An Unexpected Role for Cell Density Rather than IgM in Cell-Surface Display of the Fc Receptor for IgM on Human Lymphocytes

The Fc receptor for IgM (FcµR or FcMR) was identified in 2009 and remains the only known IgM-exclusive Fc receptor (1, 2). Cell-surface FcMR has been confirmed on human B, T, and NK cells and on mouse B cells (1–5) with debated expression in myeloid lineages of both humans and mice (3, 6–11). Although FcMR is the only Fc receptor constitutively expressed by human T cells (3), its role in these cells remains unclear. Indeed, many aspects of FcMR function remain to be elucidated. Because regulation of expression often hints at where and when a protein may be functionally relevant, we decided to characterize patterns of FcMR expression in human lymphocyte populations at the level of cell-surface display.

FcMR surface abundance is regulated by cellular activation status and composition of the surrounding tissue milieu (1, 3, 5). At baseline, lymphocyte cell-surface FcMR is reported to be low, but detectable, in blood and peripheral lymphoid organs (1, 3, 12). One hypothesized contributor to this low baseline surface level is the presence of the FcMR ligand, IgM (1, 3, 5), supported by findings that FcMR internalizes after binding to multimeric IgM (12–15). Indeed, preincubation of cells in IgM-deficient media has been recommended to raise surface FcMR to levels adequate for study, particularly for human T cells (1, 5, 12). Studies of primary lymphocytes allowed to recover FcMR in this manner have led to important insights whose recurring theme seems to be that FcMR plays markedly different roles in B versus T cells. For example, B cell activation increases surface FcMR, whereas T cell activation decreases it (1, 12). Furthermore, FcMR limits “tonic” BCR signaling by reducing transport of IgM-BCR to the surface of mouse B cells in vivo (13), whereas in cultures of primary human T cells, FcMR engagement by exogenous IgM resulted in increased transport of the TCR signaling complex to the cell surface (12).

Hypotheses regarding the function of FcMR are more advanced for B cells relative to T cells, presumably because its expression in T cells is species specific and thus cannot be evaluated in Fcmr^−/− mice. From mouse models, a consensus appears to be forming around a B cell “rheostat” hypothesis (16) in which FcMR-mediated effects on BCR signal strength enhance detection of self-antigens by immature B cells during development as well as of foreign Ags by mature B cells in secondary lymphoid organs. In contrast, hypotheses to explain why FcMR is expressed by human T cells are far less comprehensive and currently center around the tonic effect, noted above, that FcMR internalization was observed to have on the TCR complex and costimulatory molecules. In this model, naive T cells are envisioned as encountering abundant IgM upon entry into lymph nodes or spleen that will internalize as complexes with FcMR, resulting in enhanced surface expression of the TCR complex and costimulatory molecules, thus preparing the cells for cognate interactions (12). Less consideration has been given to potential functions of FcMR in T cell populations outside of lymphoid organs, perhaps because the increase in surface display of FcMR observed after culturing PBMCs in IgM-deficient media implies low abundance prior to harvesting the cells from whole blood, that is, while in circulation.

Changes in FcMR expression by circulating lymphocytes observed in some disease states such as chronic lymphocytic leukemia (CLL) may provide clues about its function. In CLL patients, FcMR is elevated compared with healthy counterparts not only on leukemic B cells, but also on non-leukemic B and T cells (3, 17). Abnormally elevated FcMR expression by leukemic cells is thought to be due to Ag-independent cross-linking of BCR in cis, which is ultimately mitogenic (1, 3). In the same patients, the mechanism by which FcMR expression is increased on non-leukemic B and T cells is unclear, but it may reflect the fact that CLL patients commonly experience a global deficiency in serum IgM such that comparatively less IgM is available to drive internalization of FcMR (3, 17). However, in patients with selective IgM deficiency, surface display of FcMR is unchanged on most lymphocyte subsets (and is actually decreased on circulating naive marginal zone B cells) relative to healthy controls (18). The lack of a consistent relationship between IgM abundance and FcMR surface expression in these two disease states suggests a need for further characterization of the regulatory mechanisms that determine when and where FcMR is available for functional interactions.

In our study, we confirmed that FcMR surface levels are upregulated on PBLs cultured in IgM-deficient media but also found that, surprisingly, FcMR upregulation was equally robust after culture in the presence of human serum, which contains an average of 1.5 mg/ml IgM, indicating that IgM exposure has little effect on steady-state surface FcMR expression (19). Cell-surface FcMR was instead strongly affected by culture at higher cell densities. Downregulation of FcMR independent of IgM abundance occurred through a mechanism requiring close cell–cell proximity that does not appear to require the presence of a particular cell type or soluble factor and thus remains undefined. We also found that ex vivo processing of whole blood decreases surface expression of FcMR, implying that circulating lymphocytes express it at significantly higher levels than previously believed despite continuous exposure to IgM. Collectively, our findings suggest that the physiological environments in which FcMR is available for functional interactions, especially for T cells, are different from previously thought, which has implications for the role FcMR may play in the human immune response.

Between 15 and 300 ml of venous blood was collected by standard phlebotomy techniques from consented donors who reported themselves to be healthy at the time of the blood draw. Blood collection was approved by the University of Louisville Institutional Review Board under expedited review. Donors were between 20 and 60 y of age, 25% female and 75% male, with some donors repeated across different experiments such that 28% of experiments were performed with female cells.

Blood was collected in a total of 6 mM K₃EDTA as anticoagulant. In some experiments, blood cells were stained under conditions of minimal manipulation in which flow cytometric Abs were added immediately before or after lysing RBCs prior to flow cytometry, as described below. In all other experiments, PBMCs were isolated using SepMate PBMC isolation tubes (STEMCELL Technologies, 85450) as directed by the manufacturer. Briefly, fresh anticoagulated blood was diluted 1:1 with PBS lacking calcium and magnesium (PBS^−/−, Thermo Fisher Scientific, 10010-023) and then layered into SepMate tubes preloaded with 15 ml of Lymphoprep density gradient medium (STEMCELL Technologies, 07801). Tubes were centrifuged at 1200 × g for 10 min at room temperature (RT). After centrifugation, PBMCs were poured off into fresh 50-ml tubes and washed twice with PBS^−/− by centrifugation first at 600 × g for 10 min and then at 300 × g for 10 min. Cell yield was determined using the count per microliter feature of a Cytek Northern Lights flow cytometer. Counts were performed by generating a 1:10 dilution of the original cell suspension in 200 μl of PBS^−/−, running the diluted suspension on the flow cytometer, and using the count per microliter of events in a cell gate generated on a forward scatter/side scatter plot in SpectroFlo for subsequent calculations. PBMCs were used immediately for flow cytometric staining or culture procedures or were further manipulated to purify or deplete blood cell subsets.

Cell depletions were performed via immunomagnetic selection using EasySep kits with the EasyEights magnet (STEMCELL Technologies, 18103) and the manufacturer’s recommended medium (PBS^−/− supplemented with 2% FBS and 1 mM K₃EDTA). For depletion of NK, B, or CD8 T cells, an EasySep biotin-positive selection kit II (STEMCELL Technologies, 17683) was used in combination with anti-CD56 biotinylated Ab, anti-CD19 plus anti-CD20 biotinylated Abs, or anti-CD8 biotinylated Ab, respectively (Supplemental Table I). For CD14⁺ monocyte depletion, an EasySep human CD14-positive selection kit II (STEMCELL Technologies, 17858) was used following a modified protocol provided by the manufacturer (20). Briefly, after traditional addition of reagents per the manufacturer’s instructions, dwell time in the EasyEights magnet was doubled and then unbound cells were transferred to fresh tubes for a second round of depletion before final collection. Mock depletions were performed in parallel by adding isotype prematched control Abs (NK, B, and CD8 T depletions) or without addition of isolation mixture (CD14⁺ monocyte depletion), as the mixture composition is proprietary.

For platelet depletion, PBMCs were first isolated using a modified SepMate protocol in which the top platelet-enriched fraction was pipetted off and discarded after density gradient separation before pouring the PBMC fraction into a new tube. Two wash steps were then performed with centrifugation at reduced speed to preferentially pellet nucleated cells (120 × g 10 min at RT). Platelets were further depleted using magnetic separation in the EasyEights magnet with an EasySep human platelet removal cocktail (STEMCELL Technologies, 19369C component of 19359). Mock platelet depletion was performed in parallel by processing PBMCs without removal of the top fraction after density gradient separation, without low-speed wash steps, as well as without addition of the platelet depletion mixture in subsequent steps.

Tonsil tissue was collected from pediatric patients at Norton Children’s Hospital undergoing tonsillectomy performed by S.K.C. Prior to tonsillectomy, the patient’s legal guardian signed an informed consent following Institutional Review Board ethical guidelines. All patients between ages 2 and 18 y presenting for tonsillectomy with or without adenoidectomy were eligible. Consecutive patients were invited into the study. Of the four donors used for this study, three were male and one was female. All patients were in the 4–11 y age range and undergoing tonsillectomy for management of sleep-disordered breathing with no other immune system-impacting comorbidities.

Immediately following tonsillectomy, one half of each of the right and left tonsil were placed into a cold solution of 35 ml of tonsil buffer made with 1 mM K₃EDTA, 100 U/ml penicillin with 100 μg/ml streptomycin (Thermo Fisher Scientific, 15140-122), 5 μg/ml gentamicin (Thermo Fisher Scientific, 15710-064), and 0.5 μg/ml amphotericin B (Sigma-Aldrich, A2942) in PBS^−/− for storage on ice up to 4 h prior to processing.

Tonsils were processed to obtain tonsil mononuclear cells (TMCs) using an optimized version of a previously published protocol (21). Tonsil tissue was minced in a sterile 100-ml petri dish while being kept wet with tonsil buffer solution. Once 1- to 3-mm fragments were obtained, the tissue was then transferred to a metal cell strainer sitting in a fresh 100-ml petri dish with additional tonsil buffer solution. Tissue was gently pushed through the strainer using the plunger of a 5-ml syringe. The tissue in the cell strainer was periodically washed with tonsil buffer to ensure it stayed wet and to encourage release of cells. The resulting cell suspension was then passed through 40-μm plastic cell strainers into fresh 50-ml tubes and pelleted by centrifugation at 600 × g for 10 min at RT. The pellet was resuspended in 20 ml of fresh tonsil buffer and split such that each of two 10-ml cell suspensions was layered on top of 25 ml of Lymphoprep density gradient medium (STEMCELL Technologies, 07801) in a 50-ml tube. This density gradient suspension was centrifuged at 800 × g for 20 min at RT with the brake off. Mononuclear cell layers at the resulting interfaces were pipetted into new 50-ml tubes and washed twice using tonsil buffer solution by RT centrifugation at 600 × g for 10 min. Cells were pooled by donor and mononuclear cell yield was determined using the count per microliter feature of a Cytek Northern Lights flow cytometer as described above. TMCs were used immediately for flow cytometric staining or for culture procedures.

For experiments requiring use of fresh serum, venous blood (5–15 ml) was collected in BD Vacutainer SST serum separation tubes (BD Biosciences, 368013) at the time of blood collection for PBMCs from either the same blood donor (autologous) or a separate donor (non-autologous). Serum was isolated according to the manufacturer’s protocol in which collected blood was mixed with clotting agents by inversion of the tubes and incubated at RT for 30 min before centrifugation at 1200 × g for 10 min at RT. Serum was collected from above the polymer gel plug. An additional centrifugation at 1200 × g for 10 min at RT was used as necessary to remove any remaining RBCs.

For all experiments, unless otherwise indicated, PBMCs or TMCs were cultured in suspension of 2.5 million cells/ml plated at 200 μl/well in 96-well Falcon U-bottom tissue culture–treated plates (Corning, 353072). In most experiments, PBMCs or derivatives were cultured in complete RPMI generated using RPMI 1640 media (Thermo Fisher Scientific, 21870-076) with the addition of 1× GlutaMAX (Thermo Fisher Scientific, 35050-061), 100 U/ml penicillin and 100 μg/ml streptomycin (Thermo Fisher Scientific, 15140-122), and 10% sterile-filtered, heat-inactivated male AB serum (Sigma-Aldrich, H3667). TMCs were cultured in complete RPMI supplemented with 5 μg/ml gentamicin (Thermo Fisher Scientific, 15710-064) and 0.5 μg/ml amphotericin B (Sigma-Aldrich, A2942).

For serum dilution experiments, complete RPMI was supplemented with 0, 10, 20, 40, or 70% human serum. For 100% human serum cultures, only antibiotics and GlutaMAX were added at the same concentrations as in complete RPMI. Human serum in these experiments was either fresh and autologous/non-autologous, collected as described above, or sterile-filtered, heat-inactivated male AB serum (Sigma-Aldrich, H3667). IgM concentrations in these sera ranged from 0.36 to 2.3 mg/ml, as determined by ELISA.

For density dilution experiments, cells were cultured at densities of 0.2, 0.5, 1, 1.5, 2, or 4 million cells/well in 200 μl of complete RPMI. In some experiments, BD Falcon flat-bottom tissue culture–treated plates (Corning, 353077) were used alongside U-bottom plates for these cultures.

For transwell culture experiments, HTS Transwells (Corning, 3388) were used with a total of 200 μl per well. The bottom “receiver” wells were first loaded with 100 μl of either complete RPMI alone or containing 0.2 or 2 million cells. After loading the bottom wells, the top “insert” wells were loaded with 100 μl of complete RPMI containing either 0.2 or 2 million cells to yield a total of 200 μl/well with top and bottom cell amounts as specified in Fig. 7B.

Transwell plate wells are flat-bottomed, so a muted density-dependent effect in bottom wells would be expected based on our results. However, in some experiments cell density effects for the cells in the transwell bottom wells were muted compared even to what we saw in BD Falcon flat-bottom plates. As controls, FcMR expression on cells plated in the bottom wells were assessed to determine whether the expected density-driven effect had occurred, and thus whether the results could be meaningfully compared with those observed when using other plate formats. These controls were assessed by calculating the ratio between FcMR mean fluorescence intensity (MFI) corrected for background (ΔMFI) measured for the bottom well 0.2 million and 2 million cell densities and comparing this ratio to the range of ratios observed in four experiments performed in traditional BD Falcon flat-bottom tissue culture–treated plates (Corning, 353077). In two experiments with transwell plates we found that ratios were well outside this range. Failure of these positive controls led to the exclusion of one experiment for all subsets, and exclusion in one other experiment for the B and NK cell subsets only.

To test for the presence of secreted factors, cell-free culture supernatants (SUPs) were collected after 24-h cultures. SUPs were generated by centrifuging culture plates at 860 × g for 3 min at RT, transferring media SUPs to 1.5-ml Eppendorf tubes, and storing at −80°C until use in subsequent experiments. Immediately prior to use, SUP tubes were centrifuged at 10,000 × g for 3 min at RT to remove cell debris. Cells tested for responses to potential soluble factors were either cultured in 200 μl of fresh complete RPMI with 10% human serum alone (no SUP) or with 100 μl of complete RPMI mixed with 100 μl of culture SUPs (50% SUP, collected from cultures of 0.2 or 2 million cells/well).

All cells were analyzed after 24-h culture or, for time course experiments, after 0.5, 1, 2, 4, 6, 12, and 24 h of culture in a standard 5% CO₂ humidified incubator (Thermo Fisher Scientific, Heracell 150i CO₂ incubator). Tests of the effects of oxygen tension were performed by culturing cells in a standard incubator (18–20% O₂) in parallel to cells cultured at approximate physioxia [5% O₂ (22, 23)] in a humidified, 5% CO₂, nitrogen-controlled incubator (Sanyo O₂/CO₂ incubator, MCO-5M).

Cells were stained for flow cytometric analysis in the same 96-well U- or flat-bottom plates in which they had been cultured. Zero-hour measurements were performed by mock plating cells in 96-well U-bottom plates. Cells in these 96-well plates were pelleted and then washed twice with PBS^−/− prior to viability staining. All pelleting and wash steps were performed by 4°C centrifugation at 860 × g for 3 min. To assess viability, cells were resuspended in 100 μl of a viability stain containing eBioscience fixable viability dye (Thermo Fisher Scientific, 65-0865-14) in PBS^−/− lacking added serum or protein. Cells plated at 0.2 or 0.5 million cells per well were resuspended directly in 100 μl of viability stain whereas cells plated at higher densities were first split to achieve a uniform density of 0.5 million cells per 100 μl of stain. After addition of viability stain, cells were incubated 20–30 min at 4°C, washed once with PBS^−/−, and washed again with stain buffer [PBS^−/− containing 0.09% sodium azide and 2% human serum to prevent binding of fluorescent Abs to FcR (24)]. After the second wash, cells were resuspended in 100 μl of stain master mix. Stain master mix was generated by first mixing staining Abs (Supplemental Table I) in BD Horizon Brilliant Stain Buffer (BD Biosciences, 566385 or 563794) according to the manufacturer’s instructions, then adding standard stain buffer to reach a cumulative 100 μl. Cells were incubated in this stain master mix for 20–30 min at 4°C, washed twice with standard stain buffer, and resuspended in cold stain buffer containing 1% formaldehyde for fixation prior to transfer to 12 × 75-mm polystyrene tubes for flow cytometric analysis.

For flow cytometry with rabbit polyclonal anti-FcMR Abs (RpAbs), cells were stained in a buffer containing TruStain FcX (BioLegend, 422302) diluted in PBS^−/− with 2% FBS and 0.09% sodium azide according to the manufacturer’s instructions. After primary incubation with RpAbs, two washes with stain buffer were performed prior to secondary incubation with an Ab master mix supplemented with BV421-labeled goat anti-rabbit staining Ab, as described above.

Flow cytometry was performed with a Cytek Northern Lights three-laser flow cytometer. Spectral profiles were unmixed accounting for autofluorescence using SpectroFlo software and appropriate single-stain and unstained controls. Processed data files were then analyzed in FlowJo (BD Biosciences), with additional fluorescence compensation performed using FlowJo compensation matrices when needed. Gating strategies used to identify cell populations are shown in Supplemental Fig. 1. FcMR stain MFI of gated cell populations was determined using FlowJo and corrected for background to calculate ΔMFI as FcMR MFI of each technical replicate minus the average MFI of all isotype control replicates in the same experiment. For the rabbit polyclonal anti-FcMR stains, ΔMFI was calculated as FcMR MFI of each experimental replicate minus the average MFI of all fluorescence-minus-one control replicates in the same experiment. When the calculated ΔMFI was negative, indicating that a replicate MFI was below that of the average isotype for the experiment, its ΔMFI was recorded as zero.

For experiments involving direct staining of leukocytes without using density gradients for isolation of PBMCs, BD Pharm Lyse (BD Biosciences, 555899) solution at 1× concentration was used to lyse RBCs. Three variations of the manufacturer’s protocol were performed, all starting with whole blood supplemented with 6 mM K₃EDTA as anticoagulant and performed at room temperature unless otherwise indicated:

1. Stain then lyse: 100 μl of freshly drawn whole blood was added to 12 × 75-mm polystyrene flow cytometry tubes. Staining was performed by adding 50 μl of a stain master mix containing staining Abs (Supplemental Table I) in BD Horizon Brilliant Stain Buffer (BD Biosciences, 563794). Cells were incubated for 15–30 min at 4°C in the dark. After incubation, 2 ml of 1× BD Pharm Lyse was added, and the tubes were gently vortexed and then incubated for 10 min in the dark at RT for RBC lysis. Tubes were centrifuged at 300 × g for 5 min at RT, washed once with 3 ml of stain buffer, and then resuspended in PBS^−/− containing 1% formaldehyde for fixation. Cells were kept on ice prior to flow cytometric analysis.

2. Lyse once and then stain: 1 ml of whole blood was added to 15-ml polypropylene tubes. RBC lysis was performed by adding 10 ml of 1× BD Pharm Lyse, gently vortexing the mixture, and incubating for 15 min. After incubation, tubes were centrifuged at 300 × g for 5 min. The cell pellet was resuspended in 2 ml of staining buffer, split into a 96 well U-bottom plate at 200 μl/well, and centrifuged at 860 × g for 3 min at 4°C. Cells were stained in 96-well plates as described above for PBMCs but without viability staining. After staining, cells were resuspended in a 1% formaldehyde solution for fixation prior to transfer into 12 × 75-mm polystyrene flow tubes. Cells were kept on ice prior to flow cytometric analysis.

3. Lyse twice and then stain: initial RBC lysis and cell centrifugation was performed as described above. The cell pellet after first lysis was resuspended in 1 ml of stain buffer, and RBC lysis was repeated by adding 10 ml of 1× BD Pharm Lyse, gently vortexing the mixture, and incubating for 15 min. After incubation, tubes were centrifuged at 300 × g for 5 min. This pellet was resuspended in 2 ml of staining buffer, split into a 96-well U-bottom plate at 200 μl/well, and stained and then fixed for flow cytometric analysis as described for the lyse once and then stain procedure.

Serum IgM concentrations were measured using the HRP/TMB-based human IgM ELISA Kit (Thermo Fisher Scientific, 88-50620) following the manufacturer’s protocol. All sera were tested in triplicate at dilutions of 1:4,000, 1:8,000, and 1:16,000. Measurements were taken using an EMax Precision Microplate Reader (Molecular Devices) at 450-nm wavelength. Triplicate values from 1:4,000 dilutions, which consistently had the lowest coefficient of variability compared with other dilutions, were averaged and used to calculate serum IgM content in milligrams per milliliter. Values plotted in Fig. 3C show final IgM concentrations in culture media containing 10% human serum.

Statistical significance was assessed using GraphPad Prism software version 9.2.0 with the test specified in each figure legend. Normality was tested with the same software by assessing linear fit of the data set to a normal Q-Q plot generated using GraphPad Prism. For tonsil data, values for technical replicates outside of 3 SD of the dataset mean were deemed outliers and excluded from figures and calculations. This led to the exclusion of one technical replicate in one experiment.

Previous studies reported that overnight culture in IgM-deficient media increased FcMR surface levels on human B, T, and NK cells (1, 12). We confirmed this pattern using flow cytometry to measure surface FcMR on peripheral blood B, T, and NK cells that were stained after processing whole blood into PBMCs and then again after culture. Using a BV421-conjugated anti-FcMR mAb (clone HM14), surface FcMR was found to be increased after 24-h culture in serum-free media for CD4 T cells, CD8 T cells, B cells, and NK cells (Fig. 1). No sex-specific differences in FcMR surface expression were observed, either at 0 h or after 24-h culture (data not shown). To validate the specificity of FcMR staining in controlling for epitope- or fluorochrome-specific effects, staining was also performed with a second anti-FcMR mAb (clone HM7) conjugated to allophycocyanin. In addition, FcMR expression was measured with RpAbs as a primary stain and BV421-conjugated anti-rabbit IgG Fc goat polyclonal Abs as a secondary stain. HM7 and RpAb staining of FcMR showed identical patterns of expression for CD4 T cells and CD8 T cells (Fig. 1A, 1B) and NK cells (Fig. 1D). B cells cultured in serum-free media displayed more variable amounts of surface FcMR such that only mAb clone HM14 registered a statistically significant increase when comparing no serum culture to 0-h samples (Fig. 1C). However, all anti-FcMR Ab stains showed roughly the same magnitude of increases during the 24-h culture period, including for these B cell populations. Our observations of similar staining patterns with three anti-FcMR formats and two different fluorochromes suggest that the differences we observe are not artifacts of Ab affinity, epitope specificity, or fluorochrome.

To assess whether the presence of human serum, and thus IgM, would suppress or reduce cell-surface display of FcMR, PBMCs were cultured with media containing 10% human serum in parallel to cultures with no serum. Unexpectedly, on all four cell types, FcMR upregulation from 0 h occurred in the presence of human serum to the same extent when measured with all three Ab stain formats (Fig. 1, red overlays and bars). Furthermore, for all cell types and all stains there was no difference in the observed FcMR level after culture in serum-free media compared with culture in the presence of 10% human serum (Fig. 1, red compared with blue overlays and bars). These results show that the use of IgM-deficient culture media is not needed to detect FcMR expression and suggest that IgM may not determine the steady-state amount of FcMR on the cell surface to the extent reported elsewhere.

Baseline surface FcMR on PBLs has been reported previously to be low (1, 3, 12). However, most measurements were taken using PBMCs prepared by processing whole blood to remove RBCs and granulocytes through density gradient separation and multiple wash steps. To test whether manipulation of blood lymphocytes affects surface FcMR levels, three separate direct-from-blood staining procedures were performed to determine which allowed reliable FcMR detection with as little cell handling as possible (Fig. 2). The average amounts of FcMR on the surfaces of each lymphocyte cell type were the highest on the least manipulated cells (Fig. 2, stain then lyse, green bars and overlays) and decreased with increasing manipulation. These decreases did not reach statistical significance but were consistent across all cell types. Hence, the comparatively intense manipulation needed to process whole blood into PBMCs with a density gradient is likely to underestimate the amount of FcMR displayed by blood lymphocytes in vivo. Moreover, the average FcMR ΔMFI obtained by the stain then lyse method likely gives a more accurate estimate of FcMR present on the surfaces of circulating lymphocytes in vivo than do values obtained after preparation of PBMC fractions by standard methods.

To further test for any linkage between IgM content and surface display of FcMR, PBMCs were cultured in media supplemented with human sera from multiple donors that contain naturally variable amounts of IgM (Fig. 3A). For these experiments, PBMCs from the same donor were tested so that IgM content was the primary variable. Five separate serum sources were used including two separate lots of filtered and heat-inactivated sera, one non-autologous serum, and two separate collections of autologous sera. The IgM content in these sera ranged from 0.36 to 2.3 mg/ml as measured by ELISA, and therefore the dose range in 10% culture was 0.036–0.23 mg/ml. For all subsets of lymphocytes, no correlation was found between the amount of IgM in culture media and surface FcMR after 24-h culture (see (Fig. 3A for r and p values), indicating that no dose-responsive relationship exists, at least within this dose range.

We next asked whether regulation of surface FcMR requires more IgM than is present in culture media containing 10% human serum. This was done by culturing PBMCs with increasing amounts of human serum and assessing FcMR display after culture. For all lymphocyte subsets tested, no consistent or significant differences in surface levels of FcMR were observed when cells were cultured for 24 h with 0, 10, 40, 70, or even 100% human serum (Fig. 3B). This indicated that even after culture in physiologic amounts of IgM, FcMR surface levels could be maintained to a similar extent as when IgM was absent. Although not seen consistently, some experiments did show a trend toward decreasing FcMR with increasing serum content, especially in cultures with 70 or 100% human serum. To determine whether a culture medium with low (30%), or no, RPMI present had limited buffering capacity we measured pH after 24 h of use in culture and found that both 70 and 100% sera media were acidified well outside of the physiologic pH range (data not shown). We speculate that this is because RPMI is buffered specifically for tissue culture in incubators with 5% CO₂ such that human serum becomes acidified without added buffers. For this reason, 40% serum was the maximum amount used in subsequent experiments.

To investigate the kinetics of FcMR surface display and determine whether increasing serum amounts influenced surface FcMR at time points earlier than at 24 h, as tested thus far, time course measurements were performed with PBMC cultures containing 0, 10, 20, or 40% human serum. For B and T cells it appeared that the changes in FcMR surface levels were biphasic, with an initial short-term increase that reached a limited plateau as early as 1 h after culture initiation before a secondary increase that began sometime after 6 h (Fig. 3C). This biphasic pattern was observed in B and T cell populations regardless of the amount of serum added. When serum was absent, FcMR surface display on CD4 and CD8 T cells lagged that of the serum-replete cultures, indicating that there is no early time point at which FcMR is of greater abundance with IgM absent. Interestingly, the initial plateau reached for both CD4 and CD8 T cells in 10–40% serum culture was similar to the ΔMFI measured for minimally manipulated blood cells (e.g., Fig 2, stain then lyse samples), which may provide a more accurate estimate of surface FcMR display by T cells while in circulation. These patterns collectively suggest that T cells cultured in the presence of 10–40% human serum quickly recover surface FcMR that was lost during cell processing, whereas in the absence of serum the recovery is comparatively slow rather than accelerated. Ultimately, these data further support a limited or absent role for IgM in regulating FcMR surface display of circulating lymphocytes, which appear to have more FcMR available for functional interactions than previously appreciated.

During pilot experiments to optimize culture conditions, we observed surprisingly low surface display of FcMR when cells were plated at higher densities. To rigorously test how high-density culture might affect FcMR surface display, PBMCs were cultured at cell densities ranging from 0.2 to 1.5 million cells per well and surface FcMR levels were measured after 24 h. As in the pilot experiments, when PBMCs were cultured at higher densities, FcMR levels on all lymphocyte cell types did not reach the same levels as when PBMCs were cultured at lower densities (Fig. 4A). For all cell types the relationship between FcMR display and cell density in 24-h culture fit with a regression line that had a significantly non-zero negative slope, strongly indicative of a cell density–dependent mechanism.

To determine whether higher cell density in culture influenced the kinetics of FcMR expression, PBMCs were cultured at cell densities ranging from 0.2 to 2 million cells per well and tested at time points ranging from 0.5 to 24 h. As in the previous time courses, B and T cells cultured at low densities of 0.2 or 0.5 million cells per well had a biphasic increase in FcMR display with an initial plateau and subsequent increase between 6 and 24 h of culture (Fig. 4B). The initial plateau in surface FcMR expression reached by T cells again approximated the amount estimated to be present on circulating cells in vivo, whereas T cells cultured at higher densities of 1 or 2 million cells per well never recovered to reach this level and, furthermore, remained low at times when the second upregulation of FcMR expression was evident in lower cell density cultures.

FcMR expression by B cells was not as clearly influenced by cell density with significant differences between the 0.2 and 2 million cell groups observed only at one early time point, 2 h, which became amplified at the later time points of 12 and 24 h. NK cells did not show a discernable pattern, with no significant differences found between the two high and low cell density groups at any time point. Collectively, the findings depicted in (Fig. 4 indicate a density-driven suppressive effect on surface FcMR for B and T cells occurring early in culture, most notably for T cells at early time points when it is plausible that increases in surface FcMR are due primarily to recycling of FcMR from internalized pools (12).

One possible explanation for cell density effects is that oxygen becomes limiting when greater numbers of cells are cultured together. Standard incubators keep cultures at close to room oxygen (around 20% O₂), which is hyperoxic compared with approximate physiologic oxygen tension of 5% O₂ (22, 23). To assess whether oxygen tension might regulate surface FcMR display, cells were cultured at different densities in either normoxia (room oxygen) or physioxia (5% O₂). However, the same patterns of lower surface FcMR after culture at higher cell densities were observed for all lymphocyte cell types regardless of oxygen tension (Supplemental Fig. 2). Statistically only one difference was observed between the surface FcMR of cells cultured in normoxia compared with physioxia in NK cell cultures containing 1 million cells per well. All other comparisons showed no differences as a function of oxygen tension, indicating that the density-dependent effect on surface FcMR expression is probably not a reflection of differences in oxygen availability.

PBL subsets have a different composition than those of lymphoid organs, with the latter possessing high cell densities as well as activated and specialized subsets of cells. To determine whether the lymphocyte subsets in one such lymphoid organ, the tonsil, exhibit similar patterns of surface FcMR display in culture to those of PBMCs, TMCs were cultured for 24 h with varying amounts of human serum (Fig. 5A). Interestingly, the maximum surface FcMR on tonsil lymphocytes was lower than that of corresponding PBMC subsets, whether cultured for 24 h with no serum or with 10% human serum. Analyses of NK cells were not included due to low counts in the TMC preparations. For all tonsil lymphocyte subsets observed after 24-h TMC culture without serum, there were no significant increases in surface FcMR compared with those observed at 0 h. In cultures with 10% human serum, a significant increase in surface FcMR from 0 to 24 h was only observed for the T follicular helper cell subset. However, for all subsets there were significant increases in surface FcMR from 0 h for 40% serum cultures. The relative difficulty in observing increases in surface FcMR in culture with lower serum levels could reflect the more extensive manipulation needed to prepare tonsil cells for culture, as evidenced by the higher serum amounts required to recover FcMR and sustain cell health in our experiments. Regardless, the lack of expected surface FcMR differences after culture with increasing serum amounts and the failure to increase FcMR in cultures without serum suggest that IgM does not affect FcMR display on tonsil lymphocyte subsets.

Tonsil-derived lymphocytes were also tested for the same density-dependent regulation of surface FcMR in culture as previously observed with PBMCs. Unlike their PBMC counterparts, a linear regression showed no significant relationship between the cell density and surface FcMR when TMCs were cultured for 24 h (see (Fig. 5B for r and p values). These data suggest that 24-h culture is insufficient for FcMR to recover from high cell density in vivo or from the extensive manipulation needed to isolate TMCs from intact tonsils. Alternatively, the regulation of surface FcMR in TMC culture may differ from that of PBMCs.

We next asked whether a particular cell type present in PBMCs might be responsible for cell density–dependent suppression of FcMR surface display. For these experiments we chose to focus on FcMR display by CD4 T cells, as these cells consistently exhibited more pronounced responsiveness than did CD8 or B cells. To assess whether a specific cell type in PBMC cultures was necessary for the density-driven regulation of surface FcMR on CD4 T cells, five different cell types were independently depleted from PBMCs prior to culture at varying cell densities (Fig. 6). All targeted cell types were depleted by at least 10-fold compared with mock depleted controls (platelets, CD14⁺ monocytes, B cells, CD8 T cells, or NK cells; see Supplemental Fig. 3), but in no depleted culture did surface FcMR fail to be downregulated on CD4 T cells under conditions of high culture density. Although these findings were not definitive, they suggested that density-dependent regulation of CD4 T cell surface FcMR is not likely to reflect the activity of any specific cell type but may instead reflect bulk cell density.

As no specific cell type was readily identified as responsible for the effect of cell density on FcMR display, we turned to evaluations of cell proximity and secreted factors as explanations. We first tested whether the spatial relationships of cells cultured together played a discernible role by comparing FcMR expression after culture over a range of cell densities in U- versus flat-bottom wells (Fig. 7A). The magnitude of differences attributable to well shape seemed greater for T and B cells than for NK cells, in which FcMR surface expression is generally low under any conditions. FcMR abundance on cells cultured in flat-bottom microplates trended higher than after culture in U-bottom microplates, in which cells settle into multilayer cell clusters at comparatively lower cell densities. Statistically, differences were significant at intermediate PBMC densities; for example, FcMR was increased more on CD4 T cells in flat-bottom wells containing 1 or 2 million cells per well, but not 0.2, 0.5, or 4 million cells. This pattern was suggestive of cell-to-cell contact playing a larger role than secreted factors but was not definitive.

To test directly for a soluble factor produced in high-density cultures we first used a transwell culture system that allows exchange of soluble factors between separated populations of cells. PBMCs were plated at high density (2 million cells per well) and low density (0.2 million cells per well) in opposite transwell chambers, and surface FcMR was measured after 24 h of culture. No significant differences were observed in FcMR surface display on cells in low-density cultures that had been continuously exposed to factors secreted by high-density cultures (Fig. 7B), suggesting that no soluble factors played a role. However, failure of a positive control for PBMC density-dependent effects on CD8 T and NK cells indicated that this conclusion was more justified when considering soluble factors that could affect FcMR display on CD4 T and B cells. Specifically, compared with previously significant differences in FcMR display for low- versus high-density cultures when using traditional U-bottom wells, flat-bottom transwell control cultures showed no significant differences in FcMR display between high- and low-density culture for NK or CD8 T cells (although the difference was trending for CD8, p = 0.057).

To further test for the presence of a soluble factor in high-density culture affecting FcMR display, PBMCs were cultured at 0.2 million cells per well in media with 1:1 addition of cell-free SUP collected previously from high-density (2 million cells per well) or low-density (0.2 million cells per well) cultures (Fig. 7C). In these experiments, control cultures plated at high density, without SUP added, exhibited the expected downregulation of surface FcMR. Surface expression of FcMR by all cell types tested was not affected by the addition of SUP from a high-density culture compared with those cultured with SUP from a low-density culture. These results support the transwell data in finding no evidence that a soluble factor produced during high-density culture regulates surface FcMR display. In combination with data demonstrating that the density-driven inhibition of FcMR display is muted in flat-bottom compared with U-bottom wells, these findings suggest that the density-driven effect on surface FcMR in culture is dependent on cell proximity rather than a soluble factor.

In this study, we report that FcMR expressed on the surfaces of human lymphocytes is not decreased in the presence of IgM, contrary to the prevailing hypothesis of a role for ligand-dependent downregulation of FcMR display (1, 5, 12, 15). We further observed higher surface FcMR on blood lymphocytes when stained immediately ex vivo, without prior RBC lysis or processing to isolate PBMC fractions, suggesting that FcMR is expressed on the surfaces of circulating lymphocytes at higher levels than previously appreciated (1, 3, 12). We also report a novel cell density effect that strongly restricts FcMR surface display in culture, which we believe to be mediated by a yet unknown cell proximity–dependent mechanism. Our findings alter the current understanding of factors influencing FcMR display and suggest that more FcMR is available at the surface of circulating lymphocytes for functional interactions than had been considered.

The idea that IgM abundance is not correlated with cell-surface FcMR is supported by our finding that FcMR levels were as high after 24-h culture with IgM-containing human serum present as in serum- and IgM-free media. We additionally found no evidence of a relationship between FcMR display and IgM content when accounting for natural variation among different sources of human sera. In time course studies, we saw no time point at which surface FcMR levels were reduced as a function of increasing serum supplementation. In fact, for T cells, the only cultures in which FcMR display lagged were those lacking serum entirely where, based on previous reports (1, 5, 12), the absence of IgM should have increased surface FcMR the most.

As we found no evidence that serum IgM influenced FcMR display, we investigated cell manipulation–related decreases as an alternative explanation for the low amounts of surface FcMR observed on lymphocytes in PBMC fractions of whole blood. To this end, we tested for and found higher FcMR levels on cells after direct-from-blood Ab staining, suggesting that circulating lymphocytes have higher surface FcMR display than previously reported by investigators who may have inadvertently stressed cells when using density gradient centrifugation and multiple washes immediately prior to flow cytometric analysis (1, 3, 12). This conclusion seems strongest in the case of peripheral T cells, whose surface FcMR rebounded quickly upon culture within 1 h and then stabilized at levels approximating those measured by direct-from-blood Ab staining.

Although we did not test for it in the present study, FcMR receptor recycling on T cells has been previously reported to take place in culture (12). We speculate that cell processing to isolate PBMCs triggers the internalization of FcMR, causing the apparent amount of FcMR on circulating B, T, and NK cells to seem artificially low, and we further speculate that display returns to true circulating levels due to receptor recycling in unperturbed PBMC cultures. Sometime after 6 h of culture, surface expression of FcMR began to increase again, which is suggestive of de novo synthesis. Hence, we propose that tests of FcMR function in human T cells may best be performed with PBMCs that have been rested in complete culture medium for 1–6 h, diverging from prior recommendations of overnight preculture in serum-free media (1, 5, 12), which were based on studies performed before mAbs for the receptor became available. Fluorescently labeled IgM was therefore used to stain what would come to be known as FcMR (1, 25), and because serum IgM can block binding by labeled IgM, serum-free medium was thought to be necessary. Our results using mAbs to stain the receptor reveal that the presence of serum IgM does not alter true receptor display.

In contrast to circulating cells, lymphocytes in secondary lymphoid organs may truly have very low surface FcMR in vivo. T cells have been shown to have low surface FcMR both in our tests and those of others of tonsil T cells from TMCs, where more extensive cell processing may admittedly be a confounding factor, but also in tonsil thin section microscopy (1, 12). This idea is further supported by our evaluation of high-density PBMC cultures, which are believed to approximate the lymphoid environment in part because they yield better T cell activation outcomes in studies of the optimal conditions needed to prepare T cells for adoptive transfer immunotherapy (26–29). We found that high-density culture markedly downregulated surface expression of FcMR by T cells, especially in U-bottom microplates in which cells cluster at the bottoms of the wells. At no time point tested did T cells in high-density cultures (1–2 million cells/well) reach FcMR amounts we estimate to be present on the surfaces of circulating lymphocytes, a stark comparison with their low-density (0.2–0.5 million cells/well) counterparts, which quickly reached this level. Assuming that high-density PBMC cultures favor at least some cell-to-cell interactions more typical of lymphoid organs, an intriguing implication is that FcMR may be more functionally relevant in populations of circulating T cells.

Interestingly, TMC T cells did not display cell density–dependent regulation of FcMR and consistently showed limited increases from a comparatively low starting point for all varieties of cultures tested. This pattern may indicate that mechanisms acting on T cells in high-density environments in vivo can continue to suppress surface FcMR display for as long as 24 h in culture. Thus, a crowded cellular environment may not only be a novel contributing factor to low surface FcMR on T cells within lymphoid organs, but exposure to such environments could have lingering effects.

Understanding the mechanism by which high cell density downregulates surface display of FcMR remains an elusive goal. We found no evidence that oxygen tension or secreted factors, including IgM, played a role, and further tested multiple cell types in systematic depletion experiments and found none that could explain the effect. We cannot rule out cell types present in PBMC fractions that were not tested. However, based on the existing data, we speculate that high cell density downregulates FcMR by a mechanism that is not cell-type specific but is driven by close cell proximity, likely requiring either cell-to-cell contact or production of a short-range factor, or both. Future studies are needed to decipher the mechanisms involved, as well as to evaluate the significance of the effect for human immune responses, especially those involving T cells.

In summary, our findings reveal a novel cell density–dependent effect on FcMR surface display in culture experiments and support the idea that FcMR expression, and therefore functional relevance, is likely to be greater on circulating lymphocytes than has been previously appreciated. A new understanding of FcMR regulation may in turn contribute to generation of new hypotheses that advance efforts to decipher when, where, and why this unusual receptor is expressed by human T cells.

We gratefully acknowledge Zachary van Winkle for phlebotomy as well as Caleb S. Whitley and Dr. Carolyn R. Casella for constructive critiques of the submitted research.

The authors have no financial conflicts of interest.