Background Circulating plasmablasts/plasma cells and activated B and T cells are increased in systemic lupus erythematosus (SLE). Interleukin (IL)-6 induces differentiation of B cells into antibody-forming cells and of T cells into effector cells.
Objective To examine the hypothesis that blocking IL-6 would reverse some of the immune abnormalities present in SLE.
Methods Fifteen patients with SLE with mild-to moderate disease activity were treated with biweekly infusions of tocilizumab, a humanised anti-IL-6 receptor monoclonal antibody for 12 weeks. Lymphocyte subsets (analysed by flow cytometry) and serum immunoglobulin levels were compared at baseline and at weeks 6 and 12.
Results Tocilizumab decreased activated T and B cells, the frequency of CD27highCD38highIgD− plasmablasts/plasma cells and IgD−CD27+ post-switched memory B cells as well as IgG+ memory B cell, whereas it increased the frequency of IgD+CD27− antigen-inexperienced B cells. Among antigen-inexperienced IgD+CD27− B cells, CD38low mature naïve B cells increased significantly and CD38IntermediateCD5+ pre-naïve B cells showed a decreasing trend, whereas CD38highCD5+ transitional type 1 B cells did not change. Most of the changes occurred in patients who had abnormal values at baseline. IgG, IgA, IgG1 and IgG3 serum levels decreased albeit within the normal range. The frequency of CD4+CD45RA+CCR7+ naïve T cells increased.
Conclusions In vivo blockade of the IL-6 receptor decreases lymphocyte activation and restores B and T cell homoeostasis by either blocking differentiation and/or trafficking in patients with SLE and leads to normalisation of the abnormal B and T cell subsets seen at baseline.
Statistics from Altmetric.com
Systemic lupus erythematosus (SLE) is characterised by increased T and B cell activation and abnormal distribution of lymphocyte subsets. Several abnormalities in peripheral B cell homoeostasis have been associated with active SLE. B cell lymphopenia is common in active SLE with a decrease in naïve B cells, expansion of circulating immunoglobulin secreting cells (plasmablasts/plasma cells)1–4 and memory B cell subsets.5 An increased proportion of transitional type 1 and CD5 pre-naïve B cells has been reported in subjects with SLE.6 ,7
Perturbation of T cell homoeostasis has also been identified in patients with SLE and has been hypothesised to play an important role in disease progression and pathology. Activated T cells are elevated in active disease.8–11 An increase of memory and a decrease of naïve CD412 ,13 and CD8 T cells14–17 together with an expansion of TCRαβ+CD4−CD8− double-negative T cells18 ,19 and abnormalities in regulatory T cells have also been reported in patients with SLE.20
Interleukin (IL)-6 is a pleiotropic cytokine with a wide range of biological activities which may contribute to many of these abnormalities.21 The critical role of IL-6 as a B cell differentiation factor has been extensively described.22–24 Plasmablasts/plasma cells represent <1% of peripheral blood B cells in normal subjects. Normal circulating plasmablasts/plasma cells strongly express IL-6 receptor (IL-6R) and IL-6 is a critical survival factor for these cells.25 ,26 Neutralising IL-6 or blocking IL-6 signalling with anti-IL-6R monoclonal antibodies inhibited plasma cell differentiation and immunoglobulin secretion by induction of plasmablast apoptosis.25 ,27 SLE B cells spontaneously secrete IL-6 and constitutively express IL-6R, inducing an autocrine hyperactivity of B cells.28 Additionally, it was shown that IL-6 plays a critical role in anti-DNA antibody production.29
IL-6 also induces both the growth and the differentiation of T cells into memory/effector cells.30–32 In addition, the apoptosis of in vivo activated human memory T cells could be partially inhibited by IL-6.33
Tocilizumab is a humanised anti-IL-6 receptor monoclonal antibody. Recently, we reported the results of a pilot study of tocilizumab in SLE, showing preliminary data for clinical efficacy and normalisation of the increased number of circulating plasmablasts/plasma cells.34 To further evaluate the immunological effects of in vivo blockade of IL-6 in patients with SLE, we performed a detailed analysis of the impact of tocilizumab treatment on the distribution of peripheral lymphocyte subsets in subjects with SLE.
Materials and methods
Patient samples and study design
All patients fulfilled the American College of Rheumatology classification criteria for SLE and had mild to moderate disease activity. Patients were enrolled in a pilot clinical study of tocilizumab (Chugai Pharma, USA, San Diago, CA). They were treated with one of three doses (2 mg/kg, n=4; 4 mg/kg, n=6; 8 mg/kg, n=5) of tocilizumab, every 2 weeks for 12 weeks. Patients were receiving stable, low-dose prednisone (<0.3 mg/kg/day) but no other immunosuppressive treatment. One of 16 patients was withdrawn after the first dose because of leucopenia; the other 15 patients (13 women and two men, age 38.5±10.2 (mean±SD)) completed the study and are included in this analysis. Control blood samples were obtained from 10 female and four male healthy blood donors, (age 41.5±13 (range 20–54)). There were no differences between female and male donors in any of the analyses. This study was approved by the institutional review board of the National Institute of Arthritis and Musculoskeletal Diseases. Written informed consent was obtained from all subjects.
Peripheral blood mononuclear cells (PBMCs) were isolated from heparinised whole blood (20 ml) by Ficoll–Hypaque density-gradient centrifugation and serum samples were obtained at pretreatment (week 0), and at weeks 6 and 12.
Flow cytometric analysis
Immunofluorescence labelling for flow cytometric analysis was performed by staining PBMCs with PerCP/Cy5.5-labelled anti-CD19, CD4 and CD8, fluorescein isothiocyanate-labelled anti-CD3, IgD, HLA-DR, CD21, IgA, TCRαβ and CD45RA; APC-labelled anti-CD38, CD5, CD8, CD44, CD11c and CD45RO; phycoerythrin-labelled anti-CD27, CD25, CD38, CCR7, CD69, CD123, CD56, IgM and IgG antibodies (BD Pharmingen, San Diego, California, USA) and mouse isotype control IgG1 coupled with each of the fluorochromes as negative controls. PBMCs were stained in phosphate-buffered saline/1% bovine serum albumin and 5% fetal bovine serum on ice for 30 min. Cells were washed twice and fixed in phosphate-buffered saline/1% paraformaldehyde. Flow cytometric analysis was performed using a FACSCalibur flow cytometer (BD Biosciences, San Jose, California, USA) and CellQuest software (BD Biosciences). A total of 70 000 events were collected for each analysis.
Activated B cells and T cells were identified by expression of CD69, HLA-DR and CD44. Each lymphocyte subset and its markers are shown in supplementary table S1. B cells were identified based on CD19 expression. CD19-negative lymphocytes were identified as non-B cells. Additional B cell subsets were identified as CD27−IgD+ antigen-inexperienced B cells, CD27+IgD− post-switch memory B cells35 and CD27highIgD−CD38high plasmablasts/plasma cells. Antigen-inexperienced IgD+CD27− B cells are divided into three subsets: CD38low mature naïve B cells, CD38IntermediateCD5+pre-naïve B and CD38highCD5+ transitional T1 B cells.7 Cell surface immunoglobulins, including IgM, IgG and IgA, were evaluated on B cells. T cells were identified based on CD4 or CD8 expression and additional subsets were identified as naïve (CCR7+CD45RA+CD45RO−), central memory (CCR7+CD45RA−CD45RO+) and effector memory (CCR7−CD45RA−CD45RO+) in CD4 and CD8 T cells, whereas terminal effector memory T cells (CCR7−CD45RA+CD45RO−) were identified only in CD8 T cells.36 TCRαβ+CD4−CD8− cells were identified as double-negative T cells.
Additionally, CD123+CD11c−HLA-DR+plasmacytoid dendritic cells, CD56+CD3−natural killer (NK) and CD56+CD3+ NK T cells were evaluated.
Frequencies of lymphocyte subpopulations were calculated using Flowjo software (Flowjo, LLC Ashland, Oregon, USA). Lymphocyte subsets of patients with SLE (week 0) were compared with values of healthy controls by unpaired t test and results given as mean plus SD. Patients were divided into two groups based on their baseline values: the ‘within normal limits (WNL)’ group was within the mean of healthy control±1 SD, whereas the ‘outside normal limits (ONL)’ group had values that were outside the means±1 SD. Pretreatment values were compared with weeks 6 and 12 values and statistical significance was assessed by the repeated-measures analysis of variance with the Bonferroni–Dunn adjustment for multiple comparisons. For unadjusted analyses p values <0.05 were considered significant, whereas a p<0.0167 was considered significant after the Bonferroni–Dunn adjustment.
Baseline comparison of patients with SLE and healthy controls
Patients and healthy controls were of similar age, gender and race. The mean modified (excluding hypocomplementaemia) SELENA-SLEDAI (Safety of Estrogens in Lupus Erythematosus National Assessment-Systemic Lupus Erythematosus Disease Activity Index) score was 8 (range 4–15) for the patients at baseline.34 There were no significant differences in total lymphocyte numbers and the absolute numbers and frequencies of major B and T cell subsets (CD3, CD4, CD8 and CD19 cells) between patients with SLE and healthy controls (data not shown). Consistent with previous reports, the proportions of CD27+IgD− post-switch memory B cells and CD27highIgD−CD38high plasmablasts/plasma cells were increased in SLE compared with controls (p<0.05). There was a trend to lower proportions of CD27−IgD+antigen-inexperienced B cells (p=0.06) (figure 1A,B). Consistent with a previous report, among the CD27−IgD+ antigen-inexperienced B cells, the frequency of CD38low mature naïve B cells was significantly (p<0.05) lower whereas the proportion of CD38Intermediate pre-naïve B cells was significantly (p<0.05) higher in SLE. There was also a trend toward higher proportions of CD38high transitional T1 B cells in SLE (p=0.09).
In absolute cell numbers, CD27+IgD+ pre-switch memory B cells were decreased in SLE but there was no significant difference in other subsets compared with controls (see supplementary figure S1A and B).
There were no significant differences in the frequencies of CD4 and CD8 T cell subsets (not shown). Patients with SLE had increased activation of B cells as demonstrated by a significantly higher frequency of CD69 B cells, both in the pre-switch and post-switch compartments and increased density of CD44, gauged by mean fluorescence intensity. There were no significant differences in the frequencies of CD69 non-B, CD4+HLA-DR+ and CD8+HLA-DR+ cells between controls and patients with SLE (figure 1C,D), but similarly to B cells, non-B cells displayed increased densities of CD44 (figure 1D). The absolute numbers of CD69 non-B, CD4+HLA-DR+ and CD8+HLA-DR+ cells were significantly higher in SLE (p<0.05) (see supplementary figure S1C). There were no differences in the baseline levels of double-negative T, NK-T, NK cells and plasmacytoid dendritic cells, between controls and SLE (not shown).
A closer examination of the data suggested that for many subsets, there were two distinct populations in the SLE group at baseline: some were within the normal range (similar to the controls) and others were outside the normal range. Therefore, we divided the patients with SLE into two groups at baseline and analysed them separately. Those whose values were within one SD of the mean of the controls were considered to be WNL whereas those who were outside this range were considered to be ONL. A patient might have had normal values in some and abnormal values in other subsets. To evaluate whether tocilizumab had a differential effect on those two subgroups we analysed both subsets separately and the cohort as a whole.
No apparent differences were noticed among the three different doses of tocilizumab, therefore the results are presented for all three doses as a whole.
Tocilizumab treatment decreases the activation state of PBMCs
Tocilizumab treatment significantly decreased the frequencies (p<0.0167) (figure 2A,B) and absolute numbers (see supplementary figure S2A,B) of activated CD69 B cells both in the pre-switch (IgD+) and post-switch (IgD−) compartments, mainly by affecting those with an abnormally elevated frequency at baseline. The frequency of CD69-activated non-B cells also decreased, primarily because of a decrease among those with an abnormally elevated frequency at baseline (p<0.0167) (figure 2C). A similar decrease was noted in the frequency of activated CD4 and CD8 T cells using HLA-DR as an activation marker (figure 2D,E) and this tendency was also shown in the absolute numbers (see supplementary figure S2C–E).
Expression levels of CD44 in B cells significantly decreased with tocilizumab and this effect was largely confined to the subsets with abnormally elevated CD44 at the baseline (figure 2F). Expression levels of CD44 in non-B cells were above the controls in all patients at baseline and decreased with tocilizumab treatment (figure 2G).
Effect of tocilizumab treatment on abnormalities in peripheral B cell subsets
As we have previously reported34 tocilizumab significantly decreased the frequency of abnormally elevated plasmablasts/plasma cells and CD27+IgD− post-switch memory B cells. In contrast, it increased the abnormally reduced proportion of antigen-inexperienced B cells. The high percentage of CD27highIgD− plasmablasts/plasma cells and CD27+IgD− post-switch memory B cells decreased and CD27−IgD+ antigen-inexperienced B cells increased with tocilizumab treatment (figure 3A–D). Within the antigen-inexperienced B cell subset, CD38low mature naïve B cells increased significantly (p<0.0167), whereas CD38Intermediate pre-naïve B cells showed a decreasing trend in patients with baseline values ONL (figure 3F,G). Transitional T1 B cells did not change with the treatment (not shown).
These changes were primarily caused by the effect of tocilizumab on those patients who were ONL at baseline with only a minimal effect on the group with normal values. For absolute cell numbers, abnormally elevated plasmablasts/plasma cells decreased and abnormally reduced CD27−IgD+ antigen-inexperienced B cells increased (see supplementary figure S3A,B). In the antigen-inexperienced B cell subset, abnormally reduced CD38low mature naïve B cells increased significantly, whereas the number of pre-naïve B cells did not change significantly (supplementary figure S3C,D).
Naïve B cells express IgM and IgD, whereas IgG and IgA are markers of class-switched memory B cells. To provide further evidence that tocilizumab leads to a shift to a more naïve B cell phenotype, we examined the expression of cell surface immunoglobulin isotypes by B cells. The frequency of IgG-expressing B cells decreased significantly by week 12 with the treatment (p<0.0167), whereas the frequency of IgM-expressing B cells showed a rising trend (figure 4A–C). IgA expression did not change significantly (not shown).
Changes in immunoglobulin isotypes and IgG subtypes in serum
There was a small but statistically significant decrease in serum levels of IgG34 and IgA, with all isotypes remaining within the normal range. There was a decreasing trend among the IgG subclasses but only IgG1 and IgG3 decreased significantly during treatment (figure 5). As previously reported, the proportional decrease of anti-dsDNA antibody levels was several-fold higher than with total IgG levels, suggesting a preferential effect on autoantibody-producing cells.34
Effect of tocilizumab treatment on peripheral T cell subsets
Tocilizumab primarily affected CD4 T cells with no significant effect on CD8 T cells. The overall effect on CD4 T cells was similar to that of B cells, resulting in a shift towards a more naïve phenotype. The frequency of naïve CD4 T cells increased (p<0.0167) with treatment (figure 6), primarily in patients whose baseline values were below the normal range (p<0.0167). A similar pattern was seen in the absolute number of this subset (see supplementary figure S4). No change in double-negative T cells was noted with tocilizumab (not shown).
Effect of tocilizumab treatment on other subsets
There were no significant changes in the frequency or number of plasmacytoid dendritic cells, NK cells and NK T cells (not shown).
These data indicate that in vivo blockade of the IL-6 receptor decreases lymphocyte activation, alters B and T cell homoeostasis in patients with SLE and leads to normalisation of the abnormal B and T cell subsets observed at baseline. Specifically, tocilizumab treatment led to an increase in naïve B cells and a decrease in memory B cells and plasmablasts/plasma cells. Similarly, the frequency and absolute numbers of abnormally reduced naïve CD4 T cells increased. Importantly, most of these changes occurred in subjects with abnormal pretreatment proportions of these subsets.
IL-6 is a homoeostatic cytokine which promotes the terminal differentiation of B cells23 ,24 and contributes to plasma cell survival.26 Both memory B cells and plasmablasts/plasma cells strongly express the IL-6 receptor. Therefore, blocking the IL-6 receptor is expected to prevent plasma cell differentiation or plasma cell survival and immunoglobulin secretion by induction of plasmablast and memory B cell apoptosis.25 ,27 Additionally, anti-IL-6 antibodies inhibited pokeweed mitogen-induced immunoglobulin production from PBMCs without affecting cell proliferation, indicating that IL-6 is required for antibody production in B-cells.37 SLE B cells may have autocrine hyperactivity caused by the spontaneous secretion of a large amount of IL-6 and constitutive expression of the IL-6R. Consistent with this, IL-6R blockade or anti-IL-6 inhibited the production of polyclonal immunoglobulins and anti-DNA autoantibodies by lupus cells in vitro.28 ,38 IL-6 is involved in the proliferation and differentiation of helper T cells. IL-6 enhances T cell proliferation and anti-IL-6 antibody and IL-6R blockade suppresses anti-CD3/CD28 antibody-induced CD4 T cell proliferation.39 In this study the significant decrease in memory B cells, plasmablasts/plasma cells and serum immunoglobulin levels, together with the previously reported decline in dsDNA antibody levels in the same cohort,34 are consistent with the expected direct effect of IL-6 blockade.
The decrease in the memory B cell compartment is consistent with those reported with tocilizumab in patients with rheumatoid arthritis.40 Unlike in rheumatoid arthritis, however, we also observed a significant decrease in plasmablast/plasma cells, probably because of the more pronounced abnormality of these sells in SLE. The naïve peripheral lymphocyte pool is also dysregulated in lupus patients, as reflected by the lower frequency and numbers of CD27−IgD+ B cells. Remarkably, both the numbers and frequency of CD27−IgD+ B cells normalised in response to tocilizumab. Notably, we found that the changes in the IgD+CD27− B cells were attributed to the recovery of reduced CD38low normal mature naïve B cells and to a lesser degree to a decline in CD38Intermediate pre-naïve B cells.
The effect of tocilizumab on T cell subsets was predominantly noted in the CD4 T cells. As with the B cells, tocilizumab led to a shift from memory to more naïve phenotype.
CD44 is involved in the migration of T cells to target organs41 and contributes to the pathophysiology of autoimmune diseases.42 The inhibition of T cell activation by blocking CD44 led the significant improvement of disease intensity in lupus-prone mice.43 Similarly to previous observations,16 ,17 in this cohort, expression of CD44 was abnormally high at baseline and decreased remarkably with tocilizumab treatment. Our data indicate that IL-6 blockade may lead to a reduction of pathogenic trafficking of lymphocytes by decreasing CD44 expression.
The patients in this study represent the milder end of the spectrum of disease activity in lupus. Therefore, it was reassuring that the immunological changes were seen primarily in patients who had abnormal subsets before treatment with no significant effect on those who had normal values.
It is unlikely that purely placebo effects would lead to such widespread changes consistent with the biological functions of IL-6. It is likely that the immunological effects of tocilizumab would be more pronounced in a population with higher disease activity and more immunological abnormalities at baseline.
The overall decrease in activated T and B cells supports a broad immunomodulating role of tocilizumab. Further studies are needed to better understand whether the effect of tocilizumab is caused by a direct effect on B and T cells or is an indirect effect mediated by other cytokines such as BAFF44 or a reflection of improvement in disease activity. Since tocilizumab administration was the only therapeutic intervention in this study our data suggest that many of the T and B cell abnormalities seen in SLE may be linked to abnormal production of IL-6 and can be reversed by blocking the IL-6 receptor. Although we cannot exclude the possibility that some of the changes observed are due to improvement in disease activity and only indirectly related to IL-6 blockade, larger studies with a comparator group and responders and non-responders in both arms may be able to distinguish the effect of treatment from the effect of disease improvement.
The data presented here indicate that in vivo blockade of the IL-6 receptor significantly improved the abnormal B and T cell homoeostasis in lupus. Together with the previously reported clinical data they provide a rationale to further evaluate tocilizumab in SLE.
We thank Drs Ruth Fritsch-Stork, Gary Sims, Rachel Ettinger and Xavier Valencia for their helpful advice in designing the antibody panels for the flow cytometry. The study was done under a Collaborative Research and Development Agreement between the National Institute of Arthritis and Musculoskeletal and Skin Diseases and Chugai, Pharma USA and was supported by the intramural research programmes of the National Institute of Arthritis and Musculoskeletal and Skin Diseases and the National Institute of Dental and Craniofacial Research.
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Files in this Data Supplement:
Contributors YS: planning the study, collecting, analysing and interpreting data and writing the manuscript. CY, T-YP: data collection, manuscript review; RF (deceased): planning of study, data collection; PL: planning the study, interpreting data and drafting the manuscript; GGI: planning the study, analysing and interpreting data and writing the manuscript.
Competing interests None declared.
Ethics approval Institutional review board of the National Institute of Arthritis and Musculoskeletal Diseases and the National Institute of Diabetes Digestive and Kidney Diseases.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.