UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Interaction of viruses with human joint material Miki, Nancy Patricia Hana 1990

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


831-UBC_1990_A6_7 M54.pdf [ 11.09MB ]
JSON: 831-1.0098396.json
JSON-LD: 831-1.0098396-ld.json
RDF/XML (Pretty): 831-1.0098396-rdf.xml
RDF/JSON: 831-1.0098396-rdf.json
Turtle: 831-1.0098396-turtle.txt
N-Triples: 831-1.0098396-rdf-ntriples.txt
Original Record: 831-1.0098396-source.json
Full Text

Full Text

INTERACTION OF VIRUSES WITH HUMAN JOINT MATERIAL by NANCY PATRICIA HANA MIKI B.Sc. (Microbiology), University of British Columbia, 1986 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FORTHE DEGREE OF MASTER OF SCIENCE in THE FACULTY OF GRADUATE STUDIES Division of Medical Microbiology in the Department of Pathology We accept this thesis as conforrning to the required standard: THE UNIVERSITY OF BRITISH COLUMBIA August 1990 © Nancy Patricia Hana Miki, 1990 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of PATHOLOGY The University of British Columbia Vancouver, Canada Date August 29, 1990 DE-6 (2/88) ABSTRACT An in vitro cell and organ culture system for the study of human arthrotropic viruses was established to deter-mine the permissiveness of joint cells and tissue to five strains of rubella virus (RV), adenovirus type 2 (Ad) and human parvovirus (B19). A semi-continuous line of fetal chondrocytes (FC) and primary synovial membrane/cartilage (SMC) cells were used to investigate virus/joint cell interactions. In addition, the SMC cells were used to determine the ability of a virus to establish a persistent infection in cells of joint origin. In-tact joint tissue containing both synovial membrane and cartilage was also infected to determine the ability of the viruses to replicate in non-dividing cells and also their invasive capability in the presence of an extracel-lular matrix. Results showed that all RV strains and Ad were able to replicate in FC, SMC cells and intact joint tissue. The only exception to this was that replication of RV strain Cendehill was not detected in joint tissue. In contrast, there was no indication, by either DNA-DNA hybridization or immunoperoxidase staining, that B19 was able to replicate in SMC cells. Most importantly, the behaviour of the rubella strains in this model was found to correlate with the clinical data concerning the incidence of complications associated with rubella infection or vaccination. This suggests that arthritogenicity is related to the ability of a particular virus to infect and per-sist in joint tissue and establishes the in vitro model as a useful system to evaluate the arthrotropicity of fu-ture rubella vaccines. in TABLE OF CONTENTS Abstract ii List of figures vi Acknowledgements viii I. INTRODUCTION 1 A Arthritis Overview 1 B. Viral Arthritis 1 1. Mechanisms of induction of arthritis by viruses 2 2. Animal models of arthritis 4 C. Arthrotropic Viruses 5 1. Rubella virus 5 2. Adenovirus 9 3. Parvovirus 11 4. Others 15 D. Purpose of Study 16 II. ABBREVIATIONS 18 III. MATERIALS 21 A Materials 21 B. Solutions 22 IV. METHODS 26 A Cell Maintenance 26 B. Synovial Cell Production 26 C. Preparation of Joint Tissue 27 D. Virus Stock Production 27 1. Rubella virus 27 iv 2. Adenovirus 27 E. Virus Titration 27 F. Cell and Tissue Fixation 29 G. Viral Growth Curves 29 H. Persistent Infections 30 I. Immunoperoxidase GP) Staining 30 1. Cells 30 2. Tissue sections 31 J. Preparation and Purification of Anti-rubella Antisera 32 K. Probe Preparation and Labelling 32 L. DNA Extraction 33 V. RESULTS AND DISCUSSION 34 A Rubella Virus 34 1. Single time-point infections 34 a. Antigen expression 34 b. Virus production 41 2. Viral growth curves 41 3. Persistence in synovial membrane/cartilage cells 51 a. Virus production 51 b. Antigen expression 51 4. Infection of joint tissue 66 a. Virus production 66 b. Antigen expression 66 V 5. Summary and discussion 73 B. Adenovirus 83 1. Viral growth curves 88 a. Antigen expression 88 b. Virus production 88 2. Persistence in synovial membrane/cartilage cells 97 3. Infection of joint tissue 97 4. Summary and discussion 97 C. Human Parvovirus B19 105 1. Virus growth 105 2. Antigen expression 108 3. Summary and discussion 111 VI. CONCLUSION 112 VII. REFERENCES 114 VI LIST OF FIGURES Figure 1. Antigen Expression of Rubella Virus Strains in Vero Cells 36 2. Antigen Expression of Rubella Virus Strains in Fetal Chondrocytes 38 3. Antigen Expression of Rubella Virus Strains in SMC Cells 40 4. Growth Curves of Supernatant Rubella Virus Strains in SMC and Vero Cells 45 5. Decay Kinetics for Rubella Virus Strains Cendehill and RA27/3 48 6. Growth Curves of Cell-Associated Rubella Virus Strains in SMC and Vero Cells 50 7. Rubella Virus Strain Comparison in SMC Cells 53 8. Persistent Infection of SMC Cells with Rubella Virus Strains 55 9. SMC Cell Persistently Infected with Rubella Virus Strain Cendehill 57 10. SMC Cell Persistently Infected with Rubella Virus Strain RA27/3 59 11. SMC Cell Persistently Infected with Rubella Virus Strain HPV77/DE5 61 12. SMC Cell Persistently Infected with Rubella Virus Strain Therien 63 13. SMC Cell Persistently Infected with Rubella Virus Strain M33 65 14. Joint Tissue Infected with Rubella Virus Strains 68 15. Uninfected Joint Tissue 70 16. Infection of Joint Tissue with Rubella Strain RA27/3 72 17. Infection of Joint Tissue with Rubella Strain HPV77/DE5 75 18. Infection of Joint Tissue with Rubella Strain Therien 77 19. Infection of Joint Tissue with Rubella Strain M33 79 vii 20. Penetration of Rubella Virus Strain Therien into Cartilage Tissue 81 21. Adenovirus Infected Fetal Chondrocytes 85 22. Adenovirus Infected SMC Cells 87 23. Progression of Adenovirus Antigen Expression in Hep-2 Cells 90 24. Progression of Adenovirus Antigen Expression in SMC Cells 92 25. Growth Curves of Adenovirus in Hep-2 and SMC Cells 94 26. Decay Kinetics of Adenovirus 96 27. SMC Cells Persistently Infected with Adenovirus 99 28. Persistent Infection of SMC Cells with Adenovirus 101 29. Joint Tissue Infected with Adenovirus 103 30. Co-cultivation of Adenovirus Infected Joint Tissue with Hep-2 Cells 107 31. Detection of B19 DNA from Infected SMC and BM Cells 110 ACKNOWLEDGMENTS My gratitude to my family is impossible to articulate; not only are they responsible for encouraging (forcing?) me towards university, but in addition, they have provided tremendous support (of all kinds) throughout However, they are no doubt relieved to see the conclusion. Heart-felt thanks also go to my supervisor, Jaki, for more than just providing a place to hang my labcoat. In her laboratory I learned a great deal from the scientific method to self-control. And in between I also learned the finer points of packing and moving an entire laboratory. Her laboratory provided an atmosphere of shar-ing, trust and freedom all of which were respected greatly and defended fiercely. Part and parcel of these positive aspects are GeeTee and KayEll, two of the best people to share a lab with, not just because they're lit-tle, but because they're cool. Also to JayBee for always being so willing to drop everything and come to my aid. I tried the patience of my committee members (Drs. RTA MacGillivray, JR O'Kusky, PE Reid and FTufaro) on various occasions but they were either ignoring me or being very tolerant; I strongly suspect the latter since they were always ready with assistance and suggestions. Thanks are also given to Drs.D Owen and N Quenville from Anatomical Pathology at Vancouver General Hospital for providing tissue specimens. Without the assistance of Jennifer Leung and Sam Trepanier from the same department, this work still would have been done (probably), but I shudder to think how long it would have taken. As rewarding as research can be, there is still the very practical side to life and I thank the Heighway Fund for supporting me. Finally, to APWS, inexpressible thanks for taking all the *&%$#@ and making me look so good. 1 I. INTRODUCTION A. Arthritis Overview Rheumatic diseases (disorders involving connective tissue) form a broad spectrum of illnesses which vary in duration and severity. The most prevalent symptom of the rheumatic diseases is arthritis which is simply defined as inflammation of a joint (46). The joint inflammation may be present on its own, affecting one or many joints, or be part of a more complex syndrome. In addition, the symptoms may be acute without se-quelae or chronic and degenerative resulting in severe deformities. In its various manifestations, arthritis at-tacks one in every seven Canadians, regardless of age (15). Because of its prevalence, a great deal of research has been committed towards the therapy, pathogenesis and of course, the cause of arthritis. Putative causes of arthritis which have been investigated include en-vironment, genetics, immunological responses and diet (15). The potential involvement of infectious agents as the cause or trigger of arthritis has also received a great deal of attention. The latter is, on its own, an enor-mous field of research including the study of bacteria, fungi and viruses (74,117,126). As the research con-tained within this thesis is concerning the latter, the remainder of the introduction will be focused on the role of viruses in inflammatory joint disease. B. Viral Arthritis There are a variety of different viral infections in humans and animals that are accompanied by arthritis (43, 74,82,115,117,131). The incidence of joint involvement may be infrequent (mumps virus, herpesviruses, Coxsackie B virus, adenovirus, and enteroviruses) or common (rubella virus, human parvovirus, and hepatitis B virus). In general the arthritis is acute and without sequelae (115) but there are notable excep-tions in which the joint inflammation has become chronic or recurrent, lasting for several years (3,4, 111). There is ample evidence in the literature to indicate that viruses do play a role in the induction of arthritis but what is unfortunately lacking at this time is information regarding the pathogenic mechanisms involved. However, consideration of virus-host interactions leads to a number of different possibilities. 2 1. Mechanisms of induction of arthritis by viruses a. Direct tissue invasion The most likely mode of access to the joint for any virus is via circulating lymphocytes during the viremic stage of the infection (43). Once within the joint, the most direct way for a virus to cause acute arthritis would be for it to infect and damage cells in the synovial lining causing inflammation as the host attempts to clear the infection and repair the damage (108,131). However, depending on the nature of the virus, or if the host is unsuccessful at clearing the infection, the virus may set up a persistent infection (107). The result of a persistent infection within a joint may be to provoke a localized and chronic inflammatory response through the expression of foreign antigens or the deposition of immune complexes (131). b. Infection at an extra-articular site Infection may also be at an extra-articular site such as in lymphoreticular cells. This would provide a virus with the opportunity to re-infect a joint and thus account for the recurrent nature of some viral arthritides. This latter possibility is examined in the following section concerning rubella virus. Alternatively, infection of another organ, such as the liver with hepatitis B virus, can give rise to immune complexes which may then settle in joints and cause inflammation (4,131). c. Autoimmunity Examples of autoimmune diseases with a suspected viral etiology include diabetes mellitus, multiple sclerosis, rheumatoid arthritis (RA), and systemic lupus erythematosus (SLE) (120). Two mechanisms through which a virus might induce autoimmunity are polyclonal B-cell activation and molecular mimicry (81,120,171). It has been demonstrated that humans and animals normally possess low levels of autoantibodies (an-tibodies directed against self-antigens) in their sera (44,87). During viral infection these antibodies often in-crease but usually subside when the infection is cleared (54,73,162). Ahmed and Oldstone (119) proposed the following as possible mechanisms of B-cell activation. Several viruses have demonstrated mitogenic activity upon B-cells that only requires adherence to the cell surface such as vesicular stomatitis virus (65), adenovirus (62), influenza A virus (23), Sindbis virus (64) and African swine fever virus (162). Herpes simplex virus and Epstein-Barr virus may also act as B-cell mitogens, but in-fection of the B-cell is required (137). Alternatively, a virus may lead to increased immunoglobulin synthesis by inactivating the suppressor T-cells responsible for switching off their synthesis. Finally, from studies with lymphocytic choriomeningitis virus in mice, the authors suggested that virus-specific helper T-cells may lead to B-cell activation by releasing non-specific B-cell growth and differentiation factors (BCGF and BCDF). They based this upon several observations. Many viruses induce strong T-cell responses that result in the for-mation of virus specific helper T-cells. These latter cells secrete non-specific B-cell stimulators which in turn affect primed B-cells (but not resting B-cells). Therefore B-cells with a variety of specificities may be stimu-lated including those which react with self-antigens. This in itself can cause direct damage to cells and tis-sues, especially in the event of a persistent infection. In addition, disease may be mediated by immune complex formation and deposition. Autoimmunity may also be a result of antigenic similarity between an infectious agent and host epitopes (2, 81,120,137). Examples of such cross-reactivity have been found with bacteria (41,49,137,171), protozoa (137) and viruses (59,66,85,136,137) and several of these have implications in rheumatic diseases. The occurrence of HLA-B27 in association with reactive arthritis and ankylosing spondylitis is extremely high (approximately 75% in the former and 90% in the latter) (171). Cross-reactivity between this MHC class I molecule and both Yersinia entercoliHca and Klebsiella pneumoniae have been documented (171). Also anti-DNA antibodies from mice and humans with SLE have been shown to bind to mycobacteria (137). The involvement of Epstein-Barr virus 03BV) in the development of RA through molecular mimicry has been reviewed (59). Patients with RA have decreased T-cell suppression of EBV-infected B-cells, increased num-bers of EBV-infected cells (approximately three times that of normal individuals) as well as higher titres of anti-RANA antibodies (rheumatoid arthritis nuclear antigen which is EBV specific). Fox et al. (59) have demonstrated that a 62 kd cellular protein shares an epitope with a virus encoded protein, EBNA-1 (Epstein-Barr nuclear antigen 1) but the significance of this in the pathogenesis of RA is still not clear. 4 2. Ariimal models of arthritis To study any human disease, the presence of a suitable animal model is always desirable since it allows for the investigation of certain disease aspects that would be unethical in humans. This is especially true of infec-tious diseases. Unfortunately, there is no appropriate animal model for the study of the human arthrotropic viruses (104) since almost all of them can only replicate in cells of human origin. However, there are a few ex-amples of arthritis in animals, two of which are naturally occurring and have an identified viral cause. a. Caprine arthritis-encephalitis This is a multi-system disease of goats caused by a lentivirus, caprine arthritis encephalitis virus (CAEV) (35, 38). The clinical manifestations of this disease depend upon the age of the goat, with arthritis being the most common in adult goats (greater than one year) and leukoencephalomyelitis being the predominant complica-tion in younger animals (88). The result of infection in young goats is often so severe that the animals either die or need to be destroyed within a week of infection (88). However, those that survive often go on to develop arthritis. Studies by Crawford et al. (38) clearly demonstrated the causal role of CAEV in the arthritis of goats by satisfying the postulates set forth by Koch over 100 years ago (24). Firstly, the virus was isolated from synovial membrane explant cultures from afflicted animals and grown in fetal synovial membrane cells. When this virus was inoculated into healthy goats they developed arthritis and the virus was again isolated from the affected tissue in culture. One of the more interesting aspects of this disease is that afflicted animals are completely devoid of neutraliz-ing antibodies to the virus. Experimental infection results in detectable antibody responses that remain for up to nine months after infection. Despite the fact that antibodies are produced for all major structural proteins, neutralizing activity could not be demonstrated (88). Lack of neutralizing antibodies in the presence of other virus specific antibodies, may allow greater dissemination of the virus by enhancing infec-tivity (119) but at present the mechanism of induction of arthritis is not understood. b. Avian viral arthritis The only other reported example of naturally occuring virus induced arthritis in animals is avian viral arthritis, which affects chickens (63). The etiologic agent is a reovirus that has been isolated from affected 5 birds. This disease has received little attention and this is probably due to the presence of a very effective immunization programme (104). C. Arthrotropic Viruses Many different viral infections may be associated with joint inflammation during the acute stage of disease. The frequency of arthritis following infection varies from a few isolated cases (human cytomegalovirus, adenovirus, Coxsackie virus B) to large numbers of cases described in population studies (human parvovirus B19 and rubella virus). The viruses included in the research of this thesis (rubella virus, adenovirus and human parvovirus B19) will be reviewed separately here and other arthrotropic viruses will be discussed together at the end. 1. Rubella Virus Within the Family Togaviridae there are four genera: Alphavirus, Rubivirus, Arterivirus and Pestivirus (130, 158). All togaviruses possess positive sense, single-stranded RNA genomes that are approximately 10-12 kb in size (169). The genome is enclosed within a protein capsid which is in turn surrounded by a lipid en-velope. All of these viruses have common structural and replicative characteristics but they differ with respect to host range (158). All the alphaviruses replicate in arthropods in addition to humans or animals, whereas arteriviruses and pestiviruses are only pathogenic in animals. Rubella virus (RV), the only member of the genus Rubivirus, is believed to only infect humans (169). Rubella was recognized in 1881 by the International Congress of Medicine in London (18) as a distinct dis-ease that mainly affected children, but it was not until 1941 that the association between RV infection in early pregnancy and congenital defects was observed (33,72). This sparked a great deal of interest in this virus, resulting in the introduction of a live, attenuated vaccine in 1969 in the United States (18). Also referred to as "German measles", uncomplicated RV disease is marked by mild fever, rash, sore throat and arthralgia, but often the infection is subclinical (71,135). When symptoms occur, they are usually tran-sient and mild but post-rubella complications such as thrombocytopenia, encephalopathy and arthritis do occur (19,164). 6 The association of arthritis in conjunction with RV infection was first noticed by Sir William Osier, a Canadian physician in 1906 (110). Since that time a tremendous amount of research has been focused on the arthritis following infection with both wild-type and vaccine strains. At present there are no definitive answers concerning the mechanism of joint inflammation elicited by RV, but the more recent reports sug-gest persistent infection (either within the joint or in circulating lymphoreticular cells) to be the cause of chronic forms of rubella-associated arthritis (RAA) (28,29,30,67,68,155). The arthritis associated with RV is usually of abrupt onset and appears one to five days following infection or two to four weeks post-vaccination (55,126). The symptoms appear at, or near, the time of seroconversion, often coinciding with the rash (55,82,131). Most frequently, RAA is an acute, migratory polyarthritis that is symmetrical in distribution (82) and while it is most often transient, chronic or recurrent symptoms occur in a proportion of patients. The joints most commonly affected are Cm descending order) hands, wrists, knees, ankles, elbows and feet (55). Factors predisposing individuals to RAA are sex and age; the highest incidence being in women over 25 years of age. The incidence of RAA is also higher following infection with wild-type strains as compared to vaccine strains but there is also variation between different vaccine strains. Since the time of Osier's initial report, numerous publications have appeared which comment on these aspects, and these are summarized in a recent report by Tingle et al. in 1986 (154). These investigators had the unique op-portunity to examine subjects during an epidemic of rubella in addition to subjects involved in a rubella vac-cination programme. Their results confirmed earlier reports (97) of the increased incidence of RAA in women following wild-type infection (52.2% of women compared to 8.7% of men). In addition they demonstrated the reduced frequency of arthritis following immunization with RA27/3 as compared to natural infection (only 13.6% of vaccinated women experienced joint inflammation). No statement regarding gender differences following vaccination could be made, as all the vaccinees were female. To examine the chronic or recurrent nature of RAA, all of the subjects from both groups were followed for at least 18 months. By doing this, Tingle et al. were able to conclude that increased severity of initial symptoms (objective arthritis compared to arthralgia) preceded joint symptoms of longer duration for both natural and vaccine infections. Of the women vaccinated 15.9% experienced two or more episodes of joint involvement, 9% were still experiencing arthralgia or arthritis six months post-immunization, and 4.5% were still 7 symptomatic 18 months later. As might be expected the incidence of prolonged symptoms increased in the group infected with wild-type RV, where 43.5% of females experienced two or more episodes of joint symptoms. By six months post-immunization 39.1% were still symptomatic and at the end of the study, 30.4% were still experiencing arthralgia or arthritis. The RA27/3 strain of RV used in this immunization programme is the strain currently in use throughout North America, and while it is associated with far fewer and less severe complications than wild strains or two of the older vaccine strains (HPV77/DE5 and HPV77/DK12) (118), there are at least 2 other strains (Cendehill and To336) that have demonstrated an even lower level of complications (22,53,55). Since RAA is much less common in men and pre-pubescent females, the question of hormonal involvement is often raised with consideration to menstrual cycle or the use of oral contraceptives. Reports by Fox et al. (58) and Weibel et al. (165) suggested that the use of oral contraceptives offered protection from RAA but others (22,34,95,102) disagreed. Similarly, there is controversy concerning the correlation between joint complications and the optimal time of vaccination during the menstrual cycle. Monto et al. (102) and Copper et al. (34) found no correlation at any given point, while others (22,58,103,149) found that there were times within the menstrual cycle when vaccination resulted in decreased joint symptoms. However, the ideal time for vaccination was found to be different for each group. The numerous observations that RAA may become chronic or recurrent led several groups to search for the involvement of RV in chronic arthritides such as RA and juvenile rheumatoid arthritis (JRA). In 1968 Mar-tenis et al. (100) presented a case study of a woman diagnosed with definite RA, by American Rheumatology Association (ARA) (156) criteria, which began seven days after a typical rubella infection. The authors sug-gested either repeated exposure to, or persistent infection with RV to be the trigger of the RA due to the lengthy duration of high hemagglutination-inhibition titres. An earlier study (83) reported that 90% of RAA cases were rheumatoid factor (RF) positive compared to 28.6% of uncomplicated rubella infections. The authors did not indicate if their subjects met any other requirements for the diagnosis of RA other than the presence of serum RF and other reports (56,75,105) have failed to show a definite relationship between rubella infection and RA. However, there is good evidence to suggest a role of RV in JRA In 1975 Ogra et al. (105) showed that patients with JRA below the age of 10 had anti-RV IgG titres four to eight times higher than controls. In addition RV antigen was found in synovial fluid smears in nine out of 25 JRA patients but not in controls with RA, osteoarthritis or septic arthritis. The smears that were positive for RV were negative for mumps virus, polio virus and measles virus antigens. Ten years later Chantler et al. (28) were able to isolate RV from mononuclear cells in seven of 19 children with chronic arthritis (five with JRA and two with seronegative spondyloarthritis). In three cases RV was isolated from both peripheral blood and synovial fluid mononuclear cells (PBMC and SFMC), in two cases from PBMC, and in two cases from SFMC only. It is im-portant to note that the diagnostic criteria for RA and JRA, as set forth by the ARA, are very specific. For this reason many cases of RAA may not be placed into these categories despite their chronic nature. After an infectious agent is firmly associated with disease, the immediate questions which follow concern the mechanism of pathogenesis, which are crucial in order to design programmes of therapy and prevention. A few possible mechanisms were outlined at the beginning of this introduction. The development of RAA often coincides, or immediately follows the typical rubelliform rash (34,159) and since both these symptoms occur at the same time as RV specific antibodies develop, immune complexes have been suspected to cause both (77). The conclusion by Vergani et al. in 1980 (159) favoured the role of immune complexes since 100% of the vaccinees who developed joint symptoms possessed circulating im-mune complexes but only one asymptomatic patient was positive by the Clq binding assay. In contrast, Ziola et al.(172) examined the circulating immune complexes in 24 patients with acute rubella infection and found that the prolonged presence of circulating immune complexes in two patients was not associated with any symptoms. This was confirmed by Singh et al. (141) who found no correlation between increased levels of cir-culating immune complexes and the development of arthritis or arthralgia in patients infected with wild-type or RA27/3 virus. All of these studies examined immune complexes in sera and therefore were unable to comment on the pos-sibility of immune complexes localized to the joints. Given that synovial tissue mononuclear cells have been shown to secrete RV specific antibodies (80), it is conceivable that a persistent or recurrent infection within the joint could result in the localized production of immune complexes. 9 The ability of RV to directly invade the joint is provided by the virus' ability to infect lymphoreticular cells (26) and the passage of the latter into the synovial fluid (43). The ability of RV to establish persistent infec-tions in vivo in various cell-types (122), including cartilage (96,144) has been documented and Cunningham et al. (40) established human synovial cells persistently infected with RV in vitro. These reports, taken together with the numerous reports of RV isolation from PBMC and SFMC (27-30,56,67,68,78,106,145, 155) from patients with various forms of chronic arthritis, suggest that RAA is the result of persistently in-fected lymphoreticular cells which transport the virus to the joint. Once in a joint, it may then replicate and/or persist giving rise to inflammation. 2. Adenovirus The human adenoviruses (Ads) belong to the genus Mastadenovirus in the Family Adenoviridae. This group of viruses has been extensively studied with respect to molecular biology and mechanisms of pathogenesis (for review see reference 80). Since the initial isolation in 1953 from the tonsils of a child, 42 distinct serotypes have been identified. The virus particles of all Ads are non-enveloped icosahedrons 65-80 nm in diameter (116). Within the protein coat is a double-stranded DNA genome that is approximately 30 kb in size (116). Covalentiy attached to both 5' ends of the viral genome is a 55 kd terminal polypeptide which is required for viral DNA synthesis (80). There are approximately 11-15 different proteins constituting Ad virions and at least five other virus-encoded proteins that are not assembled into complete particles. These latter proteins are mainly involved in DNA synthesis (80). There are several different classification schemes used to group Ads. Based upon hemagglutination patterns, this original classification yields three different groups. Ads within the same hemagglutination group were shown to have similar oncogenic potential in animals (all Ads show positive transformation capability in cell culture) and the oncogenicity may be related to the G+C content (143). After grouping by hemagglutination, oncogenicity and G+C content, Ads are assigned specific serotypes using monoclonal antibodies to specific epitopes found on two structural proteins (hexon and fiber). 10 Ads give rise to a variety of different illnesses in humans as a result of their ability to replicate in epithelial cells at different sites within the body (80). The more common Ad-associated diseases include acute respiratory disease, pneumonia, conjunctivitis, gastroenteritis, and cystitis (52). Several Ad serotypes are highly malignant in animals (143) but there has been no confirmation of Ad involvement in human tumors (20,142). Joint involvement is not a common complication of Ad infection but the literature contains a least four separate reports on seven individuals. A review of these separate cases indicates that when Ad is involved in the arthritic process it may do so via different mechanisms. The earliest report of Ad-associated arthritis was presented by Panush in 1974 (112). Three weeks after vac-cination with Ad types 4 and 7, a 19 year old military recruit developed a flu-like illness accompanied by swell-ing of the hands, wrists and elbows. Virus cultures from synovial fluid, rectal swabs and skin lesions were all negative but serum analysis indicated a 10-fold increase in Ad type 7 specific antibodies between acute and convalescent specimens. There was no similar increase in antibody litres to Ad types 3,4 or 5. This patient fully recovered in four weeks and the final diagnosis was inflammatory synovitis due to systemic Ad type 7 in-fection. The increase in the serum complement component C3 and a decrease in C4 during the symptomatic period, together with the absence of virus in the synovial fluid, suggested that the immune complexes were responsible for the symptoms rather than direct invasion of the virus into the joints. The next report described a nine year old girl who was repeatedly hospitalized during a 20 month period (121). The first diagnosis was recurrent viral pericarditis, when Ad type 7 was isolated from pericardial fluid. During this time there was joint pain, but true arthritis did not appear until eight months later. The arthritis lasted for approximately one year, after which she recovered and suffered no further sequelae. The authors suggested that the initial joint pain, which coincided with the virus isolation favoured direct invasion of the joint and that the subsequent arthritis was immune complex mediated but this is purely speculative since they report no evidence in support of either hypothesis. The report by Utsinger in 1977 (157) of three adult women who developed arthritis shortly after an upper respiratory illness, is highly suggestive of immune complex mediated disease. Ad infection in these patients 11 was confirmed by direct isolation from throat swabs and complement fixation. All three patients possessed cryoglobulins (antibodies that precipitate at low temperatures) and immune complexes in both serum and synovial fluid. These cryoglobulins reacted with antiserum specific for Ad. In addition two out of three had decreased levels of serum C3 and C4, the classical indicator of immune complex mediated disease (166). The most recent report of Ad-associated arthritis is also the most convincing. In 1985, Fraser et al. (60) reported a 29 year old male who was diagnosed as having acquired hypogammaglobulinemia. Clinical tests indicated rheumatoid-like arthritis involving fingers, wrists, elbows, shoulders and knees. His right knee re-quired total replacement 16 months after a synovectomy was performed on the same joint. Tissue samples from both procedures were cultured and when these culture media were used to infect other cells, cytopathology appeared. This cytopathic effect was specifically neutralized by antibodies to Ad type 1. Similar cultures of peripheral lymphocytes were negative. It is quite likely that the immune deficiency of this patient contributed to his disease perhaps by allowing greater dissemination or recurrent infection with the virus. This is a valuable report because the direct isolation of a virus from affected joint tissue is not a com-mon occurrence. What is also important is that the virus was isolated twice over a 16 month period of time, demonstrating that the synovial infection was not transient. Moreover, the absence of the virus in the circula-tion suggests a tropism for the joints. Two conclusions to be drawn from these reports are that one, arthritis is not a common complication of Ad in-fection and two, when it does occur the mechanism of disease induction may vary depending on the condi-tion of the patients immune system. When the patient has no underlying immunological dysfunction, the arthritis is more likely to be a result of immune complex deposition which appears to be acute and self-limited. However, when the infection is not impeded by the host's immune defences, the virus may invade and cause complete destruction of the joint. 3. Human parvovirus, B19 The virus family, Parvoviridae, comprises three genera: Parvovirus, Dependovirus and Densovirus. Depen-doviruses, (formerly adeno-associated viruses) gain their name from the fact that they require the "help" of a co-infecting adenovirus or herpesvirus for complete replication (21). Densoviruses, on the other hand, repli-cate autonomously in insects (21). Although several dependoviruses have been shown to infect humans, 12 they have not yet been associated with any disease (113) and so neither they nor the densoviruses will be dis-cussed further. Within the genus Parvovirus, only the one designated B19 has been firmly associated with human disease. There are reports of two other potential human parvoviruses but at present their identification is somewhat dubious (7,140,147). One of these, RA-1, was presumed to have been isolated from patients with RA, but there is still some debate as to whether the virus came from the joint material or from the mouse brains through which the joint extract was passaged for viral amplification. Stierle et al. (147) performed similar ex-periments but their results have further confused the matter. The first isolate by Simpson et al. (140), was shown to be antigenically distinct from other known parvoviruses; RA-1 did not react with antibodies to other parvoviruses nor did anti-RA-1 antisera react with other parvoviruses. The confusion arises when considering the second isolate (147) which was also antigenically distinct but was shown to be morphologically different. Electron microscopy indicated that the virus isolated by Stierle et al. was approximately half the size of the one isolated by Simpson et al. Further studies are obviously necessary to settle the controversy. Human parvovirus B19 (hereafter called B19) is a relatively new virus with an interesting history. It was only identified 15 years ago and is unusual because it was isolated and identified before it was associated with dis-ease. B19 was discovered by accident in 1975 by Cossart et al. (36) during antigen screening tests for hepatitis B virus (HBV). Unexpected precipitin lines from serum of HBV negative blood donors resulted in its detection. It was placed in the genus Parvovirus based on its physicochemical properties (37,148). It is a small (20-25nm) (36), non-enveloped virus (21) containing a linear, single-stranded DNA genome of 5.5 kb (148). It is capable of complete replication without a co-infecting virus but, like all the autonomous par-voviruses, it is restricted to rapidly dividing cells, requiring the S-phase of the cell-cycle in order to replicate (124,151). Two diseases, aplastic crisis and erythema infectiosum, were suspected as having a viral etiology based upon epidemiological and clinical observations. Six years after the discovery of B19, Pattison et al. (114) dis-covered B19 in the sera of two children with hemolytic anemia who were undergoing aplastic crisis. Sub-sequent reports (13,93,134) confirmed that up to 90% of the cases of aplastic crisis were attributable to this virus due to the fact that it is cytotoxic to erythroid progenitor cells (111, 170). Two years later Anderson et al. (11) confirmed that B19 was the infectious agent responsible for the pediatric disease, erythema infec-tiosum, (also referred to as Fifth disease or 'slapped cheek' syndrome due to the characteristic rash) (6). This is a mild exanthematous disease that may be clinically indistinguishable from rubella infection (153). Symptoms associated with adult infection tend to be more severe and often the rash is absent. Many aspects of B19 disease were clarified by Anderson et al. (10) during examination of experimentally in-fected adults. Results from these investigations indicated that the route of transmission is via the upper respiratory tract and that viremia occurs one week after infection but only in IgG negative volunteers. The viremia lasts for approximately five days and may reach titres of 1011 genome copies per millilitre of serum and at the same time virus may detected in nasal washes. The clinical manifestations of B19 infection ap-peared in two phases. The first phase occurred at the time of viremia and was characterized by fever, malaise, myalgia, chills and itching. The second phase began 17 days post-inoculation and included a maculopapular rash on the trunk and limbs, with joint symptoms following 24 hours after the rash appeared. This study also confirmed that erythropoiesis was interrupted resulting in a slight drop in hemoglobin in nor-mal individuals. The most notable complication of B19 infection in normal individuals is arthritis. The literature contains numerous reports associating this virus with inflammatory joint disease and these are summarized below (1,8,10-14,17,32,36,45,47,48,50,70,84,89,93,94,99,113,123,128,129,133,161,163,167). Two studies (1,161) in the late 1960's were of the first to report an association between joint involvement and erythema infectiosum, almost 10 years prior to the identification of the causative agent. Since then many similar case reports and larger epidemiological studies have followed which have defined many of the charac-teristics of B19-associated arthritis and shown that there are striking similarities to RAA. Firstly, the frequency of B19 arthritis is much higher in adults than in children; one early study of erythema infectiosum indicated that only 5% of children less than nine years of age experienced joint pain whereas 77% of adults greater than 20 years old had similar symptoms (1). One study involving children (123) reported that acute arthritis was of a longer duration in children as compared to adults (average of 54 days in the 14 children and 13 days in the adults) but of the reported cases of chronic arthritis (greater than 2 months) all have been adult women with the exception of one 11 year old boy (123,167). A second generalization is that in adults, women are affected much more often than men and when it does occur in men it is usually less severe and of shorter duration. Regrettably not all the reports distinguished males from females with respect to symptoms, but of the reports including specific information concerning af-fected men, (47,84,128,161) the average duration of joint symptoms was nine days compared to the average of 60 days seen in adult women. B19-associated arthritis is usually symmetrical and may involve a variety of joints in different patients regard-less of age or sex. The most frequently involved joints are, On descending order) knees, fingers, hands, wrists and ankles (1,13,17,45,47,48, 50, 70,84,99,123,128,129,133,161,167). Less commonly affected joints include elbows, back, feet, hips, shoulders and neck. In all the literature reviewed, there were no reports of joint destruction. All of the above reports utilized serum antibody analysis to confirm B19 infection at or near the time of arthritis and collectively they have clearly demonstrated that B19 is involved in the pathogenesis of arthritis. With its role firmly established, more in-depth studies concerning pathogenic mechanisms are beginning to appear. Induction of joint inflammation by B19 is largely believed to be immune complex mediated because the arthritis usually coincides with the increase in antibody litres in a manner similar to RAA. However, this is lar-gely speculative since synovial fluid analysis for B19 antigen or DNA are lacking in all but two of the in-cluded studies. The first of these reports (32) examined 76 patients with RA and inflammatory arthritis. They identified six patients with evidence of recent B19 infection (presence of IgM antibodies specific for B19) but did not find either B19 antigen or DNA within the synovial fluid. However, two years later, Dijkmans et al. (45) reported the detection of B19 DNA in the synovial fluid of a 33 year old woman with B19-associated arthritis. The nega-tive results of the former group may be a reflection of sampling time. They examined samples taken up to five months after the onset of symptoms whereas the positive result was achieved with a specimen that was 15 taken only 24 hours after the arthritis appeared. Although the results reported by Dijkmans et al. do not dis-tinguish between direct viral invasion and immune complex mediated disease, it is important to note that when arthritis does occur in association with B19 infection it is after the viremic phase (10). Given the high incidence of arthritis associated with B19 infection, it is not surprising that several groups have attempted to link the virus to RA but the results of these efforts require confirmation. Cohen et al. (32) demonstrated that 4 out of 69 patients with RA experienced B19 infection at the onset of their disease but as there was no indication of chronic infection, they concluded that B19 was the trigger in a genetically predisposed population. Another study (94) concluded that B19 was not involved in the pathogenesis of RA because B19-specific IgG levels were similar between controls and patients. However, these latter findings may be the result of insensitive techniques and poor sampling times (113,134,153). Several rheumatic diseases are associated with specific HLA haplotypes; HLA-B27 is present in up to 90% of patients with ankylosing spondylitis and approximately 75% of reactive arthritis cases (171). In addition, there is also a correlation between RA and the presence of HLA-DR4 (5,126). It was therefore of interest to deter-mine whether B19 arthritis was associated with a specific HLA haplotype. Unfortunately, these results, like those searching for a connection between B19 and RA, were ambiguous. Dijkmans et al. (48) reported that out of five patients with B19 associated arthritis only one was DR4 positive and all five were B27 negative. In contrast, Klouda et al. (89) examined 18 patients with B19-associated arthritis and found that 67% were DR4 positive as compared to 36% of controls. Interestingly, all of these patients were negative for HLA-DR1 (compared to 20% of controls) suggesting that this antigen may provide resistance to B19 infection. The current literature provides more than sufficient documentation to conclude that B19 is involved in the arthritic process, but the conflicting reports concerning its involvement in RA, HLA association and its mechanism of action make it clear that further investigation is required. 4. Other Arthrotropic Viruses Arthritis is a symptom of infection by various other viruses including hepatitis B virus, herpesviruses, mumps virus, enteroviruses, alphaviruses and smallpox virus. At present there is no unifying theory for the 16 pathogenic mechanisms of viral arthritis. Because the arthrotropic viruses themselves, are so diverse, it would seem likely that their modes of induction of arthritis will also vary widely. Hepatitis B virus (HBV) infection was one the first viral infections reported to be associated with joint inflam-mation (69) and this is probably one of the best examples of immune complex mediated disease (132,138). Arthritis is not associated with either hepatitis A virus or hepatitis non-A non-B infections and due to diagnos-tic confusion between these three agents it is difficult to comment on the frequency of joint inflammation as-sociated specifically with HBV (4,82). Five alphavirues (Family Togaviridae) are also strongly associated with arthritis (57,152) but due to their geographic distribution (Africa, South America and the South Pacific) they have not received as much atten-tion. These five viruses (chikungunya virus, o'nyongnyong virus, Mayaro virus, Ross river virus and Sindbis virus) are all transmitted to humans via mosquito vectors. The symptoms of alphavirus-associated arthritis are similar to RAA (57) and that the frequency of their association with joint inflammation is high is reflected in the names of two of them. In the language of East African tribes, chikungunya means "that which bends" and o'nyongnyong means "joint-breaker". One report of arthritis following an outbreak of chikungunya fever (57) noted that four out of five of the patients with chronic arthritis were HLA-B27 positive. The ability to stimulate B-cells (see Introduction), makes EBV a good candidate for an etiologic agent respon-sible for rheumatic diseases such as RA and SLE, due to the myriad of autoantibodies present in these dis-eases (42,168). The occurrence of arthritis has been documented for other human viruses (herpesviruses, enteroviruses, mumps virus and smallpox virus) but the numbers of reported cases are few. For more detail than may be in-cluded here concerning these viruses, there are numerous reviews (3,4,25,43,55,74,82,86,115,117,126, 131). D. Purpose of the Study The literature contains numerous accounts of the arthritogenic nature of many viruses but unfortunately the pathogenic mechanisms involved in the induction and perpetuation of joint inflammation are poorly under-stood. At least in part, this is due to the fact that there has been no model system to study the interaction of 17 viruses with joint tissue. An animal model would be ideal, as it would allow for the investigation of im-munopathological mechanisms but as previously mentioned, the majority of human arthritogenic viruses are restricted to cells of human origin. However, the mechanism of direct invasion may be studied in vitro assum-ing that if a virus is unable to replicate in joint cells under ideal conditions, (i.e., in the absence of an intact im-mune system) it is unlikely that it could do so in vivo. A culture system using cells and tissue of joint origin was therefore established to examine the replicative abilities of three different viruses, two of which are highly arthrotropic (RV and B19), and a third whose as-sociation with arthritis is much less frequent (Ad). In addition, the study of RV was expanded to a com-parison of eight different strains (three vaccine and five wild-type) whose association with joint inflammation was known to vary. The permissiveness of both primary synovial membrane/cartilage cells and tissue to these viruses has been examined and the ability of the viruses to produce long-term persistent infection also assessed. The validity of the model system has been determined by comparing the results obtained with clini-cal data on the relative arthritogenicity of the various test viruses. II. ABBREVIATIONS Ad = adenovirus ALCM = agar leukocyte conditioned medium B19 = human parvovirus BM = bone marrow BSA = bovine serum albumin CPE = cytopathic effect DAB = diaminobenzidinetetrahydrochloride DMEM = Dulbecco's minimal essential medium dUTP = 2'-deoxyuridine5'-triphosphate EBV = Epstein-Barr virus EDTA = emylenediarnine tetraacetic acid EPO = erythropoetin FBS = fetal bovine serum FC = fetal chondrocytes FFU = focus forming unit fg = femtogram, 10"15 gram HBSS = Hanks balanced salt solution HCl = hydrochloric acid Hep-2 = human epithelial cells HIFBS = heat-inactivated fetal bovine serum HLA = human lymphocyte antigen IgG = immunoglobulin gamma IP = immunoperoxidase JRA = juvenile rheumatoid arthritis kb = kilobase kd = kilodalton MHC = major histocompatibility complex MOI = multiplicity of infection PBS = phosphate buffered saline PBS + = phosphate buffered saline with calcium and magnesium p.i. = post-infection PF = paraformaldehyde RA = rheumatoid arthritis RAA = rubella-associated arthritis RPMI = Roswell Park Memorial Institute RV = rubella virus SDS = sodium dodecyl sulphate SLE = systemic lupus erythematosus SMC = synovial membrane/cartilage TAE = Tris, acetate, EDTA buffer TCID = tissue culture infectious dose 20 TE = Tris, EDTA buffer TES = Tris, EDTA, SDS buffer TNE = Tris, NaCl, EDTA buffer Tris = trisOiyoVoxymemyOammomem Vero = African green monkey kidney 21 III. MATERIALS A. Materials 1. Cells Vero (African green monkey kidney) cells, and Hep-2 (human epithelial) cells were obtained from American Type Culture Collection (ATCC). FC, (fetal chondrocytes) were a gift from Dr. L Sokoloff, New York, and SMC (synovial membrane/cartilage) cells were prepared from human fetal knee joints. Bone marrow (BM) cells from a patient with chronic myelogenous leukemia were kindly provided by Dr.K Humphries, Terry Fox Laboratories. See Methods for details of cell maintenance. 2. Viruses a. Rubella virus strains Cendehill - Rohm Pharma RA27/3 - Meruvax II Merck, Sharp and Dohme HPV77/DE5-MeruvaxI Merck, Sharp and Dohme M33-ATCC wild type Therien - adapted in Finland Thomas - congenital rubella syndrome isolate 1B2 - plaque purified patient isolate 5209R- B.C. Provincial Health Laboratory wild-type b. Adenovirus type 2 - Dr. B Ziola, University of Saskatchewan c. Human Parvovirus B19 - Dr. P Tattersall, Yale University 3. Cell culture All cell culture materials were obtained from Gibco (Burlington, Ontario) except for Biorich medium (Flow Laboratories, Mississauga, Ontario) 4. Antibodies The mouse monoclonal antibody to adenovirus 5 was purchased from Biocan Scientific (Mississauga, On-tario) and recognizes a groups specific antigen. The rabbit polyclonal sera to B19 fusion proteins were kindly provided by Dr. P Tattersall. See Methods for rubella virus antiserum preparation. 5. Miscellaneous All photographs were taken with Kodak 160 ASA Tungsten film. B. Solutions 1. PBS pH 7.0 3. 4. PBS+ pH7.3 TNEpH7.4 = TAE pH 8.5 5. PF 6. Hematoxylin 150mM NaCl 7.5mM Na2HP04 2.5mM NaH2PO4.H20 140mM NaCl 2.7mM KC1 8.3mM NaH2P04 1.5mM K2PO4 0.9mM CaCb 0.5mM MgCl2 lOmMTris lOOmM NaCl ImM EDTA 40mM Tris-acetate 2mM EDTA = 4% (w/v) paraformaldehyde in 0.1M phosphate pH 7.4 = 25% (v/v) 1,2 ethanediol 0.2% (w/v) hematoxylin 0.4mM NaI04 12.6mMAl2(S04)3.15H20 2% (v/v) acetic acid 23 7. Eosin = 0.05%(w/v)EosinY 0.25% (v/v) acetic acid 70% (v/v) ethanol 8. HBSS = Hanks Balanced Salt Solution 5.4mMKCl 0.44mM KH2P04 136.9mM NaCl 4.2mM NaHCOs 0.33mMNa2HP04.12H20 5.5mM D-glucose 9. HBSS+ = 1.26mMCaCl2 0.81mM MgS04 0.001% (w/v) phenol red HBSS (as above) 10. Transport medium = HBSS + 300U/ml penicillin 300ug/ml streptomycin 0.75ug/ml amphotericin 11. Collagenase buffer = 1% (w/v) collagenase II (Worthington Biochemical Corp., Freehold, New Jersey) = 0.1M Tris pH 7.5 = 0.1M CaCU 12. DE buffer pH 7.2 = 32mMTris 25mM HC1 50mMNaCl lOmM NaN3 13. Luria broth = 1% (w/v) tryptone 0.5% (w/v) yeast extract 0.5% (w/v) NaCl ImM NaOH 14. glucose/Tris/EDTA = 50mM glucose 25mMTrisClpH 8.0 15. T E = 10mMTris.Cl(ph7.5) 16. Lysis buffer ImM EDTA (pH 8.0) 4ml TES buffer 2ml 3M NaCl 200ug/ml Proteinase K 17. TES buffer = 10mM Tris.Cl (ph 7.5) lOOmM EDTA (pH 8.0) 6% w/v SDS 25 18. Culture medium - maintenance of cells a. Vero cells - Biorich + 5% FBS b. Hep-2 cells - Biorich + 3% FBS c. FC-RPMI + 15%FBS d. SMCcells-RPMI + 9%FBS e. BM cells - Iscove's DMEM + 10% FBS The concentration of antibiotic/antimycotic was the same for all cell types: 100U/ml penicillin, lOOug/ml streptomycin and 0.25ug/ml amphotericin B. 19. Culture medium - for infected cells When the above cells were infected, the FBS was heat-inactivated (56°C for 60 minutes) and used at half the normal concentration. The antibiotic/antimycotic concentration remained the same. IV. METHODS A. Cell Maintenance All cell and organ cultures were incubated in an atmosphere of 5% CO2 unless otherwise indicated. The majority of cell-lines used were adherent and therefore were passaged in the same manner. Confluent monolayers were washed with phosphate buffered saline (PBS) and incubated at 37°C in 0.25% trypsin (w/v in Hanks Balanced Salt Solution, HBSS) until the cells detached. The cells were then thoroughly resuspended in culture medium and distributed to new tissue culture flasks or plates. The split ratio varied from 1:2 for the fetal chondrocytes (FC) to 1:20 for the human epithelial GTep-2) cells. African green monkey kidney (Vero) cells and Hep-2 cells were maintained as continuous lines and the FC were maintained until ap-proximately passage 21 when their growth slowed dramatically. The synovial membrane/cartilage (SMC) cells were primary cells and were only used until passage number 3. Bone marrow (BM) cells, stored in liquid nitrogen, were rapidly thawed and slowly added to approximately 10ml of Iscove's DMEM without fetal bovine serum (FBS). The cells were pelleted (lOOOg for 10 minutes) and washed in Iscove's DMEM without FBS. After centrifugation, the cells were resuspended in complete medium (Iscove's DMEM, 20% FBS, 5% ALCM and 1.5 units/ml EPO) and incubated for 2.5 hours to allow dissociation of clumped cells prior to infection (see Viral Growth Curves). B. Synovial Membrane/Cartilage Cell Production Human fetal knee joints (ranging in age from 15-20 weeks gestation) were collected in transport medium from Anatomical Pathology at Vancouver General Hospital. Skin, muscle and fat tissue were aseptically removed to expose the joint. The entire joint was cut into 2-3 pieces and placed in RPMI, 4.5% HIFBS plus an-tibiotic/antimycotic and incubated for 15 hours. The medium was then removed and the pieces were in-cubated in collagenase for 3 hours at 37°C. The collagenase solution was replaced with 0.25% trypsin (w/v in 0.6mM EDTA) and the tissue was agitated with a magnetic stirrer for 15 minutes at 35°C in the absence of CO2. The released cells were removed and placed in RPMI, 9% FBS and the tissue was treated again with trypsin/EDTA 3 times. The cells were pooled from each cycle, pelleted (lOOOg for 10 minutes) and washed in PBS. The cells were resuspended in RPMI medium, 9% FBS and allowed to settle overnight in 60 x 35 mm tissue culture plates. The next day the cells were washed 2 times in PBS and incubated until confluent. At that time the cells were passaged 1:3 and this was referred to as passage 1. C. Preparation of Intact Tissue The fetal knee joints were collected and prepared as described above except that the joints were cut into 5-8 pieces depending on their size. These pieces were infected without further treatment. D. Virus Stock Production 1. Rubella Virus (RV) Strains All 8 strains of RV (see Materials for origins) were prepared by infecting a single subconfluent flask of Vero cells at a multiplicity of infection (MOI) of 1-10 for an adsorption period of 4 hours at 35°C. The inoculum was removed and the cells incubated in Biorich medium 2.5%, heat-inactivated fetal bovine serum (HIFBS) for 3 days or until 50% of the cells were exhibiting cytopathic effect (CPE). At that time the culture supernatant was distributed to 5 fresh flasks of subconfluent Vero cells. Again the cultures were incubated for 3 days or until 50% CPE was present At that time the medium was removed, replaced with 10ml of fresh medium and incubated for 24 hours. The virus stock was prepared by clarifying the culture supernatant of cell debris by centrifugation (3000g for 10 minutes) and storage in aliquot parts at -70°C. 2. Adenovirus (Ad) Subconfluent monolayers of Hep-2 cells were infected at an MOI of 1-5 for an adsorption period of 60 minutes at 37"C. The inoculum was removed and replaced with Biorich medium, 3% HIFBS. When 50-60% of the cells were exhibiting CPE, the culture medium was replaced with 10ml of fresh medium and 24 hours later, cell debris was removed from the culture medium by centrifugation (3000g for 10 minutes) and stored in aliquot portions at -70°C. E. Virus Titration 1. Rubella Virus Strains All strains of RV were titred in duplicate by infecting subconfluent Vero cells with serial 10-fold dilutions of virus in Biorich medium, 1% HIFBS. After a 4 hour adsorption period, the inoculum was replaced with an overlay of 0.5% agarose (Seakem, Mandel Scientific, Guelph, Ontario) in 199 medium. The focus forming 28 units (FFU) were counted 7-10 days post-infection (p.i.) for HPV77/DE5 and the wild-type strains. In the case of Cendehill and RA27/3, the monolayers received a second overlay containing neutral red on the third or fourth day following infection. These were counted the day after addition of the neutral red. 2. Adenovirus Five and ten-fold dilutions of virus in Biorich, 1% HIFBS were distributed into 24-well plates (0.5ml/well) in quadruplicate. 4 X104 Hep-2 cells were then added in an equal volume of 3% HIFBS medium. Wells with CPE were marked 10 days p.i. and the virus tifre determined by the Karber method (39) where, logioTCID = 0.5 +D-S TCID = tissue culture infectious dose D = logio of the highest dilution where all wells show CPE S = sum of me proportions of the wells demonstrating CPE Sample Calculation: Dilution logio # Wells with CPE Proportion _ __ _ . 2X10-5 -4.7 4/4 1 10-5 -5.0 4/4 1 2X10-6 -5.7 4/4 1 10-6 -6.0 3/4 0.75 2X10-7 -6.7 1/4 0.25 10-7 -7.0 1/4 0.25 logio TCID = 0.5+ (-5.5)-1.25* = -6.25 TCID = 10*25 TCID = 1.78xl06/0.5ml = 3.5xl07ml *Note: S is calculated using the proportions from dilutions separated by a factor of 10 F. Fixation of Cells and Tissue 1. Paraformaldehyde (PF) fixation Cells to be fixed for imrnunoperoxidase (IP) staining were washed in PBS and covered with PF for 5 minutes at room temperature. The PF was replaced with 0.15M ethanolamine for one minute. This was removed and the cells were stored at 4"C under 70% ethanol until tested. Pieces of tissue were washed in PBS and covered with PF for 3 hours on ice. The PF was replaced with 0.15M ethanolamine for 20 minutes and the tissue was stored at 4"C in PBS until sectioned. 2. Acid/alcohol fixation Cells to be fixed were washed in PBS and covered in a solution of 5% acetic acid in 95% ethanol for 10 minutes at -20"C. The cells were washed four times in PBS and examined immediately for virus antigen by IP staining. G. Viral Growth Curves Subconfluent monolayers of cells (Vero, Hep-2 or SMC cells) in tissue culture plates (10 x 30 mm) were in-fected with virus (RV strains or Ad) diluted in PBS. An extra aliquot of virus was prepared at the time of infec-tion so that the input virus titre could be determined. At various times (see Results) one plate of cells was harvested to prepare supernatant and cell-associated virus, and another was fixed with PF for subsequent viral antigen detection by IP staining. Supernatant virus was prepared by removing the culture medium and freezing at -70°C. Cell-associated virus was obtained by trypsinizing the cells and washing the plates 3-4 times with PBS containing calcium and mag-nesium (PBS+) once the cells had detached. The cells were then pelleted (3000g for 10 minutes) and resuspended in 0.5ml of 0.02M Tris for 5 minutes. They were then sonicated for 5 seconds, mixed with an equal volume of TNE (Tris, NaCl, EDTA) buffer and stored at -70°C. Extra plates were established at the out-set so that viable cell counts (trypan blue exclusion) could be performed at various times during the infec-tion. 30 To infect the BM cells with B19, the cells were pelleted, resuspended in 0.5ml of Iscove's DMEM (no FBS) and divided into 2 tubes. Ten microlitres of B19 were added to one tube resulting in an MOI of approximate-ly 20. The virus was allowed to adsorb for 1 hour at 35"C. After the inoculum was removed, the cells were washed in medium without additives, and finally resuspended in 9ml of complete medium (Iscove's DMEM, 20% FBS, 5% ALCM and 1.5 units/ml EPO). To harvest the viral DNA the cells were gently resuspended with a pipette and 3ml was removed. This aliquot was centrifuged to separate the cells from the culture medium and the DNA was extracted from both (see Section L). H. Persistent Infections 1. Synovial Membrane/Cartilage Cells Two infected and two control cultures of SMC cells were established in RPMI medium, 4.5% HIFBS. Samples of supernatant medium were taken 3,6, and 9 days p.i. On day 9, one infected and one control plate were fixed for viral antigen detection by IP staining and the other set was passaged 1:2 (passage 1) into RPMI, 9% FBS. The cells were passaged every week for 80 days, or as long as the cultures could be maintained. 2. Tissue Pieces of intact joint tissue were infected in undiluted stock virus for 4 hours and then maintained in RPMI medium, 4.5% HIFBS. Uninfected pieces derived from the same tissue specimen were used as controls. The medium was changed every 3-4 days and the removed medium saved for virus titration. At the indicated times the tissue was fixed. Paraffin embedding and sectioning were done by Anatomical Pathology at Van-couver General Hospital. The sections were mounted onto glass slides that were acid cleaned (1% HC1 in 70% ethanol for 20 minutes) and then dipped in a 0.5% solution (v/v in distilled water) of Kodak Photo-flo to im-prove tissue adherence (139). I. Immunoperoxidase Staining 1. Cells Endogenous peroxidase activity was eliminated by incubating the cells in 3% H2O2 in methanol for 30 minutes. Non-specific binding was then blocked by incubating the cells in a solution of 5% (w/v) bovine serum al-bumin 03SA) in PBS for 60 minutes. The cells were then incubated for 60 minutes in the primary antibody 31 . (mouse or rabbit in origin) which was diluted in 0.5% NP40 in PBS. After 3 washes in PBS, the cells were in-cubated for 60 minutes with a biotinylated secondary antibody which was directed against the primary an-tibody (either goat anti-mouse or goat anti-rabbit). The avidin-biotin complex, ABC, (Vectastain, Mississauga, Ontario) was used at half the recommended concentration for 60 minutes after excess secon-dary antibody had been removed. A10 minute incubation in a solution of 0.05% diaminobenzidine tetrahydrochloride (DAB), 0.001% H2O2 in PBS followed removal of excess ABC. Finally the cells were counterstained with hematoxylin and eosin, rinsed quickly in distilled water and air dried. To ensure that observed staining was not a result of non-specific binding during the different incubations, two controls were included with every experiment. Uninfected cells were examined with the same primary antibody to control for specificity of the anti-viral antibody and infected cells incubated with all reagents but the primary antibody were examined to ensure that subsequent reactions were not due to non-specific bind-ing. When either of these controls demonstrated stain (background) the appropriate reagent was absorbed and/or diluted until the controls were free of stain. In addition, the use of 0.5% NP40 in PBS as the diluent for the primary antibody helped to prevent background staining. Acid/alcohol was also used as an alternative fixative to PF for RV infected cells. However, it use was discontinued as it eliminated the antigen stain of Cen-dehill and RA27/3 strains. 2. Tissue Sections Duplicate sections on slides were warmed in a 70°C oven to soften the paraffin. Without allowing them to cool, the sections were placed in warm (70°C) xylene (2 changes of 5 minutes each). The sections were then rinsed in 95% ethanol for 5 minutes and incubated in a solution of 3% H2O2 in methanol for 30 minutes. The sections were then hydrated by successive 5 minute incubations in 70% ethanol followed by PBS prior to the blocking step with 5% BSA in PBS (30 minutes at room temperature or overnight at 4°C). The remainder of the staining procedure was the same as described for the cells except that more rigorous washing proce-dures were used. Instead of 3 changes of PBS, the sections were washed in 0.1% SDS (sodium dodecyl sul-phate) in PBS for 2 minutes followed by 2 washes in PBS for 5 minutes each. These more stringent washes were required to eliminate background staining in tissue. The duplicate sections on each slide were used to provide a test sample and a secondary antibody control on consecutive sections of joint tissue. J. Preparation of Rubella Antiserum New Zealand white rabbits were hyperirnmunized with RV antigen which had been prepared by the ulfracentrifugation of large volumes (10 litres) of virus derived from the culture fluid of 8-10 roller bottles of infected Vero cells. The rabbit serum was ammonium sulphate precipitated and passed through a column of DEAE-Sephacel (Pharmacia, Baie D'Urfe, Quebec) to enrich for IgG antibodies. The column was equi-librated and washed with DE buffer (see Solutions) and material eluted in the void volume was retrieved. The peak fractions were pooled and concentrated with polyethylene glycol to approximately half the volume and then dialyzed against DE buffer overnight. Following dialysis, the anti-sera were stored at -20°C in ali-quot parts. These anti-sera were absorbed to uninfected cells fixed in PF or air-dried and the dilution of each preparation giving optimal IP staining was determined using RV-infected Vero cells. K. Probe Preparation and Labelling E.coli, strain HB101, transformed with the vector pAT153 containing a B19 specific sequence was kindly sup-plied by Dr. MJ Anderson (9). The procedure for large scale plasmid isolation (16) was slightly modified. The plasmid containing bacteria were grown overnight at 37°C in 40ml Luria broth containing 12ug/ml tetracycline. The cells were pelleted at 7800g for 10 minutes in a Sorvall SS-34 rotor using 30ml Corex tubes. After resuspension in 2.5ml glucose/Tris/EDTA 5.0ml of 1% (w/v) SDS in 0.1M NaOH was added. After a 10 minute incubation on ice, 3.75 ml of potassium acetate (pH 4.8) was added and the solution was again in-cubated on ice for 10 minutes. Chromosomal DNA was pelleted by centrifuging at 7800g for 20 minutes. Two volumes of ice-cold absolute ethanol were added to the supernatant and the plasmid was allowed to precipitate overnight at -20°C. The plasmid was pelleted (7800g, 20 minutes), dried and resuspended in 1ml TE buffer. LiCl (0.5ml of 5M) was added to precipitate the RNA which was then pelleted by centrifuging at 14,000g for 5 minutes. One volume of isopropanol was added to the supernatant which was then incubated for 10 minutes at room temperature. The plasmid DNA was pelleted (14,000g, 5 minutes), dried and then resuspended in lOOul of Tris/EDTAbuffer (TE), re-precipitated with 250ul absolute ethanol, pelleted, dried and resuspended inTE. The plasmid preparation was restricted with Pst I (Pharmacia) and separated from the vector by electrophoresis in an agarose gel (1.2% w/v agarose in TAE with lmg/ml ethidium bromide) at 60V. The in-sert was isolated from a gel slice using GeneClean (BioCan Scientific, Mississauga, Ontario) which utilizes Nal to dissolve the agarose. In the presence of this salt, DNA binds to glass powder. The DNA is then washed to remove salt and then eluted into sterile distilled water by incubating at 50"C for 3-5 minutes (160). The purified insert was labelled via random priming with digoxigenin-ll-dUTP according to the instructions supplied by Boehringer Mannheim (Laval, Quebec). L. DNA Extraction and Dot Blots B19 DNA was extracted from the infected cells and culture supernatants by a modified method of Hirt (127). The virus particles were precipitated from the culture media by the addition of polyethylene glycol (MW 3500) to a concentration of 10% (w/v) and incubated for 4 hours at 4°C prior to centrifugation at 7800g for 20 minutes. The precipitated pellets were treated in the same manner as the cells. After washing in cold PBS 3 times, 1ml of lysis buffer was added and the viscous solution was transferred to a 15ml Corex tube and in-cubated at 50°C for one hour. The samples were then stored at -20°C so that all samples could be extracted at the same time. After thawing, an equal volume of phenol/chloroform (the latter was 24 parts chloroform and 1 part iso-amyl alcohol) was added and the phases were separated by centrifugation (12,000g, 15 minutes at 4°C). The aqueous layer was carefully removed and again treated with phenol/chloroform. The DNA was recovered by ethanol precipitation (2 volumes absolute ethanol for 20 minutes at -20°C). The precipitated DNA was washed 3 times with 70% ethanol and transferred to a 1.5ml microcentrifuge tube. After vacuum drying, the DNA was dissolved in lOOul sterile distilled water. The samples were denatured by heating to 70"'C for 30 minutes in 250mM NaOH. After neutralization with an equal volume of 2M ammonium acetate, the samples were applied to Hybond N nylon from Amersham (Oakville, Ontario) using a 96-well dot blot apparatus from Schleicher and Schull (Toronto, Ontario). After air drying, the DNA was cross-linked to the membrane by exposure to U.V. light (312nm) for 5 minutes. Pre-hybridization, hybridization, blocking and development were performed according to the instructions sup-plied with the Boehringer Mannheim kit. IV. RESULTS AND DISCUSSION A. Rubella Virus 1. Single time-point Infections Initially, experiments were carried out to compare the replication of eight strains of RV in Vero cells, FC and SMC cells. Each cell-type was infected for three or four days and a comparison was made of the degree of CPE and antigen expression and the yield of progeny virus. a. Antigen expression All strains of RV studied expressed viral proteins in all three cell types but variations were seen as shown in Figures 1-3. Considerable variability was found between the strains with respect to both the degree of CPE in-duced and antigen expression. Figure 1 represents Vero cells infected with eight strains of RV and it shows that while all strains expressed viral protein, those which induced the most severe CPE also expressed the most protein (HPV77/DE5, Therien, M33 and Thomas). Of particular note is M33 which induced CPE that is very characteristic of this strain. Heavily stained, star-shaped cells were observed in all cell-types (B2 in Figures 1-3). Figures 2 and 3 indicate that some strains, most notably the vaccine strains, were restricted in the cells of joint origin. Fetal chondrocytes infected with Cendehill, RA27/3, Thomas and 5209R demonstrated little or no CPE but the pattern of antigen staining was different The two wild-type strains (Thomas and 5209R) exhibited a granular stain in contrast to the diffuse stain observed with the vaccine strains. In addition, this diffuse stain was located in the perinuclear region of the Cendehill and RA27/3 in-fected cells. The remaining strains all demonstrated moderate to severe CPE and granular staining patterns throughout the cytoplasm. The results for infected SMC cells (Figure 3) were generally similar. One difference is that HPV77/DE5 demonstrated more perinuclear stain in the SMC cells compared to the FC. If examined at earlier times, other strains also exhibited a predominandy perinuclear stain, but by 4 days p.i. antigen expression progressed throughout the cytoplasm (data not shown). This perinuclear staining pattern may represent the passage of certain viral proteins through the Golgi complex where membrane proteins are glycosylated, but double labelling for both RV antigen and Golgi markers would be required to confirm this. The difference found between the strains in the nature of the antigen expression may represent an alteration 35 Figure 1. Antigen Expression of Rubella Virus Strains in Vero Cells Subconfluent monolayers of Vero cells were infected at an MOI of 5-10 for 4 hours at 35°C and were then in-cubated for 4 days (except for M33 which was incubated for 3 days). The plates were fixed in PF and ex-amined by IP staining using rabbit anti-Therien polyclonal antiserum as the primary antibody except for M33 and 1B2 which were examined with rabbit anti-RA27/3. Positive staining is achieved for all strains with either of the antisera. Al —Cendehill A2 —RA27/3 A3 —HPV77/DE5 Bl—Therien B2 —M33 B3 —1B2 CI—Thomas C2 —5209R C3 — Uninfected Figure 2. Antigen Expression Rubella Virus Strains in Fetal Chondrocytes Subconfluent monolayers of FC were infected at an MOI of 5-10 virus for 4 hours at 35°C. After incubating for 4 days, the cells were fixed with PF and examined by IP staining using rabbit anti-RA27/3 polyclonal an-tiserum as the primary antibody. Al — Cendehill A2 —RA27/3 A3 —HPV77/DE5 Bl —Therien B2 — M33 B3 — 1B2 CI — Thomas C2 — 5209R C3 — Uninfected 3S 39 Figure 3. Antigen Expression of Rubella Virus Strains in Synovial Membrane/Cartilage Cells Subconfluent monolayer of SMC cells were infected at an MOI of 5-10 for 4 hours at 35°C. After incubating for 4 days the cells were fixed with PF and examined by IP staining using rabbit anti-RA27/3 polyclonal an-tiserum as the primary antibody. Al —Cendehill A2 —RA27/3 A3 —HPV77/DE5 BI—Therien B2 — M33 B3 — 1B2 CI — Thomas C2 — 5209R C3 — Uninfected 41 in the glycosylation process in Cendehill and RA27/3 resulting in retardation of the growth cycle and lower viral yields. b. Virus production The culture media from the infections described above were assayed for infectious virus and the results are given in Table I. Again strain variation was distinctly seen. The litres from the Vero cells represent virus production under ideal conditions in a permissive cell line, with which the joint-cell yields can be compared. The results for both FC and SMC cells were similar for each RV strain and in most cases the viral yield was lower than that obtained in Vero cells. However, whereas only up to a 50-fold difference in virus litre was seen in FC and SMC cells infected with wild-type strains and HPV77/DE5, viral yields obtained with the RA27/3 and Cendehill strains were up to 2400-fold less in cells of joint origin compared to Vero cells. Indeed, the levels of virus seen at 4 days p.i. were so low for these 2 strains that they may be explained as residual virus from the inoculum. 2. Viral growth curves A comparison of the growth curves of different strains of RV in SMC and Vero cells was carried out to deter-mine whether virus replicated more slowly in joint cells but would eventually attain the same yield. In addi-tion it was important to confirm whether Cendehill and RA27/3 replicated at all in these cells. If replication of virus is occurring, a lag period (eclipse), followed by an increase in supernatant virus should be observed. As the level of cell-association of a virus may have implications in disease pathogenesis, it was also of interest to examine intracellular virus litres. Figures 4 and 6 represent the supernatant and cell-associated virus levels for five RV strains in two cell types. Cendehill, RA27/3 and HPV77/DE5 were selected for study because they are the common vaccine strains, and clinical data are available concerning their arthritogenicity, and M33 and Therien were selected as examples of wild-type viruses. The primary SMC cells were used as they are heterogeneous cell populations that would more closely approximate the cells within an intact joint than the FC cell-line. The experiments were also done using Vero cells to provide a basis for assessing the level of viral restriction in the SMC cells. 42 Table I. Rubella Virus Production in Cells of Joint Origin The culture media from the infections described in Figures 1-3 were assayed for virus by microfocus assay (see Methods under Virus Titrations). 43 Rubella Virus Production In Cells of Joint Origin Virus Titre (FFU/mL) Rubella Strain Vero Cells Synovial Membrane Cartilage Cells Fetal Chondrocytes Cendehill 1.0 x 106 4.1 x 102 1.8 x 103 RA27/3 5.1 x 106 7.5 x 10* 1.2 x 104 HPV77/DE5 2.7 x 106 6.3 x10 5 2.9 x 105 Therien 4.6 x 106 7.6 x 105 1.8 x 106 M33 1.5 x10 6 1.0 x 106 1.4 x 106 1B2 5.0 x 106 5.2 x 105 1.7 x 105 Thomas 1.0 xlO 7 6.3 x10 5 1.5 x 105 5209R 1.0 x 106 4.6 x 105 2.0 x 104 44 Figure 4. Growth Curves of Supernatant Rubella Virus Strains in SMC and Vero Cells Growth curve experiments were performed as described in Methods. The curves were drawn using the average values from replicate experiments and the vertical bars represent 1 standard deviation (cm-1). The value on Day 0 represents the level of input virus.... _ 4-5 Growth Curves of Supernatant Rubella Virus Strains in SMC and Vero Cells H 1 1 1 1 • 1 • 1 o H 1 1 1 1 1 1 1 1 0 2 4 6 8 0 2 4 6 8 8 2H H 1 1 • 1 • 1 • 1 o H • 1 • 1 • 1 • 1 0 2 4 6 8 0 2 4 6 8 Vero Cells SMC Cells I »—i 1 1 ' 1 • i 0 2 4 6 8 X Axis = Time Post Infection (days) Y Axis = Virus Titre (LoglO FFU/mL) The supernatant virus levels shown in Figure 4 indicate that within the time period of acute infection (seven days), each strain of RVwas restricted in the SMC cells to varying degrees. In addition, the time at which this restriction became apparent varied. For Cendehill, there was a difference in virus titre of 60 fold between the different cell types and this was apparent by 2 days p.i. By the end of the experiment the difference had increased to over four orders of magnitude. Similarly, for RA27/3, the differences in viral titres between the two cell types became apparent by 2 days p.i. and the difference at the end of the experiment was 500 fold. The other strains also showed differences but they appeared later and were not as pronounced. The two wild-type strains and HPV77/DE5 demonstrated the type of growth curve expected in a permissive system. In con-trast, the curves for Cendehill and RA27/3 are such that de novo synthesis can neither be demonstrated nor excluded. For this reason the decay kinetics (fall in virus titre when incubated in the absence of cells) of these two strains were examined and the results are shown in Figure 5. For RA27/3 it can be seen that during the incubation, there is a steady decline in virus infectivity, whereas in the presence of SMC cells, the virus titre is stable for six days following an early decrease. This indicates that viral replication was occurring in these cells to a level that was equal to the rate of decay, resulting in a horizontal growth curve. For Cen-dehill, however, the situation was not as clear. The decay curve shows that a steady decline to the limit of detection (10 FFU/ml) occurs after five days of incubation. Likewise, the supernatant virus titre from in-fected SMC cells has fallen to this level by the same time p.i. However, the shape of the curve from SMC cells is different and together with subsequent results (see Persistence in SMC cells) indicates that very low levels of virus replication were occurring with this strain. Similar strain variations were seen when levels of cell-associated virus were examined (Figure 6). The most apparent differences were again observed with the Cendehill and RA27/3 strains where no intracellular infec-tious virus above the limit of detection (10 FFU/ml) was observed. The cell-associated virus levels for the other three strains were also lower in the SMC cells as compared to the Vero cells but the differences were small and the fall in these titres with time was most likely due to cell death. The absence of detectable cell-as-sociated Cendehill and RA27/3 may indicate that these two strains utilize a different mechanism for obtain-ing their viral envelope in the SMC cells compared to the Vero cells. If these two strains acquired their 47 Figure 5. Decay Kinetics For Rubella Virus Strains Cendehill and RA27/3 Stock virus was incubated under the same conditions but in the absence of cells and aliquots were removed at the indicated times and titred. 48 Decay Kinetics for Rubella Virus Strains Cendehill and RA27/3 Virus Titre (log 10 FFU/ml) 1 0 Incubation Time (days) Virus Thre (log 10 FFU/ml) 1 0 Incubation Time (days) Figure 6. Growth Curves of Cell-Associated Rubella Virus Strains in SMC and Vero Cells Growth curve experiments were performed as described in Methods. The curves were drawn using the average values from replicate experiments and the vertical bars represent 1 standard deviation (an-1). 50 Growth Curves of Cell-Associated Rubella Virus Strains in SMC and Vero Cells Cendehill •f • 1 • 1 • 1 — • 1 0 2 4 6 6 RA27/3 0 2 4 6 8 51 envelopes exclusively from the plasma membrane (rather than from an internal membrane) in the SMC cells, then detecting a cell-associated, infectious virion would be a rare event. Figure 7 is a compilation of the data from Figures 4 and 6 to enable a direct comparison of RV strains in the SMC cells to be made. In the case of the cell-associated virus, the results split into two groups with Cendehill and RA27/3 showing no observable titre, and the other three strains exhibiting similar high viral yields. The supernatant virus litres also fell into the same groups with one representing the permissive viral strains (HPV77/DE5, Therien and M33) and those that were restricted (Cendehill and RA27/3). It is evident that al-though it is a vaccine strain, HPV77/DE5 behaved like a wild-type virus in this culture system. 3. Persistent infection in SMC cells The ability of RV to induce chronic arthritis suggests that the virus may be capable of persisting in joint tis-sue. For this reason the ability of the five strains to establish persistent infections in the SMC cells was inves-tigated. The results indicated that each strain was capable of persisting within these cells, but the characteristics of chronic infections were different. a. Virus production Viral yields obtained from SMC cells infected over a period of 80 days are shown in Figure 8. Similar to results obtained for the cell-associated virus during acute infections, there was a clear separation into two groups with HPV77/DE5 exhibiting a pattern like the wild-type strains although it showed much greater fluc-tuation than either M33 or Therien. The only infection that could not be maintained for 80 days was with the Therien strain Respite careful treatment designed to favour the cells, such as increased concentrations of FBS and more frequent media changes. The litres of both Cendehill and RA27/3 were dramatically lower (at least 1000 fold) compared to the wild-type group but it is clear that these cells were producing virus, al-though at low levels. b. Antigen expression Approximately once a week, the persistently infected cells were examined by IP staining and the results are displayed in Figures 9-13. The results for Cendehill and RA27/3 demonstrate the lower limit of detection of viral antigen by this technique. A positive staining pattern is lost by the third passage in Cendehill infected 52 Figure 7. Rubella Virus Strain Comparison in SMC Cells The data from Figures 4 and 5 (SMC cells only) has been compiled to enable a direct comparison of viral yields with different strains to be made. 5-3 Rubella Virus Strain Comparison in SMC Cells Virus Titre 4 (log 10 FFU/ml) -« 1 1 r 2 4 6 Time Post Infection (days) Cend RA HPV Therien M33 Virus Titre (log 10 FFU/ml) — Cendehill - * — RA27/3 -+— HPV77/DE5 -o— Therien - • — M33 Time Post Infection (days) Figure 8. Persistent Infection of SMC Cells with Rubella Virus Strains Supernatant samples for titration were collected every 3-4 days when the media was changed or when the cells were passaged. The titre of the input virus is shown on Day 0. Persistent Infection of SMC Cells Time Post Infection (days) Figure 9. SMC Cells Persistently Infected with Rubella Virus Strain Cendehill Approximately every 7 days one control and one infected tissue culture plate of persistently infected cells were fixed for IP staining and rabbit anti-RA27/3 polyclonal antiserum was used as the primary antibody. Antigen was not detected in cells from passages 4-8 and are therefore not shown. A—Control Pl B — Cendehill Pl C —Control P3 D — Cendehill P3 P # = passage # 57 Synovial Membrane/Cartilage Cells Persistently Infected with Rubella Virus Strain Cendehill Figure 10. SMC Cells Persistently Infected with Rubella Virus Strain RA27/3 Approximately every 7 days one control and one infected tissue culture plate of persistendy infected cells were fixed for IP staining and rabbit anti-RA27/3 polyclonal antiserum was used as the primary antibody. Antigen was not detected in cells from passages 5-8 and are therefore not shown. A—Control Pl B — RA27/3 Pl C —Control P3 D —RA27/3P3 E — Control P4 F — RA27/3 P4 P # = passage # Synovial Membrane/Cartilage Cells Persistently Infected Rubella Virus Strain RA27/3 Figure 11. SMC Cells Persistently Infected with Rubella Virus Strain HPV77/DE5 Approximately every 7 days one control and one infected tissue culture plate of persistently infected cells were fixed for IP staining and rabbit anti-RA27/3 polyclonal antiserum was used as the primary antibody. The appearance of the cells from passages 4-8 did not differ significantiy from those presented. A—Control Pl B — HPV77/DE5 Pl C —Control P3 D — HPV77/DE5 P3 E —Control P9 F — HPV77/DE5 P9 P # = passage # Synovial Membrane/Cartilage Cells Persistently Infected Rubella Virus Strain HPV77/DE5 61 Figure 12. SMC Cells Persistently Infected with Rubella Virus Strain Therien Approximately every 7 days one control and one infected tissue culture plate of persistently infected cells were fixed for IP staining and rabbit anti-RA27/3 polyclonal antiserum was used as the primary antibody. A—Control Pl B —Therien Pl C — Control P3 D — Therien P3 E —Control P4 F—Therien P4 P # = passage # Synovial Membrane/Cartilage Cells Persistently Infected Rubella Virus Strain Therien 63 Figure 13. SMC Cells Persistently Infected with Rubella Virus Strain M33 Approximately every 7 days one control and one infected tissue culture plate of persistently infected cells were fixed for IP staining and rabbit anti-RA27/3 polyclonal antiserum was used as the primary antibody. The appearance of the cells from passages 4-6 were not significandy different from those presented. A — Control Pl B — M33 Pl C —Control P3 C —M33P3 D —Control P7 E —M33P7 P # = passage # Synovial Membrane/Cartilage Cells Persistently Infected Rubella Virus Strain M33 65 66 cells and by the fourth passage in RA27/3 infected cells. In neither instance did the loss of antigen staining correlate with a decrease in virus titre, suggesting that at later times a greater number of cells were infected, each one producing very small numbers of virus particles, whereas in the beginning, there were fewer cells producing the same amount of virus thereby making antigen detection possible because it was concentrated in foci of infected cells. There was also no CPE induced by Cendehill during the 80 day infection, and RA27/3 only began to exhibit CPE after the seventh passage (data not shown). In contrast, the other three strains demonstrated considerable CPE from the first passage and this was maintained, with fluctuations, for the duration of the experiment. 4. Infection of joint tissue Intact joint tissue was infected with the five strains of RV to determine if they were capable of replicating in the relatively quiescent cells of organ culture. This would be closer to the in vivo situation where an infecting virus would encounter non-dividing cells and also extra-cellular matrices which might inhibit viral attach-ment. a. Virus production Figure 14 shows the results of virus titrations performed every three or four days on the culture media of seg-ments of infected joint tissue in organ culture over a period of 50-60 days. Once again HPV77/DE5 behaved like the wild-type strains Therien and M33, all three producing large amounts of virus for prolonged periods. The falling litres at later times p.i. were probably a result of cell death since the uninfected sections demonstrated that after 40 days in culture there were fewer viable cells (see Figure 15) as judged by the num-ber of intact nuclei (91). There was no evidence that Cendehill was able to replicate at all (limit of detection, 1FFU/ml) but RA27/3 showed that following a 16-day decline, low but stable titres were maintained for ap-proximately 50 days. b. Antigen expression As expected, no antigen was detected in the joint tissue infected with the Cendehill strain (data not shown). Despite the low levels of virus produced by RA27/3, sections positive for antigen were observed (Figure 16) but the foci of infection remained small and discrete. In contrast, extensive staining and CPE was seen for 67 Figure 14. Joint Tissue Infected with Rubella Virus Strains The culture supernatants were assayed for virus every 3-4 days when the media was changed. For each infec-tion control tissue derived from the same specimen was maintained in parallel. The media from all the con-trols was also assayed and found to be negative. TheJitre of the input virus is shown on Day 0. 69 Figure 15. Uninfected Joint Tissue Samples of control tissue were cultured, fixed, and examined in the same manner as the infected cultures. IP staining was performed using rabbit anti-Therien polyclonal antiserum as the primary antibody. A—0 days p.i. B — 7 days p.i. C — 21 days p.i. D — 60 days p.i. 0 days p.i. = fixed at the time of infection Figure 16. Infection of Joint Tissue with Rubella Virus Strain RA27/3 At the times indicated below, samples of infected tissue were fixed, sectioned and examined by IP staining using rabbit anti-Therien polyclonal antiserum as the primary antibody. A duplicate section treated in the same manner but without the primary antibody, is also presented. A—6daysp.i. B — Control for A C — 20 days p.i. D — Control for C E — 38 days p.i. F — Control for E Infection of Joint Tissue with Rubella Virus Strain RA27/3 73 HFV77/DE5, Therien and M33 (Figures 17-19). The antigen stain of these three strains was first observed in small foci on the edges of the synovial membrane (Figure 17A) which increased in size with time. All three strains demonstrated the ability to penetrate deeply into the synovial membrane, but they varied in the amount of tissue disruption produced, with M33 causing the most damage (Figure 19F). Variation was also observed regarding the ability of the different strains to express antigen within cartilage tissue. Even when the synovial membrane demonstrated extensive antigen stain, the cartilage was usually completely clear. In some cases a very few perichondrocytes fining the cartilage surface were found to express antigen. The ex-ception to this was with the Therien strain. Figure 19A shows how this strain is unable to penetrate the car-tilage at early times (7 days p.i.). Even by 28 days p.i. (Figure 20A) only perichondrocytes demonstrated Therien antigen. However, by 60 days p.i., a massive invasion of cartilage tissue had occurred (Figure 20C). 5. Summary and discussion The experiments outlined have attempted to answer the question of whether RV strains can replicate in cells and tissue of joint origin, thereby providing evidence for one possible mechanism of induction of RAA. For six out of eight strains (Cendehill and RA27/3 excluded) the answer was clear; the virus titres from FC and SMC cells at 3 or 4 days p.i. were similar to those from Vero cells and this was accompanied by strong an-tigen stain and/or CPE in the cells. To clarify the question concerning the replicative capabilities of Cen-dehill and RA27/3, the more detailed growth curve experiments were undertaken. These results indicated a lack of growth of these two strains since the supernatant virus levels did not increase. However, when the decay kinetics of each virus were determined i.e., fall in titre when incubated in the absence of cells, it ap-peared that there was viral growth but at low levels. The ability of RV strains to persist within joint cells and tissue is of interest because of the chronic nature of RAA. All strains (with the exception of Therien) were able to persist within the SMC cells for 80 days. How-ever, M33 and HPV77/DE5, induced severe CPE and careful monitoring was required to maintain the cells for the duration of the experiment. In contrast, RA27/3 and Cendehill induced little or no CPE but continued to produce levels of virus so low that they were undetectable by IP staining. These results, on their own, might suggest that Cendehill and RA27/3 would be the ideal candidates for causing chronic arthritis because they were able to persist in a non-destructive manner that might elicit a weak but chronic inflammatory Figure 17. Infection of Joint Tissue with Rubella Virus Strain HPV77/DE5 At the times indicated below, samples of infected tissue were fixed, sectioned and examined by IP staining using rabbit anti-Therien polyclonal antiserum as the primary antibody. A duplicate section treated in the same manner but without primary antibody, is also presented. A — 6 days p.i. B — Control for A C — 20 days p.i. D — Control for C E — 38 days p.i. F — Control for E 75 Infection of Joint Tissue with Rubella Virus Strain HPV77/DE5 Figure 18. Infection of Joint Tissue with Rubella Virus Strain Therien At the times indicated below, samples of infected tissue were fixed, sectioned and examined by IP staining using rabbit anti-Therien polyclonal antiserum as the primary antibody. A duplicate section treated in the same manner but without the primary antibody, is also presented. A—7 days p. i. B — Control for A C — 28 days p.i. E — 60 days p.i. D F — Control for C — Control for E 77 Infection of Joint Tissue with Rubella Virus Strain Therien A C E i - i r . . ' " , • " ; • * / * ' " ' . E: S I B D F Figure 19. Infection of Joint Tissue with Rubella Virus Strain M33 At the times indicated below, samples of infected tissue were fixed, sectioned and examined by IP staining using rabbit anti-Therien polyclonal antiserum as the primary antibody. A control section treated in the same manner but without primary antibody is also presented. A—7 days p.i. B — Control for A C —14 days p.i. D — Control for C E — 60 days p.i. F — Control for E Infection of Joint Tissue with Rubella Virus Strain M33 Figure 20. Penetration of Rubella Virus Strain Therien into Cartilage Tissue At the times indicated below, samples of infected tissue were fixed, sectioned and examined by IP staining using rabbit anti-Therien polyclonal antiserum as the primary antibody. A duplicate section treated in the same manner but without the primary antibody, is also presented. A—28 days p.i. B — Control for A C — 60 days p.i. D — Control for C 81 Penetration of Rubella Virus Strain Therien Into Cartilage Tissue response. However, clinical data indicate there is a much higher incidence of RAA following wild-type infec-tion compared with vaccination. In addition, the clinical data allow different vaccine strains to be placed in the following order of decreasing arthritogenicity: HPV77/DK12, HPV77/DE5, RA27/3, Cendehill and To336 (22, 53, 55,118). HPV77/DK12 was one of the first rubella vaccines used in the United States in 1970 but its use was discontinued due to the high number of associated complications (53). To336 is a vaccine used in Japan that appears to be highly effective and arguably the best candidate for future vaccine use in adult women, as it induces protective immunity with the fewest complications (22). The results of the experiments in joint tissue indicated that the different strains also varied in their ability to penetrate through the extracellular joint matrix. The matrix of cartilage consists of collagen (types II, DC, X, and other unclassified types) which is secreted by chondrocytes. In addition, these cells manufacture proteoglycans and structural glycoproteins. These secreted materials form a three dimensional structure that gives cartilage its elasticity and shock-absorbing qualities (15,51). One function of the synovial membrane is the production of synovial fluid which lubricates the joints which is accomplished by the B-type cells (126). These cells secrete some proteins and hyaluronic acid and proteoglycans surround the cells (91). The wild-type strains and HPV77/DE5 demonstrated invasive capabilities accompanied by increased virus titres, whereas replication of Cendehill was not observed. RA27/3 antigen was restricted to foci near the sur-face of the synovial membrane and thus demonstrated some replication but very limited ability to penetrate the tissue. The amount of virus produced was much lower than the "wild" group, tissue invasion was slower and less extensive, and no tissue disruption was seen. The five strains of RV that were studied in the most detail were selected because clinical data concerning their reaction in patients were also available. The results have shown that the behaviour of the three vaccine strains studied in joint tissue correlates with the reported incidence of complications with which they are as-sociated. More specifically, those vaccine strains which are associated with the most frequent and severe joint complications exhibited more aggressive behaviour in this model system. HPV77/DE5 produced levels of virus comparable to the wild-type strains in all infections but the extent of CPE and tissue disruption were 83 not as extensive. Cendehill was less productive than RA27/3 in the growth curve experiments and in addi-tion, it was unable to replicate in joint tissue. The inability of Cendehill to infect joint tissue may be explained in part by its restriction in SMC cells resulting in low viral yields incapable of overcoming the further obstruc-tion provided by the extracellular matrix of joint tissue. Infection with wild-type RV is associated with the highest levels of joint symptoms (55) and both wild-type strains studied demonstrated the most aggressive behaviour. SMC cells persistendy infected with Therien could not be maintained beyond 45 days despite careful maintenance and both Therien and M33 caused con-siderable damage to joint tissue. It is known that there is variation between different wild-type strains with respect to the incidence of joint complications in separate outbreaks of rubella, but in general wild-type infec-tion induces the highest level and most severe symptoms. Unfortunately, HPV77/DK12 and To336 were not available for study but it would be of interest to see if the correlation is maintained with these other vaccine strains. In conclusion, this study has shown that direct infection of joint tissue is a possible mechanism of induction of arthritis by RV. Moreover, a model system to investigate the arthrotropicity of RV strains, which correlates well with known clinical data, has been established. This system should prove valuable in estimating the ability of future vaccine strains to cause joint damage. B. Adenovirus To determine the permissiveness of joint cells to adenovirus (Ad), initial experiments involved analysis of FC and SMC cells at a single time-point for viral antigen expression and yield. The cells were infected in the same manner as Hep-2 cells, incubated for four days and fixed with PF prior to IP staining. The results are represented in Figures 21 and 22. At first the FC and SMC were thought to be totally non-permissive to Ad as the culture appeared to express no antigen at all (Figures 21D and 22B). However, closer examination iden-tified occasional cells, approximately 1 in 100, which expressed viral protein and exhibited a characteristic cytopathology (Figures 21B and 22C). By this time, in infected Hep-2 cells, every cell would be expressing an-tigen in addition to CPE involving rounding up and clumping of the cells prior to their detachment from the surface. Figure 21. Adenovirus Infected Fetal Chondrocytes Subconfluent monolayers of FC were infected with Ad at an MOI of 5-10 for 60 minutes at 37°C. After in-cubating for 4 days, the cells were fixed with PF and examined by IP staining using a monoclonal antibody to an Ad group specific antigen as the primary antibody. A—Uninfected B — 4 days p.i. C — 4 days p.i., no primary antibody D — 4 days p.i. Adenovirus type 2 Infected Fetal Chondrocytes 86 Figure 22. Adenovirus Infected SMC Cells Subconfluent monolayers of FC were infected with Ad at an MOI of 5-10 for 60 minutes at 37°C. After incubat-ing for 4 days, the cells were fixed with PF and examined by IP staining using a monoclonal antibody to an Ad group specific antigen as the primary antibody. A—Uninfected B — 4 days p.i. C — 4 days p.i. Adenovirus type 2 Infected Synovial Membrane/Cartilage Cells 87 A c B 1. Viral growth curves Since the preliminary experiments demonstrated that Ad was restricted in SMC cells and FC, more detailed time course experiments were performed to examine antigen expression and virus production in the two cell types. a. Antigen expression SMC and Hep-2 cells infected with Ad were fixed for IP staining daily for seven days. Two early time-points, immediately following virus adsorption and 0.5 days p.i., were also examined so that the onset of viral protein expression would be detected. Figure 23 depicts the progression of antigen expression and CPE in Hep-2 cells. By 2 days p.i. virtually every cell was expressing antigen, and 24 hours later early CPE was observed. By 5 days p.i. total destruction of the monolayer had occurred. In comparison, the restriction of Ad infection in SMC cells is shown in Figure 24. A low number of single cells expressing viral antigen could be detected 2 days p.i. Spread to surrounding cells occurred, forming foci of infection which slowly enlarged, but widespread infection of the monolayer did not occur. In addition, the CPE induced in the SMC cell cultures was very different from that seen with the Hep-2 cells which became more refractile, round and eventually detached. In contrast, the antigen posi-tive SMC cells remained attached and many of these developed large vacuolated nuclei of unusual shape (Figures 21 and 22). This type of effect was also seen in the Hep-2 cultures, at early times p.i. (Figure 23). b. Virus production At the same time that cells were fixed for IP staining, supernatant and cell-associated virus titres were also determined. Virus titres from each cell type are depicted in Figure 25. This shows that SMC cells were capable of supporting complete Ad infection, but that they were not as permissive as the Hep-2 cells. The levels of supernatant and cell-associated virus did not vary appreciably for a single cell type but the titres were approximately 10-fold less in the SMC cells than the Hep-2 cells. Cell-associated virus titres beyond 5 days p.i. from the Hep-2 cells were not obtainable because the monolayer was destroyed by this time. How-ever, since this virus is stable in culture media (Figure 26), supernatant virus titres were obtained for later times p.i. Figure 23 Progression of Adenovirus Antigen Expression in Hep-2 Cells Subconfluent monolayers of Hep-2 cells were infected with Ad at an MOI of 5-10 for 60 minutes at 37"C. At the indicated times, cells were fixed for IP staining using a monoclonal antibody to an Ad group specific an-tigen as the primary antibody. Al — 0 days p.i. A2 — 0.5 days p.i. A3 — 1 days p.i. BI —2 days p.i. . B2 — 3 days p.i. B3—4 days p.i. CI — 5 days p.i. C2 — 0 days p.i. C3—5 days p.i. Control Control 0 d.p.i. = cells were washed and fixed immediately following virus adsorption 91 Figure 24. Progression of Adenovirus Antigen Expression in SMC Cells Subconfluent monolayers of SMC cells were infected with Ad at an MOI of 5-10 for 60 minutes at 37°C. At the indicated times, cells were fixed for IP staining using a monoclonal antibody to an Ad group specific antigen as the primary antibody. Al — 0 days p.i. A2 — 0.5 days p.i. A3 — 1 day p.i. A4 — 2 days p.i. BI — 3 days p.i. B2 — 4 days p.i. B3 — 5days p.i. B4 — 6 days p.i. Cl — 7 days p.i. C2 — 0 days p.i. C3 — 3daysp.i. C4 — 7 days p.i. Control Control Control 0 d.p.i. = cells were washed and fixed immediately after virus adsorption Adenovirus type 2 Antigen Expression in Synovial Membrane/Cartilage Cells Figure 25. Growth Curves of Adenovirus in Hep-2 and SMC Cells Growth curve experiments were performed as described in Methods. The curves were drawn using the mean values of replicate experiments. Error bars represent 1 standard deviation (an-1). The value on Day 0 for the supernatant virus represents the titre of mputvirus. Growth Curves of Adenovirus in HEP-2 and SMC Cells Time Post Infection (days) Figure 26. Decay Kinetics of Adenovirus Ad was incubated under the same conditions but in the absence of cells. At the times indicated, samples were removed and titred. Adenovirus Decay Kinetics 10 Virus Titre (loglOTCID/ml) 8 H 4H 2H 10 20 Incubation Time (days) An estimation of the productivity (infectious particles/cell) may be calculated by dividing the total number of infectious particles by the total number of infected cells. This assumes that a cell must be expressing detec-table antigen to be producing virus because the number of infected cells is determined by counting positive cells following IP staining. To an approximation, it was found that at 4 days p.i. each infected Hep-2 cell was producing approximately 6000 infectious particles/cell whereas each infected SMC cell produced only about 400 infectious particles/cell. 2. Persistence in SMC cells The restricted nature of the Ad infection in the SMC cells suggested that this virus would readily establish persistent infection in these cells. However, after the first passage (Figure 27) the infection became much more aggressive with every cell expressing antigen. The CPE continued to increase until the fourth and final passage (data not shown) and this was accompanied by an increase in virus titre (Figure 28). 3. Infection of joint tissue Since the persistent infection in the SMC cells increased so dramatically after the first passage, it was con-cluded that Ad prefers cells that are mitotically active. For this reason it seemed very unlikely that this virus would be able to productively infect intact joint tissue since these cells are viable but not rapidly dividing. It was therefore surprising to observe the levels of virus produced in infected tissue (Figure 29) despite the fact that the culture media was changed every three to four days. There was considerable variability in the titres detected in separate infections which was likely due to variation between tissue specimens. In addition, the tissue from these experiments was sectioned and examined by IP staining for viral antigen using a monoclonal antibody. Even at late times p.i. (63 days) no Ad antigen was detected in any of the different tis-sues (data not shown). 4. Summary and discussion The infection of Ad in cells of joint origin proved to be very different from that observed in the permissive Hep-2 cell line. Figures 21 and 22 represent preliminary experiments that were at first considered to demonstrate a lack of antigen expression. However, on careful examination, occasional cells (approximately 1 in 100) which were heavily stained by IP were seen. It appeared that there was a block at some stage of 98 Figure 27. SMC Cells Persistently Infected with Adenovirus Approximately every 7 days one control and one infected tissue culture plate of persistently infected cells were fixed for IP staining using a monoclonal antibody to a group specific Ad antigen as the primary antibody. A—Control Pl B — Ad 9 d.p.i. C—AdPl D—AdP2 P# = passage* 99 Synovial Membrane/Cartilage Cells Persistently Infected With Adenovirus type 2 100 Figure 28. Persistent Infection of SMC Cells with Adenovirus When the media was changed or when the cells were passaged, samples of the supernatant were assayed for virus. The value on Day 0 represents the titre of the input virus. Time Post Infection (days) 102 Figure 29. Joint Tissue Infected with Adenovirus Samples of the culture media from infected and control cultures were assayed for virus every 3-4 days. Con-trol samples were all negative. The curves represent the results from separate infections where the value on Day 0 indicates the titre of the input virus. -Adenovirus Infected Joint Tissue 10 8 Virus Titre (log lOTCID/ml) 2H o H . 1 1 1 1 1 • 1 0 20 40 60 80 Time Post Infection (days) replication following viral protein translation which prevented the release of infectious progeny since the number of stained cells increased very slowly and mainly to adjacent cells suggesting spread by direct cell to cell contact. However, the growth curve experiments proved that infectious virus was produced and released in sufficient quantity to infect every cell. The long-term infections demonstrated that Ad could infect every cell but this did not occur until the cells were passaged. The procedure of trypsinizing, diluting and re-plating cells obviously causes changes in these cells that render them more permissive to infection. After any cells are passaged, there is a lag period of cell growth as they settle down and re-attach. During this time there is little or no increase in cell number but there are increases in metabolic activities such as DNA synthesis and cytoskeleton construction (61). All infections with Ad were initiated within 18 hours of passaging suggesting that infection is not cell-cycle dependent since at this time the cells are approximately synchronized (61). However, cell-cycle dependency cannot be entirely excluded as different cell types (SMC cells are a mixed population) are likely to have different cell-cycles. Another possible explanation for the rapid increase in infec-tion following the first passage is that during trypsinization existing cell-surface proteins were altered or new ones expressed that functioned as the viral receptor. Even if this alteration or expression were only transient, as virus is present throughout the procedure, infection of the cells might occur. In order to determine if infec-tion is truly cell-cycle dependent, more efficient methods of cell synchronization could be used such as DNA synthesis inhibition or nutritional deprivation (61), prior to infection. To more closely examine the role of trypsin, cells could be lightly trypsinized (not enough to cause them to detach) to determine if a viral recep-tor was being exposed by this treatment. The fact that high titres of virus were produced from infections of intact tissue shows that at least a few cells susceptible to infection were present as with the SMC cell cultures. The inability to detect these infected cells in the tissue sections may be due to several factors. There may be a very low number of infected cells producing large quantities of virus and they may not have been detected since only a few sections were ex-amined. If this were the reason, each piece of tissue would have to be completely sectioned and examined by IP staining in order to find the foci of infected cells. It is also possible that the antigenicity of the virus was al-tered by passage in joint tissue but this is unlikely since Ad antigen was undetectable even at early times before adaptation was likely to have occurred. In addition, when a piece of Ad infected tissue (63 days p.i.) 105 was co-cultivated with Hep-2 cells for several days, these cells exhibited the same antigen stain and CPE as was observed before (Figure 30). The most plausible explanation is that when the tissue pieces were processed for paraffin embedding and sectioning, the viral antigens were altered. Even a small alteration to the proteins, if located within the appropriate epitope, would be sufficient to prevent binding of the monoclonal antibody. If this were the case, then infected cells processed in the same manner as the tissue segments should also be negative when examined with the same antibody. In addition, the sections could be examined with a polyclonal antiserum which would be much less sensitive to small structural changes in in-dividual proteins. Adenovirus has clearly demonstrated that it can replicate within cells of the joint. The most likely mode of ac-cess to the joint for any virus is via the hematogenic route, possibly as passengers in circulating lym-phoreticular cells. Adenovirus replicates in human lymphoid cell lines (92,139) in a cell-cycle dependent manner (139). However, evidence to demonstrate its ability to replicate in primary human lymphocytes is scanty (79) and may explain why so few Ad infections are complicated by arthritis. Once it has reached the joint, the results described above indicate that a low number of cells may be suscep-tible to infection. It would seem reasonable that in a normal host, this limited infection would be cleared quickly with a relatively short inflammatory response, but in an immunocompromised host the infection may spread, resulting in prolonged inflammation associated with joint damage as has been found in a clinical case study (60). C. Human Parvovirus, B19 To date there is no in vitro system that will support complete replication of B19 (170), leaving serum from viremic patients as the only source of infectious virus. For this reason supplies of B19 are limited, and there-fore only a few experiments with this virus were possible. 1. Virus growth The ability of SMC cells to support the replication of B19 was examined by quantitating intracellular and supernatant viral DNA levels at three time-points p.i. by dot-blotting. Bone-marrow (BM) cells, infected in parallel were used as the positive control to ensure that the assay procedure was working. 106 Figure 30. Co-cultivation of Adenovirus Infected Joint Tissue with Hep-2 cells Two pieces of joint tissue infected with Ad (63 days p.i.) were incubated with subconfluent Hep-2 cells for 3 days. The tissue was discarded and the cells were washed, fixed and examined by IP staining using a monoclonal antibody to an Ad group specific antigen as the primary antibody. A—Uninfected B— Co-cultivation 107 Co-cultivation of Adenovirus type 2 Infected Joint Tissue with Hep-2 Cells 108 Figure 31 represents the results of a time course of B19 infected SMC cells extending over a four day period and it shows that SMC cells do not support B19 DNA replication to a level that is detectable in this system (Columns A and B). The sensitivity of the detection system was determined by extracting DNA from a known number of virus particles and applying dilutions of the samples obtained to nylon filters. Strips of these standards were used in each test. The sensitivity was found to be approximately 5 x 103 virus particles (M-II c) which is equivalent to approximately 14.5fg of DNA, indicating that while this is a sensitive system for DNA detection, it may not sufficient for detecting a virus with a small genome. A small amount of B19 DNA was seen in the cell-associated DNA preparation from SMC cells (Figure 31, BI). This may be due to the fact that virus particles were non-specifically adhering to the cell surface. Alternatively, it may represent virus which has been taken into the cells but is later degraded intracellularly as it is unable to replicate. The latter is more likely for 2 reasons. Firstly, the cells were washed once after the infection and twice before the DNA extraction. Secondly, if virus particles were attaching non-specifically to a cell it is likely that at least some of them would detach and be detected in the supernatant. The positive control for infection in this experiment was DNA extracted from infected BM cells. The amount of B19 DNA was shown to increase in both the supernatant media and cell-associated samples (Columns E and F) between 8 and 24 hours p.i. but was more readily detected in the latter. The slight increase in the amount of B19 DNA observed in the supernatant media (Column E) may not represent the release of progeny virus, but rather input virus that was released into the culture medium because its initial attachment to the cells was of low affinity. No B19 DNA was detected in uninfected SMC and BM cells (Columns C, D, G and H) indicating that the probe was specific and the hybridization conditions correct. 2. Antigen expression SMC cells infected with B19 for four days were also examined for antigen expression by IP staining after fixa-tion in PF. Antisera to B19 capsid and non-structural proteins were used as primary antibodies and both re-quired extensive absorbing to eliminate non-specific staining. Once this background staining was elirninated there was no evidence of antigen expression (data not shown). 109 Figure 31. Detection of B19 DNA from Infected SMC and BM Cells Experimental details are included in the Methods section. The time-points for infected SMC cells were 1 = Oh.p.i. 2 = 17h.p.i. 3 = 96h.p.i. The time-points for infected BM cells were 1 = 8h.p.i. 2 = 48h.p.i. 3 = 96h.p.i. Column A—SN B19 SMC B —CA B19 SMC C —SN Control SMC D —CA Control SMC SN = supernatant DNA, CA = cell-associated DNA Column E — SN B19 BM F —CA B19 BM G —SN Control BM H —CA Control BM M-II = source of infectious B19. Dilutions of extracted DNA served to determine the limit of detection and were equivalent to the number of genomes indicated below. a = 5xl05 b = 5 x 104 c = 5xl03 d = 5xl0z e = 5xl0 l 110 Detection of Human Parvovirus B19 DNA SMC BM B19 c 1 1 Big A 8 C D F 6 H 2 3 M i l £ £ • I c i t I l l 3. Summary and discussion The above results suggest that the mechanism of B19 induced arthritis is not through direct invasion of joint tissue since under conditions designed to favour the virus, it failed to replicate. It is conceivable that B19 an-tigen was not detected because of the fixative used (see rubella virus results), but this is unlikely since the more sensitive technique of DNA hybridization was also negative. These results are not necessarily con-tradictory to those of Dijkmans et al. (45) who detected B19 DNA in the synovial fluid of a woman with B19-associated arthritis (see Introduction). Rather, when taken together, they serve to suggest immune complex-mediated joint inflammation as the mechanism of B19-induced arthritis. 112 VI. CONCLUSION An in vitro system to examine the interaction of viruses with joint cells was established to investigate one of the postulated mechanisms of virus-induced arthritis, namely direct invasion and persistence of viruses in joint tissue. One of the advantages of using an in vitro system over an animal model, is that virus replication may be quantified. Also by examining infected tissue, the invasive capabilities of different viruses may be as-sessed and compared. The ability of a virus to replicate in cells and tissue of joint origin suggests that, given the opportunity (access to a joint), infection leading to inflammation may occur. An extension of the ability to infect, is the ability to persist which may be the cause of chronic viral arthritides. However, the ability to repli-cate does not negate the possible use of other mechanisms; a virus may utilize a variety of pathogenic path-ways which may vary with the individual. If, on the other hand, a virus is completely incapable of replicating in a system designed to favour its growth in joint tissue, then it would be more likely that immunopathogenic mechanisms are involved in its induction of arthritis. The results for RV strains indicated that the growth of each strain was restricted. The degree of restriction was quite pronounced for Cendehill and RA27/3 and present, but to a lesser degree for the other strains. This pattern of virus restriction according to strain is significant because it correlates with clinical data regarding the incidence of joint symptoms, suggesting that this may be a valuable tool for assessing the arthritogenicity of vaccine candidates. This model would then predict that RV strain HPV77/DK12 (a vaccine strain associated with severe side effects) would exhibit more aggressive behaviour in SMC cells or joint tis-sue than HPV77/DE5 and thatTo336 (the vaccine strain with the fewest reported clinical complications) would show a level of restriction equal to, or greater than, Cendehill. It would therefore be of great interest to examine the behaviour of these two strains to determine if the correlation is maintained. The results indicate that B19 probably utilizes an immunopathogenic mechanism to induce joint inflamma-tion since there was no indication of its replication in this system. It is interesting that RV and B19 behaved so differently in vitro when their clinical spectrums overlap so greatly. The most common result of infection with RV is an exanthematous diseases which may be clinically indistinguishable from that induced by B19. Likewise, the joint complications associated with both viruses share many clinical and epidemiological char-acteristics. Given the tremendous biochemical differences between RV and B19, perhaps it is more surpris-ing that they share anything in common at all. It is conceivable that low levels of B19 replication went undetected due to the limit of hybridization assay. It would therefore be useful to either examine these cultures with more sensitive techniques such as polymerase chain reaction, or to attempt to amplify the virus with multiple passaging of the cells. The behaviour of Ad in this model was intriguing but its significance in the in vivo situation is less clear. Its growth, which was initially restricted in the FC and SMC cells, seemed completely uninhibited after the cells were trypsinized suggesting that the virus receptor on these cells was either altered or exposed during the treatment. The literature reports of Ad-associated joint inflammation (see Introduction), together with the results presented here, suggest that Ad may cause arthritis by direct invasion of joint tissue particularly in im-munocompromised individuals. However, if faced with an intact immune system, is more likely to be as-sociated with the deposition of immune complexes. Given the prevalence of this virus in the community and the low incidence of joint involvement, it would seem that, in most cases, the host's defences are sufficient to prevent joint invasion. The potential value of this model culture system extends beyond its uses here. Firstly, it could be used to ex-amine the arthrotropicity of other viruses and secondly, the use of a heterogeneous cell population (SMC cells) may indicate if certain cells are preferentially infected. Thirdly, because it is a controlled system, the ef-fects of viral infection may be assessed in ways that are impossible to do in animals. For example, now that the growth patterns of three different viruses have been established, it will be possible to examine other viral effects such as the functioning of joint cells. For example, the levels of inflammatory mediators (interleukins and prostaglandins) or changes in cell structure and function (proteoglycan composition and collagen secre-tion) may have subtle effects on the maintenance of joint structure with long-term pathological consequences. 114 REFERENCES 1. Ager EA, Chin TDY and Poland JD. Epidemic erythema infectiosum. New England Journal of Medicine. 275:1326-1331,1966. 2. Ahmed R and Oldstone MBA Mechanisms and biological implication of virus-induced polyclonal B-cell activation. Chapter 33 in Concepts in Viral Pathogenesis. Notions AL and Oldstone MBA editors, Springer-Verlag, New York, 1984. 3. Alarcon GS and Castaneda 0. Arthritis associated with alphavirus infections. Chapter 16 in Infections in the Rheumatic Diseases. Espinoza L editor, Grune and Stratton, Toronto, 1988. 4. Alarcon GS. Hepatitis and arthritis. Chapter 13 in Infections in the Rheumatic Diseases. Espinoza L editor, Grune and Stratton, Toronto, 1988. 5. Alarcon GS. Rheumatoid arthritis. Chapter 28 in Rheumatology and Immunology., second edition. Cohen AS and Bennet JC editors, Grune and Stratton Inc., Toronto, 1986. 6. Anderson LJ. Role of parvovirus B19 in human disease. Pediatric Infectious Diseases. 6:711-718,1987. 7. Anderson MJ. Parvoviruses as agents of human disease. Progress in Medical Virology. 34:55-69,1987. 8. Anderson MJ. Human parvovirus infections. Journal of Virological Methods. 17:175-181,1987. 9. Anderson MJ, Jones SF, Minson AC. Diagnosis of human parvovirus infection by dot-blot hybridization using cloned viral DNA. Journal of Medical Virology. 15:163-172,1985. 10. Anderson MJ, Higgins LG, Davis LR, Williams JS, Jones SE, Kidd IM, Pattison JR and Tyrell DAJ. Experimental parvoviral infection in humans. Journal of Infectious Diseases. 152:257-265,1985. 115 11. Anderson MJ, Lewis E , Kidd IM, Hall SM and Cohen BJ. An outbreak of erythema infectiosum associated with human parvovirus infection. Journal of Hygiene. 93:85-93,1984. 12. Anderson MJ, Jones SE, Fisher-Hoch SP, Lewis E, Hall SM, Bartiett CLR, Cohen BJ, Mortimer PP and Pereira MS. Human parvovirus, the cause or erythema infectiosum (Fifth disease)? Lancet 1:1378, 1983. 13. Anderson MJ, Davis LR, Hodgson J, Jones SE, Murtaza L, Pattison JR, Stroud CE and White JM. Occurrence of infection with a parvovirus-like agent in children with sickle cell anaemia during a two-year period. Journal of Clinical Pathology. 35:744-745,1982. 14. Anonymous. Arthritis and parvovirus infection. Lancet. 1:436438,1985. 15. Arthritis Society, The. Basic Facts About Arthritis. 16. Ausubel FM, Brent R, Kingston RE, Moore DD, Seidman JG, Smith JA and Struhl K, editors, Current Protocols in Molecular Biology. Greene Publishing Associates, New York, p 1.7.1,1987. 17. Balfour HH. Fifth disease and arthritis: common immune mediated responses to parvovirus infection. New York State Journal of Medicine. 86:556-557,1986. 18. Banatvala JE and Best JM. Chapter 94 in Topley and Wilson's Principles of Bacteriology, Virology and Immunity., 7th edition. Wilson G, Miles A and Parker MT editors, Williams and Wilkins, Baltimore, 1984. 19. Bechar M, Davidovich S, Goldhammer G, Machley I and Gadoth N. Neurological complications following rubella infection. Journal of Neurology. 226:283-287,1982. 116 20. Benjamin T and Vogt PK. Cell transformation by viruses. Chapter 14 in Virology., second edition. Fields BN and Knipe DM editors, Raven Press Ltd., New York, 1990. 21. Berns KI. Parvoviridae and their replication. Chapter 62 in Virology., second edition. Fields BN and Knipe DM editors, Raven Press Ltd., New York, 1990. 22. Best JM, Banatvala JE and Bowen JM. New Japanese rubella vaccine: comparative trials. British Medical Journal. 3:221-224,1974. 23. Botchko GM, Armstrong RB, Martin WJ, and Ennis FA. Influenza A viruses of the H2N2 subtype are lymphocyte mitogens. Nature. 271:66-67,1978. 24. Brock TD. Biology of Microorganisms., third edition. Prentice-Hall Inc., New Jersey, 1979. 25. Butler JE. Mumps Arthritis. Chapter 15 in Infections in the Rheumatic Diseases. Espinoza L editor, Grune and Stratton, Toronto, 1988. 26. Chantler JK and Davies MA The effect of antibody on rubella virus infection of human lymphoid cells. Journal of General Virology. 68:1277-1288,1987. 27. Chantler JK, da Roza DM, Bonnie ME, Reid GD and Ford DK. Sequential studies on synovial lymphocyte stimulation by rubella antigen and rubella virus isolation in an adult with persistent arthritis. Annals of the Rheumatic Diseases. 44:564-568,1985. 28. Chantler JK, Tingle AJ and Petty RE. Persistent rubella virus infection associated with chronic arthritis in children. New England Journal of Medicine. 313:1117-1123,1985. 29. Chantler JK, Ford DK and Tingle AJ. Persistent rubella infection and rubella-associated arthritis. Lancet. 1:1323-1325,1982. 117 30. Chantler JK, Fork DK and Tingle AJ. Rubella-associated arthritis: rescue of rubella virus from peripheral blood lymphocytes two years post-vaccination. Infection and Immunity. 32:1274-1280,1981. 31. Chattopadhyay H, Chattopadhyay C, Natvig JB, Wiger D and Mellbye OJ. Demonstration of anti-rubella antibody secreting cells in rheumatoid patients. Scandinavian Journal of Immunology. 10:47-54, 1979. 32. Cohen BJ, Buckley MM, Clewley JP, Jones SE, Puttick AH and Jacoby RK. Human parvovirus infection in early rheumatoid and inflammatory arthritis. Annals of the Rheumatic Diseases. 45:832-838,1986. 33. Community Task Force on Maternal and Child Health. An immunization strategy for Manitoba: the prevention of congenital rubella and fetal damage. 1982. 34. Cooper LZ, Ziring RP, Weiss HJ, Matters BA and Krugman S. Transient arthritis after rubella vaccination. American Journal of Diseases of Childhood. 118:218-225,1969. 35. Cork LC, Hadlow WJ, Crawford TB, Gorham JR and Piper RC. Infectious leukoencephalomyelitis of young goats. Journal of Infectious Diseases. 129:134-141,1974. 36. Cossart YE, Cant B, Field AM and Widdows D. Parvovirus-like particles in human sera. Lancet. 1:72-73,1975. 37. Cotmore SF and Tattersall P. Characterization and molecular cloning of a human parvovirus genome. Science. 226:1161-1165,1984. 38. Crawford TB, Adams DS, Cheevers WR and Cork LC. Chronic arthritis in goats caused by a lentivirus. Science. 207:997-999,1980. 118 39. Cruickshank R, Duguid JP, Marmion BP and Swain RHA editors. Medical Microbiology., 12th edition, Volume II. Churchill Livingstone, New York, p 315,1975. 40. Cunningham AL and Fraser JRE. Persistent rubella virus infection of human synovial cells cultured in vitro. Journal of Infectious Diseases. 151:638-645,1985. 41. Dale JB and Beachey EH. Epitopes of streptococcal M proteins shared with cardiac myosin. Journal of Experimental Medicine. 162:583-591,1985. 42. Davis JS. Rheumatoid factors. Chapter 7 in Rheumatology and Immunology., second edition. Cohen AS and Bennet JC editors, Grune and Stratton Inc., Toronto, 1986. 43. Denman AM. The viral theory of connective tissue diseases: a review. Medical Biology. 56:61-84,1975. 44. Dighiero G, Guilbert B and Avrameas S. Naturally occurring antibodies against 9 common antigens in human sera. Journal of Immunology. 128:2788-2792,1982. 45. Dijkmans BAC, van Elsacker-Niele AMW, Salimans MMM, van Albada-Kuipers GA, de Vries E and Weiland HT. Human parvovirus B19 DNA in synovial fluid. Arthritis and Rheumatism. 31:279-281,1988. 46. Dorland's Pocket Medical Dictionary, 23rd edition WB Saunders Co. Toronto, 1982. 47. Dougados M, Amor B, Lefrere JJ and Courcouce AM. Human parvovirus arthropathy. Arthritis and Rheumatism. 29:575-76,1986. 48. Dykmans BAC, Breedveld FC and de Vries RRP. HLA antigens in human parvovirus arthropathy. The Journal of Rheumatology. 13:1192-1193,1986. 119 49. Ebringer A, Baines MJ and Ptaszynska T. Spondylarthritis, uveitis, HLA-B27 and Klebsiella. Immunological Reviews. 86:101-116,1985. 50. Everett MT. Parvovirus infection causing arthralgia. Journal of the Royal College of General Practitioners. 35:540,1985. 51. Fassbender HG. Connective tissue. Chapter 2 in Rheumatology and Immunology., second edition. Cohen AS and Bennet JC editors, Grune and Stratton Inc., Toronto, 1986. 52. Fenner F and White DO. Adenoviridae. Chapter 15 in Medical Virology., second edition. Academic Press, New York, 1976. 53. Fogel A, Gerichter CB, Barnea B, Handsher R and Heeger E. Response to experimental challenge in persons immunized with different rubella vaccines. Journal of Pediatrics. 92:26-29,1978. 54. Fong S, Tsoukas CD, Frincke LA, Lawrance SK, HolbrookTL, Vaughan, JH and Carson DA. Age-associated changes in EBV-induced human lymphocyte autoantibody responses. Journal of Immunology. 126:910-915,1981. 55. Ford DK, Tingle AJ and Chantler JK. Chapter 14 in Infection in the Rheumatic Diseases. Espinoza L editor, Grune and Stratton Inc., Toronto, 1988. 56. Ford DK, da Roza DM, Reid GD, Chantler JK and Tingle AJ. Synovial mononuclear cell responses to rubella antigen in rheumatoid and unexplained persistent knee arthritis. Journal of Rheumatology. 9:420423,1982. 57. Fourie ED and Morrison JGL. Rhuematoid arthritis syndrome after chikungunya fever. South African Medical Journal. 56:130-132,1979. 120 58. Fox JP, Rainey HA, Hall CE, Ray CG and Patterson MJ. Rubella vaccine in post-pubertal women. Journal of the American Medical Association. 236:837-843,1976. 59. Fox RI, Lotz M, Rhodes G and Vaughan JH. Epstein-Barr virus in rheumatoid arthritis. Clinics in Rheumatic Diseases. 11:665-688,1985. 60. Fraser KJ, Clarris BJ, Muirden KD, Fraser RE and Jack I. A persistent adenovirus type 1 infection in synovial tissue from an immunodeficient patient with chronic rheumatoid-like polyarthritis. Arthritis and Rheumatism. 28:455458,1985. 61. Freshney RI. Culture of Animal Cells. Alan R. liss Inc., New York, 1983. 62. Gibson M, Tiensiwakul P and Khoobyarian N. Adenovirus fiber protein (FP) functions as a mitogen and an adjuvant. Cellular Immunology. 73:397-403,1982. 63. Glass SE, Naqi SA, Hall CF and Kerr KM. Isolation and characterization of a virus associated with arthritis of chickens. Avian Diseases. 17:415-424,1973. 64. Goodman-Snitkov GW and McSharry JJ. Mitogenic activity of Sindbis virus and it isolated glycoproteins. Infection and Immunity. 38:1242-1248,1982. 65. Goodman-Snitkov GW and McSharry JJ. Activation of mouse lymphocytes by vesicular stomatitis virus. Journal of Virology. 35:757-765,1980. 66. Gould EA, Chanas AC, Buckley A, and Clegg GS. Monoclonal immunoglobulin M antibody to Japanese encephalitis virus that can react with a nuclear antigen in mammalian cells. Infection and Immunity. 41:774-779,1983. 121 67. Grahame R, Armstrong R, Simmons N, Wilton JMA, Dyson M, Laurent R, Millis R and Mims CA. Chronic arthritis associated with the presence of intra-synovial rubella virus. Annals of the Rheumatic Diseases. 42:2-13,1983. 68. Grahame R, Simmons NA, Wilton JMA, Armstrong R, Mims CA and Laurent R. Isolation of rubella virus from synovial fluid in five cases of seronegative arthritis. Lancet. 2:649-651,1981. 69. Graves RJ. Clinical Lectures on the Practise of Medicine. Fannen and Co., p 937,1843. 70. Green DH, Bellingham AJ and Anderson MJ. Parvovirus infection in a family associated with aplastic crisis in an affected sibling pair with hereditary spherocytosis. Journal of Clinical Pathology. 37:1144-1146,1984. 71. Green RH, Balsamo MR, Giles JP, Krugman S and Mirick GS. Studies of the natural history and prevention of rubella. American Journal of Diseases of Children. 110:348-365,1965. 72. Gregg NM. Transactions of the Opthalmological Society of Australia. 3:35-46,1941. 73. Haire M. Fibrillar anti-cellular antibody associated with mumps and measles infection. Clinical and Experimental Immunology. 12:335-341,1972. 74. Hamerman D. Evidence for an infectious etiology of rheumatoid arthritis. Annals of the New York Academy of Sciences. 256:25-38,1975. 75. Hart H and Marmion BP. Rubella virus and rheumatoid arthritis. Annals of the Rheumatic Diseases. 36:3-12,1977. 76. Haspel MV, Ondera T, Prabhakar BS, Horita M, Suzuki H and Notions AL. Virus-induced autoimmunity: monoclonal antibodies that react with endocrine tissues. Science. 220:304-306,1983. 122 77. Heggie AD and Robbins FC. Natural rubella acquired after birth: clinical features and complications. American Journal of Diseases of Children. 118:12-17,1969. 78. Hildebrandt HM and Maassab HF. Rubella synovitis in a one-year old patient. New England Journal of Medicine. 274:1428-1431,1966. 79. Horvath J and Weber J. Adenoviruses in human lymphocytes. Section 1 in Viral Mechanisms of Immunosuppression. Gilmore N and Wainberg MA editors, Alan R Liss Inc., New York, 1985. 80. Horwitz MS. Adenoviridae and their replication. Chapter 60 in Virology., second edition, Fields BN and Knipe DM editors. Raven Press Ltd., New York, 1990. 81. Huppert J and Wild RG. Virus related pathology: is the continued presence of the virus necessary? Advances in Virus Research. 31:357-385,1986. 82. Hyer FL and Gottlieb NL. Rheumatic disorders associated with viral infection. Seminars in Arthritis and Rheumatism. 8:17-31,1978. 83. Johnson RE and Hall AP. Rubella arthritis: report of cases studied by latex tests. New England Journal of Medicine. 258:743-745,1958. 84. Joseph PR Fifth disease: the frequency of joint involvement in adults. New York State Journal of Medicine. 86:560-563,1986. 85. Kagnoff MF, Austin RK, Hubert JJ, Bernardin JE and Kasarda DD. Possible role for a human adenovirus in the pathogenesis of celiac disease. Journal of Experimental Medicine. 160:1544-1557, 1984. 123 86. Kendall LG. Other viral arthritides. Chapter 18 in Infections in the Rheumatic Diseases. Espinoza L editor, Grune and Stratton, Toronto, 1988. 87. Kinseti E, Guilbert BJ, Bornens, M and Avrameas S. Antibodies to tubulin in normal nonimmunized animals. Proceedings of the National Academy of Science. 74:3997-4001,1977. 88. Klevjer-Anderson P and McGuire TC. Neutralizing antibody response of rabbits and goats to caprine arthritis-encephalitis virus. Infection and Immunity. 38:455461,1982. 89. Klouda PR, Corbin SA, Bradley BA, Cohen BJ and Woolf AD. HLA and acute arthritis following human parvovirus infection. Tissue Antigens. 28:318319,1986. 90. Koopman WJ. Immunopathogenetic mechanisms in viral disease. Chapter 11 in Infections in the Rheumatic Diseases. Espinoza L. editor, Grune and Stratton Inc., Toronto, 1988. 91. Krey PR, Cohen AS, Smith CB and Finland M. The human fetal synovium. Arthritis and Rheumatism. 14:319-341,1971. 92. Lavery D, Fu SM, Lufkin T and Chen-Kiang S. Productive infection of cultured human lymphoid cells by adenovirus. Journal of Virology. 61:1466-1472,1987. 93. Lefrere JJ, Courouce AM, Bertrand Y, Girot R and Soulier J. Human parvovirus and aplastic crisis in chronic hemolytic anemias: a study of 24 observations. American Journal of Hematology. 23:271-275, 1986. 94. Lefrere JJ, Meyer 0, Menkes CJ, Beaulieu MJ and Courouce AM. Human parvovirus and rheumatoid arthritis. Lancet. 1:982,1985. 124 95. Lerman SJ, Nankervis GA, Heggie AD and Gold E. Immunologic response, virus excretion and joint reactions with rubella vaccine. Annals of Internal Medicine. 74:67-73,1971. 96. London WT, Fuccillo DA, Anderson B and Sever JL Concentration of rubella virus antigen in chondrocytes of congenitally infected rabbits. Nature. 226:172-173,1970. 97. Loudon ISL. Polyarthritis in rubella. British Medical Journal. 1:1388,1953. 98. Lozovskaya LS, Desyatskova RG, Mitchenko AF, Pokryshkina EL, Shigina OM and Khairetdinova TB. Antibodies for rubella virus in blood sera and synovial fluid of children with rheumatoid arthritis. Voprosy Virusologii. 25:468473,1980. 99. MacAulay JC. Human parvovirus infection with arthropathy. British Journal of Dermatology. 116:883-884,1987. 100. Martenis TW, Bland JG and Phillips CA. Rheumatoid arthritis after rubella. Arthritis and Rheumatism. 11:683-687,1968. 101. Mims CA, Stokes A and Grahame R. Synthesis of antibodies, including anti-viral antibodies in the knee joints of patients with arthritis. Annals of the Rheumatic Diseases. 44:734-737,1985. 102. Monto AS, Cavallaro JJ and Whale EH. Frequency of arthralgia in women receiving one of three rubel-la vaccines. Archives of Internal Medicine. 126:635-639,1970 103. Nakazono N, Fujimoto S, Wakisaka A Ishii K, Ichinow K and Aizawa M. Factors associated with clinical reactions to rubella vaccination in women. International Journal of Gynaecology and Obstetrics. 25:207-2161987. 125 104. O'Sullivan FX. Viruses and animal models of arthritis. Chapter 12 in Infection in the Rheumatic Diseases. Espinoza L editor, Grune and Stratton Inc., Toronto, 1988. 105. Ogra PL, Ogra SS, Chiba Y, Dzierba JL and Herd JK. Rubella virus infection in juvenile rheumatoid arthritis. Lancet. 1:1157-1161,1975. 106. Ogra PL and Herd JK. Arthritis associated with induced rubella infection. Journal of Immunology. 107:810-813,1971. 107. Oldstone MBA. Viral Persistence. Cell. 56:517-520,1989. 108. Oldstone MBA and Dixon FJ. Immunopathology of viral infections. Chapter 16 in Textbook of Immunopathology., Volume I, second edition. Miescher PA and Muller-Eberhard HJ editors, Grune and Stratton, New York, 1976. 109. Orenstein WA, Bart SW, Bart KJ, Sirotkin B and Henman AR. Epidemiology of rubella and its complications. Chapter 5 in Vaccinating Against Brain Syndromes. Gruenberg EM, Lewis C and Goldston SE editors, Oxford University Press, New York, 1986. 110. Osier W. The Principles and Practice of Medicine., 6th edition, Appelton, New York. 1905. 111. Ozawa K, Kurtzman G and Young N. Replication of the B19 parvovirus in human bone marrow cell cultures. Science, 233:883-886,1986. 112. Panush RS. Adenovirus arthritis. Arthritis and Rheumatism. 17:534-536,1974. 113. Pattison J. Parvoviruses - medical and biological aspects. Chapter 63 in Virology., second edition. Fields BN and Knipe DM editors, Raven Press Ltd., New York. 1990. 114. 126 Pattison JR, Jones SE, Hodgson J, Davis LR, White JM, Stroud CE and Murtaza L. Parvovirus infections and hypoplastic crisis in sickle-cell anaemia. Lancet 1:664-665,1981. 115. Petty RE and Tingle AJ. Arthritis and viral infection. Journal of Pediatrics. 113:948-949,1988. 116. Philipson L, Pettersson U and Lindberg U. Molecular biology of adenoviruses. Volume 14 in Virology Monographs. Gard S and Hallaver C, editors, Springer-Verlag, New York, 1975. 117. Phillips PE. Infectious agents in the pathogenesis of rheumatoid arthritis. Seminars in Arthritis and Rheumatism. 16:1-10,1986. 118. Polk BF, Modlin JF, White JA and DeGirolami PC. A controlled comparison of joint reactions among women receiving one of two rubella vaccines. American Journal of Epidemiology. 115:19-25,1982. 119. Porterfield JS. Antibody-dependent enhancement of viral infectivity. Advances in Virus Research. 31:335-355,1986. 120. Rager-Zisman B. Virus-induced autoimmunity. Israel Journal of Medical Sciences. 24:363-365,1988. 121. Rahal JJ, Millian SJ, Noriega ER. Coxsackie virus and adenovirus infection. Association with acute febrile and juvenile rheumatoid arthritis. Journal of the American Medical Association. 235:2496-2501, 1976. 122. Rawl, WE and Melnick, JL. Rubella virus carrier cultures derived from congenitally infected infants. Journal of Experimental Medicine. 123:795-816,1966. 123. Reid DM, Brown T, Reid TMS, Rennie JAN and Eastmond CJ. Human parvovirus associated arthritis: a clinical and laboratory description. Lancet. 1:422425,1985. 127 124. Rhode SL. Replication process of the parvovirus H-l. Journal of Virology. 11:856-861,1973. 125. Rickinson AB, Yao QY and Wallace LE. The Epstein-Barr virus as a model of virus-host interactions. British Medical Bulletin. 41:75-79,1985. 126. Rodnan GP and Schumacher HR editors. Primer on the Rheumatic Diseases., eighth edition. Arthritis Foundation, Atlanta, 1983. 127. Rosenthal LJ, Crutchfield DB, Panitz PJ and Clanton DJ. Isolation of human cytomegalovirus DNA from infected cells. Intervirology. 19:113-1209,1983. 128. Rowlands S. Arthralgia from parvovirus infection. Journal of the Royal College of General Practitioners. 36:288-290,1986. 129. Saarinen UM, Chorba TL, Tattersall P, Young NS, Anderson LJ, Palmer E and Coccia PF. Human parvovirus B19-induced epidemic acute red cell aplasia in patients with hereditary hemolytic anemia. Blood. 67:1411-1417,1986. 130. Schlesinger S and Schlesinger MJ. Replication of Togaviridae and Flaviviridae. Chapter 25 in Virology., second edition. Fields BN and Knipe DM editors, Raven Press Ltd., New York, 1990. 131. Schnitzer TJ. Viral Arthritis. Chapter 98 in Textbook of Rheumatology., second edition. Kelly WN, Harris ED and Ruddy S editors. WB Saunders Co., Philadelphia, 1985. 132. Schumacher AR and Gall EP. Arthritis in acute hepatitis and chronic active hepatitis. American Journal of Medicine. 57:655-664,1974. 133. Semble EL, Agudelo CA and Pegram PS. Human parvovirus B19 arthropathy in two adults after contact with childhood erythema infectiosum. The American Journal of Medicine. 83:560-562,1987. 128 134. Serjeant GR, Mason K, Topley JM, Serjeant BE, Pattison JR, Jones SE and Mohamed R. Outbreak of aplastic crises in sickle cell anaemia associated with parvovirus-like agent Lancet. 2:595-597,1981. 135. Sever JL, Hardy JB, Nelson KB and Gilkeson MR. Rubella in the collaborative perinatal research study. American Journal of Diseases of Children. 118:123-132,1969. 136. Sheshberadaran H and Norrby E. 3 monoclonal antibodies against measles virus F protein cross-react with cellular stress proteins. Journal of Virology. 52:995-999,1984. 137. Shoenfeld Y and Cohen IR Infection and autoimmunity. Chapter 4 in Hie Antigens., volume VII. Sela M editor, Academic Press, Toronto, 1987. 138. Shumaker JB, Goldfinger SE, Alpert E and Isselbacher KJ. Arthritis and rash. Archives of Internal Medicine. 133:483485,1974. 139. Silver L and Anderson CW. Interaction of human adenovirus serotype 2 with human lymphoid cells. Virology. 165:377-387,1988. 140. Simpson RW, McGinty L, Simon L, Smith CA Godzeski CW and Boyd RJ. Association of parvoviruses with rheumatoid arthritis of humans. Science. 223:1425-1428,1984. 141. Singh VK, Tingle AJ and Schulzer M. Rubella-associated arthritis. II. Relationship between circulating immune complex levels and joint manifestations. Annals of the Rheumatic Diseases. 45:115-119,1986. 142. Sinkovics JG and Harris JE. Immunology and immunotherapy of human tumors. Chapter 5 in The Immunology of Malignant Disease., second edition. Harris JE and Sinkovics JG, editors, CB Mosby Co., St Louis, 1976. 129 143. Sinkovics, JG. Immunology of tumors in experimental animals. Chapter 2 in The Immunology of Malignant Disease., second edition. Harris JE and Sinkovics JG editors, CB Mosby Co., St Louis, 1976. 144. Smith JL, Early EM, London WR Fuccillo DA and Sever JL. Persistent rubella virus production in embryonic rabbit chondrocyte cell cultures. Proceedings of the Society for Experimental Biology and Medicine. 143:1037-1041,1973. 145. Soboleva VD, Lozovskaya LS and Mitchenko AF. Viremia in children with rheumatoid arthritis. Voprosy Virusologii. 28:207-210,1983. 146. Spruance SL, Metcalf R, Smith CB, Griffiths MM and Ward JR Chronic arthropathy associated with rubella vaccination. Arthritis and Rheumatism. 20:741-747,1977. 147. Stierle G, Brown KA, Rainsford SG, Smith CA, Hamerman D, Stierle HE and Dumonde DC. Parvovirus associated antigen in the synovial membrane of patients with rheumatoid arthritis. Annals of the Rheumatic Diseases. 46:219-223,1987. 148. Summers J, Jones SE and Anderson MJ. Characterization of the genome of the agent of erythrocyte aplasia permits its classification as a human parvovirus. Journal of General virology. 64:2527-2532,1983. 149. Swartz TA, Klinberg W, Goldwasser RA, Minberg MA, Goldblum N and Hilleman MR Clinical manifestations according to age among females given HPV77 duck rubella vaccine. American Journal of Epidemiology. 94:246-250,1971. 150. Syrjanen S and Syrjanen K. An improved in situ DNA hybridization protocol for detection of human papilloma virus (HPV) DNA sequences in paraffin-embedded biopsies. Journal of Virological Methods. 14:293-304,1986. 151. Tattersall P. Replication of the parvovirus MVM. Journal of Virology. 10:586-590,1972. 130 152. Tesh RB. Arthritides caused by mosquito-borne viruses. Annual Reviews of Medicine. 33:31-40,1952. 153. Thurn J. Human parvovirus B19: historical and clinical review. Reviews of Infectious Diseases. 10:1005-1011,1988. 154. Tingle AJ, Allen M, Petty RE, Kettyls GD and Chantler JK. Rubella-associated arthritis. I. Comparative study of joint manifestations associated with natural rubella infection and RA27/3 rubella immunisation. Annals of the Rheumatic Diseases. 45:110-114,1986. 155. Tingle AJ, Chantler JK, Pot KH, Paty DW and Ford DK. Postpartum rubella immunization: association with development of prolonged arthritis, neurologic sequelae and chronic rubella viremia. Journal of Infectious Diseases. 152:606-612,1985. 156. Utsinger, PD and Katz WA. Rheumatoid arthritis-diagnosis. Chapter 29 in Diagnosis and Management of Rheumatic Diseases., second edition. Katz WA editor, JB Lippincott Co., New York, 1988. 157. Utsinger RD. Immunologic study of arthritis associated with adenovirus infection. Arthritis and Rheumatism. 20:138,1977. 158. van Berlo MF, Spaan WJM and Morzinek MC. Replication of equine arteritis virus (EAV): a comparative review. Chapter 7 in The Molecular Biology of the Positive Strand RNA Viruses. Rowlands DJ, Mayo MA and Mahy BWJ editors, Academic Press, Toronto, 1987. 159. Vergani D, Morgan-Capner P, Davies ET, Anderson AW, Tee DEH and Pattison JR. Joint symptoms, immune complexes and rubella. Lancet 2:321-322,1980. 160. Vogelstein B and Gillespie D. Preparative and analytical purification of DNA from agarose. Proceedings of the National Academy of Science. 76:615-619,1979. 161. Wadlington WB and Riley HD. Arthritis and hemolytic anemia following erythema infectiosum. Journal of the American Medical Association. 203:473-475,1968. 162. Wardley RC. Effect of African swine fever on lymphocyte mitogenesis. Immunology. 46:215-220,1982. 163. Ware R. Human parvovirus infection. Journal of Pediatrics. 114:343-348,1989. 164. Waxham MN and Wolinsky JS. Rubella virus and its effects on the central nervous system. Neurologic Clinics. 2:367-385,1984. 165. Weibel RE, Stokes J, Buynak EB and Hilleman MR Live rubella vaccines in adults and children. American Journal of Diseases of Children. 118:226-229,1969. 166. Whaley K. Complement in immune complex diseases. Chapter 7 in Complement in Health and Disease. Whaley K editor, MTP Press Ltd., Lancaster, 1987. 167. White DG, Mortimer PP, Blake DR, Woolf AD, Cohen BJ and Bacon PA. Human parvovirus arthropathy. Lancet. 1:419421,1985. 168. Wilson MR and Tan EM. Anti-nuclear antibodies and the LE cell phenomena. Chapter 8 in Rheumatology and Immunology., second edition. Cohen AS and Bennet JC editors, Grune and Stratton Inc., Toronto, 1986. 169. Wolinsky JS. Rubella. Chapter 28 in Virology., second edition. Fields BN and Knipe DM editors, Raven Press Ltd., New York, 1990. 170. Young N, Harrison M, Moore J, Mortimer P and Humphries RK. Direct demonstration of the human parvovirus in erythroid progenitor cells infected in vitro. Journal of Clinical Investigation. 74:2024-2032, 1984. 132 171. Yu D.T.Y., Choo S.Y. and ShaackT. Molecular mimicry in HLA-B27 related arthritis. Annals of Internal Medicine. 111:581-589,1989. 172. Ziola B, Lund G, Meurman 0 and Salmi A Circulating immune complexes in patients with acute measles and rubella virus infections. Infection and Immunity. 41:578-583,1983. 


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items