Hum. Reprod. Advance Access published online on July 17, 2007
Human Reproduction, doi:10.1093/humrep/dem214
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Genome-wide expression analysis of cultured trophoblast with trisomy 21 karyotype
1 Department of Internal Medicine ‘T’, Tel Aviv Sourasky Medical Center, Tel Aviv 64239, Israel 2 Genetic Department, Sackler Faculty of Medicine, Tel Aviv University, Tel Aviv 69978, Israel 3 Prenatal Diagnosis Unit, Genetic Institute, Tel Aviv Sourasky Medical Center, 6 Weizmann Street, Tel Aviv 64239, Israel 4 Lis Maternity Hospital, Tel Aviv Sourasky Medical Center, Tel Aviv 64239, Israel
5 Correspondence address. Prenatal Diagnosis Unit, Genetic Institute, Tel Aviv Sourasky Medical Center, 6 Weizmann Street, Tel Aviv 64239, Israel. Tel: +972-3-6973921; Fax: +972-3-6974555; E-mail: yyaron{at}tasmc.health.gov.il
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionAcknowledgementsReferences |
|---|
BACKGROUND: The pathologic features of Down syndrome are assumed to be the result of over-expression of genes located on chromosome 21 and/or a more global transcriptional misregulation that crosses chromosomal borders.
METHODS: To address this issue, four RNA samples from trisomy 21 placentas and four samples from normal first trimester pregnancies were analyzed using Affymetrix U95v2 microarray. Statistical and bioinformatic analyses were employed to compare global gene expression, functional classes, and pathways to differentiate between placentas taken from trisomy 21 and from normal pregnancies.
RESULTS: About 750 genes were significantly over-expressed in trisomy 21. This list contains an 
4.5-fold over-abundance of genes that map to chromosome 21, compared to that which could be expected for this chromosome, on the microarray. Among the classes of genes that best discriminated the trisomy 21 and normal karyotype, we found genes that are also implicated in Alzheimer disease and genes that are associated with ubiquitination and proteosomal degradation. Finally, using the top 10 most discriminating genes, eight samples taken from a different database were correctly classified as either trisomy 21 or normal.
CONCLUSIONS: Our results demonstrate that gene expression in trisomy 21 affected placentas significantly differs from that of chromosomally normal placentas, and this difference is only partially explained by over-expression of genes from chromosome 21. Our findings suggest that specific highly discriminatory genes may be potential targets for further research and development of novel prenatal diagnosis techniques.
Key words: trisomy 21/gene-expression/microarray/bioinformatics/CVS
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionAcknowledgementsReferences |
|---|
Down syndrome (DS) is the most frequent cause of mental retardation due to chromosomal aberration (Merrick, 2000
). The disorder is caused by a complete or partial triplication of human chromosome 21, and is associated with heart and gastrointestinal malformations, endocrine and immunological deficiencies, typical facies and an increased risk of leukemia and early onset Alzheimer disease (AD) (Epstein et al., 1991; Korenberg et al., 1994).
The gene dosage hypothesis relates the phenotypic abnormalities of DS to over-expression of genes located on chromosome 21 (Kurnit, 1979; Korenberg et al., 1994). An alternative model suggests that a small number of over-expressed genes from the 21 trisomic chromosome brings about a secondary, generalized transcriptional misregulation (Korenberg, 1990; FitzPatrick et al., 2002).
Recent developments in microarray technology allow direct assessment of the validity of these models, on a genome-wide basis.
Several studies support the gene dosage effect: Mao et al. (2003) demonstrated up-regulation of genes localized to chromosome 21 from cerebral cortex extracts of fetuses with DS. Similarly, gene expression dosage analysis from a DS mouse model (Ts1Cje) demonstrated an average of 1.5-fold increased expression of genes on a segment of mouse chromosome 16, which is syntenic to human chromosome 21 (Amano et al., 2004). Finally, Giannone et al. (2004) analyzed cDNA from T lymphocytes of DS individuals and demonstrated consistent over-expression of genes located on chromosome 21, also supporting the gene dosage hypothesis.
In contrast, FitzPatric et al. (2002) analyzed amniocytes obtained from DS pregnancies, at 16–18 weeks of pregnancy, but failed to demonstrate over-expression of genes on chromosome 21 (FitzPatrick et al., 2002). Similarly, Chung et al. (2005) analyzed 102 genes in cells taken from trisomy 21 and normal karyotype pregnancies at gestational weeks 16–18, and found differential expression of only 2 out of the tested 24 genes that originate in the 21 chromosome (Chung et al., 2005).
These two hypotheses are not mutually exclusive and it is possible that the DS phenotype is caused in part by genes that are over-expressed from chromosome 21, and in part by generalized misregulation.
This is the first study that addresses this question using genome-wide analysis of placental samples. This tissue carries potential advantages over other tissue sources: first, the placenta is very active metabolically, and thus has a potential of producing many transcripts. Second, the placenta represents a rather early stage of embryonic development, potentially allowing detection of primary alterations, as opposed to tissues at a later stage, in which secondary changes may mask the underlying phenomena. Finally, the study of first trimester placenta may provide targets for development of novel prenatal diagnosis techniques, such as up-regulated or down-regulated genes, the protein products of which could potentially serve as early biochemical markers of aneuploidy.
In the current study, we applied the genome-wide approach to address these issues, using placental samples obtained from women undergoing first trimester chorionic villus sampling (CVS). RNA extracts from cultured trophoblast from trisomy 21 and normal pregnancies were subjected to genome-wide expression analysis using expression microarrays.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionAcknowledgementsReferences |
|---|
Ethics
The study was approved by the local Institutional Review Board (IRB) Helsinki Committee and the National High Committee for Genetic Studies. All patients gave their informed consent.
Samples
Samples were obtained from women who underwent CVS for diagnostic purposes, at the Prenatal Diagnosis Unit, Genetic Institute, Tel Aviv Sourasky Medical Center. Samples were obtained from four pregnancies each with a normal male fetus and four pregnancies each with a trisomy 21 male fetus. Samples were cleaned of potential maternal-cell contamination and were cultured on Chang Medium (Irving Scientific, CA, USA). Each sample was subjected to standard cytogenetic analysis by G-banding. Since only cases with male fetuses were included, cytogenetic analysis of all the samples demonstrated either 46XY or 47XY + 21 karyotypes, so that any significant degree of maternal-cell contamination could be excluded.
The ethical limitations posed by the IRB, mandated that only excess material be used, after initial cytogenetic results were obtained. Thus, excess cell cultures were subjected to 2–4 passages, and were then collected in phosphate-buffered saline (PBS) and immediately stored at –70°C.
Affymetrix GeneChip expression analysis
RNA samples from trisomy 21 and normal CVSs were tested using Affymetrix HG-U95v2 microarrays (Affymetrix Inc., Santa Clara, CA, USA) as described previously (Bar-Shira et al., 2002). Briefly, 10 µg of total RNA were used to synthesize double-stranded cDNA with the Superscript® Choice System (Invitrogen, Carlsbad, CA, USA) and 100 pmole of oligo-dT primer attached to a T7 promoter region (Genset, Paris, France). The double-stranded cDNAs were phenol/chloroform extracted by Phase Lock Gels (Eppendorf, Hamburg, Germany). In vitro transcription (IVT) of the double-stranded cDNA products was performed with an Enzo BioArray High Yield Transcript labeling kit in the presence of biotin-labeled nucleotides (Affymetrix, Inc.). Following purification by an RNeasy mini-column, the biotin-labeled cRNA products were submitted to alkaline treatment (200 mM Tris-acetate pH 8.1, 500 mM potassium acetate, 150 mM magnesium acetate) in order to obtain products below 300 bases. The fragmented cRNAs were hybridized overnight at 45°C to the HG_U95v2 arrays which contains 12,500 transcripts from known genes and ESTs. The array was washed, stained with strepavidin-phycoerythrin and scanned with an Affymetrix Gene-Array® Scanner.
Quantitative real-time PCR
Real-time PCR (RT–PCR) analyses were performed to determine the expression of MEST (mesoderm specific transcript), LOX (lysyl oxidase), MAT2A (methionine adenosyltransferase II, alpha) and GAPDH (glyceraldehyde-3-phosphate dehydrogenase) genes in CVS samples. The amplifications were carried out using LightCycler (Roche Biochemicals, Mannheim, Germany) as described previously (Goldstein et al., 2006). Briefly, a total reaction volume of 10 µl contained FastStart Master mix, 3 mM MgCl2 and 0.5 µM of each primer (see Table 1 for primer sequences and reaction conditions). Fluorescence quantification was calculated using LightCycler software, version 3.01 (Roche Biochemicals). The expression of MEST, LOX and MAT2A genes was normalized using GAPDH expression levels, as described previously (Goldstein et al., 2006).
|
Data analysis and bioinformatics
The algorithm, implanted in Affymetrix Suite Version 5.0 (MAS5, Affymetrix, Inc.) generates a ‘signal value’ (which designates a relative measure of the abundance of the transcript), a ‘detection P-value’ (which indicates the reliability of the transcript's signal values) and a detection call (present, absent or marginal). The ‘detection calls’ were calculated based on ‘detection P-values’ as follows: probe sets with P-value >0.06 were designated as ‘absent’, 0.06> P-value >0.04 as ‘marginal’ and P-value <0.04 as ‘present’. For interarray comparisons, the data from each array was scaled using MAS5 software. The mean intensity for each array was corrected by a scaling factor to a set target intensity of 150.
The bioinformatics analysis was carried out using GeneSpring® version7 software (Silicon Genetics, Redwood City, CA, USA).
Normalization procedure
Values below 0.01 were set to 0.01. Each measurement was divided by the 50th percentile of all measurements in that sample (per chip normalization). The ‘signal value’ for each gene was divided by the median of its measurements in all samples (per gene normalization).
Filtering
Only genes that had a ‘present’ detection call in at least four out of eight samples were chosen for further analysis (5334 genes).
Statistical analysis
Non-supervised hierarchical clustering was applied to the 5334 genes that were initially filtered. Supervised analysis using ANOVA with the source of tissue (trisomy 21/normal) as the independent variable was conducted. The Benjamini and Hochberg false discovery rate (FDR) and Bonferroni methods were used to correct for multiple comparisons.
The results were subjected to further clustering of genes and samples, using hierarchical clustering.
Functional annotation analysis
NetAffex database was used to extract relevant probe sets according to annotation demands (http://www.affymetrix.com/analysis/index.affx).
The analysis was carried out using two programs: WebGestalt tool (http://bioinfo.vanderbilt.edu/webgestalt/) (Zhang et al., 2005), and ‘Pathway Explorer’ (https://pathwayexplorer.genome.tugraz.at) (Mlecnik et al., 2005). With these programs, genes that significantly distinguished between trisomy 21 and normal samples were classified according to function, based on the gene ontology (GO) database. Fisher's exact test was used to detect classes significantly enriched with discriminating genes. In addition, selected known pathways, published on GenMapp (http://www.genmapp.org) and BioCarta (http://www.biocarta.com/genes/index.asp) were also explored.
Validation by class prediction of a published dataset
The 10 most discriminating genes were then used for class prediction of a different published dataset (http://pevsnerlab.kennedykrieger.org/index_ds.htm), to determine the validity of our observations. This dataset contains Affymetrix U133A microarray expression results of cerebral cortex samples from fetuses with either trisomy 21 (n = 4) or normal karyotype (n = 4) (Mao et al., 2003). Analysis was performed using the ‘Support Vector Machine’ algorithm, embedded in the GeneSpring software.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionAcknowledgementsReferences |
|---|
Bioinformatics
Non-supervised analysis
Non-supervised hierarchical clustering, using 5334 genes that were filtered on a non-parametric basis, correctly separated the trisomy 21 samples from the normal ones (Fig. 1A).
|
Supervised analysis
ANOVA was applied separately for each of the 5334 genes that were initially filtered, with the source of tissue (trisomy 21/normal) as the independent variable. This analysis resulted in a subset of 996 genes that demonstrated a significantly different expression pattern in trisomy 21 versus normal samples. Of these, 750 were over-expressed and 246 were under-expressed. Correction for multiple comparisons was performed using the Benjamini and Hochberg FDR algorithm, first with an FDR of 0.1, yielding 315 genes of which 251 were over-expressed and 64 were under-expressed. This list of genes was used for clustering and annotation analysis (discussed later). Subsequently, correction for multiple comparisons was performed again, with more stringent criteria, using an FDR of 0.05. This resulted in 65 genes that had significant differential expression (51 over-expressed and 14 under-expressed, Fig. 1B Table 2). It is of note that none of the first or second trimester maternal serum markers of trisomy 21 was found among the list of genes that were differentially expressed.
|
Annotation analysis by chromosomal location
The ‘gene dosage hypothesis’ predicts that more genes from the trisomic chromosome would be over-expressed, than predicted by chance. Indeed, if the expression pattern in trisomy 21 is random, the number of over-expressed genes for each chromosome would be expected to be proportional to the relative size of the chromosome (i.e. the number of genes on a particular chromosome represented in the array over the total number of genes on the array). Comparison of the observed ratio of over-expressed genes from each chromosome versus the expected ratio may serve as a crude estimate of the validity of the ‘gene dosage hypothesis’.
The microarray contains 12,500 genes, of which 140 map to chromosome 21. Of the 750 genes that were found to be significantly over-expressed in the trisomy 21 samples, 41 genes map to chromsome 21. This shows a 4.58-fold over-representation of genes that map to chromosome 21, than would be expected by chance (
2 = 95.70 df = 1 P < 0.001) (Table 3).
|
GO annotation analysis
Genes that significantly differentiate between trisomy 21 and normal samples were categorized into GO according to their molecular functions, associated biological processes and cellular components. Fisher exact test with a P-value threshold of 0.01 was applied in order to detect the significantly over-represented GO categories. This analysis yielded 65 GO categories that were over-represented. Of note were categories that involve various biological processes connected to oxidative phosphorylation (observed—6 genes, expected—1.7 genes; P < 0.007), the ubiquitine cycle (observed—17 genes, expected—8.7; P < 0.006), purine nucleotide biosynthesis (observed—9 genes, expected—1 gene; P < 0.00001) and rRNA metabolism (observed—4 genes, expected—0.81 genes; P = 0.004).
Biological pathways anotation analysis
The association of the differentially expressed genes to various biological pathways, as published in BioCarta, GenMap and KEGG, was analyzed using PathwayExplorer. This analysis detected 335 genes that map to known biological pathways. Several of these pathways are of particular interest, as they may be related to the pathophysiology of DS: two genes that have been implicated in the pathogenesis of Alzheimer disease (AD) were found to be significantly over-expressed in trisomy 21 trophoblasts: the amyloid beta precursor protein (APP) which maps to chromosome 21 and the amyloid beta precursor protein binding protein 1 (ABPBP1) which maps to chromosome 16. Another pathway that may be of clinical importance is the proteasome degradation pathway, which was found to be significantly enriched with genes that are differentially expressed in trisomy 21.
Validation of microarray results with quantitative RT–PCR
Of the genes that demonstrated the most significant differential expression, three were chosen for further validation by quantitative RT–PCR (qRT–PCR), MEST was chosen because it demonstrated the most significant fold increase. The other two genes (LOX and MAT2A) were chosen arbitrarily from the top 10 most significant genes.
In agreement with the microarray results, MEST was 13.6-fold over-expressed in the trisomic samples compared with chromosomally normal controls (P < 0.036). Similarly, LOX was 3.7-fold over-expressed in the trisomic samples (P = 0.025). In contrast, although MAT2A was found to be under-expressed/down-regulated in trisomy 21 samples in the microarray, qRT–PCR showed this gene to be expressed at a 1.7-fold higher rate in trisomy 21 samples (P = 0.01), (Fig. 2).
|
Validation by class prediction of a published dataset
Classification of a published dataset that contains Affymetrix U133A microarray expression results of cerebral cortex samples from fetuses with either trisomy 21 (n = 4) or normal karyotype (n = 4) (Mao et al., 2003
) was generated by the ‘Support Vector Machine’ algorithm using the genes identified in our current study. On the basis of the 10 most discriminating genes, the algorithm correctly assigned all samples as either trisomy 21 or normal karyotype. |
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionAcknowledgementsReferences |
|---|
It has long been established that the DS phenotype is the result of an extra chromosome 21. To this date, however, the precise mechanisms by which the extra genetic material results in the clinical manifestations are unknown. Over-expression of genes on the extra chromosome 21 has been implicated in the heterogeneous phenotype characteristic of DS. Alternatively, or perhaps additionally, it was also hypothesized that DS is the consequence of misregulation of genes that are not necessarily located on chromosome 21.
In the current study, we addressed this issue by genome-wide expression analysis in cultured, first trimester trophoblast from trisomy 21 and normal pregnancies. Our results demonstrate a substantial difference in gene expression patterns. This was particularly evident from the non-supervised hierarchical clustering analysis. With this approach, the algorithm correctly assigned blinded samples to either ‘trisomy 21’ or ‘normal’ karyotypes, based on the expression pattern on 5334 genes. Furthermore, supervised hierarchical clustering analysis revealed that 20% of the expressed genes (996 of 5334) had a differential expression pattern. Even after correcting for multiple comparisons, using standard bioinformatics methods, 315 genes still demonstrated a statistically significant differential expression pattern.
Over-expressed genes that map to chromosome 21 were found to be over-represented by 4.5-fold than would be expected by chance. Thus, our study only partially supports the ‘gene dosage hypothesis’, as previously proposed (Mao et al., 2003; Giannone et al., 2004). Indeed, most of the differentially-expressed genes (some of which were >10-fold over-expressed) do not map to chromosome 21. This observation calls for alternative models, such as generalized misregulation of gene expression.
The vast number of differentially expressed genes allowed further analyses using functional annotation analysis, which maps differentially expressed genes to known GOs and biological pathways. The results of these analyses offer new insights and perspectives regarding the biomolecular mechanisms underlying the DS phenotype. Interestingly, several key genes that were over-expressed in trisomy 21 trophoblast had a similar expression pattern to that found in brains of AD elderly patients. For example, the amyloid precursor protein (APP, MIM *104760), a key element in the pathogenesis of AD and the source of amyloid plaque deposition, was significantly over-expressed in trisomy 21 trophoblast by 2-fold. Likewise, APP binding protein (APPBP) was also over-expressed in trisomy 21 trophoblast. Moreover, LOX, which has previously been shown to be 3-fold over-expressed in neuritic plaques from AD patients (Gilad et al., 2005), was also over-expressed by 3-fold in trisomy 21 trophoblast. The association of these findings is particularly intriguing, since as many as 50% of DS patients develop AD by the age of 50 (Robbins, 1994).
MEST expression demonstrated the most significant difference between normal and trisomy 21 samples. This was further validated by qRT–PCR, with a 13.7-fold increase in trisomy 21 trophoblast. MEST, also known as paternally expressed gene-1 (Peg-1), is usually expressed in mesoderm-derivatives. No association of MEST with DS has previously been reported. It would be of interest to explore a possible association between over-expression of MEST in DS and the structural abnormalities in mesenchymal-derived organs (heart, intestines) characteristic of DS.
The ubiquitin cycle was one of the GO annotation functional classes highly enriched with genes that discriminate between trisomic and normal samples (expected = 8.7, observed = 17, P < 0.006). This is of particular relevance to the pathogenesis of DS, which may not only be the result of altered gene expression, but may also be caused by epigenetic mechanisms such as post-translational modification by ubiquitination. For example, SIM2 (single-minded 2 gene) is a transcription factor that plays a physiological role in brain development following post-transcriptional ubiquitination (Okui et al., 2005). Moreover, SIM2 has also been associated with etiology of the DS phenotype (Meng et al., 2006). In our study, however, SIM2 was not over-expressed in trisomy 21 trophoblast. It may thus be postulated that abnormal brain development in DS may be influenced, in part, by an altered ubiquitination of SIM2.
Affymetrix gene-expression analysis demonstrated a significant under expression of MAT2A in trisomy 21 samples, whereas qRT–PCR demonstrated a 1.7-fold over-expression of this gene in the same samples. A possible explanation for this discrepency may be attributed to non-specific hybridization of MAT2A transcripts to the Affymetrics microarray. BLAST analysis demonstrated that the MAT2A probe, represented on the Affymetrix array, shares complete identity to a sequence in the early endosome antigen 1 gene on chromosome 12. This may have resulted in a skewed MAT2A expression pattern, and is a limitation of the Affymetrix microarray.
The discovery of differentially expressed genes may yield novel markers for maternal serum biochemical screening of aneuploidy. Gross et al. (2002) used microarray technology on second trimester amniocytes from normal and trisomy 21 pregnancies, and detected seven genes with at least 1.7-fold difference between samples. They suggested that such an approach may contribute to the identification of additional maternal serum biochemical markers in aneuploid pregnancies (Gross et al., 2002). In our study, therefore, we chose to study placental gene expression, since it is a major source of many compounds that are found in maternal serum. Alas, none of the known first or second trimester biochemical markers of aneuploidy was differentially expressed. One possible explanation is that some of these markers are not produced by the placenta: 
-fetoprotein (AFP) is produced in pregnancy by the fetal liver, therefore a difference in placental AFP gene expression is not expected. Similarly, estriol (E3) is a steroid and not a protein encoded for by a gene. Of the three probes for human chorionic gonadotropin (hCG) represented on the Affymetrix U95V2 array, two failed to demonstrate any signal in all of the samples. A possible explanation may be due to the fact that hCG is primarily produced by the syncytiotrophoblast, and following 2–4 passages in culture, cytotrophoblast cells that do not produce hCG, predominate. As for pregnancy associated plasma protein A (PAPP-A), its gene is not represented on the U95V2 microarray, at all.
Our study clearly demonstrated significant differential expression in trisomy 21 as compared with normal trophoblast. In addition, our findings may shed some light on the pathophysiology leading to the DS phenotype. Nonetheless, the study has certain limitations: first, the expression pattern was analyzed in cultured trophoblast. This was both an ethical constrain and a requirement for obtaining sufficient amount of high-quality mRNA. Nevertheless, culturing the trophoblast may have resulted, to some extent, in an altered gene expression profile. However, since all the samples were treated in a similar manner, we safely assume that any potential artifact would not jeopardize our ability to make a valid comparison. Second, we used samples from male fetuses only. This was necessary to rule out maternal-cell contamination. Theoretically, this may have limited our ability to make generalizations regarding female fetuses, but since none of the major pathological phenomena in DS are gender-related, one may safely assume that the effect of this limitation is minimal. Another possible limitation of this study is the use of the U95V2 microarray. Newer generation microarrays may potentially provide additional information (e.g. PAPP-A is represented on the newer U133A microarray).
Using only the 10 most discriminatory genes, it was possible to correctly classify eight samples (trisomic or normal) from a different published dataset (Mao et al., 2003). This validates the significance of our results and suggests that these highly discriminatory genes may be potential targets for further research and development of novel prenatal diagnosis techniques. Moreover, the list of differentially expressed genes could serve as a starting point in the search for such markers. For example, the protein product of integral membrane protein-2B (ITM2B) gene is a type 2 membrane protein that was found to play a role in brain and neuronal development, and is also associated with some forms of dementia (Mead et al., 2000). This gene was found to be almost 2-fold over-expressed in trisomy 21 trophoblast. If, like many other type 2 membrane proteins, a soluble form exists, this study provides a solid basis for exploring its role as a candidate serum marker.
|
Acknowledgements |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionAcknowledgementsReferences |
|---|
The authors thank Prof. Avi Orr-Urtreger, Director of the Genetic Institute, for his invaluable advice and assistance. The authors also thank Ms. Serena Rosner for her insightful comments. This work was supported by M.K.Humanitarian Fund.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionAcknowledgementsReferences |
|---|
Amano K, Sago H, Uchikawa C, Suzuki T, Kotliarova SE, Nukina N, Epstein CJ, Yamakawa K. Dosage-dependent over-expression of genes in the trisomic region of Ts1Cje mouse model for Down syndrome. Hum Mol Genet (2004) 13:1333–1340.
Bar-Shira A, Pinthus JH, Rozovsky U, Goldstein M, Sellers WR, Yaron Y, Eshhar Z, Orr-Urtreger A. Multiple genes in human 20q13 chromosomal region are involved in an advanced prostate cancer xenograft. Cancer Res (2002) 62:6803–6807.
Chung IH, Lee SH, Lee KW, Park SH, Cha KY, Kim NS, Yoo HS, Kim YS, Lee S. Gene expression analysis of cultured amniotic fluid cell with Down syndrome by DNA microarray. J Korean Med Sci (2005) 20:82–87.[Web of Science][Medline]
Epstein ML, Poulsen KT, Thiboldeaux R. Formation of ganglia in the gut of the chick embryo. J Comp Neurol (1991) 307:189–199.[CrossRef][Web of Science][Medline]
FitzPatrick DR, Ramsay J, McGill NI, Shade M, Carothers AD, Hastie ND. Transcriptome analysis of human autosomal trisomy. Hum Mol Genet (2002) 11:3249–3256.
Giannone S, Strippoli P, Vitale L, Casadei R, Canaider S, Lenzi L, D'Addabbo P, Frabetti F, Facchin F, Farina A, et al. Gene expression profile analysis in human T lymphocytes from patients with Down Syndrome. Ann Hum Genet (2004) 68:546–554.[CrossRef][Web of Science][Medline]
Gilad GM, Kagan HM, Gilad VH. Evidence for increased lysyl oxidase, the extracellular matrix-forming enzyme, in Alzheimer's disease brain. Neurosci Lett (2005) 376:210–214.[CrossRef][Web of Science][Medline]
Goldstein M, Meller I, Issakov J, Orr-Urtreger A. Novel genes implicated in embryonal, alveolar, and pleomorphic rhabdomyosarcoma: a cytogenetic and molecular analysis of primary tumors. Neoplasia (2006) 8:332–343.[CrossRef][Web of Science][Medline]
Gross SJ, Ferreira JC, Morrow B, Dar P, Funke B, Khabele D, Merkatz I. Gene expression profile of trisomy 21 placentas: a potential approach for designing noninvasive techniques of prenatal diagnosis. Am J Obstet Gynecol (2002) 187:457–462.[CrossRef][Web of Science][Medline]
Korenberg JR. Molecular mapping of the Down syndrome phenotype. Prog Clin Biol Res (1990) 360:105–115.[Medline]
Korenberg JR, Chen XN, Schipper R, Sun Z, Gonsky R, Gerwehr S, Carpenter N, Daumer C, Dignan P, Disteche C, et al. Down syndrome phenotypes: the consequences of chromosomal imbalance. Proc Natl Acad Sci USA (1994) 91:4997–5001.
Kurnit DM. Down syndrome: gene dosage at the transcriptional level in skin fibroblasts. Proc Natl Acad Sci USA (1979) 76:2372–2375.
Mao R, Zielke CL, Zielke HR, Pevsner J. Global up-regulation of chromosome 21 gene expression in the developing Down syndrome brain. Genomics (2003) 81:457–467.[CrossRef][Web of Science][Medline]
Mead S, James-Galton M, Revesz T, Doshi RB, Harwood G, Pan EL, Ghiso J, Frangione B, Plant G. Familial British dementia with amyloid angiopathy: early clinical, neuropsychological and imaging findings. Brain (2000) 123:975–991.
Meng X, Peng B, Shi J, Zheng Y, Chen H, Zhang J, Li L, Zhang C. Effects of overexpression of Sim2 on spatial memory and expression of synapsin I in rat hippocampus. Cell Biol Int (2006) 30:841–847.[CrossRef][Web of Science][Medline]
Merrick J. Incidence and mortality of Down syndrome. Isr Med Assoc J (2000) 2:25–26.[Web of Science][Medline]
Mlecnik B, Scheideler M, Hackl H, Hartler J, Sanchez-Cabo F, Trajanoski Z. PathwayExplorer: web service for visualizing high-throughput expression data on biological pathways. Nucleic Acids Res (2005) 33:W633–W637.
Okui M, Yamaki A, Takayanagi A, Kudoh J, Shimizu N, Shimizu Y. Transcription factor single-minded 2 (SIM2) is ubiquitinated by the RING-IBR-RING-type E3 ubiquitin ligases. Exp Cell Res (2005) 309:220–228.[CrossRef][Web of Science][Medline]
Robbins LS. Pathologic Basis of Disease. Text Book (1994) 5th edn. 154–157.
Zhang B, Kirov S, Snoddy J. WebGestalt: an integrated system for exploring gene sets in various biological contexts. Nucleic Acids Res (2005) 33:W741–W748.
Submitted on January 18, 2007; resubmitted on June 10, 2007; accepted on June 15, 2007.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  D. K. Slonim, K. Koide, K. L. Johnson, U. Tantravahi, J. M. Cowan, Z. Jarrah, and D. W. Bianchi Functional genomic analysis of amniotic fluid cell-free mRNA suggests that oxidative stress is significant in Down syndrome fetuses PNAS, June 9, 2009; 106(23): 9425 - 9429. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Thomas and M. Fenech Chromosome 17 and 21 aneuploidy in buccal cells is increased with ageing and in Alzheimer's disease Mutagenesis, January 1, 2008; 23(1): 57 - 65. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||