Evaluation of Resistance Determinants and Clonal Relationship of Extended-Spectrum Beta-Lactamase Positive Gram-Negative Bacterial Strains by PCR and Raman Spectrophotometry
Gülnur TARHAN1*, Tülin DEMİR2, Uğur TAMER3, Merve ERYILMAZ3 and Mümtaz DADALI2
1Department of Medical Microbiology, Adiyaman University, Turkeys
2Ahi Evran University Training and Research Hospital, Turkey
3Department of Analiytical Chemistry, Gazi University, Turkey
Submission:September 22, 2016; Published: November 30, 2016
*Corresponding author: Gülnur Tarhan, Department of Medical Microbiology, Adiyaman University, Faculty of Medicine, Siteler Mahallesi, Atatürk Bulvar, No: 411 Adiyaman, Turkey, Tel: 90505 9448568; Emailgulnur.firstname.lastname@example.org
How to cite this article: Gülnur T, Tülin D, Uğur T, Merve E, et al.Evaluation of Resistance Determinants and Clonal Relationship of Extended-Spectrum Beta-Lactamase Positive Gram-Negative Bacterial Strains by PCR and Raman Spectrophotometry. Adv Biotech & Micro. 2016; 1(4): 555570. DOI: 10.19080/AIBM.2016.01.555570
Background: ESBLs are enzymes which are capable of hydrolysing penicillins, broad-spectrum cephalosporins and monobactams. They are generally derived from TEM and SHV-type enzymes.
Materials and methods: We investigated the presence of TEM, SHV, CTX-M, OXA and IBC beta-lactamase genes by PCR in 300 E. coli, 130 Klebsiella spp., 100 P. aeruginosa ve 23 Enterobacter spp. isolated from various clinical specimens recovered from inpatient and outpatients. Additionally, epidemiological relationships were evaluated by Raman spectromicroscopy method.
Results: According to the PCR results; blaTEM (49.3%) and blaCTX-M (52.6%) were found at high rates, but blaIBC was not detected in any isolates. When PCR test results were evaluated according to the origin of bacteria, the highest rates of blaTEM (69.3%) ve blaCTX-M (79.3%) were detected in E. coli isolates, while the lowest rate (1%) was found in P. aeruginosa. With raman spectromicroscopy, it was identified 9 cluster in E. coli, 6 cluster in K. pneumoniae and K. oxytoca, 11 cluster in P. aeruginosa and 3 cluster in E. Cloacae
Conclusion: Although we obtained valuable results in the comparison of the raman spectrums in positive for resistance genes, there is a great necessity for standardisation of the studies prior to the usage of the test routinely.
Extended-spectrum beta-lactamases (ESBL) were first described soon after the use of wide spectrum cephalosporins in the early 1980s. These enzymes are prevalent globally with a wide scale of variation between countries, hospitals wards and patient groups [1-8]. These are mainly described on plasmids but can be either plasmid or chromosomally-mediated causing resistance to ampicillin, carbenicillin, ticarcillin, cefalotin and cefamandole but have no effect on monobactam, cefamicin and susceptible to beta-lactamases inhibitors, cefoxitin and cefotetan [2,8-10]. ESBLs are derived from TEM-1, TEM-2 or SHV-1 gene loci by mutations  and over 400 types of different
ESBLs have been identified with the common plasmid-encoded types including TEM, SHV, CTX-M and OXA .
TEM-beta-lactamases are the most common type of enzyme among Enterobacteriaceae and were also detected in P. aeruginosa [12-14]. While TEM-1 and TEM-2 could hydrolyse penicillin and first generation cephalosporins. SHV-1 beta-lactamases, generally encoded chromosomally in most strains, were first detected among K. pneumoniae and spread to other Enterobacteriacea species [13,15]. SHV-1 is resistant to ampicillin, ticarcillin and piperacillin but has no effect on oximinocephalosporins. CTX-M beta lactamase originates from chromosomal AmpC enzymes of Klyuvera ascorbata due
to horizontal gene transfer and mutations and it shows 40%
similarity with TEM and SHV enzymes. These are more active
against cefotaxime and ceftriaxone than ceftazidime, even
though point mutations can increase the activity towards
ceftazidime . Recently, it is reported that CTX-M15 is the
most common enzyme globally. While most ESBLs are detected
among Enterobacteriaceae, OXA types are commonly found in P.
These are commonly spread with plasmid and transposon
yielding resistance to aminopenicillin and ureidopenicillin. They
have the ability to hydrolize oxacillin, cloxacillin and methicillin
[13,14]. Inhibitor-resistant beta-lactamases (IRT), a variant
due to mutations of SHV and TEM, has no ESBL activity and can
not hydrolyze third generation cephalosporins and they are
resistant to SAM and AMC but susceptible to TPZ . IRT was
commonly found among E. coli but also reported among other
Enterobacteriaceae [14,15]. IBC-1, integron related Class A wide
spectrum beta-lactamases, is highly resistant to ceftazidime,
intermediate susceptible to cefotaxime, cefepime, aztreonam and
less susceptible to clavulanic acid and piperacillin-tazobactam
compared to other wide spectrum beta-lactamases [15,17]. It
was first detected in E. cloacae and then in E. coli [11,18,19].
Recently, IBC-2, a variant of IBC-1 has been detected among
strains. Additionally, non-TEM and non-SHV ESBLs
such as PER, VEB, GES, TLA were reported .
New approaches of bacterial identification have been
considered recently for the rapid and accurate identification
of bacteria. Vibrational spectroscopic techniques, infrared (IR)
and Raman spectroscopy (RS), are commonly used in chemistry,
since vibrational information is specific to the chemical bonds
and symmetry of molecules. The mechanism of these systems
is based on an intense beam of laser in the visible or infrared
or ultraviolet region focused on the sample and detecting
the scattered beam to get information about the vibration
modes of the sample molecules. RS is a powerful molecular
fingerprinting technique by which the molecule and bacteria
can be identified through the interaction of coherent light and
the sample’s molecules. It has recently gained popularity as an
attractive approach for the biochemical characterization, rapid
identification, and accurate classification of a wide range of
bacterial species and strains [20-24]. This method is in clinical
use in some advanced microbiology laboratories in recent years
with the advantages of the ease for sample preparation, faster
test results, reproducibility and higher discrimination power
compared to other phenotypic and genotypic methods.
Studies have demonstrated that Raman spectra generated
from bacterial and fungal colonies give sufficient information to
identify and differentiate microorganisms and also for biofilm
detection [18,20-22]. Raman signals obtained from bacterial
samples suffer from weakness and a huge background. Surface
enhanced Raman spectroscopy (SERS) is the most common and widely used way to amplify the weak Raman signal is to attach
the sample to a metallic rough surface [18,20,23].
In this study, we aimed to determine the frequencies of
beta-lactamase genes, TEM, SHV, CTX-M, OXA ve IBC by PCR and
epidemiological clonal relationship by Raman spectromicroscopy
among ESBL-positive E. coli, K. pneumoniae, Enterobacter spp.
and P. aeruginosa recovered from various clinical specimens.
Between January 2009 and December 2012, a total of 553
non-duplicated ESBL-positive strain [300 E. coli, 130 Klebsiella
spp. (88 K. pneumoniae, 42 K. oxytoca), 100 P. aeruginosa ve 23
Enterobacter spp. (20 E. cloacae, three E. aerogenes) recovered
from clinical specimens of the patients admitted to Ahi Evran
University Research and Training Hospital, Kirsehir, Turkey were
included in the study. Identification to species level was carried
out using the VITEK-2 Compact automated system (bioMérieux,
France) and conventional biochemical tests.
Testing of susceptibility to ampicillin (AMP, 10 μg), amikacin
(AMK, 30 μg), amoxicillin–clavulanic acid (AMC, 20/10 μg),
aztreonam (ATM, 30 μg), cefepime (FEP, 30 μg), cefotaxime
(CTX, 30 μg), ceftazidime (CAZ, 30 μg), ceftriaxone (CRO, 30
μg), cefuroxime (CXM, 30 μg), ciprofloxacin (CIP, 5 μg), cotrimoxazole
(SXT, 1.25/23.75 μg), fosfomycin tromethamine
(FOF, 200 μg), gentamicin (GEN, 10 μg), imipenem (IPM, 10
μg), and piperacillin–tazobactam (TZP, 100/10 μg) (Oxoid Ltd,
Basingstoke, UK) was determined by Kirby–Bauer disk diffusion
test method in accordance with Clinical and Laboratory Standards
Institute (CLSI) guidelines  and the VITEK-2 Compact
system. E. coli ATCC 25922 and Pseudomonas aeruginosa ATCC
27853 were used as quality control strains. ESBL screening was
performed by disk synergy test, and results were confirmed by
cefotaxime, ceftazidime, cefotaxime–clavulanic acid (CTC, 30/10
μg), and ceftazidime–clavulanic acid (CZC, 30/10 μg) disks, in
accordance with CLSI guidelines . The minimum inhibitory
concentration (MIC) for imipenem was determined by gradient
strip method (bioMérieux, France) following the manufacturer’s
instructions, for strains resistant or intermediately resistant
to imipenem by disk diffusion test. Additionally, the MBL
gradient strip (bioMérieux, France) was used to determine MBL
production for the strains resistant or intermediately resistant
DNA templates for polymerase chain reaction (PCR) were
obtained from isolates on nutrient agar. DNA extraction was
performed by phenol:chloroform:isoamyl method . The ESBL
genes (blaIBC, blaOXA, blaCTX-M, blaTEM and blaSHV) were identified
by PCR as previously described . Primers used in this study were shown on Table 1. E. coli ATCC 25922, E. cloacae HT9
[blaIBC-1(+) control] (Strain was kindly provided by Eva Tzelepi,
from Hellenic Pasteur Institute, Bacteriology Department),
blaTEM-8, blaTEM-1 and blaSHV-3 and blaCTX-M-15 were used
in each run.
Preparation of bacterial cells: All strains were cultured in
a nutrient broth at 37°C overnight, then 1 μL of the suspension
was streaked on nutrient agar and incubated at 37 °C for 24
hours. Single colonies were harvested from the plates using
an inoculating loop and suspended in 1.5 ml of deionized
water. These aliquots were centrifuged at 5000 g for 3 min.
The supernatant was discarded and pellet was used for the
test. 10 μL of each of the suspensions was transferred to a lowfluorescence
quartz microscope slide and allowed to dry at room
Preparation of the solutions: Delta Nu Examiner Raman
Microscopy System (Deltanu Inc., Laramie, WY) was used in
the analysis of bacterial strains. Parameters were set on the
programme of 20x objective, 30 μm laser spot area, 220 mW
laser power and 60 seconds of data collection time. Spectrums
were obtained between 200-2000 cm-1 range. CTAB-stable
golden nanorods were used in this study. Gold nanorods were
synthesized with the minor revisions in the technique of the
nucleus magnification. Nucleus solution was prepared with 7.5
ml CTAB and 250 μl AuCl4. Afterwards, 600 μl of NaBH4 solution
that was previously prepared in the ice bath was added to the
mixture quickly and left in room temperature for 30 minutes
prior to use. Magnification solution was prepared using 4.75 ml
CTAB, 500 μl AuCl4 and 60 μl AgNO3 and 100 μl ascorbic acid. 10
μL of nucleus solution was added to magnification mixture and
left in room temperature for an hour prior to use.
Aluminium surface of the TLC paper was used for SERS
measurement surface. SERS spectrum for each bacterial strain
was obtained by dropping 2 μl of bacterial suspension onto TLC
paper and interacting to the paper with 1 μl of gold nanorod
solution. Median values of the signals from at least three different analysis area were evauated. Data analysis were performed by
MATLAB version 7.1 (Mathworks, Natick, MA).
Cluster analysis of the spectral sets were performed by
SpectraCell RA software (River Diagnostics). MATLAB version
7.1 (The MathWorks, USA) programme was used for histogram
and correlation matrix. The similarities between spectrums
were calculated using Pearson correlation coefficient analysis
(R2) and was multiplied by 100 to express percentage. Data
were analyzed using SPSS software 15.0 (SPSS, Inc., Chicago, IL,
USA). Comparisons of categorical variables were done using Chisquare
tests, although Fisher’s exact test was used when data
were sparse. Significance was set at p < 0.05 using two-sided
A total of 553 ESBL-producing Gram-negative clinical
bacterial isolates (300 E. coli, 130Klebsiella spp. (88 K.
pneumoniae, 42 K. oxytoca), 100 P. aeruginosa ve 23 Enterobacter
spp. (20 E. cloacae, 3 E. aerogenes) recovered from various
infection sites were included in the study. Of all strains tested 402
were recovered from urine, 71 from tracheal aspirate, 48 from
skin and soft-tissue infection. Overall the isolates tested; rates of
strains carrying only SHV, TEM, CTX-M and OXA gene loci was 19
(3.4%), 63 (11.4%), 97 (17.5%) and 8 (1.4%), respectively. Out
of 533 ESBL-producing strains 154 (27.8%) harbored none of
the gene loci. blaCTX-M (n=291; 52.6%) was the most common
enzyme type followed by blaTEM (n=274; 49.5%) and blaSHV
(n=70; 12.6%) among all isolates tested.
While blaCTX-M (79.3%) followed by blaTEM (69.3%) gene
loci were frequent among E. coliM isolates, blaTEM (44.6%) and
blaSHV (41.5%) gene loci were the frequent among Klebsiella
spp. In this study, blaIBC was not detected in any of the strains.
The most co-existence of the gene loci was blaTEM and blaCTX-M with 31.4% of the strains. blaSHV, blaTEM, blaCTX-M gene loci
was found in 1% of the P. aeruginosa strains but blaOXA gene
was not detected. blaTEM positivity was 30.4% for Enterobacter
spp. and blaSHV, blaCTX-M and blaOXA gene loci was not
detected. Overall the E. coli isolates tested (n=300), 157 was
positive for blaTEM+blaCTX-M gene and eight was positive for
SHV+TEM+CTX-M. In a K. pneumonae isolate recovered from
aspirate (0.8%), all four gene loci except IBC were detected.
Presence and distribution of blaSHV, blaTEM, blaCTX-M, blaOXA
and blaIBC gene was shown on Table II and III.
While among urinary isolates the common gene loci
was CTX-M (61.2%) followed by TEM (57.5%); 41.2% of the
bloodstream isolates were positive for TEM/CTX-M. Any gene
loci was not detected among throat and ear samples. Evaluation
of the gene loci positivity according to the patient admission,
SHV, followed by CTX-M positivity were the most common gene
loci among both inpatient and outpatient group (Figure 1, Image
In the second part of the study, subtype analysis of the
strains positive for any of beta-lactamase gene loci by PCR were
evaluated using Raman spectromicroscopy with five different
concentration. Five μL of bacterial suspension was added to 1X
nanoparticule suspansion prepared as 100 μL, 125 μL, 150 μL,
175 μL, 200 μL. Raman spectromicroscopy did not show marked
difference in the spectrum among the prepared solutions in
some strains. As shown in Figure 2, it is observed that the best
bacterial concentration was 5 μL bacteria added to 100 μL of
nanoparticule suspension. Same results were achieved in the
following four repeated tests performed with same samples
and same test conditions. Concentration of 5 μL bacterial
suspension to 100 μL of nanoparticule suspension was used for
the evaluation of the strains positive for any of blaSHV, blaTEM,
bla CTX-M and bla OXA gene loci.
The raman spectrums of the strains with lack of gene loci, E.
coli E120, was shown on Figure 2 and positive for three (blaSHV, blaTEM and blaCTX-M) of the five gene loci, E. coli E113, E. coli
E66 ve K. pneumoniae K9 strains were shown on Figure 3-5.
Weak signal was detected in E120 strain that any of the gene
loci was not detected. A strong and discriminative spectrum
was detected among E. coli E113, E. coli E66 and K. pneumoniae
K9 strains that were positive for blaSHV, blaTEM and blaCTX-M
by PCR. All bacterial strains showed similar results in this
evaluation (Figure 3-6).
Bacterial typing with Raman spectrum revealed that nine
cluster among E. coli strains, six cluster among Klebsiella spp,
11 cluster among P. aeruginosa and three cluster among E.
cloacae were detected. Cluster analysis was not performed for E.
aerogenes strains due to the low number of the strains.
Extended-spectrum beta-lactamase producing and
multidrug-resistant Gram-negative bacteria related infections
are one of the major growing concerns worldwide. Thus,
rapid and accurate detection of resistance mechanism and
determinants has critical importance in the control of the
infections and determining the treatment options .
In this study, a total of 553 ESBL-positive Gram-negative
bacteria recovered from various infection sites were evaluated
for beta-lactamase gene presence including bla IBC, bla TEM,
bla SHV, bla CTX-M by PCR and raman spectroscopy. Among all
isolates tested blaTEM 49.3% and blaCTX-M 52.6% gene loci
was detected but bla IBC presence was not detected. Evaluation
of the gene loci presence by bacterial strains revealed that bla
TEM (69.3%) and bla CTX-M (79.3%) genes were higher among
E. coli and blaSHV (41.5%) and bla TEM (44.6%) gene was higher
among Klebsiella strains. While 1% of P. aeruginosa strains were
positive for blaSHV, bla TEM, bla CTX-M genes, bla OXA was not
detected. The most co-existence of the gene loci was blaTEM and
bla CTX-M with 31.4% of the strains. Of 300 E. coli strain tested
165 were positive for bla TEM and bla CTX-M.
In a study conducted by Paterson et al.  including 455
K. pneumoniae strains recovered from 12 hospitals from seven
countries were evaluated for ESBL presence and among ESBLpositive
strains 67.1% were positive for SHV, 16.4% for TEM,
23.3% for CTX-M. This is worrisome, especially in Turkey where
ESBL prevalence is very high [6,25]. In the past decade blaCTX-M
gene has replaced blaSHV and blaTEM genes in Canada, Europe
and Asia as the most common ESBL type in these bacteria
similar with our findings. The CTX-M beta-lactamases are
now widespread in both nosocomial and community-acquired
pathogens. The blaTEM gene has a high frequency compared to
bla SHV gene which is similar to our finding.
Raman spectroscopy is a new method in the determination
and typing of infectious microorganisms. Although it has
been used for a long time for the chemical characterization of
different materials, it has just lately been applied to the study
of biological samples in order to provide a rapid identification
and discrimination of pathogenic organisms [18, 20- 22].
Data concerning issue is scarce worldwide. In our study,
epidemiological typing with Raman spectrums was performed
for the strains found to be positive for any of the five gene loci and
found that E. coli has nine cluster, K. pneumoniae and K. oxytoca
has six cluster, P.aeruginasa has eleven cluster and E. cloacae has
three clusters. Due to the low number of the strains E. aerogenes
strains were not included for typing. Comparison of the raman spectrums of the strains showed that higher spectrums were
detected among strains positive for resistance genes. There is
a great necessity for standardisation of the studies prior to the
usage of the test routinely. Although, this method gives valuable
information, it doesn’t seem that this method could be alternative
to PFGE, gold standart for epidemiologial typing and it should
be reevaluated with a reference method. Further studies should
be conducted for better understanding and standardised the
The use of some first line treatment antibiotics such
as penicillin and trimethoprim/sulfamethoxazole seem
inappropriate. Antibiotics resistance surveillance and the
determination of molecular characteristics of ESBL isolates are
primordial to ensure the judicious use of antimicrobial drugs.
The prevalence of beta-lactamase producing isolates and their
isolation from life-threatening infections, is increasing at an
alarming rate worldwide. It was shown in this study that betalactamase
producing E. coli strains are an emerging threat in
hospitals and should be supervised by implementation of timely
identification and strict isolation methods that will help reduce
their severe outcomes and mortality rate in this patients. In
conclusion, this study has confirmed the potential of Raman
spectroscopy to identify bacteria.