OUP user menu

Comparison between a bovine and a human enterohaemorrhagic Escherichia coli strain of serogroup O26 by suppressive subtractive hybridization reveals the presence of atypical factors in EHEC and EPEC strains

Marjorie Bardiau, Bernard Taminiau, Jean-Noël Duprez, Sabrina Labrozzo, Jacques G. Mainil
DOI: http://dx.doi.org/10.1111/j.1574-6968.2012.02542.x 132-139 First published online: 1 May 2012


Enterohaemorrhagic Escherichia coli (EHEC) strains are responsible for food poisoning in humans in developed countries via consumption of vegetal and animal foodstuffs contaminated by ruminant feces. The clinical conditions caused by EHEC strains vary from undifferentiated diarrhea to hemorrhagic colitis with, in a few cases, the appearance of the hemolytic uremic syndrome, which can lead to death. Most EHEC strains can be found in the gut of healthy ruminants, but some of the strains, belonging to O26, O111, O118 serogroups, for example, are also responsible for digestive disorders in calves. The aim of this research was to study the genomic differences between two EHEC strains of serogroup O26 isolated from a young calf and a human with diarrhea, to identify specific sequences of the bovine strain that could be implicated in initial adherence or host specificity. No sequence implicated in host specificity was found during our study. Finally, several factors, not usually present in EHEC strains of serogroup O26, were identified in the bovine strain. One of them, the PAI ICL3 locus initially presented as a marker for LEE-negative VTEC strains, was found in 11.3% of EPEC and EHEC strains.

  • enterohaemorrhagic Escherichia coli
  • suppressive subtractive hybridization
  • serogroup O26
  • PAI ICL3


In humans, enterohaemorrhagic Escherichia coli (EHEC) is responsible for the production of diarrhea, generally accompanied by hemorrhagic colitis with, in a few percent of cases, the occurrence of renal sequelae (hemolytic uremic syndrome), which can lead to death. EHEC strains were recognized as a distinct class of pathogenic E. coli in 1983 after two outbreaks in the United States (Wells et al., 1983). Today, they represent a significant problem for public health in developed countries. Indeed, large outbreaks are caused by EHEC strains (Nataro & Kaper, 1998), and transmission often occurs via consumption of vegetal and animal foodstuffs contaminated by feces of adult ruminants (mainly cattle), which can be healthy carriers (Caprioli et al., 2005). The most common EHEC serotype is O157:H7, which causes disease worldwide, but other serogroups such as O26, O111, and/or O103 are also of high epidemiological importance in some countries (Brooks et al., 2005; Bettelheim, 2007). In the veterinary field, different serogroups of EHEC strains (O5, O26, O111, O118, for example) are directly associated with diarrhea in 2-week to 2-month-old calves (Moxley & Francis, 1986; Stordeur et al., 2000; Hornitzky et al., 2005). The consequences are economic losses owing to a delay in growth and weakness of the calves.

Some pathogenic E. coli are host specific, based upon the production of host-specific properties, in particular adhesins and colonization factors (for example, human typical EPEC, rabbit-EPEC, and porcine-VTEC). However, the actual situation about the host specificity regarding those EHEC serogroups (e.g. O26 and O111) infecting both humans and young calves, and present in healthy adult ruminants, is unknown: Do some isolates possess some degree of host specificity or can all isolates in fact infect all the hosts? As with host-specific pathogenic E. coli, the basis of any host specificity of those EHEC strains may be related to the production of specific colonization factors, although such adhesins of EHEC strains have not yet been identified (Bardiau et al., 2009).

The aim of this study was (1) to explore the genomic differences, using suppressive subtractive hybridization (SSH), between two EHEC strains of serogroup O26, one isolated from a young calf and the other isolated from a human with diarrhea, to identify specific sequences of the bovine strain; (2) to analyze the bovine strain-specific sequences regarding their potential implication in adherence to epithelial cells; and (3) to study the prevalence of these strain-specific sequences in a collection of human and bovine EHEC and EPEC strains.

Materials and methods

Bacterial strains

Subtractive suppressive hybridization (SSH) was performed between the bovine EHEC strain 4276 of serogroup O26 isolated in Ireland from a diarrheic calf (Kerr et al., 1999) and the human EHEC strain 11368 of serogroup O26 isolated in Japan from a human suffering from diarrhea (Ogura et al., 2009). The distribution of the specific sequences was investigated in additional EHEC (n = 44) and EPEC (n = 27) strains of serogroup O26 isolated from humans (n = 27) and from cattle (n = 44). Most of the strains have been described previously (Szalo et al., 2004; Bardiau et al., 2009), and their characteristics are described in the supplemental Table S1.

Pulsed field gel electrophoresis (PFGE)

PFGE was performed as already described (Cobbaut et al., 2009; Ooka et al., 2009) on most of the tested strains. In brief, bacterial cells were embedded in 1.8% Certified Low Melt Agarose (Bio-Rad Laboratories, Inc., Tokyo, Japan), lysed with a buffer containing 0.2% sodium deoxycholate, 0.5% N-lauroylsarcosine, and 0.5% Brij-58, and treated with 100 µg mL−1 proteinase K. XbaI-digested genomic DNA was separated using CHEF MAPPER (Bio-Rad Laboratories, Inc.) with 1% Pulsed Field Certified Agarose (Bio-Rad Laboratories, Inc.) at 6.0 V cm−1 for 22 h and 18 min with pulsed times ranging from 47 to 44.69 s. Size of each DNA band was estimated by Biogene (Vilber Lourmat, France). The banding patterns were analyzed using the Dice coefficient, with an optimization and position tolerance of 1%. Dendrograms were prepared by the unweighted-pair group method using arithmetic average algorithm (UPGMA).


Genomic DNA was extracted from E. coli strain 4276 and E. coli strain 11368 using the cetyltrimethylammonium bromide procedure described by Ausubel et al. (1994). Subtractive hybridization was carried out using the PCR-Select Bacterial Genome Subtractive kit (Clontech) as recommended by the manufacturer. The bovine EHEC strain 4276 was the tester, and the human EHEC strain 11368 was the driver. The PCR products obtained were cloned into the pGEM-T Easy Vector System (Promega) and transformed into E. coli JM109. The recombinant clones were plated onto LB plates containing ampicillin (100 µg mL−1), 0.2 mM IPTG (isopropyl-β-d-thiogalactopyranoside), and 40 µg X-Gal mL−1 (5-bromo-4-chloro-3-indolyl-β-d-galactopyranoside). White colonies containing recombinant plasmids with inserts were picked up, grown overnight at 37 °C in LB broth with ampicillin (100 µg mL−1), and stored in a freezer (−20 °C) until further use.

DNA sequencing

The plasmid inserts were amplified by PCR with specific primers (nested primers 1 and 2R from the Clontech protocol), and the DNA fragments were purified using the NucleoSpin Extract II kit (Macherey-Nagel) according to the manufacturer's instructions. Sequencing of the two DNA strands was performed by the dideoxynucleotide triphosphate chain termination method with a 3730 ABI capillary sequencer and a BigDye Terminator kit version 3.1 (Applied Biosystems) at the GIGA (Groupe Interdisciplinaire de Génoprotéomique Appliquée, University of Liège, Belgium). Sequence analysis was performed using Vector NTI 10.1.1 (Invitrogen). DNA sequences were further examined for homology with the National Center for Biotechnology Information (NCBI) blastn and blastx programs (http://www.ncbi.nlm.nih.gov/BLAST/). The expectation value of 0.001 was chosen as the cutoff.

DNA colony hybridization

Several EHEC strain 4276–specific sequences were chosen for extended analysis. Their distribution was searched for in the collection of 71 bovine and human EHEC and EPEC strains by DNA colony hybridization at 65 °C on Whatman 541 paper filters (Whatman), as previously described (Mainil et al., 1997). The DNA probes were derived by PCR from plasmid inserts obtained with SSH. The DNA probe fragments were purified using the NucleoSpin Extract II kit (Macherey-Nagel), according to the manufacturer's instructions, and labeled with α32P-dCTP (Perkin-Elmer) by random priming using the Ready-To-Go dCTP-labeling beads (Amersham Biosciences). Labeled DNA probes were purified with Microcon-YM30 spin columns (Millipore).

PCR reactions

All primers and PCR conditions used in this study have been described previously (China et al., 1996; Shen et al., 2004; Durso et al., 2005) (Supporting Information, Table S2). DNA extraction was carried out by a boiling method as described previously by China et al. (1996). For the PCR, the following mixture was used: 1 U of Taq DNA polymerase (New England Biolabs), 5 µL of 2 mM deoxynucleoside triphosphates, 5 µL of 10× ThermoPol Reaction Buffer (20 mM Tris–HCl (pH 8.8, 25 °C), 10 mM KCl, 10 mM (NH4)2S04, 2 mM MgS04, 0.1% Triton X-100), 5 µL of each primer (10 µM), and 3 µL of a DNA template in a total volume of 50 µL.

Statistical analysis

A Fisher's exact test was performed to assess statistical differences (P < 0.01).



PFGE profiles were obtained for 60 of the 73 tested strains. Others strains did not present any restriction profile for XbaI or could not be tested. The 60 distinct electrophoresis profiles were used for dendrogram construction (Fig. S1). The dendrogram showed five clusters, assuming a cutoff of 45% of similarity. When a cutoff of over 80% of similarity was adopted, 38 different clusters were found, indicating the high genetic variability among the strains.

Identification of the bovine EHEC strain 4276–specific genes in the subtractive library

A total of 1920 clones resulting from the SSH process were obtained, of which 772 were randomly sequenced, resulting in 296 contigs after removal of redundant sequences. The specificity of the contigs to the bovine EHEC strain (strain 4276) was determined by a blastn search with the human EHEC strain (strain 11368) genome sequenced by Ogura et al. (2009). Of the 296 nonredundant DNA contigs, 115 contained genes different from those of the human EHEC strain (strain 11368).

Analysis of the bovine EHEC strain 4276–specific genes

BLASTN and BLASTX against the GenBank were searched for the 115 contigs specific to the bovine strain (Table 1 and Table S3). Several groups of genes were revealed by more than one clone: colicin resistance genes, multiple antibiotic resistance region from Salmonella enterica, phages P1 and P7, pathogenicity island (termed PAI ICL3) described in the VTEC O113:H21 E. coli CL3 (containing putative adhesins and hemolysins), genes from the genomic islands GEI 3.21 described in E. coli O111:H−, transposase from Enterobacter cloacae, E. coli and Acinetobacter baumanii, predicted type I restriction-modification enzyme from E. coli 0127:H6 E2348/69, DEAD/DEAH box helicase from Nitromonas europea, SNF2 family helicase from E. coli strain E24377A, plasmid pO111_2 from E. coli O111:H−, and plasmid pSMS35_8 from E. coli SMS-3-5. BLASTN revealed six sequences that are not homologous to any annotated DNA sequences in GenBank. The other sequences were detected in only one clone and corresponded to genes specific to Klebsiella pneumoniae, Pseudomonas aeruginosa, Citrobacter rotendium, Shigella sonnei, Erwinia sp., Desulfurispirillum indicum, Dickeya zeae, Pantoea ananatis, and several strains of E. coli.

View this table:
Table 1

Results of the blastn against the GenBank searched for the 115 contigs specific to the bovine strains. Sequences in bold were chosen for further characterization

FunctionNumber of nonredundant and specific contigsblastn resultsSpecies
Antibiotic resistance9Colicin resistanceEscherichia coli
3Multiple antibiotic resistance regionSalmonella enterica, Klebsiella pneumoniae
Mobile functions7TransposaseEnterobacter cloacae, E., Acinetobacter baumanii
1ExcisionaseE. coli
Genomic island1Genomic island GEI1.94E. coli
1Genomic island AGI-5E. coli
7Genomic island GEI3.21E. coli
Unknown function14Hypothetical proteinE. coli, Pseudomonas aeruginosa, Citrobacter rodentium
6No homology
Adherence related7Pathogenicity island I (PAI ICL3)E. coli
1Putative hemolysin/hemagglutininC. rodentium
1espPE. coli
1tonBShigella sonnei
Metabolism3Predicted type I restriction-modification enzyme, S subunitE. coli
2N-6 DNA methylaseDesulfurispirillum indicum, Dickeya zeae
1Galactosyl transferaseErwinia sp.
4DEAD/DEAH box helicaseNitrosomonas europaea, P. aeruginosa
3SNF2 family helicaseE. coli
2ABC transporterDickeya zeae, Nitrosomonas europaea
Phage related8Enterobacteria phage P7 or P1Enterobacteria phage P7 or P1
1Putative tail fiber assembly proteinS. sonnei
1Putative phage repressor proteinE. coli
Other1FhaBPantoea ananatis
1avrAE. coli
23Plasmid pO111_2E. coli
1Plasmid pCRP3C. rodentium
3Plasmid pSMS35_8E. coli
1Plasmid pHUSEC41-1E. coli
1Plasmid pO145-NME. coli

Distribution of specific sequences in a collection of EHEC and EPEC isolates

Several sequences (in bold in Table 1 and Table S3) were chosen for further characterization based upon the frequency of the contigs in the subtractive library or upon the putative involvement in adherence to the eukaryotic cells or in host specificity: genes from PAI ICL3, four sequences with no homology, genes from P1 and P7 phages, genes from genomic island GEI 3.21, hypothetical proteins from E23477A strain, DEAD/DEAH box helicase from Nitromonas sp., genes from E. coli O111:H− strain 11128, transposase from A. baumanii., ABC transporter from D. zeae, and avrA genes from E. coli strain CB769. The regions of DNA homologous to that previously identified in the subtractive library were searched for in EHEC and EPEC strains of serogroup O26 isolated from human and from cattle using DNA colony hybridization (Table 2) or using specific PCR for PAI ICL3 locus (Table 3).

View this table:
Table 2

Distribution of specific sequences resulting from SSH analyses

Number of strains
Number and percentage of strains found positive for
LG100GammaD11LG53LG31LG106LG88LG69LG109LG18Epsilon G6Alpha F3Dzeta H6LG28LG45LG12Nu G3LG2Espilon E1LG92LG16
No. of tested strainsNo homologyNo homologyNo homologyNo homologyEnterobacteria phage P7Recombination enhancement functionPlasmid pO111_2Plasmid pO111_2Putative phage repressor proteinHypothetical proteinHypothetical proteinHypothetical proteinABC transporterDEAD/DEAH box helicaseDEAD/DEAH box helicaseavrAGenomic island GEI3.21Genomic island GEI3.21Predicted type I restriction-modification enzyme, S subunitTransposase
Total no. of strains753 (4%)6 (8%)52 (69%)2 (3%)14 (19%)15 (20%)14 (19%)5 (7%)12 (16%)0 (0%)2 (3%)23 (31%)1 (1%)1 (1%)4 (5%)36 (48%)24 (32%)16 (21%)12 (16%)23 (31%)
Bovine strains443 (7%)4 (9%)31 (70%)1 (2%)9 (20%)10 (23%)10 (23%)1 (2%)7 (16%)0 (0%)1 (2%)11 (25%)1 (2%)1 (2%)4 (9%)25 (57%)15 (34%)11 (25%)3 (7%)13 30%)
Human strains270 (0%)2 (7%)21 (78%)1 (4%)5 (19%)5 (19%)4 (15%)4 (15%)5 (19%)0 (0%)0 (0%)12 (44%)0 (0%)0 (0%)0 (0%)11 (41%)9 (33%)5 (19%()9 (33%)10 (37%)
EPEC strains270 (0%)2 (7%)19 (70%)1 (4%)2 (7%)1 (4%)2 (7%)1 (4%)2 (7%)0 (0%)1 (4%)6 (22%)0 (0%)0 (0%)0 (0%)8 (30%)15 (56%)12 (44%)4 (15%)7 (26%)
EHEC strains443 (7%)4 (9%)33 (75%)1 (2%)12 (27%)14 (32%)12 (27%)4 (9%)10 (23%)0 (0%)1 (2%)17 (39%)1 (2%)1 (2%)4 (9%)28 (64%)9 (20%)4 (9%)8 (18%)16 (36%)
View this table:
Table 3

Distribution of the genes carried by the PAI ICL3 locus in the eight positive strains

Similar protein (% identity)Genes
Z1635, unknown protein, E. coli EDL933 (97)OI-48Z1635++++++++
Z1636, unknown protein, E. coli EDL933 (96)Z1636++++++++
Z1637, unknown protein, E. coli EDL933 (95)Z1637++++++++
YPO2491, putative hemolysin activator, Y. pestis CO92 (64)GSIS1++++++++
RS02573, putative hemolysin activating-like protein, R. solanacearum (57)S2++++++++
YPO2490, putative hemolysin (53)S3++++++++
YPO0599, putative adhesin, Y. pestis CO92 (50)
YPO2490, putative hemolysin (41)GSIIS4++++++++
YPO0599, putative adhesin, Y. pestis CO92 (39)
YPO0599, putative adhesin, Y. pestis CO92 (83)S5++++
Y2435, putative transposase, Y. pestis KIM (38)S6++++++
TnpA, transposase, P. syringae (60)S7
FN0835, hypothetical protein, F. nucleatum ATCC 25586 (27)S8
YozI, unknown protein, B. subtilis (32)S9++
Z4322, unknown protein, E. coli EDL933 (94)OI-122S10
Z4321, unknown protein, E. coli EDL933 (98)S11++++
Orf1, similarity with helicase, S. enterica (40)S12+
ST0071, hypothetical esterase Sulfolobus tokodaii (30)S13+
Y2679, hypothetical protein, Y. pestis KIM (39)S14+++
Z1640, unknown protein, E. coli EDL933 (90)
Z1641, unknown protein, E. coli EDL933 (96)OI-48S15++++++++
Z1642, unknown protein, E. coli EDL933 (99)
Z1643, unknown protein, E. coli EDL933 (97)Z1643++++++++
Z1644, unknown protein, E. coli EDL933 (98)Z1644++++++++
  • B, bovine; H, human.

Statistical analyses were performed to assess differences in the presence of the fragments according to host specificity (human or bovine) and/or pathotype (EHEC or EPEC). Two sequences, both homologous to the genomic island GEI 3.21 from E. coli O111:H−, were statistically associated with EPEC strains in comparison with EHEC strains. One of the fragments homologous to P1 phage was statistically associated with EHEC in comparison with EPEC. The sequence homologous to the predicted type I restriction-modification enzyme from E. coli O127:H6 strain E2348/69 was statistically associated with strains isolated from humans in comparison with strains isolated from bovines. All the other fragments were associated with neither pathotype nor host.

Distribution of the PAI ICL3 locus in human and bovine EPEC and EHEC strains

Shen et al. (2004) first described the PAI ICL3 locus in the O113:H21 VTEC strain CL3. PAI ICL3 is a hybrid genomic region composed of genes similar to EDL933 (serotype O157:H7) O islands 122 and 48, Yersinia pestis, Ralstonia solanacearum, Pseudomonas syringae, Fusobacterium nucleatum, Bacillus subtilis, S. enterica, and Sulfolobus tokodaii (Table 3). To date, PAI ICL3 has been detected only in eae-negative VTEC strains associated with diseases in humans and never in any other pathogenic or commensal E. coli, and it may therefore be used as a new marker for those strains (Girardeau et al., 2009). As several genes of PAI ICL3 have been identified here in the bovine EHEC strain 4276 of serogroup O26, their distribution was studied with specific PCRs in the collection of human and bovine EHEC and EPEC strains.

Eight strains (three human EPEC and five human and bovine EHEC strains) were found to be positive for several PCRs targeting different genes of the PAI ICL3 locus (Table 3). According to their PFGE pattern, these eight strains are not closely related. Indeed, they are present in the five clusters revealed by the PFGE dendrogram with a similarity of 45%, suggesting that these genes were horizontally acquired. No statistical difference was associated with the pathotype and/or the host origin (P < 0.01). This genomic island can in fact be divided into four parts: two genomic segments (GS-I inserted and GS-II including two genes of OI-122) bordered by OI-48 segments either side (Shen et al., 2004). The eight strains were tested positive here with the PCRs for the three genes of GS-I and for all six genes of the two OI-48 segments. To verify whether Z1640 gene is intact or not, we performed two PCRs: one PCR targeting the Z1640-1 and Z1640-3 sequences (using Z1640-F and Z1640-R primers) and one PCR targeting the Z1640-1 and S1 sequences (using Z1640-F and S1-bis-R primers). The eight strains were positive only with the Z1640/S1 PCR. On the other hand, only the S4 gene of GS-II was detected in all eight strains, while the other genes (including S10 and S11 genes of OI-122) were detected in none to six strains only.


Several serogroups of EHEC strains (e.g. O5, O26, O111, O118) can infect both humans and calves and can also be found in healthy cattle. Factors implicated in host specificity have been identified for some other pathogenic E. coli strains, but not for EHEC strains. Such factors could be based on proteins intervening in the colonization stage (adhesins, for example). Therefore, we wanted to explore the genomic differences between a bovine EHEC strain of serogroup O26 and a human EHEC strain of serogroup O26 using SSH to identify specific sequences of the bovine strain. This study aimed to explore the potential implication in initial adherence or host specificity of the specific sequences.

In the SSH library, we obtained 115 unique fragments that were specific to the bovine strain. These fragments include sequences with homology to genes or pathogenicity islands (PAIs) present only in other specific E. coli pathotypes (e.g. VTEC) or other species (e.g. Klebsiella, Nitromonas), which are not known to be present in EHEC strains of serogroup O26. This heterogeneity supports the hypothesis of a horizontal acquisition of genomic regions from other pathogenic bacteria (Brzuszkiewicz et al., 2009; Juhas et al., 2009; Kelly et al., 2009). Moreover, it reflects the genomic plasticity of EHEC and/or E. coli strains. This finding supports the hypothesis of Mokady et al. (2005), suggesting that this variation in the genome contents of E. coli could indicate that its evolutionary strategy tends to create a mixed assortment of virulence factors coming from various pathogenic strains. This combination leads to a unique set of such factors, which helps the bacteria to better survive.

The PAI ICL3 locus, first described by Shen et al. (2004) in the VTEC O113:H21 E. coli CL3, was found in 11.3% of the tested EHEC and EPEC strains of serogroup O26. These results are surprising when compared to those obtained by Girardeau et al. (2009), suggesting that PAI ICL3 is unique to LEE-negative VTEC strains and that this locus thus provides a new marker for such strains. We have reported here that the locus could also be present in eae-positive strains belonging to a major serogroup involved in human diseases. Girardeau et al. (2009) have suggested that PAI ICL3 used to be present in most E. coli pathotypes but that many of these pathotypes have undergone extensive deletions [probably via homologous recombination between insertion sequences (IS) elements, which removed almost the entire locus]. We can assume that our positive strains were not deleted for this locus. Another possible explanation is that these strains have recently acquired the PAI ICL3 locus via horizontal transfer, which hypothesis is supported by the fact that the PAI ICL3-positive strains are not closely related.

Concerning host specificity, only one sequence appears to be statistically specific to human strains in comparison with bovine strains. Nevertheless, this sequence is only present in a few strains (7% of bovine strains and 33% of human strains) and therefore could not represent a host-specific marker. Moreover, three sequences were statistically associated with the pathotype (EHEC or EPEC), but these sequences were not present in more than half of the EPEC strains. However, host-specific factors could, perhaps, not be detected by SSH for one of the following reasons: (1) the subtraction is nonexhaustive and host-specific factors were not detected; (2) this host specificity is not based on the presence/absence of specific factors/genes; and (3) there is no host specificity.

In conclusion, our findings support the hypothesis of the acquisition of genomic regions from other pathogenic bacteria (E. coli or others) by horizontal transfers and reflect the genomic plasticity of EHEC or even E. coli strains. This variation in the genome contents of E. coli, suggested as a evolutionary strategy to better survive by Mokady et al. (2005), could lead to serious problems in public health and to the emergence of highly virulent new strains if one strain could acquire several strong virulence systems from different pathogenic bacteria, as it was dramatically illustrated by the 2011 Shiga toxin–producing E. coli O104:H4 German outbreak (Denamur, 2011; Rasko et al., 2011).

Supporting Information

Additional Supporting Information may be found in the online version of this article:

Fig. S1. Dendrogram of 60 EPEC and EHEC strains constructed by PFGE data (UPGMA).

Table S1. Origin, pathotypes and serotypes of the E. coli strains of serogroup O26 tested for the distribution of specific sequences of the bovine EHEC strain 4276 (n.i., no information; w, week; m, month; y, year).

Table S2. Primers used in this study.

Table S3. BLASTN and BLASTX results of the 115 contigs specific to the bovine strain.


During this study, Marjorie Bardiau was a PhD fellow of the ‘Fonds pour la formation à la Recherche dans l'Industrie et dans l'Agriculture’ (FRIA). This study was funded by the Federal Public Service of Health, Food Chain Safety and Environment (contract RF 6172), the European Network of Excellence EADGENE (European Animal Disease Genomics Network of Excellence for Animal Health and Food Safety) for the sequencing, and a grant ‘Crédits aux chercheurs’ FNRS (Fonds de Recherche Scientifique) 2008, no. 1363.


View Abstract