Abstract [Objectives] This study was conducted to explore the genetic evolution of Escherichia fergusonii in different countries and regions， and to clarify the genetic relationship of E. fergusonii in different countries and regions. [Methods] Bioinformatics method and bacterial 16S rRNA sequencing technology were used to sort out and prune 16S rRNA genes isolated in laboratory and searched in NCBI database to construct a molecular evolutionary tree for analysis and comparison. [Results] The direction of evolution of E. fergusonii has broken through regions， and there was cross evolution among continents. The origin of E. fergusonii was the Asian continent， and its adaptability to arid climate was not strong. [Conclusions] This study revealed the genetic evolution laws of E. fergusonii in the spread and mutation of livestock and poultry diseases， and provides a theoretical reference for the prevention and treatment of the disease.
　　Key words Escherichia fergusonii; Genetic evolution relationship; Evolutionary tree; Homology; Kinship; Origin
　　Escherichia fergusonii belongs to Escherichia in Enterobacteriaceae. It was named by Famer et al.[1] in 1985. It exists in the natural environment and the intestines of humans and animals， and is a rare conditional pathogenic bacterium in humans and animals[2]， which can cause human and animal bacteremia， trauma infection， pleural infection and diarrhea diseases. At present， there are not many reports about this bacterium at home and abroad. There are two reasons： one is that the bacterium is only pathogenic to patients with low immune function such as malignant tumors， and is not very invasive to normal people， and the other is that people have little or no knowledge of the bacterium and laboratory bacterial identification is not systematic and complete and lacks simple and practical routine identification paths[2]. 16S rRNA is a type of ribosomal RNA necessary for protein synthesis in all prokaryotes. In the long process of biological evolution， its gene sequence changes very slowly， and it shows a high degree of conservation in terms of base composition， nucleotide sequence， high-level structure and biological function， and is called the "fossil" of microorganisms[3-4]. 16S rRNA is composed of about 1 550 nucleotides， and its length can not only express enough interspecies polymorphism， but also facilitate sequence analysis. It can be used to mark the evolutionary distance and genetic relationship of organisms[5-6]， and is the method of species classification and identification based on the evolution of bacteria， which is suitable for inter-species identification within genera， and is known as the "gold standard" in taxonomy[3]. In this study， the 16S rRNA of various E. fergusonii strains searched in the laboratory and NCBI database were screened and sorted， and a molecular evolutionary tree was constructed through homologous sequence comparison analysis to explore the evolution of E. fergusonii in different regions. The study revealed the genetic evolution laws of E. fergusonii in the spread and mutation of livestock and poultry diseases， and provides a theoretical reference for the prevention and treatment of the disease. 　　Materials and Methods
　　Materials
　　After sending diseased dog samples to our laboratory for bacterial isolation and culture， a strain was obtained， which was sent to Sangon Biotech （Shanghai） Co.， Ltd.， obtaining a 16S rRNA sequence of 1 483 bp with the GenBank sequence number MK729001.1 （joe）. The remaining 29 groups of 16S rRNA sequences were from the National Center for Biotechnology Information （NCBI） by entering the NCBI homepage and downloading the desired sequences.
　　BioEditv7.0.5 software was downloaded from http：//www.mbio.ncsu.edu/BioEdit/bioedit.html. MEGA X software was downloaded from https：//www.megasoftware.net. DNASTAR software was downloaded from https：//www.dnastar .com.
　　Methods
　　Gene sequence screening
　　The Nucleotide BLAST alignment tool was run in the search box of the NCBI website， to retrieve the 16S rRNA sequences in the NCBI database. In this experiment， MK729001.1 （joe） was used as the query code for BLAST search. The target gene was input into the nucleotide search column of the NCBI homepage， and the online BLAST comparison tool was used to perform sequence comparison analysis. The sequences with high homology similarity and sequence identity of more than 95% were selected among the BLAST alignment results[7]. Meanwhile， other 16S rRNA sequences of E. fergusonii that had not been aligned were searched， to expand the geographical scope and diversity of the tested strains. The sequences were downloaded， and sequence information was collected in NCBI， including GenBank sequence number， strain number， sequence source region， and sequence length. The data was reduced， and non-E. fergusonii 16S rRNA genes and very short length sequences were excluded.
　　Homology comparison and molecular evolutionary tree construction
　　Gene sequence processing by BioEdit software
　　The BioEdit software ClustalW Multiple alignment method was applied to integrate， align， and analyze 30 sets of gene sequences， and a set of shorter sequences that made the comparison effect ideal was selected. From the left and right ends to the beginning and end of the sequences， respectively， the less homologous part of the gene sequences at both ends were cut off， and all the gene segments with large differences in sequence alignment were deleted. Sequence splicing was performed to obtain a gene sequence of about 1 439 bp.
　　Gene homology comparison
　　The MegAlign program in the DNASTAR software package was used to import the processed MK729001.1 （joe） and the sequences downloaded from NCBI GenBank， and analyze the gene homology therebetween to obtain a homology comparison table. 　　Molecular evolutionary tree construction by MEGA software
　　In the MEGA software， the Neighbor-Joining （NJ， neighbor-joining method） and p-distance models were used to perform molecular evolutionary tree construction according to the original parameters of the software， so the pruned and aligned 16S rRNA gene sequence generated a neighbor-joining tree. It was estimated that it could guide and support 100 000 times of repeated detection[7]. Bootstrap 1 000 was used to test the confidence of each branch of the molecular phylogenetic tree. In order to ensure the diversity of nucleotides in gene sequences， an online confidence limit calculator was used to estimate the 95% confidence limit[7-8].
　　Results and Analysis
　　Gene identification and screening
　　With MK729001.1 （joe） obtained by isolation and culture in our laboratory as the target gene， through the Nucleotide BLAST option in BLAST， a total of 100 sets of similar gene sequences were obtained. Meanwhile， 50 sets of unaligned 16S rRNA gene sequences of E. fergusonii were searched in the NCBI database. After further screening， 30 16S rRNA gene sequences with clear sources and moderate length were obtained[9]， as shown in Table 1.
　　Gene homology comparison and molecular evolutionary tree construction
　　According to the data in Table 1， the MegAlign program in the DNASTAR software package was used for homology comparison， as shown in Fig. 1. The MEGA X software was used for multiple sequence alignment， and the neighbor joining method was applied to construct a neighboring molecular evolutionary tree[10]， as shown in Fig. 2.
　　It can be seen from Fig. 2 that the 30 E. fergusonii strains showed two large branches in the evolutionary direction， and each large branch was composed of multiple different small branches. Most of the strains were clustered into one large branch， including 27 strains， which were from different countries and regions， and other 3 strains belonged to another large branch. Among them， the two Brazilian strains MG429704.1 （S204） and MH304302.1 （72431） had very close genetic distances， sharing a homology of 97.8%; three Chinese strains MK038938.1 （UTI 2）， MK287772.1 （D3-10） and MK168572.1 （CICC 24137） showed very close genetic distances， and the homology was 98.3% （UTI 2 and D3-10）， 99.2% （D3-10 and CICC 24137）， and 98.8% （UTI 2 and CICC 24137）， respectively; and the two Mexican strains MH158274.1 （38CB6） and MH158272.1 （31CB1） was very close in genetic distance， sharing a homology of 98.4%， and their genetic distances with the French strain NR 074902.1 （ATCC 35469） were also very close. In addition， from Fig. 2， we also found on the large branch of strains concentrated in distribution， 12 strains from different regions： MG429704.1 （S204 Brazil）， MH304302.1 （72431 Brazil）， MK038938.1 （UTI 2 China）， MK287772.1 （D3-10 China）， MK168572.1 （CICC 24137 China）， NR 027549.1 （ATCC 35469 Canada）， KF938590.1 （Sw2 Korea）， KJ626246.1 （E45 Netherlands）， KR822244.1 （Sum4 Pakistan）， KR905687.1 （K3 Pakistan）， KY357308.1 （GT2 Iran）， LC462163.1 （JCM 5899 Japan）， had very close genetic distance as well. From Fig. 1， it can be found that the 12 E. fergusonii strains shared very high homology. 　　Agricultural Biotechnology2020
　　Meanwhile， it can be seen from Fig. 2 that in the branches of the molecular evolutionary tree， two Chinese strains of MH819636.1 （HBUAS54403） and MK729001.1 （joe） had a long genetic distance， and were far away from other three Chinese strains; there was a longer genetic distance between the two Japan strains LC462163.1 （JCM 5899） and NR 114079.1 （NBRC 102419）， of which LC462163.1 （JCM 5899） had a closer distance with HQ214033.1 （KLU01 India）， and NR 114079.1 （NBRC 102419） were clustered with France NR 074902.1 （ATCC 35469） and two Mexican strains： MH158274.1 （38CB6） and MH158272.1 （31CB1）， into one branch; the US KJ726590.1 （Gen8） and Saudi Arabia MG818962.1 （FA-5） were clustered into one branch; India MK598698.1 （UTIEF2） and Chinese MK729001.1 （joe） were clustered into one branch; and the two Dutch strains KJ626246.1 （E45） and KJ626262.1 （E3） were in different branches of the molecular phylogenetic tree and had a long genetic distance.
　　In the other large branch， only three strains of E. fergusonii gathered， and they were far apart in the molecular evolutionary tree. The South Korea MF973086.1 （KC-Tt-R3） was a single branch， and was in completely different large branches with KF938590.1 （Sw2 Korea）， so the genetic distance was very far. Meanwhile， the Indian strain FN433037.1 （CCM27B） on this large branch was genetically far apart from MK598698.1 UTIEF2 （India）. There was only one strain of MK584171.1 （EsF） in Tunisia at the northern end of Africa， which was far away from other strains in the molecular evolutionary tree， and shared homology between 91% and 98.1% with them， less than 95% mostly.
　　Conclusions and Discussion
　　Thirty strains of E. fergusonii in different countries and regions were divided into two major branches in the molecular evolutionary tree[10]. Most of the strains were clustered into one branch， including 27 strains， and a few strains were clustered into another branch， including 3 strains， indicating that the general evolutionary direction of the bacterium was determined and there are a few strains maintaining the original evolutionary direction， which might be related to their own ability to resist external environmental interference.
　　In many countries and regions， including China， the bacterium showed the closest genetic evolution in the same country and region， indicating that the bacterium was still highly conserved in the same region[11]. It can be found from Fig. 1 and Fig. 2 that the genetic distances between 12 strains of this bacterium from different regions of the world were relatively close， and they had high homology， indicating that the genetic differences of the isolated strains in different regions were small[13]. It is speculated that the bacterium has not yet rapidly mutated due to the widespread spread of the bacterium worldwide， and the degree of variability between strains is not large. It still has a high degree of conservation in the world， and the evolutionary direction has broken through regions[7，12]. 　　The two Chinese strains， MH819636.1 （HBUAS54403） and MK729001.1 （joe）， were genetically far apart. One of them formed a single branch， and the other was clustered with HQ214033.1 （KLU01 India）. Meanwhile， of the two strains in the Netherlands KJ626246.1 （E45）， KJ626262.1 （E3）， one formed a single branch， and the other was clustered into the group of the 12 strains from different regions. In addition， strains from Japan， France and Mexico were clustered into one branch， strains from the United States and Saudi Arabia were also clustered into one branch， and they finally clustered into a large branch. It is speculated that these strains had a common ancestor a long time ago， evolved from one strain， and gradually formed many new mutant strains in the subsequent evolutionary process[7，12-13]. It can be seen from Fig. 2 that strains from different regions had complex distribution in the molecular evolutionary tree， and interlaced and clustered into different branches， which were either on the same continent or on different continents. For example， the US KJ726590.1 （Gen8） and the Saudi Arabia MG818962.1 （FA-5） were clustered into one branch in North America and Eurasia， and it is speculated that this bacterium has cross-evolution between continents[7]. Meanwhile， the branch node of strains on the Asian continent was relatively higher than those on other continents. For example， among the large branches with fewer strains， the MF973086.1 （KC-Tt-R3 Korea） node ranked the highest. Among the large branches with more strains， MK729001.1 （joe China） ranked higher than KJ726590.1 （Gen8 USA）， KX673992.1 （113-3T Switzerland）， and MG576033.1 （OX1012 Iceland）， and its original origin might be in the Asian continent[7].
　　When searching for data and literatures， there are fewer examples of the bacterium found in Africa. In Fig. 1 and Fig. 2， a Tunisian strain， MK584171.1 （EsF）， was far from other strains in the molecular evolutionary tree， and shared a homology less than 95% with most strains. It is speculated that this bacterium is not adaptable to the arid climate environment in Africa.
　　With the use of bacterial antibiotics， bacterial resistance has continued to increase. Recent studies have reported that E. fergusonii has gradually tolerated commonly used antibacterial drugs for the treatment of intestinal bacterial infections. Because its biological classification and drug-resistant phenotype are very similar to those of E. coli， it may cause intestinal infections of different hosts[14]. As an important zoonotic pathogen， this bacterium is a serious threat to human health and production， but there are few reports on its related research in China[2，14]. In order to better prevent the spread of E. fergusonii， quarantine at ports or customs should be strengthened， and foreign populations or commodities that may carry pathogenic bacteria should be strictly examined. Meanwhile， the government should strengthen the prevention of biological hazards， do a good job in epidemic prevention， and continuously improve relevant laws， regulations and preventive measures， so as to prevent the spread of E. fergusonii in different regions. 　　References
　　[1] FARMER JJ， FANNING GR， DAVIS BR， et al. Escherichia fergusonii and Enterobacter taylorae， two new species of Enterobacteriaceae isolated from clinical specimens[J]. Journal of Clinical Microbiology， 1985， 21（1）： 77-81.
　　[2] TAO YY， SUN MY， LI JH， et al. Study of the biological characteristics of Escherichia fergusonii and 16S rRNA detection[J]. Journal of Pathogen Biology， 2013， 8（3）： 229-232. （in Chinese）
　　[3] YU C， GUO HY， WEI JL， et al. Application of 16S-23S rRNA gene sequence in bacterial identification[J]. China Animal Husbandry & Veterinary Medicine， 2012， 39（2）： 57-60. （in Chinese）
　　[4] LIU C， LI JB， RUI JP， et al. The applications of the 16S rRNA gene in microbial ecology： Current situation and problems[J]. Acta Ecologica Sinica， 2015， 35（9）： 2769-2788. （in Chinese）
　　[5] LI X， GAO Q. Application of 16S rRNA gene sequence analysis in clinical microbiology[J]. Journal of Microbes and Infections， 2006（3）： 184-186. （in Chinese）
　　[6] HARMSEN D， KARCH H. 16S rDNA for diagnosing pathngens： a living tree[J]. ASM News， 2004（70）： 19-24. （in Chinese）
　　[7] LIANG Y， LIU YF， YANG JD. Analysis of genetic evolution of Actinobacillus pleuropneumoniae[J]. Heilongjiang Animal Science and Veterinary Medicine， 2018（9）： 103-105， 110. （in Chinese）
　　[8] WANG TT， CONG YH， LIU JG， et al. Cloning and functional analysis of a WRKY28-like gene in soybean[J]. Acta Agronomica Sinica， 2016， 42（4）： 469-481. （in Chinese）
　　[9] ZHU PP， LI YM， HOU YZ， et al. Construction and analysis of molecular phylogenetic tree of Rosaceae plant based on PGIP gene[J]. Hubei Agricultural Sciences， 2016， 55（21）： 5672-5676， 5728. （in Chinese）
　　[10] WANG LL， ZHENG L， LU C， et al. Phylogenetic analysis of porcine circovirus type 2 isolated from Tianjin and its surrounding areas[J] China Animal Husbandry and Veterinary Medicine， 2018， 45（5）： 1331-1340. （in Chinese）
　　[11] ILLIASSOU HS， BODO L， HARVILL ET. Environmental origin of the genus Bordetella[J]. Frontiers in microbiology， 2017（8）： 28.
　　[12] CAI ZH， CHEN Q， ZHAO R， et al. Analysis of genetic evolution of two H9N2 subtype avian influenza virus isolates in Xiamen[J]. Animal Husbandry and Veterinary Medicine， 2018， 50（12）： 73-79. （in Chinese）
　　[13] CHANG HT， WANG XW， LIU HM， et al. Identification and genetic evolution analysis on the Omp P2 gene of Haemophilus parasuis from local aardvark[J]. Journal of Anhui Agricultural University， 2014， 41（3）： 371-374. （in Chinese）
　　[14] LU Y， ZHAO HY， QI XL， et al. Identification and drug sensitivity test of Escherichia fergusonii[J]. Progress in Veterinary Medicine， 2013， 34（5）： 48-51. （in Chinese）