currents
Electrophoretic method for broad MW-range proteome separation
trophoresis is run vertically, and the sample is loaded at the top of the tube; the proteins migrate downward. By positionConceived of more than 40 years ago, continuous-elution ing the tube horizontally during the separation, the researchtube-gel electrophoresis traditionally has suffered from long ers designed a simple apparatus with an opening at the top separation times and signifi of the collection chamber that cant dilution of the analytes. allowed them to easily remove Seeking an alternative to SDS100 μL fractions and replace PAGE for the fractionation of the buffer. This configuration proteomics mixtures, John Tran is an improvement over tradiand Alan Doucette of Dalhoutional continuous-elution elecsie University (Canada) recently trophoresis, in which large MW devised gel-eluted liquid fracproteins that migrate slowly are tion entrapment electrophoreincreasingly diluted during the sis (GELFrEE), a modified form collection process. of continuous-elution tube-gel A short tube gel and a relaelectrophoresis that delivers tively high voltage allowed the separated proteins in buffer team to cut down the run time. rather than in gel. Thus, sample Within 90 minutes, a 1-cmlosses are reduced because inlong cross-linked polyacrylgel digestions are not necesamide tube gel resolved prosary with this electrophoresis teins over a 10–150 kDa range method. from a bacterial proteome exProtein separation. Photo of the GELFrEE device. With GELFrEE, proteins are tract. Tran and Doucette also separated into discrete mass used a mixture of protein stanranges in a gel-filled glass tube, which empties into a coldards to show that high-mass fractions could be eluted from lection chamber that contains a molecular-weight (MW) cutthe tube gel and then analyzed by LC/MS/MS. To increase off membrane to trap the proteins. The volume of the collecthroughput, they have recently constructed an eight-column tion chamber can be controlled by the distance between the GELFrEE device that they plan to integrate with other separaend of the tube and the membrane. Normally, tube-gel election platforms. (Anal. Chem. 2008, DOI 10.1021/ac702197w)
Proteins that interact with amyloid precursor protein Amyloid precursor protein (APP) is a putative transmembrane protein that has been reported to interact with so many other proteins that researchers have a hard time deducing APP’s real physiological role. To find proteins that probably interact in vivo with APP, Gerold Schmitt-Ulms and co-workers at the University of Toronto, New York University, and the University of Alberta (Canada) performed time-controlled transcardiac perfusion cross-linking (tcTPC) on mice. With tcTPC, proteins that are interacting together essentially are frozen in time and linked together with formaldehyde, a cross-linking reagent. The researchers pump formaldehyde through the mouse’s circulatory system to fix the proteins in its tissues. To stop the crosslinking process, the target organ (brain) is removed and frozen in liquid nitrogen. Several immunoprecipitations (IPs) were conducted to discover APP interac© 2008 American Chemical Society
tors. First, an IP with an antibody against the C-terminus of APP and a mock purification were performed. However, the C-terminus of APP is similar to the C-termini of the amyloid precursor-like proteins called APLP1 and APLP2. Therefore, an IP also was performed with an antibody against a specific region of APP. Finally, IPs with APLP1- and APLP2-specific antibodies were conducted. Purified complexes were digested with trypsin and analyzed by LC/MS/MS. Known and novel interactors of APP were identified. In the next set of experiments, the researchers focused on one of these proteins, called LINGO-1, because it has been shown to bind to p75, a receptor that is processed by the same complex that processes APP. Reciprocal IPs, staining with antisense in situ hybridization probes, and siRNA knockdowns further suggest a physiological link between these proteins. The investigators say that more work must be done to figure out whether LINGO-1 and APP interact
directly. (Mol. Cell. Proteomics 2007, DOI 10.1074/mcp.M700077-MCP200)
LbL-coated memobeads for quantification in whole blood Because ELISAs (the gold-standard immunoassays) are labor-intensive, require many reagents, and detect only one protein at a time, many researchers have been investigating alternative techniques. One such method, developed by Stefaan De Smedt and colleagues at the University of Ghent and Memobead Technologies (both in Belgium), involves the use of coated, nonmagnetic microspheres called memobeads. The new multiplex immunoassay performed almost as well as ELISAs but had the added advantage of being able to detect proteins in whole blood without washes. Fluorescent polystyrene microspheres are given digital bar codes with selective photobleaching methods so that they can be distinguished in multiplexed assays. With the layer-by-layer (LbL) ap-
Journal of Proteome Research • Vol. 7, No. 2, 2008 479
currents Toolbox RefDIC To integrate transcriptomic and proteomic information on the immune system, Osamu Ohara and colleagues at the RIKEN Research Center for Allergy and Immunology and the Kazusa DNA Research Institute (both in Japan) developed a database called reference genomics database of immune cells (RefDIC). The database features quantitative mRNA profiles of human and mouse immune cells and tissues as well as quantitative protein profiles of mouse immune cells. A query interface allows researchers to visualize and analyze the data. For example, various inputs such as a gene name or probe set IDs can retrieve mRNA annotations and heat maps. Protein profiling data for multiple genes also can be visualized. Because the protein profiles are derived from 2DE analyses, users can obtain the apparent pIs and molecular masses and other information. In addition, the entire 2DE gel image can be retrieved. Finally, both mRNA and protein profiling data can be viewed simultaneously for a single gene to see whether the amounts correlate. RefDIC is accessible at http://refdic.rcai.riken.jp. (Bioinformatics 2007, 23, 2934–2941)
PHOSIDA Matthias Mann and co-workers at the Max Planck Institute of Biochemistry have developed a phosphorylation site database called PHOSIDA. The database contains thousands of high-confidence phosphosites from large-scale proteomics studies. For many entries, quantitative and time-resolved data are available. The researchers studied the structures of phosphosites and found that, in agreement with previous reports, phosphosites are in highly accessible and flexible regions of proteins. With tools in PHOSIDA, phosphoproteins from many species can be aligned, and the color of each phosphosite indicates the degree of conservation. Although regions containing phosphosites generally were less conserved than other protein regions, Mann and co-workers determined that five amino acids on either side of a phosphosite were highly conserved. Finally, a support vector machine in PHOSIDA can predict phosphosites. (Genome Biol. 2007, 8, R250)
proach, alternate coatings of oppositely charged polyelectrolytes are placed onto memobeads. These coatings help position the beads for decoding without interfering with the bar-code reading process. In addition, the layers provide many sites for antibody attachments. To characterize the LbL-coated memobeads, De Smedt and colleagues ran them through a battery of tests. Beads with antibodies against tumor necrosis factor (TNF-α) were incubated in various TNF-α solutions. The fluorescence of the beads was proportional to the TNF-α concentration and was stable for at least 20 days. When these beads were added to serum or a plasma–buffer mixture
Immunomagnetic diffractometry for biomarker detection
spiked with TNF-α, slightly lower sensitivities were observed. Next, two sets of beads with antibodies against TNF-α or P24 were added to a plasma–buffer mixture to test the multiplexing capabilities of the assay. TNF-α and P24 were specifically bound by the appropriate beads with no cross-reactivity. Finally, TNF-α and IL-2 were specifically detected in whole-blood samples spiked with TNF-α and IL-2. Unlike other immunoassays, the new test did not require any washes. The researchers say that the assay would be ideal as a microfluidic, point-of-care device because it is actually more sensitive when fewer beads are used. (Anal. Chem. 2008, 80, 80–89)
(a)
In the quest for an improved biomarker deDiffra cti tection method, Cagri +2 + on patter n 1 0 –1 –2 Savran and colleagues (b) at Purdue University and the Mayo Clinic have designed a new technique called immunomagnetic diffractometry. Their approach combines the immunomagnetic capture of an analyte on beads, Immunomagnetic diffractometry FR assay. (a) Microin situ assembly of an contact-printed F–BSA patterns; (b) FR–beads forming in optical diffraction gratsitu diffraction grating. Laser illumination yields a characing, and measurement teristic diffraction pattern dependent on the density of the attached beads. of the diffraction. In this method, magnetic beads capture the target from the serum sample and bind to a surface The researchers observed that the to form the diffraction gratings. By virFR–beads attached specifically to the tue of their size, the beads enhance the F–BSA on the surface. The packing diffraction signal, so no further signal density of the bound FR–beads intensiamplification or labeling is needed. The fied with increasing FR concentrations target chosen for the proof-of-principle (700 fM to 11 nM). After creating a calstudy was the folate receptor (FR), a ibration curve, the researchers meapotential serum biomarker for cancer. sured the FR concentration in the seWith microcontact printing, the rerum of cancer patients. The detection searchers deposited alternating 15 μm limit of immunomagnetic diffractomelines of folate-coupled bovine serum try was lower than that of several othalbumin (F–BSA) onto a gold surface. er biomarker assays such as ELISAs. Magnetic beads derivatized with the Immunomagnetic diffractometry has FR antibody (FR–beads) captured FR the additional advantages of speed, rofrom serum and subsequently bound bustness, low cost, and ease of miniato the F–BSA lines. The bound moleturization. The researchers say that the cules and beads effectively formed a assay is applicable to other biomarkgrating that diffracted the incident laers. (J. Am. Chem. Soc. 2007, 129, ser radiation. 15,824–15,829)
480 Journal of Proteome Research • Vol. 7, No. 2, 2008
currents Three-step proteomics method to analyze cells in whole saliva Many groups are investigating saliva for candidate biomarkers of oral diseases. Although several researchers have studied whole saliva for markers of oral squamous cell carcinoma (OSCC), few have focused on the cells in this fluid, and none have applied proteomics methods to the study of these cells. So, Timothy Griffin and colleagues at the University of Minnesota and the University of Arkansas for Medical Sciences decided to do just that. They developed a three-step fractionation method to analyze the cellular proteins in whole-saliva samples from OSCC patients. Human and bacterial proteins that could play a role in OSCC were identified. In previous work, Griffin’s group had applied free-flow electrophoresis (FFE) and reversed-phase (RP) LC to proteomics samples. However, they began to wonder whether the addition of another technique, strong-cation exchange (SCX) LC, would be helpful. To find out, they pooled four OSCC patient samples, pelleted the cells, and digested
Finding a function for every gene
the extracted proteins with trypsin. The peptides were separated by FFE, and then one fraction was analyzed by either RPLC/MS/MS or SCX/RP/MS/MS. To increase the confidence of peptide matches after a database search, the researchers used the peptide pI information obtained from the FFE step. The three-step method resulted in ~6× more peptide identifications and ~5× more protein identifications than that of the two-step protocol. When the pooled sample was run through the entire three-step fractionation and MS/MS procedure, the researchers identified many human proteins that were known to be involved in OSCC progression, as well as some human and bacterial proteins that also could play a role in the disease. In addition, three of the identified proteins were validated by immunoblotting whole-saliva samples from three other OSCC patients. The scientists conclude that the three-step fractionation protocol could be valuable for larger proteomics studies of cells in whole saliva. (Mol. Cell. Proteomics 2007, DOI 10.1074/mcp. M700146-MCP200)
B0 To figure out what each Fixation gene does in model organ54.7o isms, researchers knock 18 mm out the DNA sequence of the gene or use RNA interGoing for a ride. Worms like these were placed in a ference. When most genes rotor and spun to acquire 1H-HRMAS spectra. (Adapted are inactivated, however, with permission. Copyright 2007 National Academy of no major changes are deSciences, U.S.A.) tected. Therefore, Lyndon Emsley and co-workers at the Centre National de la Recherche ysis, they could distinguish the three Scientifique, Ecole Normale Supéristrains on the basis of strain-specific eure de Lyon, Université Lyon 1, and metabolic features. Institut National de L’Environnement To ensure that the sod-1 metaboIndustriel et des Risques (all in France) type was caused by the mutation, the turned to metabonomics. researchers tested several biological Spectra of C. elegans were generand analytical factors. Statistical analated by 1H high-resolution magic-anyses revealed that the effects were gle spinning (1H-HRMAS) NMR specmostly caused by biological variation troscopy. Three strains were analyzed: and that the biological factors of age wild-type; dpy-10, a collagen mutant and genetics produce distinct patwith a visible phenotype; and sod-1, a terns. Thus, strains with no observsuperoxide dismutase mutant with no able phenotypes still may have altered obvious phenotype. When Emsley and metabotypes that are detectable with 1H-HRMAS and statistical methods. co-workers applied supervised multivariate statistical modeling with par(Proc. Natl. Acad. Sci. U.S.A. 2007, tial least-squares discriminant anal104, 19,808–19,812)
Toolbox PICR Because every protein database has a different type of accession number or identifier, researchers typically have a hard time merging protein lists that were annotated with information from several sources. Although some major databases cross-reference the identifiers of other large databases, smaller repositories often are left out. Also, many existing tools for the mapping of identifiers cover a limited number of species or are labor-intensive. So, Henning Herm jakob and co-workers at the European Bioinformatics Institute (U.K.) developed the protein identifier cross-referencing service (PICR). PICR uses UniParc, which archives sequences from many databases daily, as a central data warehouse. This feature allows investigators to query many identifiers and sequences against multiple sources. The mapping algorithm in PICR finds appropriate protein entries from queried sequences on the basis of 100% sequence similarity. Accession numbers also can be queried. Then, the algorithm compiles all of the known cross-references for every entry. PRIDE and the IntAct databases are the first repositories to offer this service. (BMC Bioinformatics 2007, 8, 401)
KEGGanim To help researchers get the most out of the pathway information listed in KEGG, Jaak Vilo and colleagues at the Estonian Biocentre; the University of Tartu; and QureTec, Inc. (all in Estonia), developed KEGGanim. Although similar tools exist for other databases, no animation program was previously available for KEGG. With KEGGanim, investigators select or upload DNA microarray or proteomics data sets, select a pathway, and set the animation in motion. Colored boxes represent the levels of certain proteins or mRNAs as conditions change. For example, the animation can loop between expression levels in healthy and diseased tissue or over several time points. Vilo and colleagues say that KEGGanim will enable the community to identify important modules and key genes. The tool can be accessed at http://biit.cs.ut.ee/KEGGanim for free. (Bioinformatics 2007, DOI 10.1093/bioinformatics/btm581)
Journal of Proteome Research • Vol. 7, No. 2, 2008 481
