Behold Cytometrists: One Block Is Not Enough! <scp>Cyanine‐Tandems</scp> Bind <scp>Non‐Specifically</scp> to Human Monocytes

Abstract

Recently, we made the intriguing observation that peripheral blood mononuclear cells (PBMCs) from healthy donors, especially the monocytes, stained slightly positive for the macrophage mannose receptor (CD206). These experiments were performed using a mouse anti-human CD206 PE-Cy7 antibody (BioLegend, clone 15–2). However, it is well established in the literature that monocytes do not express CD206 (1-3). This discrepancy prompted us to perform additional experiments to further investigate the phenomenon. The results from a titration experiment, which initially raised our concern, showed CD206 staining of monocyte-derived macrophages (MDMs) (as expected), but also positive monocytes and lymphocytes (not expected) (Fig. 1A,B). Surprisingly, when repeating the experiment on PBMCs from the same donor using the same anti-CD206 clone (but APC conjugated, BioLegend), the results showed CD206 expression on MDMs, but not on PBMCs (Fig. 1C). Data from another experiment using another clone (anti-CD206 PE, clone 19–2, BD Biosciences) also showed CD206 expression on MDMs, but not on PBMCs (Fig. 1D). This led us to suspect non-specific antibody binding of the PE-Cy7 conjugated anti-CD206 antibody, which was not removed by Fc-receptor blocking with human IgG (100 μg/ml) (4). We investigated whether the phenomenon was caused by antibody aggregates, but centrifugation (5,000xg for 5 min) of the CD206 PE-Cy7 antibody before use did not alleviate the erroneously positive PBMCs (data not shown). Thus, we suspected that the PE-Cy7 fluorochrome was causing the issue. This was in agreement with a previous screening of all murine IgG1 monoclonal isotype controls available in our lab (n = 19) on PBMCs from one healthy donor (2 μg/ml, Supporting Information Table S1). Here, the term “isotype control” is used for murine monoclonal antibodies against a target irrelevant for human samples. This screening showed that all the IgG1 isotype controls displaying high level of non-specific binding to monocytes were conjugated to PE-Cyanine tandems (especially PE-Cy5 and PE-Cy7). Several other IgG1 isotype controls showed limited or no non-specific binding, for example, PerCP-Cy5.5, AlexaFluor 647 (AF647) and BV510, and one PE-Cyanine tandem (PC5, isotype 11) (Supporting Information Fig. S1). Non-specific binding is a well-known challenge in flow cytometry, especially when analyzing myeloid cells. Cyanine acceptors in tandem dyes including PE-Cy5, PE-Cy7, PerCP-Cy5.5, and APC-Cy7 have been shown to bind to monocytes and macrophages (5-7). Studies indicate that this non-specific binding is, at least in part, due to binding to the human high-affinity Fc receptor CD64 (5-7). Importantly, phosphorothioate-oligodeoxynucleotides could block this non-specific binding (5-7), but the mechanism has not been fully understood. Reviews focusing on pitfalls in flow cytometry rarely mention the issue (8, 9), even though BioLegend has demonstrated non-specific staining of PE-Cy7, PerCP-Cy5.5, APC-Cy7, PE-Cy5, PE-Dazzle594, and APC-Fire750 conjugated antibodies to monocytes, which could be eliminated by their True-Stain Monocyte Blocker (10). When reviewing all flow cytometry papers published in Cytometry Part A during the past 5 years (2015–2019), excluding non-relevant papers such as CyTOF studies and non-human cell-based work, we found that of the 74 identified papers, 59 (80%) used at least one Cyanine-tandem conjugate, and 52 (70%) used at least one PE-tandem. Further, of all 74 papers only 15 (20%) used a blocking reagent, which in all cases were Fc receptor blocking reagents, whereas none of the papers mentioned blocking of non-specific binding caused by tandem fluorochromes. Therefore, we decided to compare the blocking ability of 100 μg/ml human IgG (“Fc Block”) (4), True-Stain Monocyte Blocker (BioLegend), and phosphorothioate-oligodeoxynucleotides (here called “Oligo-Block,” Sigma-Aldrich, St. Louis, MO) as described by Jahrsdörfer and colleagues (7). We investigated whether the blocking reagents could eliminate the observed non-specific binding, presumably caused by Cyanine-tandem fluorochromes. For further information concerning Materials and Methods see supporting information S1. To investigate the effect of the mentioned blocking reagents, we performed a screening of eight IgG1 isotype controls on PBMCs from three healthy donors. This clearly showed inadequate blocking of PE-Cyanine-tandems when using Fc Block alone. In contrast, there was a marked decrease in non-specific binding for all PE-Cyanine-tandems (PE-Texas Red, PE-Cy5, and PE-Cy7) by both True-Stain and Oligo-Block (Supporting Information Fig. S2). Indeed, True-Stain and Oligo-Block removed the majority of non-specific binding, but staining intensity did not completely reach the level of unstained (Live-dead only). There was moderate inter-donor variability in the effectiveness of blocking by both reagents (see Fig. 2 and Supporting Information Table S2). Furthermore, we observed some PE-Cyanine-tandem binding to SSClow cells (lymphocytes). Backgating of this population did not show any obvious differences from the negative cells (Supporting Information Fig. S4B). Notably, this binding was not removed by neither Fc Block, True-Stain, nor Oligo-Block, which should be investigated further. Finally, on PBMCs from five different healthy donor, we investigated the ability of the three blocking reagents (Fc Block, True-Stain, and various Oligo-Block concentrations) to eliminate the non-specific binding of three different mouse (anti-human) PE-Cyanine-tandem conjugated antibodies: anti-CD206 PE-Cy7 (BioLegend, clone 15–2), IgG1 PE-Cy5 (DAKO, clone DAK-GO1), and anti-CD56 PE-Cy7 (BD Biosciences, clone B159). For these three antibodies, Fc Block caused only a slight reduction in monocyte MFI levels, compared to the unblocked sample. Notably, Oligo-Block at increasing concentrations gradually reduced the non-specific binding of all three antibodies to monocytes, reaching a plateau around ≥5 μg/ml. Importantly, results for anti-CD56 PE-Cy7 showed a decrease in MFI on the “false” CD56pos monocytes, whereas the specific staining of CD56pos NK cells remained (Fig. 2B). Thus, Oligo-Block (and True-Stain) prevented non-specific binding of PE-Cyanine-tandem conjugated antibodies to monocytes, but did not affect specific anti-CD56 binding to NK cells (Fig. 2D). To test whether the blocking reagents affected multicolor panels, we evaluated the effect of Oligo-Block in two different multicolor flow cytometry panels and found no major changes in specific staining levels (MFI) of non-Cyanine-tandem antibodies CD14 V500, CD16 AF647, HLA-DR FITC, TLR2 BV786, and CD45 AF700, but clear reduction in monocyte MFI for anti-TLR2 BB700 (Supporting Information Fig. S3). Thus, we were not surprised to discover that the new Brilliant Blue 700 (BB700) dye is actually also a Cyanine containing tandem dye (communication with BD Biosciences). In summary, we found that different Cyanine-tandems show non-specific binding to monocytes, which could not be removed by standard Fc blocking alone. However, both the commercial True-Stain Monocyte Blocker and phosphorothioate-oligodeoxynucleotides (here called “Oligo-Block”) removed most of this non-specific binding without affecting the specific binding of antibodies. Two papers investigating the mechanism of non-specific binding indicated that, at least in part, Cyanine-dye binding is due to binding to high affinity Fc receptor CD64 (5, 7). However, the mechanism is still not fully understood. On their website, BioLegend indicates that also non-Cyanine-based tandems may bind non-specifically to monocytes/macrophages. This should be investigated further. Considering the wide use of tandem conjugated antibodies in flow cytometry, especially Cyanine dye-based tandems, combined with incomplete use of blocking reagents, we wish to draw attention to the possibility of non-specific staining of human monocytes when using PE-Cyanine-tandem conjugated antibodies (and likely other Cyanine-based tandems). Importantly, this phenomenon may lead to falsely positive staining. Thus, we strongly recommend evaluating tandem conjugates (and other antibodies) for non-specific staining, and to prevent the occurrence by using a blocking reagent to eliminate this binding (e.g., the described “Oligo-Block”). Further, we also wish to emphasize the importance of “standard” Fc blocking, as described by our group and by others (4, 11, 12), since our literature review indicated that also “standard” Fc blocking was not performed in the majority of published studies. All experiments were performed using the LSRFortessa flow cytometer (BD Biosciences) or NovoCyte Quanteon (ACEA Biosciences) at the FACS Core Facility, Aarhus University, Denmark. Funding was received from the Danish Cancer Society, the Memorial Foundation of Eva and Henry Frænkel and the Memorial Foundation of Max and Inger Wørzner. Mie Kristensen: Data curation; formal analysis; writing-original draft; writing-review and editing. Sofie Kejlberg Jensen: Data curation; writing-review and editing. Anne Sofie Sørensen: Data curation; writing-review and editing. Thomas Vorup-Jensen: Data curation; writing-review and editing. Tue Wenzel Kragstrup: Data curation; writing-review and editing. Marianne Hokland: Data curation; formal analysis; writing-review and editing. Morten Andersen: Conceptualization; data curation; formal analysis; methodology; supervision; writing-original draft; writing-review and editing. All authors declare no conflict of interest. Figure S1. Isotype Screening Fluorescence Minus One (FMO, here Live-Dead only) and stained samples for the 19 different IgG1 isotype controls tested are shown. Live single cells were gated as shown in Figure 2. Isotype 12–15 showed highly significant non-specific binding as compared with respective FMOs. Other isotype controls showed much less or no non-specific binding. Note that PC5 and PC7 fluorochromes are special formulations of PE-Cy5 and PE-Cy7, respectively, created by Beckmann-Coulter. Thus, one of the investigated PE-Cyanine tandem isotype controls (isotype 11) did not display high degree of bi nding tomonocytes. Further information about used isotype controls is listed in Supporting Information Table S1. Figure S2. Isotype Screening Representative example of healthy donor PBMCs (n = 3). Shown PBMC samples include Live-Dead only, unblocked and three differently blocked samples for each of the eight isotype antibodies tested. Live single cells were gated as shown in Figure 2. The blocking reagents include Fc Block, TrueStain Monocyte Blocker and Oligo-Block. The included antibodies are number 9, 10, 12, 13, 14, 15, 18, and 19 from Supporting Information Figure S1. Further information about the isotype controls is listed in Supporting Information Table S1. Figure S3. Blocking reagents do not affect specific staining Influence of blocking reagents on binding of antibodies in two different multi-color panels (A + B). (A) We found no effects of the three blocking reagents on binding of CD16 AF647 (Biolegend, clone 3G8, cat. 302020, lot. B201807), TLR2 BV786 (BD Biosciences, clone 11G7, cat. 742771, lot. 81022544), HLA-DR FITC (BD Biosciences, clone G46-6, cat. 555811, lot. 8046823), and CD14 V500 (BD Biosciences, clone MΦP9, cat. 562693, lot. 8075909) (B) We found no effect of Oligo-Block or Fc Block on binding of CD14 V500 (BD Biosciences, clone MΦP9), while for CD16 PE-Cy7 (BD Biosciences, 3G8) and TLR2 BB700 (BD Biosciences, clone 11G7) there was some decrease in PE-Cy7 and BB700 MFI with Oligo- Block, similarly to the tandem antibodies tested in this study. Notably, BB700 is a new cyanine tandem dye created by BD Biosciences. Figure S4. Effect of Blocking Reagents on PE Tandem Binding to SSC low cells. Live single cells were gated as in Figure 2. (A) Gating of SSClow positive cells for CD206 PE-Cy7pos cells in samples with or without the tested blocking reagents (Fc Block, Oligo-Block, and True-Stain). The same gate was applied in all samples. (B) Backgating of SSClow positive cells shown in a representative sample. The positive cells showed similar characteristics as negative cells the investigated parameters. (C) Effect of blocking reagents on CD206 PE-Cy7 MFI for SSClow positive cells (normalized to show MFI in % of the unblocked sample). There was no clear decrease in non-specific antibody binding with increasing concentration of OligoBlock or TrueStain. (D) Similar analyzes as in (B) but with an IgG1 PE-Cy5 isotype control. All analyzes were performed on healthy donor PBMCs (n = 5). Representative data are shown in A. Table S1. Antibody information. Table S2. Overview of raw MFI data for individual donors (Fig. 2). Supplementary Materials. 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.

Topics & Concepts