Download Genes differentially expressed in medulloblastoma and fetal brain

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
1Department of Pediatric Oncology, Emma Kinderziekenhuis/Academic Medical Center;
and 2Neurozintuigen Laboratory and 3Department of Clinical Epidemiology,
Academic Medical Center, 1100 DE Amsterdam, The Netherlands
serial analysis of gene expression; brain tumors
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
Received 17 March 1999; accepted in final form 29 June 1999.
Article published online before print. See web site for date of
publication (http://physiolgenomics.physiology.org).
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 9–10 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 38 terminus [near the poly(A)
tail] is isolated. These 38 fragments of cDNA are ligated
to linkers and cleaved with a type II restriction enzyme
to release a short sequence (9–10 bp) of the original
cDNA (tags). The tags are ligated to ditags and PCRamplified. 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 50,000–100,000 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 241⁄2-wk fetal
brain and a medulloblastoma and compared their patterns of gene expression. Medulloblastoma is a central
nervous system tumor predominantly of childhood and
1094-8341/99 $5.00 Copyright r 1999 the American Physiological Society
83
Downloaded from http://physiolgenomics.physiology.org/ by 10.220.33.4 on August 9, 2017
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: 83–91, 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 ␹-square test statistics
were calculated. Northern blot was used to confirm some of
the results obtained by SAGE. For 16 tags, the ␹-square 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.
84
SAGE OF FETAL BRAIN AND MEDULLOBLASTOMA
MATERIALS AND METHODS
Fetal brain and medulloblastoma tissue. Fetal brain was
obtained from a 241⁄2-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. 18267–013),
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 300–600 bp
were used for cloning in the Sph I site of pZero (Invitrogen)
and transformation into TOP 10F8 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 38 rapid amplification of cDNA
ends (RACE)-PCR according to the procedure described by
Frohman et al. (13), with 58-GCATGCCAGAATTCTGGATCC-38
as a reverse primer. Template cDNA was made of mRNA of the
SAGE medulloblastoma, with 58-GCATGCCAGAATTCTGGATCCTTTTTTTTTTTTTTTTTT-38 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 2719–2739 of the mRNA (forward) and
position 3019–3000 (reverse); for secretogranin I (GenBank
accession no. Y00064), position 1449–1468 (forward) and
2094–2075 (reverse); for GAPDH (GenBank accession no.
M33197), position 245–264 (forward) and 536–517 (reverse).
For OTX2 two different probes were constructed: one of ⬃300
bp using 58-GCAAAATTCAGAGCAACTGAG-38 as forward
and 58-ATCTGCCAAATCCAGGAAGAA-38 as reverse primer
and one of ⬃600 bp using 58-TGGGAACAGGATCCAGATTTC-38
as forward and 58-CTCGACTCGGGCAAGTTGA-38 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).
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 ␹-square 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 ␹-square statistics were calculated, one for
each test statistic. Within the same pair of distributions,
more pronounced differences in expression will lead to higher
␹-square 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.
http://physiolgenomics.physiology.org
Downloaded from http://physiolgenomics.physiology.org/ by 10.220.33.4 on August 9, 2017
arising in the cerebellum. Five-year survival of affected
children is poor (39–70%) 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.
85
SAGE OF FETAL BRAIN AND MEDULLOBLASTOMA
Table 1. List of tags occurring at least 10 times in
medulloblastoma with corresponding GenBank entry
Accession
No.
Frequency
GenBank Entry
CCC ATC GTC
75
CGT
CAC
GGG
GGC
GAG
ATT
ATC
CAG
69
52
44
41
TCC GTG GTT
30
GCT
CTA
CTT
CCT
TAC
AGG
ATT
TCA
ACA
TCG
CAA
TCC
27
25
25
21
21
20
CGA CCC CAC
GAA GCA GGA
GGA TTT GGC
20
20
20
GAG
TGG
CAG
GAA
GGG
GTG
GGA
TGT
TTG
AAA
CTG
CTG
GTT
TGA
TGG
TGG
GGG
AAT
19
18
17
17
17
17
TTA
AAG
GCC
GGA
TGG
CAG
GTG
GAG
CCA
CCC
CTC
GAG
GAA
CTG
CAC
17
16
16
16
16
TTG
GCA
GTG
TTG
AGC
GAA
GCC
TGC
CCC
GCC
GGC
TAA
TTG
GTC
TAA
AAG
GTG
ACC
CAC
GTG
ACG
TGG
TTC
AAG
GGA
GAG
CTC
TAG
GCA
AAG
TCC
ATC
TCC
TTT
GTT
CAA
AAG
GCT
ATC
16
15
15
15
14
14
14
14
13
13
13
13
13
AAG
AAG
CAA
CAC
TCC
TTC
ACA
GAG
CTA
AAA
TGC
AAT
GTG
ATG
ATT
CGC
CCC
AAA
12
12
12
12
12
12
TTG GGG TTT
ACT TTG TCC
12
11
ATT
CCA
CCT
CGC
TAA
CCA
CAG
GCC
AAA
AAC
AGG
TCC
11
11
11
11
11
10
CTG GGG TAA
10
Cytochrome c oxidase,
subunit II
Elongation factor 1-␣
Mitochondrial ATPase
Thymosin ␤-10
Calmodulin binding protein (Mac Marcks)
Neuronal tissue-enriched
acidic protein NAP 22
␤-Actin
Cytoskeletal ␥-actin
Secretogranin I
T-cell cyclophilin
GAPDH
Ubiquinol-cytochrome c
reductase complex subunit VI requiring protein
Apolipoprotein E
Nonmuscle isoform cofilin
Acidic ribosomal phosphoprotein P2
Ribosomal protein L27a
Ribosomal protein S18
Neuronatin
Laminin-binding protein
Ribosomal protein L29
Nonmuscle/smooth
muscle myosin alkali
light chain
ZIC1 protein
Ribosomal protein L18a
Ribosomal protein S12
Ribosomal protein L3
Homo sapiens Opa-interacting protein OIP3
Ribosomal protein L41
Ribosomal protein L21
Ribosomal protein S3a
Thymosin ␤-4
Elongation factor 2
Ribosomal protein L19
Ribosomal protein S6
Ribosomal protein L32
Ferritin L chain
Nuclear p68 protein
Ribosomal protein L27
Ribosomal protein S26
NADH:ubiquinone oxidoreductase MLRQ subunit
Ribosomal protein L37a
Ribosomal protein L31
TRPM-2
Ribosomal protein S27
Parathymosin
Acidic ribosomal phosphoprotein P1
Ferritin H chain
Glial fibrillary acidic protein GFAP
Ribosomal protein L17
Ribosomal protein S9
Ribosomal protein L38
Ribosomal protein L4
Ribosomal protein S8
Neuron-specific ␥-2 enolase
Ribosomal protein S19
GTT
CTA
GAA
AGC
TTT
GCC
ATG
AGC
CAT
GCT
CTC
GTG
CGG
CGG
TAA
CGT
Tag
Frequency
GenBank Entry
GCC GGG TGG
10
GTG GCA GGC
10
TAG GTT GTC
10
TGA GGG AAT
10
Collagenase stimulatory
factor
Pancreatic zymogen
granule membrane protein GP-2
Translationally controled
tumor protein
Triosephosphate isomerase
X15759
X16869
AA782865
M92381
X70326
AF039656
X00351
X04098
Y00064
Y00052
M33197
M73791
K00396
X95404
M17887
U14968
X69150
U25034
X61156
U10248
M22918
D76435
L05093
X53505
X73460
AF025439
AF026844
U14967
M77234
M17733
M19997
X63527
M20020
X03342
M11147
X15729
L19527
X69654
U94586
X66699
X69181
M64722
U57847
M24398
M17886
M97164
J04569
X55954
U14971
Z26876
X73974
X67247
M22349
M81757
Accession
No.
L10240
U36221
X16064
M10036
Only tags that match with mRNA sequences of known genes
containing a poly(A) signal are listed. Frequency, no. of times the tag
is found in the medulloblastoma library (total no. of tags, 10,229).
TRPM-2, testosterone-repressed prostatic message-2.
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 38-untranslated
region (38-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 lowabundance 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.
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 ␹-square 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 ␹-square 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
http://physiolgenomics.physiology.org
Downloaded from http://physiolgenomics.physiology.org/ by 10.220.33.4 on August 9, 2017
Tag
Table 1.—Continued
86
SAGE OF FETAL BRAIN AND MEDULLOBLASTOMA
Table 2. List of tags occurring at least 10 times in fetal
brain with corresponding GenBank entry
Tag
Mitochondrial ATPase
Elongation factor 1-␣
Laminin-binding protein
Cytochrome c oxidase, subunit II
NADH dehydrogenase 1
Ribosomal protein S18
␤-Tubulin class III isotype
Acidic ribosomal phosphoprotein P2
Ribosomal protein L29
Thymosin ␤-10
Ubiquinol-cytochrome c
reductase complex subunit
VI requiring protein
Nonmuscle isoform cofilin
Elongation factor 2
Ribosomal protein L21
Ribosomal protein S12
Ribosomal protein L41
Neuronal tissue-enriched
acidic protein NAP 22
Neuronatin
Ferritin H chain
Calmodulin binding protein (MacMarcks)
Ribosomal protein S4
(RPS4X) isoform
Ribosomal protein S27
Ribosomal protein L4
Ribosomal protein S8
Ribosomal protein L31
Ribosomal protein L27a
Cytoskeletal ␥-actin
Ribosomal protein L18a
Cytochrome oxidase subunit 1
Creatine kinase-B
Ribosomal protein L6
Ribosomal protein S29
Ribosomal protein L18
Ribosomal protein S3a
Ribosomal protein L30
Ribosomal protein L11
Homolog of yeast ribosomal
protein S28
Ribosomal protein S16
Ribosomal protein L3
Pancreatic zymogen granule
membrane protein GP-2
Cystatin C
Elongation factor 1-␥
Thymosin ␤-4
Ribosomal protein L37a
Ribosomal protein L9
Ribosomal protein L17
Ribosomal protein L37
Ribosomal protein L19
Elongation factor 1 ␣-2
CTA
GTT
AAA
ATC
ATT
GAG
TGG
GTC
67
63
37
35
ACC
TGG
AAC
GGA
CTT
TGT
GAC
TTT
GGC
TGA
CTG
GGC
33
27
25
25
GGG CTG GGG
GGG GAA ATC
AGG GCT TCC
24
24
23
GAA
AGC
GCA
GCC
TTG
TCC
GGA
TCC
TAG
GAA
CTC
GTT
23
21
21
21
22
20
CAG TTG TGG
TTG GGG TTT
GGC AGC CAG
19
19
18
TCA GAT CTT
19
CAC
CGC
TAA
AAG
GAG
TCA
AAG
ATT
AAA
CGG
TAA
GAG
GGA
GGG
GTG
TGA
CGG
AAC
AGG
ATG
GTT
CTG
GAG
GAA
17
17
17
17
17
15
14
14
CAC
TAC
ATA
GGA
GTG
CCA
CGC
CTG
CCC
AAG
ATT
GTG
AAG
GAA
TGG
TTG
TGA
AGG
CTT
GAC
GCA
CAG
TTC
GTG
14
14
13
13
13
12
12
12
CCG TCC AAG
GGA CCA CTG
GAG AAA CCC
11
11
11
TGC
TGG
TTG
AAG
ATC
ATT
CAA
GAA
TCT
11
11
11
11
10
10
10
10
10
CTG
GCA
GTG
ACA
AAG
CTC
TAA
CAC
GCA
CAC
AAG
AAG
GTG
GGT
CAG
ATG
ATC
CCT
AA782865
X16869
X61156
X15759
X93334
X69150
U47634
M17887
U10248
M92381
M73791
X95404
M19997
U14967
X53505
AF026844
AF039656
U25034
M97164
X70326
M22146
U57847
X73974
X67247
X69181
U14968
S57813
L05093
X93334
M16364
X69391
L31610
L11566
M77234
L05095
L05092
D14530
M60854
X73460
U36221
X05607
Z11531
M17733
X66699
U09953
X55954
D23661
X63527
X70940
Only tags that match with mRNA sequences of known genes
containing a poly(A) signal are listed. Frequency, number of times the
tag is found in the fetal brain library (total no. of tags, 10,692).
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
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.
http://physiolgenomics.physiology.org
Downloaded from http://physiolgenomics.physiology.org/ by 10.220.33.4 on August 9, 2017
GenBank Entry
CAC
TGT
GAA
CCC
GCA
ACC
TAA
GAG
GTC
GTG
Accession
No.
Frequency
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 38 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 sequencetagged 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
87
SAGE OF FETAL BRAIN AND MEDULLOBLASTOMA
Table 3. Results of comparison of gene expression in fetal brain and medulloblastoma
Tag
MB
GenBank Entry
ESTs
ESTs*
Secretogranin I
ESTs
ZIC1 protein
ESTs*
Cytochrome c oxidase subunit II
Cytoskeletal ␥-actin
ESTs*
NADH dehydrogenase 1
Alu repeat sequence
Alu repeat sequence
No match
SOX4/?
Apolipoprotein E
␤-Actin
ESTs
ESTs = OTX2
ESTs moderately similar to homology with squid
retinal binding protein
Cytoskeletal ␥-actin
Calmodulin binding protein (Mac Marcks)
PACAP type-3/VIP type-2 receptor?
Elongation factor 1 ␣-2
Vascular endothelial growth factor-D
ESTs*
GFAP
ESTs*
ESTs*
Ribosomal protein L27
NADH dehydrogenase 3
Alu repeat sequence
␤-Tubulin class III isotype (␤-3)
No match
Collagenase stimulatory factor
ESTs*
T-cell cyclophilin
GAPDH
Alu repeat sequence
Alu repeat sequence
Thymosin ␤-10
ESTs*
ESTs*
ESTs*
Alu repeat sequence
Ribosomal protein L 38
Alu repeat sequence
Laminin-binding protein
H3 histone, family 3A
ESTs*
Cytochrome oxidase subunit 1
ESTs, highly similar to neuritin (R. norvegicus)
ESTs, weakly similar to PCBP-2 protein (H. sapiens)
ESTs
Glutamine synthase
ESTs*
Activin receptor/80 K-L protein
ESTs*
ESTs*
Mitochondrial DNA, RNA 5
ESTs*
Brain-specific tyrosine and tryptophane hydrolase
activator
ESTs*
No match
Phosphatidylinositol synthase
ESTs*
mac-2 binding protein
DAP-1
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
CCG
ACT
CTT
TCT
TTA
TTA
CCC
TCA
CCC
ACC
GCG
CCT
AGC
CAG
CGA
GCT
ATG
ACC
GAC
CTG
TTT
ATG
AAC
CAG
ACC
ATC
GGG
GTC
CTT
AAA
GTA
CCG
GCT
CCC
TTT
TCA
AAC
TCT
CGT
TCA
ACA
AAC
CTC
CTC
GTC
CTG
CGG
GGC
CCC
ATC
TGA
TTT
CAC
ATT
CGA
TGG
GGG
2
28
0
0
0
0
35
15
28
33
26
71
0
0
4
8
1
0
0
35
1
22
21
17
16
75
0
6
9
6
34
11
11
20
27
13
10
10
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
CTA
GGC
ACT
TCT
GAA
GTG
ACT
CCG
CCT
GGC
AGC
CCA
AAC
GAG
GCC
AGG
CCT
TAC
GCA
GTG
GGG
TGG
CTT
TGT
GTG
CCT
TTG
GAA
CCA
CAA
CAT
CTG
GGG
GTT
TAC
TCA
TTT
CCA
TAG
AGA
GCA
TCA
GCC TCA
AGC CAG
AAC ACC
GCA CCT
TGG CAG
GGG ACG
TTG TCC
GCC GCC
GTA ATT
AAG AAG
CCT ACA
CTG CAC
GAC CTG
AAA TTA
GGG TGG
TGG CAA
AGC TGG
CAT CAA
AAA CCC
AAA CCC
GAA ATC
GAA GTG
TTC AGC
TTG TTG
GG GGC
CGG AAA
GCC AGG
AAA TGG
AAC GTG
ATC CAA
TTG TAA
CTA TGG
CAG GGG
TCT TCC
AGT ATG
CCT TAG
TGT AAA
GGA GGA
GGC AAT
CCC ACA
GTC ATT
ATC AAG
8
18
13
10
0
0
1
1
9
2
12
59
25
1
1
8
7
7
17
99
24
12
7
7
16
2
23
37
6
0
0
0
0
0
0
0
0
13
1
6
6
6
26
41
1
0
9
9
11
11
0
13
30
30
8
10
10
0
21
21
4
62
44
2
0
0
4
11
8
17
18
6
6
6
6
6
6
6
6
3
8
0
0
0
62
63
64
65
66
67
TGG
TCC
AAG
AGG
ATG
CAT
GGA
CGT
GCA
CGA
CTC
CTG
6
5
0
0
0
0
0
15
5
5
5
5
TGT
ACA
CAG
GAT
CCT
TGA
Accession
No.
AI199006
Y00064
X93334
D76435
X15759
M19283
X93334
X70683
K00396
X00351
AA121202
H17804
AA600960
X04098
X70326
U18810
X70940
AJ000185
J04569
L19527
X93334
U47634
L10240
Y00052
M33197
M92381
Z26876
X61156
M11354
X93334
AI143163
AI364320
AI193246
X59834
D10522
X93334
S80794
AF014807
L13210
X76105
␹Square
30.98
24
23.02
21.97
17.78
16.74
16.46
14.36
13.31
12.7
11.65
11.52
11.5
11.5
11.41
11.2
10.84
10.46
10.46
10.36
10.05
9.773
9.572
9.411
9.411
8.791
8.791
8.614
8.571
8.553
8.249
8.038
7.775
7.775
7.657
7.647
7.647
7.494
7.002
6.826
6.715
6.699
6.699
6.688
6.643
6.622
6.569
6.554
6.273
6.273
6.273
6.273
6.273
6.273
6.273
6.273
5.823
5.764
5.742
5.742
5.742
5.742
5.46
5.228
5.228
5.228
5.228
Continued
http://physiolgenomics.physiology.org
Downloaded from http://physiolgenomics.physiology.org/ by 10.220.33.4 on August 9, 2017
FB
88
SAGE OF FETAL BRAIN AND MEDULLOBLASTOMA
Table 3.—Continued
Tag
MB
GenBank Entry
ESTs*
Histone H2A.Z
ESTs
␤-2-Microglobulin
ESTs*
Preferentially expressed antigen of melanoma
ESTs
ESTs*
mln 51 gene
Alu repeat sequence
Ribosomal protein S26
NADH:ubiguinone oxidoreductase MLRQ subunit
ESTs*
T cell leukemia/lymphoma 1 gene
Clone 23867 (infant brain)
Alu repeat sequence
Clone 23707 (infant brain)
ESTs*/repeat sequence
ESTs*/repeat sequence
ESTs*/repeat sequence
ESTs*
ESTs*/repeat sequence
ESTs*
Alu repeat sequence
ERK activator kinase (MEK2)
H3 histone, family 3B (H3.3B)
ESTs*/repeat sequence
Homo sapiens okadaic acid-inducible phosphoprotein (OA48-18) mRNA
ESTs*/repeat sequence
nonmuscle/smooth muscle myosin alkali light chain
ESTs*
Ribosomal protein L18
TRPM-2 gene
ESTs*
ESTs*
ESTs*/repeat sequence
Ribosomal protein S11
ESTs*
Cystatin C (cysteine proteinase inhibitor precursor)
ESTs
ESTs*
UDP-galactose transporter related isozyme 1
Cysteine- and glycine-rich protein 2
Phosphomevalonate kinase
Ubiquitine carboxyterminal hydrolase
ESTs
No match
Protein phosphatase 1 catalytic subunit
mRNA mapping to 22q13
ESTs
Cyclin protein
No match
ESTs*/repeat sequence
hLIM-1/?
No match
Human surface antigen
Alu repeat sequence
Apolipoprotein C-1
GST1-Hs mRNA for GTP-binding protein
ESTs*
Homo sapiens mRNA for GEF-2 protein
ESTs*
Phospholipase A2
ESTs*
Homo sapiens zinc finger protein 216 spice variant 2
(ZNF216) mRNA
Homo sapiens Opa-interacting protein OIP3
Nuclear p68 protein
ESTs*/repeat sequence
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
CCA
GAC
GTG
GTT
TAG
TAG
TAG
TCT
TGT
TTC
TAA
TTG
TCT
GAA
GAG
ACC
AAT
AGC
CCT
GTG
GTG
TCA
TCA
TTA
CAG
AGA
GAA
TGG
CAT
GTG
TCG
GTG
CTG
GAG
TAA
CCA
GGT
AGC
GGA
GAG
TCC
AAA
GAT
CCG
GTT
CAC
GTG
AAA
AAA
CTG
TAG
GCC
GAA
GCC
AAA
CTT
TGC
TGG
GGG
GTT
CTG
TTA
TAG
GGA
GGT
CTG
GCT
ATC
AGG
AAA
GGT
GGA
AAA
TGC
ATC
CCA
TTC
CAC
TTC
AGG
CGG
AGC
TTT
GCT
0
0
0
0
0
0
0
0
0
0
4
4
8
10
3
17
5
5
5
5
5
5
5
5
2
1
1
1
5
5
5
5
5
5
5
5
5
5
13
13
1
2
11
6
0
0
0
0
0
0
0
0
9
7
7
7
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
ATT
GTG
ACA
GGA
CAA
ACC
GGT
GTG
TCT
GAG
TGC
AAG
AGC
CAG
CCC
CTC
CTT
CTT
GAA
GAG
GCC
GGA
GGC
GGT
TAC
TCG
TCG
TCT
TCT
TGG
TTA
TTC
TTG
TTT
ACT
GGG
TAC
GTT
CTG
AAC
GTG
CTA
AAT
GAG
GCG
GTA
AAA
CTG
AAG
CGG
TGG
AGT
TTC
CTG
GGG
GTT
CCT
CAG
AAG
GTG
TGG
CCC
GCC
TAA
GGT
GTT
CCC
ATA
AGT
GAC
GCC
TTT
GTC
AGA
TAT
AAT
TTA
GAC
ATT
GAA
ACC
CAG
CAC
CCC
CAC
ACT
GCT
GTG
AAG
AGG
TGT
AGG
CTG
TGG
CCC
TTC
AAC
TAG
TGA
GAG
CGA
CTG
AAT
CAG
AAA
GCT
TGA
CTA
CTG
TGG
GGG
7
7
15
13
4
7
7
7
7
11
11
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
9
9
3
17
17
5
4
12
1
1
1
1
3
3
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
2
2
10
133
134
135
TGG CCC CAC
GCC TTC CAA
AAA AAT AAA
7
5
14
16
13
5
http://physiolgenomics.physiology.org
Accession
No.
M37583
AI362197
AA148104
U65011
AI217125
X80199
X69654
U94586
X82240
U79287
U79270
L11285
AA703312
AF069250
M22918
L11566
M64722
X06617
M27891
AI363755
D87989
U57646
L77213
D80012
AI024305
S57501
AL021682
M15796
U14755
M60922
X00570
X17644
AI391666
AF077046
M86400
AF062347
AF025439
X15729
␹Square
5.228
5.228
5.228
5.228
5.228
5.228
5.228
5.228
5.228
5.228
5.178
5.178
5.144
4.99
4.938
4.793
4.785
4.785
4.785
4.785
4.785
4.785
4.785
4.785
4.775
4.774
4.774
4.774
4.629
4.629
4.574
4.38
4.367
4.242
4.242
4.242
4.424
4.229
4.229
4.182
4.182
4.182
4.182
4.182
4.182
4.182
4.182
4.182
4.182
4.182
4.182
4.182
4.182
4.182
4.182
4.182
4.182
4.182
4.182
4.182
4.182
4.182
4.154
4.154
4.09
3.938
3.924
3.88
Downloaded from http://physiolgenomics.physiology.org/ by 10.220.33.4 on August 9, 2017
FB
89
SAGE OF FETAL BRAIN AND MEDULLOBLASTOMA
Table 3.—Continued
Tag
136
137
138
CAC CCC TGA
TAC AAG AGG
CCG CCT CCG
FB
MB
GenBank Entry
Accession
No.
␹Square
14
14
2
5
5
8
Creatine kinase B
Ribosomal protein L6
Small nuclear ribonucleoprotein particle N
M16364
X69391
U41303
3.88
3.88
3.874
The 138 tags for which the ␹-square test statistic was associated with a P value ⬍0.05 are listed with corresponding GenBank entry. FB, no. of times
the tag is found in fetal brain library; MB, no. of times the tag is found in medulloblastoma library; GenBank entry, name of corresponding hit in
GenBank; *ESTs, no hits with a known gene in GenBank but matches with multiple expressed sequence tag (EST) sequences (position of last CATG in
EST sequence in relation to poly(A) tail is unknown; therefore tag assignment is ambiguous); ␹-square, ␹-square value for pairwise ␹-square statistics.
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
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.
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 38-UTR is also known. The EST library has to be
screened to identify ESTs that carry the entire 38-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-38 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
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;
3–8, 6 other medulloblastomas.
http://physiolgenomics.physiology.org
Downloaded from http://physiolgenomics.physiology.org/ by 10.220.33.4 on August 9, 2017
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).
90
SAGE OF FETAL BRAIN AND MEDULLOBLASTOMA
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.
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: [email protected]).
REFERENCES
1. 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: 1651–1656, 1991.
http://physiolgenomics.physiology.org
Downloaded from http://physiolgenomics.physiology.org/ by 10.220.33.4 on August 9, 2017
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-38 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 9–10 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
SAGE OF FETAL BRAIN AND MEDULLOBLASTOMA
24. 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:
469–471, 1984.
25. 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:
587–591, 1990.
26. 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: 4042–4047, 1997.
27. 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: 134–144, 1991.
28. 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: 1797–1801, 1989.
29. 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: 196–203, 1996.
30. Shena, M., D. Shalon, R. W. Davis, and P. O. Brown.
Quantitative monitoring of gene expression patterns with a
complementary DNA microarray. Science 270: 467–470, 1995.
31. Simeone, A. Otx1 and Otx2 in the development and evolution of
the mammalian brain. EMBO J. 17: 6790–6798, 1998.
32. Simeone, A., D. Acampora, M. Gulisano, A. Stornaiuolo, and E.
Boncinelli. Nested expression domains of four homeobox genes in
developing rostral brain. Nature 358: 687–690, 1992.
33. 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: 2735–2747, 1993.
34. Stevenson, L., and R. Echlin. The nature and origin of some
tumors of cerebellum (medulloblastoma). Arch. Neurol. Psychiatry 31: 93–109, 1934.
35. Sutcliffe, J. G. mRNA in the mammalian central nervous
system. Annu. Rev. Neurosci. 11: 157–198, 1988.
36. 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:
464–469, 1990.
37. 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: 639–643, 1991.
38. Velculescu, V. E., L. Zhang, B. Vogelstein, and K. W. Kinzler.
Serial analysis of gene expression. Science 270: 484–487, 1995.
39. 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:
243–251, 1997.
40. Waard, V. de, B. M. van den Berg, J. Veken, R. SchutzHeienbrok, H. Pannekoek, and A. J. Zonneveld. Serial
analysis of gene expression to assess the endothelial cell response to an atherogenic stimulus. Gene 226: 1–8, 1999
41. 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: 377–383, 1996.
42. 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:
1268–1272, 1997.
43. Zimmerman, H. M. Brain tumors: their incidence and classification in man and their experimental production. Ann. NY Acad.
Sci. 159: 337–359, 1969.
44. 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. 205–208, 1979.
http://physiolgenomics.physiology.org
Downloaded from http://physiolgenomics.physiology.org/ by 10.220.33.4 on August 9, 2017
2. 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. 498–499.
3. 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: 105–117, 1994.
4. Benham, F. J., and S. Povey. Members of the human glyceraldehyde-3-phosphate dehydrogenase-related gene family map to
dispersed chromosomal locations. Genomics 5: 209–214, 1989.
5. 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: 139–147, 1989.
6. 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: 91–101, 1988.
7. Church, G. M., and W. Gilbert. Genomic sequencing. Proc.
Natl. Acad. Sci. USA 81: 1991–1995, 1984.
8. Cogen, P. H., L. Daneshvar, A. K. Metzger, and M. S.
Edwards. Deletion mapping of the medulloblastoma locus on
chromosome 17p. Genomics 8: 279–285, 1990.
9. Duffner, P. K., M. E. Cohen, M. H. Myers, and H. W. Heise.
Survival of children with brain tumors: SEER Program, 1973–
1980. Neurology 36: 597–601, 1986.
10. Enoch, T., K. Zinn, and T. Maniatis. Activation of the human
␤-interferon gene requires an interferon-inducible factor. Mol.
Cell. Biol. 6: 801–810, 1986.
11. 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: 572–582, 1990.
12. 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: 5725–5740, 1994.
13. 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: 8998–9002, 1988.
14. Hedrick, S. M., D. I. Cohen, E. A. Nielsen, and M. M. Davis.
Isolation of cDNA clones encoding T cell-specific membraneassociated proteins. Nature 308: 149–153, 1984.
15. 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: 5690–5694, 1998.
16. 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: 12–19, 1992.
17. Kleihues, P., P. C. Burger, and B. W. Scheithauer. The new WHO
classification of brain tumours. Brain Pathol. 3: 255–268, 1993.
18. Liang, P., and A. B. Pardee. Differential display of eukaryotic
messenger RNA by means of the polymerase chain reaction.
Science 257: 967–971, 1992.
19. 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: 1079–1085, 1997.
20. 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: 455–463, 1984.
21. 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: 40–49, 1993.
22. 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: 433–440, 1991.
23. 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: 684–690, 1995.
91