Aberrant Expression of a Proliferation-Inducing Ligand Underlies Autoimmune Mechanisms in Immune Thrombocytopenia

Purpose To study the relationship between surface membrane-bound APRIL and ITP. Methods The peripheral blood of all subjects, 50 patients diagnosed with ITP and 25 healthy controls, was collected. Flow cytometry was used to detect the expression of membrane-bound APRIL on immune cells and platelets. ELISA was used to detect the content of soluble APRIL in plasma. Results Membrane-bound APRIL was only expressed on the surface of platelets in both ITP patients and controls. APRIL expression on the platelet surface was significantly lower in newly diagnosed (P < 0.001) and chronic (P < 0.001) ITP patients than in controls. Platelet surface APRIL level was significantly enhanced in patients with complete remission after treatment (P = 0.02) but not in those with no response after treatment. Platelet surface APRIL level in ITP patients was negatively correlated with serum APRIL level (r = −0.09765, P = 0.0424). Conclusions Platelet surface APRIL may play a key immunoregulative role. Platelet surface APRIL is likely to be one source of the excessive serum APRIL in ITP patients. The effectiveness of treatment may be measured by determining the platelet surface APRIL levels in ITP patients.


Introduction
Primary immune thrombocytopenia (ITP) is an antibodymediated autoimmune disease characterized by accelerated platelet destruction and aberrant platelet production. While it may be asymptomatic for some patients, others can suffer mild, moderate, or even severe bleeding. The pathophysiology of ITP is complex but may be caused by multiple factors such as impaired megakaryopoiesis leading to reduced platelets [1], the proliferation of autoreactive T cells particularly the upregulation of Th17 and downregulation of Tregs [2], increased antiplatelet autoantibodies by B lymphocytes [3], and loss of peripheral immune tolerance [4].
APRIL, also known as tumor necrosis factor superfamily [TNFSF] 13, is secreted by a range of cells including myelocytes [5], activated platelets [6], and activated T and B cells [5]. Structurally and functionally similar to B cell-activating factor (BAFF), APRIL is believed to play a pivotal immunoregulatory role in patients with various autoimmune diseases including [7][8][9]. Increased levels of APRIL have been found in different autoimmune diseases such as systemic lupus erythematosus [10], rheumatoid arthritis (RA) [11], Sjogren's syndrome (SS) [12], and ITP [13,14]. Previous studies have shown that the pathogenesis of ITP may be related to an increased level of a proliferation-inducing ligand. In ITP patients, the serum level of APRIL cannot be lowered by glucocorticoid therapy which can significantly reduce BAFF [15,16], but may be effectively reduced by vitamin D3 [15]. Consensus has not been reached as to the origin of the increase in serum APRIL level. Gu et al. proposed that APRIL originates from increased gene expression. By studying the predisposing gene in familial or sporadic ITP, Peng et al. show that the p.G76S mutation on the TNFRSF13B gene in ITP patients may lead to enhanced binding ability of APRIL ligands to B cells as well as increased BAFF levels [17]. In contrast, Kamhieh-Milz et al. conclude that APRIL has nothing to do with increased gene expression [15].
Since APRIL is found to contribute to platelet production as an autocrine growth factor produced by megakaryocytes [18], its role for ITP patients can be essential. However, studies on the relationship between APRIL and ITP are few and far between. Of the few published articles on the topic, three are particularly relevant to the current paper. The first is Gu et al. which suggests that plasma APRIL level in ITP patients was significantly higher than that in healthy controls and correlated closely with platelet count [14]. Gu and his colleagues also attempted to understand the reason of the increased APRIL in the plasma of ITP patients by performing realtime quantitative polymerase chain reaction to determine the mRNA expression of APRIL. Based on the result of mRNA expression of APRIL by PBMNCs, they speculated that PBMNCs may be a source of the excessive APRIL. While Gu and his colleagues have shed important light on the role of APRIL for ITP patients, more research is needed on this topic for three reasons.
Firstly, they overlooked surface membrane-bound APRIL, since they only examined soluble APRIL in plasma. There are two forms of APRIL: soluble and surface membrane-bound [19]. It is believed that surface membrane-bound APRIL is hardly detectable due to its extremely low expression [19]. This is likely the reason that these researchers focused only on soluble APRIL. However, more recent studies detected surface APRIL in patients suffering from diseases such as leukemia and RA [20][21][22]. Therefore, researchers interested in the relationship between APRIL and ITP need to take surface membrane-bound APRIL into consideration. Secondly, the method used by Gu and his colleagues to explore the reason behind increased APRIL in the plasma of ITP patients, real-time quantitative polymerase chain, seems somewhat circuitous. It may be more straightforward to detect that by flow cytometry. Thirdly, they did not compare the changes in APRIL levels before and after treatment in ITP patients. This information is vital because it shows disease activity in relation to changes in APRIL.
Kamhieh-Milz et al. [15], another study on the relationship between APRIL and ITP, compared serum levels of BAFF and APRIL in active ITP patients before and after treatment with GC. The former decreased significantly by glucocorticoid treatment alone, while the latter only responded well after adding vitamin D3 to prednisolone. While GC has remained the primary therapy for ITP patients, it is frequently used in combination with other medications such as DMN, IVIg. In some cases, doctors even give up GC due to uncontrollable bleeding. Probably, for this reason, Álvarez Román et al., a third study relevant to the current paper, investigated the effect of TPO on the level of plasma APRIL in ITP patients and revealed that this medication significantly reduced plasma APRIL level [21]. However, it remains questionable whether the findings of Kamhieh-Milz et al. and Butta et al. can be readily generalized to patients who receive medications other than treatment in their studies.
In this study, we examine the relationship between APRIL and ITP by focusing on surface membrane-bound APRIL with two objectives in mind: (1) to find out whether APRIL is expressed on the surface of immune cells and platelets in ITP patients and normal controls; (2) if yes, how surface membrane-bound APRIL respond to treatment; and (3) to explore how surface membrane-bound APRIL is related to soluble APRIL. Treatment response was determined based on a consensus of Chinese haematologists [23]. That is, complete response (CR) was defined as a platelet count ≥100 × 109/l and absence of bleeding; no response (NR) was rated as platelet count <30 × 109/L or less than 2-fold increase of baseline platelet count or bleeding.

Materials and Methods
The control group comprised of healthy, age-matched subjects. The baseline characteristics of the patients and controls are summarized in Table 1.

Preparation of Platelet Suspension.
The EDTA blood was centrifuged at 150 × g for 10 minutes to obtain platelet-rich plasma samples, which were then transferred to polypropylene test tubes. PRP was centrifuged at 2,300 × g for 5 minutes to obtain platelet pellets, which were then resuspended in ammonium oxalate (1%) for erythrocyte lysis. After centrifuging at 1100 × g for 5 minutes to discard the supernatant, the pellet was washed twice in phosphate-buffered saline (PBS) containing EDTA (0.009 mol/l Na2EDTA,0.0264 mol/l Na2HPO4, 0.07 mol/l NaCl). The concentration of platelets in PBS-EDTA buffer was 1 × 10 7 /ml [24][25][26]. Peripheral blood was collected into EDTA-anticoagulant vacuum tubes. PBMCs were isolated using lymphoprep density gradient centrifugation (Haoyang, Tianjin, China).

Determination of Plasma APRIL Levels.
Peripheral blood was collected into EDTA-anticoagulant vacuum tubes. Plasma was separated from blood cells by centrifugation (2600 g, 5 min) and was frozen at −80°C until analysis. PBMCs were isolated using lymphoprep density gradient centrifugation (Haoyang, Tianjin, China). APRIL levels in patient and control plasma were detected by antigen capture enzyme-linked immunosorbent assay using a commercial kit (BioLegend) according to the manufacturer's instructions. Plasma APRIL concentration was calculated from a standard curve generated using purified human APRIL supplied with the kit.

Statistical
Analysis. Data are expressed as mean ± standard error of the mean and were analyzed using SPSS v.13.5 (SPSS Inc., Chicago, IL, USA). Factor variance analysis or t-test was used for comparison of sample mean within the group. The Pearson correlation coefficient or Spearman rank correlation was used to analyze the correlation between variables. The results before and after treatment were analyzed by paired t-test. P < 0:05 was considered statistically significant. All experimental analyses were performed using the GraphPad Prism 7.0 software.

Results
The APRIL level of both CR and NR ITP patients is shown in Table 2. Below, we elaborate the major findings.

APRIL Was Only Expressed on the Surface of Platelet.
Compared with isotype controls, surface APRIL could be clearly detected only in platelets in all ITP patients and normal controls. The mean fluorescence intensity (MFI) of platelet surface APRIL staining in ITP patients and normal controls was 631 and 1241, respectively. In contrast, little surface APRIL was observed in other immune cells. Specifically, the MFI of APRIL on the surface of CD19 + B cells in ITP patients and normal controls was 230 and 225, on that of CD3 + T cells being 224 and 272 and on that of CD14 + monocytes being 184 and 282, respectively. Figure 1 shows 3 Journal of Immunology Research surface APRIL and isoform staining of platelets (Figure 1(a)), B cells (Figure 1(b)), T cells (Figure 1(c)), and monocytes ( Figure 1(d)).

Platelet Surface APRIL Expression Was Downregulated in ITP Patients
. Compared with isotype controls, surface APRIL can be clearly detected on the surface of gated platelets in all ITP patients and normal controls (Figure 2(b)). A graph comparing APRIL stained MFI on platelet surfaces from the 25 healthy controls and 50 ITP patients is provided (Figure 2(c)). APRIL expression on the platelet surface was significantly reduced in newly diagnosed (P < 0:001) and chronic (P < 0:001) ITP patients relative to controls.
Correlation analysis of the MFI of the APRIL on platelet surface of ITP patients and the platelet count suggests that they were positively correlated with each other (r = 0:1811, P = 0:0013) (Figure 2(d)).

Effective Treatment Enhanced Platelet Surface Level.
The changes of platelet surface APRIL were monitored before and after treatment in active ITP patients (n = 18). In patients with complete remission, APRIL MFI on platelet was higher after treatment than before treatment in the CR group

Discussion
In this study, we investigated the relationship between APRIL and ITP, generating some interesting results. We began with exploring whether and how APRIL was expressed on the surface of immune cells and platelets in ITP patients and normal controls. We found that APRIL was saliently expressed only on the surface of platelets in both ITP patients and normal controls. Compared with controls, platelet surface APRIL was significantly reduced in ITP patients. This trend was obviously reversed in ITP patients with complete remission after treatment. In contrast, for patients who did not respond to treatment, the platelet surface APRIL level remained similar before and after treatment. We interpret our findings below. We were triggered by recent studies that detected surface APRIL in patients who suffered from diseases such as leukemia and RA [20][21][22] to ask whether membrane-bound ARPIL expresses on the cell surface of ITP patients. To answer this question, we measured the expression of APRIL on the surface of platelets and peripheral mononuclear cells including T cells, B cells, monocytes, and dendritic cells from ITP patients and normal controls by flow cytometry. We found that membrane-bound APRIL was only expressed on the surface of platelets in both ITP patients and controls. Meanwhile, surface platelet APRIL in patients remained significantly lower than that in normal controls. Thus, we join in Weldon et al. to challenge the proposition that APRIL was hardly expressed on cell surface [18,19]. In the study of Weldon and his colleagues, the surface form of APRIL has been identified in lymphoid and bone marrow mononuclear cells and macrophages in synovial tissue of RA patients [22,27]. They explained the mechanism for cell surface APRIL expression by referring to alternative APRIL splice forms and TWE-PRIL [28]. While that study revealed that RA patients manifested higher levels of surface form of APRIL than normal controls, ITP patients in this study showed a lower level of platelet surface APRIL than normal controls. We speculate that this contrast may result from the difference in disease types.
Our finding that the platelet surface APRIL level in the ITP patients fell significantly below the controls may be explained by five reasons. Firstly, it may be related to the imbalance of furin, since APRIL is processed intracellularly by furin convertase. While research on ITP on this issue is lacking, it has been reported that furin protease dysregulation is associated with increased levels of BAFF and APRIL in immune cells of patients with advanced atherosclerotic plaque [29]. The second possible reason is that some surface platelet APRIL shed into the plasma due to unknown pathophysiological reasons [14,17]. If this reason can be confirmed in the future, we can say that surface platelet APRIL is another source of soluble APRIL increase in plasma in ITP patients. The third possible reason is that megakaryocytes can transfer MHC class I molecules loaded with foreign antigen to proplatelets in vitro [30]. The fourth possible reason for the decrease of surface APRIL in the platelet of ITP patients is the preactivation of platelet, which can significantly reduce the expression of platelet surface markers with little change in platelet function [31]. An additional explanation for the plummet of platelet surface APRIL level in the ITP patients may be related to ADAM17 [32]. It has been reported that metalloproteinases ADAM17 can trigger the combination of BAFFR and TACI. Inhibition of ADAM17 can increase the expression level of BAFFR on germinal center B cells. In a similar vein, the inhibition of ADAM17 may also increase the expression level of APRIL in plasma. ADAM17 stimulates the shedding of platelet receptor (e.g., platelet-derived growth factor receptor β), TNF and its receptors (TNFR-1 and TNFR-2), and Glycoprotein V [33][34][35]. Since the mechanism of the cleavage effect of ADAM17 on APRIL remains unclear, further research is desirable.
Next, we explored how treatment affected the expression of surface APRIL in ITP patients. We found that, for CR patients, the platelet surface APRIL level after treatment was significantly higher than that before treatment. In contrast, for NR patients, platelet surface APRIL level remained similar before and after treatment. This finding seems to suggest that surface APRIL may also play a role in immunoregulation together with soluble APRIL, again challenging Lopez-Fraga et al. who suggested that only soluble APRIL has an immunoregulation role [18]. At present, the exact mechanism for the role of surface APRIL in immunoregulation remains unclear. Regardless of this, our finding may be clinically significant as the effectiveness of treatment may be measured by determining surface APRIL levels in ITP patients.
To sum up, membrane-bound APRIL was only expressed on the surface of platelets in both ITP patients and normal controls with platelet surface APRIL in the former significantly lower than that in the latter. Platelet surface APRIL was reversed by treatment for CR patients but not for NR ones. Platelet surface APRIL may be a source of the excessive APRIL in the plasma of ITP patients. Aberrant APRIL Non-responders APRIL (ng/ml) (f) Figure 3: Plasma APRIL levels in ITP patients. (a) Plasma APRIL levels were higher in chronic and newly diagnosed ITP patients than in normal controls (n = 25 per group), with no significant difference between the two patient groups. (b) APRIL concentration was negatively correlated with platelet count in ITP patients (n = 50). (c) Correlation between platelet surface APRIL and serum APRIL levels in ITP patients. (d-f) Differences in Plasma APRIL levels before and after treatment in the chronic ITP patient group (n = 25) (d) in the complete remission group (n = 10) (e) and the nonresponse group (n = 8) (f).