Role of the urokinase plasminogen activator receptor in mediating impaired efferocytosis of anti-SSA/Ro-bound apoptotic cardiocytes: Implications in the pathogenesis of congenital heart block

Abstract

Rationale: Binding of maternal anti-Ro/La antibodies to cognate antigen expressed on apoptotic cardiocytes decreases clearance by healthy cardiocytes, which may contribute to the development of autoimmune associated congenital heart block and fatal cardiomyopathy.

Objective: Given recent evidence implicating the urokinase plasminogen activator receptor (uPAR) as a "don't eat me" signal during efferocytosis, experiments addressed whether surface bound anti-Ro antibodies inhibit apoptotic cell removal via an effect on the expression/function of the urokinase-type plasminogen activator protease uPA/uPAR system.

Methods and results: As assessed by flow cytometry and confocal microscopy, uPAR colocalizes and interacts with Ro60 on the surface of apoptotic human fetal cardiocytes. Blocking of uPAR enhances phagocytosis of apoptotic cardiocytes by healthy cardiocytes and reverses the anti-Ro60-dependent impaired clearance of apoptotic cardiocytes. Binding of anti-Ro60 antibodies to apoptotic cardiocytes results in increased uPAR expression, as well as enhanced uPA activity. The binding of anti-Ro60 did not alter other surface molecules involved in cell recognition (calreticulin, CD31, or CD47).

Conclusions: These data suggest that increased uPAR expression and uPA activity induced by anti-Ro60 binding to the apoptotic fetal cardiocyte provide a molecular basis by which these antibodies inhibit efferocytosis and ultimately lead to scar of the fetal conduction system and working myocardium.

Figures

Figure 1

uPAR is expressed on human fetal cardiocytes (A) 20 µg of whole cell lysates of primary fetal cardiocytes and macrophages were subjected to SDS-page under reducing conditions, transferred to nitrocellulose and probed with murine anti-uPAR antibody (upper panel) or isotype control antibody (middle panel). Equal loading was confirmed by probing with anti-actin (bottom panel). (B) Flow cytometry of non-permeabilized healthy fetal cardiocytes demonstrates the presence of uPAR but not SSA/Ro or SSB/La on the cell surface. Healthy fetal cardiocytes were incubated with either uPAR specific antibody and the appropriate isotype control or CHB-IgG and nl-IgG respectively for 1 hr. The cells were then washed and incubated with FITC-conjugated secondary antibodies and flow cytometry was performed. A representative histogram of each condition is shown. X axis: logarithm of fluorescent intensity; Y axis: relative cell number. In the graph summarizing collective FACs data bars represent median of MFU±SEM and n=5. (C) Confocal microscopy of paraformaldehyde fixed cardiocytes demonstrates surface staining of uPAR by uPAR specific antibody (red) but not isotype control (D). Nuclear localization of Ro60 is detected in green by affinity purified anti-Ro60 specific antibody and the nucleus is stained with Hoerscht dye (blue). Results reflect representative cells from > 5 independent experiments. Scale bar represents 10 µm. (E) Immunofluoresence analysis shows fetal cardiocytes stain with the specific α-sarcomeric actinin marker (red).

Figure 2

uPAR and Ro60 are expressed on the surface of apoptotic cardiocytes (A) Apoptotic fetal cardiocytes were incubated with murine uPAR specific antibody and the appropriate isotype control or CHB-IgG and nl-IgG respectively for 1 hr. The cells were then washed and incubated with FITC-conjugated secondary antibodies and flow cytometry was performed. A representative histogram of each condition is shown. X axis: logarithm of fluorescent intensity; Y axis: relative cell number. In the graph summarizing collective FACs data bars represent median of MFU±SEM (n=5). (B) Confocal microscopy demonstrates colocalization of uPAR (red) with Ro60 (green) on the surface of non-permeabilized apoptotic cardiocytes in discrete membrane areas resembling apoptotic blebs. Cells were double labeled with mouse monoclonal anti-uPAR and affinity purified anti-Ro60 antibodies followed by Alexa Fluor conjugated secondary antibodies. All fluorescence images taken were the maximum projection composites of at least 8 optical sections. The main image shown is one raw image stack. Orthogonal view is used see the inside of the main image stack (X-Y). The two lines that cross in the X-Y image are two cuts, revealing the side (X-Z) and lower (Y-Z) view of the stack. Arrowheads point to the areas of colocalization. Cells are representative of data derived from 5 independent experiments. Scale bars represent 10 µm. (C) Triton-X soluble whole cardiocyte lysates were immunoprecipitated with mouse species isotype control or mouse anti-uPAR antibody and immunoprecipitated proteins detected with affinity purified anti-Ro60 antibody. Antibody reactivity at the expected (~ 60kD) region was observed (top panel) whereas no reactivity was detected with a control nl-IgG (middle panel). Confirmation of immunoprecipitated uPAR is shown in the bottom panel. (D) In a separate experiment anti-uPAR immunoprecipitant was not reactive with affinity purified antibodies against Ro52 (top) or La48 (bottom).

Figure 3

Blocking uPAR reverses the anti-Ro60 inhibition of efferocytosis (A) Apoptotic cardiocytes were treated with rabbit isotype IgG and rabbit anti-uPAR antibody or (B) mouse isotype control, mouse anti-uPAR antibody or mouse anti-HLA antibody. Cells were subsequently washed, co-cultured with healthy cardiocytes for 5 hr, washed, fixed and stained using TUNEL-peroxidase. Representative images of each condition are shown and numerical data summarized in the graphs presented below the images (n=4 for A and n=4 for B). The Y axis represents the phagocytic index, which is expressed as the percentage of cardiocytes that contain at least one ingested apoptotic cardiocyte. (C) Apoptotic cardiocytes were treated with nl-IgG, CHB-IgG or mouse anti-uPAR antibody after preincubation with CHB-IgG. Cells were subsequently washed and co-cultured with healthy cardiocytes for 5 hrs, washed, fixed, and stained using TUNEL-peroxidase as in (A). Representative images of cells treated with nl-IgG, CHB-IgG or anti-uPAR antibody after preincubation with CHB-IgG are shown. Below is a graph summarizing the data (n=6). (D) Apoptotic cardiocytes were treated with chicken monoclonal ScFv1, ScFv2, or anti-uPAR antibody after preincubation with ScFv1 or ScFv2. Cells were subsequently washed and cocultured with healthy cardiocytes for 5 hrs, extensively washed, fixed and stained using TUNEL-peroxidase as in (A). A representative image of ScFv1 (left) or ScFv1 subsequently treated with uPAR antibody and a graph summarizing the data are shown (n=4). (E) Apoptotic cardiocytes were treated with ScFv1 or mouse anti-uPAR or anti-HLA antibody after preincubation with ScFv1. Cells were subsequently washed and cocultured with healthy cardiocytes for 5 hr, washed, fixed and stained using TUNEL-peroxidase as in (A). A representative image of ScFv1, ScFv1 subsequently treated with uPAR antibody, or ScFv1 subsequently treated with HLA antibody is shown. A summary is presented in the graph (n=4).

Figure 3

Blocking uPAR reverses the anti-Ro60 inhibition of efferocytosis (A) Apoptotic cardiocytes were treated with rabbit isotype IgG and rabbit anti-uPAR antibody or (B) mouse isotype control, mouse anti-uPAR antibody or mouse anti-HLA antibody. Cells were subsequently washed, co-cultured with healthy cardiocytes for 5 hr, washed, fixed and stained using TUNEL-peroxidase. Representative images of each condition are shown and numerical data summarized in the graphs presented below the images (n=4 for A and n=4 for B). The Y axis represents the phagocytic index, which is expressed as the percentage of cardiocytes that contain at least one ingested apoptotic cardiocyte. (C) Apoptotic cardiocytes were treated with nl-IgG, CHB-IgG or mouse anti-uPAR antibody after preincubation with CHB-IgG. Cells were subsequently washed and co-cultured with healthy cardiocytes for 5 hrs, washed, fixed, and stained using TUNEL-peroxidase as in (A). Representative images of cells treated with nl-IgG, CHB-IgG or anti-uPAR antibody after preincubation with CHB-IgG are shown. Below is a graph summarizing the data (n=6). (D) Apoptotic cardiocytes were treated with chicken monoclonal ScFv1, ScFv2, or anti-uPAR antibody after preincubation with ScFv1 or ScFv2. Cells were subsequently washed and cocultured with healthy cardiocytes for 5 hrs, extensively washed, fixed and stained using TUNEL-peroxidase as in (A). A representative image of ScFv1 (left) or ScFv1 subsequently treated with uPAR antibody and a graph summarizing the data are shown (n=4). (E) Apoptotic cardiocytes were treated with ScFv1 or mouse anti-uPAR or anti-HLA antibody after preincubation with ScFv1. Cells were subsequently washed and cocultured with healthy cardiocytes for 5 hr, washed, fixed and stained using TUNEL-peroxidase as in (A). A representative image of ScFv1, ScFv1 subsequently treated with uPAR antibody, or ScFv1 subsequently treated with HLA antibody is shown. A summary is presented in the graph (n=4).

Figure 3

Blocking uPAR reverses the anti-Ro60 inhibition of efferocytosis (A) Apoptotic cardiocytes were treated with rabbit isotype IgG and rabbit anti-uPAR antibody or (B) mouse isotype control, mouse anti-uPAR antibody or mouse anti-HLA antibody. Cells were subsequently washed, co-cultured with healthy cardiocytes for 5 hr, washed, fixed and stained using TUNEL-peroxidase. Representative images of each condition are shown and numerical data summarized in the graphs presented below the images (n=4 for A and n=4 for B). The Y axis represents the phagocytic index, which is expressed as the percentage of cardiocytes that contain at least one ingested apoptotic cardiocyte. (C) Apoptotic cardiocytes were treated with nl-IgG, CHB-IgG or mouse anti-uPAR antibody after preincubation with CHB-IgG. Cells were subsequently washed and co-cultured with healthy cardiocytes for 5 hrs, washed, fixed, and stained using TUNEL-peroxidase as in (A). Representative images of cells treated with nl-IgG, CHB-IgG or anti-uPAR antibody after preincubation with CHB-IgG are shown. Below is a graph summarizing the data (n=6). (D) Apoptotic cardiocytes were treated with chicken monoclonal ScFv1, ScFv2, or anti-uPAR antibody after preincubation with ScFv1 or ScFv2. Cells were subsequently washed and cocultured with healthy cardiocytes for 5 hrs, extensively washed, fixed and stained using TUNEL-peroxidase as in (A). A representative image of ScFv1 (left) or ScFv1 subsequently treated with uPAR antibody and a graph summarizing the data are shown (n=4). (E) Apoptotic cardiocytes were treated with ScFv1 or mouse anti-uPAR or anti-HLA antibody after preincubation with ScFv1. Cells were subsequently washed and cocultured with healthy cardiocytes for 5 hr, washed, fixed and stained using TUNEL-peroxidase as in (A). A representative image of ScFv1, ScFv1 subsequently treated with uPAR antibody, or ScFv1 subsequently treated with HLA antibody is shown. A summary is presented in the graph (n=4).

Figure 4

Anti-Ro60 binding to apoptotic cardiocytes increases uPAR expression and induces the formation of a high molecular weight conjugate (A) Apoptotic fetal cardiocytes were preincubated with nl-IgG, 3 different CHB-IgG, RNP-IgG or anti-HLA. After washing with PBS 5% goat serum, cells were stained with rabbit anti-uPAR antibodies and analyzed by FACs. A representative histogram of each condition is shown. X axis: logarithm of fluorescent intensity; Y axis: relative cell number. Light grey trace represents uPAR expression after preincubation with the appropriate antibody and dark trace is uPAR antibody alone. In the graph summarizing collective FACs data, bars represent median of MFU±SEM and n=5. (B) Apoptotic fetal cardiocytes were preincubated with ScFv1, ScFv2, ScFv48 or ScFv52. After washing with PBS 5% goat serum, cells were stained with mouse anti-uPAR and analyzed by FACs. A graph summarizing the data as in (A) in which the Y axis represents uPAR expression (n=5) and a representative histogram for each condition are shown. (C) Apoptotic cardiocytes were incubated with the crosslinker DTSSP. Following quenching with TBS, lysates were analyzed by SDS-page and uPAR was immunodetected with mouse uPAR antibody. In the presence of the crosslinker a high molecular weight band was observed at 100kD. Data are representative of three independent experiments. (D) Apoptotic cardiocytes were incubated with nl-IgG or CHB-IgG for 30 min at 18°C. Following washing, cells were lysed, analyzed via SDS-Page and uPAR was immunodetected. A 100kD band indicative of endogenous crosslinking was observed in apoptotic cardiocytes opsonized with CHB-IgG. Data are representative of three independent experiments. (E) Apoptotic cardiocytes were incubated with ScFv1 or ScFv2 for 30 min at 18°C. Following washing, cells were lysed, analyzed via SDS-Page and uPAR was immunodetected. A 100kD band indicative of endogenous crosslinking was observed in apoptotic cardiocytes opsonized with ScFv1. Data are representative of three independent experiments.

Figure 5

Opsonization of cardiocytes with CHB-IgG does not modulate CRT, CD31 or CD47 expression (A) Apoptotic fetal cardiocytes were preincubated with nl-IgG or CHB-IgG. After washing with PBS/ 5% goat serum, cells were stained with antibodies against CD31 (top), CRT (middle), CD47 (bottom panel) and analyzed by FACs. A representative histogram of each condition is shown. X axis: logarithm of fluorescent intensity; Y axis: relative cell number. Light grey trace represents uPAR expression after preincubation with the appropriate antibody and dark trace is uPAR antibody alone. In the graph summarizing collective FACs data bars represent median of MFU±SEM and n=3. (B) Apoptotic cardiocytes were incubated with nl-IgG or CHB-IgG for 30 min at RT. Following washing, cells were lysed, analyzed via SDS-Page and CD31, CRT, CD47 was immunodetected via the respective antibodies.

Figure 6

uPA is present on the surface of apoptotic cardiocytes and promotes plasminogen activity following binding of anti-Ro60 antibodies. (A) Non-permeabilized apoptotic cardiocytes were treated with anti-uPA antibody for an hour and analyzed by flow cytometry. A representative histogram showing uPA surface expression is presented Light grey trace represents anti-uPA whereas dark trace is isotype control. (B) Non-permeabilized apoptotic cardiocytes were treated with nl-IgG, CHB-IgG, ScFv1, or ScFv2 for 30 min, plated in triplicate in a 96-well plate and plasminogen activation determined using a chromogenic assay as described in Materials and Methods. The Y axis represents plasminogen activation (uPA enzymatic activity units) (n= 5). (C) Non-permeabilized apoptotic cardiocytes were treated with CHB-IgG followed by anti-uPA, anti-uPAR or preincubated with β2GPI (5 µg/ml) and subsequently treated with CHB-IgG. Plasminogen activation was measured as in (B). The Y axis represents the % inhibition of plasminogen activation in each condition compared to plasminogen activation by CHB-IgG (n=3). Also shown is a representative histogram of non-permeabilized apoptotic cardiocytes incubated with nl-IgG or CHB-IgG in the presence or absence of β2GPI (5 µg/ml) for an hour. Light trace represents CHB-IgG and dark nl-IgG. β2GPI inhibits the binding of CHB-IgG as previously described .

Figure 7

This cartoon illustrates our current interpretation. During physiologic apoptosis in cardiac development, Ro60 translocates to the cell surface. There is a balance between anti-Ro60 and β2GPI binding to Ro60. The binding of anti-Ro60 increases uPAR expression. This blocks efferocytosis by healthy cardiocytes and also results in stimulation of uPA enzymatic activity, which generates plasmin. In turn plasmin cleavage of β2GPI favors increased anti-Ro60 binding. Thus, an amplification cycle is generated which results in accumulation of apoptotic cardiocytes as demonstrated in histologic evaluation of CHB hearts.