Detection of the Receptor for the Human Urokinase-type Plasminogen Activator Using Fluoresceinated uPA

Abstract

The urokinase-type plasminogen activator (uPA) is a serine protease that plays a crucial role in blood coagulation and in tumor invasion and metastasis. uPA is a relatively large polypeptide and binds the uPA receptor (uPAR) with high affinity and specificity. Therefore, it was a good candidate for direct labeling with a fluorochrome for detection of the uPAR. We have produced a fluorescein (FITC)-labeled human uPA using a conjugation procedure that did not significantly alter its binding characteristics to the uPAR. Thirty nM FITC–uPA efficiently stains 2 × 10⁵ uPAR-transfected mouse cells in suspension, as determined by flow cytometric analysis. One μg of FITC–uPA efficiently stains 2 × 10⁵ uPAR transfectants grown on slides and analyzed by fluorescence optical microscopy. Human cell lines expressing the endogenous uPAR were stained with similar efficiency. Fixation in paraformaldehyde only slightly reduced the efficiency of staining of both transfectants and cell lines. These characteristics allow the use of FITC–uPA in both static and dynamic morphological studies of uPAR-expressing cells.

Keywords

urokinase-type plasminogen activator fluorescein labeling cell surface receptors fluorescence microscopy flow cytometry

The serine protease urokinase-type plasminogen activator (uPA) is synthesized as a soluble single-chain inactive precursor (pro-uPA) and is activated by pro-teolysis after binding to the uPA receptor (uPAR). Active uPA specifically proteolizes plasminogen into plasmin, which in turn degrades extracellular matrix proteins (fibronectin, laminin, and proteoglycans) or activates matrix metalloproteases and the pro-uPA itself (Mignatti and Rifkin 1993). uPA and the uPAR play a crucial role in blood coagulation, participating in the plasma procoagulant and fibrinolytic systems (Simon et al. 1996). Moreover, they influence the procoagulant and/or fibrinolytic activity of circulating cells (Simon et al. 1996). uPA can also enhance tumor invasive and metastatic properties in vivo. Tumor cells frequently secrete uPA (Danø et al. 1985; Hearing et al. 1989; Ossowski et al. 1991; Pyke et al. 1991; Mignatti and Rifkin 1993) and highly invasive human carcinomas frequently show high levels of uPA (Duffy et al. 1988). The murine melanoma B16-F10 cell line with low metastatic potential can acquire a much greater capacity to form lung colonies in mice when transfected with human uPA (huPA) (Schultz et al. 1992). Conversely, the metastatization of different tumor cell lines can be strongly inhibited by anti-uPA antibodies (Mignatti and Rifkin 1993; Ossowski et al. 1991; Ossowski and Reich 1983). A main mechanism of action of the uPA–uPAR complex appears to be the degradation of the extracellular matrix (ECM), which permits malignant cells to migrate away from their original site of growth. Indeed, mouse L-fibroblasts transfected with the uPAR (Mignatti and Rifkin 1993) and grown on a reconstituted ECM show proteolytic properties in the presence of uPA and migrate through the ECM. Similarly, when mouse L-cells transfected with the human uPA are co-cultivated with L-cells transfected with the huPAR on chorioallantoic membrane (CAM) or on ECM, functional uPA–uPAR complexes are reconstituted and both types of transfectants invade the substrate (Quax et al. 1991; Mignatti and Rifkin 1993). Consistent with this mechanism of action, antibodies and inhibitors of uPA activity inhibit both ECM degradation and invasion of either the amniotic membrane or of reconstituted basement membrane (Mignatti and Rifkin 1993).

Because uPA, together with other proteases, can play a major role in blood coagulation and tumor invasion, it would be of interest to be able to efficiently follow the uPA–uPAR complex, e. g., during its formation, internalization or shedding, and interaction with its substrate. Of further use would be detection of the uPA–uPAR complex on living cells and following it dynamically over time. To reach this goal, we decided to produce a directly labeled uPA that could be used with high efficiency in fluorescence microscopy and flow cytometry. The huPA has a K_d of 2–20 × 10^-9 M for the uPAR (Roldan et al. 1990), i. e., it binds uPAR with a similar if not higher affinity than mono- or polyclonal antibodies (Nisonoff et al. 1975). The high K_d indicated the feasibility of direct detection of the uPAR using its ligand, as we could expect a similar if not higher efficiency of detection than with antibodies. Moreover, uPA was expected to be more specific than antibodies, because it detects the uPAR binding site with high affinity and species specificity, whereas antibodies commonly recognize a small cluster of residues that might appear by chance in totally unrelated molecules (Bendayan 1995).

The feasibility of direct labeling of uPA was also suggested by the efficient and nondestructive labeling of other proteic ligands, e. g., IL-2 (De Jong et al. 1995) and other interleukins, G-CSF, GM-CSF, TNFα, TGFβ (British Biotechnology; Oxford, UK), growth hormone (Bentham et al. 1994), insulin-like growth factor-I (Bentham et al. 1993), EGF (Azevedeo and Johnson 1990), and transferrin (Uriel et al. 1990). Most proteins are either fluorescein (FITC)- or biotin-conjugated (Hardy 1986; Azevedeo and Johnson 1990; Uriel et al. 1990; Bentham et al. 1993, 1994; De Jong et al. 1995). Biotin conjugation has the theoretical advantage of allowing a flexible use of streptavidin conjugated to different fluorochromes (Parks et al. 1986). However, FITC offers the great advantage of direct labeling and visualization of the target molecules with a single incubation step, i. e., more conveniently and with better control of the binding conditions. Therefore, we tested the feasibility of FITC conjugation of uPA. Here we show that uPA can be efficiently labeled with FITC and binds the uPA receptor with high specificity and essentially the same affinity as unlabeled uPA. We prove also that FITC–uPA can efficiently label the uPA receptor on both living and fixed cells and can be used in both flow cytometry and fluorescence microscopy.

Materials and Methods

Tissue Culture

Murine LB6 cells co-transfected with the phuPAR (containing an expressable human uPAR cDNA) and the pRSVneo selectable vector or with the pRSVneo vector alone (LC) were obtained from Dr. F. Blasi through the courtesy of Dr. P. Mignatti (Roldan et al. 1990). HT-1080 and U937 cells were obtained from Drs. A. De Blasi and A. Poggi. The cells were cultured in DMEM or RPMI (Gibco BRL; Paisley, UK) supplemented with 1 g/liter glucose, sodium pyruvate, glutamine, penicillin, streptomycin, 10% fetal calf serum (Hy-Clone Laboratories; Logan, UT) (culture medium). Culture medium for transfected cells included 200 μg/ml of geneticin (Sigma Chemical; St Louis, MO) as selecting agent. All cells were maintained in a 5% CO₂ incubator at 37C.

FITC Conjugation of uPA

Purified mature, i. e., double-chain human uPA (Ukidan; Serono, Rome, Italy) was kindly supplied through the courtesy of Dr. B. Donati. uPA 400 μg in 0.5 ml PBS were salt-exchanged to carbonate buffer (1.7 g Na₂CO₃ and 2.8 g NaHCO₃ in 100 ml H₂O), pH 9.4, using a PD-10 column (Pharmacia; Uppsala, Sweden). FITC freshly dissolved in DMSO was added to uPA in bicarbonate buffer at 27 μg/ml per mg of protein. The FITC–uPA was incubated in the dark under continuous rocking for 3 hr at room temperature (RT). Free FITC was removed from the FITC-uPA containing solution by salt-exchanging on a PD-10 column (Pharmacia). The absorbance of FITC–uPA was measured at 280 nm and 495 nm to determine the fluorochrome to protein (F/P) ratio, using the following formula (Hardy 1986):

F = OD_495nm/68,000 (FITC molar extinction coefficient

P=[OD_280nm-31% OD_495nm]/0.96 (uPA molar extinction coefficient) × 51,800 (uPA mw)

The molar extinction coefficient of uPA was determined from the absorbance of known amounts of purified uPA in PBS at 280 nm.

FITC–uPA Binding Assay in Flow Cytometry

Cells were collected with PBS 0.6 μM EDTA, centrifuged for 5 min at 1200 RPM at 4C, and resuspended in staining medium (SM). SM is 50% 1 × PBS, 50% 1 × HBSS, 20 mM HEPES, pH 8.1, supplemented with essential and nonessential amino acids, sodium pyruvate (Gibco BRL), 3% fetal calf serum (HyClone), and 0.1% NaN₃. Preliminary assays allowed to determine conditions of staining that resulted in high signal on uPAR-expressing cells and no background on control cells (Dell'Arciprete et al. 1996) (see Results). Efficient staining was achieved when 2 × 10⁵ cells were incubated for 1 hr on ice with 0.3 μg FITC–uPA in 200 μl of SM, i. e., in 28.9 nM uPA final concentration. After three washes with SM the cells were resuspended in 200 μl of SM with 0.5 μg/ml propidium iodide (SM-PI) and analyzed by flow cytometry. For assays on fixed cells, collected cells were resuspended in 1% paraformaldehyde (PFA) (Sigma Chemical) in PBS and incubated for 10 min on ice. Residual PFA was quenched with 50 mM NH₄Cl in PBS for 10 min on ice. After washing, the fixed cells were resuspended in SM without PI.

uPAR Competition Assay in Flow Cytometry

The binding specificity and affinity of FITC–uPA were assayed in a competition assay using unlabeled huPA. Briefly, 0.3 μg of uPA conjugated to FITC was mixed with increasing amounts of unlabeled uPA, i. e., 0.3 μg, 2.7 μg or 29.7 μg. Each premixed solution was added to 2 × 10⁵ living or fixed cells in 200 μl of SM and incubated for 1 hr on ice. After three washes with SM they were resuspended in 200 μl of SM-PI or SM, respectively, and analyzed by flow cytometry. The CV and net fluorescence values of the profiles obtained were compared with the expected values assuming unchanged affinity of uPA after FITC conjugation (see below).

Fluorescence Analysis and Cell Sorting

Fluorescence analysis was performed on a FACSTAR (Becton Dickinson; Sunnyvale, CA), used essentially as described (Parks et al. 1986). Calibration of the sorter was performed before each experiment using fluorescent beads. Linearity of the response of the flow cytometer log. amplifiers over four decades was routinely checked, to ensure reproducible quantitation of the results obtained. The figures present log. fluorescence profiles of 5000 cells. Subtraction of cell auto-fluorescence (Alberti et al. 1987) and fluorescence overcompensation in the red channel were performed as described (Alberti et al. 1991). Forward scatter, side scatter, and propidium iodide gating were routinely used to eliminate dead cells and debris from the analysis. We found it useful to compare different experimental groups by their net fluorescence (Dell'Arciprete et al. 1996). We defined the net fluorescence of transfected cells, expressed in arbitrary units (AU), as

arithmetic mean fluorescence of the stained uPAR transfected cells — arithmetic mean fluorescence of the stained pRSVneo transfected control cells

The net fluorescence is an indicator of the total amount of uPA bound to the cells. Importantly, it does not depend on a specific procedure for estimating positive cells and includes dimly expressing cells. The net fluorescence values presented in Table 1 were calculated on 5000 cells.

Table 1

Competition assay between FITC-uPA and unlabeled uPA^a

	No unlabeled uPA	1/1^b	1/9	1/99
% reduction of staining (living cells)	0	45.5^c	(50)^d	92.8 (90)	95.5 (99)
% reduction of staining (fixed cells)	0	48 (50)	91 (90)	ND^e

Flow cytometric analysis of L/uPAR cells. Transfectants were stained with 0.3 μg of FITC–uPA premixed with the indicated amounts of unconjugated uPA.

FITC–uPA/unlabeled uPA ratio.

100 – (net fluorescence of cells stained with FITC and unlabeled uPA/net fluorescence of cells stained with FITC–uPA alone). Values are expressed as percentages.

Expected values of the residual net fluorescence, expressed as percentages, assuming equal affinity of FITC–uPA and unlabeled uPA for the uPAR. Thus, an equal amount of unlabeled uPA was expected to lead to a 50% reduction of staining, ninefold higher unlabeled uPA should have lead to a 90% reduction, and 99-fold higher unlabeled uPA should have reduced the staining by 99%. No significant differences of observed vs expected means, CV, and distribution shapes were observed (data not shown).

Not determined.

Fluorescence Microscopy

Two hundred thousand cells were plated on 20 × 20-mm glass coverslips in 6-well tissue culture plates. The next day the coverslips were washed twice in SM and the cells were incubated for 1 hr on ice with 1 μg of FITC–uPA in 50 μl SM. In competition assays, cells were stained with 1 μg of FITC–uPA together with either 1 μg, 9 μg, or 99 μg of unlabeled uPA. Stained coverslips were washed in SM and fixed in the dark for 10 min at RT in 4% paraformaldehyde (PFA) (Sigma Chemical) in PBS. Unreacted PFA was quenched with 50 mM NH₄Cl for 10 min at RT. After two washes with SM the coverslips were mounted in Mowiol (Calbiochem-Novabiochem; La Jolla, CA) or in 90% glycerol–10% PBS, and were observed with a Zeiss Axiophot fluorescence microscope.

Results

To reduce the chances of inactivating the binding region of uPA for uPAR, uPA was conjugated with FITC using a mild procedure adapted from a FITC conjugation protocol for antibody molecules (Hardy 1986). Preliminary tests defined the optimal concentration of FITC and ratio with uPA to obtain a one-to-one ratio of FITC to uPA (F/P). The largest batch of FITC–uPA that was used throughout this work was labeled with 27 μg FITC/ml/mg of protein and possessed an F/P ratio of 0.82.

Figure 1

Flow cytometric analysis of FITC–uPA-stained cells. (A) Cells stained using a freshly thawed aliquot of FITC–uPA. (B) Cells stained using a repeatedly frozen and thawed aliquot of FITC–uPA. L/uPAR cells in B were maintained in culture for over 1 year longer than in A. Solid line; L/uPAR-expressing cells; dotted line; LC control cells.

The binding affinity and specificity of FITC–uPA towards the uPAR was assessed on mouse L-cells transfected with the human uPAR. Quantitation of binding was performed by flow cytometry. The K_d of uPA for uPAR transfectant LB6 cells is about 2–20 × 10^-9 M (Appella et al. 1987). Therefore, nanomolar concentrations of uPA were predicted to be adequate for transfectant staining. Titration assays were performed by flow cytometry. We used amounts of FITC-uPA corresponding to 2.89 nM, 28.9 nM, and 100 nM for 1 hr on ice. Twenty-nine nM uPA for 1 hr on ice stained 2 × 10⁵ uPAR- transfected LB6 cells to levels essentially identical to 100 nM, i. e., to levels close to saturation. No staining was detectable on LC cells transfected with the pRSVneo vector alone or on untransfected cells incubated with FITC–uPA up to 100 nM (Figure 1A; and data not shown). The essential lack of staining of mouse cells that do express the mouse uPAR (Møller 1993) confirmed the species specificity of uPA binding to uPAR (Møller 1993). The relative binding affinity of FITC–uPA was assessed in a competition assay with unlabeled uPA. uPAR-transfected or control LC living cells were stained with different ratios of FITC–uPA/unlabeled uPA, i.e., adding no unlabeled uPA or an equal amount or 9–99 fold higher amounts (Table 1). If the affinity of uPA remained constant after labeling with FITC, the reduction in the net cell fluorescence was expected to be proportional to the amount of unlabeled uPA added. Indeed, an equal amount of unlabeled uPA reduced FITC–uPA binding to living cells by 45% vs the expected 50%. Analogously, a 9-fold or a 99-fold higher unlabeled uPA reduced FITC–uPA binding by 93% or 96%, respectively, which is remarkably close to the predicted values of 90% and 99%, respectively (Table 1). Similar results were obtained on fixed cells. The profiles obtained and the expected profiles essentially overlapped, i.e., showed log-normal distribution with essentially constant coefficient of variation, together with the expected shift of the mean values (data not shown).

Figure 2

Flow cytometric analysis of FITC–uPA stained cells. (A, C) U937 cells. (B, D) L/uPAR cells. Cells in A and B were fixed in PFA. Cells in C and D are living cells. Solid line; FITC–uPA-stained cells; dotted line; unstained cells.

We also tested the efficiency of staining of U937 and HT-1080, i. e., human cell lines normally expressing uPAR (Møller 1993). U937 showed levels of surface staining comparable to the uPAR-transfected LB6 cells (Figure 2C), and competition of FITC–uPA with unlabeled uPA confirmed the specificity of the binding observed. HT-1080 cells were also specifically stained with FITC–uPA, albeit at lower levels than U937 (data not shown).

FITC-uPA efficiently stained both uPAR-transfected LB6 and U937 and HT-1080 cells fixed in 1% PFA (Figures 2A and 2B). Competition of the FITC–uPA staining with unlabeled uPA confirmed the specificity of these findings (Table 1).

FITC-uPA can efficiently detect uPAR-expressing cells in fluorescence microscopy (Figure 3B). FITC–uPA staining is mostly distributed on the plasma membrane with a ring-like pattern. However, brighter spots in button-like structures are also evident, suggesting accumulation of FITC–uPA on potential cell–substrate adhesion areas (Estreicher et al. 1990; Møller 1993). Control cells incubated with equal amounts of FITC–uPA were not detectably stained (Figure 3A), confirming the specificity of the observed staining. uPA–uPAR complexes can be internalized under physiological conditions with identical kinetics in transfected murine LB6 cells and human U937 cells (Møller 1993). However, staining of cells with FITC–uPA on ice in the presence of 0.1% NaN₃ efficiently prevents internalization of the uPA–uPAR complex (unpublished observation).

FITC–uPA was stored at −80C with no adverse effects for more than 2 years. Repeated freezing and thawing of the same aliquot of FITC–uPA did not affect binding or staining efficiency, and we did not observe reduction of the apparent titer of FITC–uPA over time due to autoproteolysis. An example is shown in Figures 1A vs 1B. The essentially overlapping profiles obtained on different batches of stably transfected cells, either freshly thawed or continuously kept in culture, confirm the reproducibility of the staining procedure presented and the stability of the FITC–uPA.

Discussion

uPA and the uPAR play a crucial role in the coagulation cascade (Simon et al. 1996) and in the process of tumor invasion and metastasis (Mignatti and Rifkin 1993). Therefore, we wanted to develop a labeled uPA that could detect uPAR not only in conventional fluorescence microscopy but also in real time microscopy on living cells and in flow cytometry. uPA is a relatively large proteic ligand and binds to its cell surface receptor with high affinity. Therefore, it appeared to be a good candidate for direct conjugation to a chromophore. Other receptor ligands have been labeled with either FITC, biotin, or phycoerythrin (Hardy 1986; Azevedeo and Johnson 1990; Uriel et al. 1990; Bentham et al. 1993,1994; De Jong et al. 1995). FITC had the theoretical advantage of allowing single-step labeling of uPAR and has a potentially lower steric hindrance than phycoerythrin or the biotin–avidin complex. Therefore, we tested the possibility of directly labeling huPA with FITC.

Figure 3

Fluorescence microscopic analysis of cells stained with FITC–uPA as described. (A) LC control cells; (B) L/uPAR-transfected cells. Bar = 2 μm.

The main uPAR binding site of human uPA is between residues 12–32 (Appella et al. 1987). FITC binds the ∊-amino group of arginine and lysine residues (Hardy 1986), and no arginines and only one lysine (Lys23) are present in this region. Therefore, it was possible to label uPA with a sufficiently low number of FITC molecules per uPA to essentially eliminate the chances of inactivating the binding region of uPA for uPAR. uPA was conjugated with FITC using a mild procedure adapted from a FITC conjugation protocol for antibody molecules (Hardy 1986) that allowed us to produce FITC-uPA with an F/P close to 1.

In spite of the apparent simplicity of the production and use of labeled receptor ligands and of the theoretical advantages of their use, only relatively few of these have been produced to date (Hardy 1986; Azevedeo and Johnson 1990; Uriel et al. 1990; Bentham et al. 1993,1994; De Jong et al. 1995). This probably reflects adverse effects of the labeling procedure on the ligand structure or binding efficiency. Therefore, it was crucial to prove that FITC-conjugated uPA maintained high affinity and specificity towards the uPAR. As a test system we used mouse L-cells transfected with the human uPAR. Transfectants were chosen to avoid as rigorously as possible potential staining artifacts, because the uPAR transfected vs control cells only differ in the presence of the human uPAR cDNA. This enabled us to exclude binding to other potential uPA binding sites on the cells surface, e. g., PAI-1 or plasminogen bound to their receptors (Møller 1993), simply by proving lack of FITC-uPA staining on the control cells. The use of mouse cells also enabled us to confirm the species specificity of the binding of FITC-uPA to uPAR (Møller 1993). The binding of FITC-uPA to its receptor was quantitated by flow cytometry and was confirmed by optical microscopy. The K_d of uPA for uPAR-transfectant LB6 cells is about 2–20 × 10^-9 M (Appella et al. 1987). Therefore, nanomolar concentrations of uPA were predicted to be adequate for transfectant staining. Titration assays were performed by flow cytometry. In our hands, 28.9 nM uPA for 1 hr on ice efficiently stained 2 × 10⁵ uPAR-transfected LB6 cells. No staining was detectable on control cells. uPAR- transfected LB6 cells express about 2.5 × 10⁴ receptors per cell (Roldan et al. 1990). For comparison, tumor cells commonly show number of receptors per cell up to 3 × 10⁵ (Roldan et al. 1990; Mignatti and Rifkin 1993). Therefore, an efficient detection of uPAR with no LB6 transfectants indicates good sensitivity of the assay. To quantitate possible adverse effects of FITC conjugation on the uPA binding affinity for uPAR, we determined the relative binding affinity of FITC–uPA in a competition assay with unlabeled uPA and proved that it is essentially unchanged compared with the native molecule.

We extended these results to human cell lines expressing the endogenous uPAR. Efficient and specific staining was obtained on the human monocytic U937 cells and on the HT-1080 ovarian carcinoma cells, albeit at lower levels, proving that FITC–uPA can detect the native human uPAR (data not shown). We also demonstrated efficient staining of PFA-fixed cells, indicating that FITC–uPA can be of potential use on histopathological samples.

FITC–uPA can efficiently detect uPAR-expressing cells in fluorescence microscopy. FITC–uPA staining is mostly distributed on the cell surface membrane with a ring-like pattern. However, brighter spots in buttonlike structures are also evident, suggesting accumulation of FITC–uPA on potential cell–substrate adhesion areas (Estreicher et al. 1990; Mignatti and Rifkin 1993; Møller 1993). Control cells incubated with equal amounts of FITC–uPA were not detectably stained, confirming the specificity of the staining observed, as did the specific displacement of FITC–uPA by unlabeled uPA.

FITC–uPA is a stable conjugate and can be used for accurate quantitative comparisons over a considerable length of time. We have stored FITC–uPA aliquots at −80C with no adverse effects for over 2 years. Repeated freezing and thawing of the same aliquot of FITC–uPA did not affect binding and staining efficiency, and we did not observe reduction of the apparent titer of FITC–uPA over time due to autoproteolysis. An example is shown in Figures 1A vs 1B. The essentially overlapping profiles obtained on different batches of stably transfected cells, either freshly thawed or continuously kept in culture, confirm the high reproducibility of the staining procedure presented and the stability of the FITC–uPA. This is also an example of use of FITC–uPA to verify the stability of expression of the uPAR on transfectants. The same transfectants were also successfully sorted for predefined levels of uPAR by flow cytometry (unpublished observation). FITC–uPA maintains its physiological proteolytic activity after binding to uPAR (data not shown). Therefore, dynamic analysis by confocal microscopy of the redistribution and proteolytic activity of the FITC–uPA–uPAR complex on binding to a substrate matrix is also allowed by the use of labeled uPA.

Footnotes

Acknowledgements

Supported by the Italian Association for Cancer Research and by the Italian National Research Council, Convenzione CNR–Consorzio Mario Negri Sud, ACRO contract no. 94.01319.39, and Progetto Bilaterale contract no. 93.00870.CT04.

We thank Drs P. Mignatti and B. Donati for material and support during the course of this work.

References

Alberti

Bucci

Fornaro

Robotti

Stella

(1991) Immunofluorescence analysis in flow cytometry: better selection of antibody-labeled cells after fluorescence overcompensation in the red channel. J Histochem Cytochem 39:701–706

Alberti

Parks

Herzenberg

(1987) A single laser method for subtraction of cell autofluorescence in flow cytometry. Cytometry 8:114–119

Appella

Robinson

Ullrich

Stoppelli

Corti

Cassani

Blasi

(1987) The receptor-binding sequence of urokinase. A biological function for the growth-factor module of proteases. J Biol Chem 262:4437–4440

Azevedeo

Johnson

(1990) Temperature-dependent lateral and transverse distribution of the epidermal growth factor receptor in A431 plasma membranes. J Membr Biol 118:215–224

Bendayan

(1995) Possibilities of false immunocytochemical results generated by the use of monoclonal antibodies: the example of the anti-proinsulin antibody. J Histochem Cytochem 43: 881–886

Bentham

Aplin

Norman

(1994) Histochemical detection of binding sites for human growth hormone using biotinylated ligand. J Histochem Cytochem 42:103–107

Bentham

Ohlsson

Lindahl

Isaksson

Nilsson

(1993) A double-staining technique for detection of growth hormone and insulin-like growth factor-I binding to rat tibial epiphyseal chondrocytes. J Endocrinol 137:361–367

Danø

Andreasen

Grøndal–Hansen

Kristensen

Nielsen

Skriver

(1985) Plasminogen activators, tissue degradation and cancer. Adv Cancer Res 44:139–266

De Jong

Rozemuller

Bauman

JGJ

Visser

JWM

(1995) Biotinylation of interleukin-2 (IL-2) for flow cytometric analysis of IL-2 receptor expression. J Immunol Methods 184:101–112

10.

Dell'Arciprete

Stella

Fornaro

Ciccocioppo

Capri

Naglieri

Alberti

(1996) High-efficency expression gene cloning by flow cytometry. J Histochem Cytochem 44:629–640

11.

Duffy

O'Grady

Devaney

O'Siorain

Fennelly

Liynen

(1988) Urokinase plasminogen activator, a marker for aggressive breast carcinomas. Cancer 62:531–533

12.

Estreicher

Muhlhauser

Carpentier

Orci

Vassalli

(1990) The receptor for urokinase type plasminogen activator polarizes expression of the protease to the leading edge of migrating monocytes and promotes degradation of enzyme complexes. J Cell Biol 111:783–792

13.

Hardy

(1986) Purification and coupling of fluorescent proteins for use in flow cytometry. In Weir

Herzenberg

Black-well

, eds. Handbook of Experimental Immunology. 104th ed. Edinburgh, Blackwell Scientific Publications, 9.11–91.10

14.

Hearing

Law

Corti

Appella

Blasi

(1989) Modulation of metastatic potential by cell surface urokinase of murine melanoma cells. Cancer Res 48:1270–1278

15.

Mignatti

Rifkin

(1993) Biology and biochemistry of proteinases in tumor invasion. Physiol Rev 73:161–195

16.

Møller

(1993) Structure and function of the urokinase receptor. Blood Coagul Fibrinol 4:293–303

17.

Nisonoff

Hopper

Spring

(1975) Nature of the active site of an antibody molecule and the mechanism of antibody–hapten interactions. In Dixon

FJJ

Kunkel

, eds. The Antibody Molecule. New York, Academic Press, 43–57

18.

Ossowski

Clunie

Masucci

Blasi

(1991) In vivo paracrine interaction between urokinase and its receptor: effect on tumor invasion. J Cell Biol 115:1107–1112

19.

Ossowski

Reich

(1983) Antibodies to plasminogen activator inhibit human tumor metastasis. Cell 35:611–619

20.

Parks

Lanier

Herzemberg

(1986) Flow cytometry and fluorescence activated cell sorting (FACS). In Weir

Herzenberg

Blackwell

, eds. Handbook of Experimental Immunology. 104th ed. Edinburgh, Blackwell Scientific Publications, 29.1–29.214

21.

Pyke

Kristensen

Ralfkiaer

Grondahl Hansen

Eriksen

Blasi

Danø

(1991) Urokinase-type plasminogen activator is expressed in stromal cells and its receptor in cancer cells at invasive foci in human colon adenocarcinomas. Am J Pathol 138:1059–1067

22.

Quax

PHA

Pedersen

Masucci

Weening–Verhoeff

Danø

Verheijen

Blasi

(1991) Complementation between urokinase-producing and receptor-producing cells in extracellular matrix degradation. Cell Regul 2:793–803

23.

Roldan

Cubellis

Masucci

Behrendt

Lund

Danø

Appella

Blasi

(1990) Cloning and expression of the receptor for human urokinase plasminogen activator, a central molecule in cell surface, plasmin dependent proteolysis. EMBO J 9:467–474

24.

Schultz

Zhang

(1992) The role of urokinase and urokinase inhibitor in tumor cell metastasis. Fibrinolysis 6:23–29

25.

Simon

Rao

Wei

Majdic

Ronne

Kobzik

Chapman

(1996) Mac-1 (CD11b/CD18) and the urokinase receptor (CD87) form a functional unit on monocytic cells. Blood 88:3185–3194

26.

Uriel

Lunardi–Iskandar

Laborda

Torres

Naval

Georgoulias

Lowi

Jasmin

Mishal

(1990) Defective uptake of alpha-fetoprotein (AFP) and transferrin (Tf) by PHA-activated peripheral blood lymphocytes from patients with AIDS and related syndromes. AIDS Res Hum Retrovir 6:401–410