Multilocus Variable Number of Tandem Repeat Analysis for molecular typing of uropathogenic Escherichia coli


 Background: The aim of this study was genotyping of Uropathogenic Escherichia coli (UPEC) based on Variable Number of Tandem Repeats (VNTRs) sequences. Methods: E. coli strains isolated from urine samples were included in this study. Seven VNTR loci were subjected to Multilocus variable-number tandem repeat analysis (MLVA) based on PCR amplification. Then data was analyzed via online mlvaplus software and the information was displayed in the form of MST analysis. Results: A total of 100 E. coli strains were isolated and subjected to the study. MLVA was able to differentiate 56 different genotypes. Also, the technique could classify E. coli isolates in 5 clonal complexes. Based on UPGMA dendrograms, E. coli isolates were classified into 4 clusters (clusters A to D). The strains associated with Complex No. 1 appeared to be dominant pathogens of UPEC in Tehran's patients. The present study provides valuable insights into the genetic relationships of E. coli isolates recovered from clinical cases in a major hospital in Iran. Conclusions: The analysis of MLVA profiles using the MST algorithm showed the usefulness of the MLVA method in the classification of uropathogenic E. coli collected in different periods. We evaluated MLVA in a laboratory equipped with simple molecular equipment. Based on these results, it has been assumed that the E. coli strains were derived from a limited number of clones that have undergo a small genetic change during this period.


Background
Escherichia coli (E. coli), a agellated member of Enterobacteriaceae, is an essential member of the normal ora in the human gut, but many kinds of symptoms are caused by various E. coli pathotypes with a wide range of virulence factors. For instance, septicemia causing E. coli (SCEC) and neonatal meningitis producing E. coli (NMEC) are ExPEC pathotypes as the main cause of septicemia and pediatric meningitis respectively. Another important member of EXPEC pathotypes is uropathogenic E. coli (UPEC) which causes urinary tract infections (UTIs) [1][2][3][4][5][6].
Different molecular techniques such as Repetitive element sequence-based PCR (rep-PCR), Random Ampli ed Polymorphic DNA (RAPD)-PCR, Ribotyping, Pulsed-eld gel electrophoresis (PFGE), and Multiple Locus Variable-number Tandem Repeat Analysis (MLVA) have been used for genotyping of different strains of E. coli up to now. Among them, the MLVA method has been proposed as a new molecular method for bacterial genotyping [16,17]. In MLVA technique, only a PCR-based method is required, while other sophisticated method such as PFGE need specialized and expensive equipment's. In this way, MLVA is a low-cost alternative to PFGE, especially in developing countries [18]. It can also be used in simple laboratories that are only equipped with a PCR machine [19]. In this study, we tried to design a new, fast, and low-cost method for genotyping of uropathogenic E. coli by choosing suitable VNTR repeat sequences.

Methods
Sample collection and E. coli isolation Over a 12-month sampling period (from October 2018 to October 2019), E. coli isolates were recovered from patients with UTI in a major hospital in Tehran, Iran and subjected to the current study. Isolation of the E. coli was done by conventional standard biochemical and serological methods. Well-isolated colonies of puri ed E. coli were resuspended in trypticase soy broth with 20% glycerol and stored in 70 °C for long-term storage [8]. DNA preparation A pure culture of E. coli was plated on nutrient agar (Merck, Germany) and incubated overnight at 37 °C. A single colony was removed from the plate, suspended in 200 µl of sterile deionized water, and boiled for 15 min. After centrifugation at 6,000 g for 8 min, the supernatant was transferred into a new tube for subsequent PCR analysis [20]. Purity (A260/A280) and concentration of extracted DNA were then checked (NanoDrop, Thermo Scienti c, Waltham, MA, USA). The truth of the DNA was assessed on a 2% agarose gel stained with ethidium bromide (0.5 µg/mL) (Thermo Fisher Scienti c, St. Leon-Rot, Germany).

MLVA assay
For genotyping, using the MLVA technique, seven loci VNTRs were used based on the protocol of O Gorgé, S Lopez, V Hilaire, O Lisanti, V Ramisse and G Vergnaud [21]. These seven VNTR loci were: ms06, ms07, ms09, ms11, ms21, ms23 and ms32. The primer sets for PCR ampli cation of these VNTR loci were previously reported by O Gorgé, S Lopez, V Hilaire, O Lisanti, V Ramisse and G Vergnaud [21] (Table 1). PCR was performed in 25 µl volume including 1X PCR buffer (50 mmol/L KCl, 10 mmol/L Tris, pH = 9), 2.5 mmol/L MgCl2, 0.2 mmol/L of each primer with 1 U of TaqDNA polymerase (CinnaGen Co., Iran), and 4 µl of the crude DNA extract. We introduced the number of replicates for each locus in each isolate, along with other characteristics of the isolates into Microsoft Excel, and then put these data on the WWW.mlvaplus.net site. The frequency of VNTR loci repeats is entirely different from each other by electrophoresis on the agarose gel. MST is a convenient complementary tool to cluster multiple isolates and visualize the relative diversity within different lineages.
A dendrogram of genetic relationships was also generated using the unweighted pair group method with arithmetic averages (UPGMA) method [22]. Furthermore, Simpson's index of diversity (D) and 95% con dence intervals (CI) for each VNTR locus were calculated using version 2.0 phyloviz software (http://www.phyloviz.net). For dendrogram analysis, the UPGMA method and Dice similarity coe cient were used to analyze VNTR data [23,24].

MLVA analysis
One-hundred E. coli strains were identi ed, of these, 70 isolates were belonged to the patients with complete biographical information and subjected to MLVA. Figure 1   The analysis of the number of VNTR repetitions using the MST algorithm showed that in 70 isolates of E. coli, 56 different genotypes were observed. However, due to the high similarity of the pro le of isolated E. coli with each other and also for the better separation of complex colonies, the criterion of difference in a locus for grouping isolates (multiple clones) was considered. Accordingly, E. coli strains were classi ed in 5 clonal complexes.

Evolutionary relationship between isolates
Each complex group represents the evolutionary relationship between isolates. In other words, isolates within each group have a genetic correlation that is closer to each other and is likely to have the same origin. In our study, Complex No. 1 was the largest clonal group (Fig. 2). E. coli isolates were classi ed in 4 clusters (A cluster to D) based on the similarity (Fig. 3).

Discussion
In this study, we successfully applied the MLVA method to analyze uropathogenic E. coli strains isolated from clinical samples.
MLVA method does not require a high level of expertise and is capable to provide reasonable results about monitoring of outbreaks and clonal spread of bacterial isolates over a short period of time. It has been successfully applied to investigate the clonal relationship and epidemiology of clinical E. coli isolates [21,25,26]. In this study, the MLVA technique was able to detect 56 different genotypes among 70 E. coli strains isolated from patients with UTI in a major hospital in Tehran, Iran, indicating a high diversity of UPEC isolates genotypes.
Also, MLVA technique could classify the stains in 5 clonal complexes. In examining UPGMA dendrograms, uropathogenic E. coli isolates were classi ed according to the similarity in 4 clusters (clusters A to D). Each complex group represents an evolutionary relationship between the isolates. In other words, isolates belonging to each group have a genetic association with each other and are likely to have the same origin. In our study, Complex No. 1 was the largest clonal group. The strains associated with Complex No. 1 appeared to be the dominant pathogens of UPEC in the hospital under study.
In the present study, we used markers that can be easily separated by electrophoresis on the agarose gel.
On the other hand, the evolution rate of different VNTRs differentiated the clonal relationships. Typically, loops that have larger replication sizes are slower than circuits that have smaller duplicate sizes. But loci with a slower evolutionary velocity are suitable for the investigation of clonal relationships between isolates that have evolved over a more extended period. Therefore, the type of VNTR locus is also in the analysis [27,28].
The applied technique could classify E. coli isolates in 5 clonal complexes. Each complex group represents the evolutionary relationship between isolates. In other words, isolates belonging to each group have a genetic link with each other and are likely to have the same origin. In our study, Complex No.
1 was the largest clonal group. The strains in Complex No. 1 appeared to be the prevalent strains of UPEC disease, with about two-thirds of the population receiving all strains.
The ms32 has the highest number of alleles. The ms06 and ms23 loci have the smallest amount of alleles, and the ms06, ms09, ms21, and ms23 loci have null alleles. In the dendrogram analysis, The UPGMA of uropathogenic E. coli isolates is classi ed according to the similarity in 4 clusters (clusters A to D). We also found that to better evaluate VNTR loci for the study of epidemics or genotyping, more and more numbers of E. coli are needed. Besides, the use of a higher number of VNTR loci will be more bene cial for better differentiation of clonal groups.
Mellor et al. used the MLVA method to determine the genetic diversity of E. coli O157 producing Shiga toxin venom. There was some evidence that E. coli O157 strains collected from different regions may also show genetically diversity. To study the amount of this variation, Shiga toxin bacteriophage insertion sites (SBI), lineage-speci c polymorphisms (LSPA-6), multilocus variable-number tandem repeat analysis (MLVA), and a tir 255T > A polymorphism were used to enquire 606 isolates from Australian and U.S. cattle and human. All analyses of collected data showed a signi cant connection between the source of origin and multilocus genotypes (P < 0.0001) and it con rms the evolution of enterohemorrhagic E. coli O157 in Australia and the United States [29].
MLVA was also used by Bustamante et al. for subtyping of 202 STEC isolates from different sources to get information about the genetic diversity of native STEC in Argentina. Two different MLVA protocols were used in their investigation, one for O157 and the other for a generic E. coli assay. As a result of their studies, MLVA was mentioned as a very sensitive tool to STEC subtyping and it showed the diversity in many serotypes [30].
Naseer et al. used the MLVA and multi-locus sequence typing (MLST) for subtyping of a total of 100 E. coli isolates from Spain, UK, Sweden, and Norway (n = 19, 20, 24 and 37 respectively) and the obtained data showed equivalence between the MLST pro le and MLVA typing results [31].
In a study in Norway by Sekse et al, on 142 E. coli O26 isolated from 491 fecal, PFGE and MLVA was performed for investigation of the genetic relationship between the strains and 63 different PFGE pro les and 22 MLVA pro les were identi ed. Despite some differences in result between MLVA and PFGE, comparison by partition mapping showed good overall agreement between two methods. A few isolates of EPEC O26: H11 had a close relationship to STEC O26: H11 [32].
VNTRs regions are widely used for bacterial subtyping but it should not be overlooked that the hyper variability in VNTR loci can be problematic in trying to predict the relationships between isolates. It is also worth noting that knowing the rate of VNTR mutation and its effective factors can introduce the MLVA for legal epidemiological or microbiological research [33].
In a study in Japan on a total of 57 isolates from patients, 20 types were identi ed by using MLVA while they were classi ed into 23 PFGE types and the results were almost the same. In another study in the same city, a total of 24 isolates from different sources and one outbreak occurred in central parts of the city was investigated by MLVA an PFGE. The results showed a good correlation between these both methods. However, the MLVA typing proved to be a much more comfortable and more rapid method for the analysis of E. coli O157: H7 strain relatedness to identify transmission paths. Hence, MLVA method was appeared to be a convenient method for typical typing in many laboratories, including public health agencies and even in hospitals [33][34][35][36].

Conclusions
The study population consisted of E. coli strains isolated from patients with UTI in a major Hospital in Tehran. The analysis of MLVA pro les using the MST algorithm showed the usefulness of the MLVA method in the classi cation of uropathogenic E. coli collected in different periods. We also found that MLVA technique was a fast, easy and low-cost method for genotyping of uropathogenic E. coli. We evaluated this technique in a laboratory equipped with simple molecular equipment. The present study provides valuable insights into the genetic relationships of E. coli isolates recovered from clinical cases in a major hospital in Tehran, Iran. Based on these results, it has been assumed that the uropathogenic E. coli strains were derived from a limited number of clones that have undergo a small genetic change during this period.

Consent for publication
There was no consent for publication.

Availability of data and materials
All data generated or analyzed throughout this research are included in this published article.

Competing interests
The authors declare that they have no competing interests Funding Not applicable.
Authors' contributions RR and FSD designed the study and carried out the PCR genetic alignment and typing. MM and OF supported the study and carried out the samples collection, bacterial isolation and statistical analysis. RR carried out the writing and drafting of the manuscript. All authors road and approved the nal manuscript. Figure 1 Polymorphism of 3 VNTR loci in different E. coli isolates. This image illustrates how the number of repeats can be directly deduced by manual reading.