USF Libraries
USF Digital Collections

Evaluation and implementation of a molecular-based protocol for the identification of enteroviruses at the Florida Depar...


Material Information

Evaluation and implementation of a molecular-based protocol for the identification of enteroviruses at the Florida Department of Health - Tampa Laboratory
Physical Description:
Smith, Matthew Adams
University of South Florida
Place of Publication:
Tampa, Fla.
Publication Date:


Subjects / Keywords:
Enterovirus B, Human   ( mesh )
Epidemiology, Molecular   ( mesh )
Polymerase Chain Reaction   ( mesh )
Transcription, Genetic   ( mesh )
Laboratory Techniques and Procedures   ( mesh )
molecular epidemiology
Dissertations, Academic -- Public Health -- Masters -- USF   ( lcsh )
Dissertations, Academic -- Public Health -- Masters -- USF   ( mesh )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )


ABSTRACT: The Enterovirus genus within the family Picornaviridae contains over 100 serotypes, of which sixty-four are known to be human pathogens. Infection with this group of RNA viruses produces a myriad of clinical conditions including poliomyelitis, meningitis, encephalitis, respiratory illnesses, and hand-foot-and-mouth disease. Outbreaks have been documented worldwide; significant morbidity and mortality exist to warrant laboratory surveillance. Traditionally, enteroviruses have been identified to the level of serotype by the serum neutralization assay. However, numerous problems are associated with this assay. The serum neutralization assay is labor intensive, results are often ambiguous, and reagents are becoming difficult to obtain. Recently, molecular-based typing protocols have been described that are cost effective and produce results that are more reliable. The overall objective of this thesis was to implement a molecular-based typing protocol to replace the serum neutralization method currently used. Three specific aims were identified to reach this objective. First, a database cataloging all enteroviruses isolated at the Florida Department of Health - Tampa Branch Laboratory from 1981 through 2002 was created. Serotype prevalence, specimen submission rates, and temporal trends were analyzed to demonstrate the public health importance of enterovirus surveillance. Next, five oligonucleotide primer sets were compared with respect to sensitivity, specificity, and overall utility in molecular typing protocols developed to accurately determine enterovirus type. Finally, the most effective molecular assay was used to conduct two basic molecular epidemiological analyses of intratypic variation of Coxsackievirus B5 isolates, and of intratypic variation of successive Echovirus 9 passages. The results from this study show that implementation of a molecular-based typing system for enteroviruses would be an improvement over current enterovirus serotyping methods. Results are obtained more rapidly and are more reliable. The implementation of such a system would improve the surveillance capabilities of the State of Florida Department of Health.
Thesis (M.S.P.H.)--University of South Florida, 2003.
Includes bibliographical references.
System Details:
System requirements: World Wide Web browser and PDF reader.
System Details:
Mode of access: World Wide Web.
Statement of Responsibility:
by Matthew Adams Smith.
General Note:
Title from PDF of title page.
General Note:
Document formatted into pages; contains 86 pages.

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001441491
oclc - 54018096
notis - AJM5931
usfldc doi - E14-SFE0000164
usfldc handle - e14.164
System ID:

This item is only available as the following downloads:

Full Text
xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam Ka
controlfield tag 001 001441491
003 fts
006 m||||e|||d||||||||
007 cr mnu|||uuuuu
008 031203s2003 flua sbm s000|0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0000164
b SE
QW 169.5.P4
S655e 2003
1 100
Smith, Matthew Adams.
0 245
Evaluation and implementation of a molecular-based protocol for the identification of enteroviruses at the Florida Department of Health Tampa Laboratory
h [electronic resource] /
by Matthew Adams Smith.
[Tampa, Fla.] :
University of South Florida,
Thesis (M.S.P.H.)--University of South Florida, 2003.
Includes bibliographical references.
Text (Electronic thesis) in PDF format.
System requirements: World Wide Web browser and PDF reader.
Mode of access: World Wide Web.
Title from PDF of title page.
Document formatted into pages; contains 86 pages.
ABSTRACT: The Enterovirus genus within the family Picornaviridae contains over 100 serotypes, of which sixty-four are known to be human pathogens. Infection with this group of RNA viruses produces a myriad of clinical conditions including poliomyelitis, meningitis, encephalitis, respiratory illnesses, and hand-foot-and-mouth disease. Outbreaks have been documented worldwide; significant morbidity and mortality exist to warrant laboratory surveillance. Traditionally, enteroviruses have been identified to the level of serotype by the serum neutralization assay. However, numerous problems are associated with this assay. The serum neutralization assay is labor intensive, results are often ambiguous, and reagents are becoming difficult to obtain. Recently, molecular-based typing protocols have been described that are cost effective and produce results that are more reliable. The overall objective of this thesis was to implement a molecular-based typing protocol to replace the serum neutralization method currently used. Three specific aims were identified to reach this objective. First, a database cataloging all enteroviruses isolated at the Florida Department of Health Tampa Branch Laboratory from 1981 through 2002 was created. Serotype prevalence, specimen submission rates, and temporal trends were analyzed to demonstrate the public health importance of enterovirus surveillance. Next, five oligonucleotide primer sets were compared with respect to sensitivity, specificity, and overall utility in molecular typing protocols developed to accurately determine enterovirus type. Finally, the most effective molecular assay was used to conduct two basic molecular epidemiological analyses of intratypic variation of Coxsackievirus B5 isolates, and of intratypic variation of successive Echovirus 9 passages. The results from this study show that implementation of a molecular-based typing system for enteroviruses would be an improvement over current enterovirus serotyping methods. Results are obtained more rapidly and are more reliable. The implementation of such a system would improve the surveillance capabilities of the State of Florida Department of Health.
Co-adviser: PhD, Ann DeBaldo
Co-adviser: PhD, Lillian Stark
molecular epidemiology.
Dissertations, Academic
x Public Health
2 650
Enterovirus B, Human.
Epidemiology, Molecular.
Polymerase Chain Reaction.
Transcription, Genetic.
Laboratory Techniques and Procedures.
Dissertations, Academic
Public Health
t USF Electronic Theses and Dissertations.
4 856


Evaluation and Implementation of a Molecular-B ased Protocol for the Identification of Enteroviruses at the Florida Department of Health – Tampa Laboratory by Matthew Adams Smith A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science in Public Health Department of Environmental and Occupational Health College of Public Health University of South Florida Co-Major Professor: Lillian Stark, Ph.D. Co-Major Professor: Ann DeBaldo, Ph.D. Azliyati Azizan, Ph.D. Date of Approval: November 13, 2003 Keywords: echovirus, coxsackievirus, surveillance, vp1, rt-pcr, molecular epidemiology Copyright 2003, Matthew Smith


i Table of Contents List of Tables iii List of Figures iv List of Abbreviations v Abstract vi Literature Review 1 Introduction 1 Taxonomy 2 Illnesses 7 Epidemiology 10 Infection Cycle 12 Molecular Biology 14 Capsid Structure 16 Detection and Surveillance 18 Objectives 23 Materials and Methods 24 Database Creation 24 Virus Panel Preparation 25 Nucleic Acid Extraction 27 Nucleic Acid Amplification 28 Detection and Sample Preparation for Sequencing 30 Nucleotide Sequencing 31 Sequence Analysis 32 Results 34 Specific Aim #1 (Descriptive Epidemiology) 34 Specific Aim #2 (RT-PCR Comparison) 39 Specific Aim #3 (Sequence Analysis) 41 Discussion 50 References 58


ii Appendices 68 I Master Mix Components and Thermal Cycling Parameters 69 II RT-PCR Comparisons: 2 x 2 Tables 72 III Alignment Reports 74 IV Possible Algorithm for Molecu lar-Based Enterovirus Serotyping 78


iii List of Tables Table 1. Members of the Picornavirus Family 2 Table 2. Currently Recognized Enterovirus Serotypes 5 Table 3. Isolates Used in this Study 26 Table 4. Oligonucleotide Primer Sets 29 Table 5. RT-PCR Results 41 Table 6. BLAST Results 43


iv List of Figures Figure 1. Dendrogram of Enterovirus P1 Nucleotide Sequence 7 Figure 2. Routes of Enterovirus Infection 13 Figure 3. Enterovirus Genomic Organization 16 Figure 4. Enterovirus Capsid Structure 18 Figure 5. VP1 Amplicon Locations 30 Figure 6. Serotype Distribution 36 Figure 7. Enterovirus Isolations by Year 37 Figure 8. Enterovirus Diagnostic Isolations 38 Figure 9. Specimen Submission by Type 38 Figure 10. Enterovirus Is olation by Month 39 Figure 11. RT-PCR Assay Comparison 40 Figure 12. Megalign Phylogenetic Tree 45 Figure 13. MEGA Phylogenetic Tree 47 Figure 14. Portion of E9 Passage Alignment 49


v List of Abbreviations BLAST Basic Local Alignment Search Tool CAV Coxsackievirus A CBV Coxsackievirus B CDC Centers for Disease Control CPE Cytopathic Effect DNA Deoxyribonucleic Acid E Echovirus EBSS Eagle’s Balanced Salt Solution EMEM Eagle’s Minimum Essential Medium EV Enterovirus HAV Hepatitis A Virus MEGA Molecular Evolutionary Genetics Analysis PV Poliovirus RFLP Restriction Fragment Length Polymorphism RNA Ribonucleic Acid RT-PCR Reverse Transcription-Polymerase Chain Reaction SN Serum Neutralization UPGMA Unweighted Paired Group Mean Arithmetic WHO World Health Organization


vi Evaluation and Implementation of a Molecu lar-Based Protocol for the Identification of Enteroviruses at the Florida Department of Health – Tampa Laboratory Matthew Adams Smith ABSTRACT The Enterovirus genus within the family Picornaviridae contains over 100 serotypes, of which sixty-four are known to be human pathogens. Infection with this group of RNA viruses produces a myriad of clinical conditions including poliomyelitis, meningitis, encephalitis, respiratory illnesses, and hand-foot-and-mouth disease. Outbreaks have been documented worldwide; significant morbidity and mortality exist to warrant laboratory surveillance. Traditionally, enteroviruses have been identified to the level of serotype by the serum neutralization assay. However, numerous problems are associated with this assay. The serum neutralization assay is labor intensive, results are often ambiguous, and reagents are becoming difficult to obtain. Recently, molecular-based typing protocols have been described that are cost effective and produce results that are more reliable. The overall objective of this thesis was to implement a molecular-based typing protocol to replace the serum neutralization method currently used. Three specific aims were identified to reach this objective. First, a database cataloging all enteroviruses isolated at the Florida Department of Health – Tampa Branch Laboratory from 1981 through 2002 was created. Serotype prevalence, specimen submission rates, and temporal trends were analyzed to demonstrate the public health importance of enterovirus


vii surveillance. Next, five oligonucleotide primer sets were compared with respect to sensitivity, specificity, and overall utility in molecular typing protocols developed to accurately determine enterovirus type. Finally, the most effective molecular assay was used to conduct two basic molecular epidemiological analyses of intratypic variation of Coxsackievirus B5 isolates, and of intratypic variation of successive Echovirus 9 passages. The results from this study show that implementation of a molecular-based typing system for enteroviruses would be an improvement over current enterovirus serotyping methods. Results are obtained more rapidly and are more reliable. The implementation of such a system would improve the surveillance capabilities of the State of Florida Department of Health.


1 Literature Review Introduction The group of viruses collectively known as the enteroviruses has historically been, and continues to represent, a significant public health threat throughout the world. In the United States alone, infections with enteroviruses result in an estimated fifteen million symptomatic illnesses each year (98). Enterovirus infections produce a myriad of clinical conditions ranging from minor respiratory infections to myocarditis, to severe neurological conditions such as encephalitis, meningitis, and poliomyelitis. The Enterovirus family includes dozens of members, most notably poliovirus, perhaps the most studied of all known viruses. In fact, the field of modern virology literally began with the discovery of poliovirus. It was not only the first virus to be propagated in vitro but poliovirus was also the first to be conclusively linked to the etiology of a human disease (80). Although fiercely debated by some, the development of vaccines against poliovirus by Jonas Salk and Albert Sabin is generally viewed as a significant victory over infectious disease (50). In the last fifty years poliovirus transmission has been drastically reduced worldwide leaving the non-polio enteroviruses as major contributors morbidity and mortality. The genus Enterovirus is one of the largest members of the family Picornaviridae Along with the Rhinoviruses, these positive-sense, single-stranded RNA (+ssRNA) viruses are responsible for many human infectious diseases. Other genera in


2 this family include Apthovirus, Cardiovirus, Hepatovirus, and Parechovirus. These vertebrate pathogens are summarized in Table 1. The Picornaviruses share many physical and biological properties such as virion size/shape, buoyant density, receptor usage, and infectious cycle. The origin of the term “picornavirus” can be interpreted as either: a) “pico” (Greek for “small”) RNA virus, or b) the acronym P olio, I nsensitive to ether, C oxsackie, O rphan, R hino, RNA genome – viruses (7). All Picornaviruses share a common genomic organization with subtle variations that produce the vast phenotypic variation of the family. Table 1: Members of the Picornavirus Family Genus Serotypes Species Diseases Rhinovirus 103 Human; Bovi ne Respiratory illnesses Enterovirus 89 Human; Bovine; Porcine Poliomyelitis; Meningitis; Febrile illness Apthovirus 8 Bovine; Porcine Foot and Mouth Disease Cardiovirus 2 Human; Murine Encephalitis, Myocarditis Hepatovirus 2 Human; Simian Acute hepatitis Parechovirus 2 Human Meningitis Adapted from (80). In the following sections enterovirus taxonomy, illness, epidemiology and infection cycles will be discussed. A general overview of the molecular biology and structure of the virus capsid will follow, as an understanding of these two areas is fundamental to the success of molecular methods. Finally, current detection and identification methods will be reviewed. Taxonomy The Enterovirus genus “officially” include s eighty-nine serotypes of which sixtyfour are known to be pathogenic in humans (89). Poliovirus, initially isolated in 1919, was the first enterovirus to be characterized. It was later discovered that three distinct variants existed (Polio 1, 2, and 3) and though all produced similar pathology, they could


3 not be neutralized by the same anti-serum. Laboratory advances in culture techniques during the 1940’s and 1950’s enabled virologists to isolate new viruses that appeared morphologically similar to the polioviruses but were pathologically and antigenically distinct. The first of these new ‘serotypes’ was isolated in Coxsackie, New York during the summer of 1948 from a child suffering from polio-like symptoms. The ability of this virus to produce acute flaccid paralysis in suckling mice led to the establishment of the Coxsackie A grouping (77). Subsequently, is olates were obtained that shared this biological trait, yet exhibited significant antigenic variation. Thus, the Coxsackie A (CAV) group was expanded. These isolates were numbered sequentially CAV1 through CAV24, with the first isolate of each serotype considered to be the prototype strain. Currently twenty-three serotypes are recognized after it was discovered that CAV23 had previously been described as Echovirus 9 (36). The Coxsackie B viruses (CBVs) differ from CAVs in regard to their pathological effects. The CBVs were observed to produce a more generalized infection in mice that involves not only the central nervous system, but cardiac and adipose tissues as well (72). It was later discovered that CBVs grow readily in primary monkey kidney cells, while many CAVs do not (31). Six CBV serotypes were identified from symptomatic patients. As more enteroviruses were isolated, it became obvious that no correlation existed between clinical disease states and serotype. Echoviruses were routinely isolated from healthy individuals throughout the 1950’s. These E nteric, C ytopathogenic, H uman O rphan viruses were generally found in stool samples (21). An early suggestion was to re-classify these viruses [presumably to


4 one of the CV groups] once a disease state was discovered (31). As more and more echovirus serotypes were identified, it became apparent that the term “orphan” had been applied in error. These viruses were identified as causative agents of meningitis, encephalitis, and numerous other clinical illnesses (72). By 1969, thirty-four echoviruses had been identified and the issue of viral disease etiology remained unresolved. Since 1970, new enteroviruses have simply been assigned a number, beginning with enterovirus 68 (52). Several enteroviruses have been re-classified since their initial description. The genus Parechovirus, created in 1998, contains two members, formerly echovirus 22 and echovirus 23, that differ markedly from other enteroviruses. This re-classification was based upon both genotypic and phenotypic di screpancies (96). Echovirus 10 was renamed Reovirus in 1959 and enterovirus 72 was changed to Hepatitis A virus in 1991 based upon similar observations (31). Other taxonomic rearrangements have occurred based on the fact that multiple serotypes have been demonstrated to actually be strains within the same serotype. Table 2 summarizes the currently recognized serotypes, along with their prototype strains and geographic locations of initial isolation. In recent years, molecular data has contributed substantially to viral taxonomy. The availability of cost-effective nucleic acid sequencing techniques has resulted in the accumulation of an enormous amount of data that was unobtainable just a decade ago. One distinct advantage of the use of sequence-derived data in virus taxonomy is that phylogeny can more readily be inferred (48). By comparing nucleotide sequences it is possible to determine the extent to which a given group of viruses are related. For instance, the aforementioned re-classifications within the Enterovirus genus become


5 obvious when sequence data are compared and analyzed (57). Several new enterovirus serotypes (EV73 – EV78) have been identified using molecular methods, but have yet to be officially recognized (54,56,65). Table 2: Currently Recognized Enterovirus Serotypes Virus Prototype Origin Virus Prototype Origin CA1 Tompkins New York E7 Wallace Ohio CA2 Fleetwood Delaware E8 [=E1’] Ohio CA3 Olson New York E9 Hill Ohio CA4 High Point North Carolina E10 [=Reovirus] CA5 Swartz New York E11 Gregory Ohio CA6 Gdula New York E12 Travis Philippines CA7 Parker New York E13 Del Carmen Philippines CA8 Donovan New York E14 Tow Rhode Island CA9 Bozek New York E15 CH 96-51 West Virginia CA10 Kowalik New York E16 Harrington Massachusetts CA11 Belgium-1 Belgium E17 CHHE-29 Mexico City CA12 Texas-12 Texas E18 Metcalf Ohio CA13 Flores Mexico E19 Burke Ohio CA14 G-14 South Africa E20 JV-1 Washington, D. C. CA15 G-9 South Africa E21 Farina Massachusetts CA16 G-10 South Africa E22 [=Parechovirus1] CA17 G-12 South Africa E23 [=Parechovirus2] CA18 G-13 South Africa E24 DeCamp Ohio CA19 NIH-8663 Japan E25 JV-4 Washington, D. C. CA20 IH-35 New York E26 Coronel Philippines CA21 Kuykendall California E27 Bacon Phillipines CA22 Chulman New York E28 [=Rhinovirus1] CA23 [=Echo 9] E29 JV-10 Washington, D. C. CA24 Joseph South Africa E30 Bastianni New York CB1 Conn-5 Connecticut E31 Caldwell Kansas CB2 Ohio-1 Ohio E32 PR-10 Puerto Rico CB3 Nancy Connecticut E33 Toluca-3 Mexico CB4 JVB New York E34 DN-19 [=CA24'] Texas CB5 Faulkner Kentucky EV68 Fermon California CB6 Schmidt Philippines EV69 Toluca-1 Mexico E1 Farouk Egypt EV70 J670 Japan E2 Cornelis Connecticut EV71 BrCr California E3 Morrisey Connecticut EV72 [=Hepatitis A] E4 Pesascek Connecticut P1 Brunhilde Maryland E5 Noyce Maine P2 Lansing Michigan E6 D'Amori Rhode Island P3 Leon California Grey rows indicate serotype s that have been re-classi fied. Adapted from (9,72) Based on molecular phylogeny the sixty-four human enteroviruses segregate into five distinct clusters. Five human pathogen species (corresponding to these clusters) are recognized within the genus Enterovirus: HEV-A, HEV-B, HEV-C, HEV-D, and PV.


6 These species are demarcated largely according to molecular data. The International Commission on Taxonomy of Viruses considers seventy percent amino acid homology in three proteins (P1, 2C, and 3CD) as a defining characteristic of each species (36). [Note: The structural and functional roles of these proteins will be described in more detail in subsequent sections.] HEV-B is the largest species containing CAV9, all CBVs, all echoviruses, and EV69 for a total of thirty-six serotypes. HEV-D, the smallest species, contains only two serotypes, EV68 and EV70. HEV-C contains ten CAVs and is closely related to PV, which contains all three poliovirus serotypes (75). Phylogenetically, the PV species cluster tightly with the HEV-C species, however the species remain distinct based upon differences in clinical manifesta tions of infection (31). HEV-A includes eleven CAVs and EV71 (11). Current data suggest that CAV4 and CAV6, which are not assigned to a species, be included with the HEV-A (61). Variation within serotypes is significant at the molecular level and the term “quasi-species” is often used to describe individual strains (23). Figure 1 depicts a dendrogram of the five human enterovirus species. The use of an un-rooted dendrogram implies that a common ancestor has not been identified and, consequently, the direction of evolution cannot be inferred.


7 Figure 1: Dendrogram of Enter ovirus P1 Nucleotide Sequence Genetic relatedness of enteroviruses based on the nucleotide sequence of the entire P1 protein. Note: PV is not shown as a separate species, but its genetic relations hip to HEV-C is evident. The E cluster is included as an outgroup to demonstrate the divergence of the bovine species which do not cause illness in humans. Used with permission (55). Illnesses Most severe of all enterovirus infections is poliomyelitis. Evidence suggests that poliovirus is truly an ancient human pat hogen. Hieroglyphic drawings have been found dating back to approximately 1400 BC that depict a young man suffering the effects of poliomyelitis (72). The disease was first described in 1789 by Michael Underwood as a “debility of the lower extremities” that typica lly affected children under the age of five (50). Outbreaks have been described sin ce the 1800’s and the World Health Organization is currently conducting a global eradication program that is approaching its goal (35). As


8 of 2002, poliomyelitis due to poliovirus was endemic to only seven countries and wild transmission of PV2 had been eradicated throughout the world (19). Classical poliomyelitis is a neurological condition that leads to “acute flaccid paralysis.” Poliomyelitis is the result of virus-induced disruption of central nervous system (CNS) function. Often recovery is slow and one extremity is rendered “lame,” while the corresponding limb is unaffected (50). Atrophy and developmental delay are consequences of CNS impairment. Numerous neurological complications can also arise, such as blood pressure irregularities and abnor malities of secretory and excretory organs. Acute illness often precedes CNS involvement with fever, headache, and lethargy occurring first followed by more sever the neurological symptoms (72). Other enterovirus illnesses are generally less extreme than those caused by the polioviruses. Enteroviruses are the most common etiologic agents of aseptic (nonbacterial) meningitis worldwide (99). Viral in fections of the CNS are secondary in nature and their mechanisms have not been elucidated (85). Echovirus 30 is frequently associated with outbreaks of meningitis (6,58). Unlike that of bacterial origin, enterovirus-derived meningitis is generally self-limiting, with adults often remaining symptomatic longer than children (87). Fever, headache, and malaise are often observed. Encephalitis is a more serious condition but its incidence is markedly lower. Global neurological depression (i.e. confusion, weakness, and irritability) can occur suddenly, preceding seizure and even coma in extreme cases (86). Respiratory illnesses are another characteristic clinical outcome of enterovirus infection. Up to fifteen percent of all upper and lower respiratory infections can be traced to an enterovirus origin (20). Croup, “common cold,” and epiglottis (upper respiratory


9 infections) are most often observed (72). General “flu-like” symptoms are quite common manifestations of enterovirus infection. These illnesses are quite often misdiagnosed or undiagnosed due to the vagueness of symptoms (87). Herpangina and hand-foot-andmouth disease affect children and are usually accompanied by characteristic “rashes” (72). Despite their namesake, enteroviruses do not routinely cause enteric illnesses (51). One notable exception is Hepatitis A (formerly known as enterovirus 72), which is commonly implicated in food-borne outbreaks. Disseminated infections can lead to a variety of pathologies including myocarditis, pleurodynuria, and conjunctivitis. All of these conditions are generally selflimiting, although chronic infections are sometimes observed. Selenium deficiency has been reported to be a risk factor for myocarditis, leading to regions of high endemicity in China (73). Enterovirus etiology of rheumatic heart disease has also been described (40). Pleurodynia, or muscle disease, is not commonly encountered, but can be quite a serious condition affecting the musculature of the ribcage. Conjunctivitis is caused primarily by EV70 and CA24, and generally is transmitted via inoculation into the eye (51). There are reports of an association of chronic en terovirus infection with diabetes, but the epidemiological data is not entirely conclusive (68). Newborns suffer the greatest risk of poor clinical outcomes. Multi-system hemorrhagic disease of infants or “sepsis-like” disease is a potentially fatal disorder often caused by E11 and E19 (44). Infants can experience elevated morbidity and mortality due to diminished immune system function, especially in the absence of maternally acquired anti-bodies (1). Perinatal infections can be quite serious and have been shown to be associated with neurodevelopmental delays (25).


10 Epidemiology Enteroviruses cause numerous disease states and do not follow clearly recognizable epidemiological patterns. Infections are generally self-limiting and often asymptomatic, reflecting an evolutionary balance between virus and host. Over ninety percent of poliovirus infections have been shown to lack clinical symptoms (50). On the other hand, virulent strains of other enteroviruses have been documented that produce a disproportionate amount of serious illness (58,92). Enterovirus transmission is known to proceed by four mechanisms. Infections have been shown to occur via the following fecal-oral, respiratory, inoculation, and blood-borne routes (51). Fecal contaminati on of fomites and perinatal transmission to newborns are responsible for a large percentage of enterovirus outbreaks (72). This model can also be applied to developing countries where water quality and general hygiene are often inadequate to prevent infections (51). Enteroviruses are shed in stool at high concentrations, are capable of persisting in the environment for extended periods, and are transmissible via sewage-contaminated water (72). These modes of transmission produce some general temporal and demographic features of enterovirus infection. In temper ate climates, enterovirus infections are most prevalent during the summer months, but can occur year-round in tropical and subtropical climates (51). Children are often at higher risk of infection due to increased contact, poor hygiene, and absence of IgG ge nerated from previous exposure (21). It is known that serotype prevalence varies annually. National trends have been documented, but are skewed due to under/over-reporting in various regions (98). In 2002, E18 and E13 accounted for over sixty percent of all enteroviruses isolated in the


11 United States as reported by the National Enterovirus Surveillance System (18). However, from 1993 to 2001, these two serotypes accounted for less than five percent of all isolates (16,17). Data from this passive surveillance mechanism is several biases. Serotypes that grow readily in culture and are easily typed via serum neutralization will tend to be isolated and identified more frequently. In addition, specimen submission for typing will be skewed in favor of serotypes that cause more severe illness. Echovirus 30, E11, and CB5 are prevalent perennially and have been linked to numerous outbreaks worldwide (37,52,63). Variant strains of E30 have been described which exhibit distinct epidemiological patterns (70). Several EV71 outbreaks have occurred in the recent years in the South-Pacific. Hand-foot-and-mouth disease, the primary clinical syndrome caused by EV71, is generally non-life-threatening and occurs mainly in children. The virus has caused some outbreaks of polio-like illness, complicating eradication and surveillance efforts (22). Encephalitis and pulmonary edema have been observed in a disproportionate number of cases during these outbreaks (92). These epidemics have been characterized by abnormally high infection rates and increased case-fatality rates (49). Between 19972001, 159 deaths due to EV71 infection have occurred in Malaysia and Taiwan alone (42). Numerous molecular epidemiological studies have been conducted, but neither the mechanism nor origin of this increase in virulence is understood (12,30). It has been hypothesized that the differential cell tropism of EV71 relative to other enteroviruses may be a key factor in these outbreaks (104).


12 Infection Cycle Despite wide variation in clinical manifestations, the primary route of enterovirus infection is through the alimentary tract. Since fecal-oral transmission is the predominant mode of enterovirus transmission, virions must be capable of surviving in the gut at least long enough to initiate an infection. Enteroviruses are resistant to both low pH (<2) and enzymatic degradation (74). Figure 2 depicts the general course of enterovirus infection. Receptor binding is the first of several steps in the course of infection. Numerous receptors, with which enteroviruses interact, have been identified. Polioviruses recognize the molecule CD155, also known as the “poli ovirus receptor” (83). The CAVs and CBVs both recognize different receptors: intercellular adhesion molecule (ICAM-1) and coxsackie-adenovirus receptor (CAR), respectively. Echoviruses recognize other molecules including delay-accelerating factor (DAF) and CD55 (82). These molecules tend to be membrane-spanning glycoproteins with multiple extra-cellular domains (8). Variation in receptor structure is well documented and most certainly plays a role in not only tissue tropism, but clinical disease as well (4). Genes encoding these various receptors have been shown to exhibit extreme variation. It is likely that different allelic forms of receptors are correlated with severity of disease (34). Upon binding to an appropriate receptor, enteroviruses must insert their nucleic acid into the host cell. Individual enterovirus virions undergo a conformational change upon receptor recognition (29). These particles have altered physical properties that allow cellular uptake of the virus (79). Endosome-mediated transport delivers enterovirus particles to the cytoplasm where they are quickly degraded to release the infectious RNA (95).


13 Figure 2: Route of Enterovirus Infection Route of entry and locations of enterovirus infection. Used with permission (79). Pocket factors have been identified that appear to be necessary for viral entry. These compounds are generally long-chain fatty acids that stabilize the altered conformation facilitating efficient delivery of the genome into the cell (94). Competitive


14 inhibition of these molecules is a promising area of anti-viral research (84). A series of molecules known as the WIN compounds (so named because they were developed by Sterling-Winthrop Pharmaceuticals) have been shown to drastically reduce the infectivity of enteroviruses in vitro (87). Synthetic peptides that block the virus-receptor interaction (through similar mechanisms) have also been proposed as therapeutic options (26,78). Enterovirus replication proceeds in the cytoplasm of infected cells (27). Viral RNA is first transcribed to a negative strand intermediate, which is then transcribed to the positive strand RNA (33). As viral RNA accumulates in infected cells, pathological changes begin to occur, although their biochemical mechanisms are not currently understood (2). Most notable of these alterations are the inhibition of both host cell transcription and translation, and the increase in permeability of the cell membrane (100). The increased permeability of the host cell membrane leads to lysis and is generally considered the underlying cause of cytopathic effect observable in infected, cultured cells (90). Upon lysis, viral progeny are released into the bloodstream and either initiate secondary viremias or are excreted. Thus, it is not surprising that enteroviruses can be isolated from a variety of tissues and sample types. Molecular Biology Genome organization and capsid structure are highly conserved throughout the family Picornaviridae and especially within the genus Enterovirus. Slight differences demarcate species, but generally these are relatively minor variations pertaining to receptor recognition, protein processing, or lengths of non-translated regions. The enterovirus genome is approximately 7500 nucleotides and is arranged in a single open reading frame (ORF). This multiple protein coding sequence is flanked on both the 5’


15 and 3’ ends by non-translated regions (NTR s). Although these regions do not code for proteins, their functions are vital to the infection cycle. While a 3’-poly-adenine tail is present, enteroviruses lack a 5’-methyl-cap. The 5’NTR is comprised of approximately 740 nucleotides with a highly conserved secondary structure (stem loop) that is often referred to as the “cloverleaf” (28). Downstream of this loop is the internal ribosome entry site (IRES), which is involved in the initiation of cap-independent translation (106). Induction of this alternate form of translation is crucial in host cell translation shutoff and allows viral progeny to be preferentially propagated (105). The 3’NTR, approximately seventy to one hundred nucleotides in length, forms a pseudoknot-like element (PKLE). This structure is thought to be involved in (-) strand RNA synthesis, the first step in viral replication (106). The 3’-poly-adenine tail is of variable length (thirty to one hundred nucleotides), and is required for infectivity (80). Both non-translated regions are highly conserved due to the functional aspects of their secondary structure (76). The single ORF codes for all structural and functional proteins. Proteolytic processing produces multiple proteins from the polypeptide. Virally-encoded proteases cleave nascent chains in precise locations which are highly conserved throughout the family Picornaviridae (7). Several processing intermediates play vital functional roles, reflecting an evolutionary trend towards efficient utilization of a relatively compact genome (28). Figure 3 summarizes the genomic organization of the enteroviruses.


16 Figure 3: Enterovirus Genomic Organization 1A 1B 2B 1D 1C 2A 2C 3A 3C 3D VP4VP2VP1 VPgVP3Structural Functional 5’ 3’ P1 P3 P2 The enterovirus genome, 5’ to 3’, w ith the relative size of each protein shown to scale. The 5’and 3’NTRs are not shown to scale. Three distinct regions comprise the ORF, which is approximately 6700 nucleotides in length. The P1 region codes for the viral capsid proteins, which are described below. These proteins are translated first and self-assemble to form capsids within the cytosol (29). The P2 and P3 regions encode proteases and other functional proteins crucial to the replication cycle. The 2A protease initially separates the P1 region from the P2-P3 peptide via an autocatalytic mechanism (80). The 3C and 3CD proteinases carry out subsequent cleavages ( 33). The 3D RNA-dependent Polymerase is the last protein translated and synthesizes (-) strand RNA (28). This enzyme lacks “proofreading” and its high error rate reinforces enterovirus variation (24). The VPg protein (also known as 3B) is crucial in linking replicated RNA to the capsid (33). Other proteins (2B, 2C 3A) play minor, but crucial roles in the replication cycle. Capsid Structure The ninety-seven kD P1 region represents roughly forty percent of the ORF and is cleaved to form the four capsid proteins VP1VP4 (29). These proteins assemble to form the 30nm diameter icosahedron capsid. VP1, VP2, and VP3 are all approximately the


17 same size, while VP4 is much smaller. Sixty copies of each protein are present in the capsid. This number is absolutely critical to capsid geometry and does not vary (29). The protomer is the building block of the capsid. Each protomer is composed of one copy each of the VP1-VP4 proteins. VP1, VP2, and VP3 form the exterior surface, while VP4 lies on the interior. The external peptides (VP1-VP3) are composed of eight anti-parallel -sheets and two -helices on each terminus; the internal peptide (VP4) is composed of only two -sheets (7). Five protomers assemble to form a pentamer; twelve pentamers combine to form the capsid. A diagram of the enterovirus capsid is provided in Figure 4. Two distinct types of symmetry axes are formed: three-fold and five-fold. Numerous depressions are formed on the capsid surfaces due to cumulative molecular interactions of the peptides. Near the five-fold symmetry axis a groove is formed which, according to the “canyon hypothesis,” facilitates receptor binding (93). Capsid structure, although highly ordered, is significantly divergent across the genus Enterovirus and throughout the family Picornaviridae The P1 region as a whole is much more divergent than the P2 and P3 regions (29). In fact, it is the variation in the capsid proteins that define serotypes (80). The ability to elicit an immune reaction with a specific monoclonal antibody is the hallmark trait of a serotype (72). For instance, the same mono-clonal antibody will not recognize both CB1 and CB2. Antibodies bind to specific protein structures based upon the sp ecificity conferred by their long and short chains. The ability of specific antibodies to neutralize enteroviruses has been the primary means of determining serotype for decades (described in more detail below).


18 Figure 4: Enterovirus Capsid Structure Three proteins are involved in protomer formation. These protomers assemble to from the capsid. Adapted from (3). Neutralization sites have been characterized for many enteroviruses. Techniques such as X-ray crystallography and cryo-electron microscopy have elucidated antibodyvirus interaction at extremely high resolutions (97). The VP1 protein has been shown to contain the majority of these neutralization sites (39). The structures of “loops” protruding from the capsid surface confer serot ype specificity (47). These loops link the eight -sheets and are named accordingly with the sheets they connect. For example, the AB loop links the A and B sheets, both of which are embedded in the capsid surface. VP1 has three such loops, while VP2 and VP3 have two each (7). The BC loop in particular has been shown to be involved in neutralization (54). Detection and Surveillance While some may argue that enterovirus detection alone is sufficient, determination of serotype is often advantageous. It is important to note that clinical and public health needs are often not identical. The ability to differentiate between polio and VP2 VP 1 V P 3x 5 x 12 ProtomerPentamer Capsid


19 non-polio enteroviruses is critical in regions where polio is endemic and active surveillance is being conducted (52). Differentiation among PV strains (intratypic variation) is also desirable as certain strains may differ in their ability to produce serious illness (101). Identification of epidemics is quite beneficial from a public health standpoint. Some enterovirus serotypes tend to be perennially endemic in certain areas, while others are associated with infrequent outbreaks (72). Finally, the isolation and characterization of new serotypes is only possible if identification (subsequent to detection) is pursued (52). Traditionally serological assays have been the method of choice in enterovirus identification. Inoculation onto cultured cells subsequent to appropriate specimen preparation has been used in clinical virology laboratories to identify virus infection for decades. A large number of different assays can be used to identify the infectious agent. Three of the most common include complement fixation, fluorescent antibody, and serum neutralization. Each of these assays exploits a specific antigen-immune response relationship and the use of each is indicat ed indifferent occasions. Depending on the particular assay design, the presence/ absence of an immune reaction is indicative of infection with a particular agent. Viral isolation from culture is generally used to detect enterovirus infection (103). Due to their large range of tissue tropism, enteroviruses can be isolated from a variety of different specimens, including: stool, CSF, blood, serum, and nasal/throat swabs (88). Enteroviruses produce a distinct and easily recognizable CPE in cultured cells. Most enteroviruses will grow readily in BGM cultures, although some serotypes are difficult to culture (5). Additionally, many of the CAVs propagate only in suckling mice (79).


20 The serum neutralization assay is considered the “gold standard” to identify serotype once an enterovirus has been isolated in culture. This assay, perfected in the 1950’s, uses eight intersecting pools of antisera to determine serotype (41). Once CPE is observed in the initial culture, the suspected enterovirus is titrated to a concentration of 100 TCID50. This concentration of virus is sufficient to demonstrate a neutralization reaction, but is not so high as to overcome the antibodies present in the antisera. Incubation of virus in the presence of homologous antisera will cause the virus to be neutralized (i.e. diminished observable CPE). The serum-virus mixture is then inoculated into multiple cultures of host systems. Depending on which of the eight anti-sera pools produce neutralization reactions, serotype can be inferred by means of a chart. These reactions can be performed in either test tubes or ninety-six well plates. Significant drawbacks exist with the serum neutralization assay. Most significant (at least in terms of clinical utility of laboratory diagnosis) is the time required for detection and identification. Some serotypes take weeks to produce observable CPE. This coupled with time required to titrate the virus and perform the actual neutralization test can lead to diagnostic “lags” of over one month. Blind passages are often required, adding to the time and cost of the assay (13). Even when performed correctly, the neutralization assay may fail due to antigenic drift, virus aggregation, or the presence of multiple serotypes (66). The test is expensive, and the results may be difficult to interpret (102). The handling of live, infectious viruse s requires extensive staff training and can be dangerous if performed improperly. The maintenance of continuous cell culture lines is prohibitively expensive to many laboratories. Even reagents are difficult to obtain as they are produced by WHO and distributed so mewhat sparingly (52). These WHO pools


21 only include anti-sera to forty serotypes. Supplemental pools and individual serotypespecific antisera are required for the identification of all sixty-four serotypes (64). In a quality assessment survey of twelve European virology laboratories conducted in 1995, the difficulties of enterovirus identification were demonstrated (102). Each laboratory was sent a proficiency pane l of ten specimens, which contained one, two, or no enteroviruses. The serotypes represented in this panel (P1, P2, P3, CA7, E4, E6, E30, EV71, CB3, and E11) were all common serotypes, known to produce strong CPE in susceptible cell lines. Clearly, there is a need to develop more reliable, unambiguous, time-efficient methods to identify enteroviruses. It is not surprising that a host of new diagnostic tools have been utilized in recent years. Immunoperoxidase tests, hybridization assays, RFLP, and microchip array technology have been used for the detection and/or serotyping of enteroviruses from a variety of specimens (10,46,91). Among these, RT-PCR is currently the most widely used procedure. Increased sensitivity and turnaround time are two of the major advantages of RT-PCR. Some virologists advocate that routine diagnostic cell culture should be abandoned in favor of nucleic acid amplification methods (13). This position, however, is strongly opposed by others who claim that elimination of cell culture would lead to the inability to recognize novel viruses (67). RT-PCR assays for enterovirus detection often utilize oligonucleotide primer sets that anneal to the highly conserved 5’NTR region. These “pan-enterovirus” primers can detect virtually all enterovirus serotypes, as well as some rhinoviruses (88). However, due to the lack of variation in the amplicon, these primers are of limited utility for molecular typing. It has been shown that no reliable correlation exists between 5’NTR sequence and serotype (38).


22 Numerous RT-PCR assays have been developed that utilize oligonucleotide primers that anneal to sequences encoding the capsid proteins. Sequence analysis of the amplicon can be used to determine serotype. Primer design, however, is crucial: primers must anneal to conserved regions, but produce a relatively divergent amplicon. The VP2 region sequence was shown to poorly correlate with serotype since expected branch points were not observed when phylogenetic trees were constructed (62). The VP4 region has been used with some success in Japan, but likely suffers some limitations because VP4 is located on the interior of the enterovirus capsid and is not exposed (32). The VP1 region has been used extensively for molecular typing of enteroviruses. The sequence of the VP1 region has been documented to highly correlate with serotype due to the large number of neutralization sites present on the VP1 surface. Often these assays utilize highly degenerated oligonucleotide primer sets that contain “wobbles” and deoxyinosine residues. These primer modifications allow the primer to tolerate some variation in target sequence. Subsequentl y, these primers are capable of amplifying many different serotypes. The deoxyinosine residues will bind to any base, but reaction kinetics involved lead to decreased sensitivity (60). Thus, it is often difficult to obtain “sequencable” amplicons from clinical specimens. Most VP1 assays still require that an enterovirus be cultured in order to obtain virus concentrations significantly higher than those present in clinical specimens (69).


23 Objectives The scope of this study pertains to enteroviruses isolated in the Virology Section of the Florida Department of Health – Tampa Branch Laboratory from 1981 to present. The primary objective of this project was to implement a molecular-based protocol for the identification of enterovirus serotypes that could replace the serum neutralization assay currently in use. Three specific aims were outlined for the study: 1) Descriptive Epidemiology – Data on enterovirus isolation since 1981 was analyzed for descriptive epidemiological parameters, and compared to national trends. Seasonality, serotype distribution, year-to-year variation, and specimen submission trends were evaluated. 2) RT-PCR Assay Comparison – Five VP1 oligonucleotide primer sets were examined for their efficacy in enterovirus detection relative to a 5’NTR assay. Assays were compared on the basis of sensitivity, specificity, positive predictive value, and negative predictive value. 3) Sequence Analysis – The best-performing molecular assay was used to determine serotype of enteroviruses via nucleotide sequencing of amplicons. Various analysis algorithms were compared relative to serum neutralization results. Two additional analyses were performed to exemplify the overall utility of the developed protocol; one investigating intra-serotypic divergence, and the other examining the effects of sequential in vitro passages.


24 Materials and Methods Database Creation Microsoft Access was used to create a database cataloging all enteroviruses isolated at the Florida Department of Hea lth – Tampa Branch Laboratory. Two thousand one hundred fifty specimens from 1981 through 2002 were entered. The fields included were: lab identification number, specimen type, date inoculated, and serotype (determined via SN assay). Where applicable, multiple specimens from the same patient were noted as “duplicate” to ensure that double counting did not occur. Of the 2150 isolates, 159 such duplicates were removed from analysis, leaving 1991 unique isolates. Performance evaluation samples were not entered into the database. Although only viruses isolated from patients were used, not all isolates were from symptomatic cases. In accordance with the IRB Category Four exemption status (IRB #100880), only laboratory identification numbers were used in this study. No client information was available to the investigator. Microsoft Excel was used to analyze the above data for general descriptive epidemiology. Serotype frequency was determined for the entire period as well as multiple intervals within the twenty-two year window. Temporal dynamics (i.e. seasonality) were explored using various queries to the original database and subsequent export to Excel. Sample submission was analyzed in a similar manner. Data obtained was compared with nationwide data reported by CDC.


25 Virus Panel Preparation A panel of fifty-six distinct enterovirus is olates was prepared for use in this study. The prepared panel included: twelve CAV serotypes (CAV2-10, CAV13-14, and CAV16), six CBV serotypes (CBV1-6), twenty-seven echovirus serotypes (E1-6, E6”, E7, E9, E11, E11’, E12-E21, E24-25, E27, E30, E31, and E33), three EV serotypes (EV68 and EV70-71), three PV serotypes (P13), and five serotypes which have since been re-classified (E8 E1’, E10 Reovirus, E22-23 Parechovirus1-2, and EV72 Hepatitis A). The re-classified serotypes were included for reference because at one time they were all considered to be enteroviruses. No archived isolates were available for fifteen serotypes (CA1, CA11-12, CA15, CA17-22, CA24, E26, E29, E32, and EV69). Financial restrictions and infectious substance shipping regulations made obtaining these serotypes impractical. Table 3 lists all serotypes used in this study. Eight serotypes (CA2-6, CA8, CA10, and CA14) were originally propagated in suckling mice and therefore not cultured. Hepatitis A was not cultured due to a lack of a cytopathic strain. Coxsackievirus B6 was not cultured because no archive was available; a performance evaluation isolate was used instead. Although these ten viruses were not cultured, they were still included in molecular testing as described below. Forty-six serotypes were propagated in buffalo green monkey kidney (BGM), rhesus monkey kidney cells (RMK), or normal fetal human lung (MRC-5) cell lines. BGM and MRC-5 cells are cultivated in-house by a dedicated cell-culture technologist. These are continuous cell lines routinely used for diagnostic isolations at the Tampa Branch Laboratory. RMK cells are received weekly from Viromed (Minnekota, MN; catalog # 14-309).


26 Table 3: Isolates Used in This Study Serotype Cultured Serotype Cultured Serotype Cultured Serotype Cultured CA2 No CB3 Yes E10 (Reo) Yes E23 (PEV2) Yes CA3 No CB4 Yes E11 Yes E24 Yes CA4 No CB5 Yes E11' Yes E25 Yes CA5 No CB6 No E12 Yes E27 Yes CA6 No E1 Yes E13 Yes E30 Yes CA7 Yes E2 Yes E14 Yes E31 Yes CA8 No E3 Yes E15 Yes E33 Yes CA9 Yes E4 Yes E16 Yes EV68 Yes CA10 No E5 Yes E17 Yes EV70 Yes CA13 Yes E6 Yes E18 Yes EV71 Yes CA14 No E6" Yes E19 Yes EV72 (HAV) Yes CA16 Yes E7 Yes E20 Yes P1 Yes CB1 No E8 (E1’) Yes E21 Yes P2 Yes CB2 No E9 Yes E22 (PEV1) Yes P3 Yes Isolates were removed from the -70 C archive freezer and rapidly thawed in 37 C water bath. Prior to inoculation, cell cultures in 15 x 125cm2 screw cap culture tubes were washed with 2ml Eagle’s Balanced Salt Solution (EBSS). Excess fluid was poured off and 100 l of undiluted stock solution virus was inoculated into tube using sterile 1ml pipettes. Tubes incubated in inclined racks for two hours on a rocking platform at 37 C. Eagle’s Minimum Essential Media (EMEM), 5% fetal calf serum (2ml) was added to each tube. Tubes were incubated in a rotating rack at 37 C. Negative controls (noninoculated tubes) were always in cluded when performing inoculations. Cells were examined with an inverted light microscope seventy-two hours postinoculation. Additional observations were performed daily. Cytopathic effect was scored on a scale of “0” to “4.” This scale refers to observable changes that occur in cultured cells upon virus infection: “0” represents no visible cytopathic effect, “1” 1025% of cells appear lysed and shriveled, “2” 25-50%, “3” 50-75%, and “4” represents 75100% appear lysed and shriveled. A “+/-“ is used to signify the upper and lower limits of


27 each range. Results were tallied in a Microsoft Excel spreadsheet. When cells reached a level of “3/3+,” tubes were frozen at -70 C for harvest. If CPE was insufficient (<3) after one week post-inoculation, an additional 2ml EMEM was added to the culture tube which was then returned to the incubator. Cells were frozen for passage if CPE was < 3 two weeks post-inoculation or non-viral induced cell deterioration were observed. Passages were performed in the same manner as inoculations using 200 l inocula. Nucleic Acid Extraction Total RNA was isolated from cells using a Qiagen RNeasy MiniKit (Valencia, CA; catalog # 74104). Cell culture supernatants (140 l) were lysed in 350 l Buffer RLT and 350 l 70% ethanol. Lysates were transferred to RNeasy spin columns and centrifuged at 10,000 x g for 1min. Columns were washed successively with 700 l Buffer RW1 and 500 l Buffer RPE with centrifugation between each wash. Viral RNA was eluted from spin column membrane in RNase-free water. Elution volumes were either 100 l or 35 l. This depended on both the amount of RNA required for subsequent analysis and the desired final concentration. RNA used for RT-PCR assay comparisons was eluted in 100 l water because multiple amplifications were required. RNA extracted solely for sequencing was eluted in 35 l water to yield a more concentrated product and facilitate sequence analysis (see below). Nucleic Acid Amplification RNA was amplified using RT-PCR by either a one-step or two-step procedure, depending on oligonucleotide primer pairs. Oligonucleotide primers were obtained from


28 Operon (Alameada, CA) as lyophilized powders that were re-hydrated in 100 l RNasefree water upon receipt. Multiple 20 l aliquots of 100 M working solutions were prepared and stored at -20 C to eliminate excessive freeze/thaws of stock solution. Six different primer sets were utilized and are summarized in Table 3 and Figure 5. Qiagen One-Step RT-PCR Kits (catalog # 210212) were us ed for five primer pairs. A two-step RT-PCR system using Promega Reverse Transcriptase (Madison, WI; catlog # M900A) and Fisher Scientific Taq Polymerase (Fair Lawn, NJ; catalog # FB60045) was used for one primer set. All temperature cycling was performed in a Perkin Elmer PE9700 thermal cycler. Master mix components and thermal cycler profiles are included in Appedndix I. Five different oligonucleotide primer pairs were analyzed as candidates for inclusion in a molecular serotyping protocol to be implemented at FL DOH – Tampa. Two of these primer sets were developed by Dr. Steven Oberste and his colleagues at CDC (59,64). These primer sets ( Oberste “ A ” and “ B ”) were used in conjunction with a Qiagen One-Step RT-PCR kit. The Casas primer pairs (referred to as “ 1st” and “ 2nd”) were designed to be used in a nested assay that is stated to be capable of producing “sequencable” amplicons from clinical sample s (15). The ability to identify virus identity directly from clinical specimens could eliminate the need to isolate the virus in culture, leading to significant savings in both expense and “turnaround time.” These two primer pairs were used both alone and in nested am plification reactions with Qiagen One-Step RT-PCR reactions. The final primer set, Caro consisted of six oligonucleotides, which were used in a two-step reaction. Despite numerous attempts, the Caro primers failed to produce correctly-sized amplicons with the Qiagen One-Step RT-PCR kit.


29 Cell culture, extraction, and amplification negative controls were included in all experiments. All assays were validated prior to comparison by demonstrating the ability to amplify several common serotypes and produce intense bands on agarose gels. Annealing temperature optimization experiments were initiated, but abandoned due to the necessity of a relatively low annealing temperature (~42 C) required with deoxyinosine bases (60). Table 4: Oligonucleotide Primer Sets Primer Sequence (5’ 3’) Name (Reference) Amplicon Length ENT3 CCTCCGGCCCCTGAATG Screening (88) 196bp ENT4 ACCGGATGGCCAATCCAA 222 CICCIGGIGGIAYRWACAT Oberste A 338bp 292 MIGCIGYIGARACNGG 012 ATGTAYGTICCICCIGGIGG Oberste B 450bp 040 ATGTAYRTICCIMCIGGIGC 011 GCICCIGAYTGITGICCRAA VP1-1A TGIGGAYTGRTAYCTIKYKGGRTARTA Casas 1st (15) 803bp VP1-1S GGTTYGAYITGGARITIACITTYGT VP1-2A CCIGTKKWRCAAIYYRCAYCTIGC Casas 2nd (15) 609bp VP1-2S ARWTWATGTAYRTICCICCIGGIG EUC2 TTTGCACTTGAATATGTA 1452bp EUC2a GGTTCAATACGGCATTTGGA Caro (14) EUC2b GGTTCAATACGGTGTTTGCT EUG3a TGGCAAACTTCCWCCAACCC EUG3b TGGCAAACATCTTCMAATCC EUG3c TGGCAGACTTCAACHAACCC Standard IUB Ambiguity codes used.


30 Figure 5: VP1 Amplicon Locations Location of amplicons shown. Primers anneal to the flanking sides. Lines represent five expected amplicons from the VP1 assays used. “Screening” primers, which anneal to the 5’NTR, are not shown. Detection and Sample Preparation for Sequencing All RT-PCR products (10-20 l) were visualized via gel electrophoresis. One percent agarose gels (0.435g NuSieve agar, 0.215g SeaKem agar, 65ml 1X TAE) were utilized, each containing 5 l 0.1% ethidium bromide. Electrophoresis was conducted at 125mV for approximately fifty minutes with 5 l lane marker (Promega, catalog # G316A) used in each gel to ensure that products were of the expected size. QIAquick Gel Excision kits (Qiagen, catalog #28706) were used to “clean-up” amplification products. Bands to be sequenced were excised with a scalpel, which was washed with 10% bleach and 70% ethanol between each use to prevent crosscontamination. Bands were placed in 1.5ml Eppendorf tubes and weighed using an analytical balance. Bands were re-suspended in GC Buffer at a 3:1 buffer volume : band


31 mass ratio. Dissolved bands were transferred to spin columns, washed with PE Buffer, and centrifuged in microcentrifuge at 20,000 x g for 2 minutes. Nucleic acid was eluted from filter by the addition of 30 l Buffer EB and subsequent centrifugation at 8,000 x g These products (5 l) were analyzed on a second 1% agarose gel along with 2 l and 4 l lane marker in order to provide a rough estimate of dsDNA concentration. One l lane marker contains roughly 7ng dsDNA. Between 5ng and 20ng are required for sequencing. By comparing the band intensity of the “clean-up” product to that of the two lane markers it is possible to estimate the amount of dsDNA present. This “quasiquantification” step is crucial to determine the amount of template to add to the sequencing reaction. Insufficient or excess quantities of dsDNA will adversely affect the sequencing reaction. Nucleotide Sequencing Nucleotide sequencing of RT-PCR amplic ons was performed with a BeckmanCoulter (Fullerton, CA) CEQ8000 Automated Sequence Analysis System in accordance with manufacturer’s protocol. It should be noted that although the enteroviruses are ssRNA viruses, it is the RT-PCR dsDNA amplic on that is being analyzed. Therefore DNA, not RNA, is sequenced. Briefly, this instrument uses the Sanger method of dyeterminators incorporated into an enzyme-m ediated amplification reaction. Bases with these dye-terminators effectively halt polym erization in a PCR-like reaction. Linear products of varying lengths (ranging from 1 to n where n is the length of the amplicon) are obtained. This product is precipitated and loaded onto the instrument whereby it is passed through a poly-acrylamide gel and separa ted based on size. A laser within the


32 instrument detects the wavelength of each dye and makes a “base call.” Each dye is covalently-linked to a specific base [black = G, red = T, blue = C, green = A]. Software computes the sequence, based on user-supplied parameters, which can then be exported and stored in a database. Sequencing reactions were carried out using DTCS Quick Start kits (BeckmanCoulter, catalog # 608126). Twenty l reactions were prepared in either 48or 96-well plates. Thermal cycling was performed in a PE9700 thermal cycler. Appendix I includes master mix components and thermal cycling profile. Glycogen (20mg/mL), 3M Sodium Acetate, and 100mM EDTA were comb ined in a 1:2:2 ratio. Five l of this “stop solution” were added to all wells to ensure product extension was terminated. Well contents were then ethanol-precipitated to re move salts and other debris. One wash was performed with cold (-20 C) 95% ethanol prior to centrifugation at 6,000 x g for 5 minutes. Two additional washes with cold (-20 C) 70% ethanol and two centrifugations at 6,000 x g for 3 minutes were performed. Forty l SLS solution (included with DTCS Quick Start kit; SLS is a de-ionized formamid e solution) was added to each well. After 10 minutes incubation at room temperature, plates were vortexed lightly and each well was overlayed with mineral oil. Plates were added to instrument and run was initiated. Method LFR (long fragment read) was used for all runs. Sequence Analysis Sequence data was exported to the Seqm anII module of the Lasergene software suite (DNAStar, Madison, WI) in the form of a “.scf” file. These files contain not only the analyzed data (“base calls”), but the ra w sequence data as well (“ferragrams”). The


33 SeqmanII module is an assembly program that allows the assembly of contigs, which are consensus sequences derived from bi-direc tional sequencing using both the forward and reverse Oberste A primers, 292 and 222 respectively. This provides a level of quality control and allows for better coverage of the amplicon region. Assembled contigs were saved as “.seq” files which can be used in all modules of Lasergene. Two methods were compared for sequence alignment, both of which utilize the Clustal W (slow/accurate option) algorithm. One of these, Megalign, is included within the Lasergene suite. The other is located on a freely available, web-based application (URL: ). These are both progressive pair-wise alignment algorithms, which attempt to align homologous bases and insert gaps where homology is not presen t [Note: The Oberste method developed at CDC uses a more sophisticated alignment algorithm, however, it requires a dedicated UNIX-based server and a $20,000+ software package]. Alignments were used to create phyloge netic trees. Again, two methods were compared. The Megalign module automatically constructs a phylogenetic tree when an alignment is produced. Another freel y-downloaded software program, MEGA (molecular evolutionary genetic analysis), was also used. This program is available at URL: The MEGA program allows the user to specify preferences, while the Megalign module doe s not. A neighbor-joining tree was used because, unlike the UPGMA (Unweighted pa ired group mean arithmetic) tree option, a constant rate of evolution is not assumed. This is the more appropriate model for RNA viruses because the rate of evolution is not constant and, in fact, differs considerably among different genomic regions (of the same virus) due to variation of selection


34 pressures (23). Phylogenetic trees from bot h methods were evaluated for accuracy of branch points and phyletic clusters. The use of basic local alignment search tool (BLAST) searches to identify enterovirus serotypes was also evaluate d. BLAST utilizes an NIH-supported, freely available sequence database (l ocated at the following URL: ). Briefly, this program compares any sequence data (nucleotide or protein) against an e normous database that is maintained by the National Library of Medicine, commonly refe rred to as GenBank. Sequences obtained were queried against GenBank and evaluated for accuracy.


35 Results Specific Aim #1 (Descriptive Epidemiology) Enteroviruses have been routinely isolated at the Florida Department of Health – Tampa Laboratory (FL DOH – Tampa) in the past several decades. Submission of samples for diagnostic culture has resulted in the isolation of over two thousand enteroviruses since 1981. The laboratory has been engaged in virus isolation since the 1960s; however, data was unavailable for years prior to 1981. The distribution of serotypes isolated at FL DOH Tampa is representative of national trends. A total of fifty serotypes were isolated (fifty-five if the re-classified E8, E10, E22, E23, and E34 are counted). The CBVs were isolated somewhat frequently. Other serotypes isolated quite regularly included: CA9, E6, E7, E9, E11, E30, and E31. The PVs were isolated quite often, but this is not indicative of the prevalence of illness. Inoculation with the Sabin, live-attenuated vaccine led to most (if not all) of these viral isolates. The three non-pathogenic PV strains readily propagate in cell culture, but do not usually cause human illness. Figure 6 shows the distribution of enterovirus serotypes isolated at FL DOH – Tampa since 1981. The distribution of serotypes isolated is inherently biased. Isolates causing more serious illnesses will tend to be over represented. Obviously, asymptomatic infections will not result in viral isolations because specimens will not be submitted. Also, the ease


36 with which serotypes replicate in vitro will skew isolate distribution. This is evident by the high number of CBV isolates and the low number of higher numbered CAVs and echoviruses. Enteroviruses were numbered chronologically, therefore, it follows that the higher numbered serotypes may often be more difficult to isolate and identify. Echovirus 30 is a notable exception, but its higher rate of isolation is likely due to its frequent association with meningitis. Figure 6: Serotype Distribution 0 50 100 150 200CA2 CA3 CA4 CA5 CA6 CA7 CA8 CA9 CA10 CA12 CA13 CA14 CA16 CB1 CB2 CB3 CB4 CB5 CB6 E1/8 E2 E3 E4 E5 E6 E7 E9 E10 E11 E12 E13 E14 E15 E16 E17 E18 E19 E20 E21 E22 E22/23 E23 E24 E25 E26 E27 E30 E31 E33 E34 EV68 EV70 EV71 P1 P2 P3SerotypeIsolates The distribution of enterovirus serotypes isolated at FL DOH – Tampa, 1981-2002 (humans only). Serotype determined by serum neutralization. Sero types not represented indicate no isolations over the time period. E10, E22, E23, and E34 are included since they were still considered unique serotypes at the time of isolation. E22/E23 represents samples that could not be differentiated due to their similar characteristics. Enterovirus isolation has been a significant portion of the workload at FL DOH – Tampa between 1981 and the mid 1990’s. Between 1981 and 2002 an average of just under ninety-eight enteroviruses were is olated per year. From 1989-1991 FL DOH – Tampa was involved in a research study that resulted in a large increase in specimen submission. Since 1995, there has been a decrease in specimens submitted for


37 enterovirus isolation. This drop is almost certainly not a result of lower prevalence and several explanations are described in the following section. Figure 7 shows the number of enterovirus isolations by year. Enteroviru ses have been isolated from nearly fifteen percent of all non-herpes diagnostic viral isolation specimens since 1981. Figure 7: Enterovirus Isolations by Year 0 50 100 150 200 250 3001981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002YearIsolates Enteroviruses isolated at FL DOH – Tampa, 1981-2002. The large increase in isolations from 1989-1991 corresponds to years that the laboratory was involved in a research study evaluating septic systems. This was a prospective study comparing enterovirus isolations from asymptomatic patients in two groups, as defined by type of water treatment. Figure 8 depicts enterovirus isolations as a percent of all non-herpes virus isolation diagnostics at FL DOH Tampa. Specimens positive for enteroviruses are primarily stool or CSF (which are the clinical standard). It is quite common to obtain both stool and CSF from the same patient. At least seventy-five percent of all enteroviruses isolated were from these two types of specimens. Figure 9 displays the large variety of specimen types from which en teroviruses have been isolated at FL DOH – Tampa. Seasonality of enterovirus isolation is depicted in Figure 10. July through November appears to be the height of “enterovirus season” in Florida, although


38 September had lower totals. Figure 8: Enterovirus Diagnostic Isolations 0% 5% 10% 15% 20% 25% 30% 35%1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002YearPercent Figure 8 shows enterovirus isolates as a percentage of all human diagnostic results at FL DOH – Tampa, 1981-2002. Includes virus isolation samples submitted during the time pe riod, excluding herpes samples. Figure 9: Specimen Submission by Type 62% 13% 2% 6% 4% 11% 2% Feces CSF Tissue Necropsy Throat Nasal Not Specified Other Specimen types from which enteroviruses were isol ated (N=2150) among specimens submitted for virus isolation to FL DOH – Tampa, 1981-2002.


39 Figure 10: Enterovirus Isolation by Month 0 50 100 150 200 250 300 JanFebMarAprMayJunJulAugSepOctNovDecMonthIsolates Enterovirus isolation totals for each month (based on date of serum ne utralization assay results) at FL DOH – Tampa, 1981 – 2002. Specific Aim #2 (RT-PCR Assay Comparison) Relative to the standard screening assay, none of the VP1 assays ( Oberste A/B Casas 1st/2nd/Nested Caro ) performed as well in detecting the presence of enterovirus RNA. Specificity and positive predictive value (PPV) were all 100% because the panel of isolates used was all (originally classified as) enteroviruses. Only two isolates (Reovirus and Hepatitis A) did not amplify with the “screening” primer set, which anneals to the highly conserved 5’NTR re gion, and none of the five VP1-annealing primer sets amplified any of these samples. Therefore, sensitivity and negative predicative value were more indicative of utility since these values did exhibit variation. Figure 11 illustrates the predictive values of each primer set. Clearly, the performance of the Oberste A oligonucleotide primer set was superior to the others. This primer set had the highest sensitivity (85.18%) and highest NPV (20.00%). Only one isolate (PV3) was amplified with another VP1 primer set that failed to amplify with the Oberste A set.


40 Table 5 shows a side-by-side comparison of all RT-PCR results from the enterovirus panel and the two-by-two tables are included in appendix II. Figure 11: RT-PCR Assay Comparison 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%Oberste A Oberste B Casas 1st Casas 2nd Casas Nested CaroAssayPercentage Sensitivity Specificity PPV NPV Predictive values obtained from 2 x 2 tables relative to “screening” assay. Each primer set was used to amplify RNA from a panel of fifty-six virus isolates The “screening” primers anneal to the highlyconserved 5’NTR and amplify virtually all enterovirus serotypes.


41 Table 5: RT-PCR Results Virus S OA OB C1 C2 CN R Virus S OA OB C1 C2 CN R CA2 + + # Reo CA3 + + + + E11 + + + + # + CA4 + E11' + # + + # + CA5 + + + E12 + + + + + CA6 + + E13 + + + + + + # CA7 + + + + + E14 + + + # CA8 + + + + # E15 + # + + + CA9 + + + + + E16 + + # + + CA10 + # E17 + + + + + + + CA13 + + # # # E18 + + + + + + CA14 + + + + E19 + + + + CA16 + # # # E20 + + + + + # + CB1 + + + + # E21 + + + + CB2 + + + # E22 + CB3 + + E23 + + + # + CB4 + + # E24 + + + + CB5 + + + + # E25 + # # + + CB6 + + + + + E27 + + + + + + E1 + + # E30 + + + + + + E2 + + + # + + + E31 + E3 + + + # E33 + + + E4 + + + + + + + EV68 + E5 + + + + # + + EV70 + # # E6 + + + + + + + EV71 + E6" + + # + + HAV E7 + + + + # + P1 + + + E8 + + + # P2 + + + + + E9 + + + + + + + P3 + + + Notes: A = Oberste A; OB = Oberste B; C1 = Casas 1st; C2 = Casas 2nd; CN = Casas Nested; R = Caro; S = Screening + = strong positive; # = weak positive; = negative Specific Aim #3 (Sequence Analysis) Nucleotide sequencing was attempted with all amplicons obtained from RT-PCR with the Oberste A primer set (N=46). Sequence results were obtained from forty-three amplicons (93.47%). Of these forty-three sequences, thirty-four (79.07%) correctly matched the original SN data. The nine that were discrepant were submitted to CDC for confirmatory analysis. All molecular results obtained at FL – DOH concurred with CDC analysis. Several possibilities for this apparent discrepancy are explored in the next section.


42 BLAST scores above 165 and E-values below e-40 are quite reliable for enterovirus typing based on observations obtai ned in this study assuming only the highest score is used. Often several serotypes will match at varying scores. The output of BLAST generates a list of possible matches (referred to as “hits” in the module) based on the query sequence. This output gives a graphical depiction of the alignment, a score, and an e-value. A graphic is provided depicting the alignment of the user-supplied query sequence and where it matches to the GenBank database sequence. Alignments with large interruptions (gaps) should be acknowledged with skepticism. The score is computed based on the length of the alignment and the absence of gaps. Thus, the higher a score is, the better the match. The e-value represents the odds that a given match could have occurred by chance. Consequently, as “E” approaches zero, the confidence in the match approaches 100%. The alignment can also be viewed “base-by-base.” This base-by-base match percentage should be used cautiously because this percentage does not incorporate gap lengths or length of alignment. For example, the CA8 query sequence aligned with 87.8% homology to the GenBank CA8 sequence, but the score was only 76; the alignment was only 65/74 bases. Table 6 depicts Blast results obtained in this study.


43 Table 6: BLAST Results DOH SN Isolated Length BLASTScoreE-Value Match % CDC ID CA10 1991 337 CA10 494 1.00E-138 285/297 96.00% CA10 CA14 1987 330 CA4 363 1.00E-97 279/311 89.70% CA4 CA2 1990 320 CA2 165 9.00E-40 243/297 81.80% CA2 CA3 1986 314 CA3 248 7.00E-65 258/302 85.40% CA3 CA5 1981 N/A N/A N/A N/A N/A N/A N/A CA6 1978 302 CA10 482 1.00E-135 276/287 96.20% CA10 CA7 1982 336 CB2 478 1.00E-134 287/302 95.00% CB2 CA8 1991 305 CA8 76 8.00E-13 65/74 87.80% CA8 CA9 1993 341 CA9 299 3.00E-80 289/335 86.50% CA9 CB1 1995 309 CB1 363 1.00E-97 216/227 95.20% CB1 CB2 1995 353 CB2 402 1.00E-111 105/339 90.00% CB2 CB3 1994 292 CB3 424 1.00E-118 262/278 94.20% CB3 CB4 1993 321 CB4 422 1.00E-115 274/293 93.50% CB4 CB5 1992 318 CB5 502 1.00E-141 283/293 96.60% CB5 CB6 N/A 348 CB6 640 0 326/327 99.70% CB6 E1 1993 330 E8 186 3.00E-46 269/326 82.50% E8 E11 1992 347 E11 607 1.00E-172 330/338 97.60% E11 E11' 1989 325 E11 509 1.00E-143 307/321 95.60% E11 E12 1989 325 E12 192 5.00E-48 244/293 83.30% E12 E13 1985 319 E13 194 6.00E-47 254/306 83.00% E13 E14 1992 321 E14 204 7.00E-50 256/307 83.40% E14 E15 1990 N/A N/A N/A N/A N/A N/A N/A E16 1991 356 E16 288 9.00E-77 290/339 85.50% E16 E17 1983 274 E17 406 1.00E-112 262/278 94.20% E17 E18 1994 326 E18 474 1.00E-131 299/319 93.70% E18 E19 1987 315 E19 256 4.00E-67 219/249 88.00% E19 E2 1990 357 E2 232 7.00E-60 285/341 83.60% E2 E20 1990 322 E30 603 1.00E-171 316/321 98.40% E30 E21 1994 319 E21 392 1.00E-108 276/302 91.40% E21 E23 1987 360 E30 628 1.00E-177 336/341 98.50% E30 E24 1981 330 E30 571 1.00E-162 306/311 98.40% E30 E25 1991 358 E25 246 2.00E-62 275/324 84.90% E25 E27 1991 316 E9 347 7.00E-93 277/324 85.50% E9 E3 1993 328 E3 543 1.00E-152 301/310 97.10% E3 E30 1993 326 E30 591 1.00E-168 310/314 98.70% E30 E33 1984 327 E30 551 1.00E-156 306/314 97.50% E30 E4 1988 N/A N/A N/A N/A N/A N/A N/A E5 1994 359 E5 287 6.00E-75 286/333 85.90% E5 E6 1994 311 E6 198 8.00E-50 260/312 83.30% E6 E6" 1989 327 E6 299 3.00E-80 61/70 87.10% E6 E7 1993 337 E7 577 1.00E-163 315/323 97.50% E7 E8 1973 313 E8 174 8.00E-42 229/276 83.00% E8 E9 1994 319 E9 315 3.00E-83 273/307 88.90% E9 EV70 1988 348 E9 337 2.00E-91 296/338 87.60% E9 P1 1991 316 P1 545 1.00E-154 295/299 98.70% P1 P2 1990 356 P2 662 0 337/338 99.70% P2


44 This method of analysis has correctly identified several isolates this year at FL DOH – Tampa. Two College of American Pa thologists performance evaluation samples (VR9 and VR11) were identified as CB5 and E11, respectively, in concordance with SN results. Two recent clinical diagnostic isolates were correctly identified as E9, which again corresponded to SN results. An additional sample, which had been indeterminate via SN, was determined to be an E9 isolate. Examination of the phylogenetic tree produced by Megalign for the isolates used in this study reveals several key relationships. For the most part, the alignments appear to be accurate relative to published data. Due to varying lengths of sequences obtained, a 287bp segment was examined. All fourteen CB5 isolates cluster together [note: CB5 and clinical-n are the same sequence]. Both of the recent diagnostic samples, clinical-a and clinical-b, cluster with E9, as do the erroneous, according to BLAST results, E27 and EV70. The cluster of CA7 with CB2 also concurs with BLAST results. The E30 cluster contains E20, E23, E24, and E33, again confirming BLAST results. Finally, the parings of CA6 with CA10 and CA14 with CA4 also support BLAST results. Thus, the phylogenetic tree supports the BLAST results that are discordant with the SN data. The homologous serotypes E1/E8, E6/E6”, and E11/E11’ all cluster tightly with one another. The HEV-A, HEV-B, and PV species branch at appropriate locations. The HEV-C and HEV-D species (higher numbered CAVs and EV68/EV70) are not represented among the isolates assayed.


45 Figure 12: Megalign Phylogenetic Tree Phylogenetic tree created in Megalign of a secondary alignment of 287 bases created from initial alignment in order to correct for variation in sequence lengths. Nucleotide Substitutions ( x100 ) 0 61.4 10 20 30 40 50 60 CB5 clinical_n clinical_l clinical_m clinical_o clinical_c clinical_d clinical_e clinical_k clinical_j clinical_g clinical_h clinical_i clinical_f CB1 CB4 CB3 CA7 CB2 CB6 E20 E33 E30 E24 E23 E21 E25 E1 E8 E13 E6 E6'' E5 E12 E3 E7 E19 CA9 E11 E11' E18 E17 E2 E27 E9 EV70 clinical_a clinical_b E16 E14 P2 P1 CA3 CA8 CA6 CA10 CA14 CA2


46 The phylogenetic tree produced with ClustalW and MEGA was quite similar to that obtained with the Megalign module. No secondary alignment was necessary because the MEGA software program does not count gaps in its similarity analysis. Thus, sequences of varying lengths can be readily compared for relatedness. The three species (HEV-A, HEV-B, PV) branched in the same manner as the Megalign tree. All of the CB5 isolates clustered tightly with one another. The discordant results clustered in the same patterns. E70 and E27 clustered with E9, while E20, E23, E24, and E33 clustered with E30. The three homologous pairs (E1/E8, E6/E6”, and E11/E11’) again grouped with one another. Some slight differences were observed. For instance, E17 branched with E18 off the E9 cluster, while it had been an orphan branch between HEV-B and PV in the Megalign tree. Also, CA9 branched off of the CB5 cluster rather than from E7. These differences are likely due to differences in the algorithm used to construct the tree. Published results of VP1 phylogenetic analyses suggest that portions of both trees are correct. CA9 should cluster most closely with E7, E19 E11, and E11’, which is observed with the Megalign (but not the MEGA) tree (61). On the other hand, E17 should cluster most closely with E18, which is observed with the MEGA (but not the Megalign) tree (64). Without sufficient replicates of each serotype (as in the CB5 isolates) it is impossible to determine which phylogenetic tree most closely resembles the exact genetic relationship of these serotypes. Finally, the inclusion of sequence data from the fifteen serotypes not represented in this study would greatly enhance the accuracy of both trees.


47 Figure 13: MEGA Phylogenetic Tree CB5 clinical n clinical o clinical m clinical l clinical k clinical e clinical g clinical j clinical c clinical d clinical f clinical h clinical i CA9 CB3 CB1 CB4 CB6 CA7 CB2 E6 E6 2 E13 E3 E12 E7 E19 E11 E11 2 E16 E5 E14 E18 E17 clinical a clinical b EV70 E27 E9 E25 E21 E24 E23 E30 E20 E33 E1 E8 P1 P2 CA2 CA14 CA10 CA6 CA3 CA8 0.1 Neighbor-joining tree created in MEGA; bootstrap valu e = 100 psuedo-replicates; Kimura-2 parameters. Thirteen CB5 isolates, originally isolated between 1985 and 1992, were


48 sequenced and aligned using the Megalign module. Similar studies have been conducted using as few as seven isolates (43). A performance evaluation sample from 2003 (VR9) was included in the analysis as a reference. Due to varying lengths of sequences obtained, a 256bp segment was examined for overall homology. Similarity indexes among the thirteen DOH isolates ranged from 91.4% and 98.8%. Interestingly, the VR9 isolate (whose BLAST and SN results indicate CB5 serotype) was only between 75.8% and 77.7% similar to the thirteen DOH isolates. It is probable that this is the prototype strain and does not reflect the divergence that has occurred since its initial isolation in the 1950’s. There were 12 C A substitutions, 38 G A substitutions, 8 T A substitutions, 3 G C substitutions, 62 T C substitutions, and 6 T G substitutions. The higher totals for G A and T C are expected since purine-purine and pyrimidine-pyrimidine substitutions are more readily tolerated due to similar structures. The entire alignment report of the clinical isolates is shown in Appendix II. Ten serial passages of E9 were performed to examine the sensitivity of the sequencing analysis and to determine how frequently mutations occur in vitro The relatively high error frequency of the enterovirus RNA polymerase coupled with the bottle neck effect of sequential passages was expected to produce numerous base substitutions. Due to varying lengths of sequences obtained, a 259bp segment was examined for overall homology. Only one base change was observed, an A-G substitution after the third passage. Figure 14 shows a portion of the alignment report for the ten passages.


49 Figure 14: Portion of E9 Passage Alignment Portion of the alignment report created in Megalign from 10 sequential passages of an E9 isolate. The A-G transition (which occurred between the third and fourth passages) is seen at position 201 of the alignment. + Majorit GATCACTAGTTACCAACATCGAGGAACAAT Majority 190 200 210 GATCACTAGTTACCAACATCGAGGAACAAT 271 Passage 9 GATCACTAGTTACCAACATCGAGGAACAAT 277 Passage 1 0 GATCACTAGTTACCAACATCAAGGAACAAT 181 Passage 2 GATCACTAGTTACCAACATCAAGGAACAAT 268 Passage 3 GATCACTAGTTACCAACATCGAGGAACAAT 274 Passage 4 GATCACTAGTTACCAACATCGAGGAACAAT 274 Passage 5 GATCACTAGTTACCAACATCGAGGAACAAT 274 Passage 6 GATCACTAGTTACCAACATCGAGGAACAAT 274 Passage 7 GATCACTAGTTACCAACATCGAGGAACAAT 274 Passage 8 GATCACTAGTTACCAACATCAAGGAACAAT 273 Passage 1


50 Discussion Enterovirus infections produce serious illness throughout the world and surveillance for them should be an important function of public health laboratories. The “gold standard” assay for identification of enteroviruses has considerable drawbacks. The test is labor intensive, requires long “diagnostic lags,” and often fails to yield conclusive results. Due to these and other factors, specimen submissions to FL DOH – Tampa for enterovirus isolation and identification have dropped dramatically in recent years. By no means, however, should this be interpreted as a decrease in either prevalence of enteroviral illness or transmission. The workload in the virology section of FL DOH – Tampa has evolved in the past several years. West Nile Virus, Norovirus, and Herpesvirus testing has soared in recent years. For instance, between 1997 and 2002, over 1700 specimens (49.76% of diagnostic virus isolation submissions) were determined to be positive for Herpesvirus via diagnostic virus isolation. Yet, between 1981 and 1996, less than 200 specimens (1.31% of submissions) were positive using the same assay. The advent of commercially available molecular reagents and kits has transformed virology labs across the country. This is evident at FL DOH – Tampa, where molecular assays for West Nile and Norovirus detection have been added since 1999. Currently, Influenza and SARS surveillance are also being conducted using molecular methods. These assays are performed not only as clinical diagnostics, but also to provide surveillance data. Clearly,


51 public health laboratories are dynamic institutions that must be capable of responding to the needs of the citizenry. Perhaps one of the biggest factors involved in the considerable decrease in enterovirus diagnostic specimens at FL DOH – Tampa is the turnaround time for reporting of results. To isolate and identify an enterovirus takes at least two weeks even under optimal circumstances. This assumes: A sample is received late in the week and appropriate culture tubes (BGM/RMK/MRC-5) are available. The specimen produces CPE within one week so that the titration and serum neutralization can be performed the following Friday No cross-neutralization occurs and the SN results are unambiguous. The enterovirus is one of the forty-two serotypes included in the WHO anti-sera pools The use of monotypic anti-sera to individual serotypes allows identification of the remaining twenty-two serotypes, but adds considerable time and cost to the procedure. The presence of mixed serotypes, aggregation of virus particles, and the use of multiple passages to produce CPE can (and often do) affect the outcome of the SN assay. When such complications arise the turnaround time can easily increase to well over one month. At this point, the diagnosis is of little clinical utility. Often clinicians simply do not bother to order the test. Several other factors also help explain the decrease in enterovirus specimen submission in recent years. The re-classification of meningitis to a “non-reportable” disease is likely responsible for at least a portion of the decline. Lack of treatment


52 options renders the determination of serotype a moot point in the minds of many clinicians. The reasoning being, “why order a test if the result will not affect patient outcome?” Health insurance companies may not reimburse for such tests using the same logic. From a public health standpoint, it makes sense to be able to identify enterovirus serotype accurately and quickly. Outbreaks of EV71 (hand-foot-and-mouth-disease) in Southeast Asia and E30 (meningitis) throughout Europe and the Americas demonstrate the ability of the enteroviruses to produce significant morbidity and mortality. Early recognition of infection patterns allows a more rapid and effective public health response. This thesis evaluated several molecular protocols capable of identifying enterovirus serotype. Amplicon sequencing of RT-PCR products of the VP1 region of the enterovirus genome has been demonstrated as a reliable method of identifying enteroviruses. The nucleotide sequence of th e VP1 region correlates highly with serotype because a large percentage of neutralization sites are located on the VP1 portion of the enterovirus capsid. The Oberste A oligonucleotide (292/222) exhibited the best overall performance. When compared to the screening assay (which is highly sensitive and specific, but produces an amplicon whose nucleotide sequence does not correlate with serotype) this primer set produced higher sensitivity and higher negative predictive value than all other VP1 primer sets. Subsequently, these amplicons were sequenced and evaluated for their ability to predict serotype. Of the fifty-four serotypes amplified by the screening primers, forty-six amplified with the Oberste A primers (85.19%). The E22 isolate was not expected to amplify, but the other seven negatives may be a result of individual strain variability, because the


53 Oberste A primer set has been reported to be capable of amplifying all prototype strains (64). Forty-three sequence results were obtained (93.48%) with these primers. Two isolates, E4 and E15, produced only very faint bands via gel electrophoresis in three separate attempts; thus, amplicons were “unsequencable.” Another isolate, CA5, produce a very intense amplicon with a secondary band slightly above it (~550bp). Nucleotide sequencing yielded a “peak-under-peak” ferragram, which is characteristic of multiple templates. Therefore, the CA5 isolate may actually be a combination of multiple serotypes, which was not revealed in the initial SN assay. Overall, there was good correlation between sequencing results and SN results. Thirty-four of the forty-three (79.07%) sequencing results yielded equivalent results when BLAST searches were employed. The use of BLAST searches to identify serotype seems to be a useful and reliable method for identifying enteroviruses. Construction of phylogenetic trees yielded slightly different results when two tree creation algorithms were employed. Discrepant results were repeated (re-extr action of original isolate, new RT-PCR, new sequencing) and sequences were sent to CDC for confirmation. Results reported reflect those obtained from the second anal ysis. Confirmation by CDC concurred with molecular results, not SN, in all cases. Of the results in which a discrepancy arose, several explanations are possible. Contamination is always a possibility when performing molecular assays. However, cell culture, extraction, and amplification negative controls all performed as expected. If contamination were the source of error it would be expected that only one or two serotypes would be represented, not five (E30, CA10, CB2, CA4, and E9). In addition,


54 no divergence would be expected in the phylogenetic trees (as in CB5 and clinical n, which are the same sample). The SN-determined EV70 isolate clustered most closely with E9, while in published results EV70 should have branched separately from all isolates used in this study due to its membership in the HEV-D species (71). It is possible that the initial SN results were inaccurate, archive records were mislabeled, or these isolates were actually multiple serotypes but when resurrected from archive and passaged, one serotype propagated preferentially. One isolate, E23, was almost certainly archived incorrectly. The CPE was observed in BGM culture was not consistent with that expected for E22/E23. Furthermore, the VP1 region in these two Picornaviruses is highly divergent from other enteroviruses and amplification should not have occurred with any of the VP1 oligonucleotide primers. A CDC official states that the discrepancy between SN and molecular results is “not surprising” (60). Time limitations prevented these nine isolates from being “re-serotyped” using the serum neutralization assay at FL DOH Tampa. Despite these discrepancies, the advantages of using sequencing in place of serum neutralization cannot be denied. Nucleotide sequence data provides a wealth of data that could be used in molecular epidemiology studies in the future. Results of both the CB5 isolate alignment and the E9 passage experiment demonstrate the utility of sequence data. The thirteen CB5 isolates obtained were shown to be closely-related to one another, yet quite divergent from the performance evaluation sample (prototype CB5 strain). It is quite possible that CB5 strains circulating in Flor ida are divergent from CB5 strains in other areas. The bi-directionality of nucleotide substitutions with respect to time likely indicates that multiple strains are present which evolve independently. A directional


55 change in base substitutions with respect to time would have indicated a continually circulating CB5 strain. Although amino acid sequence analysis was not conducted (as it was beyond the scope of this project), it would be of interest to examine which base substitutions resulted in codon changes. Sequence analysis provides a means of analyzing intra-serotypic variation that is simply not possible wit the serum neutralization assay. This is evident in the next experiment, although conservation, not divergence, is observed. The E9 passage experiment demonstrates both the accuracy of sequencing and the inherent stability of viruses cultured in vitro Only one substitution was observed in the amplicon across ten passages. All sequence data aligned perfectly demonstrating the robustness of sequence analysis, i.e. the same sequence (not including the substitution) was obtained from all ten isolates. Due to the high error rate of the enterovirus RNA polymerase, it was expected that more substitutions would have occurred. Perhaps the amplicon sequence under investigation is involved in receptor recognition and/or binding and, thus, is conserved. It is also possible that the passages required to produce consistent, strong CPE (five) produced a strain that was uniquely adapted to in vitro culture. However, even if the both of the above were true some substitutions could be tolerated due to the degeneracy of the genetic code. It is likely that ten passages simply was not enough to produce significant change in this portion of the enterovirus genome. After all, it took Albert Sabin hundreds of passages to produce the avirulent phenotypes of the three polioviruses used in the live, attenuated oral polio vaccine (50). These “mini experiments” do not yield a ny significant findings on their own; they actually raise many more questions than they answer. However, it cannot be denied that


56 sequence data allows analyses to be performed that are simply not possible with serum neutralization. Aside from the direct benefits of a molecular-based serotyping protocol relative to serum neutralization, several intangible benefits exist. Through the course of this study, Dr. Steven Oberste at CDC has offered guidan ce and support. He has supplied copies of his protocol, offered “trouble-shooting” advice, and performed confirmatory analysis of results. This enhancement of the working relationship between CDC and FL DOH – Tampa can only be viewed as an asset. In addition, all reagents required to perform this assay are available commercially (Qiagen or Beckman-Coulter). Reagent cost to perform the sequencing assay was less than twenty dollars per specimen. Considerable “hands on” technician time is required; roughly, six hours is needed to perform all tasks from harvest of CPE-producing cultures to sequence data analysis. This is somewhat higher total time than the “hands on” time required to perform the serum neutralization assay (approximately three hours). These costs are tolerable when the benefits of a molecular-based approach are considered. The method described in this study could easily be incorporated into routine use at a public health virology laboratory. With the purchase of an appropriate oligonucleotide probe, a real-time screening assay (using the 5’NTR primers) could be incorporated into diagnostic assays already performed for other viruses (West Nile, Influenza, St. Louis Encephalitis). Such probes have been described in the literature (103). Appendix IV provides an algorithm for implementing such a protocol. This study demonstrates that nucleotide sequencing of enterovirus isolates, coupled with BLAST searches, allows for accu rate and time-effective identification of


57 serotype. A thermal cycler, gene sequencer, and an Internet browser are the only hardware required. Expensive software progr ams, while beneficial, are not required to accomplish serotype identification. In 2003, this method has correctly identified all clinical isolates received, as well as two performance evaluation samples. The archive holdings of enterovirus isolates at FL DOH – Tampa are extensive and a more complete database of VP1 sequences will be created. This will provide the opportunity to carry out larger-scale molecular epidemiology studies, which are useful in public health outbreak surveillance. The forty-eight to seventy-two hour “turn-around” time eliminates the diagnostic lag inherent with the serum neutralization assay. The capability of performing enterovirus identifications faster and from a larger number of specimens will improve the infectious disease surveillance capabilities of the Florida Department of Health.


58 References 1. Abzug, M. J. 1995. Perinatal Enterovirus Infections, p. 221-238. In H. A. Rotbart (ed.), Human Enterovirus Infections. ASM Press, Washington, DC. 2. Agol, V. I., G. A. Belov, K. Bienz, D. Egger, M. S. Kolesnikova, N. T. Raikhlin, L. I. Romanova, E. A. Smirnova, and E. A. Tolskaya. 1998. Two Types of Death of Polio-Infected Cells: Ca spase Involvement in the Apoptosis but Not Cytopathic Effect. Virology 252: 343-353. 3. Airaksinen, A. 2000. The VP1 Intracapsid Hook and Uncoating of Enteroviruses. University of Helsinki, Helsinki. 4. Andino, R., N. Boddeker, D. Silvera, and A. V. Gamarnik. 1999. Intracellular Determinants of Picornavirus Replication. Trends in Microbiology 7: 76-82. 5. Arola, A., J. Santti, O. Ruuskanen, P. Halonen, and T. Hyypia. 1996. Indentification of Enteroviruses in Clinical Specimens by Competitive PCR Followed by Genetic Typing Using Sequence Analysis. Journal of Clinical Microbiology 34: 313-318. 6. Bailly, J. L., D. Brosson, C. Archimbaud, M. Chambon, C. Henquell, and H. Pegigue-Lafeuille. 2002. Genetic Diversity of Echovirus 30 During a Meiningitis Outbreak, Demonstrated by Direct Mol ecular Typing from Cerebrospinal Fluid. Journal of Medical Virology 68: 558-567. 7. Berchoff, R. P. 1987. Picornaviruses at the Molecular Level, p. 197-215. In R. P. Berchoff (ed.), The Molecular Basis of Viral Replication. Plenum Press, New York. 8. Bernhardt, G., J. Harber, A. Ziebert, M. deCrombugghe, and E. Wimmer. 1994. The Poliovirus Receptor: Identification of Domains and Amino Acids Critical for Virus Binding. Virology 203: 344-356. 9. Bourlet, T., V. Caro, S. Minjolle, I. Jusselin, B. Pozzetto, R. Crainic, and R. Colimon. 2003. New PCR Test that Recognizes All Human Prototyoes of Enterovirus: Application for Clinical Di agnosis. Journal of Clinical Microbiology 41: 1750-1752.


59 10. Bourlet, T., J. Gharbi, S. Omar, M. Aouni, and B. Pozzetto. 1998. Comparison of a Rapid Culture Method Combining an Immunoperoxidase Test and a Group Specific Anti-VP1 Monoclonal Antibody with Conventional Virus Isolation Techniques for Routine Detection of Enteroviruses in Stools. Journal of Medical Virology 54: 204-209. 11. Brown, B., M. S. Oberste, K. Maher, and M. A. Pallansch. 2003. Complete Genomic Sequencing Shows that Polioviruses and Members of Human Enterovirus Species C Are Closely Related in the Noncapsid Coding Region. Journal of Virology 77: 8973-8984. 12. Cardosa, M. J., D. Perera, B. A. Brown, D. Cheon, H. M. Chan, K. P. Chan, H. Cho, and P. McMinn. 2003. Molecular Epidemiology of Human Enterovirus 71 Strains and recent Outbreaks in the Asia-Pacific Region: Comparative Analysis of the VP1 and VP4 Genes. Emerging Infectious Diseases 9: 461-468. 13. Carman, B. 2001. Molecular Techniques Should Now Replace Cell Culture in Diagnostic Virology Laboratories. Reviews in Medical Virology 11: 347-349. 14. Caro, V., S. Guillot, F. Delpeyroux, and R. Crainic. 2001. Molecular Strategy for 'Serotyping' of Human Enteroviruses. Journal of General Virology 82: 79-91. 15. Casas, I., G. F. Palacios, G. Trallero, D. Cisterna, M. C. Freire, and A. Tenorio. 2001. Molecular Characterization of Human Enteroviruses in Clinical Samples: Comparison Between VP2, VP1, and RNA Polymerae Regions Using RT Nested PCR Assays and Direct Seque ncing of Products. Journal of Medical Virology 65: 138-148. 16. CDC. 1997. Nonpolio Enterovirus Surveillance United States, 1993-1996. Morbidity and Mortality Weekly Report 46: 748-750. 17. CDC. 2000. Enterovirus Surveillance United States, 1997-1999. Morbidity and Mortality Weekly Report 49: 913-916. 18. CDC. 2002. Enterovirus Surveillance United States, 2000-2001. Morbidity and Mortality Weekly Report 51: 1047-1049. 19. CDC. 2003. Progress Toward Global Eradication of Poliomyelitis, 2002. Morbidity and Mortality Weekly Report 52: 366-369. 20. Chonmaitree, T., and L. Mann. 1995. Respiratory Infections, p. 255-270. In H. A. Rotbart (ed.), Human Enterovirus Infections. ASM Press, Washington, DC. 21. Dagan, R., and M. A. Menegus. 1995. Nonpolio Enteroviruses and the Febrile Infant. In H. A. Rotbart (ed.), Human Enterovirus Infections. ASM Press, Washington, DC.


60 22. DaSilva, E. E., M. T. Winkler, and M. A. Pallansch. 1996. Role of Enterovirus 71 in Acute Flaccid Paralysis After the Eradication of Poliovirus in Brazil. Emerging Infectious Diseases 2: 231-233. 23. Domingo, E., C. M. Ruiz-Jarabo, S. Sierra, A. Arias, N. Pariente, E. baranowski, and C. Escarmis. 2002. Emergence and Selection or RNA Virus Variants: memory and Extinction. Virus Research 82: 39-44. 24. Duggal, R., and E. Wimmer. 1999. Genetic Recombination of Poliovirus In Vitro and In Vivo : Temerpature-Dependent Alteration of Crossover Sites. Virology 258: 30-41. 25. Euscher, E., J. Davis, I. Holzman, and G. J. Nuovo. 2001. Coxsackie Virus Infection of the Placenta Associated with Neurodevelopmental Delays in the Newborn. Obstetrics & Gynecology 98: 1019-1026. 26. Haarmann, C. M., P. L. Schwimmbeck, T. Mertens, H. P. Schultheiss, and B. E. Strauer. 1994. Identification of Serotype-Specific and Nonserotype-Specific B-Cell Epitopes of Coxsackie B Virus Using Synthetic Peptides. Virology 200: 381-389. 27. Haller, A. A., and B. L. Semler. 1995. Translation and Ho st Cell Shutoff, p. 113134. In H. A. Rotbart (ed.), Human Enterovirus Infections. ASM Press, Washington, DC. 28. Hellen, C., and E. Wimmer. 1995. Enterovirus Genetics, p. 25-72. In H. A. Rotbart (ed.), Human Enterovirus Infections. ASM Press, Washington, DC. 29. Hellen, C., and E. Wimmer. 1995. Enterovirus Structure and Assembly, p. 155174. In H. A. Rotbart (ed.), Human Enterovirus Infections. ASM Press, Washington, DC. 30. Herrero, L. J., C. S. Lee, R. J. Hurrelbrink, B. H. Chua, K. B. Chua, and P. C. McMinn. 2003. Molecular Epidemiology of Enterovirus 71 in Peninsular Malaysia, 1997-2000. Archives of Virology 148: 1369-1385. 31. Hyypi, T., T. Hovi, N. J. Knowles, and G. Stanway. 1997. Classification of Enteroviruses Based on Molecular and Biological Properties. Journal of General Virology 78: 1-11. 32. Ishiko, H., Y. Shimada, M. Yonaha, O. Hashimoto, A. Hayashi, K. Sakae, and N. Takeda. 2002. Molecular Diagnosis of Human Enteroviruses by Phylogeny-based Classification by Use of the VP4 Sequence. Journal of Infectious Disease 185: 744-754. 33. Johnson, K. L., and P. Sarnow. 1995. Viral RNA Synthesis, p. 95-112. In H. A. Rotbart (ed.), Human Enterovirus Infections. ASM Press, Washington, DC.


61 34. Karttumen, A., T. Pyry, O. Vaarala, J. Ilonen, T. Hovi, M. Roivainen, and T. Hyypi. 2003. Variation in Enterovirus Receptor Genes. Journal of Medical Virology 70: 99-108. 35. Kew, O. M., M. N. Mulders, G. Y. Lipskaya, E. E. d. Silva, and M. A. Pallansch. 1995. Molecualr Epidemiology of Polioviruses. Seminars in Virology 6: 401-414. 36. King, A. M. Q., Brown, F., Christian, P., Hovi, T., Hyypi, T., Knowles, N.J., Lemon, S.M., Minor, P.D., Palmenberg, A.C., Skern, T. and Stanway, G. 2000. Picornaviridae, p. 657-673. In M. H. V. Van Regenmortel, Fauquet, C.M., Bishop, D.H.L., Calisher, C.H., Carsten, E.B., Estes, M.K., Lemon, S.M., Maniloff, J., Mayo, M.A., McGeoch, D.J., Pringle, C.R. and Wickner, R. B. (ed.), Virus Taxonomy. Seventh Report of the International Committee for the Taxonomy of Viruses. Academic Press, New York. 37. Kopecka, H., B. Brown, and M. Pallansch. 1995. Genotypic Variation in Coxsackievirus B5 Isolates from Three Different Outbreaks in the United States. Virus Research 38: 125-136. 38. Kubo, H., N. Iritani, and Y. Seto. 2002. Molecular Classification of Enteroviruses Not Identified by Neutralization Tests. Emerging Infectious Diseases 8: 298-304. 39. Kunkel, U., and E. Schreier. 2000. Genetic Variability Within the VP1 Coding Region of Echovirus Type 30 Isolates. Archives of Virology 145: 1455-1464. 40. Li, Y., Z. Pan, Y. Ji, T. Peng, L. C. Archard, and H. Zhang. 2002. Enterovirus Replication in Valvular Tissue from Patients with Chronic Rheumatic Heart Disease. European Heart Journal 23: 567-573. 41. Lim, K. A., and M. Benyesh-Melnick. 1960. Typing of Viruses by Combinations of Antiserum Pools. Application to Typing of Enteroviruses (Coxsackie and ECHO). Journal of Immunology 84: 309-317. 42. Lin, T. Y., S. J. Twu, M. S. Ho, L. Y. Chang, and C. Y. Lee. 2003. Enterovirus 71 Outbreaks, Taiwan: Occurrence and Recognition. Emerging Infectious Diseases 9: 291-293. 43. Lindberg, A. M., and C. Polacek. 2000. Molecular Analysis of the Prototype Coxsackievirus B5 Genome. Archives of Virology 145: 205-221. 44. Lukashev, A. N., V. A. Lashkevich, G. A. Koroleva, J. Ilonen, G. G. Karganova, V. I. Reznik, and A. E. Hinkanen. 2003. Molecular Epidemiology of Enteroviruses Causing Uveitits and Multisystem Hemorrhagic Disease of Infants. Virology 307: 45-53.


62 45. Makela, M. J., T. Puhakka, O. Ruuskanen, M. Leinonen, P. Saikku, M. Kimpimaki, S. Blomqvist, T. Hyypi, and P. Arstila. 1998. Viruses and Bacteria in the Etiology of the Common Cold. Journal of Clinical Microbiology 36: 539-542. 46. Manayani, D. J., R. V. Shaji, G. J. Fletcher, T. Cherian, N. Murali, N. Sathish, T. Solomon, C. Gnanamuthu, and G. Sridharan. 2002. Comparison of Molecular and Conventional Methods for Typing of Enteroviral Isolates. Journal of Clinical Microbiology 40: 1069-1070. 47. Mateu, M. G. 1995. Antibody Recognition of Picornaviruses and Escape from Neutralization: A Structural View. Virus Research 38: 1-24. 48. Mayo, M. A., and C. R. Pringle. 1998. Virus Taxonomy 1997. Journal of General Virology 79: 649-657. 49. McMinn, P. C. 2002. An Overview of the Evolution of Enterovirus 71 and Its Clinical and Public Health Significance. FEMS Microbiology Reviews 26: 91107. 50. Modlin, J. E. 1995. Poliomyelitis and Poliovirus Immunization, p. 195-220. In H. A. Rotbart (ed.), Human Enterovirus Infections. ASM Press, Washington, DC. 51. Morens, D. M., and M. A. Pallansch. 1995. Epidemiology, p. 3-24. In H. A. Rotbart (ed.), Human Enterovirus Infections. ASM Press, Washington, DC. 52. Muir, P., U. Kammerer, K. Korn, M. N. Mulders, T. Pyry, B. Weissbrich, R. Kandolf, G. M. Cleator, and A. M. v. Loon. 1998. Molecular Typing of Enteroviruses: Current Status and Futu re Requirements. Clinical Microbiology Reviews 11: 202-227. 53. Nairn, C., and G. B. Clements. 1999. A Study of Enterovirus Isolations in Glasgow from 1977 to 1997. Journal of Medical Virology 58: 304-312. 54. Norder, H., L. Bjerregaard, L. Magnius, B. Lina, M. Aymard, and J. J. Chomel. 2003. Sequencing of 'Untypable' Enteroviruses Reveals Two New Types, EV-77 and EV-78, within Human Enterovirus Type B and Substitutions in the BC Loop of the VP1 Protein for Known Types. Journal of General Virology 84: 827-36. 55. Norder, H., L. Bjerregaard, and L. O. Magnius. 2001. Homotypic Echoviruses Share Aminoterminal VP1 Sequence Homology Applicable for typing. Journal of Medical Virology 63: 35-44. 56. Norder, H., L. Bjerregaard, and L. O. Magnius. 2002. Open Reading Frame Sequence of an Asian Enterovirus 73 Strain Reveals that the Prototype from California Is Recombinant. Journal of General Virology 83: 1721-1728.


63 57. Oberste, M. S. 1998. Complete Sequence of Echovirus 23 and its Relationship to Echovirus 22 and Other Human Enteroviruses. Virus Research 56: 217-223. 58. Oberste, M. S. 1999. Molecular Epidemiology and Genetic Diversity of Echovirus Type 30 (E30): Genotypes Correlate with Temporal Dynamics of E30 Isolation. Journal of Clinical Microbiology 37: 3928-3933. 59. Oberste, M. S. 1999. Typing of Human Enteroviruses by Partial Sequencing of VP1. Journal of Clinical Microbiology 37: 1288-1293. 60. Oberste, M. S. 2003. Personal communication. 61. Oberste, M. S., K. Maher, D. R. Kilpatrick, and M. A. Pallansch. 1999. Molecular Evolution of the Human Enteroviruses: Correlation of Serotype with VP1 Sequence and Application to Picornav irus Classification. Journal of Virology 73: 1941-1948. 62. Oberste, M. S., K. Maher, and M. A. Pallansch. 1998. Molecular Phylogeny of All Human Enterovirus Serotypes Based On Comparison of Sequences at the 5' End of the Region Encoding VP2. Virus Research 58: 35-43. 63. Oberste, M. S., W. A. Nix, D. R. Kilpatrick, M. R. Flemister, and M. A. Pallansch. 2003. Molecular Epidemiology and Type-Specific Detection of Echovirus 11 Isolates from the Americas, Europe, Africa, Australia, Southern Asia and the Middle East. Virus Research 91: 241-248. 64. Oberste, M. S., W. A. Nix, K. Maher, and M. A. Pallansch. 2003. Improved Molecular Identification of Enteroviruses by RT-PCR and Amplicon Sequencing. Journal of Clinical Virology 26: 375-377. 65. Oberste, M. S., D. Schnurr, K. Maher, S. al-Busaidy, and M. A. Pallansch. 2001. Molecular Identification of New Picornaviruses and Characterization of a Proposed Enterovirus 73 Serotype. Journal of General Virology 82: 409-416. 66. Oberste, M. S., K. Maher, M. R. Flemister, G. Marchetti, D. R. Kilpatrick, and M. A. Pallansch. 2000. Comparison of Classic and Molecular Approaches for the Identification of Untypeable Enteroviruses. Journal of Clinical Microbiology 38: 1170-1174. 67. Oglivie, M. 2001. Molecular Techniques Should Not Now Replace Cell Culture in Diagnostic Virology laboratories. Reviews in Medical Virology 11: 351-354. 68. Paananen, A., P. Ylipaasto, E. Rieder, T. Hovi, J. Galama, and M. Roivainen. 2003. Molecular and Biological Analysis of an Echovirus 9 Strain Isolated from a Diabetic Child. Journal of Medical Virology 69: 529-537. 69. Palacios, G. 2002. personal communication.


64 70. Palacios, G., I. Casas, D. Cisterna, G. Trallero, A. Tenorio, and C. Freire. 2002. Molecular Epidemiology of Echovirus 30: Temporal Circulation and Prevalence of Single Lineages. Journal of Virology 76: 4940-4949. 71. Palacios, G., I. Casas, A. Tenorio, and C. Freire. 2002. Molecular Identification of Enteroviruses by Analyzing a Partial VP1 Genomic Region with Different Methods. Journal of Clinical Microbiology 40: 182-192. 72. Pallansch, M. A., and R. P. Roos. 2001. Enteroviruses: Polioviruses, Coxsackieviruses, Echoviruses, and Newer Enteroviruses, p. 723-775. In D. M. Knipe and P. M. Howley (ed.), Fields Virology, 4th ed, vol. 1. Lippincott, Williams & Wilkens, Philadelphia. 73. Peng, T., Y. Li, Y. Yang, C. Niu, P. Morgan-Capner, L. C. Archard, and H. Zhang. 2000. Characterization of Enterovirus Isolates from Patients with Heart Muscle Disease in a Selenium-Deficient Area of China. Journal of Clinical Microbiology 38: 3538-3543. 74. Piirainen, L., T. Hovi, and M. Roivainen. 1998. Variability in the Integrity of Human Enteroviruses Exposed to Various Simulated In Vivo Environments. Microbial Pathogenesis 25: 131-137. 75. Pyry, T., T. Hyypi, C. Horsnell, L. Kinnunen, T. hovi, and G. Stanway. 1994. Molecular Analysis of Coxsackievirus A16 Reveals a New Genetic Group of Enteroviruses. Virology 202: 982-987. 76. Pyry, T., L. Kinnunen, T. Hyypi, B. Brown, C. Horsnell, T. Hovi, and G. Stanway. 1996. Genetic and Phylogenetic Clustering of Enteroviruses. Journal of General Virology 77: 1699-1717. 77. Pulli, T., P. Koskimies, and T. Hyypi. 1995. Molecular Comparison of Coxsackie A Virus Serotypes. Virology 212: 30-38. 78. Pulli, T., M. Roivainen, T. Hovi, and T. Hyypi. 1998. Induction of neutralizing Antibodies by Synthetic Peptides Representing the C Terminus of Coxsackievirus A9 Capsid Protein VP1. Journal of General Virology 79: 22492253. 79. Racaniello, V. R. 1995. Early Events in Infection: Receptor Binding and Cell Entry, p. 73-94. In H. A. Rotbart (ed.), Human Enterovirus Infections. ASM Press, Washinton, DC. 80. Racaniello, V. R. 2001. Picornaviridae: The Viruses and Their Replication, p. 685-722. In D. M. Knipe and P. M. Howley (ed.), Fields Virology, 4th ed, vol. 1. Lippincott, Williams & Wilkens, Philadelphia.


65 81. Rohll, J. B., D. H. moon, D. J. Evans, and J. W. Almond. 1995. The 3' Untranslated Region of Picornavirus RNA: Features Required for Efficient Genome Replication. Journal of Virology 69: 7835-7844. 82. Rossman, M. G. 1994. Viral Cell Recognition and Entry. Protein Science 3: 1712-1725. 83. Rossman, M. G., J. Bella, P. R. Kolatkar, Y. He, E. Wimmer, R. J. Kuhn, and T. S. Baker. 2000. Cell Recognition and Entry by Rhinoand Enteroviruses. Virology 269: 239-247. 84. Rossman, M. G., Y. He, and R. J. Kuhn. 2002. Picornavirus-Receptor Interactions. Trends in Microbiology 10: 324-331. 85. Rotbart, H. A. 1995. Meningitis and Encephalitis, p. 271-289. In H. A. Rotbart (ed.), Human Enterovirus Infections. ASM Press, Washington, DC. 86. Rotbart, H. A. 2002. Treatment of Picornavirus Infections. Antiviral Research 53: 83-98. 87. Rotbart, H. A., J. F. O'Connell, and M. A. McKinlay. 1998. Treatment of Human Enterovirus Infections. Antiviral Research 38: 1-14. 88. Rotbart, H. A., and J. R. Romero. 1995. Laboratory Diagnosis of Enteroviral Infections, p. 401-418. In H. A. Rotbart (ed.), Human Enterovirus Infections. ASM Press, Washington, DC. 89. Santti, J., R. Vainonpaa, and T. Hyypi. 1999. Molecular Detection and Typing of Human Picornaviruses. Virus Research 62: 177-183. 90. Schlegel, A., and K. Kirkegaard. 1995. Cell Biology of Enterovirus Infection. In H. A. Rotbart (ed.), Human Enterovirus Infections. ASM Press, Washington, DC. 91. Shih, S. R., Y. W. Wang, G. W. Chen, L. Y. Chang, T. Y. Lin, M. C. Tseng, C. Chiang, K. C. Tsao, C. G. Huang, M. R. Shio, J. H. Tai, S. H. Wang, R. L. Kuo, and W. T. Liu. 2003. Serotype-specific Detection of Enterovirus 71 in Clinical Specimens by DNA Microchip A rray. Journal of Virological Methods 111: 55-60. 92. Singh, S., V. T. K. Chow, K. P. Chan, A. E. Ling, and C. L. Poh. 2000. RTPCR, Nucleotide, Amino Acid, and Phylogenetic Analyses of Enterovirus 71 Strains from Asia. Journal of Virological Methods 88: 193-204. 93. Smith, T. J., and A. G. Mosser. 1997. Antibody-Mediated Neutralization of Picornaviruses, p. 134-156. In W. Chiu, R. M. Burnett, and R. L. Garcea (ed.), Structural Biology of Viruses. Oxfo rd University Press, New York.


66 94. Smyth, M., T. Pettitt, A. Symonds, and J. Martin. 2003. Identification of the Pocket Factors in a Picornavirus. Archives of Virology 148: 1225-1233. 95. Stanway, G. 1990. Structure, Function and Evolution of Picornaviruses. Journal of General Virology 71: 2483-2501. 96. Stanway, G., and T. Hyypi. 1999. Parechoviruses. Journal of Virology 73: 5249-5254. 97. Stewart, P. L., and G. R. Nemerow. 1997. Recent Structural Solutions for Antibody Neutralization of Viruses. Trends in Microbiology 5: 229-233. 98. Strikas, R. A., L. J. Anderson, and R. A. Parker. 1986. Temporal and Geographic Patterns of Isolates of Nonpolio Enteroviruses in the United States, 1970-1983. Journal of Infectious Diseases 153: 346-351. 99. Thoelen, I., P. Lemey, I. Van Der Donck, K. Beuselinck, A. M. Lindberg, and M. Van Ranst. 2003. Molecular Typing and Epidemiology of Enteroviruses Identified from an Outbreak of Aseptic Meningitis in Belgium during the Summer of 2000. Journal of Medical Virology 70: 420-429. 100. Tracy, S., N. M. Chapman, R. J. Rubocki, and M. A. Beck. 1995. Host Immune Responses to Enterovirus Infections, p. 175-194. In H. A. Rotbart (ed.), Human Enterovirus Infections. ASM Press, Washington, DC. 101. Van Der Avoort, H., B. P. Hull, T. Hovi, M. A. Pallansch, O. M. Kew, R. Crainic, D. J. wood, M. N. Mulders, and A. M. V. Loon. 1995. Comparative Study of Five Methods for Intratypic Diffe rentiation of Polioviruses. Journal of Clinical Microbiology 33: 2563-2566. 102. Van Loon, A. M., G. C. Cleator, and A. Ras. 1999. External Quality Assessment of Enterovirus Detection and Typing. Bulletin of the World Health Organization 77: 217-222. 103. Verstrepen, W. A., S. Kuhn, M. M. Kockx, and M. E. V. d. Vyvere. 2001. Rapid Detection of Enteroviruses RNA in Cerebrospinal Fluid Specimens with a Novel Single-Tube Real-Time Reverse Transcription-PCR Assay. Journal of Clinical Microbiology 39: 4093-4096. 104. Wen, Y. Y., T. Y. Chang, S. T. Chen, C. Li, and H. S. Liu. 2003. Comparative Study of Enterovirus 71 Infection of Human Cell Lines. Journal of Medical Virology 70: 109-118.


67 105. Yang, D., P. Cheung, Y. Sun, J. yuan, H. Zhang, C. M. Carthy, D. R. Anderson, L. Bohunek, J. E. Wilson, and B. M. McManus. 2003. A ShineDelgarno-like Sequence mediates in Vitro Ribosomal Internal Entry and Subsequent Scanning for Translation Initiation of Coxsackievirus B3 RNA. Virology 305: 31-43. 106. Zell, R., and A. Stelzner. 1997. Application of Genome Sequence Information to the Classification of Bovine Enteroviruses: The Importance of the 5'and 3'Nontranslated Region.


68 Appendices


69 Appendix I Master Mix Components Oberste A One-Step RT-PCR Component (Final Conc.) Volume Stock Conc. RNase-free H2O 12.275 l 5X Buffer 5 l dNTPs 1 l 10mM total (2.5mM each) 30pmol primer 292 0.3 l 100 M (100pmol/ l) 30pmol primer 222 0.3 l 100 M (100pmol/ l) 10U RNasin (RNase inhibitor) 0.125 l 40units/ l Enzyme Mix 1 l Proprietary Template 5 l Total 25 l Oberste B One-Step RT-PCR Component (Final Conc.) Volume Stock Conc. RNase-free H2O 11.975 l 5X Buffer 5 l dNTPs 1 l 10mM total (2.5mM each) 30pmol primer 040 0.3 l 100 M (100pmol/ l) 30pmol primer 011 0.3 l 100 M (100pmol/ l) 30pmol primer 012 0.3 l 100 M (100pmol/ l) 10U RNasin (RNase inhibitor) 0.125 l 40units/ l Enzyme Mix 1 l Proprietary Template 5 l Total 25 l Casas 1st One-Step RT-PCR Component (Final Conc.) Volume Stock Conc. RNase-free H2O 12.275 l 5X Buffer 5 l dNTPs 1 l 10mM total (2.5mM each) 30pmol primer VP1-1A 0.3 l 100 M (100pmol/ l) 30pmol primer VP1-1S 0.3 l 100 M (100pmol/ l) 10U RNasin (RNase inhibitor) 0.125 l 40units/ l Enzyme Mix 1 l Proprietary Template 5 l Total 25 l


70 Casas 2nd One-Step RT-PCR Component (Final Conc.) Volume Stock Conc. RNase-free H2O 12.275 l 5X Buffer 5 l dNTPs 1 l 10mM 30pmol primer VP1-2A 0.3 l 100 M (100pmol/ l) 30pmol primer VP1-2S 0.3 l 100 M (100pmol/ l) 10U RNasin (RNase inhibitor) 0.125 l 40units/ l Enzyme Mix 1 l Proprietary Template 5 l Total 25 l Casas Nested One-Step RT-PCR First RT-PCR identical to Casas 1st; Second RT-PCR identical to Casas 2nd Product (5 l) of Casas 1st used as template for nested RT-PCR. Dilutions performed ranging from 1:10 1:10,000, depending on intensity of amplicon obtained from first-round amplification. Caro Two-Step RT-PCR: RT Component Volume Stock Conc. 10X Buffer 2 l MgCl2 4 l 25mM dNTPs 8 l 10mM RNasin (RNase inhibitor) 0.25 l 40units/ l 50pmol EUC2a primer 0.5 l 100 M (100pmol/ l) 50pmol EUC2b primer 0.5 l 100 M (100pmol/ l) AMV RT enzyme 0.32 l 24units/ l Template 5 l Total 20.57 l Caro Two-Step RT-PCR: PCR Component Volume Stock Conc. 10X Buffer 8 l MgCl2 4 l 25mM RNase-free H2O 64.93 l Taq polymerase 0.5 l 5units/ l 50pmol EUC2 primer 0.5 l 100 M (100pmol/ l) 50pmol EUG3b primer 0.5 l 100 M (100pmol/ l) 50pmol EUG3b primer 0.5 l 100 M (100pmol/ l) 50pmol EUG3c primer 0.5 l 100 M (100pmol/ l) Template (RT reaction) 20.57 l Total 100 l Sequencing Reaction Component Volume Stock Conc. DTCS Maste Mix 3 l Proprietary 10pmol primer 1 l 10 M (10pmol/ l) RNase-free H2O (16-x) l Template (variable) x l Total 20 l


71 Thermal Cycler Profiles Oberste A and B 1 Cycle 40 Cycles 1 Cycle 38 C 10 minutes 95 C 30 seconds 70 C 5 minutes 50 C 40 minutes 42 C 45 seconds 4 C 95 C 3 minutes 65 C 30 seconds Casas 1st and 2nd 1 Cycle 40 Cycles 1 Cycle 38 C 10 minutes 95 C 30 seconds 70 C 5 minutes 50 C 40 minutes 46 C 1 minute 4 C 95 C 3 minutes 65 C 30 seconds Caro RT 1 Cycle 42 C 30 minutes 95 C 5 minutes 4 C Caro PCR 40 Cycles 1 Cycle 95 C 20 seconds 72 C 10 minutes 45 C 1 minute 4 C 72 C 1 minute Sequencing 35 Cycles 96 C 20 seconds 42 C 20 seconds 60 C 3 minutes 4 C


72 Appendix II Two by Two Tables Oberste A Screening Standard positive negative 222/292 positive 46 0 1-Step negative 8 2 sensitivity 0.8518519 specificity 1 PPV 1 N PV 0.2 Oberste B Screening Standard positive negative 011/012/040 positive 36 0 1-Step negative 18 2 sensitivity 0.6666667 specificity 1 PPV 1 N PV 0.1 Casas 1st Screening Standard positive negative 1A/1S positive 30 0 1-Step negative 24 2 sensitivity 0.5555556 specificity 1 PPV 1 N PV 0.0769231


73 Casas 2nd Screening Standard positive negative 2A/2S (unnested) positive 16 0 1-Step negative 38 2 sensitivity 0.2962963 specificity 1 PPV 1 N PV 0.05 Casas Nested Screening Standard positive negative 1A/1S-2A/2S positive 20 0 1-Step x 2 negative 34 2 N ested sensitivity 0.3703704 specificity 1 PPV 1 N PV 0.0555556 Caro Screening Standard positive negative EUC/EUG positive 40 0 2-Step negative 14 2 sensitivity 0.7407407 specificity 1 PPV 1 N PV 0.125










78 Appendix IV Possible Algorithm for Molecular-Based Enterovirus Serotyping Real-Time Screening Assay in conjunction with Arbovirus Taqman RT-PCR BGM/RMK Cell Culture Harvest at 3/3+ CPE RNeasy Extraction, One-Step RT-PCR Band excision, QIAquik Clean-up, quantification gel Sequencing Contig assembly, BLAST search