Evaluation of external RNA controls for the assessment of microarray performance.

External RNA controls (ERCs), although important for microarray assay performance assessment, have yet to be fully implemented in the research community. As part of the MicroArray Quality Control (MAQC) study, two types of ERCs were implemented and evaluated; one was added to the total RNA in the samples before amplification and labeling; the other was added to the copyRNAs (cRNAs) before hybridization. ERC concentration-response curves were used across multiple commercial microarray platforms to identify problematic assays and potential sources of variation in the analytical process. In addition, the behavior of different ERC types was investigated, resulting in several important observations, such as the sample-dependent attributes of performance and the potential of using these control RNAs in a combinatorial fashion. This multiplatform investigation of the behavior and utility of ERCs provides a basis for articulating specific recommendations for their future use in evaluating assay performance across multiple platforms.

Using RNA sample titrations to assess microarray platform performance and normalization techniques.

We have assessed the utility of RNA titration samples for evaluating microarray platform performance and the impact of different normalization methods on the results obtained. As part of the MicroArray Quality Control project, we investigated the performance of five commercial microarray platforms using two independent RNA samples and two titration mixtures of these samples. Focusing on 12,091 genes common across all platforms, we determined the ability of each platform to detect the correct titration response across the samples. Global deviations from the response predicted by the titration ratios were observed. These differences could be explained by variations in relative amounts of messenger RNA as a fraction of total RNA between the two independent samples. Overall, both the qualitative and quantitative correspondence across platforms was high. In summary, titration samples may be regarded as a valuable tool, not only for assessing microarray platform performance and different analysis methods, but also for determining some underlying biological features of the samples.

The MicroArray Quality Control (MAQC) project shows inter- and intraplatform reproducibility of gene expression measurements.

Over the last decade, the introduction of microarray technology has had a profound impact on gene expression research. The publication of studies with dissimilar or altogether contradictory results, obtained using different microarray platforms to analyze identical RNA samples, has raised concerns about the reliability of this technology. The MicroArray Quality Control (MAQC) project was initiated to address these concerns, as well as other performance and data analysis issues. Expression data on four titration pools from two distinct reference RNA samples were generated at multiple test sites using a variety of microarray-based and alternative technology platforms. Here we describe the experimental design and probe mapping efforts behind the MAQC project. We show intraplatform consistency across test sites as well as a high level of interplatform concordance in terms of genes identified as differentially expressed. This study provides a resource that represents an important first step toward establishing a framework for the use of microarrays in clinical and regulatory settings.

The balance of reproducibility, sensitivity, and specificity of lists of differentially expressed genes in microarray studies.

BACKGROUND: Reproducibility is a fundamental requirement in scientific experiments. Some recent publications have claimed that microarrays are unreliable because lists of differentially expressed genes (DEGs) are not reproducible in similar experiments. Meanwhile, new statistical methods for identifying DEGs continue to appear in the scientific literature. The resultant variety of existing and emerging methods exacerbates confusion and continuing debate in the microarray community on the appropriate choice of methods for identifying reliable DEG lists. RESULTS: Using the data sets generated by the MicroArray Quality Control (MAQC) project, we investigated the impact on the reproducibility of DEG lists of a few widely used gene selection procedures. We present comprehensive results from inter-site comparisons using the same microarray platform, cross-platform comparisons using multiple microarray platforms, and comparisons between microarray results and those from TaqMan - the widely regarded "standard" gene expression platform. Our results demonstrate that (1) previously reported discordance between DEG lists could simply result from ranking and selecting DEGs solely by statistical significance (P) derived from widely used simple t-tests; (2) when fold change (FC) is used as the ranking criterion with a non-stringent P-value cutoff filtering, the DEG lists become much more reproducible, especially when fewer genes are selected as differentially expressed, as is the case in most microarray studies; and (3) the instability of short DEG lists solely based on P-value ranking is an expected mathematical consequence of the high variability of the t-values; the more stringent the P-value threshold, the less reproducible the DEG list is. These observations are also consistent with results from extensive simulation calculations. CONCLUSION: We recommend the use of FC-ranking plus a non-stringent P cutoff as a straightforward and baseline practice in order to generate more reproducible DEG lists. Specifically, the P-value cutoff should not be stringent (too small) and FC should be as large as possible. Our results provide practical guidance to choose the appropriate FC and P-value cutoffs when selecting a given number of DEGs. The FC criterion enhances reproducibility, whereas the P criterion balances sensitivity and specificity.

Mutations causing severe combined immunodeficiency: detection with a custom resequencing microarray.

PURPOSE: Mutation diagnosis of severe combined immunodeficiency is challenging because of the multiplicity of disease genes and large number of disease-causing mutations, including unique ones that continue to be found. A resequencing microarray could facilitate mutation detection, increasing the chance of diagnosing infants early for optimal rescue by hematopoietic stem cell transplantation. METHODS: After analyzing cumulative mutations, we developed a custom Affymetrix GeneChip microarray including probes representing exons and flanking regions of severe combined immunodeficiency disease genes. DNA samples were analyzed by array versus standard dideoxy genomic sequencing. We tested males and their mothers with X-linked IL2RG variants and patients and carriers with autosomal variants in IL7R, JAK3, and DCLRE1C. RESULTS: New, unique severe combined immunodeficiency mutations are frequent. Resequencing array call rates of 95-98% exceeded GeneChip product specifications, and all of 47 point mutations in known samples were detected, as were the sites of 12 of 22 disease-causing insertions and deletions. Each gene had particular nucleotides that were often not called correctly and had to be excluded from analysis; exclusion rates ranged from 0.4% (hemizygous IL2RG) to 9.2% (heterozygous JAK3). CONCLUSION: Microarray resequencing is a promising technology for severe combined immunodeficiency mutation diagnosis that can detect both known and new mutations. Future customization of probe sequences and analysis algorithms could increase the number of accurately called nucleotides.

Expression of serum- and glucocorticoid-inducible kinase is regulated in an experience-dependent manner and can cause dendrite growth.

The interaction of an animal with its environment during a critical period in early postnatal life has lifelong effects on the structure and function of sensory and motor systems. To gain insight into the molecular mechanisms of experience-dependent development, we challenged young rats to adapt to a new environment that engenders novel motor behavior. Rats born in the gravitational field (1G) of the earth subsequently were reared for 2 weeks either in the absence of gravity (microgravity) or at 1G. A comparison of gene expression using microarrays led to the identification of a panel of differentially regulated transcripts. We report here that the abundance of serum- and glucocorticoid-inducible kinase (SGK) is increased in spinal cord tissue from animals reared in microgravity in comparison with 1G-reared controls. The induction of SGK expression also can be achieved by administration of glucocorticoids to animals at 1G or neurons in vitro. Expression of constitutively active SGK in neurons leads to the elaboration of neuronal dendrites and their branching. Glucocorticoids also lead to dendrite elaboration, and this effect can be abrogated by inhibiting SGK activity. Changes in the level of expression of SGK could be part of the mechanism for experience-dependent acquisition of mature neuronal properties.

New developments in high-throughput resequencing and variation detection using high density microarrays.

We developed a high-throughput method for resequencing for single nucleotide polymorphism (SNP) discovery using high-density microarrays. Over the two-year course of this study a number of improvements in sample preparation methods, hybridization assay, array handling, and analysis method were developed and implemented. DNA from 40 unrelated individuals of three different ethnic origins was amplified, labeled, and hybridized to arrays designed with probes representing genomic, coding, and regulatory regions. Protocol improvements including the use of long PCR and semi-automation reduced labeling and fragmentation costs by 33%. Automation improvements include the development of a scanner autoloader for arrays, a faster array wash station, and a linked laboratory tracking and data management system. Validation of a smaller feature size, 20 x 24 microns, allowed the simultaneous screening of 30-kb sense and 30-kb antisense DNA on each microarray, increasing throughput to 1.4 Mb per day per two laboratory personnel. More than 15,000 SNPs were identified in 8.3 Mb of the human genome using high-density resequencing and variation detection arrays (microarrays).

Comparison of frozen and RNALater solid tissue storage methods for use in RNA expression microarrays.

BACKGROUND: Primary human tissues are an invaluable widely used tool for discovery of gene expression patterns which characterize disease states. Tissue processing methods remain unstandardized, leading to unanswered concerns of how to best store collected tissues and maintain reproducibility between laboratories. We subdivided uterine myometrial tissue specimens and stored split aliquots using the most common tissue processing methods (fresh, frozen, RNALater) before comparing quantitative RNA expression profiles on the Affymetrix U133 human expression array. Split samples and inclusion of duplicates within each processing group allowed us to undertake a formal genome-wide analysis comparing the magnitude of result variation contributed by sample source (different patients), processing protocol (fresh vs. frozen vs. 24 or 72 hours RNALater), and random background (duplicates). The dataset was randomly permuted to define a baseline pattern of ANOVA test statistic values against which the observed results could be interpreted. RESULTS: 14,639 of 22,283 genes were expressed in at least one sample. Patient subjects provided the greatest sources of variation in the mixed model ANOVA, with replicates and processing method the least. The magnitude of variation conferred by processing method (24 hours RNALater vs 72 hours RNALater vs. fresh vs frozen) was similar to the variability seen within replicates. Subset analysis of the test statistic according to gene functional class showed that the frequency of "outlier" ANOVA results within each functional class is overall no greater than expected by chance. CONCLUSIONS: Ambient storage of tissues for 24 or 72 hours in RNALater did not contribute any systematic shift in quantitative RNA expression results relative to the alternatives of fresh or frozen tissue. This nontoxic preservative enables decentralized tissue collection for expression array analysis without a requirement for specialized equipment.

Accurate and reproducible gene expression profiles from laser capture microdissection, transcript amplification, and high density oligonucleotide microarray analysis.

Gene expression profiling using high density oligonucleotide arrays is a powerful method to generate an unbiased survey of a cell's transcriptional landscape. Increasingly complex biological questions require that this approach be applicable to the small numbers of cells that are obtained from sources such as laser capture microdissection (LCM) of solid tissues. In this report, we demonstrate that two rounds of transcript amplification can generate accurate and reproducible gene expression profiles using high density oligonucleotide microarrays, starting with as little as 10 ng of total RNA. Biased amplification of the 3' end of transcripts does not have a major impact on the overall transcript profile due to the 3' bias of probe sets incorporated in the array design. Furthermore, greater than 95% of all genes detected demonstrate less than a twofold difference in expression when independent tissue dissections of identical cell populations are compared. The accuracy and technical reproducibility of the method suggests that expression profiling using transcript amplification and high density oligonucleotide microarrays can be used on a routine basis.

Distinctive proliferative phase differences in gene expression in human myometrium and leiomyomata.

OBJECTIVE: To gain a comprehensive view of the gene expression and regulation involved in uterine leiomyomata and matched normal myometrium using oligonucleotide microarray-based hybridization analysis. DESIGN: Retrospective analyses of tissue obtained in a prospective randomized clinical study. SETTING: Academic institution. PATIENT(S): Seven patients with leiomyomata scheduled for surgery during the proliferative phase. INTERVENTIONS(S): Seven paired samples of leiomyomata and adjacent myometrium were obtained from patients undergoing hysterectomy. MAIN OUTCOME MEASURE(S): The total RNA extracted from leiomyomata and myometrium was used for gene expression profiling of 6800 human genes using high-density oligonucleotide microarrays. In addition, reverse transcriptase-semiquantitative polymerase chain reaction and immunohistochemistry were used to validate tumor-specific gene expression. RESULT(S): A comparison of expression patterns in each paired sample revealed 68 genes significantly up- or down-regulated in each paired tissue sample, of which 23 genes showed increased expression and 45 showed decreased expression in leiomyomata compared with normal myometrium. Cluster analysis supported the relevance of these candidate genes for distinguishing between normal myometrium and leiomyomata biologic activity. CONCLUSION(S): Expression profiling of uterine leiomyomata using high-density oligonucleotide microarrays yields signature patterns that reflect the distinctive differences between normal human myometrium and leiomyomata during the proliferative phase. These observations suggest that a number of genes are involved in the tumorigenesis of leiomyomata.

Genomic dissection for characterization of cancerous oral epithelium tissues using transcription profiling.

Genome-wide and high-throughput functional genomic tools offer the potential of identifying disease-associated genes and dissecting disease regulatory patterns. There is a need for a set of systematic bioinformatic tools that handles efficiently a large number of variables for extracting biological meaning from experimental outputs. We present well-characterized statistical tools to discover genes that are differentially expressed between malignant oral epithelial and normal tissues in microarray experiments and to construct a robust classifier using the identified discriminatory genes. Those tools include Wilks' lambda score, error rate estimated from leave-one out cross-validation (LOOCV) and Fisher Discriminant Analysis (FDA). High Density DNA microarrays and Real Time Quantitative PCR were employed for the generation and validation of the transcription profile of the oral cancer and normal samples. We identified 45 genes that are strongly correlated with malignancy. Of the 45 genes identified, six have been previously implicated in the disease, and two are uncharacterized clones.

Tissue handling and specimen preparation in surgical pathology: issues concerning the recovery of nucleic acids from formalin-fixed, paraffin-embedded tissue.

CONTEXT: Expression profiling by microarrays and real-time polymerase chain reaction-based assays is a powerful tool for classification and prognostication of disease; however, it remains a research tool, largely reliant on frozen tissue. Limiting the utility of expression profiling is the isolation of quality nucleic acids from formalin-fixed, paraffin-embedded tissue. The collection, handling, and processing of tissue directly impacts the biomolecules that can be recovered from it. High-quality nucleic acids can be obtained from formalin-fixed, paraffin-embedded tissue, but greater attention to all steps in the process of tissue handling and preparation is required. OBJECTIVE: To summarize the current state-of-the-art of preanalytic factors in tissue handling and processing as they impact the quality of RNA obtainable from formalin-fixed, paraffin-embedded tissue. The goals are to provide recommendations that will improve RNA quality for expression profiling from formalin-fixed, paraffin-embedded tissue and highlight areas for additional research. Tissue is an analyte and it must be handled in a standardized fashion to provide consistent results. DATA SOURCES: The literature was reviewed. Consultation with industry and academic leaders in the use of RNA for expression profiling was obtained to identify areas for additional research. CONCLUSIONS: Development of RNA-based assays from formalin-fixed, paraffin-embedded tissue is feasible. Greater attention to tissue handling and processing is essential to improve the quality of biospecimens for the development of robust RNA-based assays. Standardization of procedures and vigorous testing of alternative protocols are required to ensure that these assays function as designed.

Microarray analysis of differentially expressed genes in vaginal tissues from women with stress urinary incontinence compared with asymptomatic women.

BACKGROUND: The pathophysiology of pelvic floor dysfunction resulting in stress urinary incontinence (SUI) in women is complex. Evidence suggests that there is also a genetic predisposition towards SUI. We sought to identify differentially expressed genes involved in extracellular matrix (ECM) metabolism in vaginal tissues from women with SUI in the secretory phase of menses compared with asymptomatic women. METHODS: Tissue samples were taken from the periurethral vaginal wall of five pairs of premenopausal, age-matched SUI and continent women and subjected to microarray analysis using the GeneChip Human Genome U133 oligonucleotide chip set. RESULTS: Extensive statistical analyses generated a list of 79 differentially expressed genes. Elafin, keratin 16, collagen type XVII and plakophilin 1 were consistently identified as up-regulated ECM genes. Elafin, a serine protease inhibitor involved in the elastin degradation pathway and wound healing, was expressed in pelvic fibroblasts and confirmed by Western blot, quantitative competitive PCR and immunofluorescence cell staining. CONCLUSIONS: Genes involved in elastin metabolism were differentially expressed in vaginal tissue from women with SUI, suggesting that elastin remodelling may be important in the molecular aetiology of SUI.

Screening candidate genes for mutations in patients with hypogonadotropic hypogonadism using custom genome resequencing microarrays.

OBJECTIVE: The purpose of this study was to determine the consistency of calling single nucleotide polymorphisms (SNPs) by custom genome resequencing microarrays compared with capillary DNA sequencing. STUDY DESIGN: Amplified genomic DNA from 23 patients with hypogonadotropic hypogonadism was hybridized to microarrays containing 30 kilobases of sequence from 6 different candidate genes. Capillary DNA sequencing was performed in 10 patients. RESULTS: For 10 patients with > or =90% of bases called, 49 SNPs in 5 of 6 genes were identified. Of the 490 bases, 75 were ambiguous (read as "N"), and 415 were able to be called an A, C, G, or T. Of 415 called, 401 (96.6%) sequences were confirmed by DNA sequencing. All homozygotes (285/285) were called identically, while sequence from 89.2% (116/130) of heterozygotes agreed by both methods. The level of agreement between microarray calls and capillary DNA sequencing demonstrated substantial accuracy. CONCLUSION: Custom genome resequencing microarrays are highly consistent with capillary sequencing in calling individual bases in genomic DNA from patients with human disease.

The External RNA Controls Consortium: a progress report.

Standard controls and best practice guidelines advance acceptance of data from research, preclinical and clinical laboratories by providing a means for evaluating data quality. The External RNA Controls Consortium (ERCC) is developing commonly agreed-upon and tested controls for use in expression assays, a true industry-wide standard control.

Menstrual phase-dependent gene expression differences in periurethral vaginal tissue from women with stress incontinence.

OBJECTIVE: The mechanical stability of the genitourinary tract is dependent on intact collagen fibers that support the bladder neck, urethra, and pelvic organs. We hypothesize that genetic differences in collagen metabolism may contribute to stress urinary incontinence. Because sex hormones have substantial influence on the female lower urinary tract throughout adult life, we investigated the gene expression of vaginal tissue of women with stress incontinence compared with women with no stress incontinence in the proliferative phase of the menstrual cycle. STUDY DESIGN: Quantitative competitive polymerase chain reaction was used to verify that the gene expressions were similar between periurethral vaginal tissue and pelvic ligamentous tissue. Labeled complementary RNA was obtained from periurethral vaginal tissue in five pairs of age- and menstrual phase-matched, premenopausal women with and without stress urinary incontinence. The vaginal tissues were then hybridized on HuGeneFL arrays that contained probes representing 6800 full-length human genes. The Student t test and Mann-Whitney ranking were used independently to select candidates with probability values <.05. Hierarchical clustering analysis was performed on the selected candidates to assess the ability of these genes to discriminate between normal and affected individuals. RESULTS: Tissue inhibitor of metalloproteinases-1 and estrogen receptor-alpha messenger RNA expressions were found to be similar between uterosacral ligament and periurethral vaginal tissue in six participants. Of the 90 candidate genes that were identified, 62 genes were up-regulated and 28 were down-regulated in the stress urinary incontinence group. Genes that were involved in extracellular matrix activity in the up-regulated group include transforming growth factor-beta3, laminin, and collagen type VI. Down-regulated genes that may participate in collagen metabolism include laminin-related protein, collagen XVII, serine/threonine protein kinase, type II interleukin-1 receptor, and platelet-derived growth factor-associated protein. CONCLUSION: In this preliminary study, we identified differential gene expressions that may contribute to extracellular matrix remodeling in pelvic tissue from women with stress urinary incontinence in the proliferative phase versus continent control subjects. The alteration in expression of these candidate genes suggests that they should be targets for further investigation.

BRCA1 transcriptionally regulates genes involved in breast tumorigenesis.

Loss of function of BRCA1 caused by inherited mutation and tissue-specific somatic mutation leads to breast and ovarian cancer. Nearly all BRCA1 germ-line mutations involve truncation or loss of the C-terminal BRCT transcriptional activation domain, suggesting that transcriptional regulation is a critical function of the wild-type gene. The purpose of this project was to determine whether there is a link between the role of BRCA1 in transcriptional regulation and its role in tumor suppression. We developed a cell line (in which BRCA1 can be induced) and used microarray analysis to compare transcription profiles of epithelial cells with low endogenous levels of BRCA1 vs. transcription profiles of cells with 2-4-fold higher induced levels of expression of BRCA1. At these levels of expression, BRCA1 did not induce apoptosis. Undirected cluster analysis of six paired experiments revealed 373 genes, the expression of which was altered significantly and consistently by BRCA1 induction. Expression of 62 genes was altered more than 2-fold. BRCA1-regulated genes associated with breast tumorigenesis included the estrogen-responsive genes MYC and cyclin D1, which are overexpressed in many breast tumors; STAT1 and JAK1, key components of the cytokine signal transduction pathway; the extracellular matrix protein laminin 3A; ID4, an inhibitor of DNA-binding transcriptional activators, which in turn negatively regulates BRCA1 expression; and the prohormone stanniocalcin, expression of which is lost in breast tumor cells. Coordinated expression of BRCA1 with ID4 and with stanniocalcin was confirmed in primary breast and ovarian tumors.

Hepsin and maspin are inversely expressed in laser capture microdissectioned prostate cancer.

PURPOSE: Recent studies have shown that hepsin, a serine protease, is over expressed in prostate cancers, implicating hepsin activity in tumor invasion. Using microarray technology we have previously identified 22 genes that were up-regulated in high grade prostate cancers compared with benign prostatic hyperplasia. Of them hepsin was the most differentially over expressed. In the current report we compare hepsin to maspin (BD Transduction Laboratories, San Diego, California), a serine protease inhibitor (serpin), to measure the balance between levels of serine proteases and serpins, which are considered to be a critical determinant of net proteolytic activity. MATERIALS AND METHODS: We combined the technique of laser capture microdissection with gene expression monitoring by micro-array analysis to investigate the gene expression profiles of prostate cells of different histological types. We also studied maspin immunohistochemically. RESULTS: We observed that hepsin as well as 7 of 22 previously reported up-regulated genes demonstrated a pattern of increasing expression with increasing malignant phenotype. In contrast, the expression of maspin (a serpin) decreased with increasing malignancy of prostate cancers. Using immunohistochemistry we observed that maspin protein is expressed strongly in benign prostatic tissues and slightly in grade 3 prostate cancers, and is absent in grade 4/5 cancers. CONCLUSIONS: We conclude that the increased ratio of hepsin-to-maspin may have an important role in prostate cancer progression and invasion.