| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Molecular Biology and Genetics |
Istituto di Tecnologie Biomediche, Consiglio Nazionale delle Ricerche, Via Fratelli Cervi, 93, 20090 Segrate, Milan, Italy [A. M., D. Z., A. V., P. V.]; Genetics Branch, Center for Cancer Research, National Cancer Institute/NIH, Bethesda, Maryland 20892-0913 [C. M., T. R.]; Centro Ricerche Parco Scientifico "E. Menni," Ospedale Poliambulanza, 25100 Brescia, Italy [D. Z.]; Istituto Zooprofilattico Sperimentale delle Regioni Lazio e Toscana, Dipartimento di Firenze, 50010 Florence, Italy [E. I.]; Dipartimento di Oncologia Biologia e Genetica, Università di Genova, I.S.T., 16132 Genoa, Italy [O. B.]; and Istituto di Fisiologia Clinica, Consiglio Nazionale delle Ricerche, 56124 Pisa, Italy [L. C.]
 |
ABSTRACT |
|---|
 Top ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
According to this view, by gaining or losing entire chromosomes, aneuploidy could simulate both overexpression and loss of specific loci relevant to cell growth, thus providing the substrate for selection of more aggressively growing cells (11 , 13) . Therefore, a role for selection in pathogenesis of cancer (14) is not restricted to the mutation first hypothesis but is shared by the aneuploidy first one.
Recently, the molecular and genetic investigation of human colon cancer led to recognize two different forms of genetic instability: (a) the elevated frequency of loss or gain of one or a few whole chromosomes or large parts thereof within the cells of an individual clone (CIN3 ); and (b) MIN (15, 16, 17) .
The genetic basis of MIN has been identified (18) , whereas the search for genes responsible for CIN has been essentially negative (3) . This raises the possibility that germ-line abnormalities are not responsible for CIN; on the contrary, it could be the exposure to damaging agents, resulting in checkpoint activation, growth arrest, and damage repair, that cause a few cells to abnormally segregate their chromosomes, yet escape apoptosis. These compounds could interfere with molecular complexes involved in spindle checkpoint, resulting in aneuploid cells, some of which could acquire a growth advantage. In this case, aneuploidy would be the first detectable change on which subsequent genomic anomalies would take place.
Human fibroblasts have been used to investigate mechanisms of tumorigenicity. They have a limited proliferative potential, and a combination of two oncogenes and hTERT activity is needed for immortalization (4 , 5) , although data challenge this view (19) . In the present study, we set up an in vitro model to test the hypothesis that aneuploidy could occur early in transformation in humans. We therefore explored antisense-mediated selective and temporary gene inactivation, i.e., inhibition of gene function without gene mutation. Therefore, this technique would circumvent the use of genotoxic drugs, which act in complex and uncontrolled ways in tumorigenesis. We exposed human fibroblast to antisense oligonucleotides directed against 13 different genes involved in chromosome segregation. We demonstrated that inhibition of several of these genes gives rise to aneuploid cells, most of which are eliminated by apoptosis. After inhibition of BUB1, whose product is involved in the spindle checkpoint, some of these cells are able to escape programmed death and give rise, at low frequency, to AI clones. Analysis of these clones shows chromosomal instability and suggests that even temporary interfering with spindle checkpoint genes could be relevant to neoplastic transformation in humans.
|
MATERIALS AND METHODS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
60%. The population doubling was calculated as log2 (D/D0), where D is the density of cell when harvesting, and D0 is the density of cells when seeding. Senescence occurred at estimated population doublings of 64–69. At this stage, cells failed to reach confluence despite regular refeeding, and they showed the characteristic unlarged, flattened shape of senescent cells. In addition to these parameters, both molecular and biochemical markers were investigated (see below). In addition, a second cell line (subject 2), established from a Caucasian woman, was used for antisense inhibition studies only.
Antisense Oligonucleotide Methodology.
The general experimental approach to antisense has been described (20)
. For inhibition studies, three oligodeoxynucleotides were synthesized (sequences available on request): (a) antisense oligomer designed as a complementary sequence at the 5' end of the coding region; (b) sense oligomer from the identical region; and (c) scrambled oligomer with the same nucleotides used for the antisense oligomer. We performed two different protocols with antisense oligonucleotide treatment: (a) short term and (b) long term. In short-term culture, cells were treated with 40 µg/ml each antisense, sense, and scrambled oligonucleotides (diluted in DMEM) for 24 h and an additional 20 µg/ml for another 24 h. In long-term culture, after the first treatment with 40 µg/ml, cells were treated with 20 µg/ml each antisense, sense, and scrambled oligonucleotides for an additional 8 days. In addition, mock controls, by adding DMEM without oligonucletides, was performed.
Ribozyme Design, Synthesis, and Treatment.
The ribozyme design was performed according to an original approach, which makes use of a computer method for the analysis of target mRNA sequences (21)
. A series of ribozyme sequences were correspondingly deduced. The final analysis accounting for the correct structure activity relationship enabled us to select the hammerhead ribozyme addressed to AUA triplet located at the 2711 position on the coding region of BUB1 mRNA. The ribozyme sequence is: 5'-CAGGGGCUGAUGAGGCCGAAAGGCCGAAAUUUUUAUAGA-3'(Rz-BUB1). In addition to the active form of Rz-BUB1, a mutated catalytically inactive ribozyme sequence was synthesized (inact-Rz-BUB1) carrying the mutation G5 > A5 in the active core. Rz-BUB1 and inact-Rz-BUB1 were synthesized by introducing chemical modifications conferring resistance to RNase A, as described previously (22)
. At the end, the crude product was processed and purified as already described (23)
. Ribozyme was purified maintaining sterile RNase-free conditions in a MonoQ-HR 16/10 (Amersham Pharmacia Biotech) strong ion-exchange high-performance liquid chromatography preparative chromatography. For culture treatment, 1 µM Rz-BUB1 or inact-Rz-BUB1 was added to cell culture for 48 h.
siRNA Synthesis and Cell Treatment.
siRNA corresponding to BUB1 mRNAs was designed as recommended (24)
with two base overhangs (Xeragon). The following gene-specific sequence was used: si-BUB1 5'-CAUCACUUUUUCAGGGGUA dTT-3'. Cells (at 40–60% confluence) were transfected with 200 nM si-BUB1 RNA by using Oligofectamine Reagent (Invitrogen). Cells were analyzed for aneuploidy frequency 48-h post-transfection.
Immunoprecipitation and Western Blots.
Analysis with p21, p53 (Santa Cruz Biotechnology), SMC1 (Bethyl), and BUB1 (a rabbit polyclonal antibody; a kind gift of T. Yen, Fox Chase Cancer Institute, Philadelphia, PA) antibodies was performed according to a published protocol (25)
. Briefly, specific antibody was incubated at 4°C with protein extracts for 1 h. Protein A-agarose (Santa Cruz Biotechnology) was added overnight, followed by four washings with buffer lysis. Samples were boiled in sample buffer and separated by SDS-PAGE. The proteins were transferred to nitrocellulose membrane (Amersham) and incubated with primary antibody (1:250–1:2000 dilution). After removal of the unbound primary antibody, membranes were incubated with secondary antibody-peroxidase conjugate (Sigma), processed for detection by chemiluminescence (Amersham), and imaged on Biomax film (Kodak). Actin and tubulin antibodies (Santa Cruz Biotechnology) were used as internal controls.
ARP1 Expression by RT-PCR Analysis.
Total RNA was extracted from cells exposed to ARP1 antisense oligonucleotides by the SV Total RNA Isolation System (Promega) and used for cDNA synthesis with random hexamers. The following primers were used to amplify ARP1 cDNA (spanning from exon 1 to 6): Arp1-F: TGT CGT GAT CGA CAA CGG AT, ARP1-R: AAG TCG TAG CCC TCC TTA C. The presence of ARP1 transcripts was analyzed by PCR performed combining both a variable number of amplification cycles and different dilutions of primers. Hypoxanthine phosphoribosyltransferase was used as internal standard.
Cytogenetic Analysis.
Exponentially growing fibroblasts were treated with colcemid (0.05 µg/ml; Life Technologies, Inc.), harvested, incubated with 0.075 M KCl, and fixed in methanol:acetic acid 3:1. Chromosome preparations were G-banded according to the trypsin digestion procedure, and aneuploidy cells, i.e., any variation from diploid number, were scored by direct microscopic examination.
Assays for Mutation to Ouabain and 6-Thioguanine Resistance.
Control and antisense-treated cells were seeded at 105 cells into 100-mm dishes and at 200 cells into 60-mm dishes to determine the plating efficiency. Twenty-four hours later, either ouabain (Sigma), at a concentration of 1 µM, or 6-thioguanine (Sigma), at a concentration of 3 x 10-5 M, was added to the 100-mm dishes to determine the number of resistant colonies. Mutant colonies were fixed, stained, and scored 3 weeks after the initial seeding. The mutation frequency was calculated as: (resistant colonies)/(cells seeded x S), where S is the plating efficiency of cells.
MMC Treatment.
AI clones were treated with MMC (30 µM, diluted in PBS; Sigma) or with an equivalent amount of PBS for 24 h.
Soft Agar Assay.
Cells were suspended in 3 ml of 0.3% agar (Difco) supplemented with complete growth medium. This cell suspension was allowed to solidify at room temperature on 4 ml of a 0.5% agar base layer containing growth medium in 60-mm dishes.
FISH.
In situ hybridization with a biotin-labeled probe recognizing all of the human centromeres (Appligene Oncor) was performed according to a published protocol (26)
. DNA was counterstained with 0.1 µg/ml propidium iodide.
SKY.
SKY was performed as described (27)
. After in situ hybridization, images were acquired using an epifluorescence microscope (DMRXA; Leica) connected to an imaging interferometer (SD200; Applied Spectral Imaging) and analyzed using SKYView v1.6 software (Applied Spectral Imaging). Six to 10 metaphases were analyzed for each cell line.
CGH.
For CGH, DNA was prepared using high salt extraction and phenol purification and labeled by nick translation using biotin-11-dUTP (Roche). Biotin-labeled DNA and digoxigenin-labeled normal donor DNA (sex matched) was cohybridized to sex-matched normal human lymphocyte metaphase chromosomes. Images were acquired with a Leica DMRXA epifluorescence microscope (Leica) using fluorochrome-specific filters (Chroma Technologies). Quantitative fluorescence imaging and CGH analysis was performed using Leica CW4000CGH software (Leica Microsystem Imaging Solutions).
Immunohistochemistry.
Cells were cultured on chamber slides (Falcon), washed in PBS, fixed with ice-cold methanol for 10 min, and washed again in PBS. Incubation with a monoclonal anti-
-tubulin antibody (Sigma) diluted 1:1000 in 3% goat serum/PBS was performed overnight at 37°C. The antibody complexes were detected with tetramethylrhodamine isothiocyanate-conjugated goat antirabbit IgG (Sigma) and counterstained with 4',6-diamidino-2-phenylindole. Gray level images were acquired using a charged-couple device camera (CH250; Photometrics) mounted on a Leica DMRXA epifluorescence microscope and pseudo-colored using Leica Q-Fish software. Five hundred cells were analyzed for each cell line. Additional details for CGH, SKY, and immunohistochemistry protocols can be found on the Internet.4
Apoptosis Assay.
TUNEL assay for DNA fragmentation was done using an "In Situ Death Detection" kit (Roche) according to manufacturer’s recommendations.
Sequencing of p53 and BUB1.
Both p53 and BUB1 coding region were amplified by RT-PCR and directly sequenced. Total RNA was extracted by SV Total RNA Isolation System (Promega) and used for cDNA synthesis with random hexamers. The RT-PCR products were purified on an agarose gel and sequenced using Thermosequenase (Amersham).
Telomerase Assay.
Cellular extracts were assayed for telomerase activity with a PCR-based telomeric repeat amplification protocol (TRAPeze kit; Intergen) assay. PCR products were electrophoresed on 10% polyacrylamide gel and stained with silver staining kit (Amersham).
Tumorigenicity Assay.
Cells were trypsinized, resuspended in PBS, and inoculated s.c. in nude mice (Charles River Breeding Laboratories) at a concentration of 3 x 106 in 0.3 ml. As positive control, human Kaposi’s sarcoma cells were injected s.c. in nude mice at the same concentration. At least three mice for each cell line were injected, and mice were followed 
2 months to allow cell growth. At the end of this period, or when tumor mass reached 0.5 cm3, animals were sacrificed, and completed necropsy was performed.
|
RESULTS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Primary human fibroblasts at passage 4 were mock treated or treated with each of these antisense and control oligonucleotides, which include both sense and scrambled oligonucleotides, and thereafter, the effect on correct chromosome segregation was evaluated by analysis of chromosome number, the presence of micronuclei, and apoptosis.
As shown in Table 1
, antisense oligonucleotides directed against 9 of the 13 genes induced chromosome aneuploidy, although to a different extent. Inhibition of genes coding for SMC1, ARP1, and those involved in the formation of the mitotic spindle or its checkpoint (BUB1, BUB3,CENP-E, MAD2, INCENP, ZW10, and INTZW10) had a high, statistically significant number of aneuploid mitoses. In fact, the frequency of aneuploid cells ranges from 15% with BUB1 to 26% with BUB3 and SMC1 antisense oligonucleotides when compared with 2% in control cells. With the exception of CENP-B, very similar results were obtained when antisense inhibition was performed on a second human fibroblast cell line. However, treatments with CENP-A, CENP-C, and CENP-F antisense oligonucleotides resulted only in a slight increase in aneuploid cells. As the aim of our preliminary experiments was to perform a screening for the identification of genes giving rise to aneuploidy, we did not further investigate whether the results obtained with the four CENP family genes reflected a nonessential role in the maintenance of genome integrity or whether they were attributable to a failure of the antisense technology. Interestingly, the identification of putative genes conferring CIN will allow to test a recent mathematical model developed to study the effect of CIN on the evolution of cancer (28)
. Treatment with SMC1 antisense oligonucleotide, besides aneuploidy, also induced chromosome aberrations in 20% (subject 1) and 22% (subject 2) of the cells (Table 1)
, including gaps, whereas breaks were very rare. These data are consistent with the role of SMC proteins, which contribute to both condensin and cohesin complexes and with the phenotype of smc mutants in yeast, which show an increased rate of aneuploidy, defects in nuclear division, and partial chromosome decondensation (29)
.
|
. All these cell cultures showed a variable but increased number of micronuclei when compared with controls as shown in Fig. 2A
. ZW10 antisense-treated cells had the highest number (17%) of cells with micronuclei, followed by INTZW10 (10%), INCENP (9%), and BUB3 (9%) antisense-treated cells. Treatments with antisense against the other genes gave only a slight increase when compared with the baseline level of 3% observed in control cells. Additional defects observed in antisense-treated cells included occasional nuclear bridges. When FISH was performed using a pancentromeric probe, the micronuclei contained at least one centromere signal (Fig. 1, C and D)
, providing evidence that the micronuclei represented lagging chromosomes attributable to mis-segregation.
|
|
. All treatments, except CENP-A, CENP-B, CENP-C, and CENP-F antisense oligonucleotides, induced massive apoptosis ranging from 25% with ZW10 and MAD2 to 13% with INCENP (Fig. 2B)
. In contrast, control cells had only 3% of apoptotic cells. Thus, the high incidence of aneuploid cells correlates with an increase in apoptotic rate and micronuclei formation, suggesting that most aneuploid cells undergo elimination by programmed cell death.
Effect of ARP1, BUB1, and SMC1 Antisense Oligonucleotide Treatments on Long-term Cell Cultures.
As reported above, abnormal chromosome segregation leads to apoptosis, although its effects on long-term cultures have not been evaluated in vitro. Indeed, some cells with altered chromosome number could escape programmed death, and their fate is unknown. To investigate this point, we started long-term cultures of three antisense treated cells, ARP1, BUB1, and SMC1, according to an experimental protocol outlined in Fig. 3
. These genes were selected because they are thought to damage normal mitosis through different mechanisms, yet caused significant aneuploidy in our short-term test (Table 1)
. In addition, BUB1 has already been shown to be involved in human tumorigenesis (30)
. Fibroblast cells were treated for 9 days with antisense oligonucleotides and subsequently cultured as usual. With this protocol, a down-regulation of BUB1 and SMC1 protein synthesis was obtained, as shown by Western blotting with specific antibodies (Fig. 4, A and B)
. Reduction of the levels of ARP1 transcripts was investigated by RT-PCR, because by reducing primer concentration, a specific ARP1 band was seen only in control cells (Fig. 4C)
.
|
|
. Apoptosis correlated with the frequency of micronuclei (Fig. 2D)
and the number of apoptotic cells and remained elevated even 4 days after the treatment was discontinued. In these experimental conditions, SMC1 antisense oligonucleotide treatment induced the highest level of micronuclei and apoptotic cells when compared with ARP1 and BUB1. The presence of chromosomes in micronuclei was also confirmed by FISH (data not shown). To investigate whether, as a consequence of aneuploidy, a few cells could acquire AI growth features, we assessed the ability of cells to grow in soft agar. At the 25th passage, BUB1 antisense oligonucleotide-treated cells were able to form colonies in soft agar, whereas no colony was found in ARP1- and SMC1-treated and control cells. The colonies were transferred to standard culture conditions and are still in culture (passage 63).
Investigation of Consequences of BUB1 Inhibition.
Inhibition of a specific function by antisense technology needs to be thoroughly substantiated. In addition, although human fibroblasts usually do not give rise to AI colonies, a single clone could have arisen by chance. To exclude this possibility, we used alternative approaches to modify gene transcription by treating fibroblasts both with a specific ribozyme and by siRNA in the short-term assay (for 48 h, see "Materials and Methods") and counted aneuploid cells. As a matter of fact, 58.8% with ribozyme and 35% with siRNA of cells were aneuploid, whereas untreated or control ribozyme and siRNA-treated cultures showed the normal rate of aneuploidy (2–4%), thus confirming the specificity of the antisense results. These treatments do not cause BUB1 abnormality, because sequencing of this gene in three AI clones did not detect any mutation.
As stated above, inhibition of BUB1 expression at the protein level by Western blot was shown to occur in antisense-treated cells collected 9 days after the exposure to oligonucleotides. These data were also confirmed by a time course experiment where BUB1 synthesis was analyzed at different times during antisense treatment. In this experiment, specific bands were evident after 6 and 12 h of treatment but not after 24 and 48 h (Fig. 4D)
, whereas actin protein levels were unaffected. As actin is degraded during apoptosis (31)
, these results suggest that BUB1 inhibition is specific and not caused by aspecific degradation triggered by apoptotic events.
To confirm that inhibition of BUB1 function occurred in fibroblasts, we investigated whether the spindle checkpoint, in which BUB1 plays a fundamental role, is abrogated in antisense-treated cells. Antisense-treated cells were exposed to the spindle-destabilizing agent nocodazole, and their mitotic index was recorded. Consistent with the loss of spindle checkpoint, BUB1 antisense oligonucleotide-treated cells showed no peak at any time, whereas a peak at 18 h was seen in all control populations (Fig. 4E)
.
To confirm that antisense treatment can give rise to clones with an AI phenotype, we repeated the experiments with five independent fibroblast cultures. Two independent clones growing in soft agar were identified in antisense-treated cultures, whereas no growth was seen in soft agar plates, where untreated and control oligonucleotide-treated cells were seeded. These two clones were transferred to standard culture conditions and are still growing. All together, the frequency of anchorage independence after antisense treatment was 1.5 x 10-6 (in still ongoing experiments performed with ribozyme-treated cultures, the estimated frequency was 3 x 10-6; data not shown).
Cytogenetic and Molecular Analysis of Polyclonal and Clonal Fibroblast Cells.
Karyotype analysis by G banding of untreated and BUB1 antisense-treated fibroblasts was performed during in vitro cultures at various passages. Although untreated and control oligonucleotide-treated cells showed a low percentage of aneuploid cells, fibroblasts treated with BUB1 antisense show aneuploidy (8%, compared with 1–2% of the other populations). Apparently, temporary inhibition of BUB1 is sufficient to confer an aneuploid phenotype that persists in culture. However, no recurrent chromosomal abnormality was identified.
Aneuploidy was also detected in the three AI clones, which were also investigated by SKY and CGH. SKY analysis showed no chromosome rearrangements, i.e., chromosome translocations or other structural defects, in AI clones (Fig. 5A)
. Moreover, the analysis of 22 metaphases showed a widespread variability in chromosome number that range from 36 to 74 with a predominance of chromosome losses. SKY analysis of individual AI clones suggests specific chromosome aneuploidy, such as the recurring loss of chromosome 6 in 5 of 8 metaphases in clone 1 and loss of chromosome 21 in 4 of 6 metaphases in clone 2. This finding, however, could not be corroborated with CGH analysis, because the results essentially revealed a balanced genome (Fig. 5B)
. CGH can detect imbalances when present in >50% of the cells. Taken together, these data suggest that the development of AI phenotype is not attributable to specific chromosome rearrangements but that the acquisition of the AI phenotype is accompanied by a manifestation of genome instability of the chromosomal type (CIN). A selection for specific chromosomal aneuploidies, as observed in cancer cells, has not yet occurred.
|
, providing evidence that AI clones have not yet reached a senescent stage, whereas control and antisense oligonucleotide-treated cells did so. Senescent polyclonal populations (passage 44) showed no specific chromosome aneuploidy when analyzed by CGH, whereas the level of micronuclei and apoptosis was very low (0.2%), in agreement with published data on senescent cells (32)
.
|
.
|
, suggesting that in AI clones, p53 function is maintained. Moreover, no mutations were found in AI clones by sequencing the entire protein coding region of the p53 gene (data not shown). This is consistent with the consideration that antisense oligonucleotide treatment does not induce nucleotide mutations. Furthermore, we found no significant increase in either ouabain or 6-thioguanine resistance of AI cells, because the number of colonies growing in selective media did not appreciably vary when compared with control cells (0.2 versus 0.33 x 10-6 cells for ouabain and 1 versus 1.4 x 10-6 cells for 6-thioguanine resistance). These values are in agreement with the frequencies reported for normal cells (33
, 34) , suggesting that BUB1 is not involved in specific processes that prevent the formation of mutations at nucleotide levels.
Another mechanism which has been suggested to lead to CIN is the presence of centrosome abnormalities (17)
. To further investigate this possibility, centrosomes were visualized microscopically by labeling the cells with antibody against -tubulin. In both AI clones and polyclonal fibroblasts, centrosomes had visible, clearly defined, dot-shaped structures (Fig. 7, B and C)
, and the number of cells with abnormal centrosomes was similar in all populations (data not shown). It is, however, important to note that these cells have not yet acquired a neoplastic phenotype, because they were unable to grow when injected s.c. in nude mice (data not shown).
|
DISCUSSION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Although extremely interesting, data acquired in rodents cannot necessarily be transferred to humans. Differences exist between these species (5) , and the data obtained in mice must be independently corroborated by human studies. To accomplish this task, most of the approaches have investigated the growth potential of cells cultured in vitro. Historically, fibroblast cells have been the tissue of choice (35) . Fundamental studies have suggested that human fibroblast cells have a limited life span (Hayflick limit) and that, different from rodent cells, the rate of spontaneous transformation is considerably lower. This proliferative block, which is thought to represent a genetic guard to neoplastic transformation (36) , is at least in part related to the loss of the ability to faithfully replicate telomeric ends (37 , 38) . Human fibroblasts transfected with the catalytic subunit of telomerase are immortalized, growing continuously in vitro, but are not aneuploid, do not acquire AI ability, and do not grow in nude mice (19 , 38 , 39) . Additional genetic changes are needed to confer the complete neoplastic phenotype to fibroblasts (4) .
Most solid tumors are aneuploid (40) , yet aneuploidy is rarely found in normal cells. It is generally believed that aneuploidy in normal cells is commonly followed by apoptosis. However, the possibility exists that these aneuploid cells can escape apoptosis and show a selective growth advantage, which could represent the substrate to achieve additional genetic changes. In this case, although being insufficient for acquisition of complete tumoral phenotype, aneuploidy could be a very early step in transformation. To investigate the role of aneuploidy in tumorigenesis has been difficult even in the mouse model because animals in which several genes interfering with normal chromosome segregation have been inactivated often die in utero because of profound derangement of normal cellular replication (41) . However, it has recently been shown that haploinsufficiency in Mad2 spindle checkpoint gene in mice gives rise to chromosomal instability and high lung tumor incidence (42) . This suggests that aneuploidy could play a role in the early phase of transformation at least in rodent cells.
To investigate whether aneuploidy could play a role also in early tumorigenesis in humans, we set up a gene inhibition system in human fibroblasts. Both senescence and transformation have been extensively studied in this model, and many correlations between these two processes have been identified. We reasoned that by interfering with chromosome segregation, we could obtain aneuploid cells without causing any concomitant DNA damage at the sequence level and that the fate of these cells could then be followed in vitro by direct analysis of cultured cells. Inhibition by antisense technology, however, must be strongly corroborated. We think that our approach achieved substantial inhibition of BUB1 function for the following reasons: (a) the same pattern of aneuploidy without chromosome breakage, consistent with the known BUB1 role in spindle checkpoint, was obtained in 13 independent experiments, using both antisense oligonucleotides and ribozyme but not with control oligonucleotides; the greater number of aneuploid cells obtained with ribozyme could be attributable to a stronger inhibition usually associated with the ribozyme technique compared with the oligonucleotide approach; (b) similar results were also obtained with the use of the siRNA approach, a new technology for inhibition of gene expression; (c) using an antibody against BUB1, no protein was immunoprecipitated in antisense-treated cells after 24 h of treatment, and its disappearance followed a decrease after 6 and 12 h; and (d) nocodazole exposure performed during treatment with antisense showed loss of the ability of treated cells to stop the mitotic process, demonstrating functional inactivation of the spindle checkpoint.
For all these reasons, we think that the effects found in our cell cultures reported in the present study can be directly related to a substantial interference with the function of BUB1 in spindle checkpoint. With this approach, we were able to demonstrate that aneuploidy can have different effects on the fate of the cells. As shown by our data, most aneuploid cells undergo apoptosis, but a substantial portion of them are able to survive without profound alterations in their proliferative potential and still able to senesce at the appropriate time, suggesting that the genetic program leading to senescence in normal fibroblast is not automatically altered in aneuploidy. However, our data raise the possibility that with rare frequencies, aneuploid cells could acquire different proliferative abilities that confer to them the ability to grow without anchorage dependence and postpone their senescence.
Delayed senescence of human fibroblasts has been achieved by exogenous telomerase expression. In mice, several genes, including p53, are known to play a role in replicative or Ras-induced senescence. However, our analysis did neither find abnormalities in p53 nor telomerase activity, although we cannot exclude a telomerase-independent alternative lengthening of telomeres. These data would suggest that the mechanisms leading to the insurgence of AI clones are different from those involving these genes. Likewise, centrosome abnormalities were absent in most cells, suggesting that they are not responsible for chromosomal instability in our system. However, we cannot exclude the formal possibility that mutations in other genes contributed to the acquisition of AI phenotype, although we think that this is unlikely in view of the results obtained in oubain and thioguanine resistance experiments.
The clones identified here have some features reminiscent of CIN. Theoretically, temporary inhibition of spindle checkpoint could allow the insurgence of chromosome instability, but the cells should switch to a stable aneuploid state when BUB1 activity is restored. The reason why this does not occur in our clones is not clear at the moment, but it could be that, by escaping the checkpoint for several mitoses, the cells could become insensitive to the restoration of the checkpoint. The mechanisms underlying this phenomenon need to be further investigated.
CIN could be the first determinant of neoplastic transformation, allowing different combinations of chromosomes to take place, therefore providing a substrate for further selection on the basis of small growth advantage based on a specific set of chromosome content. Alternatively, CIN could be a late event in cancer pathogenesis, representing a stage in which all of the checkpoints have been inactivated and mitoses proceed completely abnormally. Our finding that both the antisense-treated aneuploid polyclonal fibroblast population and the three AI clones present a largely variable chromosome content suggests that CIN, in selected situations, could be an early event in cell transformation. A likely possibility at this point is that both mutations and aneuploidy could be the first event in the multistep process leading to cancer.
In colon cancer, CIN appears to be mutually exclusive with MIN. Thus far, despite many investigations, although genes responsible for MIN have been identified, no germ-line inactivation of spindle checkpoint genes has been identified in human cancer (15) . In addition, most colon cancers with CIN appear to be sporadic and not related to familial cancer syndromes. We demonstrate here that temporary inactivation of checkpoint is sufficient for CIN to occur in cultured cells and that this characteristic is maintained after the toxic insult has been eliminated. Interference with normal spindle checkpoint can be caused by many toxic agents, which can be found in diet or are endogenously produced in the colonic mucosa. The possibility exists that, although MIN is mainly of genetic origin, CIN arises in colon cells after environmental-driven, temporary inhibition of normal gene function. CIN could then provide few cells a peculiar chromosome content, allowing a small but sufficient growth advantage, which could be the substrate for additional genetic and/or epigenetic changes. Interestingly, this "hit-and-run" mechanism of genetic interference would leave no signature, which could be identified by genetic analysis.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
1 Supported in part by grants from "Stem 2001" from Ministero della Salute, Fondi per gli Investimenti della Ricerca di Base (FIRB) (RBNE019J9W), and "Progetto Genomica Funzionale" from Ministero dell’Istruzione dell’Università e della Ricerca (MIUR) (to P. V.) and FIRB (RBNE01RNN7; to A. M.). hBUB1 rabbit antibody was a kind gift of T. Yen (Fox Chase Cancer Institute, Philadelphia, PA). A. Musio is a fellow of Associazione Rimini Genoma Onlus. 
2 To whom requests for reprints should be addressed, at Istituto di Tecnologie Biomediche, Consiglio Nazionale delle Ricerche, Via Fratelli Cervi, 93, 20090 Segrate, Milan, Italy. Phone: 39 02 26422632; Fax: 39 02 26422660; E-mail: antonio.musio{at}itb.cnr.it
3 The abbreviations used are: CIN, chromosome instability; MIN, microsatellite instability; AI, anchorage-independent; CGH, comparative genomic hybridization; SKY, spectral karyotyping; MMC, mitomycin C; siRNA, small interfering RNA; RT-PCR, reverse transcription-PCR; FISH, fluorescence in situ hybridization; TUNEL, terminal deoxynucleotidyltransferase-mediated dUTP nick end labeling.
4 Internet address: http://www.riedlab.nci.nih.gov/.
Received 1/17/03. Accepted 3/27/03.
|
REFERENCES |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  P. Morciano, C. Carrisi, L. Capobianco, L. Mannini, G. Burgio, G. Cestra, G. E. De Benedetto, D. F.V. Corona, A. Musio, and G. Cenci A conserved role for the mitochondrial citrate transporter Sea/SLC25A1 in the maintenance of chromosome integrity Hum. Mol. Genet., November 1, 2009; 18(21): 4180 - 4188. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E Burum-Auensen, P M DeAngelis, A R Schjolberg, J. Roislien, S N Andersen, and O P F Clausen Spindle proteins Aurora A and BUB1B, but not Mad2, are aberrantly expressed in dysplastic mucosa of patients with longstanding ulcerative colitis J. Clin. Pathol., December 1, 2007; 60(12): 1403 - 1408. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I. Perez de Castro, G. de Carcer, and M. Malumbres A census of mitotic cancer genes: new insights into tumor cell biology and cancer therapy Carcinogenesis, May 1, 2007; 28(5): 899 - 912. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Musio, C. Montagna, T. Mariani, M. Tilenni, M. L. Focarelli, L. Brait, E. Indino, P. A. Benedetti, L. Chessa, A. Albertini, et al. SMC1 involvement in fragile site expression Hum. Mol. Genet., February 15, 2005; 14(4): 525 - 533. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S.C. Reshmi and S.M. Gollin Chromosomal Instability in Oral Cancer Cells Journal of Dental Research, February 1, 2005; 84(2): 107 - 117. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Myung, S. Smith, and R. D. Kolodner Mitotic checkpoint function in the formation of gross chromosomal rearrangements in Saccharomyces cerevisiae PNAS, November 9, 2004; 101(45): 15980 - 15985. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. C. Deocaris, S. C. Kaul, K. Taira, and R. Wadhwa Emerging Technologies: Trendy RNA Tools for Aging Research J. Gerontol. A Biol. Sci. Med. Sci., August 1, 2004; 59(8): B771 - B783. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. A. Lee, M. K. Keutmann, M. L. Dowling, E. Harris, G. Chan, and G. D. Kao Inactivation of the mitotic checkpoint as a determinant of the efficacy of microtubule-targeted drugs in killing human cancer cells Mol. Cancer Ther., June 1, 2004; 3(6): 661 - 669. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Hu, T. Wang, G. D. Stormo, and J. I. Gordon RNA interference of achaete-scute homolog 1 in mouse prostate neuroendocrine cells reveals its gene targets and DNA binding sites PNAS, April 13, 2004; 101(15): 5559 - 5564. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Cotsiki, R. L. Lock, Y. Cheng, G. L. Williams, J. Zhao, D. Perera, R. Freire, A. Entwistle, E. A. Golemis, T. M. Roberts, et al. Simian virus 40 large T antigen targets the spindle assembly checkpoint protein Bub1 PNAS, January 27, 2004; 101(4): 947 - 952. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Cancer Research | Clinical Cancer Research |
| Cancer Epidemiology Biomarkers & Prevention | Molecular Cancer Therapeutics |
| Molecular Cancer Research | Cancer Prevention Research |
| Cancer Prevention Journals Portal | Cancer Reviews Online |
| Annual Meeting Education Book | Meeting Abstracts Online |
), control oligonucleotide (
), and antisense-treated cells (
; E).
