Journal of Applied Microbiology ISSN 1364-5072

ORIGINAL ARTICLE

Investigating possible association between multidrug

resistance and isolate origin with some virulence factors of
Escherichia coli strains isolated from infant faeces and
fresh green vegetables
R.N. Haddadin1 , A.M. Assaf1, A. Homsi1, P.J. Collier2 and A. Shehabi3
1 School of Pharmacy, The University of Jordan, Amman, Jordan
2 Arduus ad Solem Consultancy, Manchester, UK
3 School of Medicine, The University of Jordan, Amman, Jordan

Keywords
Escherichia coli, ERIC-PCR, heat-labile
enterotoxin, heat-stable enterotoxin, human
faecal isolates, multidrug resistance, plant
isolates, siderophores, virulence factors.
Correspondence
Randa N. Haddadin, Department of Pharma-
ceutics and Pharmaceutical Technology,
School of Pharmacy, The University of Jordan,
Amman, Jordan.
E-mail: r_haddadin@ju.edu.jo
2019/0074: received 11 January 2019, revised
13 April 2019 and accepted 25 April 2019
doi:10.1111/jam.14296
Abstract
Aims: In this study, the association between multidrug resistance (MDR) and
the expression of some virulence factors were evaluated in Escherichia coli
strains isolated from infant faeces and fresh green vegetables. The effect of
isolate origin on associated virulence factors was evaluated. In addition, genetic
fingerprinting of a sample of these isolates (10 isolates from each group) was
studied in order to detect any genetic relatedness among these isolates.
Methods and Results: Escherichia coli isolates were divided into four groups
based on their origin (human faeces or plant) and their antibiotic resistance
(multiresistance or susceptible). PCR was used to investigate heat-labile and
heat-stable enterotoxin genes, and four siderophore genes (aerobactin,
enterobactin, salmochelin and yersiniabactin). Genetic fingerprinting of the
isolates was performed using enterobacterial repetitive intergenic consensus
PCR. Siderophore production was measured by a colorimetric method. Biofilm
formation was evaluated by a crystal violet assay. The results of the study
showed that the expression of MDR is not significantly associated with an
increase in these virulence factors or with biofilm formation. However, the
origin of isolates had a significant association with siderophore gene
availability and consequently on the concentrations of siderophores released.
Genetic fingerprinting indicated that human and plant isolates have the same
clonal origin, suggesting their circulation among humans and plants.
Conclusion: Antibiotic-susceptible strains of E. coli may be as virulent as
MDR strains. Results also suggest that the environment can play a potential
role in selection of strains with specific virulence factors.
Significance and Impact of the Study: Antibiotic-susceptible isolates of
Escherichia coli from plant or human origin can be as virulent as the multidrug
resistance (MDR) ones. Genetic relatedness was detected among the isolates of
plant and human origin, indicating the circulation of these bacteria among
human and plants. This could imply a potential role for environmental
antimicrobial resistant bacteria in human infection.

intestines and usually remains there harmlessly through-

Introduction
Escherichia coli is the most predominant facultative anaer-
obic Gram-negative bacteria in the human colon. Within
a few hours of birth, E. coli colonizes the infants’
out life (Nataro and Kaper 1998). In humans and warm-
blooded animals, E. coli lives in gut commensals, where a
mutual beneficial association with the hosts exists. In
addition, E. coli is frequently found in environments such

88 Journal of Applied Microbiology 127, 88--98 © 2019 The Society for Applied Microbiology
R.N. Haddadin et al.
as water, soil and vegetables as a result of faecal contami-
nation (Shehabi et al. 2006; Burjaq and Shehabi 2013). In
fact, plants and vegetables are increasingly acknowledged
as secondary reservoirs of E. coli (Ingham et al. 2004).
In contrast, E. coli is among the most common oppor-
tunistic pathogens of animals and humans. It is responsi-
ble for a wide spectrum of diseases ranging from limited
mucosal surface infections to disseminated systemic infec-
tions. The ability of E. coli to cause infections relies on
the expression of several virulence factors. Among these
are fimbriae, colonization factors, adhesins, capsule, hae-
molysin, siderophores, toxins, biofilm formation and
others (Nataro and Kaper 1998; Allocati et al. 2013).
Based on the type of virulence factors present and the
clinical symptoms of the infected host, E. coli are classi-
fied into several pathogenic types. One of these classes is
the enterotoxigenic E. coli (ETEC), which is the most
common cause of travellers’ diarrhoea. The disease ranges
from mild to severe watery diarrhoea (Allocati et al.
2013). It is caused by the secretion of enterotoxins, as
either heat-labile (LT) and/or heat-stable (ST) enterotox-
ins, in the intestinal lumen of the patient (Nataro and
Kaper 1998).
Apart from E. coli virulence, multidrug resistance
(MDR) is emerging in these bacteria. The widespread dis-
tribution of MDR strains is becoming a global health
concern. In addition to diseased hosts and healthcare set-
tings, MDR E. coli strains can be detected in healthy peo-
ple, animals, plants, surface- or ground-water and in the
environment (Shehabi et al. 2006; Burjaq and Shehabi
2013; Su et al. 2016). Infections by MDR E. coli can be
difficult to treat and may pose a real threat to public
health. However, there are few studies which have investi-
gated the spread of virulence factors in association with
antimicrobial resistance determinants.
In this study, we investigated the possibility of an asso-
ciation between MDR and the expression of some viru-
lence factors of E. coli. The E. coli isolates were collected
from two distinct sources: human infants, considered as
primary reservoirs of E. coli, and fresh green vegetables.
Infants were selected in this study, since they are rela-
tively newly colonized by bacteria. Moreover, their expo-
sure to antibiotics or other environmental conditions that
could pose a stress on microbial population is expected
to be lower than that of adults. Fresh green vegetables are
considered as secondary reservoirs of E. coli (Beuchat
2002). These vegetables may be consumed raw and hence
pose a significant infection risk. The virulence factors
studied were evaluated at the genetic level (e.g. LT, ST
and siderophore genes) and phenotypic level (e.g. sidero-
phore production and biofilm formation). Furthermore,
the genetic relatedness between the two sources of isolates
was investigated.
Journal of Applied Microbiology 127, 88--98 © 2019 The Society for Applied Microbiology
Association between MDR and virulence factors in E. coli
Materials and methods
Media and reagents
All media used were obtained from Oxoid (Hampshire,
UK) and Liofilchem (Teramo, Italy). Chemical reagents
used were from Sigma (St Louis, MO). Modified M9 med-
ium (6 g l
1 Na2HPO4, 3 g l
1 KH2PO4, 1 g l
1 NH4Cl,
05 g l
1 NaCl, 1 mmol l
1 MgSO4, 01 mmol l
1 CaCl2,
01% glucose, 00025% nicotinic acid, 02% casein amino
acids and 165 lg ml
1 thiamine in water) was prepared as
previously described (Blango et al. 2014).
Escherichia coli strains
Escherichia coli (70 strains) were isolated originally from
fresh green vegetables (parsley, lettuce and mint) that
were grown in Jordan and delivered on daily basis to the
vegetable market. They were collected, isolated and iden-
tified in a previous work (Burjaq and Shehabi 2013).
Escherichia coli faecal strains (68 strains) were isolated
from healthy infants (<1 year old). The isolates were col-
lected, isolated and identified in a previous work (Badran
et al. 2016). In that work, the collection of the isolates
was performed after obtaining the necessary ethical
approval and consent forms (Badran et al. 2016). Escheri-
chia coli Nissle 1917, a probiotic nonpathogenic micro-
organism was the kind gift of Ardeypharm GmbH (Her-
decke, Germany). This strain is known to carry the genes
for four siderophores (aerobactin, enterobactin,
salmochelin and yersiniabactin) (Searle et al. 2015). All
the isolates were stored in 15% v/v glycerol in tryptic
soya broth (TSB) at
80°C for long-term storage and at

20°C for short-term storage.
Antibiotic susceptibility test
To confirm the previously reported antibiotic susceptibil-
ity results (Burjaq and Shehabi 2013; Badran et al. 2016),
the isolates were tested for antibiotic susceptibility
according to CLSI (2016) guidelines using the following
antibiotics: amoxicillin, amoxicillin-clavulanic acid,
cefuroxime, ciprofloxacin, imipenem, amikacin, trimetho-
prim-sulfamethoxazole and tetracycline. These antibiotics
represent the major classes which have known activity
against E. coli and are used clinically. Escherichia coli
ATCC 25922 was used as a quality control strain to vali-
date the method.
Detection of siderophore production
Siderophore production was measured using SideroTec
kit (Emergen Bio, Maynooth, Ireland), which is a
89
Association between MDR and virulence factors in E. coli R.N. Haddadin et al.

colorimetric method for the quantitative determination

of siderophores. In the presence of siderophore, the col-
our of the reagent changes from blue to pink due to
transfer of ferric iron from the reagent to the sidero-
phore. In this assay, overnight cultures of E. coli isolates
were grown in modified M9 medium at 37°C. After 72 h
incubation, aliquots of the culture were centrifuged and
the supernatants were tested by the kit according to man-
ufacturer’s instructions. A calibration curve (not shown
here) was constructed using standard reagents provided
with the kit after preparing suitable dilutions. The absor-
bance at 630 nm was read for the samples and standard
solutions. Each run was performed in duplicate. Escheri-
chia coli Nissle 1917 was used as a positive control to val-
idate the method. The concentration (lg ml
1) of
siderophores in each sample was back calculated from the
calibration curve.
DNA extraction
From overnight cultures of each strain grown on tryptic
soya agar, three to four colonies were suspended in TSB
and incubated overnight. Bacterial cells from 1 ml of cul-
ture were pelleted at 14 000 g for 2 min (Microcen-
trifuge; Sigma 1-15K, Gilligham, UK). The supernatant
was aspirated and genomic DNA was extracted from the
pellet using a Wizardâ Genomic DNA Purification Kit
(Promega, Madison, WI), by applying the manufacturer’s
procedure. DNA concentration in each sample was mea-
sured spectrophotometrically. Extracted DNA from E. coli
Nissle 1917 was used as a positive control to aerobactin,
enterobactin, salmochelin and yersiniabactin genes.
Extracted DNA from E. coli (strain 10407 from Professor
Asem Shehabi’s laboratory, University of Jordan) was
used as a positive control for LT and ST genes.

Biofilm assay Primers and qPCR conditions

Biofilm formation was assessed by the crystal violet
method (Stepanovic et al. 2000; Haddadin et al. 2010b)
using a 96-well microtitre plate. The test was performed
in five replicates for each isolate. The absorbance was
observed at 570 nm using a microplate spectrophotome-
ter (BioTek, Winooski, VT). The average optical density
(OD) for each strain was reported and the cut-off value
ODc (three standard deviations above the mean) of the
blank was calculated. The strength of adhesion was classi-
fied as either non-adherent, weak, moderate or strong by
comparing the average OD of each strain using the fol-
lowing criteria (Stepanovic et al. 2000): OD ≤ ODc non-
adherent, ODc < OD ≤ 2 ODc weakly adherent, 2
ODc < OD ≤ 4 ODc moderately adherent and OD > 4
ODc strongly adherent.
qPCR was performed to detect siderophore genes, LT
enterotoxin and ST genes. The targets and primers used
to detect the siderophore genes, aerobactin, enterobactin,
salmochelin and yersiniabactin, in addition to LT and ST
enterotoxin genes are presented in Table 1. Escherichia
coli-specific gene yccT was used as an internal standard
(Clifford et al. 2012). This gene was chosen to confirm
the identity of E. coli isolates that were obtained from
other studies (Burjaq and Shehabi 2013; Badran et al.
2016) as mentioned earlier, in addition to its role as an
internal standard. The reaction mixture (20 ll) contained
the following: DNA (2 ll), 400 nmol l
1 final concentra-
tion of ST primers (200 nmol l
1 for siderophores, LT,
yccT), 10 ll KAPA SYBR Fastâ Universal qPCR (KAPA
Biosystems, Wilmington, MA). The amplification was

Table 1 Targets and primers used to detect

Amplicon siderophore genes and the heat-labile and
Target Primer sequence 50 –30 length Reference heat-stable enterotoxin genes. yccT gene
was used as an internal control
Aerobactin F: CTGCCGGTCGGATTTATTTA 138 bp Searle et al. (2015)
R: ATAAGGGAAATAGCGCAGCA
Enterobactin F: CGAGCGTTTTAGCTCCATTC 143 bp Searle et al. (2015)
R: CCTCTTTCATCGCCTGAGTC
Salmochelin F: TATACCGGTCGTGATGCAAA 150 bp Searle et al. (2015)
R: ATACTCGGCGGTGTTACGTC
Yersiniabactin F: TAAAACTGAAGCCGGGTCAC 122 bp Searle et al. (2015)
R: CCGTTGTGTCACCAGAAATG
Heat-stable F: TCT GTA TTG TCT TTT TCA CC 186 bp Bischoff et al. (2005)
enterotoxin R: TTA ATA GCA CCC GGT ACA AGC
Heat-labile F: GGC GAC AGA TTA TAC CGT GC 750 bp Bischoff et al. (2005)
enterotoxin R: CCG AAT TCT GTT ATA TAT GTC
yccT F: GCATCGTGACCACCTTGA 59 bp Clifford et al. (2012)
R: CAGCGTGGTGGCAAAA

90 Journal of Applied Microbiology 127, 88--98 © 2019 The Society for Applied Microbiology
R.N. Haddadin et al.
performed using a BioRad cycler (USA) according to the
following conditions: For aerobactin, enterobactin,
salmochelin and yersiniabactin: denaturation at 940°C
for 5 min, 40 cycles at 940°C for 30 s, 550°C (60°C for
yersiniabactin) for 30 s, 720°C for 40 s and then one
cycle at 720°C for 5 min. For yccT: denaturation at
950°C for 5 min, 40 cycles at 950°C for 10 s, 560°C for
10 s and 720°C for 35 s. For LT and ST: denaturation at
950°C for 10 min, 40 cycles at 950°C for 6 s, 520°C
(500°C for ST) for 6 s and 720°C for 35 s. For melting
curve analyses, the samples were cooled to 65°C for 30 s
and then heated to 95°C (60 cycles).
Genomic fingerprinting
In order to investigate possible genetic relatedness among
the different groups of isolates, 10 strains from each
group were subjected to enterobacterial repetitive inter-
genic consensus polymerase chain reaction (ERIC-PCR).
The test was performed according to Versalovic et al.
(1991). Briefly, 25 ll of PCR reaction was performed
using 50 pmol of each of the primers (ERIC-1: 50 -ATG-
TAAGCTCCTGGGGATTCAC-30 and ERIC-2: 50 - AAG-
TAAGTGACTGGGGTGAGCG-30 ), 100 ng of template
DNA, 125 ll of PCR Master Mix 2x (GoTaqâ Green
Master Mix; (Promega, Madison, WI). The volume was
made up to 25 ll using nuclease-free water. The amplifi-
cation was performed using a PCR thermocycler (MJ
research-Inc., St. Bruno, Canada) according to the follow-
ing conditions: denaturation at 95°C for 7 min; 30 cycles
at 90°C for 30 s, annealing at 52°C for 1 min and exten-
sion at 65°C for 8 min with a single final extension at
65°C for 16 min. Each PCR reaction was electrophoresed
on 1% w/v agarose gel and visualized by (UVP) system
(Alpha Imagerâ, San Jose, CA) using Redsafe (Intron
Biotechnology, Seongnam, Korea).
Band size estimates and genotype analyses were done
using the software BioNumerics fingerprint data software
package (Applied Maths, Sint-Martens-Latem, Belgium).
Cluster analysis was performed using the same software.
Dendrograms were generated based on the Dice similarity
coefficient and the unweighted pair group method using
arithmetic averages (UPGMA), with 1% optimization and
1% position tolerance. Isolates with a similarity of not
less than 90% were considered to be clonally related.
Ethical considerations
Similar to previous study (Badran et al. 2016), and in
order to collect the samples from infant faeces, ethical
approval was obtained from the Institutional Ethical
Review Boards at Jordan University Hospital and the
Deanship of Scientific Research at the University of
Journal of Applied Microbiology 127, 88--98 © 2019 The Society for Applied Microbiology
Association between MDR and virulence factors in E. coli
Jordan. In addition, an informed consent was obtained
from the mother of each infant investigated (Badran
et al. 2016).
Statistical analysis
The chi-squared test, Mann–Whitney U test and Kruskal–
Wallis test were used to analyse different sets of data. Post
hoc analysis was applied when relevant. The analyses were
performed using IBM SPSS 23 package.
Results
Antibiotic susceptibility test
Escherichia coli (70 isolates) from fresh green vegetables
and 68 isolates from healthy infant faeces were screened
for antibiotic resistance using eight different antibiotic
classes or subclasses. Out of the 70 plant isolates, only 19
isolates were found to be MDR (i.e. resistant to three or
more antibiotics belonging to different classes). Therefore,
in order to make a balanced study, an equivalent number
of isolates, that is, another 19 plant isolates that were
non-MDR (i.e. either susceptible to all antibiotics or
resistant to two or less antibiotics), were selected and
included in the study. Accordingly, the total plant origin
isolates included in the study were 38. Out of the 68 fae-
cal isolates, only 25 isolates were found to be MDR.
Thus, another 25 non-MDR faecal isolates were selected
and included in the study. Consequently, a total of 50
faecal origin isolates were included in the study
(Table S1). Based on antibiotic susceptibility and origin,
isolates were placed into four groups: MDR strains-plant
isolates (R-Plant), susceptible strains-plant isolates (S-
Plant), MDR strains-faecal isolates (R-Faecal) and suscep-
tible strains-faecal isolates (S-Faecal).
Heat-labile and heat-stable enterotoxin genes
Figure 1 shows the prevalence of LT (Fig. 1a) and ST
(Fig. 1b) genes in the four groups. The prevalence of the
LT and ST genes among the four groups is not signifi-
cantly different (v2, P = 0260, 0344 respectively).
Siderophore genes
The prevalence of the four siderophore genes in the
plant and faecal isolates are presented in Fig. 2. The
prevalence of enterobactin gene in plant isolates is
comparable to that in faecal isolates (v2, P = 0244;
Fig. 2a). The prevalence of salmochelin is higher in fae-
cal isolates, but it is not significant (v2, P = 0082;
Fig. 2c). However, there is a substantial difference
91
Association between MDR and virulence factors in E. coli R.N. Haddadin et al.

(a) (b)
30 30
LT gene prevalence %

ST gene prevalence %
25 25 Figure 1 (a) Heat-labile (LT) enterotoxin and
(b) heat-stable (ST) enterotoxin percentage
20 20
gene prevalence in the four groups of isolates
15 15 studied; multidrug resistance (MDR) strains-
10 10 plant isolates (R-Plant), non-MDR (susceptible)
strains-plant isolates (S-Plant), MDR strains-
5 5 faecal isolates (R-Faecal) and non-MDR
0 0 (susceptible) strains-faecal isolates (S-Faecal).
R-Plant S-Plant R-Faecal S-Faecal R-Plant S-Plant R-Faecal S-Faecal [Colour figure can be viewed at wileyonlinelib
Isolate type Isolate type rary.com]

(a) (b)
100 100

80 80
Enterobactin gene

Aerobactin gene
% prevalence
% prevalence

60 60

40 40

20 20

0 0
Plant Faecal Plant Faecal
Isolate type Isolate type

(c) 100 (d)

100
80
Salmochelin gene

Yersiniabactin gene

80
% prevalence

prevalence %

60
60
40
40
Figure 2 The percentage prevalence of the
20
20 four siderophore genes in plant and faecal
isolates: (a) enterobactin gene, (b) aerobactin
0
Plant Faecal 0 gene, (c) salmochelin gene and (d)
Plant Faecal yersiniabactin gene. [Colour figure can be
Isolate type Isolate type viewed at wileyonlinelibrary.com]

between the two sources (plant and faecal) for the enterobactin and salmochelin genes (Fig. 3a,c), there is
other siderophore genes. Aerobactin and yersiniabactin no significant variation in the prevalence of these genes
genes are more prevalent in faecal isolates than in among the four groups (v2, P = 0293 and 0183 respec-
plant isolates (v2, P < 0001; Fig. 2b,d). tively).
Figure 3 presents the prevalence of siderophore genes Previous studies (Lisiecki 2014; Zhang et al. 2017) have
in the four groups of isolates tested. Aerobactin gene shown a correlation between fluoroquinolone resistance
prevalence (Fig. 3b) is significantly higher in R-Faecal and siderophore production; hence, we investigated the
isolates than in the other groups (P < 0001, post hoc chi- possible correlation between resistance to ciprofloxacin
squared, Bonferroni-corrected P value for multiple com- and siderophore gene prevalence. Our results (Table S2)
parisons = 0008). Similarly, yersiniabactin gene is more exhibited a significant association between ciprofloxacin
prevalent in R-Faecal isolates than in the other groups resistance and the production of aerobactin gene (v2,
(P < 0001, post hoc chi-squared, Bonferroni-corrected P P = 0000), while no significant association was detected
value for multiple comparisons; Fig. 3d). While for (P > 005) for the other siderophore genes.

92 Journal of Applied Microbiology 127, 88--98 © 2019 The Society for Applied Microbiology
R.N. Haddadin et al. Association between MDR and virulence factors in E. coli

(a) (b)
100 100

80 80
Enterobactin gene
prevalence %

Aerobactingene
prevalence %
60 60

40 40

20 20

0 0
R-Plant S-Plant R-Faecal S-Faecal
R-Plant S-Plant R-Faecal S-Faecal
Isolate type Isolate type
(c) (d)
100 100

80 80

Yersiniabactingene
Salmolchelingene
prevalence %

prevalence %
60 60

40 40

20 20

0 0
R-Plant S-Plant R-Faecal S-Faecal R-Plant S-Plant R-Faecal S-Faecal

Isolate type Isolate type

Figure 3 The percentage prevalence of the four siderophore genes: (a) enterobactin, (b) aerobactin, (c) salmochelin and (d) yersiniabactin, in the four
groups of isolates studied; multidrug resistance (MDR) strains-plant isolates (R-Plant), non-MDR (susceptible) strains-plant isolates (S-Plant), MDR
strains-faecal isolates (R-Faecal) and non-MDR (susceptible) strains-faecal isolates (S-Faecal). [Colour figure can be viewed at wileyonlinelibrary.com]

from ciprofloxacin-susceptible strains, this difference is

Siderophore production
The concentrations of the released siderophores from
the isolates are presented in Fig. 4. The concentration
of siderophores released from faecal isolates is signifi-
cantly higher than that released from plant isolates
(Mann–Whitney U = 365, P < 0001, Fig. 4a). Fig-
ure 4b shows that the highest amounts of siderophores
released are from faecal isolates, whether MDR or sus-
ceptible ones. While the amount of the released sidero-
phores from these isolates was significantly higher than
R-Plant isolates (Kruskal–Wallis test, 92(2) = 271,
P < 0001), it was also higher than that released from
S-plant isolates, but the latter was not statistically sig-
nificant. The average concentrations of siderophores
released from ciprofloxacin-resistant strains and sus-
ceptible ones (regardless of their origin) were 184 and
53 lg ml
1 respectively (Table S2). Although the con-
centration of the released siderophores from the cipro-
floxacin-resistant isolates was higher than that released
not significant (Mann–Whitney U = 437, P = 0076).
Biofilm formation
The strength of biofilm adhesion was classified as strong,
moderate, weak and non-biofilm former. The majority of
the plant isolates (71%) and faecal isolates (78%) are weak
or non-biofilm formers (Table 2). There is no significant
difference between the plant isolates and the faecal isolates
in the strength of biofilm formed (v2, P = 025). Moreover,
there is no significant difference among the four groups
tested in the strength of biofilm formed (v2, P = 038).
Genomic fingerprinting
The cluster analysis profile of the 40 selected isolates is
shown in Fig. 5. ERIC-PCR grouped the 40 isolates into 10
different clusters. One major cluster grouped 21 isolates
(G2), followed by a smaller cluster which grouped 6 isolates

Journal of Applied Microbiology 127, 88--98 © 2019 The Society for Applied Microbiology 93
Association between MDR and virulence factors in E. coli R.N. Haddadin et al.

(a)
20
Siderophores concentration

15
µg ml–1

10

0
Plant Faecal
Isolate type
(b)
20
Siderophores concentration

15 Figure 4 The mean concentrations (lg ml
1)

of siderophores produced from different
µg ml–1

isolates: (a) plant and faecal isolates, (b)

10 multidrug resistance (MDR) strains-plant
isolates (R-Plant), non-MDR (susceptible)
5 strains-plant isolates (S-Plant), MDR strains-
faecal isolates (R-Faecal) and non-MDR
(susceptible) strains-faecal isolates (S-Faecal).
0
Error bars represent standard error of the
R Plant S Plant R Faecal S Faecal
data set. [Colour figure can be viewed at wile
Isolate type yonlinelibrary.com]

Table 2 Distribution of the four groups of

Total faecal S-Faecal, n R-Faecal, n Total plant, S-Plant, n R-Plant, n Biofilm isolates according to strength of biofilm
(n) (%) (%) (n) (%) (%) adhesion adhesion
1 1 (4%) 0 (0%) 4 3 (16%) 1 (5%) Strong
10 5 (20%) 5 (20%) 7 4 (21%) 3 (16%) Moderate
32 17 (68%) 15 (60%) 19 8 (42%) 11 (58%) Weak
7 2 (8%) 5 (20%) 8 4 (21%) 4 (21%) None
50 25 (100%) 25 (100%) 38 19 (100%) 19 (100%) Total

(G5). Other minor clusters grouped three or less isolates.
The largest cluster G2 was particularly diverse and included
isolates from the both origins (plant and human), both
antibiotic resistance characteristics (MDR and susceptible),
both ciprofloxacin sensitivity characteristics (resistant and
susceptible), and variable siderophore gene content
(Table S3). The second largest cluster G5 included isolates
from human faecal origin only, which were MDR and sus-
ceptible, with variable ciprofloxacin susceptibility and side-
rophore gene content. The G1 cluster included three plant
isolates only, which were all MDR.
Discussion
The aim of this study was to investigate the potential
presence of an association between MDR and selected
virulence factors of E. coli.
The presence of LT and or ST enterotoxins in E. coli
classifies it as an ETEC (Nataro and Kaper 1998). ETEC
is a pathogenic strain belonging to the diarrhoeagenic
E. coli. In this study, faecal isolates were taken from
healthy infants. Even though ETEC prevalence is associ-
ated with diarrhoea due to the production of LT and or
ST genes, the isolation of these strains from healthy peo-
ple is not uncommon (Hien et al. 2008; Rivera et al.
2010).
In this study, we observed no significant difference in
LT or ST gene prevalence among the four isolate groups
tested (R-Plant, S-Plant, R-Faecal and S-Faecal; Fig. 1).
This finding indicates that the resistance or susceptibility
to antibiotics is not associated with LT or ST gene. Hien
et al. (2008),observed high resistance among diar-
rhoeagenic E. coli (including ETEC) to antibiotics that
are used against these bacteria. Unfortunately, their

94 Journal of Applied Microbiology 127, 88--98 © 2019 The Society for Applied Microbiology
R.N. Haddadin et al. Association between MDR and virulence factors in E. coli

80

ERIC PCR
81·5

82
3·3

84
84·8

85·5
0·7

86

1·8
87·3
2·5
88·0

88·5
88

1·2

88·7

89·1
89·3

0·2
0·8

89·5
0·3

0·5

90·0
0·5
1·7 90

1·5
91·0
2·2

91·1
18·5

91·5

4·3
6·0

3·9

92·0
92·0

0·9
0·5

92·5
92

92·7

0·5
1·5
15·2

2·4

0·7
93·0

93·0

93·0

93·0
0·5
93·5

93·5

6·0
0·8

1·3
93·8
94·0

12·7

94·3
2·3
94
12·0

0·4
1·5
95·0

95·0

1·8

3·0

3·0
95·3

10·0
1·7
2·1

96·0

96·0

96·0

96·0
9·0

9·0

96·3

96·5
96
8·0

8·0

2·2

0·2
7·0

7·0

7·0

7·0

7·0

7·0

6·5

6·5

97·5

6·2
6·0

6·0

5·0

5·0

5·0

5·0

4·7

2·5
98

4·0

4·0

4·0

4·0

4·0

4·0

4·0

4·0
3·7
1·5

3·5
99·0 1·0

99·0 1·0
2·5

1·0

1·0
100
0

ERIC PCR
G6 G5 G4 G3 G2 G1

20

40

60

80

100
A32-FR
D17-FR
A97-FS
A59-FS
C25-FS
A35--FS
C89-FR
C20-FS
D38-FR
C63-FR
C19-FS
B25-FS
PC32-PR
C67-FR
C12-FS
C38-FR
PC28-PS
C35-FS
16-FS
PC33-PS
PA7-PS
PA27-PR
PB31-PS
PB16-PS
PC30-PS
PA2-PS
PC41-PS
PB40-PR
PA42-PS
P1C34-PR
C87-FR
PC16-PR
C88-FR
PC25-PR
A30-FR
PB50-PR
PA9-PS
PA36-PR
PC35-PR
P1B50-PR
Figure 5 Cluster analysis (dendrogram) of the 40 isolates tested. Minimum similarity was set at 80%. Node information and branch distance are
shown. Arrows indicate the clusters containing two or more isolates. The number under arrow indicates the number of the cluster. The suffix in
the codes of isolates is as follows: PS: plant sensitive (non-MDR), PR: plant resistant, FS: faecal sensitive (non-MDR), FR: faecal resistant.

observation was not statistically evaluated. In our study, from humans and animals. Many studies have found a
although MDR was detected in some ETEC strains, other strong correlation between siderophore production and
ETEC strains were antibiotic susceptible and there was no bacterial virulence (Lamont et al. 2002; Holden and
statistical difference in the prevalence of MDR or non- Bachman 2015). These compounds help the bacteria to
MDR ETEC strains (v2, P = 082; results not shown). acquire iron that is needed for growth and survival in
Siderophores are iron-chelating molecules released by iron depleted environment. In an animal-based host envi-
bacteria to scavenge ferric ions from the environment. In ronment, the bacteria have to overcome the limitation in
this study, the prevalence of siderophore genes varied iron availability posed by host iron-chelating mechanisms
between isolates from both sources (plants and faeces, in tissues and blood such as transferrin, lactoferrin and
Fig. 2). Aerobactin and yersiniabactin genes showed a siderocalin (Raymond et al. 2003; Correnti and Strong
higher prevalence in faecal isolates than in plant isolates 2012). Hence, siderophore production is expected to be
(Fig. 2b,d). This finding was reflected in the amounts of higher in host associated isolates when compared to free
the released siderophores, where the concentrations of living isolates. In contrast, in order to capture iron and
siderophores produced by faecal isolates were higher than other metals from the environment, some plants are
those observed in plant isolates (Fig. 4a). These results known to produce phytosiderophores, while others utilize
are consistent with previously reported data by Searle symbiotic microbial siderophores to perform this task
et al. (2015). In their study, the production of sidero- (Ahmed and Holmstr€ om 2014). This would suggest that
phores in plant-associated isolates was less than those the E. coli isolated from fresh green vegetables in our

Journal of Applied Microbiology 127, 88--98 © 2019 The Society for Applied Microbiology 95
Association between MDR and virulence factors in E. coli
study were also present in a competitive environment
that necessitates the production of siderophores for their
survival. This was reflected in the presence of various
siderophore genes in the plant isolates and the produc-
tion of these compounds from these isolates. However, it
has been hypothesized that E. coli associated with plants
could have other systems of iron acquisition than the use
of siderophores, such as the direct capturing of fer-
richrome or ferric citrate molecules, which are found in
the plant phyllosphere and rhizosphere (Reid et al. 1984;
Crowley 2007; Searle et al. 2015). Hence, siderophore
production in plant isolates was less than that in human
isolates. Our results could not elucidate an association
between MDR and siderophore gene prevalence.
Although the prevalence of aerobactin and yersiniabactin
genes was statistically more significant in R-faecal isolates
compared to non-MDR isolates, the other siderophores
were not. Moreover, no siderophore genes were prevalent
in R-plant isolates when compared to susceptible ones.
Some researchers have reported a correlation between flu-
oroquinolone resistance and siderophore production in
some bacterial species (Lisiecki 2014; Zhang et al. 2017).
In these reports, fluoroquinolone-resistant strains were
associated with larger amounts of siderophore when com-
pared with fluoroquinolone-sensitive strains. In this
regard, our data shows that although the average concen-
trations of siderophores released from ciprofloxacin-resis-
tant strains were higher than that released from
ciprofloxacin-susceptible strains (regardless of their ori-
gin), this difference was not statistically significant. Based
on these results, we investigated if there is a relationship
between ciprofloxacin resistance and any siderophore
gene prevalence. Our results showed a significant associa-
tion between ciprofloxacin resistance and the presence of
aerobactin gene in the isolates studied regardless of their
origin. The suggested explanation for the increase in side-
rophore production in ciprofloxacin-resistant isolates
assumes that siderophore release inhibits oxidative stress
response of the bacteria, thus enhancing ciprofloxacin
resistance (Zhang et al. 2017).
Biofilm formation is among the most important viru-
lence factors exhibited by microbes. This study shows
that the strength of attachment of the biofilms was not
associated with isolate origin (faecal or fresh green veg-
etable). In addition, biofilm formation was comparable
among the four groups of isolates studied (R-Plant, S-
Plant, R-Faecal and S-Faecal). These results are in line
with those previously reported by Haddadin et al.
(2010a). However, in that study they could not detect a
correlation between MDR or non-MDR Pseudomonas
aeruginosa and the strength of attachment to stainless
steel surfaces. In another study, Di Domenico et al.
(2017) observed that 80% of clinical isolates were able to
96 Journal of Applied Microbiology 127, 88--98 © 2019 The Society for Applied Microbiology
R.N. Haddadin et al.
form biofilm. For these strains, the level of biofilm pro-
duction between MDR isolates and non-MDR isolates
was comparable and no significant difference was
detected. Nevertheless, a few studies (Drenkard and
Ausubel 2002; Mah et al. 2003) have been able to identify
indirect correlation between biofilm formation of some
species and resistance to some antibiotics. In these stud-
ies, a deletion or overexpression of specific genes that
affects biofilm formation were found to have an impact
on specific traits in the bacteria that impact their resis-
tance against some antibiotics.
Genetic fingerprinting analysis in this study (Fig. 5)
has shown that many plant and human E. coli isolates
belong to the same clone. This indicates that the bacteria
are circulating in the environment and moving from
humans to plants and vice versa. This finding is alarming
since it implies the contamination of vegetables with
human faeces, despite the fact that in Jordan, sewage
wastewater is used for irrigation only after being treated
(Khreisat and Abu-Sharar 2018). Accordingly, it is possi-
ble that MDR micro-organisms could disseminate among
humans as a result of the consumption of uncooked veg-
etables. Therefore, further studies should be undertaken
to consider the role of environmentally occurring MDR
micro-organisms in subsequent human infection.
In conclusion, in this study, we could not elucidate a
clear association between MDR and the increase in any of
the virulence factors studied, whether at the genetic level
(siderophores, LT and ST genes) or at the phenotypic level
(siderophore production and biofilm formation) using the
strains isolated from two different environments (infant fae-
ces and fresh green vegetables). However, we observed a sig-
nificant association between isolate origin and siderophore
gene prevalence and siderophore production. This indicates
that the ecological adaptation of micro-organisms could
have an influence on their virulence, resulting in strains
with specific genetic traits or virulence. In addition, it is
possible to have antibiotic susceptible strains that harbour
comparable virulence to MDR strains.
Acknowledgements
The authors thank The Deanship of Scientific Research,
The University of Jordan, for funding this research.
Also, the authors thank Dr Khaldoon Alsamman,
Imam Abdulrahman Bin Faisal University, Dammam,
Saudi Arabia for performing the cluster analysis of the
isolates using BioNumerics fingerprint data software
package.
Conflict of Interest
No conflict of interest declared.
R.N. Haddadin et al.
References
Ahmed, E. and Holmstr€ om, S.J.M. (2014) Siderophores in
environmental research: roles and applications. Microb
Biotechnol 7, 196–208. https://doi.org/10.1111/1751-7915.
12117.
Allocati, N., Masulli, M., Alexeyev, M. and Di Ilio, C. (2013)
Escherichia coli in Europe: an overview. Int J Environ Res
Public Health 10, 6235–6254. https://doi.org/10.3390/ije
rph10126235.
Badran, E.F., Qamer Din, R.A. and Shehabi, A.A. (2016) Low
intestinal colonization of Escherichia coli clone ST131
producing CTX-M-15 in Jordanian infants. J Med
Microbiol 65, 137–141. https://doi.org/10.1099/jmm.0.
000210.
Beuchat, L.R. (2002) Ecological factors influencing survival
and growth of human pathogens on raw fruits and
vegetables. Microbes Infect 4, 413–423. https://doi.org/10.
1016/S1286-4579(02)01555-1.
Bischoff, C., L€ uthy, J., Altwegg, M. and Baggi, F. (2005) Rapid
detection of diarrheagenic E. coli by real-time PCR. J
Microbiol Methods 61, 335–341. https://doi.org/10.1016/j.
mimet.2004.12.007.
Blango, M.G., Ott, E.M., Erman, A., Veranic, P. and Mulvey,
M.A. (2014) Forced resurgence and targeting of
intracellular uropathogenic Escherichia coli reservoirs. PLoS
ONE 9, e93327. https://doi.org/10.1371/journal.pone.
0093327.
Burjaq, S. and Shehabi, A. (2013) Fresh leafy green vegetables
associated with multidrug resistant E. coli. Int Arab J
Antimicrob Agents 3, 1–7.
Clifford, R.J., Milillo, M., Prestwood, J., Quintero, R.,
Zurawski, D.V., Kwak, Y.I. and Mc Gann, P. (2012)
Detection of bacterial 16S rRNA and identification of four
clinically important bacteria by real-time PCR. PLoS ONE
7, e48558. https://doi.org/10.1371/journal.pone.0048558.
Clinical and Laboratory Standard Institute (CLSI). (2016). M
100S, Performance Standards for Antimicrobial Susceptibility
Testing (26th edn). Wayne, PA: CLSI.
Correnti, C. and Strong, R.K. (2012) Mammalian
siderophores, siderophore-binding lipocalins, and the
labile iron pool. J Biol Chem 287, 13524–13531. https://d
oi.org/10.1074/jbc.R111.311829.
Crowley, D.E. (2007) Microbial siderophores in the plant
rhizosphere. In Iron Nutrition in Plants and Rhizospheric
Microorganisms ed. Barton, L.L. and Abadia, J. pp. 169–
198. Dordrecht: Springer.
Di Domenico, E., Farulla, I., Prignano, G., Gallo, M.,
Vespaziani, M., Cavallo, I. and Ensoli, F. (2017) Biofilm is
a major virulence determinant in bacterial colonization of
chronic skin ulcers independently from the multidrug
resistant phenotype. Int J Mol Sci 18, 1077. https://doi.org/
10.3390/ijms18051077.
Drenkard, E. and Ausubel, F.M. (2002) Pseudomonas biofilm
formation and antibiotic resistance are linked to
Journal of Applied Microbiology 127, 88--98 © 2019 The Society for Applied Microbiology
Association between MDR and virulence factors in E. coli
phenotypic variation. Nature 416, 740–743. https://doi.
org/10.1038/416740a.
Haddadin, R.N., Saleh, S.A., Mahmoud, R.A. and Shehabi,
A.A. (2010a) Multiple drug resistance and strength of
attachment to surfaces in Pseudomonas aeruginosa isolates.
Lett Appl Microbiol 51, 48–53. https://doi.org/10.1111/j.
1472-765x.2010.02857.x.
Haddadin, R.N.S., Saleh, S., Al-Adham, I.S.I., Buultjens, T.E.J.
and Collier, P.J. (2010b) The effect of subminimal
inhibitory concentrations of antibiotics on virulence
factors expressed by Staphylococcus aureus biofilms. J Appl
Microbiol 108, 1281–1291. https://doi.org/10.1111/j.1365-
2672.2009.04529.x.
Hien, B.T.T., Scheutz, F., Cam, P.D., Serichantalergs, O.,
Huong, T.T., Thu, T.M. and Dalsgaard, A. (2008)
Diarrheagenic Escherichia coli and Shigella strains isolated
from children in a hospital case-control study in Hanoi,
Vietnam. J Clin Microbiol 46, 996–1004. https://doi.org/10.
1128/JCM.01219-07.
Holden, V.I. and Bachman, M.A. (2015) Diverging roles of
bacterial siderophores during infection. Metallomics 7,
986–995. https://doi.org/10.1039/c4mt00333k.
Ingham, S.C., Losinski, J.A., Andrews, M.P., Breuer, J.E.,
Breuer, J.R., Wood, T.M. and Wright, T.H. (2004)
Escherichia coli contamination of vegetables grown in soils
fertilized with noncomposted bovine manure: garden-scale
studies. Appl Environ Microbiol 70, 6420–6427. https://doi.
org/10.1128/AEM.70.11.6420-6427.2004.
Khreisat, A. and Abu-Sharar, T.M. (2018) Reuse of reclaimed
wastewater in irrigation: review of Jordan’s experience.
Jordan J Agr Sci 14, 91–103.
Lamont, I.L., Beare, P.A., Ochsner, U., Vasil, A.I. and Vasil,
M.L. (2002) Siderophore-mediated signaling regulates
virulence factor production in Pseudomonas aeruginosa.
Proc Natl Acad Sci USA 99, 7072–7077.
Lisiecki, P. (2014) Antibiotic resistance and siderophore
production in enterococci. Med Dosw Mikrobiol 66, 1–10.
Mah, T.-F., Pitts, B., Pellock, B., Walker, G.C., Stewart, P.S.
and O’Toole, G.A. (2003) A genetic basis for Pseudomonas
aeruginosa biofilm antibiotic resistance. Nature 426, 306–
310. https://doi.org/10.1038/nature02122.
Nataro, J.P. and Kaper, J.B. (1998) Diarrheagenic Escherichia
coli. Clin Microbiol Rev 11, 142–201.
Raymond, K.N., Dertz, E.A. and Kim, S.S. (2003)
Enterobactin: an archetype for microbial iron transport.
Proc Natl Acad Sci USA 100, 3584–3588. https://doi.org/
10.1073/pnas.0630018100.
Reid, R.K., Reid, C.P., Powell, P.E. and Szaniszlo, P.J. (1984)
Comparison of siderophore concentrations in aqueous
extracts of rhizosphere and adjacent bulk soils.
Pedobiologia 26, 263–266.
Rivera, F.P., Ochoa, T.J., Maves, R.C., Bernal, M., Medina,
A.M., Meza, R. and Lanata, C.F. (2010) Genotypic and
phenotypic characterization of enterotoxigenic Escherichia
coli strains isolated from Peruvian children. J Clin
97
Association between MDR and virulence factors in E. coli
Microbiol 48, 3198–3203. https://doi.org/10.1128/JCM.
00644-10.
Searle, L.J., Meric, G., Porcelli, I., Sheppard, S.K. and
Lucchini, S. (2015) Variation in siderophore biosynthetic
gene distribution and production across environmental
and faecal populations of Escherichia coli. PLoS ONE 10,
e0117906. https://doi.org/10.1371/journal.pone.0117906.
Shehabi, A.A., Odeh, J.F. and Fayyad, M. (2006)
Characterization of antimicrobial resistance and class 1
integrons found in Escherichia coli Isolates from human
stools and drinking water sources in Jordan. J Chemother
18, 468–472. https://doi.org/10.1179/joc.2006.18.5.468.
Stepanovic, S., Vukovic, D., Dakic, I., Savic, B. and Svabic-
Vlahovic, M. (2000) A modified microtiter-plate test for
quantification of staphylococcal biofilm formation. J
Microbiol Methods 40, 175–179.
Su, Q., Guan, T., He, Y. and Lv, H. (2016) Siderophore
biosynthesis governs the virulence of uropathogenic
Escherichia coli by coordinately modulating the differential
metabolism. J Proteome Res 15, 1323–1332. https://doi.org/
10.1021/acs.jproteome.6b00061.
Versalovic, J., Koeuth, T. and Lupski, J.R. (1991) Distribution
of repetitive DNA sequences in eubacteria and application
98
R.N. Haddadin et al.
to fingerprinting of bacterial genomes. Nucleic Acids Res
19, 6823–6831.
Zhang, W., Zhang, Y., Wang, X., Ding, F., Fu, Y., Zhao, J. and
Chen, X. (2017) Siderophores in clinical isolates of Klebsiella
pneumoniae promote ciprofloxacin resistance by inhibiting
the oxidative stress. Biochem Biophys Res Commun 491, 855–
861. https://doi.org/10.1016/j.bbrc.2017.04.108.
Supporting Information
Additional Supporting Information may be found in the
online version of this article:
Table S1. Antibiogram of the isolates included in the
study
Table S2. Escherichia coli isolates, their susceptibility to
ciprofloxacin, presence of siderophore genes and concen-
tration of siderophores released
Table S3 Isolates grouped in each of the two major
clusters G2 and G5, their resistance (R) or susceptibility
(S) to ciprofloxacin (Cip), the presence of the different
siderophore genes (enterobactin, aerobactin, salmochelin
and yersiniabactin)
Journal of Applied Microbiology 127, 88--98 © 2019 The Society for Applied Microbiology