Aggregation Aggravation

Parkinson's disease (PD), dementia with Lewy bodies (DLB), and multiple systemic atrophy (MSA) are neurodegenerative disorders characterized by aggregated α-synuclein deposits in the brain. These deposits have been linked to almost complete loss of dendritic spines, suggesting that synaptic dysfunction in DLB and possibly PD is caused by presynaptic accumulation of α-synuclein aggregates. Further investigation into the mechanisms of these diseases would be greatly facilitated by a simple biochemical method for detecting α-synuclein aggregates. The development of such an assay has turned out to be very difficult due to the insoluble nature of the aggregates. Digestion with proteinase K, extraction with various detergents, and pretreatment with guanidinium hydrochloride, formic acid, or urea all failed to solubilize α-synuclein aggregates. This made Western blot analysis ineffective since the majority of immunoreactivity appeared in the stacking gel, with the remainder smeared across the lanes. While this pattern is generally interpreted as an aggregate, it cannot be differentiated from oligomeric or unspecific immunoreactivity. Mass spectrometry, capillary electrophoresis, and reversed-phase high-performance liquid chromatography are equally unsuitable due to their dependence on soluble proteins. In the issue, Michael Kramer, Christina Behrens, and Walter Schulz-Schaeffer at the University of Göttingen (Göttingen, Germany) describe a protein aggregate filtration (PAF) assay based on the separation of aggregated insoluble α-synuclein from the soluble isoform by microfiltration. This method takes advantage of the insolubility of α-synuclein in detergents by dissolving pelleted homogenates in sodium lauryl sarcosinate and then passing the solution through a membrane. Oligomers will pass through into the supernatant leaving the membrane to be developed like an immunoblot and the immunoreactivity detected by chemiluminescence. The PAF assay involves simpler sample preparation than Western blot analysis, allows the direct comparison of the amounts of α-synuclein aggregate, and shows specificity for the detection of α-synuclein aggregates in human and mouse homogenates and cell cultures. The authors combined the PAF assay with subcellular fractionation of brain homogenates to demonstrate for the first time that α-synuclein aggregates exist at presynaptic terminals in DLB.

(See “Selective detection, quantification, and subcellular location of α-synuclein aggregates with a protein aggregate filtration assay” on page 403.)

Fuzzy Thoughts About Sticky Peptides

The identification of novel peptides regulating cell behavior via interaction with cell surface receptors is important for the development of peptide-conjugated biomaterials for such purposes as tissue engineering or drug delivery. Given that directly screening all possible permutations of 20 amino acids is impractical even for the shortest peptide sequence, researchers have been developing computationally assisted methods using different algorithms to increase the efficiency of peptide design and screening. Many of these methods, however, are limited by the fact that their predictive rules have typically been validated with the datasets initially used to generate them and have not been experimentally confirmed by repeated rounds of peptide interaction assays. In this issue, Chiaki Kaga and colleagues working in the laboratory of Hiroyuki Honda at Nagoya University (Nagoya, Japan) describe a screen for synthetic cell-binding peptides (to be used in artificial tissue culture scaffolds) in which the design rules for the tested peptides were computationally generated through a fuzzy neural network (FNN) algorithm and then tested by multiple rounds of a high-throughput cell attachment assay. Fuzzy neural networks generate a model of the hidden relationships between input and output data through learning, and this model is then interpreted by the algorithm's “fuzzy layer” to extract predictive “if-then” rules. Kaga et al. employed an FNN algorithm to design the pentameric or tetrameric peptides comprising the peptide array used in the cell attachment assay. They screened a first round of 643 peptide pentamers for attachment of NIH/3T3 cells. Sets of the best and worst ranked peptides were chosen and numerical indices for each amino acid in these peptides were derived from their position and three structural parameters (hydrophobicity, isoelectric point, and size). Using this input data, their FNN algorithm produced rules that distinguished positive peptides from negative peptides and these rules guided the synthesis and testing of the next array of 270 peptides. This next round resulted in an increased yield of peptides positive for cell adhesion, and the additional rules derived from this round were used to design 50 peptides in the third round, resulting in an even more significant increase in the number of positive peptides, including eight novel tetramers with stronger binding affinities than the known integrin ligand RGDS.

(See “Computationally assisted screening and design of cell-interactive peptides by a cell-based assay using peptide arrays and a fuzzy neural network algorithm” on page 393.)

Knocking Out More Knockout Mice

The International Mouse Knockout Consortium plans to create a knockout mouse for every gene in the recently sequenced C57BL/6 mouse genome. The creation of knockout mice is slowed by low efficiency at several steps, including producing appropriate blastocysts, generating chimeras, and achieving germline transmission. Gabriela Pacholczyk and colleagues from the Medical College of Georgia (Augusta, Georgia) aimed to improve the efficiency of generating knockout mice in a pure C57BL/6 background by optimizing blastocyst production and chimera generation capacity using their recently described C57Bl/6NHsd embryonic stem (ES) cell line, LK1. By testing potential blastocyst donor strains, the authors found that the rate of C2H × BALB/c blastocyst production was 2-3-fold higher than that of BALB/c females when mated to BALB/cAnNCr males. The resulting blastocysts successfully produced litters and showed levels of germline transmission in up to 100% of the male pups. While it was possible to generate chimeras using these blastocysts, the degree of chimerism could not be easily estimated by coat color, necessitating the development of a peripheral blood screening technique for this purpose. The degree of chimerism in male chimeric pups was determined by quantification of peripheral blood cells by flow cytometry to detect ES cell-specific markers. The degree of chimerism varied between pups, but even pups showing as low as 8% ES cell contribution achieved germline transmission in their first litter. Therefore, optimization of the blastocyst donor strain, in connection with use of the LK1 ES cell line, resulted in significantly increased numbers of blastocysts and a high degree of germline competency. While these results are dependent on the use of the LK1 ES cells, the authors achieved success using an ES cell clone from the Bruce4 C57BL/6 cell line as well. This indicates the potential of C2H × BALB/c blastocysts to provide substantial increases in capacity to generate knockout mice directly in a C57BL/6 background with a variety of available C57BL/6 ES cell lines.

(See “Generation of C57BL/6 knockout mice using C3H × BALB/c blastocysts” on page 413.)

Have a Heart, in 3-D

For the in vitro modeling of cardiac muscle differentiation and maturation, it has become increasingly clear that standard cell culture techniques are inadequate. Correct formation of the electrical and mechanical connections between mature cardiomyocytes requires a three-dimensional (3-D) environment that allows for proper cell-to-cell and cell-to-extracellular matrix interactions and provides a continuous mechnical load for developing cardiac fibers to contract against, conditions which are missing when cardiomyocytes are grown on conventional two-dimensional (2-D) tissue culture plates. While researchers have recently developed elaborate methods for growing cardiac tissue in vitro in three dimensions, these require costly and elaborate equipment that are typically restricted to laboratories specializing in tissue engineering. To overcome these limitations, K. Bakunts et al. in the Sarvazyan laboratory at The George Washington University (Washington, DC, USA) have developed a simple approach using standard cell culture equipment to produce a 3-D net of cardiac fibers that can contract against a mechanical load and are easily accessible to physiological monitoring and immunohistochemical staining. Their technique uses the commercially available basement membrane preparation Matrigel, which self-polymerizes into a tissue culture scaffold that closely resembles in vivo basement membrane. Bakunts et al. created a two-layer Matrigel pillow consisting of a layer of rat neonatal cardiomyocytes suspended in diluted Matrigel overlaid on a drop of previously polymerized concentrated Matrigel. The looser structure of the top layer facilitates the formation of the cardiac fibers while the stiffer base layer allows the spontaneously beating fibers to contract against a mechanical load as well as each other. They optimized the concentration of Matrigel in the top layer as well as the seeding concentration of the cardiomyocytes to obtain a range of thin and thick fibers. These fibers were easily stained for contractile and membrane proteins, could be loaded with indicator dyes such as calcium dyes to monitor calcium flow during fiber beating, and the larger fibers could be externally paced.

(See “Formation of cardiac fibers in Matrigel matrix” on page 341.)