Analysis of global Aeromonas veronii genomes provides novel information on source of infection and virulence in human gastrointestinal diseases.

BACKGROUND: Aeromonas veronii is a Gram-negative rod-shaped motile bacterium that inhabits mainly freshwater environments. A. veronii is a pathogen of aquatic animals, causing diseases in fish. A. veronii is also an emerging human enteric pathogen, causing mainly gastroenteritis with various severities and also often being detected in patients with inflammatory bowel disease. Currently, limited information is available on the genomic information of A. veronii strains that cause human gastrointestinal diseases. Here we sequenced, assembled and analysed 25 genomes (one complete genome and 24 draft genomes) of A. veronii strains isolated from patients with gastrointestinal diseases using combine sequencing technologies from Illumina and Oxford Nanopore. We also conducted comparative analysis of genomes of 168 global A. veronii strains isolated from different sources.
RESULTS: We found that most of the A. veronii strains isolated from patients with gastrointestinal diseases were closely related to each other, and the remaining were closely related to strains from other sources. Nearly 300 putative virulence factors were identified. Aerolysin, microbial collagenase and multiple hemolysins were present in all strains isolated from patients with gastrointestinal diseases. Type III Secretory System (T3SS) in A. veronii was in AVI-1 genomic island identified in this study, most likely acquired via horizontal transfer from other Aeromonas species. T3SS was significantly less present in A. veronii strains isolated from patients with gastrointestinal diseases as compared to strains isolated from fish and domestic animals.
CONCLUSIONS: This study provides novel information on source of infection and virulence of A. veronii in human gastrointestinal diseases.

Introduction

Aeromonas veronii is a Gram-negative rod-shaped motile bacterium that inhabits mainly freshwater environments such as ground water, lakes and river [1]. It has also been isolated from chlorinated and untreated drinking water [2-6]. Several Aeromonas species including A. veronii are pathogens of aquatic animals, causing diseases such as skin ulceration and systemic hemorrhagic septicemia in fish, which is a great concern in aquaculture globally [7-9].

A. veronii and several other Aeromonas species also cause human diseases. The most common diseases caused by Aeromonas species in humans are gastroenteritis, soft-tissue infections and bacteremia [1]. Aeromonas species associated human gastroenteritis are mainly caused by three Aeromonas species including A. veronii, Aeromonas caviae and Aeromonas hydrophila, with A. veronii being the most commonly isolated species [10]. Aeromonas species caused human gastrointestinal infections are positively associated with increasing age [10]. Aeromonas species caused gastroenteritis may present with acute or chronic courses [11-15] While most patients can recover without medical treatment, those with severe symptoms and chronic infections often require hospital admission and antibiotic therapy [14]. In addition to gastroenteritis, Aeromonas species were often detected in patients with inflammatory bowel disease [16].

Several studies have examined the genomes of A. veronii strains isolated from dairy cattle, fish, and environmental samples [17, 18]. However, limited genomic data from A. veronii strains isolated from patients with gastrointestinal diseases are available. In order to better understand the pathogenicity of A. veronii in human diseases, there is a need to examine the genomes of A. veronii strains isolated from patients with gastrointestinal diseases.

In this study, we sequenced, assembled and analysed 25 genomes of A. veronii strains isolated from fecal samples of patients with gastrointestinal diseases, including one complete and 24 draft genomes. These 25 A. veronii strains were identified in our previous study based on the sequences of seven housekeeping genes including gyrB, rpoD, gyrA, recA, dnaJ, dnaX and atpD [10]. Comparative genome analysis of 168 A. veronii strains isolated from different sources in 18 countries were also conducted.

Results

The complete and draft genomes of 25 A. veronii strains isolated from fecal samples of patients with gastroenteritis

We successfully obtained the complete genome of A. veronii strain A29V through hybrid assembly of the data obtained from Illumina MiSeq sequencing and Oxford Nanopore sequencing. The complete genome of A. veronii strain A29V had a size of 4.54 Mb, with a GC content of 58.8%. Two plasmids, designated as pAV1K and pAV7K, were identified in strain A29V, consisting of 1740 and 7073 bp respectively, with each encoding one and five proteins respectively. The pAV1K was found in another four A. veronii strains (pamvotica, NK02, CNRT12, and NK07), as well as other Aeromonas species including Aeromonas popoffii (strain CIP 105,493), Aeromonas sobria (strains 2014–10,509-27–20 and PAQ091014-19), and Aeromonas allosaccharophila (strain Z9-6), while pAV7K was only found in one additional A. veronii strain UDRT09. No potential virulence factors were identified in these two plasmids.

The detailed information of the 25 A. veronii genomes sequenced in this study are shown in Table 1.

Table 1

Summary of the 25 Aeromonas veronii genomes sequenced and assembled in this study

Strain name	Source	Level	Size (Mb)	GC%	N50b (bp)	No. of contigs	Plasmid	Fold coverage	Accession number of genome assembly	Accession number of raw data
A20	Human feces, gastroenteritis	Draft	4.57	58.7	199,035	42		112	JAIEXX000000000	SRR15464912
A20-10	Human feces, gastroenteritis	Draft	4.51	58.6	268,460	36		198	JAIEYE000000000	SRR15464890
A20-12	Human feces, IBD	Draft	4.46	58.9	159,675	56		291	JAIEYF000000000	SRR15464889
A20-14	Human feces, gastroenteritis	Draft	4.59	58.5	106,766	76		186	JAIEYG000000000	SRR15464888
A20-17	Human feces, gastroenteritis	Draft	4.46	58.9	116,957	70		117	JAIEYH000000000	SRR15464910
A20-5	Human feces, gastroenteritis	Draft	4.5	58.8	186,941	40		159	JAIEYI000000000	SRR15464909
A20-8	Human feces, gastroenteritis	Draft	4.4	59.0	255,850	40		208	JAIEYJ000000000	SRR15464908
A21	Human feces, gastroenteritis	Draft	4.49	58.8	118,245	80		120	JAIEXY000000000	SRR15464911
A21-10	Human feces, gastroenteritis	Draft	4.42	58.6	159,089	57		175	JAIEYK000000000	SRR15464907
A21-11	Human feces, gastroenteritis	Draft	4.44	58.9	165,092	54		143	JAIEYL000000000	SRR15464906
A21-13	Human feces, gastroenteritis	Draft	4.37	58.9	119,909	55		258	JAIEYM000000000	SRR15464905
A21-14	Human feces, gastroenteritis	Draft	4.65	58.4	104,270	82		129	JAIEYN000000000	SRR15464904
A21-15	Human feces, gastroenteritis	Draft	4.51	58.8	110,908	68		251	JAIEYO000000000	SRR15464903
A21-16	Human feces, gastroenteritis	Draft	4.37	58.8	260,229	37		325	JAIEYP000000000	SRR15464902
A21-19	Human feces, gastroenteritis	Draft	4.66	58.6	189,060	37		233	JAIEYQ000000000	SRR15464901
A21-4	Human feces, gastroenteritis	Draft	4.61	58.7	115,259	76		207	JAIEYR000000000	SRR15464899
A21-5	Human feces, gastroenteritis	Draft	4.64	58.8	255,727	48		262	JAIEYS000000000	SRR15464898
A21-6	Human feces, gastroenteritis	Draft	4.49	58.5	224,448	43		310	JAIEYT000000000	SRR15464897
A21-8	Human feces, gastroenteritis	Draft	4.52	58.8	169,064	50		136	JAIEYU000000000	SRR15464896
A26	Human feces, gastroenteritis	Draft	4.52	58.8	252,545	34		106	JAIEXZ000000000	SRR15464900
A27	Human feces, gastroenteritis	Draft	4.47	58.7	159,194	57		174	JAIEYA000000000	SRR15464894
A29V	Human feces, gastroenteritis	Complete	4.54	58.8	N/A	3	2	111	CP080630-CP080632	SRR15464895
A7	Human feces, gastroenteritis	Draft	4.52	58.9	118,736	80		130	JAIEYB000000000	SRR15464893
A8	Human feces, gastroenteritis	Draft	4.42	58.8	169,095	60		143	JAIEYC000000000	SRR15464892
A9	Human feces, gastroenteritis	Draft	4.54	58.8	166,550	51		146	JAIEYD000000000	SRR15464891

The 25 A. veronii strains listed in Table 1 were isolated in Australia. IBD inflammatory bowel disease, N/A not applicable

Phylogenetic analysis of global A. veronii genomes

A total of 168 A. veronii genomes were used for analysis in this study, including 25 A. veronii genomes sequenced in this study and 143 A. veronii genomes obtained from public databases (Table 2). The A. veronii genomes in the public databases were obtained from National Center for Biotechnology Information (NCBI) genome database and their genome details and isolation sources were recorded. The core genome of the 168 A. veronii strains contained 1315 genes. Based on the maximum likelihood phylogenetic tree constructed from the core genome of the 168 A. veronii strains, three distinctive phylogenetic clusters were observed (Fig. 1). Cluster 1 contained 149 A. veronii strains (bootstrap value 99), which were from 18 countries. Cluster 2 (bootstrap value 100) had 11 A. veronii strains, which were from five countries including Australia (four strains), China (four strains), Israel (one strain), India (one strain) and USA (one strain). Cluster 3 (bootstrap value 100) contained the remaining eight A. veronii strains, which were from seven countries including Australia (two strains), Turkey (one strain), South Africa (one strain), India (one strain), Germany (one strain), Spain (one strain) and China (one strain). All three clusters contained strains from different sources, including humans, animals and environmental samples (Fig. 1).

Table 2

The 143 Aeromonas veronii strains in the public databases that were used in this study

Strain names	Country	Source	Level	Size (Mb)	GC%	N50 (bp)	No. of contigs	Plasmid	Ref
BC88	Australia	Human feces, dysentery	Draft	4.60	58.5	215,763	155
FC951	India	Human feces, diarrhea	Complete	4.86	58.7	N/A	2	1
126–14	China	Human feces, diarrhea	Draft	4.37	58.6	72,935	146
312 M	Brazil	Human feces, gastroenteritides	Draft	4.57	58.6	502,756	14
VBF557	India	Human feces, gastroenteritides	Draft	4.70	58.4	19,666	526
ERR1305902-bin.15	Denmark	Human feces, diarrhea	Draft	4.11	59.4	32,267	226
CN17A0013	China	Human feces	Draft	4.45	58.9	167,760	49
CN17A0029	China	Human feces	Draft	4.60	58.8	2,737,631	19
CN17A0031	China	Human feces	Draft	4.42	58.9	156,076	45
CN17A0036	China	Human feces	Draft	4.45	58.9	230,866	38
CN17A0040	China	Human feces	Draft	4.44	58.9	302,838	32
CN17A0049	China	Human feces	Draft	4.30	58.9	302,677	31
CN17A0054	China	Human feces	Draft	4.33	58.9	217,304	57
CN17A0059	China	Human feces	Draft	4.26	58.9	164,457	48
CN17A0067	China	Human feces	Draft	4.55	58.7	180,288	60
CN17A0087	China	Human feces	Draft	4.58	58.6	145,832	80
CN17A0093	China	Human feces	Draft	4.35	58.6	120,695	64
CN17A0097	China	Human feces	Draft	4.52	58.6	141,638	88
CN17A0102	China	Human feces	Draft	4.47	58.7	126,005	75
CN17A0103	China	Human feces	Draft	4.43	58.8	110,083	102
CN17A0114	China	Human feces	Draft	4.43	58.9	237,189	35
CN17A0120	China	Human feces	Draft	4.52	58.6	82,803	128
CN17A0122	China	Human feces	Draft	4.48	58.7	196,730	33
CN17A0154	China	Human feces	Draft	4.44	59.0	260,206	33
ADV102	France	Human feces	Draft	4.52	58.6	108,450	87		[19]
AMC34	USA	Human intestinal tract	Draft	4.58	58.4	219,183	1
MGYG-HGUT-02529	China	Human gut	Draft	4.70	58.4	119,499	124
ZJ12-3	China	Human rectal swab	Draft	4.70	58.4	119,499	124
AVNIH1 (GCA_001634325)	USA	Human perirectal swab	Complete	4.96	58.5	N/A	2	1	[20]
AVNIH2	USA	Human perirectal swab	Draft	4.52	58.9	211,774	50		[20]
1708–29,120	China	Human cholangiolithiasis bile	Complete	4.50	58.9	N/A	1
C198	Thailand	Human blood, septicaemia	Draft	4.58	58.6	4,550,752	3
FDAARGOS_632	USA	Human	Complete	4.56	58.9	N/A	2	1
CECT 4257	USA	Human sputum	Draft	4.52	58.9	148,348	52		[21]
AER39	USA	Human blood	Draft	4.42	58.8	188,051	4		[21]
AER397	USA	Human blood	Draft	4.50	58.8	645,709	5		[21]
BVH37	France	Human blood	Draft	4.46	58.8	115,181	55		[19]
BVH46	France	Human blood	Draft	4.51	58.8	215,038	39		[22]
BVH47	France	Human blood	Draft	4.64	58.9	96,732	108		[19]
AK247	France	Human forehead abscess	Draft	4.55	58.8	260,691	36		[19]
AMC35	USA	Human wound	Draft	4.57	58.5	351,392	2		[21]
CCM 4359	USA	Human sputum, drowning	Draft	4.51	58.9	245,067	56
TTU2014-108AME	USA	Dairy cattle feces	Draft	4.53	58.7	162,342	62		[17]
TTU2014-108ASC	USA	Dairy cattle feces	Draft	4.53	58.7	187,473	58		[17]
TTU2014-113AME	USA	Dairy cattle feces	Draft	4.66	58.6	74,547	122		[17]
TTU2014-115AME	USA	Dairy cattle feces	Draft	4.53	58.7	205,013	53		[17]
TTU2014-115ASC	USA	Dairy cattle feces	Draft	4.53	58.7	233,487	52		[17]
TTU2014-125ASC	USA	Dairy cattle feces	Draft	4.68	58.6	168,256	58		[17]
TTU2014-130AME	USA	Dairy cattle feces	Draft	4.68	58.6	189,668	64		[17]
TTU2014-130ASC	USA	Dairy cattle feces	Draft	4.68	58.6	247,513	49		[17]
TTU2014-131ASC	USA	Dairy cattle feces	Draft	4.68	58.6	187,444	70		[17]
TTU2014-134AME	USA	Dairy cattle feces	Draft	4.68	58.6	204,478	50		[17]
TTU2014-134ASC	USA	Dairy cattle feces	Draft	4.68	58.6	193,661	59		[17]
TTU2014-140ASC	USA	Dairy cattle feces	Draft	4.68	58.6	148,012	81		[17]
TTU2014-141AME	USA	Dairy cattle feces	Draft	4.68	58.6	223,907	48		[17]
TTU2014-141ASC	USA	Dairy cattle feces	Draft	4.68	58.6	241,272	45		[17]
TTU2014-142ASC	USA	Dairy cattle feces	Draft	4.68	58.6	247,560	45		[17]
TTU2014-143AME	USA	Dairy cattle feces	Draft	4.68	58.6	204,478	59		[17]
TTU2014-143ASC	USA	Dairy cattle feces	Draft	4.68	58.6	202,296	54		[17]
A31	South Africa	Pig rectal swab	Draft	4.64	58.5	114,767	84
A5	South Africa	Pig rectal swab	Draft	4.77	58.2	230,041	33
A86	South Africa	Pig rectal swab	Draft	4.64	58.5	206,004	42
A34	South Africa	Pig rectal swab	Draft	4.64	58.5	139,859	82
A136	South Africa	Pig rectal swab	Draft	4.67	58.4	213,730	41
Ae52	Sri Lanka	Carassius auratus	Draft	4.56	58.7	158,595	80
CL8155	China	Carp gut, healthy	Draft	4.68	58.6	284,020	50
JC529	China	Carp sepsis	Complete	4.83	58.3	N/A	1
MS 17–88	USA	Catfish	Draft	5.18	58.2	1,334,815	13
MS-18–37	USA	Catfish	Complete	4.68	58.6	N/A	1
ML09-123	USA	Catfish	Draft	4.75	58.4	299,782	32
VCK_1	Greece	Dicentrarchus labrax kidney, diseased	Draft	4.63	58.6	68,239	120
PDB	Greece	Dicentrarchus labrax kidney, diseased	Draft	4.72	58.5	72,590	141
AG_5.28.6	Greece	Dicentrarchus labrax kidney, diseased	Draft	4.61	58.6	85,872	98
NS	Greece	Dicentrarchus labrax kidney, diseased	Draft	4.71	58.5	67,042	140
NS2	Greece	Dicentrarchus labrax kidney, diseased	Draft	4.72	58.5	69,902	143
NS_6.15.2	Greece	Dicentrarchus labrax kidney, diseased	Draft	4.72	58.5	66,300	149
NS22	Greece	Dicentrarchus labrax kidney, diseased	Draft	4.74	58.4	61,224	172
NS13	Greece	Dicentrarchus labrax kidney, diseased	Draft	4.67	58.6	72,418	139
BIOO50A	Turkey	Dicentrarchus labrax kidney, diseased	Draft	4.61	58.6	73,700	109
17ISAe	South Korea	Discus spleen	Complete	4.66	58.5	N/A	2	1
A8-AHP	India	Labeo rohita, diseased	Complete	4.77	58.4	N/A	4	3
UBA1835	Spain	Anguilla anguilla epidermal mucus	Draft	4.11	59.0	17,609	323
ZfB1	China	Fish	Complete	4.71	58.5	N/A	1
PhIn2	India	Fish intestinal	Draft	4.30	58.8	3789	1899		[21]
CB51	China	Grass carp	Complete	4.58	58.6	N/A	1		[17]
XH.VA.1	China	Ictalurus punctatus	Draft	5.36	56.5	259,638	62
XH.VA.2	China	Ictalurus punctatus	Draft	4.91	58.1	259,509	48
X11	China	Megalobrama amblycephala	Complete	4.28	58.8	N/A	1
X12	China	Megalobrama amblycephala	Complete	4.77	58.3	N/A	1
Aer_WatCTCBM21	Brazil	Oreochromis niloticus	Draft	4.60	58.7	317,324	45
CNRT12	Thailand	Oreochromis sp.	Draft	4.90	58.1	265,081	479
NK01	Thailand	Oreochromis sp.	Draft	4.56	58.5	171,547	95
NK02	Thailand	Oreochromis sp.	Draft	4.80	58.2	110,255	400
NK07	Thailand	Oreochromis sp.	Draft	4.78	58.6	214,996	46
UDRT09	Thailand	Oreochromis sp.	Draft	4.61	58.5	169,295	186
BAQ071013-135	USA	Perch head kidney	Draft	4.62	58.9	167,400	50
B44	Brazil	Pseudoplatystoma corruscans kidney	Draft	4.61	58.6	290,712	51
B48	Brazil	Pseudoplatystoma corruscans kidney	Draft	4.73	58.7	284,404	49
WB12	China	Carassius auratus intestine, sick	Draft	4.52	58.8	282,522	40
AVNIH1 (GCA_009834065)	South Korea	Silurus asotus	Complete	4.81	58.5	N/A	1
TH0426	China	Tachysurus fulvidraco	Complete	4.92	58.3	N/A	1
XU1	Greece	Xiphophorus helleri kidney	Draft	4.80	58.0	206,195	92
XhG1.2	India	Xiphophorus hellerii	Draft	4.57	58.7	305,294	34
HX3	China	Alligator	Complete	4.76	58.5	N/A	2	1
CQ-AV1	China	Andrias davidianus liver	Draft	4.78	58.5	204,972	36
161	China	Channa argus	Draft	4.51	58.7	312,206	28
LMG 13,067	USA	Frog	Draft	4.74	58.4	91,946	72		[21]
S00030	USA	Heterelmis comalensis	Draft	4.51	58.7	237,167	21
Hm21	Turkey	Hirudo verbena digestive tract	Complete	4.77	58.7	N/A	2	1	[21]
CMF	India	Insect gut	Draft	4.56	58.7	40,276	200
CIP 107,763	India	Mosquito gut	Draft	4.43	58.8	188,049	64		[21]
AK241	France	Snail	Draft	4.60	58.6	215,278	42		[19]
B565	China	Aquaculture pond sediment	Complete	4.55	58.7	N/A	1		[18]
22	Brazil	Combined sewer	Draft	5.09	58.3	66,851	185
28	Brazil	Combined sewer	Draft	4.97	58.5	94,322	108
CECT 7059	Spain	Drinking water	Draft	4.81	58.4	188,889	31
RU31B	USA	Duckweeds	Draft	4.53	58.7	73,776	93
CECT 4902	Germany	Environment	Draft	4.64	58.4	347,677	29		[19]
AK236	France	Lake water	Draft	4.41	58.8	412,126	26
Colony604	Thailand	Food	Draft	4.57	57.8	7493	1
Colony111	Thailand	Food	Draft	4.58	57.8	7510	1
Colony512	Thailand	Food	Draft	4.60	58.5	16,905	1
Colony125	Thailand	Food	Draft	4.58	58.0	7702	1
pamvotica	Greece	Lake Pamvotis surface sentiment	Draft	4.92	58.1	739,151	21
A134	Israel	Lake Kinneret microcystis bloom	Draft	4.41	58.7	50,812	151
S50-1	USA	Organic kale	Draft	4.56	58.5	104,479	130
CTe-01	Peru	Oxidation pond	Draft	4.68	58.6	111,555	200
ARB3	Japan	Pond water	Draft	4.54	58.8	205,115	63		[21]
Z2-7	China	Pork	Draft	4.41	58.7	265,145	48
KLG7	UK	River Don	Draft	4.55	58.8	139,212	104
KLG5	UK	River Don	Draft	4.74	58.5	280,270	103
KLG8	UK	River Don	Draft	4.59	58.6	198,583	76
KLG9	UK	River Don	Draft	4.61	58.7	180,084	74
CECT 4486	Germany	Surface water	Draft	4.41	58.9	90,706	66		[21]
CCM 7244	Germany	Surface water	Draft	4.42	58.9	185,495	74		[17]
A29	South Africa	Surface water	Draft	4.48	58.8	165,894	54
AK227	France	Wastewater treatment plant	Draft	4.40	58.7	105,208	67
WP2-S18-CRE-03	Japan	Wastewater treatment plant	Complete	4.94	58.6	N/A	4	3
WP3-W19-ESBL-03	Japan	Wastewater treatment plant	Complete	4.98	58.7	N/A	6	4
WP8-S18-ESBL-11	Japan	Wastewater treatment plant	Complete	4.91	58.7	N/A	4	3
WP8-W19-CRE-03	Japan	Wastewater treatment plant	Complete	4.79	58.5	N/A	6	4
WP9-W18-ESBL-04	Japan	Wastewater treatment plant	Complete	4.93	58.7	N/A	5	4
D	South Africa	Water	Draft	4.43	59.0	54,053	149

There are two different strains of which both named AVNIH1, the corresponding accession numbers are indicated in brackets. N/A not applicable

Fig. 1

Phylogenetic tree generated based on Aeromonas veronii core genome. The phylogenetic tree was generated based on the core genome of 168 A. veronii strains isolated from different sources globally using maximum likelihood method by FastTree. The 168 A. veronii strains formed three clusters. Cluster 1 (shaded light grey colour, bootstrap value 99) contained 149 A. veronii strains, Cluster 2 (shaded yellow colour, bootstrap value 100) contained 11 strains and Cluster 3 (shaded pink colour, bootstrap value 100) contained eight strains. Within Cluster 1, strains isolated from the same environmental or animal sources often formed small groups. The genomes of A. veronii strains with blue colour were sequenced in this study

Within Cluster 1, A. veronii strains isolated from environmental samples or domestic animals from the same geographic locations often formed small groups (Fig. 1). For example, 13 of the 17 A. veronii strains isolated from dairy cattle were in the same group (bootstrap value 100). The five strains isolated from pig rectal swabs from South Africa (A31, A5, A86, A34 and A136) were in the same group (bootstrap value 100). Similarly, the nine strains (PDB, NS2, NS6.15.2, NS22, NS13, NS, VCK1, AG5.28.6 and BIOO50A) isolated from Dicentrarchus labrax fish from Greece also formed their own group (bootstrap value 100) (Fig. 1).

The average nucleotide identity (ANI) values of each A. veronii strain against the other 167 A. veronii strains were mostly over 95%. An exception was strain WP2-S18-CRE-03, which was isolated from a wastewater treatment plant in Japan. This strain had ANI values 91- 92% against the other 167 A. veroniis strains.

Strains closely related to A. veronii strains isolated from fecal samples of patients with gastrointestinal diseases

Strains that are closed related to the 31 A. veronii strains isolated from patients with gastrointestinal diseases were identified based on the highest ANI values. Twenty-two (71%, 22/31) closely related A. veronii strains were from fecal samples of other human individuals, 19 of these 22 individuals had recorded gastrointestinal diseases. Nine closely related A. veronii strains (29%, 9/31) were from freshwater fish or domestic animals (cattle and pig) (Table 3). Of the 26 A. veronii strains isolated from patients with gastrointestinal diseases in Australia, 16 strains (61.5%, 15/26) had closely related strains from patients in Australia, four strains (15.4%, 4/26) had closely related strains isolated from intestinal tract of individuals in other countries (one patient had gastroenteritis and the clinical conditions of the remaining three individuals were not known), the remaining six A. veronii strains (23%) had closely related strains from various sources including freshwater fish, domestic animals, leech and surface water (Table 3).

Table 3

Strains that are most closely related to the 31 Aeromonas veronii strains isolated from patients with gastrointestinal diseases

A. veronii strains isolated from fecal samples of patients with gastrointestinal diseases	Most closely related strain (ANI value)
aA20	Hm21(97.09)
aA20-10	aA7 (96.769)
aA20-12	aA21-8 (96.74)
aA21-14	cCECT 4257 (96.7)
aA20-17	aA21-13 (96.55)
aA20-5	aA7 (96.79)
aA20-8	cCN17A0029 (96.84)
aA21	aA21-5 (96.58)
aA21-10	A29 (96.60)
aA21-11	aA21-5 (96.7)
aA21-13	aA21-15 (96.66)
aA21-14	cCECT 4257 (96.7)
aA21-15	aA21-13 (96.58)
aA21-16	A8-AHP (97.5)
aA21-19	A136 (96.69)
aA21-4	aA9 (99.30)
aA21-5	aA21-11 (96.7)
aA21-6	cAMC34 (97.85)
aA21-8	TTU2014-130AME (98.1)
aA26	aA21-11 (96.64)
aA27	aA7 (96.72)
aA29V	XH.VA.2 (99.48)
aA7	aA20-5 (96.81)
aA7	aA20-5 (96.81)
aA9	aA21-4 (99.25)
aBC88	aA20-10 (96.56)
bFC951	XU1 (96.55)
b121-14	XU1 (96.71)
b312M	161 (97.16)
bVBF557	aA21-8 (96.41)
bERR1305902-bin.15	aA8 (96.84)

aA. veronii strains isolated from feces of patients with gastrointestinal diseases in Australia

bstrains isolated from diarrheal feces of patients from other countries

cstrains isolated from feces of human individuals without clear clinical information

Secretion systems

Secretion systems in the genomes of 168 A. veronii strains were examined. Five types of secretion systems, including Type I Secretion System (T1SS), T2SS, T3SS, T4SS and T6SS were identified in A. veronii (Additional file 1).

T1SS system was found in all A. veronii strains except strain ERR1305902-bin.15. T2SS secretion system was found in all 168 A. veronii strains.

T3SS was found in 106 of the 168 A. veronii strains (63.1%). A. veronii strains isolated from freshwater fish, environmental samples, domestic animals (cattle and pigs) and other animals had T3SS positivity of 84% (32/38), 60% (15/25), 100% (22/22) and 70% (7/10) respectively. The ‘other animals’ group included A. veronii strains isolated from mosquito gut, insect gut, hirudo verbena digestive tract, grass carp, Heterelmis comalensis, Xiphophorus helleri, frog, snail, Andrias advidianus and alligator.

The T3SS positivity in A. veronii strains isolated from patients with gastrointestinal diseases, bacteremia and other human samples was 48% (15/32), 83% (5/6) and 30% (9/30) respectively. The ‘other human sample’ group included A. veronii strains isolated from sputum, wound infection, bile of gallstone and fecal samples of individuals without clinical information. The T3SS positivity in A. veronii strains isolated from patients with gastrointestinal diseases was significantly lower than that in A. veronii strains isolated from freshwater fish (p = 0.002) and domestic animals (p < 0.0001). The other statistical analysis data are shown in Fig. 2A. The negativity of T3SS was confirmed by searching the franking genes in the T3SS negative strains.

Fig. 2

The Aeromonas veronii genomic island AVI-1 containing the T3SS system. The AVI-1 genomic island identified in this study contains genes encoding T3SS, which is found in 106 of the 168 strains examined in this study. A The prevalence of T3SS in strains isolated from patients with gastrointestinal diseases was significantly lower than that in A. veronii strains isolated from freshwater fish (p = 0.0125) and domestic animals (p < 0.0001). B Comparison of the A. veronii genomes with T3SS (representative strain A29V) and without T3SS (representative strain FC951) shows that the AVI-1 genomic island is located adjacent to a gene encoding crossover junction endodeoxyribonuclease (red). The identical proteins in these two strains are shaded in grey. C Genes in the AVI-1 genomic island that encodes T3SS components. *Indicates statistical significance (*p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001). Other human samples include A. veronii strains isolated from sputum, wound infection, cholangiolithiasis bile and fecal samples of individuals without clinical information. Other animals include A. veronii strains isolated from mosquito gut, insect gut, Hirudo verbena digestive tract, grass carp, Heterelmis comalensis, Xiphophorus helleri, frog, snail, Andrias advidianus and alligator. The food group included strains isolated from various food

A number of T4SS components were found in several A. veronii strains, mainly strains isolated from dairy cattle in USA. T6SS was found in 55 of the 168 A. veronii strains examined (32.7%) and it did not show a statistical significance in strains isolated from different sources (Additional file 1).

T3SS in A. veronii is located in a genomic island that is highly similar to plasmids in Aeromonas salmonicida

Comparison of the genomes of 23 complete A. veronii genomes (11 T3SS positive and 12 T3SS negative) revealed that T3SS in A. veronii is located on a genomic island, which we named A. veronii genomic island-1 (AVI-1) (Fig. 2B). AVI-1 genomic island has a size of 26,064 bp and GC content of 60%. The AVI-1 island is adjunct to a gene encoding crossover junction endodeoxyribonuclease, an enzyme involving in homologous recombination. The components of A. veronii T3SS were shown in Fig. 2C.

Blast search against all bacterial genomes in public databases showed that the AVI-1 genomic island was also found in some A. hydrophilia and Aeromonas salmonicida strains. For example, the AVI-1 island is in the chromosome of A. hydrophila strains 23-C-23 and WCX23 (97% query coverage and 95.57% identity). In A. salmonicida, the AVI-1 island is in plasmids, for example plasmid pS44-3 in strain S44 and plasmid pS121-3 in strain S121 (97% query coverage and 94.85% identity).

Virulence factors

Two hundred and ninety-nine putative virulence factors were identified in the complete genome of A. veronii strain A29V, including molecules involved in adherence, colonization, invasion, secretion systems, mobility, immune evasion, antiphagocytosis and others (Fig. 3).

Fig. 3

Putative virulence factors in Aeromonas veronii strain A29V. Putative virulence factors in the complete genome of A. veronii strain A29V, a strain isolated from fecal sample of a patient with gastroenteritis, was identified through searches of the Virulence Factors Database. A total of 299 putative virulence was identified. A Percentages of virulence factors in different categories. B) Virulence genes in each virulence factor category

Toxins produced by the 31 A. veronii strains isolated from patients with gastrointestinal diseases were further examined. Two secreted toxins, aerolysin and microbial collagenase, were found in all 31 strains (Fig. 4). The aerolysin proteins in different A. veronii strains were highly similar, with the overall protein sequence identity being 75% among the 31 strains (Additional file 2). The protein sequences of aerolysin in A. hydrophila showed some variations, the sequence identity between A. veronii aerolysin and A. hydrophila aerolysin varied between 69 and 98%. Shiga toxin 1 (Stx1) and Shiga toxin 2 (Stx2) were not found in any of these strains. Zonula occludens toxin (Zot) was found in 11 of the 31 strains (35.5%). The Zot proteins in A. veronii and Vibrio cholerae shared 36% of protein sequence identity.

Fig. 4

Prevalence of toxins in Aeromonas veronii strains isolated form fecal samples of patients with gastrointestinal diseases. Toxins identified in A. veronii strain A29V were further examined in other A. veronii strains by BLASTp. Conserved protein motifs were confirmed by pfam. Aerolysin and microbial collagenases (shaded in yellow) are secreted toxins

Discussion

In this study, we sequenced and assembled 25 genomes of A. veronii strains isolated from fecal samples of patients with gastrointestinal infections in Australia and conducted comparative genome analysis of 168 global A. veronii strains, including the 25 A. veronii genomes that we have sequenced and additional 143 A. veronii strains isolated from different sources in 18 countries in Asia, Europe, Africa, Oceania, North and South America.

Twenty-five genomes, including one complete genome and 24 draft genomes of A. veronii strains isolated from patients with gastrointestinal diseases were successfully obtained in this study (Table 1). Despite the increasing importance of A. veronii in causing human gastrointestinal diseases, only six genomes of A. veronii strains isolated from patients with gastrointestinal diseases were available in public databases prior to this study. Our 25 A. veronii genomes will provide a useful source for future research on A. veronii.

Global A. veronii strains including 168 strains from 18 countries were used for phylogenetic analysis (Table 2). These 168 A. veronii strains formed three phylogenetic clusters based on the core genome (Fig. 1). Each cluster had A. veronii strains from different sources, showing the ancestors of these three clusters were not determined by the isolation sites. Most of the A. veronii strains (88.7%) from various sources in different countries were in Cluster 1, showing that the majority of A. veronii strains globally were derived from a common ancestor. Strains isolated from the same environmental or animal sources often formed small groups within Cluster 1, most likely reflecting variations in A. veronii isolates obtained from a single site.

The majority of the 31 A. veronii strains (71%) isolated from fecal samples of patients with gastrointestinal diseases were closely related to strains isolated from fecal samples of the other human individuals, most of whom had gastrointestinal diseases (Table 3). Only 29% of A. veronii strains isolated from fecal samples of patients with gastrointestinal diseases were closely related to strains isolated from freshwater fish and domestic animals. This interesting finding suggests that the main source for human gastrointestinal infections of A. veronii was not from freshwater fish or domestic animals, although they can serve as potential sources of infection. In addition to freshwater fish, domestic animals and environmental samples, A. veronii has also been frequently isolated from drinking water and fresh water [2-6]. Human Aeromonas gastrointestinal infections most often occur in warm weather [1, 10]. Aeromonas species and their load in different types of drinking water and fresh water that is used for preparation of food should be monitored during different seasons, which will provide further information on the main sources that cause human Aeromonas gastrointestinal infections.

More than half of the 168 A. veronii strains (63.1%) examined in this study had T3SS. T3SS is used by pathogenic bacteria to directly inject effector proteins into eukaryotic host cells, which facilitates bacterial infection of host cells or causes host cell apoptosis [23]. T3SS in A. veronii is located in the AVI-1 genomic island (Fig. 2). The AVI-1 genomic island is also present in the chromosome of A. hydrophila strains and plasmids in A. salmonicida, suggesting that A. veronii most likely has acquired T3SS via horizontal gene transfer from other Aeromonas species. An additional interesting finding from this study was that T3SS was significantly less present in A. veronii strains isolated from fecal samples of patients with gastrointestinal diseases as compared to strains isolated from freshwater fish and domestic animals (Fig. 2). This further supports the view that most of the A. veronii strains causing infections in human gastrointestinal tract were from a different source.

Nearly 300 putative virulence factors were identified in the complete genome of A. veronii strain A29V (Fig. 3). This shows that multiple virulence factors contribute to the pathogenesis of gastrointestinal diseases caused by A. veronii. We further examined toxins in the 31 A. veronii strains isolated from patients with gastrointestinal diseases. Aerolysin, a secreted toxin, is a common virulence factor presenting in all A. veronii strains (Fig. 4). Aerolysin is a pore-forming toxin promoting osmotic lysis of host cells. Aerolysin in A. hydrophila was shown to perturb human intestinal epithelial tight junction integrity and cell lesion repair [24]. The second secreted toxin, microbial collagenase, was also found in all 31 A. veronii strains isolated from patients with gastrointestinal diseases (Fig. 4). Bacterial collagenases degrade collagen in animal cell extracellular matrix and are important bacterial virulence factors. Microbial collagenase in A. veronii is involved in the pathogenesis of diseases caused by this bacterium in fish[25]. Its pathogenic role in human diseases requires further characterization. A previous study reported detection of Stx1 and Stx2 toxin genes in some human Aeromonas isolates [25]. However, we did not find these toxin genes in any of the 31 strains isolated from patients with gastrointestinal diseases. Zot protein was found in 35.5% A. veronii strains. V. cholerae Zot protein damages intestinal epithelial barrier tight junctions and Campylobacter concisus Zot protein causes intestinal epithelial cell death [26, 27]. Multiple hemolysins in A. veronii were identified, which were demonstrated to be virulent to host cells in other bacterial species. The levels of toxins produced by different A. veronii strains remain to be further examined, which may contribute to their ability in causing human gastrointestinal diseases of different severity.

Conclusions

In summary, we report 25 genomes of A. veronii strains isolated from fecal samples of patients with gastrointestinal diseases, including one complete genome and 24 draft genomes. Analysis of 168 global A. veronii genomes including those we have sequenced show that the global A. veronii strains formed three clusters and the majority of A. veronii strains from various sources were from a common ancestor. Most of the A. veronii strains isolated from patients with gastrointestinal diseases were closely related to each other, with only a small percentage of these strains were closely related to A. veronii strains isolated from freshwater fish, domestic animals or environmental samples. Nearly 300 putative virulence factors were identified. Aerolysin, microbial collagenase and multiple hemolysins were present in all strains isolated from patients with gastrointestinal diseases. Zot toxin was only present in some strains. T3SS in A. veronii was in the AVI-1 genomic island identified in this study, and most likely acquired via horizontal transfer from other Aeromonas species and was significantly less present in A. veronii strains isolated from patients with gastrointestinal diseases as compared to strains isolated from freshwater fish and domestic animals. These findings provide novel information on source of infection and virulence of A. veronii in human gastrointestinal diseases.

Materials and methods

genomes used in this study

A total of 168 A. veronii genomes were analysed in this study, including 25 genomes sequenced in this study and 143 genomes publicly available. Currently, there are 156 A. veronii genomes available in the public databases, 13 genomes were excluded from this study due to lack of information on isolation hosts or country of isolation. The 25 A. veronii strains sequenced in this study were isolated from fecal samples of patients with gastrointestinal diseases at the Douglass Hanly Moir Pathology laboratory in Sydney, Australia, during routine diagnostic procedure.

Draft genome sequencing of 25 A. veronii strains

Sequencing and assembly of draft genomes of 25 A. veronii strains were conducted as described in our previous study [28]. Briefly, bacterial DNA was extracted using Gentra Puregene Yeast/Bacteria Kit (Qiagen, Chadstone, Victoria, Australia). Briefly, the DNA libraries were sequenced via the 150 bp or 250 bp paired-end sequencing chemistry on the MiSeq Personal Sequencer [29]. Reads were assembled using Shovill (v 1.0.5), and genome coverage was calculated using qualimap (v 2.2.1) [30]. Sequencing of the draft genome was performed in the Marshall Centre for Infectious Diseases Research at the University of Western Australia.

Complete genome sequencing of A. veronii strain A29V

A. veronii strain A29V was also subjected to genome sequencing using Oxford Nanopore sequencing technique. Bacterial DNA used for this part of genome sequencing was extracted with phenol–chloroform. Libraries were prepared using the Native Barcoding Expansion kit (EXP-NBD104, Nanopore) and the Ligation Sequencing Kit (SQK-LSK109, Nanopore). The libraries were then loaded onto a R9.4 flow cell (FLO-MIN106) and sequenced on the GridION sequencing device (Nanopore). The nanopore sequencing of A. veronii strain A29V genome was performed at the Ramaciotti Centre for Genomics at the University of New South Wales. Basecalling were performed using Guppy (v 4.0.14). Statistics of the reads were generated using Nanostat (v 1.5.0) and genome coverage was estimated using Minimap2 (v 2.17) and qualimap (v 2.2.1) [30].

To obtain the complete genome of A. veronii strain A29V, the reads of A. veronii generated by nanopore and Illumina MiSeq were used for hybrid assembly using Unicycler (v 0.4.7). The details of hybrid assembly were described in our previous study [31].

Annotation of the A. veronii genomes sequenced in this study

The complete genome of A. veronii strain A29V and 24 draft A. veronii genomes sequenced in this study were annotated using the NCBI Prokaryotic Genome Annotation Pipeline, Rapid Annotation using Subsystem Technology, and Prokka (v 1.14.5) [32-34].

Phylogenetic analysis

Core genome was generated using Roary (v3.12.0) [35]. The maximum likelihood phylogenetic tree based on core genome was generated using FastTree (v 2.1.11) [36]. The ANI values of each A. veronii genome against the genomes remaining 167 A. veronii strains were calculated using FastANI (v 1.32) [37].

Secretion systems were examined in the genomes of 168 A. veronii strains. Prokka annotated protein files of the 168 A. veronii strains were submitted to MacSyFinder, all available protein secretion systems were searched using the default settings [38]. Visualisation of T3SS was generated using EasyFig [39]. The nucleotide sequences of A. veronii T3SS were searched against the genomes of all bacterial strains in NCBI non-redundant nucleotide database using BLASTn [40].

Identification of A. veronii strains that were closely related to A. veronii strains isolated from fecal samples of patients with gastrointestinal diseases

In this study, 31 A. veronii strains that were isolated from fecal samples of patients with gastrointestinal diseases, including the 25 A. veronii strains that we have sequenced and additional six A. veronii strains in the public databases. The six A. veronii strains from public databases were strain ERR1305902-bin.15 from Denmark, strain 126–14 from China, two strains (FC951 and VBF557) from India, strain 312 M from Brazil, and a previously reported strain (BC88) from Australia.

Among the 168 A.veronii strains, the strain that had the highest ANI value against each of the 31 A. veronii strains isolated from fecal samples of patients with gastrointestinal diseases was identified as the most closely related strain.

Putative virulence factor in A. veronii strains isolated from patients with gastrointestinal diseases

Putative virulence factors in the complete genome of A. veronii strain isolated from a patient with gastroenteritis that was sequenced in this study were firstly identified through searches of the Virulence Factors Database (VFDB) [17, 41]. The presence of toxins in the 31 A. veronii strains isolated from patients with gastrointestinal diseases was then searched using BLASTp, and conserved protein motifs were confirmed using pfam [40, 42].

Statistical analysis

Fisher’s exact test (two-tailed) was used for analysis of the presence of T3SS in A. veronii strains isolated from different sources. p < 0.05 was considered to be statistically significant. Statistical analysis was performed using GraphPad Prism 7.

Additional file 1.

Additional file 2.