DEVELOPMENTAL GENETICS 11:410417 (1990)

Nonsense Suppression in Dictyostelium discoideum THEODOR DINGERMANN, NORBERT REINDL, THOMAS BRECHNER, HERBERT WERNER, AND KATHENERKE Znstitut fur Biochemie der Medizinischen Fakultat, Uniuersitat Erlangen-Nurnberg, Fahrstrasse 17,0-8520 Erlangen, Federal Republic of Germany ABSTRACT We describe the generation of Dictyosteliurn discoideurn cell lines that carry different suppressor tRNA genes. These genes were constructed by primer-directed mutagenesis changing a tRNATrP(CCA)gene from D. discoideurn to a tRNATrP(amber)gene and changing a tRNAG1"(UUC)gene from D. discoideum to o tRNAGl"(ochre) as well as a tRNAGl"(amber) gene. These genes were stably integrated into the D. discoideurn genome together with a reporter gene. An actin 6::lacZgene fusion carrying corresponding translational stop signals served as a reported. Active P-galactosidase is expressed only in D. discoideum strains that contain, in addition to the reporter, a functional suppressor tRNA. Both amber suppressors are active in D. discoideurn without interfering significantly with cell growth and development. We failed, however, to establish cell lines containing a functional tRNAG"'(ochre) suppressor. This may be due to the fact that nearly every message from D. discoideurn known s o far terminates with UAA. Therefore a tRNA capable of reading this termination codon may not be compatible with cell growth.

Certainly for Escherichia coli and yeast the availability of a significant collection of defined suppressors was crucial for the rapid development of genetics in these systems, and nonsense suppression a s a general conditional-lethal system will be of great value for higher eukaryotes a s well. With the isolation of tRNA genes and the development of site-specific mutagenesis, it became possible to generate efficiently tRNA genes that code for nonsense suppressors. Cell lines can be established that carry such genes stably integrated into their genomes. Over the past, we have concentrated in isolating and characterizing tRNA genes from the cellular slime mold Dictyostelium discoideum [Dingermann et al., 1985, 1986, 1988a,b]. D. discoideum has a well established genetic system [Loomis, 19821, but suppressor genetics was not available yet. Here we present such a system. Different tRNA genes were converted to amber and ochre suppressors by primer directed mutagenesis. These were stably integrated into the D. discoideum genome together with mutated versions of the lacZ gene from E . coli. This gene can be expressed in D. discoideum controlled by a homologous actin 6 promoter [Dingermann et al., 1989bl. Amber suppression is fairly efficient, and suppressor tRNA genes seem not Key words: Suppressor tRNA genes, east, lacZ to interfere with cell growth. We failed, however, reexpression in D. discoideurn, tRNAY;p genes, peatedly to establish strains carrying ochre supprestRNAG1"genes sors, suggesting that such tRNAs cause lethality for D. discoideum. MATERIALS AND METHODS Transformation INTRODUCTION D. discoideum Ax-2 cells used for transformation Nonsense suppressors are defined as tRNAs that in- were routinely grown in HL5 medium [Watts and Ashsert amino acids at positions in a message containing a worth, 19701 a t 22°C. Transformants were obtained as translational stop codon. Nonsense suppressor tRNA described [Nellen et al., 1984; Early and Williams, genes have been extensively studied in bacteria [Celis 19871 using pNeo-IT (corresponds to A15T) or pDNeo-2 and Smith, 1979; Steege and SOH, 19791but also in the eukaryotes Saccharomyces cerevisiae [Sherman, 19821 and Schizzosaccharomyces pombe [Munz et al., 1983; Hottinger et al., 19821. Nonsense suppressors were isolated also from Caenorhabditis elegans [Wills et al., Received for publication July 26, 1990. 1983; Hodgkin, 1985; Fire, 19861, and recently suppresAddress reprint requests to Dr. Th. Dingermann, Institut fur Biosor tRNA genes were expressed in Drosophila melano- chemie, Med. Fak., Fahrstrasse 17, D-8520 Erlangen, West Germany. gaster [Doerig et al., 19881 and in human cell lines [Laski et al., 1982; Temple et al., 1982; Laski et al., Thomas Brechner is now a t the Department of Immunology, Research 1984; Capone et al., 19851. Institute of Scripps Clinic, La Jolla, CA 92037.

0 1990 WILEY-LISS, INC.

NONSENSE SUPPRESSION IN D. DISCOIDEUM

411

TABLE 1. Synthetic Oligonucleotides*

Name lacZ(UAG) ZacZ(UAA) gZu(ochre)

glu(amber)

trp(CCA)iG trp(amber)

Seauence

Stringent hybridization temDerature

5' -GAATTCCCTAACCATCCA-3' 5'-GAATTCCpAACCATCCA-3' 5' -CCAGTGnAAAGACTAG-3' 5'4CAGTGnAGAGACTAG-3' 5' -CTAACTTCCAACGCTTTTG-3 ' 5'-CAACGTTTTAGAGACAA-3'

52°C 50°C 48°C 50°C 54°C 46°C

*The nucleotides introducing mutations are underlined.

[Cohen et al., 1986; Knecht et al., 1986; Dingermann et al., 1989133 as selection plasmid. Plasmids carrying the lacZ reporter gene or the suppressor tRNA gene did not contain selectable markers. Routinely 5 pg of each plasmid DNA was used for cotransformation. Primary colonies were picked from the Petri dish and recloned on SM-agar plates with Klebsiella aerogenes [Welker and Williams, 19821. After 2 days of growth a t 22"C, single colonies were transferred to axenic medium and subsequently tested for P-galactosidase activity. Yeast transformation was done as described [Dingermann et al., 1988133 using strains YH-G3 (arg 4-1 7, his 4-38,lys 1-1, met 8-1, trp 1-1) and NH-A6 (ade 2-1, arg 4-1 7, his 4 , leu 2-1, met 8-1, trp 1-1 )>which were kindly provided by Dr. H. Hottinger.

Preparation of DNA and RNA Genomic DNA was prepared from isolated nuclei [Dingermann et al., 19881. Nuclei were lysed at 65°C in 8 ml buffer containing 200 mM EDTA, pH 8.4, and 2% sarcosyl. CsCl (8 g) plus 200 pl ethidium bromide (10 mgiml) were added, and after centrifugation (42,000 rpm, 20°C, 48 h) the DNA was collected, ethanol precipitated, and dissolved in Low TE buffer (10 mM Trisi HCl, pH 8, 0.1 mM EDTA). Total cellular RNA was prepared from 5 x lo7 to 1 x 10' washed cells. Cells were lysed in 3.5 ml 4M guanidinium isothiocyanate and quickly drawn several times through a syringe fitted with a needle until the viscosity of the solution decreased. The RNA was pelleted through a cushion of 1.5 ml of 4.7M CsCl (35,000 rpm, 18"C, 16 h), dissolved in 360 pl TES buffer (10 mM TrisiHCl pH 7.4, 5 mM EDTA, 1% SDS), and precipitated with 2.5 volumes of ethanol at -80°C. Purified RNA was stored in sterile water at -80°C. Hybridization Conditions DNA restriction fragments were size-fractionated in 1% agarose gels, and RNA was separated by formaldehyde gel electrophoresis [Lehrach et al., 19771 in 1.2% agarose gels. Nucleic acids were immobilized on GeneScreen plus@membranes and hybridized with end-labelled oligonucleotides used to generate the corresponding mutation in 4 x SSC, 0.1% SDS, 0.02% BSA, 0.02% Ficoll, 0.02% PVP, 0.1% Na-pyrophospat a t

stringent temperatures (Table 1).Filters were washed once in 4 x SSC, 0.1% SDS at room temperature and three times 15 minutes a t the stringent temperature. Dried filters were exposed to x-ray films with intensifying screens for 10 days at -80°C.

Enzyme Assay for P-Galactosidase Washed cells were resuspended in phosphate buffer (14.7 mM KH,PO,, 2 mM Na,HPO, pH6) to a density of 1 x 10' cellsiml. After a freezeithaw cycle, the suspension was vortexed for 1minute and cleared by centrifugation for 5 minutes in a n Eppendorf centrifuge; then 2 pl of extract was used to determine protein concentrations by the dye-binding metliod (Bio-Rad). PGalactosidase was assayed from 100 p1 extract; 300 pl Z-buffer (60 mM Na,HPO,, 40 mM NaH,PO,, 10 mM KCI, 1 mM MgSO, pH 71, 50 mM P-mercaptoethanol, and 200 p1 of a solution containing 4 mgiml o-nitrophenyl-P-D-galactopyranoside(ONPG) in 100 mM phosphate buffer pH 7 were added and assays were incubated a t 22°C. Reactions were stopped by adding 400 pl of 1M Na,CO, and cleared for 5 minutes in a microcentrifuge. An aliquot of 500 pl was diluted with 500 pl Z-buffer, and the optical density was determined at 420 nm. Specific activities are given as katimg protein. One kat is defined as the enzymatic activity which hydrolyzes one mol of ONPG per second at 22°C E , , ~ = 18.5 x lo3 [I . mol-' . cm-']). Alternatively, a s a quick screening assay, 100 pl lysate was placed into a well of a microtitre dish and 200 pl of CPRG-solution (0.04 mg chlorophenolred-p-D-galactopyranosid, 100 mM Hepes, pH 7.5, 10 mM MgCI,, 150 mM NaCl, 0.05% Tween 20, 1 mM P-mercaptoethanol) were added. The red-colored product can be quantitated a t 578 nm E~~~ = 75 x lo3 [l . mol-' . cm-'I). P-Galactosidase can also be detected after SDS-polyacrylamide gel electrophoresis [Teeri et al., 1989; Dingermann et al., 1989bI. Extracts corresponding to 50 pg protein were mixed with cold loading buffer and applied directly to a n 8% SDS-polyacrylamide gel [Laemmli, 19701. After electrophoresis at 4"C, the separation gel was washed in cold Z-buffer and twice in Z-buffer for 15 minutes a t room temperature. The substrate 4-methyl-umbelliferyl-~-~-galactopyranoside was dissolved at 20 mgiml in dimethyl sulfoxide and

DINGERMANN ET AL.

412

--+

Ha

s I

fill in Hindlll linker

H

SH I I

1

ligation into M1SmpBiHindlll primer directed mutagenesis

I

E B HiP H 1

1

S H Ha I I

I I I

I

Haelll, EcoRl fill in EcoRl linker

ligation into YRpl7iEcoRI

E B HIP H 1 1

I l l

+

S H ECE I I

I I !

Fig. 1. Construction of a n amber derivative of a tRNAnP(CCA) gene from D.discoideum. Location ofHaeIII (Ha),Sau3A ( S ) ,HindIII (H), PstI (P),EcoRI (E), BarnHI (B), Hind11 (Hi), ClaI (C) is indicated. To obtain mutant Trp5, two consecutive rounds of mutagenesis were performed using oligonucleotides trp(CCA)iG and trp(amber) (Table 1) [Dingermann et al., 1988131.

diluted 1,000-fold in Z-buffer containing 50 mM p-mercaptoethanol. The gel was agitated for 10 minutes in substrate solution at room temperature and rinsed with water, and fluorescent enzyme bands were photographed under ultraviolet (UV) illumination through a pale yellow filter (Kodak Wratten 2E).

RESULTS AND DISCUSSION Construction of Amber and Ochre Su+ Genes Derived from a D. discoideum tRNATrpGene and a D. discoideum tRNAG'" gene tRNAnP(CCA) genes can easily be converted by a single-base mutation to amber (UAG) suppressor variants. Similarly, tRNAG'"(UUC) genes can be converted to ochre (UAA) suppressor genes, which can be further mutated to amber WAG) suppressors. The construction of typtophane-inserting suppressors has been reported [Dingermann et al., 1988bl. Protruding ends of a 210 bp HaeIIIISau3A fragment containing a D. discoideum tRNAnp gene [Peffley and Sogin, 19811 were filled in, HindIII linkers were attached, and the fragment was inserted into the HindIII site of M13mp8. The Trp5 mutant used in this study was generated through two consecutive rounds of primer-directed mutagenesis [Dingermann et al., 1988b], and mutations were confirmed by sequencing. Relative to the wild-type gene, Trp5 contains the amber mutation and a n A to G transition in its 13 bp intron. Fragments were isolated and ligated into the EcoRI site of YRpl7 (Fig. 1).This vector can be propagated in E. coli as well as in s. cerevisiae [Hottinger

et al., 19821 and can also be integrated into the D.discoideum genome. From our collection of eight isolated nonallelic tRNAG1"(UUC)genes [Dingermann et aZ., 1989a1, we selected gZu2 (tRNAG1"-N376)to be mutated to suppressor variants because i t is encoded on the smallest (1.15 kb) EcoRI fragment. In a first round of mutagenesis, this tRNA gene was converted to a n ochre suppressor, which in a second round was further converted to a n amber suppressor gene. After mutagenesis the EcoRI fragments were inserted into the EcoRI site of YRpl7 as well. Cloverleaf structures of the wild-type genes and of their suppressor variants are shown in Figure 2.

Expression and Analysis of D. discoideum Suppressor tRNA Genes in S. cereuisiae Because a system allowing analysis of suppressor tRNA genes in D.discoideum was not available a t the time, we decided to analyze these mutants in the heterologous organism, S. cerevisiae. It has been shown repeatedly that heterologous tRNA genes can be expressed [Dingermann et al., 1988133 and do function [Hottinger et al., 19821 in S. cerevisiae. Our analyses revealed that the tRNAnP(amber) suppressor gene was not able to suppress the corresponding nonsense mutation in the auxotrophic marker gene met 8-1 of strain YH-G3 (arg 4-1 7, his 4-38, lys 1-1, met 8-1, trp 1-1) (Fig. 3d). This was surprising, because the tRNAnP (amber) gene from D. discoideum is transcribed and processed in vivo in yeast [Dingermann et al., 1988133. Meanwhile, the homologous tRNAnP gene has be converted to a tRNATTP(amber)suppressor, and the corresponding tRNA suppresses efficiently different amber mutations [Kim and Johnson, 19881. A likely explanation for the failure of D.discoideum tRNATTpsuppressors to function in yeast is the inability of the yeast trp-tRNA synthetase to charge the heterologous tRNAs. In contrast, both amber and ochre tRNAG'" suppressor variants from D. discoideum are active in yeast. The amber mutation in the met 8-1 allele (Fig. 3c) and the ochre mutations in the arg 4-1 7 and leu 2-1 alleles (Fig. 3b) of strain YH-G3 are efficiently suppressed by the corresponding suppressor tRNAs. The ochre mutation in the ade 2-1 allele, however, cannot be suppressed by the tRNAG1"(ochre)(Fig. 3a). These results gave us enough confidence to establish a detection system for nonsense suppression in D.discoideum. P-Galactosidase as Reporter for Suppressor tRNA Gene Function in D. discoideum Recently we described the construction of plasmids allowing p-galactosidase expression in D.discoideum [Dingermann et al., 1989bl. The plasmid pA6PTlac.l carries the lacZ gene from E. coli fused in frame to the N-terminal coding region of the actin 6 gene from D. discoideum. Additionally, the plasmid contains a tran-

NONSENSE SUPPRESSION IN D. DISCOIDEUM

~ R N A ~ ~

u

413

~RNA~"

(;

U A (;

c

c

C u A (amber)

Fig. 2. Cloverleaf structure of tRNAnP(CCA)- and tRNAG'"(UUC)-genesfrom D. discoideum. Mutations generated to yield amber and ochre suppressors are indicated. In this study, we used the tRNAnP gene variant previously termed Trp5 [Dingermann et al., 1988b1, which carries in addition to the amber mutation a n A to G transition in the 13 bp intron region.

a

re)

his, met his, met, ade.

his, arg, ade, leu

AT% ber

)

his, arg, ade, leu

Fig. 3. Phenotypic anaysis of yeast transformants carrying different tRNA gene variants from D. discoideum. The cereuisiae strain NH-A6 (ade 2-1, arg 4 -1 7 , his 4 , leu 2-1, met 8-1, trp 1-1) was transformed with plasmids carrying the indicated tRNA gene variants from D. discoideum. Positive transformants were selected on the basis of trp 1-1 complementation conferred by the trpl gene located on the transformation vector YRpl7 [Hottinger et al., 19821. These cells were grown in YPD medium (1% yeast extract, 2 4 bacto peptone, 2% glucose), washed, and counted. Cells were applied in serial dilutions

(1 x lo7 to 1 x 10') to SD agar plates (0.67% yeast nitrogen-base wio amino acids, 2% glucose, 1.2% agar, plus indicated amino acids) and incubated a t 30°C. Pictures were taken after 2 days. The tRNA'"". (ochre) is not able to suppress the ochre mutation in the ade 2-1 allele (a) but can suppress mutations in agr 4-17, and leu 2-1 (b). The tRNA"'"(amber) variant suppresses the amber mutation in the met 8-1 allele (c), whereas tRNATTP(amber)suppressors are inactive in yeast (d) [Dingermann et al., 1988bl.

scription terminationlpolyadenylation signal derived from the Dictyostelium actin 8 gene. This plasmid, if stably integrated into the genome, allows efficient ex-

pression of functional (3-galactosidase in D.discoideum (Table 2). To introduce nonsense mutations into the N-terminal part of the fusion gene, a HindIIIiMstII

s.

414

DINGERMANN ET AL. TABLE 2. P-Galactosidase Activities in D. discoideum Transformants*

Strain AX-2 716 512 112 111 912 314 712

Integrated genes A6::lacZ A6::lacZ.UAA A6::lacZ.UAG A6::lacZ.UAG trp(amber) A6::lacZ.UAG elu(amber)

+ trp(amber) + glu(amber)

Specific activity (pkatlmg protein) 0,07 i 0,003 800 2 30 0,12 2 0,009 0,42 i 0,015 64,7 2 3,5 0,09 i 0,003 18,6 t 2,l 0.1 ? 0.003

*Specific enzymatic activities are given as pkatlmg protein. 1 kat is defined as the catalytic activity which hydrolyzes 1 mole of substrate (ONPG) per second at 22°C. Mean values were calculated from at least four individual tests. ATG GAT GGT GAG GGA ATT CAT CGG GGC AAT Actin 6 a I R-gal

fragment was isolated from pA6PTlac.l and cloned into a HindIIIlMstII digested M13mp8 vector. Amber and ochre nonsense mutations were introduced by primerdirected mutagenesis at positions indicated in Figure 4. Fragments were back-ligated into pAGPTlac.1, resulting in variants pA6PTlac.UAG and pA6PTlac. UAA. Under normal growth conditions, neither of these lac2 variants produces any significant amounts of P-galactosidase when stably integrated into the D. discoideum genome (Table 2).

Amber Suppressors are Functional in D. discoideum Suppression in D.discoideurn was tested by cotransformation of three plasmids. One plasmid, pNeo-IT, contained as selectable marker the neo‘, gene under transcriptional control of a n actin 15 promoter and a n actin 15 terminator [Cohen et al., 1986; Knecht et al., 19861. This plasmid confers G418 resistance to D.discoideurn cells and is the only plasmid in our experiments that is kept under selective pressure. The second plasmid contained the mutated lac2 reporter gene, and the third plasmid carried the suppressor tRNA gene. Transformants were selected upon G418 resistance and were recloned on bacterial lawns. Resulting clones were further tested for P-galactosidase activity. All transformants carrying amber suppressor tRNA genes and the corresponding reporter gene expressed significant-although different-amounts of P-galactosidase activity (Table 2). P-galactosidase-negative transformants turned out not to contain either the reporter gene or the suppressor tRNA gene. Two positive clones, in addition to some control transformants, were tested in more detail. Total cellular RNA and genomic DNA were isolated and analyzed. Northern blots were hybridized with nick-translated fragments specific for the lac2 gene and for a not-yet-characterized endogenous mRNA of 350 bp (kindly provided by H. Kersten and U. Hoja). The latter probe served as control to evaluate the amount of mRNA applied to the gel. As shown in Figure 5, the

4

EcoRl

TAG

4

BAA

Fig. 4. Schematic representation of plasmid pA6PTlac. 1, which carries the lac2 reporter gene, controlled by a D. discoideum actin 6 promoter and an actin 8 transcription terminationipolyadenylation signal [Dingermann et al., 1989133. The cross-hatched region corresponds to the actin 6 promoter core, whereas the light-stippled region !t,ermed A6P*) corresponds t,o t,he act.in 6 promoter upstream element (PUE) [Nellen et al., 19861. The coding region contains three actin codons and a short linker sequence to which the entire lac2 coding region was fused in frame. A Glu (GAG) codon in front of the lac2 coding region was mutated to a UAG (amber) and to a UAA (ochre) codon. Resulting plasmids were termed pA6PTlac.UAG and pAGPTlac.UAA, respectively.

highest concentration of la&-specific mRNA is present in clone 716, which carries a n unmutated reporter gene. This clone serves as our general reference strain because it expresses much higher than average amounts of P-galactosidase (800 pkatlmg protein, see Table 2). Clone 112 also contains high levels of lac2 mRNA. However, due to the amber mutation in this message, no significant P-galactosidase activity is detectable. Control clones 912 and 712 carry only a suppressor tRNA gene but no reporter gene (Fig. 6, lanes e and g). Consequently, neither a lac2 specific mRNA (Fig. 5, lanes f and g) nor P-galactosidase activity (Table 2) is detectable in these clones. Clones 111 as well as 314 express the lac2 reporter gene, although steady-state levels of la&-specific mRNAs are significantly lower compared with 716 control cells. This explains in part the differences in P-galactosidase activity between clones 111 and 314 on the one hand and control strain 716 on the other hand and must be considered if one tries to evaluate suppression efficiency in D. discoideurn. Proof that P-galactosidase activity detectable in clones 111 and 314 is indeed a result of active suppression is provided in Figure 6. This blot was first hybrid-

NONSENSE SUPPRESSION IN D. DIsC0IDEU.M

415

in two consecutive rounds with oligonucleotide trp(amber) (Fig. 6b, upper panel) and with oligonucleotide P gZu(amber) (Fig. 6b, lower panel). Under stringent conE m ditions, only the suppressor tRNA genes and not the resident gene copies are picked up. Apparently, more E c tRNAnP(amber) genes than tRNAG'"(amber) genes are 5 (3 integrated in the corresponding clones, as judged from U U n L Q 3 3 Q signal intensities. Whether this reflects the less effin n ri 0 0 cient lac2 expression in clone 314 (18,6 pkatlmg proE m E m m tein) compared with clone 1/1 (64,7 pkatlmg protein) .. 0 m Yx (D.. (D 3 (D has not been determined. Alternatively, the insertion U a U a c w of tryptophane at the mutated position in the reporter 716 111 314 112 912 712 gene may be preferable over glutamic acid. This, however, seems unlikely because the mutated codon used a b c d e f g to be a glutamic acid codon, and i t resides upstream of the lac2 coding region. Activity levels can be visualized directly by activity staining after SDS-polyacrylamide gel electrophoresis (Fig. 7). In summary, amber suppression works efficiently in D. discoideum. Low mRNA levels rather than inefficient suppression account for the differences in P-ga3 kb -b lactosidase activity measurable in clones 1/1 and 3/4 compared with 7/6 control cells. This conclusion is supported by the fact that clones were recently selected that express much higher levels of p-galactosidase from a mutated reporter gene than those presented here. 0.35 kb Most of the transformants did not show any abnormalities with respect to growth or development. Only a few clones turned out to be developmentally deficient, but these strains were obtained with a frequency that Fig. 5. Northern-blot analysis of selected D. discoideum transfor- is expected if three plasmids are used for transformamants. Whole-cell RNA was prepared and separated by formaldehyde tion. The unaffected growth rate was somewhat unexgel electrophoresis [Lehrach et al., 19771 on 1.2% agarose gels. After pected, as prokaryotic as well as eukaryotic Su' blotting, membranes were hybridized with a nick-translated ClaIi XhoI fragment carrying the C-terminal part of the lac2 gene [Dinger- strains frequently grow a t a significantly slower rate than their wild-type counterparts. mann et al., 1989bl plus a nick-translated fragment that recognizes a n L

Q

Y

h

L

-.. -..

-..

e -

Y

W

+

yet-uncharacterized 350 bp D. discoideum mRNA (unpublished and kindly provided by H. Kersten and U. Hoja). The latter probe served as control to estimate the amount of RNA loaded to the gel. mRNA from untransformed Ax-2 cells is analyzed (lane a);Clone 716 (lane b) was obtained after transformation with pA6PTlac. 1;while clones 111 (lane c ) , 314 (lane d), and 112 (lane e) carry the pA6FTlac.UAG variant. Control clones 912 (lane D and 712 (lane g ) contain no reporter gene but plasmids coding for tRNA?amber) and tRNA"'"(amber), respectively. The A6::lacZ fusion gene message is approximately 3 kb long. The signal representing the 350 bp message serves as a control to estimate the relative amount of mRNA applied to the gel.

Ochre Suppressors are Apparently Lethal for D. discoideum Using the same approach as for the analysis of amber suppressor tRNA genes, we also tried to establish D. discoideum cell lines carrying functional ochre suppressor tRNA genes. However, we never succeeded in establishing such a strain. Although clones were initially visible 5 to 7 days after transformation, many died after a few generations. Stable G418-resistant clones turned out not to contain the suppressor tRNA gene. We therefore conclude that, in contrast to amber ized with the end-labelled oligonucleotide lacZ(UAG) suppressors, the presence of a n active ochre suppressor and washed at 48°C and 52°C (Fig. 6a). At nonstringent is not compatible with growth in D. discoideum. This 48"C, the oligonucleotide recognizes the mutated a s result is rationalized by the fact that D. discoideum well a s the unmutated lac2 gene, whereas at stringent uses an extreme codon bias [Sharp and Devine, 19891. 52°C only the mutated genes are recognized. As stan- The genome is very A/T-rich and, in almost all indard for estimating the copy number, we applied about stances, UAA is used a s termination codon. A ochre 10 copy equivalents of plasmid DNA (Fig. 6a, lanes i suppressor is able to read these codons causing the synand j). According to these standards, the reporter gene thesis of many oversized proteins. On the other hand, this might explain why amber is amplified roughly between 50 and 100 times. After removing the radioactive probe, filters were reprobed suppressors are tolerated so well in D. discoideum. Be-

DINGERMANN ET AL.

416

B L

a

n

5

v

E

c

L

L

a

m

A

0

? a

L

al n

s E . U

0

-m

-m

W

W

0.

U

c L

a 716

n

n

5

v

f

v

-z a m W

112 111 9 1 2 3 1 4

7/2

Y

a

a a

-m..

: z -.. -.. a

m

(0

u)

(D

0

? X

a

q 716

a

b

a a

c

d

e

f

g

U

c

U

b

a

?

L

L

al

al

n

n

Y

v

-.. -.. W d d V

a

El

El

*.

..

-ual

E c.

V

(0

1 1 2 111 912 314 712

c

d

a

f

g

h

i

1

48°C

52°C

e-

Fig. 6. Southern-blot analysis of HzndIII-digested nuclear DNA from selected D. discoideum clones. To estimate the copy number of the integrated reporter gene 10 copy equivalents of Hind111 digested plasmids pA6PTlac.l (i) and pA6PTlac.UAG were applied to the same gel as well. A Filters were hybridized with the end-labelled oligonucleotidelacZ(UAG)(Table 1)and subsequently washed at 48°C (upper panel) and 52°C (lower panel). At 52°C the probe hybridizes only to the mutated genes, whereas a t unstringent 48°C also the wildtype version is picked up by the probe, although with reduced inten-

sity. Therefore, the copy number of the reporter gene in clone 716 is significantly higher than in all other analyzed clones: B After removal of the probe, the filter was rehybridized first with end-labelled primer trp(amber), yielding a 250 bp signal in clones 111 and 912 (upper panel). Thereafter the filter was hybridized with end-labelled primer gZu(amber) yielding faint signals of about 2 kb in clones 314 and 712 (lower panel). Hybridication and washing were a t the stringent temperatures that prevented hybridization of the oligonucleotides to the resident gene copies.

cause this organism uses UAG very infrequently as termination signal, a tRNA that is able to read this rare codon does not do any harm. This makes D . discoideum a favorable organism to apply nonsense suppressor genetics. Although amber suppressors can be used in a positive sense, ochre suppressors might be useful a s well. Currently we are working on a system that will allow control of the expression of a eukaryotic tRNA gene. If this system has been established, ochre suppressors can serve a s conditional lethal mutations.

ACKNOWLEDGMENTS The authors thank Prof. Walter and Helga Kersten for continuous support and the people in the lab for many stimulating discussions and suggestions. This work was supported by grants from the Deutsche Forschungsgemeinschaft, from the Johannes und Frieda Marohn Stifung, and from Fonds der Chemischen Industrie. K.N. is a recipient of a post-doctoral fellowship from the Deutsche Forschungsgemeinschaft.

NONSENSE SUPPRESSION IN D. DISCOIDEUM

7 a a

b

l i l I 1 2 314 912 7 1 2

c

d

e

f

g

716

h

4-0-Galactosidase

Fig. 7 . In situ staining of a-galactosidase after SDS polyacrylamide gel electrophoresis. Crude protein extracts (50pg protein) from individual clones were loaded onto a SDS polyacrylamide gel [Laemmli, 19701. After electrophoresis the gel was processed as described in Materials and Methods. In lane a purified p-galactosidase (0.5 pkat) was analyzed as control. According to Table 2, the following activities were loaded lane c, 3 pkat; lane e, 1 pkat; lane h, 40 pkat.

REFERENCES Capone JP, Sharp PA, RajBhandary UL (1985): Amber, ochre and opal suppressor tRNA genes derived from a human serine tRNA gene. EMBO J 4:213-221. Celis J E , Smith J D (eds) (1979): Nonsense mutations and tRNA suppressors. New York: Academic Press. Cohen SM, Knecht D, Lodish HF, Loomis WF (1986):DNA sequences required for expression of a Dictyostelium actin gene. EMBO J 5: 3361-3366. Dingermann T, Amon-Bohm E, Bertling W, Marschalek R, Nerke K (1988a):A family of non-allelic tRNAVa'(GUU)genes from the cellular slime mold Dictyostelium discoideum. Gene 73:373-384. Dingermann T, Bertling W, Brechner T, Nerke K, Peffley DM, Sogin ML (1986):Structure of two tRNA genes from Dictyostelium discoideum. Nucleic Acids Res 14:1127. Dingermann T, Bertling W, Pistel F, Amon E (1985): Characterization of a Dictyostelium discoideum DNA fragment coding for a putative tRNAV"'(GUU)gene. Eur J Biochem 146:449-458. Dingermann T, Brechner T, Marschalek R, Amon-Bohm E, Welker DL (1989a):tRNA"'"(GAA) genes from the cellular slime mold Dictyostelium discoideum. DNA 8:193-204. Dingermann T, Nerke K, Blocker H, Frank R (198813): Structural requirements for the synthesis of tRNAT'p from Dictyostelium discoideum in yeast. Biochimie 70:711-719. Dingermann T, Reindl N, Werner H, Hildebrandt M, Nellen W, Harwood A, Williams J , Nerke K (1989b): Optimization and in situ detection ofEscherichia coli a-galactosidase gene expression in Dictyostelium discoideum. Gene 85353-362. Doerig RE, Suter B, Gray M, Kubli E (1988): Identification of an amber nonsense mutation in the rosy"'" gene by germ line transformation of an amber suppressor tRNA gene. EMBO J 7:25792584.

417

Early AE, Williams J G (1987): Two vectors which facilitate gene manipulation and a simplified transformation procedure for Dictyostelium discoideum. Gene 59:99-106. Fire A (1986):Integrative transformation of Caenorhabditis elegans. EMBO J 5:2673-2680. Hodgkin J (1985):Novel nematode amber suppressors. Genetics 111: 287-310. Hottinger H, Pearson D, Yamao F, Gamulin V, Cooley L, Cooper T, So11 D (1982):Nonsense suppression in Schizosaccaromyces pombe: the S . pombe Sup3-e tRNAS"'(UGA) gene is active in S. cereuisiae. Mol Gen Genet 188:219-224. Kim D, Johnson J (1988): Construction, expression, and function of a new yeast amber suppressor, tRNARPA.J Biol Chem 263:73167321. Knecht DA, Cohen SM, Loomis WL, Lodish HF (1986): Developmental regulation of Dictyostelium discoideum actin gene fusions carried on low-copy and high-copy transformation vectors. Mol Cell Biol 6:3973-3157. Laski FA, Belagaje R, Hudziak RM, Cappecchi MR, Palese R, RajBhandary UL, Sharp PA (1984): Synthesis of an ochre suppressor transfer RNA gene and expression in mammalian cells. EMBO J 3:2445-2452. Laski FA, Belagaje R, RajBhandary UL, Sharp PA (1982): An amber suppressor tRNA gene derived by site directed mutagenesis: Cloning and expression in mammalian cells. Proc Natl Acad Sci USA 795813-5817. Lehrach HD, Diamond D, Wozney JM, Boedtker H (1977): RNA molecular weight determinations by gel electrophoresis under denaturing conditions, a critical reexamination. Biochemistry 16:47434751. Laemmli UK (1970): Cleavage of structure proteins during assembly of the head of bacteriophage T4. Nature 227:680-685. Loomis WF (lY82): The development of' Dictyostelium discoideum. New York: Academic Press. Munz P, Dorsch-Hasler K, Leupold U (1983): The genetic fine structure of nonsense suppressors in Schizosaccharomyces pombe. 11. sup8 and suplo. Curr Genet 7:lOl-108. Nellen W, Silan C, Firtel RA (1984): DNA-mediated transformation in Dictyostelium discoideum: Regulated expression of an actin gene fusion. Mol Cell Biol 4:2890-2898. Nellen W, Silan C, Saur U, Firtel RA (1986): Regulatory sequences in the promoter of the Dictyostelium actin 6 gene. EMBO J 5:33673372. Peffley DM, Sogin ML (1981): A putative + RNARPgene cloned from Dictyostelium discoideum: Its nucleotide sequence and association with repetitive desoxyribonucleic acid. Biochem 20:4015-4021. Sharp PM, Devine KM (1989):Codon usage and gene expression level in Dictyostelium discoideum: Highly expressed genes do "prefer" optimal codons. Nucleic Acids Res 175029-5039. Sherman F (1982):Suppression in the yeast Saccharomyces cereuisiae. In Strathern JN, Jones EW, Broach J R (eds):"Molecular Biology of the Yeast Saccharomyces: Metabolism and Gene Expression." Cold Spring Harbor, New York: Cold Spring Harbor Laboratory, pp 463486. Steege DA, Sol1 D (1979): Suppression. In Goldberg RF (ed): "Biological Regulation and Development." New York: Plenum Press, vol I, pp 433-485. Teeri TH, Lehvaslaiho H, Franck M, Uotila J, Heino P, Palva ET, Van Montagu M, Herrera-Estrella L (1989):Gene fusions to Lac2 reveal new expression patterns of chimeric genes in transgenic plants. EMBO J 8:343-350. Temple GF, Dozy AM, Roy KL, Kan YW (1982): Construction of a functional human suppressor tRNA gene: an approach to gene therapy of thalassemia. Nature 296537-540. Watts DJ, Ashworth JM (1970): Growth of myxamoebae of the cellular slime mould Dictyostelium discoideum in axenic culture. Biochem J 119:171-174. Welker DL, Williams KL (1982): A genetic map of Dictyostelium discoideum based on mitotic recombination. Genetics 102:691-710. Wills N, Gestetland RF, Karn J, Barnett L, Bolten S, Waterston RH (1983):The genes sup-7X and sup-5111 of C. elegans suppressor amber nonsense mutations via altered transfer RNA. Cell 33575-583.