DNA methyltransferase 1
We have determined that abnormal DNA methylation in T cells coincides with the development of autoimmunity, using a mouse model that exhibits an age-dependent lupus-like disease (MRL/lpr mice). Splenic CD4+ T cells were isolated from these mice at 5 and 16 wk of age (before and after autoimmunity is established) and the expression of DNA methyltransferase 1 (Dnmt1) and the methylation-sensitive gene Tnfsf7 (CD70) was measured. Bisulfite DNA sequencing was used to monitor the methylation status of the Tnfsf7 gene. We found that Dnmt1 steady-state mRNA levels were significantly lower in 16-wk-old MRL/lpr mice, which had established autoimmunity, compared to the 5-wk-old MRL/lpr mice. Furthermore, the expression of CD70 was higher in MRL/lpr mice at 16 wk. CD70 was overexpressed in MRL/lpr mice compared to age- and sex-matched MRL+/+ controls. Bisulfite DNA sequencing of the Tnfsf7 gene in MRL/lpr mice revealed that at 16 wk, CG pairs were hypomethylated compared to 5-wk-old mice, and that Tnfsf7 from MRL/lpr mice was hypomethylated at 16 wk relative to age-matched MRL+/+ controls. Our data indicate that decreased expression of Dnmt1 and the corresponding T cell DNA hypomethylation correlate with the development of age-dependent autoimmunity in MRL/lpr mice.
Systemic lupus erythematosus is a chronic relapsing autoimmune disease characterized by the production of autoantibodies directed against a host of nuclear antigens. Lupus is more common in females and can affect multiple organ systems including the skin, kidney, lung, heart, and the central nervous system. The pathogenesis of lupus is incompletely understood; however, multiple genetic and environmental factors have been implicated in the disease etiology 1–3. A growing body of evidence suggests a role for abnormal T cell DNA methylation in causing both drug-induced and idiopathic lupus (reviewed in 4). In general, genes that are methylated are silenced while genes that are hypomethylated are transcriptionally active 5. It is known that decreased T cell DNA methylation results in overexpression of methylation-sensitive genes in T cells such as ITGAL (CD11a), PRF1 (Perforin), and TNFSF7 (CD70), and overexpression of these genes leads to T cell autoreactivity in vitro and autoimmunity in vivo6–10.
The MRL/lpr mouse strain is a prominent and well-characterized animal model of lupus. These mice develop glomerulonephritis, lymphadenopathy and a number of autoantibody specificities such as anti-dsDNA antibody and anti-Sm antibody, all of which are hallmarks of lupus in humans 11. The lpr mutation, a null mutation in the apoptosis gene fas, accelerates the lupus phenotype on the MRL genetic background. Autoimmunity is not evident in MRL/lpr mice at birth and develops as they approach adulthood. Anti-dsDNA antibodies can be detected as early as ∼6 wk of age and are followed by lymphadenopathy and progressive renal disease at ∼12 and 16 wk, respectively 12. Similarly, autoantibodies are not produced at birth in humans, but their production precedes the development of lupus 13. Events leading to the break in immune tolerance in the MRL/lpr mouse are not completely clear, but the model provides a means to address changes in gene expression occurring before and after this transition. In this report, we investigate the link between the hypomethylation and overexpression of genes implicated in lupus and the correlation of these events with the development of age-dependent autoimmunity in MRL/lpr mice.
Defective DNA methyltransferase 1 expression in MRL/lpr CD4+ T cells
Normal DNA methylation is maintained by DNA methyltransferase 1 (Dnmt1), and decreased activity of Dnmt1 in T cells from active lupus patients results in T cell autoreactivity 4. We therefore measured Dnmt1 in MRL/lpr mice, hypothesizing that a reduction in the expression of Dnmt1 would correlate with loss of immune tolerance and development of autoimmunity. CD4+ T cells were isolated from MRL/lpr mice at 5 wk of age, before development of autoimmune disease, and at 16 wk, after autoimmunity had been established. Dnmt1 expression, as measured by mRNA levels, was ultimately reduced by ∼80% (p = 0.016) as mice developed increasing titers of anti-dsDNA antibody, measured as an indicator of autoimmunity (Fig. 1A, B).
Increased expression of Tnfsf7 (CD70) in MRL/lpr CD4+ T cells
To determine whether reduced Dnmt1 expression affected expression of any methylation-sensitive genes, we measured mRNA levels from a methylation-sensitive gene in MRL/lpr CD4+ T cells: Tnfsf7, which encodes CD70. CD70 expression was significantly higher in MRL/lpr mice at 16 wk compared to 5 wk of age [4.35 ± 1.27 versus 0.62 ± 0.23 (mean ± SEM), p = 0.016] (Fig. 1C). Moreover, in a different experiment, we detected significantly higher expression of CD70 in 16-wk-old MRL/lpr mice compared to MRL+/+ controls [0.92 ± 0.17 versus 0.17 ± 0.09 (mean ± SEM), p = 0.003] (Fig. 2A). Changes in CD70 expression were confirmed by flow cytometry (Fig. 2B, C). Measurements of CD70 in CD4+ cells, which simultaneously express the activation marker CD69, indicated that CD70 expression remained significantly higher in MRL/lpr mice compared to MRL+/+ controls. This suggested that CD70 expression is induced, at least in part, by a mechanism other than T cell activation in MRL/lpr mice (Fig. 3).
Tnfsf7 (CD70) is hypomethylated in MRL/lpr mice when autoimmunity is established
To determine the relationship between overexpression of Tnfsf7 (CD70) and the observed reduction in the Dnmt1 expression, we performed bisulfite DNA sequencing of the mouse CD70 gene in young and old MRL/lpr mice, to determine the methylation status of CG pairs surrounding the transcription start sites (Fig. 4). Using the MethPrimer software 14, one CpG island was identified in the region from –1000 to +1000 relative to the CD70 transcription start site (Fig. 4). Bisulfite treatment followed by direct DNA sequencing revealed hypomethylation of CD70 in old versus young MRL/lpr mice, and in MRL/lpr mice relative to age-matched MRL+/+ controls (Fig. 4). MRL/lpr mice showed significant hypomethylation compared to MRL+/+ mice, with an average difference in hypomethylated CG fractions of 8.57 ± 1.90% (mean ± SEM), p = 0.0004. MRL/lpr mice also showed a significant hypomethylation in the CD70 gene at 16 wk compared to 5 wk of age [difference of 10.83 ± 1.64% (mean ± SEM), p = 0.0001].
DNA methylation refers to the addition of a methyl group to position 5 of the cytosine ring. This process, catalyzed by a group of DNA methyltransferases, occurs in CG pairs, usually clustered within or around promoter sequences to form CpG islands. As one of multiple epigenetic mechanisms, DNA methylation is involved in the regulation of gene expression in mammalian cells (reviewed in 15). While genes that are methylated are not expressed, hypomethylated gene sequences are typically available for transcription. In lupus, a number of methylation-sensitive T cell genes that are linked to T cell autoreactivity are overexpressed 4. Moreover, T cells treated with DNA methylation inhibitors, such as 5-azacytidine, procainamide or hydralazine, overexpress the methylation-sensitive genes ITGAL (CD11a), PRF1 (Perforin), and TNFSF7 (CD70), similar to T cells from active lupus patients 4. Studies demonstrated that the overexpression of methylation-sensitive genes in lupus T cells results from decreased activity of Dnmt1 16, 17. Dnmt1 maintains normal DNA methylation patterns in postnatal life. This contrasts with the DNA methyltransferases 3a and 3b which are, for the most part, involved in de novo DNA methylation during fetal development 18.
In this report, we present evidence consistent with the hypothesis that abnormal T cell DNA methylation, induced by the presence of the lpr mutation, plays a role in the pathogenesis of lupus in the MRL/lpr lupus-prone mouse. We show that expression of Dnmt1 is significantly reduced in MRL/lpr mice when autoimmunity is established. We also show that CD4+ T cells from 16-wk-old MRL/lpr mice express CD70 at significantly higher levels than age-matched MRL+/+ mice and 5-wk-old MRL/lpr mice. Increased CD70 in older MRL/lpr mice is predicted in light of reduced Dnmt1 expression in these animals, as Tnfsf7 is methylation sensitive. The product of Tnfsf7, which is also known as CD27 ligand (CD27L), is a member of the tumor necrosis factor family that is expressed on activated T cells and has B cell costimulatory functions 19–21. Indeed, the addition of the anti-CD27 monoclonal antibody could inhibit T cell-dependent pokeweed mitogen-driven B cell production of IgG 22. CD70-transgenic MRL/lpr mice develop massive T cell expansion and differentiation into IFN-γ-producing cells. In addition, these mice develop bone marrow failure and severe liver pathology by 4 wk of age 23.
Our data confirm and extend the findings of Oelke et al. who showed that CD70 is overexpressed on polyclonal as well as cloned human CD4+ T cells treated with a panel of DNA methylation inhibitors 24. In addition, the group demonstrated that CD70-overexpressing polyclonal and cloned CD4+ T cells overstimulate B cell IgG production and the stimulation can be inhibited with anti-CD70 monoclonal antibodies 24.
To confirm that reduced Dnmt1 in mature MRL/lpr mice correlates with CD70 overexpression, we demonstrate that the overexpression of CD70 observed is associated with a methylation defect in the CD70 gene. We show hypomethylation of CG pairs in older compared to both younger MRL/lpr mice and to age-matched MRL+/+ mice. Hypomethylation in the CD70 promoter sequence has been reported in CD4+ cells from lupus patients 25.
In summary, even though the etiology of lupus is still incompletely understood, a growing body of evidence is linking DNA methylation defects, particularly in T cells, to the pathogenesis of this disease. Herein, we report that a defect in CD4+ T cell DNA methylation, presumably resulting from decreased expression of the DNA-methylating enzyme Dnmt1, correlates with the development of lupus in MRL/lpr mice. Furthermore, we demonstrated that the methylation-sensitive gene Tnfsf7 (CD70) is overexpressed in MRL/lpr mice as they get older and develop autoimmunity. CD70 overexpression is linked to T cell autoreactivity in vitro, and thus its overexpression in our experimental system is a likely contributor to the development of autoimmunity. Our observation that CD4+ T cells from MRL/lpr mice overexpress CD70 and that DNA methylation is defective in CD4+ T cells in this lupus mouse model are both novel findings. This validates the MRL/lpr lupus-prone mouse as a model system to test potential therapeutic interventions that can alter DNA methylation or block the CD70/CD27 pathway in lupus.
Materials and Methods
MRL/lpr and MRL+/+ mice were purchased from Jackson Laboratories (Bar Harbor, ME) and maintained at the Oklahoma Medical Research Foundation animal facility under specific pathogen-free conditions. All protocols were approved by the Institutional Animal Care and Use Committee at the Oklahoma Medical Research Foundation. Mice were matched for sex in all experiments comparing young and old MRL/lpr mice, and were matched for both age and sex in experiments comparing MRL/lpr and MRL+/+ mice.
CD4+ cell separation
CD4+ lymphocytes were isolated from the spleens of MRL/lpr and MRL+/+ mice via magnetic bead separation using direct labeling (Miltenyi Biotec, Auburn, CA) following the manufacturer's protocol. Cell purity was checked by flow cytometry.
RNA and DNA isolation
The RNeasy kit (Qiagen, Valencia, CA) was used to isolate RNA from CD4+ T cells according to the manufacturer's instructions. DNA digestion was performed using Turbo DNA-free (Ambion, Austin, TX) following the manufacturer's instructions.
DNA was isolated from spleen CD4+ T cells of MRL/lpr and MRL+/+ mice using the DNeasy Tissue Kit (Qiagen, Valencia, CA).
Quantitative real-time RT-PCR
One-step quantitative real-time RT-PCR was performed using the QuantiTect SYBR Green RT PCR kit (Qiagen, Valencia, CA) and the Rotor-Gene 3000 real-time thermocycler (Corbett Research, Australia). A total of 150 ng of RNA was used for each reaction. The following amplification conditions were used: 30 min at 50°C, 15 min at 95°C, 54 cycles of 15 s at 94°C and 20 s at 56°C and 30 s at 72°C. Quantitation was performed by comparison with internal standards prepared by serial dilutions. The expression levels of Dnmt1 and CD70 were normalized for the house-keeping gene GAPDH. The following primers were used: mouse GAPDH forward: 5′-CAACGACCCCTTCATTGACCTC-3′; reverse: 5′-GCCTCACCCCATTTGATGTTAGTG-3′; mouse Dnmt1 forward: 5′-GGAAGGCTACCTGGCTAAAGTCAAG-3′; reverse: 5′-ACTGAAAGGGTGTCACTGTCCGAC-3′; mouse CD70 forward: 5′-TGGCTGTGGGCATCTGCTC-3′; reverse: 5′-ACATCTCCGTGGACCAGGTATG-3′. All primers were purchased from Integrated DNA Technologies, Inc. (Coralville, IA).
Flow cytometry analysis was performed at the Oklahoma Medical Research Foundation flow cytometry core facility using a FACSCalibur machine. The following antibodies were used to assess cell surface expression of various molecules: APC-conjugated anti-mouse CD4, PE-conjugated anti-mouse CD70, and FITC-conjugated anti-mouse CD69. All antibodies were purchased from BD Pharmingen (San Diego, CA).
Bisulfite DNA sequencing
A region from –1000 to +1000 bp in the mouse CD70 gene was screened for the presence of CpG islands using the MethPrimer software 14. A CpG island was defined as a stretch of DNA more than 200 bp in size with a CG content of more than 50% and an observed/expected CG ratio of more than 0.6. DNA isolated from CD4+ cells from MRL/lpr and MRL+/+ mice was bisulfite treated using the EZ DNA Methylation kit (Zymo Research, Orange, CA). Following PCR, DNA was purified using the QIAquick Gel Extraction columns (Qiagen, Valencia, CA). Direct sequencing was performed at the Oklahoma Medical Research Foundation sequencing core facility. The following PCR primers were used: mouse CD70 CpG island forward: 5′-GAGGTTTTAAGGGTAGGTTAAGGTAG-3′; mouse CD70 CpG island reverse: 5′-AATAATAAACACCACCAACAACAAC-3′.
Differences between groups were tested using Student's t-test; p values of less than 0.05 were considered statistically significant.
This publication was made possible by NIH Grant Number P20-RR015577 from the National Center for Research Resources and by funding from the University of Oklahoma College of Medicine (AHS). We would like to thank Dr. Bruce Richardson, Dr. Susan Kovats, and Dr. John B Harley for very helpful suggestions and comments. We would also like to thank Dr. John Knight for reading and editing our manuscript.