The genus Shewanella spp. belongs to the phylum Pseudomonadota, class Gammaproteobacteria, order Alteromonadales, family Shewanellaceae, and genus Shewanella1. To date, over 70 species of Shewanella have been identified2. Among them, those that cause disease in humans are Shewanella algae(S. algae), S. putrefaciens, and S. xiamenensis, while species such as S. indica and S. chilikensis have been identified as closely related to S. algae and belong to the algae clade3. Shewanella spp. produces tetrodotoxin (TTX) and belongs to the TTX-producing bacteria4. All S. algae are haemolytic. Khashe et al. confirmed that S. algae is more pathogenic than S. putrefaciens in mice and speculated that the haemolytic activity of S. algae may be an important virulence factor5. The isolation of Shewanella spp. from food poisoning cases or intestinal samples of diarrhea patients has been previously reported6,7,8,9.

The incidence of Shewanella spp., an opportunistic pathogen newly included in China’s List of Pathogenic Microorganisms Transmitted From Human to Human(2023 edition), in clinical infections has increased worldwide. The widespread transmission of Shewanella spp. poses a significant challenge to public health and clinical anti-infective treatment. Yu et al.10 classified diseases of Shewanella infections into eight broad categories based on the site of infection, namely ear, nose, and throat (E.N.T) disorders, central nervous system (CNS) disorders, chest infections, cardiovascular diseases, bloodstream infections (bacteraemia and sepsis), intra-abdominal infections, osteoarthropathy, and skin and soft-tissue infections (SSTIs). In recent years, reports of sporadic cases of infections caused by Shewanella spp. have been on the rise due to improved clinical microbiological testing techniques11. Among patients with Shewanella infections, 43.59% had been exposed to the marine environment. The population infected with Shewanella spp. is mostly elderly and neonates, with a male to female ratio of 2.84:110. This study was performed based on the investigation and pathogen characterization of seven Shewanella spp. positive cases detected in diarrhea surveillance in Beijing, China in 2023. It aims to establish a foundational basis for subsequent research regarding diarrhea caused by Shewanella spp.

Results

Epidemiological findings of Shewanella spp.-positive cases

A total of 354 stool samples from 354 diarrhea patients were collected, and the pathogen detection rates are as follows: Shewanella spp. 1.98% (7/354), Salmonella spp. 1.98% (7/354), Shigella spp. 0% (0/354), diarrheagenic Escherichia coli 8.19% (29/354), Vibrio parahaemolyticus 3.67% (13/354), Campylobacter jejuni 3.67% (13/354), Campylobacter coli 2.26% (8/354), Yersinia enterocolitica 1.41% (5/354), norovirus 29.94% (106/354), sapovirus 7.06% (25/354), rotavirus 0.03% (1/354), enteric adenovirus 8.47% (30/354), and astrovirus 8.19% (29/354). Seven strains of Shewanella spp. (named as S1-S7) were isolated from seven patients (named as P1-P7). The onset of P1 was July 1, 2023, while that of P2-P7 was July 17–22, 2023. No pathogens other than Shewanella spp. were detected in P1-P7.

Among the six patients with clustered onset times (P2-P7), there were five males and one female, and the six patients did not know each other. They were aged between 15 and 58 years old, and the incubation period ranged from 8 to 12 h. The frequency rates of clinical symptoms was 100% (6/6) for diarrhea, 50.00% (3/6) for nausea, 33.33% (2/6) for fever, 16.67% (1/6) for abdominal pain, 16.67% (1/6) for dehydration, 16.67% (1/6) for thirst, 16.67% (1/6) for vomiting, and 16.67% (1/6) for rectal tenesmus. The frequency of diarrhea episodes ranged from three to 50 times/day, and the faecal characterization was watery stools in 83.33% (5/6) patients and bloody purulent stools in 16.67% (1/6) patients, respectively. Suspected contaminated food consumed by P3 and P6 was purchased from the same market (Market X in Beijing). P4 and P5 both travelled to the tourist attraction Y in city B on the same day (the attraction is 240 km away from Beijing, and P4 and P5 did not know each other, were not in the same tour group, nor had any history of dining together) and dined at the attraction site. The suspected contaminated food product was seafood for both patients. The suspected contaminated food products that the six patients were exposed to were all purchased from the catering service establishment or retail market. The suspected contaminated food products consumed were seafood for three patients, cold drinks for two patients, and cold watermelon for one patient. P4, P5, and P7 all dined with another individual, who also contracted diarrhea (Table 1). The epidemiological data were unavailable for P1 as the call was disconnected.

Identification of isolated Shewanella spp.

Based on the average nucleotide identity (ANI) results, strains S1, S2, and S4 were identified as S. indica, strains S3, S6, and S7 were identified as S. chilikensis, and S5 was identified as S. algae(Fig. 1). The identification results of S1-S7 based on the three methods of matrix-assisted laser desorption ionisation-time of flight mass spectrometry (MALDI-TOF MS), fully automated bacterial biochemical identification instrument. Both MALDI-TOFMS and ANI results are shown in Table 2.

All seven Shewanella strains were of novel sequence typing (ST) types, with S1 being ST77, S2 and S4 being ST78, S3, S6, and S7 being ST79, and S5 being ST81, as also shown in Table 2. A total of 125,738 single nucleotide polymorphism (SNPs) were identified in the core genomes of the seven Shewanella strains, where the number of paired SNPs for S2-S4 and S3-S7 was 0, and the number of paired SNPs between S3-S7-S6 was 0–1, as shown in Fig. 2A. In the maximum likelihood phylogenetic tree constructed based on core genome single nucleotide polymorphism (cgSNP), S3, S6, and S7 were in the same genetic branch and were close to the reference strain S. chilikensis KCTC 22540T. S5 was in an independent genetic branch and was close to the reference strain S. algae JCM 21037T. S1 was in an independent genetic branch, and S2 and S4 were in the same genetic branch, which were all close to the reference strain S. indica KCTC 23171T (Fig. 2B).

Analysis of virulence genes of Shewanella spp.

A total of 26 virulence-related genes in five categories were identified in the genomes of seven Shewanella strains, among which chemotaxis- and flagella-related genes were the most abundant (26.92%, 7/26), followed by secretion system- and serum resistance-related genes at 23.08% (6/26) and 15.38% (4/26), respectively. Among them, multiple virulence genes related to the type VI secretion system (T6SS) and the effector delivery system were predicted in S5, including hcp_1, hcp_2, vipB, and exeG. Virulence genes related to the effector delivery system and motility were predicted in the other six strains. A total of three virulence gene profiles were obtained from the seven strains. The virulence gene profile of S5 was in a separate cluster, which differed significantly from the composition of the virulence gene profiles of the remaining six strains, as it contained cheY, flgG, vipA/mglA, and katB genes. The virulence gene profiles of S7, S3, and S6 clustered together (Cluster 1), while those of S1, S2, and S4 clustered together (Cluster 2). The tviB, vasA, and vipB/mglB genes were detected only in Cluster 1, while xcpR was detected only in Cluster 2 (Fig. 3).

Prediction of resistance genes and phenotypic resistance

A total of six classes of resistance genes were identified in the seven Shewanella strains, including genes resistant to aminoglycosides [aph(3’’)-Ib, aph(6)-Id], β-lactams (blaOXA-55), quinolones (qnrA1, qnrA2), sulfonamides (sul2), chloramphenicol (floR), and tetracyclines [tet(59)]. All seven strains carried the gene resistant to β-lactam (blaOXA-55). Moreover, all strains carried the quinolone resistance gene, with four strains (S3, S5, S6, and S7) carrying qnrA1 and three strains (S1, S2, and S4) carrying qnrA2. In addition, one genetic branch (including strains S3, S6, and S7) also carried genes resistant to aminoglycosides [aph(3’’)-Ib, aph(6)-Id], sulfonamides (sul2), and tetracyclines [tet(59)]. However, S3 and S6 carried a chloramphenicol resistance gene (floR), whereas S7 did not, as shown in Fig. 4. Based on the phenotypic resistance test, three resistance profiles were detected for the seven Shewanella strains, among which S1, S2, and S4 were sensitive to all tested antibiotics, S3 and S6 were resistant to tetracycline (TET) + streptomycin (STR) + florfenicol (FLO), S7 were resistant to TET + STR, and S5 was resistant to ampicillin (AMP) + Colistin (CT), as shown in Fig. 4.

Discussion

In this study, Shewanella spp. were detected as a target microorganism together with other common diarrhea-causing pathogens during a 1-year diarrhea surveillance program. The detection rate of Shewanella spp. was 1.98% (7/354) among total diarrhea patients. In the samples that were positive for Shewanella spp., no other diarrhea-causing pathogens were detected, indicating that these were cases of single infection by Shewanella. Among them, six cases were reported on July 17–22, 2023, demonstrating a cluster of the time of disease onset. Based on the ANI analysis results, among the seven isolated Shewanella strains, there were three strains of S. chilikensis, three strains of S. indica, and one strain of S. algae(Fig. 1). They were all members of the algae clade3, all of which were new STs (Table 2). S. indica isolated from P2 and P4 were strains of the same clonal group, and S. chilikensis isolated from P3, P6, and P7 were strains from another group of clones (Fig. 2).

Epidemiological investigations showed that the suspicious food described by P3 and P6 was purchased from the same market in Beijing, with an interval of four days between purchases (Table 1). P4 and P5 did not know each other previously but had travelled to the same scenic spot on the same day (the scenic spot was 240 km away from Beijing) and ate seafood at the scenic spot (Table 1). Moreover, diarrhea was seen in seven and 20 of the companies who dined together with P4 and P5, respectively (Table 1). The presence of six cases unknown to each other that did not originate from the same foodborne disease outbreak, combined with the fact that the Shewanella spp. isolates were all members of the algae clade and that there were two identical clones suggested that the novel ST of the algae clade isolated in this study may be a potentially circulating strain with a certain level of pathogenicity. We hypothesize that even a single food contaminated with Shewanella may harbor a mixture of multiple Shewanella species (or polyclonal strains), exhibiting microbial diversity. When humans consume foods contaminated with Shewanella exhibiting such diversity, different cases may carry distinct dominant species (or clones) in their stools. These findings also suggest the emergence and dispersed outbreak of the algae clade in Beijing. More attention should be paid to diarrhea surveillance and regular surveillance should be enhanced to improve the early warning of diarrhea. This study enriched the genomic database of Shewanella spp. and provided basic information for epidemiological studies related to human diarrhea caused by Shewanella spp..

In this study, food suspected of causing infection described by six patients (P2-P7) included seafood, beverages, and cold watermelon (Table 1). Given the presence of an incubation period intervening between the exposure to contaminated food and the manifestation of illness, and considering the highly subjective nature of the patients accounts regarding the suspected contaminated food, the actual contaminated food responsible for inducing diarrhea might deviate from the information provided by the patients. In previous epidemiological studies on the algae clade, seafood should be a priority food category, and seafood related vending establishments should be identified as priority monitoring sites.

Among the six cases (P2-P7), the incubation period ranged from 8 to 12 h, and watery diarrhea was the most important clinical symptom (Table 1). In particular, P5, from whom S. algae was isolated, showed bloody purulent stool, with diarrhea episodes up to 50 times/day and the clinical symptom of rectal tenesmus, a common symptom in bacillary dysentery (Table 1). A previous study has also reported on the isolation of S. algae from bloody purulent stool, with the patient diagnosed with bacillary dysentery8. Based on the number of diarrhea episodes and faeces traits, P5 was found have more severe symptoms among the six patients mentioned above, which was in line with reports that S. algae may be correlated with high human pathogenicity5. The whole genome prediction of S5 (S. algae) showed 24 virulence genes in five categories, higher than the other six strains (Fig. 3).

Virulence genes related to motility were detected in the majority of Shewanella strains in this work. By tracking bacterial surface communities at single-cell resolution, Lee et al. also found that appendages such as pili and flagella of Shewanella had a significant impact on the growth rate diversity, and cell size homeostasis during its surface colonization process12. Biofilm formation plays a crucial role in the survival and colonization of Shewanella13. It can affect the intestinal colonization of Shewanella by regulating the methylation levels of genes related to lactate and iron homeostasis in the shrimp intestine14.

The chemokine-related virulence gene cheY, flagellum-related virulence gene flgG, effector delivery system-related virulence genes vipA/mglA, and enzyme-related virulence gene KatB were only predicted in S5 (Fig. 3). The CheY protein expressed by the chemokine-related virulence gene cheY is a key component of the two-component system (TCS), which senses environmental changes and regulates the corresponding cellular responses, critical for the pathogenicity of Helicobacter pylori15. When H. pylori is exposed to certain chemicals, the CheA protein is activated and transfers phosphate groups to CheY. Phosphorylated CheY then further interacts with other proteins to enable the bacteria to move towards or away from specific chemical stimuli, helping the bacteria to find a suitable site for colonisation. In Escherichia coli, Salmonella, and other Gram-negative bacteria, the CheY protein may assist the bacteria in sensing nutrient concentration gradients in their surroundings, as well as the presence of harmful compounds, thereby efficiently seeking favourable conditions for survival and avoiding unfavourable ones16,17. Therefore, we hypothesised that S5 had a higher capacity in environmental adaptation, which enhanced its pathogenicity compared to that of the other strains.

Iron metabolism factors are also a crucial class of virulence genes in Gram-negative bacteria. They strictly regulate intracellular iron homeostasis through processes such as iron uptake, efflux, and storage. Previous studies have shown that S. putrefaciens forms iron bodies through the fez operon, which may play a role in the adaptation to iron starvation under anaerobic conditions18. A recent study reported that in Shewanella, the synthesis of siderophores was regulated at multiple levels by multiple factors such as BarA/UvrY, SsoR, and Fur19. Different signals are integrated into the regulatory network of siderophore synthesis, providing new insights into the regulatory mechanism of bacterial siderophore synthesis.

The secretion system-related virulence genes vasA and vipB/mglB were only detected in strains in Cluster 1 (Fig. 3). The vasA gene encodes an ATPase that is a key component for the energy supply of T6SS and is essential for the assembly and activation of T6SS. T6SS is found in many Gram-negative bacteria, including Vibrio cholerae and Brucella spp., and can aid in the competition between bacteria as well as attack on host cells. The vipB/mglB gene encodes a regulatory protein that is associated with the type IV secretion system (T4SS). T4SS translocates DNA or proteins across both the inner and outer membranes and participates in the exchange of genetic material between bacteria or the injection of effector molecules into the host cell20,21,22. Therefore, it can be hypothesised that the strains in Cluster 1 had a more thorough capacity to invade the organism, interact with the host, and fight other microorganisms, leading to a higher pathogenic potential than the strains in Cluster 2.

The seven Shewanella strains in this study included three phenotypic resistance profiles and four genetic resistance profiles (Fig. 4). Previous studies have reported that β-lactam resistance genes are associated with specific species. For example, blaOXA-48-like may be associated with S. xiamenensis, blaOXA-729 with S. algae, and blaOXA-900 with S. putrefaciens23. Although all seven strains belonged to the algae clade and all carried blaOXA-55, only S5 (S. algae) was resistant to AMP (a β-lactamase inhibitor combination) and CT (Fig. 4). Kang et al.9 reported that S. algae isolated from sporadic cases of diarrhea in Beijing also had a high rate of resistance to AMP and CT. Several studies have reported that Shewanella spp. isolated from neonatal patients with sepsis and wounds of patients bitten by cobras were resistant to CT24,25. The quinolone resistance determinant qnrA was prevalent in Shewanella spp23. in our study four of the seven Shewanella strains (S3, S5, S6, and S7) carried qnrA1, while the remaining three strains (S1, S2, and S4) carried qnrA2(Fig. 4). However, all were sensitive to quinolone antibiotics (nalidixic acid and ciprofloxacin), suggesting that there are some differences in the resistance genes and phenotypic resistance against quinolone antibiotics among Shewanella spp. of the algae clade. The phylogenetic tree analysis clearly indicated that the three strains of S. chilikensis were of the identical clone (Fig. 2). Notably, both S3 and S6 harbored the resistance gene floR, whereas S7 conspicuously lacked this specific gene. And the phenotype of florfenicol resistance was consistent with the situation of the resistance gene (Fig. 4). Researchers have confirmed that the floR gene carried by Klebsiella pneumoniae and E. coli may be located on plasmids26,27. If the floR gene carried by the isolates of the algae clade in this study was also on the plasmid, the gene fragment on the plasmid may not be available based on whole-genome sequencing, resulting in the failure of detection of floR in S7. The floR gene mediates bacterial resistance to florfenicol and was identified on a plasmid in the fish pathogen Photobacterium damselae subsp. piscicida in 199628. All three strains of S. chilikensis were found to be resistant to TET and STR. Moreover, these strains were found to possess the resistance genes [tet(59)], [aph(3’’)-Ib], and [aph(6)-Id], suggesting a clear congruence between the observed phenotypic traits and the underlying genetic determinants (Fig. 4). Furthermore, unlike the S. algae strains isolated from Hainan, China29the isolates from Beijing were sensitive to both imipenem and colistin, and there was no relevant resistance genes were detected.

Our study was subject to certain limitations. The sample size was relatively small, only seven out of 354 diarrheal patients tested positive for Shewanella spp., which may not comprehensively reflect the true situation of Shewanella in diarrheal cases. Additionally, patients’ descriptions of suspected contaminated foods are influenced by subjective factors, and the presence of an incubation period can easily lead to misjudgments about the actual contaminated foods, thus limiting the accuracy of source tracing. Future studies should include more cases and optimize the investigation methods to explore the relationship between Shewanella and diarrhea. These strains of the algae clade were all identified as S. algae based on fully automated bacterial biochemical identification. By contrast, these strains were all identified as S. putrefaciens by MALDI-TOF MS, which differed from the results of the ANI analysis to a certain extent (Table 2). This highlights the persistent shortcomings of both the fully automated bacterial biochemical identification and MALDI-TOF MS in the identification of Shewanella strains. The above two conventional bacterial identification methods will need to be improved in order to enhance the accurate identification of various species of Shewanella.

Conclusions

In this study, Shewanella spp. were detected in patients with diarrhea at a certain level during one year of diarrhea surveillance. Moreover, the time of infection may have been clustered due to the presence of dispersed outbreaks. Seafood should be the key food category for monitoring and seafood markets should become a key monitoring site for Shewanella spp. The novel ST of the algae clade isolated from diarrhea patients in this study may be potentially helping for tracking circulating strains. Taken together, further in-depth investigations are required to precisely elucidate the correlation between Shewanella infections and human diarrhea, as well as the pathogenic characteristics of this infection.

Materials and methods

Sources of diarrhea surveillance samples and investigation of Shewanella spp. Positive cases

Sentinel hospitals were set up for the surveillance of diarrhea in Beijing, China. Cases were defined as patients with episodes of diarrhea ≥ 3 times/day and faecal characteristics of watery, loose, mucous, or bloody stools. Surveillance lasted from January 1, 2023 to December 31, 2023. Sentinel hospitals were responsible for collecting the patient information and stool samples. Stool samples were kept in sterile containers and forwarded to the regional centres for disease control (CDC) for pathogenetic testing within 24 h. In the case of a positive detection of Shewanella spp., a telephone survey was conducted to obtain further patient details.

Bacterial culture and viral real time polymerase chain reaction of stool samples

The stool samples of diarrhea cases were sent to the laboratory and subjected to isolation and culture of Shewanella spp., Salmonella spp., Shigella spp., diarrheagenic Escherichia coli, Vibrio parahaemolyticus, Campylobacter jejuni, Campylobacter coli, and Yersinia enterocolitica. The isolation and culture of bacteria in addition to Shewanella spp. were performed as previously described30. Meanwhile, real time polymerase chain reaction (PCR) was performed for all samples for norovirus, sapovirus, rotavirus, enteric adenovirus, and astrovirus.

The brief description of the isolation of Shewanella spp. is as follows: 200 mg of faeces was inoculated in 3% NaCl alkaline peptone water for enrichment culture at 37 °C for 24 h. A 10-µL inoculation ring was used to inoculate the enrichment solution in thiosulfate citrate bile salts sucrose (TCBS) agar and a Vibrio chromogenic plate, respectively, to culture at 37 °C for 24 h. The single black colony on TCBS agar or colourless and transparent in Vibrio chromogenic plates with dominant growth was transferred to a TSA plate to culture at 37 °C for 24 h. MALDI-TOF MS (Autoflex™ Speed, Bruker) and bacterial biochemical identification (VITEK 2 COMPACT, BioMerieux) were performed to identify the isolated colonies, and those identified as Shewanella spp. strains were stored9,31. A total of seven Shewanella strains were isolated.

Whole genome sequencing and bioinformatic analysis of Shewanella strains

Whole genome sequencing of the strains was entrusted to a third-party company. Bacterial DNA extraction was performed using the Promega Wizard™ Genomic DNA Purification Kit. The Illumina Hiseq 2000 platform was used as the sequencing platform, with a sequencing depth of 150×. Paired-end sequencing was performed and reads with a length of 150 bp were used for library construction. The quality control of raw sequencing reads was performed using FastQC. Low quality reads were removed, and genome frame assembly was performed using SOAP de novo(version 2.04, https://github.com/aquaskyline/SOAPdenovo2)32.

The ANI of strains was analysed using fastANI33. The six Shewanella model bacteria selected in this study were S. algae JCM21037T, S. carassii 08MAS2251T, S. chilikensis KCTC22540T, S. indica KCTC23171T, S. putrefaciens ATCC8071T, and S. xiamenensis JCM16212T. Two strains were determined to be the same species when the ANI value > 95%.

Genome annotation was performed using Prokka (version 1.12)34 Prodigal (version 2.6.3)35and RAST (https://rast.nmpdr.org/). Based on seven housekeeping genes (16SrRNA, gyrA, gyrB, infB, recN, rpoA, and topA), the allele numbers of the housekeeping genes were obtained from the PubMLST (https://pubmlst.org/organisms/shewanella-spp/) online analyses platform to determine the ST. SNP and insertion deletion (InDel) analyses were performed with the S. algae JCM 21,037T genome as the reference sequence using Snippy (version 4.3.6)36. Moreover, a phylogenetic tree was constructed for the strains using the maximum likelihood method based on cgSNP using iqtree (version 2.0.6)37.

The virulence factor database (VFDB, http://www.mgc.ac.cn/VFs/) and the PathogenFinder 1.1 database (https://cge.cbs.dtu.dk//services/PathogenFinder/) were used in combination to predict the potential virulence-related genes in the genome of Shewanella spp.. The amino acid sequences of the whole genome were searched against the database using BLASTp, with the cutoff values set to identity ≥ 60%, coverage ≥ 60%, and an E value of 1e-5. Potential resistance genes were predicted by BLASTp homology search using the ResFinder database and the comprehensive antibiotic resistance database (CARD), with parameters set to identity ≥ 90%, coverage ≥ 80%, and an E-value of 1e-5.

Antibiotic susceptibility testing of Shewanella strains

Antibiotic susceptibility testing was performed using the broth microdilution method to obtain the minimum inhibitory concentration (MIC). Susceptibility was determined as sensitive (S), intermediate (I), or resistant (R) according to the CLSI M100 Performance Standards for Antimicrobial Susceptibility Testing(florfenicol with reference to Campylobacter and others test antibiotics with reference to Vibrio). The test antibiotics and the cut-off points for determining resistance were: ampicillin (AMP, > 32 µg/mL), tetracycline (TET, > 16 µg/mL), colistin (CT, > 4 µg/mL), ciprofloxacin (CIP, > 1 µg/mL), azithromycin (AZI, > 32 µg/mL), chloramphenicol (CHL, > 32 µg/mL), nalidixic acid (NAL, > 32 µg/mL), streptomycin (STR, > 32 µg/mL), sulfamethoxazole tablets (SXT, > 4/76 µg/mL), amikacin (AMK, > 64 µg/mL) and florfenicol (FLO, > 8 µg/mL).

Fig. 1
figure 1

Heatmap of ANI analysis of the seven strains of Shewanella spp.. Six type strains (S. algae JCM 21037T, S. carassii 08MAS2251T, S. chilikensis KCTC 22540T, S. indica KCTC 23171T, S. putrefaciens ATCC 8071T, S. xiamenensis JCM 16212T ) were used as reference genomes. The color from blue to red represents the value of ANI from <80 to 100.

Full size image
Fig. 2
figure 2

SNP difference matrix of seven Shewanella strains (A) and the phylogenetic tree based on cgSNP (B). Three type strains (S. algae JCM 21037T, S. chilikensis KCTC 22540T, S. indica KCTC 23171T) were used as reference genomes. The SNP difference of the same strain is 0, highlighted in blue in matrix (A). The SNP difference of the same clonal group is 0 or 1, include same clonal group of S. indica(S2 and S4), and same clonal group of S. chilikensis(S3, S6 and S7), were highlighted in color orange in matrix (A).

Full size image
Fig. 3
figure 3

Distribution of virulence genes of seven Shewanella strains. Colors grey and blue represent the absence and presence of virulence genes, respectively.

Full size image
Fig. 4
figure 4

Heatmap of phenotypic resistance and resistance genes of seven Shewanella spp. strains. Color grey represents the absence of phenotypic resistance and resistance genes, and color orange represents the presence of phenotypic resistance and resistance genes. AMP = ampicillin, TET = tetracycline, CT = colistin, CIP = ciprofloxacin, AZI = azithromycin, CHL = chloramphenicol, NAL = nalidixic acid, STR = streptomycin, SXT = sulfamethoxazole tablets, AMK = amikacin, FLO = florfenicol.

Full size image
Table 1 Patient information, clinical symptoms, and distribution of suspected contaminated foods of six Shewanella spp.. Positive patients and a concentrated period of disease onset.
Full size table
Table 2 Identification and multilocus sequence typing analysis of seven Shewanella spp. strains.
Full size table