Dr. Tsokos has received consulting fees, speaking fees, and/or honoraria from Millennium Pharmaceuticals, Merck, MedImmune, Genentech, and Eli Lilly (less than $10,000 each) and has performed editorial work for Elsevier.
Systemic lupus erythematosus (SLE) T cells display a hyperactive calcineurin/NF-AT pathway. The aim of this study was to determine whether this pathway is responsible for the aberrant SLE T cell function and to test the effectiveness of the recently recognized calcineurin inhibitor dipyridamole in limiting SLE-related pathology.
T cells and mononuclear cells were isolated from the peripheral blood of SLE patients and healthy individuals. Murine cells were isolated from the spleens and lymph nodes of lupus-prone MRL/lpr mice and control MRL/MpJ mice. Cells were treated in vitro with tacrolimus, dipyridamole, or control. MRL/lpr mice were injected intraperitoneally with 50 mg/kg of dipyridamole 3 times a week for 3 weeks.
MRL/lpr T cells, especially CD3+CD4–CD8– cells, displayed a robust calcium influx upon activation and increased levels of NF-ATc1. MRL/lpr T cells (both CD4+ and CD3+CD4–CD8– cells) provided help to B cells to produce immunoglobulin in a calcineurin-dependent manner. Dipyridamole treatment of SLE T cells significantly inhibited CD154 expression, interferon-γ, interleukin-17 (IL-17), and IL-6 production, and T cell–dependent B cell immunoglobulin secretion. Treatment of MRL/lpr mice with dipyridamole alleviated lupus nephritis and prevented the appearance of skin ulcers.
NF-AT activation is a key step in the activation of SLE T cells and the production of immunoglobulin. Dipyridamole inhibits SLE T cell function and improves pathologic changes of the disease in lupus-prone mice. We propose that dipyridamole can be used in treatment regimens for patients with SLE.
Upon engagement of the T cell receptor (TCR), T cells from patients with systemic lupus erythematosus (SLE) show a robust calcium response (1) and mobilize the transcription factor NF-ATc2 into the nucleus at higher rates than do T cells from healthy volunteers (2, 3). Aggregation of lipid rafts (4), substitution of the CD3 ζ-chain/ZAP-70 signaling duet by Fc receptor γ (FcRγ)/Syk (5, 6), and mitochondrial hyperpolarization (7) have been claimed to be the cause of the enhanced and accelerated early T cell response.
It is currently accepted that SLE T cells provide aberrant help to B cells to produce pathogenic autoantibodies through increased expression of costimulatory molecules, such as CD154, the ligand for CD40 (8). At the same time, activated SLE T cells directly invade tissues, such as the kidneys and the skin. These tissue-infiltrating T cells can be either CD4+ or CD4–CD8– (double-negative) T cells, and they produce the proinflammatory cytokine interleukin-17 (IL-17) (9). Double-negative T cells are expanded in SLE patients and in the lupus-prone MRL/lpr mouse and appear to be of pathogenic importance.
The degree to which the enhanced early signaling events contribute to the observed SLE T cell helper and effector functions is unclear to date. In this study, we demonstrate that MRL/lpr mouse T cells provide aberrant help to normal mouse B cells in a calcineurin-dependent manner, linking SLE T cell hyperactivity to T cell helper function. Moreover, we show that the enhanced calcium/calcineurin/NF-AT pathway in human and murine SLE T cells can be suppressed in the presence of dipyridamole, a recently recognized specific inhibitor of calcineurin–NF-AT interactions (10). Finally, administration of dipyridamole to MRL/lpr mice improves the disease pathology.
PATIENTS AND METHODS
Patients (n = 19) who fulfilled the American College of Rheumatology criteria for the diagnosis of SLE (11) were enrolled in the study through the donation of 50 ml of blood, which was drawn into heparin/lithium-containing tubes. Of the 19 SLE patients, 95% were female, their mean age was 36 years (range 23–54 years), and 42% were white, 42% were African American, and 11% were Asian American. Three healthy individuals who were taking no medications were also enrolled in the study. These subjects served as controls.
Disease activity was calculated using the Systemic Lupus Erythematosus Disease Activity Index (SLEDAI) (12). The mean SLEDAI score was 5.5, with 3 patients having quiescent disease (SLEDAI score of 0) and 5 patients having high disease activity (SLEDAI score >8). Of the 19 SLE patients, 60% were taking corticosteroids at a mean dosage of 22 mg/day (range 5–60 mg/day). Other immunosuppressive medications being taken at the time of the study included hydroxychloroquine (75%), azathioprine (20%), and mycophenolate mofetil (50%). Two patients were receiving intravenous monthly cyclophosphamide, 1 was receiving cyclosporine, and 1 was receiving weekly methotrexate at the time of the study. Prednisone was withheld for at least 12 hours prior to the blood draw.
The Institutional Review Board of Beth Israel Deaconess Medical Center (BIDMC) approved the study protocol. Informed consent was obtained from all of the study subjects.
MRL/lpr/2J (MRL/lpr) and MRL/MpJ were purchased from The Jackson Laboratory and were housed in the barrier animal facility of BIDMC. Tartaric acid and dipyridamole were injected intraperitoneally under sterile conditions using 29-gauge needles. The Institutional Animal Care and Use Committee at BIDMC approved all animal-related procedures.
The urine that was collected from individual mice was analyzed for protein, blood, and white cells using Multistix 10 SG reagent strips and a Clinitek Status analyzer (Bayer Healthcare). All analyses were done using a semiquantitative method. Proteinuria was graded as + = 30 mg/dl, ++ = 100 mg/dl, +++ = 1,000 mg/dl, and ++++ >2,000 mg/dl.
Longitudinal sections of the skin were obtained, fixed in 4% buffered paraformaldehyde, and then embedded in paraffin. The skin sections were subsequently cut at 3 μm thickness. The sections were stained with hematoxylin and eosin.
Cell isolation, culture, stimulation, and proliferation assays.
The blood was incubated for 30 minutes with a mixture of antibodies against CD14, CD16, CD19, CD56, and glycophorin A that attaches non–T cells to erythrocytes. Ficoll-containing Lymphoprep gradient (Nycomed) was subsequently used to separate these complexes from T cells. Using flow cytometry, we established that the purified cells were >98% positive for CD3. During culture, the cells were incubated in RPMI 1640 medium with 10% (volume/volume) heat-inactivated fetal calf serum (FCS; Sigma-Aldrich) supplemented with L-glutamine and 100 units of penicillin and 100 μg of streptomycin per ml. These incubations took place in a culture incubator at a temperature of 37°C in a humidified atmosphere containing 5% CO2. Where indicated below, T cells were stimulated with plate-bound anti-CD3 antibody (2 μg/ml) and anti-CD28 antibody (2 μg/ml). For the drug treatment experiments, purified T cells were incubated with the drug at 37°C for half an hour prior to stimulation. Peripheral blood mononuclear cells were isolated using the above protocol without the antibody mixture and were cultured for up to 13 days in RPMI 1640 medium as above.
Cells were extracted from murine spleens and lymph nodes by filtering the tissue through a 70-μm BD Falcon cell strainer. The extracts were centrifuged at 1,200 revolutions per minute for 5 minutes. ACK lysing buffer (Quality Biological) solution was added in the cell pellet in order to lyse the red blood cells. The treated cell pellet was subsequently washed once with DMEM with 10% (volume/volume) heat-inactivated FCS supplemented with 50 μM mercaptoethanol, 1 mM sodium pyruvate, nonessential amino acids, L-glutamine, and 100 units of penicillin and 100 μg of streptomycin per ml. Cells were placed in culture incubator and incubated at 37°C in a humidified atmosphere containing 5% CO2. For cell sorting, CD3, CD4, CD8, and CD19, antibodies were used, and the sorting was done at the cell sorting facility at BIDMC. CD4+ naive cells were isolated from murine splenic extracts using a MACS CD4+CD62L+ T Cell Isolation Kit II (Miltenyi Biotec).
For the proliferation assays, lymphocytes were suspended at a concentration of 5 × 106/ml in phosphate buffered saline (PBS)/2% FCS. Then, 5,6-carboxyfluorescein succinimidyl ester (CFSE; Molecular Probes) was added at a final concentration of 10 μM, and the cells were incubated at 37°C for 5 minutes. At the end of the incubation period, the cells were washed 3 times in cold complete RPMI 1640 and were then stimulated for 3–5 days.
Protein purification and Western blotting.
After several washings, the cells were treated initially on ice for 15 minutes with a 200-μl concentration of lysis buffer (10 mM HEPES, pH 7.9, 10 mM KCl, 0.1 mM EDTA, 0.1 mM EGTA supplemented with freshly added 1 mM DTT, 0.5 mM phenylmethylsulfonyl fluoride [PMSF], 2 mM aprotinin, 1 mM leupeptin, 10 mM NaF, and 2 mM Na3VO4). At the end of the incubation, Nonidet P40 was added to the reaction mixture at a concentration of 0.6%. The reaction mixture was vortexed for 10 seconds and then centrifuged at 13,000 rpm for 15 seconds. The supernatant was saved as the cytoplasmic extract. The pellet was resuspended in 25 μl of buffer (20 mM HEPES, pH 7.9, 0.4M NaCl, 1 mM EDTA, 1 mM EGTA, 10 mM NaF, 1 mM Na3VO4, 1 mM PMSF, 2 mM aprotinin, and 1 mM leupeptin) and then shaken for 15 minutes at 4°C. After centrifugation for 5 minutes at 13,000 rpm, the supernatant was stored as nuclear extract. We followed the manufacturer's instructions for the Western blotting (enhanced chemiluminescence; Amersham). A digital photograph was obtained, and the density of each band was calculated with QuantityOne software (Bio-Rad).
Immunoglobulin and cytokine measurements.
Human and murine IgG was measured by enzyme-linked immunosorbent assay (ELISA) according to the manufacturer's instructions, using a kit from Immunology Laboratories. Human CD154 and murine CD154 were measured using an ELISA kit from R&D Systems and an ELISA kit from PromoCell, respectively, according to the manufacturers' instructions. Human and murine cytokines were measured using flow cytometry–based cytokine bead array systems (BD Biosciences).
Calcium flux experiments.
Murine lymphocytes (5 × 106) were isolated, diluted in RPMI 1640 with 1% heat-inactivated FCS and stained with fluorescent anti-CD4 and anti-CD8 antibodies as described above. The cells were then washed and incubated for 30 minutes at 37°C with 1 mg/ml of Indo 1-AM (Molecular Probes). After washing the cells again, the samples were run for 45 seconds in an LSRII flow cytometer (BD Biosciences), and anti-CD3 antibody (10 mg/ml) was then added. The samples were run on the flow cytometer for another 45 seconds, and hamster anti-rat crosslinking antibody (SouthernBiotech) was added. Following that step, the samples were run for 5 minutes. The ratio of violet to blue emission of Indo 1-AM, which is directly proportional to the free cytosolic Ca2+, was then recorded and analyzed using FlowJo software version 9.0.1 (Tree Star).
Cells were isolated from the peripheral blood of SLE patients and healthy controls. In addition, murine cells were isolated from the lymph nodes and spleens of mice. The cells were stained for 20 minutes at room temperature using an antibody conjugated with a fluorescent dye, washed, and then analyzed by flow cytometry using the LSRII flow cytometer. In all experiments, at least 10,000 events were recorded and analyzed (FlowJo software version 9.0.1).
Anti–NF-ATc2, anti–NF-ATc1, antiactin, horseradish peroxidase (HRP)–conjugated anti-rabbit antibody, HRP-conjugated anti-goat antibody, HRP-conjugated anti-mouse antibody, and hamster anti-rat IgG were purchased from Santa Cruz Biotechnology. Anti-human and anti-mouse CD3 and CD28 were purchased from BioLegend. Anti-human CD3, CD8, and CD4, anti-mouse CD3, CD8, and CD4, anti-mouse CD154, and anti-human CD154 fluorescent antibodies were purchased from BD Biosciences. Fluorescent (allophycocyanin) anti-mouse CD86 was purchased from eBioscience. Annexin V was obtained from BioLegend and propidium iodide from Invitrogen. Tartaric acid was purchased from Ricca, and tacrolimus and dipyridamole were purchased from Sigma-Aldrich.
The analysis was done using Graph Pad Prism 5.0 software. The unpaired (or paired, where appropriate) 2-tailed t-test was used. P values less than 0.05 were considered significant.
Calcium flux is robust and NF-ATc1 is overexpressed in MRL/lpr lymphocytes.
The full repercussions of the robust calcium mobilization following the engagement of TCR (1) in human SLE T cells are unclear. To answer this question, we used lymphocytes from MRL/lpr mice with established nephritis (15 weeks old) and from control MRL/MpJ mice. Engagement of CD3 with an anti-CD3 antibody followed by a crosslinking antibody resulted in higher and earlier calcium flux in MRL/lpr T cells as compared to control T cells, with a mean ± SD time to peak calcium content of 32.8 ± 14.3 seconds in CD3+CD4–CD8– cells from MRL/lpr mice, 97.2 ± 18.2 seconds in CD4+ cells from MRL/lpr mice, and 228.1 ± 11.1 seconds in CD4+ cells from MRL/MpJ mice (P < 0.0001 for each comparison versus T cells from MRL/MpJ controls) (Figures 1A and B).
Among the major T cell subtypes, CD4–CD8– cells displayed the most robust response. Notably, CD3+CD4–CD8– cells displayed calcium influx prior to the addition of the crosslinking antibody, probably due to the fact that lipid rafts are preaggregated on the surface membranes of these cells (4, 13).
Next, we wanted to determine whether MRL/lpr T cells express higher levels of NF-AT than do control mouse T cells. Indeed, we found that NF-ATc1 was increased in the cytoplasmic and the nuclear fractions of freshly isolated lymphocytes from MRL/lpr mice as compared to those from MRL/MpJ control mice (Figure 1C). In contrast to patients with SLE (2), the levels of NF-ATc2 in T cells from MRL/lpr mice were not increased as compared to the levels in T cells from control mice (data not shown).
We also compared calcium flux and NF-ATc1 levels in naive T cells isolated from young (9-week-old) MRL/lpr mice with preclinical disease and from MRL/MpJ mice. We found that naive CD4+ cells from MRL/lpr mice had a faster calcium flux than those from MRL/MpJ mice and that both groups had very low levels of NF-ATc1. This finding suggests that while the calcium hyperresponsiveness is an early feature of lupus T cells, the accumulation of NF-ATc1 happens gradually as the T cells acquire an effector phenotype.
Overall, our data suggest that MRL/lpr T cells show a hyperactive phenotype akin to that of human SLE T cells. Interestingly, this phenotype is prominent in the double-negative population of MRL/lpr T cells, a population that is greatly expanded in these mice and is known to produce IL-17A rather than IL-2 and interferon-γ (IFNγ).
MRL/lpr T cells help B cells to produce immunoglobulin in a calcineurin-dependent manner.
NF-AT activation leads to increased transcription of an array of genes, such as CD154, a costimulatory molecule that is important for T cell–directed autoantibody production by B cells (8, 14), in patients with SLE. To assess the role of the calcineurin/NF-AT pathway in T cell–dependent B cell antibody production, we cultured sorted CD3+CD4+ and CD3+CD4–CD8– cells from 15–17-week-old MRL/lpr mice and control mice with B cells from control mouse in the presence of anti-CD3/anti-CD28 antibodies. As shown in Figure 1D, only T cells from MRL/lpr mice provided help to normal B cells for the production of immunoglobulin. There was no statistically significant difference between CD4–CD8– and CD4+ T cells from MRL/lpr mice in their ability to stimulate IgG production by normal B cells. This T cell–directed B cell antibody production was completely abrogated in the presence of the calcineurin inhibitor tacrolimus. These results expand our previous observations in patients with SLE and suggest that the calcineurin/NF-AT pathway is important in the T cell–dependent overproduction of immunoglobulin in SLE.
Dipyridamole inhibits calcium-dependent signaling events in SLE T cells.
It was recently reported that dipyridamole, a drug widely used to limit platelet activation, inhibits the calcineurin/NF-AT pathway (10). We therefore wanted to determine whether dipyridamole inhibits the production of inflammatory cytokines and costimulatory molecules by human SLE T cells. First, normal human T cells were activated with anti-CD3/anti-CD28 antibodies for 18 hours, and the levels of soluble CD154 in the supernatant were measured. As shown in Figure 2A, dipyridamole at a concentration of 50 μM prevented the secretion of CD154 by T cells (P = 0.04) similar to the effects of tacrolimus. In contrast to its effects on T cells, dipyridamole did not affect B cell activation, as measured by the expression of CD86 on B cells (Figure 2B). Tacrolimus, on the other hand, partially inhibited B cell activation (Figure 2B).
Given these results, we then asked whether dipyridamole could limit the expression of CD154 by SLE T cells and, consequently, the production of immunoglobulin by B cells. Indeed, as shown in Figure 3A, dipyridamole at a concentration of 50 μM blocked the secretion of CD154 by human SLE T cells almost as efficiently as tacrolimus (mean ± SD 396 ± 125 pg/ml of soluble CD154 in tartaric acid–treated samples [n = 12], 31 ± 13 pg/ml in tacrolimus-treated samples [n = 5], and 53 ± 31 in dipyridamole-treated samples [n = 9]; P = 0.0172 for tartaric acid versus dipyridamole treatment). Similarly, as shown in Figure 3B, dipyridamole decreased the expression of surface CD154 on activated human SLE T cells, suggesting that the decrease in CD154 in the cell culture supernatants is the result of a decrease in the expression of CD154 by activated SLE T cells and not the result of decreased CD154 shedding (mean ± SD percentage of CD154+ activated T cells 6.8 ± 1.6% in tartaric acid–treated samples versus 2.1 ± 0.7 in dipyridamole-treated samples [n = 9 per group]; P = 0.0172).
In a different set of experiments, we evaluated the effect of dipyridamole on SLE T cell–directed B cell immunoglobulin production. Peripheral blood mononuclear cells (PBMCs) from SLE patients were activated in vitro with anti-CD3/anti-CD28 antibodies in the presence of tartaric acid (control) or dipyridamole (50 μM). As seen in Figure 3C, after 13 days of activation, the immunoglobulin produced by SLE PBMCs in response to T cell activation was decreased by more than 50% in the presence of dipyridamole (P = 0.0098; n = 11 per group).
Dipyridamole inhibits cytokine production by SLE T cells.
To determine whether dipyridamole inhibits the production of cytokines, SLE T cells were incubated with either 50 μM dipyridamole or 0.4% tartaric acid and then activated with anti-CD3/anti-CD28 antibodies for 18 hours. The levels of cytokines produced were determined in culture supernatants. As shown in Figure 4A, IFNγ, tumor necrosis factor α (TNFα), and IL-2 production were inhibited by dipyridamole. Similarly, the secretion of IL-17 (Figure 4B) and IL-6 (Figure 4C) was inhibited by dipyridamole treatment. Of note, only SLE T cells, but not normal T cells (data not shown), produced IL-17.
Since dipyridamole inhibited T cell cytokine production, we postulated that it might also affect cell proliferation and/or apoptosis. Dipyridamole at a concentration of 50 μM did not increase apoptosis in T cells from either healthy individuals or SLE patients. In contrast, 50 μM dipyridamole prevented the proliferation of human SLE T cells to a level similar to that of tacrolimus (data available upon request from the author). These results suggest that dipyridamole prevents SLE T cell activation, proinflammatory cytokine production, and cell proliferation.
Dipyridamole inhibits the production of IL-6 and delays the emergence of nephritis and skin disease in lupus-prone mice.
We next wanted to determine whether dipyridamole alters disease progression in lupus-prone mice. For this analysis, lupus-prone MRL/lpr mice (11 weeks old; n = 5 per group) were injected intraperitoneally 3 times a week with either dipyridamole 50 mg/kg or tartaric acid 0.4%. This dose of dipyridamole is the mouse equivalent of a dose of 4 mg/kg in humans (15), a daily dose that is used routinely in clinical practice (oral administration). We found that during 3 weeks of treatment, the dipyridamole-treated animals had slower progression of nephritis than did control animals, as measured by the development of proteinuria and pyuria (Figures 5A and B). None of the dipyridamole-treated animals developed skin disease, whereas 60% of the animals treated with vehicle developed inflammatory skin ulcers (Figures 5C and D). At the end of the 3-week treatment, only 40% of the tartaric acid–treated control mice were still alive, whereas none of the dipyridamole-treated animals had died. At the end of the treatment period, the size of the spleen and lymph nodes did not differ in the two treatment groups.
The splenocyte population in the dipyridamole-treated group was characterized by a lower percentage of CD3+ cells than in the controls (mean ± SD 43.3 ± 6.5% versus 87.7 ± 0.6% in controls; P = 0.02). Among the CD3+ cells, the CD3+CD4–CD8– population was significantly decreased in the dipyridamole-treated group, with a reciprocal relative increase in the numbers of CD4+ and CD8+ cells (Figure 6A). We observed a relative increase in the total B cell population in the dipyridamole-treated group. Total IgG and double-stranded DNA antibody levels in sera from the two groups were not different. IL-6 levels in the serum of vehicle-treated animals were increased significantly as compared to those in the dipyridamole-treated group (Figure 6B). Levels of IFNγ, IL-2, and TNFα did not differ between the two treatment groups or between the treatment groups and normal control mice (data not shown). It should be noted that the animals in the control group that did not survive beyond the third week and, presumably, were the sickest (higher proteinuria levels and worse skin ulcers) were not included in these comparisons. We therefore conclude that beyond its anti–T cell activity in vitro, dipyridamole may be a useful anti-SLE medication at doses that are within the therapeutic window for patients.
The aberrant phenotype that characterizes SLE T cells results in a skewing of the T cell response toward heightened effector functions and deficient regulatory functions (16). We have previously shown that SLE T cells display increased calcium influx following TCR engagement. The exact functional repercussions of this observation are fully explored for the first time in this study. We showed that T cells from MRL/lpr mice, especially the double-negative population, are characterized by robust calcium influx following in vitro activation and increased levels of NF-ATc1. We have previously shown that NF-ATc2 is increased in T cells from SLE patients (2), whereas herein, we report that NF-ATc1 is increased in T cells from MRL/lpr mice. Given the similar functions of the 2 isomers, we assume transferability between the data derived from human and murine studies.
Previously thought to be anergic, CD3+CD4–CD8– cells do not produce IL-2 and IFNγ, but are expanded in patients with SLE and in lupus-prone mice, and they produce IL-17 in an IL-23–dependent manner (9, 17, 18). In the present study, we showed that double-negative T cells are characterized by robust and early calcium flux and that they provide help to B cells in a calcineurin-dependent manner. This is consistent with previous data derived from human SLE T cell lines (19, 20).
Given these data, we evaluated the usefulness of inhibiting the NF-AT pathway in patients with SLE. The traditional calcineurin inhibitor tacrolimus was highly effective in blocking NF-AT activation in SLE T cells (Figure 1D), but it has significant side effects, including hypertension and renal toxicity, when used in patients with SLE. We provide herein definitive evidence that dipyridamole, the newly recognized inhibitor of calcineurin–NF-AT interaction (10), can suppress human and murine SLE T cell function.
Dipyridamole is a phosphodiesterase inhibitor that leads to an increase in intracellular cAMP. It is used in various conditions, including strokes and certain types of nephritis, primarily for its antiplatelet and vasodilator (21). More recently, it has been recognized that dipyridamole has antiinflammatory properties (e.g., inhibits the production of metalloproteinases) and that it can act synergistically with corticosteroids (22).
Herein, we show that dipyridamole suppressed the production of cytokines and costimulatory molecules by SLE T cells and limited their proliferation. We also show in our in vivo experiments that lupus-related pathology was suppressed in dipyridamole-treated lupus-prone mice. It should be noted, however, that the effect of dipyridamole on T cell proliferation was not as striking in vivo as the in vitro results would have suggested, possibly reflecting pharmacodynamic and/or pharmacokinetic limitations. Another explanation could be that dipyridamole also limits the production of cytokines, such as IL-2 and TNFα, that may play a role in immune homeostasis (e.g., IL-2 is important in Treg cell activity). Further studies in animals with lupus are needed to define the role and, more importantly, the timing of dipyridamole administration in order to minimize its potential effects on immune homeostasis.
B cell activation does not seem to be affected by dipyridamole, which is probably the reason why immunoglobulin levels and B cell numbers in the MRL/lpr mice were not affected by this treatment. It is possible, though, that if dipyridamole is administered to lupus-prone animals early in their disease course, it may suppress the generation and activation of B cells by blocking T cell help.
In conclusion, the demonstrated ability of dipyridamole to block cytokine production, to block T cell–mediated help to B cells, and to control pathology in the MRL/lpr mouse urge the performance of definitive clinical trials in patients with SLE in which dipyridamole is added to standard treatment for the disease.
All 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. Dr. Kyttaris 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. Kyttaris, Zhang, Kampagianni, Tsokos.
Acquisition of data. Kyttaris, Zhang, Kampagianni.
Analysis and interpretation of data. Kyttaris, Zhang, Kampagianni, Tsokos.
We would like to thank Dr. Madhukar Shinde for help with the cell separation, culture, and Western blots.