|Home | About | Journals | Submit | Contact Us | Français|
The study of temperature sensitive (Ts) mutant phenotypes is fundamental to gene identification and for dissecting essential gene function. In this chapter we describe two “shuffling” methods for producing Ts mutants using a combination of PCR, in vivo recombination, and transformation of diploid strains heterozygous for a knockout of the desired mutation. The main difference between the two methods is the type of strain produced. In the “plasmid” version, the product is a knockout mutant carrying a centromeric plasmid carrying the Ts mutant. In the “chromosomal” version, The Ts allele is integrated directly into the endogenous locus, albeit not in an entirely native configuration. Both variations have the ir strengths and weaknesses, which are discussed here.
The study of temperature sensitive (Ts) mutant phenotypes has proven to be a fundamental approach both for the identification of gene sets essential for various aspects of biology and for obtaining a detailed understanding of essential gene function. While the observation that temperature sensitive mutations represent a general class of mutation was recognized in the ‘50s (Horowitz, 1950), the first targeted screen, isolation, and analysis of Ts mutants (382 mutations located in 37 genes scattered widely over the bacteriophage T4 genome) was by Edgar and Lielausis in 1963 (Edgar and Lielausis, 1964). Lee Hartwell, in 1967, reported the isolation of 400 Ts mutations in S. cerevisiae, which caused defects in essential processes including cell division, and protein, RNA, and DNA synthesis (Hartwell, 1967). Over the past 40 years, the isolation and analysis of Ts mutations in essential genes has been a linchpin technology for investigating the genetics and molecular biology of essential processes in all experimental organisms.
Ts mutations are typically missense mutations, which retain the function of a specific essential gene at standard (permissive) low temperature, lack that function at a defined high (non-permissive) temperature, and exhibit partial (hypomorphic) function at an intermediate (semi-permissive) temperature. Such mutants make possible the analysis of physiologic changes that follow controlled inactivation of a gene or gene product by shifting cells to a non-permissive temperature, offering a powerful approach to the analysis of gene function.
Essential genes, by definition, encode critical cellular functions that are not buffered by redundant functions or pathways (Hartman et al., 2001). Essential genes have been shown to be highly dense hubs within genetic interaction networks and are involved in all aspects of basic cellular function (Jeong et al., 2001). Furthermore, essential genes tend to be more highly conserved in evolution; 38% of essential yeast proteins have easily identifiable counterparts in humans, versus 20% for nonessential genes (Hughes, 2002).
Despite their importance, the functions of many essential yeast proteins have not been studied. In part, this is due to the absence of essential gene representation in the genome wide haploid mutant collections, which cover all of the ~5000 non-essential yeast genes. Thus, no comparable systematic haploid mutant collection currently exists for the ~1000 essential genes in S. cerevisiae. The frequency of sites mutable to a reduced or conditional function is highly gene-specific; for example, for highly conserved proteins, random single missense mutation would be expected, for the vast majority of positions within the protein, to cause complete loss of function when mutated. Therefore, genetic screens using a random mutagenesis approach rarely reach saturation because “mutability” varies widely among genes.
Here we report detailed protocols for two methodologies that allow the systematic isolation of Ts alleles in essential genes of interest. The first method is plasmid-based, and the second is genome integration based, and each has its specific advantages depending on the application. Both methods exploit features of the “haploid convertible” heterozygous diploid collection, which allows introduction of the library of mutagenized essential gene copies into the heterozygous diploid and subsequent direct selection of haploids which are deleted for the target essential gene and that carry individual members of the mutagenized essential gene library, using the “diploid shuffle” technique (see below). Several other useful corollary methods for transferring extant Ts alleles or specific gene constructs (eg, fusion proteins) are also presented. Finally, we note that our laboratories are in the process of generating a complete set of Ts alleles for each of the essential genes in S. cerevisiae, which will be distributed as a resource to the scientific community when completed (see Ben Aroya etal 2008 for details). For specific essential genes under study in individual laboratories, however, it may be useful, using the methods described here, to generate an additional series of independent Ts alleles for detailed functional analysis. Furthermore, mutagenized libraries of specific essential or non-essential genes can be screened for conditional viability under a variety of conditions that are normally sub-lethal in the wildtype strain (eg, sublethal doses of drugs) using the methods described here.
Much like traditional plasmid-shuffling methods, mutants generated by this version of the diploid shuffle are plasmid-borne alleles that can be very easily transferred to and tested in different strain backgrounds. The experimental procedure is outlined in Fig. 1. First, the endogenous promoter (including the 5′ untranslated region, 5′UTR, ~500bp) and the terminator (3′UTR, ~500bp) of a gene of interest (or your favorite gene, YFG) are PCR amplified and cloned in tandem onto a centromere-based yeast-E. coli shuttle vector, which contains URA3 as the selectable marker in yeast cells, such as pRS416. The resultant promoter/terminator clone is subsequently linearized with an endonuclease, typically NotI, which cuts at a site pre-engineered between the promoter and terminator. Simultaneously, the sequence of YFG, including the whole open reading frame (ORF), complete with the promoter and terminator regions, is randomly mutagenized with error prone PCR. The linearized promoter/terminator plasmid and PCR products are then combined and co-transformed into a haploid-convertible heterozygous diploid deletion mutant (YFG/yfgΔ::kanMX4) in the same gene being mutagenized. The mutagenized PCR product is thereby cloned into the URA3 plasmid via recombination mediated by the terminal homologous DNA sequences of both the PCR products and the linearized vector. The Ura+ transformants are subsequently cultured in a sporulation medium and converted into haploid cells by growing on a medium that allow growth of only haploid MATa G418R Ura+ cells. In these cells, the chromosomal wild-type copy of YFG is deleted, allowing direct observation of any phenotypes of the plasmid-borne alleles. To screen for conditional alleles, such haploid cells are first grown under a permissive condition such as low temperature. Colonies formed are subsequently replica-plated to fresh plates at permissive and nonpermissive conditions. Conditional alleles are identified as those grow under the permissive but not the non-permissive condition and subsequently verified. This method has been used to create thermosensitive (Ts) alleles of multiple essential genes as well as a large collection of methyl methanesulfonate (MMS) hypersensitive alleles of POL30 (Huang et al., 2008; Lin et al., 2008). Here we will outline the detailed methods to generating and verifying Ts alleles of an essential gene.
Haploid selection synthetic medium SC–Ura–Leu–His–Arg+G418+Can: dextrose, 20 g/L; yeast nitrogen base without amino acids and ammonium sulfate, 1.7 g/L; SC–Ura–Leu–His–A dropout mix, 2g/L; sodium glutamate, 1g/L; G418, 200 mg/L; L-canavanine (Sigma, Cat# C1625), 60 mg/L; Agar, 2%. The sodium glutamate is substituted for ammonium sulfate as the nitrogen source and makes the G418 selection more reliable on the minimal medium.
SC–Ura: dextrose, 20g/L; yeast nitrogen base without amino acids and ammonium sulfate, 1.7g/L; SC–Ura dropout mix, 2g/L; ammonium sulfate, 5g/L; Agar, 2%.
Liquid YPD: yeast extract, 10g/L; peptone, 20g/L; dextrose, 20g/L.
Solid and liquid sporulation medium: potassium acetate, 10g/L; zinc acetate 0.05g/L, with or without 2% agar respectively.
Solid Luria Broth (LB) plus carbenicillin: yeast extract, 10g/L; Tryptone, 5g/L; sodium chrolide, 10g/L; carbenicillin, 50mg/L; Agar, 2%.
Liquid LB plus ampicillin: yeast extract, 10g/L; Tryptone, 5g/L; sodium chrolide, 10g/L; ampicillin, 50mg/L.
The haploid-convertible heterozygous diploid knockout mutants (MAT aα/ura3Δ0/ura3Δ0 leu2Δ0/leu2Δ0 his3Δ1/his3Δ1 lys2Δ0/LYS2 met15Δ0/MET15 can1Δ::LEU2-MFA1pr::His3/CAN1 YFG/yfgΔ::KanMX; OpenBiosystems Cat# YSC4428) (Pan et al., 2006) is used to screen for Ts alleles. Chemically competent DH5α cells prepared as described (Inoue et al., 1990) are used for cloning and plasmid recovering from yeast.
It is essential for this method to use a plasmid vector that contains the YFG promoter and terminator separated by a unique endonuclease (typically NotI) recognition site. Due to the limited auxotrophic markers available in the haploid-convertible heterozygous diploid knockout mutants, we normally use plasmids containing URA3 as the selectable marker such as pRS416 (Brachmann et al., 1998; Sikorski and Hieter, 1989) and YCplac33 (Gietz and Sugino, 1988). Other URA3 CEN vectors should also work.
A genomic DNA sample isolated from the wild-type yeast strain BY4743 MATa/α (Brachmann et al., 1998) was used as the template for cloning the promoter and terminator of YFG and for mutagenizing its entire sequence with PCR.
In the past, we mostly used endonuclease restriction enzyme digestion and ligation to construct the promoter/terminator clone. First, the promoter and terminator of YFG are separately PCR-amplified using primers that contain endonuclease recognition sites, for example, HindIII/NotI for the promoter and NotI/BamHI for the terminator. The PCR products are then digested with HindIII/NotI and NotI/BamHI, respectively, and ligated to pRS416 (or YCplac33) digested with HindIII/BamHI in a 3-piece ligation reaction. The ligation products are transformed into DH5α competent cells and candidate clones are selected on solid LB plus carbenicillin. More recently, we have adopted a modified version of the sequence and ligation independent cloning (SLIC) procedure (Li and Elledge, 2007) (Figure 2). This method does not require endonuclease digestion of the inserts and thus greatly simplifies primer designs and experimental procedures, especially when a large number of genes are processed simultaneously. This is also a relatively new cloning technique and is thus described below in greater detail.
Mutagenesis of YFG using error-prone PCR is performed essentially as described previously (Leung et al, 1989). Again, a genomic DNA sample of the wild-type yeast strain BY4743a/α is used as the DNA template. The PF and TR primers described above are used to amplify the full-length gene and ~500bp flanking sequences. TaKaRa Ex Taq (Cat# RR001A) or LA Taq (Cat# RR002B), which are ~4 times more accurate than the normal Taq polymerase, are used here due to their robustness. Induction of mutation rates is achieved by adding Mn2+ in the PCR at a final concentration of 10–150 μM that are arbitrarily defined, with higher concentrations for smaller genes (genes’ sizes range from 0.5–5kb).
Two μg of plasmid DNA of the promoter/terminator clone is digested with NotI (NEB, Cat# R0189) in NEB buffer 3 in a 20μl reaction. The reaction is incubated at 37°C for an overnight and subsequently at 65°C for 20 minutes to inactivate NotI. A small aliquot of the digestion product is examined by agarose gel electrophoresis to ensure complete digestion of the plasmid.
The mutagenized PCR products (approximately 10–20 μg in 200 μl) and linearized plasmid DNA of the promoter/terminator clone (~2 μg) are next combined and concentrated by ethanol precipitation.
The concentrated DNA sample of PCR products and the linearized vector is next transformed into the corresponding haploid-convertible heterozygous diploid yeast knockout mutant to create a mutagenized library of YFG.
The rest of the yeast transformation reaction is either incubated in 50 ml of fresh liquid SC–Ura at 30°C for 2 days to allow propagation of the library or can be sporulated immediately to convert the transformants into a library of haploid spores that harbor mutant alleles of YFG on a plasmid (see below).
After the titer of MATa G418R Ura+ haploid cells is determined, the library is screened for potential Ts mutants. We typically screen ~4,000 clones for each gene.
Candidate mutants are picked and re-streaked onto the same haploid selection media and re-tested for the Ts phenotypes by incubating at 25°C and 37°C. The plasmids are next recovered from those confirmed to be Ts in this initial assay and re-introduced individually into the same haploid-convertible heterozygous diploid mutant to test whether the Ts phenotype is linked to the plasmids.
The “diploid shuffle” chromosome method has now been used to systematically screen for missense mutations that result in temperature sensitive (Ts) alleles of hundreds of essential genes, with each allele directly integrated at its endogenous chromosomal location and flanked with the “barcodes” of the corresponding yeast knockout mutant (Fig 3; (Ben-Aroya et al., 2008), and unpublished data). First, YFG, including its promoter and terminator regions, is mutagenized with error-prone PCR (Fig. 3A). The mutagenized PCR product is next cloned into SB221+Topo-TA (Fig. 3B). This plasmid contains the URA3 gene flanked by the 5′ and 3′ regions of KanMX. The Topo-TA cloning site (Invitrogen) has been inserted in between the KanMX 5′ end and the URA3 gene. This site allows direct cloning of each of the PCR products, without the need for any further modifications. The result of the cloning step is a library of mutagenized YFG, which is then transformed into E. coli, and digested to release linear fragments (following DNA purification) (Fig. 3B). The linear fragments are directly transformed into the corresponding strain from the haploid-convertible heterozygous YFG/yfgΔ::kanMX diploid YKO collection by selecting for Ura+ transformants (Fig. 3C). The ~700bp KanMX5′ and KanMX3′ fragments direct the mutagenized YFG library into the yfgΔ::KanMX genomic locus via homologous recombination, with retention of the original bar codes flanking each gene (Fig. 3C). Pools of Ura+ cells containing the mutant alleles are sporulated (Fig. 3E). Spores thus formed are spread on a haploid selective medium and incubated at 25°C for colony formation. Only haploid MATa Ura+ spores containing the integrated mutant allele of YFG can grow on this medium. Colonies formed on selective medium are replica-plated and incubated at 37°C (Fig. 3F). Colonies growing at 25°C but not at 37°C are selected as potential Ts alleles, and retested. In summary, the final product of the diploid shuffle approach is a confirmed MATa strain from the YKO collection genetic background containing a URA3 marked Ts allele of a specific gene integrated into its endogenous locus and flanked by both barcodes. In addition, each strain contains a LEU2-MFA1pr-HIS3 reporter integrated at the CAN1 locus.
In addition to creating Ts alleles, the diploid shuffle-chromosomal method can also be used to transfer existing alleles into the knockout strain background from other strain backgrounds and vice versa. Using the Topo-TA plasmid and protocol described in Fig. 1, any extant Ts allele can be easily transferred to the deletion collection genetic background (referred to as “allele transfer-in”). The result is an integrated allele, marked by URA3, and flanked by the appropriate barcodes. Moreover, using the Topo-TA plasmid, any PCR product (mutant allele, fusion protein, heterologous gene expression cassette, etc.) can be introduced at any of the 6000 genomic sites carrying a KanMX replacement cassette as the integration site, depending on the specific deletion mutant chosen as the recipient strain. By using primers that are external to the primers used for the original mutagenesis, the URA3 marked Ts-allele or gene construct can be easily transferred from the deletion set genetic background to any other strain of interest, and replace the wild-type copy in the recipient strain by homologous recombination (referred to as “allele transfer-out”, see Fig. 4). Thus, each Ts mutation or gene construct can be analyzed for more specific phenotypes of interest in a variety of genetic contexts.
Haploid selection medium SC–Ura–Leu–His–Arg+Can: dextrose, 20g/L; yeast nitrogen base without amino acids and ammonium sulfate, 1.7g/L; SC-Ura-Leu-His-Arg dropout mix, 2g/L; sodium glutamate, 1g/L; L-canavanine, 60mg/L; Agar, 2%. YPD: yeast extract, 10g/L; peptone, 20g/L; dextrose, 20g/L; Agar, 2%. YPD+G418: yeast extract, 10g/L; peptone, 20g/L; dextrose, 20g/L; Agar, 2%; G418, 200 mg/L.
Diploid selection medium SC-Ura+ClonNAT: dextrose, 20 g/L; yeast nitrogen base without amino acids and ammonium sulfate, 1.7 g/L; SC–Ura dropout mix, 2g/L; sodium glutamate, 1g/L; ClonNAT, 200 mg/L; Agar, 2%. The sodium glutamate is substituted for ammonium sulfate as the nitrogen source and makes the ClonNAT selection more reliable on the minimal medium.
Others are the same as described in Part I.
The genotype of haploid-convertible heterozygous diploid strains are similar to those described in Part I. Haploid Ts strains have the following genotype: MATa ura3Δ0 leu2Δ0 his3Δ1 lys2Δ0 (or LYS2) met15Δ0 (or MET15) can1Δ::LEU2-MFA1pr::HIS3 yfg-ts::URA3.
BY4742-ade2101-NatMX is a MATα wild-type haploid strain in which the NatMX gene is linked to the ade2-101 ochre mutation. This strain is used to confirm the MATa Ura+ Ts candidate: MATα his3Δ1 leu2Δ0 lys2Δ0 ura3Δ0 ade2-101-NatMX
OneShot® TOP 10 Electrocomp Cells (Invitrogen Cat #C4040-52)
SB221 was derived from M4758 (Voth et al., 2003). In M4758, the BglII/KpnI and SphI/EcoRI fragments contain the TEF promoter (388bp) and terminator (262bp) respectively of Ashbya gossypii. In SB221 these fragments were replaced by a BamHI/KpnI PCR fragment (731bp) containing the TEF promoter plus half of the KanMX gene, and a SphI/EcoRI fragment (751bp) containing the other half of the KanMX gene and the TEF terminator. In both cases the template for PCR products was the KanMX gene used for constructing the heterozygous diploid collection. Finally the BamHI site of SB221, was adjusted with a Topo-TA site (invitrogen) to create SB221-Topo-TA.
This can be carried out as described in Part I. However, we have successfully used a slightly different condition for all the experiments using this diploid shuffle chromosomal method. Here we have exclusively used LA Taq DNA polymerase (Cat# RR002B) and 150 μM MnCl2. One PCR of 50 μl, instead of four, is normally set up for each gene. Two primers, which allow amplification of the entire coding region, 250–300bp of the 5′UTR, and 150–200bp of the 3′UTR of each gene, are used for PCR.
The mutagenized PCR products are purified and cloned into E. coli cells via electroporation.
Purified plasmid DNA sample of the random mutagenesis library is next digested with NotI or EcoRI (NEB, Cat#R0101T, 100u/μl) and transformed into the corresponding haploid-convertible heterozygous knockout diploid mutant. The mutant alleles will be integrated into the endogenous locus via homologous recombination.
Amplified yeast library is then sporulated to convert the heterozygous diploid into haploid spores similarly as described in Part I. However, a slightly different haploid selection medium, SC–Ura–Leu–His–Arg+Can, is used to determine the efficiency of producing haploid MATa Ura+ cells from the sporulation culture.
After the plating efficiency is determined, typically ~6000 haploid MATa Ura+ colonies are screened for candidate Ts alleles for each gene.
Candidate Ts mutants are re-streaked on two haploid selection media plates, and incubate at 25°C and 37°C. Once the Ts phenotype is confirmed, backcross the MATa Ura+ Ts candidate to a wild-type BY4742-ade2101-NatMX MATα strain. In this strain the NatMX gene (which provides resistance to the drug ClonNAT) is linked to the ade2-101 ochre mutation. Diploid cells are selected by streaking on SC-Ura+ClonNAT media, sporulated, and dissected. The dissected tetrads are replicated to YPD (25°C and 37°C), SC-Ura (25°C), and YPD supplemented with G418 (25°C). This confirms that: 1) the temperature sensitivity segregates in a Mendelian manner (2:2), and indicates that the Ts phenotype depends on a single mutated gene, 2) the Ts phenotype is linked to the URA3 gene, and therefore co-segregates with the mutated PCR product, 3) the mutagenized PCR product was integrated at the correct genomic locus rendering the cells G418 sensitive (5′kanMX::yfg-ts-URA3::3′kanMX).
This is carried out similarly as described above but with subtle modifications. A pre-existing Ts allele (or other gene construct) is first PCR-amplified with the appropriate plasmid or genomic DNA as the template using a proofreading-competent polymerase (e.g. LA Taq DNA polymerase) that generate an “A” overhang on each 3′ end of the PCR product. The PCR products are cloned into SB221 using Topo TA cloning. The cloned PCR products are then released together with the URA3 marker and the KanMX5′ and KanMX3′ fragments from the vector backbone and trans formed into the corresponding haploid convertible heterozygous diploid mutant. Ura+ yeast transformants are sporulated as a population and plated on SC–Ura–Leu–His–Arg+Can to select for haploid MATa Ura+ cells at an appropriate colony density. Single colonies are then tested for the phenotype of interest (e.g. temperature sensitivity). Candidate clones are then backcrossed to a wild-type MATα strain and analyzed with tetrad dissection to further confirm the phenotype.
Here we describe how to transfer a Ts allele generated by the diploid shuffle method to any other ura3 strain background. Unless otherwise stated, methodologies are as mentioned above.
The recipient strain can also be diploid. In this case, Ura+ transformants are selected after Step 3, sporulated, and characterized by tetrad analysis to ensure that the Ts and Ura+ phenotypes co-segregate. This will also allow testing whether the Ts allele is viable in the particular genetic context of the recipient strain. It is also possible that this allele is no longer Ts in this strain background. If so, representative Ura+ transformants will need to be characterized with diagnostic PCR or sequencing to ensure that the 5′kanMX::yfg-ts-URA3::3′kanMX cassette is indeed integrated at the right locus.
The two variations of the “diploid shuffle” are both highly efficient methods for making Ts mutants. It is essentially always possible, by screening enough mutants using the methods outlined here, to find such alleles. An adaptation of the methods outlined here will be required for making Ts mutants in very large essential genes (in this case, the gene would be mutagenized in sections). The relative advantages of the methods described here are summarized in Table 4.
We have chosen to move forward with the chromosomal method for the generation of a genome-wide collection of Ts mutants as a community resource because it was felt that most users would prefer integrated copies that would not fluctuate or be lost from a subpopulation of cells at each division due to their being on an episome.
The ability to generate a genome wide collection compares favorably with other attempts to generate genome wide resources for the study of essential genes, such as Tet-regulated alleles (Hartman et al., 2001; Mnaimneh et al., 2004) and dAMP (Schuldiner et al., 2005). Both of those approaches, while having their distinct advantages, only produced well-behaved alleles in about 30% of the cases. Disadvantages of Ts mutants include the fact they are not uniform with regard to nonpermissive temperature and leakiness, and the fact that the needed temperature shifts may induce heat shocks or other side effects that could potentially cloud phenotypic analyses. Nevertheless these alleles have been the bastion of traditional genetic analyses of essential genes. The Ts alleles we have sequenced include a mix of single amino acid substituions and multi amino acid substitutions; however, we have not sequenced enough of these to develop extensive statistics on this. Studies of collections of Ts mutants in a variety of genes have allowed the empirical determination of their typical characteristics. On this basis, a schemes for predicting Ts mutants has been developed (Ye, P., Dymond, J., Shi, X., Lin, Y.-Y., Pan, X., Boeke J.D. and Bader, J.S., submitted). This is likely to prove very useful for designing Ts mutants for use in other organisms in which extensive screening is impractical.