FVB/NCr mice were purchased from Charles River Laboratories (Wilmington, MA, USA). C57BL/6 mice were purchased from the National Cancer Institute (Frederick, MD, USA) or Taconic (Hudson, NY, USA). Transgenic mice expressing hGPA (24) on the RBCs on FVB genetic background (H2-q) generously provided by the New York Blood Center and HOD mice that have RBC-specific expression of hen egg lysozyme, ovalbumin, and human Duffy b (28) or the KEL glycoprotein (29) at a high density (“KEL^hi”) were bred in a non-pathogen-free facility at Yale University. OTII mice on a CD45.1 C57BL/6 (H2b) background whose CD4⁺ T-cells have transgenic T-cell receptors specific for ovalbumin were also bred at Yale. Transfusion-recipient mice were 8–12 weeks of age and on FVB background unless specified otherwise. All procedures and protocols were approved by Emory and Yale University’s Institutional Animal Care and Use Committee.

For CD4 depletion experiments, a previously optimized dosing schedule was utilized (30): mice were given two IP injections of 200 µg anti-mouse CD4 monoclonal antibody (clone: GK1.5, BioXcell, West Lebanon, NH, USA), saline, or an isotype-matched control 2 days apart. 48 h after the second injection, mice were transfused. On day 7 post-transfusion, another injection of 300 µg GK1.5 antibody was given.

For CD40L blocking experiments, another previously optimized dosing schedule was utilized (31): all mice were given IP injections of 250 µg anti-mouse CD40L(CD154) monoclonal antibody (clone: MR1, BioXcell, West Lebanon, NH, USA), saline, or an isotype-matched control every 2 days starting on the day of transfusion until 6 days after the transfusion. Additional MR1 was also given on days 9 and 12.

Mice were transfused (IV injection) with 75 µL of fresh packed transgenic hGPA or HOD RBCs that had been collected in the anticoagulant preservative solution CPDA-1 (citrate phosphorus dextrose adenine, Jorgensen Labs, Henry Schein, Melville, NY, USA) and filter leukoreduced over a Pall (East Hills, NY, USA) syringe filter. Some recipients were IP injected with 100 µg poly IC, polyinosinic-polycytidylic acid (Invivogen, San Diego, CA, USA), approximately 3–4 h before transfusion. FVB recipient mice were used for hGPA experiments as the hGPA donors are on an FVB background, and C57BL/6 mice were used for HOD experiments as the HOD donors are on a C57BL/6 background.

After collection and leukoreduction, donor hGPA or wild-type FVB/NCr RBCs were labeled with chloromethylbenzamido 1,1′-dioctadecyl-3,3,3′,3′-tetramethylindocarbocyanine perchlorate or 3,3′-dihexadecyloxacarbocyanine perchlorate according to the manufacturer’s instructions (Molecular Probes, Eugene, OR, USA) and as previously described (32). Recipient mice were transfused via IV tail vein with 75 µL of hGPA RBCs and a similar amount of wild-type RBCs. Survival of the transfused RBCs was determined by comparing the ratio of circulating hGPA RBCs to control RBCs in recipients longitudinally post-transfusion.

Single splenic cell suspensions from 8 to 10 weeks old female donor CD45.1 OT-II mice were prepared using gentle mechanical disruption, followed by RBC lysis with AcK buffer (0.15 M NH4Cl, 10 mM KHCO3, and 0.1 mM EDTA). CD4⁺ T-cells were isolated using a mouse CD4⁺ T-cell-negative isolation selection kit (Stemcell Technologies, Vancouver, BC, Canada). Purified OTII CD4⁺ T-cells were retro-orbitally injected into recipient mice. Recipient C57BL/6 mice were transfused with HOD RBCs 24 h following adoptive transfer.

For analysis of germinal centers (GCs), spleens were cryopreserved in optimal cutting temperature medium, sectioned on a cryostat, and stained. Frozen spleens were sectioned at 8-µm thickness fixed in 4% paraformaldehyde (Electron Microscopy Sciences, Hatfield, PA, USA) washed in ice-cold PBS, and stained at room temperature for 60 min with anti-B220 (AF488), anti-CD4 (PE), and biotinylated PNA in 5% fetal calf serum/PBS solution. After primary staining, the slides were washed in ice-cold PBS three times and stained with streptavidin AF647. In some circumstances, AF647-conjugated GL7 was used for GC staining instead of PNA and observed similar staining patterns. For consistency, near middle sections of the spleens were used for staining and were then analyzed using an automated wide-field microscope (Nikon Eclipse Ti) and CCD camera (Qimaging Retiga 2000R) with NIS elements software. Number of GC was counted in the entire field of view of the spleen sections.

All statistical analyses were performed using Graph Pad Prism software (San Diego, CA, USA). The Mann–Whitney U test was used to determine significant differences between two groups, and the Kruskal–Wallis with Dunn’s post-test or a two-way ANOVA with Tukey’s multiple comparisons test were used when comparing more than two groups as indicated. Error bars represent one SD, and significance was determined by a p-value of <0.05.

Mice treated with poly IC approximately 4 h prior to hGPA RBC transfusion generated anti-hGPA IgG, whereas those transfused in the absence of poly IC did not form detectable anti-hGPA IgG at any evaluated time point (Figure 1A) (26).

We next sought to characterize GC formation and other immune cell responses in alloimmunized and non-alloimmunized animals. Splenic GC cells were not readily characterized in alloimmunized animals in this hGPA model by flow cytometry (Figure S1A in Supplementary Material). In parallel to the flow cytometric studies, we also imaged the spleen post-transfusion for well-defined GC areas. Consistent with the flow data, we found no statistically significant differences in the number of GCs between groups within the fields evaluated (Figure S1B in Supplementary Material). Additional studies were completed after repeat transfusion with no significant GC formation identified in alloimmunized animals (Figure S1C in Supplementary Material); antigen-specific B-cell evaluation completed using RBC membrane ghosts from hGPA donors also failed to show GC formation in this model (data not shown). Although it cannot be ruled out that anti-hGPA IgG is formed through a GC-independent pathway (34), our approach to detect GC B-cells (in the absence of being able to identify antigen-specific cells) in the hGPA system lacks sensitivity.

Along with GCs, other cells in the spleen and BM were characterized longitudinally after transfusion of hGPA RBCs in the presence or absence of poly IC treatment. The frequencies of total BM plasma cells (PCs) were not significantly different between mice transfused in the presence or absence of poly IC at any studied time point (Figure S2A in Supplementary Material). Foxp3⁺ Tregs as well as TFH cells (gated as CXCR5^hi PD1⁺ out of total TCRβ⁺ CD4⁺ T-cells) were evaluated at multiple time points (days 2–14) post-transfusion in the presence or absence of poly IC pretreatment; no differences in total cell numbers were observed between the groups at any studied time point (Figures S2B,C in Supplementary Material). Finally, CD5⁺ B-cells were evaluated, with no differences observed in gated cell numbers between any group studied (Figure S2D in Supplementary Material).

Antibodies of the IgG class that are specific for glycophorin may be clinically significant in humans, and we hypothesized that the anti-hGPA IgG generated by animals treated with poly IC prior to RBC transfusion would also be capable of leading to premature RBC clearance in our murine model. To investigate RBC clearance patterns, we labeled hGPA RBCs with one lipophilic dye and labeled wild-type FVB RBCs with a different lipophilic dye; a mixture of RBCs was then transfused in alloimmunized or non-alloimmunized animals and the ratio of circulating labeled FVB to hGPA RBCs was evaluated longitudinally post-transfusion (Figure 1B). In alloimmunized mice, there was rapid preferential clearance of 70–90% of transfused hGPA RBCs within the first 24 h, with almost all hGPA RBCs cleared by day 4 post-transfusion. In contrast, there was no preferential clearance of transfused hGPA RBCs in non-alloimmunized or naive mice by day 4 post-transfusion (Figure 1C) or beyond. These data highlight the clinical significance of the humoral immune response to hGPA RBCs in the presence of poly IC and also suggest the lack of “cellular immune”-mediated RBC clearance in animals that failed to generate an anti-hGPA IgG response after prior hGPA transfusion.

To determine whether the formation of RBC alloantibodies in the hGPA system was dependent on CD4 cells, we treated mice with CD4-depleting monoclonal antibody GK1.5 or an isotype-matched control antibody prior to and during primary transfusion with hGPA RBCs and poly IC (Figure S3 in Supplementary Material shows CD4 status at the time of transfusion). Anti-hGPA IgG was detectable only in isotype-matched control treated and not CD4-depleted animals on day 14 post-transfusion (Figure 2A).

Next, we sought to investigate the response to a secondary hGPA RBC challenge in mice previously transfused in a CD4-depleted status. To do this, mice previously treated with GK1.5 or an isotype control antibody during the primary transfusion were rested for 3 weeks; the CD4 status at the time of primary and re-transfusion is shown in Figure S3 in Supplementary Material. The animals were then treated with poly IC and re-challenged with hGPA RBCs. Despite having CD4 T-cells present during this second hGPA RBC transfusion, mice that saw the hGPA antigen initially in a state of CD4 depletion remained non-responders to the secondary RBC exposure (Figure 2B). In contrast, the isotype control antibody-treated mice had a significant boostable anti-hGPA IgG response to the secondary transfusion.

Finally, we investigated the fate of transfused hGPA RBCs in these animals. There was essentially no preferential hGPA RBC clearance over co-transfused antigen-negative FVB RBCs in animals treated with GK1.5 prior to their initial but not secondary RBC transfusion. In contrast, approximately 90% of transfused hGPA RBCs cleared in the control group within the first 24 h post-transfusion, with almost all hGPA RBCs being cleared within 5 days post-transfusion (Figure 2C).

Taken together, these data suggest that poly IC-driven hGPA alloimmunization is CD4 dependent, that conditions around the initial hGPA RBC exposure dictate secondary responses (e.g., CD4 depletion at primary RBC exposure induces tolerance to secondary exposures), and that only animals with detectable anti-hGPA IgG have premature clearance of incompatible transfused RBCs.

Since CD40-CD40L signaling is critical in both antigen-presenting cell (APC):CD4 interactions (35) and T-cell:B-cell interactions (36, 37), we sought to determine if blocking this signaling could prevent hGPA RBC alloantibody formation. Mice were treated with the anti-CD40L monoclonal antibody, MR1, or an isotype-matched control, prior to treatment with poly IC and transfusion with hGPA RBCs. MR1-treated mice failed to generate detectable anti-hGPA IgG (Figure 3A shows data from 14 days post-transfusion). As we observed with CD4-depleted animals, mice initially transfused in the presence of CD40L blockade failed to generate detectable anti-hGPA IgG when treated with poly IC and re-transfused with hGPA 5 weeks later (Figure 3B). In contrast, the isotype control antibody-treated mice demonstrated a boostable anti-hGPA IgG response. These data suggest that the transfusion recipient’s CD40/CD40L interaction at first hGPA RBC exposure dictates subsequent hGPA responses.

Next, splenic GC responses were evaluated in animals 8 days following their secondary RBC exposure. Animals transfused in the presence of MR1 had essentially no detectable GL7⁺ CD95⁺ B-cells (Figure 3C). In contrast, those transfused in the presence of an isotype type-matched controls had GCs (albeit no more than those found in naïve animals or in animals treated with poly IC alone).

To investigate whether CD40L blockade during initial hGPA RBC exposure led to diffuse RBC non-responsiveness or to antigen-specific RBC non-responsiveness, we exposed alloimmunized and non-alloimmunized mice to a third-party antigen (KEL glycoprotein) on transfused RBCs. The immune response induced by the KEL-expressing RBCs used for these experiments (KEL^hi RBCs) is T-cell dependent (data not shown, manuscript in preparation). Serum was collected longitudinally post-transfusion, with flow cytometric crossmatch assays completed using transgenic KEL or wild-type RBCs as targets. MR1-treated mice, isotype control mice, and naïve WT mice had similar peak levels of anti-KEL antibodies (Figure 3D), suggesting antigen-specific non-responsiveness.

To investigate the potential mechanisms of action of CD40L blockade using MR1 Ab in a transfusion setting, we turned to a transgenic murine model that enables the study of antigen-specific CD4⁺ T-cells. HOD transgenic mice have RBC-specific expression of the hen egg lysozyme, ovalbumin, and the human Duffy antigen, and OTII CD4⁺ T-cells have been shown to proliferate in recipients transfused with HOD RBCs (12, 38). For these experiments, naïve C57BL/6 mice were adoptively transferred with 10,000 CD45.1⁺ OTII CD4⁺ T-cells and then transfused with HOD RBCs and poly IC with or without MR1 treatment. MR1-treated mice demonstrated significant blunting of peak anti-HOD IgG responses compared to those transfused in the absence of MR1 treatment (Figure S4 in Supplementary Material).

To determine the effect of MR1 treatment on the expansion of antigen-specific T-cells, we evaluated the frequency of antigen-specific CD45.1⁺ OTII cells in transfusion recipients at multiple time points post-transfusion (Figure S4 in Supplementary Material and Figure 4A show representative gating). The frequency of OTII cells was significantly lower in the MR1-treated group compared to the isotype-matched control treated group on days 7 and 14 post-HOD transfusion (Figure 4A). Given prior studies of MR1 and Tregs in graft-versus-host disease (GVHD) models (39), we evaluated these cells in our transfused animals. We found no significant differences in the number of endogenous or antigen-specific (CD45.1⁺ OTII) Foxp3⁺ CD4⁺ Tregs between groups (Figure 4B), with there being extremely few antigen-specific OTII cells to characterize in animals treated with MR1. To further investigate whether Tregs may play a critical role in the mechanism of action of MR1 alloantibody mitigation, we depleted Tregs using PC61 antibody (BioXcell), treated animals with MR1, transfused with transgenic RBCs, and evaluated alloantibody response 2 weeks later. MR1 treatment prevented RBC alloimmunization in both Treg-depleted and Treg-replete mice in this pilot experiment (data not shown), suggesting that Treg expansion is not necessary for the mechanism of action of MR1 in this system. We also evaluated endogenous and antigen-specific TFH cells in the transfused animals and found significantly decreased TFH cell numbers in animals treated with MR1 compared to those treated with the isotype-matched control (Figure 4C).

Finally, we evaluated splenic GC responses in these transfusion recipients. HOD RBCs were able to induce robust GC formation following a single transfusion. Treatment with MR1, however, fully prevented GC formation in transfused animals (Figures 5A,B).