Jizu Zhi

Analysis techniques for microarray time-series data

We introduce new methods for the analysis of short-term time-series data, and apply them to gene expression data in yeast. These include (1) methods for automated period detection in a predominately cycling data set and (2) phase detection between phase-shifted cyclic data sets. We show how to properly correct for the problem of comparing correlation coefficents between pairs of sequences of different lengths and small alphabets. In particular, we show that the correlation coefficient of sequences over alphabets of size two can exhibit very counter-intuitive behavior when compared with the Hamming distance. Finally, we address the predictability of known regulators via time-series analysis, and show that less than 20% of known regulatory pairs exhibit strong correlations in the Cho/Spellman data sets. By analyzing known regulatory relationships, we designed an edge detection function which identified candidate regulations with greater fidelity than standard correlation methods.

Analysis techniques for microarray time-series data.

We address possible limitations of publicly available data sets of yeast gene expression. We study the predictability of known regulators via time-series analysis, and show that less than 20% of known regulatory pairs exhibit strong correlations in the Cho/Spellman data sets. By analyzing known regulatory relationships, we designed an edge detection function which identified candidate regulations with greater fidelity than standard correlation methods. We develop general methods for integrated analysis of coarse time-series data sets. These include 1) methods for automated period detection in a predominately cycling data set and 2) phase detection between phase-shifted cyclic data sets. We show how to properly correct for the problem of comparing correlation coefficients between pairs of sequences of different lengths and small alphabets. Finally, we note that the correlation coefficient of sequences over alphabets of size two can exhibit very counterintuitive behavior when compared with the Hamming distance.

Pheochromocytomas in Nf1 knockout mice express a neural progenitor gene expression profile

Pheochromocytomas are adrenal medullary tumors that typically occur in adult patients, with increased frequency in multiple endocrine neoplasia type 2, von Hippel-Lindau disease, familial paraganglioma syndromes and neurofibromatosis type 1 (NF1). Pheochromocytomas arise in adult mice with a heterozygous knockout mutation of exon 31 of the murine Nf1 gene, providing a mouse model for pheochromocytoma development in NF1. We performed a microarray-based gene expression profiling study comparing mouse pheochromocytoma tissue to normal adult mouse adrenal medulla to develop a basis for studying the pathobiology of these tumors. The findings demonstrate that pheochromocytomas from adult neurofibromatosis knockout mice express multiple developmentally regulated genes involved in early development of both the CNS and peripheral nervous system. One of the most highly overexpressed genes is receptor tyrosine kinase Ret, which is known to be transiently expressed in the developing adrenal gland, down-regulated in adult adrenals and often overexpressed in human pheochromocytomas. Real-time polymerase chain reaction validated the microarray results and immunoblots confirmed the overexpression of Ret protein. Other highly expressed validated genes include Sox9, which is a neural crest determinant, and Hey 1, which helps to maintain the progenitor status of neural precursors. The findings are consistent with the recently proposed concept that persistent neural progenitors might give rise to pheochromocytomas in adult mouse adrenals and suggest that events predisposing to tumor development might occur before formation of the adrenal medulla or migration of cells from the neural crest. However, the competing possibility that developmentally regulated neural genes arise secondarily to neoplastic transformation cannot be ruled out. In either case, the unique profile of gene expression opens the mouse pheochromocytoma model to new applications pertinent to neural stem cells and suggests potential new targets for treatment of pheochromocytomas or eradication of their precursors.

Human MLPA Probe Design (H-MAPD): a probe design tool for both electrophoresis-based and bead-coupled human multiplex ligation-dependent probe amplification assays

BackgroundMultiplex ligation-dependent probe amplification (MLPA) is an efficient and reliable technique for gene dosage analysis. Currently MLPA can be conducted on two platforms: traditional electrophoresis-based, and FlexMAP bead-coupled. Since its introduction in 2002, MLPA has been rapidly adopted in both clinical and research situations. However, MLPA probe design is a time consuming process requiring many steps that address multiple criteria. There exist only one or two commercial software packages for traditional electrophoresis-based MLPA probe design. To our knowledge, no software is yet available that performs bead-coupled MLPA probe design.ResultsWe have developed H-MAPD, a web-based tool that automates the generation and selection of probes for human genomic MLPA. The software performs physical-chemical property tests using UNAFold software, and uniqueness tests using the UCSC genome browser. H-MAPD supports both traditional electrophoresis-based assays, as well as FlexMAP bead-coupled MLPA.ConclusionH-MAPD greatly reduces the efforts for human genomic MLPA probe design. The software is written in Perl-CGI, hosted on a Linux server, and is freely available to non-commercial users.

Lrp binds to two regions in the dadAX promoter region of Escherichia coli to repress and activate transcription directly

The dadAX operon is expressed by multiple promoters that are repressed by leucine‐responsive regulatory protein (Lrp) and activated by cyclic AMP‐CRP. In previous work, we found that alanine or leucine acted as inducers to antagonize Lrp repression of the three major promoters directly. Here, we identify 11 Lrp binding sites located within 350 bp of dad DNA. A mutational analysis, coupled with in vivo and in vitro transcription experiments, indicated that Lrp sites that overlap the dad promoters were involved in repression. In contrast, sites upstream of the promoters did not appear to be necessary for repression, but were required for activation by Lrp plus alanine or leucine of one of the major dad promoters, P2. This activation by alanine or leucine was not simply relief of repression, as P2 transcription from a constitutive template was increased fivefold compared with the basal level of transcription found in the absence of Lrp and the co‐activator cyclic AMP‐CRP. Alanine or leucine decreased the affinity of Lrp to repressor sites, while having little or no effect on the binding of Lrp to activator sites. This differential effect of alanine and leucine on Lrp binding helps to explain how these modifiers influence both repression and activation of the dad operon.

Differential expression of neuroleukin in osseous tissues and its involvement in mineralization during osteoblast differentiation.

Osteoblast differentiation is a multistep process that involves critical spatial and temporal regulation of cellular processes marked by the presence of a large number of differentially expressed molecules. To identify key functional molecules, we used differential messenger RNA (mRNA) display and compared RNA populations isolated from the defined transition phases (proliferation, matrix formation, and mineralization) of the MC3T3‐E1 osteoblast‐like cell line. Using this approach, a complementary DNA (cDNA) fragment was isolated and identified as neuroleukin (NLK), a multifunctional cytokine also known as autocrine motility factor (AMF), phosphoglucose isomerase (PGI; phosphohexose isomerase [PHI]), and maturation factor (MF). Northern analysis showed NLK temporal expression during MC3T3‐E1 cell differentiation with a 3.5‐fold increase during matrix formation and mineralization. Immunocytochemical studies revealed the presence of NLK in MC3T3‐E1 cells as well as in the surrounding matrix, consistent with a secreted molecule. In contrast, the NLK receptor protein was detected primarily on the cell membrane. In subsequent studies, a high level of NLK expression was identified in osteoblasts and superficial articular chondrocytes in bone of 1‐, 4‐, and 8‐month‐old normal mice, as well as in fibroblasts, proliferating chondrocytes, and osteoblasts within a fracture callus. However, NLK was not evident in hypertrophic chondrocytes or osteocytes. In addition, treatment of MC3T3 cells with 6‐phosphogluconic acid (6PGA; a NLK inhibitor) resulted in diminishing alkaline phosphatase (ALP) activity and mineralization in MC3T3‐E1 cells, especially during the matrix formation stage of differentiating cells. Taken together, these data show specific expression of NLK in discrete populations of bone and cartilage cells and suggest a possible role for this secreted protein in bone development and regeneration.

Proline-rich transcript of the brain (prtb) is a serum-responsive gene in osteoblasts and upregulated during adhesion

To characterize the temporal expression of genes that play a functional role during the process of osteoblast adhesion, we used differential display (DD‐PCR) on mRNA isolated from attached vs. suspended osteoblasts. A 200‐bp fragment displaying upregulated expression after 30 and 60 min adhesion was isolated, sequenced, and showed 97% homology to prtb, previously showed to be expressed in mouse brain. Northern analysis confirmed a two‐fold increase in prtb message during adhesion to tissue culture polystyrene, both in the presence or absence of surface‐adsorbed serum proteins. Serum stimulation alone was also able to induce prtb expression, although to a lesser extent, in suspension cells. Strong prtb expression was also detected in both brain and bone of adult rats. Furthermore, prtb expression analysis during MC3T3‐E1 cell differentiation revealed high expression levels independent of proliferation (day 0–7), matrix maturation (day 7–14), and mineralization (day 14–31). Time course analysis of prtb expression during adhesion of sensitized osteoblasts to serum‐protein coated surfaces showed robust mRNA expression at 5 min post‐plating and a peak at 10 min. The two known serum‐inducible immediate early genes c‐fos and c‐jun showed similar expression kinetics, with c‐jun mRNA levels peaking at 15 min and c‐fos at 20 min. Based on these data, we hypothesize that prtb may function as an immediate early, serum‐responsive, and adhesion‐inducible gene with possible involvement in processes such as cell cycle control, adhesion, and proliferation. J. Cell. Biochem. 84: 301–308, 2002.

In vitro and in vivo characterization of three major dadAX promoters in Escherichia coli that are regulated by cyclic AMP-CRP and Lrp

Abstract We have shown previously that the dadAX operon of Escherichia coli expresses multiple transcripts, which are repressed by leucine-responsive regulatory protein (Lrp). Here we used site-directed mutagenesis and in vitro and in vivo transcription assays to show that each of the three major dad transcripts requires a specific promoter. These promoters, P1-P3, overlap and are positively regulated in vivo by cyclic AMP-CRP. DNase I footprinting experiments localized two CRP binding sites in this region: CRP1, which is positioned upstream of P1-P3, and CRP2, which is located within the promoters. Site-directed mutagenesis of each site provided evidence that CRP1 is necessary for the effects of cyclic AMP-CRP on dad expression in vivo and in vitro, and that CRP2 probably plays little or no role in this process.

Identification of the microRNA transcriptome during the early phases of mammalian fracture repair.

Fracture repair is a complex process that involves multiple biological processes requiring spatiotemporal expression of thousands of genes. The molecular regulation of this process is not completely understood. MicroRNAs (miRNAs) regulate gene expression by promoting mRNA degradation or blocking translation. To identify miRNAs expressed during fracture repair, we generated murine bone fractures and isolated miRNA-enriched RNA from intact and post-fracture day (PFD) 1, 3, 5, 7, 11, and 14 femurs. RNA samples were individually hybridized to mouse miRNA microarrays. Results indicated that 959 (51%) miRNAs were absent while 922 (49%) displayed expression in at least one sample. Of the 922 miRNAs, 306 (33.2%) and 374 (40.6%) were up- and down-regulated, respectively, in the calluses in comparison to intact bone. Additionally, 20 (2.2%) miRNAs displayed combined up- and down-regulated expression within the time course and the remaining 222 (24%) miRNAs did not exhibit any changes between calluses and intact bone. Quantitative-PCR validated the expression of several miRNAs. Further, we identified 2048 and 4782 target genes that were unique to the up- and down-regulated miRNAs, respectively. Gene ontology and pathway enrichment analyses indicated relevant biological processes. These data provide the first complete analysis of the miRNA transcriptome during the early phases of fracture repair.

MrNomogram: A web-based multivariable pediatric renal nomogram

UNLABELLED Multivariable Pediatric Renal Nomogram, or MrNomogram, is a web-based clinical tool for the evaluation of pediatric renal length. Unlike other available age-adjusted renal nomograms, MrNomogram takes into consideration the fact that the renal length is influenced by multiple demographic variables. It provides a more accurate prediction of pediatric renal length, given the patients demographic characteristics, such as age, gender, height, etc. ( AVAILABILITY https://www.prevmed.sunysb.edu/jjc/MrNomogram).