Development of a statistically robust quantification method for microorganisms in mixtures using oligonucleotide microarrays

Accurate predictions of postmortem interval using linear regression analyses of gene meter expression data

In criminal and civil investigations, postmortem interval is used as evidence to help sort out circumstances at the time of human death. Many biological, chemical, and physical indicators can be used to determine the postmortem interval - but most are not accurate. Here, we sought to validate an experimental design to accurately predict the time of death by analyzing the expression of hundreds of upregulated genes in two model organisms, the zebrafish and mouse. In a previous study, the death of healthy adults was conducted under strictly controlled conditions to minimize the effects of confounding factors such as lifestyle and temperature. A total of 74,179 microarray probes were calibrated using the Gene Meter approach and the transcriptional profiles of 1063 genes that significantly increased in abundance were assembled into a time series spanning from life to 48 or 96h postmortem. In this study, the experimental design involved splitting the transcription profiles into training and testing datasets, randomly selecting groups of profiles, determining the modeling parameters of the genes to postmortem time using over- and/or perfectly-defined linear regression analyses, and calculating the fit (R²) and slope of predicted versus actual postmortem times. This design was repeated several thousand to million times to find the top predictive groups of gene transcription profiles. A group of eleven zebrafish genes yielded R² of 1 and a slope of 0.99, while a group of seven mouse liver genes yielded a R² of 0.98 and a slope of 0.97, and seven mouse brain genes yielded a R² of 0.95 and a slope of 0.87. In all cases, groups of gene transcripts yielded better postmortem time predictions than individual gene transcripts. The significance of this study is two-fold: selected groups of gene transcripts provide accurate prediction of postmortem time, and the successfully validated experimental design can now be used to accurately predict postmortem time in cadavers.

Microbial signatures of oral dysbiosis, periodontitis and edentulism revealed by Gene Meter methodology

Conceptual models suggest that certain microorganisms (e.g., the "red" complex) are indicative of a specific disease state (e.g., periodontitis); however, recent studies have questioned the validity of these models. Here, the abundances of 500+ microbial species were determined in 16 patients with clinical signs of one of the following oral conditions: periodontitis, established caries, edentulism, and oral health. Our goal was to determine if the abundances of certain microorganisms reflect dysbiosis or a specific clinical condition that could be used as a 'signature' for dental research. Microbial abundances were determined by the analysis of 138,718 calibrated probes using Gene Meter methodology. Each 16S rRNA gene was targeted by an average of 194 unique probes (n=25nt). The calibration involved diluting pooled gene target samples, hybridizing each dilution to a DNA microarray, and fitting the probe intensities to adsorption models. The fit of the model to the experimental data was used to assess individual and aggregate probe behavior; good fits (R²>0.90) were retained for back-calculating microbial abundances from patient samples. The abundance of a gene was determined from the median of all calibrated individual probes or from the calibrated abundance of all aggregated probes. With the exception of genes with low abundances (<2 arbitrary units), the abundances determined by the different calibrations were highly correlated (r~1.0). Seventeen genera were classified as 'signatures of dysbiosis' because they had significantly higher abundances in patients with periodontitis and edentulism when contrasted with health. Similarly, 13 genera were classified as 'signatures of periodontitis', and 14 genera were classified as 'signatures of edentulism'. The signatures could be used, individually or in combination, to assess the clinical status of a patient (e.g., evaluating treatments such as antibiotic therapies). Comparisons of the same patient samples revealed high false negatives (45%) for next-generation-sequencing results and low false positives (7%) for Gene Meter results.

Effect of non-specific species competition from total RNA on the static mode hybridization response of nanomechanical assays of oligonucleotides

We investigate here the nanomechanical response of microcantilever sensors in real-time for detecting a range of ultra-low concentrations of oligonucleotides in a complex background of total cellular RNA extracts from cell lines without labeling or amplification. Cantilever sensor arrays were functionalized with probe single stranded DNA (ssDNA) and reference ssDNA to obtain a differential signal. They were then exposed to complementary target ssDNA strands that were spiked in a fragmented total cellular RNA background in biologically relevant concentrations so as to provide clinically significant analysis. We present a model for prediction of the sensor behavior in competitive backgrounds with parameters that are indicators of the change in nanomechanical response with variation in the target and background concentration. For nanomechanical assays to compete with current technologies it is essential to comprehend such responses with eventual impact on areas like understanding non-coding RNA pharmacokinetics, nucleic acid biomarker assays and miRNA quantification for disease monitoring and diagnosis to mention a few. Additionally, we also achieved a femtomolar sensitivity limit for online oligonucleotide detection in a non-competitive environment with these sensors.

Identification of non-specific hybridization using an empirical equation fitted to non-equilibrium dissociation curves

Non-equilibrium dissociation curves (NEDCs) have the potential to identify non-specific hybridizations on high throughput, diagnostic microarrays. We report a simple method for the identification of non-specific signals by using a new parameter that does not rely on comparison of perfect match and mismatch dissociations. The parameter is the ratio of specific dissociation temperature (T(d-w)) to theoretical melting temperature (T(m)) and can be obtained by automated fitting of a four-parameter, sigmoid, empirical equation to the thousands of curves generated in a typical experiment. The curves fit perfect match NEDCs from an initial experiment with an R(2) of 0.998±0.006 and root mean square of 108±91 fluorescent units. Receiver operating characteristic curve analysis showed low temperature hybridization signals (20-48°C) to be as effective as area under the curve as primary data filters. Evaluation of three datasets that target 16S rRNA and functional genes with varying degrees of target sequence similarity showed that filtering out hybridizations with T(d-w)/T(m)<0.78 greatly reduced false positive results. In conclusion, T(d-w)/T(m) successfully screened many non-specific hybridizations that could not be identified using single temperature signal intensities alone, while the empirical modeling allowed a simplified approach to the high throughput analysis of thousands of NEDCs.

Beyond Affymetrix arrays: expanding the set of known hybridization isotherms and observing pre-wash signal intensities

Microarray hybridization studies have attributed the nonlinearity of hybridization isotherms to probe saturation and post-hybridization washing. Both processes are thought to distort ‘true’ target abundance because immobilized probes are saturated with excess target and stringent washing removes loosely bound targets. Yet the paucity of studies aimed at understanding hybridization and dissociation makes it difficult to align physicochemical theory to microarray results. To fill the void, we first examined hybridization isotherms generated on different microarray platforms using a ribosomal RNA target and then investigated hybridization signals at equilibrium and after stringent wash. Hybridization signal at equilibrium was achieved by treating the microarray with isopropanol, which prevents nucleic acids from dissolving into solution. Our results suggest that (i) the shape of hybridization isotherms varied by microarray platform with some being hyperbolic or linear, and others following a power-law; (ii) at equilibrium, fluorescent signal of different probes hybridized to the same target were not similar even with excess of target and (iii) the amount of target removed by stringent washing depended upon the hybridization time, the probe sequence and the presence/absence of nonspecific targets. Possible physicochemical interpretations of the results and future studies are discussed.

Simultaneous quantification of multiple nucleic acid targets in complex rRNA mixtures using high density microarrays and nonspecific hybridization as a source of information

To date, it has been problematic to accurately quantify multiple nucleic acid sequences, representing microbial targets, in multi-target mixtures using oligonucleotide microarrays, primarily due to nonspecific target binding (i.e., cross-hybridization). While some studies ignore the effects of nonspecific binding, other studies have developed approaches to minimize nonspecific binding, such as physical modeling to design highly specific probes, subtracting nonspecific signal using mismatch probes, and/or removing nonspecific duplexes by scanning through a range of wash stringencies. We have developed an alternative approach that, in contrast to previous approaches, uses nonspecific target binding as a source of information. Specifically, the new approach uses hybridization patterns (fingerprints) to quantify specific nucleic acid targets in complex target mixtures. We evaluated the approach by mixing together in vitro transcribed 28S rRNA targets at varying concentrations (up to 1.0 nM), and hybridizing the 24 mixtures to microarrays (n=3160 probes, in duplicate). Three independent Latin-square-designed experiments revealed accurate quantification of the targets. The regression between actual concentration of targets and those determined by the approach were highly positively correlated with high R(2) values (e.g., R(2)=0.90, n=6 targets; R(2)=0.84, n=8 targets; R(2)=0.82, n=10 targets).

Revision of the nonequilibrium thermal dissociation and stringent washing approaches for identification of mixed nucleic acid targets by microarrays

Microarray experiments typically involve washing steps that remove hybridized nonspecific targets with the purpose of improving the signal-to-noise ratio. The quality of washing ultimately affects downstream analysis of the microarray and interpretation. The paucity of fundamental studies directed towards understanding the dissociation of mixed targets from microarrays makes the development of meaningful washing/dissociation protocols difficult. To fill the void, we examined activation energies and preexponential coefficients of 47 perfect match (PM) and double-mismatch (MM) duplex pairs to discover that there was no statistical difference between the kinetics of the PM and MM duplexes. Based on these findings, we evaluated the nonequilibrium thermal dissociation (NTD) approach, which has been used to identify specific microbial targets in mixed target samples. We found that the major premises for various washing protocols and the NTD approach might be seriously compromised because: (i) nonspecific duplexes do not always dissociate before specific ones, and (ii) the relationship between dissociation rates of the PM and MM duplexes depends on temperature and duplex sequence. Specifically for the NTD, we show that previously suggested use of reference curves, indices of curves and temperature ramps lead to erroneous conclusions.