Regulation of primary alloantibody response through antecedent exposure to a microbial T-cell epitope

Abstract

Humoral alloimmunization to red blood cell (RBC) antigens is a clinically significant problem that can lead to transfusion reactions and difficulty in locating future compatible blood for transfusion. However, factors regulating responder/nonresponder status are only partially understood. Herein, we identify a series of microbes with 100% identity in 8- to 9-amino acid peptides containing the variant amino acids in Kell, Kidd, and Duffy antigens. To test the hypothesis that infection with such a microbe could predispose to RBC alloimmunization, a mouse model was developed using murine polyoma virus expressing a defined CD4(+) T-cell epitope ovalbumin(323-339) ((OVA)(323-339)) and subsequent transfusion with RBCs expressing a B-cell epitope (hen egg lysozyme [HEL]) fused to (OVA)(323-339). Whereas infection alone induced no detectable anti-HEL, subsequent RBC transfusion induced 100- to 1000-fold more anti-HEL in mice that had been previously infected compared with control mice. This effect did not occur with wild-type polyoma virus or RBCs expressing HEL alone. Together, these data indicate that prior exposure to a pathogen with small peptide homology to RBC antigens can lead to an enhanced primary alloantibody response. As such priming is not detectable by current clinical tests, it is unknown to what extent this occurs in human alloimmunization.

Figure 1

Experimental design and presentation of (OVA)_323-339 peptide after exposure to either PyV.OVA-II or HOD RBCs. Recipient mice were infected with wild-type polyoma virus (PyV.WT) or polyoma virus expressing a known T-cell epitope (ovalbumin (OVA)_323-339), PyV.OVA-II. Two weeks after infection, recipient mice were transfused with RBCs from control FVB donors (expressing no HEL epitope), mHEL donors (expressing HEL on RBCs), or HOD donors (expressing RBC-specific HEL fused to OVA). (A) Diagram of experimental design: Sera were collected prior to transfusion and at 7 and 14 days after transfusion. (B) Diagram of B- and T-cell epitopes expressed in PyV.WT and PyV.OVA-II, as well as in FVB RBCs, mHEL RBCs, and HOD RBCs. Presence of the (OVA)_323-339 epitope is indicated by the green line. (C) OT-II Thy1.1 splenocytes were adoptively transferred into C56BL/6 mice followed by infection with either PyV.WT (C left) or PyV.OVA-II (C right). At 8 days after infection, splenocytes were harvested and stained for CD4 and Thy1.1. (D) OT-II splenocytes were adoptively transferred into C56BL/6.PL-Thy1.1 mice followed 24 hours later by transfusion with either HOD RBCs or FVB RBCs. At 4 days after transfusion, splenocytes were harvested and stained for CD4 and Thy1.2. Groups included mice receiving OT-II cells alone (left), OT-II cells and HOD transfusion (right), or OT-II cells and FVB transfusion (bottom).

Figure 2

Prior infection with PyV.OVA-II significantly enhances alloimmunization to a subsequent HOD RBC transfusion. C57BL/6 mice were infected with wild-type polyoma virus (PyV.WT) or polyoma virus expressing (OVA)_323-339 (PyV.OVA-II). Additional control mice were uninfected. All groups received subsequent transfusion with HOD RBCs. Alloimmunization was assessed 2 weeks later by anti-HEL ELISA. (A) A representative experiment with 5 mice/group is shown, with mean ± SEM shown at sera dilutions of 1:50 to 1:50 000. (B) Enhancement was also evident by flow cytometric cross-matching with FVB or HOD RBC targets. These experiments have been reproduced 3 times (5 mice/group per experiment) with similar results.

Figure 3

Quantitative real-time PCR enumeration of viral genome copy number. DNA was extracted from splenic tissue taken 7 to 10 weeks after infection from recipient mice infected with wild-type polyoma virus (PyV.WT) or polyoma virus expressing (OVA)_323-339 (PyV.OVA-II) and then transfused with HOD RBCs. Quantitative RT-PCR was used to determine the polyoma virus genome copy number. No statistically significant difference in genome copy number between the 2 groups was seen. A compilation of data from 3 individual experiments (3-5 mice/group per experiment) is shown; uninfected mice had an undetectable viral genome copy number (data not shown).

Figure 4

Enumeration of CD8⁺ T-cell antiviral responses to both PyV.WT and PyV.OVA-II. Intracellular cytokine staining was performed on splenocytes 7 to 10 weeks after infection with either wild-type PyV.WT or PyV.OVA-II; uninfected controls were also analyzed. Splenocytes were restimulated with and without a dominant MHC class I–restricted PyV peptide (LT_359-368) and stained with anti-IFNγ. Representative flow plots are shown (A), and compiled data are presented from a representative experiment (B). This experiment has been repeated 3 times (3-5 mice/group per experiment). In all cases, CD8⁺ T cells from infected mice expressed IFNγ upon peptide stimulation. In 2 of 3 experiments, a greater percentage of IFNγ-producing CD8⁺T cells was seen in PyV.OVA-II–infected mice compared with PyV.WT-infected mice; the data shown are from a representative experiment displaying this difference.

Figure 5

Enhanced alloimmunization to HOD transfusion by PyV.OVA-II requires linkage of (OVA)_323-339 to HEL on the antigen. C57BL/6 mice were infected with PyV.WT or PyV.OVA-II and transfused with mHEL or HOD RBCs (uninfected and/or untransfused control groups were also included). One week after transfusion, sera were analyzed for anti-HEL IgG. Similar results were observed at 14 days (data not shown). A representative experiment with 5 mice/group is shown (mean ± SD depicted, with sera at 1:50 dilution); this experiment has been repeated twice (using C57BL/6 as well as C57BL/6 × B10.BR recipients) with similar results.

Figure 6

Schematic of proposed enhanced alloimmunization. Response to infection. Peptides containing a polymorphism (designated by A) from a microbe will be processed and presented by host antigen-presenting cells; peptides presented in MHC II will be recognized by CD4⁺ T cells. However, in the absence of a B-cell epitope, the B cells will not be able to receive CD4⁺ T-cell help to generate an antibody response. Response to transfusion. Upon a second antigenic exposure (transfusion) that contains the same polymorphism A, the polymorphism in the blood group constitutes not only a CD4⁺ T-cell epitope but also a de novo B-cell epitope. Through receptor-mediated endocytosis, naive B cells phagocytose the polymorphism-containing blood group molecule. The polymorphism is then presented on the MHC II of B cells to the preformed helper or memory CD4⁺ T cells generated against the microbial infection. The B cells are then stimulated to differentiate into plasma cells that secrete antibodies against the polymorphism.