Identification and validation of endogenous reference genes for expression profiling of T helper cell differentiation by quantitative real-time RT-PCR.

Real-time RT-PCR method was exploited to identify endogenous reference genes in differentiating human T helper cells. When using this technology in our experimental system, finding a set of genes whose mRNA expression levels would not change appeared to be very challenging. Our initial plan to use the expression level of GAPDH in normalizing the results failed, because the mRNA expression of GAPDH underwent significant changes during the cell culture. Additional studies on the transcription of several other classical housekeeping genes led to similar results. Our second approach was to use results from an extensive survey of gene expression done by oligonucleotide microarrays and to select another panel of genes for testing. This resulted in the identification of three genes whose expression was relatively stable in our experimental system and, therefore, suitable as endogenous reference genes in these cells. The results indicate that the expression level of a constitutively expressed gene may change during the cell culture in vitro, which emphasizes again the importance of carefully validating endogenous control genes for comparative quantification.

High-throughput variation detection and genotyping using microarrays.

The genetic dissection of complex traits may ultimately require a large number of SNPs to be genotyped in multiple individuals who exhibit phenotypic variation in a trait of interest. Microarray technology can enable rapid genotyping of variation specific to study samples. To facilitate their use, we have developed an automated statistical method (ABACUS) to analyze microarray hybridization data and applied this method to Affymetrix Variation Detection Arrays (VDAs). ABACUS provides a quality score to individual genotypes, allowing investigators to focus their attention on sites that give accurate information. We have applied ABACUS to an experiment encompassing 32 autosomal and eight X-linked genomic regions, each consisting of approximately 50 kb of unique sequence spanning a 100-kb region, in 40 humans. At sufficiently high-quality scores, we are able to read approximately 80% of all sites. To assess the accuracy of SNP detection, 108 of 108 SNPs have been experimentally confirmed; an additional 371 SNPs have been confirmed electronically. To access the accuracy of diploid genotypes at segregating autosomal sites, we confirmed 1515 of 1515 homozygous calls, and 420 of 423 (99.29%) heterozygotes. In replicate experiments, consisting of independent amplification of identical samples followed by hybridization to distinct microarrays of the same design, genotyping is highly repeatable. In an autosomal replicate experiment, 813,295 of 813,295 genotypes are called identically (including 351 heterozygotes); at an X-linked locus in males (haploid), 841,236 of 841,236 sites are called identically.

Molecular genetic profiling of Gleason grade 4/5 prostate cancers compared to benign prostatic hyperplasia.

PURPOSE: Because Gleason grade 4/5 cancer is the primary cause of failure to cure prostate cancer, we examined the molecular profiles of this high grade cancer in search of potentially new therapeutic interventions as well as better serum markers than prostate specific antigen. MATERIALS AND METHODS: We compared the gene expressions in fresh frozen tissues from 9 men with Gleason grade 4/5 cancer to 8 men with benign prostatic hyperplasia (BPH) treated with radical retropubic prostatectomy. Labeled complementary RNA from each of the 17 tissues was applied to HuGeneFL probe arrays representing approximately 6,800 genes (Affymetrix, Inc., Santa Clara, California). After removing all genes undetectable in BPH and grade 4/5 cancers, and transforming the data into a parametric distribution, we chose only those up and down-regulated genes with a p difference in fluorescence between grade 4/5 cancer and BPH of p <0.0005. This value reduced the data set to 40 up-regulated and 111 down-regulated genes. We then eliminated all genes that were not expressed in all 8 BPH and 9 grade 4/5 tissues, which produced a final set of 86 genes, of which 22 were up-regulated and 64 were down-regulated. RESULTS: Cluster analysis cleanly separated men with grade 4/5 cancers from those with BPH. Only 17 of the 86 candidate genes (20%) were known to be prostate cancer related and 42 (49%) were related to other cancers. The most up-regulated gene is Hepsin, a trypsin-like serine protease with its enzyme catalytic domain oriented extracellularly. Prostate specific membrane antigen is the second most up-regulated gene (all other reports on prostate specific membrane antigen have been at the protein level). The genes for prostate specific antigen (hK3) and human glandular kallikrein2 (hK2) showed equivalent expression levels 10 times the average of other genes. Complete lists of all 22 up-regulated genes and 64 down-regulated genes, together with their locus on the chromosome, are presented in rank order. CONCLUSIONS: We characterize for the first time 64 down-regulated and 22 up-regulated genes in Gleason grade 4/5 cancer, using the gene profile from BPH as control tissue. A number of interesting new genes, previously undescribed in prostate cancer, are presented as possibilities for further study.

Oral cancer in vivo gene expression profiling assisted by laser capture microdissection and microarray analysis.

Large scale gene expression profiling was carried out on laser capture microdissected (LCM) tumor and normal oral epithelial cells and analysed on high-density oligonucleotide microarrays. About 600 genes were found to be oral cancer associated. These oral cancer associated genes include oncogenes, tumor suppressors, transcription factors, xenobiotic enzymes, metastatic proteins, differentiation markers, and genes that have not been implicated in oral cancer. The database created provides a verifiable global profile of gene expression during oral carcinogenesis, revealing the potential role of known genes as well as genes that have not been previously implicated in oral cancer.

Identification of a mammalian long chain fatty acyl elongase regulated by sterol regulatory element-binding proteins.

Fatty acids are synthesized de novo from acetyl-CoA and malonyl-CoA through a series of reactions mediated by acetyl-CoA carboxylase (ACC) and fatty acid synthase (FAS). In rodents, the principal fatty acid produced by FAS is palmitic acid (16:0). Sterol regulatory element-binding proteins (SREBPs) enhance the transcription of many genes responsible for fatty acid synthesis. In transgenic mice that overexpress SREBPs in liver, the rate of fatty acid synthesis is markedly increased, owing to the activation of these biosynthetic genes, which include ATP citrate lyase, ACC, FAS, and stearoyl-CoA desaturase. The fatty acids that accumulate in livers of SREBP transgenic mice are 18 carbons rather than 16 carbons in length, suggesting that the enzymes required for the elongation of palmitic to stearic acid may be induced. Here, we report the cDNA cloning of a murine long chain fatty acyl elongase (LCE) that was identified initially by oligonucleotide array analysis of mRNA from SREBP transgenic mouse livers. LCE mRNA is highly expressed in liver and adipose tissue. The cDNA encodes a protein of 267 amino acids that shares sequence identity with previously identified very long chain fatty acid elongases. Cells that overexpress LCE show enhanced addition of 2-carbon units to C12-C16 fatty acids. We provide evidence that LCE catalyzes the rate-limiting condensing step in this reaction. The current studies suggest that mouse LCE expression is increased by SREBPs and that the enzyme is a component of the elusive mammalian elongation system that converts palmitic to stearic acid.

A compendium of gene expression in normal human tissues.

This study creates a compendium of gene expression in normal human tissues suitable as a reference for defining basic organ systems biology. Using oligonucleotide microarrays, we analyze 59 samples representing 19 distinct tissue types. Of approximately 7,000 genes analyzed, 451 genes are expressed in all tissue types and designated as housekeeping genes. These genes display significant variation in expression levels among tissues and are sufficient for discerning tissue-specific expression signatures, indicative of fundamental differences in biochemical processes. In addition, subsets of tissue-selective genes are identified that define key biological processes characterizing each organ. This compendium highlights similarities and differences among organ systems and different individuals and also provides a publicly available resource (Human Gene Expression Index, the HuGE Index, http://www.hugeindex.org) for future studies of pathophysiology.

Comparison of human adult and fetal expression and identification of 535 housekeeping/maintenance genes.

Gene expression levels of about 7,000 genes were measured in 11 different human adult and fetal tissues using high-density oligonucleotide arrays to identify genes involved in cellular maintenance. The tissues share a set of 535 transcripts that are turned on early in fetal development and stay on throughout adulthood. Because our goal was to identify genes that are involved in maintaining cellular function in normal individuals, we minimized the effect of individual variation by screening mRNA pooled from many individuals. This information is useful for establishing average expression levels in normal individuals. Additionally, we identified transcripts uniquely expressed in each of the 11 tissues.

The cytokine portion of p43 occupies a central position within the eukaryotic multisynthetase complex.

Multicellular eukaryotes contain a macromolecular assembly of nine aminoacyl-tRNA synthetase activities and three auxiliary proteins. One of these, p43, is the precursor of endothelial monocyte-activating polypeptide II (EMAP II), an inflammatory cytokine involved in apoptotic processes. As a step toward understanding this paradoxical association, the EMAP II portion of p43 has been localized within the rabbit reticulocyte multisynthetase complex. Immunoblot analysis demonstrates strong reaction of anti-EMAP II antiserum with p43, as well as cross-reactivity with isoleucyl-tRNA synthetase. Electron microscopic images of immunocomplexes show two antibody binding sites. The primary site is near the midpoint of the multisynthetase complex at the intersection of the arms with the base. This site near the lower edge of the central cleft is assigned to the C-terminal cytokine portion of p43. The secondary site of antibody binding is in the base of the particle and maps the location of isoleucyl-tRNA synthetase. These data allow refinement of the three-domain model of polypeptide distribution within the multisynthetase complex. Moreover, the central location of p43/EMAP II suggests a role for this polypeptide in optimizing normal function and in rapid disruption of essential cellular machinery when apoptosis is signaled.

Conditional expression of a Gi-coupled receptor causes ventricular conduction delay and a lethal cardiomyopathy.

Cardiomyopathy is a major cause of morbidity and mortality. Ventricular conduction delay, as shown by prolonged deflections in the electrocardiogram caused by delayed ventricular contraction (wide QRS complex), is a common feature of cardiomyopathy and is associated with a poor prognosis. Although the G(i)-signaling pathway is up-regulated in certain cardiomyopathies, previous studies suggested this up-regulation was compensatory rather than a potential cause of the disease. Using the tetracycline transactivator system and a modified G(i)-coupled receptor (Ro1), we provide evidence that increased G(i) signaling in mice can result in a lethal cardiomyopathy associated with a wide QRS complex arrhythmia. Induced expression of Ro1 in adult mice resulted in a >90% mortality rate at 16 wk, whereas suppression of Ro1 expression after 8 wk protected mice from further mortality and allowed partial improvement in systolic function. Results of DNA-array analysis of over 6,000 genes from hearts expressing Ro1 are consistent with hyperactive G(i) signaling. DNA-array analysis also identified known markers of cardiomyopathy and hundreds of previously unknown potential diagnostic markers and therapeutic targets for this syndrome. Our system allows cardiomyopathy to be induced and reversed in adult mice, providing an unprecedented opportunity to dissect the role of G(i) signaling in causing cardiac pathology.

Structural analysis of the multienzyme aminoacyl-tRNA synthetase complex: a three-domain model based on reversible chemical crosslinking.

A subset of eukaryotic aminoacyl-tRNA synthetases (a-RS) are contained in a multienzyme complex for which little structural detail is known. Three reversible chemical crosslinking reagents have been used to investigate the arrangement of polypeptides within this particle as isolated from rabbit reticulocytes. Identification of the crosslinked protein pairs was accomplished by two-dimensional SDS diagonal gel electrophoresis. Seventeen neighboring protein pairs have been identified. Eight are seen with at least two reagents: K-RS:p38, D-RS:K-RS, R-RS dimer, K-RS dimer, K-RS:Q-RS, E/P-RS:K-RS, E/P-RS:I-RS, and Q-RS with one of the nonsynthetase proteins. Nine more are observed with one reagent: D-RS dimer, R-RS:p43, D-RS:Q-RS, D-RS:M-RS, K-RS:L-RS, I-RS:R-RS, D-RS:E/P-RS, I-RS:Q-RS, I-RS:L-RS. One trimeric association is seen: E/P-RS:I-RS:L-RS. The observed neighboring protein pairs suggest that the polypeptides within the aminoacyl-tRNA synthetase complex are distributed in three structural domains of similar mass. These can be arranged in a U-shaped particle in which each "arm" is considered a domain and the third forms the "base" of the structure. The arms have been termed domain I (D-RS, M-RS, Q-RS) and domain II (K-RS, R-RS), with domain III (E/P-RS, I-RS, L-RS) assigned to the base. The smaller proteins (p38, p43) may bridge the domains. This proposed spatial relationship of these domains, as well as their compositions, are consistent with earlier studies. Thus, this study provides an initial three-dimensional working model of the arrangement of polypeptides within the multienzyme aminoacyl-tRNA synthetase complex.

A contiguous high-resolution radiation hybrid map of 44 loci from the distal portion of the long arm of human chromosome 5.

A contiguous high-resolution map of 44 loci from a 35-Mb portion of the distal region of the long arm of human chromosome 5, q21-q35, was produced using radiation hybrid (RH) mapping in conjunction with a natural deletion mapping panel. The map includes 30 genes, four sequence-tagged site (STS) loci, and 10 DNA markers. Newly mapped markers fill two gap regions that were present in previous maps, between markers FER-IL4 and IL3-IL9. Identifying the position of genes on the physical map aids in positional cloning efforts and contributes to our understanding of the overall organization of the human genome.

Progressive myoclonus epilepsy EPM1 locus maps to a 175-kb interval in distal 21q.

The EPM1 locus responsible for progressive myoclonus epilepsy of Unverricht-Lundborg type (MIM 254800) maps to a region in distal chromosome 21q where positional cloning has been hampered by the lack of physical and genetic mapping resolution. We here report the use of a recently constituted contig of cosmid, BAC, and P1 clones that allowed new polymorphic markers to be positioned. These were typed in 53 unrelated disease families from an isolated Finnish population in which a putative single ancestral EPM1 mutation has segregated for an estimated 100 generations. By thus exploiting historical recombinations in haplotype analysis, EPM1 could be assigned to the approximately 175-kb interval between the markers D21S2040 and D21S1259.

Construction of a 750-kb bacterial clone contig and restriction map in the region of human chromosome 21 containing the progressive myoclonus epilepsy gene.

The gene responsible for progressive myoclonus epilepsy of the Unverricht-Lundborg type (EPM1) is located on human chromosome 21q22.3 in a region defined by recombination breakpoints and linkage disequilibrium. As part of an effort to clone the EPM1 gene on the basis of its chromosomal location, we have constructed a 753-kb bacterial clone contig that encompasses the region containing the gene. Because DNA markers from the region did not identify intact yeast artificial chromosome (YAC) clones after screening several libraries, we built the contig from cosmid clones and used bacterial artificial chromosome (BAC) and bacteriophage P1 clones to fill gaps. In addition to constructing the clone contig, we determined the locations of the EcoRI, SacII, EagI, and NotI restriction sites in the clones, resulting in a high-resolution restriction map of the region. Most of the contig is represented by a level of redundancy that allows the orders of most restriction sites to be determined, provides multiple data points supporting the clone orders and orientations, and allows a set of clones with a minimum degree of overlap to be chosen for efficient additional analysis. The clone and restriction maps are in excellent agreement with maps generated of the region by other methods. These ordered bacterial clones and the mapping information obtained from them provide valuable reagents for isolating candidate genes for EPM1, as well as for determining the nucleotide sequence of a 750 kb region of the human genome.

Mutations in the gene encoding cystatin B in progressive myoclonus epilepsy (EPM1)

Progressive myoclonus epilepsy of the Unverricht-Lundborg type (EPM1) is an autosomal recessive inherited form of epilepsy, previously linked to human chromosome 21q22.3. The gene encoding cystatin B was shown to be localized to this region, and levels of messenger RNA encoded by this gene were found to be decreased in cells from affected individuals. Two mutations, a 3' splice site mutation and a stop codon mutation, were identified in the gene encoding cystatin B in EPM1 patients but were not present in unaffected individuals. These results provide evidence that mutations in the gene encoding cystatin B are responsible for the primary defect in patients with EPM1.

High-resolution physical mapping of human 5q31-q33 using three methods: radiation hybrid mapping, interphase fluorescence in situ hybridization, and pulsed-field gel electrophoresis.

Three physical mapping methods, radiation hybrid (RH) mapping, pulsed-field gel electrophoresis (PFGE), and fluorescence in situ hybridization (FISH) of interphase nuclei, were used to determine the order and relative distances between 12 loci in the q31-q33 region of human chromosome 5. The information obtained by each of the methods was compared to determine whether they gave consistent results. Based upon a combination of data, the predicted order of the 12 loci is cen-ADRB2-PDEA-CSF1R-RPS14-ANX6-SPARC++ +-GLRA1-GLUR1-ADRA1B-IL12- GABRG2-GABRA1-tel. Over the 5-Mb region spanned by these loci, we determined that an RH mapping unit (centiray, cR6500) is equivalent to 21 kb, in good agreement with previous estimates of 27-34 kb/cR6500 using the same set of radiation hybrids to map other regions. Of the three methods, RH mapping was by far the easiest and most efficient method for determining both order and distances. In addition, the range of resolution of RH mapping was the broadest, ranging from approximately 200 kb to several megabasepairs. Interphase FISH was the most informative method for clarifying the order of closely linked markers. The FISH distances obtained in this study will be useful as an internal reference for high-resolution mapping studies of other regions on the long arm of human chromosome 5. In contrast to RH mapping and FISH, PFGE was the least reliable of the three methods, producing some data that were inconsistent with the distances determined by the other two methods.

5q- chromosome. Evidence for complex interstitial breaks in a case of refractory anemia with excess blasts.

Interstitial loss of the long arm of chromosome 5 (5q-) is an anomaly frequently seen in myelodysplasia (MDS) and acute myelogenous leukemia (AML). Although the limits of the interstitial deletions vary among patients, there is a critical region of overlap at 5q31 that is consistently deleted in most cases. The order of genes in the critical 5q31 region is centromere, interleukin gene cluster, an anonymous polymorphic locus D5S89, early growth response factor, CSF1 receptor, telomere. Fluorescence in situ hybridization of specific 5q31 probes to metaphases with del(5) (q11q31) from a patient with secondary refractory anemia with excess blasts in transformation demonstrates that the interstitial deletion is not contiguous. The 5q- chromosome has lost the D5S89 and CSF1R loci while retaining some of the sequences in between. A probe derived from a 300-kbp yeast artificial chromosome containing the D5S89 locus is interrupted on the normal chromosome 5 of this patient. Data presented in this report are consistent with (i) presence of a critical gene within the YAC and (ii) more than a single interstitial break within the 5q- chromosome. These results, while pinpointing one of the critical 5q31 loci, also provide evidence for a second telomeric locus.

Genes encoding adrenergic receptors are not clustered on the long arm of human chromosome 5.

The distal portion of the long arm of chromosome 5 (5q) contains a large number of genes encoding membrane receptors belonging to various gene families, including G protein-coupled adrenergic receptors. Previous reports indicated that the genes for two of the adrenergic receptors, ADRB2 and ADRA1B, were within 300 kb of one another on 5q. In an effort to determine if a third adrenergic receptor assigned to 5q, ADRA1A, was physically close to the genes encoding the other adrenergic receptors, we attempted to place all three loci on a radiation hybrid map of 5q. The results conflicted with previous mapping results in two ways. First, ADRA1B is on 5q but is several million bases, rather than a few hundred thousand bases, from ADRB2. Second, ADRA1A is not on chromosome 5, but rather on chromosome 20. Thus, even though 5q contains an extraordinary number of genes encoding receptors for various hormones, growth factors, and neurotransmitters, there is no particular clustering of genes encoding adrenergic receptors in this region.

Consistent loss of the D5S89 locus mapping telomeric to the interleukin gene cluster and centromeric to EGR-1 in patients with 5q- chromosome.

Interstitial deletions of the long arm of chromosome 5 are common in a number of disorders of leukemic and preleukemic myeloid disorders. Although the limits of these deletions vary among patients, a region of cytogenetic overlap that includes band 5q31 is deleted consistently, suggesting loss of 5q31 loci critical for normal myeloid differentiation and leukemogenesis. An anonymous genomic DNA segment D5S89, previously mapped to 5q21-31, detects consistent loss of alleles in cases showing the 5q- chromosome at presentation or relapse. Analysis of a panel of natural-deletion somatic-cell hybrids in conjunction with irradiation hybrids containing fragments of human chromosome 5q shows that the D5S89 locus is telomeric to the interleukin (IL) genes (IL-3, IL-4, IL-5, IL-9, and granulocyte-macrophage colony-stimulating factor [GM-CSF]) and interferon response factor-1 (IRF-1) gene and centromeric to the early response transcription factor (early growth response gene-1 [EGR-1]) on 5q31. To further define the principal region of loss, we have isolated and characterized yeast artificial chromosomes (YACs) spanning D5S89. The presence of several CpG islands within the 300-kb YAC is suggestive of multiple transcription units. However, IL-4, IL-5, IRF-1, IL-3, GM-CSF, and EGR-1 genes were not detected in the YAC clone spanning D5S89, implying that none of these genes are in the vicinity of the D5S89 marker. Further characterization of these YACs should facilitate the isolation of novel candidate genes that may play a role in the evolution of the abnormal phenotype associated with 5q- chromosome.

A physical map of 15 loci on human chromosome 5q23-q33 by two-color fluorescence in situ hybridization.

The q23-q33 region of human chromosome 5 encodes a large number of growth factors, growth factor receptors, and hormone/neurotransmitter receptors. This is also the general region into which several disease genes have been mapped, including diastrophic dysplasia, Treacher Collins syndrome, hereditary startle disease, the myeloid disorders that are associated with the 5q-syndrome, autosomal-dominant forms of hereditary deafness, and limb girdle muscular dystrophy. We have developed a framework physical map of this region using cosmid clones isolated from the Los Alamos arrayed chromosome 5-specific library. Entry points into this library included 14 probes to genes within this interval and one anonymous polymorphic marker locus. A physical map has been constructed using fluorescence in situ hybridization of these cosmids on metaphase and interphase chromosomes, and this is in good agreement with the radiation hybrid map of the region. The derived order of loci across the region is cen-IL4-IL5-IRF1-IL3-IL9-EGR1-CD1 4-FGFA-GRL-D5S207-ADRB2-SPARC-RPS14+ ++-CSF1R- ADRA1, and the total distance spanned by these loci is approximately 15 Mb. The framework map, genomic clones, and contig expansion within 5q23-q33 should provide valuable resources for the eventual isolation of the clinically relevant loci that reside in this region.