by Jeffrey M. Perkel
It's been nearly 30 years since W. Neal Burnette first coined the term "Western blotting" to describe the electrophoretic transfer of protein gels onto nitrocellulose membranes1, and in many ways the technique remains exactly the same. The gel transfer "sandwich" Burnette describes in detail, for instance, would be familiar to anyone running a Western blot today.
Yet much also is different. Where Burnette detected murine leukemia virus proteins using radioactive iodine-labeled protein A (which binds antibody molecules), today's researchers can take their pick of far safer methods.
Perhaps the most popular is chemiluminescent detection, in which an enzyme-conjugated secondary antibody (that is, an antibody that targets another, primary, antibody, which itself targets the protein of interest) drives a light-generating reaction, which can be detected either on film or in an imager.
Exceptionally sensitive and popular, chemiluminescence nevertheless suffers from two major drawbacks. The first is that it is, "at best," semi-quantitative, says Steve Shiflett, technical product manager for Thermo Fisher Pierce Protein Detection Products, precisely because of its reliance on enzyme kinetics.
Additionally, and perhaps more importantly, chemiluminescence is a single-plex reaction, says Peter Chiang, product manager at Cell Biosciences, a company that sells Western blot imagers. "Traditional chemiluminescence gives off light, [but] you cannot distinguish light from light." As a "one-color" technology, chemiluminescence cannot be used to detect multiple overlapping proteins simultaneously.
Enter fluorescent Western blotting.
Thanks to an extensive color palate of available fluorochromes, researchers can probe multiple proteins (typically two or three) simultaneously, enabling easy quantification by comparing one protein's abundance to an internal control. Alternatively, by using both phospho-specific and "pan" antibodies, researchers can measure the fraction of a given protein that is phosphorylated, something that is not otherwise possible without first stripping and reprobing the blot, because the masses of the two isoforms are so close together.
"If you want to multiplex and normalize your signal with a housekeeping protein, you can only do that reliably with fluorescence," observes Åsa Hagner McWhirter, a senior scientist at GE Healthcare.
Fluorescence is also more quantitative than other approaches. Chemiluminescent signal intensity varies with substrate incubation time and exposure time, but that is not true of fluorescence.
David Chimento, antibody validation group leader at Rockland Immunochemicals, whose team runs both fluorescent and chemiluminescent Westerns, says, "It's very difficult to compare [chemiluminescence] experiments. Small differences during setup, such as timing, freshness of reagents, etc., can change the total light signal output, whereas with fluorescence, it's fairly constant."
That's because a given fluorophore will produce a specific amount of light – no more, no less. Results can therefore be compared across experiments, even those separated in time by many months. "When using a stable fluorophore, I can do a blot, image it, store it for three months, and image it again, and be pretty confident that the signal will be the same," Chimento says.
The problem with chemiluminescence can be even worse for those researchers who use film rather than imaging systems, because film has a relatively restricted linear dynamic range; if a researcher "blows out" or overexposes the signal in the film in the vicinity of a highly abundant protein, it becomes impossible to compare that abundance with another protein in the same experiment, especially if they differ greatly. By contrast, fluorescent applications typically have a relatively wide linear dynamic range—about three orders of magnitude, according to Chiang. "And because you have a broader linear dynamic range, [fluorescence] is more quantitative and more accurate," he says.
Though the Western blotting workflow is more or less identical regardless of detection method, there are products specifically designed for fluorescence applications—the blotting membrane itself, for instance. Most standard nitrocellulose and PVDF filters autofluoresce under excitation illumination.
"It's important to pick a membrane with very low inherent fluorescent properties," says Ryan Short, imaging marketing manager for Bio-Rad Laboratories. "If you don't … the fluorescence from your secondary antibody may be drowned out by the background fluorescence."
A number of companies offer low-fluorescence blotting membranes, including Millipore's Immobilon FL (recommended by LI-COR for use with its Odyssey imagers); Thermo Scientific’s Low-Fluorescence PVDF Transfer Membrane; and GE Healthcare's HyBond LFP (PVDF) membranes.
Another potential source of autofluorescence is blocking buffer. Millipore's Blok-FL blocking reagent was "specifically formulated for fluorescence," says Sahar Sibani, product manager in Millipore's protein detection group. A protein-free formulation (as opposed to common blocking reagents, such as bovine serum albumin or powdered non-fat milk), this buffer is stable at room temperature, says Sibani.
Other fluorescence-optimized blocking reagents include Rockland Immunochemicals' protein-based Blocking Buffer for Fluorescent Western Blotting, LI-COR's non-mammalian-protein Odyssey® Blocking Buffer, and GE Healthcare's ECL Advanced Blocking Reagent.
You'll also need fluorescently labeled antibodies. Usually, it's the secondary antibodies that are labeled, but labeled primaries may also be used.
Thermo Scientific DyLight Western Blotting Kits contain fluorescently labeled secondary antibodies to rabbit and mouse primary antibodies, as well as fluorescence-friendly blotting filters; one kit uses fluorophores in the red/far-red region of the spectrum, the other, near infrared.
Similarly, GE Healthcare's ECL Plex CyDye-conjugated antibodies include mouse and rabbit secondaries labeled with the company's Cy2, Cy3, and Cy5 fluorophores, and users can add homemade CyDye-labeled primaries, as well, says McWhirter. (Antibody labeling kits are available from a variety of vendors, including GE Healthcare.)
However, if researchers plan to use homemade labeled antibodies, McWhirter recommends using them for internal controls, such as housekeeping proteins. "As you will be labeling your primary antibody, it is most appropriate to use them for the detection of a protein that is relatively abundant, since this skips the signal amplification generated by usage of a secondary antibody," she says.
For all its advantages, there is one metric by which fluorescence doesn't measure up to chemiluminescence: sensitivity. Chimento estimates that chemiluminescence "is maybe 10-to-100-times more sensitive," meaning it might be a better choice for extremely low abundance proteins. Recently, however, GE Healthcare developed a protocol for "three-layer probing," offering somewhat increased sensitivity, according to McWhirter.
Typical Westerns are so-called "two-layer" assays: the primary antibodies are not labeled, but the secondary antibodies are (thus, there are two layers of reagents on the blot). In a three-layer experiment, one secondary is labeled not with a fluorophore but with biotin. This is then detected by a third reagent—the third layer—which in this case is either fluorescently tagged streptavidin or a fluorescent anti-biotin antibody.
According to McWhirter, the resulting 15-fold boost in sensitivity stems from the signal amplification that occurs as each successive layer is applied. "The third layer adds more signal," she explains, "because you can bind many streptavidin molecules per antibody and many CyDyes [fluorophores] per streptavidin."
The final element required for fluorescence Western blotting is an imager. Laser-based scanners with photomultiplier tube detectors and CCD camera-based epi-illumination systems are available, including GE Healthcare's Typhoon FLA 9000 and LI-COR's Odyssey (laser-based), and Bio-Rad's VersaDoc, Cell Biosciences' FluorChem Q, and GE Healthcare's ImageQuant LAS 4000 (camera-based).
In general, such systems are not limited to fluorescence Western blotting imaging, and consideration should be given to what other applications the lab needs when making a purchase. Bio-Rad's VersaDoc, for instance, can handle chemiluminescent, fluorescent, and colorimetric Westerns, as well as fluorescent gel imaging. Similarly, GE Healthcare's Typhoon FLA 9000 can read fluorescently labeled two-dimensional protein gels, for example, as well as store phosphor screens for imaging radioactive gels.
Other useful tools for fluorescent Westerns include labeled protein markers (from GE Healthcare and Thermo Fisher Scientific), a stripping buffer formulated for fluorescent blots (Thermo Fisher Scientific), and even chemicals designed to enhance protein binding to membranes, which can improve signal (Millipore's ChemiLucent Plus).
Whatever particular combination of samples, antibodies, buffers, and membranes you choose, though, your best bet is to optimize your experiment first. "That's a general truth of Western blotting," says Margaret Dentlinger, marketing product manager at LI-COR Biosciences.
Bio-Rad's Criterion Stain Free Gel Imaging System uses specially formulated, dye-infused pre-cast gels to enable researchers to assess the efficiency of protein transfer before committing precious antibodies to it. Coomassie blue is incompatible with Western blotting, whereas stains such as Ponceau S Red can label the blot, but not the gel. With Criterion, says Short, using only the stain in the gel itself "you can check the protein load, check the transfer efficiency, and then you can decide if you really want to go through the Western blotting process."
Once you've made that decision, LI-COR's MPX Multiplexer Blotting System enables users to tweak up to 24 conditions simultaneously in a gasket-sealed multiplexing incubator, so they can get the most fluorescent bang for their buck.
"It's a very cool little tool," Dentlinger says. "It's a quick way to optimize without using too much material."
1W.N. Burnette, "“Western Blotting”: Electrophoretic transfer of proteins from sodium dodecyl sulfate-polyacrylamide gels to unmodified nitrocellulose and radiographic detection with antibody and radioiodinated protein A," Anal. Biochem, 112:195-203, 1981.