| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | ADVANCED SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Department of Integrative Life Science, Tokyo Metropolitan Institute of Medical Science, Bunkyo-ku, Tokyo 113-8613, Japan
 |
Abstract |
|---|
 Top Abstract IntroductionResultsDiscussionExperimental proceduresReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionResultsDiscussionExperimental proceduresReferences |
|---|
Belmont's group has carried out pioneering studies demonstrating dynamics of large-scale chromatin using a tandem array of lac operators (lac O), which was visualized using GFP-lac I and its derivatives (Belmont & Straight 1998; Li et al. 1998; Tumbar et al. 1999; Tsukamoto et al. 2000; Ye et al. 2001). A similar investigation utilized a tandem array of MMTV (mouse mammary tumour virus) promoters driving a ras reporter, and this was visualized by GFP-GR (glucocorticoid receptor) (Muller et al. 2001). Although heterogeneity of chromosome structure were observed in these systems as well, little analysis has been carried out regarding the spontaneous fluctuation of chromosome unfolding and of transcriptional activity.
In the course of preparing stable transformants expressing GFP-fused ASK (activator of cdc seven kinase) under the tetracycline inducible promoter (Sato et al. 2003), we have found that stable clones show stochastic induction of GFP-ASK. Intensive investigation has revealed that stochastic expression in this system coincides with stochastic fluctuation in chromatin unfolding of a target gene array with an isogenic background. In our stable transformants, tens of copies of a plasmid vector containing the GFP-ASK gene were integrated into the chromosome in tandem, and this has permitted us to visualize and assess alterations of higher order chromatin structure by FISH. We examined whether the stochastic change of chromosome unfolding is associated with significant fluctuations in gene expression under circumstances in which the relevant transcription factor is expressed constantly.
|
Results |
|---|
|
TopAbstractIntroductionResultsDiscussionExperimental proceduresReferences |
|---|
We previously investigated effects of over-expression of GFP-ASK and Cdc7 on cell cycle progression (Sato et al. 2003). Although over-expression of both huCdc7 and ASK results in elevated phosphorylation of endogenous MCM2 protein, it does not cause any significant effects on cell cycle progression. While preparing stable transformants expressing GFP-ASK under a tetracycline responsive promoter, we found that some clones expressed various levels of GFP-ASK, in spite of being re-cloned by limiting dilution (Fig. 1A). The GFP-ASK expression profile was monitored by GFP fluorescence using a flow cytometer. The number of GFP positive cells began to increase 9 h after the addition of 2 µg/mL doxycycline, and it reached a maximum at around 16 h, which was retained for several hours. As long as cells were re-plated and replenished with new additions of doxycycline, the positive cell population increased gradually. The response of cells to doxycycline became saturated at around 2 µg/mL, and concentrations higher than 2 µg/mL did not make the number of GFP positive cells increase anymore. At the steady state in the presence of 2 µg/mL doxycycline, cell populations with high and undetectable levels of GFP-ASK expression were 43 ± 11% and 45 ± 10% (mean ± SD from 13 experiments), respectively.
|
Heterogeneous expression does not arise from cellular variation of rTetR-VP16 expression level
We questioned whether such heterogeneous GFP-ASK expression in the clonal population arises from various levels of the transcriptional transactivator, rTetR-VP16, which is driven by a constitutive cytomegalovirus promoter in individual cells. Immunofluorescence analysis using anti-VP16 antibody revealed that the expression level of rTetR-VP16 does not vary from cell to cell, regardless of the presence or the absence of doxycycline (Fig. 1C). It indicates that the heterogeneity of GFP-ASK expression seen in this clonal population is not due to different levels of rTetR-VP16 expression. Interestingly, a strong nuclear spot was detected by anti-VP16 only in the cell expressing GFP-ASK in the presence of doxycycline, while rTetR-VP16 localized relatively homogeneously both in the cytoplasm and in the nuclei without doxycycline (Fig. 1C,D). The VP16 nuclear spot seen in the presence of doxycycline was co-localized with the locus of the integrated reporter gene (Fig. 4B: see below). It appears that, if the amount of binding is high, doxycycline dependent binding of rTetR-VP16 to TRE (tetracycline responsive element) is visualized as a nuclear spot.
|
Next we asked if the heterogeneous GFP-ASK expression arises from cell cycle differences. As shown in Fig. 2, GFP-undetectable (S1) and GFP-highly positive (S2) fractions were individually isolated using a cell sorter, and their cell cycle profiles were compared. Prior to sorting, cells were incubated for one hour in the presence of 20 µM BrdU. After sorting, cells were fixed in ethanol and further stained with an FITC-conjugated anti-BrdU antibody. DNA was counter-stained with propidium iodide (PI). As shown in Fig. 2A, the DNA content (upper) and the BrdU incorporation profile (lower) were very similar between S1 and S2 fractions, indicating that there was no difference in cell cycle profile between those two fractions.
|
Expression profile is reversibly altered within one cell cycle
We next investigated the kinetics of turning on and off GFP-ASK expression to establish the steady-state heterogeneity in the presence of doxycycline. As shown in Fig. 3A, each sorted fraction was further cultured to trace its doxycycline-responsiveness. After sorting, cells were cultured in the absence of doxycyline, except for 20 h before re-examination of GFP-ASK expression at day 1 and 3. More than 40% became GFP-ASK positive in the originally negative (S1) fraction within one day. Because the doubling time of these cells is around 22–24 h, the expression conversion occurs within one cycle of cell division. Because, once expressed, GFP signals remain for at least two days even though the transcription is turned off (not shown), the appearance of GFP-undetectable cells in the originally positive (S2) cells was examined three days after sorting. In the S2 fraction, the steady-state heterogeneity was recovered within three days. Thus, the steady-state heterogeneity of the GFP-ASK expression in the isogenic cell population appears to be maintained by repeated activation and inactivation of the integrated gene copies occurring rapidly within a few cell cycles.
|
An expression event is associated with unfolding of the template chromosome
Because chromosome unfolding is thought to be generally responsible for gene activation, we have asked if alteration of chromosome structure is involved in this stochastic expression. To examine the unfolding/condensed state of this locus, the interphase nuclear array of this GFP-ASK reporter gene was visualized by DNA-FISH. To preserve the three-dimensional structure, sorted cells were grown on a poly D-lysine-coated glass bottom microwell dish for one hour before fixation with paraformaldehyde. The FISH signal appears as a single spot in nonexpressing cells without doxycycline (Fig. 1B(3)), suggesting that the integrated array is tightly folded before activation. Although the array is folded in the absence of doxycycline, variation was observed, albeit small, in size of the array. In the presence of doxycycline, the appearance of FISH signal (i.e. degree of chromatin unfolding) was totally different in S1 and S2 fractions (Fig. 4A) left). The array exhibited a small dot-like shape in the S1 fraction as is in cells without doxycycline, and this was drastically extended in the S2 fraction. Arrays are often located next to or in the vicinity of the nucleolus. The degree of chromatin unfolding was estimated by integrating the volume, which was encompassed by the DNA FISH signals (see Experimental procedures and Fig. 6, upper panel). To compare the array size between those two fractions, distribution of the array size of each fraction is shown in a histogram (Fig. 4A(right)). The array size of S1 was small (median: 0.389) whereas that of S2 was large (median: 1.490), suggesting a strong correlation between array size and gene expression activity. In other words, the integrated loci spanning several hundred kbp in total were tightly packed in one population and were extensively unfolded in another. Furthermore, the chromosome folding and unfolding conversion was totally reversible and fluctuated. Thus, stochastic alteration of chromosome unfolding was reflected in stochastic gene expression activities.
|
Asynchronous chromosome conversion of a pair of daughter cells between folding and unfolding
To further confirm the correlation between the array size and transcription activity, we performed RNA-DNA-FISH in a synchronized cell population in the presence of doxycycline. Mitotic cells, collected by the ‘shake off’ method, were incubated on a poly D-lysine-coated glass bottom microwell dish for three hours, until they progressed into early to-mid G1, and were fixed with paraformaldehyde. As shown in Fig. 5, only a trace of primary transcripts was seen at the condensed array, while a large number of transcripts were detected at the unfolded array. Thus, there was a strong correlation between the degree of chromatin unfolding and the transcription activity.
|
|
Discussion |
|---|
|
TopAbstractIntroductionResultsDiscussionExperimental proceduresReferences |
|---|
1 month), implying that the array is not controlled by the strong pressure of heterochromatin formation. It will be interesting to examine whether RNAi mediated gene silencing works at the site of the integrated repeat in our system, however, the observed induction of gene inactivation is not the direct result of the preceding transcription. Rather, activation and/or inactivation occurs stochastically (alternately) due to the plasticity of the chromosome structure. The chromatin of the integrated array goes back and forth between folded and unfolded state. The probability of gene expression in response to doxycycline was about 40% of the total population in the presence of 0.05 mg/mL Hygromycin B. The probability of expression indeed decreases as the Hygromycin B concentration is lowered, although it takes about one week to alter the probability. Although the precise organization has not been analysed, FISH analysis revealed the Hygromycin B resistant gene was inserted inside the array (data not shown). It indicates that the adjacent gene activity may affect, but slowly, the unfolding probability of the array. Importantly, the effect of Hygromycin B is reversible. Even though once the expression probability is lowered by the deprivation of Hygromycin B, it returns to the original expression level by re-addition of Hygromycin B. Thus, although the factors that determine the probability of either folded or unfolded state remains to be clarified, the chromatin environment surrounding the array could be involved in the establishment of equilibrium in the chromosome structure. Such being the case, we have performed all the experiment presented here under the same concentration of Hygromycin B. The main point of this report is that the chromosome of the integrated array unfolds stochastically even under the constant culture conditions. As we observed asynchrony of GFP-ASK expression in the paired daughter cells, we asked if it is caused by the difference of unfolding state between each pair of daughter cells due to random fluctuation of chromosome. By comparing a pair of daughter cells at early to-mid G1 phase, we have shown that the mode of the chromosome-unfolding is not always transmitted into daughter cells in the same manner (Fig. 6). Because global chromosome positions are transmitted through mitosis (Gerlich et al. 2003), it might also be interesting to dissect the relationship between the regulation of higher order chromosome unfolding and positioning. It has been recognized that there is significant variation in the level of gene expression, even in the clonal (isogenic) cell populations. It is generally thought, both in prokaryotes and eukaryotes, that fluctuations of gene activity arise principally from thermodynamic properties of both transcription and translation reactions. In eukaryotes, the modes of chromatin folding could be also responsible for producing variation in the level of gene expression. In the Escheirichia coli system, Elowitz et al. 2002) has shown that both stochasticity inherent in the biochemical process of gene expression (intrinsic noise) and fluctuations in other cellular components (extrinsic noise) contribute substantially to overall variation in gene expression. The sources of extrinsic noise include concentrations, states, and locations of regulatory proteins and polymerases, and fluctuations in the amount or activity of these molecules. They succeeded in discrimination between intrinsic and extrinsic noise under the same intracellular environment; however, the nature of intrinsic noise has not been well defined (Elowitz et al. 2002). Our system has several features in common with other systems set up to study large-scale chromatin unfolding using a tandem gene array (Li et al. 1998; Muller et al. 2001; Tsukamoto et al. 2000; Tumbar et al. 1999; Ye et al. 2001). In all of these systems, there is a strong correlation between chromosome unfolding size and gene expression activity. Muller et al. (2001) have shown that hormone treatment results in large-scale MMTV array chromatin unfolding within three hours, leading to succeeding recondensation later, as visualized by GFP-GR, and that they concluded that polymerase plays a role in producing and maintaining decondensed chromatin. Belmont and colleagues have presented a series of intensive studies on the behaviour of tandem lac O repeats. The chromatin itself, visualized by GFP-lac I, which lacks unfolding activity, alters its conformation in a cell cycle-dependent manner within a length range of 1–3 µm (Li et al. 1998). In this case, although there was little detailed information about cellular heterogeneity, the chromatin conformation did not change through the mid G1 (from 2 h after mitosis) to mid S phase of the cell cycle (Li et al. 1998). In a system developed later, they showed that large-scale chromatin unfolding was caused by binding to chromatin of protein factors that contain unfolding activity, such as VP16 and BRCT (Li et al. 1998; Tumbar et al. 1999). Chromatin unfolding caused by a transcription factor supplied in trans could be also visualized by YFP-lac I (Tsukamoto et al. 2000). VP16-mediated unfolding is associated with chromatin remodeling and histone acetylation activities (Memedula & Belmont 2003). Importantly, they showed that chromatin unfolding itself would not be necessarily accompanied with ongoing transcription activity (Tumbar et al. 1999). Similarly, we found in our system that chromatin unfolding was not blocked by 5,6-dichloro-1-ß-D-ribofuranosylbenzimidazole (DRB) (not shown). Notwithstanding the fact that they admit there is cellular heterogeneity in terms of chromatin unfolding in all these systems described above, they speculate that this heterogeneity is caused by variations in the cellular expression levels of chromatin-binding factors. This is because such factors were introduced transiently or were conditionally expressed in most of their experimental systems. In contrast, in our system using stable transformants, cellular expression levels of rTetR-VP16 are constant (Fig. 1) and its DNA binding can be controlled by the addition of doxycycline. A heterogeneous cell population with different degrees of chromosome folding was produced without cellular variation in the rTetR-VP16 expression level. Furthermore, the fluctuation of chromosome folding is not cell cycle dependent as shown in Fig. 2A. Stochastic conversion of chromosome folding might not be necessarily coupled to a mitotic event (Figs 3 and 6). Indeed, the chromosome folding state is not always identical between a pair of daughter cells as shown in Fig. 6. Thus, chromatin structure fluctuates at the basal steady-state level, regardless of the cell cycle stage, expression level of activator and ongoing transcription, which could be responsible for producing variation in gene expression.
In reporting the advantages of a tandem reporter gene array, we have been able to assess the steady-state alteration of higher order chromosome structure by FISH. This may be the first direct demonstration of stochastic fluctuation of chromosome unfolding. We believe that this fluctuation may be attributed to the nature of in vivo chromosomes.
|
Experimental procedures |
|---|
|
TopAbstractIntroductionResultsDiscussionExperimental proceduresReferences |
|---|
HeLa Tet-on cells were kindly provided by Dr M. Ohtsubo. The plasmid pBI-GFP-ASK + WT CDC7 (Sato et al. 2003) was transfected into HeLa Tet-on cells together with the CAG-IRES-Hygr plasmid using Lipofectamine (Gibco BRL). Transformants were selected in medium containing 1.5 mg/mL Hygromycin B and then maintained thereafter in medium containing 0.05 mg/mL Hygromycin B. Metaphase synchronization was performed by ‘mitotic shake-off’ after a brief nocodazole treatment, as previously described (Sato et al. 2003). Cell sorting was performed using a FACS (Fluorescence Activated Cell Sorter) Vantage flow cytometer (Becton Dickinson).
FISH and immuno-FISH
Cells were grown on a poly D-lysine-coated glass-bottom microwell dish (MatTek, MA, USA). To induce GFP-ASK expression, 2 µg/mL of doxycycline was added for 18–24 h. Cells were washed with PBS and CSK buffer (10 mM PIPES; pH 6.8, 100 mM NaCl, 300 mM sucrose, 3 mM MgCl2, 1 mM EGTA) followed by incubation with CSK buffer containing 0.5% Triton X-100 for 2 min on ice. After washing with CSK buffer, cells were fixed with 3.7% paraformaldehyde in PBS for 10 min. Cells were further permeabilized for 5 min with PBS containing 0.5% Triton X-100 and 0.5% saponin. For DNA-FISH, cells were then washed with PBS and treated with 100 µg/mL RNase A for 30 min and then washed with PBS three times. DNA was denatured by incubating at 90 °C for 10 min in 70% formamide in 2 x SSC. Cells were then dehydrated sequentially in 70%, 90% and 100% ethanol for 5 min each. Denatured nick-translated plasmid probe was mixed with 3 µg of human Cot-1 DNA and 10 µg of salmon sperm DNA at 80 °C for 10 min. This was then mixed with hybridization buffer (2 x SSC, 50% formamide, 10% dextran sulphate, 4% BSA (Bovine serum albumin)), and then spotted on cells. Hybridization was carried out overnight at 37 °C in a humidified chamber. After three washes with 50% formamide in 2 x SSC and one wash with 1 x SSC for 5 min at 42 °C, the samples were incubated for 30 min in buffer (4% BSA, 0.1% Tween-20 in 4 x SSC) containing either FITC conjugated anti-digoxigenin or avidin-rhodamine, or both. Then samples were washed three times with 4 x SSC, 0.1% Tween-20 for 5 min. Finally, mounting solution (Vectashield, Vector) including DAPI (4,6-diamidino-2-phenylindole) was added and the microwells were sealed with nail polish. For RNA-FISH, RNase treatment and denaturation processes were omitted before hybridization as described (Kagotani et al. 2002; Nutt et al. 1999).
For anti-VP16 staining, cells were washed with PBS twice, fixed with 3.7% paraformaldehyde in PBS for 10 min and permeabilized for 5 min with PBS containing 0.5% Triton X-100 and 0.5% saponin. After washing with PBS twice, the samples were incubated with blocking buffer (4% BSA, 0.1% Tween 20 in PBS) for 30 min. Then the samples were incubated with blocking buffer containing rabbit anti-VP16 antibody (BD Bioscience 3844–1, USA) for 2 h and washed three times with blocking buffer. The samples were incubated with rhodamine conjugated goat anti-rabbit antibody for 30 min and washed three times with blocking buffer and once with PBS. Before proceeding to DNA-FISH, the samples were fixed with 3.7% paraformaldehyde in PBS for 30 min and treated with RNase A for 30 min. DNA-FISH was further performed as mentioned above.
Southern blot hybridization
Genomic DNA was extracted using a Wizard genomic DNA purification kit (Promega) and digested with restriction enzymes. After agarose gel electrophoresis, DNA was transferred on to Hybond-N (Amersham). Plasmid DNA pBI-GFP-ASK +WT CDC7, labelled with digoxigenin, was used as a probe. Hybridization was performed using QuickHib (Stratagene). Detection was performed using a DIG luminescent detection kit (Roche).
Image analysis
3D-images were collected as series of optical sections using the Resolve 3D data collection program (Applied Precision Inc.) with a fluorescence microscope (Olympus IX70 with a 100x, 1.4 NA Plan Apo oil immersion lens). Optical sections were collected through entire nuclei at 0.2-µm focal intervals; the pixel size was 0.066 µm, and 512 x 512 pixel images were taken. The optical sections were deconvolved using an iterative constrained deconvolution algorithm. Regions defining the area encompassed by the DNA- or RNA-FISH signal were identified by thouresholding. For an assessment of chromosome array size, the volume encompassed by the DNA-FISH signal was integrated by multiplying the total pixel number by the area size of each pixel and the focal interval length. For an assessment of primary transcript expression levels, pixel intensities were summed to yield the total intensity of the RNA-FISH signal.
Cell cycle analysis
Cells were incubated in the presence of 20 µM BrdU for one hour prior to harvest. Cells were fixed with 70% ethanol, incubated with FITC-conjugated anti-BrdU antibody and counter stained with propidium iodide (PI) as described by the manufacturer's instruction (Pharmingen). Fluorescence intensity was measured using a FACScan flow cytometer (Becton Dickinson).
|
Acknowledgements |
|---|
|
Footnotes |
|---|
* Correspondence: E-mail: nrksato{at}rinshoken.or.jp
|
References |
|---|
|
TopAbstractIntroductionResultsDiscussionExperimental proceduresReferences |
|---|
Elowitz, M.B., Levine, A.J., Siggia, E.D. & Swain, P.S. (2002) Stochastic gene expression in a single cell. Science
297, 1183–1186.
Gerlich, D., Beaudouin, J., Kalbfuss, B., Daigle, N., Eils, R. & Ellenberg, J. (2003) Global chromosome positions are transmitted through mitosis in mammalian cells. Cell 112, 751–764.[CrossRef][Medline]
Gottschling, D.E., Aparicio, O.M., Billington, B.L. & Zakian, V.A. (1990) Position effect at S. cerevisiae telomeres: reversible repression of Pol II transcription. Cell 63, 751–762.[CrossRef][Medline]
Kagotani, K., Nabeshima, H., Kohda, A., Nakao, M., Taguchi, H. & Okumura, K. (2002) Visualization of transcription–dependent association of imprinted genes with the nuclear matrix. Exp. Cell Res. 274, 189–196.[CrossRef][Medline]
Li, G., Sudlow, G. & Belmont, A.S. (1998) Interphase cell cycle dynamics of a late-replicating, heterochromatic homogeneously staining region: precise choreography of condensation/decondensation and nuclear positioning. J. Cell Biol.
140, 975–989.
Lundgren, M., Chow, C.M., Sabbattini, P., Georgiou, A., Minaee, S. & Dillon, N. (2000) Transcription factor dosage affects changes in higher order chromatin structure associated with activation of a heterochromatic gene. Cell 103, 733–743.[CrossRef][Medline]
Memedula, S. & Belmont, A.S. (2003) Sequential Recruitment of HAT and SWI/SNF Components to Condensed Chromatin by VP16. Curr. Biol. 13, 241–246.[CrossRef][Medline]
Muller, W.G., Walker, D., Hager, G.L. & McNally, J.G. (2001) Large-scale chromatin decondensation and recondensation regulated by transcription from a natural promoter. J. Cell Biol.
154, 33–48.
Nutt, S.L., Vambrie, S., Steinlein, P., et al. (1999) Independent regulation of the two Pax5 alleles during B-cell development. Nature Genet. 21, 390–395.[CrossRef][Medline]
Parra, I. & Windle, B. (1993) High resolution visual mapping of stretched DNA by fluorescent hybridization. Nature Genet. 5, 17–21.[CrossRef][Medline]
Sato, N., Sato, M., Nakayama, M., Saitoh, R., Arai, K. & Masai, H. (2003) Cell cycle regulation of chromatin binding and nuclear localization of human Cdc7-ASK kinase complex. Genes Cells 8, 451–463.[Abstract]
Tsukamoto, T., Hashiguchi, N., Janicki, S.M., Tumbar, T., Belmont, A.S. & Spector, D.L. (2000) Visualization of gene activity in living cells. Nature Cell Biol. 2, 871–878.[CrossRef][Medline]
Tumbar, T., Sudlow, G. & Belmont, A.S. (1999) Large-scale chromatin unfolding and remodeling induced by VP16 acidic activation domain. J. Cell Biol.
145, 1341–1354.
Ye, Q., Hu, Y.F., Zhong, H., Nye, A.C., Belmont, A.S. & Li, R. (2001) BRCA1-induced large-scale chromatin unfolding and allele-specific effects of cancer-predisposing mutations. J. Cell Biol.
155, 911–921.
Received: 27 February 2004
Accepted: 30 April 2004
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | ADVANCED SEARCH | TABLE OF CONTENTS |
