Making Polycarbonate More Plastic

Polycarbonate is widely employed in microfluidic devices because it is relatively easy and inexpensive to work with. However, polycarbonate is not very flexible when it comes to methods for biofunctionalization. The malleability of polycarbonate that makes it so convenient also makes this plastic sensitive to heat and chemical treatments that are often used in surface functionalization. Carion et al,. show that the limitations of polycarbonate can be overcome with the help of silica nanoparticles, for which a variety of functionalization methods have been described. The process begins by cutting polycarbonate sheets into microscope slide-shaped pieces, which can then be placed on a standard noncontact piezoelectric microarrayer. Semicarbazide silica nanoparticles are then spotted onto the surface in the desired micropattern. Adding a glyoxylyl peptide spurs a reaction with the semicarbazide moiety, thereby covalently attaching the peptide to the adsorbed nanoparticles. These immobilized peptides might then serve as probes for the presence of antibodies in biological samples. The authors tested the potential of their system for biosensor applications by immobilizing hemagglutinin and FLAG® tag peptides, exposing them to negative control and target antibodies, and then detecting by fluorescently labeled secondary antibodies. The approach proved selective and sensitive. Selectivity varied according to the size of the nanoparticles (diameters from 27 to 304 nm were tested); sensitivity measurements revealed a detection limit of 10 ng/mL. This new workflow should be of real benefit to those who are looking for alternatives to gold- or silica-based microfluidic devices and for those interested in using compact discs for biosensing applications.

-Carion et al. 2007. Chemical micropatterning of polycarbonate for site-specific peptide immobilization and biomolecular interactions. Chembiochem 8(3):315-322.

The Search Is Over

While it may be too early to declare the end of existing computational methods for interpreting shotgun proteomics data, a recent paper from Lam et al,. suggests that an alternative, library-based strategy may be reaching maturity. At the moment, most researchers with shotgun proteomics data turn to a tool such as SEQUEST to turn spectra into identifications. SEQUEST and related tools take protein sequences from public databases, digest these in silico, predict spectra for the putative peptides, and compare these computationally derived results against the experimental findings in order to deduce peptide sequence. This is a highly evolved process, but the heavy computational demands entail significant time and resource costs. Moreover, many of the in silico predictions may be biologically irrelevant—just because a peptide could exist in theory does not mean it is observed in practice—and therefore much time and effort is spent on futile searching. The appeal of directly comparing experimentally derived spectra with a library of reference data has been long recognized, but compilation and validation of an appropriate set of reference spectra has not yet been convincingly shown. While Lam et al. acknowledge that it will take time for spectral library matching to replace traditional search methods, they describe a library of 30,000 yeast peptide spectra and a search tool that enables fast, sensitive, and accurate peptide identification. In direct comparisons with SEQUEST, the new methodology proved about three orders of magnitude faster. In addition, because the library method uses real spectra, empirically observed sequence-specific peak intensity variations contribute to the reliability of the matching process. This property and related features of the spectral library matching program allow it to significantly outperform SEQUEST in measures of sensitivity and selectivity. The advantages of the spectral matching method are obviously contingent on the coverage of the library. The authors therefore conclude their report with a call for proteomics researchers to share peptide spectra via online data repositories. These databases serve as the raw material for adding reference spectra to libraries and will be crucial to extending spectral library matching to other species.

-Lam et al. 2007. Development and validation of a spectral library searching method for peptide identification from MS/MS. Proteomics [Epub ahead of print, February 13, 2007].

Mother Knows Best

Because aneuploidy is a common genetic abnormality, safe and sensitive methods for prenatal screening are of significant interest. Amniocentesis and chorionic villus sampling can detect imbalances in chromosome dosage, but as invasive procedures they can never be completely risk-free for the fetus. It would be much more desirable to use a maternal blood draw as a proxy of fetal health. It is well known that nucleic acids of fetal origin are found at low but detectable frequencies into plasma of pregnant women. Unfortunately, the ratio of maternal to fetal DNA works against any efforts to detect fetal aneuploidy by DNA analysis. Lo et al,. have now demonstrated the feasibility of using RNA for this purpose instead. The authors sought a gene that is expressed by fetal cells throughout pregnancy, but which is silent in maternal cells. Because they were interested in detecting Down syndrome, they narrowed their search to genes on chromosome 21. They then focused their attention on a SNP within the selected gene. If the fetus is heterozygous at this locus, aneuploidy would be expected to cause a deviation of the one-to-one ratio between the transcripts of the two different genotypes. The assay itself involves RNA isolation from maternal plasma, reverse transcription, primer extension, mass spectrometry-based differentiation between the two alleles, and calculation of the allelic ratio. The authors found the single marker gave them a sensitivity of 90% and a specificity of just under 97%. It is likely that this assay could be further improved by using a panel of SNPs; in addition, the general strategy may prove applicable to blood-based detection of gene dosage abnormalities indicative of cancer.

-Lo et al. 2007. Plasma placental RNA allelic ratio permits noninvasive prenatal chromosomal aneuploidy detection. Nature Medicine 13(2):218-223.

Making It

Success in determining the natural product that is produced by an uncharacterized biosynthetic gene cluster can be elusive. More and more of these so-called orphan gene clusters are coming to researchers’ attention because of the rapid accumulation of microbial genome sequences. Determining the identity of the corresponding natural products may yield bioactive compounds with commercial or medicinal value. To date, biochemical methods have dominated approaches for identifying the products of orphan gene clusters. For instance, one strategy is to try and purify the metabolite after expressing the gene in heterologous host. In cases where a convenient assay for the metabolite is available, assay-guided fractionation can be used. Genomic mining for natural products has also been described; in this case, metabolic profiling and mutagenesis are used to link a given gene to the corresponding natural product. Gross et al,. now describe what they call a “genomisotopic” approach. This method begins with detailed bioinformatic analysis of the inferred amino acid sequence coded by the orphan ORF. In the example discussed in the paper, the orphan gene cluster (from Pseudomonas fluorescens) encodes a nonribosomal peptide synthetase that is predicted by homology to have a cyclic lipopeptide (CLP) as its product. Crucially, the amino acids in the CLP can be deduced from certain residues in particular domains of the synthetase. By looking at the amino acids that are predicted to occur in the natural product, the authors selected an isotopically labeled amino acid that is expected to be incorporated in the CLP and that would be absent in other metabolites produced within the bacterial strain. In this way, the natural product of interest can be isolated with the assistance of NMR, which is used to trace the target during biochemical fractionation. By following this procedure, the authors report the discovery of orfamide A, a novel CLP that is active against zoospores produced by a plant pathogen. More significantly, however, their study is an important demonstration of the power of this methodology to isolate secondary metabolites that can be predicted via bioinformatics.

-Gross et al. 2007. The genomisotopic approach: a systematic method to isolate products of orphan biosynthetic gene clusters. Chemistry & Biology 14(1):53-63.