View
0
Download
0
Category
Preview:
Citation preview
1
1
Diagnostic potential of monoclonal antibodies specific to the unique O-antigen of 2
multi-drug resistant epidemic E. coli clone ST131-O25b:H4 3
4
Valéria Szijártó1, Jolanta Lukasiewicz2, Tomasz K. Gozdziewicz2, Zoltán Magyarics1, 5
Eszter Nagy1, Gábor Nagy1# 6
1 Arsanis Biosciences GmbH, Helmut Qualtinger Gasse 2, 1030 Vienna, Austria 7
2 Department of Immunochemistry, Ludwik Hirszfeld Institute of Immunology and 8
Experimental Therapy, Polish Academy of Sciences, R. Weigla 12, 53-114 Wroclaw, Poland 9
10
11
# Corresponding author. E-mail: gabor.nagy@arsanis.com Tel.: +43-676898543500 12
13
Running title: E. coli O25b specific mAbs 14
Word count for abstract: 234 15
16
17
18
19
CVI Accepts, published online ahead of print on 30 April 2014Clin. Vaccine Immunol. doi:10.1128/CVI.00685-13Copyright © 2014, American Society for Microbiology. All Rights Reserved.
on April 19, 2021 by guest
http://cvi.asm.org/
Dow
nloaded from
2
Abstract 20
21
The Escherichia coli lineage ST131-O25b:H4 is a globally spread multi-drug resistant clone 22
responsible for a significant proportion of extraintestinal infections. Driven by the high 23
medical need associated with this successful pathogenic lineage, we generated murine 24
monoclonal antibodies against its lipopolysaccharide (LPS) O25b antigen in order to develop 25
quick diagnostic tests. Murine mAbs were generated by immunization of mice with whole 26
killed non-encapsulated ST131-O25b E. coli cells and screening hybridoma supernatants for 27
binding to purified LPS molecules obtained from an E. coli ST131-O25b clinical isolate. The 28
mAbs selected for further study bound to the surface of live E. coli O25b strains irrespective 29
of the capsular type expressed, while they could not bind to bacteria or purified LPS from 30
other serotypes - including the related classical O25 antigen (O25a). Using these specific 31
mAbs we have developed a latex bead-based agglutination assay that has greater specificity, 32
more rapid and simpler than the currently available typing methods. The high specificity of 33
these mAbs can be explained by the novel structure of the O25b repeating unit elucidated in 34
this paper. Based on comparative analysis by NMR and mass spectrometry, the N-acetyl-35
fucose in the O25a O-antigen had been replaced by O-acetyl-rhamnose in the O25b repeating 36
unit. The genetic determinants responsible for this structural variation were identified by 37
alignment of corresponding genetic loci, and were confirmed by trans-complementation of a 38
rough mutant by the sub-serotype specific fragments of the rfb operons. 39
40
on April 19, 2021 by guest
http://cvi.asm.org/
Dow
nloaded from
3
Introduction 41
42
Infections by Gram-negative multi-drug resistant (MDR) bacteria represent an increasing 43
health care problem worldwide. Whilst there are still some new antibiotics with some efficacy 44
against Gram-positive bacteria, the pipeline of novel drugs being developed to treat Gram-45
negative pathogens is essentially empty. The potential spread of MDR pathogens therefore 46
poses a threat, which could result in levels of morbidity and mortality similar to those 47
associated with infectious diseases in the pre-antibiotic era. A particular concern is the recent 48
emergence of clonal lineages that can balance the normally mutually exclusive phenotypic 49
properties of being MDR with the retention of high virulence potential, a feature which is 50
generally unusual among MDR strains. 51
E. coli ST131-O25b:H4 is a well-characterized multi-drug resistant clonal lineage that has 52
spread globally (1-3) in the last few years since it was first described in 2008 (4). This clone 53
alone is responsible for more than 10% of all extraintestinal E. coli infections, and accounts 54
for the greatest majority of E. coli strains resistant to clinically relevant antibiotics (5). The 55
vast majority of ST131-O25b isolates are resistant to fluoroquinolones. Moreover, approx. 56
50% of isolates producing an extended spectrum beta-lactamase (ESBL) that confers 57
resistance to all beta-lactam antibiotics except the carbapenems, originate from this clone. 58
Even more alarmingly there are several recent reports, which describe representative strains of 59
this lineage expressing various carbapenemases (6-8). Consequently, infections by ST131-60
O25b:H4 strains are a growing concern with very limited therapeutic options. 61
The obvious pathogenic success of this lineage is conferred by the MDR phenotype and 62
retained virulence potential. However, the range of factors contributing to virulence have still 63
to be fully elucidated (9). In a murine model of ascending urinary tract infection, a 64
on April 19, 2021 by guest
http://cvi.asm.org/
Dow
nloaded from
4
representative ST131 strain was shown to outcompete prototype UPEC isolates (10). 65
Conversely, other studies have reported that ST131 isolates are not more (11) or even less 66
virulent (12) in various animal models than fully susceptible E. coli isolates. This could be 67
explained by the lower number of average virulence factors expressed in ST131 isolates in 68
comparison to non-MDR strains (11). However, when compared with other lineages of ESBL-69
producing E. coli isolates, ST131 strains were shown to carry significantly higher number of 70
virulence genes (9,13,14). High metabolic potential was recently suggested to contribute to its 71
overall success (10,15). Finally, the substantial reduction of core genome recombination 72
events showed recently for this clone (16) results in a phylogenetically distinct and stable 73
pathogenic clone that is expected to remain an important extraintestinal pathogenic E. coli 74
lineage (ExPEC). 75
In spite of the high medical importance, detection of this specific clone among clinical 76
isolates of E. coli is not performed routinely. This is partly due to the lack of reliable and rapid 77
diagnostic assays. For epidemiological studies, ST131-O25b isolates are identified by multi 78
locus sequence typing and the detection of the specific lipopolysaccharide (LPS) O-antigen 79
repeating unit (RU). For the latter, two methods are used: i) agglutination with O25 rabbit 80
typing serum, and ii) detection of a serotype-specific gene segment within the rfb locus 81
encoding O-antigen synthesis by PCR. Sensitivity and specificity of the immune assay is 82
suboptimal, and the PCR based method is not practical for routine clinical microbiology 83
testing. 84
In this paper we describe the discovery of mAbs with specificity towards a sugar epitope that 85
is unique to the O25b O-antigen carried by ST131 strains. We demonstrate that these mAbs 86
function as reliable diagnostic tools in a convenient agglutination assay that is more sensitive 87
and more easily applicable to routine use than the currently available typing methods. 88
89
on April 19, 2021 by guest
http://cvi.asm.org/
Dow
nloaded from
5
Materials and methods 90
91
Bacteria and growth conditions 92
Two representative, genotypically and phenotypically confirmed ST131-O25b:H4 clinical 93
isolates (17) were used routinely for the in vitro studies. For the agglutination assay, a larger 94
panel (n=44) of ST131-O25b isolates , as well as non-related (i.e. expressing O-antigens other 95
than O25b) E. coli strains were used, kindly provided by Agnes Sonnevend (Al Ain, UAE), 96
Aranzazu Valverde (Madrid, Spain), Franz-Josef Schmitz (Minden, Germany) or obtained 97
from commercial strain collections (ATCC, NCTC, Polish Collection of Microbes - PCM). 98
The prototype sequenced O25a strain E47a was obtained from NCTC. Strain 509A is a human 99
fecal isolate (18), that was confirmed by O-typing (Hungarian Epidemiology Center) to 100
express O2 antigen. Mouse immunization experiments (see below) were performed with an 101
isogenic knock out mutant of the representative ST131 strain 81009 that was generated by 102
deletion of the whole kps cluster encoding capsular synthesis. A rough derivative of 81009 103
was generated by deletion of the gene encoding O-antigen ligase (waaL). E47aΔrfb3 is a 104
rough derivative of E47a lacking the 3’ end of the rfbO25a locus. All mutants were generated 105
by the Red recombinase method (19) using oligonucleotides listed in Table S1. 106
Bacteria were grown in Luria Bertani (LB) broth (Fischer Scientific) or agar plates. When 107
appropriate, selective media were containing ampicillin (100 µg/ml), kanamycin (100 µg/ml) 108
or chloramphenicol (25 µg/ml). 109
110
111
on April 19, 2021 by guest
http://cvi.asm.org/
Dow
nloaded from
6
LPS purification 112
The LPS of E. coli strain 81009 was isolated by the hot phenol/water method and purified by 113
dialysis, proteinase K digestion and ultracentrifugation. All other LPS molecules were 114
purified using a commercial kit (LPS Extraction Kit, Intron). 115
116
Molecular cloning 117
The approximately 3 and 7 kb fragments of the O25a and O25b rfb loci, respectively, were 118
amplified (Phusion PCR mix, New England Biolabs) using primers O25 control fw + rev and 119
O25b-spec1 + spec-2 (Table S1), and were directly cloned into high copy number expression 120
vector pJET1.2 (Fermentas) giving rise to p3O25a and p3O25b, respectively. The identical 121
orientation (negative strand with respect to the T7 promoter) was confirmed by PCR and 122
sequencing. Plasmids were purified and transformed following standard protocols. 123
124
Immunizations and hybridoma generation 125
6-8-week-old female BALB/c mice were immunized with approximately 108 CFU of formalin 126
killed cells of strain 81009Δkps subcutaneously 3 times with 3-week intervals. Four days 127
following a final intravenous boost, splenocytes of selected mice were isolated and subjected 128
to hybridoma fusion. Fusion and sub-culturing of hybridomas were performed in the 129
Monoclonal Antibody Facility (MAF) at the University of Vienna, Austria. Culture 130
supernatants of hybridoma clones were tested by enzyme-linked immunosorbent assay 131
(ELISA), flow cytometry and immunoblots, based on which the specific clones were isolated. 132
Isotype of the purified mAbs obtained from the selected clones was determined by IsoQuick 133
mouse isotyping kit (Sigma-Aldrich). 134
on April 19, 2021 by guest
http://cvi.asm.org/
Dow
nloaded from
7
Immunoblotting 135
Purified LPS (1 µg) was separated in 12% polyacrylamide gels (Bio-Rad) at constant 35 mA 136
current. LPS was transferred to 0.2 μm polyvinylidene difluoride (PVDF) membranes with 137
Trans-Blot® TurboTM Blotting system (Bio-Rad) with high MW program (Bio-Rad, 1.3 A up 138
to 25 V 10 min). Following overnight blocking in 5% BSA (PAA, Austria), membranes were 139
reacted with 1 μg monoclonal antibody or hyperimmune rabbit serum (E. coli O25, Statens 140
Serum Institut - SSI) in 1:1000 dilution for 1h at room temperature (RT). Following 3 washes 141
in 0.05% Tween 20 (Fisher Scientific)-TBS buffer for 10 min, membranes were incubated 142
with horseradish peroxidase (HRP)-conjugated secondary antibody (goat anti-mouse IgG, 143
Southern Biotech, 1:20.000 dilution or anti-rabbit IgG, Southern Biotech, 1:5.000 dilution) for 144
1h at RT. Following repeated washing, the membranes were developed with ECL Prime 145
Western blotting reagent (GE Healthcare). 146
147
Silver staining 148
LPS samples were separated as described above for immunoblots. Silver staining was 149
performed as published previously (20). Briefly, following o/n fixation in 25% isopropanol 150
and 7% acetic acid (Fisher Scientific) solution, the gel was oxidized by 0.7% periodic acid 151
(Sigma-Aldrich) in 40% ethanol and 5% acetic acid. Following repeated washing in distilled 152
water, the gel was stained with 0.8% silver-nitrate (Sigma-Aldrich) in 1.4% ammonium-153
hydroxide and 200 mM sodium-hydroxide solution and developed with 0.019% formaldehyde 154
in 0.005% citric acid buffer. 50 mM EDTA was used to stop the reaction. 155
156
157
on April 19, 2021 by guest
http://cvi.asm.org/
Dow
nloaded from
8
Flow cytometry 158
For surface staining, bacteria inoculated from overnight cultures in LB medium were grown 159
to mid-log phase (OD600≈0.5), washed twice and re-suspended in Hank's Balanced Salt 160
Solution (HBSS, Gibco Life Technologies) without Ca2+ and Mg2+. Bacteria (106 CFU) were 161
stained with the indicated hybridoma supernatants diluted 2.5-fold in HBSS buffer 162
supplemented with 0.5% BSA and 0.01% sodium azide, and incubated for 30 minutes on ice. 163
Following two washing steps in HBSS buffer with BSA (PAA, Austria) and sodium azide 164
(Sigma-Aldrich), bacteria were stained with 3 µg/ml of Alexa Fluor 488 conjugated goat 165
F(ab’)2 secondary antibody against mouse IgG (Jackson ImmunoResearch) for 30 min on ice. 166
Then, bacteria were washed twice, re-suspended in HBSS buffer containing 5 µM SYTO-62 167
nucleic acid stain (Life Technologies), and incubated for 10 min at room temperature. 168
Samples were measured by Eclipse flow cytometer (i-Cyt / Sony Biotechnology) and list 169
mode data were analyzed using FCSExpress software Version 4 (De Novo Software). 170
171
ELISA 172
ELISA was performed using 96-well plates coated with lysates of E. coli with different O-173
serotypes (O1, O4, O7, O12, O15, O16, O17, O18, O25, O75, O105, O157). Bacteria were 174
cultured in LB medium overnight at 37˚C. Following washing with PBS containing Ca2+ and 175
Mg2+ (PAA, Austria) bacteria were lysed at 100˚C for 1h. Plates were coated with lysates of 176
bacteria (108 CFU/well) in PBS with Ca2+ and Mg2+ overnight at 4˚C, followed by blocking 177
with 2% BSA (PAA) in PBS for 1h at RT. After washing 3-times with 0.05% Tween 20 178
(Fisher Scientific)-PBS, plates were reacted with O25b-specific mouse mAb (1 μg/ml) or with 179
mAb9004 (1 μg/ml, Glycobiotech, Germany) for 1h at RT, then washed 3-times. As secondary 180
antibody HRP conjugated goat anti-mouse IgG (SouthernBiotech) was used in 1:10.000 181
on April 19, 2021 by guest
http://cvi.asm.org/
Dow
nloaded from
9
dilution. ABTS solution (Novex® LifeTechnologies) was added to the washed plates, 182
incubated for 30 min in dark and absorbance was measured at 405 nm with Synergy HT 183
reader (Bio-Tek). 184
185
Agglutination assays 186
O-serotyping was performed with rabbit serum specific to E. coli O25 (SSI) according to the 187
manufacturer instructions. The O25b specific mAb was coupled to latex beads with passive 188
absorption. 100 μl of 2.5% slurry of red polystyrene beads with 1 μm diameter (Polysciences) 189
were washed and incubated with 100 μg antibody for 30 min at 37˚C with 400 rpm shaking. 190
After washing and centrifugation, beads were re-suspended in PBS containing 0.1% BSA and 191
0.05% Tween 20 to obtain a 1% bead suspension. Agglutination with live bacteria was tested 192
with freshly prepared antibody coated beads; 5-10 μl bead suspension was mixed with 193
bacterial mass of a single colony. Agglutination (clumping of the beads as well as clearance of 194
the background) was read by naked eye within 30sec. 195
196
Sequencing 197
Genomic DNA of strain 81009 was purified with Wizard® Genomic DNA Purification Kit 198
(Promega). The rfb operon was amplified with KlenTaq® LA DNA Polymerase (Sigma-199
Aldrich) with corresponding primers listed in Table S1, sequenced with primer walk 200
technique at Microsynth AG and assembled with CLC Main Workbench 6.7.1. The O-antigen 201
biosynthesis cluster of strain 81009 has been deposited in GenBank (accession number: 202
KF277146). 203
204
on April 19, 2021 by guest
http://cvi.asm.org/
Dow
nloaded from
10
Isolation of poly- and oligosaccharides 205
LPS (200 mg) was degraded by treatment with 1.5% acetic acid at 100°C for 30 min, frozen 206
at -20°C, and then the hydrolysis was continued for 20 min. Soluble poly- (PS) and 207
oligosaccharides (OS) were fractionated on a column (1.6 cm x 100 cm) of Bio-Gel P-10 208
equilibrated with 0.05 M pyridine/acetic acid buffer at pH5.6. Selected fractions were 209
analyzed by 1H NMR spectroscopy, electrospray ionisation (ESI) and matrix-assisted laser 210
desorption ionization time-of-flight (MALDI-TOF) mass spectrometry (MS). 211
212
Compositional analysis 213
Methylation of PS and OS was performed according to the method described by Ciucanu and 214
Kerek (21). Alditol acetates and partially methylated alditol acetates were analysed by GC–215
MS with Thermo Scientific TSQ system using RX5 fused-silica capillary column (0.2 mm × 216
30 m) and a temperature program of 150–270ºC at 12 ºC/min. The absolute configuration of 217
sugar residues was assigned according to method of Gerwig (22). Sililated butyl glycosides of 218
L and D sugars were analysed with a temperature program of 100–270ºC at 5 ºC/min. 219
220
NMR spectroscopy 221
All NMR spectra were recorded on Bruker Avance III 600 MHz spectrometer. NMR spectra 222
of PS and OS samples were obtained for 2H2O solutions at 25°C using acetone (δH 2.225, δC 223
31.05) as an internal reference. The samples were first repeatedly exchanged with 2H2O 224
(99%). The data were acquired and processed using standard Bruker software. The processed 225
spectra were assigned with the help of SPARKY (T. D. Goddard and D. G. Kneller, SPARKY 226
3, University of California, San Francisco). The signals were assigned by one- and two-227
on April 19, 2021 by guest
http://cvi.asm.org/
Dow
nloaded from
11
dimensional experiments (COSY, dqf-COSY, clean-TOCSY, NOESY, ROESY, HMBC, 228
HSQC-DEPT, HSQC, HSQC-TOCSY, and 1H, 13C, and 31P HMBC). In the clean-TOCSY 229
experiments, the mixing times of 30, 60, and 100 ms were used. The delay time in the HMBC 230
was 60 ms and the mixing time in the NOESY and ROESY experiments 200 ms. 231
232
Mass spectrometry 233
Negative ion mode MALDI-TOF MS of PS and OS was carried out on a Bruker Reflex III 234
time-of-flight instrument. As matrix 2,5-Dihydroxybenzoic acid (10 mg/ml, acetonitryl/water, 235
1:1) was used. Spectra were scanned in the range of m/z 800–6000. External calibration in the 236
negative-ion mode was applied using the Peptide Calibration Standard II (Bruker Daltonics, 237
Germany). ESI-MS experiments were carried out on a micrOTOF-Q II spectrometer (Bruker 238
Daltonics, Germany) in the positive ion mode. The samples were dissolved in acetonitrile-239
water-formic acid solution (50:50:0.5 [vol/vol/vol]; 50 µg/ml). Source parameters were as 240
follows: sample flow - 3 µl/min, ion source temperature - 180˚ C, nitrogen flow at 4 l/min and 241
at a pressure of 0.4 bar. Spectra were scanned in 50-3000 m/z range. External calibration in 242
positive-ion mode was applied using ESI Low TuneMix mixture (Agilent) in quadratic-243
regression mode. All structures were drawn and their molecular weights were calculated with 244
the use of GlycoWorkbench software (23). 245
246
247
on April 19, 2021 by guest
http://cvi.asm.org/
Dow
nloaded from
12
Results 248
249
Generation of O25b specific mAbs 250
Murine mAbs specific for the E. coli O25b LPS antigen were generated by standard 251
hybridoma technique using spleens of mice immunized with formalin killed bacterial cells as 252
described in the materials and methods. Selection of antibodies that can be beneficial for 253
serotype determination (i.e. diagnostics) was based on their capacity to bind to the surface of 254
live E. coli ST131 cells. The flow cytometry-based staining confirmed accessibility of the 255
epitope recognized by three different antibodies (mAbs 6D1-1B2, 8D5-1G10, 8A1-1G8) 256
using two representative ST131 strains (81009 and 80503) expressing different capsular 257
polysaccharides (Fig. 1 upper and middle panel). E. coli strains expressing non-O25 antigens 258
(data not shown), as well as the related O25a (strain E47a) antigen, were not labelled by these 259
mAbs (Fig. 1 lower panel). Specificity of the mAbs was further confirmed by immunoblot 260
analysis using purified LPS molecules (Fig. 2). The murine mAbs recognized the LPS 261
molecules purified from ST131 strains containing the O25b antigen, however, did not react 262
with those containing the O25a or other O-antigens (Fig. 2B). Furthermore, LPS purified from 263
an isogenic mutant of strain 81009 lacking O-antigen ligase WaaL and hence expressing no 264
O-antigen on the surface (rough /’R’/ mutant), was not detected by these antibodies (Fig. 2B 265
lane 3). Immunoblot staining with the commercial O25 typing serum confirmed cross-266
reactivity between O25a and O25b antigens, although with significantly lower reactivity to 267
the latter (Fig. 2C). 268
Similarly, in an ELISA using bacterial lysates for coating, only strains expressing O25b 269
antigen were detected, while none of the common extraintestinal pathogenic E. coli (ExPEC) 270
serotypes investigated reacted with the O25b specific mAb (Fig. 3A-C). Nevertheless, all 271
on April 19, 2021 by guest
http://cvi.asm.org/
Dow
nloaded from
13
strains reacted strongly with the inner-core specific (i.e. cross-reactive) mAb9004 272
(Glycobiotech, Germany), confirming availability of LPS antigens in the assay (Fig. 3D). 273
All these data together suggest that the mAbs are highly specific to the O25b antigen. 274
Moreover, our data provide experimental evidence that the structure of O25b antigen indeed 275
differs from that of classical O25 (termed O25a) antigen (24). 276
277
Structural analysis of O25b antigen 278
The novel structure of the O25b O-antigen repeating unit (RU) was elucidated and K-12 type 279
of the core oligosaccharides was identified by 1H and 13C NMR spectroscopy, MALDI-TOF 280
and ESI mass spectrometry, and sugar and methylation analyses. The LPS of E. coli O25b 281
(strain 81009) was isolated by the hot phenol/water method with the yield of 2.6%. The O-282
specific PS and OS were released by mild acidic hydrolysis of the LPS and isolated by gel 283
filtration. Four fractions were obtained (fractions 1-4), and due to complexity of fraction 1, 284
fractions 2, 3, and 4 were used for structural analysis. ESI and/or MALDI-TOF MS showed 285
that fraction 4 consisted of unsubstituted core OS (a heptasaccharide, Fig. S1), fraction 3 286
consisted of the core OS substituted with one repeating unit (RU) of the O-specific PS (Fig. 287
4B, Fig. S2), while fraction 2 of the core OS substituted with 1-4 RUs (Fig. 4C). 288
Structure of the RU of the O25b antigen and its linkage to the core OS was determined 289
with the use of NMR spectroscopy. The complete assignment of fraction 3 1H and 13C 290
resonances (Table S2) was assigned by one- and two-dimensional NMR experiments. The 291
inter-residue connections between adjacent sugar residues were observed by NOESY and 292
HMBC experiments (Table S3). Additionally all residues, besides Kdo and phosphate-293
substituted monosaccharides, were identified by basic sugar and methylation analyses (data 294
not shown). As the core oligosaccharide region of fraction 3 was identified as known K-12 295
on April 19, 2021 by guest
http://cvi.asm.org/
Dow
nloaded from
14
type glycoform assignments were compared with published data (25). The spectra indicated a 296
tetradecasaccharide structure of fraction 3 (Fig. 4A, Tables S1 and S2) containing a 297
pentasaccharide RU built of terminal β-D-Glcp (residue A), terminal α-L-Rhap (residue B), 298
→3,6)-α-D-Glcp (residue C), 3-substituted α-L-Rhap2OAc (residue D), and →3)-β-D-299
GlcpNAc (residue E). Analysis of fraction 2 allowed identification of the biological repeating 300
unit of the O-specific PS, which is a pentasaccharide with →3)-β-GlcpNAc (residue E) as a 301
RU constituent substituting the first residue of the core OS: →7)-α-L,D-Hepp (residue F). 302
Methylation analysis performed on fractions 1 and 2 indicated the presence of 3,4,6-303
substituted-Glcp, what indicates a position C-4 of residue C as a place of substitution of 304
subsequent RU of the O-specific PS. Thus, the structure of the O25b RU of ST131 LPS 305
differs from the O25a RU by a single sugar residue: →3)-α-L-FucpNAc (in O25a) replaced 306
by →3)-α-L-Rhap2OAc (in O25b) (Fig. 4A grey box). 307
The sequences of the identified tetradecasaccharide (core+RU) and unsubstituted core OS 308
glycoform (fraction 4) were confirmed with the use of ESI-MSn (Fig. 4B and Fig. S1 C, inset 309
structures). All fragment ions were interpreted on the basis of the herein elucidated structure 310
of the O25b RU and previously identified glycoforms of K-12 core OS (25,26) and according 311
to the nomenclature of Domon and Costello (27). It was shown that the purified LPS 312
molecules contained two main alternative core OS glycoforms. The type of glycoform is 313
dependent on the presence or absence of the O-specific PS. Prevailing glycoform of the 314
unsubstituted core OS is a truncated version of K-12 core oligosaccharide, which is devoid of 315
outer core region →7)-α-Hepp-(1→6)-α-Glcp disaccharide. 316
Molecular weight of the identified O25b RU was confirmed for fraction 2. The MALDI-TOF 317
MS spectrum (Figure 4C) showed clusters of ions with following prevailing ions: m/z 2797.2, 318
m/z 3658.8, m/z 4520.8, and m/z 5382.6 attributed to core OS (with P and/or PPEtn) 319
substituted with 1, 2, 3, and 4 RU, respectively. Average mass difference among these ions 320
on April 19, 2021 by guest
http://cvi.asm.org/
Dow
nloaded from
15
was 861.9 Da that matched the calculated monoisotopic mass of the O-specific PS repeating 321
unit (861.3 Da, RU-H2O). 322
323
Genetic analysis of rfbO25b 324
To determine the genetic basis of the structural differences observed between the O25a and 325
O25b O-antigen RU-s, we analysed the rfb gene clusters responsible for their synthesis. We 326
sequenced the complete rfbO25b locus from ST131:O25b isolate 81009 (GenBank accession 327
number: KF277146) and compared to the corresponding rfbO25a locus from O25a strain E47a 328
(accession number: GU014554). We found that the approx. 9 kb-long 5’ regions 329
encompassing 8 genes are highly homologous, while the 3’ ends downstream of the wzy genes 330
are dissimilar (Fig. 5). In case of O25a, the unique region includes 7 genes predicted to be 331
involved in N-acetyl-fucose synthesis and transfer. This region is completely replaced by 3 332
putative genes in the O25b operon. The first gene, wbbJ, has a potential O-acetyl transferase 333
function, the two downstream genes show homology with putative glycosyltransferases of E. 334
coli O16 (accession number: U00096). The 5’ 533 bp of the putative O-acetyl transferase gene 335
is also present in the O25a rfb operon, however due to a frame shift mutation a premature stop 336
codon was introduced at the 8th codon. Furthermore, the 3’ end of the gene is truncated (73 bp 337
deletion) due to the insertion sequence insC, which might have been involved for the lateral 338
transfer of the downstream O25a-specific genes. Interestingly, the O25b-specific region 339
displays a GC content that is similar to that of the shared immediate upstream region, while 340
that in the O25a-specific part is significantly higher (Fig. 5B) implying more recent 341
recombination event within rfbO25a. 342
In order to prove that the observed differences within the rfb operons are the sole 343
determinants of the structural differences in the O-antigen subunits, complementation studies 344
on April 19, 2021 by guest
http://cvi.asm.org/
Dow
nloaded from
16
were performed. The approx. 3 and 7 kb-long 3’ serotype specific fragments were cloned into 345
an expression vector giving rise to p3O25b and p3O25a, respectively. These plasmids were 346
used for complementing E47aΔrfb3, an O25a mutant having the 3’ variable end of the rfb 347
operon deleted. The mutant exhibited a rough LPS phenotype as shown by lack of 348
agglutination by O25 serum as well as by silver staining of purified LPS (Fig. 6A). As 349
expected, complementation with the homologous region encoded on p3O25a restored the 350
smooth O25a phenotype. In contrast, complementation of the same mutant with p3O25b 351
resulted in expression of polymers of O25b repeat units as confirmed by agglutination and 352
reactivity in immunoblots with O25b specific mAb (Fig. 6B). These data corroborate that the 353
distinct 3’ region within the rfbO25b cluster is the exclusive determinant of the observed 354
serological difference. 355
356
Diagnostic assay with O25b-specific mAbs 357
To explore the applicability of the O25b specific mAbs for a clinically useful diagnostic assay, 358
mAb 6D1-1B2 was coupled to latex beads and tested for its ability to agglutinate E. coli 359
strains belonging to the ST131-O25b lineage (Table 1). We observed highly specific reaction 360
with all ST131 O25b strains tested, and no visible agglutination with any other strains, such 361
as those expressing the O25a, O2, O4, and O16 antigens (Fig. 7). 362
We compared the specificity and sensitivity of this assay to those currently used as state of the 363
art typing assays, i.e. PCR detection of a specific region within rfbO25b (28) and agglutination 364
with O25-specific rabbit serum. Agglutination with commercial typing sera requires heat 365
treatment of the E. coli cells in order to expose the target O-antigen as well as to denature 366
potential reactive protein targets. No such pre-treatment was necessary to enable positive 367
agglutination with the mAb-coupled beads (Table 1). Even in the case of heat killed cells the 368
on April 19, 2021 by guest
http://cvi.asm.org/
Dow
nloaded from
17
sensitivity of the mAb coupled-bead assay was found superior to that of the serum-based 369
typing. 370
Interestingly, one strain that was found to be positive with the O25b-specific PCR did not 371
agglutinate with the O25b mAb coated beads. We determined that this particular strain 372
displayed rough LPS phenotype by silver staining (data not shown) due to a yet unidentified 373
mutation (strain 81010 in Table 1). 374
375
on April 19, 2021 by guest
http://cvi.asm.org/
Dow
nloaded from
18
Discussion 376
377
Infections by E. coli strains from the ST131-O25b lineage represent an increasingly 378
significant medical concern due to their extended multi-drug resistant phenotype. By 379
definition, all strains from this clone express the O25b LPS repeating unit. Therefore, we 380
aimed at generating mAbs against this specific antigen to develop improved diagnostics. To 381
obtain specific mAbs, mice were immunized with whole killed cells of an isogenic non-382
encapsulated mutant of an ST131-O25b clinical isolate in order to avoid generation of 383
antibodies against the bulky capsular polysaccharide. We selected three mAbs with unique 384
CDR sequences that specifically recognized the O25b antigen expressed on the surface of the 385
multi-drug resistant ST131 clone, but showing no cross-reactivity with other E. coli strains 386
expressing different O-antigens. Specificity was further demonstrated by the lack of binding 387
to LPS molecules obtained from an isogenic rough mutant as well as to other smooth LPS 388
molecules from several E. coli serotypes including the related O25a antigen. The fact that 389
O25a polyclonal serum is cross-reactive with O25b antigens (representing one of the state of 390
the art detection of O25b strains), while these mAbs do not cross-react implies that these 391
mAbs recognize a unique epitope present in the O25b RU, however not available within the 392
O25a antigen. 393
394
In order to support this notion, the structure of the O25b repeating unit was elucidated. As 395
expected, chemical composition of the O25b antigen was found to be different from that of 396
the related O25a antigen. The main alteration was the presence of L-Rhap2OAc vs. L-397
FucpNAc in the central position of the repeating units. Since O-acetylated carbohydrate 398
structures are considered to be immuno-dominant it is not surprising that all 3 mAbs selected 399
on April 19, 2021 by guest
http://cvi.asm.org/
Dow
nloaded from
19
from the hybridoma fusion were highly specific to the O25b antigen (i.e. do not cross-react 400
with O25a). Next, we tested reactivity of these mAbs to E. coli cells expressing O-antigens 401
with a similar sugar composition, i.e. containing an O-acetylated rhamnose residue (29). None 402
of these antigens, nor any of the common ExPEC O-types were detected by any of the 3 403
mAbs corroborating that their epitope is restricted to the unique O25b antigen structure 404
presented in this paper. 405
The distinctiveness of O25b structure is supported by the genetic background. We have 406
sequenced the entire rfb cluster of the clinical isolate that had been used for the immunization 407
as well as the LPS structural analysis. The rfb locus was found to be essentially identical to 408
that of ST131-O25b strain EC958 reported earlier (30). Comparison of rfb clusters from O25a 409
and O25b, however, revealed completely dissimilar regions at the 3’ end of the operons 410
following a shared approx. 7 kb-long 5’ region. The unique rfbO25b region consists of 3 genes, 411
including 2 with putative glycosyltransferase function. The third gene encodes a putative O-412
acetyl transferase, which is in good agreement with the structure described above. As O25a is 413
a common serotype of E. coli (both among ExPEC and intestinal pathogenic strains) (31), it 414
could be postulated that the O25b variant has evolved in order to evade pre-existing immunity 415
prevalent in the community. However, genetic analysis implies the opposite order of genetic 416
rearrangement. In rfbO25a the gene encoding the putative O-acetyl transferase is truncated and 417
flanked by a completely different 3’ region, adjacent to an insertion sequence suggesting a 418
potential site for recombination events. Therefore, we hypothesize that the locus encoding 419
O25a RU evolved by the integration of a different 3’ region within the archetype O25 rfb 420
operon (currently known as rfbO25b), which was also supported by analyzing the GC contents 421
of the determinants. Interestingly, the O25a specific region encoding FucNAc synthesis and 422
transfer shows high homology to determinants of other O-types (e.g. O4 and O26) also 423
on April 19, 2021 by guest
http://cvi.asm.org/
Dow
nloaded from
20
containing this sugar, hence those serotypes might have served as the source of the O25a 424
specific genes. 425
Since it is known that, in enterobacterial pathogens, phage-derived sequences carried 426
outside of the rfb operon can modify O-antigen backbone structures (e.g. with O-acetyl 427
groups), we considered it important to show that the 3 specific genes present within rfbO25b 428
are the sole determinants of the unique O25b RU structure. This was shown by creating an 429
isogenic mutant from E. coli O25a whose O antigen synthesis was lost upon deletion of the 3’ 430
end of its rfb locus (i.e. the O25a specific portion). Subsequently, this mutant (E47aΔ3rfb) 431
was trans-complemented with either the O25a or O25b specific regions of the corresponding 432
operons, both of which retained expression of smooth LPS. Nevertheless, unlike p3O25a, the 433
carriage of p3O25b in an O25a genetic background elicited reactivity to O25b specific mAbs, 434
corroborating the exclusivity of this region in the expression of the specific epitope. 435
Next, we tested the diagnostic potential of the specific mAbs in comparison to those 436
assays currently used for the detection of ST131-O25b strains. We have collected a panel of 437
isolates confirmed to be ST131-O25b from various geographical regions. All of these strains 438
gave a clear positive result with the agglutination assay using O25b specific murine IgG3. 439
Agglutination by mAbs coupled to commercial latex beads (i.e. the assay described in this 440
paper) appeared to be superior to the currently used agglutination test with O25 hyperimmune 441
rabbit serum with respect to specificity and sensitivity. Furthermore, in case this assay is 442
aimed to be used as a companion diagnostic tool for a prospective immunotherapy targeting 443
the O25b antigen, it is essential that only the isolates in fact expressing the antigen should be 444
detected. In this respect the agglutination assay has a clear advantage over PCR based 445
detection of O25b genes, which is illustrated by the PCR positive but rough strain we 446
identified. The frequency of such mutants carrying the rfbO25b operon but not expressing the 447
antigen is unknown and requires further investigation. Still, such strains identified by PCR 448
on April 19, 2021 by guest
http://cvi.asm.org/
Dow
nloaded from
21
would give a false positive indication for an O25b targeting therapy. Besides these scientific 449
considerations, the convenience, resource and time requisite of the latex agglutination assay 450
described is clearly superior to any of the currently used tools for the detection of ST131-451
O25b strains. Interestingly, environmental isolates belonging to ST69-D clonal group were 452
shown to express O25b antigen as suggested by rfb-specific PCR positivity (32). In case these 453
strains in fact express the O25b antigen, the phenotypic test described above would also detect 454
them Nevertheless the low virulence gene content within this clone (32) predicts rare 455
association of this clone with human infections. Recently, one clinical isolate from Denmark 456
was identified (13) as ST69-D-O25:H4, however, it is uncertain, which variant of O25 antigen 457
this isolate expressed. 458
Having shown the accessibility of O25b epitopes on the surface of encapsulated ST131-459
O25b isolates, it is tempting to speculate that humanized mAbs with such specificities might 460
be developed as efficacious novel immune therapeutics against this MDR clone. In this case 461
the described rapid and reliable agglutination assays could serve as an invaluable companion 462
diagnostic tool for the identification of relevant infections. 463
464
on April 19, 2021 by guest
http://cvi.asm.org/
Dow
nloaded from
22
Acknowledgements 465
466
The technical assistance of Srijib Banerjee and Pallavi Banerjee in mAb purification and 467
molecular typing of strains, respectively, is gratefully acknowledged. We thank to Agnes 468
Sonnevend (Al Ain, United Arab Emirates), Aranzazu Valverde (Madrid, Spain), and Franz-469
Josef Schmitz (Minden, Germany) for providing clinical isolates for this study. We are 470
grateful to Fraser Leslie for critical reading of the manuscript. 471
The research work of the Vienna Team was substantially supported by the General 472
Programme of the FFG - Austrian Research Promotion Agency. 473
VS, ZM, EN, and GN are employees of Arsanis Biosciences GmbH, a privately owned 474
biotechnology company. These authors own shares of the company. 475
476
on April 19, 2021 by guest
http://cvi.asm.org/
Dow
nloaded from
23
477
References 478
479
1. Peirano G, Pitout JD. 2010. Molecular epidemiology of Escherichia coli producing 480
CTX-M beta-lactamases: the worldwide emergence of clone ST131 O25:H4. 481
Int.J.Antimicrob.Agents 35:316-321. 482
2. Rogers BA, Sidjabat HE, Paterson DL. 2011. Escherichia coli O25b-ST131: a 483
pandemic, multiresistant, community-associated strain. J.Antimicrob.Chemother. 66:1-484
14. 485
3. Woodford N, Turton JF, Livermore DM. 2011. Multiresistant Gram-negative bacteria: 486
the role of high-risk clones in the dissemination of antibiotic resistance. FEMS 487
Microbiol.Rev. 35:736-755. 488
4. Nicolas-Chanoine MH, Blanco J, Leflon-Guibout V, Demarty R, Alonso MP, 489
Canica MM, Park YJ, Lavigne JP, Pitout J, Johnson JR. 2008. Intercontinental 490
emergence of Escherichia coli clone O25:H4-ST131 producing CTX-M-15. 491
J.Antimicrob.Chemother. 61:273-281. 492
5. Johnson JR, Johnston B, Clabots C, Kuskowski MA, Castanheira M. 2010. 493
Escherichia coli sequence type ST131 as the major cause of serious multidrug-resistant 494
E. coli infections in the United States. Clin.Infect.Dis. 51:286-294. 495
6. Bonnin RA, Poirel L, Carattoli A, Nordmann P. 2012. Characterization of an IncFII 496
plasmid encoding NDM-1 from Escherichia coli ST131. PLoS.One. 7:e34752. 497
on April 19, 2021 by guest
http://cvi.asm.org/
Dow
nloaded from
24
7. Dimou V, Dhanji H, Pike R, Livermore DM, Woodford N. 2012. Characterization of 498
Enterobacteriaceae producing OXA-48-like carbapenemases in the UK. 499
J.Antimicrob.Chemother. 67:1660-1665. 500
8. Morris D, Boyle F, Ludden C, Condon I, Hale J, O'Connell N, Power L, Boo TW, 501
Dhanji H, Lavallee C, Woodford N, Cormican M. 2011. Production of KPC-2 502
carbapenemase by an Escherichia coli clinical isolate belonging to the international 503
ST131 clone. Antimicrob.Agents Chemother. 55:4935-4936. 504
9. Peirano G, Mulvey GL, Armstrong GD, Pitout JD. 2013. Virulence potential and 505
adherence properties of Escherichia coli that produce CTX-M and NDM beta-506
lactamases. J.Med.Microbiol. 62:525-530. 507
10. Vimont S, Boyd A, Bleibtreu A, Bens M, Goujon JM, Garry L, Clermont O, 508
Denamur E, Arlet G, Vandewalle A. 2012. The CTX-M-15-producing Escherichia coli 509
clone O25b: H4-ST131 has high intestine colonization and urinary tract infection 510
abilities. PLoS.One. 7:e46547. 511
11. Johnson JR, Porter SB, Zhanel G, Kuskowski MA, Denamur E. 2012. Virulence of 512
Escherichia coli clinical isolates in a murine sepsis model in relation to sequence type 513
ST131 status, fluoroquinolone resistance, and virulence genotype. Infect.Immun. 514
80:1554-1562. 515
12. Lavigne JP, Vergunst AC, Goret L, Sotto A, Combescure C, Blanco J, O'Callaghan 516
D, Nicolas-Chanoine MH. 2012. Virulence potential and genomic mapping of the 517
worldwide clone Escherichia coli ST131. PLoS.One. 7:e34294. 518
on April 19, 2021 by guest
http://cvi.asm.org/
Dow
nloaded from
25
13. Olesen B, Hansen DS, Nilsson F, Frimodt-Moller J, Leihof RF, Struve C, Scheutz F, 519
Johnston B, Krogfelt KA, Johnson JR. 2013. Prevalence and characteristics of the 520
epidemic multiresistant Escherichia coli ST131 clonal group among extended-spectrum 521
beta-lactamase-producing E. coli isolates in Copenhagen, Denmark. J.Clin.Microbiol. 522
51:1779-1785. 523
14. van der Bij AK, Peirano G, Pitondo-Silva A, Pitout JD. 2012. The presence of genes 524
encoding for different virulence factors in clonally related Escherichia coli that produce 525
CTX-Ms. Diagn.Microbiol.Infect.Dis. 72:297-302. 526
15. Gibreel TM, Dodgson AR, Cheesbrough J, Bolton FJ, Fox AJ, Upton M. 2012. High 527
metabolic potential may contribute to the success of ST131 uropathogenic Escherichia 528
coli. J.Clin.Microbiol. 50:3202-3207. 529
16. McNally A, Cheng L, Harris SR, Corander J. 2013. The evolutionary path to 530
extraintestinal pathogenic, drug-resistant Escherichia coli is marked by drastic reduction 531
in detectable recombination within the core genome. Genome Biol.Evol. 5:699-710. 532
17. Szijarto V, Pal T, Nagy G, Nagy E, Ghazawi A, al-Haj M, El KS, Sonnevend A. 533
2012. The rapidly emerging ESBL-producing Escherichia coli O25-ST131 clone carries 534
LPS core synthesis genes of the K-12 type. FEMS Microbiol.Lett. 332:131-136. 535
18. Benckert J, Schmolka N, Kreschel C, Zoller MJ, Sturm A, Wiedenmann B, 536
Wardemann H. 2011. The majority of intestinal IgA+ and IgG+ plasmablasts in the 537
human gut are antigen-specific. J.Clin.Invest 121:1946-1955. 538
on April 19, 2021 by guest
http://cvi.asm.org/
Dow
nloaded from
26
19. Datsenko KA, Wanner BL. 2000. One-step inactivation of chromosomal genes in 539
Escherichia coli K-12 using PCR products. Proc.Natl.Acad.Sci.U.S.A 97:6640-6645. 540
20. Nelson D, Neill W, Poxton IR. 1990. A comparison of immunoblotting, flow cytometry 541
and ELISA to monitor the binding of anti-lipopolysaccharide monoclonal antibodies. 542
J.Immunol.Methods 133:227-233. 543
21. Ciucanu I, Kerek F. 1984. A simple and rapid method for the permethylation of 544
carbohydrates. Carbohydrate Research 131:209-217. 545
22. Gerwig GJ, Kamerling JP, Vliegenthart JFG. 1978. Determination of the d and l 546
configuration of neutral monosaccharides by high-resolution capillary g.l.c. 547
Carbohydrate Research 62:349-357. 548
23. Ceroni A, Maass K, Geyer H, Geyer R, Dell A, Haslam SM. 2008. GlycoWorkbench: 549
a tool for the computer-assisted annotation of mass spectra of glycans. J Proteome Res 550
7:1650-1659. 551
24. Kenne L, Lindberg B, Madden JK, Lindberg AA, Gemski P, Jr. 1983. Structural 552
studies of the Escherichia coli O-antigen 25. Carbohydr.Res. 122:249-256. 553
25. Muller-Loennies S, Lindner B, Brade H. 2003. Structural analysis of oligosaccharides 554
from lipopolysaccharide (LPS) of Escherichia coli K12 strain W3100 reveals a link 555
between inner and outer core LPS biosynthesis. J.Biol.Chem. 278:34090-34101. 556
on April 19, 2021 by guest
http://cvi.asm.org/
Dow
nloaded from
27
26. Duda KA, Lindner B, Brade H, Leimbach A, Brzuszkiewicz E, Dobrindt U, Holst 557
O. 2011. The lipopolysaccharide of the mastitis isolate Escherichia coli strain 1303 558
comprises a novel O-antigen and the rare K-12 core type. Microbiology 157:1750-1760. 559
27. Domon B, Costello CE. 1988. A systematic nomenclature for carbohydrate 560
fragmentations in FAB-MS/MS spectra of glycoconjugates. glycoconj. 5:397-409. 561
28. Clermont O, Dhanji H, Upton M, Gibreel T, Fox A, Boyd D, Mulvey MR, 562
Nordmann P, Ruppe E, Sarthou JL, Frank T, Vimont S, Arlet G, Branger C, 563
Woodford N, Denamur E. 2009. Rapid detection of the O25b-ST131 clone of 564
Escherichia coli encompassing the CTX-M-15-producing strains. 565
J.Antimicrob.Chemother. 64:274-277. 566
29. Stenutz R, Weintraub A, Widmalm G. 2006. The structures of Escherichia coli O-567
polysaccharide antigens. FEMS Microbiol.Rev. 30:382-403. 568
30. Totsika M, Beatson SA, Sarkar S, Phan MD, Petty NK, Bachmann N, Szubert M, 569
Sidjabat HE, Paterson DL, Upton M, Schembri MA. 2011. Insights into a multidrug 570
resistant Escherichia coli pathogen of the globally disseminated ST131 lineage: genome 571
analysis and virulence mechanisms. PLoS.One. 6:e26578. 572
31. Clermont O, Johnson JR, Menard M, Denamur E. 2007. Determination of 573
Escherichia coli O types by allele-specific polymerase chain reaction: application to the 574
O types involved in human septicemia. Diagn.Microbiol.Infect.Dis. 57:129-136. 575
on April 19, 2021 by guest
http://cvi.asm.org/
Dow
nloaded from
28
32. Colomer-Lluch M, Mora A, Lopez C, Mamani R, Dahbi G, Marzoa J, Herrera A, 576
Viso S, Blanco JE, Blanco M, Alonso MP, Jofre J, Muniesa M, Blanco J. 2013. 577
Detection of quinolone-resistant Escherichia coli isolates belonging to clonal groups 578
O25b:H4-B2-ST131 and O25b:H4-D-ST69 in raw sewage and river water in Barcelona, 579
Spain. J.Antimicrob.Chemother. 68:758-765. 580
33. Novais A, Viana D, Baquero F, Martinez-Botas J, Canton R, Coque TM. 2012. 581
Contribution of IncFII and broad-host IncA/C and IncN plasmids to the local expansion 582
and diversification of phylogroup B2 Escherichia coli ST131 clones carrying blaCTX-583
M-15 and qnrS1 genes. Antimicrob.Agents Chemother. 56:2763-2766. 584
34. Valverde A, Coque TM, Sanchez-Moreno MP, Rollan A, Baquero F, Canton R. 585
2004. Dramatic increase in prevalence of fecal carriage of extended-spectrum beta-586
lactamase-producing Enterobacteriaceae during nonoutbreak situations in Spain. 587
J.Clin.Microbiol. 42:4769-4775. 588
589
590
591
on April 19, 2021 by guest
http://cvi.asm.org/
Dow
nloaded from
29
Figure legends 592
593
Figure 1: Surface staining. 594
Binding of 3 different O25b-specific murine mAbs to different ST131-O25b strains 595
expressing various capsular polysaccharides (K5 and K2) as well as control strain E47a 596
(O25a). As a control irrelevant isotype matched murine antibody was used. 597
598
Fig. 2. Immunoreactivity of purified LPS molecules to various anti-O25 reagents. 599
LPS samples were separated by SDS-PAGE and silver stained (A), or blotted and developed 600
by O25b specific mAb 8D5-1G10 (1 ug/ml) (B), or blotted and developed by commercial 601
O25 typing serum (SSI, 1:1000 dilution) (C). 1: 81009 (O25b), 2: 80503 (O25b), 3: 602
81009∆waaL (rough), 4: E47a (O25a), 5: H54 (O25a), 6: 509A (O2), M: Molecular weight 603
marker. 604
605
Fig. 3. ELISA reactivity of O25b mAbs to bacterial lysates of different serotypes. 606
Reactivity of different dilutions of O25b specific mAbs 6D1-1B2 (A), 8A1-1G8 (B), 8D5-607
1G10 (C), as well as cross-reactive E. coli mAb 9004 as a control (D) was determined to heat 608
killed lysates of E. coli cells expressing various LPS O-antigens. 609
610
Fig. 4. Structural analysis of the fraction 3 isolated from LPS ST131. 611
A. Structure of the fraction 3 built of the K-12 type core OS substituted with O25 RU (framed 612
with a solid line). The uppercase letters refer to carbohydrate residues identified by NMR 613
analysis. Residue D discriminate O25b RU (marked with gray box) from O25a RU (α-L-614
FucpNAc). *Non-stoichiometric substituent. **Residue C is 3,4,6-substituted in fractions 2 615
and 1, and subsequent RU is placed at position 4. B. Positive ion mode ESI-MS2 of the 616
on April 19, 2021 by guest
http://cvi.asm.org/
Dow
nloaded from
30
fraction 3 glycoform represented by the ion at m/z 1338.886 (an inset structure). Interpretation 617
of the observed fragment ions is presented in the inset structures and based on the 618
nomenclature of Domon and Costello (27). # Noninterpreted ions. C. Negative ion mode 619
MALDI-TOF mass spectrum of the fraction 2. m/z values represent monoisotopic masses. 620
621
Fig. 5. Comparison of genetic loci encoding O25a and O25b subunits. 622
A. Schematic representation (not drawn to scale) of the gene composition of the different rfb 623
operons. The genes flanking the rfb loci (galF and gnd) are represented by the empty arrows. 624
B. GC/AT content across the rfb operons encoding O25a and O25b RUs. 625
626
Fig. 6. Immunoblot analysis of purified LPS molecules from recombinant E. coli strains. 627
LPS samples were separated by SDS-PAGE and silver stained (A), or blotted and developed 628
by O25b specific mAb 8D5-1G10 (1 ug/ml) (B), or blotted and developed by commercial 629
O25 typing serum (SSI, 1:1000 dilution) (C). Lane 1: E47a (O25a), 2: E47a∆rfb3 (R), 3: 630
E47∆rfb3/p3O25a clone #1, 4: E47a∆rfb3/p3O25a clone #2, 5: E47∆rfb3/p3O25b clone#1, 6: 631
E47∆rfb3/p3O25b clone #2, 7: 81009 (O25b). 632
633
Fig. 7. Latex agglutination. 634
Detection of O25b antigen expressing E. coli strains with agglutination assay using mAb 635
6D1-1B2 coupled to latex beads. A loopful of bacteria was mixed with 10 µl of a 1% bead 636
solution in PBS. Reactions were developed by gentle agitation for about 10s. 637
638
on April 19, 2021 by guest
http://cvi.asm.org/
Dow
nloaded from
31
Table 1. Validation of the O25b latex bead assay with ST131 isolates collected from 639
various geographical regions 640
strain
Clinical
specimen
/infection
origin ESBL
Ciprofl
oxacin
resistan
ce
core type O25b
PCR
agglutinati
on with
O25 typing
serum
agglutin
ation
with
latex
beads
coupled
O25b
mAb
Remarks
FJS 020 Blood Germany + R K-12 + + +
FJS024 Blood Germany + R K-12 + + +
FJS 053 Urine Germany + R K-12 + + +
FJS 059 Blood Germany - R K-12 + + +
FJS 072 Blood Germany + R K-12 + + +
FJS 095 VAP Germany + R K-12 + + +
FJS 098 VAP Germany + R K-12 + + +
80503 Urine UAE + R K-12 + + +
80505 Urine UAE + R K-12 + + +
80703 Urine UAE + R K-12 - - - O16 ST131
80907 Urine UAE + R K-12 + + +
81010 Urine UAE + R K-12 + - - rough
on April 19, 2021 by guest
http://cvi.asm.org/
Dow
nloaded from
32
81012 Urine UAE + R K-12 + + +
81102 Urine UAE + R K-12 + + +
90103 Urine UAE + R K-12 + - +
90105 Urine UAE + R K-12 + - +
90108 Urine UAE + R K-12 + - +
90306 Urine UAE + R K-12 + + +
90309 Urine UAE + R K-12 + + +
90310 Urine UAE + R K-12 + + +
90405 Urine UAE + R K-12 + + +
90409 Urine UAE + R K-12 + + +
90416 Urine UAE + R K-12 + + +
306-0838 Urine UAE - S K-12 + - +
B15 Urine Hungary - R K-12 + + +
DE22404 Blood Hungary - S K-12 + + +
DE8881 Blood Hungary - R K-12 + + +
SE40 Blood Hungary - S K-12 - - - O16 ST131
SE42 Blood Hungary - R K-12 + + +
SE6 Blood Hungary + R K-12 + + +
8 JN (33) Wound Spain + n.d. K-12 + + +
on April 19, 2021 by guest
http://cvi.asm.org/
Dow
nloaded from
33
2N (33) Urine Spain + n.d. K-12 + + +
5 JN (33) Urine Spain + n.d. K-12 + + +
10AR
(33) Urine Spain + n.d. K-12 + + +
7 O (33) Urine Spain + n.d. K-12 + + +
3 O (33) Blood Spain + n.d. K-12 + + +
1 N (33) Blood Spain + n.d. K-12 + + +
10 JN
(33) Urine Spain + n.d. K-12 + + +
3MR (33) Urine Spain + n.d. K-12 + + +
10J (33) Urine Spain + n.d. K-12 + + +
8MI (33) Urine Spain + n.d. K-12 + + +
6MI (33) Exudate Spain + n.d. K-12 + + +
FEC250
(34) Stool Spain + n.d. K-12 + + +
Phylogenetic
group D
C70 Urine Spain + n.d. K-12 + + + Phylogenetic
group D
641
642
on April 19, 2021 by guest
http://cvi.asm.org/
Dow
nloaded from
Recommended