High-throughput method for mammalian protein function
Endogenous yeast genes are so easy to tag, it’s almost like magic. Mammalian genes are much more difficult to tag, however. They can be more complex and require more regulatory sequences than yeast genes do. Current methods for mammalian tagging involve the time-consuming and costly generation of complementary DNA transgenes that do not include all of the necessary sequences for wild-type regulation. Thus, the tagged transgene may be expressed at different levels or at different times than the endogenous gene would be. This situation prompted A. Francis Stewart, Jan-Michael Peters, Frank Buchholz, Anthony Hyman, and co-workers of the MitoCheck consortium (www.mitocheck.org) at several research institutions in Germany, Austria, Canada, the U.K., and the U.S. to develop a high-throughput method called BAC TransgeneOmics for the tagging of mammalian genes.
With the new method, the researchers tag bacterial artificial chromosomes (BACs). The advantage of BACs is that they are large enough to contain entire genes with their normal regulatory sequences. Because BACs often are generated during genome sequencing efforts, BAC libraries are available for many organisms.
Tags are efficiently introduced into BACs by a process called recombineering, which involves the homologous recombination of the tag plasmid and the BAC in E. coli. This step is performed in 96-well plates to ramp up the throughput. Then, tagged BACs are transfected into mammalian cells.
Various tests verified that most of the stably transfected cell lines expressed the correct gene at nearly endogenous levels. In addition, the proteins encoded by 11 of 15 genes selected for immunofluorescence assays were visible at the expected subcellular locations. Similarly, affinity purification of tagged proteins pulled down known binding partners. Chromatin immunopurification experiments showed that known DNA-binding proteins that were tagged by BAC TransgeneO-mics still bound to the correct DNA sequences. Finally, the researchers transfected a BAC transgene into mouse embryonic stem cells. The transgene was detectable in 3 of the 10 resulting embryos. The scientists say that the method is automatable and can be performed with almost any model system. (Nat. Methods 2008, 5, 409–415)
The effect of antimitotic chemotherapy drugs on mitotic proteins
Some microtubule poisons, such as Taxol, are used successfully in chemotherapy regimens to treat cancer. However, other drugs, such as colchicine and nocodazole, that seem to act in a similar way do not have a therapeutic effect. To better understand this phenomenon, Marc Kirschner and colleagues at Harvard Medical School, Children’s Hospital Boston, Brigham and Women’s Hospital, and the University of Heidelberg (Germany) studied the phosphorylation patterns of proteins involved in mitosis. They found that various mitotic proteins were differentially phosphorylated when cells were treated with different drugs. In addition, the mitotic arrest induced by some drugs was not as static as researchers had previously assumed.
To detect drug effects, the phosphosites of proteins in the anaphase-promoting complex (APC) were monitored by MS for nine conditions, including seven drug treatments that arrest cells in mitosis as well as cells in normal mitosis and G1 states. Kirschner and colleagues mapped 71 distinct phosphosites of APC subunits and related spindle checkpoint proteins that co-immunoprecipitated with APC. Most of the phosphorylation occurred when cells were treated with drugs.
Researchers routinely treat cells with nocodazole to arrest them in mitosis. To see whether normal mitosis and drug-induced mitotic arrest are equivalent, the investigators compared APC phosphorylation in nocodazole-treated cells and cells undergoing mitosis normally. Surprisingly, the phosphorylation patterns of some APC subunits varied between these conditions; this result suggests that a drug-induced arrest does not provide an accurate snapshot of the normal mitotic state. In addition, Kirschner and colleagues used isotope-free quantitation of the extent of modification (known as iQEM) to determine phosphorylation stoichiometries during a nocodazole arrest. APC phosphorylation changed throughout the time course, so the investigators say that “arrest” is no longer an adequate term for nocodazole-treated cells.
Although microtubule poisons appear to act in the same way as nocodazole by keeping cells in a mitotic-like state, they have different effects on APC phosphorylation. For example, some APC phosphosites were reproducibly specific for certain drugs. Also, some sites were phosphorylated only when treated with drugs and not when cells were undergoing normal cell division. Although the scientists don’t fully understand all of the effects of the microtubule poisons, they say that these types of experiments could lead to better cancer therapies. (Proc. Natl. Acad. Sci. U.S.A. 2008, 105, 6069–6074)
Proteome map of Arabidopsis thaliana
Katja Baerenfaller, Wilhelm Gruissem, Sacha Baginsky, and colleagues at ETH Zurich, the University of Zurich, the Functional Genomics Center Zurich, the University of Manchester (U.K.), and Tel Aviv University (Israel) report that they have created a proteome map for the model plant Arabidopsis thaliana. Protein extracts from six organs (roots, leaves, flowers, and so on) and cultured A. thaliana cells were separated by 1DE and in-gel digested. The peptides were analyzed by LC/MS/MS. A total of 13,029 proteins were identified with a false discovery rate <1%. As expected, rare proteins, such as those involved in transcription and signaling, were underrepresented in the data set, whereas abundant proteins, such as those involved in metabolism, were overrepresented.
With these data, the researchers discovered that some identified peptides did not match the predicted gene models in the Arabidopsis Information Resource, release 7 (known as TAIR7). Some of the peptides originated from regions that were not known to produce proteins. For example, peptides encoded by predicted introns, regions between genes, and pseudogenes were discovered. In other cases, a new start or stop site or a different open reading frame was identified for a particular gene. Evidence for 57 new or alternative gene models was obtained in the study.
When the scientists analyzed the Gene Ontology terms for the identified proteins by organ, they found that a different proteome map could be assigned to every organ. Of the proteins identified in the study, 571 were unique to an organ, and these were called organ-specific biomarkers. Also, 14,867 organ-specific proteotypic peptides were discovered. The data are freely available at the PRIDE database, and an enhanced version is located on the researchers’ server at www.AtProteome.ethz.ch. (Science 2008, DOI 10.1126/science.1157956)
Figuring out in vivo protein–protein interactions
To discover protein–protein interactions (PPIs), many researchers pull down one protein from a lysate with an antibody, then analyze the copurified proteins by MS. Other investigators perform two-hybrid assays, in which two tagged proteins must come together in the nucleus for the experiment to work properly. In both cases, transient interactions may be missed, and even with the two-hybrid assay, proteins may no longer be in their normal cellular environment or compartment. Therefore, Stephen Michnick and co-workers at the University of Montreal and McGill University (Canada) developed an in vivo assay for PPIs in which proteins remain in their typical environment. With this approach, they compiled a new yeast interactome map that includes several new PPIs.
In the protein-fragment complementation assay (PCA), two proteins are tagged with fragments of a reporter protein. If the two tagged proteins interact, then the attached reporter fragments come close enough to each other to fold and become active. Unlike in the two-hybrid screen, proteins in the PCA method are not required to move into a special cellular compartment. Thus, other (unlabeled) proteins in a complex that may be required for the tagged proteins to interact also are present. The researchers used PCA to map in vivo yeast binary interactions on a genome scale. (Science 2008, DOI 10.1126/science.1153878)
How ibuprofen may protect against Alzheimer’s disease
Recent reports suggest that some nonsteroidal anti-inflammatory drugs (known as NSAIDs), such as ibuprofen, may prevent Alzheimer’s disease (AD). To figure out how ibuprofen exerts this effect, Wei Ning Chen and colleagues at Nanyang Technological University (Singapore) conducted a proteomics study. Several proteins, such as enzymes involved in antioxidant processes, were differentially regulated in ibuprofen-treated and control neuroblastoma cells. In addition, the levels of reactive oxygen species (ROS) were low in ibuprofen-treated cells.
Neuroblastoma cells were treated with the S enantiomer of ibuprofen (the more effective form of the drug), the R enantiomer (the less effective form), or a mixture, or were left untreated. Peptides from the samples were quantified with isobaric tagging for relative and absolute quantitation (iTRAQ) and MS/MS. A total of 13 proteins, including metabolic enzymes, signaling molecules, and cytoskeletal proteins, were differentially regulated in the ibuprofen-treated and control cells. In addition, mRNA levels generally correlated with protein levels.
Enzymes involved in oxidative stress were diffentially expressed in treated and untreated cells, so the investigators took a closer look at this process. ROS levels were measured in cells treated with the S enantiomer, the R enantiomer, or the mixture, as well as in untreated cells. The lowest ROS levels were observed in cells incubated with the S enantiomer. This finding was consistent with the observation from the iTRAQ experiment in which antioxidant enzymes were up-regulated in these cells. Chen and colleagues say that ibuprofen may protect against AD by reducing ROS levels, but additional mechanisms are possible. (Proteomics 2008, DOI 10.1002/pmic.200700556)
Identifying wheat allergens
Grains are important sources of nutrition, but wheat causes allergic reactions in many people. Several research teams have suggested that members of the α-amylase inhibitor family play a role in wheat allergies. To find additional candidate allergens, Petr Šotkovský and colleagues at the Czech Academy of Sciences, the University of Defence, Imumed, and Bulovka Hospital (all in the Czech Republic) took a proteomics approach. They discovered many possible wheat antigens and developed a new ELISA to test two α-amylase inhibitors.
Salt-soluble proteins were extracted from the wheat cultivar Sulamit, which is frequently used by the food industry. Proteins were separated by 1DE, blotted onto nitrocellulose membranes, and incubated with sera from individuals with wheat allergies or pollen allergies (a disease control) as well as from healthy subjects. The immunoblotting experiment confirmed that individuals with wheat allergies, but not healthy subjects or those with pollen allergies, produce antibodies that bind to wheat proteins.
Next, wheat proteins were separated by 2DE, and the gel was silver-stained. Immunoblotting showed that ~80 spots were recognized by patient sera. Wheat proteins may be recognized in the body after they are digested in the gastrointestinal tract, however. So, to mimic this process, the researchers treated wheat proteins with pepsin and repeated the experiment. In this case, ~20 spots were visualized. In total, 19 novel and known wheat allergens were identified, including α-amylase inhibitors. Western blotting confirmed that these inhibitors are major wheat allergens.
Šotkovský and colleagues developed an ELISA and observed significant differences in the binding of antibodies in sera of wheat allergy patients, disease controls, and healthy controls to a wheat extract and two α-amylase inhibitors (types 1 and 3). Finally, the conventional test for blood basophil activation was conducted with the two α-amylase inhibitors. The type 1 inhibitor produced an activation response in sera from half of the patients, whereas the type 3 inhibitor did not produce a response. Therefore, the scientists say that the type 1 α-amylase inhibitor may work in combination with other wheat proteins to activate the immune system. (Proteomics 2008, 8, 1677–1691)
Advertisements
Return to Top
Copyright © American Chemical Society, 1155 Sixteenth Street N.W., Washington, DC 20036
