Escherichiacoli DERIVED FROM DIFFERENT SOURCES SHARE ANTIGENIC CHARACTERISTICS WITH Shigellaboydii 18AND VIRULENCE FACTORS WITH ENTEROTOXIGENIC E. coli

1. Department of Public Health, Faculty of Medicine, Universidad Nacional Autónoma de México. 2. Laboratorio de Patogenicidad Bacteriana, Unidad de Hemato-Oncología e Investigación; Hospital Infantil de México Federico Gómez/División de Investigación, Facultad de Medicina, Universidad Nacional Autónoma de México. 3. Department of Microbiology and Parasitology, Faculty of Medicine, Universidad Nacional Autónoma de México. 4. Precision Global Health, Seattle, WA, USA. ...................................................................................................................... Manuscript Info Abstract ......................... ........................................................................ Manuscript History

DNA Extraction. Extraction of DNA from the strains was carried out using the boiling method reported by Islam (2006). Briefly, the strains were inoculated into Luria-Bertani (LB) broth and incubated at 37°C for 18-24 h. One mL of the bacterial suspension was centrifuged at 13000g for 5 mins. The supernatant was decanted and the pellet was homogenized with 200 µL ultrapure water. The resulting suspension was boiled for 10 mins and placed in an ice 631 bath for 5 mins. The suspension was mixed by mechanical shaking with a vortex and centrifuged at 13000g for 10 mins before taking 100 µL of the supernatant to be preserved frozen (-20°C) until use.
Genes to differentiate between E. coli and Shigella. As reported by Horáková (2006), primers were used in a multiplex PCR form (Table 1) to determine the lacZ,uidA and cyd genes. The primers for ipaH and lacY were designed in the laboratory with the nucleotide sequence for the ipaH gene being obtained from the complete Shigella genome analyzed by the ShiBase database (http://www.mgc.ac.cn/ShiBASE/). With regards to the primers for lacY, the complete E. coli O157:H7 Sakai genome was generated with the Genbank access number BA000007.2 (Makino, 1999;Ohnishi, 2000).The size of the amplicon was analyzed and compared against ShiBase BLAST. This analysis corresponded to the ipaH-5 from S. sonnei strain 046, ipaH-1 from S. boydii 4 strain 22, ipaH-7 from S. flexneri 2ª strain 301, ipaH-6 from S. dysenteriae 1 strain 197, and finally ipaH-2 (a pseudogene). All of the PCR primers mentioned here were designed by the free software program PRIMER3 (http://primer3.ut.ee/). For the pair of primers lacY and ipaH, a duplex PCR was used with the following parameters: 35 amplification cycles with initial denaturationat 94°C for 30 sec, annealingat 60°C for 25 sec, extension at 72°C for 30 sec and final extension at 72°C for 5 mins. In addition, the primers were used to detect the ltA, sth, stp, cfaI, cs1, cs3 y cs21 genes from ETEC employing the previously described conditions (Bekal, 2003;Rodas, 2009; Mazariego-Espinosa, 2010; Chattopadhyay, 2012).
The same PCR technique was used to determine the presence of the wzx (flippase) and wxy (polymerase) genes, which relate to the biosynthesis of the somatic antigen of S. boydii 18. The nucleotide sequences of the wzx and wzy primers (Table 1) were obtained from the complete S. boydii 18 deposited in the GenBank with access number AY948196 (Feng, 2005). For this PCR, the following parameters were used: 30 amplification cycles with denaturing at 95°C for 1 min, annealing at 58°C for 1 min, extension at 72°C for 30 sec and final extension at 72°C for 5 min (Table 1).
Phylogenetic Groups.Primers and conditions for quadruple PCR reported by Clemont (2013) ( Table 1) were used to define the phylogenetic group of E. coli strains. In each case, the amplification products of the DNA obtained by PCR were analyzed by electrophoresis in an agarose gel at 1.8% with 100 Volts and the DNA was stained with GelRed (BioLabs). The amplicons were viewed using UV light in an image reader (Biosens SC805, Gel Imagine Sistems). Amplicons were considered to be positive if they presented the same size of pair bases as those presented by the positive controls.
Results:-Biochemical Identification.All 23 strains presented a biochemical profile typical of E. coli in terms of fermentation of glucose, lactose, maltose, raffinose, sorbitol, xylose and sucrose; lysine decarboxylation; and gas and indole production.
Absorption Tests. Inorder to confirm that the reaction against S. boydii 18 was specific, the sera against S. boydii 18 and E. coli 44037 antigens were absorbed with heterologous antigens. The absorption tests also included an antiserum against E. coli O152 that reacted against the S. boydii 18 antigen at a low titer level (1:100). However, antisera against S boydii 18 and E.coli 44037 did not react against the E. coli O152 antigen. Absorption of the S. boydii 18 serum with the E. coli O152 antigen did not change the reaction against the antigens from S. boydii 18 and E. coli 44037. In contrast, when the S. boydii 18 serum was absorbed with the 44037 antigen, agglutination against E. coli 632 44037 and S. boydii 18 was eliminated completely. In a similar way, absorption of the E. coli 44037 antiserum with the S. boydii 18 antigen, completely eliminated the reaction against the previously mentioned antigens (Table 2).
Virulence Genes to Differentiate Shigella and E. coli. PCR detection of the wzx and wzy genes was positive for both genes in all 23 (100%) strains of E. coli and S. boydii 18, while the same test for E. coli O152 was negative. The PCR test to detect lacZ, uidA, cyd,lacY, and ipaH was positive for the first four genes in 23 strains and for E. coli O152 but negative for the ipaH gene ( Table 3). The same test for S. boydii 18 showed a positive PCR reaction for the uidA, cyd and ipaH genes and a negative reaction for the lacZ and lacY genes. The PCR test to determine the presence of the ETEC ltA and stp genes was positive in 14 (61%) and 6 (26%) of strains, respectively. With regards to the colonization factors cfal and cs3, 21 (91%) and 13 (57%) of the strains were positive, respectively (Table 3).
Phylogenetic Groups. Analysis to define the phylogenetic group of the strains showed that 5 (22%) corresponded to group A, 5 (22%) to Group B1, 8 (35%) to Clade I, 1 (4%) to Group D1 and 3 (13%) to Group B2. The phylogenetic group was unable to be determined for 4 (17%) strains. Both the S. boydii 18 and E. coli O152 strains corresponded to group A.
Antimicrobial Resistance. Of the E. coli strains in the study, 6 showed resistance to one and three antimicrobials (Table 4). Of these six, two (33%) of the strains, both from Egypt, showed resistance against one antimicrobial (NA or TE), 1 (17%) was resistant to FOX, NA and TE, and another one (17%) was resistant to CAZ, NA and TE. Finally, only 2 (33%) of the isolated strains from dairy cows were resistant to NA or TE.

Discussion:-
This study characterized the phenotypic and genotypic profiles of 23 E. coli strains isolated in different years from children with and without diarrhea in Egypt and Mexico. In addition, E. colistrains isolated from dairy cattle from a herd in Jalisco State, Mexico were also analyzed. All of the strains presented a biochemical profile characteristic of E. coli that reacted specifically against a rabbit serum prepared against the somatic antigen of S. boydii 18. However, in contrast toS. boydii 18, which is a non-motile strain, the majority of theE. coli strains isolated in Egypt, presented the flagellar antigen H2, while the strains from Mexico presented the H2, H3, H9, H16, H48 antigens or were nonmotile (H-).
PCR tests used to identify the wxy and wxz genes, which are related to the biosynthesis of the O antigen of S. boydii 18, showed that these two genes were present in all E. coli strains. Together with the microagglutination tests, these results suggest that the 23 strains form a new sero-variety of E. coli. These results were similar to those from other studies that reported the presence of Shigella antigens in strains of E. coli (Navarro, 2010;Iguchi, 2011;Iguchi, 2015). Recently, Iguchi (2015) reported antigenic relationships between E. coli O38 and S. dysenteriae 8; E. coli O169 and O183 with S. boydii 6 and 10. Although 21 shared O antigens were recognized between E. coli and Shigella (Iguchi, 2015), in the case of S. boydii 18 no antigenic cross-reaction was observed with any other E. coli or Shigella somatic antigen (Liu, 2008). Previously, our laboratory reported E. coli strains isolated from three geographic zones with antigens from S. boydii 16 showing characteristics of ETEC (Navarro, 2010). In addition, antigenic relationships between S.dysenteriae 10 and E. coli have been found that presented characteristics of E. coli strains producing the Shiga toxin (STEC) (Iguchi, 2011). The presence of the wzx and wzy genes in the E. coli strains confirm the existence of common epitopes similarto those found in the linear pentasaccharide of the repeat units of the S. boydii 18 O antigen. This linear pentasaccharide consists of three carbohydrate residues made up of rhamnose, a residue of alpha-d-galacturonic (D-GalA) acid and a residue of N-acetylgalactosamine (N-GalNAc) (Feng, 2005).
The PCR results to determine the E. coli pathotype showed that a significant number of strains belong to the ETEC group due to the fact that they contained ltA and stpgenes. The presence of both ItA and biosynthesis of the S. boydii 18 O antigen suggests the acquisition of these genes by a horizontal transfer system (Reid, 2000;Gogarten, 2002). The E. coli strains with S. boydii 18 antigens were found in fecal samples from children and dairy cattle. These cows could provide natural reservoirs of this bacteria and be related to the transmission of pathogens associated with diarrheal diseases.
Since the E. coli strains in this study presented an antigen identical to S. boydii 18, further investigation was made to see if these strains presented the uidA, cyd, lacY, lacZ and ipaH genes in order to establish whether the strains contained Shigella genes. The results showed that all the strains were positive for uidA, cyd, lacY and lacZ genes 633 but negative for ipaH. The presence of these four genes in the strains indicated that their genotypic identity was E. coli. In contrast, S. boydii 18 presented ipaH, uidA and cyd genes but lacked the lacY y lacZ genes. Horakova (2008) reported that the lacY gene, which is responsible for lactose fermentation, is a molecular marker to identify Shigella strains, and in this study, the strains lacked this gene. However, the lacY gene is present in E. coli as well as in Enterobactercloacae and Citrobacterfreundii. The other important marker that differentiates between Shigella and E. coli is the uidA (-glucuronidase) gene that is present in both types of strains, including enteroinvasive E. coli (Pavlovic, 2011).
Analysis of the phylogenetic groups showed that as many strains isolated from children in Mexico and Egypt, as well as strains from dairy cattle, belonged to Clade I. Reports indicate that Clade I group strains that present microbiological characteristics similar to those of E. coli that making them indistinguishable from this bacteria. However, genotypic analysis shows that they are considered as a divergent group of typical E. coli but that finally they are classified as phylogroups of E. coli (Luo, 2011;Clermont, 2013).
Further to the Clade I strains, our study identified strains belonging to groups A and B1. These groups comprised mainly of intestinal type E. coli strains that form part of the commensal microbiota of the human intestine, while groups B2 and D are recognized as comprising strains originating from outside the intestine with both pathogenic capacity and more virulence factors than those in groups A and B1 (Duriez, 2001;Nowrouzian, 2005). However, group B1 is found more frequently in herbivorous animals, such as cows, sheep and goats (Baldy-Chudzik, 2008; Carlos, 2010;Ziebell, 2008). In addition, group B2 is found in the human intestine (Carlos, 2010) forming part of the resident microbiota that can colonize the intestine of humans for anumber of weeks (Nowrouzian, 2005). The presence of B2 strains in the human intestine could indicate that these strains are more adapted to the human intestine.
The lack of resistance to the antimicrobials that was found in themajority of the strains in this study was interesting. Only one of the strains from Mexico presented resistance (CAZ, NA and TE) to the antimicrobials used, in contrast to three strains from Egypt that showed resistance. Of these three, two were resistant to one antimicrobial (NA or TE) and the other to three antimicrobials (FOX, NA and TE). Of the strains isolated from dairy cows, only two presented resistance to NA or TE. This lack of resistance overall correlates with the results from a previous study of E. coli with an S. boydii 16 antigen (Navarro, 2010) but different from other results (Estrada-García, 2005; Amábile-Cuevas, 2010) that arose from studies in Mexico in which resistance to ciprofloxacin, ampicillin, trimethoprim/sulfamethoxazole and tetracycline was reported in E. coli strains isolated from environmental and clinical samples. An important note with regards to the Mexican strains is that these were isolated in the mid-1980's in a rural region (Cravioto, 1990).
The results from this study suggest that E. coli 44037:H-, E. coli 44037:H2, E. coli 44037:H3, E. coli 44037:H9, E. coli 44037:H16 and E. coli 44037:H48 strains belong to a serogroup with at least 5 serotypes that have a somatic antigen identical to that of S. boydii 18 with some exhibiting characteristics of ETEC strains with a wide geographic distribution. The presence of this type of strain opens up a new discussion as to whether these strains represent a variant of S. boydii 18 or if they are actually ETEC strains with an S. boydii 18 antigen that acquired the genes for the biosynthesis of this antigen through a horizontal transfer mechanism. These results need to be confirmed with strains from other geographical regions from similar sources.