J. Keith Vass

Regulation of a multigenic invasion programme by the transcription factor, AP-1: re-expression of a down-regulated gene, TSC-36, inhibits invasion.

The transcription factor AP-1 (activator protein-1) is required for transformation by many oncogenes, which function upstream of it in the growth factor-ras signal transduction pathway. Previously, we proposed that one role of AP-1 in transformation is to regulate the expression of a multigenic invasion programme. As a test of this proposal we sought to identify AP-1 regulated genes based upon their differential expression in 208F rat fibroblasts transformed by FBR-v-fos (FBR), and to determine if they functioned in the invasion programme. Subtracted cDNA libraries specific for up- or down-regulated genes in FBRs compared to 208Fs were constructed and analysed. Northern analysis revealed that the cDNAs in both libraries represented differentially expressed genes. Nucleic acid sequence analysis of randomly selected cDNA clones from each library coupled with searches of nucleic acid and amino acid sequence databases determined that many of the cDNAs represented proteins that function in various aspects of the invasion process. Functional analysis of one the down-regulated genes, TSC-36/follistatin-related protein (TSC-36/Frp), which has not previously been associated with invasion, demonstrated that its expression in FBRs inhibited in vitro invasion. These results support the proposal that AP-1 in transformed cells regulates a multigenic invasion programme.

Divergent routes to oral cancer.

Most head and neck squamous cell carcinoma (HNSCC) patients present with late-stage cancers, which are difficult to treat. Therefore, early diagnosis of high-risk premalignant lesions and incipient cancers is important. HNSCC is currently perceived as a single progression mechanism, resulting in immortal invasive cancers. However, we have found that approximately 40% of primary oral SCCs are mortal in culture, and these have a better prognosis. About 60% of oral premalignancies (dysplasias) are also mortal. The mortal and immortal tumors are generated in vivo as judged by p53 mutations and loss of p16(INK4A) expression being found only in the original tumors from which the immortal cultures were derived. To investigate the relationships of dysplasias to SCCs, we did microarray analysis of primary cultures of 4 normal oral mucosa biopsies, 19 dysplasias, and 16 SCCs. Spectral clustering using the singular value decomposition and other bioinformatic techniques showed that development of mortal and immortal SCCs involves distinct transcriptional changes. Both SCC classes share most of the transcriptional changes found in their respective dysplasias but have additional changes. Moreover, high-risk dysplasias that subsequently progress to SCCs more closely resemble SCCs than nonprogressing dysplasias. This indicates for the first time that there are divergent mortal and immortal pathways for oral SCC development via intermediate dysplasias. We believe that this new information may lead to new ways of classifying HNSCC in relation to prognosis.

Effect of iron and retinoic acid on the control of transferrin receptor and ferritin in the human promonocytic cell line U937

The effect of changes in iron availability and induction of differentiation on transferrin receptor expression and ferritin levels has been examined in the promonocytic cell line U937. Addition of iron (as 200 micrograms/ml saturated transferrin) or retinoic acid (1 microM) both caused approx. 70% reduction in the average number of surface transferrin receptors, while the iron chelator desferrioxamine caused an 84% increase. Comparable changes also occurred in the levels of transferrin receptor mRNA. Neither iron nor retinoic acid significantly altered the half-life of transferrin receptor mRNA in the presence of actinomycin D (approx. 75 min) but a 10-fold increase in stability occurred in the presence of desferrioxamine. Iron and retinoic acid both caused an increase in intracellular ferritin levels (approx. 4-and 3-fold, respectively), while desferrioxamine reduced ferritin levels by approx. two-thirds. The effect of iron and retinoic acid added together did not differ greatly from that of each agent alone. None of the treatments greatly affected levels of L-ferritin mRNA. Virtually no H-ferritin mRNA was detected in U937 cells. These results show that changes in ferritin and transferrin receptor caused by treatment with retinoic acid are similar to those induced by excess iron, and suggest that changes in these proteins during cell differentiation are due to redistribution of intracellular iron into the regulatory pool(s), rather than to iron-independent mechanisms.

Hypoxia-mediated control of HIF/ARNT machinery in epidermal keratinocytes.

Transcriptional activity of hypoxia-induced factor 1 (HIF1) - a heterodimer of HIF1α and ARNT (HIF1β) - is essential for cellular adaptation to environmental stress and plays an important role in skin development, wound healing, tumorigenesis and barrier function. Using primary mouse and human epidermal keratinocytes at ambient or hypoxic (1% O(2)) conditions we studied effects of hypoxia upon HIF protein expression. Significant nuclear levels of ARNT and HIF1α along with high HIF1 activity in normoxic keratinocytes suggest an as yet uncharacterised oxygen-independent role for HIF pathway in the epidermis. Acute hypoxia results in an instant but transient increase of HIF1α protein accompanied by a gradual decrease in its mRNA, while ARNT expression remains unchanged. In prolonged (chronic) hypoxia both HIF1α and Arnt are downregulated along with decline of HIF1 activity. However, expression of classical HIF1 targets such as Selenbp1 and Vegfa remains high. Thus, keratinocytes respond to acute hypoxia with immediate block of HIF1α protein degradation and concomitant increase of HIF activity, while under chronic hypoxia pro-angiogenic signalling is maintained through HIF1-independent pathway(s). Decline of HIF1α during chronic exposure is controlled at both mRNA and protein levels, while Arnt is downregulated post-translationally. Distinct transcription levels of Hif1α and Hif3α splice variants under normoxia and their differential response to hypoxia suggest functional diversity of Hif-α isoforms and highlight the complexity of HIF machinery control in epidermal keratinocytes.

Simultaneous Non-Negative Matrix Factorization for Multiple Large Scale Gene Expression Datasets in Toxicology

Non-negative matrix factorization is a useful tool for reducing the dimension of large datasets. This work considers simultaneous non-negative matrix factorization of multiple sources of data. In particular, we perform the first study that involves more than two datasets. We discuss the algorithmic issues required to convert the approach into a practical computational tool and apply the technique to new gene expression data quantifying the molecular changes in four tissue types due to different dosages of an experimental panPPAR agonist in mouse. This study is of interest in toxicology because, whilst PPARs form potential therapeutic targets for diabetes, it is known that they can induce serious side-effects. Our results show that the practical simultaneous non-negative matrix factorization developed here can add value to the data analysis. In particular, we find that factorizing the data as a single object allows us to distinguish between the four tissue types, but does not correctly reproduce the known dosage level groups. Applying our new approach, which treats the four tissue types as providing distinct, but related, datasets, we find that the dosage level groups are respected. The new algorithm then provides separate gene list orderings that can be studied for each tissue type, and compared with the ordering arising from the single factorization. We find that many of our conclusions can be corroborated with known biological behaviour, and others offer new insights into the toxicological effects. Overall, the algorithm shows promise for early detection of toxicity in the drug discovery process.

Methylation Linear Discriminant Analysis (MLDA) for identifying differentially methylated CpG islands

BackgroundHypermethylation of promoter CpG islands is strongly correlated to transcriptional gene silencing and epigenetic maintenance of the silenced state. As well as its role in tumor development, CpG island methylation contributes to the acquisition of resistance to chemotherapy. Differential Methylation Hybridisation (DMH) is one technique used for genome-wide DNA methylation analysis. The study of such microarray data sets should ideally account for the specific biological features of DNA methylation and the non-symmetrical distribution of the ratios of unmethylated and methylated sequences hybridised on the array. We have therefore developed a novel algorithm tailored to this type of data, Methylation Linear Discriminant Analysis (MLDA).ResultsMLDA was programmed in R (version 2.7.0) and the package is available at CRAN [1]. This approach utilizes linear regression models of non-normalised hybridisation data to define methylation status. Log-transformed signal intensities of unmethylated controls on the microarray are used as a reference. The signal intensities of DNA samples digested with methylation sensitive restriction enzymes and mock digested are then transformed to the likelihood of a locus being methylated using this reference. We tested the ability of MLDA to identify loci differentially methylated as analysed by DMH between cisplatin sensitive and resistant ovarian cancer cell lines. MLDA identified 115 differentially methylated loci and 23 out of 26 of these loci have been independently validated by Methylation Specific PCR and/or bisulphite pyrosequencing.ConclusionMLDA has advantages for analyzing methylation data from CpG island microarrays, since there is a clear rational for the definition of methylation status, it uses DMH data without between-group normalisation and is less influenced by cross-hybridisation of loci. The MLDA algorithm successfully identified differentially methylated loci between two classes of samples analysed by DMH using CpG island microarrays.

Hematopoietic Lineage Transcriptome Stability and Representation in PAXgene Collected Peripheral Blood Utilising SPIA Single-Stranded cDNA Probes for Microarray.

Peripheral blood as a surrogate tissue for transcriptome profiling holds great promise for the discovery of diagnostic and prognostic disease biomarkers, particularly when target tissues of disease are not readily available. To maximize the reliability of gene expression data generated from clinical blood samples, both the sample collection and the microarray probe generation methods should be optimized to provide stabilized, reproducible and representative gene expression profiles faithfully representing the transcriptional profiles of the constituent blood cell types present in the circulation. Given the increasing innovation in this field in recent years, we investigated a combination of methodological advances in both RNA stabilisation and microarray probe generation with the goal of achieving robust, reliable and representative transcriptional profiles from whole blood. To assess the whole blood profiles, the transcriptomes of purified blood cell types were measured and compared with the global transcriptomes measured in whole blood. The results demonstrate that a combination of PAXgene™ RNA stabilising technology and single-stranded cDNA probe generation afforded by the NuGEN Ovation RNA amplification system V2™ enables an approach that yields faithful representation of specific hematopoietic cell lineage transcriptomes in whole blood without the necessity for prior sample fractionation, cell enrichment or globin reduction. Storage stability assessments of the PAXgene™ blood samples also advocate a short, fixed room temperature storage time for all PAXgene™ blood samples collected for the purposes of global transcriptional profiling in clinical studies.

Multidimensional partitioning and bi-partitioning: analysis and application to gene expression data sets

Abstract Eigenvectors and, more generally, singular vectors, have proved to be useful tools for data mining and dimension reduction. Spectral clustering and reordering algorithms have been designed and implemented in many disciplines, and they can be motivated from several different standpoints. Here we give a general, unified derivation from an applied linear algebra perspective. We use a variational approach that has the benefit of (a) naturally introducing an appropriate scaling, (b) allowing for a solution in any desired dimension, and (c) dealing with both the clustering and bi-clustering issues in the same framework. The motivation and analysis is then backed up with examples involving two large data sets from modern, high-throughput, experimental cell biology. Here, the objects of interest are genes and tissue samples, and the experimental data represents gene activity. We show that looking beyond the dominant, or Fiedler, direction reveals important information.

Discretization Provides a Conceptually Simple Tool to Build Expression Networks

Biomarker identification, using network methods, depends on finding regular co-expression patterns; the overall connectivity is of greater importance than any single relationship. A second requirement is a simple algorithm for ranking patients on how relevant a gene-set is. For both of these requirements discretized data helps to first identify gene cliques, and then to stratify patients. We explore a biologically intuitive discretization technique which codes genes as up- or down-regulated, with values close to the mean set as unchanged; this allows a richer description of relationships between genes than can be achieved by positive and negative correlation. We find a close agreement between our results and the template gene-interactions used to build synthetic microarray-like data by SynTReN, which synthesizes “microarray” data using known relationships which are successfully identified by our method. We are able to split positive co-regulation into up-together and down-together and negative co-regulation is considered as directed up-down relationships. In some cases these exist in only one direction, with real data, but not with the synthetic data. We illustrate our approach using two studies on white blood cells and derived immortalized cell lines and compare the approach with standard correlation-based computations. No attempt is made to distinguish possible causal links as the search for biomarkers would be crippled by losing highly significant co-expression relationships. This contrasts with approaches like ARACNE and IRIS. The method is illustrated with an analysis of gene-expression for energy metabolism pathways. For each discovered relationship we are able to identify the samples on which this is based in the discretized sample-gene matrix, along with a simplified view of the patterns of gene expression; this helps to dissect the gene-sample relevant to a research topic - identifying sets of co-regulated and anti-regulated genes and the samples or patients in which this relationship occurs.

Discovering bipartite substructure in directed networks

Bipartivity is an important network concept that can be applied to nodes, edges and communities. Here we focus on directed networks and look for subnetworks made up of two distinct groups of nodes, connected by “one-way” links. We show that a spectral approach can be used to find hidden substructure of this form. Theoretical support is given for the idealised case where there is limited overlap between subnetworks. Numerical experiments show that the approach is robust to spurious and missing edges. A key application of this work is in the analysis of high-throughput gene expression data, and we give an example where a biologically meaningful directed bipartite subnetwork is found from a cancer microarray dataset.