Molecular Methods for Bacterial Strain Typing

MM11-A Vol. 27 No. 10 Replaces MM11-P Vol. 26 No. 14

Molecular Methods for Bacterial Strain Typing; Approved Guideline

This guideline examines the biology behind molecular strain typing and the process of characterizing and validating typing systems. The prevalent methods are described with particular attention to pulsed-field gel electrophoresis (PFGE) and multilocus sequence typing (MLST). A guideline for global application developed through the Clinical and Laboratory Standards Institute consensus process.

Clinical and Laboratory Standards Institute Advancing Quality in Healthcare Testing Clinical and Laboratory Standards Institute (CLSI, formerly NCCLS) is an international, interdisciplinary, nonprofit, standards-developing, and educational organization that promotes the development and use of voluntary consensus standards and guidelines within the healthcare community. It is recognized worldwide for the application of its unique consensus process in the development of standards and guidelines for patient testing and related healthcare issues. Our process is based on the principle that consensus is an effective and cost-effective way to improve patient testing and healthcare services. In addition to developing and promoting the use of voluntary consensus standards and guidelines, we provide an open and unbiased forum to address critical issues affecting the quality of patient testing and health care. PUBLICATIONS A document is published as a standard, guideline, or committee report. Standard A document developed through the consensus process that clearly identifies specific, essential requirements for materials, methods, or practices for use in an unmodified form. A standard may, in addition, contain discretionary elements, which are clearly identified. Guideline A document developed through the consensus process describing criteria for a general operating practice, procedure, or material for voluntary use. A guideline may be used as written or modified by the user to fit specific needs.

Most documents are subject to two levels of consensus— “proposed” and “approved.” Depending on the need for field evaluation or data collection, documents may also be made available for review at an intermediate consensus level. Proposed A consensus document undergoes the first stage of review by the healthcare community as a proposed standard or guideline. The document should receive a wide and thorough technical review, including an overall review of its scope, approach, and utility, and a line-by-line review of its technical and editorial content. Approved An approved standard or guideline has achieved consensus within the healthcare community. It should be reviewed to assess the utility of the final document, to ensure attainment of consensus (i.e., that comments on earlier versions have been satisfactorily addressed), and to identify the need for additional consensus documents. Our standards and guidelines represent a consensus opinion on good practices and reflect the substantial agreement by materially affected, competent, and interested parties obtained by following CLSI’s established consensus procedures. Provisions in CLSI standards and guidelines may be more or less stringent than applicable regulations. Consequently, conformance to this voluntary consensus document does not relieve the user of responsibility for compliance with applicable regulations. COMMENTS

The CLSI voluntary consensus process is a protocol establishing formal criteria for:

The comments of users are essential to the consensus process. Anyone may submit a comment, and all comments are addressed, according to the consensus process, by the committee that wrote the document. All comments, including those that result in a change to the document when published at the next consensus level and those that do not result in a change, are responded to by the committee in an appendix to the document. Readers are strongly encouraged to comment in any form and at any time on any document. Address comments to Clinical and Laboratory Standards Institute, 940 West Valley Road, Suite 1400, Wayne, PA 19087, USA.

•

the authorization of a project

VOLUNTEER PARTICIPATION

•

the development and open review of documents

•

the revision of documents in response to comments by users

•

the acceptance of a document as a consensus standard or guideline.

Report A document that has not been subjected to consensus review and is released by the Board of Directors. CONSENSUS PROCESS

Healthcare professionals in all specialties are urged to volunteer for participation in CLSI projects. Please contact us at [email protected] or +610.688.0100 for additional information on committee participation.

Volume 27 Number 10

MM11-A ISBN 1-56238-634-4 ISSN 0273-3099

Molecular Methods for Bacterial Strain Typing; Approved Guideline Robert D. Arbeit, MD Judy C. Arbique, ART(CSMLS), CLS(NCA) Bernard Beall, PhD Ian A. Critchley, PhD Frederic J. Marsik, PhD Sophie Michaud, MD, MPH, CSPQ, FRCP(C) Christine Steward, MPH, MT(ASCP) Fred C. Tenover, PhD, ABMM David L. Trees, PhD

Abstract Molecular strain typing has become an essential tool for the analysis of bacterial pathogens obtained during investigations of epidemiologic outbreaks, laboratory contamination, and recurrent infection. A wide variety of strain typing methods have been described using contemporary DNA-based technologies. However, developing methods and generating data have proven easier than defining robust approaches for interpreting the results. Clinical and Laboratory Standards Institute document MM11-A—Molecular Methods for Bacterial Strain Typing; Approved Guideline examines the biology behind molecular strain typing and the process of characterizing and validating typing systems. The prevalent methods are described with particular attention to pulsed field gel electrophoresis (PFGE) and multilocus sequence typing (MLST). Specific issues in analyzing typing data derived from these methods are discussed. The guideline offers a general approach, suitable for use in the situations commonly encountered in clinical laboratories, for interpreting and reporting molecular typing results. For selected bacterial pathogens, the application of molecular typing systems and the insights derived are considered in detail. Clinical and Laboratory Standards Institute (CLSI). Molecular Methods for Bacterial Strain Typing; Approved Guideline. CLSI document MM11-A (ISBN 1-56238-634-4). Clinical and Laboratory Standards Institute, 940 West Valley Road, Suite 1400, Wayne, Pennsylvania 19087-1898 USA, 2007. The Clinical and Laboratory Standards Institute consensus process, which is the mechanism for moving a document through two or more levels of review by the healthcare community, is an ongoing process. Users should expect revised editions of any given document. Because rapid changes in technology may affect the procedures, methods, and protocols in a standard or guideline, users should replace outdated editions with the current editions of CLSI/NCCLS documents. Current editions are listed in the CLSI catalog, which is distributed to member organizations, and to nonmembers on request. If your organization is not a member and would like to become one, and to request a copy of the catalog, contact us at: Telephone: 610.688.0100; Fax: 610.688.0700; E-Mail: [email protected]; Website: www.clsi.org

(Formerly NCCLS)

Number 10

MM11-A

Copyright ©2007 Clinical and Laboratory Standards Institute. Except as stated below, neither this publication nor any portion thereof may be adapted, copied or otherwise reproduced, by any means (electronic, mechanical, photocopying, recording, or otherwise) without prior written permission from Clinical and Laboratory Standards Institute (“CLSI”). CLSI hereby grants permission to each individual member or purchaser to make a single reproduction of this publication for use in its laboratory procedure manual at a single site. To request permission to use this publication in any other manner, contact the Executive Vice President, Clinical and Laboratory Standards Institute, 940 West Valley Road, Suite 1400, Wayne, Pennsylvania 19087-1898, USA.

Suggested Citation (Clinical and Laboratory Standards Institute. Molecular Methods for Bacterial Strain Typing; Approved Guideline. CLSI document MM11-A [ISBN 1-56238-634-4]. Clinical and Laboratory Standards Institute, 940 West Valley Road, Suite 1400, Wayne, Pennsylvania 19087-1898 USA, 2007.)

Proposed Guideline May 2006

Approved Guideline April 2007

ISBN 1-56238-634-4 ISSN 0273-3099

ii

Volume 27

MM11-A

Committee Membership Area Committee on Molecular Methods Roberta M. Madej, MS, MT Chairholder Roche Molecular Systems, Inc. Pleasanton, California

Uwe Scherf, PhD FDA Center for Devices and Radiological Health Rockville, Maryland

Frederick S. Nolte, PhD Vice-Chairholder Emory University Hospital Atlanta, Georgia

Michael A. Zoccoli, PhD Celera Diagnostics Alameda, California Advisors

Zhimin Cao, MD, PhD New York State Dept. of Health Albany, New York Maurizio Ferrari, MD International Federation of Clinical Chemistry Milan, Italy Carolyn Sue Richards, PhD, FACMG Oregon Health Sciences University Portland, Oregon

Leslie Hall, MMSc Mayo Clinic Rochester, Minnesota Timothy J. O’Leary, MD, PhD Biomedical Laboratory Research and Development Service Department of Veterans Affairs Washington, District of Columbia

Janet A. Warrington, PhD Affymetrix, Inc. Santa Clara, California Judith C. Wilber, PhD XDX, Inc. San Francisco, California Laurina O. Williams, PhD, MPH Centers for Disease Control and Prevention Atlanta, Georgia Janet L. Wood, MT(ASCP) BD Diagnostic Systems Sparks, Maryland

Mario Pazzagli, PhD University of Florence Florence, Italy

Subcommittee on Bacterial Strain Typing Robert D. Arbeit, MD Chairholder Paratek Pharmaceuticals, Inc. Boston, Massachusetts

Fred C. Tenover, PhD Centers for Disease Control and Prevention Atlanta, Georgia

David L. Trees, PhD Centers for Disease Control and Prevention Atlanta, Georgia

Bernard Beall, PhD Centers for Disease Control and Prevention Atlanta, Georgia

Advisors

Staff

Judy C. Arbique, ART (CSMLS), CLS Arbique-Rendell Onsite Training & Consulting Halifax, Nova Scotia, Canada

Clinical and Laboratory Standards Institute Wayne, Pennsylvania

Ian A. Critchley, PhD Replidyne, Inc. Louisville, Colorado Frederic J. Marsik, PhD Food and Drug Administration Rockville, Maryland Sophie Michaud, MD, MPH Universite de Sherbrooke Sherbrooke, Quebec, Canada

Sangeeta M. Rataul, PhD US Food and Drug Administration Bothell, Washington Lucy Scarborough, MPH, MT (ASCP) South Carolina Department of Health and Environmental Control Columbia, South Carolina

John J. Zlockie, MBA Vice President, Standards Lois M. Schmidt, DA Staff Liaison Donna M. Wilhelm Editor Melissa A. Lewis Assistant Editor

Christine Steward, MPH, MT(ASCP) Andover, Kansas

iii

Number 10

iv

MM11-A

Volume 27

MM11-A

Contents Abstract ....................................................................................................................................................i Committee Membership........................................................................................................................ iii Foreword .............................................................................................................................................. vii 1

Scope..........................................................................................................................................1

2

Standard Precautions..................................................................................................................1

3

Terminology...............................................................................................................................1 3.1 3.2 3.3

4

The Biology Behind Molecular Strain Typing ..........................................................................5 4.1 4.2 4.3 4.4

5

Visual Analysis of Electrophoresis Gels.....................................................................26 Analyzing Electrophoresis Gels by Software .............................................................26 Population Genetics and the Analysis of PFGE Patterns: a Cautionary Note ............32

Analyzing Sequence Data ........................................................................................................33 8.1 8.2 8.3 8.4

9

Controls.......................................................................................................................11 Pulsed-Field Gel Electrophoresis (PFGE) ..................................................................11 Ribotyping ..................................................................................................................20 Sequence-Based Strain Typing ...................................................................................21 Repetitive Sequence-Based PCR (rep-PCR) ..............................................................24

Analyzing Electrophoretic Typing Data ..................................................................................26 7.1 7.2 7.3

8

Reproducibility .............................................................................................................8 Discriminatory Power ...................................................................................................9 Requirements for Characterizing a Molecular Typing System...................................10 Assessment of Competency for Molecular Strain Typing ..........................................11

Methods for Molecular Strain Typing .....................................................................................11 6.1 6.2 6.3 6.4 6.5

7

Sources of Genetic Variation ........................................................................................5 Population Structure of Bacterial Species.....................................................................5 Impact of Selective Pressure on Diversity ....................................................................6 Applications of Molecular Strain Typing .....................................................................6

Validation of Typing Methodologies.........................................................................................8 5.1 5.2 5.3 5.4

6

Definitions ....................................................................................................................2 Comments on Microbiological and Epidemiologic Definitions ...................................3 Abbreviations and Acronyms .......................................................................................4

Percent Identity ...........................................................................................................33 Pattern Recognition.....................................................................................................34 BURST .......................................................................................................................34 Sequence Analysis Methods for Evolutionary Genetics.............................................34

Interpreting Variation in Molecular Typing.............................................................................35 9.1 9.2 9.3 9.4 9.5

Categories of Genotypic Relatedness in Molecular Strain Typing.............................36 Step One: Identify the “Reference Isolate” or Type That Focuses the Question........37 Step Two: Compare Each Isolate to the Reference Isolate .........................................37 Translating Genotypic Relatedness Into Epidemiologic and Clinical Relatedness ....37 Comparison to the “Tenover Criteria” ........................................................................38 v

Number 10

MM11-A

Contents (Continued) 10

Reporting Molecular Typing Results.......................................................................................38

11

General Technical Issues .........................................................................................................39 11.1 11.2

12

Identifying Isolates .....................................................................................................39 Archiving Isolates—Freezing .....................................................................................40

Examples of Molecular Typing of Bacterial Species...............................................................40 12.1 12.2

Streptococcus pyogenes ..............................................................................................40 Streptococcus pneumoniae..........................................................................................45

References.............................................................................................................................................48 Summary of Delegate Comments and Committee Responses ..............................................................59 The Quality Management System Approach ........................................................................................68 Related CLSI/NCCLS Publications ......................................................................................................69

vi

Volume 27

MM11-A

Foreword Colloquially phrased, the central question in bacterial strain typing is deceptively simple: Are two isolates the “same” or “different”? A more rigorous construction begins to identify the complexity: Using a well-characterized molecular technique, are the genotypes of two isolates sufficiently similar to conclude that they represent the same strain, or sufficiently different to conclude they represent different strains? The goal of this guideline is to provide a basis for answering this question. Among the many molecular strain typing methods that have been described, relatively few have been rigorously analyzed to define their performance characteristics. The technical and biologic reproducibility of every typing system needs to be quantitatively defined to inform the user of the reliability of clear results and the implications of the inevitable ambiguous results. Because clinically important species may differ substantially in their population structure, methods may need to be explicitly characterized for different species. The concepts underlying molecular strain typing are derived from both evolution and epidemiology. The isolates comprising a bacterial species generally display substantial genetic diversity as a result of evolutionary divergence. Isolates that are epidemiologically related (e.g., obtained from an outbreak or during the course of infection in a single patient) are presumed to be directly and recently descended from a common ancestor and thus represent a discrete lineage or genotype. Thus, molecular strain typing can distinguish among unrelated isolates because there is evolutionary diversity within a species, and can help identify epidemiologically related isolates because they are expected to be genetically closely related. From this perspective, there is a clear tension between the biological imperative toward divergence and variation, and the analytic need for stability and consistency. This tension defines the limits of any particular molecular strain typing procedure for resolving a given epidemiologic question. Understanding those limits is critical both to choosing a technique and interpreting the results. Key Words Molecular epidemiology, molecular strain typing, nucleotide sequencing, population genetics, pulsedfield gel electrophoresis

vii

Number 10

viii

MM11-A

Volume 27

MM11-A

Molecular Methods for Bacterial Strain Typing; Approved Guideline 1

Scope

Bacterial strain typing is now performed in a wide range of venues, including hospital-based clinical microbiology laboratories; federal, state, and local reference laboratories; as well as industrial and commercial laboratories. Similarly, the results of bacterial strain typing are now used in many different contexts, including clinical care settings; public health investigations, particularly of emerging infections; the food and pharmaceutical industries; and environmental analyses. The goal of this guideline is to provide a framework that will facilitate consistency in reporting bacterial strain typing and will assist both the laboratories performing these studies and the professionals applying the results. A general approach to the analysis of molecular typing data will be presented, as well as specific criteria for interpreting typing results obtained with the most commonly used methods. This guideline will focus on techniques that analyze bacterial chromosomal DNA, particularly pulsed-field gel electrophoresis (PFGE) and nucleotide sequencing; phenotypic techniques (e.g., serotyping, phage typing) and plasmid-based methods will not be addressed.

2

Standard Precautions

Because it is often impossible to know what isolates or specimens might be infectious, all patient and laboratory specimens are treated as infectious and handled according to “standard precautions.” Standard precautions are guidelines that combine the major features of “universal precautions and body substance isolation” practices. Standard precautions cover the transmission of all infectious agents and thus are more comprehensive than universal precautions, which are intended to apply only to transmission of blood-borne pathogens. Standard and universal precaution guidelines are available from the U.S. Centers for Disease Control and Prevention (Garner JS, Hospital Infection Control Practices Advisory Committee. Guideline for isolation precautions in hospitals. Infect Control Hosp Epidemiol. 1996;17(1):53-80). For specific precautions for preventing the laboratory transmission of all infectious agents from laboratory instruments and materials and for recommendations for the management of exposure to all infectious disease, refer to the most current edition of CLSI document M29—Protection of Laboratory Workers From Occupationally Acquired Infections.

3

Terminology

A Note on Terminology CLSI, as a global leader in standardization, is firmly committed to achieving global harmonization wherever possible. Harmonization is a process of recognizing, understanding, and explaining differences while taking steps to achieve worldwide uniformity. CLSI recognizes that medical conventions in the global metrological community have evolved differently in the United States, Europe, and elsewhere; that these differences are reflected in CLSI, ISO, and CEN documents; and that legally required use of terms, regional usage, and different consensus timelines are all challenges to harmonization. Despite these challenges, CLSI recognizes that harmonization of terms facilitates the global application of standards and is an area that needs immediate attention. Implementation of this policy must be an evolutionary and educational process that begins with new projects and revisions of existing documents. The following section provides formal definitions plus explanatory notes for key terms used in this document. During the next scheduled revision, these definitions will be reviewed again for consistency with international use, and revised as needed. ©

Clinical and Laboratory Standards Institute. All rights reserved.

1

Number 10

3.1

MM11-A

Definitions

accuracy (of measurement) – closeness of the agreement between the result of a measurement and a true value of the measurand (VIM93).1 analyte – component represented in the name of a measurable quantity (ISO 17511)2; NOTE: This includes any element, ion, compound, substance, factor, infectious agent, cell, organelle, activity (enzymatic, hormonal, or immunological), or property, the presence or absence, concentration, activity, intensity, or other characteristics of which are to be determined. different – in bacterial strain typing, the results for two isolates are described as “different” based on predefined criteria; this characterization implies that the isolates are not derived from a common (recent) ancestor; NOTE 1: The process for establishing such criteria is discussed in Section 9; NOTE 2: See also indistinguishable, similar. discriminatory power – in bacterial strain typing, the probability that two random, epidemiologically unrelated isolates will be distinguished by the typing method (i.e., identified as different strain types); NOTE: Ideally, each unrelated isolate is detected as unique, but, in practice, some are indistinguishable. epidemiologically related isolates – isolates cultured from specimens (e.g., patients, fomites, the environment) at a discrete time and place as part of an epidemiologic investigation of an outbreak and that are presumed to be related based on the epidemiologic data collected during the investigation.3 genotype – in bacterial strain typing, the results obtained by applying a DNA-based typing system to an isolate; NOTE: Isolates assigned the same genotype represent a strain.3 indistinguishable – in bacterial strain typing, the results for two isolates are described as “indistinguishable” when both isolates give the same results; by definition, such isolates represent the same strain within that typing system; NOTE 1: The term “identical” may be used for precise typing systems, such as nucleotide sequencing, but is not appropriate for analog systems, such as those based on restriction fragment length polymorphisms, which have inherent limits of resolution; NOTE 2: See also different, similar. isolate – a pure culture of bacteria obtained by subculture of a single colony taken from a primary culture plate and presumed to be derived from a single organism; NOTE: “Strain” should not be used as a synonym for “isolate.”3 lineage – in bacterial strain typing, a set of isolates that are derived from a common ancestor and are thereby related; NOTE: Isolates that are distantly derived might show some genetic divergence; isolates that are recently derived would be expected to be closely genetically related and thus to have a common genotype by a variety of typing methods (i.e., to represent a strain).3 measurand – particular quantity subject to measurement (VIM93)1; NOTE: The measurand describes what is causing the result of the measurement, and the analyte describes the particular component of interest to the patient. outbreak – in bacterial strain typing, the increased incidence of an infectious disease in a specific place and time that is above the baseline rate for that place and time.3 outbreak strain – in bacterial strain typing, a set of isolates of the same species that are both epidemiologically related and genotypically related (i.e., represent a single strain).3

2

©


Volume 27

MM11-A

reproducibility – in bacterial strain typing, the closeness of agreement between results obtained when the same method is applied (a) to multiple isolates representing a single strain (as defined by strict epidemiologic criteria) or (b) for replicates derived from subculture of a single isolate; the multiple assessments may be made over time in a single laboratory or in different laboratories; NOTE 1: This is similar to the VIM definition, i.e., the closeness of the agreement between the results of measurements of the same measurand carried out under changed conditions of measurement (VIM93)1; NOTE: 2: Reproducibility in bacterial strain typing is influenced by both technical and biologic factors. similar – in bacterial strain typing, the results for two isolates are described as “similar” if they fail to meet the criteria for either indistinguishable or different; NOTE 1: Isolates with similar genotypes are presumed to be genetically related at some level; NOTE 2: The relative size of this “gray zone” between the two defined categories varies for different typing systems and, potentially, for the same typing system applied to different species (see Section 9); NOTE 3: See also different, indistinguishable above. strain – a set of isolates that, as analyzed by a typing system, give the same typing result (i.e., are “indistinguishable” from each other) and that can be distinguished (i.e., are “different”) from other isolates of the same genus and species; such isolates are thereby inferred to be genotypically related to each other; NOTE 1: A single isolate with a distinctive typing result also represents a strain; NOTE 2: A strain is a descriptive subdivision of a species and is operationally dependent on the characteristics of the typing system; thus, a set of isolates representing a single strain in one typing system might be resolved into more than one strain by another system3; NOTE 3: See also isolate above. typeability – in bacterial strain typing, the ability of a typing system to give a definite result for a given isolate (i.e., to assign a type to the isolate). validation – in bacterial strain typing, the process by which the reproducibility and discriminatory power of new typing methods are shown to be appropriate for resolving epidemiologic relatedness among isolates; NOTE 1: This is similar to the ISO 9000 definition of validation, i.e., confirmation through the provision of objective evidence, that requirements for a specific intended use or application have been fulfilled (ISO 9000)4; NOTE 2: In the context of bacterial strain typing, there is currently no consensus regarding the levels of discriminatory power and reproducibility required for this purpose or the process for demonstrating a particular level of performance; in Section 5.3 of this guideline, specific criteria are proposed.

3.2

Comments on Microbiological and Epidemiologic Definitions

A common vocabulary is essential to any discipline, but has proved remarkably elusive in strain typing. Refer to Section 3.1 for definitions of terms used in this document; of particular importance are isolate, strain, genotype, lineage, indistinguishable, similar, and different. The most problematic word in the current lexicon is “clone.” Although often used to describe a set of isolates with the same genotype, clone is formally defined as an organism that is genetically identical to its parent. This can be readily confirmed for limited DNA sequences, but not for whole bacterial chromosomes. Molecular strain typing techniques sample only a small fraction of the genome; it is inappropriate to designate two isolates as clones because they have a common genotype by a particular technique. Potentially even more confusing is using “clone” as a synonym for lineage, that is, to describe a set of isolates that have obvious genetic differences, but share distinctive genetic features, typically virulence factors, that indicate they are closely related. The use of “clonal” as an adjective to describe a particular population structure is discussed below. Two metaphors are commonly used to describe the characteristics of different typing systems: the “clock speed” of evolutionary markers and the “tree” as a topographic display of phylogenetics. Carl Woese’s study of rRNA sequences to define evolutionary relationships among all living organisms can be viewed ©


3

Number 10

MM11-A

as molecular typing at the species level.5 Ribosomal sequences are useful for identifying and ordering species precisely because they vary exceedingly slowly; epidemiologically relevant strain typing requires features that show variation over far shorter time scales, such as changes in number and sequence of repetitive elements or point mutations in noncoding regions, leading to restriction fragment polymorphisms. The variation in ribosomal sequences defines the divergence of the three kingdoms, typically depicted as the very deepest trunks in the evolutionary tree. Molecular strain typing for epidemiologic purposes seeks to define the smallest, most peripheral branches, representing recent divergence within a species. No single method can be expected to be well suited to all tasks.

3.3

Abbreviations and Acronyms

AIDS AP-PCR BLAST BURST CDC cDNA CHEF DNA eBURST EDTA ERIC GAS HEPES IS MLEE MLST MLVA MRSA NCBI ORF PCR PFGE PMEN QC REP rep-PCR RFLPs RIDOM rRNA SC SLVs SNP ST TAE TBE TE TIFF UPGMA UV VNTR

4

acquired immunodeficiency syndrome arbitrary-primed PCR Basic Local Alignment Search Tool Based Upon Related Sequence Types Centers for Disease Control and Prevention (organization in the United States of America) complementary deoxyribonucleic acid contour-clamped homogeneous electric field deoxyribonucleic acid enhanced version of BURST ethylenediaminetetraacetic acid enterobacterial repetitive intergenic consensus Group A Streptococci 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid insertion sequence multilocus enzyme electrophoresis Multilocus sequence typing multiple loci variable number tandem repeat assays methicillin-resistant Staphylococcus aureus National Center for Biotechnology Information (organization in the United States of America) open reading frame polymerase chain reaction pulsed-field gel electrophoresis Pneumococcal Molecular Epidemiology Network quality control repetitive extragenic palindromic repetitive sequence-based PCR restriction fragment length polymorphisms Ribosomal Differentiation of Medical Microorganisms ribosomal ribonucleic acid similarity coefficient single-locus variants single nucleotide polymorphism sequence type Tris-acetate-EDTA Tris-borate-EDTA Tris-EDTA tagged image file format unweighted pair group method using arithmetic averages ultraviolet variable-number tandem repeat

©


Volume 27

4 4.1

MM11-A

The Biology Behind Molecular Strain Typing Sources of Genetic Variation

The main sources of genetic variation are mutation and recombination, including insertions and deletions. While molecular strain typing is not primarily concerned with the sources of diversity, it is useful to appreciate that different typing systems may detect substantially different events and, therefore, may yield discrepant results. Point mutations—base pair substitutions—can occur anywhere in the chromosome. Within coding regions, such changes may be silent (synonymous) or may result in a change in the translated amino-acid sequence (nonsynonymous), and thereby directly impact the fitness of an organism. Point mutations in noncoding regions typically have little impact on the organism. Genetic recombination involves a rearrangement of DNA, either within a single DNA structure (e.g., a chromosome or a plasmid) or between structures (e.g., a chromosome and a bacteriophage or transposon).6 Although some recombination events may not alter the phenotype of the organism, the insertion of additional DNA into the chromosome can change clinically relevant characteristics, for example, by adding virulence factors or antimicrobial resistance determinants. Studies using gene arrays representing the genome of a bacterial species indicate that insertion and deletion of entire genes is a major source of genetic variation among isolates, with as much as 22% of the genome being dispensable.7,8 Molecular strain typing techniques have been developed that exploit almost every source of genetic diversity; however, no single method short of whole genome sequencing can detect all variations. Pulsedfield gel electrophoresis (PFGE), currently the most widely used method, is based on analysis of large restriction fragments generated by digestion of the chromosome with infrequent cutting restriction enzymes. This approach detects point mutations that create or eliminate those restriction sites and also emphasizes recombination events that involve larger (>20 kb) DNA sequences (e.g., the insertion or excision of lytic phage). Southern blots of insertion elements (e.g., IS6110 typing of Mycobacterium tuberculosis) assess both changes in the number and location of those elements as well as changes in flanking restriction sites. Multilocus sequence typing (MLST) (http://www.mlst.net) analyzes the sequences of seven or more “housekeeping” genes that encode metabolic enzymes.9 Obviously, multiple techniques applied to the same set of isolates may yield different strain assignments. It is more important to appreciate the basis of the differences than to argue that a particular approach is “correct” or “best.”

4.2

Population Structure of Bacterial Species

The accumulation of genetic variation results in the diversity within a bacterial species. Maynard-Smith et al described several classes of population structures based on statistical analysis of genotypes defined by multilocus enzyme electrophoresis.10 In a clonal structure, the genotypes demonstrate strong linkage disequilibrium (i.e., nonrandom association of alleles). Clonal structure results when the rates of interchromosomal recombination (horizontal transfer of alleles among lineages) are low relative to the rates of events generating divergence within lineages (e.g., point mutations, intrachromosomal recombination). Individual genotypes are globally distributed and can be recovered from epidemiologically unrelated hosts. Escherichia coli and Salmonella represent clonal species. The clonal structure is appropriately represented by a tree. Panmictic species, such as Neisseria gonorrhoeae, represent the opposite extreme, with essentially random association of alleles at different chromosomal loci resulting from high rates of recombination among isolates representing a functionally sexual population. This structure is best depicted as a lattice or mesh. ©


5

Number 10

MM11-A

An epidemic structure results from the recent, explosive expansion of a single genotype in an otherwise panmictic species. This effect can be resolved by statistical analysis of the distribution of genotypes. The topology of the epidemic structure is a mesh with occasional nodes showing a burst of offshoots; examples include N. meningitidis and S. pneumoniae. Molecular strain typing to identify epidemiologically related isolates can be performed without reference to the underlying population structure, which describes evolutionary processes. However, the correlation of results from different typing systems may be greatly impacted by structure. Among isolates from a clonal population, more discriminatory techniques will unambiguously subdivide sets of isolates that were indistinguishable by less discriminatory techniques, with little “crossover.”11 However, among panmictic populations, the relationships defined by different typing systems are less predictable and may be incongruent, particularly for methods that assess limited numbers of chromosomal loci.

4.3

Impact of Selective Pressure on Diversity

Virulence factors and antimicrobial resistance determinants represent powerful selective pressures acting on bacterial populations. Consequently, the human pathogens within a species may represent only a subset of genotypes within the species as a whole, which can include clinical, commensal, animal, and environmental strains. For example, the genetic diversity among invasive isolates of E. coli cultured from patients with pyelonephritis or bacteremia is relatively limited compared to isolates colonizing the intestinal tract.12 The pathogenic isolates are characterized by virulence factors that enhance colonization, persistence, and invasion at the infecting site; such virulence determinants are absent from most commensal (i.e., fecal) isolates.13,14 Further, particular bacterial genotypes characterized by unique concatenations of virulence genes (or alleles) may emerge and disseminate relatively rapidly, resulting in an abrupt increase in the frequency and/or severity of particular infections.15 While these infections may have no apparent epidemiologic relationship, the isolates may have indistinguishable genotypes because dissemination has occurred more quickly than accumulation of readily detectable genetic changes. The most striking recent example of this is the emergence of “community-acquired” strains of methicillin-resistant S. aureus (MRSA). MRSA were isolated from 249 (59%) of 422 patients with acute purulent skin and soft-tissue infection seen in emergency departments in 11 different cities during August 2004. Of the 218 MRSA isolates submitted for molecular typing, 212 (97%) were closely related to a single PFGE type (designated USA300) and 156 (74%) represented a single pattern (were indistinguishable).16 Similarly, a single strain of Clostridium difficile with distinctive virulence and resistance genotypes has been recently associated with geographically and temporally diverse nosocomial outbreaks.17-19 Other examples of this process include the Streptococcus pyogenes isolates associated with toxic-shocklike syndrome, antibiotic-resistant S. pneumoniae,20 and E. coli O157:H7. Strains of MRSA that became endemic in hospitals in the 1980s and 1990s often represented a limited number of lineages.8,21-23 In some instances, an existing, but unappreciated pathogen will emerge in association with a new opportunistic niche, such as the “epidemic” of vaginal toxic-shock syndrome associated with new classes of tampons.8,15 While the precise identification of such lineages lies in the realm of population genetics, the phenomenon impacts strain typing because isolates that have no meaningful epidemiologic relationship may be closely related genetically and, thus, difficult to differentiate.

4.4 4.4.1

Applications of Molecular Strain Typing Episodes of Infection Within an Individual Patient

The organisms representing an acute, invasive bacterial infection are typically monoclonal, consistent with arising from a single organism.24 In patients with multiple episodes of infection due to the same 6

©


Volume 27

MM11-A

species, the demonstration that each episode represents a different strain is most consistent with reinfection.25-27 Finding that the infecting isolates represent the same strain is consistent with relapsing infection, but does not rule out sequential infections from a common source, including respiratory or gastrointestinal colonization. Conversely, isolation of multiple different strains during a single putative episode of infection (e.g., among isolates of coagulase-negative staphylococci from different blood cultures28) is considered evidence of contamination rather than true infection. However, in some instances, particularly patients with acquired immunodeficiency syndrome (AIDS), true polyclonal infection (concurrent invasive infection by multiple different strains) has been documented.29 4.4.2

Outbreaks

The application of molecular strain typing to outbreak isolates is based on a number of idealized assumptions as emphasized in the Definitions section (see Section 3.1). •

Isolates representing an outbreak are recently derived from a single precursor.

•

Consequently, outbreak isolates will have the same genotype (i.e., represent a single strain).

•

Epidemiologically unrelated isolates will have different genotypes.

These assumptions require that the typing system have appropriate discriminatory power and reproducibility for the outbreak species. These performance characteristics are impacted by the genetic diversity of the bacterial species and by its population structure. Quantitative aspects of these characteristics are discussed in detail below. As noted, genetic variation can occur during the course of an outbreak, resulting in isolates that are similar but no longer indistinguishable. If an acute outbreak becomes an endemic problem, as exemplified by hospital-associated MRSA, then inevitably the processes of genetic variation cause the genotypes of successive isolates to diverge. Different techniques and approaches to interpretation may be required for analyzing isolates in situations that extend over greater time and space. Most outbreaks, particularly those in hospitals, represent a single strain of a single species. Food and water-related outbreaks may comprise isolates representing two strains of two different species.30 The principles described here can readily be applied to each species separately. Polyclonal infection due to two different strains of the same species has been described29,31 and can generally be reliably resolved within the context of an individual patient. Conceivably, a single outbreak involving multiple persons could involve two different strains of the same species; analysis of this very rare situation is beyond the scope of this document. 4.4.3

Surveillance

The use of molecular strain typing in conjunction with conventional epidemiologic surveillance is an area of active development. PulseNet, coordinated by the Centers for Disease Control and Prevention (CDC, Atlanta, Georgia, USA), is a network of public health laboratories that share the results of PFGE typing of multiple pathogens, including E. coli O157:H7, typhoidal and nontyphoidal Salmonella serotypes, Listeria monocytogenes, Campylobacter, Vibrio cholera, and Shigella. Data are acquired using standardized protocols and can be compared directly with the central database via the Internet 32 (http://www.cdc.gov/pulsenet/). This system has successfully resolved the widespread emergence of specific strains of enteropathogens.33,34 Databases of PFGE profiles for MRSA isolates have recently been developed.23,35 The Health Protection Agency of the United Kingdom provides typing databases in support of the outbreak studies conducted in ©


7

Number 10

MM11-A

Europe by Enter-Net, which focuses on human enteric pathogens, and HARMONY, which emphasizes antibiotic-resistant nosocomial pathogens (http://www.hpa.org.uk/cfi/bioinformatics/dbases.htm). Typing systems based on nucleotide sequences (e.g., MLST, gene arrays,8 emm typing for group A streptococci [http://www.cdc.gov/ncidod/biotech/strep/emmtypes.htm]) are more readily compiled into shared databases. The application of such systems in large-scale collaborative surveillance programs is expected to increase substantially in the near future. However, the routine strain typing of nosocomial isolates is a substantially different, and more problematic, question. Some reports suggest the effort is cost-effective and can reduce the incidence of nosocomial infection. However, rigorous studies are lacking. There are intrinsic limitations, including the presence of endemic strains contributing to background rates of infection, the restricted diversity of nosocomial pathogens due to selective pressures of antibiotic use, the natural tendency of organisms to diverge over time, and the challenge of deciphering infecting versus colonizing organisms in patients with complex illnesses. In addition, valid study designs for investigating the value of routine strain typing remain undefined, particularly with regard to blinding and controls. Thus, while strain typing can be extremely useful in evaluating putative outbreaks identified by conventional hospital surveillance, blind testing of all isolates in the absence of epidemiologic data is not recommended as a routine component of infection control. 4.4.4

Population Genetics

Molecular strain typing for epidemiologic studies can be successfully practiced as an empiric art. That is, the techniques need to be adequate for addressing a particular type of problem, but do not need to meet particular a priori requirements. In contrast, methods for defining the population structure of a bacterial species should assess variation under neutral selection at multiple independent loci around the chromosome and should be validated using large, diverse isolate collections and rigorous statistical techniques.36 The previous method of choice—multilocus enzyme electrophoresis—has been supplanted by MLST9 (see Section 6.4) and gene arrays.8 The molecular strain typing methods most commonly used in epidemiology—those based either on restriction fragment polymorphisms or on variations in the amplification products of a polymerase chain reaction (PCR)—are not applicable to population genetics (see Section 7.3). One of the features that makes sequence-based methods increasingly attractive is that they are potentially effective for both purposes.37

5 5.1

Validation of Typing Methodologies Reproducibility

Reproducibility refers to the ability of a technique to yield the same result when the same strain is tested repeatedly. Reproducibility is influenced by both technical and biologic factors. The former is commonly assessed by performing multiple assays of replicate aliquots of a single isolate. In general, technical reproducibility is excellent for methods based on restriction digests and nucleotide sequences, but can be less consistent for PCR-based approaches. Biologic variation results in a strain typing system demonstrating genotypic differences among independent isolates that, based on a priori epidemiologic criteria, should represent a single strain. The genotypic differences observed are typically minimal and often reflect a single genetic event (e.g., a sequence change that alters a restriction site; recombination with insertion or deletion of DNA). Thus, the genotypes, while no longer indistinguishable, remain similar or closely related. Such variation can constitute a challenge when interpreting and reporting typing results. Systematic assessment of biologic variation is thus an important aspect of characterizing and validating a typing system. Because of the substantial differences in growth conditions in vitro and in vivo, simple laboratory passage does not adequately reflect the variation likely to be encountered during infection and transmission. 8

©


Volume 27

MM11-A

Sets of isolates suitable for assessing biologic variation include (a) isolates representing a bona fide outbreak; and (b) isolates representing a single episode of infection in one patient. Recent studies of M. tuberculosis suggest that the rate of variation during the course of an infection may be lower than the rate during transmission of infection.38 Unfortunately, truly validated sets are difficult to obtain, particularly for some species, and even compelling epidemiologic evidence can be incorrect. Both clinical and experimental evidence indicates that acute invasive infection typically arises from a single organism.24,39 Nevertheless, isolates from a single episode of infection often show detectable differences29 and, occasionally, remarkable variation.40 A suitable set of isolates can be derived from cultures of independent specimens from sterile sites (e.g., multiple blood cultures). Multiple isolates derived from a single specimen may be used if they represent subcultures of independent colonies isolated directly on a primary culture plate. Due to potential selection biases, multiple independent isolates cannot be reliably obtained from a single specimen inoculated into liquid (broth) medium. Assessing reproducibility requires attention to both the frequency of variation and the extent of variation. Consider a set comprising 20 independent isolates of a single strain. At the simplest level, reproducibility can be expressed as the proportion of isolates that are indistinguishable from the expected profile. For example, if restriction profiles of 18 isolates were indistinguishable with the remaining two isolates showing detectable variation, then the reproducibility would be 0.90 (18/20). Describing the extent of variation depends on the properties of the typing system; in general, the greatest variation observed should be noted. For PFGE and other methods based on analysis of restriction fragments, the total number of different fragments is determined (see Section 6.2). Thus, to continue the example above, if one isolate differed from the modal pattern by two fragments and the second isolate by three fragments, then the extent of variation detected by the reproducibility assessment would be three fragments. As illustrated by Tenover et al, a single genetic event can result in differences of two or three fragments in PFGE typing.3 In nucleotide sequencing, the relationship between sequence differences and genetic events is typically more transparent. At this time, there is no method-independent metric for describing the extent of variation observed among isolates.

5.2

Discriminatory Power

Discriminatory power refers to the ability to differentiate among epidemiologically unrelated isolates. Ideally, each unrelated isolate is detected as unique, but, in practice, some are indistinguishable. Hunter and Gaston described the use of Simpson’s index of diversity to calculate discriminatory power as the probability that two epidemiologically unrelated isolates are distinguished by the typing system (i.e., designated as different strain types).41,42 The formula is:

D =1

1 N (N - 1)

K

Σ n (n

i= l

i

i

- 1) ,

where K is the number of distinct types obtained by the method, N is the total number of isolates examined, ni is the number of isolates of the ith type, and D is the index of diversity (range 0 to 1). If D is >0.90, the typing system would be considered to have effective discriminatory power.41 Hunter subsequently described a modified formula to accommodate typing systems in which typing results may be distinguishable, but the difference is not considered reliable for designating a distinct, different strain type. For example, in phage typing, two isolates that differ only in their reaction to a single phage cannot be reliably considered epidemiologically “different” strains.42 This emphasizes a critical element of this process—assessment of discriminatory power must be based on distinct types; that is, those that meet criteria for “different.” The typing results for two isolates may be distinguishable, but the differences may not be meaningful if comparable variation is observed among isolates that meet a priori criteria for being epidemiologically related (see Sections 5.1 and 5.3.1 for ©


9

Number 10

MM11-A

further discussion). Consequently, reproducibility and discriminatory power are intimately connected.42 Specifically, for a given typing method, as reproducibility decreases, whether due to biological or technical variation, discriminatory power is also likely to decrease, since two genotypes must demonstrate a greater number of differences in order to be reliably defined as representing “different” types (i.e., distinct strains). An important corollary of this relationship is that while applying multiple typing methods (or multiple restriction enzymes) may increase the number of distinguishable types, this cannot be assumed to improve discriminatory power; the impact on reproducibility must also be considered. While excellent discriminatory power is required for a method to be broadly applicable, a poorly discriminatory method may be suitable when the index isolate proves to have a unique or rare type.

5.3

Requirements for Characterizing a Molecular Typing System

Over the past decade, there have been literally hundreds of reports describing DNA-based strain typing systems. Unfortunately, many systems are inadequately characterized. It is clearly insufficient to evaluate a typing system by comparing the results for a single, well-defined outbreak with a handful of epidemiologically unrelated “control” strains. Almost any system will have sufficient reproducibility and discriminatory power to indicate that the outbreak isolates are more similar to each other than to a few random unrelated isolates. Struelens et al emphasized the need for specific criteria for evaluating typing systems.43 Although formally validated metrics for discriminatory power and reproducibility have not been established, the following guidelines, drawn in part from the suggestions of Struelens et al, are proposed as operational standards for characterizing any molecular typing system. Good experimental design suggests that all the characterizations should be performed blinded. 5.3.1

Characterizing Reproducibility

Technical reproducibility should be assessed by two independent laboratories analyzing at least ten replicate aliquots of a pure subculture derived from a single isolate. All analyses are expected to yield indistinguishable genotypes (i.e., 100% concordance). If any variations are observed, then replicates from at least five to ten different strains should be examined to provide a valid estimate of technical reproducibility. Biologic reproducibility should be assessed by analyzing at least ten sets of isolates, where each set comprises at least five independent isolates recently derived in vivo from a common precursor based on a priori epidemiologic criteria (minimum, 50 isolates in total). As discussed above, the most practical and perhaps reliable sets will be obtained from single episodes of acute invasive infection. Based on current experience, it is expected that within one or more sets, there will be some genotypic variation (i.e., 1000 10-19 20-300 15-20 5-500 20-25 5-500

Alternative Enzymes* ApaI AsnI, DraI SmaI SstII, NruI SpeI ApaI NotI, SfiI RsrII AsnI NotI SmaI, NotI, AscI XbaI, DraI, AsnI BglII, SpeI, SfiI DraI, SspI, XbaI BluI BluI SstII, CspI, EagI SstII, KpnI DraI, SpeI SfiI, ApaI SfiI NotI

These are generally the most effective enzymes for epidemiologic studies of the species indicated; the alternative enzymes may be useful in particular analyses. † For E. coli O157:H7, BluI is recommended.

Prior to digestion, plugs should be equilibrated into a small volume of the appropriate restriction enzyme buffer for 30 minutes at the reaction temperature recommended by the manufacturer. Enzyme (typically 10 to 50 U per plug) in appropriate buffer is then added, mixed, and the plug is incubated for two to four hours at the appropriate temperature. Plugs are carefully inserted in the gel slots and sealed with 1% lowmelting temperature agarose.46 Alternatively, plugs can be melted by heating to 65 °C for five minutes and then the tubes held at 40 °C to maintain the liquid state until the melt is loaded very gently into the well; in this procedure, it is not necessary to seal the wells.74 Care should be taken not to overheat the plugs, as high temperature can degrade the DNA.

14

©


Volume 27

MM11-A

Detailed protocols that have been validated in multiple laboratories have been published for several epidemiologically important species including S. aureus 72 as well as E. coli, Salmonella, Shigella and V. cholera (http://www.cdc.gov/pulsenet/protocols.htm). 6.2.1.2

Running the Gel

The optimal running parameters for resolving the DNA digests depend both on the PFGE apparatus and on the anticipated size range of the fragments, which is generally predictable for isolates of the same species digested with a particular enzyme. The most popular PFGE system applied to bacterial strain typing uses a contour-clamped homogeneous electric field (CHEF), in which a hexagonal array of electrodes alternates between uniform fields with an angle of reorientation of 120°. DNA fragments migrate through the agarose in a zigzag course as the electric field changes direction. Larger molecules require more time to reorient and therefore migrate less distance “down” the gel. For a given set of conditions (e.g., agarose, voltage, and temperature), the range (or ramp) of time intervals between changes of field direction (or pulse time) is the primary factor determining the size range of DNA resolved.75 The other factors affecting the mobility of the DNA fragments include DNA concentration, type and concentration of the agarose, ionic strength of the electrophoresis buffer, voltage, temperature, and run time.76 Changes in any of these factors can alter the profiles obtained, and thus complicate the analysis, both within a single gel as well as between different gels. The following conditions are generally appropriate for resolution of macrorestriction digests of bacterial genomes, which typically include fragments of 50 kb to 1 Mb: •

switch times – a ramp ranging from five to 40 seconds provides optimal separation of DNA fragments from 50 to 600 kb; increasing the pulse time to 75 seconds extends the separation to 1 Mb46;

•

agarose – 1% concentration, special PFGE-grade, with a high physical strength and low electroendosmosis to permit more rapid electrophoresis without loss of resolution68;

•

buffer – low ionic strength, e.g., 0.25x or 0.5x TAE (Tris-acetate-EDTA) or TBE (Tris-borateEDTA), to reduce heat generation and shorten the run times75;

•

voltage – 6 V/cm;

•

temperature – 14 °C; maintained by a system that both cools and recirculates the buffer to prevent the generation of temperature gradients within the gel during the run46; and

•

run time – 18 to 22 hours.

Gels provide greater resolution and are easier to interpret if the fragments are dispersed over as much of the length of the gel as possible. The dynamic range of the gel—the distance from the largest to the smallest fragment—depends on the size of the fragments in the digest and the range of switch times used. While published protocols provide a useful guideline, a laboratory implementing PFGE should determine the impact of varying switch times in its own system, while holding other parameters constant. Following electrophoresis, the gel is stained in 0.5 µL/mL ethidium bromide in water (typically achieved by a 1:10 000 dilution of a stock solution) for 10 to 30 minutes, destained by one to two washes of 15 to 30 minutes each in distilled water,67,68,70 and the DNA fragments visualized under ultraviolet (UV) light. Alternatively, gels can be stained with proprietary reagents, some of which are reported to be more sensitive than ethidium bromide. The image may then be photographed or captured digitally.46

©


15

Number 10 6.2.2

MM11-A

Standards and Controls

There are three classes of controls related to quality control and interpretation of PFGE gels. •

Control for preparation and digestion of DNA. This is typically a strain that is the same species as the isolates under study, that has been successfully processed in the past and whose PFGE pattern is known. The strain should be grown and processed in parallel with the study isolates.

•

Size standards serve to confirm that the pulsing pattern has successfully resolved fragments over the expected size range. These standards are essential for computer-assisted gel analysis (see Section 7.2); they also aid in identifying genotypes and considering the molecular basis for differences. Size standards are most useful when run every fifth lane, so no study isolate is more than two lanes from a standard. Size standards may be obtained commercially (e.g., lambda concatemers) or prepared locally using a bacterial strain for which the sizes of the restriction fragments have been defined, ideally by using a strain whose genome has been fully sequenced.

•

Reproducibility standard. When performing computer analysis across multiple gels, the molecular size standards are used to normalize each gel to permit the comparison of fragment patterns for isolates run in different gels (see Section 7.2). An additional, independent standard (i.e., a single strain that is run in every gel) is needed to quantitate the technical reproducibility across the gels included in the analysis (see Section 7.2.1 for additional discussion). The variation reported by the computer analysis among the patterns of this reproducibility standard defines the minimal level of variation that might be expected of an outbreak strain. The DNA for the reproducibility strain can be prepared in batches, stored in aliquots, and individual portions digested at the time each gel is run. Alternatively, this strain can be prepared fresh for each gel and can serve as the methodology control as well.

6.2.2.1

Quality Control (QC) of Equipment Temperature

The processing and digestion of the DNA and the resolution of the fragments during electrophoresis are both strictly dependent on temperature. The water baths or controllers used to maintain temperature should be monitored using independent, validated thermometers. Observed temperatures for each run should be recorded on data sheets. This may prove particularly useful in troubleshooting. 6.2.3

Troubleshooting the Gel

A good quality gel shows clear separation of digested components (good resolution) and is easily amenable to analysis either by eye or an image analysis system. Artifacts can be caused by a variety of factors; among the most common are delayed or incomplete inactivation of endogenous nucleases, incomplete lysis of the organism, incomplete removal of proteinase K, partial restriction enzyme digestion of the DNA, and variations in the electrophoresis conditions (see Figure 1).

16

©


Volume 27

MM11-A 1

2

3

4

5 6

7

8

9 10

11 12

13 14 15

Figure 1. Pulsed-Field Gel Electrophoretic Patterns. Lanes 1, 5, 7, 9, and 11 represent appropriately digested DNA; the problems associated with the remaining lanes are detailed below. •

DNA degradation – Bacteria contain endogenous nucleases that can rapidly degrade DNA. Even partial degradation produces faint blurred digests that may be difficult to analyze (see Figure 1, lanes 2 and 3); more extensive degradation can yield a broad, uninterpretable smear in the lower half of the gel (see Figure 1, lane 4). Failure to chill and wash the bacteria promptly when harvesting the broth culture is a common cause of degradation, and the plugs must be prepared smoothly and rapidly to assure that endogenous nucleases are promptly inactivated by the proteases and the EDTA.46,66 DNA degradation frequently occurs with isolates of Clostridium spp., and occasionally with C. jejuni or P. aeruginosa. Some C. jejuni PFGE protocols recommend treatment of the bacterial suspension in 10% formaldehyde for 15 minutes to inhibit DNase activity.67,77 Alternative approaches for preventing this problem include adding thiourea in the running buffer78 or using HEPES (4-(2-hydroxyethyl)-1piperazineethanesulfonic acid) as running buffer.79 Another potential source of DNA degradation is bacterial contamination of the running buffer used in the electrophoresis apparatus. In particular, Acinetobacter and Pseudomonas spp. can grow in such buffers, often unnoticed because they cause little opacification, and release DNases. This problem should be suspected when all the DNA in the gel is degraded, including the molecular weight and reproducibility standards. Cleaning the entire chamber with 70% ethanol is recommended both to denature the DNases as well as decontaminate the apparatus.

•

Incomplete restriction digestion – Partial restriction digests typically include a single, unusually large fragment, as well as intermediate-sized fragments of variable intensity between the expected welldefined DNA fragments (see Figure 1, lane 6).66 In some circumstances, the extra fragments can be misinterpreted as indicating a valid difference in the restriction pattern. Incomplete digestion can occur if any of the reagents used to prepare the DNA (proteinase K, EDTA, detergent) are not completely removed, resulting in loss or inhibition of the restriction enzyme; the extensive washes in water and TE are essential to removing all traces of these agents from the plugs.74 In addition, some types of agarose are reported to contain inhibitors for certain restriction enzymes.74 If the cell lysis is believed to be adequate, but the restriction digestion is incomplete, the plugs can be washed thoroughly with TE and redigested with fresh restriction enzyme.66 Preparing completely digested DNA also depends on having sufficient quantity of restriction enzyme, in the appropriate buffer, for adequate duration. Most restriction enzymes should be used only with

©


17

Number 10

MM11-A

the manufacturer’s specific buffer, typically provided as 10x concentrates. Using other buffers may result in decreased activity or nonspecific digestion. Diffusion of the enzyme through agarose is obviously slower than in solution, and incubation times in excess of one hour are needed to achieve complete digestion. •

Improper plug handling – The appearance of the restriction fragments can be distorted by mechanical problems in preparing or inserting the plugs, including plugs that are unevenly cut, only partially inserted into the gel, or damaged during insertion (see Figure 1, lanes 8, 10, and 12).

•

Distortions due to electrophoresis – Comparing PFGE patterns requires that all the lanes of the gel be straight, parallel, and represent comparable DNA migration. Curved lanes (see Figure 1, lanes 13 to 15) indicate heterogeneity in either the electric fields (e.g., because of the failure of several electrodes) or the temperature across the gel.75 Variations in temperature often reflect mechanical problems including inadequate circulation of buffer or failure to have the gel or the chamber level. The depth of buffer around the gel directly impacts the flow of current; where there is a thinner layer of buffer, there is increased resistance and, consequently, increased temperature. As a result of these gradients, the DNA fragments in different lanes migrate at different rates. Gel distortions due to a similar process can occur if the gel is not fastened securely during electrophoresis or if air bubbles are trapped underneath the gel.

•

Incomplete lysis and/or incomplete protein digestion – Complete lysis of the organism and digestion of the proteins are essential to fully exposing the genomic DNA so it can be properly restriction digested. Incomplete lysis may leave large cell-wall fragments that impede movement of the restriction fragments out of the plug, resulting in little DNA migrating into the gel. Evidence of incomplete lysis includes the presence of intense staining of the material in the well; the migration of a single, very large DNA fragment; and very faint chromosome bands (not shown).46,66 Agarose plugs containing incompletely lysed bacteria can be reprocessed by using the TE washout procedure to remove the proteinase K and then recycling the plugs through the entire procedure beginning with lysis solution.66 Incomplete protein digestion typically manifests as smearing proximal to the highest DNA band as well as between the DNA fragments (not shown). This can be rectified by increasing the amount of proteinase K or the length of digestion.

6.2.4 6.2.4.1

Strengths and Limitations of PFGE Strengths

•

Experience – PFGE is currently the most widely used strain typing system for epidemiologic studies and there is extensive literature attesting to its utility for analyzing common pathogens.50 Consequently, many scientists in academia and industry are familiar with the technique, experienced in analyzing the gels, and represent a source of advice and assistance for laboratories interested in implementing PFGE.

•

Adaptability – PFGE is based on clear, simple principles and can be readily adapted to uncommonly encountered species.80

•

Flexibility – The DNA in agarose is stable for years at 4 °C, facilitating the use of internal standards and ongoing studies. Moreover, the DNA can be easily released into solution for use in other protocols, including Southern blots and PCR.50 Prolonged storage of digested DNA is not recommended, because the smaller fragments may diffuse out of the plug into the storage buffer.

18

©


Volume 27 6.2.4.2

MM11-A

Limitations

•

Cost – PFGE requires relatively expensive, specialized equipment, costing from $15,000 to $20,000 (USD).50,81

•

Time – The procedure is technically demanding with a turnaround time of 24 to 72 hours, even with recent simpler protocols using shorter incubation periods.46,67-69,72,82

•

Poor scalability – Additional effort is required to achieve sufficient reproducibility to compare patterns across multiple gels and particularly among different laboratories. Specifically, to ensure adequate normalization of patterns and accurate fragment size estimates, suitable molecular size markers should be run in every fifth lane; and to quantify intergel variability, an additional, independent reproducibility isolate should be included in each gel.46,83

•

Species-specific limitations – Certain organisms (e.g., Clostridium difficile and Aspergillus spp.) cannot be reliably typed by PFGE because their DNA cannot consistently be extracted intact.81 Some pathogens, e.g., MRSA, H. influenzae type b, and E. coli O157:H7, represent genetically restricted subsets of strains within a species84-86 and consequently, epidemiologically unrelated isolates may have very similar genotypes and may be indistinguishable by PFGE.50,51,87 However, this issue represents an inherent problem for all typing methods.

•

Impact of nonchromosomal elements – While the primary sources of variation in PFGE patterns are chromosomal changes (e.g., insertions, deletions, rearrangements, point mutations), in some instances extrachromosomal elements can impact the size and number of fragments. Examples include lysogenic phages in S. aureus88 and, less commonly, large (>75 kb) plasmids in gram-negative bacilli.89 Because these elements can be acquired or lost or both in short time frames, the variation introduced may complicate an epidemiologic analysis.

6.2.5 6.2.5.1

Application of PFGE Outbreaks

PFGE typing has been applied to the following: local epidemiologic investigations of suspected outbreaks of infection; resolving relapse and reinfection in individual patients; and studies of the clonality of acute and chronic infections and colonization, including distinguishing reinfection vs. recurrence.24,29,31,67,90 6.2.5.2

Surveillance

PFGE strain typing has also been successfully applied to national and international surveys designed to detect the emergence, geographical spread, and molecular evolution of epidemic clones of MRSA, 91,92 S. pneumoniae,93-95 Salmonella enterica Typhimurium,96 Pseudomonas aeruginosa,97 and Neisseria meningitidis.98 PFGE, in combination with standardized protocols, appropriate internal controls, and computer-assisted databases, has been successfully used by laboratory consortia to facilitate large-scale national and international studies of epidemic organisms. PulseNet, coordinated by CDC, is a network of public health laboratories throughout North America that share the results of PFGE typing of multiple pathogens, including E. coli O157:H7, nontyphoidal Salmonella serotypes, Listeria monocytogenes, and Shigella. Data are acquired using standardized protocols and can be compared directly with the central database via the Internet32 (http://www.cdc.gov/pulsenet). European isolates predominate in databases available through the Health Protection Agency in the United Kingdom in support of Enter-Net, which focuses on human enteric pathogens ©


19

Number 10

MM11-A

(http://www.hpa.org.uk/hpa/inter/enter-net_outbreaks.htm), and HARMONY, which emphasizes antibioticresistant nosocomial pathogens (http://www.hpa.org.uk/srmd/bioinformatics/harmony/harmonydb.htm). 6.2.5.3

Population Genetics

PFGE is not suitable for determining the population genetics of bacterial species. Although computer programs may construct dendrograms that appear similar to evolutionary trees, the analysis is inherently unreliable (see Section 7.3).

6.3

Ribotyping

Ribotyping is a strain typing method based on Southern blot analysis of the restriction fragment length polymorphisms (RFLPs) associated with the ribosomal operon(s).50,99,100 Operons are clusters of genes whose expression is controlled by a single promoter. The genes encoding the three species of rRNA (23S, 16S, and 5S) present in 70S ribosomal particles are organized into a polycistronic transcription unit, the rrn operon, which in many bacterial species is present as multiple copies distributed around the chromosome. A Southern blot prepared from a routine chromosomal restriction digest and hybridized with a specific probe provides a pattern of 7 to 15 restriction fragments (referred to as a ribotype) that can be used for strain identification. Bacterial rRNA genes have been highly conserved during evolution; consequently, probes made from a variety of sources will hybridize to the ribosomal operons of many bacterial species.101 Such probes include the rRNA of Escherichia coli itself99; DNA fragments derived from the cloned rrn operon102; and cDNA obtained by reverse transcription of E. coli 16S and 23S rRNA.103 A variety of different labels, e.g., 32P 104 or ligands (digoxigenin, biotin),96,105 may be incorporated into the probes and the hybridized restriction fragments detected by autoradiography, enzymatic color reactions, or chemiluminescence.106 Several detailed protocols have been published.107-109 A device that performs ribotyping and is virtually fully automated is commercially available, although at considerable expense. Some investigators have proposed that it is a cost-effective strain typing system for large institutions, especially when used as a screening tool and supplemented by PFGE.110 6.3.1

Strengths and Limitations of Ribotyping

The major advantage of ribotyping is that the basic method is directly applicable to a wide range of bacterial species, since ribosomal operons are universal and highly conserved. However, the method also has several major limitations. •

Difficulties in interpretation – The ribotyping patterns often include some faint bands, presumably reflecting restriction fragments that contain only short sequences of hybridizing DNA.50 Detection of these fragments can be variable due to technical factors and thereby complicate interpretation.

•

Modest discriminatory power – This is the major limitation of the method. Ribotyping has proven less discriminatory than PFGE or sequence-based methods for several species, including S. aureus,56 Pasteurella haemolytica,111 Enterococcus spp.,112,113 and Streptococcus pyogenes.114 The restricted discriminatory power derives at least in part from the highly conserved nature of sequences within the ribosomal operon. Ribotyping is also poorly discriminatory for those species with reduced numbers of rnn operons. For example, Mycobacterium spp., Mycoplasma spp., and Borrelia burgdorferi typically have only a single rnn operon in their genome.115 Although discriminatory power can be increased by preparing multiple digests, each using a different restriction enzyme, this markedly increases the time and effort involved.116

20

©


Volume 27 6.3.2

MM11-A

Application of Ribotyping

Ribotypes have been shown to be reasonably effective at identifying nosocomial outbreaks117-119 and appear highly effective for some species (e.g., Salmonella typhi).102 As noted above, some investigators have proposed that the automated system is useful as a screening tool for surveillance studies. In addition, some studies have suggested that specific sets of fragments within a ribotype pattern can be useful taxonomic markers for identification of genera, species, or subspecies. However, sequencing is now the method of choice for this purpose. Ribotyping, like PFGE, is not suitable for population genetics studies.

6.4

Sequence-Based Strain Typing

Nucleotide sequencing is emerging as the method of choice for molecular strain typing of bacterial pathogens. This application has become practical by exploiting PCR and automated sequencers. A simple crude extract can be readily prepared from a fraction of a single colony and provides sufficient excess chromosomal template to ensure any incorporation errors that do occur during amplification will be below the limits of detection in final DNA sequences. In contrast, such errors can become fixed in cloned DNA. Detailed guidance for performing PCR and sequencing are considered in the most current editions of CLSI document MM3—Molecular Diagnostic Methods for Infectious Diseases and CLSI/NCCLS document MM9—Nucleic Acid Sequencing Methods in Diagnostic Laboratory Medicine, respectively. Currently, there are at least four approaches for molecular strain typing based on nucleotide sequencing. •

Multilocus sequence typing (MLST) involves the analysis of DNA sequences of multiple metabolic genes.

•

Multiple loci variable number tandem repeat assays (MLVA) are based on the analysis of multiple loci that contain variable numbers of tandem repeats (VNTRs).

•

In some instances, effective strain typing can be achieved by sequencing a single highly polymorphic locus, such as the staphylococcal protein A gene (spa typing) or emm typing of S. pyogenes.

•

Patterns of single nucleotide polymorphisms (SNPs) have also been used for strain typing.

6.4.1

MLST

MLST represents the application of nucleotide sequencing using the principles established by multilocus enzyme electrophoresis (MLEE).12 Isolates are classified based on polymorphisms in the coding sequences of seven to ten different metabolic (“housekeeping”) genes; the set of loci to be sequenced must be identified and validated for each bacterial species. Like its predecessor MLEE, the primary application of MLST is analyzing population genetics. When used for strain typing of isolates associated with outbreaks, its discriminatory power is typically less than PFGE. The “clock speed” or secular rate of change in the coding regions used in MLST is relatively slow, as befits a system designed to analyze the population structure of entire bacterial species. In contrast, PFGE is sensitive not only to nucleotide sequence changes affecting the target restriction site, but also to chromosomal insertions, deletions, and rearrangements. Hence, PFGE typically has a higher “clock speed” and greater discriminatory power. MLST databases have been established for a variety of organisms and a website (http://www.mlst.net) is available to help analyze sequence data generated for a number of bacterial species, ranging from N. meningitidis9 to S. aureus.21 MLST may be useful in some epidemiologic studies, particularly if discriminatory power is enhanced by examining additional loci.

©


21

Number 10 6.4.2

MM11-A

MLVA

Genomic analysis has revealed that many bacterial species have loci with VNTRs, although the structure and function of the repeats differ widely. VNTR may represent direct or inverted repeats and may be responsible for antigenic variation or gene regulation.120 PCR amplicons produced using specific primers can be analyzed directly for variation in primary nucleotide sequence. Alternatively, variation in the sizes of the PCR products can be assessed by agarose gel or capillary electrophoresis, possibly coupled with computer software to facilitate fragment analysis. Multiplex PCR assays can be developed that allow products from multiple loci to be assessed simultaneously. Note that when fragment size rather than sequence is used, the datasets have the limitations in scalability and portability inherent in any gel-based technique (see Section 6.2.4.2). MLVA has been used successfully to type highly monomorphic organisms, such as Bacillus anthracis and Yersinia pestis121 and to differentiate among lineages of common pathogens,120 such as Shiga-toxinproducing Escherichia coli,122 Salmonella typhi,123 and S. aureus.124,125 Initially, the MLVA assay for B. anthracis was focused on a single variable repeat region, vrrA, which consisted of a series of 12 base pair tandem repeats.126 The addition of other chromosomal loci, as well as sites on the two virulence plasmids (pXO1 and pXO2), has provided remarkable discrimination among isolates of this highly monomorphic pathogen. This system of typing was used during the anthrax outbreak in 2001. MLVA was developed for the common bacterial pathogens explicitly to provide a typing method that was more rapid than PFGE but had comparable discriminatory power appropriate for analyzing outbreaks.122 Several different approaches to VNTR analysis have been applied to strain typing of staphylococci. Nishi et al used PCR primers based on the direct repeat unit (dru) of the hypervariable region associated with the methicillin-resistance determinant and determined the number of repeats present among isolates of S. aureus, S. epidermidis, and S. haemolyticus.127 The method was successful for all staphylococcal isolates carrying mecA. Sabat et al selected five different VNTR loci (sdr, clfA, clfB, spa, and sspA) to permit typing both methicillin-susceptible and -resistant S. aureus.124 Because of the similarity of the sequences associated with sdrC, sdrD, and sdrE, multiplex PCR yields six to seven bands and appears quite discriminatory. Studies by Malachowa et al indicate that MLVA typing results for S. aureus correlate well with those of PFGE.125 6.4.3

Sequence Analysis of a Single Locus

Epidemiologically useful bacterial strain typing has also been achieved by analyzing sequence polymorphisms at a single genetic locus. Notable examples include (a) the Streptococcus pyogenes genes emm and sic (see Section 12.1)128,129; (b) the polymorphic 24-bp variable-number tandem repeat (VNTR) within the 3′ coding region of the S. aureus gene for protein A (spa)37,130; and (c) the variable repeat regions of the S. aureus coagulase gene (coa). Spa typing is particularly effective for typing MRSA from healthcare institutions, providing higher resolution than coa typing,130 with results that parallel those of PFGE.130-132 There are two websites to facilitate standardization of spa nomenclature. One (http://tools.egenomics.com/) is free but password protected; the other (http://www.ridom.de/staphtype) focuses on a proprietary, commercial software system. As noted above, VNTRs are widespread, suggesting that sequence analysis of a single locus could potentially offer a rapid, portable method for molecular strain typing in many bacterial species.56 6.4.4

SNP Analysis

SNP analysis using either real-time PCR or pyrosequencing methods has proven to be a useful typing method for several applications, including differentiating closely related species of bacteria, such as Burkholderia mallei and B. pseudomallei,133 and resolving two closely related serotypes of S. pneumoniae.134 Utilizing SNPs as a method of bacterial strain typing evolved in at least one instance out 22

©


Volume 27

MM11-A

of a search for a simplified method of performing MLST. Robertson et al applied specialized software135 to existing MLST databases for Neisseria meningitidis and S. aureus and defined sets of SNPs in each species that identified several key sequence types with high precision.136 This technique is more rapid and efficient than MLST, but, obviously, has less discriminatory power and cannot identify novel sequence types. SNPs can by detected by several different methods, including kinetic (allele-specific) PCR136 and pyrosequencing of biotinylated PCR products.137 Arnold et al applied the latter approach to isolates of the M. tuberculosis complex.137 Among 99 M. tuberculosis isolates, 68% of isoniazid-resistant organisms had mutations at position 315 of katG and 92% of rifampin-resistant isolates had mutations at positions 516, 526, or 531 of rpoB. SNPs have also been used to identify the three major genotypes of M. tuberculosis and to differentiate M. tuberculosis and M. bovis. Zhao et al also developed SNP-based pyrosequencing assays for detecting drug-resistant isolates of M. tuberculosis.138 A critical advantage of the combination of PCR and pyrosequencing is the ability to provide precise and informative molecular data about species and drug resistance within as little as six hours. This technology offers the potential for more timely and appropriate selection of antituberculosis therapy by providing rapid molecular susceptibility testing of the isolates. 6.4.5

Strain Typing Using Gene Arrays

With the development of high-throughput sequencing, complete bacterial genome sequences are increasingly available; for many pathogenic species (e.g., S. aureus, M. tuberculosis, B. anthracis), the genomes of several different strains have been sequenced. These data can be used to prepare microarrays representing sequences from each of the 1000 to 3000 open reading frames (ORF) identified for that species. Studies using such “genome chips” indicate that insertion and deletion of entire genes is a major source of genetic variation among isolates, with as much as 22% of the genome being “dispensable.”8,139 Thus, there are literally hundreds of “strain-specific” genes that may be present in or absent from individual isolates and may provide a uniquely discriminatory typing system.8 This approach is currently a research tool due to the limited number of species for which multiple genomes have been completed, as well as the expense of the arrays and the equipment. 6.4.6 6.4.6.1

Strengths and Limitations of Sequence-Based Strain Typing Strengths

•

Precise, accurate data – Nucleotide sequences represent discrete, objective, well-defined data. While sequencing errors do occur, these can be readily detected and resolved.

•

Reproducibility – Technical reproducibility is essentially 100%; biological reproducibility is also high and any variation can be expressed quantitatively.

•

Discriminatory power – The genomic sequence is, by definition, the genotype of an isolate. In practice, highly discriminatory results can be obtained from a limited subset of loci.

•

Portability – Sequence data can be shared unambiguously; with contemporary Internet technology, extensive databases can be compiled and accessed in real time.

•

Quantitative analytic methods – Data analysis is also highly scalable with well-developed, robust computer programs providing quantitation of the differences across even large sets of isolates. Sophisticated algorithms integrating sequencing data, mutation rates, and the prevalence of different subtypes within a population can be used to estimate the temporal as well as evolutionary distance between two isolates. This approach is particularly useful in forensic microbiology.

©

23


Number 10 •

MM11-A

Core methodology – While sequenced-based bacterial strain typing has emerged relatively recently, nucleotide sequencing is a mature core technology found in essentially every contemporary biological research institution. Moreover, sequencing techniques continue to improve, new methodologies explicitly designed for genotyping are under development, and the technology is increasingly available in clinical settings.

6.4.6.2

Limitations

•

Cost – Automated sequencers remain a large capital expense. However, service laboratories that can provide results for an entire 96-well plate of PCR amplicons in a few days at modest cost are now widely accessible.

•

Data analysis – Data derived from VNTR loci are more complex and the analytic software currently available is less flexible.

6.4.6.3

Future Expectations

Sequencing technologies continue to develop rapidly, with advances in hardware, software, and reagents. It is expected that with decreasing cost and complexity, sequencing will become increasingly available to clinical as well as research laboratories. Given the advantages of working with sequence data as detailed above, sequence-based approaches will likely become the methods of choice for molecular strain typing. 6.4.7

Application of Nucleotide Sequencing

6.4.7.1

Outbreaks

Sequence-based typing using a hypervariable single-locus can be completed more rapidly than MLST and is thus more appropriate for acute epidemiologic studies. Published reports have described the use of spa typing for S. aureus140 and emm, sic, and sof typing for S. pyogenes.129,141 Undoubtedly, single loci suitable for epidemiologic typing will be identified shortly in additional species. In general, the discriminatory power of MLST is lower than that of PFGE.21,22,142 6.4.7.2

Surveillance

Sequence-based typing is highly effective for identifying the emergence and spread of particularly virulent or transmissible strains. Large-scale collaborations can be established once the target sequence and associated PCR primers are determined and multiple laboratories can readily contribute results to a central database (such as the Ribosomal Differentiation of Medical Microorganisms (RIDOM) or MLST databases), permitting both data acquisition and analysis to proceed essentially in real time. 6.4.7.3

Population Genetics

MLST and gene arrays are the current methods of choice for studies of population genetics21 (http://www.mlst.net). The relative contributions of point mutation and recombination to the initial stages of clonal diversification within different species can be assessed from MLST data, by identifying those sequence types (STs) that are very closely related, i.e., differing at only one of the seven MLST loci (single-locus variants [SLVs]). The sequences of the alleles at the single altered locus are then analyzed to distinguish whether the change in the housekeeping gene has occurred by recombination or by mutation.143

6.5

Repetitive Sequence-Based PCR (rep-PCR)

Rep-PCR employs primers based on short extra- or intergenic repetitive sequences, which are typically present at many sites around the bacterial chromosome.144 When two sequences are located near enough 24

©


Volume 27

MM11-A

to each other (e.g., within a few kilobases), then the DNA fragment between those sites (referred to as an “interrepeat” fragment) is effectively amplified. The fragments comprising the amplicon are readily resolved using agarose gel electrophoresis. Since the number and locations of the repetitive sequences are quite variable, the number and size of the interrepeat fragments generated can similarly vary from strain to strain. Three classes of repetitive elements have been commonly targeted: palindromic sequences (REP [repetitive extragenic palindromic]), consensus sequences (ERIC [enterobacterial repetitive intergenic consensus]), and BOX elements. REP elements are 38-bp extragenic sequences consisting of six degenerate positions and a 5-bp variable loop between each side of a conserved palindromic stem.145 ERIC sequences are 126-bp intergenic elements that contain a highly conserved central inverted repeat.146 BOX elements, which are located in intergenic regions and can form stem-loop structures,147 are mosaics of three subunit sequences referred to as box A, B, and C comprising 59, 45, and 50 bp, respectively. Because one or more of these elements are present in most bacterial species, the method can be applied to a wide range of organisms. Rep-PCR can be performed using published primers and generic PCR and gel electrophoresis equipment. Commercial implementations based on rep-PCR are available. 6.5.1

Strengths and Limitations of rep-PCR

Rep-PCR has several operational advantages: •

Speed – Rep-PCR can be completed within four hours.

•

Modest technical requirements – PCR and gel electrophoresis equipment are now widely available and familiar.

•

Broad applicability – Rep-PCR is versatile and has been applied to a wide range of microbial pathogens, including bacteria and fungi.

•

Moderate discriminatory power – Rep-PCR has been reported to be more discriminatory than plasmid profiling, multilocus enzyme electrophoresis, biochemical characterizations, or ribotyping,148 but less discriminatory than PFGE.149-152

The major, potentially significant, limitation of rep-PCR is poor reproducibility. Even within a single laboratory, strict attention to procedural details is required to achieve reliable results,153 and poor interlaboratory agreement has been observed in some studies.149 A possible explanation for this problem is suggested by an experiment performed by Plikaytis et al, who analyzed the fragments generated by a repPCR system.154 A Southern blot was prepared from the agarose gel resolving the amplified fragments and was probed with a portion of repetitive sequence that should have been present in all products generated if the system performed as predicted. In fact, many fragments apparent in the gel, including some of the most prominent products, were not detected on the Southern blot. This observation suggests that such fragments were generated as a result of “mispriming,” that is, randomly amplified. To the extent that the PCR amplicons produced using primers directed at repetitive sequences actually represent arbitrary priming, then the reproducibility of the approach is likely to suffer appreciably. The commercial rep-PCR systems are reported to have improved reproducibility.150,155 6.5.2

Application of rep-PCR

Rep-PCR has been reported for molecular strain typing of a wide variety of bacteria and fungi, including staphylococci, enterococci, mycobacteria, pathogenic Neisseria spp. enteric gram-negative rods,

©


25

Number 10

MM11-A

Burkholderia cepacia, Acinetobacter spp., Legionella pneumophila, Aspergillus spp., Candida spp., and Fusarium spp.148,155

7

Analyzing Electrophoretic Typing Data

For PFGE and other DNA-based strain typing methods that use gel electrophoresis (e.g., Southern blot and PCR-based techniques), each isolate is represented by a pattern of DNA fragments (colloquially referred to as “bands”). The general requirements for successful analysis of these gels include the following (more stringent requirements for computer-based analysis are detailed in Section 7.2.1): •

complete DNA restriction digests;

•

a good quality gel showing well-resolved fragments without distorting artifacts (e.g., due to bubbles or temperature variances);

•

adequate numbers of molecular size standards—minimally, one in each outside lane plus additional lanes through the middle of the gel; and

•

an image documentation system, either conventional photography (cassettes of rapidly developing 4 in. x 6 in. [10.2 cm x 15.2 cm] film are convenient) or electronic image capture.

7.1

Visual Analysis of Electrophoresis Gels

The simplest procedure for analyzing a strain typing gel is to examine it visually. A single gel can have as many as 20, 30, or even 45 lanes, although up to 20% of the lanes in each gel may be needed for molecular size standards. The largest gels are most useful for initial screening, particularly to identify “indistinguishable” isolates and thereby permit subsequent runs to focus on isolates that show variation. Smaller gels with 10 to 12 isolates plus three to four standards are easier to analyze and often provide better resolution. For sets of isolates that can be encompassed by one or two gels, the patterns may be efficiently analyzed by simple visual inspection (“by eye”). This includes evaluations of small outbreaks as well as studies of individual patients. However, even with internal size standards, analyzing patterns across multiple gels can be problematic, and additional gels may be required to provide side-by-side comparisons of DNA digests.

7.2

Analyzing Electrophoresis Gels by Software

For larger series of isolates requiring multiple gels, visual analysis is unwieldy and unreliable, and the use of computer-assisted systems is essential. The image of the gel is digitized so the patterns of fragments resolved in the gel can be analyzed by the computer using mathematical algorithms. Images representing the profiles of thousands of isolates can be stored and analyzed, thus permitting large multicenter studies and ongoing surveillance programs that would be impossible using manual systems.46 Closely related patterns can be reliably identified using the computerized system, although final determination that two patterns are explicitly indistinguishable may require running both side by side in a new gel. Several specialized software packages for fragment analysis are commercially available.156,157 These programs are generally similar, although they sometimes differ in analytic strategy as well as the terminology used to identify the software functions. The process of converting the patterns of fragments into suitable datasets involves multiple steps and is inherently complex; subtle, often esoteric, elements can have unexpected impact.157,158 Results obtained using computer analysis should always be confirmed by visual inspection of the restriction fragment patterns; in some instances, new gels in which selected isolates are run side by side may be needed. This discussion will be oriented toward PFGE gels, but the principles are applicable to analysis of bands generated by Southern blotting and PCR. 26

©


Volume 27 7.2.1

MM11-A

Running the Gel and Obtaining the Images

Gels intended for computerized analysis should have two independent sets of controls. •

Every fifth lane should contain a bacterial strain with precisely sized restriction fragments to serve as molecular size standards and permit computerized “normalization” to correct for inevitable distortions and inconsistencies across the gel. Lambda concatemers may be less useful for this purpose because of limited resolution of the larger fragments.

•

One lane in each gel should contain a “reproducibility standard,” i.e., replicate digests of a single independent strain, preferably of the same species as the test. Analysis of this strain across all gels in the series serves to define the level of technical variation.

A common misconception is that the molecular size standards can also serve as the reproducibility standard. This is incorrect; it is essential that two independent strains with different restriction patterns be used for these two purposes. During the process of normalization, the software explicitly adjusts the entire gel image so as to minimize, if not eliminate, differences among the molecular size standards. Determining the residual variance due to technical factors across the gels requires analysis of lanes in each gel that are subject to—not the basis of—the computerized adjustments. The level of variation reported among the reproducibility lanes across a set of gels defines the minimum difference that is likely to have biologic significance. For example, if the reproducibility lanes have a similarity of 0.94, then sets of two or more isolates with a similarity of 0.95 are almost certainly indistinguishable. In unusual circumstances, visual inspection may indicate a single distinct change. The image of the gel is entered into the computer (“acquired”) either directly using a digital camera or indirectly by scanning a conventional photograph with a flatbed document scanner. Satisfactory resolution of an 8 in. x 10 in. (20.3 cm x 25.4 cm) gel can be obtained using a camera capable of five megapixels or a 1200 dpi scan of a 4 in. x 6 in. (10.2 cm x 15.2 cm) photograph. Images should have eight-bit optical density depth (256 gray values) and be saved in a standardized format, e.g., as TIFF files (tagged image file format), which can then be imported into the analysis software as well as shared with other investigators. 7.2.2

Computerizing the Pattern of Fragments

The initial steps in the process involve considerable operator participation. The process described here represents a composite; individual steps may be automated (i.e., not under operator control) or excluded in particular computer analysis programs. 7.2.2.1

Defining Lanes

While the components of the gel are intuitive to the user, they need to be precisely defined for the software so artifacts are not misinterpreted. The first step is to define the boundaries of each individual lane (“zone of interest”) in the gel and designate it as a separate entry in the database.76 Often the software can automatically mark each lane by a rectangle, with provision for the operator to manually adjust the position, length, and width of the zone if there is any distortion present.159 Each zone should be long enough to encompass all resolved fragments and wide enough to enclose each fragment completely. Ideally, the size and shape of the zone is the same for all the lanes in the gels belonging to a single experiment, since it may influence the subsequent normalization of the lanes. Once set, the boundaries of the lanes should not be changed.

©


27

Number 10 7.2.2.2

MM11-A

Optimizing the Image

Next, the appearance of the image should be adjusted to optimize the definition of the fragments, e.g., to enhance weak bands, or to increase contrast in areas of darkness. The “tone curve editor” is a more powerful enhancement tool that acts at the level of the original tagged image file format (TIFF) image and applies to the entire gel through all subsequent steps.159 Specifically, the tone curve editor shows the distribution plot of the densitometric (grayscale) values in the TIFF file over the available range. The image can be optimized by applying a mathematical function (linear or logarithmic) to rescale that distribution. For an eight-bit image (minimum, 0; maximum, 255), a linear curve typically provides the best correction, resulting in a brighter image. The tone curve can also be adjusted by increasing or decreasing the zero level and/or contrast to enhance weak bands or sharpen dark ones. In addition, brightness and contrast settings can be adjusted for each lane individually, independent of the remainder of the gel.159 7.2.2.3

Defining Densitometric Curves

After the tone curve adjustments are completed, the program then automatically identifies the densitometric curves representing the fragments within each defined lane. However, distortion at the edges of the bands (sometimes referred to as “smiling”) should be excluded. Options available at this stage include background subtraction and filtering to optimize use of the gray scale.76 “Filtering” defines a method used to average the densitometric values across the width of the lane at a given vertical position. Arithmetic averaging is the simplest, and generally the preferred, algorithm. Filters based on median and mode represent more sophisticated methods that generally provide less overall noise reduction, but may be useful to reduce peak-like artifacts in gel, such as interfering spots.159 The optimal settings for background and filtering adjustments can be determined using the spectral analysis of the isolate patterns in a gel. The signal/noise ratio provides a measure of the overall quality of the gel and, if possible, should exceed 50. Least square averaging smoothes the profiles by removing background noise, typically seen as irregular peaks, from the profile of real (broader) peaks. The Wiener cutoff scale determines the optimal setting for the least square filtering. The background scale is an estimate of the size of the virtual disk used to scan the parts of the profiles outside the bands to determine background subtraction. Note that these parameters impact all patterns in the gel; adjustments should be applied with caution and the impact verified visually.159 They should be determined for every new set of analyses. 7.2.2.4

Normalizing the Gel

During the normalization phase, the size standards repeated throughout the gel are used to adjust the alignment of the entire gel, including the patterns of the intervening test isolates (i.e., the isolates being investigated). To minimize interpolation and avoid the problems of extrapolation, the reference pattern should be present in the lanes outside the test isolates and optimally in every fifth lane across the gel. In general, for PFGE, the outermost lane on each side is most subject to distortion and is left empty. It is also necessary for the reference pattern to span the entire range of fragments to be analyzed for the test isolates. Normalization of fragments larger or smaller than the extremes of the reference pattern will yield unpredictable results. All gels in a given analysis set should use the same size standard. After the size standard lanes are identified by the user, the fragments within the lanes are marked, either manually or automatically by the software. In some programs, each fragment within the standard is designated as a “reference position.” After the normalization algorithm is applied, any mismatches (i.e., a fragment in one reference lane that is incorrectly labeled relative to the corresponding fragments in the other reference lanes of the gel) will be readily apparent by examining the distortion bars, which indicate the local adjustments made to the gel during the alignment.159 28

©


Volume 27 7.2.2.5

MM11-A

Identifying Fragments

Identifying the fragments (“bands”) within each test isolate pattern is a critical step in computer-assisted analysis. It can be done manually by the user, automatically by the program, or by a combination of both approaches. Automatic detection is based on the band search filter. The sensitivity threshold for a potential band is set by two parameters in this filter: (a) the minimum area, which is the smallest percentage value one fragment may represent relative to the total area of all fragments in the lane; and (b) the minimum profile, which is the percentage increase in the density of the fragment relative to the surrounding background. If these threshold values are set too low, then artifacts will be incorrectly identified as bands (false positives); if the thresholds are set too high, then real bands will be ignored (false negatives). In contrast with the previous settings, these adjustments for identifying fragments directly impact the definition of an isolate profile and thereby affect the final relatedness assigned to different isolates. Consequently, although software programs provide automatic band identification, the user must always confirm the validity of the results.159 This participation is potentially subjective, particularly for profiles with partial digests, doublets, or artifacts. To minimize bias, it is best to avoid settings that require the user to both add and delete bands on a single gel. It is preferable to set the thresholds slightly low, so all real bands are identified automatically and the user is then editing out a small number of false-positive identifications. 7.2.2.6

Alternative Strategies

In the sequence described above, the size standard, which was present in multiple lanes of the gel, was used to normalize the patterns of the test isolates within and across gels. The analysis then proceeds using the normalized patterns. An alternative approach is to use the size standards to define the size of each fragment in the digests of the test isolates and then analyze the distribution of fragments using the calculated sizes. Both approaches involve similar steps and yield similar outputs and results; neither is inherently superior. 7.2.3

Cluster Analysis

Cluster analysis involves quantitating the relatedness within and between sets (“clusters”) of isolates based on their molecular strain type (e.g., RFLP pattern or sequence).76 The different clustering algorithms are all iterative processes with two basic components: •

a formula for calculating the “similarity coefficient” (SC) between any pair of existing clusters; and

•

a method for using the SC to determine which existing clusters should be combined to define the next higher level of clusters.

At the start of the process, each isolate in the analysis represents a separate cluster. At the conclusion, the entire analysis set has been ordered into a sequential (hierarchical) union of clusters, typically depicted as a dendrogram labeled with the similarity value associated with each union. The user chooses the method for calculating the SC and forming the clusters. The program then performs the analysis automatically and presents the result. Based on examination of the results for the reproducibility standard, the user may want to adjust the “optimization” and “tolerance” settings for comparing individual genotypes and/or to exclude certain gels as described below (see Section 7.2.4). After any such adjustment, the cluster analysis must be repeated.

©


29

Number 10 7.2.3.1

MM11-A

Similarity Coefficients (SCs)

A similarity coefficient (SC) is a quantitative description of the relatedness between two genotypes; the values are normalized to range between 1 (indistinguishable or identical) and 0 (no relatedness at all). For gel electrophoresis patterns, the SCs are based on comparison of the densitometric curves representing the fragments in each lane (i.e., for each isolate). In the preferred approach for analyzing PFGE profiles, each fragment in a pattern has a designated position (or size); in comparing two patterns, each fragment is treated as a binary data element that is either present (value, 1) or absent (value, 0). The SC is then calculated using the algorithms of Dice or Jaccard.46 The Dice coefficient comparing two patterns is calculated by multiplying the number of matching (“common”) fragments (Ncommon) by two and dividing by the sum of the number of fragments in each pattern (N1, N2). (When this formula is applied to RFLP data, it is sometimes referred to as the coefficient of Nei and Li.46) Dice =

2 x Ncommon (N1 + N2)

The Jaccard coefficient is calculated as the number of matching bands (Ncommon) divided by the total number of bands (Ntotal) defined by the two patterns. Note that Ntotal is equivalent to [(N1 + N2) - Ncommon]. Jaccard =

Ncommon Ntotal

The Dice coefficient is the most reliable method for calculating similarity between two fragment patterns and is the preferred approach for analyzing PFGE data.52 There is no difference between the Dice and the Jaccard methods with regard to the order of clustering in the dendrogram. Since Dice gives more weight to matching bands, whereas Jaccard gives more weight to differences, the SC values obtained using the Dice formula are greater than those with the Jaccard method.76 In comparing the PFGE profiles of two isolates recently derived from a common precursor, variations associated with single genetic events (e.g., insertions, deletions, or gain or loss of a single restriction site) result in differences of ≤3 bands.34 Thus, for a typical PFGE profile comprising 15 fragments, the gain (or loss) of a restriction site results in a three-band difference and a Dice coefficient of 0.90, calculated as [2 • 14/(15+16)]. Consequently, in PFGE analyses a pair of isolates whose Dice coefficient is ≥0.90 are likely to be genetically related. Nevertheless, in studies of large numbers of isolates, such putative matches typically need to be confirmed by direct side-by-side comparisons. In an alternative approach, the similarity between two PFGE profiles is expressed as Pearson’s correlation coefficient calculated using the entire densitometric curve, including such factors as the intensity of band staining, band shape, and gel background. Because larger DNA fragments bind more ethidium bromide stain than smaller fragments, this method is inherently biased when applied to PFGE patterns; it is also more influenced by gel artifacts.46 7.2.3.2

Clustering Algorithms

The first step in proceeding from a set of SCs representing the pairwise comparison of all the isolates to the final dendrogram is to arrange the SCs in a similarity matrix. For a set of N isolates, the similarity matrix, D, is an N x N array, where the element Di,j represents the SC of the patterns for isolates i and j. The sequence of the isolates along each axis of the array should be the same, although the actual order is arbitrary and is generally set automatically by the program.76 By definition, each isolate is identical to itself; consequently, the elements Dii on the main diagonal of the matrix are all 1 (i.e., 100% similarity). For each of the other elements of the matrix, the SC value is a number between 0 (no similarity at all) and 1. 30

©


Volume 27

MM11-A

Each stepwise application of the clustering algorithm groups elements of the similarity matrix. At the first step, each element of the matrix is an individual isolate, but as the process progresses, isolates are replaced by groups (clusters) of isolates, and subsequently by clusters of clusters. Thus, after each clustering step, the dimension of the matrix decreases and, theoretically, the clustering is a finite process.76 Several different clustering methods are available. The “unweighted pair group method using arithmetic averages” (UPGMA) is the most neutral algorithm for clustering patterns of DNA fragments and the one used most commonly. This method finds the matrix element (Dij) with the greatest SC, merges the two components (i and j) of that element into a single newly defined cluster, and then recalculates the similarity matrix by comparing that new entry with all the remaining entries. Matrix elements whose components include a cluster are calculated as the arithmetic average of the pairwise SCs of the individual isolates comprising the cluster. The process is then repeated using the new, smaller matrix. It is important to appreciate that the UPGMA method can generate dendrograms that are different in structure but equally valid. When two matrix elements have the same best value, the software acts on the first element encountered. Since the original order of the isolates was arbitrary, the same isolates arranged in a different initial order could give a different dendrogram. The algorithm has no basis for distinguishing among these alternative solutions. In general, these differences are minimal for UPGMA analysis of PFGE patterns of closely related isolates and, as discussed below (see Section 7.3), there are independent factors that confound the quantitative values assigned to more distant relationships. The Ward method defines successive clusters to minimize the increase in the total within-group variance. This method often results in a cluster of aberrant isolates that in fact have nothing in common with each other except that each is quite dissimilar to the other isolates in the data set.160 The neighbor-joining method represents the difference between isolates (or clusters) as a horizontal branch; more distantly related elements are connected by longer branches. Clusters are defined by minimizing intergroup branch length. The resulting dendrograms facilitate visualization of hierarchical groups of isolates and are particularly useful in analyzing nucleotide sequence data. When applied to PFGE and other data sets based on restriction polymorphisms, they should not be inferred to depict phylogenetic structure46 (see Section 7.3). 7.2.4

Using the Reproducibility Replicates to Assess the Cluster Analysis

The dendrogram generated by the above process should be assessed for the distribution of the reproducibility isolate whose profile was included in every gel. Even with the most careful technique and the application of the reference size standards to align patterns within and among gels, there is inevitably variation among the different replicates of the reproducibility isolate. The SC for the reproducibility replicates quantitates the technical reproducibility across the set of gels. As discussed above (see Section 5.1), the level of reproducibility constrains the level of discriminatory power. Consequently, the lower the SC, the poorer the discriminatory power. Ideally, all the reproducibility replicates will be assigned to a single well-defined cluster with SC >0.95. If the cluster has an SC 95% identity within a defined 160 bp portion of the 5′ region of emm.129,165,166 Since structural genes are rarely subject to premature truncation due to frameshifts or nonsense mutations, sequence data for typing systems using such loci should be confirmed using translation and mapping programs. For example, ©


33

Number 10

MM11-A

subtypes within the emm typing system are defined by any base changes within the type-specific 50 codons; base substitutions in this region almost always result in amino acid changes.167,168 Percent identity between two sequences can be determined using BLAST or FASTA; the latter is easier to use and is preferred for searching large local databases, but is slower. BLAST together with current protein and DNA sequence databases are available online through the National Center for Biotechnology Information (NCBI) (http://www.ncbi.nlm.nih.gov/BLAST/).

8.2

Pattern Recognition

For typing methods based on regions with variable repeats, the type is often designated as simply the pattern and/or number of repeats observed. In spa typing, a single base change in one of the 24 bp repeat sequences or the insertion/deletion of one of 2 to 16 repeats is informative.37 In basic epidemiologic analyses, strain types can be represented by a single number; assigning isolates to phylogenetic lineages in population genetics studies requires the application of a global multiple sequence alignment program that considers the presence of identical repeats representing shared exact nucleotide polymorphisms.37

8.3

BURST

BURST (Based Upon Related Sequence Types) is a novel clustering algorithm for analyzing microbial MLST data (http://outbreak.ceid.ox.ac.uk/software.shtml); an enhanced version (eBURST) is available online (http://eburst.mlst.net/1.asp). In MLST, isolates that have the same sequences at the multiple loci examined (typically seven) are referred to as clonal. A clonal complex comprises genetically related isolates that differ at only one or two loci (termed single and double locus variants, respectively).21 The primary founder of a complex is defined as the ST that differs from the largest number of other STs at only a single locus (i.e., the ST that has the greatest number of single locus variants). Frequently for biologically important and prevalent complexes, such variants represent alleles that differ at a single base. BURST is the preferred algorithm for determining relationships among MLST profiles, because it uses a simple but appropriate model of bacterial evolution in which an ancestral (or founding) genotype diversifies to produce a cluster of closely related descendant genotypes.169 Such evolutionary relationships are poorly represented by UPGMA-generated trees (dendrograms). In further contrast to UPGMA, BURST ignores most of the pair-wise relationships among the genotypes in the population and does not make inferences about the relationships among the distantly related genotypes. Even clonal bacterial populations may have numerous recombination events (horizontal transfer) among distantly related genotypes, and consequently, those relationships may be more appropriately described by a network (as for a panmictic species) than by a branching tree (dendrogram). BURST analysis of MLST data for a large collection of methicillin-susceptible and -resistant S. aureus resolved the intricate evolutionary histories of MRSA.21 In a large population-based surveillance study of S. pneumoniae, this approach was useful in examining and depicting the genetic structure within each serotype.170

8.4

Sequence Analysis Methods for Evolutionary Genetics

Several additional analytic approaches have been developed over the past few years for using nucleotide sequences to define evolutionary relationships. However, in the context of molecular strain typing, these typically offer little, if any, advantage and will be considered only briefly. 8.4.1

Sequence Alignment

This approach emphasizes the functional and evolutionary consequences of sequence variation, but can also be adapted for epidemiologic studies. Sequence alignment begins with aligning the encoded amino 34

©


Volume 27

MM11-A

acid sequences; deletions, insertions, and other events may require manual adjustment. The analysis then proceeds to assess variance in the underlying nucleotide sequence. 8.4.2

Neighbor-Joining Method

This method also requires aligned sequences, but does not require that all lineages have diverged equally (i.e., that the data be ultrametric); it is therefore particularly suited for data sets comprising lineages with large variation in rates of evolution171 (http://www.icp.ucl.ac.be/~opperd/private/neighbor.html). (Other commonly used algorithms, such as UPGMA, are not appropriate for such data.) The neighbor-joining method is favored by molecular evolutionists based on several features. It is fast and thus suitable for deducing a single dendrogram from large, diverse data sets (e.g., ribosomal sequences); it provides for bootstrap analyses; and it permits correction for multiple substitutions at individual bases in the sequence. However, the neighbor-joining method also has some inherent limitations; specifically, sequence information is reduced by the analysis and the method yields only a single possible tree. 8.4.3

Maximum Parsimony Method

In maximum parsimony analysis, the phylogenetic (cladistic) relationships among isolates are inferred by an algorithm that minimizes the total number of evolutionary steps required to explain the variances within a data set (http://www.icp.ucl.ac.be/~opperd/private/parsimony.html). The method is based on shared vs. derived “characters” and is applicable to nucleotide and amino acid sequences as well as other data. In focusing on evolution, maximum parsimony evaluates different alternative trees and may provide insights into ancestral sequences. Limitations of the method are that only informative (variant) nucleotide positions are considered, not all the sequence data; consequently, only branch points (nodes) are determined, not branch lengths. In addition, the method is slower than the distance methods discussed above and is poorly scalable for larger data sets. 8.4.4

Maximum Likelihood Method

Maximum likelihood also assumes a phylogenetic approach. Given a specific hypothesis for an evolutionary history, it seeks to determine the most likely relationships that would explain the available sequence data. The implicit assumption is that evolutionary events would take the most likely, i.e., most efficient, course (http://www.icp.ucl.ac.be/~opperd/private/max_likeli.html). The maximum likelihood uses all available sequence data, evaluates multiple tree topologies, and is very statistically robust. However, the results are dependent on the evolutionary model chosen and require extensive computations; consequently, the method is slow.

9

Interpreting Variation in Molecular Typing

The technical aspects of performing molecular strain typing and handling the data have been refined over the past decade. However, the challenge of interpreting the results has proven less tractable—particularly interpreting the significance of subtle variations among genotypes. At least some part of the difficulty is inherent in acquiring data that reflect natural, stochastic processes and attempting to provide categorical responses to relatively artificial questions (i.e., “Are these isolates part of an outbreak?”). The approach to interpreting and reporting typing results as described here and in the next section, respectively, is most appropriate for sets of isolates submitted by clinicians and infection control personnel. Such isolate sets typically share the following features: ©


35

Number 10

MM11-A

•

There is a specific question or hypothesis. “Do these isolates represent an outbreak?” “Is this a relapse or new infection?” Note that these questions reflect a common paradigm: Are two isolates the “same” or “different”?

•

The set typically comprises a limited number of isolates. Evaluating a patient with successive episodes of infection may involve as few as two isolates; but even investigations of acute nosocomial outbreaks will rarely require the analysis of more than 30 isolates.

Molecular strain typing is applied in many other contexts, such as: •

public health investigations of water- and food-borne outbreaks, which may involve 100 or more isolates;

•

large-scale surveillance studies; and

•

research studies of bacterial population genetics.

The concepts described here remain relevant, but these large data sets involve additional questions and analytic methods that are beyond the scope of this document.

9.1

Categories of Genotypic Relatedness in Molecular Strain Typing

As emphasized above: •

a strain is defined operationally in the context of a particular technique (see Section 3.1); and

•

the performance characteristics of a typing system are influenced both by technical features of the method and by the genetic diversity of the species being examined (see Section 5.3).

Consequently, a relevant and accurate assessment of the relatedness of two genotypes requires quantitative information of the performance characteristics of the typing method for that species. In that context, a pair of genotypes can be assigned to one of three mutually exclusive categories of relatedness. 9.1.1

Indistinguishable

Two genotypes are defined as “indistinguishable” if the typing system demonstrates no variation or difference between them. While two sequences that have no differences could be described as “identical,” that term is not appropriate for electrophoretic profiles and is therefore not generally applicable. If electrophoretic gels are assessed by visual inspection, then indistinguishable is a subjective judgment. When an image-analysis system is applied, “indistinguishable” is defined by the level of variation observed among the reproducibility replicates; smaller differences are consistent with technical variation and, in general, cannot be considered biologically meaningful. By definition, isolates with indistinguishable genotypes represent the same strain. 9.1.2

Different

Two genotypes are defined as “different” if the variation between them is inconsistent with clinical or epidemiologic relatedness. That threshold represents the biologic reproducibility of the typing system as established by the variation observed among a set of independent isolates that, based on a priori 36

©


Volume 27

MM11-A

epidemiologic criteria, should represent a single strain. Such data are a finite sample; predictably, the range of variation observed will be greater for more extensively evaluated typing systems. The paradoxical consequence is that an established system may appear to be less discriminatory and to have a more stringent definition of “different” than a newly developed system. Both producers and consumers of strain typing data should appreciate that the overall reproducibility of a typing system—including technical and biologic variation—is arguably the single most important characteristic in interpreting results, but is often inadequately assessed. 9.1.3

Similar

This category is defined by exclusion; two genotypes are “similar” if they differ (i.e., are distinguishable), but not sufficiently to be classified as “different.” Thus, “similar” is the gray zone in the middle between the other two categories.

9.2

Step One: Identify the “Reference Isolate” or Type That Focuses the Question

The first step in interpreting molecular strain typing data is to identify the reference or index isolate that is the focus of the question. Ideally, those submitting the isolates for analysis will designate the index isolate. In epidemiologic investigations, this is often the first isolate in the putative outbreak. For multiple isolates from an individual patient, the earliest isolate from a sterile site is likely the most reliable reference point. In the absence of any specific information, the laboratory may wait until the strain typing results are available and then identify the modal (most common) strain type and designate the first isolate of that type as the reference isolate.

9.3

Step Two: Compare Each Isolate to the Reference Isolate

Once the reference isolate is identified, the genotype of each test isolate is compared to the genotype of the reference, visually or by computer analysis (see Section 7), and the relatedness of the two isolates is categorized as defined above.

9.4

Translating Genotypic Relatedness Into Epidemiologic and Clinical Relatedness

The last step in the process of interpreting molecular strain typing data is to describe the epidemiologic and clinical implications of the categories of genotypic relatedness (see Table 2). Table 2. Interpretation of Typing Results Within Different Contexts Context Category of Genotypic Relatedness Indistinguishable Similar Different Indicates isolates Clinical – multiple Consistent with a Consistent with represent >1 infecting isolates from one single (monoclonal) variation during a strain individual infection single (monoclonal) infection Epidemiologic – multiple isolates representing putative outbreak

©

Consistent with epidemiologically related isolates representing an outbreak strain

Consistent with variation during an outbreak; additional microbiologic and epidemiologic correlation required


Indicates isolates represent epidemiologically unrelated strains

37

Number 10

MM11-A

Several aspects of this approach deserve emphasis. •

All these interpretations represent probabilities rather than absolutes. They reflect estimates based on the experience with the typing method, but since even unlikely events will occur, the estimates may be incorrect on occasion.

•

“Different” is inherently more reliable than the other categories.

•

The interpretations at this level are based entirely on the typing results. The final conclusions regarding the primary clinical or epidemiologic question must integrate the typing data with other available information. If additional strain typing is performed, a completely independent method should be considered (e.g., adding sequence data to a PFGE analysis is much more likely to be useful than repeating the PFGE with a different enzyme).

9.5

Comparison to the “Tenover Criteria”

In 1995, Tenover et al published an approach to interpreting PFGE data for molecular strain typing.3 As in the method discussed above, a reference isolate (e.g., a putative outbreak strain) and an investigational (“unknown”) isolate were postulated. The algorithm linked the number of observed fragment differences between the two profiles with the number of underlying genetic changes and the likelihood of epidemiologic relatedness. Four categories were defined: in addition to “indistinguishable” and “different,” there were “probably related” and “possibly related,” essentially two subsets within “similar.” The approach of starting from the number of deduced genetic changes had the virtue of connecting the final interpretation to the underlying molecular mechanisms. However, it imposed a step that, in practice, was often difficult as well as imprecise. Consequently, in many instances, interpretation proceeded based on the number of fragment differences, which, unfortunately, has no simple linear relationship to the number of molecular events or the likelihood of epidemiologic relatedness. Over time, other limitations of this approach have become apparent. Neither the scientific validity nor the practical utility of defining two subcategories of “similar” has ever been established. The approach was explicitly geared to PFGE profiles and offered little guidance for interpreting other strain typing data. Even within PFGE studies, the approach assumed that a single algorithm was applicable to all bacterial species. The approach described here has the virtue of being applicable to different typing methods and sensitive to differences between species. However, more often than not, robust data for defining the range and boundaries of “similar” are not currently available. Hopefully, the value of the proposed approach will encourage studies to develop the required data.

10 Reporting Molecular Typing Results The molecular strain typing report represents a critical item of communication between the epidemiologist or clinician who requested the typing (the consumer) and the laboratory performing the work (the provider). It should be comprehensive, self-contained, and readable. The guidelines presented here are most applicable to a clinical service laboratory, although the concepts remain relevant in other contexts. Ideally, the report should include the following components: •

A statement of the question being addressed. This is essential if the report is to draw conclusions that are both practical and relevant.

•

A summary of the relevant information for each isolate. This would include the identifying number assigned by the source laboratory; the identifying number assigned by the typing laboratory;

38

©


Volume 27

MM11-A

collection date of the specimen; type of specimen (e.g., blood, urine, food); source of specimen (patient, environment, fomite); and location of source (e.g., hospital ward, food serving). •

A comprehensive description of the methods employed is essential to providing the reader an objective basis for understanding the interpretation. The following materials should be presented as an appendix to the report or made available online: –

–

The technique used. This can be brief if a published protocol can be cited. The performance characteristics of the test, i.e., the discriminatory power and the reproducibility.

These may be drawn from a publication or from the laboratories’ own studies and, if possible, should relate to the species of the isolates being examined.

–

The criteria for the interpretative categories.

•

The primary data; for example, a copy of the image of the gel or a table of the relevant nucleotide sequences. A graphical representation of the data (e.g., a “bar-code” drawing of a gel or a dendrogram) may be added, but should not substitute for the primary data.

•

A summary of the typing results for the individual isolates, including the strain type assigned to each isolate and the interpretation of the relatedness among the isolates. This may be integrated with the summary of isolate information or presented separately.

•

An overall interpretation of the data set that directly addresses the primary question; for example, which isolates represent an outbreak strain or whether a patient has had reinfection with a new strain or possible relapsing infection with the same strain. The conclusions should explicitly indicate where the results of strain typing provide probabilistic rather than absolute insights.

The report formally captures the actual typing data and the most reasonable interpretation applicable to the question posed. It should be prepared independently to maximize the contribution of the typing studies in addressing the overall clinical or epidemiologic issue, Nevertheless, strain typing, like any single laboratory test, may not be definitive. Resolving the primary question may require independent information from the field, additional isolates, supplementary laboratory studies, and even repeat typing to address the possibility of error. Further discussion between the laboratory and the clinician or epidemiologist is often essential and always desirable.

11 General Technical Issues 11.1 Identifying Isolates A clear, concise system for uniquely identifying every isolate is required, together with a log for recording the date, source (hospital, laboratory, patient, specimen, subculture, etc.), and other pertinent information. Routine clinical specimen numbers are likely to be insufficient for a typing laboratory, since the isolate identification system needs to include provision for processing multiple isolates from a single specimen or multiple different subcultures derived from the same colony. A mix-up among isolates is a potential and relatively common cause of disagreement between the typing results and the clinical epidemiology; such problems may also become apparent as logical inconsistencies between the results of two independent typing methods.11 Therefore, the isolate register should unambiguously identify repeated, independent samples representing the same initial organism. The isolate register should also permit consistent identification of organisms provided by other institutions, either for primary analysis or for comparison.

©


39

Number 10

MM11-A

11.2 Archiving Isolates—Freezing In many situations, it is useful, even essential, to archive isolates for future reference, for example, isolates collected pre- and post-therapy from patients participating in clinical trials. The recommended procedure for preserving isolates is to prepare a dense suspension in Brucella broth supplemented with 20% glycerol, aliquot the suspension to cryovials, and freeze the vials at -80 oC. Retrieval of stored material is greatly facilitated by a well-maintained directory, either electronic or manual, of the specific freezer location of each isolate. Monitoring alarms and back-up power or CO2 systems are highly desirable.

12 Examples of Molecular Typing of Bacterial Species 12.1 Streptococcus pyogenes S. pyogenes (Group A streptococci [GAS]) are common clinical pathogens that are responsible for millions of superficial infections of the upper respiratory tract or the skin annually. These sites serve as principal reservoirs for rarer, but potentially lethal, invasive disease. GAS also cause postinfective sequelae, most notably rheumatic fever and acute glomerulonephritis. 12.1.1 Phenotypic Strain Typing Based on M Protein Serotype The Lancefield M virulence protein serotyping scheme, established by Dr. Rebecca Lancefield during 1928 to 1966,172 was the gold standard for strain typing this complex bacterial pathogen for more than 60 years. By the 1980s, there were more than 80 established and provisional M serotypes, although recent investigation indicates that a few of these designations were actually incorrectly assigned in the postLancefield era on the basis of the antiopacity factor reaction (dictated by hypervariable sof gene product).165 Several factors make identification of M protein of particular importance in studies of GAS. M protein is a critical virulence factor of this organism, and M serotypes are nonrandomly represented among diseasecausing strains.141,173-176 Promising multivalent vaccines based on M protein are in development, and rigorous evaluation of their efficacy will require accurate assessments of the M protein types among infecting strains.177-179 Further, decades of epidemiologic studies have been based on M serotyping and could be disconnected from current research by implementation of another typing system. However, the classical Lancefield methodology has significant technical limitations. These include the effort required to prepare and distribute M typing sera for the numerous individual serotypes; variation in expression of M protein associated with growth on laboratory media; and the subjectivity inherent in performing and interpreting serotyping reactions. One particularly undesirable consequence of these factors was that epidemiologic studies of GAS were limited by the capacity of the few international reference laboratories that could perform comprehensive serotyping of GAS isolates. 12.1.2 Genotypic Strain Typing Based on Sequencing Genes Encoding for M Protein GAS strain typing has been transformed by the demonstration that the M serotype of an isolate can be deduced from the sequence of the emm gene. The use of emm sequence types has freed investigators from the technical constraints of serotyping without sacrificing its historical and pathogenic perspective. Each validated serotype (designated sequentially from M1) correlates with a specific sequence type (correspondingly designated from emm1).180,181 The one-to-one relationship between M protein serotype and emm gene sequence type has been validated using the classical Lancefield M serotype reference strains (identified between 1928 and 1966), later M serotype reference strains (identified 1967 to 1999), as well as current clinical isolates.165,180-183 40

©


Volume 27

MM11-A

emm genotypes were initially defined based on 160 base pairs of the 5′ portion of the gene, which encode the 21 to 23 leader peptide residues plus the N terminal 30 amino acids of the processed M protein. A new definition is based solely on the variable coding region of the processed N terminus (http://www.cdc.gov/ncidod/biotech/strep/assigning.htm). Each validated emm type is represented by multiple, independent clinical isolates recovered from serious infections. Isolates representing an individual emm type generally demonstrate little sequence variation, but show at least 5% sequence divergence (and usually >15%) from the next closest matching reference strain emm sequence within the type-specific region. The Lancefield classification scheme has been successfully extended based on sequence studies. Currently, there are 113 validated M serotypes, >180 emm sequence types, and >800 emm specific subtype sequences.128,182,183 The proposed (provisional) sequence types are designated using the prefix st rather than emm, pending international validation.182,183 The complete set of emm sequences is available online (http://www.cdc.gov/ncidod/biotech/strep/emmtypes.htm). Although the emm typing system is most advanced for GAS, a large percentage of clinical isolates of group G and C beta-hemolytic streptococci can be typed using the exact same protocols and definitions. There are currently 45 distinct emm sequence types identified among these organisms [http://www.cdc.gov/ncidod/biotech/emmtypes.htm], and population-based surveillance for invasive isolates using emm typing is ongoing.184,185 12.1.3 emm Patterns Based on the chromosomal arrangement of the emm genes, there are five major emm patterns (designated A through E). The emm pattern A-C strains are typically isolated from patients with pharyngitis; pattern D strains are predominantly from impetigo lesions; pattern E strains are commonly associated with both sites of infection. With rare exception, each emm type is associated with a single pattern class.186 Overall, 17% of 156 emm types represented pattern A-C; 39%, pattern D; and 41%, pattern E. 12.1.4 MLST, emm Types, and the Population Genetics of Group A Streptococcus MLST of GAS is based on partial (405 to 498 bp) sequences of seven metabolic (“housekeeping”) genes187; the database is available online (http://www.mlst.net). Recent analyses indicate 36 to 66 alleles for each of the seven target gene sequences and >300 multilocus allelic profiles or sequence types (ST).186 Genetic variation at these loci is due to interstrain recombination at least 1.4 times more frequently than to point mutation. Overall, >95% of the STs are associated with only a single emm type, and the distribution of STs by pattern class is congruent with the distribution of emm types, with 18%, 36%, and 47% of STs representing patterns A-C, D, and E, respectively. The 12 emm-variable STs appear due to recent recombinational replacements and represent 30 (19%) of the 158 emm types in the collection.186 Of note, the different emm types within each of these STs almost always represents the same pattern class; multiple emm types of pattern A-C are found in three STs, of pattern D in eight STs, and of pattern E in only one ST. In contrast, recombinational events among the housekeeping loci are much more common among pattern D and E strains than among pattern A-C strains. These findings suggest that within GAS, the tissue-defined subpopulations may differ in their population genetic structure, perhaps reflecting different host immune selection pressures. 12.1.5 Implications of the Range of Variation Among emm Sequences Comparison of emm sequences reveals a remarkable mix of diversity and consistency. Relative to the limited allelic variation within the individual metabolic genes used for MLST, the N terminal sequences of the emm genes are truly hypervariable. Yet, within defined emm types, the extent of allowed variation approaches the low variation observed within MLST loci. This suggests there are undefined M protein ©


41

Number 10

MM11-A

functions that have strict structural requirements, also undefined despite identification of more than 160 emm types. Remarkably, silent base substitutions within the sequence encoding processed N terminal 50 residues are almost never observed, suggesting a high degree of selective pressure. For example, among 1064 consecutively emm typed invasive isolates representing 41 types and 80 subtypes, not one instance of a silent substitution within the 50 codon subtype determining region was found.168 Among 1975 pharyngitis isolates, only one subtype (represented four times within a specific surveillance area) exhibiting a silent substitution was found.167 During the initial period application of emm gene sequencing, new types (defined as < 95% identity with the relevant N terminal sequences of the existing emm types) were discovered with some regularity. However, over the past several years the rate has drastically decreased despite expanded worldwide molecular surveillance of thousands of GAS isolates by multiple investigators. This observation suggests that emm sequence types represent a finite set, rather than an unlimited “continuum” and is also consistent with critical structural requirements for M protein function. The majority of isolates causing invasive disease and uncomplicated pharyngitis in the United States represent a relatively limited number of GAS clones.167 This association has been shown to be conserved between strains isolated more than 60 years apart.128 This assumption does not always hold true at the global level where many emm types can be shown to be shared between widely diverging clones.188 12.1.6 Molecular Strain Typing of Group A Streptococcus Based on Sequencing emm and Other Loci Molecular strain typing based on emm sequencing has been used to identify isolates representing an outbreak strain and to resolve unrelated strains of GAS during surveillance studies.141,165,189-197 emm typing protocol – A standardized PCR primer set has been defined that anneals to the emm genes of all S. pyogenes strains. The resulting amplicon includes the required 150 base pairs encoding the processed type/subtype-determining region and a portion of the leader peptide. A detailed protocol is available online (http://www.cdc.gov/ncidod/biotech/strep/protocols.htm). emm sequence database – A database comprising trimmed type-specific sequences of 150 bases encoding the processed N terminus of the M protein is now available online (ftp://ftp.cdc.gov/pub/infectious_diseases/biotech/emmsequ/). This database also includes more than 800 emm subtypes of S. pyogenes, which allow additional specificity in typing, as well as more than 60 emm gene types and subtypes from groups C and G beta-hemolytic streptococci (Streptococcus dysgalactiae subspecies equisimilis).198 Isolates of this species carry homologs of GAS genes, including emm,199 sof, 200 major hemolysin gene, 201 and pyrogenic exotoxin genes. 185,202 Users who encounter subtypes not in the database are encouraged to submit sequence traces and strain information to the curator. emm types – An isolate has been assigned an emm type if there is > 95% nucleotide sequence identity within bases 1 to 160 compared to the CDC emm sequence type reference strains. Recently, a simpler, more objective definition has been adopted that avoids the subjective problem of identifying base 1 and eliminates consideration of 60 to 70 bases of conserved leader peptide sequence. New types are identified on the basis of sharing less than 92% sequence identity over the first 90 bases encoding the deduced processed M protein of the type reference strain. As mentioned previously, a single interruption of the reference sequence reading frame (through frameshift, in frame deletion or insertion) by no more than seven codons is tolerated, although each interrupting codon is assigned a penalty of 0.5%. This new definition has almost no effect upon previously designated emm type. Subtypes are still assigned on the basis of any changes within the coding region of the processed N terminus.

42

©


Volume 27

MM11-A

Sequence searches should be performed online using the CDC subtype-specific database (http://www.cdc.gov/ncidod/biotech/strep/strepblast.htm) or, when perfect matches for 150 bases are not found, using the BLAST function of GenBank (http://www.ncbi.nlm.nih.gov/BLAST/). Longer query sequences extending into the conserved 3′ portions of the emm gene may give artificially high scores when screening the untrimmed emm database. emm subtypes – Subtypes within an emm type are defined by any DNA sequence changes within the coding region for mature residues 1 to 50203 (http://www.cdc.gov/ncidod/biotech/strep/strepblast.htm). Variations within this sequence generally represent single missense substitutions or small indels and are likely to be stable markers (http://www.cdc.gov/ncidod/biotech/strep/800emms.htm). To assist in determining the coding sequence for the proposed N-terminus, a program for determining the predicted leader peptide cleavage site for gram-positive, gram-negative, or eukaryotic-derived peptides is provided online (http://www.cbs.dtu.dk/services/SignalP/). The repertoire of M signal cleavage sites is limited, and the sites are usually easily recognized. For a small number of emm types (e.g., emm5 and emm6), display analysis of the first 50 codons encoding the mature protein indicates an unusually high frequency and number of different subtypes. These variations largely reflect different patterns of excision of tandem repeats that overlap this hypervariable region. As a result of this instability, some subtypes of emm5 and emm6 may not serve as reliable subtype markers. Similarly, sequence alterations downstream of the subtype-determining region within types emm5, emm6, and other types are often caused by different patterns of tandem repeats. Nevertheless, for many emm types, single base differences downstream of the subtype-determining region define reliable subtypes and resolve epidemiologically unrelated isolates. Sequence subtypes may be particularly useful in providing an additional discriminatory power among isolates representing the common emm types, which account for a large proportion of isolates. For example, the isolate from a fatal group A streptococcal infection in an allograft recipient was the very common type emm3, but an extremely rare subtype. The same rare subtype was isolated from the tissue donor, further substantiating the epidemiologic link.204 sof genes – The serum opacity factor (sof) gene, which is present in about half of all clinical isolates, encodes a highly variable extracellular matrix-binding protein expressed on the cell surface that also functions to opacify mammalian serum. These proteins can be serotyped using specific antisera that abolish the serum opacity factor reaction. Historically, this system was used as a surrogate marker to assign M types. However, it is now appreciated that this association is unreliable. Just as the same emm type can be found among different MLST lineages, isolates of the same emm type may also have different sof genes.165 Issues regarding common emm types – Some emm types are endemic and widespread, limiting the utility of emm typing alone. Among 3424 invasive isolates obtained during population-based surveillance of multiple states in the United States from 1995 to 2001, over 92% represented only 28 of the >160 known emm types (http://www.cdc.gov/ncidod/biotech/images/emmdistr.gif). The four most common types (emm1, emm3, emm12, and emm28) each caused 7 to 20% of the cases and collectively accounted for a total of 35%. Obviously, during an outbreak investigation, isolates representing these prevalent genotypes cannot be reliably inferred to be epidemiologically related, and additional genotypic data are required. emm1 – emm1 represents a globally disseminated strain that in a given year typically constitutes up to 20% of GAS isolates recovered from normally sterile sites in the United States, and an approximate equal number of noninvasive isolates associated with pediatric pharyngitis.167 These isolates are sof-negative; about 98% display no sequence variation within the emm1 gene sequence, and most have the same MLST genotype.187 Other molecular strain typing methods (including PFGE, ribotyping, RAPD) seldom distinguish among type emm1 isolates. However, the hypervariable sic gene provides an additional ©


43

Number 10

MM11-A

discriminating target to resolve epidemiologically related isolates.194,205-209 A set of primers for amplifying a suitable sequencing template has been described.209 NOTE: The designations emm1-1, emm1-2, and emm1-4 in the CDC database (originally designated as emm1 subtypes emm1.1, emm1.2, and emm1.4) are not emm1 subtypes as defined in this document. In addition, they do not represent the same MLST genotype as the common global type emm1 strain. These current designations for these emm sequences are an attempt to maintain historical continuity as well as conform to current type and subtype nomenclature used for other emm types. emm3 – emm3 is a common sof-negative lineage that, in the United States, represents approximately 9% of sterile-site isolates recovered, and a similar proportion of noninvasive isolates associated with pediatric pharyngitis.167 Among these isolates, 75% carry the emm3.1 allele, 19% carry emm3.4, and the remainder carry miscellaneous other alleles characterized by differences within mature residues 1 to 50. The emm3.1 and emm3.4 alleles differ by single base substitutions that appear to be stable subtype markers.167,168 Most type emm3 isolates represent a single MLST genotype.187 emm12 – Type emm12 typically comprises 7 to 9% of US sterile-site isolates and up to 18% of noninvasive isolates associated with pediatric pharyngitis.167 Type emm12 isolates show very little variation in emm sequence, or in MLST genotype.187 All appear to carry a highly conserved, defective sof gene.210 emm28 – Type emm28 is associated with a common sof-positive isolate that displays little variation in emm or sof gene sequences165 or in MLST.187 emm28 was the most common type associated with cases of invasive postpartum GAS infection in the United States during 1995 to 2000,211 and accounts for about 10% of isolates associated with pediatric pharyngitis.167 12.1.7 PFGE PFGE has proven to be an effective method for molecular strain typing of GAS isolates165,192-194,212-222; detailed protocols have been published.193,214 PFGE typing is consistent with, but more discriminatory than, emm sequence typing. That is, isolates with the same PFGE genotype also have the same clonal type and emm sequence type; but PFGE is sometimes able to distinguish among epidemiologically unrelated isolates of the same emm type.191 For example, the PFGE patterns of several CDC M type reference strains originally isolated decades ago, including those for the common types M1, M4, and M12, are unambiguously distinct (seven or more band differences) from recent clinical isolates that have the same emm type and same MLST profile (unpublished data). PFGE generally does not provide more discrimination than sequence analysis using multiple loci as discussed above (see Section 6.4). 12.1.8 Applying Molecular Strain Typing Methods for GAS The technical advantages and limitations of the individual typing methods—MLST, emm typing (an example of sequencing a hypervariable locus), and PFGE—have been considered in detail in Section 6. The application of these methods to GAS offers a useful perspective on their effectiveness in different context. 12.1.8.1 Outbreaks PFGE resolves the greatest number of distinct genotypes and thus is the most discriminatory technique. Nevertheless, epidemiologically unrelated isolates may have indistinguishable or closely related PFGE profiles. This reflects the strongly clonal population structure of GAS, with the most common lineages (e.g., organisms of serotype M1 and sequence type emm1) being broadly disseminated and yet displaying little or no genotypic variation. This characteristic represents an inherent limitation for all GAS strain typing. 44

©


Volume 27

MM11-A

In many analyses, emm typing has proven almost as discriminatory as PFGE; resolution can be improved by determining subtypes based on downstream nucleotide variation and by sequencing additional hypervariable loci, e.g., sic (present in type emm1 isolates) or sof (present in ~50% of GAS isolates). As previously emphasized, emm typing identifies the M protein type, and thus provides additional information relevant to epidemiology, virulence, and clinical outcome. Further, sequence data are more portable, more precise, and can be analyzed more objectively than RFLP data. In investigating acute outbreaks, MLST offers little, if any, advantage over typing emm and other hypervariable loci. As noted, the two approaches are highly congruent and MLST is generally not more discriminatory.187,191 12.1.8.2 Surveillance of Large Populations emm sequence typing is the current method of choice for ongoing population-based surveillance. •

emm typing identifies the M protein serotype and thus provides important information relevant to virulence and vaccine development, as well as providing continuity to past experience.141 The advantages of emm typing relative to classical serotyping were detailed above (see Section 12.1.2).

•

Types can be immediately identified by comparison to validated reference sequences using online databases (see Section 12.1.5).

•

For most isolates, emm type is predictive of the genomic lineage as defined by MLST.9,187

12.1.8.3 Population Genetics Studies Individual lineages of GAS can maintain well-defined, closely related (“clonal”) populations for decades. However, analysis of the relationships among emm types and MLST types reveals an extensive recombination within this species; more than half of the recent changes at housekeeping loci in GAS are due to recombination rather than point mutation.186 Thus, MLST is the most rigorous single method for defining genomic lineages within GAS.223

12.2 Streptococcus pneumoniae S. pneumoniae is a major human respiratory tract pathogen responsible for a substantial burden of morbidity and mortality worldwide. Pneumococcal infections vary in severity, ranging from otitis media and sinusitis to pneumonia complicated by septicemia and meningitis. 12.2.1 Phenotypic Typing Based on Capsular Polysaccharide Serotype The capsular polysaccharide represents a critical pneumococcal virulence factor. Ninety distinct serotypes were defined by classical immunologic techniques; however, the distribution of serotypes among clinical isolates is skewed, analogous to M protein serotypes among isolates of GAS. Of the pneumococci that cause invasive disease in older children and adults, 88% are serotypes included in the 23-valent polysaccharide vaccine. Prior to implementation of the pneumococcal 7-valent conjugate vaccine in young children in 2000, over 80% of invasive isolates in children 1 serotypes, including ten STs in which the shared serotypes represent a subset of 7-valent conjugate vaccine serotypes or related (within the same serogroup) types that frequently exhibit antibiotic resistance (6A, 46

©


Volume 27

MM11-A

6B, 9V, 14, 19A, and 23F).246 These exceptions represent serotype changes resulting from horizontal gene transfer.241,247 Such serotype changes within a particular lineage enable a genotype with other virulence factors to “escape” from the immunity provided by current vaccines. The combination of MLST and serotyping is required to resolve these events. Of note, MLST has been used to perform strain typing of pneumococci present in samples from which the organism cannot be cultured in vitro (e.g., cerebrospinal fluid after empiric treatment).239 12.2.5 PFGE PFGE has proven highly effective for molecular strain typing of pneumococci.231,233,248-255 A recent study examined the correlation between genetic lineages defined by MLST and PFGE clusters defined as profiles with a Dice coefficient of >0.80 using UPGMA.170 Among 1168 isolates representing 19 serotypes, there were 40 PFGE clusters; MLST analysis of 165 selected isolates identified 121 STs. For 37 (92.5%) of these clusters, all isolates for which MLST was performed within each cluster had identical alleles for at least five of seven loci sequenced (i.e., represented a clonal complex). 12.2.6 Applying Molecular Strain Typing Methods for S. pneumoniae 12.2.6.1 Outbreaks PFGE is currently the method of choice for analyzing putative outbreaks. However, MLST facilitates correlating studies from different laboratories. For example, serologically nontypeable isolates from a recent conjunctivitis outbreak were identified as ST448 (http://www.mlst.net) and thereby associated with a nontypeable pneumococcal carriage strain recovered in the United Kingdom, as well as with additional conjunctivitis cases within the United States. 256 12.2.6.2 Surveillance Pneumococcal surveillance is conducted to provide information relevant to treatment, vaccine development, and epidemiology. Prior to DNA-based techniques, programs were limited to capsular serotyping and antimicrobial susceptibility testing. Through the collaborative efforts of the PMEN246 and the MLST typing group (http://www.mlst.net), a combination of PFGE and MLST may prove the most cost-effective approach to the timely identification of globally disseminated genotypes. In this approach, isolates are initially characterized using PFGE and compared to online profiles of 26 reference strains (http://www.sph.emory.edu/PMEN) that encompass a significant proportion of the prevalent antibiotic-resistant genotypes. Although comparing PFGE profiles is subjective and inherently limited, isolates closely related to the reference strains can be tentatively identified. To obtain a precise, quantitative measure of relatedness and unambiguous lineage assignment, MLST is applied to a representative of each PFGE cluster. 12.2.6.3 Population Genetics Due to the level of recombination within S. pneumoniae, genomic lineage cannot be reliably determined by sequencing a single locus, as is possible for many isolates of GAS (emm typing) or S. aureus (spa typing). MLST is the only robust method for population studies.

©


47

Number 10

MM11-A

References 1

ISO. International Vocabulary of Basic and General Terms in Metrology. Geneva: International Organization for Standardization; 1993.

2

ISO. In Vitro Diagnostic Medical Devices – Measurement of Quantities in Biological Samples – Metrological Traceability of Values Assigned to Calibrators and Control Materials. ISO 17511. Geneva: International Organization for Standardization; 2003.

3

Tenover FC, Arbeit RD, Goering RV, et al. Interpreting chromosomal DNA restriction patterns produced by pulsed-field gel electrophoresis: criteria for bacterial strain typing. J Clin Microbiol. 1995;33(9):2233-2239.

4

ISO. Quality management systems – Fundamentals and vocabulary. ISO 9000. Geneva: International Organization for Standardization; 2005.

5

Woese C, Kandler O, Wheelis M. Towards a natural system of organisms: proposal for the domains Archaea, Bacteria, and Eucarya. Proc Natl Acad Sci USA. 1990;87(12):4576-4579.

6

Jain R, Rivera MC, Lake JA. Horizontal gene transfer among genomes: the complexity hypothesis. Proc Natl Acad Sci USA. 2002;96:38013806.

7

Salama N, Guillemin K, McDaniel TK, Sherlock G, Tompkins L, Falkow S. A whole-genome microarray reveals genetic diversity among Helicobacter pylori strains. Proc Nat Acad Sci USA. 2000;97:14668-14673.

8

Fitzgerald JR, Sturdevant DE, Mackie SM, Gill SR, Musser JM. Evolutionary genomics of Staphylococcus aureus: insights into the origin of methicillin-resistant strains and the toxic shock syndrome epidemic. Proc Natl Acad Sci USA. 2001;98:8821-8826.

9

Maiden MC, Bygraves JA, Feil E, et al. Multilocus sequence typing: a portable approach to the identification of clones within populations of pathogenic microorganisms. Proc Natl Acad Sci USA. 1998;95(6):3140-3145.

10

Maynard-Smith J, Smith NH, O’Rourke M, Spratt BG. How clonal are bacteria? Proc Natl Acad Sci USA. 1993;90:4384-4388.

11

Kristjansson M, Samore MH, Gerding DN, et al. Comparison of restriction endonuclease analysis, ribotyping, and pulsed-field gel electrophoresis for molecular differentiation of Clostridium difficile strains. J Clin Microbiol. 1994;32(8):1963-1969.

12

Selander RK, Musser JM, Caugant DA, Gilmour MN, Whittam TS. Population genetics of pathogenic bacteria. Microb Pathog. 1987;3:1-7.

13

Arthur M, Arbeit RD, Kim C, et al. Restriction fragment length polymorphisms among uropathogenic Escherichia coli isolates: pap-related sequences compared with rrn operons. Infect Immun. 1990;58:471-479.

14

Johnson J, Kuskowski M, O’Bryan T, Maslow J. Epidemiological correlates of virulence genotype and phylogenetic background among Escherichia coli blood isolates from adults with diverse-source bacteremia. J Infect Dis. 2002;185(10):1439-1447.

15

Musser JM. Molecular population genetic analysis of emerged bacterial pathogens: selected insights. Emerg Infect Dis. 1996;2:1-17.

16

Moran GJ, Krishnadasan A, Gorwitz RJ, et al. Methicillin-resistant S. aureus infections among patients in the emergency department. N Engl J Med. 2006;355:666-674.

17

Samore M, Killgore G, Johnson S, et al. Multicenter typing comparison of sporadic and outbreak Clostridium difficile isolates from geographically diverse hospitals. JID. 1997;176:1233-1238.

18

McDonald LC, Killgore GE, Thompson A, et al. An epidemic, toxin gene-variant strain of Clostridium difficile. N Eng J Med. 2005;353:2433-2441.

19

Loo VG, Poirier L, Miller MA, et al. A predominantly clonal multi-institutional outbreak of Clostridium difficile-associated diarrhea with high morbidity and mortality. N Eng J Med. 2005;353:2442-2449.

20

Klugman KP. The role of clonality in the global spread of fluoroquinolone-resistant bacteria. Clin Infect Dis. 2003;36:783-785.

21

Enright MC, Robinson DA, Randle G, Feil EJ, Grundmann H, Spratt BG. The evolutionary history of methicillin-resistant Staphylococcus aureus (MRSA). Proc Natl Acad Sci USA.. 2002;99:7687-7692.

22

Oliveira DA, Tomasz A, de Lencastre H. Secrets of success of a human pathogen: molecular evolution of pandemic clones of methicillinresistant Staphylococcus aureus. Lancet Infect Dis. 2002;2:180-189.

23

Simor AE, Ofner-Agostini M, Bryce E, McGeer A, Paton S, Mulvey MR. Laboratory characterization of methicillin-resistant Staphylococcus aureus in Canadian hospitals: results of 5 years of National Surveillance, 1995-1999. J Infect Dis. 2002;186:652-660.

24

Arbeit R, Arthur M, Dunn R, Cheung K, Selander RK, Goldstein R. Resolution of recent divergence among Escherichia coli from related lineages: the application of pulsed field gel electrophoresis to molecular epidemiology. J Infect Dis. 1990;161:230-235.

25

Maslow JN, Mulligan ME, Arbeit RD. Recurrent Escherichia coli bacteremia. J Clin Micro. 1994;32:710-714.

26

Wendt C, Messer SA, Hollis RJ, Pfaller MA, Wenzel RP, Herwaldt LA. Molecular epidemiology of gram-negative bacteremia. Clin Infect Dis. 1999;28:605-610.

48

©


Volume 27

MM11-A

27

Kruuner A, Pehme L, Ghebremichael S, Koivula T, Hoffner SE, Mikelsaar M. Use of molecular techniques to distinguish between treatment failure and exogenous reinfection with Mycobacterium tuberculosis. Clin Infect Dis. 2002;35:146-155.

28

Archer GL, Karchmer AW, Vishniavsky N, Johnston JL. Plasmid-pattern analysis for the differentiation of infecting from non-infecting Staphylococcus epidermidis. J Infect Dis. 1984;149:913-920.

29

Arbeit RD, Slutsky A, Barber TW, et al. Genetic diversity among strains of Mycobacterium avium causing monoclonal and polyclonal bacteremia in patients with AIDS. J Infect Dis. 1993;167:1384-1390.

30

Outbreak of Escherichia coli O157:H7 and Campylobacter among attendees of the Washington County Fair – New York, 1999. MMWR. 1999;48:803.

31

Hakim A, Deplano A, Maes N, Kentos A, Rossi C, Struelens MJ. Polyclonal coagulase-negative staphylococcal catheter-related bacteremia documented by molecular identification and typing. Clin Microbiol Infect. 1999;5(4):224-227.

32

Waminathan B, Barrett TJ, Hunter SB, Tauxe RV. PulseNet: the molecular subtyping network for foodborne bacterial disease surveillance, United States. Emerg Infect Dis. 2001;7(3):382-389.

33

Gupta A, Fontana J, Crowe C, et al. Emergence of multidrug-resistant Salmonella enterica serotype Newport infections resistant to expanded-spectrum cephalosporins in the United States. J Infect Dis. 2003;188(11):1707-1716.

34

Jay MT, Garrett V, Mohle-Boetani JC, et al. A multistate outbreak of Escherichia coli O157:H7 infection linked to consumption of beef tacos at a fast-food restaurant chain. Clin Infect Dis. 2004;39:1-7.

35

McDougal L, Steward C, Killgore G, Chaitram J, McAllister S, Tenover F. Pulsed-field gel electrophoresis typing of oxacillin-resistant Staphylococcus aureus isolates from the United States: establishing a national database. J Clin Microbiol. 2003;41(11):5113-5120.

36

Spratt B. Exploring the concept of clonality in bacteria. Methods Mol Biol. 2004;266:323-352.

37

Koreen L, Ramaswamy SV, Graviss EA, Naidich S, Musser JM, Kreiswirth BN. spa typing method for discriminating among Staphylococcus aureus isolates: implications for use of a single marker to detect genetic micro- and macrovariation. J Clin Micro. 2004;42:792-799.

38

Glynn JR, Yates MD, Crampin AC, et al. DNA fingerprint changes in tuberculosis: reinfection, evolution, or laboratory error. J Infect Dis. 2004;190:1158-1166.

39

Rubin LG. Bacterial colonization and infection resulting from multiplication of a single organism. Rev Infect Dis. 1987;9:488-493.

40

van Eldere J, Peetermans WE, Struelens M, Deplano A, Bobbaers H. Polyclonal staphylococcal endocarditis caused by genetic variability. Clin Infect Dis. 2000;31:24-30.

41

Hunter PR, Gaston MA. Numerical index of the discriminatory ability of typing systems: an application of Simpson’s index of diversity. J Clin Microbiol. 1988;26:2465-2466.

42

Hunter PR. Reproducibility and indices of discriminatory power of microbial typing methods. J Clin Microbiol. 1990;28:1903-1905.

43

Streulens MJ and the Members of the European Study Group on Epidemiological Markers. Consensus guidelines for appropriate use and evaluation of microbial epidemiologic typing systems. Clin Micro Infect. 1996;2:2-11.

44

Kremer K, van Soolingen D, Frothingham R, et al. Comparison of methods based on different molecular epidemiological markers for typing of Mycobacterium tuberculosis complex strains: interlaboratory study of discriminatory power and reproducibility. J Clin Microbiol. 1999;37(8):2607-2618.

45

Schwartz DC, Cantor CR. Separation of yeast chromosome-sized DNAs by pulsed field gradient gel electrophoresis. Cell. 1984;37(1):6775.

46

Struelens M, de Ryck R, Deplano A. Analysis of microbial genomic macrorestriction patterns by pulsed-field gel electrophoresis typing. In: Dijkshoorn L, Towner K, Struelens M, eds. New Approaches for the Generation and Analysis of Microbial Typing Data. Elsevier; 2001;159-176.

47

Tenover FC, Arbeit RD, Goering RV, and the Molecular Typing Working Group of the Society for Healthcare Epidemiology of America. How to select and interpret molecular strain typing methods for epidemiological studies of bacterial infections: a review for healthcare epidemiologists. Infect Control Hosp Epidemiol. 1997;18(6):426-439.

48

Struelens MJ, De Gheldre Y, Deplano A. Comparative and library epidemiological typing systems: outbreak investigations versus surveillance systems. Infect Control Hosp Epidemiol. 1998;9(8):565-569.

49

Olive DM, Bean P. Principles and applications of methods for DNA-based typing of microbial organisms. J Clin Microbiol. 1999;37(6):1661-1669.

50

Arbeit RD. Laboratory procedures for the epidemiologic analysis of microorganisms. In: Murray PR, Baron EJ, Pfaller MA, Tenover FC, Yolken RH, eds. Manual of Clinical Microbiology. Washington, D.C.: ASM Press; 1999:116-137.

©

49


Number 10

MM11-A

51

Tenover FC, Arbeit R, Archer G, et al. Comparison of traditional and molecular methods of typing isolates of Staphylococcus aureus. J Clin Microbiol. 1994;32(2):407-415.

52

Struelens MJ. Consensus guidelines for appropriate use and evaluation of microbial epidemiologic typing systems. Clin Microbiol Infect. 1996;2(1):2-11.

53

Grundmann H, Schneider C, Hartung D, Daschner FD, Pitt TL. Discriminatory power of three DNA-based typing techniques for Pseudomonas aeruginosa. J Clin Microbiol. 1995;33(3):528-534.

54

Murray BE, Singh KV, Heath JD, Sharma BR, Weinstock GM. Comparison of genomic DNAs of different enterococcal isolates using restriction endonucleases with infrequent recognition sites. J Clin Microbiol. 1990;28(9):2059-2063.

55

Poh CL, Yeo CC, Tay L. Genome fingerprinting by pulsed-field gel electrophoresis and ribotyping to differentiate Pseudomonas aeruginosa serotype O11 strains. Eur J Clin Microbiol Infect Dis. 1992;11(9):817-822.

56

Prevost G, Jaulhac B, Piemont Y. DNA fingerprinting by pulsed-field gel electrophoresis is more effective than ribotyping in distinguishing among methicillin-resistant Staphylococcus aureus isolates. J Clin Microbiol. 1992;30(4):967-973.

57

Struelens MJ, Bax R, Deplano A, Quint WG, Van Belkum A. Concordant clonal delineation of methicillin-resistant Staphylococcus aureus by macrorestriction analysis and polymerase chain reaction genome fingerprinting. J Clin Microbiol. 1993;31(8):1064-1070.

58

Bannerman TL, Hancock GA, Tenover FC, Miller JM. Pulsed-field gel electrophoresis as a replacement for bacteriophage typing of Staphylococcus aureus. J Clin Microbiol. 1995;33(3):551-555.

59

Kersulyte D, Struelens MJ, Deplano A, Berg DE. Comparison of arbitrarily primed PCR and macrorestriction (pulsed-field gel electrophoresis) typing of Pseudomonas aeruginosa strains from cystic fibrosis patients. J Clin Microbiol. 1995;33(8):2216-2219.

60

Seifert H, Gerner-Smidt P. Comparison of ribotyping and pulsed-field gel electrophoresis for molecular typing of Acinetobacter isolates. J Clin Microbiol. 1995;33(5):1402-1407.

61

Hermans PW, Sluijter M, Hoogenboezem T, Heersma H, van Belkum A, de Groot R. Comparative study of five different DNA fingerprint techniques for molecular typing of Streptococcus pneumoniae strains. J Clin Microbiol. 1995;33(6):1606-1612.

62

Gori A, Espinasse F, Deplano A, Nonhoff C, Nicolas MH, Struelens MJ. Comparison of pulsed-field gel electrophoresis and randomly amplified DNA polymorphism analysis for typing extended-spectrum-beta-lactamase-producing Klebsiella pneumoniae. J Clin Microbiol. 1996;34(10):2448-2453.

63

Kerouanton A, Brisabois A, Denoyer E, et al. Comparison of five typing methods for the epidemiological study of Listeria monocytogenes. Int J Food Microbiol. 1998;43(1-2):61-71.

64

Wassenaar TM, Newell DG. Genotyping of Campylobacter spp. Appl Environ Microbiol. 2000;66(1):1-9.

65

Finney M. Pulsed-field gel electrophoresis. In: Ausubel FM, Brent R, Kingston RE, et al, eds. Current Protocols in Molecular Biology. New York: Greene-Wiley; 1993;2.5.9-2.5.17.

66

Maslow JN, Slutsky AM, Arbeit RD. Application of pulsed-field gel electrophoresis to molecular epidemiology. In: Persing DH, Smith TF, Tenover FC, White TJ, eds. Diagnostic Molecular Microbiology: Principles and Applications. Washington, D.C.: ASM Press; 1993;563572.

67

Michaud S, Arbeit RD, Gaudreau C. Molecular strain typing of Campylobacter jejuni by pulsed-field gel electrophoresis in a single day. Can J Microbiol. 2001;47(7):667-669.

68

Gautom RK. Rapid pulsed-field gel electrophoresis protocol for typing of Escherichia coli O157:H7 and other gram-negative organisms in 1 day. J Clin Microbiol. 1997;35(11):2977-2980.

69

Ribot EM, Fitzgerald C, Kubota K, Swaminathan B, Barrett TJ. Rapid pulsed-field gel electrophoresis protocol for subtyping of Campylobacter jejuni. J Clin Microbiol. 2001;39(5):1889-1894.

70

Turabelidze D, Kotetishvili M, Kreger A, Morris JG, Jr., Sulakvelidze A. Improved pulsed-field gel electrophoresis for typing vancomycinresistant enterococci. J Clin Microbiol. 2000;38(11):4242-4245.

71

Goering RV, Winters MA. Rapid method for epidemiological evaluation of gram-positive cocci by field inversion gel electrophoresis. J Clin Microbiol. 1992;30(3):577-580.

72

Mulvey MR, Chui L, Ismail J, et al. Development of a Canadian standardized protocol for subtyping methicillin-resistant Staphylococcus aureus using pulsed-field gel electrophoresis. J Clin Microbiol. 2001;39(10):3481-3485.

73

McClelland M, Jones R, Patel Y, Nelson M. Restriction endonucleases for pulsed field mapping of bacterial genomes. Nucleic Acids Res. 1987;15(15):5985-6005.

74

Birren B, Lai E. Pulsed Field Gel Electrophoresis: A Practical Guide. San Diego, CA: Academic Press; 1993.

50

©


Volume 27

MM11-A

75

Birren BW, Lai E, Clark SM, Hood L, Simon MI. Optimized conditions for pulsed-field gel electrophoretic separations of DNA. Nucleic Acids Res. 1988;16(15):7563-7582.

76

Heersma HF, Kremer K, van Soolingen D, Hauman J. Setting up intra- and inter-laboratory databases of electrophoretic profiles. In: Dijkshoorn L, Towner K, Struelens M, eds. New Approaches for the Generation and Analysis of Microbial Typing Data. Elsevier; 2001:4775.

77

Gibson JR, Sutherland K, Owen RJ. Inhibition of DNAse activity in PFGE analysis of DNA from Campylobacter jejuni. Lett Appl Microbiol. 1994;19:357-358.

78

Römling U, Tümmler B. Achieving 100% typeability of Pseudomonas aeruginosa by pulsed-field gel electrophoresis. J Clin Microbiol. 2000;38:464-465.

79

Koort JM, Lukinmaa S, Rantala M, Unkila E, Siitonen A. Technical improvement to prevent DNA degradation of enteric pathogens in pulsed-field gel electrophoresis. J Clin Microbiol. 2002;40:3497-3498.

80

Morris T, Brecher SM, Fitzsimmons D, Durbin A, Arbeit RD, Maslow JN. A pseudoepidemic of septic arthritis. Infect Cont Hosp Epidem. 1995;16:82-87.

81

Weber S, Pfaller MA, Herwaldt LA. Role of molecular epidemiology in infection control. In: Rutala WA, Weber DJ, eds. Nosocomial Infections. Saunders; 1997:257-278.

82

McEllistrem MC, Stout JE, Harrison LH. Simplified protocol for pulsed-field gel electrophoresis analysis of Streptococcus pneumoniae. J Clin Microbiol. 2000;38(1):351-353.

83

Michaud S, Ménard S, Gaudreau C, Arbeit RD. Comparison of SmaI-defined genotypes of Campylobacter jejuni examined by KpnI: a population-based study. J Med Microbiol. 2001;50:1075-1081.

84

Kreiswirth B, Kornblum J, Arbeit RD, et al. Evidence for a clonal origin of methicillin resistance in Staphylococcus aureus. Science. 1993;259(5092):227-230.

85

Musser JM, Kroll JS, Moxon ER, Selander RK. Evolutionary genetics of the encapsulated strains of Haemophilus influenzae. Proc Natl Acad Sci USA. 1988;85(20):7758-7762.

86

Whittam TS, Wachsmuth IK, Wilson RA. Genetic evidence of clonal descent of Escherichia coli O157:H7 associated with hemorrhagic colitis and hemolytic uremic syndrome. J Infect Dis. 1988;157(6):1124-1133.

87

Bohm H, Karch H. DNA fingerprinting of Escherichia coli O157:H7 strains by pulsed-field gel electrophoresis. J Clin Microbiol. 1992;30(8):2169-2172.

88

O’Neill GL, Murchan S, Gil-Setas A, Aucken HM. Identification and characterization of phage variants of a strain of epidemic methicillinresistant Staphylococcus aureus (EMRSA-15). J Clin Microbiol. 2001;39:1540-1548.

89

Buchrieser C, Weagant SD, Kaspar CW. Molecular characterization of Yersinia enterocolitica by pulsed-field gel electrophoresis and hybridization of DNA fragments to ail and pYV probes. App Environ Microbiol. 1994;60:4371-4379.

90

Maslow JN, Brecher S, Gunn J, Durbin A, Barlow MA, Arbeit RD. Variation and persistence of methicillin-resistant Staphylococcus aureus strains among individual patients over extended periods of time. Eur J Clin Microbiol Infect Dis. 1995;14(4):282-290.

91

Deplano A, Witte W, van Leeuwen WJ, Brun Y, Struelens MJ. Clonal dissemination of epidemic methicillin-resistant Staphylococcus aureus in Belgium and neighboring countries. Clin Microbiol Infect. 2000;6(5):239-245.

92

Simor AE, Ofner-Agostini M, Bryce E, et al. The evolution of methicillin-resistant Staphylococcus aureus in Canadian hospitals: 5 years of national surveillance. CMAJ. 2001;165(1):21-26.

93

Ip M, Lyon DJ, Yung RW, Chan C, Cheng AF. Evidence of clonal dissemination of multidrug-resistant Streptococcus pneumoniae in Hong Kong. J Clin Microbiol. 1999;37(9):2834-2839.

94

Greenberg D, Speert DP, Mahenthiralingam E, Henry DA, Campbell ME, Scheifele DW. Emergence of penicillin-nonsusceptible Streptococcus pneumoniae invasive clones in Canada. J Clin Microbiol. 2002;40(1):68-74.

95

Richter SS, Heilmann KP, Coffman SL, et al. The molecular epidemiology of penicillin-resistant Streptococcus pneumoniae in the United States, 1994-2000. Clin Infect Dis. 2002;34(3):330-339.

96

Tassios PT, Gazouli M, Tzelepi E, et al. Spread of a Salmonella typhimurium clone resistant to expanded-spectrum cephalosporins in three European countries. J Clin Microbiol. 1999;37(11):3774-3777.

97

Tassios PT, Gennimata V, Maniatis AN, Fock C, Legakis NJ. Emergence of multidrug resistance in ubiquitous and dominant Pseudomonas aeruginosa serogroup O:11. The Greek Pseudomonas aeruginosa Study Group. J Clin Microbiol. 1998;36(4):897-901.

98

Van Looveren M, Vandamme P, Hauchecorne M, et al. Molecular epidemiology of recent Belgian isolates of Neisseria meningitidis serogroup B. J Clin Microbiol. 1998;36(10):2828-2834.

©

51


Number 10

MM11-A

99

Grimont F, Grimont P. Ribosomal ribonucleic acid gene restriction patterns as potential taxonomic tools. Ann Inst Pasteur/Microbiol. 1986;137B:165-175.

100

Stull TL, LiPuma JJ, Edlind TD. A broad-spectrum probe for molecular epidemiology of bacteria: ribosomal RNA. J Infect Dis. 1988;157:280-286.

101

Pace NR, Olsen GJ, Woese CR. Ribosomal RNA phylogeny and the primary lines of evolutionary descent. Cell. 1986;45:325-326.

102

Altwegg M, Hickman-Brenner FW, Farmer III JJ. Ribosomal RNA gene restriction patterns provide increased sensitivity for typing Salmonella typhi strains. J Infect Dis. 1989;160:145-149.

103

Wachsmuth IK, Kiehlbauch JA, Bopp CA, et al. The use of plasmid profiles and nucleic acid probes in epidemiologic investigations of foodborne diarrheal diseases. Int J Food Microbiol. 1991;12:77-90.

104

Woods TC, Heisel LO, Swaminathan B, et al. Characterization of Neisseria meningitidis serogroup C by multilocus enzyme electrophoresis and ribosomal DNA restriction profiles (ribotyping). J Clin Microbiol. 1992;30:132-137.

105

Blumberg HM, Kiehlbauch JA, Wachsmuth IK. Molecular epidemiology of Yersinia enterocolitica O:3 infections: use of chromosomal DNA restriction fragment length polymorphism of rRNA genes. J Clin Microbiol. 1991;29:2368-2374.

106

Gustaferro CA, Persing DH. Chemiluminescent universal probe for bacterial ribotyping. J Clin Microbiol. 1992;30:1039-1041.

107

Jordens JZ. Genomic DNA digestion and ribotyping. In: Woodford N, Johnson AP, eds. Molecular Bacteriology Protocols and Clinical Applications. Totowa, NJ: Humana Press; 1998;17-31.

108

Grimont F, Grimont P. Determination of rRNA gene restriction patterns. In: Howard JJ, Whitcombe D, eds. Methods in Microbiology: Diagnostic Bacteriology Protocols. Totowa, NJ: Humana Press; 1995;46:181-200.

109

Popovic T, Bopp CA, Olsvik O, et al. Ribotyping in molecular epidemiology. In: Pershing DH, Smith TF, Tenover FC, et al, eds. Diagnostic Molecular Microbiology Principles and Applications. Washington, D.C.: ASM Press; 1993;573-583.

110

Hollis R, Bruce J, Fritschel S, Pfaller M. Comparative evaluation of an automated ribotyping instrument versus pulsed-field gel electrophoresis for epidemiological investigation of clinical isolates of bacteria. Diagn Microbiol Infect Dis. 1999;34(4):263-268.

111

Kodio A, Villard L, Bizet C, et al. Pulsed-field gel electrophoresis is more efficient than ribotyping and random amplified polymorphic DNA analysis in discrimination of Pasteurella haemolytica strains. J Clin Microbiol. 1999;37:380-385.

112

Gordillo ME, Singh KV, Murray BE. Comparison of ribotyping and pulsed-field gel electrophoresis for subspecies differentiation of strains of Enterococcus faecalis. J Clin Microbiol. 1993;31:1570-1574.

113

Hall LM, Duke B, Guiney M. Typing of Enterococcus species by DNA restriction fragment analysis. J Clin Microbiol. 1992;30:915-919.

114

Bingen E, Denamur E, Lambert-Zaechovsky N, et al. Mother-to-infant vertical transmission and cross-colonization of Streptococcus pyogenes confirmed by DNA restriction fragment length polymorphism analysis. J Infect Dis. 1992;165:147-149.

115

Postic D, Edlinger C, Richaud C, et al. Two genomic species in Borrelia burgdorferi. Res Microbiol. 1990;141:465-475.

116

Patra G, Fouet A, Vaissaire J, et al. Variation in rRNA operon number as revealed by ribotyping of Bacillus anthracis strains. Res Microbiol. 2002;152:139-148.

117

Bingen EH, Denamur E, Elion J. Use of ribotyping in epidemiological surveillance of nosocomial outbreaks. Clin Microbiol Rev. 1994;7:311-327.

118

LiPuma JJ, Fisher MC, Dasen SE, et al. Ribotype stability of serial pulmonary isolates of Pseudomonas cepacia. J Infect Dis. 1991;164:133-136.

119

Rabkin C, Jarvis W, Anderson R, et al. Pseudomonas cepacia typing systems: collaborative study to assess their potential in epidemiologic investigations. Rev Infect Dis. 1989;11:600-607.

120

Lindstedt BA. Multiple-locus variable number tandem repeats analysis for genetic fingerprinting of pathogenic bacteria. Electrophoresis. 2005;26:2567-2582.

121

Klevytska AM, Price LB, Schupp JM, Worsham PL, Wong J, Keim P. Identification and characterization of variable-number tandem repeats in the Yersinia pestis genome. J Clin Microbiol. 2001;39:3179-3185.

122

Lindstedt BA, Heir A, Gjernes E, Vardund T, Kapperud G. DNA fingerprinting of Shiga-toxin producing Escherichia coli O157 based on multiple-locus variable-number tandem-repeats analysis (MLVA). Ann Clin Microbiol Antimicrob. 2003;2:12-16.

123

Liu Y, Lee MA, Ooi EE, Mavis Y, Tan AL, Quek HH. Molecular typing of Salmonella enterica serovar typhi isolates from various countries in Asia by a multiplex PCR assay on variable-number tandem repeats. J Clin Microbiol. 2003;41:4388-4394.

52

©


Volume 27

MM11-A

124

Sabat A, Krzyszton-Russjan J, Strzalka W, et al. New method for typing Staphylococcus aureus strains: multiple-locus variable-number tandem repeat analysis of polymorphism and genetic relationships of clinical isolates. J Clin Microbiol. 2003;41:1801-1804.

125

Malachowa N, Sabat A, Gniadkowski M, et al. Comparison of multiple-locus variable-number tandem-repeat analysis with pulsed-field gel electrophoresis, spa typing, and multilocus sequence typing for clonal characterization of Staphylococcus aureus isolates. J Clin Microbiol. 2005;43:3095-3100.

126

Andersen GL, Simchock JM, Wilson KH. Identification of a region of genetic variability among Bacillus anthracis strains and related species. J Bacteriol. 1966;178:377-384.

127

Nishi JI, Miyanohara H, Nakajima T, et al. Molecular typing of the methicillin resistant determinant (mec) of clinical strains of Staphylococcus based on mec hypervariable region length polymorphisms. J Lab Clin Med. 1995;126:29-35.

128

Beall B, Facklam R, Thompson T. Sequencing emm-specific PCR products for routine and accurate typing of group A streptococci. J Clin Microbiol. 1996;34:953-958.

129

Hoe NP, Fullerton KE, Liu M, et al. Molecular genetic analysis of 675 group A Streptococcus isolates collected in a carrier study at Lackland Air Force Base, San Antonio, Texas. J Infect Dis. 2003;188:818-827.

130

Shopsin B, Gomez M, Montgomery SO, et al. Evaluation of protein A gene polymorphic region DNA sequencing for typing of Staphylococcus aureus strains. J Clin Microbiol. 1999;37:3556-3563.

131

Harmsen D, Claus H, Witte W, et al. Typing of methicillin-resistant Staphylococcus aureus in a university hospital setting by using novel software for spa repeat determination and database management. J Clin Microbiol. 2003;41:5442-5448.

132

Tang YW, Waddington MG, Smith DH, et al. Comparison of protein A gene sequencing with pulsed-field gel electrophoresis and epidemiologic data for molecular typing of methicillin-resistant Staphylococcus aureus. J Clin Microbiol. 2000;38:1347-1351.

133

U’Ren JM, Van Ert MN, Schupp JM, et al. Use of a real-time PCR TaqMan assay for rapid identification and differentiation of Burkholderia pseudomallei and Burkholderia mallei. J Clin Microbiol. 2005;43:5771-5774.

134

Pai R, Limor J, Beall B. Use of pyrosequencing to differentiate Streptococcus pneumoniae serotypes 6A and 6B. J Clin Microbiol. 2005;43:4820-4822.

135

Robertson GA, Thiruvenkataswamy V, Shilling H, et al. Identification and interrogation of highly informative single nucleotide polymorphism sets defined by bacterial multilocus sequence typing databases. J Med Microbiol. 2004;53:35-45.

136

Huygens F, Stephens AJ, Nimmo GR, Giffard PM. mecA locus diversity in methicillin-resistant Staphylococcus aureus isolates in Brisbane, Australia, and the development of a novel diagnostic procedure for the Western Samoan phage pattern clone. J Clin Microbiol. 2004;42(5):1947-1955.

137

Arnold C, Westland L, Mowat G, Underwood A, Magee J, Gharbia S. Single-nucleotide polymorphism-based differentiation and drug resistance detection in Mycobacterium tuberculosis from isolates or directly from sputum. Clin Microbiol Infect. 2005;11:122-130.

138

Zhao JR, Bai YJ, Wang Y, Zhang QH, Luo M, Yan XJ. Development of a pyrosequencing approach for rapid screening of rifampin, isoniazid and ethambutol-resistant Mycobacterium tuberculosis. Int J Tuberc Lung Dis. 2005;9:328-332.

139

Salama N, Guillemin K, McDaniel TK, Sherlock G, Tompkins L, Falkow S. A whole-genome microarray reveals genetic diversity among Helicobacter pylori strains. Proc Nat Acad Sci USA. 2000;97:14668-14673.

140

Shopsin B, Gomez M, Montgomery SO, et al. Evaluation of protein A gene polymorphic region DNA sequencing for typing of Staphylococcus aureus strains. J Clin Micro. 1999;37(11):3556.

141

O’Brien K, Beall B, Barrett N, et al. Epidemiology of invasive group A Streptococcus disease in the United States, 1995-1999. Clin Infect Dis. 2002;35:268-276.

142

Peacock SJ, de Silva GDI, Justice A, et al. Comparison of multilocus sequence typing and pulsed-field gel electrophoresis as tools for typing Staphylococcus aureus isolates in a microepidemiological setting. J Clin Micro. 2002;40:3764-3770.

143

Feil EJ, Smith JM, Enright MC, Spratt BG. Estimating recombinational parameters in Streptococcus pneumoniae from multilocus sequence typing data. Genetics. 2000;154:1439-1450.

144

Versalovic J, Koeuth T, Lupski JR. Distribution of repetitive DNA sequences in eubacteria and application to fingerprinting of bacterial genomes. Nucleic Acids Res. 1991;19:6823-6831.

145

Stern MJ, Ames GF, Smith NH, Robinson EC, Higgins CF. Repetitive extragenic palindromic sequences: a major component of the bacterial genome. Cell. 1984 37:1015-1026.

146

Hulton CS, Higgins CF, Sharp PM. ERIC sequences: a novel family of repetitive elements in the genomes of Escherichia coli, Salmonella typhimurium and other enterobacteria. Mol Microbiol. 1991;5:825-834.

©


53

Number 10

MM11-A

147

Koeuth T, Versalovic J, Lupski JR. Differential subsequence conservation of interspersed repetitive Streptococcus pneumoniae BOX elements in diverse bacteria. Genome Res. 1995;5:408-418.

148

Olive DM, Bean P. Principles and applications of methods for DNA-based typing of microbial organisms. J Clin Microbiol. 1999;37:16611669.

149

Deplano A, Schuermans A, Van Eldere J, et al. Multicenter evaluation of epidemiological typing of methicillin-resistant Staphylococcus aureus strains by repetitive-element PCR analysis. The European Study Group on Epidemiological Markers of the ESCMID. J Clin Microbiol. 2000;38:3527-3533.

150

Ross TL, Merz WG, Farkosh M, Carroll KC. Comparison of an automated repetitive sequence-based PCR microbial typing system to pulsed-field gel electrophoresis for analysis of outbreaks of methicillin-resistant Staphylococcus aureus. J Clin Microbiol. 2005;43:56425647.

151

Barbier N, Saulnier P, Chachaty E, Dumontier S, Andremont A. Random amplified polymorphic DNA typing versus pulsed-field gel electrophoresis for epidemiological typing of vancomycin-resistant enterococci. J Clin Microbiol. 1996;34:106-109.

152

Liu PY, Wu WL. Use of different PCR-based DNA fingerprinting techniques and pulsed-field gel electrophoresis to investigate the epidemiology of Acinetobacter calcoacetics-Acinetobacter baumannii complex. Diagn Microbiol Infect Dis. 1997;28:19.

153

Johnson JR, O’Bryan TT. Improved repetitive-element PCR fingerprinting for resolving pathogenic and nonpathogenic phylogenetic groups within Escherichia coli. Clin Diagn Lab Immunol. 2000;7(2):265-273.

154

Plikaytis BB, Marden JL, Crawford JT, Woodley CL, Butler WR, Shinnick TM. Multiplex PCR assay specific for the multidrug-resistant strain W of Mycobacterium tuberculosis. J Clin Microbiol. 1994;32:542-1546.

155

Healy M, Huong J, Bittner T, et al. Microbial DNA typing by automated repetitive-sequence-based PCR. J Clin Microbiol. 2005;43(1):199.

156

Gerner-Smidt P, Graves LM, Hunter S, Swainathan B. Computerized analysis of restriction fragment polymorphisms patterns: comparative evaluation of two commercial software packages. J Clin Micro. 1998;36:1318-1323.

157

Duck W, Steward C, Banerjee S, McGowan J, Tenover F. Optimization of computer software settings improves accuracy of pulsed-field gel electrophoresis macrorestriction fragment pattern analysis. J Clin Microbiol. 2003;41(7):3035-3042.

158

Cardinali G, Martini A. Critical observations on computerized analysis of banding patterns with commercial software packages [letter]. J Clin Micro. 1999;37:876-877.

159

BioNumerics: The Integral Study of Biological Relationships. Kortrijk, Belgium: Applied Maths; 2000;230.

160

van Ooyen A. Theoretical aspects of pattern analysis. In: Dijkshoorn L, Towner K, Struelens M, eds. New Approaches for the Generation and Analysis of Microbial Typing Data. Elsevier: 2001;31-45.

161

Lina B, Bes M, Vandenesch F, Greenland T, Etienne J, Fleurette J. Role of bacteriophages in genomic variability of related coagulasenegative staphylococci. FEMS Microbiol Lett. 1993;109:273-278.

162

Chachaty E, Saulnier P, Martin A, Mario N, Andremont A. Comparison of ribotyping, pulsed-field gel electrophoresis and random amplified polymorphic DNA for typing Clostridium difficile strains. FEMS Microbiol Lett. 1994;122(1-2):61-68.

163

Mathiesen DA, Oliver JH Jr, Kolbert CP, et al. Genetic heterogeneity of Borrelia burgdorferi in the United States. J Infect Dis. 1997;175(1):98-107.

164

Jolley KA, Chan MS, Maiden MCJ. mlstdbNet – distributed multi-locus sequence typing (MLST) databases. BMC Bioinformatics. 2004;5:86-94.

165

Beall B, Gherardi G, Lovgren M, Facklam RR, Forwick BA, Tyrrell GJ. Emm and sof gene sequence variation in relation to serological typing of opacity-factor-positive group A streptococci. Microbiol. 2004;39:325-332.

166

Beall B, Facklam R, Hoenes T, Schwartz B. Survey of emm gene sequences and T-antigen types from systemic Streptococcus pyogenes infection isolates collected in San Francisco, California; Atlanta, Georgia; and Connecticut in 1994 and 1995. J Clin Microbiol. 1997;35(5):1231-1235.

167

Shulman ST, Tanz RR, Kabat W, et al. US Streptococcal Pharyngitis Surveillance Group. Group A streptococcal pharyngitis serotype surveillance in North America, 2000-2002. Clin Infect Dis. 2004;39:325-332.

168

Li Z, Sakota V, Jackson D, Franklin AR, Beall B, Active Bacterial Core Surveillance/Emerging Infections Program Network. Array of M protein gene subtypes in 1064 recent invasive group A streptococcus isolates recovered from the active bacterial core surveillance. J Infect Dis. 2003;188(10):1587-1592.

169

Feil EJ, Li B, Aanensen DM, Hanage WP, Spratt BG. eBURST: inferring patterns of evolutionary descent among clusters of related bacterial genotypes from multilocus sequence typing data. J Bact. 2004;186:1518-1530.

54

©


Volume 27

MM11-A

170

Gertz RE Jr, McEllistrem MC, Boxrud DJ, et al. Clonal distribution of invasive pneumococcal isolates from children and selected adults in the United States prior to 7-valent conjugate vaccine introduction. J Clin Microbiol. 2003;41:4194-4216.

171

Swofford DL, Olsen GJ. Phylogeny reconstruction. In: Hillis DH, Moritz C, eds. Molecular Systematics. Sunderland, MA: Sinauer Associates; 1990:411-501.

172

Lancefield RC. The antigenic complex of Streptococcus haemolyticus. I. Demonstration of a type-specific substance in extracts of Streptococcus haemolyticus. J Exp Med. 1928;47:91-103.

173

Cleary PP, Kaplan EL, Handley JP, et al. Clonal basis for resurgence of serious Streptococcus pyogenes disease in the 1980s. Lancet. 1992;339:518-521.

174

Musser JM, Hauser AR, Kim MH, Schlievert PM, Nelson K, Selander RK. Streptococcus pyogenes causing toxic-shock-like syndrome. Proc Nat Acad Sci USA. 1991;88:2668-2672.

175

Carapetis J, Robins-Browne R, Martin D, Shelby-James T, Hogg G. Increasing severity of invasive group A streptococcal disease in Australia: clinical and molecular epidemiological features and identification of a new virulent M-nontypeable clone. Clin Infect Dis. 1995;21:1220-1227.

176

Cockerill FR III, Thompson RL, Musser JM, et al, and Southeastern Minnesota Streptococcal Working Group. Molecular, serological, and clinical features of 16 consecutive cases of invasive streptococcal disease. Clin Infect Dis. 1998;26:1448-1458.

177

Hu MC, Walls MA, Stroop S, Reddish M, Beal B, Dale JB. Immunogenicity of a 26-valent group A streptococcal vaccine. Infect Immun. 2002;70:2171-2177.

178

Brandt ER, Sriprakash KS, Hobb RI, et al. New multi-determinant strategy for a group A streptococcal vaccine designed for the Australian Aboriginal population. Nature Medicine. 2000;6:455-459.

179

Kotloff KL, Corretti M, Palmer K, et al. Safety and immunogenicity of a recombinant multivalent group A streptococcal vaccine in healthy adults: phase 1 trial. JAMA. 2004;292:709-715.

180

Facklam R, Beall B, Efstratiou A, et al. emm typing and validation of provisional M types for group A streptococci. Emerg Infec Dis. 1999;5:247-253.

181

Whatmore AM, Kapur V, Sullivan DJ, Musser JM, Kehoe MA. Non-congruent relationships between variation in emm gene sequences and the population genetic structure of group A streptococci. Mol Microbiol. 1994;14:619-631.

182

Facklam R, Beall B, Efstratiou A, et al. Nomenclature of new emm types of group A streptococci. In: Martin DR, Tagg JR, eds. Streptococci and Streptococcal Diseases - Entering the New Millennium. Proceedings XIV Lancefield International Symposium on Streptococci and Streptococcal Diseases. Securacopy, New Zealand: 2000;797-803.

183

Facklam R, Efstratiou A, Kriz P, et al. Continued extension of the Lancefield classification scheme for group A streptococci by addition of 22 new M protein gene sequence types: emm103 to emm124. Clin Infect Dis. 2002;34:28-38.

184

Cohen-Poradosu R, Jaffe J, Lavi D, et al. Group G streptococcal bacteremia in Jerusalem. Emerg Infect Dis. 2004;10:1455-1460.

185

Igwe EI, Shewmaker PL, Facklam RR, Farley MM, van Beneden C, Beall B. Identification of superantigen genes speM, ssa, and smeZ in invasive strains of beta-hemolytic group C and G streptococci recovered from humans. FEMS Microbiol Lett. 2003;229:259-264.

186

McGregor KF, Spratt BG, Kalia A, et al. Multilocus sequence typing of Streptococcus pyogenes representing most known emm types and distinctions among subpopulation genetic structures. J Bacteriol. 2004;186:4285-4294.

187

Enright MC, Spratt BG, Kalia A, Cross JH, Bessen DE. Multilocus sequence typing of Streptococcus pyogenes and the relationships between emm type and clone. Infect Immun. 2001;69:2416-2427.

188

McGregor KF, Bilek N, Bennett A, et al. Group A streptococci from a remote community have novel multilocus genotypes but share emm types and housekeeping alleles with isolates from worldwide sources. J Infect Dis. 2004;189:717-723.

189

Beall B, Facklam R, Hoenes T, Schwartz B. Application of emm gene sequencing and T antigen serology for typing group A streptococci systemic isolates: survey of random and outbreak-related isolates. Adv Exp Med Biol. 1997;418:307-311.

190

Phillips G, Efstratiou A, Roworth M, Beall B, Ferguson J, Tanna A. An outbreak of skin sepsis in abattoir workers caused by an ‘unusual’ strain of Streptococcus pyogenes. J Med Microbiol. 2000;49:371-374.

191

Dicuonzo G, Gherardi G, Lorino G, et al. Group A streptococcal genotypes from pediatric throat isolates in Rome, Italy. J Clin Microbiol. 2001;39:1687-1690.

192

Teixeira L, Barros R, Castro A, et al. Genetic and phenotypic features of Streptococcus pyogenes strains isolated in Brazil that harbor new emm sequences. J Clin Microbiol. 2001;39:3290-3295.

193

Haukness HA, Tanz RR, Thomson RB, et al. The Heterogeneity of endemic community pediatric group A streptococcal isolates and their relationship to invasive isolates. J Inf Dis. 2002;185:915-920.

©


55

Number 10

MM11-A

194

Perea-Mejia LM, Inzunza-Montiel AE, Cravioto A. Molecular characterization of group A Streptococcus strains isolated during a scarlet fever outbreak. J Clin Microbiol. 2002;40(1):278-280.

195

Brandt CM, Honscha M, Truong ND, et al. Macrolide resistance in Streptococcus pyogenes isolates from throat infections in the region of Aachen, Germany. Microb Drug Resist. 2001;7(2):165-170.

196

Perez-Trallero E, Marimon JM, Montes M, Orden B, de Pablos M. Clonal differences among erythromycin-resistant Streptococcus pyogenes in Spain. Emerg Infect Dis. 1999;5(2):235-240.

197

Brandt CM, Spellerberg B, Honscha M, Truong ND, Hoevener B, Lutticken R. Typing of Streptococcus pyogenes strains isolated from throat infections in the region of Aachen, Germany. Infection. 2001;29(3):163-165.

198

Vandamme P, Pot B, Falsen E, Kersters K, Devriese LA. Taxonomic study of Lancefield streptococcal groups C, G, and L (Streptococcus dysgalactiae) and proposal of S. dysgalactiae subsp. equisimilis subsp. nov. Int J Syst Bacteriol. 1996;46:774-781.

199

Collins CM, Kimura A, Bisno AL. Group G streptococcal M protein exhibits structural features analogous to those of class I M protein of group A streptococci. Infect Immun. 1992;60:3689-3696.

200

Courtney HS, Hasty DL, Li Y, Chiang HC, Thacker JL, Dale JB. Serum opacity factor is a major fibronectin-binding protein and a virulence determinant of M type 2 Streptococcus pyogenes. Mol Micro. 1999;32:89-98.

201

Humar D, Datta V, Bast DJ, Beall B, De Azavedo JCS, Nizet V. Streptolysin S and necrotising infections produced by group G streptococcus. Lancet. 2002;359:124-129.

202

Sachse S, Seidel P, Gerlach D, et al. Superantigen-like gene(s) in human pathogenic Streptococcus dysgalactiae, subsp. equisimilis: genomic localisation of the gene encoding streptococcal pyrogenic exotoxin G (speGdys). FEMS Immunol Med Microbiol. 2002;34:159167.

203

Espinosa LE, Li Z, Barreto DG, et al. M protein gene type distribution among group A streptococcal clinical isolates recovered in Mexico City, Mexico, from 1991 to 2000, and Durango, Mexico, from 1998 to 1999: overlap with type distribution within the United States. J Clin Microbiol. 2003;41:373-378.

204

Invasive Streptococcus pyogenes after allograft implantation - Colorado, 2003. MMWR. 2003;52:1173-1176.

205

Hoe NP, Vuopio-Varkila J, Vaara M, et al. Distribution of streptococcal inhibitor of complement variants in pharyngitis and invasive isolates in an epidemic of serotype M1 group A Streptococcus infection. J Infect Dis. 2001;183(4):633-639.

206

Hoe NP, Nakashima K, Lukomski S, et al. Rapid selection of complement-inhibiting protein variants in group A Streptococcus epidemic waves. Nat Med. 1999;5(8):924-929.

207

Hoe N, Nakashima K, Grigsby D, et al. Rapid molecular genetic subtyping of serotype M1 group A Streptococcus strains. Emerg Infect Dis. 1999;5(2):254-263.

208

Mejia LM, Stockbauer KE, Pan X, Cravioto A, Musser JM. Characterization of group A Streptococcus strains recovered from Mexican children with pharyngitis by automated DNA sequencing of virulence-related genes: unexpectedly large variation in the gene (sic) encoding a complement-inhibiting protein. J Clin Microbiol. 1997;35(12):3220-3224.

209

Stockbauer KE, Grigsby D, Pan X, et al. Hypervariability generated by natural selection in an extracellular complement-inhibiting protein of serotype M1 strains of group A Streptococcus. Proc Natl Acad Sci USA. 1998;95(6):3128-3133.

210

Jeng A, Sakota V, Li Z, Datta V, Beall B, Nizet V. Molecular genetic analysis of a group A Streptococcus operon encoding serum opacity factor and a novel fibronectin-binding protein, SfbX. J Bacteriol. 2003;185(4):1208-1217.

211

Chuang I, Van Beneden C, Beall B, Schuchat A. Population-based surveillance for postpartum invasive group A streptococcus infections, 1995-2000. Clin Infect Dis. 2002;35:665-670.

212

Pournaras S, Efstratiou A, Douboyas J, George RC, Tsakris A. Genetic relatedness of group A streptococci of the newly designated serotype M90 causing a food-borne outbreak and sporadic infections. Epidemiol Infect. 2000;125(2):299-301.

213

Mogielnicki NP, Schwartzman JD, Elliott JA. Perineal group A streptococcal disease in a pediatric practice. Pediatrics. 2000;106(2 Pt 1):276-281.

214

Rutishauser J, Funke G, Lutticken R, Ruef C. Streptococcal toxic shock syndrome in two patients infected by a colonized surgeon. Infection. 1999;27(4-5):259-260.

215

Chatellier S, Ihendyane N, Kansal RG, et al. Genetic relatedness and superantigen expression in group A streptococcus serotype M1 isolates from patients with severe and nonsevere invasive diseases. Infect Immun. 2000;68(6):3523-3534.

216

Descheemaeker P, Chapelle S, Lammens C, et al. Macrolide resistance and erythromycin resistance determinants among Belgian Streptococcus pyogenes and Streptococcus pneumoniae isolates. J Antimicrob Chemother. 2000;45(2):167-73.

56

©


Volume 27

MM11-A

217

Mascini EM, Jansze M, Schouls LM, Fluit AC, Verhoef J, van Dijk H. Invasive and noninvasive group A streptococcal isolates with different speA alleles in The Netherlands: genetic relatedness and production of pyrogenic exotoxins A and B. J Clin Microbiol. 1999;37(11):3469-3474.

218

Shah R, Green M, Barbadora KA, et al. Comparison of M and T type antigen testing to field inversion gel electrophoresis in the differentiation of strains of group A streptococcus. Pediatr Res. 1995;38(6):988-992.

219

Martin JM, Wald ER, Green M. Field inversion gel electrophoresis as a typing system for group A streptococcus. J Infect Dis. 1998;177(2):504-507.

220

Kiska DL, Thiede B, Caracciolo J, et al. Invasive group A streptococcal infections in North Carolina: epidemiology, clinical features, and genetic and serotype analysis of causative organisms. J Infect Dis. 1997;176(4):992-1000.

221

Marin DR, Single LA. Molecular epidemiology of group A streptococcus M type 1 infections. J Infect Dis. 1993;167:1112-1117.

222

Single LA, Martin DR. Clonal differences within M-types of the group A Streptococcus revealed by pulsed field gel electrophoresis. FEMS Microbiol Lett. 1992;70(1):85-89.

223

Feil EJ, Holmes EC, Bessen DE, et al. Recombination within natural populations of pathogenic bacteria: short-term empirical estimates and long-term phylogenetic consequences. Proc Natl Acad Sci USA. 2001;98:182-187.

224

Brito DA, Ramirez M, De Lencastre H. Serotyping Streptococcus pneumoniae by multiplex PCR. J Clin Microbiol. 2003;41:2378-2384.

225

Coffey TJ, Dowson CG, Daniels M, et al. Horizontal transfer of multiple penicillin-binding protein genes, and capsular biosynthetic genes, in natural populations of Streptococcus pneumoniae. Mol Microbiol. 1991;5(9):2255-2260.

226

Rubin LG, Rizvi A. PCR-based assays for detection of Streptococcus pneumoniae serotypes 3, 14, 19F and 23F in respiratory specimens. J Med Microbiol. 2004;53:595-602.

227

O’Halloran DM, Cafferkey MT. Multiplex PCR for identification of seven Streptococcus pneumoniae serotypes targeted by a 7-valent conjugate vaccine. J Clin Microbiol. 2005;43:3487-3490.

228

Pai R, Gertz RE, Beall B. Sequential multiplex PCR approach for determining capsular serotypes of Streptococcus pneumoniae. J Clin Microbiol. 2006;44:124-131.

229

Dowson CG, Hutchison A, Brannigan JA, et al. Horizontal transfer of penicillin-binding protein genes in penicillin-resistant clinical isolates of Streptococcus pneumoniae. Proc Natl Acad Sci USA. 1989;86(22):8842-8846.

230

Henriques Normark B, Kalin M, Ortqvist A, et al. Dynamics of penicillin-susceptible clones in invasive pneumococcal disease. J Infect Dis. 2001;184(7):861-869.

231

Figueiredo AM, Austrian R, Urbaskova P, Teixeira LA, Tomasz A. Novel penicillin-resistant clones of Streptococcus pneumoniae in the Czech Republic and in Slovakia. Microb Drug Resist. 1995;1(1):71-87.

232

McDougal LK, Facklam R, Reeves M, et al. Analysis of multiply antimicrobial-resistant isolates of Streptococcus pneumoniae from the United States. Antimicrob Agents Chemother. 1992;36(10):2176-2184.

233

Lefevre JC, Faucon G, Sicard AM, Gasc AM. DNA fingerprinting of Streptococcus pneumoniae strains by pulsed-field gel electrophoresis. J Clin Microbiol. 1993;31(10):2724-2728.

234

Zhou J, Enright MC, Sprat G. Identification of the major Spanish clones of penicillin-resistant pneumococci via the Internet using multilocus sequence typing. J Clin Microbiol. 2000;38:977-986.

235

Enright MC, Fenoll A, Griffiths D, Spratt BG. The three major Spanish clones of penicillin-resistant Streptococcus pneumoniae are the most common clones recovered from recent cases of meningitis in Spain. J Clin Microbiol. 1999;37:3210-3216.

236

Enright M, Spratt BG. A multilocus sequence typing scheme for Streptococcus pneumoniae: identification of clones associated with serious invasive disease. Microbiology. 1998;144,3049-3060.

237

Coffey TJ, Dowson CG, Daniels M, Spratt BG. Genetics and molecular biology of beta-lactam resistant pneumococci. Microb Drug Resist. 1995;1:2934.

238

Beall B, McEllistrem MC, Gertz RE Jr, et al, for the Active Bacterial Core Surveillance/Emerging Infections Program Network. Emergence of a novel penicillin resistant clone of serotype 35B Streptococcus pneumoniae in the United States. J Inf Dis. 2002;185:118-122.

239

Enright MC, Knox K, Griffiths D, Crook DWM, Spratt BG. Multilocus sequence typing of Streptococcus pneumoniae directly from cerebrospinal fluid. Eur J Clin Microbiol Infect Dis. 2000;19:627-630.

240

Shi ZY, Enright MC, Wilkinson P, Griffiths D, Spratt BG. Identification of three major clones of multiply antibiotic-resistant Streptococcus pneumoniae in Taiwanese hospitals using multilocus sequence typing. J Clin Microbiol. 1998;36:3514-3519.

©


57

Number 10

MM11-A

241

Coffey TJ, Daniels M, Enright MC, Spratt BG. Serotype 14 variants of the Spanish penicillin-resistant serotype 9V clone of Streptococcus pneumoniae arose by large recombinational replacements of the cpsA-pbp1a region. Microbiology. 1999;145:2023-2031.

242

Ip M, Lyon DJ, Yung RW, Tsang L, Cheng AF. Introduction of new clones of penicillin-nonsusceptible Streptococcus pneumoniae in Hong Kong. J Clin Microbiol. 2002;40(4):1522-1525.

243

Ho PL, Yung RW, Tsang DN, et al. Increasing resistance of Streptococcus pneumoniae to fluoroquinolones: results of a Hong Kong multicentre study in 2000. J Antimicrob Chemother. 2001;48(5):659-665.

244

Sa-Leao R, Tomasz A, de Lencastre H. Multilocus sequence typing of Streptococcus pneumoniae clones with unusual drug resistance patterns: genetic backgrounds and relatedness to other epidemic clones. J Infect Dis. 2001;184(9):1206-1210.

245

Llull D, Lopez R, Garcia E. Clonal origin of the type 37 Streptococcus pneumoniae. Microb Drug Resist. 2000;6(4):269-275.

246

McGee L, McDougal L, Zhou J, et al. Nomenclature of major antimicrobial-resistant clones of Streptococcus pneumoniae defined by the pneumococcal molecular epidemiology network. J Clin Microbiol. 2001;39:2565-2571.

247

Coffey TJ, Enright MC, Daniels M, et al. Recombinational exchanges at the capsular polysaccharide biosynthetic locus lead to frequent serotype changes among natural isolates of Streptococcus pneumoniae. Mol Microbiol. 1998;27:73-83.

248

Soares S, Kristinsson KG, Musser JM, Tomasz A. Evidence for the introduction of a multiresistant clone of serotype 6B Streptococcus pneumoniae from Spain to Iceland in the late 1980s. J Infect Dis. 1993;168(1):158-163.

249

Barnes DM, Whittier S, Gilligan PH, Soares S, Tomasz A, Henderson FW. Transmission of multidrug-resistant serotype 23F Streptococcus pneumoniae in group day care: evidence suggesting capsular transformation of the resistant strain in vivo. J Infect Dis. 1995;171(4):890896.

250

Lefevre JC, Bertrand MA, Faucon G. Molecular analysis by pulsed-field gel electrophoresis of penicillin-resistant Streptococcus pneumoniae from Toulouse, France. Eur J Clin Microbiol Infect Dis. 1995;14(6):491-497.

251

Gasc AM, Geslin P, Sicard AM. Relatedness of penicillin-resistant Streptococcus pneumoniae serogroup 9 strains from France and Spain. Microbiology. 1995;141 (Pt 3):623-627.

252

Pato MV, Carvalho CB, Tomasz A. Antibiotic susceptibility of Streptococcus pneumoniae isolates in Portugal. A multicenter study between 1989 and 1993. Microb Drug Resist. 1995;1(1):59-69.

253

Moreno F, Crisp C, Jorgensen JH, Patterson JE. The clinical and molecular epidemiology of bacteremias at a university hospital caused by pneumococci not susceptible to penicillin. J Infect Dis. 1995;172(2):427-432.

254

McDougal LK, Rasheed JK, Biddle JW, Tenover FC. Identification of multiple clones of extended-spectrum cephalosporin-resistant Streptococcus pneumoniae isolates in the United States. Antimicrob Agents Chemother. 1995;39(10):2282-2288.

255

Lefevre JC, Gasc AM, Lemozy J, Sicard AM, Faucon G. Pulsed field gel electrophoresis for molecular epidemiology of penicillin resistant Streptococcus pneumoniae strains. Pathol Biol (Paris). 1994;42(5):547-552.

256

Martin M, Turco JH, Zegans ME, et al. An outbreak of conjunctivitis due to atypical Streptococcus pneumoniae. N Engl J Med. 2003;348:1112-1121.

58

©


Volume 27

MM11-A

Clinical and Laboratory Standards Institute consensus procedures include an appeals process that is described in detail in Section 8 of the Administrative Procedures. For further information, contact CLSI or visit our website at www.clsi.org.

Summary of Delegate Comments and Committee Responses MM11-P: Molecular Methods for Bacterial Strain Typing; Proposed Guideline General 1.

This document is more a “best practice guideline” than a “standard” when compared to documents such as M7A7 or M2-A9 (the susceptibility testing documents).

•

This document is a guideline for best practices in bacterial strain typing, not a standard.

2.

I suggest stressing that the typing methods should be validated in the context they are going to be used and that the interpretation the epidemiological context should also be taken into account, along with all available background epidemiological information.

•

The issue of context is raised several times in the comments. This response will be cited for those subsequent comments as well.

•

The guidelines are explicitly directed at the interpretation of molecular typing data. That is, they are focused on the assessment, comparison, and characterization of genotypes. That process is not context dependent. It does require data about the specific species and the specific typing method. The application of genotype observations to the larger clinical, epidemiologic, or population genetic question at hand is context dependent. It is also beyond the scope of this document to cover all those possible applications. The importance of multiple inputs beyond genotype analysis is emphasized in the appropriate sections.

3.

This is a very ambitious document, and making guidelines for molecular subtyping that covers all methods and all bacteria in any epidemiological context is deemed to fail. This is not the case with this document, although it is somewhat confusing here and there. Maybe you should consider restricting it to outbreak surveillance and investigations and the problem of reinfection/persistence/new infection.

•

See the response to comment 2 above.

Foreword 4.

First paragraph, first sentence: I suggest stating that the answers to the two questions depend on the context in which they are asked (e.g., “same” in outbreak context may not be the same in reinfection/new infection or population genetic context).

•


Section 3.1, Definitions 5.

The Definitions section includes “accuracy.” Most standards now use and define the term “trueness.”

•

Because the term “trueness” is not used in the document, it is not included in the Definitions section.

6.

“Outbreak” is not the definition that is used later on. This definition includes multistrain and multisource outbreaks. Later on, the document addresses single-strain, single-source outbreaks.

•

See the response to comment 9 below.

©


59

Number 10

MM11-A

Section 3.3, Abbreviations and Acronyms 7.

If there is going to be an abbreviations/acronyms section, all acronyms and abbreviations that are used in the document should be included (e.g., SNP or define the term the first time it is used, especially if it is a section heading).

•

The Abbreviations and Acronyms section has been expanded to include all abbreviations and acronyms used in the document, and the meaning of the abbreviation or acronym is included with its first usage in the document.

Section 4.1, Sources of Genetic Variation 8.

Second paragraph: Point mutations do not occur at random.

•

The first sentence of the second paragraph has been revised to read: “Point mutations—base pair substitutions—can occur anywhere in the chromosome.”

Section 4.4.2, Outbreaks 9.

First bullet: This implies that you are talking about a single-source, single-strain outbreak. This is NOT the way outbreak is defined in Section 3.1.

•

This concern is addressed with the addition of the following paragraph: “Most outbreaks, particularly those in hospitals, represent a single strain of a single species. Food- and water-related outbreaks may comprise isolates representing two strains of two different species. The principles described here can readily be applied to each species separately. Polyclonal infection due to two different strains of the same species has been described and can generally be reliably resolved within the context of an individual patient. Conceivably, a single outbreak involving multiple persons could involve two different strains of the same species; analysis of this very rare situation is beyond the scope of this document.”

10. Third paragraph: We do see multistrain outbreaks among food- and water-borne infections. Water-borne outbreaks commonly involve more than one species and may involve even different viruses, parasites, and bacterial species at the same time. It is important always to keep the multistrain outbreaks in mind. If such an outbreak occurs and you don’t realize it, you are bound to draw the wrong conclusions. •

The comment confuses strains within a species with strains from different species. The second sentence clearly references outbreaks involving multiple different species (even kingdoms); this is directly addressed in the text. The concept of outbreaks involving two different strains of the same species is the situation characterized as “rare.” We would be pleased to consider publications indicating otherwise.

Section 4.4.3, Surveillance 11. First paragraph, second sentence: PulseNet also collects data on typhoid salmonellae. The protocol was developed for nontyphoidal Salmonella, but it works for S. typhi and PulseNet has a Typhi database. PulseNet now also has protocols and databases for Campylobacter and V. cholerae. •

The sentence has been revised to read: “PulseNet, coordinated by the Centers for Disease Control and Prevention (CDC, Atlanta, Georgia, USA), is a network of public health laboratories that share the results of PFGE typing of multiple pathogens, including E. coli O157:H7, typhoidal and nontyphoidal Salmonella serotypes, Listeria monocytogenes, Campylobacter, Vibrio cholera, and Shigella.”

12. First paragraph, second sentence: Enter-Net does NOT collect molecular subtyping information. It should be

removed from the document. If you want to mention molecular surveillance networks for food-borne infections, you could mention the international PulseNet networks: PulseNet Canada, Europe, Latin America, and AsiaPacific.

60

©


Volume 27 •

MM11-A

The Health Protection Agency of the United Kingdom, which hosts Enter-Net, also hosts typing databases that are readily found on the same website. The last sentence in the first paragraph has been revised to reflect this detail. The revised sentence reads: “The Health Protection Agency of the United Kingdom provides typing databases in support of the outbreak studies conducted in Europe by Enter-Net, which focuses on human enteric pathogens and HARMONY, which emphasizes antibiotic-resistant nosocomial pathogens (http://www.hpa.org.uk/cfi/bioinformatics/dbases.htm).”

13. Molecular subtyping is also used for microbiologic attribution analysis of food-borne pathogens (mapping the sources of the infections through comparison of the subtypes of the human isolates and isolates obtained from their food sources). This is a new discipline and the interpretation of the data is completely different from outbreak surveillance (i.e., the differentiation criteria are generally much more relaxed than with the latter). •

See comment 2 above about context.

Section 4.4.4, Population Genetics 14. Last sentence: Another advantage of the sequencing-based methods is that it is possible with information about mutation rates and prevalences of the different subtypes in a given population to calculate a probability that two different strains have the same origin. This is impossible with nonsequence-based methods like PFGE. The sequencing-based methods are therefore more powerful tools in forensic microbiology. •

This point has been addressed in Section 6.4.6.1 with the addition of the following sentence: “Sophisticated algorithms integrating sequencing data, mutation rates, and the prevalence of different subtypes within a population can be used to estimate the temporal as well as evolutionary distance between two isolates. This approach is particularly useful in forensic microbiology.”

Section 5.2, Discriminatory Power 15. I suggest replacing this formula with the one described in Hunter PR. Reproducibility and indices of discriminatory power of microbial typing methods. J Clin Microbiol. 1990;28(9):1903-1905. You could also bring both formulas. The one that you describe assumes that all subtypes are distinguishable from each other, which is not always the case. The formula from the Hunter paper takes this into account. •

The section has been revised to address this additional level of complexity.

16. Second paragraph, fourth sentence: This is why the formula quoted should not be used. If the other formula from the recommended reference is used, the reproducibility of the method may be taken into account, and Dvalue thus obtained is a much more reliable measure of the discrimination. •


Section 5.3.2, Characterizing Discriminatory Power 17. First Paragraph: How you should evaluate your typing method first of all depends on the context you want to use it in. This needs to be stated. If you are studying the epidemiology of a nosocomial pathogen in one hospital, you need to obtain the strains to validate that method from that hospital. If you want to study strains in one country, you should only use strains from that country. If you want to use it on the global scale, you should use strains from all over the world. If you want to solve a problem occurring here and now, the strains should all be recent and you should then be aware that if you want to use the method again ten years later or on a historical set of strains, the performance of the method may not be the same. •


18. Second paragraph: You may want to add here that the discriminatory power of a typing method should also be evaluated by studying the distribution of the subtypes in the population (e.g., PFGE doesn’t have a high D-value with Salmonella enteritidis because a couple of subtypes are displayed by more than 50% of strains in the population. However, if you have a cluster of strains with one of the rarer PFGE subtypes, PFGE will still be an ©


61

Number 10

MM11-A

excellent method to delineate such an outbreak. See, for example, Blanc DS, Hauser PM, Francioli P, Bille J. Molecular typing methods and their discriminatory power. Clin Microbiol Infect. 1998;4(2):61-63. •

This point has been added to Section 5.2.

Section 5.4, Assessment of Competency for Molecular Strain Typing 19. Third bullet: I am not quite sure what it is you want to do here. Is it evaluation of a laboratory’s proficiency to

subtype or is it a validation of a subtyping method in more laboratories? In PulseNet, we use three strains for the former purpose and at least 50 for the latter. What you need will depend on the clonality of the organism you study. •

All comments about clarity are appreciated and have been taken very seriously. But it is hard to see the source of confusion here. The section is titled “assessment of competency”; the first sentence is about laboratories; the third sentence says the isolates used should be a subset of those used for characterization of the method. We suggest 100 isolates for characterization, which is slightly more conservative than the 50 proposed by PulseNet. The subcommittee considered three isolates insufficient to assess competency.

Section 6.2, Pulsed-Field Gel Electrophoresis (PFGE) 20. Second paragraph, last sentence: PFGE also reflects the content of large plasmids in the cells. This is important to consider with some pathogens (e.g., E. coli and Salmonella). •

This point has been addressed in Section 6.2.4.2 with the addition of the following paragraph: “Impact of nonchromosomal elements – While the primary sources of variation in PFGE patterns are chromosomal changes (e.g., insertions, deletions, rearrangements, point mutations), in some instances extrachromosomal elements can impact the size and number of fragments. Examples include lysogenic phages in S. aureus (O’Neill GL, et al, 2001) and, less commonly, large (>75 kb) plasmids in gramnegative bacilli (Buchrieser C, et al, 1994). Because these elements can be acquired or lost or both in short time frames, the variation introduced may complicate an epidemiologic analysis.”

Section 6.2.1.1, Preparing and Digesting DNA 21. First paragraph, first sentence: change kilovolts to kilobases. •

The correction has been made.

22. Lysis incubation temperatures can be 37 °C, 50 °C, or 56 °C depending on the method. •

Additional lysis incubation temperatures with a supporting reference have been incorporated in the third paragraph of Section 6.2.1.1 as stated below: For gram-negative bacteria, effective lysis can typically be achieved by incubation at 56 °C in a solution comprising 0.5 to 1% N-lauryl sarcosine, 50 to 100 mM EDTA (pH 8.0), and 0.5 to 1.0 mg/mL proteinase K and RNase. Occasionally, gram-negative organisms may require preceding treatment at 37 °C with lysozyme solution for adequate lysis. Gram-positive microorganisms usually require such treatment with cell-wall digesting enzymes (e.g., lysozyme, lysostaphin for staphylococci; mutanolysin for Streptococcus pyogenes and enterococci).

23. Lysozyme is often added to the lysis solution for Gram negatives.

•

62

This point has been added.

©


Volume 27

MM11-A

24. The PK step is often done after the lysis step and can be at a different temperature than the lysis step (i.e., 37 °C lysis for MRSA and 50 °C PK treatment). •

The suggested addition with a supporting reference has been incorporated.

25. Washing can be in water or TE (change “and” to “or”). •

The subcommittee considered that the sequential washing procedure gave more consistent results. The sentence was revised for clarification.

26. Equilibration in restriction enzyme buffer for ten minutes is adequate. •

The subcommittee considered that the method proposed was more consistent.

27. Plugs can be melted or inserted or plugs can be stuck to the comb when the gel is poured. •

This is true. It is also quite technically demanding. The subcommittee considered this inappropriate for a broad general guideline.

28. The enzyme chart would be more useful if it described how many fragments of what size will be obtained (this appears in Tenover’s commentary in the Journal of Clinical Microbiology, 1995;33(9):2233-2239). •

Table 1 was revised to include: typical number of restriction fragments and typical fragment size range (kb).

29. The gel is stained in 0.5 µL/mL ethidium bromide in water or running buffer. •

Running buffer is not needed at this stage; further, making buffer involves time and expense.

30. We seldom have a need to destain. •

This is a decision appropriate for an advanced worker. It is more reliable to destain; it involves minimal effort and, thus, was deemed more appropriate for a general guideline.

31. Third paragraph: The document specifies “low-melting temperature agarose.” Most of the laboratories in Canada, ours included, use a specific type called “SeaKem Gold,” and we use this product because it gives far better resolution of the bands. I don’t know if the committee preparing the document did not want to mention specific reagents, but in this case, there are changes that may need to be made to the run conditions if the low melting point agarose is used. •

Citing proprietary reagents is against CLSI policy.

32. Add SeaKem Gold agarose to low melt agarose for plug preparation. •


33. Table 1: It is dangerous to insert a full and comprehensive table from a source unless all the information brought in it still is de facto standards. A CLSI protocol is by most scientists interpreted as a de facto standard. Not all the enzymes and their prioritization presented in the table are state of the art. The table should be deleted or amended to contain only the information that is still recognized as state of the art. In this case, you should add the PulseNet recommendations for the food-borne pathogens. •

©

This is a guideline, not a definitive protocol. PulseNet and other validated protocols have been cited.


63

Number 10

MM11-A

Section 6.2.3, Troubleshooting the Gel 34. Figure 1. It is not clear from the figure or the paragraphs below which lanes are controls. Not all lanes are

discussed in the paragraphs below. It would be easier to follow the discussion if the strips were in the same order as the discussion points. •

The lanes in Figure 1 have been reordered and the text following the figure has been revised for greater consistency and clarity.

35. This section is good, but needs to be clarified with regard to the figures. We could not find in the document what the lanes 1, 5, and 15 represented in Figure 1 (we think they could be controls, but this needs to be stated). •


36. Second paragraph: Incomplete digestion commonly results in faint differences that may be difficult/impossible to recognize as such (e.g., one or two extra bands). •

This point has been added.

37. Last paragraph: Don’t you want to mention DNA degradation in the gel resulting in smears of some strains? It may be prevented by using thiourea (Römling U, Tümmler B. Achieving 100% typeability of Pseudomonas aeruginosa by pulsed-field gel electrophoresis. J Clin Microbiol. 2000;38:464-465) in the running buffer or using HEPES as running buffer (Koort JM, Lukinmaa S, Rantala M, Unkila E, Siitonen A. Technical improvement to prevent DNA degradation of enteric pathogens in pulsed-field gel electrophoresis. J Clin Microbiol. 2002;40:3497-3498). •

The citations are appreciated and have been added.

38. The troubleshooting section could be more extensive and better presented. For instance, in PFGE, the buffer solutions used in the CHEF should be sterile. If they are not, the solution may grow Acinetobacter or Pseudomonas in the lines. These organisms produce DNases that cause all the lanes including the molecular weight markers to present as smears. It is important to differentiate between DNase contamination of the plugs vs. the chamber. They could also mention rinsing the chamber and lines or using 70% ethanol to remove the DNases. The importance of the correct electrical current and its relationship to buffer concentration are also not explained adequately. •

The problem of bacterial contamination of running buffer has been noted. In general, aseptic solutions should be satisfactory.

39. I would prefer that the troubleshooting section have a picture of the effect side by side with a picture of a lane that is fine, followed by the probable cause(s) and possible solutions. It is cumbersome to have it all in lengthy paragraphs. •

The suggested format is difficult to achieve. The text has been revised to make the connection to the figure clearer.

Section 6.2.4.1, Strengths 40. First bullet: I would suggest writing “almost any pathogen” not just “nosocomial pathogens.” •

The text has been revised by deleting the term “nosocomial.”

Section 6.2.4.2, Limitations 41. The cost of PFGE apparatus should be referenced re: US dollars and the year. It is not useful out of context. It is also rather a US mind-set. •

64

The sentence has been revised with the addition of (USD) following the cost figures. ©


Volume 27

MM11-A

Section 6.2.5.2, Surveillance 42. Second paragraph: The Web reference should be http://www.cdc.gov/pulsenet. •

The suggested revision has been made.

43. Third paragraph: As mentioned before, Enter-Net does NOT systematically collect and store molecular subtyping data; any mentioning of Enter-Net should be deleted from the document. •


Section 6.4.6.1, Strengths 44. You should be more careful with what you write here. What you state about true sequencing is probably correct. With the methods where you don’t do actual sequencing (e.g., MLVA or whole gene arrays), what you write is probably an over-statement. These methods are young and not fully explored. •

Gene array was noted in the discussion as a research method. The limitations of MLVA were also noted. The section is considered appropriate as an overall summary.

Section 7.2, Analyzing Electrophoresis Gels by Software 45. This section should be heavily reduced in size to only contain general statements (about gel quality, standards and reference strains, image resolution: What is a good gel?) and nothing you can read about in the software manuals. Most importantly, you need to state that all results generated by computer software need to be confirmed by visual inspection. •

The subcommittee, as well as other reviewers, feels the section is appropriate as written. The importance of visual inspection has been emphasized with the addition of the following statement: “Results obtained using computer analysis should always be confirmed by visual inspection of the restriction fragment patterns; in some instances, new gels in which selected isolates are run side by side may be needed.”

Section 7.2.1, Running the Gel and Obtaining the Images 46. First bullet: Lambda concatemers are not useful for computer analysis of PFGE gels due to problems of resolution of the larger fragments, and should not be used. •

The first bullet has been revised to read: “Every fifth lane should contain a bacterial strain with precisely sized restriction fragments to serve as molecular size standards and permit computerized “normalization” to correct for inevitable distortions and inconsistencies across the gel. Lambda concatemers may be less useful for this purpose because of limited resolution of the larger fragments.”

47. Second bullet: The molecular standard may also serve as a reproducibility standard. •

This is simply incorrect and, unfortunately, a common misconception. The fallacy in this reasoning is now emphasized with the addition of a paragraph that reads: “A common misconception is that the molecular size standards can also serve as the reproducibility standard. This is incorrect; it is essential that two independent strains with different restriction patterns be used for these two purposes. During the process of normalization, the software explicitly adjusts the entire gel image so as to minimize, if not eliminate, differences among the molecular size standards. Determining the residual variance due to technical factors across the gels requires analysis of lanes in each gel that are subject to—not the basis of—the computerized adjustments. The level of variation reported among the reproducibility lanes across a set of gels defines the minimum difference that is likely to have biologic significance. For example, if the reproducibility lanes have a similarity of 0.94, then sets of two or more isolates with a similarity of 0.95 are almost certainly indistinguishable. In unusual circumstances, visual inspection may indicate a single distinct change.”

©


65

Number 10

MM11-A

48. Second paragraph, last sentence: 1200 DPI is much too high a resolution for most applications; 400 to 500 dots from the top to the bottom of the gel is usually sufficient. •

Using a suboptimal resolution is a risk most appropriately made by the user after experience.

Section 8.1, Percent Identity 49. Sentence structure: a noun is missing in the first paragraph, third sentence. Add “section” or “region” after 160 bp or restructure the sentence. •

The sentence has been edited.

Section 9, Interpreting Variation in Molecular Typing 50. Second bullet: Many food-borne outbreaks have more than 100 cases. •

The bullet has been revised to note the potential difference between nosocomial and food-borne outbreaks.

Section 9.1, Categories of Genotypic Relatedness in Molecular Strain Typing 51. It needs to be emphasized here (and also in the beginning of the document) that interpretation of subtyping data does vary between different epidemiological contexts (species, its transmission, geography, time, outbreak vs. surveillance vs. attribution) and that all available background epidemiological information always needs to be taken into consideration (“the big picture”). •


Section 9.1.1, Indistinguishable 52. First paragraph, second sentence: This “no” is relative, so maybe you should weaken the statement by changing it to “no discernible.” •

It is relative for gel-based systems; this point has been emphasized.

53. Second paragraph: This is correct but the software settings (position tolerance and optimization) are just as or more important. You should never accept a computer-generated result without confirming it visually. •

The level of variation seen in the reproducibility isolates accounts for all these factors. The need for visual confirmation has been emphasized previously. (See the response to comment 45 above.)

Section 9.4, Translating Genotypic Relatedness Into Epidemiologic and Clinical Relatedness 54. First paragraph, second sentence: This is easy in theory but may be very difficult in practice. This needs to be stated. •

The text has been revised, although the subcommittee doubts that anyone who has read to page 35 thinks molecular typing is easy.

Section 10, Reporting Molecular Typing Results 55. First paragraph, second sentence: Note: An outbreak investigation is teamwork. During an outbreak, the interpretation may change. It is crucial that there is a continuous dialogue between the typing microbiologists and the epidemiologists doing the outbreak investigation. This needs to be stated. Face-to-face communication between the microbiologists and the epidemiologists in an outbreak investigation is crucial to the investigation of at least most food-borne outbreaks.

66

©


Volume 27 •

MM11-A

This section is directed toward producing a laboratory report of the genotypic observations. Comments about relating these observations to a larger question have been added at the end of the section.

Section 12, Examples of Molecular Typing of Bacterial Species 56. What is the purpose of this section? There are many other important pathogens with problems that are different from these two organisms but just as important. I would suggest removing everything that does not have general applicability or the whole section. •

©

The section is titled “Examples,” not “Definitive Review.”


67

Number 10

MM11-A

The Quality Management System Approach Clinical and Laboratory Standards Institute (CLSI) subscribes to a quality management system approach in the development of standards and guidelines, which facilitates project management; defines a document structure via a template; and provides a process to identify needed documents. The approach is based on the model presented in the most current edition of CLSI/NCCLS document HS1—A Quality Management System Model for Health Care. The quality management system approach applies a core set of “quality system essentials” (QSEs), basic to any organization, to all operations in any healthcare service’s path of workflow (i.e., operational aspects that define how a particular product or service is provided). The QSEs provide the framework for delivery of any type of product or service, serving as a manager’s guide. The quality system essentials (QSEs) are: Documents & Records Organization Personnel

Equipment Purchasing & Inventory Process Control

Information Management Occurrence Management Assessments—External and Internal

Process Improvement Customer Service Facilities & Safety

X

MM3

Facilities & Safety

Customer Service

Process Improvement

Assessments —External and Internal

Occurrence Management

Information Management

Process Control

Purchasing & Inventory

Equipment

Personnel

Organization

Documents & Records

MM11-A addresses the quality system essentials (QSEs) indicated by an “X.” For a description of the other documents listed in the grid, please refer to the Related CLSI/NCCLS Publications section on the following page.

M29

M29 MM3 MM9

Adapted from CLSI/NCCLS document HS1—A Quality Management System Model for Health Care.

Path of Workflow A path of workflow is the description of the necessary steps to deliver the particular product or service that the organization or entity provides. For example, CLSI/NCCLS document GP26⎯Application of a Quality Management System Model for Laboratory Services defines a clinical laboratory path of workflow, which consists of three sequential processes: preexamination, examination, and postexamination. All clinical laboratories follow these processes to deliver the laboratory’s services, namely quality laboratory information. MM11-A addresses the clinical laboratory path of workflow steps indicated by an “X.” For a description of the other Clinical and Laboratory Standards Institute documents listed in the grid, please refer to the Related CLSI/NCCLS Publications section on the following page.

Results review and follow-up

Interpretation

MM3 MM9

MM3 MM9

MM3 MM9

MM3 MM9

MM9

MM3 MM9

Sample management

Examination

MM3 MM9

Results reporting and archiving

Sample receipt/processing

Postexamination

Sample transport

Examination

Sample collection

Examination ordering

Preexamination

MM9

Adapted from CLSI/NCCLS document HS1—A Quality Management System Model for Health Care.

68

©


Volume 27

MM11-A

Related CLSI/NCCLS Publications∗ M29-A3

Protection of Laboratory Workers From Occupationally Acquired Infections; Approved Guideline— Third Edition (2005). Based on US regulations, this document provides guidance on the risk of transmission of infectious agents by aerosols, droplets, blood, and body substances in a laboratory setting; specific precautions for preventing the laboratory transmission of microbial infection from laboratory instruments and materials; and recommendations for the management of exposure to infectious agents.

MM3-A2

Molecular Diagnostic Methods for Infectious Diseases; Approved Guideline—Second Edition (2006). This guideline addresses topics relating to clinical applications, amplified and nonamplified nucleic acid methods, selection and qualification of nucleic acid sequences, establishment and evaluation of test performance characteristics, inhibitors, and interfering substances, controlling false-positive reactions, reporting and interpretation of results, quality assurance, regulatory issues, and recommendations for manufacturers and clinical laboratories.

MM9-A

Nucleic Acid Sequencing Methods in Diagnostic Laboratory Medicine; Approved Guideline (2004). This document addresses automated, PCR-based, dideoxyterminator, and primer extension sequencing done on gel- or capillary-based sequencers. Topics covered include specimen collection and handling; isolation of nucleic acid; amplification and sequencing of nucleic acids; interpretation and reporting results; and quality control/assessment considerations as appropriate.

∗

Proposed-level documents are being advanced through the Clinical and Laboratory Standards Institute consensus process; therefore, readers should refer to the most current editions. ©


69

Active Membership (as of 1 April 2007) Sustaining Members Abbott Laboratories American Association for Clinical Chemistry AstraZeneca Pharmaceuticals Bayer Corporation BD Beckman Coulter, Inc. bioMérieux, Inc. CLMA College of American Pathologists GlaxoSmithKline Ortho-Clinical Diagnostics, Inc. Pfizer Inc Roche Diagnostics, Inc. Professional Members AABB American Academy of Family Physicians American Association for Clinical Chemistry American Association for Laboratory Accreditation American Association for Respiratory Care American Chemical Society American College of Medical Genetics American Medical Technologists American Society for Clinical Laboratory Science American Society for Microbiology American Type Culture Collection, Inc. ASCP Associazione Microbiologi Clinici Italiani (AMCLI) British Society for Antimicrobial Chemotherapy Canadian Society for Medical Laboratory Science - Société Canadienne de Science de Laboratoire Médical COLA College of American Pathologists College of Medical Laboratory Technologists of Ontario College of Physicians and Surgeons of Saskatchewan Elkin Simson Consulting Services ESCMID Family Health International Hong Kong Accreditation Service Innovation and Technology Commission Italian Society of Clinical Biochemistry and Clinical Molecular Biology International Federation of Biomedical Laboratory Science International Federation of Clinical Chemistry JCCLS Joint Commission on Accreditation of Healthcare Organizations National Society for Histotechnology, Inc. Ontario Medical Association Quality Management Program-Laboratory Service RCPA Quality Assurance Programs PTY Limited SDS Pathology SIMeL Sociedad Espanola de Bioquimica Clinica y Patologia Molecular Sociedade Brasileira de Analises Clinicas Sociedade Brasileira de Patologia Clinica Turkish Society of Microbiology Washington G2 Reports World Health Organization Government Members Association of Public Health Laboratories BC Centre for Disease Control Caribbean Epidemiology Centre Centers for Disease Control and Prevention Centers for Disease Control and Prevention – Tanzania Centers for Medicare & Medicaid Services Centers for Medicare & Medicaid Services/CLIA Program Chinese Committee for Clinical Laboratory Standards

Department of Veterans Affairs FDA Center for Devices and Radiological Health FDA Center for Veterinary Medicine Health Canada Massachusetts Department of Public Health Laboratories Ministry of Health and Social Welfare - Tanzania National Center of Infectious and Parasitic Diseases (Bulgaria) National Health Laboratory Service (South Africa) National Institute of Standards and Technology National Pathology Accreditation Advisory Council (Australia) New York State Department of Health Ontario Ministry of Health Pennsylvania Dept. of Health Saskatchewan Health-Provincial Laboratory Scientific Institute of Public Health; Belgium Ministry of Social Affairs, Public Health and the Environment University of Iowa, Hygienic Lab Industry Members AB Biodisk Abbott Diabetes Care Abbott Laboratories Abbott Molecular Inc. Abbott Point of Care Inc. Access Genetics ACM Medical Technologies, Inc. Acupath AdvaMed Advanced Liquid Logic Advancis Pharmaceutical Corporation Affymetrix, Inc. Agilent Technologies Ammirati Regulatory Consulting Anapharm, Inc. Anna Longwell, PC Aptium Oncology ARK Diagnostics, Inc. Arpida Ltd A/S ROSCO AstraZeneca Pharmaceuticals Aviir, Inc. Axis-Shield POC AS Bayer Corporation – Tarrytown, NY Bayer Corporation – West Haven, CT Bayer HealthCare, LLC, Diagnostics Div. – Elkhart, IN BD BD Biosciences – San Jose, CA BD Diabetes Care BD Diagnostic Systems BD Vacutainer Systems Beckman Coulter, Inc. Beth Goldstein Consultant (PA) Bioanalyse, Ltd. Bio-Development S.r.l. Bio-Inova Life Sciences International Biomedia Laboratories SDN BHD bioMérieux (NC) bioMérieux, Inc. (MO) Bio-Rad Laboratories, Inc. Bio-Rad Laboratories, Inc. – France Bio-Rad Laboratories, Inc. – Irvine, CA Bio-Rad Laboratories, Inc. – Plano, TX Black Coast Corporation – Health Care Systems Consulting Blaine Healthcare Associates, Inc. Center for Measurement Standards/ITRI Cepheid Chen & Chen, LLC Comprehensive Cytometric Consulting Control Lab Copan Diagnostics Inc. Cosmetic Ingredient Review Cubist Pharmaceuticals Cumbre Inc. Dade Behring Inc. – Cupertino, CA Dade Behring Inc. – Deerfield, IL Dade Behring Inc. – Glasgow, DE Dade Behring Inc. – Marburg, Germany Dade Behring Inc. – Sacramento, CA David G. Rhoads Associates, Inc. Decode Genetics, Inc. Diagnostic Products Corporation

Diagnostica Stago Digene Corporation Eiken Chemical Company, Ltd. Elanco Animal Health Electa Lab s.r.l. Enterprise Analysis Corporation Eomix, Inc. Eurofins Medinet FasTraQ Inc. (NV) Future Diagnostics B.V. Gavron Group, Inc. Gen-Probe Genaco Biomedical Products, Inc. Genomic Health, Inc. Gentris Corporation Genzyme Clinical Specialty Laboratory Genzyme Diagnostics GlaxoSmithKline Gluco Tec, Inc. GluMetrics, Inc. Greiner Bio-One Inc. HistoGenex N.V. Immunicon Corporation Instrumentation Laboratory IT for Small Business Janssen Ortho-McNeil Pharmaceutical Japan Assn. of Clinical Reagents Industries Johnson & Johnson Pharmaceutical Research and Development, L.L.C. K.C.J. Enterprises LabNow, Inc. Laboratory Specialists, Inc. LifeScan, Inc. (a Johnson & Johnson Company) Maine Standards Company, LLC Medical Device Consultants, Inc. Merck & Company, Inc. Micromyx, LLC MicroPhage MultiPhase Solutions, Inc. Mygene International, Inc. Nanogen, Point-of-Care Diagnostics Div. NeED Pharmaceuticals Srl Nippon Becton Dickinson Co., Ltd. Nissui Pharmaceutical Co., Ltd. NovaBiotics (Aberdeen, UK) Novartis Institutes for Biomedical Research Nucryst Pharmaceuticals Olympus America, Inc. Optimer Pharmaceuticals, Inc. Orion Genomics, LLC Ortho-Clinical Diagnostics, Inc. (Rochester, NY) Oxonica (UK) Panaceapharma Pharmaceuticals Paratek Pharmaceuticals Pathology Services Inc. PathWork Informatics Pfizer Animal Health Pfizer Inc Pfizer Italia Srl Phadia AB Powers Consulting Services PPD, Inc. Primera Biosystems, Inc. QSE Consulting Radiometer America, Inc. Radiometer Medical A/S Reliance Life Sciences Replidyne Rib-X Pharmaceuticals Roche Diagnostics GmbH Roche Diagnostics, Inc. Roche Laboratories Roche Molecular Systems Sanofi Pasteur Sarstedt, Inc. Schering Corporation Seneca Medical, Inc. Sequenom, Inc. SFBC Anapharm Sphere Medical Holding Streck Laboratories, Inc. Sysmex America, Inc. (Long Grove, IL) Sysmex Corporation (Japan) Tethys Bioscience, Inc. The Clinical Microbiology Institute TheraDoc Therapeutic Monitoring Services, LLC Theravance Inc. Third Wave Technologies, Inc. Thrombodyne, Inc. Transasia Bio-Medicals Limited Trek Diagnostic Systems, Inc. TrimGen Corporation Watin-Biolife Diagnostics and Medicals Wyeth Research XDX, Inc. YD Consultant

Trade Associations AdvaMed Japan Association of Clinical Reagents Industries (Tokyo, Japan) Associate Active Members 3rd Medical Group (AK) 48th Medical Group/MDSS (APO, AE) 59th MDW/859th MDTS/MTL (TX) Aberdeen Royal Infirmary (Scotland) Academisch Ziekenhuis -VUB (Belgium) Acibadem Labmed Clinical Laboratory (Turkey) ACL Laboratories (IL) ACL Laboratories (WI) Akron’s Children’s Hospital (OH) Alameda County Medical Center (CA) Albany Medical Center Hospital (NY) Albemarle Hospital (NC) Alfred I. du Pont Hospital for Children (DE) All Children’s Hospital (FL) Allegheny General Hospital (PA) Allina Labs (MN and WI) Alton Memorial Hospital (MN) American Hospital Dubai (UAE) American University of Beirut Medical Center (NY) Arkansas Methodist Medical Center (AR) Arnett Clinic, LLC (IN) Asante Health System (OR) Aspirus Wausau Hospital (WI) Associated Regional & University Pathologists (UT) Atlantic Health System (NJ) Augusta Medical Center (VA) AZ Sint-Jan (Belgium) Azienda Ospedale Di Lecco (Italy) Balfour Hospital (Scotland) Baptist Hospital for Women (TN) Barbados Reference Laboratory (Barbados) Bassett Army Community Hospital (AK) BayCare Health System (FL) Baystate Medical Center (MA) BC Biomedical Laboratories (Surrey, BC, Canada) Borders General Hospital (Scotland) Boulder Community Hospital (CO) British Columbia Cancer Agency – Vancouver Cancer Center (BC, Canada) Broward General Medical Center (FL) Cadham Provincial Laboratory – MB Health (Canada) Calgary Laboratory Services (Calgary, AB, Canada) Canterbury Health Laboratories (New Zealand) Cape Breton Healthcare Complex (Canada) Cape Fear Valley Medical Center Laboratory (NC) Capital Health - Regional Laboratory Services (Canada) Capital Health System Fuld Campus (NJ) Capital Health System Mercer Campus (NJ) Catholic Health Initiatives (KY) CDC/HIV (APO, AP) CDPH (CO) Central Baptist Hospital (KY) Central Kansas Medical Center (KS) Central Texas Veterans Health Care System Centralized Laboratory Services (NY) Centura Laboratory (CO) Chang Gung Memorial Hospital (Taiwan) Chesapeake General Hospital (VA) Chester County Hospital (PA) Children’s Healthcare of Atlanta (GA) Childrens Hospital of Wisconsin (WI) Christiana Care Health Services (DE) CHUM Hopital Saint-Luc (Canada) City of Hope National Medical Center (CA)

Clarian Health – Clarian Pathology Laboratory (IN) Cleveland Clinic Health System Eastern Region (OH) Clovis Community Hospital (CA) CLSI Laboratories (PA) Commonwealth of Kentucky Community Care 5 (OH) Community Hospital (IN) Community Hospital (OH) Connolly Hospital (Ireland) Covance CLS (IN) Creighton Medical Laboratories (NE) Creighton University Medical Center (NE) Crosshouse Hospital (Scotland) Cumberland Regional Health Care Centre (Canada) Danish Institute for Food and Veterinary Research (Denmark) Darwin Library NT Territory Health Services (Australia) David Grant Medical Center (CA) Daviess Community Hospital (IN) Dekalb Memorial Hospital (IN) DeWitt Healthcare Network (USA Meddac) (VA) DHHS NC State Lab of Public Health (NC) Diagnofirm Med Labs Diagnostic Laboratory Services, Inc. (HI) Diagnostic Services of Manitoba (Canada) Dianon Systems/Lab Corp. (OK) Dr. Everette Chalmers Regional Hospital (NB) Driscoll Children’s Hospital (TX) DSI of Bucks County (PA) DUHS Clinical Laboratories (NC) Dumfries and Galloway Royal Infirmary (Scotland) Edinburgh Royal Infirmary (Scotland) EMH Regional Medical Center (OH) Emory University Hospital (GA) Evangelical Community Hospital (PA) Evanston Hospital (IL) Exeter Hospital (NH) Federal Medical Center (MN) Firelands Regional Medical Center (OH) First Health of the Carolinas Moore Regional Hospital (NC) Fisher-Titus Memorial Hospital (OH) Flaget Memorial Hospital (KY) Fleury S.A. (Brazil) Forum Health Northside Medical Center (OH) Fox Chase Cancer Center (PA) Gamma Dynacare Medical Laboratories (Ontario, Canada) Garden City Hospital (MI) Geisinger Medical Center (Danville, PA) Geisinger South Wilkes Barre Laboratory (PA) Geisinger Wyoming Valley Medical Center (Wilkes-Barre, PA) Genesis Healthcare System (OH) Gilbert Bain Hospital (Scotland) Glasgow Royal Infirmary (Scotland) Good Samaritan Hospital (NE) Hagerstown Medical Laboratory (MD) Hamad Medical Corporation (Qatar) Harris Methodist Fort Worth (TX) Hartford Hospital (CT) Health Network Lab (PA) Health Partners Laboratories (VA) Health Waikato (New Zealand) Heidelberg Army Hospital (APO, AE) High Desert Health System (CA) Hoag Memorial Hospital Presbyterian (CA) Holy Cross Hospital (MD) Holy Family Medical Center (WI) Holy Spirit Hospital (PA) Holzer Medical Center (Gallipolis, OH) Holzer Medical Center (Jackson, OH) Hopital Cite de La Sante De Laval (Canada)

Hôpital Maisonneuve - Rosemont (Montreal, Canada) Hôpital Sacré-Coeur de Montreal (Quebec, Canada) Hôpital Sainte - Justine (Quebec) Hopital Santa Cabrini Ospedale (Canada) Hospital Albert Einstein (Brazil) Hospital de Dirino Espirito Santa (Portugal) Hospital De Sousa Martins (Guarda) (Portugal) Hospital for Sick Children (Toronto, ON, Canada) Hôtel Dieu Grace Hospital Library (Windsor, ON, Canada) Humility of Mary Health Partners (OH) Hunterdon Medical Center (NJ) Icon Laboratories (Ireland) IGate Clinical Research Intl., Pvt., LTD (India) Indiana University - Chlamydia Laboratory (IN) Inova Fairfax Hospital (VA) Institut fur Stand. und Dok. im Med. Lab. (Germany) Institut National de Santé Publique du Quebec Centre de Doc. – INSPQ (Canada) Integrated Regional Laboratories South Florida (FL) International Health Management Associates, Inc. (IL) Island Hospital (WA) Jackson Hospital & Clinic, Inc. (AL) Jackson Purchase Medical Center (KY) Jacobi Medical Center (NY) John C. Lincoln Hospital (AZ) John F. Kennedy Medical Center (NJ) John H. Stroger, Jr. Hospital of Cook County (IL) John Muir Medical Center (CA) Johns Hopkins Medical Institutions (MD) Johns Hopkins University (MD) Kadlec Medical Center (WA) Kaiser Permanente (CA) Kaiser Permanente (MD) Kangnam St. Mary’s Hospital (Korea) Kenora-Rainy River Reg. Lab. Program (Canada) King Fahad Medical City (Saudi Arabia) King Faisal Specialist Hospital (MD) Kosciusko Laboratory (IN) LabCorp (NC) Laboratory Alliance of Central New York (NY) LabPlus Auckland Healthcare Services Limited (New Zealand) Lakeland Regional Medical Center (FL) Landstuhl Regional Medical Center (APO, AE) Langlade Memorial Hospital (WI) Legacy Laboratory Services (OR) Lewis-Gale Medical Center (VA) L’Hotel-Dieu de Quebec (Quebec, Canada) Licking Memorial Hospital (OH) LifeBridge Health Sinai Hospital (MD) Lions Gate Hospital (BC, Canada) Long Beach Memorial Medical Center (CA) Long Island Jewish Medical Center (NY) Los Angeles County Public Health Lab. (CA) Louis A Johnson Medical Center (WV) Madison Parish Hospital (LA) Magruder Memorial Hospital (OH) Main Line Clinical Laboratories, Inc. (PA) Malmo University Hospital (Sweden) Manipal Acunova (India) Marshfield Clinic (WI) Martin Luther King/Drew Medical Center (CA) Martin Memorial Health Systems (FL) Marymount Medical Center (KY)

Massachusetts General Hospital (Microbiology Laboratory) Maxwell Air Force Base (AL) MDS Metro Laboratory Services (Burnaby, BC, Canada) Mease Countryside Hospital (FL) Mease Dunedin Hospital (FL) Medical Centre Ljubljana (Slovenia) Medical College of Virginia Hospital (VA) Medical Univ. of South Carolina (SC) Memorial Health Center, Inc. (WI) Memorial Hospital (OH) Memorial Hospital Miramar (FL) Memorial Hospital Pembroke (FL) Memorial Hospital West (FL) Memorial Regional Hospital (FL) Mercy Medical Center (CO) Mercy Medical Center (ID) Mercy Medical Center (OR) Methodist Hospital (MN) Methodist Hospital (TX) Methodist Hospital Pathology (NE) Monklands Hospital (Scotland) Montreal General Hospital (Canada) Morgan Hospital & Medical Center (IN) Morton Plant Hospital (FL) Mount Sinai Hospital (NY) Mountainside Hospital (NJ) National Serology Reference Laboratory, Australia Naval Hospital Cherry Point Laboratory (NC) Naval Hospital Oak Harbor (WA) NB Department of Health & Wellness (New Brunswick, Canada) The Nebraska Medical Center New England Fertility Institute (CT) New Lexington Clinic (KY) New York City Department of Health and Mental Hygiene (NY) The New York Hospital Medical Center of Queens (NY) New York-Presbyterian Hospital (NY) New York University Medical Center (NY) Ninewells Hospital (Scotland) North Bay Hospital (FL) North Mississippi Medical Center (MS) North Shore Hospital Laboratory (Auckland, New Zealand) North Shore-Long Island Jewish Health System Laboratory (NY) Northern Plains Laboratory (ND) Northridge Hospital Medical Center (CA) Northwest Texas Hospital (TX) Northwestern Memorial Hospital (IL) Norton Healthcare (KY) Ochsner Clinic Foundation (LA) Oklahoma Heart Hospital, LLC (OK) Onze Lieve Vrouw Ziekenhuis (Belgium) Orlando Regional Healthcare System (FL) Our Lady of the Way Hospital (KY) Overlook Hospital (NJ) Palisades Medical Center (NJ) Pathologists Associated (IN) Pathology and Cytology Laboratories, Inc. (KY) Pathology Associates Medical Laboratories (WA) PathWest (Australia) PCA Southeast (TN) Pediatrix Screening Inc. (PA) Pennsylvania Hospital (PA) Penrose St. Francis Health Services (CO) Perry County Memorial Hospital (IN) Physicians Reference Laboratory (KS) Pitt County Memorial Hospital (NC) Powell River General Hospital (BC, Canada) PPD (KY) Prince George Medical Lab (Prince George, BC) Providence Health Care (Canada) Provincial Health Services Authority (Vancouver, BC, Canada) Provincial Laboratory for Public Health (Edmonton, AB, Canada) Queen Elizabeth Hospital (Canada) Queensland Health Pathology Services (Australia) Quest Diagnostics, Inc Quest Diagnostics, Inc (San Juan Capistrano, CA) Quintiles Laboratories, Ltd. (GA)

Raigmore Hospital (UK) Redington-Fairview General Hospital (ME) Régie Régionale Dela Santé Beaséjour (Canada) Regional Health Authority - Central Manitoba Inc (Canada) Regional Health Authority Four (RHA4) (Canada) Regions Hospital (MN) Research Medical Center (MO) Richmond General Hospital (BC, Canada) Riverside Methodist Hospital (OH) Riverview Hospital (WI) Riyadh Armed Forces Hospital (Riyadh, Saudi Arabia) Robert Wood Johnson University Hospital (NJ) Roxborough Memorial Hospital (PA) Royal Alexandra Hospital (Scotland) Sahlgrenska Universitetssjukhuset (Sweden) Saint Elizabeth Regional Medical Center (NE) Saint Francis Hospital & Medical Center (CT) St. Agnes Healthcare (MD) St. Anthony Hospital Central Laboratory (CO) St. Anthony’s Hospital (FL) St. Barnabas Medical Center (NJ) St. Christopher’s Hospital for Children (PA) St. Eustache Hospital (Quebec, Canada) St. Francis Medical Center (MN) St. John Hospital and Medical Center (MI) St. John’s Hospital & Health Ctr. (CA) St. Joseph Medical Center (MD) St. Joseph Mercy (WI) St. Joseph Mercy Hospital (MI) St. Joseph’s Hospital (FL) St. Joseph's Medical Center (CA) St. Joseph's Regional Medical Center (NJ) St. Jude Children’s Research Hospital (TN) St. Louis Children’s Hospital (MO) St. Luke’s Hospital (PA) St. Margaret Memorial Hospital (PA) St. Mary Corwin Regional Medical Center Laboratory (CO) St. Mary Medical Center (CA) St. Mary’s Health Center (MO) St Mary’s Healthcare (SD) St. Mary’s Hospital (BC, Canada) St. Mary’s Medical Center (IN) St. Rose Dominican Hospitals (NV) St. Thomas More Hospital (CO) San Antonio Community Hospital (TX) San Francisco General HospitalUniversity of California San Francisco (CA) Santa Monica Hospital Med. Ctr. (CA) Seoul Clinical Laboratories (Korea) Shands at the University of Florida Shape Healthcare Clinic (APO, AE) Sheik Kalifa Medical City (UAE) Shore Memorial Hospital (NJ) SJRMC Plymouth Laboratory (IN) Sonora Quest JV (AZ) South Bend Medical Foundation (IN) South Dakota State Health Laboratory (SD) South Florida Baptist Hospital (FL) South Texas Laboratory (TX) Southern Health Care Network (Australia) Southwest Nova District Health Authority (Canada) Specialty Laboratories, Inc. (CA) Squamish General Hospital (BC, Canada) Starke Memorial Hospital Laboratory (IN) State of Connecticut Department of Public Health (CT) State of Washington Public Health Labs Steele Memorial Hospital (ID) Stirling Royal Infirmary (Scotland)

Stony Brook University Hospital (NY) Stormont-Vail Regional Medical Center (KS) Stratford General Hospital (Canada) Sunnybrook & Women’s College Health Sciences Centre (Toronto, Ontario) Sunnybrook Health Science Center (ON, Canada) Swedish Medical Center (CO) Sydney South West Pathology Service (Australia) Taipei Veterans General Hospital (Taiwan) Taiwan Society of Laboratory Medicine Tampa General Hospital (FL) Temple Univ. Hospital - Parkinson Pav. (PA) Texas Department of State Health Services (TX) The Bermuda Hospitals (Bermuda) The Community Hospital (OH) The Hospital for Sick Children (Canada) The Nebraska Medical Center (NB) The New York Hospital Medical Center of Queens (NY) The Ottawa Hospital (Canada) The Permanente Medical Group (CA) The Public Health Laboratory, H47 (AR) The Toledo Hospital (OH) The University of Texas Medical Branch (TX) The Wisconsin Heart Hospital (WI)

Thomason Hospital (TX) Timmins and District Hospital (Canada) Touro Infirmary (LA) TPMG Inc. (CA) Tri-Cities Laboratory (WA) Tripler Army Medical Center (HI) Tuen Mun Hospital (Hong Kong) Tufts New England Medical Center Hospital (MA) Tuttle Army Health Clinic (GA) UBC Hospital (BC, Canada) UCLA Immunogenetics Lab (CA) UCLA Medical Center (CA) UCSD Medical Center (CA) UCSF Medical Center China Basin (CA) UMC of Southern Nevada (NV) UNC Hospitals (NC) Union Clinical Laboratory (Taiwan) United Clinical Laboratories (IA) Unity HealthCare (IA) Universita Campus Bio-Medico (Italy) Universitair Ziekenhuis Antwerpen (Belgium) University of Colorado Health Sciences Center (CO) University of Colorado Hospital University of Illinois Medical Center (IL) University of Maryland Medical System University of Medicine & Dentistry, NJ University Hosp. (NJ) University of MN Medical Center Fairview

University of Missouri Hospital (MO) University of MS Medical Center (MS) University of the Ryukyus (Japan) University of Virginia Medical Center University of Washington U.S. Army Health Clinic – Vicenza (APO) US LABS, Inc. (CA) U.S.A. Meddac (Pathology Division) (MO) U.T. Health Center (TX) UZ-KUL Medical Center (Belgium) VA (Asheville) Medical Center (NC) VA (Bay Pines) Medical Center (FL) VA (Cincinnati) Medical Center (OH) VA (Colmery-O’Neil) Medical Center (KS) VA (Des Moines) Central Iowa Healthcare Systems (IA) VA (Fargo) Medical Center (ND) VA (Fayetteville) Medical Center (AR) VA (Iowa City) Medical Center (IA) VA (Lincoln) Nebraska Western Iowa Healthcare System (NE) VA New Jersey Health Care System (NJ) VA (Phoenix) Medical Center (AZ) VA (San Diego) Medical Center (CA) VA (Tucson) Medical Center (AZ) Valley Health (VA) Vancouver Hospital and Health Sciences Center (BC, Canada) Virga Jessezieukenhuis (Belgium) Virginia Regional Medical Center (MN)

OFFICERS Robert L. Habig, PhD, President Abbott Laboratories Gerald A. Hoeltge, MD, President-Elect The Cleveland Clinic Foundation Wayne Brinster, Secretary BD W. Gregory Miller, PhD, Treasurer Virginia Commonwealth University Thomas L. Hearn, PhD, Immediate Past President Centers for Disease Control and Prevention Glen Fine, MS, MBA, Executive Vice President

ViroMed Laboratories (LabCorp) (MN) Waianae Coast Comprehensive Health Center (HI) Walter Reed Army Medical Center (DC) Warren Hospital (NJ) Washington Hospital Center (DC) Waterbury Hospital (CT) Waterford Regional Hospital (Ireland) Wellstar Health Systems (GA) West China Second University Hospital, Sichuan University (P.R. China) West Valley Medical Center Laboratory (ID) Westchester Medical Center (NY) Western Isles Hospital (Scotland) Wheaton Franciscan & Midwest Clinical Laboratories (WI) Wheeling Hospital (WV) Whistler Health Care Centre (BC, Canada) Whitehorse General Hospital (Canada) William Beaumont Hospital (MI) Winchester Hospital (MA) Winn Army Community Hospital (GA) Womack Army Medical Center (NC) Women’s Health Laboratory (TX) Woodlawn Hospital (IN) York Hospital (PA) Yorkshire Hospital (Scotland)

BOARD OF DIRECTORS Susan Blonshine, RRT, RPFT, FAARC TechEd

Gary L. Myers, PhD Centers for Disease Control and Prevention

Maria Carballo Health Canada

Valerie Ng, PhD, MD Alameda County Medical Center/ Highland General Hospital

Russel K. Enns, PhD Cepheid Mary Lou Gantzer, PhD Dade Behring Inc. Lillian J. Gill, DPA FDA Center for Devices and Radiological Health Prof. Naotaka Hamasaki, MD, PhD Nagasaki International University

Janet K.A. Nicholson, PhD Centers for Disease Control and Prevention Klaus E. Stinshoff, Dr.rer.nat. Digene (Switzerland) Sàrl James A. Thomas ASTM International

940 West Valley Road Suite 1400 Wayne, PA 19087 USA PHONE 610.688.0100 FAX 610.688.0700 E-MAIL: [email protected] WEBSITE: www.clsi.org ISBN 1-56238-634-4