c-Myc-Induced Extrachromosomal Elements Carry Active Chromatin

Abstract

Murine Pre-B lymphocytes with experimentally activated MycER show both chromosomal and extrachromosomal gene amplification. In this report, we have elucidated the size, structure, and functional components of c-Myc-induced extrachromosomal elements (EEs). Scanning electron microscopy revealed that EEs isolated from MycER-activated Pre-B+ cells are an average of 10 times larger than EEs isolated from non-MycER-activated control Pre-B- cells. We demonstrate that these large c-Myc-induced EEs are associated with histone proteins, whereas EEs of non-MycER-activated Pre B- cells are not. Immunohistochemistry and Western blot analyses using pan-histone-specific, histone H3 phosphorylation-specific, and histone H4 acetylation-specific antibodies indicate that a significant proportion of EEs analyzed from MycER-activated cells harbors transcriptionally competent and/or active chromatin. Moreover, these large, c-Myc-induced EEs carry genes. Whereas the total genetic make-up of these c-Myc-induced EEs is unknown, we found that 30.2% of them contain the dihydrofolate reductase (DHFR) gene, whereas cyclin C (CCNC) was absent. In addition, 50% of these c-Myc-activated Pre-B+ EEs incorporated bromodeoxyuridine (BrdU), identifying them as genetic structures that self-propagate. In contrast, EEs isolated from non-Myc-activated cells neither carry the DHFR gene nor incorporate BrdU, suggesting that c-Myc deregulation generates a new class of EEs.

Introduction

c-Myc is an important regulator of growth, proliferation, differentiation, and development. Deregulation of the c-Myc oncoprotein has been reported in apoptosis (for review, see Ref. [2]), cellular transformation (for reviews, see, Refs. [3–5]), and in malignancies of lymphoid and nonlymphoid origin [6–9]. c-Myc is involved in repressing, enhancing, and initiating transcription [5–11]. Recent evidence links c-Myc to angiogenesis [12–14].

The deregulation of c-Myc is a common feature in many tumors, where it is frequently found translocated and/or amplified [15,16]. c-Myc promotes replication [17] and genomic instability [18–25]. This study addresses the mechanisms of c-Myc-dependent genomic instability. Genomic instability refers to genetic changes that affect the normal organization and function of genes and chromosomes. These alterations may be structural or numerical. Factors such as genotoxic stimuli, growth factors, and oncogenic deregulation are potent inducers of genomic instability (for reviews, see, Refs. [26–28]). It was found previously that the deregulation of c-Myc expression under both in vitro and in vivo conditions leads to karyotypic (chromosomal) and locus-specific genomic instability [18–25]. Earlier, we showed c-Myc-dependent amplification of the dihydrofolate reductase (DHFR) gene for hamster and rat fibroblasts, as well as in human breast cancer and mouse plasmacytoma cell lines [20]. DHFR was also amplified in vivo during the first week of c-Myc-dependent induced plasmacytomagenesis in BALB/c mice [29], as well as in fully developed primary mouse plasmacytomas [30]. More recently, we have described the c-Myc-dependent generation of extrachromosomal elements (EEs) in a mouse Pre-B lymphocyte cell line (designated Pre-B) [22,31,32]. To date, this mouse cell line has been shown to carry the following amplified genes extrachromosomally and intrachromosomally upon c-Myc deregulation: DHFR [31], cyclin D2 (CCND2) [21], ribonucleotide reductase R2 (R2) [22], and ornithine decarboxylase (ODC) [33].

EEs have been found in all cells, including primary cells [34]. For immortalized human fibroblasts, EEs range in number from 10 to 60 molecules per cell and average 0.8 µm in diameter [34]. These smaller circular molecules are also called small polydispersed DNA (spcDNA) and carry repetitive DNA sequences [35,36]. In contrast, large (averaging 1–2 µm in diameter) chromatin structures, called episomes and double minutes, which are included in the broad category of EEs [34], have long been known to harbor amplified genes (for review, see Refs. [37–40]). This class of EEs is larger than 250 kb in size and can number in the thousands, depending on which tumor cell line or tumor is analyzed. In 1979, Kaufman et al. [41] first showed the generation of EEs carrying amplified DHFR in unstable methotrexate-resistant mouse cell lines. More recent studies on EEs have shown that these circular double-stranded DNA molecules are prevalent in cancer cells and harbor oncogenes, in addition to drug resistance genes [42–44].

There have been very few studies addressing the mechanism of EE formation and whether or not these molecules function as “active chromatin units.” Studies by Itoh and Shimizu [45] showed that EEs localize to the nuclear interior upon entry of the cell into S-phase. Interestingly, Solovei et al. [46] showed that EEs found in the neuroblastoma cell line HD-N-16 associate with the interchromosomal domain (ICD). It was suggested that this ICD harbors macromolecular complexes that are involved in transcription, splicing, DNA replication, and repair [47,48]. However, compelling data on these cellular functions and their association with EEs have been lacking.

Here, we have analyzed for the first time the size, structure, and functional composition of c-Myc-induced EEs. We report that EEs generated in MycER-activated [1] Pre-B cells carry competent and/or active chromatin. In addition, we show that only c-Myc-induced EEs harbor genes such as DHFR. Moreover, we describe that these c-Myc-induced EEs have the potential to replicate, whereas EEs isolated from non-c-Myc-activated Pre-B cells do not have this ability. The genetic composition of c-Myc-induced EEs is nonrandom; DHFR is present on EEs whereas cyclin C is not. Importantly, EEs from non-c-Myc-induced control cells do not contain these genes.

Materials and Methods

Cell Lines

Pre-B mouse lymphocyte line (designated Pre-B) [21,22] and Pre-B (abl) cells [49] were cultured in RPMI 1640 medium supplemented with 10% (vol/vol) fetal bovine serum, 1% (vol/vol) sodium pyruvate, 50 IU/ml penicillin, 50 µg/ml streptomycin, 5.5x10^-5 M β-mercaptoethanol, and 2 mM glutamine (Canadian Life Technologies, Burlington, Ontario, Canada). The Pre-B cells were seeded for 24 hours, followed by c-Myc activation for 72 hours using 100 µM 4-hydroxytamoxifen (4-HT) (Sigma, Oakville, Ontario, Canada) dissolved in 100% ethanol. The 4-HT-activated cells were designated Pre-B+ and the non-activated Pre-B cells were designated Pre-B-, which in MycER activation experiments received an equal volume of 100% ethanol. MycER activation was determined as described earlier, using quantitative fluorescent immunohistochemistry (fIHC) techniques [22].

fIHC of Purified EEs

fIHC was performed as described previously on Pre-B+ and Pre-B- EEs [31], which were isolated using the Hirt procedure [50]. The purity of the EE preparations was examined using the fluorochrome 4′,6′-diamidino-2-phenylindole (DAPI), which specifically intercalates into AT-rich regions of DNA. Subsequent analysis of EEs from Hirt extracts by fluorescent microscopy identified whether or not there was genomic contamination. Scanning electron microscopy (SEM) was used to confirm these staining results. As a result, EE preparations containing DNA fibers, as detected by either of the two methods, were excluded from our analysis [31]. Only purified EEs devoid of contamination by genomic or apoptotic DNA were used in all of the experimental procedures listed below.

Where indicated, cross-linking of EEs and protein was carried out for 10 minutes at room temperature with 1% formaldehyde in 1x phosphate-buffered saline/50 mM MgCl₂. Sheep anti-core histone antibody (United States Biological, SwampScott, MA) was used at 5 ng per slide and visualized with a FITC-conjugated donkey anti-sheep IgG antibody (Sigma) at 2.75 ng per slide. The anti-histone H3 phosphorylation-specific antibody used is a histone H3 phosphoserine monoclonal antibody from Juan et al. [51]. It was used at 4.0 ng per slide and visualized with a Texas Red conjugated goat anti-mouse IgG₁ antibody (Southern Biotechnology Associates, Birmingham, AL) at 2.5 ng per slide. Image acquisition was performed using a Zeiss Axioplan2 microscope (Carl Zeiss Canada, Toronto, Ontario, Canada) under a 63x oil immersion objective and a UV filter. The images were assessed using Northern Eclipse 5.0 software (Empix Imaging, Mississauga, Ontario, Canada) and a Hamamatsu CCD camera (East Syracuse, NY).

Bromodeoxyuridine (BrdU) Replication and Gene Expression Analysis of EEs

A total of 2.0x10⁶ Pre-B cells was stimulated with 100 µM 4-HT or an equal volume of 100% ethanol and grown for 72 hours, followed by a BrdU pulse (Sigma) at a concentration of 10 µM for 30 minutes before harvesting. The EEs were then isolated using the Hirt extraction procedure [50] as outlined previously [31,53]. The purified EEs were subjected to fluorescent in situ hybridization (FISH) as described previously [31], using the following probes: dihydrofolate reductase (DHFR) [20] and cyclin C [20] conjugated to either digoxigenin (DIG) or biotin, respectively. The DIG-labeled probes were detected using a 1:200 dilution of anti-DIG-FITC anti-body (Roche Diagnostics, Laval, Quebec, Canada), whereas the biotin-labeled probes were detected using a 1:200 dilution of monoclonal mouse anti-biotin antibody (Roche Diagnostics), followed by a 1:400 dilution of goat anti-mouse IgG Texas Red secondary antibody (Southern Biotechnology, Associates, Birmingham, AL). In parallel, BrdU that was incorporated into the EEs was detected as reported previously [32]. Analyses were performed using a Zeiss Axiophot microscope (Carl Zeiss Canada) under a 63x oil immersion objective and UV/FITC and Texas Red filters coupled to a Photometrics CH250/A CCD camera (Eastman Kodak, Rochester, NY). The images were assessed using the IP Lab version 3.1 software (Scanalytics, Fairfax, VA).

Sodium Dodecyl Sulphate Polyacrylamide Gel Electrophoresis (SDS-PAGE) and Silver Staining of SDS-PAGE Gels

Silver staining of 15% SDS-PAGE gels was performed as outlined by the manufacturer of the silver stain oxidizer, silver reagent, and developer solutions (Bio-Rad Laboratories, Hercules, CA). Fifteen percent SDS-PAGE gels were made, loaded, and electrophoresed as outlined by Laemmli [52].

Protein gels for silver stain and Western blot analysis were loaded with 2 µg of low molecular weight standard; either 5 or 10 µg of total histone extract from chicken erythrocytes; 40 µg of ovalbumin and glyceraldehyde-3-phosphate dehydrogenase (GAPDH); 20 µg of bovine serum albumin (BSA) and anti-mouse whole IgG antibody; either 25 µl, 50 µl, 40 µg, or 0.5 µg (based on DNA concentration) of Pre-B+/- EEs; and 0.5 µg (based on DNA concentration) of C57BL6 mouse splenocyte EEs.

Western Blot Analysis of EEs

Formaldehyde fixed or nonfixed Pre-B-/+ and Pre-B (abl)-/+ EEs as well as EEs isolated from primary C57BL6 mouse splenocytes were isolated using the Hirt procedure [50] and subjected to three rounds of dialysis using ddH₂O as the exchange buffer. Dialysis tubing (Canadian Life Technologies) (3/4 in. diameter; 12,000 kDa molecular weight exclusion limit) was used to dialyze a large volume of EEs, followed by their concentration using a Microcon 50 (50,000 kDa cutoff) concentrator (Amicon, Bedford, MA). After centrifuging repeatedly at 13,000 rpm, the EEs were examined for genomic contamination as described above. The concentration of the protein containing EEs was then determined by using either a standard Bradford protein assay [54] using BSA (Sigma) as a concentration standard, or the DNA concentration was determined by using a fluorometric assay. In this assay, the DNA concentration of the EEs was determined by resuspending a purified EEs aliquot in a buffer containing 25 mM sodium phosphate buffer (monobasic) and 0.01 M DAPI, pH 7.5. The plasmid pcDNA3 (Invitrogen, Carlsbad, CA) was used as the concentration standard. The concentration was then determined by using the Titerek Fluoroskan II fluorometer (Labsystems, McLean, VA) at an absorbance/emission spectra specific to DAPI. The measured fluorescence was then analyzed using Spectrosoft software (Labsystems). Depending on the Western blot experiment performed, either a 40-µg quantity of protein containing EEs (as determined by the Bradford assay) or 0.5 µg of EEs (as determined by the fluorometric assay) was then loaded on a 15% SDS-PAGE gel, along with total histone isolated from chicken erythrocytes. Negative controls such as ovalbumin (Sigma ImmunoChemicals, Oakville, Ontario, Canada), GAPDH (Sigma), BSA (Sigma), and anti-mouse IgG (Sigma) were loaded. After electrophoresis of the protein samples for 2 hours at 100 V, protein was transferred to a Hybond-C Super membrane (Amersham Pharmacia Biotechnologies, Piscataway, NJ) using a Bio-Rad (Mississauga, Ontario, Canada) semidry transfer apparatus Transblot SD. To optimize the transfer of highly basic proteins such as histones, a 3-[cyclohexylamino]-1-propanesulfonic acid (CAPS) transfer buffer was utilized (25 mM CAPS, pH 10, 20% methanol), allowing overnight transfer at 50 V (Bio-Rad). Blocking was then performed the next day for 15 minutes using 3% skim milk powder (wt/vol) in 1x/0.2% Tween-20 buffer. For detection of acetylation-specific species of H4 histone, a 1:1000 unit dilution of rabbit penta H4-hyperacetylated specific antibody (Upstate Biotechnology, Hornby, Ontario, Canada) was used for 1-hour incubation under rocking conditions, followed by three 5-minute washes in 1x TBS/0.2% Tween-20 solution. This was followed by detection using a rabbit-specific IgG antibody conjugated to horseradish peroxidase (concentration/time specified by the company; ECL Kit, Amersham Pharmacia Biotechnologies). Visualization of the Western blot was done using Hyperfilm ECL high-performance chemiluminescence film. The film was then developed using the CP1000 developer (AGFA, Etobicoke, Ontario, Canada).

SEM of EEs

The procedure for SEM of EEs was done as outlined previously [55,56], except for the following changes. Suspensions of Hirt-isolated EEs mixed with an equal volume of fresh methanol/acetic acid (3:1 vol/vol) were dropped on cooled cover slips and then permitted to dry. As indicated, some EE samples were treated for 24 hours at 568C with 100 µg (10 µg/µl) of proteinase K before SEM analysis. The coverslip was then plunged into 50% acetic acid solution for 1 second followed by subsequent 1-second incubations in coplin jars filled with a mixture of ethanol:distilled water:methanol:acetic acid (20:20:9:1 parts by volume), 70%, 90%, and 100% ethanol solutions, respectively. Coverslips were then cut with a diamond pencil and mounted on stubs using liquid carbon. The samples were then coated with 15 nm each of Gold in a Hummer VII (Anatech, Alexandria, VA) sputter coater and analysed at 10 kV with a JEOL JSM-6400 scanning electron microscope (Peabody, MA) and photographed on Kodak TMAX black-and-white film.

Results

Ultrastructural Analysis of EEs Isolated from MycERActivated and Control Pre-B Cells

SEM was employed to examine the ultrastructure of c-Myc-induced EEs. A direct comparison of EEs isolated from non-c-Myc deregulated cells (Pre-B-) and of EEs isolated from c-Myc deregulated cells (Pre-B+) was performed. The Pre-B- EEs were 0.1–0.2 µm in diameter (Figure 1, A and B), whereas c-Myc-activated Pre-B+ EEs ranged in size from 1.0 to 20 µm (Figure 1C), with 90% of Pre-B+ EEs being approximately 1.0 µm in diameter. Thus, the EEs isolated from MycER-activated cells were an average of 10-fold larger in diameter than Pre-B- EEs isolated from nonactivated cells, whereas some Pre-B+ EEs were up to 200-fold larger.

Figure 1.

SEM of Pre-B-/+ EEs. Representative examples of EEs isolated from Pre-B- or Pre-B+ cells (A, B, and C, respectively). Arrows in (A) and (B) point to small protein-free EEs (see alsoRef. [34]). Arrows in (C) point to examples of “globular” structures found on the surface of the c-Myc- induced Pre-B+ EEs. The magnification of the structures shown in (A)–(C) is 30,000x, 30,000x, and 3500x, respectively. The scale bar found at the bottom of (A) and (B) represents 1 µm, while the scale bar on the bottom of panel C represents 10 µm.

Our subsequent analyses of c-Myc-induced EEs focused on their protein content, their transcriptional activity, and genetic composition, as well as on their potential to replicate.

Identification of Histone Proteins on c-Myc-Induced Pre-B+ EEs

The protein-like “globular” structures on Pre-B+ EEs were not usually apparent on the Pre-B- EEs (Figure 1, A–C). Previous reports have shown that smaller EEs found in primary cells do not contain proteins on their surface [34]. However, because previous SEM studies have acknowledged similar structures on EEs isolated from tumorigenic cells as possibly being composed of proteins [55,56], we wanted to determine the nature of the “globular” structures found on c-Myc-induced Pre-B+ EEs. Therefore, we pretreated samples of Pre-B+ EEs with proteinase K before SEM analysis. Subsequent analysis showed that the overall structure of the Pre-B+ EEs disintegrated as a result of proteinase K treatment, suggesting that Pre-B+ EEs carry proteins (data not shown).

Histone proteins constitute one of the major components of nuclear chromatin material (for review, see Ref. [57]). Many EEs can replicate autonomously [38–40] and coordinately move with mitotic chromosomes [58], such that they can become accessible to the cellular replication and transcription machinery. Therefore, it is very likely that some of the protein components of EEs are core histone proteins (i.e., H2A, H2B, H3, and H4) (for review, see Ref. [57]). To date, there has been no identification of the protein components found on EEs, except for the identification of lamin B found on EEs that were isolated from COLO 320DM cells [59].

To examine the potential presence of histone proteins on c-Myc-induced EEs (Pre-B+ EEs), we performed fIHC on purified Pre-B+ EEs. Using a pan-histone antibody, which detects the core histone proteins, we identified that 25.0±3.3% of Pre-B+ EEs (Figure 2, A–D, see arrows) contain core histone protein (Figure 2B, green signals, see arrows, and Table 1). To determine whether the Pre-B+ EEs contained active chromatin, we performed fIHC using an anti-histone H3 phosphorylation (H3P)-specific antibody. This antibody detects a modification of histone H3 that is associated with transcriptionally active chromatin [51]. Our data show that 20.0±3.2% of Pre-B+ EEs contained H3P protein (Figure 2C, red signals, see arrows, and Table 1). Eighty percent of the H3P signal overlayed with the core histone signal (Figure 2D, see arrows). When EEs in non-c-Myc-activated control cells (Pre-B- EEs) were analyzed, we found that only 9.3±2.4% of the EEs contain core histone (Table 1). In addition, only 9.0±1.5% of Pre-B- EEs analyzed showed a H3P-specific signal (Table 1). These data suggest that the majority of c-Myc-induced EEs carry transcriptionally competent and/or active chromatin.

Figure 2.

Immunohistochemistry of EEs isolated from nonformaldehyde-treated Pre-B+ cells using both core histone and H3 phosphorylation-specific antibodies. (A) DAPI-stained Pre-B+ EEs. (B) Pre-B+ EEs that were detected with a fluorescein (FITC)-conjugated core histone antibody. (C) Pre-B+ EEs that were detected with a Texas Red (TR)-conjugated H3 phosphorylation-specific antibody. (D) Overlay of panels A–C. The arrows shown in each panel depict representative examples of EEs that are highlighted by either DAPI or histone antibodies. Due to the overlay in (D), each dot appears white in color. This represents a colocalization of signal from panels A to C. Inserts within the individual pictures highlight the respective EEs. For quantification of these findings, see Table 1.

Table 1.

Core Histone and H3-Phosphorylated Protein Found on EEs.

Sample of EEs	4-HT treatment of cells	Formaldehyde treatment of EEs	Percent of EEs with core histone	Percent of EEs with H3P histone
Pre-B	-	-	9.3±2.4	9.0±1.5
	+	-	25.0±3.3	20.0±3.2
	-	+	28.7±2.0	28.0±1.3
	+	+	71.0±10.6	62.0±4.2
Pre-B (abl)	-	-	11.6±5.2	10.7±2.8
	+	-	7.5±2.3	6.8±2.8
	-	+	28.1±2.1	18.4±6.5
	+	+	25.3±2.6	9.0±2.7

This table outlines all the different EE preparations that were used for these studies along with their corresponding core histone and H3 phosphorylation protein levels as measured by fIHC. The percentages and standard errors calculated are based on a total number of 300 EEs that were counted for each sample. The analysis of 100 EEs in each of three individual experiments was performed.

To eliminate the possibility that 4-HT, used to activate the Myc-ER construct in the Pre-B cells (Materials and Methods section), caused the generation of transcription-associated histones on these EEs, we performed fIHC on EEs isolated from Pre-B cells that did not harbor the MycER construct. The cells examined, designated Pre-B (abl), contained, similar to the Pre-B cells studied here [21], the Abelson leukemia virus [49]. A total of 11.6±5.2% of the EEs isolated from non-4-HT-treated Pre-B (abl) cells contained core histone protein and 10.7±2.8% of these EEs show H3P specific signal (Table 1). Similarly, 7.5±2.3% of EEs isolated from 4-HT-treated Pre-B (abl) cells contained core histone protein and 6.8±2.8% of these EEs showed a H3P-specific signal (Table 1). In conclusion, histone proteins were found at low frequencies on Pre-B- EEs and Pre-B (abl)-/+ EEs irrespective of the treatment with 4-HT. In contrast, a significant amount of c-Myc-induced Pre-B+ EEs carried histones. Thus, 4-HT does not contribute towards the generation of EEs, but rather the deregulation of c-Myc induces the generation of EEs with immunohistochemical evidence of active chromatin.

Effects of Formaldehyde Cross-Linking of Pre-B Cells on the Amount of Histone-Carrying EEs Recovered During Hirt Extraction of EEs

To determine the relative proportions of core histone protein and H3P that remain associated with EEs after Hirt isolation [50], we cross-linked EEs and protein prior to the Hirt isolation procedure using formaldehyde (Materials and Methods section). Formaldehyde leads to cross-linking of proteins to DNA [60]. To this end, EE isolation and fIHC analysis were carried out following cross-linking with 1% formaldehyde. As a result, 71.0±10.6% of the Pre-B+ EEs analyzed contained core histone protein (Figure 3B, see arrows, and Table 1). In addition, 62.0±4.2% of the Pre-B+ EEs contained H3P-specific protein (Figure 3C, see arrows, and Table 1) and the H3P signals overlayed with 85% of the core histone signals (Figure 3D, see arrows). In contrast, only 28.7±2.0% of Pre-B- EEs that were isolated from formaldehyde-pretreated Pre-B cells contained core histone protein (Table 1) and 28.0±1.3% of these EEs contained H3P-specific protein (Table 1). Interestingly, Pre-B (abl) EEs isolated from formaldehyde cross-linked 4-HT-treated cells gave similar results to Pre-B- EEs, whereby 25.3±2.6% of EEs contained core histone protein and 9.0±2.7% EEs contained H3P-specific protein. These data suggest that the formaldehyde cross-linking procedure increases the amount of histone protein bound to EEs, providing evidence that the Hirt isolation procedure caused a loss in the number of histone-bound EEs. In addition, using this experimental approach, we have shown that the histone protein content of EEs isolated from either formaldehydetreated Pre-B- cells or Pre-B (abl) cells is not affected by pretreating these cells with 4-HT. Therefore, c-Myc activation specifically affects the number of core histone and H3P histone proteins found on EEs isolated from Pre-B+ cells.

Figure 3.

Immunohistochemistry of EEs isolated from formaldehyde-treated Pre-B+ cells using both core histone and H3 phosphorylation-specific antibodies. The fIHC experiment was performed and analyzed as outlined for Figure 2. (A)–(D) represent similar pictures as in Figure 2, except the EEs used in this experiment were isolated from formaldehyde-treated Pre-B+ cells. Inserts within the individual pictures highlight the respective EEs. For quantification of these findings, see Table 1.

Further Molecular Analyses of Histone Proteins in c-Myc-Induced Pre-B+ EEs

Because the fIHC data suggest that the Pre-B+ EEs contain core histone proteins, we set out to analyze the histone protein content further using Western blot analysis. Because formaldehyde pretreatment of cells increased the amount of core histone protein that was identified by fIHC, all further analyses were performed using EEs isolated from Pre-B cells that had been pretreated with formaldehyde.

Silver stain analysis of four independent samples of Pre-B+ EEs revealed the presence of major protein complexes at approximately 14 kDa, which were absent in four corresponding samples of Pre-B-EEs (Figure 4A, lanes 1–4 and 5–8, respectively, see arrow). This major protein complex migrated at the same molecular weight (14 kDa) as total histone extract isolated from chicken erythrocytes (Figure 4A, lane 9). Silver stain analysis of Pre-B- and Pre-B+ EEs samples also identified many less abundant proteins ranging in molecular weight from less than 14 kDa to greater than 66 kDa.

Figure 4.

Silver stain and Western analysis of Pre-B-/+ and primary spleen cell EEs using H4 acetylation-specific antibody. (A) Fifteen percent silver-stained SDS-PAGE gel. Pre-B+/- EEs (lanes 1–4 and 5–8, respectively), histone extract (lane 9), and low molecular weight standard were loaded as outlined in the Materials and Methods section. (B) Western blot using an anti-rabbit H4 acetylation-specific antibody. Low molecular weight standard, histone extract, ovalbumin, GAPDH, anti-mouse whole IgG antibody (lanes 1–5, respectively), and Pre-B+/- EEs (lanes 6 and 7, respectively) were loaded as outlined in the Materials and Methods section. (C) Western blot of the same membrane used in (B), but it was stripped and reprobed with an anti-mouse whole IgG antibody. (D) Western blot using H4 acetylation-specific antibody. Histone extract and BSA (lanes 1 and 2, respectively), no sample (lanes 3 and 4), Pre-B-EEs, Pre-B+ EEs, and C57BL6 mouse spleen EEs (lanes 5–7, respectively) were loaded as outlined in the Materials and Methods section. (E) Silver-stained 15% SDS-PAGE gel of the Western blot shown in panel D.

We chose to study the 14-kDa proteins on EEs due to their predominance in Pre-B+ cells and their absence in Pre-B- cells. An antibody specific to the modified H4-acetylated histone protein was used for this analysis because it was previously shown that this H4 histone modification is associated with transcriptionally active chromatin [60]. We did not use the histone H3P antibody because it does not recognize histone H3P protein in Western blot analyses. We found that H4-acetylated histone protein was present in the Pre-B+ EE sample (Figure 4B, lane 6), whereas no H4-acetylated protein was detected in Pre-B- EEs (Figure 4B, lane 7). The specificity of the H4 acetylation-specific antibody was confirmed by identifying the IgG heavy chains (50 kDa) and the IgG light chains (25 kDa) of an anti-mouse IgG antibody that was used to reprobe the original blot (Figure 4C).

We next performed Western analyses using EE samples based on their DNA content (Figure 4D). Under these experimental conditions, we observed a faint band for Pre-B- EEs, indicating that Pre-B- EEs contain some H4-acetylated protein (Figure 4D, lane 5). A stronger band was detected for Pre-B+ EEs (Figure 4D, lane 6). To investigate whether the immortalization process of the Pre-B cells might be the reason why some H4-acetylated protein was found in the Pre-B samples [34], we examined EEs isolated from primary mouse C57BL6 splenocytes. No measurable amount of histone H4-acetylated protein was detected in this sample (Figure 4D, lane 7). To control for loading, the gel in Figure 4D was silver-stained (Materials and Methods section) (Figure 4E). Equal loading was observed for Pre-B-, Pre-B+, and C57BL6 EEs, respectively (Figure 4E, lanes 5–7, respectively). We conclude that c-Myc-induced Pre-B+ EEs contain significantly more H4-acetylated protein than Pre-B- EEs.

c-Myc-Induced EEs Carry Genes and Can Replicate

To prove that not only Pre-B+ EEs contain transcriptionally competent and/or active DNA but also genes, we performed FISH of EEs (FISH-EEs) [31] on both Pre-B- and Pre-B+ EEs. To assess their potential to replicate, we pulse-labeled the cells with BrdU prior to the isolation of EEs (Materials and Methods section). The simultaneous analysis of BrdU incorporation signals and DHFR FISH signals allowed us to conclude that: 1) Pre-B+ EEs, but not Pre-B- EEs, incorporated BrdU; 2) Pre-B+ EEs that contained BrdU also contained the DHFR gene, but not the cyclin C control gene; and 3) in contrast, Pre-B- EEs did not incorporate BrdU, and they did not contain the DHFR gene (Table 2). We conclude that c-Myc activation in mouse Pre-B cells leads to the generation of EEs that contain DHFR, which has the ability to replicate. It is likely that other genes are also present on these EEs because previous studies have identified genes such as R2 [22] and cyclin D2 (CCND2) [21] as being found amplified on Pre-B+ EEs. However, only the DHFR gene was analyzed in this study.

Table 2.

BrdU Incorporation and Detection of Genes on EEs.

Samples of EEs	Experiment 1	Experiment 2		Experiment 3
	Percent EEs incorporating BrdU	Percent EEs incorporating BrdU	Percent EEs containing DHFR	Percent EEs incorporating BrdU	Percent EEs containing Cyclin C
Pre-B-	7.6 ±3.2	3.8±1.2	3.4±1.8	7.5±2.6	4.5±1.4
Pre-B+	56.3±6.5	48.0±6.6	30.2±3.3	51.4±9.5	10.2 ±6.0

This table outlines the percentage of Pre-B- and Pre-B+ EEs that were detected for BrdU incorporation as well as the genes DHFR and cyclin C (as determined by FISH-EEs) [31]. The percentages and standard errors calculated are based on a total number of 110 EEs that were counted for each sample in each specific experiment.

Discussion

In 1990, Gaubatz [34] reported that all cells, including primary cells, have a heterogeneous population of EEs carrying repetitive DNA sequence motifs. Extrachromosomal DNA is a common form of DNA amplification that is seen in a variety of different tumors [61]. For example, in acute myelogenous leukemia, myelodysplastic syndrome, and acute lymphoblastic leukemia, generation of EEs is associated with disease progression [62,63]. The significance of EEs in tumorigenicity was shown by studies outlining the elimination of double minute-encoded c-myc genes from the colon cancer cell line COLO 320DM, which reduced the tumorigenicity in mice [64]. Recently, a group of patients with advanced ovarian cancer was treated with a noncytotoxic dose of hydroxyurea. These patients had a 45% to 60% reduction in the number of EEs, and two patients had a stabilized disease for 6 and 10 months, respectively [65]. The role of extrachromosomal c-myc in tumorigenesis was also documented in a translocation-negative mouse plasmacytoma: c-myc was transcribed from c-myc/IgH-containing EEs, suggesting that the functional extrachromosomal juxtaposition of c-myc and IgH replaced the need for a chromosomal c-myc/IgH translocation [66].

Prior to this study, the size, ultrastructure, and protein composition of c-Myc-induced EEs were unknown. Here, we have shown that c-Myc-induced EEs are 10-to 200-fold larger than EEs found in non-MycER-activated Pre-B cells (Figure 1). To date, only two ultrastructural SEM studies on EEs have been performed: one using EEs isolated from COLO 320DM cells and another using EEs isolated from epidermoid KB-V1 carcinoma cells [55,56]. The EEs isolated from these two tumor cell lines have been well characterized, in which the c-myc and multidrug resistance genes have been identified, respectively. The extrachromosomal DNA analyzed from COLO 320DM cells and KB-V1 cells was 1 and 0.5 µm in diameter, respectively. These results are similar to our own SEM analysis shown in Figure 1, in which 90% of our Pre-B+ EEs isolated were 1 µm in diameter.

We show for the first time that c-Myc deregulation leads to the de novo generation of EEs. These EEs are larger than the ones found in non-c-Myc-activated cells [34]. Thus, c-Myc deregulation can also promote genomic instability by generating a pool of extrachromosomal DNA substrates for recombination. Moreover, this study has shown that many of these EEs carry histones and genes and have the ability to self-replicate.

We demonstrated that the percentage of c-Myc-induced EEs in Pre-B+ cells that carried transcriptionally active histones was significantly higher than the percentage found in Pre-B- cells and Pre-B (abl) cells (62.0±4.2%, 28.0±1.3%, and 9.0±2.7%, respectively). However, not all of the Pre-B+ EEs have histone proteins on their surface. This finding illustrates the heterogeneity of EEs isolated from a particular cell type [34]. Western blot analysis further confirmed the potential transcriptional capability of Pre-B+ EEs (Figure 4). These data are similar to studies using the Rat1 fibroblast cell line containing the same MycER construct as the one found in the Pre-B cell line, in which c-Myc activation occurs before H4 acetylation [67]. In addition, it has been shown that the activation of c-Myc in mouse fibroblast 3T3 cells leads to the acetylation of histones H3 and H4 at the cyclin D2 promoter [68]. Therefore, it seems likely that c-Myc is directly involved in preparing the DNA for transcription through the acetylation of histone proteins not only on the chromosomes, but also on EEs.

The Western blot analyses data in Figure 4D indicate that at least some of the non-c-Myc-activated Pre-B- EEs are transcriptionally competent and/or active due to a detectable level of acetylated histone H4 protein. Because the Pre-B cell line is immortalized, we examined EEs directly isolated from mouse C57BL6 primary spleen cells. We did not identify any acetylated histone H4 protein in EEs from these primary C57BL6 cells (Figure 4D, lane 7). This result suggests that cellular immortalization may promote the formation of EEs (see also Ref. [34]), some of which carry H4-acetylated protein as was detected on Pre-B- EEs.

If c-Myc-activated Pre-B+ EEs contain transcriptionally competent and/or active DNA, then it is likely that they contain active genes and that those may be able to replicate. Although not all genes have been identified in Pre-B+ EEs, those identified to date are important for DNA synthesis and cell cycle progression. In fact, we have previously shown that Pre-B+ EEs contain genes such as R2 [22], cyclin D2 [21], and DHFR [20], whereas Pre-B- EEs do not carry these genes. In addition, cell lines such as human breast ductal carcinoma T47D, mouse plasmacytoma MOPC 265, and MOPC 460D and human colorectal carcinoma line 320 HSR have all shown c-Myc-dependent generation of DHFR-containing EEs [20]. The present results suggest a possible correlation between histone-containing EEs and genecontaining EEs. We recently confirmed this correlation by using anti-core histone antibody to isolate core histone-containing EEs, followed by FISH analysis [53]. In the present study, we have identified Myc-induced EEs that carry histones and genes and are able to replicate (Table 2).

Based on our current knowledge, we propose a model of c-Myc-dependent genomic instability (Figure 5), in which c-Myc deregulation can lead not only to genomic instability at the chromosomal level, but also at the extrachromosomal level, with the potential for further genomic instability and neoplasia. The novelty of the present study lies in the fact that the structure of c-Myc-induced EEs is comparable to the structure of chromosomes, and that c-Myc-induced EEs appear to be functional genetic units, which are able to transcribe and replicate the genes they carry and thus able to contribute to cellular transformation. Therefore, by targeting transcriptionally active genes that are found on EEs, it might be possible to reduce or even eliminate the tumorigenic potential of transformed cells in which c-Myc deregulation plays a major role.

Figure 5.

Model of c-Myc-dependent genomic instability and neoplasia. This scheme outlines pathways of c-Myc-dependent genomic instability and neoplasia. c-Myc deregulation initiates genomic instability in the form of intrachromosomal and extrachromosomal amplification (modified from Ref. [30]). At every step in this model, neoplastic transformation can be abrogated if the cells exhibiting genomic instability undergo apoptosis and/or necrosis.

Abbreviations

BSA

bovine serum albumin

BrdU

bromodeoxyuridine

CCND2

cyclin D2

CCNC

cyclin C

DAPI

4′,6′-diamidino-2-phenylindole

DIG

digoxygenin

DHFR

dihydrofolate reductase

EEs

extrachromosomal elements

FISH

fluorescent in situ hybridization

FISH-EEs

fluorescent in situ hybridization of extrachromosomal elements

4-HT

4-hydroxytamoxifen

fIHC

fluorescent immunohistochemistry

GAPDH

glyceraldehyde-3-phosphate dehydrogenase

H3P

phosphorylated histone H3

ICD

interchromosomal domain

kDa

kilodalton

MycER

Myc estrogen receptor construct (1)

ODC

ornithine decarboxylase

Pre-B- cells

non-4-HT-activated murine Pre-B lymphocyte cells

Pre-B + cells

4-HT-activated murine Pre-B lymphocyte cells

Pre-B (abl)

murine Pre-B lymphocyte cells that do not harbor the MycER construct but only the Abelson virus

R2

small subunit of ribonucleotide reductase

SDS-PAGE

sodium dodecyl sulphate polyacrylamide gel electrophoresis

SEM

scanning electron microscopy

spcDNA

small polydispersed DNA

Txred

Texas Red fluorochrome