1* _. ,‘ z - nigi’m‘é abé. a. ‘f I '7 ‘HA 3:5; {NV-u.) ., ~ A . . 3. ”'53? :., 4: . r: {M ~' 45“.. ' “fl: ? < "d .0 : .f J-A . 4;”: 3.3”. -- “.4:- ‘ 259-}; - A W}: Va“ . : law.“ » n ' ‘ .. This is to certify that the thesis entitled REAL-TIME RT-PCR OF FELINE CALICIVIRUS AND OPTIMIZATION FOR DETECTION OF VIRUS IN FELINE URINE presented by BRIAN ALAN SCANSEN has been accepted towards fulfillment of the requirements for the MS. degree in Small Animal Clinical Sciences fi/g-r-J ajor Professor’s Siéfiature W 3'!” @3__ Date MSU is an Affinnative Action/Equal Opportunity Institution fivi- LIBRARY Michigan State University PLACE IN RETURN Box to remove this checkout from your record. TO AVOID FINES return on or before date due. MAY BE RECALLED with earlier due date if requested. DAT E DUE DATE DUE DATE DUE APR {3 2005 L/ ,/~ A ;' 6/01 cJCIRC/OateDue.p65-p.15 REAL-TIME RT-PCR OF FELINE CALICIVIRUS AND OPTIMIZATION FOR DETECTION OF VIRUS IN FELINE URINE By Brian Alan Scansen A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Small Animal Clinical Sciences 2004 ABSTRACT REAL-TIME RT-PCR OF FELINE CALICIVIRUS AND OPTIMIZATION FOR DETECTION OF VIRUS IN FELINE URINE By Brian Alan Scansen Investigating the role of feline calicivirus (FCV) in idiopathic cystitis may be facilitated by a reverse-transcriptase polymerase chain reaction (RT-PCR) assay optimized for detection of FCV urinary tract infections. Two FCV RT-PCR assays were developed; a p5.6 gene-based qualitative assay and a p30 gene-based quantitative real-time SYBR® Green | assay. The p5.6 gene assay was highly sensitive and specific, but was not efficiently adapted to real-time RT-PCR. The real-time p30 gene assay was sensitive, specific, and linear over a wide range of template concentrations, and had a reaction efficiency of 95%. The p30 gene assay detected all 51 North American FCV field isolates tested. To optimize detection of FCV in urine by RT-PCR, viral RNA was prepared from urine by dilution and thermal inactivation, polyethylene glycol precipitation, isolation with oligo(dT)2s-coated magnetic beads, or extraction with two silica gel-based columns. The FCV real-time p30 gene assay performed significantly better when using RNA isolated from feline urine with either of the silica gel-based columns. Copyright by Brian Alan Scansen 2004 This work is dedicated to my wife, Kimbedy, for her constant support and unwavering praise and for showing me how to succeed in medicine, And, To my parents for continually reaffirming the value of a strong education and for instilling in me the desire to achieve all of my goals. iv ACKNOWLEDGEMENTS This work would not have been possible without the guidance of my major professor, Dr. John Kruger. He deserves much of the credit for making the combined DVM/MS program a reality and for motivating me throughout the research. For designing courses, improving my methodology, editing manuscripts, and dealing with administrative issues, this work is as much a reflection of his ability as a mentor as it is of mine as a student. Dr. Roger Maes also deserves acknowledgement for his guidance in virological methodology and continued support of my research from inception to finish. His knowledge of diagnostic virology was seemingly limitless and was crucial to the success of the project. Similarly, Dr. Pat Venta’s understanding of genetics was invaluable in dealing with the intricacies of molecular diagnostics and in troubleshooting problems as they arose. Dr. John Baker also deserves recognition for his guidance, suggestions, and input throughout the program. A significant amount Of assistance was also provided by Dr. Annabel Wise and I am indebted to her for her expertise and assistance in diagnostic virology. This work was generously funded by the Michigan State University Companion Animal Fund and the Center for Feline Health and Well-Being as well as by the Geraldine R. Dodge Foundation in concert with the Kenneth A. Scott Charitable Trust. TABLE OF CONTENTS LIST OF TABLES ............................................................................................... viii LIST OF FIGURES ............................................................................................... xi KEY TO ABBREVIATIONS ................................................................................ xiv INRODUCTION ..................................................................................................... 1 CHAPTER 1 LITERATURE REVIEW ......................................................................................... 5 Overview .................................................................................................... 5 Historical Perspectives ............................................................................... 6 Vesicular Exanthema of Swine Virus ............................................... 6 Feline Calicivirus .............................................................................. 6 San Miguel Sea Lion Virus .............................................................. 8 Norwalk Virus .................................................................................. 8 Sapporo Virus .................................................................................. 9 Rabbit Hemorrhagic Disease Virus ................................................ 10 Other Caliciviruses ......................................................................... 11 Biology of FCV ......................................................................................... 11 Physical Properties ........................................................................ 1 1 Molecular Biology .......................................................................... 12 Replication ..................................................................................... 15 Antigenic Relationships ................................................................. 15 Transmission ................................................................................. 16 Pathogenesis ................................................................................. 17 FCV-Induced Clinical Diseases ................................................................ 19 Upper Respiratory Tract Disease ................................................... 19 Arthropathy .................................................................................... 21 Abortion ......................................................................................... 21 Chronic Gingivitis ........................................................................... 22 Virulent Systemic Disease ............................................................. 23 Lower Urinary Tract Disease ......................................................... 24 Diagnosis of FCV Infection ....................................................................... 27 Virus Isolation ................................................................................ 27 Serology ........................................................................................ 28 Electron Microscopy, Fluorescent Antibody, and lmmunohistochemistry ............................................................ 29 Molecular Techniques .................................................................... 3O Diagnosis of FCV in Urine ........................................................................ 40 vi CHAPTER 2 p5.6 GENE-BASED RT-PCR ASSAY ................................................................. 43 Abstract .................................................................................................... 44 Introduction ............................................................................................... 46 Materials and Methods ............................................................................. 49 Viruses ........................................................................................... 49 Conventional Qualitative RT-PCR Assay ....................................... 50 Real-Time RT-PCR Assay with a Dual-Labeled DNA Hybridization Probe ............................................................... 51 Real-Time RT-PCR Assay with SYBR® Green | ............................ 52 Results ..................................................................................................... 52 Conventional Qualitative RT-PCR Assay ....................................... 52 Real-Time RT-PCR Assay with a Dual-Labeled DNA Hybridization Probe ............................................................... 53 Real-Time RT-PCR Assay with SYBR® Green I ............................ 53 Discussion ................................................................................................ 55 CHAPTER 3 p30 GENE-BASED RT-PCR ASSAY .................................................................. 61 CHAPTER 4 COMPARISON OF RNA PREPARATION METHODS ........................................ 73 Abstract .................................................................................................... 75 Introduction ............................................................................................... 76 Materials and Methods ............................................................................. 78 Collection and Preparation of Samples .......................................... 78 RNA Isolation Methods .................................................................. 80 FCV RT-PCR Assay ...................................................................... 85 Data Analyses ............................................................................... 86 Results ..................................................................................................... 87 Urine Specimens ........................................................................... 87 Effect of RNA Isolation Method on RT-PCR Performance ............. 87 Effect of Urine Variables on RT-PCR Performance ....................... 89 Discussion ................................................................................................ 92 APPENDIX A EFFECTS OF STORAGE TEMPERATURE ON DETECTION OF FCV IN URINE AND TISSUE CULTURE MEDIUM BY VIRUS ISOLATION AND RT-PCR ...................................................................... 99 APPENDIX B RESULTS OF CDNA SEQUENCING OF THE P30 GENE OF8 FCV ISOLATES ...................................................................... 102 LIST OF REFERENCES ................................................................................... 107 vii Table 1. Table 2. Table 3. Table 4. LIST OF TABLES Complete FCV Sequences. FCV strains for which the complete genome sequence is available on GenBank with their accession number, year and country of isolation, and the corresponding reference. NR = Not Reported. Feline calicivirus RT-PCR assays. This table comprises all previously reported FCV RT-PCR assays, together with the two assays reported here. Listed are the primers given by the individual authors (labeled as in the corresponding reference), the primer sequence (5’-3’), the size of the amplicon, the region of the FCV genome amplified (see Figure 2), the type of RT-PCR assay (gel- based or real-time), the diagnostic range of the assay as reported in the corresponding reference (listed as number of positive isolates / number tested), the corresponding nucleotides of strain FCV-Urbana (GenBank Accession number = L40021) covered by the assay, and the reference in which the assay was reported. * This paper utilized a semi-nested RT-PCR with P1 and P2 in the first round of amplification and P2 and P4 in the second round of amplification. The size of this amplicon was then 393bp. NR = Not reported Characteristics of urine specimens Obtained from six 9- month-old female specific-pathogen-free cats used in the study. A small quantity of RNase—free water was added to each urine specimen from each cat to obtain the final volume of 15.75ml per cat required for analyses. Mean lower detection limits of the FCV RT-PCR assay using RNA prepared by four RNA isolation methods. Urine and blood specimens were obtained from six 9- month-old female specific-pathogen-free cats. Unaltered urine, urine with added whole or hemolyzed blood, centrifuged urine supernatant, and tissue culture medium were spiked with FCV and serially diluted. Viral RNA was isolated from samples with each of 4 preparation methods and amplified with the FCV p30 gene-based RT-PCR assay. viii 34 79 88 Table 5. Table 6. Table 7. Table 8. Table 9. Table 10. Geometric mean detection threshold cycle (CI) of the FCV RT-PCR assay using RNA prepared by four RNA isolation methods. Urine and blood specimens were obtained from six 9-month-Old female specific-pathogen- free cats. Unaltered urine, urine with added whole or hemolyzed blood, centrifuged urine supernatant, and tissue culture medium were spiked with FCV and serially diluted. Viral RNA was prepared from samples with each isolation method and amplified in duplicate with the FCV RT-PCR assay. Number of positive results for virus isolation and RT-PCR for detection of FCV in urine or tissue culture medium stored at 4C or —7OC. Final lower detection limits (expressed as TCleo/sample) for detection of FCV in urine by virus isolation or RT-PCR after storage at 4C and —70C. Descriptions of FCV strains used for genotypic comparisons of cDNA and amino acid sequences of the FCV p30 gene. Results of bidirectional automated sequencing of a 90 base pair portion of the p30 (BA-Like) gene of ORF1 of 8 FCV isolates Obtained from the Michigan State University Diagnostic Center for Population and Animal Health. The cDNA sequences correspond to nucleotides 2433 to 2522 and to amino acids 806 to 834 of the Urbana reference strain. Percent nucleotide identity between the FCV p30 gene CDNA sequences of 8 FCV isolates Obtained from the Michigan State University Diagnostic Center for Population and Animal Health and 6 selected FCV reference strains (see Table 8 for isolate descriptions). Comparisons were made using a 90 base pair region of the FCV p30 gene corresponding to nucleotides 2433- 2522 Of the Urbana reference strain. ix 91 101 101 102 103 104 Table 11. Percent amino acid identity between the FCV p30 protein sequences of 8 FCV isolates obtained from the Michigan State University Diagnostic Center for Population and Animal Health and 6 selected FCV reference strains (see Table 8 for isolate descriptions). Comparisons were made using a 29 amino acid region of the FCV p30 protein corresponding to amino acids 806-834 of the Urbana reference strain. 105 Figure 1. Figure 2. Figure 3. LIST OF FIGURES Schematic diagram of the genomic organization of FCV. (A) The three open reading frames (ORF) of FCV (numbers correspond to nucleotides of the FCV-Urbana strain) with protein VPg at 5’ end and poly(A) tail at 3’; (B) the nonstructural proteins of ORF1; and (C) the organization of the capsid gene (ORF2) into 6 regions with C and E being hypervariable (adapted from Green et al. 2002, Sosnovtsev & Green 2002). Genomic location of reported FCV RT-PCR assay targets. The target location of the reported FCV RT-PCR assays are shown in relation to the complete genome and its three open reading frames (ORF 1, ORF2, ORF3). Products shown are: (1) Seal 1994, (2) Radford et al. 1997, (3) Geissler et al. 1997, (4) Baulch-Brown et al. 1999, (5) Sykes et al. 1998, (6) Horimoto et al. 2001, (7) Doyle et al. 1996, (8) Scansen et al. 2002, (9) Helps et al. 2002, (10) Scansen et al. 2004. Agarose gel (1.5% w/v) electrophoresis of amplification products showing the effect of variable primer concentration on performance of a FCV p5.6 gene-based RT-PCR assay. Varying the concentration of the primers indicated optimal assay performance when the fonNard (conserved) primer concentration was decreased in comparison to the reverse (degenerate) primer concentration: (A) Lanes 1-4 contain equal concentration (0.6pM) of the forward and reverse primers; (B) Lanes 1- 4 contain 0.6pM of the fonNard primer and 0.6].1M (lane 1), 0.3pM (lane 2), 0.15pM (lane 3), and 0.06pM (lane 4) of the reverse primer; (C) Lanes 1-4 contain 0.6pM of the reverse primer and 0.6pM (lane 1), 0.3pM (lane 2), 0.15pM (lane 3), and 0.06pM (lane 4) of the forward primer, and lane 5 contains 0.6uM of the forward primer and 1.2pM of the reverse primer. The amplicon size of the assay was 118bp. Lane M = Molecular size marker (123bp DNA ladder). xi 37 Figure 4. Figure 5. Figure 6. Figure 7. Polyacrylamide gel (10% w/v) electrophoresis of amplification products from serial dilutions of FCV (titer = 9.2 x 108 TCIDso/ml) reverse-transcribed into cDNA and amplified by a p5.6 gene-based RT-PCR assay. Dilutions shown here include the 104 (lane 4), 10‘5 (lane 5), 10'6 lane 6), 10‘7 (lane 7), 10'8 (lane 8), 10'9 (lane 9), and 10'1 (lane 10). The highest detectable dilution was 109, corresponding to a TCIO50 of 9.2 x 10'1 per ml. The amplicon size was 118bp. Lane M = Molecular size marker (123bp DNA ladder), lane + = positive control, lane - = negative control. Lower detection limit of the p5.6 real-time RT-PCR assay for FCV with the dual-labeled DNA hybridization probe. A graph of fluorescence vs. cycle number illustrating amplification of serial FCV dilutions with the threshold cycle varying inversely with virus concentration. Serial log dilutions of virus were made starting from a stock with a titer of 9.2 x 108 TCIDso/ml (0), progressing through 10‘2 (I), 10'3 (A), 104 (e), 10'5 (O), 10 (D) dilutions, and a negative control (A). All samples were amplified in duplicate. Optimization Of SYBR® Green I dye concentration. A graph of fluorescence vs. cycle number indicating optimal amplification at dilute dye concentrations and inhibitory effects of higher concentrations. SYBR® Green I dye was added to the p5.6 gene-based assay at variable concentrations: 1240,000 (O), 1:35,000 (I), 1:30.000 (A), 125,000 (0), 1:20.000 (0), and 1215,000 (El). Lower detection limit of the real-time RT-PCR assay for FCV. Serial log dilutions of virus were made starting from a stock with a titer of 5.2 x 108 TCleo/ml (0), pro ressing through 10'1 (I), 10’2 (A), 10'3 (O), 10“ (O), 10' (El), 10'6 (A), 10'7 (O), and 10'8 (X) dilutions. All samples were amplified in triplicate (A) A graph Of fluorescence vs. cycle number illustrating amplification of serial FCV dilutions, with the threshold cycle varying inversely with virus concentration. (B) The melting temperature curve illustrating the difference in melting temperature between specific FCV amplification products (~82C) and non-specific products (<76C). Specific FCV products were observed with amplification of the 10’6 dilution. xii 55 56 57 67 Figure 8. Figure 9. Figure 10. Linearity of the FCV p30 gene-based real-time RT-PCR. Serial 10-fold dilutions of FCV RNA were amplified in triplicate. The CT was calculated for each reaction, plotted against the dilution factor (expressed as the log1o of the dilution), and evaluated by linear regression analysis. A schematic representation of study design. Urine and blood specimens were obtained from six 9-month-old female specific-pathogen-free cats. Unaltered urine, urine with added whole or hemolyzed blood, centrifuged urine supernatant, and tissue culture medium were spiked with FCV and serially diluted. Viral RNA was prepared from samples with each isolation method and amplified in duplicate with the FCV RT-PCR assay. Effect of individual cat urine specific gravity on threshold cycle (CI). The association between urine specific gravity and C, was evaluated by linear regression analysis. xiii 69 82 90 ANOVA BHV4 CDNA CFl CPE Ct DNA DT dTMB FCV FCV-R FHV FlV IC FAT lHC KCDV HPF MAP MSU-DCPAH KEY TO ABBREVIATIONS Analysis of Variance Bovine Herpesvirus-4 Complementary Deoxyribonucleic Acid California Feline Isolate Cytopathic Effect Threshold Cycle Deoxyribonucleic Acid Dilution and Thermal Inactivation Oligo(dT) Magnetic Beads Feline Calicivirus Feline Calicivirus, Respiratory Strain Feline Herpesvirus Feline Immunodeficiency Virus Idiopathic Cystitis Fluorescent Antibody Test lmmunohistochemistry Kidney Cell Degenerating Virus High Power Field Magnesium Ammonium Phosphate Michigan State University’s Diagnostic Center for Population and Animal Health xiv NV ORF PCR PEG QA RHDV RN RNA RT SMSV SPF SV TCIDso TCM TNTC USG URTD VESV VI NonIvalk Virus Open Reading Frame Polymerase Chain Reaction Polyethylene Glycol Silica Gel-Based Extraction Column (Qiagen QlAamp Viral RNA Mini Kit®) Rabbit Hemorrhagic Disease Virus Silica Gel-Based Extraction Column (Qiagen RNeasy Mini Kit®) Ribonucleic Acid Reverse-Transcriptase San Miguel Sea Lion Virus Specific Pathogen Free Sapporo Virus Tissue Culture lnfective Dose Tissue Culture Medium Too Numerous To Count Urine Specific Gravity Upper Respiratory Tract Disease Vesicular Exanthema of Swine Virus Virus Isolation XV INTRODUCTION The family Calicivin'dae comprises a group of small non-enveloped viruses known to cause a myriad Of clinical syndromes in humans, cats, dogs, rabbits, pigs, cattle, and marine mammals. Several genera Of calicivirus have been identified and these have been associated with a wide variety of disorders including vesicular lesions, stomatitis, rhinitis, pneumonia, arthropathies, abortion, gastroenteritis, and hemorrhagic systemic disease (Murphy et al. 1999). One of the earliest identified members of Calicivin'dae was feline calicivirus (FCV), an agent that induces upper respiratory tract disease primarily afflicting young cats (Fastier 1957, Crandell & Madin 1960). Although best characterized as a respiratory pathogen, FCV infection has also been associated with lameness (Pedersen et al. 1983), oral ulcerations (Tenorio et al. 1991), abortion (Ellis 1981), and a severe and highly fatal hemorrhagic-fever—like syndrome (Pedersen et al. 2000). Studies have also implicated FCV as a potential etiologic agent in feline idiopathic cystitis (IC), a common urinary disorder of cats resulting in hematuria, pollakiuria, strangury, and urethral obstruction (Rich & Fabricant 1969, Kruger et al. 1996, Rice et al. 2002). The role of FCV in the pathogenesis of feline IC is uncertain. The inability to establish a causative link between FCV and the development of IC thus far may have been the result of insensitive or inappropriate methods of virus detection (Kruger et al. 1996). Detection of the virus in urine by standard methods, such as virus isolation, requires viable virus and the absence of substances that are inhibitory to viral replication or that are toxic to cell culture (Sykes et al. 2001, Rice et al. 2002). FCV is extremely sensitive to the effects of urine, storage temperature, and freeze-thaw cycles and feline urine may be toxic to standard cell lines used for feline virus isolation (Appendix A, Komolate et al. 1976, Sykes et al. 1998, Rice et al. 2002). Diagnosis of FCV has historically relied upon detection of viable virus by virus isolation in cell-culture or detection of FCV neutralizing antibodies in serum (Bittle et al. 1960, Gillespie & Scott 1973). Virus isolation is time consuming, labor intensive and requires viable virus in a substrate that is not toxic to standard cell culture lines (Rice et al. 2002). Serologic detection of FCV antibodies only indicates exposure and can be confounded by vaccination status or prior viral exposure (Kruger et al. 1996). Other reported methods of FCV diagnosis include direct identification of FCV antigens by fluorescent antibody testing (Gillespie et al. 1971) and immunohistochemistry (Dick et al. 1989), or virions by electron microscopy (Studdert 1978). These methods are similarly laborious and may yield false negative and/or false positive results. Molecular diagnostic techniques have become increasingly popular in the diagnosis of viral infections. In particular, the polymerase chain reaction (PCR), or its variant, the reverse-transcriptase PCR (RT-PCR), have been increasingly used in diagnostic virology for direct detection of viral DNA and RNA respectively in a broad range of clinical samples. Compared to conventional methods, PCR- based diagnostic assays have the advantages of enhanced specificity and sensitivity, increased scalability, more rapid throughput, and lower per-sample cost (Murphy et al. 1999). In addition, direct sequencing of PCR products allows for rapid detection of genomic variation between viral strains. Several RT-PCR assays have been reported for FCV, of which the majority amplify regions of the capsid protein gene (Seal 1994, Radford et al. 2000). Fewer FCV RT-PCR assays have been developed specifically for diagnostic purposes, and fewer still report sensitivity, specificity, and diagnostic range. Nearly all previously reported FCV RT-PCR assays rely on gel electrophoresis for verification of nucleic acid amplification. A relatively new development in molecular detection Of viral nucleic acids is real- time RT-PCR (Mackay et al. 2002). Real-time RT-PCR Offers the advantage over traditional RT-PCR of increased sensitivity, faster throughput, increased reproducibility, decreased risk of laboratory contamination, and better quantification (Mackay et al. 2002). Real-time RT-PCR also eliminates the need for post-PCR processing for detection of amplification products by ethidium bromide staining and agarose gel electrophoresis. A p30 gene-based real-time RT-PCR assay for FCV has recently been reported that was comparable in sensitivity to virus isolation and that had a broad diagnostic range when used to detect FCV isolates from the United Kingdom (Helps et al. 2002). Molecular diagnostics, such as RT-PCR, may circumvent many of the difficulties associated with detecting FCV urinary tract infections in cats. While molecular diagnosis of FCV infection has been reported (Seal 1994, Radford et al. 2000), the substrate on which testing is performed can play an important role in assay performance. Urine, in particular, is known to inhibit PCR amplification of viral and bacterial agents (Khan et al. 1991, Behzadbehbahani et al. 1997, Echavarria et al. 1998, Biel et al. 2000). Consequently, preparation of nucleic acids becomes a critical step that not only serves to concentrate and purify nucleic acids, but also to remove or inactivate PCR inhibitors. Studies comparing RNA isolation methods for their ability to remove or inactivate RT-PCR inhibitors and preserve FCV RNA integrity in feline urine specimens have not been reported. The overall goal Of the following studies was to develop a more sensitive and rapid means of detecting FCV urinary tract infections in cats. We hypothesized that a real-time RT-PCR assay for FCV could circumvent many problems associated with conventional urine virus isolation and RT-PCR methods, and facilitate clinical and laboratory investigations into the causative role of FCV in feline IC. The specific aims of the following studies were to develop a real-time quantitative RT-PCR assay and optimize viral RNA preparation methods for detection Of FCV nucleic acids in feline urine. LITERATURE REVIEW Overview The family Caliciviridae consists of numerous viruses of veterinary and human medical importance. The caliciviruses are organized into four genera, which are Vesivirus, Lagovims, Norovirus (previously known as small round structured viruses or Norwalk-like viruses), and Sapovims (previously known as classical caliciviruses or Sapporo-like viruses) (Biichen-Osmond 2003). All caliciviruses share a similar structure with a single linear positive-sense RNA genome 7.4- 8.3kb in length. The genome has a covalently attached protein (VPg) at the 5’ end and a polyadenylated tail at the 3’ end. The nucleic acid is embedded within a 35-40nm non-enveloped icosahedral capsid. The name calicivirus is derived from the Latin calyx, for cup, and reflects the 32 cup-shaped indentations visible on the capsid surface (Green et al. 2000). Numerous species of calicivirus have been identified and these are associated with a wide variety of disorders including vesicular lesions, stomatitis, rhinitis, pneumonia, arthropathies, abortion, gastroenteritis, and hemorrhagic-like systemic disease (Murphy et al. 1999). The prototypical caliciviruses include vesicular exanthema Of swine virus (VESV, genus Vesivirus), San Miguel sea lion virus (SMSV, genus Vesivirus), Nonrvalk virus (NV, genus Norovirus), Sapporo virus (SV, genus Sapovirus), rabbit hemorrhagic disease virus (RHDV, genus Lagovirus), and feline calicivirus (FCV, genus Vesivirus). Historical Perspectives Vesicular Exanthema of Swine Virus The first reported outbreak of disease caused by a calicivirus occurred in 1932 at swine farms in southern California (Traum 1936). Clinical signs associated with the initial outbreak, and subsequent similar outbreaks over the next three years, were characterized by pyrexia and vesicle formation around the snout, mouth, and digits and were clinically indistinguishable from foot-and-mouth disease or vesicular stomatitis. However, the disease was found to be distinct from foot- and-mouth-disease and vesicular stomatitis, based on its pattern Of infectivity in test animals, and was therefore designated VESV (Traum 1936). VESV outbreaks occurred sporadically in California until 1952 when it was diagnosed for the first time in Nebraska and then throughout 42 states by September of 1953 (Fenner et al. 1993). Implementation Of a slaughter program and enforcement of garbage cooking laws resulted in eradication of the disease by 1956. It has not been detected anywhere in the world since that time. Although VESV was believed to be viral in origin, it was not until 1968 that the agent was confirmed to be a calicivirus (although still classified as a picomavirus at that time) and found to be structurally similar to FCV (Wawrzkiewicz 1968). Feline Calicivirus The discovery of feline calicivirus was an unintended event. While attempting to grow feline panleukopenia virus in cell culture, a new and unrelated agent was isolated which produced severe cytopathogenic changes in kitten kidney cells (Fastier 1957). The agent was described at that time as the kidney cell- degenerating virus (KCDV). KCDV growth in vitro was supported by a wide range of feline tissues. When inoculated intravenously into kittens, KCDV induced only mild diarrhea and anorexia approximately one week after exposure. However, seroconversion was noted on paired sera. The inability of the virus to produce any overt clinical disease prompted Fastier (1957) to label KCDV a virus “in search of a disease.” KCDV was later determined to be FCV by serum neutralization using antiserum from feline viral isolates from cats with upper respiratory infection (Bittle et al. 1960) and feline “picomavirus” isolates (Takahashi et al. 1971). The discovery of KCDV in an animal co-infected with feline panleukopenia, and the inability of KCDV to produce notable clinical disease suggested that this strain was relatively avirulent. The first report in the literature of a feline virus consistent with what we now recognize clinically as FCV-induced disease occurred in 1960 (Crandell & Madin 1960). Designated the California feline isolate (CF I), the virus was isolated from blood and oropharyngeal secretions obtained from a nine-week-old female kitten from the San Francisco Bay area that presented with fever, tachypnea, rales, anorexia, lethargy, and a stiff gait. The CFI isolate showed characteristic cytopathogenic changes in kitten kidney cell culture. Furthermore, upper respiratory clinical signs and pyrexia were reproducible in other kittens exposed to the virus, both intranasally and via contact with experimentally infected cats. Serologic cross-neutralization did not occur between CFI and feline panleukopenia virus, feline rhinotracheitis virus, or feline pneumonitis (now designated Chlamydophila felis). Throughout the next decade, similar agents were isolated from cats with upper respiratory tract disease (URTD) in other areas of the United States and from Europe, Australia, and Japan (Gillespie & Scott 1973, Takahashi et al. 1971 ). San Miguel Sea Lion Virus In 1972, a series of abortions was studied among sea lions on San Miguel Island, California. From one of the aborting animals a virus, indistinguishable from VESV, was isolated and termed SMSV (Smith et al. 1973). The similarity in structure Of SMSV to VESV coupled with the ability of SMSV to cause vesicular lesions when injected intradennally into the lips of swine led Smith et al. (1973) to theorize that the outbreaks of VESV in California in the 1930’s were the result of transmission of SMSV from its natural marine host to swine via contaminated garbage feeding. Since the discovery of SMSV, similar viral agents have been isolated from a variety of other marine mammals (Fenner et al. 1993). Norwalk Virus The first human disease attributed to a calicivirus occurred at a school in Norwalk, Ohio in October 1968. Fifty percent of the students and teachers developed gastroenteritis, primarily characterized by vomiting and nausea, for which no etiological agent was determined (Kapikian 2000). Four years later, immune electron microscopy detected 27nm viral particles in infectious stool filtrates derived from the Nonrvalk outbreak (Kapikian et al. 1972). The Norwalk virus (NV), although of appropriate size, did not have the typical appearance of the caliciviruses and came to be classified with a number of other small round structured viruses. However, detection of a single major structural protein, a feature of the Calicivin'dae, was compelling evidence that the NV was a calicivirus (Greenberg, et al. 1981). Finally, sequencing of the NV genome confirmed its placement in Calicivin'dae (Jiang et al. 1990). NV is considered the predominant cause of acute gastroenteritis in humans, accounting for an estimated 23 million cases annually, or roughly half of all foodbome outbreaks (CDC 2004). Research on NV has been hindered due to the inability to grow the virus in tissue culture. Current trends indicate that FCV may be a convenient model for NV due to the ease with which FCV can be cultivated in the laboratory, even though the two viruses are in different genera (Bidawid et al. 2003). Sapporo Virus In contrast to the small round structured viruses such as NV, the first human viral agents having the prototypical cup-shaped depressions of caliciviruses were isolated in 1976 from the fecal samples Of children in Glasgow, Scotland (Madeley & Cosgrove 1976). Further work on the “human caliciviruses”, as they came to be known, occurred as a result of an outbreak of gastroenteritis at an orphanage in Sapporo, Japan in 1977 (Chiba et al. 1979) and subsequent outbreaks between 1977 and 1982 (Nakata et al. 1985). Viral particles from these outbreaks were morphologically identical to known animal caliciviruses, such as FCV, and came to be called Sapporo Virus (SV) (Chiba et al. 2000). SV has now been shown genetically to be a member of the Caliciviridae, although distinct from NV, and is now the archetypal virus Of the genus Sapovirus (Numata et al. 1997). Although closer in morphologic appearance to FCV than to NV, SV has proven resistant to cultivation in the laboratory. Rabbit Hemorrhagic Disease Virus Rabbit hemorrhagic disease (RHDV) was detected as an emerging epizootic beginning in China in 1984 and extending to Western Europe and Africa by 1988 (Murphy 1999). The outbreaks of this disease were severe with morbidity approaching 100% and mortality rates of 90% in adult animals, with pathologic signs of systemic hemorrhage (Fenner et al. 1993). The causative agent of the disease was determined to be a calicivirus by electron microscopy and physical properties consistent with Caliciviridae (Parra & Prieto 1990, Ohlinger et al. 1990). A similar disease causing large-scale deaths of brown hares, European brown hare syndrome, was described at the same time as rabbit hemorrhagic disease and determined to be caused by a calicivirus as well (Chasey & Duff 1990). Although both viruses are in the genus Lagovirus, the etiologic agents of rabbit hemorrhagic disease and European brown hare syndrome are believed to be distinct species. 10 Other Caliciviruses The caliciviruses described above are the best characterized of the family Caliciviridae. Numerous other caliciviruses have been identified in a variety of species. Bovine enteric calicivirus, in the genus Norovinrs, has now been described as an agent of calf diarrhea (Liu et al. 1999). Yet to be adequately classified, other known caliciviruses include canine calicivirus, mink calicivirus, porcine enteric calicivirus, walrus calicivirus, lion calicivirus, chicken calicivirus, and other caliciviruses of birds (Murphy et al. 1999). As a family, members of Caliciviridae have fairly restricted host ranges, but a broad range of target organ systems and tissues. Biology of Feline Calicivirus Physical Properties Much of the early work on FCV detailed the virus’s physico-chemical properties. The virion size was estimated to be 37-40nm by filtration and electron microscopy (BiJrki 1965, Zwillenberg & Biirki 1966, Gillespie 8. Scott 1973). The virus is non-enveloped and the capsid is comprised of 180 copies of a single structural protein assembled as 90 homodimers in a T=3 icosahedral symmetry (Prasad et al. 1994). The sedimentation coefficient Of FCV was determined to be 1548 in sucrose gradients (Studdert 1978), with a virion density in cesium chloride estimated to be 1.37-1.38 g/ml based on the density of VESV (Wawrzkiewicz et al. 1968). Due to its morphology as an unenveloped virus, FCV is resistant to lipid solvents such as ether and chloroform and remains 11 partially stable in acidic environments to pH 4, but is almost completely inactivated at pH 3 (Gillespie & Scott 1973). FCV is heat inactivated at 500 for 30 minutes and MgClz does not stabilize FCV against heat inactivation (Studdert 1978). FCV is fairly resistant in the environment, able to persist for a week or more, but is susceptible to hypochlorite (bleach) solutions (Gaskell & Dawson 1998). FCV displays no hemagglutinating activity (Studdert 1978). The physical and chemical properties of FCV and the other members of Caliciviridae were sufficient to separate these viruses from their original classification as members Of the Picomavin'dae into a new family of viruses, the Caliciviridae. Genomic analyses, beginning in the 19905, would confirm this division (Green et al. 2000). Molecular Biology Complete sequencing of the genome of the FCV-F9 strain, the prototypical vaccine strain, was accomplished by primer extension cloning in 1992 (Carter et al. 1992). The complete genome of several other strains of FCV, from differing geographic locale and year of isolation, have subsequently been sequenced (Table 1). The FCV genome consists of nearly 7.7kb of positive sense RNA with a covalently-linked protein (VPg) at the 5’ end (Herbert et al. 1997) and a poly(A) tall at the 3’ end (Carter et al. 1992). Three Open reading frames (ORF) are present in the FCV genome (Figure 1A) (Green et al. 2002). ORF1 encodes nonstructural proteins labeled according to their sequence similarity to 12 picornaviruses, or to protein expression studies. These nonstructural proteins, which are cleaved by the viral proteinase, include p5.6; p32; p39, a 2C-like nucleoside triphosphatase; p30, a 3A-like protein; 38 (VPg); a 3C-Iike cysteine protease; and a 3D-like RNA-dependent RNA polymerase (Figure 13) (Glenn et al. 1999, Green et al. 2002, Sosnovtsev et al. 2002). ORF2 encodes the single capsid protein, which is subdivided into 6 regions based on strain variation with regions C and E being hypervariable and regions A, B, D and F showing greater conservation between strains (Figure 10) (Seal et al. 1994). The third ORF encodes a protein, labeled VP2, of unknown function that is incorporated with mature virions (Green et al. 2002). This genomic organization is shared with the Vesivirus and Norovirus genera and is in contrast to the Lagovirus and Sapovirus genera in which there exist only two ORFs — the first comprising the nonstructural polyprotein together with the capsid and the second containing VP2 (Green et al. 2000) Table 1. Complete FCV Sequences. FCV strains for which the complete genome sequence is available on GenBank with their accession number, year and country Of isolation, and the corresponding reference. NR = Not Reported. . A ' seem 33:33" $3.223. 32328.?" Reference F CV-F9 M86379 1960 USA Carter et al. 1992 FCV-F4 031836 1971 Japan Oshikamo et al. 1994 FCV-CFl/68 U13992 1960 USA Neill 1994 FCV-Urbana NC_001481 late 1960's USA Sosnovtsev & Green 1995 FCV-F65 AF109465 1990 UK Glenn et al. 1999 FCV-2024 AF479590 NR NR Thumfart & Meyers 2002 13 A 7317 7637 ORF3 5311 , 20 7683 5 VPg '- ORFI — ORF2 ——1 (14)“3’ 5314 7320 B ........................................ Nucleotide 2.0 161 1016 2078 2903 3216 4054 531:1 , 2C-like 3B 3C-Iike 3D-like 5 VPg P5 p32 NTPase 3A VPg Proteinase Polymerase Amino acid 1 47 332 686 961 1072 1763 Protein p5.6 p32 p39 p30 p13 p29? p47? H—J k v 4 p43 p76 C ORF1 ORF2 VP2 5’ VPg" r Capsid Protein ORF3—{(1911 3’ [—— 2004 bp—l , .................. Regions 360bp 828bp 15b‘ 72bpg 285bp 444bp A B C D E F VP2 Figure 1. Schematic diagram of the genomic organization of FCV. (A) The three open reading frames (ORF) of FCV (numbers correspond to nucleotides of the FCV-Urbana strain) with protein VPg at 5’ end and poly(A) tall at 3’; (B) the nonstructural proteins of ORF1; and (C) the organization Of the capsid gene (ORF2) into 6 regions with C and E being hypervariable (adapted from Green et al. 2002, Sosnovtsev & Green 2002). 14 Replication FCV is a lytic virus that replicates in the cytoplasm of host cells (Fenner et al. 1993). Production of both the complete genome transcript as well as a subgenomic transcript corresponding to the 3’ end of the genome occurs following infection (Fenner et al. 1993). The polyprotein translated from the complete genome transcript is cleaved to yield the individual viral proteins, while the subgenomic transcript serves as a template for translation of the capsid protein precursor (Neill & Mengeling 1988). The capsid protein precursor is a 75- kDa protein that is processed posttranslationally to generate the mature 62-kDa capsid protein after removal of 124 amino acids from the N terminus (Sonovtsev et al. 1998, Geissler et al. 1999). Antigenic Relationships Numerous strains of FCV have been isolated with widely varying serologic relationships (Bittle et al. 1960, Burki 1965, Takahashi et al. 1971). The first serologic survey of FCV reported two distinct groups of FCV: one group of eight isolates that were antigenically identical to the first American FCV isolate (CFI) (Crandell 8 Madin 1960) and a second group of 15 isolates that were antigenically unrelated (Bittle et al. 1960). Evaluation of seven additional isolates resulted in identification of five provisional FCV serotypes with little cross- reactivity (Burki 1965). Finally, 27 Japanese strains of FCV were compared serologically and divided into five serologic subgroups. When these 27 Japanese FCV strains were compared to 15 American FCV isolates, only two of 15 the American strains were neutralized by Japanese strain antisera (Takahashi et al. 1971). These results suggested wide antigenic variation between FCV strains and discouraged initial hopes for a vaccine. However, it was later shown that although significant variability existed between strains, most isolates could be neutralized with antiserum derived from the FCV-F9 strain (Kalunda et al. 1975). This led to the development of an intramuscular FCV vaccine based on the FCV- F9 strain. Vaccination began in earnest in the mid-1970’s. Although the FCV vaccine was apparently effective at reducing clinical signs, infection still occurred and viable FCV could be isolated from vaccinated cats after virulent challenge (Bittle & Rubic 1976, Kahn & Hoover 1976b). This was the first evidence that persistently infected, asymptomatic carrier cats (vaccinated or unvaccinated) likely played a critical role in the pervasiveness Of FCV. Transmission Transmission of FCV is via close contact with the nasal, oral, or ocular secretions from an infected animal, either symptomatic or an asymptomatic animal persistently shedding the virus (Sykes 2001 ). Environmental contamination and fomite transmission are also possible means of viral spread (Sykes 2001). Direct aerosol transmission is not believed to occur over distances greater than 4 feet (Hurley et al. 2004). 16 Pathogenesis Kahn and Gillespie (1971) further characterized the clinical syndrome induced by FCV. Aerosol exposure to FCV resulted in a biphasic temperature response in some kittens, characterized by a transient rise in temperature 24 hours after exposure and a second rise nearly one week later. Clinical signs persisted for seven to ten days and included conjunctivitis, rhinitis, pneumonia in some cases, and oral ulcerations on the tongue and hard palate. Pulmonary lesions were believed to result from cytolytic infection of the respiratory epithelium, particularly within the terminal bronchi and alveoli, with alveolar edema and seropurulent exudates following due to secondary bacterial pneumonia (Kahn & Hoover, 1976a). Mortality was 30% with all deaths occurring within two weeks Of viral exposure and 64% of deaths occurring within the first 5 days (Kahn & Gillespie 1971). Virus was isolated from the lungs, pharynx, conjunctival sac, nasal turbinates, third eyelid, trachea, and, infrequently, from the blood. Virus was consistently isolated from the tonsils until 34 days after exposure suggesting persistent shedding. Later studies confirmed an asymptomatic carrier state associated with chronic FCV infection that was characterized by persistent viral shedding up to 2.5 years following natural exposure (Povey et al. 1973). Because pathogenicity may vary considerably between FCV strains, the clinical picture of FCV-induced URTD can range from mild or subclinical infection to severe disease characterized by oral ulceration, anorexia, pneumonia, and even peracute death in young kittens (August 1984). Although most cats recover and 17 cease to shed FCV, 50% of cats will still be shedding the virus 75 days after infection and some will become persistent asymptomatic carriers, continuing to perpetuate the disease in susceptible feline populations. The percentage of asymptomatic cats shedding FCV is estimated to be 40% of colony cats, 25% of show cats, and 8% of household pets (Gaskell & Dawson 1998). Although primarily believed to affect the oral mucosa and respiratory tract, FCV has been isolated from the spleen, blood, intestine, feces, muscle, joints, kidney, urinary bladder, conjunctiva, tonsils, oral mucosa, and lungs of affected cats (Kahn & Gillespie 1971, Kahn & Hoover 19763, Truyen et al. 1999). The effect of FCV on these non-respiratory tract tissues is largely unknown as pulmonary and oral/respiratory mucosal lesions predominate during experimental and natural infection (Kahn & Hoover 1976a). FCV-induced URTD is often self-limiting and the course of infection typically lasts 5 to 7 days, if uncomplicated by secondary bacterial infection (Kahn & Hoover, 1976a). Hematological abnormalities usually involve a lymphopenia at 2-4 days post-infection. Resolution Of clinical signs coincides with development of FCV neutralizing antibodies at day five post- infection (Truyen et al. 1999). FCV is widely recognized as a cause of URTD. However, the variability in tissue tropism of FCV coupled with the ability of the virus to mutate rapidly is a cause for concern and makes future, as yet unknown, manifestations of FCV infection possible. 18 FCV-Induced Clinical Diseases Upper Respiratory Tract Disease The prevalence of FCV infection was first estimated to be 40% of all cases of URTD in cats, with FCV and feline herpes virus (FHV) combining equally to account for 80% of all cases (Kahn & Hoover 1976a). However, a retrospective study from the United Kingdom showed a much higher frequency of FCV isolation compared to FHV isolation from 1980 to 1989 (Harbour et al. 1991). In this analysis of 6,866 oropharyngeal swabs, FCV was isolated from 1,364 (19.9%) while FHV was isolated from 285 (4.2%) suggesting a ratio of FCV to FHV infection of 48:1 and giving a proportion of FCV among positive isolates of 82.7%. Furthermore, of the 1,180 cats with acute URTD, 348 (29.5%) were shedding FCV compared to 162 (13.7%) shedding FHV. Lastly, of the 963 cats positive for FCV for which vaccination history was known, 417 (43.3%) were fully vaccinated. This data led Harbour and coworkers to conclude that the increased prevalence of FCV compared to FHV was likely due to widespread use of respiratory vaccines, which reduced the shedding of FHV while having little impact on the shedding of FCV (Harbour et al. 1991 ). Today, FCV remains an important and common agent in feline upper respiratory tract disease. Recent reports of the prevalence of FCV showed 33% of cats with respiratory disease were FCV positive by virus isolation from oropharyngeal swabs; 21% of healthy animals were similarly positive (Binns et al. 2000). Despite extensive FCV vaccination over the last 3 decades, these results are 19 comparable to estimates of FCV prevalence prior to and for 10 years after introduction of a FCV vaccine (Kahn & Hoover 1976a, Harbour et al. 1991). The fact that FCV incidence is unchanged in the face of vaccination can be explained by one or more of the following possibilities: (1) FCV has evolved resistance to the vaccine, (2) the vaccine (live forms) may cause disease or chronic shedding, or (3) vaccination prevents clinical symptoms, but not the carrier state (Pedersen 8 Hawkins 1995). TO evaluate the relative contribution of each scenario to the persistence of FCV, Pedersen & Hawkins (1995) evaluated the degree of protection afforded, the persistence of shedding, and the pathogenicity of three current vaccine strains as well as several field isolates of FCV. This study found that immunization with the original FCV-F9 strain resulted in production of antibodies that neutralized most field strains. It was also apparent that oral administration Of vaccine strains could cause mild to moderate disease and persistent shedding of virus. Furthermore, this study demonstrated that cats exposed to field or vaccine strains Of FCV developed less severe disease when secondarily challenged, but were not protected against the carrier state. They therefore concluded that the persistence of FCV as a cause of feline URTD could not be explained by the evolution of vaccine resistant strains and that the vaccine strain may contribute to acute infection and the development of a persistent carrier state (Pedersen & Hawkins 1995). Future research is needed to develop effective strategies to reduce the prevalence and eventually eradicate this enigmatic pathogen of the feline respiratory system. 20 Arthropathy Although well recognized as a respiratory pathogen, FCV has been shown to be the cause of several other clinical syndromes. In the early 1980’s, FCV was isolated from two kittens from separate catteries (California and Ontario) with a similar presentation of high fever and shifting limb lameness (Pedersen et al. 1983). Studies of these viruses showed that clinical signs induced by the Californian strain could be prevented by vaccination, whereas clinical signs induced by the Canadian strain could not. Both isolates induced clinical disease characterized by pyrexia, depression, anorexia, reluctance to move, and limping in specific-pathogen-free (SPF) cats after oral or oronasal inoculation (Pedersen et al. 1983). However, only mild oral ulceration was observed after resolution of the lameness; signs of upper respiratory disease were not observed. FCV was cultured directly from synovial fluid Obtained from a 12-week-old kitten with signs of anorexia and severe shifting-limb lameness (Levy & Marsh 1992). This kitten, and three littermates, also had signs of URTD, which were not seen in the field or experimental cases of Pedersen et al. (1983). In all reported cases of FCV- associated lameness, the clinical signs resolved within two to four days with no known recurrence. Abortion FCV has also been associated with abortion in a small number of cats. FCV- related abortion was first observed when FCV was isolated from autolysed fetuses recovered during necropsy of a one-year-Old Siamese queen in Australia 21 (Ellis 1981). Interestingly, two other cats in the same household as the aborting Siamese had recently recovered from URTD. More recently, FCV was isolated from pooled organs of a fetus from an unvaccinated queen with a bloody vaginal discharge and four dead fetuses with petechial hemorrhage on the skin along their backs (van Vuuren et al. 1999). Experimental studies of FCV-infected pregnant queens have not been performed so the exact pathogenesis of FCV- induced abortion uncertain. However, these two reports suggest that FCV can be transmitted transplacentally and may be abortigenic. Chronic Gingivitis Dual infection with FCV and feline immunodeficiency virus (FIV) has been implicated as a cause of chronic gingivitis and pharyngitis in cats (Tenorio et al. 1991). In this study Of 226 cats from a veterinary teaching hospital, a shelter, and a purebred cattery, oral cavity disease was present in 43% to 100% of the animals from each group. Chronic FCV shedding was detected in roughly 20% of the animals in each group. Cats infected with FCV alone did not have a greater risk of oral lesions. However, cats that were FIV positive, and coinfected with FCV, had greater prevalence and severity of oral disease than cats solely infected with FIV. This finding of more severe disease in the presence of FCV suggests that although FCV may not play a primary role in chronic feline oral disease, it may enhance the severity of disease in immunocompromised individuals (Tenorio et al. 1991). 22 Virulent Systemic Disease An apparently new clinical manifestation of FCV infection has begun to emerge in the last decade. Initially described as a hemorrhagic-like fever (Pedersen et al. 2000), several outbreaks of highly pathogenic strains of FCV have been reported with a high mortality rate and systemic symptoms (Schorr-Evans et al. 2003, Hurley et al. 2004). The first outbreak occurred in Northern California among six cats with exposure to a private veterinary practice and was transmitted to four other healthy adult cats via personnel in an isolation ward (Pedersen et al. 2000). Clinical manifestations in these cats included cutaneous edema of the face and limbs, cutaneous lesions with crusting, erythema, and epilation on the face and pinnae, fever, signs consistent with URTD, and pancreatitis. Mortality rates in field cases, experimentally infected cats, and inadvertently infected cats ranged from 33% to 50%. Prior FCV vaccination did not appear to be effective for field cases, but did lessen the severity of signs associated with experimentally- induced disease. Similar outbreaks of highly virulent FCV have been subsequently reported in Southern California (Hurley et al. 2004), Pennsylvania, Tennessee, Nevada, and Massachusetts (Schorr-Evans et al. 2003). In these outbreaks, FCV infection was associated with severe edema, ulceration, icterus, pancreatitis, and often death. The Southern California strain was shown to be genetically and serologically distinct from the FCV strain involved in the original outbreak in Northern California, and also from the vaccine strain FCV-F9, with nucleotide homology of 73.4-76.5% for a portion of the hypervariable region of 23 the viral capsid gene (Hurley et al. 2004). All reported outbreaks were self- limiting and did not spread far beyond the index cases. Lower Urinary Tract Disease Disorders of the feline lower urinary tract have long been recognized as resulting in clinical signs of hematuria, periuria (urination in inappropriate locations), dysuria, and pollakiuria (Kirk 1925, Osbaldiston & Taussig 1970). Known causes of these clinical signs include uroliths, urethral plugs, infections (bacterial, fungal, or parasitic), anatomic abnormalities (congenital or acquired), and iatrogenic causes (Kalkstein et al. 1999b). Unfortunately, the cause of lower urinary tract signs cannot be determined in many cases, and these cats are diagnosed as having idiopathic lower urinary tract disease or, alternatively, idiopathic cystitis (IC). Idiopathic cystitis accounts for roughly two-thirds of all feline patients presenting for signs of hematuria, pollakiuria, and dysuria (Kruger et al. 1991, Buffington et al. 1997). In cats with urethral Obstruction, a cause could not be identified in 29% Of cases (Kruger et al. 1991). Nonobstructive IC has no sex or breed predilection and is typically encountered in young to middle-aged cats (mean age: 3.5 years, range of 0.5 to 17.5 years) from multi-cat households (Kalkstein et al. 1999a). In contrast, obstructive IC is encountered almost exclusively in male cats and is most likely related to formation of crystalline- matrix urethral plugs that more readily occlude the narrow penile urethra (Osborne et al. 1997). In the majority of cats with nonobstructive IC, clinical signs persist for five to ten days and then spontaneously resolve with or without 24 treatment (Osborne et al. 2000). This biological behavior has resulted in the speculation that feline IC may have a viral etiology (Kruger & Osborne 1990). The notion of a viral etiology for feline IC is not a new one. Hemorrhagic cystitis in people has been associated with a number of viral agents including adenovirus, herpes simplex, herpes zoster, cytomegalovirus, polyomavirus, influenza A, and human immunodeficiency virus (Kruger et al. 1996). In cats, a viral etiology for IC was first proposed in 1969 after urethral obstruction was experimentally produced in male cats following bladder inoculation of centrifuged, filtered, and bacteriologically sterile urine from male cats with naturally occurring Obstruction (Rich & Fabricant 1969). Shortly thereafter, a FCV was isolated from a Manx cat with spontaneous urethral obstruction and this isolate induced urethral obstruction in 80% of conventionally reared cats following urinary bladder, aerosol, or contact exposure (Rich et al. 1971). The role of FCV in cystitis was questioned, however, when (1) FCV could not be reisolated beyond the fourth day post-infection, (2) a second virus (feline synctia-forming virus) was isolated from all obstructed cats, and (3) no significant serum-neutralizing antibody response was detected in experimentally infected cats (Fabricant 1984). The role of FCV in viral-induced cystitis then came to be considered secondary, perhaps being able to incite latent herpesviruses into producing cellular injury and producing the clinical syndrome. This was supported by the isolation of a gamma herpesvirus, bovine herpesvirus-4 (BHV4), from a kitten who died of spontaneous urethral obstruction and who also experienced FCV-induced URTD 25 one month previously (Fabricant 1973, Kruger et al. 1989). Subsequent experimental studies were also supportive of this theory as urethral obstruction and cystitis were induced in specific-pathogen-free (SPF) male cats after urinary bladder or intravenous inoculation with BHV4 alone or BHV4 in combination with FCV (Fabricant 1977). However, other investigators were unable to reproduce cystitis and urethral Obstruction in SPF cats infected with the same BHV4 isolate (Kruger et al. 1990). The role of FCV in feline IC came increasingly into question when other investigators were unable to isolate the virus from cats with spontaneous forms of lower urinary tract disease (Kruger & Osborne 1990). However, these studies relied primarily on standard cell-culture-inoculation techniques. More recently, FCV-like particles were detected by electron microscopy in 35 of 92 (38%) crystalline-matrix urethral plugs (Kruger et al. 1996). This Observation led to the notion that virus isolation by conventional means may not be a unifome sensitive method of detecting FCV in feline urine. Virus isolation requires the presence of viable virus and the absence of substances that are toxic to cell culture or that inhibit viral replication (Sykes et al. 2001, Rice et al. 2002). Furthermore, FCV is extremely sensitive to the effects of urine, storage temperature, and freeze-thaw cycles and feline urine may be toxic to standard cell lines used for feline virus isolation (Appendix A, Komolate et al. 1976, Sykes et al. 1998, Rice et al. 2002). 26 The development of a modified virus isolation technique designed to circumvent some of the difficulties of FCV detection in feline urine has recently enabled the isolation Of two novel FCV strains from cats with IC (Rice et al. 2002). Designated FCV-U1 and FCV-U2, phylogenetic analyses of the capsid protein genes revealed that these strains were sufficiently different from standard vaccine strains to preclude the possibility that they represent urinary shedding of vaccine virus (Rice et al. 2002). They may, however, represent coincidental urine shedding of a non-pathogenic wild-type FCV strain or they may represent true uropathogens. Further studies are required to determine if FCV-U1 or FCV- U2 play a causative role in the pathogenesis of feline IC. The isolation of two new FCV strains, coupled with the finding of FCV-like particles in a large proportion of cats with obstructive IC, justifies reexamination of the potential role of FCV in feline IC. Although conventional viral diagnostic techniques have proven inconsistent for detection of FCV urinary tract infections, the ability Of molecular diagnostics to detect virus in a wide range of biologic substrates makes PCR-based detection methods an attractive means of reexamining the role of FCV in feline IC. Diagnosis of FCV Infection Virus Isolation Viral isolation has historically been the primary means of establishing a specific diagnosis of FCV infection. FCV grows rapidly in standard feline cell lines (Le. 27 Crandell-Reese feline kidney cells), exerting a rapid cytopathic effect in 1 to 3 days (Gillespie & Scott 1973). While very useful for FCV-induced URTD, standard cell-culture techniques may be less Optimal for isolation Of FCV from other biological substrates, such as feline urine. Virus isolation requires the presence of viable virus and the absence of substances that are toxic to cell culture or that inhibit viral replication (Sykes et al. 2001, Rice et al. 2002). FCV is extremely sensitive to the effects of urine, storage temperature, and freeze-thaw cycles and feline urine may be toxic to standard cell lines used for feline virus isolation (Appendix A, Komolate et al. 1976, Sykes et al. 1998, Rice et al. 2002). Virus isolation also has the drawbacks of being expensive and labor intensive, and may require days to weeks for final results (Fenner et al. 1993). Serology Detection of viral neutralizing antibodies offers an indirect method of identifying FCV infection and was one of the first methods employed to characterize FCV isolates (Bittle et al. 1960). In particular, a 4-fold increase in antibody titers in paired sera from the acute and convalescent phases of infection is indicative of recent exposure (Fenner et al. 1993). However, because of the often asymptomatic and persistent nature of FCV infection, a cause-and-effect relationship between viral antibody titer and a specific disease state should be made with caution (Kruger et al. 1996). Furthermore, the widespread use of FCV vaccines makes interpretation of serological data difficult in determining the role of FCV in a given disease (Kruger et al. 1996). 28 Electron Microscopy, Fluorescent Antibody, and lmmunohistochemistry Electron microscopy Offers direct morphologic identification of FCV virions due to the characteristic spherical shape, 37-40nm diameter virion size, and 32 dark spots visible due to negative stain filling the cup-shaped depressions on the surface of the virion (Studdert 1978). While electron microscopy Offers good specificity, it lacks sensitivity as it requires greater than 5 logs of virus per milliliter for detection (Castro 1992). The fluorescent antibody test (FAT) is another diagnostic tool employed for the direct detection of FCV antigens in tissues or fluids. Antiviral antibody bound to FCV antigens is detected by a fluorescein-conjugated immunoglobulin derived from cats hyperimmunized to FCV (Fenner et al. 1993). FAT has been described for FCV (Gillespie et al. 1971). Unfortunately, FAT testing is also labor-intensive and costly, and is less sensitive than virus isolation (Sykes 2001). FAT testing also may be associated with false positive results due to nonspecific fluorescing debris (Sykes 2001). lmmunohistochemistry (lHC) has also been used for direct identification of FCV antigens in tissues. lmmunohistochemistry is similar to FAT testing in that both methods utilize a primary anti-FCV antibody. However, lHC relies on an enzyme-labeled antiviral antibody and the production of a colored insoluble precipitate in the presence of the enzyme (Fenner et al. 1993). lHC has been used in the detection of FCV-infected tissues in conjunction with virus isolation 29 (Dick et al. 1989). Like IFA, immunohistochemistry may be associated with false- positive results (Fenner et al. 1993). Molecular Techniques Molecular diagnostic techniques such as nucleic acid hybridization and the polymerase chain reaction (PCR) have become increasingly popular for diagnosis of viral infections. DNA or RNA hybridization using labeled probes has proven to be a sensitive, specific, and fairly rapid method Of identifying viral nucleic acids in cells, tissue sections, or membrane fixed nucleic acids (Fenner et al. 1993). A dot blot DNA hybridization assay using a radiolabeled cDNA probe specific for the capsid protein of FCV has been reported (Dawson et al. 1994). This assay was found to be nearly twice as sensitive as conventional virus isolation in detecting FCV in tissues of infected cats (Dawson et al. 1994). PCR has become increasingly useful in the field of diagnostic virology due to this test’s sensitivity, specificity, scalability, and low cost (Fenner et al. 1993). In particular, PCR amplification allows for the direct detection of viral DNA. Direct sequencing of PCR products allows for detection of variations in genomic sequences between strains. PCR methods also can be employed to detect RNA viruses by incorporating a preliminary step in which the enzyme reverse transcriptase converts viral RNA to cDNA, which is then amplified by the PCR. Reverse transcription PCR (RT-PCR) assays have been developed for a wide variety of RNA viruses, including FCV. 30 RT-PCR assays for FCV initially focused on analysis Of the capsid protein gene and the genotypic determination of antigenic variability between FCV strains. The first reported primers for RT-PCR amplification of FCV were those given by Seal (1994), whose design was based on the FCV-CFI/68 strain. These primers amplified a 670bp portion of the capsid protein gene (Table 2) beginning at the C-terminus of region B and ending at the N-terminus of region F (Figure 2). Twelve different FCV isolates were amplified using this assay and all amplifications were verified by agarose gel electrophoresis, including amplification of 7 isolates that did not cross-hybridize with any of 5 cDNA capsid gene clones from known strains of FCV (Seal 1994). This assay was an effective means to evaluate nucleotide variation in the hypervariable region Of the capsid gene among divergent FCV strains. However, neither the sensitivity nor specificity of the assay was reported and the diagnostic range was limited to evaluation of only 12 FCV isolates. In 1996, a FCV polyprotein gene-based RT-PCR assay was reported (Doyle et al. 1996). This assay amplified a 477bp region of the p5.6 and p32 genes of ORF1 at the 5’ end of the genome (Figure 2, Table 2). This assay detected 21 of 24 (88%) field isolates of FCV and sensitivity was not reported. Although this assay was based in a highly conserved region of the FCV genome, its diagnostic utility was limited by its inability to detect all isolates of FCV that it was tested against. 31 One of the most frequently cited FCV RT-PCR assays is the nested RT-PCR of Radford and coworkers (Radford et al. 1997). This assay also amplified a portion of the FCV capsid protein gene with the outer primers delineating a 529bp product spanning from region B to region F, and with the inner primers delineating a 235bp product covering the 3’ end of region D and the 5’ portion of region E (Table 2, Figure 2). A variation of this assay has also been reported, which altered the combination of primers to increase the size of the final amplicon to 393bp (Table 2) and create a semi-nested RT-PCR assay (Radford et al. 2001). This assay and the nested RT-PCR have been repeatedly used to evaluate antigenic variation in genes encoding portions of the capsid protein, which form neutralizing antibody epitopes (hypervariable region E of the capsid protein gene). Epidemiologically related isolates were typically less than 5% distant and epidemiologically unrelated isolates were 20-40% distant from each other (Radford et al. 2000). There are no published reports of this assay’s sensitivity, specificity, or diagnostic range. The assay did, however, amplify RNA from 15 wild-type FCV isolates and 3 FCV vaccine strains (Radford et al. 1997). This assay was also used by Pedersen and coworkers (2000) to detect FCV RNA from an outbreak of highly virulent FCV, indicating that the diagnostic range ' of the assay is sufficiently broad to cover a new disease phenotype. The diagnostic range of this assay cannot, however, be quantitatively determined based on the data available. A nested RT-PCR provides greater sensitivity, but can be hampered by a higher probability of environmental contamination and false positive results, and requires more time and labor to perform. 32 Table 2. Feline calicivirus RT—PCR assays. This table comprises all previously reported FCV RT-PCR assays, together with the two assays reported here. Listed are the primers given by the individual authors (labeled as in the corresponding reference), the primer sequence (5’-3’), the size of the amplicon, the region of the FCV genome amplified (see Figure 2), the type of RT-PCR assay (gel-based or real-time), the diagnostic range of the assay as reported in the corresponding reference (listed as number of positive isolates / number tested), the corresponding nucleotides of strain FCV-Urbana (GenBank Accession number = NC_001481) covered by the assay, and the reference in which the assay was reported. * This paper utilized a semi-nested RT-PCR with P1 and P2 in the first round of amplification and P2 and P4 in the second round of amplification. The size of this amplicon was then 393bp. NR = Not reported 33 828 we. - 8 e e OeOOOOOeOO_ 865x Ne w v8-2 E: 8 o.-. O we no: _ OOmmeOL 8 m c _ m $3 8% me u f O @5595 SN _ OOOL ., A II 1 I n I 5 R35 - ewe - e 5.8 e _I» oeOn_ S _ . O BE 38 «:2 e e. O 355% e88 _ <On_ ...r .I J” . . v . . . . 5 Eeemm OO)10'5(Cl), 10*3 (A). 10 (O), and 10'8 (X) dilutions. All samples were amplified in triplicate (A) A graph of fluorescence vs. cycle number illustrating amplification of serial FCV dilutions with the threshold cycle varying inversely with virus concentration. (B) The melting temperature curve illustrating the difference in melting temperature between specific FCV amplification products (~820) and non- specific products (<76C). Specific FCV products were observed with amplification of the 1045 dilution. 67 Assay sensitivity was determined in triplicate by extracting and amplifying viral RNA from serial 10-fold dilutions of a FCV respiratory disease strain (FCV-R; starting titer = 5.2 x 108 TCIDso/ml) isolated in the virology section of the Michigan State University Diagnostic Center for Population and Animal Health (MSU- DCPAH). Specific FCV amplification products from all triplicate samples from the 10'1 to 10'5 serial FCV-R dilutions were detected Optically and visualized on agarose gels (Figure 7). One of three samples was positive for specific FCV amplification products at the 10'6 FCV-R dilution; none of the reactions at the 10'7 dilution gave a positive result. The threshold cycle (07) of each dilution with specific product varied inversely with virus concentration when plotted against the dilution factor (R2 = 0.974; data not shown). The lower detection limit of the RT-PCR assay corresponds to a FCV-R titer of between 1.2 x 101 to 1.2 x 102 TCID5o/ml. Our results are similar to those of previous studies of p30 gene- and capsid protein gene-based assays which suggest that RT-PCR assays are comparable in sensitivity to conventional virus isolation (Sykes et al. 1998, Helps et al. 2002). Assay linearity and efficiency were determined by amplifying serial 10-fold dilutions of FCV RNA in triplicate (Helps et al. 2002, Meijerink et al. 2001). The assay appeared linear over a dilution range of 6 logs, with a regression coefficient of 0.978 and a reaction efficiency of 95% (Figure 8). The RT-PCR reaction efficiency was higher than that reported for a recently described FCV p30 gene-based real-time RT-PCR assay (Helps et al. 2002). Increased reaction 68 35 7 30 - 2 25 t 0 5‘ 20 - 2 g 15 4 g 10 - y = 3.4526x + 8.1778 '- R’=0.9912 5 -l O T I I I I F I I -1 0 1 2 3 4 5 6 7 Dilution Factor (10910) Figure 8. Linearity of the FCV p30 gene-based real-time RT-PCR. Serial 10-fold dilutions of FCV RNA were amplified in triplicate. The CT was calculated for each reaction, plotted against the dilution factor (expressed as the IOQ1O of the dilution), and evaluated by linear regression analysis. 69 efficiency lowers the CT value, which may increase sensitivity in reactions with low template concentrations (Meijerink et al. 2001). Assay specificity was verified by bidirectional automated fluorescent sequencing of 8 FCV amplification products. Tissues infected with canine calicivirus (genus Vesivirus), a rabbit calicivirus (genus Lagovirus), and a bovine calicivirus (genus Norovirus) obtained from the MSU-DCPAH were used as additional specificity controls. The diagnostic range of the assay was evaluated by duplicate extraction and amplification of RNA from 51 randomly selected wild-type FCV strains isolated between 1990 and 2002 in the virology section Of MSU-DCPAH. The cDNA sequences of 8 FCV isolates had greater than 93% identity with one or more published FCV p30 gene sequences. When compared to consensus sequence. nucleotide substitutions were Identified in 13 positions that were nearly identical to those described for 17 FCV isolates from the United Kingdom (Helps et al. 2002). However, deduced amino acid sequences were highly conserved with substitutions in only two positions (S for T in 4 isolates and A for T in 2 isolates in positions corresponding to amino acid residues 826 and 828 of the Urbana strain respectively). Specific FCV amplification products were detected optically and visualized on agarose gels for all 51 North American FCV isolates; specific amplification products for the canine, rabbit, or bovine calicivirus isolates were not detected. Amplicon melt temperatures Of the FCV isolates ranged from 80.5C to 84C (mean 82.20 i 0.77). Specific FCV amplification products were easily distinguished from nonspecific products by examination of 70 the RT-PCR product melt curves (Figure 7). The diagnostic range of our assay exceeds that of previously reported FCV RT-PCR gel-based assays (Sykes et al. 1998, Horimoto et al. 2001, Radford et al. 2001) and is similar to that of a recently described real-time RT-PCR assay in which p30 gene-specific primers amplified RNA from 60 FCV field isolates collected in the United Kingdom (Helps et al. 2002). In our study, we evaluated specimens obtained from cats residing in the upper Midwestern United States. Strains of FCV originating from the same geographic area tend to be genetically related (Rice et al. 2002). Genetic clustering of isolates by geographic region could result in over-estimation of the effective diagnostic range Of the assay. However, our results, and those from the United Kingdom. suggest that p30 gene-based RT-PCR assays have a broader diagnostic utility compared to other FCV RT-PCR assays. Additional studies of FCV isolates obtained from diverse geographic locations and representing different disease phenotypes are indicated to further assess the diagnostic utility of FCV p30 gene-based RT-PCR assays. Variation in FCV amplicon melting temperature amongst clinical isolates was an intriguing finding in this study. Similar variation in FCV p30 gene amplicon melting temperatures has been observed by others (Helps et al. 2002). The shape and position of the cDNA melting curve is dependent on amplicon GC content, length, and sequence (Ririe et al. 1997). Slight variation in p30 gene nucleotide sequences observed between FCV isolates observed by us and others could alter the binding energy of the amplified cDNA and most likely 71 accounts for the minor differences Observed in specific FCV product melting temperatures. In conclusion, the primers used in this study amplified a large number of North American FCV isolates and further confirm the diagnostic utility of p30 gene- based real-time RT-PCR for detection of FCV. The p30 gene-based real-time SYBR Green l® RT-PCR assay described in this study is considered to be a rapid, accurate. and affordable method of detection of FCV infections. Footnotes 8 Doyle SA, MD Sussman, JM Kruger, RK Maes. 1997. A reverse-transcriptase polymerase chain reaction assay for detection of feline caliciviruses. Abstract. Journal of Veterinary lntemal Medicine; 11:148. b Scansen BA, JM Kruger, PJ Venta, et al. 2002. Development of a one-step reverse-transcriptase poly-merase chain reaction assay for detection of feline calicivirus. Abstract. Journal of Veterinary lntemal Medicine; 16:365. ° RNeasy Mini Kit®, Qiagen, Incorporated. Valencia, CA d QuantiTectTM SYBR Green RT-PCR Kit, Qiagen, Incorporated, Valencia, CA e iCyclerTM iQ® System with detection software V2.38, Bio-Rad Laboratories, Hercules, CA 72 COMPARISON OF RNA PREPARATION METHODS Scansen BA. JM Kruger, AG Wise, PJ Venta, P Bartlett. RK Maes. Comparison of RNA preparation methods for detection of feline calicivirus in urine by RT- PCR. Journal of Veterinary lntemal Medicine; submitted May 3'“. 2004. 73 Comparison of RNA Preparation Methods for Detection of Feline Calicivirus in Urine by RT-PCR Brian A. Scansen, BS1 John M. Kruger DVM, Phi)1 Annabel G. Wise DVM, Pho2 Patrick J. Venta PhD"2 Paul Bartlett DVM, PhD, MPH3 Roger K. Maes DVM, PhD2 Key Words: Running Title: Comparison of RNA Preparation Methods From the Departments of Small Animal Clinical Sciences‘, Microbiology and Molecular Geneticsz, Michigan State University College of Veterinary Medicine, East Lansing MI. An abstract of this work was presented at the 22nd Annual Veterinary Medical Forum, Minneapolis MN, June 2004. This study was supported by the Michigan State University Companion Animal Fund and Center for Feline Health and Well-Being. Reprint Requests: John M. Kruger DVM, PhD. Department of Small Animal Clinical Sciences, Room D208 Veterinary Clinical Center, Michigan State University, East Lansing, MI 48824-1314; (517) 355-6568; e-mail: kr0$r@cvm.msu.edu. 74 Abstract Investigating the causative role of feline calicivirus (FCV) in idiopathic cystitis may be facilitated by PCR-based diagnostic methods for detection of FCV urinary tract infections. The purpose of this study was to compare methods of RNA preparation from feline urine for amplification with a FCV p30 gene-based real- time reverse-transcriptase PCR (RT-PCR) assay. Urine and blood were Obtained from 6 specific-pathogen-free cats. Unaltered and centrifuged urine, and urine with added whole or hemolyzed blood, from each cat were spiked with FCV, and serially diluted in urine. FCV serially diluted in tissue culture medium served as positive controls. Viral RNA was prepared for RT-PCR by dilution and thermal inactivation (DT), polyethylene glycol precipitation (PEG), isolation with Oligo(dT)25-coated magnetic beads (dTMB), or extraction with two silica gel- based columns (RN and QA). All RT-PCR amplifications were performed in duplicate. Lower detection limit and mean RT-PCR threshold cycle (Ct) values for each RNA preparation method and for each sample type were compared with a mixed-effects model ANOVA. Because RNA prepared with DT yielded negative results, it was eliminated from final analyses. The lower detection limits (expressed as TCleo/sample) for the assay in urine were 1,950 for PEG, 104 for dTMB. 11 for RN, and 7 for QA. Mean Cr values for RN and QA were similar and significantly lower than those for dTMB and PEG (p<0.01). Urine modifications did not affect performance when samples were prepared with dTMB, RN, or QA. In conclusion, results of the RT-PCR assay were significantly better for RNA isolated from feline urine with either RN or QA. (257) 75 Introduction Feline caliciviruses (FCV) have been associated with a variety of clinical manifestations including, rhinitis, conjunctivitis. stomatitis, ulcerative glossititis. faucitis, pneumonia. enteritis, lameness, and abortion (Baulch-Brown et al. 1997, Sykes 2001). In addition, there is evidence supporting a potential causative role for FCV in the pathogenesis of feline idiopathic cystitis. Isolation of FCV from urine from cats with nonobstructive idiopathic cystitis, and observation of FCV- like particles in 38% of 96 urethral plugs obtained from male cats with obstructive idiopathic cystitis supported the concept of a viral etiology (Rich 8. Fabricant 1969, Kruger et al. 1996, Rice et al. 2002). However, large-scale epidemiologic studies and studies of experimentally induced FCV urinary tract disease have been hindered by lack of a sensitive and efficient means of detecting FCV urinary tract infections. Previous studies have relied upon recovery of live virus from urine using tissue culture-based virus isolation methods (Rich & Fabricant 1969, Kruger & Osborne 1990, Rice et al. 2002). Although virus isolation has been the “gold standard” for FCV detection, virus isolation is time consuming, expensive, and requires viable virus in the specimen and absence of substances that are toxic to cell culture or that inhibit viral replication (Sykes et al. 2001). Molecular diagnostic methods, such as reverse-transcriptase polymerase chain reaction (RT-PCR) circumvent many of the difficulties associated with conventional virus isolation methods and are increasingly being used for rapid detection of FCV (Sykes et al. 1998, Helps et al. 2002, Mackay et al. 2002). We recently reported development of a p30 gene-based real-time RT-PCR assay for detection of FCV 76 (Scansen et al. 2004). The p30 gene-based FCV RT-PCR assay was comparable to virus isolation in sensitivity and diagnostic range, and offered the advantages of reduced sample size, faster throughput, and better quantitation (Scansen et al. 2004). Although RT-PCR assays are sensitive and rapid methods for virus detection, it is also recognized that that urine is a particularly difficult substrate for amplification of nucleic acids (Demmler et al. 1988, Kahn et al. 1991, Behzadbehbahani et al. 1997, Echavarria et al. 1998, Biel et al. 2000). Urinary substances may compromise RT-PCR assay performance most likely by interfering with enzymatic reverse-transcription Of viral RNA or cDNA amplification by DNA polymerase. Consequently, preparation of nucleic acids becomes a critical step that not only serves to concentrate and purify nucleic acids. but also to remove or inactivate PCR inhibitors. Studies on human urine specimens indicate that the nucleic acid preparation method significantly influences the ability of PCR-based assays to detect viruses in urine and other complex biological specimens (Demmler et al. 1988. Kahn et al. 1991, Behzadbehbahani et al. 1997, Echavarria et al. 1998. Biel et al. 2000). Unfortunately. the optimal method of nucleic acid preparation varies. depending on the nature of the specimen, the type and quantity of inhibitory substances present in the sample, the physical, biochemical. molecular, and antigenic properties of the virus. and the susceptibility Of individual PCR assay components to inhibition. Studies investigating amplification of cytomegalovirus 77 DNA from human urine indicated that the inhibitory effects of urine on PCR performance could be effectively removed by simple ultrafiltration (Kahn et al. 1991). However, results of pilot studies in our laboratory revealed that use of a similar ultrafiltration device8 to remove urea and concentrate FCV in feline urine, resulted in concomitant concentration of unidentified substances that substantially inhibited the FCV RT-PCR (Rice et al. 2002). To our knowledge, studies comparing RNA isolation methods for their ability to remove or inactivate RT-PCR inhibitors, and preserve FCV RNA integrity in feline urine specimens have not been reported. Therefore, the purpose of this study to was to evaluate RNA preparation methods for their ability to recover FCV RNA from feline urine for real-time RT-PCR testing. Materials and Methods Collection and Preparation of Samples Urine and whole blood were Obtained from each of six 9-month-Old specific- pathogen-free female cats”. All cats were negative for FCV neutralizing antibodies. Approximately 15 ml of urine was collected aseptically by cystocentesis from each cat. In addition, 5ml of acid citrate dextrose anticoagulated whole blood was collected from each cat by jugular venipuncture. A complete urinalysis was performed on each urine sample. If necessary. 0.75mi to 1.25mi of RNase-free water was added to the urine specimen to obtain a total volume 15.75ml of urine for each cat (Table 3). This volume of urine was required for preparation of serial dilutions Of FCV in urine. 78 Table 3. Characteristics of urine specimens obtained from six 9-month-old female specific-pathogen-free cats used in the study. A small quantity of RNase- free water was added to each urine specimen from each cat to Obtain the final volume of 15.75ml per cat required for analyses. Cat Identification Number Variable 1 2 3 4 5 6 '11:?" ”me V°'”'“e 15 15 14.75 14.75 15 14.5 Water Added (ml) 0.75 0.75 1 1 0.75 1.25 m?" ”me VO'Ume 15.75 15.75 15.75 15.75 15.75 15.75 USG Predilution 1.031 1.020 1.014 1.026 1.025 1.004 IUSG Postdilution 1.030 1.019 1.013 1.024 1.023 1.003 Urine pHa 7.5 7.0 7.0 7.5 7.5 7.0 Urine Occult Blooda 3+ Neg 3+ 3+ 3+ 3+ Urine Protein"I 1+ Trace Trace Trace Trace Trace Urine RBC (per “320 D 100-200 0-2 TNTC 1-4 8-10 2-5 Urine WBC (per hpf)b OCC 0'2 0'5 0'2 1‘3 1‘3 lUrlne Epithelial Cells Occ 0-2 Occ 0 cc 0 cc O c c (Per hpj) Urine C stals Many Occ Occ Many ,(ger my)? MAP MAP ”99 MAP MAP Neg pr = high power (400x) field; MAP = magnesium ammonium phosphate; Neg = negative; Occ = occasional; RBC = red blood cells; TNTC = too numerous to count; USG = urine specific gravity; WBC = white blood cells a. Semiquantitatively estimated by reagent test strips. b. Median number per high-power field (400x) from urine sediment. Urine from each cat was divided into 4 aliquots Of 3.75mi each (Figure 9). One urine aliquot was centrifuged at 300 x g for 15min at 4C. The cell-free urine supernatant was removed and used for subsequent analyses. The second urine aliquot was unaltered. TO the third urine aliquot, 3.75pl of whole blood from the same cat was added to simulate gross hematuria (Osborne & Stevens 1999). To the fourth urine aliquot, 3.75ul of hemolyzed blood from the same cat was added 79 to simulate hemoglobinuria. Hemolyzed blood was prepared by freezing and thawing whole blood 4 times. Serial 10-fold dilutions (10‘1 to 10”) of a FCV respiratory disease strain (FCV-R; titer = 5.2 x 108 TCIDso/ml), isolated in the Virology Section of the Michigan State University Diagnostic Center for Population and Animal Health, were made in each urine specimen and in tissue culture mediumc (positive control). Nucleic acids were isolated from each of the 10“ through 10'7 dilutions of each urine and tissue culture medium sample using each of the RNA isolation methods described below (Figure 9). A sample of uninfected tissue culture medium served as a negative control for each RNA extraction. RNA Isolation Methods Based on studies in other species, 5 methods of preparing viral RNA from feline urine were selected for study. These methods included (1) dilution and thermal inactivation (DT), (Biel et al. 2000) (2) polyethylene glycol precipitation (PEG), (Behzadbehbahani et al. 1997) (3) a commerciai mRNA extraction kitd incorporating Oligo(dT)25-coated magnetic beads (dTMB) to capture FCV polyadenylated RNA, (Chiodi et al. 1992, Dynal Handbook 2000) (4) a commercial silica gel-based extraction column methode (RN) designed for isolation of total RNA from highly cellular material, (RNeasy Handbook 2001) and (5) a commercial silica gel-based extraction column methodf (QA) designed for 80 Figure 9. A schematic representation of study design. Urine and blood specimens were obtained from six 9-month-old female specific-pathogen-free cats. Unaltered urine, urine with added whole or hemolyzed blood, centrifuged urine supernatant, and tissue culture medium were spiked with FCV and serially diluted. Viral RNA was prepared from samples with each isolation method and amplified in duplicate with the FCV RT-PCR assay. 81 N x MUm-HM N x MUQQR N x Moguhm N x MMmFM % A _ a _ 85552552355<2m _ <0 _ 72m _ _m:>:.u_ .8225 fiam ES: <2: 35> 5: em |_ _H:: :: :: :::: l: 1: 1:1: 1: 5-21:1: l: 1: 1:1: no“ em: “.21: no“ 93 i: we >Umuc< >Umcv< >Um Eu< >Um 33¢. >Um 3;. 3.5.30 55:52 0mm UoNhBEo: + Um:— + 0:23 __ 0.5:.— _ 95.;— o..===U 2.3:. V 05.5 touztbauflsz cuwaonuaoZ , “EMF—2:59.32 wowzfihaob a cubes—o: 586on BED mmooobm coca < $15 so “am 82 isolation of viral nucleic acids from body fluids with low-cellularity (Echavarria et al. 1998, QlAamp Handbook 1999). For DT, urine was diluted 1:10 in RNase-free water, incubated at 95C for 5 minutes, cooled, and used directly for RT-PCR (Behzadbehbahani et al. 1997, Biel et al. 2000). However, RNA prepared from higher dilutions of FCV in urine, but not in tissue culture medium, with DT yielded negative results with the RT- PCR assay (data not shown). Since samples prepared with PEG, dTMB, RN, and QA consistently yielded positive RT-PCR results in urine at higher dilutions, DT was not evaluated in subsequent studies. Isolation of viral nucleic acids from urine using PEG has been described (Behzadbehbahani et al. 1997). Briefly, 150pl of sample was mixed with 50pl of 30% w/v polyethylene glycol9 in 3M sodium chloride, incubated for 30 min on ice, and centrifuged at 10,000 x g for 15min. The supernatant was discarded and the pellet resuspended in 20pl of a 10mM Tris-HCl buffer (pH — 7.6) with 0.5% v/v nonionic detergent“. The suspension was then incubated for 10 min at room temperature and stored at —8OC. For dTMB, viral polyadenylated RNA was isolated from 100 pl of sample using Oligo(dT)25-Coated magnetic beadsd according to the manufacturer's instructions for viral poly A+ RNA isolation (Dynabeads® mRNA DIRECTTM Kit Handbook 2000). Briefly, 100p! of urine sample was mixed with 300pl of lysis/binding buffer 83 (100mM Tris-HCI, 500mM LiCl, 10 mM EDTA, 1% LiDS, and 5 mM dithiothreitol) and 25pl of preconditioned magnetic beads, and incubated at room temperature for 10min. The beads were then washed twice with 500pl of buffer containing 10mM Tris-HCI, 0.15mM LiCl, 1mM EDTA, and 0.1% LiDS, and twice more with 500ul of buffer containing 10mM Tris-HCI, 0.15mM LiCl, and 1mM EDTA. After a final wash with Soul of 10mM Tris-HCI, viral RNA was eluted from the magnetic beads by incubation in 25pl of 10mM Tris-HCI at 65C for 2min. Eluted RNA was then stored at —8OC prior to use. For RN, total RNA was isolated from urine samples according to the manufacturer’s instructions for animal cellse (RNeasy Mini Handbook 2001). Briefly, 200p| of urine sample was mixed with 350pl of a proprietary lysis buffer containing guanidine thiocyanate and B-mercaptoethanol, and vortexed at room temperature for one minute. The lysate was then mixed with an equal volume of 70% ethanol, applied to the silica gel-based extraction spin column, centrifuged, washed once with a proprietary buffer containing guanidine thiocyanate and ethanol, and washed twice with a second proprietary buffer. Total RNA was then eluted in 50pl of RNase-free water and stored at —800 until use. For QA. viral RNA was isolated from urine samples according to the manufacturer’s instructionsf (QlAamp Viral RNA Mini Kit Handbook 1999). Briefly, 140ul of sample was mixed with 560 pl of a proprietary lysis buffer containing guanidine thiocyanate and 10pg/ml of carrier RNA, and incubated at 84 room temperature for 10 minutes. The lysate was then mixed with 560 pl of 100% ethanol, applied to the silica gel-based extraction spin column, centrifuged, washed once with a proprietary buffer containing guanidine hydrochloride and ethanol, and washed once with a second proprietary buffer. Viral RNA was then eluted in 60pl of RNase-free water containing 0.04% sodium azide, and stored at —8OC until use. FCV RT-PCR Assay Viral RNA from each sample was assayed in duplicate using a FCV p30 gene- based real-time RT-PCR protocol and a one-step RT-PCR systemi (Scansen et al. 2004). Briefly, RT-PCR was performed in a 50pl reaction volume containing 25pl of 2x buffer (a proprietary buffer containing Taq DNA polymerase, SYBR® Green I, dNTPs, and 5.0 mM MQCI2),i 0.5pl of mixed reverse transcriptases,i 0.5 pM of the fonNard primer 5’-TGGATGAACTACCCGCCA, 0.5 pM of the reverse primer 5’-GCACATCATATGCGGCTC, 5 pl of sample RNA, and RNase-free water. Real-time RT-PCR amplification was performed in a thermal cycler with an integrated real-time optical detection system.j Cycling conditions consisted of reverse transcription at 500 for 30 minutes and a preliminary denaturation step at 95C for 15 minutes, followed by 38 cycles of 94C for 303, 53C for 308, 720 for 603, and 780 for 128. These cycling conditions are a variation from those previously reported (Scansen et al. 2004) with the addition of a data acquisition step at 78C for 12s to reduce the fluorescent signal from non-specific product formation (Mackay et al. 2002). Amplification of cDNA was continuously 85 monitored in real-time by quantifying the amount of fluorescence emitted at 530nm at each data acquisition step (780). Following a post-amplification step at 55C for one minute, amplicon melt temperatures were determined by raising the temperature in 0.50 increments from 550 to 95C. Post-PCR analysis was performed using iCycler detection software]. Samples giving peak fluorescence between 80.5C and 84.50 were determined to be FCV-specific product, while samples with peaks below 760 were determined to be the result of non-specific amplification. Samples with discordant RT-PCR results were assayed a third time and the majority result used in data analysis. Data Analyses The lower detection limit (expressed as TCleo/sample) for a given RNA extraction method was defined as the highest sample dilution at which FCV RNA was amplified from 3 or more of the 6 samples. The final lower detection limit of the system was determined by factoring the volume of the initial sample and the proportion of the extracted RNA used in the amplification. In addition, the logz of the mean RT-PCR detection threshold cycle (Ct) value for each RNA preparation method and for each urine specimen modification and positive control were determined at the 10“ dilution. The lower detection limits and mean 0, values for each RNA preparation method for each urine specimen modification and positive control were compared with a mixed-effects model ANOVA, with cat as a random effect variable and 3 fixed-effect variables (urine specimen, RNA preparation method, and their two-way interaction) (SAS/STAT 1989). 86 Results Un'ne Specimens The volume of urine collected from each cat ranged from 14.5 to 15 ml (Table 3). Mean urine specific gravity prior to dilution was 1.020 t 0.009. Dilution of urine specimens to a final volume of 15.75 mi resulted in a small decrease in urine specific gravity of approximately 0.001. All urine specimens were neutral to slightly alkaline in pH. Reagent test strips were positive for occult blood or hemoglobin in 5 of 6 (83%) urine specimens, and for mild proteinuria in one of 6 (17%) urine specimens. Urine sediment examination revealed microscopic hematuria in 3 of 6 (50%) specimens and struvite crystalluria in 4 of 6 (67%) specimens; pyuria and bacteriuria were not detected any sample. Effect of RNA Isolation Method on RT—PCR Performance The lower detection limit of the RT-PCR assay for urine specimens varied significantly by RNA isolation method (mixed-effects ANOVA p<0.0001). The detection limit of the RT—PCR assay was lower when RNA was prepared with QA (Table 4). However, the lower detection limits for QA, RN, and dTMB were not statistically different. In contrast, the detection limits for these 3 methods were significantly lower than that of PEG (Mixed-effects ANOVA, p<0.01). There was no significant interaction between sample type, isolation method, and the lower detection limit of the RT-PCR assay. 87 Table 4. Mean lower detection limits of the FCV RT-PCR assay using RNA prepared by four RNA isolation methods. Urine and blood specimens were obtained from six 9-month-old female specific-pathogen-free cats. Unaltered urine, urine with added whole or hemolyzed blood, centrifuged urine supernatant, and tissue culture medium were spiked with FCV and serially diluted. Viral RNA was isolated from samples with each of 4 preparation methods and amplified with the FCV p30 gene-based RT-PCR assay. RNA Mean Lower Detection Limit (TClD5olsample)“ Fragifigon Certhlriiaueged Unaltered Urine + Urine + 5:32;: Supernatant Urine Whole Blood Hemolyzed Blood Medium PEG 1950 1950 1950 1950 195 dTMB 104 104 104 104 104 RM 11 1 1 11 11 11 QA‘ 6 6 6 6 6 DTMB = oligo(dT)«coated magnetic beads (Dynabeads mRNA Direct Kit®) PEG = polyethylene glycol precipitation; QA = silica-gel-based extraction column (QlAamp Viral RNA Mini Kit®); RN = silica-gel-based extraction column (RNeasy Total RNA Mini Kit®); TCleo- median tissue culture infectious dose. a. Lower detection limit defined as the highest sample dilution at which FCV RNA was amplified from 3 or more of the 6 samples. The final lower detection limit of the system was determined by factoring the volume of the initial sample and the proportion of the extracted RNA used in the amplification. Similarly, mean C. values for the RT-PCR assay varied significantly by RNA isolation method (Mixed-effects ANOVA, p<0.0001). The C. value corresponds to the PCR cycle number at which the fluorescence increases from a low background level to a detectable level, and reflects starting template concentration and amplification efficiency (Mackay et al. 2002). Lower Ct values imply higher starting template concentration, more efficient amplification, or both. Although QA resulted in the lowest mean Ct values over all sample types, there was no significant difference between mean Ct values for samples prepared with 88 QA or RN (Table 5). Amplification of RNA prepared with dTMB resulted in a significantly lower Ct value compared to PEG, but a significantly higher Ct value compared to QA and RN (mixed-effects ANOVA, p<0.01). In addition, analysis of mean Ct values revealed a significant interaction between isolation method, sample type, and mean Ct value (mixed-effects ANOVA, p<0.0001). Effect of Urine Variables on RT—PCR Performance There was no apparent association between the urine specific gravity of each individual cat’s urine sample and the Ct values of their respective amplifications using RNA prepared with QA, RN, and dTMB (Figure 10). With the PEG method, however, Ct tended to increase with increasing USG (linear regression, R2 = 0.56; Figure 10). Although sample type did not significantly affect the lower detection limit, sample type significantly affected mean C, values for the RT-PCR assay (mixed-effects ANOVA, p<0.0001). Compared to unmodified urine, modification of urine by centrifugation, or addition of whole or hemolyzed blood did not significantly influence mean Ct values for any of the RNA preparation methods (Table 5). When compared to tissue culture medium positive control preparations, there were no significant effects of urine or any urine modification on mean Ct values for QA, RN and dTMB (Table 5). However, the mean Ct value for the tissue culture medium positive control prepared with PEG was significantly lower than the mean Ct values for any urine specimen similarly prepared (mixed-effect 89 38 7 3 PEG 2 _ 36 g R — 0.56 34 ~ 32 - U E 30 — a) ... 2 28 = . 26 - ’ 9 o 24 1 ‘ ‘ . ‘ ‘ 22 1 ' I I . . o 20 T l 1 l T j 1.000 1.005 1.010 1.015 1.020 1.025 1.030 Urine Specific Gravity Figure 10. Effect of individual cat urine specific gravity on threshold cycle (Ct). The association between urine specific gravity and Ct was evaluated by linear regression analysis. 90 ANOVA, p<0.01; Table 5). The mean Ct values for the tissue culture control preparation prepared with PEG where 9.4 to 10.6 cycles lower than those for any urine specimen (Table 5). This increase in mean Ct value represents an approximately 1,024-fold decrease in assay performance when amplifying RNA prepared from urine with PEG compared to tissue culture medium positive controls prepared in a similar manner. Table 5. Geometric mean detection threshold cycle (Ct) of the FCV RT-PCR assay using RNA prepared by four RNA isolation methods. Urine and blood specimens were obtained from six 9-month-old female specific-pathogen-free cats. Unaltered urine, urine with added whole or hemolyzed blood, centrifuged urine supernatant, and tissue culture medium were spiked with FCV and serially diluted. Viral RNA was prepared from samples with each isolation method and amplified in duplicate with the FCV RT-PCR assay. RNA_ Geometric Mean Detection Cycle Threshold (Ct)" Prapiaatcllon Centrifuged Urine Unaltered Urine + Urine + Tissue Culture 6 o Supernatant Urine Whole Blood Hemolyzed Blood Medium PEG 34.8 34.7 35.7 35.9 25.3 dTMB 26.5 25.6 26.5 24.9 24.1 RN 23.3 23.7 22.6 23.1 23.2 QA 22.6 22.5 21.7 21.5 22.6 Ct = detection threshold cycle; dTMB = Oligo(dT)-coated magnetic beads (Dynabeads mRNA Direct Kit®); PEG = polyethylene glycol precipitation; QA = silica-gel-based extraction column (QlAamp Viral RNA Mini Kit®); RN = silica- gel-based extraction column (RNeasy Total RNA Mini Kit®) a. The geometric mean RT-PCR detection Ct value for each RNA preparation method and for each urine specimen modification and positive control were determined at the 10“ dilution. 91 Discussion Our results indicate that RNA preparation method significantly influences the ability of the RT-PCR assay to detect FCV in feline urine. Thermal inactivation and dilution of urine has proven effective for enhancing detection of polyomavirus, cytomegalovirus, and Chlamydia DNA by PCR from human urine specimens (Vinogradskaya et al. 1995, Mahony et al. 1998, Toye et al. 1998, Biel et al. 2000). In contrast, our initial studies indicated that DT was unable to alleviate the inhibitory effects of feline urine on RT-PCR performance. The inability of DT to enhance FCV detection may be due to species-related differences in urine composition or variations in the susceptibility of the assay to the effects of inhibitory substances. Interestingly, studies examining thermal inactivation for improved viral detection in fecal specimens suggest that the efficacy of thermal inactivation is species-dependent (Uwatoko et al. 1996). It is plausible that such a difference exists between feline urine and human urine samples as well. Of the four remaining preparation methods, PEG was the most rapid, economical and technically least demanding. However, isolation of viral nucleic acids from urine with PEG resulted in the poorest RT-PCR assay performance. Diminished assay performance may be due to more limited removal or inhibition of urine RT- PCR inhibitors, lower RNA yield, or both. The relative decrease in RT-PCR performance for all PEG-prepared urine specimens, compared to PEG-prepared tissue culture medium specimens, suggests that inhibitory substances present in 92 feline urine were not effectively removed or inactivated by PEG. Our findings are in contrast to those of other studies, in which PEG appeared to be one of the methods of choice for preparation of urine for PCR-based detection of human cytomegalovirus and polyomavirus DNA (Yamaguchi et al. 1992, Vinogradskaya et al. 1995, Behazadbehbahani et al. 1997). The reason for this discrepancy is unknown. Although feline urine would be expected to contain similar inhibitory substances, it is probable that quantitative and qualitative differences in composition exist between feline and human urine. In addition, it is possible that the reverse transcriptase required in our assay, but not in those for cytomegalovirsus or polyomavirus, was inhibited by urine substances that were not inactivated or removed by PEG. The poor RT-PCR performance associated with RNA prepared with PEG may also be due to decreased RNA yield. Although RNA yield was not quantified, the observation that assay performance was significantly decreased for tissue culture medium positive controls prepared with PEG compared to other preparation methods, suggests that PEG resulted in a lower RNA yield. It is likely therefore, that a combination of decreased RNA yield and failure to remove or inactivate all urine RT-PCR inhibitors were responsible for poor RT-PCR assay performance with PEG prepared specimens. The Oligo(dT)25-coated magnetic bead method of RNA preparation relies on direct base pairing between the poly-adenylated tail of viral genomes and the 93 oligo dT sequences bound to the surface of magnetic beads (Kingsley & Richards 2001). This method has been successfully used for purification of human calicivirus (Norwalk virus), hepatitis A virus RNA from shellfish (Kingsley & Richards 2001), and HIV RNA from human cerebral spinal fluid specimens (Chiodi et al. 1992). Since the FCV genome is similarly composed of single stranded RNA with a poly-adenylated tail, it was logical to hypothesize that dTMB would be effective for isolation of FCV RNA from urine. Although the magnetic bead method significantly improved RT-PCR performance compared to PEG, isolation of FCV RNA from urine with dTMB was not as effective as either of the silica gel-based column methods. Since the RT—PCR assay performed equally well with RNA prepared from urine or tissue culture medium, the relative decrease in RT—PCR performance associated with samples prepared with dTMB compared to QA and RN, was most likely due to decreased RNA yield. It is possible that RNA yield with magnetic beads could be improved by using alternative nucleic acid capture strategies. Human calicivirus (Nonrvalk-like) RNA has been successfully prepared from environmental samples for RT-PCR by using magnetic beads coated with virus-specific antibodies (immunomagnetic- bead separation), or streptavidin-coated magnetic beads and biotinylated oligonucleotides (Loisy et al. 2000, Myremel et al. 2000). However, immunomagnetic—beads were not equally effective for isolating all antigenic types of Norwalk-like caliciviruses (Myremel et al. 2000). Antigenic variation among FCV isolates could similarly limit the use of immunomagnetic-beads for isolation of FCV from biological samples (Baulch-Brown et al. 1997). It may be of value to 94 investigate whether capture with specific biotinylated FCV oligonucleotides, or nonspecific biotinylated oliogo dT, and streptavidin-coated magnetic beads would enhance the performance of the FCV RT-PCR assay. Preparation methods using silica-gel-based membranes rely on selective nucleic acid binding by silica. These methods have been used extensively for isolation of viral RNA or DNA from urine, feces, CSF, serum, and other complex biological samples (Biel et al. 2000, Echavarria et al. 1998). Since feline urine may contain variable numbers of cells, and since FCV may be intracellular, extracellular, or both, we evaluated the performance of two silica-gel-based membrane methods: RN, designed to isolate total RNA from highly cellular preparations, and QA, designed to maximize recovery of viral RNA from cell-free fluids. Both methods use chaotropic salts to lyse cells and virus particles, and to inactivate RNases prior to membrane binding. However, QA incorporates carrier RNA in the lysis step to improve viral RNA binding and to competitively limit viral RNA degradation due to residual RNase activity. Performance of the RT-PCR was significantly better using samples prepared with QA and RN than with samples prepared with PEG and dTMB. Improved RT-PCR performance associated with QA and RN was most likely due to a combination of higher RNA yield per extracted volume of sample and more effective removal or antagonism of urine RT-PCR inhibitors. Further studies are needed to determine whether RN and QA are equally suitable for preparing viral RNA under all extremes of sample conditions encountered in cats with or without urinary tract disorders. 95 To the extent that we attempted to modify urine by removal of urine sediment, or addition of whole or hemolyzed blood, we were unable to detect significant differences between dTMB, QA, and RN on RT-PCR performance. Several substances have been shown to inhibit PCR and RT reactions including urea (Kahn et al.1991), red blood cells (Panaccio & Lew 1991), heme compounds, (Bymes et al. 1975, Tsutsui 8 Mueller 1987, Levere et al. 1991, Klein et al. 1997, Morata et al. 1998, Al-Soud & Radstrom 2001) leukocytes (Morata et al. 1998, Al-Soud & Radstrom 2001), immunoglobin G (Al-Soud et al. 2000) and crystalluria (Mahony et al. 1998). Some or all of these inhibiting substances may be present in urine specimens obtained from normal cats or cats with urinary tract disorders. Although there were no apparent sample effects in this study, our results should be interpreted within the context of a limited sample size. A type II statistical error (i.e. failure to reject the null hypothesis) can occur when the sample size is inadequate to detect small, but clinically meaningful, effects. Likewise, small sample size in the present study precluded analysis of cat- specific variables (eg. urine specific gravity, proteinuria, crystalluria) that may affect RNA isolation and RT-PCR performance. The influence of specific feline urine components on RNA preparation and FCV RT-PCR assay performance requires further investigations. In conclusion, there are notable differences between RNA isolation methods for recovery of FCV nucleic acids from feline urine. The FCV p30 gene-based RT- PCR assay performed significantly better when using RNA isolated from feline 96 urine with either of two silica gel-based extraction column methods. Our results underscore the need for species-specific studies to determine the optimal method of nucleic acid preparation from a particular clinical sample for a particular assay system. A RT-PCR assay system optimized for detection of FCV in feline urine may be useful for large-scale epidemiologic and experimental studies of FCV-induced urinary tract diseases. Footnotes a — Ultrafree-CL ultrafiltration device, Millipore Corporation, Bedford, MA b — Harlan Bioproducts, Indianapolis, IN c — Eagle’s minimum essential medium, Gibco BRL, Grand Island, NY d — Dynabeads mRNA DIRECT Micro Kit®, Dynal Inc., Lake Success, NY e — RNeasy Mini Kit®, Qiagen, Incorporated, Valencia, CA f- QlAamp Viral RNA Mini Kit®, Qiagen, Incorporated, Valencia, CA 9 — Polyethylene glycol, 8,000 average molecular weight, Sigma-Aldrich, St. Louis, MO h - lgepal CA-630, Sigma-Aldrich, St. Louis, MO i — QuantiTectTM SYBR Green RT-PCR Kit, Qiagen, Incorporated, Valencia, CA j — iCyclerTM iQ® System with detection software V2.38, Bio-Rad Laboratories, Hercules, CA 97 APPENDICES 98 APPENDIX A Effects of Storage Temperature on Detection of FCV in Urine and Tissue Culture Medium by Virus Isolation and RT-PCR. Material and Methods - Feline calicivirus (FCV; strain FVC-R, starting titer 2.4 x 108 TCID5o/ml) was serially diluted in pooled urine (urine specific gravity 1.057) obtained aseptically by cystocentesis from five 1.25 to 2 year-old female specific- pathogen-free cats, or in tissue culture medium. One aliquot of each 10“ to 10‘9 dilution was stored at 4C for 24 hours prior to assay. One additional aliquot of each dilution was stored at —7OC for 5 days prior to assay. For virus isolation, 200 pl of each 10“ to 10'9 dilution was added to 800 pl of pooled urine or tissue culture medium and concentrated approximately 10-fold by centrifugation in an ultrafiltration device (Ultrafree-Cl ultrafiltration device, 10,000MW cutoff; Millipore Corp, Bedford MA). Material retained within the filter device was resuspended in 1 ml of tissue culture medium, gently sonicated, and inoculated into 2 tissue culture roller tubes as previously described (Rice et al. 2002). For the FCV RT- PCR assay, viral RNA from 140 pl of each 10“ to 10'9 dilution was isolated using a silica-gel-based extraction column (QlAamp Viral RNA Mini Kit®; Qiagen Incorporated, Valencia CA), and amplified in triplicate using a 1-step real-time FCV RT-PCR assay previously described (Scansen et al. 2004). The lowest detection dilution for virus isolation was defined as the highest sample dilution at which FCV CPE was observed in one or more roller tubes. The lowest detection dilution for the RT-PCR was defined as the highest sample dilution at which FCV RNA was amplified from 2 or more of the 3 RT-PCR replicates. The final lower 99 detection limit of each system was determined by factoring the volume of the initial sample and the proportion of the extracted RNA used in the amplification. Results- Virus isolation and RT-PCR had similar lower detection limits for detecting FCV in urine and tissue culture medium stored at 4C (Tables 6 and 7). Storage of FCV in tissue culture medium at —70C did not appear to substantially affect virus titer compared to samples stored at 4C (Table 7). In contrast, storage of FCV in urine at —70C resulted in a substantial (greater than 1000-fold) loss of virus titer as determined by virus isolation when compared to tissue culture medium controls (Table 7). Likewise, storage of FCV in urine at -70C resulted in a 10-fold loss of sensitivity of the RT-PCR assay compared to tissue culture controls. Conclusions - Virus isolation and the RT-PCR assay were similar in their ability to detect FCV in urine and tissue culture medium stored for 24 hours at 4C. However, detection of FCV in urine specimens by virus isolation may be profoundly compromised by storage at —70C. Similarly, but to a lesser extent, detection of FCV in urine by RT-PCR also may be affected by storage of samples at —70C. These results indicate that proper sample handling is an essential prerequisite for detection of low concentrations of FCV in urine by virus isolation and RT-PCR. 100 Table 6. Number of positive results for virus isolation and RT-PCR for detection of FCV in urine or tissue culture medium stored at 4C or —7OC. Log of FCV Dilution starting titer 2.4 x 108 TCID5o/ml) Sample Negative 10“ 10'5 1045 10'7 10‘8 10'9 Control Stored at 4C Virus Isolation Urine 2/2 2/2 2/2 1/2 neg neg neg TCM 2/2 2/2 2/2 2/2 neg neg neg RT-PCR Urine 3/3 3/3 3/3 neg neg neg neg TCM 3/3 3/3 3/3 1/3 neg neg neg Storage at -7OC Virus Isolation Urine neg neg neg neg neg neg neg TCM 2/2 2/2 2/2 1/2 neg neg neg RT-PCR Urine 3/3 3/3 neg neg neg neg neg TCM 3/3 3/3 3/3 neg neg neg neg FCV= feline calicivirus; RT-PCR= reverse-transcriptase polymerase chain reaction; TCM= tissue culture medium Table 7. Final lower detection limits (expressed as TCleo/sample) for detection of FCV in urine by virus isolation or RT-PCR after storage at 4C and —70C. . . Vol. Used RT- . Dilution Final Lower for Vl or PCR . . . . FCV Conc. . . . Detection Limit Detection Extraction Dilution Sample Dilution (TCleo/ml) (ml) 1= actor (TCID5o/sample) Stored at 4C Virus Isolation Urine 1 x10'7 24 0.2 NA 5 TCM 1 x10'7 24 0.2 NA 5 RT-PCR Urine 1 x10'6 240 0.14 0.083 3 TCM 1 x10'6 240 0.14 0.083 3 Stored at -70C Virus Isolation Urine >1 x 104 >24 x 104 0.2 NA >4,800 TCM 1 x10"7 24 0.2 NA 5 RT-PCR Urine 1 x 10'5 2.4 x 103 0.14 0.083 28 TCM 1 x10'6 240 0.14 0.083 3 FCV= feline calicivirus; NA: not applicable; RT-PCR= reverse-transcriptase polymerase chain reaction; TCM= tissue culture medium; VI= virus isolation 101 APPENDIX B Results of cDNA Sequencing of the p30 Gene of 8 FCV isolates Table 8. Descriptions of FCV strains used for genotypic comparisons of cDNA and amino acid seflences of the FCV p30 gene. Case Strain No.‘ Date Location Biotype GenBankb Ref MSU 2.2 1459461 1994 USA URTD NA Scansen 04 MSU 3 1467733 1994 USA URTD NA Scansen 04 MSU 4 1524133 1995 USA URTD NA Scansen 04 F CV-U1 1903867 1998 USA ILUTD NA Rice 02 FCV-U2 2089726 1999 USA ILUTD NA Rice 02 MSU 7 2611408 2002 USA URTD NA Scansen 04 FCV-R 2351774 2000 USA URTD NA Scansen 04 MSU 9 1027025 1990 USA URTD NA Scansen 04 CFI/68 NA 1960 USA URTD U13992 51:11 3% F4 NA 1971 Japan URTD 031836 Oshikamo 94 F65 NA 1990 UK OD-Arthrop AF109465 Glenn 99 F9 NA 1958 USA URTD M86379 Carter 92 FCV2024 NA Germany AF479590 Thumfart 02 Urbana NA 15%: USA URTD-OD L40021 Sosnovtsev 95 Arthrop= arthropathy; ILUTD= idiopathic lower urinary tract disease; NA= not applicable; NR= not reported; OD=oraI disease; URTD= upper respiratory tract disease a. Identification number assigned by the Michigan State University Diagnostic Center for Population and Animal Health. b. GenBank Accession Number 102 Table 9. Results of bidirectional automated sequencing of a 90 base pair portion of the p30 (3A-Like) gene of ORF1 of 8 FCV isolates obtained from the Michigan State University Diagnostic Center for Population and Animal Health. The cDNA sequences correspond to nucleotides 2433 to 2522 and to amino acids 806 to 834 of the Urbana reference strain. gga cag ttt caa gca etc a Isolate cDNA Sequences Deduced Amino Acid Sequences at caa cat gtg gta acc gtt aac MSU tcg gta ttt gat ttg gcc tgg gct ‘ ,, ,1.” 2.2 ctt cgc cga cat ctg aca cta act QH LrgéFgflAU-{RH ggg cag ttt caa gca etc a at cag cat gtg gta acc gtt aat tcg gtg ttt gat ttg gcc tgg gct , n 'T" MSU 3 ctt cgt cgt cac ctt aca ctg gca QH LrLs’Qggé‘SXX/A‘LRRH gga cag ttc caa gct etc a at caa cat gtg gta acc gtt aat tcg gtg ttt gat ttg gcc tgg gct MSU 4 ctt cgc cgc cac ctt tca cta act QHVVYgaggééfiALRRH gga cag ttc caa gca atc a at caa cat gtg gta acc gtt aac FCV- tcg gtg ttt gat ttg gcc tgg gct , n '1'" U1 ctt cgc cgt cac cta tcg cta act QH LgfiggééxALRRH gga cag ttt caa gca ate a at caa cat gtg gta acc gtt aat FCV- tcg gtg ttt gat ttg gcc tgg gct , n 'T‘ , U2 ctt cgc cgc cac ctt acg ctg gca QH L-FLiggé-SXXALRRH ggg cag ttt caa gcc etc a at caa cat gtg gta acc gtt aat tcg gtg ttt gat ttg gcc tgg gct , n 'T" MSU 7 ctt cgc cga cac ctc acg cta QH LyééFgégXYALRRH aca gga cag ttt caa gca etc a at caa cat gtg gta acc gtt aat _ tcg gtg ttt gat ttg gcc tgg gct , n 'T" FCV R ctt cgc cgt cac ctc tca cta act QH Lgfiggé‘ngLRRH gga cag ttc caa gct atc a at caa cat gtg gta acc gtt aac tcg gtg ttt gat ttg gcc tgg gct , n 'T" MSU 9 ctt cgt cgc cac ctg tca cta act QH ngrggé‘é‘fiALRRl-i 103 Table 10. Percent nucleotide identity between the FCV p30 gene CDNA sequences of 8 FCV isolates obtained from the Michigan State University Diagnostic Center for Population and Animal Health and 6 selected FCV reference strains (see Table 8 for isolate descriptions). Comparisons were made using a 90 base pair region of the FCV p30 gene corresponding to nucleotides 2433-2522 of the Urbana reference strain. Percent cDNA Identity . MSU MSU MSU FCV FCV MSU FCV MSU CFI FCV Stra'“ 2.2 3 4 01 02 7 R 9 68 F4 F9 F65 2024 ”'9 MSU 2.2 MSU 3 85 ”f” 91 9o FCV U, 91 86 95 FCV 02 88 93 89 89 M?” 93 9o 93 95 94 F9 90 91 97 95 87 93 ”‘3” 94 88 96 96 87 91 94 CF' 89 9o 93 89 87 90 93 91 68 F4 88 84 89 9o 88 93 99 87 87 1:9 87 91 96 94 91 95 96 94 86 83 F65 93 91 96 94 91 95 96 94 94 94 88 FCV 87 88 87 84 93 9o 86 85 85 85 87 89 2024 um 87 95 87 89 98 93 87 87 89 87 94 91 91 URB= Urbana 104 Table 11. Percent amino acid identity between the FCV p30 protein sequences of 8 FCV isolates obtained from the Michigan State University Diagnostic Center for Population and Animal Health and 6 selected FCV reference strains (see Table 8 for isolate descriptions). Comparisons were made using a 29 amino acid region of the FCV p30 protein corresponding to amino acids 806-834 of the Urbana reference strain. Percent Amino Acid Identity MSU MSU MSU FCV FCV MSU FCV MSU CFI FCV St'a‘" 2.2 3 4 U1 U2 7 R 9 68 F4 F9 F65 2024 ”'9 MSU 2.2 MSU 3 97 ”f” 97 93 FCV U, 97 93 100 FCV U2 97 100 93 93 M?” 100 97 97 97 97 F3 97 93 106 100 93 97 MS” 97 93 100 100 93 97 100 CF' 97 93 93 93 93 97 93 93 68 F4 93 90 9o 90 90 93 9o 90 90 F9 93 97 9o 90 97 93 9o 90 9o 93 F65 100 97 97 97 97 100 97 97 97 93 93 FCV 9o 93 86 86 93 9o 86 86 86 83 9o 90 2024 um 97 100 93 93 100 97 93 93 93 9o 97 97 93 Urb= Urbana 105 LIST OF REFERENCES 106 REFERENCES AI-Soud WA, LJ J0nsson, P Radstrtim. 2000. Identification and characterization of immunoglobulin G in blood as a major inhibitor of diagnostic PCR. Joumal of Clinical Microbiology, 38(1):345-350. AI-Soud WA and P Radstrom. 2001. Purification and characterization of PCR- inhibitory components in blood cells. Journal of Clinical Microbiology, 39(2):485-493. August JR. 1984. Feline viral respiratory disease: the carrier state, vaccination, and control. Veterinary Clinics of North America: Small Animal Practice; 14(6):1159-1171. Baulch-Brown C, DN Love, J Meanger. 1997. Feline calicivirus: a need for a vaccine modification? Australian Veterinary Journal; 75:209-213. Baulch-Brown C, D Love, J Meanger. 1999. Sequence variation within the capsid protein of Australian isolates of feline calicivirus. Veterinary Microbiology, 68: 1 07-1 1 7. Behzadbehbahani A, PE Klapper, PJ Vallely, GM Cleator. 1997. Detection of BK virus in urine by polymerase chain reaction: a comparison of DNA extraction methods. Joumal of Virological Methods; 67:161-166. Bidawid S, N Malik, O Adegbunrin, et al. 2003. A feline kidney cell line-based plaque assay for feline calicivirus, a surrogate for Norwalk virus. Journal of Virological Methods; 107:163-167. Biel SS, TK Held, 0 Landt, et al. 2000. Rapid quantification and differentiation of human polyomavirus DNA in undiluted urine from patients after bone marrow transplantation. Journal of Clinical Microbiology, 38(10):3689- 3695. Binns SH, S Dawson, AJ Speakman, et al. 2000. A study of feline upper respiratory tract disease with reference to prevalence and risk factors for infection with feline calicivirus and feline herpesvirus. Journal of Feline Medicine and Surgery, 22123-133. Bittle JL, CJ York, JW Newberne, M Martin. 1960. Serologic relationship of new feline cytopathogenic viruses. American Journal of Veterinary Research; 21:547-550. Bittle JL and WJ Rubric. 1976. Immunisation against feline calicivirus infection. American Journal of Veterinary Research; 37:275-278. 107 BCichen-Osmond C. (Ed). 2003. 00.012. Caliciviridae. In: lCTVdB - The Universal Virus Database, version 3. lCTVdB Management, The Earth Institute, Biosphere 2 Center, Columbia University, Oracle, AZ, USA Buffington CAT, DJ Chew, MS Kendall, et al. 1997. Clinical evaluation of cats with nonobstructive urinary tract diseases. Journal of the American Veterinary Medical Association; 210(1):46-50. Biirki F. 1965. Picomaviruses of cats. Archiv fl'ir die gesamte Virusforschung; 15:690-696. Byrnes JJ, KM Downey, L Esserman, AG So. 1975. Mechanism of hemin inhibition of erythroid cytoplasmic DNA polymerase. Biochemistry, 14(4):796-799. Carter MJ, ID Milton, J Meanger, et al. 1992. The complete nucleotide sequence of a feline calicivirus. Virology, 190:443-448. Castro AE. 1992. Isolation and identification of viruses. In: Veterinary Diagnostic Virology: A Practitioner’s Guide (AE Castro & WP Heuschele, eds.). Mosby - Year Book, Inc., St. Louis, MO; 3-8. CDC. 2004. CDC technical fact sheet about noroviruses. From, http://www.cdc.gov/ncidod/dvrd/revb/gastro/norovirus-factsheet.htm. Chang D, M Wang, WC Ou, et al. 1996. A simple method for detecting human polyomavirus DNA in urine by the polymerase chain reaction. Journal of Virological Methods; 58:1 31 -1 36. Chasey D and P Duff. 1990. European brown hare syndrome and associated virus particles in the UK. Veterinary Record; 126:623-624. Chiba S, Y Sakuma, R Koasaka, et al. 1979. An outbreak of gastroenteritis associated with calicivirus in an infant home. Journal of Medical Virology, 4:249-254. Chiba S, S Nakata, K Numata-Kinoshita, et al. 2000. Sapporo virus: History and recent findings. Journal of Infectious Diseases; 18(Suppl 2)28303-308. Chiodi F, B Keys, JAlbert, etal. 1992. Human immunodeficiency virus type-1 is present in the cerebrospinal fluid of a majority of infected individuals. Journal of Clinical Microbiology. 30(7): 1768- 1771. Crandell RA and SH Madin. 1960. Experimental studies on a new feline virus. American Journal of Veterinary Research; 21 :551-556. 108 Dawson S, D Bennett, SD Carter, et al. 1994. Acute arthritis of cats associated with feline calicivirus infection. Research in Veterinary Science; 56:133. Demmler GJ, GJ Buffone, CM Schimbor, RA May. 1988. Detection of cytomegalovirus in urine from newborns by using polymerase chain reaction DNA amplification. The Journal of Infectious Diseases; 158(6):1177-1184. Dick CP, RP Johnson, S Yamashiro. 1989. Sites of persistence of feline calicivirus. Research in Veterinary Science; 47(3): 367-373. Doyle SA, MD Sussman, JM Kruger, RK Mass. 1996. A reverse-transcriptase polymerase chain reaction assay for detection of feline caliciviruses. Abstract. Proceedings of the 7‘" Annual Phi Zeta Research Day, East Lansing, MI. Dynabeads® mRNA DIRECTTM Kit Handbook. 2000. Dynal Incorporated, Lake Success, New York; 10,14-21. Echavarria M, M Forrnan, J Ticehurst, et al. 1998. PCR method for detection of adenovirus in urine of healthy and human immunodeficiency virus-infected individuals. Journal of Clinical Microbiology, 36(11):3323-3326. Ellis TM. 1981. Jaundice in a Siamese cat with in utero feline calicivirus infection. Australian Veterinary Journal; 57:383-385. Fabricant CG. 1973. Urolithiasis: a review with recent viral studies. Feline Practice; 3:22-30. Fabricant CG. 1977. Herpesvirus-induced urolithiasis in specific-pathogen-free male cats. American Journal of Veterinary Research; 38:1837-1842. Fabricant CG. 1984. The feline urologic syndrome induced by infection with a cell-associated herpesvirus. Veterinary Clinics of North America: Small Animal Practice; 14(3):493-502. Fastier, LB. 1957. A new feline virus isolated in tissue culture. American Journal of Veterinary Research; 18:382-389. Fenner FJ, EPJ Gibbs, FA Murphy, et al. 1993. Caliciviridae. In: Veterinary Virology, 2"d Edition. Academic Press, San Diego, California; 425-430. Forbes BA & KE Hicks. 1996. Substances interfering with direct detection of Mycobacterium tuberculosis in clinical specimens by PCR: effects of bovine serum albumin. Journal of Clinical Microbiology, 34(9):2125-2128. 109 Gaskell R and S Dawson. 1998. Feline respiratory disease. In: Infectious Diseases of the Dog and Cat (CE Greene, ed.), 2nd edition. W.B. Saunders; Philadelphia, PA; 97-106. Geissler K, K Schneider, G Platzer, et al. 1997. Genetic and antigenic heterogeneity among feline calicivirus isolates from distinct disease manifestations. Virus Research; 48:193-206. Geissler K, K Schneider, A Fleuchaus, et al. 1999. Feline calicivirus capsid protein expression and capsid assembly in cultured feline cells. Journal of Virology, 73(1):834-838. Gillespie JH, AB Judkins, DE Kahn. 1971. Feline viruses XIII. The use of immunofluorescent test for the detection of feline picomaviruses. The Cornell Veterinarian; 61:172. Gillespie JH and FW Scott. 1973. Feline viral infections. Advances in Veterinary Science and Comparative Medicine; 17:163-200. Glenn M, AD Radford, PC Turner, et al. 1999. Nucleotide sequence of UK and Australian isolates of feline calicivirus (FCV) and phylogenetic analysis of FCVs. Veterinary Microbiology, 67:175—193. Green KY, T Ano, MS Balayan, et al. 2000. Taxonomy of the caliciviruses. Journal of Infectious Diseases; 181 (Suppl 2):S322-330. Green KY, A Mory, MH Fogg. et al. 2002. Isolation of enzymatically active replication complexes from feline calicivirus-infected cells. Journal of Virology, 76:8582-8595. Greenberg HB, JR Valdesuso, AR Kalica, et al. 1981. Proteins of Nonlvalk virus. Journal of Virology, 37:994-999. Hale AD, J Green, DWG Brown. 1996. Comparison of four RNA extraction methods for the detection of small round structured viruses in faecal specimens. Journal of Virological Methods; 57:195-201. Harbour DA, PE Howard, RM Gaskell. 1991. Isolation of feline calicivirus and feline herpesvirus from domestic cats 1980 to 1989. Veterinary Record; 128:77-80. Helps C, P Lait, S Tasker, D Harbour. 2002. Melting curve analysis of feline calicivirus isolates detected by real-time reverse transcription PCR. Journal of Virological Methods; 106:241-244. 110 Herbert TP, l Brierley, TD Brown. 1997. Identification of a protein linked to the genomic and subgenomic mRNAs of feline calicivirus and its role in translation. Joumal of General Virology, 78:1033-1040. Horimoto T, Y Takeda, K lwatsuki-Horimoto, et al. 2001. Capsid protein gene variation among feline calicivirus isolates. Virus Genes; 23:171-174. Hurley KF, PA Pesavento, NC Pedersen, et al. 2004. An outbreak of virulent systemic feline calicivirus disease. Joumal of the American Veterinary Medical Association; 224(2):241-249. Jiang X, DY Graham, K Wang, et al. 1990. Norwalk virus genome: cloning and characterization. Science; 250:1 580-1583. Kahn DE and JH Gillespie. 1971. Feline viruses: pathogenesis of picomavirus infection in the cat. American Journal of Veterinary Research; 32:521- 531. Kahn DE and EA Hoover. 1976a. Infectious respiratory diseases of cats. Veterinary Clinics of North America: Small Animal Practice; 6(3):399-413. Kahn DE and EA Hoover. 1976b. Feline caliciviral disease: experimental immunoprophylaxis. American Journal of Veterinary Research; 37:279- 283. Khan G, HO Kangro, PJ Coates, RB Heath. 1991. Inhibitory effects of urine on the polymerase chain reaction for cytomegalovirus DNA. Journal of Clinical Pathology, 44:360-365. Kalkstein TS, JM Kruger, CA Osborne. 1999a. Feline idiopathic lower urinary tract disease. Part I: clinical manifestations. The Compendium on Continuing Education for the Practicing Veterinarian; 21 (1 ):1 5-26. Kalkstein TS, JM Kruger, CA Osborne. 1999b. Feline idiopathic lower urinary tract disease. Part II: potential causes. The Compendium on Continuing Education for the Practicing Veterinarian; 21 (2):148-1 54. Kalunda M, KM Lee, DF Holmes, JH Gillespie. Serologic classification of feline caliciviruses by plaque-reduction neutralization and immunodiffusion. American Journal of Veterinary Research; 36:353-356. Kapikian AZ, RG Wyatt, R Dolin, et al. 1972. Visualization by immune electron microscopy of a 27-nm particle associated with acute infectious nonbacterial gastroenteritis. Journal of Virology, 10(5):1075-1081. 111 Kapikian AZ. 2000. The discovery of the 27-nm Nonrvalk virus: An historic perspective. Journal of Infectious Diseases; 18(Suppl 2):S295-302. Kingsley DH and GP Richards. 2001. Rapid and efficient extraction method for reverse transcription-PCR detection of Hepatitis A and Norwalk-like viruses in shellfish. Applied and Environmental Microbiology, 67(9):4152- 4157. Kirk H. 1925. Urino-genital diseases. In: The Diseases of the Cat. Eger, Chicago, Illinois; 261-273. Klein A, R Barsuk, S Dagan, et al. 1997. Comparison of methods for extraction of nucleic acid from hemolytic serum for PCR amplification of hepatitis B virus DNA sequences. Journal of Clinical Microbiology, 35(7):1897-1899. Komolate 00. 1976. Effect of storage on the integrity of purified feline calicivirus particles. Microbios; 26:137-146. Kruger JM and CA Osborne. 1990. The role of viruses in feline lower urinary tract disease. Journal of Veterinary lntemal Medicine; 4(2):71-78. Kruger JM, CA Osborne, SM Goyal, et al. 1991. Clinical evaluation of cats with lower urinary tract disease. Journal of the American Veterinary Medical Association; 199(2):211-216. Kruger JM, CA Osborne, PJ Venta, MD Sussman. 1996. Viral infections of the feline urinary tract. Veterinary Clinics of North America: Small Animal Practice; 26(2):281-296. Kruger JM and SA Doyle. 1997. Unpublished observations. Kruger JM, BA Scansen, RK Maes. 2002. A pilot study of feline calicivirus- induced urinary tract infection in SPF cats. Abstract. In Proceedings: 12‘" Annual Phi Zeta Research Symposium, East Lansing, MI. p. 31. Lanciotti RS, AJ Kerst, RS Nasci, et al. 2000. Rapid detection of West Nile virus from human clinical specimens, field-collected mosquitoes, and avian samples by a TaqMan reverse transcriptase-PCR assay. Journal of Clinical Microbiology; 38:4066-4071. Leutenegger CM, D Klein, R Hofmann-Lehmann, et al. 1999. Rapid feline immunodeficiency virus provirus quantitation by polymerase chain reaction using the TaqMan fluorogenic real-time detection system. Joumal of Virological Methods; 78: 1 05-1 16. 112 Levere RD, Y Gong, A Kappas, et al. 1991. Heme inhibits human immunodeficiency virus 1 replication in cell cultures and enhances the antiviral effect of zidovudine. Proceedings of the National Academy of Sciences; 88:1756-1759. Levy JK, A Marsh. 1992. Isolation of calicivirus from the joint of a kitten with arthritis. Journal of the American Veterinary Medical Association; 201 (5):753-755. Liu BL, PR Lambden, H GiJnther, et al. 1999. Molecular characterization of a bovine enteric calicivirus: relationship to Norwalk-like viruses. Journal of Virology, 73(1 ):819-825. Livak KJ, SJ Flood, J Marmaro, et al. 1995. Oligonucleotides with fluorescent dyes at opposite ends provide a quenched probe system useful for detecting PCR product and nucleic acid hybridization. PCR Methods and Applications; 42357-362. Loisy F, P Le Carin, M Pommepuy, F Le Guyader. 2000. An improved method for the detection of Norwalk-like caliciviruses in environmental samples. Letters in Applied Microbiology; 31 :41 1-415. Mackay IM, KE Arden, A Nitsche. 2002. Real-time PCR in virology. Nucleic Acids Research; 30: 1 292-1 305. Madeley CR and BP Cosgrove. 1976. Caliciviruses in man. Lancet; 1:199-200. Mahony J, S Chong, D Jang, et al. 1998. Urine specimens from pregnant and nonpregnant women inhibitory to amplification of Chlamydia trachomatis nucleic acid by PCR, ligase chain reaction, and transcription-mediated amplification: identification of urinary substances associated with inhibition and removal of inhibitory activity. Journal of Clinical Microbiology, 36(11):3122-3126. Meijerink J, C Mandigers, L van de Locht, et al. 2001. A novel method to compensate for different amplification efficiencies between patient DNA samples in quantitative real-time PCR. Journal of Molecular Diagnostics; 3:55-61. Morata P, MI Queipo-Ortufio, J de Dios Colmenero. 1998. Strategy for optimizing DNA amplification in a peripheral blood PCR assay used for diagnosis of human brucellosis. Journal of Clinical Microbiology, 26(9):2443-2446. 113 Morrison TB, JJ Weis, CT Wittwer. 1998. Quantification of low-copy transcripts by continuous SYBR® Green I monitoring during amplification. Bio Techniques; 24:954-962. Murphy FA, EPJ Gibbs, MC Horzinek, et al. 1999. Caliciviridae. In: Veterinary Virology, 3rd Edition. Academic Press, San Diego, California; 533-541. Myremel M, E Rimstad, Y Wasteson. 2000. lmmunomagnetic separation of a Norwalk-like virus (genogroup 1) in artificially contaminated environmental water samples. lntemational Journal of Food Microbiology, 62:17-26. Nakata S, S Chiba, H Terashima, et al. 1985. Humoral immunity in infants with gastroenteritis caused by human calicivirus. Journal of Infectious Diseases; 152(2):274-279. Neill JD and WL Mengeling. 1988. Further characterization of the virus-specific RNAs in feline calicivirus infected cells. Virus Research; 11:59-72. Neill JD. 1990. Nucleotide sequence of a region of the feline calicivirus genome which encodes picomavirus-like RNA-dependent RNA polymerase, cysteine protease and 2C polypeptides. Virus Research; 17:145-160. Neill JD. 1994. GenBank Accession No. U13392. Numata K, ME Hardy, S Nakata, et al. 1997. Molecular characterization of morphologically typical human calicivirus Sapporo. Archives of Virology, 142:1537-1552. Ohlinger VF, B Haas, G Meyers, et al. 1990. Identification and characterization of the virus causing rabbit hemorrhagic disease. Journal of Virology, 64(7):3331-3336. Osbaldiston GW and RA Taussig. 1970. Clinical report on 46 cases of feline urological syndrome. Veterinary Medicine, Small Animal Clinician; 65:461-468. Osborne CA, JM Kruger, JP Lulich, et al. 1997. Feline lower urinary tract disease: The Minnesota experience. Proceedings of the 15th ACVIM Forum; 338-339. Osborne CA and JB Stevens. 1999. Biochemical analysis of urine: indications, methods, interpretation. In: Urinalysis: A Clinical Guide to Compassionate Patient Care. Bayer Corporation, Shawnee Mission, KS; 106-109. 114 Osborne CA, JM Kruger, JP Lulich, et al. 2000. Feline lower urinary tract diseases. In: Textbook of Veterinary lntemal Medicine (SJ Ettinger & ED Feldman, eds.), 5th Edition. W.B. Saunders, Philadelphia, PA; 1710-1747. Oshikamo R, Y Tohya, Y Kawaguchi, et al. 1994. The molecular cloning and sequence of an open reading frame encoding for non-structural proteins of feline calicivirus F4 strain isolated in Japan. Journal of Veterinary Medical Science; 56(6):1093-1099. Panaccio M and A Lew. 1991. PCR based diagnosis in the presence of 8% (v/v) blood. Nucleic Acids Research; 19(5):1151. Parra F and M Prieto. 1990. Purification and characterization of a calicivirus as the causative agent of a lethal hemorrhagic disease in rabbits. Joumal of Virology, 64(8):4013-4015. Pedersen NC, L Laliberte, S Ekman. 1983. A transient febrile “limping” syndrome of kittens caused by two different strains of feline calicivirus. Feline Practice; 13(1):26-35. Pedersen NC and KF Hawkins. 1995. Mechanisms for persistence of acute and chronic feline calicivirus infections in the face of vaccination. Veterinary Microbiology, 472141 -1 56. Pedersen NC, JB Elliott, A Glasgow, et al. 2000. An isolated epizootic of hemorrhagic-like fever in cats caused by a novel and highly virulent strain of feline calicivirus. Veterinary Microbiology, 73:281-300. Povey RC, RC Wardley, H Jessen. 1973. Feline picomavirus infection: the in vivo carrier state. The Veterinary Record; 92:224-229. Prasad BV, DO Matson, AW Smith. 1994. Three—dimensional structure of calicivirus. Journal of Molecular Biology, 240:256-264. QIAamp Viral RNA Mini Kit Handbook. 1999. Qiagen Incorporated, Valencia, California; 18-19,25. Radford AD, M Bennett, F McArdle, et al. 1997. The use of sequence analysis of a feline calicivirus (FCV) hypervariable region in the epidemiological investigation of FCV related disease and vaccine failures. Vaccine; 15(2):1451-1458. Radford AD, PC Turner, M Bennett, et al. 1998. Quasispecies evolution of a hypervariable region of the feline calicivirus capsid gene in cell culture and in persistently infected cats. Journal of General Virology, 7921-10. 115 Radford AD, S Dawson, C Wharrnby, et al. 2000. Comparison of serological and sequence-based methods for typing feline calicivirus isolates from vaccine failures. Veterinary Record; 146:117-123. Radford AD, LM Sommerville, S Dawson, et al. 2001. Molecular analysis of isolates of feline calicivirus from a population of cats in a rescue shelter. Veterinary Record; 149:477-481. Rice CC, JM Kruger, PJ Venta, et al. 2002. Genetic characterization of 2 novel feline caliciviruses isolated from cats with idiopathic lower urinary tract disease. Journal of Veterinary lntemal Medicine; 16:293-302. Rich LJ and CG Fabricant. 1969. Urethral obstruction in male cats: transmission studies. Canadian Journal of Comparative Medicine; 33:164-165. Rich LJ, CG Fabricant, JH Gillespie. 1971. Virus induced urolithiasis in male cats. The Cornell Veterinarian; 61 :542-553. Ririe KM, RP Rasmussen, CT Wittwer. 1997. Product differentiation by analysis of DNA melting curves during the polymerase chain reaction. Analytical Biochemistry, 245:1 54-160. RNeasy Mini Handbook, 3rd ed. 2001. Qiagen Incorporated, Valencia, California; 30-35. SAS/STAT Users Guide: Statistics, Version 8, 4th ed. 1989. SAS Institute Incorporated, Cary, North Carolina. Scansen BA, JM Kruger, PJ Venta, et al. 2002. Development of a one-step reverse-transcriptase polymerase chain reaction assay for detection of feline calicivirus. Abstract. Journal of Veterinary lntemal Medicine; 16:365. Scansen BA, AG Wise, JM Kruger, et al. 2004. Evaluation of a p30 gene-based real-time reverse transcriptase polymerase chain reaction assay for detection of feline caliciviruses. Journal of Veterinary lntemal Medicine; 18:135-138. Schorr-Evans EM, A Poland, WE Johnson, NC Pedersen. 2003. An epizootic of highly virulent feline calicivirus disease in a hospital setting in New England. Journal of Feline Medicine and Surgery, 5:217-226. Seal BS, JF Ridpath, WL Mengeling. 1993. Analysis of feline calicivirus capsid protein genes: identification of variable antigenic determinant regions of the protein. Journal of General Virology, 74:2519-2524. 116 Seal BS. 1994. Analysis of capsid protein gene variation among divergent isolates of feline calicivirus. Virus Research; 33:39-53. Smith AW, TG Akers, SH Madin, NA Vedros. 1973. San Miguel sea lion virus isolation, preliminary characterization and relationship to vesicular exanthema of swine virus. Nature; 244:108-110. Sosnovtsev S and KY Green. 1995. RNA transcripts derived from a cloned full- length copy of the feline calicivirus genome do not require VpG for infectivity. Virology, 210:383-390. Sosnovtsev SV, SA Sosnovtseva, KY Green. 1998. Cleavage of the feline calicivirus capsid precursor is mediated by a virus-encoded proteinase. Journal of Virology, 72(4):3051-3059. Sosnovtsev S, M Garfield, KY Green. 2002. Processing map and essential cleavage sites of the nonstructural polyprotein encoded by the ORF1 of the feline calicivirus. Journal of Virology, 76:7060-7072. Studdert MJ. 1978. Caliciviruses: brief review. Archives of Virology, 58:157- 191. Sykes JE, VP Studdert, GF Browning. 1998. Detection and strain differentiation of feline calicivirus in conjunctival swabs by RT-PCR of the hypervariable region of the capsid protein gene. Archives of Virology, 143:1321-1334. Sykes JE. 2001. Feline upper respiratory tract pathogens: herpesvirus-1 and calicivirus. The Compendium on Continuing Education for the Practicing Veterinarian; 23(2): 1 66—1 75. Sykes JE, JL Allen, VP Studdert, GF Browning. 2001. Detection of feline calicivirus, feline herpesvirus 1 and Chalmydia psittaci mucosal swabs by multiplex RT-PCR/PCR. Veterinary Microbiology, 81295-108. Takahashi E, S Konishi, M Ogata. 1971. Studies on cytopathogenic viruses from cats with respiratory infections. Japanese Journal of Veterinary Science; 33:81 -87. Tenorio AP, CE Franti, BR Madewell, NC Pedersen. 1991. Chronic oral infections of cats and their relationship to persistent oral carriage of feline calici-, immunodeficiency, or leukemia viruses. Veterinary Immunology and Immunopathology, 29:1 -14. Thumfart J0 and G Meyers. 2002. Feline calicivirus: recovery of wild-type and recombinant viruses after transfection of cRNA or cDNA constructs. Journal of Virology, 76(12):6398-6407. 117 Tohya Y, N Yokoyama, K Maeda, et al. 1997. Mapping of antigenic sites involved in neutralization on the capsid protein of feline calicivirus. Journal of General Virology, 78:303-305. Toye B, W Woods, M Bobrowska, K Ramotar. 1998. Inhibition of PCR in genital and urine specimens submitted for Chlamydia trachomatis testing. Journal of Clinical Microbiology, 36(8):2356-2358. Traum J. 1936. Vesicular exanthema of swine. Joumal of the American Veterinary Medical Association; 88:316-327. Truyen U, K Geissler, J Hirschberger. 1999. Tissue distribution of virus replication in cats experimentally infected with distinct feline calicivirus isolates. Berliner und MUnchener tierarztliche Wochenschrift; 112:355- 358. Tsutsui K and GC Mueller. 1987. Hemin inhibits virion-associated reverse transcriptase of murine leukemia virus. Biochemical and Biophysical Research Communications; 149(2):628-634. Uwatoko K, M Sunairi, A Yamamoto, et al. 1996. Rapid and efficient method to eliminate substances inhibitory to the polymerase chain reaction from animal fecal samples. Veterinary Microbiology, 52:73-79. van Vuuren M, K Geissler, D Gerber, et al. 1999. Characterisation of a potentially abortigenic strain of feline calicivirus isolated from a domestic cat. Veterinary Record; 144:636-683. Verkooyen RP, A Luijendijk, WM I-Iuisman, et al. 1996. Detection of PCR inhibitors in cervical specimens by using the AMPLICOR Chlamydia trachomatis assay. Journal of Clinical Microbiology, 34(12):3072-3074. Vinogradskaya GR, lY Goryshin, YA Berlin, VA Lanzov. 1995. Optimization of PCR-based diagnostics for human cytomegalovirus. Journal of Virological Methods; 53: 1 03-1 12. Wawrzkiewicz J, CJ Smale, F Brown. 1968. Biochemical and biophysical characteristics of vesicular exanthema virus and the viral ribonucleic acid. Archiv fiir die gesamte Virusforschung; 25:337-351. Weidbrauk DL, JC Werner, AM Drevon. 1995. Inhibition of PCR by aqueous and vitreous fluids. Journal of Clinical Microbiology, 33(10):2643-2646. 118 Wittwer CT, MG Herrmann, AA Moss, RP Rasmussen. 1997. Continuous fluorescence monitoring of rapid cycle DNA amplification. BioTechniques; 22:130-138. Yamaguchi Y, T Hironaka, M Kajiwara, et al. 1992. Increased sensitivity for detection of human cytomegalovirus in urine by removal of inhibitors for the polymerase chain reaction. Journal of Virological Methods; 37:209- 218. Zwillenberg L0 and F Biil‘ki. 1966. On the capsid structure of some small feline and bovine RNA viruses. Archiv fl'Jr die gesamte Virusforschung; 19:373- 384. 119 IIiiiijjjiijjijiiiijjji