ANALYTICAL

BIOCHEMISTRY

190,48-56

(1990)

Detection of Rat Ceruioplasmin Receptors Using Fluorescence Microscopy and Microdensitometry Robert V. Stern and Earl Frieden Chemistry Department, Florida State University, Tallahassee, Received

March

Florida 32306

26,199O

The plasma membrane receptor for ceruloplasmin (Cp; EC 1.16.3.1) was first discovered in chick heart and aorta by Stevens et al. (1) and has since become more firmly established. Receptors have been found in rat liver (2), in rat brain (3), on human blood leukocytes (4), and on human and rabbit erythrocytes (5-7). Attempts

to determine the function for the receptor have initially focused on its role in mediating copper delivery to the cell. In the copper-deficient chick aorta (8) and in human K562 cells (9), delivery of Wu from Cpl to an intracellular enzyme, superoxide dismutase (CuZn-SOD), has been shown. Along with its role in copper transport, the Cp receptor may serve a protective role in preventing lipid peroxidation by increasing the effective concentration of the antioxidant Cp at the membrane surface. Although this role has not yet been demonstrated, the mechanism of Cp alone as an extracellular superoxide dismutase has been elucidated (10). Investigating tissues for the Cp receptor initially began with the use of ‘251-labeled Cp (1,2,4-7) and more recently 6’Cu-labeled Cp (3,8,9). The radiolabeled Cp was used in plasma membrane binding experiments and Scatchard analysis (which requires saturable, reversible, and specific binding by the ligand) of the data revealed a receptor. Following the discovery of the receptor, the next step was to localize it using cell-specific binding studies. Cells of the liver were separated by collagenase digestion, incubated with a latex minibeadbound Cp, and examined by electron microscopy (EM) (2). Of the several tissues found to contain Cp receptors, to date only for rat liver (2) and human leukocytes (4) has the cell-specific binding been determined (unfortunately no binding to human erythrocytes was observed with EM). One avenue that has not yet been explored in localizing the receptor is that of labeling Cp with a fluorescent probe and using fluorescence microscopy to observe its binding. Rat Cp (rCp) has been labeled using rhodamine B isothiocyanate (RBITC) and the resulting conjugate (RBITC-rCp) showed membrane specific binding to rat erythrocytes as evaluated by microdensitometry. Evidence for a platelet receptor is also suggested.

1 Abbreviations used: MCID, microcomputer imaging device; Cp, ceruloplasmin; SOD, superoxide dismutase; EM, electron microscopy; RBITC, rhodamine B isothiocyanate; PAGE, polyacrylamide gel electrophoresis; FITC, fluorescein isothiocyanate; PPD, p-phenylenediamine.

Isolation of Rat Ceruloplasmin Blood was obtained from Sprague-Dawley rats (200 to 400 g) under ether anesthesia via the abdominal aorta

Rat ceruloplasmin (rCp) has been labeled with the fluorophores fluorescein and rhodamine by using the isothiocyanate derivatives (FITC, RBITC). High p-phenylenediamine oxidase activity of the resulting conjugates was observed (70-90% of native activity). Polyacrylamide gel electrophoresis showed fluorescein-labeled rCp (FITC-rCp) had an increased mobility, while rhodamine-labeled rCp (RBITC-rCp) showed no increase in mobility when compared to native rCp. RBITC-rCp was tested as a probe for ceruloplasmin receptors on rat erythrocytes using fluorescence microscopy to detect membrane binding. Film negatives of blood smears exposed under identical conditions were analyzed by microdensitometry to give relative optical densities for the amount of RBITC-rCp bound per unit areaof the plasma membrane. With this technique, binding of rCp was observed to be saturable, reversible, and specific. Competition of the protein ligands superoxide dismutase, catalase, and unlabeled rCp against RBITCrCp already bound on erythrocytes showed that only rCp could displace the bound RBITC-rCp with 32% specific binding being observed. Low levels of membrane binding were seen after the erythrocytes had been trypsin treated. Platelet binding was also detected and it was saturable, reversible, and trypsin sensitive. However, only 20% of the bound RBITC-rCp could be displaced by rCp. These studies demonstrate a versatile technique for detection and localization of Cp receptors. 0 1990 Academic Press, Inc.

MATERIALS

AND

METHODS

48 All

Copyright 0 1990 rights of reproduction

0003-2697/90 $3.00 by Academic Press, Inc. in any form reserved.

DETECTION

OF

CERULOPLASMIN

and collected in a plastic syringe containing a few crystals of disodium ethylenediaminetetraacetate (EDTA). The rats had been previously injected with turpentine to boost the Cp levels (11). The plasma collected after centrifugation (200 ml) was diluted with 1400 ml of water and applied to a Sepharose CL-6B ethylamine affinity column (5.5 X 4.2 cm) at a flow rate of 200 ml/h. (All steps unless noted were carried out at 4°C.) A blue band formed over the upper third of the gel. The column was washed at 100 ml/h with 20 mM potassium phosphate (KPi), pH 7.2, until Azso was less than 0.1. The Cp was eluted with 300 InM KPi, pH 7.2, at 180 ml/h and all blue fractions were combined. To the Cp solution was added 2 M sodium acetate (NaAc) + 1.25 M NaCl, pH 5.6, to achieve a final concentration of 0.2 M NaAc + 0.125 M NaCl. Then the Cp was precipitated with chloroform:ethanol(1:9) using a ratio of Cp solution to chloroform:ethanol of 1 to 3. The chloroform:ethanol was added over 30 min and stirred for 1 h. After centrifugation the blue pellet was dissolved in 155 mM NaCl, pH 7.0, at 24 * 1°C (120 ml NaCl stirred for 30 min). After centrifugation the supernatant was dialyzed against 4 liters of 155 mM NaCl, pH 7.0, overnight. The resulting solution was concentrated using Amicon ultrafiltration with an XM-100 membrane. The ratio A,,,: A,, of this preparation was greater than 0.040 and was judged to be of high purity showing less than 25% apoceruloplasmin on native polyacrylamide gel electrophoresis (PAGE) and virtually no other bands (see Fig. 2).

Sepharose CL-6B Ethylamine

Column

Upon the basis of the procedure of Calabrese et al. (12) and established techniques (13) an affinity gel was constructed using Sepharose CL-6B to facilitate the isolation of rat Cp. First, 100 ml of washed Sepharose CL-6B 200 was suspended in 100 ml of 2.0 M NaOH containing 2 mg/ml NaBH,. To this was added 0.20 mol of the diepoxide 1,4-butanediol diglycidyl ether (Sigma), and the mixture was swirled on a gyro-shaker for 10 h at 24 -+ 1°C. The gel was then washed thoroughly in distilled water and 0.40 mol of chloroethylamine (Aldrich) in 150 ml of 0.4 M Na,CO,, pH 11.3 (freshly prepared), was added. This reaction was carried out for 30 h at 24 + 1°C with shaking while the pH was maintained between 11 and 12 by the addition of 6 M NaOH. The gel was then washed with distilled water and unreacted epoxide groups were blocked by adding 100 ml of 1 M hydroxylamine, pH 7.0, overnight at 4°C. Before use, the gel was washed with distilled water, the eluting buffer 300 mM KPi, pH 7.2, distilled water, and then with 20 mM KPi, pH 7.2.

RECEPTORS

USING

49

FLUORESCENCE

Labeling Rat Cp with FITC and RBITC (Modified from (14)) For fluorescein isothiocyanate (FITC; Sigma) conjugation, 16 ~1 of rCp (31.5 mg/ml) was brought to 300 ~1 with 0.2 M NaHCO,, pH 9.0. To this was added 200 ~1 of FITC (2.5 pg/pl) which had been dissolved in the same bicarbonate buffer to achieve a final ratio of 0.5 mg FITC to 0.5 mg protein. The reaction was carried out in a brown 1.5-ml microfuge tube for 15-, 30-, and 60-min intervals at 24 -t 1°C. To remove unreacted FITC, the reaction mixture was passed over a Sephadex G-25 Pasteur pipet column (8.0 X 0.5 cm) equilibrated with 20 mM KPi + 155 mM NaCl, pH 7.2. The initial yellow fractions were pooled up to the fraction that showed a rise in pH (15). For RBITC (Sigma) conjugation, 16 ~1 of rCp (31.5 mg/ml) was brought to 490 ~1 with the bicarbonate buffer and to this was added 10 gl of RBITC (5.0 pg/pl) which had been dissolved in dimethylformamide to achieve a final ratio of 0.05 mg of RBITC to 0.5 mg protein. Incubation and removal of the free RBITC were the same as for the FITC conjugation, except the pooled fractions were purple. To ensure the complete removal of free fluorophore, the conjugated protein solution was treated with activated charcoal (14), 2 mg of charcoal per 0.6 ml of the conjugate (0.5 to 0.8 mg/ml protein) for 1 h on ice.

Protein

Assay and Activity Assay

Protein concentrations of the conjugates were determined using the Bio-Rad protein assay. Standards using bovine serum albumin were made with 155 mM NaCl, pH 7.0, and the conjugates were diluted 1:l with the same saline solution and run in duplicate. Good linearity in the assay was observed to 50 pg of protein and the dilution of the conjugate allowed reading within this range. p-Phenylenediamine oxidase activity was assayed by the method of Rice (16) using 0.17% p-phenylenediamine dihydrochloride (PPD. 2HCl) (Eastman Kodak or Sigma) and an enzyme concentration of 0.40-0.60 mg/ml.

Polyacrylumide

Gel Electrophoresis

and Densitometry

Discontinuous gel electrophoresis was performed by the method of Ornstein and Davis (17) on a Bio-Rad Protean vertical slab apparatus using a 3% stacking gel and a 7% running gel with a Tris-glycine buffer. Half of the gel was stained for oxidase activity by incubating the gel in 1.2 M NaAc buffer, pH 5.2, containing 0.4% PPD * 2HCl until bands developed. The other half of the same gel was stained for protein with Coomassie blue R-250.

50

STERN

AND

FRIEDEN

Densitometry of the gels and of black and white photographs of gels was performed using a Bio-Rad Model 620 videodensitometer interfaced with a Bio-Rad Model 3392A integrator. A point-to-point baseline was set, and the image expanded and then integrated with no filtering or enhancement. Initial scans showed no differences in apoprotein to native protein ratios between the actual gel and photographs of the gel, so later work was carried out on photographs exclusively. Blood Cell Acquisition and Slide Preparation

Whole blood was obtained from rats under ether anesthesia by cutting a 2- to 3-mm section of the tail with a razor blade. The blood was collected in a plastic microfuge tube containing a few crystals of EDTA. Blood smears were fixed 5 min in cold absolute ethanol. After drying, wells were constructed at the feathered end of the smear with rubber cement applied from a l-ml pipetman. Incubations with labeled rCp (50-100 ~1; 0.40.8 mg/ml) were carried out at 24 + 1°C. Saturating conditions were ensured as a minimum of 0.15 nmol of RBITC-rCp was added to a given well enclosing thousands of cells, whiIe a previous study (5) found that 0.15 pmol ‘251-labeled Cp readily saturated 22.5 million erythrocytes. The slides were washed by carefully removing the rubber cement and then flooding the slide with 20 mM KPi + 155 mM NaCl, pH 7.2 (PBS). For competition studies the wells were reconstructed and a 2 mg/ml solution of the appropriate protein (bovine superoxide dismutase (Calbiochem), catalase (Sigma)) in PBS buffer was added. Poly-l-lysine-coated multitest slides (Roboz Scientific) were prepared by loading the wells with a 0.5% solution of poly-1-lysine (M, greater than 300,000; Sigma) for 30 min and then washing the wells with PBS (18). Cells were then incubated for 30 min, washed with PBS, air-dried, ethanol fixed, and then incubated with labeled rCp. Cover slips were mounted using a glycerollp-phenylenediamine solution (19) to reduce fading. Fluorescence Microscopy and Photomicrography

Slides were examined and photographed on a Nikon Microphot-FX microscope equipped with an epi-fluorescence attachment. The filters selected were filter block B for FITC and filter block G for RBITC allowing for specific excitation of the label. Cells were photographed with the 40x objective using Kodak Technical Pan film and an ASA of either 50 or 100. All slides for a given experiment were manually exposed for the same length of time, 2-5 s, so they could be directly compared. The film was developed in Kodak HC-110 developer using dilution B for 6 min. At least three slides were photo-

i

0

I,

I,

I,

I,

I

10

20

30

40

50

Reaction

FIG. 1.

Labeling as described under was determined by the range observed for FITC-rCp; 558

,

,

60

time (minutes)

rCp with FITC and RBITC. Reaction carried OLI Materials and Methods. Protein concentration the Bio-Rad protein assay. Error bars represen for three to five trials. 494 nm is the X, observe{ nm is the X, observed for RBITC-rCp.

graphed per time point and only uniformly of the slide were used.

labeled areas

Computer Analysis

A microcomputer imaging device (MCID; Imaging Research Inc.) was used to determine the extent of exposure of the film per unit area of erythrocyte membrane. The system consisted of a Northern Light Precision Illuminator Model B90, a Sierra Scientific High Resolution CCD camera, and an IBM compatible microcomputer (80286 processor) that ran the brs2 software. The negative to be digitized was placed on the light table and the camera was focused with f-stop 2.8. The window size selected for analysis was 3 X 3 which just covered the width of the membrane (see Fig. 4), and at least eight windows were sampled per cell. The background was also sampled to ensure that the slide had been uniformly treated and to aid in comparing slides. Twenty cells per slide were read from various regions of the slide for a total of at least 60 cells per time point.

RESULTS

Conjugation of Fluorophores with rCp and Conjugate Characterization

Both conjugates FITC-rCp and RBITC-rCp showed shifted X maximums from that of the respective free fluorophore. A shift of X,, = 492 nm (FITC) to h,, = 494 nm (FITC-rCp) and of X,,, = 551 nm (RBITC) to

DETECTION

a b c

d

e

OF

CERULOPLASMIN

a’ b’ c’ d’ e’

FIG. 2. PAGE of native rCp and conjugates (FITC-rCp and RBITC-rCp). (Lanes a-e) Activity stained with PPD * 2HCl; (Lanes a’-e’) stained with Coomassie blue R-250.35 pg of protein was loaded per lane. Lane assignments: (a, a’) native rCp; (b, b’) FITC-rCp 15min labeling reaction; (c, c’) FITC-rCp 60-min labeling reaction; (d, d’) RBITC-rCp 15-min labeling reaction; (e, e’) RBITC-rCp 60-min labeling reaction.

x max= 558 nm (RBITC-rCp) was observed in PBS. The ratio of the absorbance at the appropriate X maximum divided by the protein concentration (mg/ml) was used to indicate the extent of incorporation of the label (see Fig. 1). Each time point represents the mean of at least three independent experiments and the error bars represent the range observed for those trials. Rechromatography of the labeled protein on Sephadex G-25 showed only one band, however, charcoal treatment yielded a decrease in the absorbance/protein concentration ratio of lo-30%, which demonstrated that this step was necessary in removing noncovalently bound label. p-Phenylenediamine oxidase activity for the native enzyme (938 IU/mg protein) was reduced in the conjugates. Freshly prepared rCp conjugates showed activities of 80-90% of native activity, while stock rCp kept a few weeks at 4°C at concentrations of 10 mg/ml in 20 mM PBS (KPi) showed reduced conjugate activity (7080%). Conjugates resulting from labeling reactions of 30 min or less showed the same level of activity for a given preparation of rCp; however, longer reaction times led to further losses of activity. Loss of activity could not be solely attributed to label incorporation, because control experiments without fluorophore showed that the reaction conditions of bicarbonate buffer, pH 9.0, alone and buffer plus dimethylformamide led to similar declines in activity. Also, very little copper loss was observed. In order to check the protein integrity of the conjugates, native PAGE was run (see Fig. 2). The left half of the gel was stained for activity using PPD. 2HCl (lanes a-e), while the right half of the gel (lanes a’-e’) was stained with Coomassie blue R-250. The native rCp (lanes a and a’) showed two major protein bands in the ratio of upper band (1825%):lower band (75-82%) from densitometry. The lower band showed activity (native enzyme), while the upper did not (apoenzyme).

RECEPTORS

USING

FLUORESCENCE

51

This is a typical band pattern seen for rCp (20,21), except that using this isolation procedure more of the enzyme remains native. The fluorescein-labeled conjugates are shown in lanes b/b’ and c/c’. Again the lower band showed all the activity but was now shifted to a lower region, which indicated an increased negative charge. The longer the labeling reaction (lanes b and b’, 15-min reaction; lanes c and c’, 60-min reaction), the more incorporation of fluorescein label with its net negative charge, and so the increased migration down the gel. The rhodamine-labeled rCp (lanes d and d’, 15min reaction; lanes e and e’, 60-min reaction), however, showed no increase in migration, since rhodamine carries no net charge. The binding of Cp to a plasma membrane receptor may involve an electrostatic component and for this reason the FITC-rCp was not used further. The RBITC-rCp used in the following experiments was based on a 20-min labeling reaction.

Fluorescence Microscopy Attempts to first visualize the binding of RBITC-rCp to erythrocyte membrane were carried out in a hemacytometer with limited success. The fluorescence was too weak to photograph and there was no way to save the result. Second, poly-l-lysine-coated slides were tried, because the cells would adhere and these could then be observed in a fixed or an unfixed state (22). Unfortunately, the cells on these treated slides were distorted by the binding to the slide (see Fig. 3a) and the whole cell fluoresced. Finally, blood smears were attempted and these proved the most useful. The cells on the smear had to be fixed, because unfixed cells washed off the slide. Milder fixatives were tried such as 4% paraformaldehyde or 70% ethanol, but in each case the cells still washed off the slide. Absolute ethanol worked well and is a mild fixative. The result of an experiment with RBITC-rCp binding to rat erythrocyte membrane can he seen in Fig. 3b. The “halo” around each cell represents high levels of binding by the protein. In control experiments a faint halo was seen for free RBITC incubated with erythrocytes, but RBITC hydrolyzed under the labeling reaction conditions for 2 h showed little or no halo on the erythrocytes. There were high levels of background fluorescence for both free RBITC and the hydrolyzed form, while RBITC-rCp showed much lower levels. In order to check that receptor binding was due to a protein-protein interaction, whole blood was incubated with 0.67 mg/ml added trypsin (12,200 units/mg; Sigma) for 1 h at 24 f 1°C before slide preparation. No halos were evident in these trypsin treated cells (see Fig. 3~). Trypsin

52

STERN

AND

FRIEDEN

FIG. 3. Various slides of RBITC-rCp bound on cells. Photographed using the 40x objective. (a) Poly-l-lysine-coated slide ethanol fixed; erythrocytes incubated with RBITC-rCp for 60 min (2-s exposure). Bar = 20 pm. (b) Whole blood smear ethanol fixed; erythrocytes and platelets incubated with RBITC-rCp for 60 min. Platelet indicated by arrow (3-s exposure). Bar = 20 pm. (c) Trypsin-treated whole blood smear ethanol fixed; erythrocytes and platelets incubated with RBITC-rCp for 60 min. Platelet indicated by arrow (5-s exposure). Bar = 20 pm.

treatment for shorter periods revealed incomplete digestion of the receptor, as some cells with halos were observed. Platelets showed the most binding as they are the brightest cell observed (see Fig. 3b). This binding may be protein-protein in nature as trypsin treatment reduced the fluorescence of the platelets; however, when photographed they still appeared brightly fluorescent (see Fig. 3~). This may be explained by noting that the slide had been exposed longer because its overall fluorescence was low. Whether this binding was only membrane associated was not possible to determine due to the small size of the cell which concentrated the fluorescence and exposed the film over the entire cell body.

Specificity

Determined

Using Computer

Imaging

With the MCID, the gray levels of the digitized negative could be converted into relative optical density units (23). So, for a given area of the erythrocyte mem-

brane the intensity of the fluorescence, recorded on the negative, could be evaluated at various lengths of incubation time before and after competition with various protein ligands. The display of the computer monitor (see Fig. 4) shows the erythrocytes with black halos since this is the digitized negative and the platelets appear as the small, black bodies. Using the mouse, windows on the cell membrane were selected taking care not to sample junctions between cells. Background readings were also made to establish that the slide had been uniformly treated which was helpful in comparing slides. The results of a competition experiment are shown in Fig. 5. For the first 60 min the RBITC-rCp was allowed to bind to the cells on the smears; after removing the wells and washing the free RBITC-rCp away, the wells were reconstructed and the competing ligands, bovine superoxide dismutase (SOD), catalase, and rCp, were added. Each time point is the mean of three different slides (60 erythrocytes), with the error bar showing the standard deviation for the 60 cells. Neither the SOD nor

DETECTION

OF CERULOPLASMIN

FIG.

RECEPTORS USING FLUORESCENCE

3-Continued

53

54

FIG. , 4. Computer red membrane for mierodensitometry

STERN

AND

monitor from MCID showing a digitized negative binding while the smaller platelets (indicated by arrow) indicated by white blocks. Bar = 20 pm.

the catalase displaced the bound RBITC-rCp, while the rCp displaced an average of 32% of the bound RBITCrCp. This result compares well with the 40% displacement seen in our laboratory for lz51-labeled bovine Cp (5) and with the 36% displacement reported by Stevens et al. (1) for ‘251-labeled chick Cp. DISCUSSION

In establishing a Cp receptor, certain criteria for it had to be met, i.e., saturability, reversibility, and specificity for Cp. In testing a new ligand (RBITC-rCp) for the receptor the same criteria must be met by the ligand itself. The RBITC-rCp showed (i) saturation in binding (see Fig. 5), (ii) reversibility in binding (see Fig. 5), (iii) specificity (as proteins other than Cp could not displace it), and (iv) that a protein-protein interaction is required (as trypsin-treated cells showed greatly reduced binding). In addition the RBITC-rCp maintained a high level of oxidase activity which is not necessarily the case in radiolabeled Cp or gold/minibead-bound Cp. Of course the cells in these experiments were ethanol fixed,

FRIEDEN

of RBITC-rCp hound on erythrocytes showed fluorescence over the entire

and platelets. cell body. Window

Erythrocyr tes size sampl .ed

but ethanol is a mild fixative and probably does not drastically affect protein-protein interactions. Ethanol fixed cells have often been used in antibody binding to protein antigens on plasma membrane (15). Indeed, often it is necessary to fix cells before protein binding can be observed due to lipid interference. Although Fig. 1 accurately represents the incorporation of the fluorescent labels, it is simpler to have an approximate number of fluorophores bound per Cp molecule. With a method (14) based upon extinction coefficients of 74,000 M-l cm-’ for FITC and of 73,000 M-l cm-l for RBITC, average ratios for FITC:rCp of 13:l and for RBITC:rCp of 5:l were obtained for a 60-min labeling reaction. These are rough approximations because the values for the extinction coefficients were derived for labeled y-globulin in PBS. The decreased labeling with RBITC can be explained by the fact that although the aromatic ring systems of FITC and RBITC are similar, RBITC has four additional ethyl groups that may sterically hinder its incorporation. In previous studies (1,6) protein aggregation of Cp has been mentioned as a potential problem in doing re-

DETECTION

OF CERULOPLASMIN

,.- ;;;,.:;.. _._“,” ,,.. ” !

Catalase SOD

"-1. __ -P

-

0

20

Binding+--.--Competition.

40

60

rCP

,---j

80

100

120

FIG. 5. Results from microdensitometry of proteins competed against RBITC-rCp bound on erythrocytes. Relative optical density derived from computer gray-level analysis of film negatives. RBITCrCp was incubated for 60 min, removed, and washed, and the competing ligands (SOD, catalase, and rCp) were added for 60 min. Error bars represent the standard deviation for three slides (60 erythrocytes), except SOD trial two slides (40 erythrocytes).

ceptor binding experiments. If protein aggregation does occur when excess unlabeled Cp is added to RBITCrCp, then quenching of fluorescence may be expected. This is a relevant concern, because in the competition experiment fluorescence quenching could be misinterpreted as displacement of the ligand. Since the addition of the proteins, SOD, and catalase did not show quenching of the fluorescence of the bound RBITC-rCp, it appears that these proteins do not associate with Cp under these conditions. Self association had not been ruled out, so a fluorescence experiment (results not shown) was conducted in which RBITC-rCp at the same concentration used in the incubations was excited at 558 nm and the emission followed at 583 nm in a PerkinElmer LS-5 fluorescence spectrophotometer. Unlabeled rCp (2 mg/ml) was added and no quenching was observed when dilution was corrected. As for existing data for Cp binding to erythrocytes, the calcium dependence reported using 1251-Cp (6,7) was not observed in this system. None of the buffers contained calcium salts and since EDTA was used as an anticoagulant, there should have been no free calcium ion in the blood before slide preparation. No fluorescence was seen inside the cell, which supports earlier observations (6) that Cp does not enter the erythrocyte. Experiments by Saenko et al. (7) have recently been carried out to determine the role of the carbohydrate in the Cplreceptor binding, in which Cp showed 87% of control specific binding in trypsin-treated erythrocytes. This result is difficult to reconcile with our data and the data of Stevens et al. (1) which showed very little binding after trypsin treatment, unless there was incomplete digestion of the receptor in the experiments of Saenko et al. (7).

55

RECEPTORS USING FLUORESCENCE

RBITC-rCp may prove helpful in determining sites of Cp binding in organs without extensive tissue preparation. Frozen sections of a tissue incubated with RBITCrCp and photographed could be digitized and stored by the MCID, and then nonspecifically labeled images (ones incubated with excess unlabeled rCp) could be subtracted from images of total binding to give areas of Cp specific binding. This technique could prove useful in finding and localizing receptors quickly and relatively easily. Of course the areas of specific binding in the in uitro experiments would have to be further evaluated as to whether the Cp, a plasma protein, could actually reach those areas in viuo. Scatchard binding experiments may be possible using RBITC-rCp in conjunction with a fluorometer which would yield quantitative values for dissociation constants and numbers of receptor sites without the complications normally associated with radiolabeled Cp. In studying the erythrocyte Cp receptor using RBITC-rCp, strong binding to the platelets was noted which was trypsin sensitive (indicating a possible receptor protein). Using the same slides from which the erythrocyte data was collected, the platelets were evaluated for RBITC-rCp binding using the MCID and 20% specific binding was found (results not shown). The high level of nonspecific binding could be explained by the nature of the platelet membrane which is highly adhesive under the appropriate conditions and which is well dressed with a variety of receptor proteins. The ethanol fixed platelet membrane may be permeable to Cp, and so RBITC-rCp may have entered the cell which would also contribute to a high nonspecific binding. Human platelets do contain a CuZn-superoxide dismutase (24), and so a platelet receptor for Cp in mammalian systems is probable. Platelet aggregation is abnormal in patients with Wilson’s disease (25), which may indicate that Cp bound to a platelet receptor is a necessary component for the aggregation reaction. Radiolabel studies and/or more fluorescent data with purer populations of platelets are required before a platelet receptor for Cp is more conclusively established. ACKNOWLEDGMENTS We would like to thank Dr. Charles C. Ouimet and Barbara N. Baker for the use of and assistance with the microdensitometer, Sandra H. Silvers and Kimberly A. Riddle for their technical assistance with the fluorescence microscope, and Michael W. Davidson for his advice with the photography.

REFERENCES 1. Stevens,

M. D., DiSilvestro,

R. A., and Harris,

E. D. (1984)

Rio-

Cell Res. 155,

232-

chenistry23,261-266. 2. Kataoka, 240.

M.,

and Tavassoli,

M.

(1984)

Exp.

56

STERN

3. Orena,

S. J., Goode,

C. A., and

Linder,

M.

C. (1986)

AND

Biochem.

Biophys. Res. Commun. 139,822-829. 4. Kataoka,

M.,

and Tavassoli,

M.

(1985)

Exp. Hematol. 13, 806-

810. 5. Barnes,

mun.

FRIEDEN 15. Kawamura, A., Jr. (1977) in Fluorescent Antibody and Their Applications, 2nd ed., pp. 37-114, Univ. Press, Tokoyo, Japan. 16. Rice, E. W., Wagman,

G., and Frieden, 125, 157-164.

6. Saenko,

E. L., Basevich,

E. (1984)

Biochem. Biaphys. Res. ComYaropolov,

A. I. (1988)

Bio-

V. V., and Yaropolov, 53, 1140-1145.

A. I. (1988)

Bio-

V. V., and

chemistry (Engl. Transl.) 53,274~278. 7. Saenko,

E. L., Basevich,

chemistry (Engl. Transl.) 8. Dameron,

C. T., and Harris,

675. N&r. 119, 779-784. 10. Vasil’ev, V. B., Kachurin, A. M., and Soroka, N. V. (1988) Biochemistry (Engl. Transl.) 63. 1756-1762. 9. Percival,

S. S., and Harris,

11. DiSilvestro, R. A., Barber, (1988) Biol. Trace Elem. 12. Calabrese,

L., Mateescu,

(1963) Biochem. Int. 16,

E. D. (1989)

E. F., David,

J.

E. A., and Cousins,

R. J.

Res. 17, l-9. M. A., Carbonaro, 199-208.

13. Dean, P. D. G., Johnson, W. S., and Middle, ity Chromatography: A Practical Approach, ford, England. 14. Nairn, R. C. (1976) in Fluorescent Protein Churchhill Livingstone, New York.

M., and Mondovi,

B.

F. A. (1985) in Affinpp. 34-35, IRL, Ox-

Y. (1963) Stand. Meth-

E., and Takenaka,

ods Clin. Chem. 4, 39-46. 17. Ornstein, 321-427.

L., and Davis,

18. Nakstad,

B., Lyberg,

B. J. (1964) T.,

Baklein,

Ann. N.Y. Acad. Sci. 121, K.,

and

Boye,

N.

(1989)

J.

Pathol. 157,99-107. 19. Johnson,

Biochem. J. 248, 669-

E. D. (1987)

Techniques of Tokoyo

G. D., and De C. Nogueirn

Araujo,

G. M. (1981)

J. Zm-

munol. Methods 43,349-350. 20. Zakharova, lovskii, M. 21. Legras, Cloarec,

B., Durou, L. (1980)

V. B., Gorbunova,

V. N., and Shav-

Biochemistry (Engl. Transl.) 48,1475-1485. M.

R.,

Cottencin,

M.,

Esvant,

J. Y., and

C. R. Sciences Sot. Biol. Ses. Fil. 174, 82-86. D., Schatten, G., and Sale, W. (1975) J. Cell Biol. 66, 198-

22. Mazia,

200. 23. Ostrowski, Localization: C. A., Eds.), 24. Sinet,

E. T., Vasil’ev,

N. (1983)

N. L., Pert, C. B., and Pert, A. (1988) in Receptor Ligand Autoradiography (Leslie, F. M., and Altar, pp. 148-150, A. R. Liss, New York.

P. M., Lavelle,

F., Michelson,

A. M., and Jerome,

H. (1975)

Biochem. Biophys. Res. Commun. 67,904-909. Tracing,

pp. 21-67,

25. Owen,

C. A., Goldstein,

Intern. Med. 136,148-152.

N. P., and Bowie,

E. J. W.

(1976) Arch.