Empirical Evaluation of Oligonucleotide Probe Selection for DNA Microarrays

DNA-based microarrays are increasingly central to biomedical research. Selecting oligonucleotide sequences that will behave consistently across experiments is essential to the design, production and performance of DNA microarrays. Here our aim was to improve on probe design parameters by empirically and systematically evaluating probe performance in a multivariate context. We used experimental data from 19 array CGH hybridizations to assess the probe performance of 385,474 probes tiled in the Duchenne muscular dystrophy (DMD) region of the X chromosome. Our results demonstrate that probe melting temperature, single nucleotide polymorphisms (SNPs), and homocytosine motifs all have a strong effect on probe behavior. These findings, when incorporated into future microarray probe selection algorithms, may improve microarray performance for a wide variety of applications.     

Does light intensity affect self‐feeding and food wastage in group‐held rainbow trout and white‐spotted charr?

The self‐feeding rhythms of rainbow trout Oncorhynchus mykiss and white‐spotted charr Salvelinus leucomaenis were studied when group‐held fishes ( n  = 10 per group) were fed using self‐feeders under two different light intensities (50 lx, 16 μW cm⁻² and 700 lx, 215 μW cm⁻²) during the light phase of the light‐dark cycle. Food wastage was also measured. At 50 lx, all groups of rainbow trout learned to operate the self‐feeder within 4 days, whereas it took up to 25 days for all groups at 700 lx. In contrast, all groups of white‐spotted charr learned self‐feeding within 17 days, irrespective of light intensity. These results, although non‐significant, suggest that lower light intensities can stimulate instrumental learning in rainbow trout, but not white‐spotted charr. In rainbow trout, the total number of trigger actuations for the entire experimental period was significantly higher at 50 rather than 700 lx, although this may have been related to delayed learning at 700 lx. There was no significant effect in white‐spotted charr. Growth rate (assessed using the thermal growth coefficient) was also higher in rainbow trout but not white‐spotted charr at 50 rather than 700 lx, although this difference was non‐significant. Light intensity had no significant effect on food wastage in either rainbow trout or white‐spotted charr, and it did not appear to affect the proportion of trigger actuations during the light phase. Clear diurnal feeding rhythms were observed in both species and these were classified into four categories: uniform, dawn, dusk and crepuscular. At 50 lx, fish from both species generally fed in temporally localized periods at either dawn and dusk, whilst feeding was predominantly uniform during the light phase at 700 lx.     

Light-directed synthesis of peptide nucleic acids (PNAs) chips

We report herein the light-directed synthesis of peptide nucleic acids (PNAs) microarray using PNA monomers protected by photolabile protecting groups and a maskless technique that uses a digital micromirror array system to form virtual masks. An ultraviolet image from the virtual mask was cast onto the active surface of a glass substrate, which was mounted in a flow cell reaction chamber connected to a peptide synthesizer. Light exposure was followed by automatic chemical coupling cycles and these steps were repeated with different virtual masks to grow the desired PNA probes in a selected pattern. In a preliminary experiment, an array of PNA probes with dimensions of 4.11 mm × 4.11 mm was generated on each slide. Each synthesis region in the final array measured 210 μm × 210 μm for a total of 256 sites. The center-to-center space was 260 μm. It was observed from the hybridization pattern of the fluorescently labeled oligonucleotide targets that the fluorescence intensities of the matched, and mismatched sequences showed substantial difference, demonstrating specificity in the identification of complementary sequences. This opens the way to exploit processes from the microelectronics industry for the fabrication of PNA microarrays with high densities.     

The development and characterization of a 57K single nucleotide polymorphism array for rainbow trout

In this study, we describe the development and characterization of the first high‐density single nucleotide polymorphism (SNP) genotyping array for rainbow trout. The SNP array is publically available from a commercial vendor (Affymetrix). The SNP genotyping quality was high, and validation rate was close to 90%. This is comparable to other farm animals and is much higher than previous smaller scale SNP validation studies in rainbow trout. High quality and integrity of the genotypes are evident from sample reproducibility and from nearly 100% agreement in genotyping results from other methods. The array is very useful for rainbow trout aquaculture populations with more than 40 900 polymorphic markers per population. For wild populations that were confounded by a smaller sample size, the number of polymorphic markers was between 10 577 and 24 330. Comparison between genotypes from individual populations suggests good potential for identifying candidate markers for populations' traceability. Linkage analysis and mapping of the SNPs to the reference genome assembly provide strong evidence for a wide distribution throughout the genome with good representation in all 29 chromosomes. A total of 68% of the genome scaffolds and contigs were anchored through linkage analysis using the SNP array genotypes, including ~20% of the genome assembly that has not been previously anchored to chromosomes.     

Characterisation of a low pathogenic form of Gyrodactylus salaris from rainbow trout

Gyrodactylus salaris was isolated from rainbow trout in a Danish freshwater trout farm, and a laboratory population of this particular parasite form was established on rainbow trout. Challenge infections were performed using different salmonid strains and species, including East Atlantic salmon Salmo salar (from the Danish River Skjern?), Baltic salmon S. salar (from the Swedish River Ume Alv) and rainbow trout Oncorhynchus mykiss (from the Danish rainbow trout farm Fousing). These were compared to infection studies on the Norwegian Laerdalselva parasite form kept under exactly the same conditions in the laboratory. The Danish G. salaris form had low virulence towards both Atlantic and Baltic salmon, whereas rainbow trout proved susceptible to the parasite. The Danish G. salaris form was able to maintain a very low infection on East Atlantic salmon, but not on the Baltic salmon, which eliminated the infection within 2 wk. Rainbow trout developed infection intensities ranging up to several hundred parasites per host. The host colonization patterns of the parasite differed clearly from those of previous studies on microhabitats of the Norwegian form of G. salaris. A comparative study on morphological characters (opisthaptoral hard parts) from the Danish parasite form and Norwegian G. salaris showed no significant differences. Selected genes comprising internal transcribed spacers 1 and 2 (ITS), ribosomal RNA intergenic spacer (IGS) and cytochrome c oxidase subunit I (COI) regions were cloned and sequenced. Five sequenced ITS clones from 5 individuals of the Danish strain consistently revealed a single base substitution compared to ITS sequences from all other known species and strains of Gyrodactylus. Mitochondrial COI gene sequences demonstrated that the Danish G. salaris form is closely similar to the Laerdalselva parasite form found in Norway. The IGS sequences were highly variable, but very similar to those obtained from German isolates of G. salaris.     

Isolation and preliminary characteristics of β-N-acetylglucosaminidase in the sperm of Siberian sturgeon (Acipenser baerii) and rainbow trout (Oncorhynchus mykiss)

The aim of this study was to characterize the enzyme β- N -acetyglucosaminidase (β-NAGase) in the milt and spermatozoa extracts from Siberian sturgeon and rainbow trout. After ion exchange chromatography one protein peak showed β-NAGase activity in sturgeon milt plasma and sperm extracts of both species. Surprisingly, two protein peaks showing β-NAGase activity were found in rainbow trout milt plasma. The molecular mass of β-NAGase was estimated by gel filtration as 127 kDa for rainbow trout spermatozoa, 271 kDa for sturgeon spermatozoa, and 74 kDa for milt plasma from both species. The kinetic parameters were determined for milt plasma and sperm extracts. The optimum pH of the β-NAGases was 3.8 for sturgeon milt plasma, 4.4 for sturgeon sperm extract, and 4.4–4.8 for milt plasma and sperm extract from rainbow trout. K _m value of the β-NAGases was 0.212, 0.563, 0.779 m m for sturgeon milt plasma, sturgeon sperm extract or rainbow trout extract, respectively. The β-NAGase from sperm extracts in both species showed 100% activity even after incubation at 56°C by 20 min, whereas its activity was decreased to 23% in sturgeon milt plasma and to 2% in trout milt plasma.     

Influence of temperature on silver accumulation and depuration in rainbow trout

To assess the influence of water temperature on silver uptake, rainbow trout Oncorhynchus mykiss ( c . 50 g; held at 13° C) were exposed to 0·1 μM AgNO₃ in ion‐poor water for 1 week at 4 and 16° C without previous temperature acclimation. To assess the influence of temperature on elimination of previously accumulated Ag, rainbow trout were exposed to 0·1 μM AgNO₃ in ion‐poor water for 1 week at 12° C, then were randomly divided amongst two Ag‐free water containers, differing only in temperature (3 and 16° C), for 2 months. In the uptake study greater accumulation of Ag was seen in the gills, plasma and especially the livers and bile of 'warm' rainbow trout (16° C) compared to 'cold' rainbow trout (4° C), which can be explained by the higher metabolic rates of the warmer fish. In the depuration study there was no net elimination of Ag from the livers and bile but there was biphasic elimination of Ag from the gills and plasma of 'warm' and 'cold' fish, but with few differences between them. This indicated that temperature‐dependent processes were less important in Ag elimination than in Ag uptake. Toxicokinetic modelling of Ag uptake by livers indicated four‐fold greater uptake of Ag by 'warm' rainbow trout compared to 'cold' rainbow trout (one compartment uptake model). Elimination of previously accumulated Ag from the plasma was best fitted by a two compartment rate‐constant based model, with approximately half the plasma Ag load eliminated within 24 h, followed by slower elimination of Ag over 2 months.     

Development and validation of a resistance and virulence gene microarray targeting Escherichia coli and Salmonella enterica

A microarray was developed to simultaneously screen Escherichia coli and Salmonella enterica for multiple genetic traits. The final array included 203 60-mer oligonucleotide probes, including 117 for resistance genes, 16 for virulence genes, 25 for replicon markers, and 45 other markers. Validity of the array was tested by assessing inter-laboratory agreement among four collaborating groups using a blinded study design. Internal validation indicated that the assay was reliable (area under the receiver-operator characteristic curve = 0.97). Inter-laboratory agreement, however, was poor when estimated using the intraclass correlation coefficient, which ranged from 0.27 (95% confidence interval 0.24, 0.29) to 0.29 (0.23, 0.34). These findings suggest that extensive testing and procedure standardization will be needed before bacterial genotyping arrays can be readily shared between laboratories.     

A Second Generation Integrated Map of the Rainbow Trout (<Emphasis Type="Italic">Oncorhynchus mykiss</Emphasis>) Genome: Analysis of Conserved Synteny with Model Fish Genomes

DNA fingerprints and end sequences from bacterial artificial chromosomes (BACs) from two new libraries were generated to improve the first generation integrated physical and genetic map of the rainbow trout (Oncorhynchus mykiss) genome. The current version of the physical map is composed of 167,989 clones of which 158,670 are assembled into contigs and 9,319 are singletons. The number of contigs was reduced from 4,173 to 3,220. End sequencing of clones from the new libraries generated a total of 11,958 high quality sequence reads. The end sequences were used to develop 238 new microsatellites of which 42 were added to the genetic map. Conserved synteny between the rainbow trout genome and model fish genomes was analyzed using 188,443 BAC end sequence (BES) reads. The fractions of BES reads with significant BLASTN hits against the zebrafish, medaka, and stickleback genomes were 8.8%, 9.7%, and 10.5%, respectively, while the fractions of significant BLASTX hits against the zebrafish, medaka, and stickleback protein databases were 6.2%, 5.8%, and 5.5%, respectively. The overall number of unique regions of conserved synteny identified through grouping of the rainbow trout BES into fingerprinting contigs was 2,259, 2,229, and 2,203 for stickleback, medaka, and zebrafish, respectively. These numbers are approximately three to five times greater than those we have previously identified using BAC paired ends. Clustering of the conserved synteny analysis results by linkage groups as derived from the integrated physical and genetic map revealed that despite the low sequence homology, large blocks of macrosynteny are conserved between chromosome arms of rainbow trout and the model fish species.     

Evaluating oligonucleotide properties for DNA microarray probe design

Most current microarray oligonucleotide probe design strategies are based on probe design factors (PDFs), which include probe hybridization free energy (PHFE), probe minimum folding energy (PMFE), dimer score, hairpin score, homology score and complexity score. The impact of these PDFs on probe performance was evaluated using four sets of microarray comparative genome hybridization (aCGH) data, which included two array manufacturing methods and the genomes of two species. Since most of the hybridizing DNA is equimolar in CGH data, such data are ideal for testing the general hybridization properties of almost all candidate oligonucleotides. In all our data sets, PDFs related to probe secondary structure (PMFE, hairpin score and dimer score) are the most significant factors linearly correlated with probe hybridization intensities. PHFE, homology and complexity score are correlating significantly with probe specificities, but in a non-linear fashion. We developed a new PDF, pseudo probe binding energy (PPBE), by iteratively fitting dinucleotide positional weights and dinucleotide stacking energies until the average residue sum of squares for the model was minimized. PPBE showed a better correlation with probe sensitivity and a better specificity than all other PDFs, although training data are required to construct a PPBE model prior to designing new oligonucleotide probes. The physical properties that are measured by PPBE are as yet unknown but include a platform-dependent component. A practical way to use these PDFs for probe design is to set cutoff thresholds to filter out bad quality probes. Programs and correlation parameters from this study are freely available to facilitate the design of DNA microarray oligonucleotide probes.     

Identification of a molecular marker for type A spermatogonia by microarray analysis using gonadal cells from pvasa-GFP transgenic rainbow trout (Oncorhynchus mykiss)

The spermatogonia of fish can be classified as being either undifferentiated type A spermatogonia or differentiated type B spermatogonia. Although type A spermatogonia, which contain spermatogonial stem cells, have been demonstrated to be a suitable material for germ cell transplantation, no molecular markers for distinguishing between type A and type B spermatogonia in fish have been developed to date. We therefore sought to develop a molecular marker for type A spermatogonia in rainbow trout. Using GFP-dependent flow cytometry (FCM), enriched fractions of type A and type B spermatogonia, testicular somatic cells, and primordial germ cells were prepared from rainbow trout possessing the green fluorescent protein (GFP) gene driven by trout vasa regulatory regions (pvasa-GFP rainbow trout). The gene-expression profiles of each cell fraction were then compared with a microarray containing cDNAs representing 16,006 genes from several salmonid species. Genes exhibiting high expression for type A spermatogonia relative to above-mentioned other types of gonadal cells were identified and subjected to RT-PCR and quatitative PCR analysis. Since only the rainbow trout notch1 homologue showed significantly high expression in the type A spermatogonia-enriched fraction, we propose that notch1 may be a useful molecular marker for type A spermatogonia. The combination of GFP-dependent FCM and microarray analysis of pvasa-GFP rainbow trout can therefore be applied to the identification of potentially useful molecular markers of germ cells in fish.