Clinical features
We collected 1,985 stool samples from people over 14 years old with acute diarrhoea at a general hospital in Beijing, China between April and October 2018. An epidemiological questionnaire was completed by the participants to assess clinical history and physical fitness. There were ultimately 349 diarrhoea cases related to bacteria, which included seven cases (2.0%) caused by seven different strains of V. furnissii. This was less than the numbers of cases caused by Aeromonas spp. (165, 47.3%), Vibrio parahaemolyticus (93, 26.6%), Salmonella spp. (30, 8.6%), Vibrio fluvialis (25, 7.2%), Vibrio cholerae (20, 5.7%) and Plesiomus shigelloides (8, 2.3%), but more than the single case caused by Shigella spp. (1, 0.3%). The use of agar base with sheep erythrocytes and AMP also helped isolate more strains of V. furnissii and other Vibrio species (Table 1).
The clinical and epidemiological characteristics of the seven patients with diarrhoea caused by V. furnissii strains are shown in Table 2. The sex ratio (male:female) was 2.5. Among these patients, one also had vomiting, two had abdominal pain, three had watery stool and none had fever. Erythrocytes and leukocytes were present under high magnification (× 40) in respectively one and two of the stool samples collected from these seven patients.
Table 1
Comparison of Vibrio numbers isolated from TCBS and AMP blood agar
Group | V. parahaemolyticus (n = 93) | V. fluvialis (n = 25) | V. cholerae (n = 20) | V. furnissii (n = 7) |
A | 25 (26.9%) | 2 (8.0%) | 6 (30.0%) | 0 (0.0%) |
B | 28 (30.1%) | 16 (64.0%) | 6 (30.0%) | 3 (42.9%) |
C | 40 (43.0%) | 7 (28.0%) | 8 (40.0%) | 4 (57.1%) |
A: Strains were isolated only from TCBS, B: Strains were isolated only from AMP blood agar, C: Strains could be isolated from both TCBS and AMP blood agar |
Table 2
Clinical characteristics of seven patients with diarrhoea caused by seven different strains of V. furnissii
Strain | Sexa | Age | Month | Feverb | Abdominal pain | Stool consistency | Vomiting | Erythrocytes No./Hpc | Leukocytes No. /Hpc |
VFBJ01 | M | 45 | May | - | - | watery | - | 0 | 0 |
VFBJ02 | F | 65 | June | - | - | loose | - | 0 | 0 |
VFBJ03 | M | 15 | July | - | - | watery | + | 0 | 1 |
VFBJ04 | M | 62 | July | - | - | loose | - | 0 | 0 |
VFBJ05 | F | 36 | July | - | + | loose | - | 3 | 30 |
VFBJ06 | M | 64 | July | - | - | watery | - | 0 | 0 |
VFBJ07 | M | 72 | August | - | + | loose | - | 0 | 0 |
a M, male; F, female. |
b Defined as axillary temperature > 37.7°C. |
c Erythrocyte or leukocyte number in stool identified under high-magnification field. |
Bioinformatic analysis
As shown in Table 3, the seven V. furnissii strains contained chromosomes of 4,802,519 bp to 5,109,544 bp in length with GC contents of 50.46–50.78% and 4,594 to 4,727 genes. They contained no plasmids, but had 8 to 15 GIs (Genomic islands) and 1 to 5 prophages in the genome at the time of sequencing.
The genome of V. furnissii contained several virulence factor-encoding genes and antibiotic resistance genes. As shown in Table 4, all seven genomes contained the following virulence operons: polar flagella and lateral flagella gene clusters, T2SS (type II secretion system), T3SS (type III secretion system), VFH (V. fluvialis haemolysin) [5], HupO (hemin-binding outer membrane protein) [6] and metalloprotease VFP (V. fluvialis protease )[7]. In addition, all strains other than VFBJ06 contained a T6SS (type VI secretion system). VFBJ07 had a series of antibiotic resistance genes, including those for aminoglycoside phosphotransferase (strA and strB), SUL resistance (sul1 and sul2), TET resistance (tetA and tetB), florfenicol/CHL resistance (floR) and quinolone resistance aac(6′)-IIa. VFBJ05 also had these antibiotic resistance genes except for sul1, tetB and aac(6′)-IIa, and VFBJ01 had the antibiotic resistance genes strA, strB and sul2. Furthermore, we found an insert sequence (IS) through BLAST in the corresponding database and obtained the transposon islands containing the antibiotic resistance genes in VFBJ05 and VFBJ07. As shown in Figs. 1 and 2, one transposon island in VFBJ05 contained strB, strA, tetA and sul2 and another contained floR, while one island in VFBJ07 contained strB, strA, tetA, sul1, sul2 and aac(6′)-Ⅱa and another contained tetB.
Table 3
Genome characteristics of seven V. furnissii strains
Sample ID | Genome size (bp) | Sequence GC% | No. of Genes | No. of GIs | No. of Prophages | No. of tRNA | No. of rRNA | No. of sRNA |
VFBJ01 | 5109544 | 50.46 | 4727 | 15 | 2 | 74 | 1 | 16 |
VFBJ02 | 4906640 | 50.75 | 4596 | 8 | 4 | 86 | 7 | 15 |
VFBJ03 | 5062567 | 50.78 | 4725 | 12 | 3 | 61 | 4 | 14 |
VFBJ04 | 4904255 | 50.58 | 4594 | 13 | 2 | 95 | 6 | 21 |
VFBJ05 | 4965951 | 50.56 | 4656 | 15 | 2 | 99 | 8 | 17 |
VFBJ06 | 4802519 | 50.74 | 4489 | 11 | 1 | 71 | 6 | 15 |
VFBJ07 | 5016530 | 50.56 | 4686 | 12 | 5 | 87 | 6 | 20 |
Table 4
Genomic presence/absence of main CDSs of toxins and antibiotic resistance in seven V. furnissii strains
Putative protein CDS a | VFBJ01 | VFBJ02 | VFBJ03 | VFBJ04 | VFBJ05 | VFBJ06 | VFBJ07 |
Appendages | | | | | | | |
Polar flagellum | + | + | + | + | + | + | + |
Lateral flagellum | + | + | + | + | + | + | + |
Toxins and exoenzymes | | | | | | |
T2SS | + | + | + | + | + | + | + |
T3SS | + | + | + | + | + | + | + |
T6SS | + | + | + | + | + | - | + |
VFP | + | + | + | + | + | + | + |
VFH | + | + | + | + | + | + | + |
HupO | + | + | + | + | + | + | + |
Antibiotic resistance | | | | | | |
StrA | + | - | - | - | + | - | + |
StrB | + | - | - | - | + | - | + |
Sul1 | - | - | - | - | - | - | + |
Sul2 | + | - | - | - | + | - | + |
tetA | - | - | - | - | + | - | + |
tetB | - | - | - | - | - | - | + |
floR | - | - | - | - | + | - | + |
aac(6′)-IIa | - | - | - | - | - | - | + |
a CDS: coding sequence |
Comparative genome analysis and phylogenetic tree
V. furnissii was first identified mass spectrometrically using VITEK MS and then identified biochemically using VITEK II (BioMerieuX, France). We found that the seven V. furnissii strains were misidentified as V. fluvialis at a rate of 85–99% by VITEK II, so we further identified these strains by whole-genome sequencing.
A phylogenetic tree of the seven isolates of V. furnissii was constructed based on SNPs and aligned with V. furnissii reference isolates CIP102972, NCTC13120 and NCTC11218 and V. fluvialis reference isolate ATCC33809. This revealed three major clades, with clades I, II and III consisting of one, two and eight isolates, respectively (Fig. 3). Clade I solely contained V. fluvialis ATCC33809, Clade II contained V. furnissii strains CIP102972 and NCTC13120, and Clade III contained all seven collected V. furnissii isolates as well as reference isolate NCTC11218.
Susceptibility to antibiotics
The resistance profiles of the seven isolates of V. furnissii to 19 antibiotic agents are shown in Table 5. High resistance to CFZ and TET was found in respectively 100.0% and 57.1% of the isolates, and intermediate resistance to AMS and IMI occurred at respective rates of 85.7% and 85.7%. Only three strains, VFBJ02, VFBJ05 and VFBJ07, exhibited multiple-drug resistance patterns against more than one of the 19 antibiotic agents: VFBJ02 was resistant to AMP, CFZ, IMI, MEM and TET, and VFBJ05 and VFBJ07 were resistant to CFZ, TET and STR.
Table 5
Antibiotic susceptibilitya patterns of all V. furnissii strains
Antibiotic1 | VFBJ01 | VFBJ02 | VFBJ03 | VFBJ04 | VFBJ05 | VFBJ06 | VFBJ07 |
AMP | I2 | R | I | I | I | R | I |
AMS | I | I | I | I | S | I | I |
CFZ | R | R | R | R | R | R | R |
CAZ | S | S | S | S | S | S | S |
FEP | S | I | S | S | S | S | S |
AZM | S | I | S | S | S | S | S |
AMI | S | S | S | S | S | S | S |
GEN | S | S | S | S | S | S | S |
CIP | S | S | S | S | S | S | S |
LEV | S | S | S | S | S | S | S |
IMI | I | R | I | I | I | I | I |
MEM | S | R | S | S | S | S | S |
TET | R | R | S | S | R | I | R |
DOX | I | S | S | S | S | S | I |
CHL | S | S | S | S | I | S | S |
SUL | S | S | S | S | S | S | S |
SXT | S | S | S | S | S | S | S |
AZI | S | S | S | S | S | S | S |
STR | R | S | I | I | R | S | R |
1 R: Resistant; I: Intermediate; S: Sensitive. |
2 Most breakpoints for defining susceptibility (µg/mL) are based on the CLSI (clinical and laboratory standards institute) M45-A3 standards for Vibrio spp., AMP: S ≤ 8, I = 16, R ≥ 32; AMS: S ≤ 8, I = 16, R ≥ 32; CFZ: S ≤ 2, I = 4, R ≥ 8; CAZ: S ≤ 4, I = 8, R ≥ 16; FEP: S ≤ 2, I = 4–8, R ≥ 16; AMI: S ≤ 16, I = 32, R ≥ 64; GEN: S ≤ 4, I = 8, R ≥ 16; CIP: S ≤ 1, I = 2, R ≥ 4; LEV: S ≤ 2, I = 4, R ≥ 8; IMI: S ≤ 1, I = 2, R ≥ 4; MEM: S ≤ 1, I = 2, R ≥ 4; TET: S ≤ 4, I = 8, R ≥ 16; CHL: S ≤ 8, I = 16, R ≥ 32; SUL: S ≤ 256, -, R ≥ 512; SXT (trimethoprim 2 µg/mL, sulfamethoxazole 38 µg/mL): S ≤ 2/38, R ≥ 4/76. Some breakpoints are based on the CLSIM100-ED31 criteria for Enterobacteriaceae, AZM: S ≤ 4, I = 8, R ≥ 16; DOX: S ≤ 4, I = 8, R ≥ 16. Breakpoint of AZI (S ≤ 2) is based on the CLSIM45-A3 standards for V. cholerae. Breakpoint of STR (S ≤ 4, I = 8, R ≥ 16) is based on the CLSI M45-A3 standards for Y. pestis. |