Ploidy Regulation of Gene Expression

See allHide authors and affiliations

Science  09 Jul 1999:
Vol. 285, Issue 5425, pp. 251-254
DOI: 10.1126/science.285.5425.251


Microarray-based gene expression analysis identified genes showing ploidy-dependent expression in isogenic Saccharomyces cerevisiaestrains that varied in ploidy from haploid to tetraploid. These genes were induced or repressed in proportion to the number of chromosome sets, regardless of the mating type. Ploidy-dependent repression of some G1 cyclins can explain the greater cell size associated with higher ploidies, and suggests ploidy-dependent modifications of cell cycle progression. Moreover, ploidy regulation of the FLO11 gene had direct consequences for yeast development.

Changes in the number of chromosome sets occur during the sexual cycle, during metazoan development, and during tumor progression. Organisms with a sexual cycle double their ploidy upon fertilization and reduce their ploidy by half at meiosis. In the development of almost all plants and animals, specialized polyploid and polytene cell types arise though endocycles, cell cycles lacking cell division (1). Aberrant cell cycle control during tumor progression is thought to result in polyploidy and altered cell behavior (2).

Cells of different ploidy typically show very different developmental, morphological, and physiological characteristics. However, a lack of isogenic controls obscures the contribution of ploidy to these differences. For example, in the yeastSaccharomyces cerevisiae, haploid cells of opposite mating type, MAT a andMATα, mate to produce aMAT a/MATα diploid. These three cell types have different phenotypes, many of which (such as mating, meiosis, and budding pattern) are directly attributable to their different genotypes at the mating-type locus rather than ploidy (3). However, one can parse the effects of mating type from those of ploidy by constructing isogenic sets of yeast strains that vary only in their ploidy, and then subjecting these strains to whole-genome expression analysis.

We generated an isogenic set of 11 strains varying in ploidy from haploid to tetraploid. There were three series, one for each mating-type genotype (4). Comparison of strains in which the mating type differs and ploidy is the same shows the role of mating type. Comparison of strains in which ploidy varies and the mating type is the same shows the effects of ploidy.

We looked for genes whose expression, relative to total gene expression, increased or decreased in proportion to ploidy. Using oligonucleotide-probe microarrays (5), we monitored the mRNA levels of all yeast genes in our strain set during exponential growth (6). We set two criteria (7) to identify ploidy-regulated genes: (i) There should be significant correlation with an idealized expression pattern either directly or inversely proportional to ploidy. Correlation thresholds were set such that random data are expected to produce less than one false positive in the entire genome (Fig. 1). (ii) Minima were set for relative and absolute changes in expression for each gene across all 11 strains. There were 17 genes satisfying both criteria; 10 were ploidy-induced and 7 were ploidy-repressed (Fig. 2) (8).

Figure 1

Correlation thresholds. We compared correlation coefficients for experimental data and randomized data (within-gene permutations of experimental data) with a reference pattern proportional to ploidy. The hatched distribution shows the Fisher-transformed Pearson correlation coefficients, Fisher(r) = 0.5 ln[(1 + r)/(1 – r)], of a randomized data set. The variance of Fisher(r) for our randomized data (0.123) matched the theoretical variance of Fisher(r) under the null hypothesis for no correlation [n = 11, Var = 1/(n– 3) = 0.125]. The shaded distribution behind the hatched distribution represents Fisher(r) of experimental data against the reference pattern. A correlation threshold was set at Fisher(r) = 1.26 (r = 0.851), corresponding to 3.6 standard deviations of the random distribution. The same analysis was carried out for a reference pattern inversely proportional to ploidy [not shown; r = 0.858, Fisher(r) = 1.29, 3.7 standard deviations]. These thresholds should permit less than one false positive in the entire yeast genome.

Figure 2

Ploidy regulation. Ploidy-regulated genes are shown with color-coded expression levels normalized by standardization (mean = 0, SD = 1). The genes encoding known functions (23) are as follows: CTS1, endochitinase; NDI1, NADH-ubiquinone oxidoreductase;YPS4, GPI-anchored aspartyl protease; CTR3, high-affinity copper transporter; FLO11, cell surface flocculin; CLN1 and PCL1, G1 cyclins; and GIC2, control of actin cytoskeletal organization.

Ploidy-regulated genes showed substantial differences in expression within isogenic ploidy series. Comparing expression in haploids and tetraploids, CTS1 was elevated by a factor of 12 and FLO11 was repressed by a factor of 11. Ploidy-regulated genes had an unbiased distribution of locations in the yeast genome. They tended to have complex promoters; the average upstream intergenic space was longer than 1300 base pairs, whereas the genome-wide average intergenic space was 500 base pairs. The identities of some ploidy-regulated genes predicted specific ploidy-dependent cell type differences. For example, the ploidy-induced CTS1 gene encodes a secreted endochitinase involved in the separation of mother cells and daughter buds; cts1 mutants form large clumps of cells (9). Polyploids, with greater CTS1expression, were less clumpy than haploids (10).

Expression patterns uncovered by microarray data analyses were examined by Northern blot analysis of RNA samples (11) (Fig. 3). These results confirmed the existence of distinct ploidy-dependent and mating type–specific gene expression patterns. Most genes, like ACT1, showed no change in mRNA levels (relative to total RNA) in response to either ploidy or mating type. Of 32 mating type–specific genes [for example, STE2 andFUS3 (10)], none was among the ploidy-regulated genes of Fig. 2. Ploidy-regulated genes, such as CTS1 andFLO11, showed ploidy-dependent expression regardless of the mating type. The magnitudes of the ploidy-dependent and mating-type effects observed in microarray experiments were in excellent agreement with quantitative phosphorimager analyses of Northern blots (10).

Figure 3

Northern blot analysis of representative mRNAs, showing ploidy regulation of gene expression and mating-type control.

Measurements of the cells in our isogenic series demonstrated that cell size in yeast increases with increasing ploidy (Fig. 4). The association of increased ploidy with increased cell size has been observed in bacteria, fungi, plants, and animals (1,12). The relation between cell size and ploidy could be explained by the fact that G1 cyclins, Cln1 and Pcl1, are repressed as ploidy increases (Fig. 2). Yeast cells grow continuously during the G1 phase and pass through START, the entry point into the cell cycle, at a cyclin-dependent critical size. Low expression of the G1 cyclins results in START passage at a larger size; higher expression of G1 cyclins results in START passage at a smaller size (13). According to this interpretation, the lower G1 cyclin mRNA levels in tetraploids would cause them to pass through START at a larger size than haploids.

Figure 4

Ploidy and cell morphology. IsogenicMATα cells varying in ploidy from haploid (n) to tetraploid (4n) were sampled from exponential phase cultures (6) and observed with Nomarski optics. Morphological quantities (mean ± 95% confidence, n = 50) were calculated from length and width measurements of budded mother cells (the buds provided orientation) assuming rotational symmetry about the long axis. Cell volumes: haploid, 72 ± 1 μm3; diploid, 111 ± 2 μm3; triploid, 152 ± 3 μm3; tetraploid, 289 ± 6 μm3. Cell length/width ratios: haploid, 1.20 ± 0.01; diploid, 1.24 ± 0.01; triploid, 1.29 ± 0.01; tetraploid, 1.39 ± 0.02.

In addition to increased size, yeast cells showed greater elongation with increasing ploidy (Fig. 4). The basis for the connection between ploidy and morphology is suggested by ploidy-regulated genes. The ploidy-repressed Gic2 protein interacts with the Rho-type guanosine triphosphatase Cdc42. These proteins, along with Gic1, control actin organization and polarized cell growth, including bud site selection, bud emergence, and mating projection (shmoo) formation (14). Ynr067C is a ploidy-induced homolog (57% similar) of Acf2, a protein involved in polarized cortical actin assembly (15). The control of genes likeGIC2and YNR067C suggests ploidy-dependent cytoskeletal organization.

Ploidy-dependent gene expression in yeast has important implications for the sexual cycle, development, and tumor biology. In mammalian cells, polyploidy in specialized cell types and hyperploidy in tumor cells may control cell physiology, morphology, and behavior. Although hyperploidy in tumor cells is usually viewed as a consequence of aberrant cell cycle control, altered ploidy may actually be a cause of altered cell properties.

We observed ploidy regulation of invasiveness, a developmental response in yeast. Invasive growth is a model for fungal virulence involving a change in budding pattern and the formation of chains of yeast cells that invade an agar substrate (16). On rich media,MAT a and MATα haploids exhibit much more vigorous invasive growth than MAT a/α diploids (16). Haploid flo11 mutants fail to invade, and overexpression of FLO11inducesMAT a/α diploids to invade more vigorously (17). Thus, Flo11, a serine-threonine–rich cell wall protein (18) whose molecular function is unknown, is a major determinant of this developmental response.

The ploidy repression of the FLO11 gene was reflected in invasive growth (Fig. 5). As ploidy increased, the expression of FLO11 and invasiveness decreased. In addition,MAT a/α cells showed less FLO11 expression (Figs. 2 and 3) and less invasiveness (Fig. 5, C and D) than did mating-type homozygotes of the same ploidy. Thus, ploidy (as well as mating type) governs this key developmental process.

Figure 5

Control of invasive growth by ploidy and mating type through FLO11 expression. Invasive growth of strains [(A), genotypes] was visualized by washing the surface cells from a rich-medium petri plate with water (16). The plate is shown before washing (B), after a brief wash (C), and after extensive washing (D). AFLO11 + multicopy plasmid, but not the vector plasmid (pRS202, 2 μm URA3), restores invasiveness to a haploid flo11 null mutant (as a control) and to a tetraploid [(E), genotypes, all strains are MAT a; (F), after washing].

The deficit of FLO11 expression in cells of higher ploidy is directly connected with diminished invasion and is not simply a coincidental effect. A FLO11 multicopy plasmid restored invasiveness to a tetraploid (Fig. 5, E and F). Thus, ectopic overexpression of FLO11 can counteract the loss of invasiveness resulting from increased ploidy.

Patterns of gene expression associated with ploidy are not likely to be caused by alterations in growth rate or viability, or to be manifestations of differences such as the time of entry to diauxic shift during culture growth. For all of our strains, the exponential growth rates were similar and cell viabilities were greater than 95% (6). Expression patterns were monitored in cells at the same stage of exponential-phase growth. Furthermore, there is only one gene, YER067W, in common between the set of 17 ploidy-regulated genes and 98 genes induced or repressed early in diauxic shift (19).

There are several mechanisms by which gene transcription could respond to ploidy. Here we consider two: sensing gene dosage (20) and sensing total DNA content. In our ploidy series there are no changes in the dosage of any gene relative to other genes. However, the absolute number of each gene per cell is increasing and could be sensed by transient pairing of homologous chromosomes. Such transient pairing has been observed in humans, Drosophila, andSaccharomyces (21). Homologous pairing has been implicated in transvection, dominant position-effect variegation, and gene silencing, all of which involve alterations in gene expression (22). If pairing is responsible for ploidy-dependent regulation, then the presence of more homologs may increase the number of transient pairing interactions. Alternatively, the increase in the amount of DNA per cell presumably results in increased nuclear size and a reduction in the nuclear surface area/volume ratio. These changes may affect the import and nuclear concentration of regulatory proteins, which could result in the alterations in transcription we observe.

  • * To whom correspondence should be addressed. E-mail: fink{at}


View Abstract

Navigate This Article