Genes differentially expressed in medulloblastoma and fetal brain
E. M. C. MICHIELS1,
E. OUSSOREN2,
M. VAN GROENIGEN2,
E. PAUWS2,
P. M. M. BOSSUYT3,
P. A. VOÛTE1 and
F. BAAS2
1 Department of Pediatric Oncology, Emma Kinderziekenhuis/Academic Medical Center
2 Neurozintuigen Laboratory
3 Department of Clinical Epidemiology, Academic Medical Center, 1100 DE Amsterdam, The Netherlands
 |
ABSTRACT
|
---|
Michiels, E. M. C., E. Oussoren, M. van Groenigen, E. Pauws, P. M. M. Bossuyt, P. A. Voûte, and F. Baas. Genes differentially expressed in medulloblastoma and fetal brain. Physiol. Genomics 1: 8391, 1999.Serial analysis of gene expression (SAGE) was used to identify genes that might be involved in the development or growth of medulloblastoma, a childhood brain tumor. Sequence tags from medulloblastoma (10229) and fetal brain (10692) were determined. The distributions of sequence tags in each population were compared, and for each sequence tag, pairwise
2 test statistics were calculated. Northern blot was used to confirm some of the results obtained by SAGE. For 16 tags, the
2 test statistic was associated with a P value < 10-4. Among those transcripts with a higher expression in medulloblastoma were the genes for ZIC1 protein and the OTX2 gene, both of which are expressed in the cerebellar germinal layers. The high expression of these two genes strongly supports the hypothesis that medulloblastoma arises from the germinal layer of the cerebellum. This analysis shows that SAGE can be used as a rapid differential screening procedure.
serial analysis of gene expression; brain tumors
 |
INTRODUCTION
|
---|
GENE EXPRESSION IN MAMMALIAN CELLS is highly complex: the genome is estimated to contain about 50,000 to 100,000 genes, and the complexity of their transcription is dependent on the type of tissue. Brain is thought to express more than 30,000 genes. Earlier studies on the reannealing kinetics of cDNA suggest that most of the cerebral mRNA molecules are present in few copies per cell (35). Several methods have been used to determine the complexity of gene expression and identify genes that tissues express differentially. Many techniques of subtractive hybridization and differential display (14, 18) are used to identify differences in expression among samples but do not inform us about the abundance of a certain gene or its expression pattern. Moreover, all these methods are technically demanding and time-consuming. Methods based on sequencing of expressed sequence tags (ESTs) (1) only allow us analysis of a limited number of genes. Until recently it was not feasible to obtain information on the majority of genes expressed in cells.
Two new techniques for the analysis of gene expression are serial analysis of gene expression (SAGE) and microarray hybridization. These can now be used because the sequence of tens of thousands of human mRNAs as ESTs is known. In microarray hybridization, a two-color hybridization technique developed by Shena et al. (30), known cDNAs are spotted on a surface and hybridized with differentially labeled cDNA. In this way the expression patterns of clones present on the microarray are analyzed. A limitation of the microarray approach is that only previously identified sequences can be analyzed. This is not a problem for the SAGE technique. SAGE, described by Velculescu et al. (38), is based on the principle that a nucleotide sequence of 910 bp can uniquely identify a transcript, if the position of the sequence within it is known. Briefly, a biotinylated oligo(dT) primer is used to synthesize cDNA from mRNA, and after digestion with a restriction enzyme, the most 3' terminus [near the poly(A) tail] is isolated. These 3' fragments of cDNA are ligated to linkers and cleaved with a type II restriction enzyme to release a short sequence (910 bp) of the original cDNA (tags). The tags are ligated to ditags and PCR-amplified. These ditags are then ligated to form long concatamers, which are cloned and sequenced. In this way, one sequence reaction yields information about the distribution of many different tags. Finally, the calculation of the abundance of the different tags and the matching of the tags in GenBank are done using the necessary computer software.
SAGE was used to compare yeast gene expression in different stages of the cell cycle (39). Because the yeast genome has been completely sequenced and the number of open reading frames amounts to only 7,000 vs. about 50,000100,000 in humans, analysis of a limited number of cDNA tags by SAGE will yield quantitative data on gene expression in yeast. In the mammalian system SAGE was used to compare gene expression profiles in lung cancer (15) and p53-transformed cells (19) in gastrointestinal tumors (42). Because of the complexity of the mammalian genome a large number of tags were analyzed to generate detailed transcription profiles. However, if one were only interested in differentially expressed genes, analysis of a limited number of tags might suffice. De Waard et al. (40) used only 12,000 tags to identify genes induced by atherogenic stimuli in endothelial cells.
In this study we applied SAGE as a differential screening in a mammalian system and confirm that the identification of major differences in gene expression does not require exhaustive analysis of sequences expressed. As model tissues we chose a 24
-wk fetal brain and a medulloblastoma and compared their patterns of gene expression. Medulloblastoma is a central nervous system tumor predominantly of childhood and arising in the cerebellum. Five-year survival of affected children is poor (3970%) despite very intensive therapy consisting of neurosurgery, radiotherapy, and sometimes chemotherapy (9, 11, 22, 27, 36). If the children survive, they often suffer from serious late effects of the tumor and, not in the least, of the treatment. Numerous previous studies have described alterations in the DNA content of medulloblastomas (5, 6, 8, 16, 21, 23, 25, 26, 28, 29, 37). Until now, no studies have been published on gene expression patterns in medulloblastoma. In this study we used SAGE to search for genes that are overexpressed in medulloblastoma and therefore may affect growth or development of the tumor. Identification of transcripts that are highly expressed in medulloblastoma may also reveal specific markers that may assist in the diagnosis.
 |
MATERIALS AND METHODS
|
---|
Fetal brain and medulloblastoma tissue.
Fetal brain was obtained from a 24
-wk-old female fetus (partus immaturus) that showed no abnormalities at obduction.
Medulloblastoma tissue from an 11-yr-old girl with a posterior fossa tumor was flash-frozen in liquid nitrogen immediately after neurosurgical removal and stored at -70°C. Diagnosis of medulloblastoma was confirmed histopathologically according to the World Health Organization classification (17).
SAGE.
The procedure was performed as described by Velculescu et al. (38, 39). From normal brain and tumor tissues total RNA was isolated using Trizol reagent according to the manufacturer's protocol (GIBCO BRL). Poly(A) RNA was obtained using the PolyATract mRNA isolation kit according to the manufacturer's protocol for small-scale mRNA isolation (Promega).
With a cDNA synthesis kit (GIBCO BRL no. 18267013), double-stranded cDNA was synthesized with a biotinylated oligo(dT) primer. The subsequent steps were performed as described by Velculescu et al. (38, 39) until the first PCR of ditags. Of this PCR, 25 cycles were performed. The PCR was analyzed by PAGE, and the desired product was excised. No additional PCR cycles were done. These ditags were kept on ice, and salt (50 mM NaCl) was added to prevent melting of double-stranded ditags. The PCR products were then cleaved with Nla III, and the band containing the ditags was excised from gel and self-ligated. After ligation the concatenated ditags were separated by PAGE, and products of 300600 bp were used for cloning in the Sph I site of pZero (Invitrogen) and transformation into TOP 10F' E. coli electrocompetent cells. Colonies were picked and inoculated into 50 µl of liquid SOB medium containing Zeocin in 96-well plates and grown overnight at 37°C. Two microliters of this culture were used in a PCR with M13 forward and reverse primers. PCR products were run on an agarose gel to check for the presence of an insert before sequencing. Sequencing was done on an ABI 377 XL automatic sequencer (Perkin Elmer) using a DYEnamic ET-T7 primer (Amersham), following the manufacturer's protocol. Analysis of the sequence results was performed using software especially designed for SAGE purposes by Velculescu et al. (38). GenBank release 100.0 and subsequent updates were used to identify matches with known gene sequences and ESTs.
RACE-PCR.
Tags with no homology to known sequences were used as forward primers in a 3' rapid amplification of cDNA ends (RACE)-PCR according to the procedure described by Frohman et al. (13), with 5'-GCATGCCAGAATTCTGGATCC-3' as a reverse primer. Template cDNA was made of mRNA of the SAGE medulloblastoma, with 5'-GCATGCCAGAATTCTGGATCCTTTTTTTTTTTTTTTTTT-3' as primer. The PCR product was run on a 1.5% agarose gel, and the band was isolated and sequenced as described below.
Cloning and sequencing of the probes.
Probes for the Northern blot were obtained by standard PCR on cDNA using the following primers: for ZIC1 (GenBank accession no. D76435), position 27192739 of the mRNA (forward) and position 30193000 (reverse); for secretogranin I (GenBank accession no. Y00064), position 14491468 (forward) and 20942075 (reverse); for GAPDH (GenBank accession no. M33197), position 245264 (forward) and 536517 (reverse). For OTX2 two different probes were constructed: one of ~300 bp using 5'-GCAAAATTCAGAGCAACTGAG-3' as forward and 5'-ATCTGCCAAATCCAGGAAGAA-3' as reverse primer and one of ~600 bp using 5'-TGGGAACAGGATCCAGATTTC-3' as forward and 5'-CTCGACTCGGGCAAGTTGA-3' as reverse primer. Fragments were cloned into a pGEM-T Easy Vector (Promega) according to manufacturer's protocol, and the sequence was confirmed by sequence analysis.
Northern blot.
For RNA blotting 10 µg of total RNA was separated on a glyoxal gel according to standard procedures (2). Total RNA of the fetal brain and medulloblastoma of which the SAGE libraries were constructed and RNA of six other medulloblastomas were used in Figs. 2 and 3, respectively. A
-actin probe was used as control for RNA loading of the lanes (10). Hybridization of the probes to the RNA blots was performed according to Church and Gilbert (7). Hybridized probe was visualized on a PhosphorImager (Molecular Dynamics).

View larger version (43K):
[in this window]
[in a new window]
|
Fig. 2. Northern blot analysis of transcripts that show different expression in fetal brain (FB) and medulloblastoma (MB) by serial analysis of gene expression (SAGE). RNA of fetal brain and medulloblastoma from which the SAGE libraries were constructed was used. Each lane contains 10 µg of total RNA. -Actin was used as control for RNA loading. GAPDH, glyceraldehyde-3-phosphate dehydrogenase.
|
|

View larger version (51K):
[in this window]
[in a new window]
|
Fig. 3. Northern blot analysis of ZIC1 and OTX2 expression in fetal brain and 7 medulloblastomas. Each lane contains 10 µg of total RNA. -Actin was used as control for RNA loading of the lanes. Lanes: FB and M, fetal brain and medulloblastoma from the SAGE library; 38, 6 other medulloblastomas.
|
|
Statistical analysis.
Statistical analysis was done in two steps. First we analyzed whether the two sets of SAGE data had a different distribution of tags. An overall
2 test statistic was calculated for which a P value was obtained through Monte Carlo simulation (StatXact 3 for Windows). The tag distribution in fetal brain and medulloblastoma were found to be statistically different (P < 0.001). In the second step, pairwise
2 statistics were calculated, one for each test statistic. Within the same pair of distributions, more pronounced differences in expression will lead to higher
2 values. Because of the large number of comparisons, the simple P values that correspond to those
-square statistics cannot reliably be used.
 |
RESULTS
|
---|
We constructed and compared SAGE libraries of a fetal brain and a medulloblastoma. We sequenced ~10,000 tags from each tissue. Of the resulting tag populations we then selected the tags that showed a marked difference in expression.
Medulloblastoma.
Of the medulloblastoma, 10,229 tags were sequenced. They represented 5,799 different tags. Of these, 273 appeared five times or more, 1,074 were seen between one and five times, and 4,452 tags occurred only once. Of the 273 tags appearing five times or more, two tags turned out to be linker sequences, occurring 102 and 75 times, respectively. They were excluded from further analysis, which brought the total of tags seen five times or more to 271. Although only 271 of 5,797 different tags (4.7%) occurred five times or more, they represent 29% of the total mass of tags. The low-abundance tags represent 95% of the different tags but only 69% of the total tag mass.
As could be expected, more abundant tags were more likely to match a known gene in GenBank. No matches with known genes were found in ~80% of the single tags, in 49% of tags occurring three times or more, and in 39% of tags occurring at least five times. Table 1 gives an overview of the tags occurring at least 10 times, with their corresponding gene and GenBank accession number. Only the tags that matched with a known gene, for which the complete 3'-untranslated region (3'-UTR) and thus the position in the mRNA sequence are known, are listed.
Fetal brain.
For fetal brain, 10,692 tags were analyzed. These represented 6,423 different tags, of which 264 appeared five times or more, 1,055 tags appeared between one and five times, and 5,104 tags were detected once. The same linker sequences as seen in the medulloblastoma library were found here in a frequency of, respectively, 48 and 28 times, which makes a total of 262 tags occurring five times or more.
As in medulloblastoma, the 262 tags appearing five times or more, which are only 4% of 6,421 different tags, represent 2,994 of the 10,692 tags, which are 28% of the total mass of tags. In contrast, the low-abundance tags (<5 times) represent 96% of the 6,421 different tags but only 71% of the total mass of tags. Of the single tags, ~86% showed no match with a known gene, whereas this was the case for only 47% of tags occurring at least three times and 33% of tags occurring five times or more. Table 2 lists the tags occurring 10 times or more, with their corresponding gene match.
As shown in Fig. 1, more than one-half of the tags that occur at least five times are found in fetal brain as well as in medulloblastoma and thus represent genes that are highly expressed in both tissues.

View larger version (10K):
[in this window]
[in a new window]
|
Fig. 1. Tags occurring at least 5 times in the fetal brain (FB) and medulloblastoma (MB) libraries. The numbers 118, 144, and 127 represent the number of tags that occur in fetal brain only, in both libraries, and in medulloblastoma only, respectively.
|
|
Comparison of medulloblastoma and fetal brain.
The distributions of sequence tags in each population were compared and found to differ significantly (P < 0.001). For each sequence tag, pairwise
2 test statistics were calculated. These test statistics were sorted to obtain a ranking of differences in expression level. Table 3 lists the 138 tags for which the
2 test statistic was associated with a P value < 0.05, with the corresponding GenBank entries. Of these, 67 (54%) matched to a gene in GenBank, although for three of them only the clone number or chromosomal mapping site is known and not the gene itself (tag nos. 82, 84, and 116). Of the 138 tags, 20 contained repeat sequences (mostly an Alu repeat) that resulted in numerous hits. In 45 cases, no match with a known gene was found, but the tags matched to multiple EST sequences. Because the poly(A) tail was not present in these ESTs, we cannot conclude that the tag identified is in fact located at the 3' end of the last Nla III site preceding the poly(A) tail. In one case, only one EST was found to be a possible hit (tag no. 4). Six tags did not match with any known gene, EST, or sequence-tagged site sequence and were designated as "no match" (tag nos. 13, 33, 63, 114, 119, 122); three others matched with a known gene, but these sequences in GenBank contained no poly(A) tail, so proof is missing that this is the correct corresponding transcript (tag nos. 14, 22 and 121). These last three tags are also marked with a question mark in Table 3. As shown in Table 3, two tags match with cytoskeletal
-actin (tag nos. 8 and 20). This turns out to be due to a polymorphism in the last CATG, which is the recognition site of the restriction enzyme used to generate the tags (see DISCUSSION).
Among the genes that show significant higher expression in medulloblastoma is the gene for ZIC1 protein, which is known to be selectively expressed in a very thin layer of brain cells and in medulloblastoma. One of the tags that shows homology to multiple ESTs was analyzed further (tag no. 18). As it was not possible to identify the correct EST because of lack of a poly(A) tail, we performed a RACE-PCR and found a sequence of ~600 bp. The sequence of these 600 bp showed homology with a rat and mouse homeobox gene called Otx2. We cloned and sequenced the full-length mRNA. Translation of this mRNA revealed an amino acid sequence similar to the OTX2 protein already known in mouse and rat and identical to the human protein sequence (12, 32, 33).
Northern blot analysis.
The SAGE results for secretogranin I, ZIC1 protein, OTX2, and GAPDH were checked by Northern blot analysis. As shown in Figs. 2 and 3, the Northern blots indeed show a higher expression of secretogranin I, ZIC1, and OTX2 in medulloblastoma. Only the expression level of GAPDH is slightly higher in fetal brain on Northern blot.
To examine whether the higher expression of ZIC1 and OTX2 also holds true for other medulloblastomas, Northern blot analysis was performed on six other tumors from which RNA was available. Results are shown in Fig. 3. All tumors show a high ZIC1 expression, and four of six medulloblastomas have high OTX2 expression. On a multiple tissue blot, we detected OTX2 in the medulloblastoma and a weak signal in lung. Esophagus, kidney, thyroid gland, testis, peripheral nerve, liver, gallbladder, thymus, tonsil, prostate, adrenal, ovary, muscle, duodenum, brain, stomach, skin, cervix, spleen, breast, colon, and salivary gland show no expression. Adult cerebral cortex V17/V18, white matter, and cerebellum are also negative for OTX2 on Northern blot (data not shown).
 |
DISCUSSION
|
---|
The purpose of this analysis was to identify differences in gene expression between medulloblastoma and fetal brain. More than 30,000 genes are believed to be expressed in human brain (35). In view of this number, a huge number of SAGE tags need to be sequenced to obtain statistically significant information on the complete gene expression profile of human brain cells. However, if one is only interested in major differences in gene expression, a much smaller sample size should be sufficient. As initial comparison for medulloblastoma we chose fetal brain from a partus immaturus that showed no malformations at autopsy. Because medulloblastoma is believed to be derived from the cells of the external germinal layer (EGL), the internal granular layer (IGL), or both (20, 34, 43, 44), a comparison of the EGL and the IGL would be of interest. Because of the very small amount of cells in the EGL and IGL, this is currently technically not possible. Thus, as a first step, we compared the expression pattern of medulloblastoma with that of fetal brain, which contains developing neuronal cells at a stage where there is no myelination. A comparison with adult cerebellum or adult whole brain would result in genes involved in myelination. Sequencing 20,917 tags in total yielded 138 tags that showed significantly different count in the two samples. Fifty-four percent of these tags matched to a known gene in GenBank. Northern blot analysis was used to correlate tag count with expression level.
Some pitfalls have to be considered when analyzing the SAGE data. First, when a hit is found with a known gene in GenBank and the complete coding sequence is known, that does not necessarily mean that the complete 3'-UTR is also known. The EST library has to be screened to identify ESTs that carry the entire 3'-UTR up until the poly(A) tail. Only in that way can one check whether the hit that was found initially is really located at the last cleaving site of the anchoring enzyme that was used to construct the SAGE library. The SAGE software searches for the sequences adjacent to the most-3' located CATG. As a consequence, multiple ESTs are identified as hits. Because for most ESTs a poly(A) tail is lacking and no precise information on the orientation of the clone is given, the majority of EST hits will be false positive. The clones we could not identify further by rescreening GenBank for polyadenylation signals or poly(A) tails are marked with an asterisk in Table 3.
Another pitfall is shown in Table 3. Tag no. 8, which occurred 15 times in fetal brain and not at all in medulloblastoma, matched in GenBank with cytoskeletal
-actin, as did tag no. 20, which on the contrary showed a much higher expression in medulloblastoma (26 times vs. 8 times in fetal brain). Further examination of these hits shows that tag no. 8 matched with the DNA for cytoskeletal
-actin and represented indeed the last CATG in front of the poly(A) tail. Tag no. 20 also matched with cytoskeletal
-actin, but in this case with the mRNA sequence. However, of the most-3' CATG in the DNA sequence the CATG was changed to CGTG in the mRNA for
-actin and thus was not recognized as a Nla III recognition site. The preceding CATG in the mRNA sequence was seen as the last one, and thus a different tag was found. Apparently this represents a polymorphism, for which the individual from whom the fetal brain library is constructed was heterozygous, and the medulloblastoma patient was homozygous. By counting the frequencies of both the tags, the total tag counts are 23 and 26 (for fetal brain and medulloblastoma, respectively), and the difference is no longer significant. This is also confirmed by Northern blot analysis (Fig. 3).
A similar phenomenon might play a role in the results that were obtained for GAPDH. As shown in Table 3, the tag corresponding to the mRNA for GAPDH (tag no. 37) was seen 7 and 21 times in fetal brain and medulloblastoma, respectively. However, Northern blot analysis showed a slightly higher expression in fetal brain (Fig. 3). Several GAPDH pseudogenes exist (4, 24). A database search identified a GAPDH pseudogene, in which the last CATG was polymorphic. This emphasizes that one should be aware of polymorphisms in the last CATG and in the following 910 bases. This will result in multiple different tags for the same gene and also possible wrong assignment of a tag and underscores the necessity of Northern blot confirmation of SAGE results. Except for GAPDH, Northern blot analysis of differentially expressed genes showed a good correlation with the SAGE data (see below).
The tag count for ZIC1 protein is significantly higher in medulloblastoma (Table 3), and this was confirmed by Northern blot (Figs. 2 and 3). ZIC1 is known to be expressed very selectively in cells of the EGL and IGL and from cells migrating from one layer to another, and in medulloblastoma (41). As the granular layers form only a very small part of the total fetal brain, their expression was "diluted" in the expression pattern of the total fetal brain. In contrast, the expression was very high in medulloblastoma. In the zinc finger region ZIC1 is highly homologous (>70%) to the Drosophila pair-rule gene Opa, and ZIC1 is the putative mammalian homolog. In the Drosophila embryo, Opa is required for the activation of wingless and engrailed (3). High ZIC1 expression in medulloblastoma was not accompanied by high expression of the mammalian homologs of wingless (Wnt-1, -2, -3, -4, and -5) and engrailed (En-1 and -2). Only for En-2 was one tag found. For the other genes no tags were detected. Thus high mRNA levels of ZIC1 do not imply that Wnt and En homologs are highly expressed. This could be due to the fact that either the hierarchy seen in Drosophila is not conserved in humans or that ZIC1 is not the human homolog of Opa. We cannot of course exclude the possibility that the expression of Wnt and En homologs is controlled by ZIC1 but that the expression levels are not high enough to be picked up by the analysis of 10,000 tags. The Wnt and En homologs did not appear in fetal brain.
More detailed analysis of tag no. 18 revealed identity to OTX2. OTX2 is related to the Drosophila homeobox gene orthodenticle. It is expressed in the developing head of the fruit fly and involved in the development of rostral brain regions. Its expression pattern is well studied in different developmental stages in mouse and rat (12, 32, 33), but regional expression in fetal or adult human brain has not been examined, as far as we know. In mouse, Otx2 -/- embryos show defective development of the rostral neuroectoderm, resulting in a headless phenotype (Ref. 31 and references therein). Just like Zic1, Otx2 is expressed in the EGL, the IGL, and cells migrating to the IGL. These layers are among candidate sites of origin of medulloblastoma (20, 34, 43, 44). The high expression of both ZIC1 and OTX2 in medulloblastoma strongly supports this hypothesis. In rats, Otx2 is also expressed in the granule neurons of the EGL as well as their precursor cells (12, 31). Normally the EGL disappears at ~1 yr of age in humans. Nests of precursor cells that fail to disappear might be the cause of medulloblastoma in later life. The activation of genes, such as OTX2 and ZIC1, that are important in the development of these layers must be strictly controlled. Inappropriate activation might cause malignant transformation of these cells. Further analysis is necessary to clarify the relationship between these genes and the development of medulloblastoma.
The finding that OTX2 is expressed in the majority of medulloblastomas tested may provide us with a tool helpful in molecular pathological diagnostics. Thus far, we have not seen OTX2 expression in adult tissues, including brain.
 |
ACKNOWLEDGMENTS
|
---|
We thank Dr. D. Troost for providing the fetal brain and reviewing the medulloblastoma slides, and the neurosurgeons of the Academic Medical Center for providing the tumor tissue. We thank our colleagues at the Neurozintuigen Laboratory, Drs. R. Versteeg, H. Tabak, J. M. B. V. de Jong, and E. Hettema, for critical comments.
This work was supported by the Stichting Kindergeneeskundig Kankeronderzoek and the European Cancer Center.
Address for reprint requests and other correspondence: F. Baas, Neurozintuigen Laboratory, Academic Medical Center, P.O. Box 22700, 1100 DE Amsterdam, The Netherlands (E-mail: f.baas{at}amc.uva.nl).
 |
FOOTNOTES
|
---|
Article published online before print. See web site for date of publication (http://physiolgenomics.physiology.org).
 |
REFERENCES
|
---|
-
Adams, M. D., J. M. Kelley, J. D. Gocayne, M. Dubnick, M. H. Polymeropoulos, H. Xiao, C. R. Merril, A. Wu, B. Olde, R. F. Moreno, A. R. Kerlavage, W. R. McCombie, and J. C. Venter. Complementary DNA sequencing: expressed sequence tags and human genome project. Science 252: 16511656, 1991.[Medline]
-
Ausubel F., R. Brent, R. Kingston, D. Moore, J. Seidman, J. Smith, and K. Struhl. Current Protocols in Molecular Biology. New York: Wiley, 1995, p. 498499.
-
Benedyk, M. J., J. R. Mullen, and S. DiNardo. Odd-paired: a zinc finger pair-rule protein required for the timely activation of engrailed and wingless in Drosophila embryos. Genes Dev. 8: 105117, 1994.[Abstract]
-
Benham, F. J., and S. Povey. Members of the human glyceraldehyde-3-phosphate dehydrogenase-related gene family map to dispersed chromosomal locations. Genomics 5: 209214, 1989.[Medline]
-
Biegel, J. A., L. B. Rorke, R. J. Packer, L. N. Sutton, L. Schut, K. Bonner, and B. S. Emanuel. Isochromosome 17q in primitive neuroectodermal tumors of the central nervous system. Genes Chromosomes Cancer 1: 139147, 1989.[Medline]
-
Bigner, S. H., J. Mark, H. S. Friedman, J. A. Biegel, and D. D. Bigner. Structural chromosomal abnormalities in human medulloblastoma. Cancer Genet. Cytogenet. 30: 91101, 1988.[Medline]
-
Church, G. M., and W. Gilbert. Genomic sequencing. Proc. Natl. Acad. Sci. USA 81: 19911995, 1984.[Abstract]
-
Cogen, P. H., L. Daneshvar, A. K. Metzger, and M. S. Edwards. Deletion mapping of the medulloblastoma locus on chromosome 17p. Genomics 8: 279285, 1990.[Medline]
-
Duffner, P. K., M. E. Cohen, M. H. Myers, and H. W. Heise. Survival of children with brain tumors: SEER Program, 19731980. Neurology 36: 597601, 1986.
-
Enoch, T., K. Zinn, and T. Maniatis. Activation of the human ß-interferon gene requires an interferon-inducible factor. Mol. Cell. Biol. 6: 801810, 1986.[Medline]
-
Evans, A. E., R. D. T. Jenkin, R. Sposto, J. A. Ortega, C. B. Wilson, W. Wara, I. J. Ertel, S. Kramer, C. H. Chang, S. L. Leikin, and G. D. Hammond. The treatment of medulloblastoma: results of a prospective randomized trial of radiation therapy with and without CCNU, vincristine, and prednisone. J. Neurosurg. 72: 572582, 1990.[Medline]
-
Frantz, G. D., J. M. Weimann, M. E. Levin, and S. K. McConnell. Otx1 and Otx2 define layers and regions in developing cerebral cortex and cerebellum. J. Neurosci. 14: 57255740, 1994.[Abstract]
-
Frohman, M. A., M. K. Dush, and G. R. Martin. Rapid production of full-length cDNAs from rare transcripts: amplification using a single gene-specific oligonucleotide primer. Proc. Natl. Acad. Sci. USA 85: 89989002, 1988.[Abstract]
-
Hedrick, S. M., D. I. Cohen, E. A. Nielsen, and M. M. Davis. Isolation of cDNA clones encoding T cell-specific membrane-associated proteins. Nature 308: 149153, 1984.[Medline]
-
Hibi, K., Q. Liu Q, G. A. Beaudry, S. L. Madden, W. H. Westra, S. L. Whage, S. C. Yang, R. F. Hetmiller, A. H. Bertelsen, D. Sidransky, and J. Jen. Serial analysis of gene expression in non-small cell lung cancer. Cancer Res. 58: 56905694, 1998.[Abstract]
-
Karnes, P. S., T. N. Tran, M. Y. Cui, C. Raffel, F. H. Gilles, J. A. Barranger, and K. L. Ying. Cytogenetic analysis of 39 pediatric central nervous system tumors. Cancer Genet. Cytogenet. 59: 1219, 1992.[Medline]
-
Kleihues, P., P. C. Burger, and B. W. Scheithauer. The new WHO classification of brain tumours. Brain Pathol. 3: 255268, 1993.[Medline]
-
Liang, P., and A. B. Pardee. Differential display of eukaryotic messenger RNA by means of the polymerase chain reaction. Science 257: 967971, 1992.[Medline]
-
Madden, S. L., E. A. Galella, J. Zhu, A. H. Bertelsen, and G. A. Beaudry. SAGE transcript profiles for p53-dependent growth regulation. Oncogene 15: 10791085, 1997.[Medline]
-
Nagashima, K., K. Yasui, J. Kimura, M. Washizu, K. Yamaguchi, and W. Mori. Induction of brain tumors by a newly isolated JC virus (Tokyo-1 strain). Am. J. Pathol. 116: 455463, 1984.[Abstract]
-
Neumann, E., D. K. Kalousek, M. G. Norman, P. Steinbok, D. D. Cochrane, and K. Goddard. Cytogenetic analysis of 109 pediatric central nervous system tumors. Cancer Genet. Cytogenet. 71: 4049, 1993.[Medline]
-
Packer, R. J., L. N. Sutton, J. W. Goldwein, G. Perilongo, G. Bunin, J. Ryan, B. H. Cohen, G. D'Angio, E. D. Kramer, R. A. Zimmerman, L. B. Rorke, A. E. Evans, and L. Schut. Improved survival with the use of adjuvant chemotherapy in the treatment of medulloblastoma. J. Neurosurg. 74: 433440, 1991.[Medline]
-
Phelan, C. M., L. Liu, M. H. Ruttledge, K. Muntzning, P. A. Ridderheim, and V. P. Collins. Chromosome 17 abnormalities and lack of TP53 mutations in paediatric central nervous system tumours. Hum. Genet. 96: 684690, 1995.[Medline]
-
Piechaczyk, M., J. M. Blanchard, S. Riaad-El Sabouty, C. Dani, L. Marty, and P. Jeanteur. Unusual abundance of vertebrate 3-phosphate dehydrogenase pseudogenes. Nature 312: 469471, 1984.[Medline]
-
Raffel, C., F. E. Gilles, and K. I. Weinberg. Reduction to homozygosity and gene amplification in central nervous system primitive neuroectodermal tumors of childhood. Cancer Res. 50: 587591, 1990.[Abstract]
-
Reardon, D. A., E. Michalkiewicz, J. M. Boyett, J. E. Sublett, R. E. Entrekin, S. T. Ragsdale, M. B. Valentine, F. G. Behm, H. Li, R. L. Heideman, L. E. Kun, D. N. Shapiro, and A. T. Look. Extensive genomic abnormalities in childhood medulloblastoma by comparative genomic hybridization. Cancer Res. 57: 40424047, 1997.[Abstract]
-
Roberts, R. O., C. F. Lynch, M. P. Jones, and M. N. Hart. Medulloblastoma: a population-based study of 532 cases. J. Neuropathol. Exp. Neurol. 50: 134144, 1991.[Medline]
-
Rouah, E., D. R. Wilson, D. L. Armstrong, and G. J. Darlington. N-myc amplification and neuronal differentiation in human primitive neuroectodermal tumors of the central nervous system. Cancer Res. 49: 17971801, 1989.[Abstract]
-
Schutz, B. R., W. Scheurlen, J. Krauss, S. du Manoir, S. Joos, M. Bentz, and P. Lichter. Mapping of chromosomal gains and losses in primitive neuroectodermal tumors by comparative genomic hybridization. Genes Chromosomes Cancer 16: 196203, 1996.[Medline]
-
Shena, M., D. Shalon, R. W. Davis, and P. O. Brown. Quantitative monitoring of gene expression patterns with a complementary DNA microarray. Science 270: 467470, 1995.[Abstract]
-
Simeone, A. Otx1 and Otx2 in the development and evolution of the mammalian brain. EMBO J. 17: 67906798, 1998.[Abstract/Free Full Text]
-
Simeone, A., D. Acampora, M. Gulisano, A. Stornaiuolo, and E. Boncinelli. Nested expression domains of four homeobox genes in developing rostral brain. Nature 358: 687690, 1992.[Medline]
-
Simeone, A., D. Acampora, A. Mallamaci, A. Stornaiuolo, M. R. D'Apice, V. Nigro, and E. Boncinelli. A vertebrate gene related to orthodenticle contains a homeodomain of the bicoid class and demarcates anterior neuroectoderm in the gastrulating mouse embryo. EMBO J. 12: 27352747, 1993.[Abstract]
-
Stevenson, L., and R. Echlin. The nature and origin of some tumors of cerebellum (medulloblastoma). Arch. Neurol. Psychiatry 31: 93109, 1934.
-
Sutcliffe, J. G. mRNA in the mammalian central nervous system. Annu. Rev. Neurosci. 11: 157198, 1988.[Medline]
-
Tait, D. M., H. Thornton-Jones, H. J. Bloom, J. Lemerle, and P. Morris-Jones. Adjuvant chemotherapy for medulloblastoma: the first multi-centre control trial of the International Society of Paediatric Oncology (SIOP I). Eur. J. Cancer 26: 464469, 1990.[Medline]
-
Thomas, G. A., and C. Raffel. Loss of heterozygosity on 6q, 16q, and 17p in human central nervous system primitive neuroectodermal tumors. Cancer Res. 51: 639643, 1991.[Abstract]
-
Velculescu, V. E., L. Zhang, B. Vogelstein, and K. W. Kinzler. Serial analysis of gene expression. Science 270: 484487, 1995.[Abstract]
-
Velculescu, V. E., L. Zhang, W. Zhou, J. Vogelstein, M. A. Basrai, D. E. Bassett, Jr., P. Hieter, B. Vogelstein, and K. W. Kinzler. Characterization of the yeast transcriptome. Cell 88: 243251, 1997.[Medline]
-
Waard, V. de, B. M. van den Berg, J. Veken, R. Schutz-Heienbrok, H. Pannekoek, and A. J. Zonneveld. Serial analysis of gene expression to assess the endothelial cell response to an atherogenic stimulus. Gene 226: 18, 1999[Medline]
-
Yokota, N., J. Aruga, S. Takai, K. Yamada, M. Hamazaki, T. Iwase, H. Sugimura, and K. Mikoshiba. Predominant expression of human zic in cerebellar granule cell lineage and medulloblastoma. Cancer Res. 56: 377383, 1996.[Abstract]
-
Zhang, L., W. Zhou, V. E. Velculescu, S. E. Kern, R. H. Hruban, S. R. Hamilton, B. Vogelstein, and K. W. Kinzler. Gene expression profiles in normal and cancer cells. Science 276: 12681272, 1997.[Abstract/Free Full Text]
-
Zimmerman, H. M. Brain tumors: their incidence and classification in man and their experimental production. Ann. NY Acad. Sci. 159: 337359, 1969.
-
Zu Rhein, G. M., and J. N. Varakis. Perinatal induction of medulloblastomas in Syrian golden hamsters by a human polyoma virus (JC). Natl. Cancer Inst. Monogr. 205208, 1979.