• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of neoplasiaLink to Publisher's site
Neoplasia. Aug 1999; 1(3): 241–252.
PMCID: PMC1508077

Chromosomal and Extrachromosomal Instability of the cyclin D2 Gene is Induced by Myc Overexpression

Abstract

We examined the expression of cyclins D1, D2, D3, and E in mouse B-lymphocytic tumors. Cyclin D2 mRNA was consistently elevated in plasmacytomas, which characteristically contain Myc-activating chromosome translocations and constitutive c-Myc mRNA and protein expression. We examined the nature of cyclin D2 overexpression in plasmacytomas and other tumors. Human and mouse tumor cell lines that exhibited c-Myc dysregulation displayed instability of the cyclin D2 gene, detected by Southern blot, fluorescent in situ hybridization (FISH), and in extrachromosomal preparations (Hirt extracts). Cyclin D2 instability was not seen in cells with low levels of c-Myc protein. To unequivocally demonstrate a role of c-Myc in the instability of the cyclin D2 gene, a Myc-estrogen receptor chimera was activated in two mouse cell lines. After 3 to 4 days of Myc-ERTm activation, instability at the cyclin D2 locus was seen in the form of extrachromosomal elements, determined by FISH of metaphase and interphase nuclei and of purified extrachromosomal elements. At the same time points. Northern and Western blot analyses detected increased cyclin D2 mRNA and protein levels. These data suggest that Myc-induced genomic instability may contribute to neoplasia by increasing the levels of a cell cycle-regulating protein, cyclin D2, via intrachromosomal amplification of its gene or generation of extrachromosomal copies.

Keywords: Myc, cyclin D2, genomic instability, expression, extrachromosomal elements

Introduction

The oncoprotein Myc plays a crucial role in transformation of a wide variety of cell types (1–3). In a subset of these tumors, and particularly in neoplasms of B lymphocytes, c-Myc expression is constitutively upregulated as a result of one of three processes associated with genomic instability: 1) gene amplification (4–5); 2) retroviral (6) or transposon (7) insertion; or 3) chromosomal translocation of c-Myc to immunoglobulin (Ig) loci. The last process has been observed in chicken bursal lymphomas, human Burkitt and other non-Hodgkin's B-cell lymphomas, mouse plasmacytomas, and rat immunocytomas (8–10). In several forms of neoplasia, c-Myc gene copy numbers and protein levels have been used as prognostic markers (11). The molecular basis for the link between c-Myc overexpression and transformation has not been fully elucidated in these systems.

Previous studies have suggested that Myc contributes to neoplasia by affecting cell cycle progression (12–15). c-Myc expression is tightly controlled in normal cells at transcriptional, posttranscriptional, translational, and posttranslational levels (3). In diploid primary cells, c-Myc is upregulated in G1 and downregulated shortly after the entry into the S-phase (16;17). c-Myc antisense oligonucleotides have been shown to block the transition from G1 to S (18). Cells with c-Myc overexpression have shown shortened G1-phases (19;20). On the other hand, cells with one disrupted c-Myc allele had a prolonged G1-phase (21), and disruption of both c-Myc alleles in a cell line prolonged the G2-phase as well (22), prolonging the cell cycle duration significantly and proving lethal in vivo (23).

The regulated expression of cyclins and cyclin-dependent kinases (CDKs) is critical to the normal progression through the cell cycle of untransformed cells. In contrast, immortalized and transformed cells often exhibit dysregulated expression of cyclins, CDKs, and CDK inhibitors. It was reported earlier that Myc plays a role in the expression of cyclins A, D1, and E (12–14). Moreover, c-Myc expression and cell cycle progression are linked through the activation of G1 cyclins and CDKs (15). It has been reported that Myc, when heterodimerized with Max, transactivates the Cdc 25 gene, which encodes a CDK-activating phosphatase, suggesting a mechanism through which Myc could influence the cell cycle (24).

The focus of our work has been on Myc-dependent genomic instability. We recently demonstrated that c-Myc overexpression is associated with the nonrandom amplification and rearrangement of the dihydrofolate reductase gene (Dhfr; Refs (25,26)) and the gene encoding the R2 subunit of ribonucleotide reductase (R2, Kuschak and colleagues, now in press, Gene).

In the present study, we show evidence of Myc-dependent genomic instability of the cyclin D2 gene, with an increase in intrachromosomal copy numbers or extrachromosomal elements bearing cyclin D2 sequences. Both amplification events occur concomitant with an increase in cyclin D2 gene products. These findings link c-Myc overexpression and cell cycle regulation for the first time at the level of genomic instability of this G1 cyclin. Based on these findings, we propose a model of Myc-dependent genomic instability and neoplasia.

Materials and Methods

Cell Lines and Tissue Culture

Human breast ductal adenocarcinoma T47D and mouse B lymphoma WEHI 231 were obtained from the American Type Culture Collection, Rockville, MD. Mouse plasmacytomas, MOPC 265 and MOPC 460D; the human colorectal carcinoma line, COLO320HSR; and primary human fibroblasts, GL30/92T, have been previously described (26–28). The spectrum of mouse B-lymphocytic cell lines has been presented in detail earlier (29). Cells were propagated in RPMI 1640 (Biofluids, Inc, Rockville, MD), supplemented with 10% heat-inactivated (30 minutes, 56°C) fetal bovine serum (Gibco/BRL, Germantown, MD), 2 mmol/L glutamine, 100 U/mL penicillin, and 100 µg/mL streptomycin. Culture medium for B-lymphoid cell lines also contained 50 µmol/L 2-mercaptoethanol. We generated an in vitro line of mouse pre-B lymphocytes by transformation of BALB/c bone marrow cells with A-MuLV (30). These cells were subsequently transfected with LXSN-bcl-2, a mouse bcl-2-expressing vector (31), and pBabePuroMyc-ERTm, an expression vector (32) with which the human Myc protein can be activated by 100 nmol/L 4-hydroxytamoxifen (4HT, Research Biochemicals International, Natick, MA). We also produced a line of mouse fibroblasts in which Myc could be upregulated by 4HT owing to stable transfection of pBabePuroMyc-ERTm into Ψ2 cells (33).

Cloning and Sequencing of Mouse Cyclin D2 cDNA and 5′ Genomic Flank

A cDNA library of the mouse pre-B cell, 18–81, in lambda ZAP-2, was screened under relaxed conditions with Cyl1 (34), a partial cDNA for murine cyclin D1, a generous gift of Dr. Charles Sherr. Several clones that encoded mouse cyclin D2 were isolated, rescued as pBlueScript clones, characterized, and sequenced. A probe derived from the clone with the longest (1255-bp) insert was sequenced and found to have a coding region identical to the mouse cyclin D2 cDNAs in the literature (35). This probe was used to screen a partial EcoRI library of BALB/c liver DNA in EMBL-4 arms, a generous gift of Drs. Linda Byrd and Konrad Huppi. One positive clone that contained a 17.1-kb insert was isolated, purified, and digested to completion with EcoRI. Only one of the three EcoRI fragments that were generated from two internal EcoRI sites, a 5.4-kb fragment, hybridized with the 5′ end of the cyclin D2 cDNA probe, and it was subcloned into pBlueScript for further study. Partial sequencing of this fragment revealed that the 3′ 505 base pairs were identical to the 5′ portion of our cyclin D2 cDNA (and that of Kiyokawa and colleagues [35]). The 3′ 194 base pairs contained an AUG, followed by an open reading frame, and the adjacent 301 upstream base pairs contained the 5′ untranslated sequence of the cDNA. The remainder was considered 5′ flank in which regulatory motifs might be located. A more complete sequence of the mouse 5′ flank is being generated and will be reported elsewhere.

Probes

A 700-bp PstI-fragment of our mouse cyclin D2 cDNA was used to probe Southern and Northern blots, and the 5.4-kb genomic clone (discussed earlier) was used as a probe for fluorescent in situ hybridization (FISH). A strategy similar to that described above was used to isolate cDNAs for mouse cyclins D1 and D3 ((36), generous gifts from Paul Hamel, University of Toronto). Each was used as an 800-bp EcoRI/HindIII fragment that contained the entirety of the coding region. Each of the cyclin D cDNAs has a 100-bp region that is 90% identical. Thus, repeated and sequential hybridization of RNA blots was necessary to determine which bands were unique to the cyclin D member being probed. Only the unique RNA bands are shown in the figures. The human cyclin D2 cDNA was a kind gift from Gordon Peters (37). It was used as a 1.2-kb NotI-Xhol-fragment. The R1 probe was a 1.5-kb BamHI fragment of mouse ribonucleotide reductase R1 subunit cDNA (38). The cDNA clone for mouse c-Myc, pMc-Myc54 (39), was the kind gift of Kenneth B. Marcu, from which a 0.6-kb Hind III-Sst I fragment was used as an exon I-specific probe and a 1.0-kb Sst I-Hind III fragment was used as an exons 2+3 probe for the Myc sequences expressed in pBabePuroMyc-ERTm (32). Mouse cyclins C and E probes were gifts of Steven Reed. They were used as 1.1- and 2.5-kb EcoRI fragments, respectively, for hybridization. The cDNA for glyceraldehyde-3-phosphate-dehydrogenase (GAPDH) was the kind gift of Dr. Marc Piechaczyk (40). It was used as a 1.3-kb PstI fragment. Northern and Southern blots were probed by labeling the aforementioned probes with 32P by using Nick Translation (GIBCO/BRL, Germantown, MD) or random priming by standard methods.

Assays for Genomic Instability: Gene Amplification and Extrachromosomal Elements

Gene dosage was examined by using Southern blot analysis (41) and FISH of metaphase and interphase chromosomes (26,42). Evaluation of metaphase spreads and interphase nuclei was performed with a Zeiss Axiophot microscope and a CCD camera (Photometrics, Tucson, Az). One hundred to 150 metaphases and interphases were evaluated in each of three independent experiments. Extrachromosomal fluorescent signals were considered specific when they also stained with 4′, 6′ diamidino-2-phenylindole (DAPI) (1 µg/mL) or propidium iodide (PI) (1 µg/mL).

RNA Isolation and Northern Blotting

Total RNA or Poly(A)+RNA was isolated from cells as previously reported (29). Five micrograms of Poly(A)+RNA or 15 µg of total RNA was fractionated on a 1% agarose gel containing formaldehyde. The RNA was transferred to a Hybond-N membrane (Amersham, Arlington Heights, IL) by capillary blotting and hybridized with 32P-labeled cDNA probes, as indicated in the figure legends. Radioisotopic labeling was performed with the Nick Translation System (GIBCO/BRL) according to the manufacturer's protocol. The membranes were hybridized overnight with 3x106 dpm/mL probe, washed with 0.1x Standard Saline Citrate, 0.1% Sodium Dodecyl Sulfate at 20°C and exposed to X-ray film overnight at -80°C. For sequential hybridization of the same blot with different probes, membranes were stripped with boiling water.

Western Blotting

Western blots were performed on lysates of pre-B cell cultures as previously described (43) except that protein concentration was determined with the BCA Protein Assay (Pierce, Rockford, IL), and 10 µg was loaded per lane. The immunoreactive bands were identified by the ECL Western blotting detection system (Amersham, Arlington Heights, IL). The anti-cyclin D2 (M-20) and anti-cyclin D3 (C16) were polyclonal antibodies purchased from Santa Cruz Biotechnology (Santa Cruz, CA). The anti-actin (mouse monoclonal, clone AC-40) was from Sigma ImmunoChemicals (St. Louis, MO). The HRP-Goat anti-Rabbit IgG and HRP-Goat anti-Mouse IgG and IgM were purchased from Axell (Westbury, NY).

Extrachromosomal Elements

Cells were subjected to a procedure designed to separate chromosomal from nonchromosomal DNA (44). Cells were suspended at 1–5x107/mL and lysed in 0.6% SDS/0.01 mol/L EDTA at room temperature for 10 to 20 minutes. Then 5 mol/L NaCl was added to bring the suspension to 1 mol/L NaCl, and precipitation was allowed to proceed at 4°C overnight. The supernatant from a centrifugation at 35,000g for 30 minutes at 4°C (“Hirt extract”) was applied to glass slides for FISH (see next section) or to grids for electron microscopy (EM) examination. For EM, Hirt extracts were diluted to a concentration of approximately 1 µg/mL in 20 mmol/L MgCl2/30 mmol/L triethanolamine (45). Aliquots were placed on formvar/carbon-coated grids, fixed with 0.1% glutaraldehyde in White's saline (46), and negatively stained with 3% uranyl acetate. Grids were allowed to air dry and then shadowed with tungsten to enhance resolution. Grids were examined in a Philips Model 420 transmission electron microscope at 80 kV.

FISH on Purified Extrachromosomal Elements

Extrachromosomal elements (EEs) were isolated according to a modification (47) of the protocol by Hirt (44). They were diluted 1 to 1 in freshly prepared methanol to acetic acid (3:1) and dropped onto slides. Fixations were carried out as described earlier (25,26). FISH was performed after RNAse and pepsin treatments. DAPI was used to stain the DNA of the EEs. Hybridization signals were considered specific if they colocalized with DAPI-stained EEs. Hybridizations were carried out with a panel of probes, including Dhfr and ribonucleotide reductase subunit 2 (R2), that had been shown to be present on EEs of 4HT-activated pre-B lymphoma cells (47). Relative fluorescence intensities were measured with IPLab Spectrum software, version 3.1 (Scanalytics, Fairfax, VA).

Results

Northern Blot Analysis of Cyclin D2 Expression in Murine B-Lymphocytic Lines With Different Degrees of B-Cell Maturation

Figure 1 shows a blot of poly(A)+RNA from a series of mouse B-cell lymphoma cell lines of increasing maturation from left to right (29). This blot was sequentially hybridized with four murine G1 cyclin probes. The cyclin D3 probe revealed strong 2.3-kb bands in four cell lines of early B lymphocytes but barely detectable levels in plasmacytomas. The cyclin D1 probe showed a very strong 3.9-kb band in the myeloid-pro-B line in lane 1, strong bands in two B-cell lines, lanes 5 and 8, but very low levels in the remaining RNA samples. Cyclin E transcripts were virtually undetectable (not shown). When normalized to the GAPDH control hybridization signals, the highest level of expression of the predominant cyclin D2 mRNA species (6.5-kb) was seen in the four plasmacytomas (lanes 11–14). In addition, several smaller cyclin D2 mRNA species are prominent, chiefly in these four lanes. Plasmacytomas are known to be rich in Myc mRNA (48), and this is the case for these four lanes. There is not a clear correlation between levels of Myc and cyclin D2 mRNA, but a similar pattern of high levels of Myc and cyclin D2 mRNA was also seen in blots of RNA from 45 additional plasmacytomas (data not shown). This series included plasmacytomas with t(12;15) and t(6;15) translocations and tumors without translocations but with Myc upregulation resulting from stable integration of Myc-expressing recombinant retroviruses (28).

Figure 1
Cyclin expression in mouse B-lymphocytic tumors. Poly(A)+ RNAs (5 µg) from a series of mouse B-lymphocytic cell lines (28) are arranged from left-to-right in increasing degree of maturation. HAFTL-1 3g4 and HAFTL-1 are two related clones of pro-B ...

Established Mouse and Human Tumors With c-Myc Overexpression Have Southern Blot Evidence of Amplification of the Cyclin D2 Locus

The stability of the cyclin D2 locus was characterized in two mouse B-lymphoid lines, MOPC 460D, a plasmacytoma that constitutively overexpresses c-Myc due to Myc/Ig chromosome translocation (28) , and WEHI 231, a lymphoblastoid tumor with low Myc protein levels (26) . Southern blot analyses showed that the cyclin D2 gene was chromosomally amplified in MOPC 460D (Figure 2, upper left panel, filled arrowheads), but not in WEHI 231 cells. Mouse ribonucleotide reductase subunit R1 (R1, Figure 2, lower panels) and cyclin C (not shown), genes that are retained as single-copy genes irrespective of Myc protein levels, were used as reference genes.

Figure 2
Southern blot analyses of mouse and human cell lines hybridized with murine and human cyclin D2 cDNA probes, as indicated (upper panels). Mouse lines: WEHI 231 B-cell lymphoma (low Myc) and MOPC 460D plasmacytoma (high Myc); Human lines: GL30/92T primary ...

These analyses were extended to human cell lines: The colon carcinoma line COLO320HSR, a classic example of c-Myc gene amplification and overexpression (28-fold higher Myc protein levels than GL30/92T primary human fibroblasts), and the breast cancer line T47D, which expresses 11 times higher c-Myc protein levels than GL30/92T, a fibroblast of normal genotype (26). COLO320HSR and T47D displayed chromosomally amplified bands of cyclin D2 gene hybridization in Southern blots (Figure 2, upper right panel, filled arrowheads), compared with the control R1 hybridizations (Figure 2, lower right panel), whereas primary human fibroblasts did not shown cyclin D2 amplification. In T47D, the cyclin D2 gene is partially deleted from the chromosomes as indicated by missing genomic bands in the Southern blot (open arrowheads in Figure 2). Such deletions appear to reflect an additional form of genomic instability of this locus in cells that overexpress Myc.

Cyclin D2 Amplification Involves the Generation of Extrachromosomal Elements in COLO320HSR and MOPC 460D

DAPI or PI staining of metaphase chromosome spreads of COLO320HSR showed the presence of EEs. FISH studies of these preparations showed amplified signals of the cyclin D2 gene on chromosomes and the EEs (Figure 3A). Paired small white arrows indicate the wild-type, endogenous cyclin D2 loci. Larger arrows point out the cyclin D2-containing EEs. A similar analysis of the BALB/c plasmacytoma MOPC 460D also showed EEs that contained cyclin D2 sequences (Figure 3B).

Figure 3Figure 3
FISH studies with cyclin D2 probes and detection with fluorescein isothiocyanate-labeled antidigoxigenin antibody. The small arrows (frequently paired) point to dots that indicate the positions of the endogenous, germ-line chromosomal cyclin D2 loci. ...

Extra- and Intrachromosomal Amplification is Gene-Specific

EEs contain only a subset of gene sequences. The Southern blot data showed amplification of cyclin D2 but not of the negative controls, cyclin C or R1. Similarly, FISH studies of COLO320HSR and MOPC 460D showed no extrachromosomal elements that hybridized with cyclin C (data not shown).

Induced Upregulation of Myc Activity in Mouse Pre-B Cells Results in Cyclin D2-Containing Extrachromosomal Elements and Increased Cyclin D2-mRNA and Protein After 3 Days

A mouse pre-B cell line derived from bone marrow cells by transformation with A-MuLV was stably transfected with pBabePuroMyc-ERTm, an inducible Myc expression vector that constitutively expresses Myc-ERTm, a chimeric protein that contains human c-Myc, linked to a mutated estrogen receptor (32). The Myc of the chimera is activated by addition of 4HT. After 3 days of stimulation by 4HT, numerous EEs could be seen in metaphase chromosome spreads after Giemsa staining of chromatin (not shown) and DAPI staining of DNA (Figure 3D). Some of the EEs were shown by FISH analysis to carry cyclin D2 sequences (indicated by large white arrows). This evidence of genomic instability involving the cyclin D2 gene was not seen in the same cells without tamoxifen stimulation (Figure 3C) nor in 4HT-treated pre-B cells that lack Myc-ERTm (data not shown). In the untreated cells, only single-copy, endogenous cyclin D2 signals (indicated by small arrows) were seen.

Northern blots were prepared from total RNA from the A-MuLV-transformed pre-B cells before and after stable introduction of pBabePuroMyc-ERTm and after different periods of stimulation by 4HT. Figure 4A shows the results of successive hybridizations of this blot with cyclin D2 and other germane probes. Cyclin D2 message expression is clearly elevated after 4 and 6 days of Myc activation by 4HT, when compared with the ethidium bromide-stained 28S ribosomal RNA bands in each lane. Transcripts from the retroviral pBabePuroMyc-ERTm in the stably transfected line, shown in the right panels, are detected with a Myc exon 2+3 probe. As often happens (3) endogenous c-Myc expression (shown by a 2.4-kb band that hybridizes with Myc exon 1 probe) decreased after 4HT activation of the exogenous Myc-ERTm (which lacks exon 1). There was no detectable mRNA for cyclin D1 (not shown), and no significant change in mRNA level of cyclin D3 was seen during the treatment with 4HT.

Figure 4
Cyclin D2 expression after Myc activation. A. Northern blots of total RNA (15 µg) from cultures of mouse pre-B cells stably transduced with v-Abl (A-MuLV) or with v-Abl plus murine Bcl-2 plus pBabePuroMycERTm. Both cell lines were treated with ...

Western blots of lysates from the 4HT-treated cells just described were probed with anti-cyclin D2 and anti-cyclin D3 antibodies and, as a loading control, with anti-actin antibody. As shown in Figure 4B, levels of cyclin D2 protein, but not cyclin D3, gradually rose in parallel with the levels of cyclin D2 mRNA after Myc activation by 4HT in the pBabePuroMyc-ERTm-containing cells but not in the pre-B cells that lack this vector. Thus, both elevated cyclin D2 mRNA and protein levels appear simultaneously with the cyclin D2 genomic amplification, not prior to it.

Induced Upregulation of Myc Activity in Mouse Fibroblasts Also Leads to the Generation of Cyclin D2-Containing Extrachromosomal Elements and Increased Cyclin D2 mRNA

A mouse fibroblast line, derived by transfection of Ψ2 cells with pBabePuroMyc-ERTm, was stimulated with 4HT to induce increased Myc activity. After 3 days of stimulation by 4HT, numerous cyclin D2-containing EEs (indicated by large white arrows) could be seen in metaphase chromosome spreads and in interphase nuclei (Figure 3G and H). This evidence of genomic instability and of cyclin D2 gene amplification were not seen in the same cells without prolonged tamoxifen stimulation (see Figure 3F). A control FISH study of 4HT-stimulated Ψ2 cells that do not bear the Myc-ERTm expression vector showed no cyclin D2-hybridizing EEs (Figure 3E). In both control panels, Figures 3E and and3F,3F, only single-copy, endogenous cyclin D2 loci could be seen in metaphases and interphase nuclei.

Northern blots were prepared from total RNA from the fibroblasts with and without stable integration of pBabePuroMyc-ERTm and after different periods of stimulation by 4HT. Figure 5C shows the results of successive hybridization of this blot with cyclin D2 and other probes. As with the pre-B cells, cyclin D2 message expression is clearly elevated after several days of Myc activation by 4HT, when compared to the GAPDH loading control. Thus the fibroblast data parallel the results seen in pre-B cells, with increased cyclin D2 expression accompanying, but not preceding, appearance of cyclin D2-containing extrachromosomal elements.

Figure 5
Electron microscopy of EEs from pBabePuroMycEReTm-transduced pre-B cells prepared according to Hirt (44) before (A) and after 3 days of Myc activation by 4HT (B–D). These extracts were placed on formvar-covered grids, stained with uranyl acetate, ...

Myc-Induced Extrachromosomal Elements are DNA-Containing Circles, Some of Which Bear Cyclin D2 Sequences

Figure 5 shows EM images of the EEs found in Hirt extracts (44) prepared from pre-B cells bearing pBabePuroMyc-ERTm before (Figure 5A) and after (Figure 5B–D) 3 days of 4HT treatment. These initial EM studies repeatedly and reproducibly yielded the unusual images shown here. Their significance is only beginning to be understood, and our tentative interpretations are as follows: Figure 5A shows small, extrachromosomal, DNA-containing, irregular, asymmetrical, circular elements (diameter, <0.10 µm), believed to contain repetitive sequence motifs only, which is characteristic of normal cells (49). Figure 5B to D are electronmicrographs taken at the same magnification as the upper panel, showing three independent examples of the larger (diameter, 0.15–0.35 µm), more discrete circles that are found after 4HT activation of pBabePuroMyc-ERTm-bearing pre-B cells and thought to contain amplified genes.

To confirm the presence of cyclin D2 on EEs in 4HT-treated cells, we developed a method to examine the total population of EEs purified by the Hirt procedure (47). This protocol involved affixing EEs isolated by the method of Hirt (44) to glass slides that were then processed for FISH and counterstained with DAPI. FISH hybridization signals are considered specific only if they colocalize with DAPI-stained EEs. When FISH was performed on the slides (Figure 6B and C), we found that about 10% of the DAPI-stained EEs contained cyclin D2-sequences (indicated by large white arrows). The sizes of the cyclin D2-hybridizing EEs were shown to be 10 to 20 pixels by IPLab software (Scanalytics, Fairfax, VA) analysis of the fluorescence images. The finding that only a fraction of the DAPI-positive spots hybridized to the cyclin D2 probe was evidence for the specificity of this hybridization. FISH studies of Hirt extracts of normal cells showed only a few, extremely tiny DAPI-staining dots (pixel size <5; data not shown), similar to those seen in uninduced pre-B cells (Figure 5A), which did not hybridize with the cyclin D2 probe. These data establish that cyclin D2 hybridizes specifically to a subset of the EEs seen in nuclei, metaphases, and on microscopic preparations of EEs. We conclude from these data that cyclin D2 is one of an unknown number of targets of c-Myc-induced genomic instability found on EEs.

Figure 6
FISH on extrachromosomal elements. Hirt extracts from pBabePuroMyc-ERTm-transduced pre-B cells were affixed to glass slides and stained with DAPI to identify EEs by their DNA content (about 90 dots shown in A). The slide was then hybridized to the cyclin ...

Discussion

A direct role for Myc in cyclin D2 gene amplification in this study was first suspected when a coupling was observed between Myc overexpression and amplification of the cyclin D2 gene in established tumors. This is not unprecedented, because earlier work has demonstrated that Myc can influence replication (50). Amplification of cyclin D2 was first seen in Southern blots of two human cell lines, COLO320HSR and T47D, which were known to have c-Myc amplification and overexpression. Similar evidence of cyclin D2 amplification was also found in mouse plasmacytomas that did not have c-Myc gene amplification but which did have constitutive expression of c-Myc because of chromosomal translocations.

The cyclin D2 amplification that was detected in mouse plasmacytomas was accompanied by enhanced mRNA and protein levels on RNA and protein blots. More transcripts were found in plasmacytomas than in other B-cell lines that did not have c-Myc-activating chromosome translocations. Such increased expression of other members of the G1 cyclins, cyclins D1, D3 and E was not found in plasmacytomas, indicating that this was a special attribute of cyclin D2.

To directly implicate Myc levels in the induction of cyclin D2 amplification, we studied the effects of inducible overexpression of Myc in mouse pre-B cells with a tamoxifen-activated pBabePuroMyc-ERTm chimeric expression vector. Because amplification of genes occurs gradually over successive replication cycles, we did not study the potential short-term effects of c-Myc upregulation of the cyclin D2 gene via transcription activation. Instead, we concentrated on the state of the locus and its expression over several days of 4HT stimulation. 4HT had no effect on the cyclin D2 of parental A-MuLV-transformed pre-B cells. FISH studies showed no evidence of genomic instability, and mRNA expression remained very low. In the cells with an activated Myc-ERTm chimera, EEs, which have also been referred to as double-minutes, polydispersed circular DNA, episomes, and extrachromosomal DNA (49; 51–53), which hybridized with the cyclin D2 probe, appeared after 3 to 4 days, indicating increased genomic instability. At these same time points, blots of RNA and cell lysates isolated from these cells began to show increased expression of cyclin D2 mRNA and protein.

Data obtained to date do not require upregulation of either RNA transcription or changes in RNA stability. It is possible that simple status-quo rates of expression could yield increased steady-state levels of mRNA and protein if the template were increased, such as by the amplification that we have demonstrated. Such a mechanism could also be responsible for the high levels of cyclin D2 mRNA in plasmacytomas, secondary to their constitutive expression of high levels of c-Myc mRNA and protein. It is interesting to note that Southern blots of DNA from pre-B cells did not show increased cyclin D2 hybridization signals (data not shown) after 3 days of 4HT-induction like those seen in well-established tumor cells that have experienced high Myc levels for many generations. This finding is not surprising, because extrachromosomal DNA is generally not visualized in conventional Southern blots (54,55). Moreover, extrachromosomal DNA molecules tend to be unstable, because they usually do not contain centromeres and may be lost during mitosis. With time, some have been shown to integrate into the chromosomes to take the appearance seen in established cell lines (52,56,57).

We do not know yet whether the EEs are the source of the increased cyclin D2 mRNA that appears simultaneously with the appearance of these elements. We have determined, however, that both the number of cyclin D2-containing EEs per cell and the level of cyclin D2 mRNA decreased when 4HT was removed for 4 days from cultures of pre-B cells and fibroblasts that had been stimulated with 4HT for 6 days (data not shown). Preliminary studies have indicated that the EEs contain protein and DNA, because they stain with Giemsa and DAPI and disappear when treated with DNAase. Some preparations of DNA from EEs can be digested with restriction endonucleases, and we are presently optimizing the isolation of EEs with undegraded DNA to attempt its cloning and sequencing. In addition, we will examine whether the EEs are transcribed to yield cyclin D2 mRNA.

A causal connection between Myc levels and cyclin D2 amplification is probably not limited to B lymphocytic tumors, because we saw amplified cyclin D2 in human colorectal and breast carcinomas. In addition, we also found a gradual increase in cyclin D2 expression in mouse fibroblasts when Myc is overexpressed and activated by 4HT treatment of cells that bear the Myc-ERTm expression vector.

We do not know yet why cyclin D2 is amplified when Myc expression is high. We found four CACGTG Myc/Max-binding E-box motifs upstream of exon one of the cyclin D2 gene (to be published in full elsewhere), and we speculate that these E boxes may play a role in targeting such genes for Myc-induced amplification.

This Myc-associated genomic instability may be the cause of the frequent aneuploidy and instability seen in long-term cultures (57) or extensively passaged experimental tumors. More specifically, it has been reported that tumor-specific, nonrandom chromosomal translocations become increasingly difficult to recognize with repeated passages of plasmacytoma lines induced by Myc-activating chromosome translocations, owing to accumulations of additional, presumably random, chromosomal aberrations (58). Although it has been shown that excess Myc activity can elicit overall karyotypic instability (57) and increased tumorgenicity (59), it is important to emphasize that our data show that Myc-associated gene amplification is locus-specific. Extra- and intrachromosomal amplification has been demonstrated previously for Dhfr (26), and in this report for cyclin D2, but we have also determined that high Myc expression produces no such amplification in the genes encoding ornithine decarboxylase, syndecan-2, GAPDH, or cyclin C (26).

Is it possible to construct a hypothetical model for how the genes that are amplified in the presence of Myc overexpression might work together toward neoplasia? We propose that increased Myc activity leads to a redundant expression of genes that promote cell cycle progression and cell proliferation. This effect produces a potent combination favoring induction, promotion, or progression of neoplastic transformation if apoptotic pathways are bypassed. Overexpression of Myc has been shown to shorten the G1 phase of the cell division cycle (19,20), which favors further mutations by curtailing the period available for cells to assess and repair DNA damage before it is duplicated in S phase. Such mutations might allow the cells to escape apoptosis, which is frequently associated with increased Myc activity. A similar effect would be expected from overexpression of cyclin D2, an important G1 cyclin. High levels of such cyclins could also foreshorten G1 and rush cells prematurely into S by titrating out CDK inhibitors such as p21 and p27. Perhaps such changes are responsible for the transformed characteristics that are induced by overexpression of cyclin D1 in fibroblasts (60). Cyclin D1 amplification and overexpression is a well-known step in various cancers (61–63). Amplification and/or overexpression of cyclin D2 may have similar effects. Overexpression of cyclin D2, along with D1 and D3, has been found in mouse skin neoplasms and has been associated with tumor progression (64). Similar to our finding of cyclin D2 gene amplification in COLO320HSR, Leach and colleagues (65) reported that this cyclin gene was amplified in a subgroup of colorectal carcinomas. What is more, inappropriate expression of cyclin D2 also occurs as a result of retroviral integration in retrovirus-induced rodent T-cell lymphomas (66). Finally, the expression of G1 cyclins and their control of the cell division cycle is known to vary between normal and transformed cells (38).

Another gene that is amplified by Myc overexpression is Dhfr. It is a key enzyme of folate metabolism, and it is essential for DNA synthesis. High levels of the product of this gene may contribute to maintenance of cell proliferation. High copy number of Dhfr genes and overexpression of the enzyme, e.g., after amplification, have been correlated with the metastatic potential of tumor cells in a rat carcinoma model (67). Thus we propose that Myc induces a locus-specific instability, and additional steps of selection will determine which cell(s) become malignant clone(s). This makes it possible, and indeed likely, that such cells that survive in this new regulatory setting will accumulate additional genomic alterations and will have an increased potential to complete the multistep process of neoplastic transformation. This concept has recently received support from experiments that demonstrated that c-Myc-induced instability allowed the outgrowth of tumors in athymic mice after their subcutaneous inoculation with fibroblasts that exhibited Myc-mediated instability (59).

In summary, cyclin D2 is one of a growing list of genes targeted for genomic instability by high Myc levels. We are in the process of determining the magnitude (number of genes involved) of the genomic instability induced by Myc overexpression.

Acknowledgements

We thank our colleagues for many valuable discussions, probes, cell lines, and libraries of clones. We thank Ms. E. McMillan-Ward for electron microscopy. This work has been supported by grants to S. M. from the National Science and Engineering Research Council (NSERC), the Manitoba Health and Research Council (MHRC), and the Thorlakson Foundation Fund. The Basel Institute for Immunology was founded and is supported by F. Hoffmann-La Roche, Basel, Switzerland.

Abbreviations

CDK
cyclin-dependent kinase
Dhfr
dihydrofolate reductase
DAPI
4′, 6′diamidino-2-phenylindole
PI
propidium iodide
4HT
4-hydroxytamoxifen
R1
ribonucleotide reductase R1 subunit
R2
ribonucleotide reductase R2 subunit
FISH
fluorescent in situ hybridization
EEs
extrachromosomal elements
A-MuLV
Abelson murine leukemia virus
GAPDH
glyceraldehyde-3-phosphate-dehydrogenase

References

1. Cole MD. The myc oncogene: its role in transformation and differentiation. Annu Rev Genet. 1986;20:361–384. [PubMed]
2. Erisman MD, Scott JK, Watt RA, Astrin SM. The c-myc protein is constitutively expressed at elevated levels in colorectal carcinoma cell lines. Oncogene. 1988;2:367–378. [PubMed]
3. Marcu KB, Bossone SA, Patel AJ. Myc function and regulation. Annu Rev Biochem. 1992;61:809–860. [PubMed]
4. Citri Y, Braun J, Baltimore D. Elevated myc expression and c-myc amplification in spontaneously occurring B lymphoid cell lines. J Exp Med. 1987;165:1188–1194. [PMC free article] [PubMed]
5. Feo S, Liegro CD, Jones T, Read M, Fried M. The DNA region around the c-myc gene and its amplification in human tumour cell lines. Oncogene. 1994;9:955–961. [PubMed]
6. Hayward WS, Neel BG, Astrin SM. Activation of a cellular oncogene by promoter insertion in ALV-induced lymphoid leukosis. Nature. 1981;290:475–480. [PubMed]
7. Morse B, Rotherg PG, South VJ, Spandorfer JM, Astrin SM. Insertional mutagenesis of the myc locus by a LINE-1 sequence in a human breast carcinoma. Nature. 1993;333:87–90. [PubMed]
8. Taub R, Kirsch I, Morton C, Lenoir GM, Swan D, Tronick S, Aaronson S, Leder P. Translocation of the c-myc gene into the immunoglobulin heavy chain locus in human Burkitt lymphoma and murine plasmacytoma cells. Proc Natl Acad Sci USA. 1982;79:7837–7841. [PMC free article] [PubMed]
9. Shen-Ong GLC, Keath EJ, Piccoli SP, Cole MD. Novel myc oncogene RNA from abortive immunoglobulin-gene recombination in mouse plasmacytomas. Cell. 1982;31:443–480. [PubMed]
10. Pear WS, Wahlstrom G, Nelson SF, Axelson H, Szeles A, Wiener F, Bazin H, Klein G, Sumegi J. 6;7 chromosomal translocation in spontaneously arising rat immunocytomas: Evidence for c-myc breakpoint clustering and correlation between isotype expression and the c-myc target. Mol Cell Biol. 1988;8:441–451. [PMC free article] [PubMed]
11. Yokota Y, Tsunetsugu-Yokota Y, Battifora CL, Cline MJ. Alterations in myc, myb, ras Ha proto-oncogenes in cancers are frequent and show clinical correlation. Science. 1986;231:261–265. [PubMed]
12. Jansen-Duerr P, Meichle A, Steiner P, Pagano M, Finke K, Botz J, Wessbecher J, Draetta G, Eilers M. Differential modulation of cyclin gene expression by MYC. Proc Natl Acad Sci USA. 1993;90:3685–3689. [PMC free article] [PubMed]
13. Daksis JI, Lu RY, Facchini LM, Marhin WW, Penn LJZ. Myc induces cyclin D1 expression in the absence of de novo protein synthesis and links mitogen-stimulated signal transduction to the cell cycle. Oncogene. 1994;9:3635–3645. [PubMed]
14. Philipp A, Schneider A, Vaestrik I, Finke K, Xiong Y, Beach D, Alitalo K, Eilers M. Repression of cyclin D1: A novel function of MYC. Mol Cell Biol. 1994;14:4032–4043. [PMC free article] [PubMed]
15. Steiner P, Philipp A, Lukas J, Godden-Kent D, Pagano M, Mittnacht S, Bartek J, Eilers M. Identification of a Myc-dependent step during the formation of active G1 cyclin-cdk complexes. EMBO J. 1995;14:4814–4826. [PMC free article] [PubMed]
16. Waters CM, Littlewood TD, Hancock DC, Moore JP, Evan GI. c-myc protein expression in untransformed fibroblasts. Oncogene. 1991;6:797–805. [PubMed]
17. Mai S, Jalava A. c-Myc binds to 5′ flanking sequence motifs of the dihydrofolate reductase gene in cellular extracts: role in proliferation. Nucleic Acids Res. 1994;22:2264–2273. [PMC free article] [PubMed]
18. Heikkila R, Schwab G, Wickstrom E, Loke SL, Pluznik DH, Watt R, Neckers LM. c-Myc antisense oligodeoxynucleotide inhibits entry into S phase but not progress from G0 to G1. Nature. 1987;328:445–449. [PubMed]
19. Karn J, Watson JV, Lowe AD, Green SM, Vedeckis W. Regulation of cell cycle duration by c-myc levels. Oncogene. 1989;4:773–787. [PubMed]
20. Alexandrow MG, Moses HL. c-myc-enhanced S phase entry in keratinocytes is associated with positive and negative effects on cyclin-dependent kinases. J Cell Biochem. 1998;70:528–542. [PubMed]
21. Hanson KD, Shichiri M, Follansbee MR, Sedivy JM. Effects of c-myc expression on cell cycle progression. Mol Cell Biol. 1994;14:5748–5755. [PMC free article] [PubMed]
22. Mateyak MK, Obaya AJ, Adachi S, Sedivy JM. Phenotypes of c-Myc-deficient rat fibroblasts isolated by targeted homologous recombination. Cell Growth Differ. 1997;8:1039–1048. [PubMed]
23. Davis AC, Wims M, Spotts GD, Hann SR, Bradley A. A null c-myc mutation causes lethality before 10.5 days of gestation in homozygotes and reduced fertility in heterozygous female mice. Genes Dev. 1993;7:671–682. [PubMed]
24. Galaktionov K, Chen X, Beach D. Cdc25 cell-cycle phosphatase as a target of c-myc. Nature. 1996;382:511–517. [PubMed]
25. Mai S. Overexpression of c-myc precedes amplification of the gene encoding dihydrofolate reductase. Gene. 1994;148:253–260. [PubMed]
26. Mai S, Hanley-Hyde J, Fluri M. c-Myc overexpression-associated DHFR gene amplification in hamster, rat, mouse and human cell lines. Oncogene. 1996;12:277–288. [PubMed]
27. Jaffe BM, Eisen HN, Simms ES, Potter M. Myeloma proteins with anti-hapten antibody activity: epsilon-2,4- dinitrophenyl lysine binding by the protein produced by mouse plasmacytoma MOPC-460. J Immunol. 1969;103:872–878. [PubMed]
28. Mushinski JF. c-Myc activation and chromosomal translocation in BALB/c plasmacytomas. In: Klein G, editor. Cellular Oncogene Activation. New York and Basel: Marcel Dekker; 1988. pp. 181–211.
29. Mushinski JF, Davidson WD, Morse HC. Activation of cellular oncogenes in human and mouse leukemia-lymphomas: spontaneous and induced oncogene expression in murine B lymphocytic neoplasms. Cancer Invest. 1987;5:345–368. [PubMed]
30. Rosenberg N, Baltimore D. A quantitative assay for transformation of bone marrow cells by Abelson murine leukemia virus. J Exp Med. 1976;143:1453–1463. [PMC free article] [PubMed]
31. Gurfinkel N, Unger T, Givol D, Mushinski JF. Expression of the bcl-2 gene in mouse B lymphocytic cell lines is differentiation stage specific. Eur J Immunol. 1987;17:567–570. [PubMed]
32. Littlewood TD, Hancock DC, Danielian PS, Parker MG, Evan GI. A modified oestrogen receptor ligand-binding domain as an improved switch for the regulation of heterologous proteins. Nucleic Acids Res. 1995;23:1686–1690. [PMC free article] [PubMed]
33. Mann R, Mulligan RC, Baltimore D. Construction of a retrovirus packaging mutant and its use to produce helper-free defective retrovirus. Cell. 1983;33:153–159. [PubMed]
34. Matsushime H, Roussel MF, Ashmun RA, Sherr CJ. Colony-stimulating factor 1 regulates novel cyclins during the G1 phase of the cell cycle. Cell. 1991;65:701–713. [PubMed]
35. Kiyokawa H, Busquets X, Powell CT, Ngo L, Rifkind RA, Marks PA. Cloning of a D-type cyclin from murine erythroleukemia cells. Proc Natl Acad Sci USA. 1992;89:2444–2447. [PMC free article] [PubMed]
36. Hamel PA, Hanley-Hyde J. G1 cyclins and control of the cell division cycle in normal and transformed cells. Cancer Investigation. 1997;15:143–152. [PubMed]
37. Sinclair AJ, Palmero I, Peters G, Farrell PJ. EBNA-2 and EBNA-LP cooperate to cause G0 to G1 transition during immortalization of resting human B lymphocytes by Epstein-Barr virus. EMBO J. 1994;13:3321–3328. [PMC free article] [PubMed]
38. Thelander L, Berg P. Isolation and characterization of expressible cDNA clones encoding M1 and M2 subunits of mouse ribonucleotide reductase. Mol Cell Biol. 1986;6:3433–3442. [PMC free article] [PubMed]
39. Stanton LW, Watt R, Marcu KB. Translocation, breakage and truncated transcripts of c-myc oncogene in murine plasmacytomas. Nature. 1983;303:401–406. [PubMed]
40. Fort P, Marty L, Piechaczyk M, El Sabrouty S, Dani C, Jeanteur P, Blanchard JM. Various rat adult tissues express only one major mRNA species from the glyceraldehyde-3-phosphate-dehydrogenase multigenic family. Nucleic Acids Res. 1985;13:1431–1437. [PMC free article] [PubMed]
41. Southern EM. Detection of specific sequences among DNA fragments separated by gel electrophoresis. J Biol Chem. 1975;253:5852–5860.
42. Mai S, Hanley-Hyde J, Coleman A, Siwarski D, Huppi K. Amplified extrachromosomal elements containing c-Myc and Pvt 1 in a mouse plasmacytoma. Genome. Genome. 1995;38:780–785. [PubMed]
43. Mischak H, Goodnight J, Kolch W, Martiny-Baron G, Schaechtle C, Kazanietz MG, Blumberg PM, Pierce JH, Mushinski JF. Overexpression of protein kinase C-δ and -ε in NIH 3T3 cells induces opposite effects on growth, morphology, anchorage dependence and tumorigenicity. J Biol Chem. 1993;268:6090–6096. [PubMed]
44. Hirt B. Selective extraction of polyoma DNA from infected mouse cell cultures. J Mol Biol. 1967;26:365–369. [PubMed]
45. Coggins LW. In: Electron Microscopy in Molecular Biology: A Practical Approach. Sommerville J, Scheer U, editors. Oxford: IRL Press at Oxford University Press; 1987.
46. Israels SJ, Gerrard JM. In: Platelets: A Practical Approach. Watson SP, Authi KS, editors. Oxford: IRL Press at Oxford University Press; 1996.
47. Kuschak TI, Paul JT, Wright J, Mushinski JF, Mai S. FISH on purified extrachromosomal DNA molecules. Technical Tips Online. 1999 (in press)
48. Mushinski JF, Bauer SR, Potter M, Reddy EP. Increased expression of myc-related oncogene mRNA characterizes BALB/c plasmacytomas induced pristane or Abelson murine virus. Proc Natl Acad Sci USA. 1983;80:1073–1077. [PMC free article] [PubMed]
49. Gaubatz JW. Extrachromosomal circular DNAs and genomic sequence plasticity in eukaryotic cells. Mutation Res. 1990;237:271–292. [PubMed]
50. Classon M, Henriksson M, Sumegi J, Klein G, Hammaskjold ML. Elevated c-myc expression facilitates the replication of SV40 DNA in lymphoma cells. Nature. 1987;330:272–274. [PubMed]
51. Flores SC, Sunnerhagen P, Moore TK, Gaubatz JW. Characterization of repetitive families in mouse heart polydisperse circular DNAs: age-studies. Nucleic Acids Res. 1988;16:3889–3906. [PMC free article] [PubMed]
52. Wahl GM. The importance of circular DNA in mammalian gene amplification. Cancer Res. 1989;49:1333–1340. [PubMed]
53. Cohen S, Regev A, Lavi S. Small polydispersed circular DNA (spcDNA) in human cells: association with genomic instability. Oncogene. 1997;14:977–985. [PubMed]
54. VanDevanter DR, Piaskowski VD, Casper JT, Douglass EC, Von Hoff DD. Ability of circular extrachromosomal DNA molecules to carry amplified MYCN proto-oncogenes in human neuroblastomas in vivo. J Natl Cancer Institute. 1990;82:1815–1821. [PubMed]
55. Hahn PJ. Molecular biology of double-minute chromosomes. Bioessays. 1993;15:477–484. [PubMed]
56. Kaufman RJ, Brown PC, Schimke RT. Amplified dihydrofolate reductase genes in unstably methotrexate-resistant cells are associated with double minute chromosomes. Proc Natl Acad Sci USA. 1979;76:5669–5673. [PMC free article] [PubMed]
57. Mai S, Fluri M, Siwarski D, Huppi K. Genomic instability in Myc-ERTm-activated Rat1 a Myc-ERTm cells. Chromosome Research. 1996;4:365–371. [PubMed]
58. Coleman AE, Schröck E, Weaver Z, du Manoir S, Yang F, Ferguson-Smith MA, Ried T, Janz S. Previously hidden chromosome aberrations in BALB/c plasmacytomas uncovered by multicolor spectral karyotyping (SKY) Cancer Res. 1997;57:4585–4592. [PubMed]
59. Felsher DW, Bishop JM. Transient excess of MYC activity can elicit genomic instability and tumorigenesis. Proc Natl Acad Sci USA. 1999;96:3940–3944. [PMC free article] [PubMed]
60. Jiang W, Kahn SM, Zhou P, Zhang YJ, Cacace AM, Infante SD, Santella RM, Weinstein IB. Overexpression of cyclin D1 in rat fibroblasts causes abnormalities in growth control, cell cycle progression and gene expression. Oncogene. 1993;8:3447–3457. [PubMed]
61. Motokura T, Arnold A. Cyclin D and oncogenesis. Curr Opin Genet Dev. 3:5–10. [PubMed]
62. Wang TC, Cardiff RD, Zuckerberg L, Lees E, Arnold A, Schmidt EV. Mammary hyperplasia and carcinoma in MMTV-cyclin D1 transgenic mice. Nature. 1994;369:669–671. [PubMed]
63. Zhou P, Jiang W, Zhang Y, Kahn SM, Schieren I, Santella RM, Weinstein IB. Antisense to cyclin D1 inhibits growth and reverses the transformed phenotype of human esophageal cancer cells. Oncogene. 1995;11:571–580. [PubMed]
64. Zhang S-Y, Liu S-C, Goodrow T, Morris R, Klein-Szanto AJP. Increased expression of G1 cyclins and cyclin-dependent kinases during tumor progression of chemically induced mouse skin neoplasms. Mol Carcinog. 1997;18:142–152. [PubMed]
65. Leach FS, Elledge SJ, Sherr CJ, Willson JKV, Markowitz S, Kinzler KW, Vogelstein B. Amplification of cyclin genes in colorectal carcinomas. Cancer Res. 1993;53:1986–1989. [PubMed]
66. Hanna Z, Jankowski M, Tremblay P, Jiang X, Milatovich A, Francke U, Jolicoeur P. The Vin-1 gene, identified by provirus inertional mutagenesis, is the cyclin D2. Oncogene. 1993;8:1661–1666. [PubMed]
67. Luecke-Huhle C. Permissivity for methotrexate-induced DHFR gene amplification correlates with the metastatic potential of rat adenocarcinoma cells. Carcinogenesis. 1994;15:695–700. [PubMed]

Articles from Neoplasia (New York, N.Y.) are provided here courtesy of Neoplasia Press
PubReader format: click here to try

Formats:

Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...

Links

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...