| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Regular Article |
From the Cancer Research Unit,*
The Medical School,
University of Newcastle, Newcastle-upon-Tyne; the Departments of Child
Health
and
Pathology,
Royal Victoria Infirmary,
Newcastle-upon-Tyne; and the Department of
Statistics,
University of Newcastle,
Newcastle-upon-Tyne, United Kingdom
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
An important mechanism of intrinsic chemoresistance in many tumor cells is an abnormality in the p53 tumor suppressor gene pathway.3 p53 is mutated in up to 60% of human cancers, leading, in most cases, to accumulation of a non-functional protein.3 p53, a nuclear phosphoprotein, is usually present at low levels in the cell due to a short half-life of 30 minutes, but accumulates in response to cellular stress such as DNA damage from irradiation or alkylating agents. It binds DNA in a sequence-specific manner to activate the transcription of a number of genes, including p21WAF1, MDM2 and BAX. WAF1 inhibits G1 cyclin-dependent kinases, blocking cell cycle progression from G1 into S phase. MDM2 binds to p53 and blocks its ability to function as a transcription factor so creating an autoregulatory feedback loop to tightly regulate p53 levels. Tumors with mutant p53 cannot usually bind to DNA and up-regulate MDM2 and consequently, there is a lack of MDM2 to bind to p53 and target it for ubiquitin-mediated degradation, resulting in p53 accumulation.4
p53 can also respond to cellular stress by inducing apoptosis, which may be transcriptionally dependent or independent depending on the cell type. In some cell types, p53 transcriptionally induces BAX, a pro-apoptotic gene that forms mitochondrial pores leading to cytosolic release of cytochrome c, which activates caspases and leads to apoptosis.4 BAX may also form heterodimers with BCL2, an anti-apoptotic membranous protein, which may be transcriptionally repressed by p53. Whether a cell undergoes growth arrest, or apoptosis, is dependent on the cell type, phase of the cell cycle, differentiation status, the presence of other oncogenic abnormalities and external growth and survival factors, and the level of DNA damage and p53 induced. Since p53 induction can lead to apoptosis following DNA damage, acquisition of p53 mutations might promote tumorigenesis and lead to chemoresistance. Several studies screening for p53 mutations by single strand conformational polymorphism analysis followed by direct sequencing, have shown that p53 mutations in neuroblastoma tumors5-8 and cell lines6,9 are rare. However, p53 is readily detectable in neuroblastoma tissue10,11 and cell lines9,12,13 where it has a prolonged half-life of 6 to 10 hours, and there is increased transcription.9,12 Stabilization of p53 can occur through mechanisms other than mutation such as binding to viral oncoproteins, but in neuroblastoma cell lines p53 has not been found to bind to the SV40 viral large T antigen or heat shock protein-70 using co-immunoprecipitation.9
Accumulation of p53 can also occur in conjunction with altered subcellular localization, and this has been proposed as a mechanism for the accumulation and functional inactivation of p53 in neuroblastoma. An immunocytochemical (ICC) study of frozen neuroblastomas reported cytoplasmic sequestration of p53 in undifferentiated neuroblastomas.14 In contrast, two other ICC studies on formalin-fixed, paraffin-embedded tissue reported nuclear localization of p53.10,11 In neuroblastoma cell lines, some studies report nuclear,15,16 and others cytoplasmic, localization of p53.17,18 There is further controversy over whether cytoplasmic localization renders p53 non-functional, with some studies suggesting that it does17,18 and others suggesting that it does not.19,20
This study investigated the hypothesis that p53 accumulation in neuroblastoma, in the absence of mutation, is associated with functional inactivation, which interferes with downstream mediators of p53 function. The current study showed that p53 is both nuclear and functional in wild-type p53 neuroblastoma cell lines and a poorly differentiated ex-vivo neuroblastoma in short term culture. The response of wild-type p53 neuroblastoma cell lines to irradiation was heterogeneous in terms of the transcriptional activity of p53, cell cycle arrest and level of apoptosis. In particular it was observed that MYCN-amplified neuroblastoma cell lines with wild-type p53 underwent predominantly G2 arrest after irradiation despite induction of WAF1. The role of MYCN in affecting the downstream response to p53 in neuroblastoma should be further explored.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Six human Mycoplasma-free neuroblastoma cell lines were studied: IMR32,21 SKNSH,22 SHSY5Y (a neuroblastic (N) type clone of SKNSH), SHEP (a substrate adherent (S) type clone of SKNSH),22 NGP23 and SKNBE(2c).22 Three human control cell lines were also studied; MCF7 breast cancer cell line as a positive control for functional wild-type p5324 and two negative controls, the p53 mutant neuroepithelioma/Ewing’s tumor cell line NB/CHP100,9,25 and the p53 null SAOS2 osteosarcoma cell line.26 All cell lines were grown as monolayers in RPMI 1640 medium (Life Technologies, Paisley, United Kingdom) supplemented with 10% fetal calf serum in a 37°C, 5% CO2, humidified incubator. Karyotypic analysis of all neuroblastoma cell lines and the NB100 cell line was performed to confirm previously published karyotypes and ensure the cell lines were genuine.
Three primary untreated tumors were obtained within 1 hour of surgical resection for ex vivo studies in short-term culture. Two were from children aged 2 years with stage 1 and 4 non-MYCN amplified neuroblastomas (tumors 1 and 3, respectively), and one was from a two-year-old child with a MYCN-amplified, 1p deleted stage 4 neuroblastoma (tumor 2).
p53 Sequencing of Cell Lines and Tumors
Total cellular RNA was extracted from frozen cell pellets
(
5 x 107
cells) using RNAzol B
(Biogenesis, Poole, UK). RNA was used as a template to synthesize full
length p53 cDNA using the RT-1 primer CGG GAG GTA GAC (5 µg/ml),
Superscript II reverse transcriptase (Life Technologies), first strand
buffer, 10 mmol/L DTT, dNTPs (4 mmol/L) and 2 µg RNA heated to 65°C
for 10 minutes. The 20 µl reaction mix was incubated for 1 hour at
46°C.
The p53 cDNA was amplified by the polymerase chain reaction (RT-PCR) in a thermal cycler (Perkin-Elmer, Norwalk, CT) using forward and reverse primers with phosphorothioate linkages: 5'-ATT TGA TGC TGT CCC CGG ACG ATA TTG AA-s-C-3' (5 µg/ml) and 5'-ACC CTT TTT GGA CTT CAG GTG GCT GGA GT-s-G-3' (5 µg/ml), Pfu DNA polymerase (Stratagene, La Jolla, CA), 10x reaction buffer (Stratagene) and dNTPs (0.2 mmol/L) (Pharmacia, Herts, United Kingdom), in a 20 µl reaction mix. The PCR reaction involved an initial denaturation step at 95°C for 5 minutes followed by 35 cycles of subsequent denaturation at 94°C for 30 seconds, annealing at 65°C for 60 seconds, and elongation at 78°C for 80 seconds. The products were analyzed by electrophoresis on a 1.8% (w/v) low melting point agarose gel (NuSieve, Maine) in Tris acetate buffer. Products were excised and purified using a QIAquick gel extraction kit (Qiagen, West Sussex, United Kingdom). The purified p53 cDNA was then sequenced using the above primers and two additional internal ones 5'-TGG CCA TCT ACA AGC AGT CA-3' and 5'-GGG CAC CAC CAC ACT ATG TC-3', by dideoxy chain termination thermocycle sequencing (Perkin-Elmer). The sequences obtained span exons 4–10 of the human p53 gene and were compared with the published p53 sequence using DNAStar computer software (Madison, WI). Any cell line with a p53 mutation detected in the RT-PCR product had the mutation confirmed by genomic DNA sequencing of the relevant exon using methods described previously.27 Genomic sequencing of DNA from the 3 primary untreated tumors was also performed by these methods for exons 4–9 of the p53 gene.27
ICC
Cell pellets (3 x 107 cells) were prepared as cytoblocks using a cytospin (Shandon, Pittsburgh, PA). Three micrometer sections were cut from formalin-fixed, paraffin-embedded cell cytoblocks and tumors and placed on adhesive coated 3–(aminopropyltriethoxysilane) (Sigma, St Louis, MO) microscope slides. Endogenous peroxidase activity was blocked by incubation in 0.5% hydrogen peroxide/methanol for 10 minutes and antigen retrieval performed by incubating slides in 0.01 mol/L citrate buffer pH 6.0 in a 650W microwave oven for two 5 minute incubations. ICC was also performed on cell lines from cells grown on Labtek chamber glass slides (Nunc, Naperville, IL) and cytospins of cell pellets, after fixation with 1:1 acetone:methanol for 5 minutes, followed by air drying. Nonspecific binding of antibody was prevented by blocking for 10 minutes in normal rabbit serum diluted 1:10 with Tris-buffered saline (TBS). ICC was performed using the streptavidin-biotin peroxidase technique. Three mouse monoclonal primary antibodies were used which recognize epitopes at the amino terminal of wild-type and mutant p53. NCL-Pab1801 (amino acids 46–55) at 1:40, NCL-DO-7 (both NovoCastra, Newcastle-upon-Tyne, UK), (amino acids 20–25) at 1:50 and DO-1 (hybridoma supernatant) (amino acids 20–25) at 1:4. Other primary antibodies used for ICC were mouse monoclonal antibodies for Ki67 (MIB1), WAF1, and BCL2 (all NovoCastra) at dilutions of 1:200, 1:20, and 1:50 respectively. Rabbit anti-mouse biotinylated secondary antibody (Dako, Glostrup, Denmark) was used at 1:500, streptavidin-biotinylated horseradish peroxidase complex (Dako) at 1:100, the chromagen 3,3-diaminobenzidine tetrachloride (Sigma) in 0.3% hydrogen peroxide solution in TBS and hematoxylin as a nuclear counterstain. The labeling index (LI), percent of immunostained cells, was determined by counting 1000 cells in the areas of densest immunostaining using an Olympus CX microscope at x40 magnification.
Irradiation
Exponentially growing cells from all cell lines were plated in 28
cm2
tissue culture dishes (Nunc) at 5 x
105/dish. When 70% confluent cells were
treated with 4 Gy 
-irradiation from a
137cesium -irradiator (Gamma Cell 1000 Elite;
Nordion International Inc., Ontario, Canada) at 3.64 Gy/minute and
returned to the incubator to be harvested at specific times afterward.
The fresh tumors were placed in RPMI 1640 medium with 10% fetal calf
serum immediately after surgical resection. After finely chopping they
were placed in dishes as above and half the sample irradiated as above
and then incubated for 6 hours before harvesting with non-irradiated
controls.
Western Blotting
Cells were harvested at 0, 1, 2, 4, 6, and 24 hours after irradiation, lysed in boiled Laemmli lysis buffer [62.5 mmol/L Tris-HCl (pH 6.8), 2% SDS, 10% glycerol],28 boiled for 10 minutes, sonicated (Soniprep 150, MSE), and the protein content estimated (Pierce, Rockford, IL). Before loading the gel, 5% 2-mercaptoethanol and 0.005% bromophenol blue were added, the samples were boiled again for 5 minutes, centrifuged briefly, and 50 µg of protein loaded on to a 4 to 20% Tris-HCl precast sodium dodecyl sulfate-polyacrylamide gel (Biorad, Hercules, CA) for electrophoresis. Molecular weight markers (Novex, San Diego, CA) or ECL markers (Amersham, UK) were used on every gel. The proteins were transferred electrophoretically overnight to Hybond C nitrocellulose membrane (Amersham). Primary antibodies used were mouse monoclonal NCL-p53DO-7 at 1:1000, MDM2 (Ab-1) (Calbiochem, Cambridge, MA) at 1:100, WAF1 (Ab1) (Calbiochem) at 1:100, NCL-BCL2 at 1:100, MYCN (NCMII100)29 at 1:10, polyclonal rabbit BAX antibody (Pharmingen, San Diego, CA) both at 1:1000, and mouse monoclonal anti-actin ascitic fluid (Sigma) at 1:1000 as a loading and transfer control. Blots were reprobed for other proteins without prior stripping. Secondary antibodies used were peroxidase-conjugated, affinity isolated, goat anti-mouse (Dako), goat anti-rabbit immunoglobulin (Dako) or biotinylated rabbit anti-mouse (Dako). Secondary antibodies were used at 1:1000, followed by peroxidase conjugated streptavidin (Dako) at 1:5000 after the biotinylated secondary. Protein detection was performed using enhanced chemiluminescence (Amersham). Densitometry was performed on immunoblots using a Bio Image capture system (Millipore, MI) and CCT camera linked to a Sun View computer (Sun View Microsystems, CA). Where >1 band was present only the band of the correct protein size was used in the analysis.
Cell Cycle Analysis
Adherent cells (2 x 106) were harvested 6 and 24 hours after irradiation by trypsinization, fixed with ice-cold 70% ethanol in PBS, and stored at 4°C, followed by resuspension in PBS with RNase (0.1 mg/ml) and propidium iodide (40 µg/ml) (Sigma), and incubation at 37°C for 30 minutes.30 DNA was analyzed by fluorescence activated cell sorting (Becton-Dickinson, Oxford, UK) and LYSII or Cell Quest computer software used to calculate the proportions of cells in different phases of the cell cycle. Results shown are mean of three experiments ± SD.
Apoptosis Measurements
Cells (2 x 106) were harvested 1, 2, 4, 6, 24, 48, 72, 96, and 108 hours after irradiation by trypsinization, fixed in Carnoy’s (3:1 v/v methanol:acetic acid), centrifuged and resuspended in Carnoy’s and an equal volume of Hoechst 33258 (8 µg/ml). They were examined under a microscope using a UV filter (Zeiss) and the proportion of apoptotic nuclei scored out of 500 counted.
Clonogenic Assays
Cells (1 x 106) were trypsinized immediately after irradiation and a single cell suspension was prepared. After counting with a hemocytometer, between 200 and 40,000 were seeded into 10-cm diameter petri-dishes with 10 ml of medium. They were placed in an incubator and the colony formation (>50 cells) examined after 2 to 3 weeks. The number of colonies per dish was counted after fixing with Carnoy’s and staining with 0.4% crystal violet.
Statistical Analyses
These were performed using Minitab version 11.2 1996 (Minitab Inc., PA) and Stata for Windows 95 version 5.0 1996 (Stata Corporation, Texas) statistical software.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
p53 cDNA sequencing confirmed that the IMR32, SKNSH, SHEP, SHSY5Y, and NGP neuroblastoma cell lines, together with the breast cancer MCF7 cell line, were wild-type for p53. The p53 status of MDM2-amplified NGP cells31 has not previously been reported but MDM2 gene amplification is associated with wild-type p53 in other cell types.32 In NB100 neuroepithelioma cells, a 4-bp insertion between codon 232 and 233 in exon 7 was confirmed resulting in a translational stop codon at position 240.9 The presence of a missense mutation in the SKNBE(2c) neuroblastoma cell line in exon 5 was confirmed at codon 135: G to T transversion; TGC (cysteine) to TTC (phenylalanine).33
All cell lines with wild-type p53 expressed a 53-kd protein on Western
blotting, but not p53 null SAOS2 and p53 mutant NB100 cells (Figure 1)
. The highest level of p53 protein
expression was in p53 mutant SKNBE(2c) cells, which expressed a mean of
2.6 times more p53 than IMR32 cells, which expressed higher levels of
p53 than any of the other neuroblastoma cell lines (Figure 1)
. The
intense p53 band in SKNBE(2c) was still present when the blots were
subsequently reprobed for actin (40kd) (Figure 1)
. All cell lines with
wild-type p53 showed predominantly nuclear p53 by ICC of
formalin-fixed, paraffin-embedded cell pellets (Figure 2a)
, cystospins (data not shown), and
cells grown on glass slides (Figure 2b)
. SKNSH cells are composed of two
types, a neuroblastic (N type) population, small cells with delicate
processes (neuropil) and a larger
epithelioid or substrate adherent cell (S type). Some cytoplasmic p53
was present in SKNSH cells using 3 p53 antibodies (DO-7, DO-1 and
1801), particularly in N type cells (Figure 2b
C and D), but
cytoplasmic staining of controls without primary antibody, was also
observed (Figure 2b, A and B)
. Some mild cytoplasmic staining was also
seen in N type cells with other nuclear antigens WAF1 (Figure 2b, E and F)
and Ki67 (Figure 2b, I and J)
, in contrast to strong BCL2
cytoplasmic immunostaining (Figure 2b, G and H)
.
|
|
. These cell lines also had a high proportion of cells
in G1 (data not shown). NGP cells
(MDM2-amplified) expressed the highest level of MDM2
protein, followed by MCF7 cells, which have been reported to
over-express MDM2 (Figure 1)
.32
BCL2 was expressed at
highest levels in SKNSH cells, but was absent in SHEP cells as
previously reported (Figures 1 and 2b
G).34
MYCN was
readily detectable in the three MYCN-amplified cell lines
(IMR32, NGP, and SKNBE(2c)) (Figure 1)
.35 p53 Transcriptional Function after Irradiation of Cell Lines
All neuroblastoma cell lines studied with wild-type p53 (ie,
SKNSH, SHSY5Y, SHEP, NGP, IMR32 and the MCF7 positive control cell
line) showed nuclear accumulation of p53 after 4 Gy irradiation (Table 1
and Figure 2
). In SKNSH, IMR32, and
MCF7 cells, DO-1 and Pab1801 were also used for ICC and both of these
antibodies also showed predominantly nuclear p53 before and after
irradiation (data not shown).
|
and Figure 3a
). p53
induction varied among the cell lines from 2- to 3-fold in the SHEP and
NGP cell lines to 20- to 25-fold in the SKNSH and IMR32 cell lines. The
p53 mutant NB100 and null SAOS2 cell lines showed no induction of p53
post irradiation (Table 1)
and the p53 mutant SKNBE(2c) cell line
showed no increase above high basal p53 expression (Table 1
and Figure 3a
).
|
, Figures 2 and 3a
).
WAF1 and MDM2 were induced from 4 hours, and MDM2 peaked at 4 or 6
hours after irradiation. WAF1 remained elevated 24 hours after
irradiation in all cell lines with wild-type p53, whereas MDM2 had
returned to pretreatment levels by this time (Figure 3a)
. The relative
increase in WAF1 varied from twofold in SHEP cells to 19-fold in
SHSY5Y, and for MDM2, from twofold in the MDM2-amplified NGP
cell line, to 15-fold in SKNSH cells (Table 1)
. BAX expression
increased more than twofold in two cell lines with wild-type p53-SKNSH
and IMR32, and there was a reduction in BCL2 expression only in IMR32
and p53 null SAOS2 cells (data not shown). There was no change in MYCN
expression after irradiation in any of the cell lines studied with
MYCN amplification (data not shown). Cell Cycle Arrest and Apoptosis after Irradiation of Cell Lines
Cell cycle analysis showed that three wild-type p53 neuroblastoma
cell lines (SKNSH, SHSY5Y, and SHEP) and MCF7 cells underwent strong
G1 arrest, as shown by an increase in G1/S
ratio at 24 hours compared with control, and the DNA histograms (Figure 3b
, i and iii). Two cell lines with wild-type p53, IMR32 and NGP, did
not show G1 arrest despite evidence of WAF1 induction,
however they underwent a striking G2 arrest, as shown by
the DNA histograms and an increased proportion of cells in G2 at 24
hours compared with control (Figure 3b
ii and iii). Mutant p53 NB100 and
null p53 SAOS2 cells did undergo some G2 arrest but it was
not as strong as IMR32 and NGP (Figure 3b, ii)
.
Apoptosis as measured by Hoechst nuclear staining was detectable from 6
hours after irradiation and measured up to 108 hours later (Table 1)
.
Using analysis of variance, the rate of induction of apoptosis ie, the
slope of the apoptosis versus time curve, the maximum value and
the time from irradiation at which the maximum value occurred, were
compared in cell lines wild-type or mutant for p53. After allowing for
differences between individual cell lines, there was no significant
difference between cell lines that were wild-type and those that were
mutant for p53 (rate of induction, P = 0.66; maximum
value, P = 0.40; and time of maximal induction,
P = 0.30). Clonogenic survival varied between cell
lines in response to 4 Gy irradiation with IMR32 cells being the most
sensitive and SHEP cells the most resistant (Table 1)
. Comparing the
clonogenic survival between individual cell lines using a Poisson
log-linear rates model, there was no significant difference between
cell lines that were mutant and those that were wild-type for p53
(P = 0.08).
Histology, p53 Sequencing, ICC, and Irradiation of Primary Neuroblastoma Tumors
Tumor 1 was a Schwannian stroma-poor (<50% tumor tissue),
differentiating neuroblastoma with >5% neuronal differentiation
toward ganglion cells and calcification (favorable histology;
International Neuroblastoma Pathology Classification)36
(Table 2)
. Tumor 2 was a Schwannian
stroma-poor, poorly-differentiated neuroblastoma, with a high
(327/5000) mitosis-karyorrhexis index and no calcification (unfavorable
histology) (Figure 4
and Table 2
). Tumor
3 was a nodular ganglioneuroblastoma (unfavorable histology). A
non-nodular area of stroma-rich, differentiating neuroblastoma was
studied, but there was insufficient material to study a nodule as well
(Figure 4
and Table 2
).
|
|
and Table 2
), together with
high Ki67 (a proliferation marker) expression (Table 2)
. In contrast,
tumors 1 and 3 expressed low levels of nuclear p53 and Ki67. In tumor 1
some differentiating neuroblastoma cells did express cytoplasmic p53
with all three antibodies, but cytoplasmic staining was also present in
the control without primary antibody (data not shown) and p53 was
undetectable on Western blotting (Figure 5)
, suggesting this was nonspecific
binding to neuropil.
|
. p53 expression increased by a
mean of ninefold (range, 3–13, n = 4), 6 hours after
irradiation with concomitant detection of WAF1 and a 90kd MDM2 protein,
but no change in MYCN expression. The 2 differentiating neuroblastomas
(tumors 1 and 3) did not express p53, MYCN, or a 90-kd MDM2 protein by
Western blotting before or after irradiation, but there was baseline
expression of WAF1, which was unchanged following irradiation (Figure 5)
. BCL2 and BAX expression were unchanged after irradiation in all of
the tumors studied (data not shown). |
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Five neuroblastoma cell lines studied were found to be p53 wild-type (SKNSH, SHSY5Y, SHEP, IMR32, and NGP), after sequencing exons 4–10 of p53 cDNA. p53 mutations outside exons 4–9 are rare,3,4 which is consistent with the structure and function of p53 and where mutations leading to functional inactivation would be expected.3
In wild-type p53 neuroblastoma cell lines immunofluorescence studies have reported the cellular localization of p53 in untreated cells to be wholly cytoplasmic,17 cytoplasmic in N type cells and both nuclear and cytoplasmic in S type cells,18 or wholly nuclear.15,16 Most studies reporting cytoplasmic p53 used the 1801 antibody, which is believed to cross-react.15 In the current study, three different methods of cell preparation were used for ICC, which gave similar results showing predominantly nuclear p53 in wild-type p53 neuroblastoma cell lines with three different antibodies (DO-7, DO-1, and 1801). In neuroblastoma cells grown on glass slides, some cytoplasmic p53 was detectable with all three antibodies. However, the presence of punctate cytoplasmic staining in controls without primary antibodies, particularly in N type SKNSH cells, suggests that some of this may be nonspecific binding to neuropil produced by neuroblastoma. Cell fractionation studies have also been used to localize p53 in neuroblastoma and studies report both nuclear and cytoplasmic p53,17-20 particularly in N type cells, whereas S type cells have been reported to have mainly nuclear p53.18,37 It is likely that p53 in neuroblastoma is both nuclear and cytoplasmic, but predominantly nuclear, and the ease of detecting one or the other may be determined by the nuclear to cytoplasmic ratio of the cell and the amount of neuropil produced.
Although p53 was detectable in all wild-type p53 cell lines, WAF1 was only detectable in SHEP and MCF7 cells, suggesting that, under normal growth conditions, apparent over-expression of wild-type p53 in most neuroblastoma cell lines is not up-regulating WAF1, and is consistent with active cellular proliferation. This situation is analogous to F9 murine teratocarcinoma cells, which have been reported to express high basal levels of p53 but low levels of WAF1 and MDM2, and it has been shown that the p53 is inactive unless DNA damage occurs.38 Neuroblastoma cells treated with 4 Gy ionizing radiation show further accumulation of p53 protein and transcriptional activation of target genes, particularly WAF1 and MDM2 consistent with previous studies.16,19 These effects all occur in a similar time course and manner to MCF7 cells, which were used as positive control, as they have previously been shown to have functional wild-type p53.24
p53 induction in SHEP cells was less than other wild-type p53 neuroblastoma cell lines. They failed to undergo apoptosis, consistent with previous reports,18,37 despite lacking BCL2, and were the most resistant cell line in clonogenic assays. The low level of apoptosis and radiation resistance of SHEP cells probably accounts for the failure to detect a significant difference between apoptosis and clonogenicity in mutant p53 cell lines compared with the wild-type ones studied in this report. In the MDM2-amplified cell line NGP, the induction of p53 and p53 responsive genes was also less than for the other wild-type p53 cell llines suggesting attenuation of p53 transcriptional function by MDM2 amplification, an effect observed in another MDM2-amplified neuroblastoma cell line NB-1691.16
Four out of six cell lines with wild-type p53 underwent G1 arrest (SHSY5Y, SHEP, SKNSH, and MCF7) 24 hours after irradiation, whereas the other two wild-type p53 cell lines (IMR32 and NGP) failed to show evidence of G1 arrest despite induction of WAF1 at this time. Lack of G1 arrest after irradiation despite induction of WAF1 has previously been reported in a Burkitt’s lymphoma cell line,39 in which it has been shown that this results from instability of WAF1. In the current study WAF1 expression remained maximally elevated 24 hours after irradiation in both IMR32 and NGP cells. Both these cell lines did, however, undergo strong G2 arrest, unlike the mutant or null p53 cell lines, which did not arrest strongly in G2 or G1. IMR32 and NGP were the only MYCN-amplified wild-type p53 cell lines, and this may have affected cell cycle distribution. It is possible that MYCN overexpression in the MYCN-amplified cell lines is able to override the effect of WAF1 and impair G1 arrest. Overexpression of c-myc has been recently shown to cause p53- and WAF1-dependent G2 arrest of normal fibroblasts.40 MYCN-transfected SHEP cells have increased DNA synthesis, a shortened G1 phase, and accelerated G0 to S progression after mitogenic stimulation,41 and recent evidence suggests that activation of MYCN in neuroblastoma causes an increase in the retinoblastoma protein inhibitor, Id2, leading to cell cycle progression.42
The p53 response pathway was also studied in three wild-type p53 tumors in short term culture. p53 was iduced after ex vivo irradiation of the poorly-differentiated tumor, together with upregulation of MDM2 and WAF1, but not in the two differentiating tumors in which there was baseline expression of WAF1, consistent with cell growth arrest, which did not change following irradiation. These preliminary findings need to be confirmed by a much larger study of ex-vivo neuroblastomas. Nevertheless, low level p53 expression in differentiating neuroblastoma is consistent with cell line studies showing a reduction in p53 expression following retinoic acid induced differentiation,9,13 and pretreatment of neuroblastoma cell lines with retinoic acid confers resistance to p53 dependent apoptosis.43,44 If p53 function is impaired in differentiating neuroblastoma, it may have implications for the design of clinical trials of neuroblastoma where differentiation agents, such as retinoids, are used alongside conventional chemotherapy. Furthermore, tumor 3 was a nodular ganglioneuroblastoma, a tumor composed of hemorrhagic neuroblastic nodules coexisting with a stroma rich ganglioneuroblastoma component.36 Nodular ganglioneuroblastoma carries a poor prognosis, which is thought to be due to the presence of biologically aggressive clones. If the differentiating component of these tumors can resist p53-dependent cytotoxic therapy, and later de-differentiate, this would provide a route for subsequent relapse. In the absence of continued differentiation stimuli in vitro, neuroblastoma cells de-differentiate and begin to proliferate.45 Transdifferentiation of neuroblastoma cell lines from the N to the S type and vice versa46 can occur spontaneously resulting in altered sensitivity to cytotoxic agents including irradiation as reported here.
In conclusion, this study examined the p53 pathway in detail after irradiation in six human neuroblastoma cell lines and three ex vivo tumors. p53 transcriptional function was shown to be intact in neuroblastoma cell lines with wild-type p53, and the poorly-differentiated neuroblastoma in short term culture. The lack of G1 arrest despite WAF1 induction in two MYCN-amplified wild-type p53 neuroblastoma cell lines is suggestive of an abnormality downstream of p53 in MYCN-amplified cells which warrants further study.
|
Acknowledgements |
|---|
|
Footnotes |
|---|
Supported by the North of England Children’s Cancer Research Fund and the UK Cancer Research Campaign.
Accepted for publication February 16, 2001.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-irradiation of human lymphoma cells. Cancer Res 1995, 55:2387-2393This article has been cited by other articles:
 |
|
 |
|
  L. Chen, N. Iraci, S. Gherardi, L. D. Gamble, K. M. Wood, G. Perini, J. Lunec, and D. A. Tweddle p53 Is a Direct Transcriptional Target of MYCN in Neuroblastoma Cancer Res., February 15, 2010; 70(4): 1377 - 1388. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. A. Daniel, A. L. Rozanska, H. D. Thomas, E. A. Mulligan, Y. Drew, D. J. Castelbuono, Z. Hostomsky, E. R. Plummer, A. V. Boddy, D. A. Tweddle, et al. Inhibition of Poly(ADP-Ribose) Polymerase-1 Enhances Temozolomide and Topotecan Activity against Childhood Neuroblastoma Clin. Cancer Res., February 15, 2009; 15(4): 1241 - 1249. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H C Moore, K M Wood, M S Jackson, M A Lastowska, D Hall, H Imrie, C P F Redfern, P E Lovat, F Ponthan, K O'Toole, et al. Histological profile of tumours from MYCN transgenic mice J. Clin. Pathol., October 1, 2008; 61(10): 1098 - 1103. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Xue, M. Haber, C. Flemming, G. M. Marshall, R. B. Lock, K. L. MacKenzie, K. V. Gurova, M. D. Norris, and A. V. Gudkov p53 Determines Multidrug Sensitivity of Childhood Neuroblastoma Cancer Res., November 1, 2007; 67(21): 10351 - 10360. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. Sugihara, M. Kanai, S. Saito, T. Nitta, H. Toyoshima, K. Nakayama, K. I. Nakayama, K. Fukasawa, M. Schwab, H. Saya, et al. Suppression of Centrosome Amplification after DNA Damage Depends on p27 Accumulation. Cancer Res., April 15, 2006; 66(8): 4020 - 4029. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Carr, E. Bell, A. D.J. Pearson, U. R. Kees, H. Beris, J. Lunec, and D. A. Tweddle Increased Frequency of Aberrations in the p53/MDM2/p14ARF Pathway in Neuroblastoma Cell Lines Established at Relapse Cancer Res., February 15, 2006; 66(4): 2138 - 2145. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J.-F. Lavoie, L. LeSauteur, J. Kohn, J. Wong, O. Furtoss, C. J. Thiele, F. D. Miller, and D. R. Kaplan TrkA Induces Apoptosis of Neuroblastoma Cells and Does So via a p53-dependent Mechanism J. Biol. Chem., August 12, 2005; 280(32): 29199 - 29207. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Torkin, J.-F. Lavoie, D. R. Kaplan, and H. Yeger Induction of caspase-dependent, p53-mediated apoptosis by apigenin in human neuroblastoma Mol. Cancer Ther., January 1, 2005; 4(1): 1 - 11. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. J. Rorie and B. E. Weissman The Ews/Fli-1 Fusion Gene Changes the Status of p53 in Neuroblastoma Tumor Cell Lines Cancer Res., October 15, 2004; 64(20): 7288 - 7295. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Lindskog, C. Spenger, J. Jarvet, A. Graslund, and P. Kogner Predicting Resistance or Response to Chemotherapy by Proton Magnetic Resonance Spectroscopy in Neuroblastoma J Natl Cancer Inst, October 6, 2004; 96(19): 1457 - 1466. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S.X. Liang and D.R. Richardson The effect of potent iron chelators on the regulation of p53: examination of the expression, localization and DNA-binding activity of p53 and the transactivation of WAF1 Carcinogenesis, October 1, 2003; 24(10): 1601 - 1614. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. P. McKenzie, M. K. Danks, R. W. Kriwacki, and L. C. Harris p21Waf1/Cip1 Dysfunction in Neuroblastoma: A Novel Mechanism of Attenuating G0-G1 Cell Cycle Arrest Cancer Res., July 1, 2003; 63(13): 3840 - 3844. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
