Christine Steinhoff

Gene Expression Changes in the Course of Neural Progenitor Cell Differentiation

The molecular changes underlying neural progenitor differentiation are essentially unknown. We applied cDNA microarrays with 13,627 clones to measure dynamic gene expression changes during the in vitro differentiation of neural progenitor cells that were isolated from the subventricular zone of postnatal day 7 mice and grown in vitro as neurospheres. In two experimental series in which we withdrew epidermal growth factor and added the neurotrophins Neurotrophin-4 or BDNF, four time points were investigated: undifferentiated cells grown as neurospheres, and cells 24, 48, and 96 hr after differentiation. Expression changes of selected genes were confirmed by semiquantitative RT-PCR. Ten different groups of gene expression dynamics obtained by cluster analysis are described. To correlate selected gene expression changes to the localization of respective proteins, we performed immunostainings of cultured neurospheres and of brain sections from adult mice. Our results provide new insights into the genetic program of neural progenitor differentiation and give strong hints to as yet unknown cellular communications within the adult subventricular zone stem cell niche.

Analyzing Gene Expression Time-Courses

Measuring gene expression over time can provide important insights into basic cellular processes. Identifying groups of genes with similar expression time-courses is a crucial first step in the analysis. As biologically relevant groups frequently overlap, due to genes having several distinct roles in those cellular processes, this is a difficult problem for classical clustering methods. We use a mixture model to circumvent this principal problem, with hidden Markov models (HMMs) as effective and flexible components. We show that the ensuing estimation problem can be addressed with additional labeled data partially supervised learning of mixtures - through a modification of the expectation-maximization (EM) algorithm. Good starting points for the mixture estimation are obtained through a modification to Bayesian model merging, which allows us to learn a collection of initial HMMs. We infer groups from mixtures with a simple information-theoretic decoding heuristic, which quantifies the level of ambiguity in group assignment. The effectiveness is shown with high-quality annotation data. As the HMMs we propose capture asynchronous behavior by design, the groups we find are also asynchronous. Synchronous subgroups are obtained from a novel algorithm based on Viterbi paths. We show the suitability of our HMM mixture approach on biological and simulated data and through the favorable comparison with previous approaches. A software implementing the method is freely available under the GPL from http://ghmm.org/gql.

Gene expression profile of mouse bone marrow stromal cells determined by cDNA microarray analysis

Bone marrow stromal cells (BMSC) have gained increased attention because of their multipotency and adult stem cell character. They have been shown to differentiate into other cell types of the mesenchymal lineage and also into non-mesenchymal cells. The exact identity of the original cells, which are isolated from bone marrow by their selective adherence to plastic, remains unknown to date. We have established and characterized mouse BMSC cultures and analyzed three independent samples by cDNA microarrays. The expression profile was compared with two previous expression studies of human BMSC and revealed a high degree of concordance between different techniques and species. To gain clues about the positional context and biology of the isolated cells within the bone marrow stroma, we searched our data for genes that encode proteins of the extracellular matrix, cell adhesion proteins, cytoskeletal proteins and cytokines/cytokine receptors. This analysis revealed a close association of BMSC with vascular cells and indicated that BMSC resemble pericytes.

Long-Term Maintenance of Hematopoietic Stem Cells Does Not Require Contact with Embryo-Derived Stromal Cells in Cocultures

We recently established that two midgestation‐derived stromal clones—UG26‐1B6, urogenital ridge–derived, and EL08‐1D2, embryonic liver–derived—support the maintenance of murine adult bone marrow and human cord blood hematopoietic repopulating stem cells (HSCs). In this study, we investigate whether direct HSC‐stroma contact is required for this stem cell maintenance. Adult bone marrow ckit+ Ly‐6C− side population (K6‐SP) cells and stromal cells were cocultured under contact or noncontact conditions. These experiments showed that HSCs were maintained for at least 4 weeks in culture and that direct contact between HSCs and stromal cells was not required. To find out which factors might be involved in HSC maintenance, we compared the gene expression profile of EL08‐1D2 and UG26‐1B6 with four HSC‐nonsupportive clones. We found that EL08‐1D2 and UG26‐1B6 both expressed 21 genes at a higher level, including the putative secreted factors fibroblast growth factor‐7, insulin‐like growth factor‐binding proteins 3 and 4, pleiotrophin, pentaxin‐related, and thrombospondin 2, whereas 11 genes, including GPX‐3 and HSP27, were expressed at a lower level. In summary, we show for the first time long‐term maintenance of adult bone marrow HSCs in stroma noncontact cultures and identify some secreted molecules that may be involved in this support.

Different Molecular Mechanisms Underlie Placental Overgrowth Phenotypes Caused by Interspecies Hybridization, Cloning, and Esx1 Mutation

To obtain a deeper insight into the genes and gene networks involved in the development of placentopathies, we have assessed global gene expression in three different models of placental hyperplasia caused by interspecies hybridization (IHPD), cloning by nuclear transfer, and mutation of the Esx1 gene, respectively. Comparison of gene expression profiles of approximately 13,000 expressed sequence tags (ESTs) identified specific subsets of genes with changed expression levels in IHPD, cloned, and Esx1 mutant placentas. Of interest, only one gene of known function and one EST of unknown function were found common to all three placentopathies; however, a significant number of ESTs were common to IHPD and cloned placentas. In contrast, only one gene was shared between IHPD and Esx1 mutant, and cloned and Esx1 mutant placentas, respectively. These genes common to different abnormal placental growth genotypes are likely to be important in the occurrence of placentopathy. Developmental Dynamics 230:149–164, 2004.

Transcriptional regulation of the human LINE-1 retrotransposon L1.2B.

Although LINE-1 (L1) sequences constitute the most important family of retrotransposons in the human genome, their transcriptional regulation is poorly understood. Specifically, their unusual internal promoter is incompletely characterized. Current promoter prediction programs fail to identify the promoter in the 5′UTR of the active LINE-1 element L1.2B. Experimental investigation of this promoter using reporter gene assays in various human and murine cell types confirmed that the promoter consists of two segments, and demonstrated that the distal portion is essential for cell-type-independent activity. No differences in promoter activity were found between normal and transformed cells. The complete promoter was shown to possess ≈20% of the activity of the strong early promoter of cytomegalovirus, and to be capable of directing the expression of levels of p53 sufficient to kill normal and transformed human cells. Thus, active LINE-1 elements contain highly active promoters capable of driving cell-type-independent expression, which are of potential use in mammalian expression constructs. In vitro methylation of the promoter at Hpa II sites decreased its activity independently of cell type, but this repression was alleviated in MBD2−/− cells. Surprisingly, mutation of specific Hpa II sites was also found to reduce promoter activity. Thus, efficient repression of the L1.2B promoter by DNA methylation may involve MBD2 binding, but at least one Hpa II site also appears to be involved specifically in transcriptional activation. Since neither promoter activity nor the efficiency of repression by methylation differed between normal and tumor cells, the re-activation of LINE-1 sequences observed in tumor cells is probably caused by hypomethylation of the promoter.

Prospectively isolated CD133/CD24‐positive ependymal cells from the adult spinal cord and lateral ventricle wall differ in their long‐term in vitro self‐renewal and in vivo gene expression

In contrast to ependymal cells located above the subventricular zone (SVZ) of the adult lateral ventricle wall (LVW), adult spinal cord (SC) ependymal cells possess certain neural stem cell characteristics. The molecular basis of this difference is unknown. In this study, antibodies against multiple cell surface markers were applied to isolate pure populations of SC and LVW ependymal cells, which allowed a direct comparison of their in vitro behavior and in vivo gene expression profile. Isolated CD133+/CD24+/CD45−/CD34− ependymal cells from the SC displayed in vitro self‐renewal and differentiation capacity, whereas those from the LVW did not. SC ependymal cells showed a higher expression of several genes involved in cell division, cell cycle regulation, and chromosome stability, which is consistent with a long‐term self‐renewal capacity, and shared certain transcripts with neural stem cells of the embryonic forebrain. They also expressed several retinoic acid (RA)‐regulated genes and responded to RA exposure. LVW ependymal cells showed higher transcript levels of many genes regulated by transforming growth factor‐β family members. Among them were Dlx2, Id2, Hey1, which together with Foxg1 could explain their potential to turn into neuroblasts under certain environmental conditions.

The RNA binding protein Musashi1 regulates apoptosis, gene expression and stress granule formation in urothelial carcinoma cells

The RNA‐binding protein Musashi1 (MSI1) is a marker of progenitor cells in the nervous system functioning as a translational repressor. We detected MSI1 mRNA in several bladder carcinoma cell lines, but not in cultured normal uroepithelial cells, whereas the paralogous MSI2 gene was broadly expressed. Knockdown of MSI1 expression by siRNA induced apoptosis and a severe decline in cell numbers in 5637 bladder carcinoma cells. Microarray analysis of gene expression changes after MSI1 knockdown significantly up‐regulated 735 genes, but down‐regulated only 31. Up‐regulated mRNAs contained a highly significantly greater number and density of Musashi binding sites. Therefore, a much larger set of mRNAs may be regulated by Musashi1, which may affect not only their translation, but also their turnover. The study confirmed p21CIP1 and Numb proteins as targets of Musashi1, suggesting additionally p27KIP1 in cell‐cycle regulation and Jagged‐1 in Notch signalling. A significant number of up‐regulated genes encoded components of stress granules (SGs), an organelle involved in translational regulation and mRNA turnover, and impacting on apoptosis. Accordingly, heat shock induced SG formation was augmented by Musashi1 down‐regulation. Our data show that ectopic MSI1 expression may contribute to tumorigenesis in selected bladder cancers through multiple mechanisms and reveal a previously unrecognized function of Musashi1 in the regulation of SG formation.

Robust inference of groups in gene expression time-courses using mixtures of HMMs

MOTIVATION Genetic regulation of cellular processes is frequently investigated using large-scale gene expression experiments to observe changes in expression over time. This temporal data poses a challenge to classical distance-based clustering methods due to its horizontal dependencies along the time-axis. We propose to use hidden Markov models (HMMs) to explicitly model these time-dependencies. The HMMs are used in a mixture approach that we show to be superior over clustering. Furthermore, mixtures are a more realistic model of the biological reality, as an unambiguous partitioning of genes into clusters of unique functional assignment is impossible. Use of the mixture increases robustness with respect to noise and allows an inference of groups at varying level of assignment ambiguity. A simple approach, partially supervised learning, allows to benefit from prior biological knowledge during the training. Our method allows simultaneous analysis of cyclic and non-cyclic genes and copes well with noise and missing values. RESULTS We demonstrate biological relevance by detection of phase-specific groupings in HeLa time-course data. A benchmark using simulated data, derived using assumptions independent of those in our method, shows very favorable results compared to the baseline supplied by k-means and two prior approaches implementing model-based clustering. The results stress the benefits of incorporating prior knowledge, whenever available. AVAILABILITY A software package implementing our method is freely available under the GNU general public license (GPL) at http://ghmm.org/gql

Expression profile and transcription factor binding site exploration of imprinted genes in human and mouse

BackgroundIn mammals, imprinted genes are regulated by an epigenetic mechanism that results in parental origin-specific expression. Though allele-specific regulation of imprinted genes has been studied for several individual genes in detail, little is known about their overall tissue-specific expression patterns and interspecies conservation of expression.ResultsWe performed a computational analysis of microarray expression data of imprinted genes in human and mouse placentae and in a variety of adult tissues. For mouse, early embryonic stages were also included. The analysis reveals that imprinted genes are expressed in a broad spectrum of tissues for both species. Overall, the relative tissue-specific expression levels of orthologous imprinted genes in human and mouse are not highly correlated. However, in both species distinctive expression profiles are found in tissues of the endocrine pathways such as adrenal gland, pituitary, pancreas as well as placenta. In mouse, the placental and embryonic expression patterns of imprinted genes are highly similar. Transcription factor binding site (TFBS) prediction reveals correlation of tissue-specific expression patterns and the presence of distinct TFBS signatures in the upstream region of human imprinted genes.ConclusionImprinted genes are broadly expressed pre- and postnatally and do not exhibit a distinct overall expression pattern when compared to non-imprinted genes. The relative expression of most orthologous gene pairs varies significantly between human and mouse suggesting rapid species-specific changes in gene regulation. Distinct expression profiles of imprinted genes are confined to certain human and mouse hormone producing tissues, and placentae. In contrast to the overall variability, distinct expression profiles and enriched TFBS signatures are found in human and mouse endocrine tissues and placentae. This points towards an important role played by imprinted gene regulation in these tissues.