Microbial Pathogenesis 77 (2014) 73e77
Contents lists available at ScienceDirect
Microbial Pathogenesis journal homepage: www.elsevier.com/locate/micpath
Integrons in uropathogenic Escherichia coli and their relationship with phylogeny and virulence ~a María Eloisa Poey*, Magela Lavin 4225, Montevideo 11.400, Uruguay n Fisiología & Gen Seccio etica Bacterianas, Facultad de Ciencias, Igua
a r t i c l e i n f o
a b s t r a c t
Article history: Received 3 September 2014 Received in revised form 21 October 2014 Accepted 3 November 2014 Available online 4 November 2014
Uropathogenic Escherichia coli (UPEC) comprise a heterogeneous group of strains. In a previous epidemiological survey performed on 230 UPEC isolates, five virulence profiles were described, each one defined by the presence of some virulence determinants and by the absence of others. Phylogenetic groups and antibiotic resistances distributed non-randomly among the isolates with different profiles. Based on these results, the presence of class 1 and 2 integrons was now investigated in these UPEC isolates in order to analyze the distribution of integrons among the phylogenetic groups and virulence profiles. As detected by PCR reactions targeted to the corresponding integrase genes, the class 1 integrons prevailed (22%) followed by those of class 2 (8%). Integrons distributed unevenly among the four main E. coli phylogenetic groups: class 1 integrons predominated in the isolates belonging to group D while class 2 were almost absent in this group. In relation to virulence, integrons frequently appeared in some virulence profiles and were particularly scarce in others. Concerning the class 1 integrons, the most notable findings were that they highly concentrated in isolates presenting one of the virulence profiles (profile V) and were absent in isolates bearing the K1 capsule. The analysis of the Pc promoter variants of the class 1 integrons revealed that all isolates with virulence profile V contained the same Pc version; PcH1. Findings in this work support the idea that, among UPEC strains, integrons would encounter constraints for their installation in some genetic backgrounds while other backgrounds would be propitious for their permanence. © 2014 Elsevier Ltd. All rights reserved.
Keywords: Uropathogenic Escherichia coli Integron Virulence Phylogeny
1. Introduction Uropathogenic Escherichia coli (UPEC) constitute a heterogeneous group of strains which lack a common denominator in their phylogeny and virulence assortment. This fact appears to indicate that UPEC is composed of several pathogenic subgroups. In a previous study, we identified five conserved virulence profiles among UPEC isolates, each one characterized by the presence of some virulence genes and by the absence of others. Genes surveyed were related to capsules, cytotoxins, fimbriae, iron uptake and microcins. The main phylogenetic groups of E. coli (A, B1, B2 and D) as well as antibiotic resistances distributed non-randomly among the isolates, depending on their virulence profile (Table 1). It was then proposed that the virulence assortment of UPEC strains would not result from the simple addition of acquired virulence genes but
abbreviations: UPEC, uropathogenic Escherichia coli. * Corresponding author. E-mail addresses: [email protected] (M.E. Poey), [email protected] ~ a). (M. Lavin http://dx.doi.org/10.1016/j.micpath.2014.11.002 0882-4010/© 2014 Elsevier Ltd. All rights reserved.
instead would be the result of a selective pressure exerted by the level of compatibility among the virulence functions, i.e. that only strains with compatible gene combinations would succeed to thrive [1]. Considering the uneven distribution of antibiotic resistances among the strains with different virulence profiles, we wondered if integrons related to antibiotic resistance would also be submitted to constraints exerted by the genetic background of the UPEC cells, including their phylogeny and virulence profile. Integrons are genetic elements able to integrate and express diverse open reading frames included in gene cassettes. They have received much attention because many of them carry antibiotic resistance genes, being frequently associated with multidrug resistance in Gram-negative bacteria. Integrons contain two parts: a conserved platform and a variable region. The platform invariably harbors three elements: an integrase gene (intI), an attachment site (attI) and a promoter (Pc). The variable region contains one or more gene cassettes that were captured by site-specific recombination and integrated into the attI site. Once integrated, these genes are expressed under the Pc promoter. Integrons are classified into several classes based on the sequence of their intI genes. Those of clinical importance belong to two main classes, 1 and 2. They are
~ a / Microbial Pathogenesis 77 (2014) 73e77 M.E. Poey, M. Lavin
74
Table 1 Virulence profiles, phylogeny and antibiotic resistances in the UPEC isolatesa Variables
Percentage of variables in virulence profilesb Prof. I (n ¼ 21)
Phylogeny A 0 B1 0 D 0 B2 100 Capsule K1 0 Fimbriae/adhesin P 90 S 100 Dr 0 Iron uptake Yersiniabactin 100 Aerobactin 43 Salmochelin 100 Microcins M 100 H47 67 ColV 0 Antibiotic resistancec AMP 24 CEF 9 CXM 0 NAL 9 CIP 0 SXT 5 a b c
Prof. II (n ¼ 9)
Prof. III (n ¼ 20)
Prof. IV (n ¼ 46)
Prof. V (n ¼ 27)
No prof. (n ¼ 107)
Total (n ¼ 230)
56 33 11 0
40 30 20 10
28 0 39 33
0 0 100 0
39 23 19 19
30 15 30 25
0
10
2
41
9
10
11 0 0
10 0 0
4 0 100
100 0 0
7 3 0
25 10 20
33 56 100
50 75 100
93 87 0
100 100 0
42 47 5
65 63 24
0 100 0
0 0 100
0 0 0
0 0 0
4 1 0
11 10 9
0 0 0 0 0 11
75 25 10 30 25 50
91 48 7 26 4 78
70 30 0 7 7 55
43 20 4 18 17 41
55 25 4 18 12 46
Data were extracted from Poey et al., 2012 [1]. UPEC isolates were from pregnant women (N ¼ 112) and from children with urinary tract abnormalities (N ¼ 118). Percentages are calculated on the totals shown above in each column. Markers in each profile are in bold. AMP, ampicillin; CEF, cefalotin; CXM, cefuroxime; NAL, nalidixic acid; CIP, ciprofloxacin; SXT, trimethoprim-sulfamethoxazole.
generally carried by mobile elements such as transposons and conjugative plasmids, and thus are able to spread horizontally to other bacteria [2,3]. Class 1 integrons are the most prevalent among clinical isolates and have been extensively studied. Their platform is divided into two parts separated by the variable region: the 50 conserved segment, which bears the recombination and expression loci, and the 30 conserved segment, which contains two genes, qacED1 and sulI [2,3]. Although the platform is highly conserved, there is a variable stretch in its 50 segment which contains two divergent promoters: the Pc promoter and the integrase promoter PintI1. The Pc promoter has several variants with different strength, which differ in the 35 and 10 hexamers. Three of them have been found most frequently: i) weak (PcW), TGGACAN17TAAGCT; ii) strong (PcS), TTGACAN17TAAACT, and iii) hybrid 1 (PcH1), TGGACAN17TAAACT. In addition, a few class 1 integrons contain a second promoter (P2) 90 nucleotides downstream of Pc, which appears when an insertion of three G residues increases to 17 bp the spacing between two hexamers able to function as 35 and 10 sequences (TTGTTAN14G3TACAGT). P2 is most frequently associated with PcW, both directing the transcription of genes in the variable region; this combination PcW-P2 is several times more active than PcW alone. P2-related sequences are also involved in the expression of the intI1 gene: when there is no P2, a LexA box is found that mediates the SOS regulation of the integrase gene expression [4e7]. Few reports are available concerning the association of integrons with phylogeny in E. coli, and disparities of the bacterial collections studied make comparative analyses difficult. For instance, some authors found that group B2 strains tend to carry less integrons than those belonging to other phylogroups [8,9], others describe the opposite result [10], others found that strains belonging to group D carry significantly more integrons than those from other groups [11,12], and still others did not find significant differences in the distribution of integrons among the phylogenetic
groups [13,14]. Studies of integrons versus virulence in pathogenic E. coli are particularly scarce and point to an association with determinants for siderophore receptors and for S fimbriae [13,15]. In this work, the presence of the class 1 and 2 integrons was surveyed in two collections of UPEC isolates previously studied for their phylogeny, virulence and antibiotic resistance. The aim of the study was to assess the possible association of integrons with phylogenetic and virulence traits of UPEC strains. Particular emphasis was put on the analysis of class 1 integrons, focusing on their platform through the analysis of the Pc promoter variants. 2. Materials and methods 2.1. Bacterial strains Two collections of UPEC isolates coming from a hospital in Montevideo were examined: 112 from pregnant women (PW collection) and 118 from children with urinary tract abnormalities (CH collection). In brief, isolates came from all urine cultures that were positive for E. coli with a high colony count and/or with a clinical record of urinary tract infection, which were collected during a three-month period in 2007 for the PW collection and during 2008 and 2009 for the CH collection. When more than one isolate came from the same patient, only those that were different were kept. PW isolates corresponded to community-acquired urinary tract infections and CH isolates came from ambulatory chronic patients. All the isolates had previously been characterized for their phylogeny, virulence profile and antibiotic resistance (Table 1) [1]. 2.2. Detection of class 1 and 2 integrons, and of the Pc promoter variants of the class 1 integrons Assays were performed by PCR employing as template total genomic DNA extracted with the “Wizard genomic DNA
~ a / Microbial Pathogenesis 77 (2014) 73e77 M.E. Poey, M. Lavin
purification system” (Promega). The presence of class 1 and 2 integrons was analyzed using primers designed to detect the corresponding integrase genes: HS463a and HS464 for intI1 [16], and HS501 and HS502 for intI2 [17]. In the subset of isolates that were positive for intI1, the Pc promoter region was amplified using primers Pre-Pc (50 GATGCGTGGAGACCGAAACCTT30 ) and Post-Pc (50 GCGACTGCCCTGCTGCGTAACAT 30 ). Amplifications were performed in a volume of 30 ml containing 1X buffer, 200 mM of each deoxynucleotide triphosphate, 500 nM of each primer and 1.25 U of U-Taq DNA polymerase (SBS Genetech), plus 20e50 ng of template DNA. The conditions used were: 2 min at 94 C, followed by 30 cycles of 94 C for 30 s, the annealing temperature for 30 s and 72 C for 30 s, and a final extension step at 72 C for 2 min. The annealing temperatures and product sizes were: 55 C to amplify a segment of 473 bp of intI1, 58 C for a segment of 789 bp of intI2, and 62 C for a segment of 362e365 bp containing the Pc promoter. The promoter region in the amplicons was sequenced at Macrogen Inc. (Seoul, Korea) using primer Pre-Pc. 2.3. Statistical analysis The generalized Fisher's exact test was applied supposing the hypothesis of homogeneous categories [18]. P-values equal to or less than 0.05 were considered significant. 3. Results and discussion 3.1. Prevalence of integrons and their relationship with antibiotic resistances The prevalence of class 1 and 2 integrons in the pregnant women (PW) and children (CH) collections was analyzed by PCR reactions targeted to the corresponding intI genes. From a total of 230 isolates, 50 contained a class 1 integron (22%) and 19 a class 2 integron (8%); four of them had both class 1 and 2. In general terms, these results agree with the data in the literature and, in particular, the prevalence of class 1 integrons was almost the same as that previously found in uropathogenic isolates from another hospital in Montevideo [17]. The proportion of integron-bearing isolates was higher in the CH collection (class 1 and 2 in 25% and 10%, respectively) than in the PW one (class 1 and 2 in 18% and 6%, respectively). This was an expected result since children repeatedly underwent antibiotic treatments while the majority of the pregnant women did not. In fact, antibiotic resistance to the tested antibiotics was always more frequent in the CH collection than in the PW one [1]. Considering all the isolates from both collections, resistances appeared more frequently among the integron-bearing isolates than in those without integron. In particular, resistance to
Table 2 Distribution of antibiotic resistances depending on the presence of integrons in the UPEC isolates intIa e No of isolates
Number of resistant isolates (%)b AMP
CEF
CXM
NAL
CIP
SXT
intI þ 65 intI e 165 p-valuesc
56 (86) 70 (42)
Contents lists available at ScienceDirect
Microbial Pathogenesis journal homepage: www.elsevier.com/locate/micpath
Integrons in uropathogenic Escherichia coli and their relationship with phylogeny and virulence ~a María Eloisa Poey*, Magela Lavin 4225, Montevideo 11.400, Uruguay n Fisiología & Gen Seccio etica Bacterianas, Facultad de Ciencias, Igua
a r t i c l e i n f o
a b s t r a c t
Article history: Received 3 September 2014 Received in revised form 21 October 2014 Accepted 3 November 2014 Available online 4 November 2014
Uropathogenic Escherichia coli (UPEC) comprise a heterogeneous group of strains. In a previous epidemiological survey performed on 230 UPEC isolates, five virulence profiles were described, each one defined by the presence of some virulence determinants and by the absence of others. Phylogenetic groups and antibiotic resistances distributed non-randomly among the isolates with different profiles. Based on these results, the presence of class 1 and 2 integrons was now investigated in these UPEC isolates in order to analyze the distribution of integrons among the phylogenetic groups and virulence profiles. As detected by PCR reactions targeted to the corresponding integrase genes, the class 1 integrons prevailed (22%) followed by those of class 2 (8%). Integrons distributed unevenly among the four main E. coli phylogenetic groups: class 1 integrons predominated in the isolates belonging to group D while class 2 were almost absent in this group. In relation to virulence, integrons frequently appeared in some virulence profiles and were particularly scarce in others. Concerning the class 1 integrons, the most notable findings were that they highly concentrated in isolates presenting one of the virulence profiles (profile V) and were absent in isolates bearing the K1 capsule. The analysis of the Pc promoter variants of the class 1 integrons revealed that all isolates with virulence profile V contained the same Pc version; PcH1. Findings in this work support the idea that, among UPEC strains, integrons would encounter constraints for their installation in some genetic backgrounds while other backgrounds would be propitious for their permanence. © 2014 Elsevier Ltd. All rights reserved.
Keywords: Uropathogenic Escherichia coli Integron Virulence Phylogeny
1. Introduction Uropathogenic Escherichia coli (UPEC) constitute a heterogeneous group of strains which lack a common denominator in their phylogeny and virulence assortment. This fact appears to indicate that UPEC is composed of several pathogenic subgroups. In a previous study, we identified five conserved virulence profiles among UPEC isolates, each one characterized by the presence of some virulence genes and by the absence of others. Genes surveyed were related to capsules, cytotoxins, fimbriae, iron uptake and microcins. The main phylogenetic groups of E. coli (A, B1, B2 and D) as well as antibiotic resistances distributed non-randomly among the isolates, depending on their virulence profile (Table 1). It was then proposed that the virulence assortment of UPEC strains would not result from the simple addition of acquired virulence genes but
abbreviations: UPEC, uropathogenic Escherichia coli. * Corresponding author. E-mail addresses: [email protected] (M.E. Poey), [email protected] ~ a). (M. Lavin http://dx.doi.org/10.1016/j.micpath.2014.11.002 0882-4010/© 2014 Elsevier Ltd. All rights reserved.
instead would be the result of a selective pressure exerted by the level of compatibility among the virulence functions, i.e. that only strains with compatible gene combinations would succeed to thrive [1]. Considering the uneven distribution of antibiotic resistances among the strains with different virulence profiles, we wondered if integrons related to antibiotic resistance would also be submitted to constraints exerted by the genetic background of the UPEC cells, including their phylogeny and virulence profile. Integrons are genetic elements able to integrate and express diverse open reading frames included in gene cassettes. They have received much attention because many of them carry antibiotic resistance genes, being frequently associated with multidrug resistance in Gram-negative bacteria. Integrons contain two parts: a conserved platform and a variable region. The platform invariably harbors three elements: an integrase gene (intI), an attachment site (attI) and a promoter (Pc). The variable region contains one or more gene cassettes that were captured by site-specific recombination and integrated into the attI site. Once integrated, these genes are expressed under the Pc promoter. Integrons are classified into several classes based on the sequence of their intI genes. Those of clinical importance belong to two main classes, 1 and 2. They are
~ a / Microbial Pathogenesis 77 (2014) 73e77 M.E. Poey, M. Lavin
74
Table 1 Virulence profiles, phylogeny and antibiotic resistances in the UPEC isolatesa Variables
Percentage of variables in virulence profilesb Prof. I (n ¼ 21)
Phylogeny A 0 B1 0 D 0 B2 100 Capsule K1 0 Fimbriae/adhesin P 90 S 100 Dr 0 Iron uptake Yersiniabactin 100 Aerobactin 43 Salmochelin 100 Microcins M 100 H47 67 ColV 0 Antibiotic resistancec AMP 24 CEF 9 CXM 0 NAL 9 CIP 0 SXT 5 a b c
Prof. II (n ¼ 9)
Prof. III (n ¼ 20)
Prof. IV (n ¼ 46)
Prof. V (n ¼ 27)
No prof. (n ¼ 107)
Total (n ¼ 230)
56 33 11 0
40 30 20 10
28 0 39 33
0 0 100 0
39 23 19 19
30 15 30 25
0
10
2
41
9
10
11 0 0
10 0 0
4 0 100
100 0 0
7 3 0
25 10 20
33 56 100
50 75 100
93 87 0
100 100 0
42 47 5
65 63 24
0 100 0
0 0 100
0 0 0
0 0 0
4 1 0
11 10 9
0 0 0 0 0 11
75 25 10 30 25 50
91 48 7 26 4 78
70 30 0 7 7 55
43 20 4 18 17 41
55 25 4 18 12 46
Data were extracted from Poey et al., 2012 [1]. UPEC isolates were from pregnant women (N ¼ 112) and from children with urinary tract abnormalities (N ¼ 118). Percentages are calculated on the totals shown above in each column. Markers in each profile are in bold. AMP, ampicillin; CEF, cefalotin; CXM, cefuroxime; NAL, nalidixic acid; CIP, ciprofloxacin; SXT, trimethoprim-sulfamethoxazole.
generally carried by mobile elements such as transposons and conjugative plasmids, and thus are able to spread horizontally to other bacteria [2,3]. Class 1 integrons are the most prevalent among clinical isolates and have been extensively studied. Their platform is divided into two parts separated by the variable region: the 50 conserved segment, which bears the recombination and expression loci, and the 30 conserved segment, which contains two genes, qacED1 and sulI [2,3]. Although the platform is highly conserved, there is a variable stretch in its 50 segment which contains two divergent promoters: the Pc promoter and the integrase promoter PintI1. The Pc promoter has several variants with different strength, which differ in the 35 and 10 hexamers. Three of them have been found most frequently: i) weak (PcW), TGGACAN17TAAGCT; ii) strong (PcS), TTGACAN17TAAACT, and iii) hybrid 1 (PcH1), TGGACAN17TAAACT. In addition, a few class 1 integrons contain a second promoter (P2) 90 nucleotides downstream of Pc, which appears when an insertion of three G residues increases to 17 bp the spacing between two hexamers able to function as 35 and 10 sequences (TTGTTAN14G3TACAGT). P2 is most frequently associated with PcW, both directing the transcription of genes in the variable region; this combination PcW-P2 is several times more active than PcW alone. P2-related sequences are also involved in the expression of the intI1 gene: when there is no P2, a LexA box is found that mediates the SOS regulation of the integrase gene expression [4e7]. Few reports are available concerning the association of integrons with phylogeny in E. coli, and disparities of the bacterial collections studied make comparative analyses difficult. For instance, some authors found that group B2 strains tend to carry less integrons than those belonging to other phylogroups [8,9], others describe the opposite result [10], others found that strains belonging to group D carry significantly more integrons than those from other groups [11,12], and still others did not find significant differences in the distribution of integrons among the phylogenetic
groups [13,14]. Studies of integrons versus virulence in pathogenic E. coli are particularly scarce and point to an association with determinants for siderophore receptors and for S fimbriae [13,15]. In this work, the presence of the class 1 and 2 integrons was surveyed in two collections of UPEC isolates previously studied for their phylogeny, virulence and antibiotic resistance. The aim of the study was to assess the possible association of integrons with phylogenetic and virulence traits of UPEC strains. Particular emphasis was put on the analysis of class 1 integrons, focusing on their platform through the analysis of the Pc promoter variants. 2. Materials and methods 2.1. Bacterial strains Two collections of UPEC isolates coming from a hospital in Montevideo were examined: 112 from pregnant women (PW collection) and 118 from children with urinary tract abnormalities (CH collection). In brief, isolates came from all urine cultures that were positive for E. coli with a high colony count and/or with a clinical record of urinary tract infection, which were collected during a three-month period in 2007 for the PW collection and during 2008 and 2009 for the CH collection. When more than one isolate came from the same patient, only those that were different were kept. PW isolates corresponded to community-acquired urinary tract infections and CH isolates came from ambulatory chronic patients. All the isolates had previously been characterized for their phylogeny, virulence profile and antibiotic resistance (Table 1) [1]. 2.2. Detection of class 1 and 2 integrons, and of the Pc promoter variants of the class 1 integrons Assays were performed by PCR employing as template total genomic DNA extracted with the “Wizard genomic DNA
~ a / Microbial Pathogenesis 77 (2014) 73e77 M.E. Poey, M. Lavin
purification system” (Promega). The presence of class 1 and 2 integrons was analyzed using primers designed to detect the corresponding integrase genes: HS463a and HS464 for intI1 [16], and HS501 and HS502 for intI2 [17]. In the subset of isolates that were positive for intI1, the Pc promoter region was amplified using primers Pre-Pc (50 GATGCGTGGAGACCGAAACCTT30 ) and Post-Pc (50 GCGACTGCCCTGCTGCGTAACAT 30 ). Amplifications were performed in a volume of 30 ml containing 1X buffer, 200 mM of each deoxynucleotide triphosphate, 500 nM of each primer and 1.25 U of U-Taq DNA polymerase (SBS Genetech), plus 20e50 ng of template DNA. The conditions used were: 2 min at 94 C, followed by 30 cycles of 94 C for 30 s, the annealing temperature for 30 s and 72 C for 30 s, and a final extension step at 72 C for 2 min. The annealing temperatures and product sizes were: 55 C to amplify a segment of 473 bp of intI1, 58 C for a segment of 789 bp of intI2, and 62 C for a segment of 362e365 bp containing the Pc promoter. The promoter region in the amplicons was sequenced at Macrogen Inc. (Seoul, Korea) using primer Pre-Pc. 2.3. Statistical analysis The generalized Fisher's exact test was applied supposing the hypothesis of homogeneous categories [18]. P-values equal to or less than 0.05 were considered significant. 3. Results and discussion 3.1. Prevalence of integrons and their relationship with antibiotic resistances The prevalence of class 1 and 2 integrons in the pregnant women (PW) and children (CH) collections was analyzed by PCR reactions targeted to the corresponding intI genes. From a total of 230 isolates, 50 contained a class 1 integron (22%) and 19 a class 2 integron (8%); four of them had both class 1 and 2. In general terms, these results agree with the data in the literature and, in particular, the prevalence of class 1 integrons was almost the same as that previously found in uropathogenic isolates from another hospital in Montevideo [17]. The proportion of integron-bearing isolates was higher in the CH collection (class 1 and 2 in 25% and 10%, respectively) than in the PW one (class 1 and 2 in 18% and 6%, respectively). This was an expected result since children repeatedly underwent antibiotic treatments while the majority of the pregnant women did not. In fact, antibiotic resistance to the tested antibiotics was always more frequent in the CH collection than in the PW one [1]. Considering all the isolates from both collections, resistances appeared more frequently among the integron-bearing isolates than in those without integron. In particular, resistance to
Table 2 Distribution of antibiotic resistances depending on the presence of integrons in the UPEC isolates intIa e No of isolates
Number of resistant isolates (%)b AMP
CEF
CXM
NAL
CIP
SXT
intI þ 65 intI e 165 p-valuesc
56 (86) 70 (42)
