INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

All eukaryotic cells have three distinct RNA polymerases that normally transcribe different sets of nuclear genes. RNA polymerase I (Pol I) is unique in that in most eukaryotic organisms, its sole function is the transcription of genes for large rRNAs (rDNA). Dedication of a separate RNA polymerase to rDNA transcription is a feature unique to eukaryotes and must have had strong selective advantages for eukaryotic organisms in evolution. However, we have recently discovered that mutants of the yeast Saccharomyces cerevisiae which are defective in transcription factor UAF (upstream activation factor) give rise to variants which now grow by transcribing endogenous rDNA by RNA polymerase II (Pol II) (39). (In this paper, we use the term transcription of rDNA to imply transcription of the gene encoding the 35S precursor rRNA, although the 5S RNA gene transcribed by RNA polymerase III is a part of the rDNA repeat unit in S. cerevisiae.) Thus, yeast cells have an inherent ability to use Pol II for rDNA transcription, but this transcription activity is apparently silenced in normal cells. Studies of the processes which enable yeast cells to grow without using the Pol I machinery may be important for understanding the normal Pol I machinery and for gaining insight into the significance of its evolution. In this paper, we describe our finding that the switch to growth using the Pol II system consists of two steps; the first step is a mutational alteration in UAF, and the second step is an expansion of chromosomal rDNA repeats. The first step, a UAF mutation, is sufficient to allow Pol II transcription of rDNA, but the overall efficiency of rRNA synthesis is apparently not sufficient for sustained cell growth. The second step, rDNA repeat expansion, represents an adaptation process, which probably involves a selection for faster-growing cells and leads eventually to a semistable state of rDNA with an increase in repeat numbers and an altered nucleolar location and morphology.

Like other eukaryotic rDNA promoters, the promoter for the gene encoding 35S precursor rRNA in S. cerevisiae consists of two elements, the upstream element and the core element. Basal transcription of yeast rDNA requires the core element and two transcription factors, Rrn3p and CF (core factor), in addition to Pol I. CF consists of three proteins encoded by RRN6, RRN7, and RRN11. For high levels of transcription, two additional factors, UAF and TBP (TATA binding protein), as well as the upstream element are required in addition to the components required for basal transcription (18, 19, 37). UAF contains three Pol I-specific protein subunits encoded by RRN5, RRN9, and RRN10, histones H3 and H4, and the uncharacterized protein P30 (17). In apparent agreement with the in vitro function of UAF, genes RRN5, RRN9, and RRN10 are not absolutely required for cell growth and yeast strains with mutations in these genes can grow, albeit very slowly (19).

Slowly growing UAF mutants give rise to faster-growing variants which do not require intact Pol I and synthesize rRNA using Pol II (39). The slowly growing mutants defective in Pol I-specific UAF components are unstable because of the appearance of faster-growing variants, but they can be maintained stably by introducing a helper plasmid, e.g., pNOY103, which carries the 35S rRNA coding region fused to the galactose-inducible GAL7 promoter. These cells can grow fairly well on galactose but extremely poorly on glucose due to repression of the fusion gene. The faster-growing variants can grow, with or without a helper plasmid, both on galactose and on glucose and were previously shown to synthesize rRNA by transcribing endogenous rDNA by Pol II (39). These cells were called PSW (polymerase switched for growth) cells, thus defining the PSW state. The original UAF mutant cells carrying the helper plasmid, which were unable to grow on glucose, were called non-PSW (referred to as N-PSW in this paper), thus defining the N-PSW state (39). The PSW state, once established, is fairly stable and can be inherited through mitosis and meiosis, but spontaneous reversion to the original N-PSW state can be easily demonstrated (39). We have now studied the reversible changes between the N-PSW and PSW states and discovered that they represent expansion and contraction of rDNA repeats.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Strains, plasmids, and media. Yeast strains and plasmids used are listed in Table 1. Disruption of RPA135 was done as previously described (40). For disruption of SIR2, a DNA fragment covering the SIR2 chromosome region, containing a 1.8-kb deletion removing most of the SIR2 coding region except for the first 18 amino acids and replaced by a 1.6-kb fragment containing LEU2, was used. For disruption of SIR3, a DNA fragment covering the SIR3 chromosome region, containing a 2.5-kb BglII-XhoI deletion within the SIR3 coding region replaced by the 2.5-kb BglII-SalI fragment containing LEU2, was used. For disruption of SIR4, a DNA fragment covering the SIR4 chromosome region, containing a 330-bp BglII-BamHI deletion inactivating SIR4 function (15) and replaced by the 3.1-kb BglII-BglII fragment containing LEU2, was used. Disruption of SIR2, SIR3, and SIR4 genes was confirmed by the nonmating phenotype of the strains constructed. Disruption of FOB1 was carried out by using the EcoRI fragment obtained from pUC-fob1::LEU2 as previously described (20). Disruption of RRN5 was done by using the 3.4-kb fragment carrying rrn5::LEU2 described previously (19).

Spot test for PSW and N-PSW phenotype. Individual colonies formed on YEP-galactose media were picked and suspended in 100 µl of H₂O, and 5-µl aliquots of 10-fold serial dilutions were spotted on YEP-galactose and YEPD plates. They were usually incubated at 30°C for 7 days.

Analysis of chromosome XII by contour-clamped homogeneous electric field (CHEF) electrophoresis. Chromosomal DNA from yeast strains was isolated as previously described (35) and electrophoresed in 0.8% agarose (SeaKem LE; FMC BioProducts, Rockland, Maine)-0.5× Tris-borate-EDTA buffer by using a CHEF Mapper (Bio-Rad, Richmond, Calif.), programmed for a switch time of 300 to 900 s for 68 h at 14°C and with an included angle of 120° (21). After electrophoresis, the gel was stained with 0.5 mg of ethidium bromide per ml for 30 min at room temperature, destained in water for 1 to 2 h, and then photographed. The gel was transferred to a nylon membrane (Zeta-Probe GT; Bio-Rad) and then analyzed by Southern hybridization with ³²P-labeled rDNA and SIR3 probes (24). The rDNA probe used was the 613-bp SmaI-EcoRV fragment spanning positions 210 to +403 (+1 is the Pol I transcription start site). The SIR3 probe used was the 754-bp PstI-HindIII fragment within the SIR3 coding region.

EM and FISH. Yeast strains were grown at 25°C, and electron microscopy (EM) analysis was carried out as previously described, using a JEOL 100CX electron microscope (28). Fluorescence in situ hybridization (FISH) analysis was also carried out as described previously (28) except that the rDNA probe used was a 6.9-kb DNA carrying the 35S rRNA coding region (+1 to +6922) with extra 15 nucleotides derived from a multicloning site of a plasmid vector.

Other methods. Isolation of DNA and Southern hybridization analysis were carried out as described by Maniatis et al. (24). Analysis of 5' ends of precursor rRNA by primer extension was carried out as described previously (18, 39), using the primer 5'-ACACGCTGTATAGAGACTAGGC-3', which hybridizes to 35S precursor rRNA 130 nucleotides downstream of the Pol I start site. Quantification in both analyses was done with a PhosphorImager (Molecular Dynamics, Sunnyvale, Calif.).

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Transition from the N-PSW state to the fully established PSW state is an adaptation process. Yeast strains carrying a deletion of RRN9 were analyzed for growth by spot testing serial dilutions on galactose and glucose plates. N-PSW strain NOY703 is able to grow well on galactose due to the presence of a helper plasmid carrying GAL7-35S rDNA (Fig. 1A Gal, rows a and b). Conversely, NOY703 grows poorly on glucose with colonies appearing after a long incubation (7 to 10 days at 30°C) at a frequency of approximately 10³ to 10⁴ (Fig. 1A Glu, rows a and b). When colonies were picked from the glucose plate, resuspended in water, and spot tested again on glucose and galactose plates, growth on galactose was still better than that on glucose as indicated by colony size, and only a small fraction (ca. 1% in the example in Fig. 1A) of the cells retained the ability to grow on glucose (Fig. 1A, rows c and d). However, after repeated streaking on glucose plates, larger colonies were easily obtained, and these clones showed better growth on glucose than on galactose. Spot test analysis of one such established PSW strain, NOY878, which was derived from NOY703 and was used in the present and previous work (39), is shown in Fig. 1B (rows c and d). These observations demonstrate that between the N-PSW and PSW states are intermediates which are able to grow weakly on glucose and moreover are unstable such that they tend to lose the ability to form colonies on glucose. Thus, switching from N-PSW to PSW states is an adaptation process that occurs through selection for better-growing variants on glucose plates.

View larger version (62K): [in this window] [in a new window]   FIG. 1.   Spot test of rrn9 N-PSW strain NOY703 and PSW variants derived from it. (A) Two independent colonies of NOY703 formed on a galactose plate were analyzed by spotting aliquots of 10-fold serial dilutions of suspension of colonies on YEP-galactose (Gal) and YEPD (Glu) (rows a and b). Several discrete colonies, which had just formed on YEPD by plating large numbers of NOY703 cells on glucose (similar to colonies shown in rows a and b), were combined and similarly analyzed (rows c and d). (B) Two independent colonies of NOY703 (rows a and b), NOY878 (rows c and d), and an N-PSW revertant derived from NOY878 (rows e and f) formed on galactose plates were analyzed by the spot test as for panel A. Plates were incubated at 30°C for 10 days.

Correlation of expansion of rDNA repeats with the switch from N-PSW to PSW states. In the course of analysis of rDNA in PSW strains, we discovered that the amount of rDNA in these strains is much higher than that in N-PSW strains. The results of a typical Southern analysis of rDNA in the wild-type (WT; NOY556), rrn9 N-PSW (NOY703), and rrn9 PSW (NOY878) strains are shown in Fig. 2 (lanes 1 to 3). SIR3, a single-copy gene on chromosome XII, the same chromosome that carries the rDNA repeats, was used for normalization. It was found that the rrn9 N-PSW strain showed about a 2-fold reduction and the rrn9 PSW strain showed about a 2.6-fold increase in rDNA copy number relative to the WT strain. The reduction in the copy number of rDNA repeats in the rrn9 N-PSW strain was not unexpected, since a twofold reduction in rDNA copy number was previously observed for a strain (NOY408-1a) carrying a deletion in an essential Pol I subunit gene (RPA135) and growing on galactose using the GAL7-35S rDNA fusion gene on a helper plasmid (21). This strain was also analyzed, together with its control RPA135 strain (NOY408-1b), and the previous finding, a twofold reduction in rDNA copy number, was confirmed (Fig. 2, lanes 4 and 5). The absolute copy number of rDNA in this control strain was previously estimated to be approximately 150 per genome. Using this number for the current control strain (NOY556), one can calculate that the rrn9 N-PSW strain has approximately 80 copies and the rrn9 PSW strain has approximately 400 copies. Thus, the switch from the N-PSW to the PSW state is accompanied by an approximately fivefold increase in rDNA copy number.

View larger version (71K): [in this window] [in a new window]   FIG. 2.   Comparison of chromosomal rDNA copy numbers of NOY556 (WT), NOY703 (9,N-PSW), NOY878 (9,PSW), NOY408-1a (A135), and NOY408-1b (WT). DNA from these five strains were digested with HindIII and PstI, and the digests were subjected to agarose gel electrophoresis, followed by transfer to a nylon membrane and hybridization with a mixture of 32P-labeled rDNA probe and SIR3 probe. An autoradiogram is shown with an inserted gap between lanes 3 and 4. Radioactivity in each DNA band was quantified with a PhosphorImager. The values for chromosomal rDNA (and plasmid rDNA) were first normalized to the values for reference SIR3 DNA. These values obtained for NOY703 (lane 2) and NOY878 (lane 3) were then divided by corresponding values for control strain NOY556 (lane 1), and the ratios calculated are shown below the pertinent bands. Similarly, the normalized values obtained for NOY408-1a (lane 4) were divided by corresponding values for control strain NOY408-1b (lane 5), and the ratios calculated are shown below the pertinent bands. It should be noted that growth of both NOY703 (lane 2) and NOY408-1a (lane 4) is achieved by transcription of GAL7-35S rDNA on helper plasmids (NOY103 and NOY102, respectively), and copy numbers of the helper plasmids were higher than those carried by the control strains due almost certainly to selection.

View larger version (58K): [in this window] [in a new window]   FIG. 3.   Correlation between the sizes of chromosome XII as analyzed by CHEF electrophoresis and PSW/N-PSW phenotypes. (A) Chromosomal DNA was isolated from strains NOY505 (lane 1) and NOY703 (lane 2), PSW strain NOY878 (lane 3), an N-PSW strain derived spontaneously from NOY878 (lane 4), and strain NOY408-1a (lane 5). The size of chromosome XII was then analyzed by CHEF electrophoresis. Size markers (lane M) are Hansenula wingei chromosomes, and their sizes are indicated in megabase pairs. Left, chromosome patterns revealed by staining with ethidium bromide; middle and right, autoradiograms obtained after hybridization with a SIR3 probe and an rDNA probe, respectively. (B) Chromosomal DNA was isolated from the following strains and analyzed by CHEF electrophoresis as for panel A: NOY505 (lane 1), NOY703 (lane 2), NOY877 (lane 3) (NOY703 and NOY877 are parents [P] of diploids [D] shown in lanes 5 to 7), a diploid obtained after the cross of N-PSW strain NOY876 and N-PSW strain NOY769 (lane 4), three independent diploid clones obtained after the cross of N-PSW strain NOY703 and PSW strain NOY877 (lanes 5 to 7), a diploid strain obtained after the cross of two PSW strains NOY877 and NOY878 (lane 8), and two haploid segregants (DH) from the cross of N-PSW strain NOY703 and PSW strain NOY877, one showing the N-PSW phenotype (lane 9) and the other showing the PSW phenotype (lane 10). Autoradiograms obtained after hybridization with a SIR3 probe and an rDNA probe, respectively, are shown. It should be noted that in panel B, a portion of the gel containing the initial sample plugs corresponding to lanes 1 to 5 was inadvertently lost before the gel was subjected to autoradiography. Therefore, radioactive signals in sample wells representing chromosome XII, which was present in incompletely digested cells or spheroplasts and failed to enter the gel, are not seen in these lanes.

Two rDNA states with different rDNA repeat numbers are inherited through meiosis. We have previously shown that the PSW state can be inherited not only through mitosis but also through meiosis (39). In crosses between PSW and N-PSW strains, both PSW and N-PSW haploid segregant clones were obtained after growth as diploids, followed by sporulation and tetrad dissection. Although an exact 2:2 segregation pattern was not always observed because of an apparently increased frequency of switching in the PSW/N-PSW diploid state and perhaps during meiosis, a clear segregation of the phenotypes indicated their association with a chromosome, presumably chromosome XII (39). We have now examined the state of chromosome XII in such diploids as well as haploid segregants in such a cross. As shown in Fig. 3B, independent diploid clones obtained from the cross showed a copy of chromosome XII with expanded rDNA repeats and a copy with decreased rDNA repeats (lanes 5 to 7). Such diploid clones, analyzed after the cross without extensive colony purification, almost always showed the PSW phenotype (39). Thus, the copy of chromosome XII with the expanded rDNA repeats derived from the PSW parent strain apparently functions as a dominant allele. The two chromosome XII species with different rDNA repeats can coexist within the same diploid nucleus and can be segregated into individual haploid spores through meiosis, as confirmed by CHEF analysis of clones of haploid segregants; segregant clones with the PSW phenotype showed a chromosome with expanded rDNA repeats and those with the N-PSW phenotype showed a chromosome with reduced rDNA repeats (Fig. 3B, lanes 9 and 10). Thus, the expanded state of rDNA on chromosome XII is stable enough to be maintained through meiosis. Figure 3B also includes the analysis of two control diploids, one obtained by a cross between two N-PSW strains and the other obtained by a cross between two PSW strains. The former showed the chromosome XII with decreased rDNA repeats (lane 4), and the latter showed the chromosome with expanded rDNA repeats (lane 8). These results give further support to the correlation between the expanded rDNA state on chromosome XII and the PSW phenotype. Interestingly, the shorter chromosome XII derived from the rrn9 N-PSW parent in the PSW/N-PSW diploids showed a band with a size distribution more homogeneous than that of the N-PSW haploid parent or the N-PSW diploid (Fig. 3B; compare lanes 5 to 7 with lanes 2 and 4) or the N-PSW haploid segregants (lane 9), and their sizes varied depending on the diploid clone examined (lanes 5 to 7). A possible reason for this observation is discussed below.

Mutation of FOB1 decreases the frequency of switching from the N-PSW state to the PSW state. The observed correlation of expansion of rDNA repeats with the PSW phenotype suggests that rDNA repeat expansion is necessary to attain the PSW state. Alternatively, the expansion of rDNA could simply be a consequence of PSW. To distinguish between these two alternatives, we examined the effects of deletion of the FOB1 gene on the switching from N-PSW to PSW states. The FOB1 gene was previously demonstrated to be required for expansion/contraction of rDNA repeats (21). It was found that strain NOY408-1a, which carries the rpa135 mutation (and helper plasmid pNOY102), showed a reduction in the number of rDNA repeats to about 80 repeats on average (Fig. 3A, lane 5) and that the introduction of the missing RPA135 gene induced a gradual increase in repeat number back to the normal level, about 150. Derivatives of the rpa135 strain carrying a fob1 deletion were constructed; these showed a reduced rDNA repeat number with more homogeneous repeat number distribution. These fob1 rpa135 strains did not show a significant increase in repeat numbers upon reintroduction of the RPA135 gene, demonstrating a requirement of FOB1 for rDNA expansion (21). It should be noted that DNA replication fork blocking aided by Fob1 protein at the site distal to the 35S rRNA coding region (20) appears to stimulate recombination, leading to a stimulation of rDNA expansion/contraction. Thus, rDNA expansion and contraction take place almost certainly during DNA replication and may require many generations of cell growth to attain large changes in repeat numbers.

4

3

5

View larger version (62K): [in this window] [in a new window]   FIG. 4.   Effect of a fob1 mutation on the frequency of switch from the N-PSW to PSW states. (A) Chromosomal DNA was isolated from the following strains and analyzed by CHEF electrophoresis: NOY505 (lane 1), NOY703 (lane 2), four independent fob1 deletion isolates obtained from NOY703 by disruption of FOB1 (lanes 3 to 6), and a fob1 deletion isolate obtained from the control strain NOY505 (lane 7). The left and right panels show autoradiograms obtained after hybridization with a SIR3 probe and a rDNA probe, respectively. (B) Three independent colonies of NOY703 (9, N-PSW) and those of a fob1 mutant derived from NOY703 were analyzed by spot test on YEP-galactose (Gal) and YEPD (Glu).

rrn9 fob1

rrn9 fob1

rrn9 FOB1

Effects of mutations in SIR genes and genes for Pol I and CF subunits. We have previously shown that in addition to UAF, both Pol I and transcription factor CF are important in the maintenance of normal yeast nucleolar structures (27-29). In addition, there are some proteins known to be involved in the maintenance of rDNA chromatin structure. For example, Sir2 protein is present in the yeast nucleolus in addition to being present at telomere regions (11). Sir2p is known to be required for silencing of some Pol II reporter genes inserted into rDNA (1, 10, 36) and for decreasing the rate of recombination within the rDNA repeats (12). Therefore, we examined mutations in some of these genes with respect to switching to the PSW state.

rrn9 rpa135

rrn9 rpa135

View larger version (56K): [in this window] [in a new window]   FIG. 5.   Efficiency of switching from the N-PSW to PSW states. (A) Two independent colonies of strains carrying rrn9 (NOY703; N-PSW), rpa135 (NOY408-1a), and rrn9 rap135 (NOY896) were analyzed by spot test on YEP-galactose and YEPD. (B) Two independent colonies of strains carrying rrn6 (NOY566), rrn6 sir2 (NOY918), and rrn6 rrn5 (NOY919) were analyzed by spot test on YEP-galactose and YEPD. (C) N-PSW strains carrying rrn9 (NOY703), rrn9 sir2 (NOY901), rrn9 sir3 (NOY911), and rrn9 sir4 (NOY912) were grown on a YEP-galactose plate. Three single colonies from each strain were analyzed by spot test.

rrn6::LEU2

rrn9 sir2

4

1

4

3

Altered localization of the nucleolus in polymerase switched strains. We examined the structure of the nucleolus in PSW strains by using EM and immunofluorescence microscopy (IFM). Since the presence of helper plasmids such as pNOY103, which allows transcription of the GAL7-35S rDNA fusion gene by Pol II, may lead to formation of several mininucleoli (27, 28), thus complicating the analysis, we examined PSW strains without such helper plasmids. Figure 6 shows electron micrographs of thin sections of cells of a rrn9 rpa135 PSW strain (NOY794) and cells of a wild-type strain (NOY505). The control cells showed electron-dense nucleolar materials at or near the nuclear periphery, forming the normal crescent-shaped nucleolus (Fig. 6A). In contrast, the nucleolus of the rrn9 rpa135 PSW strain (without any helper plasmid) showed a round nucleolus, which is distant from the nuclear periphery. Quite often, the nucleolus in the PSW strain revealed two different parts, one with higher and the other with lower electron density. Although we have not studied the basis of the presence of two subnucleolar regions, a similar feature as well as the interior localization of the nucleolus was observed for strains with chromosomal rDNA deletions complemented by the GAL7-35S rDNA fusion gene on a multicopy plasmid (28). IFM using antibodies against nucleolar protein Ssb1p (3) also supported the interior localization of the nucleolus in PSW strains, which was clearly different from the crescent structure along the nuclear periphery seen in the control strains (data not shown).

View larger version (64K): [in this window] [in a new window]   FIG. 6.   EM analysis of the nucleolus in control strain NOY505 (A) and PSW strain NOY794 (B). The strains were grown in YEPD at 25°C to an A600 of about 0.5, and samples were prepared for EM analysis. The nuclear envelope is marked with arrows to serve as a point of reference, and the nucleolus (electron-dense areas within the nucleus) is indicated as N. The vacuole is indicated as V. Bars, 1 µm.

View larger version (43K): [in this window] [in a new window]   FIG. 7.   FISH analysis of rDNA in PSW strain NOY852 and control strain W303-1a. Yeast strains were analyzed for rDNA and DNA as described in Materials and Methods. Images of rDNA and DNA were pseudocolored green and red, respectively, giving overlapped regions yellow in overlay. Individual images are shown in black and white. Note that in the PSW strain, many of the DAPI-stained nuclei have a hole with decreased DAPI staining and that rDNA appears to surround these holes. In the control strain, such a hole surrounded by rDNA was rarely seen.

Transcription of rDNA by Pol II in N-PSW rrn9 strains. Cells which can grow reasonably well on glucose (i.e., are able to form colonies on glucose) are defined to be in the PSW state, whereas cells that cannot grow or can grow only extremely slowly on glucose (no visible colonies after 7 to 10 days of incubation) are defined to be in the N-PSW state (39). By defining the N-PSW state in this way, we originally assumed that transcription of rDNA by Pol II takes place only in the PSW state and not in the N-PSW state. Weak residual transcription observed in the UAF mutants under conditions of repression of the GAL7 promoter was interpreted to be due to the basal transcription of rDNA by Pol I, as was observed in in vitro experiments (19, 39). We have now examined this previous assumption and found that it is incorrect; transcription of rDNA by Pol II actually takes place in rrn9 strains even in the N-PSW state, though very weakly.

View larger version (49K): [in this window] [in a new window]   FIG. 8.   Primer extension analysis of primary transcripts for detection of rDNA transcription by Pol II. RNA was prepared from the following strains growing on synthetic galactose medium supplemented with Casamino Acids, tryptophan, and adenosine (lanes G) and 1 h after shift from galactose to glucose synthetic medium with the same supplements (lanes D): W303-1a (lanes 1, 2, 9, and 10), NOY408-1a (lanes 3 and 4), NOY920 (lanes 5 and 6), NOY921 (lanes 7 and 8), NOY566 (lanes 11 and 12), NOY918 (lanes 13 and 14), and NOY919 (lanes 15 and 16). Primer extension reactions were done in parallel, but gel electrophoresis and autoradiography were done in two separate groups (shown in A and B) with the same WT samples included. Autoradiograms (exposure times, 12 h [A] and 18 h [B]) are shown. Lane 8' is the same as lane 8 after a longer exposure, which was equivalent to ~30 h. Positions indicated as +1, G, and P correspond to the start site for the Pol I rDNA promoter, that for the GAL7 promoter, and a major site (29) among the 5' ends identified for rDNA transcripts in PSW strains, respectively. Three independent experiments were carried out, and the amounts of Pol II-specific precursor rRNAs found in NOY921 in glucose (lane 8 or 8') were 15.4% ± 2.0% of those in control PSW strain NOY920 (lane 6).

rrn6 sir2

rrn6 rrn5

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

Model for RNA polymerase switch. The results presented in this report demonstrate that switching to the PSW state, i.e., the state that allows growth using the Pol II system, involves two steps: a mutation in UAF and an expansion of rDNA repeats. Table 2 summarizes information obtained on the WT, N-PSW, and PSW states. We found that some mutational alterations of UAF are sufficient to allow Pol II transcription of rDNA, although transcription is weak (~15% of the control PSW strain) and is apparently not sufficient to allow cellular growth. After repression of the GAL7 promoter by glucose, N-PSW UAF mutants can continue to grow and divide at least for several generations by using preexistent ribosomes, and possibly also some new ribosomes, that would continue to be synthesized by the weak rDNA transcription by Pol II. This residual growth may be sufficient to allow UAF mutants to continue FOB1-dependent expansion and contraction of rDNA repeats at least for a while so that a small fraction of cells (10³ to 10⁴) can form colonies, enabling some of them to establish a fully competent PSW state.

View this table: [in this window] [in a new window]   TABLE 2.   Summary of information on rrn9 N-PSW, rrn9 PSW, and control RRN9 (WT) strains

Significance of the requirement of rDNA repeat expansion for establishment of the PSW state. There are two different models to explain the requirement of rDNA repeat expansion for establishing the PSW state. The first model assumes that in rrn9 N-PSW strains, all of the rDNA repeats are accessible to Pol II and transcribed in a productive way leading to ribosome formation and that a higher gene dosage is required simply to increase the overall rate of rDNA transcription to meet the need for growth. In this case, one has to explain the formation of a distinct nucleolar structure in PSW cells, which is localized at an interior site rather than at the original site at the nuclear periphery. Perhaps UAF is a key element for retaining rDNA at the nuclear periphery and rrn9 cells in the N-PSW state have rDNA (and the Pol II-specific nucleolus) at an interior site. Expansion of rDNA repeats may simply increase the rRNA synthesis rate without changing the location of the nucleolus. Unfortunately, it has been difficult, for technical reasons, to study the nucleolus in rrn9 N-PSW cells which do not carry a helper plasmid, and thus, we have no information on this question (Table 2).

Relation to rDNA silencing of Pol II reporter genes. Silencing of certain Pol II reporter genes inserted into rDNA has been reported (1, 10, 36). One feature of this silencing system is the requirement of SIR2 but not SIR3 and SIR4. Similarly, SIR2, but not SIR3 or SIR4, was shown to play a role in decreasing the rate of recombination within rDNA repeats (12). In the present system, the second step in switching to the PSW state is stimulated by a sir2 mutation but not by a sir3 or sir4 mutation. Thus, a rDNA chromatin structure containing Sir2p postulated to be important for rDNA silencing of reporter genes or inhibition of recombination appears to play a role in decreasing switching to the PSW state by decreasing the rate of expansion and contraction of rDNA.

Significance of large tandem repeat numbers of rDNA genes. In most eukaryotes, rDNA genes are tandemly repeated at one or a few chromosomal loci, the nucleolar organizers. The repeat number at a locus is generally large but is highly variable among organisms, ranging from less than 100 to over 10,000 per haploid genome. The repeat numbers often vary significantly not only among related species but also among different strains of the same species (reviewed in reference 22; for variations within S. cerevisiae, see references 2 and 30). It was often assumed that the presence of large numbers of rDNA genes reflects a demand for high rates of rRNA synthesis to meet cellular growth needs. However, the large variations of gene number among closely related organisms are difficult to explain on this basis. In addition, a given organism does not appear to require the presence of all rDNA gene copies for normal growth rates. We have previously described the construction of a yeast strain which carries only 25% (i.e., ca. 40 repeats) of the normal rDNA repeats (21). This strain was constructed by first deleting RPA135 in the presence of a helper plasmid, which caused reduction of the rDNA repeat numbers followed by deletion of FOB1, preventing expansion and contraction of the repeats, and finally by reintroducing the missing RPA135 gene. This yeast strain and a control yeast strain with ~150 rDNA repeats showed identical growth rates (21) and rRNA synthesis rates (16). Thus, it is clear that normal yeast cells use only a fraction of the ~150 rDNA repeats to synthesize rRNA by Pol I, even under conditions of near maximum growth rates. Previous analyses of rDNA chromatin structure using psoralen cross-linking also showed that only a fraction of the rDNA copies is transcribed in actively growing cells in a variety of systems including yeast (5, 8). As we discussed above in connection with the requirement of rDNA repeat expansion for the formation of a new Pol II-specific nucleolar structure, extra rDNA repeats might be present simply to form suitable nucleolar structures rather than to function as template for rRNA synthesis. Perhaps the number of rDNA repeats unique to each organism reflects the presence of particular nucleolar structures unique to these organisms (and environmental or developmental conditions).