”4”, :4 .
£2

 

.
.wm»: .

u. .

 

c. ea

E

.. Axum;
3.

do a...

nu .

;

 

 

 

, 9:“. In,

‘ 3111....
1:...r1.... by“
to!
($33....
If)

tiff... .x «{v’: a... (91.13:: ‘ . V ‘ 3:... .x .I’ .3113... .

I. q 5 at. _ .. . «1.25.9. ”.2 . o . . , limajmfunna,.¢c4wwuw..r..unwur
.33 . . a ... . J. , . . a??? fiiwaf _ 3%...ng 52m

. 1.,

 

 

L _LIBFIARY
W Michigan State
University

 

 

 

This is to certify that the
dissertation entitled

A MOLECULAR APPROACH FOR THE DETECTION
AND CHARACTERIZATION OF NOVEL
TETRACYCLINE RESISTANCE GENES, INTEGRONS,
AND INTEGRON-ENCODED ANTIBIOTIC
RESISTANCE DETERMINANTS IN THE
ENVIRONMENT

presented by

 

Carlos Miguel Rodriguez Minguela

has been accepted towards fulfillment
of the requirements for the

Crop and Soil Sciences and
Doctoral degree in Environmental Toxicology

 

14 42 (rel/,—
v7 Maigf Professor'ségature
j cit/1 (1/, 2005

Date

 

 

MSU is an Affirmative Action/Equal Opportunity Institution

PLACE IN REIURN BOX to remove this checkout from your record.

To AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

§‘\
DATE 39E

DATE DUE

DATE DUE

 

SEW 4 2007

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

”mammxmwmmwms

gfifivng?

Illu Him - -

7..st-ri‘:t§iirt -.

A MOLECULAR APPROACH FOR THE DETECTION AND
CHARACTERIZATION OF NOVEL TETRACYCLINE RESISTANCE GENES,
INTEGRONS, AND lNTEGRON-ENCODED ANTIBIOTIC RESISTANCE
DETERMINANTS IN THE ENVIRONMENT

By

Carlos Miguel Rodriguez Minguela

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Crop and Soil Sciences and Environmental Toxicology Programs
Department of Crop and Soil Sciences

2005

ABSTRACT
A MOLECULAR APPROACH FOR THE DETECTION AND

CHARACTERIZATION OF NOVEL TETRACYCLINE RESISTANCE GENES,

INTEGRONS AND lNTEGRON-ENCODED ANTIBIOTIC RESISTANCE

DETERMINANTS IN THE ENVIRONMENT
By
Carlos Miguel Rodriguez Minguela
The large scale use of antimicrobial drugs in agriculture and in human

medicine has been implicated as a major cause of the antibiotic resistance
(ABR) problem. However, most of the research has focused on the following:
(i) the clinical setting, (ii) resistance mechanisms prevailing under selective
conditions and, (iii) the culturable fraction of resistant bacteria. These trends
limit our understanding of the extent of the ABR problem by overlooking
resistance determinants in non-culturable bacteria residing outside the clinical
setting and the role they may have in future clinical threats. Additionally,
knowledge on the diVersity and environmental reservoirs of genetic elements
involved in the capture and dissemination of ABR determinants is still limited.
In order to address these problems, a PCR-based, culture-independent
approach was applied to the detection, characterization, tracking, and
quantification of ABR determinants and integrons (a genetic system implicated
in the uptake of multiple antibiotic resistance genes) in total DNA extracted
from a variety of environmental samples. Four classes of tetracycline
resistance genes: tet (M), tet (O), tet (Q), and tet (36), encoding ribosomal

protection proteins (RPP) were detected in DNA from soil supplemented with

manure from swine fed tetracycline as growth promoter. A novel mosaic gene

and two new putative classes of RPP genes were also detected. Tetracycline
resistance genes were not detected in non-agricultural reference soils. A real-
time PCR assay was developed for tracking and quantifying the novel Tet 36
determinant and to determine functionality of putative RPP genes. The target
sequences were traced to the animals receiving tetracycline and they also
appeared to encode functional genes since their frequency increased in
correlation with exposure to tetracycline. Nine variants (one new) of
aminoglycoside nucleotydiltransferases and three genotypes (one new) for
resistance to quaternary ammonium drugs were detected in association with
class 1 integrons. lntegrons were prevalent and diverse among the tested
soils, but integrons encoding ABR genes were only detected in manured soils.
lntegron-encoded ABR genes persisted in manured soil after one month of
manure application, while RPP determinants became less prevalent. Samples
from marine sediments, permafrost, and subtropical soils were screened for
the presence of integron-encoded integrases. A total of 126 novel integrons
were uncovered. lntegrons were not detected in permafrost samples. These
findings indicate that PCR-based techniques are effective tools for the
detection and tracking of novel resistance loci in environmental DNA and that
the integron module is highly diverse and prevalent in the environment and not

restricted to only bacteria recovered from the clinical setting.

Copyright by
CARLOS MIGUEL RODRIGUEZ MINGUELA
2005

Dedication

To the memory of my grandmother Estebania Rivera Rodriguez

ACKNOWLEDGEMENTS

I thank my major advisor Dr. James M. Tiedje for his instruction,
guidance, and all the opportunities he made possible for me. I also thank the
members of my advisory committee: Dr. Terence L. Marsh, Dr. Clayton Rugh,
and Dr. John Linz for their guidance, and support in my academic and
professional development. I am indebted to Dr. Lee Jacobs from the Crop and
Soil Science Department at Michigan State University for his invaluable
assistance concerning the collection of samples and characterization of the
local cropland fields analyzed in this work. I also thank Gary Zehr for facilitating
the sampling of agricultural fields. I thank Dr. Marilyn Roberts from the
University of Washington and Dr. Anne Summers from the University of Georgia
for providing the strains used as controls for the ribosomal protection protein
genes and integron-targeted PCR assays as well as Prusha Patel for her
assistance in the optimization of PCR procedures.

I am really grateful to my friend Dr. Hector Ayala del Rio for the help,
technical advice, and useful discussions provided throughout all these years. I
also extend my gratitude to Dr. Geshe Braker, Veronica Gruntzig, Dr. Alban
Ramette, Dr. Kostas Konstantinidis, Claribel Cruz, Dr. Barbara O’Kelly, Dr.
Tamara Tsoi, and Joyce Wildenthal for their friendship, good will and helpful
advice provided at different stages of this work.

I am very grateful to my mother Nydia Minguela Rivera, my father Miguel

Rodriguez Toro, my brother Miguel Rodriguez Minguela, Leyda Vézquez, and

vi

the Salcedo family, for their love, encouragement, and total support. I also want
to thank the friends I made in Michigan; Encarnita Figueroa, Mark Sullivan,
Miguel Figueroa, Marisol Figueroa, Jaime Graulau, Andres Chong, Pedro
Torres, Rubén Estrada, Rosa Mary Feliu, Carolina Giugliano and the Ramos
family for all the great times. My deepest sense of gratitude goes to my wife
Kenya for all of her love, encouragement, sacrifices made, and the blessings
from the birth of our son Alejandro. Finally, I want to give my very special
thanks to my parents-in-law Linda and Edgar Leén for all of their unconditional
support and constant motivation.

Financial support was provided by the NSF Center for Microbial Ecology,
the Michigan State University Minority Competitive Doctoral Fellowship, and the
Reservoir of Antibiotic Resistance Network of the Alliance for the Prudent Use

of Antibiotics.

vii

TABLE OF CONTENTS

LIST OF FIGURES ........................................................................... x
LIST OF TABLES ............................................................................ xvii
CHAPTER I: THE ANTIBIOTIC RESISTANCE PROBLEM .......................... 1
Overview ........................................................................................... 1
Dissemination of Antibiotic Resistance ..................................................... 2
Genetic Factors Promoting the Persistence of Antibiotic Resistance ............... 8
Environmental Impact of the Anthropogenic Use of Antibiotics ....................... 10
Major Objectives of this Study ............................................................... 13
References ....................................................................................... 1 5

CHAPTER II: DIVERSITY AND PERSISTENCE OF TETRACYCLINE
RESISTANCE DETERMINANTS CONFERRING RIBOSOMAL PROTECTION,
INTEGRONS, AND CLASS 1 INTEGRON-RELATED ANTIBIOTIC

RESISTANCE GENES IN AGRICULTURAL SOIL .................................... 19
Abstract ............................................................................................ 19
Introduction ....................................................................................... 20
Materials and Methods ......................................................................... 28
Soil samples and DNA extraction ......................................................... 28
PCR analyses .................................................................................. 29
Ligation and Transformation ............................................................... 33
Screening of clone libraries ................................................................. 33
Sequence analysis ........................................................................... 34
Results ............................................................................................. 35
DNA extraction from soil .................................................................... 35
Molecular characterization of RPP genes .............................................. 35
Molecular characterization of ABR gene cassettes encoded by class 1
integrons ........................................................................................ 44
Molecular characterization of integron-encoded integrases ....................... 53
Discussion ........................................................................................ 56
References ....................................................................................... 65

CHAPTER III: DEVELOPMENT OF A QUANTITATIVE PCR METHOD FOR
THE DETECTION OF TET (36) AND INFERRING FUNCTION OF PUTATIVE
RIBOSOMAL PROTECTION PROTEIN GENOTYPES RECOVERED FROM

AGRICULTURAL SOIL ....................................................................... 71
Abstract ............................................................................................ 71
Introduction ....................................................................................... 72
Materials and Methods ......................................................................... 78
Design of real-time PCR primer sets .................................................... 78
Standards ....................................................................................... 79
Soil microcosms ............................................................................... 79
Quantitative real-time PCR conditions .................................................. 80

viii

Isolation of tetracycline resistant strains ................................................ 81
Molecular screening of tetracycline resistant strains for the presence of

putative RPP ...................................................................................... 81
Results ............................................................................................. 82
Optimization of real-time PCR primers .................................................. 82
Sensitivity of real-time PCR assay ........................................................ 83
Abundance of tested RPP genotypes and validation of quantitative assay...86
Molecular characterization of tetracycline resistant isolates ....................... 96
Discussion ........................................................................................ 99
References ...................................................................................... 109

CHAPTER IV: MOLECULAR CHARACTERIZATION OF NOVEL INTEGRON
ENCODED INTEGRASES RECOVERED FROM ENVIRONMENTAL DNA VIA

PCR ............................................................................................... 1 1 3
Abstract .......................................................................................... 1 13
Introduction ..................................................................................... 1 14
Materials and Methods ....................................................................... 123
Results ............................................................................................ 125
Discussion ....................................................................................... 135
References ...................................................................................... 142
CHAPTER V: CONCLUSIONS AND FUTURE DIRECTIONS .................... 145
Conclusions ..................................................................................... 145
Future directions ............................................................................... 147
References ...................................................................................... 151

LIST OF FIGURES

Figure 1.1. Model for the transfer of transposon-associated ABR genes via
transduction, conjugation and natural transformation. Redrawn and modified
from Levy and Marshall, 2004 ................................................................. 5

Figure 1.2. A: Schematic representation of a Class 1 integron presenting the
region common to all integrons (5’-CS) which generally consists of an integrase
coding gene (intl), the site-specific recombination sequence attl, and two
promoters. Expression of the integrase relies on the Pint promoter while the PC
promoter drives the expression of inserted gene cassettes. The region
designated as 3’-CS is characteristic of Class 1 integrons and includes genes
conferring resistance against quaternary ammonium compounds (qacEA1) and
sulfonamide drugs (sul1). B: Insertion of a trimethoprim resistance gene
cassette ((1er through a site specific recombination mechanism between the
aft! site of the integron and the attC sequence present in the cassette. Redrawn
and modified from Sabaté and Prats, 2002 ................................................ 6

Figure 1.3. Stockpiling of gene cassettes conferring resistance against
aminoglycosides (aadA), and trimethoprim (derV). Multiple insertions result

in the displacement of other cassettes that were previously captured,

leading to the formation of multicassette arrays in which most cassettes are
flanked by attC sequences. Redrawn and modified from Sabaté and Prats,
2002 .................................................................................................. 7

Figure 2.1. Chemical structure of tetracycline ........................................... 21

Figure 2.2. PCR amplification of RPP genes from total community DNA
extracted from manured soil and analyzed one week and four weeks after
manure application using two different reaction volumes. Positive control
replicates contained a cloned tet (M) gene as template while negative control
replicates lacked template DNA ............................................................. 36

Figure 2.3. Neighbor joining dendrogram based on amino acid sequence data
showing the relationship among unique RPP recovered from site 21 (one week
after manure application) with respect to others previously reported. The
sequences of elongation factors were included to demonstrate the affiliation of
environmental clones within the RPP cluster. Values represent the percentage
of 1000 replicate trees supporting the branching order ............................... 39

Figure 2.4. Global optimal alignment (71% identity) of the amino acid

sequence derived from clone 397against that of Tet 32. The positions of
identical residues are contained inside black blocks while similar residues are
highlighted in gray and mismatches are in white spaces. The alignment was
created with the Bioedit Sequence Alignment Editor software (Version 5.0.9,
Hall, 1999) ......................................................................................... 40

Figure 2.5. Global optimal alignment (85% identity) of the amino acid

sequence derived from clone 492 against that of Tet OIW. The positions of
identical residues are contained inside black blocks while similar residues are
highlighted in gray and mismatches are in white spaces. The alignment was
created with the Bioedit Sequence Alignment Editor software (Version 5.0.9,
Hall, 1999) ......................................................................................... 41

Figure 2.6. Global optimal alignment corresponding to the last 144 amino acid
residues from clone 492 against Tet ONV (panel A) and Tet 32 (panel B). The
positions of identical residues are contained inside black blocks while similar
residues are highlighted in gray and mismatches are in white spaces.
Alignments were created with the Bioedit Sequence Alignment Editor software
(Version 5.0.9, Hall, 1999). Panel C depicts a proposed recombinant mosaic
structure for clone 492 consisting of segments corresponding to RPP
determinants belonging to classes 0, W, and 32 ....................................... 42

Figure 2.7. Global optimal alignment of the amino acid sequence derived

from clone 25-A6 against that of Tet 32. The positions of identical

residues are contained inside black blocks while similar residues are
highlighted in gray and mismatches are in white spaces. The alignment was
created with the Bioedit Sequence Alignment Editor software (Version 5.0.9,
Hall, 1999) ........................................................................................ 43

Figure 2.8. Diagram illustrating the approach used for the retrieval of full-length
ABR gene cassettes inserted in class1 integrons. Previously described primers
complementary to conserved regions flanking the insertion site were used to
amplify intact ABR gene cassettes ......................................................... 44

Figure 2.9. Amplification products of the variable region of class 1 integrons
generated from DNA extracted from manured soil analyzed one week after
manure application (panel A). No amplicons were generated when using DNA
from a reference soil not exposed to animal waste (panel B). Amplification
reactions were carried out at different annealing temperature to rule out false
negative results for the detection of integron—related genes ........................ 45

Figure 2.10. Amplification products of the variable region of class 1 integrons
generated from DNA extracted from manured soil analyzed four weeks after
manure application (panel A). No amplicons were generated when using DNA
from a reference soil not exposed to animal waste (panel B). Amplification
reactions were carried out at different annealing temperature to rule out false
negative results for the detection of integron-related genes ......................... 46

xi

Figure 2.11. Neighbor Joining tree presenting the affiliation of amino acid
sequences derived from environmental clones relative to that of previously
characterized aminoglycoside nucleotidyltransferases (ANT) described in
cultured bacteria (cluster I). The arrow highlights a novel ANT variant
represented by clone 23l64C-88. Clusters II and Ill represent amino acid
sequences of an aminoglycoside acetyltransferase and an aminoglycoside
phosphotransferase described in Mycobacterium bovis and Enterococcus
faecalis respectively. GenBank accession numbers are shown in parenthesis.
Values represent the percentage of 1000 replicate trees supporting the
branching order .................................................................................. 48

Figure 2.12. Global optimal alignment (86% identity) of the amino acid
sequence derived from clone 23I64C-88 against that of an ANT gene
described in E. coli (GenBank accession no. NP_863002.1). The positions of
identical residues are contained inside black blocks while similar residues are
highlighted in gray and mismatches are in white spaces. The alignment was
created with the Bioedit Sequence Alignment Editor software (Version 5.0.9,
Hall, 1999) ......................................................................................... 49

Figure 2.13. Neighbor Joining tree presenting the affiliation of amino acid
sequences derived from environmental clones relative to that of previously
characterized aminoglycoside nucleotidyltransferases. GenBank accession
numbers are shown in parenthesis. Values represent the percentage of 1000
replicate trees supporting the branching order .......................................... 50

Figure 2.14. Global optimal alignment (70% identity) of the amino acid
sequence derived from clone MS4-F4 against that of the qacEA1 gene
described in A. baumannii (GenBank accession no. AAK72475). The positions
of identical residues are contained inside black blocks while similar residues are
highlighted in gray and mismatches are in white spaces. The alignment was
created with the Bioedit Sequence Alignment Editor software (Version 5.0.9,
Hall, 1999) ........................................................................................ 51

Figure 2.15. Abundance of integron-encoded ABR gene cassettes (panel A)
and RPP determinants recovered from clone libraries corresponding to the
sampling of site 21 after four weeks of manure application (panel B) ............. 52

Figure 2.16. Hypothetical integron structure scheme depicting the binding site
of degenerate primers used for the detection of integron-encoded integrases.
These primers were designed for the amplification of a 500 bp amplicon by
targeting conserved regions located approximately at the middle and close to
the downstream end of the integrase genes of class 1, class 2 and class 3
integrons (White et al., 2000) ................................................................ 53

xii

Figure 2.17. Relationship of environmental and known integron—encoded
integrases. O = site 21, one week after manure application, (03: site 22, no
manure, reference soil for site 21, 9= site 3, two days after manure application,
0= site 21, four weeks after manure application, <>= site 4, no manure,
reference soil for site 3 ......................................................................... 55

Figure 3.1. Examples of quantitative PCR amplification curves depicting the
number of cycles plotted against fluorescence. The region of a curve
corresponding to the exponential phase of the amplification reaction and the
baseline level used for detecting threshold cycles (Ct) are shown ................. 75

Figure 3.2. A: curve of several replicates of amplification reactions illustrating

a decrease in fluorescence as the dissociation of double stranded DNA
proceeds (panel A). B: derivative plot of the dissociation curves of replicate
reactions showing a peak at the melting temperature of the resulting amplicons
from each run .................................................................................... 76

Figure 3.3. Dissociation curve analysis depicting the detection of multiple
amplicons by differences in melting temperature ....................................... 77

Figure 3.4. AzPCR product generated with primers specific for tet (36). B: PCR
product generated with primers specific for clone 397. C: PCR product
generated with primers specific for clone 492, M=100 bp ladder; the following
numbers indicate the nature of the DNA used as template and control ractions:
1= target sequence used as template, 2= no template, 3=target sequence
mixed with other cloned RPP genes, 4=mixture of RPP clones lacking the target
sequence, 5 and 6 = DNA from manured soils 7=DNA extracted from swine
manure (pit), 8=DNA from reference soil lacking manure. The mixtures of
cloned RPP genes contained the following determinants: tet (M), tet (O), tet (Q)
tet (36) clone 397 and clone (492). Clones 397, 492 and tet (36) were excluded
appropriately in amplification reactions containing a mixture of RPP clones as
template ............................................................................................ 84

Figure 3.5. Standard curves depicting the threshold cycle (0,) values plotted
against the log of copy numbers of the target sequences tet (36) (panel A),
clone 397 (panel B) and clone 492 (panel C). Two replicates of each dilution
were amplified .................................................................................... 85

Figure 3.6. Quantification of tet (36) in soil microcosms at five days intervals.
The frequency of the target sequence was monitored in three different
scenarios: (i) non agricultural soil lacking manure and tetracycline, (ii)
agricultural soil supplemented with swine manure (no tetracycline) and (iii)
manure-supplemented agricultural soil plus tetracycline (40pg/g) .................. 87

xiii

Figure 3.7. Quantification of clone 397 in soil microcosms at five days

intervals. The frequency of the target sequence was monitored in three different
scenarios: (i) non agricultural soil lacking manure and tetracycline, (ii)
agricultural soil supplemented with swine manure (no tetracycline) and (iii)
manure-supplemented agricultural soil plus tetracycline (40ug/g) .................. 88

Figure 3.8. Quantification of clone 492 in soil microcosms after 10 days of
exposure to 0, 20, and 40 ug of tetracycline per gram of soil ........................ 89

Figure 3.9. Dissociation curve analysis of real time PCR reactions carried

out with 0.2 pg of plasmid DNA containing the tet (36) sequence (panel A); 10
ng of template DNA extracted from manured soil (panel B); 10 ng of template
DNA extracted from non-agricultural soil (panel C); no template DNA added
(panel D) .......................................................................................... 90

Figure 3.10. Dissociation curve analysis of real time PCR reactions carried out
with 0.2 pg of plasmid DNA containing the putative RPP sequence 397 (panel
A); 10 ng of template DNA extracted from manured soil (panel B); 10 ng of
template DNA extracted from non-agricultural soil (panel C); no template DNA
added (panel D) .................................................................................. 91

Figure 3.11. Dissociation curve analysis of real time PCR reactions carried out
with 0.2 pg of plasmid DNA containing the putative RPP sequence 492 (panel
A); 10 ng of template DNA extracted from manured soil (panel B); 10 ng of
template DNA extracted from non-agricultural soil (panel C); no template DNA
added (panel D) .................................................................................. 92

Figure 3.12. Alignment of nucleotide sequences corresponding to the amplicon
generated under quantitative real-time PCR conditions and the tet (36)
sequence. Identical base positions are highlighted in black. Only positions with
base calls having a quality score of > or = 20 (less than a 1 in 100 chance that
the base call is incorrect, as determined with the phred algorithm, Ewing et al.,
1998) were used in the alignment .......................................................... 93

Figure 3.13. Alignment of nucleotide sequences corresponding to the amplicon
generated under quantitative real-time PCR conditions and that of clone 397.
Identical base positions are highlighted in black. Only positions with base calls
having a quality score of > or = 20 (less than a 1 in 100 chance that the base
call is incorrect, as determined with the phred algorithm, Ewing et al., 1998)
were used in the alignment ................................................................... 94

xiv

Figure 3.14. Alignment of nucleotide sequences corresponding to the amplicon
generated under quantitative real-time PCR conditions and that of clone 492.
Identical base positions are highlighted in black. Only positions with base calls
having a quality score of > or = 20 (less than a 1 in 100 chance that the base
call is incorrect, as determined with the phred algorithm, Ewing et al., 1998)
were used in the alignment ................................................................... 95

Figure 3.15. BOX PCR patterns (A, B, and C) of aerobically grown tetracycline
resistant strains isolates which also presented an RPP amplicon. 1 = strain
NA25-5, 2 = strain NA25-6, 3 = strain NA25-12, 4 = strain NA25-18, 5 = strain
TH10-10, 6 = strain TH50-12 ................................................................ 97

Figure 3.16. Neighbor joining tree based on partial sequencing analysis of 16S
rRNA genes depicting the genetic relationship of tetracycline resistant isolates
carrying ribosomal protection protein genes. GenBank accession numbers
precede the designation of reference strains ............................................ 98

Figure 4.1. Formation of a composite transposon by the insertion of two copies
of an insertion element into the same target DNA molecule (panel A). Panel B
illustrates the mobilization of the resulting composite transposon and the
captured ABR gene into a new target DNA. Redrawn and modified from Snyder
and Champness, 1997 ....................................................................... 116

Figure 4.2. Diagram depicting the linear structure of gene cassettes associated
with class 1 integrons, which consists of two components, a coding segment
and a specific recombination sequence (attC).The attC site consists of an
internal region of variable sequence and length whose ends are delimited by a
core and a inverse core site which are the targets recognized by the integron-
encoded integrase ............................................................................ 1 18

Figure 4.3. Diagram depicting the PCR assay used to amplify a 500 bp region
common to integron-encoded integrases. Comparative analysis of amino acid
sequences derived from sequences retrieved from clone libraries were used to
survey the diversity of integron encoded integrases across different
environments. The primers used were previously described by White et al.,
2000 ............................................................................................... 123

Figure 4.4. Neighbor-joining dendrogram based on partial amino acid sequence
analysis of integron-encoded integrases amplified from sediment samples
retrieved from the Arctic and Pacific Ocean. The dendrogram depicts the
relatedness of the detected integrases with respect to others previously
described. Representative sequences (XerC and XerD) of the closest relatives
of integron encoded integrases group within the DNA breaking-rejoining
enzyme super family are included as references. Values represent the
percentage of 1000 replicate trees supporting the branching order. GeneBank
accession numbers precede the designation of previously described

sequences ....................................................................................... 127

Figure 4.5. Dendrogram depicting the relationships of integron-encoded
integrases detected in soil collected from Mona Island and those previously
descnbed ........................................................................................ 130

Figure 4.6. Dendrogram depicting the relationships of integron-encoded
integrases detected in pristine soil collected in Hawaii and those previously
descnbed ........................................................................................ 131

Figure 4.7. Panel A shows the results for the amplification of 16S rRNA genes
from permafrost soil samples. M = 1Kb ladder, + = positive control, - =negative
control lacking template DNA, 1 = DNA from permafrost loam soil recovered
from the tundra forest at a depth of 16.5 m, estimated geological age: 2-3
million years. 2 = DNA from permafrost collected from the sea coast of the
tundra zone at a depth of 3 m, estimated geological age: 5K years. Panel B
presents the results of the amplification of integrase genes from permafrost
samples. Plasmid DNA containing a cloned class 2 integrase gene was used a
positive control ................................................................................. 133

Figure 4.8. Neighbor joining dendrogram illustrating the relationship of
integron—encoded integrases recovered from a variety of marine and terrestrial
habitats. The codes explaining the source of each sequence are described in
Table 4.1 ......................................................................................... 134

xvi

LIST OF TABLES

Table 2.1. Mechanisms of resistance for characterized tetracycline resistance

determinants. Modified from Chopra and Roberts, 2001 ............................. 23
Table 2.2. History of animal waste exposure, antibiotic use, and

characterization of agricultural and reference soils .................................... 31
Table 2.3. Oligonucleotides used for PCR analyses .................................. 32

Table 2.4. Diversity and abundance of RPP sequences recovered from clone
libraries constructed with DNA amplified one week and four weeks after manure
application ........................................................................................ 38

Table 3.1. Oligonucleotides designed for quantitative real-time PCR
assays ............................................................................................. 79

Table 4.1. Description of samples screened for the presence on integron-

encoded integrases and used to designate their respective clones (NA. = not
available) ........................................................................................ 124

xvii

CHAPTER I
THE ANTIBIOTIC RESISTANCE PROBLEM

Overview. For more than 50 years antibiotics have been widely used in
human and veterinary medicine for treating bacterial infections. Moreover, sub-
therapeutic concentrations of antibiotics are currently applied by farmers in
animal and crop production systems (Schnabel and Jones, 1999; Schmidt et al.,
2001; Evans, 2003; Ghee-Sanford et al., 2001). Many of these drugs are also of
medical importance to humans. Hence, these practices have raised serious
concerns about the consequences of the excessive use of antibiotics relative to
the emergence of antibiotic resistant pathogens (Ferber, 2003). Furthermore,
habits of self-medication, antibiotic overuse by physicians and the utilization of
counterfeit, expired or substandard antibiotics influence the progression of
antibiotic resistance (Okeke et al., 1999). It is a well established fact that the
selective pressure exerted on bacterial populations by a sustained exposure to
antibiotics has a major role in promoting the enrichment, fixation and spread of
antibiotic resistance (ABR) genes through the transfer of mobile genetic
elements among unrelated bacteria (Levy and Marshall, 2004).

The combined effects of the above mentioned factors have resulted in an
unprecedented high incidence of resistance traits among commensal and
pathogenic organisms which seriously compromises the effectiveness of drugs
currently used against life threatening conditions (Levy and Marshall, 2004). For
instance, during the last two decades, the frequency of nosocomial infections

caused by multidrug resistant (MDR) Enterococci has significantly increased

(Mundy et al., 2000). Additionally, the occurrence of MDR strains of classical
pathogens like Klebsiella pneumoniae, and Mycobacterium tuberculosis as well
as opportunistic pathogens like Pseudomonas aeruginosa, and Acinetobacter
baumannni has been reported at a global scale (Levy and Marshall, 2004).
Hospital acquired strains have been increasingly reported to be resistant to the
last resort antibiotics and accounts on reduced susceptibility to the latest
generations of the B-lactam and fluoroquinolone drugs are on the rise (Levy and
Marshall, 2004). However, the spread and prevalence of antibiotic resistance
are no longer considered a problem only restricted to clinical settings.

Drug resistance and novel virulence factors are also emerging in cases
of community acquired infections (Vandenesch et al., 2003). Moreover, the
presence of antibiotic resistant bacteria in natural habitats is being documented
with increasing frequency especially at locations receiving wastes from animal
feeding operations (Ghee-Sanford et al., 2001; Smith et al., 2002; Nandy et al.,
2004; Smith et al., 2004). A similar trend has been observed in sewage
treatment plants and in environments impacted by discharges coming from
these facilities (Baya et al., 1986; Andersen and Sandaa, 1994).

Dissemination of Antibiotic Resistance. Antibiotics are biologically
active chemical compounds that interfere with processes that are vital for
prokaryotic cells. The elimination of organisms susceptible to the activity of a
given chemotherapeutic agent results in the selection of those having a
resistance mechanism that inhibits the action of the administered drug. A

continuous exposure to antimicrobial chemicals provides a powerful selection

for variants harboring resistance traits. These traits may be mobilized to
different environments either by movement of the host cells or by transfer of
their resistance traits to other unrelated strains. Besides being caused by
spontaneous mutations, antimicrobial resistance can be intrinsic or acquired.
Intrinsic resistance refers to the existence of inherent characteristics in a
bacterium that allows it to thrive in the presence of an antibiotic. An example of
this type of resistance would be the lack of membrane permeability or
differences in target organelles or biomolecules that prevent a drug from
exerting its antimicrobial properties. In contrast acquired resistance involves the
gain of accessory genetic material encoding resistance traits from an extrinsic
source. The transfer mechanisms involved in the dispersion of ABR genes
include: transduction, the delivery of foreign bacterial DNA through a viral
vector; conjugation, mating and subsequent transfer of extrachromosomal DNA
(plasmids) from a donor cell into a recipient strain; and natural transformation, a
naturally occurring physiological state that allows some bacteria to capture, and
express extraneous DNA that has been released into its surroundings
(Figure1.1).

Bacteriophages, plasmids, and transposons are mobile genetic elements
implicated in the transfer and dissemination of antibiotic resistance
determinants among bacteria (Figure 1.1). These may harbor resistance genes
against a single family of antibiotics but also may accumulate and carry multiple
resistance determinants that suppress the activity of different types of drugs.

Lately, integrons have been recognized as an important gene capture and

expression system involved in the prevalence and dissemination of ABR genes.
lntegrons are classified based on differences in their integrase gene (Rowe-
Magnus and Mazel, 2002). Class 1 integrons are distinguished by harboring
antibiotic resistance genes that are incorporated via a recombination
mechanism. The two key features of class 1 integrons are the conserved
regions designated as 3’-CS and 5’-CS (Figure 1.2). Several gene cassettes
can be captured by a single integron resulting in the formation of multiple
resistance integrons (MRIs), (Figure 1.3). Once assimilated by an integron,
multiple antibiotic resistance genes can be further mobilized if the integron is
captured by a transposon, which can subsequently merge with a conjugative or

transmissible plasmid (Figure 1.1).

   

Transposon

  
 
   

Bacteriophage

 

 

Chromosome

Chromosome “Free” DNA

Figure 1.1. Model for the transfer of transposon-associated ABR genes
via transduction, conjugation and natural transformation. Redrawn and
modified from Levy and Marshall, 2004.

 

 

 

 

 

 

 

 

 

 

C
Intl attl qacEA1 suI1
A L I
IDint I
3’-CS
B:
IntI attl qacEA1 sul1

 

 

 

 

 

 

 

 

 

sul1

 

In 1‘! attl

 

 

 

 

 

 

 

inserted gene cassette

Figure 1.2. A: Schematic representation of a Class 1 integron presenting the
region common to all integrons (5'-CS) which generally consists of an integrase
coding gene (intl), the site-specific recombination sequence attl, and two
promoters. Expression of the integrase relies on the Pm, promoter while the Pc
promoter drives the expression of inserted gene cassettes. The region designated
as 3’-CS is characteristic of Class 1 integrons and includes genes conferring
resistance against quaternary ammonium compounds (qacEAI) and sulfonamide
drugs (sul1). B: Insertion of a trimethoprim resistance gene cassette ((1er
through a site specific recombination mechanism between the attl site of the
integron and the attC sequence present in the cassette. Redrawn and modified
from Sabaté and Prats, 2002.

 

 

 

 

 

 

lntl attl ’ A V ’ " sun

 

 

 

 

 

 

 

 

Figure 1.3. Stockpiling of gene cassettes conferring resistance against
aminoglycosides (aadA), and trimethoprim (derV). Multiple insertions
result in the displacement of other cassettes that were previously captured,
leading to the formation of multicassette arrays in which most cassettes are
flanked by attC sequences. Redrawn and modified from Sabaté and Prats,

2002.

Additionally, mobilization of integron-associated ABR genes can also be
accomplished by the fusion of an integron with a conjugative transposon (Toma,
et al., 2005). Conjugative transposons are capable of intracellular transposition
and intercellular transfer via conjugation of a plasmid-like, circular intermediate
that does not undergo replication (Salyers, et al.,1995). Contrary to typical
transposons like the Tn5 and Tn10 elements, conjugative transposons do not
duplicate the sequence of the insertion site after integration. Conjugative
transposons may harbor ABR genes in addition to those encoded by a captured
integron. For example, tetracycline resistance genes encoding ribosomal
protection are typically associated with conjugative transposons, reinforcing the
dispersion of ABR genes among strains of clinical concern and commensals
(Salyers, et al.,1995).

Genetic Factors Promoting the Persistence of Antibiotic
Resistance. It has been claimed that the absence of a sustained antibiotic
selective pressure causes the loss of extrachromosomal elements harboring
ABR genes since keeping them becomes an unnecessary metabolic burden
that stresses the host by reducing its fitness and ability to compete against
other populations. However, the existence of certain genetic mechanisms that
perpetuate the prevalence of ABR independent of antimicrobial selective
pressure has been documented. It has been demonstrated in laboratory studies
that compensatory chromosomal mutations can counteract the fitness costs
associated with maintaining a load of accessory genetic material resulting in its

conservation within the host’s genome (Lenski et al., 1994).

Plasmid addiction systems can also promote the persistence of antibiotic
resistance traits under nonselective conditions since plasmid free descendants
are eliminated. These addiction systems require at least two genes, one coding
for a long-lived lethal protein (poison), the other coding for a short-lived antidote
(either an antisense RNA or protein). The cell that keeps a copy of the plasmid
will produce enough antidote to inhibit the activity of the toxin and survives. In
contrast, the cell without the plasmid will contain in its cytoplasm a finite amount
of the long-lived poison and the short-lived antidote molecules. Eventually
cellular enzymes will degrade the antidote, while the toxin will remain intact and
functional causing cell death (Zielenkiewicz and Ceglowski, 2001). Hence, after
exposure to an antibiotic selecting for a plasmid encoded resistance
determinant, the plasmid will prevail. This poison-antidote system is widely
distributed among transmissible plasmids of Gram negative bacteria (Summers,
2002)

Linkage of antibiotic and heavy metal resistance genes into the same
genetic element can also influence the selection of resistance mechanisms
unrelated to an agent exerting an antimicrobial pressure. Since multiple
resistance determinants are physically associated among themselves, exposure
to a single drug will result in the co-selection of all of them independently of the
type of resistance mechanism each of these determinants may encode. Natural
and polluted habitats in which heavy metals are present can serve as
environmental reservoirs in which co-selection through linkage for antibiotic

resistance traits might be operating considering the widespread distribution of

plasmids and heavy metals in the environment. Heavy metal and multiple drug
resistance are common in strains isolated from polluted locations and from
natural environments indicating a genetic link between these two traits
(Morozzi, et al., 1986; Sabry et al., 1997; U9 and Ceylan, 2003; Venna et al.,
2004).

Genetic linkage has been implicated with the high prevalence of ABR
within the clinical setting where increasing levels of MDR strains have been
observed regardless of the implementation of alternating cycles in the use of
antimicrobial agents (Summers, 2002). Hence, policies and guidelines based on
the assumptions that selection for resistance will occur only for the antibiotic
being used must be reexamined, particularly those that dictate the agricultural
use of antibiotics founded on whether the antibiotic in question (or a related
one) is also used in human medicine.

Environmental Impact of the Anthropogenic Use of Antibiotics.
Clinical and agricultural overuse of antimicrobial drugs, are the major factors
associated with the widespread release and distribution of antibiotic resistant
strains from hospitals and agricultural settings into natural habitats. Several
antibiotics are currently used as growth promoting agents in beef, poultry and
swine production facilities. Tetracyclines and macrolides (tylosin) are among the
most widely used drugs for growth efficiency purposes (McEwen and Fedorka-
Cray, 2002). These compounds are applied for years in sequential generations
of animals (McEwen and Fedorka-Cray, 2002) and derivatives of these two

antibiotic families are also used in human medicine (Ellison, 1992; Endberg et

10

al., 2001; Maender and Tyring, 2004). The extensive use of antibiotics in animal
husbandry has been associated with an increase of antimicrobial resistance in
foodborne pathogens and commensal bacteria detected in raw meat and
fermented meat products, raising concerns about the potential for transmission
of ABR through the food chain (Ledergerber et al., 2003; Gevers et al., 2003).
Another potential dispersal pathway for ABR reservoirs is the application of
animal waste as a fertilizer and soil conditioner. Animal waste lagoon systems
used for manure storage have been reported to promote the mobilization of
antibiotic resistance genes from the animal waste microflora into bacterial
populations residing in ground water and soil (Ghee-Sanford et al., 2001).
Sewage waste from industrial, hospital, and domestic sources also
contribute to the aggravation of the ABR problem. It has been reported that the
volume of antibiotics used in hospitals and private households released into
effluent and municipal sewage may reach and exceed the Mleo of susceptible
pathogenic bacteria (K0mmerer and Henninger, 2003). A higher incidence of
antibiotic resistant bacteria has been detected in waste water biofilms and
fecally polluted settings relative to undisturbed sites (T imoney et al., 1978; Baya
et al., 1986; Andersen and Sandaa; 1994; Goni-Urriza et al., 2000; Schwartz et
al., 2003). Additionally, gene transfer of ABR determinants among bacteria is
known to occur in sewage water treatment plants (Mach and Grimes, 1982;
Marcinek et al., 1998), and in natural environments (Sandaa, and Enger, 1994;

Paul et al., 1991).

11

Resistant bacteria in the environment may not remain confined to a
specific location, but can be disseminated by abiotic or biotic means and
ultimately can gain access to our food chain or the clinical environment. Most
importantly, commensal bacteria can serve as reservoirs for resistance genes,
collecting them and keeping them for future transmission. Pathogens and
commensals may have frequent interactions that increase the likelihood that
resistance genes residing in harmless bacteria may be transmitted to those
capable of causing disease.

Tracking antibiotic resistance genes and understanding their reservoirs
has traditionally been done through use of bacterial isolates only. However, it is
well known that antibiotic resistance determinants reside in commensal
bacteria, native environmental populations, injured pathogens or othenrvise
difficult to culture bacteria. The scenario is further supported by the observation
that less than 1% of the prokaryotic species in nature have been cultivated
(Tiedje and Stein, 1999). Much of the work done for characterization of
antibiotic resistance determinants has been based on culture techniques carried
out under the selective pressure exerted by the activity of the antibiotic. This
leads to the classical problem of recovering only the strains that are the fittest
competitors under the conditions used for enrichment and isolation. The
diversity of genotypes present in a particular environment is obscured by this
fact, which is an obstacle to the discovery of novel resistance determinants and
reservoirs. Now it is becoming more widely recognized that understanding the

origin and fate of antibiotic resistance will require more understanding of the

12

ecology of reservoirs that are not in the easily cultured clinical strains (Nandi, et

al., 2004). Further investigation on the dimensions of the pool of ABR genes

residing in environmental reservoirs is necessary to gain a comprehensive
understanding of the ABR problem and its implications to human health.

Major Objectives of this Study. To gain a more comprehensive and
accurate assessment of ABR genes residing in human impacted agricultural soil
and in marine and other terrestrial environments, I used culture-independent
DNA-based technologies. The detection and characterization of inconspicuous
reservoirs and novel determinants was of special interest. I accomplished this
by achieving the following goals:

1. Determine the diversity and distribution of Tetracycline resistance genes
encoding ribosomal protection proteins (RPP) by constructing and analyzing
clone libraries of PCR-amplified RPP genes from total community DNA
isolated from samples of manure-supplemented cropland relative to non-
agricultural soil.

2. Determine the diversity and distribution of integrons and integron-encoded
ABR genes by constructing and analyzing clone libraries of PCR-amplified
ABR gene cassettes from total community DNA isolated from samples of
manure-supplemented cropland relative to non-agricultural soil.

3. Evaluate the persistence of RPP and integron-related genes under field

conditions after manure application.

13

4. Develop a real-time quantitative-PCR assay for selected ABR determinants
and use this method to determine whether the detected RPP genes appear
to be functional.

5. Determine the diversity and biogeography of integrons in human impacted

and non-impacted habitats by analyzing the microbial DNA.

14

REFERENCES

Andersen, SR, and RA. Sandaa. 1994. Distribution of tetracycline resistance
determinants among gram- negative bacteria isolated from polluted and
unpolluted marine sediments. Appl. Envir. Microbiol. 60:908-912.

Baya, A.M., P.R. Brayton, V.L. Brown, D.J. Grimes, E. Russek-Cohen, and RR.
Colwell. 1986. Coincident plasmids and antimicrobial resistance in
marine bacteria isolated from polluted and unpolluted Atlantic Ocean
samples. Appl. Environ. Microbiol. 51:1285-1292.

Ghee-Sanford, J.C., R.l. Aminov, l.J. Krapac, N. Garrigues-Jeanjean, and R.l.
Mackie. 2001. Occurrence and diversity of tetracycline resistance genes
in lagoons and groundwater underlying two swine production facilities.
Appl. Environ. Microbiol. 67:1494-1502.

Ellison, M.J. 1992. Vancomycin, metronidazole, and tetracyclines. Clin. Podiatr.
Med. Surg. 92425-42.

Engberg, J., F.M. Aarestrup, D.E. Taylor, P. Gerner-Smidt, and l. Nachamkin.
2001. Quinolone and macrolide resistance in Campy/obacterjejuni and
C. coli: resistance mechanisms and trends in human isolates. Emerg.
Infect. Dis. 7:24-34.

Evans, JD. 2003. Diverse origins of tetracycline resistance in the honey bee
bacterial pathogen Paenibacillus larvae. J. lnvertebr. Pathol. 83:46-50.

Ferber, D. 2003. WHO advises kicking the livestock antibiotic habit. Science.
301 :1027.

Gevers, D., M. Danielsen, G. Huys, and J. Swings. 2003. Molecular
characterization of tet (M) genes in Lactobacillus isolates from different
types of fermented dry sausage. Appl. Environ. Microbiol. 69:1270-1275.

Goni-Urriza M., M. Capdepuy, C. Arpin, N. Raymond, P. Caumette, and C.
Quentin. 2000. Impact of an urban effluent on antibiotic resistance of
riverine Enterobacteriaceae and Aeromonas spp. Appl. Environ.
Microbiol. 66:125-132.

Ktimmerer, K., and A. Henninger.2003. Promoting resistance by the emission of
antibiotics from hospitals and households into effluent. Clin. Microbiol.
Infect. 9:1203-1214.

Ledergerber, U., G. Regula, R. Stephan, J. Danuser, B. Bissig, and K. Stark.

2003. Risk factors for antibiotic resistance in Campy/obacter spp.
isolated from raw poultry meat in Switzerland. BMC Public Health. 3:39.

15

Lenski, R.E., S.C. Simpson, and T.T. Nguyen. 1994. Genetic analysis of a
plasmid-encoded, host genotype-specific enhancement of bacterial
fitness. J. Bacteriol. 176:3140-3147.

Levy, 8.8., and B. Marshall. 2004. Antibacterial resistance worldwide: causes,
challenges and responses. Nat. Med. 10:S122-129.

Mach, PA, and DJ. Grimes. 1982. R-plasmid transfer in a wastewater
treatment plant. Appl. Environ. Microbiol. 44:1395-403.

Maender, J.L., and SK. Tyring. 2004. Treatment and prevention of rickettsial
and ehrlichial infections. Dermatol. Ther. 17:499-504.

Marcinek, H., R. Wirth, A. MuschoIl-Silberhom, and M. Gauer. 1998.
Enterococcus faecalis gene transfer under natural conditions in

municipal sewage water treatment plants. Appl. Environ. Microbiol.
64:626-632.

McEwen, SA, and P.J. Fedorka-Cray. 2002. Antimicrobial use and resistance
in animals. Clin. Infect. Dis. 34:893-8106.

Morozzi, G., G. Cenci, G. Caldini, G. Losito, and A. Morosi. 1986. Relationship
between environment spread and presence in hosts of Escherichia coli

strains resistant to antibiotics and metals. Zentralbl. Bakteriol. Mikrobiol.
Hyg. 182:393-400.

Mundy, L.M., F. Sahm, and M. Gilmore. 2000. Relationships between
Enterococcal virulence and antimicrobial resistance. Clin. Microbiol.
Rev.132513-522.

Nandi, S., J. Maurer, C. Hofacre, and AC. Summers. 2004. Gram-positive
bacteria are major reservoir of class 1 antibiotic resistance integrons in
poultry litter. Proc. Natl. Acad. Sci. USA. 101:7118-7122.

Okeke, IN, A. Lamikanra, and R. Edelman. 1999. Socioeconomic and
behavioral factors leading to acquired bacterial resistance to antibiotics
in developing countries. Emerg. Infect. Dis. 5:18-27.

Paul, J.H., M.E. Frischer, and J.M.Thurmond. 1991. Gene transfer in marine
water column and sediment microcosms by natural plasmid
transformation. Appl. Environ. Microbiol. 57:1509-1515.

Rowe-Magnus, DA, and D. Mazel. 2002. The role of integrons in antibiotic
resistance gene capture. Int. J. Med. Microbiol. 292:115-125.

16

Sabaté, M., and G. Prats. 2002. Estructura y funcion de Ios integrones. Enferm.
Infecc. Microbiol. Clin. 20: 341-345.

Sabry S.A., H.A. Ghozlan, D.M. Abou-Zeid. 1997. Metal tolerance and antibiotic
resistance patterns of a bacterial population isolated from sea water. J.
Appl. Microbiol. 82:245-252.

Salyers, A., N. Shoemaker. A.M. Stevens, and L.Y. Li. 1995. Conjugative
transposons :an unusual and diverse set of integrated gene transfer
elements. Microbiol. Rev. 59:579-590.

Sandaa, R., and O. Enger. 1994. Transfer in Marine Sediments of the Naturally
Occurring Plasmid pRAS1 Encoding Multiple Antibiotic Resistance. Appl.
Environ. Microbiol. 60:4234-4238.

Schmidt, A., M.S. Bruun, l. Dalsgaard, and J.L. Larsen. 2001. Incidence,
distribution, and spread of tetracycline resistance determinants and
integron-associated antibiotic resistance genes among motile

Aeromonads from a fish farming environment. Appl. Environ. Microbiol.
67:5675-5682.

Schnabel, EL, and AL. Jones. 1999. Distribution of tetracycline resistance
genes and transposons among phylloplane bacteria in Michigan apple
orchards. Appl. Environ. Microbiol. 65:4898-4907.

Schwartz, T., W. Kohnen, B. Jansen,and U. Obst. 2003. Detection of antibiotic-
resistant bacteria and their resistance genes in wastewater, surface
water, and drinking water biofilms. FEMS Microbiol. Ecol. 43:325-335.

Smith, D.|., A.D. Harris, J.A. Johnson, E.K Silbergeld, and J. Glenn Morris.
2002. Animal antibiotic use has an early but important impact on the
emergence of antibiotic resistance in human commensal bacteria. Proc.
Natl. Acad. Sci. 99:6434-6439.

Smith, M.S., R.K.Yang, C.W. Knapp, Y. Niu, N. Peak, M.M. Hanfelt, J.C.
Galland, and D.W. Graham. 2004. Quantification of tetracycline
resistance genes in feedlot lagoons by real-time PCR. Appl. Environ.
Microbiol. 70:7372-7377.

Summers, A. 2002. Generally overlooked fundamentals of bacterial genetics
and ecology. Emerg. Infect. Dis.342885-892.

Tiedje, J.M., and J.L. Stein. 1999. Microbial Biodiversity: Strategies for its
recovery. pp.682-692. In A.L. Demain and J. Davies (eds). Manual of
Industrial Microbiology and Biotechnology. Amer. Soc. Microbiol.,
Washington, DC.

17

Timoney, J.F., J. Port, J. Giles, and J. Spanier. 1978. Heavy-metal and
antibiotic resistance in the Bacterial flora of sediments of New York
Bight. Appl. Environ. Microbiol. 36:465-472.

Toma, C., N. Nakasone, T. Song, and M. lwanaga. 2005. Vibrio cholerae SXT
Element, Laos. Emerg. Infect. Dis. 11: 346-347.

Ug, A., and O. Ceylan. 2003. Occurrence of resistance to antibiotics, metals,
and plasmids in clinical strains of Staphylococcus spp. Arch. Med. Res.
34:130-136.

Vandenesch, F., T. Naimi, M.C. Enright, G. Lina, G.R. Nimmo, H. Heffernan, N.
Liassine, M. Bes, T. Greenland, M.E. Reverdy, and J. Etienne. 2003.
Community-acquired methicillin-resistant Staphylococcus aureus
carrying Panton-Valentine Ieukocidin genes: worldwide emergence.
Emerg. Infect. Dis. 11:346-347.

Verma, T., P.W. Ramteke, S.K. Garg. 2004. Occurrence of chromium resistant
thermotolerant coliforms in tannery effluent. Indian J. Exp. Biol. 4221112-
1 1 16.

Zielenkiewicz, U., and P. Ceglowski. 2001. Mechanisms of plasmid stable
maintenance with special focus on plasmid addiction systems. Acta
Biochim. Pol. 48: 1 003-1 023.

18

CHAPTER II

DIVERSITY AND PERSISTENCE OF TETRACYCLINE RESISTANCE
DETERMINANTS CONFERRING RIBOSOMAL PROTECTION, INTEGRONS,

AND INTEGRON-ENCODED ANTIBIOTIC RESISTANCE GENES IN

AGRICULTURAL SOIL
ABSTRACT
A PCR-based approach was used to track tetracycline resistance genes

and integron-related resistance determinants in swine manure supplemented
soil. Tetracycline was a supplement in the swine feed. The diversity and
prevalence of resistance genes were monitored for four weeks after manure
application and compared to unsupplemented control soils. Genes encoding
tetracycline resistance via ribosomal protection proteins (RPP) were PCR-
amplified from total soil DNA. Class 1 integrons were also amplified to assess
the presence and diversity of antibiotic resistance (ABR) gene cassettes.
Finally, conserved regions of integron integrase genes were amplified, cloned
and sequenced to evaluate their diversity in soils. Antibiotic resistance genes
were detected only in manured soils (not in control soils). Sequence analysis
identified Tet M, Tet O, Tet Q, and Tet 36 RPP determinants. A new mosaic
variant was found along with two potential novel classes of RPP. Screening for
integron cassettes revealed the presence of nine variants of aminoglycoside
nucleotidyltransferases (ANT) including a novel determinant. Three genotypes
(one new) for resistance to quaternary ammonium drugs were also found. The
abundance of RPP genes declined greatly after four weeks of manure addition

while ANT genes remained prevalent as judged by PCR detection and

presence of the respective sequences in clone libraries. Sequencing of cloned

19

integrases showed that this gene acquisition and expression system is
prevalent and diverse among all the tested soils. Moreover, most of the cloned
integrases differed from those reported previously. These results indicate that in
the absence of an obvious selective force, animal waste application may
influence the abundance and persistence of integron-related ANT genes with
respect to background levels. Hence, understanding the true impact of the
agricultural antibiotic use requires a more extensive assessment of the pool of

ABR genes in this environment.

INTRODUCTION

The use of animal waste as a soil fertilizer is a common practice in crop
production. However, sub-therapeutic or prophylactic concentrations of
antibiotics are also applied by farmers as feed additives for increasing
productivity in animal agriculture, which has raised concerns about the effect of
these practices on the emergence of antibiotic resistant (ABR) pathogens.
Tetracycline antibiotics (Figure 2.1) are widely used as growth promoters in the
swine industry (Chopra and Roberts, 2001). This drug is poorly absorbed and
metabolized and it has been estimated that 25 to 75% of the tetracycline
administered to a feedlot is released unmodified in animal wastes (Elmund et
al., 1971). Residual amounts of tetracycline ranging from 42 to 698 ug/L have
been detected in samples of swine manure slurries (Sengelov et al., 2003).
Additionally, it has been reported that tetracycline present in animal wastes may
reach significant concentrations and persist in soil after repeated applications of

liquid manure (Hamscher et al., 2002).

20

 

OH O OH O 0

Figure 2.1. Chemical structure of tetracycline

The tetracyclines are a group of broad-spectrum antibiotics produced by
Streptomyces spp. which are among the most used chemotherapeutic agents
worldwide because of their low toxicity, effective oral absorption, and low cost.
By 1980, over a thousand tetracycline analogs were synthesized or isolated and
the approximate global production of these compounds was about 500 metric
tons (Sambrook and Russell, 2001). Tetracycline disrupts protein synthesis by
impeding the incorporation of aminoacyl-tRNA into the A-site of the prokaryotic
ribosome. Binding of tetracycline to the 30S subunit distorts the ribosome’s
structure preventing proper alignment between the anticodons in the charged
tRNA and the codons in the mRNA bringing translation to a halt. Photoaffinity
labeling and chemical footprinting studies have indicated that the ribosomal
protein S7 and 16rRNA bases G693, A892, U1052, C1054, and G1338 are
implicated in this binding interaction (Chopra and Roberts, 2001). Three
different mechanisms conferring resistance against tetracycline have been

described. These include genes encoding efflux pumps, chemical inactivation,

21

and ribosomal protection proteins (RPP). The actual conventions regarding the
nomenclature and characterization of novel resistance determinants involve the
use of Arabic numerals to designate a new class for new determinants with an
amino acid sequence identity of 5 79% with respect to that of other previously
characterized determinants (Levy et al., 1999). However, a reexamination and
expansion of current nomenclature guidelines has been recently proposed
based on the recent detection of hybrid RPP genes (Stanton et al., 2005).

The efflux proteins are the most widely studied tetracycline resistance
mechanisms. These determinants code for a 46 kDa membrane-associated
protein which mobilizes tetracycline outside the cell reducing the intracellular
concentration of the antibiotic. There are currently 23 known classes of
tetracycline efflux pumps (Table 2.1). Most efflux proteins are particularly
distributed within gram-negative bacteria in association with large conjugative
plasmids (Roberts, 1996).

Three determinants have been reported to encode proteins capable of
catalyzing the chemical modification and subsequent inactivation of tetracycline
(Table 2.1). Two of them, tet (X) and tet (37) are NADP-dependent
oxidoreductases while tet (34) is related to a xanthine-guanine phosphoribosyl
transferese (Chopra and Roberts, 2001; Diaz-Torres et al., 2003; Nonaka and
Suzuki, 2002). The tet (X) and tet (34) were identified in strains of Bacteroides
and Vibn'o respectively, whereas tet (37) was recovered through the analysis of

a metagenomic library from oral microflora.

22

Table 2.1. Mechanisms of resistance for characterized
tetracycline resistance determinants. Modified from Chopra
and Roberts, 2001.

Genes

 

 

Efflux pumps

tet (A), tet (B), tet (C), tet (D), tet (E), tet (G), tet (H),

tet (J), tet (V), tet (Y), tet (Z), tet (30), tet (31), tet (33),
tet (35), tet (39), tet (K), tet (L), tet (38), tet A(P), otr (B),
otr (C), tcr3

Ribosomal protection proteins
tet (M), tet (O), tet (S), tet (W), tet (32), tet (Q), tet (T),
tet (36), otr (A), tet B(P), tet

Enzymatic inactivation
tet (X), tet (37), tel (34)

 

The RPP have been distinguished as the most widespread tetracycline
resistance mechanism among diverse groups of bacteria, except for the
enterics (Chopra and Roberts, 2001). Eleven classes within the RPP group
have been reported (Table 2.1) typically in association with conjugative
transposons (Roberts, 1996). These determinants code for a 75 kDa
cytoplasmic protein, which shares structural similarities (GTP-binding domains)
with various elongation factors (Roberts, 1996). The Tet M determinant is the
best characterized member of this group. It has been found that this protein
forms a complex with GTP and then attaches to the bacterial ribosome where
GTP hydrolysis takes place. The energy released from GTP hydrolysis is used
to remove tetracycline from the ribosome, which restores the ribosome’s normal
shape and perpetuates translation in the presence of the drug. Based on

similarities in the amino acid sequences it has been considered that other

23

members of the RPP family have a similar interaction with tetracycline as that
observed for Tet M (Chopra and Roberts, 2001).

Horizontal gene transfer of antibiotic resistance genes is mediated by
several mobile genetic elements including plasmids, transposons and
conjugative transposons. However, integrons are another group of genetic
elements that have recently received special interest, as they are able to
mobilize multiple antibiotic resistance genes. Most of the integrons currently
known have been detected in clinical strains particularly in those belonging to
the Enterobacteriaceae family. Over 60 gene cassettes conferring resistance
against chloramphenicol, erythromycin, trimethoprim, rifampicin,
aminoglycosides and B-lactam antibiotics have been described among Gram-
negative strains (Hall and Collis, 1998). The resistance mechanisms encoded
by these gene cassettes include enzymatic chemical inactivation, efflux pumps,
and allelic variants of target enzymes which are less sensitive to the activity of
antibiotics (Bissonette et al.,1991; Adrian et al., 1998; Fluit and Schmitz, 1999;
Mingeot—Leclercq et al., 1999).

Gene cassettes conferring resistance against aminoglycoside antibiotics
are commonly found in association with class 1 and class 2 integrons (Nandi et
al., 2004; F luit and Schmitz, 2004 ). Aminoglycoside antibiotics are extensively
used for the treatment of infections caused by Gram-positive and Gram-
negative bacteria including tuberculosis (Jana and Deb, 2005; White et al.,
2000). Their mode of action involves the disruption of protein synthesis by

blocking the A-site of the bacterial ribosome (Jana and Deb, 2005). Resistance

24

against aminoglycoside is conferred by lntegron-associated gene cassettes
encoding enzymes that catalyze the chemical modification of these drugs.
Three major groups of these enzymes have been described which include the
N-acetyltransferases (AAC), the O-phosphotransferases (APH), and the O-
nucleotidyltransferases (ANT). Variants of the AAC group use acetyl co-enzyme
A as a donor to modify amino groups in the antibiotic molecule while APH and
ANT use ATP as a second substrate to modify hydroxyl groups by the
incorporation of AMP (Mingeot-Leclercq et al., 1999). Recent studies have
indicated that integrons are distributed outside the clinical setting and also have
been detected among Gram-positive bacteria. (Rosser and Young, 1999; Neild
et al., 2001; Stokes et al., 2001; Tennstedt et al., 2003; Nandi et al., 2004).
Nonetheless, knowledge on the diversity and distribution of integrons and the
role this system may have in the dispersal of ABR genes in natural
environments is still limited.

Class 1 integrons are the best characterized group of this genetic
system. The primary constituents of class 1 integrons include an integrase gene
and the attl recombination site located at the 5’-CS region. This integron class
is further characterized by signature open reading frames present in the 3’-CS
conserved region designated as qacEA1, and sul1. The qacEA1 gene codes for
a membrane-associated efflux pump that confers resistance against quaternary
ammonium drugs and ethidium bromide while the sul1 determinant codes for an
alternative dihydrofolate reductase (a key enzyme in the synthesis of folic acid)

that is not inhibited by sulfa drugs. The 5’-CS and the 3’-CS regions are

25

common to all class 1 integrons and flank an insertion site which may contain a
variable array of ABR gene cassettes.

Most studies on the emergence and dissemination of antibiotic
resistance in livestock-related environments have traditionally been done
through culture-based methods that target indicator species or specific
pathogens. For instance, tetracycline resistant enterococci and E. coli have
been extensively used as indicators in studies exploring the risks associated
with the use of this drug in animal production (Phillips et al., 2004; Sengelov et
al., 2003). Nevertheless, results from these types of culture-based studies do
not represent an extensive or unbiased characterization of ABR reservoirs and
could have overlooked dominant populations of commensals residing in the
sampled environment, which can also be carriers of resistance genes.
Moreover, special emphasis has been given to the environmental study of
determinants conferring resistance against antibiotics known or suspected to
exert a selective pressure in a specific setting, which results in the exclusion of
other important reservoirs of ABR genes that might have been overlooked while
still being prevalent under nonselective conditions. This lack of information
hinders a comprehensive understanding of the true extent of the ABR gene
pool, reservoirs, potential for transfer and possible health hazards. However,
DNA-based techniques can be used to examine the diversity and composition
of the ABR gene pool in environmental sources without the selection bias linked

to isolation techniques.

26

The existence of conserved regions among ribosomal protection genes
has allowed the design of degenerate primers that facilitate the simultaneous
detection of several classes of the RPP family, resulting in the discovery of
novel resistance genotypes when screening bacterial strains via PCR (Barbosa
et al., 1999). Similarly, Oligonucleotides complementary to the conserved
regions among class1 integrons (the integrase and the quaternary ammonium
drugs resistance loci) have been used to characterize the content and order of
inserted ABR gene cassettes in clinical isolates (Levesque et al., 1995; Arduino
et al., 2003; Ebner et al., 2004).

The potential of these PCR-based techniques can be further expanded
by screening total environmental DNA, which can allow the retrieval of new
ABR genotypes and the detection of inconspicuous reservoirs, including those
residing within the unculturable microbial groups. The application of molecular
assays targeting integrons in the environment can be particularly informative
since this system is mobile and represents a potential source of multiple
antibiotic resistance traits. A more refined perception of these factors can lead
to a better understanding of the basis of the antibiotic resistance problem and
the identification of future treats in the clinic. In order to provide a more
comprehensive assessment of ABR reservoirs in environmental samples, this
study was aimed at the following objectives: (i) to determine the diversity and
distribution of tetracycline resistance genes encoding ribosomal protection
proteins, integrons, and integron-encoded ABR genes in total community DNA

isolated from samples of manure-supplemented cropland relative to non-

27

agricultural soil, and (ii) to evaluate the persistence of RPP and integron-related

genes under field conditions after four weeks of manure application.

MATERIALS AND METHODS

Soil samples and DNA extraction. Composite soil samples (1 kg) were
randomly collected from each of 11 agricultural plots having different histories of
swine manure application and antibiotic use in animal production (Table 2.2).
The agricultural fields were analyzed in parallel with respective reference soils
(same soil type not supplemented with manure). Total microbial DNA was
extracted using the FastDNA® SPIN Kit (Qbiogene Inc., Carlsbad, CA) with the
following modifications of the manufacturer’s protocol. Previous to processing,
excess moisture was removed from each of the soil samples (10 g) by placing
them into sterile disposable petri dishes that were left partially opened and
allowed to air-dry at room temperature under aseptic conditions in a biological
hood for 1-2 hrs. This facilitated the normalization of soil mass among the
samples and prevented the dilution of reagents to be added for DNA isolation.
The dried soil was ground aseptically and approximately 800 mg from each
sample were transferred into MULTIMIX 2 Tissue Matrix Tubes.
These tubes were placed on a vortex apparatus using a Mo Bio vortex adapter
(Mo Bio laboratories Inc., Carlsbad, CA) and processed for 5 min at maximum
speed followed by a centrifugation step of 10 min at 10,000 rpm in an
Eppendorf 5415D microcentrifuge. After the addition of the protein precipitation

solution (PPS reagent) the supernatants were centrifuged at 13,200 rpm for 15

28

min and mixed with 1 ml of the Binding Matrix Suspension for 10 min. The resin
was transferred into a SPINTM Filter column and washed twice with 500 pl of
SEWS-M solution. Finally, the DNA was eluted in 150 pl of DNAse/pyrogen free
water, quantified by optical density at. 260 nm and stored at —20°C until
analyzed.

PCR analyses. PCR procedures were optimized for the detection of
tetracycline resistance determinants encoding ribosomal protection, gene
cassettes associated with class 1 integrons, and integron-encoded integrases.
All PCR reactions were carried out in a Robocycler Gradient 96 thermalcycler
(Stratagene, La Jolla, CA). The primers used for PCR are listed in Table 2.3.
Amplification reactions consisted of 150 pmoles of each primer, 1X Taq buffer
(Promega, Madison, WI), 3.5 mM MgCl2,1 mM deoxynucleoside triphosphate
mix (lnvitrogen, Carlsbad, CA) 20 pg of BSA (Roche, Indianapolis, IN) 5 U of
Taq polymerase (Promega, storage buffer B) and approximately 100 ng of soil
DNA in a total reaction volume of 100 pl. This master mix was downscaled to
12.5 pl reaction volume and approximately 50 ng of soil DNA were used for the
amplification of RPP genes after four weeks of manure application and for the
detection of integron-encoded integrases and ABR genes after one and four
weeks of animal waste application. This lower reaction volume increased the
sensitivity of assay in optimization tests. As a quality control measure, small-
subunit (16S) rRNA genes were amplified from the extracted DNA to determine
if it was suitable for PCR. Amplification of 16S rRNA genes was conducted

using the following cycling parameters: initial denaturation at 95°C for 3 min

29

followed by 25 cycles of melting at 95°C for 45 s, 45 s of annealing at 57°C, 1

min 30 s of extension at 72°C and a final extension step of 7 min at 72°C.

30

Table 2.2. History of animal waste exposure, antibiotic use,
and characterization of agricultural and reference soils.

 

 

Site Source Manure application Antibiotic used Soil type
1 wheat field 1 year before sampling CI-tetracycline1 Capac loam
2 non-agricultural forest soil no Capac loam
near wheat field
corn field 2 days before sampling Cl-tetracycline‘ Riddle sandy loam
non-agricultural soil, no Riddle sandy loam
end of corn field
5 soybean field 9 months before sampling Mecadox2 Riddle sandy loam
6 non-agricultural soil, no Riddle sandy loam
end of soybean field
soybean field 16 months before sampling ASP—2502 Owosso sandy loam
8 non-agricultural soil, no Owosso sandy loam
end of soybean field
9 corn field 1 year before sampling Tilmicosin, Tiamulin Owosso marlette
Tylosin1 sandy loam
10 non-agricultural soil, no Owosso marlette
end of corn field sandy loam
11 com field 6 years before sampling Tylan 401 Aubbeenaubbe
capac loam sandy
12 non-agricultural soil, no Aubbeenaubbe
end of corn field capac loam sandy
13 wheat field 1 year before sampling Cl-tetracycline1 Marlette
Tylan 401
14 non-agricultural soil, no Marlette
end of wheat field
15 com field 2 years before sampling NA Capac marlette loam
16 non-agricultural soil, no Capac marlette loam
end of corn field
17 com field 4 months before sampling Cl-tetracycline1 Celina loam
Tylan 401
18 non-agricultural soil, no Celina loam
end of corn field
19 com field 4 months before sampling CI-tetracycline1 Miami loam
Tylan 401
20 non-agricultural soil, no Miami loam
end of corn field
21 wheat field 1 week before sampling Cl-tetracycline1 Capac loam
22 non-agricultural soil no Capac loam
near wheat field

 

1 routinely used as swine feed supplement (growth promoting agent)
2 occasional therapeutic use

31

Table 2.3. Oligonucleotides used for PCR analyses.

 

 

Set Application Primer Sequence (5' to 3') Reference
1 amplification of 16S 8F AGA GTT TGA TCM TGG CTC AG Giovannoni, 1991
rRNA genes 1392R ACG GGC GGT GTG TAC A Amman et al., 1995
2 amplification of RPP tet1 (FWD) GCT CAY GTT GAY GCA GGA A Barbosa et al., 1999
genes tet2 (REV) AGG ATT TGG CGG SAC TTC KA
3 amplification of class 1 5'-CS (FWD) GGC ATC CAA GCA GCA AG Levesque et al., 1995

integrons gene cassettes 3'-CS (REV) AGC ccc ATA CCT ACA AAG CC Arduino et al., 2003

4 amplification of integron- hep35 (FWD) TGC GGG TYA ARG ATB TKG ATT T White et al., 2000
encoded integrases hep36 (REV) CAR CAC ATG CGT RTA RAT

 

Nucleotide base codes for degenerate positions (Comish-Bowden, 1985):Y=
CorT, M=AorC, S=C orG, K= GorT, R=AorG, B= C, GorT.

The cycling conditions used for the amplification of RPP genes consisted
of an initial denaturation at 95°C for 5 min followed by 25 cycles of melting at
94°C for 35 s, 2 min of annealing at 50 °C, 1min 50 s of extension at 72°C and a
final extension step of 10 min at 72°C. Amplification reactions were prepared in
triplicate for each template and those presenting a band were pooled,
concentrated and purified by gel extraction using the Qiagen gel extraction kit
(Valencia, CA). A cloned tet (M) gene (provided by Dr. Marilyn Roberts,
University of Washington, Seattle, WA) was used as a positive control.

Conditions for the amplification of integron-gene cassettes were
performed using the gradient mode of the Robocycler Gradient 96 thermalcycler
as follows: initial denaturation at 94°C for 2 min followed by 35 cycles of melting
at 94°C for 30 s, 1 min of an annealing temperature gradient ranging from 54 to
65°C, 5 min 30 s of extension at 72°C and a final extension step of 7 min at
72°C. The cycling parameters used for the amplification of integron-encoded

integrases were: initial denaturation at 94°C for 2 min followed by 35 cycles of

32

melting at 94°C for 30 s, 1 min of annealing at 52°C, 45 s of extension at 72°C
and a final extension step of 7 min at 72°C. Strains of E. coli harboring class 1
and class 2 integrons were used as positive controls. These were provided by
Dr. Anne Summers (University of Georgia, Athens, GA).

Ligation and Transformation. Approximately 20 ng of the purified PCR
products were cloned into the pCR®4-TOPO® plasmid vector using the TOPO
TA Cloning® for Sequencing kit (lnvitrogen, Carlsbad, CA) as described in the
manufacturer's instructions with the following modifications: the ligation was
carried out for 3 hrs and chemically competent cells were incubated in ice for
two hrs after the addition of the ligation reaction and subsequently heat-
shocked at 42°C for two min. The entire transformation reaction was plated in
LB plates supplemented with ampicillin (100 pg/ml).

Screening of clone libraries. RPP clone libraries were screened by
PCR using primers (T3 and T7) flanking the multiple cloning site using these
cycling parameters: initial denaturation at 94°C for 4 min followed by 25 cycles
of melting at 94°C for 1 min s, 1min of annealing at 46°C, 1min of extension at
72°C and a final extension step of 7 min at 72°C. PCR products were digested
with Mse l and clones presenting unique banding patterns were selected for
sequencing analysis. Plasmid DNA (200ng) from selected clones was submitted
to the Michigan State University Genomics Technology Support Facility (East
Lansing, MI) for sequencing analysis. Full length (1.3 kb) nucleotide sequences
were obtained for all unique RPP genotypes using an ABI PRISM® 3100

Genetic Analyzer (Applied Biosystems, Foster City, CA). Sequencing reactions

33

were performed for both DNA strands using 35 pmoles of the T3 and T7 primer
set. Transforrnants from clone libraries of integron-encoded integrases and
ABR genes were sent for high throughput sequencing analysis to Macrogen
lnc.(Gasan-dong, Seoul, Korea).

Sequence analysis. Amino acid sequences were predicted from
nucleotide data using the JavaScript DNA Translator software version 1.1
(Perry "I, 2002). Hypothetical protein products were compared to protein
databases using BLAST (Altschul et al., 1997). Each putative ABR determinant
or integron integrase was aligned against matching sequences using the
ClustalX program, version 1.81 (Thompson et al., 1997). Protein sequence
alignments were computed using the following alignment parameters: gap
opening and extension penalties were set at 35 and 0.75, respectively, for
pairwise alignments; for multiple alignments, gap opening and extension
penalties were set at 15 and 0.3, respectively. The generated alignments were
further optimized manually with the Bioedit software version 5.0.9 (Hall, 1999)
followed by additional analysis with CIustalX. Comparative amino acid
sequence analyses were carried out by constructing Neighbor-joining
dendrograms using the MEGA2 software, version 2.1 (Kumar et al., 2001).
Dendrograms were computed using the p-distance model and tested by 1,000
bootstrap replications. All alignment positions containing gaps were excluded

from the analysis.

34

RESULTS

DNA extraction from soil. The modified version of the DNA extraction
protocol allowed the recovery of total genomic DNA of high molecular weight
from each of the tested soils. Yields averaged approximately 11 pg of DNA per
800 mg of soil. Small subunit (16S) rRNA genes were amplified from the
extracted templates. However, consistent amplification was only obtained when
supplementing the PCR master mix with BSA, which was sufficient to overcome
the activity of co-extracted PCR inhibitors (Kreader, 1996).

Molecular characterization of RPP genes. Ribosomal protection
determinants were only detected in two different composite soil samples to
which pig manure of animals fed with tetracycline as a growth promoter was
recently applied (sites 3 and 21, Table 2.2). Soil samples nearby, but from
where no manure was used, yielded no tet PCR products. Negative results
were also obtained for samples from nine agricultural soils that had not received
manure for at least four months. The resulting PCR products were of the
expected size (1.3 Kb). A clone library was constructed from the PCR product
generated with template DNA extracted from sites no. 3 and no. 21 (Table 2.2).
Site no. 21 was also sampled after one month of manure application and
screened for the presence of RPP genes by PCR.

Ribosomal protection genes were not detected at site 21 after one month
of animal waste application when PCR reactions where carried out using a

reaction volume of 100 pl. However, downscaling the PCR master mix to a

reaction volume of 12.5 pl resulted in the detection of a weak band of

35

approximately 1.3 Kb (Figure 2.2). PCR products from four replicate reactions
carried out at this lower volume were pooled, concentrated, purified by gel
extraction, cloned and finally sequenced. The 1.3 Kb band was not detected in
sites supplemented with manure four months or longer before sampling when

using a 12.5 pl reaction volume.

100 pL reaction vol.

I
| I

after 1 after 1 after 1month
month 12.5 pL reaction vol.
I |

I ll II , I

+control -control week

 

Figure 2.2. PCR amplification of RPP genes from total community
DNA extracted from manured soil and analyzed one week and four
weeks after manure application using two different reaction volumes.
Positive control replicates contained a cloned tet (M) gene as template
while negative control replicates lacked template DNA.

The RPP sequences derived from clone libraries presented an average
length of approximately 440 amino acid residues. Almost every environmental
clone of RPP sequences amplified from sample 21 (collected one week after
manure application) was affiliated to other previously described determinants:

Tet M, Tet O, Tet Q, Tet ONV and Tet 32 and Tet 36 (Table 2.3). Nevertheless

36

complete homology to published RPP sequences was observed only in three
instances (Figure 2.3). From all the cloned PCR products the ones designated
as 397, 476 and 492 were the most different relative to other RPP sequences
available in GenBank.

Clones 397 and 476 were retrieved from different clone libraries
constructed from PCR products that were independently amplified from DNA
extracted from sample 21. However these two clones presented identical amino
acid sequences and 71% of identical amino acid residues with respect to the
Tet 32 determinant, which is their closest match (Figure 2.4). Clone 492
presented an 85% amino acid sequence identity relative to Tet O/W, a mosaic
RPP gene. However, a segment corresponding to the last 135 amino acid
residues of clone 492 were 98% identical to Tet 32 (Figures 2.5 and 2.6).

The abundance of RPP genes decreased significantly over time as
judged by the observation of weak signals from PCR amplification (Figure 2.2)
and low representation of RPP sequences in clone libraries. Ribosomal
protection proteins constituted 94% (N=98) of the detected genotypes in a clone
library constructed after one week of manure application. In contrast, RPP
comprised only 14% (N=50) of the sequences recovered after one month of
exposure to animal waste. The RPP determinants detected after four weeks
were Tet M, Tet Q, Tet ONV, and Tet 36. A novel RPP variant presenting a 65%

of amino acid identity relative to Tet 32 was uncovered (Table 2.4; Figure 2.7).

37

omnz wmnz

 

 

 

 

an 6: s: F .28
$09 _. mm E
an at $8 m .96: $3 8 2<o a:
$8? _. on “we oboe F O “we
£69 _. O «me $8 78 o O “8.
$8 N 2 C8. $3-3 mm s_ E
8538 has: 220 E EmEELEco mocmsccm ESQ: 98.0 c_ EmEELQcQ
Eom oEEm $0. cozflccmoamm 28 05:5 #0. conflcmmcacm
cozmganm 23:9: Co uses; 538:9“; 235E Co x695
So.— Lctm “.8022... mocmm new. mac Leta “58023 moccm mam

 

.cosmgaam 23:9: Leta 9.603 58 new x82, 95 umEEEm <20 53> 862550
8:99: 6:20 Eo: 69989 $0538 aim Co cosmNthoEmco ucm muscucsnm 58520 .34. win...

38

Bacillus Tn916 TetM
6" 341
93 E. faecalis Tn916 TetM
95 362
Gardnerella vaginalis Tet M
Streptococcal Tn 1545 Tet M
434
466
99 385
Clostn'diaceae Tet 32
324
93 S. pneumoniae Tet O
356
C. jejuni Tet O

  
  
    
  
   
   

 

 

78

 

 

 

r‘ 100 ,—— 492
L— M. elsdenii Tet O/W

'___,{ Bacteriodes sp. 136 Tet 36
100 319 . .
I I3- thetarotaomrcron Tet 0
I314
* Caulobacter crescent EF-G
a,

 

 

 

 

Bacillus anthracis EF-Tu
Clostridium perfringens putative RPP

 

 

I-———I
0.2

Figure 2.3. Neighborjoining dendrogram based on amino acid sequence data
showing the relationship among unique RPP recovered from site 21 (one
week after manure application) with respect to others previously reported. The
sequences of elongation factors were included to demonstrate the affiliation of
environmental clones within the RPP cluster. Values represent the percentage
of 1000 replicate trees supporting the branching order.

39

Tet 32
clone 397

Tet 32
clone 397

Tet 32
clone 397

Tet 32
clone 397

Tet 32
clone 397

Tot 32
clone 397

Tet 32
clone 397

Tet 32
clone 397

1 MKIINLGI
1 ---PRFA-

 

61
57

121
117

181
177 . .

241
237

301
297

361 _ . ,

IWAS 445
lGE 441

 

421
417 ,

 

Figure 2.4. Global optimal alignment (71 % identity) of the amino acid
sequence derived from clone 397against that of Tet 32. The positions
of identical residues are contained inside black blocks while similar
residues are highlighted in gray and mismatches are in white spaces.
The alignment was created with the Bioedit Sequence Alignment
Editor software (Version 5.0.9, Hall, 1999).

40

Tet O/W
clone 492

Tet O/W
clone 492

Tet O/w
clone 492

Tet 0/w
clone 492

Tet O/W
clone 492

Tet O/W
Clone 492

 

Tet O/W
clone 492

 
 

Tet O/w
clone 492

WAS 445
. GB 438

Figure 2.5. Global optimal alignment (85% identity) of the amino acid
sequence derived from clone 492 against that of Tet OIW. The positions of
identical residues are contained inside black blocks while similar residues are
highlighted in gray and mismatches are in white spaces. The alignment was
created with the Bioedit Sequence Alignment Editor software (Version 5.0.9,
Hall, 1999).

41

Tet 0/w
clone 492

An Tet O/W
' clone 492

Tet O/w
clone 492

Tet 32
clone 492

BI Tet 32
' clone 492

Tet 32
Clone 492

 

1 303 438
l ,, . .
C: clone 492 I ‘

 

 

 

Tet ONV (98% ID) Tet 32 (98% ID)

Figure 2.6. Global optimal alignment corresponding to the last 144 amino
acid residues from clone 492 against Tet ONV (panel A) and Tet 32 (panel
B). The positions of identical residues are contained inside black blocks
while similar residues are highlighted in gray and mismatches are in white
spaces. Alignments were created with the Bioedit Sequence Alignment
Editor software (Version 5.0.9, Hall, 1999). Panel C depicts a proposed
recombinant mosaic structure for clone 492 consisting of segments
corresponding to RPP determinants belonging to classes 0, W, and 32.

42

clone 25-A6
Tot 32

clone 25-A6
Tet 32

clone 25-16
Tet 32

clone 25-16
Tot 32

clone 25-A6
Tot 32

clone 25-16
Tot 32

clone 25-16
Tot 32

clone 25-16
Tat 32

10 20 30 (0 50 60
--'-I----I'---l'---l-'-°|--'-I----l----l---'l----l---:I----I

l --CGANR 4
1 MKIINLGI '

    

 

Figure 2.7. Global optimal alignment of the amino acid sequence derived
from clone 25-A6 against that of Tet 32. The positions of identical
residues are contained inside black blocks while similar residues are
highlighted in gray and mismatches are in white spaces. The alignment
was created with the Bioedit Sequence Alignment Editor software
(Version 5.0.9, Hall, 1999).

43

Molecular characterization of ABR gene cassettes encoded by
class 1 integrons. Oligonucleotides targeting conserved regions flanking
the insertion site of gene cassettes in class1 integrons (Levesque, et
al.,1995; Arduino et al., 2003) were used to detect ABR determinants
associated with this genetic system as explained in Figure 2.8, and Table
2.2. The manure-supplemented crop field designated as site 21 (Table 2.2)
was sampled one week and four weeks after manure application along with
a reference soil lacking manure exposure. Amplification of soil DNA
extracted one week after the application of animal waste yielded a PCR
product of approximately 1.4 Kb. No PCR product was detected when
amplifying DNA extracted from the reference soil (Figure 2.9). Screening for
integron cassettes after four weeks of manure fertilization also generated a
1.4 Kb amplicon. The reference soil was also sampled at this time point and

no amplification product was detected after PCR analysis (Figure 2.10).

——9 5’-CS FWD primer (Levesque et al., 1995)

 

 

 

 

 

“h:..;i:"£";~“ '.....|Olnun ? IIIIIIIIIIIOI qaCEA1 SUI1

 

 

 

 

 

 

3’-CS REV primer (Arduino et al., 2003) <—

Figure 2.8. Diagram illustrating the approach used for the retrieval of
full-length ABR gene cassettes inserted in class1 integrons.
Previously described primers complementary to conserved regions
flanking the insertion site were used to amplify intact ABR gene
cassettes.

44

A: 65°764° 63° 62° 61° 60° 59° 58° 57° 56° 55° 54°M
N

1636 hp
1018 bp

 

65’ 64" 63“ 62‘ 61“ 60" 59' 58“ 57" 56’ 55‘ 54°M

1636 bp
1018 bp

 

Figure 2.9. Amplification products of the variable region of class 1
integrons generated from DNA extracted from manured soil analyzed
one week after manure application (panel A). No amplicons were
generated when using DNA from a reference soil not exposed to animal
waste (panel B). Amplification reactions were carried out at different
annealing temperature to rule out false negative results for the detection
of integron—related genes.

45

"o

1636 bp

1018bp

 

1636 bp

1018 bp

 

Figure 2.10. Amplification products of the variable region of class 1
integrons generated from DNA extracted from manured soil analyzed
four weeks after manure application (panel A). No amplicons were
generated when using DNA from a reference soil not exposed to
animal waste (panel B). Amplification reactions were carried out at
different annealing temperature to rule out false negative results for
the detection of integron-related genes.

46

Comparative analysis of derived amino acid sequences revealed the
presence of nine variants of aminoglycoside-modifying enzymes closely related
to determinants conferring resistance against streptomycin and spectinomycin.
Six of these variants (clones designated as 23l64-B9, 23l64-A11, 23l64-F3,
23l64-BB, 23l64-H10, and 23l64-D4) were detected after one week of soil
fertilization while genotypes designated as clones 25l64-D1, 25l64-F7 and
25l64-F10 were recovered after four weeks. All nine variants with the exception
of clone 23l64-B8 presented an amino acid sequence identity equal or greater
than 95% relative to previously described members of the family of
aminoglycoside nucleotidyltransferases (ANT). Clone 23l64-38 emerged as the
most divergent variant sharing 86% of amino acid sequence identity relative to

published ANT sequences. (Figures 2.11 and 2.12).

47

Salmonella enterica (AAMSZ926)

6‘3 aommes

.00 1| 01.an 2364A"
Pseudomonasaeruginosa M07739

67 - CLONE 25640 D1

aONE 2354cr=3

 

 

 

100
*I

ammon
1oo ——1_00I mmmgim (MW)

 

 

100 ,- atoms mo mo
_‘l I uncultured bacterium (MN-11411)
88

 

 

CLOIIE m F10

100 CLONE m or
95 Cormebacten'um yutamicum (04012230) j
Mycobacterium bovis (IIID 853933)}11

 

 

 

 

Entarococcushecdist-‘OOSSO ]-HI

 

01
Figure 2.11. Neighbor Joining tree presenting the affiliation of amino acid sequences derived
from environmental clones relative to that of previously characterized aminoglycoside
nucleotidyltransferases (ANT) described in cultured bacteria (cluster I). The arrow highlights
a novel ANT variant represented by clone 23l64C-88. Clusters II and 1]] represent amino
acid sequences of an aminoglycoside acetyltransferase and an aminoglycoside
phosphotransferase described in Mycobacterium bovis and Enterococcus faecalis
respectively. GenBank accession numbers are shown in parenthesis. Values represent the
percentage of 1000 replicate trees supporting the branching order.

48

clone 23164C-88
Aedh I. coli

clone 23164C-88
Andh E. coli

clone 23164C-88
Roda I. coli

clan. 23164C-88
Leda I. coli

clone 231646-38
Leda I. coli

 

Figure 2.12. Global optimal alignment (86% identity) of the amino acid
sequence derived from clone 23l64C-BB against that of an ANT gene
described in E. coli (GenBank accession no. NP_863002.1). The positions
of identical residues are contained inside black blocks while similar
residues are highlighted in gray and mismatches are in white spaces. The
alignment was created with the Bioedit Sequence Alignment Editor
software (Version 5.0.9, Hall, 1999).

49

Screening for ABR determinants associated with class1 integrons also
revealed the presence of three variants of gene cassettes encoding efflux
pumps conferring resistance against QAC. Two of them presented a 70% of
amino acid identity relative to a QAC gene detected in Acinetobacter baumannii
while the remaining one was 96% identical to a QAC determinant found in
Enterobacter aerogenes Figures 2.13, and 2.14) The comparison of the
abundance of RPP genes relative to integron-encoded ABR gene cassettes
(based on the composition of clone libraries) after four weeks of manure
application to cropland demonstrated that the former group becomes less

prevalent while the latter is still prominent under field conditions (Figure 2.15).

00 I CLONE 23540-04
CLONE MS4-F4
99 75lI CLONE 25640-05
qacEdeltat A. baumannii (AAK72475)
H qacEdelta1 0. freundli (AAR12141)
00I__ qacEdelta1 P. aeruginosa (AAM08185)
qacE2 P. aeruginosa (0AA 11475)

HE. I'AAG45712
fl l——qac cor( )

99 qac S. marcescens (AAK40353)
99 E qacF E. aerogenes (Q9X2N9)
55 CLONE 23l640-E5

EmrE M. avium (NP 962061)

 

 

100

 

 

 

 

 

 

 

0.05

Figure 2.13. Neighbor Joining tree presenting the affiliation of amino acid sequences
derived from environmental clones relative to that of previously characterized
aminoglycoside nucleotidyltransferases. GenBank accession numbers are shown in
parenthesis. Values represent the percentage of 1000 replicate trees supporting the
branching order.

50

anon: use-r4 1
qucxdutta1 1

 

7O 80 90 100 110 120

I I I I I I I I
CLO“ “84-!" 61 LRANSFKP VPLVRVNS VI 120
qocldeltal 61 ..................... 96

uuuuuuuuuuuuuuuuuu

130
I I I I
01.0!!! 1484-24 121 VSCV H TS HKV 142
qecldeltal 97 AF--- S L TPW 115

Figure 2.14. Global optimal alignment (70% identity) of the amino acid
sequence derived from clone MS4-F4 against that of the qacEA1 gene
described in A. baumannii (GenBank accession no. AAK72475). The
positions of identical residues are contained inside black blocks while
similar residues are highlighted in gray and mismatches are in white
spaces. The alignment was created with the Bioedit Sequence Alignment
Editor software (Version 5.0.9, Hall, 1999).

51

hypothetical/putative

/\/\r,io ABR function
(7)

     
  

quaternary
A . ammonium
- drugs resistance
( aminoglyCOSIde

resistance
(39)

GTP-binding
(2)

mtion X '
transporters nbosomal

(4) f protection
)

   
 

Figure 2.15. Abundance of integron-encoded ABR gene cassettes
(panel A) and RPP determinants recovered from clone libraries
corresponding to the sampling of site 21 after four weeks of manure
application (panel B).

52

Molecular characterization of integron-encoded integrases.
Degenerate primers homologous to integron-encoded integrases were used to
evaluate the diversity of these elements after one and four weeks of manure
application in agricultural soil in which class1 integrons ABR genes were
previously detected (sample 21). A non-agricultural soil was used as a
reference. Additionally, DNA from a second crop field sampled two days after
manure application (site 3, Table 2.2) and its respective reference soil (site 4,
Table 2.2) were analyzed. The primers used were complementary to conserved
regions among class 1, class 2, and class 3 integron integrases (Figure 2.16).
Manure-supplemented and reference soils yielded positive results for the
presence of integrons. A PCR product of 500 bp was amplified from DNA

extracted from these sources.

F. hep35

I Intl? attl m attC .
._..__J

hep36+—I

Figure 2.16. Hypothetical integron structure scheme depicting the binding site of
degenerate primers used for the detection of integron-encoded integrases. These
primers were designed for the amplification of a 500 bp amplicon by targeting
conserved regions located approximately at the middle and close to the downstream
end of the integrase genes of class 1, class 2 and class 3 integrons (White et al.,
2000).

    

 

 

3110 " attC ORF1 ]0RF2 ]

 

53

Twenty five unique integrase sequences were identified among clone
libraries constructed for each of the sites analyzed. Manure—supplemented
fields presented integron variants that were highly related among themselves.
Nine unique variants were retrieved from the sampling site no. 3. Five of these
variants were closely related among each other and to a class 2 integrase
(>90% amino acid identity) described in a Shigella sonnei isolate (Figure 2.17).
The remaining four integrase types detected at this location were significantly
different from any other integron-encoded integrase described in the publicly
available databases (_5 65% amino acid identity). However, clone MS4A1-INT
was grouped within a cluster containing representative sequences of class 1
and class 3 integrons. lntegrases recovered from sample no. 21 were almost
identical (97-100% identity) and are represented by clone 23lnt-C12 (Figure
2.17), which clustered as a sister lineage relative to the class 2 integron
assemblage. However, it was observed that integrons became more diverse
after four weeks of manure application at site no. 21, since four variants that
clustered away from the class 2 group were detected at this time point.

lntegrons were more diverse in samples not exposed to animal waste.
Six variants were recovered from sample 5 that were very divergent from other
known integrases (_<_ 65% amino acid identity). None of these variants
presented a close affiliation to any particular integrase group. A comparable

tendency was observed among integrase sequences found in sample 22.

54

Inti2 Shigella wnnel AA 172891

70
rl_I::154132 INT 0
MS4E2 INT 0 Class 2

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

I: MS4A02 INTO lntegron
100 M8402 INTO Group
— — 251m 03 O
100
* M8402 INTO
68
231m 012 O
59 Int! Vibrio mlmicus AAD55407.2
Super
5NMInt F6 0 lntegron
251m 05 O lntegrase
Int uncultured bacterium AAN16072
M8451 INT.
MS4D1 INT 0
MS4F1 INT.
Int! Xanthomonas cam earls AAM39663 Super
p lntegron
hfl Methylobaclllus flagellatus ZP 0017 lnteg rages
ht! Geobacter metallireducens ZP 002991
5lnt H11 0
24mm @ Class 3
Int!!! laebalella pneumoniae AA 032355 / lntegron
MS4A 1 INT.
5NMlnt H8 o
In!” Pseirdomonas aeruginose AF263519 Class 1
WOLF-m" Escherichia coli AA 016665 lntegron
69 mm Salmonella enterica CA882887.1 GIOUP
59 25lnt F10 0
24m: A3 @
5NMlnt C3<>
Inti6 uncultured bacterium AA K00307
5NMInt £12 0
241m F9 ©
100 |—— 241m 03 ©
, Inna uncultured bacterium AF314189
"I42 I 5NMInt 59°
46 Int! uncultured bacterium AA P37597
lnIl7 uncultured bacterium AF314190 311997 0 05
/ lntegron -

 

lntIPac Pseudomonas alcaligenes AAK73287 lntegrase

s—aL 5NMIntC8 0

Figure 2.17. Relationship of environmental and known integron-encoded integrases. O =
site 21, one week after manure application, 9: site 22, no manure, reference soil for site 21,

9= site 3, two days after manure application, 9: site 21, four weeks after manure
application, <>= site 4, no manure, reference soil for site 3.

 

 

55

DISCUSSION

Ribosomal protection determinants were only detected in two different
composite soil samples to which manure from swine fed with tetracycline as a
growth promoter was recently applied (sites 3 and 21, Table 2.2). Soil samples
nearby but from where no manure was used yielded no tet PCR products.
Negative results were also obtained for samples from all agricultural soils that
had not received manure for at least four months. Amplicons from the manured
soil were of the expected size (1.3 Kb) and coded for an amino acid sequence
consisting of approximately 440 residues related to known RPP genes and
contained a GTP-binding domain architecture very similar and consistent to that
of previously described RPP. Ribosome protection genes have been shown to
encode peptides with N-terminal having amino acid sequence related to
translation elongation factors EF-Tu and EF-G which also possess GTP-binding
and hydrolysis activity (Aminov et al., 2001).

The branching patterns of the amino acid sequence-based dendrogram
demonstrated the existence of fluctuating degrees of relatedness within
sequences clustered with respect to a specific RPP class, indicating that a class
is composed of different variants. Clones designated as 341, 362, 434, 466,
and 385 were closely related to several members of the Tet M family (95-98%
identity). Clones 341 and 362 were very similar to Tet M genes detected in
Gardnerella vagina/is, a Gram-variable bacterium associated with bacterial
vaginosis (Huang et al., 1997), and to a Tet M allelic variant identified among

strains affiliated to the Bacillus cereus group which were recovered from

56

manure and manure-supplemented fields (Agerso et al., 2002). Tet M variants
reported in Entrococcus faecalis (F lannagan et al., 1994), an agent commonly
involved in nosocomial infections (Burdet, 1990; Mundy et al., 2000) and
Streptococus pneumoniae (Martin et al., 1986) a major cause of pneumonia,
meningitis, and otitis media (Doherty et al., 2000) also resembled sequences
included in this cluster. Clones 434, 466, and 385 appeared as a sister group
within the Tet M cluster while clones 32 and 356 were closely related to Tet O
determinants found in Streptococus pneumoniae (Widdowson et al., 1996) and
in Campy/obacterjejuni (Manavathu, et al., 1988), the most commonly reported
bacterial cause of food borne infections(Flannagan et al.,1994), in the United
States (Altekruse et al., 1999).

Clones 397 and 476 were retrieved from different clone libraries derived
from site 21 but presented identical amino acid sequences. These clones had
an amino acid sequence identity of 71% with respect to the Tet Ol32l0
determinant, which is their closest match and was discovered in a Clostn'dium-
related human colonic anaerobe (Melville et al., 2001). Clones 356, 314, and
319 were related to Tet O (100% identity), Tet Q (100% identity), and Tet 36
(100% identity) respectively. Tet Q was originally reported by Nikolich et
al.,1992 in a clinical strain of Bacteroides thetaiotaomicron. The Tet 36 class
was originally described by Whittle et al., 2003 in swine manure isolates
belonging to different phylogenetic lineages (Cytophaga-Flavobacter-

Bacteroides group, Gram-positive bacteria, and Gram-negative proteobacteria).

57

Clone 492 presented an amino acid sequence identity of 85% relative to
a hybrid RPP and was an abundant genotype recovered from recently fertilized
soil. The occurrence of novel interclass hybrid genes has been recently
reported by Stanton and coworkers (2003, 2004). These studies described the
presence of mosaic RPP genes in swine—associated strains of Megasphaera
elsdenii, a commensal anaerobe prevalent in the gastrointestinal tracts of
ruminant and non-ruminant mammals, including humans. These novel genes
were designated as tet (ONV) and tet (O/W/O) and apparently emerged from
single and double crossover recombination between determinants from the
class 0 and class W groups. Moreover, M. elsdenii strains carrying hybrid tet
(ONV/O) variant presented the highest levels of tetracycline resistance (Mle
128 to > 256 pg/ml) compared to M. elsdenii isolates harboring the Tet O and
Tet W determinants. These findings lead to the reclassification of an RPP
determinant previously known as Tet 32 (Melville et al., 2001) as a recombinant
RPP containing portions of a tet (O) and an unknown gene designated as tet
(32). The designation of tet (0/32/0) has been proposed to the RPP class
previously known as tet (32) along with a reexamination of current classification
guidelines of RPP determinants (Stanton et al., 2005). Sequence alignment
analyses of clone 492 revealed that this genotype is a novel mosaic variant
homologous to Tet O, Tet W, and Tet 32. The first 303 amino acid residues of
clone 492 share a 98% identity with the Tet O/W mosaic while the remaining

135 residues are 98% identical to the Tet 32 region of the Tet 0/32/0 hybrid.

58

The analysis of composite samples retrieved from site 21 after four
weeks of manure application revealed that the frequency of RPP genes
decreased under field conditions. RPP genes were not detectable when
conducting the PCR amplification using a reaction volume of 100 pl but an
amplicon of 1.3 Kb was discernable by ethidium bromide staining when the
reaction volume was decreased to 12.5 pl. Nevertheless, the intensity of the
detected 1.3 Kb amplicon was weak suggesting low frequency of target
sequences. This was in agreement with a low representation of RPP sequences
in a clone library constructed from the DNA amplified at the time point
corresponding to four weeks. These findings suggest that tetracycline resistant
intestinal flora carrying RPP genes is viable or prevalent for a limited time
following the spread of manure. This could be attributed to differences in the
prevailing conditions between the soil, manure and the animal gut environment
as well as to competition against soil endogenous bacteria. Anaerobes in
particular would be affected when mobilized from environments lacking or
having low concentrations of oxygen (i.e. manure pits, animal gut) into a crop
field. It has been previously reported that manure-associated intestinal bacteria
persist for a limited time in manure supplemented soil (Haack and Andrews,
2000; Lau and Ingham, 2001) and that tetracycline resistance levels in farmland
are temporarily influenced by the addition of hog manure slurry (Sengelov et al.,
2002)

Subsequent plowing after manure application in addition to further soil

conditioning and cultivation can result in the dilution of the applied animal

59

waste, its associated microflora, and any residual antibiotic present. In this
study the crop field 21 was further manipulated and already cultivated when
samples corresponding to the four weeks period were collected. This might
have also contributed to a decrease in frequency of RPP genes.

In contrast to the above scenario which describes the occasional
application of animal waste to agricultural fields, Ghee-Sanford and coworkers
(2001) demonstrated that groundwater and soil in long-term contact with animal
waste leaking from manure storage lagoons can become reservoirs of bacterial
populations harboring ABR genes. Furthermore, they also found evidence
indicating the mobilization of tetracycline resistance genes into populations
residing in groundwater. Limited transfer of tetracycline resistance genes from
intestinal bacteria to soil-associated microflora has been also reported to occur
in manure-supplemented fields (Agerso, 2002).

The PCR based screening targeted to the variable region of class 1
integrons revealed the presence of two major integron populations having
different ABR gene arrays. One of these two major groups appeared to be
comprised of integrons carrying a single gene cassette encoding several
variants of aminoglycoside modifying enzymes. Nine genotypes of
aminoglycoside nucleotidyltransferases (ANT) were distinguished by
sequencing analysis. The predicted amino acid sequence of one of these
variants (clone 23164C-88) presented an 86% of identitiy relative to a plasmid
encoded streptomycin adenyltransferase described in Escherichia coli

(GeneBank accession number: NP_863002). Novel classes of aminoglycoside

60

modifying enzymes have been reported on the basis of a 5 to 9% difference in
amino acid sequence identity relative to previously described determinants
(Sandvang, 1999; White et al 2000). Hence, clone 231640-B8 appears to be a
suitable candidate to be proposed as a novel ANT determinant although its
function remains to be confirmed.

The other type of class 1 integron detected contained two genes
encoding a QAC transmembranal efflux pump, one residing at the gene
cassette insertion site and another at the 3’-CS region, serving as the target for
the reverse primer. Three variants of QAC resistance genes were identified in
clone libraries constructed from PCR product amplified from manured soils.
One variant was represented by two identical sequences retrieved from two
different fields fertilized with swine manure (clones designated as 25l64C-CS
and MS4-F4). A second variant corresponding to clone 23l640-C4 was 99%
identical to the above mentioned clones. These three sequences were clustered
in a group which presented 70% of amino acid identity relative to a qacEA1
determinant described in a class 1 integron of an Acinetobacter strain
associated with nosocomial infections (Houang et al., 2003). Clones 23l64C-
C4, 25l640-CS, and MS4-F4 presented predicted amino acid sequence 27
residues longer than that of the qacEA1 determinant which is a functional
deletion derivative of the qacE gene (Ploy et al., 1998). It has been
hypothesized that the qacEA1 gene at the 3’-CS region of class 1 integrons
could have originated from an unusual recombination event catalyzed by the

integron integrase at a secondary site, which did not involve the homologous

61

sequences typically used for the integration of gene cassettes, resulting in a
truncated but functional protein (Ploy et al., 1998). Alternatively, the insertion of
a DNA segment encoding a sulfonamide resistance gene could have also
generated a deletion in the qacE gene (Paulsen et al., 1993). The remaining
QAC genotype corresponding to clone 23l64C-E5 shared a high degree of
similarity (96% of amino acid sequence identity) with a gene cassette encoding
resistance against QAC described in an Enterobacter aerogenes strain isolated
from a human urine sample (Ploy et al., 1998).

Contrary to RPP determinants, ABR gene cassettes associated with
class 1 integrons had a tendency to persist under field conditions after four
weeks of manure application as indicated by robust amplification signals and a
high frequency of ANT and QAC sequences in clone libraries. Although genes
encoding ANT and tetracycline efflux pumps have been reported to coexist in
plasmids (Tauch et al., 2002), results from this study suggest that the detected
RPP and aminoglycoside resistance genes are not linked into the same genetic
element since RPP genes become less prevalent while ANT and QAC
determinants are still prominent under field conditions. Detecting integron-
associated ABR genes not linked with RPP determinants only in manured soil
and not in the reference soil indicates that a variety of genetic elements
harboring ABR traits not associated with tetracycline selective pressure are
present in animal wastes. Hence, the impact of antibiotic resistance associated

with agricultural practices goes beyond that of ABR mechanisms operating

62

against an obvious selective force such as the use of tetracycline as a growth
promoter.

Sequencing of cloned integron-encoded integrases showed that this
gene acquisition and expression system is prevalent and diverse across
manured and non-agricultural soils. Since exposure to antibiotics could not be
correlated with their distribution, it seems that integrons are a widespread
mechanism for bacterial evolution beyond the context of antibiotic resistance.

These results indicate that the pool of ABR genes and genetic elements
associated with their dispersal is diverse and that the ABR problem should not
be assessed solely on the basis of the persistence of culturable indicator
species or the prevalence of a specific resistance mechanism suspected to
operate under selective conditions. It was also demonstrated that ABR
determinants in manure-supplemented cropland are diverse and that potentially
novel variants can be recovered from environmental DNA. This study also
indicates that the agricultural use of antibiotics is linked to an increase in the
frequency of specific RPP genes and integron—related ABR determinants
previously reported in pathogenic species. Hence, bacterial hosts of RPP genes
and integron-encoded ABR traits introduced into the environment may have the
potential to transfer antibiotic resistance traits to pathogens or commensal
microflora residing in soil or associated with the animals. Detectable levels of
RPP genes in soil DNA appeared to be transient and associated with sporadic
but recent exposure to manure from swine receiving tetracycline as a growth

promoter. However, previous reports on the feasibility of gene transfer between

63

bacteria residing in animal waste and soil indicate that a constant exposure to
animal waste plays a significant role for maintaining and spreading ABR
resistance in the environment (Ghee-Sanford et al., 2001). Additionally, the
detection of an abundant mosaic RPP (clone 492, Table 2.4.)‘in recently
fertilized soil indicates that a sustained exposure to tetracycline may contribute
to the enrichment and increase in frequency of novel gene combinations that
could confer superior resistance traits.

In contrast, the persistence of integrons in manured and in reference
soils may influence the transfer of ABR gene cassettes residing in integrons
carried by bacteria associated with animal waste into integrons residing within
the soil bacterial community. Coliforrn bacteria have been reported to be able to
replicate in contaminated soil from tropical and subtropical regions (Desmarais
et al., 2002), which further aggravates the potential impact of the ABR problem
in terms of the establishment of reservoirs of resistance traits in the
environment and subsequent horizontal gene transfer. The use of antibiotics not
only enriches and selects for specific resistance determinants, but also selects
for mobile genetic elements which can have a wide host range. Additionally, the
ability of commensal bacteria (residing in animal waste) carrying ABR genes to
colonize and become established in the rhizosphere could be another factor
influencing the persistence of ABR traits in the environment. Hence, the spread
of animal waste into crop fields could be contributing to the dispersal and
potential establishment of ABR genes from animal production facilities into the

environment.

64

REFERENCES

Adrian, P.V., M. DU Plessis, K.P. Klugman, and S.G. Amyes. 1998. New
trimethoprim-resistant dihydrofolate reductase cassette, derV, inserted
in a class 1 integron. Antimicrob. Agents Chemother. 42:2221-4.

Agerso, Y., L.B. Jensen, M. Givskov, and MC. Roberts. 2002. The identification
of a tetracycline resistance gene tet (M), on a Tn916-like transposon, in
the Bacillus cereus group. FEMS Microbiol. Lett. 214:251-256.

Altekruse, S.F., N.J. Stern, P.l. Fields, and D.L. Swerdlow. 1999.
Campy/obecterjejuni—an emerging foodborne pathogen. Emerg. Infect.
Dis. 5:28-35.

Altschul, S.F., T.L. Madden, A.A. Schaffer, J. Zhang, Z. Zhang, W. Miller, and
DJ. Lipman. 1997. Gapped BLAST and PSI-BLAST: a new generation of
protein database search programs. Nucleic Acids Res. 25:3389-3402.

Aminov, R.l., N. Garrigues-Jeanjean, and RI. Mackie. 2001. Molecular ecology
of tetracycline resistance: development and validation of primers for
detection of tetracycline resistance genes encoding ribosomal protection
proteins. Appl. Environ. Microbiol. 67:22-32.

Amman, R.l., W. Ludwig, and K.H. Schleifer. 1995. Phylogenetic identification
and in situ etection of individual microbial cells without cultivation.
Microbiol. Rev. 59:143-169.

Arduino, S.M., M. Catalano, B.E. Orman, P.H. Roy, and D. Centrén. 2003.
Molecular epidemiology of orf513-bearing class 1 integrons in
multiresistant clinical isolates from argentinean hospitals. Antimicrob.
Agents Chemother. 47:3945-3949.

Barbosa T.M., K.P. Scott, and H.J. Flint. 1999. Evidence for recent intergeneric
transfer of a new tetracycline resistance gene, tet (W), isolated from
Butyn'vibn’o fibn'solvens, and the occurence of tet (O) in ruminal bacterial.
Environ. Microbiol. 1:53-64.

Bissonnette, L., S. Champetier, J.P. Buisson, and PH. Roy. 1991.
Characterization of the nonenzymatic chloramphenicol resistance (cmlA)
gene of the ln4 integron of Tn1696: similarity of the product to
transmembrane transport proteins. Bacteriol. 173:4493-502.

Burdett, V. 1990. Nucleotide sequence of the tet(M) gene of Tn916. Nucleic
Acids Res. 18:6137.

65

Chee-Sanford, J.C., R.l. Aminov, l.J. Krapac, N. Garrigues-Jeanjean, and RI.
Mackie. 2001. Occurrence and diversity of tetracycline resistance genes
in lagoons and groundwater underlying two swine production facilities.
Appl. Environ. Microbiol. 67:1494-1502.

Chopra, l., and MC. Roberts. 2001. Tetracycline antibiotics: mode of action,
applications, molecular biology and epidemiology of bacterial resistance.
Microbiol. Mol. Biol. Rev. 65:232-260.

Cornish-Bowden, A., 1985. Nomenclature for incompletely specified bases in
nucleic acid sequences: recommendations 1984. Nucleic Acids Res.
13:3021-3030.

Desmarais, T.R., H.M. Solo-Gabriele, and OJ. Palmer. 2002. Influence of soil
on fecal indicator organisms in a tidally influenced subtropical
environment. Appl. Environ. Microbiol. 68:1165-1172.

Diaz-Torres, M.L., R. McNab, D.A. Spratt, A. Villedieu, N. Hunt, M. Wilson, and
P. Mullany. 2003. Novel tetracycline resistance determinant from the oral
metagenome. Antimicrob. Agents Chemother. 47:1430—1432.

Doherty, N., K. Trzcinski, P. Pickerill, P. Zawadzki, and CG. Dowson. 2000.
Genetic diversity of the tet (M) gene in tetracycline-resistant clonal

lineages of Streptococcus pneumoniae. Antimicrob. Agents Chemother.
44:2979-2984.

Ebner, P., K. Garner, and A. Mathew. 2004. Class 1 integrons in various
Salmonella enterica serovars isolated from animals and identification of
genomic island SGI1 in Salmonella enterica var. Meleagridis. J.
Antimicrob. Chemother. 53: 1 004 - 1009.

Elmund, G.K., S. M. Morrison, D. W. Grant, and M. P. Nevins. 1971. Role of
excreted chortetracycline in modifying the decomposition process in
feedlot waste. Bull. Environ. Contam. Toxicol. 62129-135.

Flannagan, S.E., L.A. Zitzow, Y.A. Su, and DB. Clewell. 1994. Nucleotide

sequence of the 18-kb conjugative transposon Tn916 from Enterococcus
faecalis. Plasmid 32:350-354.

FluitA.C., and F-J. Schmitz. 1999. Class 1 integrons, gene cassettes, mobility,
and epidemiology. Eur. J. Clin. Microbiol. infect. Dis. 18:761-770.

Fluit, AC, and F-J. Schmitz. 2004. Resistance integrons and super-integrons.
Clin. Microbiol. Infect. 10:272-288.

66

Giovannoni, S.J., 1991. The polymerase chain reaction. In Nucleic acids
techniques in bacterial systematics. E. Stackebrandt. NY. John Wiley &
Sons, 177-201.

Haack, B.J. and RE. Andrews. 2000. Isolation of Tn916-like conjugal elements
from swine lot effluent. Can. J. Microbiol. 46:542—549.

Hall, RM, and CM. Collis. 1998. Antibiotic resistance in gram-negative
bacteria: the role of gene cassettes and integrons. Drug Resist. Updat.
1:109-1 19.

Hall, TA. 1999. BioEdit: a user-friendly biological sequence alignment editor
and analysis program for Windows 95l98/NT. Nucl. Acids. Symp. Ser.
41:95-98.

Hamscher, G., S. Sczesny, H. Hoper, and H. Nau. 2002. Determination of
persistent tetracycline residues in soil fertilized with liquid manure by
high performance liquid chromatography with electrospray ionization
tandem mass spectrometry. Anal Chem. 74:1509-1518.

Houang, E.T.S., Y.W. Chu, W.S. Lo, K.Y. Chu, and A.F.B. Cheng. 2003.
Epidemiology of rifampln ADP-ribosyltransferase (arr-2) and metallo-
beta-lactamase (bla(lMP-4)) gene cassettes in class 1 lntegrons in
Acinetobacter strains isolated from blood cultures in 1997 to 2000.
Antimicrob. Agents Chemother. 47: 1 382-1 390.

Huang, R., D.M. Gascoyne-Binzi, P.M Hawkey, M. Yu, J. Heritage, and A. Eley.
1997. Molecular evolution of the tet (M) gene in Gardnerella vagina/is. J.
Antimicrob. Chemother. 40:561-565.

Jana, S., and J.K. Deb. 2005. Molecular targets for design of novel inhibitors to
circumvent aminoglycoside resistance. Curr. Drug Targets. 6: 353-361.

Kreader, CA, 1996. Relief of amplification inhibition in PCR with bovine serum
albumin or T4 gene 32 protein. Appl. Environ. Microbiol. 62:1102-1106.

Kumar, S., K. Tamura, l. B. Jakobsen, and M. Nei. 2001. MEGA2: molecular
evolutionary genetics analysis software. Bioinformatics. 12:1244-1245.

Lau, MM, and SC. lngham. 2001. Survival of fecal indicator bacteria in bovine
manure incorporated into soil. Lett. Appl. Microbiol. 33:131—136.

67

Levesque, C., L. Piche, C. Larose, and PH. Roy. 1995. PCR mapping of
integrons reveals several novel combinations of resistance genes.
Antimicrob. Agents Chemother. 39: 1 85-1 91 .

Levy, S.B., L.M. McMurry, T. Barbosa, V. Burdett , P. Courvalin, W. Hillen, M.C.
Roberts, J. Rood, and D. Taylor. 1999. Nomenclature for new
tetracycline resistance determinants. Antimicrob. Agents. Chemother.
43:1523-1524.

Manavathu, E.K., K. Hiratsuka, and DE. Taylor, 1988. Nucleotide sequence
analysis and expression of a tetracycline-resistance gene from
Campy/obacterjejuni. Gene. 62: 17-26.

Martin, P., P. Trieu-Cuot, and P. Courvalin. 1986. Nucleotide sequence of the
tetM tetracycline resistance determinant of the streptococcal conjugative
shuttle transposon Tn1545. Nucleic Acids Res. 14:7047-7058.

Melville, C.M., K.P. Scott, D.K. Mercer, and H.J. Flint. 2001. Novel tetracycline
resistance gene, tet (32), in the Closm'dium-related human colonic
anaerobe K10 and its transmission in vitro to the rumen anaerobe
Bulyn'vibrio fibrisolven. Antimicrob. Agents Chemother. 45:3246-3249.

Mingeot-Leclercq, M-P., Y. Glupczynski and PM. Tulkens. 1999.
Amonoglycosides: activity and resistance. Antimicrob. Agents
Chemother. 43:727-737.

Mundy, L. M., F. Sahm, and M. Gilmore. 2000. Relationships between
Enterococcal virulence and antimicrobial resistance. Clin. Microbiol.
Rev. 1 3:51 3-522.

Nandi, S., J. Maurer, C. Hofacre, and AC. Summers. 2004. Gram-positive
bacteria are major reservoir of class 1 antibiotic resistance integrons in
poultry litter. Proc. Natl. Acad. Sci. USA. 101:7118-7122.

Nield, B.S., A.J. Holmes, M.R. Gillings, G.D. Recchia, B.C. Mabbutt, K.M.
Nevalainen, and H.W. Stokes. 2001. Recovery of new integron classes
from environmental DNA. FEMS Microbiol. Lett. 195159—65.

Nikolich, MR, NE. Shoemaker, and AA Salyers. 1992. A Bacteroides
tetracycline resistance gene represents a new class of ribosome
protection tetracycline resistance. Antimicrob. Agents Chemother.
36:1005-1012.

Nonaka L., and S. Suzuki. 2002. New MgZ+-dependent oxytetracycline
resistance determinant tet 34 in Vibn'o isolates from marine fish intestinal
contents. Antimicrob. Agents Chemother. 46: 1 550-1 552.

68

Paulsen, l.T., T.G. Littlejohn, P. Radstrom, L. Sundstrom, O. Skold, G.
Swedberg, and R. Skurray. 1993. The 3’ conserved segment of integrons
contains a gene associated with multidrug resistance to antiseptics and
disinfectants. Antimicrob. Agents Chemother. 37:761-768.

Perry lll, W.L. 2002. JavaScript DNA translator: DNA-aligned protein
translations. BioTechniques. 33:1318-1320.

Phillips, l., M. Casewell, T. Cox, B. De Groot, C. Friis, R. Jones, C. Nightingale,
R. Preston, and J. Waddell. 2004. Does the use of antibiotics in food
animals pose a risk to human health? A critical review of published data.
J. Antimicrob. Chemother. 53:28-52.

Ploy, M.C., P. Courvalin, and T.Lambert. 1998. Characterization of ln40 of
Enterobacter aerogenes BM2688, a class 1 integron with two new gene
cassettes, cmlA2 and qacF. Antimicrob. Agents Chemother. 42:2557-
2563.

Roberts, M.C.1996. Tetracycline resistance determinants, mechanism of action,
regulation of expression, genetic mobility, and distribution. FEMS
Microbiol. Rev. 19:1-24.

Rosser, S.J., and H.K. Young. 1999. Identification and characterization of
class1 integrons in bacteria from aquatic environment. Antimicrob.
Agents Chemother. 44:1 1-18.

Sambrook, J., and D.W. Russell. 2001. Molecular cloning: a laboratory manual.
3rd ed. Cold Spring Harbor, NY. Cold Spring Harbor Laboratory Press.
p.17.52.

Sandvang, D. 1999. Novel streptomycin and spectinomycin resistance gene as
a gene cassette within class 1 integron isolated from Escherichia coli.
Antimicrob. Agents Chemother. 43:3036-3038.

Sengelov, G., Y. Agerso, B. Halling-Soresen, S.B. Baloda, J.S. Andersen, and
LB. Jensen. 2003. Bacterial antibiotic resistance levels in Danish
farmlands as a result of treatment with pig manure slurry. Environ. Int.
28:587-595.

Stanton, TB, and SB. Humphrey. 2003. Isolation of tetracycline-resistant
Megasphaera elsdenii strains with novel mosaic gene combinations of let
(0) and tet (W) from swine. Appl. Envir. Microbiol. 69:3874-3882.

69

Stanton, T.B., J.S. McDowall, and MA. Rasmussen. 2004. Diverse tetracycline
resistance genotypes of Megasphaera elsdenii strains selectively
cultured from swine feces. Appl. Envir. Microbiol. 70:3754-3757.

Stanton, TB, 8.8. Humphrey, K.P. Scott and H.J. Flint. 2005. Hybrid tet genes
and tet nomenclature: request for opinions. Antimicrob. Agents
Chemother. 49:1265-1266.

Stokes, H.W., A.J. Holmes, B.S. Nield, M.P. Holley, K.M. Helena Nevalainen,
B.C. Mabbutt and MR. Gillings. 2001. Gene cassette PCR: sequence-
independent recovery of entire genes from environmental DNA. Appl.
Environ. Microbiol. 67:5240-5246.

Tauch, A., S. thker, A. PiJhler, J. Kalinowski, and G. Thierbach. 2002. The
27.8-kb R-plasmid pTET3 from Corynebacten‘um glutamicum encodes
the aminoglycoside adenyltransferase gene cassette aadA9 and the
regulated tetracycline efflux system Tet 33 flanked by active copies of
the widespread insertion sequence l86100. Plasmid. 48:117-129.

Tennstedt, T., R. Szczepanowski, S. Braun, A. Puehler, and A. Schlueter. 2003.
Occurrence of integron-associated resistance gene cassettes located on
antibiotic resistance plasmids isolated from a wastewater treatment
plant. FEMS Microbiol. Ecol. 45:239—252.

Thompson, J.D., T.J. Gibson, F. Plewniak, F. Jeanmougin, and 0.6. Higgins.
1997. The ClustalX windows interface: flexible strategies for multiple
sequence alignment aided by quality analysis tools. Nucleic Acids Res.
24:4876-4882.

White, P.A., C.J. Mclver, Y.M. Deng and W.D. Rawlinson. 2000.
Characterisation of two new gene cassettes, aadA5 and dfrA17. FEMS
Microbiol. Lett. 182: 265-269.

Whittle, G., T.R. Whitehead, N. Hamburger, N.B. Shoemaker, M.A. Cotta, and
AA. Salyers. 2003. Identification of a new ribosomal protection type of
tetracycline resistance gene, tet (36), from swine manure pits. Appl.
Envir. Microbiol. 69:4151-4158.

Widdowson, C.A., K.P. Klugman, and D. Hanslo. 1996. Identification of the
tetracycline resistance gene, tet(O), in Streptococcus pneumoniae.
Antimicrob. Agents Chemother. 40:2891-2893.

70

CHAPTER III
DEVELOPMENT OF A QUANTITATIVE PCR METHOD FOR THE
DETECTION OF TET (36) AND INFERRING FUNCTION OF PUTATIVE
RIBOSOMAL PROTECTION PROTEIN GENOTYPES RECOVERED FROM
AGRICULTURAL SOIL
ABSTRACT
The transfer of antibiotic resistance (ABR) genes from commensal
bacteria to pathogens is a major concern surrounding the widespread use of
antibiotics. Another important aspect of this subject is the prevalence and
diversity of ABR genes in non-culturable reservoirs. Tetracycline antibiotics are
extensively used as growth promoting agents in swine production. Recently, a
new ribosomal protection protein (RPP) designated as Tet 36 was described
among bacteria isolated from swine manure pits. Although the presence of Tet
36 has not been reported yet in human-related strains, the early availability of
molecular tools capable of tracking the environmental distribution of this and
other potentially novel variants should facilitate a more realistic assessment of
human health hazards. A real-time PCR assay was developed for quantification
of tet (36) in total DNA extracted from manure-amended soil (tetracycline-
supplemented swine feed). This assay was also used to determine whether two
novel RPP genotypes (previously recovered from manured soil DNA via PCR)
could possess an antimicrobial resistance function, by monitoring their
abundance in microcosms containing manured soil treated with 20 and 40 ug/g
of tetracycline. Two soil microcosms, one lacking animal waste exposure, the

other supplemented with manure but no antibiotic, were used as controls. The

novel RPP variants were also monitored with respect to tet (36). In parallel,

71

tetracycline resistant bacterial isolates were screened by PCR and colony
hybridization in order to detect culturable reservoirs of these novel RPP
sequences.

After optimization and validation of real-time PCR conditions, the target
sequences were only detected in DNA from manure-supplemented soil. No
amplification was detected in DNA from non-agricultural soil and in controls
lacking DNA. After ten days of exposure to tetracycline, the frequency of two of
the environmental RPP sequences was nearly three times higher in systems
amended with of tetracycline (40 pg/g) while the frequency of the remaining
target sequence increased over one order of magnitude. None of the recovered
isolates carried the tet (36) or the putative RPP sequences. Detecting putative,
novel tetracycline resistance sequences only in manured soils, and finding that
they increased in quantity following tetracycline amendment suggest that these
genes are functional and that novel ABR genes can be recovered directly from
environmental DNA.

INTRODUCTION

The family of tetracycline antibiotics have been extensively used for
decades in animal agriculture for prophylaxis and growth promotion leading to a
remarkable increase in the incidence of tetracycline resistant bacteria. The
characterization of the diversity, location, and dispersal of antibiotic resistance
reservoirs is a subject of research interest, since knowledge on the spread of
antibiotic resistance (ABR) among different ecosystems and potential

implications to human health is limited. The recently discovered tetracycline

72

resistance determinant tet (36) is a suitable model for the application of
environmental surveillance methods. This ribosomal protection-encoding gene
has not yet been described in strains of importance to humans, but it has been
shown to be distributed among bacteria of diverse phylogenetic affiliations
(Whittle et al., 2003). Hence, the development of molecular techniques for the
detection and quantification of emerging antibiotic resistance determinants in
the environment should advance our understanding of the dimensions of the
pool of antibiotic resistance genes, their reservoirs, and potential health
hazards. Molecular methods and DNA-based culture-independent techniques
are increasingly vital procedures to characterize microbial populations important
in issues related to public health and agricultural practices (Aminov et a., 2001,
Ghee-Sanford et al., 2001, Aminov et al., 2002 ). The advent of quantitative
real-time PCR technology has provided a highly sensitive method for the
detection and precise quantification of targeted DNA molecules. This technique
has been successfully applied to the detection of specific bacterial populations
and genes in fecal and environmental samples (Baldwin et al., 2003; Kolb et al.,
2003; Gueimonde et al., 2004; He and Jiang, 2005).

SYBR Green I is the most widely used reporter dye currently used in
quantitative real-time PCR applications and comprises a simpler alternative to
TaqMan-based systems, since it does not require the design of an internal
probe, which in some instances might not be feasible due to the lack of a highly
conserved region for hybridization in the target sequence (Baldwin et al., 2003).

In essence, this dye randomly binds to the minor groove of double stranded

73

DNA and it can be excited with blue light having a wavelength of 480 nm. SYBR
Green chemistry is applicable for monitoring the increase of the amount of PCR
product since the attached dye has a fluorescence signal more than 1000-fold
higher than that of the free form of the dye (Wilhelm and Pingoud, 2003). The
fluorescence signal emitted by dye molecules intercalated into the newly
synthesized double stranded DNA is monitored throughout the progression of
each amplification cycle, contrary to conventional PCR in which detection of
amplified product occurs at the end-point of the reaction. Real-time PCR
reactions of samples containing higher amounts of the intended target
sequence will reach a fluorescence detection threshold or baseline at earlier
cycles, while it will take longer for those samples containing lower amounts of
the target sequence to generate enough double stranded DNA to reach a
threshold level, resulting in the detection of a fluorescence signal at later cycles.
The point or cycle at which florescence is detected above the baseline is
referred as the threshold cycle (CT). Threshold cycles are detected during the
exponential phase of the reaction, where exact doubling of DNA molecules is
occurring due to optimal activity and availability of reagents (Figure 3.1). Since
the detected fluorescence is proportional to the amount of PCR product
generated, the initial amount of target molecules in an experimental reaction
mixture can be quantified relative to a standard curve consisting of dilutions of
known quantities of target DNA.

The specificity of the reaction can be assessed by dissociation curve

analyses, which can be programmed to be performed automatically by the real-

74

time thermalcycler at the end of an amplification run. Melting curve analysis
characterizes the resulting PCR product based on its melting temperature (T m).
The procedure consists of measuring the fluorescence of the amplified PCR
product while heating the sample from 60 to 95°C. A decrease in fluorescence
occurs as the SYBR green dye dissociates from double stranded DNA (Figure
3.2, panel A). The derivative of these fluorescence measurements is used to
generate a curve whose peak depicts the temperature at which the amplified
DNA melts (Figure 3.2, panel B). A consistent single peak across several
replicates indicates amplification of a pure product while the presence of
different peaks in a dissociation curve indicates that amplicons of different

sequence and/or length were generated (Figure 3.3).

10"2

 

1 0" 1
d)
o
c:
a)
o .
8 exponential phase
L
8 102-0
C

threshold
10"-1
0 5 1 0 15 2 0 25 30 35 4 0
cycle

Figure 3.1. Examples of quantitative PCR amplification curves depicting the number of
cycles plotted against fluorescence. The region of a curve corresponding to the exponential
phase of the amplification reaction and the baseline level used for detecting threshold cycles
(Ct) are shown.

75

A. 0.35—

(13 .

‘5

(12.. ith)
0.15 "“x‘

—

RFU

OJ

(105.. M\

 

 

 

l I I -_
55 60 65 70 75 80 85

temperature(°C)

E3. (135__

Oil.

0.25_ M
0.2_ ;
0.15_ /.l \‘

dRFU/dT

0.05- ,, -. - l

 

I
55 60 65 70 75 80 85
temperature(°C)

Figure 3.2. A: curve of several replicates of amplification reactions
illustrating a decrease in fluorescence as the dissociation of double
stranded DNA proceeds (panel A). B: derivative plot of the
dissociation curves of replicate reactions showing a peak at the
melting temperature of the resulting amplicons from each run.

76

dRFU/dT

 

 

 

I ‘ ' I ' ' '
55 so 65 70 75 so 85 90 95
temperature(°C)

Figure 3.3. Dissociation curve analysis depicting the detection of
multiple amplicons by differences in melting temperature.

77

This study was carried out to generate a quantitative real-time procedure
to quantify and track the tet (36) determinant along with two novel variants of
ribosomal protection proteins recovered from manure-supplemented cropland
via PCR followed by cloning into a plasmid vector. The real-time PCR
technique was also applied as an approach to acquire evidence supporting the
hypothesis that these two putative RPP are functional by monitoring their
abundance in manured-soil microcosms supplemented with tetracycline. In
parallel, the molecular characterization of tetracycline resistant cultures
isolated in a variety of media under aerobic and anaerobic conditions was
carried out in order to detect culturable reservoirs of these novel RPP
genotypes.

MATERIALS AND METHODS

Design of real-time PCR primer sets. The amino acid sequences
derived from tet (36), clone 397 and clone 492 were aligned against those of
previously described RPP genes. Regions corresponding to GTP-binding
domains were excluded while variable regions were used for the design of
primers specific for each of the targeted genotypes. All primers were designed
using the PrimerQuest and the OligoAnalyzer 3.0 programs (Integrated DNA
Technologies, Coralville, IA USA) and were synthesized by Integrated DNA

Technologies. For a description of each oligonucleotide refer to Table 3.1.

78

Table 3.1. Oligonucleotides designed for quantitative real-time PCR assays.

 

Start Amplicon Target GenBank
Pn‘mer Sequence (5'-3') G+C% Tm (°CLposition size (bp) sequence accession no.

319-1014 (FWD) ATT CCT AGT CCT GCT CTC AAA TCG 45.8 57.6 1003 163 bp tel (36) AY265739
319-14 (REV) GAG TTG CTC CAT AAA GCG AAA CC 45.8 59 1165

397-2 (FWD) CAA TCG GAA CTG TGC GGG TAT GTA 50 60.3 724 187 bp clone 397 AY265740
397-2 (REV) CGG TGT CTG TTG GGA CGA TTT CTC 54.2 60.8 910

492-8 (FWD) CAC AGA GAT GCG TAT TCC ATC CA 47.8 57.9 852 180 bp clone 492 AY265741
492-8 (REV) TCT ACC GTT GTC CGA AGT AAT 66 47.8 57.5 1031

 

Standards. PCR-amplified RPP genes were cloned into TOPO TA
vectors (Invitrogen Carlsbad, CA) and the mass of a single copy of the
construct was estimated and expressed in femtograms. Plasmid DNA
containing the target sequences was extracted from E. coli clones and
quantified by optical density at 260nm. The ratio of the total mass of plasmid
DNA per volume unit against the mass of a single copy was used to estimate
the number of gene copies per volume unit. Dilutions of purified DNA having
500 pg, 10 pg, 0.2 pg, 4 fg, and 0.08 fg per pl were prepared in DNAse/RNAse
free water (Sigma St. Louis, MO, USA) and used as standards that
corresponded to 8.3 x 107, 1.6 x 106, 3.3 x 104, 6.6 x 102, and 13 copies per pl
respectively. Quantification of target sequences was expressed in number of
copies per gram of soil

Soil microcosms. Two experimental systems were prepared with 100 g
of manure—treated soil. One was wetted with 20 ml of a tetracycline solution (20
pg/ml) while the other was amended with 20 ml of tetracycline at a ratio of 40
pg/ml. Two soil microcosms (100 g of soil), one lacking animal waste exposure,

the other supplemented with manure but no antibiotic, were wetted with 20 ml of

79

water and used as controls. All systems were mixed an incubated at room
temperature in the absence of light to prevent inactivation of tetracycline by light
exposure. Microcosms were sampled every five days for 15 days for DNA
extraction. DNA was extracted as described in Chapter II.

Quantitative real-time PCR conditions. Real-time assays were
conducted using SYBR Green I-based chemistry (SYBR® Green PCR Master
Mix, Applied Biosystems, Foster City, CA) and carried out in an ABI PRISM®
7700 Sequence Detection System (Applied Biosystems, Foster City, CA).
Amplification reactions of DNA extracted from soil microcosms were conducted
using a reaction volume of 25 pl. Each reaction contained 1X SYBR Green PCR
buffer, 1.5 mM MgCl2, 0.2 mM of dNTP mixture, 0.2 U of Amperase UNG, 150
nM of each primer, 10 pg of BSA (Roche Diagnostics Corporation, Indianapolis,
IN), 1.25 U of Amplitaq Gold, and 10 ng of soil DNA. In order to perform a
quantification assay representative of a soil scenario, amplification reactions
were carried out with standards containing the intended target sequences
spiked with 10 ng of DNA from soil not treated with animal waste in which RPP
determinants were undetectable by conventional PCR. The cycling conditions
used were as follows: initial denaturation (2 min at 98°C) followed by 40 cycles
of melting (40 sec at 95°C), annealing (40 sec at 60°C) and extension (25 sec at
60°C). Dissociation curves were performed to evaluate the specificity of the
amplification reaction. Additionally, the resulting amplification product was

cloned and sequenced to confirm the detection of the intended sequences.

80

Isolation of tetracycline resistant strains. In order to determine if
culturable reservoirs of the novel RPP genotypes (clones 397 and 492) could
be recovered, 1 g of manure supplemented soil was serially diluted (102 to 101°)
and plated in R2A, Nutrient, Todd Hewitt, and CDC blood agar media
containing 25 pg/l of tetracycline. Plates of R2A, Nutrient, and Todd Hewitt
media were incubated under aerobic and anaerobic conditions, while CDC agar
plates were incubated anaerobically only. Since tetracycline is light sensitive,
plates were protected from light exposure during the incubation period.
Anaerobic incubations were done in a glove box unit (Coy Manufacturing Corp.
Ann Arbor, MI) under an atmosphere consisting of 97% N2 (02 free) and 3% H2
(10% Nz-balanced).

Molecular screening of tetracycline resistant strains for the
presence of putative RPP. Axenic cultures recovered under antibiotic
selection were screened by PCR and colony hybridization for the presence of
the 397 and 492 clone sequences. Cultures presenting RPP amplicons were
further characterized by BOX PCR and comparative analysis of partial 16S
sequences. The detected RPP determinants were also sequenced. Genomic
DNA for PCR assays was extracted by transferring colonial growth in 0.2 ml
PCR reaction tubes containing 100 pl of molecular biology grade water (Sigma,
St. Louis, MO). Cell suspensions were placed in a thermalcycler and boiled at
95°C for 10 minutes. About 1/4 of the total volume of the PCR reaction tubes
was filled with 0.1mm zirconia silica beads (Biospec Products Inc. Bartlesville,

OK). The tubes were taped on to a Mo Bio vortex adapter (Mo Bio Laboratories

81

Inc. Carlsbad, CA) and vortexed for 5 min at full speed. Cell debris along with
zirconia beads were pelleted by centrifugation. The supernatant was transferred
into a fresh PCR tube and stored at —20°C until analyzed. PCR assays for RPP
and 16S rRNA genes. were carried out as described in Chapter II, while BOX
PCR was performed following the procedures described by (Dombek et al.,
2000)

Colony lifts, cell lysis and DNA fixation for colony hybridization
experiments were done as described by Buluwela et al., 1989. Colony
hybridizations were performed by chemiluminescence detection using the
AIkPhos Direct“ system (Amersham Biosciences, Piscataway, NJ) as
indicated by the manufacturer. Fragments of 1.3 kb were PCR-amplified from
plasmid clones containing the sequences designated as 397 and 492. The
resulting amplicons were purified by gel extraction, labeled, and used as
probes. Hybridizations were carried out at medium stringency (60°C) to allow

the detection of various RPP genotypes among the screened colonies.

RESULTS
Optimization of real-time PCR primers. The optimal annealing
temperatures were empirically determined for each primer by gradient PCR and
their respective specificity was tested by challenging the primers against an
array of different templates that lacked or had the intended target sequences
(Figure 3.4). For each primer the set of template mixtures consisted of plasmid

DNA containing the sequence from which the primers were derived, a mixture

82

of plasmid DNA having different RPP genes cloned but not the target sequence,
the target sequence mixed with other RPP clones, DNA extracted out of the
manured soil from which the putative novel RPP sequences were retrieved,
DNA from a reference soil not exposed to manure and DNA extracted from the
swine manure that was applied to the sampled fields. In all instances a single
amplicon of the expected size was observed for the templates spiked with the
target sequences and for those consisting of DNA from swine manure and
manured soils. No band was observed in negative controls lacking template
DNA or in PCR reactions where DNA from the reference soil was used.
Sensitivity of real-time PCR assay. Fifty-fold serial dilutions of plasmid
DNA containing the target sequences were used as templates in amplification
experiments carried out to evaluate the sensitivity of the method. An inversely
proportional linear relationship was observed between the log of gene copy
numbers and their corresponding CT values. For each of the tested primer sets,
fluorescence signals beyond the threshold levels were detected throughout
serial dilutions differing by seven orders of magnitude which corresponded to a

range of 13 to 8.3 X 107 gene copies (Figure 3.5).

83

M12345 678M

A:

200bp

 

M12345 678M

200bp

 

M12345678M

200bp

 

Figure 3.4. A:PCR product generated with primers specific for tet (36). B: PCR
product generated with primers specific for clone 397. C: PCR product generated
with primers specific for clone 492, M=100 bp ladder; the following numbers indicate
the nature of the DNA used as template and control ractions: 1= target sequence
used as template, 2= no template. 3=target sequence mixed with other cloned RPP
genes, 4=mixture of RPP clones lacking the target sequence, 5 and 6 = DNA from
manured soils 7=DNA extracted from swine manure (pit), 8=DNA from reference
soil lacking manure. The mixtures of cloned RPP genes contained the following
determinants: tet (M), tet (O), tet (Q) tet (36) clone 397 and clone (492). Clones 397,
492 and tet (36) were excluded appropriately in amplification reactions containing a
mixture of RPP clones as template.

40

 

tet (36)

35-
30-
25-
5201
15-
ml
5-l

 

 

 

o ‘F Y v v v T 1
1.00800 1.00801 1.00802 1.00803 1.00804 1.00805 1.00806 1.00807 1.00808

Gene copy number

 

4o clone 397
35+
30 «
25 1

5 20 ~
15 4
10 ~

5.

 

 

O r v 1 v v 1 I
1.00800 1.00801 1.00802 1.00803 1.00804 1.00805 1.00806 1.00807 1.00808

 

Gene copy number

40
35 , clone 492

 

30-
251
520-
15-l
10~
5.1

 

 

0 v T I v 1 1 1
1.00800 1.00801 1.00802 1.00803 1.00804 1.00E+05 1.00806 1.00807 1.00808

 

Gene copy number

Figure 3.5. Standard curves depicting the threshold cycle (Ct) values plotted
against the log of copy numbers of the target sequences tet (36) (panel A),
clone 397 (panel B) and clone 492 (panel C). Two replicates of each dilution
were amplified.

85

Abundance of tested RPP genotypes and validation of quantitative
assay. Quantitative real-time PCR assays revealed that clone 492 was the
most abundant of the three tested genotypes (nearly 20,000 copies per gram of
soil) in the manured soil sample followed by clone 397 and tet (36) with about
2,000 and 1,000 copies per gram of soil respectively. An increment of nearly
two orders of magnitude in copy number was observed after 15 days of
incubation for the tet (36) genotype in the manured soil microcosm lacking
tetracycline. The observed frequency of tet (36) in the microcosm supplemented
with 40 pg/g of antibiotic was slightly lower than that reported in the manured
soil lacking tetracycline. The tet (36) determinant was not detected in soil not
treated with manure (Figure 3.6) After ten days of exposure to tetracycline, the
frequency of sequences 397 and 492 was nearly three times higher in systems
amended with 40 ngg of tetracycline with respect to a control consisting of
manured soil lacking antibiotic (Figures 3.7 and 3.8). None of these two
sequences were detected in non-agricultural reference soil. Except in the case
of clone 492, no significant increase in the frequency of the analyzed genotypes
was found when supplementing soil microcosms with 20 pg/g of tetracycline
(Figure 3.8).

Melting curve analysis demonstrated that in all cases a single amplicon
was generated. In all instances no amplification product was detected in control
reactions containing 10 ng of DNA extracted from soil not treated with animal
waste and in those lacking template DNA. The PCR product produced with

primers corresponding to tet (36) presented a melting temperature of 758°C

86

(Figure 3.9), while amplicons generated with Oligonucleotides targeting
sequences 397 and 492 presented melting temperatures of 77.5 °C and 78.1 °C
respectively (Figures 3.10 and 3.11). Sequence analysis of the amplicons
generated during the quantitative real-time PCR run further validated the
specificity of the primers. Each primer set yielded an amplicon corresponding to

the targeted regions (Figures 312-314).

 

   

 

 

 

100000
tet (36)
10000 - —— ——-- _ --—-—
"i3.
'6 1000 - I no manure
9 I manured
g. 100 - El manured + tet 40pg/g
8
10
1

 

 

 

 

 

 

 

 

 

days

Figure 3.6. Quantification of tet (36) in soil microcosms at five days
intervals. The frequency of the target sequence was monitored in
three different scenarios: (i) non agricultural soil lacking manure and
tetracycline, (ii) agricultural soil supplemented with swine manure
(no tetracycline) and (iii) manure-supplemented agricultural soil plus
tetracycline (40pg/g).

87

1 2000

 

10000

8000

6000 «

copieslgram of soil

2000 ~

4000 -l -

 

clone 397

 

 

 

 

 

I no manure
I manured
c] manured + tet 40pglg

Figure 3.7. Quantification of clone 397 in soil microcosms at five days
intervals. The frequency of the target sequence was monitored in three
different scenarios: (i) non agricultural soil lacking manure and
tetracycline, (ii) agricultural soil supplemented with swine manure (no
tetracycline) and (iii) manure-supplemented agricultural soil plus
tetracycline (40pg/g).

88

copies/g of soil

60000

 

clone 492

50000 ~
40000 , ~ 7
30000 -
20000 .

10000 ~

 

 

 

no manure no tet manured no tet manured tet 20 manured tet 40

microcosm systems

Figure 3.8. Quantification of clone 492 in soil microcosms after 10 days of
exposure to 0, 20, and 40 pg of tetracycline per gram of soil.

89

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1.55-01. ...._-. . 83118.3,
105-01. A - /./
5.0E-02 /
0.0E+00 M“‘*""'“‘"""” "T"? . \e........-........-.......--._..--.-- .---..__-._..
70 75.8 80 90
155-01 -.
1.05-01 B N
.0E-02 -- .. .
70 75.8 80 90
1.5E-01. . w
1.05-01 C ‘
5.05-02- -- .
70 75.8 80 9‘0
1.5E-01
10501 D
5.05-02
0.0E+00 , . _
70 75.8 80 90

 

 

 

Figure 3.9. Dissociation curve analysis of real time PCR reactions carried out with 0.2
pg of plasmid DNA containing the tet (36) sequence (panel A); 10 ng of template DNA
extracted from manured soil (panel B); 10 ng of template DNA extracted from non-
agricultural soil (panel C); no template DNA added (panel D).

90

 

 

1.5E-01
1.0E-01 , . ~
5.0E-02~ -
0.08-00

 

 

 

 

 

 

 

90

 

 

1.55-01 -. _.
1.0E-01 B. 1
5.0E-02~

 

 

 

 

0,0500%, W

 

 

 

90

 

 

 

1.5E-01 ,_

1.0E-01 ,
5.0E—02.. .

 

0.0E+00.. . . , 5,;

 

 

 

 

70 77.5 30 90

 

 

 

1.5E-01 . - 7

105-01. .9
5.05-02

 

 

 

0.0500,, .4 _ .. ”g.

1
i

 

 

 

 

70 77.5 30 90

 

 

Figure 3.10. Dissociation curve analysis of real time PCR reactions carried out with
0.2 pg of plasmid DNA containing the putative RPP sequence 397 (panel A); 10 ng
of template DNA extracted from manured soil (panel B); 10 ng of template DNA
extracted from non-agricultural soil (panel C); no template DNA added (panel D).

91

 

1.5E-01

0.0E+00

1.05-01‘ ‘
5.05-02“ , _

 

 

V

 

 

 

 

 

70

78.1 80 90

 

 

 

1.5E-01
1.0E-01 _
5.0E-02
.0E+00

 

 

 

 

 

70

78.1 80 90

 

 

1.5E-01
1.0E-01

0.0E+00

 

5.0E-02 . ‘-

 

 

70

78.1 80 90

 

 

 

1.0E-01
5.0E-02

 

.0E+00 ,—.

 

1.5E-01, V

 

l

 

70

 

78.1 80 90

Figure 3.11. Dissociation curve analysis of real time PCR reactions carried out with
0.2 pg of plasmid DNA containing the putative RPP sequence 492 (panel A); 10 ng of
template DNA extracted from manured soil (panel B); 10 ng of template DNA
extracted from non-agricultural soil (panel C); no template DNA added (panel D).

92

 

 

 

 

amplicon
tet (36)

amplicon
tet (36)

amplicon
tet (36)

amplicon
tet (36)

961

38
1021

 

98
1081

 

158 ------------------------------------------------------- 161
1141 IACGGGAGGTGATTTTGACTTTATTGGAAGAGAGATTTTCGGTAGATGCTTACTTT 1200

 

Figure 3.12. Alignment of nucleotide sequences corresponding to the
amplicon generated under quantitative real-time PCR conditions and the
tet (36) sequence. identical base positions are highlighted in black. Only
positions with base calls having a quality score of > or = 20 (less than a 1
in 100 chance that the base call is incorrect, as determined with the phred
algorithm, Ewing et al., 1998) were used in the alignment.

93

amplicon 1 ----------------------------- 31
clone 397 781

amplicon 32
clone 397 841

 

amplicon 92 --------------------------------------------------- 99
clone 397 901 GCTGATTCCGGGGAGATTGTTATTTTATCTGACAATACATTAAAACTAAAT 960

Figure 3.13. Alignment of nucleotidesequences corresponding to the
amplicon generated under quantitative real-time PCR conditions and that
of clone 397. Identical base positions are highlighted in black. Only
positions with base calls having a quality score of > or = 20 (less than a 1
in 100 chance that the base call is incorrect, as determined with the phred
algorithm, Ewing et al., 1998) were used in the alignment.

94

amplicon
clone 492

amplicon
clone 492

amplicon
clone 492

amplicon
clone 492

850 860 870 880 890 900

1

 

 

17o ------------------------------------------------- 179
102 1 -GCCGCAAAAGCCGGAGCAAAGGGAAGCCCTGTTAAATGCCCTCGCAGAG 1 o 8 0

Figure 3.14. Alignment of nucleotide sequences corresponding to the
amplicon generated under quantitative real-time PCR conditions and that
of clone 492. Identical base positions are highlighted in black. Only
positions with base calls having a quality score of > or = 20 (less than a 1
in 100 chance that the base call is incorrect, as determined with the phred
algorithm, Ewing et al., 1998) were used in the alignment.

95

Molecular characterization of tetracycline resistant isolates. A total
of 83 cultures were selected from isolation experiments on R2A, Nutrient and
Todd Hewitt agar media containing 25 pg of tetracycline per ml grown under
aerobic conditions and screened by PCR. Genes encoding RPP were detected
in eleven of these strains. Three distinctive BOX PCR patterns (A, B, and C)
were identified within this subgroup (Figure 3.15)

Strains designated as, NA25-5, NA25-6, and TH10-10 were selected for
sequencing analysis of RPP and 168 rRNA genes. Comparative analysis of
partial 16$ rRNA sequences revealed that isolates NA25-5, NA25-6, and TH10-
10 were related to Corynebacten'um ammoniagenes, Paenibacillus lautus, and
Enterococcus saccharolyticus, respectively. The tet (M) RPP determinant was
detected in strains NA25-6, and TH10-10 while tet (O) was found in isolate

NA25-5. (Figure 3.16)

96

 

 

:15.” 5.

 

 

 

Figure 3.15. BOX PCR patterns (A, B, and C) of
aerobically grown tetracycline resistant strains isolates
which also presented an RPP amplicon. 1 = strain
NA25-5, 2 = strain NA25-6, 3 = strain NA25-12, 4 =
strain NA25-18, 5 = strain TH10-10, 6 = strain TH50-12.

97

 
  
   
 

 

 

100

 

 

 

100

 

 

100

 

 

95 Enterococcus flavewens Y18294

Enterococcus gallinarum AFD39898
Enterococcus faeclum Y18294
Enterococcus saccharolyticus AF061004

—coce TetM (_—

1 82—w0-2 TetM (—

 

11110-12 Tet M (—

Paenibacillus azotofixan X77848
Paenibacillus polymyxa AJ223988
Bacillus vortex AF039409

Paenibacillus Iautus D78473

90 NAzs-s Tet M (.—
CIosIndlum hIMIOnIS A8023971

Clostn'dum dfliclle AF072474
.— Peptostreptococcus arraembius A Y326462

CDC-1TetM {-—

 

100 '- Peptoniphilus indolicus AY153431

 

 

100' .
100

 

7- Peptoniphllus asaochaclyrrcus AF54228

— airline mature bacterium AY167907
— coo-16 Tet M ('—

Mycobacten’um ratidaonense AF055331
100 I

 

 

 

 
 
  

Corynebacmn'um segmentasrm X84437

Corynebacterium themssenii AF10474
~— Corynebactarium ammoniagenesX82056

 

99 ———NA25-5 TetO (.—

 

0.02

Escherichia coli Z8320!

Figure 3.16. Neighbor joining tree based on partial sequencing analysis of 16S
rRNA genes depicting the genetic relationship of tetracycline resistant isolates
carrying ribosomal protection protein genes. GenBank accession numbers
precede the designation of reference strains.

98

The colony hybridization technique was used for the large scale screening
of RPP hosts cultured anaerobically in CDC blood agar. Four unique BOX-PCR
profiles were detected among 90 colonies presenting hybridization signals.
These were strains CDC-1, CDC-2. CDC-3, and CDC-16. Isolate CDC-1 carried
a tet (M) gene and was affiliated to the Peptostreptococcus group. Cultures
designated as CDC-2 and CDC-3 were related to the enterococci group and
also carried the tet (M) gene. A swine manure isolate related to the
Peptoniphilus group was the closest relative of strain CDC-16 which also
contained a Tet M class RPP.

PCR Screening of isolates recovered under anaerobic conditions using
R2A, Nutrient, and Todd Hewitt agar media resulted in the identification of RPP
hosts harboring the tet (M) determinant that were isogenic to strains TH10-12
and CDC-1. The tet (M) and (O) determinants detected among the isolated
populations presented over 95% of amino acid identity to previously reported
RPP.

DISCUSSION

The 16S rRNA gene characterization of tetracycline resistant isolates
revealed that these cultures were related to Gram-positive bacteria. None of the
novel RPP genotypes were detected among the recovered bacterial populations
despite incubations being carried out under aerobic and anaerobic conditions
and the use of several media types targeting oligotrophs (R2A agar),
heterotrophs (Nutrient Agar) and fastidious organisms (Todd Hewitt and CDC

blood agar). The observed moderate diversity among resistant strains and the

99

determinants that they harbored seems to be due to enrichment bias and
limitations of culture techniques in fulfilling the growth needs of only a fraction of
the total populations comprising a microbial community. It is also possible that
bacteria carrying RPP genes different from those detected confer resistance
against tetracycline concentrations lower than those used for isolation in this
study (25 pg/ml). Although only Gram-positive bacteria were found to carry
either tet (M) or tet (O), the tet (M) gene was distributed among different
phylogenetic groups indicating possible lateral transfer among bacterial
populations. These results are in agreement with previous reports showing the
predominance of RPP genes among Gram-positive bacteria and their
association with conjugative transposons (Chopra and Roberts, 2001). Several
isolates that did not yield a PCR product with the RPP-specific primers
presented a resistance phenotype indicating the presence of alternative
resistance mechanisms, presumably efflux pumps, which are known to have
nucleotide sequences unrelated to those of RPP genes and are prevalent
among enterics (Chopra and Roberts, 2001).

PCR screening was a more efficient approach for detecting diverse hosts
of RPP genes compared to colony hybridization. Although the colony
hybridization method allows in principle the screening and processing of
virtually all colonies recovered from agar plates, the efficiency of the technique
appears to be compromised when applied to environmental samples containing
various populations with different colony textures. For instance, raised and

moist colonies had a higher binding capacity to the nylon filters when

100

performing the colony lifts that those having a dry and hard texture. This
imparted an initial limitation to this type of screening. Additionally, variability in
lysis susceptibility among the tested colonies can also affect the detection of
nucleic acids that are still compatible with the used probe, but not present in
enough quantity to generate a signal above background levels. In contrast, the
combination of boiling and bead beating allowed the recovery of DNA from
Gram-positives including those having a dry and hard colony texture as that
presented by isolate NA25-5.

The characterized axenic cultures were affiliated with the
Corynebacten'um, Paenibacillus, Enterococcus, Peptoniphilus and
Peptostreptococcus genera. These genera have been described as prevalent in
farm environments as well as in the microflora of human and animal stools
(Cooke and Keith, 1927; Farrow, 1984; Ezaki et al., 2001; Leung and Topp,
2001) and hence may have the potential for transfer of resistance traits to
clinical strains, since tetracycline resistance has been also reported in human
and animal pathogens taxonomically related to the above mentioned groups
(Rockhill et al., 1982; Bentorcha et al., 1991; Evans, 2003).

During the last 20 years, the frequency of nosocomial infections resulting
in life-threatening conditions caused by multiple-antibiotic resistant strains of
enterococci has increased. Clinical infections are typically caused by at least 12
species, but particularly Enterococcus faecalis and E. faecium. In rare instances
E. gallinan’um has been described as the causative agent of clinical infections

and E. saccharolyticus may possibly be among the group of infectious strains

101

(Mundy et al., 2000). Isolates TH50-12, CDC-2 and CDC-3 were grouped close
to E. saccharolyticus, E. faecium and E. gallinan'um. In contrast, strain NA25-5
clustered next to Corynebacterium ammoniagenes, a non-pathogenic strain that
has been reported in smear-ripened cheeses, feces of infants and piggery
wastes. Corynebacteria are well known, abundant soil microflora indicating that
antibiotic resistant members of this group might be reservoirs of antibiotic
resistance. Paenibacillus has been reported to be one of the most abundant
phylotypes in swine manure incubated aerobically under laboratory conditions
(Leung and Topp 2001).The presence of tetracycline and vancomycin
resistance determinants has been described in species affiliated to this genus
(Evans, 2003; Guardabassi et al., 2004). Paenibacilli are also effective
rhizosphere colonizers and their persistence and survival in soil after manure
applications can be enhanced by their ability to form endospores and their
facultatively anaerobic physiology, potentially allowing their establishment as a
reservoir of antibiotic resistance determinants. Strain NA25-6 is closely related
to Paenibacillus Iautus.

Isolates CDC-1 and CDC-16 which carried the tet (M) determinant
clustered within the Peptostreptococcus and Peptoniphilus groups, respectively.
Species of the genus Peptostreptococcus are anaerobic, non-spore-forrning
bacteria, which are widely distributed among the normal microflora of mucosal
surfaces and skin in humans (Hill et al., 2002) and are also prevalent in animals
(Salanitro et al., 1974; Robinson et al., 1981). Clinical strains have been

recovered in cases of infections of soft tissues, lungs, bones, and the female

102

genital tract (Brook et al., 1994; Murdoch et al., 1997; Song et al., 2003). Isolate
CDC-1 was related to Peptostreptococcus anaerobius, the type species of the
genus Peptostreptococcus. Peptostreptococcus anaerobius has been
commonly linked to poiymicrobial infections and isolated from diverse clinical
specimens (Riggio and Lennon, 2002). Tetracycline resistance conferred by
efflux pumps (Tet K) and RPP determinants (Tet M and Tet O) has been
previously reported in PeptoStreptococcus isolates (Roberts, 1991).

The taxonomy of the genus Peptostreptococcus has been recently
revised resulting in the subdivision of this group into three new additional
genera designated as Anaerococcus, Gallicola and Peptoniphilus (Ezaki et al.,
2001). Peptoniphilus asaccharolyticus, and Peptoniphilus indolicus along with a
hyper-ammonia producing swine manure isolate (Whitehead and Cotta, 2004)
were the taxa most related to isolate CDC-16. Peptoniphilus asaccharolyticus
strains have been frequently cultured from specimens of vaginal discharges and
ovarian and peritoneal abscesses while Peptoniphilus indolicus was originally
isolated from summer mastitis of cattle (Ezaki et al., 2001).

The isolation of tetracycline resistant bacteria related to clinical,
opportunistic, and human commensal strains, suggests that the agricultural use
of tetracycline promotes the prevalence of resistance traits in microbes
associated with animal waste that might be linked to human health hazards.
Moreover, the observed intergeneric distribution of the tet (M) determinant

across Gram-positive bacteria suggests that lateral gene transfer is influencing

103

the dispersal of resistance traits and that the mobilization of these genes into
pathogens could be feasible.

A quantitative real-time PCR protocol was developed for the detection of
selected RPP. The tet (36) determinant was chosen because it is the most
recently described class of the RPP family to date and has been described only
in commensals isolated from a swine production facility, constituting a suitable
model for the environmental surveillance by molecular means of a novel gene
which might be a potential health risk. Since the RPP genotypes 397 and 492
were not detected in any of the retrieved cultures, the quantitative PCR
approach was also applied as an alternative, culture-independent method to
compile evidence indicating an antibiotic resistance function in these genes.

Each of the primer sets was capable of detecting their respective target
sequence in preliminary tests carried out under conventional PCR conditions
and ethidium bromide-based detection in agarose gels. Detecting the putative
tetracycline resistance sequences and the tet (36) variant only in manured soils
and in swine manure indicates that the presence of these genes was
specifically linked to the animal waste and not to the fraction of the total DNA
co-extracted from the soil. Hence, this conventional PCR assay allowed tracing
the source of the tested determinants by the molecular screening of
environmental samples.

The specificity and sensitivity of the real time assay was confirmed by
the high correlation coefficients (r2 > 0.99) obtained for each standard curve and

the detection of as few as 13 copies of each target gene in serial dilutions of

104

cloned RPP sequences spiked with 10 ng of soil DNA in which RPP genes were
not detectable. To ensure that quantification of target genes was carried out
under antibiotic selective pressure and to compensate for sorption of
tetracycline to soil and organic matter, microcosms were amended with
tetracycline concentrations two (20 pg/g of soil) and four (40 pg/g of soil) times
higher than that previously reported by Haack and Andrew, 2000 (10 pglml of
medium) for the enrichment and isolation of tetracycline resistant bacteria from
swine effluent. The amount of tetracycline added to soil microcosms was also
approximately three orders of magnitude higher than that reported by Smith et
al., 2004, which was correlated with an increase in the frequency of RPP genes
in feedlot lagoons (1.95 pgll of bioavialable tetracycline as estimated by ELISA).
After 10 days of incubation, mycelial-like growth was evident in microcosms
supplemented with 40 pg/g of tetracycline and absent in the others, indicating a
shift in the composition of the microbial community.

The starting gene copy numbers of the tested RPP genotypes were
calculated by comparing C, values from the microcosm samples with those from
the standard curve. Approximately 1000 gene copy numbers per gram of
manured soil of the tet (36) determinant were detected at the beginning of the
incubation period. This gene reached a frequency of nearly 80,000 copies after
15 days. Under tetracycline selective conditions tet (36) reached a frequency
close to 50,000 gene numbers. This lower copy number appears to be related
to a slower growth rate of tet (36) hosts due to stress imparted by the presence

of tetracycline. The RPP resistance mechanisms requires GTP hydrolysis to

105

remove tetracycline molecules bound to the bacterial ribosome (Chopra and
Roberts, 2001), hence a higher demand on energy-releasing substrates could
have affected the growth of resistant populations.

The frequency of clone 397 was stable during the first 10 days but
increased from about 2,500 copies to 10,000 copies in the microcosm that
lacked tetracycline. in contrast, a significant increase in copy numbers was
noticeable in the tetracycline—amended microcosm after 5 and 10 days of
incubation when this sequence reached a frequency of 3,000 and 6,000 copies,
respectively. This increase in copy numbers indicates that populations carrying
this determinant were able to grow in the presence of tetracycline, which is
consistent with the implication that clone 397 confers a resistance function
against this drug.

The frequencies of clone 492 were basically the same under tetracycline
selection and antibiotic-free conditions during the incubation period except that
by the sampling point corresponding to day ten, higher gene copy numbers
were detected in both tetracycline supplemented systems relative to the control
microcosms which were not amended with the antibiotic. After 10 days of
exposure to tetracycline frequencies of 32,000 and 54,000 gene copy numbers
were detected in the microcosms containing 20 and 40 pg of tetracycline per
gram of soil respectively, while the target sequence was undetected in the
microcosm prepared with non-agricultural soil and was present in the manured
soil microcosm that lacked antibiotic at a frequency of nearly 22,000 gene

copies per gram of soil. The observed correlation between an increase in the

106

frequency of clone 492 associated with an increase in antibiotic exposure
indicates that this determinant could also posses an antibiotic resistance
function. This trend was evident for each of the tested sequences. However the
quantification assays revealed that each sequence behaves differently under
tetracycline selective conditions. The tet (36) determinant remained prevalent
after 15 days of exposure to the antibiotic and its frequency increased by over
one order of magnitude, while the frequencies of clone 397 and clone 492 had
their respective highest numbers after 10 days and then declined. This indicates
that these determinants may reside in different hosts that respond differently to
other ecological factors as environmental changes and competition with other
antibiotic resistant populations.

It can not be ruled out that the putative RPP determinants tested in this
study might reside along with other tetracycline resistance genes inside the
same host. For instance efflux pump and RPP determinants have been
detected in combination among 4% (n = 295) of Enterococcus faecium strains
isolated from chicken caecum samples (Petsaris et al., 2005). Hence, if this
scenario were true, complementation of these findings with in vivo tests would
be necessary to confirm the function of putative RPP genes. Construction of
metagenomic libraries might be an alternative approach for the recovery of full
length sequences of novel RPP genes detected by PCR if other isolation efforts
are not successful. Nevertheless, this study demonstrated that the real-time

PCR technique is a sensitive approach applicable to the detection,

107

quantification, and surveillance of antibiotic resistance determinants in

environmental samples.

108

REFERENCES

Aminov, R.l., J.C. Chee-Sanford, N. Garrigues, B. Teferedegne, l.J. Krapac,
B.A. White, and RI. Mackie. 2002. Development, validation, and
application of PCR primers for detection of tetracycline efflux genes of
gram-negative bacteria. Appl. Envir. Microbiol. 68:1786-1793.

Aminov, R.l., N. Garrigues-Jeanjean, and R. l. Mackie. 2001. Molecular ecology
of tetracycline resistance: development and validation of primers for
detection of tetracycline resistance genes encoding ribosomal protection
proteins. Appl. Envir. Microbiol. 67:22-32.

Baldwin, B.R., C.H. Nakatsu, and L. Nies. 2003. Detection and enumeration of

aromatic oxygenase genes by multiplex and real-time PCR. Appl. Envir.
Microbiol. 69:3350-3358.

Bentorcha, F., G. De Cespedes, T. Horaud. 1991. Tetracycline resistance
heterogeneity in Enterococcus faecium. Antimicrob. Agents Chemother.
35:808-812.

Brook, l. 1994. Peptostreptococcal infection in children. Scand. J. Infect. Dis.
26:503-510.

Buluwela, L., A. Forster, T. Bohem, and TH. Rabbitts. 1989. A rapid procedure
for colony screening using nylon filters. Nucleic Acids Res. 17:452.

Chee-Sanford, J.C., R.I. Aminov, l.J. Krapac, N. Garrigues-Jeanjean, and RI.
Mackie. 2001. Occurrence and diversity of tetracycline resistance genes
in lagoons and groundwater underlying two swine production facilities.
Appl. Envir. Microbiol. 67:1494-1502.

Chopra, |., and MC. Roberts. 2001. Tetracycline antibiotics: mode of action,

applications, molecular biology and epidemiology of bacterial resistance.
Microbiol. Mol. Biol. Rev. 65:232-260.

Cooke, JV, and HR. Keith.1927. A type of urea-splitting bacterium found in
the human intestinal tract. J. Bacteriol. 13:315-319.

Dombek, P.E., L.K. Johnson, S.T. Zimmerley, and M.J. Sadowsky. 2000. Use of
repetitive DNA sequences and the PCR to differentiate Escherichia coli

Isolates from human and animal sources. Appl. Envir. Microbiol. 66:
2572-2577.

109

Evans, JD. 2003. Diverse origins of tetracycline resistance in the honey bee
bacterial pathogen Paenibacillus larvae. Invertebr. Pathol. 83:46-50.

Ewing, B., and P. Green. 1998. Base-calling of automated sequencer traces
using Phred ll error probabilities. Genome Res. 8:186-194.

Ezaki, T., Y. Kawamura, N. Li, Z. Y. Li, L. C. Zhao, and S. E. Shu. 2001.
Proposal of the genera Anaerococcus gen. nov., Peptoniphilus gen. nov.
and Gallicola gen. nov. for members of the genus Peptostreptococcus.
Int. J. Syst. Evol. Microbiol. 51:1521-1528.

Farrow, J.A. 1984.Taxonomic studies on Streptococcus bovis and
Streptococcus equinus: Description of Streptococcus alactolyticus
sp.nov. and Streptococcus saccharolyticus sp.nov. Syst. Appl. Microbiol.
52467-482.

Guardabassi, L., H. Christensen, H. Hasman, and A. Dalsgaard. 2004.
Members of the genera Paenibacillus and Rhodococcus harbor genes
homologous to enterococcal glycopeptide resistance genes vanA and
vanB. Antimicrob. Agents Chemother.48:4915-4918.

Gueimonde, M., S. lekkd, T. Korpimaki, and S. Salminen. 2004. New real-time
quantitative PCR procedure for quantification of bifidobacteria in human
fecal samples. Appl. Envir. Microbiol. 70:4165-4169.

Haack, B.J., and RE. Andrews. 2000. Isolation of Tn916-like conjugal elements
from swine lot effluent. Canadian J. Microbiol. 46:542-549.

He J-W., and S. Jiang. 2005. Quantification of Enterococci and human
adenoviruses in environmental samples by real-time PCR. Appl. Envir.
Microbiol. 71:2250-2255.

Hill, K.E., C.E. Davies, M.J. Wilson, P. Stephens, MA. 0. Lewis, V. Hall, J.
Brazier, and D.W. Thomas. 2002. Heterogeneity within the gram-positive
anaerobic cocci demonstrated by analysis of 168—238 intergenic
ribosomal RNA polymorphisms. J. Med. Microbiol. 51:949-957.

Kolb, S., C. Knief, S. Stubner, and R. Conrad. 2003. Quantitative detection of
methanotrophs in soil by novel pmoA-targeted real-time PCR assays.
Appl. Envir. Microbiol. 69:2423-2429.

Leung, K., and E. Topp. 2001. Bacterial community dynamics in liquid swine
manure during storage: molecular analysis using DGGEIPCR of 16S
rDNA. FEMS Microbiol. Ecol. 38:169-177.

110

Mundy, L.M., F. Sahm, and M. Gilmore. 2000. Relationships between
Enterococcal virulence and antimicrobial resistance. Clin. Microbiol.
Rev.13z513-522.

Murdoch, D.A., M.D.Collins, A. Willems, J.M. Hardie, K.A. Young, and J.T.
Magee. 1997. Description of three new species of the genus
Peptostreptococcus from human clinical specimens: Peptostreptococcus
harei sp. nov., Peptostreptococcus ivorii sp. nov. and
Peptostreptococcus octavius sp. nov. Int. J. Syst. Bacteriol. 47:781—787.

Petsaris, O., F. Miszczak, M. GicqueI-Bruneau, A. Perrin-Guyomard, F.
Humbert, P. Sanders, and R. Leclercq. 2005. Combined antimicrobial
resistance in Enterococcus faecium isolated from chicken. Appl. Environ.
Microbiol. 71:2796-2799.

Riggio, M.P., and A. Lennon. 2002. Development of a PCR assay specific for
Peptostreptococcus anaerobius. J. Med. Microbiol. 51 :1097-1 101 .

Roberts, MC, 1991. Tetracycline resistance in Peptostreptococcus species.
Antimicrob. Agents Chemother. 35: 1682-1684.

Robinson, l.M., M.J. Allison, J.A. Bucklin. 198. Characterization of the cecal
bacteria of normal pigs. Appl. Environ. Microbiol. 41 :950-955.

Rockhill, R.C., Sumanno, H. Hadiputranto, S.P. Siregar, and B. Muslihun. 1982.
Tetracycline resistance of Corynebacterium diphtheriae isolated from
diphtheria patients in Jakarta, Indonesia. Antimicrob. Agents Chemother.
21 :842-843.

Salanitro, J.P., I.G. Blake, and PA. Muirhead. 1974. Studies on the cecal
microflora of commercial broiler chickens. Appl. Microbiol. 28:439-447.

Smith, M.S., R.K. Yang, C.W. Knapp, Y. Niu, N. Peak, M.M. Hanfelt, J.C.
Galland, and D.W. Graham. 2004. Quantification of tetracycline
resistance genes in feedlot lagoons by real-time PCR. Appl. Envir.
Microbiol. 70:7372-7377.

Song, Y., C. Liu, M. McTeague, and SM. Finegold. 2003. 168 ribosomal DNA
sequence-based analysis of clinically significant Gram-positive anaerobic
cocci. J. Clin. Microbiol. 41:1363-1369.

Whitehead, TR, and MA. Cotta. 2004. Isolation and identification of hyper-
ammonia producing bacteria from swine manure storage pits. Curr.
Microbiol. 48:20-26.

111

Whittle, G., T.R. Whitehead, N. Hamburger, N.B. Shoemaker, M.A. Cotta, and
AA. Salyers. 2003. Identification of a new ribosomal protection type of
tetracycline resistance gene, tet (36), from swine manure pits. Appl.
Envir. Microbiol. 69:4151-4158.

Wllhelm, J., and, A. Pingoud. 2003. Real-time polymerase chain reaction.
ChemBioChem. 421120-1128.

112

CHAPTER IV

DIVERSITY OF INTEGRON-ENCODED INTEGRASES
RECOVERED FROM ENVIRONMENTAL DNA VIA PCR

ABSTRACT

Advances in prokaryotic genomics have revealed that mobile elements
may comprise up to one fourth of an organism’s genome. Resistance integrons
(RI) are mobile elements capable of capturing and expressing multiple antibiotic
resistance genes. Early culture-based studies suggested that RI resided
primarily in enteric strains from clinical settings. Recently, Rl have been
detected in environments not exposed to antibiotics and Gram positives have
been reported as their major reservoirs. Nevertheless, most of the information
on the diversity and distribution of RI is derived from culture-based studies and
knowledge on their prevalence and diversity in the environment is still limited. In
order to assess the distribution and diversity of integrons, a molecular approach
was used for the characterization of new integron encoded integrases in total
DNA extracted from different environmental sources. Degenerate primers
targeting a conserved region among the integrase genes of class 1, class 2,
and class 3 integrons were used in a PCR-based screening of DNA extracted
from marine sediments of the Pacific and the Arctic Ocean, tropical soils '
(Puerto Rico and Hawaii) and ancient permafrost (Russia). All the samples
were positive for the presence of integrase genes except those from the
permafrost soil. The resulting PCR products were screened by sequencing and
unique derived amino acid sequences (598% amino acid identity) were selected

for comparative analyses. A selected group of sequences previously recovered

113

from agricultural and non-agricultural soil samples from Michigan were also
analyzed. A total 126 novel integron-encoded integrases were uncovered. Each
new sequence presented protein motifs specific to integron encoded integrases
and most of them were significantly different (SO—70% identical) from previously
described integron-encoded integrases. The recovery of such a diverse array of
new environmental integrons supports the hypothesis that these mobile
elements are ubiquitous among prokaryotes and that their distribution is not

restricted only to clinical multiresistant strains or pathogens.

INTRODUCTION

During the past few years, advances in bacterial genomics have
provided an unprecedented opportunity for the in depth study of the physiology
and molecular evolution of bacterial species. Recently, a comparative analysis
of 70 fully sequenced and closely related genomes revealed that about 65% of
the differences in gene content within species involve the presence of
bacteriophage and transposase elements, indicating that transferable genetic
material plays a significant role in bacterial speciation (Konstantinidis and
Tiedje, 2005). Additionally, a diverse array of mobile or exogenously acquired
genetic elements has been reported to comprise over a quarter of the genome
of Enterococcus faecalis V583, the first vancomycin resistant clinical isolate
reported in the United States (Paulsen et al., 2003). Although some bacteria
may be intrinsically resistant to antimicrobial drugs, the acquisition of

horizontally transferred resistance traits (rather than the occurrence of point

114

mutations in resident genes) has been described as a major mechanism
accountable for the extensive distribution of ABR determinants across
phylogenetically diverse groups (Rowe-Magnus and Mazel, 2001).
Transposons and conjugative plasmids are major dispersal mechanisms
of ABR traits. Transposons encoding ABR genes are formed when two identical
insertion elements are inserted flanking a region of DNA containing an ABR
gene resulting in the formation of a composite transposon (Figure 4.1, panel A).
If the transposase encoded by the insertion sequences acts on the inverted
repeats located at the distal ends of the composite transposon, the whole
module will be transposed mobilizing the entrapped ABR gene (Figure 4.1,
panel B). Similarly, the entrapment of an integron loaded with antibiotic
resistance gene cassettes by a composite transposon would result in the further
dispersal of not only the resistance traits but also in the dispersal of a platform
that allows its host to obtain additional resistance determinants. lntegrons
associated with transposable elements and conjugative plasmids have been
typically implicated in the dispersal of multiple resistance traits in clinical

isolates (Leverstein-van Hall et al., 2003; Fluit and Schmitz, 2004).

115

A:

 

 

 

 

 

       

 

 

IS element IS element
I l I i J I
Donor DNA .. gig: Donor DNA Donor DNA tnp p” j Donor DNA
Transposition Transposition

 

Target DNA l ABR gone ‘I Target DNA

l

Target DNA .fifi-fl; ABR gene .fi tnp “-1 Target DNA

l l
I

Composite transposon

 

 

 

 

Transposase

l

 

 

 

 

Donor DNA 811‘ g}? Donor DNA

   

ABR gene

l

Transposition into new target DNA (plasmid, chromosomal)

 

Figure 4.1. Formation of a composite transposon by the insertion of two copies of an
insertion element into the same target DNA molecule (panel A). Panel B illustrates the
mobilization of the resulting composite transposon and the captured ABR gene into a
new target DNA. Redrawn and modified from Snyder and Champness, 1997.

116

Contrary to transposons, integrons lack repeat sequences at their ends
and genes associated with self transposition. Instead, they consist of a phage—
related integrase gene belonging to the tyrosine recombinase family that
catalyzes the capture of mobile elements known as gene cassettes through a
site specific recombination mechanism which involves the integron-receptor site
(attl) and a cassette-associated attC site, also known as the 59-base element
(Hall and Collins, 1995). Gene cassettes can not be replicated or transcribed
unless they are inserted into an integron. Expression of the entrapped gene
cassette is driven by an integron-associated promoter (PC) located upstream the
integron receptor site (attl). Several gene cassettes can be stockpiled and
accommodated in tandem arrays within the integron insertion site (Chapter I).
However, those cassettes located farther away from the Pc promoter are
expressed less frequently than those closer to it due to early termination of
transcription (Collis and Hall, 1995).

Gene cassettes are circular, promoteriess DNA molecules consisting of
two components, a gene and a recombination site termed as the attC site,
which is located downstream the gene (Figure 4.2). The origin of the genes and
the attC recombination sites that constitute integron cassettes remains unclear.
However, since attC sites present complementary sequences capable of
forming stem-loop secondary structures and gene cassettes lack a promoter
sequence, it has been hypothesized that cassettes might have originated from
reverse transcribed RNA where the attC site was also serving as a transcription

terminator (Recchia and Hall, 1997). in contrast, examination of nucleotide

117

sequence data of gene cassettes has revealed that the 8th sites can be
diverse among closely related genes and vise versa, suggesting that the coding
region and the attC element are assembled from separate pools of these
genetic elements (Recchia and Hall, 1997). Moreover, the marked differences
in codon usage and G+C% among ABR determinants residing within the same
integron also supports that ABR gene cassettes have diverse origins that could
be associated with the recruitment of general house keeping genes capable of
modifying several substrates including antibiotics (Rowe-Magnus and Mazel,
1999). Resistance genes from antibiotic-producing organisms have been also
considered an alternative source for the resistance determinants detected in

integrons associated with clinical strains (Rowe-Magnus and Mazel, 1999).

 

 

coding sequence

 

 

I” 54.1 35nt ‘\\
x I
.' I l .
5’- RYYYAAC 7 variable region GTTRRRY , -3’

inverse core site core site

 

 

 

 

 

 

Figure 4.2. Diagram depicting the linear structure of gene cassettes associated with
class 1 integrons, which consists of two components, a coding segment and a specific
recombination sequence (attC).The attC site consists of an internal region of variable
sequence and length whose ends are delimited by a core and a inverse core site
which are the targets recognized by the integron-encoded integrase.

118

Based on the context in which they have been discovered and on
specific structural features, integrons have been classified as either resistance
integrons (RI) or super integrons (SI). Resistance integrons were the first to be
discovered. These were found to be the elements encoding the resistance
determinants associated with plasmids and transposons carried by antibiotic
resistant clinical strains (Stokes and Hall, 1989). By 2003, over 70 different ABR
gene cassettes have been identified within integrons (Rowe-Magnus et al.,
2003). As their designation suggests, most of the genes present in RI encode
resistance against several types of antibiotics. As many as eight ABR
determinants have been detected in a single integron carried by a multiresistant
clinical isolate of E. coli (Nass et al., 2001). Based on the similarities of the
integrase genes (39—58% identity), five classes of RI have been identified
(Rowe-Magnus et al., 2003). Besides being highly mobile due to their
association with plasmids and transposable elements, RI have been also
reported to be transferred via transduction in Salmonella enterica (Schmeiger
and Schicklmaier, 1999). Resistance integrons are also characterized by
harboring gene cassettes with variable attC sites and have been described
among Gram-negative and Gram-positive bacteria (Fluit and Schmitz, 2004).

In contrast to their RI counterparts, SI are significantly larger, not mobile,
and reside in the bacterial chromosome only. The first description of a SI was
reported in the genome of Vibrio cholerae (Mazel et al., 1998). It was found that
this element consisted of 179 ORFs from which nearly 100 represent genes of

unidentified function. Super integrons have been also described in other Viino

119

species, the pseudomonads, and strains belonging to the genera Geobacter,
Shewanella, Nitrosomonas, Treponema, and Xanthomonas (Fluit and Schmitz,
2004). The sedentary nature and stability of these superstructures has been
explained on the basis of the detection of a captured gene cassette carrying two
open reading frames (ORFs) encoding an active toxin/antidote system similar to
that described in plasmid addiction systems (Rowe-Magnus et al., 2003). The
toxin ORF presented a 42% amino acid identity to a gyrase inhibiting protein
reported in F plasmids while the antidote ORF was 42% identical to the antidote
protein found in Escherichia coli O157:H7 (Rowe-Magnus et al., 2003). Most of
the characterized Sl-associated gene cassettes have amino acid sequences
related to virulence factors, although some cassettes encoding antibiotic
resistance function have been found (Fluit and Schmitz, 2004). Contrary to what
has been observed in RI, the recombination sites of SI gene cassettes are
almost identical and seem to be species specific. Due to the high degree of
sequence conservation that these sites present, they were originally designated
as repeated sequences termed Vibrio cholerae repeats when they were first
described in this organism (Mazel et al., 1998). Moreover, in some instances it
has been found that SI cassettes may contain their own promoter and are
encoded in the inverse orientation relative to the recombination sequence
(Rowe-Magnus and Mazel, 1999).

Although several unique features have been identified between RI and SI
some structural similarities shared by these elements have been detected

which have led to the hypothesis that RI integrons evolved from SI. For instance

120

the attC sites of the Rl-encoded cassettes CARB4 and dfrVI, which confer
resistance against carbenicillin and trimethoprim respectively, are very similar to
the recombination sites of the gene cassettes found in the V. cholerae super
integron, suggesting that gene cassettes originally located in SI were removed
from their original genetic context by RI. Contrary to SI, the RI are characterized
by having gene cassettes of variable attC sites which suggests that they evolve
by capturing gene cassettes from different super integrons. Another finding
supporting the evolution of RI from SI is the fact that SI have been detected in a
Vibn'o isolate that predates the advent of antibiotics (F luit, and Schmitz 2004).
Moreover, the entrapment of a Sl-encoded integrase gene and its attl site by a
composite transposon (as explained in Figure 4.1), coupled to the selective
pressure applied by the large scale use of antibiotics for nearly half a century
could have provided the optimal scenario for the evolution of a module now
capable of moving across different bacterial species and also able to recruit the
appropriate resistance determinants from a diverse pool of Sl-associated gene
cassettes, resulting in the RI presently found in the clinical setting (Rowe-
Magnus and Mazel, 1999). The recruitment of Sl-encoded ABR gene cassettes
by RI have been demonstrated by Rowe-Magnus and coworkers (2002).
Research and surveillance on the environmental impact of the antibiotic
resistance problem have mostly focused on determinants conferring resistance
against a specific drug and on the prevalence of resistance phenotypes in
indicator species (usually pathogens). This has narrowed our vision of the true

extent of the antibiotic resistance problem since other resistance mechanisms

121

and their respective reservoirs are ignored under the above mentioned
approach. It is a known fact that commensal bacteria can be important
reservoirs of ABR traits that can be potentially transferred into pathogens (Hart,
1998). Another area that has been overlooked is the environmental distribution
and diversity of elements implicated in the active uptake, expression and
mobilization of resistance determinants which are the essence of the antibiotic
resistance problem. The sequential addition of a gene capture and expression
system such as the integron platform into mobile genetic elements like
transposons and conjugative plasmids has received special attention during the
last few years due to the high prevalence of multiresistance integrons among
clinical strains. Although there are some recent reports documenting the
existence of integrons and novel integron integrases in environmental samples
(T ennstedt et al., 2003; Nemergut et al., 2004), most of the information on the
diversity and distribution of integrons is derived from culture-based studies and
knowledge on their prevalence and diversity in the environment is still limited. In
this study a culture-independent approach was used to survey the diversity and
distribution of integrons by probing total community DNA extracted from

different terrestrial and marine habitats located around the globe.

122

MATERIALS AND METHODS

A PCR-based assay targeting a 500 bp conserved region among
integron-encoded integrases was used to survey the diversity of integrons in
different environments (Figure 4.3) Extraction of total community DNA from soil
and marine sediment samples, PCR amplification of integrase genes,
construction of clone libraries, screening of environmental sequences, and
comparative sequence analyses were conducted as described in Chapter II.
The samples analyzed and the codes used for the retrieved clones are

described in Table 4.1.

r5 hep35 (FWD)

 

lntegrase gene

 

 

 

hep36 (REV)4—I

l |
' 500 bp '

Figure 4.3. Diagram depicting the PCR assay used to amplify a 500
bp region common to integron-encoded integrases. Comparative
analysis of amino acid sequences derived from sequences retrieved
from clone libraries were used to survey the diversity of integron
encoded integrases across different environments. The primers
used were previously described by White et al., 2000.

123

 

23> coon can 523.com

 

.<.z £83... 6:3...» 5%.. En .23... .280 no» 62.2.5: emu 3
2mc> :2...Em-~ Home .8.mo.ooa

.<.z 983. .26...» 58.. an». .._o» .86. 825. 82.952 tn. m.
8.... 2.2... 2.5 0:26 a .0

>3.» ».:. 6:2». 20.2 2.» 2:8. .66: .»2o. .5 _.o» (20.2 N.

.<.z (m: .526: $2852 6:25.: 63.5.23: .6» 6:...»3 .I S
.58.. :5» :

.<.z 5.80 2.2... 25 .85.. ace... .»mw .:oE.oo» 0:..2: 9.9.. o.
.58.. :58 a

Bow ...m .6 95:20 :800 ocean. 6..» .38 56:23.5 .5586» 655:. o<>> m
565:8; . .6.:. two

68 .._o .o 95:26 880 ocean. .2» 23» 89.“. .8566» 2:2. 62. a
£33 .588

<.z :mcoo o..2< 2.» 3.68 395m .:eE.ue» 3.5.: 0%.... n
.58.. 58 :

.<.z :mooo o..2< 2.» SEE 323m Eocene» 656E 0mm... m
235E 2.3» 0.

>3.» »_:. <w: 5:95.... 2:»098 .o 9.02.. So. .6» .2:..:o...mm .53 m

. 5.82.9.6 25:2: 0: .

>3.» »_... <w: 6:95.... ..o» 8:222 .6» 23.3595: E.K v
25:2: 6:?» 0.

>3.» »E. <w: .595... 2:896 .o .663 8:0 :0» .2383: .53 m
528.3% 23:2. 0:

>3.» »E. <m... 53.5.2 .6» 3:222 :0» 23.3325: 2.22m N
25:2. 6:?» 0.

>3.» »..... <w: amazes. 2:898 .o »>au N ..o» .2333: .555. .

8:223. 595 5.553.. .2650 6%. 2.58 8:06 .o. .o: 28.3
no»: 830

 

.Am.nm._m>m .o: n .<.z. »o:o_o 62.0032 :6... 06:92.... 0. no»:
mcuoo ucm 3352:. 300316525 :0 8:32: 2.. .o. 3:020» mmaEm» .o 5.5.630 ....v 2......

124

RESULTS

Fifty three new integrase genotypes were detected in the screening of
five samples of marine sediments (Figure 4.4). The clusters formed in the
constructed dendrogram had a very diverse composition, most of the marine
integrases were widely distributed across the different branches regardless of
their origin indicating a high degree of diversity within this set of sequences.
Discrete clusters consisting of sequences from a particular source were not
prevalent. Only two clusters (Figure 4.4, “A and IIB) were found to exclusively
consist of sequences recovered from a particular sample and these were from a
depth of 180m at the Barrow Canyon site (Arctic Ocean). In general the marine
integrases presented a higher degree of divergence relative to previously
described integrases retrieved from terrestrial environments; most of the
reference sequences were grouped together (Figure 4.4, cluster I) and
appeared to be related to only two lineages recovered from the Arctic region.
Only a sequence recovered from the Washington coast site emerged as a sister
group of a class 2 integrase (51% identity) described in Shigella sonnei.
Similarly, only a single clone from the Arctic region (ADBC-AOZ) was the closest
match to a class 1 integrase carried by a Citrobacter freundii isolate (60%
identity). However, several sequences (ADBC-A12, ADBC-G1 1, ASBC-CO4,
ASBC-G10, PSCl-FO4, PSCl-FOB, PSCl-H01, WAC-001, and WAC-G02)
representative of four of the five marine sites analyzed were affiliated in a
cluster that contained a super integron integrase described in Vibn'o vulnificus

(Rowe-Magnus et al., 2003) a marine bacterium known to cause

125

gastrointestinal illness and bloodstream infections associated with the
consumption of raw sea food. Clone WAC-602 was the closest relative of the

V. vulnificus SI integrase (53% identity).

126

The analysis of the environmental integrases recovered from Mona Island
(47 miles from the west coast of Puerto Rico) also revealed a high degree of
diversity among the recovered genotypes (Figure 4.5) Twenty eight unique
sequences were detected at this location and only six of them grouped near
reference sequences. Clones MONA-lnt-A3 and MONA-lnt-A3 presented an
amino acid identity of approximately 66% relative to the integrases detected in
the genomes of Thiobacillus denitrificans and Geobacter metallireducens, while
clones MONA-Int—D10 and MONA-lnt-E7 which are 91% identical relative to
each other were related (65% identity) to integrases previously recovered from
environmental DNA using a PCR-based approach (Nemergut et al., 2004). The
sequences corresponding to clones MONA-lnt-A8 and MONA-Int—H6 presented
amino acid sequence identity of 70 and 67% relative to Int|7 and lntl6
respectively. The latter reference sequences were recovered from uncultured
bacteria via PCR (Nield et al., 2001). Only two sequences corresponding to
clones MONA-lnt-A2 and MONA-Int A4 appeared to be related to the the XerD
and XerC reference sequences.

Most of the environmental integrases recovered from pristine Hawaiian

soil were also very different from those previously described (Figure 4.6).
Twenty one integrase variants were distinguished by comparative sequence
analysis. The sequence corresponding to clone Hl-lnt-BQ was the only
genotype closely related to the integrase of a known resistance integron. This
clone was 91% identical to the class 2 integrase present in Shigella sonneii.

The sequences corresponding to clones Hl-lnt-F10, B4, and H3, were 67, 91,

128

and 66% identical to integron-encoded integrases recovered from uncultured
bacteria, while an integrase from the Thiobacillus denitrfficans genome was the

closest match (65% identity) of clone Hl-lnt-H10.

129

59 M ONA-Int-Df

  
 
  
  
 
 
 
  

70 M ONA-Int-Ff
M DNA-In t-BO
M ONA-Int-HJ
M ONA-Im-DQ

M ONA-Int-E1

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

M ONA-Int-BG
F"— M ONA-Int-A3
53 ‘
M ONA-lnt-Ca
34 Int T. denitrificans ZP 00335637
Int G. motallircducens ZP 00299196
Intl uncultured bacterium AA P3 7599
100 l— M ONA-Int-Dfa
l—M ONA-Int-E7
In t/ uncultured bac tormm A A P3 75 9 7
34 M ONA-Int-FZ
M ONA-Int-Fa
99 M ONA-Int-B‘f
M ONA-Int-H2
94 M ONA-lnt-AO
‘2 u ONA-Int-Cfo
69
44 In (I? uncultur 0d bacterium A A K00305
I
3* M ONA-Int-AB
In t/6 uncultured bacterium AA K0030?
2
M ONA-lnt-Ho
98
100 6 M ONA-Int-54
98 M ONA-Int-D3
MONA-lnt-Hf
MONA-lnt-B7
MONA-Int-Gl
100
sil‘i— M ONA-Int-C4
— M ONA-Int-EO
100 J— M ONA-lnt-AZ
I l- " ONA-Int-A4
l 100 XerD B. subtllis P463az

 

 

XerC H.1nfluonzao P44818

0.1
Figure 4.5. Dendrogram depicting the relationships of integron-encoded integrases detected in
soil collected from Mona island and those previously described.

 

 

130

 

 

53 Int/7 uncultured bacterium AAK0030

77 HI-Int-F 10

 

HI-lnt-E1

 

 

HI-ln t-B 1 2

 

HI-Int-D1 1

 

 

49 HI-Int-Gz

Hl-ln t-C7

 

 

HI-Int-AG

 

 

lntl6 uncultured bacterium AAK0030

I'll-In t-B1 1

 

70
Hl-ln t-F5

 

(0
{CI

 

 

th— HI-Int-G5

Hl-lnt-C1

 

11' 56 Hl-lnt-H3

 

 

lntl uncultured bacterium AAP3759

l__

lnt T. denitrificans ZP 00335637

F—l—

HI-Int-F 1 1
92

 

Int Geobacter metallireducens ZP 00299196

 

n ,
34
I'll-In t-H 10

98 Int Geobacter metallireducens ZP 002991 '

75

 

 

 

Int Met/i lobacillus f/EK ellatus ZP 0017

 

HI-Int-HO

 

 

L Int/2 Shic e/la sonnei AA T72891
99' XerD B. subtilis P4635

1 XerC H. influenzae P4481.

 

 

 

 

l J
I

0.1

Figure 4.6. Dendrogram depicting the relationships of integron-encoded integrases
detected in pristine soil collected in Hawaii and those previously described.

131

 

The presence of integron-encoded integrases among bacterial
populations residing in permafrost soil could not be confirmed with the approach
and experimental conditions used herein. No PCR product corresponding to the
expected 500 bp size was detected after several amplification attempts.
However small subunit (16S) rRNA genes were amplifiable from DNA extracted
from permafrost samples indicating that the isolated DNA was suitable for PCR
(Figure 4.7).

Twenty two sequences of environmental integrases recovered from
agricultural (manure-supplemented) and non-agricultural soil samples (Chapter
II) were selected and compared along with those detected in marine and
subtropical environments (Figure 4.8). Contrary to what was observed in Figure
4.4, the reference sequences were more distributed among those of an
environmental origin when genotypes from terrestrial habitats were included in
the clustering analysis. However, sequences from marine sediments still
presented a strong tendency to group among themselves which is obvious in
the nearly symmetrical arrangement of the sequences in Figure 4.8. The left
half of the circular dendrogram consists almost exclusively of integrases from
marine habitats. In contrast, sequences from subtropical environments are
dominant in the right half of the dendrogram. Interestingly, sequences from soil
exposed to animal waste were the ones most similar to integrases of resistance
integrons such as the class 1, class 2, and class 3 integrons. However, most of
the clusters generated had a diverse composition demonstrating similarities

among integrases of different origin.

132

1.3kb —
500bp ~

   

Figure 4.7. Panel A shows the results for the amplification of 168
rRNA genes from permafrost soil samples. M = 1Kb ladder, + =
positive control, - =negative control lacking template DNA, 1 =
DNA from permafrost loam soil recovered from the tundra forest at
a depth of 16.5 m, estimated geological age: 2—3 million years. 2 =
DNA from permafrost collected from the sea coast of the tundra
zone at a depth of 3 m, estimated geological age: 5K years. Panel
B presents the results of the amplification of integrase genes from
permafrost samples. Plasmid DNA containing a cloned class 2
integrase gene was used a positive control.

133

053 AAK26251

 

  

-lnf—D3

A = known resistance integrons
= prevalence of integrases from terrestrial environments

MONA

= prevalence of integrases from marine environments

.0
o

 

N
i
/ \ONA‘ 1
A C'HO9
o c
. X‘HONA‘ r‘0
I / .
$15) lntl6
64 Hi.i,,,‘_’g§"”ured AAK00307

PSCI—D10
\ Hl-lnt'Ab‘

 

665
.00“ AAC}16 /
E . a cABtiZfiaz)2 9’” .

terlC DBC'A J? H/ 4 Inns
A 8 \ 416/0154? 7

91 ‘5“ \\ ’77. ”440:.
\ 40311577 H2

700 § ’08 C1003

   

(H
.5an [NT 19
MS 2355 1 1399 6%
.ae AA%BBC,DQOA XM ,fs C/
l . A .
2 m0“ P‘ 090 ,FOM / \ 4:" Q6009
\ P‘ 0609C," ,\ a“) \ ”( g0. 07
\\ “he“ r gob Q to \ 6:326)? ”8006,
o .
\‘ 952% (3,09% Kg & m ’5 m 0:” ‘9 <>6’od/°AZ&7:”0%
N \ o
\\ p‘poesdyk‘oo‘b . L2,, d7 / (3’2007, 0,, 8,0 \
\ Q 50 \yohkbt L8: ‘96 /¢,:¢A’1g, .9 40,9 003 #
/ t
\ 90 0' 99 09K 9:0 4,7. sit/’0’: 356:;
\ 900?” (w) 6.260.996b0) 4’0
‘\ V0 9% $66 "Q 0“ a) £8) 6%3‘6/491359G/f60 ‘92.?
‘ V t“ id? “é“ 23%;0 ea?” We. ‘9
I ' " Q’ ¢ ., to a:
\\ \>Q7000g’~’z~@§fboik m>'0'°3%0021v‘7>¢0 «owevv
\ ‘bQ‘b 0 HZ‘QQN" kkk >memozzvv,a¢ o «04';
\‘~V‘§°‘g’z§7é’039988322 BEIE$§>Zg€’¢"o% / a) 00) ”33.9
\\ V? ”gmg’Oé “4.x. 9§Whgli a $6)? )
‘ YV‘O‘U‘W: lama: >°s~ .0 .90,
~\\ VV§Q (7)-v3) Dgam - 00 0
\ \ E (é; E 00 r39
\9)
‘90

      

 
 

Figure 4.8. Neighborjoining dendrogram illustrating the relationship of
integron—encoded integrases recovered from a variety of marine and
terrestrial habitats. The codes explaining the source of each sequence

are described in Table 4.1.

 

134

 

DISCUSSION

Mobile genetic elements influence the dispersal of traits that allow
bacterial populations to evolve and adapt to selective environments. It has been
recognized that the integron platform is the system responsible for the presence
of antibiotic resistance determinants frequently detected in transposons and
plasmids carried by multiresistant strains in the clinic environment. The recent
discovery of super integrons and resistance integrons across different bacterial
species as well as the detection of integrons in the environment has
demonstrated that the distribution of this genetic system is not restricted to
Gram-negative bacteria of clinical origins as early culture-based studies
indicated. However, knowledge on the extent of the diversity and distribution of
integrons as well as on their role in nature is still limited.

The molecular screening of 14 environmental samples described herein
uncovered the presence of 126 novel integron-encoded integrases in a variety
of terrestrial and marine habitats. The recovery of such a diverse array of new
environmental integrons supports the hypothesis that these mobile elements
are ubiquitous among prokaryotes and that their distribution is not restricted
only to clinical multiresistant strains or pathogens. All the environmental
integrases presented signature protein motifs specific to integron—encoded
integrases but in general, the predicted amino acids sequences were
considerably different among each other and relative to any previously
described integron sequence, indicating that the environmental pool of these

elements is highly diverse. The comparative analysis of the amino acid

135

sequences of environmental integrases relative to previously known integron
and super integron integrase sequences demonstrated that all of them grouped
together in a clade that excludes the XerC and XerD recombinases which are
the closest relatives of integron integrases from all the members belonging to
the tyrosine recombinase family. This observation supports the relatedness of
the novel sequences with characterized integrons. This clustering trend was
consistently observed when constructing alignments and dendrograms with
sequences retrieved from specific locations as well as in analyses that included
sequences from different geographic sources.

In general integrase clone libraries derived from soil samples exposed to
animal waste did not present a high degree of variability among the recovered
sequences and these were closely related to class 2 integrons which have been
implicated in the transfer and high prevalence of ABR genes among clinical
isolates (Fluit and Schmitz, 2004). In contrast, sequences similar to resistance
integrons were rare or absent among clone libraries derived from non
agricultural soil and marine sediments suggesting a link between animal waste
and the dissemination of class 2 integrons. Additionally samples not associated
with fecal material yielded clone libraries that were more diverse suggesting
that specific integron classes are enriched within the mammal gut environment
presumably because it might be a more stable environment relative to soil and
marine sediments in which a diverse pool of elements capable of facilitating the
adaptation of bacterial populations to different selective forces might be

desirable and favored.

136

The samples from permafrost soil were the only ones from which
integron integrases could not be recovered despite the fact that the extracted
DNA was suitable for PCR as indicated by the successful amplification of 16S
rRNA genes. A possible explanation for these results is that if present,
integrons residing in permafrost populations might have a nucleotide sequence
incompatible with that of the primers used. Alternatively, integrons might not be
as prevalent in permafrost samples as in warmer environments and populations
carrying them may not be dominant members of the total community resulting in
the availability of a suboptimal amount of target sequences for PCR
amplification. A more sensitive method such as real-time PCR or nested PCR
could be more useful under these circumstances.

lntegrons are mainly disseminated by transfer mechanisms such as
conjugation, transformation, and transduction which would require a dynamic
environment facilitating the movement of water, materials and high numbers of
cells, therefore promoting chances for physical contact between vectors,
donors, and recipient strains. Although the existence of viable permafrost
bacteria in unfrozen water films associated with mineral particles is well
documented, conditions of subzero temperatures, lower cell numbers (relative
to warmer environments) and low water activity are known to prevail in Siberian
permafrost (Gilichinsky et al., 1993). Permafrost bacteria are suspected to exist
under a cryobiosis state which would shut down cellular processes necessary
for the active uptake of DNA from an external source. They are also

characterized for the presence of thick cell envelopes (presumably of a

137

 

protective function against freezing, Gilichinsky et al., 1993 ) that might interfere
with the uptake of exogenous DNA. Additionally if methylation sensitive
restriction systems are widely spread among permafrost bacteria, this could
also prevent the incorporation of foreign DNA, hampering the uptake of mobile
elements carrying integrons.

The presence of a diverse pool of integrons in natural environments
certainly enhances the evolutionary potential of bacterial populations through
the exploitation of this system. Although the role of integrons in the lateral

transfer of ABR traits in the clinical setting is well documented, how these

 

genetic elements may contribute to the perpetuation of the antibiotic resistance
problem in environmental locations and how this may be linked to the clinical
setting requires further investigation. The recent discovery of SI and their
massive pool of associated gene cassettes in conjunction with comparative
molecular analyses on the activity of SI and RI integrases have provided new
insight on the role of integrons in the evolution and the expansion of the
adaptive capabilities of prokaryotes.

Besides catalyzing the site specific excision and recombination of gene
cassettes between the integron associated attl site and the cassette associated
attC region, the class 1 integron integrase has been also reported to carry out
recombination reactions between two cassette associated attC sites, and two
attl segments of unrelated integrons (Partridge et al., 2000). Furthermore, the
class 1 integrase is also capable of site specific recombination with secondary

targets (Collis et al., 2001). Although these less conservative recombination

138

events occur at lower frequencies than the typical attl X attC reaction, these
findings are indicative that the integron module has alternative means of
introducing novel traits providing its host with a very versatile adaptive
mechanism.

Recently Rowe-Magnus and coworkers (2002) demonstrated that the
integrase of a class 1 RI is capable of randomly recruiting gene cassettes
residing in SI including a cryptic chloramphenicol resistance gene.
Demonstrating that gene transfer among different integrons is feasible and that
phenotypically sensitive bacteria can be reservoirs of transferable resistance
traits. The cryptic nature of this chloramphenicol resistance gene was found to
be due to its native position within the SI which is too distant from the Pc
promoter that drives the expression of inserted gene cassettes. Placement of
the cryptic gene next to a strong promoter by genetic manipulations involving
deletions of the genes in front of it, resulted in consistent expression of
chloramphenicol resistance.

Comparative studies between class 1 integron and the Vibn'o cholerae SI
integrase activities have demonstrated that the SI integrase recognizes a
restricted range of attC sites compared to the class 1 integrase (Biskri et al.,
2005). This is in agreement with previous observation of a high degree of
similarity among the attC sites associated with gene cassettes residing in SI
versus the great variability present in the attC region of cassettes inserted in

resistance integrons (F luit and Schmitz, 2004). However, these findings are

139

 

additional evidence in support of the notion of integrons as flexible systems that
promote the evolution of bacterial genomes.

Considering the high diversity of integrons from environmental sources
reported herein, the potential number of gene cassettes associated to super
integrons residing in nature, and the versatility of integron integrases in
recognizing diverse integron and gene cassette-associated recombination sites,
all these factors create a scenario of a vast reservoir of genetic functions and
mechanisms available to prokaryotes to adapt to a variety of selective forces
including antibiotic resistance.

lntegron hosts have been reported to withstand treatments meant to
reduce the microbial load in waste water treatment plants and poultry
processing environments. Tennstedt and coworkers (2003) were able to identify
ABR gene cassettes associated with class 1 integrons present in exogenously
and directly recovered plasmids from bacteria residing in the activated sludge
and from the final effluents of a waste water treatment plant. Additionally, class
1 and class 2 integrons were detected in poultry carcasses after treatment in a
chlorinated chiller tank system developed to reduce bioburden and prevent
cross-contamination in poultry processing facilities (Roe et al., 2003).
Eventually this may result in the environmental discharge of integron hosts and
potential ABR reservoirs that might find their way back to humans through
several possible routes. Hence, implications of anthropogenic activities

involving the large scale use on antibiotics and the manipulation of materials

140

 

with high microbial loads with human health risks, should be evaluated from a

broader perspective.

141

 

REFERENCES

Biskri, L., M. Bouvier, A-M. Guérout, S. Boisnard, and D. Mazel. 2005.
Comparative study of class 1 integron and Vibn'o cholerae superintegron
integrase activities. J. Bacteriol. 187:1740-1750.

Collis, CM, and RM. Hall. 1995. Expression of antibiotic resistance genes in
the integrated cassettes of integrons. Antimicrob. Agents Chemother.
39:155-162.

Collis, CM, 6.0. Recchia, M-J. Kim, H. W. Stokes, and RM. Hall. 2001.
Efficiency of recombination reactions catalyzed by class 1 integron
integrase lntl1. J. Bacteriol. 183:2535-2542.

Fluit, AC, and F-J. Schmitz. 2004. Resistance integrons and super-integrons.
Clin. Microbiol. Infect. 10:272-288.

Gilichinsky, D.A., V.S. Soina, and MA. Petrova. 1993. Cryoprotective properties
of water in the earth cryolythosphere and its role in exobiology. Origins of
life and evolution of the biosphere. 23:65-75.

Griintzig, V., 8.0. Nold, J.Z. Zhou, and J.M. Tiedje. 2001. Pseudomonas
stutzen' nitrite reductase gene abundance in environmental samples
measured by real-time PCR. Appl. Environ. Microbiol. 67:760-768.

Hall, R. and, C. Collins. 1995. Mobile gene cassettes and integrons: capture
and spread of genes by site specific recombination. Mol. Microbiol.
15:593-600.

Hart, CA. 1998. Antibiotic resistance: an increasing problem? Br. Med. J.
316:1255—1256.

Konstantinidis, K.T., and J.M. Tiedje. 2005. Genomic insights that advance the
species definition for prokaryotes. PNAS. 102:2567-2572.

Lerverstein-van Hall, M.A., H.E. Blok, A.R. Donders, A. Paauw, A.C. Fluit, and
J. Verhoef. 2003. Multidrug resistance among Enterobacteriaceae is
strongly associated with the presence of integrons and is independent of
species or isolate origin. J. Infect.Dis. 187:251-259.

Mazel, D., B. Dychinco, V.A. Webb, and J. Davies. 1998. A distinctive class of
integron in the Vibn'o chloera genome. Science. 280:605-608.

142

Naas, T., Y. Mikami, T. lmai, L. Poirel, and P. Nordmann. 2001.
Characterization of of ln053, a class 1 plasmid- and composite
transposon-located integron of Escherichia coli which carries an unusual
array of gene cassettes. J. Bacteriol. 183:235-249.

Nemergut, D.R., A.P. Martin, S.K. Schmidt. 2004. lntegron diversity in heavy-
metal-contaminated mine tailings and inferences about integron
evolution. Appl. Environ. Microbiol. 70:1160-1168.

Nield, B.S., A.J. Holmes, M.R. Gillings, G.D. Recchia, B.C. Mabbutt, K.M.
Nevalainen, and H.W. Stokes. 2001. Recovery of new integron classes
from environmental DNA. FEMS Microbiol. Lett. 19:59-65.

Partridge, S.R., G.D. Recchia, C. Scaramuzzi, C.M. Collis, H.W. Stokes, and
RM. Hall. 2000. Definition of the attl1 site of class 1 integrons.
Microbiology. 146:2855—2864.

Paulsen, l. T., L. Banerjei, G. S. A. Myers, K. E. Nelson, R. Seshadri, T. D.
Read, D. E. Fouts, J. A. Eisen, S. R. Gill, J. F. Heidelberg, H. Tettelin, R.
J. Dodson, L. Umayam, L. Brinkac, M. Beanan, S. Daugherty, R. T.
DeBoy, S. Durkin, J. Kolonay, R. Madupu, W. Nelson, J. Vamathevan, B.
Tran, J. Upton, T. Hansen, J. Shetty, H. Khouri, T. Utterback, D. Radune,
K. A. Ketchum, B. A. Dougherty, and C. M. Fraser. 2003. Role of mobile
DNA in the evolution of vancomycin-resistant Enterococcus faecalis.
Science. 299:2071-2074.

Recchia, G., and RM. Hall. 1997. Origins of the mobile gene cassettes found in
integrons. Trends Microbiol. 52389-394.

Roe, M.T., J.A. Byrd ll, D.P. Smith, and SD. Pillai. 2003. Class 1 and class 2
integrons in poultry carcasses from broiler house and poultry processing
environments. J. Food Prot. 66:1426-1431.

Rowe-Magnus, D.A., A.M. Guerout, and D. Mazel. 2002. Bacterial resistance
evolution by recruitment of super-integron gene cassettes. Mol.
Microbiol. 43:1657—69.

Rowe-Magnus, D.A., A-M. Guerout, L. Biskri, P. Bouige, and D. Mazel. 2003.
Comparative analysis of superintegrons: engineering extensive genetic
diversity in the Vibrionaceae. Genome Res. 13:428-442.

Rowe-Magnus, DA, and D. Mazel. 2001. lntegrons: natural tools for bacterial
genome evolution. Curr. Opin. Microbiol. 4:565-569.

Rowe-Magnus, DA, and D. Mazel. 1999. Resistance gene capture. Curr. Opin.
Microbiol. 22483-488.

143

 

Schmeiger, H.,and P. Schicklmaier. 1999. Transduction of multiple drug
resistance of Salmonella enterica serovar typhimun'um DT104. FEMS
Microbiol. Lett. 170:251-256.

Snyder, L., and W. Champness. 1997. Molecular genetics of bacteria. ASM
Press. Washington DC. p.199.

Stokes, H., and R. Hall. 1989. A novel family of potentially mobile elements
encoding site-specific gene—integration functions: integrons. Mol.
Microbiol. 3: 1669-1683.

Tennstedt, T., R. Szczepanowski, S. Braun, A. Ptihler and A. Schliiter. 2003.
Occurrence of integron-associated resistance gene cassettes located on

antibiotic resistance plasmids isolated from a wastewater treatment
plant. FEMS Microbiol. Ecol. 45:229-238.

White, P.A., C.J. Mclver, Y.M. Deng and W.D. Rawlinson. 2000.

Characterisation of two new gene cassettes, aadA5 and dfrA17. FEMS
Microbiol. Lett. 182:265-269.

144

 

CHAPTER V
CONCLUSIONS AND FUTURE DIRECTIONS
Conclusions
The study of the distribution and diversity of reservoirs of antibiotic

resistance determinants and the genetic elements involved in their capture and
dissemination can advance our understanding of the antibiotic resistance
problem and the development of strategies to contain and track current and
future threats to human health. The goal of this work was the application of
culture-independent PCR-based technologies to gain a comprehensive
assessment of the diversity of ABR genes in a variety of terrestrial and marine
habitats, with particular emphasis on the detection and characterization of
inconspicuous reservoirs and novel resistance determinants. The main findings
of this study are summarized below:

1. The frequency of tetracycline resistance determinants encoding ribosomal
protection is higher in cropland recently supplemented with animal waste
from production facilities using tetracycline as a growth promoter relative to
soil not exposed to animal waste.

2. The pool of RPP genes detected was diverse and the application of
molecular techniques resulted in the discovery of novel variants of RRP
determinants.

3. A quantitative real-time PCR assay was successfully developed and applied
for the environmental detection of the novel tet (36) determinant and to infer

functionality of novel variants of putative RPP genes. This assay was also

145

successfully applied to tracing the origin of novel RPP genotypes to swine
being feed with tetracycline.

4. The integron-targeted PCR assay revealed the presence of class 1 integron-
associated genes encoding resistance against aminoglycoside antibiotics
and quaternary ammonium drugs in manured soils indicating that animal
waste is also a reservoir of additional resistance determinants not implicated
in providing protection against an agent exerting a selective constraint
(tetracycline).

5. lntegron-encoded resistance genes were found to persist under field
conditions after one month of manure application while RPP determinants
became less prevalent indicating that these genes were not linked into the
same genetic element or reservoir.

6. The integron platform was prevalent in manure-supplemented cropland and
also in non-agricultural reference soils. In general, integrons detected in
manure fields were closely related to class 1 and class 2 resistance
integrons while integrons detected in reference soils were not.

7. Many of the RPP genes and integron-encoded genes were highly related and
in some instances identical to those reported in clinical or disease causing
strains suggesting lateral transfer of these traits.

8. The molecular analysis of the distribution and diversity of integrons in DNA
extracted from marine sediments, permafrost, and tropical soils revealed
that these elements are prevalent and highly diverse in marine and

subtropical habitats but undetectable in permafrost. Contrary to early

146

 

culture-based studies these findings demonstrate that integrons are not
confined to the clinical setting and are widely distributed in a variety of
natural environments. Hence, they should play a relevant role in bacterial
evolution and adaptation.

Future Directions

This study demonstrated that commensal bacteria are reservoirs of
antibiotic resistance genes and that novel antibiotic resistance traits could be
retrieved from total community DNA extracted from environmental sources
avoiding culture bias. It was also demonstrated that integrons are prevalent and
widely distributed in natural habitats but the identity of the specific species or
bacterial populations harboring the novel resistance traits and detected
integrons remains unknown. However, additional DNA-based techniques can
be used to determine the genetic diversity and composition of a microbial
community without the selection bias linked to isolation techniques. The
fractionation of total community bacterial DNA based on its guanine plus
cytosine (G+C) content has been successfully applied to the analysis of
complex prokaryotic communities and to improve the detection of populations of
limited abundance (Holben et al., 2004).

This method involves mixing environmental DNA with a dye (bis-
benzimidazole) which binds to adenine and thymidine (A+T) residues and
changes the buoyant density of the stained DNA in proportion to its A+T load.
This allows separation of DNA from different populations by equilibrium density-

gradient (CsCl) centrifugation into aliquots representing bacterial groups having

147

different G+C contents. This both uncovers DNA in populations swamped by
that from dominant populations and relates the recovered fractions to taxonomy
since G+C content corresponds to taxonomy (Vandamme et al., 1996). For
instance, the enterococci, a classical indicator frequently monitored to evaluate
antibiotic resistance in animal agriculture (Petersen and Dalsgaard, 2003;
Ozawa et al., 2002) have a very narrow G+C content (47-48%, Apajalahti et al.,
1998); therefore by coupling a PCR based screening targeting 16$ rRNA genes
and novel determinants with a comparative sequence analysis, this approach
can facilitate the identification of reservoir populations. Similarly, the distribution
of genetic elements responsible for lateral transfer of ABR determinants (e.g.
integrons) can also be evaluated. Additionally, since potential reservoir
populations are identified through the G+C fractionation technique, the
application of this approach provides the opportunity to make informed
decisions regarding the best approaches, media, and optimal culture conditions
for the recovery of the identified reservoirs for further studies. For instance,
additional research on the lateral transfer of ABR traits from commensal
isolates into pathogens can also provide useful information for understanding
how resistance determinants find their way into the clinical environment.

The findings reported herein also demonstrated that quantitative real-
time PCR assays are a highly sensitive and specific method for tracing and
quantifying target genes in environmental DNA, especially in difficult templates
such as those extracted from manure-supplemented soil. This technique can be

a valuable resource for monitoring the environmental prevalence of emerging

148

threats like the novel Panton-Valentine virulence factor which is known to cause
severe necrotizing pneumonia and has been recently described in community
acquired infections caused by methicillin-resistant Staphylococcus aureus
(Vandenesch, et al., 2003). Quantitative PCR, could also be applied for tracing
genes conferring resistance against last resort antibiotics (vancomycin), and
newly developed drugs in environmental settings impacted by the
anthropogenic use of antibiotics.

The further development of molecular technologies for the environmental
detection of novel resistance determinants and mobile genetic elements holds
promise for the generation of useful data for elucidating how these traits are
dispersed among bacteria residing in different environments. Nucleotide
sequence data derived from novel resistance genes and associated replicons
can facilitate the development of oligonucleotides and probes for the
characterization of novel plasmids and unknown reservoirs (Smalla and
Sobecky, 2002).

The implementation of metagenomic techniques is another resource
available for analyzing the pool of antibiotic resistance genes in environmental
sources. The direct extraction and cloning of environmental DNA fragments
from microbial communities is also independent of culture techniques and
allows the recovery of intact genes, circumventing PCR and the need of
previously existing sequence data for primer design. Moreover, the incubation
of environmental samples with antibiotics previous to total DNA extraction may

facilitate the recovery of resistance mechanisms targeting a specific drug.

149

 

Hence, antibiotic resistance genes, for which no previous knowledge is
available, can be recovered through this technique also allowing to test for the
functionally of the recovered determinants. Construction of these libraries is not
limited to the use of bacterial artificial chromosomes (Riesenfeld et al., 2004) or
cosmids. Commercially available plasmid vectors have been successfully used
for the cloning and expression in Escherichia coli of fragments of DNA extracted
from oral samples, leading to the discovery of a novel tetracycline resistance
gene which inactivates this antibiotic (Diaz-Torres et al., 2003).

A more comprehensive characterization of resistance mechanisms can
lead to a new approach in overcoming antibiotic resistance. In addition to
developing novel drugs insensitive to existing resistance mechanisms, and
identifying novel drug targets, targeting the resistance mechanisms themselves
has emerged as an attractive alternative to prolong the useful life of agents
currently administrated to threat bacterial infections and prevent the eventual

loss of those newly developed (Poole, 2001).

150

REFERENCES

Apajalahti, J.H.A., L.K. Sérkilahti, B.R.E. Maki, J.P. Heikkinen, P.H. Nurminen,
and WE. Holben. 1998. Effective recovery of bacterial DNA and percent-
guanine-plus-cytosine—based analysis of community structure in the
gastrointestinal tract of broiler chickens. Appl. Envir. Microbiol. 6424084-
4088.

Diaz-Torres, M.L., R. McNab, D.A. Spratt, A .Villedieu, N. Hunt, M. Wilson, and
P. Mullany. 2003. Novel tetracycline resistance determinant from the oral
metagenome. Antimicrob. Agents Chemother. 47:1430-1432.

Holben. W.E., K.P. Feris, A. Kettunen, and J.H.A. Apajalahti. 2004. GC
fractionation enhances microbial community diversity assessment and
detection of minority populations of bacteria by denaturing gradient gel
electrophoresis. Appl. Envir. Microbiol. 70:2263-2270.

Ozawa, Y., K. Tanimoto, T. Nomura, M. Yoshinaga, Y. Arakawa, and Y. lke.
2002. Vancomycin-resistant enterococci in humans and imported
chickens in Japan. Appl. Envir. Microbiol. 68:6457-6461.

Petersen, A., and A. Dalsgaard 2003. Species composition and antimicrobial
resistance genes of Enterococcus spp, isolated from integrated and
traditional fish farms in Thailand. Environ. Microbiol. 5:395-402.

Poole, K., 2001. Overcoming antimicrobial resistance by targeting resistance
mechanisms. J. Pharm. Pharrnacol. 53:283-294.

Riesenfield, C.S., R.M. Goodman, and J. Handelsman. 2004. Uncultured soil
bacteria are a reservoir of new antibiotic resistance genes. Environ.
Microbiol. 6:981 -989.

Smalla, K., and PA. Sobecky. 2002. The prevalence and diversity of mobile
genetic elements in bacterial communities of different environmental
habitats: insights gained from different methodological approaches.
FEMS Microbiol. Ecol. 42:165-175.

Vandenesch, F., T. Naimi, M.C. Enright, G. Lina, G.R. Nimmo, H. Heffernan, N.
Liassine, M. Bes, T. Greenland, M-E. Reverdy, and J. Etienne. 2003.
Community-acquired methicillin-resistant Staphylococcus aureus
carrying Panton-Valentine Ieukocidin genes: worldwide emergence.
Emerg. Infect. Dis. 92978-984.

151

Frail: . .lilili-»..a. ill. “I.I I

l Lil‘s .

 

IIIIIIIIIIIIIIIIIIIIIIIIIIII

lllllillllillllHlilllllllll‘illlllllillll'llllillllllllll