Polyinosinic: polycytidylic acid induced inflammation enhances while lipopolysaccharide diminishes alloimmunity to platelet transfusion in mice

Abstract

Introduction: Alloimmune responses against platelet antigens, which dominantly target the major histocompatibility complex (MHC), can cause adverse reactions to subsequent platelet transfusions, platelet refractoriness, or rejection of future transplants. Platelet transfusion recipients include individuals experiencing severe bacterial or viral infections, and how their underlying health modulates platelet alloimmunity is not well understood.

Methods: This study investigated the effect of underlying inflammation on platelet alloimmunization by modelling viral-like inflammation with polyinosinic-polycytidylic acid (poly(I:C)) or gram-negative bacterial infection with lipopolysaccharide (LPS), hypothesizing that underlying inflammation enhances alloimmunization. Mice were pretreated with poly(I:C), LPS, or nothing, then transfused with non-leukoreduced or leukoreduced platelets. Alloantibodies and allogeneic MHC-specific B cell (allo-B cell) responses were evaluated two weeks later. Rare populations of allo-B cells were identified using MHC tetramers.

Results: Relative to platelet transfusion alone, prior exposure to poly(I:C) increased the alloantibody response to allogeneic platelet transfusion whereas prior exposure to LPS diminished responses. Prior exposure to poly(I:C) had equivalent, if not moderately diminished, allo-B cell responses relative to platelet transfusion alone and exhibited more robust allo-B cell memory development. Conversely, prior exposure to LPS resulted in diminished allo-B cell frequency, activation, antigen experience, and germinal center formation and altered memory B cell responses.

Discussion: In conclusion, not all inflammatory environments enhance bystander responses and prior inflammation mediated by LPS on gram-negative bacteria may in fact curtail platelet alloimmunization.

Keywords: alloantibodies; alloimmunization; inflammation; lipopolysaccharide; platelets; poly(I:C); underlying health.

Figure 1

Experimental strategy and anti-class I and II MHC antibody screen. (A, B) Schematic diagram modelling the effect of underlying inflammation on platelet alloimmunization. B6 donor platelet was administered I.V. into recipient BALB/c mice on day zero (D0). Four hours prior (-4 Hrs) to platelet transfusion (Tx), LPS or poly(I:C) was injected into recipient BALB/c mice by intraperitoneal (I.P.) route to model underlying inflammation. Blood for serum, spleen, and lymph nodes from recipient mice were harvested on day fourteen (D14) after transfusion for analysis. (C) Schematic diagram of anti-class I MHC antibody screen. Serum from recipient BALB/c mice was used as a primary stain on B6 splenocytes, followed by a secondary antibody (2° ab) stain with a cocktail of α-Igκ, α-IgM, α-IgG1, α-IgG2a, α-IgG2b, and α-IgG3 antibodies to detect total and isotype-specific alloantibodies by flow cytometry. (D) Cells were gated on singlet, lymphocyte, and then T cells (B220^-/TCRβ⁺) to identify bound alloantibodies on a non-Igκ, non-class II MHC population that expresses class I MHC. α-Igκ only shown as illustration. Median fluorescence intensity (MFI) of total α-Igκ alloantibodies for untreated (U) and with transfusion (Tx) sample are shown. (E) Schematic diagram of anti-class II MHC antibody screen. Protocol was similar to anti-class I MHC antibody screen except cultured dendritic cells (DC) from the bone marrow of β2M-KO mice on the B6 background were used as target cells. (F) Cells were gated on singlet, lymphocyte, and then dendritic cells (CD11c⁺/F4/80^-) to identify bound alloantibodies on non-class I MHC population that expresses class II MHC.

Figure 2

Poly(I:C) increases and LPS decreases the alloantibody response to non-leukoreduced platelet transfusion. (A) Total and isotype-specific anti-class I MHC antibody and (B) anti-class II MHC antibody responses at 14 days post non-leukoreduced platelet transfusion. For IgG1 for anti-class I MHC antibody (A), 100 was added to the raw values prior to normalization to adjust for negative values. Normalized MFI in mice that were untreated (), received transfusion alone (), poly(I:C) alone (), poly(I:C) followed by transfusion (), LPS alone (), or LPS followed by transfusion (). (p)<0.05 (*), p<0.01 (**), p<0.001 (***), and p<0.0001 (****).

Figure 3

The effects of underlying inflammation on leukoreduced platelet transfusion are consistent with non- leukoreduced platelet transfusion, although the magnitude of the responses is smaller. Poly(I:C) increases and LPS decreases the alloantibody response to leukoreduced platelets. Normalized MFI in mice that were untreated (), received leukoreduced (LR) platelet transfusion alone (), poly(I:C) followed by leukoreduced transfusion (), LPS followed by leukoreduced transfusion (), and non-leukoreduced (NLR) platelet transfusion alone (). 10 mice per group in a single experiment. For IgG1 and IgG3, 200 was added to the raw values prior to normalization to adjust for negative values. (p)<0.05 (*), p<0.01 (**), p<0.001 (***), and p<0.0001 (****). Tick line represents mean of the group that received a leukoreduced transfusion only.

Figure 4

Platelet transfusion specifically induces allo-B cell responses. (A) B cell subsets of different specificities were examined for responses in untreated mice (U) and after platelet transfusion (Tx), including B cells specific for the MHC tetramer (Tet, allo-B cells), syngeneic decoy (Decoy), single fluorophore (APC or PE), or the remaining B cell population (Neg). Normalized data shows the frequency of (B) the respective B cell subset out of total B cells, (C) CD86⁺ activated B cells of the respective B cell subset (see Figure 6 for gating), (D) IgD^- antigen-experienced B cells of the respective B cell subset (see Figure 7 for gating), and (E) CD95⁺/CD38^- germinal center (GC) B cells of the respective B cell subset (see Figure 8 for gating). (B–E) Results for splenic (top) and lymph node (bottom) cells were stained with MHC class I (left) and MHC class II (right) tetramers. Normalization was achieved by taking the average of the respective untreated group. Tick line falls on mean of untreated group of Neg subset. p<0.01 (**) and p<0.0001 (****).

Figure 5

Transfusion-induced expansion of alloreactive B cells is impaired following exposure to LPS. Frequency of allo-B cells (out of total B cells without syngeneic decoy events) were identified with MHC tetramers by flow cytometry two weeks after platelet transfusion. (A) Cells were first gated on singlet, lymphocytes, live cells, non-B cell^-/B220⁺ cells, and CD19⁺/syngeneic decoy^- B cells, then evaluated for tetramer staining. A two fluorophore (APC and PE)-one MHC tetramer strategy was used with allo-B cells identified as double positive tetramer (tet) events. (B–E) Shown is allo-B cell frequency in mice that were untreated (), received transfusion alone (), poly(I:C) alone (), poly(I:C) followed by transfusion (), LPS alone (), or LPS followed by transfusion (). Splenic and lymph node cells were stained with MHC class I (B, C) and MHC class II (D, E) tetramers. Tick line represents mean of untreated group. (p)<0.05 (*), p<0.01 (**), p<0.001 (***), and p<0.0001 (****).

Figure 6

Transfusion-induced activation of alloreactive B cells is diminished following LPS exposure. (A) Frequency of activated CD86⁺ allo-B cells out of total allo-B cells was evaluated with representative examples shown of untreated and platelet transfused mice. (B–E) Shown is activated allo-B cell frequency in mice that were untreated (), received transfusion alone (), poly(I:C) alone (), poly(I:C) followed by transfusion (), LPS alone (), or LPS followed by transfusion (). Splenic and lymph node cells were stained with MHC class I (B, C) and MHC class II (D, E) tetramers. Tick line represents mean of untreated group. (p)<0.05 (*), p<0.001 (***), and p<0.0001 (****).

Figure 7

Transfusion-induced down-regulation of IgD is impaired with prior LPS exposure. (A) Frequency of antigen experienced IgD^- allo-B cells out of total allo-B cells was evaluated with representative examples shown of untreated and platelet transfused mice. (B–E) Shown is antigen experienced allo-B cell frequency in mice that were untreated (), received transfusion alone (), poly(I:C) alone (), poly(I:C) followed by transfusion (), LPS alone (), or LPS followed by transfusion (). Splenic and lymph node cells were stained with MHC class I (B, C) and MHC class II (D, E) tetramers. Tick line represents mean of untreated group. (p)<0.05 (*), p<0.01 (**), p<0.001 (***), and p<0.0001 (****).

Figure 8

Germinal center response to transfusion is diminished with prior LPS exposure. (A) Frequency of germinal center (GC) CD95⁺/CD38^- allo-B cells out of total allo-B cells was evaluated with representative examples shown of untreated and platelet transfused mice. (B–E) Shown is germinal center allo-B cell frequency in mice that were untreated (), received transfusion alone (), poly(I:C) alone (), poly(I:C) followed by transfusion (), LPS alone (), or LPS followed by transfusion (). Splenic and lymph node cells were stained with MHC class I (B, C) and MHC class II (D, E) tetramers. (p)<0.05 (*), p<0.01 (**), and p<0.0001 (****).

Figure 9

LPS non-specifically enriches for marginal zone B cells. (A) Representative gating selected for CD93^- mature B cells, CD95^-/CD38⁺ non-GC (GC^Neg) B cells, and IgM⁺/CD23^- marginal zone (MZ) B cells. Shown is frequency of total MZ B cells out of total B cells in spleen (B) and lymph node (C). Frequency of marginal zone B cells within the allo-B cell subset population out of total B cells is also shown in the spleen which were stained with MHC class I (D) and MHC class II (E) tetramers. Mice were untreated (), received transfusion alone (), poly(I:C) alone (), poly(I:C) followed by transfusion (), LPS alone (), or LPS followed by transfusion (). Tick line represents mean of untreated group. (p)<0.05 (*), p<0.01 (**), p<0.001 (***), and p<0.0001 (****).

Figure 10

Underlying inflammation induced by poly(I:C) enriches for allo-B memory cells. (A) Gating strategy for allo-B memory cells selected for CD95^-/CD38⁺ non-GC (GC^Neg) B cells, B220⁺/IgD^- antigen (Ag) experienced B cells, and CD93^- mature B cells which represents total memory B cells. Total memory B cells were then selected for CD80^-/PD-L2^- double negative and CD80⁺/PD-L2⁺ double positive subsets of allo-B memory cells. (B–I) Shown is frequency out of total B cells. Splenic and lymph node cells were stained with MHC class I (B, C, F, G) and MHC class II (D, E, H, I) tetramers. (B–E) Frequency of total allo-B memory cells. Frequency of CD80^-/PD-L2^- double negative allo-B memory cells (F, H) and CD80⁺/PD-L2⁺ double positive allo-B memory cells (G, I). Mice were untreated (), received transfusion alone (), poly(I:C) alone (), poly(I:C) followed by transfusion (), LPS alone (), or LPS followed by transfusion (). Tick line represents mean of untreated group. (p)<0.05 (*), p<0.01 (**), p<0.001 (***), and p<0.0001 (****).