by Jeffrey M. Perkel
Protein activity depends on shape. Each protein has a characteristic shape, of course, which is a function of its sequence. Sometimes, though, the cell needs to tweak that shape to modulate the protein's activity. Very often, the cell accomplishes this by the addition of a phosphate group.
"Phosphorylation has for years been recognized as the major way that proteins get turned off and on," says Michael Browning, President of PhosphoSolutions, a small biotech firm in Aurora, Colorado.
Consider the glucose transporter, for instance, which pulls glucose out of blood and into cells. "You don't want it doing that all the time," Browning explains, "you only want it doing that when glucose is high. So that protein has its function regulated by the addition or subtraction of a phosphate group to a particular amino acid."
Three amino acids are often modified by the addition of a phosphate group (PO3-)—serine, threonine, and tyrosine—and this change so alters the size and charge landscape of the protein that its shape, and thus its function, changes. For many proteins, therefore, phosphorylation state is a better indicator of activity than is abundance.
A traditional method of detecting protein phosphorylation involves incubating cells with a radioactive isotope of phosphorus (32P). Following treatment with a stimulus, you lyse the cells to produce a protein extract, immunoprecipitate with an antibody to the protein of interest, run the products on a gel, and expose the gel to x-ray film. If the protein of interest lights up, it has been phosphorylated.
The method is a classic, and still used, says Natalie Ahn of the University of Colorado. "32P is very sensitive," she says. "You can often draw conclusions using 32P-labeling that you couldn't really make with other approaches."
On the other hand, many researchers are reluctant to work with 32P. And there's much the approach doesn't tell you—particularly, where on the protein the phosphate group was added. PhosphoSolutions', well, solution to this problem involves so-called phosphospecific antibodies.
As their name implies, phosphospecific antibodies target only specifically phosphorylated forms of a protein, such as CaM Kinase II with a phosphate group on threonine-305. The unphosphorylated form of CaM Kinase II (or for that matter, a differently phosphorylated form) would not react to this reagent, making it extremely effective for untangling cell signaling events.
"Once you have one of those antibodies, it can specifically bind to the protein only when it has phosphate on one specific amino acid on the protein, and that's what makes these antibodies so powerful," says Browning. "Not only does it tell you in a quantitative way how big a change there is in phosphorylation, but also it tells you in a specific way which site on the protein was actually phosphorylated. Both those pieces are very critical information and neither was possible with the prior techniques of radioactive labeling."
PhosphoSolutions offers about 150 antibodies, the bulk of which are phosphospecific and target neurological proteins. Other companies, such as Invitrogen, target a broader range of pathways. According to Jeffrey Croissant, senior marketing manager for cellular analysis at Invitrogen, the company currently offers "close to 400 unique antibody specificities for phosphorylated proteins," targeting such processes as "immunology, neuroscience, apoptosis, and cell cycle regulation."
Many of these antibodies can be used in several applications, including Western blotting, ELISAs, flow cytometry, and cell immunostaining (not every antibody will function in every assay). Yet these are mostly "single-plex" experiments—that is, each can only detect one (or as many as three or four) particular phosphorylated target. Luminex's xMAP technology, however, has no such limitation.
Like a cross between an antibody microarray and an ELISA, xMAP is entirely solution-based, performing the ELISA "sandwich" assay (which generates an antibody-antigen-antibody complex) on the surface of tiny beads. The technology enables researchers to assay for 100 different analytes in a single reaction simultaneously—in theory, at least.
"The problem with a lot of these, is you start running into antibody cross-reactivity and background," says Rick Wiese, R&D supervisor for cellular pathways at Millipore. In practice, most Luminex bead sets contain far fewer than 100 different bead types; Wiese says the biggest he's ever seen had 42. Yet xMAP still provides more bang per microliter of sample than other similar assays.
Invitrogen offers several phosphoantibody-based panels, including the "Akt Pathway Phospho 7-Plex Panel" with seven different beads. Millipore's line includes the phosphospecific "7-plex Human T Cell Receptor Signaling Kit—Phosphoprotein." According to Wiese, Millipore is set to release a 10-plex phosphospecific Luminex set for the MAPK pathway by the end of the year.
Despite their promise, there are at least three problems with phosphospecific antibodies: they are expensive, relatively hard to make, and by definition produce biased data—that is, they only probe the proteins against which they were made. For those who want a more unbiased, broader survey of the phosphorylation status of the cell—the so-called phosphoproteome—Ahn suggests another approach: mass spectrometry. "Mass spectrometry is extremely useful for large-scale studies," she says.
Browning says, "From my perspective, the two games in town [for phosphoprotein research] are mass spectrometry for discovery and phosphoantibodies for quantitative analysis."
Ahn uses an approach called "-79 Da precursor ion scanning" to probe large swaths of the phosphoproteome. Unlike more traditional approaches, which employ the instrument's positive-ion mode, this strategy uses the often more efficient phosphopeptide ionization in negative ion mode to scan for phosphorylated peptides. "Many phosphopeptides show stronger signals as negative ions than they do as positive ions," she says.
The method, first described by Steven Carr, Michael Huddleston, and Roland
Annan, works as follows. As a set of negatively charged peptide ions enters the mass spectrometer, they pass into a collision cell, where they are fragmented by collision with a gas. This collision-induced dissociation (CID) process causes the phosphate group to fall off by labile bond breakage. In positive mode, phosphate falls off as an H3PO4 neutral ion; in negative mode, PO3- is released as an ion of m/z -79. Once that signature is detected, the instrument then determines the mass of its precursor ion, switches on the fly into positive ion mode, measures the mass again, and sequences the positively charged peptide ion. The net result is to filter a complex mixture to select for only the molecules you want. "You are using the instrument to select for phosphopeptides to sequence," she explains.
As it turns out, sifting through the peptides of a complex mixture to find only those that are phosphorylated is a key challenge in phosphopeptide analysis by mass spectrometry. According to Barb Kaboord, R&D manager at Thermo Fisher Scientific, 30% of proteins in a cell may be phosphorylated at any given time. "Often, many of the proteins of interest are not abundant, so it's helpful to enrich in order to have any chance of detecting your phosphoprotein of interest by mass spec or other method," she says.
Thermo Fisher Scientific offers a pair of products specifically for this purpose, says Kaboord. The Phosphoprotein Enrichment Kit and the Phosphopeptide Enrichment Kit are both based on metal-affinity chromatography, she explains, with protocols and reagents optimized for the isolation of phosphorylated proteins or peptides, respectively.
For those interested in surveying global phosphoproteome changes without a mass spectrometer, Thermo Fisher Scientific and Invitrogen also offer phosphoprotein-specific SDS-PAGE gel stains. Though based on a similar principle, Thermo's GelCode and Invitrogen's Pro-Q Diamond stain do differ in practice: Pro-Q Diamond is a fluorescent dye, while GelCode is colorimetric.
According to Ahn, one problem with phospho-specific stains is that you cannot be certain that you are detecting only phosphorylation events. Nevertheless, these products do have a place, she says, especially if you are absolutely certain that your protein is phosphorylated, and yet cannot for whatever reason generate a phosphospecific antibody.
"I think it is useful when you need a probe for phosphorylation, and you have no better antibody probe," says Ahn, "but you have to be very careful to characterize what the staining signal means."