Rapid Genomic-Scale Analysis of Escherichia coli O104:H4 by UsingHigh-Resolution Alternative Methods to Next-Generation Sequencing

Scott A. Jackson, Michael L. Kotewicz, Isha R. Patel, David W. Lacher, Jayanthi Gangiredla, and Christopher A. Elkins

Division of Molecular Biology, Center for Food Safety and Applied Nutrition, United States Food and Drug Administration, Laurel, Maryland, USA

Two technologies, involving DNA microarray and optical mapping, were used to quickly assess gene content and genomic archi-tecture of recent emergent Escherichia coli O104:H4 and related strains. In real-time outbreak investigations, these technologiescan provide congruent perspectives on strain, serotype, and pathotype relationships. Our data demonstrated clear discrimina-tion between clinically, temporally, and geographically distinct O104:H4 isolates and rapid characterization of strain differences.

The recent combined magnitude of illness, hemolytic uremicsyndrome, and death resulting from an uncommon O104:H4

Shiga toxin-producing Escherichia coli (STEC) pathogen in Eu-rope surpassed that of previous outbreaks with known STEC eti-ology (http://ecdc.europa.eu/en/aboutus/organisation/Director%20Speeches/201109_MarcSprenger_STEC_ICAAC.pdf). Thisnewly observed pathogen contains a rare combination of two vir-ulence pathotypes involving enterohemorrhagic E. coli (EHEC)and enteroaggregative E. coli (EAEC) attributes that manifest inShiga toxin production and adherence to intestinal epithelial cellsvia the production of specialized fimbrial structures, respectively(6, 11). An additional hallmark of this outbreak was the concom-itant research interest in competing next-generation sequencingplatforms for genomic investigation. Within days of pathogenidentification, the first raw genome sequence was released usingthe Ion Torrent platform. Two weeks later, the first whole-genome assembly was available incorporating Illumina HiSeqdata followed by subsequent efforts employing the Pacific Biosci-ences platform. In aggregate, the whole-genome sequencing ef-forts produced nearly identical de novo genome sequences from anumber of outbreak isolates (3, 6, 11, 14, 17).

It is noteworthy that the utility and speed of these technologiesare typically limited by initial access to bona fide outbreak isolateswith sound epidemiological linkage to the clinical illness. How-ever, from a practical applied and regulatory perspective, whole-genome sequencing and the subsequent bioinformatic identifica-tion of strain/outbreak-specific attributes are tempered by theneed for rapid front end identification, subtyping, and sourcetracking. In the past, we have used microarray genotyping data totriage outbreak strains for further secondary analyses. For exam-ple, array data from 250 clinical and bovine isolates from the 2006spinach-associated E. coli O157:H7 outbreak were clustered into asingle major group and two outliers for mapping and sequencing(10). More recent examples include clustering of highly clonalSalmonella enterica serovar Enteritidis from an egg outbreak, S.enterica serovar Tyhimurium in peanut butter/paste, and S. en-terica serovar Montevideo from pepper spice using microarrayand optical mapping (unpublished data).

Here, we present a rapid genomic-scale analysis regimen foremerging food-borne outbreaks as applied to this recent O104:H4serotype, using custom high-density microarray genotyping inconcert with high-resolution optical restriction fragment map-ping. Respectively, these techniques provide disparate but com-plementary perspectives on unique gene content and genomic

architecture with spatial relationships anchored to genetic land-marks and novel insertions and deletions. Both techniques arefacilitated by reference genome sequences as well as internallaboratory-generated genomic-scale databases for rapid (2- to3-day) comparative assessments that limit de novo annotation andpostbioinformatic manipulation.

Strains. In-depth laboratory analysis was conducted on inde-pendent O104:H4 isolates associated with the German outbreak(M-GOS and C-GOS) and two additional O104:H4 isolates (RG1and RG2) derived from patients from the Republic of Georgia in2009. Other comparisons involved two O104:H21 isolates, MON1and MON2, from a 1994 milk-associated outbreak in Montana(7), and an O139:H28 enterotoxigenic E. coli (ETEC) strain,E24377A (GenBank accession no. NC_009801) (13). Two in silicoreference strains included EAEC 55989 (GenBank accession no.NC_011748), an O104:H4 isolate from the Central African Re-public (2, 12, 18), and O157:H7 Sakai (GenBank accession no.NC_002695).

Genome-scale genotyping. The FDA E. coli-Shigella (ECSG)custom high-density Affymetrix DNA microarray platform con-tains the entire gene contents derived from 31 chromosomal and46 plasmid sequences from a diverse set of E. coli and Shigellaisolates. In total, over 21,000 unique genes are represented on theECSG array, corresponding to each of the 137,600 chromosomalgenes and 2,868 plasmid-borne genes from the sequences avail-able at the time of design. The gene content of individual strainscan be determined in an absolute manner without a comparativereference strain using the Affymetrix probe set design strategyincorporating 11 perfect match and 11 mismatch probes for eachgene target and a rigorous custom probe set summarization tech-nique (9) (see the supplemental material).

Hierarchical clustering of array data from 380 temporally andgeographically diverse E. coli and Shigella isolates representingseveral pathotypes and over 100 serotypes illustrates the broad

Received 4 November 2011 Accepted 21 December 2011

Published ahead of print 30 December 2011

Address correspondence to Christopher A. Elkins, chris.elkins@fda.hhs.gov.

S.A.J., M.L.K., and I.R.P. contributed equally to this article.

Supplemental material for this article may be found at http://aem.asm.org/.

Copyright © 2012, American Society for Microbiology. All Rights Reserved.

doi:10.1128/AEM.07464-11

0099-2240/12/$12.00 Applied and Environmental Microbiology p. 1601–1605 aem.asm.org 1601

on July 11, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

genomic diversity that exists within this species (Fig. 1A). Thebranches of the dendrogram are color coded into eight clusters,and isolates of the O104 serogroup examined in this study fallwithin the diverse cluster 6. The O104:H4 GOS and RG isolatesform a very distinct subgroup within this cluster (Fig. 1B). Simi-larly, the O104:H21 MON isolates are unique and distinct withinthis cluster (Fig. 1B). In the FDA microarray database, the closestneighbor to the German outbreak strain is the O139:H28 ETECstrain E24377A. This is the same close association reflected by thewhole-genome phylogeny by Mellmann et al. (11).

At the tip of the hierarchical tree, the branch lengths corre-spond to the number of gene differences visualized by scatter plotcomparisons of strain pairs where each spot represents the robustmultiarray average (RMA)-summarized probe set intensity (4).The examination of these presence/absence calls for a paired set ofstrains is an effective comparative genome graphic, especiallywhen the outbreak strain is used as the reference (Fig. 2). Whenthe array scatter plot data for isolate pairs are examined, M-GOSand C-GOS are genotypically indistinguishable and have corre-spondingly minimal branch lengths on the tree (Fig. 1B and 2A).Similarly, the two MON isolates have minimal branch lengths andare indistinguishable. In contrast, the RG isolates have numerousdetectable gene differences, as demonstrated by their branchlengths and scatter plot comparisons (Fig. 1B and 2B). In aggre-gate, the pairwise scatter plots display only some 10 to 200 genedifferences among these four O104:H4 strains (Fig. 2A to D).Much larger numbers of gene differences are apparent betweenthe M-GOS strain and the more distant MON1 and E24377A iso-lates (Fig. 2E and F, respectively).

A descriptive analysis of the O104:H4 German strain and thetwo Republic of Georgia strains revealed 253 gene differences (forstrain-specific comparisons, see Table S1 in the supplemental ma-terial). Of these, 97 had functional annotation, with most (�50%)of the annotated differences being phage related. A noteworthydifference in the probe signals was observed for the tetracyclineand mercury resistance probe sets, which were absent in the RG2strain. We followed this observation with PCR amplification todetect six genes (dfrA, sulI, sulII, strA, merA, and tetA) found at theselC insertion hot spot in the genome sequence of the Germanoutbreak strain (for primer sequences, see Table S2 in the supple-mental material). Primer pairs for tetA and merA failed to amplifyPCR products from RG2, but generated amplicons of the correctsizes from M-GOS, C-GOS, and RG1 (data not shown).

Chromosome mapping. Optical mapping provides a rapid,broad overview of the arrangement of a bacterial chromosome,including prominent inversions, insertions, and deletions, thecontent of which can be quickly surmised when comparativelyassessed with reference genomes (10). Bacterial isolates were gen-tly lysed to generate high-molecular-weight DNA that was immo-bilized, digested with BamHI, and stained with fluorescent dye.Restriction fragments were assembled into overlapping contigu-ous regions to created a closed, circular map with a lower limit ofdetection of 750 to 1,000 bp using the Argus optical mappingsystem (OpGen, Inc.; see the supplemental material).

Pairwise map comparisons to the German outbreak strain,M-GOS, revealed differences ranging from 0.4% to 75%. TheGeorgian strains RG1 and RG2 are 0.4% and 1.1% different fromM-GOS, respectively (Fig. 3A and B) and 1% different from eachother (Fig. 3C). Strain 55989, an O104:H4 EAEC isolate shownpreviously (11) to be closely related to M-GOS, was 6% different

Jackson et al.

1602 aem.asm.org Applied and Environmental Microbiology

on July 11, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

(Fig. 3D). In contrast, the optical maps of the O104:H21 MON1,O139:H28 E24377A, and O157:H7 Sakai isolates are 21.5%,30.6%, and 75% different from the M-GOS map, respectively (Fig.3E to G).

While map differences of �10% generally reflect insertionsand deletions (indels), larger map differences (�25%) containsignificant amounts of syntenic variant regions (SVRs), compara-bly sized but unaligned fragment blocks distinct from indels.These unaligned segments usually represent slightly divergent re-striction fragment profiles of otherwise similar DNA segmentsand account for the majority of differences in comparisons ofstrains with map differences greater than about 25%.

Genomic landscape hallmarks. The optical maps of the Ger-man outbreak strain (M-GOS) and Georgian (RG1 and RG2)O104:H4 isolates all contain similar unique insertions at hallmarkwrbA-serX and selC chromosomal loci that differentiate the EAECO104:H4 cluster (our Fig. 3A and see Fig. 3 in reference 11). Theseinsertions, in addition to a number of other differences, can bedistinguished in the comparison of the M-GOS optical map andthe EAEC 55989 in silico map (Fig. 3D). The German and Geor-gian isolates contain a large 146-kbp insertion near the wrbA lo-cus, a site often occupied by lambdoid prophages in other patho-genic E. coli strains. Sequence data from a number of Germanoutbreak isolates show a 61-kbp Shiga toxin 2 gene (stx2)-

FIG 1 (A) Hierarchical clustering of RMA-summarized microarray data employing a database of 380 historical reference and outbreak strains from internal FDAcollections integrated with O104 outbreak strains analyzed in this study. The resulting dendrogram resolves into eight large color-differentiated clusters. (B)Enlargement of cluster 6 region containing O104 isolates. Note branch heights, which are representative of the number of gene differences occurring betweenindividual strains.

FIG 2 Microarray scatter plots of pairwise comparisons demonstrating gene-level differences between strains analyzed in this study.

Rapid Genomic-Scale Analysis of E. coli O104:H4

March 2012 Volume 78 Number 5 aem.asm.org 1603

on July 11, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

containing prophage at wrbA, with the remainder of the 146-kbpinsertion attributable to a tellurite resistance island inserted 30kbp downstream from wrbA at serX. This 86-kbp insertion at serXis similar to but distinct from a tellurite resistance cassette found atserX in O157:H7 strains. The third hallmark is a 48-kbp insertionfound at selC, a common island integration site in several patho-types of E. coli. Sequence data from the German outbreak identifythis region as a multidrug resistance island (17).

Although the two Georgian isolates share unique insertionsfound in the German O104:H4 outbreak strain, they have severalmarquee differences that account for a 1% difference in their op-tical maps. Relative to M-GOS, both RG1 and RG2 have a 14-kbpdeletion near thrW, the site of a number of cryptic prophages inother E. coli strains (Fig. 3A and B, arrows). In addition, there arethree chromosomal differences between RG1 and RG2 (Fig. 3C,arrows). RG2 contains a 49-kbp insertion near potB and a 17-kbpinsertion near argW; both loci are common prophage integrationsites in other pathogenic E. coli strains. The third difference is a9.4-kbp deletion at selC in RG2. Notably, this partial deletion ofthe larger selC multidrug resistance island correlates with arrayand PCR data for the tetracycline and mercury resistance genesmissing in RG2.

Finally, although MON1 is stx2 positive (7), there is no inser-

tion at the wrbA locus comparable to the insertions observed inthe M-GOS and RG maps. Instead, MON1 contains a 53-kbp in-sertion near yehV, a common locus for stx phages and a likely sitefor the stx2 genes in this strain (Fig. 3E).

Conclusions. The 2011 German O104:H4 outbreak strain rep-resents a chimera with a collection of virulence attributes that aresimilar to those of the 2009 Georgian strains at key chromosomalloci, unlike EAEC 55989. Thus, serotyping information alone pro-vides little to no insight into strain relatedness nor fundamentalinformation about a strain’s virulence attributes, despite its use asa primary characteristic in outbreaks. To emphasize the point,phylogenetic studies have shown that the independent acquisitionof virulence genes creates pathotypes that extend across the E. coliphylogenetic landscape (15). Consequently, development ofgroup-, clade-, or gene-specific assays for prototypic classes ofpathogenic E. coli must be balanced by the need to delineate pro-ductive virulence combinations for surveillance across previouslyplacid serogroups (1, 5).

Our comparative genomic studies demonstrate rapid identifi-cation of similar and dissimilar isolates, informative differencemeasures, as well as scoring of novel combinations of virulencefactors and genes in prophages and pathogenicity islands. Thepower of our analyses involves comparative database profiling.

FIG 3 Pairwise alignments of optical and in silico sequence-based reference maps. Aligned restriction fragments are green, and unaligned fragments are white.Locations of the wrbA, serX, and selC islands are indicated by the red, blue, and orange boxes, respectively. The location of the yehV island in the O104:H21 MON1strain is shown by the yellow box. The stx1 prophage at yehV and the locus of enterocyte effacement (LEE) island at selC in the O157:H7 Sakai strain are depictedas purple and light blue boxes, respectively. Regions of difference among the M-GOS, RG1, and RG2 O104:H4 strains described in the text are indicated by thearrows.

Jackson et al.

1604 aem.asm.org Applied and Environmental Microbiology

on July 11, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

Moreover, when employed in concert, the current value-addedbenefits of these two analyses also involve speed and instant anno-tation of genetic content. We note that in addition, the rapid gen-eration of optical maps can be used to facilitate closure of whole-genome sequencing efforts by providing scaffolds to anchorsequence reads as a powerful tool for detecting misassemblies (8,16). Regardless of database development (robustness) and main-tenance for any high-resolution format, these techniques shouldprove independently useful as companions for whole-genome se-quencing pipelines to save time, effort, and expense.

ACKNOWLEDGMENTS

We thank the Massachusetts Department of Public Health (MDPH) StateLab for a clinical isolate of the German sprout-associated outbreak strain(M-GOS), in addition to the CDC as an independent source for theO104:H4 MDPH strain (C-GOS) and strains from the Republic of Geor-gia (RG1 and -2). We also thank Peter Feng (FDA-CFSAN) for the O104:H21 isolates (MON1 and -2) and for being a direct connection to theMDPH for the M-GOS isolate. Finally, we acknowledge Joan Shields(FDA-CFSAN) for facilitating timely transfer of the O104:H4 strains fromthe CDC.

Seminal funding and support were received from the National Biofo-rensics Analysis Center and the Department of Homeland Security formicroarray and optical mapping program development; we acknowledgeJoseph E. LeClerc (FDA-CFSAN) for contributions to this work. Thisstudy was funded by the Center for Food Safety and Applied Nutrition.

The views presented in this article do not necessarily reflect those ofthe Food and Drug Administration.

REFERENCES1. Bae WK, et al. 2006. A case of hemolytic uremic syndrome caused by

Escherichia coli O104:H4. Yonsei Med. J. 47:437– 439.2. Bernier C, Gounon P, Le Bouguénec C. 2002. Identification of an ag-

gregative adhesion fimbria (AAF) type III-encoding operon in enteroag-gregative Escherichia coli as a sensitive probe for detecting the AAF-encoding operon family. Infect. Immun. 70:4302– 4311.

3. Bielaszewska M, et al. 2011. Characterization of the Escherichia coli strainassociated with an outbreak of haemolytic uraemic syndrome in Ger-many, 2011: a microbiological study. Lancet Infect. Dis. 11:671– 676.

4. Bolstad BM, Irizarry RA, Astrand M, Speed TP. 2003. A comparison of

normalization methods for high density oligonucleotide array data basedon variance and bias. Bioinformatics 19:185–193.

5. Bosilevac JM, Koohmaraie M. 2011. Prevalence and characterization ofnon-O157 Shiga toxin-producing Escherichia coli isolates from commer-cial ground beef in the United States. Appl. Environ. Microbiol. 77:2103–2112.

6. Brzuszkiewicz E, et al. 2011. Genome sequence analyses of two isolatesfrom the recent Escherichia coli outbreak in Germany reveal the emergenceof a new pathotype: entero-aggregative-haemorrhagic Escherichia coli(EAHEC). Arch. Microbiol. 193:883– 891.

7. Feng P, Weagant SD, Monday SR. 2001. Genetic analysis for virulencefactors in Escherichia coli O104:H21 that was implicated in an outbreak ofhemorrhagic colitis. J. Clin. Microbiol. 39:24 –28.

8. Giongo A, Tyler HL, Zipperer UN, Triplett EW. 2010. Two genomesequences of the same bacterial strain, Gluconacetobacter diazotrophicusPAl 5, suggest a new standard in genome sequence submission. Stand.Genomic Sci. 15:309 –317.

9. Jackson SA, Patel IR, Barnaba T, Leclerc JE, Cebula TA. 2011. Investi-gating the global genomic diversity of Escherichia coli using a multi-genome DNA microarray platform with novel gene prediction strategies.BMC Genomics 12:349.

10. Kotewicz ML, Mammel MK, LeClerc JE, Cebula TA. 2008. Opticalmapping and 454 sequencing of Escherichia coli O157:H7 isolates linked tothe US 2006 spinach-associated outbreak. Microbiology 154:3518 –3528.

11. Mellmann A, et al. 2011. Prospective genomic characterization of theGerman enterohemorrhagic Escherichia coli O104:H4 outbreak by rapidnext generation sequencing technology. PLoS One 6:e22751.

12. Mossoro C, et al. 2002. Chronic diarrhea, hemorrhagic colitis, andhemolytic-uremic syndrome associated with HEp-2 adherent Escherichiacoli in adults infected with human immunodeficiency virus in Bangui,Central African Republic. J. Clin. Microbiol. 40:3086 –3088.

13. Rasko DA, et al. 2008. The pangenome structure of Escherichia coli:comparative genomic analysis of E. coli commensal and pathogenic iso-lates. J. Bacteriol. 190:6881– 6893.

14. Rasko DA, et al. 2011. Origins of the E. coli strain causing an outbreak ofhemolytic-uremic syndrome in Germany. N. Engl. J. Med. 365:709 –717.

15. Reid SD, Herbelin CJ, Bumbaugh AC, Selander RK, Whittam TS. 2000.Parallel evolution of virulence in pathogenic Escherichia coli. Nature 406:64 – 67.

16. Reslewic S, et al. 2005. Whole-genome shotgun optical mapping of Rho-dospirillum rubrum. Appl. Environ. Microbiol. 71:5511–5522.

17. Rohde H, et al. 2011. Open-source genomic analysis of Shiga toxin-producing E. coli O104:H4. N. Engl. J. Med. 365:718 –724.

18. Touchon M, et al. 2009. Organised genome dynamics in the Escherichiacoli species results in highly diverse adaptive paths. PLoS Genet.5:e1000344.

Rapid Genomic-Scale Analysis of E. coli O104:H4

March 2012 Volume 78 Number 5 aem.asm.org 1605

on July 11, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from