Storage differentially impacts alloimmunization to distinct red cell antigens following transfusion in mice

Abstract

Introduction:

The impact of blood storage on red blood cell (RBC) alloimmunization remains controversial, with some studies suggesting enhancement of RBC-induced alloantibody production and others failing to observe any impact of storage on alloantibody formation. Since evaluation of storage on RBC alloimmunization in patients has examined antibody formation against a broad range of alloantigens, it remains possible that different clinical outcomes reflect a variable impact of storage on alloimmunization to specific antigens.

Methods:

RBCs expressing two distinct model antigens, HEL-OVA-Duffy (HOD) and KEL, separately or together (HOD × KEL), were stored for 0, 8, or 14 days, followed by detection of antigen levels prior to transfusion. Transfused donor RBC survival was assessed within 24 h of transfusion, while IgM and IgG antibody production were assessed 5 and 14 days after transfusion.

Results:

Stored HOD or KEL RBCs retained similar HEL or KEL antigen levels, respectively, as fresh RBCs, but did exhibit enhanced RBC clearance with increased storage age. Storage enhanced IgG antibody formation against HOD, while the oppositive outcome occurred following transfusion of stored KEL RBCs. The distinct impact of storage on HOD or KEL alloimmunization did not appear to reflect intrinsic differences between HOD or KEL RBCs, as transfusion of stored HOD × KEL RBCs resulted in increased IgG anti-HOD antibody development and reduced IgG anti-KEL antibody formation.

Conclusions:

These data demonstrate a dichotomous impact of storage on immunization to distinct RBC antigens, offering a possible explanation for inconsistent clinical experience and the need for additional studies on the relationship between RBC storage and alloimmunization.

1 |. INTRODUCTION

Strategies to prevent red blood cell (RBC) alloimmunization, most notably among transfusion-dependent patients like those with sickle cell disease or hematologic malignancies, are needed to improve patient care. To reduce RBC alloimmunization in transfusion-dependent patients, a greater understanding of the factors that directly influence the generation of RBC alloantibodies is needed.^1–5 Increased RBC storage time has been suggested to impact transfusion outcomes in general, as well as rates of alloimmunization more specifically.^6,7 Indeed, some clinical reports demonstrate increased alloimmunization associated with transfusion of RBC units stored for longer periods of time,⁸ which have been supported by experimental murine studies demonstrating increased alloimmunization associated with RBC storage.^9–11 Nevertheless, other clinical studies have not demonstrated such an association,^12–14 leading to uncertainty about the role or need for modification in current storage protocols and clinical practice to prevent or reduce RBC alloimmunization.¹⁵ As clinical studies have examined collective alloimmunization rates toward a variety of distinct alloantigens,^9–14 apparent discrepancies regarding the impact of storage on RBC alloimmunization may reflect the distinct consequences of storage on alloimmunization against unique alloantigens. Here we present data testing this hypothesis using a murine model of RBC transfusion. Our findings demonstrate that RBC storage can enhance the generation of IgG antibodies to one RBC antigen, yet decrease IgG antibody formation to a different RBC antigen.

2 |. STUDY DESIGN AND METHODS

2.1 |. Donor red blood cell collection, storage and transfusion

RBC units were generated from male and female C57BL6 mice expressing two distinct model antigens, HEL-OVA-Duffy (HOD) and KEL, separately or an F1 cross that expresses both antigens (HOD × KEL) previously shown to induce anti-HOD and anti-KEL antibodies.¹⁶ Blood was collected into CPDA followed by storage for 0, 8 (D8), or 14 (D14) days as outlined previously.¹⁷ Antigen levels were examined prior to transfusion of fresh (D0) or stored (D8 or D14) RBCs by staining each respective antigen at D0 (fresh) immediately following collection and comparing antigen levels observed at days 8 and 14 following storage as percentage of the levels observed at D0.¹⁸ An equivalent of one unit of RBCs (50 μl) in a total volume of 300 μL phosphate-buffered saline (PBS) were transfused by lateral tail vein intravenous injection and donor RBC survival and antibody production were assessed in female WT C57BL6 recipients as done previously.^{11,16,17,19,20}

2.2 |. Assessment of antibody development

Serum was collected from recipients on D5 and D14 post-transfusion and anti-HOD or anti-KEL IgM and IgG antibodies were detected via indirect immunofluorescence. Specifically, flow cytometric crossmatch was performed by incubating sera with RBCs expressing the HOD antigen, the KEL antigen, or antigen-negative syngeneic RBCs as a background control, washed and then incubated with anti-mouse IgM-APC (D5) or IgG-APC (D14) as outlined previously.^21–23 All flow cytometry samples were assessed using a FACSCalibur flow cytometer and FlowJo software, with statistical significance determined by one-way ANOVA with Tukey’s multiple comparisons tests and p < 0.05.

3 |. RESULTS AND DISCUSSION

As antigen density is known to impact red cell alloimmunization,²³ we first assessed levels of HOD and KEL antigens on fresh or stored RBCs. Pre-transfusion levels of the HOD antigen were not significantly different on HOD RBCs at baseline (D0) or after storage (Figure 1A,B,C). Similarly, red cell storage did not significantly impact levels of KEL antigen on KEL RBCs (Figure 1B,C). We next examined the possible impact of storage on post-transfusion donor RBC survival, as outlined previously.^9–11 Storage resulted in decreased post-transfusion RBC survival of both HOD and KEL RBCs, with longer storage resulting in more pronounced reductions (Figure 1D).

FIGURE 1.

Pre-transfusion storage produces opposite outcomes on RBC-induced alloantibody formation following HOD or KEL RBC transfusion. (A) Schematic of experimental overview of HOD or KEL RBC storage followed by analysis of antigen levels, post-transfusion RBC survival, and alloantibody formation. (B,C) Representative flow cytometric histograms (B) and quantification of antigen levels of fresh (D0) or stored (D8 or D14) HOD or KEL RBCs (C). (D) 24-h post-transfusion survival of fresh (D0) or stored (D8 or D14) HOD or KEL RBCs. (E) IgM and IgG antibody formation 5 days (D5) or 14 days (D14) following transfusion of fresh or stored HOD RBCs. (F) IgM and IgG antibody formation 5 days (D5) or 14 days (D14) following transfusion of fresh or stored KEL RBCs. Results are combined from two experiments with at least five mice per group. ns, not significant, *p < 0.05, **p < 0.01, ****p < 0.0001

To define the impact of storage on RBC-induced alloimmunization, we examined antibody production following transfusion of fresh or stored RBCs. Very little data exist regarding the impact of storage on IgM antibody formation following RBC transfusion. Unlike prior reports on IgG antibody formation, storage largely failed to alter IgM antibody development following HOD or KEL RBC transfusion (Figure 1E,F). In contrast, storage of HOD RBCs resulted in stepwise increase in IgG anti-HOD antibody development, similar to previous studies (Figure 1E).^10,11 Storage of KEL RBCs produced the opposite outcome, with RBC storage resulting in a reduced IgG anti-KEL antibody response (Figure 1F).

The opposite outcomes of storage on HOD or KEL RBC immunization could reflect distinct transgenic origins of each individual murine strain irrespective of their target antigen. To control for this possibility, we crossed HOD and KEL mice to generate offspring that produced RBCs expressing both HOD and KEL antigens (HOD × KEL RBCs) (Figure 2A). Similar to HOD or KEL RBCs,HOD × KEL RBCs experienced little change in HOD or KEL antigen levels following storage, although some reductions in each antigen were observed with increased storage duration (Figure 2B,C). HOD × KEL RBCs like wise demonstrated decreased post-transfusion survival with increased storage length (Figure 2D), similar to that observed following storage and transfusion of HOD or KEL RBCs.

FIGURE 2.

The impact of pre-transfusion storage on antigen-specific alloantibody responses is largely retained when distinct RBC antigens are expressed on the same cells. (A) Schematic of experimental overview of the storage of RBCs expressing both HOD and KEL antigens (HOD × KEL RBCs), followed by analysis of antigen levels, post-transfusion RBC survival, and alloantibody formation. (B,C) Representative flow cytometric histograms (B) and quantification of antigen levels of fresh (D0) or stored (D8 or D14) HOD × KEL RBCs (C). (D) 24-h post-transfusion survival of fresh (D0) or stored (D8 or D14) HOD × KEL RBCs. (E) IgM and IgG antibody formation against HOD 5 days (D5) or 14 days (D14) following transfusion of fresh or stored HOD × KEL RBCs. (F) IgM and IgG antibody formation against KEL 5 days (D5) or 14 days (D14) following transfusion of fresh or stored HOD × KEL RBCs. (G) Analysis of HOD and KEL antigens on WT or HOD × KEL (H × K) RBCs as shown. Results are combined from two experiments with at least five mice per group. ns, not significant, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001

We next examined antibody production following transfusion of fresh or stored HOD × KEL RBCs to determine whether the impact of storage on HOD or KEL-induced immunization was retained when both antigens were present on the same cell. IgM anti-HOD or anti-KEL antibodies were observed following HOD × KEL RBC transfusion, with antibody levels largely failing to exhibit any decrease regardless of pre-transfusion storage. IgG antibody formation following transfusion of fresh or stored HOD × KEL RBCs largely followed similar trends observed after transfusion of fresh or stored RBCs expressing the HOD or KEL antigen, with the anti-HOD antibody response demonstrating a positive correlation with storage, while pre-transfusion storage of HOD × KEL RBCs either failing to similarly increase or resulting in reduced IgG anti-KEL antibody formation (Figure 2E,F). Importantly, the ability of HOD × KEL RBCs to recapitulate HOD or KEL RBCs alone did not appear to reflect the presence of HOD or KEL single-positive populations among transfused HOD × KEL RBCs, as RBCs were double positive for both antigens (Figure 2G).

The ability of storage to differentially impact RBC-induced alloimmunization in an antigen-specific manner adds to growing evidence that distinct RBC antigens may engage unique immune pathways when inducing alloantibody formation. Prior studies have demonstrated that the role of complement, innate immune signaling events and even the relative requirement of CD4 T cells can fundamentally differ between distinct RBC alloantigens.^18,24–26 These outcomes appear to be driven in part by the unique antigens themselves^18,24–26; this is further evident by the present results, as the impact of RBC storage on alloimmunization outcomes was largely preserved even when HOD and KEL were present on the same RBC. As with any approach, this study is not without limitations. Whether responses to HOD or KEL antigens in mice recapitulate alloimmunization to RBC antigens in humans remains unknown. The HOD and KEL systems represent transgenic founders, where the HOD and KEL antigens are expressed on the RBC surface under a beta-globin promoter. While no difference in RBC survival or other phenotypic variation in RBCs isolated from these founders has been observed,¹⁶ whether this expression displaces endogenous mouse Duffy or KEL is not known.

Furthermore, while this RBC storage model is well-established,^9–11,17 whether it reflects clinical storage outcomes, which can be quite variable and may result in significant differences in post-transfusion survival between donors,^19,27,28 remains to be fully determined. This is especially important to consider when examining the outcome of RBC transfusion following 14 days of storage, where post-transfusion recovery was approximately 25%, far lower than that expected for FDA-approved storage conditions.²⁹ Whether this reflects “poor storers” observed clinically remains to be tested.²⁹ Intriguingly, when HOD and KEL are expressed on the same cell the impact of antigen-specific storage outcomes was preserved, strongly suggesting that even with mixed sex donor RBCs, storage can differentially impact alloimmunization. However, strain-specific modifications of storage in addition to donor sex, which can likewise impact RBC storage,^19,30 may also impact antigen-specific storage outcomes. Despite these limitations, the ability of this system to model key features of RBC alloimmunization provides important insight into general principles of RBC storage that may shape the likelihood of alloimmunization. As such, our study provides a possible explanation for inconsistencies observed in clinical studies and may enlighten future investigations aimed at preventing RBC alloimmunization.