Neutrophils Inhibit Synthesis of Mineralized Extracellular Matrix by Human Bone Marrow-Derived Stromal Cells In Vitro

Abstract

Although controlled local inflammation is essential for adequate bone regeneration, several studies have shown that hyper-inflammatory conditions after major trauma are associated with impaired fracture healing. These hyper-inflammatory conditions include the trauma-induced systemic inflammatory response to major injury, open fractures, and significant injury to the surrounding soft tissues. The current literature suggests that increased or prolonged influx of neutrophils into the fracture hematoma may mediate impairment of bone regeneration after hyper-inflammatory conditions. The underlying mechanism remains unclear. We hypothesize that high neutrophil numbers inhibit synthesis of mineralized extracellular matrix (ECM) by bone marrow stromal cells (BMSCs). We therefore studied the effect of increasing concentrations of neutrophils on ECM synthesis by human BMSCs in vitro. Moreover, we determined how high neutrophil concentrations affect BMSC cell counts, as well as BMSC osteogenic activity determined by alkaline phosphatase (ALP) expression and ALP activity. Co-culture of BMSCs with neutrophils induced a 52% decrease in BMSC cell count (p < 0.01), a 64% decrease in the percentage of ALP+ cells (p < 0.001), a 28% decrease in total ALP activity (p < 0.01), and a significant decrease in the amount of mineralized ECM [38% decrease after 4 weeks (p < 0.05)]. Co-cultures with peripheral blood mononuclear cells and neutrophils within transwells did not induce a significant decrease in ALP activity. In conclusion, our data shows that a decreased amount of mineralized ECM became synthesized by BMSCs, when they were co-cultured with high neutrophil concentrations. Moreover, high neutrophil concentrations induced a decrease in BMSC cell counts and decreased ALP activity. Clarifying the underlying mechanism may contribute to development of therapies that augment bone regeneration or prevent impaired fracture healing after hyper-inflammatory conditions.

Keywords: alkaline phosphatase; bone marrow stromal cell; bone regeneration; fracture healing; multipotent stromal cell; neutrophils; stromal cells.

Figure 1

(A) Surface antigen expression of bone marrow stromal cells (BMSCs) isolated from the talus bone marrow using flow cytometry. >95% of cells were negative for CD45 and CD14, and >99% of cells were negative for CD19 and CD34. In addition, >95% were positive for CD73, CD90, CD105, and CD140b. Since plastic adherence is a well-established and validated technique to isolate multipotent stromal cells (MSCs), we have only characterized one BMSC donor using flowcytometry instead of all donors. The blue lines are stained cells and the gray lines are negative (unstained) controls. Adapted from Croes et al. (25). (B) Fluorescence-activated cell sorting (FACS) gating strategy used to isolate granulocytes/neutrophils from peripheral blood leukocytes. Granulocytes were either isolated from unlabeled leukocytes using gate 1 (G1) within the forward/sideward scatter (FSC/SSC). Alternatively, leukocytes were stained using CD3, CD14, CD193, and CD123. Within the FSC/SSC of these labeled cells, debris was first excluded [gate 2 (G2)]. Subsequently, CD3+ cells (lymphocytes) and CD14+ (monocytes) were excluded [gate 3 (G3)]. In addition, CD193+ cells (eosinophils) and CD123+ cells (basophils) were excluded [gate 4 (G4)]. The remaining CD3− CD14− CD193− CD123− cells were defined FACS-sorted neutrophils (G2+, G3+, G4+ sorted neutrophils). Re-analysis of FACS-sorted neutrophils shows adequate exclusion of lymphocytes and monocytes based on their FSC/SSC. (C) Images of BMSCs obtained by array scanning after 7 days of culture. BMSCs (2,000 BMSCs/well) were seeded and imaged after 7 days of culture in basic medium (BM), expansion medium (EM), osteogenic medium (OM), and after co-culture with neutrophils (16,000 neutrophils/well) in BM. Nuclei were stained with Hoechst (blue) and alkaline phosphatase (ALP) was stained (red) using Vector Red, which is a marker of osteogenic activity. All images within each experiment had similar exposure times and were not manipulated after capture with the array-scanner. (D) Quantification algorithm used to quantify cell count and the percentage of ALP positive cells in the adherent BMSC population after 7 days of culture. The blue rings within the algorithm images show identification of nuclei, the green rings are the area of interest around each nucleus in which Vector Red fluorescence was measured and each red pixel reflects Vector Red fluorescence above the threshold used to determine whether cells are ALP positive. The same protocol was used to quantify ALP+ cells in all experiments.

Figure 2

(A) The effect of neutrophils on bone marrow stromal cells (BMSCs) cell count in vitro (mean ± SEM/6 microscopy fields). Co-culture of BMSCs with different neutrophil concentrations resulted in decreased BMSC counts after 7 days of culture. Neutrophils were isolated from unlabeled leukocytes based on granulocyte-specific forward/sideward scatter (FSC/SSC) (Figure 1B) from three donors and cultured with three different BMSC donors [reamer/irrigator/aspirator (RIA) (n = 2) and fracture hematoma (FH) (n = 1)] in a 96-well plate ***p < 0.001, **p < 0.01 compared to BMSCs cultured in BM without neutrophils. BMSCs cultured without neutrophils in expansion medium (EM) are illustrated by the dark gray bar. (B) The effect of neutrophils on osteogenic activity of BMSCs in vitro (mean ± SEM/6 microscopy fields). Co-culture with different neutrophil concentrations induced a decreased percentage of alkaline phosphatase (ALP) positive cells after 7 days of culture. The same cells and number of donors were used as described in panel (A). The percentage of ALP+ FH and RIA-derived BMSC was 32 and 29%, respectively (cultured without neutrophils). FH- and RIA-derived BMSCs cultured without neutrophils were pooled (BM). All other conditions are depicted relative to BM. Therefore, the mean percentage of ALP+ of BM was set to 100%. BMSCs cultured without neutrophils in OM are illustrated by the dark gray bar.***p < 0.001 compared to BMSCs cultured in BM without neutrophils. (C) The effect of fluorescence-activated cell sorting (FACS) sorted neutrophils, peripheral blood mononuclear cells (PBMCs), and neutrophil transwell co-culture on osteogenic activity of BMSCs in vitro after 1 week of culture (mean ± SEM). FACS-sorted CD3− CD14− CD123− CD193− neutrophils (three donors, Figure 1B) were co-cultured with bone marrow-derived BMSCs (two donors) in a 24-well plate containing basic medium (BM), which induced a significant decrease in osteogenic activity (160N = 160,000 neutrophils/well). By contrast, Ficoll isolated PBMCs did not induce a significant decrease in ALP activity (160P = 160,000 PBMCs/well in BM). Moreover, transwell experiments in which neutrophils and BMSCs did not have cell–cell contact also did not significantly inhibit osteogenic activity [160N (TW) = 160,000 neutrophils/transwell insert in BM]. (D) The effect of FACS sorted neutrophils on extracellular matrix (ECM) in vitro after 4 weeks of culture (mean ± SEM). FACS sorted neutrophil co-culture with BMSCs in osteogenic medium (OM) induced a significant decrease in ECM mineralization after 4 weeks of culture as analyzed by Alizarin Red staining compared to BMSCs that were cultured in OM alone. *p < 0.05. (E) Representative images of Alizarin Red stained monolayers of BMSCs after 4 weeks of culture with and without neutrophils.

Figure 3

(A) Bone marrow stromal cell (BMSC) co-cultured with CFSE labeled neutrophils. BMSCs were co-cultured with CFSE-labeled neutrophils (green) and imaged using fluorescence microscopy at day 0 and after 1 and 2 days of culture with BMSCs to determine whether BMSCs phagocytosize neutrophils. BMSCs did not become CFSE positive after co-culture with CFSE labeled neutrophils. This finding suggests that phagocytosis of (apoptotic) neutrophils by BMSCs in vitro is not likely. (B) BMSCs LIVE/DEAD™ assay. Representative images of BMSCs stained with a LIVE/DEAD™ kit before and after culture with and without neutrophils. Viable BMSCs are green and non-viable BMSCs are red. Triton-X killed BMSCs were used as positive control. BMSCs (M+) were imaged on day 0, day 1, and day 2, cultured with (N+) or without (N−) neutrophils. (C) BMSCs viability and BMSC cell count (mean ± SEM). BMSC was stained with the LIVE/DEAD™ kit before and after 1 and 2 days of culture with and without neutrophils. Three neutrophil donors and two BMSC donors (two arthroplasty) were used (n = 6 conditions). The total BMSC count per microscopy image and the percentage of viable (green) BMSCs is depicted. BMSC counts increased during the first 2 days of culture without neutrophils and practically all cells remained viable (green). By contrast, the number of BMSCs decreased during the first 2 days after co-culture with neutrophils. The percentage of viable BMSCs did not significantly differ at day 1 and day 2 when cultures with and without neutrophils were compared. ***p < 0.001, **p < 0.01, *p < 0.05; ns, not significant. (D) Neutrophil viability assay. Neutrophils were stained with Annexin and 7-AAD before and after 1 and 2 days of culture with and without BMSCs. Three neutrophil donors and four BMSCs donors (one talus and three arthroplasty) were used (n = 12 conditions). The graphs depict the percentage of viable cells (Annexin and 7-AAD double negative cells), apoptotic cells (Annexin positive, 7-AAD negative), and permeable cells (Annexin and 7-AAD double positive cells). As has been described previously, we found that BMSCs induced a significant increase in the percentage of viable neutrophils at day 1 of culture. However, after 2 days of culture, practically all neutrophils were non-viable and there was no significant difference between neutrophils cultured with and without BMSCs at day 2. The median fluorescence of Annexin and 7-AAD was pooled and depicted as mean ± SEM. ***p < 0.001, **p < 0.01, *p < 0.05; ns, not significant.

Figure 4

Expression of surface markers on neutrophils. Multicolor flow cytometry was used to quantify expression of CD62L, CD64, CBRM1/5, CD16, CD14, CD35, CD11b, CD49d, and CD 66b on neutrophils before and after 24 h culture with and without bone marrow stromal cells (BMSCs) in BM. Neutrophils were isolated from three donors and cultured with BMSCs isolated from four donors (one talus and three arthroplasty) (n = 12 combinations). The black bars represent uncultured neutrophils and therefore without BMSCs [day 0 (M−)]. The gray bars represent neutrophils cultured for 24 h with [day 1 (M+), dark gray] or without [day 1 (M−), light gray] BMSCs in BM. The median fluorescence of each surface marker was pooled and depicted as mean ± SEM. ***p < 0.001, **p < 0.01, *p < 0.05; ns, not significant.