Bioscience Reports, Vol. 25, Nos. 1/2, February/April 2005 (Ó 2005) DOI: 10.1007/s10540-005-2845-1

Protein Detection Methods in Proteomics Research Reiner Westermeier1,3 and Rita Marouga2

In proteomics research chemical as well as physical methods are used to detect proteins subsequently to their separation. Physical methods are mostly applied after chromatography. They are either based on spectroscopy like light absorption at certain wavelengths or mass determination of peptides and their fragments with mass spectrometry. Chemical methods are used after two-dimensional electrophoresis and employ staining with organic dyes, metal chelates, fluorescent dyes, complexing with silver, or pre-labeling with fluorophores. In some cases autoradiography is still used. Since all of these techniques are very different in terms of sensitivity, their usefulness for quantitative determinations varies significantly. This review will describe the various protein detection methods applied to electrophoresis gels. KEY WORDS: Blotting; protein staining; protein labeling; two-dimensional electrophoresis; quantification; DIGE. ABBREVIATONS: 2-D: two-dimensional; CCD: charge coupled device; DIGE: difference gel electrophoresis; EDTA: Ethylenediaminetetraacetic acid; ICAT: isotope coded affinity tag; MALDI ToF: matrix assisted laser desorption ionization time of flight; NMR: nuclear magnetic resonance; PTM: post-translational modification; PAS: periodic acid/Schiff’s reagent; PVDF: Polyvinylidenefluoride; SDS: sodium dodecyl sulphate; Tris: Tris(hydroxymethyl)-aminoethane.

INTRODUCTION The selection of the appropriate protein detection method is highly important, because the proteomics approach measures quantitative changes in expression levels in biological samples. A number of very different techniques exist. Ideally the detection limit should be as low as possible with an optimal signal to noise ratio. For proper quantification of proteins in typical proteomics samples the detection method should have a wide dynamic range and a wide linear relationship between the quantity of protein and the staining intensity. The procedure should be easy and fast to perform, non-toxic, environment-friendly, mass spectrometry-compatible, and not too expensive. At present no protein detection technique meets all these demands. It is therefore necessary to select the method which is optimal according to the biological question, sample type and other methodical requirements.

1

GE-Healthcare BioSciences, Munzinger Strasse 9, D-79111, Freiburg, Germany. GE-Healthcare BioSciences, Bjo¨rkgatan 30 S-751 82, Uppsala, Sweden. 3 To whom correspondence should be addressed. E-mail: [email protected] 19 2

0144-8463/05/0400-0019/0 Ó 2005 Springer Science+Business Media, Inc.

20

Westermeier and Marouga

PROTEIN STAINING IN POLYACRYLAMIDE GELS After two-dimensional (2-D) electrophoresis the proteins are detected in a SDS polyacrylamide gel with small pore sizes, and the proteins are unfolded. Under this condition it is easy to prevent the proteins from diffusion within and out of the matrix. Fixing with highly concentrated trichloroacetic acid is therefore not necessary.

Coomassie Brilliant Blue Staining Since its introduction by Fazekas de St. Groth et al. (1963) staining with the organic dye Coomassie Brilliant Blue is still the most frequently employed method for protein detection in polyacrylamide electrophoresis gels. This anionic triphenylmethane dye is used in two modifications: Coomassie R-250 (Red tint) and Coomassie G-250 (Green tint), which has two additional methyl groups. In presence of an acidic medium these dyes stick to the amino groups of the proteins by electrostatic and hydrophobic interactions. In the original protocols the gels were placed for a certain time into a staining solution of 0.1% Coomassie dye dissolved in 45% (v/v) methanol, 45% (v/v) water and 10% acetic acid; the background was destained with 25% (v/v) methanol, 65% (v/v) water and 10% acetic acid. Sometimes methanol was replaced by ethanol or isopropanol. Practice has shown that this procedure is not reproducible and is not very reliable for quantification. Because of the presence of alcohol some of the proteins release the dye during the background destaining process. Some proteins, like for instance collagen, destain faster than the polyacrylamide gel. Colloidal Coomassie Blue staining shows much better reproducibility and can reach very high sensitivity, when it is applied long enough or repeatedly. There is no background destaining. Because it is an endpoint method, highly reproducible results are obtained. The most established procedure is the improved protocol by Neuhoff et al. (1988). First the gel is placed in fixing solution (1.3% (w/v) o-phosphoric acid and 20% methanol (v/v) in water) for 1 hr. The staining solution is prepared by slowly adding 100 g of ammonium sulfate to 980 ml of a 2% H3PO4 (w/v) solution until it has completely dissolved. Then a solution of 1 g Coomassie Brilliant Blue G-250 solution in 20 ml of water is added. This staining solution must not be filtered, but shaken before use. Staining is carried out overnight on an orbital shaker with 160 ml of staining solution plus 40 ml of methanol, which is added during staining. Then it is placed into a neutralization solution, composed of 0.1 mol/l Tris/H3PO4 buffer, pH 6.5, for 3 min. It is rinsed for a maximum of 1 min in 25% (v/v) aqueous methanol. The protein–dye complex is stabilized in 20% aqueous ammonium sulfate for one day and then staining and the other steps are repeated for two more times. The gel can remain in staining or stabilizing solution for several days without any effect on the results. For optimal sensitivity with a detection limit of several nanogram per spot the total procedure takes about 1 week. The modification according to Anderson et al. (1991) is less labor-intensive, but should give comparable results regarding sensitivity. First the gel is fixed for at least 3 hr in 50% (v/v) ethanol/3% (w/v) phosphoric acid and then washed three times for 29 min in water. After pre-incubation for 1 hr in 34% (v/v) methanol/3% (w/v) phosphoric acid/17% (w/v) ammonium sulfate solution, 0.53 g Coomassie Brilliant

Protein Detection Methods

21

Blue G-250 powder is added per 1.5 l solution. The gel is stained for 4–5 days on an orbital shaker whereby the end-point is reached after 4–5 days. Subsequently, the gel is washed a few times in water to remove background staining. On 1.5 mm gels a detection limit of about 15 ng per spot is reached. Another useful procedure is fast hot Coomassie staining with 0.025% (w/v) Coomassie R-350 dissolved in 10% (v/v) acetic acid. Either it is performed in a metal tray with a lid heated to 50°C for 15 min or the staining solution is heated to 90°C, poured over the gel in a stainless steel tray with a lid. This tray covered with a lid is placed on a laboratory shaker for 10 min. Destaining is performed in a tray on a rocking table in 10% acetic acid for at least 2 hr at room temperature. Either the solution is changed several times, recycled by pouring it through a filter filled with activated charcoal, or a paper towel is placed into the destaining solution to absorb the Coomassie dye. Staining and destaining solutions can be used repeatedly; however, the sensitivity is not as high as with the two colloidal procedures. The advantages of Coomassie staining methods are the quantitative binding of the dye to proteins, the low price and the good reproducibility. For mass spectrometry analysis the dye is removed from the gel plug with bicarbonate prior to tryptic digestion. When a protein is detectable with Coomassie Brilliant Blue, as a rule of thump, enough protein is present for appropriate mass spectrometry analysis with MALDI ToF. The disadvantages are the long staining times, the relatively low sensitivity, and the narrow dynamic range. Copper Staining Although staining with 0.3 M copper chloride solution according to Lee et al. (1987) is faster, easier, and more sensitive than Coomassie staining, it has never become well accepted for applications in proteomics research. Negative Imidazol SDS Zinc Staining For the matter of further analysis of the detected proteins it is appealing to apply a detection method, which does not modify the protein but visualizes the background. A complex salt of zinc, SDS and imidazole precipitates in the polyacrylamide gel and forms an opaque white background, but stays soluble in presence of a protein spot (Fernandez-Patron et al., 1992). The protein zones become visible against a dark background. The sensitivity of detection (which is at a few nanograms) lies between Coomassie Blue and silver staining. The peptide yield after in-gel digestion is said to be higher then after direct staining of proteins. The salt precipitate can be dissolved with a Tris–EDTA buffer, which allows electrophoretic transfer of the proteins for Western blotting. Although it is claimed that the procedure allows quantitative determination of proteins (Ferreras et al., 1993), this statement is disputed because only the background is detected directly, but not the protein. Silver Staining Detecting proteins with silver staining is widely used, because the sensitivity is below 1 ng per spot, the costs for reagents are relatively low, and—although it is a

22

Westermeier and Marouga

multistep procedure—the results are available relatively quickly. Ideally the spots are dark brown to black on a light beige background. Since the introduction of silver staining in polyacrylamide gels by Merrill et al. (1979) about one hundred different modifications of silver staining protocols have been published. An extensive review of silver staining methods has been written by Rabilloud et al. (1994). In principle, the protocols can be divided into three groups: Colloidal fast staining in presence of tungstosilicic acid based on the first silver staining protocol in agarose gels according to Kerenyi and Gallyas (1972). These protocols have the lowest sensitivity of the three groups, but are compatible with mass spectrometry. The principle of a silver nitrate staining protocol works as follows: First the proteins are fixed; SDS and buffer components are washed out with 40% ethanol and 10% acetic acid. The next step combines cross linking of the proteins in the matrix with glutardialdehyde and sensitizing with sodium thiosulfate in presence of sodium acetate. After a number of washes with distilled water with exact timing, the gel is placed into a solution containing 0.25% silver nitrate and a low amount of formaldehyde. Development is performed with formaldehyde at basic pH value, which is achieved with sodium carbonate. The reaction is stopped by placing the gel either into 5% acetic acid or 1.5% EDTA. In practice the following two protocols (according to Heukeshoven and Dernick, 1985; Blum et al., 1987) work best and show the highest sensitivities (down to 100 pg per spot). When spots need to be analyzed further with mass spectrometry, the glutardialdehyde must be omitted from the sensitizer solution and the formaldehyde from the silver nitrate solution (Shevchenko et al., 1996). In order to obtain efficient staining, the development step

Fig. 1. 2-D electrophoresis gel stained with the silver nitrate method according to Heukeshoven and Dernick. Protein extract from Brassica oleracea (cauliflower) seeds separated in an immobilized pH gradient 4–7 over a distance of 24 cm followed by SDS polyacrylamide gel electrophoresis with 18 cm separation distance.

Protein Detection Methods

23

is then performed with the double amount of formaldehyde compared to standard protocols. This modification reduces the sensitivity to about one fourth. Figure 1 shows a 2-D electrophoresis gel of a protein extract from cauliflower seeds stained with silver nitrate according to Heukeshoven and Dernick (1985). Ammoniacal silver complex protocols look in principle like this: Fixing is often combined with the sensitizer, which in this case is naphthalene disulfonate. After several washing steps the gel is placed into ammoniacal silver solution. The development is performed under acidic conditions in citric acid with formaldehyde. The reaction is stopped with acetic acid and ethanolamine. This procedure was introduced by Eschenbruch and Bu¨rk (1982) and further improved by many different groups, mainly by Rabilloud (1992). This method is critical for forming silver precipitation on the gel surface (called ‘‘mirrors’’) caused by traces of impurities in one of the solutions, and needs a much higher amount of silver nitrate than the silver nitrate based protocols. In general, silver staining detects the proteins mainly on the gel surface. This is also the reason why further analysis of silver stained proteins is still possible. Silver staining is a multistep procedure. Because it is not an endpoint procedure, the staining intensities can vary from gel to gel. The use of a staining automat makes the procedure much more convenient and increases its reproducibility substantially. Unfortunately silver staining shows only a narrow dynamic range and is not reliable for quantification. An often observed phenomenon is that highly abundant proteins produce spots with a yellow centre. As shown in Fig. 2, these spots look like volcano craters or donuts, and create problems for image analysis software for qualitative and quantitative analysis. This effect can be repaired when the gel is submitted to blue toning after silver staining (Berson, 1983). This procedure further increases the sensitivity of silver staining.

Fig. 2. Effect of high protein concentration in some spots on silver staining. (a) 2-D view of a gel section. (b) 3-D view of the same section. The centres of these spots are yellow and lighter than the edges.

24

Westermeier and Marouga

Fluorescence Staining Fluorescence dyes show very wide linear dynamic ranges, over four orders of magnitude. The results are highly reproducible, because fluorescent staining is an endpoint method. Since the introduction of Nile Red staining (Bermudez et al., 1994) various dyes have been developed during the last few years with markedly different sensitivities and properties such as the family of SyproÒ dyes (Orange, Red, Tangerine and Ruby) (Steinberg et al., 1996; Berggren et al., 2002). Whereas the sensitivities of SyproÒ Orange, Red, and Tangerine are similar to Coomassie Blue, SyproÒ Ruby can detect down to about 1 ng of protein in a spot. Rabilloud et al. (2001) published a protocol for a ruthenium complex fluorescence staining with high sensitivity. A new dye, called Deep PurpleÔ, contains the fluorophore ‘‘epicocconone’’ from the fungus Epicoccum nigrus (Mackintosh et al., 2003) and is even more sensitive than SyproÒ Ruby, down to a few hundred picograms. Besides the high sensitivity, the latter dye has some more advantages: it does not create speckles in the background and is a natural product, which is easy to dispose of. Fluorescence staining is the optimum method for combining high sensitivity of detection with reliable quantification over a wide linear dynamic range. However, the dyes are more expensive than those used for visual staining; a fluorescence scanner is needed to visualize the spots and to acquire the image for image analysis (see section image acquisition below). Figure 3 shows the comparison of silver stained and fluorescence-labeled 2-D electrophoresis gels. At the first glance the fluorescence gel seems to show less resolution than the silver stained gel. However, if the peaks are displayed in a 3-D view as shown in Fig. 4, the wide dynamic range of the detection method becomes obvious. Specific Stains for Post-translational Modifications For studies of the two most important post-translational modifications, glycosylation and phosphorylation, specific stains can be used. The periodic acid Schiff’s

Fig. 3. Comparison of an image of a fluorescent (left hand side) and silver stained gel (right hand side). The silver staining gel image makes believe, that it has a higher resolution than fluorescent staining. Only the 3-D view (see Fig. 4) will reveal the high quality result of the fluorescence staining, which is due to its wide dynamic range.

Protein Detection Methods

25

Fig. 4. Two- (insert) and 3-D view of a few protein spot detected with fluorescent labeling.

reagent procedure for specific detection of glycoproteins, introduced by Zacharius et al. (1969) has become much more sensitive by coupling the reaction with fluorescence (Eckhardt et al., 1976). Recently, a commercial product called ‘‘Pro-QÒ-Emerald’’ (Hart et al., 2003) became available and a further fluorescence dye called ‘‘Pro-QÒ-Diamond’’ was introduced for specific detection of phosphorylated proteins (Steinberg et al., 2003). It is possible to stain a gel sequentially with different specific dyes and finally with a total stain. PRE-LABELING OF PROTEINS WITH FLUOROPHORES In 2-D electrophoresis, proteins can be pre-labeled with a fluorescent dye after the isoelectric focusing step prior to SDS polyacrylamide gel electrophoresis (Urwin and Jackson, 1991) or prior to isoelectric focusing with monobromobimane (Urwin and Jackson, 1993). The spots can be directly scanned in the gel (while it is still in the cassette) when low-fluorescent glass is used. Like other fluorescent dyes, also this label exhibits a high sensitivity and a wide dynamic range of the signal. A quantum leap in 2-D gel electrophoresis methodology has been achieved by the introduction of size and charge matched cyanine dyes (CyDyeÔ DIGE Fluors) with different excitation and emission wavelengths as protein labels for

26

Westermeier and Marouga

different samples (U¨nlu¨ et al., 1997). The method, called ‘‘difference gel electrophoresis’’ (DIGE), allows multiplexing of samples and the use of an internal standard, which is created by using one of the labels for a pooled mixture of all samples. This method leads to highly accurate qualitative and quantitative results, because gel-to-gel variations are eliminated. Figure 5 shows the workflow of difference gel electrophoresis. Two alternative concepts exist: minimum labeling by attaching the dye to the free amino group of lysine residues and saturation labeling of all cysteine residues. In the first variant only 3–5% of the total proteins will receive a label, which insures that only singly labeled proteins will be detected. In this case the labeled proteins co-migrate with the non-labeled proteins in both separation directions and the resulting 2-D image looks like the one achieved with post-staining. Typically, 400 qmol/l of one of the three available cyanine dyes is added to 50 lg protein of each sample and the pooled standard. The sensitivity of detection is comparable to a sensitive silver staining method. With the second variant a considerable increase in sensitivity is obtained. 2-D patterns resulting from down to 1 lg total protein load can be visualized. Because all available cysteine residues are labeled, many multiply labeled proteins exist. The resulting spot patterns are different from those achieved with post-stained or lysine labeled proteins. After scanning the gel at different wavelengths, the comigrated protein spots of different sample origins are co-detected using dedicated software, which allows very fast and completely automatic pattern evaluation thus avoiding any bias introduced by an operator. With the help of the pattern created by the internal standard, which is run in each gel, the patterns of different gels can easily be matched and the sample spot volumes can be normalized to the spot volume of the standard. This results in very accurate protein difference ratios with high statistical confidence (Alban et al., 2003; Friedman et al., 2004).

Fig. 5. Schematic representation of 2-D difference gel electrophoresis in one gel. Example for two samples. When more samples are to be compared, more gels are run, every time including the internal standard, which is a mixture of pooled aliquots of all samples.

Protein Detection Methods

27

ISOTOPE LABELING Radioactive Labeling In the pioneering paper by O´Farrell (1975) on high-resolution 2-D electrophoresis it was shown that autoradiography is 100–1000 times more sensitive than staining with Coomassie Brilliant Blue. The labeling is carried out predominantly in living cells, in bacteria, yeast or tissue cultures with the help of 35S-methionine and/ or 32P orthophosphate. It is also possible to apply specific labeling, for instance with 32 PO3) for the detection of phosphoproteins. Post-synthetical radiolabeling, for 4 instance by iodination with 125I or alkylation of cysteine residues with 14C- or 3 H-iodoacetamide is problematic due to problems with reproducibility and sensitivity. Because the wet gel layer quenches the radioactivity signal, the gels need to be dried prior to exposure to an X-ray film or a storage phosphor screen. Fluorography Wet gels can be used and the sensitivity can be increased with scintillation autoradiography, also called fluorography. The gel with 35S, 14C- or 3H labeled proteins is impregnated after fixation with a fluor like 2,5-diphenyloxacole. The low-energy b-particles excite the fluor molecules, which will then emit a light signal. Even though very high sensitivity of detection seems to be desirable, these methods are not very useful for proteomics applications because the spots of low expression do not contain enough protein for subsequent mass spectrometry analysis. Furthermore, because of environmental reasons, autoradiography and fluorography are becoming less important. Stable Isotope Labeling Stable isotope labeling methods for the analysis of quantitative differences between two samples can only be used when a high-resolution mass spectrometer is available. For such a differential approach, cells can be labeled by adapting techniques developed for NMR by using ammonium salt as nitrogen source with 14N vs. 15N or glucose as carbon source for 12C vs. 13C labeling. The detection is carried out by mass spectrometry of the tryptic digests of protein spots. Because ionization is equal for the same peptides containing the different isotopes, the heights of the shifted peaks can be used to measure quantitative ratios (Oda et al., 1999). For stable isotope tagging of proteins after synthesis, isotope coded affinity tag (ICAT) labeling, known from differential mass spectrometry analysis, has also successfully been applied for 2-D electrophoresis (Smolka et al., 2002). BLOTTING TECHNIQUES Alternatively to gel-based methods, proteins can be detected on immobilizing membranes, to which they have been transferred after electrophoretic separation. This transfer is usually carried out with electrophoretic force, because diffusion or capillary forces are not sufficiently efficient to mobilize the proteins from inside a polyacrylamide gel (Towbin et al., 1979). For electro blotting either the blotting

28

Westermeier and Marouga

gel-membrane sandwich is clamped between grids and inserted into a buffer tank with platinum electrode wires at the walls (‘‘tank blotting’’) or the blotting sandwich is placed between two horizontal plate electrodes made from graphite or platinumcoated steel meshes (‘‘semi-dry blotting’’). The membrane material used is mostly nitrocellulose or PVDF (Polyvinylidenefluoride). Both membrane types are available with different pore sizes and mechanical strengths, and they have different binding properties. Two technical transfer problems can limit the performance of blotting methods: Some proteins do not transfer efficiently, particularly large ones (>80,000 Da) and other proteins are not bound efficiently by the membrane and migrate through, mostly small ones. There are several possibilities of protein detection on blotting membranes: Total Protein Staining General protein staining of blotted proteins is mainly applied for locating and confirming the spot position and identity in connection with special protein probing or further protein analysis. Usually it is not employed as an alternative to staining of proteins in the gel, because it is too labor-intensive and bears too many risks of inefficient transfer. If a high sensitivity is not required, the reversible Ponceau red staining according to Salinovich and Montelaro (1986) is the standard procedure, showing a detection limit in the range of 1 lg/spot. Amido Black, Coomassie Brilliant Blue, Fast Green FCF, and Indian ink are also used for staining blotting membranes. The most sensitive techniques are colloidal gold (Moeremans et al., 1985) and Deep PurpleÔ staining with detection limits down to 3 ng/spot. SyproÒ Rose Plus and SyproÒ Ruby fluorescent dyes show sensitivities of about 12 ng/spot. It should be noted that during the blotting transfer from the separation gel to the membrane some proteins can get lost, either due to incomplete transfer or non-quantitative binding properties of the membrane. The protein detection limits given above are measured from the amount of protein loaded to the gel; they do not mean the actual amount on the blotting membrane. Special Protein Staining The specific gel stains for phosphoprotein and glycoprotein detection can also be applied on blotting membranes. The steps are much shorter than in gels, because the proteins are easily accessible on the surface of the membrane. Specific glycan detection kits for blotting membranes are offered by Roche Applied Sciences: glycoproteins are probed with digoxigenin, which is then probed with alkaline phosphatase labeled antidigoxigenin. The visualization is performed by colorimetric detection of the alkaline phosphatase with Nitroblue tetrazolium. According to the brochure by Roche, the sensitivity of detection lies in the range of several nanograms per protein band, dependent on the individual protein analyzed. This technique can be combined with selected lectins to probe for specific glycoconjugate structures. Immuno Blotting When electrophoretic protein blotting is followed by immunodetection, it is frequently called ‘‘Western Blotting’’. After the transfer the free binding sites on the

Protein Detection Methods

29

membrane must be blocked with a mixture of macromolecular substances, which do not participate in the visualization reaction. Often skim milk powder, casein hydrolysate, or fish gelatin is used. The probing for individual proteins is performed with immunoglobulins or monoclonal antibodies. Up to four times reprobing for different antigens is possible. For visualization a secondary antibody is used, which is conjugated either to gold, horseradish peroxidase, alkaline phosphatase or an avidin--biotin-enzyme complex for amplification of the signal. The conjugated enzymes are detected with the respective functional zymogram staining, which has a high sensitivity. The highest sensitivity without using radioactivity is obtained with enhanced chemiluminescence detection methods (Laing, 1986). The signal is detected by exposing the membrane to an X-ray film or placing it into a video documentation system for a certain time period, which is in principle a CCD camera in a dark cabinet. With a further developed chemiluminescence method (ECL AdvanceÔ) down to 0.1 femtogram of antigen can be detected. The blotting membranes can also be probed with a non-antibody protein to study specific protein–protein interactions (Burgess et al., 2000). The probing protein is then detected with a labeled antibody. This variant is either called ‘‘Far-Western blotting’’ or protein overlay detection. Protein Identification Blotting on PVDF or modified glass fiber membranes has been widely used for protein identification by amino acid composition analysis, N-terminal amino acid sequencing, peptide mass fingerprinting and sequencing with mass spectrometry. With the extended possibilities offered by new mass spectrometry technologies the blotting step can be omitted; proteins are digested inside the gel layer and the eluted peptides are submitted to mass spectrometry analysis. IMAGE ACQUISITION The complex protein patterns obtained in proteomics research require computeraided image analysis. Therefore the image needs to be converted into digital format. Camera Systems For chemiluminescence detection usually a camera in a dark cabinet is employed. When all proteins need to be detected, only scanning cameras can provide enough resolution for an appropriate image analysis. However, the acquired adjacent image patches need to be stitched together electronically, which can be a source of errors in the qualitative and quantitative evaluation. Scanners (Densitometers) Point and line scanners are much better suited for the acquisition of high-resolution protein patterns. Gels with visible protein zones need to be scanned with light transmission in order to obtain quantitative reliable signals. This is possible with high-end, liquid-insulated desktop line scanners. Before scanning the instrument must be calibrated. For appropriate image analysis is important to scan

30

Westermeier and Marouga

the gels in grayscale mode with at least 16 bit signal depth. For fluorescence staining or labeling special fluorescence scanners are needed. Multiple fluorescence scanners are usually point scanners with laser light sources with different wavelengths. The best instruments use confocal optics in order to exclude stray light signals and to get reliable quantitative signals with high resolution. It should also be noted that the wide dynamic range, obtained with a detection method, can only be maintained when the scanner offers a wide dynamic range as well. Multiple fluorescence scanners can also be used to measure radioactivity with the usage of storage phosphor screens. IMAGE ANALYSIS Software for image analysis should evaluate protein patterns reliably and reproducibly. Human interference must be kept to a minimum, because of two major reasons: image analysis with manual adjustments can be a severe bottleneck in the workflow and it is an additional source of variation. The main steps are spot detection, normalization and spot matching between different gels, and finally quantitative comparisons of spot volumes. The significance of changes in protein expression levels detected in different gels, and the confidence level of a measured quantitative ratio are checked with statistic tools, which are usually supplied with professional image analysis software. Before results are published, the protein spots of interest must be further analyzed. It can always happen that an electrophoresis band or spot contains more than one protein due to insufficient resolution. In such cases the measured spot volume ratios will not reflect the real situation. Usually this analysis is performed with mass spectrometry.

REFERENCES Alban, A., David, S., Bjorkesten, L., Andersson, C., Sloge, E., Lewis, S., and Currie, I. (2003) A novel experimental design for comparative two-dimensional gel analysis: two-dimensional difference gel electrophoresis incorporating a pooled internal standard. Proteomics. 3:36–44. Anderson, N. L., Esquer-Blasco, R., Hofmann, J.-P., and Anderson, N. G. (1991) A two-dimensional gel database of rat liver proteins useful in gene regulation and drug effect studies. Electrophoresis. 12:907– 930. Berggren, K. N., Schulenberg, B., Lopez, M. F., Steinberg, T. H., Bogdanova, A., Smejkal, G., Wang, A., and Patton, W. F. (2002) An improved formulation of SYPRO Ruby protein gel stain: comparison with the original formulation and with a ruthenium II tris (bathophenanthroline disulfonate) formulation. Proteomics. 2:486–498. Bermudez, A., Daban, J.-R., Garcia, J. R., and Mendez, E. (1994) Direct blotting, sequencing and immunodetection of proteins after five-minute staining of SDS and SDS-treated IEF gels with Nile Red. Biotechniques. 16:621–624. Berson, G. (1983) Silver staining of proteins in polyacrylamide gels: increased sensitivity by a blue toning Anal. Biochem. 134:230–234. Blum, H., Beier, H., and Gross, H.J. (1987) Improved silver staining of plant proteins, RNA and DNA in polyacrylamide gels. Electrophoresis. 8:93–99. Burgess, R., Arthur, T. M., and Pietz, B. C. (2000) Mapping protein–protein interaction domains using ordered fragment ladder far-Western analysis of hexahistidine-tagged fusion proteins. Meth. Enzymol. 328:141–157. Eckhardt, A. E., Hayes, C. E., and Goldstein, I. J. (1976) A sensitive fluorescent method for the detection of glycoproteins in polyacrymamide gels. Anal. Biochem. 73:192–197.

Protein Detection Methods

31

Eschenbruch, M. and Bu¨rk, R. R. (1982) Experimentally improved reliability of ultrasensitive silver staining of protein in polyacrylamide gels. Anal. Biochem. 125:96–99. Fazekas de St Groth, S., Webster, R. G., and Datyner, A. (1963) Two new staining procedures for quantitative estimation of proteins in electrophoresis strips. Biochim. Biophys. Acta. 71:377–391. Fernandez-Patron, C., Rodriguez, P., and Castellanos-Serra, L. (1992) Reverse staining of sodium dodecyl sulfate polyacrylamide gels by imidazole-zinc salts: sensitive detection of unmodified proteins. BioTechniques. 12:564–573. Ferreras, M., Gavilanes, J. G., and Garcia-Segura, J. M. (1993) A permanent Zn2+ reverse staining method for the detection and quantification of proteins in polyacrylamide gels. Anal Biochem. 213:206–212. Friedman, D. B., Hill, S., Keller, J. W., Merchant, N. B., Levy, S. E., Coffey, R. J., and Caprioli, R. M. (2004) Proteome analysis of human colon cancer by two-dimensional difference gel electrophoresis and mass spectrometry. Proteomics. 4:793–811. Hart, C., Schulenberg, B., Steinberg, T. H., Leung, W. Y., and Patton, W. F. (2003) Detection of glycoproteins in polyacrylamide gels and on electroblots using Pro-Q Emerald 488 dye,, a fluorescent periodate Schiff-base stain. Electrophoresis. 24:588–598. Heukeshoven, J. and Dernick, R. (1985) Simplified method for silver staining of proteins in polyacrylamide and the mechanism of silver staining. Electrophoresis. 6:103–112. Kerenyi, L. and Gallyas, F. (1972) A highly sensitive method for demonstrating proteins in electrophoretic, immunoelectrophoretic and immunodiffusion preparations. Clin. Chim. Acta. 38:465–467. Laing, P. (1986) Luminescent visualization of antigens on blots J. Immunol. Methods. 92:161–165. Lee, C., Levin, A., and Branton, D. (1987) Copper staining: a five-minute protein stain for sodium dodecyl sulfate-polyacrylamide gels. Anal. Biochem. 166:308–312. Mackintosh, J. A., Choi, H.-Y., Bae, S.-H., Veal, D. A., Bell, P. J., Ferrari, B. C., Van Dyk, D. D., Verrills, N. M., Paik, Y.-K., and Karuso, P. (2003) A fluorescent natural product for ultra sensitive detection of proteins in one-dimensional and two-dimensional gel electrophoresis. Proteomics. 3:2273–2288. Merril, C. R., Switzer, R. C., and Van Keuren, M. L. (1979) Trace polypeptides in cellular extracts and human body fluids detected by two-dimensional electrophoresis and a highly sensitive silver stain. Proc. Natl. Acad. Sci. USA. 76:4335–4339. Moeremans, M., Daneels, G., and De Mey, J. (1985) Sensitive colloidal metal (gold or silver) staining of proteins blots on nitrocellulose membranes. Anal. Biochem. 145:315–321. Neuhoff, V., Arold, N., Taube, D., and Ehrhardt, W. (1988) Improved staining of proteins in polyacrylamide gels including isoelectric focusing gels with clear background at nanogram sensitivity using Coomassie Brilliant Blue G-250 and R-250. Electrophoresis. 9:255–262. Oda, Y., Huang, K., Cross, F. R., Cowburn, D., and Chait, B. T. (1999) Accurate quantitation of protein expression and site specific phosphorylation. Proc. Natl. Acad. Sci. USA. 96:6591–6596. O’Farrell, P. H. (1975) High-resolution two-dimensional electrophoresis of proteins. J. Biol. Chem. 250: 4007 4021. Rabilloud, T. (1992) A comparison between low background silver diammine and silver nitrate protein stains Electrophoresis. 13:429–439. Rabilloud, T., Vuillard, L., Gilly, C., and Lawrence, J. J. (1994) Silver staining of proteins in polyacrylamide gels: a general overview. Cell. Mol. Biol. 40:57–75. Rabilloud, T., Strub, J.-M., Luche, S., van Dorsselaer, A., and Lunardi, J. (2001) A comparison between Sypro Ruby and ruthenium II tris (bathophenanthroline disulfonate) as fluorescent stains for protein detection in gels. Proteomics. 1:699–704. Salinovich, O. and Montelaro, R. C. (1986) Reversible staining and peptide mapping of proteins transferred to nitrocellulose after separation by sodium dodecyl sulfate-polyacrylamide gel electrophoresis. Anal. Biochem. 156:341–347. Shevchenko, A., Wilm, M., Vorm, O., and Mann, M. (1996) Mass spectrometric sequencing of proteins from silver stained polyacrylamide gels. Anal. Chem. 68:850–858. Smolka, M., Zhou, H., and Aebersold, R. (2002) Quantitative protein profiling using two-dimensional gel electrophoresis,, isotope coded affinity tag labeling and mass spectrometry. Mol. Cell. Proteomics. 1:19–29. Steinberg, T. H., Hangland, R. P., and Singer, V. I. (1996) Applications of SYPRO Orange and SYPRO Red protein gel stains. Anal. Biochem. 239:238–245.

32

Westermeier and Marouga

Steinberg, T. H., Agnew, B. J., Gee, K. R., Leung, W.-Y., Goodman, T., Schulenberg, B., Hendrickson, J., Beechem, J. M., Haugland, R. P., and Patton, W. F. (2003) Global quantitative phosphoprotein analysis using multiplexed proteomics technology. Proteomics. 3:1128–1144. Towbin, H., Staehelin, T., and Gordon, J. (1979) Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications. Proc. Natl. Acad. Sci. USA. 76:4350– 4354. U¨nlu¨, M., Morgan, M. E., and Minden, J. S. (1997) Difference gel electrophoresis: a single gel method for detecting changes in protein extracts. Electrophoresis. 18:2071–2077. Urwin, V. and Jackson, P. (1991) Labeling after IEF prior to SDS PAGE. Anal. Biochem. 195:30–37. Urwin, V. and Jackson, P. (1993) Two-dimensional polyacrylamide gel electrophoresis of proteins labeled with the fluorophore monobromobimane prior to first-dimensional isoelectric focusing: imaging of the fluorescent protein spot patterns using a cooled charge-coupled device. Anal. Biochem. 209:57–62. Zacharius, R. M., Zell, T. E., Morrison, J. H., and Woodlock, J. J. (1969) Glycoprotein staining following electrophoresis on acrylamide gels. Anal. Biochem. 30:148–152.