Bundle Up

Combining nucleic acids with cationic carriers enables intracellular delivery via endocytosis. Ironically, because of their stiffness and low charge density, siRNAs are harder to condense into uptake-compatible complexes than much larger DNAs. As described in Nature Materials, Mok et al. have discovered a counterintuitive solution: multimerizing the siRNAs packs them into more palatable form. In their system, the sense and antisense siRNA strands are thiol-modified at the 3′ end and, in separate reactions, chemically cross-linked to generate sense and antisense dimer preparations. These are mixed and allowed to anneal, spawning “multi-siRNAs,” which are mostly oligomers of three or more repeats. Mixing these species with linear polyethylenimine (LPEI) gives a mean complex size of 82 nm, compared with 464 nm for LPEI-bundled standard siRNA. Once in the cells, the multi-siRNAs should rapidly revert to monomeric siRNAs suitable for gene silencing, thanks to the disulfide linkages of the cross-linker and the reducing environment. At optimal multi-siRNA to LPEI ratios (balanced between greatest knock-down and least toxicity), gene expression levels were reduced 50–90%, depending upon the target, substantially outperforming monomeric siRNA. This improvement carried over to mice when the siRNA moieties were tested by intratumoral injection into green fluorescent protein (GFP)–expressing tumors. This all occurred without meaningful off-target effects. This protocol for packaging siRNA makes it possible to use a highly biocompatible cationic carrier without sacrificing silencing performance.

Mok et al. Multimeric small interfering ribonucleic acid for highly efficient sequence-specific gene silencing. Nat. Mater. [Epub ahead of print, January 24, 2010; doi: 10.1038/nmat2626].

Random Assortment

An alternative to microarrays of ordered spots is the random array, which encodes identity in target-specific barcodes instead of location. Most existing barcodes, such as those based on organic dyes or quantum dots, rely on optical detection, which limits the number of resolvable barcodes. Although DNA-based barcodes may be one solution, mass spectrometry offers another decoding approach based on using slightly different masses to uniquely identify distinct targets. In Nano Letters, Gunnarsson et al. lay the basis for combining imaging mass spectrometry with a planar random array for multiplexed DNA detection. The chip is prepared so that the generally inert surface is randomly interrupted by biotin groups. Via a NeutrAvidin bridge, biotinylated oligonucleotide probes can be affixed in these random locations. After unlabeled target DNA is added, liposomes bearing cholesterol-tagged capture oligonucleotides serve to complete the sandwich assay. Because liposomes can be prepared with different lipid components, each one's mass signature can act as a barcode when the array is analyzed by time-of-flight secondary ion mass spectrometry. By setting the instrument for maximum lateral resolution, individual liposomes are detectable, each of which, based on the assay conditions, corresponds to a single DNA target. Under conditions of maximum spatial resolution, a degree of mass resolution is sacrificed. This, combined with evidence of liposome loss during sample preparation, suggests that further optimization is needed to fully exploit the detection capabilities. However, even in its current form, imaging mass spectrometry should permit roughly 200 distinguishable barcodes. This, coupled with the system's potential for dense probe distributions, should encourage a shift away from the constraints of spatially defined microarray probes.

Barcoding by imaging mass spectrometry of oligonucleotide-conjugated liposomes with different lipid constituents. (Click to enlarge)

Gunnarsson et al. 2010. Liposome-based chemical barcodes for single molecule DNA detection using imaging mass spectrometry. Nano Lett. 10:732–737.

Vector Multiplication

There have been a number of lentiviral gene transfer vectors that have been described over the years, many exquisitely optimized for high-level transduction of exotic cell types. To the nonspecialist primarily interested in these as tools for functional genomics, committing to a particular system can be tough. The ideal would be a state-of-the-art vector system that provides an array of fluorescent and selectable markers in a flexible, modular format. The lentiviral gene ontolog y (LeGO; www.LentiGO-vectors.de) vectors were introduced in 2008 with this intent. Now, in an article in Gene Therapy from Weber et al., the LeGO team describes the full potential of the system. In its original manifestation, LeGO comprised several vectors containing different expression cassettes and various fluorescent markers. Using fluorescent proteins for marking transduced cells means that selection depends upon cell sorting, which can be inconvenient, damaging to some cell types, and prone to contamination. Antibiotic resistance markers offer an alternative for selection, and in the new work, the authors expand beyond previously described enhanced green fluorescent protein (eGFP) and dTomato fusions with the blasticidin resistance gene to offer this marker in combination with the Cerulean, Venus, and mCherry fluorescent proteins, as well as eGFP fusions with resistance genes against neomycin, hygromycin, puromycin, and zeocin. Antibiotic selection invariably produced highly purified populations within 8–10 days, even for triply transduced NIH/3T3 cells, or doubly transduced primary mesenchymal stem cells and neural stem cells. As their modular design makes additional customization straightforward, the LeGO system should be flexible enough for all types of cell tracking and functional genomics studies.