- Email: [email protected]
Sensitive fluorescent detection of protein on nylon membranes
J. Biochem. Biophys. Methods 51 (2002) 47 – 56 www.elsevier.com/locate/jbbm
Sensitive f luorescent detection of protein on nylon membranes Andrew Dubitsky *, Damien DeCollibus, Girolamo A. Ortolano Scientific and Laboratory Services, Pall Corporation, East Hills, NY, USA
Abstract Detection of antigen immobilized on membranes, as in Western transfers and dot enzyme linked immunosorbent assays (ELISAs), often employ antibody – enzyme conjugates and chemiluminescent or precipitated colored reaction products. Although chemiluminescent markers are sensitive, they are time-consuming because of their required exposure to X-ray film and the presence of background artifacts sometimes limits their use. This report demonstrates that direct fluorescent detection technique using nylon membranes that has higher sensitivity than chemiluminescent methods is easier to perform and has a uniform, low background. An alkaline phosphatase conjugated antibody was compared with antibody conjugated to a fluorescent phycobiliprotein (allophycocyanin) for sensitivity in both Western transfers and dot ELISA assays using mouse IgG as the membrane-bound antigen. Direct fluorescent detection of antigen – antibody complexes on positively charged nylon membrane provided better sensitivity and lower background than similar conditions using enzyme amplification and chemiluminescent detection on either nylon or PVDF membranes. Processing time was reduced by the elimination of steps associated with substrate incubation, washing and X-ray film exposures required for chemiluminescence detection. These data support the view that direct fluorescent detection can represent a significant improvement in assay sensitivity and reduction in time compared with more traditional chemiluminescent detection techniques employed in the conduct of Western transfers and dot ELISA studies. D 2002 Elsevier Science B.V. All rights reserved. Keywords: Fluorescent; Membrane; Protein; Arrays; Western
1. Introduction Common membrane-based protein assays include Western transfers and dot enzymelinked immunosorbent assays (ELISAs). Recent advances in proteomics indicate that *
Correponding author. 25 Harbor Park Drive, Port Washington, NY 11050, USA. E-mail address: Andrew [email protected] (A. Dubitsky).
0165-022X/02/$ - see front matter D 2002 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 5 - 0 2 2 X ( 0 1 ) 0 0 2 4 3 - 3
48
A. Dubitsky et al. / J. Biochem. Biophys. Methods 51 (2002) 47–56
Table 1 Fluorescent properties of equivalent phycobiliprotein and cyanine fluors Fluorophore
mol. wt.
Absorption max
Emission max
Extinction coefficient (cm 1 M 1)
Allophycocyanin Cy5
104,000 975
650 635
660 670
700,000 250,000
protein arrays, similar to gene expression arrays, will become a common tool for genetics research and drug discovery [1]. Therefore, the need for more sensitive, faster and less expensive protein identification and quantitative approaches is anticipated to increase. Choices for detection of antibody– antigen recognition in these assays have traditionally been enzyme systems, using chemiluminescent or chromogenic reactants [2– 7]. Fluorescent detection systems are becoming more popular. Commonly used in flow cytometry and DNA analysis [8], long wavelength excitation fluorophores can also be directly detected on membranes [1,9]. Phycobiliproteins have been documented to have the highest extinction coefficients of all available fluorophores [1]. Allophycocyanin (APC) is a phycobiliprotein with maximal excitation at 635 nm and is compatible with scanners designed for Cy5. A comparison of APC and Cy5, based on manufacturer’s specifications, is shown in Table 1. The phycobiliproteins produce a very bright signal in comparison to low molecular weight fluorophores. Their usefulness for tagging nucleic acids is probably limited by their high molecular weight. The size of the molecule is much less important when preparing conjugates of other large proteins, making them ideal labels for ultra-sensitive protein assays [9]. Pall nylon membranes have low autofluorescent background above 600 nm (unpublished observations) so that there is minimal interference with detection of far red fluorophores. This report further characterizes our preliminary observations of the potential value of combining fluorescence and nylon membrane toward improving sensitivity of detection of proteins on membranes.
2. Materials and methods 2.1. Immobilization Mouse IgG (Rockland Immunochemicals, Gilbertsville, PA) was used as the test antigen in all experiments. For Western transfers, mouse IgG was diluted in phosphate buffered saline (PBS; 20 mM PO4, 150 mM NaCl, pH 7) and sample buffer (Novex, San Diego, CA) and added to a non-reducing gel (4– 12% NuPAGE polyacrylamide, Novex) at 2 mg to 6 ng per lane. The protein was electrophoresed for 35 min at 200 V in a Novex Minicell II apparatus, using NuPAGE MES running buffer (Novex). After electrophoresis, proteins were transferred to nylon (BiodyneR Plus membrane, Pall Corporation, East Hills, NY) or PVDF membranes (FluoroTransR W PVDF membrane, Pall Corporation) in
A. Dubitsky et al. / J. Biochem. Biophys. Methods 51 (2002) 47–56
49
the same apparatus, using 25 mM Tris, 150 mM glycine, 20% methanol transfer buffer. Transfer proceeded for 1 h at 25 V. Dot blots were performed using antigen applied to Biodyne Plus nylon membrane or FluoroTrans W PVDF membrane using a PlateMateR Liquid Handling Station (Matrix Technologies, Lowell, MA) fitted with a 96 pin transfer tool. Pins apply 200 nl liquid, creating spots approximately 1 mm in diameter. Antigen concentration varied from 200 to 0.06 ng per spot. Each pin printed to four sites on the membrane, creating 384 spot patterns from 96 well master plates. Microarrays were printed with an eight pin microarray printer (Xenopore Corporation, Hawthorne, NJ) which transfers approximately 6 nl per spot. The amount of antigen varied from 17 to 0.2 pg per spot. Glycerol was added to the antigen solutions to a final concentration of 2% to aid in liquid transfer. Spots are approximately 500 mm in diameter, with 1 mm spacing. 2.2. Immune detection Non-specific binding sites were blocked before detection by immersing all membranes into a solution comprised of 0.3% Hammersten grade casein (Gallard Schlessinger, Westbury, NY) dissolved in PBS. Membranes were incubated for 30 min at room temperature and then transferred to detection solutions. Fluorescence detection was accomplished by incubating membranes for 1 h at room temperature with 1 mg/ml Goat anti-mouse IgG conjugated with allophycocyanin (Prozyme, San Leandro, CA) diluted in PBS with 0.05% Hammersten grade casein. Membranes were imaged in a StormR fluorescent scanner (Molecular Dynamics, Sunnyvale, CA) immediately after the incubation step. Membranes were then rinsed with PBS for five minutes and scanned again. Chemiluminescence detection was accomplished by immersing membranes for 1 h at room temperature into a solution of goat anti-mouse IgG conjugated with alkaline phosphatase (Sigma, St. Louis, MO) diluted 1:100,000 in PBS containing 0.05% Tween 20 and 0.2% Hammersten casein (PBST-C). After conjugate incubation, membranes were washed three times PBST-C (5-min immersions per wash). Membranes were then soaked in substrate buffer (0.1 M Tris, pH = 9.5) for 5 min followed by Lumiphos 480 substrate solution (Lumigen, Detroit, IL) for 30 min. Following this, excess liquid was removed and the membranes were transferred to clear plastic folders (Life Technologies, Bethesda, MD). Autoradiographs were prepared using Hyperfilmk MP (Amersham API, Indianapolis, IN).
3. Results Our reference method for Western transfers uses PVDF membranes and chemiluminescence (Fig. 1, panel A). For this experiment, Biodyne Plus nylon membrane was also used with chemiluminescence, resulting in slightly lower signal level than the PVDF membrane. In comparison, direct fluorescent detection with the APC conjugate provides better sensitivity and resolution on the nylon membrane. Signal with this system and the PVDF
50
A. Dubitsky et al. / J. Biochem. Biophys. Methods 51 (2002) 47–56
Fig. 1. Western transfer of mouse IgG onto PVDF and nylon membranes. Bands were detected with antibody enzyme conjugate and chemiluminescent substrate (panel A) or with antibody – PCA conjugate and direct fluorescence (panel B).
A. Dubitsky et al. / J. Biochem. Biophys. Methods 51 (2002) 47–56
51
membrane was low (Fig. 1, panel B). Overall sensitivity (major band only) in this experiment was similar for both chemiluminescent and fluorescent detection. A large increase in resolution of minor bands in the gel is evident with fluorescent detection. At high load concentrations, bloom from strong chemiluminescent signals obscured signal from minor bands on the PVDF membrane. In contrast, resolution of the minor bands using fluorescence on nylon membrane remained clear over the entire range of concentrations used.
Fig. 2. Effect of washing nylon membranes after Western transfer and incubation with fluorescent conjugate. Challenge antigen doses as in Fig. 1. Membranes were scanned immediately after the conjugate step (panel A), after two rinses in PBS (panel B) and again after air drying (panel C).
52
A. Dubitsky et al. / J. Biochem. Biophys. Methods 51 (2002) 47–56
Fig. 3. Protein dot blot. Serial dilutions of mouse IgG printed on nylon and PVDF membranes. Spot volumes were 200 nl deposited with a replica pin printer. Comparison of direct fluorescent and chemiluminescent detection on nylon membranes is shown in panel A and on PVDF membrane in panel B.
A. Dubitsky et al. / J. Biochem. Biophys. Methods 51 (2002) 47–56
53
Fig. 4. Protein microarrays printed on nylon membrane. Spot volume was approximately 6 nl deposited with a microarray pin printer. Spot size is 500 mm with 1 mm spacing. Protein concentration varied from 100 to 0.25 pg/ml, resulting in 600 – 3 fg antigen per spot. Chemiluminescent detection with 1/100,000 dilution of conjugate is shown in panel A. Fluorescent detection with 1 mg/ml of APC conjugate is shown in panel B.
54
A. Dubitsky et al. / J. Biochem. Biophys. Methods 51 (2002) 47–56
The results of employing washing protocols with nylon membranes and fluorescent detection are shown in Fig. 2. Although absolute background noise decreases after washing and air drying the membranes, these protocols did not did not appreciably improve signal to noise, sensitivity or resolution. Air-drying the membranes before scanning was not necessary. Dot blots on nylon and PVDF membranes are shown in Fig. 3. Results were consistent with the Western transfer. Using nylon membrane, sensitivity was better with direct fluorescent detection, at 0.06 pg, compared to 0.6 pg for chemiluminescent detection. There was a significant amount of bloom (spreading signal) at higher antigen concentrations with chemiluminescence that was not present using the direct fluorescent method. A protein microarray on Biodyne Plus nylon membrane is shown in Fig. 4. Sensitivity and background levels were again better with fluorescent than chemiluminescent detection. Sensitivity in this experiment was 3 fg per spot for fluorescence, compared to 30 fg per spot for chemiluminescence.
4. Discussion Chemiluminescent enzyme substrates have become a standard tool for high sensitivity detection of immobilized protein [3]. Our reference method for protein electro blotting employs FluoroTrans W PVDF membrane with chemiluminescent detection. Autofluorescence of certain nylon membranes decreases rapidly when excitation wavelength exceeds 600 nm (unpublished data), suggesting that nylon membranes are well suited to fluorescence detection with far-red fluorophores. As PVDF membranes also have low fluorescence in this region, we compared performance of the APC phycobiliprotein fluorescent conjugate on both membranes. Despite similar low background, signal levels were much higher on the nylon membrane. Resolution of closely spaced bands in Western transfers was improved with fluorescent detection. Precipitating substrates used for chemiluminescence tend to produce a bloom around high intensity signals, which could result from excess precipitate migrating away from the actual enzyme site or photons hitting the film distant to the point of origin. This bloom enlarges and distorts the shape of the original band or dot. Bloom from strong chemiluminescent signals also limits the use of chemiluminescence in high throughput screening of proteins or nucleic acids, due to strong signal or associated background artifacts obscuring weaker adjacent dots. This is especially true for highdensity membrane arrays, which can be printed with spots as small as 100 mm and with spacing of less than 150 mm [10]. In addition to better sensitivity and resolution, direct fluorescent detection also offers direct electronic signal acquisition, quicker completion, greater ease of use and higher throughput. This can be clearly seen by comparing the steps required for the chemiluminescent and fluorescent methods used for this report (Table 2). Sensitive detection of proteins with another phycobiliprotein fluorophore, PBXL-1, has been described for Westerns and dot blots on nitrocellulose membranes [9]. Maximal excitation wavelength for PBXL-1 is 540 – 560 nm, while the APC fluorophore used in our
A. Dubitsky et al. / J. Biochem. Biophys. Methods 51 (2002) 47–56
55
Table 2 Steps required for either direct fluorescent or chemiluminescent detection Direct fluorescent detection
Chemiluminescent detection
(1) (2) (3) (4)
(1) Block with casein solution (2) Incubate with conjugate (3) Wash 2 with PBS (4) Incubate with substrate (5) Expose to autoradiography film (6) Develop and fix film (7) Wash and dry film (8) Optimize film exposure with additional development and wash steps (9) Scan final film for image analysis
Block with casein solution Incubate with conjugate Rinse Scan in fluorescent imager
experiments is optimally excited at 635 nm. This suggests an opportunity for two-color detection, using antibodies tagged with either PBXL-1 or APC. Results indicate that the APC conjugate can be substituted for antibody conjugates used for immunodetection of proteins. Direct fluorescent detection on Biodyne Plus membrane should provide highest possible sensitivity and resolution along with low level homogeneous background. Since enzymes are not involved in the detection process, variations due to enzyme – substrate kinetics are eliminated. Direct fluorescent detection also requires fewer assay steps than other non-radioactive detection methods and facilitates electronic image analysis. The data presented illustrate that the combination of a positively charged nylon membrane and fluorescent detection affords increased sensitivity and improved resolution compared to conventional enzyme based techniques. This platform is particularly well suited for Western transfers, dot blots and high throughput screening methods.
References [1] Patton WF. A thousand points of light: the application of fluorescence detection technologies to two-dimensional gel electrophoresis and proteomics. Electrophoresis 2000;21:1123 – 44. [2] Alba FJ, Daban JR. Rapid fluorescent monitoring of total protein patterns on sodium dodecyl sulfatepolyacrylamide gels and western blots before immunodetection and sequencing. Electrophoresis 1998;19: 2407 – 11. [3] Bakkali L, Guillou R, Gonzague M, Cruciere C. A rapid and sensitive chemiluminescence dot-immunobinding assay for screening hybridoma supernatants. J Immunol Methods 1994;170:177 – 84. [4] Egger D, Bienz K. Protein (western) blotting. Mol Biotechnol 1994;3:289 – 305. [5] Feodorova VA, Devdariani ZL. Development, characterisation and diagnostic application of monoclonal antibodies against Yersinia pestis fibrinolysin and coagulase. J Med Microbiol 2000;49:261 – 9. [6] Hu GR, Harrop P, Warlow RS, Gacis ML, Walls RS. High resolution autoantibody detection by optimized protein G-horseradish peroxidase immunostaining in a western blotting assay. J Immunol Methods 1997;202:113 – 21. [7] Khan AS, Sears JF, Muller J, Galvin TA, Shahabuddin M. Sensitive assays for isolation and detection of simian foamy retroviruses. J Clin Microbiol 1999;37:2678 – 86.
56
A. Dubitsky et al. / J. Biochem. Biophys. Methods 51 (2002) 47–56
[8] Shapiro HM, Glazer AN, Christenson L, Williams JM, Strom TB. Immunofluorescence measurement in a flow cytometer using low-power helium-neon laser excitation. Cytometry 1983;4:276 – 9. [9] Morseman JP, Moss MW, Zoha SJ, Allnutt FC. PBXL-1: a new fluorochrome applied to detection of proteins on membranes. Biotechniques 1999;26:559 – 63. [10] Chen JW, et al. Profiling expression patterns and isolating differentially expressed genes by cDNA microarray system with colorimetry detection. Genomics 1998;51:313 – 24.
Sensitive f luorescent detection of protein on nylon membranes Andrew Dubitsky *, Damien DeCollibus, Girolamo A. Ortolano Scientific and Laboratory Services, Pall Corporation, East Hills, NY, USA
Abstract Detection of antigen immobilized on membranes, as in Western transfers and dot enzyme linked immunosorbent assays (ELISAs), often employ antibody – enzyme conjugates and chemiluminescent or precipitated colored reaction products. Although chemiluminescent markers are sensitive, they are time-consuming because of their required exposure to X-ray film and the presence of background artifacts sometimes limits their use. This report demonstrates that direct fluorescent detection technique using nylon membranes that has higher sensitivity than chemiluminescent methods is easier to perform and has a uniform, low background. An alkaline phosphatase conjugated antibody was compared with antibody conjugated to a fluorescent phycobiliprotein (allophycocyanin) for sensitivity in both Western transfers and dot ELISA assays using mouse IgG as the membrane-bound antigen. Direct fluorescent detection of antigen – antibody complexes on positively charged nylon membrane provided better sensitivity and lower background than similar conditions using enzyme amplification and chemiluminescent detection on either nylon or PVDF membranes. Processing time was reduced by the elimination of steps associated with substrate incubation, washing and X-ray film exposures required for chemiluminescence detection. These data support the view that direct fluorescent detection can represent a significant improvement in assay sensitivity and reduction in time compared with more traditional chemiluminescent detection techniques employed in the conduct of Western transfers and dot ELISA studies. D 2002 Elsevier Science B.V. All rights reserved. Keywords: Fluorescent; Membrane; Protein; Arrays; Western
1. Introduction Common membrane-based protein assays include Western transfers and dot enzymelinked immunosorbent assays (ELISAs). Recent advances in proteomics indicate that *
Correponding author. 25 Harbor Park Drive, Port Washington, NY 11050, USA. E-mail address: Andrew [email protected] (A. Dubitsky).
0165-022X/02/$ - see front matter D 2002 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 5 - 0 2 2 X ( 0 1 ) 0 0 2 4 3 - 3
48
A. Dubitsky et al. / J. Biochem. Biophys. Methods 51 (2002) 47–56
Table 1 Fluorescent properties of equivalent phycobiliprotein and cyanine fluors Fluorophore
mol. wt.
Absorption max
Emission max
Extinction coefficient (cm 1 M 1)
Allophycocyanin Cy5
104,000 975
650 635
660 670
700,000 250,000
protein arrays, similar to gene expression arrays, will become a common tool for genetics research and drug discovery [1]. Therefore, the need for more sensitive, faster and less expensive protein identification and quantitative approaches is anticipated to increase. Choices for detection of antibody– antigen recognition in these assays have traditionally been enzyme systems, using chemiluminescent or chromogenic reactants [2– 7]. Fluorescent detection systems are becoming more popular. Commonly used in flow cytometry and DNA analysis [8], long wavelength excitation fluorophores can also be directly detected on membranes [1,9]. Phycobiliproteins have been documented to have the highest extinction coefficients of all available fluorophores [1]. Allophycocyanin (APC) is a phycobiliprotein with maximal excitation at 635 nm and is compatible with scanners designed for Cy5. A comparison of APC and Cy5, based on manufacturer’s specifications, is shown in Table 1. The phycobiliproteins produce a very bright signal in comparison to low molecular weight fluorophores. Their usefulness for tagging nucleic acids is probably limited by their high molecular weight. The size of the molecule is much less important when preparing conjugates of other large proteins, making them ideal labels for ultra-sensitive protein assays [9]. Pall nylon membranes have low autofluorescent background above 600 nm (unpublished observations) so that there is minimal interference with detection of far red fluorophores. This report further characterizes our preliminary observations of the potential value of combining fluorescence and nylon membrane toward improving sensitivity of detection of proteins on membranes.
2. Materials and methods 2.1. Immobilization Mouse IgG (Rockland Immunochemicals, Gilbertsville, PA) was used as the test antigen in all experiments. For Western transfers, mouse IgG was diluted in phosphate buffered saline (PBS; 20 mM PO4, 150 mM NaCl, pH 7) and sample buffer (Novex, San Diego, CA) and added to a non-reducing gel (4– 12% NuPAGE polyacrylamide, Novex) at 2 mg to 6 ng per lane. The protein was electrophoresed for 35 min at 200 V in a Novex Minicell II apparatus, using NuPAGE MES running buffer (Novex). After electrophoresis, proteins were transferred to nylon (BiodyneR Plus membrane, Pall Corporation, East Hills, NY) or PVDF membranes (FluoroTransR W PVDF membrane, Pall Corporation) in
A. Dubitsky et al. / J. Biochem. Biophys. Methods 51 (2002) 47–56
49
the same apparatus, using 25 mM Tris, 150 mM glycine, 20% methanol transfer buffer. Transfer proceeded for 1 h at 25 V. Dot blots were performed using antigen applied to Biodyne Plus nylon membrane or FluoroTrans W PVDF membrane using a PlateMateR Liquid Handling Station (Matrix Technologies, Lowell, MA) fitted with a 96 pin transfer tool. Pins apply 200 nl liquid, creating spots approximately 1 mm in diameter. Antigen concentration varied from 200 to 0.06 ng per spot. Each pin printed to four sites on the membrane, creating 384 spot patterns from 96 well master plates. Microarrays were printed with an eight pin microarray printer (Xenopore Corporation, Hawthorne, NJ) which transfers approximately 6 nl per spot. The amount of antigen varied from 17 to 0.2 pg per spot. Glycerol was added to the antigen solutions to a final concentration of 2% to aid in liquid transfer. Spots are approximately 500 mm in diameter, with 1 mm spacing. 2.2. Immune detection Non-specific binding sites were blocked before detection by immersing all membranes into a solution comprised of 0.3% Hammersten grade casein (Gallard Schlessinger, Westbury, NY) dissolved in PBS. Membranes were incubated for 30 min at room temperature and then transferred to detection solutions. Fluorescence detection was accomplished by incubating membranes for 1 h at room temperature with 1 mg/ml Goat anti-mouse IgG conjugated with allophycocyanin (Prozyme, San Leandro, CA) diluted in PBS with 0.05% Hammersten grade casein. Membranes were imaged in a StormR fluorescent scanner (Molecular Dynamics, Sunnyvale, CA) immediately after the incubation step. Membranes were then rinsed with PBS for five minutes and scanned again. Chemiluminescence detection was accomplished by immersing membranes for 1 h at room temperature into a solution of goat anti-mouse IgG conjugated with alkaline phosphatase (Sigma, St. Louis, MO) diluted 1:100,000 in PBS containing 0.05% Tween 20 and 0.2% Hammersten casein (PBST-C). After conjugate incubation, membranes were washed three times PBST-C (5-min immersions per wash). Membranes were then soaked in substrate buffer (0.1 M Tris, pH = 9.5) for 5 min followed by Lumiphos 480 substrate solution (Lumigen, Detroit, IL) for 30 min. Following this, excess liquid was removed and the membranes were transferred to clear plastic folders (Life Technologies, Bethesda, MD). Autoradiographs were prepared using Hyperfilmk MP (Amersham API, Indianapolis, IN).
3. Results Our reference method for Western transfers uses PVDF membranes and chemiluminescence (Fig. 1, panel A). For this experiment, Biodyne Plus nylon membrane was also used with chemiluminescence, resulting in slightly lower signal level than the PVDF membrane. In comparison, direct fluorescent detection with the APC conjugate provides better sensitivity and resolution on the nylon membrane. Signal with this system and the PVDF
50
A. Dubitsky et al. / J. Biochem. Biophys. Methods 51 (2002) 47–56
Fig. 1. Western transfer of mouse IgG onto PVDF and nylon membranes. Bands were detected with antibody enzyme conjugate and chemiluminescent substrate (panel A) or with antibody – PCA conjugate and direct fluorescence (panel B).
A. Dubitsky et al. / J. Biochem. Biophys. Methods 51 (2002) 47–56
51
membrane was low (Fig. 1, panel B). Overall sensitivity (major band only) in this experiment was similar for both chemiluminescent and fluorescent detection. A large increase in resolution of minor bands in the gel is evident with fluorescent detection. At high load concentrations, bloom from strong chemiluminescent signals obscured signal from minor bands on the PVDF membrane. In contrast, resolution of the minor bands using fluorescence on nylon membrane remained clear over the entire range of concentrations used.
Fig. 2. Effect of washing nylon membranes after Western transfer and incubation with fluorescent conjugate. Challenge antigen doses as in Fig. 1. Membranes were scanned immediately after the conjugate step (panel A), after two rinses in PBS (panel B) and again after air drying (panel C).
52
A. Dubitsky et al. / J. Biochem. Biophys. Methods 51 (2002) 47–56
Fig. 3. Protein dot blot. Serial dilutions of mouse IgG printed on nylon and PVDF membranes. Spot volumes were 200 nl deposited with a replica pin printer. Comparison of direct fluorescent and chemiluminescent detection on nylon membranes is shown in panel A and on PVDF membrane in panel B.
A. Dubitsky et al. / J. Biochem. Biophys. Methods 51 (2002) 47–56
53
Fig. 4. Protein microarrays printed on nylon membrane. Spot volume was approximately 6 nl deposited with a microarray pin printer. Spot size is 500 mm with 1 mm spacing. Protein concentration varied from 100 to 0.25 pg/ml, resulting in 600 – 3 fg antigen per spot. Chemiluminescent detection with 1/100,000 dilution of conjugate is shown in panel A. Fluorescent detection with 1 mg/ml of APC conjugate is shown in panel B.
54
A. Dubitsky et al. / J. Biochem. Biophys. Methods 51 (2002) 47–56
The results of employing washing protocols with nylon membranes and fluorescent detection are shown in Fig. 2. Although absolute background noise decreases after washing and air drying the membranes, these protocols did not did not appreciably improve signal to noise, sensitivity or resolution. Air-drying the membranes before scanning was not necessary. Dot blots on nylon and PVDF membranes are shown in Fig. 3. Results were consistent with the Western transfer. Using nylon membrane, sensitivity was better with direct fluorescent detection, at 0.06 pg, compared to 0.6 pg for chemiluminescent detection. There was a significant amount of bloom (spreading signal) at higher antigen concentrations with chemiluminescence that was not present using the direct fluorescent method. A protein microarray on Biodyne Plus nylon membrane is shown in Fig. 4. Sensitivity and background levels were again better with fluorescent than chemiluminescent detection. Sensitivity in this experiment was 3 fg per spot for fluorescence, compared to 30 fg per spot for chemiluminescence.
4. Discussion Chemiluminescent enzyme substrates have become a standard tool for high sensitivity detection of immobilized protein [3]. Our reference method for protein electro blotting employs FluoroTrans W PVDF membrane with chemiluminescent detection. Autofluorescence of certain nylon membranes decreases rapidly when excitation wavelength exceeds 600 nm (unpublished data), suggesting that nylon membranes are well suited to fluorescence detection with far-red fluorophores. As PVDF membranes also have low fluorescence in this region, we compared performance of the APC phycobiliprotein fluorescent conjugate on both membranes. Despite similar low background, signal levels were much higher on the nylon membrane. Resolution of closely spaced bands in Western transfers was improved with fluorescent detection. Precipitating substrates used for chemiluminescence tend to produce a bloom around high intensity signals, which could result from excess precipitate migrating away from the actual enzyme site or photons hitting the film distant to the point of origin. This bloom enlarges and distorts the shape of the original band or dot. Bloom from strong chemiluminescent signals also limits the use of chemiluminescence in high throughput screening of proteins or nucleic acids, due to strong signal or associated background artifacts obscuring weaker adjacent dots. This is especially true for highdensity membrane arrays, which can be printed with spots as small as 100 mm and with spacing of less than 150 mm [10]. In addition to better sensitivity and resolution, direct fluorescent detection also offers direct electronic signal acquisition, quicker completion, greater ease of use and higher throughput. This can be clearly seen by comparing the steps required for the chemiluminescent and fluorescent methods used for this report (Table 2). Sensitive detection of proteins with another phycobiliprotein fluorophore, PBXL-1, has been described for Westerns and dot blots on nitrocellulose membranes [9]. Maximal excitation wavelength for PBXL-1 is 540 – 560 nm, while the APC fluorophore used in our
A. Dubitsky et al. / J. Biochem. Biophys. Methods 51 (2002) 47–56
55
Table 2 Steps required for either direct fluorescent or chemiluminescent detection Direct fluorescent detection
Chemiluminescent detection
(1) (2) (3) (4)
(1) Block with casein solution (2) Incubate with conjugate (3) Wash 2 with PBS (4) Incubate with substrate (5) Expose to autoradiography film (6) Develop and fix film (7) Wash and dry film (8) Optimize film exposure with additional development and wash steps (9) Scan final film for image analysis
Block with casein solution Incubate with conjugate Rinse Scan in fluorescent imager
experiments is optimally excited at 635 nm. This suggests an opportunity for two-color detection, using antibodies tagged with either PBXL-1 or APC. Results indicate that the APC conjugate can be substituted for antibody conjugates used for immunodetection of proteins. Direct fluorescent detection on Biodyne Plus membrane should provide highest possible sensitivity and resolution along with low level homogeneous background. Since enzymes are not involved in the detection process, variations due to enzyme – substrate kinetics are eliminated. Direct fluorescent detection also requires fewer assay steps than other non-radioactive detection methods and facilitates electronic image analysis. The data presented illustrate that the combination of a positively charged nylon membrane and fluorescent detection affords increased sensitivity and improved resolution compared to conventional enzyme based techniques. This platform is particularly well suited for Western transfers, dot blots and high throughput screening methods.
References [1] Patton WF. A thousand points of light: the application of fluorescence detection technologies to two-dimensional gel electrophoresis and proteomics. Electrophoresis 2000;21:1123 – 44. [2] Alba FJ, Daban JR. Rapid fluorescent monitoring of total protein patterns on sodium dodecyl sulfatepolyacrylamide gels and western blots before immunodetection and sequencing. Electrophoresis 1998;19: 2407 – 11. [3] Bakkali L, Guillou R, Gonzague M, Cruciere C. A rapid and sensitive chemiluminescence dot-immunobinding assay for screening hybridoma supernatants. J Immunol Methods 1994;170:177 – 84. [4] Egger D, Bienz K. Protein (western) blotting. Mol Biotechnol 1994;3:289 – 305. [5] Feodorova VA, Devdariani ZL. Development, characterisation and diagnostic application of monoclonal antibodies against Yersinia pestis fibrinolysin and coagulase. J Med Microbiol 2000;49:261 – 9. [6] Hu GR, Harrop P, Warlow RS, Gacis ML, Walls RS. High resolution autoantibody detection by optimized protein G-horseradish peroxidase immunostaining in a western blotting assay. J Immunol Methods 1997;202:113 – 21. [7] Khan AS, Sears JF, Muller J, Galvin TA, Shahabuddin M. Sensitive assays for isolation and detection of simian foamy retroviruses. J Clin Microbiol 1999;37:2678 – 86.
56
A. Dubitsky et al. / J. Biochem. Biophys. Methods 51 (2002) 47–56
[8] Shapiro HM, Glazer AN, Christenson L, Williams JM, Strom TB. Immunofluorescence measurement in a flow cytometer using low-power helium-neon laser excitation. Cytometry 1983;4:276 – 9. [9] Morseman JP, Moss MW, Zoha SJ, Allnutt FC. PBXL-1: a new fluorochrome applied to detection of proteins on membranes. Biotechniques 1999;26:559 – 63. [10] Chen JW, et al. Profiling expression patterns and isolating differentially expressed genes by cDNA microarray system with colorimetry detection. Genomics 1998;51:313 – 24.