Cell Tissue Res (1997) 288:119–126
© Springer-Verlag 1997
161815.705 030512.401 030512.110 191512.500 172101.500 190122.587 180120*006 130114*142
A simple assay for quantification of protein in tissue sections, cell cultures, and cell homogenates, and of protein immobilized on solid surfaces Angela Dieckmann-Schuppert1, Hans-Joachim Schnittler2 1 Zentrum für Hygiene und Medizinische Mikrobiologie, Arbeitsgruppe Parasitologie, Philipps-Universität Marburg, Robert-Koch-Strasse 17, D-35037 Marburg, Germany 2 Anatomisches Institut der Julius-Maximilians Universität Würzburg, Koellikerstrasse 6, D-97070 Würzburg, Germany
&misc:Received: 9 August 1996 / Accepted: 31 October 1996
&p.1:Abstract. The determination of total protein is often a key step for the quantitative analysis of various parameters in tissue and general biochemical research. The classical protocols are restricted to a few compatible buffers, and protocols for the determination of protein in solutions containing protein agglomerates or of protein immobilized on solid surfaces are not available. In such cases, quantification may be complicated. Here, we describe a simple sensitive method for protein quantification circumventing all these restrictions. Proteins in solution or suspension in any buffer are spotted onto cellulose acetate, dried, and stained with Amido Black. After washing off the excess dye, bound Amido Black is solubilized in an acidic solution and determined photometrically. Tissue slices (fixed or native), adherent cell cultures, or Western blots can also be stained and their protein content determined irrespective of the supporting material. A micro-version of the protocol for proteins in solution allows large numbers of samples to be evaluated at a time in microtitration plates and requires only 1–2 µl per sample. A linear concentration dependency (r2=0.950–0.999) was obtained for all samples in all cases investigated. The method presented here permits the exact determination of soluble protein in a large variety of buffers, of insoluble or immobilized protein present on a wide variety of supports, and even of whole cells or tissue slices. &kwd:Key words: Protein determination – Quantification in tissue slices and cell cultures – Detergents – High salt concentrations – Solid surfaces – Western blot
This work was supported by the Deutsche Forschungsgemeinschaft, SFB 355/B5 Correspondence to: Hans-J. Schnittler (Tel.: +49–931–312706; Fax: +49–931–15988; E-mail:
[email protected])&/fn-block:
Introduction The quantification of a specific biochemical signal, such as an antibody reaction, the activity of an enzyme, or the incorporation of radioactivity is a central task in anatomical, cell biological, biochemical, and pathogenetic research. Although highly sensitive indicator methods detecting extremely low levels of the respective signals, for example, enhanced chemiluminescence, or the use of radioactive substances, are available, their specific quantification in tissue sections, adherent or suspended cell cultures, cells grown on biomaterials, tissue homogenates, or on Western blots is often limited, because of an insufficient determinability of total sample protein. A method allowing the precise determination of protein not only in solution, but also in these special cases, and requiring as small a sample as possible would therefore be of great interest. The classical, widely used protocols for protein determination, such as the Lowry (Lowry et al. 1951) or Biuret (Weichselbaum 1946) methods and all their modifications, are applicable only to solutions. Moreover, these methods suffer from the limitation that they are not compatible with certain buffers, high salt concentrations, or the presence of detergents. In order to avoid these difficulties, a precipitation step is often included, which may however not always be quantitative and reproducible, especially in dilute solutions. The classical methods face even more restrictions when the protein to be determined presents itself as aggregates, whole cells, or bound to solid surfaces, i.e., in the cases mentioned above. Under these circumstances, protein determination proves extremely complicated and often the amount of protein can only be roughly estimated. The concept of staining protein spots dotted onto a solid support with Amido Black followed by elution and photometric quantification of the dye was put forward over thirty years ago (Heinzel et al. 1965), but has so far not gained wide propagation. This method has attracted more interest following the publications of Schaffner and Weissmann (1973) and Henkel and Bieger (1994),
120
because detergents such as SDS do not interfere with the determination. Sportsman and Elder (1984) have suggested circumventing the required photometric evaluation by the use of a laser gel scanner, but this equipment may not be available in every laboratory. Here, we present a modified improved method of Amido Black protein determination that exhibits high sensitivity not only in buffers containing detergent and high salt, but also with respect to tissue sections, cell cultures, cell and tissue homogenates, and protein immobilized on solid surfaces. We believe that the protocols provided here will be useful for every laboratory where sensitive protein quantification is a part of the daily routine but still an “art” depending on the respective problem. Materials and methods Materials and instruments Cellulose acetate (CA) sheets measuring 25×160 mm were obtained from Schleicher and Schüll (Dassel, Germany). Nitrocellulose (NC) was obtained from the same manufacturer and cut into pieces of a corresponding size. Phosphate-buffered saline (PBS) containing 8 g/l NaCl, 0.2 g/l KCl, 1.15 g/l Na2HPO4, 0.2 g/l KH2PO4 (pH 7.4) was obtained from Seromed (Berlin, Germany). Bovine serum albumin (BSA) was from Sigma (Deisenhofen, Germany) and Amido Black 10B from Merck (Darmstadt, Germany). All other chemicals were of the highest purity available. Polystyrene microtitration plates and multiwell cell-culture dishes were from by Becton Dickinson (Heidelberg, Germany). Cryostat sections were cut on a Frigocut 2800 (Reichert-Jung, Nußloch, Germany). Photometric analysis was performed on a Pharmacia Ultraspec 4 photometer (Pharmacia, Freiburg, Germany), and spectra were recorded with a Spekol VIS (Zeiss, Jena, Germany) automatic spectrophotometer. A Titertek Multiskan Plus MKII enzyme-linked immunosorbent assay (ELISA) reader (Eflab, Helsinki, Finnland) was used for the photometric evaluation of samples in microtitration plates. Image analysis of gels and the determination of tissue sections was performed by using the gel-blotting macro or area count of NIH-Image version 1.52 on a Macintosh computer. Suspended cells were counted and analyzed in a CASY-1 cell counter and analysis system (Schaerfe System, Reutlingen, Germany).
Tissue treatment and sections Livers from rats were removed after decapitation or after perfusion fixation with 4% formaldehyde in PBS under deep pentobarbital anesthesia via the left heart ventricle as described in detail elsewhere (Asan 1993). They were dissected into small pieces and frozen at –20°C. Cryostat sections (5 µm in thickness and approximately 3×4 mm) were cut, mounted on polyornithine-coated coverslips, and dried at 60°C for 1 h.
Cells Human umbilical vein endothelial cells (HUVEC) were isolated and cultured on 6-cm2 glass slides as described elsewhere (Schnittler et al. 1993). For the experiment shown in Fig. 6, these cells were fixed in 2% formaldehyde in PBS, followed by 3 washes in PBS, for 10 min.
PBS, and sample buffer for polyacrylamide gel electrophoresis containing 6% (w/v) SDS, 9% (v/v) glycerol, 10 mM dithiothreitol, 0.187 M TRIS-HCl at pH 6.9, supplemented with 0.02% (w/v) bromophenol blue, 1% Triton X-100, or 1% NP-40 where indicated. All samples in sample buffer were heated to 95°C for 5 min and stored at –20°C until used.
Dissolution protocol for protein quantification Macro-method. &p.2:Fields of 1 cm width were marked on CA or NC sheets with a pencil. The sheets were then mounted horizontally in a holder between magnets. Defined volumes of protein samples (10 µl) were applied onto each field. Blanks containing only the respective buffer served as the zero control. The sheets were dried for 15 min at room temperature (RT) and subsequently immersed in the staining solution Istain [0.5% (w/v) Amido Black 10B, 45% (v/v) each of methanol and water, and 10% glacial acetic acid] for 10 min. Staining was followed by three washes in solution I wash (47.5% each of methanol and water, and 5% glacial acetic acid) for 5 min each. The sheets were then dried again (5 min at RT), the individual samples cut apart, and placed into separate test tubes (2 ml Eppendorf cups). Dissolution of the CA pieces was then performed in 1 ml dissolution solution I diss (80% formic acid, and 10% each of glacial acetic acid and trichloroacetic acid) by incubation at 50°C for 15 min or 30 min at RT under shaking. The resulting blue solutions were read photometrically at 620 nm or 672 nm, respectively, against the corresponding reagent blank. All analyses were performed in triplicate or quadruplicate, and linear regression lines were calculated. Micro-method. &p.2:CA discs of 5.5 mm diameter were placed into microtitration plates (96 well), and 1 or 2 µl standard or sample protein were spotted onto the CA and further processed as described for the macro-method. Dissolution solution (100 or 200 µl) was used to dissolve the CA paper after the staining and destaining procedure. The microtitration plates were covered doubly with Parafilm to avoid evaporation of the solution. The samples were shaken for 30 min at room temperature, and the absorbance was measured at 620 nm.
Elution protocol for protein quantification The method was performed according to Henkel and Bieger (1994) as follows: staining for 3 min in solution II stain [0.1% Amido Black 10B (w/v), 45% (v/v) each of methanol and water, 10% glacial acetic acid], 2 washes for 3 min in water, then 2 washes for 3 min in solution IIwash (90% methanol, 2% acetic acid, 8% water), and again for 5 min in water. Elution of the dye from the dried sheets was accomplished by 1 ml eluent solution II elut (50% ethanol, 50% 50 mM NaOH/0.1 mM EDTA) per sample under shaking for 30 min. Where indicated, a combination of the dissolution and elution protocols was performed.
Electrophoresis and blotting SDS-polyacrylamide slab gel electrophoresis, electroblotting onto nitrocellulose, and fixation-staining of the slab gels with Coomassie blue were performed according to standard procedures as described elsewhere (Schnittler et al. 1990).
Results and discussion
Buffers tested
Staining of proteins by Amido Black, spectra, and standards
BSA stock solutions and dilutions were prepared in the following solutions: bidistilled water, PBS, 4 M urea in PBS, 1.5 M KI in
The adsorption of Amido Black to any protein is stoichiometrically dependent on the number of free amino groups
121 Table 1. Slopes of the regression lines obtained for BSA (1.25–4 mg/ml) in various buffers as obtained by the CA-dissolution method, and corresponding regression correlations. Note that the slopes increase with increasing denaturing power of the respective buffer (r2, square of correlation coefficient from linear regression analysis)&/tbl.c:& Slope
r2
Distilled water PBS Sample buffer + 1% Triton X-100 Sample buffer + 1% Triton X-100 + 1%NP-40 Sample buffer + bromophenol blue Sample buffer 4 M urea 1.5 M KI
2.03 2.18 2.37
0.997 0.996 0.996
2.47
0.996
2.48
0.996
2.49 2.76 2.81
0.994 0.990 0.980
wavelength [nm]
Buffer
Fig. 1. Spectra of Amido Black 10B (12 µg/ml) in the acidic dissolution solution Idiss (dotted line) and in the alkaline elution solution IIelut (continuous line). Note the isosbestic point at 620 nm and the absorbance maximum of the acidic sample at 672 nm&ig.c:/f
present in, and accessible on the given protein. This allows the determination of protein in the presence of chaotropic salts and/or detergents as well as in any physical status of the protein (i.e., dissolved, suspended, in membranes of membrane bound proteins, or in tissue slices). To investigate the pH-dependence of the Amido Black absorption, spectra were run in the acidic dissolution solution (Idiss) and compared with spectra recorded in the basic elution solution (IIelut), each against the corresponding solutions alone. The concentrations of Amido Black tested was between 3,25 µg/ml and 15 µg/ml. An isosbestic point is present at 620 nm (Fig. 1), rendering absorbance values recorded at this wavelength directly and quantitatively comparable, and excluding any pH-effect. Maximal absorbance in the acidic solution Idiss is, however, found at 672 nm, exceeding the value found at 620 nm by 27%. Nevertheless, all spectrophotometric recordings documented here were taken at 620 nm for the sake of comparability. It should be emphasized, however, that reading at 672 nm will further increase the sensitivity of determinations by about 25%. BSA has been shown to be a suitable representative protein for Amido Black staining (Heinzel et al. 1965; Henkel and Bieger 1994). Proteins with extreme isoelectric points will stain better or worse, corresponding to the number of free amino groups present. If such exceptional proteins are to be determined by the Amido Black method, then that particular protein or at least a similar one should be used for the generation of a standard curve. This consideration is, however, independent of the other advantages of the Amido Black method. Quantification of protein, protein-aggregates, and suspended cells in various buffers Aliquots of samples containing protein, protein-aggregates (e.g., homogenized tissue), or whole suspended
&/tbl.:
cells can be spotted onto CA, air dried, stained and, destained by using the CA-dissolution protocol as described above. In our first series of experiments, we investigated the influence of various buffers containing large amounts of salts and/or detergents on the sensitivity of the protein determination. Influence of buffer composition on the sensitivity of the CA-dissolution protocol. &p.1:Standard dilution series ranging from 1.25–40 µg/10 µl (i.e. 0.1–4.0 mg/ml) were prepared for all eight buffers listed in the Methods section. All the resulting regression lines exhibited correlation coefficients (r2) of 0.990 or above and showed similar slopes ranging from 2.0 to 2.8 [10×OD×ml/mg] (Table 1). The sensitivity of the method therefore does not vary greatly with the chemical composition of the respective buffer, rendering the method suitable for an extremely wide range of applications. The slight variation in the response factors may be a result of the different degree of protein unfolding in the respective buffers, caused by the different denaturing power of these buffers. The excellent performance of the CA-dissolution protocol in all sample buffers is probably attributable to the removal of most, if not all, buffer components by washing with solution Iwash, which contains water and methanol in equal parts but is still not acidic enough to remove the Amido Black from the protein (pH Iwash=2.5; but Idiss
