Retroviral sequences in rheumatoid arthritis synovium

Viruses and autoimmunity

Viruses have been suspected as causes and contributors of human autoimmune diseases (AID), although direct evidence for the association is lacking. However, several animal models provide strong evidence that viruses can induce AIDs as well as act to accelerate and exacerbate lesions in situations where self-tolerance is broken. Many models support the hypothesis by acting as molecular mimics that stimulate self-reactive lymphocytes. Mimicry alone is usually inadequate and with human AID, no compelling evidence supports a role for viruses that are acting as molecular mimics. Alternative mechanisms by which viruses participate in autoimmunity are non-specific, involving a mechanistically poorly understood process termed bystander activation or perhaps viral interference with regulatory cell control systems. This review briefly discusses examples where viruses are involved, taking the viewpoint that molecular mimicry is over emphasized as a critical mechanism during AID pathogenesis.

Viruses and autoimmunity

Viruses have been suspected as causes and contributors of human autoimmune diseases (AID), although direct evidence for the association is lacking. However, several animal models provide strong evidence that viruses can induce AIDs as well as act to accelerate and exacerbate lesions in situations where self-tolerance is broken. Many models support the hypothesis by acting as molecular mimics that stimulate self-reactive lymphocytes. Mimicry alone is usually inadequate and with human AID, no compelling evidence supports a role for viruses that are acting as molecular mimics. Alternative mechanisms by which viruses participate in autoimmunity are non-specific, involving a mechanistically poorly understood process termed bystander activation or perhaps viral interference with regulatory cell control systems. This review briefly discusses examples where viruses are involved, taking the view point that molecular mimicry is over emphasized as a critical mechanism during AID pathogenesis.

Retrotransposable L1 elements expressed in rheumatoid arthritis synovial tissue: association with genomic DNA hypomethylation and influence on gene expression

OBJECTIVE

Rheumatoid arthritis (RA) is characterized by a progressive destruction of joints by invasive synovial fibroblasts (SF). We searched for retroviral sequences in RA synovial fluid pellets, identified a sequence similar to that of open reading frame 2 (ORF2)/L1 retrotransposable elements, explored the expression of L1 in RA synovial tissues and cultured RA SF, and investigated the link to genomic DNA hypomethylation and the influence of functional L1 on gene expression.

METHODS

RA synovial fluid pellets were screened by reverse transcriptase-polymerase chain reaction (RT-PCR) using degenerated pol primers. The sequences were identified by GenBank search. Riboprobes to ORF2/L1 and galectin-3 and antibodies to the ORF1/L1-related p40 protein were used for in situ hybridization and immunohistochemistry of synovial tissues and cultured RA SF. Real-time quantitative RT-PCR was used for detecting ORF1 messenger RNA (mRNA). Since DNA hypomethylation occurs in inflammatory diseases, we incubated cells with the methylation inhibitor 5-aza-2'-deoxycytidine (5-azaC) and compared RA SF and osteoarthritis (OA) SF. L1-negative RA SF were transfected with the functional L1.2 construct, and differential gene expression was analyzed by subtractive hybridization combined with nested PCR.

RESULTS

RNA sequences similar to those of ORF2/L1 retrotransposable elements, THE1 transposon, human endogenous retrovirus (ERV)-E, human ERV-HC2, and gibbon ape leukemia virus pol genes were isolated from different RA synovial fluid pellets. In RA synovial tissues, ORF2/L1 transcripts were detected in the sublining layer and at sites of cartilage and bone destruction. Galectin-3 mRNA and L1-related ORF1/ p40 protein showed similar expression patterns. In contrast, OA synovial tissues in situ and cultures in vitro were negative. Real-time quantitative RT-PCR confirmed the presence of ORF1 mRNA in cultured RA SF (30-300-fold the amount in normal SF), demonstrating the existence of a nondegenerated and functional L1 element. In vitro, the majority of RA SF expressed ORF2/L1 mRNA. After incubation of SF with 5-azaC, L1 mRNA appeared in a time- and dose-dependent manner. Compared with OA SF, RA SF were more sensitive to 5-azaC. After transfection of RA SF with a functional L1.2 element, human stress-activated protein kinase 2 delta (SAPK2delta [or SAPK4]), met protooncogene, and galectin-3 binding protein genes were differentially expressed. The transcription of the SAPK2delta gene, favored also by DNA hypomethylation in vitro, was confirmed in RA synovial tissues.

CONCLUSION

Taken together, these data suggest that L1 elements and SAPK2delta pathways play a role in the activation of RA SF.

HLA association with autoimmune disease: a failure to protect?

That certain HLA specificities are associated with predisposition to autoimmune disease does not necessarily imply that self-reactive T cells restricted to particular HLA alleles are eliciting the disease. In the present essay, we argue that HLA can be a major genetic factor in the development of autoimmune diseases without T cells being primarily involved in its initiation or perpetuation. There is now ample evidence that self-reactive, regulatory T cells can protect against pernicious autoimmunity. Hereafter, we propose that extended HLA haplotypes, such as DQ3-DR4, DQ3-DR9, DQ5-DR1 and DQ5-DR10 in the case of rheumatoid arthritis, predispose to impaired T-cell-mediated immune regulation. The haplotypes associated with impaired regulation are the combination of certain class II alleles and a yet unknown 'amplifier'. In this model, products of the HLA class II region are not involved in the presentation of particular organ-specific autoantigens. Therefore, HLA does not predispose to autoimmune disease per se, but rather fails to provide efficient protection.

Activation of synoviocytes

The evaluation of molecular pathways has revealed various novel insights into rheumatoid arthritis pathophysiology during the past year. In addition, there is an increasing tendency toward analysis not merely of a single mechanism but rather of data addressing a substantial part of the cascade of events leading to cellular activation. Because synovial fibroblasts are key cells involved in joint destruction, this review outlines the events that trigger or inhibit the crucial pathways leading to their aggressive behavior. Major topics include cellular and humoral interactions (frequently modulated by cytokines), intracellular signaling and upregulation of gene transcription, and the deleterious effects on articular homeostasis.

Kinesin-like protein CENP-E is upregulated in rheumatoid synovial fibroblasts

Articular destruction by invading synovial fibroblasts is a typical feature in rheumatoid arthritis (RA),. Recent data support the hypothesis that key players in this scenario are transformed-appearing synovial fibroblasts at the site of invasion into articular cartilage and bone. They maintain their aggressive phenotype toward cartilage, even when first cultured and thereafter coimplanted together with normal human cartilage into severe combined immunodeficient mice for and extended period of time. However, little is known about the upregulation of genes that leads to this aggressive fibroblast phenotype. To inhibit this progressive growth without interfering with pathways of physiological matrix remodelling, identification of pathways that operate specifically in RA synovial fibroblasts is required. In order to achieve this goal, identification of genes showing upregulation restricted to RA synovial fibroblasts is essential.

AIMS

To identify specifically expressed genes using RNA arbitrarily primed (RAP)-polymerase chain reaction (PCR) for differential display in patients with RA.

METHODS

RNA was extracted from cultured synovial fibroblasts from 10 patients with RA, four patients with osteoarthritis (OA), and one patient with psoriatic arthritis. RAP-PCR was performed using different arbitrary primers for first-strand and second-strand synthesis. First-strand and second-strand synthesis were performed using arbitrary primers: US6 (5'-GTGGTGACAG-3') for first strand, and Nuclear 1+ (5'ACGAAGAAGAG-3'), OPN28 (5'GCACCAGGGGG-3'), Kinase A2+(5'-GGTGCCTTTGG-3') and OPN24 (5'AGGGGCACCA-3') for second strand synthesis. PCR reactions were loaded onto 8 mol/l urea/6% polyacrylamide-sequencing gels and electrophoressed. Gel slices carrying the target fragment were then excised with a razor blade, eluated and reamplified. After verifying their correct size and purity on 4% agarose gels, the reamplified products derived from the single-strand confirmation polymorphism gel were cloned, and five clones per transcript were sequenced. Thereafter, a genbank analysis was performed. Quantitative reverse transcription PCRj of the segments was performed using the PCR MIMIC technique. In-situ expression of centromere kinesin-like protein-E (CENP-E) messenger (m)RNA in RA synovium was assessed using digoxigenin-labelled riboprobes, and CENP-E protein expression in fibroblasts and synovium was performed by immunogold-silver immunohistochemistry and cytochemistry. Functional analysis of CENP-E was done using different approaches (eg glucocorticoid stimulation, serum starvation and growth rate analysis of synovial fibroblasts that expressed CENP-E).

RESULTS

In RA, amplification of a distinct PCR product suitable for sequencing could be observed. The indicated complementary DNA fragment of 434 base pairs from RA mRNA corresponded to nucleotides 6615-7048 in the human centromere kinesin-like protein CENP-E mRNA (GenBank accession No. emb/Z15005). The isolated sequence shared greater than 99% nucleic acid (P=2.9e(-169)) identify with the human centromere kinesin-like protein CENP-E. Two base changes at positions 6624 (A to C) and 6739 (A to G) did not result in alteration in the amino acid sequence, and therefore 100% amino acid identity could be confirmed. The amplification of 10 clones of the cloned RAP product revealed the presence of CENP-E mRNA in every fibroblast culture examined, showing from 50% (271.000 +/- 54.000 phosphor imager arbitrary units) up to fivefold (961.000 +/- 145.000 phosphor image arbitrary units) upregulation when compared with OA fibroblasts. Neither therapy with disease-modifying antirheumatic drugs such as methotrexate, gold, resochine or cyclosporine A, nor therapy with oral steroids influenced CENP-E expression in the RA fibroblasts. Of the eight RA fibroblast populations from RA patients who were receiving disease-modifying antirheumatic drugs, five showed CENP-E upregulation; and of the eight fibroblast populations from RA patients receiving steroids, four showed CENP-E upregulation. Numerous synovial cells of the patients with RA showed a positive in situ signal for the isolated CENP-E gene segment confirming CENP-E mRNA production in rheumatoid synovium, whereas in OA synovial tissue CENP-E mRNA could not be detected. In addition, CENP-E expression was independent from medication. This was further confirmed by analysis of the effect of prednisolone on CENP-E expression, which revealed no alteration in CENP-E mRNA after exposure to different (physiological) concentrations of prednisolone. Serum starvation also could not suppress CENP-E mRNA completely.

DISCUSSION

Since its introduction in 1992, numerous variants of the differential display method and continuous improvements including RAP-PCR have proved to have both efficiency and reliability in examination of differentially regulated genes. The results of the present study reveal that RAP-PCR is a suitable method to identify differentially expressed genes in rheumatoid synovial fibroblasts. The mRNA, which has been found to be upregulated in rheumatoid synovial fibroblasts, codes for a kinesin-like motor protein named CENP-E, which was first characterized in 1991. It is a member of a family of centromere-associated proteins, of which six (CENP-A to CENP-F) are currently known. CENP-E itself is a kinetochore motor, which accumulates transiently at kinetochores in the G2 phase of the cell cycle before mitosis takes place, appears to modulate chromosome movement and spindle elongation,and is degraded at the end of mitosis. The presence or upregulation of CENP-E has never been associated with RA.The three-dimensional structure of CENP-E includes a coiled-coil domain. This has important functions and shows links to known pathways in RA pathophysiology. Coiled-coil domains can also be found in jun and fos oncogene products, which are frequently upregulated in RA synovial fibroblasts. They are also involved in DNA binding and transactivation processes resembling the situation in AP-1 (Jun/Fos)-dependent DNA-binding in rheumatoid synovium. Most interestingly, these coiled-coil motifs are crucial for the assembly of viral proteins, and the upregulation of CENP-E might reflect the influence of infectious agents in RA synovium. We also performed experiments showing that serum starvation decreased, but did not completely inhibit CENP-E mRNA expression. This shows that CENP-E is related to, but does not completely depend on proliferation of these cells. In addition, we determined the growth rate of CENP-E high and low expressors, showing that it was independent from the amount of CENP-E expression. supporting the statement that upregulation of CENP-E reflects an activated RA fibroblast phenotype. In summary, the results of the present study support the hypothesis that CENP-E, presumably independently from medication, may not only be upregulated, but may also be involved in RA pathophysiology.