Detection of anti–Siglec‐14 alloantibodies in blood components implicated in nonhaemolytic transfusion reactions

In a previous retrospective study performed on samples from 66 blood components causing severe nonhaemolytic transfusion reactions (NHTRs), we showed that four components (6%) contained HLA antibodies and 12 components (18%) contained non-HLA leucocyte antibodies. It was notable that none of the leucocyte antibodies were against HNA-1–5. The majority of these antibodies reacted with either neutrophils only or with both neutrophils and monocytes (Matsuyama et al, 2008). These data suggest the presence of additional neutrophil antibodies that evoke immune responses and potentially cause NHTRs and emphasize the importance of clarifying the specificity of non-HLA leucocyte antibodies in blood components. Human Siglec-14 (Angata et al, 2006), a transmembrane protein with a cell-activating property, is expressed in neutrophils and monocytes. Some individuals do not express Siglec-14 due to genetic polymorphisms (Yamanaka et al, 2009) and may produce anti–Siglec-14 alloantibodies. We therefore propose that Siglec-14 may be one of the target antigens binding to non-HLA and non-HNA neutrophil alloantibodies implicated in NHTRs. Using flow cytometry, we analysed 479 healthy Japanese blood donors using an anti–Siglec-14 monoclonal antibody to determine the prevalence of subjects who did not express Siglec-14. The prevalence was found to be 10·6% (51/479). We then examined the number of blood components causing NHTR possessed anti–Siglec-14 alloantibodies. To detect anti–Siglec-14 alloantibodies, we established a cell line panel in which Siglec-14 cDNA was transduced using a retroviral vector to confer stable transgene expression in K562 cells. This technique was similar to the method we reported previously to establish cell lines expressing HNA (Yasui et al, 2007). We then compared the frequency of anti–Siglec-14 alloantibodies in the plasma of healthy donors and blood components that were administered to the NHTR patients (Table I). We noted that the frequency of anti–Siglec-14 alloantibodies in NHTR components was higher than in the plasma of healthy donors (P < 0·01). These results suggest that anti–Siglec-14 alloantibodies are novel neutrophil alloantibodies that may be involved in NHTRs. When we classified the NHTR cases further according to the AABB (American Association of Blood Banks) system, donor plasma from 5 of 29 cases of transfusion-related acute lung injury (TRALI) contained the IgG class of anti–Siglec-14 alloantibody. The frequency of these alloantibodies was significantly higher in TRALI cases than in NHTR cases (P < 0·01, Table I). Siglec-14 expression in the five TRALI patients was not determined as we were unable to obtain blood samples from these patients. We identified 4 blood donors with anti–Siglec-14 alloantibodies in their plasma (Table I). Three of these donors did not express Siglec-14, and hence it is conceivable that the lack of Siglec-14 may have caused expression of the alloantibody. The other donor expressed a comparable amount of Siglec-14 antigens and we speculate that this donor had a sequence polymorphism in the Siglec-14 antigen that resulted in the production of anti–Siglec-14 alloantibodies. Further investigations on the Siglec-14 polymorphism is needed to confirm this possibility. Because people with a transfusion history are permanently banned from donating blood in Japan, it is likely that pregnancy, rather than a transfusion, resulted in the formation of the antibody. This possibility is consistent with the fact that the four donors were all female. We also examined whether anti–Siglec-14 alloantibodies caused neutrophil activation in vitro. This involved incubating Siglec-14–expressing whole blood cells with a test plasma sample and examining the amount of Mac-1, a neutrophil activation marker, in neutrophils (Yasui et al, 2008; Hirayama, 2010). Each test plasma sample was examined using six whole blood samples and the average mean fluorescence intensity (MFI) of Mac-1 in neutrophils then calculated. Figure 1A shows the results of 10 healthy donor plasma samples without anti–Siglec-14 alloantibodies (C1–C10) and three TRALI plasma samples containing anti–Siglec-14 alloantibodies (T1–T3). Although we detected anti–Siglec-14 alloantibodies in five TRALI plasma samples, we were unable to carry out the in vitro activation assay in two TRALI plasma samples because of insufficient samples. A comparable amount of Mac-1 was observed in neutrophils treated with healthy donor plasma (C1–C10) and those treated with PBS. We determined that the TRALI plasma samples contained activated neutrophils when the average MFI of Mac-1 in six test neutrophils exceeded a specific value after stimulation in the in vitro assay. This value was defined as the average Mac-1 MFI for test neutrophils after stimulation with the plasma of 10 different healthy donors +5SD. On the basis of this definition, T1 and T2 could activate neutrophils, whereas T3 could not. However, none of the three TRALI plasma samples initiated activation of neutrophils that did not express Siglec-14 (data not shown). Involvement of anti–Siglec-14 alloantibodies in neutrophil activation. (A) Whole blood cells from healthy laboratory personnel were incubated at 37°C for 30 min with PBS as a negative control (NC), 0·1 μmol/l fMLP as a positive control (PC), healthy blood donor plasma not containing anti–Siglec-14 alloantibodies (C1–C10) or TRALI plasma containing anti–Siglec-14 alloantibodies (T1–T3). The white blood cells were then incubated on ice for 15 min with PE-conjugated anti–Mac-1 monoclonal antibody. Following removal of red blood cells and washing of the remaining cells, the MFI of Mac-1 expression in Siglec-14+ neutrophils was measured by flow cytometry. The horizontal line in each category indicates the mean value. Dashed lines in each figure indicate the mean + 5SD of MFI of Mac-1 in six blood samples stimulated with the plasma of 10 healthy blood donors. (B) Whole blood cells from Siglec-14+ healthy laboratory personnel (cells A–F) were incubated with TRALI plasma (T1) or healthy blood donor plasma (white bars). T1 plasma was preincubated at 37°C for 30 min with either KY-mock (absorption 1 × 106 KY-mock cells repeated three times, dark grey bars) or KY–Siglec-14 (absorption 1 × 106 KY–Siglec-14 cells, repeated three times, pale grey bars), or was not incubated (black bars). The MFI of Mac-1 expression in neutrophils was analysed by flow cytometry. To confirm whether neutrophils were activated by anti–Siglec-14 alloantibodies, we performed absorption experiments in which Siglec-14–expressing whole blood cells were stimulated with T1 plasma, preabsorbed by either KY–Siglec-14 cells or KY-mock cells (Fig 1B). After three series of absorption experiments using KY–Siglec-14 cells, we were unable to detect even a trace amount of anti–Siglec-14 alloantibodies in T1 plasma by the flow cytometry assay (data not shown). Treatment with T1 plasma preabsorbed by KY–Siglec-14 cells caused 53% neutrophil activation compared to non-preabsorbed plasma. On the other hand, preabsorption by KY-mock cells caused 83% neutrophil activation compared to no preabsorption (Fig 1B). We observed similar results for release of heparin-binding protein using the same experimental procedures described above (data not shown) (Yasui et al, 2008). We reported previously that neutrophils are activated by anti-HLA Class I alloantibodies via their surface Fc receptors CD16 and CD32 (Yasui et al, 2008), and accordingly, we examined whether anti–Siglec-14 alloantibodies activated neutrophils via Fc receptors. Pretreatment with blocking monoclonal antibodies for Fc receptors (Tamm & Schmidt, 1996, Aicher et al, 2000) caused 95% neutrophil activation compared to no Fc blocking (data not shown). These results indicate that anti–Siglec-14 alloantibodies are responsible for activation of neutrophils, not only by an Fc receptor-dependent mechanism, but more predominantly by an Fc receptor-independent mechanism. Yamanaka et al reported that stimulation by LPS, a ligand of Siglec-14, resulted in Siglec-14 initiating TNF-α secretion in transduced monocytic cell line. This effect was dependent on the interaction of Siglec-14 with activating adaptor protein DAP-12 (Yamanaka et al, 2009). These results suggest that DAP-12 may mediate the activation of neutrophils by anti–Siglec-14 alloantibodies. We conclude that in addition to anti-HLA and anti-HNA alloantibodies, anti–Siglec-14 alloantibodies might also be involved in the development of some NHTRs, especially TRALI.