|Home | About | Journals | Submit | Contact Us | Français|
This study was performed to quantify the expression of CD44 and variant isoforms CD44v3 and CD44v6 on T cells from patients with systemic lupus erythematosus (SLE), and to correlate the level of expression of these molecules with disease manifestations.
Information on clinical and demographic characteristics was collected and blood samples were obtained from 72 patients with SLE and 32 healthy control subjects matched to the patients by sex, race, and age. Expression of CD44 and variants CD44v3 and v6 on T cell subsets was determined by flow cytometry, and Pearson’s correlations of their expression levels with clinical variables, SLE Disease Activity Index (SLEDAI) scores, and presence of lupus nephritis were determined. Wilcoxon’s rank sum tests and conditional multivariable regression analyses were applied to identify differences in the expression of CD44 between patients with SLE and healthy controls.
Expression of CD44 was higher on CD4+ and CD8+ T cells from SLE patients compared with controls (P ≤ 0.03). Expression of CD44v3 and CD44v6 was also higher on total T cells and CD4+ and CD8+ T cells from SLE patients compared with controls (P ≤ 0.03). Cell surface levels of CD44v3 on total T cells, CD4+ T cells, and CD8+ T cells as well as cell surface expression of CD44v6 on total T cells and CD4+ T cells were correlated with the SLEDAI score (P < 0.05). The presence of lupus nephritis was associated with the expression of CD44v6 on total T cells, CD4+ T cells, and CD4−CD8− T cells (P < 0.05). Positivity for anti–double-stranded DNA antibodies was associated with the expression levels of CD44v6 on T cells (P < 0.05).
These results indicate that expression levels of CD44v3 and CD44v6 on T cells may represent useful biomarkers of SLE activity.
Infiltration of proinflammatory T cells into target organs in patients with systemic lupus erythematosus (SLE) represents a mechanism of amplification and maintenance of the autoimmune response that eventually leads to end organ damage (1,2). T cell entry into the inflamed tissue is facilitated by the increased adherence and migration capacities of T cells from patients with SLE (3). These properties have been explained, in part, by the identification of a population of T cells that bind with increased affinity to hyaluronic acid (HA) through the T cell surface molecule CD44 (4).
The CD44 glycoprotein is encoded by a highly conserved gene that is capable of generating heterogeneous proteins by alternative splicing and different posttranslational modifications (5,6). The CD44 gene includes 10 variable exons that are excluded from the standard CD44 molecule (CD44s). The inclusion of different combinations of variable exons can potentially produce a large number of variant CD44 isoforms (Figure 1). An array of these isoforms is normally expressed by epithelial cells of the skin and by cells of the nervous system (7,8). In tumor cell lines and carcinomas, the expression of certain isoforms confers metastatic capacity (9–11). Likewise, the presence of the v3 and v6 isoforms has been found to grant invasive capacity to fibroblast-type synoviocytes in patients with rheumatoid arthritis (12,13).
Migration and adherence are particularly important for immune cells. On T cells, expression of CD44 isoforms changes through different maturation and activation states (6,14). The CD44 isoforms v3 and v6 are expressed by T cells, particularly following activation, and are involved in several cellular functions (15,16). CD44v6 participates in T and B cell activation, and its blockage hampers the development of T cell–dependent and T cell–independent antibody responses. In addition, CD44v6 participates in the generation of cytotoxic T cells (17). In a previous study, we showed that T cells in the kidney infiltrates of patients with lupus nephritis expressed CD44v3 and v6 (1). We thus hypothesized that the expression of CD44v3 and v6 may be altered in patients with SLE, and that the increased expression of these molecules is associated with T cell infiltration into target organs, such as the kidneys. In this cross-sectional study, we quantified the cell surface expression of CD44s, as well as CD44v3 and v6, on CD4+, CD8+, and CD4−CD8− (double-negative [DN]) T cells from patients with SLE and healthy control subjects.
Seventy-two patients with SLE, whose diagnosis fulfilled the American College of Rheumatology 1997 revised classification criteria (18,19), were recruited from the Lupus Centers of the Brigham and Women’s Hospital (BWH) and the Beth Israel Deaconess Medical Center (BIDMC). Thirty-two subjects without SLE or without a related connective tissue disease were recruited as controls and matched to the patients by sex, race, and age (within 5 years). All subjects were recruited from the general outpatient rheumatology clinics at the same hospitals. Demographic information as well as information on family history of SLE and other autoimmune disease, cigarette smoking (current, past, never), recent infections (hospitalization or antibiotic treatment for infection during the previous month), current medications (including corticosteroids and immunosuppressive drugs), comorbid illnesses, insurance provider, and education level were collected from all subjects. In addition, the following data were collected from the patients with SLE: 1) age at SLE onset, 2) SLE manifestations and organ involvement, and 3) history of positivity for autoantibodies (determined on tests performed in the clinical immunology laboratories of BWH or BIDMC). The treating rheumatologist evaluated each patient by performing a physical examination for SLE disease activity using the Safety of Estrogens in Lupus Erythematosus: National Assessment (SELENA) version of the SLE Disease Activity Index (SLEDAI) (20), and by obtaining a history and chart review to complete an SLE organ damage assessment, using the Systemic Lupus International Collaborating Clinics/American College of Rheumatology Damage Index (SDI) (21). Blood samples were obtained from all subjects by standard antecubital venipuncture. The study was approved by the relevant Institutional Review Boards, and all participants signed informed consent forms.
Fluorescein isothiocyanate–labeled anti-CD3, phycoerythrin (PE)–Cy5-labeled anti-CD4, allophycocyanin–Cy7-labeled anti-CD8, PE-labeled anti-CD44 (clone 515), and PE-labeled mouse IgG1 were purchased from BD PharMingen. Anti-CD44v3 and anti-CD44v6 polyclonal rabbit IgG antibodies were purchased from Calbiochem. PE-labeled goat anti-rabbit antibodies were purchased from R&D Systems. The Fc receptor (FcR)–blocking reagent was purchased from Miltenyi Biotec.
Peripheral blood mononuclear cells were obtained from 10 ml of venous blood by centrifugation against a density gradient (Lymphoprep). The FcR-blocking reagent was added and the cells were stained with anti-CD3, anti-CD4, anti-CD8, and either anti-CD44, anti-CD44v3, or anti-CD44v6. After a 30-minute incubation at room temperature, cells were washed and the secondary antibody was added to the tubes stained with CD44v3 and CD44v6. Cells were incubated for 30 minutes, washed, fixed in phosphate buffered saline plus 0.3% paraformaldehyde, and analyzed in an LSRII flow cytometer (Becton Dickinson). Thirty thousand T cells were acquired.
Flow cytometry analyses were performed using FlowJo (Tree Star). Lymphocytes were gated according to the forward and side scatter patterns, and T cells were identified by CD3 expression. T cells were further divided into CD4+, CD8+, and DN T cell subsets. Levels of CD44, CD44v3, and CD44v6 were quantified on total T cells, as well as on the 3 cell subsets, with results expressed as the mean fluorescence intensity (MFI). To quantify the percentage of cells positive for CD44v3 and CD44v6, a gate was set according to the staining intensity of an isotype control.
Results are expressed as the mean ± SEM unless stated otherwise. Student’s paired t-tests were used to compare the demographic characteristics between patients and controls. The expression levels of CD44s and its variant isoforms CD44v3 and v6 were compared between patients and matched controls using Wilcoxon’s rank sum tests. Pearson’s correlation coefficients were utilized to test for associations between the levels of expression of the studied molecules and clinical parameters. Conditional logistic regression models were constructed in order to test the statistical significance of CD44 levels and clinical variables as predictors of the clinical status of patients compared with controls. Two-sided P values less than 0.05 were considered significant.
The demographic and clinical characteristics of the study population are summarized in Table 1. There were no differences in age, sex, race, smoking status, or history of recent infection between patients and controls. One-quarter of the SLE patients had a history of renal disease, 39% had anti–double-stranded DNA (anti-dsDNA) antibodies, 31% were receiving corticosteroids, and 25% were receiving immunosuppressive drugs at the time of the study.
We quantified the expression of CD44s on T cells from SLE patients and healthy controls. In agreement with our previous study findings (3), and as shown in Table 2, expression of CD44s was significantly higher in CD4+ and CD8+ T cells from patients with SLE when compared with controls (P ≤ 0.03).
Under certain conditions, cells incorporate variable exons into the mature CD44 transcript and, thus, produce the so-called variant isoforms (22). As expected, the expression levels of the individual variant isoforms were significantly lower than those of CD44s (Table 2).
CD44v3 was present in a significantly higher percentage of CD8± and DN T cells from patients with SLE than in those from controls (P < 0.01) (Table 3). When expression levels were quantified as the MFI, the expression of CD44v3 was higher on total T cells, CD4+ T cells, and CD8+ T cells from patients with SLE compared with the respective cell subsets from healthy controls (P ≤ 0.01) (Table 2). Likewise, CD44v6 was detected on a significantly higher fraction of CD8+ T cells from patients with SLE compared with those from controls (P < 0.01) (Table 3). The expression intensity of CD44v6, expressed as the MFI, was augmented on total T cells, CD4+ T cells, and CD8+ T cells from SLE patients when compared with matched controls (P < 0.05) (Table 2). These data are summarized in Figure 2.
We found no significant correlations between the levels of expression of CD44s on T cells and the disease phenotype, disease activity (according to the SLEDAI score), or SLE-associated organ damage (according to the SDI) (results not shown). Since the expression of CD44v3 and CD44v6 grants cells an increased capacity to migrate and adhere (13), and cells bearing these receptors have been observed to infiltrate lupus target organs (1), we hypothesized that increased expression levels of these isoforms would be associated with higher SLE disease activity and with the development of lupus nephritis. As shown in Table 4, the expression levels of CD44v3 on total T cells, CD4+ T cells, and CD8+ T cells showed a positive correlation with the SLEDAI score (P ≤ 0.02). Expression of CD44v6 on total T cells and CD4+ T cells also correlated positively with disease activity (P < 0.05) and with the presence of lupus nephritis (P < 0.05). Finally, high expression of CD44v3 on CD8+ T cells and of CD44v6 on total T cells, CD4+ T cells, and CD8+ T cells was more common in patients with antibodies against dsDNA (P ≤ 0.05). There was no correlation between the expression levels of CD44 variant isoforms and age, sex, race, smoking status, duration of SLE, or history of recent infection (results not shown).
Bivariate analyses detected weak (r < 0.2), but significant, positive correlations between the level of expression of CD44v3 and CD44v6 in different cell subsets and the use of corticosteroids and immunosuppressive drugs (results not shown). However, in multivariable conditional and unconditional regression models, adjustment for clinical variables, including current use of corticosteroids and immunosuppressive drugs, did not affect the statistical significance of the associations between CD44 isoform expression and disease activity (according to the SLEDAI score), presence of anti-dsDNA antibodies, or presence of lupus nephritis.
In this cross-sectional case–control study, we investigated the expression of the standard CD44 molecule and the variant isoforms CD44v3 and v6 on T cells from patients with SLE compared with matched healthy control subjects. Our results show that the expression of these molecules, especially of the variant isoforms, is significantly increased on CD4+ and CD8+ T cells from patients with SLE. These results are congruent with those reported by Estess and coworkers in a previous study (4). Moreover, our findings indicate that T cells from SLE patients express higher levels of CD44v3 and v6, and that their expression level is correlated with the extent of disease activity, the presence of nephritis, and positivity for anti-dsDNA antibodies.
Although it is known that the expression of CD44 variant isoforms alters the binding of CD44 to HA, this effect has not been well characterized (23). It has been reported that the presence of variant isoforms increases HA binding (24,25), whereas other groups have found that HA binding is decreased in the presence of variant isoforms (26–29). Aside from HA, CD44 is able to bind to several other ligands, including fibronectin, collagen, and laminin. It is likely that the structural modifications caused by inclusion of the variant exons into the CD44 molecule modifies the binding to these alternative ligands, but this has not been proven (23).
Cellular activation and effector differentiation alter the expression of CD44 variants on T cells (30). Disease activity can modify the expression of CD44 variant isoforms on lymphocytes through various mechanisms, including increased levels of T cell activation and serum factors (particularly autoantibodies), which have been shown to induce phenotypic changes in T cells, including increased adhesion and migration capacity (3,31,32). In this study, we did not quantify the expression of other T cell activation markers reported to be augmented in patients with SLE (e.g., CD69) (33), and therefore we were not able to determine whether the increased levels of CD44v3 and v6 are attributable to the enlarged pool of activated T cells that circulates in patients with SLE or are due to other processes unrelated to the cell activation status, such as altered RNA splicing (which has been shown to alter the expression of other proteins in patients with SLE) (34).
T cell migration into target organs is a necessary step in the pathogenic progression that leads to chronic inflammation and organ damage in autoimmune conditions such as SLE. The CD44 molecules are involved essentially in this process (6). T cells expressing CD44v3 have been observed in inflammatory infiltrates in the skin (35) and kidneys (1) of patients with SLE. Interestingly, CD44v6+ T cells are absent in skin infiltrates (35), but are present in inflamed kidneys (1). This underscores the importance of CD44 variant isoforms in determining the homing capacities of T cells.
The T cells that harbor increased levels of CD44 molecules have an enhanced capacity to infiltrate tissue and incite inflammation. Thus, we propose that by the quantification of the level of expression of CD44v3 and CD44v6, we can infer which cases are more likely to develop severe disease activity and nephritis. Our data suggest that there is a cross-sectional association between expression of CD44v3 and CD44v6 and disease activity and nephritis.
The CD44 molecules are expressed on the surface of T cells and can be readily quantified by standard flow cytometry methods. Our results suggest that the variant isoforms CD44v3 and v6 may represent adequate targets to be used as biomarkers of disease activity. The main limitation of this study is its cross-sectional nature and lack of longitudinal case followup. A prospective, longitudinal study will be necessary to confirm the usefulness of quantifying the expression of these molecules. Moreover, a prospective study will provide the information required to calculate the frequency with which CD44 variant isoforms must be measured and the mean time elapsed from the increment in their expression to the expected change in clinical activity (36).
In summary, the level of expression of CD44v3 and v6 on T cells may prove to be a useful biomarker in patients with SLE. We believe that the quantification of these molecules, which are directly involved in the pathogenic process, will allow us to measure disease activity in a reliable and relevant manner.
Supported by a Public Health Service grant (R01-AI-42269) to Dr. Tsokos and an Alliance for Lupus Research grant to Dr. Kostenbader.
We wish to thank Usha Govindarajulu, PhD for her assistance with the statistical analyses, and Leigh Fritz, BA for her efforts in recruiting subjects and collecting data.
Dr. Bermas has provided expert testimony on risk management for the Harvard Hospitals; she receives royalties for Up to Date. Dr. Massarotti has received consulting fees, speaking fees, and/or honoraria from Genentech (less than $10,000). Dr. Tsokos has received consulting fees, speaking fees, and/or honoraria from Genentech, Human Genome Sciences, and Celgene (less than $10,000 each).
AUTHOR CONTRIBUTIONSAll authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Drs. Costenbader and Crispín had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study conception and design. Crispín, Kyttaris, Tsokos, Costenbader.
Acquisition of data. Crispín, Finnell, Bermas, Schur, Massarotti, Karlson, Fitzgerald, Ergin, Kyttaris, Tsokos, Costenbader.
Analysis and interpretation of data. Crispín, Keenan, Tsokos, Costenbader.