Copyright 0 1987 by the Genetics Society of America

A New Type of Fusion Analysis Applicable to Many Organisms: Protein Fusions to the URA3 of Yeast

Eric Alani and Nancy Kleckner

Department of Biochemistry and Molecular Biology, Harvard University, Cambridge, Massachusetts 02138 Manuscript received April 15, 1987 Accepted June 1, 1987

ABSTRACT We have made constructs that join the promoter sequences and a portion of the coding region of the HIS4 and GALI and the E. coli lac2 gene to the sixth codon of the S. cerevisiae URA3 gene (encodes orotidine-5’-phosphate(OMP) decarboxylase) to form three in frame protein fusions. In each case the fusion protein has OMP decarboxylase activity as assayed by complementation tests and this activity is properly regulated. A convenient cassette consisting of the URA3 segment plus some immediately proximal amino acids of HZS4C is available for making URA3 fusions to other proteins of interest. URA3 fusions offer several advantages over other systems for gene fusion analysis: the URA3 specified protein is small and cytosolic; genetic selections exist to identify mutants with either increased or decreased URA3 function in both yeast (S. cerevisiae and Schizosaccharomyces pombe) and bacteria (Escherichia coli and Salmonella typhimurium); and a sensitive OMP decarboxylase enzyme assay is available. Also, OMP decarboxylase activity is present in mammals, Drosophila and plants, so URA3 fusions may eventually be applicable in these other organisms as well.

ROTEIN fusions between a gene of interest and KORNBERGand SIMMONS1955; ROSE, GRISAFIand P a gene whose activity can be monitored geneti- BOTSTEIN 1984). (C) OMP decarboxylase is small cally in vivo and assayed biochemically in vitro have (monomer 25 kD) and active as a dimer, and may been used to investigate a wide range of biological therefore be more tractable for some purposes than problems. Fusions bearing some protein of interest at &galactosidase, which is large (monomer 1 16 kD) and the amino terminus and 6-galactosidase at the carboxy functions only as a tetramer (ROSE and BOTSTEIN terminus have been successfully used to characterize 1983; BRODYand WESTHEIMER1979; CONTAXISand regulatory elements, to identify intragenic signal ex- REITHEL1971). (D) The same enzymatic activity is port elements and to distinguish transcriptional from found in Escherichia coli, Drosophila, plants and higher post-transcriptional control (SILHAVYand BECKWITH animal cells, and nucleotide analogs have been used 1985; GUARENTEand PTASHNE1981; ROSE, CASA- in some of these organisms to select against OMP DABAN and BOTSTEIN1984). A number of other genes decarboxylase function (JONES1980). Therefore, the have been identified which retain functional activity possibility exists that the URA3 fusion cassette de- in protein fusions; these include gaLK, cat, npt, phoA, scribed here can be used for genetic analysis in a wide dhfr and uidA (SILHAVYand BECKWITH1985; STUE- range of organisms. BER et al. 1984; GORMAN,MOFFAT and HOWARD 1982; JEFFERSON,BURGESS and HIRSH1986). MATERIALS AND METHODS In this paper we describe protein fusions that con- Bacterial strains: MM294 (F-, endoA, hsdR, supE44, thiA) tain the GALI, HZS4, and lacZ gene products at the was the standard strain used for plasmid amplification and amino terminus and the Saccharomyces cerevisiae URA3 DNA manipulation (GUARENTEet al. 1980). KC-8 (rx1486, gene product, orotidine-5’-phosphate (OMP) decar- M+, K12,leuB600, trpC9830, pyrF::Tn5, hisB463, AlacX74, boxylase at the carboxy terminus. In each case we find strA, galU, galK) was provided by K. STRUHL.KC-8 trans- that OMP decarboxylase is functional and that OMP formed with pNK627 (relevant genotype tet, lacif, N. Kleck- ner laboratory collection) was used to assay the lacZ-URA3 decarboxylase expression is dependent on the regu- gene fusions. lacZ-URA3 gene fusion products were identi- latory elements of the gene to whose product it is tied in NK5830 (F’ lacpro, ladL8, Ara-, Met-, NalA, Rip, fused. The yeast URA3 gene product has a number of srl::TnlO, recA56, argEam, supE, AlacproxlII) (ROBERTSet features that makes it particularly attractive for fusion al. 1985). Yeast strains: DBY745 (MATa, adel-100, leu2-3, leu2- analysis: (A) There are direct genetic selections both 112, ura3-52) was obtained from D. BOTSTEIN.NKY280 for and against URA3 function (BOEKE,LACROUTE (MATa, adel-100, leu2-3, leu2-112, ura3-52, trp1::hisG) and FINK 1984). (B) An extremely sensitive assay for bears a 1.1-kb insertion of Salmonella hisG DNA in the OMP decarboxylase activity exists (LIEBERMAN, TRPl gene of DBY745. The construction of this strain is

Genetics 117: 5-12 (September, 1987) 6 E. Alani and N. Kleckner described elsewhere (ALANI,CAO and KLECKNER 1987). by PAULMCDONALD) (RUTHER and MULLER-HILL1983), TD50 (MATa, his4-619, ura3-52) was kindly provided by PRY116 (Gal1,lO upstream regions) (YOCUM et al. 1984), TOMDONAHUE (DONAHUE, FARABAUGH and FINK 1982). YEPl 3 (2p, LEU2) (SHERMAN,FINK and HICKS1982), B115 Media: E. coli were grown in LB broth or on LB agar. E. (URA3, distal HIS4 sequence-gift from T. DONAHUE), coli used for pyrF::Tn 10 complementation experiments were pBR322 (BOLIVARet al. 1977),and pGC-1 (M-13 sequencing grown in M9 salts supplemented with 0.2% glucose and vector) (MYERSand MANIATIS 1985). Further details are 0.004% each of histidine, tryptophan and leucine. Ampicil- available upon request. lin and tetracycline were supplemented at 100 q/ml and pNKY2009, 2024: pNKY2009 was built from four other 15 pglml, respectively. Yeast were grown in either YPD or plasmids. The backbone, pNKY2003, was made by deleting minimal selective media. Selective media contained 0.7% the URA3 gene from YEP24 with a partial HindIII digest. yeast nitrogen base (Difco), 2% agar and 2% glucose. Ad- A 4.0 kb Sphl fragment encoding HIS4 (isolated from a B54 dition of 0.004% leucine, tryptophan, adenine and plasmid molecule whose SstI site had been converted into was made according to strain requirements. In the galactose an SphI site) was inserted into the SphI site of pNKY2003 induction study, yeast were grown in media containing 2% to form pNKY2005. A 2.5-kb BglII fragment bearing LEU2 of each of the following carbon sources: galactose plus was isolated from YEPl 3 and inserted into the BamHl site sucrose, galactose plus glucose and galactose. 5-Fluoro-or- of pNKY2005 to form pNKY2010. Finally, a 1.5-kb BgllI- otic acid (5-FOA)was purchased from SCM Specialty Chem- BamHI fragment encoding the URA3 gene was isolated from icals, Gainsville, Florida. 5-FOA plates were prepared as a YEP24 derivative containing a BglII linker at the EcoRl described previously (BOEKE,LACROUTE and FINK 1984). site closest to the 5' end of URA3 and inserted into the BglIl Yeast transformation: Yeast lithium acetate transforma- site of pNKY2010 to form pNKY2009. BAL31 deletions tion were performed by standard methods (ITO et al. 1983). Integration of the HIS4-URA3 fusion into NKY280 was were performed at the BglII site of pNKY2009 to form the performed as follows: pNKY54 was digested with SphI and protein fusion vector pNKY2024 (Figure 1). the 4.5-kb fragment bearing the HIS4-URA3 fusion and the B184: B115, a pBR322 derived plasmid bearing the distal TRPl gene was introduced into NKY280 by homologous region of HIS4 was digested with BglII and integrated into recombination between the 5' and 3' ends of the fragment the HIS4C locus of TD50 (his4-619). Chromosomal DNA of and the chromosomal HIS4 locus. Sixteen Trp+ transform- this integrant was made, cleaved with BamHI, ligated and ants were tested for uracil and histidine prototrophy. Eleven transformed into MM294. Plasmids bearing the entire HIS4 of these candidates were His-, suggesting that a homologous locus (12 kb) were identified by restriction mapping. recombination event occurred between the HlS4-URA3, pNKY2028: pNKY2028 was made by replacing the 1.4- TRPZ fragment and the chromosomal HIS4 locus. Ten of kb SphI-XhoI fragment which spans the HIS4 initiation co- these 1 1 His- candidates were Ura+, sensitive to 5-FOA and don in pNKY2024 with the corresponding fragment from yielded similar colony sizes on media with and without B184. uracil. Genomic DNA from Trp+,Ura+, His- transformants pNKY54: pNKY2024 was digested with SmaI and BstE2 was digested with PstI and blotted with a nick translated and a BglIl 8-bp linker was inserted to form pNKY2029. HIS4 probe. NKY280 displayed a IO-kb band while 6.3- and After a BglII linker was inserted into the EcoRI site of YRP7 3.7-kb bands were detected in the strains containing the closest to the 5' end of the TRPl, an 850-bp BamHI-BglII HZS4-URA3 integrant. This result confirmed that the ob- fragment containing an intact TRPl gene was isolated and served phenotype resulted from a single copy integration inserted into the BglIl site of pNKY2029. An Sphl fragment event within the HIS4 locus (blot not shown). from pNKY2029 which encodes the HIS44JRA3 fusion and Protein gels: A 1.0-ml culture of NK5830 transformed the TRPl gene was isolated and inserted into the Sphl site with the 1ac.Z-URA3 protein fusion plasmid (pNKY59) was of pBR322 to form pNKY54. grown to exponential phase, treated with 1 mM isopropyl p- pNKY48: In order to sequence the HIS4-URA3 fusion D-thiogalactopyranoside (IPTG) and grown for 2 hr more. junction, a HindIlI fragment from pNKY2029, which en- Cells were then suspended in 50 pl of STET (8% sucrose, codes 93 amino acids of HIS4 fused to 267 amino acids of 50 mM EDTA, 50 mM Tris, 5% Triton X-100, pH 8.0); and UKAS, was isolated and inserted into the Hind111 site of treated with 90 fig of lysozyme for 5 min on ice. After an pNKY47. The sequencing primer in pNKY48 is 290 bp equal volume of 0.1% SDS loading buffer was added, the from the HIS4-URA3 fusion junction. The backbone for the samples were boiled for 1.5 min and then applied to a 5% construct, pNKY47, is a derivative of pGC1. To make Laemmli gel. Gel was prepared and run according to stan- pNKY47, the existing Hind111 site in pCCl was destroyed dard procedures and stained with Coomassie brilliant blue by a T4 polymerase filling in reaction and a new Hind111 R250 (LAEMMLI1970). site was created in the polylinker by inserting an 8-bp Nucleic acid techniques: A11 restriction enzymes, BAL3 1, T4 DNA ligase and T4 DNA polymerase were purchased HindIII linker into the BamHI site (filled in with T4 DNA from New England Biolabs and used according to manufac- polymerase, BamHI site regenerated). turers specifications. Plasmid DNA was isolated by a cleared pNKY59,60: pNKY48 was digested with HindIlI and the lysate protocol and DNA manipulations were described 1.1-kb fragment bearing the URA3 fusion cassette was iso- previously (MANIATIS,FRITSCH and SAMBROOK1982). Di- lated and inserted into the HindIII site of pUR290. The in- deoxy sequencing was described previously (MESSING1983). frame construct is pNKY59 and the opposite orientation The sequencing primer for the pGC sequencing vectors was construct is pNKY6O. generously provided by JIM LILLIE. SOUTHERN(1 975) blot- pNKY1069: An EcoRl octamer linker was inserted into ting was performed by following standard procedures. the HindIII site of pNKY48 closest to the primer to form Plasmid constructions: All constructions described below pNKY70. pNKY70 was then digested with EcoRI and the were built from one or more of the following nine vectors 1. I-kb HZS4C-UKA3 EcoRI fragment was inserted into the (relevant genotypes or properties): YRP7 (AKSI, CENl, EcoRI site of PRY1 16. The fusion construct (pNKY71) was TRPZ) (SHERMAN,FINK and HICKS 1982). YEP24 (2p, selected and a 1.9-kb BamHI-Sal1 fragment from pNKY71 lJRA3) (BOTSTEINet al. 1979), B54 (HZS4) (DONAHUE,FAR- (bearing the CALI-URA3 fusion) was inserted into the Sal1 ABAUGH and FINK1982), pUR290 (lacPOZ, kindly provided and BamHI sites of YRP7 to form pNKY1069. Yeast URA3 Protein Fusions 7

0 2022 ------+ 1 8gl11DIGEST 1 8g/ll DIGEST 2 BAL31 DIGEST 2 BAL31 DIGEST 3 LIGATE WITH 3 LIGATE 8glll LINKER r 1 \- HIS 4C URA3 es\\‘ GTC GCA TAT AAG 530 531 6 7 LFUSIONJUNCTION 1 FIGURE 1.-BAL3 1 strategy used to make HIS4-URA3 protein fusions. Insert shows DNA sequence of the fusion junction as determined by dideoxy sequencing. P = PstI.

RESULTS desired construct and complemented the ura3-52 mu- tation upon retransformation into DBY745. DNA Construction of HZS4-URA3 gene fusion: Our goal sequencing of the fusion junction of one candidate, in this study was to first make URA3 gene fusions that pNKY2024 (parental plasmid for all subsequent ex- retained OMP decarboxylase activity and then to con- periments), revealed that the 6th codon of URA3 had struct a URA3 cassette that could be inserted into any been fused to the 531st codon of HIS4. gene of interest to create a fusion protein that ex- Expression of the URA3 gene in the HIS4-URA3 presses OMP decarboxylase. The URA3 gene, which fusion plasmid is dependent on translation initiation encodes a 267 amino acid protein, has been cloned at HZS4: a mutation in the HIS4 ATG start codon and sequenced on an 1100-bp DNA segment (ROSE, eliminates URA3 function. The HIS4 initiation codon GRISAFIand BOTSTEIN1984). Our strategy was to in the HIS4-URA3 fusion was replaced by his4-619 to make a collection of random deletions at the 5’ end result in an isogenic plasmid that bears an ATA sub- of URA3 and fuse these deletions to the 3’ end of a stitution for the HIS4 initiation codon. When the well-characterized gene and then use this fusion as a parent plasmid pNKY2024 (HIS4-URA3, 2p) was in- cassette to make other URA3 fusions. troduced into DBY745, the resulting transformants As an initial step toward constructing a URA3 fusion were uracil prototrophs and were sensitive to 5-FOA. cassette, we chose to make fusions of URA3 to the 5-FOA is a substrate analog which is toxic when acted yeast HIS4 gene. The HIS4 gene encodes a single upon by OMP decarboxylase. When the his4-619- translation product with A, B and C domains. Each domain is responsible for a separate biosynthetic step ura?, 2p plasmid (pNKY2028) was introduced into in histidine biosynthesis (DONAHUE,FARABAUGH and DBY745, the resulting transformants were unable to FINK 1982). HIS4-URA3 fusions which express OMP complement the ura3-52 mutation and were resistant decarboxylase were built in two steps (Figure 1). An to 5-FOA. intact URA3 gene was inserted into a plasmid bearing The characterization of HIS4-URA3 fusion con- the C domain of HIS4 and an initial BAL3 1 treatment structs described involved a 2p plasmid which is pres- was performed to eliminate the URA3 start signals ent in high copy. To assess more sensitively the OMP without deleting into the coding sequence. A second decarboxylase activity of the fusion protein, we ex- BAL3 1 treatment was performed on one appropriate amined the URA3 expression of a single copy HIS4- deletion variant in such a way as to simultaneously URA3 construct that had been integrated into the delete the URA3 start codon and fuse the remaining yeast genome at the HIS4 locus. We found that the codons of URA3 to HZS4C. A population of HIS4- single copy HIS4-URA3 fusion still complements the URA3 hybrid plasmid molecules resulting from a sim- ura3-52 mutation, and furthermore, cells bearing this ple BAL3 1 digestion and religation were transformed single copy fusion are still sensitive to 5-FOA. These into the yeast strain DBY745 (relevant genotype leu2- results suggest that the HIS4-URA3 fusion protein 3,leu2-112, ura3-52). Out of 3000 transformants that probably retains a reasonably high specific activity. received the LEU2 marked fusion plasmids, 28 com- URA3 protein fusions to other genes: The above plemented the ura3-52 mutation. Restriction mapping results demonstrate that HIS4-URA3 protein fusions of these 28 revealed that only two candidates both can be made which express URA3 function. To show displayed the restriction pattern consistent with the that the OMP decarboxylase is active in a URA3 fusion 8 E. Alani and N. Kleckner

SIP sp 1 HIS4 A TRPl ‘C pNKY54 I I I I 0 2.4 3.2 4.6 KB

H H FIGURE2. -Plasmids bearing po lac 2 HIS4-lJRA3, lacZ-URA3, and GALl- pNKY59 ~ URA? protein fusions. Construction J I I steps are described in MATERIALS AND I KB 0 3.2 3.5 4!5 METHODS. The extent of the open reading frame encoding each fusion protein is indicated by the solid bar. Sp = Sphl, H = HindIII, S = SalI, R = EcoRI, B = BamHI. HIS4C bridge

pNKY 1069

that bears some gene product of interest at the amino leakiness of the pyrF::TnS insertion. At levels of IPTG terminus, we made LacZ-URA3 and GALI-URA3 gene greater than 0.1 mM, the pyrF::Tn5; cells bearing the fusions (Figure 2) using a cassette derived from the in-frame fusions were able to grow on M9 media HIS4-URA3 fusion (Figure 3). Both URA3 fusions con- lacking uracil. The opposite orientation construct tain a 93 codon bridge of HIS4 sequence proximal to failed to complement the fyrF mutant in all cases. the URA3 cassette because a restriction site is not Thus the above results suggest that URA3 expression present at the exact fusion junction between HZS4C is dependent on the lac promoter. Also, the lacZ- and URA3. URA3 fusion transformants of pNKY59 grew as bright It has been previously shown that the yeast URA3 blue colonies on XG (5-bromo-4-chloro-3-indolyl-~-~- gene product can complement E. coli strains that are galactoside) plates with and without IPTG, suggesting deficient in OMP decarboxylase (pyrF::TnS) (ROSE, that the fusion also expresses P-galactosidase activity. GRISAFIand BOTSTEIN1984). We constructed lacZ- The OMP decarboxylase and P-galactosidase activities URA3 protein fusions that express OMP decarboxyl- in fusion proteins derived from pNKY59 are present ase in E. coli in response to IPTG induction. When in a single physical fusion polypeptide. The lacZ-URA3 induced, this fusion complements E. coli PyrF mutants. fusion construct was introduced into NK5830 (rele- Also, an intact fusion protein having the size predicted vant genotype Eaciq).Exponentially growing cells were of a continuous LacZ-URA3 encoded polypeptide was induced with IPTG for two hours and total cell pro- detected on Laemmli gels. tein was examined on a 5% Laemmli gel. As shown in We inserted the URA3 fusion cassette in both ori- Figure 4, a 150-kD protein which matches the size entations at the 3’ end (polylinker site) of the lacZ predicted by the lacZ-URA3 fusion protein was specif- gene of pUR290. pUR290 bears the lac2 gene under ically induced by IPTG. lac promoter and operator control and this vector is As a second example to assess the general applica- designed to generate fusions to the carboxy terminus tions of the URA3 fusion cassette, we constructed a of 1acZ. The URA3 cassette inserted into pUR290 in GALI-URA3 gene fusion that encodes a 385 amino the correct orientation is predicted to specify a 150- acid polypeptide (Figure 2) consisting of 28 amino kd polypeptide (pNKY59). The cassette insertion in acids derived from the GALl gene product and the the opposite orientation serves as a negative control remaining 357 amino acids from the URA3 fusion (pNKY6O). Both constructs were introduced into the cassette. This fusion was inserted into a high copy E. coli strain KC-8 bearing pNK627 (relevant geno- ARS vector and transformed into the yeast strain type pyrF::Tn5, AlacX74, laciq). As shown in Table 1, NKY280 (relevant genotype ura3-52, trp2-). only background growth was observed for cells grown Transcription of the wild-type GALl gene is acti- without IPTG. We believe that this growth is due to vated by galactose and repressed to undetectable lev- Yeast URA3 Protein Fusions 9

H B 'ec C N X E B

TC TAG AGG ATC CAA GCT TTG

I t wk3NElDw JucTlaw

PNUY 70: GGA ATT CCA GCT GLY ILE PRO ALA @KY 48 + EGQ RI LINKER) FIGURE3.-URA? fusion cassette. Construction of pNKY48 and pNKY70 are described in MATERIALS AND METHODS. H = HzndIII, B = BamHI, Bg = BglII, N = NcoI, E = EcoRI, C = ClaI, Xb = XbaI, X = XhoI, S = SalI.

TABLE 1 Average colony diameter (mm) of E. coli strain KC-8 (pyrF) bearing NK627 (laeig) and lacPOZ-URA3 fusion plasmids

Minimal media Minimal media + uracil Plasmid IPTC concentration (mM): 0 10-2 10-1 100 0 lo-* 10-1 100 pUR290 (lacPOZ) 0.6 0.6 0.6 0.6 2.1 2.1 2.1 2.1 YEP24 (URA3) 2.1 2.1 2.1 2.1 2.1 2.1 2.1 2.1 pNKY59 (EacPOZ-URA?) 0.6 0.6 1.0 2.1 2.1 2.1 2.1 2.1 pNKY6O (l~cP02-1~~~3) 0.6 0.6 0.6 0.6 2.1 2.1 2.1 2.1 Fresh single colony transformants of E. coli KC-8 were diluted in 0.85% NaCl and plated onto petri dishes containing M9 media, plus various concentrations of IPTG. Approximately 30 colonies were plated onto each dish and the efficiency of plating was 1.0. Colony diameter was determined after 3 days growth at 37'. els in glucose (YOCUMet al. 1984). As shown in Table fusions of URA3 product to other gene products will 2, the high copy GALl-URA3 vector complements the behave in a similar fashion. ura3-52 mutation only when the cells are grown in Advantages and applications of URA3 fusion galactose or galactose plus sucrose. The galactose plus analysis: The yeast URA3 gene product has a number sucrose carbon source was used because NKY280 of features which make it particularly attractive for grow better on this carbon source mixture and this fusion analysis. First and most important, there are combination decreases galactose induction only direct genetic selections both for and against URA3 slightly (HASHIMOTOet al. 1983). The GALl-URA3 function. Increase in function is selected by requiring fusion could not complement the urd-52 mutation growth of an OMP decarboxylase deficient mutant when the transformants were grown on media con- on defined medium in the absence of uracil. Decrease taining equal concentrations of galactose and glu- or absence of function is selected by requiring growth cose, a carbon source mixture which has been previ- of such a mutant in the presence of both uracil and ously shown to lower GAL1 expression one hundred 5-FOA. Second, there exists a sensitive assay for fold (YOCUMet al. 1984). OMP decarboxylase activity: release of I4CO2 from [ ''C]orotidine 5'-phosphate (LIEBERMAN,KORNBERG DISCUSSION and SIMMONS1955; ROSE, GRISAFIand BOTSTEIN 1984). Third, the native enzyme has several features We have constructed hybrid proteins that contain that help make derived fusion proteins especially a bacterial or yeast gene product at the amino ter- tractable: the monomer polypeptide is small (27 kD) minus and the URA3 gene of S. cerevisiae at the and the active form of the enzyme is a simple dimer carboxy terminus. A convenient URA3 cassette has and is cytosolic (BRODYand WESTHEIMER 1979; been constructed for use in generating URA3 fusions JONES 1980). to other genes of interest. Since all three of the Gene fusion analysis is frequently applied to anal- hybrid proteins described here show URA3 specified ysis of gene regulation. The URA3 fusions are clearly OMP decarboxylase activity, it seems likely that most suitable for this application, since expression of OMP 10 E. Alani and N. Kleckner TABLE 4 Average colony diamete~(mm) of yeast strain NKYSSO (uro3-52, ttpZ) bearing CALZ-URAJ fusion plasmids

Minimal media Minimal media + uracil GLU GAL GLU GAL Plasmid Carbonsource: GLU GAL GAL SUC GLU GAL GAL SUC YRP7 (TRPI) None None None None 2.5 2.5 0.7 2.5 pNKY1069 (GALI-URA3, TRPI) None None 0.5 2.5 2.5 2.5 0.7 2.5 Fresh single colony transformants of S. ccrcvisiaC NKY280 were diluted in water and plated onto yeast minimal media supplemented with the carbon sources shown above. Approximately 30 colonies were plated onto each dish and the efficiency of plating was 1.0 in all cases where colonies were visible. Colony diameter was measured after 5 days growth at 30'. pNKY 5 9 PUR 290 an average of fewer than four (active) monomers per cell per generatidn (for example, RALEIGH and + -+- +-+ - s KLECKNER1986). Only a minority of cells which by chance contain at least four monomers contribute to the measured enzymatic activity, a problem which lSO-+ KD can lead to underestimation of expression levels and KD to overestimation of induction ratios when gene expression is regulated. This problem should be less severe for URA3 fusions, where only two monomers 0-116 per cell are required for an active (dimeric) molecule (BRODYand WESTHEIMER1979). Most important for gene expression analysis, the -92 positive and negative genetic selections can be used to search for linked and unlinked regulatory muta- tions and for structural gene mutations that increase or decrease expression under conditions of interest. The one significant disadvantage of the URA3 fusion system is the absence of any colorimetric colony assay -68 for OMP decarboxylase function. FIGURE4.-IPTG dependent induction of a 150-kD protein in an E. coli strain bearing lacZ-URA3 fusion plasmids. NK5830 (loci9 Gene fusion analysis has also been applied effec- transformed with either pUR290 (IncPOZ) or pNKY59 (lacPOZ- tively to the analysis of protein localization and se- URA3) were grown to exponential phase and induced with 1 mM cretion, both in identifying and following the loca- IPTG (see MATERIALS AND METHODS).Total cell protein was run tions of particular proteins, and for selecting mutants on a 5% Laemmli gel and stained with Coomassie brilliant blue R250. The induced 1 16-kD polypeptide observed in NK5830 bear- that are altered in their localization or secretion ing pUR290 corresponds to the @-galactosidaseprotein in the behavior. URA3 fusions should be readily applicable molecular weight standard mix (S). + = IPTG induction, - = no to this type of analysis as well. It should be straight- IPTG induction. forward to raise antibodies to yeast OMP decarbox- ylase because the enzyme can be purified to homo- decarboxylase activity is properly controlled by up- geneity in only two chromatographic steps (BRODY stream regulatory regions in all three of the fusions and WESTHEIMER1979). The small size of the URA3 described above. Enzymatic assays of cells containing fusion polypeptide may make it particularly advan- a protein fusion can be used to quantitate changes in tageous for secretion and localization studies, which the level of a gene product, and with some assump- can sometimes be hampered by the failure of a large tions about the specific activity of the fusion protein, molecule such as p-galactosidase to be properly han- to obtain a rough idea about the absolute level of the dled. product present in the cell. The dimeric nature of Finally, gene fusions are often used as a tool for the active enzyme make URA3 fusions more attrac- both purifying proteins and generating antibodies to tive than the popular lac2 fusion constructs for this polypeptide sequences of interest. URA3 fusions can type of analysis. In yeast especially, genes of interest be used for this purpose. A highly specific affinity are often expressed at such low levels that protein column based on a competitive inhibitor of OMP fusions to /3-galactosidase do not express a detectable decarboxylase (6-azauridine 5'-phosphate) has been level of enzyme in standard ONPG assays (for ex- used to purify the native protein and should be ample, HUISMANet al. 1987). Also, in both yeast and equally applicable to the purification of fusion pro- in bacteria, measured &galactosidase activity is arti- teins (BRODYand WESTHEIMER1979). ficially low in whole cell assays if cells are producing General applicability of URA3 fusion analysis to Yeast URA3 Protein Fusions 11 many organisms: We have described above the anal- recombinant DNA experiments. Gene 8: 17-24. ysis of protein fusions in two organisms, the BRODY,R. S. and F. H. WESTHEIMER,1979 The purification of URA3 orotidine-5‘-phosphate decarboxylase from yeast by affinity bacterium E. coli and the yeast S. cerevisiae. The chromatography. J. Biol. Chem. 254 4238-4244. analysis described above can be applied essentially CONTAXIS,C. C. and F. J. REITHEL,1971 Studies on protein without modification to other bacteria; (Salmonella multimers. Biochem. J. 124: 623-632. typhimurium) and another yeast (Schizosaccharomyces DONAHUE,T. F., P. J. FARABAUCHand G. R. FINK,1982 The pombe). In both of these organisms mutants deficient nucleotide sequence of the HIS4 region of yeast. Gene 18: 47-59. in OMP decarboxylase have been identified and such GORMAN,C. M., L. F. MOFFATT and B. H. HOWARD, mutants can be positively selected with 5-FOA 1982 Recombinant genomes which express chlorampheni- (BOEKE,LACROUTE and FINK1984; G. FINK,personal col acetyltransferase in mammalian cells. Mol. Cell. Biol. 2: communication). 1044-1051. It seems likely that fusion analysis can even- GUARENTE,L. and M. PTASHNE,1981 Fusion of Escherichia coli URA3 lacZ to the cytochrome c gene of Saccharomyces cerevisiae. tually be applied to many other types of organisms. Proc. Natl. Acad. Sci. USA 78: 2199-2203. OMP decarboxylase activity is universal; it has been GUARENTE,L., G. LAURER,T. ROBERTSand M. PTASHNE, identified in Drosophila, mammals and plants (JONES 1980 Improved methods for maximizing expression of a 1980; G. FINK, personal communication). Also, mu- cloned gene: a bacterium that synthesizes rabbit @-globin.Cell tations which either enhance or decrease OMP de- 20: 543-553. HASHIMOTO,H., Y. KIKUCHI,Y. NOGI and T. FUKASAWA, carboxylase activity have been identified in many of 1983 Regulation of expression of the galactose gene cluster these organisms. In mouse tissue culture, for exam- in Saccharomyces cereuisiae. Isolation and characterization of ple, fluorouracil has been used to select mutants the regulatory gene GAL4. Mol. Gen. Genet. 191: 31-38. which have reduced levels of OMP decarboxylase HUISMAN,O., W. RAYMOND,K. FROEHLICH,P. ERRADA,N. activity (LEVINSON,ULLMAN and MARTIN1979). The KLECKNER,D. BOTSTEINand M. A. HOYT, 1987 A TnIO- lacZ-ha&-URA3 gene fusion transposon for insertion muta- URA3 gene should be equivalent to other reporter genesis and fusion analysis of yeast and bacterial genes. Ge- genes such as cat, dhfr, and luciferase for direct netics 116 191-199. analysis of transcript formation (STUEBERet al. 1984; ITO, H., Y. FUKUDA, K. MURATA and A. KIMURA, GORMAN,MOFFAT and HOWARD1982; OW et al. 1983 Transformation of intact yeast cells treated with alkali 1986). In organisms where suitable mutants lacking cations. J. Bacteriol. 153: 163-168. JEFFERSON, R. A., S. BURGESSand D. HIRSH, 1986 @-Glucuron- endogenous OMP decarboxylase activity can be iso- idase from Escherichia coli as a gene-fusion marker. Proc. lated, URA3 should also be as useful as other systems Natl. Acad. Sci. USA 83: 8447-8451. for assaying transient expression of the gene product JONES, M. E., 1980 nucleotide biosynthesis in ani- itself. The most unique potential advantage of the mals: genes, enzymes and regulation of UMP biosynthesis. system would be its applicability to genetic Annu. Rev. Biochem. 49: 253-279. URA3 LAEMMLI,U. K., 1970 Cleavage of structural proteins during analysis for the selection of mutations that either the assembly of the head of bacteriophage T4. Nature 227: increase or decrease the expression of OMP decar- 680-685. boxylase. LEVINSON,B. B., B. ULLMANand D. W. MARTIN, 1979 Pyrimidine pathway variants of cultured mouse lymphoma We are grateful to DAVIDBOTSTEIN for pointing out many of cells with altered levels of both orotate phosphoribosyltrans- the advantages of URA3 fusion analysis and for encouraging us to ferase and orotidylate decarboxylase. J. Biol. Chem. 254: develop the protein fusion cassette, to PATRICKERRADA, PAUL 4396-440 1. MCDONALD,RICHARD LYNN and JIM LILLIEfor advice on DNA LIEBERMAN,I., A. KORNBERC and E. S. SIMMONS,1955 sequencing and fusion protein detection, and to ANNAFERRI for Enzymatic synthesis of pyrimidine nucleotides, orotidine-5’- manuscript preparation. E.A. was funded by a National Institutes phosphate and -5’-phosphate. J. Biol. Chem. 215: of Health Genetics Training Grant awarded to the Department 403-415. of Biochemistry and Molecular Biology. MANIATIS,T., E. FRITSCHand J. SAMBROOK,1982 Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Press, LITERATURE CITED Cold Spring Harbor, N.Y. MESSING,J., 1983 New M13 vectors for cloning. Methods En- ALANI,E., L. CAO,and N. KLECKNER,1987 A method for gene zymol. 101: 20-78. disruption that allows repeated use of URA3 selection in the MYERS,R. and T. MANIATIS,1985 A general method for satu- construction of multiply disrupted yeast strains. Genetics 116: ration mutagenesis of cloned DNA fragments. Science 229 541-545. 242-247. BOEKE,J. D., F. LACROUTEand G. R. FINK, 1984 A positive Ow, D. W., K. V. WOOD,M. DELUCA,J. R. DEWET,D. R. HELINSKI selection for mutants lacking orotidine-5’-phosphate decar- and S. H. HOWELL,1986 Transient and stable expression boxylase activity in yeast: 5-fluoro-orotic acid resistance. Mol. of the firefly luciferase gene in plant cells and transgenic Gen. Genet. 197: 345-346. plants. Science 234: 856-859. BOLIVAR,F., R. L. RODRIGUEZ,P. J. GREENE,M. C. BETLACH,H. RALEIGH,E. A. and N. KLECKNER,1986 Quantitation of inser- L. HEYNEKER,H. W. BOYER,J. H. CROSAand S. FALKOW, tion sequence IS10 transposase gene expression by a method 1977 Construction and characterization of new cloning ve- generally applicable to any rarely expressed gene. Proc. Natl. hicles. 11. A multipurpose cloning system. Gene 2: 95-1 13. Acad. Sci. USA 83: 1787-1791. BOTSTEIN,D., S. FALCO,S. STEWART,H. BRENNAN,S. SCHERER, ROBERTS,D., B. C. HOOPES,W. R. MCCLUREand N. KLECKNER, D. STINCHCOMB,K. STRUHLand R. DAVIS,1979 Sterile host 1985 IS10 transposition is regulated by DNA adenine meth- yeast (SHY) a eukaryotic system of biological containment for ylation. Cell 43: 1 17-1 30. 12 E. Alani and N. Kleckner

ROSE.M. and D. BOTSTEIN,1983 Structure and function of the SILHAVY,T. J. and J. R. BECKWITH,1985 Uses of lac fusions for yeast URA3 gene. Differentially regulated expression of hy- the study of biological problems. Microbiol. Rev. 49 398- brid 8-galactosidase from overlapping coding sequences in 41 8. E. 1975 of yeast. J. Mol. Biol. 170 883-904. SOUTHERN, M., Detection specific sequences among DNA fragments separated by gel electrophoresis.J. Mol. Biol. ROSE, M., M. J. CASADABANand D. BOTSTEIN,1984 Yeast genes 98: 503-517. fused to 8-galactosidase in Escherichia coli can be expressed STUEBER,D., I. IBRAHIMI,D. CUTLER,B. DOBBERSTEINand H. normally in yeast. Proc. Natl. Acad. Sci. USA 78: 2460-2464. BUJARD,1984 Novel in vitro transcription-translation sys- ROSE, M., P. GRISAFIand D. BOTSTEIN,1984 Structure and tem: accurate and efficient synthesis of single proteins from function of the yeast URA3 gene; expression in Escherichia cloned DNA sequences. EMBO J. 3: 3143-3148. coli. Gene 29: 113-1 24. YOCUM,R., S. HANLEY,R. WESTand M. PTASHNE,1984 Use of RUTHER,U. and B. MULLER-HILL,1983 Easy identification of lacZ fusions to delimit regulatory elements of the inducible cDNA clones. EMBO J. 2: 1791-1794. divergent CALI-GAL10 promoter in Saccharomyces cerevisiae. SHERMAN,F., G. R. FINKand J. HICKS,1982 Methods in Yeast Mol. Cell. Biol. 4: 1985-1998. Genetics. Cold Spring Harbor Press, Cold Spring Harbor, N.Y. Communicating editor: D. BOTSTEIN