High-level serum B-cell activating factor and promoter polymorphisms in patients with idiopathic thrombocytopenic purpura


Florian Emmerich, Institute for Transfusion Medicine, Charité University Medicine, Augustenburger Platz 1, 13353 Berlin, Germany. E-mail: florian.emmerich@charite.de


Idiopathic thrombocytopenic purpura (ITP) is an autoimmune disorder in which platelets are opsonised by autoantibodies and destroyed by macrophages. Therefore, ITP represents a prototype of a B-cell-mediated autoimmune disorder. B-cell activating factor (BAFF) is a member of the tumour necrosis factor family and an important regulator of B-cell development. BAFF levels were determined in serum samples from 53 patients with ITP. Serum BAFF levels in patients with an active ITP were increased when compared with the healthy control group (median 1620 pg/ml vs. 977 pg/ml; P < 0·001). Moreover, immunosuppressive treatment was associated with strongly suppressed BAFF levels (median 629 pg/ml; P < 0·01). In addition, a polymorphic site was detected in the BAFF promoter region (−871) that appeared to occur more frequently in ITP patients than in healthy persons. This promoter variant was associated with very high BAFF levels in ITP patients. Our data suggest that BAFF is an important pathogenetic factor in the development of ITP.

Idiopathic thrombocytopenic purpura (ITP) is an autoimmune-specific disorder in which platelets are opsonised by autoantibodies directed against platelet glycoproteins followed by their destruction by macrophages (George, 2006). The reason for the production of autoantibodies remains unknown and the clinical course is extremely variable. Some patients remain asymptomatic, whereas others develop life-threatening bleeding episodes. Moreover, therapy including prednisolone, intravenous immunoglobulin G, anti-D and splenectomy is not always effective, and only one-third of adult patients achieve long-term remission (Bellucci et al, 1988; Berchtold & McMillan, 1989).

B-cell activating factor (BAFF) belongs to the family of tumour necrosis factor (TNF) ligands and is expressed by several cell types, including monocytes, macrophages, neutrophils, dentritic cells and T lymphocytes (Schneider et al, 1999; Mackay & Browning, 2002; Schneider & Tschopp, 2003). It binds to three receptors: BAFF receptor B-cell maturation antigen and transmembrane activator and calcium modulating cyclophilin ligand (CAML) interactor, which are primarily expressed on B lymphocytes (Thompson et al, 2001; Mackay & Mackay, 2002). BAFF plays a crucial role in B-cell development, survival and immunoglobulin production (Harless et al, 2001; Hsu et al, 2002; O'Connor et al, 2004). Moreover, excess amounts of BAFF results in the rescue of self-reactive B cells from anergy, thus implicating a role in the development of autoimmunity (Thien et al, 2004). Mice overexpressing BAFF display increased B-cell numbers and immunoglobulin levels as well as clinical features similar to that observed in patients with systemic lupus erythematosus (SLE) (Mackay et al, 1999; Gross et al, 2000; Khare et al, 2000). In accordance with such data, some studies demonstrated elevated BAFF serum levels in patients with systemic autoimmune diseases such as SLE, rheumatoid arthritis (RA) and Sjögren syndrome (Cheema et al, 2001; Zhang et al, 2001; Groom et al, 2002; Mariette et al, 2003; Tan et al, 2003). However, patients with primary biliary cirrhosis and autoimmune diabetes displayed normal BAFF levels, discounting a general association between autoimmune diseases with elevated BAFF (Mackay et al, 2002).

A possible role of BAFF in ITP has not yet been addressed. To clarify the function of BAFF, we performed a prospective study on 53 patients with chronic ITP. BAFF levels were elevated in the serum of patients with active ITP and strongly decreased following immunosuppressive treatment. Furthermore, we report on two polymorphisms in the BAFF gene promoter in patients with ITP, which were associated with very high BAFF levels.

Materials and methods

Study population

Serum BAFF levels were analysed in 53 patients diagnosed with ITP (Table I). ITP was diagnosed in accordance with the guidelines of the American Society of Haematology (George et al, 1996). Of these patients, 37 were female patients and 16 were male patients, with a median age of 46 years. Forty-four patients had an active ITP, whereas nine patients were inactive. Five patients suffered from underlying diseases (SLE and RA) (10·6%). During the course of the study, participating patients did not receive drugs known to induce thrombocytopenia. All patients were treated at the Charité University Hospital Berlin, Germany. Eighteen patients with an active ITP were pretreated with different immunosuppressive therapies (Table I). Healthy control persons (n = 24) were voluntary blood donors. Approval from the institutional review board was obtained. All patients and healthy control persons gave informed consent to the studies.

Table I.   Patients characteristics.
PatientSexAge (years)Platelet count (×109/Bl)BAFF level (pg/ml)Medication at blood sampling (mg /d)SplenectomyDuration of disease (years)Prior therapies
  1. BAFF, B-cell activating factor; Pred., prednisolone; Dexa, dexamethasone; AZA, azathioprine; MTX, methotrexate; MMF, mycophenolate mofetil; IVIgG, intravenous immunoglobulin G; CYC, cyclophosphamide; anti-D, intravenous anti-D.

 1M3430576Pred. 7·5; AZA 150, 2Pred.
 2M7617·5336Pred. 50, AZA 250 7Pred., AZA, MMF, IVIgG
 3F4058538Pred. 5 10Pred., anti-D
 4F6555605Pred. 10, AZA 150, MMF 2000 2Pred. MMF, IVIgG
 5F7832930Pred. 10 15Pred., IVIgG, anti-D
 6F3660803Pred. 10 5Pred., anti-D
 7F79115770Pred. 10 10Pred., AZA
 8m4259493Pred. 30 3None
 9F6926585Pred. 30, AZA 100 1None
10F37186335Pred. 20, MMF 500 2Pred. AZA
11M6640240Pred. 80 1Pred.
12F48164686Dex: 4 26Pred., AZA, MMF, IVIgG
13M5286875Pred. 10, AZA 100+5Pred., IVIgG
14F6975680Pred. 5, MMF 1000 5Pred., MTX., MMF, IVIgG
15F46100976Pred. 10 5Pred., AZA
16M50150247Pred. 20 mg, MTX 10 10Pred., MTX
17M6875653Pred. 10, MMF 4000+3Pred., MMF
18F351503071Pred. 2, MMF 2000 1Pred.
19M37385821None 2None
20F3626767None 15IVIgG
21M7884916None 5None
22F31861274None 2None
23F65521721None 18Pred., MMF,
24F33411820None 3None
25F7841402None 50Pred., IVIgG, anti-D
26F32781307None 3Pred., AZA
27M301041655None 3None
28F28721747None 2None
29M65231690None+7Pred., AZA, IVIgG
30F51453715None 6Pred., AZA
31F41191699None 23Pred., Cyc, IVIgG
32F68332746None 10Pred., IVIgG
33F3714891None 7Dexa.
34M41891172None 2None
35F66351427None 12Pred., AZA, CYC, IVIgG
36F53801585None 20Pred.
37F51522199None 8Pred., MMF
38F73142209None 20IVIgG, AZA
39M54601018None 3IVIgG, Pred. AZA
40F3261957None 5Dexa, AZA
41M69491213None 3None
42F87153153None 11Pred., MMF
43F32671073None 5Pred., MMF
44M4382067None 15Pred., AZA, IVIgG,
46M401701144None 1IVIgG
47F31160635None 4IVIgG
48F46487657None 5None
49F293271356None 2None
50F332601022None 2None
51F21121753None 1None
52F651521370None 10Pred., Dexa, IVIgG, anti-D
53F26147545None 10Pred., Dexa, IVIgG

Measurement of BAFF serum levels

B-cell activating factor levels were measured using a specific enzyme-linked immune assay (R&D Systems, Minneapolis, MN, USA) in accordance with the manufacturer's recommendations. Serum BAFF levels were calculated from a standard curve generated from recombinant human BAFF (R&D Systems). The Mann–Whitney U-test was used to compare the values. P < 0·05 were considered to be significant. Spearman's rank correlation was used to examine the relationship between two variables. Values are given as the median (range).

Reverse transcription-polymerase chain reaction

Total RNA was extracted from fresh whole-blood samples using a high pure RNA isolation kit (Roche Diagnostics, Mannheim, Germany). The cDNA was synthesised under the following conditions: 1 μg of total RNA was incubated with 4 μl of 5x incubation buffer, 1 mmol/l of each deoxyribonucleoside triphosphate (dNTP), 5 mmol/l of MgCl2, 50 U of Rnase inhibitor and 500 ng of Oligo-(dT)18 primer. After a denaturation step (70°C for 5 min), 20 U of Moloney murine leukaemia virus (MMuLV) was added. The mixture was incubated for 1 h at 37°C. The reaction was stopped by heating at 70°C for 10 min. The primers used for amplification of full-length BAFF cDNAs were 5′-CAG GAC ATC AAC AAA CAC AG-3′ and 5′-TAG AGG TAC AGA GAA AGG GA-3′. Polymerase chain reaction (PCR) conditions consisted of an initial denaturation step of 90 s at 95°C, 30 cycles of 40 s at 95°C, 40 s at 55°C and 60 s at 72°C. PCR buffer conditions were as follows: 1·5 μmol/l of MgCl2, 200 μmol/l of each dNTP, 10 pmol/l of each primer and 1 U of Taq polymerase (InViTek, Berlin, Germany).

Genomic PCR

Genomic DNA from fresh whole-blood samples was extracted using an automated workstation (Genovision, Vienna, Austria). To amplify the BAFF promoter, the following primers were used: 5′-CAC AGG TCC ACC AAG TCA ACA ACA GA-3′ and 5′-ATC ACT ACT TGA ACT TTG AAG GTT GG-3′ (He et al, 2003). The conditions for genomic PCR were identical to the amplification conditions used for reverse transcription (RT)-PCR.


Polymerase chain reaction products were purified using the Invisorb Spin PCRapid Kit (InViTek) and directly sequenced by the chain termination technique using fluorescence-labelled ddNTPs (BigDye; PE Applied Biosystems, Darmstadt, Germany). Sequencing reactions were analysed on an automated DNA sequencer (3130; Applied Biosystems) and the resulting sequences were compared with the sequences of the BAFF mRNA and the BAFF promoter.


Elevated BAFF serum levels in patients with chronic ITP

Serum BAFF levels in patients with untreated ITP were significantly higher [median 1620 pg/ml (range 767–5821)] than in healthy persons [median 977 pg/ml (range 470–1382); P < 0·001]. In inactive cases (n = 9), BAFF levels were observed to be similar to those of the healthy control group, suggesting a correlation between BAFF levels and disease activity [median 1022 pg/ml (range 545–3614)]. Patients receiving immunosuppressive treatment demonstrated normal or even lower BAFF levels in comparison with the control group [median 629 pg/ml (range 240–3071); P < 0·01)] (Fig 1).

Figure 1.

 Sera from 24 healthy persons and 26 untreated patients with an active idiopathic thrombocytopenic purpura (ITP), nine patients with inactive disease and 18 patients with treated ITP were analysed by enzyme-linked immunosorbent assay for B-cell activating factor (BAFF). Bars represent median BAFF levels.

A follow-up investigation was performed on 15 patients for at least 4 months following the initial visit. Ten patients demonstrated the marginal changes in BAFF expression during the observation period. In the remaining five cases, BAFF levels varied by more than 50% for at least one time-point during the observation period. This finding might be explained by the onset of immunosuppressive treatment.

Single nucleotide polymorphisms in the BAFF gene promoter

B-cell activating factor transcripts and the corresponding gene promoter were analysed in 17 patients with ITP. One polymorphism at position −871 T/C was observed in 11 patients, with five homozygous (−871 T/T) and six heterozygous alterations (−871 T/C) found. One of these patients carried an additional polymorphism at position −661 (A/G). The −871 T/T genotype was detected in only one of 10 healthy individuals (10%), whereas the heterozygous form was present in four healthy donors (40%).

To investigate whether the −871 variant correlates with enhanced BAFF serum levels, BAFF levels from patients carrying the different −871 genotypes were compared. The −871 T/T genotype was found to be associated with very high levels of BAFF (median 1721 pg/ml), whereas normal levels of BAFF were found in patients carrying the T/C genotype (median 703 pg/ml) or the C/C genotype (median 1037 pg/ml) (Fig 2).

Figure 2.

 Association between B-cell activating factor (BAFF) −871 promoter polymorphisms and serum BAFF levels in 17 patients with idiopathic thrombocytopenic purpura. Bars represent median BAFF levels.


This study showed that, for the first time, serum BAFF levels were elevated in patients with active ITP. In inactive patients, normal levels of BAFF were observed. Elevated BAFF has been previously described in systemic autoimmune diseases, such as SLE and RA. What these disorders have in common is the engagement of both auto-reactive B-cell and T-cell clones which affect a variety of tissues. In contrast, ITP is primarily mediated by self-reactive B lymphocytes. This characteristic classifies ITP as a prototype of a B-cell-mediated autoimmune disorder (Cines & Blanchette, 2002). However, a role for platelet-directed cytotoxic T cells has also been reported (Olsson et al, 2003).

The mechanism underlying the relationship between excess BAFF and thrombocyte levels remains unclear. BAFF has been shown to enhance both the expression of CD19 and the ability of the B-cell receptor to enhance and phosphorylate CD19 (Hase et al, 2004). Moreover, BAFF mediates the maturation of autoreactive B cells (Thien et al, 2004). Thus, excess BAFF may play a role in promoting the accumulation of self reactive B-cell clones directed against thrombocytes.

The molecular basis for enhanced BAFF levels remains unknown. Recent studies have reported on a number of polymorphisms (Kawasaki et al, 2002). In this study, a polymorphism at position −871 T/C was observed in ITP patients, and was recently reported in patients with SLE and B-cell chronic lymphocytic leukaemia (Kawasaki et al, 2002; Novak et al, 2006). The TT genotype was found in five of 18 patients with ITP (28%), but only in one of 10 healthy controls (10%). Moreover, the −871 T/T genotype was found in patients with the highest BAFF levels (median 1721 pg/ml). This finding is in accordance with a recent report demonstrating enhanced transcriptional activation of the BAFF gene (Novak et al, 2006). More extended studies are needed to clarify the prevalence of these and comparable alterations in ITP patients. Moreover, functional analyses with different promoter variants should clarify whether such polymorphisms influence BAFF expression in ITP patients.

Current treatment strategies in ITP consist of the administration of immunosuppressive substances, such as glucocorticoids, methotrexate and azathioprine. Our data show that BAFF was suppressed to normal levels in patients receiving these substances. Therefore, it is possible that transcriptional downregulation of BAFF represents a novel mechanism by which these substances execute their therapeutic effects. Recently, it has been shown that the BAFF promoter is controlled by nuclear factor-kappa B (NF-κB), with NF-κB activity suppressed by several immunomodulators including corticosteroids and methotrexate (Auphan et al, 1995; Wahl et al, 1998; Majumdar & Aggarwal, 2001). Immunosuppressive substances may act indirectly on BAFF expression via downregulation of these factors. However, it cannot be excluded that certain drugs bind directly to the BAFF promoter, thereby suppressing its activity.

The cellular source of excess BAFF remains unknown. A local production was suggested by the observation of accumulated BAFF in the synovial fluid of patients with RA (Cheema et al, 2001). In accordance with this, BAFF was found to be highly expressed in the salivary glands of SLE patients (Groom et al, 2002). As ITP can be effectively treated by splenectomy, it is tempting to speculate that BAFF is primarily produced by a cell population homing in the spleen. Future studies that focus on analysing the effect of splenectomy on BAFF expression and autoantibody production should clarify this issue.

Approximately, 60% of adult ITP patients develop therapy–refractory chronic disease within 12 months (Godeau et al, 2002). This illustrates the need for the development of novel therapeutic strategies. The effectiveness of BAFF inhibitors has been shown in animal models, and clinical trials have commenced in patients with SLE and RA (Baker, 2004; Ramanujam & Davidson, 2004). Hence, selective targeting of BAFF in patients with high levels of BAFF could be considered as a novel therapeutic strategy.