INVESTIGATION

Natural Variation in CDC28 Underlies MorphologicalPhenotypes in an Environmental Yeast Isolate

Hana N. Lee,* Paul M. Magwene,† and Rachel B. Brem*,1

*Department of Molecular and Cell Biology, University of California, Berkeley, California 94720and †Department of Biology, Duke University, Durham, North Carolina 27708

ABSTRACT Morphological differences among individuals in a species represent one of the most striking aspects of biology, anda primary aim of modern genetics is to uncover the molecular basis of morphological variation. In a survey of meiosis phenotypesamong environmental isolates of Saccharomyces cerevisiae, we observed an unusual arrangement of meiotic spores within the sporesac in a strain from Ivory Coast, West Africa. We mined population genomic data to identify CDC28 as the major genetic determinantof meiotic and budding cell shape behaviors in this strain. Molecular genetic methods confirmed the role of the Ivory Coast variant ofCDC28 in the arrangement of spores after meiosis, in the shape of budding cells in rich medium and in the morphology of filamentousgrowth during nitrogen limitation. Our results shed new light on the role of CDC28 in yeast cell division, and our work suggests thatwith the growing availability of genomic data sets in many systems, a priori prediction of functional variants will become an increasinglypowerful strategy in molecular genetics.

NATURAL genetic variation in morphology has given riseto Darwin’s “endless forms most beautiful” across the

tree of life, and the search for the molecular basis of mor-phological variation remains a major focus of evolutionarygenetics. To this end, unbiased genome-wide mapping isoften used to find loci that underlie trait variation withinspecies. However, the potential for high cost and limitedstatistical power of genome-wide analyses has led manyresearchers to test for causal DNA variants only in a subsetof candidate genes, those predicted from prior knowledgeabout the trait of interest (Risch 2000; Neale and Savolainen2004; Yoo et al. 2009). Classically, the candidate gene ap-proach has met with varied degrees of success, owing tolimited genotypic data and lack of detailed functional infor-mation for many genes and pathways (Tabor et al. 2002;McCarthy et al. 2003; Hirschhorn and Daly 2005; Todd2006). Recently, increases in the availability of genome se-quences and functional genomic resources have renewedinterest in hypothesis-driven molecular genetics, in which

functional variants can be predicted a priori from DNA se-quence (Moreno et al. 2008; Wang and Marcotte 2010).

The cell morphology and growth habit of budding yeasthave been the target of genetic studies in laboratory strains(Jorgensen et al. 2002; Pruyne et al. 2004; Ohya et al. 2005;Moseley and Nurse 2009) and environmental isolates (Yvertet al. 2003; Fidalgo et al. 2006; Nogami et al. 2007; Granekand Magwene 2010), but the genetic underpinnings of manynaturally occurring yeast morphologies remain unknown. Inanalysis of the genetic basis of varying traits in yeast, a re-cently reported population genomic data set (Liti et al. 2009)established a powerful test bed for the candidate gene ap-proach. In the present work, we undertook a proof of conceptfor hypothesis-driven genetic dissection of natural variation,using yeast sporulation as a model system. Upon observing anunusual arrangement of meiotic spores in an environmentalyeast isolate, we set out to mine genome sequence data andtest predictions about the underlying causal basis of this de-velopmental phenotype.

Materials and Methods

Reciprocal hemizygote strain construction

All strains used are listed in Table 1. The Ivory Coast(NCYC110) and oak (YPS606) wild isolates were obtainedfrom the National Center for Yeast Collections (NCYC) in

Copyright © 2011 by the Genetics Society of Americadoi: 10.1534/genetics.111.128819Manuscript received March 18, 2011; accepted for publication April 24, 2011Supporting information is available online at http://www.genetics.org/content/suppl/2011/04/28/genetics.111.128819.DC1.1Corresponding author: Department of Molecular and Cell Biology, 304A Stanley Hall,no. 3220, University of California, Berkeley, CA 94720. E-mail: rbrem@berkeley.edu

Genetics, Vol. 188, 723–730 July 2011 723

both diploid HO and haploid ho form. Ivory Coast · oakhybrid strains (YHL118 and YHL121) were generatedfrom single-cell matings of the haploid ho strains fromthe two parental backgrounds to each other. These hy-brid strains were subsequently used to make reciprocalhemizygotes by deleting one allele of CDC28 with aURA3 cassette. The CDC28 locus was sequenced in eachtransformant to verify that only one allele was present andto identify the allele. CDC28 hemizygotes were generatedfrom homozygous diploid parent backgrounds via thesame method.

Genomic analysis

Whole-genome alignments for a panel of genetically di-vergent strains, including the Ivory Coast and oak isolates,were retrieved from the Saccharomyces Genome Resequenc-ing Project database (Liti et al. 2009). The aligned sequencesfor the open reading frames of annotated protein-codinggenes were analyzed for the presence of nonsynonymoussingle nucleotide polymorphisms that were unique to theIvory Coast strain.

Linkage analysis

A homothallic hybrid was constructed by mating singlespores of HO wild-type Ivory Coast and oak strains to eachother. This hybrid (YHL047) was sporulated and 24 tetradswere dissected. One homothallic segregant from each tetradwas sequenced at CDC28 and analyzed for sporulation andmitotic growth phenotypes. Sporulation morphology wasquantified as the proportion of linear tetrads observed bymicroscopy, with .600 tetrads analyzed in each sample.Linkage to CDC28 was assessed by a Wilcoxon signed-ranktest to compare segregant phenotypes associated with thetwo CDC28 alleles. The percentage of variance due toCDC28 genotype was calculated as the R2 of the linearregression of the segregant phenotypes to their CDC28genotypes.

Sporulation

Cultures were grown for at least 5 days on solid minimalsporulation media (Amberg et al. 2005), prior to imagingat ·40 magnification.

Table 1 Yeast strains used

Strain name Parents Genotype Source

NCYC110 MATa/a NCYCYPS606 MATa/a NCYCNCYC110 MATa MATa ura3D::kanMX hoD::hphMX NCYCNCYC110 MATa MATa ura3D::kanMX hoD::hphMX NCYCNCYC110 MATa/a MATa/a ura3D::kanMX hoD::hphMX NCYCYPS606 MATa MATa ura3D::kanMX hoD::hphMX NCYCYPS606 MATa MATa ura3D::kanMX hoD::hphMX NCYCYPS606 MATa/a MATa/a ura3D::kanMX hoD::hphMX NCYCG30045

P1278b MATa/a ura3D0/ura3D0 his3D0/his3D0 clb2D::kanMX/

clb2D::kanMXDrees et al. (2005)

YHL047 NCYC110 · YPS606 MATa/a HO/HO This articleSegregants 1–24 YHL047 MATa/a HO/HO This articleYHL118 NCYC110 MATa ·

YPS606 MATaMATa/a ura3D::kanMX/ura3D::kanMX hoD::hphMX/hoD::hphMX This article

YHL121 NCYC110 MATa ·YPS606 MATa

MATa/a ura3D::kanMX/ura3D::kanMX hoD::hphMX/hoD::hphMX This article

YHL169 NCYC110 MATa/a MATa/a ura3D::kanMX/ura3D::kanMX hoD::hphMX/hoD::hphMXcdc28D::URA3/CDC28-NCYC110

This article

YHL170 YPS606 MATa/a MATa/a ura3D::kanMX/ura3D::kanMX hoD::hphMX/hoD::hphMXcdc28D::URA3/CDC28-YPS606

This article

YHL171 NCYC110 MATa/a ura3D::kanMX/ura3D::kanMX hoD::hphMX/hoD::hphMXcdc28D::URA3/CDC28-NCYC110

This article

YHL172 NCYC110 MATa/a MATa/a ura3D::kanMX/ura3D::kanMX hoD::hphMX/hoD::hphMXcdc28D::URA3/CDC28-NCYC110

This article

YHL173 YHL118 MATa/a ura3D::kanMX/ura3D::kanMX hoD::hphMX/hoD::hphMXcdc28D::URA3/CDC28-YPS606

This article

YHL174 YHL118 MATa/a ura3D::kanMX/ura3D::kanMX hoD::hphMX/hoD::hphMXcdc28D::URA3/CDC28-YPS606

This article

YHL175 YHL118 MATa/a ura3D::kanMX/ura3D::kanMX hoD::hphMX/hoD::hphMXcdc28D::URA3/CDC28-NCYC110

This article

YHL176 YHL121 MATa/a ura3D::kanMX/ura3D::kanMX hoD::hphMX/hoD::hphMXcdc28D::URA3/CDC28-NCYC110

This article

YHL177 YHL121 MATa/a ura3D::kanMX/ura3D::kanMX hoD::hphMX/hoD::hphMXcdc28D::URA3/CDC28-NCYC110

This article

YHL178 YHL121 MATa/a ura3D::kanMX/ura3D::kanMX hoD::hphMX/hoD::hphMXcdc28D::URA3/CDC28-NCYC110

This article

724 H. N. Lee, P. M. Magwene, and R. B. Brem

Mitotic growth

Cultures were grown to mid-log phase (OD600 �0.7) in liq-uid rich media (Amberg et al. 2005), prior to imaging at·40 magnification.

Pseudohyphal growth

Cultures were grown overnight on solid rich media. Cellsfrom a single colony were suspended in water and diluted toan OD600 of 0.01 before plating on nitrogen-limited mediawith 1% butanol to induce filamentous growth (Gimeno et al.1992). Colonies were imaged at ·10 after 4 days of growth.

Results

In wild-type yeast, meiosis is induced by nutrient limitation,and the four haploid spores resulting from meiotic divisionsform a structure called a tetrad. We surveyed meiotic phe-notypes in a panel of fully sequenced, genetically diverseyeast isolates, which fall into well-defined phylogeneticpopulations (Liti et al. 2009). Upon starvation, all strainsexhibited tetrads with a tetrahedral shape (supporting infor-mation, Figure S1), except for NCYC110, an Ivory Coast iso-late, in which we frequently observed an alternate formwhere the spores were arranged linearly in the ascus or sporesac (Figure 1A). This observation was particularly striking,given the tetrahedral tetrad form in DBVPG6044 (FigureS1), an isolate from Guinea, which is only 4.8 · 1025 percentdivergent from the Ivory Coast strain across protein-codingregions. Linear tetrads have previously been observed in lab-oratory yeast, but only rarely (Thomas and Botstein 1987;Lindsey et al. 2010; Piccirillo and Honigberg 2010), and

were used to deduce the first centromere linkage in yeast(Hawthorne 1955).

To identify candidate alleles underlying the naturallyoccurring linear tetrad form of the Ivory Coast isolate, wesearched population genomic sequence data (Liti et al.2009) for nonsynonymous coding variants unique to thisstrain. Of the six genes harboring such variants (CDC28,PAU3, SEC27, SIW14, RAS2, and TRM44), we consideredCDC28 to be the best candidate on the basis of its knownrole in cell morphology. CDC28 is a cyclin-dependent ki-nase that regulates polarization during cell division (Lewand Reed 1993). CDC28 knockouts are inviable and condi-tional loss-of-function mutations confer both meiotic andmitotic defects (Giaever et al. 2002; Kitazono and Kron2002; Benjamin et al. 2003; Kitazono et al. 2003), butvarious CDC28 point mutations give rise to enhanced po-larized growth (Edgington et al. 1999; Ahn et al. 2001).Furthermore, a laboratory strain null for a CDC28 bindingpartner, CLB2, phenocopied the tetrad morphology of theIvory Coast isolate (Figure 1). The Ivory Coast strain har-bored a unique serine-to-phenyalanine amino acid changeat residue 79 in the CDC28 sequence, an amino acid form-ing part of a region known in the human homolog, CDK1(CDC2), to be critical for CDK–cyclin interactions (Jeffreyet al. 1995; Otyepka et al. 2006). On this basis, we hypoth-esized that the Ivory Coast allele of CDC28 was a geneticdeterminant of the linear tetrad phenotype.

To test this hypothesis, we first crossed the Ivory Coaststrain to YPS606, a strain isolated from the exudate of NorthAmerican oak trees, which has high sporulation efficiency(Gerke et al. 2006) and normal tetrad form (Figure 1). We

Figure 1 Linear tetrad morphology observed in an Ivory Coast isolate ofS. cerevisiae. Each panel shows a micrograph of yeast culture taken after5 days of growth on solid minimal sporulation media. Examples of lineartetrads are indicated with arrowheads. (A) Ivory Coast and North Amer-ican oak strains. (B)

P1278b laboratory strain bearing a deletion in CLB2.

(C) Ivory Coast · oak hybrid.

Figure 2 The linear tetrad phenotype links to the CDC28 locus. Eachcolumn reports a distribution of proportions of linear tetrads across cul-tures of homozygote, homothallic yeast strains grown for 5 days ofgrowth on solid minimal sporulation media. For each distribution, themedian is reported as a thick horizontal line, the 25% quantile is shownas a box, and the extremes are shown as thin horizontal bars. Parent:homozygotes derived from wild isolates (Liti et al. 2009); distributionsrepresent sporulation phenotypes across three biological replicates. Seg-regants, Ivory Coast allele: F1 progeny from a cross between Ivory Coastand oak strains bearing the Ivory Coast allele of CDC28 (n ¼ 12). Segre-gants, oak allele: F1 progeny bearing the oak allele of CDC28 (n ¼ 12).

Natural Variation in Yeast CDC28 725

collected haploid recombinant progeny from this cross andquantified the proportion of linear tetrads in each progenystrain. The results, shown in Figure 2, revealed strong link-age between the linear tetrad phenotype and genotype atCDC28 (Wilcoxon P ¼ 3.5 · 1025). Consistent with thephenotypes of the parent strains, the Ivory Coast allele ofCDC28 was associated with a high proportion of linear tet-rads relative to the effect of the oak allele (Figure 2). In-terestingly, segregants that bore the Ivory Coast allele atCDC28 did not all phenocopy the Ivory Coast parent: thefrequency of linear tetrads varied across this strain set,reflecting the action of additional modifier loci segregatingin the cross (Figure 2). In contrast, across segregants bear-ing the oak allele at CDC28, proportions of linear tetradswere tightly distributed near zero (Figure 2). In an analysis

of all segregants, 73% of variance in the linear tetrad phe-notype was explained by the CDC28 locus. We conclude thatthe CDC28 locus is the major genetic determinant of thistrait in the cross, with additional epistatic modifiers mani-festing only in the presence of the Ivory Coast allele ofCDC28 that together give rise to the phenotype of the IvoryCoast parent strain.

To confirm the identity of CDC28 as the major causativegene underlying the linear tetrad phenotype, we generatedreciprocal hemizygotes (Steinmetz et al. 2002) for CDC28 ina diploid hybrid formed from a mating between the IvoryCoast and oak parents. This strategy produced two hybridstrains genetically identical to one another at all loci exceptCDC28, at which one hemizygote strain bore only the IvoryCoast allele and the other strain only the oak allele. Inducing

Figure 3 Natural variation inCDC28 is causative for the lineartetrad phenotype. (A) Proportionof linear tetrads in hemizygoteand wild-type strains. Each col-umn reports data from threebiological replicates (parental wild-type strains) or independentlygenerated isogenic strains (hybridsand hemizygotes). Bar heights re-port means, and error bars repre-sent one standard deviation. (B)Micrographs of CDC28 reciprocalhemizygote strains in the oak ·Ivory Coast hybrid background.(C) Micrographs of CDC28 hemi-zygote strains in homozygousoak and Ivory Coast parentalbackgrounds. Examples of lineartetrads are indicated with arrow-heads. All micrographs weretaken after 5 days of growth onsolid minimal sporulation media.

726 H. N. Lee, P. M. Magwene, and R. B. Brem

meiosis in the wild-type Ivory Coast · oak hybrid yieldeda low proportion of linear tetrads (Figures 1 and 3), indi-cating that the genetic determinants of the linear tetradphenotype in the Ivory Coast strain act in a largely recessivemanner. Consistent with predictions from linkage analysis,the hemizygote bearing only the Ivory Coast allele of CDC28had a relatively high proportion of linear tetrads when spor-ulated, and linear tetrads were rare in sporulated cultures ofthe hemizygote bearing the oak allele of CDC28 (Figure 3).We conclude that CDC28 is a major causative gene for thehigh frequency of linear tetrads in the Ivory Coast strain.

The evidence for dominance by the oak allele of CDC28 inthe genetics of the linear tetrad trait (Figure 3) suggestedthat the Ivory Coast allele at this locus was likely hypomor-phic. However, given that the oak · Ivory Coast hybrid strainproduced a modest proportion of linear tetrads when spor-ulated (Figures 1 and 3), we further hypothesized that therewas a subtle effect of haploinsufficiency at CDC28 in thehybrid relative to the oak parent. Under this model, weexpected that a reduction in CDC28 dose would also giverise to an increased frequency of linear tetrads in eitherparental strain background. To test this notion, we deletedone copy of CDC28 in the oak homozygous parent and,separately, in the Ivory Coast parent. In each case, the hemi-zygote produced more linear tetrads than did its respectivehomozygote (Figure 3), with a more dramatic effect in theIvory Coast background. Thus, the functional dose of CDC28correlated inversely with the proportion of linear tetrads;

the Ivory Coast allele of CDC28 acted as a partial loss offunction, and haploinsufficiency in CDC28 in either strainincreased the frequency of linear tetrads. Notably, the hemi-zygote bearing a given allele in the parental background didnot phenocopy the respective hemizygote in the hybrid (Fig-ure 3), consistent with the evidence from linkage analysis(Figure 2) for additional variant loci as modifiers to CDC28.

In laboratory yeast, experimentally inducing elongatedmorphology during mitotic growth gives rise to a lineartetrad phenotype upon sporulation (Thomas and Botstein1987; Lindsey et al. 2010), and a laboratory strain null forthe CDC28 binding partner CLB2 grows as elongated cells inrich medium (Ahn et al. 1999). We hypothesized that thenatural variant of CDC28 in the Ivory Coast strain wouldalso confer elongated cell shape in mitotically budding cells.Consistent with this prediction, the Ivory Coast parent andhemizygotes bearing the Ivory Coast allele of CDC28 showedelongated cell morphology during mitotic growth, in con-trast to the round cells of the oak parent and the hemizy-gotes with the oak allele of CDC28 (Figure 4). Furthermore,in the progeny of the cross between the oak and Ivory Coaststrains, segregants with hyperpolarized cells in rich mediaalso produced linear tetrads when sporulated (Figure S2).Interestingly, in contrast to our analysis of tetrads (Figure3), we found no evidence for haploinsufficiency in CDC28dose as a determinant of cell shape: oak homozygotes witha single copy of CDC28 still grew as round cells, as did thewild-type hybrid between the oak and Ivory Coast strains

Figure 4 CDC28 genotype affects cell shape during mi-totic growth. Each panel shows a representative micro-graph of culture grown to mid-log phase in rich media.(A) Ivory Coast and North American oak strains. (B) Oak ·Ivory Coast wild-type hybrid strain. (C) CDC28 reciprocalhemizygotes in the oak · Ivory Coast hybrid background.(D) CDC28 hemizygotes in homozygous oak and IvoryCoast parental backgrounds.

Natural Variation in Yeast CDC28 727

(Figure 4). We conclude that the Ivory Coast allele of CDC28is sufficient to confer elongated cell morphology, reflectinga relationship between hyperpolarized mitotic growth andlinear tetrads during sporulation.

Many strains of yeast, when starved for nitrogen,undergo unipolar budding, which generates filaments oflinked cells (Gimeno et al. 1992). Given the pronouncedfilamentous phenotype under nitrogen limitation in a labo-ratory strain mutant for CLB2, the binding partner ofCDC28 (Ahn et al. 1999), we hypothesized that the IvoryCoast allele of CDC28 would drive a similar phenotype. Inaccord with this expectation, the Ivory Coast isolate, whenstarved for nitrogen, grew as long, branching filamentscomposed of elongated cells growing from the colony edge,whereas the oak strain showed a more modest responsewith nonbranching filaments composed of round cells (Fig-ure 5). Reciprocal hemizygote analysis confirmed that theIvory Coast allele of CDC28 was sufficient to confer highlybranched, elongated filaments under nitrogen limitation(Figure 5). Again, we observed no evidence for haploin-sufficiency of CDC28 dose as a determinant of filamentousmorphology: hemizygous strains generated from the ho-mozygous parent backgrounds resembled their respectiveparents, and the wild-type hybrid between the oak andIvory Coast strains resembled the oak parent (Figure 5).Thus, the Ivory Coast allele of CDC28 underlies thedramatic filamentation behavior in this environmentalisolate, mirroring its effects on tetrad and single-cellmorphology.

Discussion

Many variations in yeast cell form and growth habit havebeen reported in environmental isolates (Dengis et al. 1995;Cavalieri et al. 2000; Casalone et al. 2005; Nogami et al.2007; Granek and Magwene 2010; St’ovicek et al. 2010),with the molecular basis known for only a small subset ofthese traits. We have shown that natural genetic variation ata single essential gene, CDC28, can lead to striking changesin yeast cell shape during meiotic and mitotic division. Therod-like tetrads, elongated cell shape, and hyperfilamentousgrowth of the environmental Ivory Coast isolate studiedhere can be attributed to variation in CDC28. Previousreports have implicated loss of function in CDC28/CLB2activity in hyperpolarized cell division and filamentousgrowth (Lew and Reed 1993; Ahn et al. 1999, 2001; Edgingtonet al. 1999), but temperature-sensitive mutations available inlaboratory strains have not been well suited to the study ofCDC28 loss of function under starvation conditions. By con-trast, our ability to map starvation-induced sporulation andfilamentation phenotypes to CDC28 highlights the utility ofanalysis of natural genetic variation when drawing relation-ships between genes and phenotypes.

The molecular mechanism by which the Ivory Coast alleleof CDC28 impacts tetrad and cell morphology remains anopen question. The phenotypic similarity between the IvoryCoast strain and the Δclb2 lab strain, and the S79F change inthe former in a known cyclin-interaction domain of CDC28,suggest that the Ivory Coast allele attenuates the binding of

Figure 5 CDC28 genotype affects morphology of the fil-amentous growth response. Each panel represents a rep-resentative micrograph of colonies after 4 days of growthon solid nitrogen limitation media. (A) Ivory Coast andNorth American oak strains. (B) Oak · Ivory Coast wild-type hybrid. (C) CDC28 reciprocal hemizygotes in theoak · Ivory Coast hybrid background. (D) CDC28 hemi-zygotes in homozygous oak and Ivory Coast parentalbackgrounds.

728 H. N. Lee, P. M. Magwene, and R. B. Brem

CDC28 to CLB2. We favor a model in which elongated cellshape is the proximate cause of the linear tetrad morphologyand dramatic filamentation phenotype attributable to theIvory Coast allele, in light of the relationship between thesetraits when cell shape is manipulated in the laboratory(Thomas and Botstein 1987; Lindsey et al. 2010). With re-spect to sporulation, one compelling idea for the underlyingmechanism is that, although meiotic cell division is orderedin yeast (Hawthorne 1955), spores arrange passively in theascus as a function of cell shape before meiosis (Piccirilloand Honigberg 2010). Among the most notable aspects ofthe Ivory Coast allele of CDC28 are its incomplete pene-trance and its sensitivity to genetic modifiers, for whichthe Ivory Coast strain’s unique coding alleles of the morphol-ogy regulator RAS2 and other genes represent strong candi-dates. As such, our data indicate that the effect of the IvoryCoast variant of CDC28 on linear tetrad formation is highlysensitized both to stochastic variation in cell shape and togenetic differences between yeast strains.

The evolutionary pressures driving variation in yeastmorphology and growth habit in the wild are as yet almostentirely unknown, although several morphological charac-ters are well-characterized targets of artificial selection inbeer and wine yeasts (Verstrepen et al. 2003). Given theorigin of the Ivory Coast strain as an isolate from ginger beer(Liti et al. 2009), its elongated cell and tetrad forms may beassociated with a fitness benefit in the wild or during do-mestication; alternatively, these phenotypes, and the variantin the essential cell cycle regulator CDC28, may representa rare genetic defect destined to be eliminated by purifyingselection. In either case, our discovery of CDC28 as the majordeterminant underscores the power of combining dense pop-ulation genomic sampling with a candidate gene/candidatenetwork approach. With the continued growth of populationgenomic sequence compendia, and increasing knowledge ofgene function and genetic interactions, this strategy holdspromise for application to many organisms and traits.

Acknowledgments

We thank C. Ellison and C. Hall for discussions and D. Lew,L. Holt, and J. Rine for helpful comments on the manuscript.This work was supported by an Ellison Medical FoundationNew Scholar Award in Aging to R.B.B., a Paul and Daisy SorosFellowship for New Americans to H.N.L., and a NationalScience Foundation grant (MCB-0614959) to P.M.M.

Literature Cited

Ahn, S. H., A. Acurio, and S. J. Kron, 1999 Regulation of G2/Mprogression by the STE mitogen-activated protein kinase path-way in budding yeast filamentous growth. Mol. Biol. Cell. 10:3301–3316.

Ahn, S. H., B. T. Tobe, J. N. Fitz Gerald, S. L. Anderson, A. Acurioet al., 2001 Enhanced cell polarity in mutants of the buddingyeast cyclin-dependent kinase CDC28. Mol. Biol. Cell. 12: 3589–3600.

Amberg, D. C., D. Burke, and J. N. Strathern, 2005 Methods inYeast Genetics: a Cold Spring Harbor Laboratory Course Manual.Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

Benjamin, K. R., C. Zhang, K. M. Shokat, and I. Herskowitz,2003 Control of landmark events in meiosis by the CDK Cdc28and the meiosis-specific kinase Ime2. Genes Dev. 17: 1524–1539.

Casalone, E., C. Barberio, L. Cappellini, and M. Polsinelli, 2005 Char-acterization of Saccharomyces cerevisiae natural populations forpseudohyphal growth and colony morphology. Res. Microbiol.156: 191–200.

Cavalieri, D., J. P. Townsend, and D. L. Hartl, 2000 Manifoldanomalies in gene expression in a vineyard isolate of Saccharomycescerevisiae revealed by DNA microarray analysis. Proc. Natl. Acad.Sci. USA 97: 12369–12374.

Dengis, P. B., L. R. Nelissen, and P. G. Rouxhet, 1995 Mechanismsof yeast flocculation: comparison of top- and bottom-fermentingstrains. Appl. Environ. Microbiol. 61: 718–728.

Drees, B. L., V. Thorsson, G. W. Carter, A. W. Rives, M. Z. Raymondet al., 2005 Derivation of genetic interaction networks fromquantitative phenotype data. Genome Biol. 6: R38.

Edgington, N. P., M. J. Blacketer, T. A. Bierwagen, and A. M. Myers,1999 Control of Saccharomyces cerevisiae filamentous growthby cyclin-dependent kinase Cdc28. Mol. Cell. Biol. 19: 1369–1380.

Fidalgo, M., R. R. Barrales, J. I. Ibeas, and J. Jimenez, 2006 Adaptiveevolution by mutations in the FLO11 gene. Proc. Natl. Acad. Sci.USA 103: 11228–11233.

Gerke, J. P., C. T. Chen, and B. A. Cohen, 2006 Natural isolates ofSaccharomyces cerevisiae display complex genetic variation insporulation efficiency. Genetics 174: 985–997.

Giaever, G., A. M. Chu, L. Ni, C. Connelly, L. Riles et al.,2002 Functional profiling of the Saccharomyces cerevisiae ge-nome. Nature 418: 387–391.

Gimeno, C. J., P. O. Ljungdahl, C. A. Styles, and G. R. Fink,1992 Unipolar cell divisions in the yeast S. cerevisiae lead tofilamentous growth: regulation by starvation and RAS. Cell 68:1077–1090.

Granek, J. A., and P. M. Magwene, 2010 Environmental and ge-netic determinants of colony morphology in yeast. PLoS Genet.6: e1000823.

Hawthorne, D. C., 1955 The use of linear asci for chromosomemapping in Saccharomyces. Genetics 40: 511–518.

Hirschhorn, J. N., and M. J. Daly, 2005 Genome-wide associationstudies for common diseases and complex traits. Nat. Rev.Genet. 6: 95–108.

Jeffrey, P. D., A. A. Russo, K. Polyak, E. Gibbs, J. Hurwitz et al.,1995 Mechanism of CDK activation revealed by the structureof a cyclinA-CDK2 complex. Nature 376: 313–320.

Jorgensen, P., J. L. Nishikawa, B. J. Breitkreutz, and M. Tyers,2002 Systematic identification of pathways that couple cellgrowth and division in yeast. Science 297: 395–400.

Kitazono, A. A., D. A. Garza, and S. J. Kron, 2003 Mutationsin the yeast cyclin-dependent kinase Cdc28 reveal a role inthe spindle assembly checkpoint. Mol. Genet. Genomics 269:672–684.

Kitazono, A. A., and S. J. Kron, 2002 An essential function of yeastcyclin-dependent kinase Cdc28 maintains chromosome stability.J. Biol. Chem. 277: 48627–48634.

Lew, D. J., and S. I. Reed, 1993 Morphogenesis in the yeast cellcycle: regulation by Cdc28 and cyclins. J. Cell. Biol. 120: 1305–1320.

Lindsey, R., Y. Ha, and M. Momany, 2010 A septin from the fila-mentous fungus A. nidulans induces atypical pseudohyphae inthe budding yeast S. cerevisiae. PLoS One 5: e9858.

Liti, G., D. M. Carter, A. M. Moses, J. Warringer, L. Parts et al.,2009 Population genomics of domestic and wild yeasts. Na-ture 458: 337–341.

Natural Variation in Yeast CDC28 729

McCarthy, M. I., D. Smedley, and W. Hide, 2003 New methods forfinding disease-susceptibility genes: impact and potential. Ge-nome Biol. 4: 119.

Moreno, C., J. Lazar, H. J. Jacob, and A. E. Kwitek, 2008 Com-parative genomics for detecting human disease genes. Adv.Genet. 60: 655–697.

Moseley, J. B., and P. Nurse, 2009 Cdk1 and cell morphol-ogy: connections and directions. Curr. Opin. Cell Biol. 21:82–88.

Neale, D. B., and O. Savolainen, 2004 Association genetics ofcomplex traits in conifers. Trends Plant Sci. 9: 325–330.

Nogami, S., Y. Ohya, and G. Yvert, 2007 Genetic complexity andquantitative trait loci mapping of yeast morphological traits.PLoS Genet. 3: e31.

Ohya, Y., J. Sese, M. Yukawa, F. Sano, Y. Nakatani et al.,2005 High-dimensional and large-scale phenotyping of yeastmutants. Proc. Natl. Acad. Sci. USA 102: 19015–19020.

Otyepka, M., I. Bartova, Z. Kriz, and J. Koca, 2006 Differentmechanisms of CDK5 and CDK2 activation as revealed byCDK5/p25 and CDK2/cyclin A dynamics. J. Biol. Chem. 281:7271–7281.

Piccirillo, S., and S. M. Honigberg, 2010 Sporulation patterningand invasive growth in wild and domesticated yeast colonies.Res. Microbiol. 161: 390–398.

Pruyne, D., A. Legesse-Miller, L. Gao, Y. Dong, and A. Bretscher,2004 Mechanisms of polarized growth and organelle segrega-tion in yeast. Annu. Rev. Cell Dev. Biol. 20: 559–591.

Risch, N. J., 2000 Searching for genetic determinants in the newmillennium. Nature 405: 847–856.

St’ovicek, V., L. Vachova, M. Kuthan, and Z. Palkova, 2010 Gen-eral factors important for the formation of structured biofilm-like yeast colonies. Fungal Genet. Biol. 47: 1012–1022.

Steinmetz, L. M., H. Sinha, D. R. Richards, J. I. Spiegelman, P. J.Oefner et al., 2002 Dissecting the architecture of a quantitativetrait locus in yeast. Nature 416: 326–330.

Tabor, H. K., N. J. Risch, and R. M. Myers, 2002 Candidate-geneapproaches for studying complex genetic traits: practical consid-erations. Nat. Rev. Genet. 3: 391–397.

Thomas, J. H., and D. Botstein, 1987 Ordered linear tetrads areproduced by the sporulation of newly formed zygotes of Saccha-romyces cerevisiae. Genetics 115: 229–232.

Todd, J. A., 2006 Statistical false positive or true disease path-way? Nat. Genet. 38: 731–733.

Verstrepen, K. J., G. Derdelinckx, H. Verachtert, and F. R. Delvaux,2003 Yeast flocculation: what brewers should know. Appl. Mi-crobiol. Biotechnol. 61: 197–205.

Wang, P. I., and E. M. Marcotte, 2010 It’s the machine that mat-ters: predicting gene function and phenotype from protein net-works. J. Proteomics 73: 2277–2289.

Yoo, Y. J., S. B. Bull, A. D. Paterson, D. Waggott, and L. Sun,2009 Were genome-wide linkage studies a waste of time? Ex-ploiting candidate regions within genome-wide associationstudies. Genet. Epidemiol. 34: 107–118.

Yvert, G., R. B. Brem, J. Whittle, J. M. Akey, E. Foss et al., 2003 Trans-acting regulatory variation in Saccharomyces cerevisiae and therole of transcription factors. Nat. Genet. 35: 57–64.

Communicating editor: M. Johnston

730 H. N. Lee, P. M. Magwene, and R. B. Brem

GENETICSSupporting Information

http://www.genetics.org/content/suppl/2011/04/28/genetics.111.128819.DC1

Natural Variation in CDC28 Underlies MorphologicalPhenotypes in an Environmental Yeast Isolate

Hana N. Lee, Paul M. Magwene, and Rachel B. Brem

Copyright © 2011 by the Genetics Society of AmericaDOI: 10.1534/genetics.111.128819

\