Structurally Sound

Few researchers would dispute that data integration is a laudable goal; nevertheless, even fewer would claim that it is easy to achieve. One challenging problem is how to meaningfully combine sequence and structural information with phylogenetics. There are highly developed applications for performing multiple sequence alignments, and equally refined tools for viewing and analyzing structures. A gap exists, however, in programs that can combine sequence and structure data in a way that enables sophisticated and flexible evolutionary analyses. To meet this need, Roberts et al., describe MultiSeq, a powerful analysis package that is intended to harness structure and sequence information in order to render an “evolutionary profile,” or a summary of the evolutionary diversity of a related group of molecules. This program combines the best of both worlds: the relative richness of sequence information (reflecting the lag time between obtaining a biomolecule's sequence and its structural description) and the power of structural information (which is better conserved than sequence and thus allows deeper probing of evolutionary relationships). MultiSeq is included in recent versions of VMD (Visual Molecular Dynamics; a visualization and simulation package) and is available in OS X, Linux, Unix, and Windows flavors. Although MultiSeq can handle both nucleic acid and protein data, its capabilities are currently more advanced for the latter. Structural data in many common formats (PDB, XYZ, NetCDF, and CHARMM) can be imported directly, as can FASTA, PHY, ALN, and other popular sequence formats. Hundreds of sequence and structure files can be uploaded, and metadata in the form of NCBI taxonomy information, Enzyme Commission (EC) numbers, and SCOP structural classifications can also be captured via the “electronic notebook” function. Alignment of imported data occurs via optimized algorithms and is quality controlled by dedicated processes for correcting any bias that might result from the composition of database entries. Visualization options for structures, aligned sequences, and phylogenetic trees are diverse and assist the user's recognition of notable patterns. Finally, although MultiSeq will most immediately appeal to evolutionary biologists, the platform is amenable to other efforts that rely on integration of sequence and structure data.

- Roberts et al. 2006. MultiSeq: unifying sequence and structure data for evolutionary analysis. BMC Bioinformatics 7:382 [Epub ahead of print, August 16, 2006].

Seeking Sulfation

Fashion-obsessed socialites aren't the only ones who feel the need to accessorize: proteins may not even be recognized by their prospective partners unless they are decorated with a particular post-translational modification (PTM). Despite the functional significance of PTMs, detecting which proteins are modified (and with what) remains a substantial challenge in proteomics. Ideally, PTM-specific antibodies could be used to isolate and probe proteins with a given modification of interest. However, antibodies that recognize the PTM alone (i.e., not within the context of the modified protein) are relatively rare. After all, with some 200 known PTMs, there are only a few modifications (nitrosylated tyrosine, phosphotyrosine, phosphoserine/threonine) for which such antibodies are available. The ubiquity of PTMs in secreted and membrane-bound proteins may make these modifications relatively nonimmunogenic, causing traditional antibody generation procedures to fail. In vitro-based approaches, such as phage antibody libraries, would not be limited in this manner. Therefore, Kehoe et al., investigated whether phage display could be used to select sequences from a phage antibody library that recognize a specific PTM independent of any sequence context. The PTM they chose to work with was tyrosine sulfation, for which no straightforward detection strategy currently exists. Initial attempts did not go smoothly: sulfated peptide antigen was not easy to generate in sufficient quantities for screening, and initial selection resulted in a preponderance of truncated clones. Thus, the authors were forced to screen nearly 8000 clones, an increase of 20- to 80-fold compared to more typical values. This vast number of clones was divided amongst several selection strategies, but the end result was a single sequence. The winning scFv was expressed in the context of full-length IgG and as a fusion with alkaline phosphatase. Both of these recombinant proteins were highly stable and showed prowess in both ELISAs and Western blot analyses. Thus, this paper not only describes a useful new reagent for proteomics, but also introduces a promising method for isolating other PTM-specific antibody fragments.

- Kehoe et al. 2006. Using phage display to select antibodies recognizing post-translational modifications independently of sequence context. Molecular & Cellular Proteomics [Epub ahead of print, September 12, 2006].

A Closer Fit

Nanoengineering certainly sounds much more cutting edge than aluminum smelting, but it turns out that a longstanding industrial process for metal passivation may be the key to an important advance in DNA-sensing nanopores. If anodized aluminum is placed in boiling water, a process known as hydrothermal sealing occurs. Metallurgists care about this phenomenon because, in partially blocking pores on the surface of the metal, it protects against corrosion. Takmakov et al., who have an interest in alumina nanopores, realized that hydrothermal sealing might be a viable strategy to improve nanoporebased DNA sensing. The basis for this hypothesis comes from the mechanism of DNA detection via nanopores. A nanopore will have different ionic conductance depending on whether it is open or partially blocked. If binding sites for a given molecule are present within the pores, the concentration of this analyte can be determined by measuring a spike in impedance. Clearly, the closer the pore diameter is to the analyte's size, the more profound the impedance change will be. Methods exist for creating such fine-scale nanopores, but they tend to be expensive and require specialized equipment. Anodized alumina membranes can be an economical source of nanopore arrays, but existing fabrication methods create pore diameters several-fold larger than would be practical for DNA detection. With this in mind, the authors' interest in pore shrinking via hydrothermal treatment becomes obvious. Pleasingly, the data obtained did strongly support a significant decrease in pore diameter. More importantly, the method seems to be applicable to DNA sensing: when a 21-nucleotide ssDNA oligonucleotide is immobilized within the pore and then exposed to a complementary sequence, DNA hybridization can be detected by an increase in impedance. This effect is specific because an unrelated oligonucleotide, which would not be expected to anneal to the ssDNA target and therefore would not block the pore, had no effect. This study is of particular note because anodized alumina membranes are very amenable to highly parallel nanopore arrays. As a consequence, it is tempting to imagine the technology could lead to a label-free electrical detection microarray in which each shrunken pore houses a different ssDNA probe.

-Takmakov et al. 2006. Hydrothermally shrunk alumina nanopores and their application to DNA sensing. The Analyst [Epub ahead of print, September 5, 2006].

NMR Gets Moving

At first glance, electrophoresis would appear to have little in common with NMR. The former technique is usually inexpensive and provides relatively general information about proteins; the latter is complicated, pricey, and generates high-resolution structural information. Even so, the two approaches have been combined in electrophoretic NMR (ENMR), which uses differing electrophoretic mobilities of proteins to untangle superimposed NMR signals obtained during probing of protein mixtures. The analysis of multiple proteins in conditions approaching their normal cellular environment does, however, mean that the buffer has high-ionic conductivity. Applying electrophoretic current under such conditions can lead to undesirable heating. Although capillary-based techniques have succeeded in allaying problematic heating, the resulting interferograms display substantial artifacts that make it difficult to accurately resolve different species separated during the electrophoresis step. Thakur and He, tackled this confounding factor by switching from a fast Fourier transform (FFT) analysis to a method using Burg's maximum entropy method (MEM), which has a successful track record for analysis of NMR spectra. To compare FFT and MEM, the authors used simulations, reanalyzed a previously published BSA/ubiquitin 2D-ENMR data set, and acquired data on a lysozyme/D₂O sample. They consistently observed a several-fold performance improvement when MEM processing was used instead of FFT-based analysis. Although the authors did not demonstrate structural acquisition in this set of experiments, the approach is an important step in the quest to use multidimensional ENMR to resolve the structures of multiple members of a protein mixture under physiological buffer conditions.

- Thakur and He. 2006. High flow-resolution for mobility estimation in 2D-ENMR of proteins using maximum entropy method (MEM-ENMR). Journal of Magnetic Resonance 183:32-40.