Packed Red Blood Cell Transfusion Modulates Myeloid Dendritic Cell Activation and Inflammatory Response In Vitro

Abstract

Transfusion of packed red blood cells (PRBCs) modulates patients' immune responses and clinical outcomes; however, the underpinning mechanism(s) remain unknown. The potential for PRBC to modulate myeloid dendritic cells (mDC) and blood DC antigen 3 was assessed using an in vitro transfusion model. In parallel, to model processes activated by viral or bacterial infection, toll-like receptor agonists polyinosinic:polycytidylic acid or lipopolysaccharide were added. Exposure to PRBC upregulated expression of CD83 and downregulated CD40 and CD80 on both DC subsets, and it suppressed production of interleukin (IL)-6, IL-8, IL-12, tumor necrosis factor-α, and interferon-gamma-inducible protein-10 by these cells. Similar effects were observed when modeling processes activated by concurrent infection. Furthermore, exposure to PRBC at date of expiry was associated with more pronounced effects in all assays. Our study suggests PRBC have an impact on recipient DC function, which may result in failure to establish an appropriate immune response, particularly in patients with underlying infection.

Introduction

Transfusion of packed red blood cells (PRBCs) is widely used as an intervention to support patients with hematological disease, severe traumatic or surgical blood loss. Although PRBC transfusion is life saving, it is not without risk. Clinical studies have associated PRBC transfusion with adverse outcomes in several patient cohorts including trauma (Ciesla and others 2005; Croce and others 2005; Netzer and others 2007; Bochicchio and others 2008; Marik and Corwin 2008), intensive care (Zilberberg and others 2007; Marik and Corwin 2008; Steiner and others 2015; Vamvakas and Carven 2000), surgery (general, cardiac, colorectal, and orthopedic) (Carson and others 1998; Engoren and others 2002; Innerhofer and others 2005; Marik and Corwin 2008; Glance and others 2011; Salvin and others 2011; Lacroix and others 2015), and oncology (von Doersten and others 1992; Shiba and others 2009; Kneuertz and others 2011). Outcomes of these studies were based on endpoints including morbidity, mortality, infectious complications, length of hospital stays, cancer recurrence, and other adverse outcomes.

PRBC is one of the major blood components derived from donor whole blood, where the RBCs are separated from platelets and plasma, and are depleted of leukocytes via filtration (Council of Europe 2008). PRBC are routinely stored for up to 42 days (date of expiry) in the presence of storage additive solution (saline-adenine-glucose-mannitol) before transfusion. During ex vivo storage at 4°C (Council of Europe 2008), RBC within the product undergo significant modifications that can alter cellular function and viability (Mitrofan-Oprea and others 2007; Glynn and others 2016). This phenomenon was originally described by Gabrio and others (1954) who demonstrated that refrigerated RBC age more rapidly than those in vivo. While storage-related morphological and biochemical modifications in stored PRBC are well documented (Van de Watering 2011), whether these changes contribute to worse patient outcomes remains controversial, with correlation of stored PRBC with poor patient outcomes in some studies (Mynster and Nielsen 2001; Leal-Noval and others 2003; Dunn and others 2012; Dhabangi and others 2015; Lee and others 2015), but not others (Yap and others 2008; Kaukonen and others 2014; Lacroix and others 2015; Steiner and others 2015; Yamal and others 2015; Heddle and others 2016). Recent prospective randomized controlled trials did not find an association between storage duration of PRBC and adverse clinical outcomes (Kaukonen and others 2014; Lacroix and others 2015; Steiner and others 2015; Yamal and others 2015; Heddle and others 2016). However, conclusions that can be drawn regarding transfusion of PRBC near date of expiry are limited as these randomized trials are restricted due to ethical considerations. Due to the study design, the average storage duration of PRBC that were considered “old” in these studies only approximates half of the actual storage period. The impact of transfusing stored PRBC near date of expiry on modulating patient outcomes remains a highly controversial issue in transfusion medicine.

As cellular processes associated with transfusion of stored PRBC remains undefined, in vitro and in vivo models have been used to study and understand the potential for PRBC to modulate recipients' immune response and impact patient outcomes. In vitro studies demonstrate modulation of monocyte (Muszynski and others 2011) and overall leukocyte (Biedler and others 2002; Karam and others 2009; Schneider and others 2009) inflammatory responses following exposure to PRBC or PRBC components. These studies consistently report suppression of tumor necrosis factor (TNF)-α and augmentation of interleukin (IL)-10 responses. Increased monocyte erythrophagocytosis of stored PRBC has also been reported (Veale and others 2014). Transfusion-related immune modulation is not limited to innate cells, with exposure to PRBC reported to increase induction of CD25⁺ regulatory T lymphocytes, reduce the number of responder T lymphocytes (Baumgartner and others 2009), and suppress B and T lymphocyte proliferation (Bernard and others 2010; Long and others 2013). To date, studies assessing the role of dendritic cells (DC) in transfusion outcomes are limited. A murine model using non-leukodepleted PRBC reported CD200⁺ DC may play a role in modulating transfusion-associated tumor growth (Clark and others 2008). Erythrophagocytosis by murine plasmacytoid DC has been reported, particularly in combination with polyinosinic:polycytidylic acid (polyI:C)-driven inflammation (Richards and others 2016). Murine studies have reported that splenic DC are important for outcomes in terms of alloimmunization (Hendrickson and others 2007; Calabro and others 2016). Despite these murine studies, our understanding of whether DC play a role in modulating patient outcomes following PRBC transfusion remains largely undefined.

DC play a crucial and central role in the immune system at the interface of innate and adaptive immune responses (Hoebe and others 2004). DC are classified into 2 major lineages, myeloid DC (mDC) and plasmacytoid DC (Wu and Liu 2007), which are further divided into myeloid blood DC antigen (BDCA)1⁺ (CD1c⁺), myeloid BDCA3⁺ (CD141⁺), plasmacytoid BDCA2⁺ (CD303⁺), and plasmacytoid BDCA4⁺ (CD304⁺) DC (Dzionek and others 2000; MacDonald and others 2002). These cells have specialized functions shaped by the differential expression of numerous pattern recognition receptors (PRRs) including toll-like receptors (TLRs) and C-type lectin receptors. These receptors have the capacity to recognize pathogen-associated molecular patterns, or damage-associated molecular patterns (DAMPs), which enable DC to detect invading pathogens, or damaged cell components respectively. mDC are the more abundant lineage and they predominantly express PRR primarily for recognition of bacteria and subsequent antigen presentation via major histocompatibility complex (MHC) class II. Interestingly, a rare subset of mDC, BDCA3⁺ DC are equipped with PRR skewed toward recognition of viruses, and these cells are specialized in dead cell-associated antigen cross-presentation (via MHC class I) (Jongbloed and others 2010; Cohn and others 2013). Following pattern recognition, DC become activated and undergo phenotypic changes essential for antigen presentation including expression of co-stimulatory molecules and secretion of cytokines and chemokines (Banchereau and Steinman 1998; Yao and others 2002). Perturbation of these DC processes has the potential to significantly affect host immune responses (Elia and others 2008; Mancino and others 2008), which may contribute to poor outcomes in transfusion recipients.

While PRBC have been reported to modulate cells of the innate and adaptive immune response, the impact of transfusion of DC responses remains largely unexplored. As central cells in the immune response, we investigated whether mDC and the specialized subset BDCA3⁺ DC were modulated following exposure to PRBC using an in vitro transfusion model. Further, the impact of PRBC on mDC and BDCA3⁺ DC capacity to respond adequately to concurrent immune challenges such as infection was modeled using TLR agonists polyI:C (“viral”) or lipopolysaccharide (LPS; “bacterial”). Herein, we sought to address the hypothesize that DC are particularly vulnerable to modulation by transfusion due to their multifaceted role in pattern recognition, phagocytosis, activation, and signaling the adaptive immune response, and propose changes in DC phenotype and function are 1 mechanism underpinning transfusion-related immune modulation.

Materials and Methods

Ethics

The Australian Red Cross Blood Service (Blood Service) Human Research Ethics Committee and University of Queensland School of Medicine Low Risk Ethical Review Committee approved this study.

Blood components

Leukodepleted PRBC units were obtained from the Blood Service. Whole blood units (450 ± 45 mL) were collected into top-and-bottom bags containing citrate phosphate dextrose (66.5 mL; Macopharma, Mouvaux, Nord, France), and processed within 24 h of collection according to standard Blood Service protocols based on the Council of Europe “Guidelines for the preparation, use and quality assurance of blood components” (Council of Europe 2008). Briefly, whole blood units were spun (3,640 g, 10 min, 22°C) and RBC were separated into the bottom bag containing storage additive (saline, adenine, glucose and mannitol solution; 105 mL) using a MacoPress component extractor (Macopharma). PRBC units (≥185 mL; 1.0 × 10⁶ leukocytes per unit) were stored at 2°C–6°C, and used “fresh” (D2) or “stored” (D42, date of expiry).

In vitro whole blood model of transfusion

An in vitro whole blood model was used to assess potential immune modulation (mDC and BDCA3⁺ DC surface antigen expression and inflammatory mediator production) in a transfusion recipient. This assay has been established to more closely model transfusion by using a whole blood approach that facilitates cell-to-cell interactions of the PRBC with leukocytes subsets without the need for cell isolation procedures involving density gradients or cell separation (Perros and others 2015). Fresh whole blood samples from healthy volunteers were collected in ethylenediaminetetraacetic acid (EDTA) blood collection tubes [Becton Dickinson (BD) Biosciences, NJ] and served as “transfusion recipient” in our model. EDTA was selected as the preferred anticoagulant in this model as it has shown to be consistent and suitable for analysis of cytokine inflammatory responses (Thavasu and others 1992; Flower and others 2000). A 25% blood replacement volume transfusion (representing 2–3 U) was modeled through culturing “recipient” whole blood (6 mL) with ABO compatible PRBC [D2 (fresh) or D42 (stored) PRBC (from different units); 3 mL] and Roswell Park Memorial Institute (RPMI) 1640 media (3 mL; Invitrogen, CA) for 5.5 h (37°C, 5% CO₂). PolyI:C (50 μg/mL, high molecular weight; Invitrogen) or LPS (1 μg/mL; Sigma-Aldrich, St. Louis, Missouri) was added in duplicate tubes to model the processes activated by concurrent viral (polyI:C) or bacterial (LPS) infections. Matched “no transfusion” controls were included in parallel with PRBC volume replaced with RPMI 1640 (included polyI:C and LPS only controls). For experiments assessing intracellular mediators, GolgiPlug (1 μg/mL; BD Biosciences) was added for the last 4.5 h. Cells were harvested and supernatant collected from cultures without GolgiPlug [stored at −80°C for later analysis using cytometric bead array (CBA)]. RBC were lysed [10 min room temperature (RT), FACS lyse; BD Biosciences] and residual leukocytes washed thrice in 3% fetal bovine serum in phosphate-buffered saline (3% FBS/PBS, both from Invitrogen) before staining for flow cytometry.

Assessment of DC surface antigen expression and intracellular inflammatory mediator production

Leukocytes were stained (15 min, 4°C) with the following DC panel of mouse anti-human mAb (all from BD Biosciences unless stated otherwise): “lineage negative”-fluorescein isothiocyanate (Lin⁻-FITC; CD3, CD14, CD19, CD20, CD34, CD56), CD45⁻ peridinin chlorophyll (PerCP), CD11c-brilliant violet (BV)421, and BDCA3⁻ allophycocyanin (APC; Miltenyi Biotec, Bergisch Gladbach, Germany). For assessment of surface antigen expression, leukocytes were also stained with CD40-phycoerythrin (PE) and CD80-BV510 or CD83-PE and CD86-BV510. Cells were washed twice (3% FBS/PBS) and resuspended in 3% FBS/PBS for flow cytometric analysis. For assessment of intracellular mediators, following staining with the DC panel outlined above, leukocytes were permeabilized (10 min, RT; FACS Perm; BD Biosciences) and labeled with intracellular PE-conjugated mAb targeting IL-6, IL-8, IL-12, TNF-α and, IL-10 (LPS panel) or, interferon-gamma-inducible protein (IP)-10 (polyI:C panel) for 30 min (RT, dark). Cells were washed (3% FBS/PBS) and resuspended in cell stabilizing fixative (BD Biosciences) for flow cytometry.

Flow cytometry gating strategies

mDC were gated as Lin⁻CD45⁺CD11c⁺ and BDCA3⁺ DC as Lin⁻CD45⁺CD11c⁺BDCA3⁺. Briefly, adopting gating strategy described in our previous study (Ki and others 2017): Side scatter versus CD45 was used to exclude granulocytes and lymphocytes, and selected mononuclear cells were displayed as Lin versus CD11c with Lin⁻CD11c⁺ cells selected as mDC. The selected mDC population was then further selected for gating of the BDCA3⁺ DC subset. Median fluorescent intensity (MFI) for each surface antigen and inflammatory mediator expressed by mDC and BDCA3⁺ DC was determined and all statistical analyses were performed on raw data (see Statistical Analysis section below).

Data presented in figures were derived from a 2-step approach. First, PRBC “transfusion” data were normalized (using a ratio of the MFI) to the matched “no transfusion” control. Specifically, data derived from “recipient” whole blood+D2 or D42 PRBC was normalized to the matched “recipient” whole blood only (ie, “no transfusion”) control. Similarly for the whole blood approach modeling processes activated by infection, data from “recipient” whole blood+D2 or D42 PRBC+polyI:C or LPS were normalized to the matched “recipient” whole blood+polyI:C or LPS only control. The MFI ratios were then calculated as binary logarithm (log2). This approach effectively “equalizes” the data around zero to display a fold change. A 2-fold increase is therefore +1 and a 2-fold decrease is therefore as −1 compared to the matched control. This approach establishes a baseline for each donor to account for donor-to-donor variation (representative data Supplementary Figs. S1 and S2; Supplementary Data are available online at www.liebertpub.com/jir). Experiment replicates were as follows: surface antigen expression (n = 16 PRBC only, n = 8 PRBC+polyI:C, n = 8 PRBC+LPS), intracellular mediators (n = 11 PRBC only, n = 6 PRBC+polyI:C, n = 5 PRBC+LPS) (Ki and others 2017).

Assessment of overall leukocyte inflammatory response

Concentrations of inflammatory mediators [IL-1α, IL-1β, IL-8, IL-6, IL-10, TNF-α, IL-12p70, IP-10, interferon (IFN)-α, IFN-γ], monocyte chemoattractant protein (MCP)-1, macrophage inflammatory proteins (MIP)-1α, MIP-1β were assessed in culture supernatant from the transfusion model (n = 16 PRBC only, n = 8 PRBC+polyI:C, n = 8 PRBC+LPS) using CBA according to manufacturer's directions (BD Biosciences). Concentrations were determined from standard curves that ran in parallel.

Assessment of erythrophagocytosis

D2 and D42 PRBC were labeled with FITC-conjugated-function-spacer-lipid (50 μg/mL, FSL-FLRO4; Kode Biotech Materials Limited, Auckland, New Zealand) as described (Blake and others 2011; Ki and others 2016). Briefly, an equal volume of PRBC and FSL-FLRO4 were vortexed and incubated (37°C, 1 h, dark), then washed and resuspended in PBS (500 μL). Peripheral blood mononuclear cells (PBMCs), isolated using Ficoll-Paque (GE Healthcare, Munich, Bavaria, Germany) density gradient centrifugation were stained with HLA-DR-PerCP, BDCA3-APC, CD11C-BV421, and CD14-V500 (15 min, 4°C, dark). FSL-FLRO4-PRBC (D2 or D42) were mixed with PBMC (10:1 ratio), and incubated (30 min, 37°C, 5% CO₂) before noninternalized PRBC were lysed (FACS lyse; BD Biosciences). Phagocytosis of PRBC (expressed as percent positive and MFI) was determined by flow cytometry in gated monocyte (CD14⁺), mDC (CD14⁻DR⁺CD11c⁺), and BDCA3⁺ DC (CD14⁻ DR⁺CD11c⁺BDCA3⁺) populations.

Flow cytometry

Three laser FACSCanto™ II flow cytometer (BD Biosciences) was used and data analyzed using FCS express V5 (De Novo, Los Angeles) or FCAP Array software (BD Biosciences).

Statistical analysis

GraphPad Prism (GraphPad Software, Santiago) was used for all statistical analyses and representation of graphs. Analysis of variance (ANOVA) with Tukey's post-tests were used on raw data to assess the differences in surface antigen expression and inflammatory mediator production following PRBC exposure. A separate ANOVA was performed for each experimental outcome, (1) assessment of changes in DC responses to PRBC transfusion [no transfusion (ie, “recipient” whole blood only) versus D2 PRBC versus D42 PRBC], (2) assessment of changes in DC responses to poly I:C in the presence of PRBC [no transfusion (ie “recipient” whole blood only)+polyI:C versus D2 PRBC+polyI:C versus D42 PRBC+polyI:C], and (3) assessment of changes in DC responses to LPS in the presence of PRBC [no transfusion (ie “recipient” whole blood only)+LPS versus D2 PRBC+LPS versus D42 PRBC+LPS]. Phagocytosis data were compared using paired t-tests. For all analyes P < 0.05 was considered significant.

Results

Exposure to PRBC modulated mDC and BDCA3⁺ DC surface antigen expression

An in vitro whole blood model of transfusion was used to assess the impact of PRBC on activation and maturation of mDC and BDCA3⁺ DC. Changes in cell-specific responses of these DC subsets were investigated using multicolor flow cytometry. Exposure to PRBC alone downregulated mDC expression of co-stimulatory molecules CD40 (P < 0.0001) and CD80 (P = 0.019), and upregulated expression of CD83 (P < 0.0001; Fig. 1A–C). When polyI:C was cocultured with PRBC to model the processes activated by viral infection, mDC expression of CD80 (P = 0.006) was downregulated and CD83 was upregulated (P = 0.006, Fig. 1C) and there was no change in expression of CD40 or CD86 compared to polyI:C alone (Fig. 1A, D). In the model of bacterial infection, mDC expression of CD86 (P = 0.030, Fig. 1D) was downregulated, and there was no change in CD40, CD80, or CD83 (Fig. 1A–C).

FIG. 1.

Assessment of mDC and BDCA3⁺ DC expression of maturation and co-stimulatory molecules following exposure to D2 and D42 PRBC. Data represent flow cytometric analysis of mDC (A–D; circle ) and BDCA3⁺ DC (E–H; square ) expression of maturation and co-stimulatory molecules after being exposed to D2 or D42 PRBC with or without polyI:C or LPS (n = 16 PRBC only, n = 8 PRBC+polyI:C, n = 8 PRBC+LPS. Raw data were analyzed using ANOVA (*P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001) with Tukey's post-tests (^ano transfusion (No Tx) versus D2 PRBC, ^bNo Tx versus D42 PRBC, ^cD2 PRBC versus D42 PRBC). Data presented were normalized (log2) to the matched no Tx control (dotted line at zero). Bars indicate mean ± SEM. ANOVA, analysis of variance; BDCA, blood dendritic cell antigen; LPS, lipopolysaccharide; mDC, myeloid dendritic cells; PRBC, packed red blood cell; SEM, standard error of mean.

For BDCA3⁺ DC, exposure to PRBC alone downregulated expression of CD80 (P = 0.006, Fig. 1F) and upregulated CD83 (P < 0.0001, Fig. 1G), with no changes in expression of CD40 or CD86 (Fig. 1E, H). In the presence of polyI:C, exposure to PRBC was associated with reduced BDCA3⁺ DC expression of CD86 (P = 0.006) (Fig. 1H). When LPS was used to model the processes activated by bacterial infection, BDCA3⁺ DC expression of CD86 (P = 0.0006) was suppressed and CD83 (P = 0.005) was increased in the presence of PRBC (Fig. 1G, F). For both DC subsets, modulation of maturation and co-stimulatory molecules were particularly evident following exposure to stored (D42) PRBC (Tukey's post-tests).

Exposure to PRBC downregulated mDC and BDCA3⁺ DC production of inflammatory mediators

Using the same whole blood approach we further assessed the impact of PRBC on cytokine and chemokine production by mDC and BDCA3⁺ DC using intracellular staining. Two panels of inflammatory mediators were used based on responses to viral or bacterial stimuli. Expression of inflammatory mediators was not significantly altered following exposure to PRBC alone for mDC (Fig. 2A–F). When PRBC were cocultured with polyI:C, mDC production of IL-6 (P = 0.004), IL-8 (P = 0.003), IL-12 (P = 0.005), and TNF-α (P = 0.003) was reduced (Fig. 2A–D). Similarly, coculture of PRBC with LPS resulted in suppression of proinflammatory cytokines IL-6 (P = 0.002) and TNF-α (P = 0.015) (Fig. 2A, D).

FIG. 2.

Assessment of mDC and BDCA3⁺ DC cytokine and chemokine production following exposure to D2 and D42 PRBC. Data represent flow cytometric analysis of mDC (A–F; circle ) and BDCA3⁺ DC (G–L; square ) inflammatory mediators after being exposed to D2 or D42 PRBC with or without polyI:C or LPS (n = 11 PRBC only, n = 6 PRBC+polyI:C, n = 5 PRBC+LPS. Raw data were analyzed using ANOVA (*P < 0.05, **P < 0.01, ****P < 0.0001) with Tukey's post-tests (^aNo Tx versus D2 PRBC, ^bNo Tx versus D42 PRBC, ^cD2 PRBC versus D42 PRBC). Data presented were normalized (log2) to the matched no Tx control (dotted line at zero). Bars indicate mean ± SEM.

For BDCA3⁺ DC, however, we found a different response profile. Expression of IL-12 (P = 0.028) and IP-10 (p = 0.047) were suppressed following exposure to PRBC alone (Fig. 2J, K). Coculture of PRBC with LPS did not result in altered production of inflammatory cytokines, however, IL-6 (P = 0.006), IL-8 (P = 0.015), IL-12 (P < 0.0001), TNF-α (P = 0.019), and IP-10 (P = 0.0002) production were significantly suppressed following coculture of PRBC and polyI:C (Fig. 2G–K). For both subsets, changes in production of inflammatory mediators were particularly evident following exposure to D42 PRBC (Tukey's post-tests).

Exposure to PRBC modulated the overall leukocyte inflammatory response

The use of the whole blood approach also allowed us to assess the impact of PRBC on the overall cytokine and chemokine production, representing changes from the mixed leukocyte population as a whole. Exposure to PRBC alone did not result in modulation of the overall leukocyte inflammatory response (Fig. 3A–F). In the presence of polyI:C, IL-8 production was augmented following exposure to D2 PRBC (P = 0.001, Fig. 3A). In the presence of LPS, exposure to PRBC upregulated IL-1β (P = 0.003), IL-6 (P = 0.001), IL-8 (P = 0.020), MIP-1α (P = 0.020), MIP-1β (P = 0.003), and downregulated IP-10 (P = 0.014) production (Fig. 3A–F). Changes in production of inflammatory mediators were particularly evident following exposure to D42 PRBC (Tukey's post-tests). No changes were observed in overall levels of IL-1α, IL-10, IL-12p70, TNF-α, MCP-1, IFN-α, or IFN-γ following exposure to PRBC alone, PRBC with polyI:C nor PRBC with LPS (data not shown).

FIG. 3.

Assessment of the overall leukocyte inflammatory response following exposure to D2 and D42 PRBC. (A–F) Concentration (pg/mL) of cytokines and chemokines secreted into culture supernatant after exposure to D2 or D42 PRBC with or without polyI:C or LPS (n = 16 PRBC only; n = 8 PRBC+polyI:C, n = 8 PRBC+LPS). IL-1α, IL-10, IL-12p70, TNF-α, MCP-1, IFN-α, and IFN-γ data not shown. Bars indicate mean ± SEM. Data were analyzed using ANOVA (*P < 0.05, **P < 0.01, ****P < 0.0001) with Tukey's post-tests (^aNo Tx versus D2 PRBC, ^bNo Tx versus D42 PRBC, ^cD2 PRBC versus D42 PRBC). IFN, interferon; IL, interleukin; MCP, monocyte chemoattractant protein; TNF, tumor necrosis factor.

PRBC storage was associated with increased erythrophagocytosis

We assessed whether phagocytosis of PRBC may be a potential mechanism driving changes in mDC and BDCA3⁺ phenotype in the transfusion model outlined above. While increased uptake of stored PRBC by monocytes has been reported, whether mDC, and BDCA3⁺ DC, contribute to PRBC clearance was unknown. We found that while monocytes were the primary cells associated with PRBC phagocytosis, mDC and BDCA3⁺ DC both phagocytosed D2 and D42 PRBC (Fig. 4A). For all 3 myeloid subsets assessed, D42 PRBC were phagocytosed more than D2 PRBC (Fig. 4B).

FIG. 4.

Assessment of erythrophagocytosis by mDC and BDCA3⁺ DC. (A) Percentage uptake and (B) MFI of phagocytosed FSL-FLRO4 labeled D2 and D42 PRBC by CD14⁺ monocytes (diamond), mDC (circle), and BDCA3⁺ DC (square). Matched donors were used for assessment of D2 and D42 PRBC, data connected by a line (n = 5). Paired t-tests *P < 0.05, **P < 0.01, ***P < 0.001. MFI, median fluorescent intensity.

Discussion

Transfusion of PRBC has been reported to modulate recipients' immune responses. However, little is known about the mechanisms underpinning transfusion-related immune modulation. As DC play a central role in the initiation and regulation of both innate and adaptive immunity, we hypothesized that PRBC transfusion modulates key DC processes, impeding the capacity of these cells to adequately respond to subsequent immune challenges. We observed significant modulation of mDC and BDCA3⁺ DC phenotype following exposure to PRBC in a number of transfusion models (PRBC only and PRBC ± polyI:C/LPS). Of particular importance for outcomes in transfusion recipients, we found changes to DC phenotype were predominantly found in our models of infection, and stored PRBC had a more pronounced effect than fresh PRBC. This study provides evidence that while PRBC transfusion is lifesaving, the potential for patients with concurrent infection to develop worse outcomes following transfusion due to the impeded DC function needs to be considered.

DC maturation and co-stimulation are an essential process for downstream immune initiation and regulation (Banchereau and Steinman 1998). We first assessed the effect of PRBC on mDC and BDCA3⁺ DC maturation and co-stimulatory surface antigen profile. We report augmentation of maturation molecule CD83 and suppression of co-stimulatory molecules (CD40, CD80, and CD86) following exposure to PRBC. CD83 is upregulated following exposure to pathogens (Banchereau and others 2000) or TLR agonists including polyI:C (TLR3) and LPS (TLR4) (Hemont and others 2013). Our finding that exposure to PRBC alone upregulates CD83 suggests that either RBCs themselves, or substances within the PRBC product have the capacity to stimulate DC maturation. Although CD83 increased, the co-stimulatory molecules were predominantly downregulated following exposure to PRBC. Given the complexity of the PRBC product and the changes that occur during storage, it is likely that multiple DC receptors are engaged resulting in differential expression of maturation and co-stimulatory molecules.

In addition to expression of maturation and co-stimulatory molecules, a balanced release of cytokines and chemokines by DC is also crucial for directing multiple immune pathways (Calabro and others 2016). We found minimal modulation of cytokine responses following exposure to PRBC alone, however, both mDC and BDCA3⁺ DC responses to polyI:C and LPS were suppressed in the presence of PRBC. Together poor co-stimulation and lack of cytokine and chemokine production may be associated with inadequate T and B lymphocyte responses (Baumgartner and others 2009; Bernard and others 2010; Long and others 2013) and our data suggest PRBC transfusion has the potential to impair downstream lymphocyte responses.

We hypothesized that changes in mDC and BDCA3⁺ DC phenotype may be in part due to phagocytosis of PRBC. Indeed we report that both mDC and BDCA3⁺ DC play a role in RBC uptake, and stored PRBC were phagocytosed more than fresh PRBC. Therefore, in addition to uptake by monocytes (THP-1) as reported previously (Veale and others 2014), mDC and BDCA3⁺ DC both preferentially recognize and remove stored PRBC. This is likely attributed to the significant changes in PRBC that occur during routine storage that result in expression of a number of DAMP (Bosman and others 2008). It has been reported that inflammation enhances erythrophagocytosis (Hendrickson and others 2007; Richards and others 2016) and the modulation of mDC and BDCA3⁺ DC phenotype observed in this study, particularly in the models processes associated with infection, may be the result of increased erythrophagocytosis.

In addition to studying mDC and BDCA3⁺ DC specific responses, we investigated the overall response of the peripheral blood leukocytes in the culture. This largely represents the expected changes in the response of the more abundant leukocyte populations in patient's plasma following PRBC transfusion. We observed minimal impact on the overall leukocyte inflammatory response in the presence of PRBC alone, and PRBC cocultured with polyI:C. However, PRBC augmented production of IL-1β, IL-8, MIP-1α, and MIP-1β and, suppression of IP-10 following PRBC exposure when modeling processes associated with bacterial infection. We found it interesting that the overall leukocyte inflammatory response was largely augmented by exposure to PRBC whereas the specific cellular responses of mDC and BDCA3⁺ DC were attenuated. These results highlight the potential for mechanisms associated with specific cell subsets to be missed when only assessing “overall inflammation,” which largely represent the response of numerous cells subsets such as neutrophils, lymphocytes, and monocytes.

We studied not only on mDC but also the rarer subset of mDC, BDCA3⁺ DC, as they have a specialized and unique role in the immune response. While mDC express PRR predominantly associated with recognition of bacteria, the PRR profile of BDCA3⁺ DC is skewed toward recognition of viruses. Following receptor ligation, mDC present antigen via the standard MHC class II pathway, but BDCA3⁺ are more superior in cross-presentation of dead cell-associated antigens onto MHC class I. BDCA3⁺ DC have been shown to be important in CD8⁺ T lymphocyte responses (Jongbloed and others 2010; Cohn and others 2013) and have been identified as a key cell involved in anti-tumor immunity (Hildner and others 2008). Whether BDCA3⁺ DC play a role in cancer recurrence in transfusion recipients would be of interest for future studies.

In summary, we report that exposure to PRBC modulated the cellular responses of both mDC and the specialized subset BDCA3⁺ DC. Importantly, we found modulation of DC inflammatory response was particularly evident in models combining transfusion and infection, and, a more prominent affect was observed following exposure to stored PRBC. These results suggest that patients who receive stored PRBC, particularly those with underlying infectious complications, may fail to mount an appropriate immune response precipitating worse patient outcomes.

Footnotes

Acknowledgments

K.K.K. is supported by an Australian Post-graduate Award from the University of Queensland and top-up scholarship from the Alex Steele Young Memorial Lions Foundation. Australian governments fund the Australian Red Cross Blood Service for the provision of blood, blood products, and services to the Australian community. We would like to thank Professor Stephen Henry (Kode Biotech Materials Limited) for provision of FSL constructs for this study.

Author Disclosure Statement

No competing financial interests exist.

References

Banchereau

, Briere

, Caux

, Davoust

, Lebecque

, Liu

, Pulendran

, Palucka

. 2000. Immunobiology of dendritic cells. Annu Rev Immunol, 18:767–811.

Banchereau

, Steinman

. 1998. Dendritic cells and the control of immunity. Nature, 392(6673):245–252.

Baumgartner

, Silliman

, Moore

, Banerjee

, McCarter

. 2009. Stored red clood cell transfusion induces regulatory T cells. J Am Coll Surg, 208(1):110–119.

Bernard

, Meier

, Ward

, Browning

, Montgomery

, Kasten

, Snow

, Manning

, Woodward

. 2010. Packed red blood cells suppress T-cell proliferation through a process involving cell-cell contact. J Trauma, 69(2):320–329.

Biedler

, Schneider

, Seyfert

, Rensing

, Grenner

, Girndt

, Bauer

. 2002. Impact of alloantigens and storage-associated factors on stimulated cytokine response in an in vitro model of blood transfusion. Anesthesiology, 97(5):1102–1109.

Blake

, Bovin

, Bess

, Henry

. 2011. FSL constructs: a simple method for modifying cell/virion surfaces with a range of biological markers without affecting their viability. J Vis Exp (54):pii: 3289.

Bochicchio

, Napolitano

, Joshi

, Bochicchio

, Meyer

, Scalea

. 2008. Outcome analysis of blood product transfusion in trauma patients: a prospective, risk-adjusted study. World J Surg, 32(10):2185–2189.

Bosman

, Werre

, Willekens

, Novotny

. 2008. Erythrocyte ageing in vivo and in vitro: structural aspects and implications for transfusion. Transfus Med, 18(6):335–347.

Calabro

, Gallman

, Gowthaman

, Liu

, Chen

, Liu

, Krishnaswamy

, Nascimento

MSL

, Xu

, Patel

, Williams

, Tormey

, Hod

, Spitalnik

, Zimring

, Hendrickson

, Stowell

, Eisenbarth

. 2016. Bridging channel dendritic cells induce immunity to transfused red blood cells. J Exp Med, 213(6):887–896.

10.

Carson

, Duff

, Berlin

, et al. 1998. Perioperative blood transfusion and postoperative mortality. JAMA, 279(3):199–205.

11.

Ciesla

, Moore

, Johnson

, Burch

, Cothren

, Sauaia

. 2005. A 12-year prospective study of postinjury multiple organ failure: has anything changed?. Arch Surg, 140(5):432–440.

12.

Clark

, Gorczynski

, Blajchman

. 2008. Transfusion-related immunomodulation due to peripheral blood dendritic cells expressing the CD200 tolerance signaling molecule and alloantigen. Transfusion, 48(5):814–821.

13.

Cohn

, Chatterjee

, Esselborn

, Smed-Sorensen

, Nakamura

, Chalouni

, Lee

, Vandlen

, Keler

, Lauer

, Brockstedt

, Mellman

, Delamarre

. 2013. Antigen delivery to early endosomes eliminates the superiority of human blood BDCA3+ dendritic cells at cross presentation. J Exp Med, 210(5):1049–1063.

14.

Council of Europe. 2008. Guide to the preparation, use and quality assurance of blood components (14th Edition). In: European Directorate for the Quality of Medicines and HealthCare of the Council of Europe (EDQM). Council of Europe Publishing, Strasbourg Cedex, France.

15.

Croce

, Tolley

, Claridge

, Fabian

. 2005. Transfusions result in pulmonary morbidity and death after a moderate degree of injury. J Trauma, 59(1):19–23.

16.

Dhabangi

, Ainomugisha

, Cserti-Gazdewich

, et al. 2015. Effect of transfusion of red blood cells with longer vs. shorter storage duration on elevated blood lactate levels in children with severe anemia: the total randomized clinical trial. JAMA, 314(23):2514–2523.

17.

Dunn

, Thiele

, Ma

, Sawyer

, Nemergut

. 2012. Duration of red blood cell storage and outcomes following orthotopic liver transplantation. Liver Transpl, 18(4):475–481.

18.

Dzionek

, Fuchs

, Schmidt

, Cremer

, Zysk

, Miltenyi

, Buck

, Schmitz

. 2000. BDCA-2, BDCA-3, and BDCA-4: three markers for distinct subsets of dendritic cells in human peripheral blood. J Immunol, 165(11):6037–6046.

19.

Elia

, Cappello

, Puppo

, Fraone

, Vanni

, Eva

, Musso

, Novelli

, Varesio

, Giovarelli

. 2008. Human dendritic cells differentiated in hypoxia down-modulate antigen uptake and change their chemokine expression profile. J Leukoc Biol, 84(6):1472–1482.

20.

Engoren

, Habib

, Zacharias

, Schwann

, Riordan

, Durham

. 2002. Effect of blood transfusion on long-term survival after cardiac operation. Ann Thorac Surg, 74(4):1180–1186.

21.

Flower

, Ahuja

, Humphries

, Mohamed-Ali

. 2000. Effects of sample handling on the stability of interleukin 6, tumour necrosis factor-alpha and leptin. Cytokine, 12(11):1712.

22.

Gabrio

, Finch

, Linde

, Rupen

. 1954. Erythrocyte preservation. I. The relation of the storage lesion to in vivo erythrocyte senescence. J Clin Invest, 33(2):242–246.

23.

Glance

, Dick

, Mukamel

, Fleming

, Zollo

, Wissler

, Salloum

, Meredith

, Osler

. 2011. Association between intraoperative blood transfusion and mortality and morbidity in patients undergoing noncardiac surgery. Anesthesiology, 114(2):283–292.

24.

Glynn

, Klein

, Ness

. 2016. The red blood cell storage lesion: the end of the beginning. Transfusion, 56(6):1462–1468.

25.

Heddle

, Cook

, Arnold

, Liu

, Barty

, Crowther

, Devereaux

, Hirsh

, Warkentin

, Webert

, Roxby

, Sobieraj-Teague

, Kurz

, Sessler

, Figueroa

, Ellis

, Eikelboom

. 2016. Effect of short-term vs. long-term blood storage on mortality after transfusion. N Engl J Med, 375(20):1937–1945.

26.

Hemont

, Neel

, Heslan

, Braudeau

, Josien

. 2013. Human blood mDC subsets exhibit distinct TLR repertoire and responsiveness. J Leukoc Biol, 93(4):599–609.

27.

Hendrickson

, Chadwick

, Roback

, Hillyer

, Zimring

. 2007. Inflammation enhances consumption and presentation of transfused RBC antigens by dendritic cells. Blood, 110(7):2736–2743.

28.

Hildner

, Edelson

, Purtha

, Diamond

, Matsushita

, Kohyama

, Calderon

, Schraml

, Unanue

, Diamond

, Schreiber

, Murphy

. 2008. Batf3 deficiency reveals a critical role for CD8alpha+ dendritic cells in cytotoxic T cell immunity. Science, 322(5904):1097–1100.

29.

Hoebe

, Janssen

, Beutler

. 2004. The interface between innate and adaptive immunity. Nat Immunol, 5(10):971–974.

30.

Innerhofer

, Klingler

, Klimmer

, Fries

, Nussbaumer

. 2005. Risk for postoperative infection after transfusion of white blood cell–filtered allogeneic or autologous blood components in orthopedic patients undergoing primary arthroplasty. Transfusion, 45(1):103–110.

31.

Jongbloed

, Kassianos

, McDonald

, Clark

, Ju

, Angel

, Chen

, Dunbar

, Wadley

, Jeet

, Vulink

, Hart

, Radford

. 2010. Human CD141+ (BDCA-3)+ dendritic cells (DCs) represent a unique myeloid DC subset that cross-presents necrotic cell antigens. J Exp Med, 207(6):1247–1260.

32.

Karam

, Tucci

, Toledano

, Robitaille

, Cousineau

, Thibault

, Lacroix

, Le Deist

. 2009. Length of storage and in vitro immunomodulation induced by prestorage leukoreduced red blood cells. Transfusion, 49(11):2326–2334.

33.

Kaukonen

, Bailey

, Ady

, Aubron

, French

, Gantner

, Irving

, Murray

, Nichol

, Pettila

, McQuilten

, Cooper

. 2014. A randomised controlled trial of standard transfusion versus fresher red blood cell use in intensive care (TRANSFUSE): protocol and statistical analysis plan. Crit Care Resusc, 16(4):255–261.

34.

, Faddy

, Flower

, Dean

. 2017. Platelet concentrates modulate myeloid dendritic cell immune responses. Platelets, 15:1–10.

35.

, Flower

, Faddy

, Dean

. 2016. Incorporation of fluorescein conjugated function-spacer-lipid constructs into the red blood cell membrane facilitates detection of labeled cells for the duration of ex-vivo storage. J Immunol Methods, 429:66–70.

36.

Kneuertz

, Patel

, Chu

, Maithel

, Sarmiento

, Delman

, Staley

, Kooby

. 2011. Effects of perioperative red blood cell transfusion on disease recurrence and survival after pancreaticoduodenectomy for ductal adenocarcinoma. Ann Surg Oncol, 18(5):1327–1334.

37.

Lacroix

, Hébert

, Fergusson

, Tinmouth

, Cook

, Marshall

, Clayton

, McIntyre

, Callum

, Turgeon

, Blajchman

, Walsh

, Stanworth

, Campbell

, Capellier

, Tiberghien

, Bardiaux

, van de Watering

, van der Meer

, Sabri

, Vo

. 2015. Age of transfused blood in critically Ill adults. N Engl J Med, 372(15):1410–1418.

38.

Leal-Noval

, Jara-Lopez

, Garcia-Garmendia

, Marin-Niebla

, Herruzo-Aviles

, Camacho-Larana

, Loscertales

. 2003. Influence of erythrocyte concentrate storage time on postsurgical morbidity in cardiac surgery patients. Anesthesiology, 98(4):815–822.

39.

Lee

, Seo

, Kim

, Lim

, Yoon

, Kim

, Ku

, Park

. 2015. Effect of intraoperative red blood cell transfusion on postoperative complications after open radical cystectomy: old versus fresh stored blood. Clin Genitourin Cancer, 13(6):581–587.

40.

Long

, Meier

, Ward

, Williams

, Woodward

, Bernard

. 2013. Immunologic profiles of red blood cells using in vitro models of transfusion. J Surg Res, 184(1):567–571.

41.

MacDonald

KPA

, Munster

, Clark

, Dzionek

, Schmitz

, Hart

DNJ

. 2002. Characterization of human blood dendritic cell subsets. Blood, 100(13):4512–4520.

42.

Mancino

, Schioppa

, Larghi

, Pasqualini

, Nebuloni

, Chen

, Sozzani

, Austyn

, Mantovani

, Sica

. 2008. Divergent effects of hypoxia on dendritic cell functions. Blood, 112(9):3723–3734.

43.

Marik

, Corwin

. 2008. Efficacy of red blood cell transfusion in the critically ill: a systematic review of the literature. Crit Care Med, 36(11):3134.

44.

Mitrofan-Oprea

, Palii

, Tissier

, Heron

, Verpoort

, Behague

, Smagghe

, Schooneman

, Huart

, Goudaliez

, Montreuil

, Bratosin

. 2007. Novel criteria for assessing red blood cell viability in blood banks. Transfus Clin Biol, 14(4):393–401.

45.

Muszynski

, Nateri

, Nicol

, Greathouse

, Hanson

, Hall

. 2011. Immunosuppressive effects of red blood cells on monocytes are related to both storage time and storage solution. Transfusion, 36(9):2667–2674.

46.

Mynster

, Nielsen

. 2001. Storage time of transfused blood and disease recurrence after colorectal cancer surgery. Dis Colon Rectum, 44(7):955–964.

47.

Netzer

, Shah

, Iwashyna

, Lanken

, Finkel

, Fuchs

, Guo

, Christie

. 2007. Association of RBC transfusion with mortality in patients with acute lung injury. Chest, 132(4):1116–1123.

48.

Perros

, Christensen

, Flower

, Dean

. 2015. Soluble mediators in platelet concentrates modulate dendritic cell inflammatory responses in an experimental model of transfusion. J Interferon Cytokine Res, 35(10):821–830.

49.

Richards

, Hendrickson

, Zimring

, Hudson

. 2016. Erythrophagocytosis by plasmacytoid dendritic cells and monocytes is enhanced during inflammation. Transfusion, 56(4):905–916.

50.

Salvin

, Scheurer

, Laussen

, Wypij

, Polito

, Bacha

, Pigula

, McGowan

, Costello

, Thiagarajan

. 2011. Blood transfusion after pediatric cardiac surgery is associated with prolonged hospital stay. Ann Thorac Surg, 91(1):204–210.

51.

Schneider

, Rensing

, Gräber

, Kreuer

, Kleinschmidt

, Kreimeier

, Müller

, Mathes

, Biedler

. 2009. Impact of platelets and fresh frozen plasma in contrast to red cell concentrate on unstimulated and stimulated cytokine release in an in vitro model of transfusion. Scand J Immunol, 70(2):101–105.

52.

Shiba

, Ishida

, Wakiyama

, Iida

, Matsumoto

, Sakamoto

, Ito

, Gocho

, Furukawa

, Fujiwara

, Hirohara

, Misawa

, Yanaga

. 2009. Negative impact of blood transfusion on recurrence and prognosis of hepatocellular carcinoma after hepatic resection. J Gastrointest Surg, 13(9):1636–1642.

53.

Steiner

, Ness

, Assmann

, Triulzi

, Sloan

, Delaney

, Granger

, Bennett-Guerrero

, Blajchman

, Scavo

, Carson

, Levy

, Whitman

, D'Andrea

, Pulkrabek

, Ortel

, Bornikova

, Raife

, Puca

, Kaufman

, Nuttall

, Young

, Youssef

, Engelman

, Greilich

, Miles

, Josephson

, Bracey

, Cooke

, McCullough

, Hunsaker

, Uhl

, McFarland

, Park

, Cushing

, Klodell

, Karanam

, Roberts

, Dyke

, Hod

, Stowell

. 2015. Effects of red-cell storage duration on patients undergoing cardiac surgery. N Engl J Med, 372(15):1419–1429.

54.

Thavasu

, Longhurst

, Joel

, Slevin

, Balkwill

. 1992. Measuring cytokine levels in blood. J Immunol Methods, 153(1):115–124.

55.

Vamvakas

, Carven

. 2000. Length of storage of transfused red cells and postoperative morbidity in patients undergoing coronary artery bypassgraft surgery. Transfusion, 40(1):101–109.

56.

Van de Watering

. 2011. Red cell storage and prognosis. Vox Sang, 100(1):36–45.

57.

Veale

, Healey

, Sparrow

. 2014. Longer storage of red blood cells is associated with increased in vitro erythrophagocytosis. Vox Sang, 106(3):219–226.

58.

von Doersten

, Cruz

, Selby

, Hilsinger

Jr . 1992. Transfusion, recurrence, and infection in head and neck cancer surgery. Otolaryngol Head Neck Surg, 106(1):60–67.

59.

, Liu

. 2007. Development of dendritic-cell lineages. Immunity, 26(6):741–750.

60.

Yamal

, Benoit

, Doshi

, Rubin

, Tilley

, Hannay

, Robertson

. 2015. Association of transfusion red blood cell storage age and blood oxygenation, long-term neurologic outcome, and mortality in traumatic brain injury. J Trauma Acute Care Surg, 79(5):843–849.

61.

Yao

, Platell

, Hall

. 2002. Dendritic cells. ANZ J Surg, 72(7):501–506.

62.

Yap

, Lau

, Krishnaswamy

, Gaskell

, Yii

. 2008. Age of transfused red cells and early outcomes after cardiac surgery. Ann Thorac Surg, 86(2):554–559.

63.

Zilberberg

, Carter

, Lefebvre

, Raut

, Vekeman

, Duh

, Shorr

. 2007. Red blood cell transfusions and the risk of acute respiratory distress syndrome among the critically ill: a cohort study. Crit Care, 11(3):R63.

Packed Red Blood Cell Transfusion Modulates Myeloid Dendritic Cell Activation and Inflammatory Response In Vitro

Abstract

Introduction

Materials and Methods

Ethics

Blood components

In vitro whole blood model of transfusion

Assessment of DC surface antigen expression and intracellular inflammatory mediator production

Flow cytometry gating strategies

Assessment of overall leukocyte inflammatory response

Assessment of erythrophagocytosis

Flow cytometry

Statistical analysis

Results

Exposure to PRBC modulated mDC and BDCA3+ DC surface antigen expression

Exposure to PRBC downregulated mDC and BDCA3+ DC production of inflammatory mediators

Exposure to PRBC modulated the overall leukocyte inflammatory response

PRBC storage was associated with increased erythrophagocytosis

Discussion

Footnotes

Acknowledgments

Author Disclosure Statement

References

Exposure to PRBC modulated mDC and BDCA3⁺ DC surface antigen expression

Exposure to PRBC downregulated mDC and BDCA3⁺ DC production of inflammatory mediators