Proteomics-based discovery of a novel, structurally unique, and developmentally regulated plasminogen receptor, Plg-RKT, a major regulator of cell surface plasminogen activation

Abstract

Activation of plasminogen, the zymogen of the primary thrombolytic enzyme, plasmin, is markedly promoted when plasminogen is bound to cell surfaces, arming cells with the broad spectrum proteolytic activity of plasmin. In addition to its role in thrombolysis, cell surface plasmin facilitates a wide array of physiologic and pathologic processes. Carboxypeptidase B-sensitive plasminogen binding sites promote plasminogen activation on eukaryotic cells. However, no integral membrane plasminogen receptors exposing carboxyl terminal basic residues on cell surfaces have been identified. Here we use the exquisite sensitivity of multidimensional protein identification technology and an inducible progenitor cell line to identify a novel differentiation-induced integral membrane plasminogen receptor that exposes a C-terminal lysine on the cell surface, Plg-R(KT) (C9orf46 homolog). Plg-R(KT) was highly colocalized on the cell surface with the urokinase receptor, uPAR. Our data suggest that Plg-R(KT) also interacts directly with tissue plasminogen activator. Furthermore, Plg-R(KT) markedly promoted cell surface plasminogen activation. Database searching revealed that Plg-R(KT) mRNA is broadly expressed by migratory cell types, including leukocytes, and breast cancer, leukemic, and neuronal cells. This structurally unique plasminogen receptor represents a novel control point for regulating cell surface proteolysis.

Figure 1

Specific plasminogen binding is enhanced by M-CSF treatment of human peripheral blood monocytes and Hoxa9-ER4 cells. Human peripheral blood monocytes were either untreated (A) or treated with 0.44 nM of M-CSF for 8 days (B) and Hoxa9-ER4 cells were either untreated (C) or treated with 20% LADMAC conditioned media (a source of M-CSF) for 2 days (D). The cells were analyzed by dual-color fluorescence-activated cell sorting (FACS) analysis for specific plasminogen binding and CD antigen expression as described. Viable (propidium iodide negative and annexin V negative) cells were gated from the nonviable cells. Histogram plots of FITC-plasminogen (left columns) or specific anti-CD antibody binding to viable cells are shown. Gray tracings indicate either FITC-plasminogen or specific anti-CD antibody; black tracings, either FITC-plasminogen + 0.2 M EACA or isotype controls.

Figure 2

Equilibrium binding of plasminogen is observed to viable human peripheral blood monocytes and viable Hoxa9-ER4 cells. Human peripheral blood monocytes (A) were either untreated (○) or treated with M-CSF (●) as in Figure 1 and Hoxa9-ER4 cells (B) were treated with M-CSF as in Figure 1 and subjected to quantitative flow cytometry with increasing concentrations of FITC-plasminogen as described in under “Quantitative flow cytometry.”

Figure 3

High interspecies homology of Plg-R_KT. Alignment of predicted amino acid sequences of mouse, human, rat, dog, and cow orthologs of Plg-R_KT (A) and the structural model of Plg-R_KT (B). Green indicates amino acids within the predicted primary transmembrane helix; orange, amino acids within the predicted secondary transmembrane helix; and red, basic amino acids.

Figure 4

Plg-R_KT behaves as a regulated integral membrane protein. (A) Membrane fractions or cytoplasmic fractions from either undifferentiated or M-CSF–treated Hoxa9-ER4 cells (30 μg/lane) were electrophoresed on 12% sodium dodecyl sulfate polyacrylamide gels under reducing conditions and Western blotted with either anti–Plg-R_KT mAb, anti–α-enolase mAb as a loading control, or isotype control IgG. (B) M-CSF–treated Hoxa9-ER4 cell membranes were solubilized in 3% Triton X-114. After heating at 37°C and separation of the phases by centrifugation, an aliquot of both phases was electrophoresed and Western blotted with anti–Plg-R_KT mAb 35B10. In controls for the method, when the cell lysates were spiked with BSA and subjected to phase partitioning, BSA was detected in the aqueous, but not the detergent phase (data not shown).

Figure 5

Plg-R_KT is dispersed over the cell surface and colocalizes with uPAR. M-CSF–differentiated Hoxa9-ER4 cells were grown on coverslips and incubated with a combination of polyclonal rabbit anti–Plg-R_KT IgG (20 μg/mL) and mouse monoclonal anti-uPAR (20 μg/mL; A). Cells were washed, fixed in 1% formaldehyde, and then stained with a combination of Alexa 488-F(ab′)₂ fragment of goat anti–rabbit IgG and Alexa 568-F(ab′)₂ fragment of goat anti–mouse IgG for 60 minutes at 20°C in PBS containing 0.001% Triton X-100. Controls are samples incubated without first antibody. In panel B, cells were preincubated with either PBS (− plasminogen) or 2μM plasminogen (+ plasminogen) for 10 minutes at 4°C. Then, the cells were fixed in 1% formaldehyde, washed, and then stained with polyclonal anti-Plg IgG or mouse anti–Plg-R_KT mAb and stained with a combination of Alexa 488-F(ab′)₂ fragment of goat anti–rabbit IgG and Alexa 568-F(ab′)₂ fragment of goat anti–mouse IgG. Cells were washed and mounted in Immuno-Fluore Mounting Medium. Images were captured using a Zeiss laser confocal scanning microscope, then imported into LSM Examiner and ImageJ for further processing as described in “Scanning confocal microscopy.” The data in panel C were quantified and the number and size of each labeled aggregate were determined as described in “Scanning confocal microscopy.” The results reflect counts (C) and colocalization correlation coefficient (M1) values (last column in panels A-B) from more than 40 cells in 2 independent experiments. Data represent mean ± SEM. *P < .001.

Figure 6

Plasminogen binds to the C-terminal peptide of Plg-R_KT. The peptide, CEQSKLFSDK, corresponding to the amino terminus of Plg-R_KT was coupled to BSA and coated onto wells of microtiter plates. Either Glu-plasminogen (A) or Lys-plasminogen (B) or t-PA (C) was then incubated with the wells, followed by antiplasminogen mAb (A-B) or anti–t-PA mAb (C) and detection with HRP-conjugated goat anti–mouse IgG (●) as described in “Plasminogen and t-PA binding assay.” The binding was specific because it was blocked in the presence of 0.2M EACA (△), consistent with the ability of EACA to inhibit plasminogen binding to differentiated Hoxa9-ER4 cells. In additional controls, nonspecific binding to either BSA (▲), or to the reverse peptide (○) was < 10% of binding to CEQSKLFSDK. (At high input concentrations of t-PA, nonspecific binding increased, but was < 10% of binding to CEQSKLFSDK at the concentration required for 50% saturation [3.2nM]). In controls for the detection method, optical density at 490 nm (OD₄₉₀) values obtained using an isotype control antibody or in the absence of added plasminogen or t-PA were < 5% of the values for plasminogen or t-PA binding to immobilized CEQSKLFSDK. (D-E) Biotinylated Glu-plasminogen (25nM) was incubated with immobilized CEQSKLFSDK in the presence of increasing concentrations of (D) the C-terminal peptide, CEQSKLFSDK (●) or a mutated C-terminal peptide with K147 substituted with alanine, CEQSKLFSDA (○) or (E) anti–Plg-R_KT mAb 35B10 (●) or isotype control (○). Biotinylated Glu-plasminogen binding was detected with HRP-streptavidin and was 96% inhibited in the presence of 0.2M EACA (not shown). Data are mean ± SEM; n = 3, for each determination.

Figure 7

Plg-R_KT regulates cell surface plasminogen activation. Plasminogen activation was determined after 12 minutes as described in “Plasminogen activation assay” in either the presence or absence of either undifferentiated Hoxa9-ER4 progenitor cells or M-CSF–differentiated Hoxa9-ER4 cells and in the presence of either rat anti–Plg-R_KT mAb 35B10 (■) or isotype control rat IgG2a (□). ***P < .001, compared with the corresponding isotype control.