-parameter model was applied.

Antibiogram of Pseudomonas anguilliseptica strains

In line with an earlier study (CLSI 2010) and using the disc diffusion method, bacterial strains were examined for their sensitivity to some antibiotics on Mueller-Hinton agar (Oxoid™). The antimicrobial agents utilized in the assessment were tetracycline (TetA) 30 µg, ciprofloxacin (Cip) 5 µg, trimethoprim/sulfamethoxazole (SXT) 1.25/23.75 µg, erythromycin (E, 15 µg), florfenicol 30 µg, gentamicin 10 µg, amoxicillin 10 µg, kanamycin 30 µg, cefotaxime 30 µg, ampicillin (Amp) 10 µg, and streptomycin 30 µg. The test was conducted in triplicates, and isolates of P. anguilliseptica were subcultured into tryptic soy broth, then incubated overnight at 30 °C, and placed onto Mueller-Hinton agar plates using a cotton swab. Following incubation for 24 h at 30 °C, the diameter of the inhibition zones was measured, and the results were interpreted based on the Clinical Laboratory Standards Institute criteria. The equation determined the multidrug resistance (MDR) index:

$$\text{MDR} = X/Y$$

x represents the number of antibiotic categories to which the isolates were resistant, while y represents the assessment of antibiotics. If the obtained MDR index is greater than 0.2, the bacterial strain is resistant to multiple antibiotics (Krumperman 1983).

Examination of water physicochemical parameters

The temperature and salinity of water samples at farm sites were analyzed using a YSI Environmental (Model EC300), while dissolved oxygen (DO) was analyzed using an Aqualytic (Model OX24), and pH was analyzed using a Thermo Orion (Model 420 A). The samples were collected in 1-L polyethylene bottles and transported on ice to the laboratory, namely, total ammonia nitrogen (TAN), unionized ammonia (NH3), nitrite (NO2), and nitrate (NO3); a UV/visible spectrophotometer (Thermo-Spectronic 300) was used for ammonia compound analyses following the recommendations of Rice and Bridgewater (2012).

Statistical examination

Different parameters were assessed using the ANOVA test and Duncan’s Multiple Range (Duncan 1955) by determining the mean and standard error of the collected data using SPSS software (SPSS 2004). At P ≤ 0.05, values were considered statistically significant.

Results

Clinical signs and post-mortem of farmed sea bream and sea bass infected with P. anguilliseptica

In Fig. 1, the moribund sea bream showed signs of septicemic bacterial infection. These were hemorrhages on the external body surface behind the pectoral fin and abdomen with ulcers under the dorsal fin. Postmortem signs were hemorrhages on the liver surface and splenomegaly (rounded surface).

Fig. 1
figure 1

Clinical signs and post-mortem of farmed sea bream infected with P. anguilliseptica. A Arrows pointed to hemorrhages behind the pectoral fin and abdomen with ulcers under dorsal fine, B brown-reddish liver and splenomegaly, C brown liver with hemorrhages and splenomegaly, D the same fish of C with clear splenomegaly (rounded) and abdominal fat

Full size image

In Figs. 2 and 3, moribund sea bass showed hemorrhages on the fish skull, eye loss, hemorrhages on the external body surface, and gills with faint red. Post-mortem revealed signs of hemorrhagic diseases: light-brownish liver, hemorrhages on liver and internal surfaces, abdominal fat with splenomegaly, and hemorrhagic-distended kidney.

Fig. 2
figure 2

Clinical signs and post-mortem of farmed sea bass infected with P. anguilliseptica. A Hemorrhages on fish skull, B hemorrhages on abdomen and behind gill cover, C light-brownish liver with splenomegaly, and D hemorrhagic and distended kidney

Full size image
Fig. 3
figure 3

Clinical signs and post-mortem of farmed sea bass infected with P. anguilliseptica. A Arrows pointed to hemorrhages on liver and internal surfaces abdominal fat and splenomegaly, B eye lost, C gills with faint red color and splenomegaly

Full size image

Bacterial growth on agar plate

The isolated bacteria were negative for gram stain, and by using API 20E, they had the number 220,004, which represented the Pseudomonas spp. The bacterial growth was evaluated in different incubation temperatures (8, 25, 37, and 42 °C) and salinity (0, 1, 3, 5 g/L) (Table 2). No growth was detected in all bacterial strains at 37 and 42 °C, while NDA0208 and NDA0209 (both are from Damietta farms) and all bacterial isolates of Kafrelshiekh farms (NKF0301-5) showed weak growth. All P. anguilliseptica strains were unable to grow in 0 g/L salinity. Also, all strains isolated from Damietta farms showed weak growth in 1 g/L salinity, while those recovered from Kafrelsheikh did not grow. All strains could grow in 3 and 5 g/L, except those recovered from Kafrelsheikh had weak growth in 5 g/L.

Sequencing of 16 S rRNA and MLST

The amplified 16 S rRNA gene was successfully performed from 14 Pseudomonas spp. isolates using the gene-specific primers pair. The purified PCR products were then sequenced to confirm the identities of Pseudomonas spp. at the species level. The successive accession numbers of these sequences (ON926995 to ON927008) were issued by the GenBank staff (Table 3). The BLAST analysis of 16 S rRNA gene sequences proved that all 14 bacterial strains belonged to the genus Pseudomonas, and all isolates were identified as P. anguilliseptica. The intrastrain nucleotide differences within the isolates of P. anguilliseptica varied from 99.80 to 99.39% with three to nine nucleotide differences. The comparative sequence analysis of the current 16 S rRNA sequences (ON926995 to ON927008) revealed a high similarity index ranging between 99.86 and 99.39% with relevant sequences of P. anguilliseptica (AB021376T, LC071931, LC194236, LC071928, LC071927, and LC071926). The alignment of these accession numbers (ON926995 to ON927008) showed 96.73–95.45% 16 S rRNA similarity to typing isolates of the following Pseudomonas spp.: P. resinovorans (NR_112062) T, P. aeruginosa (AY268175) T, P. asplenii (NR_040802) T, P. monteilii (KU057954) T, P. tolaasii (NR_114481) T, P. fluorescens (AF094731) T, P. chlororaphis (FJ652608) T, and P. putida (AF094743) T.

In Fig. 4, the phylogenetic analysis of the 16 S rRNA sequences of P. anguilliseptica isolates formed two major clades. The first clade was separated into two subclades. The first subclade involved all P. anguilliseptica, grouped with the sequence of typing P. anguilliseptica (NCIMB 1949) with a substantial 100% bootstrap value. In this study, all isolates of P. anguilliseptica were separated from other Pseudomonas spp. belonging to P. resinovorans, P. aeruginosa, P. asplenii, P. monteilii, P. tolaasii, P. fluorescens, P. chlororaphis, and P. putida. The typing strain of S. dysgalacatiae (ATCC 43,078) served as an outgroup.

Fig. 4
figure 4

The neighbor-joining phylogenetic tree based on the comparative analysis of the 16 S rRNA gene sequences, demonstrating the relationship between P. anguilliseptica isolates in this study and closely related Pseudomonas spp.

Full size image

In Table 3, MLSA allowed unambiguous identification of all recovered Pseudomonas spp. isolates as P. anguilliseptica (14 isolates). All isolates shared 99.62–99.70% concatenated sequence similarity to typing P. anguilliseptica (NCIMB 1949) strain. In addition, the alignment of concatenated sequences of these isolates showed high discriminatory power between P. anguilliseptica of this study and typical isolates of P. chlororaphis subsp. piscium ATCC 17,809 (CP027709.1)T, P. fluorescens ATCC 13,525 (QVNA00000000.1)T, P. aeruginosa ATCC 27,853 (CP015117.1)T, and P. putida ATCC 12,633 (CP101910.1)T, with similarity index ranging between 80.92 and 79.54%. On the other hand, the comparative sequence alignment of four concatenated housekeeping genes (2363 bp) demonstrated similarity scores ranging between 99.75 and 100% among all retrieved isolates.

In Fig. 5, the phylogenetic tree derived from MLSA based on concatenated sequences of 14 isolates displayed five distinct sequence types (ST1, ST2, ST3, ST4, and ST5) (Fig. 2). The phylogeny tree of MLSA revealed that all P. anguilliseptica isolates belonged to three main clusters. Clustering was correlated with the geographical source of the strains. Clusters 1 and 2 were associated with isolates originating from Damietta farms, while cluster 3 was related to isolates recovered from Kafrelsheikh farms. The first clade grouped ST1 and ST2, and the node supported with a bootstrap value of 73%. The second clade gathered ST3, with a bootstrap value of 98%. The third clade grouped ST4 and ST5 and supported with a bootstrap value of 96%.

Fig. 5
figure 5

Phylogenetic analysis of MLSA of fourteen P. anguilliseptica isolates based on the concatenated gyrB, recA, rpoB, rpoD, and aroE gene sequences

Full size image

Antibiotic resistance genes and antibiogram of P. anguilliseptica strains

Concerning the prevalence of ARG, tetracycline (tetA) was the most predominant, followed by erythromycin (ermB), while sulfonamide (sul1) and quinolone (qnrs) did not present in those isolated at Kafrelsheikh farm (Tables 4 and 5).

Table 4 Antibiogram of the isolated P. anguilliseptica strains
Full size table
Table 5 Data on multidrug resistant index and genes of the isolated bacterial strains
Full size table

In Table 4, P. anguilliseptica strains were tested against 11 antibiotic types. All P. anguilliseptica strains were resistant to the selected antibiotics tetracycline, amoxicillin, ampicillin, and kanamycin, whereas they were sensitive to gentamycin, florfenicol, and trimethoprim plus sulfamethoxazole. Antibiogram of P. anguilliseptica strains differed according to the site of collection. A set of tetracycline (TE), ampicillin (AMP), amoxicillin (AML), kanamycin (K), and E was predominant in P. anguilliseptica trimethoprim/sulfamethoxazole (SXT), ciprofloxacin (Cip), and streptomycin (S) not in NKF were presented in those recovered at Damietta farms and absent at Kafrelsheikh farm (Table 5).

Examination of water physicochemical parameters of fish farms

In Table 6, water temperature ranged between 18.03 and 19.87 °C, showing significant differences suitable for bacterial infection. The water salinity was significantly higher in Damietta farm (22.8 to 225 g/L) than Kafrelshiekh farm (19.43 g/L). The level of DO was above 4 mg/L in all farms except farm ND2 at Damietta, which was 3.43 mg/L. The pH levels, which ranged between 8.17 and 8.22, were suitable for fish culture and bacterial growth. Ammonia compounds, both TAN and total ammonium (NH3 + NH4), were significantly higher in ND2 and ND3 compared to investigated farms (0.56 and 0.58; 0.72 and 0.74 mg/L, respectively). With the same trend, NH3 was significantly higher in ND2 and ND3 (0.03 and 0.037 mg/L, respectively). There were no significant differences in NO2 and NO3 in fish farms under investigation.

Table 6 Water parameters of fish farms under investigation
Full size table

Discussion

Clinical and post-mortem signs were recorded in moribund sea bream and sea bass in this study, indicating the occurrence of septicemic-hemorrhagic bacterial disease, which were hemorrhages on the external body surface and internal organs with faint-reddish gills, splenomegaly, and hemorrhagic-distended kidney. In the field studies, general clinical symptoms such as pale skin color, loss of scales, hemorrhages, and ulcers on the body surface, and anemia in the internal organs were observed and showed similarities with previous reports for each particular bacterial fish pathogen (Akayli 2001; Austin and Austin 2016; Roberts 2012).

In this work, microbiological methods of isolation and identification confirm the presence of gram–ve, and identification number 220,004 was obtained using API 20E, indicating the presence Pseudomonas spp. unable to identify the bacteria associated with sea bream and sea bass mass mortality. Similarly, false-negative results were detected in the API 20NE (Balboa et al. 2007), with P. anguilliseptica (López-Romalde 2005; Canak and Akayli (2018). Even when Topic Popovic et al. (2007) developed API 20E kits for pathogenic fish bacteria used in field, they recorded misidentification with P. stutzeri and P. anguilliseptica isolated from moribund gilt-head sea bream. In addition, P. anguilliseptica was misidentified as P. fluorescens/putida with API 20NE kit Canak and Akayli (2018). So, the API 20NE method could be used to screen and confirm the presence of Pseudomonas spp. in sea bream and sea bass.

In this work, no bacterial growth was detected in all bacterial strains at 37 and 42 °C, and no growth was detected at 0 g/L salinity. All strains could grow in 3 and 5 g/L, except those recovered from Kafrelsheikh had weak growth in 5 g/L. Similarly, Pseudomonas spp. could tolerate a wide range of water temperature, pH, and salinity and thus could remain in the aquatic environment; however, nutrient sources are scarce (Liu et al. 2015; Palleroni 2015; Moradali et al. 2017). Inconsistency, Duman et al. (2021) found that Pseudomonas spp. could grow at NaCl concentrations of 0–45 g/L, whereas higher salinity 60, 70, 80, and 90 g/L NaCl resulted in different growth rates of 20%, 61%, 73%, and 91%, respectively, such as strains of P. haemolytica, P. proteolytica, P. lundensis, and P. weihenstephanensis. In addition, it was noted that all Pseudomonas spp. could grow at 37 °C incubation temperature, such as P. mandelii, P. meridiana, P. migulae, and Pseudomonas spp. tolerated 42 °C, and P. haemolytica was heat-tolerant up to 45 °C. It can be easily differentiated Pseudomonas sp. from P. anguilliseptica based on its ability to grow at 41 °C, which P. anguilliseptica cannot (Austin and Austin 1999).

Due to the misidentification of Pseudomonas sp. using traditional bacteria methods, ninety isolates were identified using the phylogenetic method with 4-gene multilocus sequence analysis based on the partial sequencing of the 16 S rRNA, gyrB, rpoB, and rpoD genes (Duman et al. 2021).

In this study, sequencing of 16 S rRNA confirmed that all 14 bacterial strains belonged to the genus Pseudomonas group. By BLAST analysis of 16 S rRNA, Pseudomonas strains were molecularly recognized as P. anguilliseptica (14 isolates).

However, relying on the 16 S rRNA could lead to misidentification in some instances (Elgendy et al. 2022). In this study, sequencing of 16 S rRNA allowed the differentiation of P. anguilliseptica isolates from other typing Pseudomonas spp. isolates of P. resinovorans, P. aeruginosa, P. asplenii, P. monteilii, P. tolaasii, P. fluorescens, P. chlororaphis, and P. putida and showed 96.73–95.45% similarity score index. Mulet et al. (2012) and Gomila et al. (2015) obtained similar results.

On the other hand, the MLSA assay based on four housekeeping genes (gyrB, rpoB, rpoD, and aroE) was done to confirm the identification of isolates and to study the phylogenetic relationships in P. anguilliseptica. In addition, several studies used these five housekeeping genes (16 S rRNA, gyrB, rpoB, rpoD, and aroE) to characterize and epidemiological studies of Pseudomonas spp. In this study, the concatenation of four housekeeping genes provided sufficient discriminatory power to separate P. anguilliseptica isolates regarding sequence divergences from other Pseudomonas spp. A prominent gap between the intraspecies and interspecies MLSA similarities is essential for identifying the isolates at the species level, and a threshold of 97% of MLSA similarity was suggested for species differentiation in the genus Pseudomonas (Mulet et al. 2010, 2012). In this study, P. anguilliseptica isolates shared the highest MLSA similarity (99.62–99.70%) with the typing P. anguilliseptica (NCIMB 1949) and displayed 80.92–79.54% MLSA similarity index with other typing isolates of Pseudomonas spp.: P. chlororaphis subsp. piscium ATCC 17,809 (CP027709.1)T, P. fluorescens ATCC 13,525 (QVNA00000000.1)T, P. aeruginosa ATCC 27,853 (CP015117.1)T, and P. putida ATCC 12,633 (CP101910.1)T. Therefore, the MLSA assay of four HKGs (gyrB, rpoB, rpoD, and aroE) revealed a considerable performance for the identification of P. anguilliseptica. On the other hand, MLSA is superior for resolving the intrastrain relationship among P. anguilliseptica isolates (Mjølnerød et al. 2021).

In this study, P. anguilliseptica strains were tested against 11 antibiotic types. All P. anguilliseptica strains were resistant to the selected antibiotics tetracycline, amoxicillin, ampicillin, and kanamycin, while they were sensitive to gentamycin, florfenicol, and trimethoprim plus sulfamethoxazole. Antibiogram of P. anguilliseptica strains differed according to the fish collection site. The most predominant ARG was tetA followed by ermB and sul1. There was a good correlation between the presence of resistance genes and corresponding resistance phenotypes, suggesting resistance genes, when present, were usually expressed (Randall et al. 2004). Accordingly, the antibiotics used in antibiogram tests were available in the local market; they could also be used in fish treatment (Noga 2010). Similarly, Canak and Akayli (2018) found that the isolated P. anguilliseptica was sensitive to ciprofloxacin and enrofloxacin, whereas it was resistant to 6 of the 10 antimicrobials. Other reports considered Pseudomonas spp. multidrug-resistant bacteria as they were resistant to ampicillin, trimethoprim + sulfamethoxazole, cefotaxime, aztreonam, and nitrofurantoin (Matyar et al. 2008; Devarajan et al. 2017); also, López-Romalde et al. (2003) found that all P. anguilliseptica strains were trimethoprimsulfomethoxazol. In contrast, Haenen and Davidse (2001) recorded that P. anguilliseptica were sensitive to neomycine, trimethoprim, trimethoprimqsulphadiazine, sulphonamids, oxytetracycline, flumequine, and oxolinic acid. These differences are attributed to use or lack of the use of these antimicrobials.

The intensive culture of sea bream in Egypt resulted in a high infection rate of pathogenic bacteria that grow well in brackish water ponds 23.5 to 26.5 g/L (Eisawy and Wassef 1984). In this work, water temperature ranged between 18.03 and 19.87 °C, while salinity was significantly higher in Damietta (22.8 to 22.5 g/L) than in Kafrelshiekh farm (19.43 g/L); these parameters were suitable for bacterial infection. Similarly, Haenen and Davidse (2001) isolated P. anguilliseptica at 23–25 °C in diseased European elvers that were maintained in Netherlands. These findings could be due to the decline in antibody production. Accordingly, Yamagachi et al. (1980) found that fish were immunosuppressed in response to a decline in water temperature and salinity because of a decrease in antibody formation as antibody production was seasonality dependent. In addition, the fluctuation of water temperature suppressed the fish immunity, causing them to become more vulnerable to P. anguilliseptica infection (Tort et al. 1996), so mortality ceased by raising the water temperature above 26 °C (Muroga et al. 1973; Haenen and Davidse 2001). Accordingly, bacterial disease outbreaks were recorded in fish farms at a water temperature of 16–18 °C (Kim et al. 2010), whereas the disease persisted in sea bream until the water temperature increased to 18–20 °C (Doménech et al. 1999).

The level of NH3 was stressful for fish in the examined fish farms according to standard values recommended (0.002–0.05 mg/L) by FAO (2023). Ammonia compounds were not statistically correlated with isolated bacteria strains, antibiotic resistance, MDR index, or ARGs. Accordingly, the ammonia compounds accumulated did not significantly affect the bacterial community or ARG composition; they might still cause immunosuppression status in the fish and other metabolic processes. These effects could render the fish more vulnerable to bacterial infections (Sreedevi and Ramasubramanian 2011; Rousselet et al. 2018; Sherif and AbuLeila 2022). Other findings, it was found that P. anguilliseptica infection is interrelated with winter disease (WD) syndrome in sea bream (Doménech et al. 1999); this syndrome is a multifactorial disease associated with many factors, mainly deteriorated water quality (high ammonia compound and heavy metals pollution) which could cause immunosuppressed fish (Tort et al. 1998).

Conclusion

In this study, MLSA was used to type fourteen P. anguilliseptica isolates. Bacterial infections were accompanied by low water temperature and fluctuation of water salinity. Despite the stress levels of unionized ammonia (NH3) in fish farm water, no statistical correlation was observed with bacterial infection. All P. anguilliseptica strains could grow in 3 and 5 g/L, except those recovered from Kafrelsheikh farm, which had weak growth in 5 g/L. Isolates were sensitive to florfenicol, trimethoprim, sulfamethoxazole, cefotaxime, and ciprofloxacin. In addition, tetracycline tetA and erythromycin ermB genes were the highest-presentedantibiotic-resistant genes.