Three blocks are not enough—Blocking of the murine IgG receptor FcγRIV is crucial for proper characterization of cells by FACS analysis

While binding of IgG antibodies (IgG1, IgG2a, b or c, and IgG3 in the murine system) to Fc gamma receptors (FcγR) via their Fc part is crucial for the activity of therapeutic antibodies in vivo, it is a well-known cause of erroneous results in immunological assays, such as flow cytometric analysis. Thus, blocking of the IgG-binding sites of FcγRs has been established as a preventive procedure. In mice, four different FcγRs are expressed on a wide variety of immune cells. These include FcγRI (CD64), an activating receptor with high affinity to IgG that is able to bind monomeric IgG molecules, and the lower affinity receptors FcγRIIb (CD32b; inhibitory), FcγRIII (CD16; activating), and FcγRIV (activating), which bind immune-complexes rather than monomeric IgG.1, 2 To block the Fc-dependent interaction of antibodies with FcγRs on mouse cells, mAbs directed against murine FcγRIIb and FcγRIII are commonly used. One of the most frequently used FcγR-blocking antibodies is the rat mAb 2.4G2 3. Blocking of FcγRs I, IIb, and III is achieved by binding to a common nonpolymorphic epitope on the extracellular domains of the mouse FcγRIII/IIb receptors via the Fab regions 3, 4 and binding to FcγRI likely via the Fc part of the antibody 3. When we recently set out to analyze FcγRIII/CD16 expression on murine NK cells and other peripheral blood leukocytes of C57Bl/6 mice by flow cytometry, we were surprised to find two apparently NK1.1-positive CD11b-positive cell populations with either low or intermediate CD16 expression, respectively. In contrast, when using fcgr4−/− mice or if anti-CD16-FITC staining was carried out in the presence of an FcγRIV-blocking antibody, only the CD16low population was NK1.1 positive, whereas the CD16intermediate cells were identified as NK1.1-negative nonclassical monocytes. Thus, the apparent presence of NK1.1 on these monocytes in C57Bl/6 mice could clearly be explained by antigen-independent binding of the anti-NK1.1 conjugate to FcγRIV 5 on nonclassical monocytes (also called “resident” or Ly6C/Gr-1− CD62L− CD43+ CX3CR1high monocytes 6) via its Fc-fragment. Here, we initially verified that binding of anti-NK1.1-FITC to FcγRIV+ monocytes could be inhibited by the very blocking of this Fc receptor alone. Therefore, murine peripheral blood leukocytes were stained with anti-NK1.1-FITC (clone PK136) either without FcR blocking antibodies or after pretreatment with and in the presence of anti-FcγRIV antibody (clone 9E9) or an anti-CD16/32 mAb 2.4G2 (Fig. 1) (gating strategy see Supporting Information Fig. 1). In contrast to blocking of FcγRs IIb and III (and probably I) by 2.4G2, the non-antigen-specific binding of anti-NK1.1-FITC to nonclassical monocytes was efficiently abrogated by blocking of FcγRIV. To clarify whether this was a unique effect of the anti-NK1.1 conjugate or whether also other antibodies revealed binding to FcγRIV, we tested a wide array of antibodies commonly used for flow cytometric analysis. To exclude that different fluorochromes impact our analysis we focused on antibodies labeled with FITC. (For information on antibodies used in this work see Supporting Information Table 1.) Rather than using primary murine FcγRIV-expressing cells, which usually coexpress other FcγRs, we established a test system consisting of CHO cells that were stably transfected to express FcγRIV. Parallel staining of CHO-FcγRIV cells and peripheral blood leukocytes for FcγRIV revealed comparable changes in anti-FcγRIV fluorescence intensity for CHO-FcγRIV cells and nonclassical monocytes on a logarithmic scale (Supporting information—FcγRIV expression on nonclassical monocytes and CHO-FcγRIV cells and Fig. 2). This indicates that the CHO-FcγRIV system is feasible for investigation of FcγRIV-mediated binding since it will not result in exaggerated effects in flow cytometric analyses. Using serial dilutions of anti-NK1.1-FITC we could verify that—like nonclassical FcγRIV+ monocytes, but in contrast to untransfected CHO cells–FcγRIV-expressing CHO cells readily bound this antibody conjugate. This binding was abrogated by blocking with the FcγRIV-specific antibody 9E9 (Fig. 1B). Since anti-NK1.1-binding approached a plateau level at 2.5–5 μg/mL (Fig. 1C) all further experiments with CHO-FcγRIV cells were done using 2.5 μg/mL of the respective antibody conjugate. A notable fraction of other antibody conjugates revealed binding to CHO-FcγRIV but not to untransfected control cells. The results of the flow-cytometric analyses are depicted in Fig. 1D as Δlog(MFI), i.e. the difference of the logarithm of the MFI of antibody-decorated CHO-FcγRIV and CHO cells in the FITC channel. For example, a Δlog(MFI) = 1 meant that CHO-FcγRIV and CHO cells differed in their fluorescence intensities upon antibody binding by one log rank. Upon sorting of the results according to the isotypes of the respective antibodies, a pronounced isotype prevalence for Fc-mediated binding became obvious: whereas virtually all tested conjugates of murine IgG2a or b, rat IgG2b, and hamster IgG2 antibodies revealed binding to FcγRIV, such binding was low or absent for the tested mouse and rat IgG1 and IgM antibodies as well as Armenian hamster IgG1 and polyclonal sheep sera. The latter is also true for the rat IgG2a isotype that has a higher homology to mouse IgG1 than to mouse IgG2a/b (see detailed description in Supporting Information—Homologies of Immunoglobulin G subclasses in rat and mouse and Fig. 3). These results indicate that, in fact, binding to FcγRIV is a common feature at least for antibody isotypes related to mouse IgG2. The presence or absence of FcγRIV binding both in the CHO system and to primary FcγRIV-expressing cells reflects the binding affinities of mAbs representative of the various IgG subclasses from mouse, rat, and Armenian hamster, as quantified via surface plasmon resonance in this work (Supporting Information—Fig. 4) and by others (Supporting Information Table 2). To test whether this undesirable effect of non-antigen-specific binding can be abolished by commercially available Fc-blocking mAbs, CHO-FcγRIV cells were preincubated either with such antibodies or anti-FcγRIV clone 9E9. As shown in Fig. 2 (panels A and B) pretreatment of CHO-FcγRIV cells with the commonly used Fc-block antibody 2.4G2 prior to staining with representative FITC-conjugated antibodies of murine, rat, or hamster origin barely affected their antigen-independent binding. In contrast, this binding was efficiently blocked by 9E9. In addition, two other commercially available Fc-blocking antibodies (clone 93 and SeroBlock) were tested in parallel to 2.4G2 and 9E9 as blocking agents for anti-NK1.1-FITC binding on CHO-FcγRIV cells at different concentrations. None of the anti-CD16/32 Fc-blocks could efficiently abrogate Fc-mediated binding to FcγRIV even at high concentrations, whereas 9E9 largely abolished this binding even at low concentrations (Fig. 2C). Extending the analysis to a biologically more relevant setting, the blocking capabilities of 9E9 and commercially available anti-CD16/32 Fc-blocks were compared using FcγRIV+ nonclassical monocytes in peripheral blood. This analysis showed that—like on CHO-FcγRIV cells—2.4G2, SeroBlock, and clone 93 are rather inefficient in blocking Fc-mediated binding of the anti-NK1.1 conjugate to nonclassical monocytes, in contrast to 9E9 (Supporting Information Fig. 5A and B). Finally, various antibody-FITC conjugates that had been found to reveal unspecific binding in the CHO/CHO-FcγRIV system (and anti-FcγRIV mAb 9E9 as positive control) were tested for binding to FcγRIV+ nonclassical monocytes in peripheral blood. As depicted in Supporting Information Fig. 5C, FcγR-mediated binding of different FACS antibodies to these monocytes via FcγRI/II/III (blocked by 2.4G2) or FcγRIV (which could be blocked by 9E9) was detected for the majority of these antibodies. In summary, FcγRIV expression represents a prominent risk of erroneous results in assays such as FACS analysis as the commercially available Fc blocking antibodies barely affect Fc-mediated binding to FcγRIV. Hence, inclusion of FcγRIV blockade to common Fc-blocking strategies appears critical to improve the experimental quality of immunological assays, where such Fc-dependent binding causes false-positive results. This study was supported by a grant from the German Research Foundation (SFB 643) to F.N. The authors declare no commercial or financial conflict of interest. As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer reviewed and may be re-organized for online delivery, but are not copy-edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors. Figure S1. Figure S2. Figure S3 Figure S4 Figure S5 Table S1: Antibodies used in this study Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.