|Home | About | Journals | Submit | Contact Us | Français|
This column highlights recently published articles that are of interest to the readership of this publication. We encourage ABRF members to forward information about articles they feel are important and useful to Clive Slaughter, MCG-UGA Medical Partnership, 279 William St., Athens, GA 30607-1777, USA. Tel.: (706) 369-5945: Fax: (706) 369-5936; E-mail: ude.gcm.liam@rethgualsc; or to any member of the editorial board. Article summaries reflect the reviewer's opinions and not necessarily those of the association.
Shrestha B, Vertes A. In situ metabolic profiling of single cells by laser ablation electrospray ionization mass spectrometry. Analytical Chemistry 81;2009;8265–8271.
The new ionization technique of laser ablation electrospray ionization (LAESI) is used here as an approach to metabolic profiling of individual cells with the intention of investigating the cell-to-cell variation in metabolic status that occurs with differences in the phase of the cell cycle, history, and interaction with environmental factors. Single-cell ablation is achieved by delivering to tissue pulses of laser energy at a mid-infrared wavelength at which water absorbs. The resulting plume predominantly contains neutrals. These are intercepted by an electrospray that ionizes the neutral species for mass spectrometric analysis. The laser energy is delivered to the tissue via a germanium dioxide-based glass-fiber cable with a tip sharpened by etching to provide a radius of curvature of 15 μm. By bringing the tip close to the sample surface, an area averaging as little as 30 μm diameter can be ablated. This set-up is used to study onion and daffodil bulb epidermal cells. Although still in its early stages, methodology of this type brings closer the ultimate goal of molecular imaging based on cells as the natural voxels.
Crank J A, Armstrong D W. Towards a second generation of ionic liquid matrices (ILMs) for MALDI-MS of peptides, proteins, and carbohydrates. Journal of the American Society for Mass Spectrometry 20;2009;1790–1800.
Ionic liquid matrices for MALDI have several advantages over conventional organic matrices, including absence of the need to co-crystallize them with analytes, absence of “hot-spots,” better quantification as the analyte-matrix mixture is homogeneous, and no need for harsh modifiers such as trifluoroacetic acid, which tends to disrupt molecular interactions of proteins. This paper reports a screen for new compounds to use as ionic liquid matrices for MALDI. N,N-Diisopropylethylammonium α-cyano-4-hydroxycinnamate and N-isopropyl-N-methyl-t-butylammonium α-cyano-4-hydroxycinnamate are identified as the best matrices for proteins and peptides, and N,N-diisopropylethylammonium α-cyano-4-hydroxycinnamate and N,N-diisopropylethylammonium ferulate are the best for carbohydrates. The ability to ionize analytes is shown to depend on the proton affinity and pKa of the cationic component of the matrix. Ionic liquid matrices are found to produce fewer degradation products from polysaccharides than solid matrices.
Delvolve A, Woods AS. Ammonium sulfate and MALDI in-source decay: a winning combination for sequencing peptides. Analytical Chemistry 81;2009;9585–9589.
This article provides guidelines for optimizing the conditions for in-source decay of peptides during MALDI and establishes in-source decay as an inexpensive method for acquiring amino acid sequence information. The procedure involves optimizing the choice of MALDI matrix and ion extraction delay. In confirmation of previous studies, in-source decay is shown to be promoted by addition of 3 M ammonium sulfate to the matrix.
Gardner M W, Smith S I, Ledvina A R, Madsen J A, Coon J J, Schwartz J C, Stafford G C, Brodbelt J S. Infrared multiphoton dissociation of peptide cations in a dual pressure linear ion trap mass spectrometer. Analytical Chemistry 81;2009;8109–8118.
A dual-pressure Velos linear ion trap mass spectrometer is modified for infrared multiphoton dissociation (IRMPD) of peptides. When IRMPD is performed in the low-pressure cell (3×10−4 Torr), the rate of photon absorption is greater than the rate of collisional cooling, and dissociation approaches 100% efficiency with irradiation times of <25 ms. IRMPD is found to allow observation of ions of lower mass than collision-induced dissociation (CID), including immonium ions. IRMPD provides slightly better sequence coverage, probably as this nonresonant activation method allows secondary and higher-order dissociation events. The method yields singly charged product ions predominantly, as product ions of a higher-charge state photodissociate preferentially. This has the effect of rendering less-cluttered and more-easily interpretable product ion spectra.
Madsen J A, Gardner M W, Smith S I, Ledvina A R, Coon J J, Schwartz J C, Stafford G C, Brodbelt J S. Top-down protein fragmentation by infrared multiphoton dissociation in a dual pressure linear ion trap. Analytical Chemistry 81;2009;8677–8686.
The use of the dual-pressure linear ion trap with IRMPD capability is extended here to analysis of intact proteins. The high-pressure cell provides improved trapping and isolation efficiencies. Isotopic resolution of ions with a 10+ charge state is achieved in the low-pressure cell. Fragmentation of ions by IRMPD in the low-pressure cell is compared with fragmentation by CID in the high-pressure cell (4.7 mTorr) for proteins of 8.6 to 29 kDa. IRMPD yields product ions of lower-charge states, which makes mass assignment easier. IRMPD also proves more selective for cleavage N-terminal to proline and C-terminal to acidic residues, a feature that may assist database searching to identify proteins by top-down methods.
Trimpin S, Inutan E D, Herath T N, McEwen C N. Laserspray ionization, a new atmospheric pressure MALDI method for producing highly charged gas-phase ions of peptides and proteins directly from solid solutions. Molecular and Cellular Proteomics 9;2010;362–367.
LAESI is shown here to yield multiply charged ions of intact proteins. The spectra are similar to those produced by electrospray ionization and distinct from the low-charge states of protein ions produced by MALDI. The LAESI experiments are performed with 2,5-dihydroxybenzoic acid as a MALDI matrix. The method permits the analysis of large, intact proteins using mass analyzers of circumscribed mass range, including ion-trapping instruments.
Lu Q, Zheng X, Mcintosh T, Davis H, Nemeth J F, Pendley C, Wu S-L, Hancock W S. Development of different analysis platforms with LC–MS for pharmacokinetic studies of protein drugs. Analytical Chemistry 81;2009;8715–8723.
Although mass spectrometric assays have been used extensively for pharmacokinetic measurements of small molecule drugs in serum, immunochemical methods (e.g., ELISA) have been favored for protein drugs. However, immunochemical assays are subject to a long development time and may be sensitive to competition from endogenous antibodies in patients developing immunity to the protein drug. LC-MS methods (e.g., multiple reaction monitoring) are not encumbered by these disadvantages, but enrichment is often necessary to achieve adequate sensitivity. This study tests three enrichment methods for antibody pharmaceuticals: anti-idiotypic antibody enrichment for very high specificity, protein A enrichment for moderate specificity, and albumin depletion (with SDS-PAGE) for low specificity. The albumin-depletion strategy achieved detection sensitivity for the drug of 1 ng in 30 μL serum, with linearity over five orders of magnitude. Protein A enrichment with SDS-PAGE achieved 0.02 ng sensitivity (i.e., 50× lower), and without SDS-PAGE, 10 ng sensitivity. Anti-idiotypic antibody enrichment (without SDS-PAGE) achieved 0.1 ng sensitivity with a similar dynamic range. The various methods may find complementary use. Protein A enrichment and albumin depletion are methods applicable to many drugs once developed, whereas anti-idiotypic antibodies can be used for one drug only. If anti-idiotypic antibodies are unavailable, and sensitivity requirements are not stringent, protein A enrichment without SDS-PAGE is convenient and capable of high throughput.
Gilar M, Xie H, Jaworski A. Utility of retention prediction model for investigation of peptide separation selectivity in reversed-phase liquid chromatography: impact of concentration of trifluoroacetic acid, column temperature, gradient slope and type of stationary phase. Analytical Chemistry 82;2009;265–275.
Optimizing reverse-phase separations of complex peptide mixtures is a nontrivial task that is facilitated by knowledge of how column selectivity may change under a variety of experimental conditions. The analysis in this paper addresses the changes in selectivity that may result with peptides of varying amino acid composition and molecular weight by changing the concentration of an ion-pairing reagent (trifluoroacetic acid), increasing the separation temperature, changing gradient slope, and altering the column chemistry and pore size. This information will be of interest to investigators seeking to manipulate separations to reduce analyte complexity, remove specific interfering species, or optimize the duration of separations.
Rudnick P A, Clauser K R, Kilpatrick L E, Tchckhovskol D V, Neta P, Blonder N, Billheimer D D, Blackman R K, Bunk D M, Cardasis H L, Ham A-J L, Jaffe J D, Kinsinger C R, Mesri M, Neubert T A, Schilling B, Tabb D L, Tegeler T J, Vega-Montolo L, Variyath A M, Wang M, Wang P, Whiteaker J R, Zimmerman L J, Carr S A, Fisher S J, Gibson B W, Paulovich A G, Regnier F E, Rodriguez H, Spiegelman C, Tempst P, Liebler D C, Stein S E. Performance metrics for liquid chromatography-tandem mass spectrometry systems in proteomics analyses. Molecular Cellular Proteomics 9;2010;225–241.
A major goal of proteomics is the documentation of distinct characteristics of different biological systems. These differences are detected by comparing multiple technical replicates of representative samples within and between laboratories using comparable instrument systems and methods. The critical assumption is that observed differences reflect true proteomic differences rather than differences in performance of the analytical systems. However, LC-MS-based workflows, in particular, are complex, and a lack of criteria by which their performance may be judged has hampered progress in the field. The present paper proposes a list of performance metrics for this purpose. They number 46 in total and address performance in chromatographic separation of peptides, ion-source stability, MS1 scanning, dynamic sampling, MS2 scanning, and peptide identification. These metrics are applied to replicate LC-MS/MS analyses and display high levels of consistency, and therefore, provide the means for systematic troubleshooting of even subtle problems that may lead to a decline in peptide identifications. Inter-laboratory studies indicate stability in performance across multiple sites and instrument platforms, and comparisons in which chosen system variables are held constant enable key sources of variability in capabilities to be identified.
Ow S Y, Salim M, Noirel J, Evans C, Rehman I, Wright P C. iTRAQ underestimation in simple and complex mixtures: “the good, the bad and the ugly.” Journal of Proteome Research 8;2009;5347–5355.
The availability of isobaric tag for relative and absolute quantitation (iTRAQ) data sets acquired in multiple laboratories delineating quantitative changes of protein expression in diverse systems has led to the disturbing observation that the iTRAQ technique yields fold-change values spanning a range that is considerably narrower than microarray expression studies. In this publication, the dynamic range achievable using iTRAQ is evaluated using a defined sample composed of a mixture of four standard proteins treated on its own or spiked into a bacterial cell lysate of high complexity. Quantitative ratios of 1:1, 1:1.5, 1:10, and 1:100 are selected for analysis. Factors undermining accuracy are identified. Isotopic impurities, that result from manufacture of the tags represent a source of error that can be corrected. The iTRAQ m/z 121.12 reporter is subject to contamination by immonium ions from phenylalanine-containing peptides (m/z 121.08). With appropriate correction for these factors, a dynamic range of two orders of magnitude is shown to be achievable, although quantification at ratios of 100:5 and 100:1 loses accuracy as a result of degradation of precision. Peptide cofragmentation as a result of the simultaneous selection of more than one precursor ion is identified as a source of error that is difficult to minimize presently. Finally, some experimental designs (e.g., cell or organelle separations) reflect large changes in proteome composition that may invalidate the “overall unchanged” assumption, and the resulting overcorrection of measured protein abundance differences may lead to large differences being missed or diminished.
Rachdaoui N, Austin L, Kramer E, Previs M J, Anderson V E, Kasumov T, Previs S F. Measuring proteome dynamics in vivo: as easy as adding water? Molecular and Cellular Proteomics 8;2009;2653–2663.
Protein turnover in vivo is typically studied by administering labeled amino acids. However, in long-term studies (days or weeks), continuously administering labeled amino acids may be exceedingly expensive, and in short-term studies (hours), uncertainty arises in the calculation of precursor/product labeling ratios, as the true level of precursor labeling may be altered through amino acid turnover. These problems may be overcome by administering labeled water (e.g., 2H2O) instead of labeled amino acids. Label is incorporated continually into amino acids from water, and precursor (water) labeling is readily measured. The assumption that equilibration between body water and free amino acids is faster than incorporation of free amino acids into proteins has been validated experimentally in the case of serum albumin synthesis by comparing the incorporation of 2H from 2H2O into alanine and incorporation of 2H-alanine into protein using gas chromatography-MS methods. In the present paper, the methodology is used to study albumin synthesis in steady-state and non-steady-state conditions, e.g., integrating transitions between fed and fasting state. The use of H2O18 in place of 2H2O is also investigated as a means of accelerating the protein-labeling process by incorporation of label directly from water into protein carboxyl groups during synthesis.
Hassibi A, Vikalo H, Riechmann J L, Hassibi B. Real-time DNA microarray analysis. Nucleic Acids Research 37;2009;e132.
Like the process of analyte binding to any affinity-based biosensor, analyte binding to DNA microarrays is a kinetic stochastic process that must be described as a nonlinear function of time, analyte concentration, and reaction kinetics. The authors of this paper argue that the current method of microarray analysis, which is based on a single data point from the hybridization process, is susceptible to errors arising from effects such as probe saturation, washing artifacts, and spot-to-spot variation in probe density, and is therefore incapable of dealing quantitatively with systems in which analyte concentrations vary over a wide range of concentration, as is the case in gene expression analysis. These problems would be alleviated if real-time quantification of binding were used. The difficulty with implementing such an approach is that signal must be monitored in the presence of unbound analyte. In conventional systems, fluorescent background that results from unbound analyte would swamp signal from probe spots. This problem is overcome in the present work by using fluorescent resonance energy transfer to monitor binding. Radiating donor molecules are attached to probes in each spot, and acceptor molecules are attached to the analytes. Binding is then measured as a diminution of fluorescence from each spot. In principle, this solution is applicable to any microarray system and is not restricted to nucleic acid arrays.
Carro M S, Lim W K, Alvarez M J, Bollo R J, Zhao X, Snyder E Y, Sulman E P, Anne S L, Doetsch F, Colman H G, Lasorella A, Aldape K, Califano A, Iavarone A. The transcriptional network for mesenchymal transformation of brain tumours. Nature 463;2010;318–325.
Given the large number of genes changing expression level during transformation from a normal to a cancerous cellular phenotype, is it possible to distinguish the genes causing the change in phenotype from the ones that are simply associated with it? The present paper achieves this feat in identifying a transcriptional module in high-grade gliomas that suppresses the pro-neural expression signature of glial cells and stimulates a mesenchymal expression signature that is associated with poor prognosis, local invasiveness of the tumor, and stimulation of angiogenesis. The starting point for the analysis is the long list of differentially expressed genes that constitutes the cancer signature. From this list, “Algorithm for the Reconstruction of Accurate Cellular Networks,” is deployed to infer a genome-wide repertoire of high-grade, glioma-specific interactions between transcription factors and their targets. Next, a new master regulator analysis algorithm is used to compute the statistical significance of the overlap between each transcription factor's regulon and the set of genes affected in the mesenchymal signature, and the transcription factors thus identified are ranked in order of the total number of mesenchymal signature target genes they regulate. This procedure identified six transcription factors: five activators and one repressor. Experimental follow-up then established that two of these transcription factors, C/EBPβ and STAT3, are sufficient in neural stem cells and necessary in glioma cells for mesenchymal transformation. This methodology is expected to be useful for the identification of master regulators of other phenotypic states.
Temple G, Gerhard D S, Rasooly R, Feingold L A, Good P J, Robinson C, Mandich A, Derge J G, Lewis J, Shoaf D, Collins F S, Jang W, Wagner L, Shenmen C M, Misquitta L, Schaefer C F, Buetow K H, Bonner T L, Yankie L, Ward M, Phan L, Astashyn A, Brown G, Farrell C, Hart J, Landrum M, Maidak B L, Murphy M, Murphy T, Rajput B, Riddick L, Webb D, Weber J, Wu W, Pruitt K D, Maglott D, Siepel A, Brejova B, Diekhans M, Harte R, Baertsch R, Kent J, Haussler D, Brent M, Langton L, Comstock C L G, Stevens M, Wei C, Van Baren M J, Salehi-Ashtiani K, Murray R R, Ghamsarl L, Mello E, Lin C, Pennacchio C, Schreiber K, Shapiro N, Marsh A, Pardes E, Moore T, Lebreau A, Muratet M, Simmons B, Kloske D, Sieju S, Hudson J, Sethupathy P, Brownstein M, Bhat N, Lazar J, Jacob H, Gruber C E, Smith M R, McPherson J, Garcia A M, Gunaratne P H, Wu J, Muzny D, Gibbs R A, Young A C, Bouffard G G, Blakesley R W, Mullikin J, Green E D, Dickson M C, Rodriguez A C, Grimwood J, Schmutz J, Myers R M, Hirst M, Zeng T, Tse K, Moksa M, Deng M, Ma K, Mah D, Pang J, Taylor G, Chuah E, Deng A, Fichter K, Go A, Lee S, Wang J, Griffith M, Morin R, Moore R A, Mayo M, Munro S, Wagner S, Jones S J M, Holt R A, Marra M A, Lu S, Yang S, Hartigan J, Graf M, Wagner R, Letovksy S, Pulido J C, Robison K, Esposito D, Hartley J, Wall V E, Hopkins R F, Ohara O, Wiemann S. The completion of the mammalian gene collection (MGC). Genome Research 19;2009; 2324–2333.
The MGC project, an effort by the National Institutes of Health to provide the world-wide research community with unrestricted access to sequence-validated cDNA clones of the entire protein-coding sequence of mRNA transcripts for all human and mouse genes, announces the completion of its collection. The collection now includes 92% of human genes and 89% of mouse genes with curated RefSeq transcripts, and 97% of human genes and 96% of mouse genes with curated RefSeq transcripts and at least one publication. Sequence data are of high quality, with errors below 1 in 50,000 bp. Since the inception of this project in 2000, our view of the eukaryotic transcriptome has expanded greatly to include multiple splice isoforms and non-protein-coding RNAs, many of which are as yet poorly characterized. Access to these may be expected to be satisfied largely by synthesis. However, the existing centralized collection offers benefits of scale (cost reduction and uniformity of quality), reducing wasted effort in duplication of clone preparation within the community and relieving individual laboratories of the burden of clone quality control and distribution.
Choi M, Scholl U I, Ji W, Liu T, Tikhanova I R, Zumbo P, Nayir A, Bakkaloǧiu A, Özen S, Sanjad S, Nelson-Williams C, Farhi A, Mane S, Lifton R P. Genetic diagnosis by whole exome capture and massively parallel DNA sequencing. Proceedings of the National Academy of Sciences U.S.A. 106;2009;19096–19101.
Ng S B, Buckingham K J, Lee C, Bigham A W, Tabor H K, Dent K M, Huff C D, Shannon P T, Jabs E W, Nickerson D A, Shendure J, Bamshad M J. Exome sequencing identifies the cause of a Mendelian disorder. Nature Genetics 42;2010;30–35.
Reports by two independent groups herald the inception of new methodology for discovering the genetic basis of rare Mendelian disorders or providing help where diagnosis is uncertain. The methodology consists of massively parallel sequencing of whole exomes. Exomes are chosen for study, rather than whole genomes, to realize a large reduced cost. Both groups select exomic DNA by hybridizing to NimbleGen arrays and then sequencing on an Illumina platform, achieving a mean sequence converage of >30×. Choi et al., studying a patient with the suspected diagnosis of Bartter syndrome, a renal-salt wasting disease, make the unexpected genetic diagnosis of congenital chloride diarrhea based on the finding of a homozygous mis-sense mutation at SLC26A3, the known locus for congenital chloride diarrhea. Recognition of this mutation (D652N) as a cause of the condition was assisted by the rigorous conservation of this particular residue from invertebrates to humans. The attribution was confirmed subsequently by studying five additional patients suspected of having Bartter syndrome yet lacking mutations in known genes for the disease. All were found to have homozygous deletions in SLC26A3. Ng et al. identify the gene for a rare, recessive malformation disorder, called Miller syndrome, by exome sequencing of four affected individuals in three independent kindreds. Genes having two nonsynonymous splice-site or indel sequence variants in each of the individuals were selected as candidates. Unrelated and benign variants were then filtered out by searching public Single Nucleotide Polymorphism databases and eight HapMap exomes. This procedure resulted in identification of a single candidate gene, DHODH, which encodes dihydro-orotate dehydrogenase, an enzyme of pyrimidine synthesis. The identification was confirmed subsequently by sequencing three additional, unrelated persons with the disorder and an additional sibling in one of the initial three families (making a total of 11 mutations in six families). Despite the statistical challenges inherent in such an approach, the methodology is expected to make a major contribution to the study of monogenic disorders.
Lieberman-Aiden E, Van Berkum N L, Williams L, Imakaev M, Ragoczy T, Telling A, Amit I, Lajoie B R, Sabo P J, Dorschner M O, Sandstrom R, Bernstein B, Bender M A, Groudine M, Gnirke A, Stamatoyannopoulos J, Mirny I A, Lander E S, Dekker J. Comprehensive mapping of long-range interactions reveals folding principles of the human genome. Science 326;2009;289–293.
Although chromatin folding at the level of nucleosomes is familiar and well-studied, there exists longer-range folding that brings widely separated functional elements into close spatial proximity. This process strongly affects gene activity, nuclear architecture, and the functional state of the cell. In this paper, methodology for investigating such long-range interaction on a genome-wide scale is described. The procedure, called Hi-C, consists of combining proximity-based ligation with massively parallel sequencing. Cells are cross-linked with formaldehyde, which covalently links together spatially adjacent chromatin segments. The DNA is then digested with a restriction enzyme that leaves a 5′ overhang so that when the overhang is filled, a biotin-labeled nucleotide may be incorporated to label the cut-sites. The resulting labeled, blunt-end cut sites are then ligated under dilute conditions, so as to favor ligation between cross-linked fragments. This has the effect of creating a library of pieces that were originally in close spatial proximity in the nucleus. The library is then sheared, and biotinylated fragments are selected and subjected to massively parallel paired-end sequencing to produce a catalog of interacting segments. The resulting interaction maps reveal that small, gene-rich chromosomes preferentially interact with one another. With regard to interactions within chromosomes, after normalizing to account for the tendency of segments that are close in the linear sequence to interact with one another more extensively than ones far apart in sequence, the interaction frequencies reflect the existence of two kinds of DNA regions: loosely folded, gene-rich regions that show a relatively low level of interaction with one another and densely folded, gene-poor regions that interact with one another more extensively. These regions constitute different spatial compartments in the nucleus. Furthermore, within compartments, the relationship between interaction frequency and linear distance is shown to be consistent with a fractal-type globule state, i.e., an unentangled state in which globules interact together like beads on a string, then beads-of-beads on a string, etc. This is quite distinct from the alternative, “equilibrium globule” model that calls for a densely knotted structure. Although this paper focuses on interactions on the megabase scale, the point is made that the methodology supports study of interactions on a finer scale if the number of sequence reads is increased. This is expected to enable mapping of specific long-range interactions among enhancers, silencers, and insulators.
Fullwood M J, Liu M H, Pan Y F, Liu J, Xu H I, Mohamed Y B, Orlov Y L, Velkov S, Ho A, Mei P H, Chew E G Y, Huang P Y H, Welboren W-J, Han Y, Ooi H S, Ariyaratne P N, Vega V B, Luo Y, Tan P Y, Choy P Y, Wansa K D S A, Zhao B, Lim K S, Leow S C, Yow J S, Joseph R, Li H, Desai K V, Thomsen J S, Lee Y K, Karuturi R K M, Herve T, Bourque G, Stunnenberg H G, Ruan X, Cacheux-Rataboul V, Sung W-K, Liu E T, Wei C-L, Cheung E, Ruan Y. An oestrogen-receptor-α-bound human chromatin interactome. Nature 462;2009;58–64.
Using methods similar to the above procedure of Lieberman-Aiden et al., the authors of the present paper use chromatin immunoprecipitation to focus on the chromatin interaction network bound to the estrogen receptor α. They call their method ChIA-PET (chromatin interaction analysis by paired-end tag) sequencing. The selection scheme permits much higher resolution for protein-mediated functional interactions.
Levskaya A, Weiner O D, Lim W A, Voigt C A. Spatiotemporal control of cell signaling using a light-switchable protein interaction. Nature 461;2009;997–1001.
Yazawa M, Sadaghiani A M, Hsueh B, Dolmetsch R E. Induction of protein-protein interactions in live cells using light. Nature Biotechnology 27;2009;941–945.
For the investigation of signal transduction cascades, the goal of exercising precise experimental control of protein-protein interactions in the temporal and spatial domains is of great potential benefit. It has not been realized by using small organic ligands, as the spacial and temporal resolution they permit is not high, and their specificity is often inadequate. What if it were possible to control such protein interactions in mammalian cells by means of light? The two papers here report new experimental systems that achieve this goal. Levskaya et al. construct a system using phytochrome B (PhyB) from Arabidopsis. Apo-PhyB covalently binds to the chromophore, phycocyanobilin, to form a light-sensitive holoprotein. PhyB undergoes a conformational change between two states, Pr (red-absorbing) and Pfr (far-red-absorbing): Stimulation with red light causes a transition from the Pr to the Pfr state, and in Arabidopsis, the Pfr state binds a downstream transcription factor, phytochrome interaction factor 3 (PIF3). In mammalian cells, the authors express PIF tagged with yellow fluorescent protein (YFP) and PhyB tagged with a plasma membrane-localization signal. They show that in response to red light, the YFP is translocated precisely and reversibly to the plasma membrane, and this can be made to occur with micrometer spatial resolution and on the second time-scale. Using this system, they show that upstream activators of Rho-family GTPases, which control the actin skeleton, can be made to reshape and direct the cell morphology of mammalian cells in response to light. Yasawa et al. use the light-dependent binding of the FKF1 protein from Arabidopsis to the GIGANTEA protein (GI). FKF1 binds flavin mononucleotide and when illuminated by blue light, forms a covalent bond with it via a cysteine residue on FKF1. This allows FKF1 to bind to GI. The process is reversed upon hydrolysis of the bond with cysteine. The authors express a membrane-bound form of GI in mammalian cells and a form of FKF1 tagged with YFP. They show that the YFP is recruited to the membrane in response to blue light. Using this system, they recruit the small G protein Rac1 to the membrane and induce local formation of lamellipodia in response to illumination. They also made a light-activated transcription factor by fusing domains of GI to Gal4 and FKF1 to Gal4 and VP16. Such systems are expected to be readily adapted to other systems for control of receptor activation, synapse formation, and other signaling systems.
Keiser M J, Setola V, Irwin J J, Laggner C, Abbas A I, Hufeisen S J, Jensen N H, Kuijer M B, Matos R C, Tran T B, Whaley R, Glennon R A, Hert J, Thomas K L H, Edwards D D, Shoichet B K, Roth B L. Predicting new molecular targets for known drugs. Nature 462;2009;175–181.
Although drugs are designed to interact with defined targets, and side-effects are minimized by promoting specificity, many drugs, nonetheless, have off-target interactions. In some cases, such interactions are essential to the drug's therapeutic benefit. A new method for predicting the targets to which drugs are likely to bind is described in the present article. Rather than seeking structural or sequence similarity among the targets, the approach used here is to compare targets by the similarity of the ligands already known to bind to them. The similarities are expressed as expectation values generated by modified BLAST algorithms. This similarity ensemble approach identifies commonality in ligand-binding characteristics among proteins that would otherwise be considered quite unrelated. In the present work, some 3600 drugs, including U.S. Food and Drug Administration-approved and experimental compounds, are screened for possible binding to 1400 protein targets, each target represented by its known set of ligands. Expectation of binding is scored on the basis of similarities between the drugs and the ligand ensemble for each protein. Seven thousand unanticipated associations of high probability are identified. Of these, 30 are tested experimentally, and 23 are confirmed. Five of them have dissociation constants <100 nM. Several of the confirmed associations represent previously unknown pharmaceutical effects of medical importance. The approach is likely to reveal side-effects and new indications for many drugs in the future.
Noble WS. How does multiple testing correction work? Nature Biotechnology 27;2009;1135–1137.
This article is a primer about an important issue in high-throughput experimentation, which is ignored frequently. Suppose a study of a biological process is designed to detect the involvement of hitherto unsuspected proteins. The involvement of each protein is associated with a probability score, and the magnitude of some of the scores suggests that the corresponding proteins are unlikely to have been implicated by chance alone. Nevertheless, thousands of individual proteins and therefore, thousands of separate probability scores have been generated in the study, and it is indeed possible that some or all of these probability scores might have reached the level they did simply by chance. How do you correct for this multiple testing effect? The article presents examples of methods for dealing with this problem and presents criteria for use in choosing a method, which are based on the benefits of being correct and the costs of being wrong in any particular case. References are supplied for further reading as needed.