Occurrence of Antibiotic Resistant Uropathogenic Escherichia coli Clonal Group A 1

in Wastewater Effluents 2

3

Laura A. Boczek1, Eugene W. Rice

1, Brian Johnston

2, and James R. Johnson

2* 4

1U.S. Environmental Protection Agency, Cincinnati OH 5

2*Medical Service, Veterans Affairs Medical Center and Department of Medicine, 6

University of Minnesota, Minneapolis MN 7

8

9

*Corresponding author address: 10

11

James R. Johnson, M.D. 12

Infectious Diseases (111F) 13

Minneapolis VA Medical Center 14

1 Veterans Drive 15

Minneapolis, MN 55417 (USA) 16

phone: 612-467-4185 17

fax: 612-727-5995 18

email: johns007@umn.edu 19

ACCE

PTED

Copyright © 2007, American Society for Microbiology and/or the Listed Authors/Institutions. All Rights Reserved.Appl. Environ. Microbiol. doi:10.1128/AEM.02225-06 AEM Accepts, published online ahead of print on 4 May 2007

on March 30, 2021 by guest

http://aem.asm

.org/D

ownloaded from

http://aem.asm.org/

Abstract 20

21

Isolates of Escherichia coli belonging to clonal group A (CGA), a recently 22

described disseminated cause of drug-resistant urinary tract infections in humans, were 23

present in four of seven sewage effluents collected from geographically dispersed areas 24

of the United States. All 15 CGA isolates (1% of the 1,484 isolates analyzed) exhibited 25

resistance to trimethoprim-sulfamethoxazole, accounting for 19.5 % of the 77 TMP-26

SMZ-resistant isolates. Antimicrobial resistance patterns, virulence traits, O:H serotypes, 27

and phylogenetic groupings were compared for CGA and selected non-CGA isolates. 28

The CGA isolates exhibited a wider diversity of resistance profiles and somatic antigens 29

than found in most previous characterizations of this clonal group. This is the first report 30

of recovery from outside a human host of CGA E. coli isolates with virulence factor and 31

antibiotic resistance profiles typical of human-source CGA isolates. The occurrence of 32

"human-type" CGA in wastewater effluents demonstrates a potential mode for the 33

dissemination of this clonal group in the environment, with possible secondary 34

transmission to new human or animal hosts. 35 ACCE

PTED

on March 30, 2021 by guest

http://aem.asm

.org/D

ownloaded from

http://aem.asm.org/

Introduction 36

37

Resistance to commonly prescribed antimicrobial agents is a matter of increasing 38

concern. Along with respiratory infections, urinary tract infections (UTIs) are the most 39

common bacterial infections in the United States requiring antimicrobial therapy. 40

Escherichia coli is the most frequently isolated etiological agent of UTIs and 41

trimethoprim-sulfamethoxazole (TMP-SMZ) is one of the primary antibiotics which is 42

empirically prescribed for the treatment of community-acquired UTIs (6). In the United 43

States there has been a notable increase in the isolation of uropathogenic E. coli resistant 44

to TMP-SMZ (6). This finding is of particular interest since resistance to this drug is 45

generally associated with resistance to additional antibiotics (15). Recent 46

epidemiological studies have reported on the widespread emergence of a single clonal 47

group, provisionally designated clonal group A (CGA), among TMP-SMZ-resistant 48

strains of uropathogenic E. coli (1,4,9,10,11). CGA has been reported to account for up 49

to 50% of TMP-SMZ resistant E. coli isolates from women with acute uncomplicated 50

cystitis and pyelonephritis (9). Recent human CGA isolates have typically exhibited a 51

number of traits that set them apart from other uropathogenic or drug-resistant E. coli, 52

including their characteristic virulence factor profile, several distinctive O antigens, the 53

H18 flagellar antigen, and multidrug resistance, including to ampicillin, chloramphenicol, 54

streptomycin, sulfonamides, tetracycline, and trimethoprim (9). 55

56

The widespread occurrence of CGA E. coli, and the occurrence in one community 57

of a seeming point-source outbreak of UTIs due to the same pulsotype of CGA (11), has 58

ACCE

PTED

on March 30, 2021 by guest

http://aem.asm

.org/D

ownloaded from

http://aem.asm.org/

raised questions regarding the mode of transmission of these organisms. Recent travel to 59

areas with a high prevalence of TMP-SMZ resistant E. coli has been previously cited as a 60

risk factor for acquiring TMP-SMZ-resistant E. coli strains (12). This finding suggests 61

that exposure to contaminated food or water may represent a means for the dissemination 62

of these pathogenic E. coli (10,11,14,17). The current study was designed to assess the 63

prevalence and characteristics of TMP-SMZ resistant E. coli and CGA in domestic 64

sewage samples collected from various locations throughout the United States. 65

66

MATERIALS AND METHODS 67

68

Samples and primary cultures. Two sewage effluent samples (primary and 69

secondary), prior to chlorination, were collected from each of seven geographically 70

dispersed wastewater treatment plants in the United States: San Jose, California (CA), 71

Tallahassee, Florida (FL), Lewiston, Maine (ME), Cincinnati, Ohio (OH), Virginia 72

Beach, Virginia (VA), Seattle, Washington (WA), and Milwaukee, Wisconsin (WI). 73

Samples were shipped on ice and analyzed within 24 hours of collection. The membrane 74

filtration procedure using mFC agar (BD Bioscience, Sparks, MD) and nutrient agar 75

containing 4-methylumbelliferyl beta-D-glucuronide (BD Bioscience, Sparks, MD) (3) 76

was used to detect E. coli. Approximately 200 E. coli colonies from each location were 77

picked at random and streaked for purity on heart infusion agar (BD Bioscience, Sparks, 78

MD). 79

80

ACCE

PTED

on March 30, 2021 by guest

http://aem.asm

.org/D

ownloaded from

http://aem.asm.org/

Susceptibility testing. Isolates were sub-cultured to Mueller-Hinton agar (BD 81

Bioscience, Sparks, MD) and screened for susceptibility to TMP-SMZ (BD Bioscience, 82

Sparks, MD) by the disk diffusion method (13). E. coli strain 25922 (American Type 83

Culture Collection, ATCC) was used as a reference control strain. All TMP-SMZ 84

resistant isolates (zone diameter of ≤ 10mm) were saved for further study, as was a group 85

of TMP-SMZ susceptible E. coli isolates (which were selected as described below as 86

controls for geographically matched comparisons). All isolates were confirmed as E. coli 87

using the API 20E test system (bioMerieux, Marcy’Etoile, France). Isolates were also 88

tested for susceptibility to ampicillin (AM), chloramphenicol (CH), streptomycin (ST), 89

and tetracycline (TE) by using E-test strips (AB Biodisk, Solna, Sweden), as directed by 90

the manufacturer, to determine minimum inhibitory concentration (MIC) values. Isolates 91

were classified as susceptible (s), intermediate (i), or resistant (r), based upon MIC 92

criteria as specified by the Clinical Laboratory and Standards Insititue (CLSI) (13) or 93

based on published precedent (St) (9). The following criteria were used for each 94

antibiotic (MIC µg/mL): AM (s = # 8, i = 16, r = ≥ 32), CH (s = #8, i = 16, s = ≥32), ST 95

(r = ≥ 8), TE (s = # 4, i = 8, r = ≥ 16), TMP-SMZ (s = # 2, r = ≥ 4). 96

97

Molecular analysis. All TMP-SMZ-resistant isolates were O:H serotyped by the 98

E. coli Reference Center at the Pennsylvania State University (University Park, PA) and 99

were phylotyped (for phylogenetic groups A, B1, B2, and D) using a multiplex PCR 100

procedure (2). All TMP-SMZ-resistant isolates and a group of 77 geographically matched 101

TMP-SMZ-susceptible isolates were tested for CGA status by using a gene-specific PCR 102

procedure that detects a distinctive single nucleotide polymorphism (SNP) within fumC 103

ACCE

PTED

on March 30, 2021 by guest

http://aem.asm

.org/D

ownloaded from

http://aem.asm.org/

(8). All CGA isolates and a group of geographically matched non-CGA, TMP-SMZ-104

resistant control isolates (n = 22, chosen 2:1 with respect to the CGA isolates with the 105

exception of CA, where there were only 18 TMP-SMZ-resistant non-CGA isolates), were 106

tested for 40 virulence factors of extraintestinal pathogenic E. coli by PCR as previously 107

described (7). CGA isolates were further assessed by random amplified polymorphic 108

DNA (RAPD) analysis using arbitrary decamer primer 1254 5’-ccgcagccaa-3’ (7) in 109

comparison with CGA and non-CGA reference strains, and by pulsed-field gel 110

electrophoresis (PFGE) analysis of XbaI-digested total DNA (11). 111

112

Statistical analysis. Comparisons of proportions were tested using Fisher’s exact 113

test (two-tailed). The criterion for statistical significance was P < 0.05. The BioNumerics 114

software (Applied Maths), which has band tolerance of 1.15% was used to assess the 115

digital images of the PFGE profiles. 116

117

118

RESULTS 119

120

Distribution of TMP-SMZ resistance. A total of 1,484 arbitrarily selected E. 121

coli sewage isolates (primary effluents n = 732, secondary effluents n = 754) from seven 122

locations were characterized, representing approximately 200 isolates from each location 123

(Table 1). Seventy-seven (5%) of the isolates were resistant to TMP-SMZ. The 124

prevalence of TMP-SMZ resistance varied among the various locales, with the highest 125

values being observed in CA (13%) and WI (7%), followed by OH, WA (5%), VA (2%), 126

ACCE

PTED

on March 30, 2021 by guest

http://aem.asm

.org/D

ownloaded from

http://aem.asm.org/

and ME (

[40%)], P < 0.0001). Nonetheless, while 33% of the environmental CGA isolates had 150

serotypes typical of those reported for human clinical CGA isolates (i.e. O17,77:H18, 151

O11:H18), 67% had serotypes not previously described among human clinical or fecal 152

CGA isolates. 153

154

Phylogenetic distribution. As expected, all 15 CGA isolates belonged to 155

phylogenetic group D (9). In contrast the non-CGA, TMP-SMZ-resistant isolates (n = 156

62) were approximately evenly distributed over the four phylogenetic groups: 18 (29%) 157

group A, 15 (24%) group B1, 15 (24%) group B2, and 14 (23%) group D. 158

159

Antimicrobial resistance profiles. The CGA isolates and the TMP-SMZ-160

resistant non-CGA isolates differed according to their composite resistance phenotypes. 161

Among the 15 CGA isolates, 5 resistance profiles were encountered. One isolate 162

exhibited TMP-SMZ resistance only, 2 (13%) were resistant to 2 additional 163

antimicrobials each (AM, ST and AM, TE), 11 (73%) were resistant to 3 additional 164

antimicrobials each (AM, ST, TE), and 1 exhibited resistance to all 4 additional 165

antimicrobials tested (AM, CH, ST, TE). In contrast, the 62 non-CGA, TMP-SMZ-166

resistant isolates exhibited 11 different multiple antimicrobial resistance patterns. In 167

order of descending prevalence, 25 (40%) were resistant to two additional antimicrobials, 168

20 (32%) were resistant to three additional antimicrobials, 9 (14%) were resistant to four 169

additional antimicrobials, 7 (11%) were resistant to one additional antimicrobial, and 1 170

(2%) was resistant only to TMP-SMZ. The four most commonly occurring associated 171

ACCE

PTED

on March 30, 2021 by guest

http://aem.asm

.org/D

ownloaded from

http://aem.asm.org/

resistance profiles among the non-CGA, TMP-SMZ-resistant isolates were: AM, ST, TE 172

(n = 20); AM, TE (n = 13); AM, ST (n = 8), and AM, CH, ST, TE (n = 8). 173

174

Virulence profiles. The 15 CGA isolates were compared with 22 geographically 175

matched non-CGA TMP-SMZ-resistant isolates for virulence traits, including alleles of 176

papA allele (P fimbriae structural subunit) (Table 2). The non-CGA isolates were 177

selected to provide a distribution among the four phylogenetic groups: A (n = 7), B1 (n = 178

4), B2 (n = 6), and D (n = 5). The CGA isolates were divided between the F10 (n = 9) 179

and F16 (n = 6) papA alleles, whereas none of the non-CGA isolates exhibited either of 180

these two papA alleles (P < 0.001 and P = 0.002, respectively). In contrast, 3 of the non-181

CGA isolates, but none of the CGA isolates, exhibited the F12 or F14 papA allele. 182

Additionally, all CGA isolates, but only 3 (14%) of the non-CGA isolates, exhibited 183

papEF (P fimbriae tip pilins) (P < 0.0001). The CGA and non-CGA isolates also differed 184

significantly (p < 0.0001) according to the prevalence of five non-pap virulence traits, 185

including iha (adhesin/siderophore receptor), sat (secreted autotransporter toxin), iutA 186

(aerobactin receptor), kpsM II (group II capsule), and ompT (outer membrane protease T). 187

afa/dra (Dr-antigen binding adhesins) was the only virulence trait which demonstrated a 188

significant negative association with CGA. 189

190

Interestingly, although all CGA isolates exhibited papEF (P fimbriae tip pilus), 191

only six (40%) also exhibited papA, papC (P fimbriae assembly), and papG allele II (P 192

fimbriae adhesin variant II). Moreoever, all CGA isolates exhibiting the F10 papA allele 193

(n = 9) had papEF as their only other pap element, whereas CGA isolates exhibiting the 194

ACCE

PTED

on March 30, 2021 by guest

http://aem.asm

.org/D

ownloaded from

http://aem.asm.org/

F16 papA alleles (n = 6) contained a complete copy of the pap operon, including papA, 195

papC, papEF, and papG (allele II). 196

197

RAPD and PFGE analysis. The RAPD genomic profiles of the 15 CGA isolates 198

were highly homogeneous and matched those of reference CGA isolates UMN026 and 199

SEQ102, whereas they differed substantially from those of non-CGA controls (Figure 1). 200

Interestingly, CGA isolates containing the F10 papA allele typically exhibited a one-band 201

RAPD profile difference in comparison with the CGA isolates containing the F16 papA 202

allele, consistent with these two groups possibly representing different genomic subsets 203

within CGA (Figure 1). PFGE analysis (Fig. 2) demonstrated that although the profiles 204

of the CGA isolates exhibited an obvious overall similarity, each isolate (lanes 2 - 9 and 205

11 - 17) had a distinct banding pattern, indicating that despite their considerable 206

similarity (and, in several instances, common sample of origin) no two isolates were 207

replicates of the same clone. The presence/absence of a single high molecular weight 208

band (Figure 2) corresponded closely with the split between the F10 and F16 papA alleles 209

among CGA isolates. 210

211

DISCUSSION 212

213

Among E. coli isolates from wastewater effluents from seven U.S. geographic 214

regions, the prevalence of CGA and of TMP-SMZ-resistant E. coli varied considerably 215

by region; however, no significant difference was seen between the isolates from primary 216

and secondary effluents. The highest percentage of CGA isolates was from California, 217

ACCE

PTED

on March 30, 2021 by guest

http://aem.asm

.org/D

ownloaded from

http://aem.asm.org/

which had not only the highest overall prevalence of TMP-SMZ resistance (39%), but 218

also the highest proportion of TMP-SMZ-resistant isolates accounted for by CGA (80%). 219

It is of interest that California is a region that has previously reported a high proportion of 220

CGA isolates (51%) among TMP-SMZ resistant clinical isolates (11). The explanation 221

for the observed variation in prevalence of CGA and of TMP-SMZ-resistant E. coli in 222

wastewater effluents is unknow. However, possible explanations include differences in 223

processes between wastewater treatment plants, in population densities at each treatment 224

plant, in the types of sewages found at each treatment plant (industrial, agricultural, or 225

urban), and the prevalence of CGA within the host population(s) served by each plant. 226

Variation in the prevalence of CGA among clinical isolates from different regions has 227

been documented (7, 9); it is possible that regional variation exists also among 228

environmental samples. 229

230

The increasing occurrence of TMP-SMZ-resistant uropathogenic E. coli 231

emphasizes the need to determine resistance prevalence levels within a community. 232

Determining the prevalence of antimicrobial resistance in a given locale may require 233

information beyond that obtained in clinical laboratory studies. Data on resistance rates 234

collected only from patients with UTIs may not reflect the true prevalence of resistance in 235

the local community (6,19), since specimens are often submitted for culture and antibiotic 236

susceptibility testing only after initial antimicrobial therapy has failed or because the 237

patient has had recurrent episodes of UTIs, or is otherwise considered at risk for having a 238

resistant organism. The intestinal tract serves as the primary source for uropathogenic E. 239

coli (6). E. coli found in domestic sewage comes predominantly from human fecal 240

ACCE

PTED

on March 30, 2021 by guest

http://aem.asm

.org/D

ownloaded from

http://aem.asm.org/

material (18). Thus, sewage isolates may serve as representatives of the strains of E. coli 241

which would be present within the human population in a given locale. Surveillance of 242

antibiotic resistance in coliforms from sewage has been previously suggested as a means 243

of monitoring changes in antibiotic resistance patterns in the general population (18). 244

Seneviratne and Woods (1976) specifically proposed that such a surveillance program 245

would be helpful in providing therapeutic guidance for the treatment of urinary 246

pathogens. The results of the present study suggest that such monitoring programs may 247

also provide useful information on the occurrence and dissemination of specific clonal 248

groups within a given community. 249

250

Fifteen (19.5%) of TMP-SMZ-resistant isolates in this study belonged to CGA. 251

This value is very similar to that reported in a recent national survey of clinical isolates, 252

where CGA accounted for 15% of TMP-SMZ-resistant isolates from diverse locales 253

across the United States (9). Among the sewage-source isolates, CGA E. coli were 254

distinct from geographically matched non-CGA TMP-SMZ-resistant E. coli according to 255

a broad range of bacterial characteristics, as previously reported among human clinical 256

isolates of CGA versus non-CGA E. coli (11). 257

258

Notwithstanding the overall similarity of the present sewage-source CGA isolates 259

to previously described human CGA isolates, some differences were apparent. The 260

environmental isolates exhibited a broader range of serogroups (O15, O44, O86) beyond 261

the O11, O17, O73, and O77 serogroups typically associated with clinical CGA isolates, 262

a lower prevalence of chloramphenicol resistance (7%, vs. > 25% among clinical CGA 263

ACCE

PTED

on March 30, 2021 by guest

http://aem.asm

.org/D

ownloaded from

http://aem.asm.org/

isolates), and the unique occurrence of the F10 papA allele (60% vs. 0%) (9). It is of 264

interest that the F10-positive isolates lacked most of the rest of the pap operon, so 265

presumably could not express P fimbriae, which should make them less able to colonize 266

or infect humans. Whether the F10-positive subgroup represents an environmentally 267

adapted variant of CGA, or perhaps an ancestral precursor, remains to be established. The 268

detection of this interesting subgroup in multiple locales suggests it is not an isolated 269

entity, but may be broadly prevalent in sewage effluents across the US. These findings 270

raise interesting questions regarding the ecology of these organisms outside of the human 271

host. 272

273

In summary, the sewage CGA E. coli isolates exhibited a wider diversity of 274

resistance profiles and somatic antigens than found in most previous characterizations of 275

this clonal group (France et al. 2005). However, some of these isolates had virulence 276

factor and antimicrobial resistance profiles typical of human-source CGA isolates. This is 277

the first report of the recovery of CGA E. coli with typical “human” pattern virulence and 278

resistance profiles from outside the human host. The presence of CGA E. coli in sewage 279

effluents may provide a means for dissemination of this clonal group in the environment. 280

Sewage could serve as a vehicle for entering human and non-human hosts by direct 281

contact or through contamination of drinking water supplies. 282

283

Any opinions expressed in this paper are those of the author(s) and do not, 284

necessarily, reflect the official positions and policies of the USEPA. Any mention of 285

products or trade names does not constitute recommendation for use by the USEPA. 286

ACCE

PTED

on March 30, 2021 by guest

http://aem.asm

.org/D

ownloaded from

http://aem.asm.org/

287

ACCE

PTED

on March 30, 2021 by guest

http://aem.asm

.org/D

ownloaded from

http://aem.asm.org/

Acknowledgments 288

289

This material is based on work supported by the Office of Research and Development, 290

Medical Research Service, Department of Veterans Affairs (JRJ). Dave Prentiss 291

(Minneapolis VA Medical Center) prepared the figures. 292

ACCE

PTED

on March 30, 2021 by guest

http://aem.asm

.org/D

ownloaded from

http://aem.asm.org/

REFERENCES 293

294

1. Burman, W.J., P.E. Breese, B.E. Murray, K.V. Singh, H.A. Batal, T.D. 295

MacKenzie, J.W. Ogle, M.L. Wilson, R.R. Reves, P.S. Mehler. 2003. Conventional 296

and molecular epidemiology of trimethoprim-sulfamethoxazole resistance among urinary 297

Escherichia coli isolates. Am. J. Med. 115:358-64. 298

299

2. Clermont, O., S. Bonacorsi, E. Bingen. 2000. Rapid and simple determination of the 300

301

Escherichia coli phylogenetic group. Appl. Environ. Microbiol. 66:4555-8. 302

303

3. Clesceri, L.S., A.E. Greenberg, A.D. Eaton , eds. 1998. Standard Methods for the 304

Examination of Water and Wastewater. 21st ed. Washington, DC: American Public 305

Health Assoc., 9-67-68. 306

307

4. France, A.M., K.M. Kugeler, A. Freeman, C.A. Zalewski, M. Blahna, L. Zhang, 308

C.F. Marrs, B. Foxman. 2005. Clonal group and the spread of resistance to 309

trimethoprim-sulfamethoxazole in uropathogenic Escherichia coli. Clin. Inefect. Dis. 310

40:1101-7. 311

312

5. Gupta, K., T.M. Hooton, W.E. Stamm. 2001. Increasing antimicrobial resistance 313

and the management of uncomplicated community-acquired urinary tract infections. Ann. 314

Intern. Med. 135:41-50. 315

316

317

ACCE

PTED

on March 30, 2021 by guest

http://aem.asm

.org/D

ownloaded from

http://aem.asm.org/

6. Karlowsky, J.A., L.J. Kelly, C. Thornsberry, M.E. Jones, D.F. Sahm. 2002. 318

Trends in antimicrobial resistance among urinary tract infection isolates of Escherichia 319

coli from female outpatients in the United States. Antimicrob. Agents. Chemother. 320

46:2540-5. 321

322

7. Johnson, J.R., A.R. Manges,T.T. O’Bryan, L.W. Riley. 2002. A disseminated 323

multidrug resistant clonal group of extraintestinal pathogenic Escherichia coli as a cause 324

of pyelonephritis. Lancet 359:2249-51. 325

326

8. Johnson, J.R., K. Owens, A.R. Manges, L.W. Riley. 2004. Rapid specific detection 327

of Escherichia coli clonal group A by gene-specific PCR. J. Clin. Microbiol. 42:2618-22. 328

329

9. Johnson, J.R., A.C. Murray, M.A. Kuskowski, S. Schubert, M.F. Prere, B. 330

Picard, R. Colodner, R. Raz, (A) and the Trans-Global Initiative for Antimicrobial 331

Resistance Analysis (TIARA) Investigators. 2005. A Distribution and characteristics of 332

Escherichia coli clonal group A. Emerg. Infect. Dis. 11:141-5. 333

334

10. Johnson, J.R., M.A. Kuskowski, K. Smith, T.T. O’Bryan, S. B. Tatini (B). 2005. 335

Antimicrobial-resistant and extraintestinal pathogenic Escherichia coli in retail foods. J. 336

Infect. Dis. 191:1040-49. 337

338

ACCE

PTED

on March 30, 2021 by guest

http://aem.asm

.org/D

ownloaded from

http://aem.asm.org/

11. Manges, A.R., J.R. Johnson, B. Foxman, T.T. O’Bryan, K.E. Fullerton, L.W. 339

Riley. 2001.Widespread distribution of urinary tract infections caused by a multidrug-340

resistant Escherichia coli clonal group. N. Engl. J. Med. 345:1007-13. 341

342

12. Murray, B.E., J.J Mathewson, H.L. Dupont, C.D. Ericsson, R.R. Reeves. 1990. 343

Emergence of resistant fecal Escherichia coli in travelers not taking prophylactic 344

antimicrobial agents. Antimicrob. Agents. Chemother. 34:515-8. 345

346

13. National Committee for Clinical Laboratory Standards. 2000. Supplemental 347

tables: disk diffusion. NCCLS document M100-S10. Wayne, PA. 14-18. 348

349

14. Ramchandani, M., A.R. Manges, C. DebRoy, S.P. Smith, J.R. Johnson, L.W. 350

Riley. 2005. Possible animal origin of human-associated , multi-drug resistant, 351

uropathogenic Escherichia coli Clin. Infect. Dis. 40:251-7. 352

353

15. Sahm, D.F., C. Thornsberry, D.C. Mayfield, M.E. Jones, J.A. Karlowsky. 2001. 354

Multidrug-resistant urinary tract isolates of Escherichia coli: prevalence and 355

patient demographics in the United States in 2000. Antimicrob. Agents. Chemother. 356

45:1402-6. 357

358

16. Seneviratne, E.M.E., G. Woods. 1976.Sensitivity patterns of urinary pathogens in 359

the southland community and a study of sewage flora. N. Z. Med. J. 84:15-18. 360

361

ACCE

PTED

on March 30, 2021 by guest

http://aem.asm

.org/D

ownloaded from

http://aem.asm.org/

17. Stamm, W.E., 2001. An epidemic of urinary tract infections? N. Eng. J. Med. 362

345:1055-7. 363

364

18. Sturtevant Jr, A.B., T.W. Feary. 1969. Incidence of infectious drug resistance 365

among lactose-fermenting bacteria isolated from raw and treated sewage. Appl. 366

Microbiol. 18:918-24. 367

368

19. Ti, T.Y., G. Kumarasinghe, M.B. Taylor, S.L. Tan, A. Ee, C. Chua, A. Low. 369

2003. What is true community-acquired urinary tract infection? Comparison of pathogens 370

identified in urine from routine outpatient specimens and from community clinics 371

in a prospective study. Eur. J. Clin. Microbiol. Infect. Dis. 22:242-5. 372

373

ACCE

PTED

on March 30, 2021 by guest

http://aem.asm

.org/D

ownloaded from

http://aem.asm.org/

Table 1. Occurrence of TMP-SMZ resistance and Escherichia coli clonal group A 374

(CGA) among 1,484 E. coli sewage isolates from seven regions. 375

Region Total E. coli

isolated, no.

aTMP-SMZ-resistant

E. coli, no. (% of total)

CGA E. coli, no.

(% of TMP-SMZ-

resistant)a

San Jose, CA 226 30 (13) 12 (40)

Tallahassee, FL 210 5 (2) 0 (

Table 2. Virulence factor profiles among clonal group A (CGA) and non-CGA 384

Escherichia coli sewage isolates. 385

386

Prevalence of associated trait, no (%)

Trait+

CGA

(n = 15)

non-CGA++

(n = 22)

p +++

value

F10 papA allele

9 (60) 0 (0) < 0.0001

F16 papA allele 6 (40) 0 (0) 0.0022

papEF 15 (100) 3 (14) < 0.0001

papG allele II 6 (40) 1 (4) 0.0113

afa/dra 0 (0) 14 (64) (< 0.0001)

iha 14 (93) 3 (14) < 0.0001

sat 14 (93) 4 (18) < 0.0001

iutA 14 (93) 7 (32) 0.0002

kpsM II 15 (100) 10 (45) 0.0004

ompT 15 (100) 8 (36) < 0.0001

+ Only those traits for which prevalence differences between CGA vs. non-CGA isolates 387

were statistically significant (Fisher’s exact test, p < 0.05) are shown. 388

Trait definitions: papA (P fimbriae structural subunit) 389

papEF (P fimbriae tip pilus) 390

papG allele II (P fimbriae adhesion variant II) 391

afa/dra (Dr-antigen binding adhesins) 392

iha (adhesin/siderophore) 393

ACCE

PTED

on March 30, 2021 by guest

http://aem.asm

.org/D

ownloaded from

http://aem.asm.org/

sat (secreted autotransporter toxin) 394

iutA (aerobactin receptor) 395

kpsM II (group II capsule) 396

ompT (outer membrane protease T) 397

398

Other traits present in CGA, but not significantly different in overall prevalence 399

compared with non-CGA isolates, included: papC (P fimbriae assembly), fimH (type 1 400

fimbriae adhesin), fyuA (yersiniabactin receptor), and traT (serum resistance associated). 401

Traits absent from CGA isolates, but detected among non-CGA isolate, included: F12 402

papA, papG allele III, iroN, K1 kpsM II variant, usp, iss, and malX; however these 403

differences (CGA versus non-CGA) were not statistically significantly by the Fisher’s 404

exact test. 405

++ Non-CGA TMP-SMZ-resistant control isolates were chosen in a 2:1 ratio to the CGA 406

isolates within each locale with the exception of San Jose, CA, where only 18 non-CGA 407

TMP-SMZ resistant isolates were recovered, all of which were used as controls. 408

+++ p value in parenthesis is for negative associations with CGA 409 AC

CEPT

ED

on March 30, 2021 by guest

http://aem.asm

.org/D

ownloaded from

http://aem.asm.org/

410

Figure 1. Random amplified DNA (RAPD) profiles of selected Escherichia coli 411

isolates. Lane numbers are shown below gel image. Lanes 4 and 17, 250-bp marker (M). 412

Lanes 1-3 and 5-16: clonal group A (CGA)-positive sewage isolates. Lanes 18-19: CGA-413

positive controls (human cystitis isolates SEQ102 and UMN026, respectively). Lanes 20-414

24: CGA-negative controls (lanes 20-21, non-CGA sewage isolates [from phylogenetic 415

groups B2 and D, respectively]; lane 22, strain CFT073 [from group B2]; lanes 23-24, 416

human source non-CGA isolates [from groups D and B2, respectively]). Bullets above 417

lanes indicate CGA isolates that exhibit the consensus CGA-associated virulence profile, 418

including the F16 papA (P fimbriae structural subunit) allele, and characteristic CGA-419

associated RAPD profile. CGA isolates in lanes without a bullet exhibit an atypical 420

virulence profile that includes the F10 papA allele, plus (excepting for strain 492: lane 5) 421

an atypical RAPD profile that includes an extra ~2000 bp band (vertical arrows, or "*" 422

for strain 492, lane 5). The profiles of all the CGA isolates are indistinguishable (within 423

the reproducibility limits of RAPD analysis) excepting for the ~2000 bp band, and 424

collectively are distinct from the profiles of the non-CGA isolates, each of which is 425

unique. 426

ACCE

PTED

on March 30, 2021 by guest

http://aem.asm

.org/D

ownloaded from

http://aem.asm.org/

427

428

429

Figure 2. XbaI pulsed-field gel electrophoresis (PFGE) profiles of selected 430

Escherichia coli isolates. Lane numbers are shown below gel image. Lanes 1, 10, and 431

18: marker (M) lanes, with E. coli O157:H7 strain G5244. Lanes 2-9 and 11-17: CGA-432

positive sewage isolates. Bullets above lanes indicate CGA isolates that exhibit the 433

consensus CGA-associated virulence profile, including the F16 papA (P fimbriae 434

structural subunit) allele. CGA isolates in lanes without a bullet exhibit an atypical 435

virulence profile that includes the F10 papA allele, and (excepting for strain 518: lane 8) 436

exhibit an extra large band in the PFGE profile (vertical arrows, or "*" for strain 518). 437

The profiles of the CGA isolates are all unique, yet they exhibit similarities that 438

distinguish them as a group from the E. coli O157:H7 reference strain (lanes 1, 10, 18). 439

440

441

ACCE

PTED

on March 30, 2021 by guest

http://aem.asm

.org/D

ownloaded from

http://aem.asm.org/

442

ACCE

PTED

on March 30, 2021 by guest

http://aem.asm

.org/D

ownloaded from

http://aem.asm.org/